A systems biology approach to elucidating the frequency decoding mechanism governing differential mammalian gonadotropin subunit gene expression

The synthesis of the gonadotropin-subunits is directed by pulsatile gonadotropin-releasinghormone GnRH from the hypothalamus, with the frequency of GnRH pulses governingthe differential

A SYSTEM BIOLOGY APPROACH TO ELUCIDATING THE GnRH FREQUENCY DECODING MECHANISM THAT GOVERNS DIFFERENTIAL EXPRESSION OF THE GONADOTROPIN-SUBUNIT GENES

STEFAN LIM B.Sc(Hons.), Edin U

A THESIS SUBMITTED

IN ACCORDANCE WITH THE REQUIREMENTS OF

THE NATIONAL UNIVERSITY OF SINGAPORE

FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

se-My mother, who upheld and continued the convictions of my father, and encouraged methroughout this period; if nothing else, silently praying for strength and perseverance forme.

My wife, who has remained patient and understanding throughout this time, enduringlengthy periods of loneliness when through the force of circumstances, I have had to de-vote more time to research than to her

Dr Guna, who made it possible for me to do this PhD, by first accepting me into theM.Sc in Bioinformatics programme, and then recommending me to A-Star for the award

of the Ph.D scholarship If the former hadn’t happen, I would never have entered thebeautiful world of Biology

Dr Philippa, whom I will always maintain as the best person who could ever have pervised me, who took every risk imaginable in accepting me into her lab as an ignorantintern at first, and then later, as her student Moreover, for the last half-a-year of my can-didature, when my stipend had dried up, she gave me employment in the lab, so that Iwould never have to go hungry even for a day I will never cease to respect and marvel

su-at her trust in my non-abilities, which she constantly sees as opportunities for personalgrowth and fulfillment, and to be grateful to her for the one memorable visit to Israel, themost beautiful country on earth She is truly God-sent

Prof Zvi Naor, who has inspired me a great deal not only through his published work

in this field of gonadotropin gene regulation, but also through active discussions with himduring his visits to Singapore, as well as during my visit to Israel He embodies all ofwhat great scientists ought to have - intelligence, drive, fantasy and an aura of humanity,humility and congeniality

Mingshi, who mentored me and taught me so patiently every aspect of experimental ology, who taught me the beauty of life, and who is the sole reason why I have chosen topursue a Ph.D in this field and in this lab

Bi-Stella, who was my dearest friend and god sister, and had been the constant inspiration in

my life, however hard and trying times might have been She taught me the simple truths

of selfless love and friendship, and that it was not shameful nor cowardly to cry whenthings surrounding me became overwhelmingly difficult to bear In more ways than one,and as only she would comprehend, I owe my continued existence to her

Kathy, who became my friend very late on in my PhD career, and when she was about toleave Singapore for France to pursue her own academic dreams She epitomizes every-thing of a great scientist-to-be, and is probably one of the very few people in my life whowouldn’t mind talking science with me on the subway, all the way home She re-kindled

my interest in the French language - good or bad - it is not a worthless skill, at the veryleast

Andrea and Serena, who have been inseparable in their friendship and inseparable inworking their good deeds and charm Thank you for the little card you gave me beforeyou left our lab, bearing a message that reminded me for the remainder of my time in thislab that clearing trash and dirty bottles every so often was not a thankless task after all

Sue Yuan, who was someone I tried to encourage all through her period of sorrow, butended up being encouraged by her fortitude and experiences Thank you for being such adear friend, and for the mince pies you brought back from England

Members of Philippa’s Lab, some of whom have out-stayed me, while others haven’t.Regardless, each one of them has contributed no small part to my reaching the end, andhas made the pain of each experimental failure a little less

Liu Ping, who helped me much with all the experiments involving FCCS and live cellimaging

Keng Hwee, who has at times played the role of devil’s advocate, and at other times,the author’s advocate Whichever role he assumed, he did it better than anyone else

A*star, who funded this research project and also my studies

NGS, who supported me administratively throughout the course of my studies.

Celine, who came into my life rather unexpectedly, but most timely Her extraordinaryblend of teenage innocence and youthful exuberance worked wonders for an aching heart,tormented by the mistrust of others and the despair of a rejected thesis She acted as anangel commissioned by God, who appeared, and then disappeared - but who in the fewweeks that we shared life together, became my wonderfully adorable child, my sweet anddoting kid sister, my most precious friend, and everything else I could and would everwish for in life Her charmingly facetious tendencies and insatiable appetite for food andknowledge, were a joy to behold and a pleasure to oblige She ran alongside me, encour-aged me and infused me with just enough strength to complete this final mile Withouther, I most certainly would have given up short of the finishing-line It is thus only appro-priate to reserve my final and most needful word of thanks to an earthly being for her, withwhom I was not acquainted when this thesis was first submitted, but fully and endearingly

so, by the time it was eventually re-done

God, who is the One I will have to reserve most gratitude and honor for, without whomnothing would have been possible It was He, who created our amazing universe, and allthe science that undergirds the functionality of it all The pursuit of scientific study is butonly a God-given opportunity to try and understand the beauty and wonder of creation

The synthesis of the gonadotropin-subunits is directed by pulsatile gonadotropin-releasinghormone (GnRH) from the hypothalamus, with the frequency of GnRH pulses governingthe differential expression of the common α-subunit (αGSU), luteinizing hormone β-subunit (LHβ) and follicle-stimulating hormone β-subunit (FSHβ) In many vertebratespecies, levels of these hormones vary quite dramatically throughout their life cycles ow-ing to low levels of GnRH secretion that occur during the juvenile stage, suggesting a na-tive state of gene repression Preliminary findings point to the actions of histone deacety-lases (HDACs) in repressing the gonadotropins In this study, a system biology approach

is taken to unravel the mechanisms for GnRH-frequency decoding and GnRH-inducedde-repression of the gonadotropin-subunit genes Three mitogen-activated protein kinases(MAPKs), ERK1/2, JNK and p38, are known to be contributing uniquely and combinato-rially to the expression of each of these subunit genes Using mathematical modeling andcomputer simulations, it was found that dual specificity phosphatase (DUSP) regulation ofthe activity of these MAPKs through negative feedback, forms the basis for decoding thefrequency of pulsatile GnRH Furthermore, a fourth MAPK, ERK5, whose activation ki-netics and role in FSHβ gene expression are shown, was found to enhance the preference

of FSHβ for low GnRH pulse frequencies Evidence is presented for ERK5-activation ofFSHβ gene expression through Nur77-dependent and independent mechanisms, throughinteractions with MEF2D This involves the Ca2+-activated calcineurin both in activatingNur77 transcription, as well as possibly dephosphorylating Nur77, which is required forits activity Having established that distinct sets of HDACs repress the two β-subunits, arole for GnRH-activated Ca2+/calmodulin-dependent protein kinase I (CaMKI) is eluci-

dated in the de-repression of the FSHβ gene, which primarily involves phosphorylatingcertain class IIa HDACs, critical for their nuclear export Finally, Gem, a negative reg-ulator of calcium L-type channels, is shown to be involved in regulating αGSU expres-sion through influencing ERK1/2 activation in both a Ca2+-dependent and independentway These rely on Gem’s ability both to be re-localized to the cytosol upon CaM bind-ing, and to effect cytoskeletal remodeling upon 14-3-3 binding These findings reveal acomplex interplay of signal transducers, transcription factors, and both chromatin- andcytoskeletal-remodeling proteins at different levels to orchestrate the expression of vari-ous gonadotropin-subunit genes under the diverse actions of GnRH.

1.1 The gonadotropic hormones 1

1.1.1 The hypothalamic control of pituitary action 1

1.1.2 The gonadotropins and their role in reproduction 2

1.1.3 Gonadotropin-subunit gene regulation at a glance 3

1.1.4 Understanding gonadotropin-subunit gene expression through the use of model cell-lines 4

1.2 Regulation of gonadotropin expression by pulsatile GnRH 5

1.2.1 The requirement of pulsatile GnRH for optimal gonadotropin-subunit gene expression 5

1.2.2 The GnRH receptor-stimulated network as a frequency decoder 6 1.3 Regulation of gonadotropin expression by calcium 10

1.3.1 The calcium-channel regulator Kir/Gem is induced by GnRH 12

1.3.2 Gem 13

1.3.3 Both CaM and 14-3-3 localize to lipid rafts in c-raf signaling in the gonadotropes 17

1.4 Regulation of gonadotropin expression through targeting the chromatin 17 1.4.1 The fluctuating levels of GnRH at different stages of the verte-brate life cycle reveal a possible natural state of gonadotropin-subunit gene repression 17

1.4.2 Chromatin structure and the repression of the gonadotropin-subunit genes 19

1.4.3 Histone deacetylases (HDACs) 19

1.4.4 HDAC activity is involved in the repression of the gonadotropin β-subunit genes, and is overcome by GnRH 22

1.4.5 Distinct sets of HDACs repress the gonadotropin β-subunit genes in the immature gonadotropes 23

1.4.6 GnRH activates CaMKI in immature gonadotropes 25

1.4.7 Nur77 and MEF2D de-repress the FSHβ gene 26

1.5 Frequency decoding re-visited: the search for a frequency decoding mech-anism 28

1.6 Hypothesis and aims 32

1.6.1 Hypothesis 32

1.6.2 Aims 33

2 Experimental Materials and Methods 34 2.1 Cell culture, transfection and treatment 34

2.1.1 Cell culture 34

2.1.2 Cryo-storage of cells 34

2.1.3 Recovery of cells 35

2.1.4 Transfection of cells 35

2.1.5 Chemical treatment of cells 35

2.2 Plasmid construction 36

2.2.1 SiRNA constructs 36

2.2.2 Expression vectors 38

2.2.3 Isolation, verification and plasmid preparation 39

2.3 RNA extraction and reverse transcriptase PCR 42

2.3.1 RNA isolation 42

2.3.2 First strand cDNA synthesis 42

2.3.3 PCR and gel electrophoresis analysis 42

2.4 Luciferase assay 44

2.5 Statistical analysis 44

2.6 Whole cell extraction 45

2.7 Co-immunoprecipitation 45

2.8 Western blot 46

2.9 Immuno-fluorescence/Confocal microscopy 47

2.10 Live cell imaging 48

2.11 Fluorescence cross-correlation spectroscopy (FCCS) 48

3 Computational Modeling 50 3.1 Introduction 50

3.1.1 Published models on frequency decoding of GnRH signals 50

3.1.2 Proposed scheme of model development 52

3.2 The basic model 54

3.2.1 Model development 55

3.3 The intermediate and full models 58

3.3.1 Model development 59

3.4 Computer simulations and key readouts 65

3.5 Model codes 66

3.5.1 Code for the basic model 66

3.5.2 Code to analyze the basic model 69

3.5.3 Code for the intermediate model 70

3.5.4 Code to analyze the intermediate model 73

3.5.5 Code for the full model 73

3.5.6 Code to analyze the full model 78

4 Results and Sectional Discussions 80 4.1 Elucidating the GnRH frequency decoding mechanism in the gonadotropes 80 4.1.1 MKP negative feedback gives rise to frequency-dependent differ-ential gonadotropin-subunit gene expression 80

4.1.2 Sensitivity analysis of the basic model 84

4.1.3 Differential gene expression results from phosphatase-induced in-creases in average MAPK activation with decreasing frequency of the stimulus 86

4.1.4 GnRH activates ERK5 in αT3-1 cells 90

4.1.5 ERK5 activates the murine FSHβ promoter 91

4.1.6 ERK5 up-regulates FSHβ but down-regulates GnRHR mRNA lev-els in αT3-1 cells 93

4.1.7 ERK5 enhances FSHβ expression in a concentration-dependent manner 94

4.1.8 Sensitivity analysis of the intermediate model 97

4.1.9 Differential GnRHR concentration alone appears not to give rise to full differential gonadotropin-subunit gene expression 99

4.1.10 JNK-positive feedforward without ERK5-negative feedback on GnRHR expression causes loss of differential gonadotropin-subunit gene expression in the full model 103

4.1.11 ERK5-negative feedback against GnRHR expression restores

dif-ferential gonadotropin-subunit gene expression in the full model 1074.1.12 Sensitivity analysis of the full model 1114.1.13 Discussion 1154.2 GnRH-mediated de-repression of the gonadotropin β-subunit genes 1264.2.1 GnRH causes the nuclear export of wild-type HDACs 4, 5, and 7 1264.2.2 Mutation of 14-3-3 recognition sites abolishes nuclear export of

HDACs 4 and 5 1314.2.3 GnRH activates CaMKII rapidly in the immature gonadotropes 1334.2.4 GnRH-mediated de-repression of the FSHβ but not the LHβ gene

involves activation of CaMKI 1334.2.5 GnRH-mediated de-repression of the FSHβ but not the LHβ gene

involves CaMK phosphorylaton of HDACs 4 and 5 1344.2.6 ERK5 up-regulates Nur77 mRNA levels in αT3-1 cells 1344.2.7 Phosphorylated ERK5 co-precipitates with MEF2D after GnRH

treatment of immature gonadotropes 1374.2.8 ERK5 activates the murine FSHβ promoter through interactions

with MEF2D 1384.2.9 Discussion 1404.3 The role of Gem in α-subunit expression 1464.3.1 External calcium is both necessary and sufficient for basal and

GnRH-stimulated α-subunit gene activity in αT3-1 cells 1464.3.2 Gem over-expression does not disrupt GnRH-induction of α-subunit

mRNA levels in αT3-1 cells 1484.3.3 CaM- but not 14-3-3-binding sites of Gem are necessary and suf-

ficient for both basal and GnRH-induced α-subunit gene expression149

4.3.4 14-3-3- but not CaM-binding ability is crucial for both nuclear

import and export of Gem, and for mediating GnRH-induced phological changes in αT3-1 cells 1524.3.5 Gem knockdown reduces basal murine α-subunit promoter activity 1584.3.6 GnRH causes co-diffusion of Gem with ERK close to the plasma

mor-membrane 1594.3.7 Discussion 163

List of Tables

2.1 Nucleotide sequences used for making siRNA 36

2.2 Ligation reaction mix 37

2.3 Primers used for sequencing 41

2.4 Sequencing reaction mix 41

2.5 Pre-annealing reaction mix for cDNA synthesis 42

2.6 Primers used for PCR 43

2.7 PCR mix for gel analysis of gene expression levels by RT-PCR 43

2.8 Antibody dilutions used for western blotting 47

2.9 Buffers used in western blotting 47

3.1 Glossary of variables for the basic model 54

3.2 Constants 55

3.3 Glossary of new variables for the intermediate and full models 59

3.4 Additional constants for the intermediate and full model 60

List of Figures

1.1 Layout of the human pituitary 1

1.2 The GnRH receptor-stimulated network 9

1.3 Comparison of Gem sequences among various species 14

1.4 Ontogeny of the hypothalamic-pituitary-gonadal axis 18

1.5 Interaction partners of class IIa HDACs determine their localization 21

1.6 LHβ and/or FSHβ gene expression is repressed in gonadotrope cell-lines by HDACs, and this is overcome by GnRH 22

1.7 Distinct sets of HDACs are associated with LHβ and FSHβ genes in the immature gonadotropes 24

1.8 Knockdown of the associated HDACs reveals their crucial roles in the repression 25

1.9 GnRH activates CaMKI but not CaMKIV 26

1.10 Nur77 induces expression of the FSHβ gene in the immature gonadotropes and plays a role in the GnRH de-repressive effect 27

4.1 Profiles of MAPKK used in simulation of models 81

4.2 Lack of phosphatase feedback results in no differential-gene expression with the exponential pulse 82

4.3 Lack of phosphatase feedback results in no differential-gene expression with the square pulse 83

4.4 Inclusion of phosphatase feedback results in differential-gene expression 84 4.5 Sensitivity analysis of the basic model for the exponential pulse profile 85

4.6 Sensitivity analysis of the basic model for the square pulse profile 864.7 Analysis of MAPK activation for the exponential pulse profile in the basicmodel without phosphatase feedback 874.8 Analysis of MAPK activation for the square pulse profile in the basicmodel without phosphatase feedback 884.9 Analysis of MAPK activation for the exponential pulse profile in the basicmodel with phosphatase feedback 894.10 Analysis of MAPK activation for the square pulse profile in the basicmodel with phosphatase feedback 904.11 ERK5 is activated by GnRH in αT3-1 cells 914.12 Effects of over-expression of ERK5 and constitutively-active MEK5 onthe murine FSHβ promoter activity 924.13 RT-PCR analysis of the effects of over-expression of ERK5 and MEK5(D)

on GnRHR and FSHβ mRNA levels in αT3-1 cells 934.14 Intermediate model with phosphatase feedback demonstrates differentialgene expression 944.15 Analysis of ERK5 activation in the intermediate model 964.16 Highest fold-induction of FSHβ expression is dependent on total concen-tration of ERK5 974.17 Sensitivity analysis of the intermediate model for the exponential pulseprofile 984.18 Sensitivity analysis of the intermediate model for the square pulse profile 994.19 Differential GnRHR concentration alone appears not give rise to full dif-ferential gonadotropin-subunit gene expression 1014.20 Analysis of MAPK activation in the full model 1024.21 JNK-positive feedforward without ERK5-negative feedback on GnRHRexpression results in the loss of differential gonadotropin-subunit geneexpression 104

4.22 Analysis of rms values of MAPKs in the full model with JNK feedforward

only on GnRHR expression 105

4.23 Analysis of total activation of MAPKs in the full model with JNK feed-forward only on GnRHR expression 106

4.24 ERK5-negative feedback on GnRHR expression restores differential gonadotropin-subunit gene expression 108

4.25 Analysis of rms values of MAPKs in the full model with ERK5-negative feedback only on GnRHR expression 109

4.26 Analysis of total activation of MAPKs in the full model with ERK5-negative feedback only on GnRHR expression 110

4.27 JNK-positive feedforward and ERK5-negative feedback together on Gn-RHR expression endows differential gonadotropin-subunit gene expression 111 4.28 Sensitivity analysis of the full model to k1 112

4.29 Sensitivity analysis of the full model to k11 113

4.30 Sensitivity analysis of the full model for the exponential pulse profile 114

4.31 Physiological pulse profile of GnRH compared with those used in model simulation 121

4.32 GnRH stimulates nuclear export of HDAC 4 127

4.33 GnRH stimulates nuclear export of HDAC5 128

4.34 GnRH stimulates nuclear export of HDAC7 129

4.35 GnRH does not change the localization of HDAC6 130

4.36 Mutation of 14-3-3 recognition sites abolishes nuclear export of HDAC4 by GnRH 131

4.37 Mutation of 14-3-3 recognition sites abolishes nuclear export of HDAC5 by GnRH 132

4.38 GnRH activates CaMKII rapidly 133

4.39 GnRH de-repression of the FSHβ gene involves CaMKI 135

4.40 GnRH-mediated de-repression of the FSHβ but not the LHβ gene requiresCaMK phosphorylation sites on HDACs 4 and 5 1364.41 RT-PCR analysis of the effects of over-expression of ERK5 and MEK5(D)

on Nur77 mRNA levels in αT3-1 cells 1374.42 pERK5 co-precipitates with MEF2D in αT3-1 cells 1384.43 Effects of siMEF2D together with over-expression of ERK5 and MEK5(D)

on the murine FSHβ promoter activity 1404.44 A model depicting the proposed mechanisms through which GnRH de-represses the FSHβ gene in immature gonadotropes 1454.45 RT-PCR analysis of the effects of BayK 8644 (BK) and nifedipine onα-subunit expression levels in αT3-1 cells 1474.46 Effects of BK and nifedipine on the murine α-subunit promoter activity 1484.47 RT-PCR analysis of the effects of wild type Gem over-expression on α-subunit expression levels in αT3-1 cells 1504.48 RT-PCR analysis of the effects of over-expression of mutant forms ofGem on α-subunit expression levels in αT3-1 cells 1514.49 Effect of GnRH on wild type Gem localization in αT3-1 cells 1544.50 Effect of GnRH on CaM-binding mutant Gem localization in αT3-1 cells 1554.51 Effect of GnRH on 14-3-3-binding mutant Gem localization in αT3-1 cells 1564.52 Effect of GnRH on CaM/14-3-3-binding mutant Gem localization in αT3-

1 cells 1574.53 Effect of Gem knockdown on the murine α-subunit promoter activity 1584.54 Wild type Gem appears to co-diffuse with ERK after GnRH treatment 1604.55 CaM-binding mutant of Gem fails to co-diffuse with ERK after GnRHtreatment 1614.56 Positive and negative controls for FCCS 1624.57 A model depicting the proposed mechanisms through which Gem medi-ates GnRH actions on the α-subunit gene in immature gonadotropes 169

αGSU Glycoprotein α-subunit

AP-1 Activator protein 1

BAPTA/AM 1,2-bis-(o-aminophenoxy)ethane-N,N,N’,N’-tetra-acetic acid acetoxymethyl

tetraester

BMK Big mitogen-activated protein kinase

CaM Calmodulin

CaMK Ca2+/calmodulin-dependent protein kinase

cAMP 3’-5’-cyclic adenosine monophosphate

CoA Coactivator

CsA Cyclosporine A

DAG Diacylglycerol

dnNur77 Dominant negative Nur77

DUSP1 (or 4) Dual-specificity phosphatase 1 (or 4)

EGFP Enhanced green fluorescent protein

Egr-1 Early growth factor 1

ER Endoplasmic reticulum

ERK Extracellular-signal regulated kinase

FCCS Fluorescence cross-correlation spectroscopy

FSH Follicle-stimulating hormone

FSHβ Follicle-stimulating hormone β-subunit

GAP GTPase-activating protein

Gq/11 Alpha-subunit of heterotrimeric Gq protein

Gs Adenylyl cyclase-stimulating G alpha protein

GTP Guanosine triphosphate

HAT Histone acetyltransferase

HDAC Histone deacetylase

LH Luteinizing hormone

LHβ Luteinizing hormone β-subunit

MAPK Mitogen-activated protein kinase

MAPKK MAPK kinase

MEF2 Myocyte enhancer factor-2

MEK MAPK Erk kinase

MKP1 (or 2) MAPK phosphatase 1 (or 2)

mRFP Monomeric red fluorescent protein

NFAT Nuclear factor of activated T-cells

N-CoR Nuclear receptor co-repressor

Nur77(Nr4a1) Nuclear receptor subfamily 4, group A, member 1

PAGE Poly-acrylamide gel electrophoresis

PCR Polymerase chain reaction

PIP Phosphatidylinositol 4-monophosphate

PIP2 Phosphatidylinositol 4,5-biphosphate

Pitx-1 Pituitary homeobox 1

PKA Protein kinase A

PKCs Protein kinase C isoforms

PLC Phospholipase C

RMS Root mean-square

ROK Rho kinase

RT-PCR Reverse transcriptase polymerase chain reaction

SDS Sodium dodecyl sulphate

siRNA Short interfering ribonucleic acid

SMRT Silencing mediator of retinoic and thyroid hormone receptors

Looking back on the memory of

The dance we shared ’neath the stars above

For the moment all the world was right

How could I have known that you’d ever say goodbye

And now I’m glad I didn’t know

The way it all would end

The way it all would go

Our lives are better left to chance

I could have missed the pain

But I’d have had to miss the dance

The dance

The dance

I would have missed the dance

Holding you I held everything

For a moment wasn’t I a king

But if I’d only known how the king would fall

Hey who’s to say you know I might have changed it all

And now I’m glad I didn’t know

The way it all would end

The way it all would go

Our lives are better left to chance

I could have missed the pain

But I’d have had to miss the dance

So no one wanted to supervise you immunology? and that’s why you worked on gonadotropins?!!

Celine, aged A*Teen

Chapter 1

Introduction

1.1 The gonadotropic hormones

1.1.1 The hypothalamic control of pituitary action

The pituitary gland and the hypothalamus are both located within the cranial region ure 1.1) The pituitary gland is sometimes known as the ‘master gland’ of the endocrine

(Fig-Figure 1.1: Layout of the human pituitary The hypothalamus is connected to the pituitarythrough the infundibulum Electrical and chemical signals are conducted through the infundibu-lum to regulate the production and release of the pituitary hormones (Picture from [1] Copyrightc

system, because it controls the functions of the other endocrine glands Located at thebase of the brain, it consists of two structurally and functionally distinct parts: the anterior

pituitary (adenohypophysis) and the posterior pituitary (neurohypophysis) This gland isattached to the hypothalumus by the infundibulum comprising mainly blood vessels andnerve fibers Electrical and chemical signals are conducted through the infundibulum toregulate the production and release of the pituitary hormones.

The production of hormones by the anterior pituitary is stimulated by hypothalamicreleasing hormones, and, in some cases, suppressed by inhibitory hormones, all of whichtravel to the anterior pituitary via the portal vein system [2]

1.1.2 The gonadotropins and their role in reproduction

The hypothalamus secretes the decapeptide gonadotropin releasing hormone (GnRH)(pGlu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH2), which stimulates the production ofthe pituitary gonadotropins: the follicle stimulating hormone (FSH) [3] and the luteiniz-ing hormone (LH) [4] Both these hormones have crucial roles in vertebrate reproduction:FSH stimulates the ovaries to release steroid hormones, principally estrogen, and to ma-ture eggs in preparation for ovulation, while LH stimulates the production of androgens,e.g testosterone In female humans, LH acts on thecal cells of mature Graffian follicles

to produce the androgens that serve as substrates for follicular estradiol synthesis, thuspromoting luteinization and ovulation of the follicles [5] Ovulation is initiated by a sud-den rush of LH (the “LH surge”) In males, androgens produced by the Leydig cells ofthe testes in response to LH, induce spermatogonial cell formation and spermatogenesis

in the Sertoli cells [4]

LH and FSH are heterodimers consisting of two dissimilar subunits, α and β, ated non-covalently The α-subunit is common to all the pituitary glycoprotein hormones(LH, FSH and the thyroid-stimulating hormone (TSH)), whereas the β-subunit is uniqueand confers biological specificity [6,7] Both subunits must be associated for the hormone

associ-to be biologically active

1.1.3 Gonadotropin-subunit gene regulation at a glance

Given the importance of the gonadotropins in the vertebrate life cycle, much research inthe past two decades has been centered around the regulation of their biosynthesis [8–11].Mouse and rat models, as well as cell-lines, have been employed extensively to look atvarious aspects of gonadotropin-subunit gene expression For instance, very early on,

it was discovered that the pulsatility of GnRH signals was crucial in eliciting a strongtranscriptional response from the β-subunit genes Furthermore, the optimal responsefor each subunit gene correlated with a unique GnRH pulse frequency [12–15] Thistriggered many groups to attempt to elucidate the network of signaling molecules thatcould contribute to an understanding of the downstream effects of GnRH on the subunitgenes [9, 16] Such studies began to demonstrate the role of calcium, protein kinase C(PKC) and the mitogen-activated protein kinases (MAPKs) on the transcriptional regu-lation of the gonadotropin-subunit genes [17–22] At the same time, high-throughputmicroarray experiments were also being carried out in order to quicken the discovery

of genes up-regulated by GnRH, which could ultimately influence the synthesis of thegonadotropin-subunits [23, 24] While these experiments revealed a number of well-characterized transcription factors such as early growth response protein (Egr)-1 andthe orphan steroid receptor, Nur77, they also uncovered other less familiar genes, likeKir/Gem, known then only to regulate the number of active L-type calcium channels onthe plasma membrane [25] Recently, interest has shifted to the mechanism of decodingGnRH pulse frequencies, that would give rise to the differential expression of the subunitgenes Frequency decoding may be defined as the ability of the gonadotrope cells withinthe anterior pituitary to recognize different pulse frequencies of GnRH and through itsintracellular mechanisms, allow the frequency of GnRH to dictate the predominance inthe expression of any subunit gene A number of both theoretical and experimental ap-proaches have been employed thus far to elucidate this mechanism [26–36] Although

a precise mechanism has yet to be proposed, these attempts have provided evidence thatdecoding likely occurs within the GnRH receptor-activated signaling network This might

include receptor dynamics, transduction of the GnRH pulse signal through multiple ways, negative feedback mechanisms minimally at the levels of the receptor and MAPKs,and could even include elements of chromatin re-modeling [16, 32, 35].

path-1.1.4 Understanding gonadotropin-subunit gene expression through

the use of model cell-lines

To facilitate ongoing studies in the field of gonadotropins, two major cell-lines were rived from different stages of gonadotrope development The αT3-1 cell-line represents

de-an immature gonadotrope that is not fully differentiated, de-and expresses the α-subunit, theGnRH receptor (GnRHR) and the steroidogenic factor (Sf)-1, but neither of the two β-subunit genes [37, 38] To generate the αT3-1 cell-line, immature gonadotrope cells wereimmortalized through directed oncogenesis in transgenic mice In particular, targeted ex-pression of the oncogene SV40 T-antigen was carried out using a 1.8 kb fragment of the5’ regulatory region of the human glycoprotein hormone α-subunit gene Cell-lines werethen derived from resultant pituitary tumors in these transgenic mice The αT3-1 cell-linewas created from the tumor of a sexually immature transgenic mouse, while bearing theabove-mentioned gene expression phenotype [37, 38]

On the other hand, the LβT2 cell-line represents a fully-differentiated mature nadotrope that also expresses the LHβ-, but only very low levels of the FSHβ-subunit,unless hormonally treated [38, 39] The synthesis and secretion of intact and biologicallyactive LH was found to be strongly augmented by GnRH in these cells [39] The LβT2cell-line was generated through a different targeted oncogenesis with the 5’-flanking re-gion of the rat LHβ gene [38]

go-Since the two cell-lines fail to express one or both of the β-subunit genes, they aresuitable for investigating the molecular mechanisms governing the constitutive repression

of these genes Furthermore, the presence of the GnRHR, together with other factorscritical for carrying out downstream GnRH signaling, permits studies of GnRH-inducedexpression of all gonadotropin-subunit genes [37, 38] Ever since their derivation, these

cell-lines have been routinely used in both static and perifused culture for studying theregulation of gonadotropins at various levels [20–22, 32, 36, 40–42].

1.2 Regulation of gonadotropin expression by pulsatile

GnRH

1.2.1 The requirement of pulsatile GnRH for optimal

gonadotropin-subunit gene expression

In its natural physiological setting, the stimulation of both gonadotropin expression andsecretion by GnRH is highly-dependent on the pulsatile nature of GnRH delivery to theanterior pituitary When administered in a continuous fashion, exogenous GnRH causesthe down-regulation of LH and FSH secretion in several species, including primates andthe rat [43–45] Furthermore, the frequency and amplitude of GnRH pulses secreted bythe hypothalamus, which vary during different phases of the estrous cycle, regulate differ-entially LH and FSH secretion [12, 46–48] Likewise, the levels of gonadotropin-subunitgene expression in the rat pituitary vary by 2- to 4-fold, depending on the GnRH pulsefrequency and amplitude [12–15] A pulse of GnRH every hour maintains average levels

of gonadotropin-subunit mRNA levels and secretion However, increasing the frequency

of GnRH pulses increases LHβ gene expression and the secretion of LH On the otherhand, lowering the GnRH frequency results in the decline of LHβ gene expression and

LH secretion but induces a rise in FSHβ gene expression and FSH secretion The mon α-subunit gene expression is less stringently regulated by GnRH pulse frequency,but is produced in excess and, therefore not a major determinant of the rate of FSH or LHbiosynthesis [10] Nevertheless, continuous or high-frequency GnRH ensures high levels

com-of α-subunit production [12] The ability com-of the gonadotropes to decode different cies of GnRH and subsequently translate that into differential subunit gene expression isthus evident Nevertheless, the mechanism for decoding these frequencies has yet to besatisfactorily elucidated [35]

frequen-1.2.2 The GnRH receptor-stimulated network as a frequency decoder

It is reasonable to assume that the decoding of GnRH frequencies should reside in thecellular network of gonadotropic signaling pathways activated by GnRH While differentfrequencies of GnRH determine the optimal levels of gonadotropin-subunit gene expres-sion, any single GnRH molecule activates its cognate receptor in much the same way,regardless of pulse frequency All downstream pathways would also be expected to besimilarly activated by the one instance of receptor binding by GnRH Hence, it is crucial

to gain a clear understanding of the molecular events that happen upon receptor activation,before a mechanism for GnRH frequency decoding can be purported

1.2.2.1 The GnRH receptor

The mammalian GnRH receptor (GnRHR) is composed of 327-328 amino acids and is amember of the rhodopsin-like heptahelical G-protein-coupled receptor (GPCR) family ofproteins [49,50] It is unique among rhodopsin-family receptors in that it lacks a carboxyl-terminal domain, a feature that contributes to its relatively slow internalization [51–53],

a lack of G-protein receptor kinase phosphorylation, and a lack of rapid desensitization[54–56] This would imply that receptor desensitization is unlikely to be employed as aprimary means of differentiating between continuous or high frequency GnRH pulses andslower ones for frequency decoding

Notwithstanding, a correlation between cell surface GnRH receptor concentration der different GnRH pulse frequency regimes and optimal levels of subunit gene expressionwas reported [27, 32] It was found after 20 h of stimulation, that the receptor concen-trations were highest for intermediate pulses (1 pulse/30 min) of GnRH, which coincidedwith commensurately high levels of human α-subunit, rat LHβ and rat GnRHR promoteractivity in both GH3 rat pituitary somatolactotropic and LβT2 cells Conversely, highestlevels of rat FSHβ promoter activity were achieved with comparatively lower receptorconcentrations at slower GnRH frequencies (1 pulse/2 h) The receptors themselves areknown to be partially-regulated by GnRH [40–42] Thus, while receptor desensitization

un-may not contribute significantly to GnRH frequency decoding, it is possible that over anextended period (20 h or longer), GnRH regulates differentially the three gonadotropin-subunits through controlling GnRHR gene activity and cell surface receptor concentra-tion.

1.2.2.2 Signal transduction from the GnRHR to PKC

Once GnRH binds its specific cell-surface receptor, which subsequently izes [57, 58], the dimerized receptor then interacts with a Gq/11-protein [19, 59–66],causing the production of inositol triphosphate (IP3) through phospholipase C (PLC) bycleavage of phosphatidylinositol 4-monophosphate (PIP) and phosphatidylinositol 4,5-bisphosphate (PIP2) [61, 67–74] The production of IP3 then results in calcium mobi-lization from the endoplasmic reticulum (ER), with external calcium entering throughvoltage-gated calcium channels (VGCCs) on the cell membrane at a later stage to re-plenish the internal stores [75] There is also continuous regulation of cytosolic calcium,which is either flushed out of the cell or pumped into internal storage Consequently, cal-cium levels within the cell oscillate in response to the continuous actions of the internalcalcium management system [76–80] The elevation of internal calcium concentrationwithin the cell as a result of the GnRH stimulus, activates in particular, two of the conven-tional protein kinase C isoforms, α and βII (PKCα, βII), known to be found in pituitarycells [18, 81–84] In αT3-1 cells, calcium has also been found necessary for the activa-tion of two other isoforms, PKCδ and -ε [85] All these calcium-activated PKC isoforms(PKCs), in turn, have important downstream roles in mediating the actions of GnRH ongonadotropin expression [18, 85]

homodimer-1.2.2.3 Activation of the mitogen-activated protein kinases (MAPKs)

One of the notable consequences of calcium release into the cytoplasm and the activation

of PKCs is the firing of the three major MAPK cascades, culminating in the activation

of extracellular-signal regulated kinase (ERK) 1/2, c-Jun NH2-terminal kinase (JNK) andp38 A fourth MAPK, Big MAPK (BMK) or ERK5, is also stimulated, but little is known

about its effect on gonadotropin-subunit gene expression [85] Nevertheless, ERK5 hasbeen reported to co-activate Nur77 together with the transcription factor MEF2D in T-cells [86] Since Nur77 has already been implicated in the regulation of GnRHR expres-sion along with JNK [40–42], it is conceivable that ERK5 may yet feature in the GnRHfrequency decoding process by negatively-regulating the levels of GnRHR [42].

A considerable amount of work has already been carried out to demonstrate MAPKactivation by GnRH in the gonadotrope cell-lines ERK1/2 activation in these cells ismainly PKC-dependent, as shown in studies depleting and inhibiting PKC and then ex-amining the GnRH stimulation of ERK1/2 [18, 22, 63, 87] The same series of studiesshow a further means of ERK activation through c-Src/dynamin/Ras, but this is much lessinfluential [63, 87] Both GnRH-induced JNK and p38 activation have also likewise beenfound to be PKC-dependent [85]

Simultaneous with Gq/11 activation is also the activation of another G-protein, Gs[62, 88–96] Gs activation causes a rise in cAMP levels [61, 97], leading to the activa-tion of protein kinase A (PKA) [98, 99], also implicated in ERK1/2 activation [19, 100].Therefore, following GnRH stimulation of αT3-1 cells, ERK1/2 is activated about 12 foldabove basal levels, JNK 20 to 50 fold, and p38 about 2 fold [85] GnRH also causes theup-regulation of the ERK1/2, JNK and p38-specific phosphatases, DUSP1 and 4 (alsoknown as MKP1 and 2 respectively), possibly in a frequency-dependent manner to act

as negative feedback, terminating GnRH signaling to MAPK activation [24, 101, 102].The GnRH receptor-stimulated network comprises extensive cross-talk among variouspathways, and the participation of many other molecular species have only begun to beincluded in this ever-enlarging signaling map (Figure 1.2)

1.2.2.4 The differential dependence of the gonadotropin-subunit genes on various

combinations of MAPKs

A series of publications indicated that the three subunit genes are activated by differentcombinations of the three major MAPK pathways [20–22] Essentially, while all threesubunit genes require ERK1/2 for transcriptional activation, LHβ requires in addition,

Figure 1.2:The GnRH receptor-stimulated network GnRH activates a number of signaling ways, including MEK, JNK, ERK1/2, cAMP/PKA, PKC, Ca2+- and CaM-dependent pathways Anumber of transcription factors are activated through the phosphorylation by these kinases (Figureadapted from [16].)

path-JNK FSHβ, however, requires all three MAPK pathways

The dependence of gonadotropin-subunit gene expression on the MAPKs has twolikely implications on the GnRH frequency decoding mechanism: firstly, one group sug-gested MAPK activity thresholding as a possible mechanism for differential gene activa-tion In the simplest instance, both the α-subunit and FSHβ depend heavily on ERK1/2for transcriptional activation Simultaneous with the activation of this pathway is also theGnRH-stimulated up-regulation of DUSP1, known to inactivate ERK1/2 [102] Hence,higher frequencies of GnRH mean that the level of active ERK1/2 may not cross a re-quired threshold to induce high FSHβ gene activity, although it may be sufficient forα-subunit activity In this case, slower pulses of GnRH will result in less phosphataseactivity, allowing sufficient build up of ERK1/2 to past the threshold level required forsignificant activation of FSHβ gene activity [35]

Secondly, the differential reliance of the three subunit genes on various combinations

of ERK1/2, JNK and p38 could also form a basis of GnRH frequency decoding Althoughthis has not been suggested anywhere in the literature, it can be speculated that a gene

like the α-subunit, which depends only on ERK1/2 for activation, would be optimallyexpressed at GnRH frequencies where only ERK1/2, but not JNK or p38 are most highlyactivated On the other hand, genes requiring more than one or two MAPKs for activation,would only be optimally expressed at frequencies which ensure that all requisite MAPKsare activated at the highest possible levels simultaneously Such a synchronization ofrequisite MAPK activation would likely be dictated by GnRH frequency, and furnish areasonable connection between GnRH frequency and differential subunit gene expression.Certainly, this forms a hypothesis that ought to be tested.

1.3 Regulation of gonadotropin expression by calcium

Calcium is relatively well-characterized as an important second messenger for many naling systems, and in the case of the gonadotropes, it has been implicated in a variety

sig-of downstream responses [9, 18, 20–22, 103–108] Calmodulin (CaM) acts as a Ca2+ ceptor involved in many of these downstream responses by its interaction with a number

re-of CaM-binding proteins [109] Among these are the multi-functional Ca2+dependent protein kinases (CaMKs) In both rat primary pituitary and LβT2 cells, GnRHwas reported to activate CaMKII, one of the three known isoforms, the other two beingCaMKI and -IV [110,111] Moreover, treatment with the CaMK inhibitor KN-93 reducedGnRH stimulation of α-subunit and LHβ promoter activity, as well as mRNA levels of allthree subunits [110, 111] This suggests that GnRH can act through CaMK to stimulatethe gonadotropin-subunit genes

/calmodulin-CaM has also been implicated in GnRH signaling to ERK1/2 Inhibition of /calmodulin-CaM usingW-7, a common calmodulin antagonist that binds to calcium-loaded calmodulin (Ca2+-CaM) in place of its normal physiological target proteins, was sufficient to block GnRH-induced ERK1/2, but not JNK activation in αT3-1 cells [112] It was further shown thatCaM inhibition also attenuated ERK-dependent gene reporter activity of c-fos, murineα-subunit and MAPK phosphatase (MKP)-2 promoters However, since the use of KN-

62, a CaMK inhibitor, did not recapitulate these findings, other CaM-dependent signal

transducers are likely to be involved in mediating the GnRH effects [112].

The effect of GnRH-induced increases in intracellular calcium on the expression ofeach gonadotropin-subunit gene has also been studied In the rat, elevation of cytosoliccalcium levels with the calcium ionophore, ionomycin, has been reported to stimulate arise in α-subunit, LHβ and FSHβ mRNA levels [17, 28, 113] Using the cell-permeant

Ca2+ chelator, BAPTA/AM, to inhibit the mobilization of intracellular Ca2+ in αT3-1cells surprisingly elevated basal levels of α-subunit mRNA, but markedly reduced theGnRH effect on α-subunit mRNA levels [114] In contrast, investigations using GH3rat pituitary somatolactotropic cells expressing the GnRH receptor failed to demonstratecalcium-induced increases in rat LHβ and FSHβ promoter activity, while stimulating ratα-subunit promoter activity [115] Other studies showed that chelation of external Ca2+with EGTA did not prevent GnRH from stimulating rat LHβ- or FSHβ-promoter activ-ity using chloramphenicol acetyl transferase (CAT) and luciferase assays, respectively, inLβT2 cells [20, 22] CAT assays however demonstrated significant disruption in rat α-subunit promoter activity after EGTA-chelation of Ca2+, suggesting that external calciumwas important in mediating the GnRH effect on α-subunit expression [21] This wasfurther evidenced by studies carried out on the human α-subunit promoter transfectedinto αT3-1 cells, and subsequently treated with the calcium channel agonist, BayK 8644,the calcium channel blocker, Nifedipine, or thapsigargin, which depletes intracellular cal-cium stores [116] These studies demonstrated significant stimulation of human α-subunitpromoter activity by BayK 8644, while Nifedipine essentially blocked GnRH-stimulatedpromoter activity Additionally, depletion of internal calcium stores by thapsigargin led

to marked reduction in both basal and GnRH-stimulated α-subunit promoter activity try of external calcium has already been shown to be important for sustained biphasiccalcium oscillations in gonadotropes after GnRH stimulation, which account for the non-transient increases in intracellular calcium [8, 68, 76, 117, 118] As such, results fromstudies involving depletion or forced influx of external calcium would, in general, be ex-pected to agree with those derived from directly inducing or blocking increases in levels

En-of cytosolic calcium This appears to be the case mostly for the α-subunit, but less so forthe β-subunits The reason for this could be that the activation of ERK1/2, which is thekey MAPK implied in α-subunit gene expression [21], has been shown to rely on bothcytosolic calcium-activated CaM [112], as well as external calcium influx [22].

Therefore, while the possibility of calcium mediating GnRH-induced subunit gene expression has certainly been studied to some degree, a concrete under-standing of the calcium requirement for each gonadotropin-subunit’s gene activity is stilllacking This is due mainly to the discrepancies in the cell model employed for the abovestudies, the methods of inducing or depleting calcium, or the measured phenotype More-over, the specificity of some inhibitors used in targeting various calcium-related signaltransducers has frequently been found wanting For instance, KN-62, when used at dif-ferent molar concentrations can inhibit CaMKI, -II or -IV, although it has often been cited

gonadotropin-as a specific inhibitor of CaMKII [119, 120] On the other hand, while KN-93 tions usually as a general CaMK inhibitor, one study, at least, shows that it has actionsother than CaMK inhibition [121] Hence, even with W-7 being an undisputed specificinhibitor of Ca2+−CaM, one cannot discount the possibility of non-specific actions ofthese inhibitors confounding any result obtained A more prudent approach might be toemploy other methods of blocking these signal transducers, over and above the use ofdrug-based inhibitors

func-1.3.1 The calcium-channel regulator Kir/Gem is induced by GnRH

While calcium is known to play a crucial role in gonadotropin-subunit gene sion [17, 20–22, 28, 113], some of the regulators of calcium activity known in other celltypes have yet to be studied in the context of the gonadotropes A large-scale microarrayscreen for genes up-regulated by GnRH in LβT2 cells revealed that the mRNA for Gem(also referred to as Kir), a calcium signaling pathway-associated protein, was increasedwith GnRH treatment This result was confirmed by quantitative real-time PCR [24].Additionally, it was recently shown that the Gem protein in the nuclear extracts of un-

expres-stimulated αT3-1 cells is more abundant than in LβT2 cells [122] Gem may thus tribute to the differences in the expression levels of certain genes in these two cell-linesdue to its ability to regulate calcium activity, which would influence the behavior of keytranscription factors downstream.

GEM is a small, 35-kD G-protein that is highly expressed in the pituitary, and belongs

to the small GTP-binding family of proteins within the Ras superfamily Together withRem1, Rem2 and Rad, they are sometimes referred to as the RGK (Rad, Gem, Kir) family[123–125]

1.3.2.1 Gem structure

Ras-family GTPases share a set of conserved elements, designated G1 through G5, volved in GDP/GTP binding and GTP hydrolysis (Figure 1.3) They possess high-affinityguanine-nucleotide binding activity and relatively low, but easily detectable, intrinsic GTPhydrolysis activities Other factors are required to accelerate the intrinsic GTPase activity(GTPase-activating proteins (GAPs)) or promote the formation of the GTP-bound forms(GDP-GTP exchange factors (GEFs)) [126]

in-RGK proteins have been found, in general, to contain critical substitutions in aminoacids that participate in GTP hydrolysis when compared with Ras (Figure 1.3) Moreover,Gem, in particular, displays relatively low levels of intrinsic GTPase activity [127], and

is predominantly found in the GTP-bound state as a result of the approximately ten-foldhigher concentrations of GTP relative to GDP in most cells [124] As such, it is unlikelythat Gem acts similarly to Ras in activating some of the key MAPK pathways Ratherthan being a bona fide GTPase, it is more a guanine nucleotide-binding protein

Another notable characteristic of RGK proteins is their conserved 40-residue C-terminalextension, which lacks motifs directing fatty acylation, a biochemical modification thatmediates membrane association of many Ras family proteins Membrane anchorage is

Figure 1.3:Predicted amino acid sequences encoded by cDNAs for human and murine Gem and M-Gem, respectively) and comparison to human Rad and human c-H-Ras1 The open readingframe was determined for two independent human and a single murine Gem cDNA clone Aminoacids conserved in at least three proteins are in bold type and those conserved in all four proteinsare bold and underlined Dots indicate gaps inserted to allow optimal alignment of the sequences.Numbers on the right indicate residue number Consensus sequences for GTP-binding regions areindicated in italics (Figure adapted from [123].)

(H-critical for Ras signaling since it is commonly activated by SOS, another bound signaling molecule [126] This is an additional reason why Gem likely does not

membrane-activate MAPKs in the usual fashion of Ras Nevertheless, contained in the conserved40-residue C-terminal extension are necessary serine phosphorylation sites and bindingregions for CaM and 14-3-3 [128] A 14-3-3 binding region is also found in the N-terminalregion of Gem (Figure 1.3).

1.3.2.2 The function of Gem as a calcium-channel regulator

It is mainly the CaM binding ability of Gem that gives rise to its function as a channel regulator [125, 128] Gem achieves this by influencing directly the operation ofL-type channels, which are found on the cell-surface of gonadotropes [8] These L-typechannels are one of several kinds of VGCCs that regulate entry of extracellular calciuminto the gonadotropes External calcium entry is required for downstream processes such

calcium-as transcription, and secretion of proteins [129]

The L-type channels are multisubunit complexes comprising a main pore-forming

α1-subunit, complexed with β-, α2δ- and γ-subunits The β-subunits are required forproper membrane trafficking of the α1-subunit and are thus involved in modulating theelectrophysiological properties of the channel [130] Gem was found to bind the β3-subunit (a subtype of the β-subunit) through its Ras-like core [25] This binding competeswith the α1-subunit binding of the β3-subunit, thus reducing the number of active L-typechannels on the plasma membrane, and consequently results in an inhibition of channelactivity [25, 131]

Using mutant forms of Gem where necessary serine and tryptophan residues were stituted, 14-3-3 and CaM-binding of Gem were shown to be necessary for its cytoplasmiclocalization [25, 132], although only CaM-binding proved necessary for inhibition of L-type channel activity [128] CaM binds Gem around tryptophan 269, while 14-3-3 binds

sub-at serines 23, 261 and 289 (Figure 1.3) [128], with both CaM and 14-3-3 competing forbinding to Gem [132] CaM is activated by GnRH through Ca2+, and 14-3-3 proteins havebeen shown to interact with CaM in the gonadotropes [112] It is thus possible that GnRHcauses CaM to bind to Gem in order to down-regulate L-type calcium channels through

the consequent cytoplasmic re-localization of Gem that sequesters the β3-subunits thermore, since GnRH already activates calcium, this reduction of external calcium entry

Fur-by Gem could act as a negative feedback against calcium activation Fur-by GnRH, affectingpathways downstream of calcium, particularly the MAPKs

1.3.2.3 The function of Gem as a negative regulator of the Rho-Rho kinase pathway

in cytoskeletal remodeling

Other than binding the β3-subunit of calcium L-type channels, Gem has also been found

to bind Rho kinase (ROK) β [124, 125] ROKβ is a major effector of the Rho Pase, and the formation of the ROKβ-Gem complex blocks the interaction of ROKβwith specific substrates including myosin light chain (MLC) and myosin phosphatase[133] These are required for cytoskeletal remodeling during the regulation of calcium-dependent actinomycin-based contraction of smooth muscle and non-muscle cells in theRho-Rho kinase pathway [124, 134] The phosphorylated serine residues at positions 261and 289 were found to be critical for this function of Gem In conjunction with Ser23,the bi-dentate 14-3-3 binding at these phosphorylated serine residues increases Gem pro-tein half-life [128] It is possible that this increase in half-life is brought about by aconformational change to the Gem core which 14-3-3 binding stabilizes, and this change

GT-is necessary for ROKβ binding Alternatively, as ROKβ and its substrates are localized

to membranes and the associated contractile apparatus, it is conceivable that Gem localization with ROKβ is dependent on the phosphorylation of these necessary serineresidues [124] Since 14-3-3 binding of Gem appears to be essential for re-localization

co-of Gem either to the cytoplasm in general, or specifically to sub-cellular compartmentswhere ROKβ is located, it may be that 14-3-3 proteins chaperon Gem to particular desti-nations where Gem functions

While Gem binding to ROKβ blocks the interaction with certain cytoskeletal eling substrates, it does not block ROKβ activation of LIM kinase, which in turn phos-phorylates cofillin, another structural protein [133] This suggests that Gem may alsoact by modifying the substrate specificity of ROKβ, rather than inhibiting its enzymatic