Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The trafficking and pathogenesis of cathepsin S in Sjögren’s syndrome
(USC Thesis Other)
The trafficking and pathogenesis of cathepsin S in Sjögren’s syndrome
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
Copyright 2021
Runzhong Fu
THE TRAFFICKING AND PATHOGENESIS OF CATHEPSIN S IN
SJӦGREN’S SYNDROME
By
RUNZHONG FU
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PHARMACEUTICAL SCIENCES)
August 2021
ii
ACKNOWLEDGEMENTS
I would like to thank my PI Dr. Sarah Hamm-Alvarez for guiding and training me during this
PhD. Throughout the past five years in Dr. Hamm-Alvarez’s lab, she not just provided me with
advanced scientific training but also taught me much about good leadership, effective management,
and constructive criticism. I would like to thank my committee members Dr. Curtis Okamoto and Dr.
J. Andrew Mackay for their generous advice and support. I want to appreciate Srikanth R Janga,
Dr. Maria Edman, Dr. Yaping Ju, Dr. Hao Gou, Minchang Choi, for helping with my experiments
and contributing data to my thesis. I would also like to thank Dr. Zhen Meng, Dr. Mihir Shah and
Dr. Wannita Klinngam for the training and onboarding of my PhD projects. I would like to
acknowledge all the funding resources for my research: NIH R01 grants EY011386, P30
EY029220 to the USC Ophthalmology Core Grant in Vision Research, and by an unrestricted
departmental grant to the Department of Ophthalmology from Research to Prevent Blindness
(RPB), New York, NY 10022. The APC was paid for by institutional funding available to Dr.
Sarah Hamm-Alvarez.
Finally, I want to thank my family, Dr. David R. Tyrpak, Dr. Jeff Shen, Dr. Peilan He, Stephen
Tyrpak, Camille B. Tyrpak, Jenna Tyrpak, Dr. Holly Tyrpak, Stephanie McCombs, Patrick Ho
and Ying Zhao. Their support and encouragement carried me this far.
iii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ...................................................................................................... ii
LIST OF FIGURES ................................................................................................................... vi
ABBREVIATIONS ................................................................................................................... ix
ABSTRACT ............................................................................................................................... x
PREFACE ............................................................................................................................... xiii
CHAPTER 1 Introduction .......................................................................................................... 1
1.1 Lacrimal gland and tear fluid ................................................................................................ 1
1.2 Neural regulation in lacrimal gland secretion ....................................................................... 2
1.3 Major conventional secretory trafficking pathway ............................................................... 5
1.4 Rab3D and Rab27a trafficking functions in the LGAC ....................................................... 6
1.5 Pikfyve in membrane trafficking .......................................................................................... 7
1.6 Dry eye disease ..................................................................................................................... 9
1.7 Sjögren’s syndrome (SS) .................................................................................................... 11
1.8 Cathepsin S in Sjögren’s Syndrome ................................................................................... 12
1.9 CX3CL1 characteristics and functions ............................................................................... 14
1.10 Goals ................................................................................................................................. 16
CHAPTER 2 CX3CL1, together with CTSS activity, is increased in tear of SS
murine model ............................................................................................................................ 17
2.1 CX3CL1 (fractalkine) is elevated in tears of NOD mice. .................................................. 17
2.2 CX3CL1 and CTSS are increased in NOD mouse LGAC. ................................................ 19
2.3 CX3CL1 is increased in corneal buttons from male NOD LG. .......................................... 21
2.4 Cx3cr1 expression and CX3CR1+ immune cells are increased in
male NOD LG. ........................................................................................................................ 23
CHAPTER 3 Proinflammatory cytokines and CTSS activity upregulates
CX3CL1 expression and shedding in LGACs and HCE-T cells. ............................................. 27
3.1 In vitro IFN-γ induces CX3CL1 in BALB/c LGAC. ......................................................... 27
3.2 IFN-γ treatment increases CX3CL1 in HCE-T cells and culture medium. ........................ 29
3.3 rhCTSS increases CX3CL1 expression in HCE-T cells. .................................................... 32
iv
3.4 Extracellular CTSS enhances CX3CL1 shedding in HCE-T cells. .................................... 33
CHAPTER 4 Rab27a is involved in CTSS secretion in lacrimal gland acinar cells ................ 37
4.1 CTSS and Rab27a are enriched in the same vesicles in the lacrimal gland
acinar cells in male NOD mouse .............................................................................................. 37
4.2 In Male NOD Mouse LG, Rab27a-Enriched Vesicles Are Subapically
Increased and Their Size Decreased, Relative to Rab27a-Enriched Vesicles in
Healthy Male BALB/c Mouse LG ............................................................................................ 39
4.3 Ad-mCFP-Rab27a Construct Design and Characterization ............................................... 41
4.4 CCh-Stimulated Rab27a Vesicle Fusion Detected in Primary Cultured
Rabbit LGACs .......................................................................................................................... 45
4.5 Expression of DN Rab27a Significantly Reduces CTSS and β-Hex
Secretion in Rabbit LGACs ...................................................................................................... 47
4.6 Endolysosomal Markers, together with Rab27a, Are Redistributed to
the APM with Topical CCh Stimulation in Male NOD Mouse LGAC ................................... 49
4.7 Rab27a Distribution on Endolysosomes is Increased in NOD Mice ................................. 52
CHAPTER 5 α1-Adrenergic agonist phenylephrine increases tear CTSS
secretion .................................................................................................................................... 58
5.1 Phenylephrine induces less fluid secretion in C57 mouse lacrimal glands ........................ 58
5.2 Phenylephrine significantly increased CTSS secretion in C57 mouse tears ...................... 63
5.3 Phenylephrine significantly upregulated CTSS secretion in primary mouse LGACs ........ 65
CHAPTER 6 α1-Adrenergic agonist phenylephrine induces alternative trafficking in
mouse LGACs .......................................................................................................................... 68
6.1 Phenylephrine induces less Rab3D vesicle fusion and triggers limited
calcium response in mouse LGACs ......................................................................................... 68
6.2 Carbachol induces Rab3D colocalization with endo-lysosomal markers
at the subapical area of mouse LGACs .................................................................................... 71
6.3 Dysfunctional Rab3D enhanced α1-adrenergic sensitivity and upregulates
CTSS secretion ......................................................................................................................... 73
6.4 Phenylephrine induces apical redistribution of endo-lysosomal markers
in both wild-type and Rab3D-deficient LGACs ....................................................................... 76
6.5 Phenylephrine induces apical redistribution of Vamp2 in both wild-type
and Rab3D-deficient LGACs .................................................................................................. 78
CHAPTER 7 PIKFyve inhibition reduces tear CTSS secretion ............................................... 81
7.1 Inhibition of PIKfyve reduces CTSS secretion in rabbit LGACs ...................................... 81
7.2 Inhibition of PIKfyve reduces Rab7-enriched vesicle fusion in rabbit LGACs ................. 83
7.3 Inhibition of PIKfyve increases endolysosome size but reduces apical
accumulation of endolysosomes ............................................................................................... 85
v
CHAPTER 8 Discussion .......................................................................................................... 91
8.1 Interaction between CX3CL1 and CTSS contributes to ocular inflammation ................... 91
8.2 Rab27a regulate CTSS secretion in the LGACs through both a major regulated
secretion and a minor endolysosomal pathway ....................................................................... 96
8.3 Neural regulation affects CTSS secretion ........................................................................ 102
CHAPTER 9 Conclusions ...................................................................................................... 110
CHAPTER 10 Material and methods ..................................................................................... 111
10.1 Reagents ......................................................................................................................... 111
10.2 Animals ........................................................................................................................... 112
10.3 Tear collection ................................................................................................................ 114
10.4 Tear CTSS activity measurement ................................................................................... 114
10.5 Tear β-hexosaminidase activity measurement ................................................................ 114
10.6 Mouse primary cell culture and IFN- γ treatment .......................................................... 115
10.7 HCE-T cell culture and IFN- γ treatment ....................................................................... 115
10.8 CTSS treatment in cell culture ....................................................................................... 115
10.9 Analysis of gene expression in cells and tissues ............................................................ 116
10.10 ELISA measurements of CX3CL1 ............................................................................... 116
10.11 LG tissue processing and confocal fluorescence microscopy ...................................... 117
10.12 Flow cytometry of lymphocytes infiltrating the LG ..................................................... 118
10.13 Molecular cloning of Ad-mCFP-Rab27a constructs .................................................... 118
10.14 Adenovirus production and amplification .................................................................... 120
10.15 Rabbit LGAC preparation and transfection .................................................................. 120
10.16 Live Cell Imaging of Rabbit LGACs ........................................................................... 121
10.17 In vitro secretion of primary rabbit LGAC ................................................................... 121
10.18 Western blotting for tear CTSS .................................................................................... 122
10.19 Primary C57 LGAC preparation and secretion assay ................................................... 123
10.20 Immunofluorescence of primary LGACs ..................................................................... 124
10.21Fluorescence Imaging of Intracellular Calcium in mouse LGACs ............................... 124
10.22 Apilimod treatment in rabbit LGACs for secretion assay ............................................ 125
10.23 Rab7-CFP BacMam virus transduction and imaging in rabbit LGACs ....................... 125
10.24 Apilimod treatment in mouse tissue ............................................................................. 126
10.25 Statistics ........................................................................................................................ 126
REFERENCES ....................................................................................................................... 127
vi
LIST OF FIGURES
CHAPTER1 Introduction 1
Figure 1.1 The LG acini is responsible for secreting lacrimal fluid into the tear film. 4
Figure 1.2 The LG innervated with sympathetic and parasympathetic nerves. 5
Figure 1.3 M3AchR and α1D-adrenergic signaling pathways 7
Figure 1.4 PIKfyve is a main regulator for endolysosome maturation 8
Figure 1.5 Dry eyes disease classification. 10
Figure 1.6 Symptoms and causes of Sjögren’s syndrome 12
Figure 1.7 Cathepsin S (CTSS) is implicated in many biological functions. 13
Figure 1.8
CX3CL1 (Fractalkine) is a unique protease cleavable chemokine.
15
CHAPTER2 CX3CL1, together with CTSS activity, is increased in tear of SS
murine model
17
Figure 2.1 CX3CL1 protein and CTSS activity are elevated in NOD mouse tears
relative to BALB/c mouse tears.
18
Figure 2.2 CX3CL1 and CTSS expression are elevated in NOD mouse LG relative
to BALB/c.
20
Figure 2.3 CX3CL1 is increased in a cross section of NOD mouse LG compared to
BALB/c mouse LG.
21
Figure 2.4 Cx3cl1 is elevated in the corneal button of NOD mouse. 22
Figure 2.5 CX3CR1+ immune cells are increased in NOD mouse LG. 25
CHAPTER3 Proinflammatory cytokines and CTSS activity upregulates CX3CL1
expression and shedding in LGACs and HCE-T cells.
27
Figure 3.1 Recombinant mouse IFN-γ (200 U/ml) induces Cx3cl1 gene expression
and elevates CX3CL1 protein in culture medium of cultured LGAC from
BALB/c mice.
29
Figure 3.2 Recombinant human IFN-γ (1 µg/ml) induces both CX3CL1 and CTSS
gene and protein expression in HCE-T cells.
32
Figure 3.3 rhCTSS exposure induces CX3CL1 gene and protein expression in
HCE-T cells.
34
Figure 3.4 rhCTSS increases CX3CL1 recovery in culture medium. 36
vii
CHAPTER4 Rab27a is involved in CTSS secretion in lacrimal gland acinar cells 38
Figure 4.1 Increased enrichment of CTSS and Rab27a in vesicles in LGAC from
male NOD mice.
40
Figure 4.2 Increased accumulation on subapically-enriched vesicles and reduced
vesicle diameter characterizes Rab27a-enriched vesicles in male NOD
mouse.
42
Figure 4.3 WT and DN Ad-mCFP-Rab27a constructs and characterization. 44
Figure 4.4 Ad-mCFP-Rab27a production and plasmid validation. 45
Figure 4.5 Estimated transduction efficiency for Ad-mCFP-Rab27a transduction in
rabbit LGACs.
46
Figure 4.6 CCh stimulation induces mCFP-Rab27a-enriched vesicle homotypic
fusion and vesicle depletion in in primary cultured rabbit LGAC.
48
Figure 4.7 CCh stimulation induces CTSS, β-hex and total protein secretion from
primary cultured rabbit LGACs.
50
Figure 4.8 DN Ad-mCFP-Rab27a transduction significantly reduces CTSS and β-
hex secretion in primary cultured rabbit LGACs.
51
Figure 4.9 Rab27a and endolysosomal membrane marker colocalization with the
apical actin meshwork.
53
Figure 4.10 CTSS immunofluorescence is significantly reduced by topical CCh
stimulation in NOD mouse relative to BALB/c mouse LG.
55
Figure 4.11 Rab27a is co-enriched with lysosomal fractions containing CTSS. 56
Figure 4.12 Vesicular Rab27a is dispersed into the cytoplasm in CCh-stimulated LG. 57
Figure 4.13 Rab27a enrichment with endolysosomes is increased in LG from male
NOD mice.
59
CHAPTER5 α1-Adrenergic agonist phenylephrine increases tear CTSS secretion 60
Figure 5.1 In vivo topical stimulation of LG α1-adrenergic receptor with
phenylephrine evokes lesser tear fluid secretion and more concentrated
tears relative to topical carbachol.
62
Figure 5.2 Carbachol (CCh) and phenylephrine (PE) induces both protein and fluid
secretion in tears in in vivo topical stimulation.
63
Figure 5.3 In vivo tear secretion in C57 LG stimulated by phenylephrine (PE) at
different doses.
64
Figure 5.4 LG stimulation with topical PE increases CTSS activity and CTSS
protein content of tears of C57 mice.
66
Figure 5.5 Phenylephrine stimulation specifically induces CTSS activity in mouse
primary lacrimal gland acinar cell culture.
69
viii
CHAPTER6 α1-Adrenergic agonist phenylephrine induces alternative trafficking
in mouse LGACs
70
Figure 6.1 Phenylephrine increased intracellular calcium by small amount, reflected
by minimum increase of Rab3D size.
72
Figure 6.2 Carbachol significantly increased Rab3D colocalization with endo-
lysosomal markers at the subapical area.
74
Figure 6.3 Rab3D integrated density reduced with Carbachol stimulation in C57
LGAC.
75
Figure 6.4 Dysfunctional Rab3D CTSS enhanced α1-adrenergic sensitivity and
upregulates CTSS secretion.
77
Figure 6.5 Phenylephrine induces apical redistribution of endo-lysosomal markers
in both wild-type and Rab3D-deficient LGACs
79
Figure 6.6 Phenylephrine induces apical redistribution of Vamp2 in both wild-type
and Rab3D-deficient LGACs.
81
Figure 6.7 Phenylephrine significantly increased colocalization of Vamp2 with
endo-lysosomal markers in C57 LGACs.
82
CHAPTER7 PIKFyve inhibition reduces tear CTSS secretion 83
Figure 7.1 Pharmacological inhibition of PIKfyve reduces CTSS activity recovered
in rabbit LGACs.
84
Figure 7.2 PIKfyve inhibition with apilimod reduces the fusion of Rab7 enriched
vesicles.
86
Figure 7.3 PIKfyve inhibition increases Lamp1 enriched vesicle size and reduces
apical accumulation of Lamp1 at the APM.
89
Figure 7.4 PIKfyve inhibition increases Rab7 enriched vesicle size and decreases
apical accumulation of Lamp1 at the APM.
90
Figure 7.5 Gene expression of PIKfyve, TFEB, TRPML1 in BALB/c, NOD, C57 and
Rab3DKO mouse.
92
CHAPTER8 α1-Adrenergic agonist phenylephrine induces alternative trafficking
in mouse LGACs
93
Figure 8.1 Proposed mechanisms implicated in CX3CL1 induction, cleavage and
recruitment of CX3CR1+ immune cells in autoimmune dacryoadenitis.
94
Figure 8.2 Proposed mechanism of Rab27a involvement in CTSS traffic in LGAC. 99
Figure 8.3 Proposed mechanism of phenylephrine stimulated CTSS traffic in
LGAC.
105
ix
ABBREVIATIONS
ADDE Aqueous-Deficient Dry Eye
APM Apical Plasma Membrane
β-hex β-Hexosaminidase
CCh Carbachol
CTSS Cathepsin S
DN Dominant negative
EDE Evaporative dry eye
IFN-γ Interferon-γ
LG Lacrimal Gland
LGAC Lacrimal Gland Acinar Cells
MHC II Major histocompatibility complex II
M3AchR Muscarinic M3 cholinergic receptors
MVB Multivesicular Bodies
NOD Non-obese Diabetic
PE Phenylephrine
PI3P Phosphatidylinositol-3-phosphate
PIKfyve Phosphatidylinositol-3-phosphate 5-kinase
PI(3,5)P2 Phosphatidylinositol 3,5-bisphosphate
PCM Peter’s complete medium
RA Rheumatoid arthritis
SS Sjögren’s syndrome
SG Salivary Gland
SV Secretory Vesicle
TNF-α Tumor Necrosis Factor
WT Wild-type
x
ABSTRACT
Autoimmune dacryoadenitis and altered lacrimal gland (LG) secretion are features of Sjögren’s
syndrome (SS). Activity of cathepsin S (CTSS), a cysteine protease, is significantly and
specifically increased in SS patient tears. The soluble chemokine, CX3CL1 (fractalkine), is
cleaved from membrane-bound CX3CL1 by proteases including CTSS.
In the first main part of this thesis (Chapter 2 and 3), I show that CX3CL1 is significantly elevated
by 2.5-fold in tears (p = 0.0116) and 1.4-fold in LG acinar cells (LGAC)(p = 0.0026) from male
NOD mice, a model of autoimmune dacryoadenitis in SS, relative to LGAC from BALB/c controls.
Primary mouse LGAC and human corneal epithelial cells (HCE-T cells) exposed to interferon-
gamma, a cytokine elevated in SS, showed up to 9.6-fold (p ≤ 0.0001) and 25-fold (p ≤ 0.0001)
increases in CX3CL1 gene expression, and 1.9-fold (p = 0.0005) and 196-fold (p ≤ 0.0001)
increases in CX3CL1 protein expression, respectively. Moreover, exposure of HCE-T cells to
recombinant human CTSS at activity equivalent to that in SS patient tears increased cellular
CX3CL1 gene and protein expression by 2.8-fold (p = 0.0021) and 5.1-fold (p ≤ 0.0001), while
increasing CX3CL1 in culture medium by 5.8-fold (p ≤ 0.0001). Flow cytometry demonstrated a
4.5-fold increase in CX3CR1-expressing immune cells (p ≤ 0.0001), including increased T-cells
and macrophages, in LG from NOD mice relative to BALB/c. CTSS-mediated induction/cleavage
of CX3CL1 may contribute to ocular surface and LG inflammation in SS.
The second part (Chapter 4), includes my exploration of the role of Rab27a in CX3CL1 trafficking
in the LGACs. Rab3D and Rab27a/b isoforms are effectors of exocytosis in LG, but Rab27a is
poorly studied. To investigate whether Rab27a mediates CTSS secretion, we utilized quantitative
xi
confocal fluorescence microscopy of LG from SS-model male NOD and control male BALB/c
mice, showing that Rab27a-enriched vesicles containing CTSS were increased in NOD mouse LG.
Live-cell imaging of cultured LGAC transduced with adenovirus encoding wild-type (WT)
mCFP-Rab27a revealed carbachol-stimulated fusion and depletion of mCFP-Rab27a-enriched
vesicles. LGAC transduced with dominant-negative (DN) mCFP-Rab27a exhibited significantly
reduced carbachol-stimulated CTSS secretion by 0.5-fold and β-hexosaminidase by 0.3-fold,
relative to stimulated LGAC transduced with WT mCFP-Rab27a. Colocalization of Rab27a and
endolysosomal markers (Rab7, Lamp2) with the apical membrane was increased in both stimulated
BALB/c and NOD mouse LG, but the extent of colocalization was much greater in NOD mouse
LG. Following stimulation, Rab27a colocalization with endolysosomal membranes was decreased.
In conclusion, Rab27a participates in CTSS secretion in LGAC though the major regulated
pathway, and through a novel endolysosomal pathway that is increased in SS.
The third major (Chapter 5 and 6) in my thesis involves the investigation of neural regulations in
the LGAC and its relation to CTSS trafficking. Dysfunctional neural regulation in the LG
contributes to altered tear secretion. Here, we compared secretion and trafficking of CTSS in the
mouse LG when stimulated with the M3 muscarinic receptor agonist, carbachol (CCh), versus the
α1-adrenergic agonist, phenylephrine (PE). In situ and in vitro secretion studies showed that PE
significantly increased CTSS secretion into tears compared to CCh, while CCh induced higher β-
hexosaminidase (β-hex) secretion. Calcium [Ca
2+
] imaging showed that CCh significantly
elevated intracellular [Ca
2+
] and increased the size of Rab3D-enriched vesicles. The enlarged
Rab3D is associated with increased homotypic secretory vesicle (SV)-SV fusion, heterotypic
fusion between endolysosomes and SVs, and SV fusion towards the apical plasma membrane
(APM). PE stimulation induced a lower intercellular [Ca
2+
] response and had a minimal effect in
xii
inducing Rab3D-enriched major regulated pathway. LGs deficient of Rab3D exhibited higher
sensitivity towards PE stimulation and induced higher tear CTSS activity. The colocalization of
endolysosomal markers (Lamp1, Lamp2, Rab7) with the apical actin was increased with both CCh
and PE stimulation, with PE demonstrating a more apparent upregulation. In conclusion, PE
stimulation increased CTSS secretion into tears, which is related to decreased Rab3D-enriched
major regulated and an increased minor endolysosomal seceretory pathway.
The final study (Chapter 7) investigates the role of PIKfyve in CTSS trafficking in the LGACs.
Through in vitro rabbit LGAC culture, I found that pharmacological inhibition of PIKfyve with
apilimod significantly reduced CTSS secretion as well as Rab7-vesicle fusion. PIKfyve inhibition
in ex vivo tissue also showed that inhibition of PIKfyve significantly enlarged Rab7 and Lamp1-
enriched vesicles in both C57 and Rab3DKO LG. Apilimod also significantly reduced the apical
accumulation of Rab7 and Lamp1 in the mouse acini. In conclusion, I hypothesize that PIKfyve
plays a role in endolysosome trafficking pathway, which is the main source of CTSS secretion in
the LGAC.
xiii
PREFACE
This dissertation focuses on addressing the disease biology and membrane trafficking of
cathepsin S, a tear biomarker for Sjögren’s syndrome. Accordingly, this dissertation consists of
four distinct areas: 1) CX3CL1, a downstream disease inducer activated by cathepsin S. 2) Rab27a
a major regulator for cathepsin S trafficking in lacrimal gland acinar cells. 3) Differential neural
regulation related to cathepsin S secretion in the LG. 4) PIKfyve, a possible regulator for CTSS
activity. Note that Chapters 1, 2, 3, 4, 8 and 10 are quoted verbatim quoted from my published first
author manuscripts
1-3
.
Chapter 1 gives a broad overview of the tear film, the lacrimal gland and neural functional units
essential for maintaining ocular surface homeostasis. It also introduces Sjögren’s syndrome dry eye
disease and its tear biomarker, cathepsin S. Due to the extensive biological functions of cathepsin
S, I hypothesized that it also contributes to disease etiology. Chapter 2 describes a unique
chemokine, CX3CL1, that is found to be upregulated in tears, LGACs and corneas of diseased
male NOD mouse, a model of ocular symptoms of Sjögren’s syndrome. The elevated CX3CL1 is
also suspected to induce an infiltration of CX3CR1-expressing lymphocytes infiltrating the
lacrimal gland. Chapter 3 further investigates the relationship between proinflammatory
cytokines, cathepsin S activity and CX3CL1 in both the lacrimal gland acinar cell and in HCE-T
cells which are a human corneal cell line. CX3CL1 expression was found to be upregulated by
both interferon-γ and cathepsin S activity. Interferon-γ also elevated cathepsin S expression.
Active cathepsin S further cleaves off membrane bound CX3CL1 and sheds more soluble CX3CL1
into the tears.
xiv
Despite many studies investigating the disease biology of cathepsin S, the membrane
trafficking of cathepsin S in the lacrimal gland acinar cell and how it may relate to health versus
disease remains unclear. In Chapter 4, I continue to investigate the relationship of Rab27a and
cathepsin S secretion. In our previous studies we have found that upregulated cathepsin S secretion
is likely related to dysfunctional trafficking of Rab3D-enriched secretory vesicles, instead arising
through increased through a Rab27-dependent pathway. Chapter 4 further supports this
hypothesis, finding that Rab27a likely regulates cathepsin S secretion through both a major-
regulated pathway and a minor endolysosomal pathway.
Chapter 5 and Chapter 6 explore the upstream neural regulation that may selectively
upregulate cathepsin S trafficking pathways. The lacrimal gland is innervated with various
sympathetic and parasympathetic nerves that are responsible for protein and fluid secretion. In this
study we found that α-adrenergic agonist, phenylephrine, increases cathepsin S secretion in the
lacrimal gland acinar cells. Compared to the cholinergic agonist, carbachol, phenylephrine had
minimal effect in increasing intracellular calcium and prompting Rab3D-enriched secretory
vesicle exocytosis. Phenylephrine is suspected to induce cathepsin S secretion through both
elevated minor endolysosomal pathway and a downregulated Rab3D-enriched major regulated
pathway.
Chapter 7 is a preliminary study exploring the role of the lipid kinase, PIKfyve, in regulating
cathepsin S secretion. PIKfyve is a key regulator in modulating endolysosome maturation. We
have hypothesized that cathepsin S traffics through an endolysosomal pathway. Interesting, we
found that in vitro inhibition of PIKfyve reduced cathepsin S secretion and Rab7-enriched vesicle
fusion. Therefore, we hypothesize that PIKfyve is involved in cathpsin S trafficking in the LGACs.
xv
Chapter 8 and Chapter 9 include discussions and conclusions about the four major topics.
Chapter 10 contains detailed experimental methods for the studies.
1
CHAPTER 1
Introduction
1.1 Lacrimal gland and tear fluid
The tear film is a unique thin transparent film covering the outer mucosal surface of the eye
that helps maintain the integrity and homeostasis of the ocular surface. The tear film consists of an
aqueous/mucin layer and a lipid layer. The lipids in the tear film, the majority of which are secreted
from the meibomian gland, help prevent the evaporation of tears. Most of the aqueous layer content
originates from the lacrimal gland, which secretes a complex mixture of water, electrolytes, mucin,
hydrolases, growth factors, cytokines protease and protease inhibitors. The conjunctival
epithelium is another source of electrolytes and water in tears. The mucin in the aqueous layer,
mainly shed from the conjunctival goblet cells with its transmembrane form branching out from
the corneal and conjunctiva epithelium, also help maintain tear film stability. (Figure 1.1 A) Based
on its dynamic turnover, tears can be separated into basal tears, reflex tears and closed-eye tears.
Each tear type also has its own unique biochemical feature as well as its unique function. The
complex mixture of tears contains various functions including provide nutrients and oxygen to the
cornea, regulation of inflammatory processes, protection from pathogens and regulated corneal
wound healing. On othe ne hand tear content can affect corneal health, on the other tear
composition is also a reflection of certain ocular surface conditions.
The lacrimal gland (LG) is localized anteriorly in the superolateral aspect of the orbit. The
LG primary consists of lacrimal gland acinar cells (LGAC), making up 80% (w/w) of the gland
4
.
Other cells in the gland include ductal cells, myoepithelial cells and lymphocytes. The LGACs,
held together by tight junctions, generate an acinar structure and form polarized acinar cells. The
2
polarized LGACs have distinct apical and basolateral membranes
4
. (Figure 1.1 B) The basolateral
membrane contains many ion channels, ion transport proteins and receptors for neurotransmitters,
neuropeptides, and growth factors. The receptors on the basolateral membrane can initiate
secretion processes and stimulate intracellular trafficking of vesicles stores to the apical lumen.
Figure 1.1 The LG acini is responsible for secreting lacrimal fluid into the tear film. The LG secrete
lacrimal fluid, which makes up the aqueous layer of the tear film. The LG mainly consists of polarized
secretory epithelial cells, LGAC. Polarized LGACs contain have apical and basolateral membranes where
protein and fluid are mainly secreted through the apical membrane and released into the lumen.
1.2 Neural regulation in lacrimal gland secretion
The secretion of water, electrolytes and protein from the LG is tightly regulated by the afferent
sensory nerves in the cornea and conjunctivas, and efferent nerves innervating the gland
5
. The LG
is innervated with both sympathetic and parasympathetic nerves. Parasympathetic nerves are both
anatomically and functionally the predominant system in regulating protein secretion
5,6
. Upon
3
activation, the nerves release neurotransmitters, which interact with their specific receptors on the
basolateral membrane of LGACs and regulate secretion. The major neurotransmitters associated
with secretion include the parasympathetic cholinergic agonist, acetylcholine, and the sympathetic
adrenergic agonist, norepinephrine
6
.
Figure 1.2 The LG innervated with sympathetic and parasympathetic nerves. A functional unit for
neural regulation for tear secretion includes sensory nerves localized at the cornea and conjunction and
sympathetic and parasympathetic nerves innervating the LG. The sympathetic and parasympathetic nerves
secret neurotransmitters and activate specific receptors, such as M3AchR and α1D-adrenergic receptor.
Cholinergic agonists can modulate protein secretion through both increasing intracellular Ca
2+
and activating PKC. The cholinergic agonist first interacts with muscarinic M3 cholinergic
receptors (M3AchR) on the lacrimal gland, which subsequently activates G protein Gq/11α and
enzyme phospholipase C β (PLC β)
7
. The PLC β breaks down phospholipid phosphatidylinositol
biphosphate and generates 1,4,5-insositol triphosphate (InsP3) and diacylglycerol (DAG). InsP3
4
binds to its receptors on the endoplasmic reticulum (ER), leading to rapid release of Ca
2+
from
ER
8
. Concluding the Ca
2+
signaling, DAG activates protein kinase C (PKC) α, ε and δ, which
also stimulate secretion
9
.
While the adrenergic agonist, norepinephrine, can activate both α and β-adrenergic pathways,
α1D-adrenergic is the predominant functional pathway activated by norepinephrine in mouse and
rat LG
10
. The mechanism of α1D-adrenergic-induced secretion is controversial. Some suspect that
upon stimulation, α1D-adrenergic receptors activate endothelial nitric oxide (eNOS) synthase
localized on the basolateral membrane, and produce NO. NO then activates guanylate cyclase,
generates cGMP and stimulates secretion
11
. Others suggested that α1-adrenergic agonists affect
secretion by increasing Ca
2+
levels through inducing cADP-ribose and cGMP
12
. Another proposed
mechanism involves PKCs, where α1D-adrenergic agonist activates PKCε to prompt secretion
while PKCα and δ attenuate secretion
9
. However, due to limited studies, the molecular mechanism
of α1D-adrenergic stimulated protein secretion in the lacrimal gland remains unclear.
5
Figure 1.3 M3AchR and α1D-adrenergic signaling pathways. Activation of M3AchR leads to a significant
upregulation in intracellular Ca
2+
, and together with PKC pathways promote protein secretion in the
LGACs. α1D-Adrenergic signaling contributes to protein secretion through the NO/cGMP pathway. An
inhibitory Raf/MEK/MARK pathway is present in both M3AchR and α1D-adrenergic induced secretion.
1.3 Major conventional secretory trafficking pathway
The primary method of protein secretion is by apical exocytosis of proteins stored in mature
secretory vesicles (SV) or secretory granules, also known as the conventional major regulated
pathways. The SVs are formed from traffic originating in the Golgi apparatus and trans-Golgi
network, where secretory contents are sorted. During conventional major-regulated secretion,
Rab3D-enriched mature SVs are distributed on the subapical area, withheld from fusion by the
dense actin filaments adjacent to the PM. When stimulated, increased intracellular [Ca
2+
] triggers
compound vesicle fusion, leading to enlarged SV
13
. α1-Adrenergic and cholinergic agonists can
6
both activate PKC ε, which binds to both actin and activates proteins regulating actin filament
turnover
9
. The turnover of the apical actin network increases available fusion sites at the apical
membrane, while increased intracellular [Ca
2+
] and SNAREs further induces vesicle fusion to the
PM and released of contents. Proteins secreted through regulated exocytosis include lactoferrin,
β-hexosaminidase (β-hex) and peroxidase.
1.4 Rab3D and Rab27a trafficking functions in the LGAC
Major regulated exocytosis from the SV can be modulated with multiple molecular regulators
such as Rab proteins. Rab proteins are critical regulators of regulated exocytosis. As part of the
Ras superfamily of small GTPases, Rab proteins can modulate membrane association and
disassociation through regulated binding and hydrolysis of GTP
14,15
. The Rab3 and Rab27
subfamilies are commonly associated with SV exocytosis in neurons, endocrine, exocrine and
immune cells
16
. Rab3D is the most highly expressed Rab3 protein in LGAC, and appears to
modulate regulated exocytosis through regulation of SV maturation
17
and prevention of premature
fusion
18
. Analysis of Rab3D distribution in LGAC by immunofluorescence reveals its enrichment
on an abundant network of large (0.8–1 μm) apparent SV that are depleted following stimulation
of secretion in healthy LGAC. The SS model male NOD mice exhibit reduced expression and
altered LGAC distribution of Rab3D, concurrent with increased tear CTSS
19
. Previous work has
shown that tear CTSS activity is reduced in mice lacking functional Rab27 isoforms and, moreover,
is significantly increased in tears of Rab3D knockout mice which may exhibit a compensatory
increased function of Rab27 isoforms
19
. These findings implicate Rab27 isoforms in CTSS
secretion.
Rab27 has two isoforms that share 71% homology in amino acid sequence: Rab27a and
Rab27b
20
. Both isoforms are expressed on apparent SVs in pancreatic acini
21-23
, parotid acini
24
,
7
and LGACs
19
. Rab27b is linked to regulated exocytosis in pancreatic
21
and parotid acinar cells
24
,
as well as in LGAC
19,25
, since expression of dominant-negative (DN) Rab27b in LGAC
significantly impairs the process of stimulated secretion. The role of Rab27a in regulated
exocytosis has been extensively studied in various professional secretory cells including pancreatic
and parotid acinar cells, melanocytes and cytotoxic T-lymphocytes
26
. Rab27a has been implicated
in secretion from endosomes and lysosomes in certain cell types through direct fusion of
multivesicular bodies (MVB)
27,28
and lysosomes with the plasma membrane
29,30
. Rab27a is also
reported to modulate amylase release through a minor endolysosomal secretory pathway in mouse
pancreatic acinar cells
22
. Despite findings showing partial colocalization of Rab27a on apical SVs
enriched in Rab27b and Rab3D
19
, its functional role during exocytosis in LGAC is yet unclear.
1.5 Pikfyve in membrane trafficking
The metabolism of phosphoinositide lipids on the phago/endo/lysosomes are essential for the
maturation and trafficking of these organelles
31
. Phosphatidylinositol-3-phosphate 5-kinase
(PIKfyve) is a lipid kinase that phosphorylates phosphatidylinositol-3-phosphate (PI3P) to 3,5-
bisphosphate (PI(3,5)P2) through binding PI3P to its FYVE domain. PIKfyve together with its
product PI(3,5)P2 are key regulators of membrane homeostasis, vesicle transport and cargo
trafficking in the endolysosomal pathway
32,33
. PI(3,5)P2 is crucial for sorting proteins intro MVBs
and aids intralumenal vesicle formation through interacting with ESCRT-III compound Vps24.
PI(3,5)P2 also regulates lysosomal Ca
2+
release by v-ATPase, essential for the acidification and
lysosome maturation process
34,35
. Accordingly, deficiency in PIKfyve and its regulators results in
many physiologic problems including neurodegeneration, immune malfunction and lysosomal
diseases
32,36
. These physiological defects are suspected to relate to altered protein sorting into the
8
lysosomes, impaired autophagic flux, dysfunctional lysosomal Ca
2+
and massive enlargement of
lysosomes
37
.
Despite the regulating effects of PIKfyve in endolysosome maturation, only few studies have
investigated the relationship between PIKfyve and cathepsins. In dendritic cells, both
pharmacological and dominant negative inhibition of PIKfyve disrupts the trafficking and reduces
the activity of cathepsin S. Inactive cathepsin S further slows MHC II processing and antigen
loading. Previously, we have reported that cathepsin S is likely secreted through direct exocytosis
of endolysosomes. Therefore, it is of interest to investigate the relationship between PIKfyve and
cathepsin S trafficking in the LGACs.
Figure 1.4 PIKfyve is a main regulator for endolysosome maturation. PIKfyve is a lipid kinase
responsible for converting PI3P to PI(3,5)P2 , a key component in endolysosome maturation.
9
1.6 Dry eye disease
Dry eye disease (DED) is a multifactorial disease resulting in discomfort, visual disturbance,
and tear film instability and associated with an increase in tear osmolarity and ocular surface
inflammation
38
. Ocular surface irritation and hyperosmolarity along with the presence of increased
inflammatory mediators in tears contributes to the dysfunction of afferent sensory nerves and
efferent parasympathetic and sympathetic neurons, shutting down communication between the
ocular surface and LG, and promoting tear film instability and ocular surface inflammation
39
.
Dryness is a downstream consequence of this inflammatory cycle, which is associated with
increased production of many proteases in ocular surface tissues and tears
40
. The balance between
protease and protease inhibitors plays an important role in ocular surface homeostasis in DED
41
.
DED patients can be stratified into those with aqueous tear-deficient dry eye (ADDE) or
evaporative dry eye (EDE). Aqueous tear deficiency is caused by insufficient LG tear secretion
secondary to LG destruction or dysfunction, leading to tear hyperosmolarity that can stimulate
proinflammatory cascades
38
. LG dysfunction in ADDE leads to production of proinflammatory
mediators which are delivered to the ocular surface and into tears
38
. Aqueous tear deficiency can
be subdivided into 2 categories, Sjögren’s syndrome (SS) and non-Sjögren’s syndrome (non-
SS) associated ADDE. SS is a systemic autoimmune disease that leads to dryness of many mucosal
surfaces, including the eyes and mouth. A majority of SS patients are women, and the spectrum of
clinical symptoms can involve ADDE, development of interstitial lung disease, renal failure,
cryoglobulinaemic vasculitis, infections due to compromised mucosal surfaces, cardiovascular
disease and development of B-cell lymphoma
42,43
. Non-SS ADDE is associated with obstruction
of the LG duct or with age-related DED, which occurs in otherwise healthy individuals without
autoimmune disease
38
.
10
Evaporative DED is characterized by premature evaporation of the tear film in the setting of
normal LG secretion. Blepharitis is a primary cause of EDE, and can also be categorized into 2
subclasses, anterior and posterior blepharitis
44
. Meibomian gland dysfunction, often used as a
synonym for posterior blepharitis, is associated with altered lipid composition in the meibum, and
leads to reduced nonpolar lipids in tears and increased tear evaporation
45
. These categories are
summarized in Figure 1.5.
The diagnostic criteria for both ADDE and EDE are complicated: symptoms overlap, and
diagnosis is based on patient symptoms obtained from questionnaires and objective tests such as
tear film break-up time, corneal fluorescein staining and tear secretion tests
46
. Frequently,
objective tests do not correlate with patient-reported symptoms, complicating workup of DED
47
.
Proteases and their inhibitors have the potential for serving as biomarkers for DED because of the
correlation of protease dysregulation in the tear film with DED symptoms and pathology
48
.
Figure 1.5. Dry eyes disease classification. Dry eyes diseases (DED) are categorized into evaporative dry
eye (EDE) and aqueous deficient dry eye (ADDE). EDE is mainly caused by meibomian gland dysfunction
11
(MGD)/blepharitis. ADDE is primarily based on defective secretion by the lacrimal gland (LG). ADDE is
further subdivided into Sjögren’s syndrome (SS) related and non-Sjögren’s syndrome related (non-SS).
1.7 Sjögren’s syndrome (SS)
Sjögren’s syndrome (SS) is a systemic autoimmune disease associated with lymphocytic
infiltration of lacrimal glands (LG) and salivary glands (SG), associated with dacryoadenitis and
sialoadenitis, respectively.
49
SS patients develop associated complications including reduced tear
and saliva production, blurred vision, corneal damage, dental cavities and oral thrush.
49
Inflammation in the LG promotes release of pro-inflammatory cytokines to the ocular surface,
which can further compromise tear secretion by disruption of corneal sensory and efferent nerve
responses.
50,51
This further reduces tear flow and alters tear composition, resulting in pro-
inflammatory and proteolytic tears
52
which may elicit apoptosis and autophagy to further damage
the ocular surface.
53
Reduced tear volume, tear film instability and ocular surface inflammation all
contribute to the reduced visual acuity and increased patient discomfort associated with dry eye
symptoms in SS.
50,54
The male non-obese diabetic mouse (NOD) is commonly used as a model of the autoimmune
dacryoadenitis and ocular surface inflammation characteristic of SS. While the ocular
manifestations of SS spontaneously develop in the males from 8-10 weeks of age, the female
mouse develops a later autoimmune sialoadenitis from 14-16 weeks of age.
55
The male NOD mice
share many ocular surface system manifestations with SS patients including lymphocytic
infiltration of the LG
55,56
, reduced tear flow
19,55-57
, generation of a proteolytic tear film
19,55,57
,
altered distribution and expression of Rab3D
19,58
, reduced myoepithelial cells
59
and loss of
extracellular matrix
56
and elevated cytokines in LG and tears.
60
12
Figure 1.6. Symptoms and causes of Sjögren’s syndrome. Sjögren’s syndrome is an autoimmune disease
affected by age, gender, genetics and other autoimmune diseases. The hallmarks of sjögren’s syndrome are
inflammation in the lacrimal and salivary glands. Other related symptoms include dental cavities, joint pain
itchy skin etc.
1.8 Cathepsin S in Sjögren’s Syndrome
Cathepsin S (CTSS) activity is markedly elevated in the LG and tears of male NOD mice.
61
This lysosomal cysteine protease has diverse physiological functions including degradation of
extracellular matrix, processing of major histocompatibility complex II (MHC II) for antigen
presentation, and protein catabolism.
62,63
Due to its distinct role in antigen presentation, CTSS is
implicated in autoimmune diseases including rheumatoid arthritis (RA), systemic lupus
erythematosus and multiple sclerosis.
64
Following studies in the male NOD model
61
, CTSS activity
was demonstrated as significantly elevated in SS patient tears relative to tears of patients with non-
SS dry eye or other autoimmune diseases.
57
In vitro studies suggest that elevated tear CTSS may
13
affect ocular surface homeostasis, since it can induce expression and secretion of pro-inflammatory
cytokines and matrix metallopeptidase 9 (MMP-9) in a corneal cell line.
65
In the male NOD
mouse, CTSS activity and protein are increased not only in tears but also in the protein-secreting
acinar cells of the LG, suggesting additional actions.
19,60
With knowledge of the relationship of
other protease-sensitive pathways to inflammation, we sought out potential targets of CTSS that
might be implicated in the etiology of SS.
Figure 1.7. Cathepsin S (CTSS) is implicated in many biological functions. Lysosomal enzyme, CTSS,
is involved in many biological processes including lysosome protein degradation, antigen presentation,
extracellular matrix degradation, PAR2 activation and CX3CL1 induction and shedding.
14
1.9 CX3CL1 characteristics and functions
CX3CL1 (fractalkine) is the only member of the subclass of CX3C chemokines.
66
Existing in
two forms, membrane-bound and soluble, it is one of only two chemokines to exist in both
forms.
66,67
Synthesized as the membrane-bound form, it is cleaved to the soluble form through
proteases including metallopeptidase 10 (ADAM-10)
68
, metallopeptidase 17(ADAM-17)
69
, MMP-
9
70
and CTSS
71
. Some studies have suggested that organ-specific cleavage products resulting from
different protease cleavage may have distinct actions.
70
Membrane-bound CX3CL1 has a
chemokine domain, a mucin-like stalk, a transmembrane α-helix and a cytoplasmic tail, while
soluble CX3CL1 contains only the chemokine domain and mucin-like stalk.
67
Both membrane-
bound and soluble CX3CL1 recognize the G protein-coupled receptor, CX3CR1, abundantly
expressed on immune cells including natural killer (NK) cells, dendritic cells, T-cells, monocytes
and macrophages.
72
Both forms of CX3CL1 are implicated in adhesion and chemoattraction.
73
CX3CL1 is frequently expressed on vascular endothelium, where it facilitates leukocyte
extravasation by capturing CX3CR1-expressing cells from the blood, inducing transmigration and
recruitment to tissue.
74
CX3CL1 has been implicated in systemic lupus erythematosus
75
and rheumatoid arthritis.
76
Intriguingly, CX3CL1 is also elevated in the serum and SG of primary SS patients.
77
A previous
study also showed that CX3CL1 expression was elevated in LG of thymectomized NFS/sld mice,
another model of SS-associated dacryoadenitis.
78
We hypothesized that the elevated CTSS in tears
and LG in SS might participate in the generation of soluble CX3CL1, contributing to pathogenesis
in the lacrimal functional unit.
15
Figure 1.8. CX3CL1 (Fractalkine) is a unique protease cleavable chemokine. CX3CL1 (fractalkine) is
a unique chemokine, expressed by neurons and endothelial cells, that can be cleaved by proteases including
Cathepsin S, ADAM7 and ADAM10. Both membrane-bound and soluble CX3CL1 interact with its specific
receptor CX3CR1 and initiate biological functions including immune cell recruitment, neuroprotection and
chronic neuropathic pain.
IL-1, IL-6, IL-8, TNF-α, interferon-γ (IFN-γ) and IL-17 are all increased in SS patient
tears.
79-82
In vitro exposure of LGACs to IFN-γ evokes some of the same SS-associated changes
characteristic of disease, including altered organization of Rab3D and induction of antigen
presentation.
60
Since IFN-γ also induces the expression of membrane-bound CX3CL1 and
reinforces its functions
58
, we used in vitro exposure to IFN-γ to induce CTSS and CX3CL1
expression to further study their interactions.
83
16
1.10 Goals
Despites many studies associating tear CTSS with SS pathogenesis
56,84
, the mechanism of
secretion, as well as why CTSS activity is increased in the tear is still not clear. In these studies, I
plan to 1. Elucidate how CTSS contributes to diseases, through investigating its relationship with
a downstream cytokine CX3CL1. 2. Identify Rab27a’s role in CTSS trafficking in the LGAC. 3.
Explore neural regulatory effect in CTSS trafficking in the LGACs.
17
CHAPTER 2
CX3CL1, together with CTSS activity, is increased in tear of SS
murine model
2.1 CX3CL1 (fractalkine) is elevated in tears of NOD mice.
To determine whether CX3CL1 was altered in tears of diseased male NOD mice versus
healthy age-matched BALB/c mice, we measured CX3CL1 abundance in tears and tissue from
these models. A significant increase in CX3CL1 was observed in tears of NOD mice (Figure 2.1A),
correlated with its decrease in LG after CCh-induced tear stimulation (Figure 2.1D). In
unstimulated LG, CX3CL1 abundance was not significantly different between strains (Figure
2.1C). CTSS activity was measured in the same tear samples in Figure 2.1A, revealing a marked
increase in NOD mice (Figure 2.1B) paralleling increased CX3CL1. Membrane-bound CX3CL1
has multiple cysteine protease cleavage sites and could undergo cleavage by CTSS, consistent with
increased soluble CX3CL1 in NOD mouse tears. CX3CL1 was also significantly elevated in NOD
mouse serum relative to BALB/c mice (Figure 2.1E), consistent with reports in SS patients.
77
18
Figure 2.1. CX3CL1 protein and CTSS activity are elevated in NOD mouse tears relative to BALB/c
mouse tears. A. Soluble CX3CL1 (fractalkine) was significantly increased in stimulated tears from NOD
mice compared to BALB/c mice (p=0.0116, N=3). B. CTSS activity was significantly increased in the same
reflex tears of NOD mice relative to BALB/c mice used for CX3CL1 analysis in A. (p=0.0001, N=3). C.
CX3CL1 levels in NOD and BALB/c mouse LG in the absence of topical stimulation were not significantly
different (p=0.1501, N=4). D. CX3CL1 levels in NOD mouse LG were significantly reduced relative to
BALB/c mouse LG levels by topical CCh stimulation (p=0.0057, N=4). E. CX3CL1 was significantly
increased in the corneal button from NOD mice relative to BALB/c (p=0.0059, N=4). F. CX3CL1 was
significantly elevated in NOD mouse serum relative to serum from BALB/c mice p=0.0080, N=4). In A.,
B. and e., each N value consisted of tears or cornea button from 2 mice, pooled for 1 sample, while in C.,
D., and F., N=mouse number. Panels A., C., D., E. and F. were measured using ELISA, while panel B. was
measured with a commercial CTSS fluorescence activity kit. All data are presented as mean ± SD.
19
2.2 CX3CL1 and CTSS are increased in NOD mouse LGAC.
Many proteins present in aqueous tears are produced and secreted by LGAC. To elucidate the
source of increased CX3CL1 in NOD mouse tears, we compared the expression of Cx3cl1 and
Ctss in NOD and BALB/c LG. Gene expression of Ctss, together with the pro-inflammatory
cytokines, Ifn-γ and Tnf-α, were significantly elevated in NOD mouse LG (Figure 2.2A) consistent
with previous reports.
60
No changes in Cx3cl1 gene expression were observed (Figure 2.2A). The
percentage of LGAC relative to other cells in the LG decreases as lymphocytes extensively
infiltrate the gland in disease. With this knowledge and the observation that CX3CL1 protein was
decreased in stimulated LG from NOD mice (Figure 2.1D), we analyzed its abundance specifically
in LGAC using quantitative immunofluorescence. Figure 2.3 shows a section across the entire LG
indicating increased CX3CL1 protein in NOD LG compared to BALB/c. Quantification of proteins
distributed in LGAC (Figure 2.2B) with the corrected total cell fluorescence (CTCF) method
85
revealed that CX3CL1 (Figure 2.2B, C) and CTSS protein levels (Figure 2.2B,D) were
significantly elevated in LGAC from NOD mice compared to BALB/c. Magnified images of
LGAC (Figure 2.2B) demonstrated vesicular basolateral and apical enrichment of CX3CL1 and
CTSS, consistent with possible secretion/shedding at either plasma membrane domain.
20
Figure 2.2. CX3CL1 and CTSS expression are elevated in NOD mouse LG relative to BALB/c. A.
Gene expression of Ctss and pro-inflammatory cytokines were significantly elevated in LG from NOD mice
relative to BALB/c mice, but no difference in Cx3cl1 gene expression was observed (Cx3cl1: p=0.901,
Ctss: p<0.0001, Ifng: p=0.007, Tnf: p<0.0001, N=5). Ifng gene expression was undetectable in 3 BALB/c
mice. B. Indirect immunofluorescence revealed increased levels of CX3CL1 and CTSS protein in acinar
cells in NOD mouse LG relative to levels in BALB/c. In zoomed images of NOD LGAC, apical and
basolateral CX3CL1/CTSS are marked with red and blue arrowheads, respectively, and the lumen is marked
with an asterisk. Image quantification is shown in C. with p=0.0026 for CX3CL1 and D. with p<0.0001 for
CTSS, respectively (N=4). For immunofluorescence analysis, N=mice per group while sections from each
mouse LG were evaluated in 6 ROI per sample, each of which is represented as points on the graph. All
data are presented as mean ± SD.
21
Figure 2.3. CX3CL1 is increased in a cross section of NOD mouse LG compared to BALB/c
mouse LG. Arrows indicate sites of increased CX3CL1 (green).
2.3 CX3CL1 is increased in corneal buttons from male NOD LG.
To determine whether CX3CL1 was affected on the ocular surface, we measured CX3CL1
levels in the corneal button of NOD and BALB/c mice. Cx3cl1 gene expression in NOD mouse
corneas was significantly higher compared to BALB/c (Figure 2.4A). CX3CL1 protein in the
mouse corneal button measured by ELISA similarly showed a significant increase in NOD mice
(Figure 2.4B). In corneal cross sections, we found that CX3CL1 was mainly present in the
epithelium, with no obvious differences in CX3CL1 distribution between BALB/c and NOD
mouse (Figure 2.4C).
22
Figure 2.4. Cx3cl1 is elevated in the corneal button of NOD mouse. A. Cx3cl1 gene expression was
significantly elevated in the corneal buttons of NOD mice relative to BALB/c mice (p=0.0008, N=5). N,
mRNA sample pooled from 2 mouse. B. CX3CL1 immunofluorescence in corneal epithelium, quantified
from C, showed a significant increase in NOD mice (p=0.0195, N=4). N, mouse number. 3 ROIs per sample
were quantified, presented as points on the graph. C. Confocal fluorescence microscopy shows CX3CL1
protein distribution in the plasma membrane and membrane organelles in the corneal epithelium. Green,
CX3CL1; red, rhodamine phalloidin labeling F-actin; blue, DAPI. All data are presented as mean ± SD.
23
2.4 Cx3cr1 expression and CX3CR1+ immune cells are increased in male
NOD LG.
CX3CL1 contributes to immune cell recruitment through interactions with its receptor, CX3CR1,
which is largely expressed on immune cells. We analyzed CX3CR1 expression and enrichment on
immune cells in NOD and BALB/c LG. In whole LG, Cx3cr1 gene expression was significantly
elevated in NOD mice compared to BALB/c (Figure 2.5A). To explore whether increased
CX3CR1 gene expression was related to increased numbers of CX3CR1+ cells, LGs were isolated
and processed as single cell suspensions for flow cytometry profiling. The immune cell (CD45+)
population expressing high CX3CR1 protein is shown in the gated area in Figure 2.5B and
quantified in Figure 2.5C. A significantly greater number of CD45+ immune cells were seen in
NOD mouse LG relative to BALB/c (Figure 2.5C) within same number of events. To further
identify the CX3CR1+ populations, we evaluated the presence of NK cells (CD335+ CD11b-
CD11c-), dendritic cells (CD11c+ F4/80-), macrophages (CD11b+ F4/80+ CD11c-), monocytes
(CD11b+ Ly6C+ CD11c-) and T-cells (CD3+ CD11b- CD11c-). Figure 2.5D shows the gated
populations in color overlaying the total immune cell population in grey. The histograms in Figure
2.5E validate the identity of each respective gated population by demonstrating related marker
expression. T-cells were identified by high CD3+ levels, dendritic cells by CD11c, natural killer
cells by CD335, macrophages by F4/80 and monocytes by Ly6C. With this strategy, we identified
the frequencies and percentages of cells in each CX3CR1+ population, and determined the
differences in these cell types between BALB/c and NOD mouse LG. The absolute numbers of
CX3CR1+ NK cells, CD11b+ and CD11b- dendritic cells, macrophages and T-cells were all
significantly elevated in NOD mice LG (Figure 2.5F). Among CX3CR1+ immune cell
populations in NOD LG, the percentage of T cells increased by 5-fold and macrophages by 2-fold
24
(Figure 2.5G). These results suggest a population of CX3CR1+ immune cells showing increased
migration into LG tissue in autoimmune dacryoadenitis, concurrent with CX3CL1 and CTSS
elevation in diseased LG and tears.
25
Figure 2.5. CX3CR1+ immune cells are increased in NOD mouse LG. Flow cytometry was performed
on isolated cells from LG from BALB/c and NOD mice. A. Cx3cr1 gene expression in LG was significantly
increased in NOD mice relative to BALB/c mice. p=0.0375, N=5 mice. B. CX3CR1+ cells from BALB/c
and NOD mouse LG after gating for live CD45+ singlets. C. Absolute cell numbers of CX3CR1+ cells
26
from panel B were increased in NOD mouse LG compared to BALB/c. p<0.0001, N=9 mice. D. Validation
of immune cell identities in mouse LG. Populations are demonstrated in NOD mouse LG, after gating for
live singlet CD45+. NK= natural killer cells, DC= dendritic cells, M𝜙= macrophages, Mo= Monocytes. E.
Histograms from the respective gated populations, supporting immune cell population validation. F. The
CX3CR1+ CD45+ cell population was further identified based on the gating and validation method
illustrated in panel D. and E. Absolute numbers of NK cells (p=0.0015), CD11b- (p=0.0002) and CD11b+
dendritic cells (p<0.0001), macrophages (p=0.0012) and T cells (p=0.0159) were significantly elevated in
NOD mice. N=3~6, mice per immune cell type. G. CX3CR1+ cell cells in NOD mouse LG are altered
compared to BALB/c mouse LG. All data are presented as mean ± SD.
27
CHAPTER 3
Proinflammatory cytokines and CTSS activity upregulates CX3CL1
expression and shedding in LGACs and HCE-T cells.
3.1 In vitro IFN-γ induces CX3CL1 in BALB/c LGAC.
Previously, we demonstrated that in vitro IFN-γ treatment of BALB/c mouse LGAC induces
changes comparable to those seen in NOD mice LG. Also, as in previous studies, Ifn-γ gene
expression was elevated in NOD mouse LG (Figure 2.2C). To further explore its effect on
CX3CL1 in vitro, we measured Cx3cl1 gene expression in cultured LGAC from BALB/c mice
treated with active recombinant mouse IFN-γ for 2, 4, 8 and 24 hr. CX3CL1 gene expression was
significantly elevated by 4 hr of IFN-γ treatment and peaked by 8 hr (Figure 3.1A). CX3CL1 and
CTSS protein levels in LGAC treated with IFN-γ were analyzed by quantitation of
immunofluorescence (Figure 3.1D). CX3CL1 protein in LGAC was elevated at shorter treatment
times, peaking at 4 hr and dropping to untreated levels by 24 hr (Figure 3.1B). In comparison,
CTSS protein was increased by 4 hr of treatment and remained consistently elevated through the
treatment period (Figure 3.1C). Decreased CX3CL1 in LGAC at 24 hr is consistent with cleavage
of the membrane-bound form and release of the soluble form, particularly with the increased CTSS
expression measured. In culture medium, a 4-fold increase in CX3CL1 was seen at 24 hr (Figure
3.1E). To test whether more CTSS could enhance this recovery, rhCTSS was added to LGAC for
2 hr after pretreatment with IFN-γ for 8 hr to induce higher CX3CL1 levels prior to any observed
shedding. CX3CL1 levels in cells, measured by CTCF, were significantly decreased with
extracellular CTSS treatment (Figure 3.1F), but CX3CL1 levels in the media were unchanged
with additional rhCTSS (Figure 3.1G).
28
Figure 3.1. Recombinant mouse IFN-γ (200 U/ml) induces Cx3cl1 gene expression and
elevates CX3CL1 protein in culture medium of cultured LGAC from BALB/c mice. A. Gene
expression of Cx3cl1 in cultured LGAC from BALB/c mice was increased by IFN-γ treatment.
Relative to untreated: 4 hr, p=0.0014; 8 hr, p<0.0001; 24 hr, p=0.0011. Between groups: 2 hr
vs. 4 hr, p=0.0319; 2 hr vs. 8 hr , p<0.0001; 2 hr vs. 24 hr, p=0.0238; 4 hr vs. 8 hr p=0.0001; 8 hr
vs. 24 hr, p=0.0001 (N=3). B. CX3CL1 protein in LGAC as quantified by immunofluorescence
intensity from confocal microscopy images typical of those in panel E, showed significant
increases in CX3CL1 from 2-8 hr of IFN-γ incubation. Relative to untreated: 2 hr, p=0.0025; 4 hr,
p=0.0005; 8 hr, p=0.0014 (N=3). C. CTSS in LGAC as quantified from images typical of those in
29
panel E, showed increased CTSS by 4 hr of IFN-γ sustained throughout the treatment time.
Relative to untreated: 4 hr, p<0.0001; 8 hr, p=0.0011; 24 hr, p<0.0001. Comparison between
different IFN-γ treatment times showed: 2 hr vs. 24 hr, p=0.0334 (N=3). D. Confocal fluorescence
microscopy of LGAC shows that CX3CL1 and CTSS image intensities change with time of
exposure to IFN-γ. Green, CX3CL1; red, F-actin; blue, DAPI staining. E. Soluble CX3CL1
detected by ELISA in culture media was significantly elevated by IFN-γ treatment for 24 hr.
Relative to untreated: 24 hr, p<0.0001. Comparison between different IFN-γ exposure times: 2 hr
vs. 24 hr, p<0.0001; 4 hr vs. 24 hr, p<0.0001; 8 hr vs. 24 hr; p<0.0001 (N=3). F. Membrane-bound
CX3CL1 analyzed by immunofluorescence was significantly reduced with 2 hr incubation with
active recombinant CTSS (20,000 RFU/treatment) in LGAC after pretreatment with IFN-γ for 8
hr. Ratio paired t-test, p=0.0322. G. In medium collected from the experiments in panel F., soluble
CX3CL1 was not significantly increased by additional CTSS treatment in the presence of IFN- γ.
Paired t-test: p=0.3503, From panels A. to E., comparisons between treated and untreated cells
were tested with one-way ANOVA Dunnett’s multiple comparison test. Comparison between
treatment groups was analyzed with one-way ANOVA with Tukey’s multiple comparison test. N=
preparations, Each preparation included LGAC isolated from 10 BALB/c mice. For confocal
microscopy image quantification, 5 ROIs were obtained from each preparation and represented as
points on the image. All data are presented as mean ± SD.
3.2 IFN-γ treatment increases CX3CL1 in HCE-T cells and culture medium.
In addition to CTSS, we previously demonstrated increased IFN-γ in tears of male NOD mice.
60
We analyzed whether IFN-γ affected CX3CL1 expression and abundance in corneal epithelial cells,
in parallel with CTSS. HCE-T cells were grown to 70% to 80% confluency, then starved in
supplement-free KSFM media for 16-18 hr prior to IFN-γ treatment in complete KSFM media for
0.5, 1, 1.5, 2, 4, 8, 24, and 48 hr. CX3CL1 gene expression was elevated by 1.5 hr of IFN-γ
treatment, peaked at 2 hr and remained elevated through 8 hr of treatment before dropping at 24
hr (Figure 3.2A). CTSS required longer IFN-γ exposure for increased gene expression, increasing
significantly by 4 hr of treatment and remaining elevated through 48 hr (Figure 3.2B). To correlate
CX3CL1 abundance and CTSS activity in HCE-T medium and lysates with observed changes in
gene expression, HCE-T cells were treated with IFN-γ for 2, 4, 8 and 24 hr in complete KSFM
medium after starvation. Cell medium and lysates were collected and analyzed using ELISA for
CX3CL1. CX3CL1 levels in cell lysates were significantly increased at 2 hr of IFN-γ treatment,
30
peaked at 4 hr, then decreased from 8 hr (Figure 3.2C), similar to observed patterns of gene
expression (Figure 3.2A). These effects occurred without a change in CTSS activity in cell lysates
(Figure 3.2D). However, CX3CL1 in culture medium increased with IFN-γ treatment at 4 hr and
continued throughout the 24 hr treatment (Figure 3.2E). CTSS activity in the lysates also was
significantly elevated at 2 and 4 hr, prior to the accumulation of CX3CL1 in culture medium
(Figure 3.2F).
31
Figure 3.2. Recombinant human IFN-γ (1 µg/ml) induces both CX3CL1 and CTSS gene
and protein expression in HCE-T cells. A. Gene expression of CX3CL1 was increased by IFN-
γ. Relative to untreated: 1.5 hr, p<0.0001; 2 hr, p<0.0001; 4 hr, p<0.0001; 8 hr, p<0.0001; 48 hr,
p<0.0001. N=6, B. CTSS gene expression was increased by IFN-γ. Relative to untreated: 4 hr,
p<0.0001; 8 hr, p<0.0001; 48 hr, p<0.0001. N=6. C. Soluble CX3CL1 in cell lysate measured by
ELISA was increased by IFN-γ. Relative to untreated: 2 hrs, p=0.0256; 4 hr, p<0.0001; 8 hr,
32
p=0.0068. Compared between treatments: 2 hr vs. 4 hr, p=0.0074; 4 hr vs. 8 hr, p=0.026; 4 hr vs.
24 hr, p=0.0002; 8 hr vs. 24 hr, p=0.0244. N=3. D. No change in CTSS activity in cell lysates
was elicited by IFN-γ. N=3, E. CX3CL1 measured by ELISA in culture medium was increased by
IFN-γ: 4 hr, p=0.0104; 8 hr p<0.0001; 24 hr, p<0.0001. Compared between treatments: 2 hr vs. 4
hr, p=0.0081; 2 hr vs. 8 hr, p<0.0001; 2 hr vs. 24 hr, p<0.0001; 4 hr vs. 8 hr, p=0.0051; 4 hr vs.
24 hr, p<0.0001; 8 hr vs. 24 hr, p=0.0022. N=3. F. CTSS activity in culture medium was increased
with IFN-γ treatment. Relative to untreated: 2 hr, p= 0.0049; 4 hr, p=0.0003. Compared between
treatments: 4 hr vs. 8 hr, p=0.01; 4 hr vs. 24 hrs, p=0.006. N=3. One-way ANOVA with Dunnett’s
multiple comparison was used to compare different times of IFN-γ treatments to untreated. One-
way ANOVA with Tukey’s multiple comparison was used to compare between IFN-γ treatments.
All data are presented as mean ± SD.
3.3 rhCTSS increases CX3CL1 expression in HCE-T cells.
CTSS activates the expression and secretion of other cytokines and proteases in HCE-T
cells.
65
To evaluate whether CTSS could also induce CX3CL1, HCE-T cells were exposed to
rhCTSS at activity commensurate with that in SS patient tears, and changes in CX3CL1 gene and
protein expression were measured.
65
CX3CL1 gene expression was significantly increased by 8 hr
of rhCTSS, and returned to pre-exposure levels by 24 hr (Figure 3.3A). CX3CL1 protein, detected
by immunofluorescence and quantified with CTCF method, was significantly increased by 24 hr
of rhCTSS (Figure 3.3B, C). These results suggest that extracellular CTSS activity can induce
CX3CL1 expression in HCE-T cells.
33
Figure 3.3. rhCTSS exposure induces CX3CL1 gene and protein expression in HCE-T
cells. HCE-T cells grown to 70-80% confluency was starved in supplement-free KSFM basal
medium for 16-18 hr and treated with rhCTSS (20,000 RFU/treatment) for 2, 4, 8 and 24 hr. A.
CX3CL1 gene expression was increased by CTSS. Relative to untreated: 8 hr, p=0.0021 (N=3),
Data points indicate the individual mRNA samples (3 for each N or repeat) presented as points.
B. CX3CL1 protein was increased by rhCTSS as determined by quantitation of fluorescence
intensity using immunofluorescence labeling and confocal microscopy of images such as those
in panel C. Relative to untreated: 24 hr, p<0.0001. Compared between samples: 2 hr vs. 24 hr,
p<0.0001; 4 hr vs. 24 hr, p<0.0001, 8 hr vs. 24 hr, p<0.0001 (N=3). N=experimental repeats
performed on different days, 3 ROIs were analyzed under each condition for each repeat and are
presented as points. All data are presented as mean ± SD.
3.4 Extracellular CTSS enhances CX3CL1 shedding in HCE-T cells.
CTSS can generate soluble CX3CL1 though direct proteolysis.
86
To investigate whether
active CTSS cleaves membrane-bound CX3CL1 induced in corneal epithelial cells, the abundance
of cell-associated and soluble CX3CL1 in HCE-T cell cultures was determined under conditions
of differing CTSS activity. To induce high expression of CX3CL1, HCE-T cells were pretreated
with recombinant human IFN-γ (1 µg/mL) for 2, 4, 8, and 24 hr (Fig 3.5). rhCTSS was then added
34
to IFN-γ treated HCE-T cells for 2 hr. The choice of this relatively short-term exposure of rhCTSS
to the cells was intended to minimize the influence of CTSS on gene expression (Fig 3.4), and to
focus on enzyme-mediated actions of CTSS. Cell medium and cell lysates were collected post-
treatment and analyzed by ELISA. Soluble CX3CL1 in the cell medium was significantly elevated
by rhCTSS, with the highest increases at 4 hr and 8 hr of treatment with IFN-γ (Fig 3.4A).
Consistent with elevated CX3CL1 in culture medium, CX3CL1 content of cell lysates was slightly
decreased by rhCTSS at 4 hr and 8 hr of IFN-γ induction (Fig 3.4B). To confirm that effects were
due directly to CTSS activity, HCE-T cells exposed to 24 hr IFN-γ were treated with heat-
inactivated rhCTSS (HI-rhCTSS) in the last 2 hr of treatment, which decreased CX3CL1 in culture
medium relative to rhCTSS or no CTSS. Correspondingly, CX3CL1 in cell lysates was
significantly increased with HI-rhCTSS treatment compared to rhCTSS or no CTSS (Fig 3.4C).
The amount of CX3CL1 present in culture medium and cell lysates of HCE-T cells exposed to
IFN-γ for 24 hr and then treated without or with Z-FL, a specific CTSS inhibitor, in the final 2 hr
also showed that the CX3CL1 content of the culture medium was significantly reduced by Z-FL
treatment while CX3CL1 in cell lysates was not significantly affected (Fig 3.4D). At the same
time, CTSS activity in the cell culture medium was decreased significantly with Z-FL, with no
change in CTSS in cell lysates (Fig 3.4E).
35
Figure 3.4. rhCTSS increases CX3CL1 recovery in culture medium. A. HCE-T cells
grown to 70-80% confluency and starved in supplement-free KSFM basal medium for 16-18 hrs
were treated without and with recombinant human IFN-γ (1 µg/ml) for 2, 4, 8 and 24 hrs.
Additional active rhCTSS (20,000 RFU/treatment) was added during the last 2 hr of IFN-γ
treatment. CX3CL1 in culture medium was increased by rhCTSS treatment. Relative to untreated
and at the equivalent IFN-γ treatment time: 4 hr: p<0.0001; 8 hr, p<0.0001; 24 hr, p=0.0309,
(N=3). B. CX3CL1 in cell lysates was not significantly affected by rhCTSS (N=3). C. Heat-
inactivated rhCTSS (HI-rhCTSS), at equivalent concentration to active rhCTSS as shown in A.
36
and B., was incubated for 2 hr with HCE-T cells pretreated with IFN-γ for 24 hr. Soluble CX3CL1
in the culture medium was significantly decreased by HI-rhCTSS treatment compared to active
CTSS treatment (p=0.0346) and untreated IFN-γ-induced cells (p=0.0206). CX3CL1 in cell
lysates was significantly increased with HI-rhCTSS treatment compared to CTSS treatment
(p=0.0425). (N=3) D. HCE-T cells pretreated with IFN-γ for 24 hr were exposed to the CTSS
inhibitor, Z-FL (1 µM), for the last 2 hr. Soluble CX3CL1 in the culture medium was significantly
decreased with Z-FL treatment (p=0.0232), but unaffected in cell lysates. (N=3) E. CTSS activity
in culture medium was also significantly reduced by Z-FL treatment, (p=0.0048) with no changes
in CTSS activity in cell lysates. (N=3) All N values are experimental repeats performed on
different days. All data are presented as mean ± SD.
37
CHAPTER 4
Rab27a is involved in CTSS secretion in lacrimal gland acinar cells
4.1 CTSS and Rab27a are enriched in the same vesicles in the lacrimal gland
acinar cells in male NOD mouse
We have previously reported increased CTSS activity in the tears of male NOD mouse, as
well as enhanced CTSS gene and protein expression in the LG with disease progression
56
. Analysis
of primary cultured LGAC transduced with recombinant Ad-CTSS-GFP also showed that
exogenous CTSS was first sorted into endolysosomes upon synthesis, and was eventually
trafficked to subapically-enriched vesicles enriched in Rab27b
19
. 3D-structured illumination
microscopy showed that Rab27a was partially colocalized with Rab27b on subapical SV
19
. In this
study we utilized 16–17 week old male NOD mice as a model of SS-associated dacryoadenitis,
which exhibits significant ocular manifestations of disease, without exhibiting hyperglycemia. To
determine whether CTSS was detected in Rab27a-enriched subapical vesicles in diseased LGAC,
we analyzed LG tissue from male BALB/c and NOD mice labeled with antibodies to CTSS and
Rab27a for indirect immunofluorescence, as shown in Figure 4.1. In addition to its elevated
expression in male NOD LG, as we have previously demonstrated [17,24,26], CTSS distribution
is also altered. While showing the enrichment in small puncta detected throughout the cytoplasm
seen in LGAC in BALB/c mouse LG sections, CTSS showed an increased abundance and striking
accumulation beneath both apical and basolateral membranes in NOD mouse LG. Notably, the
increased apically-enriched CTSS in the diseased male NOD mouse LGAC was colocalized with
subapical Rab27a, suggesting that the increased CTSS stores accumulating in these cells were
enriched in Rab27a-enriched vesicles.
38
Figure 4.1. Increased enrichment of CTSS and Rab27a in vesicles in LGAC from male NOD
mice. CTSS and Rab27a distribution in LGAC from LG from 16–17 week male BALB/c and NOD
LG sections were characterized by immunofluorescence. CTSS immunofluorescence intensity
(green) was clearly increased in NOD mouse LG sections, while also showing an increased
colocalization with Rab27a (red) in the subapical area. Staining adjacent to the plasma membrane
was observed for Rab27a due to secondary anti-mouse antibody reaction with residual interstitial
immunoglobulins. In order to visualize intracellular Rab27a above this high interstitial
background, the fluorescence signal in the interstitium was acquired under conditions of saturation.
N = 4, N = number of mice. *, Lumen.
39
4.2 In Male NOD Mouse LG, Rab27a-Enriched Vesicles Are Subapically
Increased and Their Size Decreased, Relative to Rab27a-Enriched Vesicles in
Healthy Male BALB/c Mouse LG
To further evaluate Rab27a distribution and expression in healthy and diseased LGAC, we
quantified the intensity of immunostaining/acinus as well as the size of Rab27a-enriched vesicles
in LGAC from male BALB/c and NOD mouse LG sections. We also analyzed Rab27a expression
in the same samples through qRT-PCR. Confocal microscopy imaging of indirect
immunofluorescence demonstrated that Rab27a-enriched vesicles were present in both male
BALB/c and NOD mouse LGAC near apical and basolateral membranes (Figure 4.2A).
Quantification of Rab27a immunofluorescence per acinar cell using the corrected total cell
fluorescence (CTCF) method
87,88
revealed that Rab27 accumulation was significantly elevated in
NOD mouse LGAC, as shown in Figure 4.2B. Moreover, the increase was apparent, by visual
inspection to be more prominent in the subapical area. No change was observed in Rab27a gene
expression through qPCR (Figure 4.2D). Since Rab27a undergoes a GTP-GDP cycle that controls
its membrane association and dissociation
89
, the increased Rab27a intensity associated with
membranes may be related to protein activation and increased membrane recruitment rather than
altered expression. Quantification of vesicle size also showed a significant reduction in Rab27a-
enriched vesicle diameter in sections from male NOD LG relative to BALB/c mouse LG (Figure
4.2C). This decreased vesicle size may result from the inability of Rab27a-enriched vesicles to
sustain normal trafficking and possibly, merger with subapical Rab3D-enriched SV that are
decreased in diseased male NOD mouse LGAC
19
. It has been reported that Rab3D is required to
maintain normally-sized SV
18
. In other secretory cells, a reduction in SV size is associated with
40
reorganization of contents and/or vesicle condensation
90
. These observations suggested possible
alteration in vesicle content as well a function of Rab27a-enriched apical vesicles in NOD mouse.
Figure 4.2. Increased accumulation on subapically-enriched vesicles and reduced vesicle
diameter characterizes Rab27a-enriched vesicles in male NOD mouse. Tissue sections from
3–4 mice per strain were quantified, 3 ROI were obtained from each section. A. Indirect
immunofluorescence microscopy of Rab27a labeling (green) revealed increased accumulation of
Rab27a-enriched vesicles near the apical lumen. B. The increased intracellular accumulation of
Rab27a was quantified per acinus, and was increased in male NOD mouse acini relative to BALB/c
mouse acini. ***, p = 0.0003 (N = 4). Red, actin filaments. Integrated density of three acini were
quantified and averaged in each ROI, pre-sented as points on the graph. C. Feret’s diameter of
Rab27a-enriched vesicles was significantly decreased in LG sec-tions from NOD mice relative to
vesicles in LG sections from BALB/c mice.****, p ≤ 0.0001 (N = 4). 30–40 vesicles from three
acini were quantified from ROI, presented as points. D. No significant changes in Rab27a gene
expression were detected in NOD versus BALB/c mouse LG. ns, p = 0.6057 (N = 6). N = mice
per group. *, Lumen.
41
4.3 Ad-mCFP-Rab27a Construct Design and Characterization
In order to study the trafficking of Rab27a in vitro in primary LGAC, we generated
monomeric (m) CFP-Rab27a WT and DN adenovirus (Ad) expression constructs. The
doxycycline-induced Tet-on Ad construct, shown in Figure 4.3A, was generated by inserting
fragments encoding WT and DN mCFP-Rab27a into a linearized pAdenoX-Tet-3G vector. DN
Rab27a, with the T23N mutation shown in Figure 4.3B, mimics the GDP-bound form of the
protein, inhibiting its function. The Ad constructs were characterized by diagnostic enzyme
digestion and DNA sequencing (Figure S1) and packaged into 293a helper cells for virus
production. High titer Ad were obtained after serial transfection and amplifications, characterized
though Western blotting and immunofluorescence of primary transduced mouse LGAC (Figure
4.4). Western blotting with antibodies to Rab27a and mCFP demonstrated a band of mCFP-Rab27a
at ~54 kDa, while the anti-Rab27a antibody also showed a lower band (~27 kDa) representing
endogenous Rab27a. Since our functional secretion studies routinely use primary cultured rabbit
LGAC, due to the superior cellular yield, we also imaged rabbit LGAC transduced with WT and
DN mCFP-Rab27a (MOI = 4–6), (Figure 4.3C) demonstrating that overexpressed Rab27a had a
similar distribution as that observed in LG tissue (Figure 4.2A) when the CFP was visualized
directly. Immunostaining for mCFP in fixed rabbit LGAC also showed vesicle-like structures
distributed beneath the apical lumen (Figure 4.4). Estimates of transduction efficiency in primary
rabbit LGAC are estimated as 40–50% (Figure 4.5). In rabbit LGACs expressing DN mCFP-
Rab27a shown in Figure 4.3D, Rab27a signal was more diffuse and mainly localized in the cytosol.
42
Figure 4.3. WT and DN Ad-mCFP-Rab27a constructs and characterization. A. Tet-on WT and DN
Ad-mCFP-Rab27a vectors were generated by subcloning the mCFP-Rab27a inserts into the in-fusion
cloning site of linearized pAdenoX-Tet-3G vector and packaged in 293a cells. B. Amino acid sequences of
WT Rab27a and DN Rab27aT23N, red indicates the sequence location of domi-nant negative mutation. C.
WT Ad-mCFP-Rab27a was transduced into primary rabbit LGACs D. DN Ad-mCFP-Rab27a was
transduced into primary rabbit LGACs. In C. and D., mCFP was directly imaged with a 458 nm laser.
Transduced cells expressing mCFP-Rab27a (green) are out-lined in white to delineate the multicellular
acinar borders. *, Lumen.
43
Figure 4.4. Ad-mCFP-Rab27a production and plasmid validation. A. WT and DN Ad-mCFP-Rab27a
were generated by infusion cloning of the respective mCFP-Rab27a inserts into a Tet-on Ad vector, which
was further packaged into 293a helper cells for virus production. B. Prior to 293a cell transduction, the WT
Ad-mCFP-Rab27a construct was verified by diagnostic enzyme digestion insert sequencing. L#1-2 are
loading ladders. The DN Ad-mCFP-Rab27a construct was similarly verified C. WT mCFP-Rab27a protein
expression was characterized by Western blotting and immunofluorescence microscopy in primary cultured
mouse LGAC from C57BL/6 mouse LG. Both anti-mCFP and anti-Rab27a antibodies labeled a CFP-tagged
Rab27a of ~54kDa. c) By immunofluorescence microscopy, CFP labeled by anti-mCFP antibody is shown
in green, actin cytoskeleton labeled by rhodamine phalloidin is shown in red, and nuclei labeled by DAPI
is in blue.
44
Figure 4.5. Estimated transduction efficiency for Ad-mCFP-Rab27a transduction in rabbit LGACs.
Based on the 3D structures of acini clusters, we estimated transduction efficiency through phase imaging
of acinar clusters and evaluating transduction efficiency of individual cells in the cluster. A. Phase imaging
(20X) of primary rabbit lacrimal gland acinar cells transduced with wild-type and dominant-negative Ad-
mCFP-Rab27a (MOI of 4~6). 60~80% of the acinar clusters were transduced. B. Representative image of
transduction efficiency quantification within each acinar cluster. Cell membranes were labeled with
CellMask plasma membrane stain (red). The acinar borders are outlined by a solid white line, while
individual cells within the acinar cluster are outlined with dashed white lines. Arrowheads show the cells
with fluorescence signal. *, Lumen.
45
4.4 CCh-Stimulated Rab27a Vesicle Fusion Detected in Primary Cultured
Rabbit LGACs
The cholinergic-muscarinic receptor agonist, CCh, stimulates protein and fluid secretion in
the LG. To observe the response of Rab27a-enriched vesicles during acinar stimulation using live
cell imaging, we transduced primary cultured rabbit LGAC with WT mCFP-Rab27a. Time series
live-cell imaging over 15 min was conducted before and after CCh stimulation. Within 4 min of
CCh addition, mCFP-Rab27a-enriched vesicles appeared to undergo a homotypic fusion process
and then to fuse with the APM, demonstrated by a transiently increased vesicle size followed by
reduced vesicle numbers Figure 4.6A. The size of mCFP-Rab27a-enriched vesicles was increased
upon stimulation, peaked by 12 min of CCh exposure and was then decreased by 15 min Figure
4.6B. The number of mCFP-Rab27a-enriched vesicles was decreased with CCh addition, with a
significant reduction after 8 min of stimulation Figure 4.6C.
46
Figure 4.6. CCh stimulation induces mCFP-Rab27a-enriched vesicle homotypic fusion and vesicle
depletion in in primary cultured rabbit LGAC. A. WT Ad-mCFP-Rab27a transduced rabbit LGAC were
imaged in time-series experiments before and after CCh stimulation (100 µM). Transduced cells expressing
mCFP-Rab27a (green) are outlined in white to delineate cell borders. B. Rab27a-enriched vesicle size was
gradually increased upon CCh stimulation, peaking at 12 min (p = 0.0182) and then decreasing by 15 min.
C. Rab27a-enriched vesicle number was concurrently and significantly decreased by 8 min after stimulation
(p = 0.0011), continuing to decrease at 12 and 15 min, p ≤ 0.0001. No significant changes in these
parameters were observed in the resting stage. N = 9, N = cell preparation. Data are presented as mean ±
SEM. *, Lumen.
47
4.5 Expression of DN Rab27a Significantly Reduces CTSS and β-Hex Secretion
in Rabbit LGACs
Based on their extensive co-enrichment in vesicles in NOD mouse LG (Figure 4.1), we
further probed whether Rab27a might be a principal regulator of CTSS secretion. Analysis of CCh-
stimulated secretion conducted in non-transduced rabbit LGAC shows that CCh significantly
induces the secretion of CTSS, β-hex, and total protein into culture medium (Figure 4.7). To test
the functional effect of Rab27a, we conducted similar secretion assays in primary cultured rabbit
LGAC transduced with WT versus DN mCFP-Rab27a. Rab27a activity, like other Rab proteins,
requires the slow hydrolysis of bound GTP to GDP; GTP-bound Rab27a binds to membranes and
is considered the activated form: WT Rab27a cycles naturally between GTP and GDP bound states
associated with membrane binding and facilitation of trafficking. DN Rab27a
T23N
sequesters GDP
and is trapped in the “inactive state”: its overexpression reduces GTP-bound Rab27a and thus
impairs Rab27a activity
89,91
. The LGACs were transduced with WT and DN mCFP-Rab27a at
~40-50% efficiency, demonstrated in Figure 4.8D and Figure 4.5. In cells transduced with WT
mCFP-Rab27a, as plotted in Figure 4.8A, B, CTSS and β-hex activity recovered in the cell
medium was increased by 3.5-fold and 5-fold upon CCh stimulation relative to their unstimulated
controls, respectively. In DN mCFP-Rab27a transduced LGAC, CCh-induced CTSS and β-hex
secretion into culture medium was significantly reduced by 0.5-fold and 0.3-fold, compared to
secretion from LGAC transduced with the WT. The total protein secreted from DN mCFP-Rab27a
transduced cells showed a 20% increase, compared to WT Rab27a transduced cells, as shown in
Figure 4.8C. β-hex is thought to be secreted through the major regulated secretory pathway
enriched in Rab3D. The reduction of β-hex secretion in LGACs expressing DN mCFP-Rab27a
suggests that Rab27a may participate in this pathway, consistent with its co-enrichment with
48
Rab27b and Rab3D on a subset of apical SV
19
. At the same time, DN mCFP-Rab27a expression
more profoundly decreased CTSS activity secreted into culture medium, consistent with our
hypothesized role of this protein in CTSS secretion. Increased total protein secreted during Rab27a
inhibition may, thus, indicate an upregulation of an alternative secretory pathway.
Figure 4.7. CCh stimulation induces CTSS, β-hex and total protein secretion from primary cultured
rabbit LGACs. CTSS, β-hex and total protein in cell culture medium without and with CCh (100 µM)
were normalized to cell protein in each well. A. CTSS activity in culture medium was increased by 5-fold,
with p=0.0041; B. β-hex activity in culture medium was increased by 10-fold, p≤ 0.0001 and C. total protein
in culture medium was increased by 6-fold, p≤ 0.0001, compared to the unstimulated controls. N=5, N: cell
preparation. 2~3 replicate measurements per preparation, represented as points on graph, mean ± SD.
49
Figure 4.8. DN Ad-mCFP-Rab27a transduction significantly reduces CTSS and β-hex secretion in
primary cultured rabbit LGACs. Secretion assays were conducted in primary rabbit LGACs transduced
with adenovirus encoding either mCFP-tagged WT or DN Rab27a Ad at a MOI of 4–6. CTSS and β-hex
activity, as well as total protein in cell culture medium were normalized to protein in cell lysate. A. CTSS
activity secreted into culture medium with CCh stimulation (100 µM) was significantly reduced by half in
cells transduced with the DN mCFP-Rab27a compared to WT Rab27a. *, p = 0.0120 B. β-hex activity in
culture medium was also significantly reduced by expression of DN mCFP-Rab27a relative to WT mCFP-
Rab27a.**, p = 0.0020. C. Total protein secreted in response to CCh was significantly increased with DN
mCFP-Rab27a relative to WT Rab27a expression. *, p = 0.0271. D. LGAC transduction was ~40–50%
efficiency, as shown by the representative image for cells transduced with WT mCFP-Rab27a, detailed in
Figure S2. N = 3, N= cell preparation. Two to three replicate measurements per preparation, represented as
points on the graph, mean ± SD.
4.6 Endolysosomal Markers, together with Rab27a, Are Redistributed to the
APM with Topical CCh Stimulation in Male NOD Mouse LGAC
Rab27a is implicated in secretion of late endosomes/MVB and lysosomes at the plasma
membrane in many cells. To explore the possibility that CCh might trigger direct secretion of
Rab27a-enriched endolysosomal membranes at the APM, Z-projection images of Rab27a and the
endolysosomal markers, Rab7 and Lamp2, were acquired in BALB/c and NOD LG sections, with
and without topical CCh (Figure 4.9). Lamp2 is localized primarily on lysosomes and late
50
endosomes
92
, while Rab7 is mainly enriched on late endosomes/MVB
93
. Upon topical CCh
stimulation, Lamp2, Rab7, and Rab27a in both BALB/c and NOD mouse LGAC were
redistributed to the APM region, denoted by the enrichment of the subapical actin meshwork
immediately underneath the APM, and as demonstrated by increased colocalization with the apical
actin (Figure 4.9B-D). In unstimulated LG, a significantly increased apical accumulation of
Lamp2 was also detected in male NOD mouse LGAC compared to male BALB/c mouse LG. With
stimulation, enhanced colocalization of Lamp2 with the apical actin meshwork was observed in
NOD LGAC relative to BALB/c LGAC. Similarly, Rab7 was colocalized with the apical actin
meshwork with stimulation to a greater extent in NOD mouse LGAC relative to BALB/c mouse
LGAC. Rab27a recovery with the apical actin meshwork at the apical lumen was equally enhanced
by topical CCh stimulation in both BALB/c and NOD mouse LG. These findings suggest that CCh
stimulation may trigger endolysosomal trafficking to the apical membrane area in LGAC, with
NOD mouse LGAC demonstrating a more substantial redistribution than BALB/c mouse LGAC.
51
Figure 4.9. Rab27a and endolysosomal membrane marker colocalization with the apical actin
meshwork. A. Enrichment of the endolysosomal markers, Lamp2 and Rab7, as well as Rab27a were
evaluated by indirect immunofluorescence in untreated and CCh-stimulated LG from male BALB/c and
NOD mice. Z-projection images are shown for Lamp2, Rab7 and Rab27a (green) and the subapical actin
meshwork, enriched immediately beneath the APM (red). Colocalization with the apical actin is
demonstrated by arrows. B. Lamp2 showed increased accumulation with apical actin in untreated LG of
male NOD mouse, compared to LG from BALB/c mouse, with ***, p = 0.0003. In stimulated LG, Lamp2
colocalization with the apical actin meshwork increased by 2-fold in LG from male NOD, compared to LG
from male BALB/c (****, p ≤ 0.0001). Lamp2 colocalization with the apical actin meshwork was increased
in both male BALB/c (p = 0.0792) and NOD (***, p = 0.0007) LG upon stimulation. C. Rab7 colocalization
with the apical actin meshwork was increased in both BALB/c (**, p = 0.0037) and NOD (****, p ≤ 0.0001)
mouse LG with stimulation. Stimulated NOD LG sections also showed a 1-fold increase (****, p ≤ 0.0001)
in colocalization with apical actin, relative to stimulated BALB/c LG sections. D. Rab27a colocalization
with the apical actin meshwork was increased in both BALB/c (***, p = 0.0001) and NOD (**, p = 0.002)
mouse LG upon stimulation. However, no obvious differences were observed between the two strains with
stimulation. N = 3–4, N = mice per group. Sections from each mouse LG were evaluated in three ROI per
sample. All data are presented as mean ± SD. *, Lumen.
52
4.7 Rab27a Distribution on Endolysosomes is Increased in NOD Mice
Previously, we reported that CTSS was localized to both endolysosomes and SVs in LG
19
.
Upon topical CCh stimulation, tear CTSS activity significantly is increased in NOD mice
compared to BALB/c mice. Correlating with the CTSS increased in tears, CTSS protein is depleted
in NOD mouse LGAC, as shown in Figure 4.10, concurrent with detection of some residual CTSS
within lumena in stimulated LG tissue in NOD mouse. Density gradient fractionation of
membranes from mouse LG showed that CTSS was enriched in a high-density lysosomal
membrane fraction along with Rab27a in healthy mouse LG (Figure 4.11). Moreover, Figure 6
demonstrated the apical redistribution of Rab27a concurrent with the apical accumulation of
endolysosomal markers in both male BALB/c and male NOD mouse LG, with a more profound
enrichment in NOD mouse LG. These findings collectively suggest Rab27a may mediate
trafficking of CTSS between endolysosomes and the APM.
53
Figure 4.10. CTSS immunofluorescence is significantly reduced by topical CCh stimulation in NOD
mouse relative to BALB/c mouse LG. CTSS Integrated density was quantified using the CTCF method
87,88
in each acini, with 5 acini quantified per field, 5 fields were obtained from each mouse LG. A. CTSS
immunofluorescence is increased in the subapical area in NOD mouse LG without stimulation. Topical
CCh stimulation depletes the subapical CTSS in the acini, with residual protein detected within the lumen
(arrowheads). CTSS labeling by immunofluorescence is in green, actin cytoskeleton labeling by rhodamine
phalloidin is shown in red, while nuclear labeling by DAPI is shown in blue. B. Quantification of A.
showed that CTSS is significantly increased in unstimulated male NOD mouse LGAC (p≤0.0001),
compared to BALB/c mouse LG. CCh stimulation significantly reduced the CTSS recovered in NOD mouse
LGAC by 60% (p≤0.0001). N=3-5, N: mouse per group. Average integrated density per acini in each field
were presented as points.
54
Figure 4.11. Rab27a is co-enriched with lysosomal fractions containing CTSS. Clarified cell lysate, the
light mitochondrial fraction (LMF) and the lysosomal fraction were prepared from C57BL/6 mouse LGs
using an Optiprep density gradient
94
. Each lane was loaded with an equal amount of protein (35 μg/lane).
Rab27a , detected by Western blotting, was enriched in the lysosomal fraction, compared to whole cell
lysate or LMF. An accumulation of CTSS was also detected in the Rab27a-enriched lysosomal fraction.
55
Figure 4.12. Vesicular Rab27a is dispersed into the cytoplasm in CCh-stimulated LG. Z-stack images
of resting and CCh-stimulated LGACs show that, in resting acini, Rab27a is localized to vesicles enriched
in the subapical area from both BALB/c and NOD mice, but NOD mouse LGAC exhibit more accumulation
near the lumen (asterisk). After CCh stimulation, Rab27a is dispersed into the cytoplasm as well as recruited
to the APM in both BALB/c and NOD.
To further elucidate this relationship, we labeled resting and topically-stimulated LG from
BALB/c and NOD mice with antibodies to Rab27a and Lamp2 (Figure 4.13A). In BALB/c mouse
LG, Rab27a was mainly subapically enriched while Lamp2 had a more basolateral distribution.
However, an apical accumulation of large Lamp2-enriched vesicles was observed in NOD mouse
LG, illustrated in the magnified images of unstimulated NOD mouse LGAC within LG sections
56
(Figure 4.13), and concurrent with increased colocalization of Rab27a and Lamp2 in unstimulated
mouse NOD LG (Figure 4.13C). With CCh stimulation, a significant reduction in Rab27a and
Lamp2 colocalization was observed in NOD mouse LG. This may result from reduced Rab27a
disassociation from membranes upon stimulation of secretion and parallels the loss in mCFP-
Rab27a enriched vesicles with CCh stimulation in vitro (Figure 4.6) and the reduction in
integrated density of Rab27a fluorescence/NOD mouse acini, shown in Figure 4.13B. Z-stack
images in Figure 4.9 and Figure 4.12 demonstrate that, with stimulation, vesicular Rab27a is both
dispersed into the cytoplasm, as well as recruited to the APM region. Decreased association with
membranes concurrent with the activation of trafficking pathways has been reported for secretory
Rab proteins, consistent with hydrolysis of bound GTP. In Rab27a-mediated exocytosis, after
plasma membrane fusion and content release, the GDP dissociation inhibitor is known to
translocate GDP-Rab27 from the membrane to the cytosol
89
. These findings suggest a recruitment
process occurring simultaneously with exocytosis of Rab27a-enriched endolysosomes upon CCh
stimulation. Lamp2, as an integral membrane protein, is present with apical membrane and in
retrieved SV membrane following exocytosis, thus remaining enriched at the apical region.
57
Figure 4.13. Rab27a enrichment with endolysosomes is increased in LG from male NOD mice. A.
Analysis of immunofluorescence by confocal fluorescence microcopy showed that the endolysosomal
marker, Lamp2, was partially colocalized with Rab27a in unstimulated acini. B. Upon CCh stimulation, the
integrated density of Rab27a within each acinar cell was decreased, with this reduction shown to be
significant in NOD mouse LG (*, p = 0.0290; **, p = 0.0040). Integrated density in three acini were
quantifed and average in each field, as described previously
87
. 4 fields were obtained in each mouse section,
presented as points. N = 4, N = mouse per group. C. Rab27a colocalization with Lamp2 was significantly
increased in unstimulated NOD mouse LG (****, p ≤ 0.0001) relative to BALB/c mouse LG, while CCh
stimulation significantly reduced this colocalization in NOD mouse LG (**, p = 0.0032). D. Magnified
images of untreated NOD mouse LGAC demonstrated an accumulation of large Lamp2-enriched vesicles
in the apical region, colocalized with Rab27a (White arrows). Colocalization was quantified in three acini
per field, four fields were aquired from each mouse section. N = 4, N: mouse per group. *, Lumen.
58
CHAPTER 5
α1-Adrenergic agonist phenylephrine increases tear CTSS secretion
5.1 Phenylephrine induces less fluid secretion in C57 mouse lacrimal glands
A lacrimal gland functional unit includes afferent nerves on the ocular surface, efferent nerves
innervating the lacrimal gland, and LGAC and ductal cells that secrete lacrimal fluid. Activation
of both parasympathetic and sympathetic nerves release neurotransmitters that controls lacrimal
fluid secretion. Despite reported involvement of both α1 and β-adrenergic pathways in sympathetic
nerve stimulated protein secretion, α1-adrenergic is the predominant functional pathway activated
in murine and rat LG. In our preliminary study (Figure 5.2) where we in situ topically stimulated
C57 mouse LG with α1-adrenergic agonist phenylephrine (PE), β-adrenergic agonist isoproterenol
(IPR) and muscarinic cholinergic agonist carbochol (CCh), we also demonstrated that IPR had
minimum effect stimulating protein and fluid secretion in mouse LG. Therefore, in this study we
mainly focused on the secretion effect of α1-adrenergic agonist, PE. Although PE have been
associated with lacrimal fluid secretion, the molecular mechanism of protein and fluid secretion
induced by PE is woefully understood. To elucidate the secretion ability of PE, we compared tear
protein and volume after in situ topically stimulating C57 LG for 15 minutes with CCh and PE at
effective dose. Effective dose of PE was selected by topically stimulating C57 LG with doses
ranging from 10 μM, 50 μM, 100 μM and 200 μM, and selecting the dose with maximum tear
protein and fluid secretion, shown in Figure 5.3. The effective dose of CCh was chosen in a similar
manner as previously reported. From topical LG stimulation we found that α1-adrenergic agonist
PE treatment significantly increased tear concentration compared to muscarinic cholinergic
agonist CCh. This is likely caused the fact that both treatments stimulate a similar amount of tear
59
proteins, while PE resulted in considerably lower fluid secretion (Figure 5.1.). Fluid and
electrolytes secretion in the lacrimal gland are commonly associated with active pumping of ions
across the plasma membrane, followed by passive osmotically driven water flow. CCh has been
reported to activate apical K
+95
, Cl
-
channels
96
and basolateral Na
+
-H
+
exchangers
97
, which
accounts for fluid secretion
5
. Since PE and CCh is reported have distinct downstream pathways,
the ability of PE to trigger fluid secretion related ion channels may be different.
60
Figure 5.1. In vivo topical stimulation of LG α1-adrenergic receptor with phenylephrine evokes lesser
tear fluid secretion and more concentrated tears relative to topical carbachol. C57BL/6J mice LG
were topically stimulated for 3 times, 5 minutes each, with 3uL 50 μM of carbachol (CCh) or 100 μM of
phenylephrine (PE) as shown in A. PE stimulation led to significantly reduced tear fluid secretion (shown
as tear volume) associated with in increased tear protein concentration, as demonstrated in B and C. ****,
p ≤ 0.0001, N=15, N, mouse count. Data presented as mean ± SD.
61
Figure 5.2. Carbachol (CCh) and phenylephrine (PE) induces both protein and fluid secretion in
tears in in vivo topical stimulation. C57 LG was stimulated for 3 times, 5 minutes each, with 3μL 50 μM
of CCh and 100 μM of PE and 100 μM of isoproterenol. A. CCh stimulated the highest tear fluid, followed
by PE, while isoproterenol (IPR) showed limited effect in stimulating tear secretion. ****, p ≤ 0.0001; ***,
p=0.0002 B. Similar to tear fluid, CCh also had the highest amount of tear protein recovered, followed by
PE, with IPR showing little effect in vivo. ****, p ≤ 0.0001; **, p=0.0097. N=5, N: mouse. Data presented
as mean ± SD.
62
Figure 5.3. In vivo tear secretion in C57 LG stimulated by phenylephrine (PE) at different doses. C57
LG was stimulated for 3 times, 5 minutes each, with 3μL CCh (50 μM) and PE at 10 μM, 50 μM, 100 μM
and 200 μM. A. Stimulation of PE obtained highest tear volume at 100 μM dose stimulation. B. Tear protein
recovered also peaked with 100 μM PE stimulation. N=5, N: mouse. Data presented as mean ± SD.
63
5.2 Phenylephrine significantly increased CTSS secretion in C57 mouse tears
Increased tear CTSS activity is associated with pathogenesis and progression of SS dry eye
disease. Elevated CTSS in tears disturbs the homeostasis of the ocular surface by upregulating
inflammatory cytokines, triggering apoptosis and autophagy in cornea epithelium, provoking
neuropathy pain and recruiting lymphocytes. It has been previously reported that in SS dry eye
aging model, the release process of neurotransmitters is impaired, as well as deficient downstream
signalling pathways of cholinergic and α1D-adrenergic agonist
98
. To elucidate the possible
relationship between CTSS secretion and LG efferent nerve stimulation, we topically stimulated
C57 LGs with CCh and PE at selected effective dose and analyzed CTSS and β-hex activity in the
stimulated tears. PE stimulation significantly elevated tear CTSS activity by 1-fold compared to
CCh, while no obvious difference in tear β-hex activity was observed between the two treatments,
demonstrated in Figure 5.4A,B. We further analyzed CTSS protein abundancy in tears stimulated
by CCh and PE through western blot. Similar to activity, CTSS protein level in tears of equal total
protein significantly increase in that stimulated by PE compared to CCh (Figure 5.4C). This
further substantiated that enhanced CTSS activity induced by PE is likely caused by increased
CTSS secretion into the tears.
64
Figure 5.4. LG stimulation with topical PE increases CTSS activity and CTSS protein content of tears
of C57 mice. C57BL/6J (C57) mice LG were topically stimulated for 3 times, 5 minutes each, with 3μL 50
μM of carbachol (CCh) and 100 μM of phenylephrine (PE). A. shows tear CTSS activity normalized to
total tear protein when tear production was stimulated with CCh or PE as above. CTSS activity/protein for
CCh-treated samples was considered as 100. ***, p ≤ 0.0009 B. No obvious change was observed in β-
hex activity normalized to total tear protein when evoked by CCh versus PE. ns, p=0.1010 C. shows CTSS
protein/total tear protein when tear production was stimulated with CCh or PE as above. *, p=0.0319. CTSS
protein/total protein for CCh-treated samples was considered as 100%. N=15, N, mouse count,
demonstrated as points. Data presented as mean ± SD.
65
5.3 Phenylephrine significantly upregulated CTSS secretion in primary mouse
LGACs
The LG is composed mainly of acinar cells, myoepithelial cells, ductal cells, and lymphocytes.
Lacrimal gland acinar cells (LGACs), consisting 80% of the gland, are largely responsible for
protein secretion in the LG. To test the direct secretion effect of CCh and PE in LGACs, we
conducted in vitro secretion assay in primary LGACs isolated from C57 mouse. After stimulating
the cells with CCh and PE for 30 minutes at effective in vitro dose, cell medium was collected and
analyzed for CTSS and β-hex activity, as well as total protein abundancy. CTSS activity in cell
media significantly increased by 50% with PE treatment, while only showing an increasing trend
with CCh stimulation, shown in Figure 5.5A. Both CCh and PE treatment significantly increased
media β-hex activity by 100% and 50% in Figure 5.5B. Similar to β-hex, total protein released
into the cell media had a significant 100% increased with CCh and a relatively lower 50% increase
with PE in Figure 5.5C. To substantiate that primary LGACs still preserved the same morphology
and secretory functions, we conducted immunofluorescence of CTSS and trafficking markers on
unstimulated LGACs, shown in Figure 5.5D. Comparable with the observation in LG tissue
samples, CTSS and endo-lysosomal marker Lamp2 are mainly distributed as puncta across the
cytoplasm. Early endosomal marker EEA1 is mainly subapically distributed with some localized
on the basolateral membrane. Mature SV marker Rab3D is localized around the apical lumen
demonstrating distinct large vesicles. Rab27a and Rab27b, associated the trafficking and released
of SV, is elongated throughout the cytoplasm exhibiting trafficking network like structure. β-hex
is a lysosomal enzyme released through major regulated exocytosis pathway that involves
exocytosis of Rab3D-enriched mature secretory vesicles. In previous studies, we have indicated
that CTSS, different from β-hex, maybe secreted through an alternative endo-lysosomal
66
pathway
1,19
. The differential secretion of β-hex and CTSS stimulated by CCh and PE in vivo
(Figure 5.4) and in vitro (Figure 5.5) not only supported the alternative secretion of CTSS, but
also suggested that different efferent nerve system can activate distinct trafficking pathways.
67
Figure 5.5. Phenylephrine stimulation specifically induces CTSS activity in mouse primary lacrimal
gland acinar cell culture. Primary LGACs were prepared from 10-15 mice and stimulated with or without
CCh (100 μM) and PE (100 μM) for 30 min. Cell media was collected and analyzed for CTSS activity, β-
hex activity and total protein. A. CTSS activity was significantly upregulated by PE stimulation by 0.5-
fold. *, p=0.0363. B. β-hex activity was induced by both CCh and PE by 1-fold and 0.5-fold, while CCh
demonstrated a higher response. ****,p ≤ 0.0001;**,p=0.0014. C. Total protein secretion was also
significantly upregulated by both CCh and PE by 1-fold and 0.5-fold. ***, p=0.0002; *, p=0.0151. N=9, N,
preparation. Triplicate samples were used for secretion with each preparation, samples are represented as
points. Data presented as mean ± SD D. Indirect immunofluorescence microscopy of CTSS and trafficking
markers, including Lamp2, EEA1, Rab3D, Rab27A and Rab27B in unstimulated primary mouse LGACs
show the cells maintained the same morphology and trafficking characteristic as in tissue.
68
CHAPTER 6
α1-Adrenergic agonist phenylephrine induces alternative trafficking
in mouse LGACs
6.1 Phenylephrine induces less Rab3D vesicle fusion and triggers limited
calcium response in mouse LGACs
Intracellular [Ca
2+
] plays a crucial role in protein secretion through modulating vesicle-vesicle
fusion and SV-PM fusion. Major regulated granule secretion involves a Rab3D-dependent
maturation process of SV and a Ca
2+
-dependent exocytosis process, shown in Figure 6.1A.
Rab3D-enriched mature SVs are distributed on the subapical area, withheld upon the dense actin
filaments adjacent to the PM. When stimulated, increased intracellular [Ca
2+
] trigger vesicle fusion,
leading to enlarged SV. Cholinergic and α1-adrenergic agonists have been reported to increase
intracellular [Ca
2+
] at different levels. To investigate the ability of CCh and PE in inducing major
regulated secretion, we measured intracellular [Ca
2+
] through calcium imaging with Fluo-4 AM in
primary C57 mouse LGACs stimulated with CCh and PE, and analyzed the change of Rab3D
vesicle size in C57 mouse LG tissue topically stimulated with CCh and PE. Fluorescent intensity
of cells was recorded for 10 minutes upon stimulation. CCh stimulation significantly increased
increased intracellular [Ca
2+
], demonstrated by increased fluorescence intensity, starting from 2
minutes and peaking at 4 minutes. PE resulting in less increase in intracellular [Ca
2+
], with slight
increase in fluorescent intensity at 4-minute time point, shown in Figure 6.1B. The change of
intracellular [Ca
2+
] was quantified by relative change of fluorescent intensity, (F-F0)/F0, as
previously reported, shown in Figure 6.1C. The accumulative [Ca
2+
] increase in Figure 6.1D
showed that CCh-increased intracellular [Ca
2+
] is 3-fold higher than that of PE. Corresponding
69
with significantly elevated intracellular [Ca
2+
], increased Rab3D-enriched fusion was observed in
the subapical area of LGACs stimulated with CCh, shown in Figure 6.1E. Increased Rab3D
vesicle fusion was further demonstrated by increased vesicle size in Figure 6.1F, where CCh
significantly increased Rab3D-enriched vesicles by 0.5-fold, while PE only had a slight 0.2-fold
increase. PE’s limited ability in triggering intracellular calcium and inducing Rab3D vesicle fusion
may suggest it have minimum effect in stimulating the major regulated secretory pathway.
70
Figure 6.1. Phenylephrine increased intracellular calcium by small amount, reflected by minimum
increase of Rab3D size. A. Proposed mechanism shows that CCh and PE have different abilities in
upregulating intracellular [Ca
2+
], which induces Rab3D-enriched SV-SV fusion and SV-APM fusion. B.
Fluo-4 AM calcium imaging were carried out in primary LGACs stimulated with CCh (100 μM) and PE
(100 μM), shown in D.. Fluorescent intensity of the cells was recorded for 10 minutes upon stimulation.
The change of fluorescent intensity was quantified by relative change of fluorescent intensity. CCh
significantly increased fluorescent intensity from 2 minutes and peaked at 4 minutes, while PE showed a
slight increase in fluorescent intensity at 4-minutes. Calcium signal response data presented as mean ±
SEM. C. The accumulative intracellular [Ca
2+
] were analyzed through calculation of area under the curve
71
of calcium response in (b). CCh stimulation induced a significantly higher intracellular [Ca
2+
]. ***,
p=0.0006. N=9, N, preparation, presented as points on the graph. Data presented as mean ± SD. E. Confocal
microscopy of Rab3D vesicles in resting and stimulated C57 LGACs showed CCh stimulation significantly
increased Rab3D vesicle sizes near the APM, while PE only exhibited a slight increase in vesicle size,
demonstrated by arrowheads and quantified in F. 30-40 vesicles were quantified in each acini, 3 fields were
acquired from each mouse section. N=4, N: mouse. ****, p ≤ 0.0001. Data presented as mean ± SD.
6.2 Carbachol induces Rab3D colocalization with endo-lysosomal markers at
the subapical area of mouse LGACs
We have previously demonstrated that CTSS is localized in both the lysosomes and SVs in
LGACs
1,19
, where it was first contained in lysosomes upon synthesis and sorting into SVs for long-
term storage
19
. In diseased male NOD mouse, where CTSS activity is significantly elevated in
tears, endo-lysosomal markers accumulated at the subapical actin meshwork underneath the APM
1
.
Asides from direct release of endo-lysosomes at APM
1
, endolysosomal content may be secreted
through an intersection with Rab3D-enriched major secretory pathways. To further elucidate the
relationship between the two secretory pathways under different stimulation conditions, we labeled
resting LG and LG topically stimulated with CCh and PE with antibodies to Rab3D, Lamp1 and
Lamp2. At resting stage, Rab3D is localized on the large SV (~1 µm diameter) clustered at the
subapical area of the acini, while Lamp1 and Lamp2 are mainly sub-basolateral distributed. During
CCh and PE stimulation, Lamp1 and Lamp2 are found to have increased accumulation near the
APM. However, CCh stimulation led to a significant increase in Rab3D colocalized with Lamp1
and Lamp2 at the subapical area, while PE only showed an increasing trend in Lamp2-Rab3D
colocalization, shown in Figure 6.2. The increased colocalization is likely associated with
increased fusion between Rab3D-enriched SVs and endo-lysosomes, driven by the drastic increase
in intracellular [Ca
2+
]
99
. In Figure 6.3, we also found that CCh significantly reduced the integrated
density of Rab3D, where vesicular Rab3D appear to be more dispersed into the cytosol. Rab3D,
72
like other Rab proteins, undergoes a GDP-GTP cycle that controls membrane association and
dissociation. During SV maturation, Rab3D is recruited to the vesicle membrane, where its further
actives into a GTP-bound form and interacts with its specific receptors. After membrane fusion
and content release, a slow hydrolysis converts the bound GTP into GDP, which was
subsequentially translocated from the membrane back to the cytosol. The increased dissociation
of Rab3D indicated a possible upregulation of Rab3D related exocytosis. These findings suggest
that CCh and PE have different abilities in initiating endolysosomal content sorting into Rab3D-
enriched vesicles and Rab3D-mediated exocytosis.
Figure 6.2. Carbachol significantly increased Rab3D colocalization with endo-lysosomal markers at
the subapical area. A. Analysis of immunofluorescence by confocal fluorescence microscopy in resting
and stimulated C57 LGACs show that Rab3D had little colocalization with endolysosomal marks Lamp1
and Lamp2 in unstimulated acini. Upon CCh stimulation, a significant increase in Rab3D and Lamp1/2
colocalization was observed at the subapical area of the acini, demonstrated by white arrowheads and
quantified in B. and C. No obvious increase was observed in acini stimulated with PE. Lamp1, Con vs
CCh:*, p-0.0187; CCh vs. PE: *, p=0.0463. Lamp2, Con vs CCh: *, p=0.0249. Colocalization was
quantified in 1-2 acini per field, 3 fields were acquired from each mouse section. N=4, N: mouse. Data
presented as mean ± SD. *, Lumen.
73
Figure 6.3. Rab3D integrated density reduced with Carbachol (CCh) stimulation in C57 LGAC.
Indirect immunofluorescence of Rab3D were carried out in resting and stimulated C57 LGACs, where Z-
stack images were obtained and quantified with corrected total cell fluorescence method. A. Rab3D is
enriched on membranes of mature SVs localized at the subapical area near the APM in resting acini. With
stimulation, Rab3D is recruited back to the cytosol, shown as dispersed signal and reduced integrated
intensity. B. CCh stimulation significantly reduced integrated intensity of Rab3D in LGACs, while PE
showed a reducing trend. *, p=0.0235 Images containing 1-2 acini was quantified in per field, 3 fields were
acquired from each mouse section. N=3~4, N: mouse. Data presented as mean ± SD.
6.3 Dysfunctional Rab3D enhanced α1-adrenergic sensitivity and upregulates
CTSS secretion
In our previous studies we reported that loss of Rab3D and disproportionate Rab27-to-Rab3D
activity contributes to enhanced CTSS release
19
. In Figure 6.1 and Figure 6.3, we concluded that
PE have limited affect activating Rab3D-enriched major-regulated secretion. To further investigate
the relationship between Rab3D and CTSS release under α1-adrenergic stimulation, we compared
the CTSS activities in tears stimulated with CCh and PE in both C57 and mouse deficient of Rab3D
74
(Rab3DKO). LGs of C57 and Rab3DKO was topically stimulated with PE doses ranging from 10
μM, 50 μM, 100 μM and 200 μM. Similar to previously reported, Rab3DKO had a significantly
higher tear CTSS with CCh stimulation at 50 μM dose, shown in Figure 6.4A. Interestingly,
Rab3DKO demonstrated different sensitivity to PE compared to C57. While C57 exhibited highest
tear CTSS activity and tear concentration with 100 μM PE stimulation, while Rab3DKO exhibited
highest response at a lower dose of 50 μM, demonstrated in Figure 6.4A and B. We further
compared CTSS abundancy through measuring corrected total cell fluorescence in the LGACs of
C57 and Rab3DKO after topically stimulating at a same dose of 100 μM PE. In C57 LGACs,
CTSS is distributed as small puncta across the cytoplasm at resting stage, as previously
demonstrated. Upon CCh and PE stimulation, CTSS mobilizes to both APM and sub-basolateral
area of the acinar cell, shown in magnified images of Figure 6.4C. Compared to CCh, PE
stimulation significantly reduced CTSS abundancy in the acini in Figure 6.4D, indicating a
degranulation process, concurring increased CTSS secretion in tears. In Rab3DKO acini, CTSS is
sprawled across the cytoplasm at resting stage. CCh stimulation mobilized CTSS to the subapical
area of the acini, while PE significantly depleted CTSS, demonstrated in Figure 6.4E. CCh is
suspected to cause the recruitment of CTSS to subapical secretory stores through upregulated
intracellular [Ca
2+
]. Deficient Rab3D impeded the maturation of SVs, which further interfered with
the release of mature SVs, leading to the accumulation of CTSS at the subapical area. PE is
suspected to stimulate an alternative pathway independent of Rab3D, which resulted in notable
reduction of CTSS protein.
75
Figure 6.4. Dysfunctional Rab3D CTSS enhanced α1-adrenergic sensitivity and upregulates CTSS
secretion. C57 and Rab3DKO LGs were topically stimulated for 3 times, 5 minutes each, with 3μL 50 μM
of CCh and PE at 10 μM, 50 μM, 100 μM and 200 μM, CTSS in tears and LG were analyzed. A. Tear CTSS
activity exhibited a higher increase with PE stimulation from 50 μM of dose, compared to CCh. CTSS
activity stimulated by PE peaked at 100 μM in C57 mice, which only peaked with 50 μM in Rab3DKO. *,
p=0.0115. B. Tear concentration exhibited a similar effect, reaching highest concentration with 100 μM PE
in C57 mice, while Rab3DKO exhibited a significantly higher response at 50 μM. ****, p ≤ 0.0001; *,
p=0.0141. N=4, N: mouse C. When comparing CTSS protein in tissue after topical stimulation with CCh
(50 μM) and PE at (100 μM), CTSS mobilized to both the APM and the sub-basoteral area of the acinar
cell with both CCh and PE stimulation. D. Change of CTSS abundancy, quantified by corrected total cell
fluorescence, showed that PE significantly depleted CTSS in C57 LGACs. *, p=0.0474; ****, p ≤ 0.0001
(e) In Rab3DKO acini, PE also significantly reduced protein abundancy in the acinar cells. **, p=0.0054
Data presented as mean ± SD. *, Lumen.
76
6.4 Phenylephrine induces apical redistribution of endo-lysosomal markers in
both wild-type and Rab3D-deficient LGACs
We have previously shown that CCh stimulation can trigger direct endolysosomal traffick to
the APM of LGACs in BALB/c and NOD mouse
1
. To explore the possibilities of CCh and PE in
initiating direct secretion of endolysosomal vesicles, we acquired Z-projection images of
endolysosomal markers, Lamp1, Lamp2 and Rab7 in LGACs stimulated with CCh and PE. Lamp1
and Lamp2 is mainly localized on late endosomes and lysosomes, while Rab7 is enriched on late
endosome/ MVBs. As previous
1
, we quantified the enrichment of endolysosomal markers at the
apical actin meshwork immediately underneath the APM in LGACs with different stimulation
conditions. Shown in Figure 6.5A, both CCh and PE stimulation resulted in a redistribution of
Lamp1, Lamp2 and Rab7 to the APM in both C57 and Rab3DKO LGACs, demonstrated by
increased colocalization with apical actin in Figure 6.5B-D. PE stimulation exhibited a
significantly higher increase in apical accumulation of Lamp 1 and Lamp2, compared to CCh, in
LGACs of both C57 and Rab3DKOs. However, CCh and PE demonstrated a comparable ability
in recruiting Rab7-enriched large vesicles to the APM, shown in Figure 6.5A and D. When
compared between strains, Rab3DKO induced a 30% increase in Rab7 apical redistribution in the
LGACs stimulated with CCh. The different responses of Lamp1 and Lamp2 with Rab7 may
suggest distinct trafficking pathways of lysosomes and MVBs. In all, these data demonstrate that
PE stimulation is able to traffic endolysosomal markers to the APM at a higher extent.
77
Figure 6.5. Phenylephrine induces apical redistribution of endo-lysosomal markers in both wild-type
and Rab3D-deficient LGACs. A. Enrichment of endolysosomal markers, Lamp1, Lamp2 and Rab7 were
evaluated by indirect immunofluorescence in resting LG and CCh, PE stimulated LG from C57 and
Rab3DKO mouse. Sum projection of Z stack images showed that CCh and PE induced an increased
accumulation of Lamp1, Lamp2 and Rab7(green) at the subapical mesh network, immediately beneath the
APM (red). Colocalization with the apical actin is demonstrated by white arrowheads. B. Lamp1 showed
an increased accumulation with the apical actin in LGACs stimulated by CCh and PE. In stimulated
LGACs, colocalization of Lamp1 and apical actin increased by 2-fold with PE stimulation in C57 and
Rab3DKO, compared to CCh. ****, p ≤ 0.0001;***, p=0.0002; **, p=0.0021;*,p=0.0322 C. The
colocalization of Lamp2 and apical actin significantly increased by 1-fold with PE stimulation in C57 and
Rab3DKO, compared to CCh. ****, p ≤ 0.0001; ***,p=0.0002; **,p=0.0056 D. Both CCh and PE
significantly increased colocalization of Rab7 with apical actin with no obvious different observed between
the two treatments. When compared between strains, under CCh stimulation Rab3DKO exhibited higher
levels of Rab7 apical redistribution. ****, p ≤ 0.0001.*, p=0.022 Black asterisk represent p values
compared between control and treatments, while blue asterisk represents comparisons between treatments,
red asterisks represent comparison between strains. Colocalization was quantified in 1-2 acini per field, 3
fields were acquired from each mouse section. N=3~4, N: mouse. Data presented as mean ± SD. *, Lumen.
78
6.5 Phenylephrine induces apical redistribution of Vamp2 in both wild-type
and Rab3D-deficient LGACs
Vesicle-associated membrane proteins (VAMPs) are a family of Soluble N-ethylmaleimide-
sensitive fusion protein-attachment protein receptor (SNARE) proteins that participates in the
exocytosis process through modulating vesicle docking and fusion. LGAC express several SNARE
proteins including Vamp2
100
. Vamp2 is localized on a subpopulation of SV, known as recruitable
SV (rSV) in the LGAC, which is dispersed throughout the cytosol and recruited to the APM upon
stimulation
13
. Vamp2, together with syntaxin3, have been identified as a marker for minor
regulated pathway in parotid cells, where it relocated to the APM with low levels of CCh (50nM)
and IPR(1μM) stimulation
101
. To explore the possibilities of an alternative release from Vamp2-
enriched rSV with PE stimulation, we compared the accumulation of Vamp2 at apical actin mesh
network in resting LGACs and those stimulated with CCh and PE. Both CCh and PE triggered
apical relocation of Vamp2 to the APM, shown in Figure 6.6. with no significant difference
between the two treatments. When compared between strains, Rab3DKO and C57 exhibited
similar magnitude of change in the apical redistribution of Vamp2, with Rab3DKO demonstrating
a slight reduction under CCh stimulation. These findings suggest that same as CCh, PE also
induced Vamp2-enriched SV traffic to the APM. Despite that fact that PE showed similar affect
in recruiting Vamp2 to the APM, PE is suspected to upregulate endo-lysosomal fusion with
Vamp2-enriched SV, demonstrated by increased colocalization of Vamp2 with Lamp2 in Figure
6.7, indicating a possible enrichment of endolysosomal content in Vamp2-enriched SV under PE
stimulation.
79
Figure 6.6. Phenylephrine induces apical redistribution of Vamp2 in both wild-type and Rab3D-
deficient LGACs. A. Indirect immunofluorescence confocal microscopy of resting and stimulated C57 and
Rab3DKO LGACs showed that Vamp2 is redistributed to the apical actin mesh network with CCh and PE
stimulation, labelled by white arrowheads. B. Colocalization of apical actin with Vamp2 showed that both
CCh and PE significantly induced enrichment of Vamp2 to the apical actin mesh network, with no
significant different between treatment. Black asterisk represents p values compared between control and
treatments, ***, p=0.0003; **, p=0.0034; *, p=0.0144. Colocalization was quantified in 1-2 acini per field,
3 fields were acquired from each mouse section. N=3~4, N: mouse. Data presented as mean ± SD. *, Lumen.
80
Figure 6.7. Phenylephrine (PE) significantly increased colocalization of Vamp2 with endo-lysosomal
markers in C57 LGACs. A. Indirect immunofluorescent images of resting and stimulated C57 LGACs
co-labelled with antibodies towards Lamp1 and Vamp2 were acquired, with colocalization quantified. PE
stimulation increased colocalization of Vamp2 and Lamp1 at the sub-basolateral area of the acini, shown
by white arrowheads. B. Colocalization quantification further substantiated that PE significantly increased
the Vamp2 colocalization with Lamp1. Colocalization was quantified in 1-2 acini per field, 3 fields were
acquired from each mouse section. N=4, N: mouse. Data presented as mean ± SD. *, Lumen.
81
CHAPTER 7
PIKFyve inhibition reduces tear CTSS secretion
7.1 Inhibition of PIKfyve reduces CTSS secretion in rabbit LGACs
Regulated secretion can be separated into a the major regulated and a minor endolysosomal
secretion in LGAC, shown in Figure 7.1A. Major regulated secretion, extensively characterized
in the rabbit LGACs, consists of SV maturation, compound fusion and exocytosis. Rab3D is a
major regulator for SV maturation and compound fusion, while intracellular [Ca
2+
] is essential for
the release of SV at the APM. In previous studies, we have shown CTSS is likely trafficked through
a Rab27a regulated endolysosome trafficking pathway in the LGACs. Since PIKfyve is a major
regulator for maturation of endolysosomal vesicles, we hypothesized that it may also contribute to
CTSS secretion in the LGACs. Apilimod is a first-in-class specific inhibitor for PIKfyve. Primary
rabbit LGACs were treated with 300 nM apilimod for 1 hour and stimulated with 100 μM CCh for
30 minutes. In the control group (0.01% DMSO), CCh stimulation induced a 14-fold increase in
CTSS activity (Figure 7.1B) and 6.5-fold increase in β-hex activity (Figure 7.1C). After 1-hour
apilimod treatment the CTSS activity recovered in the media only increased 9-fold, while β-hex
increased 12-fold. The reduced CTSS activity in medium from LGACs treated with apilimod
suggests that PIKfyve inhibition may specifically block the CTSS secretion through hindering
endolysosome trafficking.
82
Figure 7.1 Pharmacological inhibition of PIKfyve reduces CTSS activity recovered in rabbit LGACs.
Rabbit LGACs were treated with 300nM Apilimod for 1 hour and stimulated with 100 μM CCh for 30
minutes. A. Diagram demonstrates the major and minor regulated secretion pathway. The major regulated
secretion is consisted of a Rab3D-regulated SV maturation and a calcium dependent released process. A
minor regulated endolysosome secretion is present in the LGAC, where the endolysosomes are directly
released from the apical membrane. PIKfyve is involved in the maturation of endolysosomes, which can be
inhibited with apilimod treatment. B. CTSS activity recovered into the media significantly reduced in cells
pretreated with apilimod. *, p=0.0168. C. β-hex activity recovered into the media increased in the media
with apilimod treatment. *, p=0.0093. N=3, N: cell preparation. Each cell preparation utilized one rabbit.
Triplicate wells were conducted with in each cell prep. Individual wells from each cell prep were presented
as points on the graph. Data presented as mean ± SD.
83
7.2 Inhibition of PIKfyve reduces Rab7-enriched vesicle fusion in rabbit
LGACs
Rabbit LGACs were transduced with Rab7-GFP BacMam virus at rate of 10 particles per cell
and stained with CellMask membrane dye to better visualize individual acini. Shown in Figure
7.2A, in DMSO-treated controls, Rab7-enriched vesicles exhibited vesicle motion and fusion with
100 μM CCh stimulation. In cells pretreated with 300 nM apilimod for 1 hour before CCh
stimulation, Rab7-enriched vesicles appeared to be more constrained with less occurrence of
vesicle fusion. The observations were further quantified by a change in vesicle size and vesicle
number per acini. In the DMSO group, Rab7-enriched vesicles started increasing in size 3 min
after CCh stimulation and peaked at a 9-min timepoint, shown in Figure 7.2B. The alteration in
vesicle sizes with stimulation was diminished in cells pretreated with 300 nM apilimod.
Concurrent with the increased vesicle sizes, Rab7-enriched vesicles numbers decreased from 3
minutes after treatment and were consistently reduced over time in the DMSO control. When
comparing apilimod treatment with DMSO, apilimod demonstrated a minimal change with
stimulation, which significantly differed with DMSO-treated control from 3 min and lasted till 15
min post-treatment (Figure 7.2C). These observations may suggest that CCh may have direct
effects in the exocytosis of late endosome/MVB vesicles in LGACs. The inhibition of PIKfyve
reduces the vesicle fusion process which may further influence the release of these vesicles.
84
Figure 7.2 PIKfyve inhibition with apilimod reduces the fusion of Rab7 enriched vesicles. Primary
rabbit LGACs were transduced with Rab7-GFP BacMan virus at a rate of 10 particles per cell. The cells
were stained with CellMask membrane dye, stimulated with 100 μM CCh and imaged for 15 minutes. A.
Time series images show that Rab7 enriched vesicles undergo vesicle fusion with CCh stimulation,
demonstrated in DMSO control. Aplimod treatment reduces vesicle fusion process. Green labels Rab7
enriched vesicle, while red demonstrate plasma membrane. The reduction of membrane fusion is quantified
as increase and decreased vesicle size and a constant decrease in vesicle number in the LGACs, shown in
B and C. Blue asterisk compares the significance between the DMSO control at time of treatment with time
0, red asterisk compares the significance between the apilimod group at time of treatment with time 0. Black
asterisk compares the difference between the DMSO control and apilimod group. N=5, N cell preps. Data
presented as mean ± SEM.
85
7.3 Inhibition of PIKfyve increases endolysosome size but reduces apical
accumulation of endolysosomes
In Figure 6.5 we showed that CCh and PE induced an apical accumulation of endolysosomal
markers in both C57 and Rab3DKO LGACs, with a stronger colocalization in acini stimulated
with PE. I suspected that this may be correlated with increased CTSS activity stimulated with PE.
In this study, we explored the involvement of PIKfyve in the exocytosis of endolysosomes in both
C57 and Rab3DKO mice. Lamp1 is a membrane marker expressed on both late endosomes and
lysosomes, while Rab7 is abundant on late endosomes/MVBs. After being harvested from the mice,
the LG was incubated for 1 hour in apilimod 300 nM in PCM and stimulated with 100 μM CCh or
PE. As previously reported, we found that apimilod treatment significantly increased the Lamp1
(Figure 7.3B) and Rab7 vesicle (Figure 7.4B) colocalization with apical actin in both resting and
stimulated vesicles. The enlargement of endolysosome vesicles during PIKfyve inhibition has been
associate with an altered fission process. No significant difference in vesicle sizes was observed
in the LGACs between C57 and Rab3DKO. However, when comparing the stimulation effects, we
found that PE stimulation led to a reduction of Lamp1 vesicle size with apilimod treatment
compared to the resting acini. This could be caused by either an increased exocytosis of large
Lamp1-enriched endolysosomes or PE may induce a downstream pathway associated with
endolysosome fission. One hypothesis of the enlarged Lamp1-enriched vesicles is that they are
may be more associated with MVB characteristics due to the blockage of PIKfyve. When we
investigated the vesicle size changes of apilimod treated Rab7-enriched vesicles with stimulation,
we found that Rab7-enriched vesicles remained the same with both CCh and PE stimulation. From
these observations, we suspect that lysosome and MVB stimulation may have different sensitives
or degree of responses to CCh and PE. The apical accumulation of endolysosomes was analyzed
86
by measuring the % volume colocalization of endolysosomal markers with the apical actin. Same
as in vivo stimulation in Figure 6.5, the endolysosomes were relocated to the APM with CCh and
PE treatment. In C57 mice, exposure to apilimod did not significantly affect the apical
redistribution of Lamp1(Figure 7.3C), but significantly reduced the apical localization of
Rab7(Figure 7.4C). In Rab3DKO mouse, where the major regulated secretion is impaired and the
endolysosomal secretion is expected to upregulate, we found that apilimod significantly reduced
the apical redistribution of Lamp1 and Rab7 under PE stimulation. Apilimod treatment also
reduced the apical localization of Lamp1 under CCh stimulation but not Rab7. This further
substantiates our hypothesis that PE has more effect in inducing Rab7-enriched MVB release under
PIKfyve inhibition. From these ex vivo tissue data, we sugget that PIKfyve plays a role in
endolysosome exocytosis and PE may induce additional exocytosis of MVB when PIKfyve is
impaired.
87
Figure 7.3. PIKfyve inhibition increases Lamp1 enriched vesicle size and reduces apical accumulation
of Lamp1 at the APM. LGACs from C57 and Rab3DKO were harvested, pretreated with 300nM apilimod
and stimulated with 100 μM CCh and 100 μM PE. A. The LG tissue sections were stained with antibodies
towards Lamp1. B. Apilimod pretreatment significantly increased the Feret diameter of Lamp1 enriched
vesicles in both resting and stimulated LGACs. When comparing the between secretagogues, PE induced
vesicles of smaller sizes in both C57 and Rab3DKO LG pretreated with apilimod. Green represents Lamp1,
while red is rhodamine phalloidin labelling actin filaments. C. Apical accumulation of Lamp1 was
quantified by measuring the colocalization of Lamp1 with the apical actin. In DMSO control, both CCh
and PE induced a significant increase in the apical accumulation of Lamp1 in C57 and Rab3DKO LGACs,
with Rab3DKO demonstrating a more substantial increase. When treated with apilimod, LGAC deficient
of Rab3D exhibited a significant reduction in apical redistribution in LGACs stimulated with CCh and PE.
Data presented as mean ± SD.
88
Figure 7.4 PIKfyve inhibition increases Rab7 enriched vesicle size and decreases apical accumulation
of Lamp1 at the APM. LGACs from C57 and Rab3DKO were harvested, pretreated with 300nM apilimod
and stimulated with 100 μM CCh. A. The LG tissue sections were stained with antibodies towards Rab7.
Green represents Rab7, while red is rhodamine phalloidin labelling actin filaments. B. Apilimod treatment
significantly enlarged the Rab7-enriched vesicles in both resting and stimulated LGACs in C57 and
Rab3DKO. C. Apical accumulation of Rab7-enriched vesicles significantly decreased in both CCh and PE
stimulated C57 LGACs. The reduced apical accumulation was only observed in the PE-stimulated LGACs
in Rab3DKO mouse. Data presented as mean ± SD.
89
7.4 PIKfyve gene expression is decreased in Rab3DKO and NOD mouse
From data above we hypothesized that PIKfyve plays a critical role in CTSS trafficking. To
further investigate its involvement in the disease model, we looked at the gene expression of
PIKfyve and genes related to lysosome biogenesis and maturation in the LG of diseased NOD
mouse and disease trafficking model Rab3DKO. Interestingly, PIKfyve mRNA levels showed a
significant 0.5-fold decrease in both Rab3DKO and NOD mouse compared to the corresponding
healthy controls, C57 and Rab3DKO. Similarly, the lysosomal calcium channel TRPML1 gene
expression significantly decreased in both Rab3DKO and NOD mouse. However, TFEB showed
a different response, decreasing in Rab3DKO and increasing in NOD mouse. TRPML1 is a
P(3,5)P2 -gated lysosomal Ca
2+
channel that helps media phagosome lysosome fusion. PIKfyve
accelerates lysosome acidification though TRPML1 activation
102
. Therefore, it is expected that
TRPML1 demonstrates the same trend in gene expression as PIKfyve. Pharmacological inhibition
of PIKfyve activates the gene expression of TFEB
37
. Also different from what we expected, where
PIKfyve inhibition result in decreased CTSS activity, both NOD and Rab3DKO have increased
tear CTSS activity in tears. However, gene expression does not represent protein expression and
lipid kinase activity. There are also may factors that may play a role in the change of CTSS activity
including inflammation in NODs, altered trafficking pathways that may affect CTSS protein
activity, imbalance between endogenous CTSS inhibitors. Therefore, these data provide a first
glance into the PIKfyve gene expression in LGs of mouse models, which will be further
investigated with protein assays in the future.
90
Figure 7.5 Gene expression of PIKfyve, TFEB, TRPML1 in BALB/c, NOD, C57 and Rab3DKO
mouse. A. PIKfyve gene expression significantly decreased in the LG of Rab3DKO mouse compared to
C57. B. Lysosome biogenesis factor TFEB mRNA levels also significantly reduced in the LG of Rab3DKO
mouse. C. Lysosome acidification unit TRPML1 gene levels significantly decreased in Rab3DKO LG. D.
similar to Rab3DKO, PIKfyve gene expression decreased in the LG of male NOD mouse. E. TFEB levels,
however, significantly increased in NOD mouse. F. TRPML1 showed a similar reduction in NOD mouse
LG as in Rab3DKO. N=5, N: mouse count.
91
CHAPTER 8
Discussion
8.1 Interaction between CX3CL1 and CTSS contributes to ocular
inflammation
In NOD mice, autoimmune dacryoadenitis associated with elevated CTSS activity in LG and tears
is correlated with increased CX3CL1 in LG, cornea and tears, and with increased CX3CR1+
immune cells in LG. Exposure of cultured LGAC to IFN-γ coordinately induces expression of
CX3CL1 and CTSS. Moreover, rhCTSS at activities found in SS patient tears increases CX3CL1
expression in HCE-T cells and recovery of CX3CL1 in culture medium. Figure 8.1 shows the
proposed relationships between IFN-γ, CTSS, CX3CL1 and CX3CR1, highlighting mechanisms
that may occur both within LG and on the ocular surface.
The elevated CX3CL1 found in the tears of NOD mice is likely due in part to the presence of
CX3CL1 in apical secretory compartments also enriched in CTSS (Figure 2.2), which may expose
membrane-bound CX3CL1 to increased active CTSS also present in these compartments for
shedding in tears.
60
IFN-γ is also elevated in tears of NOD mice with autoimmune dacryoadenitis
and in patients with primary SS.
60
The demonstration that both IFN-γ and CTSS exposure to HCE-
T cells increase recovery of CX3CL1 in culture medium is consistent with a model where these
tear constituents continue to elicit inflammatory responses in cornea and other ocular surface
epithelia, which in turn can further enhance shedding of CX3CL1 into tears.
92
Figure 8.1 Proposed mechanisms implicated in CX3CL1 induction, cleavage and recruitment of
CX3CR1+ immune cells in autoimmune dacryoadenitis. 1. Pro-inflammatory cytokines including IFN-
γ induce gene and protein expression of CX3CL1 and CTSS. 2. CTSS activity also contributes to
upregulation of CX3CL1 expression. 3. Induced CX3CL1 is synthesized as the membrane-bound form and
trafficked to the plasma membrane where it is exposed. 4. Upregulated CTSS is secreted, leading to an
accumulation of active CTSS in areas rich in CX3CL1. 5. Active CTSS enhances shedding of membrane-
bound CX3CL1 and generates soluble CX3CL1. 6. Soluble and membrane-bound CX3CL1 interact with
CX3CR1 to recruit immune cells to sites of inflammation. 7. Intravesicular cleavage of membrane-bound
CX3CL1 by increased CTSS present in both basolaterally-targeted and apically-targeted vesicles may also
occur, increasing CX3CL1 shedding into the interstitium and tears. 8. Restrained constitutive recycling of
CX3CL1 by alterations in actin cytoskeleton may increase membrane-bound CX3CL1 on the plasma
membrane, also promoting shedding of CX3CL1. Solid arrows highlight mechanisms suggested as primary
explanations in the study while dashed arrows are alternative mechanisms.
Our finding is the first to report CX3CL1 in tears in a model of autoimmune dacryoadenitis.
CX3CL1 is elevated in tears of general dry eye
103
and evaporative dry eye
104
patients, but SS
patient tears have not been evaluated. CX3CL1 in tears could participate in ocular surface
93
pathology in SS by recruiting immune cells to the ocular surface
105,106
and possibly interfering
with corneal nerves. CX3CL1 has been previously implicated in development of autoimmune
exocrinopathy in the LG in another murine model,
78
but its function and mode of generation were
not addressed.
We propose that the accumulation of CX3CR1+ immune cells in NOD mouse LG is driven by
increased shedding of CX3CL1 from LGAC, an effect driven in part by elevated CTSS present in
the interstitium of the LG. CX3CL1 is clearly increased in basolateral membranes and basolateral
vesicles in LGAC from NOD mice (Figure 2.2B). While increased LGAC CTSS is clearly
secreted apically,
56,57
its cellular overexpression is expected to increase its abundance in tissue
interstitium through release at basolateral membranes. CTSS is normally present in all cell
lysosomes and in late endosomes and lysosomes of professional antigen-presenting cells. This
localization is associated with its mannose-6 phosphorylation (M6P), resulting in its capture by
M6P-receptors and active sorting to endolysosomal compartments. Overexpression of cathepsins
results in increased amounts of cathepsins lacking M6P tags, since this modification is rate
limiting.
107
These unmodified cathepsins can be secreted by bulk flow mechanisms to the
extracellular space where they participate in degradation of extracellular matrix and other
inflammatory functions.
64,107
CTSS is one of the only cathepsins that is equally active at a neutral
(extracellular) pH, as well as in the acidic environment of the lysosome.
64
Elevated LG CTSS
associated with increased IFN-γ or other pro-inflammatory cytokines may therefore yield
additional extracellular CTSS in proximity to overexpressed CX3CL1 that is enriched in
basolateral vesicles and membranes. Figure 8.1 summarizes this aspect of our model.
94
IFN-γ is a known inducer of CTSS, and is elevated in tears and LG of diseased NOD mice.
60
IFN-
γ exposure in vitro also modestly increases CTSS in HCE-T cells and primary LGAC.
60
The work
herein has reconfirmed these findings, further showing that IFN-γ induction increases CX3CL1
gene and protein expression. The time course for detection of increased CX3CL1 expression
relative to CTSS expression in LGAC ((Figure 3.1) and HCE-T cells (Figure 3.2) shows that
CX3CL1 gene expression precedes CTSS (Figure 3.1). However, increased recovery of CX3CL1
in culture medium lags relative to increased CTSS activity (Figure 3.2). CTSS activity itself
increases CX3CL1 in HCE-T cells, although these effects are not detectable until later times
(Figure 3.3). The initial increase in CX3CL1 expression appears related to IFN-γ exposure, rather
than to the increased expression of CTSS. The burst of CTSS activity detected in the culture
medium of HCE-T cells that precedes the increased CX3CL1 recovery in culture medium may
reflect secretion of existing cellular stores of CTSS prior to the induction of additional CTSS
expression to replenish intracellular stores.
108
In this study, we focused mainly on IFN-γ, although
SS cytokines including TNF-α and IL-1β may also affect CX3CL1.
We did not focus extensively on the role of CTSS in LGAC shedding of CX3CL1 in cultured
LGAC beyond initial findings in Figure 3.1. This choice was due in part to the large number of
mice required for cultured LGAC preparations, and in part to the lower extent of CX3CL1 protein
expression elicited in LGAC by IFN-γ (1.9-fold) relative to HCE-T cells (196-fold). However,
several findings in HCE-T cells suggest that CTSS directly cleaves CX3CL1, including
observations that added CTSS increased CX3CL1 in culture medium, while CX3CL1 was
decreased in culture medium by the CTSS inhibitor, Z-FL (Figure 3.4). Moreover, heat-
inactivated rhCTSS did not elicit CX3CL1 release into culture medium (Figure 3.4). Other
proteases upregulated in SS such as MMP-9 may also contribute to increased tissue and tear
95
CX3CL1. MMP-9 is also upregulated by exposure of HCE-T cells to rhCTSS activity but this
effect requires 24 hr of treatment.
65
We used only 2 hr of rhCTSS treatment of HCE-T cells to
minimize later effects of CTSS on other signaling pathways.
An alternative explanation for the increased CX3CL1 released into tears and possibly. interstitium
beyond active cleavage by elevated CTSS is through altered intracellular trafficking. Membrane-
bound CX3CL1 is rapidly recycled between the plasma membrane and endocytic compartments,
regulating its availability for processing.
109
Constitutive endocytosis of CX3CL1 is reported to
protect membrane-bound CX3CL1 from cleavage by metalloproteases.
110
When actin cytoskeleton
is disrupted, more interaction between CX3CL1 and metalloproteases occurs, resulting in shedding
of soluble CX3CL1.
111
The actin cytoskeleton is intimately associated with exocytosis and
endocytosis in LGAC. We have reported multiple changes in LGAC secretory pathways and actin
filament organization in NOD mouse.
19,112
Such a mechanism might account for increased
CX3CL1 in tears, and the increased CX3CL1 in interstitium implied by detection of increased
CX3CL1 at the basolateral membranes of NOD mouse LGAC and the increased accumulation of
CX3CLR+ immune cells in NOD mouse LG. Future studies utilizing selective protease inhibitors
in vivo may elucidate the role of these different processes in regulation of the CX3CL1-CX3CR1
axis in LG.
Several CTSS inhibitors are in clinical trials for autoimmune diseases, including RWJ-445380
(www.clinicaltrials.gov, Identifier: NCT00425321) and RO5459072 (www.clinicaltrials.gov,
Identifier: NCT02701985). These therapeutic studies have focused primarily on assessment of
MHC II-related endpoints such as Iip10 accumulation or H2-Ab1 gene expression as indicators of
therapeutic efficacy.
113
CX3CL1 levels and CX3CL1+ immune cell counts in tissues of interest
may serve as additional endpoints for CTSS activity. The CX3CL1-CX3CR1 axis is implicated in
96
RA pathogenesis, where CX3CL1 is expressed in fibroblast-like synoviocytes and endothelial cells
in RA synovium. Accumulation of CX3CR1+ T-cells and macrophages in RA synovial tissue is
also associated with upregulation of CX3CL1.
114,115
Since comparable changes are observed in the
LG of NOD mice, the CX3CL1-CX3CR1 axis may participate similarly in autoimmune
dacryoadenitis. CX3CL1-CX3CR1 also participates in immune cell homing to the conjunctiva
during ocular inflammation induced by benzalkonium chloride.
42
Currently, an anti-CX3CL1
monoclonal antibody therapy, KANAb001 (E6011), is undergoing phase I and II clinical trial in
Japan for RA and Crohn’s disease.
116
CX3CL1 may therefore represent another potential
therapeutic target in SS to mitigate autoimmune exocrinopathy.
8.2 Rab27a regulate CTSS secretion in the LGACs through both a major
regulated secretion and a minor endolysosomal pathway
SS-associated autoimmune dacryoadenitis is linked to elevated CTSS activity in tears and LG,
an increase which appears to contribute to ocular inflammation. Concurrent with increased tear
CTSS activity, the reduced expression of Rab3D and the redistribution of remaining Rab3D-
enriched vesicles is observed in LGAC of both SS patient
58
and male NOD mouse LG
19
. Rab3D
and Rab27a/b isoforms are both associated with tear protein exocytosis in LGAC. Previous
findings have suggested that reduced Rab3D function contributes to altered tear protein secretion
at the APM, leading to changes in tear content in SS. CTSS activity in tears is reduced in mice
lacking either Rab27a or b, but enhanced in Rab3D knockout mice
19
. We propose that Rab27a
contributes to regulation of CTSS secretion in LGAC and may be a principal driver of the increased
tear CTSS secretion seen in the diseased NOD mouse LG. Figure 8.2 shows the proposed
mechanism of Rab27a in CTSS secretion in healthy and diseased (autoimmune dacryoadenitis)
97
LGAC, highlighting its proposed involvement in both the major regulated secretion and a novel
endolysosomal secretory pathway that appears to be activated in disease.
Figure 8.2. Proposed mechanism of Rab27a involvement in CTSS traffic in LGAC. CTSS is
synthesized and initially sorted to endolysosomal compartments. Rab27a regulates CTSS secretion at the
APM though both the major regulated secretory pathway and a direct endolysosomal secretory pathway. In
healthy LGAC, Rab27a-enriched endolysosomally-derived vesicles may fuse with Rab3D-enriched SV,
generating SV enriched in Rab3D, Rab27a and Rab27b containing small amounts of CTSS and other
lysosomal enzymes that are secreted with stimulation. In SS, the depletion of Rab3D-enriched SV leads to
accumulation of Rab27a-enriched endolysosomally derived vesicles that may accumulate and fuse directly
with the APM to release increased CTSS, in the absence of a functional major regulated secretory pathway.
Solid lines highlight mechanisms supported as the most likely in the study while dashed arrows represent
alternative mechanisms. Figure created with BioRender.com 2020 version (Toronto, ON, Canada).
In healthy LGAC, we propose that Rab27a mediates the trafficking of endolysosomes to
Rab3D-enriched SV. This pathway is responsible for providing low levels of lysosomal enzymes
such as CTSS to SV for secretion at the APM into tears. This premise is also consistent with
findings by 3D-structured illumination microscopy showing that Rab27a, Rab27b, and Rab3D
reside at distinct microdomains on SV in healthy mouse LGAC
19
. This model is further supported
by the findings here that inhibition of Rab27a using the DN mCFP-Rab27a construct in vitro
98
impaired the release of β-hex into culture medium from healthy LGAC (Figure 4.8), reflecting the
involvement of Rab27a with Rab3D and Rab27b in formation and fusion of healthy SV containing
some lytic enzymes from the major regulated secretory pathway.
Direct fusion of Rab27a-enriched endolysosomal vesicles with the APM may also occur in
healthy LGAC, but appears to be less pronounced than secretion from the major regulated pathway
in LGAC when stimulated with CCh. Time lapse imaging experiments in healthy LGAC
expressing fluorescently-tagged mCFP-Rab27a showed that CCh could evoke direct exocytosis of
mCFP-Rab27a-enriched SV (Figure 4.6). This fusion appears, however, less vigorous than rates
of fusion of fluorescently-labeled Rab3D- and Rab27b-enriched SV with the APM seen in
response to CCh stimulation in vitro, suggestive that at least under conditions of engagement of
muscarinic receptors, that the direct endolysosomal secretory pathway has a secondary role.
In disease-model male NOD mouse LG, Rab3D expression is decreased, with remaining
vesicular stores redistributed to the basolateral area. At the same time, Rab27a-enriched
endolysosomal vesicle formation and trafficking of CTSS-enriched vesicles into the apical region
is increased. These Rab27a-enriched endolysosomal-derived vesicles may accumulate in the
absence of Rab3D-enriched SV, and upon stimulation, undergo homotypic fusion as well as
increased direct fusion with the APM. Rab27a’s involvement in increased CTSS secretion in
diseased LGAC is supported by findings of increased CTSS colocalized with increased subapical
Rab27a-enriched vesicles (Figures 4.1 and 4.2) in NOD mouse LGAC which are depleted
following CCh (Figure 4.10). These subapical Rab27a-enriched vesicles are suspected to be
primarily secretory vesicles with a few endolysosomes. Our previous work has shown that Rab27a
is located on the same vesicles as the traditional secretory vesicle markers, Rab27b and Rab3D
[26]. As well, the Rab27a-enriched vesicles range from 0.75–1 µM in diameter, which is the
99
approximate size for secretory vesicles. In Figure 4.13, an increased colocalization of Rab27a and
Lamp2 was observed in the resting LGACs of NOD mouse. These subpically-enriched Rab27a-
enriched vesicles in NOD mouse LGAC are also of a reduced size (Figure 4.2) relative to those
in BALB/c mouse LGAC, suggestive of either an impaired terminal maturation process with
Rab3D-enriched SV and/or a shift to the endolysosomal secretory pathway involving direct fusion
as a dominant secretory pathway.
A secretagogue-dependent minor regulated secretory pathway has been identified in both
pancreatic
22
and parotid acinar cells
101
. In pancreatic acinar cells, Rab27a was shown to mediate
this pathway which regulates apical exocytosis of endolysosomes
22
. Primary pancreatic acini
isolated from Rab27a-deficient mice exhibited a significant reduction in stimulated amylase
release
22
. At the same time, stimulation of WT pancreatic acini recruited Lamp1-enriched vesicles
to the APM, while no such effect was observed in pancreatic acini from Rab27a-deficient ashen
mouse acini, indicating impaired endolysosomal membrane fusion in the absence of Rab27a
22
.
We propose that a comparable endolysosomal secretory pathway is responsible for secretion of
endolysosomes containing CTSS in healthy LGAC, a process that is dramatically increased in
disease. Consistent with this, in otherwise healthy Rab27
ash/ash
Rab27b
-/-
mouse LG tissue lacking
both Rab27 isoform activities, lysosomal accumulation is clearly detectable
25
. Similarly to the
studies in pancreatic acini, Lamp2-enriched vesicles colocalized with Rab27a are enriched at the
APM region with CCh stimulation, but the enrichment is greater in NOD mouse LG which secrete
more CTSS (Figure 4.9).
LGAG are innervated by both parasympathetic and sympathetic nerves, which modulate
secretion through neurotransmitters and neuropeptides
5
. The interaction of these secretagogues
with their receptors serve key roles in activating specific signaling pathways through increased
100
calcium influx and activation of secondary messengers. Our studies have primarily focused on
stimulation of muscarinic receptors with CCh. It is possible that engagement of signaling pathways
evoked by other neurotransmitters or neuropeptides could more potently activate the
endolysosomal secretory pathway, perhaps in response to ocular surface pain or infection, thus
triggering the transient release of a more proteolytic tear film.
The apparent upregulation of the endolysosomal secretory pathway in autoimmune
dacryoadenitis may also be associated with alterations in signaling pathways associated with
autoimmune inflammation of the LG. Peripheral neuropathy is a key contributor to the pathology
of SS
5,117
. The LGAC of MRL/MpJ-Fas
lpr
, another murine model of SS, exhibit a denervation-like
supersensitivity to exogenously-added neurotransmitters
118
. This altered sensitivity to
secretagogues is thought to result from the presence of proinflammatory cytokines that impair
neurotransmitter release from the nerve terminals
118,119
. In aged mice suffering from reduced tear
flow, not only was neurotransmitter release defective, but the downstream signaling pathways
evoked by cholinergic and a1-adrenergic agonists were impaired
120
. The NOD mouse may
experience a similarly-impaired neural response in the LG. Altered sensitivity to signaling agonists
may contribute to its distinctive secretory features including activation of the endolysosomal
secretory pathway.
We observed what appeared to be larger Rab7-enriched and Lamp2-enriched vesicles beneath
the APM in unstimulated NOD mouse acini, relative to unstimulated BALB/c mouse acini, as
shown in Figure 4.9. Imaging of Rab27a and Lamp2 immunofluorescence further demonstrated
that these large endolysosomal structures were enriched in Rab27a (Figure 4.13). These subapical
structures may be MVB or amphisomes, which are enriched in both Rab7 and Lamp2
121
, thus
suggesting that MVB or amphisome secretion may be increased in diseased LG. Proinflammatory
101
cytokines such as interferon-γ (IFN-γ) and tumor necrosis factor (TNF-α) are highly expressed in
NOD mouse LG
60
. LGAC treated in vitro with IFN-γ exhibit characteristics of NOD mouse LGAC
including reduced Rab3D
60
, increased MHC II
60
and increased CX3CL1 expression
2
. Cultured
LGAC exposed to IFN-γ also exhibit enhanced CTSS secretion with CCh stimulation, while β-hex
release is unaffected
60
. IFN-γ has been reported to stimulate exosome secretion through Rab27a-
mediated release of amphisomes in lung epithelial cells
122
. Studies in eosinophils also reported
enhanced MVB membrane fusion and degranulation as well as increased exosome release, with
IFN-γ
123
. One factor stimulating the increased secretion of CTSS via Rab27a from NOD mouse
LGAC may thus be the local increase in proinflammatory cytokines.
Rab27a was originally thought to be functionally redundant with Rab27b in regulated
secretion, since these isoforms share some common effectors
124
. However, Rab27a can play
drastically different roles in trafficking relative to Rab27b, depending on the cell type and specific
secretory pathways under study, through interaction with distinct effectors
125
. In this study, we did
not investigate how Rab27a plays concurrent roles in major and endolysosomal secretory pathways,
as our main focus was to address whether Rab27a mediates CTSS secretion in LGAC and how
this function is impacted in disease. In future investigations we can explore the interactions of
Rab27a with its diverse effectors, and investigate their expression and regulation in disease, to
help better explain Rab27a’s ability to participate in different secretory pathways.
In conclusion, Rab27a regulates CTSS secretion though both the major regulated secretory
pathway and a novel endolysosomal secretory pathway in LGAC. This study is the first to link
Rab27a directly to CTSS secretion in the LG. Our finding is of potential clinical relevance since
CTSS is implicated in the etiology of SS
84
. Continuing functional studies and evaluation of protein
102
trafficking in disease-model and healthy mice may in future provide new targets for therapeutic
therapeutic development.
8.3 Neural regulation affects CTSS secretion
Neural regulation is key to regulating lacrimal fluid secretion. The parasympathetic and
sympathetic nerves innervating the LG stimulate lacrimal fluid secretion through neurotransmitters
activating specific receptors at the basolateral membranes of the acini. Among them,
parasympathetic cholinergic muscarinic and sympathetic α1-adrenergic receptors have been
reported to be predominant in protein secretion in murine LG. Despite many reports on the
downstream signaling pathways associated with these receptors, the link between different neural
stimulation and distinct protein trafficking is still unclear. Increased CTSS activity in tears has
been associated with the etiology of SS-associated autoimmune dacryoadenitis. Therefore, it is of
importance to elucidate the trafficking and upstream regulators modulating CTSS secretion. In this
study, we investigated the differential secretion of CTSS stimulated by α1-adrenergic agonist
phenylephrine (PE) and cholinergic muscarinic agonist carbachol (CCh). I propose that PE
increased tear CTSS secretion, possibly through a downregulated Rab3D-enriched trafficking and
an upregulated minor endolysosomal pathway. Figure 8.3 shows the proposed mechanism of
CTSS trafficking stimulated by PE in healthy LGACs.
103
Figure 8.3. Proposed mechanism of phenylephrine stimulated CTSS traffic in LGAC. CTSS is
enriched in both endolysosomes and SVs in the LGAC. Cholinergic agonist carbachol activates the
acetylcholine receptor M3 in mouse LG and stimulate a drastic increase in intracellular calcium. Increased
intracellular calcium induce Rab3D-enriched major regulated secretion of lysosomal contents, through
triggering (1) homotypic fusion between Rab3D-enriched SV, (2) stimulate Rab3D-enriched SV fusion
with the APM and (3) induce heterotypic fusion between endolysosomes and Rab3D-enrich SVs. The
stimulation of α1-adrenergic agonist phenylephrine (PE) resulted in a small increase of intracellular calcium,
where limited fusion of Rab3D-enriched SVs was observed, shown in (4). (5) Decreased Rab3D-enriched
major secretion is suspected to give rise to an increased minor endolysosomal secretion. (6) PE is also
suspected to induce increased fusion of endolysosome with Vamp2-enriched recruitable SV, which may
also account as an alternative secretion of CTSS.
During CCh stimulation, I propose that the Rab3D enriched major-regulated pathway,
accounting for low levels of CTSS secretion, is upregulated. This is likely due to a drastic increase
in intracellular calcium that leads to an increased fusion of Rab3D-enriched SV with each other
and with the APM. Though [Ca2+] imaging I showed that CCh stimulation significantly elevated
intracellular [Ca2+] in isolated mouse LGACs. Elevated intracellular [Ca2+] has been associated
with fusion of mature SVs with the APM[51]. I also observed extensive increases in Rab3D
vesicle size and apparent vesicle fusion with the APM in LGACs stimulated with CCh in Figure
104
6.1. CTSS is localized in both SVs and endolysosomes[41]. Upregulated colocalization of Rab3D
and endolysosomal markers was observed with CCh stimulation, indicating increased sorting of
endolysosomal content into the Rab3D-enriched SVs in Figure 6.2. The minor-regulated pathway,
where a subpopulation of endolysosomes can directly fuse with APM may also occur with CCh
stimulation (Figure 6.5) but appear to be less pronounced than the major regulated pathway.
Through in vivo and in vitro secretion, we found that PE significantly increased CTSS activity
in tears by elevating CTSS secretion in LGACs in Figure 4.4 and Figure 4.5. PE has been reported
to stimulate a small amount of intracellular [Ca2+] through possible upregulation of cGMP[53]
and cyclicADP-ribose[14] levels. Similarly, through [Ca2+] imaging we found that PE induced
less intracellular [Ca2+] and had minimal effect on Rab3D-enriched SV fusion. The small effect
on intracellular [Ca2+] is suspected to impede the major-regulated pathway through hindering the
exocytosis of subapically distributed SVs in Figure 6.1. In addition, in vivo and in vitro secretion
assays demonstrated that PE has a limited effect in stimulating β-hex secretion. β-hex is commonly
associated with the Rab3D-enriched major regulated secretory pathway, which further
substantiates the limited ability of PE in inducing major regulated secretion. Concurrent with
obstructed major regulated pathway, a significant accumulation of endolysosomal markers at the
APM was observed with PE stimulation in Figure 6.5. Increased CTSS secretion has been
associated with a minor regulated release through direct fusion of Rab27a-enriched endolysosomes
with the APM[41]. Another alternative secretion pathway involves a subpopulation recruitable SV
enriched with SNARE protein Vamp2[51]. where Vamp2 are mobilized to the APM under CCh
and PE stimulation (Figure 6.6). Increased colocalization of Vamp2 and endolysosomal markers
was observed in LG stimulated with PE, indicating a possible increase of endolysosomal content
sorted into the Vamp2-enriched SV (Figure 6.7). These observations suggest that PE triggers an
105
increase in CTSS release in LGACs through either a minor regulated direct endolysosomal
exocytosis or increased endolysosomal content sorting into a Vamp2-enriched secretion.
We have previously reported a possible counterbalance between Rab3D-enriched major
regulated pathway and an Rab27-dependent alternative pathway associated with CTSS
secretion[37]. Rab3D proteins are enriched on mature SVs and essential for regulating SV
maturation[43] and premature fusion[44]. Reduced expression and basolateral redistribution of
Rab3D is observed in the LGACs of both SS patients and a SS murine model, male NOD mouse,
where CTSS activity is increased in tears[37, 54]. Increased accumulation of endolysosomal
markers with Rab27a was observed in LGACs of NOD mouse, suggesting an upregulated Rab27a-
dependent minor endolysosomal secretion[41]. Additionally, increased of tear CTSS activity was
observed in mouse lacking Rab3D, while decreasing in Rab27 deficient mouse[37]. To test
whether diminished Rab3D-enriched major regulated pathway can increase CTSS secretion, we
compared that secretion profile of Rab3D knockout mouse with C57 control. Rab3DKO is
believed to have a dysfunctional major regulated pathway. CCh is expected evoke the major
regulated secretion, while PE mainly promotes the minor regulated endolysosomal secretion.
Interestingly, Rab3DKO exhibited higher tear CTSS activity with PE stimulation at lower
concentrations, indicating a possible increased sensitivity of CTSS secretion in the absence of the
major regulated secretion Figure 6.4. Both CCh and PE induced higher tear CTSS activity in
mouse lacking Rab3D. However, no significant difference was observed in the ability to mobilize
endolysosomal markers to the apical membrane with PE stimulation in LGACs of Rab3DKO and
C57. These data suggest impediment of the Rab3D-enriched major regulated pathway through
either Rab3D deficiency or limited intracellular [Ca2+] results in increased CTSS activity, while
the mechanism of increase in CTSS differs between different steps of blockage.
106
Asides from the differences in tear CTSS activity, we also observed an altered volume and
concentration in tear stimulated with different secretagogues. PE significantly reduced tear fluid
secretion, resulting in more concentrated tears, compared to CCh (Figure 5.1). This is suspected
to be related a possible differential activation of ion channels localized on the apical and basolateral
membrane responsible for electrolyte and water secretion. In general, fluid secretion can occur
through either active electrogenic secretion of an anion or an absorption of cation. In LGACs, CCh
stimulate fluid secretion through activating apical Cl-, K+ channels, basolateral Na+-H+
exchanges and an increased recruitment of Na+,K+-ATPase to the basolateral membrane[2, 55].
In this thesis, I mainly focused on the secretion profile of PE in LGACs, since it comprises 70-80%
of the LG. However, this does not eliminate the possibility that PE and CCh have different effects
on lacrimal ductal epithelial cells. Ductal cells are primarily responsible for secreting electrolytes
and water. CCh-stimulated electrolyte and water secretion in ductal cells, distinct from acinar cells,
is regulated by the activation of intermediate-conductance calcium-activated potassium channels
(IKCa1K+) at the apical membrane[55]. PE also initiates a rapid fluid secretion in isolated mouse
LG ducts, which was partially inhibited through NO/cGMP pathway inhibition and completely
depleted through [Ca2+] blockage[12]. Therefore, the reduced tear volume stimulated by PE can
be associated with differential downstream signaling and ion channel activations in both acinar
and ductal cells.
Despite the fact that CCh and PE have different abilities in stimulating intracellular [Ca2+],
we cannot conclude that intracellular [Ca2+] is the sole cause of the differential secretion between
PE and CCh. In fact, in rat LGACs, a crosstalk occurred between α1-adrenergic and P2X7
receptors, where P2X7 inhibition blocked PE-stimulated increase in intracellular [Ca2+] but not
peroxide secretion[56]. PE can also induce PKC signaling pathways, where PKCɛ stimulate
107
protein secretion through remodeling of apical actin filaments, while PKCα and –δ nhibit secretion.
In this thesis, I did not focus on investigating downstream signaling pathways activated by PE,
since my main focus was to explore the ability of PE in inducing CTSS secretion and differential
trafficking in the LGACs. Future investigations will further explore other downstream and cross-
reactive processes related to membrane trafficking.
Peripheral neuropathy is a key contributor to the pathology of SS [2,54]. Impaired LG nerve
activity was found in the LG of MRL/MpJ-Faslpr mouse, another murine model of SS. MRL/MpJ-
Faslpr exhibit a denervation-like supersensitivity to exogenously-added cholinergic and a1-
adrenergic secretogogues, with increase of lymphocytes infiltrating the LGs[57]. Upregulated
proinflammatory cytokines, including tumor necrosis factor (TNF)α, interleukin (IL)-1α and -β
have been reported to block the release of neurotransmitters from nerves, resulting in the functional
“denervation” in the LG[58]. IL-β can also induce NO production in LGACs[59], which suggests
altered trafficking with inflammation. In addition to the interference of proinflammatory cytokines,
antibodies to M3AChR were found in SS patients, which obstructs the interaction between
M3AChR and ACh. In response to the blockage of M3AChR, increased synthesis of M3AChR
was found in the labial salivary glands of primary SS[60]. In NOD mouse, infusion of antibodies
towards M3AChR significantly reduced secretion in the submandibular gland cells[61]. In a rat
aging model, where tear production decreased with age, a reduction in functional secretion
stimulated by CCh and IPR was observed in LG fragments, while PE stimulated secretion
remained the same in aged glands[62]. This may suggest that the M3AChR and α1-adenergic have
different sensitivity with disease onset. I suspect a combined effect of proinflammatory cytokines
and inhibitory antibodies towards M3AChR alters the secretion in the LG of NOD mouse.
Increased proinflammatory cytokines in NOD LG[32] possibly leads to a denervation-like
108
sensitivity and an alteration of downstream signaling, resulting in the upregulated minor
endolysosomal pathway and increased tear CTSS. Further investigations will explore the effect of
neural regulation in diseased mouse and the possible involvement of inflammation.
In conclusion, PE significantly increased CTSS activity in tears, through a downregulated
major-regulated secretion and an upregulated minor endolysosomal pathway in LGACs. This is
the first study to link neural regulation with CTSS trafficking. This study is of clinical relevance
as CTSS is implicated in the etiology of SS [29]. Elucidating the relationship between different
neural regulations and distinct protein trafficking pathways provides more insight in the disease
biology of SS and may provide more targets for novel therapeutic development.
8.4 PIKfyve may be a key regulator in CTSS secretion in the LGACs
PIKfyve is commonly associated with maturation and trafficking of endolysosomes
36
. Studies
have reported a reduction of CTSS activity during PIKfyve inhibition in T cells, which
subsequently resulted in impaired MHC II antigen presentation
126
. This provides the bases for
PIKfyve being a novel target for autoimmune diseases and cancer. Asides from reducing
intracellular CTSS activity, I also suspected that PIKfyve may play a role in CTSS trafficking in
the LGACs.
In previous studies, I substantiated that CTSS is largely secreted though a Rab27-dependent
endolysosomal trafficking pathway. Many papers also reported that PIKfyve inhibition leads to
enlarged endolysosomes, which is suspected to relate to lysosomal coalescence
37
. However, the
relationship between increased size and vesicle trafficking is still unknown. In our study we
hypothesized that there may be a reduction in endolysosome traffic with PIKfyve inhibition.
Through in vitro secretion assay in rabbit LGACs treated with PIKfyve inhibitor apilimod, we
109
found that inhibiting PIKfyve significantly reduced CTSS activity and increased β-hex activity
recovered into the media in Figure 7.1. β-hex is secreted though the Rab3D-dependent major
regulated pathway. Previously, we have suspected a counterbalance between the Rab3D-enriched
major regulated pathway and the Rab27-enriched minor regulated endolysosomal pathway
19
.
Therefore, the increased β-hex may indicate a possible upregulation in the major regulated
pathway cause by the blockage of endolysosomal pathway. Since CTSS play a major role in the
antigen presentation in the MVBs, which were the main compartment inhibited by PIKfyve in T
cells
126
. We further investigated the trafficking profile of late endosomes/MVBs in primary rabbit
LGAC treated with apilimod (Figure 7.2). The LGACs were transduced with Rab7 BacMam virus,
treated with apilimod and stimulated with CCh. Rab7-enriched vesicles exhibited vesicle fusion
under CCh stimulation, which was reduced with apilimod treatment. This further substantiated a
possible blockage in MVB released in the LGAC. To further test whether the same effects are
present in mouse LGACs, we harvested LG tissue from C57 and Rab3DKO mouse and observed
the change of endolysosomal markers with and without PIKfyve inhibition. As reported, we
observed an increase in endolysosomal vesicles with apilimod treatment in both resting and
stimulated LGACs (Figure 7.3 and Figure 7.4). In DMSO control, endolysosomes exhibited an
accumulation at the APM with both CCh and PE stimulation, same as observed in vivo topical
stimulation in Figure 6.6. However, apilimod treatment significantly reduced the apical
redistribution process. Rab3DKO and PE stimulation are both suspected to have a reduced major
regulated trafficking through deficient Rab3D and a minimal Ca
2+
response. In the DMSO control
group, Rab3DKO mouse exhibited the highest apical accumulation with PE. With apilimod
treatment, the apical mobilized endolysosomes were significantly reduced in PE-stimulated
Rab3DKO LGACs, which was more apparent than in C57 mouse LGACs. This may suggest that
110
blockage of major regulated trafficking may upregulated endolysosome exocytosis and PIKfyve
also helps modulate this process.
Finally, we compared the gene expression of PIKfyve and lysosome related factors TFEB and
TRPML1 in BALB/c, NODs, C57 and Rab3DKO. Interestingly, PIKfyve and TRPML1 gene
expression were reduced both in NOD and Rab3DKO mouse LG, while TFEB was increased in
NOD mouse LG and decreased in Rab3DKO mouse LG. These data further substantiate that there
is a change in PIKfyve and lysosomal proteins in the LG of the SS disease model as well as the
deficient trafficking model. However, it is hard to determine the role of PIKfyve in the mouse
model just based on gene expression. In further investigations, it will be of interest to further
evaluate PIKfyve’s involvement in CTSS secretion through either biochemical protein and kinase
activity assays and/or in vivo or in vitro inhibition studies.
CHAPTER 9
Conclusions
In summary, I found that a unique chemokine CX3CL1 is elevated in tears of SS model male
NOD mouse. Increased CX3CL1 level in tears is likely associated with both increased expression
and shedding induced by active CTSS and proinflammatory cytokines in both the LGAC and
corneal epithelium. Increased CX3CL1 further contributes to inflammation in the LG. The
elevation of CTSS trafficking in the diseased NOD mouse is associated with an alteration of
Rab3D and Rab27 isoforms. Through in vitro assays, we found that Rab27a is a main regulator for
CTSS secretion in the LGACs and is suspected to traffic through an endolysosomal pathway.
Investigations of upstream neural regulation in the LG found that there is an alternate protein
trafficking pathway stimulated by the α1-adrenergic agonist PE compared to CCh. PE significantly
111
induces CTSS secretion, which is suspected to be associated with an upregulated endolysosomal
pathway. PIKfyve is a key regulator for endolysosome maturation. In vitro pharmacological
inhibition of PIKfyve significantly reduced CTSS secretion, which was accompanied by decreased
Rab7 vesicle fusion, enlarged endolysosomes and a reduced apical accumulation of endolysosomal
markers at the APM.
CHAPTER 10
Material and methods
10.1 Reagents
Carbachol (CCh) (CAS No. 51832), doxycycline monohydrate (CAS no. 17086-28-1),
phenylephrine hydrochloride (CAS No.61767), Tris-Buffered Saline with 0.1% Tween
®
20
Detergent (TBST) and 4-methylumbelliferyl N-acetyl-β-d-glucosaminide (CAS no. 37067-30-4)
were from Sigma-Aldrich (St. Louis, MO, USA). Bovine serum albumin (#2905) was from
Calbiochem (Billerica, MA). The Cathepsin S activity kit (#K144) and rabbit polyclonal cathepsin
S antibody (#6686) was from Biovision Inc. (Milpitas, CA, USA). Hanks Balanced Salt Solution
(HBSS) without Mg
2+
and Ca
2+
was from Lonza Group Ltd. (Basel, Switzerland). Rabbit
polyclonal anti-Rab3D antibody was generated by Antibodies Inc. (Davis, CA, USA) as previously
reported
19
. Rabbit polyclonal anti-Rab27a antibody (#17817-1-AP) was purchased from
Proteintech (Rosemont, IL, USA). Mouse monoclonal anti-Rab27a antibody (#ab55667), rat
monoclonal anti-Lamp1 antibody (#ab25245), rabbit monoclonal anti-Vamp2
antibody(#ab181869) and rat monoclonal anti-Lamp2 antibody(#ab13524) were from Abcam
(Cambridge, MA, USA). Rabbit monoclonal Rab7 antibody was from Cell Signaling Technology
(Danvers, MA, USA). Rabbit polyclonal anti-mCFP antibody (#GTX59717) was from GeneTex
112
(Irvine, CA, USA). Rhodamine phalloidin, FITC anti-Rat, Alexa Fluor
â
488, 568, 680 secondary
antibodies and the ProLong Gold Antifade Mounting Medium were from Invitrogen (Grand Island,
NY, USA). Bio-Rad protein assay dye (#5000006) was from Bio-Rad (Hercules, CA, USA). BCA
reagents (#23235), CellMask deep red plasma membrane stain (#C10046), Fluo-4 AM dye
(F14201), 10% Triton X-100 (#NC0478124), 10% tris-glycine gel (#XP0010C), nitrocellulose
iBlot
TM
2 transfer stacks (#IB23001), TaqMan reverse transcription kit (#N8080234), and primers
to mouse Rab27a (#Mm00469997) and Gapdh (#Mm99999915_g1) were from Thermo Fisher
Scientific (Waltham, MA, USA). CloneAmp HiFi PCR Premix (#639298), Adeno-X Adenoviral
System 3 (Tet-On 3G Inducible) (#631180), NucleoSpin Gel and PCR Clean-Up kit (#740609),
In-Fusion
®
HD Cloning Kit (#638933), Stella
TM
competent cells (#63676), Adeno-X screening
primer mix 3 (#631030), Terra PCR direct red dye premix (#639286), CalPho mammalian
transfection kit (#631312), Adeno-X virus purification kit (#631533) and Adeno-X rapid titer kit
(#632250) were from Takara Bio Inc (Mountain View, CA, USA). All restriction enzymes used
were purchased from New England Bio Laboratories (Ipswich, MA, USA). Revert™ 700 total
protein staining kit (#926-11010) and IRDye
®
680RD Donkey anti-Rabbit IgG secondary antibody
(#926-68073), IRDye® 680RD Donkey anti-goat IgG secondary antibody (#926-68074) was from
LI-COR (Lincoln, NE, USA). The QIAquick gel extraction kit (#28704), RNeasy Plus Universal
Mini Kit (#73404) and Qiagen plasmid midi kit (#12145) were from Qiagen (Germantown, MD,
USA). Fluorescent blocking buffer was from Rockland Immunochemical Inc. (Limerick, PA,
USA). VWR Superfrost® Plus microslides were from VWR (Radnor, PA, USA).
10.2 Animals
BALB/c (000651) and NOD ShiLtJ (001976) mice were from Jackson Laboratories
(Sacramento, CA, USA). All mouse work used LG from male mice aged 16–17 weeks, when
113
lymphocytic infiltration of the LG as well as increased tear CTSS is established
56
. The non-obese
diabetic (NOD) mouse is an extensively-used murine model for the autoimmune exocrinopathy
characteristic of SS
55
. NOD mice spontaneously develop lymphocytic infiltration of the exocrine
glands, resulting in reduced secretory flow
55,56
. The male NOD mouse is more prone to develop
ocular manifestations of SS, and to exhibit profound LG infiltration as early as 8–10 weeks of age,
while females are more suspectable to infiltration of the SG from 14–16 week of age
2,55
. Mice
were anesthetized with a ketamine/xylazine combination at 80–100 mg/kg and 5–10 mg/kg dose,
respectively, through intraperitoneal injection. C57BL/6J (C57) mice were obtained from Jackson
Laboratories (Sacramento, CA). Rab3D deficient mice (3DKO) were generated as described and
bred in house. Experiments involving C57 and 3DKO mice used male and female mice aging from
12-18 weeks. To obtain in situ topically-stimulated glands, mouse LG were surgically-exposed
and stimulated topically with carbachol (CCh, 3 µL, 50 µM) for three times, 5 min each, after
anesthesia
2
. LG tissues were harvested from mice euthanized by cervical dislocation. Female New
Zealand White rabbits (2.1–2.5 kg) were obtained from Western Oregon Rabbit Company,
(Philomath, OR, USA). Rabbits were anesthetized through intramuscular injection of
ketamine/xylazine at 50 mg/kg and 5 mg/kg doses, respectively. After anesthesia, rabbits were
euthanized with intravenous ear injection of 1 mL Euthasol. LG were aseptically dissected from
euthanized rabbits and prepared as primary lacrimal acini cultures following published
procedures
19
. All animal procedures were conducted in accordance with the Guiding Principles for
the Care and Use of Laboratory Animals (8th edition)
127
and approved by the University of
Southern California’s Institutional Animal Care and Use Committee.
114
10.3 Tear collection
As previously described
2
, BALB/c, NOD, C57 and 3DKO mouse LGs were stimulated
topically with secretagogues including CCh (3 µL, 50 µM) OR PE (3 µL,100 µM). 2µL microcaps
pipettes are placed at the lateral canthus of the eye for tear collection. Each gland was stimulated
3 times, with collection time of 5mins each. Stimulated tear volume was recorded.
10.4 Tear CTSS activity measurement
Tear CTSS activity was measured with CTSS activity fluorometric assay kit. 200 µL of CTSS
reaction buffer was added to stimulated tears, vortexed and separated into 2 wells on a clear 96-
well plate (100 µL each). 2 µL CTSS inhibitor was added to one well, followed by the addition of
2 µL CTSS substrate to both wells. After 1-hour incubation at 37°C protected from light,
fluorescence was measured in measured in SpectraMax iD3 (Molecular Devices, San Jose, CA)
with 400/505-nm excitation/emission filters. Samples were then collected with total tear protein
was measured by Bio-Rad protein assay.
10.5 Tear β-hexosaminidase activity measurement
Tear β-hexosaminidase (β-hex) activity was measured by 4-methylumbelliferyl N-acetyl-β-
D-glucosaminide (4-MUF-NAG), a fluorogenic substrate for β-N-acetylhexosaminidase.
Collected tear samples was diluted with 220 µL of phosphate buffered saline (PBS) and added in
duplicates on a 96-well plate (60 µL per well), subsequently followed by 50 µL of 5 mM 4-MUF-
NAG in β-hex reaction buffer (133 mM sodium citrate, 133 mM sodium chloride, pH 4.3). Plate
was protected from light and incubated at room temperature for 2 hours. Enzymatic fluorescence
was measured by SpectraMax iD3 (Molecular Devices, San Jose, CA) with 360/465-nm
excitation/emission filters. Total tear protein was measured by Bio-Rad protein assay.
115
10.6 Mouse primary cell culture and IFN- γ treatment
As described,
60
LG isolated from 10 male BALB/c mice were pooled, washed with Ham’s
medium, and minced into 1 mm
3
pieces. Pieces were transferred into 150 mL Erlenmeyer flasks
and incubated with shaking in H-E medium and CHD medium at 37ºC. Supernatants were
collected after each incubation. Cells were pelleted by centrifugation and filtered through 100 μm
cell strainers. Filtrate was layered on 5% Ficoll and centrifuged at 300 × g for 10 min. Acinar cells
were pelleted, washed with Ham’s medium and resuspended in Matrigel-supplemented Peter’s
serum-free culture medium.
19
Cells were seeded at 2 × 10
6
cells per well in 12-well plates and
coverslip-coated 24-well plates. Cultured mouse LGACs were treated with 200 U/ml recombinant
mouse IFN- γ after 2 hr of seeding.
10.7 HCE-T cell culture and IFN- γ treatment
The SV-40 transformed human cornea epithelial cell line, HCE-T cells, was obtained from
the RIKEN Cell Bank, Japan (RCB2280) and cultured in complete KSFM supplemented with
human recombinant EGF, bovine pituitary extract and gentamycin. Cells were starved for 16-18
hr in supplement-free KSFM prior to treatments. Recombinant human IFN-γ (1 µg/mL) treatment
was in complete KSFM media.
10.8 CTSS treatment in cell culture
The activity of recombinant human CTSS (rhCTSS) used for treatments was equal to the
enzymatic activity level detected in the 90-95
th
percentile in SS patient tears.
57
CTSS activity was
tested prior to each treatment and maintained at 20,000 RFU per treatment.
116
10.9 Analysis of gene expression in cells and tissues
RNA from mouse primary LGACs and HCE-T cells was extracted using the RNeasy® Plus
Mini Kit. RNA from the corneal button and LG were extracted using RNeasy® Plus Universal
Mini Kit. The reverse transcription reaction used the reverse transcription kit, with 2 μg total RNA
from corneal buttons, mouse primary LGACs and HCE-T cells, and 4 μg total RNA from LG for
each 50 μL reaction. cDNA was obtained using GeneAmp® PCR System 9700 with incubation
cycles of 25°C (10 min), 48°C (30 min) and 95°C (5 min). Real-time qPCR was carried out with
the QuantStudio 12K Flex Real-Time PCR System with GAPDH as an internal control. Human
and mouse primers to CX3CL1, CX3CR1, CTSS, IFN-γ and TNF-α were used. Reaction
conditions and calculation methods were as previously described.
60
10.10 ELISA measurements of CX3CL1
For measurement of tear CX3CL1, tears collected from two mice were pooled and diluted as
above. For measurement of serum CX3CL1, blood was collected from mice via cardiac puncture
and centrifuged at 1000 × g, 10 min, 4°C for isolation of serum. For CX3CL1 measurements in
tissue, mouse corneal buttons and LG were collected, rinsed with PBS and homogenized in PBS
with a BeadBlaster 24 microtube homogenizer (Benchmark Scientific, Edison, NJ) at 7 M/S speed
for 1 min for 2 cycles at 30 sec intervals. Tissue lysates were collected and centrifuged at 5000 ×
g, 5 min, 4°C. For measurement of CX3CL1 in LGAC culture medium, medium was collected and
centrifuged at 1000 × g, 10 min, 4°C. For measurement of CX3CL1 in HCE-T cell culture medium,
medium was collected and centrifuged at 500 × g, 10 min, 4°C. Both clarified supernatants were
concentrated with protein centrifugal filters. For measurement of CX3CL1 in HCE-T cells, cells
were collected by scraping, pelleted and lysed in cell lysis buffer for 30 min on ice with constant
vortexing. Samples were spun down at 5,000 × g, 5 min, 4°C. Supernatants were concentrated with
117
centrifugal filters. CX3CL1 levels in all samples were measured using either the mouse or human
CX3CL1 Quantikine® kit as appropriate and according to the manufacturer’s protocol. Plates
were read using the SpectraMax iD3. Biorad or BCA assays were used for total protein in each
sample.
10.11 LG tissue processing and confocal fluorescence microscopy
As previously described
2
, LG from mice in which the LG was surgically exposed and treated
topically either without and with CCh or PE were retrieved and fixed in 4% paraformaldehyde and
4% sucrose PBS solution for 3 h at room temperature. After fixation, tissues were transferred to
30% sucrose PBS solution and left at 4 °C overnight. LGs were embedding in O.C.T compound
and frozen on dry ice. O.C.T blocks were sectioned at 5-µm thickness and mounted on glass micro
slides (VWR Superfrost
®
Plus microslides; VWR, Radnor, PA, USA). The cryosections were
quenched with NH4Cl (50 mM) in PBS for 5 min and permeabilized with 0.3% Triton X-100 for
30 min. Slides were then blocked with 5% BSA in 0.3% Triton X-100 for 3 h at room temperature.
The tissues were then incubated with primary antibody diluted in blocking buffer at 4 °C overnight.
On the second day, the slides were washed in PBS three times, 20 min each, followed by secondary
antibody incubation for 1 h at 37 °C. After three more thorough PBS washes, the samples were
fixed with ProLong anti-fade mounting medium and imaged the next day. Images were acquired
with a Zeiss LSM 800 with Airyscan. Z stack images with a total thickness of 4.96 µm, containing
30 slices at a 0.17-µm interval were acquired. Images were first assigned with random codes to
achieve a blinded analysis. Corrected total cell fluorescence quantification was performed using a
reported image processing pipeline using ImageJ version 2.1.0 (U. S. National Institutes of Health,
Bethesda, Maryland, USA) [47]. 3D % volume colocalization within regions of interest was
carried out using Bitplane Imaris software version 9.0.1 (Andor Technology Ltd., Belfast, UK)
118
according to the manufacturer’s instructions. The % volume colocalization was obtained by
quantifying the number of pixels containing both channels above threshold and normalizing to the
total number of pixels in the image.
10.12 Flow cytometry of lymphocytes infiltrating the LG
LG from individual mice were isolated and digested in multiple rounds of incubation in H-
E medium and CHD medium. Supernatants were collected and passed through a 70 µm cell strainer.
Cells were pelleted and resuspended in cell-staining media at a concentration of 5 to 10 × 10
6
cells/mL. Cells were treated with mouse TruStain FcX™ PLUS (1 µg per 10
6
cells) for 10 min to
block Fc-receptors. LIVE/DEAD® fixable aqua dead cell stain was added to the cell suspension
(1 µL per 10
6
cells). Fluorescently conjugated antibodies including anti-CD45, anti-CD11b, anti-
CD11c, anti-CD335, anti-CX3CR1, anti-F4/80 and anti-Ly6C were added to the cell suspension
at optimized concentrations and incubated on ice in the dark for 30 min. Cells were washed twice,
centrifuged at 350 × g for 5 min, and resuspended in cell-staining buffer for flow analysis on a BD
LSR Fortessa X20 (Franklin Lakes, NJ). Unstained live cells were used as controls to identify
positive staining.
10.13 Molecular cloning of Ad-mCFP-Rab27a constructs
mCFP-tagged Rab27a plasmid DNA was obtained by subcloning WT mouse Rab27a (678bp)
and the DN mutant (Rab27a
T23N
) into vector pCMV6-AN-mCFP (#6631, Takara Bio Inc.) at the
Hind III-Mlu I cloning site. Plasmid synthesis and subcloning were carried out by Genescript
(Piscataway, NJ) gene services. The plasmid insert mCFP-Rab27a was amplified through RT-PCR
with forward primer, 5′-GTAACTATAACGGTCATGAGCGGGGGCGAGGAG-3′ and reverse
primer 5′ATTACCTCTTTCTCCTTAACGCGTTCAACAGCCACAC-3′ using CloneAmp HiFi
PCR Premix according to the company’s instructions. Agarose gel analysis and the NucleoSpin
119
gel PCR clean up kit were used to analysis and purify the PCR inserts. The purified WT and DN
mCFP-Rab27a were then inserted into a linearized pAdenoX-Tet3G vector though in-fusion
cloning by incubation with 5´ in-fusion HD enzyme premix for 15 min at 50 °C. The recombinant
adenoviral constructs were transformed in Stella
TM
competent cells. 1.5 µL in-fusion reactions
were added to 100 µL Stella stock, incubated on ice for 30 min and heat-shocked for 45 s at 42 °C.
A total of 900 µL of the SOC medium was added to the cell mix and shaken at 37 °C for 1 h. 100
µL from each transformation was plated onto LB agar plates containing 100 µg/mL ampicillin.
Well-separated colonies were picked from the plates and screened by RT-PCR with Adeno-X
screening primer mix 3 and Terra PCR direct red dye premix. After the recommended number of
thermal cycles, 5 µL of each PCR reaction was analyzed by 1.2% agarose gel. The positive clones
were amplified by inoculating 100 mL of liquid LB/Ampicillin Medium (100 μg/mL ampicillin)
with 2–5 mL of fresh, log phase culture and shaken overnight at 37 °C at 300 rpm for 12–16 h.
The amplified bacteria were then lysed, and plasmid DNA purified with the Qiagen plasmid midi
kit according to kit instructions. Plasmid quality was evaluated through individual digestions with
Xho I and Nhe I and analysis on a 1.2% agarose gel. Bands containing mCFP-tagged WT and DN
Rab27a fragments were excised, and plasmid fragments extracted using the QIAquick Gel
Extraction Kit. The extracted fragments were sent to Retrogen Inc (San Diego, CA, USA). for
Sanger sequencing. After confirming the sequence for both mCFP-tagged WT and DN Rab27a,
the adenovirus plasmids were linearized by incubating with PacI restriction enzyme at 37 °C for 2
h to expose inverted terminal repeats. The digestion process was monitored using a 1.2% agarose
gel to monitor the migration of the resulting 3 kb fragment with completion. 60 μL 1 × TE Buffer
(pH 8.0) and 100 μL phenol: chloroform: isoamyl alcohol (25:24:1) were added to the linearized
plasmids with gentle vortexing. The mixtures were then centrifuged at 14,000 rpm for 5 min at
120
4 °C to separate phases. The top aqueous layer was removed and added to a mixture of 400 μL 95%
ethanol, 25 μL 10 M NH4OAc (or 1/10 volume 3 M NaOAc), and 1 μL glycogen (20 mg/mL). The
mixture was centrifuged at 14,000 rpm at 4 °C for 5 min, and the pellets collected, washed with
300 μL 70% ethanol, and redissolved in 10 μL sterile 1 × TE Buffer (pH 8.0).
10.14 Adenovirus production and amplification
QBI-HEK 293 cells (R70507) were obtained from Qbiogene, Adenovirus Technology
(Carlsbad, CA, USA) and cultured in high glucose DMEM supplemented with 10% FBS, 0.1 mM
MEM non-Essential amino acids (NEAA), 2 mM L-glutamine and 1% Pen-Strep. 293A cells were
plated on a 60 mm culture plate 24 h before transfection. Once cells reached 60% confluency,
PacI-digested adenovirus DNAs encoding WT and DN mCFP-Rab27a were added using the
CalPho mammalian transfection kit according to the manufacturer’s manual. The cells started to
display cytopathic effects after 2–3 days. After one week in culture, the cells were harvested and
lysed through three consecutive freeze-thaw cycles. The lysates were utilized to further infect a
fresh 60-mm culture plate of cells with cells collected, lysed after development of cytopathic
effects, and the supernatant used to infect larger cultures. Ad-transfected cells were harvested and
purified with the Adeno-X virus purification kit according to the kit instructions. Adenovirus titer
was measured by both plaque assay as previously reported
19
and with the Adeno-X rapid titer kit.
10.15 Rabbit LGAC preparation and transfection
Isolation of rabbit LGACs was as previously described
19,25
. At day 1 after isolation, LGACs
were transduced with high titer adenovirus. Cells were incubated with mCFP-tagged WT or DN
Rab27a adenovirus for 4 h at a multiplicity of infection (MOI) of 4–6. Medium containing the Ad
was removed and replaced with fresh Peter’s complete medium (PCM) containing 1 µg/mL
doxycycline to induce protein expression. Cells were cultured for another 18–20 h for optimal
121
protein expression. LGAC were kept in culture for a total of two days prior to functional secretion
and imaging experiments. During culture, cells aggregate into acinus-like structures, where distinct
apical and basolateral domains are presented, mimicking the structures and functions of acini in
vivo
19,25
.
10.16 Live Cell Imaging of Rabbit LGACs
LGACs were seeded in 35-mm glass bottomed petri dishes coated with Matrigel at a cell
density of 6 ´ 10
6
cells per dish. After transduction with WT mCFP-Rab27a and incubation with
PCM containing 1 µg/mL doxycycline, the medium was gently removed and replaced with Mg
2+
-
free HBSS. During time lapse imaging, images of single cell focal planes were acquired for 15
min at a fixed time interval of 10 s. Resting state videos were recorded by imaging cells in a 37 °C
imaging chamber for 15 min. After acquisition of these videos, 100 μM CCh in Mg
2+
-free HBSS
was added to the cells to stimulated regulated exocytosis and imaged for another 15 min. Similarly,
the same volume of Mg
2+
-free HBSS was added to the cells and used as an unstimulated blank
control. Vesicle sizes and counts were quantified by Image J. Images were acquired with a Zeiss
LSM880 using the 458 nm laser line.
10.17 In vitro secretion of primary rabbit LGAC
As previously described
19
, LGACs were seeded in 12-well plates coated with Matrigel at a
cell density of 2 ´ 10
6
per well. Cells were transduced as described above. Prior to secretion assays,
cells were pre-incubated in Mg
2+
-free HBSS at 37 °C for 1 h and the medium collected as the basal
secretion control. Cells were then treated with CCh (100 μM) in Mg
2+
-free HBSS at 37 °C for 30
min to stimulate secretion. The same volume of medium was added into the cells under the same
conditions for the untreated controls. The cell culture medium was collected post-incubation and
122
the cells lysed with NaOH. CTSS activity in the cell culture medium was measured using the CTSS
activity fluorometric assay kit according to the manufacturer’s instructions at 37 °C for 2 h.
Fluorescence was measured using 400/505 nm excitation/emission filters. β-hex activity was
measured as previously described
19
, with incubation of the reaction mixture at room temperature
for 2 h and measurement of signal using 365/460-nm excitation/emission filters in a SpectraMax
iD3 (Molecular Devices, San Jose, CA, USA). Total protein in culture medium was measured by
the Bio-Rad assay protein assay. The total protein in cell lysates was measured using the Pierce
BCA Protein Assay (Thermo Fisher Scientific, Waltham, MA, USA, #23225) as previously
reported. Activities of CTSS and β-hex, together with media protein, were normalized to total cell
protein in each well, as relative fluorescence per microgram protein.
10.18 Western blotting for tear CTSS
Mouse tears were collected and diluted with 50 µL PBS solution. Sample concentration was
measured by BioRad assay. Tear samples were left in β-mercaptoethanol reducing dye for 1 hour
at room temperature. 10 ug of sample were loaded into each well of a precast 10% PAGEr™ EX
gel and ran under constant voltage 80 V at 4 °C for 2 hours. The proteins were then transferred to
a nitrocellulose membrane with iBlot 2 dry blotting system (Thermo Fisher Scientific, Waltham,
MA, USA). The membrane was first blocked with fluorescent blocking buffer for 1 hour at room
temperature and incubated overnight in primary CTSS antibody (1:1000 dilution in blocking buffer)
at 4 °C. For secondary control, membrane was incubated with blocking buffer only overnight. On
the second day, the membrane was washed by 1X Tris-Buffered Saline with 0.1% Tween
®
20
Detergent (TBST) for 3 times, 15 minutes each. The membrane was then incubated in secondary
antibody (1:2000 dilution) containing blocking buffer for 1 hour at room temperature, followed by
123
three more TBST washes. After the final wash, the membrane was imaged with Odyssey Licor
imaging system (LI-COR Biotechnology, Lincoln, NE, USA).
10.19 Primary C57 LGAC preparation and secretion assay
As previously described
1
, LGs from 15-18 week old C57 mice were polled and washed twice
with Ham’s medium. The glands were minced into 1 mm
3
pieces and transferred into a 150 mL
Erlenmeyer flask, where they were alternately incubated and shake in CHD medium (250 U/mL)
and H-E medium (250 U/mL). Supernatants from each incubation were collected with cells
pelleted through centrifugation. The cells acquired were filtered through a 100 μm cell strainer.
The filtrate was layered on top of 5% Ficoll and centrifuged at 300 × g for 10 min. Acinar cells
were pelleted, washed with Ham’s medium and eventually resuspended in matrigel-supplemented
Peter’s serum-free culture medium. Cells were seeded at 2 × 10
6
cells/well in 12-well plates,
or 1× 10
6
cells/well in 24-well plates with coverslip. Cultured LGACs were incubated for 2 hours
prior to experiments. After removing cell culture medium, cells were gently washed with Mg
2+
-
free HBSS (Mg
2+
and Ca
2+
free HBSS supplied with 1.265 mM CaCl2). Cells were pre-incubated
in Mg
2+
-free HBSS media for 1 hour and stimulated with or without CCh (100 μM) and PE (100
μM) in Mg
2+
-free HBSS media for 30 minutes. Cell culture medium was collected after incubation.
The cells were then washed three times with PBS and lysed with RIPA buffer. CTSS and β-hex
activity was measured in the culture medium with above mentioned methods. Total protein level
in medium was measured by BioRad assay. Cell lysate protein was measured with BCA assay.
Enzymatic activity levels and medium protein levels were normalized to cell lysate. Increase in
activity and medium protein was calculated by subtracting measurements from preincubation
groups, as previously described.
124
10.20 Immunofluorescence of primary LGACs
Cultured LGACs were wash twice with PBS after gently removing culture medium. Cells
were fixed and permeabilized with ice-cold methanol and acetone (1:1), followed by two PBS
washes, 5 minutes each. The cells were then blocked with 1% BSA at room temperature for 1 hour,
subsequently followed by primary antibody and secondary antibody incubation, each for 1 hour at
37°C. Three 5-minute PBS washes were applied after each antibody incubation. After the final
wash cells were fixed with ProLong anti-fade mounting medium and imaged by Zeiss LSM 800
with Airyscan.
10.21Fluorescence Imaging of Intracellular Calcium in mouse LGACs
After removing cell culture medium, cultured mouse LGACs were gently rinsed with sterile
PBS twice. Cell were incubated in Fluo-4 AM working solution (2.5 µM in Matrigel supplemented
PCM medium) for 20 minutes at room temperature protected from light. Cells were then rinse
twice in NaCl-Ringer’s solution and left in NaCl-Ringer’s solution for 30 minutes at room
temperature. Complete Ringer’s solution was replaced with 1.18 mL Ca
2+
-free NaCl-Ringer’s
solution and image with Zeiss LSM 800. Region of interests (ROIs) were selected with stable
baseline record for 2-3 minutes. The cells were then stimulated with CCh (100 µM) and PE (100
µM) by carefully adding dropwise CCh (20 µL, 6mM in Ca
2+
-free NaCl-Ringer) and PE (20 µL,
6mM in Ca
2+
-free NaCl-Ringer’s) stock solution. Addition of 20 µL Ca
2+
-free NaCl-Ringer’s was
used as spiking control. Mean fluorescence intensity of ROIs was recorded for 15 minutes. As
described, fluorescence change was analyzed and defined as delta function, ΔF/F(t)=(F0-F(t))/F0,
F0 is the average fluorescence intensity of the ROI baseline, F(t) is the fluorescence at a given time.
125
10.22 Apilimod treatment in rabbit LGACs for secretion assay
Cells were seeded at 2 × 10
6
cells/well in 12-well plates. On day 2 of the rabbit cell culture,
PCM culture media was carefully removed. Cell were washed twice with Mg
2+
-free HBSS media
and incubated in Mg
2+
-free HBSS media for 30 minutes at 37°C. For the baseline group, Mg
2+
-
free HBSS media was added at 270 µL/ well; for untreated and stimulation groups, 243 µL/ well
was added. Apilimod 3µM stock was prepared by diluting 100mM apilimod/DMSO stock in Mg
2+
-
free HBSS media, a same volume of DMSO was prepared in a same manner and used as control.
At the 30-minute timepoint, 30 µL apilimod (3µM) was added to baseline group and 27 µL was
added into the untreated and treatment groups, making the final cell treatment concentration 300
nM apilimod. Cells were further incubated for 1 hour. After preincubation, baseline media were
collected, and 30 µL media and 30 µL (1mM) CCh was added to the media and incubated for 30
minutes. Media were collected and analyzed for CTSS and β-hex activity. Cells were lysed in
NaOH and analyzed with BCA assay.
10.23 Rab7-GFP BacMam virus transduction and imaging in rabbit LGACs
Rabbit LGACs were seeded in u-slide 8-well imaging chamber at a density of 2.5x10
5
cells/well in 260 µL PCM media. 40 µL BacMan Rab7-GFP were added to the cells and incubated
overnight in 37°C. The transduction media was discarded the next day and cells were gently
washed for 3 times in cold PBS. CellMask™ plasma membrane stain was prepared in PCM
according to company instructions and incubated with cells for 15 min at 37°C. Cells were further
washed with Mg
2+
-free HBSS media for three times with 270 µL of the media left in the cells after
the final wash. Time series images of the cells were taken for 15 mins with 10 s intervals. Cells
were stimulated by adding 30 µL 1 mM CCh, to make a final concentration of 100 µM CCh.
126
10.24 Apilimod treatment in mouse tissue
LG tissue from C57 and Rab3DKO were harvested cut in half and incubated in 270 µL 300nM
apilimod in PCM for 1 hour at 37°C. 30 µL CCh (1mM) and PE (1mM) were added to wells and
incubated for 30 minutes at 37°C. After treatment, LGs were pat dry on a Kimwipe and fixed in
4% PFA with 4% sucrose in PBS for 3 hours at room temperature. The tissue was transferred to
30% sucrose overnight at 4°C. On the next day, the tissue was embedded into OCT frozen blocks,
and sectioned and stained according to above.
10.25 Statistics
All statistical analyses were performed using Graphpad Prism 8.3.0 software (San Diego, CA,
USA). Data normality was assessed by Kolmogorov-Smirnov, D’Agostino and Person omnibus,
and Shapiro–Wilk normality tests. A two-tailed unpaired Student’s t-test was used to compare
between two treatment groups. A one-way ANOVA with Dunnett’s comparison was used to
compare multiple treatments to the untreated control group. A one-way ANOVA with Tukey’s
multiple comparison was used to compare effects among multiple treatments when comparing
more than two groups. A two-way ANOVA with Sidak’s multiple comparison test were used to
compare different treatment effects between two groups. The criterion for statistical significance
was set at p ≤ 0.05. In all graphs, ns, p > 0.05; * p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤
0.0001.
127
REFERENCES
1 Fu, R., Edman, M. C. & Hamm-Alvarez, S. F. Rab27a Contributes to Cathepsin S Secretion
in Lacrimal Gland Acinar Cells. International Journal of Molecular Sciences 22,
doi:10.3390/ijms22041630 (2021).
2 Fu, R. et al. Cathepsin S activation contributes to elevated CX3CL1 (fractalkine) levels in
tears of a Sjögren’s syndrome murine model. Sci Rep 10, 1455, doi:10.1038/s41598-020-
58337-4 (2020).
3 Fu, R., Klinngam, W., Heur, M., Edman, M. C. & Hamm-Alvarez, S. F. Tear Proteases
and Protease Inhibitors: Potential Biomarkers and Disease Drivers in Ocular Surface
Disease. Eye Contact Lens 46 Suppl 2, S70-s83, doi:10.1097/icl.0000000000000641
(2020).
4 Edman, M. C., Marchelletta, R. R. & Hamm-Alvarez, S. F. in Encyclopedia of the Eye (ed
Darlene A. Dartt) 522-527 (Academic Press, 2010).
5 Dartt, D. A. Neural regulation of lacrimal gland secretory processes: relevance in dry eye
diseases. Progress in Retinal and Eye Research 28, 155-177,
doi:10.1016/j.preteyeres.2009.04.003 (2009).
6 Dartt, D. A. Dysfunctional Neural Regulation of Lacrimal Gland Secretion and its Role in
the Pathogenesis of Dry Eye Syndromes. The Ocular Surface 2, 76-91,
doi:https://doi.org/10.1016/S1542-0124(12)70146-5 (2004).
7 Meneray, M. A., Fields, T. Y. & Bennett, D. J. Gs and Gq/11 couple vasoactive intestinal
peptide and cholinergic stimulation to lacrimal secretion. Invest Ophthalmol Vis Sci 38,
1261-1270 (1997).
8 Mauduit, P., Jammes, H. & Rossignol, B. M3 muscarinic acetylcholine receptor coupling
to PLC in rat exorbital lacrimal acinar cells. Am J Physiol 264, C1550-1560,
doi:10.1152/ajpcell.1993.264.6.C1550 (1993).
9 Zoukhri, D., Hodges, R. R., Sergheraert, C., Toker, A. & Dartt, D. A. Lacrimal gland PKC
isoforms are differentially involved in agonist-induced protein secretion. Am J Physiol 272,
C263-269, doi:10.1152/ajpcell.1997.272.1.C263 (1997).
10 Mauduit, P., Herman, G. & Rossignol, B. Protein secretion in lacrimal gland: alpha 1-beta-
adrenergic synergism. Am J Physiol 250, C704-712, doi:10.1152/ajpcell.1986.250.5.C704
(1986).
11 Hodges, R. R. et al. Nitric Oxide and cGMP Mediate α1D-Adrenergic Receptor–
Stimulated Protein Secretion and p42/p44 MAPK Activation in Rat Lacrimal Gland. Invest.
Ophthalmol. Vis. Sci. 46, 2781-2789, doi:10.1167/iovs.05-0022 (2005).
128
12 Gromada, J., Jørgensen, T. D. & Dissing, S. The release of intracellular Ca2+ in lacrimal
acinar cells by alpha-, beta-adrenergic and muscarinic cholinergic stimulation: the roles of
inositol triphosphate and cyclic ADP-ribose. Pflugers Arch 429, 751-761,
doi:10.1007/bf00374798 (1995).
13 Wu, K. et al. Molecular mechanisms of lacrimal acinar secretory vesicle exocytosis.
Experimental Eye Research 83, 84-96, doi:https://doi.org/10.1016/j.exer.2005.11.009
(2006).
14 Deneka, M., Neeft, M. & Sluijs, P. v. d. Regulation of Membrane Transport by rab
GTPases. Critical Reviews in Biochemistry and Molecular Biology 38, 121-142,
doi:10.1080/713609214 (2003).
15 Deacon, S. W. & Gelfand, V. I. Of Yeast, Mice, and Men: Rab Proteins and Organelle
Transport. Journal of Cell Biology 152, F21-F24, doi:10.1083/jcb.152.4.F21 (2001).
16 Izumi, T., Gomi, H., Kasai, K., Mizutani, S. & Torii, S. The Roles of Rab27 and Its
Effectors in the Regulated Secretory Pathways. Cell Structure and Function 28, 465-474,
doi:10.1247/csf.28.465 (2003).
17 Schlüter, O. M., Khvotchev, M., Jahn, R. & Südhof, T. C. Localization Versus Function of
Rab3 Proteins: EVIDENCE FOR A COMMON REGULATORY ROLE IN
CONTROLLING FUSION. J. Biol. Chem. 277, 40919-40929,
doi:10.1074/jbc.M203704200 (2002).
18 Riedel, D. et al. Rab3D Is Not Required for Exocrine Exocytosis but for Maintenance of
Normally Sized Secretory Granules. Mol Cell Biol 22, 6487-6497,
doi:10.1128/mcb.22.18.6487-6497.2002 (2002).
19 Meng, Z. et al. Imbalanced Rab3D versus Rab27 increases cathepsin S secretion from
lacrimal acini in a mouse model of Sjögren's Syndrome. Am J Physiol Cell Physiol 310,
C942-C954, doi:10.1152/ajpcell.00275.2015 (2016).
20 Chen, D., Guo, J., Miki, T., Tachibana, M. & Gahl, W. A. Molecular Cloning and
Characterization of Rab27a and Rab27b, Novel Human Rab Proteins Shared by
Melanocytes and Platelets. Biochemical and Molecular Medicine 60, 27-37,
doi:https://doi.org/10.1006/bmme.1996.2559 (1997).
21 Chen, X. et al. Rab27b localizes to zymogen granules and regulates pancreatic acinar
exocytosis. Biochemical and Biophysical Research Communications 323, 1157-1162,
doi:https://doi.org/10.1016/j.bbrc.2004.08.212 (2004).
22 Hou, Y., Ernst, S. A., Stuenkel, E. L., Lentz, S. I. & Williams, J. A. Rab27A Is Present in
Mouse Pancreatic Acinar Cells and Is Required for Digestive Enzyme Secretion. PLOS
ONE 10, e0125596, doi:10.1371/journal.pone.0125596 (2015).
129
23 Yi, Z. et al. The Rab27a/Granuphilin Complex Regulates the Exocytosis of Insulin-
Containing Dense-Core Granules. Mol Cell Biol 22, 1858, doi:10.1128/MCB.22.6.1858-
1867.2002 (2002).
24 Imai, A., Yoshie, S., Nashida, T., Shimomura, H. & Fukuda, M. The small GTPase Rab27B
regulates amylase release from rat parotid acinar cells. J. Cell. Sci. 117, 1945,
doi:10.1242/jcs.01048 (2004).
25 Chiang, L. et al. Rab27b regulates exocytosis of secretory vesicles in acinar epithelial cells
from the lacrimal gland. American Journal of Physiology-Cell Physiology 301, C507-C521,
doi:10.1152/ajpcell.00355.2010 (2011).
26 Tolmachova, T. et al. A General Role for Rab27a in Secretory Cells. Mol Biol Cell 15,
332-344, doi:10.1091/mbc.e03-07-0452 (2003).
27 Pfeffer, S. R. Two Rabs for exosome release. Nature Cell Biology 12, 3-4,
doi:10.1038/ncb0110-3 (2010).
28 Ostrowski, M. et al. Rab27a and Rab27b control different steps of the exosome secretion
pathway. Nature Cell Biology 12, 19-30, doi:10.1038/ncb2000 (2010).
29 van der Sluijs, P., Zibouche, M. & van Kerkhof, P. Late Steps in Secretory Lysosome
Exocytosis in Cytotoxic Lymphocytes. Front Immunol 4, doi:10.3389/fimmu.2013.00359
(2013).
30 Neeft, M. et al. Munc13-4 Is an Effector of Rab27a and Controls Secretion of Lysosomes
in Hematopoietic Cells. Mol Biol Cell 16, 731-741, doi:10.1091/mbc.e04-10-0923 (2004).
31 Levin, R., Grinstein, S. & Schlam, D. Phosphoinositides in phagocytosis and
macropinocytosis. Biochimica et Biophysica Acta (BBA) - Molecular and Cell Biology of
Lipids 1851, 805-823, doi:https://doi.org/10.1016/j.bbalip.2014.09.005 (2015).
32 McCartney, A. J., Zhang, Y. & Weisman, L. S. Phosphatidylinositol 3,5-bisphosphate: low
abundance, high significance. Bioessays 36, 52-64, doi:10.1002/bies.201300012 (2014).
33 Dove, S. K., Dong, K., Kobayashi, T., Williams, F. K. & Michell, R. H.
Phosphatidylinositol 3,5-bisphosphate and Fab1p/PIKfyve underPPIn endo-lysosome
function. Biochem J 419, 1-13, doi:10.1042/bj20081950 (2009).
34 Dong, X.-p. et al. PI(3,5)P(2) controls membrane trafficking by direct activation of
mucolipin Ca(2+) release channels in the endolysosome. Nat Commun 1, 38-38,
doi:10.1038/ncomms1037 (2010).
35 Li, S. C. et al. The signaling lipid PI(3,5)P₂ stabilizes V₁-V(o) sector interactions and
activates the V-ATPase. Mol Biol Cell 25, 1251-1262, doi:10.1091/mbc.E13-10-0563
(2014).
130
36 Min, S. H. et al. Loss of PIKfyve in platelets causes a lysosomal disease leading to
inflammation and thrombosis in mice. Nat Commun 5, 4691, doi:10.1038/ncomms5691
(2014).
37 Choy, C. H. et al. Lysosome enlargement during inhibition of the lipid kinase PIKfyve
proceeds through lysosome coalescence. J. Cell. Sci. 131, jcs213587,
doi:10.1242/jcs.213587 (2018).
38 Lemp, M. A. The definition and classification of dry eye disease: Report of the definition
and classfication subcommittee of the International Dry Eye Workshop (2007). Ocul Surf
5, 75-92, doi:10.1016/S1542-0124(12)70081-2 (2007).
39 Dartt, D. A. Neural regulation of lacrimal gland secretory processes: relevance in dry eye
diseases. Prog Retin Eye Res 28, 155-177, doi:10.1016/j.preteyeres.2009.04.003 (2009).
40 Stern, M. E. & Pflugfelder, S. C. Inflammation in dry eye. Ocul Surf 2, 124-130,
doi:10.1016/s1542-0124(12)70148-9 (2004).
41 Pescosolido, N. et al. Role of Protease-Inhibitors in Ocular Diseases. Molecules 19, 20557-
20569, doi:10.3390/molecules191220557 (2014).
42 Brito-Zeron, P. et al. Systemic activity and mortality in primary Sjogren syndrome:
predicting survival using the EULAR-SS Disease Activity Index (ESSDAI) in 1045
patients. Ann Rheum Dis 75, 348-355, doi:10.1136/annrheumdis-2014-206418 (2016).
43 Brito-Zeron, P. et al. Sjogren syndrome. Nat Rev Dis Primers 2, 16047,
doi:10.1038/nrdp.2016.47 (2016).
44 Foulks, G. N. & Bron, A. J. Meibomian Gland Dysfunction: A Clinical Scheme for
Description, Diagnosis, Classification, and Grading. The Ocular Surface 1, 107-126,
doi:10.1016/s1542-0124(12)70139-8 (2003).
45 Tong, L., Zhou, L., Beuerman, R. W., Zhao, S. Z. & Li, X. R. Association of tear proteins
with Meibomian gland disease and dry eye symptoms. Br J Ophthalmol 95, 848-852,
doi:10.1136/bjo.2010.185256 (2011).
46 Soria, J. et al. Tear proteome and protein network analyses reveal a novel pentamarker
panel for tear film characterization in dry eye and meibomian gland dysfunction. J
Proteomics 78, 94-112, doi:10.1016/j.jprot.2012.11.017 (2013).
47 Bartlett, J. D., Keith, M. S., Sudharshan, L. & Snedecor, S. J. Associations between signs
and symptoms of dry eye disease: a systematic review. Clin Ophthalmol 9, 1719-1730,
doi:10.2147/OPTH.S89700 (2015).
48 D'Souza, S. & Tong, L. Practical issues concerning tear protein assays in dry eye. Eye and
Vision 1, 1-12, doi:10.1186/s40662-014-0006-y (2014).
131
49 Brito-Zerón, P. et al. Sjögren syndrome. Nature Reviews Disease Primers 2, 16047,
doi:10.1038/nrdp.2016.47 (2016).
50 Hyon, J. Y., Lee, Y. J. & Yun, P.-Y. Management of Ocular Surface Inflammation in
Sjögren Syndrome. Cornea 26, S13-S15, doi:10.1097/ICO.0b013e31812f6782 (2007).
51 Stern, M. E. & Pflugfelder, S. C. Inflammation in Dry Eye. The Ocular Surface 2, 124-130,
doi:https://doi.org/10.1016/S1542-0124(12)70148-9 (2004).
52 Pflugfelder, S. C. & de Paiva, C. S. The Pathophysiology of Dry Eye Disease: What We
Know and Future Directions for Research. Ophthalmology 124, S4-S13,
doi:https://doi.org/10.1016/j.ophtha.2017.07.010 (2017).
53 Byun, Y.-S., Lee, H. J., Shin, S. & Chung, S.-H. Elevation of autophagy markers in Sjögren
syndrome dry eye. Sci Rep 7, 17280, doi:10.1038/s41598-017-17128-0 (2017).
54 Foulks, G. N. et al. Clinical Guidelines for Management of Dry Eye Associated with
Sjögren Disease. The Ocular Surface 13, 118-132,
doi:https://doi.org/10.1016/j.jtos.2014.12.001 (2015).
55 Ju, Y. et al. NOD and NOR mice exhibit comparable development of lacrimal gland
secretory dysfunction but NOD mice have more severe autoimmune dacryoadenitis.
Experimental Eye Research 176, 243-251, doi:https://doi.org/10.1016/j.exer.2018.09.002
(2018).
56 Janga, S. R. et al. Longitudinal analysis of tear cathepsin S activity levels in male non-
obese diabetic mice suggests its potential as an early stage biomarker of Sjögren’s
Syndrome. Biomarkers 24, 91-102, doi:10.1080/1354750X.2018.1514656 (2019).
57 Hamm-Alvarez, S. F. et al. Tear Cathepsin S as a Candidate Biomarker for Sjögren's
Syndrome. Arthritis & Rheumatology 66, 1872-1881, doi:10.1002/art.38633 (2014).
58 Bahamondes, V. et al. Changes in Rab3D expression and distribution in the acini of
Sjögren's syndrome patients are associated with loss of cell polarity and secretory
dysfunction. Arthritis & Rheumatism 63, 3126-3135, doi:10.1002/art.30500 (2011).
59 Nashida, T., Yoshie, S., Haga-Tsujimura, M., Imai, A. & Shimomura, H. Atrophy of
myoepithelial cells in parotid glands of diabetic mice; detection using skeletal muscle actin,
a novel marker. FEBS Open Bio 3, 130-134, doi:10.1016/j.fob.2013.01.009 (2013).
60 Meng, Z., Klinngam, W., Edman, M. C. & Hamm-Alvarez, S. F. Interferon-γ treatment in
vitro elicits some of the changes in cathepsin S and antigen presentation characteristic of
lacrimal glands and corneas from the NOD mouse model of Sjögren’s Syndrome. PLOS
ONE 12, e0184781, doi:10.1371/journal.pone.0184781 (2017).
61 Li, X. et al. Increased Expression of Cathepsins and Obesity-Induced Proinflammatory
Cytokines in Lacrimal Glands of Male NOD Mouse. Invest. Ophthalmol. Vis. Sci. 51, 5019-
5029, doi:10.1167/iovs.09-4523 (2010).
132
62 Shi, G.-P. et al. Cathepsin S Required for Normal MHC Class II Peptide Loading and
Germinal Center Development. Immunity 10, 197-206, doi:https://doi.org/10.1016/S1074-
7613(00)80020-5 (1999).
63 Kirschke, H. in Handbook of Proteolytic Enzymes (Third Edition) (eds Neil D. Rawlings
& Guy Salvesen) 1824-1830 (Academic Press, 2013).
64 Wilkinson Richard, D. A., Williams, R., Scott Christopher, J. & Burden Roberta, E. in
Biological Chemistry Vol. 396 867 (2015).
65 Klinngam, W., Fu, R., Janga, S. R., Edman, M. C. & Hamm-Alvarez, S. F. Cathepsin S
Alters the Expression of Pro-Inflammatory Cytokines and MMP-9, Partially through
Protease—Activated Receptor-2, in Human Corneal Epithelial Cells. International Journal
of Molecular Sciences 19, 3530 (2018).
66 Bazan, J. F. et al. A new class of membrane-bound chemokine with a CX3C motif. Nature
385, 640-644, doi:10.1038/385640a0 (1997).
67 Fong, A. M. et al. Ultrastructure and Function of the Fractalkine Mucin Domain in CX3C
Chemokine Domain Presentation. J. Biol. Chem. 275, 3781-3786,
doi:10.1074/jbc.275.6.3781 (2000).
68 Hundhausen, C. et al. The disintegrin-like metalloproteinase ADAM10 is involved in
constitutive cleavage of CX3CL1 (fractalkine) and regulates CX3CL1-mediated cell-cell
adhesion. Blood 102, 1186-1195, doi:10.1182/blood-2002-12-3775 (2003).
69 Garton, K. J. et al. TACE (ADAM17) mediates the cleavage and shedding of Fractalkine
(CX3CL1). J. Biol. Chem., doi:10.1074/jbc.M106434200 (2001).
70 Wildenberg, M. E., van Helden-Meeuwsen, C. G., Drexhage, H. A. & Versnel, M. A.
Altered fractalkine cleavage potentially promotes local inflammation in NOD salivary
gland. Arthritis Research & Therapy 10, R69, doi:10.1186/ar2441 (2008).
71 Clark, A. K., Yip, P. K. & Malcangio, M. The Liberation of Fractalkine in the Dorsal Horn
Requires Microglial Cathepsin S. The Journal of Neuroscience 29, 6945,
doi:10.1523/JNEUROSCI.0828-09.2009 (2009).
72 Lee, M., Lee, Y., Song, J., Lee, J. & Chang, S.-Y. Tissue-specific Role of CX(3)CR1
Expressing Immune Cells and Their Relationships with Human Disease. Immune Netw 18,
e5-e5, doi:10.4110/in.2018.18.e5 (2018).
73 Cambien, B. et al. Signal transduction pathways involved in soluble fractalkine–induced
monocytic cell adhesion. Blood 97, 2031-2037, doi:10.1182/blood.V97.7.2031 (2001).
74 Fong, A. M. et al. Fractalkine and CX3CR1 Mediate a Novel Mechanism of Leukocyte
Capture, Firm Adhesion, and Activation under Physiologic Flow. The Journal of
Experimental Medicine 188, 1413-1419, doi:10.1084/jem.188.8.1413 (1998).
133
75 Yajima, N. et al. Elevated levels of soluble fractalkine in active systemic lupus
erythematosus: Potential involvement in neuropsychiatric manifestations. Arthritis &
Rheumatism 52, 1670-1675, doi:10.1002/art.21042 (2005).
76 Nanki, T., Imai, T. & Kawai, S. Fractalkine/CX3CL1 in rheumatoid arthritis. Modern
Rheumatology 27, 392-397, doi:10.1080/14397595.2016.1213481 (2017).
77 Lee, J. H. et al. Role of Fractalkine in the Pathogenesis of Primary Sjögren Syndrome:
Increased Serum Levels of Fractalkine, Its Expression in Labial Salivary Glands, and the
Association with Clinical Manifestations. The Journal of Rheumatology 41, 2425,
doi:10.3899/jrheum.130892 (2014).
78 Tsubota, K. et al. The Role of Fractalkine as an Accelerating Factor on the Autoimmune
Exocrinopathy in Mice. Invest. Ophthalmol. Vis. Sci. 50, 4753-4760, doi:10.1167/iovs.08-
2596 (2009).
79 Yoon, K.-C., Jeong, I.-Y., Park, Y.-G. & Yang, S.-Y. Interleukin-6 and Tumor Necrosis
Factor-α Levels in Tears of Patients With Dry Eye Syndrome. Cornea 26, 431-437,
doi:10.1097/ICO.0b013e31803dcda2 (2007).
80 Solomon, A. et al. Pro- and Anti-inflammatory Forms of Interleukin-1 in the Tear Fluid
and Conjunctiva of Patients with Dry-Eye Disease. Invest. Ophthalmol. Vis. Sci. 42, 2283-
2292 (2001).
81 Liu, R. et al. Analysis of Th17-associated cytokines and clinical correlations in patients
with dry eye disease. PLOS ONE 12, e0173301, doi:10.1371/journal.pone.0173301 (2017).
82 Chen, X. et al. Elevated cytokine levels in tears and saliva of patients with primary
Sjögren’s syndrome correlate with clinical ocular and oral manifestations. Sci Rep 9, 7319,
doi:10.1038/s41598-019-43714-5 (2019).
83 Matsumiya, T. et al. Characterization of Synergistic Induction of CX3CL1/Fractalkine by
TNF-α and IFN-γ in Vascular Endothelial Cells: An Essential Role for TNF-α in Post-
Transcriptional Regulation of CX3CL1. The Journal of Immunology 184, 4205,
doi:10.4049/jimmunol.0903212 (2010).
84 Klinngam, W. et al. Inhibition of Cathepsin S Reduces Lacrimal Gland Inflammation and
Increases Tear Flow in a Mouse Model of Sjögren’s Syndrome. Sci Rep 9, 9559,
doi:10.1038/s41598-019-45966-7 (2019).
85 Gavet, O. & Pines, J. Progressive activation of CyclinB1-Cdk1 coordinates entry to mitosis.
Dev Cell 18, 533-543, doi:10.1016/j.devcel.2010.02.013 (2010).
86 Clark, A. K. et al. Inhibition of spinal microglial cathepsin S for the reversal of neuropathic
pain. Proceedings of the National Academy of Sciences 104, 10655-10660,
doi:10.1073/pnas.0610811104 (2007).
134
87 Tyrpak, D. Corrected-Total-Cell-Fluorescence. doi:https://10.5281/zenodo.3247836
(2019).
88 Tyrpak, D., Li, Y., Lei, S. & Mackay, J. A. SIAL: A simple image analysis library for wet-
lab scientists. The journal of open source software 5(56), 2689 (2020).
89 Imai, A. & Tsujimura, M. The small GTPase, Rab27, and its effectors and regulators
participate in granule exocytosis by parotid acinar cells. Journal of Oral Biosciences 59,
12-16, doi:https://doi.org/10.1016/j.job.2016.10.003 (2017).
90 Hammel, I., Lagunoff, D. & Galli, S. J. Regulation of secretory granule size by the precise
generation and fusion of unit granules. Journal of Cellular and Molecular Medicine 14,
1904-1916, doi:https://doi.org/10.1111/j.1582-4934.2010.01071.x (2010).
91 Yamaoka, M., Ishizaki, T. & Kimura, T. GTP- and GDP-Dependent Rab27a Effectors in
Pancreatic Beta-Cells. Biological and Pharmaceutical Bulletin 38, 663-668,
doi:10.1248/bpb.b14-00886 (2015).
92 Eskelinen, E.-L. et al. Role of LAMP-2 in lysosome biogenesis and autophagy. Mol Biol
Cell 13, 3355-3368, doi:10.1091/mbc.e02-02-0114 (2002).
93 Guerra, F. & Bucci, C. Multiple Roles of the Small GTPase Rab7. Cells 5, 34,
doi:10.3390/cells5030034 (2016).
94 Graham, J. M. Isolation of Lysosomes from Tissues and Cells by Differential and Density
Gradient Centrifugation. Current Protocols in Cell Biology 7, 3.6.1-3.6.21,
doi:https://doi.org/10.1002/0471143030.cb0306s07 (2000).
95 Lechleiter, J. D., Dartt, D. A. & Brehm, P. Vasoactive intestinal peptide activates Ca2(+)-
dependent K+ channels through a cAMP pathway in mouse lacrimal cells. Neuron 1, 227-
235, doi:10.1016/0896-6273(88)90143-2 (1988).
96 Findlay, I. & Petersen, O. H. Acetylcholine stimulates a Ca2+-dependent C1- conductance
in mouse lacrimal acinar cells. Pflugers Arch 403, 328-330, doi:10.1007/bf00583609
(1985).
97 Lambert, R. W., Bradley, M. E. & Mircheff, A. K. pH-sensitive anion exchanger in rat
lacrimal acinar cells. Am J Physiol 260, G517-523, doi:10.1152/ajpgi.1991.260.3.G517
(1991).
98 Ríos, J. D. et al. Age-dependent alterations in mouse exorbital lacrimal gland structure,
innervation and secretory response. Experimental Eye Research 80, 477-491,
doi:https://doi.org/10.1016/j.exer.2004.10.012 (2005).
99 Kiselyov, K. K., Ahuja, M., Rybalchenko, V., Patel, S. & Muallem, S. The intracellular
Ca²⁺ channels of membrane traffic. Channels (Austin) 6, 344-351, doi:10.4161/chan.21723
(2012).
135
100 Wang, Y. et al. Cytoplasmic dynein participates in apically targeted stimulated secretory
traffic in primary rabbit lacrimal acinar epithelial cells. J. Cell. Sci. 116, 2051,
doi:10.1242/jcs.00398 (2003).
101 Castle, A. M., Huang, A. Y. & Castle, J. D. The minor regulated pathway, a rapid
component of salivary secretion, may provide docking/fusion sites for granule exocytosis
at the apical surface of acinar cells. J. Cell. Sci. 115, 2963-2973 (2002).
102 Isobe, Y. et al. PIKfyve accelerates phagosome acidification through activation of
TRPML1 while arrests aberrant vacuolation independent of the Ca2+ channel. The Journal
of Biochemistry 165, 75-84, doi:10.1093/jb/mvy084 (2019).
103 Na, K.-S., Mok, J.-W., Kim, J. Y., Rho, C. R. & Joo, C.-K. Correlations between Tear
Cytokines, Chemokines, and Soluble Receptors and Clinical Severity of Dry Eye Disease.
Invest. Ophthalmol. Vis. Sci. 53, 5443-5450, doi:10.1167/iovs.11-9417 (2012).
104 Enríquez-de-Salamanca, A. et al. Tear cytokine and chemokine analysis and clinical
correlations in evaporative-type dry eye disease. Mol Vis 16, 862-873 (2010).
105 Denoyer, A. et al. CX3CL1 expression in the conjunctiva is involved in immune cell
trafficking during toxic ocular surface inflammation. Mucosal Immunol 5, 702-711,
doi:10.1038/mi.2012.43 (2012).
106 Lu, P. et al. Protective Roles of the Fractalkine/CX3CL1-CX3CR1 Interactions in Alkali-
Induced Corneal Neovascularization through Enhanced Antiangiogenic Factor Expression.
The Journal of Immunology 180, 4283, doi:10.4049/jimmunol.180.6.4283 (2008).
107 Reiser, J., Adair, B. & Reinheckel, T. Specialized roles for cysteine cathepsins in health
and disease. J Clin Invest 120, 3421-3431, doi:10.1172/JCI42918 (2010).
108 Božić, M., Verkhratsky, A., Zorec, R. & Stenovec, M. Exocytosis of large-diameter
lysosomes mediates interferon γ-induced relocation of MHC class II molecules toward the
surface of astrocytes. Cell Mol Life Sci, doi:10.1007/s00018-019-03350-8 (2019).
109 Liu, G.-Y. et al. Recycling of the Membrane-anchored Chemokine, CX3CL1. J. Biol.
Chem. 280, 19858-19866, doi:10.1074/jbc.M413073200 (2005).
110 Huang, Y.-W. et al. Constitutive Endocytosis of the Chemokine CX3CL1 Prevents Its
Degradation by Cell Surface Metalloproteases. J. Biol. Chem. 284, 29644-29653,
doi:10.1074/jbc.M109.045682 (2009).
111 Wong, H. S. et al. Cytoskeletal confinement of CX3CL1 limits its susceptibility to
proteolytic cleavage by ADAM10. Mol Biol Cell 25, 3884-3899, doi:10.1091/mbc.e13-11-
0633 (2014).
112 Jerdeva, G. V. et al. Actin and non-muscle myosin II facilitate apical exocytosis of tear
proteins in rabbit lacrimal acinar epithelial cells. J. Cell. Sci. 118, 4797,
doi:10.1242/jcs.02573 (2005).
136
113 Theron, M. et al. Pharmacodynamic Monitoring of RO5459072, a Small Molecule
Inhibitor of Cathepsin S. Front Immunol 8, doi:10.3389/fimmu.2017.00806 (2017).
114 Nanki, T. et al. Migration of CX3CR1-positive T cells producing type 1 cytokines and
cytotoxic molecules into the synovium of patients with rheumatoid arthritis. Arthritis &
Rheumatism 46, 2878-2883, doi:10.1002/art.10622 (2002).
115 Yano, R. et al. Recruitment of CD16+ monocytes into synovial tissues is mediated by
fractalkine and CX3CR1 in rheumatoid arthritis patients. Acta Medica Okayama 61, 89-98,
doi:10.18926/amo/32882 (2007).
116 Tanaka, Y. et al. Safety, pharmacokinetics, and efficacy of E6011, an antifractalkine
monoclonal antibody, in a first-in-patient phase 1/2 study on rheumatoid arthritis. Modern
Rheumatology 28, 58-65, doi:10.1080/14397595.2017.1337056 (2018).
117 Hocevar, A. et al. Parasympathetic nervous system dysfunction in primary Sjögren's
syndrome. Ann Rheum Dis 62, 702-704, doi:10.1136/ard.62.8.702 (2003).
118 Zoukhri, D. & Kublin, C. L. Impaired neurotransmitter release from lacrimal and salivary
gland nerves of a murine model of Sjögren's syndrome. Invest. Ophthalmol. Vis. Sci. 42,
925-932 (2001).
119 Zoukhri, D., Hodges, R. R., Byon, D. & Kublin, C. L. Role of proinflammatory cytokines
in the impaired lacrimation associated with autoimmune xerophthalmia. Invest.
Ophthalmol. Vis. Sci. 43, 1429-1436 (2002).
120 Ríos, J. D. et al. Age-dependent alterations in mouse exorbital lacrimal gland structure,
innervation and secretory response. Exp Eye Res 80, 477-491,
doi:10.1016/j.exer.2004.10.012 (2005).
121 Hanson, P. I. & Cashikar, A. Multivesicular Body Morphogenesis. Annual Review of Cell
and Developmental Biology 28, 337-362, doi:10.1146/annurev-cellbio-092910-154152
(2012).
122 Chen, Y.-D. et al. Exophagy of annexin A2 via RAB11, RAB8A and RAB27A in IFN-γ-
stimulated lung epithelial cells. Sci Rep 7, 5676, doi:10.1038/s41598-017-06076-4 (2017).
123 Mazzeo, C. et al. Exosome secretion by eosinophils: A possible role in asthma
pathogenesis. Journal of Allergy and Clinical Immunology 135, 1603-1613,
doi:https://doi.org/10.1016/j.jaci.2014.11.026 (2015).
124 Barral, D. C. et al. Functional redundancy of Rab27 proteins and the pathogenesis of
Griscelli syndrome. J Clin Invest 110, 247-257, doi:10.1172/jci15058 (2002).
125 Fukuda, M. Rab27 Effectors, Pleiotropic Regulators in Secretory Pathways. Traffic 14,
949-963, doi:https://doi.org/10.1111/tra.12083 (2013).
137
126 Baranov, M. V. et al. The Phosphoinositide Kinase PIKfyve Promotes Cathepsin-S-
Mediated Major Histocompatibility Complex Class II Antigen Presentation. iScience 11,
160-177, doi:https://doi.org/10.1016/j.isci.2018.12.015 (2019).
127 in Guide for the Care and Use of Laboratory Animals The National Academies Collection:
Reports funded by National Institutes of Health (ed th) (2011).
Abstract (if available)
Abstract
Autoimmune dacryoadenitis and altered lacrimal gland (LG) secretion are features of Sjögren’s syndrome (SS). Activity of cathepsin S (CTSS), a cysteine protease, is significantly and specifically increased in SS patient tears. The soluble chemokine, CX3CL1 (fractalkine), is cleaved from membrane-bound CX3CL1 by proteases including CTSS. ? In the first main part of this thesis (Chapter 2 and 3), I show that CX3CL1 is significantly elevated by 2.5-fold in tears (p?=?0.0116) and 1.4-fold in LG acinar cells (LGAC)(p?=?0.0026) from male NOD mice, a model of autoimmune dacryoadenitis in SS, relative to LGAC from BALB/c controls. Primary mouse LGAC and human corneal epithelial cells (HCE-T cells) exposed to interferon-gamma, a cytokine elevated in SS, showed up to 9.6-fold (p ??0.0001) and 25-fold (p???0.0001) increases in CX3CL1 gene expression, and 1.9-fold (p?=?0.0005) and 196-fold (p???0.0001) increases in CX3CL1 protein expression, respectively. Moreover, exposure of HCE-T cells to recombinant human CTSS at activity equivalent to that in SS patient tears increased cellular CX3CL1 gene and protein expression by 2.8-fold (p?=?0.0021) and 5.1-fold (p???0.0001), while increasing CX3CL1 in culture medium by 5.8-fold (p???0.0001). Flow cytometry demonstrated a 4.5-fold increase in CX3CR1-expressing immune cells (p???0.0001), including increased T-cells and macrophages, in LG from NOD mice relative to BALB/c. CTSS-mediated induction/cleavage of CX3CL1 may contribute to ocular surface and LG inflammation in SS. ? The second part (Chapter 4), includes my exploration of the role of Rab27a in CX3CL1 trafficking in the LGACs. Rab3D and Rab27a/b isoforms are effectors of exocytosis in LG, but Rab27a is poorly studied. To investigate whether Rab27a mediates CTSS secretion, we utilized quantitative confocal fluorescence microscopy of LG from SS-model male NOD and control male BALB/c mice, showing that Rab27a-enriched vesicles containing CTSS were increased in NOD mouse LG. Live-cell imaging of cultured LGAC transduced with adenovirus encoding wild-type (WT) mCFP-Rab27a revealed carbachol-stimulated fusion and depletion of mCFP-Rab27a-enriched vesicles. LGAC transduced with dominant-negative (DN) mCFP-Rab27a exhibited significantly reduced carbachol-stimulated CTSS secretion by 0.5-fold and ?-hexosaminidase by 0.3-fold, relative to stimulated LGAC transduced with WT mCFP-Rab27a. Colocalization of Rab27a and endolysosomal markers (Rab7, Lamp2) with the apical membrane was increased in both stimulated BALB/c and NOD mouse LG, but the extent of colocalization was much greater in NOD mouse LG. Following stimulation, Rab27a colocalization with endolysosomal membranes was decreased. In conclusion, Rab27a participates in CTSS secretion in LGAC though the major regulated pathway, and through a novel endolysosomal pathway that is increased in SS. ? The third major (Chapter 5 and 6) in my thesis involves the investigation of neural regulations in the LGAC and its relation to CTSS trafficking. Dysfunctional neural regulation in the LG contributes to altered tear secretion. Here, we compared secretion and trafficking of CTSS in the mouse LG when stimulated with the M3 muscarinic receptor agonist, carbachol (CCh), versus the ?1-adrenergic agonist, phenylephrine (PE). In situ and in vitro secretion studies showed that PE significantly increased CTSS secretion into tears compared to CCh, while CCh induced higher ?-hexosaminidase (?-hex) secretion. Calcium [Ca²?] imaging showed that CCh significantly elevated intracellular [Ca²?] and increased the size of Rab3D-enriched vesicles. The enlarged Rab3D is associated with increased homotypic secretory vesicle (SV)-SV fusion, heterotypic fusion between endolysosomes and SVs, and SV fusion towards the apical plasma membrane (APM). PE stimulation induced a lower intercellular [Ca²?] response and had a minimal effect in inducing Rab3D-enriched major regulated pathway. LGs deficient of Rab3D exhibited higher sensitivity towards PE stimulation and induced higher tear CTSS activity. The colocalization of endolysosomal markers (Lamp1, Lamp2, Rab7) with the apical actin was increased with both CCh and PE stimulation, with PE demonstrating a more apparent upregulation. In conclusion, PE stimulation increased CTSS secretion into tears, which is related to decreased Rab3D-enriched major regulated and an increased minor endolysosomal seceretory pathway. ? The final study (Chapter 7) investigates the role of PIKfyve in CTSS trafficking in the LGACs. Through in vitro rabbit LGAC culture, I found that pharmacological inhibition of PIKfyve with apilimod significantly reduced CTSS secretion as well as Rab7-vesicle fusion. PIKfyve inhibition in ex vivo tissue also showed that inhibition of PIKfyve significantly enlarged Rab7 and Lamp1-enriched vesicles in both C57 and Rab3DKO LG. Apilimod also significantly reduced the apical accumulation of Rab7 and Lamp1 in the mouse acini. In conclusion, I hypothesize that PIKfyve plays a role in endolysosome trafficking pathway, which is the main source of CTSS secretion in the LGAC.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Increased cathepsin S plays an important role in ocular surface manifestations in Sjögren’s syndrome
PDF
A study of the role of Rab27 in lacrimal gland acinar cell secretory trafficking
PDF
Investigation of mechanisms underlying development of autoimmune dacryoadenitis in lacrimal gland
PDF
Secretion of exosomes through the potential endolysosomal pathway in Rab3DKO and Non-Obese Diabetic murine models
PDF
Exploring serum and tear micro-RNA as biomarkers for early diagnosis of Sjögren’s Syndrome
PDF
Cathepsin L changes in development of autoimmune dacryoadenitis
PDF
Characterization of cathepsin S as potential biomarkers of Sjögren’s syndrome in mouse models
PDF
Secretion of exosomal miRNAs in Rab3DKO, a model of exocrine secretory deficit
PDF
Characterization of changes in metabolism and inflammation in the lacrimal gland in dry eye disorders
PDF
Development of novel immunosuppressant-based therapies to treat dacryoadenitis in a Sjögren’s syndrome mouse model
PDF
The characterization of Rab11a and trafficking mechanisms of polymeric immunoglobulin receptor (pIgR) in lacrimal gland acinar cells (LGACs)
PDF
The characterization of SNAREs and rabs in lacrimal gland acinar cells
PDF
Production and purification of recombinant mouse cathepsin S
PDF
Trafficking of targeted elastin‐like polypeptide nanoparticles in the lacrimal gland
PDF
Development and therapeutic assessment of multivalent protein polymers for cancer and eye diseases
PDF
Development of an elastin-like polypeptide-based cyclosporine A formulation that improves autoimmune-mediated dry eye characteristic of Sj鰃ren抯 syndrome
PDF
Targeting LFA-1/ICAM-1 interaction using ICAM-1 binding elastin-like polypeptide
PDF
Optimization of exosome isolation protocols for mouse serum and rabbit primary lacrimal gland acinar cells
PDF
Studies of lacrimal gland immune cell homeostasis in polymeric immunoglobulin receptor (pIgR) knockout mice and evaluation of monoclonal dimeric IgA producing rabbit hybridomas
PDF
Characterization of the class v myosin motor, Myosin 5c, in lacrimal acini
Asset Metadata
Creator
Fu, Runzhong
(author)
Core Title
The trafficking and pathogenesis of cathepsin S in Sjögren’s syndrome
School
School of Pharmacy
Degree
Doctor of Philosophy
Degree Program
Pharmaceutical Sciences
Degree Conferral Date
2021-08
Publication Date
07/26/2021
Defense Date
04/20/2021
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cathepsin S,lacrimal gland acinar cells,OAI-PMH Harvest,Sjögren’s syndrome
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Hamm-Alvarez, Sarah (
committee chair
), MacKay, J. Andrew (
committee member
), Okamoto, Curtis (
committee member
)
Creator Email
christinafu@ymail.com,runzhonf@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-oUC15622361
Unique identifier
UC15622361
Legacy Identifier
etd-FuRunzhong-9878
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Fu, Runzhong
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 author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Repository Email
cisadmin@lib.usc.edu
Tags
cathepsin S
lacrimal gland acinar cells
Sjögren’s syndrome