Transgenic Studies of Growth Hormone Secretogogue Physiology

Nina Balthasar (Dipl. Biochem.)

A thesis submitted in fulfilment of the requirements of the University of London, for the degree of Doctor of Philosophy.

November 2000

Division of Neurophysiology, National Institute for Medical Research, The Ridgeway, Mill Hill, London, U.K ProQuest Number: 10010142

All rights reserved

INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest.

ProQuest 10010142

Published by ProQuest LLC(2016). Copyright of the Dissertation is held by the Author.

All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code. Microform Edition © ProQuest LLC.

ProQuest LLC 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106-1346 Abstract

The synthetic growth hormone secretagogues (GHS) exert their growth hormone (GH) releasing effects by activating their receptor (GHS-R) in hypothalamic neurones and pituitary GH cells. The aim of this study was to investigate hypothalamic and pituitary GHS physiology by in vitro studies and transgenesis, over-expressing the GHS-R in hypothalamic GH releasing hormone (GHRH) neurones or pituitary GH cells. Initially, to analyse transgene expression driven by two targeting cosmids - a rat GHRH or a human GH cosmid - transgenic mice expressing enhanced green fluorescent protein (eGFP) were generated. In rGHRH-eGFP transgenic mice eGFP protein was visible in brightly fluorescing GHRH neurone cell bodies and median eminence nerve terminals. hGH-eGFP transgenic mice showed high levels of eGFP fluorescence limited to GH expressing pituitary cells. These mice provide useful new tools to study GHRH and GH cell physiology. In vitro studies with a stably GHS-R transfected GH cell line showed that processing of the GHS-R cDNA resulted in a functionally active receptor. In transgenic mice over­ expression of the GHS-R in GHRH neurones caused alteration of the GH axis with increased pituitary GH content and body weight. GH responses to a GHS injection seemed enhanced. However, prolonged GHS exposure did not cause enhanced body weight gain compared to WT littermates. These studies suggested that enhancement of GHS-R abundance in GHRH neurones caused alteration of the GH and GHS axis, implicating GHRH neurones as an important GHS target in GH release. In contrast, transgenic mice with 50-fold GHS-R over-expression in GH cells showed reduced pituitary GH content, body weight and GH responses to GHRH. These data suggest that somatotroph function might be compromised by high levels of GHS-R over-expression. Crossing the hGHS-R over-expressing mice with their eGFP counterpart mice will facilitate the further investigation of their physiological phenotype by enabling in vivo analysis of GHS effects directly in identified cells. Contents

Acknowledgements ix Author’s Declaration x Abbreviations xi

1 Introduction 1

1.1 Aims of the thesis 1

1.2 Anatomical aspects of the Hypothalamo-Pituitary axis 3 1.2.1 Hypothalamo-Pituitary Communication 3 1.2.2 Hypothalamus Architecture 4 1.2.3 Somatotroph Differentiation 6

1.3 Regulation of the Hypothalamo-Pituitary Growth Hormone Axis 7 1.3.1 GH - the gene, the receptor and its actions 7 1.3.2 Growth hormone releasing hormone (GHRH) 9 1.3.3 Somatostatin 13 1.3.4 GH feedback mechanisms 16 1.3.5 GH pulsatility 17

1.4 Artificial Growth Hormone Secretagogues (GHS) and Ghrelin 18 1.4.1 Historical and Structural Aspects of GHS development 19 1.4.2 GH releasing actions of synthetic GHS 21 1.4.3 GHS sites of action 24 1.4.4 The GHS Receptor (GHS-R) 30

1.5 Tissue-Specificity in Transgenesis 36 1.5.1 Regulation of gene transcription 36 1.5.2 Locus control regions 37 1.5.3 Generating targeting constructs 38

2 Materials and Methods 40

2.1 Bacterial cultures 40

2.2 Purification of nucleic acids 40

2.3 DNA preparation 40 2.3.1 Preparation of plasmid and cosmid DNA 40 2.3.2 Preparation of genomic DNA fromanimal tissue 40

2.4 Subcloning of DNA fragments 41 2.4.1 Restriction digestion 41

11 2.4.2 Blunting of DNA fragments by refilling the 5’ overhang 41 2.4.3 Vector dephosphorylation 41 2.4.4 Insertion of linkers 41 2.4.5 Gel electrophoresis of DNA fragments 41 2.4.6 Purification of DNA fragments from agarose gels 42 2.4.7 Ligation of DNA fragments 42 2.4.8 Preparation and transformation of competent cells 42 2.4.9 Packaging of cosmid DNA into bacteriophage 42

2.5 DNA sequencing 42

2.6 Polymerase Chain Reaction 43

2.7 Detection of DNA sequences by Southern blotting 45 2.7.1 Southern Blotting 45 2.7.2 Radioactive random prime labelling 45 2.7.3 Hybridisation of Southern Blots 45

2.8 Determination of protein content 45

2.9 Generation of transgenic mice 45 2.9.1 Purification of large DNA fragments for microinjection 45 2.9.2 Superovulation, microinjection, and embryo transfers 46

2.10 Radioimmunoassay (RIA) 46 2.10.1 lodination of tracers 47 2.10.2 Mouse and rat GH and Prl RIA 47 2.10.3 cGFPRIA 47 2.10.4 cAMPRIA 47

2.11 Reverse transcriptase polymerase chain reaction (RT-PCR) 48

2.12 /« hybridisation 48 2.12.1 In vitro transcription 48 2.12.2 Riboprobe Hybridisation 48 2.12.3 Oligonucleotide Hybridisation 49 2.12.4 DIG hybridisation 49 2.12.5 Image analysis 49

2.13 RNAse Protection Assay 50

2.14 Immunocytochemistry 50

2.15 Jugular vein cannulation 51

2.16 /« v/vo Confocal Microscopy 51

2.17 /« vzYro cell systems 52 2.17.1 Culture of rat GC cells 52

111 2.17.2 Cell Transfection 52 2.17.3 Treatment of GC cultures with hypothalamic factors 52 2.17.4 Phosphoinositol turnover 52 2.17.5 Culture of hypothalamic neurones 53

2.18 Statistical Analysis 54

3 In vitro studies in a GH cell line stably transfected with the human GHS-R 54

3.1 Introduction 54

3.2 Characterisation of GHS-R^''® GC cells 55 3.2.1 Generation and detection of GHS-R^''® GC cells 55 3.2.2 Phosphoinositol turnover studies in GHS-R'^''® GC cells 57

3.3 Transfection of the human GHRH-R into GHS-R^"'® GC cells 60 3.3.1 Detection of GHRH-R/GHS-R+"" GC cells 60 3.3.2 Physiology of GHRH-R/GHS-R+"' GC cells 61

3.4 Discussion 62

4 Targeting transgenes to hypothalamic growth hormone releasing hormone (GHRH) neurones 66

4.1 Transgenic mice expressing eGFP in GHRH neurones 66 4.1.1 Introduction 66 4.1.2 The rGHRH-eGFP cosmid construct 67 4.1.3 Generation of rGHRH-eGFP transgenic mice 68 4.1.4 Transgene expression in rGHRH-eGFP transgenic mice 71 4.1.5 Discussion 81

4.2 Transgenic mice over-expressing the hGHS-R in GHRH neurones 84 4.2.1 Introduction 84 4.2.2 The rGHRH-hGHS-R cosmid construct 84 4.2.3 Generation of rGHRH-hGHS-R transgenic mice 87 4.2.4 Transgene expression in rGHRH-hGHS-R transgenic mice 89 4.2.5 Discussion 92

5 Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice 97

5.1 Studies of rGHRH-eGFP transgenic mouse physiology 97 5.1.1 Introduction 97 5.1.2 rGHRH-eGFP transgenic mice are phenotypically normal 97 5.1.3 Discussion 99

5.2 Studies of rGHRH-hGHS-R transgenic mouse physiology 100 5.2.1 Introduction 100

IV 5.2.2 rGHRH-hGHS-R transgenic mice show altered GH physiology 101 5.2.3 Discussion 109

6 Targeting transgenes to pituitary GH cells 114

6.1 hGH-eGFP transgenic mice 114 6.1.1 Introduction 114 6.1.2 The hGH-eGFP cosmid construct 115 6.1.3 Generating hGH-eGFP transgenic mice 115 6.1.4 Transgene expression in hGH-eGFP transgenic mice 117 6.1.5 Discussion 118

6.2 hGH-hGHS-R transgenic mice 119 6.2.1 Introduction 119 6.2.2 The hGH-hGHS-R cosmid construct 119 6.2.3 Generating hGH-hGHS -R transgenic mice 120 6.2.4 Transgene expression in hGH-hGHS-R transgenic mice 123 6.2.5 Discussion 123

7 Physiological studies of hGH-hGHS-R transgenic mice 127

7.1 Introduction 127

7.2 hGH-GHS-R transgenic mice are GH deficient 127

7.3 Discussion 134

8 Ghrelin - a novel growth hormone secretagogue 138

8.1 Introduction 138

8.2 Cloning of mouse Ghrelin 28 and 27 140

8.3 Expression analysis of mouse Ghrelin 141

8.4 Discussion 145

9 GHS treatment studies in transgenic mice and rats 149

9.1 Oral GHS treatment of Tgr and WT rats 149 9.1.1 Introduction 149 9.1.2 Results 150 9.1.3 Discussion 154

9.2 GHS treatment of rGHRH-hGHS-R and WT mice 156 9.2.1 Introduction 156 9.2.2 GHRP-6 (25pg/day i.p.) treatment for 14 days in female rGHRH-hGHS- R transgenic mice 157 9.2.3 GHRP-6 (2xl5|ig s.c. daily) treatment for 9 days in male rOHRH-hOHS- R transgenic mice 160 9.2.4 Discussion , 164

10 Final discussion 168

10.1 Hitting the spot - Generation of targeting constructs 168

10.2 GHS Physiology in hGHS-R over-expressing transgenic mice 170 10.2.1 hGHS-R over-expression in GHRH neurones 170 10.2.2 GHSs - GH vs metabolic effects 173 10.2.3 Longer-term GHS treatment 174 10.2.4 hGHS-R over-expression in GH cells 176 10.2.5 Hypothalamic vs pituitary GHS-R over-expression 177 10.2.6 Knocking out the background 178

10.3 eGFP and hGHS-R expression in the same cell 179

List of Figures and Tables

Figure 1.1 Hypothalamo-Pituitary communication 4 Figure 1.2 Hypothalamus Architecture 5 Figure 1.3 The hGH cluster 7 Figure 1.4 Biosynthesis of GHRH 10 Figure 1.5 GHRH signalling pathway 14 Figure 1.6 Several classes of GHS 20 Figure 1.7 GHS signal transduction 26 Figure 1.8 Regulation of GH secretion 30 Figure 1.9 Predicted GHS-R transmembrane topology 32 Figure 3.1 Transfection of the human GHS-R type la cDNA into GC cells 56 Figure 3.2 Detection of hGHS-R protein in GHS-R^"'® GC cells 56 Figure 3.3 GHRP-2 response in GHS-R^"'® GC cells 57 Figure 3.4 Dose range study for GHRP-2 in GHS-R^"'® GC cells 58 Figure 3.5 Dose range study for GHRP-6 and L-163,255 in GHS-R^''® GC cells 59 Figure 3.6 Somatostatin had no effect on GHRP-2 induced PI turnover in GHS-R'^''®GC cells 60 Figure 3.7 Transfection of GHS-R^''® GC cells with the hGHRH-R 61 Figure 3.8 Physiology of GHRH-R/GHS-R+"' GC cells 62 Figure 4.1 Generation of the rGHRH-eGFP construct 69 Figure 4.2 Analysis of rGHRH-eGFP transgenic mice 70 Figure 4.3 RT-PCR analysis of hypothalamic eGFP expression in rGHRH-eGFP transgenic mice 71 Figure 4.4 In situ hybridisation of eGFP mRNA in rGHRH-eGFP transgenic mice 72 Figure 4.5 High resolution imaging of eGFP in situ hybridisation sections 73

VI Figure 4.6 In situ hybridisation of GHRH mRNA in WT mice 74 Figure 4.7 High resolution imaging of GHRH in situ hybridisation sections 74 Figure 4.8 Double in situ hybridisation for mGHRH and eGFP mRNA 75 Figure 4.9 eGFP expression in the kidney of rGHRH-eGFP transgenic mice 76 Figure 4.10 Hypothalamic eGFP protein content in rGHRH-eGFP transgenic mice 77 Figure 4.11 Confocal microscopy of live rGHRH-eGFP hypothalamic slice preparations 79 Figure 4.12 Confocal microscopy of rGHRH-eGFP hypothalamic ARC cultures 80 Figure 4.13 Construction of the hGH/hGHS-R chimeric Mlul fragment 86 Figure 4.14 Assembly of the final rGHRH-hGHS-R construct 87 Figure 4.15 Analysis of rGHRH-hGHS-R transgenic mice 88 Figure 4.16 RNAse protection analysis of the rGHRH-hGHS-R transgene 90 Figure 4.17 Hypothalamic transgene expression in rGHRH-hGHS-R transgenic mice 91 Figure 4.18 Immunocytochemistry for hGHS-R protein in rGHRH-hGHS-R transgenic and WT mice 93 Figure 5.1 Confocal microscopy of the ME in rGHRH-eGFP transgenic mice after the addition of Ipg GHRP-6 99 Figure 5.2 Body weight of rGHRH-hGHS-R transgenic and WT mice 102 Figure 5.3 Pituitary GH content in adult rGHRH-hGHS-R transgenic mice and WT littermates 103 Figure 5.4 In situ hybridisation of mouse GHRH mRNA levels in rGHRH-hGHS-R transgenic and WT mice 104 Figure 5.5 Mouse GHRH RNAse protection assay of hypothalamic extracts from WT, rGHRH-eGFP and rGHRH-hGHS-R transgenic mice 105 Figure 5.6 In vivo GH responses to GHRP-6 in male rGHRH-hGHS-R transgenic mice and WT littermates 108 Figure 5.7 In vivo GH responses to GHRH in male rGHRH-hGHS-R transgenic and WT mice 109 Figure 6.1 Construction of the hGH-eGFP transgene 116 Figure 6.2 Analysis of transgenic hGH-eGFP mice by PCR 117 Figure 6.3 Fluorescent confocal microscopy of an adult male hGH-eGFP transgenic mouse pituitary 118 Figure 6.4 Construction of the hGH-hGHS-R transgene 121 Figure 6.5 Analysis of hGH-hGHS-R transgenic mice 122 Figure 6.6 Pituitary GHS-R expression in hGH-hGHS-R transgenic and WT mice 124 Figure 6.7 Immunocytochemistry for GHS-R protein in hGH-hGHS-R transgenic and WT mouse pituitaries 125 Figure 7.1 Female hGH-hGHS-R and WT body weight sorted for founders 128 Figure 7.2 Pituitary GH content in hGH-hGHS-R transgenic and WT mice 130 Figure 7.3 Pituitary Prolactin content in hGH-hGHS-R transgenic mice 131 Figure 7.4 In vivo GH responses to GHRP-6 and GHRH in female hGH-hGHS-R transgenic and WT mice 132 Figure 7.5 In vivo GH responses to GHRP-6 and GHRH in male hGH-hGHS-R transgenic and WT mice 133 Figure 7.6 In vitro responses of female hGH-hGHS-R transgenic and WT pituitaries to GHRH treatment 134 Figure 8.1 Mouse Ghrelin genomic structure in stomach and heart 141

Vll Figure 8.2 Ghrelin expression in mouse tissues 142 Figure 8.3 High resolution imaging of mouse Ghrelin expression in stomach and gut 143 Figure 8.4 Ghrelin RNAse Protection Assay and RT-PCR 145 Figure 9.1 Body weight gain of L-163,255 treated (13.2mg/kg/day) adult Tgr and WT rats 152 Figure 9.2 Pituitary GH content of L-163,255 treated (13.2mg/kg/day) adult Tgr rats 153 Figure 9.3 ARC NPY mRNA expression in orally L-163,255 treated Tgr and WT rats 154 Figure 9.4 Body weight gain of GHRP-6 treated (25|ig i.p. daily) adult female rGHRH- hGHS-R and WT mice 158 Figure 9.5 Pituitary GH content of GHRP-6 treated (25|ig i.p. daily) adult female rGHRH-hGHS-R and WT mice 159 Figure 9.6 Soleus and fat pad weight of GHRP-6 treated (25|Xg ip daily) adult female rGHRH-hGHS-R and WT mice 160 Figure 9.7 Body weight gain of GHRP-6 treated (2x15pg s.c. daily) adult male rGHRH- hGHS-R and WT mice 161 Figure 9.8 Pituitary GH content of GHRP-6 treated (2x15pg s.c. per day) adult male rGHRH-hGHS-R and WT mice 162 Figure 9.9 Pituitary Prl content of GHRP-6 treated (2xl5|ig sc per day) adult male rGHRH-hGHS-R and WT mice 163 Figure 9.10 Epiphysial growth plate width and soleus weight in GHRP-6 treated (2x15pg s.c. per day) adult male rGHRH-hGHS-R and WT mice 164

Table 2.1 Primer sequences for DNA fragment amplification in PCR and RT-PCR 44 Table 5.1 Body weights and pituitary GH content of rGHRH-eGFP transgenic mice 98 Table 5.2 Growth parameters in adult female rGHRH-hGHS-R transgenic and WT mice 106 Table 7.1 Body weight of hGH-hGHS-R transgenic mice 128

Vlll Acknowledgements

I am very grateful to Prof. Iain Robinson for his enthusiasm, never-ending encouragement and confidence throughout these three years. His time spent on my ‘Do you have a minute?’ questions is inspiring. I would also like to thank Dr. Pam Bennett for taking me under her wings with such encouraging patience and friendship. I am greatly indebted to all the members of the Robinson lab for much needed help and advice and for creating such a pleasant group. Thanks to everyone in Neurophysiology for making it a really great place to work in.

I am grateful to Roy Smith (Baylor College, Houston, Texas) and Andy Howard (Merck Research Laboratories, Rahway, New Jersey) for kind gifts of the hGHS-R type la cDNA, several antibodies, growth hormone secretagogues, sharing ‘secret’ knowledge and most of all for funding this Ph.D. I would also like to thank Nancy Cooke for the donation of the hGH cosmid and Bill Baumbach (American Cyanide) for the GHRH-R cDNA.

My very special thanks to Matt for always, always being there for me; he surely must be the most understanding and supportive man ever.

Many, many thanks to my family: so far and yet so close!

IX Author’s Declaration

All of the work contained in this thesis was, unless otherwise stated, performed by myself and has not previously been submitted for any degree or diploma to this, or any other University examining body.

The views expressed in this thesis are my own and not those of the University.

Nina Balthasar Abbreviations Small letters in front of some abbreviations specify the species: e.g. hGH is human growth hormone. h human m mouse r rat b bovine aa amino acid aCSF artificial cerebrospinal fluid ACTH adrenocorticotrophic hormone ARC arcuate nucleus AC adenylyl cyclase AVP arginine vasopressin BAC bacterial artificial chromosome bp base pair cAMP cyclic adenosine-mono-phosphate CREB cAMP response element binding protein CRH corticotrophin releasing hormone DAG diacetyl glycerol DM dorsomedial nucleus eGFP enhanced green fluorescent protein FM family member FSH follicle stimulating hormone GPCR G protein coupled receptor GH growth hormone GH-R growth hormone receptor GHRH growth hormone releasing hormone GHRH-R growth hormone releasing hormone receptor GHRH-RP growth hormone releasing hormone related peptide GHRP growth hormone releasing peptide GHS growth hormone secretagogue GHS-R growth hormone secretagogue receptor

XI HS hypersensitive site IP3 inositol-tris-phosphate IGF-I insulin-like growth factor JAK Janus kinase Kb kilo base KO knock-out LHN lateral hypothalamic nucleus LCR locus control region LH luteinizing hormone MSG monosodium glutamate MSH melanocyte stimulating hormone MT metallothionein MTL-R motilin receptor NMU neuromedin U NPY neuropeptide Y n.s. not significant NT-R neurotensin receptor NT non-transgenic ORF open reading frame PVN paraventricular nucleus PEN periventricular nucleus PMA phorbol myristate acetate PBPCK phosphoenolpyruvate carboxykinase PI phosphoinositol Prl prolactin PKA protein kinase A PLC phospholipase C RACE rapid amplification of cDNA ends RIA radio immuno assay RPA RNAse protection assay SDR spontaneous dwarf rat SS somatostatin

X ll sstr somatostatin receptor SCN suprachiasmatic nucleus SON supraoptic nucleus TSH thyroid stimulating hormone T transgenic Tgr transgenic growth retarded rat TM transmembrane VMN ventromedial nucleus WT wild-type

Xlll Chapter 1 : Introduction

1 Introduction 1.1 Aims of the thesis Growth hormone (GH) secretion from the pituitary is regulated by the hypothalamic peptides GH releasing hormone (GHRH) and somatostatin (SS). In addition, a novel class of GH regulators has been identified: the artificial peptidyl and non-peptidyl growth hormone secretagogues (GHS) and more recently an endogenous acylated peptide named Ghrelin (Bowers et al., 1977; Kojima et al, 1999; Smith et al, 1996). Ghrelin and the GHSs bind to their G-protein coupled receptors (GHS-R) in pituitary somatotrophs and hypothalamic neurones, causing GH release by both pituitary and hypothalamic actions (Howard et al, 1996). Although both hypothalamic and pituitary GHS effects are well documented (see Chapter 1.4), the relative importance of these two sites in GH release is still unclear.

In brief, the aim of this study is to further elucidate hypothalamic and pituitary GHS physiology by in vitro studies and cell specific transgenesis. I intended to distinguish hypothalamic and pituitary GHS actions by enhancing GHS-R expression either in hypothalamic GHRH neurones or pituitary GH cells in transgenic mice. This would allow analysis of differential effects of hypothalamic or pituitary GHS-R over­ expression on the GHS axis. I aimed to investigate whether constitutive over-expression of the GHS-R in either location would have an effect on GH physiology in absence of the exogenous ligand, or whether merely GHS responses might be altered. Several studies in the hypothalamus have implicated GHRH as a major mediator of the hypothalamic GHS signal on GH release. However, endogenous GHS-R mRNA expression in the arcuate nucleus of the hypothalamus was shown to co-localise mainly with neuropeptide Y (NPY) and to a lesser extent with GHRH (Tannenbaum et al, 1998a; Willesen etal, 1999). Similarly, the percentage of GHS responsive neurones co- localising with NPY was much greater than that co-localising with GHRH (Dickson and Luckman, 1997; Kamegai et al, 1996). NPY is strongly implicated in the regulation of food intake and GHS treatment does affect this axis (Locke et al, 1995; Okada et al, 1996). However in this study the aim was to selectively analyse GHRH neurones as a direct target for GHSs to mediate their effects on the GH axis, rather than analysing the Chapter 1 : Introduction metabolic GHS axis. Therefore I aimed to target GHS-R over-expression to GHRH neurones to analyse whether the alteration of the GHS-R abundance specifically in these neurones would alter GHS responsiveness in transgenic mice. Several studies have shown that the pituitary also acts as a direct target for GHSs. One of the questions was whether selective enhancement of the pituitary GHS-R would alter GHS’ effect on GH release, or whether GH release in response to GHSs could only be altered in concert with the hypothalamic GHS action. If GHS responses in transgenic mice over-expressing the GHS-R in the pituitary were altered, would those alterations be different to those one might observe in hypothalamic GHS-R over-expression? To elucidate the importance of the pituitary GHS input into GH release I thus aimed to over-express the GHS-R in pituitary GH cells.

The intention was to target GHS-R over-expression to GHRH neurones or anterior pituitary somatotrophs by placing the hGHS-R type la cDNA under GHRH or GH promoter control. While the cosmid targeting transgenes to GH cells has previously been shown to contain all the necessary regulatory sequences for reliable transgene expression (Bennani-Baiti et al, 1998; Jones et al, 1995), for the GHRH cosmid (Flavell et al, 1996) these expression studies have not been fully carried out. Ideally, the components that were to make up the constmcts targeting GHS-R expression to these cells in vivo should first be tested in other simpler systems. I therefore aimed to test the functional activity of the hGHS-R type la cDNA in a stably transfected GH cell line. This cell system would also have the advantage of allowing further elucidation of the GHS-R signalling system in vitro. In order to analyse expression driven by the GHRH targeting vector in transgenic mice, a biologically inert transgene, not involved in GH physiology, such as enhanced green fluorescent protein (eGFP) was chosen to be expressed under GHRH promoter control. The advantage of such an approach is that these transgenic mice are very useful tools in themselves. One of the problems when working with hypothalamic neurones or pituitary cells is in identifying these cells in vivo. Transgenic mice expressing eGFP in GHRH or GH cells would overcome this problem and would greatly facilitate the study of GHRH and GH cell physiology, by visualising intracellular distribution and secretion events. This is of particular interest in GHS physiology, since secretory events, especially in hypothalamic GHRH neurones in Chapter 1 : Introduction response to GHSs, are still unclear. The analysis of transgenic mice expressing eGFP in GH cells was not a major part of this thesis and will only be described briefly in some chapters.

The reasoning behind creating transgenic mice rather than rats was to enable future experiments crossing these GHS-R over-expressing mice onto GHS-R KO mice (a generous gift from R.G. Smith, Baylor College, Houston). This cross would then completely eliminate endogenous mouse GHS-R background signalling and entirely separate the hypothalamic from the pituitary pathway of GHSs.

Though not an issue at the beginning of this project, Ghrelin was recently identified as an endogenous GH releasing ligand of the GHS-R in rats and humans (Kojima et al, 1999). I therefore aimed to analyse Ghrelin expression in mice, especially to study whether Ghrelin expression was changed in GHS-R over-expressing transgenic mice.

1.2 Anatomical aspects of the Hypothalamo-Pituitary axis

1.2.1 Hypothalamo-Pituitary Communication Hypothalamic function becomes manifest through efferent pathways to autonomic nuclei in the brain stem and spinal cord, and through an intimate relationship with the pituitary gland. Two different mechanisms of hypothalamo-pituitary communication are used. The posterior pituitary contains axon terminals of secretory, magnocellular neurones originating in nuclei of the hypothalamus (Figure 1.1). The cell bodies of these neurosecretory cells produce two hormones, oxytocin and vasopressin, which are transported down the axons to the posterior pituitary by carrier-proteins called neurophysins and stored in neurosecretory granules in the axon terminals. In response to nerve impulses the hormones are then released from the nerve endings to penetrate capillary vessels in the posterior pituitary (for review see (Leng et ai, 1999)). The anterior pituitary, more relevant to this thesis, contains at least five different endocrine cell types, with hormone production under tight influence of hypothalamic regulating hormones or factors. The hypothalamic stimulatory or inhibitory hormones are produced in the cell bodies of neurones of several different hypothalamic nuclei. Chapter 1 : Introduction from which the neurones project to the median eminence (ME) (Figure 1.1). From their terminals they release their products into the hypophyseal blood system, which delivers them to the anterior pituitary cells.

Hypothalamus

Magnocellular Neurons Median eminence Posterior Pituitary

Hypophyseal Portal Vessel Posterior Hypophyseal Veins

Target Cell

Anterior Pituitary

Figure 1.1 Hypothalamo-Pltultary communication On the left hypothalamic neurones projecting to the median eminence release their neuropeptide (yellow) into the hypophyseal blood system, which transports it to the site of action in the anterior pituitary. Magnocellular neurones (on the right) project directly to the posterior pituitary and release their product (blue) into capillary vessels.

1.2.2 Hypothalamus Architecture The hypothalamus contains two major subdivisions, a medial, nuclear component and a lateral component containing the medial forebrain bundle. The medial zone is subdivided into three regions: anterior, tuberal and mamillary. Important nuclei of the medial zone for this thesis are the periventricular (PEN) nucleus in the anterior region, and the arcuate (ARC), ventromedial (VMN) and dorsomedial (DM) nuclei of the tuberal region (Figure 1.2). Hypothalamic function is diverse, for example among others its neurones regulate several anterior pituitary hormones, water balance, energy metabolism, circadian rhythm Chapter 1 : Introduction

Anterior Zone Tubular Zone

P

SON SON ME

Figure 1.2 Hypothalamus Architecture Depicted are typical coronal sections ttirougti ttie anterior and tuberal region of the medial hypothalamus and nuclei relevant to this thesis are highlighted. Abbreviations are: supraoptic (SON), suprachiasmatic (SCN), periventricular (PEN), paraventricular (PVN), arcuate (ARC), ventromedial (VMN) and dorsomedial (DM) nucleus, median eminence (ME). and stress responses. In keeping with its array of functions, hypothalamic nuclei contain neurones expressing numerous neuropeptides, transmitters and receptors. Hypothalamic peptides regulating the anterior pituitary are corticotropin-, thyrotropin-, gonadotropin-, prolactin- and GH releasing hormone. Together they influence systems as diverse as steroid metabolism, thyroid function, reproduction, lactation and growth. Expression patterns of hypothalamic neuropeptides and receptors and the axonal projections of their neurones involved in the regulation of GH secretion, such as GH releasing hormone (GHRH), somatostatin (SS), the GH- and GHS- receptor will be discussed in chapters 1.3.1-3 and 1.4.4. In addition to expressing several of the GH regulating hormones and receptors, the hypothalamic arcuate nucleus was found to be involved in the hypothalamic control of feeding. Numerous peptides expressed in the arcuate nucleus are implemented in metabolic control, most of these under the close regulation of the anti-obesity factor leptin (Ahima et aL, 2000). Neuropeptide Y (NPY) is one of these, but in addition to its function in energy metabolism it is of importance here, because of its role in the feed-back mechanism of GH secretion (Kamegai et aL, 1994) (see chapter 1.3.4) and GHS physiology (Dickson and Luckman, 1997; Kamegai etal., 1996) (see Chapter 1 : Introduction chapter 1.4.3.2). In general, hypothalamic neurones not only project directly to the ME to release their neuropeptide from the nerve terminals, but also to other neuronal targets in and beyond the hypothalamus.

1.2.3 Somatotroph Differentiation The five endocrine cell types of the anterior pituitary and the hormones they produce are: - somatotrophs (growth hormone (GH)) - lactotrophs (prolactin (Prl) - corticotrophs (adrenocorticotrophic hormone (ACTH) and melanocyte stimulating hormone (MSH)) - thyrotrophs (thyroid stimulating hormone (TSH)) - gonadotrophs (follicle stimulating hormone (FSH) and luteinizing hormone (LH)) Elegant transgenic studies suggested that cells co-expressing two types of hormones, GH and Prl, might derive from common GH-expressing stem cells (Behringer et ai, 1988; Borrelli et al, 1989) and these were called mammosomatotrophs. Research into the molecular mechanisms underlying somatotroph differentiation has demonstrated cw-acting elements essential for cell-specific expression of the GH gene and a cell-specific transcription factor that binds these elements was isolated (Bodner et ai, 1988; Ingraham et al, 1988). Mutations in this transcription factor (Pit-1) result in dwarfism, as demonstrated for example in the Snell dwarf mutations (Li et al, 1990). These mice are deficient in GH, Prl and TSH and have no detectable somatotrophs, lactotrophs or thyrotrophs, the three cell types in which Pit-1 expression is observed in the normal pituitary (Simmons et al, 1990). The Ames dwarf mouse pituitary phenotype is identical to that of the Snell dwarf and it was shown that mutation of the gene Prop-1 (Prophet of Pit-1) caused failure to activate Pit-1 transcription (Somson et al, 1996) thereby leading to GH, Prl and TSH deficiency. These observations suggested that regulation of Pit-1 expression is essential for proliferation regulation of somatotrophs and their precursors. Chapter 1: Introduction

1.3 Regulation of the Hypothalamo-Pltultary Growth Hormone Axis 1.3.1 GH - the gene, the receptor and its actions GH is the major endocrine regulator of post-natal growth in mammals. Identified due to growth promoting effects of bovine anterior pituitary extracts in rats, bovine GH was one of the first anterior pituitary hormones to be purified (Li and Evans, 1944). Human GH followed (Li and Papkoff, 1956) and sequencing revealed it to be a 22kDa, 191 amino acid protein with two disulphide bridges (Li and Evans, 1969). Rat GH was shown to be 190 amino acids long with 66% homology to its human counterpart (Seeburg et al, 1977). Mouse GH also consists of 190 amino acids and is highly homologous to rat GH (95%) (Linzer and Talamantes, 1985). While the rat possesses a single GH gene, the human GH gene is comprised of a 5 gene cluster spanning 48kb (Barsh et al, 1983; Chen et al, 1989) (Figure 1.3). The most 5’ gene in the cluster, human GH-N, is exclusively expressed in the somatotroph and mammosomatotroph cells of the anterior pituitary, where it comprises 3% of the total pituitary mRNA (Chen et al, 1989; Lewis, 1992). Deletion or mutation of the hGH-N gene results in isolated growth hormone deficiency, causing metabolic alterations and growth failure (Cogan et a l, 1994). All four other genes in the cluster, human chorionic somatomammotropin-A (hCS-A), hCS-B, hCS-L and hGH-V are expressed only in the placenta, hCS-A and hCS-B at high expression levels, whereas hCS-L and hGH-V are expressed at trace levels (MacLeod et al, 1992; Misra-Press et al, 1994).

Pituitary Piacenta

hGH-N hCS-L hCS-A hGH-V hCS-B

Figure 1.3 The hGH ciuster The arrangement of the five genes in the hGH cluster is shown. The site of native expression, pituitary or placenta is noted above each gene. GH = growth hormone; CS = chorionic somatomammotropin Chapter 1 : Introduction

GH has both direct effects by binding to its receptor (GH-R) in target tissues and indirect effects via insulin-like growth factor I (IGF-I), thereby stimulating bone and soft tissue growth and influencing carbohydrate, protein and lipid metabolism. GH-R mRNA and protein is present in high concentrations in the liver (Leung et al., 1987) and has a widespread distribution in extra-hepatic tissues. GH-R CNS distribution will be described in connection with its role in GH feed-back (Chapter 1.3.4). The GH-R belongs to the family of cytokine receptors, whose transduction processes involve tyrosine kinases of the JAK2 (Janus Kinase) family (Argetsinger et at., 1993; Campbell et ai, 1993). Although several new signalling peptides involved in the GH signalling cascade have been identified (for review see Argetsinger and Carter-Su, 1996; Wojcik and Postel-Vinay, 1999), little is known regarding the specificity of signalling for GH and other cytokines. A truncated, soluble form of the receptor, known as GH binding protein (GHBP), lacks the transmembrane and intracellular domain of the GHR (Baumann et at., 1990; Leung et a i, 1987). While the GHBP in rodents arises at least in part from alternative splicing (Baumbach et ai, 1989), in humans GHBP is produced mostly by proteolytic cleavage of the GH-R (Sotiropoulos etal., 1993). The function of the binding protein is unclear, but it appears to modulate the level of circulating GH (Herington et al., 1991) and might play a role in transport of GH to target tissues (Fuh et al., 1992). Disruption of the GH- R/BP gene causes GH resistance with low serum IGF-I, elevated GH serum levels and very short stature in humans (Laron Syndrome) and mice expressing a disrupted GHR/BP gene (Laron et al., 1966; Zhou et al., 1997). Serum IGF-I production mainly occurs in the liver as a response to GH stimulation. However, direct effects of GH through its receptor are accompanied by local generation of IGF-I, suggesting that GH and IGF-I act in concert as dual effectors to stimulate growth (Green et al., 1985). Several groups have shown that GH and IGF-I have differential effects on the growth of individual organs; GH stimulates bone and liver growth, whereas IGF-I induces growth in other organs such as kidney and gut (Clark and Robinson, 1996). IGF-I disruption studies in mice have shown that total IGF-I deficiency is postnatally lethal with severe growth retardation (Liu et al., 1993; Powell- Braxton et al., 1993). However when IGF-I is disrupted only in the liver, overall growth in such mice was unaffected, suggesting that liver-derived IGF-I might be relatively Chapter 1 : Introduction unimportant in postnatal growth (for review see (Liu et al, 2000)). These data indicated that autocrine/paracrine-derived IGF-I might be more important for body growth. The role of IGF-I in GH feedback mechanisms will be described in chapter 1.3.4. For a schematic summary of the control of GH secretion, see Figure 1.8.

1.3.2 Growth hormone releasing hormone (GHRH) 1.3.2.1 The GHRH gene First evidence for hypothalamic control of GH secretion came from in vivo and in vitro experiments in the 1960s, demonstrating that destruction of the hypothalamus led to reduced growth in rats (Reichlin, 1960) and that hypothalamic extracts could release GH (Deuben and Meites, 1964). GHRH was isolated 20 years later from pancreatic tumours causing acromegaly in patients (thickening of bones due to GH hypersecretion) (Guillemin et al, 1982; Rivier et al, 1982) and later from human hypothalamus (Ling et al, 1984). Rat hypothalamic GHRH contains 43 amino acids and is stmcturally different from the 44 amino acid GHRH of many species (Spiess et al, 1983). Mouse GHRH consists of 42 amino acids with considerable amino acid divergence to GHRH of other species and 64% nucleic acid sequence homology to rat and human GHRH (Miki et al, 1994; Suhr et al, 1989). Nevertheless all GHRH peptides can release GH across species. The GHRH gene is a single gene; while exons 2-5 are common in all tissues, an alternative first exon is either placental- or brain-specific (Gonzalez-Crespo and Boronat, 1991). Because these mRNAs are spliced to encode identical precursor proteins, the hormone in the hypothalamus and placenta is structurally identical, but regulated in a tissue-specific manner by alternative promoter usage (Mayo et al, 1995).

The active GH stimulating GHRH is derived by mutiple posttraslational processing steps (see Figure 1.4) (Nillni et al, 1999). The signal leader sequence (amino acid (aa)l-30) and a 30 aa C-terminal peptide are cleaved from the 104 aa GHRH precursor to form the active GHRH molecule (31-73aa) (Mayo et al, 1985; Mayo et al, 1983). In humans formation of the active GHRH is followed by amidation of the terminal glycine, giving the major GHRH form produced by the hypothalamus (Frohman, L. et al, 1989). Rat and mouse GHRH are non-amidated (Frohman, M. et al, 1989; Mayo et al, 1985). The 30 aa C-terminal peptide cleaved from the 104 aa GHRH precursor was called GHRH Chapter 1: Introduction

related peptide (GHRH-RP) and shown to be a potent stimulator of stem cell factor and critical for normal spermatogenesis (Breyer et ai, 1996; Fang et ai, 2000). The active GHRH and GHRH-RP proteins are further processed into smaller peptides but their functionality has yet to be established (Nillni et ai, 1999).

ATG TGA AATAAA I Gene:

mRNA: 5'UTR S P M i g c i n i n B g l GHRH-RP 13'UTR 5'UTR S P T/GHRH //! GHRH-RP 3'UTR

precursor protein: SP K/ZqHRH GHRH-RP 1 31 73 104 I K / / .9 H " h X / Z I g h r h -r p 73 104 7 \ GHRH-RP 31 73 75 104

Amidation in humans

31 73

Figure 1.4 Biosynthesis of GHRH The structures of the GHRH gene, mRNA and posttranslational processing steps are schematically presented. PI and HI represent the first exon used in placenta and hypothalamus. Exons 2-5 are common. Use of alternative first exons results in two distinct mRNAs, encoding an identical GHRH precursor peptide. Cleavage of the precursor GHRH results in the active, GH stimulating GHRH protein (31-73aa), amidated in humans but not in rodents, and the GHRH related peptide (GHRH-RP), shown to stimulate stem cell factor(modified from (Nillni etal., 1999)).

10 Chapter 1 : Introduction

13.2.2 Expression ofGHRH Early studies, showed GHRH immunoreactivity in the ARC of the rat hypothalamus, with fibres projecting to the ME (Merchenthaler et al, 1984; Sawchenko et al, 1985) and GHRH secretion from these into the portal blood system has been measured in rats and sheep (Frohman et al, 1990; Plotsky and Vale, 1985). Low levels of GHRH immunoreactivity were found in the PVN, DM and the medial and lateral borders of the VMN. In the CNS no substantial expression was observed in brain regions outside the hypothalamus (Merchenthaler gr aA, 1984; Sawchenko et al, 1985). Retrograde tract- tracing techniques in combination with immunocytochemistry in mice (Romero and Phelps, 1997) and rats (Niimi et al, 1989) have provided further evidence for ARC neurones being the main source of GHRH-containing projections to the ME. In situ hybridisation confirmed that cells on the ventral surface of the hypothalamus with additional cells extending into the DM are the main site of GHRH synthesis (Berelowitz et al, 1992; Mayo, 1989; Zeitler et al, 1990). Hypothalamic GHRH mRNA was shown to be sexually dimorphic, with males expressing greater levels than females (Argente et al, 1991). Expression of GHRH outside the hypothalamus and placenta was measured by RT-PCR and detected in the ovary and testes (Bagnato et al, 1992; Berry and Pescovitz, 1988) and by immunocytochemistry in the pancreas and gastrointestinal tract (Bosman et al, 1984). However little is known about the mRNA transcripts or peptide forms expressed in these extrahypothalamic tissues. Increases in pituitary GH gene expression, GH synthesis and secretion (Barinaga et al, 1985a,b; Bilezikjian and Vale, 1984; Fukata et al, 1985; Gick et al, 1984; Tanner et al, 1990) are controlled by GHRH. Furthermore, GHRH induces the expression of the protooncogene c-fos and stimulates the proliferation of pituitary somatotroph cells (Billestrup et al, 1987; Billestrup et al, 1986). These data suggest a strong trophic effect of GHRH on the pituitary. Transgenic mice over-expressing human GHRH under ubiquitous metallothionein (MT) promoter control (Hammer et a l, 1985) show enhanced growth, high serum GH levels, pituitary hyperplasia and adenoma (Mayo et a l, 1988), thereby reflecting GHRH control of pituitary GH. Ectopic GHRH over­ expressing tumours in acromegalic patients cause a very similar phenotype, which wanes after removal of the tumour (Thomer et al, 1982).

11 Chapter 1 : Introduction

An important factor in GHRH mRNA control is feedback regulation through GH, mediated directly by GH-R expressed in the hypothalamus (see chapter 1.3.4). GH deficiency is associated with increased GHRH mRNA levels (Chomczynski et al, 1988), while GH treatment decreases GHRH levels (de Gennaro Colonna et al, 1988). Regulation based on GH status has been observed in genetically manipulated growth hormone deficiency and excess syndromes (for review see Phelps and Hurley, 1999). In accordance with their low GH levels, Ames and Snell dwarf mice demonstrate increased levels of GHRH expression (Phelps et al, 1993), while transgenic giant mice expressing GH under the ubiquitous MT promoter or the phosphoenolpyruvate carboxykinase (PEPCK) promoter (Steger et al, 1991) show markedly reduced GHRH mRNA levels (Phelps and Hurley, 1999). For a schematic summary of GHRH’s involvement in the control of GH secretion, see figure 1.8.

1.3.2.3 The GHRH receptor (GHRH-R) Three groups achieved the molecular cloning of anterior pituitary GHRH receptors (GHRH-R) in human, rat and mouse (Gaylinn et al, 1993; Lin et al, 1992; Mayo, 1992). The GHRH-R was found to be a 423 aa seven transmembrane G protein-coupled receptor in all three species, and its expression was localised to GH cells of the anterior pituitary (Lin et al, 1992). Female rats were shown to express markedly lower levels of GHRH-R mRNA than males (Ono et al, 1995), but Carmignac et al were unable to confirm these data (Carmignac et al, 1996). Expression levels of GHRH-R mRNA are up-regulated by GHRH itself (Miller and Mayo, 1997) and by Pit-1 (Lin et al, 1992). The GHRH-R is absent in dwarf Snell and Ames mice, which are impaired in Pit-1 expression, suggesting that GHRH-R expression might be Pit-1 dependent (Lin et al, 1992). Early studies showed that GH release in response to GHRH was associated with stimulation of cAMP production (Bilezikjian and Vale, 1984) and that extracellular Ca^^ was required for this cAMP accumulation, suggesting that Ca^^ might act independent of cAMP (Brazeau et al, 1982). Transfection of COS or human kidney 293 cells with the GHRH-R demonstrated that GHRH stimulated the accumulation of cAMP as well as a cAMP-responsive reporter gene (Gaylinn et al, 1993; Mayo, 1992), consistent with the predicted coupling of the GHRH-R to a G^ protein. Although no complete cAMP responsive elements have been found in the GH gene promoter (Bertherat et al, 1995),

12 Chapter 1 : Introduction the importance of the cAMP response element binding protein (CREB) was demonstrated by pituitary hypoplasia and dwarfism in transgenic mice over expressing an inactive CREB (Struthers et al, 1991). The model for GHRH signalling in pituitary somatotrophs is the following (see Figure 1.5): GHRH interaction with its receptor leads rapidly to GH secretion, possibly by a G-protein mediated interaction with ion channels or through ion channel phosphorylation, and to stimulation of intracellular cAMP through stimulation of adenylate cyclase (AC) by the G, protein (Mayo et al, 1996). Elevated cAMP results in the phosphorylation and activation of CREB by protein kinase A (PKA) (Gonzalez and Montminy, 1989; Sheng et al, 1991) and then enhanced Pit-1 transcription (Bodner et al, 1988; Ingraham et al, 1988; McCormick et al, 1990). Pit-1 activates transcription of the GH gene to elevate GH mRNA and protein and also activates GHRH-R transcription (Lin et al, 1992). The importance of GHRH-R function for pituitary somatotroph development is illustrated by the dwarfed little mouse. A point mutation in the GHRH-R gene is responsible for its severe GH deficiency (Godfrey et al, 1993; Lin et al, 1993). Its pituitary somatotroph (Wilson et al, 1988) and GH levels (Cheng et al, 1983) are markedly reduced, but detectable. The few somatotrophs left are unresponsive to GHRH, but do release GH in response to cAMP (Clark and Robinson, 1985; Jansson et al, 1986a). Recently Gaylinn et al reported that the little mutation did not dramatically affect the expression level, glycosylation, or cellular localization of the receptor protein, but that it blocked specific GHRH binding and therefore inhibited signalling (Gaylinn et a l, 1999). In humans, mutation of the GHRH-R causes a similar GH deficient phenotype (Baumann and Maheshwari, 1997; Wajnrajch et al, 1996).

1.3.3 Somatostatin Somatostatin was identified much earlier than GHRH (Brazeau et al, 1973). SS exists in two major forms consisting of 14 or 28 aa, derived by endopeptidase cleavage of a 116 aa precursor, pre-pro SS (Van Itallie and Femstrom, 1983). Both SS14 and SS28 inhibit spontaneous and GHRH induced GH secretion in vivo and in vitro (Brazeau et al, 1981). In contrast to GHRH, the sequence of SS has remained highly conserved between

13 Chapter 1: Introduction

GHRH

'AC •as cAMP

PKA GH AAA GH GH GHRH-R AAA CREB-P

Pit-1 GH Pit-1

AAA

Figure 1.5 GHRH signalling pathway GHRH interaction wltti its receptor leads to rapid GH secretion, pertiaps ttirough ttie receptor-induced interaction of a G protein with ion channels (10) or 10 phosphorylation. Stimulation of the adenylyl cyclase (AO) by the Gg protein results in increased intracellular cAMP accumulation and activation of the catalytic subunit of protein kinase A (PKA). PKA phosphorylates OREB leading to increased Pit-1 transcription. Pit-1 then activates transcription and synthesis of the GH and the GHRH-R gene. Blue: DMA, Pink: mRNA (Modified from (Mayo et ai, 1996)). species. SS producing neurones are present throughout the central and peripheral nervous system, the gut, the endocrine pancreas and, in smaller numbers, in the thyroid, adrenals, submandibular glands, kidney, prostate and placenta (Patel, 1990; Reichlin, 1983). Within the hypothalamus, the majority of SS neurones lie within the PEN (Alpert et al, 1976) but these are also seen in the ARC and VMN (Finley et al, 1978). 78% of SS PEN neurones project to the external layer of the ME (Kawano and Daikoku, 1988); other neurones interconnect with several hypothalamic nuclei, including the ARC and VMN. The SS axons and perikarya in the ARC and VMN, two areas where GHRH neurones originate, indicate the existence of anatomic and functional peptide interactions. Furthermore, several somatostatinergic axon varicosities cluster around GHRH synthesising cells, suggesting synaptic associations (Liposits et al, 1988). Co­

14 Chapter 1 : Introduction localisation studies finally showed that SS receptor 1 (sstrl) and sstr2 were co-localised in part with ARC neurones expressing GHRH mRNA (McCarthy et al.y 1992; Tannenbaum et al., 1998b). These data provide strong anatomical support for a direct hypothalamic interaction of GHRH and SS to modulate GH secretion. Elegant studies by Zheng et al. with sstr2 knock-out (KO) mice gave further functional evidence in favour of this hypothesis (Zheng et al., 1997). Since GH pre-treatment prevented the activation of c-fos by GH secretagogues (GHS) in ARC neurones (GHSs activate these neurones in the ARC, see chapter 1.4.3.2) in wild-type (WT) but not sstr2 KG mice, the authors suggested that GH induced SS release from PEN neurones suppressed the activity of ARC neurones. This interpretation is further supported by electrophysiological experiments, where stimulation of PEN neurones resulted in inhibition of ARC neurones (Dickson et al., 1994). Willoughby et al. demonstrated that in the PEN only very few SS neurones were closely approached by GHRH immunoreactive fibres, providing possible evidence of scant innervation of SS neurones by GHRH cells (Willoughby et al, 1989).

SS’s lack of a trophic effect on the pituitary was shown by the absence of an effect on basal levels of GH synthesis, gene transcription, mRNA or somatotroph proliferation in rat or bovine pituitary cell cultures (Barinaga et al, 1985b; Billestrup et al, 1986; Fukata et a l, 1985; Simard et al, 1986; Tanner et al, 1990). However, it was shown that SS attenuated the GHRH-induced increases in somatotroph proliferation and c-fos response in the rat (Billestrup et al, 1987; Billestrup et al, 1986). At the pituitary level, SS was shown to inhibit spontaneous GH release and GHRH induced GH secretion (Brazeau et al, 1973; Fukata et al, 1985). As reported for GHRH, there is a sex-related difference in hypothalamic SS expression, with males expressing higher SS mRNA and protein levels than females (Argente et al, 1991; Nurhidayat et al, 1999). In contrast to GHRH, SS mRNA is down-regulated in GH deficiency and up-regulated in GH excess. This is demonstrated by the previously described Snell and Ames dwarf mice, who have reduced hypothalamic SS levels, while SS is elevated in mice, over-expressing GH or GHRH under MT promoter control (for review see Phelps and Hurley, 1999). The actions of SS are mediated by a family of G-protein coupled receptors (sstrl-5; for review see Patel, 1997). All five genes are expressed by the majority of rat pituitary cell

15 Chapter 1 : Introduction subtypes, with high levels of sstr4 and 5 in rat somatotrophs and sstr2 in thyrotrophs (O'Carroll and Krempels, 1995). CNS sstr expression is wide spread, with a particular pattern for each receptor subtype (Bruno et al, 1993). The functional role of the individual sstr subtypes is however unknown. In situ hybridisation in the hypothalamus showed that all sstr are expressed in the hypothalamus, sstrl and 2 being particularly abundant (Beaudet et ai, 1995; Breder et al, 1992). sstrl, 2 and 3 have been located in the ARC (Senaris et al, 1994) and for sstrl mRNA regulation by GH has been shown, implicating its potential involvement in feedback regulation of GH secretion (Guo et al, 1996). Interestingly sstrl expression was reported to co-localise with SS expressing neurones; the sstrl might thus act as an auto-receptor (Lanneau et al, 2000). As described previously, sstr2 KO mice strongly implicated this sstr in PEN-ARC communication (Zheng et al, 1997). Sstr are coupled to a G^-protein; ligand binding results in a decrease in adenylate cyclase, cAMP and PKA levels (for review see (Patel et al, 1995)). Therefore activation of sstr causes activation of K^ channels, hyperpolarisation of the membrane and reduction of intracellular Ca^"^ (Sims et al, 1991). This effect is enhanced by the blocking of Ca^^ channels (Ikeda and Schofield, 1989). These opposite effects to GHRH thus cause inhibition of GH release. For a schematic summary of SS’s involvement in the control of GH secretion see figure 1.8.

1.3.4 GH feedback mechanisms As mentioned before, a key regulator of GH secretion is GH itself. The feed-back mechanisms of GH have been described as (1) a short loop, involving the control of hypothalamic neuropeptides by GH, and (2) a long loop, involving IGF-I. GH’s actions are mediated by the GH-R. In the CNS, the main sites of expression are the hypothalamus and hippocampus. The hypothalamic GH-R expression was found to be in PEN SS neurones and in the ARC in a small percentage of GHRH cells (Burton et al, 1992; Minami et al, 1992) but predominantly in ARC NPY neurones (Kamegai et al, 1994). This localisation of the GH-R is consistent with the idea that GH feeds back to increase SS and decrease GHRH expression, when GH expression is high. It also implies that GH feedback inhibition on GHRH may be indirect via NPY. In an in vitro hypothalamus explant system, NPY decreased GHRH release and increased

16 Chapter 1: Introduction somatostatin, consistent with NPY’s role as GH feedback mediator (Korbonits et al, 1999a). It is interesting to note that NPY expression is GH sensitive (Chan et al, 1996), NPY expression being reduced in GH deficiency (Bennett et al, 1997). GH in turn is sensitive to NPY, with i.c.v. NPY injections causing reduced plasma and pituitary GH (Rettori et al, 1990), but i.v. injections of NPY increasing GH release (Catzeflis et al, 1993). In summary, current reports suggest that in addition to a direct feedback of GH onto SS and GHRH neurones, negative GH feedback is mediated by NPY. However, one has to bear in mind that GH effects caused by NPY might be primarily in the regulation of food intake and utilisation rather than growth regulation (Bennett and Robinson, 2000). Not only are the hypothalamic neuropeptides regulated by the GH state, but hypothalamic GH-R expression itself is decreased in GH deficiency and can be increased by GH treatment (Bennett et al, 1995). Physiologically, the short-loop GH mechanism becomes apparent in chronic cannulation experiments in rats, where GH inhibits its own secretion, blocking GH pulsatility (Abe et al, 1983; Chihara et al, 1981; Clark eta l, 1988; Willoughby etal, 1980). The long-loop GH feedback mechanism is operated by IGF-I. IGF-I receptors have been found in pituitary cells (Ceda, 1995) and IGF-I binding sites have been reported in the ME (Bohannon et al, 1986) and other brain areas (Goodyer et a l, 1984). IGF-I acts directly on the pituitary to inhibit GH gene expression and GHRH induced GH release (Yamashita and Melmed, 1986)(Ceda et al, 1987) In the hypothalamus IGF-I stimulates somatostatin synthesis and secretion (Berelowitz et al, 1981; Tannenbaum et al, 1983) and inhibits GHRH synthesis and release (Sato and Frohman, 1993). By all these mechanisms, IGF-I inhibits the posttranslational processing of GH and reduces GH exocytosis. GH feed-back mechanisms, involved in the control of GH secretion are summarised schematically in Figure 1.8.

1.3.5 GH pulsatility GH secretion from the pituitary is pulsatile in all species studied (Chihara et al, 1983; Davis et al, 1977; Jansson et al, 1985; Miller et al, 1982; Steiner et al, 1978). This pulsatility is important for regulating growth (Hindmarsh et al, 1987) and metabolism (Hindmarsh et al, 1997) and requires the co-ordinated interaction of the hypothalamic

17 Chapter 1 : Introduction peptides GHRH and SS, For example, infusions of GH in dwarfed dw/dw rats (Charlton et al, 1988), that are GH deficient for a yet unknown reason, have shown that the most effective way of GH repletion is by brief, large GH pulses (Gevers et al, 1996). The male rat shows regular GH secretory episodes every 3-3.5 h with very low GH concentrations between bursts (Tannenbaum and Martin, 1976), while female rats show a more continuous profile (Clark et al, 1987; Eden, 1979). Several studies have confirmed that pulsatile release of GH is dependent on episodic release of GHRH, whilst the magnitude of the GH response to GHRH is reliant on the prevailing SS tone (Tannenbaum, 1991). Both, GHRH and SS are necessary for generating the GH pulse pattern, since anti-serum and antagonist studies for SS and GHRH have shown that abolishing either one interrupts episodic GH secretion (Baumbach, W.R., Fairhall, K.M., Carmignac, D.F. and Robinson I.C.A.F., unpublished data, (Wehrenberg et al, 1982)). Trains of GH pulses will entrain the GH pulse generator in conscious male rats (Carlsson and Jansson, 1990), whilst continuous exposure to GH will suppress GH secretion (Abe et al, 1983; Clark et al, 1988). These models oversimplify the matter by excluding other inputs, for example the synthetic GHSs and GH feed-back mechanism.

Neuroendocrine mechanisms are most likely not the only input into the pulsatility of GH secretion and electrophysiological studies of the involved neurones might give novel insights into the regulation of pulsatile GH secretion. Could, for example, co-ordinated firing in a population of synchronised hypothalamic cells result in a pulse of hypothalamic hormone release? Dickson et al have shown previously that electrical stimulation of the ARC can elicit a pulse in GH secretion (Dickson et al, 1993a). A similar system has been described for magnocellular oxytocin neurones (for review see (Leng et al, 1999)). However, the exact mechanism that determines GH’s spontaneous pulse frequency remains elusive.

1.4 Artificial Growth Hormone Secretagogues (GHS) and Ghrelin

The pulsatile release of GH is fundamental to its biological actions (Hindmarsh et al, 1987). However, in GH deficient patients, GH replacement therapy currently takes the form of single daily injections of recombinant hormone, a regimen that clearly fails to

18 Chapter 1 : Introduction echo the endogenous GH release profile. Bowers et al, in search of GHRH, first identified small peptides that released GH in vitro and in vivo (Bowers et al, 1977) and it became clear later that these peptides were distinct from GHRH and had their own endogenous signalling cascade. These experiments then encouraged studies to identify other mimics or amplifiers of GH pulse pattern, which could be used as an alternative, more physiological treatment in GH deficiency and for the convenience of the patient could be taken orally.

1.4.1 Historical and Structural Aspects of GHS development In 1977, Bowers et al reported that synthetic enkephalin opiate analogues specifically released GH in vitro (Bowers et al, 1977). In a series of structure-activity experiments and theoretical conformational studies, the first in vitro and in vivo active hexapeptide GHRP-6 (GH releasing peptide-6; HisDTrpAlaTrpPheLysNHj, Figure 1.6) was evolved. Further modification of the original GHRP-6 resulted in GHRP-2, a more active GHS (Bowers, 1993), but unfortunately these potent GH releasing agents still had very low oral bioavailability. At this point a group at Merck designed a high throughput in vitro assay with rat primary cultures to screen for small, non-peptidylic GH releasing molecules. This lead to the identification of several benzolactam derivatives, which had excellent potency, selectivity and tolerability in animals, but still displayed relatively low bioavailabitlity (DeVita and Wyvratt Jr., 1996; Leung et al, 1996; Wyvratt Jr., 1996). Therefore a new lead was sought and a novel class of GHSs was designed from spiroindanes. Encouraging data in beagle dogs demonstrated high oral bioavailability and sustained increased GH and lGF-1 levels after a single oral dose (Jacks et al, 1996; Patchett et al, 1995) and MK-0677 entered clinical trials (Figure 1.6). 20 years after the first discovery of GHSs, an endogenous receptor (GHS-R) was cloned for these artificial agents (Howard et al, 1996), suggesting the existence of an endogenous ligand involved in GHS physiology. Finally, after 22 years of GHS research, an endogenous ligand for the GHS-R that released GH in vitro and in vivo was found and termed Ghrelin (Kojima et al, 1999). Ghrelin was a surprising discovery, as it was shown to be a 27 or 28 amino acid peptide with an octanoylated SerineB (Figure 1.6), the first peptide to ever show this kind of post-translational modification. Also

19 Chapter 1: Introduction surprising was its expression pattern, with very high levels of expression in the stomach and low levels of detectable peptide in the hypothalamus. Ghrelin is discussed extensively in Chapter 8.

His - DTrp - Ala - Trp - DPhe - Lys - NH 2 NH

CH .NH, NH. NH. NH NH NH.

HN HN NHc GHRP-6

NH; C = 0 0 I

N-SOpCH

L-163,191 (MK-0677)

0=C-(CH2)6-CH3 o 1 I 10 20 28 GSSFLSPEHQKAQ(Q)RKESKKPPAKLQPR ghrelin

Figure 1.6 Several classes of GHS

The three classes of GHSs - peptidylic GHRPs, non-peptidylic GHS and the endogenous peptide Ghrelin - seem structurally very different given that they bind to the same receptor. However, computer assisted overlays demonstrated that at least the GHRPs and non-peptidylic GHSs show three-dimensional similarities (Schoen et ai, 1994). Feighner et al. showed that in a three-dimensional model of GHS docking to the GHS-R, MK-0677 and GHRP-6 had similar complementarity to the GHS-R binding pocket (Feighner et al, 1998). However, it was also observed that while some mutations in

20 Chapter 1 : Introduction

distinct regions of the receptor affected all GHSs tested, other mutations were selective for particular GHSs, suggesting that the GHSs did not need to achieve an identical position in the GHS-R binding pocket to activate the receptor. These kind of data are not yet available for Ghrelin, but it will be interesting to see how Ghrelin fits into the GHS- R binding pocket, especially with its rather bulky octanoyl group. Other active and structurally very different GHSs were designed by several pharmaceutical companies, but will not be further discussed here. In general, the GHS-R seems to be activated by a broad range of agonists.

1.4.2 GH releasing actions of synthetic GHS The GH releasing activity of GHRP-6 has been demonstrated in a number of species (Bowers etal., 1984; Croom et al., 1984; Doscher a/., 1984; Fairhall et al., 1995; Kraft et al., 1984; Malozowski et al., 1991), including humans (Bowers et al, 1990; llson et al., 1989). GHSs are active at varying doses when given as a bolus injection (Bowers et al., 1990), orally (Bowers et al., 1992) or intranasally (Hayashi et ai, 1991). Acute i.v. injections of all GHSs rapidly increase serum GH concentrations within 5-10 minutes, with the peak GH concentration being observed 15-30 minutes after injection (Bercu et ai, 1992; Bowers et al., 1990; Imbimbo et ai, 1994; Kirk et al., 1997; Massoud et ai, 1996; Tiulpakov et al., 1995). GHSs are the most efficacious GH releasing agents in experimental animals and humans; the amount of GH secreted after GHSs is much larger than that secreted after GHRH (Korbonits and Grossman, 1995; Massoud et al., 1996; Tiulpakov et al, 1995). Furthermore, GHRP-6 showed less intra­ subject variability than GHRH (Ghigo et al, 1994; Huhn et al, 1993). Although prolonged infusion of GHRP-6 in humans enhanced pulsatile GH secretion and raised plasma lGF-1 concentrations (Huhn et al, 1993; Jaffe et al, 1993), GH responses to a subsequent GHRP-6 challenge were then attenuated. Continuous infusion of GHSs caused attenuation of GH responses, but repeated daily oral or intranasal doses of non-peptidylic GHS induced no signs of such desensitisation (Chapman et al, 1996; Ghigo et al, 1996; Ghigo et al, 1994; Pihoker et al, 1997). Still, a 16 week treatment with twice daily s.c. Hexarelin, the commercial GHRP-6, resulted in a partial and reversible attenuation of the GH response (Rahim et al, 1998) and a study with beagle dogs demonstrated that a reduced amplitude of the GH response to MK-0677 might stem

21 Chapter 1 : Introduction from enhanced IGF-I levels (Hickey et al, 1997). The authors suggested that tachyphylaxis might be associated with IGF-I feedback on the GH/GHRH axis. However MK-0677 can stimulate and maintain pulsatile GH release and increase plasma IGF-I concentrations during oral treatment over hours to weeks in young, older and obese men and women (Chapman et al, 1996). Even in certain conditions of GH axis pathophysiology, such as obesity (Svensson et al, 1998), starvation (Murphy et al, 1998), critical illness (Van den Berghe et al, 1998; Van den Berghe et al, 1997; Van den Berghe et al, 1997; Van den Berghe et al, 1999) and ageing (Aloi et al, 1994; Chapman et al, 1996; Merriam et al, 1997) GHSs can release GH. GHSs might also play a role in GH pulsatility. Fairhall et al showed that GH responses to GHSs were dependent on the pattern of administration; more consistent responses were obtained with less frequent injections, whereas continuous exposure to GHSs lead to amplified endogenous secretion (Fairhall eta l, 1995). There is evidence suggesting that the age-related changes in bone and muscle mass and increased adiposity may, at least in part, depend on decreased GH secretion, as indicated by the demonstration that treatment with exogenous GH counteracts them (Rudman et al, 1990). Treatment of healthy elderly men and women with once daily oral MK-0677 increased serum GH and IGF-I dose dependently to levels normally seen in young adults (Chapman et al, 1996). Murphy et al further demonstrated that 2-9 week once daily MK-0677 treatment increased indices of bone turnover (Murphy et al, 1999). Although data point to the ability of GHSs to restore the reduced function of the GH/IGF-I axis in ageing, whether it is really useful to restore the GH/IGF-I axis to young levels remains unclear. Despite the ability of GHSs to generate impressive pulsatile GH secretion, attempts to stimulate growth in animals by GHS exposure have shown only modest body weight gain, mostly by once daily administration (Bowers e ta l, 1984; Hansen et al, 1999; McDowell et al, 1995; Svensson et al, 2000). Attempts to stimulate additional growth in animals must overcome the problem that the GH responses to GHSs do sometimes desensitise with time (Jacks et al, 1994). Growth responses to GHS are dependent on the pattern of GHS administration, with a continuous infusion being much less effective than twice daily injections (McDowell et al, 1995). Growth stimulation was demonstrated in GH deficient, but GHS responsive Tgr (Transgenic growth retarded)

22 Chapter 1 : Introduction rats (more detailed information in Chapter 9), where pulsatile infusion of GHRP-6 caused increased growth at twice the rate of saline infused Tgr littermates (Wells et al, 1997). Growth responses to GHSs will be discussed in detail in Chapter 9. One of the most attractive uses of GHSs is as diagnostic and therapeutic tools in GH deficiency. In a few GH deficient and insufficient children GHS treatment - intranasally or subcutaneously - resulted in a modest rise in growth (Laron, 1995; Mericq et al, 1995; Pihoker et al, 1997), not always accompanied by raised IGF-I levels. Other studies were less promising (Micic et al, 1999). Nevertheless, GHS treatment in GH deficient adults enhanced spontaneous GH secretion (Leal-Cerro et al, 1995) and MK- 0677 increased GH, IGF-I and IGF-BP3 in a group of severely GH deficient men (Chapman era/., 1997). A fact hampering the use of GHSs as therapeutic agents is that they are not specific for GH release. GHSs were shown to modify sleep pattern (Copinschi et al, 1997; Frieboes et al, 1995) and food intake in rats (Kuriyama et al, 2000; Locke et al, 1995; Luoni et a l, 1997; Okada et a l, 1996) and humans (Chapman et al, 1996). Furthermore prolonged MK-0677 treatment raised fasting blood glucose and insulin concentrations (Chapman et al, 1996; Gertz et al, 1993; Hickey et al, 1994). Maybe the most problematic side effect of GHSs is the small but consistent elevation in ACTH and cortisol they cause (Bowers et al, 1990; Copinschi et al, 1996; Frieboes et al, 1995; Hickey etal, 1994; Hickeyeta l, 1996; Jacks etal, 1994; Thomas et al, 1997). Several lines of investigation indicate that this effect is exerted indirectly via the hypothalamus, as hypothalamo-pituitary stalk lesions abolished the ACTH response to GHS and no ACTH was released in vitro from isolated pituitaries (Cheng et al, 1993a; Schleim et al, 1996), (Elias et al, 1995; Hickeyet al, 1996). Furthermore, studies with GHRP-6 in combination with CRH (corticotropin releasing hormone) or AVP (arginine vasopressin) by Thomas et al suggested that GHRP-6 acts via the hypothalamus to mediate the release of ACTH and that these effects were probably at least in part via the release of CRH to then synergise with AVP at the corticotroph (Thomas et al, 1997). In contrast Korbonits et al came to the conclusion that in humans and rat hypothalamic explants, the effect of hexarelin on the HP A axis does not involve CRH, but may occur through the stimulation of AVP release (Korbonits et al, 1999a; Korbonits et al, 1999b). GHSs also stimulate small, but significant amounts of Prl (Ciccarelli et al, 1996; Copinschi et

23 Chapter 1 ; Introduction al., 1996; Ghigo e ta l, 1997; Hickey et al, 1994; Kirkef a/., 1997; Massoud et al, 1996; Van den Berghe et al, 1996), an action thought to be exerted at the pituitary level, perhaps from mammosomatotrophs (Carmignac et al, 1998). Given the above data, an endogenous GHS system might well play a role in pulsatile GH control and a revision of our understanding of the control of GH pulsatility might be necessary. The clinical use of GHSs is currently hampered by our inadequate understanding of the pharmacodynamic profile required for optimal efficacy and our lack of knowledge concerning the exact molecular and physiological mechanism of action of GHSs.

1.4.3 GHS sites of action Specific binding sites for GHSs have been detected in the pituitary and hypothalamus in rats and pigs (Codd et al, 1989; Pong et a l, 1996; Veeraragavan et a l, 1992), (Sethumadhavan eta l, 1991), suggesting a central hypothalamic and a direct pituitary action for GHSs.

1.4.3.1 The direct pituitary pathway As previously described, GHSs were shown to release GH in vitro from dispersed rat pituitaries in a dose dependent manner (Bowers et al, 1977). Radiolabelled MK-0677 was synthesised (Dean et al, 1996) and shown to bind with high affinity and limited capacity to porcine and rat pituitary membranes (Pomes et al, 1996; Pong et al, 1996). Using dual fluorophor labelling for GH and [^^S]MK-0677 binding sites demonstrated that p^S]MK-0677 binding was confined to GH containing cells (Hreniuk et al, 1996), in accordance with GHRP-6 and its mimetics selectively activating somatotrophs and mammosomatotrophs (Smith etal, 1996). Initially, it was thought, that GHSs might act on GHRH-R at the pituitary level. However several lines of investigation have since proven that GHSs and GHRH act on different receptors with different signalling pathways. Binding studies demonstrated that [^^S]MK-0677 binding was competitively displaced by GHRP-6 but not GHRH, suggesting that GHRPs and non-peptidylic GHSs used the same receptor and that their receptor was different from that for GHRH (Pong et al, 1996). Studies using a perfusion system consisting of rat pituitary cells and in vivo experiments in rats and pigs showed

24 Chapter 1 : Introduction that the GHRH-R could be desensitised by prolonged exposure to GHRH, but administration of GHSs still caused an increase in GH secretion (Blake and Smith, 1991; Chang et al., 1995; Cheng et al., 1989; Clark et al., 1989). Similarly, when desensitisation against GHSs occurred, GHRH could still release GH. It was shown that the GHSs activate a G protein coupled receptor, since [^^S]MK-0677 binding was Mg^^ dependent and inhibited by GTP analogues (Cheng et al, 1991; Pong et al, 1996; Smith et al, 1996). However, in contrast to GHRH (which increases cAMP levels in somatotrophs), GHSs had no effect on intracellular cAMP (Cheng e ta l, 1989; Wu gr al, 1994), but when combined with GHRH, the effects of GHRH on cAMP production were potentiated (Cheng et al, 1989). In further experiments, Phorbol myristate acetate (PMA) mimicked these effects and protein kinase C antagonists abolished GHS signalling, suggesting that the effects of GHRP-6 were at least partially dependent on protein kinase C (Cheng et al, 1991). Phospholipase C was shown to reproduce these effects and GHSs increasing inositol trisphosphate (IP3) turnover (Adams et a l, 1995), strengthened the involvement of phospholipase C in the GHS signalling pathway. GHSs also caused depolarisation by blocking channels (McGurk et a l, 1993), leading to a transient increase in intracellular Ca^^ concentrations by opening voltage operated Ca^^ channels (Herrington and Hille, 1994). There is considerable controversy regarding the possibility that GHSs may also cause a rise in intracellular Ca^"^ by redistribution from internal stores, probably mediated by the generation of IP3 (Smith et al, 1999). The suggested signal transduction pathway of GHSs is depicted in Figure 1.7. Both GHRH and GHS signalling lead to the increased intracellular Ca^^, thereby suggesting that GH secretion is directly related to intracellular Ca^^ concentrations. The depolarisation of somatotroph membranes explains in part their functional antagonism to SS, since SS causes hyperpolarisation by increasing conductance and blocking Ca^^ channels (Koch et al, 1988). It is speculated that GHSs and GHRH synergise through cross-talk in their signal transduction pathways to increase cAMP and GH release (Smith et al, 1996). This might be mediated by interactions between the Gp^- subunit associated with the GHS-R

25 Chapter 1 : Introduction

&

Ca2+

Figure 1.7 GHS signal transduction The GHS signalling pathway involves activation of phospholipase C (PLC) by the Ga-subunit to produce DAG and IP3. IP3 might cause the release of Ca^^ from intracellular stores, however this is not clear. Protein kinase C (PKC) is then activated presumably by DAG. It is speculated that depolarisation of the membrane by inhibition of channels and activation of L-type Ca^* channels is associated with PKC mediated phosphorylation. Subsequently GH is secreted. and the Gg, subunit of the GHRH-R (Tang and Gilman, 1991). Since GHSs do not directly activate the cAMP pathway, they have no effect on GH mRNA expression levels in normal rats (Torsello et ai, 1996) or Pit-1 in vitro (Soto et ai, 1995). Interestingly Mitani etal. have shown by reverse haemolytic plaque assay that subsets of somatotrophs responded to either only GHRH, only GHRP-6 or both (Mitani et aL, 1996). A possible explanation arises from data showing that [^^S]MK-0677 binding was restricted to GH cells, but only half of the identified GH

2 6 Chapter 1 : Introduction cells showed p^S]MK-0677 binding (Hreniuk et al, 1996). The physiological significance of the functional heterogeneity of somatotrophs remains unknown. Surprisingly, the lit/lit mouse, an animal model with a point mutation in the GHRH-R, and humans with GHRH-R mutations did not show GH responses to GHRP-6 (Jansson et al, 1986b; Maheshwari et al, 1999), indicating that GHSs could not exert their GH releasing capacities without a fully functioning pituitary GHRH-R. Humans with the GHRH-R mutation retained the response of elevated plasma Prl, ACTH and cortisol levels to GHRP-6, suggesting that at least this effect is independent of GHRH-R signalling. In vitro GHSs retained their GH releasing capacity in states of complete GHRH desensitisation (Blake and Smith, 1991). The precise nature of interactions between GHSs and the GHRH-R signalling pathway are still unclear. However, further evidence of the dependence of GHS action on an intact GHRH system came from experiments in pituitary stalk lesioned pigs and humans, which showed attenuation of GHS induced GH release (Hayashi et al, 1993; Hickeyet al, 1996; Loche et al., 1995; Popovic et a l, 1995). The efficacy of GHSs is also greatly attenuated following administration of anti-GHRH serum (Bercu et at., 1992; Clark et al, 1989; Conley et al, 1995). These data not only suggested an involvement of GHRH in GHS action, but also gave the first indications of a central hypothalamic action for GHS.

1.4.3.2 The central hypothalamic GHS pathway The synergistic effect of GHRH and GHRP-6 in vivo, but additive effect in vitro (Akman et al, 1993; Blake and Smith, 1991; Bowers et al, 1984) and the fact that conscious animals showed more potent responses to GHRP-6 than anaesthetised ones (Clark et al, 1989), suggested a hypothalamic involvement in the GHS pathway. In conscious sheep and guinea pigs GHSs were much more potent when given i.c.v.; small doses that did not release GH when injected i.v. were able to release GH when given i.c.v. (Fairhall et al, 1996). The authors concluded that GHSs released GH by a specific and sensitive central action. In sheep, GHSs were also shown to directly release GHRH into the hypophyseal blood system (Fletcher et al, 1996; Guillaume et al, 1994). However there seems to be some species difference, since studies with a non-peptidylic GHS in pigs showed no effect on portal blood GHRH concentrations, but reduced SS levels (Drisko et al, 1999). In vitro in a hypothalamus explant system GHS were unable

27 Chapter 1 : Introduction to release GHRH and even blocked GHRH release at high concentrations (Korbonits et al, 1999a), suggesting desensitisation of the GHS-GHRH system at high levels of GHS.

Finally, direct evidence for hypothalamic GHS actions came from several electrophysiological and c-fos studies. In electrophysiological in vivo studies putative ARC GHRH neurones were identified by following multiple criteria e.g. projection to the ME, excited by stimulation of the basolateral amygdala, inhibited by stimulation of the PEN and rebound hyperactivation following the end of PEN stimulation (Dickson et a l, 1994). Systemic and i.c.v. injections of GHRP-6 and non-peptidylic GHSs were shown to cause a large increase in firing rate 3-5 minutes after the injection in these ARC neurones (Dickson et al, 1995a; Dickson et al, 1993b). In addition, Dickson et al found that a subpopulation of putative, non-secretory ARC neurones was inhibited by GHRP-6. In in vitro slice preparations, GHRP-6 was able to enhance the electrical activity of ARC and VMN neurones, but also inhibited a smaller population of these neurones (Hewson et al, 1999). GHRP-6 also inhibited some PEN neurones and the authors suggested that the ARC and PEN inhibited neurones might be SS positive, thereby implying that GHSs might reduce somatostatinergic tone to enhance GH release. In further experiments the immediate-early gene c-fos was used as a marker for neuronal activation and Fluorogold for retrograde tracing of neurosecretory cells, c-fos expression in response to systemic GHRP-6 expression was highly restricted to the ARC (Dickson et al, 1993b). 68-82% of retrogradely labelled cells co-localised with c-fos expressing neurones in the ARC, suggesting that this proportion of neurones projected outside the blood brain barrier (Dickson et al, 1996). When Dickson et al compared systemic injections of GHSs to i.c.v. injections, it became clear that i.c.v. GHS induced ARC c- fos expression at doses that were ineffective i.v. (Dickson et al, 1995a). The c-fos activation was shown to be independent of GH, since severely GH deficient litAit mice and dw/dw rats still responded with central ARC c-fos activation (Dickson et al, 1995b; Sirinathsinghji et al, 1995). Co-localisation studies of ARC c-fos expression in response to GHS and neuropeptide mRNA revealed, that 51% of all c-fos expressing neurones express NPY, while only 23% express GHRH, 11% POMC and 4% somatostatin (Dickson and Luckman, 1997; Kamegai et al, 1996). Activation of NPY neurones by GHSs might explain the increased feeding behaviour following GHRP-6 injection

28 Chapter 1 : Introduction

(Locke et al, 1995; Okada et al, 1996). Bennett et al also reported that a 6 day GHRP- 6 i.c.v. treatment in rats up-regulated ARC NPY mRNA expression (Bennett et al, 1998). In GH deficient dw/dw rats 14 days continuous i.c.v. GHRP-6 infusion increased food intake and body weight gain (personal communication D.F. Carmignac, P.Bennett). However, Wells et a l infused GH deficient Tgr rats for 7 days i.v. with pulsatile GHRP- 6 and found that this significantly increased their body weight gain and epiphysial plate width, but reduced NPY mRNA levels (Wells et al, 2000). The effect GHSs have on growth, feeding behavior and NPY expression might thus be administration pattern dependent and will need further investigation. Furthermore, recently Tschop et al showed that presence of NPY was not necessary for the GHS Ghrelin to exert its effects on energy metabolism (Tschop et al, 2000). Chronic central infusion of GHS for 5 days caused absence of ARC c-fos expression and and lack of c-fos and GH responses to acute GHS treatment (Bailey et al, 1999a). The authors showed that ARC NPY, GHRH and SS levels were unchanged, while SS levels in the PEN were significantly enhanced and suggested that chronic central GHS treatment caused enhanced GH feedback onto PEN SS neurones. Hypothalamic GHS action underlies SS control, since a somatostatin analogue inhibited GHS induced GH release when administered icv (Fairhall eta l, 1995). Indeed, i.v. and i.c.v. SS was able to attenuate the GHS induced c-fos response in ARC neurones (Dickson et al, 1997). Furthermore arcuate neurones that were electrically stimulated by GHS treatment were inhibited following i.v. somatostatin infusions (Leng et a l, 1998). Studies with sstr2 knock-out mice showed that GH pre-treatment prevented the activation of c-fos by GHSs in ARC neurones in WT but not sstr2 KG mice (Zheng et al, 1997). These data suggested that GH induced SS release from PEN neurones, which then suppressed the GHS responsiveness of ARC neurones. The hypothesis to date is therefore that the central action of GHSs is inhibited by SS through direct inhibitory projections onto ARC neurones. Figure 1.8 shows a schematic summary of the GH regulating pathways described in this chapter.

29 Chapter 1 : Introduction

NPY GHRH

©//©

GHS (ghrelin?)

' IGF-I

Figure 1.8 Regulation of GH secretion GHRH stimulates the production of GH, while SS inhibits GH release. At the hypothalamic level SS seems to regulate GHRH via SS receptors, while it is not clear, whether GHRH can regulate SS. GH stimulates IGF-I production, which via a feedback loop inhibits GH release either at the pituitary level or by regulating hypothalamic GHRH and SS. GH itself regulates its own secretion by control of hypothalamic GHRH (also via NPY) and SS. The exogenous GHSs increase GH release from the pituitary by a direct pituitary action and indirectly via stimulation of hypothalamic GHRH. GHSs also act on NPY. SS can inhibit hypothalamic GHS actions. Furthermore it seems, that GHSs might be involved in the GH feedback mechanism, maybe also by inhibition through IGF-I.

1.4.4 The GHS Receptor (GHS-R) Using a high specific-activity p^S]MK-0677 binding assay, a low-abundance high- affinity G protein-linked receptor of the anterior pituitary gland (2fmol/mg protein) and hypothalamus (8fmol/mg protein) was identified (Dean etal., 1996; Pong etal., 1996). Xenopus oocytes co-injected with [poly(A)H-J RNA from swine pituitary and cRNA for the G protein G„,i exhibited oscillatory increases in calcium-activated chloride currents

30 Chapter 1 : Introduction induced by MK-0677 (Howard et al, 1996). This expression system was used to screen pools of cRNA from swine and human pituitary gland cDNA for an MK-0677 inducible system. Finally the full-length pig and human GHS-R cDNA were cloned (Howard et al., 1996). The full length GHS-R is a 366 amino acid protein with seven putative transmembrane domains and a G-protein coupled receptor triplet signature sequence (Howard et al, 1996) (Figure 1.9). Sequence alignment showed that the pig and human GHS-R sequences were 93% identical and 98% similar at the amino acid level. Similarly the rat GHS-R was cloned (McKee et al, 1997a) and the sequence was found to be -95% similar to human and pig GHS-R. From all pituitaries two types of cDNA were isolated: type la encoding the full length 7 transmembrane (TM) functional protein, and type lb, encoding a protein containing TM- 1 through to TM-5 with no measurable function in cell based assays (Howard et al, 1996; McKee et al, 1997a) (Figure 1.9). These two types of cDNA were shown to arise from a single gene by alternative mRNA splicing. Type la encodes the complete 7-TM GHS-R and results from a splicing event that removes an intron at Leu^^. In type lb an alternative polyadenylation signal in the intron is used and a short 24 / SOaa (human & pig / rat) open reading frame (ORF) is fused to leucine 263 (McKee et al, 1997a). The site of the intron between TM 1-5 and TM 6-7 is highly conserved between rat, human and swine, while the intronic sequence is not (McKee et al, 1997a). Transfection of the full length cDNAs in either Xenopus oocytes (Howard et al, 1996) or HEK-293 and COS-7 (McKee et a l, 1997a) cells demonstrated that the pharmacological profile of GHS binding to the cloned GHS-R correlated with their biological activity. Due to these properties the cloned receptor was called the GHS-R.

In RNAse protection analysis experiments the GHS-R expression was limited to the pituitary, hypothalamus and hippocampus with no GHS-R expression in any other peripheral tissues (Howard et al, 1996; McKee et al, 1997a). In situ hybridisation using a type la GHS-R cDNA probe on sections of rhesus monkey and rat brains and pituitaries further localised the expression of the GHS-R to the anterior lobe of the pituitary and several regions of the brain (Guan et al, 1997). The strongest GHS-R expression was seen in hypothalamic areas including the anterior

31 Chapter I; Introduction

tOOOO) Extracellular ÎÎÎÎÎ UUA Intracellular

366aa

COOH Type 1b 289aa

Figure 1.9 Predicted GHS-R transmembrane topology Proposed two-dimensional structure of the type 1a and 1b GHS-R. The GPC-R signature sequence Glu-Arg-Tyr is depicted in blue. While the full-length type 1a GHS- R contains all 366 aa and 7 TM, the type 1b GHS-R only contains 5 TM regions and diverges beyond Leu^®^ hypothalamic area, SCN, LHN, SON, VMN, ARC, and PVN. Specific signals were also detected in the hippocampus, several nuclei of the brain stem, ventral tegmental area, median and dorsal raphe nuclei, Edinger-Westphal nucleus, laterodorsal tegmental nucleus and facial nerve (Guan et ai, 1997). The functional significance of the GHS-R signal in these regions is unclear. In another in situ hybridisation study of GHS-R expression in rat brain and pituitary, no GHS-R expression was detectable in the pituitary, but GHS-R expression in the brain was confirmed in the hippocampus and the hypothalamic ARC, VMN and PVN (Bennett et al, 1997). Double label in situ hybridisation revealed that -25% of ARC and VMN GHRH neurones co-expressed GHS-R mRNA, however these dually labelled neurones only represented a small subpopulation of GHS-R expressing ARC and VMN neurones (Tannenbaum et al, 1998). Virtually all ventromedial arcuate NPY neurones were found to also express GHS-R mRNA; the proportion of GHS-R cells that also express NPY

32 Chapter 1 : Introduction remains unclear, although a subset of GHS-R neurones in the lateral ARC was shown not to contain any NPY mRNA (Willesen et al, 1999). Only a few cells in the VMN were shown to co-express GHS-R and NPY mRNA (Willesen et al, 1999). A minor degree of SS and GHS-R mRNA overlap was found in the dorsomedial area of the ARC, where -30% of the SS cells also expressed the GHS-R (Willesen et al, 1999). This tissue distribution of the GHS-R together with the distribution of c-fos positive cells in response to GHSs is in close agreement with their physiological effects. However, the absence of a VMN c-fos signal in response to GHS treatment is somewhat surprising, since a large population of GHS-R is expressed there (Guan et al, 1997) and electrical activity in response to GHRP-6 was measured in VMN neurones (Hewson et al, 1999). The presence of GHS-R on GHRH and NPY neurones supports the notion that these neurones are direct targets for GHS action. Whether activation of these neurones by GHSs results in the same neuroendocrine effect or whether activation of NPY neurones mainly regulates food intake, while GHRH neurones mediate the GH releasing effect of GHSs, remains to be proven. Bennett et al showed by quantitative in situ hybridisation that dw/dw rats with chronic severe GH deficiency displayed an increase in hypothalamic ARC and VMN GHS-R expression, and that treatment with bovine GH reduced GHS-R abundance to even below normal levels (Bennett et al, 1997). These changes paralleled the GHRH expression and were opposite to the NPY expression in dw/dw rats with or without GH treatment and might imply that GHS-R and the GHS may be part of the GH feed-back system (Bennett et al, 1997). Similarly, elevated pituitary GHS-R levels in the spontaneous dwarf rat (SDR) could be reduced by GH treatment (Horikawa et al, 2000; Kamegai et al, 1998), while hypothalamic GHS-R levels seemed unaffected by the GH status in the SDR as measured by RT-PCR (Kamegai et al, 1998). Recently Nass et al reported that in rats with high plasma GH levels due to GH secreting tumours, GHS-R expression levels in the pituitary were down regulated (Nass et al, 2000). These data highlight the possible importance of the GHS-R in GH feed-back mechanisms. Underlining this notion are data by Bailey et al demonstrating that GH feed-back was able to alter the responsiveness of neurones involved in the central response to GHS (Bailey etal, 1999c). Hypothalamic GHS-R expression was similar in male and female rats in the ARC

33 Chapter 1 : Introduction nucleus and the hippocampus, but adult female rats displayed significantly higher GHS- R mRNA levels in the VMN (Bennett et al, 1997). However, hypothalamic GHS-R mRNA was shown to be enhanced in estrogen treated rats (Carmignac etal., 1998), maybe explaining GHRP-6’s more pronounced effect in female rats (Popovic et at., 1995; Sartor et at., 1985). Furthermore the absence of glucocorticoids in adrenalectomised rats reduced hypothalamic GHS-R mRNA levels and these returned to normal with dexamethasone (Dex) treatment (Thomas et al, 2000). Similar effects of adrenalectomy and Dex treatment were found at the level of the pituitary GHS-R expression (Tamura et ai, 2000). Support for the possible regulation of GHS-R expression by estrogen and glucocorticoids came from analysis of the 5’ region of the GHS-R, which revealed a putative estrogen-responsive element (Kaji et ai, 1998) and binding sites for inducible transcription factors for the estrogen and glucocorticoid receptor (Petersenn et ai, 1999). While Bennett et al. reported that continuous s.c. infusion of GHRP-6 for 2 weeks in normal female rats did not change ARC or VMN GHS-R expression levels (Bennett et ai, 1997), Kineman et al. found that 4h i.v. infusion of a non-peptidyl GHS decreased pituitary GHS-R levels to 50% (Kineman et al., 1999). Furthermore in Kineman’s RT- PCR analysis 4h of i.v. GHRH treatment up-regulated pituitary GHS-R mRNA levels in normal rats and SDRs, but not in vitro in cultured pituitaries (Kineman et al., 1999). However in a study by the same group in which hypothalamic GHRH neurones were destroyed by monosodium glutamate (MSG) treatment, the resulting GHRH deficiency caused an up-regulation of pituitary GHS-R and GHRH treatment had no effect on pituitary GHS-R levels (Kovacs et ai, 2000). The effect of GHRH on pituitary GHS-R expression is thus somewhat unclear. GHS-R expression in pituitary adenoma has been of considerable interest and was found not to be restricted to GH-secreting adenoma (Adams et ai, 1998; Barlier et al., 1999; de Keyzer et al., 1997; Korbonits et al., 1998; Skinner et al., 1998). GHS-R expression was present in some prolactin-, TSH- and ACTH-secreting, gonadotroph, pancreatic, endocrine bronchial and non-functional adenoma. Somatotropinomas seemed to express higher levels of GHS-R (Korbonits et al., 1998; Skinner etal., 1998), which might add an explanation to why some acromegalic patients exhibit major GH discharges to GHRP (Alster et al., 1993; Ciccarelli et al., 1996; Hanew et al., 1994; Popovic et al., 1994).

34 Chapter 1 : Introduction

Furthermore some somatotropinomas and prolactinomas were shown to respond to GHSs with enhanced PI turnover (Adams et al, 1998; Adams et al, 1996).

In addition to the endogenous GHS-R ligand Ghrelin, which will be discussed in detail in Chapter 8, Smith et al recently reported adenosine as a partial ligand for the GHS-R (Smith et al, 2000). Adenosine was shown to activate the GHS-R in vitro in GHS-R transfected HEK293/aeql7 cells, but adenosine was found to be less potent (EC50 =2(xM) than GHRP-6 and MK-0677. Furthermore, the authors suggested that adenosine binds to a different binding pocket than MK-0677 and GHRP-6 and showed that adenosine did not stimulate GH release, neither alone, nor did it amplify the effects of GHRH in vitro. The putative role of adenosine in GHS physiology remains to be established.

Recently three homologues to the human GHS-R were cloned from Pufferfish, of which one was activated specifically by several GHSs (Palyha et al, 1997). The authors suggested that the functional structure of the GHS-R ligand binding domain has been conserved for 400 million years. Apart from the GHS-R four more family members have been cloned (McKee et al, 1997b; Tan et al, 1998). The nucleotide sequence of the GHS-R is most closely related to the neurotensin receptor-1 (NT-Rl) (35% overall protein identity) (Howard et al, 1996). Two human GPC-Rs related to both the type la GHS-R and NT-Rs were cloned and characterized (McKee et al, 1997b). GPR38 and GPR39 share significant amino acid sequence identity with the GHS-R and NT-Rs 1 and 2. An acidic residue (El24) in TM-3, essential for the binding and activation of the GHS-R by structurally dissimilar GHSs (Feighner et al, 1998), was conserved in GPR38 and GPR39. GPR38 is encoded by a single gene expressed in thyroid gland, stomach, and bone marrow. Alternative splicing of the GPR38 mRNA, in a similar mechanism to the GHS-R, results in two cDNAs encoding either full-length GPR38 or a truncated 5 TM GPR38 (Feighner et al, 1999). GPR39 is encoded by a highly conserved single-copy gene, expressed in brain and other peripheral tissues. GPR 38 was recently identified as the human motilin- receptor (MTL-R) (Feighner et al, 1999). It was demonstrated that motilin and erythromycin activated the MTL-R, which was also expressed in specific interstitial cells in the human duodenum, jejunum, colon, ileum and esophagus (Feighner et al.

35 Chapter 1 : Introduction

1999). These data might explain the effect of erythromycin on GI motility. Interestingly an endogenous ligand for the GHS-R Ghrelin was found to be mainly expressed in the GI tract as well (Kojima et al, 1999), a correlation that will be discussed further in Chapter 8. The third family member of the GHS-R is FM3 in human and mouse (Tan et al, 1998). It was cloned from a mouse T-cell library due to its sequence homology to the GHS-R and was also expressed in pancreas, testis, small intestine adrenal cortex, liver and stomach (Howard et al, 2000). Rat and human FM-4 were cloned by homology to FM-3 (51% human FM-3 and -4) and predicted as a 7 TM G protein coupled receptor (Howard et al, 2000). FM-4 was found to be expressed in the rat brain in the PVN, along the wall of the third ventricle and in area CAl of the hippocampus. The endogenous ligand to both these receptors was found to be Neuromedin U (NMU) and the receptors were called NMUIR and NMU2R (Howard et a l, 2000). NMU mRNA expression itself was most abundant in the hypothalamic ARC and expression levels were reduced by fasting. I.c.v. administration of NMU markedly suppressed food intake in rats. NMU as a ligand for FM-3 has recently been confirmed by three other groups (Fujii et al, 2000; Kojima et al, 2000b; Szekeres et al, 2000).

1.5 Tissue-Specifîcity in Transgenesis

Transgenic technology has enabled the study of gene regulation in mammals in vivo (for review see (Grosveld and Kollias, 1992)). The primary aim of transgenic studies is to direct expression of particular genes to specific groups of cells within heterogeneous cellular systems. Whether constructs used for transgenesis are expressed and in which cell-type the expression occurs is dependent on several factors.

1.5.1 Regulation of gene transcription Transcriptional elements such as the core promoter, containing elements including the TATA box, TFIIB recognition element, initiator and the downstream promoter element, are sufficient to direct basal transcription initiation. Immediately up-stream of the core promoter are multiple recognition sites for a subgroup of sequence-specific DNA-

36 Chapter 1 : Introduction binding transcription factors. The proximal promoters and consensus sequences are well described (for review, see Komberg, 1996; Nikolov et al, 1996; Roeder, 1996), but inclusion of these in transgenic targeting constructs usually only confers low basal transcription. Transcriptional control regions, that regulate transcription above basal levels, often contain multiple, autonomous enhancer modules, each of which performs a specific function, such as activation of its gene in a specific cell type or at a particular stage in development (Blackwood and Kadonaga, 1998). Thus inappropriate transgene expression in ectopic sites can often be due to the absence of specific silencers or enhancers of gene expression, which are essential for the cell-specific expression of genes in the targeting construct. But even the inclusion of enhancers and silencers is sometimes not sufficient to ensure correct physiological levels or tissue-specific expression (Palmiter and Brinster, 1986). This variation has been termed position effect and is thought to result from the positive or negative effects of chromatin surrounding the transgene. Mouse lines with multi-copy transgenes exhibiting expression levels that bear no relationship to the number of integrated transgene copies are an example of such integration position effects.

1.5.2 Locus control regions The identification of locus control regions (LCR) that were able to overcome position effects was thus a major discovery. LCRs are powerful genetic regulatory elements that are characterised phenotypically by three major properties. They confer tissue-specific and physiological levels of expression on linked genes, activate the transcription of transgenes in a position- independent, copy-number dependent manner, and determine DNA replication timing and the origin utilised (Grosveld et al, 1987). Thus tissue-specific control of basal transcription might reside in the promoter, whereas tissue-specific enhancement of gene expression is a property of the LCR. The majority of the genome is packaged into a largely inaccessible chromatin conformation. This higher-order chromatin structure is locally disrupted in the vicinity of genes that are actively transcribed or ready for transcription (for review see Gross and Garrard, 1988; Wallrath et at., 1994). Copy number-dependent expression is widely

37 Chapter 1: Introduction considered to be indicative of an open chromatin structure i.e. DNA that is accessible to transcription factors. It became clear that position-independent, copy-number-dependent transgene expression was due to certain sites in an LCR called DNAse hypersensitive sites (HS). These elements may establish the functional boundaries for segments of activated chromatin, or increase the efficiency of classical enhancer elements in a chromatin environment. Thus the absence of any DNAse HS can render the site of integration of a transgene inaccessible, such that a transgene integrated into heterochromatin can be then subject to silencing (Kioussis and Festenstein, 1997). Due to the possible large distances between DNAse HS, transgenic studies have moved away from so-called ‘mini-promoters’ towards larger constructs.

1.5.3 Generating targeting constructs When designing a construct for introduction into a transgenic organism the elements of transcription control described above need to be considered to achieve appropriate transgene expression. Our current understanding of transcriptional control mechanisms suggests the presence of regulatory elements many kb away from the transcription initiation site. The use of larger DNA fragments improves the chances of obtaining an integration site-independent expression pattern, and a copy number-dependent expression level. Still, although a large amount of flanking sequences will reduce the risk of omitting important regulatory elements, too much DNA may introduce unknown and unwanted genes. The search and analysis of a full gene LCR is laborious and time consuming. Hence if this is to be avoided, a quick, and probably most appropriate, answer to the question ‘How big a piece of DNA must be used to give specific transgene expression?’ is ‘As big as possible’. Therefore, new targeting vectors that carry large genomic pieces of DNA were needed. Genomic phage or cosmid clones can be used for transgenes up to 50kb. For larger genes, bacterial artificial chromosome (BAG) clones (100-150kb) or PI genomic clones (75-lOOkb) are essential. Only recently has the manipulation of e.g. BAG clones become easier by using homologous recombination in E.coli (Zhang et al, 1998). In this thesis, genomic cosmid clones of approximately 40kb were used to maximise the chances of appropriate tissue-specific expression, but retain the ease in handling cosmid

38 Chapter 1 : Introduction clones. As described in detail in Chapter 4, the cosmid targeting transgenes to pituitary GH cells had already been proven previously to contain the complete hOH LCR (Bennani-Baiti et al, 1998; Jones et al, 1995; Su et al, 2000). For the cosmid targeting hypothalamic GHRH neurones no such data were available, except that the cosmid had previously appropriately targeted transgene expression to hypothalamic ARC neurones (Flavell et al, 1996).

39 Chapter 2: Materials and Methods

2 Materials and Methods Kits were used according to manufacturer’s instructions, unless otherwise stated. 2.1 Bacterial cultures Bacterial cultures were grown in standard LBroth (Sambrook et al, 1989) with shaking at 37°C. E. coli strains were plated on LBroth agar plates (LBroth with 15 g/1 bacto- agar). These media were supplemented with the appropriate antibiotic (Sigma) at the following working concentrations: ampicillin at lOOpg/ml, tetracyclin at 50pg/ml and kanamycin at 50pg/ml. Bacterial stocks were kept at -70°C in 15% glycerol. 2.2 Purification of nucleic acids Aqueous solutions of DNA were purified by adding an equal volume of phenol:chloroform:isoamyl alcohol (25:24:1) (Life Technologies), vortexing and centrifuging at 12000 rpm for 5 minutes in a benchtop centrifuge at room temperature. The aqueous supernatant was precipitated by adding 3M sodium acetate (pH5.2) to a final concentration of 300mM and two volumes of absolute ethanol. Samples were frozen for an hour at -20°C, centrifuged and the pellet resuspended in water. 2.3 DNA preparation 2.3.1 Preparation of plasmid and cosmid DNA Small scale DNA preparations (3 ml saturated bacterial culture) were prepared using the Qiaprep Spin Miniprep Kit (Qiagen) and DNA from large volumes of saturated bacterial cultures (300 ml) was extracted with the Plasmid Maxi Kit (Qiagen). 2.3.2 Preparation of genomic DNA from animal tissue Rat and mouse tail biopsies were taken from 18-21 day old animals and placed in ‘tail mix’ (50mM Tris-HCl pH8.0, lOOmM BDTA, lOOmM NaCl). Genomic DNA was extracted using a standard procedure (Hogan et a l, 1986). Tissues, including tail biopsies and post mortem liver samples and ears, were incubated with proteinase K (Sigma, 20-40|ig/sample) at 55°C for 4 hours, following RNase A (Sigma, 10- 20|xg/sample) incubation at 37°C for 30 minutes. Digested samples were then phenol- chloroform extracted, the DNA precipitated with 0.6 volumes of isopropanol and pellets resuspended in water.

40 Chapter 2: Materials and Methods

2.4 Subcloning of DNA fragments 2.4.1 Restriction digestion Plasmid DNA was incubated with restriction enzymes following the manufacturer’s instructions (Roche Diagnostics or New England Biolabs) for up to 2 hrs; genomic DNA was digested overnight. 2.4.2 Blunting of DNA fragments by refilling the5 ’ overhang After restriction digestion, DNA was purified by phenol extraction and ethanol precipitation and resuspended in water. After digestion with a restriction enzyme which leaves a 5’ overhang, the single stranded ends were filled in by using the Klenow fragment of E.coli DNA polymerase I (New England Biolabs). The Klenow reaction was performed in Ix Klenow buffer (New England Biolabs) supplemented with dNTPs (Amersham Pharmacia Biotech) to a final concentration of 0.25mM and 10 units of Klenow fragment at 37°C for 30 minutes and DNA was then repurified. 2.4.3 Vector dephosphorylation After digestion with a restriction enzyme, the 5’phosphate groups of DNA fragments were removed to prevent self-ligation. DNA was incubated with 5 units of calf alkaline phosphatase (Roche Diagnostics) in Ix alkaline phosphatase buffer (Roche Diagnostics) at 37°C for one hour and repurified. 2.4.4 Insertion of linkers Plasmid DNA was digested, blunted and dephosphorylated to prevent self-ligation. A phosphorylated linker was ligated to the plasmid in Ix ligase buffer (New England Biolabs) with 400 units of T4 DNA ligase (New England Biolabs) at room temperature overnight. The reaction was inactivated at 65°C for 15 minutes, the plasmid linearised with the restriction enzyme appropriate to the linker and repurified. 2.4.5 Gel electrophoresis of DNA fragments DNA fragments were separated by size in gels of varying percentages of agarose in IxTAE buffer (0.04M Tris-acetate, ImM EDTA) supplemented with 0.5ng/ml ethidium bromide. DNA bands were photographed on an ultraviolet transluminator and fragment sizes compared to markers (1Kb DNA ladder. Life Technologies).

41 Chapter 2: Materials and Methods

2.4.6 Purification of DNA fragments from agarose gels DNA fragments were excised from agarose gels and processed further using the QIAEX II Gel Extraction Kit (Qiagen). 2.4.7 Ligation of DNA fragments DNA fragments were mixed at varying molar ratios in Ix ligase buffer (New England Biolabs) and 400 units T4 DNA ligase (New England Biolabs) were added. The reaction was incubated overnight at 16°C and inactivated at 65°C for 15 minutes prior to repurification. 2.4.8 Preparation and transformation of competent cells A litre of fresh YENB medium (0.75% Bacto Yeast Extract, 0.8% Bacto Nutrient Broth, Difco Laboratories) was inoculated with a fresh overnight bacterial culture DHIOB and grown at 37°C with shaking to an OD^oo of 0.5 to 0.9. To harvest the cells, the flask was chilled on ice for 5 minutes and spun at 4000xg for 10 minutes at 4°C. The pellet was washed twice in 100 ml of cold water and centrifuged as before. The cells were finally resuspended in 3 ml of cold 10% glycerol and either used fresh or stored in aliquots at - 70°C until further use. For electrotransformation l-5pl of ligated DNA were added to a 40pl aliquot of cells, mixed and kept on ice for 1 minute. The mix was transferred to a cold 0.2cm electroporation cuvette and electroporated in a Bio-Rad Gene Puiser according to the manufacturer’s instructions. 1ml of YENB medium was added immediately and the reaction incubated at 37°C for 1 hour before plating onto selective medium plates. Commercially available competent cells (Epicuran Coli XL-1-Blue, Stratagene) were transformed by the heatshock method according to manufacturer’s instructions. 2.4.9 Packaging of cosmid DNA into bacteriophage Packaging of cosmid constructs into bacteriophage particles was performed using Gigapack III XL packaging extracts (Stratagene). E.coli strain VCS257 was infected following the manufacturer’s instructions. 2.5 DNA sequencing Double stranded DNA was sequenced on both strands on an ABI 377 automated sequencer using the ABI Prism dye termination cycle sequencing ready reaction kit (Perkin Elmer).

42 Chapter 2: Materials and Methods

2.6 Polymerase Chain Reaction Amplification of DNA fragments was performed by the polymerase chain reaction (PCR). Varying amounts of DNA template were supplemented with IxPCR reaction buffer (containing 1.5mM MgClj, Perkin Elmer), dNTPs (0,25mM, Amersham Pharmacia Biotech), two specific primers (50ng each, Oswell) and 1.25units of TaqPolymerase (Perkin Elmer). The reactions were incubated in a Hybaid PCR Express machine, using the following typical programme:

Initial dénaturation: 94°C for 5 minutes Denaturing: 94°C for 1 minute “ Annealing: 60°C for 30 seconds 30 cycles Extension: 72°C for 1 minute 30 seconds Final Extension: 72°C for 5 minutes

The annealing and extension temperatures and times were changed according to the primers and templates used. PCR products were analysed on agarose gels.

Primers for cloning transgene constructs Primer 1 F CGGAATTCGACGCGTGATCCCAAGG (EcoRUMlul sites+hGH 5’UTR) CCCAACTCC Primer 2 R TGGGCGTCGCGTTCCACATTGCAGCT (chimeric hGH 5’UTR / hGHS-R ATG) AGGTGAGCTGT Primer 3 F GACAGCTCACCTAGCTGCAATGTGG (chimeric hGH 5’UTR / hGHS-R ATG) AACGCGACGCCCA Primer 4 R CCTTGGTGACCACCACCTTG (hGHS-R BstEII restriction site) Primer 5 F GAATCTAGTATTAATACATGACTGCC (chimeric hGHS-R STOP / hGH 3’UTR) CGGGTGGCATCCCT Primer 6 R CCCTCGAGCGCCGCACGCGTATTAT (hGH 3’UTR + MluUNotUXhol sites) AGAAGGACACCTAG Primer 7 F ATTCTGTACAACATCATGT (hGHS-R BsrGl restriction site) Primer 8 R AGGGATGCCACCCGGGCAGTCATGT (chimeric hGHS-R STOP / hGH 3’UTR) ATTAATACTAGATTC Primer 9 F TCGCGGATCCCAAGGCCCAACTCC (hGH 5’UTR BamHl restriction site) Primer 10 R CCATAATATTATAGAAGGACACCTA (hGH 3’UTR Sspl restriction site) GTCA

43 Chapter 2: Materials and Methods

Primers for genotyping transgenic rGHRH-hGHS-R and hGH-hGHS-R transgenic mice hGHS-R 4 R (cDNA position bpl044) CCGAGAACTTTCATCTTTCAGAGTGG AGAG hGH 5’UTR AACCACTCAGGGTCCTGTGGACAG hGH exon 2 R CCTCTTGAAGCCAGGGCAGGCAGAG CAGGC Primers for RT-PCR eGFP R (cDNA position 411) GAGGACGGCAACATCCTGGGGCA hGHS-R IF (cDNA position bp360) TTCGTCAGTGAGAGCTGCACCTAC hGHS-R 2R (cDNA position bp833) AAATATCGCCCTACGTGGAAGG hGHRH-R F (cDNA position bp38) TGTTGAGCCCGTTACCGACC hGHRH-R R1 (cDNA position bp716) TGTCAGCACCTTTGCCGC hGHRH-R R2 (cDNA position bp1243) CAGCCACCAGAAGGCTCTCC mP-actin F (cDNA position bp95) TTGTAACCAACTGGGACGATATGG mp-actin R (cDNA position bp835) GATCTTGATCTTCATGGTGCTAGG Ghrelin F (cDNA position bp-9) ACCAAGACCATGCTGTCTTC Ghrelin R (cDNA position bp387) GCAAGAACTCACATCCGCCT Table 2.1 : Primer sequences for DNA fragment amplification in PCR and RT-PCR F=forward primer; R=reverse primer; m=mouse; h=human; r=rat. The combintations of these primers and the fragments they amplified are discussed in more detail in the results chapters.

2.7 Detection of DNA sequences by Southern blotting 2.7.1 Southern Blotting DNA (lOpg of genomic DNA or 0.5pg of plasmid DNA) was digested with the appropriate restriction enzymes, run out on agarose gels and the DNA transferred onto nylon membranes as described in the Hybond-N+ () user manual, using the capillary transfer method (Sambrook et at., 1989; Southern, 1975). 2.7.2 Radioactive random prime labelling Single stranded, gel purified DNA fragments (25ng) were radiolabelled with [a^^PJdCTP and [a^^PJdATP with the random primer labelling kit Prime-a-gene (Promega). Radiolabelled probes were cleaned by elution through Sephadex G-50 columns (Amersham Pharmacia Biotech). 2.7.3 Hybridisation of Southern Blots 20xSSC 3M NaCl, 0.3M NaCit, pH7.0 Denhardt’s solution 2% BSA, 2% Ficoll 400, 2% Polyvinyl Pyrollidine 20xSSPE 3M NaCl, 0.2M NaH^PO^, 25mM EDTA, pH7.4

44 Chapter 2: Materials and Methods

Filters from Southern blotting were washed in 2xSSC and pre-hybridised for at least 2 hours at 65°C in Prehybridisation Mix (5xSSPE, 5x Denhardt’s Solution, 0.5% SDS, 250pg/ml heterologous DNA). The prehybridisation mix was then replaced with fresh mix, supplemented with the single stranded radioactive probe and incubated at 65°C overnight. Filters were washed for one hour in O.SxSSC, 0.1%SDS at 65°C and exposed to Kodak BioMax film at -70°C. 2.8 Determination of protein content Protein content in tissue homogenates was measured with the BCA Protein Assay Reagent (Pierce, Rockford, IL). The OD^^z of samples and bovine serum albumin standards (Sigma) was measured in a Beckman DU-64 spectrophotometer. 2.9 Generation of transgenic mice 2.9.1 Purification of large DNA fragments for microinjection lOOpg cosmid DNA was digested with Notl, phenol/chloroform extracted, ethanol precipitated and resuspended in 200pl TE (lOmM Tris.HCl, ImM EDTA, pHS.O). A 5% - 25% NaCl gradient was prepared using a Hoefer SG 30 gradient former and the DNA layered on top of the gradient. The gradient was spun at 37000 rpm for 5 hours and harvested in 0.5ml aliquots, which were examined by gel electrophoresis. Aliquots containing the desired fragment were pooled, ethanol precipitated and resuspended in microinjection buffer (lOmM TrisHCl, pH7.5, O.lmM EDTA, pHS.O). 2.9.2 Superovulation, microinjection and embryo transfers Pre-pubertal 3-4 week old, superovulating, fertilised female mice ((CBA/ca x C57B1/10) FI) were received from the NIMR SPF unit. For egg collection these were culled and their oviducts removed and placed into M2 medium (Hogan et al, 1986). The oviducts were dissected to release the eggs, which were placed in M2 medium supplemented with 0.5mg/ml hyaluronidase (Sigma) to remove cumulus cells. The eggs were then washed twice in M2 and placed into M16 medium (Hogan et al, 1986) in a 37°C incubator supplemented with 5% CO2. In collaboration with K.Mathers (Biological Services, NIMR) DNA (2-lOng/pl) was injected into the pronucleus of fertilised one-cell mouse oocytes using standard methods (Hogan et a l, 1986). Viable eggs were incubated at 37°C until transfer into the oviducts of pseudo-pregnant mice (supplied by NIMR SPF

45 Chapter 2: Materials and Methods unit, (CBA/caxC57Bl/10) FI). Surgery was performed under 10% Avertin anaesthetic (O.lml/lOg bodyweight, ip; 100% solution: Ig/ml tribromoethanol in amylalcohol (Aldrich)), with 15 eggs being transferred into each infundibulum. Tail biopsies from resulting pups were analysed by PCR and Southern Blot. 2.10 Radioimmunoassay (RIA) PBS 50mM NaH2P04, lOOmM NaCl, 0.6mM Thimerosal, pH7.4 with NaOH 18%PEG 2ml/l 10% Triton-X, 1.5g y-Globulins, 330ml/l Tris Buffer (below), 667ml/l 27% Polyethylene Glycol Tris Buffer lOOmM Tris-HCl (pH 8.4), 0.6mM Thimerosal assay buffer PBS, 0.3% BSA Pituitaries were homogenised in 1ml PBS, spun in a bench top centrifuge at maximal speed for 2 minutes and diluted further in assay buffer. Hypothalamic blocks were homogenised directly in 0.5ml assay buffer and spun. In all assays standards were doubly diluted ten times in triplicate and the hormonal content of the samples was expressed in terms of the standard values. 2.10.1 lodination of tracers Recombinant proteins (5|xg rGH, rPrl, mGH and mPrl (NIDDK, Bethesda, MD)) and 5|ig eGFP (Clontech) were radioiodinated with Nal’^^ using the lodogen method as previously described (Robinson, 1980) and purified by Sephadex G75 chromatography. Tracers were used at 5-7000cpm/100|xl PBS + 0.3% BSA. 2.0’-mono succinyl cAMP tyrosine methyl ester (Sigma) was iodinated by using the lactoperoxidase method (Morrison and Bay se, 1970) and purified by a methanol gradient in 0.1% TFA on Sep-Pak C l8- cartridges (Waters, Milford, Massachusetts). It was used at 4000-6000cpm/50|il acetate buffer (50mM sodium acetate trihydrate, pH6.2 + 0.5% BSA). 2.10.2 Mouse and rat GH and Prl RIA Mouse and rat GH and Prl in pituitary or cell extracts was assayed by RIA as previously described for the rat (Bartlett et ai, 1990; Carmignac and Robinson, 1990) using reagents kindly provided by NIDDK.

46 Chapter 2: Materials and Methods

2.10.3eGFPRIA D.F. Carmignac (NIMR, London) developed a novel RIA assay for eGFP protein measurement. For assay, lOOpl of iodinated eGFP (5-7000cpm) were mixed with 100|xl of tissue extract or standard of recombinant eGFP (O.Ol-lOng) and lOOpl polyclonal antibody against GFP (Molecular Probes Inc.) at a dilution of 1:500,000 for 16b at room temperature. Bound and free fractions were separated by addition of 2 vols 18% polyethylene glycol, followed by 30 mins incubation and centrifugation. Radioactivity in the pellets was determined by gamma counting. 2.10.4cAMPRIA Samples, prepared as described in section 2.17.3, and the cAMP standard (Sigma) were acetylated by addition of 2%v/v trietbylamine and 1% acetic anhydride (Sigma). For assay acetylated extracts, made up to lOOpl with PBS or the standard (O.l-SOOfmol/tube) were mixed with lOOpl cAMP antibody (NIDDK, CV-27, final concentration 1:400,000) in acetate buffer and 50 |il of the iodinated tracer. The assay was incubated over night at 4°C and the antibody bound from free tracer was separated in the same way as described for eGFP. 2.11 Reverse transcriptase polymerase chain reaction (RT-PCR) RNA was prepared from homogenised mouse tissues using the Trizol reagent (Life Technologies). RNA was isolated from lysed GC cells using the SNAP Total RNA Isolation kit (Invitrogen). 500ng of total RNA were transcribed with 200units of M- LMV reverse transcriptase (Roche Diagnostics) in Ix M-LMV reverse transcriptase buffer (Roche Diagnostics) supplemented with Ipg random primers (Invitrogen), dNTPs (Amersham Pharmacia Biotech, 0.3mM), 40 units RNase Inhibitor (Promega) and 5mM DTT. The mixture was incubated at 37°C for 2hours and the desired cDNA amplified by PCR. 2.12 In Situ hybridisation 2.12.1 In vitro transcription Plasmids containing polymerase promoters were linearized with a restiction enzyme and repurified. Anti-sense and sense radiolabelled riboprobes incorporating [a^^S]-UTP were generated using a SP6/T7 transcription kit (Roche Diagnostics). After transcription probes were treated with RNAse free DNAse I (Roche Diagnostics) for 15 minutes at

47 Chapter 2: Materials and Methods

37°C and purified on a Sephadex G-50 Nick Column (Amersham Pharmacia Biotech). Alternatively riboprobes were non-radioactively labelled using the SP6/T7 DIG (digoxigenin) kit (Roche Diagnostics), DNAse treated and purified by phenol/chloroform extraction and ethanol precipitation. Oligonucleotides were [a^^S]-dATP labelled using the enzyme terminal transferase (Roche Diagnostics) according to manufacturer’s instructions and repurified by phenol/chloroform extraction and ethanol precipitation. 2.12.2 Riboprobe Hybridisation In situ hybridisations were performed by a previously published method (Bennett et al, 1995). Briefly, sections were fixed in 4% paraformaldehyde, acetylated, dehydrated and delipidated in chloroform. Sections were hybridised overnight at 45-50°C, in buffer containing IxlO^cpm probe/slide in 50% formamide, 0.025 M Tris, pH7.5, 0.001 M EDTA, pH 8.0, 0.4M NaCl, Ix Denhardt’s solution and 10% dextran sulfate (MW 500,000), under Nescofilm coverslips (Bando Chemical Industries, Ltd., Kobe, Japan). Following hybridisation, the slides were washed in 2xSSC/50%formamide at 45°C, digested in RNAse (40mg/ml) at 37°C for 30 minutes, rinsed in water and air dried. 2.12.3 Oligonucleotide Hybridisation Again, in situ hybridisations were performed by a previously published method (Young et al, 1986). Briefly, sections were fixed in 4% formaldehyde, acetylated, dehydrated and delipidated in chloroform. Sections were hybridised overnight at 37-42°C, in buffer containing IxlO^cpm probe/slide in 50% formamide, 4xSSC, 2mg/ml single stranded salmon sperm DNA, 4mg/ml yeast tRNA, Ix Denhardt’s solution and 10% dextran sulfate (MW 500,000), under Nescofilm coverslips (Bando Chemical Industries, Ltd., Kobe, Japan). Following hybridisation, the slides were washed in IxSSC at 55°C, rinsed in water and air dried. 2.12.4 DIG hybridisation Paraffin embedded slides were dewaxed with Histoclear (national diagnostics) rehydrated, permeabilized with lOfXg/ml proteinase K (Sigma), fixed in 4% paraformaldehyde, acetylated and incubated in prehybridisation buffer (4xSSC, 50% formamide). Sections were hybridised overnight at 45-50°C, in buffer containing 5-lOng probe/slide in 40% formamide, 10% dextran sulfate, Ix Denhardt’s, 4xSSC, lOmM DTT, 1 mg/ml yeast t-RNA and 1 mg/ml denatured and sheared salmon sperm DNA

48 Chapter 2: Materials and Methods under Nescofilm coverslips (Bando Chemical Industries, Ltd., Kobe, Japan). Slides were then washed in 2xSSC/50% formamide at 45°C, digested in RNAse (40mg/ml) at 37°C for 30 minutes, followed by a standard immunological detection protocol for DIG (Komminoth, 1992). I used sheep anti-DIG-alkaline phosphatase (Fab fragment; Roche Diagnostics) with NBT/BCIP (Nitro-blue tetrazolium/Bromo-chloro-indolyl-phosphate, Roche Diagnostics) as alkaline phosphatase substrate giving a purple colour. 2.12.5 Image analysis Slides were exposed to autoradiographic film (Biomax MR, Kodak, Rochester, NY) for several days and the resulting images analysed densiometrically, using the Image program (W. Rasband, NIH, Bethesda, MD). The values represent integrated optical density expressed in arbitrary units. For each transcript, comparisons were made with the same batch of labelled probe on sections from all animals exposed concurrently. For higher resolution images, slides were dipped into photographic G5 emulsion (Ilford Imaging, UK) according to manufacturer’s instructions. Slides were exposed for several days then developed in D-19 developer (Kodak), counter stained in cresy 1-fast violet and mounted in DPX (BDH). 2.13 RNAse Protection Assay In vitro transcription to generate [a^^P]-UTP radiolabelled anti-sense RNA probes was performed as described in section 2.12.1., then the probes were gel purified on a 5% acrylamide gel. Full size transcript bands were visualised by exposure to autoradiographic film (Biomax MR, Kodak) followed by excision and elution of the transcript. RNAse protection assays were performed using the RPA III kit (Ambion). Briefly, IxlO^cpm labelled probe and lOpg sample RNA (prepared as in section 2.11) were incubated at 42°C over night. After hybridisation, the mixture was treated with RNAse and protected fragments were separated on a 5% acrylamide gel. Gels were then dried, exposed to a storage phosphor screen (Molecular Dynamics) and digitally analysed using the ImageQuant (Molecular Dynamics) program. The amount of protected sample RNA was normalised by comparison to p-actin RNA fragments.

49 Chapter 2: Materials and Methods

2.14 Immunocytochemistry Immunocytochemistry was performed by E. Gevers and A. Sesay (NIMR). For immunocytochemistry from pituitary cell lines, cells on poly-L-lysine (Sigma) coated coverslips were incubated in 4% paraformaldehyde for 10 minutes and washed twice in IxPBS before inactivation of endogenous peroxidases. 8|im paraffin embedded sections were deparaffinsed in histoclear (national diagnostics) and rehydrated, in 100-30% graded ethanol baths. Endogenous peroxidase activity was blocked in 1% H2O2 in PBS/40% methanol at RT for 30 minutes. Sections were then washed 5 minutes in IxPBS and incubated in 0.05% trypsin (Sigma) in 0.2% CaCl2 (pH7.4) for 15 minutes. Non-specific binding was blocked by incubation in 2% sheep serum in blocking buffer (NEN Life Sciences) for 60 minutes. Sections were then incubated in a human-anti- rabbit 7 peptide C terminal GHS-R anti-body (1:75, kindly provided by A.D. Howard, Merck Research Laboratories, Rahway, NJ) at 37°C for 2 hours, followed by incubation with the second goat-anti-rabbit peroxidase labelled anti-body (1:200, Dako) at 37°C for 60 minutes. After incubation with biotin labelled tyramides (1:50 (NEN Life Sciences) in O.IM Tris/O.IM NaCl/10% dextrane sulfate/lOmM imidazole/1:10000 H2O2) for 20 minutes at RT in the dark, sections were incubated in peroxidase labelled rabbit-anti- biotin anti-body (1:300, NEN Life Sciences) for 45 minutes at 37°C. The last anti-body was visualised by diaminobenzidine (DAB) (0.05% (Sigma) in 50mM Tris/HCl/lmM imidazole/ 0.085% H2O2 for 5 minutes at RT). Finally sections were washed in tap water, counterstained with haematoxilin (BDH) and mounted in DPX (BDH). In control sections the first anti-body was omitted. For pituitary cell line immunocytochemistry all steps for tyramide enhancement were omitted and the second anti-body was visualised directly with DAB. 2.15 Jugular vein cannulation All jugular vein cannulations were performed by D.F. Carmignac, NIMR. Mice were anaesthetised with Sagatal (6mg/ml, O.lml/lOg body weight i.p., Rhone Merieux) and maintained under anaesthetic throughout the experiments. Jugular catheters were inserted for i.v. injections, and manual blood sampling was begun 1 hour later, as previously described (Carmignac and Robinson, 1990). GHRH (human GHRH-(1- 29)NH2) and GHRP-6 (Ferring, Malmo, Sweden) were dissolved in saline containing

50 Chapter 2: Materials and Methods

0.05%BSA, diluted in heparinised saline, and injected in lOOjXl, flushed in with lOOjil saline. For each peptide injection, samples of 50p.l blood were withdrawn 5 minutes before and after injection. After GHRH injection a period of 90 minutes was allowed before the next injection in the same animal, while 60 minutes were sufficient after GHRP-6 injections. Blood samples were centrifuged, and the plasma separated and frozen for hormone measurement. 2.16 In vivo Confocal Microscopy Transgenic rGHRH-eGFP mice were decapitated and the brains quickly removed and transferred to ice cold artificial cerebrospinal fluid (aCSF: 120mM NaCl, 3mM KCL, 1.2mM NaH2P04, 2mM CaClz, ImM MgSO^, 23mM NaHCOg, lOmM Glucose gassed with 95%02 and 5%C02). Coronal hypothalamic sections were cut at 150pm on a tissue vibrotome (Leica VT lOOOS) and placed in a purpose-built recording chamber mounted on a microscope stage. This chamber was perfused with gassed aCSF and maintained at 32°C. Low power images were initially collected to identify regions of fluorescent labelling. Labelled regions were then imaged at high magnification using a Zeiss Axioskop upright microscope (with an Olympus X60, 0.9NA, long-working distance water immersion objective) and a Biorad MRC 1000 scanning confocal microscope. Excitation was achieved using the 488nm line of an Argon ion laser. Emission filters were optimised for eGFP signals. Digital images were collected in COMOS (Biorad) and analysed offline in NIH image. 2.17 In vitro cell systems 2.17.1 Culture of rat GC cells All culture work was carried out in a Biomat Class II microbiological safety cabinet (Medical Air Technology Ltd.). Cells were cultured using a method modified from Kineman and Frawley (Kineman and Frawley, 1994). Cells were incubated at 37°C in a 5% C02 incubator in complete medium (D-MEM (Sigma), 15% Horse Serum (Sigma), 2.5% foetal calf serum (PAA, Austria), 2mML-Glutamine (Sigma) and Ix antibiotic antimycotic solution (Sigma)). Medium was changed every 2-3 days and cells harvested by centrifugation of the cell suspension at 130xg for 10 minutes. Cells were frozen at - 70°C before storage in liquid nitrogen.

51 Chapter 2: Materials and Methods

2.17.2 Cell Transfection GC cells were transfected with hGHS-R and hGHRH-R plasmids using Lipofectamine (Life Technologies). Briefly, 1-2 pg of plasmid DNA were diluted in lOOpl Opti-MEM (Life Technologies) and supplemented with 14pl lipofectamine in 100 pi Opti-MEM. After 45 minutes this mixture, supplemented with 0.8ml Opti-MEM, was added to 50- 80% confluent cells and incubated for -16 hours. Medium was then replaced with complete medium for 48 hours, after which selection for transfected GC cells was introduced by the addition of 250pg/ml G-418 (Life Technologies). 2.17.3 Treatment of GC cultures with hypothalamic factors Culture medium o f-80% confluent cells (GC, GHS-R+"'GC cells, GHS-R+"VGHRH-R+"" GC cells) was aspirated and replaced by treatment medium (DMEM + lOOnM GHRP- 6/GHRH). Cultures were incubated for 20 minutes to study cAMP accumulation. Medium was aspirated and stored at -20°C until analysed by RIA. Cells were washed in ice cold PBS pH7.4 and after addition of 0.5 ml ice cold acid alcohol (95% ethanol, O.lMHCl) cells were scraped off with a rubber policeman and incubated at -20°C over night. The homogenate was centrifuged at 14000rpm for 5 minutes and the supernatant stored at -20°C. For RIA, 50pl were removed, evaporated to dryness and the pellet resuspended in PBS (see section 2.10.4). 2.17.4 Phosphoinositol turnover The rate of PI hydrolysis was measured as described by Adams et ûe/.(Adams et a/., 1996). Briefly, -80% confluent cells (GC, GHS-R+"'GC cells, GHS-R+"VGHRH-R+"" GC cells) were incubated over night in DMEM supplemented with Ix antibiotic antimycotic solution, 0.5% v/v charcoal-stripped foetal calf serum and 5pCi [^H]-inositol (). Cells were then washed in serum-free DMEM containinglOmM LiCl (Sigma) and lOmM inositol and incubated in this medium supplemented with test substances for 2 hours at 37°C. The media was removed and the cells extracted with 3.3% perchloric acid. After the addition of lOM KOH supernatants were applied to anionic Dowex columns (AG 1- X8, BioRad) and the inositol phosphates (IP) eluted in IM ammonium formate (Sigma). Membrane bound phosphoinositol was determined by dissolving the extracted cells with IM NaOH and IM HCl. PI hydrolysis was then expressed as percentage of free PI over membrane bound PI 4- free IP (%PI tumover= free IP/(free IP+ bound PI)x 100%

52 Chapter 2: Materials and Methods

2.17.5 Culture of hypothalamic neurones 1-2 day old mouse pups were culled by an overdose of inhaled anaesthetic (Enfluorane), their brains removed, placed in ice cold HDMEM (25mM Hepes buffered Dulbecco’s medium, pH 7.3; Life Technologies) and the hypothalamus dissected into ice cold HEBSS (25mM Hepes buffered Ca^^ and Mg^’^ free Earles BSS pH 7.3; Life Technologies). Tissues were then transferred into a solution containing 6mg/ml papain,

1 mg/ml DL-cysteine, 250jiM EDTA, 50|liM p-mercaptoethanol in HEBSS and incubated at 37°C in a shaking waterbath for 20-25 min. The tissue was then washed twice in 5ml HEBSS and placed into a solution containing 1 mg/ml chick egg white trypsin inhibitor, 0.1 mg/ml DNAse in HDMEM. The tissue was diluted with 5ml final culture medium (1% N2 medium supplement, 10% Horse serum, 1% antibiotics, ImM Sodium Pyruvate, 20mM Glucose in Glutamine-free BME; all Life Technologies) and dissociated further by pipetting up and down in a syringe with kwill tube. After centrifugation at 200xg for 2 minutes, the supernatant was removed, the pellet resuspended in 5ml final culture medium and grown on 13mm diameter poly-D-lysine- and laminin-coated coverslips. Coverslips were removed from culture every 24 hours, fixed in 4% paraformaldehyde and mounted in Mowiol (Harco, Essex). Analysis of these rGHRH-eGFP hypothalamic cultures was performed by fluorescent confocal microscopy. 2.18 Statistical Analysis Unless otherwise stated, data are presented as mean±sem and statistical analyses were performed using the unpaired student t test. * = P<0.05, ** = P<0.01, *** = P<0.001

53 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R

3 In vitro studies in a GH cell line stably transfected with the human GHS-R

3.1 Introduction A variety of cultures from a GH- and Prl-producing rat pituitary tumour were established in 1965 (Tashjian eta l, 1968). Several clones of these GH cells have been maintained, one of them the GC clone (Yu et al, 1977). Its functional characteristics are GH production and specific responses to SS and thyroid hormone, but no response to GHRH can be measured. GC cells are similar to the original GHj cells in their regulatory responses, but have no detectable Prl production. Although GHSs release GH from freshly isolated pituitary cells (Bowers et al, 1984), most GH producing cell lines either do not respond to such secretagogues at all (Adams et al, 1998; Bercu et al, 1992), or only release minimal amounts of GH (Giustina et al, 1997) and show little or no GHS-R expression (Adams et al, 1998). Transfection of GC cells with GHRH-R cDNA failed to restore responsiveness of GC cells to GHRP-2 (Chen et al, 1998). However, GHS responses may be achieved in non-responding cells after transfection with exogenous GHS-R. Functional activity of a transfected GHS-R transcript has previously been shown in Xenopus oocytes, COS-7, 293-AEQ17 and CHO cells (Howard et al, 1996; Kojima et al, 1999; McKee et al, 1997a), but never in a GH cell line. When studying receptor physiology it seems more appropriate to do so in a cell line which endogenously provides the appropriate signal transduction proteins. GC cells are such a GH cell line, since they have most somatotroph characteristics, but are a much simpler and more controllable tool than whole pituitary cultures. GC cells transfected with the GHS-R create an ideal in vitro system to further elucidate the molecular mechanism of GHS signal transduction and GH release from GH cells. One of the main aims for these in vitro experiments was that I wished to ensure that hGHS-R type la cDNA was able to confer GHS responsiveness in stably GHS-R transfected GC cells (GHS-R^''® GC cells), before using the cDNA in transgenic studies. Appropriate GHS-R mRNA and protein expression, as well as functional activity in response to GHSs in GHS-R'^''® GC cells in vitro would encourage the use of this cDNA under exogenous promoter control in transgenic animals.

54 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R

GHRH and GHRP-6 show a marked synergistic effect in GH release in vivo, compared to an additive effect of these two peptides on GH release in vitro (for review see (Bowers, 1998)). Furthermore, it has previously been reported that GHSs have no direct effect on cAMP levels, while the combination of GHSs with GHRH markedly raised cAMP levels in vitro (Adams et al, 1996; Cheng et al, 1989; Wu et al, 1996). Augmentation of GHRH effects on intracellular cAMP levels by GHSs is perhaps mediated by cross talk between the PI-Ca^^/PKC pathway into the adenylyl cyclase system, a phenomenon previously shown in other cell types (Katada et ai, 1985; Kawabe eta l, 1994; Tang and Gilman, 1991; Yoshimasa etal., 1987). The GHS-R^''® GC cells could thus be used to study the interaction of these receptor transduction pathways in the same GH cell in vitro by transfecting GHS-R^^^ GC cells with the human GHRH-R.

3.2 Characterisation of GC ceils 3.2.1 Generation and detection ofGHS-R^^" GC cells Several lines (21) of stable GHS-R^''® GC cells were generated by A. Sesay (NIMR) by transfection of the human GHS-R type la cDNA (kindly donated by R.G. Smith and A.D. Howard, Merck Research Laboratories, New Jersey) into GC cells, using the lipofectamine method (see Chapter 2.17.2). hGHS-R expression was driven by the CMV promoter and RNA transcript stability ensured by a bovine GH polyadenylation signal; both elements are part of the pcDNAS vector (Invitrogen) as depicted in Figure 3.1. RT- PCR analysis, using hGHS-R primers 1 F and 2 R (Chapter 2.6, Table 2.1) could not detect GHS-R transcripts in untransfected GC cells but did amplify the expected 495bp human GHS-R transcript in GHS-R^"'® GC cells and a rat GHS-R cDNA control (Figure 3.1). Following the detection of hGHS-R transcripts in GHS-R^""^ GC cells, the presence of hGHS-R protein was shown by immunocytochemistry in collaboration with A. Sesay (NIMR). GC cells were transiently transfected with hGHS-R type la cDNA for easier comparison of untransfected GC cells and GHS-R'^"'® GC cells on the same slide. GHS- GC cells were detected with a hGHS-R antibody raised against the C terminal tail (a kind gift from A.D. Howard, Merck Research Laboratories, New Jersey) shown in

55 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R

Construct:

pCMV STOP bGH pA hGHS-R type 1a cDNA 1F

RT-PCR: M 1 rGHS-RcDNA

500 bp 2 no RT 395 bp 3 GC cells 4 GHS-R+ve cells

Figure 3.1 Transfection of the human GHS-R type la cDNA into GC cells The construct used to transfect GC cells drove hGHS-R expression with the CMV promoter and used the bGH poly A signal. RT-PCR of 0.5pg GC cell cDNA with primers for the hGHS-R amplified the predicted 495bp product only in GHS-R GC cells and a rat GHS-R cDNA control.

Figure 3.2. Immunocytochemistry thus confirmed that hGHS-R transcripts were translated and processed to result in hGHS-R protein. Higher magnification showed that the hGHS-R protein was concentrated in a ring around the cells, suggesting the receptor was expressed in the plasma membrane.

Figure 3.2 Detection of hGHS-R protein in GHS-R*''® GC cells Transiently hGHS-R transfected GC cells were fixed on coverslips and incubated with a C terminal 7 peptide human GHS-R primary antibody, followed by goat-anti-rabbit-HRP as secondary antibody and DAB staining. (40x magnification, scale bar = 10pm)

56 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R

3 .2 .2 Phosphoinositol turnover studies in GHS-R^'’^ GC cells Although hGHS-R mRNA and protein expression were detected in GHS-R^'"^ GC cells, the functionality of the GHS-R remained to be proven. Since the GHS-R transfers the extra-cellular receptor-mediated signal by activating phospholipase C, functional activity of the GHS-R can be measured by the hydrolysis-rate of membrane-bound phosphatidylinositol (PI) into diacylglycerol (DAG) and inositol-tris-phosphate (IP,). Initially, two GHS-R'^''® GC cell lines and untransfected GC cells were incubated ± lOOnM GHRP-2 for 2 hours and PI turnover was measured as previously described in section 2.17.4. Basal PI turnover in the absence of the ligand was elevated in GHS-R^''^ GC cells (Figure 3.3) (GC: 1.54±0.31%; GHS-R+"= 2: 2.08±0.11%; GHS-R+"" 13: 2.64±0.28%, n=3, mean±range). While GHRP-2 had no effect on PI turnover in untransfected GC cells, PI turnover was increased in GHS-R^''^ GC cells in response to GHRP-2 (Figure 3.3) (GC: 1.27±0.12%; GHS-R+"" 2: 7.64±0.36%, GHS-R+""

8 -

6 - Q) > O I c 4 - i 2 - i 0 -

Cell line GC GHS-R+''® GHS-R+''® no. 2 no. 13

Figure 3.3 GHRP-2 response in GHS-R^''® GC cells Untransfected and GHS-RGC cells were labelled with pH]-inositol, incubated ±100nM GHRP-2 and their rate of PI turnover measured. While untransfected GC cells showed no response to lOOnM GHRP-2, both lines of GHS-R^'"® GC cells responded with increased PI turnover. Basal PI turnover levels were also enhanced in GHS-R'^''® GC cells. (n=3; bars indicate the range of measurements)

57 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R

13: 6.17±0.44%, n=3, mean±range). Six of the 21 GHS-R'"''^ GC cell lines were tested for GHS responsiveness and cell line 2 proved to be the most responsive (data not shown). All further experiments were conducted with GHS-R^''^ GC cell line no.2. The PI turnover data were confirmed in 10 replicate experiments. Pooling results from GHS-R^''^ GC cell line no.2, the average PI turnover rose from 2.51±0.I4% basal to 5.27±0.34% (2.1-fold) with lOOnM GHRP-2 (n=30, P<0.0001). Basal PI turnover in untransfected GC cells and in GHS-R^''" GC cell line no.2 was 1.53±0.14% vs 2.19±0.15%(n=9, P<0.01). A dose range study demonstrated that responses to GHRP-2 were already maximal at concentrations above InM (Figure 3.4).

8 -

*** 6 - o 1 1 > o c B 4 .

2 -

0 - 0.1 1 10 100

nM GHRP-2 Figure 3.4 Dose range study for GHRP-2 in GHS-R^''® GC cells A dose range study for GHRP-2 revealed that responses were maximal from 1nM GHRP-2. Data were pooled from 4 separate experiments (n= 9-12 per dose; * P<0.05; ** P<0.01, *** P<0.001).

Similar studies were performed using GHRP-6 and a non-peptidyl GHS L-163,255 (a kind gift by A.D. Howard, Merck Research Laboratories) (Figure 3.5). The non-peptidyl GHS L-163,255 was previously shown to be an orally active GHS in pigs and was to be

58 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R used later in a feeding study with rats (Chapter 9). GHRP-6 was found to be less effective than GHRP-2 at lower doses, but maximal PI turnover was comparable at lOOnM (2.3-fold increase). L-163,255 was apparently more potent than GHRP-2, reaching maximal %PI turnover increases at O.lnM L-163,255 (Figure 3.5). The maximal %PI turnover increase was comparable to that of GHRP-2 and GHRP-6, although in this study the basal PI turnover was uncharacteristically high.

6 -

5 -

4 - % o c 3 -

2 -

1 -

0 0.1 1 10 100 0 0.1 1 10 100

nM GHRP-6 nM L-163, 255

Figure 3.5 Dose range study for GHRP-6 and L-163,255 in GHS-R'^''® GC cells While GHRP-6 was active only at lOOnM, L-163,255 reached maximal PI turnover increases already at 0.1 nM. Note that data for GHRP-6 and L-163,255 stem from different experiments. Data for these dose range studies were pooled from two experiments per compound (n=4-6).

A dose of lOOng/ml SS (maximally effective dose, personal communication E.F. Adams) alone had no effect on PI turnover and in one experiment when SS was given in combination with lOOnM GHRP-2, it did not affect the increase in PI turnover induced by GHRP-2 (Figure 3.6).

59 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R

5 -

S) > o ^ 3 -

2 -

1 -

0 - GHRP-2 SS GHRP-2 + SS

Figure 3.6 Somatostatin had no effect on GHRP-2 induced PI turnover in GHS- R+'=GC cells lOOnM GHRP-2 caused an increase in PI turnover, while neither SS (100ng/ml) on its own, nor in combination with GHRP-2 had an effect on PI turnover. (n=3; bars indicate the range of measurements)

3.3 Transfection of the human GHRH-R into GHS-R GC cells

3 . 3 . f G C GC cells have previously been shown to be GHRH unresponsive (Chen et ai, 1998), but these cells may respond when transfected with the GHRH-R. In order to study the interaction of the GHS-R and the GHRH-R transduction pathways in vitro, I attempted to restore GHRH responsiveness in GHS-R^''^ GC cells by further transfecting the cells with the human GHRH-R cDNA (kindly donated by W.Baumbach, American Cyanamid). GHS-R^''^ GC cells were transiently transfected with the hGHRH-R using the lipofectamine method (see Chapter 2.17.2). RT-PCR analysis with two different primer pairs specific to the hGHRH-R spanning across several introns (hGHRH-R F -i- hGHRH- R Rl, hGHRH-R F + hGHRH-R R2) (Chapter 2.6, table 2.1), showed hGHRH-R expression in transiently transfected GHRH-R/GHS-R^'"^ GC cells, but not in GHS-R^'^^ GC cells (Figure 3.7). (3-actin expression was used as a control for RNA quality in both cell extractions and detected in both GHS-R+'' and GHRH-R/GHS-R+'" GC cells.

60 Chapter 3; In vitro studies in a GH cell line stably transfected with the human GHS-R

GHRH-R/ GHS-R+ve G H S-R +ve hGHRH-R GC cells GC cells cDNA 1 r M

1.6kb 1 hGHRH-R primer 1+2 I.Okb 2 hGHRH-R primer 1+3 3 p-actin primers 0.5kb

Figure 3.7 Transfection of GHS-R^''® GC cells with the hGHRH-R RT-PCR analysis with two sets of primer pairs of O.Spg GC cell cDNA confirmed hGHRH-R expression only in GHRH-R/GHS-R""' GC cells.

3.3.2 of GC Transiently transfected cells were used for two separate experiments following the exposure to the appropriate ligands for these receptors: measurement of PI turnover and intracellular cAMP levels. Attempts at measuring GH release from GHRH-R/GHS-R'^''^ GC cells were made, but no reliable GH release could be measured. Baseline GH secretion also from untransfected GC cells was high and very variable and responses to GH releasing stimuli were therefore indistinguishable from baseline (data not presented). In a first experiment, expression of the hGHRH-R in a GHS-R^''^ GC cell did not influence the PI turnover in response to lOOnM GHRP-2 (basal: 1.88±0.I6%; lOOnM GHRP-2: 4.5±0.33%, n=3, mean±range; 2.4-fold increase compared to an average 2.1- fold increase in GHS-R^''^ GC cells). No increase from basal PI turnover was seen in response to lOOnM GHRH (I.54±0.23%), suggesting the GHRH-R signalling cascade did not involve phospholipase C. A combined dose of lOOnM GHRH plus lOOnM GHRP-2 also showed no difference to basal GHRH-R/GHS-R^''^ GC cell PI turnover levels (2.32±0.26%) (Figure 3.8). In fact it seemed that the addition of GHRH abolished the GHRP-2 response in GHRH-R/GHS-R+"' GC cells. The first thing that became apparent when measuring intracellular cAMP levels was that the values in treatment groups were very variable, giving large SEM. While lOOnM

61 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R

GHRP-2 did not seem to affect intracellular cAMP levels, lOOnM GHRH appeared to cause a trend towards elevated cAMP levels, which was slightly enhanced by a combined dose of lOOnM GHRH plus GHRP-2, although this did not reach significance (Figure 3.8). These data are very preliminary and need to be repeated in future with a larger number of cells per treatment group. PI turnover cAMP

5 - 25 -

4 - 20 - 0) > o c 3 - 15 1 3

1 - 5 -

0 - 0 - GHRP-2 GHRP-2 GHRH GHRP-2 GHRP-2 GHRH +GHRH +GHRH

Figure 3.8 Physiology of GHRH-R/GHS-R"^''® GC cells Transiently transfected GHRH-R/GHS-R^''® GO cells were treated with lOOnM GHRP-2, lOOnM GHRH or a combination of the two. PI turnover increased expectedly in response to lOOnM GHRP-2, but no such response was seen with lOOnM GHRH or a combination of the two. Intracellular cAMP levels showed a trend towards an increase in response to lOOnM GHRH, which seemed slightly enhanced by the addition of lOOnM GHRP-2. (n=3; bars indicate the range of measurements)

3.4 Discussion A stable GHS-R^''® GC cell line was created and analysed for the functionality of the hGHS-R protein. To my knowledge this is the first stable cell line in which a functionally active GHS-R is expressed in a GH cell. In contrast to GH3 cells (Adams et ai, 1998) no endogenous GHS-R transcripts were present in untransfected GC cells as was shown by RT-PCR. By immunocytochemistry the presence of the GHS-R protein was confirmed in GHS-R^^^ GC cells, but not in untransfected GC cells. Although the antibody was raised against the C terminus of the human GHS-R the amino acid

62 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R antibody was raised against the C terminus of the human GHS-R the amino acid sequence in that region is 86% homologous to the rat GHS-R, further strengthening the idea that no endogenous rat GHS-R protein is present in GC cells. The hGHS-R protein appeared to be more concentrated in the plasma membrane of GHS-R'^''® GC cells, suggesting its correct translation and processing as a 7 TM receptor. One of the main aims of the PI turnover experiments was to find out whether the hGHS- R cDNA, to be used subsequently in a construct for transgenic mice later, would be functionally active when driven by an exogenous promoter and stabilised by a GH polyadenylation signal. The robust > 2-fold increase in PI turnover in response to lOOnM GHRP-2 confirmed that the hGHS-R transcript and resulting protein was functionally active in GHS-R transfected cell lines. These studies encouraged me to use the hGHS-R cDNA under GH and GHRH promoter control and GH polyadenylation in my later transgenic studies (see Chapters 4-7). In PI turnover experiments I confirmed previous data by Adams et al. that like GH3 cell lines, untransfected GC cells are not responsive to GHSs (Adams et ai, 1998). GHS-R transfection not only restored GHS responsiveness, but also enhanced basal phospholipase C signalling in the absence of artificial ligand. Experiments by Koch et at. over-expressing the p^-adrenergic receptor in the myocardium of transgenic mice also resulted in enhanced activation of myocardial signalling and function in the absence of agonist (Koch et ai, 2000). The authors argued that G protein-coupled receptors are thought to exist at an equilibrium between an inactive and an active form of the receptor, with the active form being at low percentage in the absence of ligand stimulation. When these receptors are then over-expressed at high enough levels, the up-regulated previously low percentage of active receptors causes them to effectively function as though they were constitutively active. However, enhanced basal GHS-R signalling could also stem from the endogenous GHS- R ligand Ghrelin. Endogenous rat Ghrelin expression in GC cells has not yet been analysed, but judging from whole rat pituitaries (Kojima et al, 2000a), it is probably low or absent. Interestingly in these in vitro experiments the GHS-R signal transduction at least up to phospholipase C activation seemed independent of GHRH-R signalling, since this receptor is absent from GHS-R^"'® GC cells. As described in the introduction in vivo this

63 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R is clearly not the case, since litâit mice with a GHRH-R mutation do not respond to GHS treatment with GH release (Jansson et al, 1986b). Although these data might suggest that the cross-talk between the GHS-R and the GHRH-R is downstream of phospholipase C signalling, we cannot make such a conclusion in terms of GH release from GHS-R'^'"® GC cells. As mentioned previously, GC cells do not store GH well, such that basal GH secretion is high and variable. GH responses to GHSs can therefore not be measured in GHS-R'^^'® GC cells. The dose range studies for GHRP-2, GHRP-6 and L-163,255 reflect what has been reported for the potency of the three GHSs in in vitro rat pituitary assays previously, with GHRP-2 (EC5q= 3nM) being more potent at lower doses than GHRP-6 (EC5q= lOnM) and L-163,255 (ECgo=1.5nM) being the most potent amongst the three (Nargund and van der Ploeg, 1997). SS has been reported to inhibit cAMP, decrease intracellular Ca^^ (for review see Patel et al, 1995) and inhibit GHS induced GH release at the pituitary level (Bowers et al, 1984). GC cells remain responsive to SS (Yu et al, 1977). Here I demonstrated that the inhibiting effect of SS on GHS induced GH release is indeed downstream of phospholipase C activation, since SS did not block GHRP-2 induced PI turnover. These data support the idea that the antagonistic effect GHSs and SS have on each other lie in their opposite actions on somatotroph polarisation and intracellular Ca^^ levels (Bowers et al, 1994; Smith et al, 1996). The role of phospholipase C induced PI hydrolysis in the GHRH pathway has been the subject of some dispute. Although earlier reports suggested that GHRH stimulated PI turnover in cultured rat pituitaries (Canonico et al, 1983), this has been disputed in later studies (Escobar et al, 1986; French et al, 1990). Recently Ramirez et al reported PI turnover increases in response to GHRH only in a subpopulation of porcine somatotrophs (Ramirez et al, 1999). In my preliminary studies with a transiently transfected hGHRH-R in GHS-R^'"® GC cells, the PI turnover was unaffected by GHRH suggesting that phospholipase C is not involved in GHRH signalling. However, GHRH seemed to decrease GHRP-2 induced PI turnover in this experiment. Although not significant, GHRH appeared to at least enhance intracellular cAMP levels and there seemed to be a trend towards a further increase by GHRP-2, confirming previous studies, that GHSs have an enhanced effect on cAMP levels in combination with GHRH (Adams et al, 1996; Cheng et al, 1989; Cheng et al, 1993; Wu et al, 1997). This needs

64 Chapter 3: In vitro studies in a GH cell line stably transfected with the human GHS-R to be repeated before speculating on a mechanism, but one explanation might be that in the presence of a GHRH signalling system the main GHS response goes towards a cAMP response by G-protein cross talk. Another important factor is that the GHRH-R transfection was transient. To gain more consistent and robust responses to GHRH, it would be preferable to create a stable cell line that expresses both receptors, such that receptor expression is consistent among cells. Such a system would create a useful tool for analysing GHS-R and GHRH-R cross talk, for example by then using toxins specific to the G-proteins and other cascades involved. Recently Cunha and Mayo showed data in GHRH-R transfected GH3 cells, that GHRH activated the MARK pathway through a cAMP dependent but PKA independent pathway (Cunha and Mayo, 2000). Similar studies to analyse GHS cell signalling biochemistry would be possible in the stable hGHS-R'^''® GC cell line.

65 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

4 Targeting transgenes to hypothalamic growth hormone releasing hormone (GHRH) neurones This chapter describes the generation of two lines of transgenic mice both expressing transgenes under rat hypothalamic GHRH promoter control. In the first model a fluorescent reporter is expressed in hypothalamic GHRH neurones and in the second, GHS-R abundance is enhanced in these neurones. 4.1 Transgenic mice expressing eOFP in GHRH neurones 4.1.1 Introduction In 1996 Flavell et al. generated a new model of dominant GH-dependent dwarfism in the rat. Human GH (hGH) expression was targeted to hypothalamic GHRH neurones causing local feedback mechanisms to induce GH deficiency (Flavell et al., 1996). Appropriate transgene expression in the Tgr (Transgenic growth retarded) rat was achieved with a 38kb cosmid construct containing rat GHRH (rGHRH) regulatory elements. To avoid extrahypothalamic transgene expression, genomic hGH was inserted into the 5’ UTR of the rGHRH hypothalamic exon 1 (for GHRH gene structure see Figure 1.4). Transgene expression was restricted to the hypothalamic ARC and extrahypothalamic transgene expression was measurable only by RT-PCR and only at extremely low levels in the anterior pituitary, cerebellum, lung and testis, possibly reflecting endogenous GHRH expression (Matsubara et al, 1995). A major advantage of this rGHRH cosmid was that it had been engineered to allow insertion of other reporters into the cosmid using a unique Mlul restriction site. In the Tgr rats transgenic hGH protein in GHRH cells down-regulated transgene expression driven by the GHRH promoter by a short feed-back loop, thereby making transgene expression studies difficult. To facilitate further transgene expression studies with the rGHRH cosmid construct a fluorescent reporter, in this case enhanced green fluorescent protein (eGFP), was considered more useful than a physiologically active transgene. As described in chapter 1.1.1 decided to create transgenic mice, rather than rats, to allow crossing of transgenic mice with a GHS-R KG mouse in future. A transgenic mouse expressing eGFP driven by the rGHRH promoter would then help in investigating the expression of a transgene by this rat promoter in mice.

66 Chapter 4; Targeting transgenes to hypothalamic GHRH neurones

GFP is the companion protein to aequorin, cloned from the chemiluminescent protein from Aequorea jellyfish (Shimomura et al, 1962). It has been used widely as a genetic fusion partner to host proteins to monitor their localisation and fate (Tsien, 1998). In such tagging applications the gene encoding GFP is fused in frame with the endogenous protein and the resulting protein chimera are used to assay levels of gene expression or subcellular localisation. In transgenic animals, eGFP expressed from cell-specific promoters can then identify specific cell types in vivo. A transgenic mouse expressing eGFP under rGHRH promoter control would thus greatly facilitate the identification of hypothalamic neuronal subtypes in vivo. For example, to date electrophysiological studies of hypothalamic neurones require neuropeptide-characterisation of the recorded neurone after the experiment. The ability to visualise GHRH neurones would then enable us to study the physiological properties and processes of pre-identified neurones. At the beginning of our studies no such transgenic animals had been made, however, in the meantime this approach has successfully been used for other hypothalamic neurones such as oxytocin and gonadotrophin releasing hormone neurones (Spergel et al, 1999; Young et al, 1999).

4.1.2 The rGHRH-eGFP cosmid construct C. Magoulas (NIMR) engineered the transgene construct, targeting eGFP to rGHRH neurones, which I used to generate and analyse transgenic mice. Although the generation of this construct itself is thus not part of this thesis, our strategies in its design were relevant to my analysis of the transgene expression and physiology in the resulting transgenic mice and to the generation of my constructs later and are therefore discussed here in detail. When expressing eGFP in hypothalamic neurones, we aimed to target the fluorescent product to secretory vesicles, which might allow us to follow GFP’s transport to nerve terminals and subsequent release. The hGH signal peptide was previously shown to be sufficient to enable eGFP to be packaged into GH secretory vesicles in GC cell lines (Magoulas et al, 2000) and we therefore used this hGH signal peptide to target eGFP to secretory granules in GHRH neurones. Genomic sequence encoding the first 48 amino acids of the hGH gene product (signal peptide and N-terminal 22 residues of the hGH) was fused in frame via a ISmer oligonucleotide linker to the coding sequence of eGFP

67 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

(see Figure 4.1). Briefly, an Xba\-Not\ fragment (750bp) of the pEGFP-N3 CMV expression plasmid (Clontech Inc.), was blunt-ended by Klenow and ligated into the blunt ended PvwII sites of a hGH genomic clone (Flavell et al, 1996) containing 5’ and 3’ untranslated hGH sequences flanked by a Mlul linker (see Figure 4.1 !.))• This chimeric Mlul fragment was cloned into a 12kb Kpnl plasmid, containing rGHRH exons HI and 2-4, which was engineered to have a single Mlul restriction site in the 5’UTR (untranslated region) of the first hypothalamic exon (see Figure 4.1 2.)). The Kpnl fragment with the hGH/eGFP fusion cassette was packaged into a 38kb rat GHRH cosmid, containing 16 kb of 5’- and 14kb of 3’-flanking sequence (see Figure 4.1 3.))(Flavell et al, 1996). I then used the 38kb Notl fragment for microinjection.

4.1.3 Generation of rGHRH-eGFP transgenic mice The 38kb rGHRH-eGFP Notl fragment was purified on a salt-gradient and in collaboration with K. Mathers (Biological Services, NIMR) I injected the construct into fertilised oocytes of [(CBa/CaxC57Bl/10) FI] mice followed by oviductal transfer into pseudopregnant recipients. Resulting pups were assayed for the presence of the rGHRH- eGFP transgene by PCR and Southern Blot (Figure 4.2). Because of the different size intron between exon 1 and 2 in the mouse and human GH gene, PCR with primers hGH 5’UTR and hGH exon2 R (see table 2.1) resulted in two products; an endogenous mGH product 50bp smaller than the hGH transgene product (350bp). The reliability of this PCR genotyping method was tested several times by confirming the genotype of mice on a Southern Blot and the PCR method was found to be extremely reliable. In a total of 23 pups bom, 4 were transgenic for the rGHRH-eGFP construct. rGHRH- eGFP founders were set up to breed with wild-type C57Bl/10ScSn mice. Two founders, mouse 2 and 11, produced only 1 transgenic pup in 4 litters, so these positive FI pups were set up to breed instead. Both FI pups, 151 and 342, then gave 50% transgenic litters, suggesting the initial founders had been chimeric. The third founder, mouse 39, gave positive progeny and the fourth founder, 43, was sterile. As depicted in the Southern Blot in Figure 4.2, the endogenous mouse GH band was hard to visualise, possibly because the probe was designed in homology to the transgenic hGH rather than

68 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

Mlul Pvull Pvull Mlul

ATG hGH [_TAG i =

1 r Xbal Notl ATG eGFP TAG J____ — J 1 Mlul Mlul hGH-SP eGFP hGH-3'UTR 1 .) O C ATG STOP /

\ / \ / \ / \ / \ / Kpn I \ / Kpn \ / 2.) rGHRH \ \ M ul N \ N y Notl Kpn I Kpn I Notl

3.) rGHRH 1 II eGFP 1 D 1 1 1 1 1 2 3 4 5

Figure 4.1 Generation of the rGHRH-eGFP construct Essentially eGFP was fused to the hGH-signalpeptide (SP) and inserted into the first hypothalamic exon of the rGHRH gene. The rGHRH-eGFP construct was created by several steps. 1.) A 1.3kb chimeric hGH/eGFP MliA fragment was created by inserting eGFP into the coding region of a genomic hGH fragment, leaving eGFP fused to the hGH signal peptide. 2.) The chimeric hGH/eGFP Mlul fragment was inserted into a 12kb rGHRH Kpn\ fragment, which 3.) in turn was packaged into the 38kb rGHRH cosmid. Note that the fragments are not drawn to scale with each other.

69 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

the endogenous mGH. Because of relatively high sequence divergence in the first GH intron, the homology between mouse and human GH in the region of the probe was only 60%. The estimated transgene copy number from this Southern for founder 39 was measured as 9 times WT. However, since the Southern probe favours the transgene fragment due to better homology to the probe, the estimated copy number of 9 must be taken as an overestimate.

rGHRH hGH-SP hGH-5'UTR rGHRH itfS F g "

5'UTR ex2R I Southern Probe | 4.9 kb Bglll Bglll

B PCR: C Southern:

1 2 3 4

4.9kb-^ # # # 350bp. # 300bp- a

1.4kb 1 rGHRH-eGFP founder no. 39 :&■ 2 transgenic pup 3 WT littermate 4 transgenic pup

Figure 4.2 Analysis of rGHRH-eGFP transgenic mice A: Genotyping of rGHRH-eGFP transgenic mice was by PCR with primers hGH 5’UTR and hGH ex2R and by Southern blotting by digestion with 8^/11 and probing with a chimeric probe homologous to hGH-SP, ex1, ex2 and 65bp of eGFP. B: PCR resulted in a 350bp transgenic and 300bp endogenous GH product. C: Southern blot analysis revealed a strong 4.9kb transgenic hybridisation band and a much weaker 1.4kb endogenous mGH hybridisation signal.

70 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

4.1.4 Transgene expression in rGHRH-eGFP transgenic mice Transgene expression was studied by several methods: analysing eGFP mRNA and protein by RT-PCR, in situ hybridisation, radioimmunoassay, and fluorescent microscopy. To confirm eGFP expression in hypothalami of rGHRH-eGFP transgenic mice, RT-PCR with primers specific for the chimeric hGH/eGFP transgene was performed on hypothalamic cDNA from transgenic and WT mice (primers hGH 5’UTR + eGFP R, table 2.1);(Figure 4.3). The expected transgene-band was detected only in hypothalamic cDNA from transgenic mice, while p-actin cDNA was amplified in both transgenic and WT mice.

p-actin eGFP I 1 I 1 2 3 4 1.6kb ---- ► 1 rGHRH-eGFP hypo. Ikb ► 2 WT hypo. 3 no RT enzyme O.Skb— ► 4 no RNA

Figure 4.3 RT-PCR analysis of hypothalamic eGFP expression in rGHRH-eGFP transgenic mice RT-PCR primers amplifying mouse p-actin from O.Spg hypothalamic oDNA confirmed the presence of RNA in rGHRH-eGFP transgenic and WT mouse hypothalamic extracts. A hGH/eGFP chimeric fragment was only amplified in hypothalamic extracts of rGHRH-eGFP transgenic mice.

In order to locate the site of hypothalamic transgene expression, brains from rGHRH- eGFP transgenic mice were cut into 12pm sections on a cryostat and used in an in situ hybridisation assay with an eGFP anti-sense and sense riboprobe. A strong hybridisation signal was seen in the arcuate (ARC) nucleus, with a few positive neurones in the dorsomedial nucleus (DM) (Figure 4.4). Weak hybridisation was demonstrated in the zona incerta (ZI). No eGFP expression was seen in any other brain areas. No signal was observed in sections hybridised with a sense eGFP riboprobe. When anti-sense sections

71 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

were dipped in photographic emulsion and analysed at high power, distinct, heavily labelled neurones were obvious in the ARC (Figure 4.5). In the zona incerta neurones, were labelled less heavily, but specifically labelled neurones could be seen. GFP expression spread from the more central areas of the ARC laterally along the ventral region of the hypothalamus in more rostral sections. Although not measured quantitatively, comparison of transgenic mice of the three different lines - termed 39, 151 and 342 for the founder no. - showed no obvious differences in the distribution of ARC eGFP expression, but the 151 transgenic mouse brain also expressed some eGFP in the SCN, PVN and SON (data not shown).

rGHRH-eGFP Anti-sense WT Anti-sense

rGHRH-eGFP Sense

Figure 4.4 In situ hybridisation of eGFP mRNA in rGHRH-eGFP transgenic mice In the top panel an in situ hybridisation with an eGFP anti-sense and sense riboprobe showed that eGFP expression was restricted to the ARC with few positive DM neurones and a weak zona incerta (ZI) signal in rGHRH-eGFP transgenic mice (line no.39). Scale bar=1mm

72 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

'.y

* * *, . *

Figure 4.5 High resolution imaging of eGFP in situ hybridisation sections rGHRH-eGFP ARC sections from an in situ hybridisation with an eGFP anti-sense riboprobe were dipped in photographic emulsion, developed and counterstained with cresyl-fast violet. Imaging at xIO (top panel) and x40 (bottom panel) magnification showed the specific labelling of a proportion of ARC neurones with clusters of silver grains. The left panel shows the section in light field, while the right panel shows the same section analysed in dark field, (scale bar = 0.1 mm)

To compare eGFP expression in hypothalamic ARC neurones to endogenous GHRH expression, an I.M.A.G.E. mGHRH cDNA clone (1496474) was kindly donated by the UK HGMP Resource Centre, Cambridge. In situ hybridisation studies of hypothalamic mouse GHRH mRNA expression, using this clone to create a mGHRH riboprobe, revealed a strong hybridisation signal in the ARC with few positive neurones in the DM and a much weaker signal in the zona incerta (Figure 4.6). High resolution imaging of these sections after dipping in photographic emulsion demonstrated heavily labelled neurones in the ARC, spreading laterally in more rostral sections of the ARC (Figure 4.7). Few DM and zona incerta neurones were specifically labelled as well.

73 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

mGHRH Anti-sense mGHRH Sense

Figure 4.6 In situ hybridisation of GHRH mRNA in WT mice In situ hybridisation with a mouse GHRH riboprobe showed GHRH mRNA expression in the ARC of WT mice with few positive neurones in the DM and a much weaker signal in the zona incerta (ZI). Scale bar=1 mm

Bright Field Dark Field

Figure 4.7 High resolution imaging of GHRH in situ hybridisation sections Higher resolution images of the ARC nucleus of WT mice in mGHRH in situ hybridisation sections at x10 (top panel) and x 40 (bottom panel) magnification showed specific labelling of a proportion of ARC neurones. The left panel shows the same section at light field and on the right in dark field, (scale bar = O.lmm)

74 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

A double in situ hybridisation assay, using labelled eGFP and DIG labelled mouse GHRH riboprobes showed co-localisation of eGFP and GHRH mRNA in the ARC of rGHRH-eGFP transgenic mice (Figure 4.8). In rGHRH-eGFP transgenic mice from founder 39, 90% of 108 arcuate neurones expressed both mRNAs, while 6% expressed eGFP and 4% GHRH only. However, in rGHRH-eGFP transgenic mice from line 151, co-localisation of the two signals was only 81% with 16% of the arcuate neurones expressing eGFP only.

I

ARC 3V

Figure 4.8 Double in situ hybridisation for mGHRH and eGFP mRNA In situ hybridisation using a DIG-labelled mGHRH riboprobe and a ^^S-labelled eGFP riboprobe demonstrated co-localisation of the two mRNAs in the ARC of a rGHRH- eGFP transgenic mouse (line 39). Sections were dipped in photographic emulsion and analysed at high power (left panel x40 magnification, right panel x100 magnification). (scale bar = 10pm). An in situ hybridisation tissue screen for eGFP mRNA expression in transgenic rGHRH- eGFP mice from line 39 revealed no eGFP expression in pancreas, thymus, pituitary, stomach, testes, spleen, liver or heart. However, a strong hybridisation signal was obtained from the cortex of the kidney in some transgenic animals of all 3 lines (Figure 4.9). eGFP expression in the kidney was only seen in some littermates independent of sex or age of the animals. High resolution microscopy revealed heavily labelled podocyte cells in the glomerulus, but no GHRH mRNA expression was detected by in situ hybridisation in this area.

75 rGHRH-eGFP Anti-sense WT Anti-sense rGHRH-eGFP Sense 5 ■s ‘

I3 OQ f r I

O rGHRH-eGFP Anti-! o\<1 ZfHiA JL ^ ÆÊti m mm #

0 poX DC 1 O3 W

Figure 4.9 eGFP expression in the kidney of rGHRH-eGFP transgenic mice The top left panel shows eGFP mRNA expression in the cortex of the kidney of a ‘39’ rGHRH-eGFP transgenic mouse. High resolution imaging of a rGHRH-eGFP kidney section dipped in photographic emulsion on the bottom panel at x40 magnification showed that eGFP expression was restricted to cells in the glomerulus. top scale bar = 1 mm; bottom scale bar = 5pm Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

RIA showed that eGFP-immunoreactive protein was readily detectable in supernatants of hypothalamic block homogenates of rGHRH-eGFP transgenic and not WT mice. It also became apparent that the hypothalamic eGFP protein content was significantly higher in mice derived from rGHRH-eGFP founder 151. While mice from the 39 and 342 founders expressed 6.7±1.3 and 15.2±1.7ng eGFP/hypothalamic extract respectively, the eGFP content from 151 founder mice was 71.0±9.0 ng/hypothalamus (n=6, Figure 4.10). No significant difference between male and female eGFP expression was detected in transgenic mice from founder 39 {9.1 ±22 compared to 8.1±1.5 ng eGFP/hypothalamic extract, n=7, n.s., data not shown). eGFP protein was measurable only in one out of 4 rGHRH-eGFP kidney homogenates as 0.9ng/kidney.

80 - I r

60 -

% 4 0 - Q.>.

Q. Li. 20 - % U) c n.d. 0 -

39 151 342

WT rGHRH-eGFP

Figure 4.10 Hypothalamic eGFP protein content in rGHRH-eGFP transgenic mice Hypothalami of rGHRH-eGFP transgenic and WT mice were homogenised and assayed for eGFP protein content by RIA. While there was no eGFP detectable in WT mice, rGHRH-eGFP transgenic mice from founder 151 had significantly elevated hypothalamic eGFP content compared to transgenic mice from founder 39 and 342 (n=6; **P<0.01, ***P<0.001).

Having established hypothalamic eGFP mRNA expression and protein contents, the distribution of eGFP protein in the hypothalamus of rGHRH-eGFP transgenic mice was investigated. To gain an impression of the differences in fluorescence between founders.

77 Chapter 4; Targeting transgenes to hypothalamic GHRH neurones two animals of each founder were analysed by fluorescence microscopy. Hypothalamic slices of 150pm were cut on a vibrotome and the sections were mounted on slides in small chamber gaskets filled with aCSF (artificial cerebrospinal fluid). Hypothalamic slices of transgenic mice from rGHRH-eGFP founder 39 showed strong fluorescence in the ME and several fluorescent neurones in the ARC. rGHRH-eGFP mice from line 342 also showed strong fluorescence in the ME, but cell bodies in the ARC seemed not to fluoresce as strongly. Hypothalamic slices from transgenic 151 mice, however, showed very bright fluorescence in the ME and the whole hypothalamus. The hypothalamus was fluorescent with ‘clouds’ of eGFP protein and copious eGFP varicosities. On the basis of these images, rGHRH-eGFP transgenic animals from founder 39 were chosen for further analysis. Some vibrotome sections of kidneys from rGHRH-eGFP transgenic mice (line 39) showed eGFP fluorescence in the glomerulus, however, imaging of these sections was difficult due to strong autofluorescence. For fluorescent confocal microscopy live, hypothalamic slices of 150pm were cut on a vibrotome in oxygenated aCSF and in collaboration with N. Emptage (NIMR) these were analysed in chambers that were constantly perfused with oxygenated aCSF. In rGHRH-eGFP transgenic mice from founder 39 the terminal field of hypothalamic neurones in the ME was brightly fluorescent. Higher magnification of the nerve terminals in the ME showed that eGFP was transported along projections to the terminals (Figure 4.11). These higher magnification images also suggested that eGFP was indeed packaged into secretory granules. In the ARC fluorescent cell bodies were clearly visible, surrounded by non-fluorescing neurones (Figure 4.11). Higher magnification microscopy revealed that these fluorescing cell bodies had projections, which seemed to transport eGFP granules along them (Figure 4.11). Hypothalamic blocks of rGHRH-eGFP transgenic new-born mice (1-2 days, 10 pups from a heterozygous rGHRH-eGFP line 39 breeder) were dissociated and cultured on coverslips for several days, fixed in 4% paraformaldehyde every 24 hours and mounted in Mowiol. Analysis of primary cultures by fluorescent confocal microscopy showed few brightly fluorescent cells after 4 hours of culturing (Figure 4.12). After 72 hours only 1-2 fluorescing neurones were visible per coverslip, however these had clearly grown projections (Figure 4.12).

78 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

rGHRH-eGFP ME and terminal fibres

rGHRH-eGFP ARC neurones

Figure 4.11 Confocal microscopy of live rGHRH-eGFP hypothalamic slice preparations Live hypothalamic slices (150pm) from rGHRH-eGFP transgenic mice (founder 39) were placed in a perfusion chamber and analysed by fluorescent confocal microscopy. The top panel shows the terminal field of fluorescing GHRH neurones. eGFP fluorescence seemed granular and was transported along fibres to the ME. The bottom panel shows fluorescing GHRH cell bodies with freshly produced eGFP in the cytoplasm to be transported down the projections. Scale bar = 10pm

79 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

rGHRH-eGFP neurone 4h after culturing

i

%

’"ife'fe

rGHRH-eGFP neurone 72h after culturing

Figure 4.12 Confocal microscopy of rGHRH-eGFP hypothalamic ARC cultures

Hypothalami were dissected from new-born rGHRH-eGFP transgenic mice (founder 39), dissociated and cultured. The top panel shows a confocal scanning image through a single eGFP positive GHRH neurone 4 hours after culturing. The bottom panel shows a similar neurone 72h after culturing, demonstrating that eGFP GHRH neurones are detectable and viable in culture. Scale bar=10pm

80 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

4.1.5 Discussion In order to further characterise transgene expression driven by the rGHRH cosmid and to visualise and study the physiology of GHRH neurones, eGFP expression was targeted to these neurones in transgenic mice. GFP has recently been expressed in other neuroendocrine neurones, such as oxytocin or gonadotrophin releasing hormone neurones by transgenesis (Spergel et al, 1999; Young et al, 1999). However, this is the first report of transgenic mice expressing eGFP in GHRH neurones. Although minimal promoter sequences can express transgene reporters in specific cells, the expression level is often low and variable. Only by inclusion of all loci important for cell-specific expression of a transgene can position-independent copy-number-dependent transgene expression occur. Accordingly, Flavell et a l had isolated cosmids spanning large regions of the GHRH gene locus (Flavell et al, 1996) and shown that these cosmids effectively target transgenes to hypothalamic ARC neurones. It is important to note that this rGHRH cosmid has never been analysed for the presence of all DNAse hypersensitive sites involved in appropriate transgene expression and it cannot be claimed that this cosmid contains the full GHRH LCR. When using this construct to target eGFP expression to GHRH neurones, we fused the signal peptide and the first 22 amino acids of GH to the eGFP coding sequence to enable eGFP to be processed through the secretory pathway. Three transgenic rGHRH-eGFP transgenic mouse lines were established and their expression pattern studied. In two of the three lines the transgene expression pattern in the hypothalamus followed endogenous mouse GHRH expression. I found mouse GHRH expression mainly in the ARC, spreading laterally in more rostral ARC sections. These GHRH mRNA expression studies confirmed previous studies by Phelps et al in the mouse (Phelps et a l, 1993) and by Sawchenko et al in the rat (Sawchenko et al, 1985), which describe mRNA expression and immunoreactive cell bodies in these areas at similar expression levels. However, GHRH expression in a few specific neurones in the mouse zona incerta has not previously been reported. The zona incerta is a complex region containing a number of different cell body types and nerve terminals and has been implicated e.g. in the control of gonadotrophin release (Wilson et al, 1991). It is thus conceivable that GHRH as a pituitary regulating hormone might have some function in the zona incerta, but this

81 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones would need further investigation. Co-localisation studies confirmed that rGHRH-eGFP transgenic mice from line 39 co-expressed mouse GHRH and eGFP mRNA in 90% of all ARC neurones counted. The remaining 10% of ARC neurones that expressed either only eGFP or mouse GHRH might reflect hybridisation or DIG detection errors. Another possibility might be that in some neurones GHRH and eGFP might be expressed at low levels such that, due to a difference in relative sensitivity of the riboprobes only one of them is detectable. The fidelity of GFP identification of ARC GHRH neurones in rGHRH-eGFP transgenic mice from founder 39 is thus at least 94%. Only rGHRH-eGFP mouse line 151 showed transgene expression outside these regions. This was also reflected by the high eGFP protein content as detected by RIA and fluorescent microscopy. Direct comparison of copy numbers or quantitative measurement of ARC eGFP mRNA expression in the three rGHRH-eGFP transgenic mouse lines might have helped to explain the differences in eGFP expression. The difference in eGFP expression in animals from founder 151 might be due to the locus of insertion of the transgene, causing altered regulation of the transgene and therefore enhancing eGFP expression levels in areas outside endogenous GHRH. Considering that the ARC mRNA levels in rGHRH-eGFP transgenic mice from line 151 did not seem strikingly different from those of transgenic mice of line 39, the difference in ARC eGFP protein content and distribution is surprising. It seems that in 151 transgenic mice, translation and protein processing are significantly upregulated in the ARC. It has been shown previously, that levels of hypothalamic GHRH do not always change in parallel with GHRH mRNA (Chomczynski et al, 1988), because they reflect a balance between synthesis and release. For example, in this study I did not find a sexual dimorphism in eGFP expression in rGHRH-eGFP mice from founder 39, although reports suggest higher GHRH mRNA levels in males compared to females (Argente et al, 1991). A further interesting feature of rGHRH-eGFP transgenic mice is the eGFP expression in specific cells in the cortex of the kidney, which is seen in some littermates, but not others. The suggestion that there might have been problems with unspecific eGFP riboprobe cross-hybridisation in the kidney seems unlikely, since eGFP protein was measurable at least in one out of 4 transgenic kidneys. An explanation for this seemingly ectopic eGFP expression remains speculative. Since all 3 lines show kidney eGFP expression in some transgenic animals, but not others, the insertion site cannot be the

82 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones answer. Possibly the cosmid contains kidney-specific promoter sequences that are normally silenced, but why these sequences are active in some animals and not others is not easily explained. In general, the aberrant hypothalamic extra-ARC eGFP expression in rGHRH-eGFP transgenic mice from founder 151 and the eGFP expression in the kidney of all three rGHRH-hGHS-R transgenic lines shows, that the 38kb rGHRH cosmid does not contain all sequences necessary for position-independent and GHRH- specific tissue expression. Fluorescent and confocal microscopy of hypothalamic rGHRH-eGFP transgenic mice from founder 39 showed brightly fluorescing cell bodies of ARC neurones, with granular staining in the projections. The terminal field of these neurones, the ME, also showed bright fluorescence, with granular staining of projections ending in the outer layer of the MB. In 1986 Ibata et al. described GHRH immunoreactive neurones as ‘’beaded strands...particularly in the central ME’ (Ibata et al, 1986). In further electron microscopic studies they described dense granules in GHRH cell bodies and nerve terminals and suggested that GHRH was packaged into granules and transported down axons to release GHRH from the ME (Ibata et al, 1986). Data from rGHRH-eGFP transgenic mice suggested, that eGFP was also packaged into vesicles, that were transported along the projections to be secreted into the portal blood system from the nerve terminals in the ME. If so, eGFP could be said to follow the fate of endogenous mouse GHRH product. Co-localisation of eGFP and GHRH in the same granule would strengthen the idea that eGFP can be used to investigate GHRH secretory processes and could be studied by immunogold labelling and electron microscopy. Furthermore, I have shown that in primary hypothalamic cultures GHRH neurones are viable and readily detectable by eGFP fluorescence and that some neurones grow projections with continuous eGFP expression. Fluorescing cell numbers are small due to low GHRH cell numbers with even lower GHRH expression levels of new-born mouse pups (Hurley et al, 1998) and also pups were not genotyped before culturing, such that approximately 50% of the pups were not transgenic. Fluorescent-activated cell-sorting might be used to enrich the primary hypothalamic cultures for eGFP fluorescing neurones (Galbraith et al, 1999). These fluorescent and confocal microscopy data demonstrate the successful in vivo labelling of GHRH neurones with eGFP in rGHRH-eGFP transgenic mice.

83 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

4.2 Transgenic mice over-expressing the hGHS-R in GHRH neurones 4.2.1 Introduction One of the aims of this project was to distinguish between hypothalamic and pituitary GHS action by transgenic over-expression of the human GHS-R in hypothalamic GHRH neurones or pituitary GH cells. I aimed to determine whether altering the abundance of hypothalamic GHS-R in GHRH neurones alone would have an effect on GHS physiology. The successful targeting of the hGH gene to hypothalamic neurones in Tgr rats had encouraged me to consider the rGHRH cosmid to target GHS-R expression to GHRH neurones. In rGHRH-eGFP mice I demonstrated that the rGHRH cosmid did not contain the true rGHRH LCR. However, the co-localisation of the ARC mGHRH and eGFP mRNA and the confocal microscopy data showing distribution of eGFP fluorescence in specific cell bodies transported to the ME in hypothalamic slices in vivo convinced me that ARC eGFP was indeed following GHRH expression and fate and that the rGHRH cosmid would be able to target GHS-R expression to GHRH neurones. I had also tested the hGHS-R type la cDNA for functional activity in vitro and showed that transcription of the hGHS-R driven by an exogenous promoter and stabilised by the GH polyadenylation signal resulted in a functionally active receptor. The rGHRH cosmid and the hGHS-R type la cDNA were thus used to generate the rGHRH-hGHS-R construct.

4.2.2 The rGHRH-hGHS-R cosmid construct The cosmid construct targeting the hGHS-R to GHRH neurones was engineered similarly to the rGHRH-eGFP construct. Again, a chimeric Mlul fragment was used to insert the transgene into the unique M lul restriction site of the rGHRH cosmid. However, the hGH signal peptide was not included here, because the aim was not to direct GHS-R protein to secretory vesicles. Instead, the chimeric Mlul fragment was made by fusing hGH 5’ and 3’ UTR sequences directly to the hGHS-R type la cDNA without hGH signal sequences. I still wished to include the hGH 5’and 3*UTR and polyadenylation tail in the chimeric Mlul fragment, because these regions had proven to

84 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones give efficient and stable expression and processing of the hGH gene in Tgr rats (Flavell et al, 1996). Furthermore, as described in chapter 3, transcription of the hGHS-R type la cDNA in an expression vector using the bovine GH 3’ UTR as polyadenylation signal in transfected GC cells was found to produce a functionally active GHS-R. Thus a chimeric DNA fragment consisting of hGH 5’UTR, the hGHS-R type la cDNA and the hGH 3’UTR and polyadenylation signal, was engineered by PCR in several subcloning steps (Figure 4.13). In a first PCR step, 4 products were amplified: 14-2, 34-4, 54-7, 64-8. Primers 2R & 3F and 6F & 7R are entirely complementary to each other such that their products from the first PCR step overlap and self-anneal in a second PCR reaction. Therefore in a second PCR step primers 14-4 and 54-8 were used to create hGH/hGHS-R chimeric MluVBsrGl and MluVBstEH fragments respectively. Digestion and ligation with the original hGHS- R type la cDNA clone then gave the final hGH/hGHS-R chimeric Mlul product. The completed Mlul fragment (Figure 4.13) was sequenced and no base pair mismatch was found. Subcloning of the chimeric hGH/hGHS-R Mlul fragment was as described for the rGHRH-eGFP cosmid construct. The chimeric Mlul fragment was cloned into the 12kb rGHRH Kpnl fragment with the unique single Mlul restriction site in the 5’UTR of the first hypothalamic exon (see Figure 4.14 2.)). The Kpnl fragment was then packaged into the 38kb rat GHRH cosmid, containing 16 kb of 5’- and 14kb of 3’-flanking sequence (see Figure 4.14 3.)), resulting in the final rGHRH-hGHS-R construct. A restriction map confirmed the orientation of the rGHRH-hGHS-R cosmid (data not shown) and I used the 38kb Notl insert fragment for microinjection in fertilised mouse embryos.

85 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

EcoRI Mlul BsrGI IIF^ 3F V 5 F ^ 6F \ Notl

hGH ATG hGHS-R STOP hGH 1 V 4 R \ 7 P T \8 F BstEII Mlul

PCR 1+2 PCR 3+4 PCR 5+7 PCR 6+8

EcoRI BsrGI 5F Mlul

Mlul BstEII , Notl \ / I \ \ / I PCR+Mlul/BsrGI digest \ PCR + Mlul/BstEII digest ^ \ / I / hGH ATG_ hGHS-R STOP hGH - C A L ! _

Mlul BsrGI BstEII Mlul EcoRI Notl

Figure 4.13 Construction of the hGH/hGHS-R chimeric MliA fragment To create a chimeric hGH/hGHS-R Mlul fragment Primers IF and 8R introduced artificial Mlul restriction sites at the 5’ and 3’ end of the fragment. Primers 3 & 2 and 6 & 7 are chimeric and entirely complimentary to each other therefore allowing the annealing of PCR products 1+2 and 3+4, and 5+7 and 6+8 respectively in a second PCR amplification step. Digestion with the appropriate enzymes of the chimeric 5’ and 3’ fragments and ligation into the original hGHS-R clone with the EcoRI and Notl sites allowed easy access to the Mlul fragment. Please note that fragments are not drawn to scale with each other.

86 Chapter 4; Targeting transgenes to hypothalamic GHRH neurones

hGH-5'UTR hGH-3'UTR 1.) ATG hGHS-R STOPT

M ul Mlul

Kpn I Kpn I

2 .) TSHÏÏÏÏ- jjL H1 2 3 4 %

M ul N % S y Notl Kpn I Kpn I Notl

3.) rGHRH I |LhGH6-ft|| | | | || 1 2 3 4 5

Figure 4.14 Assembly of the final rGHRH-hGHS-R construct Essentially the hGHS-R type 1a cDNA was fused to hGH 5’ and 3’ UTR sequences and inserted into the first hypothalamic exon of the rGHRH gene. The rGHRH-hGHS-R construct was created by several steps. The chimeric hGH/hGHS- R Mlul fragment (1) was inserted into a 12kb rGHRH Kpn\ fragment (2), which in turn was packaged into the 38kb rGHRH cosmid (3). Note that the fragments are not drawn to scale with each other

4.2.3 Generation of rGHRH-hGHS-R transgenic mice The purified 38kb Not! rGHRH-hGHS-R fragment was microinjected into fertilised (CBa/CaxC57Bl/10)Fl mouse oocytes in collaboration with K. Mathers (Biological Services, NIMR). Surviving eggs were transferred into the oviducts of pseudopregnant recipients, and the resulting pups were assayed for the presence of the rGHRH-hGHS-R transgene by PCR and Southern Blot (Figure 4.15). Genotyping by PCR for rGHRH- hGHS-R transgenic mice was with 3 primers; as described earlier, the primers hGH 5’UTR and exon 2R were used to amplify a 300bp mGH fragment and a third primer hGHS-R 4R amplified a chimeric hGH/hGHS-R 506bp fragment with the hGH 5’UTR primer (table 2.1). Again, the reliability of this PCR genotyping method was tested

87 Chapter 4; Targeting transgenes to hypothalamic GHRH neurones

several times by confirming the genotype of mice on a Southern Blot. This PCR method proved to be extremely reliable. Of 42 pups, one female was found to be transgenic and was mated with a wild-type (CBa/CaxC57Bl/10)Fl male, giving positive progeny. These were bred further with C57Bl/10ScSn WT mice. The estimated transgene copy number from several Southerns was measured as 8 times WT. However, previous studies have shown that using a

M M 1 506bp 5UTR<------►4R 1 rGHRH 1 ■ hGHS-R ^.111 I i rGHRH 1 1 2 3 4 5 Southern probe Bglfl -► Bglll 4.7 kb

C Southern Blot:

1 2 3 4

506bp 4.7kb 300bp

1.6kb 1 transgenic founder 2 transgenic pup 3 WT litter mate 4 transgenic pup 1 transgenic founder 2 transgenic pup 3 WT littermate 4 cosmid control

Figure 4.15 Analysis of rGHRH-hGHS-R transgenic mice A; Genotyping of rGHRH-hGHS-R transgenic mice was by PCR with primers hGH 5'UTR, hGH ex2R and hGHS-R 4R and by Southern blotting by digestion with Bgl\\ and probing with a hGHS-R probe. B: PCR resulted in a 506bp transgenic and 300bp endogenous mGH product. C: Southern blot analysis showed a strong 4.7kb hGHS-R transgenic band and a weaker 1.6kb endogenous mGHS-R hybridisation signal.

88 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

human probe specific to the transgene in Tgr rats gave copy number estimates double those obtained when a rat probe was used to identify the transgene (doctoral thesis, S.Wells, NIMR). Since a human GHS-R probe was used to identify transgene copy numbers, rather than a mouse probe, the estimation of copy numbers in rGHRH-hGHS- R transgenic mice is probably an overestimate.

4.2.4 Transgene expression in rGHRH-hGHS-R transgenic mice Transgene expression in rGHRH-hGHS-R transgenic mice was probed by RNAse protection assay, in situ hybridisation and immunocytochemistry. RNA was extracted from the hypothalamus, the brain minus the hypothalamus, pituitary, heart and kidney of rGHRH-hGHS-R transgenic and WT mice and hybridised to a probe specific for the rGHRH-hGHS-R transgene. Subsequently samples were RNAse digested and separated on an acrylamide gel. Hybridisation of the probe to remaining DNA in the RNA extracts could be distinguished, because the band for DNA hybridisation was 30 bp larger. Two hypothalamic blocks and 6 pituitaries were pooled for RNA extracts. The RNA content was controlled by analysing P-actin mRNA expression in the same samples and assay. A protected transgene RNA fragment was seen only in the hypothalamus of rGHRH-hGHS-R transgenic mice (Figure 4.16). To localise transgene expression in the hypothalamus, brains of rGHRH-hGHS-R were cut into 12pm sections on the cryostat and hybridised with a hGH oligoprobe, specific to the 5’UTR hGH sequences in the transgene. Transgene expression was restricted to the ARC in transgenic mice (Figure 4.17 top panel). In situ hybridisation with a riboprobe for rat GHS-R type la cDNA was used to measure total GHS-R expression levels, since this probe equally measures human and mouse GHS-R mRNA. GHS-R expression was clearly elevated in the ARC of transgenic animals compared to WT littermates (Figure 4.17 middle panel). Hypothalamic and hippocampal WT GHS-R expression was as previously described for the rat with GHS-R expression in the VMN, CAl, CA3 and dentate gyrus of the hippocampus (Bennett et al, 1997). However, VMN GHS-R expression in mice was found to be much lower in comparison to rat VMN GHS-R expression. Low endogenous expression levels and high background variation made quantitative analysis of GHS-R over-expression in the ARC of transgenic animals

89 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

brain hypoth. pituitary minus heart kidney hyaoth.

rGHRH-hGHS-R transgene

P-actin

Figure 4.16 RNAse protection analysis of the rGHRH-hGHS-R transgene RNAse protection analysis of lOpg mRNA from tissues of rGHRH-hGHS-R transgenic and WT mice with a rGHRH-hGHS-R transgene and p-actin specific probe showed that rGHRH-hGHS-R transgene mRNA expression was restricted to the hypothalamus of transgenic mice (T= transgenic; NT = non=transgenic). compared to WT littermates by densitometry of in situ hybridisation difficult. However, when these sections were dipped into photographic emulsion and analysed by high resolution microscopy, ARC neurones of transgenic mice were clearly heavily labelled, while in WT littermates a signal above background, specifically labelling a certain neurone was difficult to identify (Figure 4.17 bottom panel). The distribution of hOHS- R labelled neurones in rGHRH-hGHS-R transgenic mice was similar to that of endogenous GHRH. As described above a quantification of hGHS-R over-expression compared to WT littermates was difficult because of low endogenous GHS-R expression levels in WT mice and variable background signals in the in situ hybridisation. However, on the basis of some sections from the rGHS-R riboprobe in situ hybridisation with equal background and the accumulation of grains in ARC neurones in dipped rGHRH-hGHS- R sections compared to WT, I estimated the over-expression of the hGHS-R in rGHRH- hGHS-R transgenic mice to 6-fold. This number will need confirmation. An in situ hybridisation screen of rGHRH-hGHS-R and WT mouse tissues with a hGHS-R type la cDNA probe showed no transgenic or endogenous GHS-R expression

90 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

in pancreas, stomach, pituitary, gut, spleen, kidney, liver or heart, confirming the absence of the transgene from pituitary heart and kidney, previously measured by RNase protection analysis. Furthermore, an in situ hybridisation with a rGHS-R riboprobe confirmed that the receptor was not expressed in the stomach or gut of WT mice.

WT rGHRH-hGHS-R

hGH-5'UTR oligoprobe

rGHS-R riboprobe

rGHS-R riboprobe

Figure 4.17 Hypothalamic transgene expression in rGHRH-hGHS-R transgenic mice Top panel: In situhybridisation with a hGH-5’UTR oligoprobe specific to the transgene showed transgene mRNA expression was limited to the ARC of rGHRH-hGHS-R transgenic mice. Middle panel: Hybridisation with a rGHS-R riboprobe measuring total GHS-R expression confirmed over-expression of the GHS-R in transgenic mice. Bottom panel: Sections from the rGHS-R hybridisation were dipped in photographic emulsion, counterstained with cresyl fast violet and analysed by dark field microscopy (x40 magnification). Highly labelled neurones appeared in the ARC of rGHRH-hGHS-R transgenic mice in a GHRH-like pattern, -pop and middle scale bar = 1mm; bottom scale bar = 50pm

91 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

For immunological analysis, I perfused mouse brains with 4% paraformaldehyde, paraffin embedded and cut them to 6|im sections. E. Gevers (NIMR) performed initial immunocytochemistry experiments with a 7 peptide C terminal hGHS-R anti-body (a kind gift from A.D. Howard, Merck Research Laboratories, Rahway, New Jersey) to localise GHS-R protein expression in rGHRH-hGHS-R transgenic mice. The mouse GHS-R protein is only laa different to the hGHS-R protein in this region. Preliminary results showed elevated staining in the ARC and ME of rGHRH-hGHS-R transgenic mice. Specific heavily labelled neurones were visible in the ARC of rGHRH-hGHS-R transgenic mice, while there was no staining in the ARC of WT mice (Figure 4.18). WT mice showed weak staining in fibres in the ME, which was clearly elevated in rGHRH- hGHS-R transgenic mice (Figure 4.18). A similar protein expression pattern with strong staining for the GHS-R in the ME was also observed in studies in the rat by Merck Research Laboratories, Rahway (unpublished data, personal communication J. Woods).

4.2.5 Discussion

I aimed to target GHS-R over-expression to hypothalamic GHRH neurones, using the previously described 38kb rGHRH cosmid (Flavell et al., 1996). In contrast to rGHRH- eGFP transgenic mice, here the reporter was a functionally active receptor and not to be targeted to the secretory pathway. Therefore the hGHS-R type la cDNA was fused directly to hGH 5’UTR sequences, omitting the hGH signal peptide, but including hGH 5’ and 3’UTR sequences that had been shown to cause efficient and stable transcription (Flavell et al, 1996). In vitro studies had confirmed that at least the GH 3’UTR would make a stable transcript, which could be translated into a functioning receptor (see Chapter 3). Indeed expression studies in rGHRH-hGHS-R transgenic mice showed that the hGH 5’and 3’ UTR/hGHS-R type la cDNA chimeric fragment was appropriately transcribed and translated. Microinjection of the final 38kb cosmid construct resulted in the generation of a rGHRH-hGHS-R transgenic mouse line. In comparison to microinjections with the rGHRH-eGFP construct, when 4 transgenic founders were created from 23 pups, the rate of success was much lower for the rGHRH-hGHS-R construct, where only one founder was obtained from 42 pups. The same DNA concentration was used for microinjection.

92 nrr "O rGHRH-hGHS-R WT

1. 3 CTO g %3 ARC Q 0 v;zr "B - ^4- ^ 1 VO oI w Q X70 X m3 c o ME

Figure 4.18 Immunocytochemistry for hGHS-R protein in rGHRH-hGHS-R transgenic and WT mice Immunocytochemistry with a 7 peptide C terminal hGHS-R anti-body on rGHRH-hGHS-R transgenic and WT paraffin embedded 6pm hypothalamic sections showed strong staining of neurones in the ARC of rGHRH-hGHS-R transgenic mice (top panel, x40 magnification)). Staining of fibres in the ME was also enhanced in these animals, compared to WT littermates (bottom panel, x40 magnification), scale bar = 0.1mm Chapter 4: Targeting transgenes to hypothalamic GHRH neurones

but the viability of the microinjected eggs is highly dependent on the cleanliness of the DNA. It is feasible that the difference in microinjection success rate stems from a difference in purity of the DNA. The female founder was mated and transferred the transgene onto its progeny. As described in chapter 1.5, the phenotype a transgenic animal displays, could be caused by the insertion of the transgene in an essential region for the expression of an endogenous gene. To eliminate the possibility that a phenotype seen in rGHRH-hGHS-R transgenic mice might be such an insertion-site effect, at least a second line of rGHRH-hGHS-R transgenic mice should be created in future. This is particularly important, since I showed that the rGHRH cosmid does not contain the full GHRH LCR. However, careful analysis of the transgenic phenotype can give an indication what the likelihood of an insertion-site effect might be. In situ hybridisation with an oligoprobe revealed that the hGHS-R transgene expression was restricted to the ARC. A riboprobe measuring endogenous and exogenous GHS-R mRNA confirmed over-expression of the GHS-R in the ARC of rGHRH-hGHS-R transgenic mice with highly labelled cells in a GHRH-like pattern. The over-expression of the GHS-R was estimated to approximately 6-fold in rGHRH-hGHS-R compared to WT littermates, but this number will need confirmation. This marks the difference in endogenous expression levels driven by the GHS-R and GHRH promoters, with GHS-R abundance being much lower than GHRH levels. Co-localisation studies for GHRH and eGFP mRNA in the ARC of rGHRH-eGFP mice had previously demonstrated their expression in the same cell. Since the expression pattern of the hGHS-R in the ARC of rGHRH-hGHS-R transgenic mice was distinctly GHRH-like, these co-localisation studies were not repeated. I had intended to specifically elucidate the role of ARC GHRH neurones as a GHS target involved in GH release. It is thus important to note that overexpression of the GHS-R was indeed only in the ARC with no enhanced GHS- R expression in the VMN, where significant levels of endogenous GHS-R are expressed. Interestingly the endogenous GHS-R expression in the VMN seems much lower in the mouse in comparison to the rat. No transgene expression was measured in the zona incerta of rGHRH-hGHS-R transgenic mice, neither with the oligo- nor the riboprobe. The copy number in rGHRH-eGFP and rGHRH-hGHS-R transgenic mice seemed similar, but transgene expression might be more efficient for the eGFP construct, which contains the hGH ATG and signal peptide. Zona incerta transgene levels might be too

94 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones low to be measured in rGHRH-hGHS-R transgenic mice. The other possibility is a difference in quality of the riboprobes. The eGFP riboprobe is a highly specific probe that consistently labels very well. The eGFP sense probe shows no hybridisation signals at all. The rGHS-R hybridisation probe however labels less well and the corresponding sense probe shows relatively high background labelling. Therefore the eGFP riboprobe might be more sensitive to lower expression levels, which the rGHS-R riboprobe will not detect. In situ hybridisation and RNAse protection tissue screens with a hGHS-R specific probe revealed no transgene expression in the brain minus the hypothalamus, heart, kidney, liver, pancreas, stomach, gut, spleen and pituitary. In rGHRH-eGFP transgenic mice I had found eGFP expression in glomerular cells in the kidney, but in rGHRH-hGHS-R transgenic mice no such expression was found. This underlines my previous finding that the rGHRH cosmid does not contain all the regulatory sequences to make it an LCR. Flavell et al. had found low transgene expression in the pituitary of the Tgr rat (Flavell et al., 1996). Neither RNAse protection nor in situ hybridisation with a hGHS-R probe elicited transgene expression in the pituitary. This finding is also important for later physiological studies in rGHRH-hGHS-R transgenic mice, since exogenous hGHS-R over-expression in the pituitary would have seriously compromised the interpretation of any effects. I found by in situ hybridisation with a rGHS-R probe, which also measured endogenous GHS-R expression, that the GHS-R is not expressed in the stomach or gut of WT mice. This is in contrast to findings by Kojima et al., who analysed GHS-R expression in the rat by RT-PCR and found its expression in the heart, lung, pancreas, gut and adipose tissue (Kojima et al., 1999). This difference is probably due to the differing sensitivity of in situ hybridisation and RT-PCR. RNAse protection assays would help to establish whether significant amounts of GHS-R expression are present in the stomach and gut. With the high levels of expression of the GHS-R ligand Ghrelin in the stomach and gut, it would be interesting to verify whether there is any relevant local expression of the receptor. Inununocytochemistry studies using a C terminal human GHS-R anti-body localised GHS-R protein to cell bodies in the ARC and terminals in the ME of rGHRH-hGHS-R transgenic mice. In WT mice no endogenous GHS-Rs were visible in cell bodies in the ARC, but light staining for the GHS-R was found in fibres of the ME. These preliminary

95 Chapter 4: Targeting transgenes to hypothalamic GHRH neurones immunocytochemistry studies suggested that a high amount of GHS-R is located in the ME. This new finding of GHS-R expression in the ME opens a new perspective in GHS physiology. Due to its close proximity to the hypophyseal blood system, the hypothalamus is one of the few areas of the brain outside the blood brain barrier. GHS-R expression in the nerve terminals might suggest that GHSs could reach their site of action by being taken up from the hypophyseal blood system and binding to the GHS-R directly. With GHS-R expression so close to the hypothalamic GHRH release site, a mechanism for the GHS-R to facilitate GHRH release directly from the nerve terminals is feasible. Interestingly the sstrl was also reported to be expressed in nerve fibres in the external zone of the ME (Helboe et al, 1998) and the authors suggested that the sstrl, co-localising with SS, might act as an auto-receptor. This is an interesting concept, considering that Ghrelin peptide, a ligand for the GHS-R, was localised to ARC neurones in the hypothalamus. Furthermore Helboe et a l reported that sstrl immunoreactive terminals made presynaptic contacts on other nerve terminals and they suggested a regulatory function of SS on the release of other neurohormones from these terminals. The nerve terminals in the ME might thus be an additional point of GHS/SS interaction in the hypothalamus. It is important to note that the quality of the GHS-R anti-bodies is not ideal, making immunocytochemical studies difficult to interpret. However, in the light of its potential importance for regulation of GH secretion, the protein expression pattern of the GHS-R in the hypothalamus of WT and transgenic animals warrants further investigation, if and when an improved GHS-R anti-body becomes available.

96 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

5 Physiological studies of rGHRH-eGFP and rGHRH- hGHS-R transgenic mice In Chapter 4 I described the generation of two transgenic mouse lines expressing transgenes in hypothalamic GHRH neurones. In both cases the transgenes, eGFP in rGHRH-eGFP transgenic mice and the hGHS-R in rGHRH-hGHS-R transgenic mice, were expressed in a GHRH-like pattern. In this chapter I describe the analysis of the resulting phenotypes. 5.1 Studies of rGHRH-eGFP transgenic mouse physiology 5.1.1 Introduction rGHRH-eGFP transgenic mice express a fluorescent marker in a specific cell type, which, other than following GHRH expression levels, packaging and transport, is not expected to impact directly on GH physiology. For example, Suter et al. have previously shown, that eGFP expression in GnRH neurones did not influence GnRH neurone’s electrophysiological properties in hypothalamic slice cultures (Suter gr a/., 2000). As described in Chapter 4, eGFP fluorescence in rGHRH-eGFP transgenic mice was shown to be granular along the projections of GHRH neurones and in the terminal field, in the external layer of the ME, suggesting eGFP was indeed being transported and secreted like GHRH. However there are possibilities of eGFP interfering with ‘normal’ GHRH neurone cell physiology. For example, transcription of the eGFP transgene could reduce endogenous GHRH levels in competition for transcription factors. Another factor is a hypothetical competition between the eGFP product and endogenous GHRH for granule packaging.

5.1.2 rGHRH-eGFP transgenic mice are phenotypically normal The reproductive axis of rGHRH-eGFP transgenic mice seemed unaffected, since mice bred easily and litter sizes were normal. Analysis of growth curves of rGHRH-eGFP transgenic animals from line no.39 compared to their WT littermates showed no significant difference in body weight at any age, neither in male nor in female transgenic mice (Table 5.1). Furthermore, no significant difference in pituitary GH content was found in rGHRH-eGFP transgenic

97 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

mice compared to adult male WT littermates (Table 5.1). The GH axis therefore appeared unaffected by eGFP expression in GHRH neurones.

GH (fig/pit) Body Weight (123-130 d)

Male Male Female

WT 37.1±5.6 35.1 ±2.2 26.1 ±0.9 (n) (5) (7) (7)

rGHRH-eGFP 41.95±5.2 35.1±1.7 25.1 ±0.8 (n) (5) (9) (5)

Table 5.1 Body weights and pituitary GH content of rGHRH-eGFP transgenic mice

GH contents were assayed in pituitary homogenates from adult male littermates. Body weights, recorded weekly from litters of mice showed indistinguishable growth curves; weights at 123-130 days are shown. Data are mean±SEM.

To analyse whether the eGFP transgene interfered with endogenous mGHRH transcription, hypothalamic GHRH mRNA content was measured by RNAse protection analysis (RPA). The RNA probe used for the RPA spanned several exon/intron boundaries of the mouse GHRH mRNA, therefore avoiding hybridisation of the probe to remaining DNA sequences in the RNA extracts. Endogenous hypothalamic GHRH expression levels are low, such that for a strong enough signal in the RPA two hypothalami were pooled into one extract. In total, 8 rGHRH-eGFP (line no.39) and WT animals were analysed in 4 RNA extracts per genotype. The RNA content was controlled by analysing p-actin mRNA expression in the same samples and assay. The expression of eGFP in GHRH neurones had no apparent effect on endogenous GHRH mRNA levels (WT: 2.45±0.50 normalised arbitrary units; rGHRH-eGFP: 2.61±0.40 normalised arbitrary units; n=4, n.s.). Figure 5.5 (page 105) shows a representative sample. Having established that the GH axis seemed unaffected, I attempted to visualise eGFP secretion from nerve terminals in response to GHSs. Hypothalamic slice preparations from rGHRH-eGFP transgenic mouse line no.39 were incubated in oxygenated aCSF

98 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

and visualised by fluorescent confocal microscopy. The flowthrough of aCSF was stopped for the time of the experiment and Ipg of GHRP-6 was added to the bath. Confocal images of the same plain and area of the slice were taken at 5 minutes and 7 minutes after the addition of GHRP-6. Some of the eGFP vesicles disappeared between 5 and 7 minutes, while new areas of eGFP fluorescence became apparent at 7 minutes (Figure 5.1). These data suggested that eGFP might be secreted in response to GHRP-6. Whilst promising, this was only a preliminary experiment in collaboration with N. Emptage (NIMR) and further studies will be needed to see if eGFP can be used as a means of following GHRH release.

5 minutes after l^g GHRP-6 7 minutes after 1pg GHRP-6

disappearing granule

appearing granule

Figure 5.1 Confocal microscopy of the ME in rGHRH-eGFP transgenic mice after the addition of Ipg GHRP-6 The ME of a rGHRH-eGFP (line no.39) live, hypothalamic slice preparation was analysed by confocal microscopy. Both panels show the same plain and area 5 and 7 minutes after the addition of 1|ig GHRP-6. White arrows indicate eGFP vesicles that were disappearing, while blue arrows show newly appearing eGFP fluorescence between 5 and 7 minutes, scale bar = 10pm

5.1.3 Discussion eGFP expression in rGHRH-eGFP transgenic mice had no discernible effects on the GH or reproductive axis. Their growth, pituitary GH content and endogenous hypothalamic mGHRH expression levels were unaltered. This suggests that packaging of endogenous

99 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

GHRH into granules and its secretion from the ME was not inhibited by the presence of a further eGFP product. Fluorescent confocal microscopy revealed that eGFP distribution seemed granular and that eGFP was possibly secreted from the MF in response to GHS treatment. As described previously, Ibata et al. were the first to report dense granules in the cell bodies and the terminal field of GHRH neurones (Ibata et al, 1986). Recently Nillni et al. described GHRH secretory granules near the nucleus and in the axonal processes of hypothalamic cultures (Nillni et al, 1999). However to date studies on the actual GHRH release from the MF were either performed in vivo by cannulating the portal vein in rats, sheep and pigs (Drisko et al, 1999; Frohman et al, 1990; Plotsky and Vale, 1985), or in vitro in hypothalamic or MF expiants (Aguila et al, 1991; Korbonits et al, 1999). Little is known about the packaging or secretory processes of GHRH. If co-localisation of eGFP and GHRH in the same granule could be shown by immunogold electron microscopy, eGFP could be used to study GHRH’s secretory processes from live nerve terminals. A direct comparison of GHRH and eGFP peptide content of GHRH neurones could reveal whether the packaging, storage and secretion mechanism is as efficient for the eGFP peptide as for endogenous GHRH. Further experiments are necessary to confirm the preliminary confocal microscopy data, showing eGFP granule movement in the MF in response to GHRP-6. Not only will studies of the GHRH terminal field benefit from the rGHRH-eGFP transgenic mice, but they will also facilitate electrophysiological experiments of pre­ identified GHRH cell bodies in slice preparations enormously. Further interesting applications of these mice in combination with other transgenic mice will be discussed in Chapter 10.

5.2 Studies of rGHRH-hGHS-R transgenic mouse physiology 5.2.1 Introduction In contrast to rGHRH-eGFP transgenic mice rGHRH-hGHS-R transgenic mice over­ express the GHS-R, a functionally active molecule that should affect GH physiology. Previously (Chapter 3) I have shown that over-expression of this receptor caused an increase in basal PI turnover in a GH tumour cell line. Overexpression of the GHS-R in

100 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice transgenic mice might therefore cause alterations in the GHS axis if this over-expression impacts on the normal physiology of GHRH neurones. There are no reports of transgenic mice or rats targeting GHS-R expression alterations. Bailey et al. recently presented data on a novel GHS-R knock-out mouse, showing residual Fos responses to a highly active class of non-peptidylic GHS (Bailey, 1999b). Their argument was that these GHSs activate a second type of GHS-R, still present in the GHS-R KG mouse. However, it is important to note that no other data are available on these GHS-R KG mice, nor was the total absence of the mouse GHS-R expression documented. GH release (Fairhall et al, 1995) and Fos responses (Dickson et al, 1995a) are sensitive to lower doses of GHSs when administered centrally rather than peripherally and higher doses of GHSs administered via the i.c.v. route can inhibit GH release (Yagi et al, 1996); the effects of GHS-R overexpression in hypothalamic GHRH neurones on GHS sensitivity were difficult to predict.

5.2.2 rGHRH-hGHS-R transgenic mice show altered GH physiology The reproductive axis of rGHRH-hGHS-R transgenic mice seemed unaffected, with no effect of the transgene on breeding or littersizes. Unlike with rGHRH-eGFP transgenic mice, analysis of growth curves for rGHRH- hGHS-R transgenic mice showed a significant, though slight body weight increase in adult male and female rGHRH-hGHS-R transgenic mice compared to their WT littermates (figure 5.2). Due to the mildness of the growth phenotype and the high variability of body weights in mice of the same age, the body weight difference between rGHRH-hGHSR transgenic and WT mice was usually only significant when a large number of animals was analysed. Within a single litter, the body weight difference was not reliably significant. Female rGHRH-hGHS-R transgenic mice displayed the body weight increase from a young age, while male transgenic mice did not show body weight differences until they reached adulthood. Analysis of the effect of hypothalamic GHS-R overexpression on the GH axis in rGHRH-hGHS-R transgenic mice was carried out by measuring the pituitary GH content. Male rGHRH-hGHS-R transgenic mice were found to have significantly elevated GH levels compared to WT littermates (WT=93.2±5.5; T=125.1±9|ig GH/pituitary; P< 0.01; n=12-14), while female transgenic mice did not show

101 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

Body Weight (32-34d) Body Weight (77-81 d)

Male Female Male Female

WT 19.2±0.9 16.1±0.3 29.7±0.5 22.4±0.4 (n) (10) (15) (8) (21)

rGHRH-hGHS-R 21.0±.07 17.5±0.4* 33.1 ±0.6** 23.8±0.3 (n) (16) (13) (18) (24)

40 -, ? **

^ 30 - O) n.s. O) % I 20 - >* E 10 -1 %

0 J WT T WT T WT T WT T 33 d 81 d 33 d 81 d

(n=10-16) (n=8-18) (n=13-15) (n=21-24)

Figure 5.2 Body weight of rGHRH-hGHS-R transgenic and WT mice Animals from several litters were weighed weakly and their weight shown before and after adulthood. While female transgenic mice showed enhanced growth from a young age, male transgenic mice were not heavier than WT mice until adulthood {Mann- Whitney Test, *P<0.05, **P<0.01).

102 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

** 150 1

« 100 - n.s. t E 50 -

0 - rGHRH- rGHRH- WT hGHS-R WT hGHS-R ? (n=12-14) (n=13) Figure 5.3 Pituitary GH content in adult rGHRH-hGHS-R transgenic mice and WT littermates GH content was measured in pituitary homogenates of adult male and female rGHRH- hGHS-R transgenic mice and their WT littermates. While male transgenic mice showed a significant enhancement of pituitary GH content (n=12-14, **P<0.01), no such difference was seen in female mice (n=13). significantly different GH levels, although there was a trend towards higher values (WT=55.6±7.7; T=62.1±7.5p.g GH/pituitary; n.s.; n=13) (Figure 5.3).

Pituitary Prolactin levels were indistinguishable in adult male and female rGHRH- hGHS-R transgenic mice compared to WT littermates (male: WT=1.74±0.10, T=1.67±0.10pg/pit, n.s. (n=20-23); female: WT=8.18±0.47, T=7.65±0.28|ag/pit, n.s. (n=8)). These data suggest that GHS-R overexpression in GHRH neurones has an effect on the pituitary specific for GH. One of the factors influencing pituitary GH content is hypothalamic GHRH, as GHRH was shown to enhance pituitary GH gene expression (Barinaga et al, 1985a; Barinaga et al, 1985b) and synthesis (Bilezikjian and Vale, 1984; Fukata et al, 1985), as well as somatotroph proliferation (Billestrup a/., 1986). The increase in GH content and body weight thus suggested a hypothalamic trophic effect on the pituitary. Therefore the

103 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice next step was to establish, whether transgene expression in GHRH neurones of rGHRH- hGHS-R transgenic mice had any effect on endogenous GHRH expression. In situ hybridisation with a mouse GHRH riboprobe on hypothalamic ARC sections of rGHRH- hGHS-R and WT littermates (6 animals per genotype, 6 arcuate sections per animal) showed a trend towards upregulated mouse GHRH mRNA expression levels (WT = 673.0±88.2 int. dens; rGHRH-hGHS-R = 1116.4±234.9 int. dens; Mann-Whitney, n.s.), but this was not significant, perhaps due to high background variability (figure 5.4).

WT rGHRH-hGHS-R

Figure 5.4 In situ hybridisation of mouse GHRH mRNA levels in rGHRH- hGHS-R transgenic and WT mice Comparison of mGHRH mRNA expression levels in rGHRH-hGHS-R transgenic and WT mice suggested a trend towards enhanced GHRH levels in rGHRH-hGHS-R transgenic mice, but this did not reach significance. Scale bar=1mm, ZI=zona incerta

In order to quantify mGHRH mRNA levels in WT and hGHRH-hGHS-R transgenic mice more accurately, RNAse protection analysis using a mouse GHRH mRNA probe was performed on hypothalamic RNA extracts as described for rGHRH-eGFP mice. In a first experiment an up-regulation of mouse GHRH mRNA expression in male transgenic mice was clearly visible (figure 5.5). Quantification, using (3-actin as a loading control, revealed mouse GHRH to be upregulated 2.5 times in transgenic mice. These are results from three individual measurements of the same hypothalamic RNA extract. Bearing in mind, the hypothalamic RNA extract consisted of only two pooled hypothalamic blocks from transgenic mice, I attempted to confirm these data with a larger number of mice. Therefore two hypothalamic blocks from female mice were pooled per RNA extract, such that in total 8 rGHRH-hGHS-R and WT animals were analysed in 4 RNA extracts per genotype. The RNA content was controlled by analysing p-actin mRNA expression in the same samples and assay. Figure 5.5 shows a representative sample. Analysis of all

104 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice protected fragments showed, that GHRH mRNA was up-regulated in rGHRH-hGHS-R transgenic mice (WT: 2.45±0.50 normalised arbitrary units; rGHRH-hGHS-R: 4.21 ±0.30 normalised arbitrary units, n=4, P<0.05) rGHRH- rGHRH- WT hGHS-R eGFP

mGHRH

p-actin

Figure 5.5 Mouse GHRH RNAse protection assay of hypothalamic extracts from WT, rGHRH-eGFP and rGHRH-hGHS-R transgenic mice 10pg hypothalamic RNA extract were hybridised with a mGHRH RNA probe and digested with RNAse afterwards. The top panel shows the protected mGHRH fragment, while the bottom panel shows the p-actin content of the hypothalamic extracts. mGHRH expression levels seemed up-regulated in rGHRH-hGHS-R transgenic mice,

However when D.F. Carmignac (NIMR) examined this issue in an in solution hybridisation assay, using a mGHRH cDNA probe, no significant differences in hypothalamic GHRH content were found, neither in male nor female rGHRH-hGHS-R transgenic and WT mice. Therefore the GHRH expression levels in rGHRH-hGHS-R transgenic mice compared to WT GHRH expression levels remain uncertain until larger numbers of mice are analysed. PEN SS mRNA expression levels were also analysed by in situ hybridisation and no significant difference was found between rGHRH-hGHS-R transgenic mice and their WT littermates (data not shown). Although the enhanced pituitary GH content in rGHRH-hGHS-R transgenic mice suggested that the enhanced body weight might result from enhanced growth, I aimed to analyse whether the increased body weight could stem from a metabolic effect with increased fat weight. Therefore a group of 7 transgenic and WT adult female mice was analysed for body weight, nose-anus length, ovarian fat pad weight and tibia length (table 5.2). Although the body weight of transgenic mice was increased, in this group the

105 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

weight difference did not reach significance. Despite this the rGHRH-hGHS-R transgenic mice showed significantly enhanced nose-anus and tibia length, while no difference in ovarian fat pads was measured (table 5.2). rGHRH-hGHS-R transgenic mice did show a trend towards higher fat pad weights, but both genotypes showed a widely varying range of fat pad weights, which made the interpretation of these data difficult.

WT rGHRH-hGHS-R (n=7) (n=7)

body weight [g] 23.4±0.8 26.3±1.3

nose-anus length [mm] 105±1 109±1 *** tibia length [mm] 18.91 ±0.08 19.49±0.06

ovarian fat pad weight [mg] 166±56 221±76

Table 5.2 Growth parameters in adult female rGHRH-hGHS-R transgenic and WT mice Although the mean weights of this group of female rGHRH-hGHS-R transgenic were higher, the difference was not significant. Their nose-anus and tibia length were significantly increased, whereas fat pad weights were indistinguishable (n=7, *P<0.05, ***P<0.001).

NPY mRNA expression levels were measured in transgenic mice and WT littermates by using a rat NPY cDNA riboprobe. Rat and mouse NPY mRNAs are very homologous and I found that it was not necessary to use a mouse specific NPY probe. Neither ARC nor cortex NPY expression levels were significantly different between rGHRH-hGHS-R and WT mice (ARC: WT=9571±2835; T=10557±1196 int.dens, CORT: WT=5049±1132; T=5553±979 int.dens, (n=7), n.s., Mann-Whitney Test).

I have shown that hypothalamic GHS-R over-expression in GHRH neurones caused enhanced pituitary GH levels and body weight, however, an important question I wanted to address was whether the over-expression inferred enhanced responses to GHSs.

106 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

Therefore I collaborated with D.F. Carmignac (NIMR), who had established a method for jugular vein cannulations and sampling in mice. Several groups of adult (10-12 weeks) rGHRH-hGHS-R transgenic and WT mice were anaesthetised and one of the jugular veins cannulated and their GH responsiveness to a high dose of GHRP-6 (500ng, determined earlier to be the maximal dose) was compared. Results from two separate experiments were combined (Figure 5.6). GH responses to 500ng of GHRP-6 were significantly elevated in rGHRH-hGHS-R transgenic mice compared to WT littermates (WT=65.0±20.3, T=134.1±24.0ng GH/ml plasma, P<0.05, n=12). Male transgenic mice also had significantly elevated basal plasma GH levels, but this was variable (WT=0.1±0.1, T=5.8±1.7ng/ml plasma, P<0.01, n=12). Basal plasma GH levels are at the RIA detection limit and therefore only interpretable with great care. Next, the sensitivity of such animals to lower doses of GHRP-6 was analysed. Two doses (50ng and 250ng of GHRP-6) were given to another group of animals consecutively after a 120 minute washout time. Again data from two separate experiments were pooled. Importantly transgenic animals were not significantly more sensitive to 50ng of GHRP-6 than WT mice (WT=8.4±1.7; T=11.2±1.9ng/ml plasma, n.s., n=ll-13), but when 250ng of GHRP-6 were given, rGHRH-hGHS-R transgenic mice responded with a significantly higher increase in plasma GH levels than WT animals (WT=21.5±3.1; T=36.0±5.3ng/ml plasma, P<0.05, n=ll-13). It would seem from these data that rGHRH-hGHS-R transgenic mice do not show increased sensitivity to GHRP-6, but do have greater GH responses to medium and high doses of GHRP-6 than WT mice. Preliminary data in female rGHRH-hGHS-R transgenic mice showed that there was a trend towards enhanced GH responses to SOOng of GHRP-6, although this did not reach significance due to a large variability of GH responses to GHRP-6 in transgenic animals (WT=77.2±11.3, T=188.6±86.5ng/ml plasma ; n.s., n=4-5, Mann-Whitney Test). In order to test whether a significantly enhanced GH response to GHRP-6 can be found in female transgenic mice, a larger number of animals would need to be analysed. Generally, we experienced quite large differences in GH responses to GHRP-6 between animals of the same genotype and experiments. The technique is thus not ideal for detecting small differences in GH responses and needs repeated experiments with large numbers of animals, to draw definite conclusions. Thus the presented data would also

107 Chapter 5; Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

50

150 - GHRP-6 40 - (0 500ng E GHRP-6 GHRP-6 % 30 - a. 100 - 50ng 250ng

I I - 20 Ü O) c 50 1 10 - **

0 -J 0 - WT I WT T WT T WT T WT T time [min] 0

Figure 5.6 In vivo GH responses to GHRP-6 in male rGHRH-hGHS-R transgenic mice and WT littermates rGHRH-hGHS-R and WT mice were anaesthetised and their jugular vein cannulated to inject GHSs and withdraw blood samples. Plasma GH levels were analysed by RIA. Left panel: rGHRH-hGHS-R transgenic mice showed enhanced basal levels of plasma GH and GH responses to SOOng of GHRP-6 (n=12). Right panel: GH responses to low doses of GHRP-6 were indistinguishable, while transgenic mice respond to 250ng of GHRP-6 with significantly increased plasma GH levels (n=11-13); **P<0.01; *P<0.05. Please note different scales in the left and right panel.

ideally need confirmation. One could argue that the enhanced GH response to GHRP-6 was simply due to the enhanced pituitary GH content in transgenic rGHRH-hGHS-R transgenic mice, rather than being caused by enhanced responsiveness to GHSs. If this was true, rGHRH- hGHS-R transgenic mice should also show enhanced responses to GHRH. To test this hypothesis, D.F.Carmignac (NIMR) analysed rGHRH-hGHS-R transgenic mice for their GH responses to several doses of GHRH. In general, responses to GHRH were more variable than responses to GHRP-6. The same 12 adult male animals from the first GHRP-6 experiment (Figure 5.6 left panel) were given lOOng of GHRH after a 120 minutes washout period. No significant difference in the GH response to GHRH was found between rGHRH-hGHS-R transgenic and WT mice (WT=261 ±75.1 ng/ml plasma; T=187.3±50.9ng/ml plasma; n.s., n=12). Also we found no significant differences in GH

1 0 8 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

responses to lower doses of GHRH between male rGHRH-hGHS-R and WT mice (Figure 5.7). Enhanced GH responses to GHRP-6 (SOOng) were thus not solely due to an increased pituitary GH content.

400 - 1 2 0 - GHRH GHRH lOng 50ng

1 0 0 - GHRH (0 300 - lOOng E 1 80 - 1 I I Q. 1 200 - 1 60 - X i e> o> 40 - c 100 -

20 - ** ** I 1 0 J 0 - WT I WT WT T WT T WT T WT T

time [min] 0 90 90

Figure 5.7 In vivo GH responses to GHRH in male rGHRH-hGHS-R transgenic and WT mice Left panel: rGHRH-hGHS-R transgenic mice showed similar GH responses to lOOng GHRH; (n=12) Right panel: GH responses to lower doses of GHRH (lOng and 50ng) were also indistinguishable, but basal GH plasma levels were elevated at 90 minutes (n=6-8); **P<0.01. Note different scales in left and right panel.

5.2.3 Discussion Over-expression of the human type la GHS-R in hypothalamic GHRH neurones was shown to alter several elements of the GH axis. Male and female rGHRH-hGHS-R transgenic mice showed a subtle phenotype with increased pituitary GH content and body weight in male transgenic mice, while female rGHRH-hGHS-R mice displayed a similar phenotype with increased body weight and tibia and nose-anus length. In male rGHRH-hGHS-R transgenic mice NPY expression was unaltered and female transgenic mice showed slightly enhanced, but not significantly different fat pad weights. Interpretation of the fat pad data is somewhat difficult due to large variances, but the other data are consistent with a true growth phenotype in rGHRH-hGHS-R transgenic mice, rather than a metabolic effect. Nevertheless, to test whether rGHRH-hGHS-R

109 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice transgenic mice might show increased body weight due to enhanced feeding, their food intake could be measured using mouse metabolic cages. Neither male nor female rGHRH-hGHS-R transgenic showed altered pituitary Prl levels, suggesting that the effect of hypothalamic hGHS-R over-expression on the pituitary is specific for GH, consistent with a possible effect via GHRH. GH content was only significantly upregulated in male rGHRH-hGHS-R transgenic mice; females had only slightly increased pituitary GH levels, but still showed the effect of enhanced growth. However, measurements of pituitary GH content only reflect a single point in time of the varying GH peptide content due to a balance between synthesis and release. An explanation for the increased body weight in spite of only slightly enhanced pituitary GH content might lie in the GH secretory pattern of female rGHRH-hGHS-R transgenic and WT mice. Technical difficulties with serial sampling from such small animals, with only limited amounts of blood make the analysis of circulating GH profiles difficult. Furthermore, sex differences in GHRH mRNA have been demonstrated, with males expressing greater levels than females (Argente et al, 1991). As the hGHS-R trangene in these animals is driven by the rat GHRH promoter, it is possible that transgene expression is lower in female rGHRH-hGHS-R transgenic mice. However, it was shown in Chapter 4 that male and female rGHRH-eGFP transgenic mice do not display a sexual dimorphism in hypothalamic eGFP protein levels. Although female rGHRH-hGHS-R transgenic mice do not show a significant increase in GH content, their growth is enhanced and they do show a trend towards enhanced GH responses to GHSs. I suggest that there might not be a significant phenotypical sexual dimorphism due to the transgene between male and female rGHRH- hGHS-R transgenic mice. In vitro studies in GHS-R transfected GC cells had shown that over-expression of the GHS-R caused enhanced PI turnover in absence of a ligand. As explained in Chapter 3, over-expression of G-protein coupled receptors is thought to cause enhanced basal activity because with the over-expression the total number of receptors in the active state in absence of a ligand has risen. If a similar mechanism occurred in the hypothalamus of rGHRH-hGHS-R transgenic mice, the enhanced pituitary GH content in male transgenic mice could be a result of an enhanced trophic effect from increased active GHRH neurones. Measurement of hypothalamic GHRH mRNA expression by several

110 Chapter 5; Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice techniques gave variable results. While in situ hybridisation and RNAse protection assay suggested enhanced GHRH mRNA levels in rGHRH-hGHS-R transgenic mice, solution hybridisation did not confirm this. Although in RNAse protection assays it is possible to control for specific hybridisation of the probe better than in solution hybridisation assays, both assays have the same problem; if dissection of the hypothalamic blocks is different between animals, the GHRH content per animal is artificially more concentrated or diluted. A better way to analyse GHRH mRNA expression might have been to perform a GHRH in situ hybridisation, dip it in photographic emulsion and count silver grains per labelled cell, to see whether GHRH mRNA content per cell is up- regulated in rGHRH-hGHS-R transgenic mice. GHS responsiveness to higher doses of GHRP-6 seemed increased in rGHRH-hGHS-R transgenic mice and a trend towards this effect was observed in female transgenic mice. Though speculative, a possible mechanism might be that increased amounts of hypothalamic binding sites for GHRP-6 cause increased GHS-R signalling in the GHRH neurones in response to which higher levels of GHRH might be released. GHRH together with GHRP-6 at the pituitary level might then cause up-regulated GH release. GHS-R over-expression did not alter the sensitivity to GHRP-6, since both transgenic and WT mice responded equivalently to a low dose of GHRP-6. However, as stated before, GH responses to GHRP-6 and GHRH were variable between animals of the same genotype and between experiments, so interpretation of the mean data from such heterogeneous groups might not be ideal. Currently our data suggest that over­ expression of the hGHS-R targeted to GHRH neurones enhanced GH responses to higher doses of GHSs, but this needs confirmation. Interestingly the rGHRH-hGHS-R transgenic animals gave more consistent GH responses to GHSs than the WT littermates. On the other hand, analysis of GH responses to GHSs by measuring plasma GH levels is fairly remote from the primary effects GHSs might have on GHRH neurones. There are many steps between GHS signalling via hypothalamic GHS-R and finally detecting GH responses in the plasma, such that the direct effect of GHS-R over-expression in the hypothalamus might not be obvious. To analyse whether GHS-R over-expression changes these first events of GHS on hypothalamic neurones, measuring a more direct effect at the hypothalamic level is probably more desirable. For example, electrophysiological studies could be used to measure whether the electrical activation of

111 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice

GHRH cells by GHSs is altered in rGHRH-hGHS-R transgenic mice. Bennett et al. had previously reported that continuous s.c. GHRP-6 treatment for 2 weeks (an experimental paradigm known to cause marked desensitisation to GHRP-6) increased ARC GHS-R expression levels in rats and suggested that GHRP-6 desensitisation might be associated with increased GHS-R levels (Bennett et al, 1998). In this study continuously up-regulated GHS-R levels in ARC GHRH neurones however did not cause desensitisation to an acute GHRP-6 injection. A recent study by Kineman et ah demonstrated that a four hour i.v. infusion of GHRH in normal rats resulted in a 10-fold increase of circulating GH levels associated with a 2- fold increase of pituitary GHS-R mRNA (as measured by semi-quantitative RT-PCR) (Kineman et al, 1999). This effect was only seen in vivo, not with in vitro primary rat pituitary cultures, suggesting the involvement of additional central or systemic factors in GHS-R mRNA regulation. If GHRH mRNA levels proved to be enhanced in rGHRH- hGHS-R transgenic mice this might cause endogenous pituitary mouse GHS-R levels to increase. Endogenous mouse GHS-R levels in the pituitary were not determined quantitatively in rGHRH-hGHS-R or WT mice. Pituitary GHS-R mRNA levels are extremely low and hardly detectable by P^S] in situ hybridisation (Bennett et al, 1997). Indeed, a tissue screen in situ hybridisation with a hGHS-R riboprobe (described in Chapter 4) showed no GHS-R hybridisation signal in the pituitary of rGHRH-hGHS-R transgenic or WT mice. If the endogenous GHS-R was up-regulated in transgenic mice, this was not to a degree where GHS-R mRNA levels become measurable by in situ hybridisation. I can however not formally exclude the possibility of small amounts of up-regulated endogenous pituitary GHS-R expression causing the increased responsiveness to GHS. We also analysed the possibility that the increased GH release from the pituitary in response to GHRP-6 might be due to the elevated GH content, thus making more GH available for release. As rGHRH-hGHS-R transgenic mice did not show altered responsiveness to GHRH at any of the tested concentrations, I conclude that the enhanced GH release in response to GHRP-6 is a specific effect for GHSs. Another factor to bear in mind is Ghrelin, a recently cloned ligand for the GHS-R (Kojima et al, 1999). Chapter 8 describes later studies on Ghrelin expression in my rGHRH-hGHS-R transgenic mice, but at this stage I could not exclude the possibility

112 Chapter 5: Physiological studies of rGHRH-eGFP and rGHRH-hGHS-R transgenic mice that altered GH physiology and GHS responsiveness might stem from altered Ghrelin expression and signalling. In summary, I have shown that over-expression of the hGHS-R in hypothalamic GHRH neurones altered GH physiology in rGHRH-hGHS-R transgenic mice. As I also showed previously in vitro, even in the absence of exogenous GHSs, alterations to the GH axis were demonstrated in vivo in rGHRH-hGHS-R transgenic mice. The phenotype of rGHRH-hGHS-R transgenic mice is rather subtle, but hypothalamic hGHS-R over­ expression seemed to specifically affect the GH axis. This over-expression of the GHS- R also seemed to increase GH release from the pituitary in response to exogenous ligand. One of my aims was to investigate GHRH neurones as a direct target for GHSs. Here I have demonstrated that GHS-R over-expression restricted to GHRH neurones indeed alters GH and GHS physiology, implying GHRH as an important direct target for GHS to mediate their GH releasing effect.

113 Chapter 6: Targeting transgenes to pituitary GH cells

6 Targeting transgenes to pituitary GH cells This chapter describes the generation of two lines of transgenic mice both expressing transgenes under human GH promoter control. The first part of the chapter describes the expression of a fluorescent reporter in pituitary GH cells and in the second part enhanced GHS-R abundance in these cells is demonstrated. 6.1 hGH-eGFP transgenic mice 6.1.1 Introduction As described earlier, anterior pituitary somatotrophs are the main endocrine cell type expressing GH and are responsible for the co-ordination of GH gene expression, synthesis, storage and release. GH secretion from the anterior pituitary is pulsatile in all species tested so far, including rats and humans (Finkelstein et al, 1972; Tannenbaum and Martin, 1976). For pulsatile release of large peaks of GH, a co-ordination and communication of somatotrophs is feasible, but it is unclear whether such a mechanism exists and how it could be activated. The study of GH secretory events in vivo in whole pituitaries would require a means of distinguishing cell types involved and the hormones secreted. To identify living GH cells we have chosen to mark GH cells specifically using eGFP expressed under GH promoter control. In 1995 Jones et al. generated transgenic mice bearing several hGH constructs with differing 5’UTR lengths (Jones et al, 1995). Of these the K2B cosmid contained 40kb of contiguous 5’ flanking regions and the coding region for hGH-N, the pituitary- specific exon of the hGH gene cluster. Only transgenic mice bearing this construct gave copy-number-dependent, position-independent and somatotroph-specific expression of hGH (Bennani-Baiti et al, 1998; Jones et ai, 1995; Su et al, 2000). Shorter constructs with less 5’UTR sequences failed to give tissue-specific hGH expression, because they lacked some of the DNAse hypersensitive sites necessary for a physiologically regulated chromatin domain for hGH transcription (Bennani-Baiti et al, 1998; Jones et a l, 1995). N. Cooke (University of Pennsylvania) kindly donated the K2B cosmid which comprised the described hGH LCR.

114 Chapter 6: Targeting transgenes to pituitary GH cells

6.1.2 The HGH-eGFP cosmid construct C. Magoulas (NIMR) engineered the transgene construct, targeting eGFP to hGH cells. The generation of this construct is thus not part of this thesis, but some elements of the hGH-eGFP transgene construction are relevant to the generation of my hGH-hGHS-R construct later and will be discussed here briefly. As described for the rGHRH-eGFP construct eGFP expression was targeted to GH granules in order to study GH trafficking and secretion in vivo in whole pituitaries. A similar approach to that used to generate rGHRH-eGFP transgenic mice was taken and eGFP was again fused to the entire hGH signal peptide sequences. An Mlul restriction site was introduced into the 40kb K2B hGH cosmid 326bp upstream of the hGH ATG (Figure 6.1, right panel). The hGH gene sequences can then be excised as a single Mlul fragment, because of an additional Mlul site in the pWElS vector. This hGH Mlul fragment was subcloned into a modified Mlul tinkered pBluescript vector for further modification (Figure 6.1 right panel). The hGH/eGFP chimeric Mlul fragment was constructed as described in Figure 6.1 on the left panel and the Mlul fragment was finally re-introduced into the K2B cosmid to create the hGH-eGFP cosmid construct. I excised the 40kb Notl fragment, purified the DNA on a salt gradient and used it for microinjection.

6.1.3 Generating HGH-eGFP transgenic mice In collaboration with K. Mathers (Biological Services, NIMR), I microinjected the purified Notl fragment of the hGH-eGFP construct into fertilised mouse oocytes and viable eggs were transferred into the oviduct of pseudopregnant recipients. The resulting pups were analysed by PCR, as described for genotyping of rGHRH-eGFP transgenic mice. The primers hGH 5’UTR and exon2R (table 2.1) amplified an endogenous 300bp mGH and a 350bp transgenic hGH fragment (Figure 6.2). Of 25 pups, 3 were positive for the transgene and bred with [(CBa/CaxC57Bl/10)Fl] mice. The progeny of these breeders was -50% transgenic.

115 9 1os EcoRi EcoRI Ic. *Mlul Mlul CTO s GH (K2B) cosm id h G H g pW15 1 o "3. EcoRI EcoRI ^ subcloning of Mlul 6. Mlul fragment into Mlul Mlul 1 h G H -S P hG H -3'U T R I llnkered pBKS O |hG H | |p W 1 5 os A pal

\ Mlul

GH| 1 eGFP

Apa

Not Not I Mlul

hGH-5'UTR

Figure 6.1 Construction of the hGH-eGFP transgene Essentially eGFP was placed under the promoter control of hGH. Right panel: An MlulsWe was introduced upstream of the hGH ATG and the resulting hGH-PW15 M/u/fragment subcloned into a modified Mlul linkered pBluescript vector. Left panel. An Xbal-Notl eGFP fragment was introduced into an EcoRI hGH subclone and the Apa/-A/of/fragment used to replace the subcloned hGH sequences in the /W/u/linkered pBluescript. The resulting A/f/u/fragment was finally introduced into the K2B cosmid to create the hGH-eGFP cosmid construct. Chapter 6; Targeting transgenes to pituitary GH cells

Notl Notl

hGH-5'UTR eGFP

5'UTR ex2R

1 WT genomic DNA 350bp 2 hGH-eGFP transgenic founder 300bp 3 WT pup 4 transgenic llttermate

Figure 6.2 Analysis of transgenic hGH-eGFP mice by PCR Genotyping of hGH-eGFP transgenic mice was by PCR with primers hGH 5’UTR and hGH ex2R (top panel). PCR resulted in a 350bp transgenic hGH and 300bp endogenous mGH product (bottom panel).

6.1.4 Transgene expression in hGH-eGFP transgenic mice Initially I examined hGH-eGFP adult male mice for eGFP fluorescence in anterior pituitaries. 12|im cryostat sections were analysed by fluorescent confocal microscopy and a major population of the anterior pituitary cells showed strong, punctate fluorescence, while there was no eGFP expression in the posterior pituitary (Figure 6.3). Further expression and physiological studies on hGH-eGFP mice were not part of my thesis, but performed by C.Magoulas, L. McGuinness, D. Carmignac (NIMR, London) and H. Christian (Oxford) and are in press (Magoulas et ai, 2000). To summarise briefly. Northern analysis and a RIA for GFP showed that transgene expression was confined to hGH-eGFP transgenic mouse pituitary cells. Co-localisation of eGFP and GH proteins in somatotroph cells was demonstrated by double immunocytochemistry. The fluorescent cells could be separated by fluorescent activated cell sorting and were shown to have a granular appearance of fluorescence by confocal microscopy. Immunogold electronmicroscopy demonstrated that eGFP protein was packaged into secretory granules. GFP is secreted from these granules in parallel with GH in response to GHRH in vitro as was shown by RIA of medium supernatants.

117 Chapter 6: Targeting transgenes to pituitary GH cells

Figure 6.3 Fluorescent confocal microscopy of an adult male hGH-eGFP transgenic mouse pituitary A pituitary from an adult male hGH-eGFP transgenic mouse was frozen and cut into 12pm slices on a cryostat. Strong eGFP fluorescence was observed by fluorescent confocal microscopy in cells of the anterior pituitary (AP) of the hGH-eGFP transgenic mouse. No eGFP expression was seen in posterior pituitary cells (PP). scale bar = 10pm

6.1.5 Discussion Transgenic mice expressing eGFP under GH promoter control were generated by fusing eGFP to the hGH signal peptide and inserting this chimeric fragment into a 40kb hGH cosmid. By fusing eGFP to the entire hGH signal peptide and the 22 first hGH N- terminal residues, eGFP was targeted to the large dense core granules in somatotrophs. The hGH LCR in the K2B cosmid further ensured high-level specific eGFP transgene expression restricted to pituitary GH cells. For my purposes this confirmed that the GH cosmid strategy could be used to target over-expression of the GHS-R to pituitary GH cells. These transgenic animals are very useful tools for analysing GH cell physiology by visualising events in identified cells and therefore enable the study of secretory processes. With FACS (fluorescent activated cell sorting) studies of primary hGH-eGFP transgenic mouse pituitary cultures somatotrophs can be isolated and further studied in vitro detached from the influence of other pituitary cells. Furthermore the labelling of these cells allows the study of co-ordinated [Ca^""]; transients in whole live pituitary

118 Chapter 6: Targeting transgenes to pituitary GH cells slices in response to exogenous secretagogues (Magoulas et al, 2000).

6.2 hGH-hGHS-R transgenic mice 6.2.1 Introduction As previously described one of the aims of this project was to investigate the relative importance of the hypothalamic and pituitary GHS actions, by enhancing GHS-R expression in a cell-specific manner. Chapter 4 and 5 described the generation and analysis of rGHRH-hGHS-R transgenic mice, where hGHS-R over-expression caused alterations in the normal hypothalamo-pituitary GH axis. The next aim was to analyse the effects of GHS-R over-expression in the pituitary GH cell and whether this would alter GH or GHS physiology. The successful targeting of the hGH and the eGFP transgene to mouse pituitary somatotrophs in transgenic mice had convinced me to use the hGH cosmid to target GHS-R expression to GH cells. I had previously demonstrated that transcription of the hGHS-R type la cDNA resulted in a functionally active receptor in vitro (Chapter 3). The hGH K2B cosmid and the hGHS-R type la cDNA were thus assembled to generate the hGH-hGHS-R construct.

5.2.2 The hGH-hGHS-R cosmid construct As described for the rGHRH-hGHS-R transgene, targeting of the hGHS-R to secretory vesicles is not useful when trying to establish physiological influences of ligands. If the receptor was packaged into dense core vesicles, no signalling of the receptor would occur. Therefore the hGH signal peptide was not included in the construct that was designed to target hGHS-R expression to pituitary GH cells. hGH 5’ and 3’ UTR sequences were fused directly to the hGHS-R type la cDNA without the hGH signal peptide. Subsequently a similar cloning strategy was used as described for the hGH- eGFP construct. Again the modified 40kb K2B cosmid with the Mlu\ restriction site 326bp upstream of the hGH ATG was used (Figure 6.4, right panel). The hGH gene sequences were then excised as a single Mlul fragment and cloned into a modified Mlul linkered pBluescript I had created for this purpose. As described in chapter 4.2.2,1 had already generated a chimeric hGH/hGHS-R Mlul

119 Chapter 6: Targeting transgenes to pituitary GH cells fragment, containing hGH 5’ and 3’UTR sequences for the rGHRH-hGHS-R construct, which I would be able to use now to introduce the hGHS-R sequences into the hGH cosmid. However, the Mlul restriction site in the hGH cosmid was 326bp upstream of the hGH ATG, which is further up-stream than the Mlul restriction site I had created 60bp up-stream of the hGH ATG site in the 5’UTR of the hGH/hGHS-R chimeric fragment used for the rGHRH-hGHS-R transgenic construct. Therefore the BamHI and SspI sites that previously had been Mlul linkered were reintroduced by PCR site-directed mutagenesis (primers 9F and lOR, table 2.1); (Figure 6.4, left panel). The resulting BamHI -SspI fragment was sequenced and no basepair mismatches were found. This BamHI-SspI hGH/hGHS-R chimeric fragment was cloned into the BamHI and SspI sites of the hGH Mlul fragment in pBluescript, to replace the hGH coding region by the hGHS-R (Figure 6.4). The resulting hGH/hGHS-R Mlul fragment was cloned back into the 40kb K2B hGH cosmid. A restriction map confirmed the accuracy of the Notl fragment to be used for microinjection.

6.2.3 Generating hGH-hGHS-R transgenic mice hGH-hGHS-R transgenic mice were generated in collaboration with K.Mathers (Biological Service, NIMR) by pronuclear microinjection of the purified Notl fragment of the hGH-hGHS-R cosmid construct into fertilised mouse oocytes and oviductal transfer into pseudopregnant mice. Of 73 resulting pups four founders were obtained, that incorporated the transgene as assessed by PCR and Southern Blot (Figure 6.5). For PCR the same 3 primer assay was used as described for rGHRH-hGHS-R transgenic mice, two hGH primers amplifying an endogenous mGH 300bp fragment and the third primer amplifying a chimeric hGH/hGHS-R 506bp fragment with the hGH 5’UTR primer. The progeny of three founders gave positive litters (B, C and D), while the fourth founder (E) only transferred the transgene once onto one pup. This pup was set up to breed instead and produced 50% transgenic litters, so founder E had probably been chimeric. The copy number of the transgenic founders was analysed by Southern Blot, using a hGHS-R probe. As mentioned previously the exact number of transgenes is then difficult to measure because the assay is biased towards the transgene. Preliminary Southerns suggested that founder B had two copies of the transgene, but quantification of the copy

120 Mlul *Mlul Mlul Mlul g

hGH ATG STOP hGH 1 hG H S-R GH (K2B) cosm id hGH g pW15 os T BamHI SspI I I PCR 3p. +BamH1/Sspl digest I OQ I 5 *Mlul subcloning of Mlul fragment Into Mlul Mlul linkered pBKS %

hGH ATG hG H S-R srop hGH | i G H | I g p W 1 5 o \ T3 BamHI SspI \

O \ X *Mlul \ Mlul \

hG HS-R \ BamHI SspI \ \ \ \ Not 'Mlul Notl Mlul

hGH hGHS-R

Figure 6.4 Construction of the hGH-hGHS-R transgene Essentially the hGHS-R type 1a cDNA was placed under hGH promoter control. Right panet. An Mlul site was introduced upstream of the hGH ATG and the resulting hGH-PW15 /W/u/fragment subcloned into a modified Mlul linkered pBluescript vector. Leftpanet. The original hGH BamHI ar\6 SspI sWes were reintroduced into the hGH/hGHS-R chimeric fragment and the BamH/-Ssp/fragment used to replace the subcloned hGH sequences in the Mlul linkered pBluescript. The resulting Mlul fragment was finally re-introduced into the K2B cosmid to create the hGH-hGHS-R cosmid construct. Chapter 6: Targeting transgenes to pituitary GH cells

numbers in other founders was difficult since the endogenous mGHS-R band was not comparably clear in all Southerns (figure 6.5). A: hGH-hGHS-R construct

Not I Notl 506bp 5'UTR 4R hGH-5'UTR hGHS-R I I Southern probe j | Bgii M ► Bgii 1.5 kb B: PCR

1 WT control 506bp -I 2 rGHRH-hGHS-R control 3 hGH-hGHS-R founder SOObp -I 4 WT pup 5 WT pup 6 hGH-hGHS-R pup

C: Southern Blot M 1 2 4 5 6 1 hGH-hGHS-R cosmid control 2 rGHRH-hGHS-R control 3 WT control 4.0kb 4 WT pup Ï 5 hGH-hGHS-R pup no.B 6 WT pup 7 hGH-hGHS-R pup no.C 1.5kb -► i

Figure 6.5 Analysis of hGH-hGHS-R transgenic mice A: Genotyping of rGHRH-hGHS-R transgenic mice was by PCR with primers hGH 5’UTR, hGH ex2R and hGHS-R 4R and by Southern Blot digesting with Bgll and probing with a hGHS-R probe. B: PCR resulted in a 506bp transgenic and 300bp endogenous mGH product. C: Southern blot analysis showed a 1.5kb hGHS-R transgenic band and a weaker 4.0kb endogenous mGHS-R hybridisation signal.

122 Chapter 6: Targeting transgenes to pituitary GH cells

6.2.4 Transgene expression in hGH-hGHS-R transgenic mice

Anterior pituitaries of adult male hGH-hGHS-R transgenic mice (founder B) were cut into 12|xm sections on a cryostat. In situ hybridisation with a rat GHS-R type la anti­ sense riboprobe, which measured endogenous mGHS-R and transgenic hGHS-R mRNA, showed a strong hybridisation signal confined to the anterior pituitary of transgenic mice (Figure 6.6). The hybridisation signal in anterior pituitaries of WT mice was only slightly elevated from the non-specific control signal of a sense riboprobe. Sections were dipped in photographic emulsion and examined at high magnification by light microscopy. This revealed that the strong hybridisation signal in transgenic mice was the result of single, heavily labelled anterior pituitary cells (Figure 6.6). No GHS-R signal was found in the posterior pituitary at this level of exposure. In WT pituitaries specific cell labelling was difficult to distinguish from background labelling. The level of hGHS- R over-expression was analysed densitometrically and found to be 50-fold in comparison to WT littermates (3 sections from two different mice per genotype).

For immunological analysis anterior pituitaries from adult male hGH-hGHS-R transgenic (founder B) and WT mice were fixed in 4% paraformaldehyde over night, paraffin embedded and cut to 6p.m sections. Immunocytochemistry (performed by B. Gevers, NIMR) with a human C terminal GHS-R anti-body showed a similar picture to that observed by in situ hybridisation. Single heavily labelled cells in hGH-hGHS-R transgenic mouse pituitaries were confined to the anterior pituitary (Figure 6.7). WT anterior pituitaries demonstrated only faint labelling above background.

6.2.5 Discussion In order to target hGHS-R expression to somatotroph cells in the anterior pituitary of transgenic mice a 40kb hGH cosmid, previously shown to comprise the full hGH LCR (Bennani-Baiti et al, 1998; Jones et al, 1995), was used. The hGHS-R type la cDNA was thus placed under hGH promoter control by directly replacing the hGH coding sequence with the hGHS-R coding sequence, thereby avoiding the hGH signal peptide and hGHS-R targeting to secretory granules.

123 Chapter 6; Targeting transgenes to pituitary GH cells

WT-antisense hGH-hGHS-R-antisense

Figure 6.6 Pituitary GHS-R expression in hGH-hGHS-R transgenic and WT mice In situ hybridisation of adult male hGH-hGHS-R transgenic (founder B) and WT mouse pituitaries with a rat GHS-R riboprobe showed that GHS-R over-expression was restricted to the anterior pituitary of hGH-hGHS-R transgenic mice (top panel). In the bottom panel high resolution imaging of dipped pituitary sections showed that GHS-R over-expression in hGH-hGHS-R mice was limited to specific anterior pituitary cells. Scale bar = 0.1 mm

Four transgenic founders were generated with the hGH-hGHS-R construct, of which founder B was found to have incorporated two copies. hGH-hGHS-R transgenic mice were the last of the animal models I generated, so I have had less time to analyse their expression pattern and phenotype. In situ hybridisation with a rGHS-R riboprobe showed strong GHS-R expression confined to the anterior pituitary of hGH-hGHS-R transgenic mice with a defined population of cells being specifically labelled. As reported previously (Bennett et al, 1997), the endogenous mouse GHS-R expression in the anterior pituitary was so low, that it was difficult to distinguish any labelling above background in WT mice.

124 Chapter 6: Targeting transgenes to pituitary GH cells

hGH-hGHS-R WT

hGH-hGHS-R no antibody

Figure 6.7 Immunocytochemistry for GHS-R protein in hGH-hGHS-R transgenic and WT mouse pituitaries Immunocytochemistry with a hGHS-R anti-body on adult male hGH-hGHS-R transgenic (founder B) and WT mouse paraffin embedded 6|im pituitaries showed strong staining in single hGH-hGHS-R anterior pituitary cells. In WT pituitaries staining was hardly above background. Scale bar = 0.1 mm

Similarly, the immunocytochemistry assay measuring GHS-R protein expression in pituitaries showed strong staining of a subpopulation of anterior pituitary cells in hGH- hGHS-R transgenic mice with only little staining above background in WT mice. The GHS-R transgene, driven by the hGH UTR, was thus appropriately expressed and translated in hGH-hGHS-R transgenic mice. The 50-fold over-expression of the hGHS- R compared to endogenous mouse GHS-R expression highlights the difference in the endogenous GHS-R and GH promoter driven expression levels, with pituitary GHS-R abundance being so much lower than GH levels. In future, transgenic animals of all four hGH-hGHS-R founders will need to be investigated, concerning their transgene copy-numbers and GHS-R pituitary expression

125 Chapter 6: Targeting transgenes to pituitary GH cells levels, to study whether in these animals the hGH LCR indeed achieved copy-number dependent and position-independent transgene expression. The pattern of hGHS-R expression confined to a subpopulation of anterior pituitary cells was reminiscent of the pattern of eGFP fluorescence I had observed previously in hGH- eGFP transgenic mice and suggested that the hGH LCR had successfully targeted hGHS-R expression to GH cells. Tissue- and cell-specificity of transgene expression driven off the hGH cosmid has been demonstrated previously for hGH-N (Bennani-Baiti et al, 1998; Jones et al, 1995) and hGH-eGFP transgenic mice (Magoulas et al, 2000), but should nevertheless be confirmed in hGH-hGHS-R transgenic mice in future. When comparing hGHS-R expression under rGHRH or hGH promoter control it became clear that the levels of transgene expression were representative of endogenous GHRH and GH expression levels. hGHS-R over-expression in hypothalamic GHRH neurones was estimated as approximately 6-fold in rGHRH-hGHS-R transgenic mice having 8 copies of the transgene incorporated into their genome. Pituitary hGHS-R over­ expression in hGH-hGHS-R transgenic mice was measured as 50-fold WT GHS-R expression levels, although only two copies of the transgene were incorporated. These data suggest that the GHS-R over-expression per tissue indeed corresponds to the relative endogenous expression levels and cell numbers of GHRH and GH.

126 Chapter 7: Physiological studies of hGH-hGHS-R transgenic mice

7 Physiological studies of hGH-hGHS-R transgenic mice 7.1 Introduction As described in the Introduction (Chapter 1.4.4) GH status clearly has an influence on pituitary GHS-R expression levels. While the SDR rat (Spontaneous Dwarf Rat, due to GH mutation) has increased pituitary GHS-R expression levels, which can be reduced by GH treatment (Kamegai et al, 1998), high plasma GH levels induced by ectopic GH tumours decrease pituitary GHS-R levels (Nass et al, 2000). However, pituitary tumours from acromegalic patients have been shown to express increased levels of pituitary GHS-R mRNA (Korbonits et al, 1998), which might explain why their responsiveness to GHSs is increased in some cases (Alster et al, 1993; Ciccarelli et al, 1996; Hanew et al, 1994; Popovic et al, 1994). It is unclear what effect altered GHS-R expression at the pituitary level would have on GH and GHS physiology. I demonstrated earlier that approximately 6-fold over-expression of the hGHS-R in hypothalamic GHRH neurones caused alterations in pituitary GH content and GHS responsiveness in male rGHRH- hGHS-R transgenic mice. In Chapter 6 I showed that over-expression of the hGHS-R was 50-fold in the anterior pituitary of hGH-hGHS-R transgenic mice, so this chapter describes initial studies investigating the implications of this level of hGHS-R over­ expression on the GH and GHS axis in these transgenic mice.

7.2 hGH-GHS-R transgenic mice are GH defîcient The reproductive axis of rGHRH-hGHS-R transgenic mice seemed unaffected, with no effect of the transgene on breeding or littersizes. Analysis of growth curves for male and female hGH-hGHS-R transgenic mice and their WT littermates revealed that female hGH-hGHS-R transgenic mice had significantly reduced body weights compared to their WT littermates (Table 7.1). While male hGH- hGHS-R transgenic mice showed no significant weight reduction at any age, female hGH-hGHS-R transgenic mice were slightly reduced in weight from day 30 (Table 7.1). At first FI pups from founders B, C and D were weighed once weekly and their body weight was recorded together. No pups from founder F have been studied yet. However, it became obvious that the phenotype differed among founders, as seen when a scatter plot was drawn for female pups from each founder in comparison to WT mice (Fig. 7.1).

127 Chapter 7: Physiological studies of hGH-hGHS-R transgenic mice

Body Weight (33-35d) [g] Body Weight (97-1 OOd) [g]

Male Female Male Female

WT 21.5±0.6 17.5±0.2 34.2±0.7 25.0±0.6 (n) (11) (18) (11) (14)

hGH-hGHS-R 20.2±0.7 16.1 ±0.3*** 33.7±0.7 21.7±0.4***

(n) (10) (13) (16) (17)

Table 7.1 Body weight of hGH-hGHS-R transgenic mice Animals from several litters were weighed weekly and their weight is displayed here before and after they reached adulthood. While female transgenic mice showed reduced weight from a young age, male transgenic mice did not display this phenotype at any age. ("**P<0.001)

30 n

♦ WT 25 - O) ♦ B O) I D ■ C 20 .

# ■

50 70 90 110

age [days]

Figure 7.1 Female hGH-hGHS-R and WT body weight sorted for founders Animals from several litters were weighed once a week and their body weights are recorded here according to their founder. A logarithmic curve was fitted to the weight data per founder. While all hGH-hGHS-R transgenic pups showed reduced weight, pups from founder C were particularly affected.

128 Chapter 7 : Physiological studies of hGH-hGHS-R transgenic mice

Figure 7.1 demonstrates that female pups from all founders were significantly reduced in weight, but pups from founder C showed more severe weight reduction than founders B and D. The body weight of female hGH-hGHS-R transgenic pups from founder B and D was reduced by 9%, while body weights of hGH-hGHS-R transgenic pups from founder C were reduced by 16% compared to WT littermates. No significant weight differences were observed for male pups from founder B or D compared to WT littermates at any age. At this stage not enough male pups from founder C were available to test whether male C-pups showed a more severe weight reduction compared to male pups from founder B and D. To analyse whether this growth retardation was a result of pituitary GH deficiency, pituitary homogenates from hGH-hGHS-R mice were analysed by GH RIA. Again initial data were based on mice from different founders and it was clear that hGH- hGHS-R transgenic mice were GH deficient (Figure 7.2). Both male and female adult hGH-hGHS-R transgenic mice showed significant pituitary GH content reduction (male: WT = 70.5±3.5; hGH-hGHS-R = 44.4±5.8|ig/pituitary, n=7; P<0.01; female: WT = 62.0±3.3; hGH-hGHS-R = 31.6±5.7 |ig/pituitary, n=7; P<0.001). Separation of the GH content of hGH-hGHS-R transgenic mice by founders showed that mice from founder C were more affected than founder D in both male and female animals (Figure 7.2). Female hGH-hGHS-R transgenic mice from founder C only had 35% of the WT mGH pituitary content, while pups from founder D had double that, 71%. In males hGH- hGHS-R transgenic mice from founder D also had 68% of the WT mGH content. Knowing now that different hGH-hGHS-R founders display different phenotypes, the GH content for mice from each founder should be measured with larger numbers and correlated with their respective transgene copy number. However, one observation was clear; over-expression of high levels of GHS-R were correlated with reduced pituitary GH contents. Prl content was measured from the same pituitaries, to analyse whether the hGHS-R over-expression specifically affected GH content. No difference in Prl levels was found between hGH-hGHS-R transgenic and WT mice, neither in male nor female adult mice, nor between transgenic animals from different founders (male: WT = 1.6±0.1, hGH-hGHS-R = 1.4±0.1|i,g Prl/ pituitary; female: WT = 10.5±0.8, hGH-hGHS- R = 11.6±1.0jig Prl/ pituitary; n=7, n.s.) (Figure 7.3). These data suggested that GHS-R over-expression caused a phenotype specific to GH cells.

129 Chapter 7: Physiological studies of hGH-hGHS-R transgenic mice

80 - Î

60 - 1C (0 **

*** I t 40- (3 5 *** 2 0 - 1 //' i Ll // 0 - I WT hGH- WT hGH- C D hGHS-R hGHS-R n= 7 7 4 3

Figure 7.2 Pituitary GH content in hGH-hGHS-R transgenic and WT mice GH content was measured in pituitary homogenates of adult male and female hGH- hGHS-R transgenic mice and their WT littermates. The red bars represent the pooled data of mice from different founders, while the coloured bars to the right represent the hGH-hGHS-R transgenic mouse pituitary GH content sorted for founders. All transgenic mice were GH deficient, with pups from founder C more affected. (Significance marks are in comparison to WT: *P<0.05, **P<0.01, ***P<0.001)

The significant and specific GH deficiency in hGH-hGHS-R transgenic mice was somewhat unexpected, as hGHS-R over-expression in GHRH neurones had caused enhanced pituitary GH content. In rGHRH-hGHS-R transgenic mice the over-expression of the hGHS-R also caused enhanced GHS responsiveness and the question was whether the GH deficiency in hGH-hGHS-R transgenic mice would alter GHS responsiveness in these animals. Therefore I again provided D.F. Carmignac (NIMR) with adult (10-12 weeks) female hGH-hGHS-R transgenic and WT mice, which were anaesthetised and one of the jugular veins cannulated. Blood samples were withdrawn before and 5 minutes after an i.v. injection of GHRP-6 (500ng) and GHRH (lOOng).

130 Chapter 7: Physiological studies of hGH-hGHS-R transgenic mice

15 - ? :

I 10- .t: a.

? 5 -

0 - □ B WT hGH- B C D WT hGH- C D hGHS-R hGHS-R n= 7 7 1 1 5 7 7 4 3

Figure 7.3 Pituitary Prolactin content in hGH-hGHS-R transgenic mice Prolactin content was measured in pituitary homogenates of adult male and female hGH-hGHS-R transgenic mice and their WT littermates. The red bars represent the pooled data of mice from different founders, while the coloured bars to the right represent the hGH-hGHS-R transgenic mouse pituitary Prl content sorted for founders. No significant difference was found between hGH-hGHS-R transgenic and WT mice or between founders.

Initially adult female hGH-hGHS-R mice from different founders were analysed together (Figure 7.4 left panel, red bars) and later sorted in accordance to their founder (4 C, pink and 3 D, yellow; figure 7.4 right panel). The results were similar, with hGH- hGHS-R transgenic mice from founder C being more affected than those from D. The response to 500ng of GHRP-6 was reduced, but not significantly (WT = 92.1 ±25.3; hGH-hGHS-R = 62.3±13.4ngGH/ml plasma, n=7, n.s.). Founder C, which was shown to have a greater GH deficiency, secreted much less GH in response to GHRP-6. Perhaps due to high SEM and low n this did not reach significance. Animals from founder D however showed normal responses to GHRP-6. A second surprising result was the poor GH response to lOOng of GHRH (WT = 158.2±25.9; hGH-hGHS-R = 40.2±16.9ng GH/ml plasma, n=7, P<0.01). Mice from both founders showed significantly reduced GH responses to GHRH. In fact the GH responses to GHRH were not even significantly different to their basal GH levels.

131 Chapter 7; Physiological studies of hGH-hGHS-R transgenic mice

200 n 200 n

GHRP-6 GHRH GHRP-6 GHRH SOOng 100ng 1 150^ SOOng lOOng 150- aI a c 100h 100 - 1 1O n.s. CO ** c 50 - n.s ** n.s 0 - 0 J I à WT T WT T WT T WT T WTC D WTC D WTC D WTC D time 0 +5 +120 +125 0 +5 +120 +125

Figure 7Ain vivo GH responses to GHRP-6 and GHRH in female hGH-hGHS-R transgenic and WT mice hGH-hGHS-R and WT mice were anaesthetised and their jugular vein cannulated to inject GHSs and withdraw blood samples. Plasma GH levels were analysed by RIA. Left panel: Female hGH-hGHS-R transgenic mice from founders C and D showed enhanced basal levels of plasma GH, slightly reduced responses to GHRP-6 (SOOng) and did not respond to GHRH (lOOng) (n=7). Right panel: Sorting of hGH-hGHS-R transgenic mice for founders demonstrated that C-mice responded less to GHRP-6 and both 0- and D-mice did not respond to GHRH at all (C n=4; D n=3). Significance marks are in comparison to WT; **P<0.01; *P<0.05.

This experiment was repeated in adult male hGH-hGHS-R transgenic and WT mice (figure 7.5). All hGH-hGHS-R transgenic mice were from founder D. GH responses to GHRP-6 were normal, but again GH responses to GHRH were somewhat low, however not significantly compared to WT responses (WT = 167.4±51.4; hGH-hGHS-R = 82.7±20.3ng GH/ ml plasma, n=6-7, n.s.).

The reduced GH responses to GHRH in female hGH-hGHS-R transgenic mice could be a direct effect at the pituitary level, but this phenotype could be influenced by hypothalamic GHS effects. To analyse the pituitary effect separated from a hypothalamic effect, D.F. Carmignac incubated pituitaries of adult female hGH-hGHS-R and WT mice (founder B and D) in vitro before and after a challenge of 1 and 5|Lig/ml GHRH and the release of GH was measured by RIA.

132 Chapter 7: Physiological studies of hGH-hGHS-R transgenic mice

250-1

200 - GHRP-6 GHRH SOOng 100ng (0 I 150. Q. II n.s I 100- (3 o> c 5 0 -

0 - WT T WT T WT T WT T

time 0 +5 +120 +125

Figure 7.5 In vivo GH responses to GHRP-6 and GHRH in male hGH-hGHS-R transgenic and WT mice Male hGH-hGHS-R transgenic mice showed slightly enhanced basal GH plasma levels, and normal GH responses to GHRP-6 (SOOng). The GH responses to GHRH (100ng) however seemed slightly reduced in hGH-hGHS-R transgenic mice, although this was not significantly different compared to WT responses (n=6-7). Significance marks are each in comparison to WT: *P<0.05

Basal GH release from hGH-hGHS-R pituitaries was reduced compared to WT pituitaries, but this did not always reach significance. GH responses to both doses of GHRH were significantly lower in hGH-hGHS-R transgenic pituitaries (Ipg: WT = 47.9±10.7, T = 21.8±3.3 ng GH/ lOOpl, 5|ig: WT = 70.5±3.2, T = 38.5±13.3ng GH/ lOOp.1, n=7, Mann-Whitney-Test, P<0.05) (Figure 7.6). However, the GH increase in response to GHRH treatment was significant for both WT and transgenic pituitaries. These data suggested that the poor GH responses to GHRH treatment in vivo were mirrored by significantly reduced responses in vitro, suggesting a direct pituitary effect of the hGHS-R over-expression.

133 Chapter 7: Physiological studies of hGH-hGHS-R transgenic mice

80 Ipg GHRH 5pg GHRH ♦ ♦♦ I I T 60 - □ WT I ■ I i 40 o O) T c

20 - T T T T * T 1 i 0 J iüiUi I ■ 30 60 90 120 150 180 210 240

minutes

Figure 7.6 In vitro responses of female hGH-hGHS-R transgenic and WT pituitaries to GHRH treatment Female hGH-hGHS-R transgenic (founder B and D) and WT pituitaries were incubated before and after a challenge of 1 and 5|ig/ml GHRH and the release of GH was measured by RIA. Basal GH release was reduced, though not always significantly, and the GH responses to both doses of GHRH were significantly reduced in hGH-hGHS-R transgenic pituitaries (n=7). Rhombic significance marks (♦) compare GH responses before and after the addition of GHRH (*P<0.05,**P<0.01,***P<0.001), while asterisk significance marks compare WT and transgenic GH release (*P<0.05).

7.3 Discussion The hGH-hGHS-R transgenic mice were the last line of transgenic mice to be created. The analysis of their physiological phenotype is therefore still very much at the beginning and I am indebted to D.F. Carmignac for analysis of GHS responsiveness in hGH-hGHS-R transgenic animals. In Chapter 6 I described four lines of hGH-hGHS-R transgenic mice I generated, with 50-fold over-expression of the hGHS-R in the anterior pituitary. To analyse whether this level of GHS-R over-expression in the GH cell had an influence on endogenous GH physiology, the pituitary GH content was measured. Both male and female hGH-hGHS- R transgenic mice of three analysed founders (B, C and D) were significantly GH

134 Chapter 7 : Physiological studies of hGH-hGHS-R transgenic mice deficient. In the females this GH deficiency coincided with significantly reduced body weight compared to their WT littermates. At the moment it is unclear, why the same percentage of GH deficiency for example in hGH-hGHS-R transgenic mice from founder D in males and females, results in a weight phenotype in females, but not in males. However, whether the absence of the weight reduction in male hGH-hGHS-R transgenic mice is really due to a sexual dimorphism in growth rate in response to GH deficiency will need further investigation, perhaps with larger amounts of male animals from all founders. The similar in vivo GH responses to GHRP-6 and reduced responses to GHRH in male and female hGH-hGHS-R transgenic mice from founder D suggests that there might not be a sexual dimorphism in the overall phenotype of hGH-hGHS-R transgenic mice. These studies highlighted an interesting difference between hGH-hGHS-R founders. While founder B and D seemed to give similar results, mice from founder C show a similar phenotype, but are clearly more affected by their hGH-hGHS-R transgene in terms of body weight, pituitary GH content and GHRP-6 and GHRH responses. A straightforward explanation for this would be a difference in copy numbers and resulting hGHS-R expression levels, since the hGH LCR has previously been reported to cause copy-number-dependent, position-independent transgene expression (Bennani-Baiti et al., 1998; Jones et al., 1995; Su et al, 2000). The quality of the Southern Blots did not permit me to make a reliable copy-number approximation per founder at this time. Extensive quantitative analysis of hGH-hGHS-R transgene expression in the anterior pituitary needs to be performed in these lines to compare expression levels between founders for example by in situ hybridisation and RNAse protection analysis. Interestingly hGH-eGFP transgenic mice were demonstrated to be significantly GH deficient as well (Magoulas et al, 2000). Both male and female hGH-eGFP transgenic mice contained only 30-50% pituitary GH compared to WT littermates, but this did not affect their growth. The reduction of GH stores in hGH-eGFP transgenic mice suggested a competition between mGH and the GH/eGFP product for granule packaging. This cannot be the case for hGH-hGHS-R transgenic mice, since their transgene product was not targeted to granules. Still, in both cases the somatotroph seems to be influenced by the expression of transgenes off the hGH LCR. Previously Jones et al have shown in transgenic mice over-expressing hGH in anterior pituitary somatotrophs, that the more

135 Chapter 7: Physiological studies of hGH-hGHS-R transgenic mice exogenous hGH was expressed, less endogenous mGH was produced (Jones et al, 1995). The additional hGH promoter in the GH cell might therefore cause a shortage of available transcription factors. Closer analysis of the hGH-hGHS-R transgenic anterior pituitary somatotrophs concerning their morphology and number might give insight into whether it is the GH content per pituitary that is affected or the GH content per cell. Prolactin was unaffected by the expression of the hGH-hGHS-R transgene, suggesting the effect of hGHS-R over-expression in the anterior pituitary was specific to GH. Based on the in vitro data presented in Chapter 3 ,1 expected that over-expression of the pituitary GHS-R would cause enhanced GH responsiveness to GHSs. The in vitro data in Chapter 3 seemed to suggest that over-expression of the GHS-R in GC cells caused enhanced PI turnover, i.e. signal transduction, basally and in response to GHSs. However, this was clearly not the case in vivo. While male and female transgenic mice from founder D showed normal GHS responsiveness, female C-mice showed a reduced response to GHRP-6 and no significant responses to GHRH was seen in any female hGH-hGHS-R transgenic mice. Incubation of whole pituitaries from female hGH- hGHS-R transgenic mice from founder D in vitro confirmed that in the absence of a hypothalamic signal, the GHRH response was significantly reduced in hGH-hGHS-R transgenic pituitaries. The reduction of GHRP-6 and GHRH responses in vivo of hGH- hGHS-R transgenic C-mice might simply be due to the fact that these animals have less GH available to release. D-mice still have a GH response to GHRP-6, but then fail to respond to GHRH. This might be, because after the GH discharge in response to GHRP- 6 the GH stores were too depleted to respond to GHRH, which might explain why pituitaries from hGH-hGHS-R transgenic mice from founder D do respond significantly to GHRH in vitro, though less than WT mouse pituitaries. It was shown previously that GH deficient little mice and the dwarfs of Sindh do not respond to GHRP-6 with GH release, although it is only their GHRH-R that is dysfunctional (Jansson et al, 1986; Maheshwari et al, 1999). Cross talk and dependence of the GHS-R signalling on an intact GHRH-R at the pituitary level were suggested, since hypothalamic neurones still responded to GHSs with c-fos activation and humans with a GHRH-R mutation still released ACTH, Prl and cortisol. Here I have shown that alteration of GHS-R expression in GH cells of the anterior pituitary also altered GHRH signalling, suggesting that the dependence of the two receptors’ intracellular signalling

136 Chapter 7; Physiological studies of hGH-hGHS-R transgenic mice pathways might be mutual. The discussion of this phenotype must remain speculative until we confirm the presented data, but it seems that hGHS-R over-expression in hGH-hGHS-R transgenic mice disrupts GHS-R and GHRH-R induced GH responses. Considering the level of GHS-R overexpression, expressing 50-fold the amount of GHS-R per somatotroph might alter internal pituitary signalling cascades, strained by this massive over-expression of the GHS-R. One way of investigating GHS-R and GHRH-R signalling is by measuring the PI turnover and cAMP accumulation in dispersed pituitary cells from hGH-hGHS-R transgenic mice, to study whether over-expression of the GHS-R has disrupted GHS and GHRH responses at this level. If transgene expression in hGH-hGHS-R transgenic mice from founder C was indeed found to be elevated compared to B and D this would suggest that more GHS-R expression per cell causes more GH deficiency and less GHS responsiveness. The somatotroph does not seem to cope with too much GHS-R expression very well. As for the rGHRH-hGHS-R transgenic mice, in this model of pituitary hGHS-R over­ expression a possible influence of Ghrelin and/or other unidentified endogenous GHS-R ligands has to be taken into account. In summary, I have shown that a 50-fold over-expression of the hGHS-R in anterior pituitary cells significantly altered GH physiology in hGH-hGHS-R transgenic mice. The observed phenotype of hGH-hGHS-R transgenic mice is somewhat surprising and these studies seem to suggest that the level of GHS-R over-expression caused disruption to GHRH responsiveness in all female hGH-hGHS-R mice and GHS responsiveness at least in hGH-hGHS-R from founder C. Over-expression of the hGHS-R at the level of endogenous GH expression seems to be problematic for the somatotroph. Here I have demonstrated that GHS-R over-expression in the anterior pituitary has a direct effect on pituitary cells, independent of a hypothalamic signal, but the physiological relevance is not clear.

137 Chapter 8: Ghrelin - a novel growth hormone secretagogue

8 Ghrelin - a novel growth hormone secretagogue 8.1 Introduction As described in the introduction, several key experiments indicated the existence of an elusive endogenous ligand for the GHS-R. Both the discovery of their own distinct signalling pathway (Pong et al, 1996) and the cloning of an endogenous receptor for an artificial class of drugs (Howard et al, 1996) suggested the existence of an endogenous GHS system. It was not until 22 years after the first discovery of GHS that finally an endogenous ligand for the GHS-R was cloned (Kojima et al, 1999). By monitoring Ca^^ changes after the addition of protein extracts from various tissues to GHS-R transfected CHO cells, ‘Ghrelin’ was purified from rat stomach extracts. Two active fractions were purified by HPLC and analysed by mass spectrometry. These two fractions were later shown to be splice variants of each other; Ghrelin (named Ghrelin 28 in this thesis) and des-Gln’"*-Ghrelin (Ghrelin 27) (Hosoda et al, 2000). Surprisingly both peptides, 28 and 27 amino acids long, were shown to be octanoylated at the serine 3 amino acid and that this was necessary for their activity, suggesting a novel posttranslational mechanism for acyl modifications of peptides. In vitro experiments showed that Ghrelin 28 was competitively inhibited by a GHRP-6 antagonist and that a high dose of Ghrelin (lO'^M) specifically released GH from primary rat cultures at BC50 values comparable to GHRH. In vivo experiments with both Ghrelin 28 and 27 further demonstrated that i.v. injections of l|Xg/250g rat and above caused an increase in plasma GH levels after 10-20 minutes. A single i.c.v. injection of Ghrelin also released GH dose-dependently, but after 12 days of continuous i.c.v. infusion of Ghrelin elevated plasma GH levels were not sustained (Date et al, 2000). These data suggest that similar to other GHSs the GH response to Ghrelin is prone to desensitisation. These GH responses to Ghrelin in anaesthetised rats were specific to GH and did not change plasma ACTH, FSH, LH, prolactin or TSH levels. However these data have been somewhat over-interpreted, since GHRP-6 does not release Prl, TSH or ACTH in anaesthetised rats either (unpublished data: R.G. Clark and I.C.A.F. Robinson). No data are yet available for the GH specificity of Ghrelin in conscious animals. It will be interesting to see, whether an endogenous ligand is more specific in the conscious state than the artificial GHS, which also release Prl and ACTH (Smith et

138 Chapter 8: Ghrelin - a novel growth hormone secretagogue al, 1997). Recently Ghrelin was shown to release GH after a single injection in humans with the GH release being higher than that after GHRH (Arvat et al, 2000). As mentioned previously Ghrelin 28 and 27 were found to be splice variants of each other derived from the 117 amino acid pre-pro Ghrelin. Ghrelin 28 was cloned from rat and human (82.5% amino acid homology), while Ghrelin 27 has yet only been reported in rat. The distribution of cDNA clones was reported as 1 (Ghrelin 27):5(Ghrelin 28), while in the stomach the peptide concentrations were 32pmol/g and 120pmol/g respectively. Kojima et al. described an intron between nucleotide 134 and 135. A tandem GAG pair serves as an alternative splice signal and thereby creates two alternative peptides from a single gene. By in situ hybridisation and immunocytochemistry the distribution of Ghrelin 28 in the stomach was shown to be localised from the neck to the base of the oxyntic gland in compact, dense core granules of cells, that due to their morphological features and distribution were classed as X/A-like cell (Nakazato et al, 2000). The authors also suggested that Ghrelin accounted for 20% of the endocrine cell population in the oxyntic gland. Immunocytochemistry further demonstrated expression of Ghrelin 28 in ARC neurones in the hypothalamus. Experiments with a NHj-terminal acyl-specific antibody revealed Ghrelin abundance to be low in cell bodies, and high in nerve terminals in the ME (Kojima et at., 2000a), a common distribution for hypothalamic secretory peptides. No in situ hybridisation data on Ghrelin mRNA expression in the hypothalamus are available. Northern Blot analysis revealed Ghrelin 28 expression in the stomach, small intestine, pancreas and a different size transcript in the heart. By RIA Ghrelin was also shown to circulate in human blood. In studies presented in abstract form (The Endocrine Society’s 82”^* Annual Meeting) Ghrelin 28 expression was analysed by RT-PCR and it was detected in the stomach, small and large intestine, pancreas, liver, heart, lung, white and brown adipose tissue, hypothalamus, pituitary, forebrain, spleen, skeletal muscle, testis, placenta, thymus, adrenals, prostate, and kidney (Iwakura et al, 2000; Komatsu et al, 2000; Mori et al, 2000; Nakazato et al, 2000; Torsello et al, 2000). Another feature Ghrelin has in common with the GHS, is its effect on NPY; i.c.v. injections of Ghrelin 28 upregulated hypothalamic NPY mRNA expression (Kamegai et al, 2000). Furthermore Bowers et al demonstrated that in rat in vivo experiments Ghrelin’s actions in combination with GHS antagonists and GHRH are very comparable

139 Chapter 8: Ghrelin - a novel growth hormone secretagogue to artificial GHS actions (Bowers et al, 2000). Sugihara et al underlined this, by showing that Ghrelin and GHRH only cause synergistic GH responses in vivo, not in vitro (Sugihara et al, 2000) and thereby also suggesting a hypothalamic component in Ghrelin action. Recently Hewson et al confirmed a hypothalamic action of Ghrelin in c- fos studies (Hewson and Dickson, 2000). Ghrelin was shown to activate ARC neurones after an i.v. injection in fed rats and interestingly more neurones were activated in response to Ghrelin and GHRP-6 in fasted rats. This unusual peptide as a novel ligand for the GHS-R and its involvement in the hypothalamo-pituitary GH axis is intriguing. As previously discussed the up-regulation of the GH axis in rGHRH-hGHS-R transgenic mice might stem from altered endogenous ligand activity. I therefore analysed Ghrelin expression in rGHRH-hGHS-R transgenic and WT mice.

8.2 Cloning of mouse Ghrelin 28 and 27 In order to analyse Ghrelin expression in rGHRH-hGHS-R transgenic mice the mouse homologue of rat and human Ghrelin had to be cloned. Initial searches in the GenBank demonstrated that a mouse peptide 95.7% identical to rat and human Ghrelin called mouse gastric peptide m46 (AJ243503) had been reported. P. LeTissier (NIMR) further analysed data from the GenBank and by comparison of the human Ghrelin cDNA to human genomic BAG clones he established the genomic structure of Ghrelin, comprised of 5 exons with the ATG in the 2"‘‘ exon and the STOP codon in the 5*. (Figure 8.1). With this information P. Le Tissier cloned mouse Ghrelin cDNA from stomach and hypothalamus by RT-PCR and 5’RACE (rapid amplification of cDNA ends). Soon it became apparent that two out of four cDNA clones contained a shorter form of Ghrelin, missing one of two Gin in position 14. Analysis of the genomic sequence showed that these two forms of Ghrelin were alternative splice forms of each other and were named mouse Ghrelin 28 and 27. This mechanism of splicing two peptides out of a single gene was later confirmed to be operational in rats as well (Hosoda et al, 2000). A further Ghrelin transcript with Ghrelin exon 2-5, but a different first exon was cloned by 5’RACE from heart (Figure 8.1). This heart transcript will make the ‘normal’ Ghrelin 28 peptide.

140 Chapter 8: Ghrelin - a novel growth hormone secretagogue

stomach

heart ------P7773 ------WÊM ------WWÊ — ------WSSm------

Figure 8.1 Mouse Ghrelin genomic structure in stomach and heart Stomach Ghrelin consists of 5 exons, with the ATG in exon 2 and STOP in exon 5. A transcript obtained by 5’RACE in heart contained an alternative exon1.

8.3 Expression analysis of mouse Ghrelin I used a full length pre-pro-Ghrelin 28 riboprobe to analyse the mouse Ghrelin expression pattern in several tissues by in situ hybridisation. In stomach I found the same pattern in mouse as reported in the rat with very strong Ghrelin mRNA expression from the neck to the base of the oxyntic gland (Figure 8.2). Dipping the sections in photographic emulsion showed that mouse Ghrelin was expressed heavily in cells at the base of the microvilli near the nerve plexus (Figure 8.3). Analysis of gut sections showed that gut Ghrelin was restricted to few scattered, but highly labelled cells (Figure 8.2 and 8.3). Stomach and gut Ghrelin in situ hybridisations were left to develop for two days, while other tissues only showed signals after two weeks of exposure to film. Ghrelin expression in mouse testis appeared restricted to some tubules, but not others (Figure 8.2). Ghrelin expression in the heart was not visible on film, but analysis at higher resolution showed only two Ghrelin positive cells in 6 heart sections (data not shown). Histology of these frozen sections was not ideal and the exact localisation or nature of these cells was unclear. Analysis of Ghrelin expression in the brain showed that mRNA expression levels were extremely low, if present at all. The anti-sense hybridisation signal was hardly distinguishable above a sense signal (Figure 8.2). Higher resolution analysis showed no specific labelling in the hypothalamus and only few scattered cells seemed to be labelled, but had no distinct localisation pattern in the brain. No Ghrelin expression was found in the kidney, thyroid, lung or pituitary with this technique.

141 Chapter 8; Ghrelin - a novel growth hormone secretagogue

Anti-sense Sense

stomach

gut

testis

brain

Figure 8.2 Ghrelin expression in mouse tissues 12pm frozen tissue sections were hybridised with a full length Ghrelin anti-sense (left panel) and sense (right panel) riboprobe. Stomach and gut were exposed for two days, while testis and brain needed two weeks exposure. Strong Ghrelin expression was apparent in the stomach and specific cells in the gut. Some tubules in the testis showed Ghrelin hybridisation as well, while in the brain no distinct hybridisation signal above background was visible. Scale bar =1 mm

142 Chapter 8; Ghrelin - a novel growth hormone secretagogue

Stomach

m â t Gut

r '

Figure 8.3 High resolution imaging of mouse Ghrelin expression in stomach and gut Ghrelin in situ hybridisation tissue sections were dipped into photographic emulsion, developed and counterstained with cresyl fast-violet. In the stomach Ghrelin was expressed strongly in specific cells at the base of microvilli near the nerve plexus. In the gut Ghrelin expression was distributed sparsely, but expression in single cells was high. Scale bar = 50pm

143 Chapter 8: Ghrelin - a novel growth hormone secretagogue

Kojima et al. have identified an acromegalic patient suffering from Ghrelin secreting carcinoids (personal communication I.C.A.F. Robinson), thus indicating that overproduction of gastric Ghrelin will have an effect on the GH axis. Since in other acromegalic patients ectopic GHRH over-expressing tumours cause GH hypersecretion, and GHRH has previously been identified in the GI tract (Bosman et al, 1984), I aimed to analyse whether GHRH was expressed in the mouse stomach and intestine and if so, if it might co-localise with Ghrelin. In situ hybridisation and subsequent high resolution imaging using a GHRH mRNA riboprobe did not detect any GHRH mRNA expression in the stomach or gut of WT mice.

In order to compare Ghrelin expression in rGHRH-hGHS-R and WT mice an RNAse protection assay measuring both Ghrelin 28 and Ghrelin 27 was established. Ghrelin 28 was used as a probe, such that hybridisation of the Ghrelin 28 probe to Ghrelin 27 would result in a base mismatch and be digested by the RNAse into a large 307 bp fragment and a smaller 115bp fragment. The assay was optimised for fragments above 150bp, so I chose to focus on the larger band for Ghrelin 27. High levels of Ghrelin expression were seen in the stomach, with no difference in stomach Ghrelin expression in rGHRH- hGHSR transgenic and WT mice (Figure 8.4). The relative expression levels of Ghrelin 27:28 in the stomach were 1:1.7 (n=4). Of the other tissues only Ghrelin expression in the gut was detectable. The levels of hypothalamic Ghrelin expression were so low that they are undetectable by this very sensitive method. However, by RT-PCR using mouse Ghrelin specific primers for exon 2 and 3 on the same samples, Ghrelin expression could be found in all tissues including hypothalamus. In both assays, RFA and RT-PCR, RNA amounts were controlled by measuring p-actin levels in the same samples. These data indicate that the amount of Ghrelin RNA in other tissues than stomach and gut is extremely low. Its physiological significance in all these tissues is thus not clear, especially since the type la GHS-R is not expressed in peripheral tissues (Howard et al, 1996; McKee et al, 1997).

144 Chapter 8: Ghrelin - a novel growth hormone secretagogue

RNAse Protection:

123456789

Ghrelin 28

Ghrelin 27

RT-PCR:

1 stomach WT 5 heart WT 2 stomach rGHRH-hGHS-R 6 testis WT 3 hypothalamus WT 7 gut WT 4 hypothalamus rGHRH-hGHS-R 8 kidney WT 9 brain WT

Figure 8.4 Ghrelin RNAse protection assay and RT-PCR For RNAse protection assay (RPA) 10|ig of RNA were loaded per tissue and hybridised with a Ghrelin RNA probe over night. RNAse digestion and subsequent gel-separation revealed that Ghrelin 27 and 28 were measurable only in WT and rGHRH-hGHS-R stomach and WT gut. RT-PCR with mouse Ghrelin specific primers on 0.5 pg cDNA showed expression in all tissues analysed.

8.4 Discussion The cloning of the novel GHS Ghrelin mainly from the rat and human stomach has raised many new questions in GHS physiology. The nature of Ghrelin, its size of 28aa and the somewhat ‘bulky’ octanoyl group essential for Ghrelin’s functional activity, is surprising, considering that the artificial GHSs are comparatively small compounds. The process of Ghrelin’s octanoylation of the serineS needs future attention as to what the posttranslational mechanism for acyl modifications of proteins is.

145 Chapter 8: Ghrelin - a novel growth hormone secretagogue

In this study P. Le Tissier cloned the mouse Ghrelin homologue and established the genomic structure of stomach Ghrelin. The existence of two alternative splice forms of Ghrelin - Ghrelin 27 and 28 - were demonstrated. I measured their relative expression in stomach as 1:1.7, while the protein levels of Ghrelin 27:28 were previously shown to be 1:4 (Hosoda et al, 2000). Of interest is the Ghrelin expression in the heart, found by 5’RACE and in situ hybridisation. An alternative exon 1 suggests possible altered transcriptional and translational regulation for the heart Ghrelin. Hexarelin, a GHRP-6 analogue, has previously been shown to bind to rat cardiac membranes and Ong et al. suggested the existence of a different GHS-R subtype (Ong et al. 1998a,b). Hexarelin binding can be displaced by GHRP-6 but not MK-0677 (Bodart et al., 1999) and there is no GHS-R type la detectable in the heart (Guan et al, 1997; Howard et al, 1996), suggesting a different type of Hexarelin receptor in the heart than the GHS-R expressed in the hypothalamus and pituitary. There are now several studies of the cardiac and hemodynamic effects of GHSs (Bisi et al, 1999; De Gennaro Colonna et al, 1997; Locatelli et al, 1999; Tivesten et al, 2000). Therefore alternatively regulated Ghrelin expression in combination with a different GHS-R subtype or even another receptor could be of clinical interest in the heart. However, Ghrelin expression is extremely low in the heart, as was shown by in situ hybridisation and by it not being detectable in RPA.

A specific signal for Ghrelin expression was absent in the hypothalamus as shown by in situ hybridisation and RNAse protection analysis. Although Ghrelin can be amplified from hypothalamic RNA extracts and many other tissues by RT-PCR, the physiological relevance of these very low levels of expression not only in the hypothalamus, but in all these tissues is questionable. Immunocytochemistry assays demonstrated by Kojima et al. had located Ghrelin in this area and a RIA for Ghrelin showed that 117.2±37.2fmol/ml circulated in the blood (Kojima et al, 2000a). In Chapter 4 the GHS-R peptide was shown to be present in fibres of the ME. Since hypothalamic Ghrelin mRNA expression is low, but Ghrelin protein was located to this area, one could speculate that Ghrelin protein might be taken up into the hypothalamic area from the hypohpyseal blood system, since this area lies outside the blood brain barrier and by this means the rather ‘bulky’ Ghrelin would not need to cross it. Hewson et al. reported c-fos

146 Chapter 8: Ghrelin - a novel growth hormone secretagogue expression in the ARC after Ghrelin injection, suggesting a direct central effect for Ghrelin (Hewson and Dickson, 2000). The very low levels of Ghrelin expression in the hypothalamus will make the analysis of Ghrelin’s possible involvement in the GH axis difficult. Whether Ghrelin is the or the only endogenous GH regulating ligand for the GHS-R remains to be proven. Cloning of Ghrelin and its GH secreting action in vitro and in vivo suggests that our current understanding of the regulation of GH pulsatility may need modification. Ghrelin’s presence in the blood strengthens the notion of Ghrelin as a third GH regulator, but measurements of Ghrelin in portal blood and their relationship to GH pulses are needed to clarify its relevance to GH physiology. Since the main expression of Ghrelin is in the stomach it also makes sense to look at its actions in the gastric area. The expression of Ghrelin in endocrine cells close to the nerve plexus suggests that these might be cells that are sympathetically innervated, thereby regulating stomach motility. Solution hybridisation experiments by D.F. Carmignac showed that Ghrelin expression in the stomach was down-regulated in starved mice, thereby implicating involvement of Ghrelin in feeding and maybe gastric motility (personal communication D.F. Carmignac). Interestingly, other family members of the GHS-R were identified as the motilin receptor (MTL-R) (Feighner et at., 1999; McKee et ai, 1997) and the NMU receptor (Fujii et al, 2000; Howard et al.y 2000; Kojima et al.y 2000; Szekeres et al.y 2000). Both Motilin and NMU themselves are expressed throughout the gastrointestinal (GI) tract, being involved in GI tract motility (Howard et al. y 2000). It is thus feasible that Ghrelin might play a similar role in the GI system and food metabolism. Indeed, recently Masuda et al. confirmed that Ghrelin acts on the GI tract possibly by sympathetic innervation (Masuda et al. y 2000). I.v. injections of Ghrelin in rats dose-dependently stimulated gastric acid secretion and motility, and these stimulatory actions on gastric function were almost completely abolished by blocking vagal nerve activity (Masuda et al.y 2000). Whether Ghrelin’s actions on the GI tract via the vagal nerve pathway are mediated through the GHS-R remains to be proven. It was further shown that once daily s.c. Ghrelin injections for two weeks induced body weight gain in mice, but not food intake (Tschop et al.y 2000). However when Ghrelin was given i.c.v. food intake was enhanced. The body weight gain was shown to result from reduced fat utilisation, which was independent of GH and NPY (Tschop et al. y

147 Chapter 8: Ghrelin - a novel growth hormone secretagogue

2000). As shown previously in c-fos studies (Hewson and Dickson, 2000) and Ghrelin stomach expression studies (unpublished data D.F. Carmignac), Tschop et al. also indicated Ghrelin to be regulated by feeding state (Tschop et al, 2000). Kamegai et al had previously demonstrated that Ghrelin upregulated hypothalamic NPY expression (Kamegai et al, 2000) but Tschop et al showed that presence of NPY was not necessary for Ghrelin to exert its effects on energy metabolism (Tschop et al, 2000). Whether the effects of Ghrelin on GH secretion and metabolism are both mediated by the GHS-R remains to be proven. In preliminary RNAse protection assays I have shown that Ghrelin expression in the stomach was not affected by hypothalamic GHS-R over-expression in GHRH neurones, so the metabolic effects of Ghrelin might not be regulated by hypothalamic GHS-R, at least not those on GHRH neurones. Furthermore, by in situ hybridisation I have been unable to detect the GHS-R in the stomach or gut, suggesting that the effects of Ghrelin on the GI tract might not be mediated through local GHS-R.

In summary, our understanding of the physiological role of Ghrelin is still very much at the beginning. Its in vitro and in vivo actions suggest involvement in the regulation of GH secretion. In addition Ghrelin’s expression pattern and regulation in the stomach had implied a further action of Ghrelin in food metabolism and possibly gastric motility, which was recently confirmed (Masuda et al, 2000; Tschop et al, 2000). It seems that the GH secreting and metabolic effects of Ghrelin are similar to those of the artificial GHSs, although sympathetic innervation of the GI tract has never been reported for GHSs. Whether both the GH secreting and the metabolic effects of Ghrelin and the GHSs are mediated through the GHS-R and what the mechanism for Ghrelin’s influence on the GI tract is, remains unclear.

148 Chapter 9: GHS treatment studies in transgenic mice and rats

9 GHS treatment studies in transgenic mice and rats 9.1 Oral GHS treatment of Tgr and WT rats 9.1.1 Introduction GHRP-6 and the non-peptide GHSs have been shown to elicit an immediate and significant increase in GH secretion (Clark et al, 1989; Hickey et al, 1996). As described in the introduction, further studies demonstrated that MK-0677 could stimulate and maintain pulsatile GH release during once daily oral treatment in humans (Chapman et al, 1996). Despite the ability of GHSs to generate impressive pulsatile GH secretion, attempts to stimulate additional growth in animals and humans have not been as impressive. Still, several studies in rats have shown modest body weight gain in response to GHS treatment, mostly by once daily oral administration (Bowers et al, 1984; Hansen et al, 1999; Svensson et al, 2000; McDowell et al, 1995). Appropriate animal models for studying the growth promoting action of GHSs in the context of dwarfism have not been available. For example, whilst GHRP-6 activates GHRH neurones in the dw/dw rat (Dickson et al, 1995b), the GH deficiency is so profound that neither GHRH (Carmignac and Robinson, 1990) nor GHRP-6 (Carmignac et al, 1998) release sufficient GH to accelerate skeletal growth. Other dwarf models, e.g. the lit/lit mouse, hypophysectomised animals or animals with a pituitary stalk lesion are not useful in a growth study because of the dependence of GHSs on an intact GHRH system. A novel dwarf animal model in which elevated body weight gain in response to GHS treatment was demonstrated, is the Tgr rat (Flavell et al, 1996; Wells et al, 1997). The Tgr rat is a model of dominant GH-dependent dwarfism, created by the expression of hGH under the control of the rat GHRH promoter. The hGH expression in GHRH neurones suppresses endogenous GHRH production through a short-loop feedback mechanism, resulting in reduced endogenous pituitary GH levels. Both male and female Tgr rats show reduced body and skeletal growth, reduced pituitary weight, GH content and reduced plasma IGF-I concentrations, but all these were more pronounced in males than in females (Flavell et al, 1996; Wells et al, 1997). The male Tgr animals display halved GH peak height, peak areas and total GH secretion compared to WT littermates, while none of these GH parameters was significantly different in females (Wells et al, 1997). In addition, while male Tgr rats show a markedly reduced response to an i.v.

149 Chapter 9: GHS treatment studies in transgenic mice and rats injection of GHRP-6 compared to WT littermates, female Tgr rats responded at least as well to GHRP-6 as their WT littermates. Male Tgr rats infused with pulsatile GHRP-6 for 7 days demonstrated elevated body weight gain compared to that seen in untreated WT animals (Wells et al., 1997). Recently Wells et al. infused Tgr rats for 7 days i.v. with pulsatile GHRP-6 and found that this significantly increased their body weight gain and epiphysial growth plate width (Wells et al, 2000). Peptidylic GHSs and the non-peptidylic benzolactam derivatives show potent GH releasing effects, but suffer from low oral bioavailability (DeVita and Wyvratt Jr., 1996; Leung et al, 1996; Wyvratt Jr., 1996). However, studies for the non-peptidylic MK- 0677 showed that it was the first in a series of potent compounds with good oral bioavailability and activity (Patchett et al, 1995), which highlighted the therapeutic potential of this GHS in GH deficiency. Since the dwarfism in Tgr rats could be corrected by GHRP-6 treatment, the Tgr rats were shown to be the only dwarf animal model in which the growth promoting effects of GHSs could be studied effectively. The aim was therefore to test the ability of non-peptidylic GHSs to promote growth when given orally to Tgr rats for 14 days. Sufficient quantities of MK-0677 were not available for the size of study we had in mind. However, studies with the non-peptidylic compound L-163,255, a close analogue of MK-0677, had demonstrated that it was an orally active GH secretagogue suitable for long-term efficacy at least in pigs (EC5o=1.5nM, compared to 1.3nM for MK-0677);(Chang etal, 1996) and L163,255 was provided to me (a kind gift by A.D. Howard, Merck Research Laboratories, Rahway, New Jersey) in sufficiently large quantities for these studies.

9.1.2 Results In order to study the ability of L-163, 255 to stimulate growth when given orally, adult (10-12 weeks) male and female Tgr and WT rats were given L-163, 255 ad libitum in their drinking water. The groups were 8 each of: Male WT control, Male WT L-163,255, Female WT control. Female WT L-163,255, : Male Tgr control, Male Tgr L-163,255, Female Tgr control. Female Tgr L-163,255. For a week before the experiment all animals were housed together with 4 rats per cage and their water intake ‘baseline’ was measured. L-163, 255 was dissolved in 10% Ethanol and further diluted to its final concentration in 0.4% Ethanol in tap water according to the amount of water consumed

150 Chapter 9; GHS treatment studies in transgenic mice and rats per cage. The final concentration was calculated such that with normal water consumption the rats would receive 20mg L-163,255/kg body weight/day. Control rats received 0.4% Ethanol in tap water. All animals were weighed every second day and at the end of the treatment their heart, kidneys and spleen were weighed and their brain and pituitary removed and frozen on dry ice. Due to the bad taste of L-163, 255, all rats receiving the drug reduced their water intake to approximately 66%, such that the dose of L-163,255 was reduced to 13.2mg/kg/day. We further have to bear in mind that the reduced water intake would possibly also result in reduced body weight gain, if associated with a concomitant reduction in food intake. Analysis of the body weight gain for male Tgr and WT rats showed that there was no significant difference between L-163,255 treated and control animals, neither in WT nor in Tgr rats (Figure 9.1) (day 14 A weight gain: WT contr.=27.1±1.5g, WT L-163,255 = 25.7±2.5g, Tgr contr.=23.8±1.2g, Tgr L-163,255=24.6±1.9g, n=8, n.s.). It was also obvious in both groups, male Tgr and WT rats, that the GHS treated animals lost body weight in the first two days, possibly due to the reduced water intake, but then caught up to ‘normal’ body weight gain. Although Tgr rats were shown to have reduced body weights compared to their WT littermates, their growth rate after 9-10 weeks is similar (Flavell et ai, 1996). In this study male control WT and Tgr rats also displayed similar body weight gain rates. Analysis of the body weight gain for female Tgr and WT rats was remarkably different to that for males (Figure 9.1). While L-163,255 treated WT rats showed no significant difference in body weight gain compared to their untreated controls, female Tgr rats displayed significantly enhanced body weight gain in response to L-163,255 from day 6 of treatment (day 14 A weight gain: WT contr.=16.8±1.6g, WT L-163,255=18.9±2.2g, n=8, n.s, Tgr contr.=15.0±1.2g, Tgr L-163,255=20.3±.7g, n=8, P<0.01). Comparison of L-163,255 treated female Tgr and WT rats showed that the weight gain of treated Tgr rats was always greater than that of WT rats, but this was only significant on day 8 (WT=9.5±1.2g, Tgr=13.7±1.0g, n=8, P<0.05). Female control WT and Tgr rats showed indistinguishable body weight gain rates.

151 Chapter 9: GHS treatment studies in transgenic mice and rats

30 1 30-1 WT L-163,255 T L-163, 255 WT control T control

2 0 - 2 0 - o> £

•5 1 0 - 10 - <

-10 -10 2 4 6 8 10 12 14 2 4 6 8 10 12 14 days days

Î

25^ 25-1 WT L-163, 255 T L-163, 255 **

2 0 . WT control 20 - T control

O) 15- ** £

I 10. 10 - <

2 4 6 8 10 12 14 2 4 6 8 10 12 14 days days Figure 9.1 Body weight gain of L-163,255 treated (13.2mg/kg/day) adult Tgr and WTrats Adult Tgr and WT rats were treated orally witti L-163,255 (13.2mg/kg/day) in their drinking water for 14 days. Their body weight was recorded every other day and is depicted here as body weight gain [gj. No significant difference was found between male treated and control, WT and Tgr rats. Female Tgr rats gained weight significantly faster in response to L-163,255 than their controls (n=8 for each group; *P<0.05, **P<0.01, ***P<0.001).

152 Chapter 9: GHS treatment studies in transgenic mice and rats

To test whether the difference in weight gain in response to L-163,255 between male and female Tgr was related to differing changes in pituitary GH content, the pituitary GH content was measured in control and treated male and female Tgr rats (Figure 9.2). Although the GH content in treated female Tgr was slightly enhanced compared to Tgr controls, no significant differences were observed between treated and control groups, neither for male nor female Tgrs (male Tgr contr.: 626±56, male Tgr L-163,255: 567±79, female Tgr contr.: 245±22, female Tgr L-163,255: 301±31|ig GH/pituitary, n=8, n.s.). The pituitary GH content of L-163,255 treated female Tgr could thus not fully explain their increased body weight gain.

800 -

600 - >, â 3 I 400 - %

200 -

%

0 - contr L-163,255 contr L-163,255

o ’ Î Figure 9.2 Pituitary GH content of L-163,255 treated (13.2mg/kg/day) adult Tgr rats GH contents were assayed in pituitary homogenates of Tgr rats by RIA. Control and L- 163,255 treated Tgr rats showed no significant differences, neither in male nor in female Tgrs. (n=8)

Previously it has been shown that GHRP-6 treatment caused enhanced food intake in rats, when given i.c.v. (Kuriyama et al., 2000; Locke et ai, 1995; Okada et al, 1996); however, no change in food intake was noted when it was given systemically (Okada et ai, 1996). Bennett et al. also showed that i.c.v. injections of GHRP-6 significantly enhanced ARC NPY expression in rats, which may explain the enhanced food intake

153 Chapter 9: GHS treatment studies in U'ansgenic mice and rats

(Bennett et al., 1998). To analyse, whether orally administered L-163,255 treatment altered ARC NPY expression the brains from all rats were cut into 12|im sections on a cryostat and their ARC NPY mRNA expression was analysed by in situ hybridisation with a rat NPY riboprobe. No significant changes in NPY mRNA expression were measured, possibly due to high background variability in all groups (Figure 9.3). Interestingly though, mean NPY mRNA was higher in all L-163,255 treated groups compared to their untreated control animals, except in the male treated Tgr rats. If this study was repeated, a grain count of NPY mRNA expression per neurone might be a more concise method than densitometry.

5000- ?

4000-

(0 c ■o

Figure 9.3 ARC NPY mRNA expression in orally L-163,255 treated Tgr and WT rats Hypothalamic ARC sections were used for an in situ hybridisation with a rat NPY riboprobe and the autorads were analysed by densitometry. Although in most L-163, 255 treated groups the NPY mRNA content seemed higher, no significant changes were found between any of the groups (n=8, n.s., Mann-Whitney Test).

9.1.3 Discussion Previously L-163,255 had been given to pigs through oral gavage and in feed and was shown to release GH from the pituitary at doses between 30 and 300p.g/kg (Chang et ai, 1996). Treatment with the drug (15mg/kg/day) in feed over 3 days showed elevated GH

154 Chapter 9: GHS treatment studies in transgenic mice and rats

AUC measurements over the initial 12-hour period and increased mean IGF-I levels over the treatment period. In addition, Tgr rats had previously been shown to respond to pulsatile s.c. GHRP-6 treatment with enhanced body weight gain. Therefore, I used Tgr rats to test the ability of the orally active MK-0677 analogue L-163,255 to promote growth. A dose of ~13mg/kg was given ad libitum to rats in their drinking water. Female treated Tgr rats showed significantly enhanced body weight gain in response to the L-163,255 treatment compared to Tgr untreated controls. Comparison of their body weight gain with treated WT rats showed that the treated Tgr weight gain was always higher, but this was only significant on day 8 of the study. Neither male WT or Tgr, nor female WT rats responded to the L-163,255 treatment with enhanced body weight gain. No significant difference was shown between male and female treated Tgr pituitary GH contents compared to their untreated controls. However, females showed a higher mean GH content than their controls, which does not fully explain, but might add to their enhanced body weight gain in response to GHS treatment. Similarly, male treated Tgr rats were the only GHS treated group that did not show slightly elevated mean NPY mRNA expression. Essentially, it remains unclear why female Tgrs responded to the treatment and male Tgrs did not. One explanation might lie in the differences in GH secretion and GH responses of Tgrs to acute i.v. injections of GHRP-6. While male Tgrs are impaired in GH secretion pattern and GH responses to GHRP-6 are attenuated compared to WT controls, female Tgr’s GH secretion and GH responses are normal. This might explain, why female Tgr rats responded better to the L-163,255 treatment than male animals. Also, data from the L-163,255 study were consistent with reports showing that GH responses to GHSs are greater and more consistent in female rats (Bercu et al, 1991; Sartor et al, 1985). Many studies with GHS treatment have therefore used female animals, e.g. (Bercu et al, 1992; Carmignac e ta l, 1998; McDowell e ta l, 1995; Svensson et al, 2000; Thomas et al, 2000). Furthermore in some cases E2 was shown to increase GH responses to GHRP-6 (Mallo et al, 1993) and Carmignac et al pointed out that E2 increased GHS-R transcription (Carmignac et al, 1998). These data gave a possible explanation as to why female Tgrs responded better than male Tgr, but then why didn’t WT females grow? Growth stimulation is best demonstrated in GH deplete animals and GH deficient Tgr rats have so far shown the

155 Chapter 9: GHS treatment studies in transgenic mice and rats most impressive body weight gain in response to GHSs, compared to other GHS treatment studies in normal rats (Wells et al, 1997). Dwarf rats {dw/dw) have previously been shown to have elevated GHS-R expression levels (Bennett et al, 1997) and if it holds true that GH deficiency in Tgr rats also causes enhanced hypothalamic GHS-R levels, their enhanced responsiveness might also be explained by enhanced hypothalamic GHS-R levels. Thus measurement of GHS-R mRNA levels in the animals of this study would have been helpful. As mentioned earlier, GHS treatment can affect body weight gain as enhanced feeding and fat mass rather than growth (Kuriyama et al, 2000; Locke et al, 1995; Okada et al, 1996). Studies giving L-163,255 i.c.v. showed that food intake was not affected in rats (personal communication D.F.Carmignac), suggesting that the body weight gain observed here in female Tgrs might indeed be growth. This cannot be stated with certainty, since in this study no measurement of food intake, fat mass or tibia length/growth plate width were performed and the slightly enhanced ARC NPY levels might suggest an effect on the metabolic axis. At the moment it is thus unclear, whether L-163,255 affected the GH or metabolic axis in female Tgrs. Especially in view of the recent study by Tschop et al, which showed that Ghrelin, an endogenous ligand for the GHS-R, given i.c.v enhanced food intake and caused increased body weight gain as a result of reduced fat utilisation (Tschop et al, 2000), more studies to investigate the growth/metabolic effect of oral L-163,255 treatment are needed. Nevertheless, data from this study showed that female, GH deficient animals were the most promising animals to elicit growth responses to GHS treatment. I suspected that this might be due to a contribution of elevated GHS-R expression levels and sensitivity to GH correction.

9.2 GHS treatment of rGHRH-hGHS-R and WT mice 9.2.1 Introduction The above described L-163,255 study suggested that enhanced GHS-R mRNA expression might contribute to enhanced responsiveness to GHS treatment. Our previous experiments suggested that transgenic mice over-expressing the GHS-R in the hypothalamus might show increased responsiveness to acute i.v. injections of GHRP-6 in vivo. I therefore aimed to investigate, whether enhanced hypothalamic GHS-R

156 Chapter 9: GHS treatment studies in transgenic mice and rats expression in GHRH neurones was able to enhance growth responses to long-term GHS treatment in rGHRH-hGHS-R transgenic mice. Previously McDowell et al had described experiments in which female rats were treated either continuously or intermittently s.c. with a GHS for 14 days, resulting in an enhanced body weight gain compared to saline treated rats, which was much greater with twice daily s.c. injections than with continuous s.c. exposure (McDowell et al, 1995). Similarly, Thomas et al confirmed that continuous s.c. infusion of GHRP-6 had an initial effect on weight gain in rats but failed to produce a sustained increase in weight gain altogether (Thomas et al, 2000). Fairhall et al. had shown that GH responses to GHSs were dependent on the pattern of GHS administration with more consistent GH responses when GHSs were administered with less frequent i.v. injections rather than continuous GHS infusion (Fairhall etal, 1995). Moderate body weight gain responses to GHSs have also been achieved in rats by once daily oral or s.c. treatment (Bowers et al, 1984; Hansen et al, 1999), pulsatile i.v. infusion (Wells et al, 2000; Wells et al, 1997) and continuous s.c. infusion (Svensson et al, 2000). Except for the Svensson et al study (Svensson et al, 2000) continuous infusion had shown desensitisation to GHS (McDowell etal, 1995; Thomas e ta l, 2000). Since in my study with oral L-162,255 administration it had proven difficult to control the exact dosage per animal per day, a once daily i.p. injection in mice was chosen instead. Furthermore, twice daily s.c. GHS injections had proven effective in rats (McDowell et a l, 1995) and I aimed to investigate the effectiveness of this injection schedule to promote body weight gain in rGHRH-hGHS-R transgenic mice. Transgenic rGHRH- hGHS-R mice and their WT littermates were thus exposed to GHRP-6 injections over 9 -14 days and the effects on body weight analysed.

9.2.2 GHRP-6 (25fig/day Ip.) treatment for 14 days in female rGHRH-hGHS-R transgenic mice Groups of age-matched 11-13 week old female transgenic rGHRH-hGHS-R and WT mice were injected i.p. daily with or without 25|ig GHRP-6 in 200)li1 saline (WT saline: n=5, WT GHRP-6: n=4, T saline: n=5, T GHRP-6: n=6). The body weight of these animals was recorded every three days. After 14 days treatment, mice were sacrificed and their ovarian fat pads and the soleus muscle weight measured. Brain and pituitary

157 Chapter 9: GHS treatment studies in transgenic mice and rats

were removed and frozen on dry ice. Over the 14 days treatment no significant difference in body weight gain between WT and rGHRH-hGHS-R transgenic mice was observed, whether they were treated with GHRP-6 or saline (Figure 9.4). Both, GHRP-6 treated rGHRH-hGHS-R and WT mice, showed a trend towards increased body weight gain compared to their saline controls, but this was only statistically significant in rGHRH-hGHS-R transgenic mice (day 14 Aweight gain: WT saline = 0.7±0.5g, WT GHRP-6 = 1.9±1.2g; n=4-5, n.s.; T saline = 1.0±0.6g; T GHRP-6 = 2.5±0.3g; n=5-6; P< 0,05).

4

3 T GHRP n=6 2 WTGHRP n=4 0» T saline n=5 O) 1 I <1 WT saline n=5 0

1

2 5 10 15 days

Figure 9.4 Body weight gain of GHRP-6 treated (25pg i.p. daily) adult female rGHRH-hGHS-R and WT mice Adult female rGHRH-hGHS-R and WT mice were treated once daily with a 25pg GHRP-6 i.p. injection. Their body weight was recorded every three days and is depicted here as body weight gain [g]. Only GHRP-6 treated rGHRH-hGHS-R transgenic mice showed a significant body weight gain in response to GHRP-6 (n=5-6; *P<0.05).

158 Chapter 9: GHS treatment studies in transgenic mice and rats

Analysis of the pituitary GH content confirmed our previous observation that female rGHRH-hGHS-R transgenic mice showed a slightly enhanced, but not significantly elevated pituitary GH content (Figure 9.5) (WT saline: 25.0±3.7; T saline: 33.0±5.8pg GH/pituitary, n=5, n.s.). The pituitary GH content in WT mice did not change in response to GHRP-6 treatment (WT saline: 25.0±3.7; WT GHRP-6: 26.8±3.3|Lig GH/pituitary, n=4-5, n.s.). However, the GH content in rGHRH-hGHS-R transgenic mice was significantly increased in the GHRP-6 treated group (T saline: 33.0±5.8; T GHRP-6: 52.2±5.4pg GH/pituitary, n=5-6, P<0.05). Pituitary GH content in GHRP-6 treated rGHRH-hGHS-R transgenic mice was also increased compared to treated WT mice (n=4-6; P<0.01).

60 n

50 -

I 40 _ ‘5 5 30 - I O o> 20 -

10 -

0 - W T W T rGHRH- rGHRH- WT hGHS-R hGHS-R

saline GHRP-6 saline GHRP-6

n= 5 4 5 6

Figure 9.5 Pituitary GH content of GHRP-6 treated (25pg i.p. daily) adult female rGHRH-hGHS-R and WT mice GH contents were assayed in pituitary homogenates from rGHRH-hGHS-R and WT mice by RIA. GHRP-6 treated rGHRH-hGHS-R transgenic mice displayed a significantly elevated GH content compared to saline treated rGHRH-hGHS-R transgenic and GHRP-6 treated WT mice. (n=4-6, *P<0.05, **P<0.01)

159 Chapter 9: GHS treatment studies in transgenic mice and rats

To determine whether there was an effect of the GHRP-6 treatment on lean body mass or fat mass, the soleus muscle and the ovarian fat pads were weighed. Although fat pad weights seemed slightly increased by GHRP-6 treatment in WT and rGHRH-hGHS-R transgenic mice, neither the fat pads nor the soleus weight were significantly different in any of the groups (Figure 9.6).

250 n

200 - o 5 1 5 O) E E 1 ^ 150 §. I 10 1 (A c 1 0 0 0)3 S 5 J 50 H

0 J rGHRH- rGHRH- WT WT S = WT WT hGHS-R hGHS-R saline GHRP-6 saline GHRP-6 saiine GHRP-6 saline GHRP-6

n= n=

Figure 9.6 Soleus and fat pad weight of GHRP-6 treated (25pg ip daily) adult female rGHRH-hGHS-R and WT mice No significant difference was found between soleus or ovarian fat pad weigfit in any of the groups (n=4-6).

9.2.3 GHRP-6 (IxlSfig s.c. daily) treatment for 9 days in male rGHRH-hGHS-R

transgenic mice Groups of age-matched (11-13 weeks) male transgenic rGHRH-hGHS-R and WT mice were injected twice daily s.c. with or without 15pg GHRP-6 in lOOpl saline (WT saline: n=6, WT GHRP-6: n=7, T saline: n=8, T GHRP-6: n=7). The body weight of these animals was recorded every other day. After 9 days treatment, mice were sacrificed, the soleus muscle weight measured and the brain, pituitary and tibiae removed and frozen on dry ice. A body weight gain curve showed that both GHRP-6 treated rGHRH-hGHS-R and WT

160 Chapter 9: GHS treatment studies in transgenic mice and rats mice gained weight significantly faster than the saline treated animals from day 4 of treatment (day 9 Aweight gain: WT saline = 0.0±0.5, WT GHRP-6 = 1.7±0.5g, n=6-7; P< 0.05; T saline = 0.0±0.5; T GHRP-6 = 2.2±0.4 g; n=7-8, P<0.01);(Figure 9.7). Although the body weight gain in rGHRH-hGHS-R transgenic mice was slightly higher in response to GHRP-6 compared to GHRP-6 treated WT animals this never reached significance over the 9-day period.

■k'k TGHRP n=7 •kit WT GHRP n=7 Ui T saline n=8 O) I WT saline n=6 <]

- 1-

2 4 6 8 days Figure 9.7 Body weight gain of GHRP-6 treated (2x15pg s.c. daily) adult male rGHRH-hGHS-R and WT mice Adult male rGHRH-hGHS-R and WT mice were treated twice daily with a 15pg GHRP-6 s.c. injection. Their body weight was recorded every other day and is depicted here as body weight gain [g]. Both rGHRH-hGHS-R transgenic and WT mice gained weight significantly faster in response to GHRP-6 than their saline injected controls (n=6-8; *P<0.05, **P<0.01).

Analysis of the pituitary GH content confirmed my previous finding that male rGHRH- hGHS-R transgenic mice have enhanced GH content compared to their WT littermates (Figure 9.8) (WT saline: 66.4±9.6; T saline: 107.9±9.6p,gGH/pituitary, n=6-8, P<0.05). GHRP-6 treatment slightly, but not significantly enhanced pituitary GH content in both rGHRH-hGHS-R and WT mice compared to their respective controls. The difference in pituitary GH content between rGHRH-hGHS-R and WT mice remained, when both

161 Chapter 9: GHS treatment studies in transgenic mice and rats groups were treated with GHRP-6 (WT GHRP-6: 76.6±4.6; T GHRP-6: 122.2±10.7|LigGH/pituitary, n=7, P

150 **

.g 100- 2 Q. I G ? 50-

WT rGHRH- rGHRH- WT hGHS-R hGHS-R saline GHRP-6 saline GHRP-6

n= a Figure 9.8 Pituitary GH content of GHRP-6 treated (2x15pg s.c. per day) adult male rGHRH-hGHS-R and WT mice GH content was assayed in pituitary homogenates from rGHRH-hGHS-R and WT mice by RIA. Saline and GHRP-6 treated rGHRH-hGHS-R transgenic mice displayed a significantly elevated GH content compared with saline treated and GHRP-6 treated WT mice respectively. (n=6-8, *P<0.05, **P<0.01)

Previously it has been shown that GHRP-6 treatment can have an effect on Prl secretion (Ciccarelii et ai, 1996; Copinschi et ai, 1996; Ghigo et ai, 1997; Hickey et ai, 1994; Kirk et ai, 1997; Massoud et al., 1996; Van den Berghe et ai, 1996). To investigate, whether GHRP-6 treatment affected pituitary Prl content in this study, it was measured in the same pituitary homogenates. Although slightly reduced in rGHRH-hGHS-R mice, the prolactin content between transgenic and WT mice was not significantly different in any of the groups (WT saline: 1.95±0.19, WT GHRP-6: 1.81 ±0.26, T saline: 1.52±0.15, T GHRP-6: 1.37±0.13 pg Prl/pituitary, n=6-8, n.s.);(Figure 9.9). However, it is interesting to note that in rGHRH-hGHS-R transgenic mice GHRP-6 treatment seems to alter pituitary GH and Prl levels in the opposite direction.

162 Chapter 9; GHS treatment studies in transgenic mice and rats

2.5-

2 - 1 1 (0 1.5- .t;a

1 - O)

0.5-

0 - rOHRH- rGHRH- WT WT hGHS-R hGHS-R saline GHRP-6 saline GHRP-6

n= 6 7 8 7

Figure 9.9 Pituitary Prl content of GHRP-6 treated (2x15|dg sc per day) adult male rGHRH-hGHS-R and WT mice Pituitary Prl contents were assayed in pituitary homogenates from rGHRH-hGHS-R and WT mice by RIA. No significant difference between treatment groups was found. (n=6- 8)

At the end of the study the tibiae of all mice were removed, decalcified in 10% EDTA for 2 weeks, paraffin embedded and cut into 6jim sections. The sections were stained with Masson’s trichrome and the width of the proliferative zone of the epiphysial growth plate was measured under the microscope using the NIH Image program. Although GHRP-6 treated animals had slightly wider growth plates, no significant difference was observed between any groups (WT saline: 102.1±5.5, WT GHRP-6: 110.4±7.2 pm, T saline: 107.3±3.9, T GHRP-6: 113.9±5.2 pm, n=6-8, n.s.); (Figure 9.10). The soleus muscle weight did not show any significant differences either (Figure 9.10), suggesting that the enhanced body weight gain in GHRP-6 treated mice must result from an overall increased weight of other tissues or perhaps water.

163 Chapter 9; GHS treatment studies in transgenic mice and rats

25 - 150 -

20 - O) £ 1 0 0 - & 15 I £O)

I I 10 - a (0 1 50 - i 8 O)2 5 -

0 - 0 - rOHRH- rGHRH- rGHRH- rGHRH- WT WT WT WT hGHS-R hGHS-R hGHS-R hGHS-R

saline GHRP-6 saline GHRP-6 saline GHRP-6 saline GHRP-6

n= 8 7 n= 6 7 8 7

Figure 9.10 Epiphysial growth plate width and soleus weight in GHRP-6 treated (2x15|ig s.c. per day) adult male rGHRH-hGHS-R and WT mice Neither the width of the proliferative zone of the growth plate nor the soleus weight showed any significant differences between groups (n=6-8).

9.2.4 Discussion With the experiments described above, I aimed to characterise further the effects of longer term exposure to GHSs in mice, which has not been described before. Two different routes of GHRP-6 administration in mice were chosen: either once daily i.p. or twice daily s.c.. The daily doses in the two experiments were comparable (lp.g/g/day) considering the differing body weight between male and female mice. In the first study, only rGHRH-hGHS-R transgenic mice significantly increased their body weight gain in response to once daily i.p GHRP-6 compared to saline treated rGHRH-hGHS-R mice. This enhanced weight gain in response to GHRP-6, however, was not significantly different from treated WT mice, which also showed a slight, but not significant increase in weight. The weight gain response to GHRP-6 was much more pronounced, when the drug was given twice daily s.c.; both, male GHRP-6 treated WT and transgenic mice, gained significantly more weight than their saline treated littermates. Although the rGHRH-hGHS-R transgenic mice again showed slightly enhanced growth compared to their treated WT littermates, this was not significant.

164 Chapter 9: GHS treatment studies in transgenic mice and rats

With these experiments I confirmed in mice what has been reported for rats; administration of GHRP-6 twice daily by s.c. injection is by far the more effective route (McDowell et al., 1995). Once daily i.p. injections of GHRP-6 might be too infrequent to generate a cumulative effect between injections. Despite their enhanced GH responses to acute GHRP-6 i.v. injections, neither male nor female rGHRH-hGHS-R transgenic mice showed enhanced weight gain in response to longer term s.c. or i.p. GHRP-6 treatment compared to treated WT littermates. However, we have no information in these experiments, whether acute s.c. or i.p. injections caused enhanced GH release in these mice. Although both male and female rGHRH-hGHS-R transgenic mice did respond to GHRP-6 with an increased body weight gain compared to saline treated transgenic animals, over-expression of the GHS-R in GHRH neurones did not seem to cause additional body weight gain compared to treated WT animals. In these studies the hypothesis that female Tgr rats might have responded better to GHS treatment due to enhanced hypothalamic GHS-R expression cannot be confirmed; at least not for enhanced ARC GHS-R expression targeted to GHRH neurones. It seems thus likely that the enhanced body weight gain in response to GHS treatment in female Tgr rats is a result of their GH deficiency. One has to bear in mind that in the first study (once daily i.p.) females were used, while the second study was performed with male mice. The differences seen in weight gain might thus not only be due to the injection schedule but might also be influenced by a sexual dimorphism in GHS responses. The increased pituitary GH content in female GHRP-6 treated rGHRH-hGHS-R transgenic mice might suggest that enhanced GHS-R signalling in the hypothalamus enhanced GHRH expression and secretion, which in turn had a trophic effect on the pituitary somatotroph (Barinaga et al, 1985a,b; Bilezikjian and Vale, 1984; Fukata et al, 1985; Gick et al, 1984; Tanner et al, 1990; Billestrup et al, 1987; Billestrup et al, 1986). GHRP-6 treatment in males however, caused more pronounced weight enhancement, but the pituitary GH levels were slightly, but not significantly up- regulated further in treated rGHRH-hGHS-R transgenic mice from the already enhanced pituitary GH levels in control male transgenic mice. In the second study, it seemed that the slight pituitary GH content increase seen in response to GHS treatment was at the

165 Chapter 9: GHS treatment studies in transgenic mice and rats expense of the pituitary Prl content. Whether this reflects a secondary effect of GHSs on mammosomatotroph differentiation is unclear and could be studied by investigating pituitary morphology. A study in beagle dogs had previously shown that after immobilisation of a hind limb, orally L-163,255 treated dogs recovered muscle strength significantly faster than untreated dogs (Lieber et al, 1997), suggesting GHS treatment had an effect on lean body mass. In neither of these studies significant differences in soleus muscle weight were noted. The increased body weight gain observed in GHS treated mice thus might result from weight changes of other muscle groups, organs or tissues. Svensson et al reported recently that continuous s.c. GHS infusion for 3 months significantly increased body weight gain and growth of the bones with increased bone dimensions (Svensson et al, 2000). Similarly Wells et al infused Tgr rats for 7 days i.v. with pulsatile GHRP-6 and found that this significantly increased their body weight gain and epiphysial plate width (Wells et al, 2000). In these studies no difference in the proliferative zone of the epiphysial growth plate was noted in response to GHS treatment. However the growth of 11-13 week old mice has ‘plateaued’ and therefore the proliferative zone of the growth plate is very small and irregular, making accurate measurements of minor changes difficult. Other measurements of growth such as bone dimensions, mineral content and density and ash weight might be more accurate and appropriate to analyse the effect of GHS treatment on bone growth. In this GHS study (twice daily s.c. GHRP-6) I therefore cannot say with certainty whether GHS treatment affected bone growth in these animals. As mentioned earlier, in view of the recent Ghrelin study by Tschop et «/., the investigation of an effect of GHS treatment on other parameters such as respiratory quotient, energy expenditure and food intake might be interesting, especially since in the first study (GHRP-6 once daily i.p.) ovarian fat pads seemed slightly increased in GHRP-6 treated animals. All in all the mechanism by which GHSs induce weight gain in mice and rats is still unclear. The studies with rGHRH-hGHS-R transgenic mice have shown that over­ expression of the GHS-R in hypothalamic GHRH neurones did not enhance increased weight gain in response to GHRP-6. Recent studies suggest that the influence GHSs have on body weight gain through skeletal growth or food intake are administration

166 Chapter 9: GHS treatment studies in transgenic mice and rats pattern and GHS dependent (Svensson et al, 2000; Wells et al, 2000; Kuriyama et al, 2000; Locke et al, 1995; Okada et al, 1996; Tschop et al, 2000) (D.F. Carmignac, unpublished data). Whether these GHS effects are all directly mediated through the GHS-R and which hypothalamic targets they are mediated by remains unclear.

167 Chapter 10: Final discussion

10 Final discussion 10.1 Hitting the spot - Generation of targeting constructs In the studies I described in this thesis, transgenes were targeted to two key cell types of the GH axis - the hypothalamic GHRH neurones and the anterior pituitary somatotrophs. Both these cell types have been targeted by transgenesis before. hGH-N, directed by its proximal promoter or by additional 4.6kb of 5’ and 26kb of 3’ flanking sequences, was either not expressed at all or only poorly so in transgenic pituitaries (Hanuner et al, 1984; Palmiter and Brinster, 1986). However, a 40kb hGH-N cosmid was shown to comprise all regulatory sequences for appropriate transgene expression restricted to anterior pituitary GH cells and to contain the full hGH LCR (Bennani-Baiti et al, 1998; Jones et al, 1995). These data highlighted the importance of including a full LCR in transgene constructs for high-level, tissue-specific transgene expression irrespective of the chromosomal site of transgene integration. Previous attempts to target GHRH neurones in transgenic mice used very short flanking sequences (900bp 5’UTR) of the human (Botteri et al, 1987) or mouse (Giraldi et al, 1994) GHRH genes expressing SV40 large T antigen. No hypothalamic expression was obtained, though ectopic expression produced thymic hyperplasia or adrenal medullary tumours. However, Flavell et a l had demonstrated appropriate transgene expression in ARC neurones by using a 38kb rat GHRH genomic cosmid construct (Flavell et al, 1996), but no studies were performed to analyse whether this cosmid contained the full GHRH LCR. Here I chose to use the largest clones available to maximise the chances of appropriate tissue-specific expression. The 40kb hGH-N cosmid and the 38kb rGHRH cosmid were used to place transgenes under GH or GHRH promoter control and two different transgenes - eGFP and hGHS-R - were expressed by each promoter. I tested transgene expression under rGHRH cosmid control using the fluorescent marker eGFP in transgenic mice, showing that the rGHRH cosmid did successfully target GHRH neurones. However, one of three transgenic mouse lines showed inappropriate hypothalamic eGFP expression, and some transgenic mice of all rGHRH-eGFP transgenic founders demonstrated eGFP expression in the kidney, suggesting the 38kb rGHRH cosmid did not behave as a full GHRH LCR. Nevertheless, in both rGHRH-

168 Chapter 10: Final discussion

eGFP (founder ‘39’) and rGHRH-hGHS-R transgenic mice, the transgenes were specifically expressed in hypothalamic ARC neurones in a GHRH-like pattern. Co­ localisation of the GHRH and eGFP mRNA in the same ARC neurones demonstrated that the rGHRH cosmid is indeed capable of targeting transgenes to hypothalamic GHRH neurones, but this may vary in fidelity from line to line. The rGHRH cosmid is thus a very useful tool when targeting transgenes to GHRH neurones, although the resulting transgenic animals do then require careful expression analysis. The fidelity of the rGHRH cosmid may also vary between species, such that reliable transcription of the rGHRH cosmid might be more efficient in transgenic rats, than mice. As described in the introduction, the GHRH gene is expressed in a number of tissues and it has been shown that two different first exons are used for placental and hypothalamic GHRH expression (Gonzalez-Crespo and Boronat, 1991). Furthermore, it has been demonstrated that the GHRH gene produced another distinct peptide with tissue-specific activities (Breyer et al, 1996). These data and my expression analysis suggest that the GHRH promoter may be more complicated containing more upstream or downstream regulatory sequences than were included in the 38kb rGHRH cosmid. In hGH-eGFP transgenic mice the hGH cosmid was shown, once more, to target transgene expression appropriately and exclusively to GH cells in the anterior pituitary. My expression studies in hGH-hGHS-R transgenic mice indicate that transgenic mice over-expressing the hGHS-R under the same promoter control showed the same expression pattern as hGH-eGFP mice. Further analysis of transgene copy numbers in hGH-hGHS-R transgenic mice with different levels of severity in the GH physiology phenotype might underline the copy-number dependent transgene expression driven by the hGH promoter. Comparison of transgene expression in GHRH neurones or GH cells suggested that the intensity of transgene expression was in keeping with endogenous GHRH or GH expression levels and cell numbers. My studies also highlight that the inclusion of differing amounts of hGH sequences targeted transgene expression to different structures of the GHRH neurone and GH cell. The eGFP transgene was shown in both cases to be targeted to secretory vesicles by fusion with the hGH signal peptide, while this was avoided for the hGHS-R by fusing it directly to hGH 5’ and 3’ UTR. As described previously by Flavell et al, my studies confirmed that the addition of some hGH 5’ and 3’ UTR sequences to the transgene

169 Chapter 10: Final discussion

coding region can insure efficient and stable transcription of transgenes (Flavell et al, 1996). Manipulation of large cosmid constructs is usually difficult, because one is dependent on the availability of appropriate restriction sites. For both the rGHRH and the hGH cosmid, it has proven to be a major advantage that the targeting cosmids were engineered to have a unique restriction site where the transgene is inserted into the cosmids. The same targeting cosmids can then be used for various transgenes, which can easily be introduced into these cosmids by use of the unique Mlul cloning sites. In general, when handling large colonies of several different strains of transgenic mice, a fast and reliable method for genotyping these large amounts of animals is vital. The genotyping used for all transgenic mice described in this thesis, with two or three primers to amplify an endogenous band together with a transgene specific band, proved extremely reliable and time-saving. However, with this PCR method the stability of the transgene from generation to generation cannot be verified. Therefore the transgene hybridisation pattern of genomic DNA from transgenic animals of different generations digested with frequently cutting restriction enzymes should be tested regularly.

10.2 GHS Physiology in hGHS-R over-expressing transgenic mice 10.2.1 hGHS-R over-expression in GHRH neurones In vitro experiments with a GH cell line expressing the hGHS-R demonstrated that over­ expression of the GHS-R caused increased GHS-R signalling, even in the absence of GHSs. Similarly, over-expression of another G protein coupled receptor, the p2" adrenergic receptor in the myocardium of transgenic mice, had resulted in enhanced activation of myocardial signalling and function in the absence of agonist (Koch et al, 2000). Over-expression of the GHS-R at high enough levels might thus raise its constitutive activity. Over-expression of the hGHS-R in hypothalamic GHRH neurones in rGHRH-hGHS-R transgenic mice indeed caused mild but specific alteration of the GH axis in the absence of exogenous GHSs. Increased pituitary GH levels and body weight suggested a hypothalamic trophic effect, possibly by GHRH. Indeed, some data in this study suggested that hypothalamic GHRH mRNA levels in rGHRH-hGHS-R transgenic mice might be up-regulated. Endogenous GHRH expression levels in WT mice are relatively low, such that over-expression of the GHS-R was only approximately

170 Chapter 10: Final discussion

6-fold in rGHRH-hGHS-R transgenic mice compared to endogenous mGHS-R expression levels. Still higher levels of GHS-R over-expression, for example in mice with higher copy numbers, might show more severe alterations of the GH axis in the absence of ligand. However this could possibly be limited by GH feed-back on the rGHRH promoter. I further investigated the phenotype of rGHRH-hGHS-R transgenic mice in the presence of exogenous GHS-R ligands, the GHSs. Over-expression of the hGHS-R in hypothalamic GHRH neurones seemed to enhance responsiveness to acute i.v. injections of GHRP-6 in vivo, while the sensitivity was unchanged. A mechanism for this action remains speculative, but taking the in vitro data into account, the hypothalamic GHRH neurone PI turnover in response to GHSs might be enhanced. Increased signalling through this pathway might enhance GHRH expression and release, which at the pituitary level causes enhanced GH release. However, other factors like the electrophysiological activity of the GHRH neurone or involvement of the zif 268 pathway (activated in response to GHS, personal communication S.Lall, Cambridge) might play a role in this mechanism. These data support the view that GHRH neurones are an important target for GHS actions on GH release and alteration of GHS-R abundance only in these neurones seemed to alter GHS responsiveness. This is the first demonstration that the hypothalamic GHS-R in GHRH neurones might have a direct influence on GHS mediated GH release. However, measurement of plasma GH concentrations is a relatively indirect way of investigating effects of hypothalamic GHS-R over-expression on GHS physiology. With this method it remains unclear whether GHS-R over-expression has direct effects on the physiology of GHRH neurones. In electrophysiological experiments, GHSs have been shown to enhance the electrical activity of ARC neurones (Dickson et al, 1993b). Similar experiments could be used to investigate whether over-expression of the GHS-R in GHRH neurones alters the electrophysiological response of single neurones to GHSs, or whether GHSs might cause co-ordinated firing of several neurones to create a GH pulse. The immediate-early gene c-fos can be used as a marker for neuronal activation, such that measurements of Fos immunoreactive neurones could be used to analyse GHS responsiveness in rGHRH-hGHS-R transgenic mice. However, Fos immunoreactivity is difficult to quantify in single neurones, so Fos analysis in WT and rGHRH-hGHS-R

171 Chapter 10: Final discussion transgenic mice might show the same number of GHS responsive neurones but differences in the magnitude of responses between neurones would be more difficult to detect. In order to measure differences in gene transcription in response to GHSs between rGHRH-hGHS-R and WT animals, mice could be injected a few times a day with GHSs and hypothalamic sections used for in situ hybridisation measurements of e.g. GHRH, c-fos, zif 268 (personal communication S.Lall, Cambridge) as well as GHS- R mRNA expression levels. Furthermore, direct measurement of intracellular Ca^"^ levels in response to GHSs would be very useful. By crossing the rGHRH-hGHS-R transgenic mice onto the rGHRH-eGFP transgenic mice, alterations of Ca^^ levels could be measured in pre-identified GHS-R over-expressing GHRH neurones, either in live slice preparations or in ARC cultures (see chapter 10.3). The rGHRH-hGHS-R transgenic mice provide an interesting system for analysing the ACTH and Prl releasing effect of GHSs. Whilst the Prl releasing effect of GHSs may also involve direct actions on the pituitary (Carmignac et al, 1998) it is thought that GHSs’ ability to release ACTH reflects a hypothalamic site of action (Hickey et al, 1996) possibly involving CRH and/or arginine vasopressin (Korbonits et al., 1999b; Thomas et ai, 1997). GHRH-R deficiency in humans showed that GH responses to GHSs were absent, but ACTH and Prl responses to this treatment were present and probably normal (Maheshwari et at., 1999), suggesting that the GHRH-R system might not be involved or crucial in these responses. It would therefore be interesting to compare Prl and ACTH release in response to GHSs in transgenic mice that exhibit hypothalamic GHS-R over-expression confined to GHRH neurones.

When analysing the physiological phenotype of rGHRH-hGHS-R transgenic mice, the endogenous GHS-R ligand Ghrelin has to be considered. Preliminary studies suggested that there was no difference in stomach Ghrelin expression between rGHRH-hGHS-R transgenic and WT mice. Ghrelin expression was not detectable in the hypothalamus by RNAse protection analysis or in situ hybridisation, therefore no further studies to quantify Ghrelin expression in WT and transgenic mice were carried out in the brain. At the moment it is unclear whether GHS-R over-expression in GHRH neurones has any influence on Ghrelin signalling. Also, the exact role of CNS vs. stomach Ghrelin in GH release or other effects remains unknown.

172 Chapter 10: Final discussion

10.2.2 GHSs - GH V5. metabolic effects Studies of growth parameters in female rGHRH-hGHS-R transgenic mice suggested that hypothalamic GHS-R over-expression specifically enhanced nose-anus and tibial length. Although not entirely conclusive, the ovarian fat pad weights in rGHRH-hGHS-R transgenic mice were slightly, but not significantly up-regulated. NPY expression was not significantly altered in rGHRH-hGHS-R transgenic compared to WT mice. Some GHSs and Ghrelin treatment both enhance food intake (Kuriyama et al., 2000; Locke et ai, 1995; Okada et ai, 1996; Tschop et ai, 2000) and Ghrelin was shown to enhance fat mass, without a change in lean body mass and bone mass (Tschop et al, 2000). Similarly 14-day i.c.v. GHRP-6 treatment in dw/dw rats caused enhanced body weight gain with increased food intake (personal communication D.F. Carmignac). However, it is unclear whether the metabolic effects of GHSs and Ghrelin are directly regulated through the GHS-R. The rGHRH-hGHS-R transgenic mice are an interesting animal model to investigate the influence of hypothalamic GHS-R over-expression targeted to GHRH neurones on metabolic parameters particularly in response to GHSs and Ghrelin treatment. In general, the artificial GHSs and Ghrelin have several effects in common (see Chapter 8), one of these being the dose-dependent GH releasing effects. However, neither the GHSs nor Ghrelin are specific for these GH releasing effects and even among the artificial GHSs a variety of effects apart from the GH release distinguishes them. For example, i.c.v. treatment of rats with GHRP-6 and KP-102 significantly increases their food intake (Kuriyama et al, 2000; Locke et al, 1995; Okada et al, 1996), while this was not the case for L-163,255 (personal communication D.F. Carmignac). Not only are GHSs’ effects on skeletal growth or metabolism dependent on the GHS used, but also the pattern by which it is administered (Svensson et al, 2000; Tschop et al, 2000; Wells et al, 2000). Ghrelin was shown to have an effect on gastric acid secretion which was abolished by vagotomy (Masuda et al, 2000), an effect that has never been reported for GHSs. Furthermore, Hexarelin, a GHRP-6 analogue, was shown to bind to receptors in the heart and have specific effects on the heart (Bisi et al, 1999; Bodart et al, 1999; Locatelli et al, 1999; Rossoni et al, 2000; Tivesten et al, 2000), but binding of Hexarelin could not be displaced by MK-0677 and MK-0677 also had no specific effect on the heart (Bodart et al, 1999). Research to date suggests that the GHSs act through

173 Chapter 10: Final discussion the GHS-R to release GH, but that GHSs also have a variety of other effects, which may stem from activation of other subtypes of GHS-R. GHS-R knock-out mice will be very useful to analyse which effects of GHSs and Ghrelin remain, after GHS-R signalling has been abolished.

10.2.3Longer-term GHS treatment In studies with the orally active GHS L-163,255 I analysed the effect of oral GHS treatment in GH deficient, but GHS responsive transgenic ‘Tgr’ rats on body weight gain and showed that only the female Tgr rats responded to GHS treatment with enhanced body weight gain. These data were in agreement with previous findings that female rats respond to GHS treatment more consistently than males (Bercu et al, 1991; Sartor et al, 1985). Other data from GH deficient animals implied that increased GHS-R mRNA expression and the GH depletion might contribute to enhanced responsiveness to GHS treatment (Bennett et al, 1997; Carmignac et al, 1998; Wells et al, 1997). In studies with acute i.v. injections of GHRP-6 I suggested that enhanced hypothalamic GHS-R levels might be responsible for more consistent GHRP-6 responses. The question was thus whether, by longer GHS treatment of mice with increased abundance of hypothalamic GHS-R, an enhanced GH release could be sustained to cause enhanced growth. This is the first report of GHS treatment (2x daily s.c.) in mice leading to enhanced body weight gain in both WT and rGHRH-hGHS-R transgenic mice compared to their saline treated littermates. However, independent of the injection schedule there was no difference in body weight gain in response to GHSs in either male or female rGHRH- hGHS-R transgenic mice compared to treated WT littermates. Although GH responses to acute GHS i.v. injection seemed enhanced in rGHRH-hGHS-R transgenic mice, increased GHS-R abundance in hypothalamic GHRH neurones did not enhance body weight gain further in response to GHSs. However, currently we have no information on the GH release after GHS s.c. injections in rGHRH-hGHS-R transgenic animals. One explanation for the absence of an enhanced growth response might be that the GH response to s.c. GHS is simply not elevated in rGHRH-hGHS-R transgenic mice. When several doses of GHRP-6 were given i.v., we saw that GHS responsiveness seemed enhanced only at higher GHS doses, so perhaps the s.c. GHS concentration in the

174 Chapter 10: Final discussion

longer-term GHS studies was too low to cause enhanced GH release. On the other hand, if the GH release in response to s.c. GHS was elevated, possibly this GH release might not be sufficient to stimulate further body weight gain in rGHRH-hGHS-R mice or an increased GH release in response to GHSs might decline to normal after several GHS injections. A mechanism remains speculative until the GH release profile after GHS s.c. injections in these mice becomes apparent. However, perhaps an initially enhanced GH response to GHSs fades, because enhanced GH and/or IGF-I levels in response to the treatment might feed-back onto the GHRH promoter that drives the GHS-R over­ expression resulting in down-regulated GHS-R expression. By this means enhanced responses to GHSs would result in self-suppression of the GHS-R in rGHRH-hGHS-R transgenic mice. Previously in humans long-term GHS treatment was shown to enhance pulsatile GH secretion and raise plasma IGF-I concentrations (Huhn et al, 1993; Jaffe et al, 1993), but GH responses to a subsequent GHRP-6 challenge were then attenuated. Furthermore, Hickey et al suggested that a reduced amplitude of the GH response to pro-longed MK-0677 exposure in beagle dogs might stem from enhanced IGF-I levels (Hickey et al, 1997). In addition, Bailey et al have recently reported that chronic central infusion of GHSs caused absence of ARC c-fos expression and lack of GH responses to a subsequent systemic GHS injection (Bailey et al, 1999a). The authors suggested that this inactivation of ARC neurones to a GHS signal might be indirect and due to enhanced GH feed-back onto PEN SS neurones, which in turn suppress ARC neurones. The rGHRH-hGHS-R transgenic mice are therefore also an interesting model to analyse GHS desensitisation. Continuous infusions of GHRP-6 have been shown to cause desensitisation (McDowell et al, 1995) and this might be due to the GHS-R being in a constant state of phosphorylation. I found that GHS treatment twice daily s.c. was more effective in causing enhanced body weight gain than once daily i.p. injections. Maybe twice a day s.c. GHS treatment gives the GHS-R enough time between injections to dephosphorylate, while once daily GHS i.p. injections might be ineffective because the time between doses is too long to achieve a cumulative effect. Using the continuous s.c. infusion schedule, which is known to cause desensitisation to GHSs in rats (McDowell et al, 1995) one could investigate whether hypothalamic GHS-R over-expression can protect rGHRH-hGHS-R transgenic mice from desensitisation.

175 Chapter 10: Final discussion

10.2.4 hGHS-R over-expression in GH cells hGH-hGHS-R transgenic mice were created with 50-fold over-expression of the hGHS- R in anterior pituitary cells. The GHS-R in vitro studies and the results from the rGHRH-hGHS-R transgenic mice had suggested that enhanced GHS-R expression in GH cells might lead to enhanced GHS responsiveness in hGH-hGHS-R transgenic mice. However, GHSs alone do not show a trophic effect on the pituitary (Soto et al, 1995; Torsello et al, 1996) and over-expression of the GHS-R at the pituitary level alone might thus not result in elevated GHS responsiveness. Especially in female hGH-hGHS-R transgenic mice it became obvious that GHS-R over­ expression in the anterior pituitary in fact caused almost the opposite effect observed in rGHRH-hGHS-R transgenic mice. Studies on the GH physiology of these mice are still preliminary and discussion of these results remains speculative. At the moment it is unclear how GHS-R over-expression mediates reduced pituitary GH content in hGH- hGHS-R transgenic mice. In vivo and in vitro experiments analysing GHS and GHRH responsiveness of these animals suggested that GHS-R over-expression at such high levels might disrupt somatotrophs, possibly by depleting G-protein dependent intracellular signalling cascades. If this was indeed the case, reduced GHRH signalling would result in the loss of the trophic GHRH effect, causing down-regulated GH transcription and synthesis (Barinaga et al, 1985a; Bilezikjian and Vale, 1984; Fukata et al, 1985; Gick et al, 1984; Tanner et al, 1990). Since a lack of stimulation of Pit-1 by GHRH might also influence GHRH-R abundance (Lin et al, 1992), analysis of the GHRH-R mRNA levels in hGH-hGHS-R somatotrophs might indicate whether a loss of GHRH signalling is indeed responsible for the GH deficiency. At the moment it is unclear whether the inhibition is specific to GHRH signalling, or whether other somatotroph responses in hGH-hGHS-R transgenic mice are compromised as well.

These data might suggest dependence of the GHRH-R signalling cascade on an intact GHS-R system, as has been suggested vice versa, but to me it seems more likely that the reduced GHRH effect on GH release stems from a GH deficient pituitary. Since Prl was unaffected, the effect of GHS-R over-expression seems to be somatotroph or GH gene specific. Thorough analysis of somatotroph morphology and number will give further

176 Chapter 10: Final discussion

insight into what causes the GH deficiency. As described previously, analysis of several signalling cascades could also help to understand why the somatotroph seems unable to cope with high levels of GHS-R over-expression. If the GH physiology is confirmed to be pathological as a result of high levels of GHS-R over-expression at the pituitary level, it might be better in future to generate mice with less GHS-R over-expression to avoid desensitisation of the GHS/GHRH system. A true effect of moderate pituitary GHS-R up-regulation might thus be over-shadowed by a GH deficient somatotroph in my hGH-hGHS-R transgenic mice. Korbonits et al. have suggested previously that elevated pituitary GHS-R expression levels in somatotropinomas from acromegalic patients might account for the enhanced GH discharge in response to GHSs in some acromegalic patients (Korbonits et al, 1998). The authors measured the GHS-R/GADPH ratio by RT-PCR and found a 2-10 fold over-expression of the GHS-R in somatotropinomas from acromegalic patients (Korbonits et al, 1998). Quantification of mRNA levels by RT-PCR, especially at such low levels as the pituitary GHS-R, is somewhat problematic. However, if these are indeed the levels of GHS-R over-expression seen in combination with enhanced GHS responsiveness, then maybe it is these moderate levels one should aim for when generating transgenic mice with less pronounced GHS-R over-expression than my hGH- hGHS-R transgenic mice.

10.2.5 Hypothalamic vs pituitary GHS-R over-expression One of the main aims of this study was to create a novel way of distinguishing hypothalamic and pituitary GHS actions by enhancing GHS-R expression either in hypothalamic GHRH neurones or pituitary GH cells in transgenic mice, followed by analysis of differential effects of hypothalamic or pituitary GHS-R over-expression on the GH and GHS axis. I succeeded in generating both mice over-expressing the GHS-R in the hypothalamus or the pituitary with over-expression in these tissues resulting in a very different effect on GH and GHS physiology. However, a direct comparison of the effects on the GHS axis in rGHRH-hGHS-R and hGH-hGHS-R transgenic mice is not straightforward. The very different levels of GHS-R over-expression in the hypothalamus and the pituitary make a comparison of the hypothalamic and pituitary GHS effect very difficult. If in hGH-hGHS-R transgenic mice pituitary GH deficiency is

177 Chapter 10: Final discussion a result of a ‘pathological’ somatotroph, a 50-fold over-expression of the GHS-R in hypothalamic GHRH neurones might potentially have the same effect of disrupting ‘normal’ GHRH neurone physiology. On the other hand my studies could highlight an essential difference in hypothalamic and pituitary GHS signalling. While in the hypothalamus enhanced GHS-R might have a direct effect to enhance GH release in response to GHSs, at the pituitary level, the signalling of the GHS-R is much complicated by its interaction with the GHRH-R. If GHS-R over-expression specifically caused inhibition of the GHS-R and GHRH pathway to regulate GH synthesis and secretion, this would highlight how essential intact cross-talk between the two receptors at the pituitary level was for an appropriately regulated GH system. Further studies with these rGHRH-hGHS-R and hGH-hGHS-R transgenic mice will elucidate important mechanisms of GHS signalling in the hypothalamus and the pituitary.

10.2.6 Knocking out the background As mentioned earlier, a long-term aim of the lab following on from the animals I produced is to cross these tissue-specific GHS-R over-expressing transgenic mice with GHS-R KO mice. This cross would enable the study of GHS-R over-expression restricted to hypothalamic GHRH neurones or pituitary GH cells with no endogenous pituitary GHS-R signalling. Thereby the pituitary and the hypothalamic GHS-R signalling pathways would be entirely separated. This would further elucidate, whether the hypothalamic and pituitary GHS signalling pathway are dependent on each other or whether GHSs can exert their GH secreting effect solely through hypothalamic or pituitary actions. However, current data from lit/lit mice and patients with GHRH-R mutations suggest that a loss of cross-talk between the pituitary GHS-R and GHRH-R, and thus a resulting loss of pituitary GHS signalling,would selectively eliminate the GH releasing effect of GHSs (Jansson et al, 1986b; Maheshwari et al, 1999). Whether elimination of the pituitary GHS-R instead of the GHRH-R would have a similar effect is unclear. Additionally it was shown that separation of the pituitary signal from the hypothalamic GHS pathway for example in stalk lesioned pigs and humans, greatly reduced the GH releasing activity of GHSs (Hayashi et al, 1993; Hickeyet al, 1996;

178 Chapter 10: Final discussion

Loche et al, 1995; Popovic et al, 1995). Another interesting experiment would be to cross both transgenic GHS-R over­ expressing mice onto the GHS-R KO together to analyse, whether in these mice with expression of the GHS-R restricted to pituitary GH cells and hypothalamic GHRH neurones only the GH axis was affected, or whether metabolic and ACTH/Prl releasing effects of GHSs were altered, too. I have already started to cross my rGHRH-hGHS-R transgenic mice onto GHS-R KG mice (kindly provided by Dr. R.G. Smith, Houston, Texas), but no data on their physiology are available as yet.

10.3 eGFP and hGHS-R expression in the same cell rGHRH-eGFP transgenic mice were not only useful to test the rGHRH cosmid, but are very practical animals in themselves and in combination with other transgenic mice, since they overcome the problem of not being able to identify neurones in vivo. I showed that eGFP neurones are readily identifiable in hypothalamic live slices and in hypothalamic cultures of rGHRH-eGFP transgenic mice making these neurones accessible to various techniques and studies to analyse their cell physiology in pre­ identified neurones. For example, until now electrophysiological studies of hypothalamic neurones required neuropeptide-characterisation of the recorded neurone after the experiment. Expression of eGFP in GHRH neurones will now facilitate electrophysiological studies in pre­ identified neurones immensely, as has been elegantly shown for GnRH-GFP transgenic mice (Spergel et al, 1999; Suter et al, 2000). rGHRH-eGFP transgenic mice will also facilitate the study of spontaneous [Ca^^]j transients in cell bodies and nerve terminals of pre-identified neurones by dual wavelength imaging. This will be particularly interesting for GHS physiology, since it is still unclear, how GHS binding to their receptors in hypothalamic neurones causes GHRH release into hypophyseal blood. If eGFP and GHRH co-localisation in the same granule in GHRH nerve terminals can be proven, rGHRH-eGFP transgenic mice will be an ideal tool to study GHRH packaging and secretion in live slices. To date, not much is known about the process of GHRH packaging and secretion. Technically more challenging would be the intracellular electrophysiological recording from single eGFP identified GHRH neurones in vivo.

179 Chapter 10: Final discussion using a ventral surgical technique to expose the ME and the base of the ARC (Dickson eta l, 1993b). It might be possible to use eGFP protein levels as a surrogate measurement for GHRH protein status, enabling us to monitor GHRH gene expression in vivo. However, there are considerable technical problems with this, concerning the lag-time of eGFP folding to fluorescence and its photobleaching. Similar techniques can be applied in hGH-eGFP transgenic mice. In addition, an interesting component is the analysis of whole pituitaries to study co-ordinated actions of many GH cells. First studies in this direction have been described by Magoulas et al., where dual wavelength imaging indicated that mouse GH cells display spontaneous rhythmic bursts of intracellular Ca^^ concentrations (Magoulas et at., 2000). This technique now enables the simultaneous measurement of Ca^^ transients in multiple GH cells. My production and analysis of eGFP and hGHS-R expression in transgenic mice now enables the crossing of these animals. Both transgenic eGFP expressing mice would be particularly interesting in combination with their hGHS-R over-expressing counterpart. Crossing of eGFP transgenic mice with hGHS-R transgenic mice would lead to eGFP and hGHS-R expression in the same cell, since both transgenes are expressed off the same promoter. Here the targeting of eGFP and the hGHS-R to different cell structures comes into play. While eGFP mimics GHRH or GH packaging and secretion, the hGHS- R is over-expressed and presumably targeted to the plasma membrane of the cell to allow appropriate cell signalling. Transgenic mice expressing both eGFP and the hGHS-R in the same cell would be very valuable to study GHS physiology in these cells in vivo. Previously recordings from putative GHRH neurones in the ARC had demonstrated that GHSs stimulated the firing of neurones in this area (Dickson et at., 1993b). In eGFP/GHS-R expressing transgenic mice one could analyse by electrophysiology in eGFP pre-identified cells whether the excitability in response to GHSs of cells over-expressing the hGHS-R was altered compared to their WT littermates. In addition, these cells could be characterised by single cell RT-PCR, to analyse the expression of channels, receptors and other genes. Furthermore the effect of GHSs on GHRH and GH processing and secretion could be analysed by eGFP tracking in hGHS-R over-expressing and WT cells in live slice

180 Chapter 10: Final discussion preparations.

The combination of transgenic studies using eGFP as a fluorescent marker and the hGHS-R to analyse GHS physiology has proven very useful. First eGFP was used for further studies of the expression pattern of transgenes driven by the rGHRH and hGH promoter and these transgenic mice can now be used to analyse ‘normal’ GHRH or GH cell physiology. Then over-expression of the GHS-R was targeted to GHRH neurones and GH cells and the effects on GH and GHS physiology were analysed in transgenic mice. A combination of the two transgenic mice will now allow the analysis of effects of GHS-R over-expression directly at the pre-identified target cell.

181 References

References

Abe, H., Molitch, M. E., Van Wyk, J. J., and Underwood, L. E. (1983). Human growth hormone and somatomedin C suppress the spontaneous release of growth hormone in unanesthetized rats. Endocrinology 113, 1319-24.

Adams, E. P., Huang, B., Buchfelder, M., Howard, A., Smith, R. G., Feighner, S. D., van der Ploeg, L. H., Bowers, C. Y., and Fahlbusch, R. (1998). Presence of growth hormone secretagogue receptor messenger ribonucleic acid in human pituitary tumors and rat GH3 cells. J Clin Endocrinol Metab 83, 638-42.

Adams, E. P., Lei, T., Buchfelder, M., Bowers, C. Y., and Fahlbusch, R. (1996). Protein kinase C-dependent growth hormone releasing peptides stimulate cyclic adenosine 3',5'- monophosphate production by human pituitary somatotropinomas expressing gsp oncogenes: Evidence for cross-talk between transduction pathways. Molecular Endocrinology 10,432-438.

Adams, E. P., Petersen, B., Lei, T., Buchfelder, M., and Fahlbusch, R. (1995). The growth hormone secretagogue, L-692,429, induces phosphatidylinositol hydrolysis and hormone secretion by human pituitary tumors. Biochem Biophys Res Commun 208, 555-61.

Aguila, M. C., Snyder, G. D., and McCann, S. M. (1991). Rat growth hormone-releasing factor stimulates cyclic GMP formation and phosphatidylinositol metabolism in the median eminence. Life Sci 49, 67-74.

Ahima, R. S., Saper, C. B., Flier, J. S., and Elmquist, J. K. (2000). Leptin regulation of neuroendocrine systems. Front Neuroendocrinol 21, 263-307.

Akman, M. S., Girard, M., O'Brien, L. P., Ho, A. K., and Chik, C. L, (1993). Mechanisms of action of a second generation growth hormone-releasing peptide (Ala- His-D-beta Nal-Ala-Trp-D-Phe-Lys-NH2) in rat anterior pituitary cells. Endocrinology 132, 1286-91.

Aloi, J. A., Gertz, B. J., Hartman, M. L., Huhn, W. C., Pezzoli, S. S., Wittreich, J. M., Krupa, D. A., and Thomer, M. O. (1994). Neuroendocrine responses to a novel growth hormone secretagogue, L-692,429, in healthy older subjects. J Clin Endocrinol Metab 79, 943-9.

Alpert, L. C., Brawer, J. R., Patel, Y. C., and Reichlin, S. (1976). Somatostatinergic neurons in anterior hypothalamus: immunohistochemical localization. Endocrinology 98, 255-8.

Alster, D. K., Bowers, C. Y., Jaffe, C. A., Ho, P. J., and Barkan, A. L. (1993). The growth hormone (GH) response to GH-releasing peptide (His-DTrp-Ala-Trp-DPhe-Lys- NH2), GH-releasing hormone, and thyrotropin-releasing hormone in acromegaly. J Clin Endocrinol Metab 77, 842-5.

182 References

Argente, J., Chowen, J. A., Zeitler, P., Clifton, D. K., and Steiner, R. A. (1991). Sexual dimorphism of growth hormone-releasing hormone and somatostatin gene expression in the hypothalamus of the rat during development. Endocrinology 128, 2369-75.

Argetsinger, L. S., Campbell, G. S., Yang, X., Witthuhn, B. A., Silvennoinen, O., Ihle, J. N., and Carter-Su, C. (1993). Identification of JAK2 as a growth hormone receptor- associated tyrosine kinase. Cell 74, 237-44.

Argetsinger, L. S., and Carter-Su, C. (1996). Mechanism of signaling by growth hormone receptor. Physiol Rev 76, 1089-107.

Arvat, E., Di Vito, L., Broglio, P., Papotti, M., Muccioli, G., Dieguez, C., Casanueva, F. P., Deghenghi, R., Camanni, P., and Ghigo, E. (2000). Preliminary evidence that Ghrelin, the natural GH secretagogue (GHS)-receptor ligand, strongly stimulates GH secretion in humans. J Endocrinol Invest 23,493-5.

Bagnato, A., Moretti, C., Ohnishi, J., Prajese, G., and Catt, K. J. (1992). Expression of the growth hormone-releasing hormone gene and its peptide product in the rat ovary. Endocrinology 130, 1097-102.

Bailey, A. R., Giles, M., Brown, C. H., Bull, P. M., Macdonald, L. P., Smith, L. C., Smith, R. G., Leng, G., and Dickson, S. L. (1999a). Chronic central infusion of growth hormone secretagogues: effects on fos expression and peptide gene expression in the rat arcuate nucleus. Neuroendocrinology 70, 83-92.

Bailey, A. R., Honda, K., Smith, R. G., and Leng, G. (1999c). Growth hormone- releasing hormone and morphine attenuate growth hormone secretagogue-induced activation of the arcuate nucleus in the male rat. Neuroendocrinology 70, 101-6.

Bailey, A. R. T. (1999b). The role of growth hormone secretagogues in the brain (Abstract). In: Annual Meeting of the British Neuroendocrine Group (London, UK).

Barinaga, M., Bilezikjian, L. M., Vale, W. W., Rosenfeld, M. G., and Evans, R. M. (1985b). Hormonal regulation of gene expression: regulation of growth hormone gene transcription by growth hormone releasing factor. Oxf Surv Eukaryot Genes 2, 49-73.

Barinaga, M., Bilezikjian, L. M., Vale, W. W., Rosenfeld, M. G., and Evans, R. M. (1985a). Independent effects of growth hormone releasing factor on growth hormone release and gene transcription. Nature 314, 279-81.

Earlier, A., Zamora, A. J., Grino, M., Gunz, G., Pellegrini-Bouiller, I., Morange-Ramos, I., Pigarella-Branger, D., Dufour, H., Jaquet, P., and Enjalbert, A. (1999). Expression of functional growth hormone secretagogue receptors in human pituitary adenomas: polymerase chain reaction, triple in-situ hybridization and cell culture studies. J Neuroendocrinol 11,491-502.

183 References

Barsh, G. S., Seeburg, P. H., and Gelinas, R. E. (1983). The human growth hormone gene family: structure and evolution of the chromosomal locus. Nucleic Acids Res 11, 3939-58.

Bartlett, J. M., Charlton, H. M., Robinson, I. C., and Nieschlag, E. (1990). Pubertal development and testicular function in the male growth hormone-deficient rat. J Endocrinol 126, 193-201.

Baumann, G., and Maheshwari, H. (1997). The Dwarfs of Sindh: severe growth hormone (GH) deficiency caused by a mutation in the GH-releasing hormone receptor gene. Acta Paediatr Suppl 423, 33-8.

Baumann, G., Vance, M. L., Shaw, M. A., and Thomer, M. O. (1990). Plasma transport of human growth hormone in vivo. J Clin Endocrinol Metab 71,470-3.

Baumbach, W. R., Homer, D. L., and Logan, J. S. (1989). The growth hormone-binding protein in rat serum is an altematively spliced form of the rat growth hormone receptor. Genes Dev 3, 1199-205.

Beaudet, A., Greenspun, D., Raelson, J., and Tannenbaum, G. S. (1995). Pattems of expression of SSTRl and SSTR2 somatostatin receptor subtypes in the hypothalamus of the adult rat: relationship to neuroendocrine function. Neuroscience 65, 551-61.

Behringer, R. R., Mathews, L. S., Palmiter, R. D., and Brinster, R. L. (1988). Dwarf mice produced by genetic ablation of growth hormone-expressing cells. Genes Dev 2, 453-61.

Bennani-Baiti, I. M., Asa, S. L., Song, D., Iratni, R., Liebhaber, S. A., and Cooke, N. E. (1998). DNase I-hypersensitive sites I and II of the human growth hormone locus control region are a major developmental activator of somatotrope gene expression. Proc Natl Acad Sci U S A 95, 10655-60.

Bennett, P. A., Levy, A., Sophokleous, S., Robinson, I. C., and Lightman, S. L. (1995). Hypothalamic GH receptor gene expression in the rat: effects of altered GH status. J Endocrinol 147, 225-34.

Bennett, P. A., and Robinson, I. C. A. F. (2000). The central nervous system as a direct target for growth hormone action. In: Human Growth Hormone: Basic and Clinical Research, R. G. Smith and M. O. Thomer, eds. (Totowa, NJ: Humana Press Inc.), pp. 145-166.

Bennett, P. A., Thomas, G. B., Howard, A. D., Feighner, S. D., van der Ploeg, L. H., Smith, R. G., and Robinson, I. C. (1997). Hypothalamic growth hormone secretagogue- receptor (GHS-R) expression is regulated by growth hormone in the rat. Endocrinology 138, 4552-7.

184 References

Bennett, P. A., Thomas, G. B., and Robinson, I. A. C. F. (1998). Growth hormone releasing peptide (GHRP-6) regulation of hypothalamic gene expression (Abstract). In: The Endocrine Society's 80th Annual Meeting (New Orleans, Louisiana).

Bercu, B. B., Weideman, C. A., and Walker, R. F. (1991). Sex differences in growth hormone (GH) secretion by rats administered GH-releasing hexapeptide. Endocrinology 129, 2592-8.

Bercu, B. B., Yang, S. W., Masuda, R., and Walker, R. F. (1992). Role of selected endogenous peptides in growth hormone-releasing hexapeptide activity: analysis of growth hormone-releasing hormone, thyroid hormone-releasing hormone, and gonadotropin-releasing hormone. Endocrinology 130, 2579-86.

Berelowitz, M., Bruno, J. F., and White, J. D. (1992). Regulation of hypothalamic neuropeptide expression by peripheral metabolism. Trends Endocr Metab 3, 127-133.

Berelowitz, M., Szabo, M., Frohman, L. A., Firestone, S., Chu, L., and Hintz, R. L. (1981). Somatomedin-C mediates growth hormone negative feedback by effects on both the hypothalamus and the pituitary. Science 212, 1279-81.

Berry, S. A., and Pescovitz, O. H. (1988). Identification of a rat GHRH-like substance and its messenger RNA in rat testis. Endocrinology 123, 661-3.

Bertherat, J., Bluet-Pajot, M. T., and Epelbaum, J. (1995). Neuroendocrine regulation of growth hormone. Eur J Endocrinol 132, 12-24.

Bilezikjian, L. M., and Vale, W. W. (1984). Chronic exposure of cultured rat anterior pituitary cells to GRF causes partial loss of responsiveness to GRF. Endocrinology 115, 2032-4.

Billestrup, N., Mitchell, R. L., Vale, W., and Verma, I. M. (1987). Growth hormone- releasing factor induces c-fos expression in cultured primary pituitary cells. Mol Endocrinol 1, 300-5.

Billestrup, N., Swanson, L. W., and Vale, W. (1986). Growth hormone-releasing factor stimulates proliferation of somatotrophs in vitro. Proc Natl Acad Sci U S A 83, 6854-7.

Bisi, G., Podio, V., Valetto, M. R., Broglio, F., Bertuccio, G., Del Rio, G., Arvat, E., Boghen, M. F., Deghenghi, R., Muccioli, G., Ong, H., and Ghigo, E. (1999). Acute cardiovascular and hormonal effects of GH and hexarelin, a synthetic GH-releasing peptide, in humans. J Endocrinol Invest 22, 266-72.

Blackwood, E. M., and Kadonaga, J. T. (1998). Going the distance: a current view of enhancer action. Science 281, 61-3.

Blake, A. D., and Smith, R. G. (1991). Desensitization studies using perifused rat pituitary cells show that growth hormone-releasing hormone and His-D-Trp-Ala-Trp-D-

185 References

Phe-Lys-NH2 stimulate growth hormone release through distinct receptor sites. J Endocrinol 129, 11-9.

Bodart, V., Bouchard, J. F., McNicoll, N., Escher, E., Carrière, P., Ghigo, E., Sejlitz, T., Sirois, M. G., Lamontagne, D., and Ong, H. (1999). Identification and characterization of a new growth hormone-releasing peptide receptor in the heart. Circ Res 85, 796-802.

Bodner, M., Castrillo, J. L., Theill, L. E., Deerinck, T., Ellisman, M., and Karin, M. (1988). The pituitary-specific transcription factor GHF-1 is a homeobox-containing protein. Cell 55, 505-18.

Bohannon, N. J., Figlewicz, D. P., Corp, E. S., Wilcox, B. J., Porte, D., Jr., and Baskin, D. G. (1986). Identification of binding sites for an insulin-like growth factor (IGF-I) in the median eminence of the rat brain by quantitative autoradiography. Endocrinology 119, 943-5.

Borrelli, E., Heyman, R. A., Arias, C., Sawchenko, P. E., and Evans, R. M. (1989). Transgenic mice with inducible dwarfism. Nature 339, 538-41.

Bosman, F. T., Van Assche, C., Nieuwenhuyzen Kruseman, A. C., Jackson, S., and Lowry, P. J. (1984). Growth hormone releasing factor (GRF) immunoreactivity in human and rat gastrointestinal tract and pancreas. J Histochem Cytochem 32, 1139-44.

Botteri, F. M., van der Putten, H., Wong, D. F., Sauvage, C. A., and Evans, R. M. (1987). Unexpected thymic hyperplasia in transgenic mice harboring a neuronal promoter fused with simian virus 40 large T antigen. Mol Cell Biol 7, 3178-84.

Bowers, C. Y. (1993). GH releasing peptides—structure and kinetics. J Pediatr Endocrinol 6, 21-31.

Bowers, C. Y. (1998). Growth hormone-releasing peptide (GHRP). Cellular and Molecular Life Sciences 54, 1316-1329.

Bowers, C. Y., Alster, D. K., and Frentz, J. M. (1992). The growth hormone-releasing activity of a synthetic hexapeptide in normal men and short statured children after oral administration. J Clin Endocrinol Metab 74, 292-8.

Bowers, C. Y., Chang, J. K., and Fong, T. T. W. (1977). A synthetic pentapeptide which specifically releases GH, in vitro (Abstract). In: The Endocrine Society's 59th Annual Meeting (Chicago, USA), pp. 2332.

Bowers, C. Y., Momany, F. A., Reynolds, G. A., and Hong, A. (1984). On the in vitro and in vivo activity of a new synthetic hexapeptide that acts on the pituitary to specifically release growth hormone. Endocrinology 114, 1537-45.

186 References

Bowers, C. Y., Reynolds, G. A., and Chang, K. (2000). Unnatural GHRP and natural Ghrelin linkage (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 170.

Bowers, C. Y., Reynolds, G. A., Durham, D., Barrera, C. M., Pezzoli, S. S., and Thomer, M. O. (1990). Growth hormone (GH)-releasing peptide stimulates GH release in normal men and acts synergistically with GH-releasing hormone. J Clin Endocrinol Metab 70, 975-82.

Bowers, C. Y., Veeraragavan, K., and Sethumadhavan, K. (1994). Atypical growth hormone peptides. In: Growth hormone II: Basic and Clinical Aspects, B.B. Bercu and R. F. Walker, eds. (New York: Springer-Verlag), pp. 203-222.

Brazeau, P., Ling, N., Esch, F., Bohlen, P., Benoit, R., and Guillemin, R. (1981). High biological activity of the synthetic replicates of somatostatin-28 and somatostatin-25. Regul Pept 1, 255-64.

Brazeau, P., Ling, N., Esch, P., Bohlen, P., Mougin, C., and Guillemin, R. (1982). Somatocrinin (growth hormone releasing factor) in vitro bioactivity; Ca++ involvement, cAMP mediated action and additivity of effect with PGE2. Biochem Biophys Res Commun 109, 588-94.

Brazeau, P., Vale, W., Burgus, R., Ling, N., Butcher, M., Rivier, J., and Guillemin, R. (1973). Hypothalamic polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone. Science 179,77-9.

Breder, C. D., Yamada, Y., Yasuda, K., Seino, S., Saper, C. B., and Bell, G. I. (1992). Differential expression of somatostatin receptor subtypes in brain. J Neurosci 12, 3920- 34.

Breyer, P. R., Rothrock, J. K., Beaudry, N., and Pescovitz, O. H. (1996). A novel peptide from the growth hormone releasing hormone gene stimulates Sertoli cell activity. Endocrinology 137, 2159-62.

Bruno, J. F., Xu, Y., Song, J., and Berelowitz, M. (1993). Tissue distribution of somatostatin receptor subtype messenger ribonucleic acid in the rat. Endocrinology 133, 2561-7.

Burton, K. A., Steiner, R. A., and Clifton, D. K. (1992). Double-label in situ hybridisation confirms that the GRF gene and the GH receptor gene are coexpressed in a subset of neurons in the arcuate nucleus (Abstract). In: The Endocrine Society's 74th Annual Meeting (San Antonio, TX), pp. 1370.

Campbell, G. S., Christian, L. J., and Carter-Su, C. (1993). Evidence for involvement of the growth hormone receptor-associated tyrosine kinase in actions of growth hormone. J Biol Chem 268, 7427-34.

187 References

Canonico, P. L., Cronin, M. J., Thomer, M. O., and MacLeod, R. M. (1983). Human pancreatic GRF stimulates phosphatidylinositol labeling in cultured anterior pituitary cells. Am J Physiol 245, E587-90.

Carlsson, L., and Jansson, J. O. (1990). Endogenous growth hormone (GH) secretion in male rats is synchronized to pulsatile GH infusions given at 3-hour intervals. Endocrinology 126, 6-10.

Carmignac, D. P., Bennett, P. A., and Robinson, I. C. (1998). Effects of growth hormone secretagogues on prolactin release in anesthetized dwarf (dw/dw) rats. Endocrinology 139, 3590-6.

Carmignac, D. P., Plavell, D. M., and Robinson, I. C. (1996). Pituitary growth hormone- releasing factor receptor expression in normal and dwarf rats. Neuroendocrinology 64, 177-85.

Carmignac, D. P., and Robinson, I. C. (1990). Growth hormone (GH) secretion in the dwarf rat: release, clearance and responsiveness to GH-releasing factor and somatostatin. J Endocrinol 127, 69-75.

Catzeflis, C., Pierroz, D. D., Rohner-Jeanrenaud, P., Rivier, J. E., Sizonenko, P. C., and Aubert, M. L. (1993). Neuropeptide Y administered chronically into the lateral ventricle profoundly inhibits both the gonadotropic and the somatotropic axis in intact adult female rats. Endocrinology 132, 224-34.

Ceda, G. P. (1995). IGPs in the feedback control of GH secretion: Hypothalamic and/or pituitary action? J Endocrinol Invest 18, 734-7.

Ceda, G. P., Davis, R. G., Rosenfeld, R. G., and Hoffman, A. R. (1987). The growth hormone (GH)-releasing hormone (GHRH)-GH-somatomedin axis: evidence for rapid inhibition of GHRH-elicited GH release by insulin-like growth factors I and II. Endocrinology 120, 1658-62.

Chan, Y. Y., Steiner, R. A., and Clifton, D. K. (1996). Regulation of hypothalamic neuropeptide-Y neurons by growth hormone in the rat. Endocrinology 137, 1319-25.

Chang, C. H., Rickes, E. L., Marsilio, P., McGuire, L., Cosgrove, S., Taylor, J., Chen, H., Peighner, S., Clark, J. N., De Vita, R., and et al. (1995). Activity of a novel nonpeptidyl growth hormone secretagogue, L-700,653, in swine. Endocrinology 136, 1065-71.

Chang, C. H., Rickes, E. L., McGuire, L., Frazier, E., Chen, H., Barakat, K., Nargund, R., Patchett, A., Smith, R. G., and Hickey, G. J. (1996). Growth hormone (GH) and insulin-like growth factor I responses after treatments with an orally active GH secretagogue L-163,255 in swine. Endocrinology 137, 4851-6.

188 References

Chapman, L M., Bach, M. A., Van Canter, E., Farmer, M., Krupa, D., Taylor, A. M., Schilling, L. M., Cole, K. Y., Skiles, E. H., Pezzoli, S. S., Hartman, M. L., Veldhuis, J. D., Gormley, G. J., and Thorner, M. O. (1996). Stimulation of the growth hormone (GH)-insulin-like growth factor I axis by daily oral administration of a GH secretogogue (MK-677) in healthy elderly subjects. J Clin Endocrinol Metab 81,4249-57.

Chapman, I. M., Pescovitz, O. H., Murphy, G., Treep, T., Cerchio, K. A., Krupa, D., Gertz, B., Polvino, W. J., Skiles, E. H., Pezzoli, S. S., and Thomer, M. O. (1997). Oral administration of growth hormone (GH) releasing peptide-mimetic MK-677 stimulates the GH/insulin-like growth factor-I axis in selected GH-deficient adults. J Clin Endocrinol Metab 82, 3455-63.

Charlton, H. M., Clark, R. G., Robinson, I. C., Goff, A. E., Cox, B. S., Bugnon, C., and Bloch, B. A. (1988). Growth hormone-deficient dwarfism in the rat: a new mutation. J Endocrinol 119, 51-8.

Chen, C., Farnworth, P., Petersenn, S., Musgrave, I., Canny, B. J., and Clarke, I. J. (1998). Growth hormone-releasing peptide-2 (GHRP-2) does not act via the human growth hormone-releasing factor receptor in GC cells. Endocrine 9, 71-7.

Chen, E. Y., Liao, Y. C., Smith, D. H., Barrera-Saldana, H. A., Gelinas, R. E., and Seeburg, P. H. (1989). The human growth hormone locus: nucleotide sequence, biology, and evolution. Genomics 4,479-97.

Cheng, K., Chan, W. W., Barreto, A., Jr., Convey, E. M., and Smith, R. G. (1989). The synergistic effects of His-D-Trp-Ala-Trp-D-Phe-Lys-NH2 on growth hormone (GH)- releasing factor-stimulated GH release and intracellular adenosine 3',5'-monophosphate accumulation in rat primary pituitary cell culture. Endocrinology 124, 2791-8.

Cheng, K., Chan, W. W., Butler, B., Wei, L., Schoen, W. R., Wyvratt, M. J., Jr., Fisher, M. H., and Smith, R. G. (1993a). Stimulation of growth hormone release from rat primary pituitary cells by L-692,429, a novel non-peptidyl GH secretagogue. Endocrinology 132, 2729-31.

Cheng, K., Chan, W. W., Butler, B., Wei, L., and Smith, R. G. (1993). A novel non­ peptidyl growth hormone secretagogue. Horm Res 40, 109-15.

Cheng, K., Chan, W. W. S., Butler, B., Barreto, A., and Smith, R. G. (1991). Evidence For a Role of Protein-Kinase-C in His-D-Trp-Ala- Lys-Nh2-Induced Growth-Hormone Release From Rat Primary Cells. Endocrinology 129, 3337-3343.

Cheng, T. C., Beamer, W. G., Phillips, J. A. d., Bartke, A., Mallonee, R. L., and Dowling, C. (1983). Etiology of growth hormone deficiency in little, Ames, and Snell dwarf mice. Endocrinology 113, 1669-78.

189 References

Chihara, K., Minamitani, N., Kaji, H., Arimura, A., and Fujita, T. (1981). Intraventricularly injected growth hormone stimulates somatostatin release into rat hypophysial portal blood. Endocrinology 109, 2279-81.

Chihara, K., Minamitani, N., Kaji, H., Kodama, H., Kita, T., and Fujita, T. (1983). Human pancreatic growth hormone-releasing factor stimulates release of growth hormone in conscious unrestrained male rabbits. Endocrinology 113, 2081-5.

Chomczynski, P., Downs, T. R., and Frohman, L. A. (1988). Feedback regulation of growth hormone (GH)-releasing hormone gene expression by GH in rat hypothalamus. Mol Endocrinol 2, 236-41.

Ciccarelli, E., Grottoli, S., Razzore, P., Gianotti, L., Arvat, E., Deghenghi, R., Camanni, G., and Ghigo, E. (1996). Hexarelin, a synthetic growth hormone releasing peptide, stimulates prolactin secretion in acromegalic but not in hyperprolactinaemic patients. Clin Endocrinol (Oxf) 44, 67-71.

Clark, R. G., Carlsson, L. M., and Robinson, I. C. (1988). Growth hormone (GH) secretion in the conscious rat: negative feedback of GH on its own release. J Endocrinol 119, 201-9.

Clark, R. G., Carlsson, L. M., and Robinson, I. C. (1987). Growth hormone secretory profiles in conscious female rats. J Endocrinol 114, 399-407.

Clark, R. G., Carlsson, L. M. S., Trojnar, J., and Robinson, I. C. A. F. (1989). The effects of a growth hormone-releasing peptide and growth hormone-releasing factor in conscious and anaesthetized rats. J Neuroendocrinol 1, 249-255.

Clark, R. G., and Robinson, I. C. (1985). Effects of a fragment of human growth hormone-releasing factor in normal and 'Little' mice. J Endocrinol 106, 1-5.

Clark, R. G., and Robinson, I. C. (1996). Up and down the growth hormone cascade. Cytokine Growth Factor Rev 7, 65-80.

Codd, E. E., Shu, A. Y., and Walker, R. F. (1989). Binding of a growth hormone releasing hexapeptide to specific hypothalamic and pituitary binding sites. Neuropharmacology 28, 1139-44.

Cogan, J. D., Phillips, J. A., 3rd, Schenkman, S. S., Milner, R. D., and Sakati, N. (1994). Familial growth hormone deficiency: a model of dominant and recessive mutations affecting a monomeric protein. J Clin Endocrinol Metab 79, 1261-5.

Conley, L. K., Teik, J. A., Deghenghi, R., Imbimbo, B. P., Giustina, A., Locatelli, V., and Wehrenberg, W. B. (1995). Mechanism of action of hexarelin and GHRP-6: analysis of the involvement of GHRH and somatostatin in the rat. Neuroendocrinology 61, 44-50.

190 References

Copinschi, G., Leproult, R., Van Onderbergen, A., Gaufriez, A., Cole, K, Y., Schilling, L. M., Mendel, C. M., De Lepeleire, L, Bolognese, J. A., and Van Cauter, E. (1997). Prolonged oral treatment with MK-677, a novel growth hormone secretagogue, improves sleep quality in man. Neuroendocrinology 66, 278-86.

Copinschi, G., Van Onderbergen, A., L'Hermite-Baleriaux, M., Mendel, C. M., Gaufriez, A., Leproult, R., Bolognese, J. A., De Smet, M., Thomer, M. O., and Van Cauter, E. (1996). Effects of a 7-day treatment with a novel, orally active, growth hormone (GH) secretagogue, MK-677, on 24-hour GH profiles, insulin-like growth factor I, and adrenocortical function in normal young men. J Clin Endocrinol Metab 81, 2776-82.

Groom, W. J., Leonard, E. S., Baker, P. K., Kraft, L. A., and Ricks, C. A. (1984). The effects of sythetic growth hormone releasing hexapeptide BI679 on semm growth hormone levels and production in lactating dairy cattle (Abstract). Journal of Animal Science 67, 109.

Cunha, S. R., and Mayo, K. E. (2000). GHRH signaling and Map Kinase Pathway Activation in GH3 cells expressing the GHRH receptor (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, Canada), pp. 161.

Date, Y., Murakami, N., Kojima, M., Kuroiwa, T., Matsukura, S., Kangawa, K., and Nakazato, M. (2000). Central effects of a novel acylated peptide, ghrelin, on growth hormone release in rats. Biochem Biophys Res Commun 275,477-80.

Davis, S. L., Ohlson, D. L., Klindt, J., and Anfmson, M. S. (1977). Episodic growth hormone secretory patterns in sheep: relationship to gonadal steroid hormones. Am J Physiol 233, E519-23. de Gennaro Colonna, V., Cattaneo, E., Cocchi, D., Muller, E. E., and Maggi, A. (1988). Growth hormone regulation of growth hormone-releasing hormone gene expression. Peptides 9, 985-8.

De Gennaro Colonna, V., Rossoni, G., Bemareggi, M., Muller, E. E., and Berti, F. (1997). Cardiac ischemia and impairment of vascular endothelium function in hearts from growth hormone-deficient rats: protection by hexarelin. Eur J Pharmacol 334, 201- 7. de Keyzer, Y., Lenne, P., and Bertagna, X. (1997). Widespread transcription of the growth hormone-releasing peptide receptor gene in neuroendocrine human tumors. Eur J Endocrinol 137, 715-8.

Dean, D. C., Nargund, R. P., Pong, S. S., Chaung, L. Y., Griffin, P., Melillo, D. G., Ellsworth, R. L., Van der Ploeg, L. H., Patchett, A. A., and Smith, R. G. (1996). Development of a high specific activity sulfur-35-labeled sulfonamide radioligand that allowed the identification of a new growth hormone secretagogue receptor. J Med Chem 39, 1767-70.

191 References

Deuben, R. R., and Meites, J. (1964). Stimulation of pituitary growth hormone release by a hypothalamic extract in vitro. Endocrinology 74, 408-414.

DeVita, J. R., and Wyvratt Jr., M. J. (1996). Benzolactam growth hormone secretagogues. Drugs Future 21, 273-281.

Dickson, S. L., Doutrelant-Viltart, O., Dyball, R. E., and Leng, G. (1996). Retrogradely labelled neurosecretory neurones of the rat hypothalamic arcuate nucleus express Eos protein following systemic injection of GH-releasing peptide-6. J Endocrinol 151, 323- 31.

Dickson, S. L., Doutrelant-Viltart, O., and Leng, G. (1995b). GH-deficient dw/dw rats and lit/lit mice show increased Eos expression in the hypothalamic arcuate nucleus following systemic injection of GH-releasing peptide-6. J Endocrinol 146, 519-26.

Dickson, S. L., Leng, G., Dyball, R. E., and Smith, R. G. (1995a). Central actions of peptide and non-peptide growth hormone secretagogues in the rat. Neuroendocrinology 61,36-43.

Dickson, S. L., Leng, G., and Robinson, I. C. (1994). Electrical stimulation of the rat periventricular nucleus influences the activity of hypothalamic arcuate neurones. J Neuroendocrinol 6, 359-67.

Dickson, S. L., Leng, G., and Robinson, I. C. (1993a). Growth hormone release evoked by electrical stimulation of the arcuate nucleus in anesthetized male rats. Brain Res 623, 95-100.

Dickson, S. L., Leng, G., and Robinson, I. C. (1993b). Systemic administration of growth hormone-releasing peptide activates hypothalamic arcuate neurons. Neuroscience 53, 303-6.

Dickson, S. L., and Luckman, S. M. (1997). Induction of c-fos messenger ribonucleic acid in neuropeptide Y and growth hormone (GH)-releasing factor neurons in the rat arcuate nucleus following systemic injection of the GH secretagogue, GH-releasing peptide-6. Endocrinology 138, 771-7.

Dickson, S. L., Viltart, O., Bailey, A. R., and Leng, G. (1997). Attenuation of the growth hormone secretagogue induction of Eos protein in the rat arcuate nucleus by central somatostatin action. Neuroendocrinology 66, 188-94.

Doscher, M. E., Baker, P. K., Kraft, L. A., and Ricks, C. A. (1984). Effect of a synthetic growth hormone releasing peptide (BI679) and growth hormone releasing factor (GRE) on serum growth hormone levels in barrows (Abstract). Journal of Animal Science 59, 218.

Drisko, J. E., Eaidley, T. D., Zhang, D., McDonald, T. J., Nicolich, S., Hora, D. E., Cunningham, P., Li, C., Rickes, E., McNamara, L., Chang, C., Smith, R. G., and Hickey,

192 References

G. J. (1999). Administration of a nonpeptidyl growth hormone secretagogue, L-163, 255, changes somatostatin pattern, but has no effect on patterns of growth hormone- releasing factor in the hypophyseal-portal circulation of the conscious pig. Proc Soc Exp Biol Med 222, 70-7.

Eden, S. (1979). Age- and sex-related differences in episodic growth hormone secretion in the rat. Endocrinology 105, 555-60.

Elias, K. A., Ingle, G. S., Bumier, J. P., Hammonds, R. G., McDowell, R. S., Rawson, T. E., Somers, T. C., Stanley, M. S., and Cronin, M. J. (1995). In vitro characterization of four novel classes of growth hormone-releasing peptide. Endocrinology 136, 5694-9.

Escobar, D. C., Vicentini, L. M., Ghigo, E., Ciccarelli, E., Usellini, L., Capella, C., and Cocchi, D. (1986). Growth hormone-releasing factor does not stimulate phosphoinositides breakdown in primary cultures of rat and human pituitary cells. Acta Endocrinol (Copenh) 112, 345-50.

Fairhall, K. M., Mynett, A., and Robinson, I. C. (1995). Central effects of growth hormone-releasing hexapeptide (GHRP-6) on growth hormone release are inhibited by central somatostatin action. J Endocrinol 144, 555-60.

Fairhall, K. M., Mynett, A., Smith, R. G., and Robinson, I. C. (1995). Consistent GH responses to repeated injection of GH-releasing hexapeptide (GHRP-6) and the non­ peptide GH secretagogue, L-692,585. J Endocrinol 145,417-26.

Fairhall, K. M., Mynett, A., Thomas, G. B., and Robinson, I. C. A. F. (1996). Central and peripheral effects of peptide and nonpeptide GH secretagogues on GH release in vivo. In: Growth Hormone Secretagogues, B. B. Bercy and R. F. Walker, eds. (New York: Springer-Verlag), pp. 219-236.

Fang, S., Steinmetz, R., Walker King, D., Zeng, P., Vogelweid, C., Cooper, S., Hangcoc, G., Broxmeyer, H. E., and Pescovitz, O. H. (2000). Development of a transgenic mouse that overexpresses a novel product of the growth hormone-releasing hormone gene. Endocrinology 141, 1377-83.

Feighner, S. D., Howard, A. D., Prendergast, K., Palyha, O. C., Hreniuk, D. L., Nargund, R., Underwood, D., Tata, J. R., Dean, D. C., Tan, C. P., McKee, K. K., Woods, J. W., Patchett, A. A., Smith, R. G., and Van der Ploeg, L. H. (1998). Structural requirements for the activation of the human growth hormone secretagogue receptor by peptide and nonpeptide secretagogues. Mol Endocrinol 12, 137-45.

Feighner, S. D., Tan, C. P., McKee, K. K., Palyha, O. C., Hreniuk, D. L., Pong, S. S., Austin, C. P., Figueroa, D., MacNeil, D., Cascieri, M. A., Nargund, R., Bakshi, R., Abramovitz, M., Stocco, R., Kargman, S., O'Neill, G., Van Der Ploeg, L. H., Evans, J., Patchett, A. A., Smith, R. G., and Howard, A. D. (1999). Receptor for motilin identified in the human gastrointestinal system. Science 284, 2184-8.

193 References

Finkelstein, J. W., Roffwarg, H. P., Boyar, R. M., Kream, J., and Heilman, L. (1972). Age-related change in the twenty-four-hour spontaneous secretion of growth hormone. J Clin Endocrinol Metab 35, 665-70.

Finley, J. C., Grossman, G. H., Dimeo, P., and Petrusz, P. (1978). Somatostatin- containing neurons in the rat brain: widespread distribution revealed by immunocytochemistry after pretreatment with pronase. Am J Anat 153, 483-8.

Flavell, D. M., Wells, T., Wells, S. E., Carmignac, D. F., Thomas, G. B., and Robinson, I. C. (1996). Dominant dwarfism in transgenic rats by targeting human growth hormone (GH) expression to hypothalamic GH-releasing factor neurons. Embo J 15, 3871-9.

Fletcher, T. P., Thomas, G. B., and Clarke, I. J. (1996). Growth hormone-releasing hormone and somatostatin concentrations in the hypophysial portal blood of conscious sheep during the infusion of growth hormone-releasing peptide-6. Domest Anim Endocrinol 13, 251-8.

French, M. B., Lussier, B. T., Moor, B. C., and Kraicer, J. (1990). Effect of growth hormone-releasing factor on phosphoinositide hydrolysis in somatotrophs. Mol Cell Endocrinol 72, 221-6.

Frieboes, R. M., Murck, H., Maier, P., Schier, T., Holsboer, F., and Steiger, A. (1995). Growth hormone-releasing peptide-6 stimulates sleep, growth hormone, ACTH and cortisol release in normal man. Neuroendocrinology 61, 584-9.

Frohman, L. A., Downs, T. R., Chomczynski, P., Brar, A., and Kashio, Y. (1989). Regulation of growth hormone-releasing hormone gene expression and biosynthesis. Yale J Biol Med 62,427-33.

Frohman, L. A., Downs, T. R., Clarke, I. J., and Thomas, G. B. (1990). Measurement of growth hormone-releasing hormone and somatostatin in hypothalamic-portal plasma of unanesthetized sheep. Spontaneous secretion and response to insulin-induced hypoglycemia. J Clin Invest 86, 17-24.

Frohman, M. A., Downs, T. R., Chomczynski, P., and Frohman, L. A. (1989). Cloning and characterization of mouse growth hormone-releasing hormone (GRH) complementary DNA: increased GRH messenger RNA levels in the growth hormone- deficient lit/lit mouse. Mol Endocrinol 3, 1529-36.

Fuh, G., Cunningham, B. C., Fukunaga, R., Nagata, S., Goeddel, D. V., and Wells, J. A. (1992). Rational design of potent antagonists to the human growth hormone receptor. Science 256, 1677-80.

Fujii, R., Hosoya, M., Fukusumi, S., Kawamata, Y., Habata, Y., Hinuma, S., Onda, H., Nishimura, O., and Fujino, M. (2000). Identification of neuromedin U as the cognate ligand of the orphan G protein-coupled receptor FM-3. J Biol Chem 275, 21068-74.

194 References

Fukata, J., Diamond, D. J., and Martin, J. B. (1985). Effects of rat growth hormone (rGH)-releasing factor and somatostatin on the release and synthesis of rGH in dispersed pituitary cells. Endocrinology 117,457-67.

Galbraith, D. W., Anderson, M. T., and Herzenberg, L. A. (1999). Flow cytometric analysis and FACS sorting of cells based on GFP accumulation. Methods Cell Biol 58, 315-41.

Gaylinn, B. D., Dealmeida, V. I., Lyons, C. E., Jr., Wu, K. C., Mayo, K. E., and Thomer, M. O. (1999). The mutant growth hormone-releasing hormone (GHRH) receptor of the little mouse does not bind GHRH. Endocrinology 140, 5066-74.

Gaylinn, B. D., Harrison, J. K., Zysk, J. R., Lyons, C. E., Lynch, K. R., and Thomer, M. O. (1993). Molecular cloning and expression of a human anterior pituitary receptor for growth hormone-releasing hormone. Mol Endocrinol 7, 77-84.

Gertz, B. J., Barrett, J. S., Eisenhandler, R., Krupa, D. A., Wittreich, J. M., Seibold, J. R., and Schneider, S. H. (1993). Growth hormone response in man to L-692,429, a novel nonpeptide mimic of growth hormone-releasing peptide-6. J Clin Endocrinol Metab 77, 1393-7.

Gevers, E. F., Wit, J. M., and Robinson, I. C. (1996). Growth, growth hormone (GH)- binding protein, and GH receptors are differentially regulated by peak and trough components of the GH secretory pattem in the rat. Endocrinology 137, 1013-18.

Ghigo, E., Arvat, E., Gianotti, S., Rizzi, G., Ceda, G. P., Boghen, M. F., Deghenghi, R., and Camanni, F. (1996). Aging and GH-releasing peptides. In: Growth Hormone Secretagogues, B. B. Bercu and R. Walker, eds. (New York: Springer-Verlag).

Ghigo, E., Arvat, E., Ramunni, J., Colao, A., Gianotti, L., Deghenghi, R., Lombardi, G., and Camanni, F. (1997). Adrenocorticotropin- and cortisol-releasing effect of hexarelin, a synthetic growth hormone-releasing peptide, in normal subjects and patients with Cushing's syndrome. J Clin Endocrinol Metab 82, 2439-44.

Ghigo, E., Arvat, E., Rizzi, G., Goffi, S., Grottoli, S., Mucci, M., Boghen, M. F., and Camanni, F. (1994). Growth hormone-releasing activity of growth hormone-releasing peptide-6 is maintained after short-term oral pretreatment with the hexapeptide in normal aging. Eur J Endocrinol 131, 499-503.

Gick, G. G., Zeytin, F. N., Brazeau, P., Ling, N. C., Esch, F. S., and Bancroft, C. (1984). Growth hormone-releasing factor regulates growth hormone mRNA in primary cultures of rat pituitary cells. Proc Natl Acad Sci U S A 81,1553-5.

Giraldi, F. P., Mizobuchi, M., Horowitz, Z. D., Downs, T. R., Aleppo, G., Kier, A., Wagner, T., Yun, J. S., Kopchick, J. J., and Frohman, L. A. (1994). Development of neuroepithelial tumors of the adrenal medulla in transgenic mice expressing a mouse

195 References hypothalamic growth hormone-releasing hormone promoter-simian virus-40 T-antigen fusion gene. Endocrinology 134,1219-24.

Giustina, A., Bonfanti, C., Licini, M., Ragni, G., and Stefana, B. (1997). Hexarelin, a novel GHRP-6 analog, stimulates growth hormone (GH) release in a GH-secreting rat cell line (GHl) insensitive to GH-releasing hormone. Regul Pept 70,49-54.

Godfrey, P., Rahal, J. O., Beamer, W. G., Copeland, N. G., Jenkins, N. A., and Mayo, K. E. (1993). GHRH receptor of little mice contains a missense mutation in the extracellular domain that disrupts receptor function. Nat Genet 4, 227-32.

Gonzalez, G. A., and Montminy, M. R. (1989). Cyclic AMP stimulates somatostatin gene transcription by phosphorylation of CREB at serine 133. Cell 59, 675-80.

Gonzalez-Crespo, S., and Boronat, A. (1991). Expression of the rat growth hormone- releasing hormone gene in placenta is directed by an alternative promoter. Proc Natl Acad Sci U S A 88, 8749-53.

Goodyer, C. G., De Stephano, L., Lai, W. H., Guyda, H. J., and Posner, B. I. (1984). Characterization of insulin-like growth factor receptors in rat anterior pituitary, hypothalamus, and brain. Endocrinology 114, 1187-95.

Green, H., Morikawa, M., and Nixon, T. (1985). A dual effector theory of growth- hormone action. Differentiation 29, 195-8.

Gross, D. S., and Garrard, W. T. (1988). Nuclease hypersensitive sites in chromatin. Annu Rev Biochem 57, 159-97.

Grosveld, F., and Kollias, G. (1992). Transgenic Animals (London: Academic Press).

Grosveld, P., van Assendelft, G. B., Greaves, D. R., and Kollias, G. (1987). Position- independent, high-level expression of the human beta-globin gene in transgenic mice. Cell 51, 975-85.

Guan, X. M., Yu, H., Palyha, O. C., McKee, K. K., Feighner, S. D., Sirinathsinghji, D. J., Smith, R. G., Van der Ploeg, L. H., and Howard, A. D. (1997). Distribution of mRNA encoding the growth hormone secretagogue receptor in brain and peripheral tissues. Brain Res Mol Brain Res 48, 23-9.

Guillaume, V., Magnan, E., Cataldi, M., Dutour, A., Sauze, N., Renard, R. H., Contedevolx, B., Deghenghi, R., and Lenaerts (1994). Growth-Hormone (Gh)-Releasing Hormone-Secretion Is New Gh-Releasing Hexapeptide in Sheep. Endocrinology 135, 1073-1076.

Guillemin, R., Brazeau, P., Bohlen, P., Esch, P., Ling, N., and Wehrenberg, W. B. (1982). Growth hormone-releasing factor from a human pancreatic tumor that caused acromegaly. Science 218, 585-7.

196 References

Guo, F., Beaudet, A., and Tannenbaum, G. S. (1996). The effect of hypophysectomy and growth hormone replacement on sstl and sst2 somatostatin receptor subtype messenger ribonucleic acids in the arcuate nucleus. Endocrinology 137, 3928-35.

Hammer, R. B., Brinster, R. L., Rosenfeld, M. G., Evans, R. M., and Mayo, K. E. (1985). Expression of human growth hormone-releasing factor in transgenic mice results in increased somatic growth. Nature 315, 413-6.

Hammer, R. E., Palmiter, R. D., and Brinster, R. L. (1984). Partial correction of murine hereditary growth disorder by germ-line incorporation of a new gene. Nature 311, 65-7.

Hanew, K., Utsumi, A., Sugawara, A., Shimizu, Y., and Abe, K. (1994). Enhanced GH responses to combined administration of GHRP and GHRH in patients with acromegaly. J Clin Endocrinol Metab 78, 509-12.

Hansen, B. S., Raun, K., Nielsen, K. K., Johansen, P. B., Hansen, T. K., Peschke, B., Lau, J., Andersen, P. H., and Ankersen, M. (1999). Pharmacological characterisation of a new oral GH secretagogue, NN703. Eur J Endocrinol 141, 180-9.

Hayashi, S., Kaji, H., Abe, H., and Chihara, K. (1993). Effect of intravenous administration of growth hormone-releasing peptide on plasma growth hormone in patients with short stature. Clin Pediatr Endocrinol 2, 69-74.

Hayashi, S., Okimura, Y., Yagi, H., Uchiyama, T., Takeshima, Y., Shakutsui, S., Oohashi, S., Bowers, C. Y., and Chihara, K. (1991). Intranasal administration of His-D- Trp-Ala-Trp-D-Phe-LysNH2 (growth hormone releasing peptide) increased plasma growth hormone and insulin-like growth factor-I levels in normal men. Endocrinol Jpn 38, 15-21.

Helboe, L., Stidsen, C. E., and Moller, M. (1998). Immunohistochemical and cytochemical localization of the somatostatin receptor subtype sstl in the somatostatinergic parvocellular neuronal system of the rat hypothalamus. J Neurosci 18, 4938-45.

Herington, A. C., Ymer, S. I., and Tiong, T. S. (1991). Does the serum binding protein for growth hormone have a functional role? Acta Endocrinol (Copenh) 124 Suppl 2, 14- 20.

Herrington, J., and Hille, B. (1994). Growth hormone-releasing hexapeptide elevates intracellular calcium in rat somatotropes by two mechanisms. Endocrinology 135, 1100- 8.

Hewson, A., and Dickson, S. (2000). Systemic administration of Ghrelin activates hypothalamic arcuate nucleus neurones: a comparison between fasted and fed rats (Abstract). In: Annual Meeting of the Growth Hormone Research Society (Gothenburg, Sweden), pp. 132.

197 References

Hewson, A. K., Viltart, O., McKenzie, D. N., Dyball, R. E., and Dickson, S. L. (1999). GHRP-6-induced changes in electrical activity of single cells in the arcuate, ventromedial and periventricular nucleus neurones [correction of nuclei] of a hypothalamic slice preparation in vitro. J Neuroendocrinol 11, 919-23.

Hickey, G., Jacks, T., Judith, F., Taylor, J., Schoen, W. R., Krupa, D., Cunningham, P., Clark, J., and Smith, R. G. (1994). Efficacy and specificity of L-692,429, a novel nonpeptidyl growth hormone secretagogue, in beagles. Endocrinology 134, 695-701.

Hickey, G. J., Drisko, J., Eaidley, T., Chang, C., Anderson, L. L., Nicolich, S., McGuire, L., Rickes, E., Krupa, D., Feeney, W., Friscino, B., Cunningham, P., Frazier, E., Chen, H., Laroque, P., and Smith, R. G. (1996). Mediation by the central nervous system is critical to the in vivo activity of the GH secretagogue L-692,585. J Endocrinol 148, 371- 80.

Hickey, G. J., Jacks, T. M., Schleim, K. D., Frazier, E., Chen, H. Y., Krupa, D., Feeney, W., Nargund, R. P., Patchett, A. A., and Smith, R. G. (1997). Repeat administration of the GH secretagogue MK-0677 increases and maintains elevated IGF-I levels in beagles. J Endocrinol 152, 183-92.

Hindmarsh, P., Smith, P. J., Brook, C. G., and Matthews, D. R. (1987). The relationship between height velocity and growth hormone secretion in short prepubertal children. Clin Endocrinol (Oxf) 27, 581-91.

Hindmarsh, P. C., Fall, C. H., Pringle, P. J., Osmond, C., and Brook, C. G. (1997). Peak and trough growth hormone concentrations have different associations with the insulin­ like growth factor axis, body composition, and metabolic parameters. J Clin Endocrinol Metab 82, 2172-6.

Hogan, B. L. H., Constantini, F., and Lacy, E. (1986). Manipulating the Mouse Embryo (Cold Spring Harbour, NY: Cold Spring Harbour Laboratory Press).

Horikawa, R., Tachibana, T., Katsumata, N., Ishikawa, H., and Tanaka, T. (2000). Regulation of pituitary growth hormone-secretagogue and growth hormone-releasing hormone receptor RNA expression in young Dwarf rats. Endocr J 47 Suppl, S53-6.

Hosoda, H., Kojima, M., Matsuo, H., and Kangawa, K. (2000). Purification and Characterization of Rat des-Gln 14-Ghrelin, a Second Endogenous Ligand for the Growth Hormone Secretagogue Receptor. J Biol Chem 275, 21995-22000.

Howard, A. D., Feighner, S. D., Cully, D. F., Arena, J. P., Liberator, P. A., Rosenblum, C. I., Hamelin, M., Hreniuk, D. L., Palyha, O. C., Anderson, J., Paress, P. S., Diaz, C., Chou, M., Liu, K. K., McKee, K. K., Pong, S. S., Chaung, L. Y., Elbrecht, A., Dashkevicz, M., Heavens, R., Rigby, M., Sirinathsinghji, D. J. S., Dean, D. C., Melillo, D. G., and Van der Ploeg, L. H. (1996). A receptor in pituitary and hypothalamus that functions in growth hormone release. Science 273, 974-7.

198 References

Howard, A. D., Wang, R., Pong, S. S., Mellin, T. N., Strack, A., Guan, X. M., Zeng, Z., Williams, D. L., Jr., Feighner, S. D., Nunes, C. N., Murphy, B., Stair, J. N., Yu, H., Jiang, Q., Clements, M. K., Tan, C. P., McKee, K. K., Hreniuk, D. L., McDonald, T. P., Lynch, K. R., Evans, J. P., Austin, C. P., Caskey, C. T., Van der Ploeg, L. H., and Liu, Q. (2000). Identification of receptors for neuromedin U and its role in feeding. Nature 406, 70-4.

Hreniuk, D. L., Howard, A. D., Rosenblum, C. I., Palyha, O., Diaz, C., Cully, D. P., Paress, P. S., McKee, K. K., Dashkevicz, M., Arena, J. P., Liu, K. K., Schaeffer, J. M., Nargund, R. P., Smith, R. G., VanderPloeg, L. H. T., and Peighner, S. (1996). Cytochemical identification and cloning of a G-protein coupled growth hormone secretagogue receptor (Abstract). In: The 26th Annual Meeting of the Society for Neuroscience (Washington, DC), pp. 1300.

Huhn, W. C., Hartman, M. L., Pezzoli, S. S., and Thomer, M. O. (1993). Twenty-four- hour growth hormone (GH)-releasing peptide (GHRP) infusion enhances pulsatile GH secretion and specifically attenuates the response to a subsequent GHRP bolus. J Clin Endocrinol Metab 76, 1202-8.

Hurley, D. L., Wee, B. E., and Phelps, C. J. (1998). Growth hormone releasing hormone expression during postnatal development in growth hormone-deficient Ames dwarf mice: mRNA in situ hybridization. Neuroendocrinology 68, 201-9.

Ibata, Y., Okamura, H., Makino, S., Kawakami, P., Morimoto, N., and Chihara, K. (1986). Light and electron microscopic immunocytochemistry of GRP-like immunoreactive neurons and terminals in the rat hypothalamic arcuate nucleus and median eminence. Brain Res 370, 136-43.

Ikeda, S. R., and Schofield, G. G. (1989). Somatostatin blocks a calcium current in rat sympathetic ganglion neurones. J Physiol (Lond) 409, 221-40.

Ilson, B. E., Jorkasky, D. K., Cumow, R. T., and Stote, R. M. (1989). Effect of a new synthetic hexapeptide to selectively stimulate growth hormone release in healthy human subjects. J Clin Endocrinol Metab 69, 212-4.

Imbimbo, B. P., Mant, T., Edwards, M., Amin, D., Dalton, N., Boutignon, P., Lenaerts, V., Wuthrich, P., and Deghenghi, R. (1994). Growth hormone-releasing activity of hexarelin in humans. A dose-response study. Eur J Clin Pharmacol 46,421-5.

Ingraham, H. A., Chen, R. P., Mangalam, H. J., Elsholtz, H. P., Plynn, S. E., Lin, C. R., Simmons, D. M., Swanson, L., and Rosenfeld, M. G. (1988). A tissue-specific transcription factor containing a homeodomain specifies a pituitary phenotype. Cell 55, 519-29.

Iwakura, H., Hosoda, K., Son, C., Sagawa, N., Itoh, H., Yura, H., Matsuda, J., Akamizu, T., Pujii, S., Hosoda, H., Kojima, M., Kangawa, K., and Nakao, K. (2000). Ghrelin gene

199 References was expressed in human placenta (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 273.

Jacks, T., Hickey, G., Judith, P., Taylor, J., Chen, H., Krupa, D., Feeney, W., Schoen, W., Ok, D., Fisher, M., and et al. (1994). Effects of acute and repeated intravenous administration of L-692,585, a novel non-peptidyl growth hormone secretagogue, on plasma growth hormone, IGF-1, ACTH, cortisol, prolactin, insulin, and thyroxine levels in beagles. J Endocrinol 143, 399-406.

Jacks, T., Smith, R., Judith, P., Schleim, K., Frazier, E., Chen, H., Krupa, D., Hora, D., Jr., Nargund, R., Patchett, A., and Hickey, G. (1996). MK-0677, a potent, novel, orally active growth hormone (GH) secretagogue: GH, insulin-like growth factor I, and other hormonal responses in beagles. Endocrinology 137, 5284-9.

Jaffe, C. A., Ho, P. J., Demott-Friberg, R., Bowers, C. Y., and Barkan, A. L. (1993). Effects of a prolonged growth hormone (GH)-releasing peptide infusion on pulsatile GH secretion in normal men. J Clin Endocrinol Metab 77, 1641-7.

Jansson, J.-O., Downs, T. R., Beamer, W. G., and Frohman, L. A. (1986b). The dwarf 'little' (lit/lit) mouse is resistant to growth hormone (GH) releasing peptide (GH-RP-6) as well as GH-releasing hormone (GRH) (Abstract). In: The Endocrine Society's 68th Annual Meeting (Anaheim, CA), pp. 397.

Jansson, J. O., Downs, T. R., Beamer, W. G., and Frohman, L. A. (1986a). Receptor- associated resistance to growth hormone-releasing factor in dwarf "little" mice. Science 232,511-2.

Jansson, J. O., Eden, S., and Isaksson, O. (1985). Sexual dimorphism in the control of growth hormone secretion. Endocr Rev 6, 128-50.

Jones, B. K., Monks, B. R., Liebhaber, S. A., and Cooke, N. E. (1995). The human growth hormone gene is regulated by a multicomponent locus control region. Mol Cell Biol 15, 7010-21.

Kaji, H., Tai, S., Okimura, Y., Iguchi, G., Takahashi, Y., Abe, H., and Chihara, K. (1998). Cloning and characterization of the 5'-flanking region of the human growth hormone secretagogue receptor gene. J Biol Chem 273, 33885-8.

Kamegai, J., Hasegawa, O., Minami, S., Sugihara, H., and Wakabayashi, I. (1996). The growth hormone-releasing peptide KP-102 induces c-fos expression in the arcuate nucleus. Brain Res Mol Brain Res 39, 153-9.

Kamegai, J., Minami, S., Sugihara, H., Higuchi, H., and Wakabayashi, I. (1994). Growth hormone induces expression of the c-fos gene on hypothalamic neuropeptide-Y and somatostatin neurons in hypophysectomized rats. Endocrinology 135, 2765-71.

200 References

Kamegai, J., Tamura, H., Ishii, S., Shuto, Y., Sugihara, H., and Wakabayashi, I. (2000). Central effects of Ghrelin, an endogenous GH Secretagogue, on hypothalamic peptide gene expression (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 170.

Kamegai, J., Wakabayashi, I., Miyamoto, K., Unterman, T. G., Kineman, R. D., and Frohman, L. A. (1998). Growth hormone-dependent regulation of pituitary GH secretagogue receptor (GHS-R) mRNA levels in the spontaneous dwarf Rat. Neuroendocrinology 68, 312-8.

Katada, T., Gilman, A. G., Watanabe, Y., Bauer, S., and Jakobs, K. H. (1985). Protein kinase C phosphorylates the inhibitory guanine-nucleotide-binding regulatory component and apparently suppresses its function in hormonal inhibition of adenylate cyclase. Bur J Biochem 151,431-7.

Kawabe, J., Iwami, G., Ebina, T., Ohno, S., Katada, T., Ueda, Y., Homey, C. J., and Ishikawa, Y. (1994). Differential activation of adenylyl cyclase by protein kinase C isoenzymes. J Biol Chem 269, 16554-8.

Kawano, H., and Daikoku, S. (1988). Somatostatin-containing neuron systems in the rat hypothalamus: retrograde tracing and immunohistochemical studies. J Comp Neurol 271,293-9.

Kineman, R. D., and Frawley, L. S. (1994). Secretory characteristics and phenotypic plasticity of growth hormone- and prolactin-producing cell lines. J Endocrinol 140, 455- 63.

Kineman, R. D., Kamegai, J., and Frohman, L. A. (1999). Growth hormone (GH)- releasing hormone (GHRH) and the GH secretagogue (GHS), L692,585, differentially modulate rat pituitary GHS receptor and GHRH receptor messenger ribonucleic acid levels. Endocrinology 140, 3581-6.

Kioussis, D., and Festenstein, R. (1997). Locus control regions: overcoming heterochromatin-induced gene inactivation in mammals. Curr Opin Genet Dev 7, 614-9.

Kirk, S. E., Gertz, B. J., Schneider, S. H., Hartman, M. L., Pezzoli, S. S., Wittreich, J. M., Krupa, D. A., Seibold, J. R., and Thomer, M. O. (1997). Effect of obesity and feeding on the growth hormone (GH) response to the GH secretagogue L-692,429 in young men. J Clin Endocrinol Metab 82, 1154-9.

Koch, B. D., Blalock, J. B., and Schonbrunn, A. (1988). Characterization of the cyclic AMP-independent actions of somatostatin in GH cells. I. An increase in potassium conductance is responsible for both the hyperpolarization and the decrease in intracellular free calcium produced by somatostatin. J Biol Chem 263, 216-25.

Koch, W. J., Lefkowitz, R. J., and Rockman, H. A. (2000). Functional consequences of altering myocardial adrenergic receptor signaling. Annu Rev Physiol 62, 237-60.

201 References

Kojima, M., Haruno, R., Nakazato, M., Date, Y., Murakami, N., Hanada, R., Matsuo, H., and Kangawa, K. (2000b). Purification and identification of neuromedin U as an endogenous ligand for an orphan receptor GPR66 (FM3). Biochem Biophys Res Commun 276,435-8.

Kojima, M., Hosoda, H., Date, Y., Nakazato, M., Matsuo, H,, and Kangawa, K. (1999). Ghrelin is a growth-hormone-releasing acylated peptide from stomach. Nature 402, 656- 60.

Kojima, M., Hosoda, H., Date, Y., Nakazato, M., Matsuo, H., and Kangawa, K. (2000a). Ghrelin, a novel growth hormone-releasing acetylated peptide from stomach (Abstract). In: 3rd Intern. Symp. on GHS (Keystone, CO).

Komatsu, Y., Yasoda, A., Sakuma, Y., Miura, A., Tanaka, K., Hosoda, H., Kojima, M., Kangawa, K., Hosoda, K., and Nakao, K. (2000c). Expression of the novel Growth Hormone Releasing Peptide, Ghrelin in the osteoblastic cells (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 274.

Komminoth, P. (1992). Digoxigenin as an alternative probe labeling for in situ hybridization. Diagn Mol Pathol 1, 142-50.

Korbonits, M., and Grossman, A. B. (1995). Growth-hormone-releasing peptide and its analogues. Novel stimuli to growth hormone release. Trends Endocr Metab 6, 43-49.

Korbonits, M., Jacobs, R. A., Aylwin, S. J., Burrin, J. M., Dahia, P. L., Monson, J. P., Honegger, J., Fahlbush, R., Trainer, P. J., Chew, S. L., Besser, G. M., and Grossman, A. B. (1998). Expression of the growth hormone secretagogue receptor in pituitary adenomas and other neuroendocrine tumors. J Clin Endocrinol Metab 83, 3624-30.

Korbonits, M., Kaltsas, G., Perry, L. A., Putignano, P., Grossman, A. B., Besser, G. M., and Trainer, P. J. (1999b). The growth hormone secretagogue hexarelin stimulates the hypothalamo-pituitary-adrenal axis via arginine vasopressin. J Clin Endocrinol Metab 84, 2489-95.

Korbonits, M., Little, J. A., Forsling, M. L., Tringali, G., Costa, A., Navarra, P., Trainer, P. J., and Grossman, A. B. (1999a). The effect of growth hormone secretagogues and neuropeptide Y on hypothalamic hormone release from acute rat hypothalamic explants. J Neuroendocrinol 11, 521-8.

Komberg, R. D. (1996). RNA polymerase II transcription control. Trends Biochem Sci 21, 325-6.

Kovacs, M., Kineman, R. D., Schally, A. V., Flerko, B., and Frohman, L. A. (2000). Increase in mRNA concentrations of pituitary receptors for growth hormone-releasing hormone and growth hormone secretagogues after neonatal monosodium glutamate treatment. J Neuroendocrinol 12, 335-41.

202 References

Kraft, L. A., Baker, P. K., Doscher, M. E., and Ricks, C. A. (1984). Effect of a synthetic growth hormone releasing hexapeptide on serum growth hormone levels in steers. Journal of Animal Science 59, 218-225.

Kuriyama, H., Hotta, M., Wakabayashi, I., and Shibasaki, T. (2000). A 6-day intracerebroventricular infusion of the growth hormone-releasing peptide KP-102 stimulates food intake in both non-stressed and intermittently-stressed rats. Neurosci Lett 282, 109-12.

Lanneau, C., Bluet-Pajot, M. T., Zizzari, P., Csaba, Z., Doumaud, P., Helboe, L., Hoyer, D., Pellegrini, E., Tannenbaum, G. S., Epelbaum, J., and Gardette, R. (2000). Involvement of the Sstl somatostatin receptor subtype in the intrahypothalamic neuronal network regulating growth hormone secretion: an in vitro and in vivo antisense study. Endocrinology 141, 967-79.

Laron, Z. (1995). Growth hormone secretagogues. Clinical experience and therapeutic potential. Drugs 50, 595-601.

Laron, Z., Pertzelan, A., and Mannheimer, S. (1966). Genetic pituitary dwarfism with high serum concentation of growth hormone—a new inborn error of metabolism? Isr J Med Sci 2, 152-5.

Leal-Cerro, A., Garcia, E., Astorga, R., Casanueva, F. P., and Dieguez, C. (1995). Growth hormone (GH) responses to the combined administration of GH-releasing hormone plus GH-releasing peptide 6 in adults with GH deficiency. Eur J Endocrinol 132,712-5.

Leng, G., Bailey, A. R. T., Dickson, S. L., and Smith, R. G. (1998). Electrophysiological studies of the arcuate nucleus: Central actions of growth hormone secretagogues. In: Growth Hormone Secretagogues in Clinical Practice, B. B. Bercu and R. F. Walker, eds.,(Marcel Dekker), pp. 173-186.

Leng, G., Brown, C. H., and Russell, J. A. (1999). Physiological pathways regulating the activity of magnocellular neurosecretory cells. Prog Neurobiol 57, 625-55.

Leung, D. W., Spencer, S. A., Cachianes, G., Hammonds, R. G., Collins, C., Henzel, W. J., Barnard, R., Waters, M. J., and Wood, W. I. (1987). Growth hormone receptor and serum binding protein: purification, cloning and expression. Nature 330, 537-43.

Leung, K. H., Cohn, D. A., Miller, R. R., Doss, G. A., Steams, R. A., Simpson, R. E., Feeney, W. P., and Chiu, S. H. (1996). Pharmacokinetics and disposition of L-692,429. A novel nonpeptidyl growth hormone secretagogue in preclinical species. Dmg Metab Dispos 24, 753-60.

Lewis, U. J. (1992). Growth hormone: what is it and what does it do? Trends Endocr Metab 3, 117-121.

203 References

Li, C. H., and Evans, H. M. (1969). Human pituitary growth hormone: The primary structure of the hormone. Arch. Biochem. Biophys. 133, 70-91.

Li, C. H., and Evans, H. M. (1944). The isolation of pituitary growth hormone. Science 99, 183-184.

Li, C. H., and Papkoff, H. (1956). Preparation and properties of human growth hormone from human and monkey pituitary glands. Science 124, 1293-1294.

Li, S., Crenshaw, E. B. d., Rawson, E. J., Simmons, D. M., Swanson, L. W., and Rosenfeld, M. G. (1990). Dwarf locus mutants lacking three pituitary cell types result from mutations in the POU-domain gene pit-1. Nature 347, 528-33.

Lieber, R. L., Jacks, T. M., Mohler, R. L., Schleim, K., Haven, M., Cuizon, D., Gershuni, D. H., Lopez, M. A., Hora, D., Jr., Nargund, R., Feeney, W., and Hickey, G. J. (1997). Growth hormone secretagogue increases muscle strength during remobilization after canine hindlimb immobilization. J Orthop Res 15, 519-27.

Lin, C., Lin, S. C., Chang, C. P., and Rosenfeld, M. G. (1992). Pit-1-dependent expression of the receptor for growth hormone releasing factor mediates pituitary cell growth. Nature 360, 765-8.

Lin, S. C., Lin, C. R., Gukovsky, I., Lusis, A. J., Sawchenko, P. E., and Rosenfeld, M. G. (1993). Molecular basis of the little mouse phenotype and implications for cell type- specific growth. Nature 364, 208-13.

Ling, N., Esch, F., Bohlen, P., Brazeau, P., Wehrenberg, W. B., and Guillemin, R. (1984). Isolation, primary structure, and synthesis of human hypothalamic somatocrinin: growth hormone-releasing factor. Proc Natl Acad Sci U S A 81, 4302-6.

Linzer, D. L, and Talamantes, F. (1985). Nucleotide sequence of mouse prolactin and growth hormone mRNAs and expression of these mRNAs during pregnancy. J Biol Chem 260, 9574-9.

Liposits, Z., Merchenthaler, I., Pauli, W. K., and Flerko, B. (1988). Synaptic communication between somatostatinergic axons and growth hormone-releasing factor (GRF) synthesizing neurons in the arcuate nucleus of the rat. Histochemistry 89, 247-52.

Liu, J. L., Yakar, S., and LeRoith, D. (2000). Conditional knockout of mouse insulin­ like growth factor-1 gene using the Cre/loxP system. Proc Soc Exp Biol Med 223, 344- 51.

Liu, J. P., Baker, J., Perkins, A. S., Robertson, E. J., and Efstratiadis, A. (1993). Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igflr). Cell 75, 59-72.

204 References

Locatelli, V., Rossoni, G., Schweiger, F., Torsello, A., De Gennaro Colonna, V., Bemareggi, M., Deghenghi, R., Muller, E. E., and Berti, F. (1999). Growth hormone- independent cardioprotective effects of hexarelin in the rat. Endocrinology 140, 4024- 31.

Loche, S., Cambiaso, P., Merola, B., Colao, A., Faedda, A., Imbimbo, B. P., Deghenghi, R., Lombardi, G., and Cappa, M. (1995). The effect of hexarelin on growth hormone (GH) secretion in patients with GH deficiency. J Clin Endocrinol Metab 80, 2692-6.

Locke, W., Kirgis, H. D., Bowers, C. Y., and Abdoh, A. A. (1995). Intracerebroventricular growth-hormone-releasing peptide-6 stimulates eating without affecting plasma growth hormone responses in rats. Life Sci 56, 1347-52.

Luoni, M., Grilli, R., Guidi, M., Deghenghi, R., Torsello, A., Muller, E. E., and Locatelli, V. (1997). Effects of acute and long-term administration of of growth hormone-releasing peptides on GH secretion and feeding behaviour in young-adult and aged rats (Abstract). In: The Endocrine Society's 79th Annual Meeting (Minneapolis, MN), pp. 152.

MacLeod, J. N., Lee, A. K., Liebhaber, S. A., and Cooke, N. E. (1992). Developmental control and alternative splicing of the placentally expressed transcripts from the human growth hormone gene cluster. J Biol Chem 267, 14219-26.

Magoulas, C., McGuinness, L., Balthasar, N., Carmignac, D. F., Sesay, A. K., Mathers, K. E., Christian, H., Candeil, L., Bonnefont, X., Mollard, P., and Robinson, I. C. A. F. (2000). A secreted fluorescent reporter targeted to pituitary growth hormone cells in transgenci mice. Endocrinology (in press).

Maheshwari, H. G., Rahim, A., Shalet, S. M., and Baumann, G. (1999). Selective lack of growth hormone (GH) response to the GH-releasing peptide hexarelin in patients with GH-releasing hormone receptor deficiency. J Clin Endocrinol Metab 84, 956-9.

Mallo, F., Alvarez, C. V., Benitez, L., Burguera, B., Coy a, R., Casanueva, F. F., and Dieguez, C. (1993). Regulation of His-dTrp-Ala-Trp-dPhe-Lys-NH2 (GHRP-6)-induced GH secretion in the rat. Neuroendocrinology 57, 247-56.

Malozowski, S., Hao, E. H., Ren, S. G., Marin, G., Liu, L., Southers, J. L., and Merriam, G. R. (1991). Growth hormone (GH) responses to the hexapeptide GH-releasing peptide and GH-releasing hormone (GHRH) in the cynomolgus macaque: evidence for non- GHRH-mediated responses. J Clin Endocrinol Metab 73, 314-7.

Massoud, A. F., Hindmarsh, P. C., and Brook, C. G. (1996). Hexarelin-induced growth hormone, cortisol, and prolactin release: a dose-response study. J Clin Endocrinol Metab 81,4338-41.

205 References

Masuda, Y., Tanaka, T., Inomata, N., Ohnuma, N., Tanaka, S., Itoh, Z., Hosoda, H., Kojima, M., and Kangawa, K. (2000). Ghrelin stimulates gastric acid secretion and motility in rats. Biochem Biophys Res Commun 276, 905-8.

Matsubara, S., Sato, M., Mizobuchi, M., Niimi, M., and Takahara, J. (1995). Differential Gene-Expression of Growth-Hormone (Gh)- (Grh) and Grh Receptor in Various Rat- Tissues. Endocrinology 136,4147-4150.

Mayo, K. E. (1992). Molecular cloning and expression of a pituitary-specific receptor for growth hormone-releasing hormone. Mol Endocrinol 6,1734-44.

Mayo, K. E. (1989). Structure and expression of growth hormone-releasing hormone (GHRH) genes. In: Advances in Growth Hormone and Growth Factor Research, E. E. Muller, D. Cocchi and V. Locatelli, eds. (Milan: Pythagora), pp. 217-230.

Mayo, K. E., Cerelli, G. M., Rosenfeld, M. G., and Evans, R. M. (1985). Characterization of cDNA and genomic clones encoding the precursor to rat hypothalamic growth hormone-releasing factor. Nature 314,464-7.

Mayo, K. E., Godfrey, P. A., De Almeida, V., and Millner, T. L. (1996). Structure, function and regulation of the pituitary receptor for growth hormone releasing hormone. In: Growth Hormone Secretagogues, B. B. Bercu and R. F. Walker, eds. (New York: Springer-Verlag), pp. 53-71.

Mayo, K. E., Godfrey, P. A., Suhr, S. T., Kulik, D. J., and Rahal, J. O. (1995). Growth Hormone-Releasing Hormone - Synthesis and Signaling. Recent Prog Horm Res 50, 35- 73.

Mayo, K. E., Hammer, R. E., Swanson, L. W., Brinster, R. L., Rosenfeld, M. G., and Evans, R. M. (1988). Dramatic pituitary hyperplasia in transgenic mice expressing a human growth hormone-releasing factor gene. Mol Endocrinol 2, 606-12.

Mayo, K. E., Vale, W., Rivier, J., Rosenfeld, M. G., and Evans, R. M. (1983). Expression-cloning and sequence of a cDNA encoding human growth hormone- releasing factor. Nature 306, 86-8.

McCarthy, G. P., Beaudet, A., and Tannenbaum, G. S. (1992). Colocalization of somatostatin receptors and growth hormone-releasing factor immunoreactivity in neurons of the rat arcuate nucleus. Neuroendocrinology 56, 18-24.

McCormick, A., Brady, H., Theill, L. E., and Karin, M. (1990). Regulation of the pituitary-specific homeobox gene GHFl by cell-autonomous and environmental cues. Nature 345, 829-32.

McDowell, R. S., Elias, K. A., Stanley, M. S., Burdick, D. J., Bumier, J. P., Chan, K. S., Fairbrother, W. J., Hammonds, R. G., Ingle, G. S., Jacobsen, N. E., and et al. (1995).

206 References

Growth hormone secretagogues: characterization, efficacy, and minimal bioactive conformation. Proc Natl Acad Sci U S A 92, 11165-9.

McGurk, J. F., Pong, S.-S., Chaung, L.-Y. P., Gall, M,, Butler, B. S., and Arena, J. P. (1993). Growth Hormone Secretagogues modulate potassium currents in rat somatotrophs (Abstract). In: The 23rd Annual Meeting of the Society of Neuroscience (Washington, DC), pp. 1559.

McKee, K. K., Palyha, O. C., Feighner, S. D., Hreniuk, D. L., Tan, C. P., Phillips, M. S., Smith, R. G., Van der Ploeg, L. H., and Howard, A. D. (1997a). Molecular analysis of rat pituitary and hypothalamic growth hormone secretagogue receptors. Mol Endocrinol 11,415-23.

McKee, K. K., Tan, C. P., Palyha, O. C., Liu, J., Feighner, S. D., Hreniuk, D. L., Smith, R. G., Howard, A. D., and Van der Ploeg, L. H. (1997b). Cloning and characterization of two human G protein-coupled receptor genes (GPR38 and GPR39) related to the growth hormone secretagogue and neurotensin receptors. Genomics 46,426-34.

Merchenthaler, I., Vigh, S., Schally, A. V., and Petrusz, P. (1984). Immunocytochemical localization of growth hormone-releasing factor in the rat hypothalamus. Endocrinology 114, 1082-5.

Mericq, F., Cassorla, F., Salazar, T., Avila, A., Iniguez, G., Bowers, C. Y., and Merriam, G. R. (1995). Increased growth velocity during prolonged GHRP-2 administration to growth hormone deficient children (Abstract). In: The Endocrine Society's 77th Annual Meeting (Washington, DC), pp. OR30-3.

Merriam, G. R., Buchner, D. M., Prinz, P. N., Schwartz, R. S., and Vitiello, M. V. (1997). Potential applications of GH secretagogs in the evaluation and treatment of the age-related decline in growth hormone secretion. Endocrine 7,49-52.

Micic, D., Casabiell, X., Gualillo, O., Pombo, M., Dieguez, C., and Casanueva, F. F. (1999). Growth hormone secretagogues: the clinical future. Horm Res 51 Suppl 3, 29- 33.

Miki, N., Ono, M., Asakawa-Yasumoto, K., Aoki, T., Murata, Y., Ishituka, Y., Demura, H., and Sasaki, F. (1994). Characterization and localization of mouse hypothalamic growth hormone-releasing factor and effect of gold thioglucose-induced hypothalamic lesions. J Neuroendocrinol 6, 71-8.

Miller, J. D., Tannenbaum, G. S., Colle, E., and Guyda, H. J. (1982). Daytime pulsatile growth hormone secretion during childhood and adolescence. J Clin Endocrinol Metab 55, 989-94.

Miller, T. L., and Mayo, K. E. (1997). Glucocorticoids regulate pituitary growth hormone-releasing hormone receptor messenger ribonucleic acid expression. Endocrinology 138, 2458-65.

207 References

Minami, S., Kamegai, J., Sugihara, H., Hasegawa, O., and Wakabayashi, I. (1992). Systemic administration of recombinant human growth hormone induces expression of the c-fos gene in the hypothalamic arcuate and periventricular nuclei in hypophysectomized rats. Endocrinology 131, 247-53.

Misra-Press, A., Cooke, N. B., and Liebhaber, S. A. (1994). Complex alternative splicing partially inactivates the human chorionic somatomammotropin-like (hCS-L) gene. J Biol Chem 269, 23220-9.

Mitani, M., Kaji, H., Abe, H., and Chihara, K. (1996). Growth hormone (GH)-releasing peptide and GH releasing hormone stimulate GH release from subpopulations of somatotrophs in rats. J Neuroendocrinol 8, 825-30.

Mori, K., Yoshimoto, A., Hosoda, K., Sawai, K., Yokoi, H., Yoshioka, T., Nagae, T., Fujinaga, Y., Makino, H., Suganami, T., Yahata, K., Mukoyama, M., Sugawara, A., Hosoda, H., Kojima, M., Kangawa, K., and Nakao, K. (2000). Ghrelin Gene Expression in the Kidney, Glomerulus, and Glomerular Cells (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 273.

Morrison, M., and Bay se, G. S. (1970). Catalysis of iodination by lactoperoxidase. Biochemistry 9, 2995-3000.

Murphy, M. G., Bach, M. A., Plotkin, D., Bolognese, J., Ng, J., Krupa, D., Cerchio, K., and Gertz, B. J. (1999). Oral administration of the growth hormone secretagogue MK- 677 increases markers of bone turnover in healthy and functionally impaired elderly adults. The MK-677 Study Group. J Bone Miner Res 14, 1182-8.

Murphy, M. G., Plunkett, L. M., Gertz, B. J., He, W., Wittreich, J., Polvino, W. M., and Clemmons, D. R. (1998). MK-677, an orally active growth hormone secretagogue, reverses diet-induced catabolism. J Clin Endocrinol Metab 83, 320-5.

Nakazato, M., Date, Y., Kojima, M., Kangawa, K., and Matsukura, S. (2000). Ghrelin, a novel growth-hormone-releasing acetylated peptide, synthesized in a distinct endocrine cell type in rats and humans (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 171.

Nargund, R. P., and van der Ploeg, L. H. T. (1997). Growth Hormone Secretagogues. Annual Reports in Medicinal Chemistry 32, 221-230.

Nass, R., Gilrain, J., Anderson, S., Gaylinn, B., Dalkin, A., Day, R., Peruggia, M., and Thomer, M. O. (2000). High plasma growth hormone (GH) levels inhibit expression of GH secretagogue receptor messenger ribonucleic acid levels in the rat pituitary. Endocrinology 141, 2084-9.

208 References

Niimi, M., Takahara, J., Sato, M., and Kawanishi, K. (1989). Sites of origin of growth hormone-releasing factor-containing neurons projecting to the stalk-median eminence of the rat. Peptides 10, 605-8.

Nikolov, D. B., Chen, H., Halay, E. D., Hoffman, A., Boeder, R. G., and Burley, S. K. (1996). Crystal structure of a human TATA box-binding protein/TATA element complex. Proc Natl Acad Sci U S A 93,4862-7.

Nillni, E. A., Steinmetz, R., and Pescovitz, O. H. (1999). Posttranslational processing of progrowth hormone-releasing hormone. Endocrinology 140, 5817-27.

Nurhidayat, Tsukamoto, Y., Sigit, K., and Sasaki, F. (1999). Sex differentiation of growth hormone-releasing hormone and somatostatin neurons in the mouse hypothalamus: an immunohistochemical and morphological study. Brain Res 821, 309- 21.

O'Carroll, A. M., and Krempels, K. (1995). Widespread distribution of somatostatin receptor messenger ribonucleic acids in rat pituitary. Endocrinology 136, 5224-7.

Okada, K., Ishii, S., Minami, S., Sugihara, H., Shibasaki, T., and Wakabayashi, I. (1996). Intracerebroventricular administration of the growth hormone-releasing peptide KP-102 increases food intake in free-feeding rats. Endocrinology 137, 5155-8.

Ong, H., Bodart, V., McNicoll, N., Lamontagne, D., and Bouchard, J. F. (1998a). Binding sites for growth hormone-releasing peptide. Growth Horm IGF Res 8 Suppl B, 137-40.

Ong, H., McNicoll, N., Escher, E., Collu, R., Deghenghi, R., Locatelli, V., Ghigo, E., Muccioli, G., Boghen, M., and Nilsson, M. (1998b). Identification of a pituitary growth hormone-releasing peptide (GHRP) receptor subtype by photoaffinity labeling. Endocrinology 139,432-5.

Ono, M., Miki, N., Murata, Y., Osaki, E., Tamitsu, K., Ri, T., Yamada, M., and Demura, H. (1995). Sexually dimorphic expression of pituitary growth hormone-releasing factor receptor in the rat. Biochem Biophys Res Commun 216, 1060-6.

Palmiter, R. D., and Brinster, R. L. (1986). Germ-line transformation of mice. Annu Rev Genet 20, 465-99.

Palyha, O., Yu, H., McKee, K., Feighner, S., Hreniuk, D., VanDerPloeg, L., Howard, A., and Guan, X. (1997). Distribution of mRNA of growth hormone secretagogue receptor in brain of the rat. Faseb Journal 11, 2271.

Patchett, A. A., Nargund, R. P., Tata, J. R., Chen, M. H., Barakat, K. J., Johnston, D. B., Cheng, K., Chan, W. W., Butler, B., Hickey, G., and et al. (1995). Design and biological activities of L-163,191 (MK-0677): a potent, orally active growth hormone secretagogue. Proc Natl Acad Sci U S A 92, 7001-5.

209 References

Patel, Y. C. (1997). Molecular pharmacology of somatostatin receptor subtypes. J Endocrinol Invest 20, 348-67.

Patel, Y. C. (1990). Somatostatin. In: Principles and Practice of Endocrinology and Metabolism, K. Becker, ed. (Philadelphia, PA: Lipincott), pp. 1297-1301.

Patel, Y. C., Greenwood, M. T., Panetta, R., Demchyshyn, L., Niznik, H., and Srikant, C. B. (1995). The somatostatin receptor family. Life Sci 57, 1249-65.

Petersenn, S., Penshom, M., Beil, F. U., and Schulte, H. M. (1999). [Molecular analysis of the human "growth hormone secretagogue "-receptor]. Med Klin 94, 202-6.

Phelps, C. J., Dalcik, H., Endo, H., Talamantes, P., and Hurley, D. L. (1993). Growth hormone-releasing hormone peptide and mRNA are overexpressed in GH-deficient Ames dwarf mice. Endocrinology 133, 3034-7.

Phelps, C. J., and Hurley, D. L. (1999). Pituitary hormones as neurotrophic signals: update on hypothalamic differentiation in genetic models of altered feedback. Proc Soc Exp Biol Med 222, 39-58.

Pihoker, C., Badger, T. M., Reynolds, G. A., and Bowers, C. Y. (1997). Treatment effects of intranasal growth hormone releasing peptide-2 in children with short stature. J Endocrinol 155, 79-86.

Plotsky, P. M., and Vale, W. (1985). Patterns of growth hormone-releasing factor and somatostatin secretion into the hypophysial-portal circulation of the rat. Science 230, 461-3.

Pomes, A., Pong, S. S., and Schaeffer, J. M. (1996). Solubilization and characterization of a growth hormone secretagogue receptor from porcine anterior pituitary membranes. Biochem Biophys Res Commun 225, 939-45.

Pong, S. S., Chaung, L. Y., Dean, D. C., Nargund, R. P., Patchett, A. A., and Smith, R. G. (1996). Identification of a new G-protein-linked receptor for growth hormone secretagogues. Mol Endocrinol 10, 57-61.

Popovic, V., Damjanovic, S., Micic, D., Djurovic, M., Dieguez, C., and Casanueva, F. F. (1995). Blocked growth hormone-releasing peptide (GHRP-6)-induced GH secretion and absence of the synergic action of GHRP-6 plus GH-releasing hormone in patients with hypothalamopituitary disconnection: evidence that GHRP-6 main action is exerted at the hypothalamic level. J Clin Endocrinol Metab 80, 942-7.

Popovic, V., Damjanovic, S., Micic, D., Petakov, M., Dieguez, C., and Casanueva, F. F. (1994). Growth hormone (GH) secretion in active acromegaly after the combined administration of GH-releasing hormone and GH-releasing peptide-6. J Clin Endocrinol Metab 79, 456-60.

210 References

Powell-Braxton, L., Hollingshead, P., Warburton, C., Dowd, M., Pitts-Meek, S., Dalton, D., Gillett, N., and Stewart, T. A. (1993). IGF-I is required for normal embryonic growth in mice. Genes Dev 7, 2609-17.

Rahim, A., O'Neill, P. A., and Shalet, S. M. (1998). Growth hormone status during long­ term hexarelin therapy. J Clin Endocrinol Metab 83, 1644-9.

Ramirez, J. L., Castano, J. P., Torronteras, R., Martinez-Fuentes, A. J., Frawley, L. S., Garcia-Navarro, S., and Gracia-Navarro, F. (1999). Growth hormone (GH)-releasing factor differentially activates cyclic adenosine 3',5'-monophosphate- and inositol phosphate-dependent pathways to stimulate GH release in two porcine somatotrope subpopulations. Endocrinology 140, 1752-9.

Reichlin, S. (1960). Growth and the hypothalamus. Endocrinology 67, 760-773.

Reichlin, S. (1983). Somatostatin. In: Brain peptides, D. T. Krieger, M. O. Brownstein and J. B. Martin, eds. (New York: Wiley), pp. 711-742.

Rettori, V., Milenkovic, L., Aguila, M. C., and McCann, S. M. (1990). Physiologically significant effect of neuropeptide Y to suppress growth hormone release by stimulating somatostatin discharge. Endocrinology 126, 2296-301.

Rivier, J., Spiess, J., Thomer, M., and Vale, W. (1982). Characterization of a growth hormone-releasing factor from a human pancreatic islet tumour. Nature 300, 276-8.

Robinson, I. C. (1980). The development and evaluation of a sensitive and specific radioimmunoassay for oxytocin in unextracted plasma. J Immunoassay 1, 323-47.

Roeder, R. G. (1996). The role of general initiation factors in transcription by RNA polymerase H. Trends Biochem Sci 21, 327-35.

Romero, M. I., and Phelps, C. J. (1997). Identification of growth hormone-releasing hormone and somatostatin neurons projecting to the median eminence in normal and growth hormone-deficient Ames dwarf mice. Neuroendocrinology 65, 107-16.

Rossoni, G., Locatelli, V., Gennaro Colonna, V., Muller, E. E., and Berti, F. (2000). Hexarelin, a growth hormone secretagogue, protects the isolated rat heart from ventricular dysfunction produced by exposure to calcium-free medium. Pharmacol Res 42, 129-136.

Rudman, D., Feller, A. G., Nagraj, H. S., Gergans, G. A., Lalitha, P. Y., Goldberg, A. F., Schlenker, R. A., Cohn, L., Rudman, I. W., and Mattson, D. E. (1990). Effects of human growth hormone in men over 60 years old. N Engl J Med 323, 1-6.

211 References

Sambrook, J., Fritsch, E. P., and Maniatis, T. (1989). Molecular Cloning. A Laboratory Manual. 2nd edition, N. Ford, C. Nolan and M. Ferguson, eds.: Cold Sprin Harbor Laboratory Press).

Sartor, O., Bowers, C. Y., Reynolds, G. A., and Momany, F. A. (1985). Variables determining the growth hormone response of His-D-Trp-Ala-Trp-D-Phe-LyS-NH2 in the rat. Endocrinology 117, 1441-7.

Sato, M., and Frohman, L. A. (1993). Differential effects of central and peripheral administration of growth hormone (GH) and insulin-like growth factor on hypothalamic GH-releasing hormone and somatostatin gene expression in GH-deficient dwarf rats. Endocrinology 133, 793-9.

Sawchenko, P. E., Swanson, L. W., Rivier, J., and Vale, W. W. (1985). The distribution of growth-hormone-releasing factor (GRF) immunoreactivity in the central nervous system of the rat: an immunohistochemical study using antisera directed against rat hypothalamic GRF. J Comp Neurol 237, 100-15.

Schleim, K. D., Jacks, T., Cunningham, P., Feeney, W., Frazier, E. G., Niebauer, G. W., Zhang, D., Chen, H., Smith, R. G., and Hickey, G. (1996). Effect of MK-677 on growth hormone, IGF-I and cortisol in beagles before and after hypophysectomy (Abstract). In: The 10th International Congress of Endocrinology (San Francisco CA), pp. 603.

Schoen, W. R., Pisano, J. M., Prendergast, K., Wyvratt, M. J., Jr., Fisher, M. H., Cheng, K., Chan, W. W., Butler, B., Smith, R. G., and B^l, R. G. (1994). A novel 3-substituted benzazepinone growth hormone secretagogue (L-692,429). J Med Chem 37, 897-906.

Seeburg, P. H., Shine, J., Martial, J. A., Baxter, J. D., and Goodman, H. M. (1977). Nucleotide sequence and amplification in bacteria of structural gene for rat growth hormone. Nature 270, 486-94.

Senaris, R. M., Humphrey, P. P., and Emson, P. C. (1994). Distribution of somatostatin receptors 1, 2 and 3 mRNA in rat brain and pituitary. Eur J Neurosci 6, 1883-96.

Sethumadhavan, K., Veeraragavan, K., and Bowers, C. Y. (1991). Demonstration and characterization of the specific binding of growth hormone-releasing peptide to rat anterior pituitary and hypothalamic membranes. Biochem Biophys Res Commun 178, 31-7.

Sheng, M., Thompson, M. A., and Greenberg, M. E. (1991). CREB: a Ca(2+)-regulated transcription factor phosphorylated by calmodulin-dependent kinases. Science 252, 1427-30.

Shimomura, O., Johnson, F. H., and Saiga, Y. (1962). J Cell Comp Physiol 59, 223-239.

212 References

Simard, J., Labrie, F., and Gossard, F. (1986). Regulation of growth hormone mRNA and pro-opiomelanocortin mRNA levels by cyclic AMP in rat anterior pituitary cells in culture. Dna 5, 263-70.

Simmons, D. M., Voss, J. W., Ingraham, H. A., Holloway, J. M., Broide, R. S., Rosenfeld, M, G., and Swanson, L. W. (1990). Pituitary cell phenotypes involve cell- specific Pit-1 mRNA translation and synergistic interactions with other classes of transcription factors. Genes Dev 4, 695-711.

Sims, S. M., Lussier, B. T., and Kraicer, J. (1991). Somatostatin activates an inwardly rectifying K+ conductance in freshly dispersed rat somatotrophs. J Physiol (Lond) 441, 615-37.

Sirinathsinghji, D. J., Chen, H. Y., Hopkins, R., Trumbauer, M., Heavens, R., Rigby, M., Smith, R. G., and van der Ploeg, L. H. (1995). Induction of c-fos mRNA in the arcuate nucleus of normal and mutant growth hormone-deficient mice by a synthetic non- peptidyl growth hormone secretagogue. Neuroreport 6, 1989-92.

Skinner, M. M., Nass, R., Lopes, B., Laws, E. R., and Thomer, M. O. (1998). Growth hormone secretagogue receptor expression in human pituitary tumors. J Clin Endocrinol Metab 83, 4314-20.

Smith, R. G., Feighner, S., Prendergast, K., Guan, X., and Howard, A. (1999). A New Orphan Receptor Involved in Pulsatile Growth Hormone Release. Trends Endocr Metab 10, 128-135.

Smith, R. G., Griffin, P. R., Xu, Y., Smith, A. G., Liu, K., Calacay, J., Feighner, S. D., Pong, C., Leong, D., Pomes, A., Cheng, K., Van der Ploeg, L. H., Howard, A. D., Schaeffer, J., and Leonard, R. J. (2000). Adenosine: A partial agonist of the growth hormone secretagogue receptor. Biochem Biophys Res Commun 276, 1306-13.

Smith, R. G., Pong, S. S., Hickey, G., Jacks, T., Cheng, K., Leonard, R., Cohen, C. J., Arena, J. P., Chang, C. H., Drisko, J., Wyvratt, M., Fisher, M., Nargund, R., and Patchett, A. (1996). Modulation of pulsatile GH release through a novel receptor in hypothalamus and pituitary gland. Recent Prog Horm Res 51, 261-85; discussion 285-6.

Smith, R. G., Van der Ploeg, L. H., Howard, A. D., Feighner, S. D., Cheng, K., Hickey, G. J., Wyvratt, M. J., Jr., Fisher, M. H., Nargund, R. P., and Patchett, A. A. (1997). Peptidomimetic regulation of growth hormone secretion. Endocr Rev 18, 621-45.

Somson, M. W., Wu, W., Dasen, J. S., Flynn, S. E., Norman, D. J., O'Connell, S. M., Gukovsky, I., Carrière, C., Ryan, A. K., Miller, A. P., Zuo, L., Gleiberman, A. S., Andersen, B., Beamer, W. G., and Rosenfeld, M. G. (1996). Pituitary lineage determination by the Prophet of Pit-1 homeodomain factor defective in Ames dwarfism. Nature 384, 327-33.

213 References

Sotiropoulos, A., Goujon, L., Simonin, G., Kelly, P. A., Postel-Vinay, M. C., and Finidori, J. (1993). Evidence for generation of the growth hormone-binding protein through proteolysis of the growth hormone membrane receptor. Endocrinology 132, 1863-5.

Soto, J. L., Castrillo, J. L., Dominguez, P., and Dieguez, C. (1995). Regulation of the pituitary-specific transcription factor GHF-1/Pit-1 messenger ribonucleic acid levels by growth hormone-secretagogues in rat anterior pituitary cells in monolayer culture. Endocrinology 136, 3863-70.

Southern, E. M. (1975). Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 98, 503-17.

Spergel, D. J., Kruth, U., Hanley, D. P., Sprengel, R., and Seeburg, P. H. (1999). GAB A- and glutamate-activated channels in green fluorescent protein-tagged gonadotropin-releasing hormone neurons in transgenic mice. J Neurosci 19, 2037-50.

Spiess, J., Rivier, J., and Vale, W. (1983). Characterization of rat hypothalamic growth hormone-releasing factor. Nature 303, 532-5.

Steger, R. W., Bartke, A., Barkening, T. A., Collins, T., Buonomo, P. C., Tang, K. C., Wagner, T. E., and Yun, J. S. (1991). Effects of heterologous growth hormones on hypothalamic and pituitary function in transgenic mice. Neuroendocrinology 53, 365-72,

Steiner, R. A,, Stewart, J. K., Barber, J., Koerker, D., Goodner, C. J., Brown, A., Illner, P., and Gale, C. C. (1978). Somatostatin: a physiological role in the regulation of growth hormone secretion in the adolescent male baboon. Endocrinology 102, 1587-94.

Struthers, R. S., Vale, W. W., Arias, C., Sawchenko, P. E., and Montminy, M. R. (1991). Somatotroph hypoplasia and dwarfism in transgenic mice expressing a non- phosphorylatable CREB mutant. Nature 350, 622-4.

Su, Y., Liebhaber, S. A., and Cooke, N. E. (2000). The human growth hormone gene cluster locus control region supports position-independent pituitary- and placenta- specific expression in the transgenic mouse. J Biol Chem 275, 7902-9.

Sugihara, H., Kamegai, J., Ishii, S., Tamura, H., Shuto, Y., Oikawa, S., Shibasaki, T., and Wakabayashi, I. (2000). Ghrelin acts synergistically with GHRH on GH secretion in urethane anesthetized rats (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 171.

Suhr, S. T., Rahal, J. O., and Mayo, K. E. (1989). Mouse growth-hormone-releasing hormone: precursor structure and expression in brain and placenta. Mol Endocrinol 3, 1693-700.

Suter, K. J., Song, W. J., Sampson, T. L., Wuarin, J. P., Saunders, J. T., Dudek, P. E., and Moenter, S. M. (2000). Genetic targeting of green fluorescent protein to

214 References gonadotropin-releasing hormone neurons: characterization of whole-cell electrophysiological properties and morphology. Endocrinology 141, 412-9.

Svensson, J., Lall, S., Dickson, S. L., Bengtsson, B. A., Romer, J., Ahnfelt-Ronne, I., Ohlsson, C., and Jansson, J. O. (2000). The GH secretagogues ipamorelin and GH- releasing peptide-6 increase bone mineral content in adult female rats. J Endocrinol 165, 569-77.

Svensson, J., Lonn, L., Jansson, J. O., Murphy, G., Wyss, D., Krupa, D., Cerchio, K., Polvino, W., Gertz, B., Boseaus, I., Sjostrom, L., and Bengtsson, B. A. (1998). Two- month treatment of obese subjects with the oral growth hormone (GH) secretagogue MK-677 increases GH secretion, fat-free mass, and energy expenditure. J Clin Endocrinol Metab 83, 362-9.

Szekeres, P. G., Muir, A. I., Spinage, L. D., Miller, J. E., Butler, S. I., Smith, A., Rennie, G. I., Murdock, P. R., Fitzgerald, L. R., Wu, H., McMillan, L. J., Guerrera, S., Vawter, L., Elshourbagy, N. A., Mooney, J. L., Bergsma, D. J., Wilson, S., and Chambers, J. K. (2000). Neuromedin U is a potent agonist at the orphan G protein-coupled receptor FM3. J Biol Chem 275, 20247-50.

Tamura, H., Kamegai, J., Sugihara, H., Kineman, R. D., Frohman, L. A., and Wakabayashi, I. (2000). Glucocorticoids regulate pituitary growth hormone secretagogue receptor gene expression. J Neuroendocrinol 12, 481-5.

Tan, C. P., McKee, K. K., Liu, Q., Palyha, O. C., Feighner, S. D., Hreniuk, D. L., Smith, R. G., and Howard, A. D. (1998). Cloning and characterization of a human and murine T-cell orphan G-protein-coupled receptor similar to the growth hormone secretagogue and neurotensin receptors. Genomics 52, 223-9.

Tang, W. J., and Gilman, A. G. (1991). Type-specific regulation of adenylyl cyclase by G protein beta gamma subunits. Science 254, 1500-3.

Tannenbaum, G. S. (1991). Neuroendocrine control of growth hormone secretion. Acta Paediatr Scand Suppl 372, 5-16.

Tannenbaum, G. S., Guyda, H. J., and Posner, B. I. (1983). Insulin-like growth factors: a role in growth hormone negative feedback and body weight regulation via brain. Science 220, 77-9.

Tannenbaum, G. S., Lapointe, M., Beaudet, A., and Howard, A. D. (1998a). Expression of growth hormone secretagogue-receptors by growth hormone-releasing hormone neurons in the mediobasal hypothalamus. Endocrinology 139,4420-4423.

Tannenbaum, G. S., and Martin, J. B. (1976). Evidence for an endogenous ultradian rhythm governing growth hormone secretion in the rat. Endocrinology 98, 562-70.

215 References

Tannenbaum, G. S., Zhang, W. H., Lapointe, M., Zeitler, P., and Beaudet, A. (1998b). Growth hormone-releasing hormone neurons in the arcuate nucleus express both Sstl and Sst2 somatostatin receptor genes. Endocrinology 139, 1450-3.

Tanner, J. W., Davis, S. K., McArthur, N. H., French, J. T., and Welsh, T. H., Jr. (1990). Modulation of growth hormone (GH) secretion and GH mRNA levels by GH-releasing factor, somatostatin and secretagogues in cultured bovine adenohypophysial cells. J Endocrinol 125, 109-15.

Tashjian, A. H., Jr., Yasumura, Y., Levine, L., Sato, G. H., and Parker, M. L. (1968). Establishment of clonal strains of rat pituitary tumor cells that secrete growth hormone. Endocrinology 82, 342-52.

Thomas, G. B., Bennett, P. A., Carmignac, D. P., and Robinson, I. C. (2000). Glucocorticoid regulation of growth hormone (GH) secretagogue-induced growth responses and GH secretagogue receptor expression in the rat. Growth Horm IGF Res 10,45-52.

Thomas, G. B., Fairhall, K. M., and Robinson, I. C. (1997). Activation of the hypothalamo-pituitary-adrenal axis by the growth hormone (GH) secretagogue, GH- releasing peptide-6, in rats. Endocrinology 138, 1585-91.

Thomer, M. O., Perryman, R. L., Cronin, M. J., Rogol, A. D., Draznin, M., Johanson, A., Vale, W., Horvath, E., and Kovacs, K. (1982). Somatotroph hyperplasia. Successful treatment of acromegaly by removal of a pancreatic islet tumor secreting a growth hormone-releasing factor. J Clin Invest 70, 965-77.

Tiulpakov, A. N., Brook, C. G., Pringle, P. J., Peterkova, V. A., Volevodz, N. N., and Bowers, C. Y. (1995). GH responses to intravenous bolus infusions of GH releasing hormone and GH releasing peptide 2 separately and in combination in adult volunteers. Clin Endocrinol (Oxf) 43, 347-50.

Tivesten, A., Bollano, E., Caidahl, K., Kujacic, V., Sun, X. Y., Hedner, T., Hjalmarson, A., Bengtsson, B. A., and Isgaard, J. (2000). The growth hormone secretagogue hexarelin improves cardiac function in rats after experimental myocardial infarction. Endocrinology 141, 60-6.

Torsello, A., Grilli, R., Luoni, M., Guidi, M., Ghigo, M. C., Wehrenberg, W. B., Deghenghi, R., Muller, E. E., and Locatelli, V. (1996). Mechanism of action of Hexarelin. I. Growth hormone-releasing activity in the rat. Eur J Endocrinol 135, 481-8.

Torsello, A., Scibona, B., Di Credico, S., Avallone, R., Bresciani, E., Deghenghi, R., Muller, E. E., and Locatelli, V. (2000). Tissue-specific expression of Ghrelin mRNA in the CNS and peripheral organs of young-adult and old rats (Abstract). In: The Endocrine Society's 82nd Annual Meeting (Toronto, CAN), pp. 273.

216 References

Tschop, M., Smiley, D. L., and Heiman, M. L. (2000). Ghrellin induces adiposity in rodents. Nature 407, 908-913.

Tsien, R. Y. (1998). The green fluorescent protein. Annu Rev Biochem 67, 509-44.

Van den Berghe, G., de Zegher, F., Baxter, R. C., Veldhuis, J. D., Wouters, P., Schetz, M., Verwaest, C., Van der Vorst, E., Lauwers, P., Bouillon, R., and Bowers, C. Y. (1998). Neuroendocrinology of prolonged critical illness: effects of exogenous thyrotropin-releasing hormone and its combination with growth hormone secretagogues. J Clin Endocrinol Metab 83, 309-19.

Van den Berghe, G., de Zegher, F., Bowers, C. Y., Wouters, P., Muller, P., Soetens, F., Vlasselaers, D., Schetz, M., Verwaest, C., Lauwers, P., and Bouillon, R. (1996). Pituitary responsiveness to GH-releasing hormone, GH-releasing peptide-2 and thyrotrophin-releasing hormone in critical illness. Clin Endocrinol (Oxf) 45, 341-51.

Van den Berghe, G., de Zegher, F., Veldhuis, J. D., Wouters, P., A wouters, M., Verbruggen, W., Schetz, M., Verwaest, C., Lauwers, P., Bouillon, R., and Bowers, C. Y. (1997). The somatotropic axis in critical illness: effect of continuous growth hormone (GH)-releasing hormone and GH-releasing peptide-2 infusion. J Clin Endocrinol Metab 82, 590-9.

Van den Berghe, G., de Zegher, F., Veldhuis, J. D., Wouters, P., Gouwy, S., Stockman, W., Weekers, F., Schetz, M., Lauwers, P., Bouillon, R., and Bowers, C. Y. (1997). Thyrotrophin and prolactin release in prolonged critical illness: dynamics of spontaneous secretion and effects of growth hormone-secretagogues. Clin Endocrinol (Oxf) 47, 599-612.

Van den Berghe, G., Wouters, P., Weekers, F., Mohan, S., Baxter, R. C., Veldhuis, J. D., Bowers, C. Y., and Bouillon, R. (1999). Reactivation of pituitary hormone release and metabolic improvement by infusion of growth hormone-releasing peptide and thyrotropin-releasing hormone in patients with protracted critical illness. J Clin Endocrinol Metab 84, 1311-23.

Van Itallie, C. M., and Femstrom, J. D. (1983). In vivo studies of somatostatin-14 and somatostatin-28 biosynthesis in rat hypothalamus. Endocrinology 113, 1210-7.

Veeraragavan, K., Sethumadhavan, K., and Bowers, C. Y. (1992). Growth hormone- releasing peptide (GHRP) binding to porcine anterior pituitary and hypothalamic membranes. Life Sci 50, 1149-55.

Wajnrajch, M. P., Gertner, J. M., Harbison, M. D., Chua, S. C., Jr., and Leibel, R. L. (1996). Nonsense mutation in the human growth hormone-releasing hormone receptor causes growth failure analogous to the little (lit) mouse. Nat Genet 12, 88-90.

217 References

Wallrath, L. L., Lu, Q., Granok, H., and Elgin, S. C. (1994). Architectural variations of inducible eukaryotic promoters: preset and remodeling chromatin structures. Bioessays 16, 165-70.

Wehrenberg, W. B., Ling, N., Bohlen, P., Bsch, P., Brazeau, P., and Guillemin, R. (1982). Physiological roles of somatocrinin and somatostatin in the regulation of growth hormone secretion. Biochem Biophys Res Commun 109, 562-7.

Wells, T., Bennett, P. A., and Scarborough, D. (2000). Pattern-dependent effects of GHRP-6 on the hypothalamo-pituitary axis and skeletal growth in Transgenic Growth Retarded rats (Abstract). In: Annual Meeting of the Growth Hormone Research Society (Gothenburg, Sweden), pp. 185.

Wells, T., Flavell, D. M., Wells, S. B., Carmignac, D. P., and Robinson, I. C. (1997). Effects of growth hormone secretagogues in the transgenic growth-retarded (Tgr) rat. Endocrinology 138, 580-7.

Willesen, M. G., Kristensen, P., and Romer, J. (1999). Co-localization of growth hormone secretagogue receptor and NPY mRNA in the arcuate nucleus of the rat. Neuroendocrinology 70, 306-16.

Willoughby, J. O., Brogan, M., and Kapoor, R. (1989). Hypothalamic interconnections of somatostatin and growth hormone releasing factor neurons. Neuroendocrinology 50, 584-91.

Willoughby, J. O., Menadue, M., Zeegers, P., Wise, P. H., and Oliver, J. R. (1980). Effects of human growth hormone on the secretion of rat growth hormone. J Endocrinol 86, 165-9.

Wilson, C. A., James, M. D., Grierson, J. P., and Hole, D. R. (1991). Involvement of catecholaminergic systems in the zona incerta in the steroidal control of gonadotrophin release and female sexual behaviour. Neuroendocrinology 53, 113-23.

Wilson, D. B., Wyatt, D. P., Gadler, R. M., and Baker, C. A. (1988). Quantitative aspects of growth hormone cell maturation in the normal and little mutant mouse. Acta Anat (Basel) 131, 150-5.

Wojcik, J., and Postel-Vinay, M. C. (1999). Signal transduction of the growth hormone (GH) receptor, and GH-binding protein. Growth Horm IGF Res 9 Suppl A, 51-5.

Wu, D., Chen, C., Zhang, J., Bowers, C. Y., and Clarke, I. J. (1996). The effects of GH- releasing peptide-6 (GHRP-6) and GHRP-2 on intracellular adenosine 3',5'- monophosphate (cAMP) levels and GH secretion in ovine and rat somatotrophs. J Endocrinol 148, 197-205.

218 References

Wu, D., Chen, C., Zhang, J., Katoh, K., and Clarke, I. (1994). Effects in vitro of new growth hormone releasing peptide (GHRP-1) on growth hormone secretion from ovine pituitary cells in primary culture. J Neuroendocrinol 6, 185-90.

Wu, D., Clarke, I. J., and Chen, C. (1997). The role of protein kinase C in GH secretion induced by GH-releasing factor and GH-releasing peptides in cultured ovine somatotrophs. J Endocrinol 154, 219-30.

Wyvratt Jr., M. J. (1996). Non-peptidyl growth hormone secretagogues. In: Growth Hormone Secretagogues, B. B. Bercu and R. F. Walker, eds. (New York: Springer- Verlag), pp. 103-117.

Yagi, H., Kaji, H., Sato, M., Okimura, Y., and Chihara, K. (1996). Effect of intravenous or intracerebroventricular injections of His-D-Trp-Ala-Trp-D-Phe-Lys-NH2 on GH release in conscious, freely moving male rats. Neuroendocrinology 63, 198-206.

Yamashita, S., and Melmed, S. (1986). Insulin regulation of rat growth hormone gene transcription. J Clin Invest 78, 1008-14.

Yoshimasa, T., Sibley, D. R., Bouvier, M., Lefkowitz, R. J., and Caron, M. G. (1987). Cross-talk between cellular signalling pathways suggested by phorbol-ester-induced adenylate cyclase phosphorylation. Nature 327, 67-70.

Young, W. S., 3rd, lacangelo. A., Luo, X. Z., King, C., Duncan, K., and Ginns, E. I. (1999). Transgenic expression of green fluorescent protein in mouse oxytocin neurones. J Neuroendocrinol 11, 935-9.

Young, W. S. d., Bonner, T. I., and Brann, M. R. (1986). Mesencephalic dopamine neurons regulate the expression of neuropeptide mRNAs in the rat forebrain. Proc Natl Acad Sci U S A 83, 9827-31.

Yu, L. Y., Tushinski, R. J., and Bancroft, F. C. (1977). Glucocorticoid induction of growth hormone synthesis in a strain of rat pituitary cells. J Biol Chem 252, 3870-5.

Zeitler, P., Argente, J., Chowen-Breed, J. A., Clifton, D. K., and Steiner, R. A. (1990). Growth hormone-releasing hormone messenger ribonucleic acid in the hypothalamus of the adult male rat is increased by testosterone. Endocrinology 127, 1362-8.

Zhang, Y., Buchholz, F., Muyrers, J. P., and Stewart, A. F. (1998). A new logic for DNA engineering using recombination in Escherichia coli. Nat Genet 20, 123-8.

Zheng, H., Bailey, A., Jiang, M. H., Honda, K., Chen, H. Y., Trumbauer, M. E., Van der Ploeg, L. H., Schaeffer, J. M., Leng, G., and Smith, R. G. (1997). Somatostatin receptor subtype 2 knockout mice are refractory to growth hormone-negative feedback on arcuate neurons. Mol Endocrinol 11, 1709-17.

219 References

Zhou, Y., Xu, B. C., Maheshwari, H. G., He, L., Reed, M., Lozykowski, M., Okada, S., Cataldo, L., Coschigamo, K., Wagner, T. E., Baumann, G., and Kopchick, J. J. (1997). A mammalian model for Laron syndrome produced by targeted disruption of the mouse growth hormone receptor/binding protein gene (the Laron mouse). Proc Natl Acad Sci U SA 94, 13215-20.

220