University of Southern California Dissertations and Theses

Connections of the dorsomedial hypothalamic nucleus in the rat

(USC Thesis Other)

Connections of the dorsomedial hypothalamic nucleus in the rat

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content CONNECTIONS OF THE DORSOMEDIAL HYPOTHALAMIC
NUCLEUS IN THE RAT
BY
Richard Holley Thompson
A DISSERTATION PRESENTED TO THE FACULTY OF
THE GRADUATE SCHOOL OF THE UNIVERSITY OF
SOUTHERN CALIFORNIA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS OF THE DEGREE
DOCTOR OF PHILOSOPHY
(BIOLOGY)
May 1997
Copyright 1996 Richard H olley Thom pson
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UNIVERSITY OF SOUTHERN CALIFORNIA
THE g r a d u a t e s c h o o l
UNIVERSITY PARK
LOS ANGELES. CALIFORNIA 90007
This dissertation, written by
R ic h a rd ^ H ^ ^
under the direction of /lis Dissertation
Committee, and approved by all its members,
has been presented to and accepted by The
Graduate School, in partial fulfillment of re
quirements for the degree of
DOCTOR OF PHILOSOPHY
Dean of C ndusu Studies
Date
DISSERTATION COMMITTEE
Gut)
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
TABLE OF CONTENTS
Chapter I. General Introduction i
Historical and Modem Concepts of Hypothalamic Organization 2
Technical Considerations 12
Literature Cited 17
Chapter II. O rganization o f Projections from the D orsom edial 24
N ucleus of the H ypothalam us: A PHAL Study in the Rat
Materials and Methods 28
Results 30
Discussion 63
Literature Cited 100
Chapter III. O rganization of Projections from the D orsom edial i ll
N ucleus of the H ypothalam us: A R eexam ination w ith
Fluorogold and PHAL in the Rat
Materials and Methods 114
Results 117
Discussion 163
Literature Cited 175
Chapter IV. General D iscu ssion 189
Connections of the DMH 189
The Organization of Hypothalammic Pathways 193
Comparison with Medial Zone Nuclei: Implication for the 197
Organization o f the Hypothalamic Periventricular and
Medial Zones
Toward an Integrated Model of Hypothalamic Function 205
Literature Cited 222
1 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LIST OF FIGURES
Chapter II
Figure 1. Cam era lucida plots of PHAL-Iabeled neurons follow
ing injections of the DMH centered in the DMHa (ex
perim ent DMH 22), DMHp (experim ent DMH 23), and
DM Hv (experiment DMH 15) ........................................................ 3 1
Figure 2. A. Brightfield photomicrograph of a thionin-stained sec
tion caudally adjacent to the injection site illustrated in B
andC .................................................................................................. 35
Figure 3. Photomicrograph and line draw ing to illustrate the distri
bution of PHAL-labeled fibers in the horizontal plane of
section in experiment DMH 34 ...................................................... 37
Figure 4. Distribution of PHAL-labeled projections from experi
m ent DMH 3 plotted onto a series of standard tem plates
of th e ra t b ra in tak en from th e a tla s of S w anson
(1992;1993) 39
Figure 5. Darkfield photom icrograph to illustrate the appearance
of PHAL labeled fibers in the rostral paraventricular
nucleus of the hypothalam us (A), the preoptic region (B),
the paraventricular nucleus of the hypothalam us, at the
level of the posterior m agnocellular p art (C), and the
periaqueductal gray (D) 4?
Figure 6. General organization of projections from the DMH .............. 80
Figure 7. Schematic representation of the projections of the medial
zone nuclei of the hypothalamus at tuberal levels ...................... 95
Chapter III
Figure 1. Brightfield photomicrographs of Fluorogold injection site
(A-C) and caudally adjacent thionin-stained sections (A'-
C') in Case 17 ................................................................................... 118
1 1 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2. Distribution of Fluorogold-labeled neurons from experi
m ent 17 plotted onto a series of standard tem plates of
the rat brain taken from the atlas of Sw anson (1992) ............ I20
Figure 3. Dark- and brightfield photom icrographs to illustrate the
appearance of retrogradely-labeled neurons following
Fluorogold injection in the DMH (experim ent 17) ................ 125
Figure 4. Cam era lucida plots of PHAL-labeled neurons follow
ing injections centered in the parastrial (PS), m edian pre
optic (MePO), anteroventral periventricular (AVPv),
anteroventral preoptic (AVP), an tero d o rsal preoptic
(ADP), lateral septum (LS), bed nucleus of the stria
term inalis (BST), and parabrachial (PB) nuclei ...................... 129
Figure 5. Distribution of PHAL-labeled projections from injection
sites illustrated in Fig. 4 to the DMH and surrounding
region .............................................................................................. 132
Figure 6. General organization of projections to the DM H ................... 16I
Chapter IV
Figure 1. Schematic diagram representing the organization of the hy
pothalamic periventricular (left) and m edial (right) zones . . . 207
IV
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
List of A bbreviations
A M anterior amygdaloid area
ACB nucleus accum bens
aco anterior commissure, oifactory limb
act anterior commissure, temporal limb
ADP anterodorsai preoptic nucleus
AHA anterior hypothalamic area
ANN anterior hypothalamic nucieus
AHNa anterior hypothalamic nucleus, anterior part
AHNc anterior hypothalamic nucleus, central part
AHNd anterior hypothalamic nucleus, dorsal part
AHNp anterior hypothalamic nucleus, posterior part
AMBd nucleus ambiguus, dorsal division
AMBv nucleus ambiguus, ventral division
AON anterior olfactory nucleus
AONm anterior olfactory nucleus, medial part
AONpv anterior olfactory nucleus, posteroventral part
AP area postrem a
AQ cerebral aqueduct
ARM arcuate nucleus hypothalamus
ARM arcuate nucleus hypothalamus
AVP anteroventral preoptic nucleus
AVPV anteroventral periventricular nucleus hypothalamus
B Barrington’ s nucleus
BLAa basolateral nucleus amygdala, anterior part
BMAa basomedial nucleus amygdala, anterior part
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
BMAp basomedial nucleus amygdala, posterior part
BST bed nuclei stria terminalis
BSTad bed nuclei stria terminalis, anterior division, anterodorsal area
BSTal bed nuclei stria terminalis, anterior division, anterolateral area
BSTav bed nuclei stria terminalis, anterior division, anteroventral area
BSTdl bed nuclei stria terminalis, anterior division, dorsolateral nucleus
BSTdm bed nuclei stria terminalis, anterior division, dorsomedial
nucleus
BSTfu bed nuclei stria terminalis, anterior division, fusiform nucleus
BSTif bed nuclei stria terminalis, posterior division, interfascicular
nucleus
BSTju bed nuclei stria terminalis, anterior division, juxtacapsular nucleus
BSTmg bed nuclei stria terminalis, anterior division, magnoceilular nucleus
BSTov bed nuclei stria terminalis, anterior division, oval nucleus
BSTpr bed nuclei stria terminalis, posterior division, principal nucleus
BSTrh bed nuclei stria terminalis, anterior division, rhomboid nucleus
BSTsc bed nuclei stria terminaiis, anterior division, subcommissural zone
BSTtr bed nuclei stria terminalis, posterior division, transverse nucleus
BSTv bed nuclei stria terminalis, anterior division, ventral nucleus
C central canal, spinal cord/medulla
CA1 field CA1, Ammon’ s horn
CA1 so field CA1, stratum oriens
CA1 spd field CA1, pyramidal layer, deep
CA1 sp s field CA1, pyramidal layer, superficial
CA1 sr field CA1, stratum radiatum
CA2 field CA2, Ammon’ s horn
VI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
CA2sr field CA2, stratum radiatum
CA3 field CA3, Ammon’ s horn
CA3so field CA3, stratum oriens
CEAc central nucleus amygdala, capsular part
CEAI central nucleus amygdala, lateral part
CEAm central nucleus amygdala, medial part
CLI central linear nucleus raphe
CM central medial nucleus thalamus
COAa cortical nucleus amygdala, anterior part
COApI cortical nucleus amygdala, posterior part, lateral zone
COApm cortical nucleus amygdala, posterior part, medial zone
CSI superior central nucleus raphe, lateral part
CSm superior central nucleus raphé, medial part
CUN cuneiform nucleus
DMH dorsomedial nucleus hypothalamus
DMHa dorsomedial nucleus hypothalamus, anterior part
DMHp dorsomedial nucleus hypothalamus, posterior part
DMHv dorsomedial nucleus hypothalamus, ventral part
DMX dorsal motor nucleus vagus nerve
DR dorsal nucleus raphé
DTN dorsal tegmental nucleus [Gudden]
ENTmv entorhinal area, medial part, ventral zone
EW Edinger-Westphal nucleus
fi fimbria
fr fasciculus retroflexus
fx columns of the fornix
VII
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
lAD interanterodorsal nucleus thalamus
lAM interanteromedial nucleus thalamus
IF interfascicular nucleus raphé
III oculomotor nucleus
IMD Intermediodorsal nucleus thalamus
INF infundlbulum
IPNa interpeduncular nucleus, apical subnucleus
IPNc Interpeduncular nucleus, central subnucleus
IPNd Interpeduncular nucleus, dorsomedial subnucleus
IPNI Interpeduncular nucleus. Intermediate subnucleus
IPNI Interpeduncular nucleus, lateral subnucleus
IPNId Interpeduncular nucleus, lateral subnucleus, dorsal part
IPNII Interpeduncular nucleus, lateral subnucleus. Intermediate part
IPNIr Interpeduncular nucleus, lateral subnucleus, rostral part
IPNIv Interpeduncular nucleus, lateral subnucleus, ventral part
IPNr Interpeduncular nucleus, rostral subnucleus
KF Kolllker-Fuse subnucleus (of PB)
LC locus coeruleus
LOT laterodorsal tegmental nucleus
LGd lateral geniculate complex, dorsal part
LGv lateral geniculate complex, ventral part
LGvl lateral geniculate complex, ventral part, lateral zone
LGvm lateral geniculate complex, ventral part, medial zone
LH lateral habenula
LHA lateral hypothalamic area
LM lateral mammillary nucleus
Vlll
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LPO lateral preoptic area
LRNm lateral reticular nucleus, magnocellular part
LS lateral septal nucleus
LSd lateral septal nucleus, dorsal part
LSI lateral septal nucleus, intermediate part
LSv lateral septal nucleus, ventral part
MaPO magnocellular preoptic nucleus
MB mammillary body
met medial corticohypothalamic tract
MDc mediodorsal nucleus thalamus, central part
MDI mediodorsal nucleus thalamus, lateral part
MDm mediodorsal nucleus thalamus, medial part
MDRNv medullary reticular nucleus, ventral part
ME median eminence
MEAad medial nucleus amygdala, anterodorsal part
MEApd-a,b,c medial nucleus amygdala, posterodorsal part, sublayers a-c
MePO median preoptic nucleus
MEPO median preoptic nucleus
MEV mesencephalic nucleus of the trigeminal
MH medial habenula
MM medial mammillary nucleus
MPN medial preoptic nucleus
MPNc medial preoptic nucleus, central part
MPNI medial preoptic nucleus, lateral part
MPNm medial preoptic nucleus, medial part
MPO medial preoptic area
IX
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MRN m esencephalic reticular nucleus
MS medial septal nucleus
mtt mammillothalamic tract
NDB nucleus of the diagonal band
NI nucleus incertus
NTS nucleus of the solitary tract
NTSco nucleus of the solitary tract, commissural part
NTS! nucleus of the solitary tract, lateral part
NTSm nucleus of the solitary tract, medial part
och optic chiasm
opt optic tract
OV vascular organ of the lamina terminalis
PA posterior nucleus amygdala
PAG periaqueductal gray
PARN parvicellular reticular nucleus
PB parabrachial nucleus
PBIc parabrachial nucleus, central lateral part
PBId parabrachial nucleus, dorsal lateral part
PBIe parabrachial nucleus, external lateral part
PBIs parabrachial nucleus, superior lateral part
PBIv parabrachial nucleus, ventral lateral part
PBme parabrachial nucleus, external medial part
PBmm parabrachial nucleus, medial medial part
PBmv parabrachial nucleus, ventral medial part
PCG pontine central gray
PD posterodorsal preoptic nucieus
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PH posterior hypothalamic nucleus
PL prelimbic area
PMd dorsal premammillary nucleus
PMv ventral premammillary nucleus
PPN pedunculopontine nucleus
PRNc pontine reticular nucleus, caudal part
PRNr pontine reticular nucleus, rostral part
PS parastrial nucleus
PSCH suprachiasmatic preoptic nucleus
PT parataenial nucleus
PVa anterior periventricular nucleus hypothalamus
PVH paraventricular nucleus hypothalamus
PVHam paraventricular nucleus hypothalamus, anterior magnocellular part
PVHap paraventricular nucleus hypothalamus, anterior parvicellular part
PVHdp paraventricular nucleus hypothalamus, dorsal parvicellular part
PVHf paraventricular nucleus hypothalamus, fornical part
PVHIp paraventricular nucleus hypothalamus, lateral parvicellular part
PVHmpd paraventricular nucleus hypothalamus, medial parvicellular part,
dorsal zone
PVHmpv paraventricular nucleus hypothalamus, medial parvicellular part,
ventral zone
PVHpml paraventricular nucleus hypothalamus, posterior magnocellular part,
lateral zone
PVHpv paraventricular nucleus hypothalamus, periventricular part
PVi intermediate periventricular nucleus hypothalamus
PVp posterior periventricular nucleus hypothalamus
XI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PVpo preoptic periventricular nucieus
PVT paraventricular nucleus thalamus
RCH retrochiasmatic area
RE nucleus reuniens
REm nucleus reuniens, median part
RH rhomboid nucieus
RL rostral linear nucleus raphé
RM nucleus raphé m agnus
RO nucleus raphé obscurus
RPA nucleus raphé pallidus
RR mesencephalic reticular nucleus, retrorubral area
SC dg superior colliculus, deep gray layer
SCH suprachiasmatic nucleus
S C sg superior colliculus, superficial gray layer
SF septofimbrial nucleus
SFO subfornical organ
SI substantia innominata
sm stria meduilaris
sm d supramammillary decussation
SO supraoptic nucleus
SOr supraoptic nucleus, retrochiasmatic part
SPFm subparafascicular nucleus thalamus, magnocellular part
SPFp subparafascicular nucleus thalamus, parvicellular part
SPVC spinal nucleus of the trigeminal, caudal part
St stria terminalis
SUBv subiculum, ventral part
Xll
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
SUBv.sp subiculum, ventral part, stratum pyramidal
SUMI supramammillary nucleus, lateral part
SUMm supramammillary nucleus, medial part
sup supraoptic commissures
TM tuberomammillary nucleus
TMd tuberomammillary nucleus, dorsal part
TMv tuberomammillary nucleus, ventral part
TTd1 -4 taenia tecta, dorsal part, layers 1 -4 TTd
TTv3 taenia tecta, ventral part, layer 3
TU tuberal nucleus
V3 third ventricle
V3p third ventricle, preoptic recess
V4 fourth ventricle
VL lateral ventricle
VMH ventromedial nucleus hypothalamus
VMHa ventromedial nucleus hypothalamus, anterior part
VMHc ventromedial nucleus hypothalamus, central part
VMHdm ventromedial nucleus hypothalamus, dorsomedial part
VMHvI ventromedial nucleus hypothalamus, ventrolateral part
VTA ventral tegmental area
Zl zona incerta
ZIda zona incerta, dopaminergic group
Xlll
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter I.
General Introduction
As the result of a long history of physiological experiments, it is now generally
accepted that the hypothalamus is involved in the expression of responses that
function to ensure the survival of the individual (homeostasis) and the species
(reproduction); they include sleep/w ake cycle, and ingestive, thermoregulatory,
and reproductive behaviors (see Ranson and Magoun, 1939; Morgane and Panksepp,
1979; Swanson, 1987; Knobü and Neill, 1988). Challenges to these capacities have a
high priority to be amended and are usually called goal-oriented, or "m otivated",
behaviors. It is a general characteristic of these behaviors that they, more than others,
have two components, visceral and somatomotor, that are integrated to produce a
unified response. Based on this understanding and because the hypothalam us has
been implicated in the control of virtually every visceral (endocrine and autonomic)
response, it seems clear that the underlying neural circuitry m ust be exceedingly
complex, as well as highly differentiated. To clarify further the neuronal circuitry
underlying these responses, a series of experim ents has been perform ed to
investigate the connections of the dorsomedial nucleus of the hypothalamus (DMH).
The results of these experim ents provide valuable insights into the overall
organization of the hypothalamus, and suggest mechanisms by which sensory
stim uli can be in teg rated to produce th e coordinated v iscero m o to r and
som atom otor responses that underlie goal-directed behavior.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Historical and M odem Concepts of Hypothalamic Organization
The com plexity of the anatom ical connections of the hypothalam us is
evidenced in the fact th a t an understanding of the anatom ical circuits that
mediate these responses has lagged far behind the physiology, prim arily due to
lim itations of the available techniques. In general, hypothalam ic axons are thin
and unm yelinated and do not travel in compact bundles, b u t instead branch
profusely along their entire course, m aking them particularly refractory to
exploration w ith classical (pre-1970) m ethods (Swanson, 1987). Nevertheless,
this w ork laid the foundation for the present understanding of hypothalam ic
organizational principles.
The current parcellation of hypothalam ic nuclei is based on the classic w ork
of G urdjian (1927) an d K rieg (1932), using cellular an d m yelin sta in in g
techniques. These descriptions were placed into an organizational fram ew ork
by LeCros C lark (1938) and Crosby and W oodburne (1940). C rosby and
W oodbume (1940) described three longitudinal zones: a periventricular zone,
adjacent the third ventricle; a medial zone, lateral to the periventricular zone
and m ed ial to the fornix; and a la te ra l zone, lateral to the fornix an d
encom passing the m edial forebrain bundle. LeCros Clark (1938) divided the
hypothalam us into four rostro-caudal levels: preoptic, supraoptic (corresponding
to the anterior level of Sw anson (1987, 1992)), tuberal, and mammillary, each
defined by a nucleus of the medial zone—that is, the m edial preoptic nucleus.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
an terio r hypothalam ic n ucleus, ventrom edial hypothalam ic n ucleus, and
m am m illary body, respectively. It should be noted that the DMH is also located
at tu b eral levels of th e h y p o th alam u s, b u t is p erh ap s the m o st difficult
hypothalam ic nucleus to distinguish cytoarchitecturally (Cajal, 1995; Gurdjian,
1927; Krieg, 1932), and th u s serves as a poor m arker of the tu b eral level,
com pared to the distinct ventrom edial nucleus. Moreover, the results presented
here argue against inclusion of the DM H in the m edial zone (see below, and
C h ap ters 2 an d 4), th ereb y considerably sim plifying this m orphological
fram ework.
Although this cytoarchitectural organization is still valid today, insights
into hypothalam ic projections, especially those of the m edial zone, w ere slow
in coming. The earliest significant progress was m ade in the hypothalam ic
periventricular zone. Pines and Greving (as cited in Ranson and M agoun, 1939)
were the first to describe a neural pathw ay from the supraoptic nucleus to the
posterior pituitary, which w as, for the next 50 years, considered to control the
release of factors from cells in the pituitary itself. In 1928, however, Ernst Scharrer
(Scharrer, 1928) noted in fish sim ilarities betw een certain nerve cells in the
hypothalam us and endocrine cells outside of the nervous system . H e then
inferred that these neurons were "glandular" in nature, and that the secretory
m aterial was synthesized by these neurons and secreted from the som a. These
finding were extended by Bargmarm (1949) who reported that neurons in the
paraventricular and supraoptic nuclei, along w ith their axons and term inals in
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the posterior pituitary, were stained selectively w ith the Gomori m ethod, which
w as commonly used to stain secretory m aterial in the endocrine 15-cells (insulin-
producing) of the pancreas. It is now clear that separate groups of magnocellular
neurosecretory neurons in b o th the paraventricular and su p rao p tic nuclei
synthesize either vasopressin or oxytocin and release them in to the vasculature
of the posterior pituitary (see Swanson, 1986).
Clinical disorders such as acromegaly and C ushing's disease provided the
first evidence for the involvem ent of the anterior pituitary in endocrine function.
These observations were followed by reports th at puncture of the hypothalam us
produced sym ptom s sim ilar to hypopituitarism and that stim ulation of the
hypothalam us induced ovulation in the rabbit, indicating n eu ral control of
pituitary function (reviewed in Swanson, 1986). However, considerable debate
ensued, due to the absence of neural projections to the anterior pituitary. This
debate was resolved by the studies of Popa an d Fielding (1933), Wislocki and
King (1936), and H arris (Green and Harris, 1947,1949; Harris, 1948,1955), who
show ed a portal capillary system running d o w n the p itu ita ry stalk and
determ ined that the direction of blood flow w as from the brain to the pituitary.
This link was firm ly e stab lish ed as n eu ro v ascu lar w hen G uillem in and
Rosenburg (1955) and Saffran and Schally (1955) independently show ed that
crude extracts of the hypothalam us or median em inence contain a factor capable
of stim ulating adrenocorticotropin release from the anterior pituitary (reviewed
in Swanson, 1986; Everett, 1988). However, it took almost 30 years before all
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
five of th e m ajor releasing factors (that is, th y ro tro p in -, gonadotropin-,
corticotropin-, and grow th horm one-releasing horm ones and som atostatin)
could be purified and characterized. In addition, it is now firmly established
that each horm one is a peptide th at is synthesized in a separate group of
parvicellular neurons in the periventricular zone of the hypothalam us and that
each of these hypophysiotropic cell groups projects to the neurohem al zone of
the m edian eminence (Swanson, 1986).
Unfortunately) sim ilar advances w ere not forthcom ing in delineating other
hypothalam ic projections. Due to the "tortuous course" of axons emanating from
the m ed ial zone nuclei, projections could not be inferred from the Golgi
technique, as they could w ith large fiber tracts (Cajal, Szentagothai et al., 1968;
Ramon y Cajal, 1995). Because of their small diameter, axons often were not
im pregnated by degeneration techniques and because of their highly-branched
nature th at left most collaterals intact following lesions, they did not undergo
significant chromalytic (retrograde) changes (Swanson, 1986). However, when
axons could be visualized, many collaterals and synaptic varicosities could be
seen along their length. This led Szentagothai et al. (1968) to conclude that "...the
m edial hypothalam us ought to be considered a neuronal netw ork of quasi
random internal connexions..." from which, "excitation can spread from a given
focus in any direction and can establish an infinite num ber of closed self
reexciting chains."
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
On a more positive note, N auta and Haym aker (1969), in their review of
hypothalam ic connections, did report bidirectional connections of the lateral
and m edial zones w ith the "limbic forebrain" and "lim bic m idbrain" (that is,
the hippocam pus, am ygdala and lateral septum, an d the periaqueductal gray
and mesencephalic reticular formation, respectively), b u t were unable to detect
any direct sensory in p u t to the hypothalam us, o th er than inputs from the
olfactory tubercle and piriform (olfactory) cortex to the lateral hypothalam ic
area. They also reported, on the basis of Golgi m aterial, projections from the
lateral hypothalam ic area into the m edial zone. This led them to conclude that
th e m ajor avenue for th e m edial an d p e riv e n tric u la r zones (in clu d in g
neuroendocrine regions) to interact w ith the rest of the brain was through a
relay in the lateral hypothalam us, which in turn relayed in mesencephalic and
bulbar reticular nuclei to effect autonom ic m otor responses. Thus a m ajor
shortcom ing of this era, w hich was in fact recognized by N auta and H aym aker
(1969), was that direct sensory input from or motor output to the visceral systems
w ith w hich the hypothalam us had been physiologically associated could n o t be
dem onstrated anatomically.
A round 1970, the introduction of new methods, prim arily autoradiographic
anterograde, horseradish peroxidase (HRP) and fluorescent retrograde tract
tracing techniques, and immimocytochemistry, led to new insights into the
anatomical organization of the hypothalam us. From this w ork it was show n
that the m edial zone receives major descending inputs from the hippocam pus,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
am ygdala, and septal complex (Swanson and Cowan, 1977; Krettek and Price,
1978; Sw anson and Cowan, 1979) and a significant though m odest input from
the periaqueductal gray and m esencephalic reticular form ation (Berk and
Finkelstein, 1981a; Kita and Oom ura, 1982; Eberhart, 1985), confirm ing the
findings reported in N auta and Haym aker (1969).
D u rin g this tim e, d irect sensory in p u ts were fo u n d , especially from
viscerosensory and visual systems. Viscerosensory parts of the nucleus of the
solitary tract receive inputs from the vagus (particularly hepatic, cardiac, and
gastric branches) and glossopharyngeal nerves (Norgren, 1978; Powley and
Berthoud, 1986; Spyer, 1990). The nucleus of the solitary tract also has been
show n to project to the ventrolateral m edulla and parabrachial nucleus (Ricardo
and Koh, 1978). All three of these regions have been show n to provide m ajor
inputs to the hypothalam us, and the paraventricular hypothalam ic nucleus in
particular (Ricardo and Koh, 1978; Fulwiler and Saper, 1984; Loewy et al., 1981,
Sawchenko and Swanson, 1982). Interestingly, the only projection outside of
the canonical periventricular zone from each region is to the DMH and the
preoptic region.
A direct input from the retina to the hypothalam ic suprachiasm atic nucleus,
located in the periventricular zone just dorsal to the optic chiasm, was reported
independently by Hendrickson et al. (1972) and Moore and Lerm (1972). It was
then show n that lesions of this small nucleus essentially abolished behavioral
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
circadian rhythm s such as the sleep/w ake cycle, feeding and drinking, as well as a
variety of endocrine rhythm s, including diurnal rhythm s of corticosterone,
luteinizing horm one, and thyroid-stimulating hormone, among others (Brown-
Grant and Raisman, 1977; Raisman and Brown-Grant, 1977; Rusack and Zucker,
1979; Turek, 1985; Moore and Eichler, 1972; Abe et al., 1979). Although the projections
of the suprachiasmatic nucleus were examined with the autoradiographic m ethod,
the inability to m ake sufficiently small injections led to the conclusion that it projects
to the paraventricular hypothalamic nucleus, among other regions (Swanson and
Cowan, 1975; Berk and Finkelstein, 1981b). It was later shown with Phaseolus vulgaris
leucoagglutanin (PHAL; for inform ation on this method, see below) that the
projections of the suprachiasmatic nucleus tend to avoid both the medial zone nuclei
and the paraventricular hypothalamic nucleus (Watts et al., 1987). The largest
hypothalamic projection is found just ventral to the hypothalamic paraventricular
nucleus, followed closely by a projection to the DMH.
Also during this time, the importance of "extrasynaptic" sensory inputs was
beginning to be realized. These include such factors as brain temperature, osmolality,
and circulating horm ones such as corticosterone and thyroid hormones (reviewed
in Sutin and McBride, 1979; Boulant, 1981; Bourque, 1994; Keller-Wbod and Dallman,
1984; Dallman et al., 1993; Ahrén, 1986; Ceccatelli et al., 1992).
Another m ajor finding was direct projections from the hypothalam us to
autonom ic sym pathetic and parasym pathetic preganglionic neurons (Saper et
8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
al., 1976a; Hancock, 1976; Hosoya, 1980; Swanson and Kuypers, 1980). Although
scattered cells w ere found in the retrochiasmatic area and perifom ical region,
the m ajority of these neurons w ere found in d istin ct groups w ith in the
h y p o th a la m ic p a ra v e n tric u la r n u c le u s. N e u ro n s p ro je c tin g to the
interm ediolateral (sympathetic motor) column or m arginal zone (somatosensory,
particularly nociception and tem perature) of the spinal cord and dorsal vagal
com plex (parasym pathetic m otor), that is the nucleus of the solitary tract
(viscerosensory) and dorsal m otor nucleus of the vagus (visceromotor), are
concentrated in the dorsal, lateral, and ventral m edial parvicellular parts of the
hypothalam ic paraventricular nucleus. These findings indicate that a direct
projection from th e h y p o th alam u s influences b o th the cen tral relay of
sy m p ath etic a n d p a ra sy m p a th e tic sensory in fo rm atio n , as w ell as the
visceromotor o u tp u t of the vagus nerve and spinal cord.
Finally, it also has been show n that, although each medial zone nucleus
sends fibers into the m edial forebrain bundle (Conrad and Pfaff, 1976a,b; Saper
et al., 1978; Swanson, 1976; Saper et al., 1976b; Krieger et al., 1979), this projection,
by-and-large is not reciprocated (Saper et al., 1979; Berk and Finkelstein, 1982),
in contrast to evidence presented in the review of N auta and H aym aker (1969).
W ith this w ealth of inform ation generated, the era culm inated in the
comprehensive review of Swanson (1987), w ho review ed all of the available
anatom ical literatu re pertinent to the hypothalam us and offered the first
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
plausible overview of hypothalam ic organization. In this view, the lateral zone
is seen as a rostral extension of the reticular form ation w ith inputs and o u tp u ts
widely distributed to all levels of the neuraxis—from cerebral cortex to spinal
cord. H ow ever, because it has lim ited in teractio n s w ith the m edial an d
periventricular zones of the hypothalam us, despite the fact that it is clearly
involved in all types of behavioral and autonom ic events w ith w hich the
hypothalam us is associated, its specific role in the expression of these functions
is beyond the scope of this discussion (for reviews of this extensive topic, see
Swanson, 1987; Bemardis and Bellinger, 1993; W inn, 1995)
The m edial zone is thought to initiate specific m otivated behaviors, rather
than visceral responses (Swanson, 1987). Anatomically, the organization is seen
as receiving massive descending inputs from the "lim bic" telencephalon (that
is, the hippocam pal form ation, prefrontal cortex, am ygdala, lateral septal
nucleus, and bed nuclei of the stria terminalis), and less dense inputs from the
brainstem . The output is seen as organized such that each m edial zone nucleus
sends an ascending and descending projection into the m edial forebrain bundle
on the one hand, while projecting to specific regions of the periventricular zone,
on the other.
The periventricular zone is the final common pathw ay for the control of the
anterior, in term ed iate, an d p o ste rio r p itu ita ry b ecau se it co n tain s th e
m otoneurons of the neuroendocrine system . E specially prom inent is the
10
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
paraventricular nucleus of the hypothalam us, which is com posed exclusively
of viscerom otor neurons projecting either to the m edian eminence to control
the anterior pituitary, the posterior pituitary directly, or brainstem and spinal
cord autonom ic cell groups, including sym pathetic and p arasy m p ath etic
preganglionic neurons. A nother notable constituent of the periventricular zone
is the suprachiasm atic nucleus, w hich has a m ajor role in p ro d u cin g and
entraining endogenous rhythms such as the sleep/w ake and reproductive cycles.
Thus the periventricular zone can be viewed as a m otor region that generates
viscerom otor and behavioral (sleep/w ake cycle) rhythms.
W ith the advent of the exclusively anterograde tracer PHAL, w hich allows
restricted injection sites and the examination of short connections, this m odel
has continued to be refined. The studies described in this thesis utilize this tech
nique alone (chapter 2) and as the complement to Fluorogold retrograde trac
ing experim ents (chapter 3) to examine in detail the connections of the DMH.
These studies show that the majority of DMH outputs are intrahypothalam ic,
w ith only lim ited projections to the telencephalon and brainstem . The m ajority
of inputs also arise in the hypothalam us, from both m edial and periventricular
zones, b u t significant inputs also originate from limited regions of the telen
cephalon and brainstem. These results contrast significantly w ith other recently
published PHAL studies of the "classic" m edial zone nuclei projections (that is,
the m edial preoptic, anterior hypothalam ic, and ventrom edial hypothalam ic
nuclei) and provide considerable insight into the overall organization of the
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
periventricular and m edial zones of the hypothalam us. Specifically, the results
of these experiments indicate that the concept of the periventricular zone should
be expanded to include prem otor as w ell as viscerom otor hypothalam ic cell
groups. Prem otor cell groups are defined as those that receive prim ary viscero
sensory in fo rm atio n a n d send a m ajor p rojection to the h y p o th alam ic
paraventricular nucleus, b u t not to the nuclei of the m edial zone. These results
also slightly modify the organization of m edial zone nuclei projections such
that each has ascending projections to the telencephalon (and thalam us), a mas
sive descending projection to the periaqueductal gray, and a significant projec
tion to the prem otor p e riv e n tricu la r zone, b u t n o t to the hypothalam ic
paraventricular nucleus, or other parts of the periventricular zone.
Overall, this conception of hypothalamic organization considerably refines and
extends the scheme originally proposed by Swanson (1987) and is consistent with a
primary role of the medial zone in initiating somatic components of goal-directed
behaviors, and with the periventricular zone in synchronizing and integrating
visceral output. The significance of these observations in relation to generating goal-
directed and circadian aspects of coordinated behavioral and autonomic responses
is considered further in the General Discussion in Chapter 4.
Technical Considerations
Before presenting th e results, it w ill be useful to consider briefly the
technique used to represent the data, i.e., m ap the projections of the DMH. The
12
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
relative merits of the PHAL m ethod have been discussed previously in detail
(Gerfen and Sawchenko, 1984; Ter Horst et al., 1984; Sawchenko and Gerfen,
1985; Wouterlood and Groenewegen, 1985). Briefly, these merits are its sensitivity
a n d selectivity. W hen injected under the rig h t conditions, PHAL is not
incorporated by fibers-of-passage and is transported only anterogradely. It fills
the entirety of the labeled cell, including dendrites and axon. From this derives
its sensitivity. Individual axons can be distinguished w ith the clarity of a Golgi
im pregnation. Telodendria, boutons, and boutons-of- passage can be clearly
visualized (see Cajal (1995) for terminology). Therefore, not only can fibers-of-
passage versus term inating fibers be easily distinguished, but also the relative
degree of ramification to produce a terminal field (that is, fewer axons branching
profusely versus greater num bers of axons branching m oderately). Similarly,
Fluorogold is very sensitive, yet relatively résistent to uptake by undam aged
fibers-of-passage as com pared to other retrograde tracers (Schmued and Fallon,
1986). Moreover, an antibody to Fluorogold is now available that allows this
tracer to be labeled by conventional im m unohistochem ical m ethods. This
provides a light-stable reaction product w ith very low background, enabling a
very precise localization of labeled neurons because they can be compared
extensively to Nissl-stained adjacent sections.
To take full advantage of these features of these m ethods, and particularly
the resolution afforded by PHAL, we use a commercially available illustration
13
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
program (Adobe Illustrator'*'” ) in conjunction with tem plates of the atlas of
Swanson (1992; 1994). These drawings allow representation of fibers at virtually
infinite resolution and detail in a standardized way. Templates provide a broader
context for in terp retatio n of the experim ental observation, b ecau se the
relationship of a nucleus is preserved and all nuclei of the relevant level are
illustrated, not just the nucleus of interest. This provides a great aid in precise
and reproducible localization of projections and the cells of origin. The tem plates
also provide a foundation for comparison across experim ents. If the initial
translation from experimental material onto the standard is reasonably accurate,
useful com parisons can be m ade across experim ents and ex perim enters
(laboratories). Having all the nuclei of an atlas present aids accuracy because it
adds additional constraints that must be met. Ideally, one looks at a term inal field
in relation to all of the surrounding nuclei to confirm the exact location and
orientation of fibers.
At first glance, it would appear that translating the experimental m aterial onto
standardized templates relies heavily on the interpretation of the experimenter.
Differences in plane of section m ust be corrected and the region of interest m ust be
identified accurately to localize the ram ifying fibers. H ow ever, w hen the
experimenter makes a camera ludda drawing of a labeled section, the information
contained in that section is the physical location and relative number of fibers. There
is nothing to identify directly which nucleus the fibers are innervating. Thus, the
judgment needed to pair observed fibers with a known nucleus is essentially the
14
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
same for placing them directly onto a template. One final concern in transforming
onto a standard template is correction for plane of section differences. The variation
in plane of section can result in large differences betw een the dorsal/ventral or
right/left surfaces. However, local changes around the nucleus of interest are
smaller, so that the context within which a nucleus is centered remains intact. By
proceeding from local change to local change, section by section, one can work
across the length or w idth of the brain to localize accurately the entire set of
projections to the plane of the atlas. It might also be argued that differences in the
plane of section cause the length of the fibers to change, depending on how close to
perpendicular or parallel to the plane of section they travel. In general, this is true,
but when terminating, fibers branch profusely into telodendria, which are not,
generally, oriented in one plane. In addition, fibers of passage are only difficult to
distinguish when they are perfectly perpendicular to the plane of section. In this
case, they can still be seen, but their numbers may be underestimated. Otherwise,
relative length and direction can be ascertained reliably. These problems are the
same as for pen and ink drawings, except with the latter, relative proportions are
generally difficult to represent, because in pen or pencil drawings the lines are too
thick to allow adequate representations of dense term inal fields. These problems
are largely surm ounted by standardizing the representation of the various kinds of
fibers (i.e., fibers-of-passage versus terminating fibers). In general, the more that
the axon resembles a fiber-of-passage, the more it is represented by a straight,
unbranched line. Also, the more the fiber is seen to be traveling parallel to the
15
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
plane of section, the longer the fiber is drawn. The more that the fibers resemble
telodendria, the more branches are given to each fiber. In this way, boutons-of-
passage can represented by a small branch on a passing fiber. Similarly, this form
allows the differentiation between what appears to be a few fibers terminating
profusely versus m any fibers with few branches or boutons (see chapter 2 PVHlp
(Fig. 4 1) versus PSch (Fig. 4 C-D). One final advantage of using atlas templates is
that there are readily available sufficient levels to describe completely a projection;
for example, four levels of the DMH are illustrated, and six of the paraventricular
hypothalam ic nucleus of the hypothalam us, representing all reported parts of
each nucleus.
16
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Literature Cited
Abe, K., J. Kroning, M.A. Greer, and V. Critchlow (1979) Effects of destruction
of the su p ra c h ia sm a tic n u c le i on the c irc ad ian rh y th m s in p lasm a
corticosterone, body te m p e ra tu re , feeding and plasm a th y ro tro p in .
Neuroendocrinology 29:119-131.
A hrén, B. (1986) Thyroid neuroendocrinology: N eural regulation of thyroid
hom one secretion. Endocr. Rev. 7:149-155.
B arg m an n , W. (1949) Ü b e r d e N e u ro se k re to risc h e V erk n iip fu n g von
H ypothalam us and N eurohypophyse. Z. Zellforsch. Mikrosk. Anat. 34:610-
634.
Berk, M L., and J.A. Finkelstein (1981a) Afferent projections to the preoptic area
and hypothalam ic regions in the rat brain. Neuroscience 6:1601-1624.
Berk, M.L., and J.A. Finkelstein (1981b) An autoradiographic determ ination of
the efferent projections of the suprachiasm atic nucleus of the hypothalam us.
Brain Res. 226:1-13.
Berk, M.L., and J.A. Finkelstein (1982) Efferent connections of the lateral
hypothalam ic area of the rat: A n autoradiographic investigation. Brain Res.
Bull. 6:511-526.
Bemardis, L.L., and L.L. Bellinger (1993) The lateral hypothalamic area revisited:
N eu ro an a to m y , b o d y -w e ig h t re g u la tio n , n e u ro e n d o c rin o lo g y an d
metabolism. Neurosci. Biobehav. Rev. 17:141-193.
Boulant, J.A. (1981) H ypothalam ic mechanisms in therm oregulation. Fed. Proc.
40:2843-2850.
B o u rq u e, C.W ., S.H R. O lie t, a n d D. R ich ard (1994) O sm o re c e p to rs,
osmoreception, and osm oregulation. Front. Neuroendocrinol. 15:231-274.
Brown-Grant, K., and G. Raisman (1977) Abnormalities in reproductive function
associated with the destruction of the suprachiasmatic nuclei in female rats.
Proc. R. Soc. Lond. [Biol.] 198:279-296.
Cajal, S. Ram on y (1995) H isto lo g y of the N ervous System of M an and
Vertebrates. Translated by N. Swanson and L.W. Swanson. Oxford.Oxford
University Press.
17
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ceccatelli, S., L. Giardino, and L. Calza (1992) Response of hypothalamic peptide
mRNAs to thyroidectomy. Neuroendocrinology 56:694-703.
Clark, W.E.L. (1938) Morphological aspects of the hypothalamus. In W.E.L. Clark,
J. B eattie, G. R id d o ch a n d N .M . D o tt (eds.): The h y p o th a la m u s.
M orphological, functional, clinical and surgical aspects. Edinburgh: Oliver
and Boyd, pp. 2-68.
Conrad, L.C.A., and D.W. Pfaff (1976a) Efferents from m edial basal forebrain
and hypothalam us in the rat. 11. An autoradiographic study of the anterior
hypothalam us. J. Comp. Neurol. 169:221-262.
Conrad, L.C.A., and D.W. Pfaff (1976b) Efferents from m edial basal forebrain
and hypothalam us in the rat. 1. An autoradiographic study of the m edial
preoptic area. J. Comp. Neurol. 169:185-220.
Crosby, E.C., and R.T. W oodbum e (1940) The com parative anatom y of the
preoptic area and the hypothalam us. Proc. Assoc. Res. N ervous Mental Dis.
20:52-169.
D allm an, M.F., A.M. Strack, S.F. Akana, M.J. Bradbury, E.S. H anson, K.A.
S crib n er, a n d M. S m ith (1993) Feast a n d fam ine: C ritic al role of
glucocorticoids w ith insulin in daily energy flow. Front. Neuroendocrinol.
14:303-347.
E b e rh a rt, J.A ., J.l. M o rre ll, M .S. K rieger, a n d D.W. P faff (1985) A n
autoradiographic study of projections ascending from the m idbrain central
gray, and from the region lateral to it, in the rat. J. Comp. Neurol. 241:285-
310.
Everett, J.W. (1988) Pituitary and hypothalam us: Perspectives and overview. In
E. Knobil and J.D. Neill (eds.): The physiology of reproduction. New York:
Raven Press, Vol. 1, pp. 1143-1159.
Fulwiler, C.E., and C.B. Saper (1984) Subnuclear organization of the efferent
connections of the parabrachial nucleus in the rat. Brain Res. 319:229-259.
Gerfen, C.R., and P.E. Sawchenko (1984) An anterograde neuroanatomical tracing
m ethod that shows the detailed m orphology of neurons, their axons and
term inals: im m unohistochem ical localization of an axonally transported
plant lectin, Phaseolus vulgaris leucoagglutinin (PHA-L). Brain Res. 290:219-
238.
18
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
G reen, J.D ., and G.W. H arris (1947) The neurovascular link b etw een the
neurohypophysis and adenohypophysis. J. Endocrinol. 5:136-146.
Green, J.D., and G.W. H arris (1949) Observations of the hypophysial portal
vessels of the living rat. J. Physiol. (Lond.) 105:359-361.
G uillem in, R., and B. Rosenburg (1955) H um oral hypothalam ic control of
anterior pituitary: a study w ith combined tissue cultures. Endocrinology
57:599-607.
Gurdjian, E.S. (1927) The diencephalon of the albino rat. J. Comp. N eurol. 43:1-
114.
Hancock, M.B. (1976) Cells of origin of hypothalm o-spinal projections in the
rat. Neurosci. Lett. 3:179-184.
Harris, G.W. (1948) N eural control of the pituitary gland. Physiol. Rev. 28:139-
179.
Harris, G.W. (1955) The function of the pituitary stalk. Bull. Johns H opkins Hosp.
97:358-375.
H endrickson, A.E., N. Wagoner, and W.M. Cowan (1972) An autoradiographic
and electronm icroscopic study of retino-hypothalam ic connections. Z.
Zellforsch. 135:1-26.
H osoya, Y. (1980) T he d istrib u tio n of spinal p ro jectio n n e u ro n s in the
hypothalam us of the rat, studied w ith the HRP m ethod. Exp. Brain Res.
40:79-87.
Keller-Wood, M.E., and M.F. Dallman (1984) Corticosteroid inhibition of ACTH
secretion. Endocr. Rev. 5:1-24.
Kita, H., and Y . O om ura (1982) An HRP study of the afferent cormections to rat
m edial hypothalam ic region. Brain Res. Bull. 8:53-62.
Knobil, E., and J.D. N eill (Eds) (1988) The physiology of reproduction. (2 Vols.).
New York:Raven Press.
Krettek, J.E., and J.L. Price (1978) A m ygdaloid projections to subcortical
structures within the basal forebrain and brain stem in the rat and cat. J.
Comp. Neurol. 178:225-254.
19
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Krieg/ W.J.S. (1932) The hypothalam us of the albino rat. J. Comp. N eurol. 55:19-
89.
Krieger, M.S., L.C.A. Conrad, and D.W. Pfaff (1979) A n autoradiographic study
of the efferent connections of the ventromedial nucleus of the hypothalam us.
J. Com p. Neurol. 183:785-816.
Loewy, A.D., J.H. W allach, and S. McKellar (1981) Efferent connections of the
ventral m edulla oblongata in the rat. Brain Res. Rev. 228:63-80.
Moore, R., and V.B. Eichler (1972) Loss of circadian adrenal corticosterone rhythm
following suprachiasm atic lesions in the rat. Brain Res. 42:201-206.
Moore, R.Y., and N.J. Lenn (1972) A retinohypothalamic projection in the rat. J.
Comp. Neurol. 146:1-14.
M organe, P.J., and J. Panksepp (Eds.) (1979) Handbook of the hypothalam us. (3
Vols.). New York:Marcel Dekker.
N auta, W.J.H., an d W. H aym aker (1969) H ypothalam ic nuclei a n d fiber
connections. In W. Haymaker, E. Anderson and W.J.H. N auta (eds.): The
hypothalam us. Springfield: Charles C. Thomas, pp. 136-209.
Norgren, R. (1978) Projections from the nucleus of the solitary tract in the rat.
Neuroscience 3:207-218.
Popa, G.T., and U. Fielding (1933) Hypophysioportal vessels and their colloid
accompanim ent. J. Anat. 67:227-232.
Powley, T.L., and H.-R. Berthoud (1986) Participaton of the vagus and other
autonom ic nerves in the control of food intake. In R.C. Ritter, S. Ritter and
C D. Bames (eds.): Feeding behavior: Neural and hum oral controls. San
Diego: Academic Press, pp. 67-101.
Raisman, G., and K. Brown-Grant (1977) The suprachiasmatic syndrome, endocrine
and behavioral abnormalities following lesions of the suprachiasmatic nuclei
in the female rat. Proc. R. Soc. Lond. [Biol.] 198:297-314.
Ranson, S.W., and H.W. M agoun (1939) The hypothalamus. Ergeb. Physiol. 41:56-
163.
Ricardo, J. A., and E.T. Koh (1978) Anatomical evidence of direct projections from
the nucleus of the solitary tract to the hypothalam us, am ygdala, and other
forebrain structures in the rat. Brain Res. 153:1-26.
20
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Rusak, B y and I. Zucker (1979) N eural regulation of circadian rhythm s. Physiol.
Rev. 59:449-526.
Saffran, M., and A.V. Schally (1955) The release of corticotropin by anterior
pituitary tissue in vitro. Can. J. Biochem. Physiol. 33:408-415.
Saper, C.B., A.D. Loewy, L.W. Sw anson, and W.M. C ow an (1976a) D irect
hypothalam o-autonom ic connections. Brain Res. 117:305-312.
Saper, C.B., L.W. Swanson, and W.M. Cowan (1976b) The efferent connections
of the ventrom edial nucleus of the hypothalam us of the rat. J. Comp. Neurol.
169:409-442.
Saper, C.B., L.W. Swanson, and W.M. Cowan (1978) The efferent connections of
the anterior hypothalam ic area of the rat, cat and monkey. J. Comp. N eurol.
182:575-600.
Saper, C.B., L.W. Swanson, and W.M. Cowan (1979) An autoradiographic stu d y
of the efferent connections of the lateral hypothalam ic area in the rat. J.
Comp. Neurol. 183:689-706.
Sawchenko, P.E., and C.R. Gerfen (1985) Plant lectins and bacterial toxins as
tools for tracing neuronal connections. TINS 8:378-384.
Sawchenko, P.E., and L.W. Swanson (1982) The organization of noradrenergic
pathw ays from the brainstem to the paraventricular and supraoptic nuclei
in the rat. Brain Res. 257:275-325.
Scharrer, E. (1928) Die Lichtempfindlichkeit blinder Elritzen (U ntersuchungen
über das Zw ischenhim der Fische. I). Z. Vgl. Physiol. 7:1-38.
Schmued, L.C., and J.H. Fallon (1986) Fluoro-Gold: a new fluorescent retrograde
axonal tracer w ith num erous unique properties. Brain Res. 377:147-154.
Spyer, K.M. (1990) The central organization of reflex circulatory control. In A.D
Loewy, and K.M. Spyer (eds.): Central regulation of autonom ic functions.
Oxford: Oxford University Press, pp. 168-188.
Sutin, J., and R.L. McBride (1979) Limbic and brainstem connections of the
hypothalam us. In P.J. Morgane and J. Panksepp. (eds.): H andbook of the
H ypothalam us. Vol. 1, Anatomy of the hypotiialam us. New York: M arcel
Dekker, pp. 555-592.
2 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Swanson, L.W. (1976) A n autoradiographic study of the efferent connecitons of
the preoptic region in the rat. J. Com p. Neurol. 167:227-256.
Sw anson, L.W. (1986) Organization of m am m alian neuroendocrine system . In
Bloom , RE. (ed.): H andbook of Physiology, The N ervous System , IV.
Baltimore, MD: W averly Press, pp. 1-125.
Sw anson, L.W. (1987) The hypothalam us. In A. Bjorklim, T. H okfelt and L.W.
Sw anson (eds.): H andbook of Chemical Neuroanatomy, Integrated system s
of the CNS, Part I. Amsterdam: Elsevier Science, Vol. 5, pp. 1-124.
S w an so n , L.W. (1992) B rain M ap s: S tru ctu re o f the R a t B rain.
Amsterdam:Elsevier.
Sw anson, L.W. (1994) Brain Maps: C om puter G raphics Files. A m sterdam :
Elsevier Science.
Sw anson, L.W., and W.M. Cow an (1975) The efferent connections of the
suprachiasm atic nucleus of the hypothalam us. J. Comp. Neurol. 160:1-12.
Sw anson, L.W., and W.M. Cowan (1977) A n autoradiographic stu d y of the
organization of the efferent connections of the hippocam pal form ation in
the rat. J. Comp. N eurol. 172:49-84.
Swanson, L.W., and W.M. Cowan (1979) The connections of the septal region in
the rat. J. Comp. N eurol. 186:621-656.
Sw anson, L.W., and H.G.J.M. Kuypers (1980) The parventricular nucleus of the
h y p o th alam u s: cy to arch itecto n ie su b d iv isio n s an d o rg a n iz a tio n of
projections to the pituitary, dorsal vagal com plex, and sp in al cord as
dem onstrated by retrograde flourescence double-labeling m ethods. J. Comp.
N eurol. 194:555-570.
Szentagothai, J., B. Flerko, B. Mess, and B. Halasz (1968) H ypothalam ic control
o f th e a n te rio r p itu ita ry An e x p e rim e n ta l-m o rp h o lo g ic a l stu d y .
Budapest:Akadém iai Kiado.
Ter H orst, G.J., H.J. Groenewegen, H. Karst, and RG. Luiten (1984) Phaseolus
vulgaris leuco-agglutinin imm unohistochemistry. A com parison betw een
autoradiographic and lectin tracing of neuronal efferents. Brain Res. 307:379-
383.
2 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Watts, A.G., L.W. Swanson, and G. Sanchez Watts (1987) Efferent projections of
the suprachiasm atic nucleus: I. Studies using anterograde transport of
Phaseolus vulgaris leucoagglutinin in the rat. J. Com p. Neurol. 258:204-229.
W inn, P. (1995) The lateral hypothalam us and m otivated behavior: A n old
syndrom e reassessed and a new perspective gained. Curr. Dir. Psychol. Sci.
4:182-187.
Wisloki, G.B., and L.S. King (1936) The perm eability of the hypophysis and the
hypothalam us to vital dyes with a study of the hypophysial vascular supply.
Am. J. Anat. 58:421-472.
W outerlood, F.G., and H.J. Groenewegen (1985) Neuroanatom ical tracing by use
of Phaseolus o«/garzs-leucoagglutinin (PHA-L): electron microscopy of PHAL-
filled neuronal somata, dendrites, axons and axon term inals. Brain Res.
326:188-191.
23
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter IL
Organization of Projections from the Dorsomedial Nucleus of the
Hypothalamus: A PHAL Study in the Rat
The dorsom edial n u cleu s of the hypothalam us (DMH) h as long been
recognized as a distinct entity in the hypothalam us and its location relative to
adjacent nuclei is well established. A rich history of investigation has implicated
this nucleus in a wide variety of responses, including those related to ingestive
(Bemardis, 1972; Dalton et al., 1981), circadian (Bellinger et al., 1976; Bellinger
et al., 1986; Buijs et al., 1993), reproductive (Gallo, 1981; Pan and Gala, 1985;
Gunnet and Freeman, 1985), cardiovascular (Anderson and Dimicco, 1990; Soltis
and Dimicco, 1991; Stotz-Potter et al., 1996), locomotor (Sinamon, 1984), and
stress / anti-anxiety functions (Soltis and Dimicco, 1992; Inglefield et al., 1994).
It is generally agreed that the DMH lies adjacent to the third ventricle caudal to
the paraventricular nucleus of the hypothalam us, dorsal to the ventrom edial
nucleus of the hypothalam us and ventral to the zona incerta. However, its caudal
and lateral borders are far less distinct. Laterally, it is bordered by the perifomical
region of th e lateral h y p o th alam ic area and cau d a lly it m erges alm ost
indistinguishably w ith the posterior periventricular nucleus. D escriptions of
its exact borders remain extremely vague and no systematic attempt has been made
to delineate them.
24
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The earliest description of the DM H as a nucleus can be found in Cajal (1995).
From Golgi im pregnated m aterial, he describes "an ovoid mass of gray m atter
that is separated from the m idline by a particularly thick fiber layer." Gurdjian
(1927) first recognized w hat is now referred to as the posterior part (the compact
p art of Faxinos and Watson, 1986), w hich he saw as m erging w ith the posterior
periventricular nucleus caudally. Krieg (1932) generally agreed with the borders
and divisions of Gurdjian (1927), b u t suggested that the lateral border extends
to the fomix. However, because m any retrograde tracing studies have show n
that the perifom ical region at the level of the DMH generates projections to the
brainstem and spinal cord (Saper et al., 1976; Hancock, 1976; H osoya, 1980;
Swanson and Kuypers, 1980; Schwanzel-Fukada et al., 1984; Tucker and Saper,
1985; Hosoya and Kohno, 1987), w e have adopted a m ore conservative view of
the lateral border- one that does n ot include this cell group (see the atlas of
Swanson, 1992).
Current views of DMH projections are based on the w ork of L uiten and
colleagues (Luiten and Room, 1980; Ter H orst and Luiten, 1986; 1987; reviewed
in Luiten et al., 1989). A lthough w e have confirm ed the m ajority of their
observations, the results of o u r experim ents are presented here for three
fundam ental reasons: differences in observed projections, differences in mode
of representation, and differences in interpretation.
25
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
First, in a prelim inary study of the DMH (Thompson et al., 1992), we noted
several discrepancies betw een projections labeled in our experim ents and those
reported earlier. These differences are addressed individually in the Discussion.
Second, the present results are illustrated in far more detail than previously,
and are sum m arized on a standard series of tem plates that facilitate comparison
w ith the projections of other regions that are plotted on the same templates.
Also, because these draw ings are generated w ith computer graphics applications
in a standard file form at, they can be incorporated eventually into electronic
databases.
A third argum ent for the present study relates to different interpretations of
the significance of certain DMH projections. Previous anatom ical studies have
em phasized either the descending DM H projections (Ter H orst and Luiten, 1986;
Luiten et al., 1989) or an intrahypothalam ic circuit said to involve reciprocal
projections from the ventrom edial nucleus and lateral hypothalam ic area to the
DMH and from the DM H to the paraventricular nucleus (Luiten and Room,
1980; Ter Horst and Luiten, 1986; 1987; Luiten et al., 1987). While we find that
DMH axons almost completely avoid the ventrom edial nucleus, we agree with
Ter H orst and Luiten (1989) that such axons pass adjacent to this nucleus and
could thus contact the distal dendrites of ventromedial neurons. However, based
on the low density of this projection, and the indeterm inate nature of the
observation (i.e., it has not been dem onstrated that axons from the DMH synapse
26
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
upon ventrom edial nucleus dendrites), little em phasis can be placed on this
observation. N evertheless, a projection to the ventrom edial nucleus has been
ad o p ted as a significant reflection of DM H function in the hypothalam us
(e.g., O om ura and Yoshimatsu, 1984; Steffens et al., 1988; D allm an et al.,
1993).
The em p h a sis on descending p ro jectio n s is b ased larg ely u p o n the
observation th a t DM H lesions change pancreatic nerve activity (Yoshim atsu
et al., 1984) and on considerable w ork show ing that DM H lesions also change
ingestive behavior (see B em ardis and Bellinger, 1987 for review ). H ow ever,
w e find th a t intrahypothalam ic projections are far m ore extensive than
descending projections. Therefore, it w ould seem m ore p arsim o n io u s to
exam ine the m ajor projections of the DM H and determ ine th eir function
first. F urtherm ore, because the DM H is bidirectionally connected w ith the
targets of subfornical organ projections, it seem s reasonable to conclude
that it m ust have som e role in body fluid hom eostasis. Finally, the projections
of the DMH are com pared w ith those of the hypothalam ic m edial zone. Based
on Phaseolus vulgaris leucoagglutanin (PHAL) studies, it is concluded that
the DM H is very dissim ilar to these nuclei and bears m ore resem blance, in
connections and function, to the periventricular zone.
27
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MATERIALS AND METHODS
Thirty-one adult male Sprague-Dawley rats (260-330 g) received a single injection
of PHAL (Vector Laboratories) into the region of the DMH. Each anim al was
anesthetized with a m ixture of ketamine and xylazine (v/v; Im l/k g b.w.), and a
single iontophoretic injection of a 2.5% solution of PHAL in 0.01 M sodium
phosphate-buffered saline (pH 7.4) was m ade through a stereotaxically positioned
glass micropipette (tip diam eter 10-15p) by applying a +5 pA current pulsed at 7-
second intervals provided by a constant current source (Midgard Electronics, model
CS3), for a period of 15 minutes. Postinjection survival times for all experiments
ranged from 10 to 14 days. The animals were perfused and the brains were processed
according to procedures described in detail elsewhere (Ganteras et al., 1992a). Briefly,
each anim al was deeply anesthetized with sodium pentobarbital (1.5 m l/k g b.w.)
and perfused transcardially with a solution of 4.0% paraformaldehyde in 0.1 M
sodium borate buffer at pH 9.5; the brains were removed immediately and postfixed
overnight in the same fixative with 10% sucrose added.
For experiments processed in the frontal plane, four series of 30 |xm thick frozen
sections were cut on a sliding microtome and collected at the following frequencies:
one-in-ten through the olfactory bulbs and forebrain rostral to the genu of the corpus
callosum, one-in-four through the hypothalamus and rostral third of the midbrain,
and one-in-ten through the remaining midbrain, hindbrain, and rostral spinal cord.
Additional sections were collected through tuberal levels of the hypothalam us that
2 8
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
contain the DMH and "quick stained" for confirmation of injection site location
(see Simerly and Swanson, 1988). For experiments processed in the horizontal plane,
30|im thick frozen sections were also cut on a sliding microtome and collected in a
one-in-three series throughout the entire depth of the brain. No "quick stain" w as
performed. All other processing of this tissue was identical to that of tissue cut in
the hrontal plane.
In both p lan e s, one com plete series of sections w as p rocessed for
im m unohistochem istry w ith an antiserum directed a g ain st PHAL (D ako
Laboratories) at a dilution of 1:1,000. The antigen-antibody complex was localized
by using a variation of the avidin-biotin complex system (ABC; H su and Raine,
1981; Hsu et al., 1981), with a commercially available kit (ABC Elite Kit, Vector
Laboratories). The sections were m ounted on gelatin-coated slides and treated w ith
osm ium tetroxide to enhance visibility of the reaction product. Slides were then
dehydrated and coverslipped with DPX. An adjacent series was always stained
w ith thionin to serve as a reference series for cytoarchitectonie purposes.
Sections were examined with a Leitz Laborlux D microscope w ith both bright-
and darkfield illumination. PHAL-labeled cells in injection sites were plotted w ith
the aid of a camera lucida on maps prepared from adjacent Nissl-stained sections.
The distribution of anterogradely labeled fibers was then transferred onto a series
of standard drawings of the rat brain (Swanson, 1993) using an Apple Macintosh
29
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
nfx and the draw ing program A dobe Illustrator (versions 3.2 and 5.5). The
parcellation of the rat brain follows the atlas of Swanson (1992).
RESULTS
Injection Sites
The parameters used in these experiments resulted in discrete injections of PHAL
that clearly labeled a large num ber of cells. In 15 injections, the labeled cells were
confined almost entirely w ithin the borders of the DMH. Of these, two w ere
primarily restricted to the anterior part, one was centered in the posterior part, and
three were centered in the ventral part, as defined by Swanson (1992). Two large
injections labeled all three parts and four injections labeled two of the three. In all
experiments there was more or less contamination in surrounding areas. However,
all injections showed the same set of projections and all fibers utilized the sam e
pathways. Therefore, the discussion below will concentrate on one case, DMH 3,
where a large amount of labeling w as evenly distributed among the three parts of
the DMH and is thus representative of all injections into the nucleus (Figs. 1 and 2).
Reference will also be made to experiment DMH 34 (Fig. 3), which was cut in the
horizontal plane, to provide useful com plem entary information about m ajor
pathways. The distribution of labeled axons in experiment DMH 3 is sum m arized
on a standardized series of drawings of the rat brain that are arranged from rostral
(A) to caudal (T) (Fig. 4). Terminology used to describe axon morphology is discussed
in Risold et al. (1994).
30
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DMH 3
n
DMH 22
•
• \
DMH 23
✓
f
/•
/
r - x \ /
I • \ \ 1
/• \ \
' - ^
t ^\
' - - y
, \ \
f M
\ ^
\ \L_.
DMH 15
PH
' 1 '
L' V \
/ PM d
it-'
‘- .V
PVp
PMv
Fig. 1. Camera lucida plots of PHAL-labeled neurons following injections of the
DMH centered in the DMHa (experiment DMH 22), DMHp (experiment DMH
23), and DMHv (experiment DMH 15). The injection in experiment DM H 3 la
beled cells in each of these subdivision and labeled projections were found to be
representative of those of the DMH as a whole. All sections are arranged from
rostral (left) to caudal (right). Every clearly labeled neuron in these 30 pm thick
sections was plotted. The distribution of labeled axons projecting from the DMH
3 injection site is illustrated in Fig. 4.
31
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Projections of the DMH
The projections of the DMH can be divided into ascending fibers innervating
areas rostral to the nucleus in the di- and telencephalon, and descending fibers
projecting to the caudal hypothalamus and brainstem. In general, DMH projections
are predominantly intrahypothalamic, ascending in a formation that surrounds,
but largely avoids, the medial zone nuclei. Although most fibers in the different
pathways converge and end at the level of the lamina terminalis, a few can be seen
as far rostral as the anterior olfactory nucleus. The total num ber of labeled
descending axons was quite small relative to the number of labeled ascending axons.
Nevertheless, the former could be traced to the midbrain, pons, and medulla. In
the following account, each of the pathways emanating from the DMH will be
described in detail.
Ascending Projections. Labeled axons that ascend from the DMH appear to
take three major pathways: a periventricular pathway, a lateral pathway that courses
through medial parts of the medial forebrain bundle, and a ventral pathw ay that
contributes some fibers to the medial zone nuclei, but mainly ascends adjacent to
the ventral surface of the brain.
Periventricular pathway. Projections from the DMH to structures near the midline
course within, or adjacent to, the periventricular nucleus of the hypothalamus. This
pathw ay can be seen clearly in both Figure 3 (horizontal) and Figure 4 (C-L). Dense
terminal fields are found in all levels of the periventricular pathway and related
32
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
periventricular nuclei except the arcuate nucleus, w hich receives only a m odest
projection. The p araventricular nucleus of the hypothalam us (Fig. 4 F-I),
suprachiasm atic preoptic nucleus (Fig. 4 D), anteroventral periventricular
nucleus (Fig. 4 C-D), and m edian preoptic nucleus (Fig. 4 C-E) receive a very
dense projection from the DMH via the periventricular pathway.
Within the paraventricular nucleus, fibers and term inals are distributed
throughout the parvicellular division, whereas m any fewer course through, or
term inate in, the m agnocellular division (Fig. 5c). It is w orthy of note that
term inal boutons are especially dense in the dorsal, ventral, and forniceal
parvicellular parts, which generate descending projections to the brainstem and
spinal cord (Saper et al., 1976; Hancock, 1976; Hosoya, 1980; Sw anson and
Kuypers, 1980; Schwanzel-Fukuda et al., 1984; Tucker and Saper, 1985; Hosoya
and Kohno, 1987). Axons innervating the forniceal p art of the paraventricular
nucleus are likely to arise also from the lateral pathw ay (see below). Fibers are
also very dense throughout the anterior parvicellular part of the paraventricular
nucleus. At the rostral pole of the paraventricular nucleus, labeled fibers are
especially dense and concentrated dorsally, largely avoiding the anterior
m agnocellular neurons (Fig. 5b). A few fibers can be seen to stray outside of
this nucleus along its length both dorsally and ventrally, b u t the adjacent
thalam ic nuclei and the subparaventricular zone, respectively, are largely
avoided.
33
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Rostral to the arcuate nucleus, a contingent of fibers descends from the
paraventricular nucleus to provide an especially dense terminal field surrounding
the caudal half of the suprachiasmatic nucleus, while sending very few axons into
the nucleus itself (Fig. 4 E-G). These fibers continue rostrally into the ventral preoptic
region to innervate densely the preoptic suprachiasmatic nucleus and primarily
superficial parts of the anteroventral periventricular nucleus, and merge with fibers
firom the other ascending pathways in undifferentiated parts of the preoptic region
(Fig.4C-E).
At the rostral pole of the paraventricular nucleus, the fibers in the dorsal
periventricular zone split into two components: one continues rostrally (adjacent
to the ventricle) to innervate the median preoptic nucleus, and another turns laterally
to innervate primarily the parastrial nucleus. Both of these pathways are very dense
and generate many varicosities and boutons. The medial component innervates
m ainly peripheral parts of the caudal m edian preoptic nucleus and the entire
ipsilateral rostral part. At the level of the rostral m edian preoptic nucleus, this
component of the periventricular pathway also appears to contribute to a strong
input to a region of the medial preoptic area surrounding the vascular organ of the
lamina terminalis, while leaving the vascular organ itself free of labeled axons (Fig.
3 and Fig. 4 C).
A much smaller component of this pathw ay skirts the rostral surface of the
anterior commissure and ascends along the m idline to traverse the dorsal median
34
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 2. A. Brightfield photom icrograph of a thionin-stained section caudally
adjacent to the injection site illustrated in B an d C. Dotted lines indicate the
boundaries of the DMH and its parts, corresponding to level 30 in the atlas of
Swanson (1992). Note large cells m edial to the fom ix w hich aid in defining the
lateral border of the DMH. B. Brightfield photom icrograph of PHAL-labeled
neurons in case DMH 3, corresponding to the cam era lucida draw ing in figure
1, panel 3. C. Darkfield photom icrograph of injection site in B. Arrows indicate
blood vessels that appear in each im age. Scale = 250 pm.
35
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m
w S “
m
m
36
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 3. Photom icrograph and line drawing to illustrate the distribution of PHAL-
labeled fibers in the horizontal plane of section in experim ent DM H 34. The
thionin-stained section is caudally adjacent to the PHAL injection site. The line
draw ing w as prepared by scanning the photom icrograph an d tracing the out
lines of the nuclei in Adobe Illustrator. The fibers were then placed on top of
this draw ing in a m anner sim ilar to that used for representing fibers on the
standard tem plates. The gray region indicates the area occupied by PHAL-la
beled neurons in the injection site. Scale = 500 pm.
37
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m
l i l f f *
m i r
AHN^ ' V '
%
500 pm
38
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
IS
us
IS
MS
l p o '
etfi
Hg.4. Distribution of PHAL-labeled projections 6om experiment DMH 3 plotted onto a
series of standard templates of the rat brain taken from the atlas of Swanson (1992;1993).
Drawings are arranged from rostral (A) to caudal (T). The shaded region within tiie DMH
indicates the area occupied by PHAL labeled cells in tiie ir^ection site (see Fig. lA).
39
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
Sf
US
BST
UPO
40
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
SFO
It s c H :
RE
LHA
41
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
VL
LHA
VL
MO
RE
LHA
42
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
T '
■/ : .
bm«>
43
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
LHA
44
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
PAG
CA1
L M
45
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
PA6
AA
« I * PAG
AÛ
M E V
C U N
tUT
PRNC
46
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 5. Darkfield photomicrograph to illustrate the appearance of PHAL labeled
fibers in the rostral paraventricular nucleus of the hypothalam us (A), the preop
tic region (B), the paraventricular nucleus of the hypothalam us, at the level of
the posterior magnocellular part (C), and the periaqueductal gray (D). Note the
significant bilateral labeling in (A) and (B), and that this labeling forms a circum
scribed bilateral input to the parastrial nucleus (B). Scale = 250 pm.
47
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
preoptic nucleus and ultimately reach the subfornical organ (Fig. 4 E-F). Along this
segment of the pathway, the fiber morphology is very simple. They travel to the
subfornical organ w ith no apparent branching, and very few varicosities, and end
mainly in peripheral parts of this nucleus.
The parastrial nucleus is one of the most striking terminal fields of the DMH
(Fig. 4 C-E, Fig. 5a). The majority of these fibers arrive by turning laterally through
the rostral paraventricular nucleus, although some fibers from the caudal m edian
preoptic nucleus m ay also add to this input. Together they form a well-circumscribed
and very dense input to the parastrial nucleus that is largely bilateral (see below).
These fibers are aligned in parallel with abundant branches and boutons. M any
continue through, and some travel dorsal to, the parastrial nucleus to innervate
surrounding regions, primarily the fusiform nucleus of the bed nuclei of the stria
terminalis (BST) and the small posterodorsal preoptic nucleus (Fig. 4 C-E). These
fibers also innervate undifferentiated parts of the medial preoptic area and may
contribute to the innervation of the anterodorsal preoptic nucleus as well.
The periventricular pathw ay carries fibers innervating extrahypothalam ic
midline structures as weU. From the level of the injection site, some axons follow
the midline dorsal to the third ventricle to provide a moderate input to the nucleus
reuniens at the level of the DMH (Fig 4 G-K). A few axons resembling fibers-of-
passage can also be seen in m edial parts of the zona incerta. Rostral to the DMH,
fibers in the nucleus reuniens and zona incerta are m uch more sparse. These few
48
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
axons resemble fibers-of-passage and generate only occasional boutons-of-passage.
A t the level of the rostral thalamus, a large num ber of fibers ascend from the
periventricular zone, passing through the rostral pole of the nucleus reuniens, to
enter the paraventricular nucleus of the thalamus (Fig. 4 G). As these fibers ascend,
a few enter the paratenial nucleus. These fibers have a simple morphology until
they reach the thalamic paraventricular nucleus, where they turn caudally and travel
im m ediately ventral and perpendicular to the ependym al surface. From this
position, they branch extensively, sending collaterals with many prominent boutons
ventrally into the paraventricular nucleus. This input varies from moderate to dense
along the entire rostrocaudal extent of the nucleus, but is generally restricted to more
dorsal regions (Fig. 4 G-M).
The fibers that ascend at the rostral pole of the thalamus are, by far, the largest
group entering the thalamic paraventricular nucleus. However, at all levels of the
diencephalon a few fibers are seen ascending directly to the nucleus through the
midline thalamus. For the most part, these fibers are few in num ber and demonstrate
few boutons or varicosities, indicating that they are predominantly fibers-of-passage.
Nevertheless, very restricted (apparent at only one anatomical level) branching
fibers were seen in both the rhomboid and mediodorsal nuclei (Fig. 4 I-L). In
addition, exiting the paraventricular nucleus laterally, a small num ber of fibers could
be seen in and around the habenula - fibers that occasionally branch and generate
boutons and varicosities, most often at the m edial border of the lateral habenula.
49
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Lateral pathway. From the injection site, a second group of labeled axons exits
the nucleus dorsal and ventral to the fomix to enter the medial forebrain bundle.
This pathway can also be seen clearly in Figures 3 and 4 (CM). As these fibers pass
laterally out of the nucleus, they give off substantial numbers of branches and
boutons, indicating that they innervate this level of the perifomical region very
heavily (Fig. 4 J-K). Amajority of the fibers turn rostrally and ascend through medial
regions of the medial forebrain bundle, although scattered fibers can be seen
throughout this area. In general, fibers in all parts of the lateral hypothalamic area
appear to be cut in cross-section and generate few boutons-of-passage, indicating
that they are mostly ascending fibers-of-passage. However, because they do generate
occasional boutons-of-passage, it m ust be considered that the DMH sparsely
iimervates neurons along this entire path.
As fibers exit the DMH to enter the m edial forebrain bundle, some remain
adjacent to the lateral aspect of the fomix and ascend as a fairly compact group.
Although most of these fibers appear to turn medially, again to enter fomiceal and
other parvicellular parts of the paraventricular nucleus of the hypothalamus, a
perifomical contingent of the lateral pathw ay can be seen throughout the anterior
hypothalamic level (Fig 4G-J). As these fibers travel adjacent to the fomix, they
brzinch profusely and generate many boutons. At more rostral levels of the emterior
hypothalamus, these fibers again merge indistinguishably with the lateral pathw ay
as it enters the preoptic region.
50
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Except for the perifomical group of fibers, the lateral pathway travels rostrally
through the tuberal and anterior levels of the lateral hypothalamic area as a fairly
diffuse group of fibers. As the lateral pathw ay reaches the preoptic region, it
aggregates into a m ore compact bundle. From this bundle of fibers, three distinct
components arise to form its major terminal fields. Perhaps the largest group of
fibers remains associated with the medial forebrain bundle, traveling in medial
parts of the lateral preoptic area and along the lateral border of the m edial preoptic
area (Fig. 4C-E). These fibers branch and provide m any boutons as they pass through
these areas, especially the medial preoptic area, which is one of the largest terminal
fields innervated by the DMH. These fibers also provide a strong input to the lateral
part of the medial preoptic nucleus, and to the anterodorsal and anteroventral
preoptic nuclei. The innervation of the medial preoptic nucleus is formed by fibers
from the medial preoptic area and is primarily restricted to its lateral part. In contrast,
the medial and central, sexually dimorphic, parts of the medial preoptic nucleus
are largely avoided (Fig. 4E-F). This branch of the lateral pathway follows the medial
preoptic region to its rostral end, continuously generating numerous branches and
boutons, and terminates at the level of the lamina terminalis, along w ith the major
components of the other ascending pathways.
A second major contingent of fibers from the lateral pathway turns dorsally to
enter the BST. Caudally, the most dense terminations (the greatest num ber of fibers
and the most prom inent boutons) in the BST are seen in the ventral nucleus and
ventral parts of the interfascicular nucleus (Fig. 4F). Significantly fewer fibers are
51
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
seen in the transverse and dorsal regions of the interfascicular nuclei. Rostrally, the
fusiform and dorsomedial nuclei appear to be innervated by num erous branching
fibers arriving from both the periventricular and lateral pathways (see above). The
remainder of the BST receives a small input from the lateral path, w ith the exception
of the principal and oval nuclei, which are largely avoided. These fibers innervate
the BST by following closely the anterior commissure rostral to its decussation.
From this position, these fibers generate a sparse in p u t to the im m ediately
surrounding BST, until its rostral pole, where they seem to end.
A third component of the lateral pathway exits the lateral preoptic area dorsally,
at levels rostral to the m edial preoptic nucleus, to enter the lateral septal nucleus
(Fig. 4B-D). fri general this is a small to moderate input seemingly provided by a
few axons that branch extensively. These fibers have the appearance of telodendria,
with many terminal branches and accompanying large boutons. To reach their
destination, these fibers skirt just rostral to the decussation of the anterior
commissure and arch dorsolaterally to provide a m oderate input to ventral and
intermediate parts of the lateral septal nucleus. This term inal field is m ost obvious
in caudal and intermediate levels of the intermediate part. Fewer fibers innervate
the ventral part and only occasional fibers can be seen in the medial septal and
septofimbrial nuclei. Fibers from this region of the lateral pathway proceed rostrally
to innervate also a small portion of the nucleus accumbens that is ventrally adjacent
to the lateral septal nucleus and medial to the lateral ventricle (Fig. 4 B). Rostral to
this level, very few fibers can be seen. Those that do remain follow the medial
52
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
surface of the brain and turn laterally to enter die anterior olfactory nucleus (Fig.
4A). These fibers, very few and witiiout obvious boutons, are the most rostral extent
of the DMH projection labeled in our experiments.
All fibers that reach die amygdala and hippocampus appear to use the lateral
pathway. Although the amygdalar innervation as a whole is quite sparse, a few
fibers can be seen taking each of two distinct pathways, the stria terminalis and the
ansa peduncularis. Fibers enter the ansa peduncularis along the length of the lateral
pathway. From the level of the injection site, a few labeled fibers traverse the lateral
hypodialamic area directly to enter the ventral supraoptic commissure system (see
Magoun and Ranson, 1942) (Fig. 4 I-J). Rostral parts of this pathway arise from
fibers that leave caudal regions of the BST and course through the substantia
iimominata, where they give rise to occasional boutons-of-passage. From these
pathways, a few fibers from the DMH could be followed to capsular and medial
parts of the central nucleus, and the medial, basomedial, basolateral, posterior, and
cortical nuclei of the amygdala.
Beyond the amygdala, a few fibers appear to extend caudally near the medial
surface of the temporal region, ending in layer 1 of the entorhinal area and in the
ventral subiculum (Fig. 4 M-O). These fibers are joined by a few axons that reach
the ventral hippocampal formation via the fimbria and end mainly in the stratum
radiatum of fields CA l and CA3 (Fig. 4 N). All fibers in this pathway and in the
hippocampal formation are generally of a sim ple type, with few apparent branches
or varicosities.
53
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ventral pathway. The ventral pathw ay appears to consist of fibers that arrive
from both the periventricular and lateral pathways. Periventricular fibers descend
adjacent to the ventricle and turn laterally dorsal to the arcuate nucleus. Some of
the fibers contributing to this portion of the pathw ay continue into the arcuate
nucleus, where they give off few, but prominent, boutons. Additional fibers separate
from the lateral pathway and course ventrolaterally, ventral to the fomix, before
turning m edially to join axons from the periventricular path (Fig. 4J-K). As these
fibers descend to form this pathway, branches and boutons can be seen along their
length, indicating that they form synaptic contacts in route. It is this pathway that
provides a few axons to the ventromedial nucleus and only a slightly greater number
to the anterior hypothalamic nucleus. By and large, the ventral pathway ascends in
the medial zone of the hypothalamus near the ventral surface of the brain. Most
fibers in this pathway appear to be fibers-of-passage. However, many more boutons-
of-passage can be seen along this route than in the lateral pathway, indicating that
the DMH provides a small to m oderate input near the ventral surface of the
hypothalamus. The only terminal fields to which this pathway specifically gives
rise are in the retrochiasmatic area (Fig. 4G-FI) and the supraoptic nucleus (Fig. 4F-
G), where a moderate num ber of boutons and branching fibers can be seen. Upon
reaching preoptic levels, these fibers condense and provide a dense terminal field,
with an immense number of boutons, just dorsal to the optic chiasm - and contribute
to the fiber plexus that surrounds the suprachiasmatic nucleus (Fig. 4E-F). These
fibers extend rostrally, where they merge indistinguishably with fibers from the
54
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
periventricular and lateral pathways, providing conjoint inputs to the preoptic
suprachiasmatic, anteroventral, and anteroventral periventricular nuclei, as well
as to other, undifferentiated, parts of the ventral preoptic region, before ending
near the rostral pole of the hypothalamus.
D escending P rojections. The descending projections of the DM H are
considerably less substantial and, as a whole, much more sparse than the ascending
projections. Nevertheless, the DMH utilizes several pathw ays to provide a modest
input to a wide variety of areas in the caudal hypothalam us and brainstem. From
the injection site, fibers begin to split into two primary pathways: a dorsal pathw ay
that ascends to enter the midbrain periventricular system and a ventral pathw ay
that courses through caudal parts of the hypothalamus and ventromedial parts of
the brainstem.
Dorsal Pathway. From the DMH, m any labeled axons leave the nucleus and
enter the dorsally adjacent posterior hypothalamic nucleus. These fibers diverge
and extend caudally in this nucleus, providing a moderately dense input w ith many
boutons (Fig. 4L-N). U pon reaching the caudal posterior hypothalamic nucleus and
rostral periaqueductal gray, these fibers turn dorsally to join the m idbrain
periventricular fiber system. As they extend dorsally, it is clear that a majority of
these fibers undergo a roughly 90 degree bend and then travel caudally, nearly
perpendicular to the transverse plane of section. However, it is possible that some
55
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of these fibers either branch or turn rostrally to enter the paraventricular nucleus of
the thalamus.
These fibers pass caudally in medial regions of the periaqueductal gray (Fig.
4N-R), providing a sparse innervation and giving off occasional fibers to
surrounding areas including the superior coUiculus, mesencephalic reticular nucleus,
and retrorubral area (Fig. 4 0 Q ). In all three regions, the num ber of fibers w as small
and their morphology simple, demonstrating very few varicosities or boutons. Fibers
from this pathway also provide a very sparse input of very simple fibers to the
dorsal nucleus of the raphe. At these levels, the input to the periaqueductal gray
becomes somewhat larger due to increased axonal branching and m any more
boutons (Fig. 4Q-R). The terminal field is now centered in ventromedial parts of
the periaqueductal gray, although some fibers can be seen to extend into dorsomedial
regions of the periaqueductal gray at the end of this pathw ay (Fig. 5D). In addition,
the ventromedial fibers also branch much more abundantly with m ore obvious
boutons, than do those in the dorsomedial region. At the caudal extreme of the
dorsal descending pathw ay, some fibers appear to turn laterally to enter the
cuneiform nucleus, although they have few boutons. Labeled fibers also provide a
comparatively dense input to Barrington's nucleus (see Satoh et al., 1978 and Imaki
et al., 1991 for localization), with many well-labeled boutons (Fig. 45). These fibers
extend into undifferentiated parts of the pontine central gray, while tending to avoid
the dorsal and lateral tegmental nuclei and the nucleus incertus. From this level.
56
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
occasional unbranched fibers arch laterally to end in the lateral part of the
parabrachial nucleus, the most caudal extent of the dorsal pathway.
Ventral Pathway: Descending axons forming the ventral pathway exit the DMH
as a diffuse group innervating all three zones (medial, lateral, and periventricular)
of the caudal hypothalamus, and reach the brainstem through the medial forebrain
bundle as it extends through the ventral tegmental area. The component of this
pathw ay that innervates the periventricular zone seems to continue into the capsule
surrounding the mammillary bodies, where it ends, w ith very few boutons. Caudal
to the mammillary bodies, there are very few axons labeled in the ventral descending
pathw ay and they descend as a discrete pathway near the midline that is not
continuous with any identifiable group of fibers exiting the injection site. Thus, it is
impossible to determine which pathw ay gives rise to these fibers. Nevertheless,
the course of this pathw ay will be described to its apparent termination in the caudal
medulla.
A t hypothalam ic levels, som e fibers leave the DMH to enter the lateral
hypothalamic area, while others descend adjacent to the third ventricle, and in
undifferentiated parts of the medial zone (Fig. 4L-M). For the most part, these fibers
nonspecifically innervate the caudal hypothalamus because very few fibers are seen
in the ventral premammillary and dorsal premammillary nuclei. The one exception
to this generalization is that the dorsal tuberomammillary nucleus receives a dense,
circumscribed plexus of fibers (Fig. 4M). These fibers spread into adjacent parts of
the posterior periventricular nucleus, which is otherw ise only m oderately
57
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
innervated. In the mammillary complex, sparse fibers, prim arily resembling fibers-
of-passage, were seen in the medial and lateral parts of the supramammillary
nucleus, and in the capsule surrounding the mammillary nuclei. Only an occasional
fiber w as seen passing through the m edial or lateral m am m illary nuclei.
Additional terminations, few in num ber but generating prom inent boutons, were
seen in the ventral tuberom am m illary nucleus (Fig. 4N), and they m ay in part
be a lateral continuation of fibers from the dorsal tuberom am m illary nucleus.
C audal to the m ammillary complex, m edial and lateral components of the
ventral pathw ay can still be seen as fibers passing through the ventral tegm ental
area and scattered through the interpeduncular nucleus (Fig. 40), respectively.
Caudally, it appears that fibers in these components join near the m idline before
entering the interfascicular, central linear, and superior central nuclei of the
raphe (Fig. 4P-R). In this way, these fibers descend through the pons and into
the nuclei raphe magnus and pallidus of the medulla, although fewer in num ber
(Fig. 4S-T). At m edullary levels, this sm all pathw ay fractionates, sending its
few fibers dorsally, laterally, and caudally. The dorsally directed fibers ascend
to the nucleus of the solitary tract, w here they provide a sparse innervation,
w ith very few boutons, of the m edial p a rt (Fig 41). A n occasional fiber can also
be seen in the commissural subnucleus of the nucleus of the solitary tract and
in the area postrema, as well. Caudally directed fibers rem ain within the nucleus
raphe pallidus, where they end.
58
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A few axons can also be identified that move laterally away from the midline at
medullary levels. These fibers travel through ventral parts of the m edullary reticular
nucleus, giving off an occasional bouton. From here, one or two distinct fibers also
can be identified descending adjacent to the spinal tract of the trigeminal nerve
and followed to the spinomeduUary junction, at which point they could no longer
be observed. It is unclear whether this is because the pathw ay ends here or because
of the difficulty in localizing a small num ber of fibers traveling nearly perpendicular
to the plane of section.
Ascending Contralateral Projections. The majority of ascending crossed projections
emerge from the periventricular path. From the injection site, a group of fibers
passes continuously over the length of the dorsal aspect of the third ventricle in
ventral regions of the thalamic gray (middle) commissure. Specifically, these axons
travel through ventral regions of nucleus reuniens to provide input primarily to
the contralateral DMH and paraventricular nucleus of the hypothalam us. The
innervation of these two structures seems fairly restricted. Thus, in the DMH the
posterior part is largely avoided (Fig. 4K), and in the paraventricular nucleus, this
is also true of the magnocellular division, whereas the medial parvicellular parts,
and the fomiceal part in particular, are preferentially innervated (Fig. 4H-I). This
pathway m ay also contribute a few axons to the contralateral lateral hypothalamic
area. However, a more obvious input to the latter derives from axons traveling
through the supraoptic commissures (Fig. 4G). These fibers begin to cross the midline
at levels of the rostral arcuate nucleus and continue through the retrochiasmatic
59
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
area, including the area surrounding the suprachiasmatic nuclei. Upon crossing
the midline, this group of fibers has three obvious components: one that turns
dorsally to travel through the periventricular zone and merges with fibers from the
thalamic commissure; another that sends a few axons into the medial and lateral
zones of die hypothalam us, and that is most obvious at intermediate levels of the
anterior hypothalamic nucleus; and a third that travels laterally along the ventral
surface of the brain, dorsal to the optic chiasm, to provide a few axons to all parts of
the supraoptic nucleus.
At the level of the rostral thalamus, the periventricular pathway sends a large
projection to the contralateral paraventricular nucleus of the thalam us through
rostroventral regions of the thalamic commissure (Fig. 4G). These fibers ascend
along the midline through the medial part of the nucleus reuniens. U pon reaching
the paraventricular nucleus of the thalamus, these fibers distribute in parallel to
the ipsilateral projection, although fewer axons are involved.
As with the ipsilateral pathways, many of the commissural fibers described
above extend rostrally into the preoptic region. Here they are joined by additional
crossing fibers to provide a significant bilateral input to most of this region (Fig.
4C-E). Perhaps the greatest num ber of these, at least in this case (see Discussion),
utilize a commissure form ed in the rostral pole of the paraventricular nucleus of
the hypothalamus and juxtaposed to the caudal aspect of the anterior commissure
(Fig. 5b). These axons provide a substantial input to the contralateral anterior
60
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
parvicellular part of the paraventricular nucleus, while almost completely avoiding
the anterior magnocellular part. It should be noted, however, that this commissure
could not be seen in every case, although the distribution of fibers was similar. This
could be due to the small diameter of the commissure, i.e., too small to be reliably
sam pled in a one-in-four series, or to variability in localization. To clarify this issue,
we examined brains sectioned in the horizontal plane. In all of these experiments at
least a few fibers were observed crossing rostrally in apposition to the anterior
commissure and in three out of five cases, this commissure appeared similar to that
seen in transverse sections. In the two experiments where the num ber of fibers
traveling through this commissiure was smaller, the apparent num ber of fibers
crossing rostral to the anterior commissure, through the length of the m edian
preoptic nucleus, increased. In all experiments, both horizontal and transverse, the
distribution of axons in the contralateral paraventricular nucleus and preoptic area
was similar.
As the fibers that cross through the paraventricular and median preoptic nuclei
m ove rostrally, they turn away from the periventricular zone to ramify in the
parastrial and posterodorsal preoptic nuclei, and m uch more sparsely in the
contralateral BST (Fig. 4F). Notably, however, the anterodorsal preoptic and
anteroventral periventricular nuclei, which receive moderate ipsilateral inputs, are
almost completely avoided on the contralateral side (Fig. 4 D). Some of these axons
passing rostral to the anterior commissure ascend to provide the few contralateral
fibers observed in the septal region and subfornical organ.
61
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As m entioned above, a significant number of fibers cross in the supraoptic
commissures (Fig. 4E-G). It is m ost likely th at these fibers travel rostrally to
su pplem ent th e innervation of the ventral preoptic region, especially the
contralateral preoptic suprachiasmatic and anteroventral preoptic nuclei. It also
appears that a smaller group of axons passes continuously dorseil to the optic chiasm
to contribute to the contralateral input. This is suggested by the continual presence
of labeled axons in the ventral ependymal layer of the third ventricle.
Descending Contralateral Projections. Contralateral descending pathways also
appear to m irror the ipsilateral pathways very closely. As opposed to the ascending
pathways, however, the greatest number of fibers appears to cross at the level of
the caudal hypothalamus. Because the ventral pathw ay descends along the midline
and is composed of very few fibers, it is impractical to differentiate ipsilateral and
contralateral components.
The major contralateral pathways in the caudal hypothalam us and brainstem
are formed by descending fibers that exit the DMH and travel through the
supram am m illary decussation and ventral parts of the posterior hypothalamic
nucleus. After crossing, they distribute ventrally to undifferentiated parts of the
caudal hypothalamic medial zone and the posterior hypothalamic nucleus. These
fibers provide an input along the midline to caudal regions of the posterior
hypothalamic nucleus and medial regions of the supramammillary nucleus (Fig.
62
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
4L-N). Occasional fibers also can be observed in the lateral hypothalamic area and
more caudaUy in the contralateral ventral tegmental area.
From the posterior h ypoth alam ic nucleus, fib ers ascend to en te r the
periaqueductal gray in a fashion similar to the ipsilateral projection. Although the
majority of axons in this pathw ay ascend through the posterior hypothalamic
nucleus, additional fibers seem to cross throughout the length of the periaqueductal
gray. This is indicated by the constant presence of fibers that appear to be moving
laterally dorsal and ventral to the cerebral aqueduct (Fig 4 N-R). Throughout the
length of the brainstem pathway, contralateral fibers m irror the distribution of the
ipsilateral pathway. The most num erous terminating fibers in the contralateral
pathw ay are located in caudal parts of the ventral central gray, similar to the
ipsilateral pathway. The contralateral dorsal pathway also appears to terminate at
the level of the parabrachial nucleus.
DISCUSSION
The results of this PHAL study clarify and extend previous findings regarding
the projections of the DMH. In contrast to some previous reports, the major
conclusion of the experiments reported here is that the majority of projections from
the DMH are intrahypothalam ic. M ore specifically, these intrahypothalam ic
projections preferentially target the periventricular zone of the hypothalamus and
the preoptic region.
63
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Projections of the Dorsomedial Nucleus
The DMH has long been recognized as a separate hypothalamic entity (Gurdjian,
1927; K rieg/1932; Cajal, 1995), and, we agree in general w ith the bounding areas
and nuclei described in earlier studies of DMH projections by Ter Horst and Luiten
(1986; 1987). However, as yet there is no precise definition of DMH borders, and
this has caused problems because laterally adjacent cell groups project to the spinal
cord (Saper et al., 1976; Hancock, 1976; Hosoya, 1980; Swanson and Kuypers, 1980;
Schwanzel-Fukuda et al., 1984; Tucker and Saper, 1985; Hosoya and Kohno, 1987)
and cerebral cortex (Saper, 1985), projections that are not typical of the bulk of DMH
projections. Therefore, we have taken a somewhat restricted view of the lateral
DM H border and have adopted the borders used in Gurdjian's (1927) original
description, and in the atlas of Swanson (1992).
Current views of DMH projections are based mostly on the PHAL studies of Ter
H orst and Luiten (1986; 1987; Luiten et al., 1987), and our results confirm m any of
their observations, although there are obvious differences as well. The latter stem
essentially from how the DMH boundaries are defined and differences in the size
of injection sites. In the previous work, injection sites were large and not clearly
illustrated. From the terminal fields illustrated, it is clear that the injection sites
spread to label neurons in surrounding areas, including the anterior hypothalamic,
ventromedial hypothalamic, and posterior hypothalamic nuclei, the perifom ical
region, and the zona incerta.
64
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
For example. Ter Horst and Luiten (1986) illustrated fibers in the peripeduncular
nucleus originating from their injection sites. However, after large injections of HRP
centered in the peripeduncular nucleus, Arnault and Roger (1987) show ed many
retrogradely labeled cells within the ventromedial and posterior hypothalamic
nuclei, but none in the DMH. One other prominent example is the possibility of a
major projection to the ventromedial nucleus from the DMH (Ter Horst and Luiten,
1986; 1987; Luiten et al., 1987). However, Üieir figure 3A (Ter Horst and Luiten,
1987, p. 194), which is said to illustrate a projection to the ventromedial nucleus
"after a PHAL injection in the anterior DMH," in fact shows a dense projection to
the dorsomedial part of the ventromedial nucleus and virtually no fibers in the
DMH. As discussed in the Results, all PHAL injections involving the DMH,
regardless of location, invariably produced extremely dense intranuclear labeling.
In addition, none of our experiments w ith PHAL injections centered in the DMH
labeled more than an occasional fiber in any part of the ventromedial nucleus.
Therefore, we suggest that the Ter H orst and Luiten injection site extended either
into th e anterior hypothalam ic nucleus, w hich projects m assively to the
ventromedial nucleus (see Risold et al., 1994, their Fig. 4), or into the dorsomedial
part of the ventromedial nucleus, which is the only part of the nucleus that exhibits
extensive intranuclear projections (Ganteras et al., 1994).
Another significant discrepancy in possible targets of the DMH involves the
circum ventricular organs. Ter H orst and Luiten (1986) reported significant
projections to the subfornical organ, vascular organ of the lamina terminalis, and
65
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
area postrema. Although we observed a few unbranched fibers in the subfornical
organ and an occasional fiber in the area postrema, we saw no labeled fibers in the
vascular organ. The exact reason for this discrepancy is unclear. However, it is clear
that the DMH sends a very dense projection to the area immediately surrounding
the vascular organ. It is possible that, because the vascular organ is small and difficult
to localize, a dense projection to surrounding areas was taken to include it. To answer
this question definitively, we conducted a series of experiments with PHAL injections
in the DMH, where the brains were sectioned in the horizontal plane. From this
work it is clear that axons from the DMH avoid this circumventricular organ, while
innervating adjacent areas very densely (Fig 3).
O ther areas reportedly receiving a significant projection where we saw no or
only sparse labeling (see Results, above) include motor, somatosensory and visceral
cortical areas, the ventral premammillary nucleus, the cerebellar cortex, the KoUiker-
Fuse nucleus, and regions in and around the locus coeruleus. Again, as discussed
below, these projections could arise from neurons adjacent to the DMH.
In addition, extensive topographically organized inputs to a variety of nuclei
have been reported. These include the preoptic region, anterior hypothalamic area,
mesencephalic reticular formation (Ter H orst and Luiten, 1986), ventromedial
nucleus of the hypothalamus, and lateral hypothalamic area (Ter Horst and Luiten,
1987). In contrast, we found that all injections centered in the DMH demonstrated
the same set of efferent targets and all used the same pathways to reach these targets.
66
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The differential involvement of adjacent nuclei mentioned above could also account
for this second set of discrepancies between the two studies. To avoid this problem,
and because the hypothalamus consists of relatively small and closely situated nuclei
with substantially different projection patterns, we used smaller injection parameters
to achieve injection sites largely restricted to the DMH, as well as each of its possible
parts (Fig. 1).
A third difficulty w ith earlier w ork is that illustration and description of DMH
efferents include little detail. For example, there was no discussion of pathways
utilized by DMH projections, and hypothalamic parcellation (e.g., the preoptic
region) w as limited so that the precise localization of projections and their context
relative to adjacent nuclei is very difficult to assess. Furthermore, the drawings
were m ade in such a way that even where the targets were identified, the relative
density of the projections is difficult to assess.
By w ay of summary, in aU experiments with PHAL deposits centered in the
DMH, regardless of rostrocaudal distribution o r concentration in various
subdivisions, labeled axons follow the same basic routes (see Fig. 6). Therefore,
axons projecting from the DMH m ay be divided into an ascending group that
innervates targets rostral to the nucleus in tiie diencephalon and telencephalon,
and a descending group to the caudal hypothalamus and brainstem. Ascending
fibers course through each of the three longitudinal zones of the hypothalamus,
but predominate in the periventricular zone. Some fibers ascending through these
67
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
pathways take a dorsal route to the thalamus, and at preoptic levels a smaller, though
significant, number of fibers ascend around the anterior commissure to enter the
septal region. A few axons in the ascending group can be seen to extend as far
rostral as the anterior olfactory nucleus. In addition, very small groups of fibers
travel through the fimbria to the hippocampus, and the stria terminalis and ansa
peduncularis to the amygdala.
Descending fibers entering the brainstem from the DMH follow two main routes:
(i) a dorsal pathway traveling through the midbrain periventricular system (Cajal's
periependymal longitudinal fascicle); and (ii) a ventral pathway that descends to
the brainstem through the medial hypothalamus and medial forebrain bundle. These
two pathways largely remain segregated throughout the brainstem. Although the
dorsal pathway appears to terminate at the level of the parabrachial nucleus, a few
axons, presumably from the ventral pathway, can be seen as far caudal as the caudal
medulla.
Ascending Projections
Intrahypothalamic projections. As mentioned above, the largest component of
fibers from the DMH innervates hypothalamic structures, particularly those of the
periventricular zone and preoptic region. Other, smaller, groups of fibers course
ventral to the medial zone and diffusely through the lateral zone. These three groups
of fibers form what are referred to here as the periventricular, medial, and lateral
pathways, respectively. In addition, as remarked in the Results, the output of the
68
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DM H is largely ipsilateral. However, as discussed below, there is a significant
contralateral component that utilizes a w ide variety of commissural systems.
In agreement w ith observations based on the Golgi method by Cajal (1995), we
found that the DMH generates very dense intranuclear projections. This observation
extends to injections m ade in certain areas surrounding the DMH, for example, the
dorsal capsule of the ventromedial nucleus and the medial perifomical region
(unpublished observations). PHAL injections in these areas or in the DMH proper
all project very densely to all parts of the DMH.
At the level of the injection site, fibers provide a dense input to the periventricular
zone, the perifomical region, and medial parts of the lateral hypothalamic area. As
these fibers enter their respective pathways, it is clear that they skirt, but almost
completely avoid, the hypothalamic ventromedial nucleus. It has been reported
that HRP injections in the ventromedial nucleus produce retrogradely labeled cells
in the DMH (Luiten and Room, 1980) and that such injections do not produce
labeling in the DMH (Kita and Oomura, 1982a). This issue is much more easily and
reliably addressed using the PHAL m ethod because axons from the DMH travel so
close to the ventromedial nucleus and it is well documented that HRP is incorporated
by fibers-of-passage as well as terminating fibers. A later PHAL study by Luiten
and colleagues (1987) is in general agreement with the present study as to the course
of fibers from the DMH, at least in relation to the ventromedial nucleus.
69
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Confirming the extent to which the DMH utilizes the periventricular pathway,
Fodor et al. (1994) showed numerous retrogradely labeled cells in the DMH after
WGA-HRP tracer injections in the anterior periventricular nucleus. It is generally
agreed that one of the major targets of the DMH is the paraventricular nucleus of
the hypothalamus. Such a projection has been shown after True Blue injections in
the paraventricular nucleus (Sawchenko and Swanson, 1983; Levin et al., 1987) and
PHAL (Ter Horst and Luiten, 1986; 1987 and Luiten et al., 1987; Levin et al., 1987) or
tritiated leucine (Sawchenko and Swanson, 1983; Ter H orst and Luiten, 1986)
injections in the DMH. All studies showing this cormection observed that the DMH
projects heavily to parvicellular parts of the paraventricular nucleus while sending
very few axons to magnocellular neurons. Sawchenko and Swanson (1983) observed
that paraventricular inputs also appear to arise from the periventricular zone and
dorsomedial regions of the medial forebrain bundle as well. Also in agreement
w ith the present findings, they reported that the dorsal and ventral m edial
parvicellular parts of the paraventricular nucleus are the m ost heavily innervated.
To this we add the observation that the DMH also significantly innervates the
fomiceal and lateral parvicellular parts of the paraventricular nucleus. These DMH
projections are of particular interest because these parvicellular parts of the
parventricular nucleus project densely to autonomic motor and prem otor nuclei of
the brainstem and spinal cord (Swanson and Kuypers, 1980). This projection to
these parts of the paraventricular nucleus seems to be the single largest output of
the DMH to an autonomic premotor structure. It is possible that m any effects of
70
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
DMH experimental manipulation on visceral motor activity may be effected by
this pathway (see below).
The DMH also projects fairly heavily to the other parvicellular groups of the
paraventricular hypothalamic nucleus. Many of these cells express hypophysiotropic
hormones, particularly corticotrophin releasing hormone, thyrotrophin releasing
hormone, and somatostatin, and project to the median eminence (see Swanson,
1987 for review). Although the appropriate double labeling studies have yet to be
performed, it has been noted (Ter H orst and Luiten, 1986) that certain DMH
projections o verlap significantly w ith th e d istrib u tio n of neurosecretory
corticotrophin releasing hormone-containing neurons (see Swanson et al., 1983).
At this level of analysis, it would seem that the projections of the DMH also overlap
significantly w ith the d istrib u tio n of h ypophysiotropic som atostatin and
thyrotrophin releasing hormone (see Johansson et al., 1984 and Lechan and Jackson,
1982, respectively) neurons as well. However, physiological evidence to support
these anatomical observations is poor.
Levin et al. (1987) demonstrated that the majority of galaninergic inputs to the
paraventricular nucleus arise from the DMH. Such fibers were most dense in anterior
parvicellular, periventricular, and m edial parvicellular parts of the nucleus.
Interestingly, while they observed many afferent fibers to both the dorsal and ventral
medial parvicellular parts, the incidence of galaninergic doubly-labeled fibers was
m uch higher in the latter than in the former. This suggests that, while the DMH
7 1
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
projects to both parts, these projections arise from different cell populations within
the nucleus. Fodor et al. (1994) later showed that many galaninergic cells can be
retrogradely labeled in the DMH after injections in the anterior periventricular
nucleus. This suggests that the axons of galaninergic cells in the DM H extend rostral
to the paraventricular hypothalamic nucleus to innervate additional targets along
the periventricular path.
We also observed a few axons among the m agnocellular neurons of the
paraventricular and supraoptic nuclei. Sawchenko and Sw anson (1983) also
observed this small projection and stated that it is found prim arily in areas where
oxytodnergic cells predominate.
Another projection that has figured prominently in DMH literature involves
the lateral hypothalamic area. In our experiments, many fibers exit the nucleus
laterally and travel through the lateral hypothalamic area, w here m ost of them
display the appearance of fibers-of-passage. However, at the level of the injection
site and in a small area following the fomix rostrally, we observed moderate terminal
fields in the lateral hypothalamic area. In addition, as this pathway enters the lateral
preoptic area, the fibers begin to turn medially and branch more as they distribute
within the preoptic and septal regions.
A DMH projection through the lateral hypothalamic area has been confirmed
w ith both anterograde and retrograde tracing studies. Retrogradely labeled cells
were first seen in the DMH after HRP placements in the lateral hypothalamic area
72
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Luiten and Room, 1980; Barone et al., 1981; Kita and Oomura, 1982b). However,
little can be gleaned from this work about the localization of DMH fibers in the
lateral zone because all injections were large, covering the entire medial to lateral
extent of the lateral hypothalamic area, hi one study (Kita and Oomura, 1982b) two
different levels of the lateral zone were injected. Each injection was very large and
both showed approximately equal numbers of labeled cells in the DMH. However,
these observations do confirm that this pathway persists throughout the length of
the lateral hypothalamic area and into the lateral preoptic area. In addition, due to
limitations of the HRP technique, it was impossible to determine w hether these
fibers innervate the lateral zone or merely pass through. In general agreem ent with
our results, subsequent PHAL studies of DMH efferents showed these lateral
projections to be largely restricted to the most medial parts of the medial forebrain
bundle (Ter Horst and Luiten, 1986; 1987).
Continuing rostrally, all pathways converge upon the preoptic region. Such DMH
projections have also been reported previously with anterograde and retrograde
tracer methods (Kita and Oomura, 1982a; Saper and Levisohn, 1983; Chiba and
Murata, 1985; Simerly and Swanson, 1986; Ter Horst and Luiten, 1986). However,
due to the small size of individual cell groups in the preoptic region (as defined by
Simerly et al., 1984) and the large size of tracer injections, these studies w ere only
able to establish that the DMH projects to this general area. In addition, the previous
PHAL study of DMH efferents did not distinguish any nuclei of the preoptic region,
including the medial preoptic nucleus. Thus, although a DMH to preoptic region
73
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
projection has been well-established, the distribution of these fibers was largely
unknown. Nevertheless, in a study of inputs to the medial preoptic nucleus (Simerly
and Swanson, 1986), an injection of tritiated leucine in the DMH labeled projections
to the preoptic region, including a very dense projection to the parastrial nucleus,
and substantial inputs to the lateral (but not m edial or central) part of the medial
preoptic nucleus, the preoptic suprachiasm atic, periventricular preoptic, and
anteroventral preoptic nuclei, and undifferentiated parts of the medial preoptic
area. From this distribution, they inferred that the primary innervation of the
preoptic region by the DMH courses through the medial forebrain bundle. While
they did see a periventricular pathw ay after DMH injection, it was not labeled
rostral to the paraventricular nucleus.
The results of our PHAL injections confirm all of the observations m ade by
Simerly and Swanson (1986) w ith two notable additions. First, we observed that
the periventricular pathway provides a major avenue from the DMH into the
preoptic region. It is from this pathw ay that projections to the parastrial, preoptic
periventricular and preoptic suprachiasmatic nuclei arise. Second, with the more
sensitive PHAL method, we also observed term inal fields in the median preoptic,
anteroventral periventricular, and anterodorsal preoptic nuclei.
It is also w orth noting that in a retrograde study of neural inputs to the median
preoptic nucleus. Saper and Levisohn (1983) concluded that the DMH does not
project to this nucleus because retrograde labeling in the DMH was not bilateral.
74
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
As m entioned above, we observed a substantial input to the m edian preoptic
nucleus. It is moderately dense and centered in lateral parts of the nucleus caudally,
b u t becomes m uch more dense rostrally, providing an especially prom inent input
to regions surrounding, but not within, the vascular organ of the lam ina terminalis.
The pattern of DMH inputs to the preoptic region as a whole bears a striking
resemblance to the distribution of neurons synthesizing gonadotrophin-releasing
horm one (see Swanson, 1986; Barry et al., 1985). However, in a preliminary double
labeling study, we were unable to find significant juxtaposition of PHAL-labeled
fibers w ith cells immimoreactive for this peptide (unpublished observation).
Therefore, although there is limited evidence for an influence of the DMH on the
estrous cycle (Gallo, 1981; Pan and Gala, 1985; G unnet and Freeman, 1985), the
evidence to date suggests that it may not be m ediated via a direct input to
gonadotrophin-releasing hormone cells.
In the caudal periventricular zone, the DM H provides a significant input to
d o rsa l regions of the p o sterio r p e riv e n tric u la r nucleus a n d th e dorsal
tuberom am m illary nucleus. It is interesting to note that, seem ingly via laterally
directed fibers, the DMH also provides a sm all, but well-form ed input to the
ventral tuberom am m illary nucleus. In addition, the general p attern of DMH
fibers in the caudal hypothalam us overlaps significantly w ith the distribution
of histam inergic cell groups (Kohler et al., 1985; W ada et al., 1991). However,
only in the dorsal and ventral tuberom am m illary nuclei do these histaminergic
75
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
cells form cytoarchitecturally distinguishable nuclei. T herefore, although
suggestive, any relation between these fibers and histaminergic cells is merely
correlational at this time.
Thalamic Projections: At the rostral and caudal poles of the diencephalon, axons
from the DMH can be seen to extend dorsally out of the hypothalamus. These fibers
seem to arise largely from the periventricular pathw ay and m ainly target the
thalamic paraventricular nucleus. The projection to this nucleus is the single largest
extrahypothalamic projection of the DMH, to which it provides a moderate to strong
input along its entire length. The innervation of this nucleus has previously been
described after both anterograde tracer injections of the DMH and retrograde tracer
placements in the thalamus. Retrograde injections involved surrounding thalamic
nuclei such as the paratenial and mediodorsal nuclei. However, w henever an
injection involved the thalamic paraventricular nucleus, num erous labeled cell were
consistently seen in the DMH (Cornwall and Phillipson, 1988; Chen and Su, 1990).
We observed that the majority of fibers in this projection ascend at the rostral pole
of the thalamus. Descending fibers to the brainstem also turn dorsally to enter the
periaqueductal gray and pass just caudal to the thalamic paraventricular nucleus.
It is therefore possible that some fibers turn rostrally at this point to enter the
paraventricular nucleus from its caudal pole. Only an occasional fiber was seen to
ascend along the thalamic midline to innervate the paraventricular nucleus directly,
and was thus not considered as a separate pathway. This description differs
76
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
somewhat from that of Ter Horst and Luiten (1986) who recognized one route to
the paraventricular nucleus at the rostral pole of the thalamus. However, they also
illustrate fibers ascending along the midline for the entire length of the diencephalon.
The source of this discrepancy remains unclear.
Telencephalic Projections: In general, telencephalic projections of the DMH are
quite small. However, they use a wide variety of pathways to distribute, primarily
to the septal region, hippocam pus, and amygdala. Except for the BST, where fibers
derive from both the periventricular and lateral pathways, all telencephalic fibers
arise from the lateral pathway. To enter the septal region, axons follow two routes
derived from the lateral pathway: one that targets mainly the BST and sends a few
fibers into the stria terminalis, and another that targets the septum. From this later
path, an occasional labeled axon can also be seen in the fimbria.
Except for the principal and oval nuclei, the BST is innervated, to varying degrees,
in its entirety. The lateral pathw ay seem s to target prim arily the ventral,
interfascicular, and transverse nuclei of the posterior division, and some of these
fibers also continue dorsally to enter the stria terminalis and end in the amygdala.
The BST also receives axons from the periventricular path. These fibers seem to
target predominately the posterodorsal and fusiform nuclei, but may also contribute
to a modest innervation of the dorsomedial, subcommissural, and anterodorsal
nuclei. There is little previous evidence for these projections. Thus, in a study of
BST inputs, no hypothalamic contributions were discussed or illustrated (Weller
77
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and Smith, 1982). However, in a PHAL study of DMH projections, some fibers were
illustrated in the BST, but their distribution was not discussed (Ter Horst and Luiten,
1986).
The lateral pathway also passes ventral to the anterior commissure. A contingent
of these fibers turns dorsally to innervate the septum, where the intermediate and
ventral parts of the lateral septal nucleus receive moderate and small projections,
respectively. Projections to the septal region have been reported previously in both
anterograde and retrograde studies. Retrograde tracer injections centered in the
"rostral" or "dorsal" lateral septum or in the septofimbrial nucleus labeled only a
few cells in the periventricular zone at the level of the DM H (Luiten et al., 1982).
This study also illustrated labeled cells just dorsal and lateral to the DMH, but
none in the nucleus proper. After a PHAL injection in the DMH, labeled fibers were
seen in the "lateral" and "dorsal" septum (Ter Horst and Luiten, 1986).
A few axons from the septal region enter the fimbria and stria terminalis. Those
from the stria terminalis primarily end in capsular and lateral parts of the central
nucleus, and in the basomedial, basolateral, posterior, and cortical nuclei of the
amygdala. These fibers are joined by a small pathway traveling through the ansa
peduncularis that exits the lateral hypothalamic area throughout anterior and tuberal
levels. It was reported previously that all fibers reaching the amygdala from the
DMH use only the latter pathw ay (Ter H orst and Luiten, 1986). As these fibers pass
to the amygdala, they generate occasional boutons and branches in the substantia
78
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
innominata, indicating synaptic contacts within this area. Evidence for this projection
has been reported w ith retrograde tracer placements in the substantia innominata
(Grove, 1988). In an HRP study of amygdalar inputs, no retrograde labeling was
seen in the DMH (Ottersen, 1980). However, Ter Horst and Luiten (1986) also showed
a few PHAL-labeled axons with a similar distribution.
Labeled axons in the fimbria travel to ventral regions of the hippocam pal
formation, where they terminate in fields CAl and CA3. Some of these fibers extend
caudal to the hippocampus, along w ith a few of the amygdalar fibers, and end as
unbranched fibers in layer 1 of the medial entorhinal area and the ventral subiculum.
These axons, although few in num ber are well documented. Wyss et al. (1979)
reported scattered cells in the DM H after HRP injections at any level of the
hippocampal formation (i.e., septal, intermediate, or temporal). Ter Horst and Luiten
(1986) reported labeled axons in field CA3 and the dentate gyrus. Here, w e saw no
fibers in the dentate gyrus and only a few in ventral regions of fields C A l and CA3.
This latter observation is confirmed by Saper (1985), who reported no cell labeling
in the DMH after large injections of the retrograde tracer True Blue in the dorsal
hippocampus. It is interesting to note, however, that at the level of the DMH,
retrograde labeling appears just dorsal and ventral to the nucleus. A projection
from the DMH to the ventral hippocampus was also observed in this study (Saper,
1985). After large injections in the ventral hippocampus that encompassed both
fields CA l and CA3, scattered labeling was seen in the DMH.
79
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 6. General organization of projections from the DMH. The relative size of
each pathw ay is roughly proportional to the thickness of the line associated
w ith it. The m ajor pathw ays used by fibers to cross the m idline are also
schem atized: 1, preoptic com m issure (Fig. 4C-D); 2, rostroventral p art of the
thalam ic com m issure (Fig. 4G); 3, supraoptic com m issures (Fig. 4E-H); 4,
caudoventral part of the thalam ic commissure (Fig. 4J-L); 5, prem am m illary
decussation (Fig. 4M); 6, supram am m illary decussation (Fig. 4N).
80
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(O
5 - ^
I
i
I
I
81
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Additional projections to the cerebral cortex were observed by Ter H orst and
Luiten (1986). They illustrated labeled fibers in w hat appear to be motor,
somatosensory and visceral cortical areas. Again, injections into these areas were
made by Saper (1985) and no retrograde labeling in the DMH was observed.
However, his study revealed num erous cells lateral and dorsal to the DMH, in the
lateral hypothalamic area and zona incerta, that were prominent after injections in
insular and motor areas.
Descending projections. Axons from the DMH to the brainstem follow two
main routes: a dorsal pathw ay that travels through the midbrain periventricular
system and a ventral pathway that courses ventrally near the midline. Although
the projections of the DMH to the brainstem are, as a whole, less numerous than its
intrahypothalamic projections, they are well-documented. Retrograde tracers have
been placed in the brainstem along each termination and pathway of the DMH.
It is fairly w ell-established th at the hypothalam us provides the largest
descending input to the periaqueductal gray and that the DMH contributes to this
input (Morrell et al., 1981; Beitz, 1982 and 1989; Marchand and Hagino, 1983, Veening
et al., 1991, Behbehani, 1995). However, the relative contribution of the DMH is
quite unclear, ranging in various reports from equal to that of the dorsal
premammillary input (Marchand and Hagino, 1983) to a diminutive input from
only a few scattered cells (Beitz, 1982). These studies have also reported projections
8 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
from the DMH to all levels of the periaqueductal gray probably because all of these
studies used HRP, w hich is known to be incorporated by fibers-of-passage.
Ter H orst and Luiten (1986) reported periaqueductal terminal fields as most
dense at the level of the oculomotor nucleus. In contrast, our PHAL experiments
are in agreement w ith the distribution reported by Veening et al. (1991), w ith the
greatest number of terminations (i.e., more boutons and increased branching) at
the level of the rostral inferior coUiculus. Interestingly, a composite diagram from
three large retrograde tracer injections in the periaqueductal gray (Veening et al.,
1991) shows a medial area in the DMH with no retrograde labeling. The work of
AUen and Cechetto (1992) supports the observation that at least lateral regions of
the DM H and the perifomical region project to the periaqueductal gray. They made
a large PHAL injection centered just m edial to the fomix at the level of the posterior
part of the DMH, and numerous cells in the DMH proper were undoubtedly labeled.
They illustrated a pattem of terminals in the central gray essentially identical to
that reported here (primarily ending in the most caudal reaches of the periaqueductal
gray, just rostral to and continuing into the pontine central gray). However, the
total num ber of fibers appears greater, which m ay be due either to a difference in
the num ber of cells labeled in the injection site, or because it is centered more lateral.
At the level of the oculomotor nucleus, a few axons can be seen to leave the
periaqueductal gray to enter the mesencephalic reticular nucleus. After m any large
retrograde tracer injections in the mesencephalic reticular nucleus, Shammah-
83
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Lagnado et al. (1983) reported only occasional cell labeling in the DMH. Essentially
in agreement with this study, Morrell et al. (1981) also observed scattered labeling,
while Ter Horst et al. (1984) reported no cell labeling in the DMH.
Caudal to the periaqueductal gray, the DMH sends a significant projection to
Barrington's nucleus and the surrounding pontine central gray, and a few axons
also extend into the parabrachial nucleus. Upon closer examination, it w ould appear
that the majority of cells projecting to Barrington's nucleus reside in the m ost lateral
regions of the DMH and adjacent perifomical region, while the major projection to
the parabrachial nucleus arises in the lateral hypothalamic area. Moga et al. (1990)
m ade a large HRP injection in the parabrachial nucleus that included both medial
and lateral com ponents, as w ell as surrounding areas (possibly including
Barrington's nucleus and the pontine central gray). Dense retrograde labeling was
observed in the lateral hypothalamic area (including the perifomical region) that
encroaches upon the DM H while leaving a label-free DM H per se. Smaller cholera
toxin placements centered in Barrington's nucleus show a similar p attern of
perifomical labeling, but m any fewer cells in other parts of the lateral hypothalamic
area (Valentino et al., 1994). These findings are supported by the work of Allen and
Cechetto (1992) who saw very little labeling in the parabrachial nucleus, b u t a very
dense projection to Barrington's nucleus after WGA-HRP and PHAL injections
centered in the perifom ical region. After a more lateral injection in the lateral
hypothalamic area (lateral to, but at the same level as, the perifomical injection),
they showed the converse: a very dense projection to the parabrachial nucleus and
84
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
only sparse labeling in the pontine central gray and Barrington's nucleus. Taken
together, these previous observations confirm the present results that the DMH
projects significantly to the pontine central gray and Barrington's nucleus and not
to the parabrachial nucleus.
A final projection suggested for the dorsal pathw ay is to the cerebellum (Ter
H orst and Luiten, 1986). In the present study, we saw no such fibers. In a study
where large HRP injections were made throughout the cerebellar cortex, scattered
labeling was observed in the lateral hypothalamic area at the level of the DMH, but
no cells were labeled medial to the fomix (Dietrichs et al., 1992).
M any retrograde tracer studies have also examined hypothalamic projections
to the brainstem raphe nuclei and consistently dem onstrated labeled neurons
(though few in number) in the DMH. The dorsal nucleus of the raphe is especially
difficult to inject selectively because of its proximity to the periaqueductal gray.
Nevertheless, whenever an injection included the dorsal raphe, at least a few cells
were reported in the DMH (Morrell et al., 1981; Kâlen et al., 1985). However, using
m uch smaller injections of HRP, Marchand and Hagino (1983) illustrated no cells in
the DMH. Studies with injections centered in the central linear and superior central
nuclei also labeled scattered cells in the DMH (Marcinkiewicz et al. 1989; Behzadi
et al., 1990). In the later study, moderate num bers of cells w ere evenly distributed
through all levels of the DMH. In addition, it is interesting that they also illustrated
a group of cells lateral to the DMH in the perifomical region and adjacent lateral
85
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
hypothalamic area. This pattern of cell labeling is very similar to that reported for
neurons projecting to Barrington's nucleus (see above). These studies also suggest
that not only does the DMH send a projection to the periaqueductal gray and raphe
nuclei, but that this is, at least partly, a glutamatergic projection, as indicated by
retrograde d-pH ] aspartate labeling (Beitz, 1989).
Continuing along the midline near the ventral surface of the brain, a few axons
can be seen in the nucleus raphe pallidus. Some such fibers turn laterally along a
semicircular path, traveling through the m edullary reticular nucleus ultimately to
end in the n u cleu s of the solitary tract. H osoya (1985) m ade num erous
autoradiographic injections into the DMH, posterior hypothalamic nucleus, and
surrounding areas. Significant projections were reported to the nucleus raphe
pallidus from areas centered dorsal and dorsolateral/caudal to the DMH, but each,
due to the size of the injection site, involves the DMH. He concluded that there is
a moderate to dense projection from the DMH to the nucleus raphe pallidus. Similar
results were reported by Carlton et al. (1983) for the nucleus raphe magnus using
retrograde tracers. They showed a prominent group of cells located just dorsal to
the DMH (although they called this cell group "DMH," they did not draw
boundaries for the nucleus). Because the cells are primarily located dorsal to the
third ventricle, it is most likely that they lie outside the boundaries of the DMH as
defined here.
At m edullary levels, a few axons from the ventral pathway turn laterally in a
semicircular course through the medullary reticular nuclei to end in the nucleus of
86
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the solitary tract. Although this projection is very small, these fibers C cin be seen to
target primarily the m edial subnucleus of the nucleus of the solitary tract and an
occasional fiber can also be seen in the area postrema. The course of the ventral
pathw ay on its w ay to these targets is documented by Ter Horst et al. (1984) and
Sofroniew and Schrell (1980). However, in the study of van der Kooy et al. (1984),
no labeled cells were reported in the DMH after large retrograde tracer injections in
the dorsal vagal complex. Ter H orst et al. (1984) m ade retrograde tracer placements
centered in the parvicellular reticular nucleus, nucleus ambiguus, and nucleus of
the solitary tract. After each injection, a few labeled cells in the DMH were observed.
In addition, Hosoya and M atsushida (1981) reported a small projection from the
DM H to the area postrema. This is in general agreement with the present results.
While we saw no clear terminals in the nucleus ambiguus, apparent fibers-of-passage
were observed in the area. The results of Sofroniew and Schrell (1980) are, however,
m ore difficult to interpret. An extremely large HRP injection centered in the medulla,
caudal to the area postrema, reportedly labeled numerous cells in the area of the
DMH, seemingly not in accord w ith this or previous studies.
Contralateral Projections. In general, the contralateral projections of the DMH
m irror the ipsilateral projections, but are far less dense. However, in comparison
w ith other hypothalamic nuclei, these projections are larger and utilize a greater
num ber of commissures (virtually every commissure available) and, as such, deserve
a more complete description. Only recently, with the development of more sensitive
techniques, has it become clear that the hypothalamus generates a significant crossed
87
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
projection. These fibers use both commissures that are normally thought to contain
long, extrahypothalam ic projections and o n e s th a t p referen tially carry
intrahypothalamic fibers.
Several discrete commissures passing in or near the hypothalamus are recognized
as major routes for axons to cross the midline; the supraoptic, anterior, and thalamic.
In general, these crossed projections have been thought to consist of long
extrahypothalamic projections (a) originating in the telencephalon and brainstem
and terminating in the contralateral basal ganglia, ventral thalamus and optic tectum
through the supraoptic commissures (Gurdjian, 1927; Krieg, 1944; Tsang, 1940;
Haym aker et al., 1969); (b) interconnecting the olfactory bulbs and temporal cortex
(Gurdjian, 1925; Haberly and Price, 1978; Hotel and Stelzner, 1981) through the
anterior commissure ; and (c) interconnecting the dorsal thalamus through the
thalamic commissures (Gurdjian, 1927; Krieg, 1944). Recently it has been recognized
that these commissures are also used to interconnect parts of the hypothalamus.
We shall now present a brief discussion of the position and composition of each
commissure, with special reference to the hypothalamic component of each.
W hat generally is referred to as the supraoptic commissure largely corresponds
to the supraoptic commissure of Gudden, which is composed of thick, myelinated
fibers in a discrete bundle at the ventral surface of the hypothalamus and intermixed
w ith the optic chiasm (Krieg, 1932; Tsang, 1940). This commissure mostly contains
axons projecting to regions in and around the superior coUiculus. Dorsal to, and
88
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
extending caudal to, Gudden's commissure are the supraoptic commissures of
M eynert and Ganser, which are thought to convey axons to the basal ganglia and
ventral thalamus in addition to the dorsal midbrain. Despite these major conclusions,
evidence has been presented that the supraoptic commissures contain hypothalamic
fibers. More dorsally , and extending farther caudally, Tsang (1940) described a
small periventricular commissure containing a few myelinated axons from medial
zone nuclei and crossing the midline only at their level of origin to innervate their
contralateral fellows. In addition, Szentagothai et al. (1968) reported degenerating
fibers in the supraoptic comm issures following lesions of the hypothalam ic
ventromedial nucleus. However, N auta and Haym aker (1969) reported that this
observation could not be confirm ed. Cajal (1995) illustrated th e supraoptic
commissure of M eynert (fig. 322, p.409 vol. H) in the retrochiasmatic region. He
suggested that m any of the fibers in this commissure arise from cells within this
region. More recently, using the PHAL method, it has been shown that each of the
three major m edial zone nuclei (medial preoptic, anterior hypothalamic, and
ventromedial) sends fibers through the supraoptic commissures (Simerly and
Swanson, 1988; Risold et al., 1994; Ganteras et al., 1994). In addition, the DMH, as
reported here, as w ell as the paraventricular (Luiten et al., 1985; L.W. Swanson,
unpublished observations), posterior hypothalamic (Vertes et al., 1995), and ventral
premammillary nuclei (Ganteras et al., 1992b) also use this commissure.
Similarly, there has been little previous evidence for axons crossing in the preoptic
region. Magoun and Ranson (1942) observed fibers crossing the midline at the dorsal
89
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
border of rostral levels of the optic chiasm in cat and monkey, but concluded that
this commissure did not contain hypothalamic elements and that it was not a part
of the supraoptic commissure system. In the opossum. Loo (1931) described fibers
crossing just dorsal to the optic chiasm at the level of the optic recess - the most
rostral extent of the median preoptic nucleus (Swanson, 1992). He considered this
commissure an extension of the hypothalamic periventricular fiber system of N auta
and Haymaker (1969) and identified fibers in it from nuclei that we now call m edian
preoptic, periventricular preoptic, and preoptic suprachiasmatic. This observation
has since b een confirm ed and extended. In several PH AL studies of the
hypothalamus, axons have been seen to cross throughout the length of the m edian
preoptic nucleus. These nuclei include the m edial preoptic, anterior hypothalamic,
DMH, and ventromedial nuclei (Simerly and Swanson, 1988; Risold et al., 1994;
Ganteras et al., 1994). In addition, other small nuclei of the preoptic region, such as
the parastrial, anterodorsal periventricular, and anteroventral periventricular nuclei,
innervate the contralateral preoptic area via this route (Simerly and Swanson, 1988;
R.H. Thompson and L.W. Swanson, unpublished observations).
We suggest that this preoptic commissure is related to the fibers that we observed
crossing in rostral parts of the hypothalamic paraventricular nucleus (see Results).
Nauta and Haymaker (1969) also observed w hat appears to be this same commissure
in the human brain. They interpreted this pathw ay as an anomalous (occurring in
approxim ately two per cent of hum an brains) com ponent of the supraoptic
com m issure of Ganser, and noted that these fibers enter the contralateral
90
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
paraventricular nucleus, but could follow them no farther. We propose that, although
an occasional fiber was observed in the anterior commissure in horizontal sections,
these fibers are part of a preoptic commissure. We have show n that this commissure
is more common in the rat (present in approximately sixty per cent of our horizontal
experiments) and has a distribution similar to other elements of the preoptic
commissure.
The DMH also iimervates the contralateral periventricular zone, including the
contralateral paraventricular nucleus and DMH. These projections are formed by
axons passing continuously dorsal to the third ventricle and then turning ventrally
upon crossing the midline. Both Gurdjian (1927) and Krieg (1944) describe a system
of commissures crossing the thalamus, but do not refer to any fibers crossing in
ventral parts. This is likely due to technical lim itations. Both reached their
conclusions on the basis of myelin-stained material and thus may have m issed the
majority of crossing unmyelinated fibers. Again, more recent studies w ith PHAL,
which labels unmyelinated axons, have shown that, in addition to the DMH, the
anterior hypothalamic (Risold et al., 1994), ventromedial (Ganteras et al., 1994),
and posterior hypothalamic nuclei (Vertes et al., 1995.) all send fibers through ventral
parts of the thalamic (middle or gray) commissure immediately dorsal to the third
ventricle to innervate the contralateral hypothalamus.
The DMH also sends fibers across the midline in and around the area of the
mammillary recess. This area has been called the premammillary commissure by
91
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Risold et al. (1994) because a large number of fibers from the anterior hypothalamic
nucleus cross the midline here to innervate the contralateral dorsal premammillary
nucleus. These fibers constitute the largest hypothalam ic com ponent of this
commissure by far, but additional fibers arise in the DMH and ventrom edial and
tuberal hypothedamic nuclei (Ganteras et al., 1994).
The supramammillary decussation contains crossing myelinated fibers readily
apparent in Nissl preparations, when viewed with darkfield illumination. It is
located dorsal to the mammillary nuclei and courses through the supramammillary
nucleus (Swanson, 1992). Traditionally, it has been thought to convey fibers from
the m idbrain and little mention can be found of hypothalamic fibers crossing in
this region (Gurdjian, 1927; Nauta and Haymaker, 1969). Recently, PHAL studies
indicate that virtually all nuclei of the hypothalamus, including the three major
medial zone nuclei (Simerly and Swanson, 1988; Risold et al., 1994; Ganteras et al.,
1994), and paraventricular nucleus (L.W. Swanson, unpublished observations), in
addition to the DMH, send a small crossed projection through the supramammillary
nucleus and ventral parts of the posterior hypothalamic nucleus.
Com parison with other M edial Zone Nuclei
This study marks the completion of a PHAL analysis of nuclei in and around
the hypothalamic medial zone (excluding the mammillary complex) by our labora
tory. Thus, we are in an excellent position to compare DMH projections w ith "classic"
medial zone nuclei (e.g., m edial preoptic, anterior hypothalamic, and ventrome-
92
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
dial, as sum m arized recently by Risold et al., 1994) in an effort to generalize about
the possible functional significance of the DMH. At the same time, these consider
ations clarify the overall structural organization of the hypothalamus.
The prim ary conclusion that emerges from a comparison of the three definitive
medial zone nuclei listed above is that they all use a restricted set of pathw ays to
innervate their targets and that this set is strikingly different from that used by the
DMH—which seems to have more in common w ith periventricular zone nuclei
than medial zone nuclei.
The intrahypothalamic pathways described here have been named on the basis
of their location and are referred to as periventricular, medial, ventral, dorsal and
lateral pathw ays (Fig. 7). The location and course of the periventricular, ventral
and lateral pathw ays have been described in considerable detail, because they are
the major pathw ays used by the DMH (see Results). Briefly, the periventricular
pathway courses adjacent to, and is coextensive with, the third ventricle in the
periventricular zone. Other medial zone nuclei that utilize this pathway include
the medial part of the medial preoptic nucleus, various other nuclei of the preoptic
area (Simerly and Swanson, 1988), and the ventral premammillary nucleus (Gan
teras and Swanson, 1992). The ventral pathway travels ventral to the m edial zone
along its length and is especially prominent in the retrochiasmatic area and ventral
preoptic region (e.g., see Fig. 4 D-K). Rostrally, this pathway enters the preoptic
region and ends at the level of the lamina terminalis. Caudally, it can be followed
93
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
adjacent to the ventral surface of the brain into premammillary levels and m ay be
the source of fibers th at end in the capsule of the m am m illary body. This path
way is used by all m edial zone nuclei except the dorsal prem am m illary nucle
us. The m edial preoptic (Simerly and Swanson, 1988), ventrom edial (Ganteras
et al., 1994), and ventral prem am m illary (Ganteras and Swanson, 1992) nuclei
contribute most heavily, whereas the dorsom edial and anterior hypothalam ic
(Risold et al., 1994) nuclei both send m oderate num bers of axons through this
route.
The lateral pathw ay is also used to varying degrees by all m edial zone nu
clei, ranging from an occasional fiber from the dorsal prem am m illary nucleus
(Ganteras and Swanson, 1992) to a dense pathw ay originating in the m edial
preoptic nucleus, especially its lateral part (Simerly and Swanson, 1988). This
path is largely restricted to m edial and ventrom edial regions of the lateral hy
pothalam ic area, and dorsally fibers appear to be concentrated im m ediately lat
eral and dorsal to the fomix. This pathw ay runs along the extent of the lateral
hypothalamic area. Rostrally, it turns laterally into the preoptic region where it
joins the ventral and periventricular paths. Gaudally, it extends into the ventral
tegm ental area, w here, in the case of the m edial preoptic nucleus, it generates a
dense term inal field (Simerly and Swanson, 1988). Axons from other m edial
zone nuclei turn m edially at this level to provide a m odest input to the supra
m am m illary nucleus.
94
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 7. Schematic representation of the projections of the medial
zone nuclei of the hypothalamus at tuberal levels. Gray shading
indicates the area through which each pathway travels. Each path
way is present at all levels of the hypothalamus, but summarized
at this level only (see text).
95
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The dorsal pathw ay lies in dorsal regions of the hypothalam us and ventral
regions of the zona incerta. Rostrally, this pathw ay appears to end at the level
of the rostral zona incerta, w here the fibers either ascend to the thalam us or
descend to end in the anterior perifom ical region and the caudodorsal preoptic
region. Caudally, this pathw ay extends through the posterior hypothalam ic
nucleus w here it turns dorsally and then caudally to enter the brainstem . This
pathw ay is m ost extensively used by the dorsal prem am m illary (Ganteras and
Swanson, 1992) and ventrom edial (Ganteras et al., 1994) nuclei, but also contains
m oderate num bers of fibers from the anterior hypothalam ic nucleus (Risold et
al., 1994).
The m edial pathw ay travels through the body of the m edial zone dorsal to
the ventral pathw ay. The only tw o hypothalam ic nuclei utilizing extensively
the m edial pathw ay are the anterior hypothalam ic and ventrom edial nuclei.
Both lie rostrocaudally adjacent to one another and share dense bidirectional
connections. However, it is considered a separate pathw ay m ainly because a
vast majority of the descending input to the hypothalam us from the am ygdala
travels through it. Most fibers in the stria term inalis pass rostral to the anterior
commissure, descend, and tu rn caudally (Heim er and N auta, 1969) to form a
circumscribed pathw ay through the medial preoptic, anterior hypothalam ic,
and ventrom edial nuclei before ending m assively in the ventral prem am m illary
nucleus. The DM H does not receive an input from the am ygdala (Ganteras et
al., 1992a; Ganteras et al., 1995).
96
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Functional Considerations
The DMH is a very enigmatic nucleus. As outlined in the introduction, it has
been implicated in a wide variety of responses, although it is m ost often associated
with the 'T)MH syndrome," i.e., disturbances in ingestive behavior (Bemardis, 1972;
Dalton et al., 1981; also see Bem ardis and Bellinger, 1987 for review) and the
descending control of autonomic m otor function (Luiten et al., 1987). Indeed, this
predom inant view led Ter Horst and Luiten (1986) to interpret their results in terms
of feeding control and pancreatic horm one release. H ow ever, because the
intrahypothalam ic projections of the DMH are m uch larger than those to the
brainstem, and because the DMH does not generate significant direct projections
to autonomic motor nuclei, it is difficult to accept this as a prim ary function of the
DMH. Nevertheless, changes in pancreatic nerve activity (Yoshimatsu et al., 1984;
Oomura and Yoshimatsu, '84) and alterations in feeding behavior occur after DMH
lesions and it remains to be determined how these changes are effected. It should
also be noted that, while the DMH has limited descending projections, it does send
moderate to dense projections to other regions that in turn contain at least some
neurons that project to autonomic preganglionic regions. These areas include the
paraventricular nucleus, the retrochiasmatic and lateral hypothalamic areas, and
the BST (Saper et al., 1976; Swanson and Kuypers, 1980; Schwanzel-Fukada et al.,
1984; Holstege et al., 1985). It is well-established that the paraventricular nucleus
has extensive descending connections and is involved in regulating ingestive
behavior (see Swanson, 1987). The lateral hypothalam ic area also has well-
97
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
established descending projections and is also thought to be important for feeding
and behavioral activation (Grossman et al., 1978; Strieker et al., 1978).
It is im portant to note that the paraventricular nucleus and lateral hypothalamic
area are involved in drinking as well as feeding behavior . Indeed, although
supporting functional data is lacking, the most obvious role for the DMH suggested
by our anatomical results is in fluid balance and drinking behavior. One of the
responses to a drop in blood volume is increased levels of circulating angiotensin
II. This horm one has a variety of actions, including vasopressin release, a
sympathetic pressor response, and the initiation of drinking behavior. Lesions of
the subfornical organ, or its projections, block these effects (see Swanson, 1987).
The m ajority of subfornical organ projections end in the paraventricular and
supraoptic nuclei, m edian preoptic, and parastrial nuclei, although other terminal
fields include the anteroventral periventricular nucleus, DMH, and adjacent
perifomical region (Lind et al., 1984; Swanson and Lind, 1986). It is known that
each of these subfomical-redpient nuclei projects to the DMH (Simerly and Swanson,
1988; Sawchenko and Swanson, 1983). In addition, our results show that the DMH
projects densely to these areas and sparsely to the subfornical organ itself. Therefore,
it is difficult to avoid the conclusion that the DMH influences the output of the
subfornical organ, and thus influences body fluid homeostasis. Functionally it is
known that rats with lesions of the DMH are hypodipsic (Bemardis, 1972;, Dalton
et al., 1981; Bellinger and Bemardis, 1982). However, these rats have been shown to
be either equally responsive or more sensitive to both osmotic and volemic thirst
98
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
challenges (Dalton et al., 1981; Bellinger and Bemardis, 1982). This result is not
surprising, however, because the subfornical organ sends a massive projection to
the median preoptic and parastrial nuclei, both of which project directly to the
paraventricular nucleus. What this does indicate is that the role of the DMH in
body fluid homeostasis is not sim ple and may depend on other aspects of the
animal's physiological state. These aspects may include variables such as the fasted/
fed state of the animal, point in the reproductive cycle, or behavioral state, vis a vis
the sleep-wake cycle. As outlined in the Introduction, there is some evidence for
each, although it remains to be determined w hether any by itself provides an
organizing theme for DMH function, or whether the DMH allows each to provide
a conditional modulatory influence, or a context within which it exerts a modulatory
influence, on eating and drinking.
99
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LITERATURE CITED
Allen, G.V., and D.F. Cechetto (1992) Functional and anatomical organization of
cardiovascular pressor and depressor sites in the lateral hypothalamic area: I.
Descending projections. J. Comp. Neurol. 315:313-332.
A nderson, J.J., and J.A. Dimicco (1990) Effect of local inhibition of gamma-
aminobutyric acid uptake in the dorsomedial hypothalamus on extracellular
levels of gamma-aminobutyric add and on stress-induced tachycardia: a study
using microdialysis. J. Phamacol. Exp. Then 255:1399-1407.
Arnault, P., and M. Roger (1987) The connections of the peripedim cular area stud
ied by retrograde and anterograde transport m the rat. J. Comp. Neurol. 258:463-
476.
Barone, F.C., M.J. Wayner, S.L. Scharoun, R. Guevara-Aguilar, and H.U. Aguilar-
Baturoni (1981) Afferent connections to ffie lateral hypothalamus: A horseradish
peroxidase study in the rat. Brain Res. Bull. 7:75-88.
Barry, J., G.E. Hoffman, and S. Wray (1985) LHRH-containing system s. In A.
Bjorklund and T. Hokfelt (eds): Handbook of Chemical Neuroanatomy, Gaba
and Neuropeptides in the CNS: Part I. Amsterdam: Elsevier, Vol. 4, pp. 166215.
Behbehani, M.M. (1995) Functional characteristics of the m idbrain periaqueductal
gray. Prog. Neurobiol. 46:575-605.
Behzadi, G., P. Kalén, F. Parvopassu, and L. Wiklund (1990) Afferents to the median
raphe nucleus of the rat: Retrograde cholera toxin and wheat germ conjugated
horseradish peroxidase tracing and selective D-^H]aspartate labeling of pos
sible excitatory amino ad d inputs. Neurosdence 37:77-100.
Beitz, A.J. (1982) The organization of afferent projections to the m idbrain
periaqueductal gray of the rat. Neurosdence 7:133-159.
Beitz, A.J. (1989) Possible origin of glutamatergic projections to the m idbrain
periaqueductal gray and deep layer of the superior coUiculus of the rat. Brain
Res. BuU. 23:25-35.
Bellinger, L.L., and L.L. Bemardis (1982) Water regulation in weanling hypodipsic
dorsomedial hypothalamic-Iesioned rats. Am. J. Physiol. 242:R285-R295.
1 0 0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Bellinger, L.L., L.L. Bemardis, and V.E. Mendel (1976) Effect of ventromedial and
dorsom edial hypothalam ic lesions on circadian corticosterone rhythm s.
Neuroendocrinology 22:216-225.
BeUinger, L.L., V.E. Mendel, L.L. Bemardis, and T.W. Castonguay (1986) Meal pat
terns of rats w ith dorsomedial hypothalamic nuclei lesions or sham operations.
Physiol, and Behav. 36:693-698.
Bemardis, L.L. (1972) Hypophagia, hypodipsia, and hypoactivity following electro
lytic lesions in the dorsomedial hypothalamic nuclei of mature rats of both
sexes. J. Neural Trans. 33:1-10.
Bemardis, L.L., and L.L. Bellinger (1987) The dorsomedial hypothalamic nucleus
revisited: 1986 update. Brain Res. 434:321-381.
Buijis, R.M., A. Kalsbeek, T.P. Van der Woude, J.J. Van Heerikhuize, and S. Shinn
(1993) Suprachiasmatic nucleus lesion increases corticosterone secretion. Am.
J. Physiol. 264: R1186-R1192.
Cajal, S. Ramon y (1995) Histology of the Nervous System of M an and Vertebrates.
Translated by N. Swanson and L.W. Swanson. Oxford: Oxford University Press.
Ganteras, N.S., and L.W. Swanson (1992) The dorsal premammillary nucleus: An
unusual com ponent of the m am m illary body. Proc. Natl. Acad. Sci. USA
89:10089-10093.
Ganteras, N.S., R.B. Simerly, and L.W. Swanson (1992a) Gormections of the posterior
nucleus of the amygdala. J. Gomp. Neurol. 324:143-179.
Ganteras, N.S., R.B. Simerly, and L.W. Swanson (1992b) Projections of the ventral
premammillary nucleus. J. Gomp. Neurol. 324:195-212.
Ganteras, N.S., R.B. Simerly, and L.W. Swanson (1994) Organization of projections
from the ventromedial and tuberal nuclei of the hypofrialamus: A PHAL study
in the rat. J. Gomp. Neurol. 348:41-79.
Ganteras, N.S., R.B. Simerly, and L.W. Swanson (1995) Organization of projections
from the medial nucleus of the amygdala: A PHAL study in the rat. J. Gomp.
Neurol. 360:213-245.
Garlton, S.M., G.R. Leichnetz, E.G. Young, and D.J. Mayer (1983) Supramedullary
afferents of the nucleus Raphe Magnus in the rat: A study using the transcannula
HRP gel and autoradiographic techniques. J. Gomp. Neurol. 214:43-58.
101
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chen, S., and H.S. Su (1990) Afferent connections of the thalam ic paraventricular
and paiataenial nuclei in the rat-a retrograde tracing study w iüi iontophoretic
application of Fluoro-Gold. Brain Res. 522:1-6.
Chiba, T., and Y. M urata (1985) Afferent and efferent connections of the medial
preoptic area in the rat: a WGA-HRP study. Brain Res. Bull. 14:261-272.
Cornwall, J., and O.T. Phillipson (1988) Afferent projections to the dorsal thalamus
of the rat as shown by retrograde lectin transport. H. The m idline nuclei. Brain
Res. Bull. 21:147-161.
Dallman, M.F., A.M. Strack, S.F. Akana, M.J. Bradbury, E.S. H anson, K.A. Scribner,
and M. Smith (1993) Feast and famine: Critical role of glucocorticoids w ith in
sulin in daily energy flow. Frontiers Neuroendocrinol. 14:303-347.
Dalton, L.D., R.G. Carpenter, and S.P. Grossman (1981) Ingestive behavior in adult
rats with dorsomedial hypothalamic lesions. Physiology and Behavior 26:117-
123.
Dietrichs, E., L. Wiklund, and D.E. Haines (1992) The hypothalamocerebellar pro
jection in the rat: origin and transmitter. Arch. Ital. Biol. 130:203-211.
Fodor, M., Z. Csaba, C. Kordon, and J. Epelbaum (1994) Grow th hormone-releasing
hormone, somatostatin, galanin and 6-aidorphin afferents to the hypothalamic
periventricular nucleus. J. Chem. Neuroanat. 8:61-73.
Fuchs, S. A., H.M. Edinger, and A. Siegel (1985) The role of the anterior hypothalamus
in affective defense behavior elicited from the ventromedial hypothalam us of
the cat. Brain Res. 330:93-107.
Gallo, R.V. (1981) Effect of electrical stimulation of the dorsom edial hypothalamic
nucleus on pulsatile LH release in ovariectomized rats. Neuroendocrinology
32:134-138.
Gerfen, C.R., and RE. Sawchenko (1984) An anterograde neuroanatom ical tracing
m ethod that shows the detailed morphology of neurons, their axons and ter
minals: immunohistochemical localization of an axonally transported plant
lectin, Phaseolus vulgaris leucoagglutinin (PHA-L). Brain Res. 290:219238.
Grossman, S.P., A.E. Halaris, T. Collier, and A. Routtenberg (1978) Aphagia and
adipsia after preferential destruction of nerve cell bodies in hypothalamus.
Science 202:537-539.
1 0 2
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Grove, E.A. (1988) Neural associations of the substantia innominata in the rat: af
ferent connections. J. Comp. Neurol. 277:315-346.
Gunnet, J.W., and M.E. Freeman (1985) The interaction of the medial preoptic area
and the dorsomedial-ventromedial nuclei of the hypothalam us in the regulation
of the mating-induced release of prolactin. Neuroendocrinology 40:232-237.
Gurdjian, E.S. (1925) Olfactory connections in the albino rat, with special reference
to the stria meduUaris and the anterior commissure. J. Comp. Neurol. 38:127-
163.
Gurdjian, E.S. (1927) The diencephalon of the albino rat. J. Comp. Neurol. 45:249-
281.
Haberly, L.B., and J.L. Price (1978) Association and commissural fiber systems of
the olfactory cortex of the rat H. Systems originating in the olfactory peduncle.
J. Comp. Neurol. 181:781-808.
Hancock, M.B. (1976) Cells of origin of hypothalmo-spinal projections in the rat.
N eurosd. Lett. 3:179-184.
Heimer, L., and W.J.H. Nauta (1969) The hypothalamic distribution of the stria
terminalis in the rat. Brain Res.13:284-297.
Holstege, G., L. Meiners, and K. Tan (1985) Projections of the bed nudeus of the
stria terminalis to the mesencephalon, pons, and m edulla oblongata. Exp. Brain
Res. 58:379-391.
Horel, J.A., and D.J. Stelzner (1981) neocortical projections of the rat anterior com
missure. Brain Res. 220:1-12.
Hosoya, Y. (1980) The distribution of spinal projection neurons in the hypothalamus
of the rat, studied with the HRP method. Exp. Brain Res. 40:7987.
Hosoya, Y. (1985) Hypothalamic projections to the ventral medulla oblongata in
the rat, with spedal reference to the nudeus raphe pallidus: a study using auto
radiographic and HRP techniques. Brain Res. 344:338-350.
Hosoya, Y, and K. Kohno (1987) The topographical organization of neurons in the
dorsal hypothalamic area that project to the spinal cord or to the nucleus raphe
pallidus in the rat. Exp. Brain Res. 66:500-506.
10 3
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Hosoya, Y., and M. Matsushita (1981) A direct projection from the hypothalamus to
the area postrema in the rat, as demonstrated by the HRP and autoradiographic
methods. Brain Res. 214:144-149.
Hsu, SM ., and L. Raine (1981) Protein A, avidin and biotin in immunohistochemistry.
J. Histochem. Cytochem. 29:1349-1353.
Hsu, S.M., L. Raine, and H. Fanger (1981) The use of avidin-biotin-peroxidase com
plex (ABC) in immunoperoxidase techniques: A comparison between ABC and
unlabeled antibody (PAP) procedures. J. Histochem. Cytochem. 29:577-580.
Inglefield, J.R., S B. Schwarzkopf, and C.K. Kellogg (1994) Alterations in behav
ioral responses to stressors following excitotoxin lesions of dorsomedial hypo
thalamic regions. Brain Res. 633:151-161.
Imaki, T., J.L. Nahan, C. Rivier, PE. Sawchenko, and W. Vale (1991) Differential
regulation of corticotropin-releasing factor mRNA in rat brain regions by glu
cocorticoids and stress. J. N eurosd. 11:585-599.
Johansson, O., T. Hokfelt, and R.P. Elde (1984) Immunohistochemical distribution
of somatostatin-like immunoreactivity in the central nervous system of the adult
rat. Neurosdence 13:265-339.
Johnson, A.K. (1985) Role of the periventricular tissue surrounding the anteroventral
third ventride (AV3V) in the regulation of body fluid homeostasis. In R.W.
Schrier (ed): Vasopressin. N ew York: Raven Press, pp. 319-331.
Kalén, P., M. Karlson, and L. W iklund (1985) Possible exdtatory amino acid affer
ents to nudeus raphe dorsalis of the rat investigated with retrograde wheat
germ agglutinin and D-[3H]aspartate tracing. Brain Res. 360:285-297.
Kita, H., and Y . Oomura (1982a) A n HRP study of the afferent connections to rat
m edial hypothalamic region. Brain Res. Bull. 8:53-62.
Kita, H., and Y . Oomura (1982b) An HRP study of the afferent connections to rat
lateral hypothalamic region. Brain Res. Bull. 8:63-71.
Krieg, W.J.S. (1932) The hypothalam us of the albino rat. J. Comp. Neurol. 55:1989.
Krieg, W.J.S. (1944) The medial region of the thalam us of the albino rat. J. Comp.
Neurol. 80: 381-415.
104
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Lammers, J.H., M.R. Kruk, W. Meelis, and A M. van der Poel (1988) Hypothalamic
substrates for brain stim ulation-induced attack, teethchattering and social
grooming in the rat. Brain Res. 449:311-327.
Lechan, R.M., and I.M.D. Jackson (1982) Immunohistochemical localization of
thyrotropin-releasing horm one in the ra t hypothalam us and pituitary.
Endocrinology 111:55-65.
Levin, M.C., RE. Sawchenko, P.R. Howe, S.R. Bloom, and J.M. Polak (1987) Organi
zation of galanin-immunoreactive inputs to the paraventricular nucleus w ith
special reference to their relationship to catecholaminergic afferents. J. Comp.
Neurol. 261:562-582.
Lind, R.W., L.W. Swanson, and D. Ganten (1984) Angiotensin n immunoreactive
pathw ays in the central nervous system of the ra t evidence for a projection
from the subfornical organ to the paraventricular nucleus of the hypothalamus.
CHn. Exp. Hypertens., Part A 6:1915-1920.
Loo, Y.T. (1931) The forebrain of the opossum, Didelphis virginiana. J. Comp. Neurol.
52:1-148.
Luiten, P.G., and P. Room (1980) Interrelations between lateral, dorsomedial and
ventrom edial hypothalamic nuclei in the rat. An HRP study. Brain Res. 190:321-
332.
Luiten, P.G.M., F. Kuipers, and H. Schuitmaker (1982) Organization of diencephalic
and brainstem afferent projections to the lateral septum in the rat. Neurosci.
Lett. 30:211-216.
Luiten, P.G.M., G.J. Ter Horst, H. Karst, and A.B. Steffens (1985) The course of
paraventricular hypothalamic efferents to autonomic structures in medulla and
spinal cord. Brain Res. 329:374-378.
Luiten, P.G.M., G.J. Ter Horst, and A.B. Steffens (1987) The hypothalam us, intrinsic
connections and outflow pathways to the endocrine system in relation to the
control of feeding and metabolism. Prog. Neurobiol. 28:1-54.
M agoun, H.W., and W. Ranson (1942) The supraoptic decussations in the cat and
monkey. J. Comp. Neurol. 76:435-459.
M archand, J.E., and N. Hagino (1983) Afferents to the periaqueductal gray in the
rat. A horseradish peroxidase study. Neuroscience. 9:95-106.
105
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Marcinkiewicz, M., R. Morcos, and M. Chretien (1989) CNS connections w ith the
median raphé nucleus: retrograde tracing with WGA-apoHRP-Gold complex
in the rat. J. Comp. Neurol. 289:11-35.
Moga, M.M., H. Herbert, K.M. Hurley, Y . Yasui, T.S. Gray, and C.B. Saper (1990)
Organization of cortical, basal forebrain, and hypothalamic afferents to the
parabrachial nucleus in the rat. J. Comp. Neurol. 295:624-661.
Morrell, J.I., L.M. Greenberger, and D.W. Pfaff (1981) Hypothalamic, other dien
cephalic, and telencephalic neurons that project to the dorsal midbrain. J. Comp.
Neurol. 201:589-620.
N auta, W.J.H., and W. Haymaker (1969) Hypothalamic nuclei and fiber connec
tions. In W. Haymaker, E. Anderson and W.J.H. Nauta(eds): The hypothalamus.
Springfield: Charles C. Thomas, pp. 136-209.
Oom ura, Y, and H. Yoshimatsu (1984) Neural network of glucose m onitoring sys
tem. J. Autonom. Nerv. Syst. 10:359-372.
Ottersen, O P. (1980) Afferent connections to the amygdaloid complex of the rat
and cat: U. Afferents from the hypothalamus and the basal telencephalon. J.
Comp. Neurol. 194:267-289.
Palkovits, M., L. Zaborsky, A. Feminger, É. Mezey, M.I.K. Fekete, J.P. Herm an, B.
Kanyicska, and D. Szabo (1980) Noradrenergic innervation of the rat hypo
thalamus: Experimental biochemical and electron microscopic studies. Brain
Res. 191:161-171.
Pan, J.T., and R.R. Gala (1985) Central nervous system regions involved in the
estrogen-induced afternoon prolactin surge. I. Lesion studies. Endocrinology
117:382-387.
Paxinos, G., and C. Watson (1986) The Rat Brain in Stereotaxic Coordinates. Sydney;
Academic Press.
Risold, P.Y., N.S. Canteras, and L.W. Swanson (1994) Organization of projections
from the anterior hypothalamic nucleus: A PHAL study in the rat. J. Comp.
Neurol. 348:1-40.
Saper, C.B. (1985) Organization of cerebral cortical afferent systems in the rat. H.
Hypothalamocortical projections. J. Comp. Neurol. 237:21-46.
Saper, C.B., and D. Levisohn (1983) Afferent connections of the m edian preoptic
nucleus in the rat: Anatomical evidence for a cardiovascular integrative mecha-
106
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
nism in the anteroventral third ventricular (AV3V) region. Brain Res. 288:21-
31.
Saper, C.B., A.D. Loewy, L.W. Sw anson, and W.M. C ow an (1976) Direct
hypothalamo-autonomic connections. Brain Res. 117:305-312.
Satoh, K., N. Shimizu, M. Tohyama, and T. Maeda (1978) Localization of the mictu
rition reflex center at dorsolateral pontine tegmentum of the rat. Neurosd.Lett.
8:27-33.
Sawchenko, RE., and L.W. Swanson (1983) The organization of forebrain afferents
to the paraventricular and supraoptic nuclei of the rat. J. Comp. Neurol. 218:121-
144.
Schwanzel-Fukuda, M., J.I. Morrell, and D.W. Pfaff (1984) Localization of forebrain
neurons which project directly to the medulla and spinal cord of the rat by
retrograde tracing with wheat germ agglutinin. J. Comp. Neurol. 226:120.
Shammah-Lagnado, S.J., J.A. Ricardo, N.T.M.N. Sakamoto, and N. Negrao (1983)
Afferent connections of the mesencephalic reticular formation: A horseradish
peroxidase study in the rat. Brain Res. 9:391-409.
Simerly, R.B., and L.W. Swanson (1986) The organization of neural inputs to the
m edial preoptic nucleus of the rat. J. Comp. Neurol. 246:312-342.
Simerly, R.B., and L.W. Swanson (1988) Projections of the m edial preoptic nucleus:
a Phaseolus vulgaris leucoagglutinin anterograde tract-tracing study in the rat.
J. Comp. Neurol. 270:209-242.
Simerly, R.B., L.W. Swanson, and R.A. Gorski (1984) Demonstration of a sexual
dimorphism in the distribution of serotonin-immunoreactive fibers in the me
dial preoptic nucleus of the rat. J. Comp. Neurol. 225:151-166.
Sinnamon, H.M. (1984) Forelimb and hindlimb stepping by the anesthetized rat
elicited by electrical stimulation of the diencephalon and mesencephalon.
Physiol, and Behav. 33:191-199.
Sofroniew, M.V., and U. Schrell (1980) Hypothalamic neurons projecting to the rat
caudal m edulla oblongata, examined by immunoperoxidase staining of retro-
gradely transported horseradish peroxidase. Neurosci. Lett. 19:257-263.
Soltis, R.P., and J.A. Dimicco (1991) GABAa and excitatory amino acid receptors in
dorsomedial hypothalamus and heart rate in rats. Am. J. Physiol. 260:R13R20.
107
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Soltis, R.P., and J.A. Dimicco (1992) Hypothalamic excitatory amino acid receptors
m ediate stress-induced tachycardia in rats. Am. J. Physiol. 262:R689-R697.
Steffens, A.B., A.J. Scheurink, P.G. Luiten, and B. Bohus (1988) Hypothalamic food
intake regulating areas are involved in the homeostasis of blood glucose and
plasm a FFA levels. Physiol, and Behav. 44:581-589.
Strieker, E.M., A.F. Swerdloff, and M.J. Zigmond (1978) Intrahypothalamic injec
tions of kianic acid produce feeding and drinking deficits in rats. Brain Res.
158:470-473.
Stotz-Potter, E.H., L.R. Willis, and J.A. Dimicco (1996) Muscimol acts in dorsomedial
but not paraventricular hypothalamic nucleus to suppress cardiovascular ef
fects of stress. J. Neurosci. 16:1173-1179.
Swanson, L.W. (1986) Organization of mammalian neuroendocrine system. In RE.
Bloom (ed); H andbook of Physiology, The Nervous System, IV. Baltimore, MD:
Waverly Press, pp. 1-125.
Swanson, L.W. (1987)The Hypothalamus. In A. Bjorklund, T. Hokfelt and L.W.
Swanson (eds): Handbook of Chemical Neuroanatomy, Vol. 5 Integrated sys
tems of the CNS, Part I. Amsterdam: Elsevier Science, pp. 1-124.
Swanson, L.W. (1992) Brain Maps: Structure of the Rat Brain. Amsterdam: Elsevier
Science.
Swanson, L.W. (1993) Brain Maps: Computer Graphics Files. Amsterdam; Elsevier
Science.
Swanson, L.W., and B.K. Hartman (1975) The central adrenergic system. An immuno
fluorescence study of the location of cell bodies and their efferent connections
in the rat utilizing dopamine-B-hydroxylase as a marker. J. Comp. Neurol.
163:467-506.
Swanson, L.W., and H.G.J.M. Kuypers (1980) The parventricular nucleus of the
hypothalamus: cytoarchitectonie subdivisions and organization of projections
to the pituitary, dorsal vagal complex, and spinal cord as dem onstrated by ret
rograde fluorescence double-labeling methods. J. Comp. Neurol. 194:555570.
Swanson, L.W., and R.W. Lind (1986) Neural projections subserving the initiation
of a specific motivated behavior in the ra t new projections from the subfornical
organ. Brain Res. 379:399-403.
108
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Swanson, L.W., RE. Sawchenko, J. Rivier, and W.W. Vale (1983) Organization of
ovine corticotropin-releasing factor immunoreactive cells and fibers in the rat
brain: an immimohistochemical study. Neuroendocrinology 36:165-186.
Szentagothai, J., B. Flerko, B. Mess, and B. Halasz (1968) Hypothalamic control of
the anterior pituitary An experim ental-m orphological study. Budapest:
Akadémiai Kiado.
Ter Horst, G.J., H.J. Groenewegen, H. Karst, and P.G. Luiten (1984) Phaseolus vul
garis leuco-agglutinin immunohistochemistry. A comparison between auto
radiographic and lectin tracing of neuronal efferents. Brain Res. 307:379383.
Ter H orst, G.J., and P.G. L uiten (1986) The projections of the dorsom edial
hypothalamic nucleus in the rat. Brain Res. Bull. 16:231-248.
Ter Horst, G.J., and P.G. Luiten (1987) Phaseolus vulgaris leuco-agglutinin tracing of
intrahypothalamic connections of the lateral, ventromedial, dorsomedial and
paraventricular hypothalamic nuclei in the rat. Brain Res. Bull. 18:191203.
Ter Horst, G.J., PG M. Luiten, and E Kuipers (1984) Descending pathways from
hypothalam us to dorsal m otor vagus and am biguus nuclei in the rat. J.
Autonom. Nerv. Syst. 11:59-75.
Thompson, R.H., N.S. Ganteras, and L.W. Swanson (1992) The efferent cormections
of the dorsomedial nucleus of the hypothalamus: a PHAL study in the rat. Soc.
Neurosd. Abstr. 18:1415.
Tsang, Y.C. (1940) Supra- and post-optic commissures in ttie brain of the rat. J. Comp.
Neurol. 72:535-567.
Tucker, D C., and C.B. Saper (1985) Specificity of spinal projections from hypotha
lamic and brainstem areas which innervate sympathetic preganglionic neu
rons. Brain Res. 360:159-164.
Valentino, R.J., M E. Page, PH. Luppi, Y . Zhu, E. van Bockstaele and G. A ston-
Jones (1994) Evidence for w idespread afferents to Barrington's nucleus, a
brainstem region rich in corticotropin-releasing hormone neurons. Neurosdence
62:125-143.
van der Kooy, D., L.Y. Koda, J.E McGinty, C.R. Gerfen, and EE. Bloom (1984) The
organization of projections from the cortex, amygdala, and hypothalam us to
the nucleus of the solitary tract in rat. J. Comp. Neurol. 224:1-24.
109
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Veening, J., P. Buma, G J. Ter Horst, T.A.P. Reeling, PG M. Luiten, and R.
Nieuwenhuys (1991) Hypothalamic projections to the PAG in the rat: Topo
graphical, immuno-electronmicroscopical and functional aspects. In A. DepauHs
and R. Handler (eds): The Midbrain Periaqueductal Gray Matter, Functional,
Anatomical and Neurochemical Organization. New York: Plenum Press, pp.
211-235.
Vertes, R.P., A.M. Crane, L.V. Colom, and B.H. Bland (1995) Ascending projections
of the posterior nucleus of the hypothalamus: PHA-L analysis in the rat. J. Comp.
Neurol. 359:90-116.
Wada, H., N. Inagaki, A. Yamatodani, and T. Watanabe (1991) Is the histaminergic
neuron system a regulatory center for whole-brain activity? TINS 14:415-418.
Weller, K.L., and D A. Smith (1982) Afferent connections to the bed nucleus of the
stria terminahs. Brain Res. 232:255-270.
Wyss, J.M., L.W. Swanson, and W.M. Cowan (1979) A study of subcortical afferents
to the hippocampal formation in the rat. Neuroscience 4:463-476.
Yoshimatsu, H., A. Niijima, Y . Oomura, K. Yamabe, and T. Katafuchi (1984) Effects
of hypothalamic lesion on pancreatic autonomic nerve activity in the rat. Brain
Res. 303:147-152.
1 1 0
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter III.
Organization of Projections from the Dorsomedial Nucleus of
the Hypothalamus: A Reexamination with Fluorogold and PHAL
in the Rat
The dorsom edial nucleus of the hypothalam us (DMH) has been im plicated
in a variety of responses, including ingestive (Bemardis, 1972; Dalton et al.,
1981, W ang, et al., 1996); reproductive (Gallo, 1981; G unnet and Freeman, 1985;
Polston and Erskine,1995; Coolen et al.,1996; also see G unnet and Freeman, 1983
and Erskine, 1995 for reviews); endocrine, autonom ie, and behavioral aspects
of stress (Keim and Shekhar, 1996; Stotz-Potter et al., 1996a,b; Inglefield et al.,
1994; Shekhar, 1993; Shekhar and Katner, 1995; Kalsbeek et al., 1996a); circadian
(Bellinger et al., 1976,1986; Buijs et al., 1993; Kalsbeek et al., 1992, 1996b); and
therm ogenic (McCarthy et al., 1993; Sagar et al., 1995; Elm quist et al., 1996).
C urrent understanding of neural inputs, w hich the DMH presum ably integrates
to produce these responses, is based prim arily on two retrograde tracer studies
in the rat w ith the horseradish peroxidase (HRP) technique (Berk and Finkelstein,
1981a; Kita and Oom ura, 1982), although tw o additional studies used HRP
injections in the DMH to examine subsets of its projections (Fahrbach et al.,
1984; Chiba and M urata 1985).
I l l
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The present reexam ination of DMH inputs w ith Fluorogold (FG) retrograde
axonal tracing and Phasiolus vulgaris leucoagglutinin (PHAL) anterograde axonal
tracing techniques was undertaken for several reasons. First, although the above
s tu d ie s in d ic ate th a t th e DMH receives in p u ts from all reg io n s of the
hypothalam us, and especially densely from the preoptic region (Kita and
O om ura, 1982; Chiba and M urata, 1985), bed nuclei of the stria term inalis (BST),
lateral septal nucleus, periaqueductal gray, and parabrachial nucleus (Berk and
Finkelstein, 1981a; Kita and Oomura, 1982), they are not in complete agreem ent.
Specifically, inputs from the ventral subiculum, lateral habenula, peripeduncular
nucleus, nucleus of the solitary tract, and ventrolateral m edulla w ere reported
by Berk and Finkelstein (1981a), but not by Kita and Oom ura (1982). Second,
because these studies em ployed relatively large injections of HRP, w hich can be
taken up and transported by fibers-of-passage, it is unclear which pathw ays
innervate specifically the DMH, and which project to surrounding regions or
pass through w ithout terminating. One obvious exam ple is the in p u t from the
m edial nucleus of the amygdala reported by Berk and Finkelstein (1981a) that
has been show n recently to project densely to the capsule surrounding the
hypothalam ic ventrom edial nucleus, but not to the DM H (Ganteras et al., 1995).
Finally, the present results are illustrated in more detail than in earlier work,
and are sum m arized on a standard series of tem plates that facilitate com parison
w ith other experim ents that are plotted on the sam e tem plates. Also, because
these draw ings are generated with com puter graphics applications in a standard
112
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
file format, they can be incorporated eventually into electronic databases (Dashti
et al., in press)
Possible in p u ts to the D M H were stu d ie d first using the fluorescent
retrograde tracer FG, and identified cell groups were then injected with PHAL
to examine the distribution of labeled axons in and around the DMH. FG is
very sensitive, and appears to be taken up less by fibers-of-passage than other
retrograde tracer m ethods (Schmued and Fallon, 1986). Moreover, an antibody
to FG is now a v a ila b le , a llo w in g it to be labeled b y c o n v e n tio n a l
immunohistochemical m ethods. This provides a light-stable reaction product
w ith very low background, enabling a very precise localization of labeled
neurons because they can be com pared to adjacent Nissl-stained sections. The
PHAL method also has several distinct advantages. It labels axons, including
te rm in a l b o u to n s an d b o u to n s-o f-p a ssa g e , w ith the c la rity of G olgi
im pregnations, is not taken u p by fibers of passage, and is an exclusively
anterograde tracer (Gerfen and Sawchenko, 1984). In addition to discrim ination
betw een passing and term inating axons, these features allow detailed analysis
and representation of intranuclear fiber distribution.
From this w ork we conclude that the m ajority of inputs to the DMH arise in
the hypothalam us, although there are a few significant projections from the
telencephalon and brainstem . Each major nucleus and area of the hypothalam us
provides inputs to the DMH, except for the m agnocellular preoptic nucleus.
113
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m agnocellular neurons in the paraventricular and supraoptic nuclei, and the
m am m illary body. Telencephalic inputs arise m ainly in the ventral subiculum ,
infralimbic area of the prefrontal cortex, lateral septal nucleus and BST. The
m ajority of brainstem inputs arise in the periaqueductal gray, parabrachial
nucleus and ventrolateral m edulla. In addition, it now seem s clear that inputs
to the DM H use only a few, discrete pathways. Descending inputs course through
a periventricular pathw ay through the hypothalamic periventricular zone, a medial
pathw ay th at follows the m edial coriticohypothalam ic tract, and a lateral
pathw ay traveling through m ed ial p arts of the m edial forebrain bundle.
Ascending inputs arrive through a midbrain periventricular pathw ay that travels
adjacent to the cerebral aqueduct in the periaqueductal gray, and through a
brainstem lateral pathw ay that travels through central and ventral m idbrain
tegm ental fields and enters the hypothalam us and then DM H from more lateral
parts of the medial forebrain bundle.
MATERIALS AND METHODS
Injections of FG (Fluorchrome Inc.) were m ade into the region of the DMH
in 19 a d u lt m ale S p rague-D aw ley ra ts (250-300g). A n tero g rad e tracin g
experim ents were perform ed in 11 rats w ith PHAL (Vector Laboratories). In all
experiments, animals were anesthetized w ith a mixture of ketamine and xylazine
(v/v; Im l/k g body weight).
114
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In FG experiments, a 3.0% solution of the tracer w as prepared in O.OIM KPBS,
and w as injected iontophoretically through a glass m icropipette (tip diameter:
20-25|im) into the DMH. For injections, a positive current (5 p,A, 7s o n /7 s off)
was applied for 2-5 min. Similar procedures were used for the injection of a
2.5% solution of PHAL into the parabrachial nucleus (n=5) or ventrolateral
m edulla (n=6), although the micropipette tip diam eter w as 10-15|im and injection
time w as 20 min. A dditional animals received injections of PHAL that were
centered in one of the following: the BST, lateral septal nucleus, or nuclei of the
preoptic area. The injection in the lateral septal nucleus has previously been
published in p art (Risold and Swanson, in press) and was provided by P.Y.
Risold. Injections of the m edial part of the m edial preoptic nucleus and preoptic
region (parastrial, anterodorsal preoptic, anteroventral preoptic, anteroventral
periventricular, and m edial part of the m edial preoptic) have been describe
previously in p art by Simerly and Swanson (1988), and were provided by Dr.
Richard Simerly. The injection in the m edian preoptic nucleus w as provided by
Dr. G. Gu.
A fter a survival time of 7 days for FG experiments and 10-14 days for PHAL
experim ents, the anim als were deeply anesthetized w ith pentobarbitol, and
perfused transcardially w ith 150 ml of 0.9% NaCl, followed by 300 ml of 4%
paraform aldehyde in O.IM borate buffer (pH 9.5) at 4°C. The brain s were
rem oved and postfixed overnight at 4°C in the same paraform aldehyde solution
115
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
plus 10% sucrose. Frozen 30 jim thick sections were collected in a one-in-four
series through the forebrain and a one-in-six series through the brainstem .
Im m unodetection of PH A L and FG w as perform ed using a procedure
described in detail elsewhere (Ganteras and Swanson, 1992; Ganteras et al., 1992).
Briefly, sections were processed for im m unohistochem istry w ith an antiserum
against FG (Ghem icon International Inc., dilution 1:5000) or PH A L (Dako
Laboratories, dilution 1:1000), and then a solution containing avidin-biotin-HRP
com plex (ABG Elite Kit, Vector Laboratories). S taining w as obtained by
processing the peroxidase histochem istry w ith a solution containing 0.05%
diam inobenzidine and 0.01% hydrogen peroxide. The sections w ere then
m ounted on gelatin-coated slides and dried overnight. The slides were then
dehydrated and coverslipped w ith DPX. An adjacent series of Nissl-stained
sections was alw ays prepared for cytoarchitectonie purposes.
Sections w ere exam ined under the microscope w ith bright- and darkfield
illum ination. FG-labeled neurons and PHAL-labeled axons were plotted onto a
series of standard draw ings of the rat brain (Swanson, 1992) w ith the aid of a
computer (Apple Power M acintosh 7100; Adobe Illustrator 6). In FG illustrations,
one dot represents one cell. H ow ever, because m ultiple sections w ere needed to
construct each atlas level, d u e prim arily to plane-of-section constraints, all cell
num bers are approxim ate. Therefore, each level represented serves to illustrate
the presence or absence of labeling, the anatomical distribution and pattern of
116
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
neuronal labeling, and the relative strength of a projection. Unless otherwise
indicated, nomenclature follows Swanson (1992). Terminology for describing the
morphology of PHAL-labeled axons has been discussed in Risold et al. (1994).
RESULTS
Injection Sites
In 6 of the 19 anim als injected, the FG deposit was restricted prim arily w ithin
the borders of the DMH. H ow ever only two of the six did not involve dorsal
parts of the capsule of the ventrom edial hypothalamic nucleus. The injection
site in experim ent 17 prim arily involved the anterior and posterior parts, and
to a lesser extent the ventral part (see Thompson et al., 1996 for parcellation of
the DMH), but did not im pinge upon the capsule of the ventrom edial nucleus
(Fig. 1). Because experim ent 17 had very low background, little am ygdalar
labeling (which was taken to be indicative of limited ventral contam ination)
and the distribution of labeled neurons was representative of all DMH injections,
only this case is represented in detail.
By way of introduction, the results of our retrograde and anterograde axonal
transport experiments indicate that the DMH receives input from geographically
w idespread areas in the forebrain and brainstem. With few exceptions, each
region of the h y p o th alam u s p ro v id es a sig n ifican t in p u t to the DM H.
Telencephalic inputs are mainly restricted to the ventral subiculum, lateral septal
117
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
B#
i - m
m
r ^ . W - f e v > J
Fig. 1. Brightfield photomicrographs of Fluorogold injection site (A-C) and cau-
dally adjacent thionin-stained sections (A'-C') in Case 17. The approximate area
of the injection site illustrated in panel B is plotted in Fig. 2H.
118
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
nucleus, an d bed nuclei of the stria terminalis (BST), w hereas brainstem inputs
originate prim arily in the ventrolateral m edulla, parabrachial nucleus, and
periaqueductal gray.
DM H Inputs
M th the few exceptions noted, our results are in general agreement with previous
studies based on the HRP m ethod (Berk and Finkelstein, 1981a; Chiba and M urata,
1985; Kita and Oomura, 1982; Fahrbach et al., 1984). However, inputs to the DMH
in this earlier work were not were not described comprehensively or in detail.
Therefore, w e have thoroughly m apped the sources of inputs to the DMH based on
FG injections (Figs. 2 and 3), m apped the distribution of labeled axons w ithin the
DMH after PHAL injections of major sources of inputs (Figs. 4 and 5), and compared
these findings with those in the literature. Inputs to the DMH, along w ith the
pathway utilized by each input, are summarized in Figure 6.
Descending Inputs
Hypothalamus. The m ajo rity of pro jectio n s to the D M H arise in the
hypothalam us. Moreover, each major nucleus or area of the hypothalam us
reliably co n tain ed at least a few retrogradely labeled neurons, w ith the
exceptions of the m agnocellular preoptic nucleus, m agnocellular parts of the
hypothalam ic paraventricular nucleus, supraoptic nucleus, and m edial and
lateral m am m illary nuclei.
119
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 2. D istribution of Fluorogold-labeled neurons from experim ent 17 plotted
onto a series of standard tem plates of the rat brain taken from the atlas of
Swanson (1992). Drawings are arranged from rostral (A) to caudal (Q). N um
bers in the upper right com er refer to the corresponding atlas level of Swanson
(1992). The large red area in H indicates the approximate limits of the Fluorogold
injection site illustrated in Fig. IB.
120
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
VL
L S i
MS
al
aco j
d m
tPO
• • •
v 3tr
och
20
LSd L S d
V L
V L
fcp r LS
SF
M S
o v
cc
BST
aco
aoo
sc
tei
kS'Vj
och J
IPO
• M P O
LPO
och
'SO
121
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2 (continued)
26
P V T
SFO
RE
I? * '
BST . t r
’ M ir...
LHA
• LHA
: s o
Ç C H
opt
M E
intt
30 27
. p h W mtt
rr6d
m
AHNp
• LHA
C .ÎH A
M E
122
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2 (continued)
34
32
L H
FAG
PH
mtt
V T A
LHA
LHA
p m
mp
MM
LM
V3m
44
P A G
OP
C U N
SO
PPN
VTA
CSm
CSI
PR N r
IPNi
123
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 2 (continued)
'48
47
CUN
AQ
m£v
CUN V 4
L O T
U 3T
PBic
O T N
P g m m
PPN
M l
PPN
CS c a
CSm
PRNr
PRNr
66
GRN
V4
LDT
OTN
70
AP
M ORN v
124
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 3. Dark- and brightfield photomicrographs to illustrate the appearance of ret-
rogradely-labeled neurons following Fluorogold injection in the DMH (experiment
17). All level references are to the atlas of Swanson (1992) and correspond to a level
also represented in Figure 2. (A), the lateral septal nucleus preoptic region (40X),
just rostral to atlas level 17; (B), preoptic region, including the p arastrial,
anteroventral periventricular, and median preoptic nuclei (40X), and approximately
corresponding to atlas level 19; (C), anterior hypothalamic area (40X), including
the suprachiasmatic nucleus and anterior parvicellular part of the hypothalamic
paraventricular nucleus, approxim ately corresponding to atlas level 22; (D),
subfornical organ (lOOX), also corresponding to atlas level 22; (E), parabrachial
nucleus (40X), corresponding to a level intermediate to atlas levels 47 and 48. La
beled neurons are predominately seen in distinct groups in the central lateral, dorsal
lateral and superior lateral parts; (F), caudal tuberal hypothalam us (40X), includ
ing the dorsal tuberamammillary and ventral premammiUary nuclei, corresponding
to atlas level 32; and (G), high power brightfield photomicrograph of neurons in
the ventral part of the medullary reticular nucleus (200X) approximately correspond
ing to atlas level 70.
125
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
126
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The preoptic region contained the largest num ber of labeled neurons in the
hypothalam us projecting to the DMH. W ithin the preoptic region, each nucleus
and area reliably contained retrogradely labeled neurons. The single greatest
density of cells was seen in the parastrial nucleus (Fig. 2B-D; Fig. 3B). Num erous
cells were also observed in the preoptic periventricular, posterodorsal preoptic,
and caudal region of the preoptic suprachiasm atic nuclei. Moderate num bers
w ere lab eled in the m ed ial p a rt of the m ed ial p re o p tic , a n te ro v en tral
p eriv en tricu lar, m edian preoptic, an tero v en tral p reo p tic (including the
ventrom edial and ventrolateral preoptic regions of Sherin et al., 1996 and
Scammel et al., 1996, respectively), and anterodorsal preoptic nuclei; and the
medial preoptic area (Fig. 2B-E; Fig. 3B).
This is in general agreem ent w ith previous retrograde tracer studies (Kita
and Oom ura, 1982; Fahrbach et al., 1984; Chiba and M urata, 1985), although
these authors did not m ap this region in detail. Sw anson (1976) m ade small
injections of tritiated am ino acids into the m edian preoptic and preoptic
suprachiasm atic nuclei. This study indicated that the preoptic suprachiasm atic
nucleus projects to the DMH, although a projection from the m edian preoptic
nucleus w as not reported.
Simerly and Swanson (1988) injected PHAL into the m edial preoptic nucleus
and reported that the m edial part sends a significant projection to the DMH,
whereas only a few axons w ithin the DMH were illustrated following injections
127
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
centered in the central and lateral p arts of the m edial preoptic nucleus. The
present results generally agree w ith these findings. N um erous cells were
observed in the m edial p art of the m edial preoptic nucleus, and fewer, scattered
cells w ere labeled in the lateral p art (Fig. 2D-E). However, virtually no cells
were labeled in the central part, indicating that fibers observed in the DMH
after PHAL injections preferentially b u t not exclusively confined to the central
part probably arise in the medial p art of the m edial preoptic nucleus. They also
reported that the anteroventral periventricular, parastrial, anterodorsal, and
anteroventral nuclei project heavily to the DMH. Gu and Simerly (in press)
extended these findings to female rats. Follow ing PHAL injections in the
anteroventral periventricular nucleus, they reported a substantial projection to
the DM H in the female w ith a distribution sim ilar to that observed in the male.
Interestingly however, they observed m ore labeled fibers in the posterior part
(see below). In this study, Gu and Sim erly (in press) also m ade discrete PHAL
injections in the vascular organ of the lam ina term inalis, m edian preoptic
nucleus, and m edial p a rt of the m edial preoptic nucleus in the female. After
each injection they reported m oderate to dense axonal labeling in the DMH.
Using the control injections of Sim erly and Swanson (1988; Fig. 4A-C and G-
O), the projections of the preoptic nuclei to the DMH have been m apped onto
the atlas tem plates of Swanson (1992) at three levels representing the rostral to
caudal extent (including all parts) of the DMH (Fig. 5). An additional control
128
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 4. Cam era lucida plots of PHAL-labeled neurons following injections cen
tered in the parastrial (PS), m edian preoptic (MePO), anteroventral periven
tricular (AVPv), anteroventral preoptic (AVP), anterodorsal preoptic (ADP), lat
eral septum (LS), bed nucleus of die stria term inalis (BST), and parabrachial
(PB) nuclei. All sections are arranged from rostral (left) to caudal (right). Every
clearly labeled neuron in these 30 pm thick sections was plotted. The distribu
tion of labeled axons projecting to the DMH is illustrated in figure 5. The entire
injection site in the m edial part of the m edial preoptic nucleus, as w ell as parts
of the injection sites in the parastrial, anteroventral periventricular, anteroventral
preoptic and anterodorsal preoptic, corresponding to panels B, H, K and N,
respectively, were previously reported in Simerly and Swanson (1988, figures 2
and 12). The injection site in the lateral septum, corresponding to panels P and
Q, was previously reported in Risold and Swanson (in press, figure 10).
129
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
MPN .
MePO
AVPv
5 4 “ / -
. . ; AVP •
.M e P O * . MPO
P
130
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Figure 4 (continued)
A p P -,
S
D
131
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 5. D istribution of PHAL-labeled projections from injection sites illustrated
in Fig. 4 to the DMH and surrounding region. Each is plotted onto a series of
standard tem plates of the rat brain representing levels 28,30, and 31 of the atlas
of Swanson (1992), as indicated in the upper left com er of the first series. Draw
ings are arranged from rostral (28) to caudal (31).
132
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
PS MePO AVPv MPNm
m
■Æ
PH /PWL
• f c
V
•••;'.--p v p . - > - > 0
i
y /.-.....v .v ^ ^
# %
133
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
injection was m ade in the m edian preoptic nucleus and its projection to the
DMH w as m apped in the sam e w ay (Fig. 4D-F; Fig. 5 MePO).
The d istrib u tio n of fibers w ithin the DM H is som ew hat sim ilar across
experim ents in that all preoptic nuclei project to the anterior part and tend to
avoid the posterior part, and none project significantly to the ventrom edial
hypothalam ic nucleus or the cell-sparse region ventral to the DMH (Fig. 5 PS,
MePO, AVPv, AVP, and ADP). However, these projections vary in apparent
density and intranuclear distribution. Overall, the density of term ination closely
m irrored the observed density of retrograde labeling. The m ost dense projection
seems to arise from the parastrial nucleus (Fig. PS), followed closely by those
from the anteroventral periventricular and m edial part of the medial preoptic
nuclei (Fig. 5 AVPv, MPNm). A t the other extreme, the m edian preoptic nucleus
provides only a m oderate input to the anterior p a rt and a m uch smaller input
to posterior and ventral parts (Fig. 5 MePO). From the preoptic region as a whole,
the parastrial nucleus provides the only substantial input to the posterior part
of the DMH, and these fibers are concentrated laterally. All preoptic nuclei send
a few axons to the posterior part, but this projection is proportionally m uch
sm aller than that to other parts of the DMH. Finally, input to the ventral p art of
the DMH seems quite varied. The parastrial, anteroventral periventricular, and
m edial preoptic nuclei provide the largest input to the ventral part. The parastrial
nucleus generates an especially dense plexus in rostral portions of the ventral
part. Interestingly, this also appears to apply to the anterodorsal nucleus,
134
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
although the num ber of fibers from the latter is less, as is a projection to caudal
regions of the ventral part (Fig. 5 PS, ADP).
Retrograde labeling was also observed in two circum ventricular organs: the
vascular organ of the lam ina term inalis and subfornical organ. O nly a few
neurons w ere seen in the vascular organ proper, but m any cells w ere labeled in
the surrounding area (Fig. 2B). Many well-labeled cells were observed bilaterally
in peripheral regions of the subfornical organ (Fig. 2E; Fig. 3D), leaving its central
region largely free of label. This annular d istrib u tio n is typical of m any
subfornical organ projections (Lind et al., 1982). This is in agreem ent with a
previous stu d y of subfornical organ projections that dem onstrated an input to
the DMH (Swanson and Lind, 1986).
Proceeding caudally from the m edial preoptic area, the DMH also receives
significant input from structures in the anterior hypothalamic area. These include
the anterior hypothalam ic, paraventricular, and suprachiasm atic nuclei, as well
as undifferentiated regions of the anterior hypothalam ic area (Fig. 2E-G; Fig.
3C).
FG-labeled neurons were seen throughout the anterior hypothalam ic nucleus
and surrounding area. The m ost dense labeling was seen in anterior regions of
the anterior hypothalam ic nucleus and anterior hypothalam ic area at the level
of the caudal pole of the medial preoptic nucleus (Fig. 2E; Fig. 3C). Significant
num bers of cells were also seen in the central part of the anterior hypothalam ic
135
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
nucleus. Here, there was a tendency for neurons to aggregate medially, just lateral
to the posterior part, whereas in the posterior and dorsal parts, only scattered
cells were seen (Fig. 2F-G). Previous studies of DMH inputs reported cells
throughout the anterior hypothalam ic nucleus. However, Fahrbach et al. (1984)
did call attention to "a prom inent cluster of labeled neurons... in the ventral
[anterior hypothalamic area], just dorsal to the optic tract." More recently, Risold
et al. (1994) m ade injections of the anterograde tracer PHAL into each p art of
the anterior hypothalam ic nucleus. They reported that the DMH receives a weak
(posterior part) to m oderate (central part) input and that these fibers term inate
predom inantly in the anterior part of the DMH.
R etrogradely labeled neurons w ere seen throughout the length of the
suprachiasm atic nucleus, and they were localized dorsomedially. In addition,
significant num bers of neurons w ere observed around the suprachiasm atic
n u c le u s, e sp ecially d o rsa lly (Fig. 2E; Fig. 3C). A p ro jectio n from th e
suprachiasm atic nucleus (Swanson and Cowan, 1975; Berk and Finkelstein,
1981b; Watts et al., 1987) and surrounding region (Watts and Swanson, 1987) to
the DMH is well-established, although it was not reported in previous studies
of DMH inputs (Berk and Finkelstein, 1981a; Kita and Oom ura, 1982). This
projections has since been show n to contain vasopressin (Watts and Swanson,
1987) and GAB A (Buijs et al., 1994). Watts et al. (1987) also m ade PHAL injections
into the subparaventricular zone, a major target of the suprachiasmatic nucleus,
and reported that this area also generates a significant projection to the DMH. In
136
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
another study. Watts and Swanson (1987) placed the retrograde tracer True Blue
into the region of the DMH. They showed labeled cells distributed throughout the
suprachiasmatic nucleus and surrounding region. Although the current results are
in general agreement w ith both Watts and Swanson (1987) and Watts et al. (1987),
w e observed retrogradely labeled m ostly w ithin dorsom edial p arts of the
suprachiasmatic nucleus. It is likely that this difference is explained by the large
size of the injection used by Watts et al. (1987), which may have labeled fibers passing
near the DMH, or by the use of True Blue, which is taken up avidly by fibers-of-
passage.
A few FG-labeled neurons were observed in the retrochiasmatic area (Fig. 2F).
A PHAL injection into the rostral retrochiasmatic area, ventral to the anterior part
of the anterior hypothalamic nucleus, indicates that this region provides a moderate
projection to the DMH, with a distribution similar to that observed in most of our
PHAL experiments, that is, a moderate to dense projection to anterior and ventral
parts of the DMH, while tending to avoid the posterior part (P.Y. Risold, personal
communication). With injections centered more caudally in the retrochiasmatic
region, only a few axons can be seen in the DMH (Risold et al., 1994).
In the hypothalamic paraventricular nucleus, dense retrograde labeling was seen
only in the anterior parvicellular part (Fig. 2E). Scattered cells were observed in the
ventral zone and caudally in the dorsal zone of the medial parvicellular part,
and in the lateral parvicellular part. Alm ost no neurons were seen in the dorsal
137
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and fomiceal parts, and m agnocellnlar neurons in both the paraventricular and
supraoptic nuclei unlabeled (Fig. 2F-G). Luiten and colleagues (Luiten et al.,
1985; Ter Horst and Luiten, 1987) injected PHAL into the paraventricular nucleus
and reported a major projection to the DMH, w ith the greatest density in the
ventral part (their ventral posterior DMH). They also reported m any branching,
varicose fibers in the rostral DMH. In contrast, after a PHAL injection centered
in the dorsal zone of the m edial parvicellular part, but including neuronal
labeling in anterior and dorsal parts, a m oderate projection to both the anterior
and posterior parts of the DM H w as found, w hile only sparse labeling w as seen
in the ventral part (L.W. Swanson and RE. Sawchenko, personal communication).
Labeled neurons were found at all levels of the arcuate nucleus, although
they were most dense near the rostral pole (Fig. 2F-H). Previous studies of DM H
inputs did not label arcuate neurons (Kita and Oom ura, 1982; Fahrbach et al.,
1984). However, Sim and Joseph (1991) reported a projection to the DM H after
injections of PHAL in both rostral and tuberal levels of the arcuate nucleus. In
addition, Magoul et al., (1994) show ed extensive fiber labeling in the DM H after
an injection of carbocyanine dye (Dil) centered in tuberal levels of the arcuate
nucleus. However, these results m ust be viewed w ith some caution because Dil
labels fibers originating from and projecting through the injection site. It is
therefore possible that this injection labeled fibers in the periventricular zone
th at subsequently turn dorsally to irmervate the DMH. The sam e caveats also
138
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
apply to the study of Bai et al. (1985) who lesioned the arcuate nucleus and
reported reduced num ber of neuropeptide Y-immunostained fibers in the DMH.
In the hypothalam ic ventrom edial nucleus, no cells were seen in the anterior
p a rt and m oderate num bers were observed in the dorsom edial, central, and
ventrolateral parts, except at the level of the injection site. At this level, greater
num bers were seen in both the central and ventrolateral parts (Fig. 2F-H). This
m ay be due in part to the fact than m any neurons in the ventrom edial nucleus
have long dendrites, some of which extend dorsally into the DMH (Millhouse,
1973). Therefore, at this level, some neuronal labeling in the ventrom edial
nucleus m ay have been produced by m eans of dendritic rather than axonal
transport.
P rev io u s H R ? retro g rad e tracer stu d ie s re p o rte d a sim ilar lab elin g
distribution (Kita and Oomura, 1982; Fahrbach et al., 1984; Ter horst and Luiten,
1987) in the ventrom edial nucleus, and also illustrate neurons ventrally and
laterally in a region corresponding to the tuberal nucleus. Recently, G anteras et
al. (1994) m ade PFIAL injections in each p art of the ventrom edial nucleus, and
in the tuberal nucleus. They reported that the central and ventrolateral parts
provide a m oderate input to the DMH, restricted m ainly to the anterior and
v e n tra l p arts. Both the anterior and dorsom edial p a rts tended to avoid
innervating the DMH, although a few axons were seen w ithin the borders, and
m any pass just lateral and dorsal to the DMH. From the estrogen and androgen
139
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
receptor-rich tuberal nucleus, they reported a m oderately dense projection to
anterior and ventral parts of the DMH.
M oderate num bers of retrogradely labeled neurons were found throughout
the ventral prem am m illary nucleus, although there was a tendency for them to
be m ore concentrated medially (Fig. 21; Fig. 3F). This is in agreem ent w ith 1er
horst and Luiten (1987) who reported moderate retrograde labeling in the ventral
prem am m illary nucleus after HRP injections, w hereas Kita and O om ura (1982)
only reported labeling in the "marginal region". However, Ganteras et al. (1992)
show ed that PHAL injections of the ventral prem am m illary nucleus label fibers
predom inantly w ithin the periventricular zone associated w ith the DMH, where
m any term inal boutons and varicosities were observed. Only sparse labeling
was present in other parts of the DMH.
Significant num bers of retrogradely labeled neurons were also seen in the
posterior periventricular and dorsal (but not ventral) tuberom am m illary nuclei
(Fig. 21; Fig. 3F). The tuberom am m illary cell group is rem arkable because it
synthesizes the putative neurotransm itter histam ine and is know n to have
w idespread projections throughout the neuraxis (Panula et al., 1989; Wada et
al., 1991). A lthough there is no direct confirmation that the tuberom am m illary
nuclei project to the DMH, the retrograde tracing study of Staines et al. (1987)
in d ic a te s th a t th is cell group projects h e a v ily th ro u g h th e p o ste rio r
h y p o th a la m u s, an d it is know n th a t th ere are a b u n d a n t h istam in e
140
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
immunoreactive fibers in the DMH (Panula et al., 1989). Because of its proximity
to the DMH, it is possible that tuberom am m illary neuronal labeling was
produced by non-axonal (perhaps dendritic) transport. However, this is unlikely
because a significant, though smaller, num ber of cells was also labeled in the
contralateral tuberom am m illary nucleus (Fig. 3F).
M any retrogradely labeled neurons were seen in rostral parts of the posterior
hypothalam ic nucleus, whereas relatively few cells w ere seen in caudal parts.
H ere, too, it is possible that these neurons were labeled by means of non-axonal
tran sp o rt because of the proxim ity of this area to the DMH injection site.
However, this does not account entirely for the observed labeling because 1) a
sim ilar distribution was seen in all experiments, including those centered slightly
m ore rostrally and ventrally; 2) labeling was observed contralaterally, although
to a m uch lesser extent than ipsilaterally; and 3) Vertes et al. (1995) reported a
m oderate projection to the DMH following injections of PHAL into the posterior
hypothalam ic nucleus, although it w as less dense th an the major projections of
the nucleus.
In all experiments, a few retrogradely labeled neurons w ere seen at each
level of the lateral hypothalam ic area. Most of these cells were observed medially
and the greatest num ber were seen betw een the levels of the DMH and the caudal
paraventricular hypothalam ic nucleus (Fig. 2E-J). A t all other levels, neuronal
labeling was sparse. These findings are in general agreem ent w ith previous
141
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
reports of DMH inputs (Luiten and Room, 1980; Kita and Oomura, 1982; Luiten
et al., 1987). In contrast, earlier studies of lateral hypothalamic projections based
on the autoradiographic m ethod show only m odest labeling in the DM H after
large injections (Saper et al., 1979, Berk and Finkelstein, 1982). M ore recent
studies with the sensitive PHAL m ethod have show n a m oderate projection
from tuberal levels of the lateral hypothalam us (Luiten et al., 1987; Allen and
Cechetto, 1993). However, projections from the most m edial lateral hypothalam ic
area, which seems to generate the densest input to the DMH, have n ot yet been
exam ined with anterograde transport m ethods.
In agreem ent w ith the w o rk of K ita an d O om ura (1982), w e found
retrogradely labeled cells in the supram am m illary nucleus, the majority of which
w ere located in the m edial part, although scattered labeling was seen in the
lateral part as well (Fig. 2J). Vertes (1992) reported a small projection to the
DM H after a PHAL injection in the lateral part, and a m oderate projection from
the m edial part and from the m edial region of the lateral supram am m illary
nucleus.
Thalamus: FG-labeled neurons w ere consistently observed in only three cell
g ro u p s of th e th alam u s: th e p a ra v e n tric u la r th alam ic n u c le u s, the
subparafascicular nucleus, a n d the lateral habenula. Cell labeling in the
paraventricular thalam ic nucleus was m ainly observed in the rostral third of
this nucleus, although a few scattered cells were seen throughout its extent (Fig.
142
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
2E-F, I). This nucleus has been reported to project to the DMH by Moga et al.
(1995). After a large PHAL injection centered rostrally, they showed dense fiber
labeling in the DMH that seemed equally to innervate the posterior, anterior,
and ventral parts, although they did n o t illustrate subdivisions. Furtherm ore,
they show ed very little axonal labeling in the DMH follow ing injections at more
caudal levels of the thalam ic paraventricular nucleus.
In each experim ent, a few retrogradely labeled neurons were seen in the
lateral habenulae (Fig. 21), as previously reported (Berk and Finkelstein, 1981a).
Following injections of tritiated am ino acids, H erkenham and N auta (1979)
show ed a very sparse projection to the region of the DMH.
A few labeled neurons were consistently observed in both the magnocellular
and parvicellular parts of the subparafascicular nucleus (Fig. 2I-J). A lthough it
is clear that the parvicellular part (and the peripeduncular nucleus) project to
the ventrom edial nucleus of the hypothalam us, retrograde tracer injections into
the ventrom edial nucleus reportedly do not label the m agnocellular p art of the
subparafascicular nucleus (LeDoux et al., 1985). In contrast, PHAL injections of
the m agnocellular p a rt label axons in dorsal but n o t ventral regions of the
hypothalam ic m edial zone, though the DMH was n o t specifically exam ined
(Yasui et al., 1992). It is therefore likely that a part of the retrograde labeling
observed here, at least for the magnocellular part, has a contribution from axons
w ithin the DMH. However, it is clear that this is not a m ajor input.
143
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Telencephalon. Projections to the DM H from the telencephalon are generally
rather sparse and arise from only a few, restricted areas. Subcortically, FG-labeled
neurons were seen mostly in the BST and lateral septal nucleus (Fig. 2B-E). A
few neurons were also present in the substemtia innominata at rostral levels of
the anterior hypothalamic area (Fig. 2E). In the cortex, labeled cells w ere seen
in the prefrontal cortex (mainly in the infralim bic area, b u t also in the prelimbic
area and tenia tecta), claustrum, and ventral subiculum.
Subcortical Telencephalon. W ithin the septum , FG-labeled neurons w ere seen
in interm ediate and ventral parts of the lateral septum. In interm ediate parts,
neuronal labeling could be seen throughout the rostrocaudal extent of the septum
w ith slightly greater num bers of cells rostrally. However, at all levels, labeling
was limited to the ventral and medial quadrant. Neuronal labeling in the ventral
p art w as greatest in the caudal two thirds. Substantially fewer cells w ere seen
in the rostral parts of the ventral lateral septum (Fig. 2B-D).
In the BST, labeling was observed at all levels (Fig. 2B-E). N eurons were
fo u n d in the a n tero d o rsal (in c lu d in g th e central core), a n te ro v e n tra l,
dorsom edial, ventral, interfascicular, transverse, and principle nuclei. Perhaps
the greatest num ber of labeled neurons in the BST w as observed rostrally,
particularly in the anterodorsal part (Fig. 2B). Virtually no cells were seen in the
oval, rhom boid, and juxtacapsular nuclei, and the anterolateral part.
144
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Scattered labeling was also seen throughout the substantia innom inata.
However, the num ber of cells at any level w as very small. Nevertheless, this
projection has been confirmed w ith PHAL and it has been show n that these
neurons project to the DMH through the m edial forebrain bundle (Grove, 1988).
Swanson and Cowan (1979) injected tritiated amino acids into the septal
region, including the lateral septal nucleus and BST. After an injection at central
levels of the BST, they reported substantial projections to the DMH, whereas
injections centered in the caudal BST selectively labeled the capsule surrounding
the ventrom edial hypothalamic nucleus. In a PHAL control injection centered
in the rostral BST (Fig. 4S-U), we have confirm ed that the BST sends a major
projection to the DMH. Although this injection labeled significant num bers of
neurons in the nucleus accum bens and adjacent lateral septal nucleus, we
conclude that the major part of this projection arises from the anterodorsal part
of the BST because of the very sparse retrograde labeling in the other regions
after PC injections in the DMH. W ithin the DMH, the m ajority of anterograde
labeling w as found in the anterior and ventral parts. W hile labeled axons
generally tended to avoid the posterior part, a significant num ber of terminals
was seen ventrom edially (Fig. 5 BST).
Swanson and Cowan (1979) also reported a m oderate projection from the
lateral septal nucleus to the DMH. In a more detailed study, Risold and Swanson
(in press) reported that the m ost dense projection from the lateral septal nucleus
145
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
to the DMH arises ventrally, particularly caudally. From this injection (their
case LSRS in the ventral lateral septal nucleus) they noted that the DM H receives
one of the most dense projections at the hypothalam ic tuberal level and that
labeled axons term in ate in the a n te rio r and v e n tra l parts, w h ile alm ost
com pletely avoiding the posterior part. This pattern is a similar to that seen
after an injection centered in more rostral parts of the ventrolateral septum,
w hich in clu d ed sig n ifican t lab e lin g in the v e n tro m ed ia l re g io n of the
interm ediate lateral septum (Fig. 4P-R). A lthough the projections from this
experim ent (LS31, in the ventral region of the ventrolateral zone of the rostral
part of the lateral septum ) were illustrated in this stu d y (and the injection site
plotted in their Figure 10), a more detailed m apping of this case is presented in
Figure 5 LS. Here it is shown that rostral regions of the anterior and ventral
parts of the DMH receive a m oderate to dense projection, whereas caudal regions
of these two parts receive m any few er axons, although significant branching
and varicosities can still be seen. Again, there is a tendency for axons to avoid
the posterior part, but fibers can be seen term inating w ithin its ventral region.
Risold and Swanson (in press) also m ade PHAL injections into various parts
of the rostral lateral septal nucleus. From this region, particularly case LS49 in
the dorsal region of the ventrolateral zone of the rostral part of the lateral septal
nucleus, they report a small projection to the anterior part of the DM H and a
dense projection to the capsule of the ventrom edial hypothalam ic nucleus,
including the part ventrally adjacent to the DMH. Because we see significant
146
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
retrograde labeling in this region, it therefore m ust be considered that these
FG-labeled cells m ay be labeled from axons in the capsule of the ventrom edial
nucleus, and not in the DMH proper. However, this seems unlikely because
very few cells were observed in parts of the am ygdala know n to project very
densely to the dorsal part of the capsule (Ganteras et al., 1995). A more likely
explanation is that because the neuronal labeling in the lateral septal nucleus
w as confined to a fairly restricted region, PHAL injections just dorsal or lateral
to this region label correspondingly fewer DM H-prejecting cells.
Cortical Telencephalon. The projection from the infralimbic area, although
small, is well docum ented (H urley et al., 1991; Sesack et al., 1989; Brittain, 1988).
N o specific reports of tenia tecta projections to the m edial hypothalam us are
available and labeling in this region was not reported in previous studies of
DM H afferents (Berk and Finkelstein, 1981a; Kita and Oom ura, 1982). Therefore,
it is impossible to determ ine w hether this area sends fibers specifically to the
DMH, or to adjacent regions.
There have been no previous reports of a projection from the claustrum to
th e m edial hypothalam us. H ow ever, large HRP injections in the lateral
hypothalam ic area just caudal to the DMH result in significant labeling of
claustral neurons (Allen et al., 1991-33). PHAL injections in the overlying anterior
insular area that include the claustrum label axons that pass dorsom edially
adjacent to the DMH (Beckstead, 1979). It is unlikely that this pattern of labeling
147
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
results from projections of the insular area because more restricted injections of
PHAL do not label fibers in the vicinity of the DMH (Yasui et al., 1991).
Retrogradely labeled cells were consistently seen in the ventral subiculum .
After injections of PHAL here. G anteras and Swanson (1992) reported that at
the level of the DM H, the majority of fibers from the ventral subiculum are
located in the capsule of the ventrom edial hypothalamic nucleus and from this
region, a proportionately sm all n u m b er turn dorsally to e n te r the DM H.
However, after a large PHAL injection centered in the ventral subiculum m ore
caudal to the level injected by G anteras and Swanson (1992), G ullinan et al.
(1993) illustrate a m uch more dense projection. Although they do not parcellate
the DMH, it w ould appear that the ventral subiculum provides input to all parts.
Ascending Inputs
Brainstem. Inputs to the DMH from the brainstem are prim arily lim ited to
the periaqueductal gray, parabrachial nucleus, and ventrolateral m edulla (the
region of the A l/G l catecholam inergic cell groups). Scattered cell were also
seen in various raphe nuclei, mesencephalic reticular nucleus, region of the A7
noradrenergic cell group, Barrington's nucleus, and pontine central gray. In
addition, a very restricted group of neurons was found in the caudal end of the
nucleus incertus (not shown). To our knowledge, no systematic investigation of
the projections of this nucleus has been perform ed and no other stu d y of DM H
inputs has reported this projection. Therefore, it rem ains to be determ ined
148
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
w h eth er the cau d al nucleus incertus projects specifically to the DM H.
Nevertheless, it is a novel finding that this region of the brainstem projects to
the hypothalam us.
In the periaqueductal gray, labeled neurons w ere found at each level.
However, at no level were they particularly dense, although there was a tendency
tow ard increasing cell num bers and label intensity caudally (Fig. 2-JM). This is
in general agreem ent w ith Kita and O om ura (1982) and Berk and Finkelstein
(1981a), who reported the m ain body of neuronal labeling ventrom edially at
inter- and inferior coUicular levels.
Eberhart et al. (1985) m ade injections of tritiated am ino acids into various
levels of the periaqueductal gray and the adjacent m esencephalic reticular
nucleus. R esultant labeling w as seen in the DM H following injections of each
level (superior, inter-, and inferior collicular), although labeling was lighter after
injections of the periaqueductal gray at the level of the superior colliculus.
Projections w ere also sparse from the m esencephalic reticular nucleus. In
c o n trast, follow ing PHAL injections into ro stral o r c au d a l dorso- an d
ventrolateral periaqueductal gray, Cameron et al. (1995) report only m odest
projections to the DMH.
Within the raphe nuclei, scattered neuronal labeling w as found in the rostral
and central linear, medial and lateral superior central, and dorsal nuclei of the
raphe. Moore et al. (1978) and Bobillier et al. (1979) reported that the DMH
149
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
receives a light projection from the superior central nucleus. Vertes (1991) m ade
sm all injections of PHAL in the dorsal raphe and reported an occasional fiber in
the DMH, confirm ing that the dorsal raphe does not generate a major input to
the DMH. A lthough the num ber of retrogradely labeled neurons was small and
the reported projections sparse, it has been shown that there is a m oderate
num ber of serotonin fibers within the DMH (Brownstein and Palkovits, 1984)
and that serotonin imm unoreactivity dim inishes in the DM H after raphe lesions
(Van de Kar and Lorens, 1979). Therefore, these small projections from the raphe
nuclei are the best candidates for generating the serotonergic input.
A sm all group of labeled neurons w as found in a region corresponding to
the A7 noradrenergic cell group: dorsal and ventral to fibers in the decussation
of the superior cerebellar peduncle and extending caudally in this region to the
level of the parabrachial nucleus (Moore and Card, 1984; Fig. 2L). Although the
requisite double labelling study w as not perform ed, and there are no studies
reporting projections from this area to the DMH, it has been reported that lesions
of this cell group result in significant term inal degeneration, and a reduction in
noradrenaline content, in the DMH (Palkovits et al., 1980).
A projection from the parabrachial nucleus has been reported previously
w ith the autoradiographic (Saper and Loewy, 1980) and PHAL (Krukoff et al.,
1993) m ethods. Based on HRP studies of DMH inputs, it is generally agreed
that this projection m ainly arises from lateral parts of the nucleus (Berk and
150
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
F inkelstein, 1981a; Kita an d O om ura, 1982; F u iw iler a n d Saper, 1984).
Furtherm ore, Fuiw iler and Saper (1984) reported projections to the DM H
specifically from the superior, dorsal and central lateral parts of the parabrachial
nucleus. O ur present findings are in agreem ent w ith these previous results. The
majority of neuronal labeling was found in the central lateral part, with scattered
cells found in the superior and dorsal parts (Fig. 2M -0; Fig. 3E). An occasional
cell was found in the external lateral part caudally, although by-and-large this
p art was devoid of labeling. Interestingly, the single largest group of labeled
neurons was found in the central lateral part im m ediately adjacent to, b u t not
within, the external lateral part. A few neurons also w ere found in the Kolliker-
Fuse nucleus. It is possible, however, that these latter neurons are part of the
caudal extension of the A7 catecholaminergic cell group, rather than part of the
Kolliker-Fuse nucleus, per se.
Because the parabrachial nucleus constitutes a m ajor input to the DM H, a
series of PHAL control injections were perform ed. O f five injections aim ed at
the parabrachial nucleus, two produced significant neuronal labeling in the
lateral division (Fig. 4V-X). The distribution of axonal labeling in the DM H
produced by the larger of these two injections is illustrated in Figure 5 PB. These
results show that the lateral parabrachial nucleus provides a m oderate projection
to the anterior p art and a sparse projection to the ventral p art of the DMH.
Interestingly, in the posterior part there is very dense term inal labeling laterally,
whereas medially the posterior part was almost completely free of labeled axons.
151
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However, because this injection w as not centered in the restricted neuronal
clusters that project to the DM H, w e cannot rule out the possibility that there is
a v ery tight topography such th a t com plete labeling o f the parabrachial
projection w ould result in axonal labeling in other parts of the DMH, even
though this pattern was seen in both cases. Nevertheless, it is clear that the
parabrachial nucleus sends a very dense projection to at least p art of the posterior
p a rt—a region avoided by m ost other inputs (see above).
A t caudal levels of the parabrachial nucleus, retrogradely labeled neurons
w ere found in Barrington's nucleus and scattered throughout the pontine central
gray (Fig. 2N -0). A lthough few neurons were seen in the locus coeruleus, other
than a small cluster near the rostral pole, previous studies consistently have
reported a connection to the D M H from the locus coeruleus in both retrograde
(DMH) and anterograde tracer (locus coeruleus) studies (Holets et al., 1988;
Berk and Finkelstein, 1981a; Jones and Moore, 1977).
In the m edulla, a few neurons w ere seen in caudal parts of the nucleus of the
solitary tract (NTS), and throughout the region of the paragigantocellular and
m edullary reticular nuclei generally referred to as the ventrolateral m edulla.
W ithin the NTS, labeled neurons are first apparent just rostral to the obex
and were present at each level caudally. There w ere no dense groups of labeled
neurons, and while most cells w ere found in the medial part, scattered labeling
w as found in the comm issural and lateral parts (Fig. 2Q). These results are
152
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
consistent w ith previous studies of NTS projections (Ricardo and Koh, 1978;
Sawchenko and Swanson, 1983; Ter horst et al., 1989), which have show n a
moderate projection to the DMH. Ter horst et al. (1989) have reported that the
caudal, viscerosensory region of the NTS projects to areas corresponding to our
anterior, posterior, and ventral parts of the DMH. In addition, Swanson and
Sawchenko (1983) reported cells in the NTS that were doubly labeled after an
injection of the retrograde tracer True Blue and im m unostaining for dopam ine-
6-h y d ro x y lase (DBH), in d ic atin g th a t th is projection is a t least p a rtly
catecholaminergic, arising from the A2 noradrenergic cell group. C onsidering
the total num ber of cells labeled throughout the length of the NTS, it seem s
likely that this nucleus provides a significant input to the DMH, part of w hich
is noradrenergic.
The m ost caudally arising projections to the DMH found in this study w ere
found in the lateral paragigantocellular and m edullary reticular nuclei (Fig. 2P-
Q). This projection w as largely bilateral, w ith the num ber of contralateral
neurons being about tw o-thirds that of the ipsilateral neurons. Labeling in this
region, on both sides of the brain, first appears just caudal to the rostral pole of
the inferior olivary com plex and continues as far caudally as the transition area
between the m edulla and spinal cord. However, no labeling could be found at
cervical spinal levels. This contrasts som ewhat w ith the reports of Giesler and
colleagues (d iffe r et al., 1991; G iesler et al., 1994) who reported a sparse
projection to the DMH after m ultiple PHAL injections into lower cervical and
153
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
upper thoracic levels of the spinal cord. Although the reason for this discrepancy
is not entirely clear, it is possible that we did not observe this projection because
sections were not collected beyond cervical levels.
It is interesting to note that the distribution of m edullary cells corresponds
very w ell to the distribution of DBH labeling in the C l/A l catecholaminergic
cell groups generally referred to as the rostral and caudal ventrolateral m edulla,
respectively (Swanson and H artm an, 1975). Indeed, significant projections to
the DM H have been reported after autoradiographic tracing experim ents
involving this region (Loewy et al., 1981; McKellar and Loewy, 1982; Sawchenko
and Swanson, 1982; Vertes et al., 1986). Moreover, following injections of the
fluorescent retrograde tracer True Blue into the DMH, Sawchenko and Swanson
(1982) reported that about 70% of the labeled cells also im m unostained w ith an
antiserum raised against DBH. A lthough this implies that there is a significant
num ber of non-catecholaminergic cells projecting to the DMH, the distribution
and num ber of cells, both ipsi- and contralaterally, in this projection indicate
th a t the v en tro lateral m edulla is a m ajor source of b o th adrenergic and
noradrenergic inputs to the DMH.
C ontralateral Inputs
In general, FG-labeled neurons contralateral to the DMH injection site showed
a sim ilar distribution to ipsilateral labeled neurons, although the num bers were
far smaller. In the forebrain, contralaterally labeled neurons in m ost regions
154
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
were very sparse, although in most cases an occasional cell could be seen. The
only exceptions w ere the m edian preoptic, rostral pole of the arcuate, and
supram am m illary nuclei where labeling could be seen along the m idline, and
in the subfornical organ where labeling w as bilateral throughout (Fig. 2E; Fig.
3D).
In the brainstem , the same is generally true. The projection from the raphe,
although sm all, was largely bilateral. The contralateral projection from the
ventrolateral m edulla, as m entioned above, consisted of about tw o-thirds of
the num ber of ipsilateral cells, providing the largest contralateral projection to
the DM H (Fig. 2Q).
Input Pathways
C om bining the present results w ith a survey of the literature, it is now
possible to sum m arize the pathw ays used by virtually all known inputs to the
DM H (Fig. 6). Although the DMH receives projections from a large num ber of
brain regions (about 40 areas, not counting the individual parts of the BST),
these inputs use a very lim ited set of pathways. Descending inputs from the
telencephalon, thalam us, and hypothalam us use only three pathw ays in total: a
periventricular pathw ay that travels through the hypothalam ic periventricular
zone, a lateral pathw ay that travels in the medial forebrain bundle (in the lateral
h y p o th a la m ic a re a), an d a m edial p a th w a y th a t follow s th e m ed ial
corticohypothalam ic tract. The medial pathw ay overlaps the periventricular
155
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
pathw ay m edially but form s a distinct pathw ay just dorsal to the anterior and
ventrom edial hypothalam ic m edial zone nuclei, where it passes through the
subparaventricular zone and capsule of the ventromedial hypothalam ic nucleus.
Ascending inputs from the caudal hypothalam us and brainstem are prim arily
lim ited to a m id b rain p e riv e n tric u la r p ath w ay that passes th ro u g h the
periaqueductal gray near the cerebral aqueduct, and a brainstem lateral pathw ay
ascending in the m edial forebrain bundle.
The m ajority of intrahypothalam ic inputs to the DMH project through the
periventricular pathway. In particular, all of the projections from the preoptic
region (m edian preoptic, parastrial, anteroventral periventricular, preoptic
suprachiasmatic, anterodorsal preoptic, anteroventral preoptic, and m edial parts
of the m edial preoptic nucleus) send a major com ponent of their axons m edially
and then cau d ally to descen d w ith in preoptic and an terio r p a rts of the
hypothalam ic periventricular zone. A t the level of the DMH, they again turn
laterally to innervate th is stru ctu re w ith a d istrib u tio n described above
(Swanson, 1976; Simerly and Swanson, 1988; Gu and Simerly, in press). It should
be noted, however, that each of these nuclei also sends a much sm aller projection
through the m edial forebrain bundle, prim arily in m edial parts of the lateral
hypothalamic area, that turns laterally to enter the DMH at the level of its anterior
part. The arcuate and v entrom edial hypothalam ic nuclei also send th eir
projections through the periventricular pathway, although in these cases, the
156
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
fibers prim arily ascend dorsally to enter the periventricular zone at the level of
the DMH (Ganteras et al., 1994; M agoul et al., 1994).
In the telencephalon, the only region to send a major projection through the
periventricular pathw ay is the BST, although a significant com ponent of these
fibers travels through m edial parts of the m edial forebrain bundle as w ell
(Swanson and Cowan, 1979). Although it is possible that different parts of the
BST selectively use different pathw ays, the results of our control injection
centered in the anterodorsal part of the BST show that this part distributes fibers
alm ost equally betw een periventricular and lateral pathw ays.
Fewer regions project to the DMH through the lateral pathw ay—m ainly the
infralimbic area, subfornical organ, and anterior hypothalam ic nucleus. In the
descending com ponent of the lateral pathway, projections to the DM H are found
prim arily in dorsom edial regions of the lateral hypothalamic area. Labeled axons
descend from the infralim bic area through interm ediate regions of the rostral
p a rt of the lateral septal nucleus before entering the lateral pathw ay. From a
position in m edial regions of the lateral hypothalam ic area, a group of these
fibers turns m edially to innervate the anterior hypothalam ic nucleus and then
extends dorsally and caudally into the DMH. Along w ith axons extending
medially from the dorsom edial lateral hypothalamic area, they generate a m odest
term inal field in the DM H (Brittain, 1988). Projections from the an terio r
hypothalam ic nucleus take a similar route. Fibers exit this nucleus dorsally and
157
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
p ass through the subparaventricular zone an d around cau d al regions of the
p arav en tricu lar nucleus to e n te r m edial regions of the zona incerta. From
here, the axons descend to innervate the D M H (Risold et al., 1994).
Axons from the subfornical organ descend ventrally along the m idline,
th ro u g h the m edian preoptic nucleus, and th en turn laterally to innervate
the p arastrial nucleus before entering the lateral pathw ay. A t the level of
the DMH, these fibers again tu rn m edially a n d pass dorsal to the fornix to
e n te r the D M H (Swanson and Lind, 1986).
Perhaps th e best characterized projection through the m edial pathw ay
an d m edial corticohypothalam ic tract originates in the v en tral subiculum .
T hese fibers e n te r the hypoth alam us th ro u g h the ro stral hypothalam ic
p a ra v e n tr ic u la r n u c le u s, tra v e l c a u d a lly as a d e n se b u n d le in th e
su b p a rav e n tric u la r zone and capsule of th e ventrom edial hypothalam ic
nucleus and en d diffusely in caudal regions of the tuberal hypothalam us.
In route, a g roup of these fibers turns dorsally from the ventrom edial nucleus
capsule to en te r the DMH (G anteras and Sw anson, 1992). In ad d itio n to the
v e n tra l su b ic u lu m , the p ro jectio n s from th e lateral s e p ta l an d s u p ra
chiasm atic nuclei, and the subparaventricular zone join the m edial p ath
w ay and tu rn dorsally from the capsule of the ventrom edial nucleus to en
te r the DM H (W atts et al., 1987; Risold and Sw anson, in press).
158
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
The majority of brainstem areas projecting to the DMH use the lateral brainstem
pathway. These areas include the ventrolateral medulla, NTS, parabrachial nucleus,
and raphe (mainly the central lateral and superior central) nuclei.
Axons from the ventrolateral m edulla and NTS ascend in the ventrolateral
tegm entum in the m edulla and pons, just ventrom edial to the spinal trigem inal
complex. In the pons, this projection bifurcates to follow two distinct routes. A
dorsal com ponent courses ventrolateral to the periaqueductal gray. The second,
sm aller com ponent takes a more ventral route through the m esencephalic
reticular nucleus, approxim ately corresponding to the ventral noradrenergic
bundle (Swanson and Hartm an, 1975). At the mesodiencephalic junction, parts
of these two fiber groups merge and continue rostrally in dorsolateral regions
of the medial forebrain bundle, from w hich they abruptly turn m edially to enter
the DMH. A sim ilar fiber group decussates prim arily w ithin the m edulla and
th e n ascends on the opposite side of the brain (Ricardo and Koh, 1978;
Sawchenko and Swanson, 1982).
Projections from the parabrachial and mesencephalic reticular nuclei also
join ascending projections from the ventrolateral m edulla, although in this case
the majority of fibers were seen in the dorsal tegm ental bundle, ventrolateral to
the periaqueductal gray. At the mesodiencephalic junction, some of these fibers
extend rostrally into the medial forebrain bundle, from which they turn medially
to enter the DM H (Saper and Loewy, 1980; Eberhart et al., 1985).
159
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ascending projections from the superior central nucleus take a slightly
different route, although they also enter the DM H from the lateral pathway.
These fibers ascend just lateral to the central linear nucleus and course through
the ventral tegm ental area to enter the medial forebrain bundle (Bobillier et al.,
1979).
Projections to the DM H from the ascending m idbrain p eriv en tricu lar
pathw ay are prim arily lim ited to nuclei of the caudal hypothalam us and the
p eriaq u ed u ctal gray. The m ajo rity of projections to the D M H from the
periaqueductal gray, labeled by injections at superior, inferior, and intercoUicular
levels, extend rostrally from the injection site to ascend adjacent to the cerebral
aqueduct. A t the m esodiencephalic junction, these fibers follow the third
ventricle to innervate the periventricular zone, including the DM H (Eberhart et
al., 1985). However, it should be noted that a com ponent of axons from each of
these injection sites exits the periaqueductal gray laterally to enter m edial regions
of the m esencephalic reticular nucleus, passes through the ventral tegm ental
area, and joins the lateral pathw ay (Eberhart et al., 1985). It is therefore likely
that part of the periaqueductal gray projection reaches the DM H through the
brainstem lateral pathway.
In the caudal hypothalam us, projections from the ventral prem am m illary
nucleus exit laterally, then tu rn rostrally and dorsally to pass through and
innervate the periventricular zone at the level of the DMH (Ganteras and
160
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 6. General organization of projections to the DMH. Inputs prim arily uti
lize three descending pathways: periventricular and m edial (1) and lateral (2),
and tw o major ascending pathways: m idbrain periventricular (3) and brainstem
lateral (4). Pathw ays that were observed in our control injections, or reliably
reported in the literature, are represented by a solid line. Regions for which the
pathw ay is uncertain are represented by a dashed line.
161
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
?»
162
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Swanson, 1992). Projections from the supram am m illary nucleus pass dorsally
and rostrally through the posterior hypothalam ic nucleus to enter the DMH
directly (Vertes, 1992). Because these fibers travel in a m anner sim ilar to
projections from the periaqueductal gray, they are included in the periventricular
pathway.
DISCUSSION
The DMH is a very enigm atic nucleus. As outlined in the Introduction, it has
been associated in some capacity w ith almost every goal-directed behavior and
visceral response associated w ith the hypothalam us. It has m ost often been
im plicated in disturbances of feeding behavior, that is, the "DMH syndrom e"
(see Bem ardis and Bellinger, 1987). The characteristics of this syndrom e are
m ainly that, following DMH lesions, animals reduce food and w ater intake and
have a low er body weight. However, these anim als defend their new body
w eight and h y d ratio n sta tu s, show ing n o rm al responses to hom eostatic
challenges such as food or w ater restriction (Bemardis, 1972; Dalton et al., 1981).
A lthough these results are not indicative of a prim ary role in ingestive
behavior, the DM H does appear to receive inform ation about the metabolic state
of the anim al via projections from the ventrolateral medulla and parabrachial
nucleus. A lthough a direct in p u t to the DM H from the NTS is sm all, the
ventrolateral m edulla and the central lateral and dorsal lateral parts of the
163
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
parabrachial nucleus, both of which contain m any retrogradely labelled neurons
after DM H FG injections, receive a major input from viscerosensory regions of
the NTS (Ricardo and Koh, 1978; Loewy et al., 1981; H erbert et al., 1990). Inputs
to the DM H from these parts of the parabrachial nucleus m ay be especially
im p o rtan t because the latter have been im plicated in feeding induced by
peripheral signals. Fatty ad d s and fructose can used as fuel directly by peripheral
tissues, but not by the brain. M ercaptoacetate and 2,5-anhydro-D-mannitol,
specific inhibitors of fatty acid and fructose utilization, respectively, stim ulate
feeding (Friedman et al., 1986; Tordoff et al., 1988). This feeding is blocked by
ibotenic acid lesions of the lateral p arab rach ial nucleus, especially those
involving the central and dorsal lateral parts (Calingasan and Ritter, 1993; Grill
et al., 1995). It is im portant to note that these lesions did not affect feeding in
resp o n se to glucose d e p riv a tio n p ro d u c e d by 2-deoxyglucose (2-DG)
adm inistration. Because glucose can be utilized (and glucose levels sensed)
directly by the brain, this finding suggests additional mechanisms regulating
brain glucose levels.
Interestingly, there is evidence to suggest that the DMH plays a role in these
central mechanisms. DM H-lesioned anim als do not increase food consum ption
in the first 4 hours after 2-DG injection, although 24 hour totals were norm al for
these rats, and glucose injections, which suppress refeeding after an overnight
fast in unlesioned anim als, did not suppress food intake in DM H-lesioned
anim als (Bemardis and Bellinger, 1987). M oreover, glucose-receptive neurons
164
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
have been reported in the DMH, although the functional contribution of such
neurons is unclear (Niijima, 1989).
Further evidence for a role of the DM H in carbohydrate metabolism and
ingestive behavior may lie in the projection from the arcuate nucleus to the
DMH. Recent w ork has show n that the neuropeptide Y-containing projections
from the arcuate nucleus, prim arily to the hypothalam ic paraventricular nucleus,
are involved in feeding behavior, particularly related carbohydrate ingestion
and insulin secretion (Akabayashi et al., 1994; Jhanw ar-U niyal et al., 1993).
However, a role for such projections to the DMH in feeding is implied by the
finding that injections of neuropeptide Y into the DM H induces expression of
the imm ediate early gene product Fos in food deprived (and refed) anim als (Li
et al., 1994). Although the study of Bai et al. (1985) suggests that the projection
from the arcuate nucleus to the DMH is at least partially neuropeptide Y-ergic,
it sh o u ld be n o te d th a t th e p e p tid e is e x te n siv e ly c o ex p ressed w ith
norepinephrine in neurons of the A1 cell group and w ith epinephrine in neurons
of the C l and C2 cell groups (Everitt et al., 1988; Sawchenko et al., 1985), and
that brainstem transactions decrease neuropeptide Y levels in the DMH (Sahu
et al., 1988).
Although few specific changes in drinking occur follow ing DMH lesions,
one of the m ost obvious roles for the DMH suggested by our anatomical results
is in fluid balance and drinking behavior. One of the responses to a drop in
165
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
blood volum e is increased levels of circulating angiotensin n. This hormone
has a variety of actions, including vasopressin release, a sym pathetic pressor
response, and the initiation of drinking behavior (Johnson, 1985; Swanson, 1987).
Lesions of the subfornical organ or its projections block these effects (Johnson,
1985). The m ajority of subfornical organ projections end in m agnocellular
neurosecretory cell groups in the paraventricular and supraoptic nuclei, and in
the median preoptic and parastrial nuclei, although the DMH, anteroventral
periventricular nucleus, and perifom ical area receive substantial projections as
w ell (Lind et al., 1984; Swanson and Lind, 1986). O ther than the magnocellular
neurons of the supraoptic and paraventricular hypothalamic nuclei, each of these
subfornical organ-recipient nuclei also project to, and receive projections from,
the DMH (Thom pson et al., 1996, present results). Based on these anatomical
observations, it is difficult to avoid the conclusion that the DMH is influenced
by, and influences the output of, the subfornical organ. These anatom ical results
m ay explain, in part, the hypodipsia observed after DM H lesions (Dalton et al.,
1981; Bellinger and Bemardis, 1982) and suggest a role for the DMH in fluid
balance and drinking behavior.
The aspect of DMH function that has received the m ost attention recently is
its involvem ent w ith circadian and stress-induced corticosterone secretion,
p resu m ab ly m e d ia te d by p a rv o c e llu la r n e u ro sec reto ry n e u ro n s in the
hypothalam ic paraventricular nucleus th a t contain corticotropin-releasing
horm one and control secretion of adrenocorticotropic horm one secretion from
166
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the anterior pituitary. Initially, it was noted that DMH lesions produce blood
lev els of c o rtico stero n e sim ilar to those o bserved after lesio n s of the
suprachiasm atic nucleus; that is, the circadian, preactivity peak w as abolished
and basal secretion was elevated (Bemardis et al., 1976). More direct evidence
w as p ro v id e d by L eibow itz et al. (1989) who show ed th at in jectio n s of
norepinephrine into the DMH stim ulate corticosterone secretion, although less
than that produced by similar injections into the hypothalam ic paraventricular
nucleus. In addition, a survey of the literature shows that the im m ediate early
gene c-fos is expressed in the DMH in response to a variety of stressors, including
footshock (Pezzone et al.,1992), restraint (Imaki et al., 1993), swim m ing (Cullinan
et al., 1996), im m une stim ulation (Sagar et al., 1995; Elmquist et al., 1996), and
ether (Cullinan et al., 1996; RE. Sawchenko, personal communication).
R ecent studies u ndertaken to exam ine further the projection from the
suprachiasm atic nucleus to the DMH, and its influence on the preactivity
corticosterone peak, indicate that the circadian com ponent is m ediated by
vaso p ressin and GABA containing axons, and th at this projection has a
significant influence upon stress-induced as well as circadian corticosterone
secretion (Buijs et al., 1993). Originally, Kalsbeek et al. (1992) show ed that
infusion of vasopressin into the region of the DMH in suprachiasm atic nucleus-
lesioned anim als produced an im m ediate decrease in circulating corticosterone
levels, w hereas the converse was true of infusions of a vasopressin antagonist
in intact anim als—a dram atic increase in corticosterone levels in response to a
167
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m ild stressor, relative to the increase în intact controls. Kalsbeek et al. (1996b)
then show ed that the effects of vasopressin antagonist adm inistration varies,
depending upon the tim e of day; there w as little effect during the dark period
and an increasing effect during the course of the light period. D uring the
preactiv ity corticosterone peak (im m ediately preceding the d a rk period),
vasopressin antagonist infusion further increase corticosterone secretion. These
re su lts in d icate th e presence of a d d itio n a l stim u la to ry an d in h ib ito ry
com ponents of circadian corticosterone secretion.
C o n c ern in g th e stim u la to ry co m p o n en t, it is w e ll-e sta b lish e d th a t
catecholam inergic cell groups in the caudal brainstem are involved in the
stimulus-specific (stress related) secretion of corticosterone (e.g., D arlington et
al., 1986; Szafarczyk et al., 1985; H erm an and Cullinan, 1997). These regions
have also been im plicated in the expression of diurnal corticosterone release.
Szafarczyk et al. (1985) show ed that, in addition to blunting the response to
peripheral stressors, catecholamine selective lesions produced by the injection
of 6-hydroxydopam ine into the ventral noradrenergic bundle degrade the
diurnal release of corticosterone into asynchronous ultradian rhythm s.
A lthough the inhibitory component m ay be explained partly by a G AB Aergic
projection from the suprachiasm atic nucleus (Buijs et al., 1994), an additional
source of tonic GAB Aergic inhibition is indicated by the work of Kalsbeek et al.
(1996a) who showed that adm inistration of the GABAa antagonist bicucculine
168
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
causes a dram atic increase in plasma corticosterone levels in both intact and
suprachiasm atic-lesioned animals. Interestingly, corticosterone secretion w as
of greater m agnitude and duration in the lesioned animals.
M any inputs to the DM H express GABA, most notably from the m edial p art
of the m edial preoptic nucleus, anterodorsal preoptic nucleus, parastrial nucleus,
and BST (Okamura et al., 1990). The preoptic region is interesting because it
receives a substantial projection from the suprachiasmatic nucleus, w hereas the
BST does not (Watts et al., 1987). However, the BST receives a dense input from
the ventral subiculum (Ganteras and Swanson, 1992; C ullinan et al., 1993), and
lesions of the ventral subicular projection in the fomix system abolishes the
circadian rhythm , and results in elevated basal and stress induced levels, of
plasm a corticosterone. O n the other hand, stim ulation of the ventral subiculum
inhibits stress induced corticosterone release (see Jacobsen and Sapolsky, 1991).
It has been noted previously that the output of the ventral subiculum is strikingly
similar to that of the suprachiasm atic nucleus (Ganteras and Swanson, 1992) in
terms of hypothalam ic pathw ays and projections.
Because it is generally agreed that the output of the hippocam pal form ation
is excitatory (Walles and Fonnum, 1980) and does not innervate directly the
paraventricular nucleus (Ganteras and Swanson, 1992), influences of the ventral
subiculum on corticosterone secretion are m ost probably di- or oligosynaptic.
C ullinan et al. (1993) have show n that the ventral subiculum projects onto
169
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
GABAergic neurons in the BST and that a population of BST neurons projects to
the hypothalam ic paraventricular nucleus. However, it should be borne in m ind
that the ventral subiculum also projects to several other regions that in turn
project to the DMH and that have been implicated in the stress response (the
preoptic region and the lateral septal nucleus; see H erm an and Cullinan, 1997),
and to the DMH itself.
In this context it should also be noted that the DMH is one of the m ost
prom inent GABAergic cell groups of the hypothalam us (Ferraguti et al., 1990,
and personal observation) and th at this nucleus generates extrem ely dense
intranuclear projections (Thom pson et al., 1996). The proportion of these
GABAergic cells that gives rise to long projections, or to local projections, or to
both, is unknow n. All that can be said at this tim e is that the GABAergic
projection from the suprachiasm atic nucleus does not ap p ear to end on
GABAergic neurons w ithin the DM H (Buijs et al., 1994-851) and the DMH does
n o t a p p e a r to be a m ajor G A B A ergic p ro jectio n to the h y p o th a la m ic
paraventricular nucleus (Roland and Sawchenko, 1993). It is tem pting to
speculate that the excitatory input from the ventral subiculum ends prim arily
on GABAergic intem eurons w ithin the DMH, w hich then inhibit the same,
presum ably excitatory, neurons that are disinhibited by the suprachiasm atic
nucleus during the circadian peak of corticosterone secretion.
170
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
However, it is clear th at the DMH is not the sole m odulator of corticosterone
secretion by p a rv ic ellu la r n eu ro secreto ry neurons in the hypothalam ic
p a rav en tricu lar n u cleu s. B rainstem catecholam inergic cell g ro u p s have
sig n ific a n t d ire c t p ro je c tio n s to p a rv ic e llu la r p a rts of h y p o th alam ic
paraventricular nucleus (Sawchenko and Swanson, 1982) and are involved in
reflex-like secretion of corticosterone in response to peripheral stressors. In
addition to the hippocam pal formation, the amygdala and prefrontal cortex also
have been im plicated in hypothalam ic-pituitary-adrenal axis responses to
"processive" or "psychogenic" stressors such as restraint (Herm an and Cullinan,
1997). Moreover, the DM H is bidirectionally connected w ith the nuclei of the
preoptic region (Thompson et al., 1996; present results) and each of these regions,
including the DMH, project densely to the region of the m edial parvicellular
part of the hypothalam ic paraventricular nucleus containing corticotropin-
releasing horm one neurons (Simerly and Swanson, 1988; Thom pson et al., 1996;
T hom pson and Sw anson, in p rep aratio n ). From this abbreviated list of
connections, it is apparent that the secretion of corticosterone is under complex
regulation. However, it is also clear th at the DMH is an im portant component
of this regulatory system .
Indeed, corticosterone secretion m ay be a major theme of DMH function in
that it is secreted in each of the behavioral conditions/states with w hich the
DMH is associated, including ingestion, reproduction, and therm ogenesis
(Buckingham et al., 1978). Viau and Meany, 1996; Rothwell, 1994). Most notably,
171
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
there is a corticosterone peak entrainable to restricted feeding schedules that is
independent of the suprachiasm atic nucleus and separable from the preactivity
surge of corticosterone (Krieger, 1974; Krieger et al., 1977; Honma et al., 1983).
Corticosterone is secreted also in response to disturbances in fluid balance
(D arlin g to n et al., 1986; A guilera et al., 1993; W atts, 1992). A lth o u g h
corticosterone secretion is not readily entrainable to w ater restriction, it has been
o b se rv ed th a t the p resen ce of b o th food and w a te r are n e ce ssa ry for
corticosterone to return to basal levels (Honma et al., 1986). Therefore, a system
is required to monitor both energy and fluid intake to produce the corticosterone
response appropriate for prevailing metabolic dem ands. It should be noted that
in addition to daily caloric and fluid intake, metabolic dem ands vary depending
on such factors as the point in the estrus cycle, exercise, or ambient tem perature.
Excessive dem ands upon homeostatic systems, w hich is a com ponent of w hat
is usually referred to as stress, include food deprivation (Kato et al., 1980),
dehydration or hemorrhage (Strieker et al., 1979; Darlington et al., 1986; Aguilera
et al., 1993; Watts, 1992), and pyresis (Harbuz and Lightm an, 1993), and ail
involve elevated levels of corticosterone.
However, the above account of DMH function is not complete. There are
also sig n ific a n t c ircad ian , m etabolic, and stress rela te d in flu en ces on
reproduction, and the discussion thus far does n o t account for reported
influences of the DMH on endocrine aspects of reproductive function, including
the secretion of prolactin on the afternoon of proestrus and the initiation and
172
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
m aintenance of a bim odal prolactin surge during pseudopregnancy (Pan and
Gala, 1985; see G unnet and Freeman, 1985, and Erskine, 1995 for reviews).
However, it is im portant to point out that no distinction w as m ade in these
reports between the DMH and the subjacent ventromedial hypothalam ic nucleus.
Nevertheless, it does seem likely that the DMH is influenced by m echanisms
controlling reproduction because it receives a direct input from the ventrolateral
p a rt of the v entrom edial hypothalam ic nucleus and shares bidirectional
connections w ith the anteroventral periventricular and m edial preoptic nuclei,
each im plicated in either behavioral or neuroendocrine aspects of reproduction
(Larsson, 1979; Pfaff, 1980; Chateau et al., 1981; W éigand and Terasawa, 1982;
Sachs and Miselis, 1988; Cohen and Pfaff, 1992)
It is also clear that possible DMH influences upon thyrotropin-releasing
horm one (TRH) secretion need to be investigated further. The DM H projects
densely to the region of the hypothalamic paraventricular nucleus containing
TRH neurons that control thyroid-stim ulating horm one secretion from the
anterior pituitary (Thompson et al., 1996), and it has been show n th at the DMH
is sensitive to circulating levels of thyroid horm ones (Ceccatelli et al., 1992).
Moreover, thyroid horm one levels demonstrate several properties indicative of
DMH influence—they are secreted in a diurnal rhythm (Abe et al., 1979), and in
response to food restriction (Campbell et al., 1977; Blake et al., 1991), and
following cold exposure (M cCarthy et al., 1993; Fregly, 1990).
173
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
By necessity, this view of DMH function is som ew hat speculative and
incomplete. However, it does indicate that the role of the DMH in corticosterone
secretion and other aspects of hom eostasis is not sim ple and m ay depend on
other aspects of the anim al's physiological state. These aspects m ay include
variables such as the metabolic state of the anim al (including hydration state,
and nutritional zmd therm ogenic requirem ents), point in the reproductive cycle,
or behavioral state vis-à-vis the sleep-wake cycle. A lthough corticosterone is
secreted under each of these conditions, it rem ains to be determ ined w hether
corticosterone secretion per se provides an organizing theme for DMH function
or is a component of more broad influences upon integrated endocrine and
autonom ic function associated w ith goal-directed behavior.
174
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
LITERATURE CITED
Abe, K., J. Kroning, M.A. Greer, and V. Critchlow (1979) Effects of destruction of
the suprachiasmatic nuclei on the circadian rhythms in plasma corticosterone,
body temperature, feeding and plasma thyrotropin. Neuroendocrinology 29:119-
131.
Aguilera, G., S.L. Lightm an, and A. Kiss (1993) Regulation of the hypothalam ic-
pituitary-adrenal axis during w ater deprivation. Endocrinology 132:241-248.
Akabayashi, A., C. Wahlestedt, J.T. Alexander, and S.F. Leibowitz (1994) Specific
inhibition of endogenous neuropeptide-Y synthesis in arcuate nucleus by
antisense oligonucleotides suppresses feeding behavior and insulin secretion.
Brain Res. Mol. Brain Res. 21:55-61.
Allen, G.V., an d D.F. Cechetto (1993) Functional and anatomical organization of
cardiovascular pressor and depressor sites in the lateral hypothalam ic area,
n. A scending projections. J. Comp. Neurol. 330:421-438.
Allen, G.V., C.B. Saper, K.M. Hurley, and D.F. Cechetto (1991) O rganization of
visceral and limbic connections in the insular cortex of the rat. J. Com p.
Neurol. 311:1-16.
Bai, F.L., M. Yamano, Y. Shiotani, P.C. Emson, A.D. Smith, J.F. Powell, and M.
Tohyama (1985) An arcuato-paraventricular and -dorsom edial hypothalam ic
neuropeptide Y-containing system which lacks noradrenaline in the rat. Brain
Res. 331:172-175.
Beckstead, R.M. (1979) A n autoradiographic exam ination of corticocortical and
subcortical projections of the m ediodorsal-projection (prefrontal) cortex in
the rat. J. Com p. Neurol. 184:43-62.
Bellinger, L.L., L.L. Bemardis, and V.E. Mendel (1976) Effect of ventrom edial
and dorsom edial hypothalam ic lesions on circadian corticosterone rhythm s.
Neuroendocrinology 22:216-225.
Bellinger, L.L., V.E. Mendel, L.L. Bemardis, and T.W. Castonguay (1986) Meal pat
terns of rats w ith dorsomedial hypothalamic nuclei lesions or sham operations.
Physiol. Behav. 36:693-698.
Berk, M.L., an d J.A. Finkelstein (1981a) Afferent projections to the preoptic area
and hypothalam ic regions in the rat brain. Neuroscience 6:1601-1624.
175
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Berk, M.L., and J.A. Finkelstein (1981b) An autoradiographic determination of the
efferent projections of the suprachiasmatic nucleus of the hypothalamus. Brain
Res. 226:1-13.
Berk, M.L., and J.A. Finkelstein (1982) Efferent connections of the lateral hypo
thalamic area of the rat: An autoradiographic investigation. Brain Res. Bull.
6:511-526.
Bemardis, L.L. (1972) Hypophagia, hypodipsia, and hypoactivity following elec
trolytic lesions in the dorsom edial hypothalam ic nuclei of m ature rats of
both sexes. J. N eural Trans. 33:1-10.
Bemardis, L.L., and L.L. Bellinger (1987) The dorsom edial hypothalam ic nucleus
revisited: 1986 update. Brain Res. 434:321-381.
Blake, N.G., M R. Johnson, D.J.A. Eckland, O.J.F. Foster, and S.L. Lightm an (1992)
Effect of food deprivation and altered thyroid status on the hypothalam ic-
pituitary-thyroid axis in the rat. J. Endocrinol. 133:183-188.
Bobillier, P., S. Seguin, A. Degueurce, B.D. Lewis, and J.F. Pujol (1979) The efferent
connections of the nucleus raphe centralis superior in the rat as revealed by
radioautography. Brain Res. 166:1-8.
Brittain, D A. (1988) The efferent connections of the infralimbic area in the rat.
Dissertation. U niversity of California, San Diego.
Brownstein, M.J., and Palkovits, M (1984) Catecholamines, serotonin, acetylcho
line, and g-am inobutyric acid in the rat brain: biochemical studies. In A.
Bjorklund and T. Hokfelt (eds.): Handbook of Chemical Neuroanatomy Classical
Transmitters in the CNS: Part I. Amsterdam: Elsevier Science,Vol. 2, pp. 23-54.
Buckingham, J.C., K.-D. Dohler, and C.A. W ilson (1978) Activity of the pitu
itary-adrenocortical system and thyroid gland during the oestrous cycle of
the rat. J. Endocrinol. 78:359-366.
Buijs, R.M., Y-X. H ou, S. Shinn, and L.P. Renaud (1994) U ltrastructural evidence
for intra- and extranuclear projections of GABAergic n eu ro n s of the
surprachiasm atic nucleus. J. Comp. Neurol. 340:381-391.
Buijs, R.M., M. M arkm an, B. Nunes-Cardoso, Y-X. H ou, and S. Shirm (1993) Pro
jections of the suprachiasm atic nucleus to stress-related areas in the rat
hypothalam us: A light and electron microscopic study. J. Com p. Neurol.
335:42-54.
176
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Calingasan, N.Y., and S. Ritter (1993) Lateral parabrachial subnucleus lesions
abolish feeding induced by m ercaptoacetate but not by 2-deoxy-D-glucose.
Am. J. Physiol. 265:R1168-R1178.
Cameron, A.A., LA. Khan, K.N. Westlund, K.D. Cliffer, and W.D. Willis (1995)
The efferent projections of the periaqueductal gray in the rat: A Phaseolus
üw/^ans-leucoagglutinin study. I. Ascending projections. J. Comp. Neurol.
351:568-584.
Campbell, G.A., M. Kurcz, S. M arshall, and J. Meites (1977) Effects of starvation
in rats on serum levels of follicle stim ulating horm one, luteinizing hormone,
thyrotropin, grow th horm one and prolactin; Response to LH-releasing hor
mone and thyrotropin-releasing horm one. Endocrinology 100:580-587.
Canteras, N.S., R.B. Simerly, and L.W. Swanson (1992) Projections of the ventral
prem am m illary nucleus. J. Comp. Neurol. 324:195-212.
Canteras, N.S., R.B. Simerly, and L.W. Swanson (1994) O rganization of projec
tions from the ventromedial and tuberal nuclei of the hypothalamus: A PHAL
study in the rat. J. Comp. N eurol. 343:41-79.
Canteras, N.S., R.B. Simerly, and L.W. Sw anson (1995) O rganization of projec
tions from the medial nucleus of the am ygdala: A PHAL study in the rat.
J.Comp.Neurol. 360:213-245.
Canteras, N.S., and L.W. Sw anson (1992) Projections of the ventral subiculum
to the am ygdala, septum , and hypothalam us: A PHAL anterograde tract-
tracing study in the rat. J. Com p. Neurol. 324:180-194.
Ceccatelli, S., L. Giardino, and L. Calza (1992) Response of hypothalam ic pep
tide mRNAs to thyroidectomy. Neuroendocrinology 56:694-703.
Chateau, D., S. Plas-Roser, and C. Aron (1981) Follicular growth, ovulatory phe
nomena and ventrom edial nucleus lesions in the rat. Endokrinologie 77:257-
268.
Chiba, T., and Y. M urata (1985) Afferent and efferent connections of the m edial
preoptic area in the rat: a WGA-HRP study. Brain Res. Bull. 14:261-272.
d iffe r, K.D., R. Burstein, and G.J. Giesler, Jr. (1991) D istributions of spino
thalamic, spinohypothalam ic and spinotelencephalic fibers revealed by
anterograde transport of PHA-L in rats. J. Neurosci. 11:852-868.
177
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Cohen, R.S., and D.W. Pfaff (1992) Ventromedial hypothalam ic neurons in the
m ediation of long-lasting effects of estrogen on lordosis behavior. Prog.
Neurobiol. 35:423-453.
Coolen, L.M., H.J.P.W. Peters, and J.G. Veening (1996) Fos imm imoreactivity in
the rat brain following consum m atory elements of sexual behavior: A sex
comparison. Brain Res. 735:67-82.
Cullinan, W.E., D.L. Helmreich, and S.J. Watson (1996) Fos expression in fore
brain afferents to the hypothalam ic paraventicular nucleus following swim
stress. J. Comp. Neurol. 365:88-99.
Cullinan, W.E., J.P. Herman, and S.J. W atson (1993) Ventral subicular interaction
w ith the hypothalam ic paraventricular nucleus: Evidence for a relay in the
bed nucleus of the stria term inalis. J. Comp. N eurol. 332:1-20.
Dalton, L.D., R.G. Carpenter, and S.P. Grossman (1981) Ingestive behavior in adult
rats w ith dorsomedial hypothalamic lesions. Physiol. Behav. 26:117-123.
D arlington, D.N., J. Shinsako, and M.F. Dallman (1986) M edullary lesions
elim inate ACTH responses to hypotensive hem orrhage. Am. J. Physiol.
251 :R106-R115.
Dashti, A.E., S. Ghandeharizadeh, J. Stone, L.W. Swanson, and R.H. Thom pson
(1997) Database challenges and solutions in neuroscientific applications.
N euroim age 5:97-115.
Eberhart, J.A., J.I. Morrell, M.S. Krieger, and D.W. Pfaff (1985) An autoradio
graphic study of projections ascending from the m idbrain central gray, and
from the region lateral to it, in the rat. J. Comp. N eurol. 241:285-310.
Elmquist, J.K., T.E. Scammell, C D. Jacobson, and C.B. Saper (1996) Distribution
of Fos-like im m unoreactivity in the rat brain follow ing intravenous li-
popolysaccharide adm inistration. J. Comp. Neurol. 371:85-103.
Erskine, M.S. (1995) Prolactin release after mating and genitosensory stim ula
tion in females. Endocr. Rev. 16:508-28.
Everitt, B.J., T. Hokfelt, L. Terenius, K. Tatemoto, V. M utt, and M. G oldstein
(1984) Differential co-existence of neuropeptide Y (NPY)-like im m unore
activity w ith catecholamine sin the central nervous system of the rat. N eu
roscience 11:443-462.
178
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fahrbach, S.E., J.I. Morrell, and D.W. Pfaff (1984) Temporal pattern of HR? spread
from an inotophoretic deposit site and description of a new HRP-gel im
p lan t m ethod. J. Com p. Neurol. 225:605-619.
Ferraguti, F., M. Zoli, M. Aronsson, L.F. Agnati, M. G oldstein, D. Filer, and K.
Fuxe (1990) D istribution of glutamic acid decarboxylase m essenger RNA-
containing nerve cell populations of the male rat brain. J. Chem. Neuroanat.
3:377-396.
Fregly, M.J. (1990) Activity of the hypothalam ic-pituitary-thyroid axis during
exposure to cold. In E. Schonbaum and P. Lomax (eds.): Therm oregulation:
Physiology and Biochemistry. New York: Pergamon Press, pp. 437-494.
Friedm an, M.I., M.G. Tordoff, and 1 . Ramirez (1986) Integrated metabolic con
trol of food intake. Brain Res. Bull. 17:855-859.
Fulwiler, G.E., and C.B. Saper (1984) Subnuclear organization of the efferent
cormections of the parabrachial nucleus in the rat. Brain Res. 319:229-259.
Gallo, R.V. (1981) Effect of electrical stim ulation of the dorsom edial hypotha
lamic nucleus on pulsatile LH release in ovariectom ized rats. N euroendo
crinology 32:134-138.
Gerfen, C.R., and P.E. Sawchenko (1984) An anterograde neuroanatom ical trac
ing m ethod that show s the detailed m orphology of neurons, their axons
and terminals: immunohistochemical localization of an axonally transported
plan t lectin, Phaseolus vulgaris leucoagglutinin (PHA-L). Brain Res. 290:219-
238.
Giesler, G.J.,Jr., J.T. Katter, and R.J. Dado (1994) Direct spinal pathw ays to the
limbic system for nociceptive information. TINS 17:244-250.
G rill, H.J., M l. F riedm an, R. N orgren, G. Scalera, a n d R. Seeley (1995)
P arabrachial nucleus lesions im pair feeding response elicited by 2,5-
anhydro-D-m annitol. Am . J. Physiol. 268:R676-R682.
Grove, E.A. (1988) Efferent connections of the substantia innom inata in the rat.
J. Com p. Neurol. 277:347-364.
Gu, G.B., and R.B. Sim erly (In Press) Projections of the sexually dim orphic
anteroventral periventricular nucleus in the female rat. J. Comp. Neurol.
179
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
G unnet, J.W., and M.E. Freem an (1983) The m ating-induced release of prolac
tin: A unique neuroendocrine response. Endocr. Rev. 4:44-61.
G unnet, J.W., and M.E. Freem an (1985) The interaction of the m edial preoptic
area and the dorsom edial-ventrom edial nuclei of the hypothalam us in the
regulation of the m ating-induced release of prolactin. N euroendocrinology
40:232-237.
H arbuz, M.S., and S.L. Lightm an (1993) Stress and the hypothalam o-pituitary-
adrenal axis: acute, chronic and immunological activation. J. Endocrinol.
134:327-339.
Herbert, H., M.M. M oga and C.B. Saper (1990) Connections of the parabrachial
nucleus w ith the nucleus of the solitary tract and the m edullary reticular
form ation in the rat. J. Comp. Neurol. 293:540-580.
H erkenham , M., and W.J.H. N auta (1979) Efferent connections of the habenular
nuclei in the rat. J. Com p. Neurol. 137:19-48.
Herm an, J.P., and W.E. C ullinan (1997) Neurocircuitry of stress: central control
of the hypothalam o-pituitary-adrenocortical axis. TINS 20:78-84.
Holets, V.R., T. Hoekfelt, A. Roekaeus, L. Terenius, and M. Goldstein (1988) Locus
coeruleus neurons in the rat containing neuropeptide Y, tyrosine hydroxy
lase or galanin and their efferent projections to the spinal cord, cerebral
cortex and hypothalam us. Neuroscience 24:893-906.
Honma, K-L, C. vonGoetz, and J. Aschoff (1983) Effects of restricted daily feed
ing on freerunning circadian rhythm s in rats. Physiol. Behav. 30:905-913.
Honm a, K., S. H onm a, T. H irai, Y. K atsuno, and T. H iroshige (1986) Food
ingestion is more im portant to plasm a corticosterone dynam ics than w ater
intake in rats under restricted daily feeding. Physiol. Behav. 37:791-795.
Hurley, K.M., H. H erbert, M.M. Moga, and C.B. Saper (1991) Efferent projec
tions of the infralimbic cortex of the rat. J. Comp. Neurol. 303:249-276.
Imaki, T., T. Shibasaki, M. H otta, and H. D em ura (1993) Intracerebroventricular
adm inistration of corticotropin-releasing factor induces c-fos m RNA expres
sion in brain regions related to stress responses: Com parison w ith pattern
of c-fos mRNA induction after stress. Brain Res. 616:114-125.
180
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Inglefield, J.R., S.B. Schwarzkopf, and C.K. Kellogg (1994) Alterations in behav
ioral responses to stressors following excitotoxin lesions of dorsom edial
hypothalam ic regions. Brain Res. 633:151-161.
Jacobson, L., and R. Sapolsky (1991) The role of the hippocam pus in feedback
regulation of the hypothalam ic-pituitary-adrenocortical axis. Endocr. Rev.
12:118-134.
Jhanwar-Uniyal, M., B. Beck, Y.S. Jhanwar, C. Burlet, and S.F. Leibowitz (1993)
N europeptide Y projection from arcuate nucleus to parvocellular division
of paraventricular nucleus: specific relation to the ingestion of carbohydrate.
Brain Res. 631:97-106.
Johnson, A.K. (1985) Role of the p eriv en tricu lar tissue su rro u n d in g the
anteroventral third ventricle (AV3V) in the regulation of body fluid hom eo
stasis. In R.W. Schrier (ed.): Vasopressin. New York: Raven Press, pp. 319-331.
Jones, B E., and R.Y. Moore (1977) Ascending projections of the locus coeruleus
in the rat. U. Autoradiographic study. Brain Res. 127:23-53.
Kalsbeek, A., R.M. Buijs, J.J. van Heerikhuize, M. Arts, and T.P. vander Woude
(1992) Vasopressin-containing neurons of the suprachiasm atic nuclei inhibit
corticosterone release. Brain Res. 580:62-67.
Kalsbeek, A., W.-J. Drijfhout, B.H.C. Westerink, J.J. van Heerikhuize, T.P. Van
Der W oude, J. van der Vliet, and R.M. Buijs (1996a) GABA receptors in the
region of the dorsom edial hypothalam us of rats are im plicated in the con
trol of m elatonin and corticosterone release. N euroendocrinology 63:69-78.
Kalsbeek, A., J.J. van Heerikhuize, J. Wortel, and R.M. Buijs (1996b) A diurnal
rhythm of stim ulatory input to the hypothalam o-piuitary-adrenal system
as revealed by tim ed intrahypothalam ic adm inistration of the vasopressin
Vj antagonist. J. Neurosci. 16:5555-5565.
Kato, H., M. Saito, and M. Suda (1980) Effect of starvation on the circadian
adrenocortical rhythm in rats. Endocrinology 106:918-921.
Keim, S.R., and A. Shekhar (1996) The effects of GABA^ receptor blockade in the
dorsom edial hypothalam ic nucleus on corticotrophin (ACTH) and corticos
terone secretion in m ale rats. Brain Res. 739:46-51.
Kita, H., and Y . Oomura (1982) An HRP study of the afferent connections to rat
m edial hypothalam ic region. Brain Res. Bull. 8:53-62.
181
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Krieger, D.T. (1974) Food and w ater restriction shifts corticosterone, tem pera
ture, activity and brain am ine periodicity. Endocrinology 95:1195-1201.
Krieger, D.T., H . Hauser, and L. Krey (1977) Suprachiasm atic nuclear lesions do
not abolish food-shifted circadian adrenal and terperature rhythmicity.
Science 197:398-399.
Krukoff, T.L., K.H. Harris, and J.H. Jham andas (1993) Efferent projections from
the parabrachial nucleus dem onstrated w ith the anterograde tracer Phaseolus
vulgaris leucoagglutinin. Brain Res.Bull. 30:163-172.
Larsson, K. (1979) Features of the neuroendocrine regulation of masculine sexual
behavior. In C. Beyer (ed.): Endocrine control of sexual behavior. New York:
Raven Press, pp. 77-163.
LeDoux, J.E., D.A. Ruggiero, and D.J. Reis (1985) Projections to the subcortical
forebrain from anatomically defined regions of the m edial geniculate body
in the rat. J. Comp. Neurol. 242:182-213.
L eib o w itz, S.F., S. D iaz, a n d D. Tem pel (1989) N o re p in e p h rin e in th e
paraventricular nucleus stimulates corticosterone release. Brain Res. 496:219-
227.
Li, B.-H., B. Xu, N.E. Rowland, and S.P. Kalra (1994) c-fos expression in the rat
brain follow ing central adm inistration of neuropeptide Y and effects of food
consum ption. Brain Res. 665:277-284.
Lind, R.W., L.W. Swanson, and D. Ganten (1984) Angiotensin II im m unoreactiv
ity in the neural afferents and efferents of the subfornical organ of the rat.
Brain Res. 321:209-215.
Lind, R.W., G.W. Van Hoesen, and A.K. Johnson (1982) An HRP study of the
connections of the subfornical organ of the rat. J. Comp Neurol. 210:265-
277.
Loewy, A.D., J.H. Wallach, and S. McKellar (1981) Efferent connections of the
ventral m edulla oblongata in the rat. Brain Res. Rev. 228:63-80.
Luiten, P.G., and P. Room (1980) Interrelations between lateral, dorsom edial
and ventrom edial hypothalam ic nuclei in the rat. An HRP study. Brain Res.
190:321-332.
182
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Luiten, P.G., G.J. Ter Horst, H. Karst, and A.B. Steffens (1985) The course of
paraventricular hypothalamic efferents to autonom ic structures in m edulla
and spinal cord. Brain Res. 329:374-378.
Luiten, P.G.M., G.J. Ter Horst, and A.B. Steffens (1987) The hypothalam us, intrinsic
connections and outflow pathways to the endocrine system in relation to the
control of feeding and metabolism. Prog. Neurobiol. 28:1-54.
Magoul, R., P. Ciofi, and G. Tramu (1994) Visualization of an efferent projection
route of the hypothalam ic rat arcuate nucleus through the stria term inalis
after labeling w ith carbocyanine dye (Dil) or proopiom elanocortin-im m im o-
histochemistry. Neurosci. Lett. 172:134-138.
McCarthy, H.D., A.P. Kilpatrick, P. Trayhum, and G. W illiams (1993) W idespread
increases in regional hypothalamic neuropeptide Y levels in acute cold-
exposed rats. Neuroscience 54:127-132.
McKellar, S., and A.D. Loewy (1982) Efferent projections of the A1 catecholamine
cell group in the rat: an autoradiographic study. Brain Res. 241:11-29.
M illhouse, O.E. (1973) The organization of the ventrom edial hypothalam ic
nucleus. Brain Res. 55:71-87.
M oga, M.M., R.P. Weis, and R.Y. Moore (1995) Efferent projections of the
paraventricular thalamic nucleus in the rat. J. Comp. N eurol. 359:221-238.
Moore, R.Y., and J.P. Card (1984) Noradrenaline-containing neuron systems. In
A. Bjorklund and T. Hokfelt (eds.): Handbook of Chem ical N euroanatom y
Vol. 2 Classical Transmitters in the CNS: Part I. Am sterdam : Elsevier Science,
pp. 123-156.
Moore, R.Y., A.E. Halaris, and B E. Jones (1978) Serotonin neurons of the m id
brain raphe: Ascending projections. J. Comp. Neurol. 180:417-438.
Niijima, A. (1989) N ervous regulation of metabolism. Prog. Neurobiol. 33:135-
147.
O kam ura, H ., M. A bitbol, J.F. Julien, S. Dum as, A. B erod, M. G effard, K.
Kitahama, P. Bobillier, J. Mallet, and L. W iklund (1990) N eurons containing
m essenger RNA encoding glutam ate decarboxylase in rat hypothalam us
dem onstrated by in situ hybridization, w ith special em phasis on cell groups
in m edial preoptic area, anterior hypothalam ic area and dorsom edial hypo
thalam ic nucleus. Neuroscience 39:675-699.
183
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Palkovits, M., L. Zaborsky, A. Feminger, É. Mezey, M.I.K. Fekete, J.P. H erm an,
B. Kanyicska, and D. Szabo (1980) Noradrenergic iim ervation of the rat
hypothalam us; Experim ental biochemical and electron microscopic studies.
Brain Res. 191:161-171.
Pan, J.T., and R.R. Gala (1985) Central nervous system regions involved in the
estrogen-induced afternoon prolactin surge. I. Lesion studies. Endocrinology
117:382-387.
Panula, P., U. Pirvola, S. Auvinen, and M.S. A iraksinen (1989) H istam ine-
im m unoreactive nerve fibers in the rat brain. Neuroscience 28:585-610.
Pezzone, M.A., W.-S. Lee, G.E. Hoffman, and B.S. Rabin (1992) Induction of c-
fos im m unoreactivity in the rat forebrain by conditioned and unconditioned
aversive stim uli. Brain Res. 597:41-50.
Pfaff, D.W. (1980) Estrogens and brain function. New York:Springer Verlag.
Polston, E.K., and M.S. Erskine (1995) Patterns of induction of the im m ediate-
early genes c-fos and egr-1 in the female rat brain following differential
am ounts of m ating stim ulation. Neuroendocrinology 62:370-384.
Ricardo, J.A., and E.T. Koh (1978) Anatomical evidence of direct projections from
the nucleus of the solitary tract to the hypothalam us, am ygdala, and other
forebrain structures in the rat. Brain Res. 153:1-26.
Risold, P.Y., N.S. Ganteras, and L.W. Swanson (1994) Organization of projec
tions from the anterior hypothalamic nucleus: A PHAL study in the rat. J.
Comp. Neurol. 348:1-40.
Risold, P.Y., and L.W. Swanson (In Press) Connections of the rat lateral septal
complex. Brain Res. Rev.
Roland, B.L., and P.E. Sawchenko (1993) Local origins of some GABAergic pro
jections to the paraventricular and supraoptic nuclei of the hypothalam us
in the rat. J. Comp. Neurol. 332:123-143.
Rothwell, N.J. (1994) CNS regulation of thermogenesis. Crit. Rev. Neurobiol.
8:1- 10.
Sachs, B.D., and R.L. M iselis (1988) The physiology of male sexual behavior. In
E. Knob il and J.D. Neill (eds.): The physiology of reproduction. N ew York:
Raven Press, pp. 1393-1485.
184
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Sagar, S.M., K.J. Price, N.W. Kasting, and F.R. Sharp (1995) Anatomic patterns
of Fos im m unostaining in rat brain following systemic endotoxin adm inis
tration. Brain Res. Bull. 36:381-392.
Sahu, A., S.P. Kalra, W.R. Crowley, and P.S. Kalra (1988) Evidence that NPY-
containing neurons in the brainstem project into selected hypothalam ic
nuclei: im plication in feeding behavior. Brain Res. 457:376-378.
Saper, C.B., and A.D. Loewy (1980) Efferent connections of the parabrachial
nucleus in the rat. Brain Res. 197:291-317.
Saper, C.B., L.W. Swanson, and W.M. Cowan (1979) An autoradiographic study
of the efferent connections of the lateral hypothalam ic area in the rat.
J.Comp .Neurol. 183:689-706.
Sawchenko, P.E., and L.W. Swanson (1982) The organization of noradrenergic
pathw ays from the brainstem to the paraventricular and supraoptic nuclei
in the rat. Brain Res. 257:275-325.
Sawchenko, P.E., L.W. Swanson, R. G rzanna, PR. Howe, S.R. Bloom, and J.M.
Polak (1985) C olocalization o f n eu ro p ep tid e Y im m u n o reactiv ity in
brainstem catecholaminergic neurons that project to the paraventricular
nucleus of the hypothalam us. J.Comp.Neurol. 241:138-153.
Scammell, T.E., J.K. Elmquist, J.D. Griffin, and C.B. Saper (1996) Ventromedial
preoptic prostaglandin E2 activates fever-producing autonomic pathw ays.
J. Neurosci. 16:6246-6254.
Schmued, L.C., and J.H. Fallon (1986) Fluoro-Gold: a new fluorescent retrograde
axonal tracer w ith num erous unique properties. Brain Res. 377:147-154.
Sesack, S.R., A.Y. Deutch, R.H. Roth, and B.S. Bunney (1989) Topographical
organization of the efferent projections of the m edial prefrontal cortex in
the rat: A n a n te ro g rad e tra c t-tra c in g stu d y w ith Phaseolus vulgaris
leucoagglutinin. J. Comp. Neurol. 290:213-242.
Shekhar, A. (1996) GABA blockade in the region of the dorsom edial hypo
thalam us regulates 'anxiety' in rats in the elevated plus-maze: I. Behavioral
measures. Brain Res. 627:9-16.
Shekhar, A., and J.S. Katner (1995) GABA^ receptors in the dorsom edial hypo
thalam us of rats regulate anxiety in the social interaction test. Pharm .
Biochem. Behav. 50:253-258.
185
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Sherin, J.E., P.J. Shiromani, R.W. McCarley, and C.B. Saper (1996) Activation of
ventrolateral preoptic neurons during sleep. Science 271:216-219.
Sim, L.J., and S.A. Joseph (1991) Arcuate nucleus projections to brainstem regions
which m odulate nociception. J. Chem. Neuroanat. 4:97-109.
Staines, Wm.A., T. Yamamoto, P.E. D oddona, and J.I. N agy (1987) The hypo
thalam us receives m ajor projections from the tuberom am m illary nucleus in
rat. Neurosci. Lett. 76:257-262.
Stotz-Potter, E.H., S.M. M orin, and J.A. Dimicco (1996a) Effect of microinjection
of m uscim ol into the dorsom edial or paraventricular hypothalam ic nucleus
on air stress-induced neuroendocrine and cardiovascular changes in rats.
Brain Res. 742:219-224.
Stotz-Potter, E.H., L.R. W illis, and J.A. Dimicco (1996b) M uscim ol acts in
dorsom edial b u t not paraventricular hypothalm ic nucleus to suppress
cardiovascular effects of stress. J. Neurosci. 16:1173-1179.
Strieker, E.M., A.H. Vagnucci, R.H. M cDonald, and F.H. Leenan (1979) Renin
and aldosterone secretions during hypovolem ia in rats: Relation to NaCl
intake. Am. J. Physiol. 237:R45-R51.
Swanson, L.W. (1976) A n autoradiographic study of the efferent connecitons of
the preoptic region in the rat. J. Comp. Neurol. 167:227-256.
Swanson, L.W. (1987) The hypothalam us. In A. Bjorklun, T. Hokfelt and L.W.
Swanson (eds.): H andbook of Chemical Neuroanatom y, Vol. 5 Integrated
systems of the CNS, Part I. Amsterdam: Elsevier Science, pp. 1-124.
S w anson, L.W. (1992) B rain M aps: S tru c tu re o f th e R at B rain.
Amsterdam:Elsevier.
Sw anson, L.W., and W.M. Cowan (1975) The efferent connections of the
suprachiasm atic nucleus of the hypothalam us. J. Com p. Neurol. 160:1-12.
Swanson, L.W., and W.M. Cow an (1979) The connections of the septal region in
the rat. J. Comp. N eurol. 186:621-656.
Swanson, L.W., and B.K. H artm an (1975) The central adrenergic system. An
immunofluorescence study of the location of cell bodies and their efferent
connections in the rat utilizing dopam ine-B-hydroxylase as a marker. J.
Comp. Neurol. 163:467-506.
186
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Swanson, L.W., and R.W. Lind (1986) N eural projections subserving the initia
tion of a specific m otivated behavior in the rat: New projections from the
subfornical organ. Brain Res. 379:399-403.
Szafarczyk, A., G. Alonso, G. Ixart, F. Malaval, and I. Assenmacher (1985) Diumal-
stimulated and stress-induced ACTH release in rats is mediated by ventral
noradrenergic bundle. Am. J. Physiol. 249:E219-E226.
Ter H orst, G.J., P. de Boer, P.G. Luiten, and J.D. van Willigen (1989) Ascending
projections from the solitary tract nucleus to the hypothalam us. A Phaseolus
vulgaris lectin tracing study in the rat. Neuroscience 32:785-797.
Ter H orst, G.J., and P.G. Luiten (1987) Phaseolus vulgaris leucoagglutinin tracing
of intrahypothalam ic cormections of the lateral, ventrom edial, dorsom edial
and paraventricular hypothalam ic nuclei in the rat. Brain Res. Bull. 23:191-
203.
Thompson, R.H., N.S. Ganteras, and L.W. Swanson (1996) Organization of pro
jections from the dorsom edial nucleus of the hypothalam us: A PHA-L study
in the rat. J. Comp. Neurol. 376:143-173.
Thompson, R.H., and L.W. Swanson (in preparation) Anatomical substrate for
hypothalam ic pattern generators. Brain Res. Rev.
Tordoff, M.G., N. Rawson, and M L Friedm an (1991) 2,5-anhydro-D-mannitol
acts in liver to initiate feeding. Am. J. Physiol. 262:R283-R288.
Van de Kar, L.D., and S.A. Lorens (1979) Differential serotonergic innervation
of individual hypothalam ic nuclei and other forebrain regions by the dor
sal and m edian m idbrain raphe nuclei. Brain Res. 262:45-54.
Vertes, R.P. (1991) A PHA-L analysis of ascending projections of the dorsal raphe
nucleus in the rat. J, Comp. Neurol. 323:643-668.
Vertes, R.P. (1992) PHA-L analysis of projections from the supram am m illary
nucleus in the rat. J. Comp. Neurol. 326:595-622.
Vertes, R.P., A.M. Crane, L.V Colom, and B.H. Bland (1995) Ascending projec
tions of the posterior nucleus of the hypothalam us: PHA-L analysis in the
rat. J .Comp. Neurol. 359:90-116.
Vertes, R.P., G.F. Martin, and R. W altzer (1986) An autoradiographic analysis of
ascending projections from the m edullary reticular formation in the rat.
Neuroscience 29:873-898.
187
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Viau, V., and M.J. M eaney (1996) The inhibitory effect of testosterone on hypo-
thalam ic-pituitary-adrenal responses to stress is m ediated by the medial
preoptic area. J. Neurosci. 16:1866-1876.
Wada, H., N. Inagaki, A. Yamatodani, and T. Watanabe (1991) Is the histaminergic
neuron system a regulatory center for whole-brain activity? TINS 14:415-418.
W alaas, I., and F. Fonnum (1980) Biochemical evidence for glutam ate as a trans
m itter in hippocam pal efferents to the basal forebrain and hypothalam us in
the rat brain. Neuroscience 5:1691-1698.
W ang, Y., S.L. Cum m ings, and D.W. Gietzen (1996) Temporal-spatial pattern of
c-fos expression in the rat b rain in response to indispensable amino acid
deficiency I. The initial recognition phase. Brain Res. Mol. Brain Res. 40:27-
34.
W atts, A.G. (1992) Disturbance of fluid hom eostasis leads to tem porally and
anatomically distinct responses in neuropeptide and tyrosine hydroxylase
m RNA levels in the paraventricular and supraoptic nuclei nuclei in the rat.
Neuroscience 46:859-879.
Watts, A.G., and L.W. Swanson (1987) Efferent projections of the suprachiasmatic
nucleus: H. Studies using retrograde transport of fluorescent dyes and si
m ultaneous peptide im m unohistochem istry in the rat. J. Comp. Neurol.
258:230-252.
W atts, A.G., L.W. Swanson, and G. Sanchez Watts (1987) Efferent projections of
the suprachiasm atic nucleus: I. Studies using anterograde transport of
Phaseolus vulgaris leucoagglutinin in the rat. J. Comp. Neurol. 258:204-229.
W iegand, S.J., and E. Terasawa (1982) Discrete lesions reveal functional hetero
geneity of suprachiasmatic structures in regulation of gonadotropin secretion
in the female rat. N euroendocrinology 34:395-404.
Yasui, Y., K. Nakano, and N. M izuno (1992) Descending projections from the
subparafascicular thalam ic nucleus to the lower brain stem in the rat. Exp.
Brain Res. 90:508-518.
188
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Chapter IV.
General Discussion
The resu lts p resen ted in this thesis describe the c o n n ectio n s of the
dorsom edial nucleus of the hypothalam us (DMH) and the pathw ays used by
the DMH and its inputs. By com paring these results w ith those concerning the
projections of the hypothalam ic m edial zone nuclei reported recently w ith the
PHAL m ethod, it will also been show n that all hypothalam ic nuclei use a
restricted set of intrahypothalam ic or "propriohypothalam ic" pathw ays.
Moreover, such a comparison leads to considerable insights into the overall
organization of the hypothalamic periventricular and medial zones and suggests
an anatom ical framework for the coordination of endogenous (diurnal) and
stim ulus-induced visceromotor responses and their integration w ith behavioral
processes.
Connections of the DMH
The DMH has been implicated in a wide variety of responses. These primarily
include endocrine and autonomic aspects of ingestive (Bemardis, 1972; Dalton
et al., 1981, Wang, et al., 1996); reproductive (Gallo, 1981; G unnet and Freeman,
1985; Polston and Erskine,1995; Coolen et al.,1996; also see G unnet and Freeman,
1983 and Erskine, 1995 for reviews); stress (Keim and Shekhar, 1996; Stotz-Potter
et al., 1996a,b; Inglefield et al., 1994-1218; Shekhar, 1993; Shekhar and Katner,
1995; Kalsbeek et al., 1996a); circadian (Bellinger et al., 1976,1986; Buijs et al.,
189
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
1993b; Kalsbeek et al., 1992, 1996b); and therm ogenic (McCarthy et al., 1993;
Sagar et al.,1995; Elmquist et al.,1996) function. These responses overlap to a
large degree at the visceral level (for exam ple, all involve v ariatio n s in
corticosterone secretion, see below) and can be better understood as a coherent
function once the circuits that involve the DMH are understood.
The connections of the DMH, b o th efferent a n d afferent, are largely
intrahypothalam ic, although significant projections are also generated in the
te le n c e p h a lo n and b ra in ste m (C h a p te rs 2 a n d 3, resu lts). W ith in the
hypothalam us, the DMH shares m oderate to dense bidirectional connections
w ith th e p a ra stria l, p reo p tic su p ra ch ia sm atic, a n te ro v e n tra l p re o p tic ,
anterodorsal preoptic, and anteroventral periventricular nuclei in the preoptic
region, anterior parvicellular part of paraventricular hypothalam ic nucleus and
dorsal tuberom am m illary nucleus, and undifferentiated regions of the m edial
preoptic area. The DMH also projects to the lateral p a rt of the m edial preoptic
n u c le u s, th ro u g h o u t th e p a rv ic e llu la r d iv isio n of the p a ra v e n tric u la r
hypothalam ic nucleus, and the retrochiasmatic area and perifom ical region. The
DM H receives additional input from the vascular organ of the lam ina term inalis,
arcuate nucleus, suprachiasmatic nucleus (and related perisuprachiasm atic area
and subparaventricular zone), posterior periventricular nucleus, and all nuclei
of the medial zone, except the dorsal premammillary nucleus—that is, the m edial
p a rt of the m edial p reo p tic, a n terio r and cen tral p a rts of th e a n te rio r
190
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
h y p o th alam ic, dorsom edial and v entrolateral p arts of the ventrom edial
hypothalam ic, and ventral prem am m illary nuclei.
W ithin the paraventricular hypothalam ic nucleus, it is interesting to note
that fibers and term inals are especially dense in the dorsal, lateral, ventral, and
fomiceal parts, which generate descending projections to the brainstem and
spinal cord (Saper et al., 1976; Hancock, 1976; Hosoya, 1980; Sw anson and
Kuypers, 1980), and in the periventricular and m edial parvicellular parts, which
contain the neurons synthesizing hypophysiotropic somatostatin, corticotrophin-
and thyrotrophin-releasing hormones (Swanson et al., 1983; Lechan and Jackson,
1982; Johansson et al., 1984) that stimulate or m odulate the secretion of growth
horm one, adrenocorticotropin, and thyroid-stim ulating hormone, respectively
(Swanson, 1986).
The connections of the DMH with the telencephalon prim arily involve the
lateral septal nucleus and bed nuclei of the stria terminalis (BST), but also include
projections from the infralim bic area and ventral subiculum. Projections to the
BST term inate m oderately in the ventral, interfascicular, and transverse nuclei
of the posterior division, and in the posterodorsal and fusiform nuclei. The
anterodorsal area receives a sparse projection. Similarly, retrograde tracer
injections in the DMH labeled moderate num bers of neurons in the anterodorsal
area and posterodorsal, and fusiform nuclei, as well as in the anteroventral area.
However, m any fewer neurons were labeled in the posterior division of the
BST.
191
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In the lateral septal nucleus, a m oderate projection from the DMH w as
observed in the rostral part and a few PHAL-labeled axons w ere found in the
ventral part. In the retrograde tracer experiments, m oderate num bers of labeled
neurons were found in both rostral and ventral parts of the lateral septal nucleus.
In comparison, it should be noted that retrograde labeling was found throughout
the length of the rostral part, w ith neurons slightly more num erous rostrally.
Descending projections from the DMH to the brainstem appear to be m uch
sm aller than ascending brainstem inputs to the DMH. The DMH sends m oderate
num bers of fibers to Barrington's nucleus, the pontine central gray, and the
periaqueductal gray, and com paratively few to the superior central and central
la te ra l ra p h e n u c le i. The D M H receives p ro m in e n t in p u ts from th e
periaqueductal gray, from th e parabrachial nucleus, and bilaterally from
ventrolateral regions in the lateral paragigantocellular and m edullary reticular
nuclei, collectively referred to as the ventrolateral m edulla. The projection from
the ventrolateral m edulla m ay be especially im portant because this region
contains the A1 and C l noradrenergic and adrenergic cell groups, respectively,
and it has been reported that about 70% of the neurons in this region that project
to the DMH also im m unostain w ith an antiserum raised against dopam ine-6-
hydroxylase, in d icatin g th a t this projection is largely catecholam inergic
(S aw chenko a n d S w anson, 1982). Few er, b u t sig n ific a n t, n u m b e rs of
retrogradely-labeled neurons w ere found in B arrington's nucleus and the
superior central and central lateral raphe nuclei. The distribution of PHAL-
192
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
labeled axons and retrogradely-labeled neurons in the periaqueductal gray was
similar. The greatest num ber of each was found caudally in ventrom edial regions
of the periaqueductal gray.
The Organization of Hypothalamic Pathways
Projections of nuclei in and around the hypothalam ic m edial zone, rostral to
the m ammillary nucleus, have been analyzed recently w ith the PHAL m ethod
(Simerly and Swanson, 1988; Risold et al., 1994; Ganteras and Swanson, 1992a;
G anteras et al., 1992b; T hom pson et al., 1996; Thom pson and Swanson, in
preparation), th u s p ro v id in g an excellent o p p o rtu n ity to com pare DMH
projections with "classic" m edial zone nuclei (that is, m edial preoptic, anterior
hypothalam ic, ventrom edial hypothalam ic, and ventral prem am m illary nuclei,
as sum m arized recently by Risold et al., 1994) in an effort to generalize about
their possible anatomical organization and functional significance. A t the same
tim e, these considerations clarify the overall structural organization of the
hypothalam us. The prim ary conclusions from such a com parison are: 1) all
hypothalam ic nuclei use a restricted set of pathw ays that are related especially
to intrahypothalam ic projections, and are th us called propriohypothalam ic
pathw ays; 2) the set of propriohypothalam ic pathw ays used by m edial zone
nuclei is strikingly different from that used by the DMH, but sim ilar to the
pathw ays used by nuclei of the periventricular zone; and 3) there are major
differences between the DMH and the hypothalam ic m edial zone nuclei, and
193
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
sim ilarities between the DMH and the periventricular zone, in term s of both
extra- and intrahypothalam ic connections, leading to the general conclusion
that the DM H has more in common w ith periventricular zone nuclei than m edial
zone nuclei.
The propriohypothalam ic pathw ays have been nam ed on the basis of their
location and are referred to as periventricular, medial, ventral, dorsal and lateral
pathw ays (Chapter 2, Fig. 7). The periventricular, ventral, and lateral pathw ays
are the m ajor pathways used by the DMH. The course, location, and relative
contribution of each hypothalamic nucleus to these pathw ays has been described
in detail (see Chapter 2, Results). Briefly, the periventricular pathw ay courses
through the periventricular zone adjacent to, and coextensive with, the third
ventricle. The ventral pathw ay travels ventral to the m edial zone along its length
and is especially prom inent in the retrochiasmatic area and ventral preoptic
region. The lateral pathw ay is restricted largely to m edial and ventrom edial
regions of the lateral hypothalam ic area, and dorsally fibers appear to be
concentrated immediately lateral and dorsal to the fom ix. This pathw ay runs
along the extent of the lateral hypothalam ic area. The dorsal pathw ay lies in
dorsal regions of the hypothalam us and ventral regions of the zona incerta.
Rostrally, this pathw ay appears to end at the level of the rostral zona incerta,
w here the fibers either ascend to the thalam us or descend to end in the rostral
perifom ical region and the caudodorsal preoptic region. Finally, the m edial
pathw ay travels through the body of the m edial zone dorsal to the ventral
194
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
pathw ay. The only two hypothalam ic nuclei extensively using the m edial
pathw ay are the anterior hypothalam ic and ventrom edial hypothalam ic nuclei.
Both lie rostrocaudally adjacent to one another and share dense bidirectional
connections (Ganteras et al., 1994; Risold et al., 1994). However, it is considered
a separate pathw ay mainly because the vast m ajority of the descending inputs
to the hypothalam us from the am ygdala travel through it (Ganteras et al., 1995).
M ost fibers in the stria term inalis pass rostral to the anterior comm issure,
descend, and turn caudally (Heimer and Nauta, 1969) to form a circumscribed
pathw ay through the medial preoptic, anterior hypothalam ic, and ventrom edial
nuclei before ending massively in the ventral prem am m illary nucleus.
Extrahypothalam ic inputs to the DMH, and presum ably other regions of the
hypothalam us, also use a restricted set of pathw ays (Ghapter 3, Fig. 6). In the
case of the DMH, descending inputs travel through periventricular, m edial, and
la te ra l p a th w a y s, w hereas a sce n d in g in p u ts tra v e l th ro u g h m id b ra in
periventricular and brainstem lateral pathw ays. The location and components
of these pathw ays have been described in detail (see Ghapter 3, Results). Briefly,
the periventricular pathw ay and intrahypothalam ic components of the lateral
pathw ay are coextensive w ith the propriohypothalam ic periventricular and
lateral pathw ays, respectively. The m edial pathw ay partially overlaps both the
m edial and periventricular propriohypothalam ic pathw ays and is defined by
the m edial corticohypothalamic tract. It travels ventral to the hypothalam ic
paraventricular nucleus, through the capsule of the ventrom edial hypothalam ic
195
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
nucleus to end diffusely in caudal regions of the tuberal hypothalam us (Ganteras
and Sw anson, 1992b). The brainstem periventricular pathw ay ascends adjacent
to the cerebral aqueduct and, at the mesodiencephalic junction, follows the third
ventricle to enter the periventricular zone. The brainstem lateral pathw ay
ascends in the lateral tegm entum along with axons in the m edial forebrain
bundle. A t the m esodiencephalic junction, this pathw ay continues rostrally in
dorsolateral regions of the lateral hypothalamic area.
C om parison of the pathw ays used by hypothalamic nuclei indicates that
there is a topographical relationship both among functionally related cell groups,
and betw een the distance of the projection and the pathw ay utilized. In general,
hypothalam ic nuclei that project to the paraventricular hypothalam ic nucleus
use the periventricular pathway. M edial zone nuclei use ventral, dorsal, and
medial pathw ays for short intrahypothalam ic connections. The lateral pathw ay
conveys axons betw een the longest intrahypothalam ic projections (for example,
the DMH projection to the lateral p art of the m edial preoptic nucleus), and is
used by all hypothalam ic nuclei as the major pathw ay for projections to both
ascending and descending extrahypothalamic regions. With the exception of
the ventral subiculum and amygdala, which use the m edial pathw ay to innervate
the hypothalam us, distant telencephalic and brainstem projections travel in the
medial forebrain bundle lateral to the propriohypothalam ic lateral pathway.
This arrangem ent of pathw ays is generally analogous to intraspinal projections
within the propriospinal tract (fasciculus proprius), located in the deepest parts
196
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
of the w hite m atter, in contrast to supraspinal-projecting axons located in more
superficial parts of the spinal w hite matter.
Comparison with Medial Zone Nuclei: Implication for the Organization of the
Hypothalamic Periventricular and Medial Zones
Because there are major differences between the connections of the DMH
and the m edial zone, one of the major conclusions of this w ork is th at the DMH
is fundam entally different from the "classic" m edial zone nuclei. However, there
are a few sim ilarities am ong the connections of all hypothalam ic nuclei. In
g e n e ra l, all h y p o th a la m ic n u clei have a sc e n d in g , d e sc e n d in g , an d
intrahypothalam ic connections, although the relative distribution and density
of each class of projection varies by hypothalamic zone of origin.
In the brainstem , a defining feature of m edial zone nuclei appears to be a
m assive projection to the periaqueductal gray. These projections from the
anterior hypothalam ic, ventrom edial hypothalamic, and dorsal prem am m illary
nuclei term inate in interm ediate and dorsal regions throughout the length of
the periaqueductal gray (Risold et al., 1994; Ganteras et al., 1994; Ganteras and
Swanson, 1992a), w hereas the projection from the m edial preoptic nucleus
prim arily term inates in caudoventral regions (Simerly and Sw anson, 1988). The
distribution of projections from the DMH to the periaqueductal gray shows
considerable overlap with that of the m edial preoptic nucleus. However, the
projection from the DMH appears to be much smaller.
197
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In contrast, a defining feature of the periventricular zone appears to be a
massive ascending catecholaminergic input from the caudal brainstem —that
is, the A l / C l and A 2/C 2 noradrenergic and adrenergic cell groups in the
ventrolateral m edulla and nucleus of the solitary tract, respectively. These
projections ascend in lateral regions of the m edial forebrain b u n d le and
specifically avoid th e ventrom edial hypothalam ic, anterior hypothalam ic, and
medial preoptic nuclei as the fibers extend medially to innervate periventricular
structures, including the paraventricular hypothalam ic nucleus and DMH
(Swanson and H artm an, 1975; Loewy et al., 1981; Sawchenko an d Swanson,
1982).
A nother defining characteristic of m edial zone nuclei is that all receive
massive inputs from the telencephalon, mostly originating in the am ygdala,
lateral septal nucleus, and infralim bic area of the m edial prefrontal cortex
(Brittain, 1988; G anteras et al., 1992a; Ganteras et al., 1995; Risold an d Swanson,
in press). In com parison, the DMH receives a m odest in p u t from these regions,
other than the am ygdala, from which it receives virtually no fibers.
The connections betw een the lateral septal nucleus and the hypothalam us
are largely bidirectional and topographically organized by hippocam pal inputs
(Risold and Swanson, 1996; Risold and Swanson, in press). It has been show n
that projections from field GAl and the subiculum innervate the rostral part of
the lateral septal nucleus. This part then projects to the lateral part of the m edial
198
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
preoptic nucleus, anterior hypothalam ic nucleus, ventrolateral p art of the
ventrom edial hypothalam ic nucleus, and tuberal nucleus. Projections from the
ventral tip of the C A l-subiculum field innervate the ventral p art of the lateral
septal nucleus, w hich in tu rn projects to the m edial part of the m edial preoptic
nucleus, the ventral prem am m illary nucleus, and the periventricular zone
(Risold and Sw anson, 1996).
Projections from the hypothalam us to the lateral septal nucleus m ainly
o rig in a te in th e a n te rio r h y p o th a la m ic , m ed ial p reo p tic, an d v e n tra l
prem am m illary nuclei, although the DMH sends a m oderate projection as well
(Ganteras et al., 1992b; Ganteras et al., 1994; Risold et al., 1994; Thom pson et al.,
1996). The anterior hypothalam ic nucleus innervates both rostral and ventral
p arts of the lateral septal nucleus, w ith a slight predom inance in the rostral
part. The opposite is true of m edial preoptic and ventral prem am m illary nuclei
term inations. They are more dense in the ventral part, where neurons express
high levels of the estrogen receptor (Simerly et al., 1990). The ventrom edial
hypothalam ic nucleus projects only sparsely to the lateral septal nucleus,
w hereas the tuberal nucleus sends a m oderate projection to the rostral part
(Ganteras et al., 1994).
The am ygdala also shares major connections w ith the hypothalam us. The
ventral prem am m illary nucleus sends a m assive projection to the posterior and
m edial nuclei of the am ygdala (Ganteras et al., 1992b). The m edial nucleus of
199
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
the amygdala receives a moderate projection from the ventrolateral and anterior
parts of the ventromedial hypothalamic nucleus and a small projection from the
m edial preoptic nucleus and central part of the anterior hypothalamic nucleus.
Similarly, the central nucleus of the amygdala receives a moderate projection from
the ventromedial hypothalamic nucleus and a m odest projection from the anterior
hypothalamic nucleus ( (Ganteras et al., 1994; Risold et al., 1994; Simerly and
Swanson, 1988). In contrast, the DMH sends only a few, scattered fibers to the
amygdala.
The amygdala projects massively through the stria terminalis (and to a lesser
extent through the ansa peduncularis) to the hypothalam us. These projections
prim arily arise in the anterodorsal and posteroventral parts the medial nucleus of
the amygdala, which densely innervate the m edial preoptic, anterior hypothalamic,
ventromedial hypothalamic, and ventral premeunmillary nuclei and define the
course of the stria terminalis (Ganteras et al., 1995). More restricted hypothalamic
inputs originate in the posterior nucleus and the posterodorsal part of the medial
nucleus of the am ygdala (which project to the m edial preoptic and ventral
premammillary nuclei; see Ganteras et al., 1992a; Ganteras et al., 1995), posterolateral
p a rt of the cortical nucleus of the am ygdala (w hich projects to the ventral
prem amm illary nucleus; see Ganteras et al., 1992a), and posterior part of the
basomedial nucleus (which projects to the ventromedial hypothalamic nucleus; see
Petrovich et al., 1996). Notably, there are no am ygdalar projections to the DMH.
200
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
A final principle underlying the organization of the hypothalamic medial and
periventricular zones is that all medial zone nuclei, with the possible exception of
the dorsal prem am m illary nucleus, send a m ajor projection to certain other
m edial zone nuclei, and a m oderate projection to the DMH and periventricular
zone (other than the paraventricular hypothalam ic nucleus). For example, the
m edial preoptic nucleus is bidirectionally connected w ith the ventrolateral part
of the ventrom edial hypothalam ic nucleus and also projects to the ventral
prem am m illary nucleus and DMH (Simerly and Swanson, 1988; Ganteras et al.,
1994), and the anterior hypothalam ic nucleus is densely and bidirectionally
co n n ected w ith the v e n tro m e d ia l h y p o th a la m ic n u c le u s an d d o rs a l
prem am m illary nucleus an d also projects to the DM H (Risold et al., 1994;
G anteras et al., 1994; R isold and Sw anson, 1995). On the other hand, the
projections of the DMH generally tend to avoid all m edial zone nuclei, other
than a m oderate input to the lateral p art of the m edial preoptic nucleus, and
terminate densely in the periventricular zone and paraventricular hypothalam ic
nucleus.
On the basis of these anatom ical observations it is concluded that the DM H
has little in common w ith the m edial zone and is thus more appropriately
considered p art of the periventricular zone. This conclusion is perhaps illustrated
best by the stark contrast betw een descending inputs from the am ygdala that
appear to define the m ed ial zone, and ascending inputs from brainstem
catecholam inergic cell g ro u p s that a p p e a r to define com ponents of the
201
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
periventricular zone (see Ganteras et al., 1995 and Swanson and H artm an, 1975,
respectively). In addition, the nuclei that receive a m assive telencephalic input
generally are associated w ith social (agonistic and reproductive) behaviors,
w hereas nuclei that receive a dense catecholaminergic in p u t from the brainstem
play a role in m ediating endocrine and autonom ic responses that are associated
w ith all types of behavior, including ingestion, which has few social components
and m any endocrine and autonom ic correlates (Swanson, 1987; Ganteras et al.,
subm itted).
Gonsidered in this light, it is interesting to compare the output of the DMH
w ith the recently proposed distinct behavioral circuits w ithin the m edial zone
of the hypothalam us (Ganteras et al., subm itted). These are prim arily sexual,
i.e., sexually dimorphic, and defensive behavior circuits. The sexually dimorphic
circuit is differentiated by in p u ts from the accessory olfactory bulb and
vom eronasal organ, relayed through the corticomedial am ygdala (Simerly et
al., 1989; Segovia and Guillamon, 1993). In addition, the elem ents of this circuit
express receptors for the gonadal steroids estrogen a n d /o r androgen (Simerly
et al., 1990). W ithin the hypothalam us, the nuclei of the sexually dim orphic
circu it p rim a rily include the m edial p reo p tic, v e n tro la te ra l p a rt of the
ventrom edial hypothalam ic, tuberal, ventral premammillary, and anteroventral
periventricular nuclei. The defensive behavior circuit has been defined primarily
on the basis of lesion and stim ulation experiments. A t this point, it is fairly
w ell-established th at the anterior hypothalam ic, d o rsom edial p a rt of the
202
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ventrom edial hypothalam ic, and dorsal premammillary nuclei are key elements
for the expression of defensive behavior (Fuchs et al., 1985; Yardley and Hilton,
1986; ham m ers et al., 1988; Ganteras et al., submitted). Interestingly, these two
circuits are extensively interconnected and share a dense, overlapping output
in the periaqueductal gray (Ganteras and Swanson, 1992a; Ganteras et al., 1994).
Risold et al., 1994). Risold and Swanson, 1995).
In contrast, the projections of the DM H almost completely avoid all members
of both the sexually dim orphic and defensive behavior circuits, except for the
anteroventral periventricular nucleus. However, it is im portant to note that the
latter nucleus plays a significant role in endocrine aspects of reproductive
function (Wiegand and Terasawa, 1982). The projection from the DM H to the
periaqueductal gray, in addition to being qualitatively smaller, does not overlap
the term ination of these circuits—it is distinctly more caudal and ventral. On
the other hand, the DM H receives an input from all nuclei in the sexually
dim orphic circuit and from the anterior hypothalamic and dorsom edial part of
the ventrom edial hypothalam ic nucleus in the defensive behavior circuit. In
fact, the only nucleus of the medial zone that does not project to the DM H is the
dorsal prem am m illary nucleus (Ganteras and Swanson, 1992a).
In sum m ary, the DM H receives a m oderate input from all nuclei of the
hypothalam ic medial zone (except for the dorsal premammillary nucleus), major
viscerosensory inputs from the nucleus of the solitary tract, parabrachial nucleus,
203
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
ventrolateral m edulla, and the subfornical organ, and photic/circadian input
from the suprachiasm atic nucleus. The viscerosensory input also projects to the
nuclei of the preoptic region an d all parts of the hypothalam ic paraventricular
nucleus (Ricardo and Koh, 1978; Ter H orst e t al., 1989; Loewy et al., 1981;
Sawchenko and Swanson, 1982). The subfornical organ projects to the m edian
preoptic an d parastrial nuclei in the preoptic region, and to vasopressin-
containing m agnocellular parts of the paraventricular hypothalam ic nucleus
(Swanson an d Lind, 1986). The other projections of the suprachiasm atic nucleus
are slightly m ore complicated. It sends major projections to the immediately
surrounding, perisuprachiasm atic region and to the subparaventricular zone
ventral to the paraventricular hypothalam ic nucleus. These target regions show
a p attern of projections sim ilar to, but m ore robust than, those from the
suprachiasm atic nucleus. T hus, all three regions project to the DMH, the
anteroventral periventricular, anteroventral preoptic nuclei and anterodorsal
preoptic nuclei, b u t send very few axons to the paraventricular hypothalam ic
nucleus directly (Watts and Swanson, 1987; W atts et al., 1987).
The m ajor projections of the DMH are to the paraventricular hypothalam ic
nucleus, w hich contains autonom ic and endocrine visceromotor neurons, and
to the nuclei of the preoptic region. To complicate m atters further, the DMH is
bidirectionally connected w ith all preoptic nuclei, w hereas the preoptic nuclei
share a com plex set of interconnections (sum m arized in Fig. 1). However, the
DMH and all preoptic nuclei project to parvicellular parts of the paraventricular
204
R eproduced w ith perm ission of the copyright owner. Further reproduction prohibited w ithout perm ission.
hypothalam ic nucleus (both hypophysiotropic and autonom ic cell groups), but
only the parastrial and m edian preoptic nuclei (that receive projections from
th e subfornical organ) project to m agnocellular parts of the paraventricular
hypothalam ic nucleus (see Chapters 1 and 2, Results; Simerly and Swanson,
1988; Thom pson and Swanson, in preparation).
Toward an Integrated Model of Hypothalamic Function
T his com bination of anatom ical observ atio n s su ggests a v ery novel
organization of DMH circuitry (Fig. 1). The input to this circuit suggests that
the DMH, and the preoptic nuclei, respond prim arily to viscerosensory stimuli
a n d circadian inform ation. Interestingly, viscerosensory stim uli are relayed
directly to the paraventricular hypothalam ic nucleus, in addition to the "pre
m o to r" n etw o rk (see below ), su ggesting a reflex circuit an alo g o u s to a
som atom otor stretch reflex where sensory fiber collaterals synapse directly on
a-m otor neurons, but also generate supraspinal projections. It is also im portant
to n o te th a t circadian in fo rm atio n reaches the DM H d ire c tly from the
suprachiasm atic nucleus and indirectly from all cell groups receiveng an input
from it. Finally, inputs from m edial zone nuclei, considered together, appear to
be substantial and probably m odulate this circuit w ithin specific behavioral
contexts (Swanson, 1987; Ganteras and Swanson, in press).
The output of the circuit w ith which the DMH is involved has two main
co m p o n en ts: 1) the DM H and p reo p tic region nuclei form a com plex,
205
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
interconnected netw ork; and 2) each com ponent of the circuit projects to the
hypothalam ic paraventricular nucleus (see Chapters 2 and 3, Results; Sawchenko
and Swanson, 1983; Simerly and Swanson, 1988; Thompson and Swanson, in
p re p a ra tio n ). T he p ro jectio n s of n e tw o rk com ponents are sim ila r an d
overlapping, but are, in fact, unique for each nucleus in term s of both inputs
and outputs (Fig. 1). Thus, it is clear that only p a rt of this circuit receives circadian
inform ation and only part receives a non-overlapping input from the subfornical
organ. The influence of these apparently segregated inputs can be integrated
e ffe c tiv e ly by th e c o n n e c tio n s w ith in th e n e tw o rk . H o w ev er, th e se
intercormections are not sym m etrical and it is probably not the case that any
in p u t w ould lead to sim ilar activation of all components. For exam ple, the
anteroventral p eriv en tricu lar nucleus, w hich is prim arily associated w ith
circadian and endocrine aspects of gonadotropin-releasing horm one secretion
(W iegand and Terasawa, 1982) projects to, b u t does not receive a projection from,
the m edian preoptic nucleus, w hich receives a m ajor projection from the
subfornical organ and is associated with body fluid hom eostasis (Swanson and
Lind, 1986; Johnson et al., 1992). This w ould seem to indicate, at least to a first
approxim ation, th at the output of the endocrine reproductive system has a
greater influence on body fluid hom eostasis than drinking does on the estrus
cycle. Indeed, it has been show n that there are slight variations in vasopressin
secretion over the estrus cycle (Forsling and Peysner, 1988), whereas norm al
thirst does not alter sexual function.
206
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Fig. 1. Schematic diagram representing the organization of the hypothalamic peri
ventricular (left) and medial (right) zones. In this view, the periventricular zone
consists of separate populations of premotor and visceromotor cell groups. Premotor
regions integrate viscerosensory, modulatory (including entraining inputs), and
behavior-specific information to synchronize the output of the paraventricular hy
pothalamic nucleus, generating the pattern of visceromotor responses; that is, the
rhythm (period, magnitude, and duration) and relative tem poral order (phase) of
the various components. The paraventricular hypothalamic nucleus is composed
exclusively of visceromotor neurons projecting either to the median eminence to
control the anterior pituitary, the posterior pituitary directly, or brainstem and spi
nal cord autonomic cell groups, including sympathetic and parasympathetic pregan
glionic neurons.
Both premotor and visceromotor components of the periventricular zone re
ceive a massive viscerosensory input. Although part of this input is functionally
and anatomically distinguishable by terminations in magnocellular (vasopressin-
and oxytodn-containing) parts of the paraventricular nucleus, the input from the
brainstem to the preoptic region and DMH has not been show n to be segregated,
although such is likely to be the case. However, it has been shown that restricted
groups in the premotor part receive differential input. Thus, only the anteroventral
periventricular (AVPv), anteroventral preoptic (AVP), anterodorsal preoptic (ADP)
nuclei and DMH receive projections from the suprachiasmatic nucleus (Sch) and
related brain regions (see text), whereas only the parastrial (PS) and median preop
tic (MePO) nuclei receive viscerosensory input from the subfornical organ (SFO)
and are the only two premotor nuclei to send a major projection to magnocellular
parts of the paraventricular hypothalam ic nucleus. All nuclei, including the
parastrial and median preoptic nuclei, project densely to parvicellular parts of the
paraventricular hypothalamic nucleus.
The medial zone receives a massive input from the telencephalon, including
olfactory-related regions of the cortex and amygdala, and the output of the hippoc
ampal formation via the lateral septal nucleus, and sends a massive projection to
the periaqueductal gray and a moderate projection to prem otor parts of the peri
ventricular zone. It is seen as integrating cortical and sensory information to ini
tiate somatomotor and visceromotor components associated with the appropriate
goal-directed behavior.
207
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C D
■ o
O
Q.
C
g
Q.
■ D
C D
vagus/glossopharyngeal
cranial nerves
C /)
C /)
C D
8
ê'
3
3 "
C D
C D
■ D
O
Q.
C
a
o
3
■ D
O
C D
Q.
■ D
C D
C /)
C /)
Stim ulate
(V is c e ro s e n s o ry )
Entrain
Synchronize
Pre(viscero)molor
Pattern Generators
Viscerom otor
■NTS
1
VLM
S ch
AVPv AVP ADP DMH
Parvicellular
,Ctx
2 sensory/olfactory
THALv
Am ygdala
Lateral Septum
Medial Z one
(Betiavloial
Initiation)
PAG
(somato- a visceromotor
pattern genaratois)
Anterior Pituitary
(GnRH, CRH, TRH, SS/GH)
ë
00
B rainstem an d
Spinal Cord
(Dorsal Vagal Complex
Intemwdlolateral Cell Column)
Posterior Pituitary
(OXY/VAS)
Because of its structure and projections, this circuit is in an ideal position to
influence all aspects of endocrine and autonom ic function, as typified by the
paraventricular hypothalam ic nucleus, and generate the patterned output of
the viscerom otor system from the level of single horm one secretion to the
integration the complex, synchronized patterns of horm onal and autonom ic
responses th at underlie m ost com plex functions, including goal-directed
behaviors.
To effect this integration, it is proposed that the DMH, in conjunction w ith
the nuclei of the preoptic region, function as visceromotor pattern generators
analogous to those in the som atom otor system that generate locomotor rhythm s
(see Grillner, 1981; Selverston and M oulins, 1985; M arder and Calabrese, 1996).
In this respect, the viscerom otor system has m uch in com m on w ith the
somatomotor system in terms of the generation and expression of motor patterns.
Both consist of ongoing (or endogenous) processes, such as the motor rhythm s
that generate breathing m ovem ents or circadian variations in horm onal or
autonomic output, and episodic processes such as chewing or stress-induced
corticosterone secretion. Each also show s varying levels of interaction betw een
components of the same system and functionally related systems. For example,
m am m als typically m ay co n tro l in d iv id u a l lim bs, w h ich can be u sed
independently for different behaviors such as scratching or grooming, but the
same limb (and presum ably some of the sam e neural circuitry) also can be used
in com bination with others to produce coordinated m otor output such as
209
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
locom otion. A nalogous exam ples also occur throughout the viscerom otor
system . A lthough visceral responses rarely occur in iso latio n , different
combinations of responses are produced to different stimuli. Each response varies
in the type (endocrine, sympathetic, and parasym pathetic), form (am plitude,
frequency, period, and duration), and num ber of components. Presum ably those
responses using sim ilar elements also use som e of the same neural circuitry.
Finally, fundam entally independent systems, perhaps all involving rhythmic
m otor components, m ay be coupled in some w ay under various conditions. An
exam ple of such coupling involves respiration and locomotion, w here higher
rates of locomotion (running) produce increased respiratory dem ands (increased
breathing rate). Similarly, goal-directed behavior has visceral and somatic motor
com ponents that m ust be coupled to produce a coordinated response. However,
sim ilar to breathing in a resting animal, viscerom otor activity is present in the
absence of overt behavior.
M echanism s responsible for generating the p atterned o u tp u t of m otor
system s have long been debated. For example, it has been proposed that they
involve either a chain of reflexes, dependent upon sustained sensory input, or
that they involve discrete sensory inputs interacting w ith oscillatory systems to
p ro d u ce a rhythm ic o u tp u t (Selverston an d M oulins, 1985; M arder and
Calabrese, 1996). M uch recent w ork in vertebrate and invertebrate systems has
suggested that the prim ary form of rhythmic m otor activity is generated by
oscillators (pattern generators) that are largely independent of sensory feedback
210
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(Selverston and M oulins, 1985; Grillner, 1995). However, sensory system s play
a m ajor role in adapting the m otor response to appropriate environm ental
conditions. In this way, sensory inputs can initiate or term inate m otor activity,
modulate the period or amplitude of the response, a n d /o r provide critical tim ing
cues. Thus, the natural expression of internally generated rhythmic m otor output
is shaped by sensory inform ation (M arder and Calabrese, 1996).
A lthough there is no direct evidence for oscillatory neuronal m echanism s
within the DMH or preoptic region, each of the functions w ith w hich the DM H
has been associated—ingestion, body fluid hom eostasis, therm oregulation,
reproduction, and stress (see C hapters 2 and 3, Discussion)—dem onstrates
prom inent oscillatory com ponents (both over the lig h t/d a rk cycle and in
response to stimuli) and synchronized visceromotor output. Thus, am ong other
elem ents, ingestion involves synchronizing autonom ic control of insulin,
glucagon, and epinephrine release, and gastric motility, and variations in the
secretion of adrenocorticotropic and thyroid-stim ulating horm ones from the
anterior pituitary (Powley and Berthoud, 1986; Grossman, 1986; Friedm an et
al., 1986; Dallm an et al., 1993; Bray, 1993). Body fluid hom eostasis involves
autonomic control of the cardiovascular system, and secretion of vasopressin
from the posterior pituitary (Sw anson and M ogenson, 1981; Johnson, 1985;
Thrasher, 1985), and adrenocorticotropin from the anterior pituitary (Strieker et
al., 1979; Watts, 1992). Therm oregulation involves autonom ic control of brow n
fat therm ogenesis, diet-induced thermogenesis, shunting of blood flow to deep
211
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(thermogenesis) or superficial (cryogenesis) capillary beds, and variations in
the secretion of adrenocorticotropin and thyroid-stim ulating horm ones from
the anterior pituitary (Boulant, 1981; Schonbaum and Lomax, 1990; Rothwell,
1994; Arancibia et al., 1996). Reproduction prim arily involves the secretion of
luteinizing horm one, follicle-stim ulating horm one and prolactin from the
anterior pituitary (Gallo, 1980; Kalra and Kalra, 1983; Knobil and Neill, 1988).
However, the secretion of corticosterone and thyroid hormones also varies across
the estrus cycle of the rat (Buckingham et al., 1978) and significant interactions
w ith autonom ic aspects of m etabolism in both m ales and females (Campbell et
al., 1977; Wade and Schneider, 1992; Wade et al., 1996) are well-established.
Finally, although the complement of endocrine and autonomic responses to stress
prim arily depend upon which system is challenged, corticosterone secretion is
generally understood to be a w orking definition of stress (Munck et al., 1984;
D allm an et al., 1987; Harbuz and Lightm an, 1992). Nevertheless, it should be
noted that under m any circumstances, prolactin is secreted in a sim ilar fashion
(Blake, 1974), whereas the secretion of thyroid-stim ulating hormone and growth
horm one is often suppressed (M artin, 1979; Morley, 1981).
In addition to these stim ulus-induced responses, most autonom ic and
endocrine activity exhibits diurnal variations as w ell (Jarrett, 1979; Van Cauter
and Honinckx, 1985; Nijiima et al., 1992; Jasper and Engeland, 1994). It is well-
established that the suprachiasm atic nucleus contains pacemaker neurons that
are entrained to the lig h t/d ark cycle by a direct retinal input and that lesions of
212
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
this nucleus disrupt m any behavioral and visceral rhythms, such as the sleep/
w ake cycle, autonom ic influences on glycolytic and gluconeogenic cycles and
b o d y tem p era tu re , and d iu rn a l rh y th m s in th e secretion of p ro lactin ,
gonadotrophin, thyroid-stim ulating horm one, and corticosterone (see Brown-
G rant and Raisman, 1977; Raism an and Brown-Grant, 1977; Rusack and Zucker,
1979; Turek, 1985; Steffens et al., 1993; N agai et al., 1993). Moreover, the w ork of
Kalsbeek et al. (1992) indicates that the suprachiasmatic nucleus not only entrains
these rhythms, b u t also plays a role in the expression of the stim ulus-induced
response, at least in the case of corticosterone secretion. However, several points
argue against a role of the suprachiasm atic nucleus as the sole m ediator of
patterned visceral output.
First, as m entioned above, the suprachiasm atic nucleus projects to the DMH
and preoptic region, but only very lightly to the paraventricular nucleus (Watts
et al., 1987). Therefore, any m ajor control over the pattern of visceral responses
exerted by the suprachiasm atic nucleus m ust be relayed through at least one
other neural structure.
Second, the output of the suprachiasm atic nucleus is essentially biphasic,
w ith the highest levels of activity occurring during the subjective day in both
nocturnal and diurnal animals (Turek, 1985; Gillette, 1991). In contrast, visceral
rhythm s not only differ in period, but also in phase. For example, gonadotropin-
stim ulating horm one is synthesized in the brain and released into the portal
213
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
vasculature of the median eminence ultimately to stimulate luteinizing horm one
secretion from the anterior pituitary in both m ales and females. Because lim ited
sex differences have been found in anatomical circuitry (Gu and Simerly, in
press), it is likely that the release of gonadotropin-releasing horm one is controlled
by similar neural (but probably n ot hormonal) mechanisms (Fink, 1977; Simerly,
1990). However, the pattern of luteinizing horm one varies w idely betw een males
and females, and between intact and orchidectom ized anim als. Luteinizing
hormone secretion in intact m ales is episodic, b ut exhibits no rhythm in period
or am plitude, nor is it entrained to the lig h t/d ark cycle (Ellis and Desjardins,
1982), whereas the period is approxim ately 20 minutes in both gonadectom ized
m ales and fem ales (Soper a n d W eick, 1980; W atts a n d F ink, 1984). In
ovariectomized, estrogen-prim ed rats, luteinizing horm one release is diurnal,
compared to a period of 4 to 5 days in intact female rats, although in both cases
secretion occurs at the same tim e of day (Turek and Van Cauter, 1988). In female
rats exposed to constant light, spontaneous ovulation and the surge of luteinizing
horm one are inhibited as these rats sw itch to reflex (stim ulation-induced)
ovulation (Critchlow, 1963). Thus, various m anipulations are capable of altering
secretion frequency (and am plitude), blocking the preovulatory surge, and
disrupting the circadian rhythm of luteinizing hormone secretion (Fink, 1988).
Other exam ples of varied tim ing in visceral activity include the 3-4 hour
p e rio d of g ro w th horm one secretio n (T annenbaum a n d M a rtin , 1976;
Tannenbaum and Ling, 1984) and the diurnal release of corticosterone and
214
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
thyroid-stim ulating horm one (Moore and Eichler, 1972; Abe et ah, 1979; Morley,
1981). In the latter case, the period of these two hormones is similar, b u t they
are less than 180° (th at is, less than 12 hours) out of phase—corticosterone
secretion peaks at the beginning of the dark period in rats, whereas thyrotropin-
releasing horm one levels peak just after midday.
Visceral activity also m ay be either facilitated or depressed in response to a
particular stimulus. For exam ple, food deprivation, among other effects, results
in increased corticosterone secretion (Honm a et al., 1983b; Honm a et al., 1984;
Dallman et al., 1993), b u t dram atic decreases in the secretion of grow th horm one,
luteinizing horm one, and thyroid-stim ulating hormone (Campbell et al., 1977;
Bruno et al., 1990; Blake et al., 1992). In contrast, both corticosterone and thyroid-
stim ulating horm one blood levels are increased by exposure to cold (Arancibia
et al., 1996).
The above observations cannot be accounted for solely by the pattern of neu
ronal activity produced w ithin the suprachiasm atic nucleus. The pattern of vis
ceral activity in a m ultifaceted response is capable of varying in period, phase,
and direction, relative to other com ponents in response to a single stim ulus,
and the pattern of activity of an individual element is capable of varying in
form in response to a variety of stimuli, indicating that visceral responses can
be differentially controlled (probably by a variety of mechanisms) and are inde
pendent of, but interact w ith, the suprachiasm atic nucleus.
215
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Finally, elem ents other than light are able to entrain visceral rhythm s in
suprachiasm atic nucleus-lesioned anim als (Honma et al., 1983a,b; Turek, 1985;
Rosenwasser and Adler, 1986). The m ost widely studied example involves the
ability of restricted feeding to entrain corticosterone rhythm s (Krieger, 1974;
Krieger et al., 1977; H onm a et al., 1983b). In suprachiasm atic nucleus-lesioned
anim als the rise in corticosterone secretion becomes conditioned to anticipate
the presentation of food and begins to decline as food is consumed. This pattern
appears to persist for at least one cycle following term ination of the feeding
schedule (Krieger, 1979).
It has been reported that ventrom edial hypothalam ic nucleus lesions, which
invariably included the DMH, abolish the feeding-entrainable corticosterone
rhythm , as well as autonom ic and body tem perature rhythms (Krieger, 1980;
Egawa et al., 1993). In addition, lesions of the "ventral noradrenergic bundle,"
w hich contains ascending catecholaminergic fibers and conveys m uch of the
viscerosensory inform ation that stim ulates feeding, am ong other responses, may
su p p re s s c irc a d ia n an d fe e d in g -a n tic ip a to ry c o rtico ste ro n e secretio n .
Szafarczyck et al. (1985) reported that injection of the catecholamine-specific
neurotoxin 6-hydroxy dopam ine (6-OHDA) into th e ventral noradrenergic
bundle degrades the diurnal release of corticosterone into unsynchronized,
ultradian rhythm s. However, H onm a et al. (1992) reported that injection of 6-
OHDA into either the paraventricular hypothalam ic nucleus or the ventral
noradrenergic bundle suppressed the feeding-anticipatory corticosterone peak,
216
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
but neither type of injection altered the diurnal peak. Although it is difficult to
explain this discrepancy, the results suggest that viscerosensory and circadian
systems m ay act independently. In any event, it is clear that they do interact
because viscerosensory inform ation can stim ulate intrinsically oscillatory
processes. Moreover, stress-induced adrenocorticotropin and corticosterone
secretion varies across the lig h t/d ark cycle such that the sam e stressor elicits a
response of greater am plitude in the m orning (at the trough of the diurnal
rhythm), com pared to evening values (Yasuda et al., 1976; H anson et al., 1994).
This effect is m odified further by the fasted or fed state of the anim al, in a
seem ingly contradictory fashion. T hus, m orning levels of stress-induced
adrenocorticotropin secretion are blunted in response to an overnight fast,
whereas corticosterone levels are enhanced (Akana et al., 1994). Previous studies
have also noted this discrepancy between adrenocorticotropin and corticosterone
responses to stress (B radbury et al., 1991; Buijs et al., 1993a). The recent
observation that sym pathetic innervation of the adrenal gland contributes to
the generation of the diurnal rhythm of corticosterone release (Jasper and
Engeland, 1994; Dijkstra et al., 1996) suggests that there are m ultiple avenues
for the control of visceral responses, and that whereas visceral and circadian
responses are norm ally coupled, they m ay become desynchronized in response
to homeostatic challenge. This finding also underscores the interplay between
endocrine and autonom ic systems at all levels of visceral response.
217
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
One possibility for the anatomical organization of a circuit to entrain feeding-
influenced corticosterone secretion in the absence of the suprachiasmatic nucleus
is that there could be pacem aker neurons in the brainstem that interact directly
w ith the paraventricular hypothalam ic nucleus. However, because no cell group
w ith this property has yet been found in the brainstem and because the 6-OHDA
in jectio n s of H o n m a e t al. (1992) p ro b a b ly d e stro y e d m uch of th e
catecholaminergic input to both the DM H and the preoptic region, a more likely
explanation is that circadian (suprachiasmatic) and feeding (viscerosensory)
system s interact at som e level w ith both hypophysiotropic corticotropin-
releasing horm one (neuroendocrine motor) neurons and autonom ic pre-m otor
cell groups, located w ithin the paraventricular hypothalam ic nucleus. However,
because viscerosensory information is not relayed directly to the suprachiasmatic
nucleus, and the suprachiasm atic nucleus does not project directly to either
viscerosensory cell groups or the hypothalam ic paraventricular nucleus, this
interaction is probably m ediated by a common target of both regions that in
tu rn projects to the p a ra v en tricu la r hypothalam ic nucleus. A n excellent
candidate for this comm on projection is found in the DMH circuit because it
has been show n that the DM H, along w ith certain nuclei of the preoptic region,
receive both ascending viscerosensory information and an entraining input from
the suprachiasm atic nucleus, and each cell group projects to the paraventricular
hypothalam ic nucleus (Fig. 1; Chapters 2 and 3, results; Loewy et al., 1981;
Sawchenko and Swanson, 1983; Simerly and Swanson, 1988; W atts, 1991; Buijs
et al., 1993b).
218
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In sum m ary, because there is significant evidence for the interaction of
circadian and other rhythm ic influences on viscerom otor function, b u t no
anatom ical evidence for direct connections betw een these system s, and because
of the variety of patterned output that the viscerom otor system exhibits, it is
p ro p o se d th a t circadian inform ation is tran sd u ced an d in teg rated w ith
viscerosensory information by the DMH and the nuclei of the preoptic region, which
generate the appropriate set of coordinated visceromotor responses underlying
homeostasis and specific goal-directed behaviors. These nuclei are especially well-
suited to this purpose because each of them has been implicated in patterned visceral
responses that, combined, form functionally overlapping sets that span the range
of visceral responses (Wiegand and Terasawa, 1982; Swanson and Lind, 1986;
Johnson et al., 1992; Joyce and Barr, 1992; Scammell et al., 1993; Li et al., 1994; Cirelli
et al., 1995; Elmquist et al., 1996; Sherin et al., 1996). These functional sets are
articulated within the network by a complex set of interconnections, and translated
into patterned motor output by the interdigitated projection of each region to the
paraventricular hypothalamic nucleus (Fig. 1). In this w ay the elements of this circuit
serve as hypothalamic visceromotor pattern generators that m ay contain intrinsic,
"dam ped" (that is, non-self-sustaining), or network-derived oscillatory processes.
Indeed, the presence of additional oscillators, or pattern generators, has been widely
hypothesized and although their location, structure, and organization is largely
unknown, virtually every model of diurnal influences on brain function incorporates
them (for example, see Krieger, 1979; Van Cauter and Honinckx, 1985; Rosenwasser
and Adler, 1986; Robinson and Dyer, 1988; Turek and Van Cauter, 1988).
219
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
In analogy to invertebrate and simple vertebrate systems, it is likely that the
elements of this netw ork can be em ployed to form a variety of functional circuits
th a t are "co n stru cted " in response to d ifferen t sensory and m o dulatory
(endogenous) inputs and recom bined as needed for the expression of the
appropriate m otor pattern (Getting, 1989; Grillner, 1991; M orton and Chiel, 1994;
M arder and Calabrese, 1996). The m odel presented here, based on the results of
the study of the organization of DMH projections, describes w hat are likely to
be fu n d am e n ta l organizing p rin cip les of the hy p o th alam ic m ed ial and
periventricular zones, and delineates the 'T)uilding blocks" (Getting, 1989) of a
n eu ral circuit th a t is likely to form a pre-viscerom otor p attern generating
network. However, before this function can be ascribed conclusively, m uch needs
to learned about the relative distribution and types of sensory input (including
inputs from osm oreceptors, tem perature sensitive neurons, horm ones, and so
on) to particular elements of the circuit, transm itters, receptors and intrinsic
m em brane properties of each neuronal type, and the num ber and types of
viscerom otor neurons w ith w hich each m akes functional connections.
The inform ation generated from such an endeavor will eventually provide
insights into such fundam ental problems as the degree to which a nucleus is
dedicated to a particular function (as suggested by "center" theories, see
M organe, 1979), or alternatively, the range and hierarchical organization of
responses in the different functional netw orks in which a nucleus participates,
and thus, w hich functions can be perform ed independently (analogous to the
220
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
concept of "m otivational time-sharing", McFarland, 1974); or the degree to which
function is p ro d u ce d by sequential or concurrent processes, such as the
interaction betw een "leader" and "follower" neurons (Robinson and Dyer, 1988),
or "m aster" an d "slave" oscillators (Turek and Van Cauter, 1988). Although
d e te rm in in g th e p h y sio lo g ic a l c h a ra c te ristic s of th is c irc u it w ill be
extraordinarily difficult and time consuming due to the vast num ber of neurons
involved, the com plexity of the mammalian brain, and the contextual nature of
functional coupling, the m odel presented here is of considerable heuristic value
in u n d e rsta n d in g the organization of the viscerom otor system , and the
organization of the hypothalam us as a whole.
221
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Literature Cited
Abe, K., J. Kroning, M.A. Greer, and V. Critchlow (1979) Effects of destruction
of th e su p ra c h ia sm a tic nuclei o n the circad ian rh y th m s in p lasm a
co rtico stero n e, b o d y tem p era tu re , feeding a n d plasm a th y ro tro p in .
N euroendocrinology 29:119-131.
Akana, S.F., A.M. Strack, E.S. Hanson, and M.F. Dallm an (1994) R egulation of
activity in the hypothalam o-pituitary-adrenal axis is integral to a larger
hypothalam ic system that determ ines caloric flow. Endocrinology 135:1125-
1134.
Arancibia, S., F. Rage, H. Astier, and L. Tapia-Arancibia (1996) N euroendocrine
a n d au to n o m o u s m echanism s u n d erly in g th e rm o re g u la tio n in cold
environm ent. Neuroendocrinology 64:257-267.
Bellinger, L.L., L.L. Bem ardis, and V.E. Mendel (1976) Effect of ventrom edial
and dorsom edial hypothalam ic lesions on circadian corticosterone rhythm s.
N euroendocrinology 22:216-225.
Bellinger, L.L., V.E. M endel, L.L. Bemardis, and T.W. Castonguay (1986) Meal
patterns of rats w ith dorsom edial hypothalam ic nuclei lesions or sham
operations. Physiol. Behav. 36:693-698.
Bellinger, L.L., and F.E. Williams (1981) The effects of liver denervation on food
and w ater intake in the rat. Physiol. Behav. 26:663-671.
Bem ardis, L.L. (1972) H ypophagia, hypodipsia, and hypoactivity follow ing
electrolytic lesions in die dorsom edial hypothalam ic nuclei of m ature rats
of both sexes. J. N eural Trans. 33:1-10.
Blake, C. A. (1974) Stim ulation of pituitary prolactin and TSH release in lactating
and proestrous rats. Endocrinology 94:503-508.
Blake, N.G., M R. Johnson, D.J.A. Eckland, O.J.F. Foster, and S.L. Lightm an (1992)
Effect of food deprivation and altered thyroid status on the hypothalam ic-
pituitary-thyroid axis in the rat. J. Endocrinol. 233:183-188.
Boulant, J.A. (1981) H ypothalam ic mechanisms in therm oregulation. Fed. Proc.
40:2843-2850.
Bradbury, M.J., C.S. Cascio, K.A. Scribner, and M.F. Dallm an (1991) Stress-
induced adrenocorticotropin secretion: Diurnal responses and decreases
222
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
d u rin g stress in the e v e n in g are not d e p e n d e n t on c o rtico stero n e.
Endocrinology 128:680-688.
Bray, G.A. (1993) Food intake, sym pathetic activity, and adrenal steroids. Brain
Res. Bull. 32:537-541.
Brittain, D.A. (1988) The efferent connections of the infralimbic area in the rat.
Dissertation. University of California, San Diego.
Brown-Grant, K., and G. Raisman (1977) Abnorm alities in reproductive function
associated w ith the destruction of the suprachiasm atic nuclei in female rats.
Proc. R. Soc. Lond. [Biol.] 198:179-196.
Bruno, J.F., D. Olchovsky, J.D. W hite, J.W. Leidy, J. Song, and M. Berelowitz
(1990) Influence of food deprivation in the ra t on hypothalam ic expression
of grow th horm one-releasing factor and som atostatin. E ndocrinology
127:2111-2116.
Buijs, R.M., A. Kalsbeek, T.P. Van D er Woude, J.J. van Heerikhuize, and S. Shinn
(1993a) Suprachiasmatic nucleus lesion increases corticosterone secretion.
Am. J. Physiol. 264:R1186-R1192.
Buijs, R.M., M. M arkman, B. Nim es-Cardoso, Y-X. H ou, and S. Shinn (1993b)
Projections of the suprachiasm atic nucleus to stress-related areas in the rat
hypothalam us: A light and electron microscopic study. J. C om p. Neurol.
335:42-54.
Cam pbell, G.A., M. Kurcz, S. M arshall, and J. M eites (1977) Effects of starvation
in rats on serum levels of follicle stim ulating horm one, luteinizing hormone,
thyrotropin, grow th horm one and prolactin; Response to LH -releasing
horm one and thyrotropin-releasing horm one. Endocrinology 100:580-587.
C anteras, N.S., S. C hiavegatto, L.E. Ribeiro d o Valle, and L.W. Sw anson
(Submitted) Severe reduction of defensive behavior to a predator by discrete
hypothalam ic chemical lesions. Brain Res. Bull,
C anteras, N.S., R.B. Simerly, and L.W. Swanson (1992a) Connections of the
posterior nucleus of the am ygdala. J. Comp. N eurol. 324:143-179.
Canteras, N.S., R.B. Simerly, and L.W. Swanson (1992b) Projections of the ventral
prem am m illary nucleus. J. Com p. Neurol. 324:195-212.
C an teras, N.S., R.B. Sim erly, a n d L.W. S w anson (1994) O rg an izatio n of
projections from the ventrom edial and tuberal nuclei of the hypothalam us:
A PHAL study in the rat. J. Com p. Neurol. 345:41-79.
223
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
C a n te ras, N.S., R.B. Simerly, and L.W. S w anson (1995) O rg an izatio n of
projections from the medial nucleus of the amygdala: A PHAL study in the
rat. J. Comp. N eurol. 360:213-245.
C anteras, N.S., and L.W. Swanson (1992a) The dorsal prem am m illary nucleus:
A n unusual com ponent of the m am m illary body. Proc. N atl. Acad. Sci. USA
69:10089-10093.
C anteras, N.S., and L.W. Swanson (1992b) Projections of the ventral subiculum
to the am ygdala, septum , and hypothalam us: A PHAL anterograde tract-
tracing study in the rat. J. Comp. Neurol. 324:180-194.
Cirelli, C., M. Pom peiano, and G. Tononi (1995) Sleep deprivation and c-fos
expression in the rat brain. J. Sleep Res. 4:92-106.
Coolen, L.M., H.J.P.W. Peters, and J.G. Veening (1996) Fos im m unoreactivity in
the rat brain following consum m atory elem ents of sexual behavior: A sex
comparison. Brain Res. 736:67-82.
Critchlow, V. (1963) The role of light in the neuroendocrine system . In A.V.
N albandov (ed.): Advances in neuroendocrinology. Urbana: University of
Illinois Press.
D allm an, M.F., S.F. Akana, C.S. Cascio, D.N. Darlington, L. Jacobson, and N.
Levin (1987) Regulation of ACTH secretion: Variations on a them e of B. In:
Recent Progress in Hormone Research. N ew York: Academic Press,Vol. 43,
pp. 113-173.
D allm an, M.F., A.M . Strack, S.F. A kana, M.J. Bradbury, E.S. H anson, K.A.
S crib n er, a n d M. Sm ith (1993) F e a st a n d fam ine: C ritic a l role of
glucocorticoids w ith insulin in daily energy flow. Front. N euroendo. 24:303-
347.
D alton, L.D., R.G. Carpenter, and S.P. G rossm an (1981) Ingestive behavior in
adult rats w ith dorsom edial hypothalam ic lesions. Physiol. Behav. 26:117-
123.
Dijkstra, I., R. Birmekade, and F.J.H. Tilders (1996) Diurnal variation in resting
le v e ls of c o rtic o ste ro n e is n o t m e d ia te d by v a ria tio n in a d re n a l
responsiveness to adrenocorticotropin but involves splanchnic nerve activity.
Endocrinology 237:540-547.
Egaw a, M., S. Inoue, S. Satoh, Y. Takamura, K. N agai, and H. N akagaw a (1993)
Acute and chronic effects of VMH lesions on circadian rhythm s in food intake
and m etabolites. Brain Res. Bull. 32:293-299.
224
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Ellis, G.B., and C. Desjardins (1982) Male rats secrete luteinizing horm one and
testosterone episodically. Endocrinology 120:1618-1627.
Ellis, G.B., and C. Desjardins (1984) Mapping episodic fluctuations in plasma
LH in orchidectom ized rats. Am. J. Physiol. 247-.E130-E135.
Elm quist, J.K., T.E. Scammell, C D . Jacobson, and C.B. Saper (1996) Distribution
of Fos-like im m unoreactivity in the rat b ra in follow ing in trav en o u s
lipopolysaccharide adm inistration. J. Comp. Neurol. 371:85-103.
Erskine, M.S. (1995) Prolactin release after mating and genitosensory stim ulation
in females. Endocr. Rev. 16:508-28.
Fink, G. (1977) Hypothalam ic pituitary ovarian axis. In J. Stallw orthy and G.
Bourne (eds.): Recent advances in obstetrics and gynaecology. Edinburgh:
Churchill Livingstone, pp. 3-54.
Fink, G. (1988) Gonadotropin secretion and its control. In E. Knobil and J.D.
N eill (eds.): The physiology of reproduction. N ew York: Raven Press, Vol.
1, pp. 1349-1377.
F orsling, M L., an d K. Peysner (1988) P ituitary and p lasm a vasopressin
concentrations and fluid balance throughout the oestrous cycle of the rat. J.
Endocrinol. 117:397-402.
Friedm an, M L, M.G. Tordoff, and I. Ramirez (1986) Integrated m etabolic control
of food intake. Brain Res. Bull. 17:855-859.
Fuchs, S.A., H .M . Edinger, and A. Siegel (1985) The role of the anterior
hypothalam us in affective defense behavior elicited from the ventrom edial
hypothalam us of the cat. Brain Res. 330:93-107.
Gallo, R.V. (1980) Neuroendocrine regulation of pulsatile luteinizing hom one
release in the rat. Neuroendocrinology 30:122-131.
Gallo, R.V. (1981) Effect of electrical stimulation of the dorsom edial hypothalamic
nucleus on pulsatile LH release in ovariectomized rats. Neuroendocrinology
32:134-138.
G etting, P.A. (1989) Emerging principles governing the operation of neural
networks. Ann. Rev. Neurosci. 12:185-204.
Gillette, M.U. (1991) SCN electrophysiology In Vitro: Rhythmic activity and
endogenous clock properties. In D.C. Klein, R.Y. Moore and S.M. Reppert
225
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
(eds.): The suprachiasm atic nucleus: The m ind's clock. Oxford: Oxford
University Press, pp. 125-143.
Grillner, S. (1981) Control of locom otion in bipeds, tetrapods and fish. In V.B.
Brooks, (ed.): H andbook of Physiology, Sect. 1. The nervous system Vol. U.
Motor control. M aryland: W averly Press, pp. 1179-1236.
Grillner, S. (1991) Recombination of m otor pattern generators. Curr. Biol. 1:231-
2331.
Grillner, S., T. Deliagina, O. Ekeberg, A. El Manira, R.H. Hill, A. Lansner, G.N.
Orlovsky and P. Wallen (1995) N eural networks that co-ordinate locomotion
and body orientation in lamprey. TENS 18:270-279.
Grossman, S.P. (1986) The role of glucose, insulin and glucagon in the regulation
of food intake and body w eight. Neurosci. Biobehav. Rev. 10:295-315.
Gu, G.B., and R.B. Simerly (In Press) Projections of the sexually dim orphic
anteroventral periventricular nucleus in the female rat. J. Comp. Neurol.
Gunnet, J.W., and M.E. Freem an (1983) The m ating-induced release of prolactin:
A unique neuroendocrine response. Endocr. Rev. 4:44-61.
Gunnet, J.W., and M.E. Freem an (1985) The interaction of the m edial preoptic
area and the dorsom edial-ventrom edial nuclei of the hypothalam us in the
regulation of the m ating-induced release of prolactin. Neuroendocrinology
40:232-237.
Hancock, M.B. (1976) Cells of origin of hypothalm o-spinal projections in the
rat. Neurosci.Lett. 3:179-184.
Hanson, E.S., M.J. Bradbury, S.F. Akana, K.S. Scribner, A.M. Strack, and M.F.
D allm an (1994) The d iu rn a l rh y th m in ad renocorticotropin (ACTH)
responses to restraint in adrenalectom ized rats is determ ined by feeding
and calorie intake. Endocrinology 134:2214-2220.
H arbuz, M.S., and S.L. Lightm an (1992) Stress and the hypothalam o-pituitary-
adrenal axis: acute, chronic and immunological activation. J. Endocrinol.
134:327-339.
Heimer, L., and W.J.H. N auta (1969) The hypothalam ic distribution of the stria
terminalis in the rat. Brain Res. 23:284-297.
Honma, K-I., S. Honma, and T. Hiroshige (1983a) Critical role of food am ount
for prefeeding corticosterone peak in rats. Am. J. Physiol. 245:R339-R344.
226
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
H onm a, K-I., C. vonGoetz, and J. Aschoff (1983b) Effects of restricted daily
feeding on freerunning circadian rhythm s in rats. Physiol. Behav. 30:905-
913.
H o n m a , K .I., S. H o n m a, an d T. H iro sh ig e (1984) F e e d in g -a sso c ia te d
corticosterone peak in rats under various feeding cycles. Am. J. Physiol.
246:R271-R726.
H onm a, K.I., Y. Noe, S. H onm a, Y. Katsimo, and T. Hiroshige (1992) Roles of
paraventricular catecholamines in feeding-associated corticosterone rhythm
in rats. Am. J. Physiol. 262:E948-E955.
Hori, T., T. Nakashima, H. Koga, T. Kiyohara, and T. Inoue (1988) Convergence
of therm al, osmotic and cardiovascular signals on preoptic and anterior
hypothalam ic neurons in the rat. Brain Res. Bull. 20:879-885.
H osoya, Y. (1980) The d istrib u tio n of sp in al p ro jectio n n e u ro n s in the
hypothalam us of the rat, studied w ith the HRP m ethod. Exp. Brain Res.
40:79-87.
Inglefield, J.R., S B. Schw arzkopf, and C.K. K ellogg (1994) A lterations in
b e h a v io ra l responses to stre sso rs fo llo w in g ex cito to x in lesio n s of
dorsom edial hypothalamic regions. Brain Res. 633:151-161.
Jarrett, R.J. (1979) R hythm s in insulin and glucose. In D.T. K rieger (ed.):
Endocrine rhythms. New York: Raven Press, pp. 247-258.
Jasper, M.S., and W.C. Engeland (1994) Splanchnic neural activity m odulates
ultradian and circadian rhythm s in adrenocortical secretion in aw ake rats.
Neuroendocrinology 59:97-109.
Johansson, O., T. Hokfelt, and R.P. Elde (1984) hnmunohistochemical distribution
of somatostatin-like im m unoreactivity in the central nervous system of the
adult rat. Neuroscience 23:265-339.
Johnson, A.K. (1985) Role of the p eriv en tricu lar tissu e su rro u n d in g the
an tero v en tral th ird ventricle (AV3V) in the reg u latio n of b o d y fluid
hom eostasis. In R.W. Schrier (ed.): Vasopressin. N ew York: Raven Press, pp.
319-331.
Johnson, A.K., A.M. Zardetto-Sm ith, and G.L. E dw ards (1992) Integrative
m echanism s and the m aintenance of card io v ascu lar and b o d y fluid
hom eostasis: The central processing of sensory in p u t derived from the
circumventricular organs of the lamina terminalis. In A. Ermisch, R. Landgraf
227
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
and H.-J. Rühle (eds.): Progress in Brain Research. Amsterdam: Elsevier
Science,Vol. 91, pp. 381-393.
Joyce, M.P., a n d G.A. B arr (1992) The ap p earan ce of Fos p ro te in -lik e
im m unoreactivity in the hypothalam us of developing rats in response to
cold am bient tem peratures. Neuroscience 49:163-173.
K alra, S.P., an d P.S. K alra (1983) N eural regulation of luteinizing hom one
secretion in the rat. Endocr. Rev. 4:311-351.
Kalsbeek, A., R.M. Buijs, J.J. van Heerikhuize, M. Arts, and T.P. vander W oude
(1992) Vasopressin-containing neurons of the suprachiasm atic nuclei inhibit
corticosterone release. Brain Res. 580:62-67.
Kalsbeek, A., W.-J. Drijfhout, B.H.C. Westerink, J.J. van Heerikhuize, T.P. Van
Der W oude, J. van der Vliet, and R.M. Buijs (1996a) G ABA receptors in the
region of the dorsom edial hypothalam us of rats are implicated in the control
of m elatonin and corticosterone release. Neuroendocrinology 63:69-78.
Kalsbeek, A., J.J. van H eerikhuize, J. Wortel, and R.M. Buijs (1996b) A diurnal
rhythm of stim ulatory in p u t to the hypothalam o-piuitary-adrenal system
as revealed by tim ed intrahypothalam ic adm inistration of the vasopressin
Vi antagonist. J. Neurosci. 16:5555-5565.
Keim, S.R., and A. Shekhar (1996) The effects of GABAa receptor blockade in
the dorsom edial hypothalam ic nucleus on corticotrophin (ACTH) and
corticosterone secretion in m ale rats. Brain Res. 739:46-51.
Knobil, E., and J.D. Neill (Eds.) (1988) The physiology of reproduction. (2 Vols.)
New York:Raven Press.
Krieger, D.T. (1974) Food and w ater restriction shifts corticosterone, tem perature,
activity and brain am ine periodicity. Endocrinology 95:1195-1201.
Krieger, D.T. (1979) Rhythms in CRF, ACTH, and corticosteroids. In D.T. Krieger,
(ed.): Endocrine Rhythm s. N ew York: Raven Press, pp. 123-142.
Kreiger, D.T. (1980) Ventrom edial hypothalam ic lesions abolish food-shifted
circadian adrenal and tem perature rhythmicity. Endocrinology 106:649-654.
Krieger, D.T., H. Hauser, and L. Krey (1977) Suprachiasmatic nuclear lesions do
not abolish food-shifted circadian adrenal and terperature rhythm icity.
Science 297:398-399.
228
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
L am m ers, J.H ., M R. K ruk, W. M eelis, and A.M. v an d e r P oel (1988)
H ypothalam ic su b strates for brain stim ulation-induced attack, teeth-
chattering and social groom ing in the rat. Brain Res. 449:311-327.
Lechan, R.M., and I.M.D. Jackson (1982) hnm unohistochem ical localization of
thyrotropin-releasing horm one in the rat hypothalam us and pituitary.
Endocrinology 111:55-65.
Li, B. H., B. Xu, N.E. Row land, and S.P. Kalra (1994) c-fos expression in the rat
brain following central adm inistration of neuropeptide Y and effects of food
consum ption. Brain Res. 665:277-284.
Loewy, A.D., J.H. Wallach, and S. McKellar (1981) Efferent connections of the
ventral m edulla oblongata in the rat. Brain Res. Rev. 228:63-80.
M arder, E., and R.L. Calabrese (1996) Principles of rhythm ic m otor pattern
generation. Physiol. Rev. 76:687-717.
M artin, J.B. (1979) Brain m echanism s for integration of grow th horm ones
secretion. Physiologist 22:23-29.
McCarthy, H.D., A.P. Kilpatrick, P. Trayhum, and G. Williams (1993) W idespread
increases in regional hypothalam ic neuropeptide Y levels in acute cold-
exposed rats. Neuroscience 54:127-132.
M cFarland, D.J. (1974) Tim e-sharing as a behavioral phenom enon. In D.S.
L ehrm ann, R.A. H inde and E. Shaw (eds.): A dvances in the Study of
Behavior. New York: Academic Press, Vol. 5, pp. 201-225.
Moore, R., and V.B. Eichler (1972) Loss of circadian adrenal corticosterone rhythm
following suprachiasm atic lesions in the rat. Brain Res. 42:201-206.
M organe, P.J. (1979) H isto rical an d M odern C oncepts of H ypothalam ic
O rganization and Function. In P.J. M organe and J. Panksepp. (eds.):
H andbook of the H ypothalam us, Anatomy of the Hypothalam us. N ew York:
Marcel Dekker, Vol. 1, pp. 1-64.
Morley, J.E. (1981) Neuroendocrine control of thyrotropin secretion. Endocr. Rev.
2:396-436.
M orton, D.W., and H.J. Chiel (1994) Neural architectures for adaptive behaviors.
TINS 17:413-420.
229
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Mimck, A., P.M. Guyre, and N.J. Holbrook (1984) Physiological functions of
glucocorticoids in stress and th eir relation to pharm acological actions.
Endocr. Rev. 5:25-44.
Nagai, K., N. Nagai, H. Kajikawa, and H. Nakagawa (1993) Suprachiasm atic
nucleus and glucose homeostasis. In H. Nakagawa, Y. Oom ura and K. Nagai
(eds.): N ew functional aspects of the suprachiasm atic nucleus of the
hypothalam us. London: John Libbey, pp. 105-116.
Niijim a, A., K. N agai, N. N agai, an d H. N akagaw a (1992) Light enhances
sym pathetic and suppresses v ag al outflow s and lesions including the
suprachiasm atic nucleus elim inate these changes in rats. J. A utonom . Nerv.
Syst. 40:155-160.
Petrovich, G.D., P.Y. Risold, and L.W. Swanson (1996) Organization of projections
from the basom edial nucleus of the amygdala: A PHAL study in the rat. J.
Comp. Neurol. 374:387-420.
Polston, E.K., and M.S. Erskine (1995) Patterns of induction of the imm ediate-
early genes c-fos and egr-1 in the female rat brain following differential
amounts of m ating stim ulation. Neuroendocrinology 62:370-384.
Powley, T.L., and H.-R. Berthoud (1986) Participaton of the vagus and other
autonomic nerves in the control of food intake. In R.C. Ritter, S. Ritter and
C.D. Bames (eds.): Feeding behavior: Neural and hum oral controls. San
Diego: Academic Press, pp. 67-101.
Raism an, G., and K. B row n-G rant (1977) The suprachiasm atic syndrom e,
e n d o crin e an d b e h a v io ra l a b n o rm a litie s fo llo w in g lesio n s of the
suprachiasmatic nuclei in the fem ale rat. Proc. R. Soc. Lond. [Biol.] 198:297-
314.
Ricardo, J.A., and E.T. Koh (1978) Anatomical evidence of direct projections from
the nucleus of the solitary tract to the hypothalam us, amygdala, and other
forebrain structures in the rat. Brain Res. 153:1-26.
Risold, P.Y., N.S. Canteras, and L.W. Swanson (1994) Organization of projections
from the anterior hypothalam ic nucleus: A PHAL study in the rat. J. Comp.
Neurol. 348:1-40.
Risold, P.Y., and L.W. Swanson (1995) Evidence for a hypothalamothalamocortical
circuit mediating pheromonal influences on eye and head movements. Proc.
Natl. Acad. Sd. USA 92:3898-3902.
230
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Risold, P.Y., and L.W. Swanson (1996) Structural evidence for functional dom ains
in the rat hippocam pus. Science 272:1484-1486.
Risold, P.Y., and L.W. Swanson (In Press) Connections of the rat lateral septal
complex. Brain Res.Rev.
Robinson, J., and R. Dyer (1988) The pulsatile secretion of pituitary horm ones:
The phenom enon and its im plications. In G. Leng (ed.): P ulsatility in
Neuroendocrine Systems. Boca Raton: CRC Press, pp. 11-33.
Rosenwasser, A.M., and N.T. A dler (1986) Structure and function in circadian
tim ing system s: E vidence for m ultiple coupled circadian oscillators.
Neurosci. Biobehav. Rev. 10:431-448.
Rothwell, N.J. (1994) CNS regulation of thermogenesis. Grit. Rev. Neurobiol.
8:1- 10.
Rusak, B., and I. Zucker (1979) N eural regulation of circadian rhythm s. Physiol.
Rev. 59:449-526.
Sagar, S.M., K.J. Price, N.W Kasting, and PR. Sharp (1995) Anatomic patterns of
FOS immunostaining in rat brain following systemic endotoxin administration.
Brain Res. Bull. 36:381-392.
Saper, C.B., A.D. Loewy, L.W. Sw anson, and W.M. C ow an (1976) D irect
hypothalam o-autonom ic connections. Brain Res. 117:305-312.
Sawchenko, P.E., and L.W. Sw anson (1982) The organization of noradrenergic
pathw ays from the brainstem to the paraventricular and supraoptic nuclei
in the rat. Brain Res. 257:275-325.
Sawchenko, P.E., and L.W. Swanson (1983) The organization of forebrain afferents
to the paraventricular and supraoptic nuclei of the rat. J. Comp. N eurol.
218:121-144.
Scammell, T.E., K.J. Price, and S.M. Sagar (1993) Hypertherm ia induces c-fos
expression in the preoptic area. Brain Res. 618:303-307.
Schonbaum, E., and P. Lomax (Eds.) (1990) Theremoregulation: Physiology and
Biochemistry. New York:Pergamon Press.
Segovia, S., and A. G uillam on (1993) Sexual dim orphism in the vom eronasal
pathw ay and sex differences in reproductive behaviors. Brain Res. Rev. 18:51-
74.
231
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Selverston, A.I., and M. Moulins (1985) Oscillatory neural netw orks. Ann. Rev.
Physiol. 47:29-48.
S hekhar, A. (1996) GABA blockade in th e reg io n o f th e d o rso m e d ia l
hypothalam us regulates 'anxiety' in rats in the elevated plus-m aze: I.
Behavioral m easures. Brain Res. 627:9-16.
Shekhar, A., and J.S. K atner (1995) GABAg receptors in the dorsom edial
hypothalam us of rats regulate anxiety in the social interaction test. Pharm.
Biochem. Behav. 50:253-258.
Sherin, J.E., P.J. Shirom ani, R.W. McCarley, and C.B. Saper (1996) Activation of
ventrolateral preoptic neurons during sleep. Science 272:216-219.
Simerly, R.B. (1990) H orm onal control of neuropeptide gene expression in
sexually dim orphic olfactory pathways. TINS 23:104-110.
Simerly, R.B., C. Chang, M. M uramatsu, and L.W. Swanson (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, R.B., and L.W. Swanson (1988) Projections of the m edial preoptic
nucleus: a Phaseolus vulgaris leucoagglutinin anterograde tract-tracing
study in the rat. J. Comp. Neurol. 270:209-242.
Simerly, R.B., B.J. Young, M.A. Capozza, and L.W. Swanson (1989) Estrogen
d ifferen tially reg u lates n europeptide gene ex p ressio n in a sexually
dim orphic olfactory pathway. Proc. Natl. Acad. Sci. USA 86:4766-4770.
Soper, B.D., and R.F. W eick (1980) H ypothalam ic and extrahypothalam ic
m e d ia tio n of p u ls a tile d isch arg es of lu te in iz in g h o rm o n e in the
ovariectom ized rat. Endocrinology 206:348-355.
Steffens, A.B., and Strubbe, J.H. (1993) Circadian rhythm icity in autonom ic
nervous system functioning. In H. Nakagawa, Y. O om ura and K. Nagai
(eds.): N ew functional aspects of the suprachiasm atic n u cleu s of the
hypothalam us. London: John Libbey, pp. 241-247.
Stotz-Potter, E.H., S.M. M orin, and J.A. Dimicco (1996a) Effect of microinjection
of m uscim ol into the dorsom edial or paraventricular hypothalam ic nucleus
on air stress-induced neuroendocrine and cardiovascular changes in rats.
Brain Res. 742:219-224.
232
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Stotz-Potter, E.H., L.R. Willis, and J.A. Dimicco (1996b) M uscim ol acts in
dorsom edial b u t n o t paraventricular hypothalm ic nucleus to suppress
cardiovascular effects of stress. J. Neurosci. 16:1173-1179.
Strieker, E.M., A.H. Vagnucci, R.H. McDonald, and F.H. Leenan (1979) Renin
and aldosterone secretions du rin g hypovolem ia in rats: Relation to NaCl
intake. Am. J. Physiol. 237:R45-R51.
Swanson, L.W. (1986) O rganization of m am m alian neuroendocrine system . In
F.E. Bloom (ed.): H andbook of Physiology, The N ervous System , IV.
Baltimore, MD: Waverly Press, pp. 1-125.
Swanson, L.W. (1987) The hypothalam us. In A. Bjorklimd, T. Hokfelt and L.W.
Swanson (eds.); H andbook of Chemical Neuroanatom y, Integrated systems
of the CNS, Part I. Am sterdam : Elsevier Science, Vol. 5, pp. 1-124.
Swanson, L.W., and H.G.J.M. K uypers (1980) The parventricular nucleus of the
h y p o th alam u s: c y to arch itecto n ie su b d iv isio n s a n d o rg a n iz a tio n of
projections to the pituitary, dorsal vagal com plex, and spinal cord as
dem onstrated by retrograde flourescence double-labeling m ethods. J. Comp.
Neurol. 194:555-570.
Swanson, L.W., and R.W. Lind (1986) N eural projections subserving the initiation
of a specific m otivated behavior in the rat: new projections from the
subfornical organ. Brain Res. 379:399-403.
Swanson, L.W., and G.J. M ogenson (1981) Neural m echanism s for the functional
coupling of autonom ic, endocrine and som atom otor responses in adaptive
behavior. Brain Res. Rev. 228:1-34.
Swanson, L.W., P.E. Sawchenko, J. Rivier, and W.W. Vale (1983) O rganization of
ovine corticotropin-releasing factor imm unoreactive cells and fibers in the
rat brain: an im m unohistochem ical study. N euroendocrinology 36:165-186.
Szafarczyk, A., G. Alonso, G. Ixart, F. M alaval, and I. A ssenm acher (1985)
D ium al-stim ulated and stress-induced ACTH release in rats is m ediated by
ventral noradrenergic bundle. Am. J. Physiol. 249:E219-E226.
Tannenbaum , G.S., and N. Ling (1984) The interrelationship of grow th horm one
(GH)-releasing factor and som atostatin in generation of the ultradian rhythm
of GH secretion. Endocrinology 215:1952-1957.
Tannenbaum , G.S., and J.B. Martin (1976) Evidence for an endogenous ultradian
rhythm governing grow th horm one secretion in the rat. Endocrinology
98:540-562.
233
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
T hom pson, R.H., N.S. Ganteras, and L.W. Sw anson (1996) O rganization of
projections from the dorsom edial nucleus of the hypothalam us; A PHA-L
study in the rat. J. Comp. Neurol. 376:143-173.
Thom pson, R.H., and L.W. Swanson (in preparation) Anatomical substrate for
hypothalam ic pattern generators.
Thrasher, T.N. (1985) Circumventricular organs, thirst, and vasopressin secretion.
In R.W. Schrier (ed.): Vasopressin. New York: Raven Press, pp. 311-318.
Turek, RW. (1985) Circadian neural rhythm s in mammals. Ann. Rev. Physiol.
47: 49-64.
Turek, RW., and E. Van Cauter (1988) Rhythms in reproduction. In E. Knobil,
and J.D. Neill (eds.): The physiology of reproduction, vol. H. New York:
Raven Press, pp. 1789-1830.
Van Cauter, E., and E. Honinckx (1985) The pulsatility of pituitary horm ones. In
H. Schulz and P. Lavie (eds.): Ultradian rhythm s in physiology and behavior.
Berlin: Springer-Verlag, pp. 41-60.
W ade, G.N., and J.E. Schneider (1992) Metabolic fuels and reproduction in female
m ammals. Neurosci. Biobehav. Rev. 16:235-272.
Wade, G.N., J.E. Schneider, and H.-Y. Li (1996) Control of fertility by metabolic
cues. Am. J. Physiol. 270:E1-E19.
Wang, Y, S.L. Cum mings, and D.W. Gietzen (1996) Tem poral-spatial pattern of
c-fos expression in the rat brain in response to indispensable amino acid
deficiency I. The initial recognition phase. Brain Res. Mol. Brain Res. 40:27-
34.
W atts, A.G. (1991) The efferent projections of the suprachiasm atic nucleus:
Anatomical insights into the control of circadian rhythms. In D C. Klein,
R.Y. Moore and S.M. Reppert (eds.): The sup rachiasm atic nucleus: The
m ind's clock. Oxford: Oxford Univ. Press, pp. 75-104.
W atts, A.G., and G. Fink (1984) Pulsatile luteinizing horm one release, and the
inhibitory effect of estradiol-17^ in gonadectom ized male and female rats:
Effects of neonatal androgen or exposure to constant light. Endocrinology
115:2251-2259.
W iegand, S.J., and E. Terasawa (1982) D iscrete lesions reveal functional
heterogeneity of sup rachiasmatic structures in regulation of gonadotropin
secretion in die female rat. Neuroendocrinology 34:395-404.
234
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Yardley, C.P., and S.M. Hilton (1986) The hypothalam ic and brainstem areas
from w hich the cardiovascular and behavioural com ponents of the defence
reaction are elicited in the rat. J. Autonom. Nerv. Syst. 15:227-244.
Yasuda, N., K. Takebe, and M.A. Greer (1976) Evidence of nycterohemeral periodicity
in stress-induced pituitary-adrenal activation. Neuroendocrinology 21:214^224.
235
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI
films the text directly from the original or copy submitted. Thus, some
thesis and dissertation copies are in typewriter &ce, while others may be
from any type o f computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality
illustrations and photographs, print bleedthrough, substandard margins,
and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete
manuscript and there are missing pages, these will be noted. Also, if
unauthorized copyright material had to be removed, a note will indicate
the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand comer and
continuing from left to right in equal sections with small overlaps. Each
original is also photographed in one exposure and is included in reduced
form at the back of the book.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6” x 9” black and white
photographic prints are available for any photographs or illustrations
appearing in this copy for an additional charge. Contact UMI directly to
order.
UMI
A Bell & Howell Information Company
300 North Zeeb Road, Ann Arbor MI 48106-1346 USA
313/761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
UMI N um ber: 9733150
UMI Microform 9733150
Copyright 1997, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized
copying under Title 17, United States Code.
UMI
300 North Zeeb Road
Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A seasonal study on the distribution of polycystine radiolarians from waters overlying the Catalina basin, southern California

PDF

Fluid balance and neuropeptides in the lateral hypothalamic area

PDF

¹⁹F-NMR studies of trifluoroacetyl insulin derivatives

PDF

A total evidence cladistic analysis of the Haliotidae (Gastropoda: Vetigastropoda)

PDF

Neurotrophic actions of TGFɑ, NT4/5 and bFGF in the developing and adult intact and lesioned rat nigrostriatal system

PDF

Biochemical evidence for calpain's involvement in long term synaptic plasticity

PDF

Corticosterone modulation of stress-induced neuropeptide gene expression in the paraventricular nucleus

PDF

Cellular proliferation in the pulmonary alveoli of the mouse and rat

PDF

Cerebellar cortical-nuclear projections and classical eyeblink conditioning

PDF

A study of the effect of glycogen on the oxidation of butyrate by rat liver slices

PDF

A biogeographic analysis of the freshwater fishes of Honduras

PDF

A new approach to the problem of measuring the properties of micelles

PDF

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

PDF

Evolution in the oceanic midwaters: comparative osteology and relationships of the lanternfishes (family Myctophidae)

PDF

Systematics and host relationships of the mites of the family Spinturnicidae In Costa Rica (Acarina: Spinturnicidae)

PDF

Building a model of organization acculturation: an interpretive study of organizational culture and stories

PDF

Spatial distribution of neuroendocrine motoneuron pools in the hypothalmic paraventricular nucleus

PDF

Computer simulation of polar solvents with the SCAAS model

PDF

Alternate models of women's health care policy in the United States

PDF

A critical analysis of ideology and discourse in two hotels

Asset Metadata

Creator Thompson, Richard Holley (author)

Core Title Connections of the dorsomedial hypothalamic nucleus in the rat

School Graduate School

Degree Doctor of Philosophy

Degree Program Biology

Degree Conferral Date 1997-05

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

Tag biology, neuroscience,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor [illegible] (committee chair), [illegible] (committee member), Watts, Alan (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c17-279854

Unique identifier UC11349888

Identifier 9733150.pdf (filename),usctheses-c17-279854 (legacy record id)

Legacy Identifier 9733150.pdf

Dmrecord 279854

Document Type Dissertation

Rights Thompson, Richard Holley

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