University of Southern California Dissertations and Theses

Spatial distribution of neuroendocrine motoneuron pools in the hypothalmic paraventricular nucleus

(USC Thesis Other)

Spatial distribution of neuroendocrine motoneuron pools in the hypothalmic paraventricular nucleus

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content SPATIAL DISTRIBUTION OF NEUROENDOCRINE MOTONEURON POOLS IN THE HYPOTHALMIC PARAVENTRICULAR NUCLEUS by Donna Marie Simmons ________________________________________ A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (NEUROSCIENCE) May 2006 Copyright 2006 Donna Marie Simmons UMI Number: 3237189 3237189 2007 Copyright 2006 by Simmons, Donna Marie UMI Microform Copyright All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, MI 48106-1346 All rights reserved. by ProQuest Information and Learning Company. Table of contents LIST OF TABLES....................................................................................................................v LIST OF FIGURES.................................................................................................................vi ABSTRACT ........................................................................................................................... xii PREFACE ............................................................................................................................ xiv SECTION I: BACKGROUND AND INTRODUCTION............................................................. 1 Chapter 1. Effector (Motor) Systems ................................................................................. 1 Overview:........................................................................................................................ 1 Somatic System: ............................................................................................................. 3 Autonomic system:.......................................................................................................... 3 Neuroendocrine system: ................................................................................................. 4 Magnocellular neuroendocrine neurons:......................................................................... 4 Parvicellular neuroendocrine neurons:............................................................................ 5 Chapter 2. The Paraventricular Nucleus............................................................................ 8 Overview:........................................................................................................................ 8 Anatomical subdivisions:............................................................................................... 10 Afferent Inputs to Magnocellular Neurons:.................................................................... 17 Afferent Inputs to Parvicellular Neurons:....................................................................... 19 Physiology:.................................................................................................................... 24 Place in behavioral circuits:........................................................................................... 25 Chapter 3. Research Rationale and Experimental Design............................................... 27 Overview:...................................................................................................................... 27 Experimental Rationale Questions:............................................................................... 29 Practical Experimental Considerations:......................................................................... 31 Specific Experimental Design: ...................................................................................... 32 SECTION II: EXPERIMENTAL DETAILS............................................................................. 35 Chapter 4. Experimental Materials and Methods............................................................. 35 Overview:...................................................................................................................... 35 General Experimental Procedures: ............................................................................... 38 Surgical Procedures:..................................................................................................... 41 Perfusion and Tissue Preparation:................................................................................ 42 Histology: ...................................................................................................................... 45 Chapter 5. Mapping Strategy........................................................................................... 48 Overview:...................................................................................................................... 48 Variability in Histological Sections:................................................................................ 49 Data Recording Procedure:........................................................................................... 51 Creating Primary Maps: ................................................................................................ 53 Conversion into Computer Graphics Files:.................................................................... 57 Cumulative Mapping from a Huge Data Set, Two Approaches: .................................... 60 Chapter 6. PVH 3-D Atlas Sub-Project ............................................................................ 75 Overview:...................................................................................................................... 75 Creating a Serial Section Atlas of PVH: ........................................................................ 76 Generating Sagittal and Dorsal Projection Outlines: ..................................................... 83 Other Information from Brain Maps, Useful for Meta-Analysis: ..................................... 88 ii Chapter 7. Meta-Data Mapping Strategy ......................................................................... 92 Overview:...................................................................................................................... 92 PVH 3-D Atlas Provides New Tools for Meta-Analysis:................................................. 94 Using PVH Outline Projections to Display Cumulative Data.......................................... 99 Accurate Mapping to a Standard Atlas:....................................................................... 104 Dealing with Distortion in Mapping to a Standard:....................................................... 109 SECTION III: RESULTS..................................................................................................... 114 Overview:.................................................................................................................... 114 Chapter 8. Somatostatin and Growth Hormone Releasing Hormone ............................ 116 Overview -- Somatostatin:........................................................................................... 116 Overview -- Growth Hormone Releasing Hormone:.................................................... 119 Somatostatin (and GRH): ne, and non-ne cell type distribution in sagittal view: ......... 121 Somatostatin (and GRH): ne, and non-ne distribution on Brain Maps Atlas Levels: ... 126 Chapter 9. TH (tyrosine hydroxylase), a marker for Dopamine....................................... 143 Overview:.................................................................................................................... 143 TH (dopamine): ne, and non-ne cell type distribution in sagittal view:......................... 144 TH (dopamine): ne, and non-ne cell type distribution on Brain Maps Atlas Levels: .... 147 Chapter 10. Thyrotropin Releasing Hormone ................................................................ 162 Overview:.................................................................................................................... 162 TRH: ne, and non-ne cell type distribution in sagittal view:......................................... 164 TRH: ne, and non-ne cell type distribution on Brain Maps Atlas Levels:..................... 168 Tables of PVH subdivisions vs. referenced papers showing cells present for TRH: ... 179 Chapter 11. Corticotropin Releasing Hormone .............................................................. 181 Overview:.................................................................................................................... 181 CRH: ne, and non-ne cell type distribution in sagittal view:......................................... 183 CRH: ne, and non-ne cell type distribution on Brain Maps Atlas Levels:..................... 186 Tables of PVH subdivisions vs referenced papers showing cells present for CRH:.... 198 Chapter 12. Oxytocin and Vasopressin ......................................................................... 200 Overview:.................................................................................................................... 200 OXY&VAS: ne, and non-ne cell type distribution in sagittal view: ............................... 202 OXY and VAS: ne, and non-ne cell type distribution on Brain Maps Atlas Levels:...... 206 Tables of PVH subdivisions showing OXY & VAS cells: ............................................. 218 Chapter 13. Composite View: all Neuroendocrine cell types Surveyed ......................... 219 Overview:.................................................................................................................... 219 Examples of combining data used in the first round of analysis:................................. 221 All neuroendocrine cell types surveyed: distribution on Brain Maps Atlas Levels: ...... 224 Tabulation of all cells recorded in Atlas Level figures:................................................. 237 Chapter 14. Staining Reproducibility and Co-Localization of Peptides ........................... 238 Overview:.................................................................................................................... 238 Staining Reproducibility............................................................................................... 238 Cell types that showed co-localization with each other: .............................................. 241 SECTION IV: DISCUSSION AND CONCLUSIONS ........................................................... 245 Chapter 15. Discussion of Results................................................................................. 245 Overview:.................................................................................................................... 245 Mapping Strategy, Creation of 3-D PVH model, and Meta-Analysis methods:............ 246 Discussion of Data-Results by Cell Type, as detailed in Chapters 8-12:..................... 249 iii Somatostatin and Growth Hormone Releasing Hormone Staining:............................. 259 Tyrosine Hydroxylase (marker for Dopamine) Staining:.............................................. 261 Thyrotropin Hormone Releasing Hormone Staining:................................................... 263 Corticotropin Hormone Releasing Hormone Staining:................................................. 265 Oxytocin and Vasopressin Staining: ........................................................................... 268 Composite Maps of Neuroendocrine Cell Type Distribution:....................................... 272 Peptide co-localization, General Discussion: .............................................................. 279 Chapter 16. Conclusions ............................................................................................... 283 Introduction: ................................................................................................................ 283 Overview:.................................................................................................................... 286 Significance of findings: .............................................................................................. 288 BIBLIOGRAPHY................................................................................................................. 294 APPENDIX I ....................................................................................................................... 317 Anatomical History of the PVH ....................................................................................... 317 Appendix I References................................................................................................ 322 APPENDIX II ...................................................................................................................... 326 Afferent Inputs to PVH and their Significance................................................................. 326 Appendix II References............................................................................................... 332 APPENDIX III ..................................................................................................................... 337 Electrophysiological Studies of the PVH......................................................................... 337 Appendix III References.............................................................................................. 344 APPENDIX IV..................................................................................................................... 351 Preliminary Experiments and Antibody Testing .............................................................. 351 Appendix V......................................................................................................................... 359 Reagents and instructions for in situ hybridization.......................................................... 359 Cryoprotectant Tissue Storage Solution for ISH (contains fixative)............................. 359 Hybridization Solution ................................................................................................. 362 ISH Solutions Preparation List .................................................................................... 363 In Situ Hybridization Supplies Checklist ..................................................................... 365 Addenda to published ISH protocol............................................................................. 367 APPENDIX VI..................................................................................................................... 368 Histology Solutions and Procedures............................................................................... 368 Fixation and Perfusion ................................................................................................ 368 Perfusion Technique (30 to 60 minutes) ..................................................................... 369 Cryoprotectant Tissue Storage Solution (“Anti-Freeze”).............................................. 370 Immunohistochemistry Procedure............................................................................... 371 Thionin (Nissl) Staining Procedure.............................................................................. 372 iv APPENDIX VII.................................................................................................................... 374 Fast Blue injection procedure ......................................................................................... 374 Appendix X......................................................................................................................... 376 Comprehensive Mapping procedure............................................................................... 376 PHASE I, from primary tissue images: 35mm projection maps................................... 376 PHASE II, from 35mm projection maps: separated layer maps .................................. 376 PHASE III, from separated layer maps: Illustrator data files ....................................... 376 PHASE IV, making Nissl drawings.............................................................................. 377 PHASE V, putting Nissl drawings [maps] in order....................................................... 377 PHASE VI: determining orientation of Experimental Nissls to Atlas............................ 378 PHASE VII: positioning Experimental Nissl data to Atlas orientation .......................... 379 PHASE VIII, fitting data layers to oriented Nissl maps ................................................ 380 PHASE IX, how to plot data onto PVH projection images ........................................... 381 PHASE IX, preparing data for PVH projection display ................................................ 382 PHASE X, displaying data (averaged 1-in-4 sections) on PVH projection images...... 385 Appendix IX........................................................................................................................ 387 Additional Experimental Data ......................................................................................... 387 APPENDIX X...................................................................................................................... 387 Numeric Tables of all Cells Mapped ............................................................................... 387 Appendix XI........................................................................................................................ 388 Background Discussion on TRH..................................................................................... 388 Appendix XI References ............................................................................................. 394 v LIST OF TABLES Table 5.01: Types of error, and sources of error in experimental neuroanatomical data...... 51 Table 8.01: From the literature, PVH subdivisions with identified Somatostatin (SS)......... 140 Table 8.02: From the literature, Growth Hormone Releasing Hormone (GRH) in PVH. ..... 141 Table 8.03: Data from this work, PVH subdivisions with identified Somatostatin (SS). ...... 142 Table 9.01: From the literature, PVH subdivisions with identified TH (dopamine) cells...... 160 Table 9.02: Data from this work, PVH subdivisions with identified TH (dopamine) cells. ... 161 Table 10.01: From the literature, PVH subdivisions with identified TRH cells. ................... 180 Table 10.02: data from this work, PVH subdivisions with identified TRH cells. ................. 180 Table 11.01: From the literature, PVH subdivisions with identified CRH cells.................... 198 Table 11.02: Data from this work, CRH cells in PVH subdivisions seen on Atlas Levels. .. 199 Table 12.01: Data from this work, PVH subdivisions with identified Oxytocin cells. ........... 218 Table 12.02: Data from this work, PVH subdivisions with identified Vasopressin cells. ..... 218 Table 13.01: Data from this work, PVH subdivisions with identified cells on Atlas Levels.. 237 Table 15.01: Cell type staining, numbers of cells per set, numbers at each Atlas Level. ... 251 Table 15.02: Cell type staining, numbers of cells (by subdivision) on Atlas Levels only..... 251 Table 16.01: Numbers of ne and non-ne cells from exemplar (1-in-4, 15µm) series.......... 288 vi LIST OF FIGURES Fig. 1.01: Three motor systems—diagram of neurons, major transmitters and endings. ....... 2 Fig. 2.01 Three functional compartments of PVH, diagramed at the ‘classic level’. ............... 9 Fig. 2.02 PVH in Nissl-stained rat brain, with corresponding Brain Maps Atlas Level.......... 11 Fig. 2.03 Vignettes of Atlas Levels to illustrate all PVH cytoarchitectonic subdivisions........ 12 Fig. 2.04 PVH hypophysiotropic cell-types and their major anterior pituitary cell targets. .... 14 Fig. 2.05 Schematic neural inputs to magnocellular & parvicellular neuroendocrine PVH.... 17 Fig. 3.01 Segregated model—five neuroendocrine cell types at the “classic level” in PVH.. 28 Fig. 3.02 Possible Motoneuron pool relationships illustrated in five Venn diagrams. ........... 31 Fig. 4.01 Schematic representation of the four stained sets of sections from one animal.... 36 Fig. 4.02 Schema for adequate mapping of all target cell types in relation to each other..... 37 Fig. 5.01 Photomicrographs of original data: Identical 35mm fields at 100X magnification.. 52 Fig. 5.02 Early (a), later (b) versions: hand-drawn data maps from two serial sections. ...... 55 Fig. 5.03 Example hand-drawn data map template, with subsequent computer maps......... 59 Fig. 5.04 Stacked Data Maps, from a single series at three anatomical levels in PVH......... 61 Fig. 5.05 Consecutive sections from a 1-in-4 series, viewed individually and all together.... 63 Fig. 5.06 Cumulative file from PVH12 section 4/4, viewed individually and all together. ...... 64 Fig. 5.07 Adding stacked data from successive sections: two virtual thick sections............. 66 Fig. 5.08 Partial data from three serial sections, shown as different virtual thin sections..... 68 Fig. 5.09 Composite file: different animals stained with same combination of antibodies. ... 69 Fig. 5.10 Composite files: “best fit” similarly stained map data from different animals. ........ 70 Fig. 5.11 Composite files: “best fit” data from different animals, one antibody in common... 71 Fig. 5.12 Composite files: “best fit” cumulative maps, data from three different animals...... 72 Fig. 5.13 Composite file: one “best fit” cumulative map, data from three different animals. . 73 Fig. 6.01 PVH subdivision outlines from Nissls, serial sections from Atlas Level 25 to 26. .. 78 vii Fig. 6.02 Virtual PVH 3-D model. Serial section outlines stacked in Atlas Level context...... 80 Fig. 6.03 Detail: Virtual PVH 3-D model: position of three functional compartments............ 81 Fig. 6.04 Three views: Virtual PVH model: assists dynamic visualization in 3-D space. ...... 82 Fig. 6.05 Measuring PVH section outlines to create sagittal and dorsal projections............. 85 Fig. 6.06 PVH sagittal outline projection: anatomical fiducial marks and Atlas Levels. ........ 86 Fig. 6.07 PVH dorsal outline projection: anatomical fiducial marks and Atlas Levels........... 87 Fig. 6.08 PVH Sagittal Projection: frozen-section proportions from celloidin-section Atlas... 89 Fig. 6.09 Large PVH midsagittal projection, frozen-section size with serial atlas sections... 90 Fig. 7.01 Determine plane-of-section: experimental data referenced to Brain Maps Atlas... 97 Fig. 7.02 Plane-of-section for three experimental brains: PVH10, PVH12, PVH14. ............. 98 Fig. 7.03 Can “1-in-4” section data approximate serial sections in a Virtual 3D Model?..... 100 Fig. 7.04 Move data to 60µm box to approximate 4-section span in Sagittal View............. 101 Fig. 7.05 Randomized data from figure 7.04, final form on PVH Sagittal Projection........... 103 Fig. 7.06 The “Geometry Problem” in transferring data to a standard Atlas Template. ...... 105 Fig. 7.07 Atlas Level 25 oriented in 3D Space, PVH Sagittal and Dorsal Projections......... 106 Fig. 7.08 Data-sections intersect sagittal and dorsal projections at Atlas Level 25. ........... 107 Fig. 7.09 Cumulative array of segmented Data Grids (from PVH12) on Atlas Level 25. .... 108 Fig. 7.10 Scaling-factor differences change rapidly on highly curved surfaces of PVH. ..... 113 Fig. 8.01 Somatostatin cell type staining: example photomicrographs from original data. . 118 Fig. 8.02 Somatostatin cell type distribution throughout PVH, in parasagittal view. ........... 122 Fig. 8.03 Somatostatin cell type distribution throughout PVH, all peptide-positive cells. .... 123 Fig. 8.04 Growth Hormone Releasing Hormone distribution throughout PVH.................... 124 Fig. 8.05 PVH subdivisions at Brain Maps Atlas Levels, reference in following figures...... 125 Fig. 8.06 Somatostatin cell type distribution in rostral PVH, on Atlas Level 22................... 126 Fig. 8.07 Somatostatin cell type distribution in rostral PVH, on Atlas Levels 23 and 24..... 127 viii Fig. 8.08 Somatostatin cell type distribution in central PVH, on Atlas Level 25.................. 128 Fig. 8.09 Somatostatin cell type distribution in central PVH, on Atlas Level 26.................. 129 Fig. 8.10 Somatostatin cell type distribution in central PVH, on Atlas Level 27.................. 130 Fig. 8.11 Growth Hormone Releasing Hormone distribution in PVH, on Atlas Level 27..... 131 Fig. 8.12 Somatostatin cell type distribution in PVH, from Atlas Level 22 & caudally......... 133 Fig. 8.13 Somatostatin cell type distribution in PVH, from Atlas Level 23 & caudally......... 134 Fig. 8.14 Somatostatin cell type distribution in PVH, from Atlas Level 24 & caudally......... 135 Fig. 8.15 Somatostatin cell type distribution in PVH, from Atlas Level 25 & caudally......... 136 Fig. 8.16 Somatostatin cell type distribution in PVH, from Atlas Level 26 & caudally......... 137 Fig. 8.17 Somatostatin cell type distribution in PVH, from Atlas Level 27 & caudally......... 138 Fig. 9.01 Neuroendocrine and non-neuroendocrine TH cell distribution throughout PVH. . 144 Fig. 9.02 TH (dopamine) cell type distribution throughout PVH, all peptide-positive cells. . 145 Fig. 9.03 TH (dopamine) staining-consistency, in three serial sets from the same case.... 146 Fig. 9.04 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures.... 147 Fig. 9.05 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 22. 148 Fig. 9.06 TH (dopamine) cell type distribution, on Atlas Level 22 & caudally. .................... 149 Fig. 9.07 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 23. 150 Fig. 9.08 TH (dopamine) cell type distribution, on Atlas Level 23 & caudally. .................... 151 Fig. 9.09 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 24. 152 Fig. 9.10 TH (dopamine) cell type distribution, on Atlas Level 24 & caudally. .................... 153 Fig. 9.11 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 25. 154 Fig. 9.12 TH (dopamine) cell type distribution, on Atlas Level 25 & caudally. .................... 155 Fig. 9.13 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 26. 156 Fig. 9.14 TH (dopamine) cell type distribution, on Atlas Level 26 & caudally. .................... 157 Fig. 9.15 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 27. 158 ix Fig. 9.16 TH (dopamine) cell type distribution, on Atlas Level 27 & caudally. .................... 159 Fig. 10.01 Neuroendocrine TRH cell type distribution throughout PVH.............................. 165 Fig. 10.02 Non-Neuroendocrine TRH cell type distribution throughout PVH. ..................... 166 Fig. 10.03 TRH cell type distribution throughout PVH, all peptide-positive cells................. 167 Fig. 10.04 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures.. 168 Fig. 10.05 TRH cell type distribution on Atlas Level 22 ...................................................... 169 Fig. 10.06 TRH cell type distribution on Atlas Levels 23 and 24......................................... 169 Fig. 10.07 TRH cell type distribution on Atlas Level 25. ..................................................... 170 Fig. 10.08 TRH cell type distribution on Atlas Level 26. ..................................................... 170 Fig. 10.09 TRH cell type distribution on Atlas Level 27. ..................................................... 171 Fig. 10.10 TRH cell type distribution in rostral PVH, on Atlas Level 22 & caudally............. 172 Fig. 10.11 TRH cell type distribution in rostral PVH, on Atlas Level 23 & caudally............. 173 Fig. 10.12 TRH cell type distribution in central PVH, on Atlas Level 24 & caudally............ 174 Fig. 10.13 TRH cell type distribution in central PVH, on Atlas Level 25 & caudally............ 175 Fig. 10.14 TRH cell type distribution in caudal PVH on Atlas Level 26 & caudally. ............ 176 Fig. 10.15 TRH cell type distribution in caudal PVH on Atlas Level 27 & caudally. ............ 177 Fig. 11.01 Neuroendocrine CRH cell type distribution throughout PVH.............................. 183 Fig. 11.02 Non-Neuroendocrine CRH cell type distribution throughout PVH..................... 184 Fig. 11.03 CRH cell type distribution throughout PVH, all peptide-positive cells. ............... 185 Fig. 11.04 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures. 186 Fig. 11.05 CRH cell type distribution in rostral PVH, on Atlas Level 22.............................. 187 Fig. 11.06 CRH cell type distribution in rostral PVH, on Atlas Level 22 & caudally. ........... 188 Fig. 11.07 CRH cell type distribution in rostral PVH, on Atlas Levels 23 and 24. ............... 189 Fig. 11.08 CRH cell type distribution in central PVH, on Atlas Level 23 & caudally............ 190 Fig. 11.09 CRH cell type distribution in central PVH, on Atlas Level 24 & caudally............ 191 x Fig. 11.10 CRH cell type distribution in central PVH, on Atlas Level 25. ............................ 192 Fig. 11.11 CRH cell type distribution in central PVH, on Atlas Level 25 & caudally............ 193 Fig. 11.12 CRH cell type distribution in caudal PVH on Atlas Level 26. ............................. 194 Fig. 11.13 CRH cell type distribution in caudal PVH on Atlas Level 26 & caudally............. 195 Fig. 11.14 CRH cell type distribution in caudal PVH on Atlas Level 27. ............................. 196 Fig. 11.15 CRH cell type distribution in caudal PVH on Atlas Level 27 & caudally............. 197 Fig. 12.01 Neuroendocrine and non-ne OXY cell type distribution throughout PVH........... 202 Fig. 12.02 OXY cell type distribution throughout PVH, all peptide-positive cells. ............... 203 Fig. 12.03 Neuroendocrine and non-ne VAS cell type distribution throughout PVH. .......... 204 Fig. 12.04 OXY cell type distribution throughout PVH, all peptide-positive cells. ............... 205 Fig. 12.05 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures. 206 Fig. 12.06 OXY and VAS cell type distribution in caudal PVH on Atlas Level 22. .............. 207 Fig. 12.07 OXY and VAS cell type distribution in rostral PVH, Atlas Level 22 & caudally. . 208 Fig. 12.08 OXY and VAS cell type distribution in caudal PVH on Atlas Levels 23 and 24.. 209 Fig. 12.09 OXY and VAS cell type distribution in rostral PVH, Atlas Level 23 & caudally. . 210 Fig. 12.10 OXY and VAS cell type distribution in rostral PVH, Atlas Level 24 & caudally. . 211 Fig. 12.11 OXY and VAS cell type distribution in caudal PVH on Atlas Level 25. .............. 212 Fig. 12.12 OXY and VAS cell type distribution in central PVH, Atlas Level 25 & caudally.. 213 Fig. 12.13 OXY and VAS cell type distribution in caudal PVH on Atlas Level 26. .............. 214 Fig. 12.14 OXY and VAS cell type distribution in central PVH, Atlas Level 26 & caudally.. 215 Fig. 12.15 OXY and VAS cell type distribution in caudal PVH on Atlas Level 27. .............. 216 Fig. 12.16 OXY and VAS cell type distribution in caudal PVH, Atlas Level 27 & caudally.. 217 Fig. 13.01 Composite file: different animals stained with same combination of antibodies. 221 Fig. 13.02 Composite files: “best fit” similarly stained map data from different animals. .... 222 Fig. 13.03 Composite files: “best fit” data: different animals, one antibody in common...... 223 xi Fig. 13.04 PVH subdivisions at Atlas Levels, for reference in the following figures............ 224 Fig. 13.05 Composite view neuroendocrine cell types, Brain Maps level 21(no cells)........ 225 Fig. 13.06 Composite view neuroendocrine cell type staining, Brain Maps level 22........... 226 Fig. 13.07 Composite color view neuroendocrine cell type staining, Brain Maps level 22.. 227 Fig. 13.08 Composite view neuroendocrine cell type staining, Brain Maps level 23 & 24. . 228 Fig. 13.09 Composite color view neuroendocrine cell type staining, Brain Maps level 23.. 229 Fig. 13.10 Composite color view neuroendocrine cell type staining, Brain Maps level 24.. 230 Fig. 13.11 Composite view neuroendocrine cell type staining, Brain Maps level 25........... 231 Fig. 13.12 Composite color view neuroendocrine cell type staining, Brain Maps level 25.. 232 Fig. 13.13 Composite view neuroendocrine cell type staining, Brain Maps level 26........... 233 Fig. 13.14 Composite color view neuroendocrine cell type staining, Brain Maps level 26.. 234 Fig. 13.15 Composite view neuroendocrine cell type staining, Brain Maps level 27........... 235 Fig. 13.16 Composite color view neuroendocrine cell type staining, Brain Maps level 27.. 236 Fig. 14.01 neSS cell type distribution throughout PVH, in different experimental animals.. 239 Fig.14.02 TH (dopamine) staining-consistency, in three serial sets from the same case... 240 Fig.14.03 Small numbers of cells showing CRH/VAS co-localized staining. ...................... 243 Fig.14.04 Apparent OXY and VAS co-localization staining with shCRH antisera............... 243 Fig.15.01 Cumulative figure (color) of all cell types stained at Atlas Level 21. ................... 252 Fig.15.02 Cumulative figure (color) of all cell types stained at Atlas Level 22. ................... 253 Fig.15.03 Cumulative figure (color) of all cell types stained at Atlas Level 23. ................... 254 Fig.15.04 Cumulative figure (color) of all cell types stained at Atlas Level 24. ................... 255 Fig.15.05 Cumulative figure (color) of all cell types stained at Atlas Level 25. ................... 256 Fig.15.06 Cumulative figure (color) of all cell types stained at Atlas Level 26. ................... 256 Fig.15.07 Cumulative figure (color) of all cell types stained at Atlas Level 27. ................... 258 xii ABSTRACT The paraventricular nucleus (PVH) is a small bilateral group of neurons in the basal hypothalamus. PVH is a key brain structure in control of eating, drinking, response to stress, and other goal-oriented behaviors essential for life. Endocrine function is controlled directly by PVH neurons—via hypophysiotropic signaling molecules released into the blood from their axon terminals in the posterior pituitary and median eminence (pituitary stalk). This work comprises a detailed “new millennium view” of spatial distribution of neuroendocrine effector cells within PVH, at much higher resolution than previously available. The major new finding is that primary neuroendocrine cell types are more intermixed within PVH subdivisions, and more widely distributed throughout the nucleus, than previously appreciated. Distribution patterns for a great number of intermixed non- neuroendocrine cells of hypophysiotropic chemical phenotype were also documented. In a secondary project, a serial section anatomical atlas of PVH was constructed from Nissl sections of the rat brain originally used for L.W. Swanson’s Brain Maps: an Anatomical Atlas of the Rat Brain. Thus, data can now potentially be recorded and interpreted at near serial section resolution within the context of a published standard. An accurate 3-D model of PVH and two new methods for analyzing high-resolution neuroanatomical data were developed from the serial section atlas. These in turn were used to elucidate important features of spatial distribution of neuroendocrine motoneurons in the paraventricular nucleus. PVH cells were identified as neuroendocrine by fluorescent retrograde tracer (Fast Blue) in the blood. Neurons that contained the hormones vasopressin and oxytocin, and pituitary stimulating and inhibiting substances dopamine, somatostatin, corticotropin- releasing-hormone (CRH), thyrotropin-releasing-hormone (TRH) and growth hormone- releasing-hormone (GRH) were identified by antibody staining in (1-in-4, 15µm) series of xiii frozen sections throughout rat PVH. All possible combinations of two different target proteins were stained in many series of sections from numerous animals, and identified with fluorescence microscopy. Maps of cell type distribution were collated (in register) within a graphics program. Detailed analysis of numbers, density, distribution and interrelationship between major neuroendocrine cell types in PVH from three exemplar animals is presented at high resolution, in a much more detailed view than previously available. xiv PREFACE The paraventricular nucleus of the hypothalamus (PVH) is a key element in control circuits for eating, drinking and other essential goal-oriented behaviors in mammals. Similar groups of cells serving the same functions exist in all vertebrate species. In the PVH, signals regulating endocrine and autonomic function are integrated with those in circuits that elicit complex behaviors crucial to maintain life. PVH controls metabolism directly via endocrine function, and physiologic responses to internal and external stressors via autonomic reflex projections. In response to sensory, metabolic and complex derived neural inputs, PVH outflow signals are adjusted dynamically moment-to- moment. Thus, exquisitely coordinated endocrine, autonomic and behavioral responses are produced according to ongoing needs of the organism. Internal physiologic homeostasis is maintained and behavior patterns appropriate to given developmental or situational needs are initiated. This small nucleus is crucial to physiological processes underlying medical conditions rampant in the modern developed world: obesity and related diabetes, and stress- related cardiovascular and gastrointestinal diseases to name but a few. The paraventricular (subventricular) nucleus was described based on Nissl stained material by Santiago Ramón y Cajal, the legendary nineteenth century Spanish neuroanatomist. He recognized PVH as a “unique mass of gray matter lying along the midline”, which he did not succeed in impregnating with the Golgi method. However, unusually, he had little else to say about this small bilateral nucleus located in the medio- basal hypothalamus. Cajal also described a “bundle of axons coursing through the infundibulum” that when followed toward its origin seemed to fan out toward a nucleus caudal to the optic chiasm (Ramón y Cajal, 1995: pp. 376, 404: Vol. II, English translation of 1911 French Edition). Thus, PVH position in close proximity to the pituitary gland hinted at an endocrine function that would be discovered in years to follow. xv Early in the twentieth century, clinical observations of aberrant autonomic and hormone function in patients with hypothalamic or pituitary tumors implicated the hypothalamus in viscero-autonomic and endocrine control. For historical review of early classic papers, see (Fulton, 1940). Based on studies performed in the 1940s, British physiologist Geoffrey Harris predicted hypothalamic control of pituitary function via axonal connections to a plexus of blood vessels in the neurohypophysis or pituitary stalk (Harris, 1948, 1955). Thus began the modern era in study of hypothalamic (PVH) control of pituitary function. In rat, the paraventricular nucleus of the hypothalamus is a compact, roughly triangular group of around ten thousand cells, about a cubic millimeter in volume on either side of the third ventricle just above the median eminence or pituitary stalk. In humans, it appears less compact and localized—seen rather as a thin strip of cells along the ventricle. Early anatomical observation of PVH revealed two cell types, magnocellular (large cell) and parvicellular (small cell) that might subserve different functions. Axons of magnocellular neurons could be traced to the posterior pituitary while projection of parvicellular neurons to the median eminence was less evident. Oxytocin and vasopressin hormones were characterized in magnocellular neurons and the posterior pituitary, and it was widely thought that hypothalamic factors might also control anterior pituitary function, especially the release of adrenocorticotropin hormone (ACTH) from corticotrope cells. In experimental studies, hypothalamic lesions and injections of hypothalamus tissue- extracts showed changes in endocrine function. Further, cessation of eating or ravenous eating and obesity after hypothalamic lesions in animals gave early indications of specific hypothalamic involvement in ingestive behavior. This led researchers to propose the existence of lateral and ventromedial “hypothalamic eating centers” (Stellar, 1954). Two studies in the 1970s spectacularly focused attention on the PVH as a primary location of xvi neural signals for eating and drinking. Paul Gold showed with an elegant series of small discrete lesions that ravenous eating could be elicited by interrupting ascending pathways to PVH rather than by lesioning cells in the ventromedial hypothalamic nucleus (Gold, 1973). A few years later Sarah Leibowitz (Leibowitz, 1975,1978) injected tiny amounts of norepinephrine into PVH to elicit immediate, specific drinking and eating responses in freely moving rats. In 1981 attention was again focused on PVH when corticotropin releasing factor (CRF, later termed CRH—corticotropin releasing hormone) was isolated from hypothalamus. Neurons of origin for this primary pituitary-stimulating hormone were localized in PVH, along with evidence that other brain cells also contained the peptide (Bloom et al., 1982; Swanson et al., 1983). Discovery and synthesis of CRH, the brain hormone controlling ACTH release (and thereby, adrenal gland function), was followed by intensified study of the physiological stress response, particularly its effects on the cardiovascular system (Gardiner and Bennett, 1989). The stage was thus set for investigation into the role of PVH in controlling ingestion (eating and drinking), response to stress, and possibly other motivated behaviors related to endocrine and autonomic function. Retrograde tracing methods combined with immunohistochemical localization of PVH cell types were employed to characterize PVH function and anatomical connections. Complex innervation of the nucleus, implications for endocrine and autonomic integration, and information about defined behavioral circuits related to PVH function were elucidated in the 1980s and 1990s with sensitive and specific anterograde and retrograde tracing and immuno-histochemical methods (Swanson and Hartman, 1980; Sawchenko and Swanson, 1983a; Swanson, 1986; Roland and Sawchenko, 1983; Moga and Saper, 1994; Jansen et al., 1995; Sawchenko et al., 1996; Thompson and Swanson, 1996). Thus, endocrine and autonomic character of PVH cells and functional xvii relation of PVH to behavioral-circuits supporting ingestion and other homeostatic functions were revealed. In 1998, Paul Sawchenko wrote an interesting, thought-provoking editorial in the Journal of Comparative Neurology: “Toward a new neurobiology of energy balance, appetite, and obesity: the anatomists weigh in”, discussing significance of current knowledge about PVH anatomy (Sawchenko, 1998). He reviewed traditional and developing ideas about hypothalamic control of ingestive behaviors mediated through endocrine and autonomic outflow from PVH, in relation to newly discovered peripheral and central peptides that have profound effects on ingestion and energy homeostasis. These areas of investigation have stimulated wider interest in PVH anatomy and functional connections. In rat, chemically characterized neuronal phenotypes are generally thought to be segregated within morphologically and functionally defined parts of PVH, based on major studies of hypothalamic chemical architecture (see for example Swanson, 1986). This work, now over two decades old, aided interpretation of neuroanatomical and physiological studies. However, as new more detailed data on PVH inputs, physiology and response to experimental manipulations accumulate, the need for a higher resolution description of specific cell-type distribution within PVH is apparent. That need, and the ability to identify more than one cell phenotype in a histological section (thus, intermixing of cell types and possible co-localization of two protein signatures within a single neuron) prompted new questions about the detailed chemical anatomy and probable function of PVH. The work presented here is intended to address some of those questions. 1 SECTION I: BACKGROUND AND INTRODUCTION Chapter 1. Effector (Motor) Systems Overview: Reference to a general model helps form a conceptual framework when discussing function and interrelationship of neuronal cell groups in the central nervous system (CNS). The effector (or motor system) model of functional organization is a good one. Instructive and easily grasped, it has provided a coherent approach for study of numerous central nuclei. The motor system model embodies the idea that input stimuli impinge on a central neuron or group of neurons, whose output is the final common path to an effector end organ, thus producing an observable function. Inputs to effector systems are typically thought of as sensory, and the end organ has been commonly characterized as a muscle that produces movement (thus, motor) in the skeletal or somatic motor system. In fact, however, it is well accepted that effector systems also receive blood-borne (humoral) and complex derived neural inputs, and they target several end-organ types that can produce different kinds of functional consequences. The musculo-skeletal or Somatic System, the Autonomic System and Endocrine System are all examples of effector or motor systems, as described below and shown diagrammatically in figure 1.01. Study of neuroendocrine (neurosecretory) cell types in the paraventricular nucleus of the hypothalamus (PVH) is approached here with the assumption that these cells provide outputs to effect endocrine function and are therefore accurately termed secretomotoneurons (Swanson, 1991; Shepherd, 1994; Markakis and Swanson, 1997; Markakis, 2002). The PVH is a complex nucleus that contains elements of both autonomic and endocrine motor systems. It also has connections to circuits controlling integrated behavioral output (skeletal motor) systems, a topic discussed below and in chapter two. 2 Fig. 1.01: Three motor systems—diagram of neurons, major transmitters and endings. a.) Somatic neuromuscular junction. b.) Visceral synapses. c.) Neuroendocrine peptides are dispersed diffusely via the bloodstream and, based on their chemical composition, have exquisitely precise effects on specific target cells. CNS—central nervous system, ACh— acetylcholine, NA— noradrenalin (norepinephrine), Oxy, Vas—oxytocin, vasopressin. 3 Somatic System: Anatomical analysis of neural control in the voluntary or somatic motor system that produces movement of an animal within its environment has provided major insight into organizing principles of CNS function; see for example (Hollyday, 1980). Thus, it is well accepted that a motoneuron in the ventral horn of the spinal cord represents the final common pathway to direct stimulation of a specific voluntary (skeletal or striated) muscle at specialized cholinergic muscle fiber synapses called neuromuscular junctions, schematized in figure 1.01, a). When its collective fibers are adequately stimulated by an innervating motoneuron pool a muscle contracts, resulting in motion e.g., of a limb (Kandel et al., 1991). Clear delineation of neuronal function in the skeletal-muscular or somatic motor system has lead to use of ‘motor’ in a general conceptual way to define CNS output in terms of effector function. Autonomic system: The autonomic or visceral motor system controls the internal milieu. It functions to control the viscera largely out-of-consciousness, unlike the purposeful voluntary activation of the somatic motor system. Autonomic functions include monitoring and control of crucial visceral processes such as blood circulation, digestion and elimination of metabolic waste. The autonomic system maintains various internal parameters (blood pressure, core temperature, glucose level, fluid balance, etc.) within an optimal (homeostatic) physiological range. It also effects physiologic adjustments required in response to changing environmental circumstances and physiological challenge. For example, blood pressure alterations during exercise are controlled by the autonomic nervous system. Central control of the autonomic system is effected via preganglionic neurons that synapse onto autonomic ganglia, which in turn make postganglionic effector synapses onto smooth muscle, cardiac muscle and exocrine glands in the viscera, as shown in figure 1.01, b). 4 Two types of autonomic ganglia, sympathetic and parasympathetic, generally subserve stimulatory and quiescent functions respectively in end organs that receive both types of innervation. Sympathetic ganglia form an interconnected chain on either side of the spinal cord that provide widely ramified adrenergic innervation to viscera for rapid response to changes in the (internal or external) environment. This is typified by the so-called “fight or flight” response elicited by exposure to a threat, or external stressor. Parasympathetic ganglia, in contrast, are located in or near their target organs and effect a predominantly “rest and digest” control of basal functions via cholinergic synapses (Kandel et al., 1991). Some parts of PVH have direct descending projections to autonomic preganglionic neurons, a subject discussed later. Neuroendocrine system: The neuroendocrine system is also accurately characterized as an effector, or motor, system. It is comprised of central magnocellular and parvicellular neurosecretory neurons. They control endocrine function via release of peptide products (and possibly other neurotransmitters) from axons that terminate on or near blood vessels, rather than at classical nerve cell synapses, as shown in figure 1.01, c). Released signal molecules enter the bloodstream and can act systemically as whole body hormones (humoral factors), or more locally as releasing hormones or inhibiting factors in the anterior pituitary. In direct response to these hypothalamic (hypophysiotropic) signal molecules, specific cells in the anterior pituitary release stimulating hormones that target distant endocrine glands (Swanson, 1986). Magnocellular neuroendocrine neurons: Magnocellular (large cell) neurosecretory neurons are located primarily in the PVH and supra optic nuclei (SON) and a few small accessory groups of cells scattered between them. Their axon terminals release one of two closely related nonapeptides, Vasopressin 5 (VAS) or Oxytocin (OXY), into blood vessels of the posterior pituitary and thus into the general circulation (Zimmerman, 1981). VAS (also called anti-diuretic hormone) controls various aspects of fluid balance physiology, predominantly in the kidney. OXY also influences fluid homeostasis, by regulation of sodium retention in the kidney, sodium ingestion (salt appetite) (Blackburn et al., 1992; Stricker and Verbalis, 2004) and through its effects on smooth muscle contraction (thus, vasodilatation and vasoconstriction of blood vessels) throughout the body. Well-studied features of smooth muscle function controlled by OXY are concerned with reproduction: contractions in the uterus during parturition, and in the mammary gland during the milk-ejection reflex of lactation. These effects are by no means exclusive or comprehensive, as males and females (outside pregnancy and lactation) have relatively similar OXY expression in neurons. For a comprehensive review of OXY related topics see (Gimpl and Fahrenholz, 2001). VAS and OXY and their receptors have also been implicated in affiliative social behaviors e.g., pair bonding (monogamy), parental behavior, and intra-species aggression (Young, et al., 1998; Gimpl and Fahrenholz, 2001). Those functions of OXY and VAS, while extremely important, will not be considered further in these studies of neuroendocrine cell type expression in PVH. Parvicellular neuroendocrine neurons: Anterior pituitary cell releasing and inhibiting factors are synthesized in subsets of medial hypothalamic parvicellular (small cell, sometimes called parvocellular) neurons. The majority of these are located in PVH. Axon terminals of PVH parvicellular neurons release their neurosecretory products into the external zone of the median eminence (pituitary stalk) where they gain access to the anterior pituitary gland via specialized vasculature. Rather than having a typical arterial blood supply, the anterior pituitary (adenohypophysis) is bathed directly by blood flowing through a specialized capillary bed in the median eminence, the adenohypophyseal portal system. Thus, very small amounts of releasing hormones and 6 inhibitory factors axonally delivered to these specialized vessels in the median eminence are available for rapid and exquisite efferent control of anterior pituitary cells. Specific cell products (typically, stimulating hormones) released into the general circulation from the pituitary elicit distant responses such as release of hormones from endocrine glands (Swanson, 1986). Gonadotropin-releasing hormone (GnRH) neurons, which also send their axons to the external median eminence, are located rostral to PVH along the third ventricle in the median preoptic area. GnRH stimulates pituitary gonadotropes to release leutinizing and follicle-stimulating hormones that subsequently stimulate the gonads. The gonadotropin system, though crucially important, will not be discussed further in this work because its primary hypothalamic neurons are located outside PVH. The majority of growth hormone- releasing hormone (GRH) cells are located in the arcuate nucleus of the hypothalamus. However some GRH cells have been reported in PVH, thus their distribution is evaluated in the work described below. GRH stimulates growth hormone release from the pituitary, resulting in growth, repair and general metabolic regulation in virtually all tissues. Neurons that contain the other primary pituitary stimulating hormones and inhibiting factors are located in the periventricular and medial parvicellular subdivisions of the PVH. They are the proteins surveyed in these studies and will be discussed in detail below An interesting speculation is that the effector or motor system analogy could be extended to include functional organization of the neuroimmune system. In some senses, the immune system and CNS can be similarly characterized as containing two populations of widely dispersed but functionally interrelated cells: neurons and circulating immune cells. Immune cells communicate by diffuse, largely humoral chemical signals that have effects in both the neural and immune systems (Shepherd, 1994). In addition to classical neural synapses and neurochemical receptors, nerve cells clearly have receptors for “input” from blood-borne immune system signals such as cytokines (Patterson, 1994). Moreover, some authors refer 7 to the cell surface interactions of T cells and antigen presenting cells as “immunologic synapses” (Bromley et al., 2001). Characterizing the neuroimmune system as a motor system may represent yet another conceptual step extending the continuum of skeletal- motor, viscero-motor and endocrine-motor systems. PVH, with its endocrine, autonomic and behavioral system connections, is positioned (anatomically and functionally) to interact with and integrate output that influences all these systems. For example, the well-known effects of stress and high circulating corticosteroid hormones in suppressing immune system function are thought to be mediated at least in part by PVH endocrine output (Rivest, 2001). That phenomenon and other neural-cytokine interactions, though beyond the scope of these studies, are the subject of much current research. They may well bear consideration for a future “immune motor system” model of neural organization that could be integrated with that for the endocrine motor system and others described above. 8 Chapter 2. The Paraventricular Nucleus Overview: The paraventricular nucleus of the hypothalamus (PVH) is clearly an effector or motor nucleus. It is the final common pathway for central nervous system control of endocrine function. It is also the central site for integration of endocrine, autonomic and related behavioral aspects of life-sustaining activities such as ingestive behavior (eating and drinking), metabolic regulation, and response to internal (physiological) perturbations and external (world environment) stressors. Three functional compartments in PVH have been defined by physiological studies and retrograde tracing of axonal projections to and from PVH neurons (Swanson and Sawchenko, 1980; Swanson, 1986). Functional cell groupings are fairly distinct in rat PVH subdivisions, as defined by cytoarchitechtonic analysis of Nissl stained material (figure 2.02). This subdivision by cell grouping is exemplified by samples from Swanson’s Brain Maps rat brain atlas (figure 2.03) and data from this work (figures 5.01, 5.03), and is illustrated diagrammatically in figure 2.01 below. PVH neuroendocrine motoneurons are the subject of study in the experimental work presented here. Parvicellular neuroendocrine neurons directly control hormone output of the anterior pituitary gland (endocrine function) via axonal projections to portal vasculature in the external layer of the median eminence. In concert, magnocellular neuroendocrine neurons control body water metabolism more directly, via axonal release of vasopressin and oxytocin hormones into the general circulation via blood vessels of the posterior pituitary. For an interesting historical review, including clinical implications and comparative biology, see the Viewpoint: Science & Society article, “One hundred years of hormones” in the European Molecular Biology Organization’s EMBO reports (Tata, 2005). 9 Fig. 2.01 Three functional compartments of PVH, diagramed at the ‘classic level’. Neuroendocrine subdivisions subserving endocrine homeostatic functions project to median eminence and posterior pituitary. Descending subdivisions project to preganglionic neurons that control autonomic function, and to midbrain reticular structures involved in behavior. This frontal section view, most commonly used for illustrative purposes, is one of several through PVH. Non-neuroendocrine PVH neurons with descending projections to sympathetic and parasympathetic preganglionic neurons in the brainstem and spinal cord control viscero- autonomic output (cardiac, digestive, etc) in response to changes in the internal and external environment. Descending projections to midbrain reticular core and other brainstem nuclei provide essential input to circuits producing complex behaviors that maintain survival, e.g., 10 seeking food, water, social interaction and a safe environment. Within this context, the topics “PVH Anatomical Subdivisions”, “Afferent Inputs”, “Physiology”, and “Place in Behavioral Circuits” are discussed below. Anatomical subdivisions: PVH is a bilateral nucleus, containing about 10,000 cells in the rat, in a cubic millimeter of tissue on both sides and immediately adjacent to the dorsal aspect of the third ventricle (Swanson and Sawchenko, 1983). It can be divided into three or four magnocellular and six or seven parvicellular subdivisions (some are further divided into zones) based on spatial and cytoarchitectonic features (Gurdjian, 1927; Armstrong et al., 1980; Swanson and Kuypers, 1980; Koh and Ricardo, 1980). (See Appendix I for an in- depth discussion of PVH parcellation.) The relatively compact nature of PVH in rodent (especially rat) makes it a very useful target for anatomical and functional studies. Figure 2.02 depicts a Nissl stained section through central PVH with its corresponding Atlas Level from Swanson’s Brain Maps: Structure of the Rat Brain, anatomical atlas of the rat brain (Swanson, 1998). Defined anatomical subdivisions (see figure 2.03) have been shown to correspond well with neuropeptide content and axon projections that relate to endocrine or autonomic function (Swanson, 1986; 1991). Neurons in magnocellular parts of PVH send axons in a laterally arching trajectory that then turns medially and ventrally, traversing through the internal layer of the median eminence to terminate in apposition to large blood vessels in the posterior pituitary. Magnocellular neurons synthesize oxytocin (OXY) or vasopressin (VAS), and the two cell phenotypes are generally regarded as separate though somewhat intermixed and functionally related populations. Virtually all studies of peptide content and axonal release support this view. However, co-localized expression of OXY and VAS messenger RNAs in the same magnocellular neuron has been shown under special physiological conditions 11 (lactation in the female rat, extreme dehydration) and using sensitive DNA amplification techniques applied to individual isolated cells (Mezey and Kiss, 1991; Xi et al., 1999). Fig. 2.02 PVH in Nissl-stained rat brain, with corresponding Brain Maps Atlas Level. Illustration from Swanson, ’98: Brain Maps: Structure of the Rat Brain. Top panel is an overview of level 26, showing half of a photographed coronal Nissl-stained section displayed opposite to the graphics map defining its named brain structures. At top right is a sagittal brain schematic diagram illustrating the rostral-caudal position of the Atlas Level displayed. Note relatively small size of intensely stained PVH in comparison with the entire brain. In the bottom panel vignettes of the Nissl section and Atlas template are enlarged to show PVH and its subdivisions (detailed in figure 2.02) more clearly. Nissl cytoarchitectonic detail is not shown well in this enlargement of the original scanned section, but the template drawing clearly shows PVH subdivisions and PVH position in relation to median eminence, the base of the brain, and other hypothalamic structures. 12 Fig. 2.03 Vignettes of Atlas Levels to illustrate all PVH cytoarchitectonic subdivisions. Illustrated are vignettes from seven levels of Brain Maps (2004, 3 rd edition) that contain PVH. At bottom, right is a mid-sagittal diagram of rat brain indicating the rostro-caudal level of Atlas Level 27. Individual levels in the atlas were selected with a frequency such that each named structure would be included on at least two levels, not necessarily equidistant from one another. This figure contains drawings showing all defined subdivisions of PVH in levels that show it. They will be used in subsequent figures illustrating data results obtained in the experiments described in this work. 13 Anterior and medial magnocellular parts of PVH contain virtually pure populations of OXY cells while the posterior magnocellular part (especially its lateral zone) is generally seen to consist of a globular core of VAS cells embedded in a surround of mostly OXY-containing neurons (Swanson, 1987). This juxtaposition of OXY and VAS cell populations seems likely significant, in terms of related functions of the two peptides in maintaining body water homeostasis. Periventricular and anterior and medial parvicellular parts of PVH control anterior pituitary function via axonal release of peptide products into the portal vasculature of the median eminence. These cells are thought to be almost exclusively neuroendocrine (secretomotor) in function, releasing their peptides from axons in apposition to blood vessels. However, there is scant evidence of a few collateral projections to nearby hypothalamic structures (Hatton et al., 1985; Rho and Swanson, 1987). Parvicellular neurons send their axons ventrally in the case of periventricular neurons or diagonally, medially and ventrally near the third ventricle in the case of medial parvicellular neurons, to terminate in the portal vasculature of the external layer of the median eminence. There is some segregation of axons containing different peptides from lateral to medial in the median eminence (Hokfelt et al., 1975; Epelbaum et al., 1981) and evidence of preferential segregation of blood circulation to medial and lateral parts of the anterior pituitary (Harris, 1955; Sowers, 1980; Ceccatelli et al., 1992). It seems plausible that this segregation in cell body distribution (and axon terminal distribution) might subserve one or more levels of endocrine control. Subtle differences in pituitary endocrine stimulation could be effected by differential input to subsets of parvicellular neurons, or by localized median eminence inputs that might be inhibitory or facilitatory to peptide release (Bondy et al., 1989; Ceccatelli et al., 1991). Figure 2.04 shows 14 a schematic representation of parvicellular neuroendocrine PVH cell types and their anterior pituitary target cell types. Fig. 2.04 PVH hypophysiotropic cell-types and their major anterior pituitary cell targets. Prolactin stimulates the mammary gland, and is elevated in certain types of acute stress. TSH stimulates the thyroid gland to release thyroid hormones, GH induces growth and tissue repair, and ACTH stimulates the cortex of the adrenal gland to produce corticosteroids. Gonadotropes, the fifth pituitary phenotype that releases leutinizing hormone and follicle stimulating hormone (LH & FSH) are not shown here, because their hypothalamic releasing hormone (GnRH) is not contained in PVH cells. Hypophysiotropic molecules have been shown to be somewhat segregated in parvicellular subdivisions of PVH as noted above. A great many somatostatin (SS) or growth 15 hormone inhibiting factor cells, and relatively fewer dopamine (DA) containing neurons are localized in periventricular PVH neurons. Thyroid hormone releasing hormone (TRH), corticotropin releasing hormone (CRH), and growth hormone releasing hormone (GRH) are reported mostly in medial parvicellular PVH neurons. TRH and CRH neuron populations are intermixed in PVHmpd, with TRH predominating medio-dorsally and CRH predominating more ventrally in the subdivision (Swanson, 1986; Elde and Parsons, 1975; Elde and Hokfelt, 1979; Vandesande et al., 1980; Lechan and Jackson, 1982; Lechan et al., 1983; Swanson et al., 1983; Chan-Palay et al, 1984; Kawano and Daikoku, 1987). Scattered GRH neurons have been reported in medial and caudal PVHmpd, although most of them are localized in the more posterior arcuate nucleus of the hypothalamus (ARH) (Sawchenko et al., 1985). See Table 8.02. Specific locations of cells containing pituitary releasing and inhibiting factors will be discussed in Chapters 8-12, which detail results from this work obtained for each of them. Dorsal parvicellular (PVHdp), the ventral part of medial parvicellular (PVHmpv) and lateral parvicellular (PVHlp) subdivisions of PVH contain large numbers of cells with descending (largely autonomic) projections, as indicated earlier in figure 2.01. They project to spinal cord (sympathetic and parasympathetic preganglionic neurons), brainstem (visceral-sensory and visceromotor cell groups), and to targets in midbrain organizing circuits (Sawchenko and Swanson, 1982). More posterior parvicellular cells in PVH (shown as lateral and forniceal parts in figure 2.03, panel e.) also send descending projections to the brainstem and reticular core (Swanson and Sawchenko, 1980; 1983). Autonomic and other descending PVH neurons do not appear to send axon collaterals to the median eminence or posterior pituitary, although they contain at least some of the same neuropeptides (notably, OXY, VAS and CRH) as other neuroendocrine PVH neurons (Swanson et al., 1980; Swanson and Kuypers, 1980; Swanson and Sawchenko, 1982). PVH neurons have also 16 been shown to co-localize many other neuropeptides not known to be pituitary hormone releasing factors (e.g. neurotensin, enkephalin, cholecystokinin, galanin, etc,) and some of these may prove to be autonomic neurotransmitters (Hokfelt et al., 1983; Mezey et al., 1985; Hokfelt et al., 1987; Ceccatelli et al., 1989; Sakanaka et al., 1989; Meister et al., 1990; Landry et al., 1991; Levin and Sawchenko, 1993; Landry et al., 1997; Merighi, 2002). There is very little substantive evidence for the presence of classical interneurons in the PVH (Swanson, 1986), although GABAergic neurons in rostral PVH that apparently project to descending parvicellular subdivisions have been reported (Roland & Sawchenko, 1993). However, ultrastructural observation of dendritic arbors with possible intrinsic axo- dendritic synapses, reports of TRH-TRH axo-dendritic synapses and visualization of recurrent CRH axons directed back into PVH suggest there could be synaptic connectivity between cells in the same or different subdivisions (van den Pol, 1982; Lechan and Jackson, 1982; Rho and Swanson, 1987). Physiological studies show that glutamate microstimulation of identified GABA cells immediately adjacent (largely ventro-lateral) to PVH in slice preparations have an inhibitory effect on PVH magnocellular and parvicellular neuron firing (Tasker and Dudek, 1993; Boudaba et al., 1996). This is consistent with known peri-nuclear inhibitory cells projecting to PVH that Roland and Sawchenko term “documented (and generally inhibitory) limbic system influences on neuroendocrine function” (Roland, and Sawchenko 1993; Oldfield et al., 1985). Thus, glutamatergic stimulation of local inhibitory circuits may be important, considering that some hypothalamic afferents preferentially innervate the PVH surround, and could therefore have either an inhibitory or an excitatory (that is, disinhibitory: inhibiting a tonic inhibitory input) effect on PVH, depending on whether their transmitters or peptides are excitatory or inhibitory in nature (Herman et al., 2002). 17 Fig. 2.05 Schematic neural inputs to magnocellular & parvicellular neuroendocrine PVH. Sagittal view of the rat brain, from midline toward the animal’s right side. Projections, as determined by retrograde and anterograde tracing studies: brainstem, midbrain, limbic areas, and almost all parts of the hypothalamus. Panel a: neural projections to VAS or vasopressin (vas) cells of PVH, as examples of inputs to magnocellular neurons projecting to posterior pituitary (pp). IX: glossopharyngeal nerve, X: vagus nerve, NTS: nucleus of the solitary tract, A1: noradrenergic ventrolateral medullary cell group (nucleus ambiguous), ne: norepinephrine-containing pathway, SFO: subfornical organ, MePO: median preoptic nucleus. Panel b: inputs to CRH or corticotropin releasing hormone (crh) cells that project to median eminence (me) and exemplify neuroendocrine parvicellular innervation. LIMBIC: limbic structure inputs, BST: bed nucleus of the stria terminalis, HYP: hypothalamus, PB: parabrachial nucleus. Afferent Inputs to Magnocellular Neurons: Afferent neural inputs to PVH tend to parcellate into different subdivisions within the nucleus, and in some cases onto specific cell types (Sawchenko and Swanson, 1983; Swanson, 1986; 1987). An obvious implication is that different neural inputs (or combinations of them) subserve specialized functional outputs of those subdivisions or cell 18 types. (See Appendix II for a more detailed discussion of PVH inputs). Innervation of magnocellular neurons is illustrated schematically in the top panel of figure 2.05. Magnocellular neurosecretory neurons were the first recognized and have been the most studied PVH cells (Scharrer and Scharrer, 1940; Zimmerman and Robinson, 1976; Hatton et al., 1976; Russell et al., 1980; Zimmerman et al., 1984; Renaud et al., 1991; Morris and Pow, 1993). Pulsatile Oxytocin release has been shown to be directly related to parturition and lactation in the female and there is a large literature on sensory and autonomic innervation subserving these functions (Lincoln and Wakerly, 1975; Sawchenko and Swanson, 1983). Cardiovascular response to hemorrhage and dehydration is mediated largely by vasopressin release in response to changes in blood volume and osmolarity (Swanson, 1987; Blair et al., 1996; Watts, 2001). It has been shown that oxytocin also participates in fluid homeostasis, via regulation of salt appetite and sodium excretion as well as vasodilatation and vasoconstriction (Stricker and Verbalis, 2004). OXY cells are preferentially innervated by serotonergic projections from several midbrain raphe nuclei, whereas VAS cells receive strong noradrenergic inputs from the A1 nucleus in the brainstem. Raphe nuclei are considered to be locomotor-related structures, and A1 relays information (e.g., baroreceptor, and chemosensory) from the vagus and glossopharyngeal nerves, and presumably from thoracic dorsal root ganglia. So these inputs may eventually be shown to subserve behavioral activation circuits related to maintenance of homeostasis in fluid balance (Sawchenko and Swanson, 1983). Magnocellular neurons (predominantly VAS) also receive afferents from the subfornical organ (SFO), which lacks a blood brain barrier, and responds electrically to changes in circulating levels of angiotensin II, a potent stimulus for drinking behavior (Lind et al., 1984; Lind, 1986). In addition, VAS and OXY neurons in the PVH receive an equal projection from median preoptic (MePO) cells that are themselves strongly innervated by the SFO (Simerly 19 and Swanson, 1988; 1988). The MePO is responsible for a complex of reproductive activities that are known to include changes in fluid balance and oxytocin release. Both types of magnocellular neurons receive a sparse projection from the suprachiasmatic nucleus (SCH), the neural circadian rhythm generator. This input very likely coordinates OXY and VAS release with the well-known circadian rhythms of other hormones. Dorsal medial hypothalamus (DMH) and the arcuate nucleus (ARH) send specific projections to OXY cells, and some subdivisions of the bed nucleus of the stria terminalis (BST) also seem to innervate magnocellular OXY cells preferentially, though not very heavily. Fibers from ARH have been shown to contain peptides (e.g., ACTH, βEndorphin, NPY, AgRP) that stimulate or inhibit feeding. DMH is also broadly associated with ingestive behavior, and the BST is considered a funnel for limbic information. The significance of BST afferents may relate to OXY release observed in response to limbic stimulation, and to the calmed affect and increased feeding associated with lactation (Sawchenko and Swanson, 1983; Thompson and Swanson, 1996; Dong and Swanson, 2006a). Afferent Inputs to Parvicellular Neurons: Afferent projections to parvicellular neurons are illustrated schematically in the bottom panel of figure 2.05. They are diffuse and similar over the subdivisions, with some interesting exceptions. Inputs from other hypothalamic nuclei are fairly sparse and distributed over all parvicellular subdivisions, with DMH and some subdivisions of the bed nuclei of the stria terminalis sending a notably denser innervation (Sawchenko and Swanson, 1983; Thompson and Swanson, 1996; Dong et al., 2001; Dong and Swanson, 2004a; 2004b; Dong and Swanson, 2006a; 2006b). Extremely sparse SCH inputs are directed primarily at the ventral periventricular portion, through which traverse axons projecting to the neurohemal zone of the median eminence (Watts et al., 1987). These may well determine much of the circadian rhythmicity in hormone levels mentioned above. 20 Autonomic and neuroendocrine parvicellular subdivisions receive generally similar afferents, with a few significant differences. Midbrain serotonergic inputs are localized to periventricular (neuroendocrine SS and TH-containing) and dorsal parvicellular (predominantly autonomic, spinal sympathetic-projecting) subdivisions (Sawchenko and Swanson, 1983). A similar pattern is seen with innervation by ACTH/beta endorphin fibers from the arcuate nucleus, except that they tend to end in all autonomic subdivisions (Sawchenko et al., 1982). Brainstem and midbrain projections to parvicellular PVH include adrenergic and noradrenergic fibers from the nucleus of the solitary tract and other dorsal vagal complex nuclei, from locus ceruleus, parabrachial nucleus, periaqueductal gray matter and the laterodorsal tegmental nucleus. All of these structures relay primary or secondary viscero- sensory information, and some of them in turn are in receipt of locomotor and descending behavioral information (Sawchenko and Swanson, 1981). Serotonergic inputs from raphe nuclei innervate autonomic divisions more strongly (as noted above), and are low in PVH relative to the immediate surround, a fact that may relate to presence of local inhibitory neurons in that area (Boudaba et al., 1996). This pattern is somewhat complementary to adrenergic fibers, which strongly innervate periventricular and dorsal parvicellular areas and are notably absent from magnocellular divisions. In contrast, noradrenergic fibers innervate parvicellular and magnocellular neurosecretory divisions as well as those with both types of autonomic projections (Sawchenko and Swanson, 1983). It is tempting (and convenient) to parcellate importance of afferents according to functions of their target cells, but this may be quite simplistic. For example, in a study of immediate early gene activation Martha Blair and co-workers proposed that cardiovascular information relayed to parvicellular PVH autonomic regions may be used to modulate behavioral, rather than homeostatic responses to dehydration (Blair et al, 1996). 21 From the forebrain, SFO, MePO, and BST projections to parvicellular divisions are much more robust than to magnocellular areas (see previously cited references). There is also a diffuse projection from the substantia inominata (SI) to the periventricular subdivision, which may correlate with its nucleus acumbens-associated connection to the meso-limbic dopamine system (Swerdlow et al., 1984; Swanson, 1991). SFO and MePO projections are especially dense in the medial parvicellular (CRH-containing) part compared to other parvicellular subdivisions, and this projection, along with that to the autonomic parts, corresponds well with studies of coordinate ACTH release and body water regulation (Lind et al., 1985). Information funneled from amygdala and subiculum via massive BST input can be related to known limbic and cognitive effects on the stress response. Diencephalic inputs include those from the paraventricular thalamus and nucleus reuniens, both of which receive projections from hypothalamic structures as detailed below, and are likely correlated with motor aspects of behavior (Swanson, 1987). As noted earlier, hypothalamic afferents to PVH are sparse and widespread, with a slight preponderance of fibers in medial subdivisions. Considered in total, however, this sparse innervation represents a rich input from virtually every hypothalamic area (Sawchenko and Swanson, 1983). All major hypothalamic subdivisions (excepting medial and lateral mammillary nuclei) send projections to parvicellular subdivisions of PVH. Specifically, projections have been demonstrated from medial preoptic (MPO) and lateral hypothalamic areas (LHA), and from anterior (AHN) (Risold et al., 1994), ventromedial (VMH) (Canteras et al., 1994) and dorsomedial (DMH) hypothalamic nuclei as noted above. VMH is involved in feeding, rage, and female sexual behaviors and is consequently much studied (Simerly and Swanson, 1988; Canteras et al., 1994). It has a heavy limbic interconnection via the BST and amygdala, and a sparse input from LHA. LHA cells are diffusely scattered rostro-caudally, and their afferents to PVH travel through the medial 22 forebrain bundle where they are in a position to interact with many other hypothalamic fibers. LHA has been “implicated in the processing of sensory information, as well as the modulation of somatomotor responses, particularly in relation to the expression of behaviors associated with hunger and thirst, aggression and reproduction” (Swanson, 1987). Further information about connections of individual subdivisions in PVH (efferent and afferent connections), can be found at the Brain Architecture Management System, or “BAMS” (http://brancusi.usc.edu/bkms). This online repository of neuroanatomical information has a set of inference engines for processing neurobiological data that can produce dynamic and comprehensive listings of relationships between named brain structures. It contains “on the order of 40,000 reports of connections between different brain structures in the rat, as collated from the literature” (Bota et al., 2003; Bota and Swanson, 2005). Many of the references cited above are represented in this database, as are most of the PHAL (phaseolus vulgaris - leucoagglutinin lectin) anterograde tract-tracing studies published by L. W. Swanson and colleagues. It is clear from the above discussion of PVH inputs that neuroendocrine and autonomic output cells of the PVH have access to information from all central sensory and motor systems. This is essential for integration of neuroendocrine and autonomic functions with behaviors that assure survival of the individual and the species. The robust innervation by DMH, relative to other hypothalamic structures is significant in regard to DMH involvement in arousal and ingestive behaviors. This may well be enhanced by the selective serotonergic and ACTH/beta endorphin innervation of periventricular and autonomic divisions of PVH. Scant innervation by the SCH circadian pacemaker seems at first surprising, in view of well-known (light-entrained) circadian rhythms in levels of various hormones. However, this projection does originate preferentially in the ventral (retinal recipient) part of the SCH. SCH projects massively and reciprocally to the sub- 23 paraventricular zone, which has been shown to have a complex inhibitory relationship with PVH neurons, and research on the SCH is beginning to reveal a complicated pattern of inhibitory connections that produce circadian rhythmicity. Another complexity of hypothalamic innervation is the possibility of parvicellular-magnocellular communication within the PVH, as noted above. In addition, hypothalamic afferents to PVH are illustrative of an important fact not mentioned earlier: most intra-hypothalamic connections are to some extent reciprocal. Thus, PVH afferent and efferent information has access to virtually constant update, which can only enhance (and is probably essential to) information for endocrine and autonomic control. Neural afferents are often considered primary in control of PVH function, however another type of input (humoral, or blood-borne) deserves important mention. The PVH is perhaps the most richly vascularized area of the brain (Craige, 1940; Swanson, 1986), and therefore seems almost certainly to be affected by changes in circulating levels of a variety of blood-borne chemicals. Obvious examples of substances that cross the blood-brain barrier with profound functional consequences are gonadal and adrenal steroids, for which PVH cells are known to express receptors (Simerly et al., 1990). Circulating corticosterone, for example, is known to exert negative feedback control on corticotropin releasing hormone (CRH) synthesis in PVH parvicellular neurons of the rat (Swanson and Simmons, 1989), and is thus a significant modulating factor in the stress response. It is also quite possible that changes in PVH capillary permeability under varying physiologic conditions could alter sensitivity to circulating messenger molecules (Hartman et al., 1980). Since microvascular control is known to be largely autonomic in origin (Raichle, 1975; Swanson et al., 1977), and highly correlated with some primary PVH functions (e.g., body water homeostasis and cardiovascular function), the significance of possible PVH responsiveness to subtle humoral signals is without doubt interrelated with specific neural inputs. 24 Physiology: There has been a great deal of work on neural versus chemical influences, and on local inhibitory inputs regulating the electrophysiological activity of various cell types in the PVH as alluded to above. (See Appendix III, for a more detailed discussion of PVH electrophysiology). Most of the resulting data support the prevalent working hypothesis that at least three major cell types, based on cytoarchitectonic morphology and functional efferent connections, exist in the PVH and can be identified by their electrophysiological signatures. Both in vivo, and in vitro slice preparations of various kinds have been used to investigate PVH electrophysiology. The earliest classic studies of PVH neurophysiology were of antidromically identified magnocellular neurons. They helped Lincoln, et al elucidate the synchronized magnocellular OXY neuron firing that preceded and predicted the milk ejection reflex in lactating rats (Lincoln and Wakerley, 1975;). OXY cells in supraoptic nucleus (SON) and PVH have been shown by dye transfer studies to contain gap junctions, implying electrical coupling (Andrew et al., 1981). Studies of glial withdrawal from basement membrane between OXY neurons and terminal processes show that direct apposition between these cells increases during osmotic manipulations (Hatton et al., 1984). These observations may well account for the mechanism(s) underlying observed coordinated firing of OXY neurons during lactation. A physiological study of response to limbic stimulation showed a convergence of inhibitory limbic input on individual neurons, and that neurosecretory cells which were influenced by limbic stimuli were also inhibited by baroreceptor activation and excited by osmotic stimulation (Ferreyra et al., 1983). Though interesting, the significance of this study is not entirely clear, since the cells were not identified as containing OXY or VAS. In addition, recent studies have elucidated dendritic release and autoreceptors for OXY and VAS in magnocellular neurons (Ludwig and Pittman, 2003). This implies a complicated and sensitive local control of OXY and VAS release under 25 normal physiologic conditions, i.e., in males, or non-pregnant, non-lactating females. In contrast, experiments in lactating animals reveal a different response to infusion of exogenous vasopressin between anterior and posterior magnocellular divisions in PVH. A state dependant effect of infusing vasopressin on firing rate of cells within a PVH subdivision (inhibition of highly active cells, excitation of relatively inactive cells, and no change in moderately active cells) has also been shown. Thus, local dendritic release of vasopressin may be a normalizing or coordinating mechanism for hormonal release from VAS neuron axon terminals (Tasker et al., 2002). Electrophysiological recordings in parvicellular PVH neurons have been technically much more difficult to obtain, but some interesting results have emerged. For example, paraventricular parvicellular neurons display low- threshold potentials that generate one or two action potentials (Hoffman et al., 1991). Daftary, Boudaba and Tasker investigated noradrenergic regulation of parvocellular neurons, identified by their localization within PVH slices and lack of staining by neurophysin (OXY or VAS precursor molecule). They suggested, “excitatory inputs to parvocellular neurons of the paraventricular nucleus are mediated mainly by an intrahypothalamic glutamatergic relay, and that only a relatively small subset of paraventricular parvocellular neurons receives direct noradrenergic inputs, which are primarily inhibitory” (Daftary et al., 2000). Place in behavioral circuits: Integration of behavior with endocrine and autonomic function implies a complex circuitry that allows access to behavioral state (nutritional status, metabolic rhythms, the sleep wake cycle), cognitive (memory, decision-making, affect or emotional tone), and viscero-sensory (cardiovascular, metabolic, digestive) information (Swanson, 1987; Swanson, 1991). Analysis of afferent connections illustrate how neural inputs to PVH might co-ordinate endocrine and autonomic function in the maintenance of homeostasis, and 26 participate in circuits which subserve behavior patterns crucial to preservation of the individual and the species (Thompson and Swanson, 2003). To that end, and as more precise and specific input data continue to accumulate, it becomes increasingly important to obtain data for a high resolution model that shows the location and comparative relationship of the major neuroendocrine cell types in the PVH. This concept leads directly to the rationale for the experimental work presented here. It focuses on characterization of the distribution of secretomotoneurons in the two neuroendocrine compartments—magnocellular and parvicellular—surveying chemically characterized cell types with identified endocrine function in subdivisions of the PVH. 27 Chapter 3. Research Rationale and Experimental Design Overview: The accepted view of chemical anatomy of neuroendocrine motoneurons in rat PVH is based on literature showing phenotypic segregation of cell types that corresponds fairly well to cytoarchitectonically defined subdivisions. It is known in general which areas of the nucleus contain which peptides or transmitters, and it is clear that many neurons contain more than one neurochemical. Much less is known about quantitative distribution patterns of individual cell types in PVH, or distribution of different cell type populations in relation to each other at high resolution. Cells containing the major hypophysiotropic releasing factors have been localized in defined areas of anterior and medial parvicellular PVH. Similarly, cells containing oxytocin and vasopressin have been demonstrated within sub-compartments of the magnocellular PVH subdivisions. This has led to a segregated model of neuroendocrine cell type distribution, illustrated schematically in figure 3.01. The segregated model is a composite based on numerous surveys of neuropeptide protein and messenger RNA content in the rat hypothalamus (Swanson, 1991), and see for example (Swanson and Simmons, 1989). A few of these studies labeled neuroendocrine cells with retrograde tracers injected into the posterior pituitary or median eminence, some were performed on paired serial sections using antibodies to different target peptides (or by the elution and re-staining procedure using antibodies to two different peptides sequentially in the same section), while others employed markers to two different peptides in the same section surveyed at selected levels through PVH (Ceccatelli et al., 1989; Meister et al., 1986; Markakis and Swanson, 1997). 28 Fig. 3.01 Segregated model—five neuroendocrine cell types at the “classic level” in PVH. Parvicellular neurons containing the major hypophysiotropic factors have been localized in cytoarchitechtonically-defined areas of PVH as described in the text. Numerous SS and a smaller number of TH cells are seen in the periventricular zone (pv). TRH and CRH cells occur in overlapping populations of the medial parvicellular zone (mpd), with TRH more concentrated toward the dorsal part and CRH more concentrated in the ventral part. VAS-containing neurons are localized toward the center of posterior magnocellular divisions (pml shown here), while OXY-containing cells tend to surround them in a “shell”, especially in the lateral part of the subdivision. Separate anterior and medial magnocellular divisions (not shown) are almost exclusively composed of OXY-containing cells. This segregated anatomic model of neuroendocrine cell phenotype distribution has a profound effect on interpretation of a variety of clinical and research data regarding the PVH. Analysis of functional, developmental and pathological studies, and anatomical and 29 physiological investigations like those mentioned above is all influenced by assumptions about neuroendocrine cell type distribution. With the aim to further refine the segregated model, experiments described in this work were designed to compare locations of defined hypophysiotropic neurons throughout PVH in rat, from one animal to another. A further goal was to produce a high-resolution, comprehensive description of the relationship between distributions of all the major neuroendocrine peptide-containing cells (neuroendocrine motoneuron pools) in the PVH. Following are discussions of “Experimental Rationale Questions”, “Practical Experimental Considerations”, and “Specific Experimental Design”. Experimental Rationale Questions: Newly described data from several types of studies show increasing specificity and high resolution of inputs to PVH. Examples are PHAL anterograde tracing, retrograde protein tracing, trans-synaptic virus studies, identified protein content of afferent fibers, and markers for gene expression patterns and putative activated cell function (e.g., c-fos) after pharmacological manipulation (Swanson, 1991; Bittencourt et al., 1991; Sawchenko and Pfeiffer, 1995; Viau et al., 1999; Thompson and Swanson, 2003; Viau et al., 2003; Fekete et al., 2004; Kovacs et al., 2004; Rinaman and Schwartz, 2004; Wen et al., 2004; Wittmann et al., 2005; Sarkar et al., 2003; Bali and Kovacs, 2003). It is apparent that more precise information about neuroendocrine cell type (chemical phenotype) distribution within subdivisions in PVH is needed to complement this high-resolution anatomical and functional data. A number of questions arise in the quest to understand what cell type or combination of cell types might be the target of a defined input. This is important in considering what coordinated output might be predicted, or in otherwise evaluating what the significance of a given input might be. How well does the distribution of neuroendocrine cell types (as defined by their endocrine stimulating or inhibiting neuropeptide content) correspond to cytoarchitechtonic subdivisions and consensus functional compartments in PVH? How 30 much, if any, intermixing of cell types occurs within their distribution patterns? Can a more precise, high-resolution delineation of anatomic distribution for neuroendocrine cell types of defined phenotype be elucidated for PVH? Applying combinations of primary antibodies using immunohistochemical staining (immuno-staining) procedures has revealed that PVH neurons of a defined phenotype can co-express a variety of other peptides under different physiological conditions as discussed above. However, surprisingly, there has been no systematic survey at high resolution using double immuno-staining staining of defined primary neuroendocrine cell types in the same section to determine their anatomical localization in relation to each other. Are all cells of presumed hypophysiotropic phenotype exclusively neuroendocrine, and are the motoneuron pools that control the major endocrine organs precisely distributed? Are they intermixed or separate, and how much if any co-localization of the major phenotypes in the same neurons can be identified? The work presented here is intended to address these questions. Figure 3.02 shows several possible models for relationships between different neuroendocrine motoneuron cell types in PVH. Preliminary experiments aimed at more precise localization of the two magnocellular cell types (OXY, VAS) in the same section throughout a high- resolution series revealed more heterogeneity than the current segregated model supports. See Appendix IV for details of those experiments. Thus, an alternative working model was formed—one that postulates greater intermixing of phenotypically defined neuroendocrine cell types within subdivisions of PVH, as schematized in figure 3.02 (panels c and d). The concept of greater PVH cell type heterogeneity shaped the design and careful observation of results in subsequent experiments presented here. 31 Fig. 3.02 Possible Motoneuron pool relationships illustrated in five Venn diagrams. For PVH the first model, schematized in a, might be postulated for periventricular and lateral magnocellular neurons, while model b could be proposed for medial parvicellular and lateral magnocellular cells. The commonly accepted “segregated” model illustrated earlier is more similar to c, with some intermixing of cell types seen especially in medial parvicellular neurons. Preliminary studies suggested that example d might represent a better model for observed heterogeneous relation of cell phenotypes across cytoarchitechtonic borders of PVH. Complete intermixing of cell types within or between subdivisions of PVH, as in e, is not expected from preliminary studies or observations reported in the literature. Practical Experimental Considerations: Distribution of defined motoneuron pools has traditionally been determined by introduction of retrograde tracers into their axons or terminal fields. Thus a cut-nerve was immersed in, or a specific muscle was injected with a tracer that traveled by retrograde axoplasmic flow to the motoneuron cell bodies of origin. There it could be visualized by virtue of its intrinsic fluorescence or by immunohistochemical reaction. See for example (Hollyday, 1980; Chiken et al., 2001). In that way a population of labeled motoneurons that supply innervation to a specified target—usually a muscle, muscle-segment or related group of muscles—was identified. However, analysis of such work is hampered by the difficulty of isolating and enclosing a cut nerve uniquely and completely so that all axons are labeled, or 32 achieving complete target injections that successfully expose all motor terminals in a given muscle to the retrograde labeling substance. Aside from potential problems of contamination from other inputs, this approach can be compromised by incomplete or questionable label of the desired target in an individual experiment. Thus an underestimate of the total motoneuron pool in question can (and probably does) result. However, the basic approach remains useful and has been extended. For example Chiken, Hatanaka, and Tokuno (above) used three fluorescent tracers to simultaneously distinguish between individual motoneuron pools in the spinal cord innervating flexor and extensor muscles involved in grasping, and several workers have used retrograde trans-neuronal transport of viral vectors to elucidate central innervation of autonomic targets (Strack et al., 1989; Rinaman and Schwartz, 2004). Specific Experimental Design: For all practical purposes the confounding problem of incomplete target label can be eliminated when studying the neuroendocrine motor system, and specifically the endocrine motoneurons within PVH. Here we have the advantage of using the fluorescent dye Fast Blue (FB) as a retrograde tracer injected peripherally in the blood stream, which gives it access to all axons terminating in association with blood vessels or areas that lack a blood brain barrier (Rho and Swanson, 1987; Markakis and Swanson, 1997). Thus, by definition, all active neuroendocrine motoneurons that release peptides into the bloodstream will be labeled via endocytosis and retrograde transport of tracer (in this case, Fast Blue) from intact and active axon terminals associated with blood vessels (Ambalavanar and Morris, 1989). Recall that PVH neuroendocrine axon trajectories have been well characterized and defined according to their projections through the internal and external lamina of the median eminence where they associate with blood vessels in the portal vasculature and posterior pituitary, as discussed above. Their cell bodies can thus be identified, based on retrogradely 33 transported Fast Blue visualized with ultraviolet fluorescence microscopy. In contrast, neurons with conventional axon terminals inside the blood brain barrier of the central nervous system remain unlabeled. Cells within defined subdivisions in PVH can be identified as hypophysiotropic neurosecretory neurons unequivocally by virtue of retrograde label content. In the same way, PVH neurons without retrograde label may be assumed to be non-neuroendocrine in function. Neurons outside PVH may be labeled due to their termination in other areas lacking a blood-brain barrier, but they are chemo-architectonically distinct from PVH (Merchenthaler, 1991). Further, within the relatively compact rat PVH, specific anterior pituitary stimulating cell types can be reliably identified by fluorescent antibody staining for the peptide phenotype of individual cells. Thus, in a closely spaced series of sections a defined hypophysiotropic neuroendocrine motoneuron pool can be determined almost completely and with little or no ambiguity. In a similar way, the two magnocellular phenotypes can be identified together, or in combination with other neuroendocrine peptide markers by double fluorescent immunohistochemical staining in the same sections. Double staining for neuroendocrine phenotypes (two antibodies to different cell types stained with different fluorescent markers in the same section) in retrogradely characterized cells was the approach used in these experiments to define the complete distribution and extent of overlap for the major neuroendocrine motoneuron pools in PVH. This work comprises a high-resolution study to refine the current distribution model of primary hypophysiotropic cell types between animals, and to systematically describe the neuroendocrine PVH chemoarchitecture much more completely than has been done in the past. Such a comprehensive, comparative survey of neuroendocrine cell type distribution has important implications for interpretation of other anatomical work and for analysis of physiological and developmental studies, since they are based on assumptions about underlying morphological information. 34 A secondary goal of this work was to determine anatomically accurate and reproducible ways to describe distribution of neuroendocrine motoneuron pools in PVH—to facilitate comparison between experimental animals and to view in reference to a standard published atlas. This is an important technical problem, whose solution is essential for evaluation of this and future work in relation to extant and rapidly accumulating literature. The procedures used to illustrate spatial distribution of cells in PVH within the context of a published atlas, Swanson’s Brain Maps, will be described more fully in Chapters 5 and 7. PVH is a compact nucleus defined by relatively visible borders seen in Nissl sections of rat brain. However, it is located within the hypothalamus where there are few if any rigid anatomical landmarks. Thus, a method to consistently identify subdivisions within PVH and their position relative to the entire hypothalamus was needed. Such an approach is conceptually different than that for studying isolated or well circumscribed and cytoarchitectonically distinct areas, such as the facial nucleus or the dentate gyrus of the hippocampus. The method presented below is offered as a possible exemplar for examining regional organization in the CNS to accommodate other areas with imprecise borders. Such areas may be typified by groups of cells within PVH that can be defined by their histochemistry or afferent input rather than typical cytoarchitectonic borders seen in a Nissl stain. 35 SECTION II: EXPERIMENTAL DETAILS Chapter 4. Experimental Materials and Methods Overview: Described here are essential details of tissue preparation methods, and staining strategy implemented to assure a complete, high-resolution survey of target neuroendocrine motoneuron pools. Spatial distribution patterns for a total of five parvicellular (CRH, TRH, SS, TH, and GRH) and two magnocellular (OXY, VAS) neuroendocrine cell types were identified in closely spaced series of histological sections throughout the PVH. Section thickness and number of series per animal were chosen to optimize accurate cell-counting and assure only slight change in cell type distribution between adjacent sections in a series, as described in “Histology”, below. It is not technically possible at present to reveal every cell type of interest, at the same time, in the same tissue section. However, with double-staining using two primary antibody probes and different color fluorescent labels one can demonstrate with certainty two peptides in the same section—in the same cell if they are co-localized. A fluorescent retrograde marker of a third color (Fast Blue) is used to show whether a cell’s axon terminals deliver neurotransmitter to the bloodstream, thus identifying it as neuroendocrine. Using this approach, distribution of neuroendocrine motoneuron pools and their relationship to each other in PVH was determined by immunohistochemical staining with various antibody pairs to target the seven phenotypes of interest. Systematic combinations of neuroendocrine peptides were antibody-stained in closely spaced histological series from fifteen experimental animals. Antibody probes for each target cell type were used two or more times in combination with those for every other target cell type: in the same animal, in different animals, and with different primary antibody source reagents for a given cell type. This assured that a consistent and reproducible staining pattern for each phenotype was 36 obtained. Schematic representations of this approach are illustrated in figures 4.01 and 4.02, and a more detailed discussion follows in the section on Histology. Some series through PVH were analyzed to confirm and optimize staining parameters and verify specificity and compatibility of antisera, while five were mapped in total to reveal cell type distribution. Details of antibody testing are in Appendix IV. Typically, one antibody probe was held constant between at least two adjacent series in a single animal to confirm staining patterns from one adjacent set to the next, as schematized in figure 4.01. Fig. 4.01 Schematic representation of the four stained sets of sections from one animal. A through D: vertically arranged rectangular boxes represent four 1-in-4 series of 15µm sections through PVH in one rat brain. The horizontal boxes with diagonal separation indicate a series (or, e.g., one example section), stained with a combination of two different antisera. A=Nissl, B=SS+TRH, C=CRH+OXY, and D=CRH+VAS. Symbols and background colors [variable grayscale in this black and white image] shown on the left are correlated with the corresponding peptide target indicated on the right. Inverted filled triangle=VAS, open box=CRH, filled circle=OXY, script ‘X’=SS, asterisk=TRH. These symbols, and some others, are used for clarity in subsequent diagrams and maps when more than three peptide cell types are illustrated together. 37 Fig. 4.02 Schema for adequate mapping of all target cell types in relation to each other. Seven neuroendocrine cell types are indicated at left: VAS, OXY, CRH, TRH, SS, TRH, and GRH. Each is boxed with a unique background color [gray-value in this black and white image] and symbol associated with it in subsequent composite maps. To the right are copies of staining schema (in the style of figure 4.01, above) for each series in an animal where that cell type was surveyed. Background colors and symbols correspond to the cell type definitions at left. For each cell type, a dotted line extending to the right indicates staining in the specific series (A-D) of individual experimental animals—PVH10, 12 13 or 14. Thus, one sees the number of complete series stained for each cell type, and which other cell types they were compared with—in the same sections, in other series from the same animal or in other animals. Two or more animals and series were surveyed for each cell type. Note this is a schematic of the experimental design, not a complete list of all cases analyzed. 38 Different combinations of antisera were used in other cases, where one or more of the former antisera were tested again. In this way, a consistent “linkage” of the staining pattern for a given peptide was obtained—between adjacent sets in the same animal and between different animals. See the staining schema diagram in figure 4.02, and example maps illustrating the principle in Chapter 5. This approach verified the cell type distribution and reproducibility of staining for a given peptide while providing additional data for eventual use in populating a 3D database. In some cases, another approach was taken: a single antibody was stained in combination with three (or four) different comparator antibodies in successive sets from the same animal (for example, TH staining in PVH14). Thus, a near- serial section resolution of cell type distribution was obtained with the first antibody, and the distribution pattern for that phenotype could be compared to a maximum number of others in a single animal. See figure 4.02. Material from three exemplar animals (PVH10, PVH12 and PVH14) was subsequently analyzed in greater detail for final presentation, as will be described in Chapter 5 and illustrated in Chapters 8-12. These cases include one or more examples of all tested cell types. Importantly, across the three cases are representative double-stained sets for key intermixed cell types in all relevant combinations. General Experimental Procedures: National Institutes of Health and Society for Neuroscience guidelines, and University of Southern California Institutional Animal Care and Use Committee approved protocols for humane animal handling and surgical procedures were used throughout. Experimental animals were adult male (325-350g) Harlan Sprague-Dawley rats maintained on a 12-hour light/dark cycle under standard vivaria conditions. All procedures were performed in the early afternoon, before the daily adrenocorticotropin hormone (ACTH) surge that precedes onset of circadian activity in these nocturnal animals. This is an important consideration for 39 experimental design, as it is known there are daily fluctuations in chemical content of at least some PVH neurons. For example, variations in RNA message for CRH have been documented over the circadian cycle (Watts and Swanson, 1989; Watts et al., 2004). Moreover, it is almost certain that synthesis and release of other molecules of interest fluctuates over time, as indicated by the fluctuating amounts of the various circulating endocrine hormones that they stimulate (Sowers, 1980). To label all neuroendocrine neurons the fluorescent retrograde tracer Fast Blue, which emits blue fluorescence under ultraviolet illumination, was injected into the bloodstream via the jugular vein (Rho and Swanson, 1987; Markakis and Swanson, 1997). After such a Fast Blue injection, axons terminating in apposition to blood vessels internalize Fast Blue by terminal endocytosis and transport it by retrograde axoplasmic flow to their cell bodies of origin. There, it accumulates and can later be detected microscopically in histological sections. This approach to characterizing neuroendocrine cell types—by definition, because of their ability to retrogradely transport various molecules from the bloodstream or via the bloodstream from the periphery (e.g., from intraperitoneal injections of tracer)—has been used by others for some time and is now well accepted (Armstrong and Hatton, 1980; Meister et al., 1988; Silverman et al., 1990; Merchenthaler, 1990, 1991; Decavel and van den Pol, 1992). Fast Blue shows no apparent damage to neurons (over at least several weeks), unlike the more soluble retrograde fluorescent tracer Fluoro-Gold, which has been shown to be neurotoxic over the long term (twenty weeks) when retrogradely transported to motor neurons from cut nerves (Garrett et al., 1991), and at the site of pressure injections directly into brain tissue but not in retrogradely labeled central neurons (Schmued et al., 1993). When performing fluorescence immunohistochemistry for many peptides and other small molecules in neurons, pretreatment with intraventricular colchicine is required 40 (Weisenberh et al., 1968; Dube and Pelletier, 1979). Colchicine is considered primarily a microtubule transport inhibitor that prevents axon transport of proteins after synthesis, allowing them to accumulate in the cell body in amounts more easily detectable by antibody staining methods. Some authors disagree on the exact mechanism of observed peptide accumulation in the cell body (Alonso et al., 1988) but concur that synthesis continues in the short term after colchicine treatment. This accumulation allows for adequate visualization in neuronal cell bodies of peptides and other protein molecules that would otherwise be rapidly transported away from the soma for release at axon terminals (Parish et al., 1981; Daikoku et al., 1985). However, colchicine is in essence a biologic toxin. Therefore, possible negative effects of its use in this work were taken into consideration. Some studies have shown that it can induce or alter expression of peptide gene products (not those surveyed in these studies) in hypothalamic neurons (Cortes et al., 1990; Kay-Nishiyama and Watts, 1999). See also Appendix IV: “Chart of Experiments”, preliminary studies from this work, for unpublished observations of pre-proENK (enkephalin) mRNA expression in hypothalamus of colchicine-treated rats compared to little or no expression in the same areas in untreated rats (quantified by Alan Watts). Colchicine shows selective neurotoxicty to dentate gyrus granule cells (and not pyramidal neurons) following intrahippocampal injections (Brady et al., 1992), or to cerebellar granule cells when injected into the lateral ventricle (Ceccatelli et al., 1997). Changes in CRH expression observed by Brady and co-workers in PVH following hippocampal colchicine injection are apparently related to subsequent alterations in afferent inputs, not to direct effects of colchicine on PVH. Ceccatelli et al. did not show concomitant effects in PVH following injections similar to the ones in the work presented here. Colchicine pre-treatment clearly produces patterns of PVH peptide staining consistent with physiological stress (Berkenbosch and Tilders, 1988; de Goeij et al., 1991; Ceccatelli et al., 1991), but it does not alter the number or spatial distribution of neurons surveyed. Parallel in situ 41 hybridization studies of mRNA distribution for the major peptides surveyed here in colchicine-treated, vs. normal control animals showed no significant difference in cell type distribution pattern. See Appendices IV and V. The exception noted above for changes in pre-proENK expression were not contraindicative for colchicine use, since enkephalin (while important in evaluating stress response) was not one of the primary peptides surveyed here. In spite of potential drawbacks, almost all previous studies showing immunohistochemical localization of hypophysiotropic proteins have used colchicine treatment. Since the work here is aimed at identifying locations of all relevant peptide expressing cells, rather than assessing physiologic changes (for example, those produced by stress) colchicine treatment was a method of choice. It is an accepted way to better visualize low abundance antibody targets when using immunohistochemical probes. Thus, synthesized proteins that might otherwise be transported for release at axon terminals accumulate in neuronal somata to a level that can be adequately visualized with antibody staining. All animals used in the study were very deeply anesthetized and killed by trans- aortic perfusion of physiologic saline followed by fixative, using a protocol optimized for antibody staining of peptides and other small protein molecules (Berod et al., 1981). The entire hypothalamic block containing PVH was frozen, and sectioned on a sliding microtome in a frontal plane closely approximating that of a published brain atlas (Swanson’s Brain Maps, 1998). Immunohistochemistry (IHC) was performed on free-floating sections with antibodies to two peptide targets using different color reporter molecules—Rhodamine or Texas Red™ (red emission under green excitation wavelength) and Fluorescein (green emission under blue excitation wavelength). Surgical Procedures: Seven to fourteen days before perfusion, animals were anesthetized with halothane and 2% oxygen, using a gas anesthesia inhalation machine for rodents. A jugular vein was 42 cannulated with a 27-gauge needle attached to a 1cc tuberculin syringe filled with 2.5mg/ml aqueous Fast Blue (Illing GMB™, or Sigma™). After assuring adequate cannulation (venous blood reflux on slight plunger withdrawal) the retrograde tracer (about 0.8ml: at 0.6mg/100g body weight for a 350g rat) was infused very gradually, over about 3 minutes. Fast Blue is moderately soluble in water and not soluble in physiologic buffers. It was therefore necessary to inject it in an aqueous solution; and slow and gentle infusion was essential to prevent osmotic shock. Preliminary experiments and the work of others in our laboratory showed that both left and right PVH nuclei labeled equally well if tracer was injected on only one side (usually the animal's left), and there was little or no pericyte label or diffusion of fluorescence from neuronal cell bodies for up to three weeks (Rho and Swanson, 1987; Markakis and Swanson, 1997). Two days before perfusion animals were anesthetized with a freshly mixed xylazine/Ketamine™ solution (equal parts 20mg/ml xylazine and 100mg/ml ketamine HCl) injected intramuscularly at 0.1ml/100g body weight. A 30-gauge cannula was placed stereotaxically in the left lateral ventricle and about 20µl of 4mg/ml colchicine (Sigma™) in physiologic saline was allowed to slowly infuse by gravity flow, so that intraventricular pressure was not artificially increased. The dose was equivalent to 23µg/100g body weight, e.g., 80µg total in 20µl for a 345-350g rat. Colchicine was instilled on the left side—opposite to that typically used for mapping and analysis, to minimize possible bilateral difference effect of colchicine or interference from any tissue distortion caused by the cannula track. Labeling in the left PVH, surveyed during preliminary evaluation of all staining series, was always qualitatively similar to that on the right (mapped for data analysis) side. Perfusion and Tissue Preparation: A few minutes before perfusion fixation, animals were placed briefly in a bell jar containing a few milliliters of halothane. This light preliminary anesthesia was performed to 43 minimize stress of handling. Rats were then deeply anesthetized with an intraperitoneal injection of 65mg/ml sodium pentobarbital (10mg/100g body weight, or about 0.55ml for a 350g rat), a dose that would otherwise be lethal after about 10 minutes. The thoracic cavity was opened to expose the heart, the descending aorta was quickly clamped, and a 13- gauge cannula attached to #16 silastic tubing (3.1mm inside diameter) was inserted through the ventricle up into the ascending aorta. The left auricle was incised for fluid drainage immediately before beginning perfusion of saline. A Cole-Parmer™ peristaltic pump produced perfusion at 25-30ml/min flow-rate. Room temperature physiologic saline (about 50-100ml, or enough to clear all blood-flow) was perfused through the ascending aorta and thus into the head and brain. The saline was immediately followed by ice cold 4% paraformaldehyde perfusion: first at pH 6.5 (250ml 4% paraformaldehyde in 0.01M sodium acetate buffer) for about 10 min, then at pH 9.5 (450ml 4% paraformaldehyde in 0.01M sodium tetraborate buffer plus 1ml of 2.5% electron microscopy grade glutaraldehyde) for about 15 minutes according to the method of Berod, et al. as modified for use in our laboratory. This perfusion fixation protocol was developed for optimal fluorescence immunohistochemistry to demonstrate tyrosine hydroxylase enzyme (TH) and small peptide protein molecules. It has also been used extensively with slight modifications for in situ hybridization and other histochemical procedures (Simmons et al., 1989; Swanson and Simmons, 1989; Simmons et al., 1990). Brains were removed after perfusion and stored at 4°C overnight in 4% paraformaldehyde, pH 8.5 (second fixative, but without glutaraldehyde) that contained 5% sucrose (cane sugar) as a cryoprotectant. The next day, brains were rinsed with phosphate buffered saline, blotted free of excess liquid and attached with Superglue™ cyanoacrylate adhesive, dorsal surface down (to approximate "level cortex", as measured in stereotaxic surgeries) and mid-sagittal plane aligned with an orienting line, on the glass surface of a brain blocking platform. The brain 44 was blocked in coronal plane using a Teflon™-coated single edge razor blade in a Kopf™ stereotaxic electrode holder. The planes of cut were perpendicular to midline, conforming to the Swanson Brain Maps atlas brain (allowing a measured 4°variation). Cuts were placed just rostral to the optic chiasm and immediately caudal to the mammillary bodies to produce a hypothalamic bloc containing PVH and at least 0.5cm tissue rostral and caudal to it. The aim was to achieve consistently parallel rostral and caudal cut surfaces as close as possible to the frontal stereotaxic plane of Swanson's rat brain atlas, Brain Maps: Structure of the Rat Brain (Swanson, 1992). The hypothalamic segment containing PVH was gently lifted away (peeled) from the glass/glue surface, marked with a cut along the left dorsal cortex and rinsed with 5% sucrose in phosphate buffer. Resting on a smooth metal platform (a rod with flattened, L-shaped spatulate end) where the caudal surface remained flat, the tissue was frozen en bloc. It was gradually lowered into a hexane bath chilled to below freezing (estimated about -100°C) in crushed dry ice, where it remained for 5 minutes. After freezing the tissue was either stored in an airtight container at -70°C to -75°C, or mounted on a copper freezing stage (cooled with crushed dry ice in absolute ethanol) attached to a Reichart-Jung™ sliding microtome. The plane of histologic section was determined by cutting a smooth platform of base-ice with the microtome blade and applying the flat-frozen caudal surface of the tissue block to it. This assured subsequent sections were obtained very close to the atlas plane established by the cut originally determined with the stereotaxic brain-blocker. Frontal (coronal) sections cut at 15 micrometers (15µm) thickness from rostral to caudal throughout the entire hypothalamus and overlying cerebral hemispheres. They were saved in sequential order in cold phosphate buffered saline (4°C) for immediate processing, or cryoprotectant solution (glycerol / ethylene glycol / buffer) for storage at -20 C (Simmons et al., 1989; Simmons, 2000). See Appendix VI for solution and storage details. 45 Sections were collected in 4 sequential series (A-D): section 1A was directly serial to section 1B, which was directly serial to section 1C, which was directly serial to section 1D, which was then in turn directly serial to section 2A, and so on. Thus, in a given series (A, B, C or D) two sequentially numbered sections (e.g., B 4/3, B 4/4) are separated by three 15µm sections—thus, a span of 45 µm is missing between them. This is a very thin, closely spaced series in comparison to published atlases and anatomical studies of cell type distribution and connectivity in the rat brain such as those cited previously. Chapters 5 and 6 will offer a more detailed discussion of the significance of section spacing and thickness in evaluating neuroanatomical data. Histology: One series of sections (usually A) was stained with 0.25% thionin for Nissl substance to identify PVH subdivisions using cytoarchitectonic criteria (Simmons and Swanson, 1993), and the other three were processed for selected double-label IHC. See Appendix VI for solution formulae and detailed staining protocols. Two series (B and D) were directly serial [contiguous] to the Nissl set and thus assumed to be extremely similar in architectonic distribution of cell types, while the third (C) was separated from the Nissl series by only 15µm. Thus, it would be expected to vary little compared to the Nissl sections preceding and following it. Parvicellular neurons in PVH average 10-20µm in diameter while magnocellular neurons average 30-60µm in diameter (Swanson, 2000). Thus, this section thickness and sampling frequency assured a very high-resolution survey of each peptide pair stained in the three IHC series. In addition, it allowed accurate identification of cytoarchitectonically-defined subdivisions throughout the nucleus with a resolution of 15- 30µm. For immuno-staining, sections were transferred to a Stain Net™, a 24-well Plexiglas tray with a nylon net bottom (Lonnie Nason, R.G., Nason Machine, Ft. Bragg, CA), and 46 rinsed 1-4hr in potassium phosphate buffered saline (KPBS) plus 0.1% Triton X-100™. Triton is an anionic detergent, a permeabilization agent typically used to ensure adequate penetration of antibody reagents into fixed tissue. Extensive cross-linking of proteins during aldehyde fixation has been shown to form a barrier to antibody penetration, and in some cases it is thought that denaturing fixation might alter the conformation of the protein epitope to which an antibody binds in target tissue. Thus, permeabilization of fixed tissue is used to improve antibody penetration for binding to target proteins. Optimal time and concentration of Triton exposure was previously determined by testing with fixation and antibody staining protocols, as noted earlier. Following two ten-minute rinses in plain KPBS, sections were incubated in “antibody cocktail” (two primary antibodies grown in different host species, each directed to a single target peptide) at appropriate concentrations in KPBS for 48-72h, at 4°C with gentle agitation on a rotator platform. They were subsequently rinsed in two changes of KPBS (ten minutes each) and incubated for one hour at room temperature (or overnight at 4°C) in secondary antibody cocktail. Red-fluorescing Rhodamine (tetra methyl isothiocyanate: TRITC), or Texas Red™ and green-fluorescing Fluorescein (fluorescein isothiocyanate: FITC) were conjugated to secondary antisera directed at IgG of the two different species in which primary antisera were raised. Optimal concentrations and mixtures of all primary and secondary antisera were determined in preliminary test-staining experiments (see Appendix IV). After three KPBS rinses, sections were mounted onto chrome alum gelatin coated (“subbed”) slides, air-dried and cover slipped with 50% glycerol/phosphate buffer mountant at pH 8.5. They were viewed immediately or stored in a dark container at -20°C. See Appendix VI for details of solutions, reagent concentrations, etc. In addition, recall figures 3.01 through 4.02 and note that IHC staining series were designed to compare neuropeptide distributions within and between animals. An overlapping pattern of stained peptides in the same animal and companion experimental 47 animals was chosen. Thus, each peptide was dual-stained with every other, in two or more sets of sections in one or more animals. Some peptides were co-stained with three different comparators in one animal using different primary antisera for the main peptide, and some peptide pairs were compared several times in different animals and with different primary antisera to confirm observations. 48 Chapter 5. Mapping Strategy Overview: Mapping strategy was a crucial element (second only to the design of individual staining series discussed earlier) in planning analysis of the extensive data collected in these experiments. A basic precept of neuroanatomical research is that similar patterns (or differences) demonstrated in normal or experimentally manipulated material can yield generalized information about anatomically connected circuits, and ultimately implications about their function. However, interpreting high-resolution neuroanatomical data even at a semi-quantitative level presents a challenge in that no two brains from members of a given species are identical, though it is agreed they share a very similar common structure. (The same problem holds as well for interpreting physiological data.) Furthermore, consideration of within-animal and between-animal differences in anatomical preparations is important when recording data that consist of maps from individual sections, a topic detailed below in “Variability of Histological Sections”. Initial analysis employed straightforward documentation of cytoarchitectonic and cell type staining patterns from series of Nissl-stained and double immunohistochemistry-stained sections as described in “Data Recording Procedure”. This approach resulted in generation of numerous detailed maps, discussed in “Creating Primary Maps” and “Conversion into Computer Graphics Files”. They reveal patterns of staining and interrelationships between those patterns for different neuroendocrine cell type motoneuron pools across dozens of sections spanning PVH. Twenty-five to thirty Nissl maps plus IHC maps (containing seven information elements) for up to three sets of dual-stained antibody pairs were generated for each experimental animal. The complexity and sheer number of individual maps (500-600 individual layers in a graphics file for some cases!) made the prospect of obtaining a meaningful representation of data throughout PVH a daunting one. In the final topic of this 49 chapter, “Cumulative Mapping from a Huge Data Set: Two Approaches”, the use of graphics files for presentation of an accurate comprehensive view of these high-resolution data sets is addressed. Dynamic analyses of cumulative data from single experimental animals and comparisons of data between animals as viewed in composite files are described in detail. This methodology was used for the first round of data analysis in these studies. Variability in Histological Sections: A major challenge in the studies described here was determining how to record and display a rich and complex data set—locations of chemically characterized cell types seen microscopically in single sections—in a consistent way for a series of sections throughout the PVH from one or more animals. The goal, of course, is to present data in an accurate and understandable way. Then, other similar or complementary data e.g., PVH inputs revealed by anterograde tracing studies, can be compared to them in anatomically precise and functionally meaningful ways. Even in the laboratory rat, a species selected for physical uniformity over countless generations, there is slight variation that results in small biologic differences between individual animals. For example, in the course of this work more than a dozen animals were examined in an attempt to demonstrate a small group of consistently located neuroendocrine magnocellular neurons slightly ectopic to the PVH. That is, cells that were isolated from the main mass of closely packed cells in the nucleus. If their consistent location could be documented, they might provide a reliable target for physiological recording or other studies. A few ectopic neuroendocrine cells were invariably seen laterally near PVH. However, they were not in the same position from one animal to another. This sort of subtle biological difference, in addition to inevitable and potentially significant distortions introduced during tissue preparation, accounts for potential sources of error when preparing anatomical maps. Some amount of error is intrinsic to the study of biological organisms and 50 cannot be eliminated. The goal then, is to minimize it where possible. What are sources of error in experiments described here, and how were they minimized? As seen in Table 5.01, major sources of error can be broadly categorized as pre mortem biological and surgically induced variability, or post mortem changes in tissue and artifacts introduced during histological processing. Types of error in sections used to create maps (aside from intrinsic variability) are linear and non-linear distortion (e.g., stretching), differences in plane of section, and sampling or partial-data error. Biological variability in these experiments was minimized by using animals of a similar age, weight, sex and strain that were housed, surgically manipulated and perfusion- fixed for histological processing under well controlled similar conditions. Lightly fixed (24h) frozen sections were used in histological preparation, rather than a dehydration and embedding technique. This minimized differential shrinkage and swelling between cellular and heavily myelinated brain areas due to fixation and processing. Every effort was made to cut sections in the same plane from animal to animal as discussed earlier in tissue preparation methods, and tissue was blocked for sectioning in a uniform way, as described above. However, some variation in plane of section was unavoidable. Inevitably, slight distortions were introduced by freezing, sectioning and the subsequent staining procedures performed on free-floating sections followed by hand-positioning of sections on microscope slides. Variation between individual microscopic sections (stretching, folding, and minor defects such as bubbles or nicks and tears in the tissue) was minimized by mounting all the sections under similar conditions of temperature, buffer composition, etc. Approaches used to minimize error in recording data from one section to the next in a series are detailed in “Data Recording Procedure”, below. Error due to plane-of-section variation and the problem of incomplete data due to partial sampling are addressed later in Chapter 7. 51 Table 5.01: Types of error, and sources of error in experimental neuroanatomical data. Items starred (*) or **bold font are most relevant in these studies, as discussed in the text. Data Recording Procedure: Sections encompassing PVH were observed on a Leitz Dialux 20™ fluorescence microscope equipped with 100W mercury vapor light source and ploempak filters. Filters used were: "A" (340-380nm ultraviolet excitation for Fast Blue emission), "I2/3" (450-490nm blue excitation for FITC green emission), and "N2.1" (530-560nm green excitation for Texas Red™ or TRITC red emission). Leitz™ 10x highpoint ocular lenses and Nikon-Fluor™ objective lenses were used for observation and photography. Because immuno- fluorescence fades on lengthy or repeated exposure to fluorescent excitation, photography is the best immediate and permanent record of staining. The left PVH was surveyed to assess general staining patterns, and to test photographic exposures. Then the entire right PVH 52 was photographed with a Wild-Leitz™ MPS-46 Photoautomat™ camera system at 100X final magnification: 10X ocular, plus 10X, 0.5 numerical aperture (n.a.) objective lens. This was followed by supplementary spot checks and photographs at 200X and occasionally at 400X (20X, 0.75 n.a. and 40X, 0.85 n.a. objective lenses, respectively). Fig. 5.01 Photomicrographs of original data: Identical 35mm fields at 100X magnification. Photographic images of four 15µm serial sections that comprise one set, showing all staining parameters at one level (4/4) from a single experimental animal (PVH12). White area at left is third ventricle (3V). Midline in these photos, framed to include maximum PVH, runs from near the top left corner diagonally downward in a line parallel to the edge of 3V in the tissue. These images are of dorsal PVH, at about Atlas Level 26 in Swanson’s Brain Maps. Each field was photographed without movement while successively exposed to UV, FITC and Rhodamine filters such that discrete photographs of each fluorescent-labeled entity in a field were obtained in exact register on three adjacent 35mm slides. In one case, a dual combination FITC-Rhodamine filter set (Omega Optical™ XF52 dual band pass filter) 53 was used to simultaneously record both fluorochromes in the same photograph. This was especially useful in photographing some OXY and VAS double-stained sections where there was virtually no expected co-localization or overlap in staining of the same cells; and none was seen in any of the cases examined. In other stained series the dual filter set allowed clear visualization of orange double-stained cells co-localizing two peptides, and single- labeled cells as red or green in the same photographic field. Some sections required a montage of overlapping 35mm fields to encompass all of PVH at the standard 100X magnification. All sections were photographed on 35mm Kodak Elite™ daylight transparency film ASA 400. An optimal fixed exposure (determined for each series) was used to assure correct visual comparison of staining intensity in different cells from the same experiment. See figure 5.01 for typical images of raw data from one experiment, showing different staining combinations in four serial sections, one from each set of a 1-in-4 series through PVH. Creating Primary Maps: Maps were hand drawn from photographic images—35mm slides projected at a fixed distance onto a tabletop tablet, so that an optimally large image of PVH was obtained for drawing in 11" x 14" format. The depth-of-focus range in the film image was used to advantage by zooming the focus slightly when evaluating closely positioned or adjacent cells. This allowed a comprehensive assessment through the depth of the section in a static photo image, and avoided increased fading or possible counting bias introduced by changing adjustment of the Z-focus plane in a ‘live’ microscopic image. Criteria for counting positive cells were the clear presence of a nucleus and/or nucleolus inside a strongly stained cell body, as assessed by comparison with preliminary staining tests for optimization of signal. (See specific details of staining tests in Appendix IV.) This assured that positive cells were counted only once, and partial cells were not counted. Since the stained cytoplasm of very 54 large cells might span the thickness of more than one 15µm section, it was important not to count the same cell in two sets of differently stained sections. In principle, one could record microscopic images electronically with a digital camera and proceed directly to the creation of vector-graphics files from large digital images. Thus, the entire data recording and analysis process might be performed exclusively on a computer workstation with few if any “hard copy” data map versions between initial microscopic observation and publication of results. However, one might lose the visual and mental interpretive gestalt provided by complete PVH overview of all data elements in the same physical map. The choice of 35mm film vs. digital image storage of primary data is, in the end, a decision based on available microscopic and photographic equipment and on personal preference. In the initial maps produced to define the data recording procedure, projected images of cell body outlines were faithfully traced, to give a true anatomical and proportional size relationship among the labeled neurons. Later, standard symbols—circles or crosses centered on the cell body—were used for consistent and efficient recording in hand-drawn maps, as shown in figure 5.02. The change to placing standard symbols on a map at the center of a cell’s image (rather than using different size symbols to correspond to true cell sizes) presented an advantage beyond mere technical mapping convenience. It allowed for accurate recording of cell position (essentially the position of its centrally-placed nucleus) in maps where the microscopic image might be confused by crowded, partially overlapping extended cell bodies. Thus, absolute cell size and shape (the character of fusiform versus more rounded or stellate cells, for instance) are not appreciated in final maps. However, the relative positions of cells in the two dimensional space observed in each microscopic section is accurately preserved as seen in figure 5.02, and further in figure 5.03. It should be noted 55 that the graphic-symbol circles used on final primary maps in Adobe Illustrator™ were equivalent to a 15-20µm-diameter cell size in the original tissue—close to the average parvicellular neuron size in PVHmpd. Therefore, apparent space between magnocellular neurons in a given map may well be more a function of symbol size rather than actual 3D packing of these large neurons. Fig. 5.02 Early (a), later (b) versions: hand-drawn data maps from two serial sections. Sections from adjacent series of similarly stained sections in the same animal. PVH outline and subdivisions were drawn from Fast Blue-defined neuroendocrine borders and deduced from characteristic positions of cell bodies seen in companion dark field views. Sections were stained for CRH and TRH, using the same TRH antibody and two different CRH antibodies (monoclonal- rat/mouse-acites IgG, or polyclonal sheep IgG) to verify cell type distribution pattern and consistent staining between different antisera (detailed in Appendix IV). Panel a: true cell outlines drawn in different colors for the two cell types, and green tick-marks indicate co-localized staining. Panel b: circles and crosses indicate different cell types stained, while neuroendocrine cells were filled with blue. Scale bar is 100µm, with notation for 90µm. Color is not apparent in these black and white scans of original maps. 56 A comprehensive description of all mapping procedure details is in Appendix VIII. Below, in brief, is the general mapping procedure, as exemplified for an initial section through central PVH (about level 26 in Swanson’s Brain Maps). a.) Draw outline of third ventricle (3V), ventral edge of section (if seen), and large blood vessels. Add 100µm scale bar and external fiducial as described below. b.) Draw outline of Fast Blue-positive (FB) cell area, for an approximate outline of PVH neuroendocrine (ne) borders and underlay for subsequent data recording. c.) Plot positive cells of first peptide (circle), distinguish ne vs. non-ne by filled or non-filled circles—with blue fill indicating FB positive cells. d.) Plot positive cells of second peptide (cross), as above, using a plain cross for non-ne cells and a cross filled with blue color for ne cells. e.) Additionally, identify co-localized peptides (cross, superimposed on circle) in ne (blue color-filled) and non-ne cells. Neuroendocrine border outlines, major blood vessels and the edge of the third ventricle (3V) were used as fiducials to align maps from one section to the next, and to help determine overlap edges in montages of two 35mm fields from the photomicrographs. For each map, a 100µm scale bar was drawn on the bottom right, from the image of a micrometer reticule photographed through the microscope under the same conditions as the data images. Independent fiducial symbols (crossed lines) were traced from one map to the next at the bottom right and top left. This provided additional confirmation of dorso-ventral positioning between adjacent maps for a set of sections that were not in direct serial order. After completion, each map was re-checked on a different day: by comparison of each mapped cell to its 35mm slide, and with spot-checks of individual sections in the microscope at 100X or 200X magnification. 57 When complete, each hand drawn primary map contained seven data elements that could be displayed subsequently in different layers of a computer graphics image. Neuroendocrine borders in relation to anatomical landmarks and cytoarchitectonic PVH subdivisions comprised the base layer. In addition, two neuroendocrine cell type distributions, two corresponding non-neuroendocrine cell type distributions and (potentially) two double-labeled cell distributions, either neuroendocrine or non-neuroendocrine were recorded. For a detailed illustration of data elements incorporated in one hand drawn primary map, see figure 5.03 below, in “Conversion into Computer Graphics Files”. Up to four series (25-30 sections each) of 15µm thick sections through the entire rostro-caudal extent of PVH were thus mapped from a given animal, and two or more such series were mapped in different animals for each neuroendocrine hormone and releasing or inhibiting factor studied. Key sets of sections from three exemplar animals (PVH 10, 12 and 14) were mapped more completely in computer graphics files. Together, they provide data in high (close to serial section) resolution for systematic analysis, as discussed below and later in the Results Section. Conversion into Computer Graphics Files: Hand-drawn data maps were converted into primary graphics maps consisting of multiple layers in Adobe Illustrator™ computer graphics files. Each hand-drawn map was photocopied (reduced 64%: from 11”x17” ledger size to 8.5"x11" standard page size), scanned and imported as a template into Adobe Illustrator™. See figure 5.03, a). Medial edge of the third ventricle and neuroendocrine borders defined by Fast Blue staining were traced using the pen tool in a layer over the template. A standardized external fiducial mark (four crossed lines, about the same length as a 100µm scale bar) was added in the space of the third ventricle, positioned low enough on rostral sections so that it would still be within the 3V outline in the caudal (more ventral PVH) sections. In some cases prominent blood 58 vessels, etc. were drawn as additional alignment aids. This first drawing served as the base-layer, termed ‘neuroendocrine borders’ map (figure 5.03, b). Each set of data elements (populations of neuroendocrine and non-neuroendocrine antibody-stained cells) was then recorded in exact register in a separate overlying layer. A standard circle, whose size was equivalent to that of an average parvicellular neuron at this magnification, was entered over each hand-drawn data symbol in the template from the primary data map. Originally, the bitmap symbol for a lower case “L” in ITC Zapf Dingbats™ font was used for ease of mapping and convenient size. In later maps, this symbol was copied into a version created as a spline-curve circle for more efficient manipulation in the graphics program. For neuroendocrine (ne) cells, the symbol was plain in color, according to the cell type. In the final version, an appropriately colored circle bounded by a thin black rim indicated non-ne cells. This system is typified as filled or empty circles in the black and white versions illustrated in figure 5.03, panels e and f. Double-labeled cells were indicated on a different layer, by a combination symbol consisting of two intersecting circles of the corresponding colors (figure 5.03, panel c). A visually distinct color was used to identify each of the surveyed target cell types: CRH=sky blue, TRH=yellow, SS=gray, TH=pink, VAS=green, OXY=red and GRH=brown. Each of the data categories listed above—Fast Blue neuroendocrine cell staining, peptide 1 (ne and non-ne), peptide 2 (ne and non-ne), and co- localization of peptides 1 and 2 (in ne or non-ne cells)—were accurately copied from the template image to separate graphical layers in exact register. 59 Fig. 5.03 Example hand-drawn data map template, with subsequent computer maps. Maps from PVH12D, 4/3—one section rostral to primary data photographs ‘D’, shown earlier in figure 5.01. Panel a: original hand-drawn data map imported as a template into Adobe Illustrator™. Panel b: outlines of third ventricle and neuroendocrine borders serve as a base layer, and a standardized fiducial mark and scale bar are added. Panels e and f show composites of neuroendocrine and non- neuroendocrine data elements for each of the two peptides (CRH and OXY) stained in series D, while panel c shows a composite of neuroendocrine and non-neuroendocrine double-labeled neurons. Note these cells are also shown as single symbols in layers depicting individual cell type staining. Panel d: Nissl image with PVH outline is section PVH12A, 4/3, the first section of this 1-in-4 set—thus it is 45µm rostral to the mapped section. Compare PVH subdivisions with those inferred from Fast Blue staining in the other panels. Original maps were in color, with symbols for CRH colored sky-blue and those for OXY colored red. All finished graphics maps were carefully compared to corresponding original hand- drawn maps. As an important final check, every cell recorded in the graphics map was compared again to the original stained section in the microscope with the same 10X objective lens used for photography (100X magnification). To resolve any ambiguities, some 60 cells were also viewed again with a 20X objective lens (200X magnification). PVH subdivisions and further delineation of external borders were defined post hoc from photomontage images of adjacent Nissl-stained sections, as shown in figure 5.03, panel d. The major advantage of transferring primary hand-drawn data maps into computer graphics maps is that cell types separated into different layers can be viewed, grouped, automatically counted and visually juxtaposed at will. It is thus much easier to compare cell type distribution within PVH in one section, and from section to section in a series as seen in figure 5.04. Spatial relationships between differently stained series in the same animal are easily appreciated in side-by-side comparisons or in composite maps, as illustrated in figure 5.05, and detailed later in figure 5.07. It follows that similar comparisons can ultimately be made between different animals, an important topic discussed further below. Cumulative Mapping from a Huge Data Set, Two Approaches: For data analysis, the power of the Adobe Illustrator™ graphics program was used in two important ways to illustrate distribution patterns of stained cell types. First, detailed data from a single animal were analyzed together in two types of cumulative file (see figures 5.04 through 5.08). This allowed a more direct comparison of staining patterns for several target cell types that were identified in the same section and in closely spaced series of differently stained sections. Second, data from different animals were combined, or collated, into composite files, as shown in figures 5.09 through 5.11. Using this approach, patterns of distribution for all cell types of interest could be further compared in composite views derived from two or more animals (see figures 5.12 and 5.13). 61 Fig. 5.04 Stacked Data Maps, from a single series at three anatomical levels in PVH. Staining patterns throughout PVH can be inferred from maps of adjacent 15µm sections from a 1-in-4 series that are stacked in register. Panel a-c: Maps from one series, PVH12B—stained all at the same time with antisera to OXY (open circles) and VAS (black circles). All data layers are not shown in this example. Selected data layers displayed here are ‘stacked’ in correct anatomical position, to yield a cumulative map spanning three data sections at three different levels through PVH (recall Fig. 2.2 for PVH subdivisions). Sections are separated anatomically by 45µm (three 15µm sections comprise the three intervening, differently stained, series). Thus, stacked maps show accurate cell type distribution, but at one-fourth expected true cell density for the rostral-caudal distance spanned. Note rostral PVH appears more dorsal, and caudal levels more ventral in absolute coordinates, as indicated by the uniform vertical position of the fiducial mark. This reflects changing position of PVH in relation to the ventral surface of the brain. A mid-sagittal view diagram of the rat brain, at bottom left, shows the rostro-caudal position of Atlas Levels 22, 24 and 26 in Swanson’s Brain Maps—corresponding to the three stacked maps in panels a-c. In the first stage of cumulative mapping, layers from individual section maps were transferred (stacked, in register) into a cumulative file to contain maps for every rostral to caudal section from a double-stained series. Maps were entered into the stacked file using the dorso-ventral and medio-lateral fiducial information recorded in the original hand drawn 62 maps and the composite fiducial symbol entered in the computer graphics maps. In these cumulative files cell type position information could then be viewed in a dynamic manner (in one or more sections from rostral to caudal) by sequentially revealing and hiding individual layers or groups of layers in the stacked graphics file. For example, graphics maps of the 30 sections from animal PVH12, series B that were double-stained for Vasopressin and Oxytocin (recall photographs in figure 5.01) were combined into one file. The goal was to obtain a stack of accurately registered maps in the same relative positions that the original sections occupied in the intact animal, as illustrated in figures 5.04, and 5.05. Accurate placement of the section maps in relation to one another was maintained by reference to anatomical landmarks, external fiducial symbols copied from one map to another in the original drawings, and the standardized fiducial symbol recorded on all the computer graphics maps. This is an important consideration, because apparent dorsal- ventral position of PVH within the brain changes from most rostral to most caudal sections of the nucleus, as illustrated in figure 5.04 and addressed further in the next chapter. Viewing data in computer files of stacked layers allows all sections in a series to be displayed in approximately the relationship seen in the intact tissue during microtome sectioning. Reference fiducials used to position one section in correct relation to the next are the edge of the third ventricle, major blood vessels, and Fast Blue staining that defines the extent of neuroendocrine cells in PVH. Layers in Illustrator™ files can be displayed or printed in any desired number or grouping. This allows data elements to be viewed in any combination for comparison, and to show their distribution throughout PVH in a visually dynamic way by turning successive rostral to caudal layers on and off in the view menu. See for example figure 5.06. Also, selected layers can be easily extracted by direct copy or exported as bitmapped images such as those used for the figures above. 63 Fig. 5.05 Consecutive sections from a 1-in-4 series, viewed individually and all together. Using graphics program features to view and display data in new and synthetic ways: cell type distribution patterns over larger distances can be inferred from maps of individual sections viewed together. Panel a-c: consecutive 15µm sections from PVH12 series B, stained for OXY (open circles) and VAS (black circles). Dotted black lines: Fast Blue stained neuroendocrine borders. Dashed gray lines: PVH subdivisions from immediately adjacent Nissl sections. Section PVH12B,4/4 was shown earlier photographed in figure 5.01, panel B. Three sections (45µm) omitted between each map in this 1-in-4 series. Thus, layers stacked in correct anatomical position in a cumulative map (panel d) give a view of cell type distribution over 180µm in rostro-caudal extent, but at one-fourth true cell density. 64 Fig. 5.06 Cumulative file from PVH12 section 4/4, viewed individually and all together. Maps of differently stained serial sections from the same animal can be interleaved in a single graphics file. Data from sections shown photographed earlier in figure 5.01. Black dotted lines: Fast Blue stained neuroendocrine borders. Gray dashed lines: PVH divisions from immediately adjacent Nissl section. Panel a: 12B,4/4—stained for OXY (open circles) and VAS (black circles), PVH outlines from rostrally adjacent 12A,4/4. Panel b: 12C,4/4—VAS and CRH (gray circles). Panel c: 12D,4/4—CRH and OXY, PVH outlines from caudally adjacent 12A,4/5. Panel d: cumulative data map comprised of all three serial sections (spanning 45µm) are ‘stacked’ in correct anatomical position. Thus, composite- staining patterns can be inferred from adjacent serial sections that were stained differently. 65 Another useful feature of computer graphics maps is the ability to view equivalent cell type data layers simultaneously in two or more accurately aligned sections from the same set, to create a visual approximation of a much thicker section—a “virtual thick section”, if you will. However, because there are three 15µm sections missing between successive sections in each set from a 1-in-4 series, the viewed image is at one quarter the cell density that would be expected in an actual thick section. See figure 5.05, panel d and figure 5.07, panel a. Viewing such virtual thick sections is valuable in revealing patterns of cell type distribution as discussed further in relation to figure 5.07 below. After creating the multi-section computer map files described above, adjacent series from the same animal stacked in a similar fashion were interleaved into an even larger cumulative file. The resultant file was effectively a reconstruction at near- or actual-serial section resolution. In this way, cell type position information from all stain combinations applied to one animal could be viewed dynamically. Figure 5.06 shows the relationship between three serial sections from sequential sets with overlapping staining combinations in such an interleaved file. Note these are maps from the primary data sections (PVH12, 4/3 - sets A through D) that were previously shown photographed in figure 5.01. Thus, serial sets from the same animal (comprising in total, surveys of several cell types) can be interleaved and combined in one large Illustrator™ file. This is important, because the relationship of all variously mapped peptides to each other at highest resolution can be easily seen and compared with co-staining in the same or adjacent sets of sections. One can, for example, compare and contrast staining patterns between three or more different cell types, in two or three adjacent sections where only one cell type is stained in common between them, as shown in figures 5.06 through 5.08. 66 Fig. 5.07 Adding stacked data from successive sections: two virtual thick sections. Enlargement of panels-d from Figs 5.05 and 5.06, with Nissl PVH subdivision borders included. Use of graphics program features to view and display data in new and synthetic ways allows composite staining patterns to be inferred from a composite of sequential sections in a series, or serial sections from adjacent differently-stained series. Panel a: 3-section stack from one series (PVH12B) stained at the same time with antisera to OXY and VAS. Sections are separated anatomically by 45µm (three 15µm sections are in the intervening series). The result is a visual impression of cell type distribution over 180µm in rostro-caudal extent, but at 1/4 expected true density. Panel b: three anatomically serial sections from different sequential series, plus the PVH outlines from the immediate rostral and caudal Nissl sections. In this example, a virtual thick section is obtained showing distribution of all cells, but illustrating different cell types in sections that were not stained together. Thus one obtains higher density information—three sections, spanning 180µm—about three different cell types in relation to each other, but not stained simultaneously in every section. Put another way—in fact, the way that these experiments were designed—one can make inferences about the comparative distribution pattern of two cell types that were not stained together, but were each stained in adjacent sections with a common reference cell type. For example, in PVH12: set B=OXY+VAS, set C=VAS+CRH, and set D=CRH+OXY. In this experiment CRH staining can be inferred in set B, OXY staining can be inferred in set C and VAS staining can be inferred in set D. Further, CRH, VAS and OXY distributions can 67 be inferred for the corresponding Nissl section in set A, while PVH subdivisions seen in the Nissl section can be inferred for the closely adjacent sections B-D. This concept is exemplified in figure 5.08. Manipulating accurately registered computer graphics files of cell type distribution maps is extremely valuable, certainly for qualitative and eventually also for quantitative data analysis. This important concept can be illustrated in different ways. Stacking images of serial sections to give a composite virtual thick section is a method illustrated in Figure 5.07, panel b. Another useful approach is to create a “virtual thin section”, as illustrated in figure 5.08. Using data from one section and adding to it data showing different cell types from immediately serial adjacent sections, one can create a virtual thin section displaying reliable cell type distribution patterns for three or more cell types, all of which were stained in the same animal but not the same section. Sometimes it is helpful in discerning general patterns and reproducibility of cell type distribution to create several virtual thin sections at a critical level through PVH. For example, in creating figure 5.08 there were six possible combinations of cell type distribution from the three serial source-sections used to create the virtual thin section shown. In the second approach to viewing cumulative data from a large data set, maps from different experimental animals were interleaved in a single file to show patterns in similar or contrasting sets of stained sections. Maps were displayed together on a ‘best fit’ basis, according to their rostro-caudal position in PVH, to yield “composite maps” for comparison and integration of data from different animals, as exemplified in figures 5.09 through 5.13. This was an encouraging modification of data analysis, because it allowed staining patterns for numerous cell types to be compared at very high resolution. For the first time, a dynamic comparison of cell type distribution was thus possible between more individual cell types 68 than could possibly be stained in the same section, or even in the same animal, at high resolution. Fig. 5.08 Partial data from three serial sections, shown as different virtual thin sections. Information in computer graphics files can be combined to create a composite virtual thin section. Here data from three adjacent 15µm serial sections are selectively combined in two ways to form unique virtual 15µm sections showing information not available together in any one of the original sections. At left, the PVH subdivisions are from section A, 4/4, while distribution of VAS is from section B, 4/4, neuroendocrine outlines and CRH are from section C, 4/4 and OXY is from section D, 4/4. At right is the corresponding virtual thin section of the remaining data, the complimentary-stained peptides from each of the original three sections. Thus, distribution of the same cell type staining is shown in a different original order: OXY in section B, 4/4, VAS in section C, 4/4 and CRH in section D, 4/4. These virtual thin sections together contain the same data as the virtual 45µm thick section (figure 5.07, b) formed from viewing all data from the three sections simultaneously, but displayed at one-third effective resolution. Cell type distribution patterns are seen to be qualitatively similar, a validation of this alternative data display concept. Note: partial data from two adjacent sections could also be combined into a virtual thin section, e.g., VAS+OXY from section B with CRH from section C, or CRH+VAS from section C with OXY from section D, etc. This would show potential double staining for two of the target cell types in the same section. The result (not shown) is anatomically more accurate since it spans only 30µm to yield a virtual 15µm section, and could reveal co-localization of staining in some cells. However, the overall effect in visualizing distribution of the three target cell types is qualitatively similar. 69 Fig. 5.09 Composite file: different animals stained with same combination of antibodies. Two examples of maps from different animals—stained with the same combination of antisera—at ‘classic’ PVH, equivalent to Brain Maps Atlas Level 26. Panel a: PVH10C stained for OXY and CRH. Panel b: PVH12D stained with the identical combination of antisera to OXY and CRH. Panel c: stacked, "best fit" composite, showing reproducibility of staining patterns between animals using the same combination of antibody pairs. 70 Fig. 5.10 Composite files: “best fit” similarly stained map data from different animals. Two examples of maps from different animals—stained similarly—at ‘classic’ PVH, equivalent to Brain Maps Atlas Level 26. Panel a: one section from PVH10D series stained with antibodies to TRH and SS. Panel b: two adjacent sections from PVH14, one stained for TH with TRH, the other stained for TH with SS. Panel c: stacked, "best fit" composite showing reproducibility of staining patterns for TRH and SS between animals. 71 Fig. 5.11 Composite files: “best fit” data from different animals, one antibody in common. Two examples (a, b) of maps from different animals—2 cases, stained with one antibody (CRH) in common—at ‘classic’ PVH, equivalent to Brain Maps Atlas Level 26. Panel a: PVH12C stained for VAS and CRH. Panel b: PVH14C stained for CRH and TH. Panel c: stacked, "best fit" composite, showing reproducibility of staining patterns for a single target cell type (CRH, in this case) between animals. 72 Fig. 5.12 Composite files: “best fit” cumulative maps, data from three different animals. Data from different experimental animals at the same level through PVH are combined in a composite file to obtain an image that is equivalent to a single virtual 15µm section stained with antisera to all the cell types under investigation. Here, some data shown earlier are illustrated in a slightly different format. Circles of various colors for different cell types have been replaced with unique symbols for display in this black and white format. At left is a composite file from PVH12B,4/3 (OXY and VAS) and PVH12B,4/3 (CRH). This is equivalent to the CRH data shown in figure 5.11a. At right is a composite file from PVH10D,4/8 (TRH and SS) and PVH14A+C,6/4 (TH and GRH). This is similar to some of the data presented as circles in figure 5.10a. 73 Fig. 5.13 Composite file: one “best fit” cumulative map, data from three different animals. Enlarged view of a composite section through the same level of PVH (corresponding to level 26 in Swanson’s Brain Maps rat brain atlas) from three different animals, illustrating together all seven cell types surveyed in these studies. Color-coded circles in the original maps have been exchanged for unique geometric symbols in this black and white display. 74 The innovation in graphics file display of data discussed and illustrated in this chapter marked an exciting advance in evaluating cell type distribution because of its dynamic and cumulative quality. Previously, analysis of stained series through PVH was limited to interpretation of individual views that were then intellectually integrated into an observer’s conceptual mental models. This new manner of display allows an informative view of a rich representation of multiple cell type distribution in the same sections; and in composite or virtual sections derived from them—all in accurate register. However, a quandary immediately arose—how to communicate these findings to others in a way that didn’t require sequential presentation of dozens of printed maps, or dynamic manipulation of views displayed for others on a single computer screen? Some solutions to this quandary are addressed in the next two chapters and illustrated in the Results Section to follow. The approach described above was used to good advantage in the first round of data analysis from these studies. Patterns of distribution for all the cell types surveyed and the relationship between them were discerned in a comprehensive fashion throughout the entire extent of PVH, thus helping to fulfill a central goal of this project. This early analysis produced insights that led to further sophistication in subsequent analytic methods, as detailed in Chapter 7. One way of visualizing this data (using mock-up images similar to the figures above) was exploited in creating a prototype computer model for displaying experimental results on overlays of templates derived from Swanson’s Brain Maps (Dashti et al., 2001). Thus, published experimental results in a central computer repository could be called up via a web-based interface for dynamic display and analysis. This interesting project, an outgrowth of work on the USC portion of the federal “Human Brain Project”, is currently being further modified and developed for general use by some of the original authors and others. 75 Chapter 6. PVH 3-D Atlas Sub-Project Overview: A major challenge in these studies was how to relate a rich and complex data set— exact locations of chemically characterized cell types within PVH—to a published atlas standard such as Swanson’s Brain Maps: Structure of the Rat Brain. In early editions of Brain Maps, PVH is shown on six levels (22-27) through the hypothalamus. It became apparent after initial data analysis described above in Chapter 5 that a higher resolution atlas would be very helpful. With such an atlas as a standard reference, perhaps a new approach would emerge to assist in subsequent meta-analysis. The original Nissl-stained slide set used to produce Brain Maps was available, and it was possible to view all the serial sections spanning PVH in microscopic detail. Thus, a secondary project was undertaken in collaboration with Dr. Swanson: creation of a serial section atlas of PVH, comprised of sequential anatomical drawings from contiguous sections within the context of Brain Maps. When viewed as stacked layers in an Illustrator™ file, such an atlas could approximate a three dimensional (3-D) representation of PVH at very high resolution. “Creating a Serial Section Atlas of PVH” details the use of microscopic sections from the original Brain Maps celloidin-embedded rat brain to generate accurate drawings of cytoarchitectonically-defined subdivisions in PVH. Cumulative views of the drawings through PVH are presented to appreciate the approximate 3-D model derived. A crude 3-D wire- frame model of PVH was also produced from these drawings, using the commercial computer-assisted-design software MicroStation™. This program had been used by others to analyze data in conjunction with Atlas Level templates from Brain Maps (Georges Tocco and Michel Baudry, personal communication; Bouteiller et al., 2000). . However, preliminary results created in MicroStation™ were deemed inadequate and impractical for these studies and they were not pursued further. More important for data 76 analysis in this work, accurate overviews of the entire PVH—outline views in planes different than the original sections—were obtained as described in “Generating Sagittal and Dorsal Projection Outlines”. Use of these outline projections and their accompanying anatomical fiducial points is discussed in “Other Information from Brain Maps, Useful for Meta-Analysis”, an introduction to Chapter 7. Creating a Serial Section Atlas of PVH: At present a realistic 3-D model of PVH—a virtual PVH, if you will—is not obtainable with cytoarchitectonic accuracy. In an exploratory project with a student of Computer Science, modeling of individual cells from experimental maps inside a Virtual Reality space was explored (Ilmi Yoon and Ulrich Neumann, personal communication). This interesting approach to visualizing cell type data proved too complicated for the methods then available. Therefore, it was not pursued further. Similarly, an interesting attempt to analyze patterns in stained sections through PVH and register them to the published atlas images with computer algorithms designed for face-recognition did not prove useful for this anatomical material (Kazunori Okada and Christoph von der Malsburg, personal communication). Another 3-D approach in future may be the use of very high-resolution magnetic resonance image (MRI). Currently, research instruments achieve a resolution of 1-10mm. Thus PVH in rat, while visible (Swanson, 2004), is seen only as a 1mm spot in the basal hypothalamus. Recent predictions estimate MRI image resolution will eventually reach 10µm. Thus, cellular level analysis will be feasible. Recent interesting work in developing enhanced MRI markers for cell function in living brain shows promise (Leergaard et al., 2003), and methods equivalent to resolution and specificity of present-day light microscopy may well become available in future. Thus, it will eventually be possible to obtain a very high-resolution MRI analysis of functional compartments within PVH in an intact animal. Such a digital (voxel-based) data set would be accurate in 3-D space and free of distortions discussed earlier regarding 77 histologic data. A high-resolution, voxel-based 3-D model of PVH (easily amenable to mathematical transforms) might allow dynamic fitting of experimental data to a standard. Although such cellular resolution would be extremely valuable, MRI of adult rat brain will very likely remain comparatively low resolution (about 100µm, at best) for some time. It is a desirable long-term goal, which is, however, far beyond the scope of this work. See the chapter “Interactive brain maps and atlases” in Computing the Brain: A Guide to Neuroinformatics (Swanson, 2001) for an interesting discussion of this topic. There is however an alternative method of analysis, obtainable with current methods and materials at hand. It is the creation of a different kind of 3-D brain atlas—a registered set of serial-section anatomical drawings, in the same plane of section as Swanson’s Brain Maps but at highest possible resolution. Thus, experimental data collected at very high resolution could be more accurately interpreted and compared because of greater accuracy in mapping to a similarly high resolution standard. Serial 40µm Nissl sections from the celloidin-embedded rat brain used to produce Brain Maps were employed to produce the PVH 3-D atlas described here. Sections spanning PVH were microscopically observed at 100X on a Lieca™DMRE microscope, photographed with a Spot Jr™ digital camera and scanned into PhotoShop™. Beginning with sections used for the Atlas Level drawings of PVH, subdivision outlines were drawn, as illustrated in figure 6.01 below. They were created as PhotoShop™ paths over Nissl images on the computer screen, with immediate confirmation of cytoarchitectonic detail in the microscope. Subdivision borders were based on criteria defined by Dr. Swanson for the original Brain Maps drawings. Then, PVH outlines for sections immediately rostral, immediately caudal, and between defined PVH Atlas Levels were similarly drawn using microscopic observation, while comparing paths from adjacent sections for reference. 78 Fig. 6.01 PVH subdivision outlines from Nissls, serial sections from Atlas Level 25 to 26. PVH subdivisions (left) drawn over Nissl-section images in PhotoShop™. PVH outlines (center) are smoothed, and stacked in register (mid-right) to form a virtual 3-D representation of PVH. Light gray=parvicellular neuroendocrine compartment, dark gray=magnocellular neuroendocrine compartment, white=descending (non-neuroendocrine) functional compartment of PVH. Original Brain Maps atlas sections 271 and 276 were used to create the Atlas Level 25 and 26 drawings (right). Note the bottom panel encompasses the Nissl image and atlas vignette illustrated earlier in figure 2.02. 79 Path drawings created in PhotoShop™ were exported into Illustrator™ for manipulation in a similar fashion as data analysis figures shown previously. Serial section drawings were entered as layers into a single file, interleaved with the relevant Atlas Level drawings from Brain Maps. Subdivision outlines were color-coded to functional compartments, as shown previously in Chapter 2, with the parvicellular neuroendocrine compartment further divided into periventricular (PVHpv) and medial parvicellular (PVHmpd) parts. See figure 6.01. Neuroendocrine parvicellular and magnocellular divisions and non-neuroendocrine descending parvicellular divisions were colored differently (light gray, dark gray and white in the figures here). Edges of the PVH celloidin-section drawings were slightly smoothed to represent a visible progression more clearly, as illustrated in figure 6.02 below. The stacked-layer depiction of serial sections is not a true 3-D volume in the same sense as a voxel-based rendering would be. However, it provides an accurate and very informative view of PVH shape and position within the rat brain. Because they are from serial sections, the outline drawings stacked in register approximate a virtual 3-D representation of total PVH volume in the rat brain used for Brain Maps. The dorso-ventral position of PVH changes, relative to the ventral surface of the brain when progressing from rostral to caudal, as shown clearly in figure 6.02. Thus, one can appreciate the 3-D extent of PVH in stacked images of opaque drawings that are viewed from a constant dorso-ventral point along the z-axis. Figure 6.03 below shows an enlarged image of PVH alone, as if floating in space, to make the visual point of 3-D representation more clearly. 80 Fig. 6.02 Virtual PVH 3-D model. Serial section outlines stacked in Atlas Level context. Right side of the rat brain, as illustrated in template outlines from Swanson’s Brain Maps. Midline is at the left edge of this drawing, with outlines of the third (midline) and lateral ventricle (arching up to the right) outlines in gray. The outer surface of the brain (in black) is drawn from Atlas Levels 17-29, with the most forward outline (Level 17) including the rounded edge of optic chiasm at its ventral surface. The base of the brain (essentially the bottom of the median eminence) appears more ventral as sections progress caudally. In the bottom left portion of the image, serial section outlines of PVH are shown in context, bordered medially by their adjacent third ventricle outlines. Note size and position of PVH, on levels 21 through 27, in relation to the whole rat brain 81 Fig. 6.03 Detail: Virtual PVH 3-D model: position of three functional compartments. PVH outlines from serial sections are shown ‘floating’ but still in register, with margin of the third ventricle at the left. Functional compartments are colored as in figure 2.2: light gray=parvicellular neuroendocrine, dark gray=magnocellular neuroendocrine, white=parvicellular (non-neuroendocrine) descending projections. The stacked-section model is shown in three different views in figure 6.04 below, to further aid conceptualization of PVH structure in 3-D space. By separating each individual layer a fixed amount to the right from that before it, one obtains a lateral-offset image of PVH (top panel) that approximates an oblique view in whole brain context. The lateral or outside edge of the nucleus is emphasized, and changes from one section to the next are more apparent. One sees clearly the small size and lateral placement of the most rostral extent of PVH. It was labeled caudal parastrial nucleus (PS) in early editions of Brain Maps, a detail corrected in Brain Maps, III (2004). In addition, a vertical extension of PVH, one section behind the anterior commissure (detailed in figure 6.09) is now apparent. 82 Fig. 6.04 Three views: Virtual PVH model: assists dynamic visualization in 3-D space. 83 By far, the most notable feature revealed in this stacked-section 3-D model is the shape and extent of the lateral part of the medial parvicellular subdivision. A small, intensely stained, disconnected group of cells labeled “pml” (posterior magnocellular division, lateral part) on Level 25 of the original Brain Maps, it is revealed in the 3-D model to be a bulbous expansion of the medial parvicellular subdivision. This expansion (PVHmpdl) extends through eight sections and is connected to PVHmpd by a thin stalk spanning no more than two sections—less than 100µm. The full extent and (by virtue of continuity) clear parvicellular identity of this part of PVH is a new and interesting observation that will be discussed again later. In the lower panels of figure 6.04, size adjustments made to the original drawings reflect frozen-section proportions more applicable to comparison with experimental data (a topic detailed in figure 6.08). In the final panel of figure 6.04, opaque section drawings are stacked in reverse order—from caudal to rostral. This shows PVH as if seen from back to front, floating in space. All these views help form an accurate conceptual 3-D model of PVH, useful when analyzing experimental data. Generating Sagittal and Dorsal Projection Outlines: Creation of a realistic, high-resolution 3-D model of PVH in the context of a published atlas was a major advance. It provided a concrete form for a mental model when observing and interpreting experimental material. Such a model allows fresh ideas about manipulation and presentation of high-resolution data in way that can be integrated with existing and new information. The question then arises, how best to use this model with a rich data set comprised of sequential frontal maps in multi-layer computer graphics files? Obviously, one could transfer stacked-section data (recall figure 5.04) onto multiple new layers of the Brain Maps atlas—essentially creating a 37 level PVH atlas from one that originally contained six or seven levels. In documenting a complete 3-D record of cell type 84 distribution, this approach might be ideal. However, it would result in production of many additional templates and maps. Thus, it does not address the challenge of representing PVH cell type information in a comprehensive, yet succinct format. Therefore, a new approach was chosen, based on a novel use of the high-resolution Nissl images from Brain Maps. As illustrated in figure 6.05, top and bottom (or medial and lateral) extent of each section were plotted “as if seen on-edge”, either from the middle of the brain looking outward (mid-sagittal view) or from the top looking downward (dorsal or horizontal view). Points defined by the individual section measurements were joined in a smooth curve to form an accurate sagittal or dorsal projection outline of PVH positioned within the whole-brain outline diagrams in computer files from Brain Maps, 2 nd Ed. (’98). 85 Fig. 6.05 Measuring PVH section outlines to create sagittal and dorsal projections. Distances from midline and dorsal, ventral, medial, and lateral extent of PVH were measured in each section. Here the Nissl drawing of the source section for level 26 is positioned over its slightly idealized outline on the atlas template for that level. Stars indicate the limits of PVH on the original section. Dorsal and ventral limits are also indicated by stars on the line representing Level 26 in the atlas brain mid-sagittal diagram above. Similar points measured on stacked serial sections define a curve bounding the dorsal and ventral contour limits of PVH in sagittal projection view, as illustrated next in figures 6.06 and 6.08. Medial and lateral contour limits in each section similarly define PVH outlines in a dorsal projection (horizontal view), as shown in figure 6.07. 86 Fig. 6.06 PVH sagittal outline projection: anatomical fiducial marks and Atlas Levels. Vignette of mid-sagittal view of rat brain detailing position and shape of PVH in sagittal projection. The line defining the shape of PVH is a closed complex curve whose anchor points were determined by the position of maximum dorsal and ventral limits of PVH on each serial section as detailed above in figure 6.05. Positions of Nissl sections (251-288) are shown as a series of fine vertical lines representing those containing PVH. Note their relationship to the wider vertical lines representing levels 15-35 in Brain Maps. Thus, section 251 is that used for the anatomical drawing of level 21 and section 276 is that used for the drawing of level 26 -- shown above in figures 6.01 and 6.05 and earlier in figure 2.02. Anatomical fiducial points used to verify the position of PVH within the hypothalamus are also shown. These are important when considering the relationship of experimental material in reference to the atlas templates, a subject discussed in detail below. 87 Fig. 6.07 PVH dorsal outline projection: anatomical fiducial marks and Atlas Levels. Using distance from midline, with medial- and lateral-extent measures from each section (as detailed for section 276 in figure 6.05), a dorsal (horizontal view) outline projection of PVH is obtained. In this projection, the shape of the newly appreciated lateral extension of parvicellular PVHmpd is even more apparent. In horizontal view, midline and lateral fiducial structures are seen to be very important, as the atlas animal was actually sectioned at a slight angle (4 degrees) from true coronal bilateral symmetry. Thus lines representing Atlas Levels 15-35 are not vertical, and the thin lines representing the serial Nissl sections through PVH are similarly at a 4-degree angle. This projection shows clearly that the most rostral extent of PVH is on level 21. It is at a widened part of the third ventricle, and positioned laterally. Level 27, the most caudal Brain Maps template showing PVH, contains the reference fiducial structure AHNd. It is seen to also encompass the most lateral extent of PVH. Thus, the virtual 3-D model represented by outlines of 37 stacked sections is converted into two “flattened” images or projection outlines of PVH, seen in its complete rostral to caudal extent. These projections were used in the Results section, below to display data throughout the entire PVH in a single, comprehensive image. 88 Other Information from Brain Maps, Useful for Meta-Analysis: Several helpful computer files are included in the second edition of Brain Maps, in addition to template files for mapping of experimental data onto each Atlas Level. For example, illustrated above in figures 6.05 to 6.07 are vignettes from projection outlines in sagittal and dorsal (horizontal) view of the entire rat brain. These files show exact relative location of each Level illustrated on templates in the atlas, and location of anatomical fiducials for further reference to experimental material. Using a file modified from Brain Maps, 2 nd Ed., figure 6.08 (with detail in figure 6.09) illustrates the proportional difference between a frozen-section and celloidin-section (original atlas) view of PVH, illustrated in mid- sagittal perspective. A number of relatively small structures (nuclei, fiber tracts and gross anatomical features) were identified throughout the brain for use as fiducial reference points easily recognized in Nissl sections. They were identified in serial sections from the atlas animal and located on mid-sagittal and dorsal projections in Brain Maps as shown in figures 6.06 and 6.07. Several sets of Nissl stained, frozen-sectioned brains prepared in a way similar to the experimental material in studies presented here were surveyed to create a similar frozen-section mapping of the same fiducials (Pierre-Yves Risold, personal communication). Measurement comparisons between frozen- and celloidin-section fiducial locations yielded consistent proportional differences based on the greater shrinkage of celloidin-sections during processing compared to that in frozen-sections. “Frozen-section proportions”, shown here in sagittal view is a 126.6%Horizontal by 161.3%Vertical re-scale of the original celloidin-section atlas material. Frozen Section proportion in the dorsal (horizontal) view shown in figure 6.07, is a 126.6%Horizontal by 138.9%Vertical re-scale of the original. (Also, the bilateral fiducials in horizontal view were used to calculate the four- degree left/right variance from true frontal plane noted in the original atlas.) 89 Fig. 6.08 PVH Sagittal Projection: frozen-section proportions from celloidin-section Atlas. Proportional difference between a frozen-section and celloidin-section (original atlas) view of PVH in mid-sagittal perspective, Modified from Brain Maps, second Edition (’98). Eleven relatively small, distinct structures (nuclei, fiber tracts and gross anatomical features) were identified throughout the brain for use as fiducial reference points easily recognized in Nissl sections. Several sets of Nissl- stained frozen-sectioned brains were surveyed to create a similar mapping of the same anatomical fiducials. Comparisons between frozen- and celloidin-section intra-fiducial distances yielded consistent proportional differences. Celloidin-sections are smaller, due to differential shrinkage in cellular and fiber tract areas during dehydration and celloidin-embedding procedures. The “Frozen Section proportions” image shown here in mid-sagittal view is re-scaled (126.6% Horizontal by 161.3% Vertical) from the original celloidin-section atlas material. Thus, sections from frozen-sectioned brains span 126.6% greater length (rostro-caudal or z-axis), and 161.3% greater height (dorso-ventral or y- axis) than the original celloidin-embedded atlas brain. That difference is apparent mostly in the dorsal- ventral aspect of the figures above. A similar set of measurements and proportional adjustments was applied to the dorsal (horizontal view) projection. The same rostral-caudal (z-axis) difference shown above was obtained for horizontal measurement, while lesser medial-to-lateral shrinkage was seen. See Brain Maps for further details. PVH projections adjusted as for “Frozen Section proportions” were used in analysis of frozen-section antibody-stained data from the work presented in this work. 90 Fig. 6.09 Large PVH midsagittal projection, frozen-section size with serial atlas sections. Detail from figure 6.08, above. Outline of PVH, scaled to frozen-section proportions, with lines representing Nissl sections #231-288 from the PVH 3-D atlas. Note two fiducial points fall exactly on PVH Atlas Levels: ‘c’=caudal anterior commisure on Atlas Level 21, and ‘h’=AHNd on Atlas Level 27. Other structures as labeled in figure 6.06 and 6.08. 91 Sections from frozen-section brains span 126.6% more length (rostro-caudal direction, or z-axis), 161.3% more height (dorso-ventral direction, or y-axis), and 138.9% more width (medio-lateral direction, or x-axis) than the original atlas brain. This is consistent with the generally observed 30-35% shrinkage of celloidin-embedded material, where greater shrinkage occurs in highly cellular areas like cerebral cortex compared to the denser, heavily myelinated areas like brainstem. Similar proportional re-scaling of the frontal Atlas Level maps would take into account only height and width (x, y, coordinate) differences. Thus, a coronal 40µm celloidin-section spans about the same rostro-caudal distance, as would a 60µm frozen-section in a rat brain of equivalent size; and each section from the atlas brain is equivalent to four 15µm frozen-sections. Therefore, one stained set in the experimental material presented here (1-in-4, 15µm sections) is anatomically equivalent in rostro-caudal extent to one atlas section, an estimate verified in section counts and measurements from the data. Use of measured proportional differences to adjust data maps to celloidin-section proportions, or (more commonly) atlas templates to frozen-section proportions is a useful feature of Brain Maps computer files. The following chapters will describe and exemplify how information derived from Brain Maps and the new 3-D PVH model atlas is used in meta-analysis. 92 Chapter 7. Meta-Data Mapping Strategy Overview: A significant effort in the analysis stage of this work was devoted to determining how best to represent the enormous body of high-resolution data for easy comparison, and for integration with the published literature. When describing cell type distributions the traditional and most accurate approach is to present series of sequential frontal section maps—surveys of twenty to thirty levels throughout PVH—for each cell type. However, it is impractical to publish many dozens of maps, and conceptually difficult to interpret data in a comprehensive way merely by seeing numerous printed maps. Turning layers on and off in an Illustrator™ file to view data dynamically proved very helpful, as discussed and demonstrated in Chapter 5. Further, interleaving of data into high-resolution composite- maps from several animals proved informative, as will be shown again in later chapters. An expedient method used in early stages of data analysis used that approach. Composite- maps showing all cell types surveyed were printed, and mounted in sequence along the wall of a long hallway. Viewing maps while walking down the hallway was like “taking a walk through the PVH”. It was conceptually quite valuable, however not amenable to sharing widely with others. This exercise provided impetus for a sub-project to create a serial- section version of a standard atlas through PVH, described above in Chapter 6. Considering the huge amount of high-resolution data generated in these experiments, it was obvious that a method for meta-analysis was sorely needed. The data are primarily graphical rather than numerical, and their spatial distribution is important. Thus, an attempt to employ methods used in multivariate analysis was not fruitful (Loren Smith, personal communication). A graphical method was clearly essential for any comprehensive analysis. Such an approach was needed to better reveal over-all patterns within PVH. If combined with accurate reference to a standard published atlas, it would 93 facilitate anatomically accurate comparison of data from different animals. With the Brain Maps atlas-based 3-D model of PVH in hand, two new approaches to analyzing data were developed and two additional rounds of analysis were undertaken. In “PVH 3-D Atlas Provides New Tools for Meta-Analysis”, rationale for the initial meta-data mapping strategy that arose from insights gained while compiling the 3D PVH atlas is discussed. By matching anatomical fiducial structures in experimental Nissl sections with their equivalent landmarks in mid-sagittal and dorsal views from Swanson’s Brain Maps, the plane of section for experimental material in relation to that of the published atlas could be determined. Thus, it became possible to display cumulative data from entire sets of sections through PVH on a standardized projection, without inaccuracies introduced because of differences in plane of section between experimental series. See figures 7.01 and 7.02. Rationale for re-arranging data for display in a spatial distribution and orientation different from original plane of section (re-slicing) is discussed with figure 7.03. “Using PVH Outline Projections to Display Cumulative Data” describes a procedure illustrated in figures 7.04 and 7.05. A midline sagittal outline of PVH is used as a template for single-image display of cumulative cell type distribution from one section out of a set containing many frontal maps. Projection images created with this approach are used to display cumulative data from an entire series, as shown later in the Results Section. The second strategy used for meta-analysis is discussed in “Accurate Mapping to a Standard Atlas”. A new geometric coordinate-based approach was devised for accurate display of high-resolution single section data from experimental animals (sectioned in different planes) in reference to a published standard. Thus, relative distribution of many neuroendocrine cell types obtained from different experimental animals can be displayed together with great anatomical accuracy on standard templates from a published atlas such as Brain Maps. This is accomplished using a grid-based coordinate system based on 94 comparing planes of section in experimental animals to that of the standard atlas, as illustrated in figures 7.07 - 7.09. Resulting template-based maps of experimental data are displayed in Results chapters 8-12. Composite Atlas Level maps, showing integrated data from several animals are detailed in Chapter 13. Finally, “Dealing with Distortion in Mapping to a Standard” addresses some practical problems encountered in meta-analyses of experimental data from this work. PVH 3-D Atlas Provides New Tools for Meta-Analysis: The visual model of PVH in 3D space derived from stacking sequential serial section outlines in register was informative, both conceptually and practically (recall figures 6.02 through 6.04). An Illustrator™ file of stacked outlines of PVH within context of the whole brain is not a true 3D model, in the sense that it cannot be rotated and re-sliced in different planes, as one might with MRI voxels. Although a voxel-based model would theoretically be most useful, as a practical matter of fact, virtually all current neuroanatomical research methods at the cellular level provide detailed two-dimensional, histological-slice graphical data—not digital data composed of voxels. Comprehensive cellular resolution in a true 3-D (voxel-based) volume of PVH or other small neural structures is still in the future. When considering physiological or behavioral results in neuroscience research, large data sets are often represented numerically. Thus, they can be binned or otherwise consolidated for meta-analysis. Unfortunately, large sets of high-resolution microscopic data such as those in these studies are not very amenable to meaningful numerical consolidation. Numbers of cells and general distribution of cell types within PVH subdivisions can be obtained, and will be useful in future for comparisons to experimentally manipulated material. (See, for example, final tables showing data from these studies in Chapters 8-12.) However, crucial spatial information about the density, possible co-localization and juxtaposition of cell types within the nucleus would be lost in numerical analyses. Therefore, it is important to 95 consider ways to communicate visually, to display experimental and observational data graphically, in the context of an accepted conceptual view of in situ anatomical structures. This is both a general and specific aim of most neuroanatomical investigations like those described here. It has been a primary goal since the earliest studies of PVH, as can be seen in pictures, diagrams and artists’ renderings in the literature (Swanson et al., 1986). The serial section stack of PVH subdivision outlines from the Brain Maps atlas animal contains intrinsic geometric coordinate (x-, y-, z-axis) information. It is much more accurate than previous conceptualizations of PVH in three dimensions or in mid-sagittal or dorsal outline view (Swanson and Sawchenko, 1980; Markakis and Swanson, 1997). In the atlas-based 3-D visual model described above, extent and contour became more apparent as the limits of PVH in each section were plotted to create mid-sagittal and dorsal outline projection views (recall figures 6.01 and 6.05). Especially interesting were the changing dorso-ventral position of the nucleus in relation to the base of the brain, and the far rostral and lateral extents of anterior and medial parvicellular divisions, shown in figure 6.04. Close examination of serial sections revealed the most rostral extent of anterior parvicellular PVH to be on Atlas Level 21, one level rostral to that seen in early editions of Brain Maps. The shape of the most lateral extent of the medial parvicellular division, which can be seen contiguously in only one of the serial sections, is newly appreciated. This most lateral part of PVHmpd is shown as a small patch of separated cells labeled “pml” (posterior magnocellular division, lateral part) on level 25 of the published atlas (recall figure 2.2). In the 3D PVH model it is revealed instead to be a bulbous lateral extension of medial parvicellular neurons connected to the main part of the subdivision by a thin stalk (PVHmpdl). The almost ovoid- shaped lateral extension is also seen much farther rostral and caudal (spanning eight serial sections) than previously appreciated. This observation prompted reconsideration of the possible contiguous relationship of parvicellular neuroendocrine cells originally thought to be 96 ectopically located, lateral to PVHmpd in some experimental sections from this study. (See for example the illustration of CRH cell type distribution at Atlas Level 25 in Chapter 11). With a serial-section PVH atlas in hand, it was now possible—and very desirable—to accurately map and integrate experimental data at equally high resolution in relation to a published standard. A hypothalamic vignette of the “Interbrain Section Plane” graphics layer from Brain Maps, Computer files (2 nd Edition) was used to calculate the plane of section for experimental brains in relation to the atlas plane of section. See figure 7.01. This hypothalamic vignette is from a mid-sagittal view, re-sized from the original celloidin section atlas measurements to frozen section proportions, as illustrated earlier in figure 6.08. Thus, location of the internal fiducial structures shown can be directly compared with the same structures in experimental brains, and plane of section can be compared between brains using geometric principles. To accurately determine plane of section in relation to the atlas, target fiducial structures are identified in specific Nissl sections from a 1-in-4 set stained from each experimental animal. Recall that one 40µm celloidin section is equivalent in thickness and cell density to 60µm in original tissue volume (essentially the same as four 15µm frozen sections) because of ~30% shrinkage during celloidin processing. Therefore, each 1-in-4 frozen section series approximates the span of one 40µm celloidin-section from the Brain Maps atlas brain. The goal in comparing atlas and experimental data is to determine an accurate triangulation—using fiducials that are as far as possible from each other in dorso- ventral and medio-lateral space, yet as close as possible to PVH. In the atlas, the location of fiducial structures has been defined from serial sections through the entire brain, and the thirty-seven sections encompassing PVH are of specific interest here. Using the “Interbrain Section Plane” computer file from Brain Maps, a sample set of lines representing serial 97 frozen-sections through PVH can be aligned over the mid-sagittal atlas image, based on fiducial locations identified in the Nissl set of experimental sections. Fig. 7.01 Determine plane-of-section: experimental data referenced to Brain Maps Atlas. Data sections containing fiducial points are microscopically confirmed. Beginning with the most rostral fiducial (e.g.,“c”=caudal anterior commisure), the corresponding section line is positioned over that symbol. The line set is grouped and rotated together to position the correct line over the next- closest rostral symbol while keeping the first symbol in register with its corresponding section. Here, “c” is in section 2/1, while “a1” is farther caudal in section 2/3. The grouped lines are rotated to fit appropriate sections to their correct fiducials; therefore, the angle of cut for the data sections is not vertical in relation to the atlas. Caudal fiducial “h” was identified in section 6/4. The rotated, grouped line set is then scaled horizontally to align 6/4 to with the “h” fiducial mark in the projection outline. This determines correct rostral-caudal line spacing for this experimental set in relation to the sagittal projection of PVH. Note, for greater clarity in this size figure, a 1-in-6 frozen section series used for initial pilot experiments is illustrated. Data shown in detail in previous chapters and analyzed in later results chapters were all generated from experiments using 1-in-4 series of 15µm sections. 98 The plane of section (for both sagittal and dorsal/horizontal views) was determined for each of the three major experimental data sets, as described in figure 7.01. Fig. 7.02 Plane-of-section for three experimental brains: PVH10, PVH12, PVH14. At left the plane of section of the Brain Maps atlas animal is illustrated in frozen-section sized hypothalamic vignettes of mid-sagittal and dorsal views. PVH outline, Atlas Levels 20 through 28, and anatomical fiducials are shown, with an overlay of lines representing the serial sections (#250-288) spanning PVH in the original atlas brain. At right, are similar projections of PVH, fiducials, and Atlas Levels, with overlays of equally spaced lines representing frozen section series of experimental data sets (PVH10, PVH12, and PVH14) chosen for further analysis. In each case the plane of section in frontal (top) and horizontal (bottom) perspectives are determined by reference to anatomical fiducial points as described in figure 7.01 and shown earlier in figure 6.07. Angle of deviation from the atlas plane of section was measured by comparison to fiducials on the underlying frontal Atlas Levels and midline zero references in this graphical representation. Plane of section angles were confirmed by reference to calculated distances between fiducial points in the data sections, compared to similar measurements in the frozen-section size atlas projection vignettes. For frozen-section size, recall the proportional transformation illustrated in figure 6.08. 99 Triangulation comparison was made between anatomical fiducial points in atlas and experimental Nissl sections, to determine the experimental plane of section in dorso- ventral and medio-lateral orientations. Using the rotate and scale tools in the Adobe Illustrator™ program, a set of lines representing the experimental sections seen on edge was aligned and positioned in correct relation to the atlas projection images. The lines were grouped, to move all together; and after the first rotation, they were held horizontally constant to prevent skew when adjusting the rostro-caudal (z-axis) distance. In this way, the plane of section (angle) in relation to the atlas standard can be reliably determined for any experimental brain. Figure 7.02 illustrates the result of this procedure for the exemplar data sets whose analysis is detailed in chapters 8-13. Using PVH Outline Projections to Display Cumulative Data One result of having an accurate projection outline of PVH in atlas coordinates was the idea of displaying all data (many maps) from an experimental set in one comprehensive sagittal- or dorsal-view outline image of PVH. This would allow an anatomically accurate overview of cell type distribution in two dimensions that would not be practical to display as a complete series of individual maps. Further, because the PVH projection outline format is constant, any registered data set could be displayed for comparison no matter how its original plane of section varied from that of the atlas. Some principles of displaying data in different orientations for eventual accurate mapping to a standard and for intra-experiment comparison are addressed in figure 7.03. One approach is to average data over a larger area than the original section, to approximate the spatial distribution that might be seen in intervening missing sections of a closely spaced but incomplete set. That was the kind of approach taken in the second round of data analysis in these studies. Data in sections viewed orthogonal to the plane of section can be seen “as if on edge” to show distribution in the z-axis derived from stacked sections cut in 100 the (x, y) plane. In this type of data re-display, care must be taken to maintain data points accurately in one axis (e.g., holding the y-coordinate constant) while moving them along another axis (e.g., x-coordinate) for spatial averaging. See for example, figure 7.04. Fig. 7.03 Can “1-in-4” section data approximate serial sections in a Virtual 3D Model? Principles of data re-arrangement for display in a different plane of view—one approach to compensating for limitations of partial data without significant alteration of true anatomical position. Single section data can be randomized in one axis (e.g., x, or y) over the area of missing adjacent sections (z-axis), while holding constant their position in the other axis. Data now seen orthogonally (as if turned to view on edge) are at correct original coordinates, but in less than expected true density. Data could also be segmented (“re-slice planes”, above) and re-combined for virtual view from a different angle than originally sectioned. Implementation of this concept, with data-example figures, is addressed in detail later in this chapter. 101 Maps from each experimental series were aligned using stereotaxic coordinate (e.g., midline vertical) orientation information determined in the first round of analysis. Then, while holding their vertical position constant, copies of data symbols were moved laterally, into a defined dorso-ventral area representing 4 sections in sagittal-view. Tools for grouping and moving symbols in Illustrator™ facilitated this process. See figure 7.04. Fig. 7.04 Move data to 60µm box to approximate 4-section span in Sagittal View. Grouped data from original map aligned with atlas PVH outline: copied and moved laterally, at constant d/v position. Data randomly distributed (≈1/4 expected actual density) in a box the height of PVH at this level and 60µm wide—to approximate the rostro-caudal area of four-15µm sections as they might be seen “on edge” in a sagittal projection. See next figure. 102 In the ideal situation, data points would distribute randomly in the newly defined space, as illustrated in figure 7.03. However, in these studies the data were “randomized by eye”, from left to right inside the defining box, to minimize complete overlap of symbols and prevent any artificially-introduced concentration toward one part of the defining area. Thus, one sees the data in much the same way as in the original section, but on edge with the lateral distance collapsed into this new 2-D box. Distribution from left to right in the box approximates density, as it is a collapsed version of the original medio-lateral distribution. This procedure is straightforward, although time-consuming, using tools in the Illustrator™ program. It may be amenable to semi-automation (and true mathematical randomization) in future studies. Data from the map of a 15µm section were thus averaged into a box 60µm wide and the height of PVH in the section as it intersected the atlas PVH projection outline. This box approximates the z-axis area that might be occupied by four consecutive sections, if they were seen “on-edge”. The averaged, grouped data were then transferred, in register at the previously calculated angle and position, to a sagittal outline projection of PVH as shown in figure 7.05, below. Data from complete sets of sequential rostral-to-caudal sections were averaged in a similar manner for transfer onto single PVH sagittal projection outlines. This meta-method—projecting composite data from an entire series of sequential maps onto a single image of PVH—was used for the second major round of analysis in these studies. Using this approach, a comprehensive view of cell type distribution from an entire stained series was generated in a plane orthogonal to the original data, but at one fourth the expected true density. Display at lower density is actually an advantage, because all the cells from serial sections viewed in such manner would confound the view of apparent distribution, due to stacking and overlap of data symbols in the medio-lateral direction. In the sagittal outline view, rostral-to-caudal (z-axis) and dorsal-to-ventral (y-axis) density 103 distribution of stained cells throughout PVH is appreciated at a glance, even though the medial-to-lateral (x-axis) distribution can no longer be seen. See figure 7.05 for a single- section example, and Chapters 8-12 for complete surveys of data. Fig. 7.05 Randomized data from figure 7.04, final form on PVH Sagittal Projection. Data from map of PVH12C, 4/3: randomly distributed in a box 60µm wide, to approximate the thickness of four-15µm sections, as they would be seen in sagittal profile. Randomized data transferred from the original map, re-sized, and rotated to accurate dorso-ventral orientation on this sagittal projection of PVH. See Chapter 11 for cumulative sagittal projection maps obtained in this way from all sections of CRH in this series. 104 In principle, the same approach can be used to generate similar horizontal or dorsal-view projection maps of entire data sets. Dorsal projections were not prepared in these studies because lateral distortion in some experiments produced by third ventricle distention from colchicine treatment made medio-lateral positioning less accurate. In dorsal view, z-axis or rostral-caudal cell type distribution would again be made apparent in a single cumulative image from an entire set of sections spanning PVH. At the same time medio- lateral extent of stained cell types would be preserved on the x-axis, while dorsal-ventral distribution originally available on the y-axis would no longer be visible. For PVH, dorsal projection view is not as informative as the sagittal projection view, since the lateral borders of the nucleus are fairly well defined in the rat. Also, in experimental neuroanatomy mid- sagittal views are frequently presented, and the mental transform from typical coronal sections is more intuitive than for a dorsal view. Accurate Mapping to a Standard Atlas: Using the plane of section angles calculated in reference to the standard atlas, primary frontal maps from experimental series can be divided into grids representing their 3D position in relation to the Atlas Levels or templates, as detailed below. Ideally, an experimental set cut at the exact plane of the atlas would completely overlap a relevant Atlas Level and would therefore contain only one square or rectangle grid. By dividing sections into grids (rectangles) based on their dorso-ventral and medio-lateral angle of intersection with the atlas plane, appropriate portions of sections can be transferred accurately onto relevant Atlas Level drawings. Using this method to combine data from different animals into a composite display on standard atlas templates is illustrated in detail in Chapter 13. Use of an anatomically accurate reference grid is different from previous methods used to display experimental data. A ‘best-fit’ approximation or mental interpolation usually guides display of data from a single experimental section onto the closest Atlas Level that 105 appears to match its underlying anatomy. This approach has serious potential pitfalls, as illustrated in figure 7.06. Fig. 7.06 The “Geometry Problem” in transferring data to a standard Atlas Template. Transferring experimental data to a standard reference atlas includes inherent error if the experimental brain is not sectioned in the exact plane of the atlas. S1-S3=three standard Atlas Levels. Test section Tn, has a midpoint anatomical structure (g) in register with the midpoint anatomical structure in S2. Panel a.: at constant angle, increased distance from the anatomically accurate intersect increases error. Data from structures (y) and (b) thus map incorrectly to reference structures (c) and (m). Panel b: adjustment for difference in plane of section by dividing the experimental section data so that it maps more accurately onto three different Atlas Levels. T(n+1, or n-1) might contain data in structure (m) and (c). Thus, triangles defining the magnitude of possible error in mapping are much smaller, although the angle of error is, in fact, constant. Modified from original drawing by L.W. Swanson. 106 Using the grid-based approach described below, data is transferred to a standard reference template in an anatomically accurate, reproducible fashion. The goal is to reduce to a practical minimum, in a systematic way, the inherent problem that experimental material is seldom if ever cut in the exact plane of an atlas. Figure 7.07 shows a conceptual model of the 3-D space visualized when dividing data maps into grids for transfer to Brain Maps Atlas Levels. Fig. 7.07 Atlas Level 25 oriented in 3D Space, PVH Sagittal and Dorsal Projections. Projections of PVH outline are displayed in 3D coordinate space. PVH sagittal outline is shown (obliquely, to indicate visual perspective) in the plane of midline zero. PVH dorsal outline is shown at the base of the x-axis as a virtual shadow that might be cast by the nucleus, when viewed from the top— caudal-to-rostral in the z-axis. (The (x-, y-) coordinate plane of transparent Atlas Level 25 (dark lines) is shown at a slight angle to enhance the 3D effect. PVH at this frontal Atlas Level is shown in outline, with dashed lines marking the borders to further emphasize the measurements originally used to create the projections. Short bars indicate the point at which the Level 25 Nissl boundary intersected the point on a curve that defined each projection outline, as detailed earlier in figure 6.05. 107 Figure 7.08 shows lines representing individual sections from PVH12 as they cross Atlas Level 25 in both sagittal and dorsal projections. The angle and spacing of lines for each section was previously determined by measurement (recall figure 7.02). The Atlas Level lines are placed in their anatomically correct position in true atlas coordinates. Level 25, depicted in the 3-D model above, is thus seen to intersect three different sections of PVH12B. More than just a slight angle discrepancy, the intersection of PVH12 is actually quite oblique because of difference in both dorso-ventral and medio-lateral plane of section. Fig. 7.08 Data-sections intersect sagittal and dorsal projections at Atlas Level 25. Data from experimental sections can be divided into grids for transfer to an Atlas template, depending on the angle of section in comparison to the reference atlas. PVH12, which varies significantly from the atlas plane, illustrates the principle quite well. As in figure 7.06, experimental sections can be divided, both vertically and horizontally so that data will accurately map to the relevant Atlas Level. Transition from one section to another on the target Atlas Level is determined by the point where the line representing each data section is equidistant from the Atlas Level. Sections are coded by individually colored boxes that span the area to be mapped onto the atlas template. See next figure. The point at which each data section line crosses, and is equidistant from the Atlas Level reference line is defined. Thus, it can be segmented for transfer of data into lengths 108 that minimize error introduced by angle difference—described earlier in figure 7.06. Points at which segments are equidistant from the reference Atlas Level then define lines drawn across the data sections. Sagittal and dorsal projection views define horizontal and vertical lines respectively, to divide data section maps into grids. Figure 7.09 illustrates how grids containing data from individual sections would plot onto Atlas Level 25. Fig. 7.09 Cumulative array of segmented Data Grids (from PVH12) on Atlas Level 25. PVH outline on Atlas Level 25, as in 3-D image of figure 7.07. Superimposed are color-coded grids (rostral is B1=lightest gray, caudal is B5=darkest gray) representing divisions from sequential 1-in-4 sections of PVH12B that map accurately to level 25, illustrated in figure 7.08. Recall concepts from figures 7.03 and 7.06, and note PVH12 plane of section is oblique to that of the atlas. Thus, only a portion of each experimental section intersects a given atlas template. Data representation in standard atlas coordinates must necessarily be a mosaic derived with geometric accuracy from the original experimental sections. Some remaining grids from these sections were mapped in a similar manner onto rostral and caudal adjacent atlas serial sections as if they were templates in a virtual 3-D PVH atlas. Using principles presented here, a semi-automated procedure can be developed to complete such a 3-D data display. 109 Note that portions of at least two, and possibly four sections are truly at Level 25. Thus if any one of them were estimated to be at Level 25, it would be an inaccurate representation of data. That is perhaps acceptable in general, for spot-samples of data. However, it is not adequate for the high-resolution data and 3-D PVH model presented here. The procedures described thus facilitate anatomically accurate, quantitative comparison of data to a standard published reference, and between experimental animals sectioned in different planes. Use of this procedure for mapping to a standard atlas comprised the third major phase of data analysis in these studies, illustrated in Chapters 8-12. It proved to be even more important than other methods, because it allowed anatomically accurate presentation of data from different animals to show relationship of different cell types in PVH. This will be shown in detail in Chapter 13. Dealing with Distortion in Mapping to a Standard: In the results presented here, accurate maps are illustrated only for the published Atlas Levels in PVH. Details of data map segmentation used for data transfer from individual exemplar experiments are shown in Chapters 8-12. In the case of PVH10 and PVH14 the frontal plane of section was not very different from the atlas, and transfer of data was fairly straightforward. PVH14 had a large difference (+9.1 degrees) in medio-lateral section angle, as revealed in the dorsal PVH outline projection shown in figure 7.06. However, TH data mapped from that animal were mostly along the third ventricle—close enough to the midline atlas-intersect so that very little potential error was introduced in transferring data. PVH10 had a very distended ventricle from colchicine treatment and thus a slightly modified approach to transfer of data from grids was used, as discussed below. PVH12, in which the frontal plane of section was most different from the atlas, presented more of a challenge for accurate mapping. As illustrated above, data from two or more different sections were 110 mapped to each Atlas Level. Data from PVH12 were transferred first to Atlas Level templates and led to some important problem-solving exercises described below. PVH12 provided a rigorous testing of the grid-based data transfer model, because of its fairly large variation from the atlas plane of section. Thus, successful and accurate portrayal of PVH12 data on Atlas Levels proved the validity of the method. PVH 10: From earlier discussion (Table 5.01), it was observed that there are several intrinsic sources of error to consider in interpreting data from histological material. Among them are lack of absolute fiducials (e.g., bony landmarks or changes in tissue-type) in brain, plane of section variations, and linear or non-linear tissue distortion due to surgical interventions or tissue processing methods. Solutions to the problem of finding appropriate fiducials in the brain were addressed in figures 6.06 and 6.07. Corrections for plane of section error have been discussed in detail above. Non-linear distortion is more difficult to address, and probably is not correctable in a systematic way. Some such distortion is accepted as a by-product of tissue processing and must ultimately be compensated for by addressing a research question using a variety of methods in comparison with each other. As in the practice of clinical histopathology, certain histologic distortions are accepted as the norm for processed tissue, and investigators adjust their interpretations to accommodate expected histologic artifact. Compensation for linear distortion was achieved in two different ways in the studies presented here. In experiment PVH14, measurements in height and width of PVH at Atlas Level intersect revealed a dorso-ventral linear stretching. Before data maps were transferred to Atlas Level templates, their height was proportionately adjusted using the scale tool in Illustrator™. Thus, at the appropriate angle and Atlas intersect, the top, bottom and lateral extent of PVH in the experimental set matched the same proportions in the atlas. In experiment PVH10, as mentioned above, there was a relatively linear distortion introduced 111 by a distended third ventricle due to colchicine treatment. The third ventricle border was not vertical and the ventricular space narrow, as in the atlas. Instead, the ventricle was ovoid in cross section, with the ventricular margin curved rather than straight. Thus, the edge of the ventricle and PVH at the midpoint of the section was displaced laterally compared to its position near the top and bottom (dorsal and ventral extent) of the nucleus. This is a problem that will probably be obviated in future experiments because staining- method sensitivity has increased to a point where colchicine treatment is seldom if ever necessary. However, PVH10 data (and pre-existing data in the literature) are quite important for their anatomical information content. It was thus important to devise a method to adjust for linear distortion caused by a curved ventricle. The approach taken was a slight modification of the grid method detailed above. After measuring to assure minimal distortion in the size and shape of PVH in the experimental sections, the mapped data were divided into a series of lateral strips within each defined grid. Data were un-grouped and the select- object tool in Illustrator™ was used to select a strip (one or two cells in height) of data symbols for movement laterally. The segment of third ventricle contained in the strip was aligned with the ventricle outline in an underlying atlas reference image. After all the strips (varying from one to four per grid, depending on the amount of curvature) were moved, the position and proportions were again checked to assure that no additional distortion was introduced. Finally, data included in the originally defined grids was transferred to the Atlas Level template. Accuracy of this approach was tested at key levels by plotting similar cell type data from a different experiment showing less ventricle distortion. Approximate overlap or agreement in general outline and density of distribution within PVH was regarded as validation of the adjustment technique. PVH12 presented an interesting test of the grid-based data transfer strategy. Data were transferred onto Atlas Levels from two or more sections as described above and 112 illustrated in figures 7.07 through 7.09. In order to test the procedure, remaining data was also transferred to one of the appropriate virtual 3-D serial-atlas outlines on either side of Level 25. This assured that all data from the crucial grids would eventually transfer to a section in the atlas plane without loss of any data points if complete 3-D volume transfer were attempted. Thus, the strategy used here does not result in accidental “lost data” from the process of dividing original data maps into grids. Another interesting potential problem in transferring PVH12 data is illustrated in figure 7.10. Because there are 45µm in distance (three 15µm sections) between each section in a stained set, there is a “step” or blank space of missing tissue between adjacent sections in a stained set. When section maps are divided into grids, the edges of a grid from one section do not exactly match the edges of a similar grid from the succeeding section. Though real, this difference is hardly discernable in practice. In PVH12, which was used to plot OXY and VAS data in the magnocellular neuroendocrine compartment, there appeared a “notch” of apparently missing data at the lateral edge of PVHpmm on Level 26. In considering this perplexing phenomenon, it was realized that the notch was real. It was a result of data missing from the spatial distribution of similar OXY and VAS cells that were located on the intervening serial sections between those of the data set being mapped. Thus, the mapped grids were not in error. The outside edge of PVHpmm on Level 26 is strongly stained, quite distinct and is very sharply curving in three dimensions—it is essentially a spherical ball of cells with defined border. Figure 7.10 illustrates the physical principles that produced the apparent discrepancy in data mapped to PVHpmm on Level 26. 113 Fig. 7.10 Scaling-factor differences change rapidly on highly curved surfaces of PVH. Distortion and apparent error might be introduced when mapping experimental data onto a standard atlas template in an area of high edge-curvature. Atlas Levels 25 and 26 in sagittal projection. Lines a-d, and w-z represent two sets of experimental sections with data that map onto atlas templates—at the same plane of section in this example. PVH boundaries are about parallel and proportional, but not overlapping. Thus, experimental data are scaled to fit the atlas template as described in the text. Distance to the edge of PVH can change rapidly; even within the thickness of a single section, as indicated by different-size triangles at the top of section lines w-z. Therefore, data from adjacent sections might appear to have a ‘notch’ or ‘step’ in distribution at the edge of a highly curved surface, an artifact that is more extreme if the curve is in three dimensions. When mapped data are taken from several 1-in-4 sections, as in figure 7.09, such an apparent distortion is amplified, and data grids must be individually scaled as described in the text. Discovering and correcting for this unusual source of error was an interesting exercise in “three-dimensional thinking”, and a good test of the grid-transfer meta-analysis method. However, for the sake of generalized representation, the data finally mapped onto Atlas Level 26 from PVH12 (shown in the Results chapters) were adjusted slightly by hand to reflect known well-documented cell type distribution and minimize the apparent (not real) gap in the experimental data surveyed. 114 SECTION III: RESULTS Overview: Results chapters 8-12 (describing data for individual cell types surveyed) each begin with a brief background introduction about the cell type in question, followed by figures illustrating the data obtained in these studies. The chapters end with a table of findings on distribution of the cell type in PVH from earlier papers, followed by a table of cell counts, by PVH subdivision at defined Atlas Levels from these experiments. Figures include a sample photomicrograph of staining for the target cell type where appropriate, followed by sagittal projection views of cell type distribution: neuroendocrine, non-neuroendocrine, and total cell type staining throughout PVH. Data illustrated in these figures are from a single exemplar animal, though the number of animals and series stained for each cell type illustrated will be noted in the text, with additional details in appendices. Following the comprehensive parasagittal views of cell type staining throughout the entire PVH, individual maps of data plotted onto standard Atlas Level templates from Swanson’s Brain Maps are presented. A schematic figure showing PVH subdivisions at each Brain Maps Atlas Level (reprise of figure 2.03) is included for reference to PVH subdivisions in individual templates. Staining at each Brain Maps Atlas Level is illustrated for section(s) from the same 1-in-4 set illustrated in the preceding sagittal projection figures. Thus, Atlas Level figures comprise essentially a frontal slice, in Brain Maps plane of section, through the data shown previously in cumulative sagittal projection figures. As detailed in Chapter 6, staining illustrated at a given Atlas Level is effectively that for a 15µm section taken through PVH at that level. Thus, over-all cell density is less than one might expect in similar maps of data onto Atlas Level templates seen elsewhere, or in a comparative section of similar thickness (40-60µm) to the Atlas reference Nissl section. 115 Following initial data presentation on Atlas Levels, the source-section(s) for each level displayed are shown in smaller composite figures, along with other intervening sections in the series used for final analysis at that Atlas Level. These figures differ slightly in form for each cell type, according to the salient features of the individual experiment. This gives a better comprehensive impression of cell type distribution between the defined Atlas Levels in the coronal plane. The number of labeled cells seen in each PVH subdivision is noted only for the Atlas Levels presented—detailed in a table of cell counts by subdivision at the end of each results chapter. Numbers of labeled cells found on all primary data maps throughout PVH are available. See Appendix X. Thus, for example, an estimate of total number of cells in PVH could be inferred for each phenotype surveyed, based on averages of cells in individual experimental sets. 116 Chapter 8. Somatostatin and Growth Hormone Releasing Hormone Overview -- Somatostatin: Somatostatin, a tetradecapeptide first isolated in 1973 from ovine hypothalamus, was named for its ability to inhibit growth hormone (somatotropin) secretion in a bioassay of cultured pituitary cells (Brazeau et al., 1973). Later studies showed that it also could inhibit release of thyroid stimulating hormone (TSH) from thyrotropes in the pituitary (Knigge et al., 1978; Kasting et al., 1981; Reichlin, 1983). Somatostatin-14 (SS) was soon synthesized (Ling et al., 1973), and effective antibodies were produced for physiological and anatomical studies (Parsons et al., 1976). In brain, the highest concentrations of SS were immediately seen in median eminence (Brownstein et al., 1975) by radioimmunoassay. This emphasized the importance of SS as a neurohypophyseal hormone (Kawano et al., 1988), although cells were later identified both inside and outside rat and human hypothalamus using immunohistochemical methods (Elde and Parsons, 1975; Hokfelt et al., 1975; Alpert et al., 1976; Desy and Peltier, 1977). By 1978, several mapping studies of somatostatin-containing cells in rat hypothalamus were available, as detailed in Table 8.01 (Brownstein and Palkovits, 1976; Hoffman and Hayes, 1979; Dierickx and Vandesande, 1979; Merchenthaler et al., 1989; Shiosaka, 1992; Markakis and Swanson, 1997). The majority of somatostatin cells, assumed to be neuroendocrine in function, were described in medial (mostly periventricular) PVH as well as in magnocellular parts of the nucleus. In 1979 Dierickx and Vandesande (cited above) published a detailed, well-controlled immunohistochemical study of cells and fibers identified with antibodies to SS in rat hypothalamus. From paraffin- embedded sections cut in parasagittal plane, they demonstrated a “main somatostatin cell mass, the location of which corresponds to that of the parvocellular part of the paraventricular nucleus”. They noted that the part of PVH described merged imperceptibly with the periventricular nucleus at that (parasagittal) level. PVH in the study was assumed to 117 represent neuroendocrine hypothalamus, though there was no retrograde confirmation that the SS cells in PVH were in fact neuroendocrine. Included in the study was an excellent series of absorption and dilution tests to control for SS antiserum specificity and cross- reactivity with neurophysin-like proteins. An interesting feature of this work was double staining for both antigens in the same section using different chromagens to differentiate them one from another. Neurophysin, a precursor to oxytocin and vasopressin was examined along with SS done to prevent a false-positive interpretation of OXY or VAS co- localization seen when using high concentrations of primary antiserum raised against the cyclic somatostatin peptide molecule. In transverse sections, separate populations of medial SS cells and lateral magnocellular (neurophysin-labeled) cells could be seen. Their work failed to reveal somatostatin staining in magnocellular PVH and SON that had been reported by others, and they generously concluded that the earlier studies might represent special physiological circumstances where magnocellular neurons express somatostatin. Subsequently, active somatostatin (SS-14) was shown to be a cleavage product of the somatostatin-28 long N-terminal form, while both forms were processed from a much larger precursor molecule (Lechan et al., 1983; Swanson, 1986). Thus, numerous antisera with slightly different specificity could be generated using various antigenic proteins. It was definitively shown that somatostatin cells are present in non-neuroendocrine parts of hypothalamus, and other areas of the brain (Bennett-Clarke et al., 1980; Finley et al., 1981; Johansson et al., 1984; Reichlin, 1983), as well as in many other body tissues (Patel and Reichlin, 1978). Somatostatin in the brain appears to also have transmitter or neuromodulator function and to serve as a neurotrophic or neurogenic factor during development (Leroux et al., 1992; Daikoku et al., 1983), interesting topics that will not be discussed further. See Table 8.01 for a synopsis of literature showing SS distribution. In the work presented here, antisera to somatostatin-14 (SS) were used to stain more than four 118 sets of sections from several animals. Both monoclonal and polyclonal SS antisera were employed in different combinations with antisera to other peptides, to detect possible peptide co-localization. Also, the different SS antisera were tested in combination with each other to assure completeness of SS cell type label as detailed in Appendix VI. There was only slight variability in staining patterns observed between the different sets of sections and different animals, a topic discussed in Chapter 14. Fig. 8.01 Somatostatin cell type staining: example photomicrographs from original data. Two images of the same microscopic field, taken with different fluorescence filters to illustrate all neuroendocrine cells (blue in the original color image) and Somatostatin-stained cells (FITC green in the original color image). This section is about mid-way between those whose data maps are illustrated later for Atlas Levels 25 and 26. See figure 8.13 for comparison to data maps. Figure 8.01 above shows a photomicrograph of SS staining in a single one-in-four series (set D) of sections from an exemplar animal (PVH10). It illustrates typical PVH SS cell type appearance and distribution pattern observed in these studies. This chapter includes a description of GRH staining in PVH from the literature in Table 8.02 and 119 concludes with Table 8.03; a survey of SS cells counted in PVH subdivisions shown at each Brain Maps Atlas Level illustrated. Overview -- Growth Hormone Releasing Hormone: Distribution of growth hormone releasing hormone (GRH) in PVH is discussed in this chapter for two reasons. First, its function is opposite to, and intimately related with that of somatostatin and crucial for the typical pulsatile release of growth hormone observed in studies of pituitary function (Sawchenko and Swanson, 1990). Second, data obtained in these studies revealed minimal GRH cell type distribution in PVH, and therefore did not warrant presentation in a separate chapter. By the mid-1970s it was increasingly clear Geoffrey Harris’ early prediction would prove true: hypothalamic factors released via the portal vasculature control anterior pituitary function (Harris, 1948). Extracts of hypothalamus were well known to induce release of growth hormone and other pituitary products (Deuben and Meites, 1964; Schally et al., 1966). Thus, it was widely assumed there were unique specific releasing factors, and possibly inhibiting factors, associated with each major anterior pituitary hormone; and discovery of somatostatin prompted an immediate search for an opposing factor that might stimulate growth hormone release from the anterior pituitary. Growth hormone releasing hormone, identified a full decade after somatostatin, was the last and most difficult of hypothalamic pituitary hormone releasing factors to isolate. In retrospect, this is perhaps not surprising since the protein is localized in a restricted area of hypothalamus in a relatively small number of cells compared to other releasing factors. GRH was first isolated from human pancreatic tumors that induced acromegaly—abnormal adult growth in bone and other tissues. Acromegaly is usually seen in patients with pituitary tumors that over-produce growth hormone. When these tumors were found not to contain growth hormone but some other factor that stimulated its production, a race ensued to 120 isolate and identify the factor (Rivier et al., 1982; Spiess et al., 1983). This last chapter in the history of hypophysiotropic hormone identification is interesting in that it also involved Roger Guilleman, one of the two major researchers in the initial race to identify TRH—the first hypothalamic-stimulating hormone characterized (The Salk Institute, 1985). Two nearly identical 40 and 44 amino acid peptide forms, both with growth hormone releasing properties, were isolated at the Salk Institute: first by Wylie Vale and co-workers and shortly thereafter by Guilleman and his collaborators (Spiess et al., 1982; Bloch et al., 1983). They were shown by immunohistochemical methods to be localized in cells of the hypothalamus, and upon sequencing proved to be encoded by the same gene (Bloch et al, 1983; Sawchenko and Swanson, 1990). Human GRH is similar to that isolated from several other mammalian species (cow, pig, goat, sheep), but differs in almost a third of its amino acids from rat GRH (Mayo et al., 1986). Rat GRH is non-amidated and contains 43 amino acids, 14 of which are divergent from the sequence of human GRH (Mayo et al., 1985; Mayo et al., 1985). In rat GRH cells are seen only in the hypothalamus, mostly in the arcuate nucleus (ARH), where they give rise to a clear axonal pathway to the neurohemal zone of the median eminence. Early confusion about somewhat wider GRH cell distribution in rat brain was resolved with antisera specific to genuine rat GRH (Vale et al., 1986). In a detailed mapping study, Paul Sawchenko described a group of GRH cells in ARH, extending forward to encapsulate the ventromedial nucleus and “through the anterior periventricular nucleus to ventral parts of the parvicellular division of the PVH” (Sawchenko et al., 1985), although they were not tested to differentiate neuroendocrine character. This distribution of GRH immunoreactive cells was confirmed by others and by complementary in situ hybridization for GRH mRNA in cell bodies (Sawchenko and Swanson, 1990). However, it was noted that cells in PVH among other areas were “difficult to label immuno-histochemically” (Bruhn et al., 1985; and Paul Sawchenko, personal communication). Indeed, a study co-authored by 121 Sawchenko (Meister et al., 1986) used the same authentic ratGRH antiserum as the ’85 Sawchenko paper, but failed to show GRH cells in PVH. See Table 8.02 for a synopsis of literature showing GRH distribution (Guillemin et al., 1982; Rivier et al., 1982; Spiess et al., 1982; Bloch et al., 1983; Spiess et al., 1983; Bruhn et al., 1985; Sawchenko et al., 1985; Mayo et al., 1985; Meister et al., 1986; Vale et al., 1986; Sawchenko and Swanson, 1990; Rodier et al., 1990). In the work presented here, numerous series from more than five different animals were stained with well-characterized antisera to rat GRH. In addition, two animals were hybridized through PVH with RNA probes complimentary to rat GRH mRNA. GRH cells were seen, as expected, in ARH and extending forward, ventral to VMH. However, GRH cells were almost never seen in PVH. Those few illustrated below in figures 8.04 and 8.09 are a composite, representing consensus distribution from four cases stained with antisera to rat GRH. Thus, no table is presented for GRH cell type distribution seen in these studies. Somatostatin (and GRH): ne, and non-ne cell type distribution in sagittal view: Sagittal projection views below show distribution of somatostatin cell type staining in PVH, determined from a single 1-in-4, 15µm series of sections from animal PVH10. They are plotted onto a projection outline of PVH as seen from midline sagittal surface looking toward the lateral (right temporal) surface (described in Chapter 7). The inset box at top right in the figures shows the experimental plane of section in relation to the Brain Maps atlas plane as true vertical. For SS data derived from experiment PVH10 the experimental plane is only slightly different from that of the atlas. For the single sagittal figure illustrating GRH data derived from several experiments (figure 8.04), the plane of section is approximated as identical to the atlas. 122 Fig. 8.02 Somatostatin cell type distribution throughout PVH, in parasagittal view. Complete 1-in-4, 15µm frontal series from PVH10. Neuroendocrine and non-ne data displayed separately with dorso-ventral position held constant, “as if seen on edge” as described in chapter 7. The entire dorso-ventral and rostro-caudal extent and density of SS cell type distribution throughout PVH can be appreciated in these sagittal projections, although medio-lateral distribution is no longer evident. Comparative distribution of neSS versus non-neSS can be seen clearly in these views. In sagittal projection view, the greatest density of SS cells is seen dorsally and ventrally in central PVH, with moderate density in mid-rostral and lowest density in caudal PVH. Few if any SS cells are in anterior PVH, rostral to Atlas Level 22. Non-neuroendocrine SS cells are seen throughout PVH, notably less dense in areas of highest concentration of neSS in central PVH. More non-ne than neSS cells are seen in caudal PVH, as expected for the mostly non-neuroendocrine, descending subdivisions. 123 Fig. 8.03 Somatostatin cell type distribution throughout PVH, all peptide-positive cells. Non-neSS cells are intermixed with neSS cells in central (endocrine) PVH. Comparatively more non- neSS cells are in caudal PVH, consistent with presence in the descending functional compartment. 124 Fig. 8.04 Growth Hormone Releasing Hormone distribution throughout PVH. Composite view, locations of the few GRH cells identified in PVH from four different cases. Most were non-neuroendocrine and one of the two neuroendocrine cells shown (the one at about Atlas Level 25) was very ventral, probably not actually inside the true borders of PVH. For reference the inset shows vertical lines representing location of Atlas Levels 21-28 that encompass PVH. Thus, indicating the relative vertical (frontal or coronal) plane of section for the Atlas. 125 Fig. 8.05 PVH subdivisions at Brain Maps Atlas Levels, reference in following figures. 126 Somatostatin (and GRH): ne, and non-ne distribution on Brain Maps Atlas Levels: The following figures illustrate somatostatin cell type distribution from the same series of sections shown above in sagittal view. Here they are illustrated on templates from Swanson’s Brain Maps rat brain atlas that encompass PVH. These Atlas Level views represent data from a single experimental level (in essence, a 15µm section) through PVH, and thus are not as comprehensive as sagittal projection views. However, they reveal medial-to-lateral distribution and cell density at a given Atlas Level, displayed in the standard frontal or coronal orientation familiar to most observers. Fig. 8.06 Somatostatin cell type distribution in rostral PVH, on Atlas Level 22. Four of five ne cells in PVHpv are quite ventral, while two others are in central PVHap near the medial corticohypothalamic tract (mct). There is one neSS cell in PVHam, an area typified by the presence of magnocellular neurons. This is the single neuron seen to co-express SS and TRH at this level, a topic that will be addressed in Chapter 14. 127 Coronal Atlas Level views are essential for evaluating cell distribution patterns in PVH subdivisions, especially in reference to functionally distinct compartments: parvicellular neuroendocrine, magnocellular neuroendocrine, and non-neuroendocrine descending subdivisions that subserve autonomic function and behavioral integration. Composite figures of Atlas Levels plus intervening data maps from the series will follow the individual Atlas Level depictions. Thus, one can gain a cumulative, relatively dynamic appreciation of subtle changes in distribution of SS cells throughout the nucleus between the Atlas Levels. Fig. 8.07 Somatostatin cell type distribution in rostral PVH, on Atlas Levels 23 and 24. At Level 23 about a third of neSS cells are in PVHpv, with most of those located in the dorsal two thirds of the subdivision. More than half of medial PVHap neSS cells are close to the dorsal periventricular zone, while eight more are distributed in a medial to lateral diagonal band across the center of PVHap. A single non-neSS cell is seen in ventral PVHap. At Level 24, in contrast, over half of total neSS cells are in the ventral two thirds of the periventricular zone, overlapping the medial magnocellular area. The remaining neSS cells (and three of only four total non-neSS cells) are located in PVHap. Almost 80% of those are in the mid-dorsal part, extending diagonally from medial to lateral, while the remaining four more ventral neSS cells in PVHap are located close to PVHpv. 128 Fig. 8.08 Somatostatin cell type distribution in central PVH, on Atlas Level 25. The highest density of neSS cells is seen at this level, with density in PVHpv slightly higher dorsally and ventrally than in the middle portion. Distribution in PVHmpd is less dense, and in the same medio- lateral diagonal orientation noticed earlier in PVHap on Level 24. 129 Fig. 8.09 Somatostatin cell type distribution in central PVH, on Atlas Level 26. At this level neSS cells are seen only in PVHpv, at lower density than more rostral Levels. 130 Fig. 8.10 Somatostatin cell type distribution in central PVH, on Atlas Level 27. On this caudal level of PVH, SS cell type distribution appears at first glance to be slightly denser than on level 26. But this is an illusion because the eight cells in PVHpv are grouped into a smaller area than the twelve cells in PVHpv on level 26. Two other neSS cells are seen in medial PVHmpd. One non-neSS cell is seen in ventral PVHpv, and ventral PVHmpd, while a third is located at the dorsal edge of medial PVHlp. The slightly increased distribution of non-neSS one might expect at this level in the descending PVHlp subdivision is not evident, perhaps due to chance in the sample taken. 131 Fig. 8.11 Growth Hormone Releasing Hormone distribution in PVH, on Atlas Level 27. Growth Hormone Releasing Hormone distribution in PVH, is illustrated only on Atlas Level 27. The very few GRH cells seen in several animals surveyed in these experiments were very sparsely distributed at approximately this caudal level in PVH. This is a representation of one 15µm section to illustrate a composite view of the actual data observed in four cases. One neSS cell was seen ventrally in medial PVHmpd close to PVHpv, and four non-neSS cells are distributed both dorsally and ventrally in the middle area of PVHmpd. Following in figures 8.12 through 8.17 are illustrations of all PVH data sections as originally mapped for somatostatin, including the one projected onto each Atlas Level shown above. Since the Atlas Level illustrations are widely spaced and represent only data from a 15µm section, these figures are provided to gain a better appreciation for the distribution of SS in the standard coronal plane of section. Thus, one can correlate medio-lateral distribution of SS throughout PVH for comparison with the sagittal projection views shown earlier in figures 8.02 and 8.03. The experimental data set was sectioned in a plane very close to that of the atlas, and data were transferred from original maps to Atlas Level 132 illustrations with a minimum of manipulation. For meta-analysis (accurate transfer of data from experimental maps to Atlas Level drawing), the data maps were adjusted in dorso- ventral height and the third ventricle was better aligned with the atlas as discussed in Chapter 07. Note that the Nissl outline mapped from the adjacent section for each data map appears shrunken in comparison to the data and margins of PVH mapped from the immuno- stained section, though the image was adjusted in size for best fit to the data section. This is the result of differences in tissue shrinkage between immuno-stained and Nissl stained material. Data transfer to Atlas Levels was guided primarily by PVH boundaries in the data section: these were determined from dark field-view outlines of all cells and fiber tracts and ultraviolet view of all ne cells retrogradely labeled by Fast Blue from the bloodstream. 133 Fig. 8.12 Somatostatin cell type distribution in PVH, from Atlas Level 22 & caudally. 134 Fig. 8.13 Somatostatin cell type distribution in PVH, from Atlas Level 23 & caudally. 135 Fig. 8.14 Somatostatin cell type distribution in PVH, from Atlas Level 24 & caudally. 136 Fig. 8.15 Somatostatin cell type distribution in PVH, from Atlas Level 25 & caudally. 137 Fig. 8.16 Somatostatin cell type distribution in PVH, from Atlas Level 26 & caudally. 138 Fig. 8.17 Somatostatin cell type distribution in PVH, from Atlas Level 27 & caudally. 139 Below are tables containing tabulated data from significant references to SS cell type distribution in hypothalamus. They comprise an historical account of the known distribution of somatostatin in PVH (Table 8.01), and a reprise of Atlas Level data presented here, but in tabular format (Table 8.03). Taken together, the figures above and tables below illustrate the greater accuracy and comprehensiveness of this study of SS distribution compared to preceding ones. An obvious finding is the intermixing and wider distribution of neuroendocrine compared to non- neuroendocrine SS cells within PVH. Even in cases where neuroendocrine versus non- neuroendocrine SS cells have previously been inferred (Merchenthaler et al., 1984) or documented (Markakis and Swanson, 1997); the total distribution of SS cells within PVH has been underestimated. As with most accepted descriptions of SS distribution in PVH (cited above), the highest concentration of SS cells is seen in the periventricular zone. Rather than a uniform distribution within this subdivision, however, data presented here reveal some differences in density of SS cells from dorsal to ventral and rostral to caudal. This variation has apparently not been noted elsewhere. 140 Table of referenced papers showing SS cells present in PVH Reference; general info Location of data described in PVH Other Brownstein & Palkovits, ‘76 Front. Neuroendocrinol. vol 4 ‘PVN’ 4.4+/-1.8ng protein (they say ARH!) - p.13 Apparently PVHpv 23.7+/-9 ng ‘m. eminence.’ 309.1+/-60.8ng Punch samples: RIA data given in ng protein per punch; sagittal-diagonal orientation diagrams. Early reference for use of antibodies to identify and anatomically localize SS in brain tissue. No cytological information on precise location, or whether cell bodies or processes were identified. Hoffman & Hayes ‘79 J Comp Neurol 186: (3) 371-91. PVHpv(?) Many “+” cells 20µm lateral to midline PVHmpd(?) labeled “PA” 10cells 240µm lateral to midline Fibers projecting to m.e.- 4 sagittal drawings Nice study of SS cells and fibers visualized in Dog diencephalon. Parasagittal oriented sections: 1 PVH photo: SS+ cells & large SS- cells. SS confirmed (-) magnos addresses earlier controversy. Dierickx & Vandesande ‘79 Cell Tissue Research 201:(3) 349-59. PVHpv 13cells/Lt 8cells/Rt PVHap sagittal Photo 172cells/Rt PVHpml? 0 cells/Lt, 4cells/Rt PVHmpd 9 cells/Lt, 22cells/Rt Good survey. Paraffin sections, 5µm; SS IHC/DAB. Sagittal map & photo, + coronal photos of PVHmpd. Sagittal photo looks like PVHap. Cell numbers above approximated from photos in article. Merchenthaler, et al. ‘89 Endocrinol.125:(6) 2812-21 70% SS are ne, mixed w/ non-ne PVHpv? 4/8 on diagram =>51/102 of those counted? ap? 4/8 on diagram =>51/102 of those counted? mpd? 50(?) of 101 counts @this level Study comparing hypophysial SS with leutinizing releasing hormone. WGA (wheat germ agglutinin lectin, a retrograde tracer) was injected into external m.e. to label ne cells; ne population labeled is restricted by size of injection site. LRH&SS IHC, colchicine, 30-35µm vibratome sections. No defined pv and ap or mpd; lack of complete diagrams confuses count results. A few MePO neSS. Shiosaka ‘92 Handbook of Chemical Neuroanatomy, Vol.10. Somatostatin, Ch. XII, p.369-98. Very nice chapter on SS expression during development -- which is different than expression in adult. Contains only one excellent photo of adult rat pv - p.387; no good maps or cell-counts. Markakis & Swanson ‘97 Brain Research Reviews 24: (2-3) 255-91 ne non pv 47 9 ap 29 18 am/mm 6 / n.a. 2 / n.a. pmm 2 1 pml 0 0 dp 1 0 mpd 9 5 mpv 0 0 lp 0 0 Best over-all comparison to the work described here; ne cells marked w/ Fast Blue from blood, one antibody stain for SS, second antibody for BrdU cell birthdates. No mm cells plotted distinct from ap or pv. No cells plotted in mpdl (lateral part). Data from Atlas Levels-only for SS neurons with e11, through e15 birthdays (peak at e12&e13). Thus smaller totals expected. Alternatively, different technique of FB injection may yield fewer labeled cells, as noted also in CRH data from this study. Table 8.01: From the literature, PVH subdivisions with identified Somatostatin (SS). 141 Table of referenced papers showing GRH cells present in PVH Reference; general info Location of data described in PVH Other Guillemin, et al ‘82 Science 218: 585-587 Initial report: “Growth hormone-releasing factor from a human pancreatic tumor that caused acromegaly” Rivier, et al ‘82 Nature 300: 276-78 Second report - Wylie Vale Lab: “Characterization of a growth hormone-releasing factor from a human pancreatic tumor” No IHC. Spiess, et al ’82 Biochem.21:6037-40 “Sequence analysis of a growth hormone-releasing factor from a human pancreatic islet tumor” (Vale Lab) Bloch, et al ‘83 Nature 301: 607-08 “Immunohistochemical detection of growth hormone releasing factor in brain”, described hpGRH in monkey & human ARC (Guillemin Lab) Spiess, et al ’83 Nature 303:532-35 “Characterization of Rat hypothalamic growth hormone-releasing factor” (Vale Lab) Bruhn, et al ‘85 Endocrinol. 117(4): 1710-2 RIA study: Rat hypothalamus = 32.5 fmol/mg dry weight, rat duodenum = 2.15 fmol/mg dry weight. Other brain regions nil. Comparison of hypothalamic and duodenal GRH by highly specific ria for ratGRF. Brain & gut peptides tested as authentic and identical. Sawchenko, et al ‘85 J Comp Neurol. 237 (1): 100-115 “Much smaller groups of neurons were localized in the parvicellular division of the paraventricular nucleus” Most cells were seen in ARC. Smaller groups in PVH and DMH, “and it is unclear whether they contribute to the plexus of rhGRF-stained fibers in the median eminence.” (rhGRF=rat hypothalamic GRH) They found dense GRH fibers in PVH, especially in areas of SS cell localization. Mayo, et al ‘85 Nature 314: 464-7 Rat GRH gene: “both complementary DNA and genomic clones encoding rat hypothalamic GHRF.” Determined to differ from human. (R. Evans Lab, Salk) “The rat GHRF gene spans nearly 10 kilobases (kb) of rat genomic DNA, contains 5 exons and encodes a 104-amino-acid precursor to the rat GHRF peptide. ... Comparison with previously characterized human GHRF cDNA and genomic clones has allowed patterns of conservation of amino-acid and nucleotide sequences between the human and rat GHRFs to be determined.” Meister, et al ‘86 Neuroendocrinol. 42: 237-47 No GRH cells seen in PVH. Only TH was seen at what might be called PVH -- actually, in ventral pv nucleus. Co-localization study with TH; didn’t find GRH in PVH using G-75 (specific anti-ratGRH) on puddled cryostat sections... technique possibly not as sensitive as using free-floating sections. G-75 used by their co-authors (Sawchenko, et al ’85, Sawchenko & Swanson ’90) to show GRH in PVH. Vale, et al ’86 In “Human Growth Hormone”, Raiti & Tolman Eds. “Cell bodies stained for rhGRF have been localized in just two regions” in ARC (ne) and below VMH (non-ne). Not in PVH By some of the same authors (PES, LWS) as Sawchenko, et al - above. From earlier data? No photos, just a composite line drawing of cells and fibers. Sawchenko & Swanson ‘90 Handbook Chem. Neuroanat. V9; Neuropeptides in the CNS, Part II “Growth Hormone Releasing Hormone, Chapter II” “Virtually all ne-GRH cells thought to be in the ARC”. (GRH fiber input to neSS neurons may well be from non-neGRH cells outside the ARC, mostly surrounding VMH.) However this chapter clearly describes cells in rat PVH by rGRH IHC and ISH. GRH first detected in m.e. of embryos at e18, about the same time GH release first observed. Early controversy on GRH distribution due to poor sequence homology between human and rat resolved. Exquisite interplay of SS and GRH release produces ultradian 3-4h peak in GH release on average, which is also associated with bouts of eating. GRH stimulates both GH release and synthesis. Rodier, et al ’90 J Comp Neurol. 291:363-72 GRH cell birthdate study. Found about 12 GRH cells per rat PVH, agrees with Sawchenko ’85, but using a different antibody and strain of rat (Long Evans). Markakis & Swanson ‘97 Brain Research Reviews 24: (2-3) 255-91 ne non pv 1 0 ap 0 0 am/mm 0 / n.a. 0 / n.a. pmm 0 1 pml 0 1 dp 0 0 mpd 0 5 mpv 0 1 lp 0 1 Best over-all comparison to the work described here; ne cells marked w/ Fast Blue from blood, one antibody stain for GRH, second antibody for BrdU cell birthdates. Data from Atlas Levels-only for GRH neurons in PVH with e11, through e14 birthdays. In agreement with the work presented here. Table 8.02: From the literature, Growth Hormone Releasing Hormone (GRH) in PVH. 142 Table of PVH subdivisions, with SS cells in data presented here subdivision pv ap am / mm pmm pml dp mpd mpd (lat) mp v lp f Atlas Level: ne / non-ne Summary Data from Atlas Levels only, as illustrated in preceding figures 22 5/2 23 9/0 24 19/0 25 98/3 26 12/1 27 8/1 22 2/0 23 19/1 24 23/4 22 1/0 24 6/1 25 5/0 26 0/2 25 3/0 26 0/0 25 25/0 26 0/0 27: 2/1 25 0/0 26 0/0 26 0/0 27 0/1 27 0/0 Neuro- endocrine 151 44 1 6 5 0 3 27 0 0 0 0 non Neuro- endocrine 7 5 0 1 0 2 0 1 0 0 1 0 subdivision pv ap am / mm pmm pml dp mpd mpd (lat) mp v lp f Table 8.03: Data from this work, PVH subdivisions with identified Somatostatin (SS). Note: the Atlas Levels are not equally spaced. Therefore, estimates of total cell counts by subdivision within PVH cannot be inferred from these atlas-level data. Cumulative cell counts in PVH as a whole are available on sequential section maps. See figures 8.10 through 8.15, and Appendix IX. Thus, if desired, total number of labeled cells within the nucleus could be calculated. 143 Chapter 9. TH (tyrosine hydroxylase), a marker for Dopamine. Overview: Tyrosine hydroxylase (TH) is the rate-limiting enzyme in the catecholamine synthesis pathway from tyrosine to dopamine (DA). Its presence in cells can be used to infer, with appropriate controls, that they synthesize dopamine and release it at axon terminals. In the PVH, cells positive for TH have been shown to be devoid of two enzymes that further modify dopamine: dopamine beta hydroxylase (DBH) and phenylethanolamine- N-methyltransferase (PNMT). DBH transforms dopamine to noradrenaline and PNMT further transforms noradrenaline to adrenaline. PVH cells positive for TH are thus assumed to be functional dopamine neurons (Swanson et al., 1981). Hypophysiotropic dopamine inhibits pituitary lactotropes (recall figure 2.04). Thus, production of prolactin is stimulated during lactation in females, as might be expected, and inhibited under non-lactating conditions. Prolactin release is also greatly increased in males and females in response to certain specific experimental stressors, notably exposure to ether. Some authors regard the major source of hypophysiotropic dopamine to be from TH cells in the arcuate nucleus (Meister et al., 1986; Kawano and Daikoku, 1987; Voogt et al., 1990), with scant mention of TH cells in PVH. TH terminals have also been seen in apposition to GRH terminals in the median eminence, with an interesting proposed model that they are positioned to modulate GRH release at the level of its terminal axons. There was no proof of source of the TH terminals (neuroendocrine or extrinsic) and location of their cell bodies was a matter of speculation (Sawchenko and Swanson, 1990). However, several investigators have reported significant numbers of TH stained neurons in PVH (concentrated in medial and periventricular parvicellular divisions) that were assumed to be neuroendocrine (Swanson et al., 1981; Chan-Palay et al., 1984; Liposits et al., 1986; Ceccatelli et al., 1989). 144 TH (dopamine): ne, and non-ne cell type distribution in sagittal view: Fig. 9.01 Neuroendocrine and non-neuroendocrine TH cell distribution throughout PVH. There are relatively few TH neurons in PVH (left) compared to other cell types surveyed. They are seen mostly in the caudal half of the nucleus, and slightly more concentrated in ventral regions. Non-ne TH neurons (right), in contrast to ne-TH, are evident throughout PVH except the very rostral part. 145 Fig. 9.02 TH (dopamine) cell type distribution throughout PVH, all peptide-positive cells. Relatively sparse distribution of TH cells in PVH is evident in this parasagittal outline view, with neTH clearly concentrated in posterior PVH compared to non-neTH seen throughout most of PVH except the most rostral and rostro-dorsal part. Neuroendocrine cells are slightly more numerous in the ventral part of caudal TH, otherwise they are seen at quite low density overall. 146 Fig. 9.03 TH (dopamine) staining-consistency, in three serial sets from the same case. The staining pattern is similar, but not identical, in sets of serial 15µm sections stained at different times. This is easily seen in a cell type with low density such as TH, but the same holds true for staining patterns seen for other cell types. At this section thickness and in serial sections, one might expect to see an almost identical distribution of a consistently localized cell type if it in fact existed. The pattern and distribution are qualitatively similar for the same cell type among sections in the same animal and between different animals. While not quantitatively the same, these consistent patterns are representative of a statistical biological similarity between experimental animals. 147 TH (dopamine): ne, and non-ne cell type distribution on Brain Maps Atlas Levels: Fig. 9.04 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures. 148 Fig. 9.05 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 22. Data divided into grids for transfer (minimal change needed for PVH14), as detailed in chapter 7. 149 Fig. 9.06 TH (dopamine) cell type distribution, on Atlas Level 22 & caudally. Distribution of TH cells in PVHap (all non-ne at this rostral level) is similar in sequential sections. No cells were seen in PVH on Atlas Level 21. Thus, it is not illustrated in this series of figures. 150 Fig. 9.07 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 23. Data divided into grids for transfer (minimal change needed for PVH14), as detailed in chapter 7. 151 Fig. 9.08 TH (dopamine) cell type distribution, on Atlas Level 23 & caudally. The few neTH cells seen on Level 23 in PVHap reflect their scarcity at this rostral level of PVH. 152 Fig. 9.09 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 24. Data divided into grids for transfer (minimal change needed for PVH14), as detailed in chapter 7. 153 Fig. 9.10 TH (dopamine) cell type distribution, on Atlas Level 24 & caudally. At this level, no neTH cells are seen, while non-neTH cells are more numerous than all the rostral levels together. Three non-neTH cells are seen in dorsal and ventral PVHpv (seeming to avoid the medial magnocellular division), and the remaining eight are clustered in ventral, mid-PVHap. 154 Fig. 9.11 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 25. Data divided into grids for transfer (minimal change needed for PVH14), as detailed in chapter 7. This is the first level at which data are derived from more than one section in the series. 155 Fig. 9.12 TH (dopamine) cell type distribution, on Atlas Level 25 & caudally. Though derived from different sections, the data distribution pattern at this level is not significantly different from that of the two source-sections. There are now more neTH than non-neTH, though the total is still quite small. Three of four cells are located ventrally, and all are in the medial third of PVH. 156 Fig. 9.13 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 26. Data divided into grids for transfer (minimal change needed for PVH14), as detailed in chapter 7. The scant data shown at this central level of PVH are taken from two adjacent sections in the set. 157 Fig. 9.14 TH (dopamine) cell type distribution, on Atlas Level 26 & caudally. The very few TH cells at this level are all neuroendocrine. Virtually no non-neTH cells are seen at this central, very metabolically active level. 158 Fig. 9.15 PVH14A grid references + TH (dopamine) cell type distribution: Atlas Level 27. Data divided into grids for transfer (minimal change needed for PVH14), as detailed in chapter 7. Three sections map to this Level, which makes little difference since data is on only one section. 159 Fig. 9.16 TH (dopamine) cell type distribution, on Atlas Level 27 & caudally. Data for Level 27 are taken from only the middle section of three that span the level. As it turns out, the section containing a few more cells in the medial part of PVH was not the one sampled for the medial part of the Atlas Level. Thus in this one case, the Atlas Level data map differs significantly from the section(s) adjacent to it. 160 As shown above, the grid-work figures for each Atlas Level are informative in further understanding this method of preparing data for accurate display on Brain Maps templates. In the case for TH, a very low-density cell type in PVH, presentation of single- image projection maps does not yield significantly different information from data displayed on Atlas Levels. It is apparent that while most TH cells are located in the medial third of the nucleus, very few—either neTH or non-neTH—are located in PVHpv. Data display on Atlas Level 26 makes the point that the apparent small cluster of neTH cells seen ventrally and caudally in the sagittal projection are located in the ventral part of the area at the PVHpv - PVHmpd border. Below are two tables. The first shows information about localization of TH cells in PVH obtained from the literature. The second shows data obtained from the Atlas Level representation shown above. Such tabular information for TH, defined by presence of cells in specific PVH subdivisions illustrated in a standard atlas, may be useful as numerical source-data for future bioinformatics analysis. Table of referenced papers showing TH cells present in PVH Reference; general info Location of data described in PVH Other Brownstein & Palkovits, ’76 Front. Neuroendocrinol. vol. 4 PVH= 10+/-1.5 ng/mg “DA is moderate in PVN” -p.13 NPV= +2 (of 1-4) -p. 13 Punch samples: RIA data given in ng protein per punch; sagittal, diagonal-orientation diagrams. They note “Some DA cells were seen by Dahlstrom & Fuxe” -- probably refers to pv cells Kawano & Daikoku ’87 J Comp Neurol 265:242-53 <10/35 me <20/35 pp (neTH/non-neTH) 4non-neTH me 4neTH pp 1neTH me 1non-neTH me 2neTH pp WGA retrograde tracer in posterior pituitary or median eminence to label ne cells; antibody stain for TH. They call “RPR” rostral periventricular area (of PVH?). In RPR 39% of TH cells are labeled (ne) w/ pp injection. With me injection “only some” 19% of RPR TH cells are labeled. “It is concluded that DA neurons involved in the hypothalamic-anterior pituitary axis are located in the arcuate nucleus; those involved in neuro-intermediate lobe function in the RPR.” Markakis & Swanson ’97 Brain Research Reviews 24: (2-3) 255-91 ne non pv 0 16 ap 1 33 am / mm 0 / n.a. 1 / n.a. pmm 1 1 pml 0 1 dp 0 0 mpd 1 0 mpv 0 1 lp 0 0 Best over-all comparison to the work described here; ne cells marked w/ Fast Blue from blood, one antibody stain for TH, second antibody stain for BrdU cell birthdates. No mm cells plotted distinct from ap or pv. No cells plotted in mpdl (lateral part). 3 or 4 neTH were ventral to PVH in pv zone. Data from Atlas Levels-only for TH neurons with e11, through e15 birthdays (peak PVHap at e13, others at e12). Thus, smaller totals expected. Different technique of FB injection may yield fewer labeled neuroendocrine cells, as noted also in CRH data from this study. Table 9.01: From the literature, PVH subdivisions with identified TH (dopamine) cells. 161 Table of PVH subdivisions, with TH cells in data presented here Subdivision pv ap am mm pmm pml dp mpd mpd(lat) mpv lp f Atlas Level: ne / non-ne Summary Data from Atlas Levels only, as illustrated in preceding figures 22: 0/0 23: 0/1 24: 0/3 25: 1/1 26: 2/0 27: 0/0 22: 0/2 23: 2/2 24: 0/8 22: 1/0 24: 0/0 25: 1/1 26: 1/0 25: 0/0 26: 0/0 25: 0/1 26: 1/0 27: 0/2 25: 0/0 26: 0/0 26: 0/0 27: 1/1 27: 0/0 Neuroendo- crine (total) 3 2 1 0 1 1 0 1 0 0 1 0 non Neuro- endocrine 5 12 0 0 1 0 0 3 0 0 1 0 subdivision pv ap am mm pmm pml dp mpd mpd(lat) mpv lp f Table 9.02: Data from this work, PVH subdivisions with identified TH (dopamine) cells. Note: the Atlas Levels are not equally spaced. Therefore, estimates of total cell counts by subdivision within PVH cannot be inferred from these atlas-level data. Data on cumulative cell counts in PVH are available from sequential section maps. (See figures 9.06, 9.08, 9.10, 9.12, 9.14, and 9.16.) Thus, an inference of total labeled cells within the nucleus could be calculated. 162 Chapter 10. Thyrotropin Releasing Hormone Overview: Thyrotropin Releasing Hormone (TRH) was the first hypothalamic substance proven to cause hormone release from the anterior pituitary gland. It stimulates pituitary thyrotropes to release thyrotropin (thyroid stimulating hormone), thereby exerting control over the thyroid gland and basal metabolism. The story of the race to purify and identify TRH and other putative hypothalamic releasing factors (now typically described as hormones) is a colorful and exciting chapter in the history of modern Neuroscience. In the 1960s TRH was purified and characterized by Schally and co-workers from porcine and bovine hypothalami and shortly afterward by Guilleman and co-workers from ovine hypothalamus—work for which they would later share a Nobel Prize (Schally et al., 1966a; Schally et al., 1966b; Burgus et al., 1969; Burgus et al., 1970). Its sequence was soon identified and active samples were synthesized for physiologic testing and production of antisera (Boler et al., 1969; Burgus et al., 1969; Folkers et al., 1969; Enzmann et al., 1971). Thus, Geoffrey Harris’ early hypothesis of primary hypothalamic control over the endocrine system via the portal circulation (Harris, 1955) was strengthened—strongly supported, at least for the pituitary- thyroid axis. TRH is a tri-peptide (thyro-glutimyl-histidyl-proline amide), a final bioactive product (one of several small peptides) cleaved from a larger precursor peptide, pre-proTRH. Perhaps due to the difficulty of obtaining a good antibody to such a small bioactive molecule, detailed data on distribution of identified TRH-containing cells in the PVH accumulated surprisingly slowly. For the most part, early papers presented descriptive surveys of TRH seen in various areas of hypothalamus. For example, Lechan & Jackson (Lechan and Jackson, 1982) referred to a 1980 paper by Johansson and Hokfelt that studied TRH at the electron microscopic (EM) level as “one of the 'remarkably limited' IHC studies of TRH 163 localization in the hypothalamus, although antibodies became available shortly [after] the original chemical characterization in 1969.” In that EM study, Johansson and Hokfelt (Johansson et al., 1980) showed TRH in large dense core vesicles (diameter ~1000Angstroms) localized within cell bodies and nerve terminals. Cell bodies were described in the periventricular, peri-paraventricular and perifornical areas. High concentrations of TRH-containing fibers were observed in the pituitary stalk, passing through to terminate primarily in the medial part of the external layer of the median eminence with some terminations seen more laterally. As shown in Table 10.01 below, there were a number of early immunohistochemical studies describing TRH cells and fibers in PVH (Winokur and Utiger, 1974; Brownstein et al., 1974; Zimmerman, 1976; Johansson and Hokfelt, 1980; Lechan and Jackson, 1982; Nishiyama et al., 1983; Nishiyama et al., 1985; Ishikawa et al., 1986; Tsuruo et al., 1987; Ceccatelli et al., 1989; Kawano et al., 1991; Markakis and Swanson, 1997; Legradi and Lechan, 1998; Fekete et al., 2000). However, for the most part they showed only a figure or two and very few distinguished neuroendocrine versus non-neuroendocrine TRH cells. Later work confirmed the consensus view (discussed earlier in Chapter 2) that most neuroendocrine TRH cells were located in PVHmpd. They were reported to be in higher concentration at the medial and dorsal parts of the subdivision, somewhat mixed with CRH neurons that were in highest concentration at middle and ventral parts of PVHmpd (Swanson, 1986). See Table 10.1 below for comments on TRH distribution described in specific references. For in-depth background discussion about TRH, see Appendix XI. In the work described here, a well-characterized rabbit anti-TRH antibody was used (gift of Dr. Martin Wessendorf). TRH is synthesized as a precursor peptide containing five copies of the sequence Gln-His-Pro-Gly, QHPG, linked by other peptides. The antibody from Dr. Wessendorf was directed against PPT (160-169, SFPWMESDVT), one of the larger 164 peptide precursors of TRH rather than to the tri-peptide itself. Although PPT does not have TRH secretagog function, conjugation to this precursor molecule produced a high-titer antibody that also showed high-specificity for staining cells proven to contain and release authentic TRH (Carr et al., 1992). Another antibody used for preliminary staining series was exhausted, and no replacement was available. Other TRH antisera tested proved inadequate for visualizing TRH-containing neurons using combined retrograde label of neuroendocrine neurons and immunohistochemical methods. See Appendix IV for details of antibody testing and preliminary experiments used to optimize experimental parameters. TRH: ne, and non-ne cell type distribution in sagittal view: Sagittal projection view illustrations of neuroendocrine and non-neuroendocrine TRH are shown below in larger size than in the previous chapter. Thus, one can clearly see the greater numbers of cells and appreciate subtle differences in their distribution patterns. 165 Fig. 10.01 Neuroendocrine TRH cell type distribution throughout PVH. Neuroendocrine TRH cells are located mostly in the caudal two thirds of PVH, concentrated more dorsally and ventrally around Atlas Levels 25 and 26, with a more uniform concentration in the middle parts of caudal levels. There are very few neTRH cells in rostral PVH. 166 Fig. 10.02 Non-Neuroendocrine TRH cell type distribution throughout PVH. Non-neuroendocrine TRH cells are seen throughout PVH, except for its most rostral levels. Density of cells in the rostral half of PVH is clearly higher than in caudal PVH. This is in sharp contrast, and somewhat complimentary, to the distribution of neTRH cells. 167 Fig. 10.03 TRH cell type distribution throughout PVH, all peptide-positive cells. In this image showing all ne and non-ne TRH cells displayed together, the complimentary density of hypophysial-projecting and non-neTRH neurons is quite obvious. The number of TRH cells (most of which are non-ne) in rostral PVH was somewhat surprising, based on references in the literature. Dorsal and ventral separation of neTRH cells at the level of PVHmpd was also unexpected, as was the large population of ne and non-ne TRH cells at more caudal levels. 168 TRH: ne, and non-ne cell type distribution on Brain Maps Atlas Levels: Fig. 10.04 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures. 169 Fig. 10.05 TRH cell type distribution on Atlas Level 22 Fig. 10.06 TRH cell type distribution on Atlas Levels 23 and 24. 170 Fig. 10.07 TRH cell type distribution on Atlas Level 25. Fig. 10.08 TRH cell type distribution on Atlas Level 26. 171 Fig. 10.09 TRH cell type distribution on Atlas Level 27. In the figures below, Atlas Level images are shown along with data maps from their adjoining sections. Because PVH10 was sectioned in a plane close to that of Brain Maps, very little alteration or adjustment in data grids was needed when transferring TRH data to Atlas Levels. Compensation for curvature of the third ventricle (moving data in thin horizontal strips) was discussed earlier. Otherwise, the section mapped to the corresponding Atlas Level is the first one shown in the figure. These illustrations are mainly presented to show more detail between Atlas Levels because of the rapid change from one section to another, even in this high-resolution data set. This is a feature that is not apparent in observing only the six Atlas Level maps above. 172 Fig. 10.10 TRH cell type distribution in rostral PVH, on Atlas Level 22 & caudally. The predominance of non-neTRH in rostral PVH is clearly shown on Level 22 and the sections caudal to it. The positive cells are distributed throughout PVHap, with no preference for PVHpv. A few cells are present in PVHam, an area usually seen to contain OXY magnocellular neurons. One of the two neTRH cells in PVHam was also positive for SS. This was one of less than 10 such double-labeled cells seen throughout the series stained with antisera to SS and TRH. 173 Fig. 10.11 TRH cell type distribution in rostral PVH, on Atlas Level 23 & caudally. At this level, three of four neTRH cells are located in PVHpv. Most positive cells are non-ne, and more than half of them are in (or quite close to) PVHpv. The remaining non-neTRH cells in PVHap are clustered in the dorsal aspect of the subdivision, occupying a diagonal or triangular-shaped area. 174 Fig. 10.12 TRH cell type distribution in central PVH, on Atlas Level 24 & caudally. At this level, non-neTRH cells still predominate, but they are no longer seen in PVHpv. Here they are distributed across most of PVHap, in a slightly diagonal array. Here, the dorsal part of PVHap is notably lacking in positive cells. Among the few cells in PVHpv are the few neTRH cells present. They seem to avoid PVHmm, another area usually seen to be occupied by OXY magnocellular neurons. Section 4/2 at Level 24 is the last section to show a preponderance of non-neTRH cells. More neTRH cells begin to appear in the next section (9 out of a total of 29). In the next section (immediately rostral to Level 25) almost half of the TRH positive cells are neuroendocrine, most of them in the medial part of the nucleus. neTRH cells are disbursed fairly evenly from dorsal to ventral, while the non-neTRH cells are seen as before in a diagonal band directed upward and laterally from mid-PVH. The odd pattern of cells in dorso-lateral PVH compared to the putative PVH outline from adjacent Nissl series may be partly a function of differential shrinkage between Nissl-stained and antibody-stained sections. 175 Fig. 10.13 TRH cell type distribution in central PVH, on Atlas Level 25 & caudally. Level 25 shows many neTRH cells in PVHmpd, as expected from the literature. They extend from a concentrated group in ventral PVHpv diagonally upward into the ventral half of PVHmpd, with a few extending into PVHmm. This is in contrast to their expected concentration in dorsal PVHmpd. Few ne cells are seen in middle PVHpv and none are located dorsally near the midline. non-neTRH cells are located in dorsal and lateral PVHmpd, with some in medial PVHdp and a few in very lateral PVHmm. In general the two populations do not overlap in space, they show rather complimentary distributions, with non-ne cells located more laterally. A few sections caudal to Level 25 the clustering of neTRH more densely in dorsal and ventral compared to central PVHmpd that was noted in the sagittal view is seen. Non-neTRH cells in the more caudal sections are now intermixed with ne cells, but they are located slightly more lateral than the main body of medially located ne cells. The lone non-neTRH seen is at the ventral border of PVHmpd near its junction with PVHmpv and PVH pml. 176 Fig. 10.14 TRH cell type distribution in caudal PVH on Atlas Level 26 & caudally. At this level, neTRH cell distribution is more like that expected from the literature, with cells distributed mostly in the medial and dorsal aspect of PVHmpd. There are no cells in the ventral or most dorsal part of the periventricular zone, but a small cluster is seen in mid-PVHpv, close to a similar grouping in PVHmpd. A few ne cells are located laterally, at the borders of PVHmpv with PVHpml, just below the center of PVHdp. At this level there are more than four times as many ne as non-ne cells. The non-ne cells are seen in two groups, one dorsal and one more ventral, located more laterally in PVHmpd than the neTRH cells. In sections caudal to Level 26 neTRH cells are seen in the medial mpd and the adjacent PVHdp, then clustered more generally in mpd as it grows smaller. Non-ne cells are mixed, generally located more ventral in caudal sections where they increase from 1/4 th to 2/5 th of total seen. 177 Fig. 10.15 TRH cell type distribution in caudal PVH on Atlas Level 27 & caudally. At this most caudal Level, neTRH cells are twice as numerous as non-ne cells and they are located centrally in the now-restricted PVHmpd. A few of each type are in PVHpv, with the balance of non-ne cells mixed among the center of the main group of neTRH cells. This is in contrast to distribution a few sections rostral, where a cluster of non-ne cells was seen ventral to the neTRH in PVHmpd. 178 In the frontal Atlas Level maps above, the patterns seen vary from those apparent in sagittal-projection view. They reveal a somewhat different pattern of distribution from that expected according to the literature. Distribution of neTRH is less restricted than expected, with many more cells caudally. In addition, the value of high-resolution mapping is appreciated when noting the rapid change in apparent distribution pattern between one section and the next, particularly at areas of highest density of TRH neurons. A surprising and consistent feature is the great number of non-neTRH neurons located in rostral areas that are almost devoid of ne cells, while in the caudal part of PVH they are intermixed with neTRH cells, although lower in number. Below is a table showing details of some reports of TRH cell distribution from the literature. Following is another table of TRH data on Atlas Levels (divided by PVH subdivision) from this work. 179 Tables of PVH subdivisions vs. referenced papers showing cells present for TRH: Table of referenced papers showing TRH cells present in PVH Reference; general info Location of data described in PVH Other Winokur & Utiger ’74 Science 185: 265-267 “hypothalamus” regional / RIA Followed immediately by the next paper showing regionalization Brownstein, et al ’74 Science 185: 267-269 Crude diagrams *punch/RIA “NPE”=~ 4ng/mg (periventricular?) m.e.=+ (no PVH) Zimmerman, ’76 Frontiers in Neuroendocrinology V4 Chapter from book on early days of Neuroendocrinology brief mention of TRH, “it has not been seen intracellularly” Johnsson & Hokfelt ’80 JHstochem&Cytochem 28: 364-366 2 minimal maps first imm-cytochem. [ap?] ~2 dots ?cells [mpd?] ~2dots?cells Lechan&Jackson ‘82 Endocrinol 111:55-61 New fixative, better figures they say data “Confirms Hokfelt” est. cell numbers from photos IHC/TRH, 60µm [vibratome] fig 8 drawing Lt-PVH@5levels 3levels 4,28,19 7,24,12 … 5levels 24cells 2 sides Lt=43 Rt=45 … 2levels 23cells “PVNm” 2lev, 169 and 115 … 3levels 30cells “PVNl” L~11 Rt=30 … 1 level 2cells “peri-fx” L=n/a Rt=27 … 2levels 11cells Nishiyama et al. ’83 Biomed Res (Suppl)4:65-74 [pv] ~15? estimate from colch photo TRH ontogenesis [mpd] ~30? estimate fr/ colch photo PVH: increasing number of + cells on e20.5; + in me e22.5 PTU in adult=>ne cells labeled Nishiyama et al. ’85 Brain Res345(2):205-18 Excellent paper! (Kawano & Daikoku group) pv table2= ~10 int ~25 PTU ~99col ap table2= ~27 int ~35 PTU ~108col “pm” no #s for pm dp table2= ~4 int. ~3 PTU ~18col “mp” table2= ~60 int ~119PTU ~210col. lp table2= ~3 int ~2 PTU ~20col Intact <<colch; colch>PTU(chemical thyroidectomy). “PTU implies only cells in pv and mp are HPT- axis TRH neurons”. 5µm paraffin sections, 1-in-20 stained. Swanson & Kuypers PVH parcellation. Includes figures with anatomic location and tables of total cell counts by subdivision. PVN totals: ~106 intact; ~193 PTU ~458 colch Ishikawa et al. ’86 Neuroendocrinol 44:54-58 septal HRP shows TRH-IHC cells that project to septum (non-ne?) [mpd] 1photo ~20 cells [pv] 1photo ~20 cells Tsuro, et al ’87 ExpBrainRes 68:213-17 (Hokfelt group) Comparison to RIA, new fixative; TRH-IHC [pv] “many cells” [ap] “CNS survey” [mpd] cells in photo [f] “around fornix” Ceccatelli et al. ’89 Neuroendocrinol 49(3):309-23 [pv] 29 [ap] 31 “dc” 2 “mp” 100 [lp] 0 [f]? 18 Hokfelt student’s Ph.D work (noted in thesis-TRH cells are GR positive). Best paper showing TRH localization in PVH, though no ne label. Co-localization study (stain, elute, re-stain): TRH/CRF, TRH/NT. 200g S-D rats, cryostat sections cut @14µm; 80 from each PVH. Swanson & Kuypers nomenclature. Counts from two diagrams showing CRH or NT stained with TRH (Cells in [f] seem medial). One NT/TRH cell seen in ventral PVHpv. Six PVH levels surveyed -- schematic maps. Kawano,et al. ’91 J CompNeurol 307(4):531-8 (Daikoku group) [pv] ~51/30 ne=58.1% [ap] ~125/2.5 ne=1.6% [pml] 4/0 faint [dp] 5.8/3 ne= 51.7%[!] “mp” ~146/~100 ne=68% ~lp 6.3/.3 ne4.8% WGA retrograde tracer in m.e.; 5µm paraffin sections at 100µm-intervals. Cell counts, plus drawings of some ne cell distribution. “anterior, medial, and periventricular parvicellular neurons” Data given as statistical number of ne vs. non-ne. Antibody to TRH prohormone used for IHC. 180 Table of referenced papers showing TRH cells present in PVH (continued) Markakis & Swanson ’97 Brain Res Rev 24(2-3):255-91 [pv] ne=1 non =9 [ap] ne=2 non =23 [am] ne=0 non =1 [dp] ne=0 non =3 [pmm &pml] ne=3 [mpd] ne=4 non =10 [mpv] non =2 [lp] ne=1 Best comparison data to work presented here: Fast Blue in blood to mark ne cells, same antibody to TRH. Study used BrdU to mark cell birthdates, IHC for TRH +BrdU. Surprisingly few neTRH cells seen and TRH over-all, perhaps due to FB injection technique &/or acid pre-treatment for Brdu IHC. Legradi & Lechan ’98 Endocrinol 139(7):3262-79 One photo, no cell counts, Typical pattern of staining in PVHmpd. This group has done a great deal of work on the HPT-axis. ARC/NPY cells projecting to TRH cells. “NPY nerve fibers densely innervate hypophysiotropic TRH perikarya and dendrites in PVN” -- but no retrograde tracer was used. They survey “NPY terminals contacting TRH perikarya and first order dendrites in the medial parvocellular and periventricular subdivisions of the PVN”, and report “about 11 terminals per cell” Little difference in density of label for animals whose GRH cell population in ARC had been ablated with fetal MSG treatment. Fekete, et al ’00 JNeurosci 20(4):1550-8 [pv] ~17cells 3 levels [ap] ~20cells 1 level “mp”~60+cells 2 levels Same research group as Legradi and Lechan (above). Color co-localization; using ISH, shows aMSH terminals innervating TRH neurons--many cells. Con-focal images, counts above from survey photos and maps. Title implies this innervation “prevents fasting-induced suppression of prothyrotropin-releasing hormone gene expression”. Table 10.01: From the literature, PVH subdivisions with identified TRH cells. Table of PVH subdivisions: TRH cells in Atlas Level data presented here Subdivision [per Atlas] pv ap am mm pmm pml dp mpd mpd (lat) mpv lp f 0/0 0/0 Atlas Level: ne / non-ne 22: 0/3 22: 0/21 22: 2/3 23: 3/1 5 23: 1/33 Summary Data 24: 3/3 24: 0/37 24: 0/1 from Atlas Levels only, as illustrated 25: 17/ 1 25: 5/4 25: 1/3 25: 21/8 25: 0/0 25: 0/0 in preceding figures 26: 14/ 0 26: 0/1 26: 0/0 26: 28/7 26: 0/0 26: 2/4 27: 2/2 27: 25/5 27: 1/1 27: 0/0 subdivision pv ap am mm pmm pml dp mpd mpd (lat) mpv lp f neuro- endocrine 39 1 3 0 5 0 1 74 0 2 1 0 non- neuro- endocrine 24 91 36 1 4 1 3 20 0 4 1 0 subdivision pv ap am mm pmm pml dp mpd mpdl mpv lp f Table 10.02: data from this work, PVH subdivisions with identified TRH cells. 181 Chapter 11. Corticotropin Releasing Hormone Overview: Corticotropin releasing hormone (CRH) is a 41-amino acid neuropeptide first characterized from extracts of sheep hypothalamus by Vale and co-workers (Vale et al., 1981). The race to isolate and characterize this peptide (as with TRH and later, GRH) provides a colorful and interesting chapter in neuroendocrinology. Because of overwhelming interest in stress-related clinical observations, CRH was quickly enlisted to study important physiological processes. Synthesized versions of the peptide and antibodies to it were produced, and many experiments using CRH as a pharmacological tool were initiated in the hope of understanding corticosterone regulation in normal and disease states (Vale et al., 1983). Antibodies to synthetic ovine CRH were produced and used to document its presence “in neurones of anterior parvocellular region of the paraventricular nucleus complex in the rat hypothalamus, as well as in nerve fibers projecting to the median eminence, basal arcuate area, and the posterior pituitary” (Bloom et al., 1982). Swanson and colleagues described CRH distribution in PVH and throughout the rat brain in detail in 1983 (Swanson et al., 1983). CRH cells and fibers were seen in many regions of brain from cortex to hypothalamus. However, it was clear the hypophysiotropic hormone was synthesized in cells of PVH whose axons projected to the external layer of median eminence and thus to the pituitary portal vasculature. As mentioned in Chapter 2, it is well accepted that neCRH cells are located primarily in PVHmpd, somewhat intermixed with TRH-containing neurons. In fact, immuno- and in situ hybridization-histochemical stains for CRH neurons are used as hallmark probes to identify the neuroendocrine parvicellular functional compartment of PVH (Swanson and Simmons, 1989). Within PVHmpd, CRH cells are thought to be somewhat ventrally located compared to TRH. Some CRH cells are seen in magnocellular subdivisions, and it is well documented 182 that CRH neurons can express other peptides (Sawchenko and Swanson, 1985; see also many citations in previous chapters). Notably, in terms of endocrine function vasopressin is a co-secretagog for ACTH that is expressed in CRH neurons under highly stressful conditions (Sawchenko et al., 1984; Swanson, 1986; de Goeij et al., 1991). This is thought to afford another level of pituitary responsiveness when an additional stressful signal is imposed acutely on CRH neurons that are at maximal expression and secretion levels. One aim of the work presented here was to document such hypophysiotropic co-localization at high resolution throughout PVH. See Chapter 13 for discussion of peptide co-localization seen in the work presented here, and Table 11.01 below for a synopsis of CRH cell type distribution described in the literature (Bloom et al., 1982; Swanson et al., 1983; Sawchenko et al., 1984; Whitnall, 1988; Ceccatelli et al., 1989; Lennard et al., 1993; Markakis and Swanson, 1997). 183 CRH: ne, and non-ne cell type distribution in sagittal view: Fig. 11.01 Neuroendocrine CRH cell type distribution throughout PVH. In this comprehensive view, many more neCRH cells are seen throughout PVH than previously appreciated. While the densest concentration of cells is seen in the area encompassed by PVHmpd, unexpected distribution of ne cells in rostral (especially) and caudal PVH is apparent. 184 Fig. 11.02 Non-Neuroendocrine CRH cell type distribution throughout PVH. In this comprehensive map, non-neCRH distribution in PVH seems to be complimentary to that of neCRH. There are virtually no non-neCRH cells in the area of highest neCRH concentration. In rostral and caudal areas, non-neCRH distribution seems similar to that seen above for neCRH. In agreement with known literature, non-neCRH cells are seen caudally, in PVH descending functional compartment. Significant numbers of non-neCRH cells seen in rostral PVH was unexpected. 185 Fig. 11.03 CRH cell type distribution throughout PVH, all peptide-positive cells. Intermixing of ne and non-ne cells at rostral and caudal areas of low concentration are emphasized in this combination view of CRH staining, while any non-ne cells in the area of highest concentration of neCRH (PVHmpd) are not visible. As expected, the most caudal cells are non-neCRH. 186 CRH: ne, and non-ne cell type distribution on Brain Maps Atlas Levels: Fig. 11.04 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures. 187 Since the exemplar experiment for CRH was PVH12, additional figures showing the derivation of data transfer on grids from the original section maps are shown immediately after figures of CRH distribution on each Atlas Level. Recall the detailed explanation of data transfer methodology for this experimental animal in Chapter 7 (Meta-Data Mapping Strategy). Because the plane of section for PVH12 is very oblique compared to that of the Brain Maps atlas animal, partial data were taken from two or three consecutive sections for anatomically accurate display on Atlas Level templates. This is shown in detail in a composite figure following each Atlas Level figure. Note: for clarity in display figures below the grid shapes over experimental maps were adjusted to reflect their eventual mapping location on the Atlas template. This is different from the actual meta-data transfer method used, but allows viewing of the original data maps in correct relation to each other. Fig. 11.05 CRH cell type distribution in rostral PVH, on Atlas Level 22. The few non-neCRH cells at this Level are at the ventral edge of PVHap. In contrast, the neCRH cells are lateral in PVHap and in PVHam, which is typified by presence of magnocellular OXY neurons. 188 Fig. 11.06 CRH cell type distribution in rostral PVH, on Atlas Level 22 & caudally. Typical data transfer section pattern for most PVH12 data transfers to Atlas Levels is seen clearly here at rostral levels containing few data points. Portions of data from three different sections are used to create an anatomically correct transfer of data to the Atlas Level template. 189 Fig. 11.07 CRH cell type distribution in rostral PVH, on Atlas Levels 23 and 24. On these Atlas Levels neCRH are distributed in both ap and pv subdivisions with a slight preponderance seen medially. On level 23, they are located dorsally in PVHpv, while on Level 24 they are concentrated ventrally. On Level 24, one ne-CRH cell is seen in the medial magnocellular division. On level 23 almost the same number of non-ne as neCRH cells are seen, while on Level 24 only two of 12 total CRH cells are non-ne. On Level 23 non-ne cells are seen mostly in the dorsal part of PVHap. (Incidentally, this is a similar pattern to that seen for non-neTRH cells in the same area.) Following are two figures (11.08 and 11.09) that show specific data transfer sections and selection grids used in plotting CRH distribution on Atlas Levels 23 and 24 shown in the figure above. Patterns of data distribution in the two sections with most data represented on the Atlas Levels are clearly similar. This, in part, visually verifies that data distribution patterns are not significantly altered in the meta-analysis data transfer procedure used to display experimental in anatomically correct position on Atlas Level templates. 190 Fig. 11.08 CRH cell type distribution in central PVH, on Atlas Level 23 & caudally. 191 Fig. 11.09 CRH cell type distribution in central PVH, on Atlas Level 24 & caudally. 192 Fig. 11.10 CRH cell type distribution in central PVH, on Atlas Level 25. At this level dense distribution of neCRH is expected in PVHmpd. However, a significant proportion (more than half) of neCRH cells are seen in the medio-dorsal part of PVHpmm (magnocellular neuroendocrine compartment), extending medially into very ventral PVHpv. The relatively few non- neCRH cells at this level (~10% of total CRH cells) are scattered at the border of mpd with PVHdp and among neCRH cells in the medial part of the pmm magnocellular division. Here and in figure 11.11 below note the neCRH cells in PVHmpdl (PVHmpd-l). These were recognized as genuine PVH neurons only after characterizing the parvicellular identity of this subdivision, (detailed in Chapter 5) and re-examining the original data. Other cells in this thin lateral extension of PVHmpd are also seen in adjacent sections shown in figure 11.11. A few neVAS cells (stained with a monoclonal antibody to VAS, but not co- localized with CRH) were seen in close proximity to the lateral neCRH cells in these sections. However they were not from the set mapped onto Atlas Levels in the figures for Chapter 12. 193 Fig. 11.11 CRH cell type distribution in central PVH, on Atlas Level 25 & caudally. Note the neCRH cells in PVHmpdl. They were identified as genuine PVH neurons only after characterizing the parvicellular identity of this subdivision, as detailed in Chapter 5. Other cells in this thin lateral extension of PVHmpd are also seen in adjacent sections. 194 Fig. 11.12 CRH cell type distribution in caudal PVH on Atlas Level 26. Level 26 shows the typical distribution of neCRH in PVHmpd, with some overlap into adjacent PVHpv and a few cells in the medial part of PVHpml. Two non-ne cells (<0.1% of total) are in central PVHpml. Level 26 is termed the “classic level” of PVH, because it illustrates all three functional compartments very clearly. This is the level usually shown to illustrate typical distribution of neCRH in PVHmpd. The distribution pattern and density shown here agree for the most part with descriptions in the literature. Density in ventral PVHmpd is slightly less than in the central part of the subdivision. Descriptions of TRH vs. CRH cell distribution usually describe CRH concentration in mpd more ventral compared to TRH cells, seen more dorsally and extending medially into PVHpv. Non-neCRH cells are almost absent (<0.1% of total) at this level of highest neCRH concentration. Unexpectedly, the two non-neCRH cells seen are located in the center of PVHpml, a magnocellular area usually seen to contain almost a pure population of neVAS cells. 195 Fig. 11.13 CRH cell type distribution in caudal PVH on Atlas Level 26 & caudally. These sections, only parts of which map to Level 26, all show the typical pattern of neCRH distribution. 196 Fig. 11.14 CRH cell type distribution in caudal PVH on Atlas Level 27. The neCRH cells at this level, though more caudal than might be expected, are still clearly in the parvicellular neuroendocrine PVHmpd. The relatively few (~2%) non-neCRH neurons are mostly among the ne cells in mpd, not in the descending compartment as with parvicellular neurons in PVHlp. 197 Fig. 11.15 CRH cell type distribution in caudal PVH on Atlas Level 27 & caudally. 198 Tables of PVH subdivisions vs referenced papers showing cells present for CRH: References showing CRH cells described in PVH Bloom, et al ’82 Regulatory Peptides 4:43-8 Fibers in external layer of median eminence. Cell bodies in parvocellular PVH give rise to a massive pathway to median eminence. First immunohistochemical report of CRH neurons in PVH: photos, but no cell counts. (Vale, et al ’82: first paper on CRF described its presence in ‘medial basal hypothalamus’ by RIA). pv 246T ap 258T am/ mm 58T / 27T “pm” 293T dp 85T “mp” 808T lp 219T Swanson, et al ‘83 Neuroendocrinology 36:165-86 T=total cells throughout PVH. Cells on six levels of PVH 54 20 11 / n/a 22 9 “mp” 248 mpdl 6 26 Antisera to ovineCRH, serial 30µm sections, colchicine treated. Trajectory of CRH fibers from PVH to median eminence described. Fibers in ext. lamina of m.e., spanning medio-lateral width. Several photos, figure of six levels through PVH (total numbers from serial sections by subdivision given in figure legend). “Some 2000 positive neurons, less than 400 (~20%) in magnocellular divisions”. In adrenalectomized animals (data not shown), ~750 PVH cells brightly CRH positive, 90% in pv,medial, and medial-lateral divisions, < 6% in magnocellular divisions: definitive distribution paper. Sawchenko,Swanson&Vale ’84 J. Neuroscience 4:1118-29 pv 23 ap 12 am / mm 2 / 6 pm 14 dp 5 mp 187 lp 5 Co-localization study of CRH with OXY or VAS in magnos (elution, re-stain). Colchicine treated, 1- in-5, 20 m sections. Six levels through PVH (fig 2) show anti-ratCRH cells in subdivisions. Totals similar to results with anti-ovineCRH in number (~2000 cells); 10% less in eluted material. In adrenalectomized rats: Sawchenko, Swanson & Vale ’84, (PNAS) showed CRH/VAS in parvos. Whitnall ’88 J.Comp.Neurol. 275:13-28 pv 10 ap 16 am/ mm 9 / n/a pm 20 dp 6 mp 205 lp 8 Six PVH levels shown (ea 200 m). 271cells mapped: Est. to be 10% of total CRH cells on one side. Implied co-localization study on 3 serial 1µm-EM immuno-sections, seeking quantitation of CRH parvicellular cells expressing VAS or Neurophysin-VAS (NP-VAS). Numbers above from fig 3 data: almost all CRH/VAS in medial parvicellular division. Eliminated CRH magnos, defined as staining also for OXY or NP-OXY (results in agreement with Sawchenko et al ’84, PNAS). Ceccatelli etal, ‘89 Neuroendocrinol. 49, 309-23 pv 32 ap 20 am / mm 8 / 5 pm 32 dp 8 mp 185 lp 7 Very good comparative PVH peptide IHC; Co-localization study of several peptides (includes NT/,CRF and CRF). CRF totals from 14 m, one ~1-in-3or4 series; numbers from maps of 6 PVH levels. Note: am and mm subdivisions shown, but not mpv; and dp was labeled ‘dc’. Lennard, et al ’93 J.Neuroendocrinology 5:(2) 175-81 pv 9 ne 24 non ap 12 ne 30 non am (mm=0) 0 ne 6 non pm 15 ne 18 non dp 0 ne 21 non “mp” 240 ne 180 non lp 3 ne 9 non Early identification of ne- vs. non-ne CRH cells; i.p. FluoroGold, CRH fluoresc. IHC. Hypophyseal CRH cells characterized as those co-localized with Fluoro-Gold. “Vast majority of CRH+/FG+ are in medial parvicellular divisions.” Both sexes were studied: [all CRH cells, not just PVH] 1-in-5, 30 m- vibratome= 65%neCRH, 1-in-10, 8 m-paraffin= 74%neCRH (p176). Only 70% of mpd CRH cells were ne, vs. 25% in other subdivisions (p179). Nice drawings of 6 levels through PVH (p180). Markakis&Swanson ’97 Brain Research Reviews 24: (2-3) 255-91 ne non pv 6 8 ap 22 32 am/mm 1/n.a. 2/n.a. pmm 1 0 pml 3 4 dp 2 4 mpd 19 26 mpv 0 3 lp 1 2 Best over-all comparison to the work described here; ne cells marked w/ Fast Blue from blood, one antibody stain for CRH, second antibody stain for BrdU cell birthdates. No mm cells plotted distinct from ap or pv. No cells plotted in mpdl (lateral part). Data from Atlas Levels-only for CRH neurons with e12, e-13,-e14 birthdays (peak at e-13). Thus, smaller totals expected. Alternatively, different technique of FB injection may yield fewer labeled cells, as noted also in TRH data from this study. Table 11.01: From the literature, PVH subdivisions with identified CRH cells. 199 Table of PVH subdivisions, with CRH cells on Atlas Levels, presented in this work subdiv. AtlasLevel pv 23-7 ap 22-4 am/ mm 22&24 pmm 25 pml 26 dp 25-6 mpd 25-7 mpdl 25-6 mpv 26 lp 27 f 27 Level: ne/ non-ne 22: 4/2 22: 3/0 23: 2/0 23: 9/9 24: 5/0 24: 3/2 24: 0/0 25: 10/2 25: 44/4 25: 0/1 25: 21/3 25: 3/0 26: 10/0 26: 11/2 26: 0/0 26: 204/0 26: 0/0 26: 0/0 summary data on Atlas Levels from preceding figures 27: 1/0 27: 18/4 27: 0/1 27: 0/0 Totals by subdivision, CRH cells on PVH Atlas Levels in figures presented above subdivision pv ap am/ mm pmm pml dp mpd mpdl mpv lp f Neuro- endocrine 28 16 3 / 0 44 11 0 243 3 0 0 0 non Neuro- endocrine 2 13 0 / 0 4 2 1 7 0 0 1 0 subdivision pv ap am/mm pmm pml dp mpd mpdl mpv lp f Table 11.02: Data from this work, CRH cells in PVH subdivisions seen on Atlas Levels. In the 15µm sections plotted, 348 neCRH cells and 30 non-neCRH (about 250 total in mpd) were counted on Atlas Levels showing PVH. This is equivalent to 25% of actual total neCRH cells that might be seen in PVH on a 60µm section at each Level in the Atlas animal. However, since distances vary between Atlas Levels, these numbers cannot be interpreted as a given percentage of CRH cells expected throughout PVH. In these studies there were 1821 total CRH cells mapped (from the exemplar series) in PVH12 as shown above, and 2002 CRH cells mapped throughout PVH in a similar series from case PVH14 (see Appendix X). This is similar to published reports (Swanson, ’83, Sawchenko, ’84), where about 2000 CRH cells were counted throughout PVH in 30 m sections. 200 Chapter 12. Oxytocin and Vasopressin Overview: Oxytocin (OXY) and Vasopressin (VAS), the earliest molecules recognized as neurosecretory products were for many years the most studied hypothalamic hormones. Recognition of OXY and VAS contributed directly to understanding of neurosecretion in general, and specifically to that of hypothalamic endocrine control. Late in the nineteenth century anatomists used the Mallory trichrome staining method to observe globules in neurons that resembled those seen in thyroid cells. It was subsequently posited that these globules were secretory and hormonal in nature (Scharrer and Scharrer, 1940; Bargmann, 1949; Scharrer, 1952). Later observations of such globules and granules in axon pathways leading from large cells in the hypothalamus into the vascular posterior pituitary led to Harris’s prediction that neural products were transported via a vascular route to influence pituitary function (Harris, 1948; Harris, 1955). Oxytocin and vasopressin are nonapeptides synthesized from similar neurophysin precursor molecules transcribed from different genes—they differ in sequence by only one amino acid residue out of the nine. They are contained in closely related but separate populations of magnocellular neurons in PVH, supra optic nucleus (SON), and small accessory nuclei (i.e., nucleus circularis) scattered between them. VAS, known as anti- diuretic hormone, is released into the general circulation from axons terminating on blood vessels in the posterior pituitary. Its primary hormonal effect is on fluid balance physiology in the kidney. OXY is released in a similar fashion and exerts general control over smooth muscle (e.g., in blood vessel walls, mammary gland and uterus), and salt retention physiology in the kidney. Both of these hormones have been implicated as effectors in neural circuits controlling pair bonding and parental affiliative behaviors (Young et al., 1998). Also, in addition to the well-known coordination of firing and pulsatile axonal release during 201 the milk ejection reflex (see for example Hatton, 1983), OXY neurons have been shown to exhibit dendritic release of peptide, which might function as a local-feedback neurotransmitter (Ludwig and Pittman, 2003). However, these very interesting subjects are beyond the scope of discussion for the work presented here. In PVH, the anterior and medial magnocellular subdivisions (PVHam and PVHmm) are characterized by presence of OXY neurons. The posterior and lateral magnocellular subdivisions (PVHpmm and pml) are in general seen as a globular mass of mostly OXY cells, enclosing (especially in the posterior division) a central ball of neurons that are almost exclusively VAS-expressing. Thus, there are more OXY- than VAS-expressing cells in PVH. As noted earlier, OXY and VAS are virtually never seen co-localized in the same neuron, except in extreme physiological conditions of altered fluid balance. In that case they are only detectable by very sensitive molecular techniques (see for example Xi et al., 1999). Thus, no significant amount of co-localized staining was expected in the work presented here (see Appendix IV for preliminary antibody staining and co-localization testing). For that reason (and to use as few exemplar animals as possible for data analysis), OXY and VAS cells were stained in the same set of sections (PVH12B) and data for each of them are presented in alternating figures below. As in Chapter 11 showing CRH results, figures of Atlas Level views below are followed by companion figures showing original data sections from which the Atlas Level views were derived. They detail the data-to-Atlas mapping technique used for PVH12, because it was the experiment whose plane of section deviated most from that of the Atlas. 202 OXY&VAS: ne, and non-ne cell type distribution in sagittal view: Fig. 12.01 Neuroendocrine and non-ne OXY cell type distribution throughout PVH. Distribution in general is as expected, with neOXY concentrated in the area of PVHpmm and pml, and rostrally in areas corresponding to PVHam and mm. However, there appear to be many more cells outside those subdivisions than previously noted. Non-neOXY cells are seen mostly in the caudal descending compartment, as expected. In addition, a few are located rostrally near neOXY cells. 203 Fig. 12.02 OXY cell type distribution throughout PVH, all peptide-positive cells. This combined image emphasizes the apparent intermixing of non-ne with neOXY cells in rostral PVH. Distribution is fairly separate in caudal (descending functional compartment) PVH. Also emphasized in this image is distribution of at least a few neOXY cells in virtually all parts of PVH. 204 Fig. 12.03 Neuroendocrine and non-ne VAS cell type distribution throughout PVH. Neuroendocrine VAS cell distribution, as expected, is predominantly in the central magnocellular area, with a few cells scattered rostrally and caudally. Only six caudal non-neVAS cells are seen. They are scattered among very few neuroendocrine cells at the caudal part of the main magnocellular cell mass. 205 Fig. 12.04 OXY cell type distribution throughout PVH, all peptide-positive cells. This combined image emphasizes separation of neVAS from the very few non-neVAS cells seen. In this projection non-ne cells appear to be in the (caudal) descending functional compartment of PVH. 206 OXY and VAS: ne, and non-ne cell type distribution on Brain Maps Atlas Levels: Fig. 12.05 PVH subdivisions at Brain Maps Atlas Levels, reference the following figures. 207 Fig. 12.06 OXY and VAS cell type distribution in caudal PVH on Atlas Level 22. Note that there are no VAS cells present at this level, in agreement with the general observation that the am subdivision contains magnocellular neurons (magnos) of the OXY phenotype. 208 Fig. 12.07 OXY and VAS cell type distribution in rostral PVH, Atlas Level 22 & caudally. Note that at this level there are no VAS cells. 209 Fig. 12.08 OXY and VAS cell type distribution in caudal PVH on Atlas Levels 23 and 24. At these levels there are very few VAS cells, but quite a number of both ne and non-ne OXY cells. Many ne and non-ne OXY cells are distributed in ap, caudal to expected distribution of OXY magnos in am, seen on Level 22. The cluster of neOXY cells in and around mm is consistent with the literature. 210 Fig. 12.09 OXY and VAS cell type distribution in rostral PVH, Atlas Level 23 & caudally. There are significant numbers of non-ne OXY, but not VAS cells at this level. Thus, open circles represent non-ne OXY cells, while gray circles represent the very few VAS cells. 211 Fig. 12.10 OXY and VAS cell type distribution in rostral PVH, Atlas Level 24 & caudally. Only ne cells are present, thus OXY is represented by black circles and VAS by open circles. 212 Fig. 12.11 OXY and VAS cell type distribution in caudal PVH on Atlas Level 25. Only neuroendocrine cells of either phenotype are present at this level. Oxy cells are seen in PVHpv and PVHmpd, in addition to expected dense distribution in PVHpmm. VAS distribution, as expected, is seen mostly in PVHpmm. 213 Fig. 12.12 OXY and VAS cell type distribution in central PVH, Atlas Level 25 & caudally. This combined image makes clear the localization of VAS neurons as a predominant cell type within a surround of OXY cells, in general agreement with the literature. In the caudal section, neOXY cells are evident in PVHdp, a subdivision known to contain large numbers of mostly spinal cord-projecting cells. 214 Fig. 12.13 OXY and VAS cell type distribution in caudal PVH on Atlas Level 26. At this central level of PVHpml, the well-known OXY ‘shell’ surrounding a dense collection of VAS cells is apparent. Also seen is unexpected distribution of both OXY and VAS cells in PVHmpd, an area considered to consist of mostly parvicellular neuroendocrine cells. OXY cells are also seen in PVHpv. 215 Fig. 12.14 OXY and VAS cell type distribution in central PVH, Atlas Level 26 & caudally. As with Level 25, the juxtaposition of OXY and VAS cells within PVH are seen more clearly in these sequential section maps. Data sampled from three different sections show similar distribution patterns. 216 Fig. 12.15 OXY and VAS cell type distribution in caudal PVH on Atlas Level 27. At this caudal level, OXY cells are seen in subdivisions usually considered to contain parvicellular neuroendocrine cells. NeOXY cells are medial, at the border of PVHpv and mpd, while non-neOXY are central, to lateral parts of PVHmpd. Two neVAS cells are in ventral PVHpv and one of the very few non-neVAS cells is in PVHmpd—apparently at the dorsal edge of the group of non-neOXY cells. 217 Fig. 12.16 OXY and VAS cell type distribution in caudal PVH, Atlas Level 27 & caudally. Locations of OXY cells in comparison with VAS cells noted in the Atlas Level 27 above are confirmed in these images of sequential sections sampled for the Atlas Level figure. Distribution patterns of both peptides are not significantly different in the three sections compared to the meta-data Atlas figure. 218 Tables of PVH subdivisions showing OXY & VAS cells: Table of PVH subdivisions, with OXY cells in Atlas Level data presented here Atlas sub- division pv ap am mm pmm pml dp mpd mpdl mpv lp f Atlas lev: ne / non 22: 10/0 22: 1/0 22: 4/0 23: 5/2 23: 50/13 Summary data from 24: 22/0 24: 7/0 24: Atlas Levels 25: 7/0 25: ?/? 25: 131/0 25: 0/0 25: 12/0 25: 0/0 25: 0/0 only, as seen in 26: 14/0 26: 77/0 26: 4/0 26: 48/0 26: 0/0 26: 5/0 preceding figures 27: 4/0 27: 9/19 27 0 27 0 Sub- division pv ap am mm pmm pml dp mpd mpd (lat) mpv lp f neuro- endocrine 62 58 4 ? 131 77 4 69 0 5 0 0 non-ne 2 13 0 ? 0 0 0 19 0 0 0 0 Table 12.01: Data from this work, PVH subdivisions with identified Oxytocin cells. Table of PVH subdivisions, with VAS cells in data presented here subdivision pv ap am mm pmm pml dp mpd mpd(lat) mpv lp f Atlas Level: ne / non-ne 22: 0/0 22: 0/0 23: 2/0 Summary Data from 24: 1/0 Atlas Levels only, as 24: 0/0 25: 63/0 25: 0/0 25: 0/0 illustrated in preceding 26: 169/0 26: 0/0 26: 35/0 26: 0/0 26: 0/0 figures 27: 2/0 27: 0/0 27: 0/0 subdivision pv ap am mm pmm pml dp mpd mpd(lat) mpv lp f Neuro- endocrine 2 3 0 0 63 169 0 35 0 0 0 0 non Neuro- endocrine 0 0 0 0 0 0 0 0 0 0 0 0 Table 12.02: Data from this work, PVH subdivisions with identified Vasopressin cells. 219 Chapter 13. Composite View: all Neuroendocrine cell types Surveyed Overview: Widespread intermixing of cell types was apparent after the first round of analysis, where composite comparisons were made by interleaving layers of sequential-section data maps in Illustrator™ files, as shown below in “Examples of combining data used in the first round of analysis“. Figures 13.01 - 13.03 are a reprise of figures 5.09 - 5.11 from Chapter 5 (Mapping Strategy). They show data from different animals and series compared together. These can be compared for consistency with subsequent figures depicting data from the exemplar series (taken from three animals) showing atlas-adjusted composite views. Further rounds of analysis provided more information about extent of distribution and intermixing among cell types. New analyses facilitated display of data from different animals in accurate anatomical relationship based on standardized projection templates of PVH outline from Brain Maps rat brain atlas. Projection images of PVH showing all data from an entire experimental series in a single image were valuable in appreciating rostro-caudal and dorso-ventral distribution of each cell type surveyed. Recall Chapter 6 (PVH 3D Atlas) and the initial figures in results Chapters 8-12. The method of meta-analysis devised to display data from different series or different animals, all in the same anatomical coordinate system, allows a more accurate assessment of cell type distribution within PVH subdivisions. Meta-analysis methods were described in Chapter 7 (Meta-Data Mapping Strategy) and shown in detail for individual cell types in results Chapters 8-12. Further, as illustrated below, data from different series can be combined with great accuracy to yield a composite image of all neuroendocrine cell types in specific PVH subdivisions at a given Atlas Level. Comparisons made between cell types in the first round of analysis can thus be more rigorously evaluated. All cell types surveyed 220 may be viewed simultaneously, as if in the same section—to better appreciate anatomical relationships between them. In “All neuroendocrine cell types surveyed: distribution on Brain Maps Atlas Levels”, illustrations of all neuroendocrine cell types displayed together in anatomical register are shown for each Brain Maps Atlas Level. Figures 13.06, 13.08, 13.11, 13.13, and 13.15 are prepared in black and white according to the requirements for the Ph.D. dissertation manuscript format. (Thus, original different-colored symbols were converted to different- shape black and white symbols— retain information when the dissertation is archived with microfiche technology.) The patterns of cell type intermixing and juxtaposition are much more obvious when the data are viewed in multi-color format. Therefore, each black and white figure is followed by its corresponding color version (figures 13.07, 13.09, 13.10, 13.12, 13.14, and 13.16). Here, color symbols for parvicellular cells are displayed alone in the left panel for clarity, and color symbols for all neuroendocrine cells together are displayed in the right panel. Thus the right panel of a color figure is equivalent to its preceding black and white figure. Illustrations 13.05 – 13.16 are presented in landscape (sideways) format, to allow maximum viewing size within the restricted page margins defined for the dissertation manuscript. The section “Tabulation of all cells recorded in Atlas Level figures” is a table showing cumulative cell counts (tabulated as both neuroendocrine and non-neuroendocrine) from all Atlas Levels surveyed in the previously illustrated cumulative figures. 221 Examples of combining data used in the first round of analysis: Fig. 13.01 Composite file: different animals stained with same combination of antibodies. Two examples of maps from different animals—stained with the same combination of antisera—at ‘classic’ PVH, equivalent to Brain Maps Atlas Level 26. Panel a: PVH10C stained for OXY and CRH. Panel b: PVH12D stained with the identical combination of antisera to OXY and CRH. Panel c: stacked, "best fit" composite, showing reproducibility of staining patterns between animals using the same combination of antibody pairs, but at higher effective density (15µm+15µm=30µm). 222 Fig. 13.02 Composite files: “best fit” similarly stained map data from different animals. Two examples of maps from different animals—stained similarly—at ‘classic’ PVH, equivalent to Brain Maps Atlas Level 26. Panel a: one section from PVH10D series stained with antibodies to TRH and SS. Panel b: two adjacent sections from PVH14, one stained for TH with TRH, the other stained for TH with SS. Panel c: stacked, "best fit" composite showing reproducibility of staining patterns for TRH and SS between animals. 223 Fig. 13.03 Composite files: “best fit” data: different animals, one antibody in common. Two examples (a, b) of maps from different animals—2 cases, stained with one antibody (CRH) in common—at ‘classic’ PVH, equivalent to Brain Maps Atlas Level 26. Panel a: PVH12C stained for VAS and CRH. Panel b: PVH14C stained for CRH and TH. Panel c: stacked, "best fit" composite, showing reproducibility of staining patterns for a single target cell type (CRH, in this case) between animals. 224 All neuroendocrine cell types surveyed: distribution on Brain Maps Atlas Levels: Fig. 13.04 PVH subdivisions at Atlas Levels, for reference in the following figures. 225 Fig. 13.05 Composite view neuroendocrine cell types, Brain Maps level 21(no cells). 226 Fig. 13.06 Composite view neuroendocrine cell type staining, Brain Maps level 22. Although PVH extends into Atlas Level 21, according to the new 3-D model derived from serial sections, the first examples of staining for the cell types surveyed were seen at this level. 227 Fig. 13.07 Composite color view neuroendocrine cell type staining, Brain Maps level 22. 228 Fig. 13.08 Composite view neuroendocrine cell type staining, Brain Maps level 23 & 24. 229 Fig. 13.09 Composite color view neuroendocrine cell type staining, Brain Maps level 23. 230 Fig. 13.10 Composite color view neuroendocrine cell type staining, Brain Maps level 24. 231 Fig. 13.11 Composite view neuroendocrine cell type staining, Brain Maps level 25. 232 Fig. 13.12 Composite color view neuroendocrine cell type staining, Brain Maps level 25. 233 Fig. 13.13 Composite view neuroendocrine cell type staining, Brain Maps level 26. 234 Fig. 13.14 Composite color view neuroendocrine cell type staining, Brain Maps level 26. 235 Fig. 13.15 Composite view neuroendocrine cell type staining, Brain Maps level 27. 236 Fig. 13.16 Composite color view neuroendocrine cell type staining, Brain Maps level 27. 237 Tabulation of all cells recorded in Atlas Level figures: Super-Table of cells seen in PVH subdivisions, on Atlas Levels 22-27 Table of all ne cell type distributions, on Atlas Levels, cumulative - by subdivision subdivision pv ap am mm pmm pml dp mpd mpdl mpv lp f neSS 151 44 1 6 5 0 3 27 0 0 0 0 neGRH 0 0 0 0 0 0 0 1 0 0 0 0 neTH 3 2 1 0 1 1 0 1 0 0 1 0 neTRH 39 1 3 0 5 0 1 74 0 2 1 0 neCRH 28 16 3 0 44 11 0 243 3 0 0 0 neOXY 62 58 4 11 131 77 4 69 0 5 0 0 neVAS 2 3 0 0 63 169 0 35 0 0 0 0 Table of all non-ne cell type distributions, on Atlas Levels, cumulative - by subdivision subdivision pv ap am mm pmm pml dp mpd mpdl mpv lp f non-neSS 7 5 0 1 0 2 0 1 0 0 1 0 non-neGRH 0 0 0 0 0 0 0 2 0 0 0 0 non-neTH 5 12 0 0 1 0 0 3 0 0 1 0 non-neTRH 24 91 36 1 4 1 3 20 0 4 1 0 non-neCRH 2 13 0 0 4 2 1 7 0 0 1 0 non-neOXY 2 13 0 0 0 0 0 19 0 0 0 0 non-neVAS 0 0 0 0 0 0 0 0 0 0 0 0 Table 13.01: Data from this work, PVH subdivisions with identified cells on Atlas Levels. Total cells seen of a given phenotype were noted for each individual Atlas Level (Fig. 13.04-13.16, above). This table shows cumulative number of cells present within a given PVH subdivision from all six Atlas Levels illustrated. One can gain an appreciation of the major concentration of each cell type, by PVH subdivision. Note, however: this table shows data from a single 15mm section at each Atlas Level, and the levels are not equidistant from each other. Therefore, no quantitative estimates can be derived here for total numbers of cells of a given phenotype throughout the full rostro-caudal extent of each PVH subdivision. Names on the two bottom rows are offset to indicate magnocellular neurons. 238 Chapter 14. Staining Reproducibility and Co-Localization of Peptides Overview: One advantage of staining with combinations of two different antisera in tissue with retrogradely identified neuroendocrine neurons is the ability to identify cells co-expressing the same chemical phenotype, whether neuroendocrine or non-neuroendocrine. As noted previously in Chapter 4, antibodies for all peptide phenotypes surveyed were combined in different sets of two in various experiments. Thus, each peptide was stained in comparison with every other more than once, in the same or different animals. Since only three sets of sections plus a Nissl-stained set were available per animal, not all peptides could be surveyed at the same time. An effort was made to keep one peptide in common among at least two of the three sets of sections, staining it against others to determine possible co-localization and patterns of intermixing within a single animal. Slight differences in fixation or antibody staining-intensity that produced qualitative differences in overall staining could thus be compared between animals by using the same antibodies. Alternatively, similar comparisons could be made by the use of several different primary antibodies to the same or different peptide targets in the same animal. Some examples showing results of this staining strategy are illustrated below. In general, surprisingly few instances of co-localization were seen in the primary neuroendocrine cell phenotypes surveyed in these studies. This is different from the well- documented co-localization of almost all these peptides with other proteins found in PVH and other hypothalamic neurons. Staining Reproducibility Reproducibility of staining is demonstrated by using the same antibody to stain different set of sections in the same animal. Alternatively, a comparison can be made of staining using the same antibody in different animals, as shown in figures below. 239 Fig. 14.01 neSS cell type distribution throughout PVH, in different experimental animals. The general pattern of neSS cell distribution is qualitatively similar between animals, though actual numbers of stained cells varies from case to case. This is one method of confirming staining consistency for a single antibody no matter the specific fixation or processing conditions. 240 Fig.14.02 TH (dopamine) staining-consistency, in three serial sets from the same case. The staining pattern is similar, but not identical, in sets of serial 15µm sections from the same animal stained at different times. This is easily seen in a cell type with low density such as TH, but the same holds true for staining patterns seen for other cell types. Cell density and distribution are qualitatively similar for the same cell type among sections in the same animal and between different animals. In these thin serial sections, one might expect an almost identical distribution for a consistently localized cell type, if it in fact existed. While not quantitatively the same, these consistent patterns are however representative of expected biological similarity between experimental animals. 241 Cell types that showed co-localization with each other: There were few cells showing co-localization of SS with another target peptide surveyed in these studies, and those were only with TRH. For example, a single cell co- localizing neSS and neTRH at Atlas Level 22 is shown in figure 8.06, followed by a single co- localized cell in each of the following sections of the set shown in fig. 8.12. One non-ne cell co-localizing SS and TRH was seen, also near Level 22 as shown in figure 8.12. A few instances of TRH/CRH co-localization seen in pilot studies (recall figure 5.02) were very intriguing, however they were not reproduced using different TRH antiserum in later studies. None of the very few GRH cells stained in PVH (see figures 8.04 and 8.11) showed co- localization with any of the other peptides studied. Co-localization staining in was never seen in neuroendocrine TH cells. However, two non-ne TH/TRH cells were seen: one in the section caudal to Level 26 (recall figure 9.14 for TH cells), the other in the last section of PVH from the same series. No co-localization was seen with antisera to OXY, except for assumed false-positive cross-reactive staining described more fully below. As expected from previous studies and reports in the literature, some CRH cells were seen to co-localize VAS staining. VAS is a known co-secretagog for ACTH, expressed in CRH cells that have been stimulated maximally so that corticosterone (CORT) levels in blood are elevated. (Corticosterone is the rat steroid hormone equivalent to human cortisol, which is released from the adrenal cortex in response to ACTH.) Thus, recruitment of VAS synthesis in these parvicellular neurons provides a measure of plasticity in the HPA stress axis so that an acute stress response can still be mounted under conditions of chronic high CORT. The number of neCRH/VAS cells seen in these studies was low, perhaps because CORT levels in blood due to colchicine stress were not high enough to stimulate recruitment of VAS expression. Figure 14.03 is a sagittal projection of co-localization staining from PVH12C, where only seven neuroendocrine and two non-neuroendocrine cells were seen to 242 co-localize CRH and VAS staining. Most of these were located in mid-PVH about at the level of PVHmpd, as expected. An example of apparent CRH co-localization illustrates a phenomenon sometimes encountered in immunohistochemical staining, that of non-specific cross reactivity of antisera to non-target antigens (see figure 14.04). Antiserum to CRH produced in sheep (shCRH, from Wylie Vale, The Salk Institute) was used with various antisera grown in rabbit for double-staining cocktails. Testing of this shCRH against rabbit-CRH and mouse/rat monoclonal-CRH (mAbCRH, from F.J. Tilders, Amsterdam) showed that it reproducibly stained all authentic CRH-containing cells (see Appendix IV). However, this shCRH antiserum was known to cross react non-specifically with rabbit anti-GRH (Appendix IV, and Newton Canteras, personal communication). As seen in figure 14.04, this shCRH antiserum showed co-localized staining with some of the cells labeled with rabbit antisera to OXY or VAS. Complimentary tests of shCRH with mouse monoclonal antibodies to VAS or of mouse/rat monoclonal antibodies to CRH with the rabbit antiserum to OXY (in series from this and other animals) did not show this pattern of co-localization. Thus, it was determined that for this experiment the shCRH cell staining was authentic while the co-localization with OXY and VAS was not. 243 Fig.14.03 Small numbers of cells showing CRH/VAS co-localized staining. PVH12C (CRH results shown in Chapter 11) shows very few cells co-localizing CRH and VAS. Fig.14.04 Apparent OXY and VAS co-localization staining with shCRH antisera shCRH was used in combination with rabbit antisera to OXY or VAS in two different animals. At left is an example of staining from the same animal shown above. With the shCRH antiserum, atypical staining in some CRH positive cells also appears to co-localize with OXY. Using shCRH with rabbit antiserum to VAS in a different animal (PVH12) produces a similar result, but with even more apparent co-localization staining. PVH12 was the animal used for data collection in Chapter 12: OXY and VAS. 244 All rabbit antisera tested, but one, were compatible with the shCRH antiserum. The sheep anti-CRH was thus used in an antibody cocktail for several dual-staining experimental sets. After complete analysis, the shCRH antibody was seen to react with some but not all cells co-stained for VAS or OXY, as shown in figure 14.04 and discussed above. Thus, in those specific cases the apparently high number of CRH co-localized cells was interpreted as a false positive result. By comparative staining, it was shown that individual staining for either of the cell types alone was consistent with cell numbers and distribution seen using other antisera. Thus, the single-staining CRH data maps from the shCRH stained sets were deemed valid for inclusion in the experimental results, while the erroneous co-localization data were not. They do, however present an interesting illustration of one of the potential problems inherent in the use of antibodies as experimental tools. Extensive testing such as that performed at the early stages of this work (see Appendix IV) is again seen to be necessary and useful. 245 SECTION IV: DISCUSSION AND CONCLUSIONS Chapter 15. Discussion of Results Overview: The work presented here was originally intended to produce a high-resolution account of spatial distribution and relationship between the primary neuroendocrine, or hypophysiotropic, neurons in PVH. This was accomplished by staining many high-resolution sets of sections with antibodies to phenotypic markers in animals treated to define hypophysiotropic neurons. In the process of detailing locations of all cell types surveyed, methods of mapping were devised to accurately present these data in the context of a published reference: Swanson’s Brain Maps: Atlas of the Rat Brain. Thus, two different but inter-related types of result are presented: 1). Development of mapping methods that are useful for data representation and integration with established literature. These, in future, will facilitate accurate recording of experimental data within a true 3D representation of PVH when it becomes available (reference Chapters 6 and 7). 2). Presentation of cumulative data showing spatial distribution of SS-, GRH-, TH-, TRH-, CRH-, OXY- and VAS-expressing neurons in PVH (reference Chapters 8-12). Presentation of data including reproducibility of staining, co-localization of peptides in the same cell, and patterns of intermixed cell types presented on Brain Maps Atlas Levels (reference Chapters 13 and 14). Results presented here comprise the first comprehensive, high-resolution survey of the primary hypophysiotropic cell types (in relation to each other) within PVH—all in the context of a published anatomical atlas standard. Details of mapping technology, data distribution, and its significance are discussed again below, under topic headings reflecting the order in which they were presented in previous chapters. 246 Mapping Strategy, Creation of 3-D PVH model, and Meta-Analysis methods: PVH neuroendocrine cell staining data, prepared at very high resolution (15µm frozen sections, sectioned as 1-in-4 series throughout PVH) were mapped by hand onto drawings from photomicrograph images. All drawings were re-confirmed by microscopic observation of the original sections. They were then transferred into Adobe Illustrator™ graphics files, where the position and character (i.e., peptide content, neuroendocrine status) of each mapped cell was recorded with a specific identifying symbol. (This data transfer was re-confirmed once again by microscopic observation of the original material.) A valuable feature of the Illustrator™ graphic program allows recording and separation of different types of information from the same map into different layers (overlays) above the original map image. Thus, very accurately mapped data elements could be grouped, colored for easy identification, and hidden or viewed at will to appreciate various elements alone or in combination. Further, this type of data recording allowed manipulations to combine and contrast data from many sections simultaneously (in accurate anatomic register), and to “interleave” data from more than one experiment into a single composite file for analysis and comparison. This method of data recording and display allowed a dynamic viewing of PVH chemical neuroanatomy never before appreciated. Early analysis using such map combinations produced interesting new information about the distribution and intermixing of neuroendocrine cell types in PVH. Significant numbers of most cell types surveyed were seen more rostrally than expected, and some cell types (e.g., OXY, CRH) were more intermixed in various PVH subdivisions than previously appreciated. After obtaining a rich and complicated data set produced at very high resolution, it became obvious that a similarly high-resolution standard reference atlas was needed. In collaboration with Dr. Larry W. Swanson, author of Brain Maps: Structure of the Rat Brain, a registered serial-section reconstruction through the full extent of PVH was prepared from the 247 complete set of Nissl sections of the brain used for the original Brain Maps atlas. These drawings of PVH subdivision outlines, based on cytoarchitectonic features, together approximate a 3-D reconstruction of PVH within the Brain Maps atlas. This is, of course, a registered set of accurate anatomical outline drawings that define limits of PVH subdivisions rather than a true 3-D atlas of PVH. (A true 3-D atlas might be comprised of high-resolution, voxel-based data that could be mathematically manipulated and/or re-sliced for viewing at different angles.) Since anatomical data are largely visual, this graphical reconstruction of PVH subdivision outlines nonetheless proved extremely useful. It helped in forming mental models of PVH anatomy and in devising practical new methods for analyzing data. Eventually, this 3-D PVH atlas representation may inform new models proposed to explain mechanisms of PVH function. Viewing the serial PVH drawings as stacked slices to prepare accurate projection outlines allowed a new appreciation for detailed PVH anatomy. Certain PVH substructure visible in the serial Nissl-stained sections was redefined, with some modified subdivision outlines subsequently incorporated into the third edition of Brain Maps (Swanson, 2004). For example, the rostral limit of PVH is farther forward, and the anterior part extends more dorsally than originally appreciated. An area labeled “ps” on Level 21 in the original Brain Maps (Swanson, 1992) is now seen to be the rostral pole of PVHap. An important new observation was the shape and size of a lateral extension of the medial parvicellular subdivision. An area of intensely stained neurons previously thought to be magnocellular (associated with posterior pituitary function) was determined rather to be an extension of the medial parvicellular division (thus, PVHmpdl or alternatively PVHmpd-l) associated with anterior pituitary control. This observation was important in final analysis of the work presented here. It is evidenced by the illustration of very lateral neCRH neurons within PVHmpdl, shown in figures 11.11, 11.12, 13.11, 13.12 and noted in Table 15.02 and Figure 248 15.05 below. When confirmed by additional studies and ongoing work in the Swanson laboratory (Joel Hahn, personal communication), this new appreciation of the lateral extent of PVHmpd has important implications for analysis of neural input data to hypophysiotropic cells in PVH. Accurate assessment of the full extent of parvicellular neurons is crucial when considering larger issues such as the role of PVH in complex neural circuits (chemical and connectional) involved in behavioral control. An important result derived from the 3-D PVH sub-project was production of two new methods for analyzing very high-resolution experimental data. The first was creation of anatomically accurate sagittal and dorsal projection outlines of PVH, registered within the coordinates of the Brain Maps atlas. A method was devised to accurately transfer large sets of data (from many rostral-to-caudal frontal sections) onto a single projection outline orthogonal to their original plane of section. Thus, all the data from an entire series of sections throughout PVH (30 or more individual maps of cell staining distribution) could be viewed in a single image. New information was gleaned at a single glance, in this summary from dozens of difficult-to-remember individual data maps. A new appreciation of the rostro- caudal (and in sagittal view, the dorso-ventral) distribution, and the relative density of specific neuroendocrine cell types within PVH was gained. The relationship of cell type distribution from very rostral to very caudal throughout the nucleus could now be compared and contrasted in a manner never before possible. A second analytic method, one for anatomically accurate meta-analysis, was produced from the 3-D PVH atlas model. Thereafter ensued a third major round of analysis of the rich data set contained in many hundreds of accurate anatomical maps of neuroendocrine cells in PVH. Using fiducial structures observed in Nissl sections and “Interbrain Section Plane” information from Brain Maps, 2 nd Ed. computer files (Swanson, 1998), a method was devised to determine accurately the plane of section of experimental 249 material in reference to the atlas. This is a major advance for analyzing experimental material, or in fact any histological brain sections, because differences in plane of section in individual brains are unavoidable. Data can easily be misinterpreted if transfer-interpolation from experimental material to reference atlas anatomy is not accurate. Almost inevitably, analysis distortions can be introduced due to differences in plane of section. These become more extreme with increasing frequency of sampling and decreasing size in the structure of interest—such as PVH. With a reliable method in hand to determine the relationship of experimental data to the reference atlas in 3-D coordinates, accurate composite data maps were made using Brain Maps Atlas Level templates. Anatomically accurate data on these Atlas Level maps show complex results from different experiments as if seen together in the plane of the standard atlas brain. Those results are reviewed in the sections below, with reference to color figures for each Atlas Level—each an informative reprise of black and white figures presented earlier. Two super-tables of results (tabulated in different ways for all cell types surveyed in chapters 8-12) are included as an overview of individual and combined cell type distribution. Discussion of Data-Results by Cell Type, as detailed in Chapters 8-12: In the following sections general staining patterns seen for all cell types studied are briefly reviewed, each discussed in the order shown in Results Chapters 8-12. Though the emphasis of this work was description of the distribution of neuroendocrine cell types, findings for both neuroendocrine and non-neuroendocrine cells are discussed, and color illustrations of all cell type data shown together (both ne and non-ne views) are presented in figures 15.01 through 15.07. Data distribution is discussed first in terms of the sagittal projection (single image) analysis of an entire experimental set (as illustrated in previous chapters). Then findings within individual PVH subdivisions at each Atlas Level of the meta- analysis data plots are discussed, as illustrated in the color figures. Tables 15.01 and 15.02 250 present tabulated data on numbers of cells from all the figures in this and previous results-chapters, for reference in subsequent discussion. Table 15.01 shows cell counts for an entire series, and at each Atlas Level for all cell types surveyed. Table 15.02 shows cell counts in PVH subdivisions (totals from Atlas Levels only) for each cell type surveyed. Final sections in this chapter will discuss cell type intermixing and peptide co- localization: from combination maps of all neuroendocrine cell type data illustrated together on PVH Atlas Levels (recalling figures from Chapter 13, and with reference to the color figures in this chapter). As noted previously, specific data discussed here are those from an exemplar set for each peptide surveyed. Thus, total numbers and distribution of cells detailed are from a single animal for each cell type. Where possible, cell types were surveyed in adjacent sections from the same animal. However, it is not technically feasible to survey all cell types in a single animal at the resolution used here (1-in-4, 15µm sections). Therefore, collective data presented are from a total of three animals. They represent consistent distribution patterns observed in all other cases surveyed (see Appendix IX). 251 Stained cells: Sagittal Projection view, And on Atlas Levels 22-27 Total cells sag. view Cell Type surveyed Atlas Level 22 Atlas Level 23 Atlas Level 24 Atlas Level 25 Atlas Level 26 Atlas Level 27 787 neSS 8 30 48 131 12 10 120 non-neSS 2 1 5 3 3 3 2 neGRH 0 0 0 0 0 1 6 non-neGRH 0 0 0 0 0 4 48 neTH 0 2 0 2 4 1 123 non-neTH 2 3 1 3 3 3 510 neTRH 2 4 3 44 44 28 543 non-neTRH 27 48 41 18 12 8 1696 neCRH 7 11 10 78 204 19 125 non-neCRH 2 9 2 10 2 5 1447 neOXY 15 55 29 150 148 13 166 non-neOXY 0 15 0 0 0 24 758 neVAS 0 2 1 63 204 2 6 non-neVAS 0 0 0 0 0 1 Table 15.01: Cell type staining, numbers of cells per set, numbers at each Atlas Level. The first column shows all cells of a given phenotype stained in a complete 1-in-4, 15µm series of sections from a single animal, as illustrated in chapters 8-12. In contrast the last six columns are cell counts for a given phenotype (represented by one 15µm section) at each individual Atlas Level. Stained cells in PVH subdivisions (only) on Atlas Levels 22-27 Subdivision=> pv ap am mm pmm pml dp mpd mpdl mpv lp f neSS 151 44 1 6 5 0 3 27 0 0 0 0 neGRH 0 0 0 0 0 0 0 1 0 0 0 0 neTH 3 2 1 0 1 1 0 1 0 0 1 0 neTRH 39 1 3 0 5 0 1 74 0 2 1 0 neCRH 28 16 3 0 44 11 0 243 3 0 0 0 neOXY 62 58 4 11 131 77 4 69 0 5 0 0 neVAS 2 3 0 0 63 169 0 35 0 0 0 0 Subdivision=> pv ap am mm pmm pml dp mpd mpdl mpv lp f non-neSS 7 5 0 1 0 2 0 1 0 0 1 0 non-neGRH 0 0 0 0 0 0 0 2 0 0 0 0 non-neTH 5 12 0 0 1 0 0 3 0 0 1 0 non-neTRH 24 91 36 1 4 1 3 20 0 4 1 0 non-neCRH 2 13 0 0 4 2 1 7 0 0 1 0 non-neOXY 2 13 0 0 0 0 0 19 0 0 0 0 non-neVAS 0 0 0 0 0 0 0 1 0 0 0 0 Table 15.02: Cell type staining, numbers of cells (by subdivision) on Atlas Levels only. Note these cell counts represent data from a 15mm section at each Atlas Level, and the levels are not equidistant from each other. Therefore, no estimates of total cells throughout the rostro-caudal extent of a given subdivision can be made. 252 Fig.15.01 Cumulative figure (color) of all cell types stained at Atlas Level 21. No cells were seen in PVH at this level. It is included for reference on PVH position in the brain (see sagittal vignette at bottom right) and to preview symbols used for all cells. 253 Fig.15.02 Cumulative figure (color) of all cell types stained at Atlas Level 22. 254 Fig.15.03 Cumulative figure (color) of all cell types stained at Atlas Level 23. 255 Fig.15.04 Cumulative figure (color) of all cell types stained at Atlas Level 24. 256 Fig.15.05 Cumulative figure (color) of all cell types stained at Atlas Level 25. 257 Fig.15.06 Cumulative figure (color) of all cell types stained at Atlas Level 26. 258 Fig.15.07 Cumulative figure (color) of all cell types stained at Atlas Level 27. 259 Somatostatin and Growth Hormone Releasing Hormone Staining: As with all cell types studied, there were no positively stained SS cells in the most rostral part of PVH—the first examples on an Atlas Level were seen at Level 22 (recall figure 8.06, and see figures 15.01 and 15.02, above). Somatostatin staining (total 787 ne, 120 non-ne cells in sagittal projection) was observed, as expected, mostly in the periventricular subdivision of PVH. More than two thirds of neSS cells were in PVHpv, however distribution was not uniform throughout the subdivision. In sagittal projection view (recall figures 8.02, 8.03), cells were most dense in central PVH with a suggestion of higher concentrations at the dorsal and ventral limits of that central area. In addition, there were clearly large numbers of more sparsely distributed neSS cells in rostral and caudal parts of PVH. Many non-ne cells (120 out of 907 total) were also seen in sagittal view. These non-neSS cells, notably less dense in central PVH (the location of most neSS cells), were found scattered among the neSS cells in rostral and caudal PVH. In coronal Atlas Level views (recall figures 8.06 - 8.10 in Chapter 8, and see color figures 15.01-15.06 above), neSS cells were seen in rostral sections to form a vague band, extending diagonally upward from medial to lateral in the anterior parvicellular part of PVH. This band continued caudally (increasing in density) into the medial parvicellular part, reaching highest density both in PVHpv and PVHmpd at Level 25 (figure 15.05). At that level, neSS cell distribution in PVHpv was slightly denser dorsally and ventrally, in agreement with observations from the sagittal view. By level 26, overall density of neSS cells was much lower than in the previous level, and almost all were in PVHpv (figure 15.06). This pattern continued on to Level 27 where there were still a fair number of cells, mostly in PVHpv, with a few in caudal PVHmpd (figure 15.07). In summary, neSS cells were distributed much more laterally in PVHap and PVHmpd than initially expected, and a small number of non-neSS cells were consistently 260 seen among them. They are clearly intermixed with CRH and TRH cells in PVHmpd, and are disbursed more rostrally and caudally than expected from previous published descriptions. This distribution is consistent with the idea that intermixing of neuronal cell bodies (and presumably the dendritic arbors) of somatostatin cells among CRH and TRH cells may well serve an integrative function in coordinating normal variations in metabolism and responses to physiological perturbations—especially in response to stress. Somatostatin inhibits growth hormone release, probably in a tonic fashion, in concert with thyroid control of metabolism (via TRH) and concurrent circadian and developmental patterns of growth hormone stimulation (via GRH). It has also been shown to inhibit the effects of CRH at the pituitary level, decreasing ACTH release and resultant elevation in circulating corticosteroids (Ceccatelli et al., 1989). Thus, intermixing of neSS, neCRH and neTRH cells in PVH may be one way to orchestrate metabolic coordination, and could be a major factor in effecting the well-known decrease in growth hormone release during acute and protracted stress (Patel and Srikant, 1986; Cintra et al., 1991). Growth Hormone Releasing Hormone positive cells (total 2 ne, 6 non-ne cells in a sagittal projection composite from four animals) were almost never seen in PVH, in spite of numerous staining attempts in several animals with a very well characterized antibody specific to rat GRH (recall figure 8.04). Two animals (with and without colchicine pre- treatment) were also prepared using in situ hybridization histochemistry with a probe to ratGRH mRNA (see Appendix V). Cells positive for pre-proGRH message were seen in arcuate nucleus, extending ventro-medially forward toward PVH. However, results of hybridization for GRH mRNA were virtually negative in PVH, similar to antibody staining results. Only six GRH cells identified by immunohistochemistry (two neGRH, six non- neGRH) were counted in a composite series from four animals. They were all located in very caudal PVHmpd and PVHlp. This is in agreement with some authors, but not with 261 others (as discussed previously) who reported GRH cells in ventral parts of PVH. As inferred in the discussion about SS cell localization above, relative paucity of GRH cells in PVH clearly does not mean that PVH plays no role in growth hormone release. Indeed, inhibition by somatostatin and modulation by TRH (and, under appropriate physiological conditions, by CRH) are very likely important mechanisms in hypothalamic control of circulating growth hormone levels. Tyrosine Hydroxylase (marker for Dopamine) Staining: TH cells in PVH were not seen as frequently as expected according to references in the literature. Sections stained for TH yielded the lowest cell count for a 1-in-4 series (total 48 ne, 123 non-ne in sagittal projection) of all cell types surveyed, and more than twice as many non-neTH as neTH cells were seen. In sagittal view, all but two neTH cells were in caudal PVH, distributed slightly more ventrally than dorsally (recall figures 9.01-9.02 in Chapter 9, and see color figures 15.01-15.06 above). They were spaced relatively far apart, except for a cluster of about a dozen ventral cells closer together at approximately Level 26. In contrast, non-neTH cells were distributed somewhat uniformly, though in low density, throughout PVH. Three series (serial section sets, recall figure 9.03) from one animal were stained at different times with different co-staining antisera to confirm staining pattern and consistency. Thus, low cell counts were not in error, at least for the experimental animal that was mapped in detail. Pilot studies conducted in other animals to compare different TH antisera and optimize staining showed similar results (see Appendix IV). On Atlas Level views, TH cells were seen mainly at or near the border between PVHpv and PVHmpd, with relatively more cells ventrally and caudally. There were a few cells more lateral at all levels, and these were invariably non-neuroendocrine. In general, TH was not different in distribution than expected, but density was somewhat less than has been reported previously. For example, Swanson, et al in a 1981 study of catecholaminergic cells 262 and fibers in PVH reported more than 500 TH-stained dopaminergic cells throughout the nucleus, mostly in medial and periventricular areas (Swanson et al., 1981). Assuming this was a bilateral total, then 250+ cells per side is still more than the total seen (170+) in the representative 1-in-4 series described here, although it might reasonably be assumed that this thin-section series reflect only one third to one fourth of the expected total. It is probable that earlier reports counted all TH positive cells, with no distinction between neuroendocrine or non-neuroendocrine neurons. However, the total number of cells seen in this work was still low in comparison, as noted above. In a study with similar methodology to the work reported here, Markakis and Swanson used Fast Blue to mark neuroendocrine cells. Using a different antibody than the one in these studies, they also showed low numbers of TH cells (52neTH and 13non-neTH) on Atlas Levels in PVH (Markakis and Swanson, 1997). For technical reasons (post-staining damage to some sections) the experimental animal used for TH survey in these studies was unique, in that the left rather than the right PVH was mapped. An interesting (speculative) possibility is that distribution of TH cells might differ between left and right PVH in the same animal and thus partially account for the lower than expected number of TH cells mapped. In a study of left versus right arterial baroreceptor nerve ligation, fewer TH axons (and less dopamine by chemical assay) were shown in left compared to right median eminence in normal control material (Alexander, et al. 1990). This observation was consistent with a difference in size of the corresponding aortic sinus nerves. The authors attribute arcuate nucleus origin to these axons, but it is possible a difference in number of TH axons arising from left and right PVH might contribute to it as well. However, the possibility of a significant left vs. right difference in TH cell distribution in the work presented here seems unlikely. When preliminary observations of left and right PVH were made before photographing sections for extensive mapping no qualitative difference in TH cell number or distribution pattern was seen. 263 Thyrotropin Hormone Releasing Hormone Staining: Over a thousand TRH cells (total 510 ne, 543 non-ne in sagittal projection) were mapped in a 1-in-4 series throughout PVH. Unexpectedly, more than half of the total were non-neuroendocrine rather than neuroendocrine. In sagittal view, almost all neTRH were seen in the caudal half of PVH (recall figures 10.01-10.03), while the clear majority of non- neTRH cells were distributed in rostral PVH. TRH cell distribution in the central region of the nucleus (around PVHmpd) was less dense than that seen for SS in the same area. Distribution of TRH cells (mostly neuroendocrine) was moderately dense in the area of central PVHmpd, as expected from the literature, with increased density in the caudal quarter of the nucleus. Density of non-neTRH cell distribution was relatively low in areas where many neTRH cells were seen, and about twice as dense in rostral PVH (where few if any neTRH were seen) compared to caudal PVH. In Atlas Level views (recall figures 10.05-10.09 in Chapter 10, and see color figures 15.01-15.06 above), the few rostral neTRH cells seen were quite medial, in or near PVHpv. Non-neTRH cells were widely distributed in PVHap in a diagonal band directed from medial to lateral and diagonally upward, dorsally on Level 23 and more ventrally in PVHap at Level 24. At levels 25 and 26, the “classic levels” usually used to illustrate PVH in numerous publications, neTRH cells were seen extending from PVHpv into PVHmpd, with a few cells in other subdivisions. Numbers of neTRH cells on the two levels were virtually the same but distribution was somewhat different. Thus, in part of PVH (Level 25, figure 15.05), neTRH cells are not seen predominantly in medial and dorsal PVH as reported previously. Also at this level there were many neTRH cells in the ventral three-quarters of PVHpv. A similar number of cells were seen in medial (but not dorsal) parts of PVHmpd, with a few in more lateral PVHmpd. Some neTRH cells were also seen ventrally in adjacent PVHpmm. At Level 26 neTRH cells were seen in a dense grouping in mid to ventral PVHpv, and dorso- 264 medially in PVHmpd (figure 15.06). Some cells ventro-lateral from the main group in PVHmpd were located near or in PVHmpv, a parvicellular subdivision usually seen to contain non-ne, descending-projection cells. There were more non-neTRH cells at Level 25 than Level 26, consistent with distribution seen on sagittal view. However, unexpectedly, about three times as many neTRH as non-neTRH cells were seen very caudal in PVH at Level 27. Many of these neTRH cells were in PVHmpd. In fact, at this caudal level in PVHmpd, TRH is the predominate neuroendocrine cell type. This is in contrast to the generally accepted view that most TRH cells are in central PVHmpd, concentrated medially and dorsally in the subdivision. In summary, TRH cells were much greater in number and more widely distributed throughout PVH than expected, especially in the rostral parts of the nucleus where large numbers of non-ne cells predominated. Similar distribution patterns for cells containing pre- proTRH messenger RNA were seen in companion in situ hybridization studies comparing distribution patterns between colchicine treated and normal control animals. (See Appendix V. Due to technical limitations of the in situ hybridization method, neuroendocrine status of labeled cells could not be evaluated and those results were not included in the data presented above.) On Atlas Level views many neuroendocrine TRH cells were seen in PVHmpd. However, they were not localized as expected from descriptions in the literature. Rather than being somewhat segregated in dorso-medial PVHmpd, neTRH cells (though relatively higher in concentration in some areas) were found to be well mixed with neCRH neurons. This intermixing probably affords a great advantage for integration of neuroendocrine response to various neural, hormonal, and humoral inputs. Thus important signals affecting general metabolism and physiologic response to stress have fairly equal access to hypophysiotropic neurons controlling thyroid hormone and adrenal corticosteroid output. The relatively high 265 concentration of neTRH cells in posterior parts of PVHmpd may reflect a subtly different metabolic and functional emphasis. These neurons are well positioned to integrate incoming neural information with nearby PVHdp, PVHmpv and PVHlp cells that have descending autonomic and behavioral-associated projections (Swanson and Sawchenko, 1980; Sawchenko and Swanson, 1981; Swanson et al., 1981; Sawchenko and Swanson, 1982; Cunningham et al., 1990). Thus, the caudal population of neTRH cells may be more functionally associated with metabolism-related autonomic regulation, a speculation that warrants future investigation. The significance of large numbers of predominantly non-neTRH cells in rostral PVH is not immediately apparent. It is likely that early workers who observed TRH cells in rostral PVH assumed them to be neuroendocrine, as with cells in PVHmpd documented to project to median eminence. Interestingly, a few recent studies from groups interested in TRH function have mentioned in passing that TRH cells in anterior parts of PVH are often not neuroendocrine (Sanchez et al., 2001; Wittmann et al., 2004). It should perhaps be noted that many non-neTRH cells were observed (but not documented in these studies) outside PVH, e.g., in lateral hypothalamus. Based on its distribution outside PVH, TRH (as with CRH and other peptides surveyed in these studies) very likely serves a neurotransmitter or neuromodulator function in at least some non-neTRH cells in the brain. Corticotropin Hormone Releasing Hormone Staining: Neuroendocrine CRH cells (total 1696 ne, 125 non-ne in sagittal projection) were seen, as expected, most densely distributed in central PVH—assumed to be in PVHmpd. They were also widely distributed rostrally and caudally, but at much less density than in PVHmpd (recall figures 11.01-11.03). Rostral and caudal distribution was less than that of neSS, but more than neTRH. An area just rostral to the densest distribution, estimated to be about Level 24, contained distinctly few CRH neurons. Non-neCRH cells were scattered 266 throughout PVH in low density, except for a conspicuous area with virtually no non- neCRH cells in the area including and just rostral to the highest density for neCRH. That is, the area just rostral to the highest density of neCRH cells was devoid of all CRH expression. In observing Atlas Level views (recall figures 11.05-11.15 in Chapter 11, and see color figures 15.01-15.06 above), the few neCRH cells seen on Level 22 were in or near PVHam, a magnocellular subdivision characterized by OXY-expressing cells. From VAS/CRH cocktail staining in this series (VAS staining data not shown here, but see Appendix IX) and OXY/CRH cocktail staining in other series from this animal, it was determined that CRH cells illustrated were parvicellular, because they did not co-express either VAS or OXY. At Level 23 there were only two neCRH cells in PVHpv (one dorsal, the other quite ventral), and similar numbers of ne and non-neCRH cells (9) were intermixed in PVHap (figure 15.03). Their distribution extended from medial to lateral through PVHap, mostly in the dorsal part. At Level 24 almost all neCRH cells were in or adjacent to PVHpv. A few of these were in or near PVHmm, the small region of mostly OXY cells embedded in the periventricular zone (figure 15.04). More than half (six) of neCRH cells in PVHpv were clustered closely together in the most ventral part. Two neCRH cells were seen among the group of neSS cells and scattered neOXY cells in dorsal PVHap. Only two non-neCRH cells were seen at Level 24, ventral to the main group of neuroendocrine cells (amidst the large group of non- neTRH cells) in ventral PVHap. At Level 25, nearly eighty neCRH cells were seen (figure 15.05). Many, but certainly not all, were in PVHmpd, as expected. There were three neCRH parvicellular neurons located in PVHmpdl, the recently defined lateral extension of PVHmpd (recall earlier comments about the 3-D PVH model). Ten neCRH cells were seen in dorsal or ventral PVHpv and more than half of the total were clearly in PVHpmm, a magnocellular subdivision thought to contain mostly OXY cells at this level. Neuroendocrine CRH cells extended 267 across the border of PVHmpd ventrally through the medial half of PVHpmm, from very ventral PVHpv to the lateral border of the magnocellular subdivision. Ten non-neCRH cells were also seen at Level 25, half in PVHmpd and half in PVHpmm. At Level 26, the expected classic distribution of neCRH was observed, with nearly 200 cells at dense uniform concentration throughout PVHmpd (figure 15.06). PVHpv contained several neCRH cells dorsally and ventrally. About ten cells, spaced farther apart, were seen extending from PVHmpd into the medial edge of magnocellular PVHpmm. Only two non-neCRH cells were seen at Level 26, both in central PVHpmm. Caudal to Level 26 density of neCRH cells fell rapidly, while the small number of non-neCRH cells increased slightly. Twenty neCRH and five non-neCRH cells were seen on Level 27, mostly in PVHmpd (figure 15.07). Thus, neCRH in caudal PVHmpd at Level 27 was midway in number and density between that seen for TRH (half as many) and SS (one third more). In summary, neCRH (at almost 1700 cells, the most numerous cell type surveyed in these 1-in-4, 15µm series) was much more widely distributed outside PVHmpd than previously realized. Neuroendocrine CRH cells were seen in all parvicellular subdivisions, at low density throughout rostral PVH, and intermixed with neurons in predominantly magnocellular subdivisions. An unexpected finding was distribution of non-neCRH cells at low density throughout rostral PVH, in numbers at least as dense as those seen caudally. Most of the caudal non-neCRH cells were not in areas with known descending projections, as might be expected. Rather, they were intermixed with neCRH cells. This wider distribution of (mostly neuroendocrine) CRH neurons and their significant intermixing with magnocellular (especially OXY) neurons is likely related to the intimate relationship of CRH stress-response function with vascular smooth muscle and salt-balance functions of OXY (Stricker and Verbalis, 2004). For example, smooth muscle contraction and fluid homeostasis (including sodium and other electrolyte balance) are altered during blood 268 pressure changes commonly seen in the physiologic response to stress. The great intermixing of neCRH with neTRH and neSS neurons in mostly parvicellular subdivisions is consistent with their complementary effects in metabolic regulation, as discussed above. It is interesting to speculate about the function of the non-neCRH neurons located in areas that are thought to be almost exclusively neuroendocrine in function. Perhaps their neuropeptide content serves as a neurotransmitter, or has a more local paracrine effect. Theoretically, these cells could possibly have an interneuron function, since the trajectory and termination (local or distant) of their axons cannot be determined in the material presented here. These speculations would be interesting to pursue using more specialized cell labeling and axon- tracing techniques, with detailed analysis at the electron microscopic level. Oxytocin and Vasopressin Staining: Oxytocin-stained cells (total 1447 ne, 166 non-ne in sagittal view) were more numerous and widely distributed in PVH than previously appreciated. This is in contrast to Vasopressin cells (total 758 ne, 6 non-ne in sagittal view), which were seen at about the density and distribution expected (recall figures 12.01-12.04 in Chapter 12). OXY cells were densely distributed in areas of known magnocellular content: centrally in the area of PVHpmm and PVHpml, and rostrally in the areas of PVHam and PVHmm. However, many other neOXY cells were seen in areas thought to be primarily parvicellular: both caudally, and at moderate density rostrally between the known magnocellular subdivisions. There was a larger concentration of neOXY in mid-rostral PVH than expected, with numerous cells in PVHap, scattered between PVHam and PVHmm. Most non-neOXY cells were observed caudally, as expected for neurons that are known to be part of the descending functional compartment. Not expected were a group of non- neOXY distributed (at about the same density as caudal cells) in the approximate area of the anterior magnocellular group. 269 As expected in sagittal view, neuroendocrine vasopressin cells were grouped densely in the area encompassed by the posterior magnocellular subdivision. However, rather than a single dense circle- or ball-shaped distribution, some cells were seen slightly separated into a ventral and dorsal group. There were also a few scattered neVAS cells rostral and caudal to the main group. Only six non-neVAS cells were seen, clustered fairly close together in caudal PVH. Atlas Level views (recall figures 12.06-12.16 in Chapter 12, and see color figures 15.01-15.06 above) provide a more detailed view of the relationship between OXY and VAS cell distribution. On Atlas Levels 22 and 24 there were no non-neOXY cells while many were seen in the anterior and medial magnocellular PVH subdivisions, as expected (figures 15.02 and 15.04). On Level 22 a majority of neOXY cells (10 out of 15) were distributed in the ventral part of PVHpv rather than in the magnocellular PVHam subdivision. On Level 24, most of the 29 neOXY cells in medial PVH were in PVHmm or ventral to it, while seven more lateral cells were distributed widely in the dorsal part of PVHap. Fifty-five neOXY cells were seen between the two anterior magnocellular subdivisions on Level 23 (figure 15.03). Most were scattered throughout PVHap, with a slightly greater density ventro-laterally, and a half dozen were seen in PVHpv. A small number of non-neOXY cells (15) were seen distributed among the neOXY cells on Level 23, mostly in central PVHap. On Level 25 no non-neOXY cells were seen, and most of the 150 neOXY cells were distributed medially, in the magnocellular subdivision PVHpmm (figure 15.05). One neOXY cell was seen at the most ventral extent of PVHpv, about seven were clustered dorsally near the junction of PVHmpd and PVHdp, and several more were distributed laterally into PVHmpd along the mpd/dp border. Seven neOXY cells were seen in mid-ventral PVHmpd, adjacent to the border with the main mass of cells in PVHmm. Clearly, many of the neOXY cells at this level were in subdivisions usually regarded as parvicellular. Compared to Level 270 25, Level 26 appears at first glance to contain many fewer OXY cells, and again no non- neOXY cells are seen (figure 15.06). However, neOXY cells are similar in number at both levels (148 at Level 26 compared to 155 at Level 25), only spread over a larger area and thus at comparatively lower density. On Level 26, the classic impression of a ring (in coronal section view) or shell of OXY cells at the margin of PVHpml is quite distinct, though there are also a few OXY cells at the center of the magnocellular subdivision. A surprising finding at this level was the number and distribution of neOXY cells in PVHmpd, the parvicellular subdivision usually illustrated as containing mostly TRH and CRH cells. In general, neOXY cells at this level in PVHmpd seem more densely distributed at the dorsal and ventral parts, and in the narrow lateral area between PVHdp and the medio-dorsal margin of PVHpml. There were also several OXY cells scattered near the third ventricle in dorsal and ventral PVHpv. The dorsal periventricular group seemed to define a loose medio-lateral band continuous with the ones in dorsal PVHmpd. Cells in the ventral half of PVHpv were distributed at about the same low density as those in ventral PVHmpd. Also, there were a few neOXY cells in PVHdp and PVHmpv, subdivisions usually considered to contain mostly neurons with descending projections. Caudal to Level 26 numbers of both OXY and VAS cells declined rapidly, consistent with the impression that the magnocellular subdivision is a densely packed ball of cells with a fairly distinct outline on Levels 25 and 26. On Level 27 a small group of neOXY cells were seen spanning the border of PVHpv and PVHmpd, while a slightly higher number of non-neOXY cells were distributed more laterally in PVHmpd (figure 15.07). Distribution of neOXY at this level is similar to that seen for neSS, in contrast to the more lateral distribution seen for both neTRH and neCRH. The great majority of vasopressin cells (total 758 ne, 6 non-ne in sagittal view) were seen on Levels 25 and 26 (figures 15.05 and 15.06). As expected on Level 25, neVAS cells were seen exclusively in the magnocellular PVHpmm subdivision. They were intermixed 271 with OXY cells, extending to the lateral border of PVHpmm but not into the OXY-rich crescent comprising the ventral, medial and dorsal margins (surrounding ring, or external shell) of the subdivision. The main concentration of neVAS cells was contained on Level 26— about four times as many as on Level 25. The majority of neVAS cells at this level were in magnocellular PVHpml (distributed throughout, but chiefly in the central part) with only a few other cell types mixed in. However, one third of neVAS cells observed extended from the main mass into the lateral third of PVHmpd, an area of high concentration of neTRH and (especially) neCRH neurons. Rostral to Level 25 and caudal to Level 26 neVAS cells are almost absent. Two neVAS cells were seen on Level 23, one on Level 24 (both in lateral PVHap, see figures 15.03 and 15.04), and two in PVHpv on Level 27. Also on Level 27 was one of only six non-neVAS cells seen throughout PVH. It was in the dorso-lateral area of PVHmpd, close to non-neOXY cells also seen at that level (figure 15.07). Intermixing of OXY and VAS neurons in PVHpmm and PVHpml is not unexpected, considering their coordinated effects in regulation of renal physiology i.e., fluid balance and sodium concentration controlled by the kidney. Since these homeostatic functions are crucial to metabolism maintenance and mobilization of energy stores it is also not surprising to see OXY and VAS cells intermixed in areas containing predominantly neTRH and neCRH neurons. As noted above in discussion of SS, TRH and CRH cell intermixing, it seems likely that at least some population of OXY and VAS cells would benefit from the same or similar efferent signals impinging on neurons controlling metabolism, growth and tissue repair, and the response to stress. Neuroendocrine OXY fibers are seen in the internal layer of the median eminence (passing to the posterior pituitary) rather than the external layer (passing to the portal vasculature of the anterior pituitary). Therefore, it seems unlikely that neOXY cells in parvicellular regions of central and rostral PVH are providing primary stimulation to anterior 272 pituitary cells. Data showing dendritic release of OXY has been implicated in autoregulation of individual OXY cell firing, and in the coordinated firing of populations of OXY cells in the well-studied milk ejection response during lactation (Armstrong et al., 2002). However, other functions of dendritic peptide release seem likely. If this phenomenon extends to OXY cells in various subdivisions of PVH, it provides interesting evidence for speculation that these neurons may play a specialized role in local paracrine signaling or functional circuit coordination (Hirasawa et al., 2004). Harold Gainer and colleagues have contributed a great deal of creative thinking and intensive research to the study of basic cell biology in the OXY neuron for more than three decades. It will be interesting to observe new insights gained in this most-studied PVH cell type as modern experimental methods such as gene transfer, quantitative single cell mRNA analysis and targeted green fluorescent protein in OXY neurons provide new tools for investigation (Keir, et al., 1999; Xi et al., 1999; Zhang et al., 2002). Composite Maps of Neuroendocrine Cell Type Distribution: After evaluating all data for each cell type, as described above, individual maps from Atlas Levels were combined into composite Atlas Level figures to show patterns of distribution and intermixing among all neuroendocrine cell types surveyed in single Atlas Level images. Recall figures 13.07-13.16 in Chapter 13, and see color figures 15.01-15.06 above. This type of data display was one of the desired goals during the planning phase of these experiments. Because of anatomical accuracy gained in using the 3-D PVH model, it is more informative than the initial use of interleaved layers from individual, uncorrected data maps. Below (following Table 15.02: numbers of cell types within PVH subdivisions) is a description of patterns of neuroendocrine cell distribution seen at Levels 22 - 27 in Brain Maps. Discussion in this section about cell type distribution in combined maps is mainly about distribution of neuroendocrine cells, the primary focus of these studies. 273 Level 22, although not the most rostral to show PVH, was the first where ne cell type data was seen (figure 15.02). It contains 8-SS, 7-CRH, 15-OXY, and 2-TRH neuroendocrine cells (recall Table 15.01). Only a third of OXY cells at this level are in PVHam, where one would expect to see magnocellular OXY neurons. Two thirds are seen grouped in mid and ventral PVHpv. The middle group is just rostral to PVHmm, so this distribution is perhaps not unusual. One of the four SS cells at this level is also in PVHam. It exemplifies one of the few neurons seen to co-localize staining with another of the primary hypophysiotropic peptides surveyed—TRH in this case. Other SS cells are seen in or close to PVHpv, as expected. The two TRH cells seen are both in PVHam, as are three of the seven CRH cells. Other CRH cells at this level are immediately dorsal to PVHam in PVHap. It is striking that many different parvicellular cell types were seen in PVHam, an area known for its content of OXY cells, although in this representative 15µm section relatively few OXY cells were labeled. At Level 23, SS, CRH, and OXY cells predominate, with 28-SS, 11-CRH, 54-OXY, 4- TRH, 2-TH, and 2-VAS cells plotted (figure 15.03). Cells in PVHpv are densest in the dorsal part, with SS predominating, extending ventrally to mid-PVHpv. Many more SS cells than expected extend laterally, away from the periventricular area, into mid-dorsal PVHap. These form a vague medio-lateral band coursing slightly upward (reflecting the general shape of PVHap), as mentioned earlier. OXY cells predominate in the ventro-lateral quadrant of PVHap at Level 24, and are also seen in PVHpv dorsally and ventrally. TRH and CRH cells are scattered among the SS cells in PVHap and in the area of SS and OXY cells at the border with PVHap in dorsal PVHpv. PVHap is not a known magnocellular area, so predominance of OXY cells, especially in the ventro-lateral part is unexpected. They could be a caudal extension of cells from PVHam. A more speculative and interesting possibility is that they serve an active metabolic integrative function with SS, TRH and CRH cells nearby. 274 This idea is supported by evidence such as dendritic peptide released discussed earlier. Another interesting observation is the distribution of non-neTRH. At more than an order of magnitude greater than neTRH (48 versus 4 cells), it is the most numerous non-ne cell type seen at this level. These cells are seen mostly in ventral PVHpv and medio-dorsal PVHap— in a distribution almost complementary to that of neOXY. The significance of this relationship (and the intermixing of non-neOXY with lateral neOXY cells) is likely meaningful but not at all apparent. Neither is the distribution of non-neCRH cells in medio-lateral PVHap at numbers equivalent to neCRH. Perhaps this somewhat complementary non-ne and ne cell type distribution is significant for integrated PVH function? (See discussion about non- neTRH in the following chapter.) Such a tempting speculation certainly merits future investigation. At Level 24 SS, CRH, and OXY again predominate—in total 48-SS, 3-TRH, 10-CRH, 29-OXY, cells and 1-VAS cell are seen (figure 15.04). The SS CRH and OXY are concentrated in ventral PVHpv (rather than dorsally as in the previous level), with a few OXY and CRH cells seen in dorsal PVHap. Many SS cells are seen extending in a band across the dorsal part of PVHap, passing between two small separated groups of OXY cells and bounded by three CRH cells at the lower extent of their distribution. The difference in distribution of neSS between PVHpv (ventral) and PVHap (mostly dorsal) is striking, especially since other ne cells seem to follow a similar pattern. Most neOXY cells are in or near PVHmm, the magnocellular subdivision embedded within PVHpv. The concentration of neOXY cells seen in ventro-lateral PVHap at Level 23 is no longer evident. TRH is by far the most numerous non-ne cell type seen at this level: at 41 cells, they comprise 84% of the total. Non-neTRH cells are seen mostly in PVHap, with more dense distribution ventrally. Again, this distribution pattern is somewhat complementary to that of neuroendocrine cell types, further supporting the speculation about functional relationships mentioned above. 275 There are 131-SS, 44-TRH, 77-CRH, 150-OXY, 63-VAS, and only 2-TH cells seen at Level 25 (figure 15.05). Many neuroendocrine TRH cells are intermixed with the high concentration of SS cells in PVHpv, especially in its ventral two thirds. Few other cell types are seen in this subdivision. (Several OXY and CRH cells are intermixed with the densely packed SS cells in the dorsal part of PVHpv, and a small cluster of CRH cells are seen in the most ventral part.) Widely spaced SS cells extend in a band across the top part of PVHmpd, intermixed with a few OXY, CRH and TRH cells. There are a few more SS cells in far lateral PVHmpd at its border with PVHdp. A slight majority of the total TRH cells seen are located in the medio-ventral part of PVHmpd, with others in the ventral part of PVHpv as noted above. A few TRH cells extend into ventro-medial PVHpmm. Surprisingly, PVHmpd contains relatively few scattered CRH neurons, and their concentration is slightly higher near its lateral border with dorsal PVHpmm. Numerous CRH cells are located in medio-ventral parts of PVHpmm, and a small somewhat separate cluster is seen in extremely ventral PVHpv. Thus, at this level TRH cells (distributed in ventrally PVHpv and medio-ventrally in PVHmpd) seem somewhat separated from CRH cells seen ventrally in PVHpmm and laterally in PVHmpd. However, this is not in agreement with the medio-dorsal PVHmpd distribution of TRH cells with CRH cell concentration in ventral PVHmpd expected from the literature. Also noted on Level 25 are three CRH cells in the most lateral extension of PVHmpd (mpdl, or mpd-l). This part of the subdivision was made obvious by construction of the 3-D serial section atlas within Brain Maps. A few more CRH cells were seen close to this subdivision on sections nearby that were not mapped exactly onto Level 25 (recall figure 11.11). VAS cells are tightly grouped in ventro-lateral PVHpmm, as expected at Level 25. They comprise the single most densely distributed cell type (approaching a pure population) of any area in PVH. However, there are numerous OXY cells, a few CRH cells and at least 276 one non-neTRH cell scattered among them. OXY cells, in contrast, are most numerous in dorso-lateral PVHpmm, mixed with CRH cells. They are also concentrated in ventro- medial PVHpmm, mixed with TRH, SS and a few CRH cells. There are a few scattered OXY cells in central PVHmpd, near sparsely distributed SS, CRH and TRH cells as noted above. PVHdp, the dorsal parvicellular subdivision known to contain mainly spinal cord projection neurons, is notably free of any neuroendocrine (or non-ne) antibody-labeled cells at this level. A small cluster of OXY cells are seen dorsally in PVHpv at the border with PVHdp, intermixed with densely distributed SS cells. In general, there is much intermixing of all cell types in all subdivisions with neuroendocrine cells, with more OXY and far fewer total TH cells seen than expected. There were a small number of non-neuroendocrine parvicellular neurons seen sparsely distributed in each subdivision on Level 25. TRH cells were the most numerous (18) and widely distributed, with about half as many (10) CRH cells and only a few SS and TH cells. Again, distribution of non-neTRH and non-neCRH cells is somewhat complementary to their corresponding neuroendocrine cell distribution. That is, the majority of non-neCRH cells are seen in neTRH-rich areas, and several non-neTRH cells are similarly located adjacent to or among neCRH cells. At Level 26, cell density in PVHpv is low compared to very dense distribution of mostly OXY VAS and CRH cells in PVHmpd and PVHpml (figure 15.06). In total 12-S, 44- TRH, 204-CRH, 148-OXY, 204-VAS, and 4-TH cells are seen at this level. The few SS cells are almost all in PVHpv along with scattered OXY cells throughout. In dorsal PVHpv there are also a number of CRH cells intermixed, while in the ventral third of the subdivision a fairly dense group of TRH cells are seen between SS and OXY cells. In fact, TRH cells are at highest concentration locally of any cells in PVHpv on this Level. In central PVHpml OXY cells are seen in a ventro-lateral crescent surrounding a dense ball of VAS cells. OXY cells at the dorso-lateral and dorsal edges of the subdivision 277 are less numerous, while at the medio-lateral border they are more numerous and mixed with a few TRH and CRH cells. A few OXY cells are seen in PVHmpv and PVHdp; however, with the exception of two TRH cells in PVHmpv these descending projection subdivisions are devoid of neuroendocrine cells. A scattered band of OXY cells extends laterally from a few in dorsal PVHpv across the top of PVHmpd quite close to PVHdp. Ventral PVHmpd contains several widely scattered OXY cells intermixed with CRH and TRH cells. Oxy cells also extend into ventral PVHpv, where they are seen generally medial to TRH and SS cells. Three of only four TH cells seen on Level 26 are located quite ventrally at the border between PVHpv and PVHmpd near CRH and OXY cells, below the main group of TRH neurons. Most TRH cells in PVHmpd are located medially, well intermixed with CRH cells. They are seen from the most dorsal part of the subdivision, extending ventrally where their concentration is highest near the dense group of TRH cells in ventral PVHpv. Some TRH cells are seen in mid-lateral PVHmpd, a few dorsal near PVHdp and several in or near PVHmpv. In contrast, CRH cells are seen densely distributed throughout PVHmpd as expected. However, there are also CRH cells in all neuroendocrine subdivisions adjacent to PVHmpd. Surprisingly, numerous VAS cells are intermixed with CRH cells in PVHmpd, but only a few CRH cells are seen in the area of high concentration of VAS cells in PVHpml. Level 26, with the highest number and concentration of neuroendocrine cells, has the lowest number of non-ne cells seen at any atlas level. Non-ne cells are widely scattered for the most part, and non-neTRH (12 out of 17 total cells) is most numerous of three cell types seen. There were no non-neOXY cells at this level, however two each SS, CRH and TRH cells are seen in lateral PVHmpl. Non-neTRH cells are sparsely distributed except for a small group in or near PVHmpv, a subdivision known to contain cells with descending projections. 278 At Level 27, 10-SS, 13-OXY, 28-TRH, 19-CRH, 2-VAS, 1-TH, and 1-GRH cell are seen (figure 15.07). It is interesting that this most caudal level contains modest numbers of every neuroendocrine cell type surveyed. The SS cells are in or near PVHpv, mixed with a few TRH cells and some of the OXY cells. Most of the OXY cells span the border of PVHpv and PVHmpd with some TRH and CRH cells among them. Unexpectedly, two VAS cells are seen at this caudal level, very ventral in PVHpv. In contrast to OXY distribution, the main group of TRH and CRH cells are tightly intermixed in a triangular pattern in central PVHmpd, with one TRH cell seen laterally in the medial part of PVHlp. One TH cell is located quite lateral in PVHlp, far from other neuroendocrine cells. The single neGRH cell shown on Atlas Level views (only 2 were seen in the composite data series surveyed) is at the ventral edge of medial PVHmpd between one CRH and two TRH cells. Thus, at Level 27 neuroendocrine SS, and OXY cells are mixed very medially with a few TRH and CRH cells, while the central part of PVHmpd at this caudal level contains a fairly dense diagonal band of intermixed TRH and CRH neurons. Examples of all non-neuroendocrine cells types were also seen at this level. In contrast to neuroendocrine cells, non-neOXY and the single non-neVAS cell were located in the lateral part of PVHmpd, fairly intermixed with non-neCRH cells. Non-neTRH cells were seen more medially in PVHmpd (with a few in PVHpv and PVHlp) where they were intermixed with neOXY and the most medial neTRH cells. It is easily seen in the combination meta-analysis figures of Chapter 13 and the color figures above that there is much more intermixing of all neuroendocrine cell types at every level of PVH than has ever before been appreciated. In addition, cell types formerly thought to be fairly localized in specific PVH subdivisions or functional compartments are shown widely distributed throughout all rostro-caudal levels and most subdivisions of PVH. Most notable in that respect is the very wide distribution of neOXY, neCRH and neTRH in rostro- caudal extent and neSS in medio-lateral extent. Magnocellular neOXY neurons are 279 scattered widely among parvicellular neurons, and mixtures of parvicellular cell types occur typically in a vague diagonal band from medial PVHpv to dorso-lateral PVHap or PVHmpd. This pattern of intermixed cell type distribution is no doubt significant. Future analysis of specific patterns of inputs to PVH will likely reveal interesting functional implications. More efficient integration of neural information to produce a coordinated endocrine output under varying physiological conditions is the most obvious functional implication of this wide-spread neuroendocrine cell type intermixing. Intriguing implications of intermixed non-neuroendocrine cells with hypophysiotropic peptide phenotype will be discussed later. Peptide co-localization, General Discussion: All possible combinations of two different peptide antibody cocktails were tested in these studies, in expectation that spatially distinct populations of cells co-localizing two primary hypophysiotropic peptides might be found. However, only a small number of peptide double-labeled cells (either neuroendocrine or non-neuroendocrine) were seen. As expected (discussed earlier in chapter 12), OXY and VAS were never found co-localized in the same cell. This is in agreement with all previous reports of tissue surveyed for either protein or messenger RNA in animals under relatively normal physiological conditions. Co- localization of OXY and VAS gene expression determined by Polymerase Chain Reaction (very high amplification of DNA) or peptide expression using super-sensitive immunohistochemistry has been reported in cells sampled under special conditions of extreme challenge to body water homeostasis as noted previously. These data reflect an unusual physiological state, manifested by metabolic requirements of lactation or response to extreme osmotic challenge. They do not appreciably controvert the accepted view that magnocellular secretory neurons in PVH, supraoptic, and accessory nuclei are composed of 280 two mutually exclusive and somewhat spatially segregated populations of OXY or VAS expressing neurons. With the exception of cells co-localizing CRH and VAS, discussed below, double- labeled cells were infrequently observed and not reproducibly localized in different animals. For example, the single neSS/TRH neuron illustrated on Atlas Level 22 (figure 15.02) was one of only three such cells seen throughout PVH. Presence of several (6-10) double- labeled neCRH/TRH cells in PVHmpd in two pilot animals (see Appendix IV) was intriguing. A few test sections from one animal and a complete series from another showed similar results. (Recall the images of hand-drawn data maps in figure 5.02.) However, the double- label staining was not observed in subsequent experiments. The tissue from pilot studies was sectioned and stained at a different frequency compared to other experimental animals (1-in-6 rather than 1-in-4), and the initial TRH antiserum used was exhausted. Therefore those sections were not completely mapped and transferred to Atlas Levels for inclusion in the more extensive data presented here. It is unknown at this time whether a different combination of antisera to TRH will confirm this interesting preliminary result, which is in contrast to previous literature citing virtually no co-localization of CRH and TRH in the same PVH parvicellular neurons (Ceccatelli et al., 1989). In future, neurons expressing both the CRH and TRH phenotype might be identified using a different approach. For example, a combination of immunohistochemistry and in situ hybridization in the same section as described by Watts and colleagues could be used to advantage (Watts et al., 1989). Thus, staining for CRH protein and pre-pro TRH mRNA in the same neuron would unequivocally identify them as expressing both the CRH and TRH phenotype. This approach would obviate need for a new, more robust anti-TRH antibody, which has so far been difficult to obtain. However, since the TRH mRNA prohormone product is further processed into several different peptides (Lechan et al., 1986; Lechan et al., 1987; Bulant et al., 1988; Carr 281 et al., 1992; Pu et al., 1996), it would not address the issue of whether a given neuron produces functional CRH and hypophysiotropic TRH protein for release at its axon terminals. The only significant peptide co-localization seen in this work was that of VAS seen in some CRH-expressing cells in PVHmpd. This is a well-known phenomenon, in agreement with a body of literature documenting CRH/VAS co-localization in PVH parvicellular neurons under various physiological conditions (Sawchenko et al., 1984; Childs, 1989). Vasopressin is a known co-secretagog for ACTH release. In fact, it was an early strong candidate in the search for a corticotrophin releasing factor produced in hypothalamus (Roger Guillemin, personal communication). Vasopressin has been shown to appear in CRH neurons after acutely increased or long lasting physiological stress. Examples are colchicine injection, as in these experiments, or (especially) adrenalectomy. Sawchenko and Swanson showed that very few (estimated as 2%) CRH neurons in PVHmpd expressed VAS under basal conditions in colchicine-treated animals, while that proportion increased to almost 70% in adrenalectomized animals (Sawchenko and Swanson, 1985). In the studies presented here, the higher numbers of CRH/VAS positive cells seen in PVHmpd may be a function of increased sensitivity in immunohistochemical staining. Some of the same antisera were used here at 3- to 10-fold higher dilution than in the 1985 study cited above. Also, use of thinner sections sampled at higher frequency throughout PVH may influence the apparent density of co-labeled cells. In general agreement with this idea is the detailed work of Whitnall, Gainer, et al, documenting discrete populations of CRH and CRH/VAS cells within PVHmpd (Whitnall, 1988). Co-localization was observed in cells of the dorsal part of central PVHmpd in colchicine-treated animals. (Their previous work showed ~50% of CRH axons in the median eminence contained vesicles that co-packaged VAS or Vasopressin specific neurophysin in normal rats while after adrenalectomy virtually all were positive for both peptides.) They showed in 200µm steps through PVH using EM-immunohistochemical 282 staining that more than half of PVHmpd CRH cells at the mid rostro-caudal level were CRH/VAS positive. However, the two types of cells had different spatial distributions: the CRH-only cells were “more spread out in the medial parvicellular subdivision in both the dorso-ventral and the rostro-caudal axes”. The CRH/VAS co-localization data presented here appears to be about intermediate in density and distribution between that described in the two studies cited above. 283 Chapter 16. Conclusions Introduction: The goal of this work was to obtain a “new millennium view” of spatial distribution of neuroendocrine motoneurons in the paraventricular nucleus of the hypothalamus. Results described here comprise the first in-depth re-examination of this topic in nearly two decades. High-resolution analysis of the numbers, density, distribution and interrelationship between major neuroendocrine cell types in PVH has produced a new and detailed view. In addition, information about distribution of non-neuroendocrine cells of hypophysiotropic peptide phenotype was obtained. Double staining to define cells (either neuroendocrine or non- neuroendocrine) containing two of the surveyed primary hypophysiotropic peptides revealed very few cells, and no new information about their distribution. In pursuit of the primary goal to define the spatial relationships between cells in neuroendocrine PVH, a secondary project was undertaken. A high-resolution anatomical atlas of PVH, referenced to the published standard: Swanson’s Brain Maps, was constructed from serial Nissl sections of the rat brain originally used to produce Brain Maps. This atlas- mapping project was time consuming, but very productive. A new anatomically accurate 3-D model of PVH was obtained, and two new methods of analyzing high-resolution neuroanatomical data were developed. These in turn were used to elucidate important features of the spatial distribution of neuroendocrine motoneurons in the paraventricular nucleus. Historically, investigations of disease states and normal systems biology have driven research related to the three functional compartments of PVH. Thus, several areas of study have developed over the years with PVH as a central anatomical focus. For example, viscero-autonomic function (descending projections), fluid homeostasis (magnocellular neuroendocrine cells), and metabolic balance (parvicellular neuroendocrine cells) represent 284 major fields of investigation. Their attendant bodies of literature are, of course, interrelated. They have expanded apace with increasing specific knowledge about the primary cell types involved with relevant functions. Cardiovascular, renal, digestive, and other viscero-autonomic functions (though importantly related) are not primary to PVH neuroendocrine organization and will not be considered further in this discussion. It became apparent when surveying background information on each of the major neuroendocrine cell types that each area of study addresses a slightly different model of PVH anatomy and function. Understandably, these models derive largely from the specific diseases, physiological phenomena or cell types of major interest. However much of this literature seems somewhat mutually exclusive. Notably, study of growth and metabolism has focused on TRH, while that surrounding study of stress and related energy mobilization has focused on CRH. Workers in each of these fields reference fairly old descriptions of neuroendocrine cell type distribution in their investigations of PVH function, and tend not to reference the same sets of modern studies. This reflects the accepted (very useful) model of functional segregation within PVH. Each group may well have a more detailed ad hoc individual laboratory atlas or anatomical model for their cell type of interest, but specific details of its distribution throughout PVH are not usually mentioned. Thus, it is difficult to (mentally or actually) translate or transpose results from one area of investigation in relation to another for comparison within a common frame of reference or standard anatomical atlas. As functional studies coalesce around broader health issues and questions of physiological function (for example diabetes, obesity, and stress related illness such as cardiovascular disease) it is apparent that the detailed model of neuroendocrine cell type distribution presented here will inform all areas of study related to PVH. Considerable new information compared to that known twenty years ago, is available about neural circuits controlling endocrine and autonomic integration and the metabolic and 285 behavioral activities that are crucially involved with PVH function. In the past fifteen years, high-resolution anterograde tracer studies of specific neural inputs to various PVH subdivisions have added to the detailed knowledge of efferent control of PVH function. Immunohistochemical studies of defined peptide terminal fields ending on specific PVH cell types, discrete injections of retrograde tracers such as PHAL and the newer trans-neuronal virus labeling methods have generated a treasure trove of PVH input data. Some of the more interesting examples from the 1990s and a few more recent studies illustrate the explosion of increased information about inputs to PVH—especially significant advances in elucidating pathways and circuits involved in stress response and feeding behaviors (Cunningham et al., 1990; Kiss and Halasz, 1990; Toni et al., 1990; Bittencourt et al., 1991; Hisano and Daikoku, 1991; Liao et al., 1991; Canteras and Swanson, 1992; Liao et al., 1992; Roland and Sawchenko, 1993; Shioda et al., 1993; Larsen et al., 1994; Moga and Saper, 1994; Akabayashi, 1994; Jansen et al., 1995; Bester et al., 1997; Champagne et al., 1998; Diano et al., 1998; Broberger, 1999; Cowley et al., 1999; Thompson and Swanson, 2003; Legradi and Lechan, 1999; Fekete et al., 2004; Kerman et al., 2006). These studies and others elucidate complex neural circuits subserving behavioral aspects of metabolic regulation, for example coping with stressors in the environment, and regulating food and water intake. During the same time investigations of energy metabolism and ingestive behavior have produced a rich literature concerning the interaction between stress and metabolism (i.e. CRH and TRH function) and their relationship to obesity and diabetes (Stricker and Woods, 2004; Dallman et al., 2005). In the past decade molecular and genetic manipulations and the discovery of key peripheral hormones that control feeding (e.g., leptin, ghrelin) have prompted new investigations of central connections involved in feeding behavior, especially arcuate nucleus projections to PVH. Important implications for PVH function in metabolic regulation, stress response, and integration of motor behaviors that 286 ensure survival of the individual and the species are many and varied—all very important considerations in human public health. The high-resolution description of anatomical relationships between primary hypophysiotropic cell types presented here does not eclipse the general model of PVH functional compartmentalization in rat, nor does it imply new functions for the cell types surveyed. However, it will add subtlety to interpretation of neuroendocrine cell type distribution and insight for other anatomical and physiological studies of the rat PVH as a general mammalian model for endocrine, autonomic and behavioral function. This information is important for integrating information from functional and clinical studies about possible mechanisms of endocrine (thus PVH) function in humans. A considerable amount of data generated in this work is still to be analyzed and further evaluated in terms of ongoing research. For example, a streamlined, possibly semi- automated, method to enter data in register on the new atlas templates constructed for PVH between the seven published Atlas Levels of Brain Maps is a future goal. This will allow visualization of an accurate 3-D model of PVH cell type distribution similar to the Atlas Level views presented here, but at virtually serial section resolution. It remains for future work to fully realize the potential utility of these data in planning research into the function of PVH. Some of the most straightforward and interesting results have been discussed above. Following are comments about their potential usefulness and some possible goals for future investigation. Overview: The major new finding in these studies is the great intermixing of primary neuroendocrine cell types within PVH subdivisions, and their wider distribution throughout the nucleus, than was previously appreciated. Contrary to expectation, high-resolution analysis did not reveal a more precise parcellation or localization of cell types within 287 functional compartments or anatomical subdivisions in PVH. Although areas of high concentration of a single cell type were apparent, they overlapped, and no distinct regions containing a pure population of one cell type (or unique mixture of cell types) were observed. Intermixing of cell types produces a probabilistic or indefinite (as opposed to a cyto- architectonically defined) border between regions of highest cell type concentration. Some cell types surveyed (e.g., neCRH, neVAS and neOXY) were found in areas of high concentration and density, in fair agreement with the commonly accepted view. However they were not limited to that area of high concentration, and every cell type surveyed was distributed widely throughout PVH. At least a few examples of all cell types were seen in all functional compartments and in two or more anatomical subdivisions. (The exception was GRH, where few if any cells were seen in any given brain series.) Essentially every cell type surveyed was found more widely distributed and far more intermixed with others than previously appreciated. Another potentially important finding that emerged from the experimental design was the discovery of a significant population of clearly non-neuroendocrine neurons with the chemical phenotype of classical hypothalamic endocrine stimulating hormones. This phenomenon is very interesting, though detailed analysis is outside the scope of this dissertation. No doubt many of these cells have for years been considered hypophysiotropic in function because of their location in PVH and chemical signature. Therefore, current interpretations about the significance of input data to certain areas in PVH containing cells of assumed neuroendocrine function may not be accurate. This interesting observation bears more study and analysis in context of PVH efferent, pharmacological and physiological data. 288 Cell number surveyed (one series): CRH > OXY > TRH > SS > VAS > TH > GRH Cell type (peptide) Neuroendocrine Non-neuroendocrine Total (one series) CRH 1696 125 1821 OXY 1447 166 1613 TRH 510 543 1053 SS 787 120 907 VAS 758 6 764 TH 48 123 171 GRH 2 6 8 (composite: 4 cases) Table 16.01: Numbers of ne and non-ne cells from exemplar (1-in-4, 15µm) series. Occurrence of non-neuroendocrine cell types within supposedly neuroendocrine PVH was most apparent for TRH, while relatively smaller numbers of non-neCRH and non-neSS cells were observed. See table 16.01 and recall figures in chapter 15. There were significant numbers of non-ne cells seen for each of these major PVH peptides, generally distributed in a manner complementary to their areas of highest neuroendocrine cell density (recall discussion in chapter 15). The great number of non-neTRH cells in rostral PVH (nearly 50% of total TRH cells) was especially striking, as they comprised the vast majority of all cells surveyed in that area. However, additional examples were distributed among various similar and contrasting neuroendocrine cell types throughout PVH, as were other cell types surveyed. Significance of findings: The classical view of neural circuit organization in relation to a motor output nucleus such as (endocrine) PVH posits that sensory or derived neural information impinges on specialized groups of cells with a supposed common functional output. Indeed, great advances in understanding PVH function since the chemical neuroanatomy revolution of the 1970s are based on analysis of specific neural inputs in relation to identified endocrine 289 and/or autonomic outflow from PVH. In fact, an early aim of this work was to find a more precise, finer grain parcellation of groups of cells with a unique neuroendocrine chemical signature. The ultimate goal was to better understand how incoming sensory, neural and behavioral-state information are integrated to produce dynamic PVH output on a day-by-day and moment-to-moment basis. Thus, an appropriate exquisitely-coordinated combination of endocrine, autonomic and behavioral responses are produced to maintain physiological integrity in the face of internal metabolic challenges and the exigencies of continued existence in an uncertain environment. The finding that primary endocrine cell types were much less separate within PVH than previously thought was at first perplexing. However, on consideration it is even more interesting than a parcellation of more precise functional subdivisions might have been. PVH is unusual in that it has few if any classical interneurons (see earlier discussion); and many neural efferents important to PVH function are directed to a surrounding inhibitory input area, rather than to specific parts of the nucleus. How then might the well-known enormous integrative functions of PVH be effected? It seems likely that intermixing of cell types within the PVH provides a spatial integration of different functional effector sites (endocrine stimulating cell types). Thus, complex derived and dispersed information has immediate broad access to whichever endocrine response cell type or combination of cell types is most appropriate at the moment. For example, the intermixing of somatostatin cells with CRH and TRH cells speaks to the changing demands on energy mobilization, growth and metabolism made by appropriate response to internal or external stressors. Further, the considerable number and distribution of non-neuroendocrine cells with a neuroendocrine effector (hypophysiotropic) phenotype is intriguing. Based on localization date from many previous studies, most clearly are not likely to be descending projection cells. Might these neurons serve as functional interneurons within PVH? Perhaps they 290 communicate directly with other PVH cells of similar or complementary chemical signature. Terminal fields of these cells are unknown; therefore defining them as interneurons (cells with axon projection, soma and dendritic arbor within the same brain area) is speculative. As discussed above, non-neTRH terminals synapsing on neTRH cells in PVH have been demonstrated, but the source of those terminals (TRH cells inside or outside PVH) is unknown. However, there is electron microscopic evidence of dendro- dendritic synapses between cells with a similar chemical phenotype (TRH) in PVH (van den Pol, 1982). Thus, neuroendocrine and non-neuroendocrine cells may well be functionally connected, even if the relationship is not that expected of classical interneurons. Another indication of possible intra-PVH connectivity derives primarily from physiological patch-clamp studies in acutely prepared hypothalamic slices. Several distinct electrophysiological cell signatures have been characterized in endocrine PVH, more than might be expected for the two main subdivisions studied. Some of these apparently have inhibitory effects and electrical elements similar to those observed in recordings from interneurons in other brain areas. Thus, physiologists have for more than a decade argued for the existence of interneurons in PVH. More recently their evidence of endocannabinoid action in PVH makes the physiological data quite relevant to stress and feeding investigations (Tasker and Dudek, 1991; Hoffman et al., 1991; Boudaba et al., 1996; Daftary et al., 2000; Herman et al., 2002; Di et al., 2003; Tasker, 2004). However, anatomists have not found evidence in PVH of the typical type of interneuron (often inhibitory, with only local connections) seen elsewhere in the brain. Whether or not they serve as interneurons, the enormous numbers of non-neTRH cells (especially those concentrated in rostral parts of PVH where neTRH cells and most other neuroendocrine cells are scarce) undoubtedly have functional significance. Those and other rostral non-ne cell types observed in this study 291 might comprise a specialized chemical relay within PVH that widely integrates information derived from neural inputs restricted to rostral parts of the nucleus. Another possible functional significance of wide cell type intermixing within PVH derives from accumulating data on dendritic release of peptides, especially oxytocin (Leng et al., 1999; Brussaard and Herbison, 2000; Ludwig and Pittman, 2003). OXY cells have been known for some time to exhibit coordinated firing, whose functional consequence is release of large pulses of OXY into the posterior pituitary and thus the general circulation during parturition and lactation. Evidence of plasticity in electrical connections between OXY cells modulated by changes in surrounding glial cells was one of the more interesting mechanisms discovered in the control of these reproductive-related functions. Subsequently, microdialysis and other studies have shown that OXY is clearly released from dendrites into the interstitial fluid, thus it can act as a local diffusible exocrine hormone as well as a transmitted neuropeptide involved in release at axon terminals. Peptides are slow acting signal molecules, compared to classical neurotransmitters like epinephrine or glutamate. When released at synapses, their effective time of action is on the order of 30 seconds or less and their receptors have been seen at sites distant from axon terminals. Thus, resulting in the apparent “peptide receptor mismatch problem” (Zupanc, 1996). Actions of diffusing peptides released from cell bodies, dendrites or even interneurons (Baraban, 2004) are thus seen at some distance and can last 20 minutes or longer. In addition, they act locally to influence accumulation of additional peptide, continued dendritic release, and alterations in probability of firing for the original neuron. Dendritic release of peptides in nearby neurons may be similarly affected. It is possible, and quite likely that many other neuropeptides act in this fashion (Ludwig and Leng, 2006). Thus, an entirely different functional dynamic may be produced from the same phenotypic cell type, as described by Ludwig and Leng in the Nature Reviews Neuroscience paper cited above: 292 “Neuropeptides that are released from dendrites, such as oxytocin and vasopressin, function as autocrine or paracrine signals at their site of origin, but can also act at distant brain targets to evoke long-lasting changes in behaviour. Oxytocin, for instance, has profound effects on social bonding that are exerted at sites that richly express oxytocin receptors, but which are innervated by few, if any, oxytocin-containing projections. How can a prolonged, diffuse signal have coherent behavioural consequences? The recently demonstrated ability of neuropeptides to prime vesicle stores for activity-dependent release could lead to a temporary functional reorganization of neuronal networks harbouring specific peptide receptors, providing a substrate for long-lasting effects.” One functional consequence of cell type intermixing demonstrated in PVH may thus be an internal modulation of cell function (e.g., peptide accumulation and probability of its release at axon terminals) within the nucleus. Possibilities for modulation of effective functional circuits (thus, optimized endocrine output) within PVH by way of dendritic release of peptides are interesting indeed. The data presented here will allow new conceptual models for PVH integration of disparate neural inputs to be formed and tested. Studies of hypothalamus-pituitary-adrenal (HPA) stress axis and hypothalamus-pituitary-thyroid (HPT) metabolic axis will eventually be integrated, with each other and with studies of neural circuits and behavioral systems supporting them. Benefits from accessing cell type distribution information presented here will surely derive for these and other areas of research into PVH function. In addition, it will be interesting and very important to integrate modern knowledge of genetics and pharmacology with PVH anatomy, to probe various facets of endocrine function. The goal of manipulating the endocrine system in a controlled way is clearly of value in treating metabolic and stress-related diseases. In this “modern world” of virtually epidemic cardiovascular disease, obesity and diabetes (to name but a few effects of endocrine or 293 autonomic imbalance) every tool we can bring to bear on answering questions regarding control of endocrine (thus PVH) physiology is more than welcome. Anatomical results presented here, and the new methods devised to analyze them, are expected to support increases in understanding significant features of PVH function. 294 BIBLIOGRAPHY Akabayashi A, Koenig JI, Watanabe Y, Alexander JT, Leibowitz SF. 1994. Galanin- containing neurons in the paraventricular nucleus: A neurochemical marker for fat ingestion and body weight gain. Proc Natl Acad Sci U S A 91:10375-10379. Alonso G, Siaud P, Assenmacher I. 1988. Immunocytochemical ultrastructural study of hypothalamic neurons containing corticotropin-releasing factor in normal and adrenalectomized rats. Neuroscience 24:553-565. Alpert LC, Brawer JR, Patel YC, Reichlin S. 1976. Somatostatinergic neurons in anterior hypothalamus: Immunohistochemical localization. Endocrinology 98:255-258. Ambalavanar R, Morris R. 1989. Fluoro-gold injected either subcutaneously or intravascularly results in extensive retrograde labeling of CNS neurones having axons terminating outside the blood-brain barrier. Brain Res 505:171-175. Andrew R, MacVicar B, Dudek F, Hatton G. 1981. Dye transfer through gap junctions between neuroendocrine cells of rat hypothalamus. Science 211:1187-1189. Armstrong W, Warach S, Hatton G, McNeill T. 1980. Subnuclei in the rat hypothalamic paraventricular nucleus: A cytoarchitectural, horseradish peroxidase and immunocytochemical analysis. Neuroscience 5:1931-1958. Armstrong WE, Hatton GI. 1980. The localization of projection neurons in the rat hypothalamic paraventricular nucleus following vascular and neurohypophysial injections of HRP. Brain Res Bull 5:473-477. Armstrong WE, Stern JE, Teruyama R. 2002. Plasticity in the electrophysiological properties of oxytocin neurons. Microsc Res Tech 56:73-80. Bains J, Ferguson A. 1995. Paraventricular nucleus neurons projecting to the spinal cord receive excitatory input from the subfornical organ. Am J Physiol 268:R625-633. Bali B, Kovacs KJ. 2003. Gabaergic control of neuropeptide gene expression in parvocellular neurons of the hypothalamic paraventricular nucleus. Eur J Neurosci 18:1518-1526. Baraban SC, Tallent MK. 2004. Interneuron diversity series: Interneuronal neuropeptides - endogenous regulators of neuronal excitability. Trends in Neurosciences 27:135-142. Bargmann W. 1949. Uber die neurosekretorische verknupfung von hypothalamus und neurohypophyse. Z Zellforsch Mikrosk Anat 34:610-634. Beaulieu J, Champagne D, Drolet G. 1996. Enkephalin innervation of the paraventricular nucleus of the hypothalamus: Distribution of fibers and origins of input. J Chem Neuroanat 10:79-92. Bennett-Clarke C, Romagnano MA, Joseph SA. 1980. Distribution of somatostatin in the rat brain: Telencephalon and diencephalon. Brain Res 188:473-486. 295 Berk M, Finkelstein J. 1982. Efferent connections of the lateral hypothalamic area of the rat: An autoradiographic investigation. Brain Res Bull 8:511-526. Berkenbosch F, Tilders FJH. 1988. Effect of axonal transport blockade on corticotropin- releasing factor immunoreactivity in the median eminence of intact and adrenalectomized rats: Relationship between depletion rate and secretory activity. Brain Res 442:312-320. Berod AK, Hartman BK, Pujol JF. 1981. Importance of fixation in immunohistochemistry: Use of formaldehyde at variable pH for the localization of tyrosine hydroxylase. J Histochem Cytochem 29:844-850. Bester H, Besson JM, Bernard JF. 1997. Organization of efferent projections from the parabrachial area to the hypothalamus: A phaseolus vulgaris-leucoagglutinin study in the rat. J Comp Neurol 383:245-281. Bittencourt J, Benoit R, Sawchenko P. 1991. Distribution and origins of substance p- immunoreactive projections to the paraventricular and supraoptic nuclei: Partial overlap with ascending catecholaminergic projections. J Chem Neuroanat 4:63-78. Blackburn RE, Demko AD, Hoffman GE, Stricker EM, Verbalis JG. 1992. Central oxytocin inhibition of angiotensin-induced salt appetite in rats. Am J Physiol 263:R1347-1353. Blair M, Piekut D, Want A, Olschowka J. 1996. Role of the hypothalamic paraventricular nucleus in cardiovascular regulation. Clin Exp Pharmacol Physiol 23:161-165. Bloch B, Brazeau P, Bloom F, Ling N. 1983. Topographical study of the neurons containing hpGRF immunoreactivity in monkey hypothalamus. Neurosci Lett 37:23-28. Bloch B, Brazeau P, Ling N, Bohlen P, Esch F, Wehrenberg WB, Benoit R, Bloom F, Guillemin R. 1983. Immunohistochemical detection of growth hormone-releasing factor in brain. Nature 301:607-608. Bloom FE, Battenberg EL, Rivier J, Vale W. 1982. Corticotropin releasing factor (CRF): Immunoreactive neurones and fibers in rat hypothalamus. Regul Pept 4:43-48. Boler J, Enzmann F, Folkers K, Bowers CY, Schally AV. 1969. The identity of chemical and hormonal properties of the thyrotropin releasing hormone and thyro-glutimyl-histidyl-proline amide. Biochem Biophys Res Commun 37:705-710. Bondy CA, Jensen RT, Brady LS, Gainer H. 1989. Cholecystokinin evokes secretion of oxytocin and vasopressin from rat neural lobe independent of external calcium. Proc Natl Acad Sci U S A 86:5198-5201. Bota M, Dong H, Swanson LW. 2003. From gene networks to brain networks. Nat Neurosci 6:795-799. Bota M, Dong HW, Swanson LW. 2005. Brain architecture management system. Neuroinformatics 3:15-48. 296 Boudaba C, Szabo K, Tasker JG. 1996. Physiological mapping of local inhibitory inputs to the hypothalamic paraventricular nucleus. J Neurosci 16:7151-7160. Bouteiller JM, Simmons DM, Baudry M. 2000. Three-dimensional reconstruction of a rat brain atlas: Application to hippocampus reconstruction. Soc. for Neurosci. Abstr. 26. Brady LS, Lynn AB, Whitfield HJ, Jr., Kim H, Herkenham M. 1992. Intrahippocampal colchicine alters hypothalamic corticotropin-releasing hormone and hippocampal steroid receptor mRNA in rat brain. Neuroendocrinology 55:121-133. Brazeau P, Vale W, Burgus R, Ling N, Butcher M, Rivier J, Guillemin R. 1973. Hypothalamic polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone. Science 179:77-79. Broberger C. 1999. Hypothalamic cocaine- and amphetamine-regulated transcript (CART) neurons: Histochemical relationship to thyrotropin-releasing hormone, melanin-concentrating hormone, orexin/hypocretin and neuropeptide Y. Brain Res 848:101-113. Bromley SK, Burack WR, Johnson KG, Somersalo K, Sims TN, Sumen C, Davis MM, Shaw AS, Allen PM, Dustin ML. 2001. The immunological synapse. Annu Rev Immunol 19:375- 396. Brownstein M, Arimura A, Sato H, Schally AV, Kizer JS. 1975. The regional distribution of somatostatin in the rat brain. Endocrinology 96:1456-1461. Brownstein M, Palkovits M. 1976. Regional distribution of somatostatin in the rat brain. In: Front neuroendocrinol. Brownstein MJ, Palkovits M, Saavedra JM, Bassiri RM, Utiger RD. 1974. Thyrotropin- releasing hormone in specific nuclei of rat brain. Science 185:267-269. Bruhn TO, Mason RT, Vale WW. 1985. Presence of growth hormone-releasing factor-like immunoreactivity in rat duodenum. Endocrinology 117:1710-1712. Brussaard AB, Herbison AE. 2000. Long-term plasticity of postsynaptic GABAA-receptor function in the adult brain: Insights from the oxytocin neurone. Trends Neurosci 23:190-195. Bulant M, Delfour A, Vaudry H, Nicolas P. 1988. Processing of thyrotropin-releasing hormone prohormone (pro-TRH) generates pro-TRH-connecting peptides. Identification and characterization of prepro-TRH-(160-169) and prepro-TRH-(178-199) in the rat nervous system. J Biol Chem 263:17189-17196. Burgus R, Dunn TF, Desiderio D, Guillemin R. 1969. [molecular structure of the hypothalamic hypophysiotropic trf factor of ovine origin: Mass spectrometry demonstration of the pca-his-pro-nh1 sequence]. C R Acad Sci Hebd Seances Acad Sci D 269:1870-1873. Burgus R, Dunn TF, Desiderio D, Ward DN, Vale W, Guillemin R. 1970. Characterization of ovine hypothalamic hypophysiotropic TSH-releasing factor. Nature 226:321-325. 297 Cajal SRy. 1995. Histology of the nervous system of man and vertebrates. New York: Oxford University Press. Canteras N, Swanson L. 1992. Projections of the ventral subiculum to the amygdala, septum, and hypothalamus: A PHAL anterograde tract-tracing study in the rat. J Comp Neurol 324:180-194. Canteras NS, Simerly RB, Swanson LW. 1994. Organization of projections from the ventromedial nucleus of the hypothalamus: A phaseolus vulgaris-leucoagglutinin study in the rat. J Comp Neurol 348:41-79. Carr FE, Fein HG, Fisher CU, Wessendorf MW, Smallridge RC. 1992. A cryptic peptide (160-169) of thyrotropin-releasing hormone prohormone demonstrates biological activity in vivo and in vitro. Endocrinology 131:2653-2658. Ceccatelli S, Ahlbom E, Diana A, Zhivotovsky B. 1997. Apoptosis in rat hippocampal dentate gyrus after intraventricular colchicine. Neuroreport 8:3779-3783. Ceccatelli S, Cintra A, Hokfelt T, Fuxe K, Wikstrom AC, Gustafsson JA. 1989. Coexistence of glucocorticoid receptor-like immunoreactivity with neuropeptides in the hypothalamic paraventricular nucleus. Exp Brain Res 78:33-42. Ceccatelli S, Eriksson M, Hokfelt T. 1989. Distribution and coexistance of corticotropin- releasing factor-, neurotensin-, enkephalin-, cholecystokinin-, galanin- and vasoactive intestinal polypeptide/peptide histidine isoleucine-like peptides in the parvocellular part of the paraventricular nucleus. Neuroendocrinology 49:309-323. Ceccatelli S, Ernfors P, Villar MJ, Persson H, Hokfelt T. 1991. Expanded distribution of mRNA for nerve growth factor, brain-derived neurotrophic factor, and neurotrophin 3 in the rat brain after colchicine treatment. Proc Natl Acad Sci U S A 88:10352-10356. Ceccatelli S, Fahrenkrug J, Villar MJ, Hokfelt T. 1991. Vasoactive intestinal polypeptide/peptide histidine isoleucine immunoreactive neuron systems in the basal hypothalamus of the rat with special reference to the portal vasculature: An immunohistochemical and in situ hybridization study. Neuroscience 43:483-502. Ceccatelli S, Lundberg JM, Fahrenkrug J, Bredt DS, Snyder SH, Hokfelt T. 1992. Evidence for involvement of nitric oxide in the regulation of hypothalamic portal blood flow. Neuroscience 51:769-772. Champagne D, Beaulieu J, Drolet G. 1998. CRFergic innervation of the paraventricular nucleus of the rat hypothalamus: A tract-tracing study. J Neuroendocrinol 10:119-131. Chan-Palay V, Zaborszky L, Kohler C, Goldstein M, Palay SL. 1984. Distribution of tyrosine- hydroxylase-immunoreactive neurons in the hypothalamus of rats. J Comp Neurol 227:467- 497. Chiken S, Hatanaka N, Tokuno H. 2001. Distribution of median, ulnar and radial motoneurons in the monkey spinal cord: A retrograde triple-labeling study. Neurosci Lett 307:143-146. 298 Childs GV, Westlund KN, Unabia G. 1989. Characterization of anterior pituitary target cells for arginine vasopressin: Including cells that store adrenocorticotropin, thyrotropin-b, and both hormones. Endocrinology 125:554-559. Cintra A, Fuxe K, Solfrini V, Agnati LF, Tinner B, Wikstrom AC, Staines W, Okret S, Gustafsson JA. 1991. Central peptidergic neurons as targets for glucocorticoid action - evidence for the presence of glucocorticoid receptor immunoreactivity in various types of classes of peptidergic neurons. J Steroid Biochem Mol Biol 40:93-103. Cortes R, Ceccatelli S, Schalling M, Hokfelt T. 1990. Differential effects of intracerebroventricular colchicine administration on the expression of mRNAs for neuropeptides and neurotransmitter enzymes, with special emphasis on galanin: An in situ hybridization study. Synapse 6:369-391. Cowley MA, Pronchuk N, Fan W, Dinulescu DM, Colmers WF, Cone RD. 1999. Integration of NPY, AGRP, and melanocortin signals in the hypothalamic paraventricular nucleus: Evidence of a cellular basis for the adipostat. Neuron 24:155-163. Craige EH. 1940. Measurements of vasculartity in some hypothalamic nuclei of the albino rat. In: Fulton JF, Ranson SW, Frantz AM, Editors. The hypothalamus and central levels of autonomic function. Baltimore: Williams & Wilkins. p 310-319. Cullinan WE, Herman JP, Watson SJ. 1992. Ventral subicular interaction with the hypothalamic paraventricular nucleus: Evidence for a relay in the bed nucleus of the stria terminalis. J Comp Neurol 322:1-20. Cunningham ET, Jr, Bohn MC, Sawchenko PE. 1990. Organization of adrenergic inputs to the paraventricular and supraoptic nuclei of the hypothalamus in the rat. J Comp Neurol 292:651-667. Cunningham ET, Jr., Sawchenko PE. 1988. Anatomical specificity of noradrenergic inputs to the paraventricular and supraoptic nuclei of the rat hypothalamus. J Comp Neurol 274:60-76. Daftary SS, Boudaba C, Tasker JG. 2000. Noradrenergic regulation of parvocellular neurons in the rat hypothalamic paraventricular nucleus. Neuroscience 96:743-751. Daikoku S, Hisano S, Kawano H, Okamura Y, Tsuruo Y. 1983. Ontogenetic studies on the topographical heterogeneity of somatostatin-containing neurons in rat hypothalamus. Cell and Tissue Research 233:347-354. Daikoku S, Okamura Y, Kawano H, Tsuruo Y, Maegawa M, Shibasaki T. 1985. CRF- containing neurons of the rat hypothalamus. Cell Tissue Res 240:575-584. Dallman M, Pecoraro N, la Fleur S. 2005. Chronic stress and comfort foods: Self-medication and abdominal obesity. Brain Behav Immun 19:275-280. 299 Dashti AE, Burns GAPC, Simmons DM, Swanson L, Ghandeharizadeh S, Shahabi C, Stone J, S. J. 2001. The neuroanatomical rat brain viewer (NeuARt): A system for registering data against brain atlases. Chapter 4.3. In: MA Arbib, Grethe JS, Editors. Computing the brain: A guide to neuroinformatics. San Diego: Academic Press. Decavel C, van den Pol AN. 1992. Converging GABA- and glutamate-immunoreactive axons make synaptic contact with identified hypothalamic neurosecretory neurons. J Comp Neurol 316:104-116. De Goeij DCE, Kvetnansky R, Whitnall MH, Jezova D, Berkenbosch F, Tilders FJH. 1991. Repeated stress-induced activation of corticotropin-releasing factor neurons enhances vasopressin stores and colocalization with corticotropin-releasing factor in the median- eminence of rats. Neuroendocrinology 53:150-159. Del Bo A, Rosina A. 1986. Potential disynaptic pathways connecting the fastigial pressor area and the paraventricular nucleus of the hypothalamus in the rat. Neurosci Lett 71:37-42. Desy L, Pelletier G. 1977. Immunohistochemical localization of somatostatin in the human hypothalamus. Cell Tissue Res 184:491-497. Deuben RR, Meites J. 1964. Stimulation of pituitary growth hormone release by hypothalamic extract in vivo. Endocrinology 74:408-414. Di S, Malcher-Lopes R, Halmos KC, Tasker JG. 2003. Nongenomic glucocorticoid inhibition via endocannabinoid release in the hypothalamus: A fast feedback mechanism. J Neurosci 23:4850-4857. Diano S, Naftolin F, Goglia F, Horvath TL. 1998. Segregation of the intra- and extrahypothalamic neuropeptide Y and catecholaminergic inputs on paraventricular neurons, including those producing thyrotropin-releasing hormone. Regul Pept 75-76:117-126. Dierickx K, Vandesande F. 1979. Immunocytochemical localization of somatostatin- containing neurons in the rat hypothalamus. Cell Tissue Res 201:349-359. Dong HW, Petrovich GD, Watts AG, Swanson LW. 2001. Basic organization of projections from the oval and fusiform nuclei of the bed nuclei of the stria terminalis in adult rat brain. J Comp Neurol 436:430-455. Dong HW, Swanson LW. 2004a. Organization of axonal projections from the anterolateral area of the bed nuclei of the stria terminalis. J Comp Neurol 468:277-298. Dong HW, Swanson LW. 2004b. Projections from bed nuclei of the stria terminalis, posterior division: Implications for cerebral hemisphere regulation of defensive and reproductive behaviors. J Comp Neurol 471:396-433. Dong HW, Swanson LW. 2006a. Projections from bed nuclei of the stria terminalis, dorsomedial nucleus: Implications for cerebral hemisphere integration of neuroendocrine, autonomic, and drinking responses. J Comp Neurol 494:75-107. 300 Dong HW, Swanson LW. 2006b. Projections from bed nuclei of the stria terminalis, anteromedial area: Cerebral hemisphere integration of neuroendocrine, autonomic, and behavioral aspects of energy balance. J Comp Neurol 494:142-178. Dube D, Pelletier G. 1979. Effect of colchicine on the immunohistochemical localization of somatostatin in the rat brain: Light and electron microscopic studies. J Histochem Cytochem 27:1577-1581. Elde R, Hokfelt T. 1979. Localization of hypophysiotropic peptides and other biologically active peptides within the brain. In: Ann. Rev. Physiol.: Annual Reviews, Inc. p 587-602. Elde RP, Parsons JA. 1975. Immunocytochemical localization of somatostatin in cell bodies of the rat hypothalamus. Am J Anat 144:541-548. Enzmann F, Boler J, Folkers K, Bowers CY, Schally AV. 1971. Structure and synthesis of the thyrotropin-releasing hormone. J Med Chem 14:469-474. Epelbaum J, Arancibia LT, Herman JP, Kordon C, Palkovits M. 1981. Topography of median eminence somatostatinergic innervation. Brain Res 230:412-416. Fekete C, Legradi G, Mihaly E, Huang QH, Tatro JB, Rand WM, Emerson CH, Lechan RM. 2000. Alpha-melanocyte-stimulating hormone is contained in nerve terminals innervating thyrotropin-releasing hormone-synthesizing neurons in the hypothalamic paraventricular nucleus and prevents fasting-induced suppression of prothyrotropin-releasing hormone gene expression. J Neurosci 20:1550-1558. Fekete C, Wittmann G, Liposits Z, Lechan RM. 2004. Origin of cocaine- and amphetamine- regulated transcript (CART)-immunoreactive innervation of the hypothalamic paraventricular nucleus. J Comp Neurol 469:340-350. Ferreyra H, Kannan H, Koizumi K. 1983. Influences of the limbic system on hypothalamo- neurohypophysial system. Brain Res 264:31-45. Finley JCW, Maderdrut JL, Roger LJ, Petrusz P. 1981. The immunocytochemical localization of somatostatin-containing neurons in the rat central nervous system. Neuroscience 6:2173- 2192. Fodor M, Csaba Z, Kordon C, Epelbaum J. 1994. Growth hormone-releasing hormone, somatostatin, galanin and beta-endorphin afferents to the hypothalamic periventricular nucleus. J Chem Neuroanat 8:61-73. Folkers K, Enzmann F, Boler J, Bowers CY, Schally AV. 1969. Discovery of the synthetic tripeptide-sequence of the thyrotropin releasing hormone having activity. Biochem Biophys Res Commun 37:123-126. Franklin TR, Druhan JP. 2000. The retrograde tracer fluoro-gold interferes with the expression of fos-related antigens. Journal of Neuroscience Methods 98:1 - 8. Fulton JF. 1940. Historical resume. In: Fulton JF, Editor. The hypothalamus and central levels of autonomic function. New York: Hafner. p xiii-xvi. 301 Gardiner SM, Bennett T. 1989. Brain neuropeptides: Actions on central cardiovascular control mechanisms. Brain Research Reviews 14:79-116. Garrett WT, McBride RL, Williams JK, Jr., Feringa ER. 1991. Fluoro-gold's toxicity makes it inferior to true blue for long-term studies of dorsal root ganglion neurons and motoneurons. Neurosci Lett 128:137-139. Gimpl G, Fahrenholz F. 2001. The oxytocin receptor system: Structure, function, and regulation. Physiol Rev 81:629-683. Gold RM. 1973. Hypothalamic obesity: The myth of the ventromedial nucleus. Science 182:488-490. Gu GB, Simerly RB. 1997. Projections of the sexually dimorphic anteroventral periventricular nucleus in the female rat. J Comp Neurol 384:142-164. Guillemin R, Brazeau P, Bohlen P, Esch F, Ling N, Wehrenberg WB. 1982. Growth hormone-releasing factor from a human pancreatic tumor that caused acromegaly. Science 218:585-587. Gurdjian ES. 1927. The diencephalon of the albino rat. J Comp Neurol 43:1-114. Harris G. 1948. Neural control of the pituitary gland. Physiol Rev 28:139-179. Harris G. 1955. Neural control of the pituitary gland. London: Edward Arnold. Hartman BK, Swanson LW, Raichle ME, Preskorn SH, Clark HB. 1980. Central adrenergic regulation of cerebral microvascular permeability and blood flow; anatomic and physiologic evidence. Adv Exp Med Biol 131:113-126. Hatton G, Hutton U, Hoblitzell E, Armstrong W. 1976. Morphological evidence for two populations of magnocellular elements in the rat paraventricular nucleus. Brain Res 108:187- 193. Hatton GI. 1983. Some well-kept hypothalamic secrets disclosed. Federation Proceedings 42:2869-2874. Hatton GI, Cobbett P, Salm AK. 1985. Extranuclear axon collaterals of paraventricular neurons in the rat hypothalamus: Intracellular staining, immunocytochemistry and electrophysiology. Brain Res Bull 14:123-132. Hatton GI, Perlmutter LS, Salm AK, Tweedle CD. 1984. Dynamic neuronal-glial interactions in hypothalamus and pituitary: Implications for control of hormone synthesis and release. Peptides 5 Suppl 1:121-138. Herman JP, Tasker JG, Ziegler DR, Cullinan WE. 2002. Local circuit regulation of paraventricular nucleus stress integration: Glutamate-GABA connections. Pharmacol Biochem Behav 71:457-468. 302 Hermes M, Buijs R, Renaud L. 1996. Electrophysiology of suprachiasmatic nucleus projections to hypothalamic paraventricular nucleus neurons. Prog Brain Res 111:241-252. Hirasawa M, Schwab Y, Natah S, Hillard CJ, Mackie K, Sharkey KA, Pittman QJ. 2004. Dendritically released transmitters cooperate via autocrine and retrograde actions to inhibit afferent excitation in rat brain. J Physiol (Lond) 559:611-624. Hisano S, Daikoku S. 1991. Existence of mutual synaptic relations between corticotropin- releasing factor-containing and somatostatin-containing neurons in the rat hypothalamus. Brain Res 545:265-275. Hoffman GE, Hayes TA. 1979. Somatostatin neurons and their projections in dog diencephalon. J Comp Neurol 186:371-391. Hoffman NW, Tasker JG, Dudek FE. 1991. Immunohistochemical differentiation of electrophysiologically defined neuronal populations in the region of the rat hypothalamic paraventricular nucleus. J Comp Neurol 307:405-416. Hokfelt T, Efendiá S, Hellerstrom C, Johansson O, Luft R, Arimura A. 1975. Cellular localization of somatostatin in endocrine-like cells and neurons of the rat with special references to the a1-cells of the pancreatic islets and to the hypothalamus. Acta Endocrinol Suppl (Copenh) 200:5-41. Hokfelt T, Fahrenkrug J, Ju G, Ceccatelli S, Tsuruo Y, Meister B, Mutt V, Rundgren M, Brodin E, Terenius L, Hulting AL, Werner S, Bjorklund H, Vale W. 1987. Analysis of peptide histidine-isoleucine/vasoactive intestinal polypeptide-immunoreactive neurons in the central nervous system with special reference to their relation to corticotropin releasing factor- and enkephalin-like immunoreactivities in the paraventricular hypothalamic nucleus. Neuroscience 23:827-857. Hokfelt T, Fahrenkrug J, Tatemoto K, Mutt V, Werner S, Hulting AL, Terenius L, Chang KJ. 1983. The phi (phi-27)/corticotropin-releasing factor/enkephalin immunoreactive hypothalamic neuron: Possible morphological basis for integrated control of prolactin, corticotropin, and growth hormone secretion. Proc Natl Acad Sci U S A 80:895-898. Hokfelt T, Fuxe K, Johansson O, Jeffcoate S, White N. 1975. Distribution of thyrotropin- releasing hormone (TRH) in the central nervous system as revealed with immunohistochemistry. Eur J Pharmacol 34:389-392. Hollyday M. 1980. Organization of motor pools in the chick lumbar lateral motor column. J Comp Neurol 194:143-170. Ishikawa K, Taniguchi Y, Kurosumi K, Suzuki M. 1986. Origin of septal thyrotropin-releasing hormone in rat. Neuroendocrinology 44:54-58. Jansen A, Wessendorf M, Loewy A. 1995. Transneuronal labeling of CNS neuropeptide and monoamine neurons after pseudorabies virus injections into the stellate ganglion. Brain Res 683:1-24. 303 Johansson O, Hokfelt T. 1980. Thyrotropin-releasing hormone, somatostatin, and enkephalin: Distribution studies using immunohistochemical techniques. J Histochem Cytochem 28:364-366. Johansson O, Hokfelt T, Elde RP. 1984. Immunohistochemical distribution of somatostatin- like immunoreactivity in the central nervous system of the adult rat. Neuroscience 13:265- 339. Johansson O, Hokfelt T, Jeffcoate SL, White N, Sternberger LA. 1980. Ultrastructural localization of TRH-like immunoreactivity. Exp Brain Res 38:1-10. Kandel ER, Schwartz JH, Jessel TM. 1991. Principles of neural science. New York, San Francisco: McGraw-Hill, Health Professions Division. Kasting NW, Martin JB, Arnold MA. 1981. Pulsatile somatostatin release from the median eminence of the unanesthetized rat and its relationship to plasma growth hormone levels. Endocrinology 109:1739-1745. Kawano H, Daikoku S. 1987. Functional topography of the rat hypothalamic dopamine neuron systems: Retrograde tracing and immunohistochemical study. J Comp Neurol 265:242-253. Kawano H, Daikoku S. 1988. Somatostatin-containing neuron systems in the rat hypothalamus: Retrograde tracing and immunohistochemical studies. J Comp Neurol 271:293-299. Kawano H, Tosuru Y, Bando H, Daikoku S. 1991. Hypophysiotropic TRH-producing neurons identified by combining immunohistochemistry for pro-TRH and retrograde tracing. J Comp Neurol 307:531-538. Kay-Nishiyama C, Watts AG. 1999. Dehydration modifies somal CRH immunoreactivity in the rat hypothalamus: An immunocytochemical study in the absence of colchicine. Brain Res 822:251-255. Keir SD, House SB, Li J, Xiao X, Gainer H. 1999. Gene transfer into hypothalamic organotypic cultures using an adeno- associated virus vector. Exp Neurol 160:313-316. Kerman IA, Akil H, Watson SJ. 2006. Rostral elements of sympatho-motor circuitry: A virally mediated transsynaptic tracing study. J Neurosci 26:3423-3433. Kiss J, Halasz B. 1990. Ultrastructural analysis of the innervation of TRH -immunoreactive neuronal elements located in the periventricular subdivision of the paraventricular nucleus of the rat hypothalamus. Brain Res 532:107-114. Kitazawa S, Shioda S, Nakai Y. 1987. Catecholaminergic innervation of neurons containing corticotropin-releasing factor in the paraventricular nucleus of the rat hypothalamus. Acta Anat (Basel) 129:337-343. Knigge KM, Joseph SA, Hoffman GE. 1978. Organization of LRF-and SRIF-neurons in the endocrine hypothalamus. Res Publ Assoc Res Nerv Ment Dis 56:49-67. 304 Koh EJ, Ricardo JA. 1980. Paraventricular nucleus of the hypoothalamus: Anatomical evidence of ten functionally discrete subdivisions. Soc, Neurosci. Abstr. 6:521. Kovacs KJ, Miklos IH, Bali B. 2004. Gabaergic mechanisms constraining the activity of the hypothalamo-pituitary-adrenocortical axis. Ann N Y Acad Sci 1018:466-476. Landry M, Roche D, Angelova E, Calas A. 1997. Expression of galanin in hypothalamic magnocellular neurones of lactating rats: Co-existence with vasopressin and oxytocin. J Endocrinol 155:467-481. Landry M, Trembleau A, Arai R, Calas A. 1991. Evidence for a colocalization of oxytocin mRNA and galanin in magnocellular hypothalamic neurons: A study combining in situ hybridization and immunohistochemistry. Brain Res Mol Brain Res 10:91-95. Larsen P, Hay-Schmidt A, Mikkelsen J. 1994. Efferent connections from the lateral hypothalamic region and the lateral preoptic area to the hypothalamic paraventricular nucleus of the rat. J Comp Neurol 342:299-319. Lechan RM, Goodman RH, Rosenblatt M, Reichlin S, Habener JF. 1983. Prosomatostatin- specific antigen in rat brain: Localization by immunocytochemical staining with an antiserum to a synthetic sequence of preprosomatostatin. Proc Natl Acad Sci U S A 80:2780-2784. Lechan RM, Jackson IMD. 1982. Immunohistochemical localization of thyrotropin-releasing hormone in the rat hypothalamus and pituitary. Endocrinology 111:55-65. Lechan RM, Lin HD, Ling N, Jackson IM, Jacobson S, Reichlin S. 1984. Distribution of immunoreactive growth hormone releasing factor(1-44)nh2 in the tuberoinfundibular system of the rhesus monkey. Brain Res 309:55-61. Lechan RM, Wu P, Jackkson IMD. 1987. Immunocytochemical distribution in rat brain of putative peptides derived from thyrotropin-releasing hormone prohormone. Endocrinology 121:1879-1891. Lechan RM, Wu P, Jackson IM, Wolf H, Cooperman S, Mandel G, Goodman RH. 1986. Thyrotropin-releasing hormone precursor: Characterization in rat brain. Science 231:159- 161. Leergaard TB, Bjaalie JG, Devor A, Wald LL, Dale AM. 2003. In vivo tracing of major rat brain pathways using manganese-enhanced magnetic resonance imaging and three- dimensional digital atlasing. Neuroimage 20:1591-1600. Legradi G, Lechan RM. 1998. The arcuate nucleus is the major source for neuropeptide y- innervation of thyrotropin-releasing hormone neurons in the hypothalamic paraventricular nucleus. Endocrinology 139:3262-3270. Legradi G, Lechan RM. 1999. Agouti-related protein containing nerve terminals innervate thyrotropin-releasing hormone neurons in the hypothalamic paraventricular nucleus. Endocrinology 140:3643-3652. 305 Leibowitz SF. 1975. Pattern of drinking and feeding produced by hypothalamic norepinephrine injection in the satiated rat. Physiol Behav 14:731-742. Leibowitz SF. 1978. Paraventricular nucleus: A primary site mediating adrenergic stimulation of feeding and drinking. Pharmacol Biochem Behav 8:163-175. Leng G, Brown CH, Russell JA. 1999. Physiological pathways regulating the activity of magnocellular neurosecretory cells. Prog Neurobiol 57:625-655. Lennard DE, Eckert WA, Merchenthaler I. 1993. Corticotropin-releasing hormone neurons in the paraventricular nucleus project to the external zone of the median eminence: A study combining retrograde labeling with immunocytochemistry. J Neuroendocrinol 5:175-181. Leroux P, Gonzalez BJ, Bodenant C, Bucharles C, Vaudry H. 1992. Somatostatin: A putative neurotropic factor with pleiotropic activity in the rat central nervous system. In: Joose J, Bujis RM, Tilders FJH, Editors. The peptidergic neuron. Elsevier. Levin M, Sawchenko P, Howe P, Bloom S, Polak J. 1987. Organization of galanin- immunoreactive inputs to the paraventricular nucleus with special reference to their relationship to catecholaminergic afferents. J Comp Neurol 261:562-582. Levin MC, Sawchenko PE. 1993. Neuropeptide co-expression in the magnocellular neurosecretory system of the female rat: Evidence for differential modulation by estrogen. Neuroscience 54:1001-1018. Liao N, Bulant M, Nicolas P, Vaudry H, Pelletier G. 1991. Anatomical interactions of proopiomelanocortin (POMC)-related peptides, neuropeptide y (NPY) and dopamine beta- hydroxylase (D beta H) fibers and thyrotropin-releasing hormone (TRH) neurons in the paraventricular nucleus of rat hypothalamus. Neuropeptides 18:63-67. Liao N, Vaudry H, Pelletier G. 1992. Neuroanatomical connections between corticotropin- releasing factor (CRH) and somatostatin (SRIF) nerve endings and thyrotropin-releasing hormone (TRH) neurons in the paraventricular nucleus of rat hypothalamus. Peptides 13:677-680. Lincoln D, Wakerley J. 1975. Factors governing the periodic activation of supraoptic and paraventricular neurosecretory cells during suckling in the rat. J Physiol (Lond) 250:443-461. Lind RW. 1986. Bi-directional, chemically specified neural connections between the subfornical organ and the midbrain raphe system. Brain Res 384:250-261. Lind RW, Swanson LW, Bruhn TO, Ganten D. 1985. The distribution of angiotensin ii- immunoreactive cells and fibers in the paraventriculo-hypophysial system of the rat. Brain Res 338:81-89. Lind RW, Swanson LW, Ganten D. 1984. Angiotensin II immunoreactive pathways in the central nervous system of the rat: Evidence for a projection from the subfornical organ to the paraventricular nucleus of the hypothalamus. Clin Exp Hypertens A 6:1915-1920. 306 Lind RW, Thunhorst RL, Johnson AK. 1984. The subfornical organ and the integration of multiple factors in thirst. Physiol Behav 32:69-74. Ling N, Burgus R, Rivier J, Vale W, Brazeau P. 1973. The use of mass spectrometry in deducing the sequence of somatostatin-- a hypothalamic polypeptide that inhibits the secretion of growth hormone. Biochem Biophys Res Commun 50:127-133. Liposits Z, Sherman D, Phelix C, Paull WK. 1986. A combined light and electron microscopic immunocytochemical method for the simultaneous localization of multiple tissue antigens. Tyrosine hydroxylase immunoreactive innervation of corticotropin releasing factor synthesizing neurons in the paraventricular nucleus of the rat. Histochemistry 85:95-106. Ludwig M, Leng G. 2006. Dendritic peptide release and peptide-dependent behaviours. Nat Rev Neurosci 7:126-136. Ludwig M, Pittman QJ. 2003. Talking back: Dendritic neurotransmitter release. Trends Neurosci 26:255-261. Markakis EA. 2002. Development of the neuroendocrine hypothalamus. Front Neuroendocrinol 23:257-291. Markakis EA, Swanson LW. 1997. Spatiotemporal patterns of secretomotor neuron generation in the parvicellular neuroendocrine system. Brain Res Brain Res Rev 24:255-291. Mayo KE, Cerelli GM, Lebo RV, Bruce BD, Rosenfeld MG, Evans RM. 1985. Gene encoding human growth hormone-releasing factor precursor: Structure, sequence, and chromosomal assignment. Proc Natl Acad Sci U S A 82:63-67. Mayo KE, Cerelli GM, Rosenfeld MG, Evans RM. 1985. Characterization of cDNA and genomic clones encoding the precursor to rat hypothalamic growth hormone-releasing factor. Nature 314:464-467. Mayo KE, Evans RM, Rosenfeld GM. 1986. Genes encoding mammalian neuroendocrine peptides: Strategies toward their identification and analysis. Annu Rev Physiol 48:431-446. McKellar S, Loewy A. 1981. Organization of some brain stem afferents to the paraventricular nucleus of the hypothalamus in the rat. Brain Res 217:351-357. Meister B, Hokfelt T, Geffard M, Oertel W. 1988. Glutamic-acid decarboxylase-aminobutyric and gamma-aminobutyric acid-like immunoreactivities in corticotropin-releasing factor- containing parvocellular neurons of the hypothalamic paraventricular nucleus. Neuroendocrinology 48:516-526. Meister B, Hokfelt T, Vale WW, Sawchenko PE, Swanson LW, Goldstein M. 1986. Coexistence of tyrosine hydroxylase and growth hormone-releasing factor in a subpopulation of tubero-infundibular neurons of the rat. Neuroendocrinology 42:237-247. Meister B, Villar MJ, Ceccatelli S, Hokfelt T. 1990. Localization of chemical messengers in magnocellular neurons of the hypothalamic supraoptic and paraventricular nuclei: An immunohistochemical study using experimental manipulations. Neuroscience 37:603-633. 307 Merchenthaler I. 1990. Retrograde labeling of hypophysiotropic neurons by local injection of wheat germ agglutinin (WGA) into the median eminence or peripheral administration of fluoro-gold. Molecular and Cellular Neurosciences 1:93-106. Merchenthaler I. 1991. Neurons with access to the general circulation in the central nervous system of the rat: A retrograde tracing study with fluoro-gold. Neuroscience 44:655-662. Merchenthaler I, Setalo G, Csontos C, Petrusz P, Flerko B, Negro-Vilar A. 1989. Combined retrograde tracing and immunocytochemical identification of luteinizing hormone-releasing hormone- and somatostatin-containing neurons projecting to the median eminence of the rat. Endocrinology 125:2812-2821. Merchenthaler I, Vigh S, Schally AV, Petrusz P. 1984. Immunocytochemical localization of growth hormone-releasing factor in the rat hypothalamus. Endocrinology 114:1082-1085. Merchenthaler I, Vigh S, Schally AV, Petrusz P. 1984. Immunocytochemical localization of growth hormone-releasing factor in the rat hypothalamus. Endocrinology 114:1082-1085. Merighi A. 2002. Costorage and coexistence of neuropeptides in the mammalian CNS. Prog Neurobiol 66:161-190. Mezey E, Kiss JZ. 1991. Coexpression of vasopressin and oxytocin in hypothalamic supraoptic neurons of lactating rats. Endocrinology 129:1814-1820. Mezey E, Reisine TD, Skirboll L, Beinfeld M, Kiss JZ. 1985. Cholecystokinin in the medial parvocellular subdivision of the paraventricular nucleus. Co-existence with corticotropin- releasing hormone. Ann N Y Acad Sci 448:152-156. Moga M, Saper C. 1994. Neuropeptide-immunoreactive neurons projecting to the paraventricular hypothalamic nucleus in the rat. J Comp Neurol 346:137-150. Morris JF, Pow DV. 1993. New anatomical insights into the inputs and outputs from hypothalamic magnocellular neurons. Ann N Y Acad Sci 689:16-33. Nishiyama T, Heike Y, Matzuzaki T, Kawano H, Daikoku S, Suzuki M. 1983. Immunoreactive trh-containing neurons in the rat hypothalamus. Biomedical Res (Suppl.):65/74. Nishiyama T, Kawano H, Tsuruo Y, M M, Hisano S, Adachi T, Daikoku S, Suzuki M. 1985. Hypothalamic thyrotropin-releasing hormone(TRH)-containing neurons involved in the hypothalamic-hypophysial-thyroid axis. Light microscope immunocytochemistry. Brain Res 345:205. Oldfield BJ, Hou-Yu A, Silverman AJ. 1985. A combined electron microscopic HRP and immunocytochemical study of the limbic projections to rat hypothalamic nuclei containing vasopressin and oxytocin neurons. J Comp Neurol 231:221-231. Parish DC, Rodriguez EM, Birkett SD, Pickering BT. 1981. Effects of small doses of colchicine on the components of the hypothalamo-neurohypophysial system of the rat. Cell Tissue Res 220:809-827. 308 Parsons JA, Erlandsen SL, Hegre OD, McEvoy RD, Elde RP. 1976. Central and peripheral localization of somatostatin immunoenzyme immunocytochemical studies. J Histochem Cytochem 24:872-882. Patel YC, Reichlin S. 1978. Somatostatin in hypothalamus, extrahypothalamic brain, and peripheral tissues of the rat. Endocrinology 102:523-530. Patterson PH. 1994. Leukemia inhibitory factor, a cytokine at the interface between neurobiology and immunology. Proc Natl Acad Sci U S A 91:7833-7835. Pu LP, Ma W, Barker JL, Loh YP. 1996. Differential coexpression of genes encoding prothyrotropin- releasing hormone (pro- TRH) and prohormone convertases (pc1 and pc2) in rat brain neurons: Implications for differential processing of pro- TRH. Endocrinology 137:1233-1241. Raichle ME. 1975. Cerebral blood flow and metabolism. Ciba Found Symp:85-96. Reichlin S. 1983. Somatostatin. In: Kreiger DT, Brownstein MJ, Martin JB, Editors. Brain peptides. New York: John Wiley and Sons. p 711-752. Renaud LP, Bourquet CW. 1991. Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog Neurobiol 36:131-169. Rho JH, Swanson LW. 1987. Neuroendocrine CRH motoneurons: Intrahypothalamic axon terminals shown with a new retrograde-lucifer-immuno method. Brain Res 436:143-147. Ricardo J, Koh E. 1978. Anatomical evidence of direct projections from the nucleus of the solitary tract to the hypothalamus, amygdala, and other forebrain structures in the rat. Brain Res 153:1-26. Rinaman L, Hoffman GE, Dohanics J, Le WW, Stricker EM, Verbalis JG. 1995. Cholecystokinin activates catecholaminergic neurons in the caudal medulla that innervate the paraventricular nucleus of the hypothalamus in rats. J Comp Neurol 360:246-256. Rinaman L, Schwartz G. 2004. Anterograde transneuronal viral tracing of central viscerosensory pathways in rats. J. Neurosci. 24:2782-2786. Risold PY, Canteras NS, Swanson LW. 1994. Organization of projections from the anterior hypothalamic nucleus: A phaseolus vulgaris-leucoagglutinin study in the rat. J Comp Neurol 348:1-40. Rivest S. 2001. How circulating cytokines trigger the neural circuits that control the hypothalamic-pituitary-adrenal axis. Psychoneuroendocrinology 26:761-788. Rivier J, Spiess J, Thorner M, Vale W. 1982. Characterization of a growth hormone-releasing factor from a human pancreatic islet tumour. Nature 300:276-278. 309 Rodier PM, Kates B, White WA, Phelps CJ. 1990. Birthdates of the growth hormone releasing factor cells of the rat hypothalamus: An autoradiographic study of immunocytochemically identified neurons. J Comp Neurol 291:363-372. Roland BL, Sawchenko PE. 1993. Local origins of some GABAergic projections to the paraventricular and supraoptic nuclei of the hypothalamus in the rat. J Comp Neurol 332:123-143. Russell JT, Brownstein MJ, Gainer H. 1980. Biosynthesis of vasopressin, oxytocin, and neurophysins: Isolation and characterization of two common precursors (propressophysin and prooxyphysin). Endocrinology 107:1880-1891. Sakanaka M, Magari S, Shibasaki T, Inoue N. 1989. Co-localization of corticotropin-releasing factor-like and enkephalin-like immunoreactivities in nerve-cells of the rat hypothalamus and adjacent areas. Brain Res 487:357-362. Sanchez E, Uribe RM, Corkidi G, Zoeller RT, Cisneros M, Zacarias M, Morales_Chapa C, Charli JL, Joseph_Bravo P. 2001. Differential responses of thyrotropin-releasing hormone (trh) neurons to cold exposure or suckling indicate functional heterogeneity of the TRH system in the paraventricular nucleus of the rat hypothalamus. Neuroendocrinology 74:407- 422. Saper C, Loewy A. 1980. Efferent connections of the parabrachial nucleus in the rat. Brain Res 197:291-317. Sarkar S, Fekete C, Legradi G, Lechan RM. 2003. Glucagon like peptide-1 (7-36) amide (GLP-1) nerve terminals densely innervate corticotropin-releasing hormone neurons in the hypothalamic paraventricular nucleus. Brain Res 985:163-168. Sawchenko P, Benoit R, Brown M. 1988. Somatostatin 28-immunoreactive inputs to the paraventricular and supraoptic nuclei: Principal origin from non-aminergic neurons in the nucleus of the solitary tract. J Chem Neuroanat 1:81-94. Sawchenko P, Swanson L. 1981. Central noradrenergic pathways for the integration of hypothalamic neuroendocrine and autonomic responses. Science 214:685-687. Sawchenko P, Swanson L. 1985. Localization, colocalization, and plasticity of corticotropin- releasing factor immunoreactivity in rat brain. Fed Proc 44:221-227. Sawchenko P, Swanson L, Rivier J, Vale W. 1985. The distribution of growth-hormone- releasing factor (GRF) immunoreactivity in the central nervous system of the rat: An immunohistochemical study using antisera directed against rat hypothalamic grf. J Comp Neurol 237:100-115. Sawchenko P, Swanson L, Vale W. 1984a. Corticotropin-releasing factor: Co-expression within distinct subsets of oxytocin-, vasopressin-, and neurotensin-immunoreactive neurons in the hypothalamus of the male rat. J Neurosci 4:1118-1129. 310 Sawchenko P, Swanson L, Vale W. 1984b. Co-expression of corticotropin-releasing factor and vasopressin immunoreactivity in parvocellular neurosecretory neurons of the adrenalectomized rat. Proc Natl Acad Sci U S A 81:1883-1887. Sawchenko P, Swanson LW. 1990. Growth hormone releasing hormone, Chapter II. In: Bjorklund A, Hokfelt T, Kuhar MJ, Editors. Handbook of chemical neuroanatomy. Elsevier. p 131-163. Sawchenko PE. 1998. Toward a new neurobiology of energy balance, appetite, and obesity: The anatomists weigh in. J Comp Neurol 402:435-441. Sawchenko PE, Brown ER, Chan RK, Ericsson A, Li HY, Roland BL, Kovacs KJ. 1996. The paraventricular nucleus of the hypothalamus and the functional neuroanatomy of visceromotor responses to stress. Prog Brain Res 107:201-222. Sawchenko PE, Pfeiffer SW. 1995. Ultrastructural localization of inhibin [beta]- and somatostatin-28-immunoreactivities in the paraventricular and supraoptic nuclei. Brain Res 694:233-245. Sawchenko PE, Swanson LW. 1982a. Immunohistochemical identification of neurons in the paraventricular nucleus of the hypothalamus that project to the medulla or to the spinal cord in the rat. J Comp Neurol 205:260-272. Sawchenko PE, Swanson LW. 1982b. The organization of noradrenergic pathways from the brainstem to the paraventricular and supraoptic nuclei in the rat. Brain Res 257:275-325. Sawchenko PE, Swanson LW. 1983a. The organization of forebrain afferents to the paraventricular and supraoptic nuclei of the rat. J Comp Neurol 218:121-144. Sawchenko PE, Swanson LW. 1983b. The organization and biochemical specificity of afferent projections to the paraventricular and supraoptic nuclei. In: The neurohypophysis: Structure, function and control. Amsterdam: Elsevier. p 19-29. Sawchenko PE, Swanson LW, Joseph SA. 1982. The distribution and cells of origin of ACTH(1-39)-stained varicosities in the paraventricular and supraoptic nuclei. Brain Res 232:365-374. Schally AV, Bowers CY, Redding TW. 1966a. Presence of thyrotropic hormone-releasing factor (TRF) in porcine hypothalamus. Proc Soc Exp Biol Med 121:718-722. Schally AV, Bowers CY, Redding TW. 1966b. Purification of thyrotropic hormone-releasing factor from bovine hypothalamus. Endocrinology 78:726-732. Schally AV, Kuroshima A, Ishida Y, Arimura A, Saito T, Bowers CY, Steelman SL. 1966. Purification of growth hormone-releasing factor from beef hypothalamus. Proc Soc Exp Biol Med 122:821-823. Scharrer E. 1952. The general significance of the neurosecretory cell. Scientia (Milan) 87:176-182. 311 Scharrer E, Scharrer B. 1940. Secretory cells within the hypothalamus. In: Fulton JF, Ranson SW, Frantz AM, Editors. The hypothalamus and central levels of autonomic function. Baltimore: Wiliams & Wilkins. Schmued LC, Beltramino C, Slikker J, William. 1993. Intracranial injection of Fluoro-Gold results in the degeneration of local but not retrogradely labeled neurons. Brain Research 626:71-77. Shepherd GM. 1994. Neurobiology. New York: Oxford University Press. Shioda S, Nakai Y. 1993. Medullary synaptic inputs to thyrotropin-releasing hormone (trh)- containing neurons in the hypothalamus: An ultrastructural study combining WGA-HRP anterograde tracing with TRH immunocytochemistry. Brain Res 625:9-15. Shioda S, Nakai Y, Tadokoro Y, Matsumoto K. 1989. Electron microscopic observations of medullary adrenergic synaptic inputs to the TRH -containing neurons in the rat hypothalamus. J Clin. Electron Microscopy 22:5-6. Shiosaka S. 1992. Ontogeny of the central somatostatinergic system. In: Bjorklund A, Hokfelt T, Tohyama M, Editors. Ontogeny of transmitters and peptides in the CNS. Elsevier. p 369- 398. Silverman AJ, Oldfield B, Hou-Yu A, Zimmerman EA. 1985. The noradrenergic innervation of vasopressin neurons in the paraventricular nucleus of the hypothalamus: An ultrastructural study using radioautography and immunocytochemistry. Brain Res 325:215-229. Silverman AJ, Witkin JW, Silverman RC, Gibson MJ. 1990. Modulation of gonadotropin- releasing hormone neuronal activity as evidenced by uptake of fluorogold from the vasculature. Synapse (New York, N.Y.) 6:154-160. Simerly R, Swanson L. 1988. Projections of the medial preoptic nucleus: A phaseolus vulgaris leucoagglutinin anterograde tract-tracing study in the rat. J Comp Neurol 270:209- 242. Simerly RB, Chang C, Muramatsu M, Swanson LW. 1990. Distribution of androgen and estrogen receptor mRNA-containing cells in the rat brain: An in situ hybridization study. J Comp Neurol 294:76-95. Simerly RB, Swanson LW. 1986. The organization of neural inputs to the medial preoptic nucleus of the rat. J Comp Neurol 246:312-342. Simmons DM. 2000. Cryoprotectant tissue storage solutions: Stability at lower temperatures, longer storage times, more versatile usage. Microscopy Today 8: no.3. Simmons DM, Arriza JL, Swanson LW. 1989. A complete protocol for in situ hybridization of messenger rnas in brain and other tissues with radiolabeled single-stranded RNA probes. J Histotech 12:169-181. Simmons DM, Swanson LW. 1993. The Nissl stain. Brain Res Prot 3-050-12:01-07. 312 Simmons DM, Voss JW, Ingraham HA, Holloway JM, Broide RS, Rosenfeld MG, Swanson LW. 1990. Pituitary cell phenotypes involve cell-specific pit-1 mRNA translation and synergistic interactions with other classes of transcription factors. Genes Dev 4:695-711. Sowers JR. 1980. Hypothalamic hormones. Stroudsburg, PA: Dowden Hutchinson & Ross. Spiess J, Rivier J, Thorner M, Vale W. 1982. Sequence analysis of a growth hormone releasing factor from a human pancreatic islet tumor. Biochemistry 21:6037-6040. Spiess J, Rivier J, Vale W. 1983. Characterization of rat hypothalamic growth hormone- releasing factor. Nature 303:532-535. Stellar E. 1954. The physiology of motivation. Psychol Rev 61:5-22. Strack AM, Sawyer WB, Hughes JH, Platt KB, Loewy AD. 1989. A general pattern of CNS innervation of the sympathetic outflow demonstrated by transneuronal pseudorabies viral infections. Brain Res 491:156-162. Strack AM, Sawyer WB, Platt KB, Loewy AD. 1989. CNS cell groups regulating the sympathetic outflow to adrenal gland as revealed by transneuronal cell body labeling with pseudorabies virus. Brain Res 491:274-296. Stricker E, Woods SC. 2004. Neurobiology of food and fluid intake. New York: Kluwer Academic/Plenum Publishers. Stricker EM, Verbalis JG. 2004. Inhibition of salt appetite in rats by central oxytocin. Am J Physiol Regul Integr Comp Physiol 287:R487; author reply R487-488. Swanson L, Hartman B. 1980. Biochemical specificity in central pathways related to peripheral and intracerebral homeostatic functions. Neurosci Lett 16:55-60. Swanson L, Sawchenko P, Berod A, Hartman B, Helle K, Vanorden D. 1981. An immunohistochemical study of the organization of catecholaminergic cells and terminal fields in the paraventricular and supraoptic nuclei of the hypothalamus. J Comp Neurol 196:271- 285. Swanson LW. 1986. Organization of mammalian neuroendocrine system. In: Bloom FE, Editor. Handbook of physiology, the nervous system. Baltimore, MD: Waverly Press. p pp. 317-363. Swanson LW. 1987. The Hypothlamus, Chapter 1. In: A Bjorklund TH, L W Swanson, Editor. Integrated systems of the CNS, Part I. Amsterdam: Elsevier Science Publishers. p 1-124. Swanson LW. 1991. Biochemical switching in hypothalamic circuits mediating responses to stress. Prog Brain Res 87:181-200. Swanson LW. 1992. Brain maps : Structure of the rat brain. Amsterdam ; New York: Elsevier. 313 Swanson LW. 1992. Spatiotemporal patterns of transcription factor gene expression accompanying the development and plasticity of cell phenotypes in the neuroendocrine system. Prog Brain Res 92:97-113. Swanson LW. 1998. Brain maps : Structure of the rat brain : A laboratory guide with printed and electronic templates for data, models, and schematics. Amsterdam ; New York: Elsevier. Swanson LW. 2000. A history of neuroanatomical mapping. In: Toga AW, Mazziotta JC, Editors. Brain mapping: The applications. San Diego: Academic Press. p 77-109. Swanson LW. 2001. Interactive brain maps and atlases. In: MA Arbib and JG Grethe, Editors. Computing the brain: A guide to Neuroinformatics. San Diego: Academic Press. p 167-177. Swanson LW. 2004. Brain Maps III: Structure of the Rat Brain : An atlas with printed and electronic templates for data, models, and schematics. Amsterdam ; Boston: Elsevier Academic Press. Swanson LW, Connelly MA, Hartman BK. 1977. Ultrastructural evidence for central monoaminergic innervation of blood vessels in the paraventricular nucleus of the hypothalamus. Brain Res 136:166-173. Swanson LW, Kuypers HG. 1980. The paraventricular nucleus of the hypothalamus: Cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J Comp Neurol 194:555-570. Swanson LW, Sawchenko PE. 1980. Paraventricular nucleus: A site for the integration of neuroendocrine and autonomic mechanisms. Neuroendocrinology 31:410-417. Swanson LW, Sawchenko PE. 1983. Hypothalamic integration: Organization of the paraventricular and supraoptic nuclei. Annu Rev Neurosci 6:269-324. Swanson LW, Sawchenko PE, Lind RW. 1986. Regulation of multiple peptides in crf parvocellular neurosecretory neurons: Implications for the stress response. Prog Brain Res 68:169-190. Swanson LW, Sawchenko PE, Rivier J, Vale WW. 1983. Organization of ovine corticotropin- releasing factor immunoreactive cells and fibers in the rat brain: An immunohistochemical study. Neuroendocrinology 36:165-186. Swanson LW, Sawchenko PE, Wiegand SJ, Price JL. 1980. Separate neurons in the paraventricular nucleus project to the median eminence and to the medulla or spinal cord. Brain Res 198:190-195. Swanson LW, Simmons DM. 1989. Differential steroid hormone and neural influences on peptide mRNA levels in CRH cells of the paraventricular nucleus: A hybridization histochemical study in the rat. J Comp Neurol 285:413-435. 314 Swerdlow NR, Swanson LW, Koob GF. 1984. Substantia innominata: Critical link in the behavioral expression of mesolimbic dopamine stimulation in the rat. Neurosci Lett 50:19-24. Tasker J. 2004. Endogenous cannabinoids take the edge off neuroendocrine responses to stress. Endocrinology 145:5429-5430. Tasker JG, Di S, Boudaba C. 2002. Functional synaptic plasticity in hypothalamic magnocellular neurons. In: Poulain D, Oliet S, Theodosis D, Editors. Vasopressin and oxytocin: From genes to clinical applications. Elsevier. Tasker JG, Dudek FE. 1991. Electrophysiological properties of neurones in the region of the paraventricular nucleus in slices of rat hypothalamus. J Physiol 434:271-293. Tasker JG, Dudek FE. 1993. Local inhibitory synaptic inputs to neurones of the paraventricular nucleus in slices of rat hypothalamus. J Physiol Lond 469:179-192. Tata JR. 2005. One hundred years of hormones. EMBO Rep 6:490-496. Ter Horst GJ, Luiten PG. 1987. Phaseolus vulgaris leuco-agglutinin tracing of intrahypothalamic connections of the lateral, ventromedial, dorsomedial and paraventricular hypothalamic nuclei in the rat. Brain Research Bulletin 18:191-203. The Salk Institute. 1985. Neuropeptides: Control of growth and the response to stress. In: Research report (Guilleman and Vale labs). La Jolla: Salk Institute for Biological Studies. p 1-6. Thompson RH, Swanson LW. 1996. Organization of projections from the dorsomedial nucleus of the hypothalamus. J Comp Neurol 376:143-173. Thompson RH, Swanson LW. 2003. Structural characterization of a hypothalamic visceromotor pattern generator network. Brain Res Brain Res Rev 41:153-202. Toni R, Jackson IM, Lechan RM. 1990. Thyrotropin-releasing-hormone-immunoreactive innervation of thyrotropin-releasing-hormone-tuberoinfundibular neurons in rat hypothalamus: Anatomical basis to suggest ultrashort feedback regulation. Neuroendocrinology 52:422-428. Toni R, Jackson IMD, Lechan RM. 1990. Neuropeptide-Y-immunoreactive innervation of thyrotropin-releasing hormone-synthesizing neurons in the rat hypothalamic paraventricular nucleus. Endocrinology 126:2444-2453. Tsuruo Y, Hokfelt T, Visser T. 1987. Thyrotropin releasing hormone (TRH)-immunoreactive cell groups in the rat central nervous system. Exp Brain Res 68:213-217. Vale W, Rivier C, Brown MR, Spiess J, Koob G, Swanson L, Bilezikjian L, Bloom F, Rivier J. 1983. Chemical and biological characterization of corticotropin releasing factor. Recent Prog Horm Res 39:245-270. 315 Vale W, Spiess J, Rivier C, Rivier J. 1981. Characterization of a 41-residue ovine hypothalamic peptide that stimulates secretion of corticotropin and beta-endorphin. Science 213:1394-1397. Vale W, Vaughn J, Sawchenko P, Siefert H, Bilezikjian L, Perrin M, Swanson L, Thorner MO, Rivier J. 1986. Chemical, anatomical, and physiological studies on human pancreatic and rat hypothalamic growth hormone-releasing factor. In: Raiti S, Tolman RA, Editors. Human growth hormone. Plenum. p 325-336. van den Pol AN. 1982. The magnocellular and parvocellular paraventricular nucleus of rat: Intrinsic organization. J Comp Neurol 206:317-345. Vandesande F, Dierickx K, Goossens N. 1980. Immunocytochemical localization of the posterior lobe hormones and of their carrier proteins. J Histochem Cytochem 28:469-471. Viau V, Chu A, Soriano L, Dallman MF. 1999. Independent and overlapping effects of corticosterone and testosterone on corticotropin-releasing hormone and arginine vasopressin mRNA expression in the paraventricular nucleus of the hypothalamus and stress-induced adrenocorticotropic hormone release. J Neurosci 19:6684-6693. Viau V, Lee P, Sampson J, Wu J. 2003. A testicular influence on restraint-induced activation of medial parvocellular neurons in the paraventricular nucleus in the male rat. Endocrinology 144:3067-3075. Voogt JL, Arbogast LA, Quadri SK, Andrews G. 1990. Tyrosine hydroxylase messenger RNA in the hypothalamus, substantia nigra and adrenal medulla of old female rats. Brain Res Mol Brain Res 8:55-62. Vrang N, Larsen P, Moller M, Mikkelsen J. 1995. Topographical organization of the rat suprachiasmatic-paraventricular projection. J Comp Neurol 353:585-603. Watts A, Swanson L. 1989. Combination of in situ hybridazion with immunohistochemistry and retrograde tract tracing. In: Methods in neurosciences: Gene probes. San Diego: Academic Press. p 127-136. Watts A, Swanson L. 1989. Diurnal variations in the content of preprocorticotropin-releasing hormone messenger ribonucleic acids in the hypothalamic paraventricular nucleus of rats of both sexes as measured by in situ hybridization. Endocrinology 125:1734-1738. Watts AG. 2001. Neuropeptides and the integration of motor responses to dehydration. Annu Rev Neurosci 24:357-384. Watts AG, Swanson LW, Sanchez-Watts G. 1987. Efferent projections of the suprachiasmatic nucleus: I. Studies using anterograde transport of phaseolus vulgaris leucoagglutinin in the rat. J Comp Neurol 258:204-229. Watts AG, Tanimura S, Sanchez-Watts G. 2004. Corticotropin-releasing hormone and arginine vasopressin gene transcription in the hypothalamic paraventricular nucleus of unstressed rats: Daily rhythms and their interactions with corticosterone. Endocrinology 145:529-540. 316 Weisenberh RC, Borisy GG, Taylor EV. 1968. The colchicine binding protein of mammalian brain and its relation to microtubules. Biochemistry 7:4466. Wen YQ, Zhu JN, Zhang YP, Wang JJ. 2004. Cerebellar interpositus nuclear inputs impinge on paraventricular neurons of the hypothalamus in rats. Neurosci Lett 370:25-29. Whitnall MH. 1988. Distributions of pro-vasopressin expressing and pro-vasopressin deficient CRH neurons in the paraventricular hypothalamic nucleus of colchicine-treated normal and adrenalectomized rats. J Comp Neurol 275:13-28. Winokur A, Utiger RD. 1974. Thyrotropin-releasing hormone: Regional distribution in rat brain. Science 185:265-267. Wittmann G, Lechan RM, Liposits Z, Fekete C. 2005. Glutamatergic innervation of corticotropin-releasing hormone- and thyrotropin-releasing hormone-synthesizing neurons in the hypothalamic paraventricular nucleus of the rat. Brain Res 1039:53-62. Wittmann G, Sarkar S, Hrabovszky E, Liposits Z, Lechan RM, Fekete C. 2004. Galanin- but not galanin-like peptide-containing axon terminals innervate hypophysiotropic TRH- synthesizing neurons in the hypothalamic paraventricular nucleus. Brain Res 1002:43-50. Xi D, Kusano K, Gainer H. 1999. Quantitative analysis of oxytocin and vasopressin messenger ribonucleic acids in single magnocellular neurons isolated from supraoptic nucleus of rat hypothalamus. Endocrinology 140:4677-4682. Young LJ, Wang Z, Insel TR. 1998. Neuroendocrine bases of monogamy. Trends Neurosci 21:71-75. Zhang BJ, Kusano K, Zerfas P, Iacangelo A, Young WS, 3rd, Gainer H. 2002. Targeting of green fluorescent protein to secretory granules in oxytocin magnocellular neurons and its secretion from neurohypophysial nerve terminals in transgenic mice. Endocrinology 143:1036-1046. Zimmerman EA. 1976. Localization of hypothalamic hormones by immunocytochemical techniques [chapter 2]. In: Martini L, Ganong WF, Editors. Frontiers in neuroendocrinology. New York: Raven Press. p 25-62. Zimmerman EA. 1981. The organization of oxytocin and vasopressin pathways. Adv Biochem Psychopharmacol 28:36-75. Zimmerman EA, Nilaver G, Hou-Yu A, Silverman AJ. 1984. Vasopressinergic and oxytocinergic pathways in the central nervous system. Federation Proceedings 43:91-96. Zimmerman EA, Robinson AG. 1976. Hypothalamic neurons secreting vasopressin and neurophysin. Kidney International 10:12-24. Zupanc G. 1996. Peptidergic transmission: From morphological correlates to functional implications. Micron 27:35-91. 317 APPENDIX I Anatomical History of the PVH Changes in anatomical concepts of the paraventricular nucleus of the hypothalamus (PVH) since the nineteenth century reflect insightful observation during several eras of methodological advance in histologic technique. Known clinical phenomena, current experimental data and theoretical hypotheses were correlated in parallel with each new method for visualizing structure and implying function in the nervous system. Thus, progress from recognition of a small group of Nissl-stained cells with no obvious function or connectivity, to the concept of PVH as a crucial hypothalamic center with complex substructure reflecting endocrine and autonomic control and integration of behaviors essential for survival has depended on coordinate advances in physiological, chemical, and anatomical Neuroscience. Santiago Ramon y Cajal (5) made specific reference to the “paraventricular (subventricular) nucleus” seen in Nissl-stained material, recognizing its uniqueness, even though he was unable to obtain adequate Golgi impregnations to reveal its fine structure and possible connectivity. Other workers in the late nineteenth and early twentieth century observed the constancy of cellular appearance and general location of PVH in numerous species (17). And there seemed to be consensus that the PVH and supraoptic (SO) nuclei contained similar cells, which in some lower species formed a continuum and were thought to have a common, possibly hormonal, function (4). One reason for lack of early attention may have been that the PVH in human is less obvious than in the rat, and did not appear strikingly organized in a region (the hypothalamus) that itself seemed lacking in cellular patterns and obvious fiber bundles when compared to cortex, cerebellum or brainstem. An early name for PVH was nucleus filiformis (filamentous or threadlike) hypothalami, possibly due to its appearance in the human as a band of strongly staining cells along the margin of the third ventricle. By 1938, Le Gros Clark (8) listed PVH along with two synonyms: n. filiformis and n. magnocellularis, and Gurdjian had reportedly divided the rat PVN into a medial parvocellular and lateral magnocellular part in 1927(17)(6), while Ingram (21) described mixing of smaller cells medially near the periventricular area in the primate PVH. Both Ingram and Scharrer (36) noted in 1940 that n. filiformis represents the medial part of PVN, while the lateral part is what had been separately called n. magnocellularis. It is interesting that these older terms stayed in use even into the sixties, with Krieg illustrating them three-dimensionally in human material (22) and Szentagothai showing anterior nucleus 318 magnocellularis separate from the PVN illustrated in his rat atlas (43). His anterior n. magnocellularis seems to refer to the modern PVH periventricular zone along with all periventricular cells along the wall of the third ventricle, and while he shows no named subdivisions of PVH he does indicate a medial-lateral distinction by a dotted line on his drawings. Clinical observations dating from the mid nineteenth century of various endocrine and autonomic conditions linked to tumors or hypertrophy of the hypothalamic area (e. g., Frolich’s syndrome (15)) prompted many experimental investigations, which heightened during the 1920s and ‘30s. Oliver Strong, in Frederick Bailey’s 1921 Textbook of Histology correlated physiological and neuropathological studies to describe “splanchnic” connective pathways that allow emotional states to cause involuntary changes in the vegetative system -- speculative interpretation at the time, but based on sound observations (2). A close association between pituitary and hypothalamus was realized from embryological studies that showed the posterior pituitary is an outgrowth of the embryonic hypothalamus (26). However, the mode of physiological interaction was not apparent, and most studies involved observations of endocrine effects from hypophysectomy and injection of pituitary extracts (43). Work with cortical and hypothalamic separations by knife cut, lesions to hypothalamic areas, and observations of reaction to electrical stimulation in the hypothalamus proved the involvement of hypothalamus in sympathetic autonomic function. And lesion experiments by Bard showing ‘sham rage’ (activation of sympathetic somatomotor responses) in decorticate cats, indicated a behavioral component to hypothalamic control of autonomic function (16). Indeed, as late as 1972 the PVH was defined merely as “related to the autonomic nervous system” in Stedman’s Medical Dictionary (46). Degenerated or chromatolytic cells in PVH were seen by several workers (8)(16) after brainstem lesions (dorsal vagal nucleus punch), sympathetic ganglia extirpation (noted in 1898!), and hypophysectomy, and pathways were traced to the midbrain and spinal cord from hypothalamic lesions that created autonomic deficit. Weigert stains for myelin, Marchi osmium treatment for degenerating myelin, and Cajal neurofibrillary stains were used to study connectivity in the hypothalamus. They revealed a complex and confusing meshwork of fine fibers whose afferent and efferent connections to specific cell groups could only be inferred very generally from clinical and experimental data. It remained for the specialized silver degeneration methods of Nauta in the ‘50s and ‘60s to define some precise efferent connectivity from experimental lesions. One important approach to the definition of PVH structural organization came from the extensive work of Ernst and Berta Scharrer (34-36). They noted and then documented in 319 numerous species the gland-like appearance of certain neurons based on their staining properties with the van Gieson method. These studies revealed the presence of protein, colloid-like droplets similar to those seen in thyroid and other glandular cells. They realized that unlike other cells in the nervous system, these magnocellular neurons were the only ones seen to vary histologically over time and with changing physiological condition of the organism, and their unusually close association with blood vessels implied a high metabolic activity (10). They felt these neurons must have a unique function and that their protein content was unlikely to be a mere metabolic byproduct. This led to the prediction that these colloid-like substances were likely secreted from the neurons with functional consequence -- as with substances secreted from gland cells elsewhere. Their hypothesis that PVH and SO neurons are “neurosecretory” has proven quite correct in principle, and is in fact a fundamental element in our understanding of neuroendocrine function. This discovery seems all the more brilliant since it was based on inferences from careful observation of histological preparations without benefit of modern knowledge of axoplasmic transport. They and later workers (3) used the Gomori stain to trace droplets of colloid substance along the axons of neurons coursing to the posterior pituitary, and to describe the intimate relation of pituicytes with the nerve terminals that release the colloid substance. The Sharrers realized that their hypothesis would be difficult to prove, and mentioned at a symposium in 1939 that one challenge would be to determine what effect would be meaningful to observe in experiments of injected hypothalamic extracts containing putative neurosecretory products (36). Identification and synthesis of oxytocin (Oxy) and vasopressin (Vas) in the 1950s provided a key tool for investigation of autonomic and endocrine function. Experiments to test their presence in posterior pituitary and in hypothalamic neurons were combined with physiological studies of cardiovascular action to prove the function of neurosecretory magnocellular neurons in the PVH. Measurement of Oxy and Vas release into the circulation became a powerful bioassay for subsequent studies in physiological endocrinology -- David Lincoln’s classic studies of Oxy cell activation in the milk ejection reflex is but one example (23). During the 1940s Geoffrey Harris, who had previously shown changes in pituitary release of reproductive hormones due to medial hypothalamic stimulation, conducted a brilliant series of experiments (some in living animals) to demonstrate the patterns of blood flow associated with the pituitary gland. By proving a flow of arterial blood through a portal capillary network from the median eminence to the anterior pituitary, he illustrated how the poorly innervated anterior pituitary could be affected by releasing factors from hypothalamic 320 axon terminals in the median eminence and posterior pituitary. Thus the connection between neurosecretion by PVH cells (other than those projecting only to the posterior pituitary) and observed hypothalamic control of pituitary function was made possible at an anatomical level. His work was also doubted at first, but his insights have proven pivotal to modern understanding of pituitary control by the PVH (33). In fact, in 1955 he predicted the ensuing search for multiple hypothalamic pituitary releasing factors that characterized experimental endocrinology well into the early 1980s(18). Findings of efferent autonomic pathways, probable neurosecretion by hypothalamic axons, and elucidation of the portal vasculature to deliver these secretions to the anterior pituitary completed the conceptual model for neural control of both the endocrine and autonomic systems. The stage was set for identification of specific cell group effectors, and interest in anatomy of the PVH was building toward a more modern subdivision based on both anatomical and functional considerations. In the 1950s and early ‘60s methods for histofluorescent identification of monoamine containing cells and fibers and of reduced silver methods for fine degenerating fibers provided two powerful tools for correlation of cytoarchitectonics and possible function and connectivity in the PVH. For example, identification of rich innervation of PVH by cholinergic fibers arising from brainstem cell groups (11) was correlated with physiological experiments where amines were injected in areas known to elicit pituitary hormone release (33). And evidence of PVH afferents, including some that might define behavioral activation circuits was compiled using the Nauta silver stains following small lesions (24). In the late sixties, powerful techniques were devised to exploit the phenomenon of axoplasmic transport of tracers without significant damage to the areas under study. Thus, cells could be retrogradely labeled from their terminal fields with injections of macromolecules like horseradish peroxidase (29), and conversely, terminal arborizations of cell groups could be labeled via injection of radioactive amino acid protein precursors among their cell bodies (9). Thus, the source and trajectory of defined fibers could be traced. Introduction in the seventies of immunohistochemical (IHC) identification of neurons and fibers containing chemically specified protein antigens (39)(45) marked a major revolution in correlating function with anatomical subdivisions in PVH -- especially when used in combination with improved tracing techniques (30)(31). Antibody staining expanded the visualization of specific fine fiber pathways (previously invisible or indiscernible by classical methods) that was begun with the use of amine histofluorescence. The confusing meshwork of fibers in the hypothalamus was seen to have chemical (thus functional) organization, and histochemical identification of specific peptides (45), enzymes (19) and putative 321 neurotransmitters (14) has since defined modern Chemical Neuroanatomy. Correlation of information from several methods is quite powerful. For example, early staining of Oxy fibers in brainstem and spinal cord specifically implicated the PVH and/or SO nuclei in well known hypothalamic autonomic function, since they were the sole known source of oxytocinergic neurons (40). A pattern of magnocellular projections to posterior pituitary, medial parvicellular projections to median eminence, and dorsal and ventral parvicellular projections to brainstem and spinal cord based on anterograde and retrograde transport studies (20)(28, 28a-abstract only)(25)(41) and IHC (7)(12)(13) began to emerge. In the early 1980s anatomical subdivisions of PVH were reviewed by several authors (1)(37)(38), in an attempt to correlate cytoarchitectonic cell types and groupings visible in a good Nissl stain with modern functional and experimental anatomical information. Three interesting new ideas about the location, boundaries and basic structural organization of PVH resulted: the subdivision of two descending parvicellular cell groups, inclusion of periventricular cells of the parvicellular division as a functional part of neuroendocrine PVH that projects to the median eminence, and inclusion of anterior and medial magnocellular subdivisions as a partial continuum of the posterior magnocellular division projecting to the posterior pituitary. The eight major subdivisions described by Swanson are now generally well accepted. Their boundaries have remained amazingly stable for almost two decades, as more detailed information about PVH cell morphology (44)(27), receptor and neurotransmitter content (42), and afferent innervation (32) has accumulated. There have been refinements, but no obvious change of boundaries visible to the educated eye in Nissl preparations (see above references). Two examples: Characterization of descending projections and evidence for collateralization was provided by the use of multiple injections of fluorescent retrograde tracers introduced by Kuypers, et al. It became apparent that the dorsal parvicellular group of descending neurons constituted an almost exclusive population of sympathetic spinal cord projections, while the ventral group sent a combination of parasympathetic and sympathetic projections to midbrain, brainstem and spinal cord. Oxy and Vas neurons are seen to segregate generally, but not completely within magnocellular divisions, but the trajectory of their projection to the posterior pituitary shows no major divergence. As hypothalamic releasing factors for anterior pituitary function were discovered and antibodies and molecular probes became available, the function of periventricular and medial parvicellular subdivisions in anterior pituitary control was shown, but no obvious further subdivision is yet indicated, since pure populations of one chemical subtype have not been shown. Afferent innervation seems to correspond generally to the basic subdivisions, 322 with some preference between Oxy and Vas magnos in certain cases (31), which likely correlates with their different but related physiological effects. It is possible that afferent innervation by certain types of neurotransmitter fibers, or the presence of yet-to-be- discovered agents may delineate new functional subdivisions in future studies -- especially when considered in tandem with patterns of receptor localization. It remains to be seen if more sensitive tracing methods and molecular biological techniques introduced in the ‘80s and ‘90s (e.g. in situ hybridization, immediate early gene activation, and gene knockout studies) will prompt another round of reevaluation of normal PVH structural organization. And perhaps new methods will reveal functional correlations with afferent and efferent connections that were previously enigmatic. Judging from surprising findings of the past, the future of anatomical research concerning the PVH is likely to remain interesting. Appendix I References 1. Armstrong, WE, Warach, S, Hatton, GI and McNeill, TH, Subnuclei in the rat hypothalamic paraventricular nucleus: a cytoarchitectural, horseradish peroxidase and immunocytochemical analysis. Neuroscience, 1980. 5(11): p. 1931-58. 2. Bailey, Frederick R, A Textbook of Histology. sixth revised ed. 1921, New York: William Wood. 3. Bargmann, W, Uber die Neurosekretorische Verknupfung von Hypothalamus und Neurohypophyse. Z. Zellforsch. Mikrosk. Anat., 1949. 34: p. 610-634. 4. Butler, Ann B and Hodos, William, Comparative Vertebrate Neuroanatomy: Evolution and Adaptation. 1996, New York: Wiley-Liss. 514. 5. Cajal, Santiago Ramón y, Histology of the Nervous System of Man and Vertebrates, Vol.II, History of Neuroscience. Vol. No. 6. 1995, New York: Oxford University Press. 806 pp. 6. Ceccatelli, S, Eriksson, M and Hokfelt, T, Distribution and Coexistence of Corticotropin- Releasing Factor-, Neurotensin-, Enkephalin-, Cholecystokinin-, Galanin- and Vasoactive Intestinal Polypeptide/Peptide Histidine Isoleucine-Like Peptides in the Parvocellular Part of the Paraventricular Nucleus. Neuroendocrinology, 1989. 49: p. 309-323. 7. Choy, VT and Watkins, WB, Immunocytochemical study of the hypothalamo- neurohypophysial system. II. Distribution of neurophysin, vasopressin and oxytocin in the normal and osmotically stimulated rat. Cell Tiss. Res., 1977. 180: p. 467-490. 8. Clark, W E, Le Gros, Morphological Aspects of the Hypothalamus, in [Four Lectures on] The Hypothalamus: morphological functional, clinical and surgical aspects, Le Gros Clark, W., Beattie, J., Riddoch, G. and Dott, N.M., Editors. 1938, Oliver and Boyd: Edinburgh. p. 1-68. 323 9. Cowan, WM, Gottlieb, DI, Hendrickson, AE, Price, JL and Woolsey, TA, The autoradio- graphic demonstration of axonal connections in the central nervous system. Brain Res.,1972.37: p. 21-51. 10. Craige, E Horne, Measurements of Vascularity in some Hypothalamic nuclei of the Albino Rat, in The Hypothalamus and Central Levels of Autonomic Function, Fulton, J.F., Ranson, S.W. and Frantz, A.M., Editors. 1940, Williams & Wilkins: Baltimore. p. 310-319. 11. Dahlstrom, A. and Fuxe, K., Evidence for the Existence of Monoamine-containing Neurons in the Central Nervous System. I. Demonstration of Monoamines in the Cell Bodies of Brainstem Neurons. Acta Pysiol. Scand. (Suppl. 232), 1964. 62: p. 1-55. 12. Defendini, R and Zimmerman, EA, The magnocellular neurosecretory system of the mammalian hypothalamus. Res Publ Assoc Res Nerv Ment Dis, 1978. 56: p. 137-54. 13. Dierickx, K, Immunocytochemical localization of the vertebrate nonapeptide neurohypophyseal hormones and neurophysins. Int. Rev. Cytol., 1980. 62: p. 119-185. 14. Elde, R and Hokfelt, T, Localization of hypophysiotropic peptides and other biologically active peptides within the brain, in Ann. Rev. Physiol. 1979, Annual Reviews, Inc. p. 587- 602. 15. Frolich, Alfred, Ein Fall von Tumor dir Hypophysis Cerebri Ohne Akromegalie, in The Hypothalamus and Central Levels of Autonomic Function, Fulton, l.J.F., Editor. 1940, Hafner: New York. p. xvii-xxviii. 16. Fulton, John F, Historical Resume, in The Hypothalamus and Central Levels of Autonomic Function, Fulton, J.F., Editor. 1940, Hafner: New York. p. xiii-xvi. 17. Gurdjian, E S, The diencephalon of the albino rat. J. Comp. Neurol, 1927. 43(1): p. 1- 114. 18. Harris, Geoffrey, Neural Control of the Pituitary Gland. 1955, London: Edward Arnold. 19. Hartman, BK, Zide, D and Udenfriend, S, The Use of dopamine-ß hydroxylase as a marker for the central noradrenergic nervous system in the rat brain. Proc. nat. Acad. Sci. (Wash.), 1972. 69: p. 2722-2726. 20. Hatton, GI, Hutton, UE, Hoblitzell, ER and Armstrong, WE, Morphological evidence for two populations of magnocellular elements in the rat paraventricular nucleus. Brain Res, 1976. 108(1): p. 187-93. 21. Ingram, W.R., Nuclear Organization and Chief Connections of the Primate Hypothalamus (Chapter V), in The Hypothalamus and Central Levels of Autonomic Function, Fulton, l.J.F., Editor. 1940, Hafner: New York. p. 195-244. 22. Krieg, Wendell J. S., Functional Neuroanatomy. Third revised and Enriched ed. 1966, Bloomington IL: W J S Krieg. 324 23. Lincoln, DW and Wakerley, JB, Factors governing the periodic activation of supraoptic and paraventricular neurosecretory cells during suckling in the rat. J Physiol (Lond), 1975. 250(2): p. 443-61. 24. Nauta, W J H and Haymaker, W, Hypothalamic nuclei and fiber connections, in The Hypothalamus, Haymaker, W., Nauta, W.J.H. and Anderson, E., Editors. 1969, Thomas: Springfield, IL. p. 136-209. 25. Ono, T , Nishino, H , Sasaka, K , Muramoto, K , Yano, I and Simpson, A , Paraventricular Nucleus Connections to Spinal Cord and Pituitary. Neuroscience Letters, 1978. 10: p. 141-146. 26. Papez, James P, The Embryological Development of the hypothalamic area in Mammals (Chapter II), in The Hypothalamus and Central Levels of Autonomic Function, Fulton, l.J.F., Editor. 1940, Hafner: New York. p. 3-29. 27. Rho, JH and Swanson, LW, A morphometric analysis of functionally defined subpopulations of neurons in the paraventricular nucleus of the rat with observations on the effects of colchicine. J Neurosci, 1989. 9(4): p. 1375-88. 28. Ricardo, JA and Koh, ET, Anatomical evidence of direct projections from the nucleus of the solitary tract to the hypothalamus, amygdala, and other forebrain structures in the rat. Brain Res, 1978. 153(1): p. 1-26. 28a. Koh EJ, Ricardo JA. 1980. Paraventricular nucleus of the hypoothalamus: Anatomical evidence of ten functionally discrete subdivisions. Soc, Neurosci. Abstr. 6:521. 29. Saper, CB, Loewy, AD, Swanson, LW and Cowan, WM, Direct hypothalamic autonomic connections. Brain Res., 1976. 117: p. 305-312. 30. Sawchenko, PE, Swanson, LW, Steinbusch, HW and Verhofstad, AA, The distribution and cells of origin of serotonergic inputs to the paraventricular and supraoptic nuclei of the rat. Brain Res, 1983. 277(2): p. 355-60. 31. Sawchenko, P.E and Swanson, L.W., The organization and biochemical specificity of afferent projections to the paraventricular and supraoptic nuclei, in Progress in Brain Research. 1983, Elsevier: Amsterdam. p. 19-29. 32. Sawchenko, P.E. and Swanson, L.W., The organization of forebrain afferents to the paraventricular and supraoptic nuclei of the rat. J Comp Neurol, 1983. 218(2): p. 121-44. 33. Sawyer, Charles H, History of the Neurovascular Concept of Hypothalamo-Hypophysial Control. Biology of Reproduction, 1978. 18: p. 325-328. 34. Scharrer, Berta and Scharrer, Ernst, A Comparison Between the intercerebralis- cardiacum-allatum system of insects and the hypothalamo-hypophyseal system of the vertebrates. Biol Bull, 1944. 87: p. 142-251. 35. Scharrer, E, The General Significance of the Neurosecretory Cell. Scientia (Milan), 1952. 87: p. 176-182. 325 36. Scharrer, Ernst and Scharrer, Berta, Secretory cells within the hypothalamus, in The Hypothalamus and Central Levels of Autonomic Function, Fulton, l.J.F., Editor. 1940, Hafner: New York. p. 170-194. 37. Swanson, LW and Kuypers, HG, The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J Comp Neurol, 1980. 194(3): p. 555-70. 38. Swanson, LW and Sawchenko, PE, Paraventricular nucleus: a site for the integration of neuroendocrine and autonomic mechanisms. Neuroendocrinology, 1980. 31(6): p.410-7. 39. Swanson, L W and Hartman, B K, The central adrenergic system. An immuno- fluorescence study of the location of cell bodies and their efferent connections in the rat utilizing dopamine-ß hydroxylase as a marker. J. Comp. Neurol., 1975. 163: p. 467-506. 40. Swanson, L.W., and McKellar, S., The distribution of oxytocin- and neurophysin-stained fibers in the spinal cord of the rat and monkey. J. Comp. Neurol., 1979. 188: p. 87- 106. 41. Swanson, L W, Sawchenko, P E, Wiegand, S J and Price, J L, Separate neurons in the paraventricular nucleus project to the median eminence and to the medulla or spinal cord. Brain Res, 1980. 198(1): p. 190-5. 42. Swanson, L W and Simmons, D M, Differential Steroid Hormone and Neural Influences on Peptide mRNA Levels in CRH Cells of the Paraventricular Nucleus: A Hybridization Histochemical Study in the Rat. J. Comp. Neurol., 1989. 285: p. 413- 435. 43. Szentagothai, Janos, Flerko, Bela, Mess, Bela and Halasz, Bela, Hypothalamic control of the Anterior Pituitary. Third ed. 1968, Budapest: Akademiai Kiado. 44. Van den Pol, A.N., The magnocellular and parvocellular paraventricular nucleus of the rat: Intrinsic organization. J. comp. Neurol., 1982. 206: p. 317-345. 45. Vandesande, F, Dierickx, K and Demey, J, The Origin of the vasopressinergic and oxytocinergic fibers of the external region of the median eminence of the rat hypophysis. Cell Tissue Res., 1977. 180: p. 443-452. 46. Williams&Wilkins, Stedman's Medical Dictionary, Illustrated. 22 ed. Stedman's Medical Dictionary, ed. Anne G. Cutler, S.M.E. 1972, Baltimore: Williams & Wilkins. 1533. 326 APPENDIX II Afferent Inputs to PVH and their Significance The paraventricular nucleus of the hypothalamus (PVH) is “a small though highly differentiated region that in essence constitutes the final common pathway for the neuroendocrine system, and has important inputs to the autonomic motor system as well.”(1). Endocrine and autonomic function of the hypothalamus was inferred by clinicians and experimenters early in this century (2), and modern neuroanatomical methods have shown PVH to be the key central structure mediating these functions. Endocrine control via synaptic release of neurohormones into the posterior pituitary (3), and of anterior pituitary control factors into the portal circulation of the median eminence is the obvious primary PVH function (4). A co-ordinate function is modulation of autonomic responses to visceral sensory input in order to integrate endocrine system function with that of other (e.g., cardiovascular, digestive, immune) physiological systems (5)(6)(7)(8). This is accomplished largely via descending efferents to visceromotor brainstem nuclei and to sympathetic and parasympathetic preganglionic neurons in the spinal cord (9)(10)(11). A more subtle, but quite crucial third function of the PVH is to provide a focal connection to circuits producing (goal oriented) behaviors that serve to maintain homeostasis and adapt to changes in the external environment (12). Many behaviors key to survival (ingestive, stress-response, etc.) can be directly related to PVH activity (13)(14). Among these are foraging or avoidance behaviors linked to evaluation of the organism’s internal and external environment - such as the search for food, water, or safety. Also crucial are metabolically stimulated reflex behaviors like shivering or panting responses to temperature change (15), and social behaviors mediated by endocrine factors e.g., aggression, grooming, care of young, and reproduction (7)(16)(17). These behaviors, traditionally thought to be due to activation of medial or lateral hypothalamic ‘centers’ for feeding, aggression, etc. (18)(19) have been shown rather to rely on visceral and forebrain input to the PVH for their expression (1). Integration of behavior with endocrine and autonomic function implies a complex circuitry that allows access to behavioral state (nutritional status, metabolic rhythms, the sleep-wake cycle), cognitive (memory, decision-making, affect or emotional tone), and viscero-sensory (cardiovascular, metabolic, digestive) information (1). Discussion of afferent connections detailed below will illustrate how neural inputs to PVH may co-ordinate endocrine and autonomic function. In analogy to the “sensory input/motor output” model of functional neural circuits, the PVH can be thought of as the motor output nucleus of the 327 neuroendocrine viscero-autonomic system. In order to encourage a systems approach to understanding their function and connectivity, PVH neurosecretory neurons have in fact been termed ‘neuroendocrine motoneurons’ by L.W. Swanson, a term that has gained widespread acceptance in the past decade. This concept (a functional one) will guide the following analysis of inputs to PVH. I will deal here with known neural afferents, although another type of probable important input (humoral) deserves mention. The PVH is perhaps the most richly vascularized area of the brain (20), and therefore seems almost certainly to be affected by changes in circulating levels of a variety of blood-borne chemicals. Obvious examples of substances that cross the blood-brain barrier with profound functional consequences are gonadal and adrenal steroids, for which PVH cells are known to express receptors (21). Circulating corticosterone, for example, has been shown to exert negative feedback control on corticotropin releasing hormone (CRH) synthesis in PVH parvicellular neurons of the rat (22), and is thus a significant modulating factor in the stress response. It is also quite possible that changes in PVH capillary permeability under varying physiologic conditions could alter sensitivity to circulating messenger molecules (23). Since microvascular control is known to be largely autonomic in origin (24)(25), and highly correlated with some primary PVH functions (e.g., body water homeostasis and cardiovascular function), the significance of possible PVH responsiveness to subtle humoral signals is without doubt interrelated with specific neural inputs. The paraventricular nucleus can be divided into three magnocellular and five parvicellular subdivisions based on spatial and cytoarchitectonic features (26)(27)(11). The defined anatomical subdivisions correspond well with neuropeptide content that relates to endocrine or autonomic function (11). Thus, magnocellular subdivisions are composed of neurons that contain Oxytocin or Vasopressin (anti-diuretic hormone) and project to the posterior pituitary where those substances are released from their terminals into the general circulation. Vasopressin is a regulator of fluid balance via vasoconstriction and alterations in kidney metabolism, and Oxytocin induces smooth muscle contraction -- most notably evidenced during parturition (28) and lactation in the female (29). Dye transfer studies indicate magnocellular neurons contain gap junctions, implying electrical coupling which may well be the mechanism of observed coordinated firing that produces pulsatile release of Oxytocin during lactation (30). Periventricular and medial parvicellular parts of PVH control anterior pituitary function via release of peptide products into the portal vasculature of the median eminence, though there is some evidence for collateral projections to nearby 328 hypothalamic structures (31). Hypophysiotropic molecules (e.g., thyroid hormone releasing hormone, corticotropin releasing hormone, somatostatin/growth hormone inhibiting factor, and dopamine: TRH, CRH, SS, DA) are relatively well segregated in parvicellular subdivisions (personal observations). Dorsal parvicellular and ventral medial parvicellular subdivisions contain autonomic efferent cells. They project to spinal cord (sympathetic and parasympathetic preganglionic neurons) and brainstem (visceral-sensory and visceromotor cell groups) respectively (11). Descending autonomic PVH neurons do not appear to send axon collaterals to the median eminence or posterior pituitary, although they do contain at least some of the same neuropeptides as other PVH neurons (notably, Oxytocin and Vasopressin)(32). PVH neurons have also been shown to co-localize a variety of other neurochemicals not known to be pituitary hormone releasing factors (e.g. neurotensin, enkephalin, cholecystokinin, galanin, etc. (33)), and some of these may prove to be autonomic neurotransmitters. There is virtually no evidence for interneurons in the PVH, although dendritic arbors within the nucleus (34) suggest that there could be some synaptic connectivity between subdivisions, and recent physiological studies indicate that GABA inputs to areas immediately adjacent to PVN may have an inhibitory effect on PVN neurons (35)(36). This fact may be important, considering some hypothalamic afferents preferentially innervate the PVH surround. As detailed below, afferent inputs to PVH tend to parcellate into different subdivisions within the nucleus, and in some cases onto specific cell types (12). The obvious implication is that different neural inputs (or combinations of them) subserve specialized functional outputs of those subdivisions or cell types. It seems logical therefore to structure the following survey of inputs according to common PVH targets, and group them according to ascending inputs from the brainstem and midbrain, descending inputs from telencephalic structures, and converging inputs from the diencephalon. Magnocellular neurosecretory neurons were the first recognized and have been the most studied PVH cells (37). Pulsatile Oxytocin release has been shown to be directly related to parturition and lactation in the female, though its effect on smooth muscle contraction in the male is less well understood. Cardiovascular response to hemorrhage and dehydration is mediated largely by Vasopressin release in response to changes in blood volume and osmolarity, while peripheral or central injection of Angiotensin II (AII, a kidney hormone produced in response to dehydration) also elicits Vas release and immediate drinking behavior (38). Oxy cells are preferentially innervated by serotonergic projections from several midbrain raphe nuclei (39) whereas Vas cells receive strong noradrenergic 329 inputs from the A1 nucleus in the brainstem (40). The raphe nuclei are considered to be locomotor-related structures, and A1 relays information (e.g., baroreceptor, and chemosensory) from the vagus and glossopharyngeal nerves, and presumably from thoracic dorsal root ganglia. Magnocellular neurons (predominantly Vas) also receive afferents from the subfornical organ (SFO), which lacks a blood brain barrier, and responds electrically to changes in circulating levels of AII (41). In addition, Vas and Oxy neurons in the PVH receive an equal projection from median preoptic (MePO) cells that are themselves strongly innervated by the SFO. The MePO is responsible for a complex of reproductive activities that are known to include changes in fluid balance (Vas) and Oxytocin release. Both types of magnocellular neurons also receive a sparse projection from the suprachiasmatic nucleus (SCN), the neural circadian rhythm generator. This input probably coordinates Oxy and Vas release with the circadian rhythms of other hormones. A few inputs unique to Oxy cells deserve mention, though their significance is not entirely clear: Dorsal medial hypothalamus (DMH) and the arcuate nucleus (ARC) send specific projections to Oxy cells, and the bed nucleus of the stria terminalis (BST) also seems to innervate magnocellular Oxy cells preferentially, though not very heavily. The ARC fibers have been shown to contain ACTH and beta-endorphin, DMH is considered broadly to be associated with feeding, and the BST is considered a funnel for limbic information. The significance of these afferents may relate to Oxy release observed in response to limbic stimulation, and to the calmed affect and increased feeding associated with lactation. A physiological study of response to limbic stimulation showed a convergence of inhibitory limbic input on individual neurons, and that neurosecretory cells, which were influenced by limbic stimuli, were also inhibited by baroreceptor activation and excited by osmotic stimulation (42). Though interesting, the significance of this data is not entirely clear, since the cells were not identified as containing Oxy or Vas. In general, afferent projections to parvicellular neurons are diffuse and fairly similar over the subdivisions, with a few interesting exceptions. Hypothalamic inputs are fairly sparse and distributed over all parvicellular subdivisions, with DMH having a notably denser innervation, while extremely sparse SCN inputs are directed primarily at the ventral periventricular portion, through which traverse axons projecting to the neurohemal zone of the median eminence (43). Midbrain serotonergic inputs are fairly localized to periventricular and dorsal autonomic (predominantly spinal sympathetic-projecting) subdivisions (39). A similar pattern is seen with innervation by ACTH/beta endorphin fibers from the arcuate nucleus, except that it tends to innervate both autonomic compartments (44). Since 330 autonomic and neuroendocrine parvicellular subdivisions receive generally similar afferents, they will be considered together in discussion of projections from different CNS areas -- with comments on significant differences. Brainstem and midbrain projections to parvicellular PVH include adrenergic and noradrenergic fibers from the nucleus of the solitary tract and other dorsal vagal complex nuclei, from locus ceruleus, parabrachial nucleus, periaqueductal gray matter and the laterodorsal tegmental nucleus. All of these structures relay primary or secondary viscero- sensory information, and some of them in turn are in receipt of locomotor and descending behavioral information (45). Serotonergic innervation from raphe nuclei innervates autonomic divisions more strongly (as noted above), and is low in PVH relative to the immediate surround, a fact that may relate to presence of local inhibitory neurons in that area (36). This pattern is somewhat complementary to adrenergic fibers, which strongly innervate periventricular and dorsal parvicellular areas and are notably absent from magnocellular divisions, while noradrenergic fibers innervate parvi- and magnocellular neurosecretory divisions as well as those with both types of autonomic projections (40). It is tempting (and convenient) to parcellate importance of afferents according to functions of their target cells, but his may be quite simplistic. For example, in a recent study of immediate early gene activation it has been proposed that cardiovascular information relayed to parvicellular PVH autonomic regions may be used to modulate behavioral, rather than homeostatic responses to dehydration (46). From the forebrain, SFO, MePO, and BST projections to parvicellular divisions are much more robust than to magnos. There is also a diffuse projection from the substantia inominata (SI) to the periventricular subdivision, which may correlate with its nucleus acumbens-associated connection to the meso-limbic dopamine system (47). SFO and MePO projections are especially dense in the medial parvicellular (CRH-containing) part, and this projection, along with that of the autonomic parts, correlates well with studies of coordinate ACTH release and body water regulation (48). Information funneled from amygdala and subiculum by massive BST input can be related to known limbic/cognitive effects on the stress response, although BST input is mostly from the medial (amygdalar recipient) part and seems evenly distributed over all PVH subdivisions. Diencephalic inputs include those from the paraventricular thalamus and nucleus reuniens, both of which receive projections from hypothalamic structures detailed below, and are likely correlated with motor aspects of behavior. 331 As noted earlier, hypothalamic afferents to PVH are sparse and widespread, with a slight preponderance of fibers in medial subdivisions. Considered in total, however, this sparse innervation represents a rich input from virtually every hypothalamic area (49). All major hypothalamic subdivisions (excepting medial and lateral mammillary nuclei) send projections to parvicellular subdivisions of PVH. Specifically, projections from medial pre optic (POm) and lateral hypothalamic areas (LHA), and from anterior (AHN), dorsomedial (DMH) and ventromedial (VMH) hypothalamic nuclei have been shown. VMH is involved in feeding, rage, and female sexual behaviors and is consequently much studied. It has a heavy limbic interconnection via the BST and amygdala, and a sparse input from LHA. LHA cells are diffusely scattered rostro-caudally, and their afferents to PVH travel through the medial forebrain bundle where they are in a position to interact with many other hypothalamic fibers. LHA has been “implicated in the processing of sensory information, as well as the modulation of somatomotor responses, particularly in relation to the expression of behaviors associated with hunger and thirst, aggression and reproduction.”(1). Thus, information from all central sensory and motor systems has access to neuroendocrine and autonomic output cells of the PVH - a key element in integration of neuroendocrine function with behaviors that assure survival of the individual and the species. The robust innervation by DMH, relative to other hypothalamic structures is significant in regard to DMH involvement in arousal and ingestive behaviors. This may well be enhanced by the selective serotonergic and ACTH/beta endorphin innervation of periventricular and autonomic divisions of PVH. Scant innervation by the SCN circadian pacemaker, seems at first surprising in view of well known circadian rhythms in levels of various hormones, but the projection does originate preferentially in the ventral (retinal recipient) part of the SCN. SCN projects massively and reciprocally to the subparaventricular zone, which has recently been shown to have a complex inhibitory relationship with PVH neurons, and research on the SCN is beginning to reveal a complicated pattern of inhibitory connections that produce circadian rhythmicity. Another complexity of hypothalamic innervation is possible parvicellular to magnocellular communication within the PVH mentioned earlier, but those connections are not at all clear. Also, hypothalamic afferents to PVH are illustrative of an important fact not mentioned earlier: most of these connections are to some extent reciprocal. Thus, PVH afferent and efferent information has access to virtually constant update, which can only enhance (and is probably key to) endocrine and autonomic control information. Analysis of PVH inputs is surprisingly complex, and I have tried to minimize long lists of structures, and maximize the importance of types of information available. Certain basic 332 principles emerge. Inputs to PVH are organized for integration of internal sensory inputs with relayed external sensory and cognitive information in order to coordinate endocrine and autonomic function. Integration of behavior with PVH function is crucial for the animal to find food, water, and shelter from the elements so that homeostasis is maintained. It is further necessary to ensure appropriate social interactions and effective response to threatening situations, for individual and species survival in an unpredictable environment. Monitoring and integration of internal signals such as metabolic status, growth and circadian cycles, and response to physiologic challenge (e.g., illness, blood loss, parturition, and other stresses) is the primary importance of viscero-sensory and osmosensitive afferents. Also important is appropriate integration of somatic-sensory and cognitive information with endocrine function, so that appropriate life-sustaining behaviors are initiated. Multiple hypothalamic and brainstem reticular inputs (behavioral activating), indirect limbic system inputs (cognitive emotional), and visual associated (circadian rhythm) afferents provide a neural substrate for this coordination. We are now only beginning to appreciate how the numerous and complex inputs to this tiny hypothalamic nucleus can orchestrate such an impressive integrative function (50). Appendix II References 1. Swanson, L.W., The Hypothlamus, Chapter 1, in Integrated Systems of the CNS, Part I, T.H. A Bjorklund, L W Swanson, Editor. 1987, Elsevier Science Publishers: Amsterdam. p. 1-124. 2. Le Gros Clark, W.E., Morphological Aspects of the Hypothalamus, in [Four Lectures on] The Hypothalamus: morphological functional, clinical and surgical aspects, W. Le Gros Clark, et al., Editors. 1938, Oliver and Boyd: Edinburgh. p. 1-68. 3. Hatton, G., et al., Morphological evidence for two populations of magnocellular elements in the rat paraventricular nucleus. Brain Res, 1976. 108(1): p. 187-93. 4. Sawyer, C.H., History of the Neurovascular Concept of Hypothalamo-Hypophysisl Control. Biology of Reproduction, 1978. 18: p. 325-328. 5. Dyball, R.E.J. and K. Koizumi, Stimuli for the release of neurohypophyseal hormones. Brit. J. Pharmacol., 1969. 33: p. 319-328. 6. Rho, J. and L. Swanson, A morphometric analysis of functionally defined subpopulations of neurons in the paraventricular nucleus of the rat with observations on the effects of colchicine. J Neurosci, 1989. 9(4): p. 1375-88. 333 7. Hefco, V., et al., The ways through which the hypothalamic paraventricular nucleus (PVH) and the medial hypothalamus affect the organism's defence function. Rom J Physiol, 1993. 30(1-2): p. 87-91. 8. Fabris, N., E. Mocchegiani, and M. Provinciali, Pituitary-thyroid axis and immune system: a reciprocal neuroendocrine-immune interaction. Horm Res, 1995. 43(1-3): p. 29-38. 9. Ono, T., et al., Paraventricular Nucleus Connections to Spinal Cord and Pituitary. Neuroscience Letters, 1978. 10: p. 141-146. 10. Swanson, L. and P. Sawchenko, Paraventricular nucleus: a site for the integration of neuroendocrine and autonomic mechanisms. Neuroendocrinology, 1980. 31(6): p. 410-7. 11. Swanson, L. and H. Kuypers, The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivisions and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J Comp Neurol, 1980. 194(3): p. 555-70. 12. Swanson, L.W., The neural basis of motivated behavior. Acta Morphol. Neerl.- Scand., 1989. 26: p. 165-176. 13. Yoshihara, T., [A neuroendocrinological study on the mechanism of feeding- associated circadian rhythm of plasma corticosterone and feeding behaviour in rats: a role of neuropeptide Y in the paraventricular nucleus]. Hokkaido Igaku Zasshi, 1995. 70(1): p. 113- 31. 14. Shibasaki, T., et al., The response pattern of noradrenaline release to repeated stress in the hypothalamic paraventricular nucleus differs according to the form of stress in rats. Brain Res, 1995. 670(1): p. 169-72. 15. Kormilitsyna, N. and I. Bazhenov, [Role of the paraventricular nuclei of the hypothalamus in the development of cold adaptation in albino rats]. Fiziol Zh SSSR, 1988. 74(3): p. 414-9. 16. Van Erp, A., et al., Periaqueductal gray lesions do not affect grooming, induced electrically in the hypothalamic paraventricular area in the rat. Behav Brain Res, 1993. 59(1- 2): p. 95-101. 17. Consiglio, A. and A. Lucion, Lesion of hypothalamic paraventricular nucleus and maternal aggressive behavior in female rats. Physiol Behav, 1996. 59(4-5): p. 591-6. 18. Stellar, E., The Physiology of Motivation. Psychol. Rev., 1954. 61: p. 5-22. 19. Grossman, S. and J. Hennessy, Differential effects of cuts through the posterior hypothalamus on food intake and body weight in male and female rats. Physiol Behav, 1976. 17(1): p. 89-102. 20. Craige, E.H., Measurements of Vasculatity in some Hypothalamic nuclei of the Albino Rat, in The Hypothalamus and Central Levels of Futonomic Function, J.F. Fulton, S.W. Ranson, and A.M. Frantz, Editors. 1940, Wiliams & Wilkins: Baltimore. p. 310-319. 334 21. Simerly, R.B., et al., Distribution of androgen and estrogen receptor mRNA- containing cells in the rat brain: An in situ hybridization study. J. Comp. Neurol., 1990. 294: p. 76-95. 22. Swanson, L.W. and D.M. Simmons, Differential Steroid Hormone and Neural Influences on Peptide mRNA Levels in CRH Cells of the Paraventricular Nucleus: A Hybridization Histochemical Study in the Rat. Journal of Comparative Neurology, 1989. 285: p. 413-435. 23. Hartman, B., D. Zide, and S. Udenfriend, The Use of dopamine-ß hydroxylase as a marker for the central noradrenergic nervous system in the rat brain. Proc. nat. Acad. Sci. (Wash.), 1972. 69: p. 2722-2726. 24. Raichle, M.E., et al., Central Noradrenergic Regulation of Cerebral Blood Flow and Vascular Permeability. Proc. nat. Acad. Sci. (Wash.), 1975. 72: p. 3726-3730. 25. Swanson, L.W., M.A. Connelly, and B.K. Hartman, Ultrastructural evidence for central monoaminergic innervation of blood vessels in the paraventricular nucleus of the hypothalamus. Brain Res., 1977. 136: p. 166-173. 26. Gurdjian, E.S., The diencephalon of the albino rat. J. Comp. Neurol, 1927. 43(1): p. 1-114. 27. Armstrong, W., et al., Subnuclei in the rat hypothalamic paraventricular nucleus: a cytoarchitectural, horseradish peroxidase and immunocytochemical analysis. Neuroscience, 1980. 5(11): p. 1931-58. 28. Tweedle, C. and G. Hatton, Magnocellular neuropeptidergic terminals in neurohypophysis: rapid glial release of enclosed axons during parturition. Brain Res Bull, 1982. 8(2): p. 205-9. 29. Lincoln, D. and J. Wakerley, Factors governing the periodic activation of supraoptic and paraventricular neurosecretory cells during suckling in the rat. J Physiol (Lond), 1975. 250(2): p. 443-61. 30. Andrew, R., et al., Dye transfer through gap junctions between neuroendocrine cells of rat hypothalamus. Science, 1981. 211(4487): p. 1187-9. 31. Rho, J. and L. Swanson, Neuroendocrine CRF motoneurons: intrahypothalamic axon terminals shown with a new retrograde-Lucifer-immuno method. Brain Res, 1987. 436(1): p. 143-7. 32. Sawchenko, P.E. and L.W. Swanson, Immunohistochemical identification of neurons in the paraventricular nucleus of the hypothalamus that project to the medulla or to the spinal cord in the rat. J Comp Neurol, 1982. 205(3): p. 260-72. 335 33. Ceccatelli, S., M. Eriksson, and T. Hokfelt, Distribution and Coexistance of Corticotropin-Releasing Factor-, Neurotensin-, Enkephalin-, Cholecystokinin-, Galanin- and Vasoactive Intestinal Polypeptide/Peptide Histidine Isoleucine-Like Peptides in the Parvocellular Part of the Paraventricular Nucleus. Neuroendocrinology, 1989. 49: p. 309- 323. 34. Van den Pol, A.N., The magnocellular and parvocellular paraventricular nucleus nucleus of the rat: Intrinsic organization. J. comp. Neurol., 1982. 206: p. 317-345. 35. Roland, B. and P. Sawchenko, Local origins of some GABAergic projections to the paraventricular and supraoptic nuclei of the hypothalamus in the rat. J Comp Neurol, 1993. 332(1): p. 123-43. 36. Boudaba, C., K. Szabo, and J. Tasker, Physiological mapping of local inhibitory inputs to the hypothalamic paraventricular nucleus. J Neurosci, 1996. 16(22): p. 7151-60. 37. Scharrer, E. and B. Scharrer, Secretory Cells Within the Hypothalamus, in The Hypothalamus and Central Levels of Autonomic Function, J.F. Fulton, S.W. Ranson, and A.M. Frantz, Editors. 1940, Wiliams & Wilkins: Baltimore. 38. Epstein, A.N., J.T. Fitzsimmons, and B.J. Rolls, Drinking induced by injection of angiotensin into the brain of the rat. J. Physiol., Lond., 1970. 210: p. 457-474. 39. Sawchenko, P.E., et al., The distribution and cells of origin of serotonergic inputs to the paraventricular and supraoptic nuclei of the rat. Brain Res, 1983. 277(2): p. 355-60. 40. Sawchenko, P.E. and L.W. Swanson, The organization and biochemical specificity of afferent projections to the paraventricular and supraoptic nuclei, in Progress in Brain Research. 1983, Elsevier: Amsterdam. p. 19-29. 41. Lind, R.W., L.W. Swanson, and D. Ganten, Angiotensin II immunoreactive pathways in the central nervous system of the rat: Evidence for a projection from the subfornical organ to the paraventricular nucleus of the hypothalamus. Clin. Expt. Theory Practice, 1984. A6: p. 1915-1920. 42. Ferreyra, H., H. Kannan, and K. Koizumi, Influences of the limbic system on hypothalamo-neurohypophysial system. Brain Res, 1983. 264(1): p. 31-45. 43. Watts, A., L. Swanson, and G. Sanchez-Watts, Efferent projections of the suprachiasmatic nucleus: I. Studies using anterograde transport of Phaseolus vulgaris leucoagglutinin in the rat. J Comp Neurol, 1987. 258(2): p. 204-29. 44. Sawchenko, P., L. Swanson, and S. Joseph, The distribution and cells of origin of ACTH(1-39)-stained varicosities in the paraventricular and supraoptic nuclei. Brain Res, 1982. 232(2): p. 365-74. 45. Sawchenko, P. and L. Swanson, Central noradrenergic pathways for the integration of hypothalamic neuroendocrine and autonomic responses. Science, 1981. 214(4521): p. 685-7. 336 46. Blair, M., et al., Role of the hypothalamic paraventricular nucleus in cardiovascular regulation. Clin Exp Pharmacol Physiol, 1996. 23(2): p. 161-5. 47. Swerdlow, N.R., L.W. Swanson, and G.F. Koob, Substantia innominata: Critical link in the behavioral expression of mesolimbic dopamine stimulation in the rat. Neurosci. Lett., 1984. 50: p. 19-24. 48. Lind, R.W., L.W. Swanson, and P.E. Sawchenko, Anatomical Evidance that Neural Circuits Related to the Subfornical Organ Contain Angiotensin II. Brain Research Bulletin, 1985. 15: p. 79-82. 49. Sawchenko, P.E. and L.W. Swanson, The organization of forebrain afferents to the paraventricular and supraoptic nuclei of the rat. J Comp Neurol, 1983. 218(2): p. 121-44. 50. ZZ-Topp, I. R., Stress Research Ballad, J. Imponderable Results, 1997. 53(4):1-16 337 APPENDIX III Electrophysiological Studies of the PVH Electrophysiology of the hypothalamus, the PVH in particular, have provided insight (and sometimes predictive hypotheses) regarding neuroanatomical and clinical observations since early in this century. Electrical recording and neuroanatomical techniques advanced in tandem, so that now a great deal can be said with certainty about the function of PVH neurons through analysis of their neurochemical content and electrical activity in the framework of known synaptic connectivity. The earliest electrophysiological studies, inspired by work of Sherrington and others, were mainly observations of physiological responses to electrical stimulation in an effort to understand hypothalamic control of the autonomic nervous system (3, 4). Introduction of single-unit extracellular and sharp-electrode intracellular recording techniques in the 1960s brought the possibility of another level of insight, and it is interesting to note that Eric Kandel among others published work on electrical properties of hypothalamic neuroendocrine cells (41)(18). In 1967 Cross and Kitay tried a new preparatory approach (15), and succeeded in showing effects that were predictive of some of the most interesting current work using patch-clamp techniques in PVH slices (6). They isolated central parts of the thalamus and hypothalamus with a cylindrical cutting device in decerebrate cats, thereby producing what was termed a ‘diencephalic island.’ They found an abundance of neurons with faster and more uniform firing rates than usual in this preparation, in which most of the extrinsic neuronal input was eliminated, thereby implying a release from inhibitory connections. They further showed these firing rates to be unaffected by peripheral nervous stimulation but slowed by intravenous barbiturates, indicating a patent blood supply in the isolated tissue. Cross and Kitay’s work referred to earlier single unit studies showing responsiveness to humoral stimuli (hypertonic saline, hyper- or hypoglycemia, and gonadal hormones), and suggested that these effects could be from either humoral or synaptic inputs to the hypothalamus. They also commented on the lack of knowledge whether their observed ‘spontaneous activity’ was intrinsic. Perhaps it was stimulated by extrinsic sources such as those from the reticular activating system, since hypothalamic single-unit firing had been correlated with synchronized EEG activity. Their paper nicely illustrates the ‘stage setting’ for the subsequent three decades of electrophysiological study of PVH neurons. A modern reading reveals foreshadowing of numerous avenues of productive neuroscience research, e.g., limbic and humoral modulation of autonomic and pituitary function, circadian 338 rhythmicity, and the current systems analysis approach to understanding neural function. There has been a great deal of work on neural versus chemical influences, and on local inhibitory inputs regulating the electrophysiological activity of the various cell types in the PVH. Much of this data appears to support the working hypothesis that three major cell types, based on cytoarchitectonic morphology and functional efferent connections, exist in the PVH. Interesting data on reproductive function in relation to altered afferent connections and firing rate changes (knife cuts in rostral but not caudal hypothalamus decrease PVH firing rates (53)) are noted, but will not be further discussed since the primary effector neurons are not localized to PVH. Magnocellular neurosecretory neurons containing oxytocin (Oxy) and the anti- diuretic hormone vasopressin (Vas) were the first recognized, and remain the best-studied cells in the PVH (68). Isolation and characterization of these two peptides in the 1950s, and subsequent studies in the 1960s using antibodies to demonstrate their presence in supraoptic (SO) and paraventricular neurons that send their axons to the posterior pituitary (17) heralded a new era of chemical neuroanatomy in endocrinology research. In the early 1970s, electrophysiologists antidromically stimulated neural lobe (posterior pituitary) to identify SO and PVH magnocellular neurosecretory neurons using collision criteria for spontaneous spike activity, and observed two populations of cells with slightly different phasic activity patterns that correlated with either oxytocin or vasopressin release (59). Changes in firing of these neurons were monitored after the microiontophoretic application of a number of neurally active chemicals known or suspected to influence their function (57)(86)(27). In order to determine possible sources of extrinsic synaptic input which were not clearly apparent at that time, single cell responses to glutamine (Glu), acetylcholine (ACh), noradrenaline (NA), and Oxy were tested by Cross and others using electrophysiological changes in identified magnocellular neurons as an assay of effectiveness. They showed that Glu excited most of the cells tested (and never produced an inhibition), ACh also had a predominantly excitatory effect (only three PVH cells showed inhibition), and in contrast NA inhibited more cells in both nuclei than it excited, while Oxy appeared to modestly excite some cells and not others (16)(14). These and similar studies (2) provided valuable information for later neuroanatomical correlation of possible function of cholinergic and peptidergic fibers in PVH (68). Hoblitzell, et al (34) cite a 1972 observation (later published in French (25)) characterizing three cell types in the PVH: “two were antidromically activated from the pituitary...of those type I was ventral to type II and exhibited a greater spontaneous activity during lactation and estrous.” The third type was possibly 339 parvicellular, but it is not clear whether it was antidromically tested. These observations were predictive of the major work to follow on electrophysiology of PVH neurons (62)(76, 77). Beginning in the mid-seventies, Wakerly, Lincoln and their associates conducted a classic series of experiments that characterized a defined physiological role for Oxy neurons (87). This was the first use of single-unit recording -- to brilliant advantage -- for study of integrated function in the PVH. Lincoln, et al used single-unit recording in antidromically- identified supraoptic and PVH magnocellular neurons to follow the development of the milk ejection reflex in response to suckling in lactating rats (47, 48). They showed that Oxy cells could be discriminated from Vas cells on the basis of differences in their firing patterns, and that effective release of Oxy at neuron terminals could be evaluated by recordings of intramammary pressure at the time of milk ejection preceded by a 20-40 fold increase in firing rate of Oxy cells. There was a clear relationship of background firing rate and intensity of stimulus (number of pups) to the subsequent firing patterns that elicited the reflex. In addition, they noticed that stimulus intensity and amount of Oxy release could be correlated with the activity of each single-unit recorded. This led to the later discovery of synchronized firing by all Oxy neurons during the milk ejection reflex. The mechanism for this synchronization was inferred by subsequent demonstration of dye transfer between Oxy cells, revealing gap junctions and implying an electrical coupling between them (1). The lactation experimental model (even though it doesn’t shed light on possible function of Oxy cells in males) has been well used for elucidation of central neural processes, and has prompted numerous technical and conceptual advances (74)(44)(37)(13)(38). One clever technique was used by Summerlee et al to corroborate findings from anesthetized animals in a more normal paradigm (72), since some critics had implied that Lincoln’s findings might be an artifact of anesthesia. It entailed permanent implantation of a PVH recording electrode cemented to the skull in a young pregnant rat, such that with normal growth the electrode slowly and with minimal trauma advanced down through the PVH during parturition and subsequent lactation. With this method recordings were made, sometimes over several hours, from single Oxy cells identified on the basis of their phasic firing patterns and correlated with the milk ejection reflex. During a period of several weeks almost 250 Oxy cells were recorded in the awake, freely moving animal, and they were found without fail to reflect the milk ejection reflex. Apparently, earlier negative findings in awake animals had been due to a stress-related inhibition of normal function in restrained rats, which agrees with earlier findings by Lincoln that in rat Oxy cells seem to be under a mild tonic inhibition. 340 Two technical innovations in the early 1980s allowed development of in vitro recording methods that, combined with microiontophoretic application of chemicals and microstimulation of synaptic inputs, have produced a wealth of detailed electrophysiological information about the PVH. One was the development of a technique by Bourque and Renaud for excising a hypothalamic block with attached infundibular stalk on the surface, and carotid arteries intact to assist perfusion of nutrient and electrolyte solutions for in vitro studies (7). Use of this preparation allowed stable recording (with no heartbeat artifact) for ten or more hours from antidromically identified neurons whose osmotic milieu and other experimental conditions could be altered at will. With this approach, results were obtained that led researchers to propose the presence of recurrent synaptic inhibition in magnocellular axon terminals (8). The validity of this hypothesis was strengthened by electron- microscopical studies of changing glial patterns around magnocellular synapses during parturition (82) and dehydration (83), and they have recently been re-demonstrated with in vivo recordings from PVH in the male rat (85). Explant studies were limited to SO due to the size of the tissue block (8x8x2mm), but along with other SO studies of Oxy and Vas cells (9), these results were found to generalize quite reliably to the PVH (40). The other important technical innovation for single-unit recording was published in 1980 by Hatton and colleagues detailing a method to sustain for several hours viable one- to four-hundred micron thick slices through the entire PVH and surround (31). They had successfully used this technique to visualize and electrically identify PVH neurons recorded intracellularly with glass micropipets loaded with salt or protein marking solutions (30). One advantage of this approach was the option to choose a slice plane preserving desired axon connectivity pathways so that discrete input-stimulus combinations could be tested (12)(79)(91). It was also possible to immunohistochemically label identified cells after electrical recordings (36, 91). This preparation, in addition to the use of dissociated cells in culture (89), has been exploited for whole cell patch-clamp analysis of PVH and other hypothalamic neurons (6)(35). For example, three types of outward potassium currents (delayed outward, Ca++ dependant, and transient) in response to peptide stimulation have been identified recently in magnocellular neurons (46). One major value of using single-unit recording in neuroscience research is the ability to correlate electrophysiological firing patterns with neural input -- by electrical stimulation of known pathways, chemical application of putative neurotransmitters, or alteration in physiological levels of metabolic signals such as electrolytes, glucose, or corticosteroids in the blood. These approaches have all been used, singly or in combination, to investigate a 341 number of known PVH functions. Following the demonstration of anterior pituitary hormone control by releasing factors from the parvicellular part of the PVH to the portal vasculature of the median eminence, antidromic stimulation of the hypophysial stalk was used to identify PVH parvicellular neuroendocrine cells and define their intermittent or slow- phasic firing nature (26). Another parvicellular cell type that sends exclusive descending projections to the brainstem or spinal cord was seen anatomically (75) and later characterized by antidromic stimulation of brainstem catecholamine-containing structures (42, 43, 63, 67). Comparative antidromic responses showed three separate populations of PVH cells that project to neurohypophysis or brainstem, and that at least some of the brainstem projecting cells respond to baroreceptor stimulation (42). One study detailing three response patterns and differing excitatory latencies of phasically firing PVH cells to A1 noradrenergic cell group stimulation tested differential reaction to various antagonists. They inferred from physiological observations that the projections were unmyelinated, and that there exist dual, alpha-adrenergic and non-adrenergic excitatory pathways, and a third beta- adrenergic inhibitory ascending pathway to PVH (78). In other work on the PVH, Sundsten had observed effects of septal stimulation on identified single-units as early as 1971 in primate (73), and in 1980 mild inhibitory limbic inputs to both Vas and Oxy neurons were confirmed in rat (58). Later, Ferreyra and co- workers identified the effects of multiple neural inputs to a single cell type, or even a single neuron in the PVH (24). These and other studies helped to clarify thinking about the significance of multiple inputs to a cell type or combination of PVH cell types that could subserve slightly different, but related functions. This concept has guided a number of research programs seeking to understand different physiological functions reflected in electrophysiology of PVH cells. Control of autonomic and endocrine function, and integration with appropriate behaviors to maintain homeostasis and insure survival are clearly crucial and complex tasks that seem to be mediated by PVH neurons. Electrophysiological studies have been conducted apace with anatomical (including continuing neurochemical and updated electron microscopic analysis (50)), behavioral, and neurochemical work. Single unit recording has been used to advantage in studying lactation and parturition, as detailed above. It has also been used to study the well-known effects of mealtime conditioning and chemical stimulation in PVH (51)(52) for onset and food preference in feeding behavior (69)(93)(71). However, it now appears from studies of Leibowitz and others that those effects are mediated through peptidergic PVH connections to the dorsal medial hypothalamic 342 nucleus, which receives ascending visceral information concerning nutritional status (92)(45). There is a large body of electrophysiological and anatomical work studying the relationship of various homeostatic and cognitive inputs for activation of the hypothalamic- pituitary-adrenal (HPA) axis in the stress response (20, 21). Recent recordings from identified PVH neurons have shown a change in PVH h-p-a spike activity from such diverse stimuli as gastric distension (84), insulin application in situ (55), alpha interferon infusion (64, 66), and systemic interleukine-1 injection (65). The latter work suggests a mechanism for immune system communication with the central nervous system -- the long sought link for well known but poorly understood immune responses to stress. Another research area related to the stress response and PVH control of the autonomic and endocrine systems is study of input from limbic structures involved in fear and defensive behavior (56)(70)(11)(61). It has also been shown that cardiovascular responses are modified by the h-p-a axis during adjustments in fluid homeostasis (90). A recent anatomical finding implicating Oxy neurons in cardiac function may also be involved in the h-p-a axis: Transneuronal tracing of the pathway to the stellate ganglion, the principle source of sympathetic supply to the heart, shows preferential input from Oxy cells in PVH (39). This finding may give some hint of Oxy function in males, and certainly bears electrophysiological investigation. Another major research area to which single-unit recording in PVH has been applied is that of circadian control of rhythms of daily activity and cycling levels of key hormones, a major recent thrust of Leo Renaud’s research group and others (33)(81)(32)(54). Space constraint prohibits more than a brief mention of these interesting studies. The largest body of literature concerning electrophysiology in the PVH is also the best example of deprivation or satiety effects on PVH single-unit activity. This is the long pursued research investigation of hypothalamic (thus PVH) control of body-water homeostasis, which electrophysiologists have entered with enthusiasm. Early anatomical (28) and electrophysiological studies had shown that Vas cells respond to water deprivation and blood osmolality changes from injections of hypertonic saline (29), blood loss (88), and direct baroreceptor stimulation (90). Peripheral or central injection of the dehydration- response peptide angiotensin-II (A-II) was known to elicit drinking behavior, and the presence of A-II had also been demonstrated histochemically in osmosensitive neurons projecting from the subfornical organ (SFO) to the PVH (49). It was not clear if PVH autonomic responses to cardiac baroreceptor input (as in hemorrhage) were related to the putative osmoreceptor information relayed from the SFO. Transection of the SFO-to-PVH 343 fiber pathway was found to eliminate drinking in response to intravenously injected A-II, but not to dehydration induced by water deprivation (19). What then, was the mechanism of SFO influence on PVH control of fluid homeostasis? An electro-physiological approach was taken recently by Ferguson and coworkers to address this perplexing question. They used both in vivo and in vitro techniques with standard intracellular and whole cell patch- clamp recordings to identify neurosecretory and descending autonomic PVH cells whose firing was excited by SFO stimulation or modulated by local A-II application. They were able to demonstrate that the non-humoral effects of SFO innervation were mediated by projections to autonomic rather than neuroendocrine cells in the PVH (5). In additional studies using whole cell patch-clamp and calcium imaging, they further demonstrated three potassium channel subtypes in PVH that respond to local A-II microdialysis, and visualized accumulation of calcium in SFO neurons stimulated by circulating A-II. This is strong evidence that A-II is at least one of the neurotransmitters responsible for stimulation of descending PVH output to the sympathetic nervous system (22, 23). These findings appear to have answered two interesting questions posed by the results of earlier anatomical studies: How is the observed SFO projection to different PVH cell types reflected in specific function? And do the osmoreceptive SFO neurons that respond to circulating A-II and also contain A-II immunoreactivity in their perikarya and axonal projections use that peptide as a transmitter? An interesting, new area of research using whole cell patch-clamp recording in identified PVH neurons is being pursued by Tasker and co-workers (6, 79). They have used the PVH slice preparation in several different planes to record from and biocytin-label identified neurons while applying micro amounts of excitatory or inhibitory agents nearby. Typical transient K+ or Ca++ currents identify putative cell types (60)(10)(80), and post hoc immunohistochemistry in PVH and in situ hybridization for GAD mRNA in small neurons in the PVH surround, helped to document a donut-shaped ring of inhibitory inputs surrounding the ventral aspect of the nucleus. PVH does not contain identifiable interneurons, but local inhibitory cells had long been suspected from anatomical tracing and electrophysiological experiments. These results clarify much of the work concerning suprachiasmatic (circadian clock) and dorsomedial hypothalamus (feeding) inputs to PVH and its immediate surround. Such powerful combination of modern anatomical and electrophysiological methods will surely continue to provide great insight on the function of PVH neurons. In conclusion, electrophysiological studies have produced data in agreement with my hypothesis of three major cell types in PVH. It is possible that finer subdivisions of these major architectonically 344 visible cell types will be elucidated with new anatomic, neurochemical, or electrophysiological methods. However, I believe these will serve to aid interpretation of complex interactions, rather than define an alternative cellular parcellation of PVH (94). Appendix III References 1. Andrew, R., MacVicar, B., Dudek, F. and Hatton, G., Dye transfer through gap junctions between neuroendocrine cells of rat hypothalamus. Science, 1981. 211(4487): p. 1187-9. 2. Arnauld, E., Cirino, M., Layton, B. and Renaud, L., Contrasting actions of amino acids, acetylcholine, noradrenaline and leucine enkephalin on the excitability of supraoptic vasopressin-secreting neurons. A microiontophoretic study in the rat. Neuroendocrinology, 1983. 36(3): p. 187-96. 3. Association for Research In Nervous and Mental Disease, The Hypothalamus and Central Levels of Autonomic Function. 1966 reprint facsimile ed. Prodeedings of the Association for Research in Nervous and Mental Disease, ed. Fulton, J.F. Vol. XX. 1940, reprint 1966, New York: Hafner. 980. 4. Association for Research In Nervous and Mental Disease, The Hypothalamus. Research Publications -- Association for Research in Nervous and Mental Disease, ed. Reichlin, S., Baldessarini, R.J. and Martin, J.B. Vol. 56. 1978, New York: Raven Press. 5. Bains, J. and Ferguson, A., Paraventricular nucleus neurons projecting to the spinal cord receive excitatory input from the subfornical organ. Am J Physiol, 1995. 268(3 Pt 2): p. R625-33. 6. Boudaba, C., Szabo, K. and Tasker, J., Physiological mapping of local inhibitory inputs to the hypothalamic paraventricular nucleus. J Neurosci, 1996. 16(22): p. 7151-60. 7. Bourque, C. and Renaud, L., In vitro neurophysiology of identified rat hypothalamic 'neuroendocrine' neurons. Neuroendocrinology, 1983. 36(2): p. 161-4. 8. Bourque, C. and Renaud, L., A perfused in vitro preparation of hypothalamus for electrophysiological studies on neurosecretory neurons. J Neurosci Methods, 1983. 7(3): p. 203-14. 9. Bourque, C. and Renaud, L., Activity patterns and osmosensitivity of rat supraoptic neurones in perfused hypothalamic explants. J Physiol (Lond), 1984. 349: p. 631-42. 10. Bourque, C.W., Transient Calcium-dependant Potassium Current in the Magnocellular Neurosecretory Cells of the Rat Supraoptic Nucleus. J. Physiol. Lond., 1988. 397:331-347. 11. Brudzynski, S.M., Eckersdorf, B. and Golebiewski, H., Regional specificity of the emotional-aversive response induced by carbachol in the cat brain: a quantitative mapping study. J Psychiatry Neurosci, 1995. 20(2): p. 119-32. 345 12. Carette, B. and Poulain, P., Vasopressin-sensitive neurons in the lateral paraventricular nucleus area in a guinea pig slice preparation. Brain Res Bull, 1989. 22(6): p. 969-74. 13. Ceccatelli, S. and Eriksson, M., The effect of lactation on nitric oxide synthase gene expression. Brain Res, 1993. 625(1): p. 177-9. 14. Cross, B., Dyball, R. and Moss, R., Stimulation of paraventricular neurosecretory cells by oxytocin applied iontophoretically. J. Physiol.(Lond), 1972. 222(1): p. 22P-23P. 15. Cross, B. and Kitay, J., Unit activity in diencephalic islands. Exp Neurol, 1967. 19(3): p. 316-30. 16. Cross, B.A., Moss, R.L. and Urban, I., Effects of Ionotophoretic Application of Acetylcholine and Noradrenaline to Antidromically identified Paraventricular Neurones. J. Physiol. (Lond.), 1971. 214: p. 28-30P. 17. Dierickx, K., Immunocytochemical localization of the vertebrate nonapeptide neurohypophyseal hormones and neurophysins. Int. Rev. Cytol., 1980. 62: p. 119-185. 18. Dyball, R.E.J. and Koizumi, K., Stimuli for the release of neurohypophyseal hormones. Brit. J. Pharmacol., 1969. 33: p. 319-328. 19. Eng, R. and Miselis, R.R., Polydipsia and abolition of angiotensin-induced drinking after transections of subfornical organ efferent projections in rat. Brain Res, 1981. 225(1):200-6 20. Feldman, S., Neural pathways mediating adrenocortical responses. Fed Proc, 1985. 44 (1 Pt 2): p. 169-75. 21. Feldman, S., Conforti, N., Itzik, A. and Weidenfeld, J., The role of limbic structures in the modulation of ACTH responses following adrenalectomy. Ann N Y Acad Sci, 1995. 771: p. 73-81. 22. Ferguson, A.V. and Bains, J.S., Electrophysiology Of the Circumventricular Organs. Frontiers In Neuroendocrinology, 1996. 17(4): p. 440-475. 23. Ferguson, A.V. and Bains, J.S., Actions Of Angiotensin In the Subfornical Organ and Area Postrema - Implications For Long-Term Control Of Autonomic Output. Clinical and experimental pharmacology and physiology, 1997. 24(1): p. 96-101. 24. Ferreyra, H., Kannan, H. and Koizumi, K., Influences of the limbic system on hypothalamo-neurohypophysial system. Brain Res, 1983. 264(1): p. 31-45. 25. Freund-Mercier, M., Richard, P. and Miro, J., Donnees fonctionnelles en faveur de l'existence de trois types de cellules dans le noyau paraventriculaire chez la ratte. Brain Research, 1975. 89: p. 149-154. 26. Gartside, S.E., Suaud, C.M.F. and Tappaz, M., Evidence that activation of the hypothalamo-pituitary-adrenal axis by electrical stimulation of the noradrenergic A1 group is not mediated by noradrenaline. Neuroendocrinology, 1995. 62(1): p. 2-12. 346 27. Gosbee, J. and Lederis, K., In vivo release of antidiuretic hormone by direct application of acetylcholine or carbachol to the rat neurohypophysis. Can J Physiol Pharmacol, 1972. 50(6): p. 618-20. 28. Gregory, W., Tweedle, C. and Hatton, G., Ultrastructure of neurons in the paraventricular nucleus of normal, dehydrated and rehydrated rats. Brain Res Bull, 1980. 5(3): p. 301-6. 29. Harris, M.C., Dreyfuss, J.J. and Legros, J.J., Excitation of Phasically firing Supraoptic Neurones During Vasopressin Release. Nature, 1975. 258: p. 80-82. 30. Hatton, G., Armstrong, W. and Gregory, W., Spontaneous and osmotically-stimulated activity in slices of rat hypothalamus. Brain Res Bull, 1978. 3(5): p. 497-508. 31. Hatton, G., Doran, A., Salm, A. and Tweedle, C., Brain slice preparation: hypothalamus. Brain Res Bull, 1980. 5(4): p. 405-14. 32. Hermes, M.L., Coderre, E.M., Buijs, R.M. and Renaud, L.P., GABA and glutamate mediate rapid neurotransmission from suprachiasmatic nucleus to hypothalamic paraventricular nucleus in rat. J Physiol, 1996. 496(Pt 3): p. 749-57. 33. Hermes, M.L. and Renaud, L.P., Differential responses of identified rat hypothalamic paraventricular neurons to suprachiasmatic nucleus stimulation. Neuroscience, 1993. 56(4): p. 823-32. 34. Hoblitzell, E., Hatton, G. and Armstrong, W., Paraventricular nucleus: changes in the medial and lateral cell groups during dehydration and rehydration in the rat. Brain Res Bull, 1976. 1(3): p. 329-32. 35. Hoffman, N.W., Kim, Y.I., Gorski, R.A. and Dudek, F.E., Homogeneity of intracellular electrophysiological properties in different neuronal subtypes in medial preoptic slices containing the sexually dimorphic nucleus of the rat. J Comp Neurol, 1994. 345(3): p. 396- 408. 36. Hoffman, N.W., Tasker, J.G. and Dudek, F.E., Immunohistochemical differentiation of 00paraventricular nucleus. J Comp Neurol, 1991. 307(3): p. 405-16. 37. Ingram, C. and Moos, F., Oxytocin-containing pathway to the bed nuclei of the stria terminalis of the lactating rat brain: immunocytochemical and in vitro electrophysiological evidence. Neuroscience, 1992. 47(2): p. 439-52. 38. Ingram, C.D., Kavadas, V., Thomas, M.R. and Threapleton, J.D., Endogenous opioid control of somatodendritic oxytocin release from the hypothalamic supraoptic and paraventricular nuclei in vitro. Neurosci Res, 1996. 25(1): p. 17-24. 39. Jansen, A., Wessendorf, M. and Loewy, A., Transneuronal labeling of CNS neuropeptide and monoamine neurons after pseudorabies virus injections into the stellate ganglion. Brain Res, 1995. 683(1): p. 1-24. 347 40. Ji, Y.P., Mei, J. and Tang, B., [Effects of intracerebroventricular administration of norepinephrine on paraventricular neurosecretory cells]. Sheng Li Hsueh Pao, 1995. 47(2): p. 120-6. 41. Kandel, E.R., Electrical Properties of hypothalamic neuroendocrine cells. J. gen. Physiol., 1964. 47: p. 691-717. 42. Kannan, H. and Yamashita, H., Electrophysiological study of Paraventricular Nucleus Neurons Projecting to the Dorsomedial Medulla and Their Response to Baroreceptor stimulation in Rats. Brain Research, 1983. 279: p. 31-40. 43. Kawano, S., Osaka, T., Kannan, H. and Yamashita, H., Excitation of hypothalamic paraventricular neurons by stimulation of the raphe nuclei. Brain Res Bull, 1992. 28(4): p. 573-9. 44. Kiss, J., Kanyicska, B. and Nagy, G., The hypothalamic paraventricular nucleus has a pivotal role in regulation of prolactin release in lactating rats. Endocrinology, 1986 119(2): p.8703. 45. Leibowitz, S.F., Brain peptides and obesity: pharmacologic treatment. Obesity Research, 1995. 3(4): p. 573S-589S. 46. Li, Z. and Ferguson, A.V., Electrophysiological properties of paraventricular magnocellular neurons in rat brain slices: modulation of IA by angiotensin II. Neuroscience, 1996. 71(1): p. 133-45. 47. Lincoln, D. and Wakerley, J., Factors governing the periodic activation of supraoptic and para-ventricular neurosecretory cells during suckling in the rat. J Physiol (Lond), 1975. 250(2): p. 443-61. 48. Lincoln, D.W. and Wakerly, J.B., Electrophysiological evidence for the activation of supraoptic nurones during the release of oxytocin. J. Physiol. Lond., 1974. 242: p. 533-554. 49. Lind, R., Swanson, L., Bruhn, T. and Ganten, D., The distribution of angiotensin II- immunoreactive cells and fibers in the paraventriculo-hypophysial system of the rat. Brain Res, 1985. 338(1): p. 81-9. 50. Liposits, Z., Ultrastructure of hypothalamic paraventricular neurons. Crit Rev Neurobiol, 1993. 7(2): p. 89-162. 51. Mitome, M., Honma, S., Yoshihara, T. and Honma, K., Prefeeding increase in paraventricular NE release is regulated by a feeding-associated rhythm in rats. Am J Physiol, 1994. 266(4 Pt 1): p. E606-11. 52. Morien, A., Wellman, P. and Fojt, J., Diurnal rhythms of paraventricular hypothalamic norepinephrine and food intake in rats. Pharmacol Biochem Behav, 1995. 52(1): 169-74. 53. Negoro, H. and Akaishi, T., Effect of anterior of posterior deafferentation of the hypothalamus on unit activity in the paraventricular nucleus of ovariectomized female rats with or without estrogen treatment. Endocrinol Jpn, 1981. 28(1): p. 37-43. 348 54. Oddie, S.D., Stefanek, W., Kirk, I.J. and Bland, B.H., Intraseptal procaine abolishes hypothalamic stimulation-induced wheel-running and hippocampal theta field activity in rats. J Neurosci, 1996. 16(5): p. 1948-56. 55. Orosco, M., Gerozissis, K., Rouch, C. and Nicolaidis, S., Feeding-related immunoreactive insulin changes in the PVN-VMH. Brain Res, 1995. 671(1): p. 149-58. 56. Petrov, T., Jhamandas, J.H. and Krukoff, T.L., Electrical Stimulation of the Central Nucleus of the Amygdala Induces fos-like Immunoreactivity in the Hypothalamus of the Rat: a Quantitative study. Brain Research, Molecular Brain Research, 1994. 22(1-4): p. 333-340. 57. Pickford, M., The Action of Acetylcholine on the Supraoptic Nucleus of the Chlorosed Dog. J. Physiol. (Lond.), 1947. 106: p. 264-270. 58. Poulain, D., Ellendorff, F. and Vincent, J., Septal connections with identified oxytocin and vasopressin neurones in the supraoptic nucleus of the rat. An electrophysiological investigation. Neuroscience, 1980. 5(2): p. 379-87. 59. Poulain, D.A., Wakerly, J.B. and Dyball, R.E.J., Electrophysiological differentiation of oxytocin- and vasopressin-secreting hormones. Proc. R. Soc. B, 1974. 196: p. 367-384. 60. Poulian, D.A. and Wakerly, J.B., Electrophysiology of Hypothalamic Neurones Secreting Oxytocin and Vasopressin. Neuroscience., 1982. 7: p. 773-808. 61. Prutskova, N.P., [An electrophysiological analysis of the morphofunctional organization of the limbic control over the magnocellular neurosecretory hypothalamic nuclei in rats]. Fiziol Zh Im I M Sechenova, 1996. 82(7): p. 18-27. 62. Renaud, L.P., Blume, H.W., Kearney, R.E., MacKenzie, B.W. and Pittman, Q.J., Physiological Anatomy of the Hypothalamus, in Current Studies of Hypothalamic Function: Metabolism and Behavior. 1978, Karger: Basel. p. 1-16. 63. Saphier, D., Electrophysiology and neuropharmacology of noradrenergic projections to rat PVN magnocellular neurons. Am J Physiol, 1993. 264(5 Pt 2): p. R891-902. 64. Saphier, D., Neuroendocrine effects of interferon-alpha in the rat. Adv Exp Med Biol, 1995. 373: p. 209-18. 65. Saphier, D. and Ovadia, H., Selective facilitation of putative corticotropin-releasing factor- secreting neurones by interleukin-1. Neurosci Lett, 1990. 114(3): p. 283-8. 66. Saphier, D., Roerig, S.C., Ito, C., Vlasak, W.R., Farrar, G.E., Broyles, J.E. and Welch, J.E., Inhibition of neural and neuroendocrine activity by alpha-interferon: neuroendocrine, electrophysiological, and biochemical studies in the rat. Brain Behav Immun. [published erratum appears in Brain Behav Immun 1994 Dec;8(4):374], 1994. 8(1): p. 37-56. 67. Saphier, D. and Zhang, K., Inhibition by the serotonin1A agonist, 8-hydroxy-2- (di-n- propylamino)tetralin, of antidromically identified paraventricular nucleus neurons in the rat. Brain Res, 1993. 615(1): p. 7-12. 349 68. Sawchenko, P.E. and Swanson, L.W., The organization and biochemical specificity of afferent projections to the paraventricular and supraoptic nuclei, in Progress in Brain Research. 1983, Elsevier: Amsterdam. p. 19-29. 69. Shibasaki, T., Oda, T., Imaki, T., Ling, N. and Demura, H., Injection of anti-neuropeptide Y gamma-globulin into the hypothalamic paraventricular nucleus decreases food intake in rats. Brain Res, 1993. 601(1-2): p. 313-6. 70. Silveira, M.C., Sandner, G., DiScala, G. and Graeff, F.G., c-fos Immunoreactivity in the Brain Following Electrical or Chemical Stimulation of the Medial Hypothalamus of Freely Moving Rats. Brain Research, 1995. 674(2): p. 265-74. 71. Smith, B., York, D. and Bray, G., Effects of dietary preference and galanin administration in the paraventricular or amygdaloid nucleus on diet self-selection. Brain Res Bull, 1996. 39(3): p. 149-54. 72. Summerlee, A.J.S., Lincoln, D.W. and Webb, A.C., Long-term electrical recordings from single neurosecretory cells in the conscious lactating rat. J. Endocrinol., 1983. 83:41P. 73. Sundsten, J.W., Septal inhibition of antidromically activated hypothalamic paraventricular neurons in the monkey. Anat. Rec, 1971(169): p. 143. 74. Sutherland, R.C. and Fink, G., The Milk Ejection Pathway in Brain Studied with the 2- Deoxyglucose Method. Brain Research, 1983. 273: p. 291-296. 75. Swanson, L., Sawchenko, P., Wiegand, S. and Price, J., Separate neurons in the paraventricular nucleus project to the median eminence and to the medulla or spinal cord. Brain Res, 1980. 198(1): p. 190-5. 76. Swanson, L.W., Neuropeptides—new vistas on synaptic transmission. Trends in Neurosci., 1983. 6: p. 294-295. 77. Swanson, L.W., The neural basis of motivated behavior. Acta Morphol. Neerl.-Scand., 1989. 26: p. 165-176. 78. Tanaka, J., Kaba, H., Saito, H. and Seto, K., Response of phasically firing paraventricular neurons to A1 noradrenergic region stimulation and its attenuation by adrenoceptor antagonists. Exp Neurol, 1985. 88(3): p. 714-25. 79. Tasker, J. and Dudek, F., Local inhibitory synaptic inputs to neurones of the paraventricular nucleus in slices of rat hypothalamus. J Physiol (Lond), 1993. 469: p. 179-92. 80. Tasker, J.G. and Dudek, F.E., Electrophysiological Properties of Neurones in the Region of the Paraventricular Nucleus in Slices of Rat Hypothalamus. J. Physiol. Lond., 1991. 434: p. 271-293. 81. Thellier, D., Moos, F., Richard, P. and Stoeckel, M., Evidence for connections between a discrete hypothalamic dorsochiasmatic area and the supraoptic and paraventricular nuclei. Brain Res Bull, 1994. 34(3): p. 261-74. 350 82. Tweedle, C. and Hatton, G., Magnocellular neuropeptidergic terminals in neurohypophysis: rapid glial release of enclosed axons during parturition. Brain Res Bull, 1982. 8(2):205-9. 83. Tweedle, C. and Hatton, G., Synapse formation and disappearance in adult rat supraoptic nucleus during different hydration states. Brain Res, 1984. 309(2): p. 373-6. 84. Ueta, Y., Kannan, H. and Yamashita, H., Gastric afferents to the Paraventricular Nucleus in Rat. Exp. Brain Res., 1991. 84(3): p. 487-94. 85. Ueta, Y., Kannan, H. and Yamashita, H., Does recurrent inhibition exist in the paraventricular magnocellular neurosecretory cells in the male rat? Jpn J Physiol, 1995. 45(5): p. 913-9. 86. Verney, E.B., The Antidiuretic Hormone and the Factors Which Affect its Release. Proc. Royal Soc. B, 1947. 135: p. 25-106. 87. Wakerly, J.B. and Lincoln, D.W., The milk ejection reflex of the rat: a 20- to 40-fold acceleration in the firing rate of paraventricular neurones during oxytocin release. J. Endocrinol., 1973. 57: p. 477-493. 88. Wakerly, J.B., Poulain, D.A., Dyball, R.E.J. and Cross, B.A., Activity of Phasic Neurosecretory Cells During Hemorrhage. Nature, Lond., 1975. 258: p. 82-84. 89. Weiss, M.L. and Cobbett, P., Intravenous injection of Evans Blue labels magnocellular neuroendocrine cells of the rat supraoptic nucleus in situ and after dissociation. Neuroscience, 1992. 48(2): p. 383-95. 90. Yamashita, H., Kannan, H., Inenaga, K. and Koizumi, K., The role of cardiovascular and muscle afferent systems in control of body water balance. J Auton Nerv Syst, 1984. 10(3-4): p. 305-16. 91. Yang, L.M., Wang, Z.A. and Ho, S.Y., [Effects of angiotensin II, atrial natriuretic peptide III and arginine vasopressin on activity of paraventricular neurons of rat hypothalamic slices]. Sheng Li Hsueh Pao, 1993. 45(4): p. 405-12. 92. Yettefti, K., Orsini, J.C., el Ouazzani, T., Himmi, T., Boyer, A. and Perrin, J., Sensitivity of nucleus tractus solitarius neurons to induced moderate hyperglycemia, with special reference to catecholaminergic regions. J Auton Nerv Syst, 1995. 51(3): p. 191-7. 93. Yoshihara, T., [A neuroendocrinological study on the mechanism of feeding-associated circadian rhythm of plasma corticosterone and feeding behaviour in rats: a role of neuropeptide Y in the paraventricular nucleus]. HokkaidoIgaku Zasshi, 1995. 70(1):113-31. 94. ZZ-Topp, I. R., Ballad for Stress Research, J. Imponderable Results, 1997. 53(4):1-16 351 APPENDIX IV Preliminary Experiments and Antibody Testing Antibodies Tested and Used (*) [see Appendix VI for immuno-staining procedure] CRH* (*) indicates used for mapped cases. mAbrCRH @ 1:1000 From FJH Tilders, Amsterdam. Rat mono- clonal, fused w/ mouse myeloma cell line clone PFU83. Our stock is ascites fluid (batch 882) containing 3 mg/ml rat IgG2a. Both anti-mouse (PVH-1) and anti-rat secondaries work for this monoclonal. CRH* rabbit anti rCRH From Wylie Vale, shows some MCH #C-70 @ 1:3000 cross-reactivity (eg lat. hypothal) that can be blocked with Pro-Ile-NH2 di-peptide amide. CRH* sheep anti rCRH From Wylie Vale and Joan Vaughn, #174-217B @ 1:750 affinity purified (no MCH x-reactivity). Needs secondary grown in donkey. Not compatible with anti GRH (G-75). GRH* rabbit anti rGRH #G-75 @ 1:5000 From Wylie Vale. Does not cocktail with sheep anti-CRH (forms immuno-ppt!), cocktails well with monoclonals. GnRH (LHRH) (#40 is not quite as good) rabbit anti (Lys8)GnRH #44 @ 1:4000 From Wylie Vale. OXY* rabbit anti OXY From Dietryx in Belgium, adsorbed against argVAS. used @ 1:8000 OXY rabbit anti OXY From Chemicon, Inc. (rec. by Paul Sawchenko). Blocking exp’ts with argVAS not yet done. SS* mAb anti SS 1-14 From Novo Biolabs, cross reacts 100%w #SOM-018 @ 1:200 human somatostatin 1-28. Ascites fluid from mouse, contains IgG1, kappa. SS* rabbit anti SS(15-28) From INCstar (formerly Inc -- Immuno Nuclear). #20067 @ 1:2000 This is our standard rabbit anti-SS. Grown against the bioactivecarboxy terminus cleaved from pro-somatostatin(1-28). May be used in future mapping exp’ts. TH* rabbit anti TH From Pelfreeze, Inc. , grown against denatured tyrosine hydroxylase. #P40101-0 Lot 03813 @ 1:1000 Rec. by Paul Sawchenko. TH* mAbTH, clone 7D6 From Pelfreeze. Grown against SDS denatured rat tyrosine hydroxylase. #P80101-0, Lot #08413 @ 1:10,000 TH rabbit anti TH From EugeneTech Int’l. Used in past, not nearly as good as Pelfreeze As. 352 Antibodies Tested and Used (*), continued TRH* rabbit anti TRH From Gordon Orning, Ctr. for Ulcer Res. and Ed. -- UCLA / Brentwood VA (ref. Y. Tache). #8964 @ 1:3000 May be used for newer PVH expts. TRH* rabbit anti pre-pro TRH From Martin Wessendorf at U. Minnesota (Minneapolis). (ppT) #R363J @ 1:5000Grown against a part of the precursor molecule (pre-pro TRH160-169) that is clipped out during post translational pro-cessing, and is not in the bioactive tri-peptide. Used to show double-labeled cells with mAbrCRH in PVH-1. TRH rabbit anti TRH From Arnel Products, Inc. - lot 10205-9 no longer available. Small sample from Paul, requires acrolein fixation. Tested briefly, but no mapping study done (but mAbCRH cocktail saved). New lot purchased did not stain well. VAS* mAbVAS clone #3-D-VII@ 1:80 From G. Nilaver, Oregon HSU. ref: Hou-Yu, et al, J. Histochem Cytochem 30:1249-1260 (1982). Our sample is TC supernatant - not very high titer. Previous sample was usable @ 1:1500 (1989). Clone not currently being grown - difficult to obtain samples. VAS rabbit anti VAS From Chemicon, Inc. (rec. by Paul Sawchenko). So far, attempts at blocking with OXY have not been successful. Secondary Antisera Used for PVH Experiments Serum IgG, grown in goat (unless otherwise indicated) Against mouse primary antisera: American Qualex affinity pure FITC anti-mouse @1:100, (LaMirada, CA). American Qualex affinity pure Rhodamine anti-mouse @1:100. Vector Biotin anti-mouse (grown in horse) @1:200, (Burlingame, CA). Against rabbit primary antisera: Tago FITC anti-rabbit @1:200, (Burlingame, CA). Tago Rhodamine anti-rabbit @1:100. Vector Biotin anti-rabbit @1:200. Amersham Biotin anti-rabbit (grown in donkey) @1:200, (Arlington Heights, IL). . Against rat primary antisera: FITC anti-rat @1:500 -- gift from Jon Lindstrom. Vector Biotin anti-rat (grown in rabbit) @1:200. Against sheep primary antisera: Jackson FITC anti-sheep (grown in donkey) @1:200, (West Grove, PA). Jackson Biotin anti-sheep (grown in donkey) @1:200 353 Data Synopsis Chart Expt Antibodies, etc Photo Map Layers comments PVH-1 mAbCRH, TRH, yes yes yes cell outlines; peptides only; 33 d FG ENK,FB, FGserial set yes yes yes FG/FB borders; common fiducial PVH-1A shCRH, TRH, FG yes yes yes +FG map, no fids; no Nissl; 1/6 ser PVH-2B ppTRH, ISH non-colch yes yes n/a 1/4 ser; FB/FG photo before Nissl PVH-2C ppCRH, ISH non-colch yes yes n/a A=Nissl; all 3 sets have borders PVH-2D ppENK, ISH non-colch yes yes n/a uv & darkiefld photos of ISH PVH-3B ppTRH, ISH yes yes n/a 1/4 ser; FB/FG photo before Nissl PVH-3C ppCRH, ISH yes yes n/a A=Nissl; all 3 sets have borders PVH-3D ppENK, ISH yes yes n/a uv & darkiefld photos of ISH PVH-10A Nissl yes n/a n/a only 2 borders done: C-4/8, D-5/2 PVH-10B mAbVas, rbCRH a few no -- blk’d C-70-good; but wimpy Vas PVH-10C shCRH, Oxy yes yes yes fiducial marks carry through all PVH-10D mAbSS, TRH yes yes yes to“define mpdv; fid. marks move PVH-11B fitc - PNMT non-colch yes no --- photos @ 20X; A=Nissl PVH-11C fitc - DBH non-colch yes one --- photos @ 20X, one test map done PVH-11D mAbVas,Oxy n/colch yes two 2 photos @ 10x/ dual cube; non-colch PVH-12A Nissl yes -- -- one border: #4/3 “fingers” on each PVH-12B mAbVas, Oxy yes yes yes common fiducials on all sets PVH-12C mAbVas, CRH yes yes yes fitc-Vas, rhod-C-70 (unblocked-NEI) PVH-12D shCRH, Oxy yes yes yes lots of triples on 4/3 PVH-13B mAbTH, GRH yes no -- good fitcTH; GRH/composite map PVH-13C mAbTH, GRH no no -- partial set, (rhod TH); A=Nissl PVH-13D mAbTH, CRH 1 test no -- unblk’d C-70, beautiful staining PVH-14A mAbTH, TRH yes yes yes dual cube photos; no Nissls PVH-14B mAbTH, SS yes yes yes all sets have common fiducials PVH-14C mAbTH, CRH yes yes yes unblk’d C-70, bkd higher than 13D PVH-14D mAbTH, GRH no no -- no GRH in pvh; cells on compos. map PVH-16A TMB - ppWGA/HRP yes yes n/a injected by Ju’s lab in China PVH-16B OXY-DAB yes yes n/a 15um: 1-in-4 cryostat sections PVH-15 (control) shCRH/mAbCRH/rbC RH cocktails a few no -- 3 sets- 1/8; 100% co-labeling; mAbCRH autonomics a bit weaker PVH-15a mAb+polyTH; mAb+polySS no -- -- 98-100% co-labeled; confirms authenticity of mAb staining PVH-4 ArnelTRH, or WsTRH, w/ mAbCRH a few doubl no -- Acrol perf for Paul’s Arnel TRH; 1/6 ser; E=Nissl; F=for later ISH PVH-5 mAbCRH, TRH one no -- std perf; 1/6 ser; ArnelTRH didn’t work well; E=Nissl; F=for later ISH. PVH-6 mAbCRH,shCRH,w/Ar nTRH; WsTRH one yes no no before CRH-ISH Acrol perf-Paul’sArnelTRH;1/6 ser; C=IHC-ISH test;E=Nissl;F=for ISH PVH-7 mAbCRH,shCRH,w/W sTRH;WsTRH some yes no no before CRH-ISH 3-mAb, 8-shCRH photo’d; 1/6 ser; C=IHC-ISH test;E=Nissl;F=for ISH 354 Chart of Experiments: Photography, Mapping, Antibody Testing, etc. Expt # Photos Maps Separated Borders Comments, details PVH- 1 Photos : 5/1- 5/12 (L& R) 5/5- 5/6 (20X) 5/10: FG in cortex Maps: 5/1- 5/12 FG doubles too (Right) 5/5 mpdv and all of 5/6 also@20X Separated layers: 5/1- 5/12, peptides, (not ne/non) same fiducial scale --no 5/1-5/11 Borders : FG/FB (no nissls) (15 d) fast blue in blood, (33 d) fluoro-gold in spinal cord, 2 d colch; 15mm, 1/6 series; 6/'90.; fitc — rb a-TRH(Wss)/mAbCRH – rhod a-mouse; both sides photo’d.; TRH, CRH, and FG cell outlines drawn, only a few blues;100µm scale fiducial on separations with shaded doubles and highlighted strong IHC cells; also bilateral 3V and bottom-of-brain tracings made; color transparencies of 80% separated maps; (ledger to letter size) for CRH and TRH cells PVH- 1 4/1- 4/11 (Right) 4/1-4/11 4/2-4/12 4/2- 4/10 “ serial set to above, stained w/ anti-metENK; ENK, FG, and FB mapped and separated; color transparencies of 80% ENK cells; (note: 2 brains stained, one analyzed in detail) PVH- 1A 1R- 10R, &1L 1R-10R#2 missing, so 1L~=2R map 1R-10R non-ne doubles #3 ventral midline- bilateral “ + #5 : peptide groups last set of sections from PVH-1 brain; stained w/ sha-CRH-fitc and rba-TRH-TexRed; CRH(o), TRH(+) and FG (gold cells)mapped - w/blues filled; n.e. borders separated. no fiducial marks; CRH/TRH doubles — bit fewer than mAbCRH #3map->ventr. midline triples (blue) PVH- 2 B-D non- colch ISH darkfield - 10X 1'/2-6, 1/1-12, 3/1-6, SON for ENK uv's, pre-Nissl (L&R) w/ borders Left, best UVs ; TRH,CRH , ppENK; Layer s n/a yes, A=nissl uv photos of FG/FB before Nissl stain non-colch FG/FB as above;15mm, 1/4; ISH for TRH(B),CRH(C),and ppENK(D); 4 slides per set, w/ 12 sections each; surviving FB and FG cells on maps w/ black dots for ISH cells, and penciled borders. (Alan Watts quantitated SON (-) ppENK data for non-colch) PVH- 3 B-D colch ISH 1/1-12, 2/1-12, 3/1-6, also SON for ppENK uv's, pre-nissl, both sides w/ borders Left; CRH, TRH, & ENK; Layer s n/a yes, A=Nissl uv photos of FG/FB pre Nissl (pvh & SON) Colch’ed FG/FB, as above; ISH for CRH, TRH, and ppENK mRNAs; slides: 1'/1-6, 2/1-12, 3/1-12, 4/1-6 (Alan Watts quantitated SON (+) ppENK data for colch) PVH-4 FB /colch Acrolein 1A & 2A 10& 20X; double exposures w/ N2-I 2/3 cubes; 1, 4, 5B right doubles w/ both TRH As!! in notebook of CRH/TRH im. for doubles not mapped; doubles noted (?) should re-check w/ dual filter set E= Nissl 7dFB/2d colch; 0.05%Acrolein in 3%para pH9.0 - post fix 2h 3%para/sucrose - rinse 22h KPBS sucrose; 15mm, 1/6 series, A-D in DEPC cryo, E-F in 2%para DEPC cryo. Set C test (Paul's) Arnel TRH/mAbCRH; set A (Paul's) Arnel TRH/ mAbCRH; Set B WssTRH/mAbCRH (std); set D test of new Arnel TRH; set F mounted for ISH. (NB:Paul Sawchenko has done shCRH/TRH study to see TRH terminals around CRH cells--not published) PVH-5 FB colch setA#5R; std cocktail & 2 nd ; pre-ISH photos DAB IHC-R&L no map ('better')7d FB/2d colch, std perf5/31/91; 15mm 1/6 series 6/4/91. Set C test TRH/mAbCRH/Arnel TRH; set A -(Paul's)Arnel TRH/mAbCRH; set B mAbCRH/TRH (orig); set D test of new Arnel TRH diln's& fitc a-rat vs rhod a-ms on mAb CRH - two doubles seen on slide 4B; set F mounted for ISH. E=Nissl (√ re-check w/ dual filter set) 355 Exp’t # Photos Maps Separated Borders Comments, details PVH-6 FB/colch; (Acrolein) set A#5R only; D- pre-ISH photos of DAB IHC (R &L) --in CRH/TRH immuno notebook not mapped; (?)should re-check w/ dual filter set E= Nissl 7dFB/2d colch; .05%Acrolein in 3%para pH9.0 - post fix 2h 3%para/sucrose - rinse 22h KPBS sucrose; 15mm, 1/6 series, A-D in DEPC cryo, E-F in 2%para DEPC cryo. One slide each stained: A- rhod a-ms, mAbCRH - (Paul's Arnel)TRH-fitc; B- fitc sh a-CRH - (Paul's Arnel)TRH-TxRed; C- fitc mAbCRH - (Paul'sArnel)TRH-Tx red; and D- a-TRH immuno + CRH in situ test. 8 sections/slide. PVH-7 FB/colch Set B #1-3R; Set C #1-8R setD - some pre-ISH shots of TRH imm. --in CRH/TRH imm notebook (spectacula r fitc mAb CRH) check w/ dual filter No maps E=nissl, section s a bit mixed up 7dFB/2d colch; 15 mm 1/6 series, A-D in DEPC cryo, E&F in 2%para-DEPC cryo. B- fitc sh a-CRH/TRH-TxRed one slide; C- fitc mAbCRH/TRH-TxRed one slide; set D a- TRH-IHC + CRH in situ testing; set F mounted for later ISH. PVH- 8 TPX/ADX (Acrolein fix) (non-colch) scope survey only; look @ 20x for CRH/TRH doubles Thyroidectomy +10 d Adx: perf w/ .04% Acrolein in 4%para pH 9.5;cut 6/21/91:15 mm, 1/6 series - sets A-D in DEPC cryo (saved w/ PVH-4); sets E&F mounted and post-fixed for later ISH. Set C stained w/ mAbCRH / (Paul's Arnel)TRH cocktail- many, many TRH fibers, some PVH cells (not much elsewhere); some CRH cells in PVH, not many fibers; no doubles seen; PVH-9 TPX/ ADX (non- colch) scope survey only (9B was 2nd test on mAbCRH) A= Nissl Thyroidectomy +10 d Adx: cut 6/21/91:15 mm, 1/6 series - sets A-D in DEPC cryo (saved w/ PVH-5); sets E&F mounted and post-fixed for later ISH. Set B mAbCRH/TRH cocktail-test of Jon's anti-rat fitc + strAvTxRed vs rhod anti-ms std 2nd; D dil'n test of cocktail saved fr/ B; set C stained w/ mAbCRH/TRH cocktail - many TRH fibers (high background) some cells in PVH - not many elsewhere, also some CRH cells in PVH; no doubles PVH10A 1/5 - 7/8 (Right) 3/1-6/1 Right n/a yes note: Nissls mixed up during mounting. Eleni M. figured it out, and provided a conversion chart to use when analyzing the immuno. PVH10B 4/4: tests of new uv bulb good CRH, wimpy Vas (12d)FastBlue/(2d)colch; 15 mm, 1/4 series. fitc-CRH (blocked C-70) / mAbVas rhod.; cocktail; good CRH stain was surveyed for stain agreement w/ shCRH; wimpy Vas staining (too dilute@1/1k), so not photographed. Trial rescue re-staining on puddled slides didn’t work. Test section photos of biotinTxRed on 1:250, 1:500 mAbVas(good!) in PVH-10 notebook PVH10C 3/1 - 6/1 some arcuate 3/1- 6/1 3/3-6/1 fiducial in common 4/8 by nissl "fingers"IHC fitc sh a-CRH(o) / Belgian a-Oxy(+) TxRed Eleni M. did GRH-ARH tests & bis benzimide Nissl tries PVH10D 1/5-7/3; 7/3-7/4 bis benzimide tests, IHC is gone 3/2-5/8 3/2-5/8; new fiducials- each section 5/2 by Nissl "fingers"IHC rhod-TRH(+)/fitc-mAbSS(o),"to define mpdv". Right side mapped; fiducial marks move from section to section; #5/2 left mapped as ~#5/1 right (no section 5/1 or 5/7) 356 Exp’t # Photos Maps Separated PVH-11 non-colch 6/1-11/5(20X) right=data;8/6- 9/3 L=test 10 &20X A= Nissl non-colch, (12 d) fast blue in blood. 9/18/91 perf-same group as PVH-10&12. fitc anti PNMT (Bohn) for fibers in PVH -10X are impossible to analyze; so right side photo'd @ 20X for data -- slides 5-12 PVH-11C non-colch 6/1-11/4 (20X) right=data 8/6-9/3 L=test 10 & 20X one test map of 9/3L; wait, for more non-colch, (12 d) fast blue in blood.; fitc mAb-DBH, fibers to be mapped @20X with uv slide of blues; to define limits of PVH. sections missing: 6/2, 6/6, 10/2 PVH-11D non-colch 6/2-10/5 (10X) right=data dual cube; SON - 6/3,7/2 non-colch, (12 d) fast blue in blood.; fitc-Belgian aOXY (1/3k)+ mAbVas(1/80)TxRed; Incr. conc of Oxy & use biotinTxRed forVas to see good staining on this non-colch tissue. Sections missing: 8/5, 8/6. #11/4=end of pvh, #11/5 to end. Rick did ABC of DMH w/ mAbDBH (didn't work...) PVH-12A 1/5- 5/8 1/6 - 5/7 n/a yes Nissl set of final dataset experiments below PVH-12B 1/6-5/8: Right also,1/5 SON fitc and N2, 1/8 red only 3/3L SON 10 & 20X dual cube filed w/ nissls 1/6-5/7; 3/2- 5/2 fiducial marks carry through all sections 4/3 – A= Nissl "fingers " IHC (12d)FastBlue/(2d)colch, 15mm, 1/4 series Rhod Belgian a-Oxy(+) / fitc mAb-Vas(o); Oxy@1/8k, Vas@1/80; note: later test of 1/250 & 1/500 mAb-Vas (48h) worked w/ biotin 2nd and TxRed StrAvidin on colch- PVH. No Vas cells seen on 1/6-3/1 or 5/3- 5/7; Spectacular SON parcellation of cells - may be a double or two (?) more likely to be cell overlap.... PVH-12C 1/6-5/8 1/6-5/8 3/1-5/2 as above fitc mAb-Vas(o)/rhodamine CRH(+)—C-70 not blocked w/ NEI dipeptide; one triple on 4/3; #1/6-2/8 & 5/3-5/8 no Vas cells, only CRH mapped, and no separations done on these. PVH-12D 1/8-5/8 3/5R SON 1/8-5/8 2/1-5/7 as above fitc shaCRH (+)/ Belgian aOxy(o) texas red; tx red shows on all filters - 1/3k Oxy too conc; Lots of triples on 4/3; spectacular OXY staining photoed in SON PVH-13B Photo entire series:1/3-7/6: Right (good fitc-TH); also photo’ed FB/colch #1-7/31; 5/3&5,6/3&5 for (?)pvh/GRH Not mapped no GRH cells in pvh; composite map on atlas plates (expt'l notebk) A= Nissl (14 d) fast blue/ (2 d) colch; 15mm, 1/4 fitc mAbTH / rhod GRH -- good fitc TH serie; some rostral GRH cells plotted on composite map. 7/6 shows good GRH in arcuate; nice blue in OVLT and SON on 1/3 note: composite map has cells from 4 of 12 brains. 8 totally(-)for GRH in rostral hypothal. PVH-13C not photoed partial set, rhod mAbTH/GRH-fitc; no GRH, few TH PVH-13D 6/6L test photo fitc-CRH/rhodTH; photo in notebook/13B; Staining beautiful (unblk'd C-70) 357 Exp’t # Photos Maps Separated Borders Comments, details PVH- 14A 3/1-9/6: Left dual cube 5/4&5 all of 3V= fiducials Map 3/1- 9/6: Left (all) 3/1-8/6 - end of TRH blues all PVH-14 same fiducials no Nissls ne borders from fast blue cells But see “D” (15d)fast blue/(2d)colch #1- 9/92;15mm, 1/4; (Wss)aTRH(o) fitc / mAbTH(+) rhod. 3/15/93; 3/6 end of a.c., 5/2 top of 3v closed; 1st TRH blue is 4/5 - end 8/6; PVH- 14B 2/3-9/3 dual cube 2/3-8/1, 8/2-9/1 top Map 2/3- 9/1 Borders as above fitc(Incstar)aSS (o)/ mAbTH (+) rhod.12/22/92; 8/1 is arcuate level PVH- 14C 2/2-8/1 dual cube 3/6- 7/6 3/6- 7/6 as above fitc a-CRH(o)/rhod mAb-TH(+). 12/21/92. C-70 w/ NEI (old) dipeptide not blocked. Background higher than PVH-13D; 3/6 is 1st section caudal to a.c.; 6/4 map reduced for trial scanning PVH- 14D fitc-GRH/mAbTH; few+GRH onto composite map; no TH map. NB: This set Stained later for Nissl –worked great, in spite of earlier Triton exposure! PVH-15 CRH - control some shCRH/ C-70: 5B, 7B, 12B; mAbCRH, C70 (ant magno) 10A&12R;10A SCH; PVH 2/5R & 2/6R w/ 4X6 print no maps-- all sections checked @ 20X on scope no Nissl -set A saved in cryo @-20 (15d)FB/(2 d)colch test rat #4 - 9/92 group; one other set used for GRH/TH, one set left; two serial,15mm, 1in4 series divided into 4 groups (= 1/8 series for each cocktail); set C-odd is fitc-mAbCRH w/ unblocked C-70-rhod; C-even is fitc-sh a-CRH w/ unblocked C-70-TxRed; set B-odd is TxRed mAb- CRH w/ fitc sh a-CRH; and B-even is held in reserve. Result: 100% co-localization-specific CRH; Autonomics are a bit weaker w/ mAbCRH. Good demo of NEI staining w/C-70 (esp C-even) -- cells typically "punctate-hairy" looking PVH-15 (a) TH&SS mAb /poly control) no maps-- all were checked 20X on scope (14 d) FB/colch 4/92 group; 15mm, saved as 1/8 series across 2 trays, so easy to use either 1/4 or 1/8 for staining. Test rat #1 - 1/10k-mAbTH/1/2k-ETI TH cocktail; 1/10k-mAbTH/1/2k-Pelfreeze TH cocktail; 1/200-mAbSS/1/2k-IncSS cocktail. SS is 100% co-labeled. TH cells are 98%+ co-labeled, 100% of (arcuate) n.e. cells are co-lableld, no fine fibers in cortex w/ mAbTH ... implies poly-clonal is probably more sensitive. PVH-16 Photo PVH; +TMB in pit Map PVH This is a posterior pit injection of WGA/HRP from Gong Ju’s lab in Xi’an China (“R2-p.p”) 15µm, 1/4 cryostat sections; set A is TMB of retrogradely labeled cells in hypothalamus, set B is anti-OXY(DAB), set C is (poorly staining) anti-CRH, and set D is Nissl. NB: the key Nissl slide is missing(!) on arrival… Darkfield TMB w/ PHA-L scope & condenser is spectacular - looks like optical segmentation - colored purple & intense aqua 358 Exp’t # Photos Maps Comments, details PVH- IHC tests 1/5-1/8, 2/1-2/4 both L&R some fitc-CRH; some fitc bleedthru to N2& (loan fr/Leica) N2.1 not mapped; doubles seen slides in CRH/ TRH ihc notebook) sections from “colch test brain”’91; test of fitc- shaCRH/TRH TxRed about 12; Lindstrom fitc- arat vs rhod-ams – mAbCRH; test mAbCRH vs C-30 & C-70 (all fitc) check fitc bleedthru to rhod filter (mAbCRH << poly,someTRH) nil for shCRH fitc; Paul’s TRH fitc toN2 is nil; rhod to I2/3 is + PVH- test rats scope survey: misc tests 4sets of 4 FB/(2d)colch rats done for GRH:; 12dFB:9/91(3 colch, 2 non-colch) #1 is PVH-10; #3 is PVH-12, #5-non is PVH-11. #2A shCRH/rbCRH 11/91 test didn't work- high bkgrd; #2B is a TH/GRH; part of #2B used for mAb/polySS and mAb/polyTH cocktails; others used for mAbVAS testing; #4 died. 14dFB:4/92-#4 is PVH-13, #1-3 lots of rbc's, save for misc testing after GRH test. 15dFB:7-8/92-v strong FB (pericytes) some GRH(?)-pvh photos of #1 Balance saved for misc testing after GRH test. 13dFB:9/92-#1 is PVH-14, #3 tested for GRH and discarded (too many rbc's), #2 & 4 saved for testing after GRH test. Other Misc brains... FG/FB#1&2:4/90RevcoTfluctuation:discarded FB/colch: #1-3 - 12/91-“hypothal blocs frozen” TPX expt rats (all have FB) scope survey: "poor IHC results" all grossly overstained... confusing - badAb? all were FB++ all std pHshift perf; 15mm 1/6 series,A-C DEPC cryo (A - mAbCRH/WssTRH, like PVH-1), D-F 2% para DEPC cryo (D-F all mounted and post fixed).; 2 ea Normal (9d FB); 2 ea ADX (9 d Adx/FB); 2 ea ADX/TPX (9 d Adx/FB) #1: set B&C also used for IHC testing; C looks similar to PVH-9C; 2 ea TPX (9 d FB) 1 ea TPXcolch (9d FB/ 1h colch) 72 h post fix, sections cut and saved as above, none used 359 APPENDIX V ISH experiments listed in Appendix IV were all performed using the following procedures. Reagents and instructions for in situ hybridization From Simmons, et al., 1989 ISH protocol paper, and Simmons 1990 “One Week Short Course in In Situ Hybridization” Baylor University College of Medicine, Neurosciences Division. Cryoprotectant Tissue Storage Solution for ISH (contains fixative) 2% paraformaldehyde in D.E.P.C. cryo-protectant solution 4/96: D.Simmons DEPC=diethylpyrocarbonate, RNAse inhibitor. use for ISH tissue (best preservation of mRNA) do not use for tissue that may be destined for immuno-staining (over-fixation blocks IHC) In a sterile graduated one liter bottle, make 4% para in 0.05M phosphate buffer: 350 ml DEPC water 20 g paraformaldehyde one pellet NaOH (about 0.1 g) 96.2 ml 0.2M sodium phosphate - dibasic (B) (use sterile 25 ml pipet, to avoid RNAse) cover tightly, and heat in 65 degree water bath until dissolved - about 45 min. (open container in hood because of formaldehyde fumes) add 28.8 ml 0.2M sodium phosphate - monobasic (A) cool to room temp. and adjust final volume to 500 ml with DEPC water if needed test pH with a drop from a sterile (RNAse free) pipet onto pH paper -- should be 7.3 - 7.4. using graduations on liter bottle, add: 300 ml ethylene glycol 200 ml glycerol mix well and store in the freezer sodium phosphate buffer 0.2 M stock solutions: A: sodium phosphate - monobasic 13.7 g NaH2PO4 in 500 ml total volume DEPC water B: sodium phosphate - dibasic 28.6 g Na2HPO4 in 1000 ml total volume DEPC water convenient to use Corning sterile storage bottles (styrene, orange plastic cap) to make these solutions -- their graduations are pretty accurate. This helps avoid ubiquitous destructive RNAses found on skin and hair and almost everywhere. RNAse: Modern ‘RNAse-Away’ products are extremely useful for cleaning counters, tissue mounting brushes, etc. RNAse –free water and buffers can be purchased. DEPC treatment is often no longer necessary. 360 (Days 1 and 2 are used for perfusion and sectioning) Day 3 Pre-Hybridization Treatment: FROM NOW ON, WEAR GLOVES TO HANDLE SLIDES; USE STERILE TECHNIQUE AND STERILE REAGENTS. (if so, its not necessary to use DEPC-treated water to inhibit ribonuclease.) A) Prehybridization Treatment: 1. 0.001% proteinase K digestion--30 minutes at 37°C (made up in 100mM Tris pH 8.0, 50mM E.D.T.A., 10µg/ml Prot. K). For 250mls: 25 ml 1M Tris pH 8.0 25 ml 0.5M E.D.T.A. (pH 8.0) 200 ml sterile D. H2O 250 µl 10 mg/ml Prot. K (1% in D.H2O-kept frozen) 2. Rinse briefly and gently in D. H2O (blue cap, de-ionized) 3. Rinse in 0.1M TEA (triethanolamine) at pH 8.0 -- 2 or 3 mins. (Pre-equilibration buffer for next step) 25 ml 1M TEA - pH 8.0 225 ml D. H2O 4. Acetylation--10min. At room temp. (to block positive charges) Add 625 µl acetic anhydride to dry slide dish Add 250 ml 0.1M TEA pH 8.0 Then, place slides in dish 5. Rinse briefly and gently in 2 X SSC (NaCl/NaCitrate, 2 changes, 2 min. Each). SSC is made in 20 X stock solution (see attached). 2X is 25 ml stock plus 225 ml d H2O. 6. Dehydrate quickly--in ascending Ethanol concentrations. 50% EtOH -- 3 min. 70% EtOH -- 3 min. 95% EtOH -- 3 min. 100% EtOH -- 3 min. 100% EtOH -- 3 min. (New solution) 7. Drain well and store with dessicant under vacuum at room temperature at least 2 hours until hybridization. Can leave here overnight. If you must wait longer to hybridize, store dessicated at -20°C. Note: for embryos or other very fragile tissues -- use dilute Triton X-100 instead of Prot. K (5ml of 25% Triton in 250ml tris-EDTA), rinse in TE, then dehydrate in EtOH. NOTE: For frozen sections of embryo at 5-10um, proteinase K digestion is much too harsh. Instead, we permeabilize the tissue by pretreating with 0.5% Triton X-100 at room temperature in TE buffer for 30 min. Sections are then rinsed in two changes of buffer and dehydrated -- no acetylation needed. For lightly fixed tissue, or if section detachment seems a potential problem, pretreatment (other than alcohol dehydration) may be omitted completely. At the opposite extreme, paraffin processed or glutaraldehyde fixed tissues may require more pretreatment for adequate probe penetration. 361 Day 4 Post-Hybridization Treatment: Radioactivity Hot 1. Cool slides, and peel off dried DPX with sharp forceps. Very Hot 2. Soak off cover glass in 4 X SSC rinse (use gentle agitation). Takes 10-20 minutes. Let cover glass slide off, don’t pull, as tissue may be fragile. Hot to Cold 3. Four 5-min. Rinses in 4 X SSC after cover glasses are off. Use rotator table for gentle agitation. 40/200 dilution of 20 X SSC--stock solution Very Hot 4. RNAse digestion--30 min. At 37°C 500 µl RNAse A:10 mg/ml (kept in freezer) 25 ml 5M NaCl 2.5 ml 1M Tris pH 8.0 500 µl 0.5M E.D.T.A. (pH 8.0) 222.5 ml D.H2O FROM NOW ON, ADD 1mM DTT TO ALL SOLUTIONS WHEN USING 35S 250 µl of 1M DTT to each 250 ml of solution Cold 5. Rinse and gradually de-salt. 2 X SSC (+DTT) -- 5 min. (25/250 dilution) 2 X SSC (+DTT) -- 5 min. 1 X SSC (+DTT) -- 10 min. (12.5/250 dilution) 0.5 X SSC (+DTT) -- 10 min. (6.5/250 dilution) Hot 6. 0.1 SSC (+DTT) -- 30 min. At 65°C (2.5/500 dilution) Cold 7. 0.1 SSC (+DTT) -- brief rinse at room temp. Cold 8. Dehydrate quickly in ethanol with salt and DTT. (1 ml of 20 x SSC and 250 µl of 1M DTT to each 250 ml EtOH.) 50% EtOH (+SSC,DTT) -- 3 min. 70% EtOH (+SSC,DTT) -- 3 min. 95% EtOH (+SSC,DTT) -- 3 min. 100% EtOH (+SSC,DTT) -- 3 min. (3 changes) 9. Drain well and vacuum dry at room temperature for 30 min. NOTES: • The above reduction in salt content increases the stringency of hybridization conditions, as does the increased temperature in the 0.1 SSC rinse. This serves to wash away most of the non-specific label. 362 Hybridization Solution We have done dilution series, and find 5 X 106 cpm/ml gives good signal with low background for probes of 300 - 700 bp. Other probes may vary, due to UTP content and/or message abundance. A. Part I of hyb solution (use 800µl per ml of final hyb solution) Make up in sterile 15 ml (or 50 ml) cent. Tube, mix well and store @ 4°C 10 ml formamide (kept frozen) or: 25 ml formamide BUFFER PART ONE 4 ml Dextran sulfate (50%) 0 ml 50% dextran sulfate 400 µl 50 X Denhardt’s Solution 1 ml Denhardt’s 40 µl 0.5M E.D.T.A. (pH 8.0) 0.1 ml E.D.T.A. (0.5M pH 8.0) 200 µl 1M Tris (pH 8.0) 0.5 ml Tris (1M, pH 8.0) 1.2 ml 5M NaCl 3 ml NaCl (5M) 160 µl d.H2O (DEPC) 0.4 ml d.H2O (DEPC) I aliquot formamide 10 ml/15 ml tube or 25 ml/50 ml tube and freeze. Then, it is easy to measure the dextran sulfate by volume increase in the graduated tube. B. PROBE (Part II) For 10 ml Hyb. Solution: (use 15 ml sterile tube) 500 µl tRNA 100 µl 1M DTT ___ µl probe (volume varies with cpm/µl probe) ___ µl d.H2O (DEPC) to total 2 ml Heat to 65°C - 5 min. C. HYBRIDIZATION SOLUTION (Part III) To the 2 ml of probe mixture (B), add hybridization buffer (A) to total 10 ml (buffer is 80% of final volume). Close tube tightly and vortex to mix well; remember, this is a radioactive substance. If using immediately, centrifuge at 2000-4000 rpm for 10 minutes to reduce bubbles and eliminate any dextran sulfate precipitate (occasionally seen with some probes). Store at -20°C. Remove small amounts as needed or (better) aliquot into sterile microfuge tubes before storage. Use the half-life of the probe signal as a guideline of how long to keep stored Hyb-Solution. Hybridization solution stored at -20°C should be heated for 10 minutes at 65°C and centrifuged before use. NOTE: Since probe transcription is usually very efficient - yielding enough for 20-50 milliliters of ISH solution, we always make at least 10ml of ISH solution, rather than the 1ml test volumes used in early work. Thus, more ISH solution can be used on each slide; and we now use a sterile plastic disposable Pasteur pipet to apply solution to the cover glass. This assures an adequate volume [est: 100ul - 150ul] to cover the tissue and prevent high background from pressing the cover glass down too firmly. It also avoids the potential of pipettor contamination with radioactive ISH solution. Once the new probe has been tested, we typically make up the entire volume into aliquots of ISH solution for use on several experiments. 363 ISH Solutions Preparation List Reference Books: (Solution Formulas & Lab Hints) “Molecular Cloning: A Laboratory Manual”; T. Maniatis et al., Cold Spring Harbor (1982) (pp. 446-448 lists most reagent formulas) “Basic Methods in Molecular Biology”; Davis, Dibner, & Battey. Excellent General/Theoretical Reference — a technical paper on probe synthesis and in situ hybridization with cRNA probes (one of the original papers on the topic): Melton, D.A., Krieg, Rebagliati, Maniatis, Zinn, & Green: Efficient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter. Nucleic Acids Research: Vol. 12 #18 (1984). _____ Newer, better, references are no doubt available, since this list was composed _____ -- General stock solutions, large volumes needed -- DEPC water == RNAse free water. Add 0.1% Diethylpyrocarbonate to D.I. water in autoclavable containers. Tip them so that the lid is treated as well. Let solution stand to de-activate the RNAses for 12-20 hours. Then, autoclave for at least 25min (the DEPC boils off -- as EtOH vapor and CO2). DEPC is a caustic solution, stored in the refrigerator—work in the hood! It does not keep well after opening: order small bottles and use entire volume at one time. Sodium phosphate buffer - monobasic and dibasic salts: make 0.2M stocks with DEPC water. For .05M working solution at pH 7.3: use 5.75ml 0.2M monobasic + 192.5ml 0.2M dibasic + 801.75ml DEPC water. To be used in cryoprotectant. Cryoprotectant 20% ethylene glycol, 20% glycerol, 50% .05M sodium phosphate buffer at pH 7.3. Use DEPC water. Include 2% paraformaldehyde (4%, dissolved in the buffer component—immerse in a water bath to dissolve at this pH) for in situ-only sections. (Post-fixation directly after cutting will improve ISH signal but decrease IHC signal). Buffered saline (KPBS): This is our standard immunohistochemistry buffer. Make 6x stock: 0.12M potassium phosphate, 5.6% NaCl: 2.96g monobasic K-phosphate; 17.2g dibasic K-phosphate; 54g NaCl in one liter total volume. Working solution is .02M buffer in physiologic saline: Dilute 1 part stock to 5 parts D.I. Water [e.g., 150ml stock +750ml D.I. Water = 900ml working buffer]. SSC-20X stock: Sodium chloride (175.3g/l), sodium citrate (88.2g/l) Use solution at pH7.0 This is a standard Mol-Bio buffer. (1X~= physiologic osmolarity and pH) Make with D.I. water and filter sterilize. 364 ISH SOLUTIONS PREPARATION LIST (continued) 5M NaCl: Make in DEPC water. This concentration is close to saturation and may be difficult to filter, so autoclave to sterilize. Tris: Tris hydroxymethyl aminomethane. Prepare a 1M pH 8.0 solution. (Sigma sells acid and base salts to combine, or “Trizma Base” to use with NaOH). Filter sterilize. EDTA: Ethylene diamine tetra acetic acid. Make 0.5M solution, adjust to pH 8.0 and filter sterilize. Tetra-sodium salt is much easier to dissolve, others may need heat. TE: Tris-EDTA: 10mM Tris pH 7.4, 1 mM EDTA. Use DEPC water, Filter sterilize. (Used in probe/hyb solution) TEA: Triethanolamine hydrochloride Make 1M stock solution with D.I. water. Adjust to pH 8.0 and filter sterilize. --General solutions, smaller volumes needed - made infrequently— 20% SDS: Sodium dodecyl sulfate (detergent) use DEPC water and measure with a sterile instrument. Dissolve at 60°C (2 hours). Keep sterile. SET: TE with 0.1% SDS . Use to dilute probe to final volume and to stop the transcription reaction. Also to equilibrate P-60 or spin column so probe RNA doesn’t degrade. 50X Denhardt’s Solution: Use 5g each of Ficoll, polyvinylpyrrolidone and BSA (RNAse free), made up to 500 ml with DEPC water. Filter sterilize, aliquot and store at -20°C. May be thawed and re-frozen many times if kept sterile. DTT: Dithiothreitol (Cleland’s reagent) make 1M with DEPC water (6ml/g -- add DEPC water to original bottle, to minimize sulfur smell and keep sterile). Stabilizer for the sulfhydryl bonds in the radiolabeled-sulfur of the probe. Freeze this smelly reagent in single-use aliquots (~2ml). tRNA: Transfer RNA from Yeast. Make 10mg/ml concentration with DEPCwater (add to vial, or use sterile cfg. tube) and aliquot for storage at -20°C. Prot. K Proteinase K for pre-treatment permeabilization of sections. Make 10mg/ml: add 10ml DEPC-water to 100mg vial. Aliquot and store at -20°C. RNAse A: Ribonuclease A (from pancreas). Dissolve in DEPC water @ 10mg/ml, (add 10ml to 100mg vial) heat to 65°C about 10 min. to kill DNAse) aliquot and store at -20°C 50% Dextran Sulfate: Make volumetrically with DEPC water, store at 4°C . This polymer is a space filling powder that dissolves slowly to a very viscous solution, so the most practical way to make it is volumetrically in a calibrated sterile wide mouth bottle larger than the desired final volume (e.g., a100ml autoclave bottle calibrated @ 40, 60, & 80ml). For 80ml total: assume the reagent is sterile, and add 40g with a sterile instrument to about 40ml of DEPC water in the bottle. Cap securely, and allow to dissolve overnight in a 65-70°C water bath with occasional mixing. Only after complete solution will you be able to add more DEPC water to achieve the correct final volume, as the powder is voluminous. NOTE: If you add 40g to 80ml of water at the outset, the final volume and concentration will not be the same! Dextran Sulfate polymer increases effective concentration of the probe at the surface of the tissue, so whichever method you use, make sure you always prepare it the same way. 365 In Situ Hybridization Supplies Checklist NEEDED FOR DAY 1/ perfusion: _ Anesthetic, and dissection instruments _ Physiologic Saline _ Fixative Solution A (make A and B fresh) _ Fixative Solution B, ( and 25% Glutaraldehyde, if used) _ Post Fixation Solution (Solution B + sucrose) NEEDED FOR DAY 2/sectioning & mounting: (CLEAN--USE GLOVES to handle): _ Cold buffer or cryoprotectant for section storage _ Subbed, Poly L-Lysine coated slides (must be dry, prepare ahead). 10/98: Fisher Brand™ “Superfrost/Plus” slides work as purchased! _ Cryoprotectant solution (“Anti-freeze”) for long-term section storage. (Use solutions made with DEPC water). 2 types: for regular IHC, or with 2% paraformaldehyde for ISH. _ New, or very clean, tissue section storage containers (24-well tissue culture plates) _ Clean brushes set aside for ISH. (Use “RNAse-Away” to clean). _ Cold 4% paraformaldehyde—pH 7.3, for post fixation of slides. NEEDED FOR DAY 3 / probe synthesis and hybridization: (All RNAse-FREE) For Probe Synthesis: _ Isotope e.g., New England Nuclear™ #NEG-039 a-thiol 35S UTP _ Linearized DNA template (preferably, gel purified - at 1mg/ml) _ 1M DTT: Promega™ kit _ Sterile D.H20: Promega™ kit _ Transcription Buffer: - Promega™ kit _ G,C,A, Stock Solutions—Promega™ kit _ RNAsin—Promega™ kit _ RNA Polymerase—Promega™ kit, or purchased individually. All the above are kept at -20°C _ S.E.T. (Stock)--0.1% SDS (Sodium dodecyl sulfate) --0.1M Tris pH 7.4 (7.2 - 8.0) --0.5M E.D.T.A. pH 8.0 _ Spin column (Boehringer Mannheim™ G50 or other brand, size appropriate for separating un-incorporated NTPs from labeled probe) _ 37°C water bath, 65°C water bath For Hybridization Solution: Amounts for 10 ml Total _ Formamide (stored @ -20°C }”hyb buffer-Part I” 5 ml _ 5M NaCl } 600 µl _ 1M Tris pH 8.0 } can be made ahead 100 µl _ 0.5M E.D.T.A. pH 8.0 } in much larger 20 µl _ Denhardt's Solution (@-20°C) } volumes 200 - 500 µl _ 50% Dextran Sulfate (@ 4°C) } 2 ml _ tRNA--10 mg/mL (@-20°C) } “hyb sol’n Part II” 500 µl->1ml _ Synthesized Probe (@-20°C) } 10 -> 900 µl _ Sterile D.H2O } (Make Fresh) 10 -> 900 µl _ 1M DTT (@-20°C) } 100 µl 366 In Situ Hybridization Supplies Checklist (continued) NEEDED FOR DAY 3/pretreatment and hybridization: (All RNAse-FREE) If slides were not post-fixed, use 4% cold paraformaldehyde to do this for one hour, then dehydrate and dry before Prot. K pretreatment below. _ Proteinase K--10 mg/mL _ sterile items: ( aliquots stored @-20°C) 1.5 or 2 ml microfuge tubes _ 1M Tris pH 8.0 pipet tips _ 0.5M E.D.T.A. pH 8.0 fine-bore disposable pipets _ Sterile D.I. water _ 37°C water bath _ 1M TEA pH 8.0 _ 65°C water bath _ Acetic Anhydride _ 55-60°C slide warming tray _ 20 x SSC (Stock) _ aluminum foil _ 50%, 70%, 95% Ethanol _ 22 mm wide cover glasses _ Absolute Ethanol (30, 40 or 60 mm long) — plastic slide staining dishes _ DPX mountant (used only for Pre-treatment) (methacrylate in toluene) NEEDED FOR DAY 4/post hybridization treatment: (Sterile) _ 20 x SSC (Stock) _ radioactive waste disposals _ RNAse A--10 mg/ml (for solids, and liquids) (aliquots stored @ -20°C) _ slide staining dishes: _ 5M NaCl marked for “RNAse-only” _ 1M Tris pH 8.0 _ 37°C and 65°C water baths _ 0.5M E.D.T.A. pH 8.0 _ 50%, 70%, 95% Ethanol _ 1M DTT (aliquots @ -20°C) _ Absolute Ethanol Stock Solutions Needed for One Experiment: (Assume 250 ml slide dishes and only one rack of slides) 1M Tris pH 8.0 32 ml Others: (≤ 1 ml) 0.5M EDTA pH 8.0 32 ml acetic anhydride 1M TEA 50 ml proteinase K 20 x SSC 375 ml RNAse A 0.5M NaCl 25 ml Formamide } hyb Sterile Water 250 ml Denhardt’s Soln. } buffer 1M DTT 2.5 ml 50% Dextran SO4 } part I 95%, 70%, 50% Ethanol 500 ml ea tRNA, EDTA, } buffer Absolute Ethanol 1000 ml Tris 8, probe } part II Kodak D-19 developer and Rapid Fixer for developing X-Ray images (make with distilled water and store in small volumes e.g., 500 ml) It’s good to plan the week before an expt as solution preparation time: Lysine coating of slides is a do-ahead project that spans a couple days, and the finished slides are good for weeks/months. NOTE: Fisher Brand™ Superfrost/Plus ISH slides work as received!! (But not if stored a long time) All except the alcohols, should be made ahead as sterile solutions, or aternatively filtered- or autoclave-sterilized after dissolving. Denhardt’s solution aliquots will last for months/years when frozen. Dextran sulfate takes up to 2 days to dissolve, but lasts for months. 367 Addenda to published ISH protocol D. M. Simmons <dsimmons@usc.edu> University of Southern California FIXATION: This protocol is optimized for performing immunocytochemistry or in situ hybridization on alternate series of sections from the same material, therefore fixation time is rather minimal. In general, for hybridization studies alone, longer fixation is better that shorter. Tissues fixed for several days to weeks before sectioning hybridize quite well, as do sections from paraffin embedded material, and immediately post-fixed cryostat sections from fresh-frozen tissue. Note that nuclear precipitating fixatives such as methanol or Carnoy’s solution do not adequately preserve mRNA in situ -- aldehyde fixatives must be used. Aldehyde cross-linkages merely trap mRNAs in a meshwork, rather than covalently bonding them, and thus they are exposed to possible diffusion from the cut surface of sections. Therefore, additional fixation after cutting is often helpful for best hybridization -- especially for low-abundance mRNA. We post fix sections either by storing free-floating frozen sections in cryo- protectant that includes 2% para-formaldehyde, or by cutting them into 0.25% paraformaldehyde in buffer for mounting the same day, then immersing the air dried mounted sections in cold 4% paraformaldehyde for 60 minutes, followed by ethanol dehydration before pretreatment or storage. We routinely dry cryostat sections of fixed tissue on a (50 degrees C) hotplate before post-fixing. Unfixed frozen sections can also be dried before fixation, to assure their firm adhesion to the slides. SLIDES: Many commercial ISH slides tested in the past gave very high background. However, we have recently begun using Fisher Brand Superfrost/Plus [#12-550-15] slides from the box as received—with good results. New lots and vendors (and slides stored a long time) should be tested. LOWERING BACKGROUND: In a study of hybridization on decalcified tissue, we have shown that immersion of mounted sections in 10% EDTA +4% paraformaldehyde at 4 degrees C. reduces background and actually improves signal to noise ratio for the low abundance message we hybridized in 10um embryo sections. (Ryan, A.F., A.G.Watts, and D.M.Simmons. Preservation of mRNA during in situ hybridization in the cochlea. Hearing Research, 56 : 148-152, 1991). If this is used as a method of lowering background for sticky probes to low abundance message, overnight treatment would probably be adequate. Note that high background may result if sections are hybridized in an oven, rather than on a hotplate. In that case, reduce the hybridization temperature by about 5 degrees C. Also, using too small a volume of hybridization solution or pressing down on the cover glass to remove bubbles may cause cover glass-static that could increase background. MICRO-TRANSCRIPTION: We have reduced the amount of isotope routinely used to a more economical 100 - 150 uCi per transcription reaction. In general, the original amount of DNA can be halved, and concentration of cold NTPs and reaction time doubled. Promega asserts that the rate limiting NTP (35S-UTP) should be 12nM for full length transcripts, but the 6-8nM we use is adequate for some full length and much partial length transcripts. They are of course specific, and for ISH actually penetrate better than many large full length transcripts. Probes may vary somewhat in transcription efficiency, but the following guidelines are generally adequate. Use a reaction volume of 7.5 - 8ul -- only take care that the final concentration of transcription buffer is correct: a 5x stock buffer concentration needs to be diluted to 1x in the final volume. (eg. add to dried isotope:1.5ul 5x txn buffer, 1ul DTT, 1ul DEPC water, 1.5ul mixed NTPs (stock 10mM A+G+C in equal amounts), 1ul RNAsin, 1ul RNA polymerase, 0.5-1ul template DNA (1ug/ul concentration) --- incubate 2hours ). By adding 50ul of SET afterward, we obtain about 60ul of probe from the spin column. That 60ul is usually 1-3 million cpm per ul. We only dilute it further if cpms are very high, so as to increase accuracy of measurement when preparing the final hybridization solution. If you do not have facilities for drying down radioisotopes, increase the volumes of transcription buffer and NTPs to an appropriate amount to balance the volume of isotope. (eg. use a 20-25ul total reaction volume, with 3ul mixed NTPs, 1ul DTT and RNAsin, no added water, and 1-1.5ul eachof RNA polymerase and template DNA. Amountof 35S-UTP (8-12ul for 100-150mCi) will determine the total volume, and thus the amount of transcription buffer needed). 368 APPENDIX VI Histology Solutions and Procedures Fixation and Perfusion From ISH protocol paper: Simmons DM, Arriza JL, Swanson LW. 1989. A complete protocol for in situ hybridization of messenger RNAs in brain and other tissues with radiolabeled single stranded RNA probes. J Histotech 12:169-181. Fixation (1 to 3 hours): Fixative Preparation Prepare fixatives early enough to allow cooling to 4°C (may be prepared the afternoon before use and stored at 4°C). Paraformaldehyde is toxic; therefore, weigh powder and heat solutions in a fume hood. We use granular “prill” paraformaldehyde (#19202, Electron Microscopy Sciences, Ft Washington, PA). Amounts given below are for perfusion of one rat. The following formulas for 4% paraformaldehyde solutions at two different pH values are for optimal immunohistochemistry on brain. Immersion fixation of small pieces of tissue in other aldehyde solutions is adequate for in situ hybridization alone. Fixative Solution A: Fixative Solution B: 250 ml distilled water (dH2O) 500 ml dH2O 1 g NaOH 2 g NaOH 10 g paraformaldehyde 20 g paraformaldehyde 3.4 g sodium acetate (CH3COONa•3H2O) 19.07 g sodium tetraborate (borax) Heat and stir to dissolve; Heat and stir to dissolve; do not exceed 65°C. Cool in ice bath. do not exceed 65°C. Cool in ice bath. When fixative solutions have cooled to at least 10°C, adjust Solution A to pH 6.5 with 50% glacial acetic acid, and Solution B to pH 9.5 with 50% HCl. Just before use: Reserve 50 ml of Fix B (without glutaraldehyde) in a specimen jar for tissue storage and post-fixation. Add 10% sucrose (5 g cane sugar) as a cryoprotectant, to prevent ice crystal artifact when frozen sections are prepared later. Then, add 1 ml of 25% glutaraldehyde (electron microscopy grade) to the balance of Fix B, mix well, and use within one hour. Addition of glutaraldehyde is optimal for fluorescent immuno of small peptides and enzymes (TH, DβH. In the past we postfixed overnight in Fix B with sucrose that included glutaraldehyde to obtain optimal immunohistochemistry. However, glutaraldehyde in the postfixative causes a shell-like “edge-effect” of nonspecific probe binding with in situ hybridization, probably due to overfixation at the surface of the tissue block. Eliminating glutaraldehyde in the postfixative abolished that artifact, and has not significantly affected the quality of immunohistochemistry. Immunostaining is often diminished in tissue with longer postfixation times, and the schedule in this protocol is based on that constraint. However, brains postfixed for one week before sectioning have given excellent in situ hybridization results in cases where immunostaining was not needed. For some purposes, Fix B alone may prove adequate (P.E. Sawchenko, personal communication). 369 Perfusion Technique (30 to 60 minutes) 1. Anesthetize the animal by intraperitoneal injection with general anesthesia of choice (e.g., nembutol); use slightly more than the usual surgical dose, for very deep anesthesia. Fill perfusion pump tubing with saline, taking care to eliminate all air bubbles. 2. Open the chest cavity, expose the heart, and visualize the ascending aorta. Ensure an open working field by retracting the rib cage or cutting all ribs laterally. Clamp the descending aorta if tissue other than brain is not required. 3. Insert cannula (a blunted 13 gauge hypodermic needle) into the base of the left ventricle, up through the left atrium, and into the ascending aorta. Clamp it in place in the aorta, and cut the right atrium. The right atrium is on your left, as you look into the ventrally-exposed chest cavity. The left ventricle is on your right and is usually easily identifiable because it contains bright red oxygenated blood and is lighter in color. 4. Begin the perfusion with physiologic saline (0.9% NaCl) at room temperature as soon as the atrium is cut. Perfuse 20-50 ml of saline, or enough to assure that the system is patent (make certain there are no air bubbles in tubing) and solution is flowing smoothly at about 40 ml/min. 5. Stop the pump and transfer intake tubing to ice-cold Fix A, being careful to avoid introduction of air bubbles. Re-start pump and watch for muscle contraction, which indicates fixation has begun. Pack the animal in crushed ice, reduce the pump speed to 20 ml/min, and perfuse for 10 minutes. Stop the pump, change to ice-cold Fix B, and continue perfusing for 15 min. Remove brain or other tissue and store in postfixative (with sucrose) overnight at 4°C. 6. We use a Cole-Parmer (Model #7520-25) Masterflex variable speed pump for perfusion. This allows precise control of fixative flow rate (20ml/min), and initial perfusion (50ml) of fixative at a slightly higher rate. Gravity perfusion is adequate, if a flow rate of 20 ml/min can be obtained. 7. If it is not convenient to cut sections the next day, freeze the tissue en bloc after postfixation, and store (well sealed to prevent dessication) at -30°C to - 80°C. 370 Cryoprotectant Tissue Storage Solution (“Anti-Freeze”) This is a slight modification of tissue collecting solution for enzyme histochemistry of axonally transported horseradish peroxidase. The original formulation used cane sugar in place of the ethylene glycol. Its cheaper but can attract insects [!]. This formula is better for both immuno and in situ procedures – thus its, our lab standard. Store sections in 24-well tissue culture dishes – stackable plastic trays with close fitting lids (~$1ea, purchased sterile; wash them for re-use). 50% 0.05M sodium phosphate buffer, pH 7.3 (see below). 30% ethylene glycol 20% glycerol It is convenient to make this solution in a plastic liter bottle, pre-marked for volume. Mix well and store in the refrigerator or freezer until used. Use reagent grade chemicals. HPLC or MolBio grades are way-expensive and not necessary for use with our fixed tissues. Sections are stored in this cryoprotectant preservative solution at -20°C (regular freezer). They retain antigenicity for months to years. They can be frozen solid at -70°C, or with dry ice, for shipping with no adverse effects. After storage, pass sections through two or three brief washes of buffered saline before immuno- staining or mounting. Some people use this as a cryoprotectant for pieces of tissue stored at ultralow temperatures. That tissue requires sectioning at extremely low temperatures (-30 or so, lower than most cryostats), else it’ll be too ‘soft’ to obtain good sections—part of the anti-freeze effect. Buffer Stock Solutions: 0.2M sodium phosphate A) 27.6 g NaH2PO4 • H2O/liter (monobasic) B) 28.6 g anhydrous Na2HPO4/liter (dibasic) (always make buffer stock solutions in volumetric flasks) Working 0.05M phosphate buffer @ pH = 7.3 for 2 liters: 115 ml Stock A. 385 ml Stock B. 1500 ml d.H2O Immunohistochemistry buffer: Phosphate buffered saline at physiologic pH - equivalent to PBS, with potassium salts. (best for use with pH shift perfusion and fluorescent immuno for small peptide molecules.) Can make one liter at a time, but better to make large amounts, to always have more than an adequate supply. 0.12M KPBS Stock Solution (6X working concentration) 2.96g KH2PO4 Potassium phosphate monobasic 17.2g K2HPO4 Potassium phosphate dibasic 54g NaCl 980 ml dH2O (Measure very carefully! Amounts are critical.) working concentration 0.02M KPBS: mix 150 ml stock with 750 ml dH2O = 900ml total. 371 Immunohistochemistry Procedure Perfuse and store brains as above. Over-fixation decreases immuno staining! Brains can be stored for 1-2 days maximum in buffer+sucrose (not fixative!) before sectioning. Frozen sections cut on a sliding microtome (15-30µm). All immunohistochemistry is performed on free-floating sections, usually in net-bottom wells with gentle agitation. Multi-well tissue culture plates work well for section storage and immuno-processing if nets are not available. Store cut sections in buffer at 4deg.C for immediate processing, or in cryoprotectant solution (above) in regular freezer (-20deg.C, remains liquid) or ultra-low freezer at (-80deg.C, freezes solid) for indefinite storage. Fluorescence antibody staining Primary antibody-cocktail incubation (see Appendix IV for antibodies and concentrations): Rinse stored sections in 2 x 10min in 0.02M KPBS—standard antibody buffer (see above) Pretreat: Incubate for one hour only in KBPS+ 0.05-0.1%TritonX-100 (Sigma™) Very dilute, mild detergent, to slightly permeabilize the cells for antibody penetration. Must be minimized (or eliminated, if possible) for fluorescent cocktail staining. Rinse well afterward in plain KPBS: 3-4 x 5 min Make primary antibody dilutions in cold KPBS with .02% sodium azide added, so it can be stored,refrigerated, and reused, if testing has shown a high enough titer for recycling {**}. Use appropriate dilutions (per previous tests, see Appendix IV): rabbit antibody to first peptide target, plus mouse monoclonal (mAb) to second peptide target (or other combinations of two antisera grown in different species) Dilute primary antibodies in 0.02 M KPBS buffer without Triton. 1% normal serum (e.g., goat) can be added to lower non-specific staining. (If needed, use serum from the source-species of the planned secondary antibody.) Primary antibody: Incubate, refrigerated, for 48-72h with constant gentle agitation Rinse well: 3-4 x 5min in KPBS. Secondary antibody: Incubate 45-60 min at room temperature secondary antibody cocktail at appropriate dilutions: Goat anti-rabbit IgG conjugated to FITC Goat anti-mouse IgG conjugated to RITC(Rhodamine) or other Red fluorchrome (secondary antibodies to the two source-species of the primary antibodies.) Both should preferably be grown in the same species for minimum background. Can use different species for the two secondaries if necessary (e.g., goat anti rabbit plus guinea pig anti mouse), test for cross reactivity beforehand. Make secondary antibody solutions in KPBS + 1% serum of the secondary animal. 372 Rinse 2x10min in KPBS. Mount from fresh KPBS onto chrom-alum coated or Plus slides and let air dry about 10 min. Coverslip with buffered-glycerol mountant (pH 8.4) [50:50 glyceerol and 0.2M NaPhosphate buffer] and observe with fluorescence microscope. View Fast Blue with ultraviolet excitation, FITC and Rhodamne with their appropriate excitation filters. Best to view asap. Store slides in flat trays in the freezer, as fluorescence fades with light, heat and exposure to the fluorescent beam of the microscope. Photograph asap for permanent record. {**} Immuno sections treated with 0.3% Triton give very poor subsequent Nissl staining. Recently we have seen 15micron frozen sections stained as above with primary antibodies recycled yield a fantastic Nissl stain after 2yr storage in the freezer (coverslipped in glycerol mountant) at -20 degrees C! Thionin (Nissl) Staining Procedure From a chapter in Elsevier’s Brain Research Protocols (now called Neuroscience Protocols) Simmons DM, Swanson LW. 1993. The Nissl stain. Neurosci Protocols 050-12-01-07. Fixed (frozen) sections should be mounted on subbed slides and well dried. air dry @ room temp(RT), and overnight @ 37-60 deg C; or several days @ RT. overnight in 70% alc before beginning stain procedure is often very helpfull --- sections adhere better (alcohol crosslinks the tissue/gelatin subbing bond) and the background is generally more uniform and free of non-specific color. If original fixation was poor, adding 10% formalin to this alcohol may be helpful (Use bottled formaldehyde concentrate from Labstore. Our usual buffered paraformaldehyde fixative might precipitate in the alcohol solution.) 95% alc -- 2 x 3 min 0.25%Thionin stain -- 5 to 20 seconds 100% alc -- 2 x 3 min fresh dist. or DI water -- 5 to 10 dips Xylene -- 2 changes, 30 min total 50% alc -- 2 min delipidizing step: 70% -- 2 min solvent substitute OK 95% -- 2 x 2 min 100% alc -- 2 x 2 min check staining; 95% alc -- 2 min use acetic alcohol, if needed--see note 70% alc -- 2 min 100% alc --2 changes, 3 -5 min each 50% alc -- 2 min Clearing agent (eg xylene) dist water -- 2 min 3 changes, 3 - 5 min each (Be sure the 2nd one is fresh) last one must be fresh Then, coverslip from last change. NOTE: If stain is still very dark after 95%alc; accelerate differentiation with 95% alcohol containing 1% acetic acid -- time will vary from a few dips to one minute. If over-differentiation results, the sections can be restained. Some labs use 1% Thionin for a one-dip, super-fast stain. However, we’ve found it doesn’t penetrate thick sections (>15µm) well enough for an even stain. Slides can be left several hours to overnight in a fresh change of clearing agent. 373 NOTE: Xylene or toluene are common organic-solvent clearing agents that have the high refractive index needed for microscopic preparations. However, they are quite toxic, and substitutes such as Hemo-D (limonene or citrus oil) have been introduced in recent years. Hemo-D has virtually no tolerence for water contamination, and takes at least twice as long as xylene to clear adequately. The 100% alcohol preceding Hemo-D must be very fresh -- If there is significant water contamination, a milky precipitate will be seen when passing the slides into Hemo-D. The problem is obvious, and you must return them to an anhydrous alcohol before trying Hemo-D clearing again. A less obvious problem is with very slight water contamination: the slides may pass through Hemo-D and look fine to the naked eye. However, under the microscope, they have a slightly “murky” appearance. Too short a time in Hemo-D may produce a similar effect. Therefore, clear at least 30 min total when using Hemo-D. The advantage of Hemo-D is its low toxicity, so that it is not essential to use a fume hood. A good compromise is 2 changes of xylene (in the hood) after 100% alc, followed by 3 changes of Hemo-D before cover slipping. Thionin Stain Formula Simmons; 6/91 Stock concentrates for buffer Solution A: 5% NaOH (w/v) -- 5g in 100ml DI water Solution B: 6% Glacial acetic acid (v/v) -- 6ml + 94ml DI water (100ml total) --------------------------------- Working buffer for dissolving thionin stain powder: 18 ml Sol’n A + 100 ml Sol’n B + 382 ml DI water (yields 500 ml total => adjust pH to 4.5) Heat the above buffer to steaming: USE a full size, ONE LITER ehrlenmyer flask! Mark 500ml accurately on the outside before heating Add 5g Thionin - for 1% (or 1.25g Thionin - for 0.25%*) Careful, this is unbelievably messy if it spatters (as from an open beaker) or spills! Bring solution to a boil, and continue to boil for one hour. Let cool, and return volume to 500 ml with distilled water. Store in a dark bottle, and filter before use. Can be used for several months. *0.25% is preferred for slightly slower, but more uniform stain. 1% tends to precipitate. 374 APPENDIX VII Fast Blue injection procedure Surgery protocols: 6/92 Swanson Lab Fast Blue injection into Jugular vein: (for retrograde labeling of neuroendocrine neurons) Dissolve Fast Blue (FB) at 2.5 mg/mL in sterile DI water (not soluble in saline or buffer). Vortex well and centrifuge briefly if using right away (so no precipitate is accidentally injected into the animal’s bloodstream). Store in fridge - solution remains good a very long time (at least a few months). FB is expensive -- so weigh carefully, empty into a 30mL sterile tube and then add the appropriate amount of water to make 2.5 mg/mL. Withdraw needed volumes with a long sterile needle on a tuberculin syringe, then change to a half inch 27 ga needle for the actual injection at room temperature. For 250-300 g rat, inject 0.7-0.8 mL, four to seven days before pH shift - perfusion. (14 days is very bright, good for ISH - but results in pericyte staining in blood vessels) ________________________________ Halothane (Metafane) anesthesia, animal on its back. Attach (tape) forelimbs out at right angle to trunk, but not so tightly that there is tension across the top of the chest. Incise skin from the level of the forelimbs about 2-3 cm toward the head, midway between sternum and shoulder. Pull skin up with rat-tooth forceps, and use large scissors. Pick up sub-dermal fat and connective tissue, and use sharp scissors to blunt dissect and expose underlying tissue compartment. Visualize flat strap-like muscle band at level of forelimbs with distended dark red jugular just above it. The extended limbs cause that muscle to put just enough pressure on the jugular to have a slight tourniquet effect. If they are taped down too tightly, the vessel is more likely to collapse when injecting. NB: Injection must be done quite slowly, as osmotic shock of injecting a large bolus of DI water can cause heartbeat and breathing to stop. Lift strap-like muscle with fine forceps and insert injection needle (bevel up) through muscle and up into the jugular. Going through the muscle provides stability and helps to hold the needle in place while injecting. Hold the syringe parallel to the table and stabilize it by resting your knuckles or fingertips there, as the injection will take several minutes. Use your other hand to move the plunger while watching that the needle doesn’t move. Draw back to confirm placement (see blood enter at hub of needle) and inject about 0.03 mL, wait for about 20 respirations or 10 sec and inject another small bolus. Continue at this rate or slower. Monitor breathing and needle tip placement. Vein may collapse (looks like it disappears) while you are injecting. Injection will continue OK if there is no perforation. If you see yellow fluid at the injection site, you are out of the vein. You may be able to re-enter it (be sure to pull back and see blood before trying to inject again), but unlikely. Better to try opening and injecting the other side -- especially if only half or less of the full amount was successfully injected. Close with wound clips. Animal recovers from gas anesthesia within a minute or two. 375 Eleni Markakis: 1994 -- Addendum to Jugular Vein Injection of Fast Blue I have found that animal survival of this procedure is increased when the solution being injected is iso-molar and of neutral pH, and warm. To that end I have developed a method which incorporates these principles. Since Fast Blue does not dissolve in saline, and since the dye is acidic when made in water, I now dissolve the dye at 5 mg/ml in sterile water (twice our normal concentration) and make a solution of 1.8% saline in .1M TRIS buffer (twice normal osmolarity). I keep these solutions warm until the time of injection, and then fill a 1cc syringe with .35cc of each solution, bringing the total injection volume to .7cc for a 250g rat (less for smaller animals), and making the resultant solution 2.5mg/ml Fast Blue, .9% saline (normal), and with a neutral pH of 7. I slowly inject this warm solution into the exposed jugular vein. Simmons annotation: The above solution mixture apparently precipitates at least to some extent upon mixing, especially if not used quickly enough (per Markakis). FB does precipitate in phosphate buffer. I haven’t tried it in Tris buffer, or with heating. The amount of neTRH staining in parvicellular PVH reported by Markakis was apparently less than that seen by Simmons when using the same rimary antibody—unlike other antisera used. Perhaps this is due to changes in IHC procedure used to label BrdU (acid pre- treatment). Or, could it be that fewer TRH cells were labeled as neuroendocrine with the above saline/Tris FB solution? Markakis quantified ne cell labeling as ~98% in SON. However, there could be a difference in ne parvicellular regions… It is possible that the extent of labeling may be less with this modified method. If used in experiments where quantitative cell counts of neuroendocrine cell types are crucial, preliminary ad hoc side-by-side comparison tests should probably be made with samples using the original method. 376 APPENDIX X Comprehensive Mapping procedure PHASE I, from primary tissue images: 35mm projection maps 1) Photo sections @standard magnification. Also photo the slide micrometer, for later determination of a 100mm scale bar in the maps. 10x fl. Nikon/10x Leica eyepiece (montage, for all of PVH) standard exposure time for each filter on a given IHC set -->important, to record accurate stain intensity [determine time by quick survey of lightest & darkest] 2) Project 35mm’s onto tabletop to map onto 11x17 ledger size paper: use prism/mirror on slide projector, [~34in from table to lens]. --DO NOT MOVE PROJECTOR during drawing [to keep scale the same] easiest: start from classic level and work forward and back [+1, -1] trace ventricle and major vessels for fiducial alignment trace outline of blue labeled cells [~PVH, w/o autonomics] map w/ + and o, for 2 peptides and ne or non-ne designation (use color codes to help in mapping) recheck each mapped cell w/ scope after map is finished [most efficient way -- to minimize fading] make an arbitrary fiducial ‘cross’ or ‘X’ to align succeeding maps Draw 100mm scale bar on map, from 35mm of slide micrometer PHASE II, from 35mm projection maps: separated layer maps 1) Do ‘layer separation maps’ by tracing from originals, on light box because original maps are too confusing to read as placed art in a.i. results in 7 (superimposable) map sheets drawn in black pen: *outline borders of 3v and blue labeled cells *ne-1st peptide *ne-2nd peptide *ne-[1st+2nd] peptides (triples) *non-ne-1st peptide *non-ne-2nd peptide *non-ne-[1st+2nd] peptides (doubles) 2) Calculate % reduction needed to convert to 8.5x11 copy -to fit into b&w flatbed scanner -Mita copier converts, if standard paper sizes are used [~67%] make reduced size images of all the separation maps 3) Scan [b&w flatbed] into Photoshop and save as PICT format files PHASE III, from separated layer maps: Illustrator data files 1)PICT scan as ‘placed art’ [dimmed image] to trace data into Illustrator make 7 different layers in Illustrator for each section: *1ea: 3v w/ solid line, and ne-blue border w/ dotted line *2ea: ne=plain filled color, Zapf Dingbat letter (‘l’==a circle ~~same size as parvis) color coded to peptides: CRH=sky blue, OXY=red, VAS=green, TRH=yellow, SS=grey, TH=pink *2ea: non-ne=filled color w/ black outline *2ea: ne-triples, and non-ne-doubles=combo[2-overlapped] 377 2) Use copy tool to place symbols over all drawn cells on layer separation Final check with actual separated maps and original map art (the 1st maps were done in individual a.i. Files, w/ few layers) Illustrator 7 or 8 allows direct mapping of many sections/layers 3) Use select all, copy, paste in front, and rotate to align a.i. layers from successive sections directly above each other in the same file These are the “stacked files” [from the same A-D set in 1 rat] In Illustrator 7 or 8 these files are composite from the beginning. Re-orientation may be needed after data entry. 4) Try to merge data by aligning sections at the classic level from 2 rats these are “combined level-maps” from different PVH expt’s make printouts for the wall with the merged data from PVH 10 & 12 (may not do in future, as the sag projection shows better registration) PHASE IV, making Nissl drawings 1) photograph all the Nissl sections (this could have been done earlier) since the Nissl sections are shrunken compared to IHC, anyway... [?] Use 6.3x objective, not 10x; to get all PVH in one field 2) scan 35mm Nissls as b&w images into Photoshop 3) adjust levels and contrast to get best image in photoshop careful not to unsharpen so much that cells turn into pixels 4) draw contours and subdivisions of PVH in photoshop layer Nissl image as background [drawing pen is pretty crude in PS] microscope next to computer screen to confirm drawing be sure to draw a vertical line for fiducial midline: ~center of 3v this may have to be estimated in some photos important - for correct orientation in space, later 5) export layer to illustrator file format PHASE V, putting Nissl drawings [maps] in order Va: stack nissl drawings: 1) Stack exported a.i. Nissl drawings in order in one file Add corresponding ‘PVH borders’ layers from data files, &lock. Set B &/or D are best, since they’re serial to Nissls[A] Use rotate and scale to fit (classic) Nissl outline to data outlines [because nissl sections are shrunken compared to IHC] Find %scale, by calculating measures [e.g. 3v to blood vessels] Do this at 2-3 levels, to check that your scale factor is consistent. This assumes all the Nissl sections have shrunk ~ the same Use same scale factor [~190%+] for all the a.i. Nissl drawings 2) Draw better outlines of 3v and subdivisions w/ illustrator pen tool This is for graphical improvement of the crude PS pen tool lines 3v as black line, PVH contours, subdivisions as dotted lines Use microscope for confirmation of outlines 378 Vb: adjust Nissls to true vertical: 1) Correct vertical orientation of Nissl drawings ->->first pass: Turn off the data layers, and start with classic Nissl *place classic section at about the middle of the artwork page Add and adjust Nissl images rostral and caudal, as w/ data 2) Use midline vertical lines to re-position for vertical orientation some data files are skewed because only part of 3v was available in the 35mm fluorescence photo [montage]. So, need ‘straight’ Nissls. 3) Remember, PVH slopes down in brain from rostral->caudal use slope as a mental guide for height of Nissl outlines [exact height will be determined from fiducial orientation] PHASE VI: determining orientation of Experimental Nissls to Atlas VIa: find fiducial structures in data set. [Fids used should be as far apart as tissue borders allow.] 1) now [if not before] determine Nissl #s for all fiducial landmarks e.g., ac, SCH, and AHNd as a minimum for the sagittal plane need 2 widest-spaced (in your data) fids for the horizontal plane look to see what you have, esp. if your block was trimmed wider spacing gives best triangulation with midline rostral fid 2) copy vertical parallel lines [~~serial frozen sections] from Atlas or, make a set, group it, and meas. @6400x to check if parallel make lines thicker by changing the point size of the stroke set width so that lines barely touch, to get even spacing “human eye is sometimes best micrometer” work in line art mode, not view image, if fat lines bother or, make lines thinner after confirming they’re parallel these A-D lines will be crucial for projection data entry 3) Need a parallel set of ~120 lines = average #PVH expt’l sections: make each 4th line (set A [Nissl]) different, e.g., 50% gray 4) Mark lines of Nissls with fiducial structures: [e.g., color hot pink] Double check this, with scope and by counting lines! Error here is semi-fatal for correct data registration! 5) make two copies of this set, for sagittal and horizontal projections VIb: determine orientation of data set to atlas. Set d/v tilt using fids in similar d/v plane [e.g., ac and SCH] Set line spacing by scaling between rostral and caudal fids. [Fids used should be as far apart as tissue borders allow.] 1) Position data line set over atlas sagittal projection vignette include red fid marks when copying vignette from Atlas file. 2) Next will be a series of rotate, fit, and scale maneuvers: rotate correct section lines onto a.c. and SCH fid marks (d/v tilt) -likely to be off the vertical [atlas level], at least a little select scale tool, and choose a.c. fid mark as the anchor point then, scale/drag laterally, so AHNd section# is on fid mark best to click-shift, so that scale/drag is true horizontal -to avoid accidentally introducing skew into the line set this determines the spacing between lines in the serial set. 379 3) Final = several small approximations: repeats of above procedure... After AHNd fid is scaled to ac fid, fit to SCH fid can move off center just re-rotate for the a.c. & SCH fitting, as above then, re-scale for the AHNd fid, as above This is much easier that trying to calculate a standard %scale one could calculate or measure the %scale afterward... 4) Even up the tops & bottoms of the “data set orientation lines”, which were originally a rectangular group, that has now been rotated off its vertical/horizontal axis. The set should extend beyond the PVH projection image from the atlas. 5) With single select, and lasso, choose all top endpoints. Select average, then horizontal: all points move to the same height. 6) Lengthen lines to cover sagittal projection image, If needed When lengthening grouped lines that are not true vertical/horizontal: Be very careful not to change the fiducial angle you’ve set! 7) Copy and paste a new set over the original, and lock it. Adjust the length of endpoints-with locked copy as a template. When the length is perfect, hide the group of lines. Unlock and delete the copy, then unhide [show all] and lock the new version. Repeat #5-7 for the bottom line-points **result should be a perfect parallelogram! 8) Repeat procedure #1-3 for the horizontal projection alignment: Have ac section cross the black line representing midline (can use SCH as ~~midline, if ac not present in expt’l sections) Orient marked sections with their lateral fid marks [left & right] Select right lat fid section as origin point for scale tool (atlas is defined to be right side, PVH data side is also right) drag the ac section into register with the [midline] fid mark Do incremental adjustments, as above to perfectly fit all fids Align endpoints to perfect parallelogram, as in #5-7, above. The above procedure sets the true mm line-spacing for the experimental set... [the graphical representation for distance between sections] with reference to the sagittal & dorsal projections of PVH in the Atlas. [differs slightly between expt’l brains, because of differing planes of section] Note: the distance between section lines determined from these two projection processes should be equivalent in the same experimental brain [representing true anatomical proportions]. The angle of the grouped line sets may, however, look quite different. This will be more pronounced if the experimental set was cut in a plane very different from the Atlas. PHASE VII: positioning Experimental Nissl data to Atlas orientation The goal here is to position Nissl drawings in d/v space, to correspond to the true position of PVH as determined in the Atlas. Since the data is recorded much larger than the Atlas PVH, a blowup of vignettes for the 2 projections should be made to correspond to the size of the PVH data. 380 VIIa: make large projection vignette from atlas PVH projection. 1) Copy Atlas sag hypothalamic vignette + Data orientation lines. Magnify uniformly, so projection PVH and Nissl data PVH are similar size. Position it off artboard, left of Nissl data images. 2) Make horizontal lines @ top & bottom of PVH Nissl image extend ends to left so they’ll overlap & intersect the vignette [use classic; or level that matches best to the frontal atlas] 3) Scale size of vignette: to height of classic level Nissl data image: goal: lines extending leftward from the Nissl outline should intersect the PVprojection image at the same points that the top and bottom intersects of the same number section lines do. measure Nissl height & sag proj height @ crossing of section# (measure vertical height, not the [slanted] length of line) calculate a % scale, and make the major enlargement (1st time may need a bit more, + or- 1% to 2% at a time) scale by %, to avoid skew if scaled by dragging between points. This procedure ‘scales’ the sagittal projection image vignette to all the Nissl data sections, assuming the classic [reference] section fits best. Total % enlargement from Atlas can be calculated post hoc. VIIb: Adjust height of Nissls using the angle on the Atlas projection. -best determination of d/v distance spacing of [1 in 4] Nissls Adjust each Nissl data layer, fore and aft of the classic level, in succession. Horizontal guides crossing the PVH projection image at equivalent section lines now DEFINE the height of the Nissl data. 1) Draw horizontal lines rightward from the top and bottom intersect points of the appropriate section line on the PVH projection and lock. 2) Select all in Nissl layer, and move image up or down to fit top and bottom of PVH on the Nissl drawing between lines from the PVH projection image. In some cases the Nissl section PVH outline may be slightly larger or smaller [depends on accuracy of reference section chosen, and individual tissue distortions]. In those cases, ‘average’ the difference and use the best fit for the d/v position of the Nissl. The goal is to eventually fit (stack) all the data sections [A-D] in order, in perfect orientation, as they would have been in the uncut tissue [~~3D...] 3) The m/l placement can be done by eye, using 3v, etc. AT THIS POINT, RECHECK THAT MIDLINE FIDUCIAL IS VERTICAL! PHASE VIII, fitting data layers to oriented Nissl maps B->D data sections will necessarily be interpolated perfectly between Nissls in the completely reconstructed, Atlas registered, oriented stack.. 1) On one completed set [B-D] rostral to caudal [if there is one finished] copy and Interleave Nissl borders [set A] to correct section levels. -Nissls have the correct d/v &m/l spacing, according to the atlas 381 2) Orient data from individual sections to fit between PVH outlines on the 2 closest Nissls. Direction of position change is known from Nissls. 3) Fit 3v and top and bottom PVH borders [ne outline] to the closest Nissl outline. B will fit best to same section# A, and D will fit best to following section# A, then C will interpolate between B and D. 4) Above are details of my convention for PVH12, largely because the data layers were first done before the sagittal projection line orientations were prepared... Thus, they needed re-rotation to fit the now accurately measured position of PVH in relation to the Atlas. Then, set C position was interpolated between B and D. Data were usually oriented to the Nissls one data set [B, C, or D] at a time, thus the same measurements for PVH height and width (see below) were used on the Nissl set and the data set serial to it. Then the middle set [C] was interpolated between them for final orientation. In future, If all data layers are oriented to the Nissl set simultaneously, it will be easier to interpolate positions [3v, cells, ne borders height and width] in serial section order. You can then maximize data manipulations [all layers from one section] and minimize measuring common distances and re-calculating proportions later. In fact, future experimental Nissl sets will likely be done before any of the data sets. Then, all the subsequent steps in the analysis procedures will be keyed to the Nissl positioning. PHASE IX, how to plot data onto PVH projection images Remembering how the PVH 3D projections from the atlas were made: 1) on each level [serial section] showing PVH in the atlas, construct a bounding box that just touches the edge of the PVH on top&bottom, and medial&lateral. Also, note fiducial: base of brain and midline of 3v. 2) on the sagittal view: mark the points where the top & bottom bounds lie on the appropriate Atlas serial section level, when plotted using the base of the brain fiducial as a reference point. 3) when all sections are thus marked, a perimeter of the outline of the PVH in the respective planes will be defined by joining the points with a smooth curved pen line. Fill the outline in a contrasting color. 4) repeat for the horizontal projection: but, rotate the bounding box 90degrees and use the midline fiducial and the m/l distance along the Atlas serial section lines. How to see data we have in the PVH projection planes? 1) Looking on edge (like a loaf of bread) from medial to lateral in the sagittal projection, a given neuropeptide data set [1-in-4 sections] would look like an interrupted set of colored lines at the appropriate height on the PVH projection image. 2) What if we could look a bit obliquely, to get an idea of what the ‘filled in’ pattern of data points would look like? Would this approximate a 3D image, or at least allow over- all viewing of data in the 2D maps? It would only work if the colored data symbols were transparent; and determining the correct oblique angle of rotation might be difficult. At this time, layers in Illustrator cannot be rotated as a group to give the illusion of a 3D image, and the data circles cannot be individually rotated into spheres to show their ‘thickness in a section. It seems useless to try to look at data we have in an orthogonal [90degree] plane. What can we do with tools at hand in Illustrator to approximate this goal? 382 3) One way to approach this might be to make a bounding box, as with the atlas sections, and collect all the data points from one section into an area [60mm] that would be occupied by it and the 3 intervening sections as projected onto the (sagittal) PVH outline image. It would be important to keep the d/v position of the data points constant, and randomly distribute them horizontally - over the area that would be occupied by all 4 sections in the set. Random slight overlap of data ‘circles’ onto the edges of the 60mm box would ensure smooth merging of the data from one set to another in the projection image. a) the center [line] of this grouped data would then be fitted to the line representing its parent section on the data orientation line set overlying the atlas PVH projection. The bounding box height would determine the point at which the center line intersected the PVH outline (at the same intersect as its line in the data set line series). b) data thus averaged from successive sections would be slightly offset one from another, and calculation of PVH height and width would accurately reflect both the primary data set and the projection image [model]. 4) this would result in a true data density of 25%, but distributed accurately in the d/v [sagittal projection] or m/l [horizontal projection] coordinate plane. 5) As it turns out, in all but the most densely labeled areas of PVH, when data points are distributed this way, there is not a significant [obscuring] amount of overlap. This means that at 25% density, you get a meaningful visual representation of the distribution of that cell type throughout the entire rostro-caudal [sagittal projection] or medio-lateral [dorsal projection] extent of PVH. 6) It should be understood that if all cells from each section were plotted in this way [100% density] much of the apparent Density [represented by lateral cells that would be ‘behind’ the more medial visible cells] would be effectively lost, or at least not viewable using opaque circles for data. PHASE IX, preparing data for PVH projection display First, decide what size images to work with: Data layers are very large compared to the Atlas PVH projection images. (We have a magnified Atlas hypothalamus vignette used to position the [A-D] data layers). However, its probably more efficient to reposition the data and reduce it by a measured % and fit to the true Atlas PVH image size--all in one procedure. Smaller [atlas size] images are also convenient for later display on a single page, to aid in visual comparisons of multiple data sets. 1) Set up a text area on the classic data layer to record section number, cell count, Nissl registration number, measurements [see below], calculations, etc. This will be copied and modified for each subsequent layer. It assures accurate records of all numeric data, calculations, %reductions, etc. used when averaging and transferring data for display on an Atlas sized PVH projection. 2) On the classic level data layer: make, group, and lock a bounding box that defines the top, bottom, lateral and medial extent of PVH [medial is at the 3v intersect, usually dorsal]. This will be fairly precise on the Nissl, but will necessarily be interpolated from mapped ne-borders on data sections. 383 On positioning the PVH data layer bounding box: The classic [or other] section is defined as the best match to the Atlas PVH, and subsequent analysis is based on that relationship. Therefore, the proportions of height and width calculated from direct measurements of section line length on the PVH projection images, and segment lengths of the data bounding box in the data layer should be about perfect. This yields a near equal %reduction [for Height & Width] used later for averaged data. Other data layer bounding boxes will likely yield slightly different %reductions in height and width. This is a result of distortion in the tissue, which will be reflected in the apparent PVH borders, and thus in the size of the bounding box. If the %reduction calculated for height and width are too different, the symbols of subsequently averaged and reduced data will be altered from their original circles into elipses. This results in an apparent [d/v] change in position of the original data in relation to the Atlas PVH reference standard. An attempt to minimize this data alteration is worthwhile. The lateral PVH margin is more likely to vary [especially in sections far rostral or caudal from the classic level], because of colchicine treatment and subsequent 3V expansion. The d/v measurement is less likely to vary in proportion to the Atlas PVH because of the taken in positioning data layers vertically in relation to each other and to the Atlas PVH. Use the following general guide (along with visual evidence from the primary data map) in determining the lateral extent of PVH in the data layers [i.e., length of the horizontal arms of the bounding box] Calculate the height/width proportion of the equivalent section lines over the PVH projection image. Multiply that fraction by the height of the PVH bounding box. Result approximates the length of the horizontal legs of the bounding box [the width of PVH in that section]. Use visual cues and measurements of calculated PVH width to set the final position of the lateral leg of the bounding box. Notice relationship of lateral bounds on nearby sections. 3) Using the 100µm scale bar measurement, construct a rectangular box equivalent to 60µm wide and the exact height of the PVH d/v boundary. Easiest to make this box in red, and have its center point on the medial bounding line. Add a vertical center line slightly longer than the PVH height, group it with the red box and lock. This red box will be copied to successive layers, and the height re-sized to fit individual bounding boxes. 384 4) On the Atlas PVH projection image, measure the distance [at right angles] between four data section lines. This will be a fixed number, representing the distance occupied by four sections in a set [A-D]. It will be used to determine the %reduction in width to apply to the 60µm wide data box when scaling it down to display averaged data onto an Atlas sized PVH projection. Measure the distance between the top and bottom intersects on the line representing the data image layer on the PVH projection. This will vary with each data section, and will be used to determine the %reduction in height when transferring averaged data to the Atlas sized PVH projection. Measure corresponding mm distances on the data layer bounding box and 60µm wide red box. -distance between four section lines on PVH projection divided by mm width of 60mm on the data layer = %width reduction of data to be transferred to Atlas size PVH projection. -Section line length on PVH projection divided by mm height of bounding box on data layer = %height reduction of data to be transferred to Atlas sized PVH projection. It is convenient at this point to make similar measurements for the horizontal projection, although data will not be plotted onto it until later. Ventricular swelling resulting from colchicine treatment introduced [predominantly] lateral distortion of PVH borders. At this time it cannot be easily compensated for, without altering spatial accuracy of the primary data. However, a visual approximation can be quickly made by averaging data along one line (see below). 5) Select all the data points [from a single data layer for that section] contained within PVH outlines and group. Count the number of cells [use selection info from file menu] and record in the text box and layer name. This object-counting feature is a major benefit of Illustrator: Until now, the data were spatially accurate, but didn’t include cell counts! Copy the data group, paste in front, and hide. 6) Group-select data symbols and move them to the left of the bounding box, while holding down the shift key to keep all the symbols at a constant [correct] horizontal level. 7) Using single-select and shift, move the data symbols individually to the right, to rest within the boundaries of the 60mm wide red box. The goal is to randomly distribute their final m/l positions within the box, such that each circle has an equal chance of resting on the area representing any one of the 4 sections in the 60mm wide set. On a random basis, let some symbols overlap the left and right edges of the red box by about 1/16 to 1/8 their diameter. (This is necessary for the data from adjacent sections to distribute evenly in the final projection image -- otherwise they will appear as discrete ‘stripes’ of dots oriented in the plane of section). An algorithm can be generated to do this truly randomly in the future. Does this mean that the automated distribution algorithm would have the red box defined as slightly wider that drawn? Perhaps an addendum would be added: “offset x% of cells at the edges by 1/16 diameter” or, by a mm distance? 385 8) Use show all to reveal the data group at its original orientation [they will all be selected]. Use command-key + j-key to see a window for join/average. Select average, horizontal, and copy. Result will be a straight line of compressed symbols that looks like a horizontal row of little bowties [each circular data symbol was a curve of 4 equidistant points]. Move this row of horizontally averaged data [use shift/select to maintain accurate m/l position] to a position just above the top horizontal bounding box line. 9) Now, [if not earlier] finish the calculations needed for %reduction of the averaged data to the Atlas sized PVH projection: This will result in a constant width % and a variable height % for the sagittal projection; and a constant height and variable width % for the horizontal projection[which must be rotated 90degrees before reducing]. 10) Unlock the red box, and the PVH bounding box and select it along with the enclosed data symbols. 11) Double click the scale tool and enter the horizontal[width] and vertical[height] % in the window and choose copy. Result will be a very small copy [~~5% of original] centered in the data window. Move it to the left of the data bounding box, so it is visible and not overlapping any data. 12) Again, select bounding boxes and averaged data, as in #10, above. Double click the rotate tool, choose 90degrees, and copy. Then, double click the scale box and enter the appropriate % numbers calculated for the horizontal projection, and enter. Move the resultant small image to the left of the data bounding box and below that for the sagittal projection. PHASE X, displaying data (averaged 1-in-4 sections) on PVH projection images 1) Now [if not earlier] copy the Atlas sized PVH projection image[s] and the grouped orientation line set[s] for the experimental data. Put them in a new layer, selected and locked individually, with the lines on top of the PVH projection outline. Label this layer to indicate the experiment, section set[A-D], and cell type of the data to be transferred onto that PVH projection image. e.g., “PVH12B neCRH sag proj” These projection images can be positioned at any convenient place on their pages [e.g., near the reduced data images to be transferred]. Their size and 2D orientation in relation to the atlas is important, but not their position on the page. In fact, their position will likelybe altered to create a montage page of different PVH projections to compare labeled cell types in the same and different animals. The averaged data from a layer can be displayed on the atlas sized projections at once, or you can wait to do several data layers at a time. 1) In a data layer, select and copy the reduced image [bounding boxes and data symbols] for the sagittal projection, and magnify to window size. Note: reduced data remains in the original data layer for reference. 2) Use select, and the rotate tool [NOT the scale tool!] to move the magnified red box with enclosed data symbols over the PVH projection image and data orientation line set. Its center line should superimpose on the orientation section line for that data set. The top and bottom of the red box [at its center line] should intersect the PVH borders at the same place as the orientation line for that section does, because %reduction was based on measurements of PVH height in the data layer and length of orientation line representing that section over the PVH projection. 386 3) If the match is off just a tiny amount, use the scale tool to shorten or lengthen the red box [and enclosed data] to fit exactly. This tiny fitting adjustment probably reflects inherent measurement error encountered when proportioning two very different- sized data set images. If it is off by more than just a tiny amount, there likely is an error somewhere in the %reduction procedure. Recheck all calculations, and the original bounding box size. Then, re-do the %reduction of averaged data, and try again. 4) When the averaged data is positioned accurately, de-select, click to choose the data bounding boxes [but not the data], and delete the boxes. 5) Select the [grouped] data (now oriented over the PVH image as it was in the experimental animal) and move it into the layer containing the PVH projection image. It should appear on top of the colored outline of PVH and the orientation line set. Note: the data will extend to the top and bottom of the PVH image outline only if there were cells at the most dorsal and ventral PVH in the original primary data maps. An approximate horizontal PVH projection can be obtained fairly quickly, without taking the time to individually average data circles inside a 60mm high box created in the lateral direction for the horizontal view. 1) Copy and position the horizontal reduced data [as in #1-5 above] over the horizontal PVH projection image and data orientation line set. Remember, the reduced data has been rotated 90degrees to agree with the position of the horizontal PVH projection. 2) In this case, the averaged ‘line’ of data [originally at top, and now at left] should be positioned directly over the data orientation line representing that section. Its [lateral] position on the line is determined by the length of the dorsal [now left] lateral bounding box segment. This procedure results in an interrupted [1-in-4] set of data- colored lines representing ‘collapsed data points’. However, it gives an accurate approximation of data positions over the PVH projection in the lateral orientation. Because of the accentuated curve in the 3V of the [colch] data set, the medial boundary will not be quite accurate, but it will be good enough to get a general impression of lateral distribution of cells. These lines of data should conform fairly well to the lateral ‘bulge’ of anterior magnos and two prominent lateral projections seen on the horizontal PVH projection. In fact, looking at the horizontal projection for PVH12 data helped determine inclusion inside the PVH data bounding box of lateral neCRH and neVAS cells that were not thought to be inside PVH when the primary maps were drawn! They comprise a group of neuroendocrine cells located in the darkly stained ‘eyebrow’ seen above the fornix in the Nissl for Atlas plate 26. These are distinct from the more caudal lateral group of supra-fornicial cells that are almost exclusively neurons with descending autonomic projections. 387 3) Transfer averaged, reduced data for an individual cell type from every level of PVH in an experimental set onto a PVH projection image layer, as above. Select all the [individually grouped] 60mm strips and include them in a new ‘super-group’ that defines the complete distribution of that cell type from that experimental set in the PVH projection orientation. The resulting ‘super-group’ will be the top or front level in the Illustrator layer containing the projection, e.g., “PVH12 neCRH sagittal proj”, with the PVH outline and data orientation line set locked below or behind it. The cell type data can now be selected and copied alone, to combine in register with other similar data sets over a new PVH projection. In this way composites of cell types from different brains can be accurately compared on the standard PVH projection. Also, the cell types can be ‘brought to the front’ in different order to make alternative contrasts and comparisons of data sets. APPENDIX IX Additional Experimental Data Original data maps not shown above in dissertation illustrations may be available by request. APPENDIX X Numeric Tables of all Cells Mapped Files containing tabulated raw data of cell counts may be available by request. 388 APPENDIX XI Background Discussion on TRH Simmons TRH essay: circa 2001 The tri-peptide Thyrotropin Releasing Hormone (TRH) was the first hypothalamic substance proven to stimulate hormone release from the anterior pituitary gland (Greer 1951) (Aizawa and Greer 1981) (Bowers, Redding et al. 1965). During the 1960s TRH was purified and characterized (Schally, Bowers et al. 1966) (Redding, Bowers et al. 1966), and its sequence was identified (Boler, Enzmann et al. 1969) (Burgus, Dunn et al. 1969) (Folkers, Enzmann et al. 1969) and active samples were synthesized (Enzmann, Boler et al. 1971). Thus Geoffrey Harris’ 1955 hypothesis (Harris 1955) of primary hypothalamic control over the endocrine system via the portal circulation was strongly supported, at least for the pituitary-thyroid axis. The discovery of TRH introduced a new era of thinking and vigorous investigation based on the concept of integration of endocrine and other physiological functions under direction of the central nervous system that continues to this day. The story of the race to purify and identify TRH and other putative “hypothalamic releasing factors” (now typically described as hormones) is one of the colorful and exciting chapters in the history of modern Neuroscience. Yet, perhaps due to the difficulty of obtaining a good antibody to such a small bioactive molecule (thyro-glutimyl-histidyl-proline amide), detailed data on the distribution of identified TRH-containing cells in the PVH accumulated surprisingly slowly. There were immunohistochemical studies describing TRH in the PVH and showing a figure or two, but for the most part they were descriptive surveys of TRH seen in various areas of hypothalamus. For example, Lechan & Jackson in a 1982 paper (Lechan and Jackson 1982) referred to a 1980 paper (Johansson and Hokfelt 1980) that studied TRH at the EM level as “one of the 'remarkably limited' IHC studies of TRH localization in the hypothalamus, although antibodies became available shortly [after] the original chemical characterization in 1969." Another reason for the slow appearance of immunohistochemical localization maps for TRH may have been the intense focus of both clinicians and researchers on the biological function of this exciting new molecule. In fact, most of the early publications regarding TRH were in regard to its experimental use as a physiologically active chemical rather than its cellular localization in the central nervous system [several refs here from NYAC annals volume on TRH]. 389 Early reports on anatomical localization of TRH in the brain indicated the presence of TRH-containing neurons in hypothalamic sites other than PVH as well as extra- hypothalamic sites. There did not seem to be a clear immediate focus on a brain area that might be a “neuroendocrine control center”. TRH cell descriptions early-on were vague, and didn’t seem to indicate an awareness of a localized source of neuroendocrine TRH, much less a reflection of the portal vessels as a possible mechanism for delivery of this powerful peptide directly to its anterior pituitary target cells. I find only four papers (Brownstein, Palkovits et al. 1976) (Nishiyama, Kawano et al. 1985) (Ceccatelli, Eriksson et al. 1989) (Kawano, Tosuru et al. 1991) that show schematic maps localizing TRH immunolabeled cells (the earliest of which was a punch sampling survey). Nishiyama’s paper gives two quite excellent surveys of TRH cells in PVH of hypothyroid, euthyroid, and colchicine treated rats. The second part of their study listed cell counts and maps (4 levels) of TRH cells in each subdivision [Swanson and Kuypers ‘80] of the nucleus. Ceccatelli surveyed TRH among numerous other PVH peptides, and did an elution/restaining study that located only one TRH/CRH/NT cell and no cells co-localizing TRH and CRH. The maps illustrated six levels, and there were two photomicrographs of TRH cells in the PVH. Kawano et al illustrated cells at 15 levels (spaced at 100um intervals) through the PVH in material co-stained for TRH and WGA-HRP transported from the median eminence. Though the actual number of cells illustrated was significantly less than other studies, the advantage was that they were identified hypophysial-projecting neurons. A recent paper (Markakis and Swanson 1997) shows mapped TRH cells in the PVH according to defined cell-birthdates determined by BrdU labeling. These maps have the advantage of being entered on standardized atlas plates [Swanson ‘89], and of showing TRH neurons that are identified as neuroendocrine by the presence of retrograde tracer form the blood. However, they represent only a subset of the total TRH neurons that can be stained in the PVH Early workers investigating the effects of TRH found an interesting, seemingly paradoxical situation, in that hypothyroidism [surgically or chemically induced] produced an increase in TRH secretion and resultant TSH production by the pituitary, while elevation of circulating thyroid hormones proved not to have a direct negative feedback effect on the secretion of TRH. Many experiments were done to elucidate the complicated relationship between TRH secretion and regulation of the pituitary-thyroid axis [several refs here]. Workers were hampered in their thinking about the system because early immunologic data on TRH localization showed it to be in several cell groups in the hypothalamus and even in 390 the brainstem, rather than just in what could be thought of as primary neuroendocrine cells affecting the pituitary. It was known from clinical data [human studies] that TRH was likely to be intimately involved with temperature regulation and general metabolic phenomena such as ingestive behavior. But no clear picture of the hypothalamic circuits involved emerged early on. While RIA assays were available soon after ‘TRF’ isolation from hypothalamic extracts, a major challenge seems to have been in creating a robust antibody to this common hormone that was nonetheless specific in its ability to immunohistochemically demonstrate TRH. The earliest reports of TRH localization in brain areas were from assays of dissected or punched samples of defined areas of fresh brain tissue. These studies, combined with the advent of better antisera, served to focus the search for the primary neuroendocrine TRH cells in the rostral hypothalamus, close by the third ventricle. By the 1980s It was known that the anterior preoptic, the paraventricular and dorsomedial nuclei of the hypothalamus had significant populations of TRH neurons. There seems to have been a rush of interest in localizing TRH-containing cells and fibers in the brain in the late 1980s - early 1990’s, as several well characterized antibodies to various and larger portions of the TRH prohormone became available (Winokur and Utiger 1974) (Brownstein, Palkovits et al. 1974). Using manipulations such as thyroidectomy and colchicine treatment (Nishiyama, Kawano et al. 1985), along with variations in fixation methodology, a consensus on TRH distribution in the central nervous system was reached by the early 90s. However, much work was left to be done in terms of identifying ne vs non-ne TRH neurons, and patterns of afferent and efferent distribution at both light and EM levels. The first molecular biological surveys of preproTRH mRNA and its deduced peptides (Lechan, Wu et al. 1986) (Segerson, Hoefler et al. 1987) brought a new and powerful technology to bear in the effort to define the central control of the pituitary-thyroid axis. In situ hybridization confirmed previous immunologic data on localization, and showed a difference in amount of preproTRH mRNA between anterior (less mRNA) and medial (more mRNA) parvicellular PVH neurons, implying a differential regulation within TRH cells in the nucleus. Lechan et al suggested that “The appearance of pro-TRH mRNA in neurons not previously recognized to contain TRH but which contain the prohormone suggests that non- TRH peptides within the TRH precursor may be preferentially expressed in certain regions of the brain.” Thus they confirmed TRH synthesis in some areas (e.g., olfactory cortex) that had been controversial when demonstrated with antibodies by some but not all studies. Later that year Segerson et al (Segerson, Kauer et al. 1987) showed by in situ hybridization and 391 immunohistochemistry that the parvicellular neurons of PVH were directly responsive to hypothyroidism - increasing both prohormone transcription and peptide synthesis in response to drops in circulating thyroid hormone. There followed further studies (Dyess, Segerson et al. 1988) using pellet implants and knife cuts to show the regulation of hypothyroid response was directly from TRH PVH neurons. Near the same time Hokfelt, et al (Tsuruo, Hokfelt et al. 1987) used a specialized fixation method to show a wider distribution of TRH cells by antibody staining than had been previously appreciated. Antibodies generated to unique deduced amino acid peptides from the cDNA sequence (Liao, Bulant et al. 1989) (Lechan and Segerson 1989) were tested for possible bioactivity and neuronal localization. These data agree quite well with in situ hybridization studies (Lechan and Segerson 1989). In addition, Merchanthaler et al (Merchenthaler, Meeker et al. 1989) using antisera from a tridecapeptide deduced from Lechan’s cDNA sequence showed staining different than TRH (though overlapping somewhat) and a western band larger that possible from proTRH. They postulated the existence of one or two additional precursors that can be processed to TRH in rat brain. More recently, (Lechan, Qi et al. 1994) TRH receptor isoforms have been identified, giving support to the theory of differential effects that may be attributed to alternatively-processed products of the prepro-TRH mRNA. There seems to be a discrepancy between ant and medial PVH parvis in terms of TRH content, in the extant literature. (Early thyroid depletion studies showed change in cell morphology w/ physiological manipulation - large nuclei, etc.) It is interesting to note that recent data (e.g., AGRP, NPY and CART input from ARC) are showing differences in synapses on the two subpopulations of PVH TRH neurons... Some general thoughts, and various notes on references from 2001: (Bowers, Lee et al. 1967) Endo.82, 75-82. “A study on the interaction of the thyrotropin-releasing factor and L-Triiodothyronine: effects of Purtomycin and Cycloheximide” --"concluded continuous protein synthesis is probably required for a maximum inhibitory action of T3 on the TRF-TSH release response and also that TRF-TSH release does not depend on formation of new protein." --example of early experiments using “TRF” to manipulate H-P-T axis, without yet localizing its production by specific hypothalamic neurons. Heavily based on clinical observations, as well as hypothesis driven experiments... This is from Shally’s group, which later published the structure of TRH. 392 (Kiss and Halasz 1990) “Ultrastructural analysis of innervation of TRH- immunoreactive neuronal elements located in the periventricular subdivision of the paraventricular nucleus of the rat hypothalamus” EM IHC and autoradiography: TRH IHC for cells and 3 H 5-HT uptake for serotonergic elements. TRH terminals were seen in synaptic contact with TRH dendrites and unlabeled dendritic branchlets. ... The findings substantiate the view that TRH neurons of the periventricular subdivision of the paraventricular nucleus may be influenced by TRH axons, serotonergic fibers and a large number of unidentified nerve terminals. (Taylor, Gyves et al. 1990) “Thyroid Hormone Regulation of TRH mRNA Levels in Rat Paraventricular Nucleus of the Hypothalamus Changes during Ontogeny” To determine the onset and extent of TRH response to low thyroid hormone levels during ontogeny, normal and [chemically] hypothyroid rats: e16, e20, P7, P21 and P56. Plasma hormones were assayed from pregnant mothers, pups (pooled) and adults. Levels of TRH mRNA were measured in the paraventricular nuclei (PVN) by ISH...’In summary, lesions in rat PVN prevented the full increase in plasma TSH, pituitary TSH beta mRNA, and alpha mRNA levels in response to hypothyroidism. Thus, factors in the PVN are important in thyroid hormone feedback regulation of both TSH synthesis and secretion.’ (Toni, Jackson et al. 1990) “Thyrotropin-releasing-hormone-immunoreactive innervation of thyrotropin-releasing-hormone-tuberoinfundibular neurons in rat hypothalamus: anatomical basis to suggest ultrashort feedback regulation” ‘Since ultrashort feedback regulation of TRH in the hypothalamus has been suggested by physiological studies’... we sought TRH synaptic contacts on TRH tuberoinfundibular neurons in the PVN. LM&EM —antiserum to the N-terminal cryptic sequence of the TRH precursor, preproTRH 25-50. At LM, in medial and periventricular subdivisions of the PVN. At EM, TRH-neurons appeared either tightly juxtaposed to TRH-immunopositive perikarya and dendrites or to establish axodendritic and axosomatic contacts suggestive of synaptic associations. These data provide a morphologic basis to support a neuroendocrine role for TRH or processed forms of proTRH in the PVN and in particular suggest their involvement as neuromodulators in an ultrashort feedback regulation of TRH tuberoinfundibular neurons.’’ (Toni, Jackson et al. 1990) “Neuropeptide-Y-immunoreactive innervation of thyrotropin-releasing hormone-synthesizing neurons in the rat hypothalamic paraventricular nucleus” At LM, a diffuse group of TRH-IR cell bodies were observed in the anterior parvocellular subdivision of the PVN and became more numerous and densely clustered in the medial and periventricular parvocellular subdivisions. --nice double immuno light and 393 EM study of NPY fibers innervating TRH [and some not TRH] neurons in all subdivisions of PVH; particularly dense in the anterior, medial, and periventricular parvocellular subdivisions”. --EM of NPY terminals are "highly reminiscent of the CA innervation to PVN" and a subpopuation probably co-localizes in CA terminals from A1, C1, C2, and C3”; and "since CA are thought to increase secretion of TRH and NPY has been demonstrated to decrease CA outflow in the hypothalamus, it is conceivable that NPY may modulate the facilitatory adrenergic and noradrenergic influences on TRH-synthesizing neurons in the PVN”. ~last paragraph: "We conclude that NPY innervation of TRH synthesizing neurons in the PVN represents a potential neuroendocrine regulatory pathway which may affect TRH tuberoinfundibular neurons. The effect of NPY may be directly on TRH neurons or indirectly through neuromodulatory effects of CA secretion co-contained in the same axon terminals. (Zoeller, Kabeer et al. 1990) “Cold Exposure Elevates Cellular Levels of messenger Ribonucleic acid Encoding Thyrotropin-Relasing Hormone in Paraventricular Nucleus Despite Elevated Levels of Thyroid Hormones” --Cold exposure increases blood levels of TSH and thyroid hormones by stimulating the secretion of TRH from the median eminence. Thyroid hormones reduce TRH release and cellular levels of TRH mRNA. By quant. ISH, cold exposure also increases cellular levels of TRH mRNA in neurons of the paraventricular nucleus (PVN), supporting the concept that TRH mRNA levels are reflective of TRH secretion in these neurons. The effect of cold appeared to be specific for TRH expression in the PVN. --Cellular levels of mRNA encoding CRH were elevated by cold exposure, which is predictable based on the cold-induced activation of the hypothalamic-pituitary-adrenal axis. There was a 24-h rhythm and a time of day difference in the effect of cold on TRH mRNA levels , but not on CRH. Cold exposure appeared to elevate TRH mRNA levels in all neurons of the PVN, indicating that the neurally mediated effect of cold on TRH expression can override the inhibitory effects of circulating T3 within the same neuronal population. (Ceccatelli 1991) (Ceccatelli 1991) ISH with oligos showed no change in TRH mRNA in response to colchicine [clear increase for CRH mRNA, and now a presence of NT and VIP mRNA in PVH parvos that wasn't seen in controls] this was a study designed primarily to examine CRH...**"for comparison the paraventricular TRH system (Ceccatelli, Eriksson et al. 1989), which is separate from the CRH neurons (Lechan and Jackson 1982), was analysed”. 394 (Blake, Eckland et al. 1991) “Inhibition of Hypothalamic Thyrotropin-Releasing Hormone Messenger ribonucleic acid during Food Deprivation” (with Lightman and seven others) (Carr, Fein et al. 1992) “A Cryptic Peptide (160-169) of Thyrotropin-Releasing Hormone Prohormone Demonstrates Bilolgical Activity in Vivo and in Vitro” --”PPT Stimulates TSH and PRL Synthesis” --TRH is synthesized as a precursor peptide containing five copies of the sequence Gln-His-Pro-Gly, QHPG, flanked by paired basic amino acids, and linked by other peptides. We tested TRH vs one cryptic peptide, PPT (160-169, SFPWMESDVT), as a possible physiological regulator of pituitary activity in vivo. PPT caused no consistent effects on either TSH or PRL secretion, while TRH stimulated the secretion of both hormones. However, PPT stimulated a dose-dependent increase in both pituitary TSH beta and PRL mRNA content at 240 min similar to TRH. -- Thus, PPT appears to be a physiological regulator of both TSH and PRL synthesis, but, unlike TRH, does not act as a secretagogue. Though there are only 2 or 3 [old] mapping studies, there is a lot of input/synapse data on TRH cells in PVH subsequent to those. Most recent papers seem to assume there is accurate mapping -- Perhaps each lab makes a working map or model for their own use? (Markakis and Swanson 1997) is probably the best recent map... It seems incomplete (low resolution?) relative to my data, even though she shows some TRH cells at each atlas level that encompases PVH. However her paper has the caveat that it is not a complete mapping, rather only a ‘snapshot in time’ to determine birthdates of phenotyped cells in a narrow time window. I wonder about all the apparently non-neTRH cells and scarcity of neTRH cells born on the same days... is this a possible effect of the tris/FB injection technique? [same TRH antibody that I used, but a different Fast Blue solvent solution] Markakis did cell counts in SON to determine/infer completeness of ne-labeling. Is it possible that there is a solubility difference of FB that shows up in parvis, but not in the more metabolically active magnos? Appendix XI References Aizawa, T. and M. A. Greer (1981). “Delineation of the Hypothalamic Area Controlling Thyrotropin Secretion in the Rat.” Endocrinology 109(5): 1731-1738. Blake, N. G., D. J. A. Eckland, et al. (1991). “Inhibition of Hypothalamic Thyrotropin- Releasing Hormone Messenger ribonucleic acid during Food Deprivation.” Endocrinology 129(5): 2714-2718. 395 Boler, J., F. Enzmann, et al. (1969). “The identity of chemical and hormonal properties of the thyrotropin releasing hormone and thyro-glutimyl-histidyl-proline amide.” Biochemical & Biophysical Research Communications 37(4): 705-10. Bowers, C. R., T. W. Redding, et al. (1965). “Effect of thyrotropin releasing factor (TRF) of ovine, bovine, porcine and human origin on thyrotropin release in vitro and in vivo.” Endocrinology 77(4): 609-16. Bowers, C. Y., K. L. Lee, et al. (1967). “A study on the interaction of the thyrotropin-releasing factor and L-Triiodothyronine: effects of Puromycin and Cycloheximide.” Endocrinology 82: 75-82. Brownstein, M. J., M. Palkovits, et al. (1974). “Thyrotropin-Releasing Hormone in Specific Nuclei of Rat Brain.” Science 185: 267-269. Brownstein, M. J., M. Palkovits, et al. (1976). Distribution of Hypothalamic Hormones and Neurotransmitters Within the Diencephalon. Frontiers in Neuroendocrinology. L. Martini and W. F. Ganong. New York, Raven Press. 4: 1-23. Burgus, R., T. F. Dunn, et al. (1969). “Molecular structure of the hypothalamic hypophysiotropic TRF factor of ovine origin: mass spectrometry demonstration of the PCA- His-Pro-NH1 sequence. [Structure moleculaire du facteur hypothalamique hypophysiotrope TRF d'origine ovine: mise en evidence par spectrometrie de masse de la sequence PCA- His-Pro-NH2].” C R Acad Sci [D] (Paris) [Comptes Rendus Hebdomadaires des Seances de l Academie des Sciences - D: Sciences Naturelles] 269(19): 1870-3. Carr, F. E., H. G. Fein, et al. (1992). “A Cryptic Peptide (160-169) of Thyrotropin-Releasing Hormone Prohormone Demonstrates Bilolgical Activity in Vivo and in Vitro.” Endocrinology 131(6): 2653-2658. Ceccatelli, S. (1991). Effect of different types of stressors on peptide mRNAs in the hypothalamic paraventricular nucleus. (Chapter IX of PhD thesis: Immunohistochemical and in situ Hybridization Studies on Peptides in the Hypothalamo-Pituitary System). Department of Histology and Neurobiology; Thomas Hokfelt, supervisor. Stockholm, Sweden, Karolinska Institute: IX-1 to IX-14, plus 5 figures. Ceccatelli, S. (1991). Immunohistochemical and in situ Hybridization Studies on Peptides in the Hypothalamo-Pituitary System [PhD Thesis]. Department of Histology and Neurobiology. Stockholm, Sweden, Karolinska Institute: 72 + appendix chapters. Ceccatelli, S., M. Eriksson, et al. (1989). “Distribution and coexistence of corticotropin- releasing factor-, neurotensin-, enkephalin-, cholecystokinin-, galanin- and vasoactive intestinal polypeptide/peptide histidine isoleucine-like peptides in the parvocellular part of the paraventricular nucleus.” Neuroendocrinology 49(3): 309-23. Dyess, E. M., T. P. Segerson, et al. (1988). “Triiodothyronine exerts direct cell-specific regulation of thyrotropin-releasing hormone gene expression in the hypothalamic paraventricular nucleus.” Endocrinology 123(5): 2291-7. 396 Enzmann, F., J. Boler, et al. (1971). “Structure and synthesis of the thyrotropin-releasing hormone.” Journal of Medicinal Chemistry 14(6): 469-74. Folkers, K., F. Enzmann, et al. (1969). “Discovery of the synthetic tripeptide-sequence of the thyrotropin releasing hormone having activity.” Biochem Biophys Res Commun 37(1): 123-6. Greer, M. A. (1951). “Evidence of hpothalamic control of the pituitary release of thyrotroin.” Proc Sod Exp Biol Med 77: 603. Harris, G. (1955). “Neural Control of the Pituitary Gland.”. Johansson, O. and T. Hokfelt (1980). “Thyrotropin-Releasing Hormone, Somatostatin, and Enkephalin: Distribution Studies Using Immunohistochemical Techniques.” Journal of Histochemistry & Cytochemistry 28(4): 364-366. Kawano, H., Y. Tosuru, et al. (1991). “Hypophysiotropic TRH-Producing Neurons identified by Combining Immunohistochemistry for Pro-TRH and Retrograde Tracing.” Journal of Comparative Neurology 307(4): 531-538. Kiss, J. and B. Halasz (1990). “Ultrastructural analysis of the innervation of TRH- immunoreactive neuronal elements located in the periventricular subdivision of the paraventricular nucleus of the rat hypothalamus.” Brain Research 532(1-2): 107-114. Lechan, R. M. and I. M. D. Jackson (1982). “Immunohistochemical Localization of Thyrotropin-Releasing Hormone in the Rat Hypothalamus and Pituitary.” Endocrinology 111(4): 55-65. Lechan, R. M., Y. Qi, et al. (1994). “Identification of thyroid hormone receptor isoforms in thyrotropin-releasing hormone neurons of the hypothalamic paraventricular nucleus.” Endocrinology 135(1): 92-100. Lechan, R. M. and T. P. Segerson (1989). Pro-TRH Gene Expression and Precursor Peptides in Rat Brain: Observations by Hybridization Analysis and Immunocytochemistry. Thyrotropin-Releasing Hormone: Biomedical Significance. G. Metcalf and I. M. D. Jackson. New York, New York Academy of Sciences. 553: 29-59. Lechan, R. M., P. Wu, et al. (1986). “Thyrotropin-releasing hormone precursor: characterization in rat brain.” Science 231(4734): 159-61. Liao, N., M. Bulant, et al. (1989). “Thyroid hormone regulation of neurons staining for a pro- TRH-derived cryptic peptide sequence in the rat hypothalamic paraventricular nucleus.” Neuroendocrinology 50(2): 217-21. Markakis, E. A. and L. W. Swanson (1997). “Spatiotemporal patterns of secretomotor neuron generation in the parvicellular neuroendocrine system.” Brain Research - Brain Research Reviews 24(2-3): 255-91. Merchenthaler, I., M. Meeker, et al. (1989). “Identification and Immunocytochemical Localization of a New Thyrotropin-Releasing Hormone Precursor in Rat Brain.” Endocrinology 124(4): 1888-1897. 397 Nishiyama, T., H. Kawano, et al. (1985). “Hypothalamic thyrotropin-releasing hormone(TRH)-containing neurons involved in the hypothalamic-hypophysial-thyroid axis. Light microscope immunocytochemistry.” Brain Research 345: 205. Redding, T. W., C. Y. Bowers, et al. (1966). “An in vivo assay for thyrotropin releasing factor.” Endocrinology 79(2): 229-36. Schally, A. V., C. Y. Bowers, et al. (1966). “Purification of thyrotropic hormone-releasing factor from bovine hypothalamus.” Endocrinology 78(4): 726-32. Segerson, T. P., H. Hoefler, et al. (1987). “Localization of Thyrotropin-Releasing Hormone Prohormone Messenger Riblnucleic Acid in Rat Brain by in Situ Hybridization.” Endocrinology 121(1): 98-107. Segerson, T. P., J. Kauer, et al. (1987). “Thyroid hormone regulates TRH biosynthesis in the paraventricular nucleus of the rat hypothalamus.” Science 238(4823): 78-80. Taylor, T., P. Gyves, et al. (1990). “Thyroid Hormone Regulation of TRH mRNA Levels in Rat Paraventricular Nucleus of the Hypothalamus Changes during Ontogeny.” Neuroendocrinology 52(3): 262-267. Toni, R., I. M. Jackson, et al. (1990). “Thyrotropin-releasing-hormone-immunoreactive innervation of thyrotropin-releasing-hormone-tuberoinfundibular neurons in rat hypothalamus: anatomical basis to suggest ultrashort feedback regulation.” Neuroendocrinology 52(5): 422-8. Toni, R., I. M. D. Jackson, et al. (1990). “Neuropeptide-Y-immunoreactive innervation of thyrotropin-releasing hormone-synthesizing neurons in the rat hypothalamic paraventricular nucleus.” Endocrinology 126(5): 2444-53. Tsuruo, Y., T. Hokfelt, et al. (1987). “Thyrotropin releasing hormone (TRH)-immunoreactive cell groups in the rat central nervous system.” Experimental Brain Research 68: 213-217. Winokur, A. and R. D. Utiger (1974). “Thyrotropin-Releasing Hormone: Regional Distribution in Rat Brain.” Science 185: 265-267. Zoeller, R. T., N. Kabeer, et al. (1990). “Cold Exposure Elevates Cellular Levels of messenger Ribonucleic acid Encoding Thyrotropin-Relasing Hormone in Paraventricular Nucleus Despite Elevated Levels of Thyroid Hormones.” Endocrinology 127(6): 2955-2962.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Initiation mechanism of estrogen neuroprotection pathway

PDF

Connections of the dorsomedial hypothalamic nucleus in the rat

PDF

Reactive synaptogenesis in the striatum of the adult rat: the role of dopamine

PDF

The ecology and behavior of the California moray eel Gymnothorax mordax (Ayres, 1859) with descriptions of its larva and the leptocephali of some other east Pacific muraenidae

PDF

The role of the sensorimotor cortical system in skill acquisition and motor learning: a behavioral study

PDF

Phylogenetics of the cutlassfishes (Scombroidei: Trichiuridae)

PDF

Experimental and computational analysis of microsatellite repeat instability in mismatch-repair-deficient mice

PDF

The histology and fine-structure of some bioluminescent organs in deep-sea fishes

PDF

The role of host proteins in retroviral cDNA integration

PDF

The importance of using domain knowledge in solving information distillation problems

PDF

The role of light organ symbionts in signaling early morphological and biochemical events in the sepiolid squid Euprymna scolopes

PDF

Three antibody-based immunotherapeutic modalities for malignancies

PDF

Moloney murine leukemia virus (MoMuLV) envelope protein: Interaction with receptor and engineering for targeted gene transfer

PDF

The design and synthesis of concurrent asynchronous systems.

PDF

The innervation of the teeth and periodontium of the rat

PDF

Identification of the first symbiotic determinants of Vibrio fischeri, the light organ symbiont of Euprymna scolopes

PDF

Cation localization in growth hormone and prolactin cells of the rat pars distalis

PDF

Studies of Drosophila sex-lethal protein interactions and cloning of a novel interactor, SIN

PDF

The effects of marketing communication on consumers' choice behavior: the case of pharmaceutical industry

PDF

The incorporation of radioactive carbons of glucose into the protein-bound amino acids of one-day-old mouse brain

Asset Metadata

Creator Simmons, Donna Marie (author)

Core Title Spatial distribution of neuroendocrine motoneuron pools in the hypothalmic paraventricular nucleus

School Graduate School

Degree Doctor of Philosophy

Degree Program Neuroscience

Degree Conferral Date 2006-05

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag biology, anatomy,biology, neuroscience,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-113164

Unique identifier UC11351271

Identifier 3237189.pdf (filename),usctheses-c17-113164 (legacy record id)

Legacy Identifier 3237189.pdf

Dmrecord 113164

Document Type Dissertation

Rights Simmons, Donna Marie

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA