The Biology of GPR50, the -Related

A thesis submitted to the University of Manchester for the degree of Doctor of Philosophy in the Faculty of Life Sciences

2011

Laura Hand

1

Table of Contents LIST OF FIGURES ...... 7 LIST OF TABLES...... 10 ABSTRACT ...... 11 DECLARATION ...... 12 COPYRIGHT STATEMENT ...... 12 ABBREVIATIONS ...... 13 ACKNOWLEDGEMENTS ...... 20 CHAPTER 1 ...... 21 INTRODUCTION ...... 21 1.1 GPR50, the Melatonin Related Receptor ...... 22 1.2 Melatonin Signalling ...... 22 1.2.1 Melatonin physiology and function ...... 22 1.2.2 Melatonin synthesis ...... 23 1.2.3 subtypes ...... 25 1.2.4 G -coupled receptors ...... 28 1.2.5 Melatonin receptor signal transduction ...... 30 1.3 Localisation of GPR50 ...... 31 1.4 GPR50 expression in the ...... 33 1.4.1 Hypothalamic control of energy balance ...... 33 1.5 GPR50 expression in the DMH ...... 37 1.6 Potential outputs regulated by GPR50: The HPT and HPA axes ...... 38 1.6.1 The HPT axis ...... 40 1.6.2 The HPA axis ...... 42 1.7 GPR50 expression in the ependymal layer of cells lining the third ventricle ...... 44 1.8 Variations in GPR50 expression ...... 48 1.9 Torpor ...... 49 1.9.1 Definition of torpor...... 49 1.9.2 Timing of daily torpor ...... 50 1.9.3 Mechanisms regulating torpor induction ...... 52 1.10 Conclusion ...... 54 1.11 Overall aims ...... 54 CHAPTER 2 ...... 55 MATERIALS AND METHODS ...... 55 2.1 Expression Constructs...... 56 2.2 Cloning ...... 56

2

2.2.1 Hemagglutin (HA) epitope tagging ...... 56 2.2.1 Polymerase chain reaction (PCR) ...... 57 2.2.2 PCR amplicon gel excision and purification ...... 58 2.2.3 Plasmid transformation ...... 58 2.2.4 Plasmid amplification and purification ...... 59 2.2.5 Sequencing and plasmid purification ...... 59 2.3 Cell Culture ...... 60 2.3.1 Maintenance of cell lines ...... 60 2.3.2 Transient transfections ...... 61 2.3.3 Stable transfections ...... 61 2.3.4 Differentiation of 3T3-L1 cells ...... 62 2.4 Protein Analysis ...... 62 2.4.1 Protein quantification by bicinchoninic acid (BCA) assay ...... 62 2.4.2 Co-Immunoprecipitation ...... 62 2.4.3 SDS-PAGE ...... 64 2.4.4 Immunoblotting (Western blotting) ...... 64 2.5 Immunocytochemistry and Immunofluorescence ...... 65 2.6 Flow Activated Cell Sorting (FACS) ...... 66 2.7 Cyclic Adenosine Monophosphate (cAMP) Accumulation Assay ...... 67 2.8 Oil-Red-O Staining of Differentiated 3T3-L1 Cells ...... 67 2.9 Intracellular Triglyceride Content and Lipolysis of Differentiated 3T3-L1 Cells ...... 68 2.10 RNA Extraction ...... 68 2.10.1 Tissue ...... 68 2.10.2 Cells ...... 69 2.11 Reverse Transcription (RNA to cDNA) ...... 69 2.12 RT-PCR of Gpr50 and Txnip in mouse tissues ...... 69 2.13 Quantitative PCR (qPCR) ...... 70 2.14 Animal maintenance ...... 72 2.14.1 Indirect calorimetry...... 73 2.14.2 Cold Chamber ...... 73 2.14.3 2-deoxyglucose (2-DG) administration ...... 74 2.14.4 Tissue collection ...... 74 2.15 Microarray ...... 76 2.16 In Situ Hybridisation ...... 76 2.16.1 Cryosectioning and preparation of tissue sections ...... 76 2.16.2 Preparation of radiolabelled probes ...... 77

3

2.16.3 Synthesis and purification of 33Pα-UTP labelled probe ...... 79 2.16.4 Hybridisation ...... 80 2.16.5 Analysis of in situ hybridisation ...... 80 2.17 Immunhistochemistry (IHC) ...... 80 2.17.1 Tissue fixing and processing ...... 80 2.17.2 Immunohistochemical staining of tissue ...... 81 2.18 Protein Extraction from Tissue ...... 82 2.19 Statistical Analysis ...... 82 CHAPTER 3 ...... 83 GPR50 SIGNALLING ...... 83 3.1 INTRODUCTION ...... 84 3.2 RESULTS ...... 85 3.2.1 Identification of a novel splice variant of GPR50 ...... 85 3.2.2 Expression of whole-length GPR50 and tGPR50 in mammalian cells ...... 85 3.2.3 Differential localisation of whole-length GPR50 and tGPR50 in mammalian cells ...... 85

3.2.4 Co-immunoprecipitation of GPR50 with Gαi1 ...... 91 3.2.5 GPR50 decreases forskolin-stimulated cAMP accumulation ...... 91 3.2.6 GPR50 decreases CREB serine133 phosphorylation ...... 94 3.3 DISCUSSION ...... 94 CHAPTER 4 ...... 100 GPR50 IN WHITE ADIPOSE TISSUE ...... 100 4.1 INTRODUCTION ...... 101 4.2 RESULTS ...... 102 4.2.1 Identification of GPR50 expression in white adipose tissue ...... 102 4.2.2 Gpr50 expression during adipogenesis...... 104 4.2.3 Triglyceride accumulation in 3T3-L1s with targeted knock-down or overexpression of Gpr50...... 104 4.2.4 Adipogenic expression in 3T3-L1s with targeted knock-down or overexpression of Gpr50...... 107 4.2.5 Lipolysis in 3T3-L1s with targeted knock-down or overexpression of Gpr50 ...... 110 4.3 DISCUSSION ...... 110 CHAPTER 5 ...... 118 HYPOTHALAMIC GENE CHANGES DURING TORPOR ...... 118 5.1 INTRODUCTION ...... 119 5.2 RESULTS ...... 120

4

5.2.1 Acute hypometabolism in Gpr50-/- mice upon fasting and selection of animals for analysis of in the hypothalamus ...... 120 5.2.2 Microarray results ...... 121 5.2.3 Gpr50 expression in the hypothalamus of Gpr50-/- mice ...... 125 5.2.4 Hprt1 expression in the hypothalamus of Gpr50-/- mice ...... 127 5.2.5 Validation of the microarray results by qPCR in the DeltaGen mice ...... 130 5.2.6 Validation of the microarray results by qPCR in the Organon mice 130 5.2.5 Modulation of period in torpor...... 130 5.2.6 Hypothalamic Fabp7 expression during torpor ...... 133 5.3 DISCUSSION ...... 137 CHAPTER 6 ...... 147 TXNIP: A NOVEL TORPOR-INDUCED GENE ...... 147 6.1 INTRODUCTION ...... 148 6.2 RESULTS ...... 150 6.2.1 Txnip expression in the hypothalamus ...... 150 6.2.2 Txnip expression in the hypothalamus in response to fasting and torpor ...... 150 6.2.3 Txnip expression in peripheral tissues in response to fasting and torpor ...... 153 6.2.4 Gpr50 expression in peripheral tissues in response to fasting ...... 153 6.2.5 Gpr50 and Txnip expression following acute cold exposure ...... 157 6.2.6 Gpr50 and Txnip expression following glucoprivation ...... 159 6.2.7 Txnip expression is increased by fasting-induced torpor ...... 159 6.2.8 Txnip expression in the hypothalamus of the Siberian hamster ...... 162 6.3 DISCUSSION ...... 166 CHAPTER 7 ...... 173 GENERAL DISCUSSION ...... 173 7.1 GPR50 SIGNALLING ...... 174 7.2 GPR50 IN THE HYPOTHALAMUS ...... 175 7.3 GPR50 IN WHITE ADIPOSE TISSUE ...... 177 7.4 GENE CHANGES DURING TORPOR ...... 179 7.4.1 Txnip, a novel torpor-induced gene...... 180 7.4.2 Gpr50 in brown adipose tissue ...... 185 7.5 THESIS OVERVIEW ...... 187 REFERENCES ...... 188 APPENDIX A ...... 227 Publications ...... 227

5

APPENDIX B ...... 236 Significantly altered genes in Gpr50-/- versus WT mice when fasted...... 236 APPENDIX C ...... 243 Significantly altered genes in Gpr50-/- versus WT mice when fed ad libitum...... 243 APPENDIX D ...... 244 Significantly altered genes in WT mice when fasted versus fed ad libitum. . 244 APPENDIX E ...... 250 Significantly altered genes in Gpr50-/- mice when fasted versus fed ad libitum...... 250 APPENDIX F ...... 279 Thermogenesis in Gpr50-/- and WT mice fasted at 15 °C...... 279 APPENDIX G ...... 280 Functional clustering of genes with mRNA differences between fasted WT and Gpr50-/- mice...... 280 APPENDIX G ...... 281 Torpor expression in second line of Gpr50-/- mice...... 281

Final word count: 75, 409

6

LIST OF FIGURES

Chapter 1 Page

Figure 1.1 Melatonin synthesis in the mammalian 24

Figure 1.2 Schematic model of GPCR and receptor mediated 26 activation

Figure 1.3 Schematic representation of the mammalian hypothalamus 34

Figure 1.4 Schematic representation of the hypothalamic-pituitary-thyroid 39 (HPT) and hypothalamic-pituitary-adrenal (HPA) axes

Figure 1.5 Schematic summarising the cell types of the ependymal layer 46 lining the third ventricle

Chapter 2

Figure 2.1 Schematic of 3T3-L1 differentiation 63

Figure 2.2 Collection of hypothalamic blocks for quantitative PCR and 75 protein analysis

Chapter 3

Figure 3.1 Identification and cloning of full length and truncated GPR50 86

Figure 3.2 Immunoblotting of GPR50 87

Figure 3.3 Subcellular localisation of GPR50 in mammalian cells 89

Figure 3.4 Plasma membrane insertion of GPR50 90

Figure 3.5 wGPR50 co-immunoprecipitates with Gαi1 92

Figure 3.6 wGPR50 decreases forskolin-stimulated cAMP accumulation 93

Chapter 4

Figure 4.1 Gpr50 expression in white adipose tissue 103

7

Figure 4.2 Gpr50 expression in 3T3L1 cells increases during 105 differentiation

Figure 4.3 Gpr50 plays an essential role in triglyceride accumulation 106

Figure 4.4 Influence of GPR50 on expression of lipogenic and lipolytic 108 genes

Figure 4.5 Influence of GPR50 on triglyceride accumulation and lipolysis 109

Figure 4.6 Diagrammatic representation of triglyceride turnover in 112 adipocytes

Chapter 5

Figure 5.1 Metabolic rates of individual mice used in the microarray study 122

Figure 5.2 Transcript changes in ad libitum fed and fasted WT and Gpr50-/- 123 mice

Figure 5.3 Functional clustering of genes with mRNA differences between 124 fasted WT and Gpr50-/- mice

Figure 5.4 Exon expression of GPR50 in ad libitum fed WT and Gpr50-/- 126 mice

Figure 5.5 Hprt1 expression in two lines of Gpr50-/- mice 128

Figure 5.6 Body temperature recordings in fasted WT and Hprt1-/- mice 129

Figure 5.7 Validation of microarray results 131

Figure 5.8 Per1 expression in the hypothalamus 134

Figure 5.9 Per2 expression in the hypothalamus 135

Figure 5.10 Fabp7 expression in the hypothalamus 136

Figure 5.11 Chromosomal locations of Gpr50, Hprt1 and Gabrq 139

8

Chapter 6

Figure 6.1 Txnip expression in the hypothalamus 151

Figure 6.2 Txnip expression in the hypothalamus 152

Figure 6.3 Peripheral Txnip expression in fed and fasted WT and Gpr50-/- 154 mice

Figure 6.4 Txnip expression in the hypothalamus and BAT 155

Figure 6.5 Gpr50 expression in fasted WT mice 156

Figure 6.6 Gpr50 and Txnip expression following acute cold exposure 158

Figure 6.7 Gpr50 and Txnip expression following 2-deoxyglucose 160 treatment

Figure 6.8 Txnip expression in female Gpr50-/- mice 161

Figure 6.9 Txnip expression in the hypothalamus during torpor 163

Figure 6.10 Peripheral Txnip expression in WT mice during torpor 164

Figure 6.11 Txnip expression in the hypothalamus of Siberian hamsters 165

Figure 6.12 Model linking Txnip expression, glycolysis and oxidative 168 phosphorylation

Appendix F Thermogenesis in Gpr50-/- and WT mice fasted at 15 °C 279

9

LIST OF TABLES

Chapter 1

Table 1.1 Classification of Gα subtypes and their effectors 29

Chapter 2

Table 2.1 Primers used for RT-PCR epitope tagging 57 Table 2.2 PCR mastermix 58 Table 2.3 Primers used for RT-PCR murine tissue profiling 70 Table 2.4 Primers used for qPCR 71 Table 2.5 qPCR mastermix 72 Table 2.6 PCR primers used for RT-PCR to generate in situ hybridisation 77 riboprobe templates Table 2.7 Mastermix for plasmid linearization 78 Table 2.8 Mastermix for riboprobe in vitro transcription 79

Chapter 7

Table 7.1 Hypothalamic transcripts with known/predicted changes in 181 expression with altered energy status/torpor

Appendix B Significantly altered genes in Gpr50-/- versus WT mice when 236 fasted

Appendix C Significantly altered genes in Gpr50-/- versus WT mice when fed 243 ad libitum

Appendix D Significantly altered genes in WT mice when fasted versus fed 244 ad libitum

Appendix E Significantly altered genes in Gpr50-/- mice when fasted versus 250 fed ad libitum

Appendix G Functional clustering of genes with mRNA differences between 280 fasted WT and Gpr50-/- mice

10

ABSTRACT The University of Manchester Submitted by Laura Hand for the degree of Doctor of Philosophy, and entitled: The Biology of GPR50, the Melatonin-Related Receptor. September 2011.

GPR50 is the mammalian ortholog of the avian/ melatonin receptor, Mel1c, yet its ligand and physiological function remain unknown. Previous studies implicate the receptor in the control of energy homeostasis; Gpr50-/- mice demonstrate elevated basal metabolic rate, reduced fat stores, and partial resistance to diet-induced obesity. This thesis examines the physiological impact of GPR50 on a cellular level as well as in two key structures involved in metabolism and energy balance; namely white adipose and the hypothalamus.

In vitro studies demonstrate GPR50’s ability to activate downstream signalling pathways, specifically, inhibition of forskolin-stimulated cAMP accumulation in HEK293 cells. Co-immunoprecipitation experiments reveal association of GPR50 with Gαi1, suggesting the receptor acts through Gi/o to elicit inhibition of cAMP production. This work further identifies a novel truncated variant of GPR50 protein, lacking transmembrane domains 6 and 7 and the cytoplasmic tail. This variant demonstrates altered subcellular localisation from wild-type protein, being retained in the endoplasmic reticulum instead of inserting into the plasma membrane, and does not possess the same signalling capability as the full-length receptor.

This work reports the expression of GPR50 in murine white adipose tissue and explores the potential significance of the receptor in this organ. Levels of GPR50 mRNA increase during differentiation of the adipocytic 3T3-L1 cell line, and stable knockdown or overexpression of GPR50 results in decreased and increased triglyceride accumulation, respectively. Knockdown of GPR50 represses transcript levels of hormone sensitive lipase, adipose triglyceride lipase, and lipoprotein lipase, whereas overexpression of GPR50 increases levels of these adipocytic enzymes, as well increasing expression of peroxisome proliferator-activated receptor gamma. Augmentation of GPR50 expression in either direction blunts forskolin-stimulated lipolysis in differentiated 3T3-L1 cells. These data implicate GPR50 in adipocyte metabolism.

Gpr50-/- mice readily enter torpor in response to fasting, and gene expression (Affymetrix) profiling of Gpr50-/- mice reveals strong induction of thioredoxin interacting protein (Txnip) in the hypothalamus during torpor, an original finding. Recent data suggests a role for Txnip as a molecular nutrient sensor important in the regulation of energy metabolism. Here, Txnip expression is shown to be strongly expressed within the ependyma lining the third ventricle, a known site of GPR50 expression. Hypothalamic Txnip expression is induced by fasting, but demonstrates exaggerated levels of expression in Gpr50-/- and also in WT mice driven into torpor. Strikingly, Txnip expression is also elevated in the hypothalamus of a model of natural seasonal torpor, the Siberian hamster. Thus, Txnip appears to have a critical role during the torpor response, potentially regulating energy expenditure and fuel utilisation during this extreme hypometabolic state.

11

DECLARATION

No portion of the work referred to in this thesis has been submitted in support of an application for another degree or qualification of this or any other university or other institute of learning.

COPYRIGHT STATEMENT i. The author of this thesis (including any appendices and/or schedules to this thesis) owns certain copyright or related rights in it (the “Copyright”) and he has given The University of Manchester certain rights to use such Copyright, including for administrative purposes.

ii. Copies of this thesis, either in full or in extracts and whether in hard or electronic copy, may be made only in accordance with the Copyright, Designs and Patents Act 1988 (as amended) and regulations issued under it or, where appropriate, in accordance with licensing agreements which the University has from time to time. This page must form part of any such copies made.

iii. The ownership of certain Copyright, patents, designs, trade marks and other intellectual property (the “Intellectual Property”) and any reproductions of copyright works in the thesis, for example graphs and tables (“Reproductions”), which may be described in this thesis, may not be owned by the author and may be owned by third parties. Such Intellectual Property and Reproductions cannot and must not be made available for use without the prior written permission of the owner(s) of the relevant Intellectual Property and/or Reproductions.

iv. Further information on the conditions under which disclosure, publication and commercialisation of this thesis, the Copyright and any Intellectual Property and/or Reproductions described in it may take place is available in the University IP Policy (see http://www.campus.manchester.ac.uk/medialibrary/policies/intellectual- property.pdf), in any relevant Thesis restriction declarations deposited in the University Library, The University Library’s regulations (see http://www.manchester.ac.uk/library/aboutus/regulations) and in The University’s policy on presentation of Theses.

12

ABBREVIATIONS

delta/change

-/- knockout

~ approximately

°C degrees Celsius

2-DG 2-deoxyglucose

5HT serotonin

AANAT arylalkylamine-N-acetyltransferase

AC

Acetyl-CoA acetyl coenzyme A

ACTH adrenocorticotropic hormone

AdipoR adiponectin receptor

AgRP agouti-related peptide

Ala alanine

AMP adenosine monophosphate

AMPK AMP-activated protein kinase

ANOVA analysis of variance

AR adrenergic receptors

ARC arcuate nucleus

ATGL adipose triglyceride lipase

ATP adenosine triphosphate

BAT brown adipose tissue

Bmal brain and muscle aryl hydrocarbon receptor nuclear translocator-like

BPAD bipolar affective disorder

13

BSA bovine serum albumin

CAM constitutive activity-inducing mutation cAMP cyclic adenosine monophosphate

CART cocaine-and amphetamine-regulated transcript

ChoRE carbohydrate response element

Co-IP co-immunoprecipitation

CRE cAMP response element

CREB cAMP response element binding protein

CRH corticotropin-releasing hormone

CSF cerebrospinal fluid

D aspartate d day

DAPI 4’,6-diamidino-2-phenylindole

Dec deleted in esophageal cancer

Dio deiodinase

DMEM Dulbecco’s modified Eagle medium

DMH dorsomedial nucleus

DNA deoxyribonucleic acid

E glutamate e(1-3) extracellular loop(1-3)

EDTA ethylyenediaminetetraacetic acid

EGFP enhanced green fluorescent protein

ER endoplasmic reticulum

FA fatty acid

14

Fabp fatty acid binding protein

FACS flow activated cell sorting

FAS fatty acid synthase

FBS fetal bovine serum

FCS fetal calf serum

FFA free fatty acids

FITC fluorescein isothyocyanate

G6P glucose-6-phosphate

Gabr gamma-aminobutyric acid (GABA) receptor

GADP glyceraldehydes-3-phosphate

GC glucocorticoid

GDP guanosine diphosphate

GLUT glucose transporter

GPCR G protein coupled receptor

GPR50 G protein coupled receptor 50

GTP guanosine triphosphate h hour

HDL high density lipoprotein

HEK293 embryonic kidney cells 293

HIOMT hydroxyindole-O-methyltransferase

HPR horseradish peroxidise

Hprt hypoxanthine-guanine phosphoribosyltransferase

HPT hypothalamic-pituitary-thyroid

HSL hormone sensitive lipase

15 i(1-4) intracellular loop(1-4)

ICC immunocytochemistry

ICV intracerbroventricular

IHC immunohistochemistry

Ile isoleucine

JAK janus kinase

LB Luria-Bertani broth

LDL low density lipoprotein

LH lateral hypothalamic nucleus

LPL lipoprotein lipase

MAPK mitogen-activated protein kinase

MBH mediobasal hypothalamus

MDD major depressive disorder

Mel melatonin

Mel1c min minute mRNA messenger ribonucleic acid

MT1

MT2

N asparagines

NAS N-acetylserotonin

NE norepinephrine

NEFA non-esterified fatty acid

NP-40 nonident P-40

16

NPY neuropeptide Y

Osbpl oxysterol-binding protein like

P phosphorylated

PAGE poly acrylamide gel electrophoresis

PBS phosphate buffered saline

PBST phosphate buffered saline with Tween-20

PBSX phosphate buffered saline with Triton-x-100

PCR polymerase chain reaction

PDE3B phosphodiesterase 3B

PEPCK phosphoenol pyruvate carboxykinase

Per period

PFA paraformaldehyde

PGC-1α PPARγ coactivator-1α

Phospho phosphorylated

PKA protein kinase A

PKC protein kinase C

PLCβ phospholipase Cβ

PPAR peroxisome proliferator-activated receptor

PT pars tuberalis

PUFA polyunsaturated fatty acids

PVN paraventricular nucleus

QCS quality control system qPCR quantitative polymerase chain reaction

R arginine

17

RNA ribonucleic acid

ROS reactive oxygen species

RXR retinoid X receptor

SAD seasonal affective disorder

SCN

SCZ

SDS sodium dodecyl sulphate sec second

SFA saturated fatty acids

SGLT sodium/glucose transporter sh short hairpin

SNS sympathetic nervous system

SP short photoperiod

SREBP sterol regulatory element binding protein

STAT signal transducers and activators of transcription

T3 3,5,3’-triiodothyronine

T4 thyroxine

Tb core body temperature

TCL total cell lysates tGPR50 truncated GPR50

Thr threonine

TM transmembrane

TRH thyrotropin-releasing hormone

TSH thyroid stimulating hormone

18

Txnip thioredoxin-interacting protein

UCP uncoupling protein

UFA unsaturated fatty acids

Val valine

VLDL very low density lipoprotein

VMN ventromedial nucleus

VO2 oxygen consumption

W tryptophan

WAT white adipose tissue

WB western blot wGPR50 whole-length GPR50

WT wild-type

Y tyrosine

α alpha

α-MSH α-melanocyte-stimulating hormone

β beta

γ gamma

θ theta

19

ACKNOWLEDGEMENTS

I would like to thank my supervisors, Professor Andrew Loudon and Professor Simon Luckman, for the opportunity to undertake this project and for encouragement during the course of my PhD. I would also like to acknowledge the financial support provided by the Biotechnology and Biological Sciences Research Council and Astra Zeneca for funding my studies.

I must thank all the members of the Loudon lab, especially Jian Li for teaching me the molecular biology techniques used in this thesis, and Sandrine Dupré and Ben Saer for assistance with the in situ hybridisations.

Finally, thanks to David Bechtold for his immeasurable help during my PhD; not only with the in vivo work, but for his continued interest in the project and invaluable experimental guidance and advice.

20

CHAPTER 1

INTRODUCTION

21

1.1 GPR50, the Melatonin Related Receptor

GPR50 is a G protein coupled receptor (GPCR), originally cloned from a human pituitary cDNA library on the basis of a moderate degree of homology to the melatonin receptors MT1 and MT2 (Reppert et al., 1996). Indeed, GPR50 is approximately 45% identical at the amino acid level to these receptors, with identity increasing to 55% when comparing just the transmembrane domains. However, despite this close structural relationship, GPR50 does not bind melatonin (Drew et al., 1998), and remains an .

1.2 Melatonin Signalling

1.2.1 Melatonin physiology and function

Melatonin is synthesised in a number of different organs and cells, including the , gastrointestinal tract, bone marrow, Harderian gland, leukocytes and the skin (Hardeland et al., 2011). However, from these sites of formation, melatonin is either poorly released or only in response to specific stimuli. In , the primary site of melatonin production and secretion is the , where it is synthesised during the hours of darkness only (Reiter, 1991). As a highly lipophilic molecule, rather than being stored in this secretory organ, melatonin is released immediately into the blood upon synthesis (Dubocovich et al., 2010), and in conjunction with its rapid metabolisation, primarily by the liver (Pandi-Perumal et al., 2008), dynamic daily changes in melatonin synthesis are translated into similar changes in circulating levels of melatonin. Thus, melatonin serves as a hormonal message faithfully reflecting the photoperiod. In this way, melatonin not only influences daily, circadian rhythms, but also mediates the regulation of seasonal rhythms, whereby many organisms exhibit circannual cycles in a host of physiological and behavioural responses. In moderate zones, changes in ambient temperature have profound effects on living conditions, and in order to adapt to these changes, seasonal rhythms in reproductive status, feeding behaviour, fur colour and quality, and readiness to hibernation increases the viability of organisms during unfavourable seasons of the year (Vanecek, 1998). Seasonal variations in day length are encoded into variations in the duration of nocturnal melatonin secretion, which are interpreted by tissues of the neuroendocrine system, such as the pars tuberalis (PT) of the anterior lobe of the , to trigger seasonal responses (Malpaux et al., 2001).

22

Diverse effects of melatonin have been reported, including a role for the hormone in sleep initiation (Kennaway and Wright, 2002), inhibition of release in the hypothalamus and retina (Zisapel, 2001), involvement in the ageing process (Karasek, 2004) and pubertal development (Salti et al., 2000), blood pressure control (Grossman et al., 2006; Scheer et al., 2004), glucose regulation (Peschke and Muhlbauer, 2010), immune response regulation (Carrillo-Vico et al., 2006), and functions (Luchetti et al., 2010). An important conceptual difficulty in melatonin research is that it is a signal of darkness, but has different functional consequences depending on the species’ time of peak activity. In nocturnal species, melatonin is associated with arousal and physical activity, whereas in diurnal species, it is associated with sleep and rest (Pandi-Perumal et al., 2008). Therefore, differential interpretation of the melatonin signal in different species must be relayed by brain regions and peripheral tissues that mediate melatonin signalling.

1.2.2 Melatonin synthesis

Daily rhythms of pineal melatonin synthesis are generated by an endogenous in the suprachiasmatic nucleus (SCN) of the hypothalamus, which is entrained to the light:dark cycle via retino-hypothalmic input from the eyes (Pevet et al., 2002). The SCN is described as the master biological clock, generating rhythmic output of neuronal activity and hormone production, which govern many aspects of physiology and behaviour. For example, the sleep-wake cycle, locomotor activity, food intake and energy homeostasis are all regulated in a circadian manner, and the generation of such rhythms allows entrainment of organisms to the light:dark cycle, appropriately phasing their biology to the time of day (Reppert and Weaver, 2002). Like other processes regulated by the SCN, under conditions of constant darkness, the circadian synthesis of melatonin ‘free-runs’ with a natural endogenous rhythm that is close to 24 hours. The SCN itself expresses a high density of melatonin receptors, and thus melatonin engages in a feedback loop to directly entrain the SCN (Pevet et al., 2006).

Rhythmic SCN activity results in activation of a multi-synaptic pathway (Figure 1.1) that terminates on the pineal gland in an extensive network of norepinephrine (NE)- containing fibres (Ganguly et al., 2002). NE released by these processes, signals through α- and β- adrenergic receptors, expressed postsynaptically in the pineal gland,

23

Figure 1.1 Melatonin synthesis in the mammalian brain. A. Schematic representation of the main neural system regulating the pineal gland. Light detected by the retina generates a signal that is transmitted by the retino- hypothalamic tract to the SCN. Monosynaptic connections between the SCN, PVN, intermediolateral cell column (IML) and superior cervical ganglia (SGL) course to the pineal gland. B. Activation of the neural systems in A results in release of norepinephrine (NE), which acts through α- and β-adrenergic receptors (AR). αAR mediates G protein activation of phospholipase C (PLC) which cleaves membrane 2+ phospholipids to yield inositol triphosphate (IP3) which stimulates the release of Ca from intracellular stores, and diacylglycerol (DAG), which together with Ca2+, activates protein kinase C (PKC), which potentiates the actions of the βAR. βAR mediates G protein activation of adenylyl cyclase (AC) to increase cAMP production, which activates protein kinase A (PKA). PKA phosphorylates (P) and activates cAMP response element (CRE) binding protein (CREB), and drives transcription of AANAT. PKA further phosphorylates AANAT, allowing it to bind 14-3-3 proteins, protecting it from proteasomal degradation. AANAT activity is thus increased, accelerating the conversion of serotonin (5HT) to N-acetylseratonin (NAS), which is subsequently converted to melatonin (MEL). Schematic by the author.

24 to activate signalling pathways, which regulate the penultimate enzyme in melatonin biosynthesis, arylalkylamine-N-acetyltransferase (AANAT) (Figure 1.1).

Melatonin is synthesised from the amino acid tryptophan, which is first converted to serotonin through hydroxylation and decarboxylation. Serotonin is then metabolised into melatonin by the sequential action of AANAT and hydroxyindole-O- methyltransferase (HIOMT) (Schomerus and Korf, 2005). A large nocturnal increase in AANAT activity within the pineal gland as a result of SCN-controlled NE-signalling is responsible for the daily rhythm of melatonin synthesis and secretion (Klein et al., 1997). The amplitude of this activity increase varies among species, along with the relative importance of mechanisms involved in its control, which may be transcriptional or post-translational (Figure 1.1). HIOMT is also subject to regulation in certain seasonal species, which show increased transcript expression when placed on short photoperiod (Simonneaux and Ribelayga, 2003). This increase in activity parallels increased amplitude of the nocturnal melatonin peak observed in these animals, suggesting that HIOMT may be involved in seasonal modulation of melatonin production, however mechanisms underlying this variation are poorly understood.

Following cessation of neural stimulation, AANAT is rapidly destroyed, ensuring prompt termination of melatonin synthesis (Ganguly et al., 2002). Light exposure suppresses neural output to the pineal gland, and in combination with the short half-life of melatonin, leads to a rapid decrease in circulating levels of the hormone. By this mechanism, differences in the duration of the night period can be reliably converted into differences in the duration of melatonin production. As a result, organisms can discern subtle variations in day length, which would not be possible if the melatonin signal was switched off gradually.

1.2.3 Melatonin receptor subtypes

Upon release from the pineal gland, melatonin enters the systemic circulation to all peripheral and central structures where it mediates most of its actions via specific membrane receptors of the GPCR superfamily, which are characterised by a common structure, composed of seven membrane-spanning helices with an extracellular N- terminus and a cytoplasmic tail (Bockaert and Pin, 1999) (see Figure 1.2). Three

25

Figure 1.2 Schematic model of GPCR and receptor mediated G protein activation. A. GPCRs have a common structure composed of seven transmembrane helices (I-VII) connected by three intracellular (i) and three extracellular (e) loops. B. Activation of a GPCR by ligand-binding causes a conformational change in the receptor which allows it to activate interacting heterotrimeric G proteins, anchored to the plasma membrane by lipid modification of the α and γ subunits. The active GPCR stimulates the release of GDP from the α subunit and subsequent binding of GTP. This promotes the dissociation of the G protein into α and βγ subunits, which are then free to interact with effectors. Schematic by the author.

26 melatonin receptor subtypes have been characterised: MT1 and MT2, which are expressed in mammals, and the non-mammalian Mel1c, which is found in , fish and (Dubocovich et al., 2010). Melatonin can also bind transcription factors belonging to the retinoic acid receptor superfamily, namely several splice variants of RORα, although their affinity to melatonin is lower, compared to MT1 and MT2 (Hardeland et al., 2011). Melatonin binding sites have also been described for the ubiquitously expressed proteins calmodulin and calreticulin, which are involved in calcium metabolism, and quinone reductase 2 (Hardeland et al., 2011). The relevance of melatonin interactions with these binding sites is still uncertain, but, if found to be functionally important, would imply another host of pleiotropic actions of this hormone.

A recent study of the phylogenetic evolution of the melatonin receptors, supported by tracking the synteny of genes surrounding Mel1c and GPR50 on their respective , suggests that GPR50 is actually the mammalian ortholog of Mel1c (Dufourny et al., 2008). However, the receptor has been greatly remodelled through evolution by mutations of numerous amino acids and the addition of a remarkably long C-terminal tail. These alterations have rendered GPR50 unable to bind melatonin, despite its retention of a conserved histidine residue in transmembrane domain (TMD) 5, demonstrated as a key amino acid in the binding of melatonin (Conway et al., 1997). Studies using chimeras of MT1 and GPR50, devised to investigate structure-function relationships of the melatonin receptor (Conway et al., 2000; Gubitz and Reppert, 2000), give clues about GPR50’s inability to bind melatonin. In essence, regions of GPR50 were substituted into the corresponding positions of MT1 and affinity for melatonin was assessed. Replacement of TM4 and external loop 2 (e2) and internal loop 1 (i1) resulted in complete loss of ligand binding (Conway et al., 2000), however substitution of TMD4 alone had no effect (Gubitz and Reppert, 2000). This suggests that sequences in e2 or i2 of MT1 affect the conformation of the melatonin binding pocket, perhaps in the case of i2 by facilitating interactions with intracellular components, and that GPR50 lacks these particular domains. Replacement of TMD6 also abolished ligand binding, and non-conserved residues between GPR50 and MT1 were targeted by site-directed mutagenesis in order to further characterise melatonin- interaction sites within this domain (Conway et al., 2000; Gubitz and Reppert, 2000). Replacement of glycine in position 258 of the human MT1 receptor with threonine, the corresponding residue in GPR50, vastly reduced affinity for the radioligand [125I]iodomelatonin. As this amino acid is predicted to reside near the upper extracellular surface of TMD6 (Barrett et al., 2003) with only a very short side chain 27 projecting into the binding pocket, prevention of melatonin entry to GPR50’s binding site by steric hindrance may be the primary reason for the inability of the receptor to bind this ligand.

1.2.4 G protein-coupled receptors

GPCRs constitute a large family of proteins, responsible for transducing a diverse range of extracellular stimuli into intracellular signals. They are generally classified into six main families based on their sequences and ligand-binding properties: Class A (-like), Class B (secretin-like), Class C (metabotropic), Class D (Fungal ), Class E (cyclic AMP), and Class F (/) (Gangal and Kumar, 2007). The melatonin receptors and GPR50 form a novel subfamily within the largest, Class A family.

Upon binding their ligand, GPCRs undergo a conformational change that enables them to activate heterotrimeric GTP (guanosine triphosphate)-binding proteins (G proteins), which serve to relay the extracellular stimulus into cellular responses (Kroeze et al., 2003). G proteins are composed of α, β, and γ subunits, and are anchored to the plasma membrane via lipid modification of the α and γ subunits (β and γ form a functional unit, only dissociable by denaturation, and thus lipid modification of β is not required for its association with the plasma membrane) (Casey, 1994). The α subunit is able to bind and hydrolyse GTP, which allows G proteins to act as molecular switches. G proteins are inactive in the GDP (guanosine diphosphate)-bound, heterotrimeric state, but become activated by receptor catalysed guanine nucleotide exchange, resulting in GDP release and GTP binding to the α subunit (Hamm, 1998). α-GTP has a reduced affinity for βγ, causing dissociation of the heterotrimeric G protein. The separated α and/or βγ subunits can then interact with their downstream effectors, until the system is reset when the α subunit hydrolyses GTP to GDP (Johnston and Siderovski, 2007). The hydrolysis of GTP to GDP is regulated by RGS (regulators of G protein signalling) proteins that enhance the intrinsic GTPase activity of the α subunit (De Vries et al., 2000).

A number of genes encoding α, β and γ subunits have been identified, allowing distinct

28

Family Subtype Effector

Gs s + Adenylyl Cyclase + Ca2+ channels – Na+ channels

olf + Adenylyl Cyclase

Gi i1 i2 i3 – Adenylyl Cyclase + K+ channels – Ca2+ channels

oA oB – Adenylyl Cyclase – Ca2+ channels + PLC

z – Adenylyl Cyclase – Ca2+ channels + K+ channels

t1 t2 + cGMP-PDE Gq q 11 14 15 16 + PLC isoforms + + G12 12 13 + Na /H exchanger-1 + RhoGEFs Table 1.1 Classification of Gα subtypes and their effectors. PDE: phosphodiesterase; PLC: phospholipase C; GEF: guanine nucleotide exchange factor. Adapted from (Kostenis et al., 2005).

G protein heterotrimers to form. Four main classes of G proteins can be distinguished:

Gαs, Gαi, Gαq and Gα12 (Kostenis et al., 2005). Each group is further divided into specific isotypes, which function to activate different effector pathways (see Table 1.1).

In general, the Gαs family couples to adenylyl cyclase to stimulate an increase in cAMP

(cyclic adenosine monophosphate), and the Gαi family primarily acts by inhibiting adenylyl cyclase (Jacoby et al., 2006). The primary effector of the Gαq family is phospholipase Cβ (PLCβ) which catalyses the hydrolysis of phosphatidylinositol-4,5- bisphosphate (PIP2) to inositol-1,4,5-triphosphate (IP3) and diacylglycerol (DAG), which both act as secondary messengers to trigger the release of Ca2+ from intracellular stores and activate protein kinase C (PKC) (Jacoby et al., 2006). Lastly, members of the Gα12 family regulate the activation of Rho guanine-nucleotide exchange factors (GEFs) (Jacoby et al., 2006). Gβγ also possess the ability to interact with and activate several effectors, including ion channels, G-protein regulated inward rectifying K+ channels (GIRKs), phosphatidylinositol 3-kinase (PI3K), phospholipases and adenylyl cyclase (Bridges and Lindsley, 2008), and this action is directed by the composition of the βγ dimer (Cabrera-Vera et al., 2003).

29

Once activated, GPCRs are desensitised by two families of proteins; the G protein coupled receptor kinases (GRKs) and the arrestins. GRKs phosphorylate - bound or activated GPCRs, which promotes binding of the inhibitory arrestin proteins (Pitcher et al., 1998). Arrestins block G protein binding by steric inhibition, and therefore prevent receptor signalling (DeWire et al., 2007). In addition to this role in GPCR desensitisation, GRK and arrestin proteins also promote receptor internalisation by endocytosis (Drake et al., 2006). Arrestins add a further level of complexity to GPCR signalling by acting as scaffolds to bring elements of different signalling pathways into close proximity, including members of mitogen activated protein kinase (MAPK) cascades (DeWire et al., 2007), allowing GPCRs to mediate downstream effectors through a G protein-independent mechanism.

1.2.5 Melatonin receptor signal transduction

GPCRs can couple to members of one or more of the G protein subfamilies to selectively modulate multiple signalling cascades. G protein coupling is determined by various intracellular receptor regions, with the second intracellular loop and the beginning and end of the third intracellular loop appearing to be the most important (Moller et al., 2001). The classical coupling of melatonin receptors is to members of the

Gi/o subfamily, whereby cloned MT1, MT2 or Mel1c all inhibit forskolin-stimulated cAMP accumulation, in a pertussis-toxin sensitive manner, when expressed in heterologous cell lines (von Gall et al., 2002). Melatonin-mediated decreases in cAMP have been observed in a number of mammalian tissues, including pituitary, SCN and cerebral arteries (Capsoni et al., 1994; Morgan et al., 1994). Parallel signalling pathways have further been demonstrated for both MT1 and MT2. Activation of MT1 also induces an elevation in cytosolic calcium and activation of PLCβ, either through αi or βγ, due to maintained sensitivity to pertussis toxin (Godson and Reppert, 1997). In addition, MAPK1/2 is stimulated by MT1 receptors and the inward-rectifier potassium channel, Kir3, is activated through a pertussis toxin-sensitive mechanism (Dubocovich et al., 2010). Further investigation of MT2 reveals it can mediate inhibition of soluble gunanylyl cyclase and cGMP formation (Petit et al., 1999), and in the SCN can cause increases in PKC activity, which is required for the phase shifting effects of melatonin on circadian rhythms (Hunt et al., 2001; McArthur et al., 1997). PKC activation could be based on the involvement of other G protein subunits rather than decreases in cAMP, and multiple G protein-dependent activation mechanisms are possible. With regard to other α variants, such as those in the Gq or G12 subfamilies, many potential interactions

30 have been proposed between members of these G protein groups and the melatonin receptors. However, these were typically observed in transfected cells, and thus, although such data identifies a spectrum of possibilities, it does not reveal the physiologically relevant route in a specific cell type in vivo. Therefore, the cellular environment will likely determine which interactions occur, specifically; different cells will vary in their level of melatonin receptor expression and G protein availability, limiting the activation of some pathways.

GPCRs form dimers or higher order oligomers, resulting in distinct functional units with altered receptor pharmacology, signalling, and regulation (Maggio et al., 2005). MT1 and MT2 receptors likewise exist as dimers, with the relative propensity to form MT1 homodimers or MT1/MT2 heterodimers comparable and that of the MT2 homodimer 3- 4 fold lower (Ayoub et al., 2004; Daulat et al., 2007). Using cloned receptor expression in cell lines, it has been suggested that GPR50 also homodimerises and has the ability to form heterodimers with MT1 and MT2 (Levoye et al., 2006). It is of considerable interest that through formation of these heterodimers, GPR50 abolishes high affinity binding of 2-[125I]iodomelatonin to MT1 and antagonises signalling through this receptor. Intriguingly, when the C-terminal of GPR50 was deleted, the mutant receptor no longer elicited these inhibitory effects on MT1. It is possible that the long tail of

GPR50 hinders recruitment of intracellular interacting partners to MT1, such as Gi proteins, which would explain the inhibition of signalling and may also be responsible for the diminished agonist binding, as high-affinity interactions of ligand with MT1 has been shown to rely on the presence of Gi (Barrett et al., 1994). Coincident expression of GPR50 and MT1 mRNAs has been observed in ovine retinal tissue and the PT (Drew et al., 2001), however other regions of GPR50 expression, such as the ependymal cell layer of the third ventricle, do not demonstrate melatonin receptor expression. This suggests that although GPR50 may have a role in melatonin function by altering binding to MT1, further functions for this receptor such as ligand-dependent signalling remain to be elucidated.

1.3 Localisation of GPR50

Melatonin receptors are expressed in various parts of the CNS and associated tissues. MT1 and MT2 are not restricted to the sites of highest density, such as SCN or PT, but their presence has also been demonstrated in the cortex, hippocampus, substantia

31 nigra, retina and choroid plexus (Jockers et al., 2008). MT1 in the human brain is further described in various nuclei of the hypothalamus besides the SCN, including the paraventricular nucleus (PVN), ventromedial nucleus (VMH) and dorsomedial nucleus (DMH) (Wu et al., 2006).

Mel1c demonstrates widespread expression in the chicken brain, with specific localisation in the retina, optic chiasm, optic tract, SCN, thalamus, cerebellum and pineal gland (Reppert et al., 1995). Interestingly, Mel1c also demonstrated expression in the ependyma of the brain ventricles and the choroid plexus, which along with the pineal gland, show no demonstration of [125I]iodomelatonin binding. The function of Mel1c at these sites is thus enigmatic, but it is suggested that the receptor gene could be translated into a non-functional truncated protein due to intronic splice variants (Reppert et al., 1995). As its mammalian orthologue, it is perhaps expected that GPR50 might share similar distributions in the brain. Indeed, in situ hybridisation studies on mouse, rat and hamster brain sections have demonstrated GPR50 mRNA at some of the same sites, including the optic nerve, optic chiasm, retina and choroid plexus (Drew et al., 2001). GPR50 mRNA expression was also demonstrated in several hypothalamic nuclei, including the lateral hypothalamic area (LHA), arcuate nucleus (ARC), paraventricular nucleus (PVN), paraventricular thalamic nucleus, median eminence and pituitary. However, the most intense sites of GPR50 mRNA expression demonstrated by these studies, along with subsequent analysis of protein expression (Sidibe et al., 2010)(Appendix A), reveals a more restricted localisation of GPR50 within the mouse, rat and human hypothalamus, with expression limited to the ependymal layer of cells lining the third ventricle and the dorsomedial nucleus.

Finally, although Mel1c has been detected in the skin of fish species (Confente et al., 2010; Sauzet et al., 2008), beyond this it appears to have limited peripheral expression, whereas in the mouse, Gpr50 expression has been demonstrated in a multitude of peripheral tissues, including the kidney, adrenal gland, intestine, heart, lung, testis and ovary (Drew et al., 2001). The functional significance of the evolution of Mel1c into GPR50 is unclear due to the lack of knowledge concerning both of their physiological roles. However, the altered pattern of expression and the loss of affinity of GPR50 for melatonin is probably not a neutral physiological event.

32

1.4 GPR50 expression in the hypothalamus

In the absence of an identified ligand, analysis of the tissue distribution of GPR50 provides valuable information relating to its potential physiological function, and the restricted expression of GPR50 within the hypothalamus, suggests some prominent role for the receptor at this site. The hypothalamus links the nervous system to the endocrine system via the pituitary gland, and is composed of distinct nuclear groups, which have been implicated in diverse physiological and behavioural responses, including control of reproduction, circadian responses, stress responses, fluid balance, ingestive behaviour, and thermoregulation.

Initial evidence using an engineered Gpr50-/- mouse tentatively points towards a potential role for this receptor in energy sensing or the regulation of energy metabolism. These animals exhibit marked metabolic alterations, being partially resistant to a hypercalorific, high fat diet, meaning they display attenuated weight gain when placed on this regime, despite consuming significantly more food per unit body weight than their wild type littermates (Ivanova et al., 2008). Moreover, Gpr50-/- mice have an elevated metabolic rate, and reduced fat stores when fed a normal chow diet. The hypothalamus is a centre of convergence and integration of multiple nutrient- related signals and reciprocally serves to coordinate neuroendocrine, behavioural and metabolic effectors of energy balance. Thus, the expression of GPR50 here, further lends support to a potential role for the receptor in the control of metabolism.

1.4.1 Hypothalamic control of energy balance

Some of the key nutrient-sensing hypothalamic neurons have been identified in the arcuate (ARC), the ventromedial (VMH) and the lateral (LH) nuclei of the hypothalamus (see Figure 1.3), and the molecular mechanisms underlying intracellular integration of nutrient-related signals in these neurons is subject to intensive investigation.

The ARC contains two distinct neuronal populations producing neuropeptide Y (NPY)/ agouti-related peptide (AgRP), and those producing α-melanocyte-stimulating hormone (α-MSH)/cocaine-and amphetamine-regulated transcript (CART) (Gao and Horvath, 2008). NPY is a potent orexigenic agent and reduces energy expenditure, whereas

33

Figure 1.3 Schematic representation of the mammalian hypothalamus. Location of the third ventricle (IIIV) and the surrounding hypothalamic nuclei with established roles in energy homeostasis; ARC, arcuate nucleus, VMN; ventromedial nucleus; LH, lateral hypothalamic nucleus; DMH, dorsomedial nucleus; PVN, paraventricular nucleus. Diagram by the author.

34

CART acts to inhibit food intake. AgRP and α-MSH have opposing actions on melanocortin receptors (MCR); α-MSH activates MC3 and 4Rs to inhibit food intake, whereas AgRP is an endogenous antagonist of these receptors, relieving inhibition of feeding and permitting stimulation of food intake (Ollmann et al., 1997).

The stimulatory and inhibitory neuropeptides expressed in the neuronal populations of the ARC, provide spatially distinct and opposing circuits to regulate energy expenditure and feeding behaviour. The ARC neurons receive blood-borne and neuronal signals that allow them to assess energy status, and the integration of these signals is reflected in the excitability of these cells (Blouet and Schwartz, 2010). In response to increased nutrient levels, anorexigenic α-MSH/CART neurons depolarise, whereas orexigenic NPY/AgRP neurons hyperpolarise, in order to inhibit feeding. The converse is true when nutrients are scarce. Further, activated NPY/AgRP neurons contact and inhibit α-MSH/CART neurons through the release of GABA, such that when this system is turned on, the opposing circuitry is silenced (Cowley et al., 2001). This unidirectional NPY/AgRP to α-MSH/CART neuronal interaction allows the tonic inhibition of satiety signals to promote feeding.

Through direct and multisynaptic projection pathways, the ARC neuronal populations can control behavioural, autonomic and neuroendocrine systems to maintain energy balance. Their projection to neurons in other hypothalamic nuclei, including those in the LH, VMN, PVN and DMH allows the produced neuropeptides to act at these sites (Gao and Horvath, 2008; Schwartz et al., 2000).

Bilateral lesions of the LH cause anorexia and weight loss (Schwartz et al., 2000), suggesting orexigenic signalling molecules are synthesised here. Indeed, glucose- inhibited neurons in the LH correspond to hypocretin/orexin neurons, while glucose- excited neurons correspond to cells containing melanin-concentrating hormone (MCH) (Blouet and Schwartz, 2010; Date et al., 1999). Both MCH and the orexins result in increased food intake when administered centrally (Edwards et al., 1999). NPY, AgRP and α-MSH immunoreactive terminals extensively contact MCH and orexin-expressing cells, and in turn, projections from the LH to the ARC allow orexins to exert direct excitatory actions on ARC NPY/AgRP neurons (van den Top et al., 2004) and indirect

35 inhibitory actions on α-MSH/CART neurons (Ma et al., 2007). In addition to the orexigenic neuropeptides, ‘satiety factors’, including bombesin, released from the gut act in the LH to promote physiological satiety (Bellinger and Bernardis, 2002).

In contrast to the LH, bilateral PVN lesions results in hyperphagia and obesity (Schwartz et al., 2000), indicating that anorexigenic signalling molecules are produced here. The PVN contains magnocellular neurosecretory cells, which project to the posterior pituitary, and synthesise the hormones oxytocin and vasopressin (Tasker and Dudek, 1993). The PVN also contains parvocellular neurosecretory cells that project to the median eminence and release hormones such as corticotropin-releasing hormone (CRH) and thyrotropin-releasing hormone (TRH) into the blood vessels of the hypophyseal portal system (Tasker and Dudek, 1993). When administered centrally CRH and TRH reduce food intake (Schwartz et al., 2000). NPY/AgRP neurons communicate with PVN neurons containing TRH and CRH, as well as neurons of the PVN that detect and integrate melanocortin signalling, to regulate their activities (Cowley et al., 1999). The PVN also directs sympathetic activity to brown adipose tissue, which alters energy metabolism by regulating thermogenic output by this tissue (Lechan and Fekete, 2006). Neurons of the PVN, similarly to those of the LH, project to the ARC to allow information between these hypothalamic sites to flow bi-directionally (Schwartz et al., 2000).

The neuropeptide phenotypes of VMN nutrient-sensing neurons remain largely unknown; however it contains a large population of glucoresponsive neurons and receives NPY, AgRP and α-MSH neuronal projections from the ARC (Blouet and Schwartz, 2010). The anorexigenic brain-derived neurotrophic factor (BDNF) is highly expressed in the VMN, and it is thought that ARC neurons have a role in activating these BDNF-expressing neurons to decrease food intake (Blouet and Schwartz, 2010).

Central glucose and fatty acid detection by neurons within the previously discussed nuclei allows the hypothalamus to sense nutrient availability (Jordan et al., 2010). Hypothalamic neurons also integrate signals provided by adipose tissue, gut hormones and pancreatic hormones, making them critical in determining whole-body nutrient availability and utilisation. Leptin is one such peptide hormone, and serves as both an

36 acute and long-term indicator of energy status, reflecting daily fasting/feeding cycles, as well as overall adipose stores. Leptin is released from peripheral white adipose tissue, circulating at levels corresponding to adipose tissue mass (Frederich et al., 1995). During nutrient abundance, leptin secretion is increased, leading to decreased food intake and increased thermogenesis, whereas leptin deficiency causes hyperphagia and obesity. Leptin acts via a specific isoform of the leptin receptor (Lep- Rb), which shows dense expression in the ARC (Schwartz et al., 2000). When circulating leptin levels are low, anorexigenic α-MSH/CART production is reduced, while expression of NPY/AgRP is stimulated (Ahima and Lazar, 2008). The reverse occurs during nutrient abundance when levels of leptin are high. Leptin receptors have also been identified in LH, PVN, VMH and DMH neurons, implying they are also targets for regulation by this adiposity signal, and along with their received input from ARC leptin receptor-expressing neurons, direct leptin’s actions on feeding and thermogenesis (Blouet and Schwartz, 2010). Insulin receptors are also highly concentrated in the ARC, and like leptin, insulin circulates at levels proportional to body fat content, and acts centrally to reduce food intake (Schwartz et al., 2000). Decreased insulin activates NPY/AgRP neurons and inhibits α-MSH/CART neurons, similarly to low plasma leptin.

1.5 GPR50 expression in the DMH

Intense expression of GPR50 mRNA and protein is demonstrated in the DMH of the hypothalamus. This nucleus receives both neural and humoral input from pathways critical in the regulation of feeding, body weight regulation and energy consumption, and in turn, projects to brain regions critical for the regulation of sleep and wakefulness, body temperature, and corticosteroid secretion (Bernardis and Bellinger, 1998). Studies using bilaterally DMH lesioned rats revealed that this causes hypophagia and resistance to diet-induced obesity in animals when maintained on a high-fat diet (Bellinger and Bernardis, 2002). The DMH contains glucose-sensitive and receptive neurons, suggesting it has a feeding system responsive to glucopenia; an idea supported by studies whereby DMH lesioned rats challenged with the glucoprivic agent 2-deoxyglucose or conversely a glucose infusion failed to increase or decrease their food intake, respectively, unlike control animals.

37

The DMH is composed of cells and fibres containing NPY, AgRP, CART, α-MSH, orexin-A and MCH, all neurotransmitters implicated in energy regulation, as previously discussed (Blouet and Schwartz, 2010). In addition, the DMH contains neurons expressing receptors for the satiety agent cholecystokinin-8 (CCK-8) (Bellinger and Bernardis, 2002), as well as the long form of the leptin receptor, so the DMH can form part of the circuitry that integrates multiple signals reflecting whole-body energy status.

Feeding and metabolic regulation are phenomena closely linked to thermoregulation, a process coordinated by the CNS through endocrine, autonomic and behavioural mechanisms. Most species respond to acute food restriction or hypercalorific feeding by lowering or elevating, respectively, their thermogenic output. Normothermia in a cold environment can be maintained through sympathetically regulated means such as the metabolic activation of brown adipose tissue (BAT) and cutaneous vasoconstriction, which serve to generate heat and limit heat loss, respectively. Involvement of the DMH in thermoregulation is well established (Dimicco and Zaretsky, 2007), and a polysynaptic connection exists between neurons of the DMH and interscapular brown adipose tissue (BAT) (Voss-Andreae et al., 2007). Moreover, the DMH also projects heavily to the PVN, a site which again has a neural link to BAT (Lechan and Fekete, 2006). Projections from the DMH to parvocellular neurons of the PVN also allow the DMH to relay signals that regulate the hypothalamic-pituitary-thyroid (HPT), and hypothalamic-pituitary-adrenal axis (HPA) (Evans et al., 2004; Mihaly et al., 2001). Therefore, GPR50-expressing neurons in the DMH could be involved in the regulation of energy homeostasis, or in sympathetic/endocrine thermoregulatory adjustments to changing energy status.

1.6 Potential outputs regulated by GPR50: The HPT and HPA axes

The expression pattern of GPR50 within the hypothalamus, along with phenotypic traits of the Gpr50-/- mice, suggests a possible influence of the receptor on the HPT and HPA axes (Figure 1.4). Mice lacking GPR50 demonstrate reduced Trh levels in the PVN (Appendix A), and elevated levels of the circulating corticosterone (Ivanova et al., 2008), suggesting abnormalities to both of these systems.

38

Figure 1.4 Schematic representation of the hypothalamic-pituitary-thyroid (HPT) and hypothalamic-pituitary-adrenal (HPA) axes.

Thyroid hormone (T3 and T4) secretion is initiated by thyrotropin-releasing hormone (TRH)-synthesising neurons of the PVN, which release TRH into the portal circulation at the median eminence. This induces release of thyroid-stimulating hormone (TSH) from the anterior pituitary into circulation which promotes the synthesis and release of

T3 and T4 from the thyroid gland. T3 and T4 exert a negative feedback effect at the level of the pituitary and hypophysiotropic TRH neurons. Glucocorticoid secretion (GC) is subject to a similar axis, where corticotropin-releasing hormone (CRH)-synthesising neurons of the PVN release CRH into the pituitary portal circulation at the median eminence and induce the release of adrenocorticotropic hormone (ATCH) from the anterior pituitary. ACTH promotes the synthesis and secretion of GCs from the adrenal cortex. GCs exert a negative feedback effect at the level of the pituitary and hypophysiotropic CRH neurons. By the author.

39

1.6.1 The HPT axis

The HPT axis plays an essential role in the maintenance of metabolic homeostasis. TRH-producing neurons of the PVN project to the median eminence (ME) where they are in close proximity to the capillaries of the hypophysial-portal system. Released TRH stimulates the anterior pituitary to secrete thyroid stimulating hormone (TSH) into the circulation, which in turn, stimulates the biosynthesis of the thyroid hormones, thyroxine

(T4) and triiodothyronine (T3) (Zoeller et al., 2007). Active thyroid hormones regulate the transcription of genes by binding nuclear receptors in cells throughout the body, and have pivitol roles in the regulation of basal metabolic rate, cardiac output, lipid metabolism and thermogenesis (O'Shea and Williams, 2002). A well-characterised negative feedback system means that TRH production is decreased in the PVN when circulating levels of thyroid hormones are high, with the reverse being true when circulating levels are depleted (Lechan and Fekete, 2006). TRH neurons are also subject to regulation by NPY, leptin, α-MSH and AgRP, and integration of these inputs provides an exact set point for the thyroid axis, when changes in nutritional status and external temperature occur (Nillni, 2010). Important afferent connections to the TRH neurons in the PVN include catecholamine neurons from the brainstem, which play a significant role in the up-regulation of Trh during cold-exposure (Zoeller et al., 2007), and neurons from the ARC and DMH, which act as metabolic sensors for the HPT axis (Fekete and Lechan, 2007).

Not all TRH-containing neurons of the PVN project to the ME (Nillni, 2010). These non- hypophysiotropic neurons therefore do not serve to direct hypophysiotropic function, but rather play a different role. Central administration of TRH decreases food and water intake, increases locomotor activity and increases body temperature (Choi et al., 2002; Schuhler et al., 2007), without necessarily altering the HPT axis, and it is thought these non-hypophysiotropic neurons control such outputs mediated by TRH. Neuronal TRH projections from the PVN to the VMN and LH have been described, and are proposed to be critical for TRH-induced anorexia, by regulating the production of neuropeptides at these sites (Nillni, 2010). Non-hypophysiotropic TRH neurons in the PVN can also act on the sympathetic nervous system to induce thermogenesis in brown adipose tissue (Lechan and Fekete, 2006) in concert with a simultaneous stimulation of the HPT axis by hypophysiotropic TRH neurons.

40

Although T4 is the major product released from the thyroid gland, T3 is the more biologically active metabolite. Therefore, the activity of thyroid hormone in target cells is determined by the local activities of iodothyronine deiodinase (Dio) enzymes (Bianco et al., 2002). T4 is converted to T3 by Dio1 and Dio2, whereas Dio3 inactivates T4 and T3 by converting them to reverse T3 and inactive diiodothyronine (T2), respectively (Bianco et al., 2002). T4 is transported into the brain much more efficiently than T3, and thus the primary source of central T3 comes from the local conversion of T4. Within the CNS, this is mainly catalysed by Dio2 (Fekete and Lechan, 2007). Within the hypothalamus, Dio2 activity is detectable in the pituitary and mediobasal hypothalamus, specifically in the ME, ARC and VMN. However, marked expression of Dio2 has been established in tanycytes of the ependymal layer lining the third ventricle, which may be crucial in the regulation of T3 availability in the hypothalamus (Rodriguez et al., 2005). These cells are ideally placed to extract T4 from the CSF of the third ventricle at their apical surface, or from the bloodstream by their end feet processes that terminate on portal capillaries and vessels of the ARC.

During fasting, reduced TRH synthesis is necessary to conserve energy expenditure, by decreasing thyroid hormone-stimulated mitochondrial respiration and thermogenesis (Fekete and Lechan, 2007). Fasting results in decreased TRH synthesis in the PVN, even though levels of circulating thyroid hormone are low. This paradoxical response is a result of transcriptional regulation of Dio2 in the hypothalamus. During fasting, increased expression of Dio2 is observed; leading to an increase in locally formed T3 and increased negative feedback within the hypothalamus, causing suppression of TRH (Diano et al., 1998).

Thyroid hormone signalling has an interesting role in the regulation of seasonal responses, first demonstrated through studies using the Japanese quail, which showed that thyroidectomy blocked many of the typical seasonal responses exhibited by this animal, which could be restored by a single injection of T4 (Follett and Nicholls, 1985).

Suppression of T3 within the hypothalamus has since been associated with reproductive quiescence in a number of seasonal breeders, achieved by different mechanisms in different species; long-day breeding Syrian hamsters decrease expression of hypothalamic Dio2 during short photoperiod, whereas Siberian hamsters increase expression of Dio3 (Murphy and Ebling, 2011). In contrast, the short-day breeding Saanen goat exhibits decreased Dio2 expression on long photoperiod (Yasuo 41 et al., 2006). Studies also suggest that a change in hypothalamic availability of thyroid hormone is the key determinant of annual weight regulation. Contrary to prior expectations, a consistent pattern of changes indicative of an animal in a state of energy deficit is not observed in hypothalamic gene expression during short-day induced weight loss in the Siberian hamster. For example, studies have failed to detect changes in CART, AgRP, NPY or orexin expression in the ARC (Morgan and Mercer, 2001), and POMC (the precursor polypeptide to α-MSH) shows strong downregulation with short-photoperiod (Rousseau et al., 2002), which seems counterintuitive given that these animals reduce food intake and energy expenditure during this time. Furthermore, lesions of the ARC do not impede photoperiodic body weight regulation (Ebling et al., 1998), suggesting that changes upstream of this energy-regulation centre must be altered in these animals. In the Siberian hamster, hypothalamic application of

T3 during short photoperiod promotes a long-day phenotype; increasing food intake and body weight, without affecting the peripheral thyroid axis (Barrett et al., 2007). It is hypothesised that T3 exerts its seasonal hypothalamic actions via structural and functional plasticity, rather than altering synthesis and secretion of neuropeptides (Murphy and Ebling, 2011). Clear evidence for neurogenesis in the adult hypothalamus demonstrates that this part of the brain retains such potential for plasticity (Kokoeva et al., 2007). Furthermore, tanycytes (the major site of Dio2 and Dio3 expressions, and thyroid hormone uptake) show the potential for marked cyclical remodelling. In rats, ultrastructural studies of gonadotropin releasing hormone (GnRH) neurons, which project to the median eminence and regulate the reproductive axis, revealed that they are enwrapped by tanycyte processes under conditions of low gonadotropin output, but during the preovulatory surge, a structural remodelling of tanycytes occurs, resulting in the release of the engulfed axons, allowing GnRH neurons to access the pituitary portal blood system (Prevot et al., 2010). So although it is clear that hypothalamic T3 availability plays a major role in seasonal cycles of reproduction, body weight and energy balance, the cellular mechanisms which ultimately generate these cycles remain unknown.

1.6.2 The HPA axis

Activated by a diverse range of physical and psychological stressors, the HPA axis plays a central role in adaptive stress responses, the end point of which controls the release of glucocorticoids (GCs) from the adrenal gland (Herman et al., 2003). GCs exert numerous effects, acting through intracellular GC receptors to modify

42 transcription of key regulatory proteins, culminating in the mobilisation of energy to allow the body to deal with the stressor. CRH expressing neurons originating in the PVN stimulate the synthesis and release of adrenocorticotropin (ACTH) from the anterior pituitary, which then acts to promote synthesis and release of GC from the adrenal gland (Watts, 2005). GCs can feedback on the HPA axis to decrease CRH and ACTH synthesis and release.

Central administration of CRH inhibits feeding and stimulates sympathetically-mediated thermogenesis and lipolysis (Hillebrand et al., 2002), by suppressing orexigenic neural pathways involving NPY. CRH inhibits NPY synthesis (Krysiak et al., 1999), and administration of CRH-antagonist potentiates NPY-induced food intake (Heinrichs et al., 1993). GCs on the other hand, stimulate food intake, and this opposing action may occur due to negative feedback of GCs on CRH neurons, thereby decreasing the anorexigenic effects of CRH (Nieuwenhuizen and Rutters, 2008). Alternatively, studies showing that adrenalectomised rodents do not increase food intake in response to exogenous NPY treatment (Stanley et al., 1989), suggest that GCs may increase feeding by acting as a permissive factor for the orexigenic effects of NPY. The relationship of NPY with the HPA axis however, is complex. NPY can stimulate this system by increasing CRH synthesis in the PVN and release in the ME (Krysiak et al., 1999), and thus a negative feedback appears between CRH and NPY. However the source of NPY might be a determining factor in the regulation of CRH expression. NPY/AgRP neurons from the ARC innervating the PVN have a chronic inhibitory effect on CRH expression, whereas NPY-expressing noradrenergic populations from the brainstem that innervate the PVN, activated in response to glucoprivation, infection and inflammation, appear to have an acute activating effect on CRH neurons (Fuzesi et al., 2007).

Studies using both rodents and have demonstrated that the HPA axis fails to be activated under fasting conditions (Rohleder and Kirschbaum, 2007). CRH expression in the PVN is inhibited during fasting, potentially due to activation of ARC NPY/AgRP neurons (Fuzesi et al., 2007). ICV infusion of α-MSH reactivates CRH gene expression in the PVN of fasting animals (Fekete et al., 2000), suggesting that attenuation of MCR signalling during fasting leads to reductions in CRH expression. Further studies revealed that leptin increases CRH expression and release in the PVN,

43 and so reduced leptin during fasting might also contribute to reduced expression of CRH (Mastorakos and Zapanti, 2004).

Elevated circulating levels of the GC corticosterone have been observed in Gpr50-/- mice (Ivanova et al., 2008), indicating that these animals may have a heightened stress response. Dysregulation of the HPA axis has been implicated in the development of several human disorders, with evidence suggesting that abnormally elevated activation of this system is associated with the pathogenesis of psychiatric disorders, such as major depressive disorder (MDD), bipolar affective disorder (BPAD) and schizophrenia (SCZ) (Phillips et al., 2006). Intriguingly, the gene encoding GPR50 is located on Xq28, a region previously implicated in BPAD (Thomson et al., 2005). Moreover, reported associations between polymorphisms in Gpr50 and various mood disorders have been revealed. Three polymorphisms that result in alteration to the C-terminal tail of GPR50 have been described; two represent single amino acid mutations (Thr532Ala and Val606Ile), and one results in an insertion of four amino acids (TTGH) at position 501. Two additional non-coding polymorphisms (rs1202874 and rs2072621) found in the intron between exons 1 and 2 of Gpr50 have also been reported. Increased risk of BPAD and MDD in a Scottish population was associated with the deletion variant 502- 505 (Thomson et al., 2005), although attempts to replicate the findings in a Northern Swedish population failed to confirm these findings (Alaerts et al., 2006). In women, the Val606Ile single-nucleotide polymorphism (SNP) has been associated with MDD (Thomson et al., 2005), the rs1202874 SNP has also been associated with BPAD (Macintyre et al., 2010), and the rs2072621 SNP has been associated with SCZ (Thomson et al., 2005). A recent report further suggests that the intronic polymorphism rs2072621 variant is also a gender-specific risk factor for seasonal affective disorder (Delavest et al., 2011). These data imply that variants in Gpr50 represent a risk for a spectrum of mood disorders, however further studies using larger samples are necessary to firmly establish links between GPR50 and the development of these affective disorders.

1.7 GPR50 expression in the ependymal layer of cells lining the third ventricle

High levels of GPR50 expression are further observed in the ependymal layer of cells lining the third ventricle (Barrett et al., 2006; Drew et al., 2001; Ivanova et al., 2008; Sidibe et al., 2010). Ependymal cells lie at the interface between the ventricular cavities

44 and the brain parenchyma. Although the functions of ependymal cells remain enigmatic, it is generally accepted that they constitute the brain-cerebrospinal fluid (CSF) barrier, serving as a selectively permeable physical and biochemical junction between the neural parenchyma and CSF. CSF is primarily generated by choroid plexus cells and probably to a small extent by diffusion of extracellular fluid from the brain, and is moved through the ventricular system by the coordinated beating of cilia on the apical surface of ependymocytes (Worthington and Cathcart, 1963). In addition to cilia, ependymal cells have numerous microvilli, increasing the surface area of ependymocytes, to aid contact with the CSF (Szele and Szuchet, 2003). The CSF may be an important route by which some nutrients reach the CNS and may also act as a means of communication within the CNS whereby it carries hormones and transmitters between different areas of the brain (Brown et al., 2004). Ependymal cells lining the third ventricle are metabolically active, possessing receptors for numerous growth factors (Mathew, 2008), along with a vast number of transmembrane transporters and enzymes involved in hormone metabolism (Bruni, 1998; Del Bigio, 1995). Expression of the long-form of the leptin receptor (Baskin et al., 1999), the glucose transporters GLUT1, GLUT2 and GLUT4 (Farrell et al., 1992; Jetton et al., 1994; Leloup et al., 1994; Ngarmukos et al., 2001), and the sodium/glucose cotransporter, SGLT1 (Briski and Marshall, 2001) have been demonstrated in these cells. Expression of key enzymes involved in energy metabolism have further been shown, including endothelial lipase (Paradis et al., 2004), glucokinase (Jetton et al., 1994; Maekawa et al., 2000), and glycogen phosphorylase (Cataldo and Broadwell, 1986; Pfeiffer et al., 1990). Intracerbroventricular (ICV) administration of glucose, or inhibitors of its metabolism, results in modulation of peripheral energy homeostasis (Flynn and Grill, 1985; Miselis and Epstein, 1975; Tordoff et al., 1988), as does ICV administration of ketones (Park et al., 2011). Similarly, ICV treatment with adipokines such as adiponectin and leptin (Qi et al., 2004) or the pancreatic hormone insulin (Foster et al., 1991; Obici et al., 2002) also modifies peripheral energy metabolism. These studies further suggest that energy status in the brain can be monitored via CSF-borne signals, and that altered availability of metabolic intermediates/end products may serve as a stimulus for CNS-mediated activity aimed at restoring homeostasis. The sidewall of the third ventricle can be divided into distinct zones based on the cellular arrangement: multilayered arcuate tanycyte zone, monolayered arcuate tanycyte zone, irregular tanycyte zone, double layered tanycyte zone, mixed cell zone, and ciliated cuboidal cell zone (see Figure 1.3) (Mathew, 2008). The multilayered arcuate tanycyte zone is located in the most ventral part of the ependymal wall, and the remaining zones extend dorsally in the order listed.

45

Figure 1.5 Schematic summarising the cell types of the ependymal layer lining the third ventricle. Regional variation in cellular organisation in different zones is denoted by the dashed lines. Ciliated ependymal cells are represented by the blue and green hexagons, tanycytes by the blue open circles. Localisation and distribution of α and β tanycytes are depicted, with β1 tanycytes comprising the median eminence-arcuate nucleus (ARC) barrier, and β2 tanycytes forming the median eminence-cerebrospinal fluid barrier. Projections of α1, α2 and β1 tanycytes to blood vessels and neurons of the ARC and ventromedial nucleus (VMH) are also shown. By the author.

46

The anatomical location and morphology of cells demonstrating GPR50 mRNA and protein expressions in the ependyma is suggestive of tanycytes. As described, tanycytes reside at the base of the third ventricle, extending approximately one-third dorsally along the ventricular wall. Hypothalamic tanycytes are bipolar cells, with cell bodies possessing cilia or microvilli at their apical surface that extend into the ventricular space and contact the CSF, and long basal processes that terminate in “end feet” that contact either the portal blood vessels or surrounding hypothalamic neurons (Rodriguez et al., 2005). In this way, tanycytes are thought to play a role in relaying signals from the CSF and may link the CSF to endocrine events. Four subtypes of tanycytes can be distinguished; α1, α2, β1, and β2, based on their spatial orientation within the ependymal layer (see figure 1.5), as well as differential expression of certain molecules, which confer functional differences (Rodriguez et al., 2005). The β2 tanycytes around the median eminence have a known role in the regulation of gonadotrophin-releasing hormone (GnRH) secretion (Prevot et al., 2010), however the roles of the more dorsal α1,2 tanycytes remain uncharacterised. Their cell bodies project towards neurons of the arcuate (ARC) and ventromedial hypothalamic nuclei (VMH), key centres in the control of food intake and energy metabolism, suggesting tanycytes may influence these circuits and have a role in nutrient sensing. Indeed, tanycytes have been implicated in glucosensing as they express several glucose transporters and hexokinase (Rodriguez et al., 2005), and recent studies demonstrate that glucose and a variety of its non-metabolisable analogues evoke intracellular Ca2+ waves along the tanycyte cell body (Frayling et al., 2011). They were similarly responsive to histamine and acetylcholine – other transmitters associated with the drive to feed. Therefore tanycytes seem ideally placed and equipped to respond to a number of neuronally-derived and circulating transmitters and metabolites and their proximity to the ARC and VMH is suggestive of possible role in the regulation of feeding and energy balance. Tanycytes are also the key site of throxine deiodinase type II (Dio2) expression, the enzyme responsible for the conversion of T4 into the biologically active product T3 (Rodriguez et al., 2005), and thus may be intrinsically involved in control of the hypothalamic-pituitary-thyroid axis and supply the primary source of local T3 in the brain.

GPR50 mRNA expression has been demonstrated in the entire lateral walls of the ventricle in caudal ARC sections of the hypothalamus, but becomes restricted to the ventral third of the wall in rostral ARC sections (Barrett et al., 2006). This limited pattern

47 of GPR50 expression could imply that signalling from the receptor has a key function in influencing the cells that surround the ependyma, specifically in these regions.

1.8 Variations in GPR50 expression

In mice, Gpr50 expression in the DMH and ependymal layer of cells lining the third ventricle is highly responsive to energy status. mRNA expression is reduced at both sites in response to fasting, and is restored within 12 hours of re-feeding (Ivanova et al., 2008). Expression of Gpr50 is similarly suppressed in mice maintained for 5 weeks on a hypercalorific, high fat diet (Ivanova et al., 2008).

In the Siberian hamster, ependymal Gpr50 expression is decreased following exposure of animals to short photoperiod (Barrett et al., 2006). The Siberian hamster exhibits robust annual rhythms, showing seasonal changes in a host of physiological and behavioural responses. In reaction to short day-lengths and lower ambient temperatures characteristic of winter, these animals spontaneously reduce food intake, and body weight thus declines by around 40% over a 12-week period, mainly in the form of mobilised fat stores (Rousseau et al., 2003). Other winter adaptations include a change in pelage colour, improved fur insulation, reproductive quiescence, and display of daily torpor, where body temperature, heart rate and metabolic rate are decreased for 4-8 hours several times a week (Berriel Diaz et al., 2004). Such changes enable Siberian hamsters to survive the harsher climate of winter, when ambient temperatures drop and food availability is limited.

Along with its hypothalamic localisation, the co-incident reduced expression of Gpr50 in Siberian hamsters in short photoperiod, at a time when they are undergoing such marked metabolic alterations, and the sensitivity of its expression to energy status in mice, suggests a possible role for this receptor in energy sensing or the regulation of metabolism.

Genetic studies further imply some role for GPR50 in mechanisms controlling energy expenditure. Three coding polymorphisms (previously discussed in section 1.6.2) in the gene encoding GPR50 have been associated with elevated circulating triglyceride

48 levels in humans (Bhattacharyya et al., 2006). An intronic Gpr50 polymorphism (rs1202874) was also significantly associated with increased triglyceride levels, as well as reduced amounts of circulating high-density lipoprotein (HDL)-cholesterol (Bhattacharyya et al., 2006). These findings suggest a role for GPR50 in the regulation of lipid metabolism.

1.9 Torpor

A striking feature of the Gpr50-/- mice is that upon fasting, they enter a state of deep torpor, characterised by a considerable drop in metabolic rate (VO2) and hypothermia during which core body temperature (Tb) falls to within a few degrees of ambient temperature. As the receptor is also decreased in winter phenotype Siberian hamsters (Barrett et al., 2006), at a time when these animals show a heightened torpor response, it can be supposed that GPR50 may contribute to key networks controlling hypometabolism.

1.9.1 Definition of torpor

Torpor has been defined as a temporary physiological state characterised by a controlled lowering of metabolic rate, followed by Tb cooling and inactivity (Geiser and

Ruf, 1995). In different species, there is great variation in the magnitude of Tb reduction and duration of the torpor bouts. Shallow, daily torpor is demonstrated by Siberian hamsters, where Tb of torpid animals remains between 12 and 22°C and torpor lasts for a matter of hours. Other species enter a state of deep torpor, or hibernation, such as the Little brown bat, which lowers Tb to between 1.3 and 9°C, and can persist in this hypometabolic state for up to 40 days (Melvin and Andrews, 2009). Torpor is a highly effective adaptation that various mammals use to cope with periods of energetic constraint generated by decreased quantity and quality of available food. Daily torpor can be induced in laboratory mice by chronic food restriction (Hudson and Scott, 1979); the subtropical nectar-eating blossom-bat enters torpor during the summer, when nectar availability is reduced (Geiser, 2004); and although Siberian hamsters express daily torpor spontaneously in winter, it may also occur at other times of the year in response to acute energy shortage (Ruby, 2003). Interestingly, in mice, daily torpor can be induced by high-foraging costs, in that animals forced to wheel run for food utilised

49 torpor in order to reduce their energy budget rather than increasing activity to get more food (Schubert et al., 2010).

1.9.2 Timing of daily torpor

In some species, annual cycles of torpor are under photoperiodic control. Exposure of Siberian hamsters to short photoperiod (SP) alone leads to the expression of daily torpor, and melatonin implants in hamsters held in long photoperiod can mimic these effects, whereas pinealectomy prevents acclimation to SP and concurrent torpor (Kortner and Geiser, 2000). However, although responsible for entrainment to photoperiod, melatonin is not directly involved in the expression of torpor, as pinealectomy in hamsters following acclimation to SP does not block the already initiated expression of torpor (Ruby et al., 1989). However, daily torpor is not strictly a seasonal event influenced by photoperiod alone. Among the changes induced by exposure to short days, Siberian hamsters spontaneously reduce food intake and exhibit marked reductions in body weight (Rousseau et al., 2003). Therefore it remains that decreased food availability and low body energy reserves appear to be the most important stimuli for the expression of daily torpor.

Typical daily changes in metabolic rate and Tb can be observed in most animals and birds, with both factors decreasing during the rest phase (Heldmaier et al., 2004). Daily torpor may therefore be viewed as an extension of the animal’s normal daily resting period, utilised to further reduce energy expenditure. As daily torpor in most heterotherms occurs during the rest phase, this implies it is a component of normal circadian organisation. There is compelling evidence for the involvement of the hypothalamic SCN in the maintenance of correctly timed daily torpor bouts. Although not essential for the expression of torpor, ablation of the SCN disrupts its temporal organisation (Ruby and Zucker, 1992). Further, torpor bouts can readily re-entrain to shifts of the photophase (Geiser and Baudinette, 1985), and in animals housed under constant conditions, free-running rhythms of torpor bouts are demonstrated, which is a typical phenomenon of processes regulated by the SCN (Ruby, 2003). The timing of the torpor bouts demonstrated by Gpr50-/- mice appear to be gated by circadian timing systems; entry into torpor occurs at approximately the same time of day, irrespective of when food restriction commences, and arousals occur prior to the introduction of food.

50

Furthermore, Gpr50-/- mice fasted for 48 hours in constant light, exhibit two torpor bouts, timed at the same relative phase each day (Appendix A).

It appears that both the photoperiod and the time of feeding affect the timing of torpor onset in Siberian hamsters. Food restriction has been demonstrated to maintain diurnal rhythmicity of daily torpor onset in SCN-ablated Siberian hamsters and also influences the circadian timing of torpor onset in intact animals (Paul et al., 2004). This suggests that timing of torpor induction is determined by the circadian pacemaker’s control over eating patterns rather than direct SCN regulation. Another hypothalamic area implicated in the organisation of torpor is the PVN. Like the SCN, ablation of this nucleus can disrupt its timing but do not eliminate its occurrence (Ruby, 1995).

Arousal from torpor in Siberian hamsters is accompanied by a brief surge in plasma concentrations of melatonin (Larkin et al., 2003). Levels peak during the early arousal phase, and subsequently decline as body temperature rises, until returning to basal levels upon achievement of euthermia. This melatonin surge occurs even during the light phase, which is surprising considering the typically rapid suppression of melatonin secretion induced by light exposure. It is thought that during arousal, strong sympathetic stimulation driven by the SCN to trigger thermogenesis in brown adipose tissue, increase heart and breathing rates, and stimulate lipolysis in white adipose tissue, also activates sympathetic fibers that terminate on the pineal gland, and this stimulation overrides the inhibitory effects of light on melatonin secretion (Larkin et al., 2003). Intact sympathetic nervous system (SNS) activity is not just essential for arousal, but also entry into torpor. Blockade of peripheral SNS signalling in Siberian hamsters eliminates the display of torpor (Braulke and Heldmaier, 2010), and dopamine-hydroxylase knockout mice, which lack norepinephrine (NE), are not able to enter fasting-induced torpor (Swoap and Weinshenker, 2008). Potentially, SNS- mediated reduction in circulating leptin, a permissive signal for mediating torpor initiation (Freeman et al., 2004), is the reason this system is required for the onset of torpor.

51

1.9.3 Mechanisms regulating torpor induction

Siberian hamsters do not begin to express spontaneous daily torpor until body mass and fat stores are at their nadir. During this time, circulating plasma leptin concentrations are low. Ob/ob mice are unable to produce leptin, and exhibit spontaneous bouts of torpor, despite having massive fat stores (Himms-Hagen, 1985; Webb et al., 1982). Exogenous leptin treatment eliminates torpor in ob/ob mice, and also blocks torpor in Siberian hamsters (Freeman et al., 2004), suggesting that reduced leptin concentrations provides a signal that mediates torpor initiation. Leptin signalling within the hypothalamus enhances thermogenesis, perhaps through a mechanism involving melanocortin signalling (Haynes et al., 1997). Inhibition of AgRP and stimulation of α-MSH turnover and release by leptin signalling leads to activation of melanocortin 3/4 receptors (MC3/4R) located in the PVN, which are inhibited by AgRP and activated by α-MSH (Mounien et al., 2010). Activated MCR signalling results in increased sympathetic outflow to BAT and enhanced thermogenic output (Ahima, 2006). Reduced thermogenesis during torpor is essential to the maintenance of decreased Tb and energy expenditure during this hypometabolic state; therefore reduced leptin signalling would appear necessary to maintain diminished BAT thermogenesis. However, although necessary for its occurrence, simply reducing leptin concentrations is not sufficient to trigger a torpor bout (Freeman et al., 2004). Further, leptin treatment fails to block torpor in A-ZIP/F1 mice, which have virtually no white adipose tissue stores, and readily enter a state of hypometabolism in response to fasting (Gavrilova et al., 1999). Therefore some other leptin-independent signal is required for entry into torpor.

Fibroblast growth factor 21 (FGF21) is a hormone produced in the liver and pancreas that promotes lipolysis, decreased serum glucose and increased serum ketone levels. Increased FGF21 expression also predisposes fasted mice toward entry into torpor, suggesting it regulates key aspects of the global torpor response (Inagaki et al., 2007).

ICV injection of NPY reliably induces torpor-like hypothermia, resembling natural torpor in hamsters (Paul et al., 2005). NPY-induced torpor is also produced by ICV administration of an NPY-Y1 receptor agonist, and can be blocked by co-injection of NPY and an NPY-Y1 , suggesting that activation of this receptor is both sufficient and necessary for NPY-induced torpor in hamsters (Dark and Pelz,

52

2008; Pelz and Dark, 2007). Interestingly, the stomach-derived hormone ghrelin, which is increased in response to fasting and acts within the ARC to increase expression of

NPY, has been shown to deepen the Tb minimum of a fasting-induced torpor bout (Gluck et al., 2006). This effect is lost in Npy-/- mice, suggesting that ghrelin requires

NPY neurons to elicit its effect on Tb, and further highlighting the importance of this circuitry in the control of torpor.

Thyroid hormones have a critical role in controlling basal metabolic rate by regulating energy expenditure in peripheral tissues, as well as acting on hypothalamic neurons of the ARC and PVN to trigger homeostatic mechanisms directly in the brain. As previously discussed, altered thyroid hormone signalling has been implicated in both seasonal and acute adaptive metabolic responses (Bechtold and Loudon, 2007;

Coppola et al., 2007; Rogers et al., 2009). Administration of T3 to mice or Siberian hamsters substantially lowers metabolic rate and Tb, along with an increase in lipid oxidation, which mimics naturally occurring torpor bouts (Braulke et al., 2008; Scanlan et al., 2004). During starvation, induction of Dio2 increases T3 availability to ARC neurons, leading to increased expression of NPY and decreased expression of TRH in the PVN (Fekete and Lechan, 2007). Repressed TRH expression serves to reduce thermogenic output to BAT (Rogers et al., 2009), and thus similar increased expression of Dio2, higher hypothalamic T3 concentrations and reduced expression of TRH could be expected during torpor.

The occurrence, depth and duration of torpor are influenced by dietary fatty acids. High levels of unsaturated fatty acids (UFAs) maintain the fluidity of cellular plasma membranes at decreased body temperatures, as they have a lower melting point than saturated fatty acids (SFAs). Acclimation to cold is generally accompanied by increased UFAs and decreased SFAs in tissues and cell membranes (Munro and Thomas, 2004). Diets rich in UFAs result in incorporation of more UFAs into body lipids, and more frequent, deeper and prolonged torpor bouts in several species, whereas the opposite is true for diets rich in SFAs. In line with this, when Siberian hamsters are exposed to short photoperiod, they automatically select a diet richer in UFAs than those maintained on long photoperiod (Hiebert, 2000).

53

1.10 Conclusion

As an orphan receptor, GPR50’s ligand and physiological function is unknown. Previous studies implicate the receptor in the hypothalamic control of energy expenditure and feeding behaviour (Ivanova et al., 2008). Specifically, Gpr50 expression in the brain is highly responsive to energy status being decreased by both fasting and high fat diet feeding, and Gpr50-/- mice demonstrate elevated metabolic rate, reduced fat accumulation, and partial resistance to diet-induced obesity. In humans, polymorphisms in Gpr50 have been linked to elevated circulating triglycerides and cholesterol levels (Bhattacharyya et al., 2006), as well as psychiatric affective disorders (Delavest et al., 2011; Macintyre et al., 2010; Thomson et al., 2005). Finally, Gpr50-/- mice are prone to fasting-induced torpor, and Gpr50 mRNA expression is suppressed in Siberian hamster exposed to SP (Barrett et al., 2006), a time when they lose body weight and express spontaneous torpor.

1.11 Overall aims

To broaden our understanding of the function of GPR50, this thesis will address the following:

Investigation of putative intracellular signalling pathways mediated by GPR50.

Characterisation of GPR50 in white adipose tissue.

Examination of gene expression within the hypothalamus to identify pathways altered in Gpr50-/- mice, potentially underlying the metabolic phenotype of these animals, and to assess gene changes during torpor.

54

CHAPTER 2

MATERIALS AND METHODS

55

2.1 Expression Constructs

The following plasmids encoding whole-length murine GPR50 (wGPR50), and a truncated form of the receptor (tGPR50) were kind gifts from Dr Jian Li, University of Manchester: pcDNA3.1/myc-His(A)-wGPR50 (wGPR50myc), pcDNA3.1/myc-His(A)- tGPR50 (tGPR50myc), pEGFP-N1-wGPR50 (wGPR50EGFP), pEGFP-N1-tGPR50 (tGPR50EGFP), pcDNA3, pHcRed1-N1/1. shRNA against Gpr50 (sh; short hair-pin) was purchased from Sigma (CCGGGCCAGCTCTAATCATCTTCATCTCGAGATGAAGATG ATTAGAGCTGGCTTTTT, TRCN0000025780) and Mission Nontarget shRNA Control Vector, SHC002 was used as control (Sigma; encodes an shRNA targeting a nonsense sequence). The following constructs, encoding human Gα subunits were donated by

Graham Davies, AstraZeneca: pcDNA3.1Gαi1, pcDNA3.1Gαi2, pcDNA3.1Gαi3, and pcDNA3.1Gαs.

2.2 Cloning

2.2.1 Hemagglutin (HA) epitope tagging

HA tags were added to the human Gα subunits and to murine wGPR50 and tGPR50. HA pcDNA3-HA-Gαi1-3/s ( Gαi1-3/s) were made by amplifying polymerase chain reactions

(PCRs) (section 2.2.1), using pcDNA3.1-Gαi1-3/s as templates and the primers listed in Table 2.1, and subsequent cloning (sections 2.2.2-2.2.5) into pcDNA3. pcDNA3-HA- wGPR50 (HAwGPR50), HA-wGPR50-pEGFP-N1 (HAwGPR50EGFP) and HA-tGPR50- pEGFP-N1 (HAtGPR50EGFP) were similarly created using RT-PCR reactions with pcDNA3.1/myc-His(A)-wGPR50, pEGFP-N1-wGPR50 and pEGFP-N1-tGPR50 as templates, respectively, and the primers listed in Table 2.1, before cloning into pcDNA3, or pEGFP-N1, as indicated in the construct names. The forward primers contained sequence encoding the HA epitope, and both forward and reverse primers contained restriction enzyme sites to allow sub-cloning of the amplicons in-frame, into the appropriate vectors. The sequences for the Gα subunits and GPR50 were taken from the genome browser website Ensembl, and all primers were obtained from MWG biotechnology, Germany.

56

Construct Sequence (5’ – 3’)

HA GC ATGTACCCATACGACGTCCCAGACTACGCTATGGGCTGCACGCTGAG Gαi1 F GAATTC

R GCCTCGAGTTAAAAGAGACCACAATCTTTTAGATT

Ha GC ATGTACCCATACGACGTCCCAGACTACGCTATGGGCTGCACCGTGAG Gαi2 F GAATTC

R GCCTCGAGTCAGAAGAGGCCGCAGTCCTTCAG

HA GC ATGTACCCATACGACGTCCCAGACTACGCTATGGGCTGCACGTTGAG Gαi3 F GAATTC

R GCCTCGAGCTAATAAAGTCCACATTCCTTTAAGTTG

HA GC ATGTACCCATACGACGTCCCAGACTACGCTATGGGCTGCCTCGGAAC Gαs F AAGCTT

R GCCTCGAGTTAGAGCAGCTCGTACTGAC

HAwGPR50 F CGGAATTCATGTACCCATACGACGTCCCAGACTACGCTATGGCCACGGTCCCCAAGAG

R GCCGCTCGAGTCACACAGCCATCTCATCAGAGCAA

HAwGPR50EGFP F CGCTCGAGATGTACCCATACGACGTCCCAGACTACGCTATGGCCACGGTCCCCAAGAG

R CGGCGGATCCCGCACAGCCATCTCATCAGAGCAATCAG

HAtGPR50EGFP F CGCTCGAGATGTACCCATACGACGTCCCAGACTACGCTATGGCCACGGTCCCCAAGAG

R CGGCGGATCCCGTGCAGTAGGCTCAGCAAACTGGTTG

Table 2.1 Primers used for RT-PCR epitope tagging. The primers used to amplify Gα subunits and w/tGPR50 are shown. F and R are the forward and reverse primers, respectively, the HA sequence is underlined and restriction enzyme sequences are in bold italics.

2.2.1 Polymerase chain reaction (PCR)

PCR reactions were performed in a volume of 25µl, with final concentrations of reagents as indicated in Table 2.2. The enzyme used was the Expand High Fidelity PCR Taq polymerase (Roche), which has a 3’-5’ exonuclease activity to ‘proof-read’ the new sequence, enhancing the reliability of the transcript produced.

Deoxynucleotide tri-phosphates were obtained from Bioline and nuclease free H2O from Ambion. The reaction mix was amplified in a thermal cycler, using the following protocol: 94°C for 2 min (denaturing), then 35 cycles of 94°C for 20 sec (melting), 56°C

57 for 30 sec (annealing), and 72°C for 40 sec (for Gα subunits, tGPR50) or 1 min 30 sec (for wGPR50) (extension).

Reagent Final Concentration Per Reaction F primer (10µM) 0.4µM R primer (10µM) 0.4µM Plasmid DNA template 100ng Deoxy nucleotide triphosphates 0.2mM (dNTP) (2.5mM) Expand High Fidelity 10x buffer 1x Expand High Fidelity Taq DNA 2.6 Units polymerase

Made up to 25µl in nuclease free H2O Table 2.2 PCR mastermix.

2.2.2 PCR amplicon gel excision and purification

The resulting PCR products were separated by molecular size on 1% (w/v) agarose (Sigma, UK) gels, containing ethidium bromide (Sigma, UK), which intercalates into the DNA and allows its visualisation under U.V. light. PCR amplicons were excised from agarose gels using a sterile scalpel blade and purified using a QIAquick Gel Extraction Kit (QIAGEN, UK), according to the manufacturer’s protocol. The DNA was digested with 10 Units of appropriate restriction enzymes (HAGαi1-3 were digested with EcoRI and XhoI; HAGαs was digested with HindIII and XhoI; HAwGPR50 was digested with EcoRI and XhoI; HAwGPR50EGFP and HAtGPR50EGFP were digested with XhoI and BamHI; all restriction enzymes were obtained from New England Biolabs), for 3 h at 37°C. Empty pcDNA3 and pEGFP-N1 vectors were similarly digested (pcDNA3 cut with EcoRI and XhoI, or HindIII and XhoI; pEGFP-N1 cut with XhoI and BamHI). Digested DNA was again run on a 1% agarose gel and purified as before. The digested fragments were then ligated into the appropriate digested vectors. Ligation was carried out at 16°C for 16 h using T4 DNA ligase (New England Biolabs), according to the manufacturer’s instructions, using a DNA to vector ratio of 4:1.

2.2.3 Plasmid transformation

The ligated vectors containing the PCR amplicon of interest were transformed into DH5α Chemically Competent E.coli (Invitrogen). 50ng plasmid was mixed with 25µl DH5α bacterial cells and incubated on ice for 30 min. The bacteria/DNA mix was then

58 heat shocked at 42°C for 45 sec and returned to ice, before the addition of 250µl SOC medium (Invitrogen) and incubation at 37°C for 1 h, shaking at 200 revolutions per minute (rpm). The transformation mixes were then spread onto LB agar plates (University of Manchester media stores) supplemented with 100µg/ml Ampicillin for selection of clones transformed with the pcDNA3 vector backbone and 100µg/ml Kanamycin for the selection of clones transformed with the pEGFP-N1 vector backbone. The plates were briefly allowed to dry before incubation upside down at 37°C for 16 h in a humidified atmosphere.

2.2.4 Plasmid amplification and purification

Potential colonies were selected and grown for 16 h in 2ml of LB medium (University of Manchester media stores), supplemented with 100µg/ml ampicillin or kanamycin. Aliquots were taken from each bacterial colony expansion, lysed and plasmids extracted using QIAgen mini prep kit, following the manufacturer’s protocol. The rest of the bacteria were stored at 4°C. Samples of extracted plasmid were digested with the appropriate restriction enzymes and run on agarose gels as described previously (section 2.2.2). Plasmids showing a fully digested band of the predicted size were then sequenced (section 2.2.5).

2.2.5 Sequencing and plasmid purification

Potential plasmids containing the inserts of interest were checked by automated DNA sequencing (University of Manchester Sequencing Department). First, the region of choice was PCR amplified using the BigDye terminator v3.1 cycle sequencing kit (Applied Biosystems) according to the manufacturer’s instructions, using the T7 forward primer (5’-TAATACGACTCACTATAGGG-3’) and for constructs using the pcDNA3 vector backbone, and the primers indicated in Table 2.1 for pEGFP-N1, as primers specific to this vector were not available. 100ng plasmid was added to 4µl of BigDye and 1µl of 10x reaction buffer, together with 10µM forward or reverse primer and made up to 10µl with nuclease free H2O. This reaction mix was then placed in a thermal cycler at 96°C for 4 min followed by 30 cycles of 98°C for 30 sec, 50°C for 15 sec and 60°C for 4 min. The amplification reaction was precipitated by adding 1µl 3M sodium acetate pH 5.2, 1µl 10mg/ml linear polyacrylamide (LPA) (both Sigma, UK), 0.5µl 15 mg/ml glycoblue (Ambion, UK) and 25µl (2.5 volumes) 100% ethanol, to the

59 reaction. The resulting mix was incubated at RT for 20 min then centrifuged at 13000xg for 10 min. The supernatant was discarded and the pellet air dried, before being sent to the in-house sequencing facility. Providing the sequence was correct, the respective bacteria were grown-up from the stocks at 4°C in 2ml LB media supplemented with amplicillin or kanamycin, for 8h at 37°C shaking at 200rpm. This starter culture was then transferred to a larger 50ml supplemented LB culture using a 1 in 1000 dilution (50µl bacteria in 50ml LB) and incubated for 16h at 37°C, shaking at 200rpm. The resulting bacteria were lysed and plasmids extracted using QIAgen midi prep kit, following the manufacturer’s instructions. Nucleotide concentration (ng/µl) and purity (ideal ratio for 260:280nm peak absorbance is 1.8 for DNA) was determined using a NanoDrop 1000 spectrophotometer (Thermo Scientific). Plasmids were stored at - 20°C, and aliquots taken from each bacterial colony were kept for storage at -80°C as a glycerol stock, using a 50:50 mix of bacteria:40% glycerol.

2.3 Cell Culture

2.3.1 Maintenance of cell lines

All cell lines used for work reported in this thesis were maintained at 37°C in a humidified 5% CO2 environment. Human Embryonic Kidney (HEK293) cells (a kind gift from Dr Jian Li, University of Manchester) were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM), containing 4.5g/litre glucose, sodium pyruvate, sodium bicarbonate, Ca2+ and Mg2+ (Gibco), and supplemented with 10% fetal bovine serum (FBS), 0.05mg/ml streptomycin, and 50 Units of penicillin. The murine 3T3-L1 cell line (ATCC, Catalogue Number CL-173) were maintained in DMEM, containing 4.5g/litre glucose, L-glutamine, sodium bicarbonate and pyroxidine HCL (Gibco), and supplemented with 10% bovine calf serum and 0.05mg/ml streptomycin and 50 Units of penicillin. To ensure a healthy cell population, cell lines were passaged every 7 days, by detaching with a 0.25% trypsin, 0.02% ethylenediaminetetraacetic acid (EDTA) solution (Sigma, UK) (1ml to detach cells in a 75cm2 dish), before quenching this reaction with the addition of fresh supplemented DMEM, and re-aliquoting cells at an appropriate density.

60

2.3.2 Transient transfections

All transfections were performed using Fugene6 (Roche), according to the manufacturer’s instructions. HEK293 cells were cultured in a sterile 6-well plate until 70% confluent (approximately 7x105 cells). A transfection mix of 97µl serum and antibiotic free DMEM and 3µl Fugene6 was incubated at RT for 5 min, before addition of 1µg of plasmid DNA. After incubation of the fugene-DNA mix at RT for a further 20 min, it was added drop wise to the cells, which were then returned to 37°C for 48 h. For SDS-PAGE/immunoblotting or co-immunoprecipitation (co-IP) experiments, cells were washed with PBS (Sigma, UK), then lysed in 200µl lysis buffer (10mM Tris-HCL, pH7.4, 150mM NaCl, 1mM EDTA, pH8.0, 0.5% Triton-X-100, 0.5% NP-40, and protease inhibitor cocktail (Roche). Where samples were to be used to identify phosphorylated proteins, phosphatase inhibitors (Phosphatase Inhibitor Cocktail 1 and 2 (Sigma, UK)) were added to the lysis solution) for 10 min at 4°C, collected and centrifuged at 12000xg for 10 min. Supernatant was stored at -20°C, or used directly for co-IP (section 2.4.2).

2.3.3 Stable transfections

HEK293 or 3T3-L1 cells were cultured in a sterile 6-well plate until 40% confluent (approximately 4x105 cells). Cells were transfected as before, using Fugene6. HEK293 cells were transfected with pcDNA3.1/myc-His(A)-wGPR50 or pcDNA3.1/myc-His(A)- tGPR50. After 48 h, medium was replaced with selective medium containing 1000µg/ml Geneticin (Roche). Individual clones were allowed to develop for 2 weeks, with selective media changed every other day to remove dead cells and debris. Individual colonies were then isolated using cloning discs (3mm, Sigma), and transferred to 12- well plates for further propagation in selective medium. The clones were expanded in T-75 flasks, and maintained in selective medium containing 500µg/ml Geneticin. 3T3- L1 cells were transfected with pcDNA3.1/myc-His(A)-wGPR50 or mock transfected with empty vector, pcDNA3.1/myc-His(A), and subject to the same protocol as used for the creation of stable HEK293 cell lines. 3T3-L1s that were transfected with plasmid containing shRNA targeting GPR50, or the non-targeting shRNA control vector SHC002, were subject to a similar protocol, but were selected using 2µg/ml puromycin (Sigma), and maintained in 1µg/ml puromycin.

61

2.3.4 Differentiation of 3T3-L1 cells

3T3-L1 pre-adipocytes were differentiated into mature adipocytes by the following method (see Figure 2.1). Cells were seeded into tissue culture plates and allowed to grow to 100% confluency in complete media (described in section 2.3.1). Cells were kept for a further 48h in this state to arrest cell division, before treatment with adipocyte differentiation media, consisting of DMEM (section 2.3.1), supplemented with 10% fetal bovine serum, 0.05mg/ml streptomycin, 50 Units of penicillin, 1µM dexamethasone, 0.5mM 3-Isobutyl-1-methylxanthine and 10µg/ml bovine insulin (all Sigma, UK). Cells were kept for 96h in this condition before replacement of the media with adipocyte maturation media consisting of DMEM supplemented with 10% fetal bovine serum, 0.05mg/ml streptomycin, 50 Units of penicillin and 10µg/ml bovine insulin. Cells were fed maturation media every 2 days. After day 8, lipid droplets inside the cells were visible, and by day 12, large lipid droplets were visible, which was the day subsequent assays were performed, unless otherwise stated.

2.4 Protein Analysis

2.4.1 Protein quantification by bicinchoninic acid (BCA) assay

A standard curve was made using known concentrations of protein standards of bovine serum albumin (BSA) (Sigma, UK). The protein standards; 0, 100, 200, 400, 800 and 1000µg/ml BSA diluted in lysis buffer, and unknown sample diluted 1:10 in lysis buffer to make a final volume of 50µl were pipetted in triplicate in 96-well plates. Then, 200µl

BCA working reagent (BCA solution (Sigma, UK) and 4% (w/v) CuSO4 (Sigma, UK), in a ratio of 1:50) were added to the sample and standard wells before incubation of the plate at 37°C for 30 min. Samples were read at 560nm on a plate reader (Bio-Rad).

2.4.2 Co-Immunoprecipitation

HEK293 cells stably transfected with wGPR50myc or tGPR50myc (described in section 2.3.3) were seeded into 6-well cell culture dishes at 70% confluency, before transient transfection (described in section 2.3.2) with a specific HAGα expressing construct, or empty vector pcDNA3 as a control. As a further control, HEK293 cells were also transiently transfected with wGPR50myc alone or a specific HAGα subunit alone. Following cell lysis, aliquots were taken to confirm expression of the transfected

62

Figure 2.1 Schematic of 3T3-L1 differentiation. Cells are grown to 100% confluency and maintained for 48h in this state to arrest cell division, before treatment with adipocyte differentiation media, supplemented with a hormonal cocktail of dexamethasone, Isobutyl-1-methylxanthine (IBMX) and insulin. IBMX is a cAMP phosphodiesterase inhibitor and increases levels of cAMP, which activates cAMP response element binding protein (CREB) to drive expression of CCAAT-enhancer binding proteins (C/EBP) β and δ, which are early transcription factors involved in the adipocyte differentiation process. Dexamethasone activates glucocorticoid receptors (GRs) to similarly drive transcription of C/EBPβ/δ, which then drive the transcription of proliferator-activated receptor γ (PPARγ), which controls the expression of a multitude of adipogenic genes. The signalling pathways mediated by insulin have not been fully delineated, but require activation of the insulin-like growth- factor receptor (IGFR) and the downstream proteins Akt and Ras. Cells are kept for 96h in differentiation media before replacement with adipocyte maturation media supplemented with insulin only, which cells are fed every 2 days. After day 8, lipid droplets inside the cells are visible, and by day 12, large lipid droplets are visible. By the author.

63 proteins, and the rest used for immunoprecipitation. 70µl of resuspended Dynabeads (Invitrogen) per experimental point were placed on a magnet to separate the beads from the solution, and the supernatant removed. 2µg rabbit α-myc polyclonal antibody (MBL International) diluted in 200µl PBS with 0.02% Tween-20 (Sigma, UK) per experimental point was then added to the beads and incubated with rotation for 30 min at RT. Tubes were placed on a magnet and the supernatant removed, before adding total cell lysates to the Dynabeads-antibody complex and incubating with rotation for 1 h at RT. The beads were then washed three times using 200µl PBS for each wash. Beads were then resuspended in 30µl sample buffer (75mM Tris-HCL pH6.8, 15% SDS, 20% glycerol, 0.001% Bromophenol blue, and 100mM Diothrothreitol; DTT) and heated at 95°C for 5 min. Samples were then subjected to analysis by SDS-PAGE, followed by immunoblotting (Section 2.4.3-2.4.4).

2.4.3 SDS-PAGE

SDS-PAGE was carried out using Mini-PROTEAN 3 apparatus (BioRad), according to maufacturer’s instructions. Briefly, resolving gels were prepared (containing 10% (v/v) bis-acrylamide, 0.1% (v/v) SDS, 0.1% ammonium persulfate, 0.375M Tris-HCL pH 8.8, 30mM TEMED) to separate total protein along with a stacking gel layer (containing 4% (v/v) bis-acrylamide, 0.1% (v/v) SDS, 0.1% ammonium persulfate, 0.25M Tris-HCL pH 6.8, 30mM TEMED) to enhance band quality. 50µg quantified protein lysates were denatured at 95°C for 5 min in the presence of 20µl sample buffer (as in section 2.4.2). These samples (or all 30µl eluted samples from co-IP experiments) were separated on the SDS-PAGE gels by running at 150 volts until the proteins were sufficiently resolved, in a Tris-SDS running buffer (containing 10x running buffer: 192mM glycine, 25mM Tris, 0.01% SDS). A pre-stained molecular weight protein marker (Precision Dual Colour Marker, BioRad) was resolved in parallel with samples to allow estimation of the molecular sizes of the separated proteins.

2.4.4 Immunoblotting (Western blotting)

The proteins separated by SDS-PAGE were transferred to a 0.2µm nitrocellulose membrane (BioRad) for 1 h at 12 volts using a semi-dry transfer cell (BioRad), in a methanol based transfer buffer (running buffer made 200ml 100% methanol and made to 1 litre with deionised H2O). To monitor transfer efficiency and equal protein loading,

64 the membrane was stained with Ponceau S (Sigma, UK), and subsequently washed with deionised H2O. Non-specific protein binding sites were blocked by incubating the membrane for 1 h at RT in 5%(w/v) skimmed milk (Marvel) in PBS (10x PBS: 1.4M

NaCl, 25mM KCl, 0.1M Na2HPO4, 20mM KH2PO4) supplemented with 0.1% Tween-20 (PBST). Primary antibody was added, diluted in 5% milk-PBST and incubated overnight with agitation at 4°C. Antibodies used were rabbit α-myc-tag polyclonal antibody (MBL International) diluted 1 in 800, mouse (12CA5) α-HA-tag monoclonal antibody (Abcam, UK) diluted 1 in 1000, mouse (JY2) α-Txnip/VDUP1 monoclonal antibody (MBL International) diluted 1 in 1000, rabbit α-CREB polyclonal antibody (Millipore) diluted 1 in 1000, rabbit α-phospho-CREB(Ser133) polyclonal antibody (Millipore) diluted 1 in 1000 and mouse α-β-actin monoclonal antibody (Abcam) diluted 1 in 5000. The membrane was washed 3 x 10 min in PBST to remove unbound antibody and the secondary antibody added diluted in 5% milk-PBST for 1h at RT, with agitation. The secondary antibodies were horseradish peroxidise (HPR)-linked α- mouse or α-rabbit (raised in sheep and donkey, respectively) whole antibody immunoglobulin G (IgG) (GE Healthcare). Excess antibody was washed off with 3 x 10 min washes in PBST before membranes were imaged using the enhanced chemiluminescence (ECL) Amersham advanced detection system (GE Healthcare) and an automatic developing/fixing system (JPI Automatic x-ray film processor, model JP- 33) with a Biomax film (Kodak). Immunoblots were quantitated by scanning the X-ray film and analysing images using SigmaScan software and are represented as expression relative to β-actin or CREB, as indicated in the figures.

2.5 Immunocytochemistry and Immunofluorescence

HEK293 cells were cultured on a poly-L-lysine coated cover slip, approximately 13mm diameter and 0.17mm thick, in 6-well tissue culture plates in DMEM, as section 2.3.1. 48 h following transfection, as in section 2.3.2, cells were washed in PBS and fixed with 10% neutral buffered formalin solution (Sigma, UK) for 10 min at RT. Coverslips were then washed x 2 with PBS. If cells were transfected with fluorescently-tagged proteins, coverslips were allowed to air-dry and were subsequently mounted onto slides using vector shield hard-set mounting medium (Vector labs), which contained the DNA stain 4’,6-diamidino-2-phenylindole (DAPI). If cells were then subject to immunofluorescence, following the wash step, cells were permeabilised with PBS containing 0.25% Triton-X-100 (PBS-X) for 10 min RT, to allow antibody penetration. Cells were blocked for 30 min RT in PBS supplemented with 5% goat serum (host of

65 the secondary antibody), before incubation 1 h RT with primary antibody, mouse (12CA5) α-HA-tag monoclonal antibody (Abcam, UK), diluted 1 in 400 in PBS supplemented with 1% goat serum. Excess antibody was washed off with 5x 2 min washes in PBS-X before incubation of coverslips in the dark with secondary antibody, Texas-red-conjugated goat α-mouse IgG secondary antibody (1:400; Vector Labs). antibody, diluted 1 in 500 in PBS supplemented with 5% goat serum. Coverslips were again washed for 5x 2 min in PBS-X, before allowing to air dry, and were subsequently mounted onto a slide, as previously described. Images for fluorescently-tagged proteins were collected using an Olympus BX51 upright widefield microscope and a 60x objective and captured using a Coolsnap ES camera (Photometrics). Images for immunoflurescence were collected using a Nikon C1 confocal using a TE2000 PSF inverted microscope and a 60x objective. Specific filter sets for DAPI (excitation 365nm, emission 397nm); FITC (excitation 450-490nm, emission 515-565nm) and Texas red (excitation 546nm, emission 615nm) were used. Images were then processed using ImageJ software.

2.6 Flow Activated Cell Sorting (FACS)

Cells were seeded and transfected with HAwGPR50EGFP+empty vector (EV; pcDNA3), HAtGPR50EGFP+EV, HAwGPR50EGFP+tGPR50myc or EV alone, in triplicate wells, as described in section 2.3.2. 48 h after transfection, media was removed from cells and 5 ml FACS buffer (PBS + 2% FCS) was added to each well and cells collected by pipetting gently over the surface of the well. Cells were counted using a Schärfe CASY cell counting system and 1 million cells were aliquoted into FACS tubes (BD Biosciences) and centrifuged 4°C 5 min 200xg. Supernatant was removed and the cell pellet was resuspended in 50µl rabbit (Y-11) α-HA antibody (Santa Cruz Biotechnology)-conjugated to Alexa fluor 647, diluted 1:200 in FACS buffer. Cells were then incubated on ice in the dark for 30 min. Cells were centrifuged 4°C 5 min 200xg and washed with 1 ml FACS buffer. This step was repeated x 3. After the final wash and centrifugation step, cells were resuspended in 400µl FACS buffer for analysis using a FACSCalibur flow cytometer (BD Biosciences). Dual excitation, using the 488nm argon-ion laser, and 635nm red diode laser, and fluorescent detection using photomultipliers fitted with 530nm or 661nm bandpass filters, to detect GFP and Alexa 647 respectively, was used to determine the percentage of cells within the population that were expressing Alexa 647 (HA), GFP or both HA and GFP. Scatter plots were

66 analysed to determine dual HA/GFP positive cell populations using BD FACStation

Mac Pro software for flow cytometry (BD Biosciences).

2.7 Cyclic Adenosine Monophosphate (cAMP) Accumulation Assay

HEK293 cells were plated in 12-well tissue culture dishes and transfected with wGPR50myc, tGPR50myc or empty vector (pcDNA3.1/myc-His(A)) (6 wells each per treatment group). After 48 h, cells were treated with 10µM forskolin (Sigma, UK) or Dimethyl Sulfoxide (DMSO) (vehicle) for 15 min and lysed in 0.1M HCL for 15 min RT with gentle agitation. Direct cAMP measurements were performed using a direct cAMP enzyme immunoassay kit (Enzo Life Sciences) according to the manufacturer’s protocol.

2.8 Oil-Red-O Staining of Differentiated 3T3-L1 Cells

3T3-L1 cells stably expressing wGPR50myc, empty vector (pcDNA3.1/myc-His(A)), plasmid containing shRNA targeting GPR50, or the non-targeting shRNA control vector SHC002, described in section 2.3.3, were seeded in 35-mm cell culture dishes and subject to the adipocyte differentiation protocol described in section 2.3.4. On day 12 of differentiation, cells were washed with PBS and fixed with 10% neutral buffered formalin solution (Sigma, UK) for 1 h RT. Cells were then washed for 2x 2 min with deionised H20, followed by incubation in 60% isopropanol 5 min RT. Isopropanol was then removed and cells allowed to dry completely at RT. Oil-Red-O stock solution (0.9mM Oil-Red-O in isopropanol, stirred overnight at RT and 0.2µ filtered) was diluted

3:2 in deionised H20, allowed to rest 20 min RT followed by filtering (0.2µ) to create Oil- Red-O working solution. Oil-Red-O working solution was added to the dried cells and incubated 10 min RT. Oil-Red-O working solution was then removed and cells were immediately washed with deionised H2O. Cells were washed x4 before images were acquired using a CoolSnap-Pro camera (Photometrics, UK) with tissue culture dishes on a light box (Jencons Scientific, UK). For quantitative Oil-Red-O analysis, cells were seeded in 6-well dishes, in triplicate, induced to differentiate and stained with Oil-Red-

O as previously described. Following the final wash with deionised H2O, all H2O was removed and cells allowed to dry completely. Oil-Red-O was then eluted by adding 1 ml 100% isopropanol and incubated at RT 10 min with gentle agitation. The isopropanol was then pipetted up and down gently to ensure all Oil-Red-O was in the

67 solution, before transfer to a suitable collection tube. OD of the eluted solution was measured in triplicate at 495nm using 100% isopropanol as blank.

2.9 Intracellular Triglyceride Content and Lipolysis of Differentiated 3T3-L1 Cells

3T3-L1 cells stably expressing wGPR50myc, empty vector (pcDNA3.1/myc-His(A)), plasmid containing shRNA targeting GPR50, or the non-targeting shRNA control vector SHC002, described in section 2.3.3, were seeded in 96-well cell culture dishes and subject to the adipocyte differentiation protocol described in section 2.3.4. On day 12 of differentiation, triglyceride levels were measured by colorimetric detection of glycerol produced by triglyceride hydrolysis, using the Triglyceride Assay Kit (TG-1-NC; Zen- Bio), according to manufacturer’s instructions. Lipolysis rates, again using day 12 differentiated cells in 96-well cell culture dishes, were determined by detection of free fatty acids and free glycerol release, using the Lipolysis Assay Kit (LIP-3; Zen-Bio), according to manufacturer’s instructions. Cells were exposed to 10µM forskolin or DMSO for 3 h before measurement of lipolysis.

2.10 RNA Extraction

2.10.1 Tissue

Tissues were dissected, after isolation from mice (section 2.14.5), and immediately frozen and stored at -80°C. Tissue was homogenised using lysing matrix D tubes (MP Biomedicals) in 1 ml TriZol reagent (Invitrogen), using a FastPrep-24 system (MP Biomedicals) at 4°C. In the case of white adipose tissue, samples were then centrifuged at 4°C 12000xg for 10 min, to separate excess lipids into a top layer which was then removed prior to phase separation, for other tissues, phase separation was carried out immediately following homogenisation. For phase separation, samples were incubated at RT for 5 min before addition of 200µl chloroform (Sigma, UK) and thorough mixing by vigorous inversion. After incubation for 3 min at RT, samples were centrifuged at 4°C 12000xg for 15 min. The resulting solution separated into 3 different phases with total RNA in the upper aqueous phase. 500µl of the aqueous phase was removed and mixed with 500µl isopropanol and incubated 10 min at RT to precipitate the RNA, before centrifugation at 4°C 12000xg for 10 min. The isopropanol solution was removed and the resulting RNA pellet washed in 70% ethanol prepared using nuclease free H2O, and centrifuged at 4°C 12000xg for 5 min. The ethanol was

68 removed and the RNA pellet air-dried before resuspension in 50-200µl nuclease free

H2O. RNA concentration (ng/µl) and purity (ideal ratio for 260:280nm peak absorbance is 2.0 for RNA) was determined using a NanoDrop 1000 spectrophotometer (Thermo Scientific). RNA samples were stored at -80°C.

2.10.2 Cells

RNA was extracted from 3T3-L1 cells using the RNeasy extraction kit (QIAgen) according to the manufacturer’s instructions, and finally eluted into sterile RNase free tubes in 30µl of nuclease free H2O, before quantification as in section 2.10.1 and storage at -80°C.

2.11 Reverse Transcription (RNA to cDNA)

2µg total RNA was used per 20µl reaction, with the assumption that all RNA was converted to cDNA, to yield a final concentration of 100ng/µl cDNA. Prior to conversion to cDNA, RNA was treated for 30 min at 37°C with 1µl RNase free DNase (Promega). The reaction was stopped with 1µl DNase stop buffer (Promega), incubated at 65°C for 10 min. A High Capacity RNA to cDNA Kit (Applied Biosystems) was then used for the conversion reaction, with the addition of 10µl 2x buffer and 1µl of 20x enzyme mix and deionised H20 so that the total volume came to 20µl. The reverse transcription reaction was incubated at 37°C for 60 min before the reaction stopped by denaturing the enzyme with incubation at 95°C for 5 min. Resulting cDNA was stored at -20°C.

2.12 RT-PCR of Gpr50 and Txnip in mouse tissues

Using the same PCR mastermix described in Table 2.2, section 2.2.1, but using 100ng cDNA instead of plasmid DNA and the primers listed in Table 2.3, PCR reactions to detect the presence of Gpr50 and Txnip in mouse tissues was carried out in a thermal cycler using the following protocol: 94°C for 2 min; then 35 cycles of 94°C for 15 sec, 55°C for 30 sec and 72°C for 1 min 30 sec. For glyceraldehyde-3-phosphate, Gapdh control, the PCR protocol was as follows: 94°C for 2 min; then 35 cycles of 94°C for 15 sec, 52°C for 30 sec and 72°C for 30 sec. The RT-PCR products were run on 1% ethidium bromide agarose gels (section 2.2.2) and imaged under U.V. light.

69

PCR Product Primer Sequence (5’ – 3’)

GPR50 F ATGGCCACGGTCCCCAAGAG

1776 bp R TCACACAGCCATCTCATCAGAGCAA

Txnip F ATGGTGATGTTCAAGAAGATC

1191 bp R TCACTGCACGTTGTTGTTG

GAPDH F CCTTCATTGACCTCAACTAC

594 bp R GGAAGGCCATGCCAGTGAGC

Table 2.3 Primers used for RT-PCR murine tissue profiling. The GPR50, Txnip and GAPDH primers used to amplify the mouse transcripts are shown. F and R is the forward and reverse primer respectively, and the amplified length of the product in base pairs (bp) is listed.

2.13 Quantitative PCR (qPCR)

SYBR Green based qPCR was used to measure levels of transcripts in tissues or 3T3- L1 cells. SYBR Green is a commonly used fluorescent DNA binding dye, binding all double-stranded DNA, which causes the SYBR to fluoresce, and detection is monitored by measuring the increase in fluorescence throughout the PCR cycle. Primers used for qPCR are shown in Table 2.4. The primers were all taken from the literature and synthesised by MWG or purchased as QuantiTect Primer assays from QIAGEN. Pilot experiments were performed to ensure specificity of the primers to produce a single product, assessed by the addition of a melting curve analysis at the end of each PCR assay, which resulted in production of a single melting peak. The efficiency of a given qPCR primer set amplification was also assessed prior to experimental use as being close to 100% and equal to that of the reference gene. This was done by serially diluting a pilot cDNA sample so that 100ng, 50ng, 25ng, 5ng, 2.5ng and 0.5ng cDNA would be included in PCR reactions with each primer set and the Ct value at each dilution was measured. A straight line was then constructed for each primer set by plotting the Ct value against log(cDNA concentration), to determine efficiencies. Generation of a slope with a gradient of between -3.1 and -3.6 indicates reaction efficiency between 90 and 110%, and if the slope generated by reactions using a particular primer set fell within this range, they were considered acceptable for use.

70

Gene Primer Sequence (5’ – 3’) Reference

18S rRNA F TCCGACCATAAACGATGCCGACT (Kitahara et al., R TCCTGGTGGTGCCCTTCCGTCAAT 2005) Pparg F AGGCCGAGAAGGAGAAGCTGTTG (Watanabe et al., R TGGCCACCTCTTTGCTCTGCTC 2011) Fabp4 F GAAAACGAGATGGTGACAAGC (van Beekum et al., R TTGTGGAAGTCACGCCTTT 2008) Txnip F CATGAGGCCTGGAAACAAAT (Forrester et al., R ACTGGTGCCATTAGGTCAGG 2009) Per1 F TGGCTCAAGTGGCAATGAGTC (Tanioka et al., R GGCTCGAGCTGACTGTTCACT 2009) Rev-erb alpha F AGCTCAACTCCCTGGCACTTAC (Tanioka et al., R CTTCTCGGAATGCATGTTGTTC 2009) AdipoR1 F ACGTTGGAGAGTCATCCCGTAT (Yamauchi et al., R CTCTGTGTGGATGCGGAAGAT 2003) Fabp7 F CTCTGGGCGTGGGCTTT (Gerstner et al., R TTCCTGACTGATAATCACAGTTGGTT 2008) Dio2 F CAGTGTGGTGCACGTCTCCAATC (Sun et al., 2011) R TGAACCAAAGTTGACCACCAG Ucp1 F GGCCAGGCTTCCAGTACCATTAG (Kitahara et al., R GTTTCCGAGAGAGGCAGGTGTTTC 2005) Gpr50 Mm_Gpr50_1_SG (Qiagen) Fas Mm_Fasn_1_SG (Qiagen) Hsl Mm_Lipe_1_SG (Qiagen) Atgl Mm_Pnpla2_1_SG (Qiagen) Lpl Mm_Lpl_2_SG (Qiagen) Hprt1 Mm_Hprt_1_SG (Qiagen) Gabrq Mm_Gabrq_1_SG (Qiagen) Osbpl11 Mm_Osbpl11_1_SG (Qiagen) Table 2.4 Primers used for qPCR. Primers used to amplify mouse transcripts are shown, F and R are forward and reverse primers, respectively.

Individual qPCR reactions were carried out in 25µl volumes in a 96-well Fast-prep plate (Applied Biosystems), using a SYBR Green PCR Master Mix (QIAGEN) and 100ng cDNA, as described in Table 2.5. qPCR amplification and detection reactions were performed using an Applied Biosystems 7500 Real-Time PCR system (Applied Biosystems), using the following protocol: 2 min at 50°C, followed by 15 min at 95°C, then 40 cycles of 15 sec at 94°C, 30 sec at 60°C and 30 sec at 72°C. Data collection was performed during the final 72°C extension step. Each sample was run in triplicate, and the data analysed using the 2- CT method (Livak and Schmittgen, 2001). Mouse

71 housekeeping gene 18S rRNA was used as an internal control, and all transcript abundances are expressed relative to this control.

Reagent Final Concentration Per Reaction

2x QuantiTect SYBR Green Master Mix 1x Either Forward Primer (10µM) 0.2µM Reverse Primer (10µM) 0.2µM 10x QuantiTect Primer Assay 1x cDNA 100ng Uracil-N-glycosylase 0.5 Units Made up to 25µl with nuclease free H2O Table 2.5 qPCR mastermix.

2.14 Animal maintenance

All experimental procedures were within the guidelines of the Animals (Scientific Procedures) Act, 1986, and received ethical approval from the University of Manchester animal welfare committee. Gpr50-/- and congenic WT mice were obtained via collaboration with AstraZeneca (Alderley Park, Cheshire, UK) who obtained the mice from DeltaGen (California, USA). Mice were partially introgressed onto a C57/B6 background and maintained as a colony at the University of Manchester. A second colony of Gpr50-/- mice on a C57/B6 background was obtained from Organon Pharmaceuticals (Motherwell, UK), although their use was limited to studies shown in Figures 5.5 and 5.7, and all other studies used the DeltaGen Gpr50-/- mice. Adult male mice were used for all experiments, apart from in studies shown in Figures 6.8, 6.9 and 6.10, which used adult female mice. All mice were maintained in a 12 h light 12 h dark lighting schedule and housed at an ambient temperature of 20-22°C, apart from those in the studies shown in Figures 6.9 and 6.10, which were housed at 16°C. Standard rodent chow and water were supplied ad libitum, except during fasting periods, in which all food was removed (for experiments requiring the induction of torpor in Gpr50-/- mice, WT and Gpr50-/- mice were typically fasted for approximately 16 h before sacrifice; for studies shown in Figures 6.9 and 6.10, WT female mice were fasted for 48 h). For all studies, tissue was collected at Zeitgeber time (ZT): 2 (selected because fasted Gpr50- /- mice are in deep torpor at this time), apart from in the studies shown in Figures 6.6 and 6.7, where tissue was collected at ZT: 8.

72

The Siberian hamster brain tissue sections used for in situ hybridisation of Txnip were kindly provided by Dr Ben Saer (University of Manchester). Briefly, adult male Siberian hamsters (Phodopus sungorus), wild-trapped from Siberia, were subsequently bred at the Veterinary University of Hannover, Germany, in the laboratory of Professor Stephan Steinlechner. Winter phenotype hamsters were obtained by housing animals outdoors, with cages exposed to natural short day (SD) photoperiod and temperature (approximately 8 h light, ambient temperature 4.9±0.4°C). A parallel set of hamsters were maintained indoors under an artificial long day (LD) light cycle (16 h light: 8 h dark) and ambient temperature (ca 18°C). All animals were individually housed and given ad libitum access to food and water. Torpid hamsters were identified by their curled up posture and inactivity. Upon sacrifice by cervical dislocation (ZT: 5), Tb was recorded and collected and frozen in RNase-free aluminium foil on dry-ice, and stored at -80°C, before cryosectioning as described in section 2.16.1.

2.14.1 Indirect calorimetry

To assess metabolic rate, age-matched Gpr50-/- and WT mice were placed in indirect calorimetric cages (Columbus Instruments, Columbus, OH, USA). Mice were acclimated to the cages for two days, following which oxygen consumption (VO2) and carbon dioxide production (VCO2) were measured every 10 min until the termination of the experiment. Standard rodent chow and water were supplied ad libitum unless where stated that mice were fasted, and all food was removed.

2.14.2 Cold Chamber

Mice used in studies shown in Figure 6.6 and Appendix F, were individually housed in a modified upright 1200 L industrial fridge (Polar Refrigeration, UK). The fridge was adapted to include an Akor ventilation unit (Akor Systems, Fance) to allow an internal- external air exchange to meet home office requirements. Internal temperature was maintained at 15°C for the study in Appendix F, or for the study in Figure 6.6, animals were exposed to 4°C for 2 h. Animals were given ad libitum access to food and water, apart from during periods of fasting, when all food was removed. Prior to being housed at 15°C, mice used in this study were implanted in the peritoneal cavity with remote telemetry probes (DataScience International, The Netherlands), to allow continuous monitoring of core-body temperature. Mice were allowed to recover for 10 days post-

73 surgery before use in the experiment. Radiofrequency signals from the implant transmitters were averaged every 5 min by receivers placed under each animal’s cage.

2.14.3 2-deoxyglucose (2-DG) administration

All food was removed from cages before 2-DG (Sigma, UK) (1500 mg/kg) or saline vehicle was administered intraperitoneally to WT mice for 1 h. Body temperature was measured using a rectal thermometer at the time of sacrifice.

2.14.4 Tissue collection

For all studies, animals were killed by cervical dislocation, tissues removed rapidly, and either frozen whole on dry ice on RNase-free aluminium foil (for in situ hybridisation (section 2.16)), placed directly in fixative (for immunohistochemistry (section 2.17)) or immediately dissected, using a sterile blade, according to tissue type (for RNA (section 2.10.1) and protein extraction (section 2.18)) before being frozen on dry ice, and stored at -80°C. For larger tissues, approximate 100mg blocks were typically dissected, apart from in the case of epidydimal white adipose tissue, where all was taken due to low RNA yield, and interscapular brown adipose tissue, where surrounding connective tissue and white adipose tissue was removed before the two lobes were taken. For studies used in Figure 6.3 and 6.5, in the case of smaller tissues, for example, the eyes, adrenal glands and pituitary, whole tissue was included in the RNA extractions. Generally, for dissection of hypothalamic blocks for RNA or protein analysis, tissue was dissected immediately rostral to the SCN and caudal to the DMH, and blocks trimmed to remove tissue dorsal to the hypothalamus and laterally adjacent to the optic tract (see Figure 2.2). Hypothalamic collection for the microarray study (Chapter 5) differed slightly in an attempt to control the area of tissue included in this study. Whole brains were cut coronally in rostral-caudally using a vibratome and a stainless steel blade (both Campden Instruments) up until the level of the SCN. A 1.5mm slice was then cut, so that the whole hypothalamus would be included in the slice, up until the caudal level of the DMH. Again, blocks were trimmed to remove tissue dorsal to the hypothalamus and laterally adjacent to the optic tract, and the hypothalamic block immediately frozen and stored at -80°C.

74

Figure 2.2 Collection of hypothalamic blocks for quantitative PCR and protein analysis. Hypothalamic blocks from WT and Gpr50-/- mice were dissected (red dotted lines) immediately rostral to the SCN and caudal to the DMH. Blocks were trimmed to remove tissue dorsal to the hypothalamus and laterally adjacent to the optic tract. Diagram by the author.

75

2.15 Microarray

To measure gene transcript changes in the hypothalamus of ad libitum fed and fasted WT and Gpr50-/- mice (n=4/group), hypothalamic tissue blocks were collected as described in section 2.14 and total RNA was extracted as in section 2.10.1, and eluted in 50µl of nuclease free H2O. Total RNA was then sent to an in-house microarray department (Genomic Technologies Core Facility, University of Manchester), where RNA quality and concentration were assessed before conversion to cDNA and subsequent analysis of transcript expression using Affymetrix mouse exon arrays (Affymetrix). The data generated was then sent for bioinformatic analysis in a further in- house department (Bioinformatics Core Facility, University of Manchester). Transcripts were considered to be significantly altered if they had a fold change >1 and statistical significance of q<0.05 (q is a false discovery rate-adjusted P-value). Functional clustering analysis was carried out using The Database for Annotation, Visualization and Integrated Discovery (DAVID) version 6.7 resource. Significantly altered transcripts (q<0.05) between genotypes with fasting were analysed for GO and KEGG pathway functional enrichment. Groupings with fold enrichment >2 , n>3 and significance p<0.01 only were included in the resulting gene lists.

2.16 In Situ Hybridisation

In situ hybridisation was used to visualise brain regions and quantify mRNA expression of Per1, Per2, Fabp7 and Txnip in WT and Gpr50-/- mice (Txnip also analysed in Siberian hamster brain).

2.16.1 Cryosectioning and preparation of tissue sections

Brains collected as described in section 2.15.5, were sectioned (12µm) using a Leica CM3050 S cryostat freezing microtome (Leica Microsystems, UK). The cryostat chamber was set at -20°C and cleaned with 70% ethanol, and the cryostat blade, brushes and chuck were RNase treated and equilibrated in the chamber in the chamber. Brains were mounted using OCT compound medium (Leica Microsystems, UK) and left to attain a specimen temperature of -18°C, Rostral to caudal coronal sections of the entire hypothalamus were taken from each animal starting from the emergence of the SCN to the end of the DMH, and dried onto polysine slides (VWR International, UK), which were stored at -80°C. To preserve tissue morphology and

76 mRNA, mounted cryosections were fixed in 4% paraformaldehyde (PFA), prepared as follows: 12g PFA (Sigma, UK) was dissolved in 300ml PBS, heated to 50°C then 2-3 drops of 5M NaOH, pH 5.2, added to clarify the solution. The resulting PFA was then cooled and filtered through 0.45µm filter paper (Whatman). Sections were then washed twice in PBS (Sigma, UK) for 5 min. Triethanolamine (TEA; Sigma, UK) and acetic anhydride (AA; Sigma, UK) was used to acetylate sections, which prevents non- specific binding and minimises background signal. The TEA/AA solution was prepared fresh for each batch of slides using 200ml 0.9% NaCl, 3ml TEA and 520µl AA. Slides were treated for 10 min, on a magnetic stirrer. Slides were washed in PBS for 5 min before dehydration in increasing grades of ethanol (70%, 80% and 90% 5 min, then 95% and 100% 10 min). Finally, the slides were placed into chloroform (Fisher Scientific) for 5 min and left to air-dry.

2.16.2 Preparation of radiolabelled probes

For in situ hybridisation labelling, mouse or Siberian hamster riboprobes were amplified by PCR using the same mastermix described in Table 2.2, section 2.2.1, but using 100ng mouse or Siberian hamster (kind gift from Ben Saer, University of Manchester) hypothalamic cDNA, and the primers listed in Table 2.6. PCR reactions were carried out in a thermal cycler using the following protocol: 94°C for 2 min; then 35 cycles of 94°C for 20 sec, 53°C for 30 sec and 72°C for 30 sec.

Gene Primer Sequence (5’ – 3’)

Txnip F GCCGCTCGAGCATGAGGCCTGGAAACAAAT (mouse/Sib.hamster) R CGGCGAATTCGCCATTGGCAAGGTAAGTGT 406bp Fabp7 F GCCGCTCGAGGTAGATGCTTTCTGCGCAAC (mouse) R CGGCGAATTCTTTCTTTGCCATCCCACTTC 304bp Table 2.6 PCR primers used for RT-PCR to generate in situ hybridisation riboprobe templates. Primers used to amplify mouse or Siberian hamster transcripts are shown, F and R are forward and reverse primers, respectively, with restriction enzyme sites indicated in bold italics. Amplicon length in base pairs (bp) is also listed.

77

The products were run on 1% ethidium bromide gels (section 2.2.2) and imaged under U.V. light prior to gel excision and purification (section 2.2.2) and digestion of the amplicons using the restriction enzymes EcoRI and XhoI (New England Biolabs), for 3 h at 37°C. Empty vector, pcDNA3 was similarly digested with EcoRI and XhoI, before digestion products were run on 1% ethidium bromide gels to establish purity, and again excised and purified, before each riboprobe amplicon was ligated into pcDNA3 as in section 2.2.2. The plasmids were then amplified, purified and sequenced using the T7 forward primer, as in sections 2.2.3-2.2.5. Plasmids for Per1 and Per2 in situ probes were kindly provided by Dr Jian Li (University of Manchester). Plasmids were linearised with an XhoI or EcoRI single digestion to allow production of antisense or sense transcripts, respectively, using the reaction setup in Table 2.7. To ensure the plasmid was fully linearised, 5µl of each reaction was checked on a 1% agarose gel, with a non- linearised plasmid control. Linearised plasmids were purified using a phenol-chloroform based extraction to precipitate the DNA. 300µl nuclease-free H2O and 400µl phenolchloroform (Sigma, UK), were added to the plasmid reaction before centrifugation 5 min 13000xg. 370µl of the upper aqueous phase was then added to 370µl chloroform (1:1 ratio) (Sigma, UK) and samples centrifuged 5 min 13000xg. The upper aqueous phase was then mixed with 10% volume 3M sodium acetate (Sigma, UK), 1µl 20mg/ml glycogen and 2x volume 100% ethanol (2 volumes). DNA was then precipitated overnight at -20°C before samples were centrifuged at 4°C 45 min 13000xg. Resulting pellets were washed with 150µl 70% ethanol (prepared with nuclease free H2O) and centrifuged at 4°C 5 min 13000xg. Supernatant was removed and pellets allowed to air-dry before being dissolved in 15µl nuclease-free H2O. DNA concentration was measured using a Nanodrop, as in section 2.2.5

Reagent Final Concentration

Plasmid DNA 20µg 10x buffer 1x Enzyme 100 Units (EcoRI for sense, XhoI for antisense) Made up to 100µl with nuclease free H2O Table 2.7 Mastermix for plasmid linearisation.

78

2.16.3 Synthesis and purification of 33Pα-UTP labelled probe

All radioactive work followed the rules and regulations associated with ionising radiation, including protection and waste disposal. For radioactive in situ hybridisation, antisense and sense riboprobes were synthesised using T7 or SP6 (Promega, UK) RNA Polymerase enzymes, respectively, in reactions setup as in Table 2.8, in the presence of 33P-uridine 5’-triphosphate (UTP; MP Biomedical USA) at 37°C for 45 min. After incubation, 3µl RQ1 DNase (Promega, UK) was added and incubated for 10 min to degrade the DNA template. The radiolabelled probes were purified using illustra ProbeQuant G-50 micro columns (GE Healthcare, UK). Columns remove unincorporated 33P-UTP, which might result in unspecific binding. The column gel matrix was resuspended and centrifuged at 735xg 1 min to remove the equilibration buffer and re-establish the matrix bed. Buffer was discarded and the column placed into a new microfuge tube. To each of the riboprobes, 27µl nuclease-free H2O was added and samples were loaded into individual columns, before centrifugation 2 min 735xg. Elute was retained and 50µl deionised formamide 99.5% (Sigma, UK) added to each sample and denatured on a 65°C heat block for 5 min, then quenched on ice for 2 min. 1µl of each probe was then added to 4ml scintillation fluid (National Diagnostics, UK) and checked on a TRI-CARB 2100CA Liquid Scintillation Analyser to measure the radiation count. To ensure even coverage of probe per slide, a measure of 15x105 counts per minute (cpm) was mixed in 100µl of hybridisation buffer (pre-heated to 60°C).

Reagent Final Concentration

Linearised Plasmid DNA 1µg

5x buffer 1x

DTT 10mM rNTP (10mM rATP, rCTP, rGTP, rUT33P) 500µM each

RNase inhibitor 30 Units

Transcription enzyme Sense (SP6) 37.5 Units

Anti-sense (T7) 22.5 Units

Made up to 20µl with nuclease free H2O

Table 2.8 Mastermix for riboprobe in vitro transcription. 79

2.16.4 Hybridisation

100µl probe/hybridisation buffer was dispensed over each slide and covered with a hybrid-slip (Sigma, UK), and incubated overnight at 60°C in a humid chamber containing moistened strips of filter paper. Hybrid-slips were removed and slides washed in a cradle containing 2x SSC buffer (0.3M NaCl, 30µM sodium citrate; Promega) with 50% formamide (Sigma, UK) at RT 10 min. The solution was replaced with pre-heated 60°C 2xSSC/50% formamide and incubated for 2x 30 min. Slides were then washed in 2xSSC/50% formamide at 60°C for 2x 15 min, followed by 0.5xSSC at 60°C for 30 min. Finally, slides were placed in increasing grades of ethanol (70%, 95% and 100%, 5 min each). Slides were fixed into development cassettes using masking tape and covered with Kodak Biomax MR Film (Sigma, UK). Cassettes were left at - 80°C for 1 week before development using a Compact Automatic Film Processor (Xograph Healthcare, UK).

2.16.5 Analysis of in situ hybridisation

Films were scanned using a CoolSnap camera (Photometrics, UK) while on a light box (Jencons Scientific, UK). SigmaScan software (Sigma, UK) was used to analyse the signal intensity from highlighted areas of the scanned images to obtain the mean optical density (mOD). The relative optical density (ROD) was then calculated by subtracting the mOD of a non-specific area of the brain section from the area being quantified. RODs were obtained for 5-10 sections per animal, before a mean ROD value was produced for each area of interest.

2.17 Immunhistochemistry (IHC)

2.17.1 Tissue fixing and processing

Intact tissue intended for IHC using antibody directed against Txnip was removed from the animal and placed directly into 4% PFA. Tissue for IHC using antibody directed against GPR50 was fixed in Bouin’s solution (Sigma, UK). Once in fixative, tissue was left for a minimum of 48 h at 4°C, followed by 72 h in 20% sucrose solution before freezing. After fixation, white adipose tissue was embedded in paraffin wax using a Microm spin STP-120 processor (UK), according to the manufacturer’s automated protocol. Tissue was cut in continuous 5µm sections using a Leica RM2155 rotary

80 microtome, and transferred to a water bath, where they were mounted onto slides and left to dry. Slides were stored at RT until stained. Fixed brains were cut into 20µm sections on a freezing microtome (Leica) using a non-toxic histological mounting solution to fix the sections to the base. Sections were cut to include slices of the hypothalamus, ranging from the level of the SCN to the DMH. Slices were then placed in PBS and stored at 4°C until stained.

2.17.2 Immunohistochemical staining of tissue

Paraffin embedded sections were de-waxed in a xylene bath for 4x 5 min and rehydrated in decreasing ethanol washed (100%, 95%, 50% and 20%) then placed in tap H2O. An antigen retrieval step was then carried out, where sections were covered in fresh citrate buffer (11mM sodium citrate with HCl added to pH 6, Sigma, UK) and placed in the microwave, at approximately 900W, for 12 min and allowed to cool. These and the free-floating sections were then subject to the following antibody staining protocol. Sections were washed with PBS before permeabilisation by incubation for 1 h RT in PBS supplemented with 0.1% Triton-X-100 (PBS-X), with gentle agitation. Sections were then blocked for 1 h at RT in PBS supplemented with 10% donkey serum (serum from host species of secondary antibody). Primary antibody was diluted in PBS supplemented with 1% serum; primary antibody used were mouse (JY2) α-Txnip/VDUP1 monoclonal antibody (MBL International) diluted 1 in 200, and goat α-GPR50 (Santa Cruz) diluted 1 in 200, and incubated sections for 16 h with gentle agitation. Excess primary antibody was removed with 3x 10 min washes in PBS supplemented with 0.05% Tween-20 (PBS-T). Tissue was then incubated for 1 h at RT with FITC-conjugated donkey IgG secondary antibody (Jackson ImmunoLabs, Suffolk, UK) diluted 1 in 400 in PBS supplemented with 5% serum. Sections were then washed

3x 10 min in PBS-T, washed briefly in deionised H2O, and mounted onto slides (free- floating sections), and coverslipped (using vector shield hard set mounting medium with DAPI). Where appropriate tissue sections were counterstained with cresyl violet by dehydrating sections in 70% ethanol before incubating in 0.5% Cresyl Fast Violet

(Sigma) and 0.01% acetic acid in deionised H20. Sections were then dehydrated in increasing grades of ethanol (70%, 95% and 100%) including an acetic acid-ethanol soak to remove excess stain from the tissue. Unstained areas were made transparent using a clearing agent (Histoclear, Fisher Scientific); before sections were coverslipped using Depex Mounting Medium (Fluka) and images collected using a Leica DM2000 upright light microscope and a Leica DFC295 digital camera. Fluorescent images were

81 collected using an Olympus BX15 upright widefield microscope and a 20x objective and captured using a Coolsnap ES camera (Photometrics). Specific filter sets for DAPI (excitation 365nm, emission 397nm) and FITC (excitation 450-490nm, emission 515- 565nm) were used and images were processed using ImageJ software.

2.18 Protein Extraction from Tissue

To analyse Txnip protein levels, hypothalamic tissue blocks and brown adipose tissue were collected as described in section 2.14.5, immediately frozen and stored at -80°C. Tissue was homogenised using lysing matrix D tubes (MP biomedicals) in 500µl Tissue Protein Extraction Reagent (T-Per; Thermo Scientific) and a FastPrep-24 system (MP Biomedicals. Samples were then centrifuged at 4°C 10 min 12000g and supernatant collected. Protein concentration was quantified as in section 2.4.1, and 80µg protein loaded onto SDS-PAGE gels and subject to immunoblotting as in sections 2.4.3 and 2.4.4.

2.19 Statistical Analysis

Data are presented as mean ± standard error (SEM). Statistical data were analysed using GraphPad Prism 4 software using unpaired student’s t-test, one-way analysis of variation (ANOVA), or two-way ANOVA with a Bonferroni’s post hoc test. The specific test used is specified in each section.

82

CHAPTER 3

GPR50 SIGNALLING

83

3.1 INTRODUCTION

Little is known about the intracellular signalling pathways downstream of GPR50, though it has been suggested that this receptor may participate in G-protein independent signalling (Levoye et al., 2006; Reppert et al., 1996). The C-terminal tail of GPR50 is one of the longest among mammalian GPCRs and analysis shows it to be proline-rich and containing multiple serine and threonine residues that may be potential targets of phosphorylation. By these means, it is possible that the intracellular tail of GPR50 acts as a scaffold for recruitment of intracellular proteins, a common feature of GPCRs (Tobin et al., 2008). Additionally, the identification of a consensus site for a serine protease within the C-terminal tail of GPR50 raises the interesting possibility that it may be a target for proteolysis, with cleaved portions of the receptor having some intracellular function (Reppert et al., 1996).

GPR50 has been shown to heterodimerise with the MT1 receptor in vitro and antagonise its melatonin-dependent signalling (Levoye et al., 2006). Intriguingly, when the C-terminal tail of GPR50 was deleted, the mutant receptor no longer elicited an inhibitory effect on MT1. Therefore it is possible that in these heterodimers, the long C- terminal tail of GPR50 hinders the recruitment of intracellular interacting partners to

MT1, such as Gi proteins, thereby inhibiting MT1 signalling. Impaired Gi recruitment could also be responsible for the apparent decline in agonist binding to MT1 receptors when engaged in heterodimers with GPR50, as high-affinity agonist binding to MT1 has been shown to rely on the presence of Gi (Barrett et al., 1994).

Despite the observed role for GPR50 in the modulation of MT1 signalling in cell lines, in vivo studies reveal that regions of GPR50 expression such as the ependymal layer of cells lining the third ventricle do not demonstrate melatonin receptor expression. This suggests that GPR50 has further functions such as ligand-dependent signalling, which remain to be discovered. This chapter describes initial experiments to identify potential intracellular signalling cascades downstream of GPR50. As this receptor shares 45% sequence identity with MT1 and MT2, and the best known signalling pathway mediated by the melatonin receptors is inhibition of cAMP formation via pertussis-toxin sensitive G proteins, I focused my attentions on this signalling cascade to assess if GPR50 can activate a similar pathway.

84

3.2 RESULTS

3.2.1 Identification of a novel splice variant of GPR50

Cloning of full length Gpr50 from murine hypothalamus and pituitary by RT-PCR using primers specific to Gpr50, identified two transcripts in each tissue (Figure 3.1, A). DNA sequencing confirmed the larger transcript to be full length Gpr50, while the lower band represented a smaller receptor cDNA, displaying a 139- deletion between nucleotide positions 720 and 860, in exon 2 of the gene (Figure 3.1, B and C). The deletion causes a frameshift in the open reading frame, generating new coding sequence for 3 further amino acids before the introduction of a premature stop codon. Examination of Gpr50’s sequence reveals a canonical splice donor (5’-GT-3’) and splice acceptor (5’-AG-3’) site, immediately after nucleotide position 720 and prior to nucleotide 860, respectively (Figure 3.1, B). This indicates that alternative splicing of Gpr50 mRNA generates a novel splice variant of Gpr50, which results in a truncated form of the GPR50 protein (herein referred to as tGPR50), lacking the last two transmembrane domains and the intracellular tail (Figure 3.1, D).

3.2.2 Expression of whole-length GPR50 and tGPR50 in mammalian cells

Western blot analysis of total cell lysates from HEK293 cells, transiently transfected with C-terminally myc-tagged full-length GPR50- or tGPR50-containing plasmids, demonstrated expression of both forms (Figure 3.2). Immunoreactive protein migrating at ~69 kDa, the expected size of GPR50, was detected, as well as protein at ~29 kDa representing tGPR50. Interestingly, immunoblotting of full-length GPR50 also revealed a lower molecular weight band at ~36 kDa, which likely reflects a proteolytic cleavage product of the C-terminal domain. Further western blot analysis using an N-terminally HA-tagged GPR50 construct (Figure 3.2), identified an N-terminal product migrating at ~39 kDa, which could be the reciprocal fragment to the C-terminal domain product.

3.2.3 Differential localisation of whole-length GPR50 and tGPR50 in mammalian cells

Using fluorescently-tagged proteins, the subcellular localisations of GPR50 and tGPR50 were examined in HEK293 cells. Full-length GPR50EGFP localised primarily to

85

Figure 3.1 Identification and cloning of full length and truncated GPR50. A. RT-PCR cloning of Gpr50 from murine hypothalamus and pituitary produced two bands reflecting full length Gpr50 (~1.8kb; wGPR50) and a novel truncated transcript (~1.7kb; tGPR50). B. The nucleotide sequence for murine wGPR50, with annotated transmembrane domains. The sequence shown in blue corresponds to that lost from tGPR50, with a canonical GT-AG splicing site in bold-italic. Sequence highlighted in pink represents the terminal coding sequence for tGPR50. The resulting protein lacks the last two transmembrane domains and contains 3 unique C-terminal amino acids. C. Schematic of Gpr50 gene showing the alternative splicing site in exon 2. D. Schematic of GPR50 protein, with common amino acids shown in black and sequence unique to wGPR50 and tGPR50 represented in grey and pink, respectively.

86

Figure 3.2 Immunoblotting of GPR50. Western blot analysis of total cell lysates from HEK293 cells, transiently transfected with empty vector (EV) control, C-terminally tagged wGPR50myc or tGPR50myc plasmids, using α-myc antibody confirmed expression of both forms. Immunoreactive protein migrating at ~69 kDa, the expected size of wGPR50, was detected, as well as protein at ~29 kDa representing tGPR50. Immunoblotting of wGPR50 revealed a lower molecular weight band at ~36 kDa, which likely reflects a proteolytic cleavage product of the C-terminal domain. Further western blot analysis using an n-terminally tagged HAGPR50 construct identified an N-terminal product migrating at ~39 kDa, which could be the reciprocal fragment to the C-terminal domain product.

87 the plasma membrane (Figure 3.3, A), while tGPR50EGFP demonstrated a cytoplasmic distribution (Figure 3.3, B). Immunocytochemistry using α-calnexin antibody and a texas-red-conjugated secondary antibody, showed co-localisation of tGPR50 with this marker of the ER (Figure 3.3, C, D). Co-expression of GPR50HcRed and tGPR50EGFP showed, as before, that full-length GPR50 localised to the plasma membrane, whereas tGPR50 accumulated intracellularly. Co-localisation of the two proteins occurs within the cell, presumably within the ER (Figure 3.3, E-G).

Truncated variants often form heterodimers with their WT counterparts and alter the intracellular trafficking of the WT receptor, thus studies were carried out to assess whether the same would be true for tGPR50 and wGPR50. To assess whether co- expression with tGPR50 affects the plasma membrane insertion of wGPR50, N- terminally tagged HAGPR50 and C-terminally tagged tGPR50EGFP plasmids were co- expressed in HEK293 cells, followed by immunocytochemistry on unpermeabilised cells using α-HA antibody and texas-red-conjugated secondary antibody. Antibody detection of the extracellular N-terminal epitope tag on full-length GPR50 confirmed its insertion into the plasma membrane (Figure 3.4, A). No positive staining was detected when cells were incubated with secondary antibody alone (data not shown). wGPR50 further demonstrated a certain degree of co-localisation with tGPR50EGFP (yellow; Figure 3.4, C), indicating that although tGPR50 accumulated predominantly within the cytosol, a small amount localised to the plasma membrane.

The potential influence of tGPR50 on surface expression of wGPR50 was assessed using FACS analysis. HEK293 cells expressing either HAGPR50EGFP or HAtGPR50EGFP alone, or co-expressing HAGPR50EGFP and tGPR50myc were incubated with α-HA Alexa Fluor 647-conjugated antibody before sorting by expression of GFP and Alexa 647 (Figure 3.4, D-E). The percentage of transfected cells (GFP+ve) that showed cell- surface expression of the HA-epitope was significantly lower in cells expressing the truncated version of GPR50 (HAtGPR50EGFP) compared to those expressing the full- length receptor (12% GFP+ve cells were also HA+ve vs. 36% for full-length GPR50). This confirms that although a small amount of this receptor does insert into the plasma membrane, most is retained within the cytosol. Importantly, co-expression of tGPR50 with HAwGPR50EGFP did not alter the degree of surface expression of the full-length protein construct, suggesting there is little or no interaction of the two isoforms. The relatively low figures obtained for plasma membrane expression of full-length GPR50 88

Figure 3.3 Subcellular localisation of GPR50 in mammalian cells. wGPR50EGFP and tGPR50EGFP plasmids were expressed in HEK293 cells. Full length GPR50 localised predominantly to the plasma membrane (A), while tGPR50 accumulated intracellularly (B). Immunocytochemistry using α-calnexin antibody, a marker of the ER, and a texas-red-conjugated secondary antibody (C), showed that tGPR50 co-localised with the ER (D, yellow). wGPR50HcRED (E) and tGPR50EGFP (F) plasmids were co-expressed in HEK293 cells (G). As before, wGPR50 localised to the plasma membrane and tGPR50 accumulated intracellularly. Co-localisation of the two proteins occurs within the cell (G, yellow), presumably in the ER. Blue = DAPI counterstaining, cells were examined with an Olympus BX51 upright widefield microscope and a x60 objective.

89

Figure 3.4 Plasma membrane insertion of GPR50. A-C. N-terminally tagged HAwGPR50 and C-terminally tagged tGPR50EGFP plasmids were co-expressed in HEK293 cells, followed by immunocytochemistry on unpermeabilised cells using α-HA antibody and a texas-red-conjugated secondary. Detection of the extracellular epitope tag confirmed insertion of wGPR50 into the plasma membrane (A) and some membrane co-localisation with tGPR50, yellow (C). To quantify surface expression, FACS analysis of HEK293 cells expressing HAGPR50EGFP or HAtGPR50EGFP alone, or co-expressing HAGPR50EGFP and tGPR50myc plasmids was carried out after incubation with α-HA Alexa Fluor 647-conjugated antibody. Cells were sorted by GFP and Alexa 647 for PE (D-E). The percentage of transfected cells (GFP+ve) that were also sorted for showing cell-surface expression of the HA-epitope was significantly lower for cells expressing HAtGPR50EGFP (12% vs 36% for HAGPR50EGFP) confirming that although a small amount of this receptor is inserted into the plasma membrane, most of it is retained intracellularly. Co-expression of tGPR50myc with HAGPR50EGFP did not alter the degree of surface expression of the full- length protein construct (E). Data shown are mean ± SEM; ***P< 0.001, significance compared to HAGPR50EGFP, one-way ANOVA with Bonferroni’s post-hoc test.

90 might be a consequence of reduced sensitivity of α-HA binding, compared to the endogenously expressed EGFP.

3.2.4 Co-immunoprecipitation of GPR50 with Gαi1

GPR50 is structurally related to the melatonin receptor family of GPCRs, which classically couple to members of the Gi/o subfamily of G proteins. To assess whether GPR50 can similarly signal through these G proteins, HEK293 cells which stably express a GPR50myc plasmid were created to examine potential interacting Gα subunits through co-immunoprecipitation (co-IP) experiments. Cells were subsequently HA transiently transfected with a specific Gα construct (including Gαi1, Gαi2, Gαi3 and

Gαs), and total cell lysates were subjected to immunoprecipitation using α-myc antibody. Western blotting of immunoprecipitates using α-HA antibody for Gα protein detection (Figure 3.5, A), showed in all lanes, a band running at ~55 kDa, representing the antibody heavy chain. Interestingly an interaction with Gα was revealed, but was HA HA HA specific to Gαi1 (band at ~40 kDa). No co-IP was observed for Gαi2, Gαi3 or Gαs. myc Gαi1 did not precipitate with tGPR50 (Figure 3.5, C) or in the absence of wGPR50 transfection (Figure 3.5, B), indicating specific association between wGPR50 and Gαi1.

3.2.5 GPR50 decreases forskolin-stimulated cAMP accumulation

Gαi signals through the cyclic AMP (cAMP) generating enzyme adenylyl cyclase, therefore levels of cAMP were measured in HEK293 cells transiently transfected with empty vector, GPR50myc or tGPR50myc. Basal levels of cAMP showed no difference with transfection of the various constructs, and stimulation of adenylyl cyclase with 10µM forskolin significantly increased intracellular cAMP accumulation in all groups (Figure 3.6, A). However, forskolin-stimulated cells transfected with full-length GPR50 had significantly lower levels of intracellular cAMP compared to cells subject to mock transfection or those expressing tGPR50. These findings indicate that GPR50 signalling via Gαi decreases forskolin-stimulated cAMP accumulation. cAMP levels in cells transfected with tGPR50 however, were comparable to that of the control group.

91

Figure 3.5 wGPR50 co-immunoprecipitates with Gαi1. A, C. HAGα subunits were expressed in HEK293 cells stably-expressing wGPR50myc or tGPR50myc constructs, and total cell lysates (TCL) were subjected to immunoprecipitation (IP) using α-myc antibody. Western blotting (WB) of HA immunoprecipitates using α-HA antibody, showed the specific association of Gαi1 myc myc HA with wGPR50 (A). tGPR50 was unable to bind Gαi1 (C). B. Immunoprecipitation of HEK293 cells transfected with the various HAGα subunits alone demonstrated that these proteins do not precipitate in the absence of wGPR50, indicating specific association between these proteins.

92

Figure 3.6 wGPR50 decreases forskolin-stimulated cAMP accumulation. A-C. HEK293 cells expressing empty vector (EV), wGPR50myc or tGPR50myc were stimulated with vehicle or 10µM forskolin before analysis of intracellular cAMP levels (A) (n=6/group, representative experiment repeated 3 times) or total cell lysates by western blotting using α-CREB and α-p133CREB antibodies (B-C) (n=3/group, representative experiment repeated 3 times). Cells transfected with wGPR50 had significantly reduced levels of intracellular cAMP compared to the mock transfected group. cAMP levels in cells transfected with tGPR50 were comparable to control (A). Forskolin significantly increased levels of p133CREB in all groups; however transfection with wGPR50 significantly reduced levels of p133CREB compared to the mock transfected group. P133CREB levels in cells transfected with tGPR50 were comparable to that of the control group. Data shown as mean ± SEM; ***P< 0.001 significance with forskolin treatment versus vehicle, ##P< 0.01 significance with transfection construct; one-way ANOVA, with Bonferroni’s post-hoc test.

93

3.2.6 GPR50 decreases CREB serine133 phosphorylation

The classical intracellular signalling pathway controlled by levels of cytosolic cAMP is directed through the enzyme PKA (cAMP-dependent protein kinase) (for a review see Mayr and Montminy, 2001). PKA resides in the cytoplasm as an inactive heterotetramer of paired regulatory and catalytic subunits. When concentrations of cAMP rise following activation of adenylyl cyclase, cAMP binds sites on the regulatory subunits, allowing release of the catalytic subunits, which passively diffuse into the nucleus and initiate gene expression by phosphorylating the basic leucine zipper transcription factor CREB (cAMP response element-binding protein) at serine133. Phosphorylated CREB then mediates the activation of cAMP-responsive genes by binding to conserved cAMP-responsive elements (CRE) located in the promoter region of target genes. As GPR50 decreases forskolin-stimulated cAMP accumulation, it was further demonstrated that this signalling impacts on the accumulation of forskolin- stimulated CREB Ser133-phosphorylation (Figure 3.6, B and C). Empty vector, GPR50myc or tGPR50myc constructs were expressed in HEK293 cells before analysis of total cell lysates by western blotting. Basal levels of p-133CREB were low and showed no difference with transfection of the different constructs. Stimulation of cells with 10µM forskolin significantly increased levels of p-133CREB in all transfection groups; however cells transfected with full-length GPR50 had significantly reduced levels of p-133CREB compared to the empty vector transfected cells or those expressing tGPR50. This presumably occurs via reduced PKA-CREB signalling, due to GPR50’s ability to couple p-133 Gαi1 and inhibit cAMP accumulation within cells. CREB levels in cells transfected with tGPR50 were similar to levels in the mock transfected cells, again demonstrating that this truncated version of GPR50 lacks the same signalling capacity as the full- length receptor.

3.3 DISCUSSION

This study aimed to identify putative intracellular signalling pathways directed by GPR50. To initiate these studies, murine Gpr50 was cloned from hypothalamus and pituitary tissue, established sites of GPR50 expression (Drew et al., 2001; Ivanova et al., 2008; Reppert et al., 1996). This revealed a variant of Gpr50 containing 732 bases, compared to the expected 1773 bases of the full length receptor cDNA. The smaller Gpr50 transcript represents a splice variant of Gpr50, in which a fragment of exon 2 has been removed. The altered splicing introduces a premature stop-codon, which results in a truncated form of GPR50 protein, lacking the last two transmembrane 94 domains and the long intracellular tail. The alternative splicing acceptor and donor sites used to generate this variant are conserved in exon 2 of the human, rat and sheep GPR50 genes, suggesting that truncated forms of GPR50 are also expressed in these species.

Two splicing variants of human Gpr50 have previously been reported (Slominski et al., 2003), including an isoform generated by a 151 base pair deletion in exon 2 of the transcript, leading to a frame shift and transcription termination. The excised sequence corresponds to the sequence lost from the murine tGpr50 identified in this study, and is presumably generated using the conserved splice donor and acceptor sites described (Figure 3.1). However, this shorter form was isolated from two melanoma cell lines, leading the authors to suggest it as a product of splicing deregulation commonly observed in cancer. In the present study, identification of tGPR50 in the tissues of healthy mice indicates this idea does not appropriately explain the alternative splicing of Gpr50, and that it is perhaps a directed mechanism.

Alternative mRNA splicing has been described for many GPCRs and introduces considerable structural diversity of expressed proteins, which can impact greatly on the properties of the receptor. Alternative forms can differ in tissue distribution, ligand- binding capability, and signalling potential (Minneman, 2001) and accumulating evidence suggests that splice variants of GPCRs occupy functionally distinct roles (reviewed by Markovic and Challiss, 2009), implying they have a significant influence on GPCR biology. Alternative splicing that introduces premature stop codons, producing proteins lacking one or more transmembrane domains, have been reported for several GPCRs (Bakker et al., 2006; Cordoba-Chacon et al., 2011; Coupar et al., 2007; Grosse et al., 1997; Harada et al., 2008; Karpa et al., 2000; Lee et al., 2000; Motomura et al., 1998; Zhu and Wess, 1998). Studies using these truncated receptors mainly report their altered intracellular localisation to their wild-type counterparts along with an ability to impair signalling through the full-length receptor in a dominant- negative fashion, by impairing its recruitment to the plasma membrane. The present work shows that GPR50 localised, as expected, to the plasma membrane, whereas tGPR50 remained primarily in the ER. However, co-expression studies revealed that tGPR50 does not block expression of wGPR50 within the plasma membrane, implying that this truncated variant does not have a role in regulating wGPR50 signalling at the cell surface by altering its subcellular trafficking. The cytosolic localisation of tGPR50 95 suggests it lacks certain motifs that improve either targeted delivery and/or anchoring to the plasma membrane. Alternatively, due to the loss of such a large proportion of the receptor, tGPR50 may be retained within the ER as misfolded protein. FACS analysis showed that a small proportion of tGPR50 does reach the plasma membrane, indicating that it can be directed here, albeit inefficiently. Recognition of proteins by the cell’s quality control system (QCS) is purely physical and does not consider the ability of a protein to function, with chaperones of the QCS recognising general errors of folding, such as exposure of hydrophobic regions in an aqueous environment, rather than specific defects of the protein, such as failure of receptors to recognise ligand (Ulloa-Aguirre et al., 2006). Thus, tGPR50’s retention in the ER is not necessarily a consequence of the protein being non-functional. Cells may use the technique of misfolding and misrouting variant proteins as a post-translational regulatory mechanism to decrease the localisation of wild-type proteins to their normal site of function, as previously discussed, or to potentially provide a protein reserve to call upon without the need for transcription or translation. Although this may appear intrinsically wasteful in light of the high metabolic cost of synthesising unused receptor, there may be a selective advantage, since there are an increasing number of examples demonstrating only partial expression of receptors at their normal locus of action (Castro-Fernandez et al., 2005). Finally, it is perhaps a classical notion that GPCRs function only at the cell surface where they access their extracellular ligands; a notion challenged by several recent studies (Head et al., 2005; Marrache et al., 2005; Miaczynska et al., 2004). For example, murine GluR5 can be found at the nuclear membrane as a functional receptor, able to mediate calcium signalling (O'Malley et al., 2003); GPR30 is localised mainly in the ER where it can rapidly mobilise calcium and activate phosphoinositide-3- kinase upon ligand-stimulation (Revankar et al., 2005; Revankar et al., 2007); and native CB1 cannabinoid receptors do not reach the cell surface, but reside primarily within endosomal/lysosomal compartments where they too can mediate signal transduction via G proteins (Rozenfeld and Devi, 2008). The broad distribution of tGPR50 mRNA (see Chapter 6, Figure 6.5, A) and conservation of the splicing site suggest a physiological significance of this variant, perhaps as a means by which wGPR50 function is regulated, which remains to be determined.

The best known intracellular signalling pathway mediated by the melatonin receptors is the inhibition of cAMP formation via pertussis-toxin sensitive G proteins. GPR50 shares 45% amino acid sequence identity with MT1 and MT2, and the current work shows that

GPR50 similarly couples to members of this G protein subfamily, specifically Gαi1. 96

GPR50 signalling via Gαi1 was demonstrated using co-IP as well as second messenger assays, showing that overexpression of GPR50 in HEK293 cells inhibited forskolin- stimulated cAMP accumulation. Although no ligands are available to specifically stimulate GPR50, it is assumed that signalling can occur during overexpression of the receptor. When overexpressed, GPCRs can achieve an active state even in the absence of agonist, as a result of a shift in equilibrium between their inactive and active conformations (Bockaert et al., 2002). Forskolin-stimulated phosphorylation of CREB is also inhibited in GPR50-transfected cells, probably by virtue of decreased activation of PKA, due to the decreased levels of cAMP. Investigation of other intracellular pathways activated by GPR50, such as coupling to other G proteins or activation of mitogen- activated protein kinase pathways, would allow a more comprehensive characterisation of the signalling properties of the receptor. tGPR50 was unable to couple to Gαi1, and did not impact on intracellular cAMP accumulation or CREB phosphorylation. This is perhaps not surprising, as tGPR50 is truncated within intracellular loop-3 (IL-3), and the general idea with many class A GPCRs is that the C-terminal part of IL-3 determines specificity of interaction with particular G proteins (Bockaert et al., 2002). Most of the intracellular surface of the receptor (including IL-4 and the C-terminal tail) are implicated in G protein coupling, again sites which are lacking in tGPR50.

Although it cannot be ruled out that the ligand for GPR50 was present in the media used to culture cells for the cAMP signalling experiments, it may be that GPR50 possesses inherent constitutive activity that is not attributable to overexpression alone. Constitutive signalling is an intrinsic property of many GPCRs, and studies using either wild-type or mutant receptors have confirmed the ability of GPCRs to signal in a constitutive agonist-independent manner (reviewed in Smit et al., 2007). Constitutive activity has been documented for more than 60 wild-type GPCRs (Seifert and Wenzel- Seifert, 2002), but in contrast to the wealth of available pharmacological data, structural information on GPCRs is still scarce, and the molecular mechanisms of GPCR activation are still not fully understood. It is proposed that the inactive conformation of the receptor is maintained through restraining intramolecular interactions impeding motion of several, if not all transmembrane (TM) domains (Farrens et al., 1996). Release of these constraints is induced by either agonist-binding or constitutive activity-inducing mutations (CAMs) within the receptor, which may be spontaneous or engineered (Smit et al., 2007). Unique features of the melatonin receptor group, which are also found in GPR50’s amino acid sequence, include an NRY sequence in TM3, rather than the typical (D/E)RY sequence (Reppert et al., 1996). Normally, an ‘ionic 97 lock’ exists between the R residue with its adjacent D/E residue and an additional D/E near the cytoplasmic end of TM6 (Ballesteros et al., 2001). This motif is a well-studied CAM site, and charge-neutralising mutation of D/E in TM3 removes the ionic interaction and increases constitutive activity of many GPCRs (Alewijnse et al., 2000; Ballesteros et al., 2001; Kim et al., 1997; Scheer et al., 1996). As GPR50 and the melatonin receptors contain a neutral N residue at this position in TM3, this is potentially responsible for the high basal activity of GPR50, and the reported spontaneous activity demonstrated by MT1 (Roka et al., 1999). Mutation of this residue to a D/E amino acid might abolish GPR50’s apparent constitutive activity. It would be useful to examine whether a physiological expression level of GPR50 is enough to evoke the constitutive activation of Gi proteins, which could be achieved by using a knock-down strategy in cells that endogenously express GPR50, and measuring basal levels of cAMP. If endogenous GPR50 is constitutively active, one would expect to see increases in basal cAMP when expression of GPR50 is inhibited. However, before any experimental strategies assessing constitutive activity of GPR50 are undertaken, it must first be established that its ligand is not present in the cell culture media used, perhaps through the development of cell-free, membrane-based assays which are prepared containing the GPCR and associated G proteins, and used to assess basal receptor activation of G proteins by measuring labelled GTP incorporation in the absence of cell culture media (Eglen, 2005).

A ligand-independent function for GPR50 has recently been proposed, acknowledged by its ability to heterodimerise with MT1 and prevent high-affinity agonist binding, G protein coupling and β-arrestin binding to this receptor (Levoye et al., 2006). Interestingly, the C-terminal tail of GPR50 is essential for this inhibition. Immunoblotting of an epitope-tagged construct showed full-length GPR50 along with a lower molecular weight band, which likely reflects a proteolytic cleavage product of the C-terminal domain. The size of the band implies cleavage of approximately the whole C-terminal tail, so speculatively; it is therefore possible that proteolytic events control the ability of GPR50 to impact on melatonin receptor signalling. Processing of the C-terminal tail of GPCRs has been previously reported for the angiotensin II type 1 receptor (Cook et al., 2007) and the Frizzled2 receptor (Ataman et al., 2006), resulting in C-terminal fragments that re-locate to the nucleus for potential transcriptional regulation events. A putative enzymatic cleavage site has previously been described within the C-terminal tail of GPR50 (Reppert et al., 1996), but the resulting fragment would be much smaller than the one identified by immunoblotting, and thus further characterisation of the 98 proteolytic site and potential enzyme responsible for this cleavage is required. Alternatively, if tGPR50 can likewise engage in heterodimers with the MT1 receptor, as it lacks the C-terminal tail it will not impair MT1 signalling, and thus regulation of expression of wGPR50 versus tGPR50 within cells could impact on melatonin receptor signalling.

99

CHAPTER 4

GPR50 IN WHITE ADIPOSE TISSUE

100

4.1 INTRODUCTION

Gpr50-/- mice exhibit lower body weights than their WT littermates, mainly attributable to reduced fat accumulation. Gpr50-/- mice further demonstrate reduced weight loss during food deprivation, and partial resistance to diet-induced obesity (Ivanova et al., 2008). In humans, polymorphisms in Gpr50 have been linked to elevated circulating triglycerides and reduced high-density lipoprotein cholesterol levels (Bhattacharyya et al., 2006). The following studies therefore aimed to investigate GPR50 outside the central nervous system, by examining receptor expression in peripheral tissues. As white adipose tissue (WAT) demonstrated expression of GPR50, further studies were undertaken in an attempt to determine a potential role for the receptor in this organ.

WAT contains adipocytes which store excess energy in the form of triglycerides inside lipid droplets, composed of a neutral lipid core surrounded by a phospholipid monolayer, interspersed with proteins of the perilipin family (Brasaemle, 2007). In addition to their role in energy storage, adipocytes perform endocrine functions by secreting and responding to hormones and cytokines that regulate energy intake and expenditure (Poulos et al., 2010). During periods of increased energy demand, lipolysis in adipocytes is activated, a process in which triglycerides are hydrolysed into glycerol and free fatty acids (FFAs). This process releases FFAs into the bloodstream, where they may be re-esterified by the adipocyte or provide other tissues with fuel for oxidation (Schweiger et al., 2006).

Significant advances regarding the molecular basis for adipocyte differentiation have occurred in recent years, along with ever-expanding lists of agents involved in the regulation of triglyceride accumulation/hydrolysis in adipocytes. Several cellular models are available to study adipogenesis, including primary and immortalised embryonic stem cells and multipotent precursor cells isolated from adult tissues (Ntambi and Young-Cheul, 2000). A well-characterised and reliable model for studying adipocyte function is the murine 3T3-L1 preadipocytic cell line, which faithfully recapitulate the critical aspects of fat formation in vivo, and when fully differentiated, possess the same ultrastructural characteristics of adipocytes, with the formation, appearance and maintenance of fat droplets mimicking that of live adipose tissue (Green and Kehinde, 1979; Novikoff et al., 1980). During the differentiation process (see Chapter 2, Figure 2.1), 3T3-L1 cells are shifted from dividing preadipocytes to growth-arrested

101 adipocytes. Growth arrest is a pre-requisite step for differentiation induction, and is achieved by contact-inhibition at confluence. Only then will a hormonal regimen induce adipocyte differentiation (Sadowski et al., 1992). This hormonal cocktail activates signalling pathways which converge on a tightly regulated cascade of transcriptional events that is only now being fully elucidated. Peroxisome proliferator-activated receptor γ (PPARγ) lies at the core of the transcriptional cascade and is both necessary and sufficient for adipogenesis, and no factor has been discovered that promotes adipogenesis in the absence of PPARγ (Rosen et al., 2000). Several CCAAT-enhancer binding proteins (C/EBP) family members also participate in adipogenesis, and these alongside PPARγ, alone or in coorperation with each other, induce the transcription of many adipocytic genes encoding proteins and enzymes involved in creating and maintaining the adipocyte phenotype. The discovery of PPARγ and the C/EBPs gave the field a starting point, but it is becoming clear that numerous transcription factors, in addition to a complex and interleaving set of cofactors, have a significant role in differentiation and maintenance of the mature adipocyte (for a review see Rosen and MacDougald, 2006).

The following studies aimed to investigate the influence of GPR50 on adipocyte differentiation and further determine the impact this receptor has on adipocyte metabolism.

4.2 RESULTS

4.2.1 Identification of GPR50 expression in white adipose tissue

Total RNAs isolated from murine WAT and day 12 differentiated 3T3-L1 cells were subject to cDNA synthesis, before RT-PCR using primers specific for Gpr50 (Figure 4.1, A). This revealed two transcripts representing full-length Gpr50 and its truncated splice variant in both WAT and the adipocytic 3T3-L1 cell line.

Immunoreactivity using antibody directed against GPR50, demonstrated receptor expression in WAT (Figure 4.1, B). No immunostaining was observed in sections incubated with secondary antibody alone (data not shown), or sections from Gpr50-/- mice (Figure 4.1, C).

102

Figure 4.1 Gpr50 expression in white adipose tissue.

A. RT-PCR cloning of Gpr50 from murine white adipose tissue (WAT) and day 12 differentiated 3T3-L1 adipocytes produced two bands representing full length Gpr50 (~1.8kb) and its truncated variant (~1.7kb). B-C. GPR50 immunoreactivity in murine WAT (B) Blue = DAPI counterstaining. Positive staining was not seen in WAT from Gpr50-/- mice (C), or when no primary antibody was included (data not shown).

103

4.2.2 Gpr50 expression during adipogenesis

3T3-L1 preadipocytes were grown to confluence. After 2 days of contact-inhibition growth arrest (i.e. day -2 to day 0), 3T3-L1 cells were induced to differentiate into mature adipocytes by treatment with a hormonal ‘cocktail’ containing the synthetic glucocorticoid dexamethasone, insulin, the phosphodiesterase inhibitor methylisobutylxanthine, and fetal bovine serum (at day 0) (Sadowski et al., 1992). After 3 days in this supplemented media (i.e. day 4 of differentiation protocol), cells were fed maturation media containing insulin every 2 days. After day 8, lipid droplets inside the cell were visible, which further enlarge so that by day 12, large droplets are visible.

To examine the time course of changes in Gpr50 expression, total cellular RNA was prepared every 2 days during the differentiation process and examined by qPCR, using primers specific for Gpr50, and relative to housekeeping 18S control expression levels. In preadipocytes, Gpr50 mRNA was detectable at low levels, and decreased from day -2 before the start of the differentiation assay up to day 0, possibly caused by growth inhibition as a result of contact inhibition (Figure 4.2, A). Between days 2 and 10 after initiation of differentiation, a progressive increase in Gpr50 level was observed. As a control for differentiation, expressions of Pparg, and Fabp4 an established PPARγ target gene, were determined, and as expected, both genes were induced during adipocyte differentiation (Figure 4.2, B-C).

4.2.3 Triglyceride accumulation in 3T3-L1s with targeted knock-down or overexpression of Gpr50

To assess the relevance of GPR50 in adipogenesis, 3T3-L1s which stably expressed plasmids encoding GPR50myc, an empty vector control, shRNA targeting Gpr50, or a control shRNA, were created. These stable cell lines were then induced to differentiate as previously described, up until day 12 of differentiation. Expression of Gpr50-specific shRNA reduced expression of Gpr50, assessed by qPCR at day 12, by approximately 60% (Figure 4.3, A) when compared to cells transfected with a non-target shRNA control vector. Ectopic expression of GPR50myc increased Gpr50 expression by approximately 200-fold at day 12 (Figure 4.3, B) compared to empty-vector transfected cells. Incredibly, knockdown of GPR50 expression in 3T3-L1s resulted in apparent impaired differentiation of these cells into adipocytes, illustrated by a marked reduction

104

Figure 4.2 Gpr50 expression in 3T3L1 cells increases during differentiation. A. Gpr50 expression in 3T3L1 cells during differentiation was quantified by qRT-PCR (n=4/group, representative experiment repeated 3 times). Gpr50 expression increases during differentiation, assessed by induction of Pparγ, a marker of differentiation (B) and Fabp4, a PPARγ target gene (C). Gene expression is reported as mean ± SEM, relative to levels at day -2. ***P< 0.001, **P<0.01, P<0.05, one-way ANOVA, with Bonferroni’s post-hoc test.

105

Figure 4.3 Gpr50 plays an essential role in triglyceride accumulation. A-D. Mouse 3T3-L1 preadipocytes stably-expressing a control or GPR50-specific shRNA construct, or empty vector or Gpr50myc construct, were differentiated into mature adipocytes. At day 12, RNA was extracted and qPCR analysis showed that knockdown of Gpr50 using shRNA reduced expression by ~60% (A) and ectopic expression of Gpr50myc increased expression by ~207-fold (B). In parallel, day 12 differentiated cells were fixed and stained for triglycerides using Oil-Red-O (C). Pictures are representative of four independent experiments. Oil-Red-O accumulation was quantified by extracting in isopropyl and measuring optical density, in triplicate. D-E. Data shown as mean ± SEM, relative to control transfections and normalised to the mouse housekeeping 18S rRNA internal control; ***P< 0.001, students’ t-test.

106 of accumulated triglycerides, as assessed by Oil-Red-O staining at day 12 of differentiation (Figure 4.3, C-D). On the other hand, overexpression of GPR50 appeared to strongly promote adipocyte differentiation, with larger amounts of triglyceride accumulated by day 12, compared to empty-vector control (Figure 4.3, C,E).

4.2.4 Adipogenic gene expression in 3T3-L1s with targeted knock-down or overexpression of Gpr50

Using qPCR analysis, the expressions of adipogenic genes (see Figure 4.6) in day 12 differentiated 3T3-L1 adipocytes, with stable elevation or repression of GPR50 expression, were measured. Expressions of Pparg as a marker of differentiation, Fas (fatty acid synthase) involved in the de novo synthesis of fatty acids, Hsl (hormone sensitive lipase) and Atgl (adipose triglyceride lipase), responsible for lipolytic mobilisation of fatty acids from intramyocellular lipid stores, and Lpl (lipoprotein lipase), which controls fatty acid uptake by hydrolysing circulating triglycerides packaged in lipoproteins, were quantified. Knockdown of GPR50 using targeting shRNA had no significant effect on PPARγ or Fas expressions (Figure 4.4, A, B), but significantly decreased the mRNA expressions of Lpl (69%), Hsl (81%) and Atgl (40%) compared to cells transfected with a non-targeting shRNA (Figure 4.4, C-E). Cells overexpressing GPR50myc demonstrated significantly higher levels of Pparg (2.3-fold), Atgl (2.6-fold), Lpl (2.2-fold) and Hsl (3.5-fold), but again, no significant alteration in Fas expression levels were demonstrated, compared with empty-vector transfected cells (Figure 4.4, F- J). These results suggest that GPR50 does not affect the differentiation of adipocytes per se, as knockdown of the receptor does not alter expression of the differentiation marker PPARγ. Therefore, the increased or decreased triglyceride accumulation demonstrated in 3T3-L1s with stably elevated or repressed GPR50 expression, respectively, might be a result of altered lipid metabolism in these cell lines. As stable elevation or repression of GPR50 expression did not influence expression of Fas, it can be concluded that de novo synthesis of fatty acids is not affected by GPR50; however opposing effects of overexpression and knockdown of the receptor on key lipolytic enzymes suggests some influence of GPR50 on triglyceride uptake and lipolysis in adipocytes.

107

Figure 4.4 Influence of GPR50 on expression of lipogenic and lipolytic genes. qPCR analysis of RNA extracted from 3T3-L1 cells at day 12 of differentiation, stably- transfected with control shRNA, shRNA targeting Gpr50, empty vector or Gpr50myc encoding plasmids. Knockdown of GPR50 had no effect on Pparγ (A) or Fas (B) expression, but significantly decreased the expression of Lpl (69%; C), Hsl (81%; D) and Atgl (40%; E), compared to cells transfected with a non-targeting shRNA. Cells overexpressing GPR50myc demonstrated significantly higher levels of Pparg (2.3-fold; F), Lpl (2.2-fold; H), Hsl (3.5-fold; I) and Atgl (2.6-fold; J), but had no effect on Fas (G) expression, compared to empty vector transfected cells. Data shown as mean ± SEM, relative to control transfections and normalised to the mouse housekeeping 18S rRNA internal control; *P<0.05, **P<0.01, students’ t-test.

108

Figure 4.5 Influence of GPR50 on triglyceride accumulation and lipolysis. A. Total intracellular triglyceride content was measured (TG-1-NC; Zen Bio) in stable 3T3-L1 cells expressing empty vector, mycGpr50, control shRNA or shRNA targeting Gpr50 at day 12 of differentiation. (Data shown are mean ± SEM; ***P< 0.001 vs. control transfections). Rates of lipolysis were assessed by measuring cellular efflux of non-esterified fatty acids (NEFA) (B) and glycerol (C) (LIP-3-NC; Zen Bio). Data shown are mean basal efflux and 10µM forskolin-stimulated lipolysis ± SEM; ***P< 0.001, *P<0.05 significance between basal and stimulated lipolysis; #P<0.05, #P<0.01, ###P<0.001 significantly different vs. appropriate control condition, two-way ANOVA with Bonferroni’s post-hoc test.

109

4.2.5 Lipolysis in 3T3-L1s with targeted knock-down or overexpression of Gpr50

To assess the effect of GPR50 on the lipolytic properties of adipocytes, total triglyceride levels and forskolin-induced lipolysis were measured in the stable 3T3-L1 cell lines (Figure 4.5). Measurement of total triglyceride levels on day 12 of differentiation showed that cells transfected with GPR50myc stored approximately 2.5 times more triglyceride than those transfected with empty-vector (Figure 4.5, A). Conversely, cells in which GPR50 was knocked-down had an 86% reduction in stored triglycerides compared to cells transfected with a non-targeting control shRNA (Figure 4.5, A). These data are consistent with the lower retention of Oil-Red-O by shRNA- Gpr50 transfected cells and increased staining in GPR50myc transfected cells.

Forskolin can be used in cell culture to simulate lipolysis, by directly activating adenylyl cyclase, to increase levels of intracellular cAMP. Elevated cAMP acts as a second messenger to activate HSL, which catalyses the hydrolysis of triglycerides to glycerol and free fatty acids. Forskolin treatment of day 12 differentiated adipocytes increased the release of free glycerol and non-esterified fatty acids (NEFA) compared to vehicle in all the stable cell lines apart from those expressing GPR50-specific shRNA (Figure 4.5, B-C). However, in cells stably transfected with GPR50myc, levels of released glycerol and NEFA were significantly reduced, compared to control EV-transfected cells (Figure 4.5, B-C), indicating impaired stimulated lipolysis in these adipocytes. Interestingly, basal levels of free glycerol and NEFA were significantly increased in adipocytes stably overexpressing GPR50 (Figure 4.5, B-C), suggesting increased constitutive levels of lipolysis in these cells compared to control. Conversely, adipocytes with stable knock-down of GPR50 demonstrated significantly reduced levels of glycerol and NEFA in the basal and stimulated state, compared to control cells expressing a non-specific shRNA (Figure 4.5, B-C), indicating that loss of the receptor leads to decreased lipolysis in adipocytes.

4.3 DISCUSSION

This work reports the expression of GPR50 in murine WAT, and explores the potential significance of the receptor in this organ. Levels of Gpr50 mRNA steadily increased during the differentiation process of the 3T3-L1 preadipocytic cell line, reaching the highest level of expression in mature adipocytes. When GPR50 expression was stably

110 knocked down in 3T3-L1s and cells induced to differentiate, the proportion of adipocytes staining positively with Oil-Red-O for lipid droplets decreased, as did assayed levels of accumulated triglycerides. The expression of Pparg was not significantly different in these day 12 differentiated cells compared to control, suggesting that GPR50 is not essential during the adipogenic process per se, but that its function in mature adipocytes is essential for the formation or storage of lipid.

Differentiated 3T3-L1s stably overexpressing GPR50 demonstrated increased triglyceride accumulation and expression of Pparg. This nuclear hormone receptor is essential for adipogenesis, and regulates the expression of multiple lipogenic genes in mature fat cells. These include those encoding the adipocyte fatty acid binding protein (FABP4), lipoprotein lipase (LPL), fatty acid transport protein, acyl-CoA synthase and phosphoenol pyruvate carboxykinase (PEPCK) (Rosen and Spiegelman, 2001). Consequently, PPARγ directs the storage of fatty acids in adipocytes. There is much supporting data highlighting the role of PPARγ in stimulating lipogenesis, for example; patients taking synthetic PPARγ activators frequently gain weight (Fuchtenbusch et al., 2000); heterozygous Pparg mutant mice exhibit smaller fat stores on a high fat diet (Kubota et al., 1999); and in the liver, where PPARγ is normally only minimally expressed, hepatic triglyceride accumulation is associated with a dramatic increase in PPARγ expression (Chao et al., 2000). Thus, in 3T3-L1s stably overexpressing GPR50, the increased triglyceride accumulation and increased levels of Lpl, may be direct results of increased Pparg expression. Interestingly, CREB inhibits hepatic PPARγ expression (Herzig et al., 2003), and overexpression of GPR50 in HEK293 cells reduced intracellular levels of cAMP and phosphorylated CREB (Chapter 3, Figure 3.6). Therefore, it is possible that in the cells transfected with GPR50, activation of CREB is reduced, relieving the inhibition on Pparg expression. This, however, is perhaps a rather simplistic view of what is actually an intricate network of intracellular signalling pathways regulating gene expression, but remains an interesting possibility.

Fat accumulation is determined by the balance between lipogenesis and lipolysis (see Figure 4.6). No change in Fas was observed in differentiated 3T3-L1 cells with stable overexpression or repression of GPR50, indicating that control of de novo fatty acid synthesis may not be affected by GPR50 signalling. The mobilisation of FAs from WAT depends on the activity of two major lipases: HSL and ATGL, which are responsible for

111

Figure 4.6 Diagrammatic representation of triglyceride turnover in adipocytes. A simplified model of triglyceride (TG) storage (lipogenesis) and breakdown (lipolysis) into free fatty acids (FFA) and glycerol, with the proteins examined by qPCR (Figure 4.4) highlighted. HSL: hormone sensitive lipase; ATGL: adipocyte triglyceride lipase; LPL: lipoprotein lipase; FAS: fatty acid synthase; PPAR: proliferator-activated receptor; SREBP: sterol regulatory element-binding protein; PGC1: PPARγ co-activator; CEBP: CCAAT-enhancer binding protein; FABP: fatty acid-binding protein; ACC: acetyl-coA- carboxylase; PEPCK: phosphoenolpyruvate carboxylase; ADRP: adipose differentiation related protein; P: phosphor. Diagram by the author.

112 approximately 95% of triglyceride hydrolase activity in adipocytes (Schweiger et al., 2006). Transcriptional levels of HSL and ATGL were increased in differentiated 3T3- L1s overexpressing GPR50, and decreased in cells with reduced expression of GPR50, potentially implicating GPR50 signalling in the regulation of lipolysis in adipocytes.

Subsequent assessment of lipolytic activity in the stable 3T3-L1 cell lines was achieved by measuring free fatty acid and glycerol release from differentiated cells. Cells stably transfected with GPR50 showed constitutively higher levels of fatty acid and glycerol release compared to empty-vector transfected cells, indicating that lipolysis of triglycerides is occurring at a higher rate in these cells. However, when stimulated, the induction of lipolysis was significantly lower, with fold induction of free fatty acids and glycerol both decreased compared to release from control cells. On the other hand, knockdown of GPR50 in 3T3-L1s resulted in a marked decrease of both basal and stimulated lipolysis. As basal lipolytic rate is increased in cells overexpressing GPR50 and decreased in cells with knockdown of GPR50 expression, it seems counterintuitive that they respectively accumulate increased and decreased levels of triglyceride. However, triglyceride accumulation is determined by the balance between lipogenesis and lipolysis, and thus it appears a new steady-state has been achieved by the transfected cell lines, where uptake and release of FAs is balanced. In those overexpressing GPR50, although lipid oxidation is increased, triglyceride accumulation is maintained at a higher level due to increased lipogenesis, or indeed increased re- esterification of fatty acids, with the reverse being true for cells with GPR50 knockdown.

As in the differentiated 3T3-L1s overexpressing GPR50, obesity is similarly associated with an increase in basal lipolysis but a decrease in catecholamine-stimulated lipolysis (Bougneres et al., 1997; Jensen et al., 1989; Jocken and Blaak, 2008). Theories to explain this include the fact that obese subjects tend to develop insulin-resistance, and so impaired sensitivity of adipocytes to insulin signalling, including the antilipolytic effects of this hormone (Choi et al., 2010), may be responsible for increased basal lipolysis. Another idea is that the increased lipolytic rate is a response to overexpression of leptin, which stimulates lipolysis (Fruhbeck et al., 1997). Correspondence with another research group (R. Jockers, unpublished work) revealed that GPR50 interacts with and alters leptin-receptor signalling, and work within our 113 group has shown Gpr50-/- mice display attenuated responses to leptin (Appendix A), therefore it seems possible that overexpression of GPR50 impacts on basal lipolytic rate by affecting leptin signalling. Finally, 3T3-L1s overexpressing GPR50 accumulated larger triglyceride stores than controls, and direct positive correlations have been found between adipocyte size (mainly dependent on the size of the accumulated lipid droplet, which accounts for >90% of the cell volume) and basal lipolytic rate (Wueest et al., 2009). The discovery that basal lipolysis is higher in larger cells was also associated with higher mRNA expression levels of ATGL and HSL (Wueest et al., 2009), both of which are increased in the GPR50 transfected cells. It has further been shown that ATGL plays a dominant role in basal rates of lipolysis (Bezaire et al., 2009; Miyoshi et al., 2008), and therefore higher levels of this enzyme in particular are probably responsible for the increased constitutive lipolysis demonstrated by the 3T3-L1s overexpressing GPR50. HSL has been shown to be the most important lipase during stimulatory conditions, but has negligible contribution to lipolysis under basal conditions (Haemmerle et al., 2003), so increased levels probably do not contribute to the increased basal lipolytic rate exhibited by these cells.

Due to the increased levels of Hsl in the GPR50-transfected 3T3-L1s, it would be assumed that upon stimulation, levels of lipolysis within these cells would be much higher, which was not the case, indicating that factors upstream of HSL activation must be augmented in this cell line. In the basal state, HSL is dispersed in the cytoplasm and perilipin proteins coat lipid droplets (Yamaguchi, 2010). When lipolysis is stimulated, phosphorylation of both perilipin and HSL facilitates access of the lipase to lipid substrates for hydrolysis (Wang et al., 2009). The main pathway leading to these phosphorylation events is the cAMP-dependent PKA pathway, where signalling by Gs- coupled receptors induces the activation of adenylyl cyclase, and the subsequent increase in intracellular cAMP leads to activation of PKA, which phosphorylates HSL and perilipins (Belfrage et al., 1981). At the organismal level, the sympathetic nervous system generates such signals by producing adrenergic stimuli that activate beta- adrenergic receptors, which signal through Gs proteins. Other major regulators of lipolysis include melanocortins and thyroid-stimulating hormone, which also act through

Gs-coupled receptors, alongside adenosine, (D)-β-hydroxybuterate, and lactate, which act through Gi-coupled receptors to inhibit lipolysis (reviewed in Chaves et al., 2011). Even the potent anti-lipolytic effects of insulin, which signals through a non-GPCR membrane receptor, results in augmented cAMP levels and PKA activity via the

114 phosphorylation and activation of phosphodiesterase 3B (PDE3B), which degrades cAMP (Ahmadian et al., 2010).

Clearly, as GPR50 couples to Gαi and inhibits forskolin-stimulated cAMP accumulation (Chapter 3, Figures 3.5 and 3.6), this signalling likely accounts for the reduction in forskolin-stimulated lipolysis demonstrated by the differentiated 3T3-L1s overexpressing the receptor. Alternatively, enhanced fatty acid re-uptake and re- esterification might be responsible for the apparent decreased stimulated lipolysis, meaning that cells are responsive to forskolin-stimulation and release fatty acids accordingly; however they reclaim them at an increased rate. This could also explain how despite demonstrating increased basal lipolysis, cells overexpressing GPR50 accumulate larger triglyceride stores. Such re-esterification of fatty acids relies on an increase in glyceroneogenesis as a result of the induction of its key enzyme PEPCK, whose expression is PPARγ-dependent (Tordjman et al., 2003). Thus, increased levels of Pparg in this cell line could be responsible for enhancing a futile cycle of fatty acid re-esterification. This could also explain why upon stimulation of lipolysis, compared to control there is a major decrease of fatty acids in the media, contrasted with a smaller difference in glycerol levels, as fatty acids but not glycerol are extracted by adipocytes for re-esterification events (Coppack et al., 1999). Further investigation of levels of PEPCK within these cell lines, alongside more elegant time-course experiments or the use of labelled fatty acids would clarify whether increased fatty acid re-esterification in 3T3-L1s overexpressing GPR50 underlies the seeming reduction of stimulated lipolysis.

Basal lipolytic rates in cells transfected with GPR50 targeting shRNA were dramatically reduced, and levels of free fatty acids and glycerol were only marginally increased in response to forskolin-stimulation. This implies that although cells with reduced levels of GPR50 expression have the capacity to respond to exogenous signals and stimulate lipolysis, their response is greatly attenuated. This is likely due to the large decrease in transcription of Hsl in these cells, responsible for lipolysis in stimulatory conditions, along with the decreased levels of Atgl being perhaps responsible for the low levels of basal lipolysis. Further, as these cells exhibit such a huge decrease in intracellular triglyceride accumulation, it may be that upon stimulation, there are only trivial amounts of lipid to mobilise, exacerbating the result demonstrated by the lipolysis assays.

115

Alongside PPARγ, transcriptional regulation of lipogenesis is mediated by sterol regulatory element binding transcription factors, SREBPs (Hua et al., 1993; Tontonoz et al., 1993; Yokoyama et al., 1993). Interestingly, it has been discovered that enlarged adipocytes exhibit a relative decrease in membrane cholesterol content, and depletion activates the expression of the cholesterol-regulated transcription factor SREBP-2 (Le Lay et al., 2001). Induction of SREBP-2 activates in turn the expression of its known target genes, including the LDL receptor, involved in the repletion of cell cholesterol by increasing its uptake. A recent yeast two-hybrid screen implicated an interaction between the C-terminal tail of GPR50 and SREBP-2 (Grunewald et al., 2009), which although unconfirmed, could indicate cross-talk of their activities within adipocytes. Validation of this interaction and investigation of GPR50’s involvement in the regulation of SREBP-2 activity would be an interesting avenue to further explore the function of GPR50 in WAT.

In summary, GPR50 is implicated in the regulation of adiposity, altering the expression of adipocytic genes such as Hsl, Atgl, Lpl and Pparg. Upregulation of Gpr50 mRNA during adipogenesis of 3T3-L1 cells is not essential for regulation of this process, as knockdown of its expression does not block pre-adipocyte differentiation and Gpr50-/- mice are also not deficient in WAT. However, upregulation of Gpr50 mRNA may be essential for modulation of the adipogenic process in mature adipocytes, by functioning as a factor that regulates lipolysis. Increased plasma fatty acid concentration, which is a hallmark of obesity and insulin resistance (Boden, 2011), can be caused by increased lipolysis and fatty acid release from adipose tissue. It seems that GPR50 may be associated with an obese phenotype, with overexpression of the receptor associated with increased triglyceride accumulation, increased basal lipolysis and impaired acute lipid mobilisation. On the other hand, loss of the receptor results in mice with lower WAT stores and a resistance to diet-induced obesity (Ivanova et al., 2008), and knockdown of Gpr50 expression in adipocytes results in drastically lower accumulation of stored triglyceride.

Polymorphisms in the gene encoding GPR50 have been associated with increased levels of circulating triglycerides and HDL (Bhattacharyya et al., 2006). The present work indicates an influence of GPR50 on the expression of Lpl in adipocytes, which

116 has a crucial role in the uptake of circulating triglycerides, by mediating the release of fatty acids from triglycerides packaged in lipoproteins (Gonzales and Orlando, 2007).

The precise molecular function of GPR50 in adipocytes is unclear, but further studies of the relationship between obesity and GPR50 should be of considerable interest. In vivo analysis of expression levels of key adipocytic enzymes in the WAT of Gpr50-/- mice would perhaps allow the delineation of inherent differences in the physiology of this tissue in these animals, and confirm the in vitro data presented here.

117

CHAPTER 5

HYPOTHALAMIC GENE CHANGES DURING TORPOR

118

5.1 INTRODUCTION

The most striking phenotype of the Gpr50-/- mice is that upon fasting, they enter a state of deep torpor, characterised by a severe drop in metabolic rate (VO2) and hypothermia during which core body temperature (Tb) falls to within a few degrees of ambient temperature (Appendix A or Figure 5.1). Torpor is not observed in WT mice even after 48 h of food deprivation, which causes marked depletion of stored energy. This establishes that the torpor bouts expressed by Gpr50-/- mice are not simply driven by the depletion of energy reserves, as these animals are lighter and have reduced WAT (Ivanova et al., 2008). Furthermore, GPR50 expression is reduced in two further models demonstrating a heightened torpor response, namely Siberian hamsters exposed to short photoperiods (Barrett et al., 2006), and ob/ob mice (Appendix A). Thus, the receptor may be implicated in the modulation of thermogenic drive.

The neuroendocrine substrates that mediate torpor initiation, maintenance and arousal remain poorly understood. Decreased circulating leptin concentration is thought to provide a permissive signal for entry into torpor; however, although necessary for the occurrence of this hypometabolic state, simply reducing leptin concentrations is not sufficient to trigger a torpor bout (Freeman et al., 2004).

Postnatal monosodium glutamate (MSG) treatments, which destroy ARC neurons, eliminates photoperiod-dependent torpor in Siberian hamsters (Pelz et al., 2008) and food deprivation-induced torpor in mice (Gluck et al., 2006), emphasising the importance of this hypothalamic area in the control of torpor. Leptin receptors are densely expressed in the ARC of the hypothalamus, and leptin signalling here inhibits neurons that produce NPY/AgRP and stimulates POMC/CART neurons (Schwartz et al., 2000). Reduced leptin concentrations would therefore disinhibit ARC NPY neurons, and turnover and release of NPY would be increased. NPY is a powerful orexigenic transmitter, which also has the capacity to decrease metabolic rate by suppressing SNS outflow to BAT, and decreasing thermogenesis (Levine et al., 2004). ICV injection of NPY or an NPY-Y1 receptor agonist reliably induces torpor-like hypothermia, resembling natural torpor in hamsters (Paul et al., 2005). This can be eliminated by co- injection of NPY and an NPY-Y1 receptor antagonist, suggesting that activation of this receptor is both sufficient and necessary for NPY-induced torpor in hamsters (Dark and Pelz, 2008; Pelz and Dark, 2007).

119

ICV NPY or Y1 receptor agonist treatments are associated with a state of central hypothyroidism characterised by markedly reduced hypothalamic concentrations of TRH mRNA (Dark and Pelz, 2008). The brain is relatively buffered against fluctuations in circulating thyroid hormone (TH), with much of the active T3 in the CNS derived from local uptake of T4 and subsequent conversion by the enzyme Dio2 (Crantz and Larsen, 1980). Regulation of deiodinase activities in tanycytes lining the third ventricle of the hypothalamus, a known site of GPR50 expression, thus serves to control central levels of hypothalamic TH (Fekete and Lechan, 2007). Increased Dio2 increases T3 availability to ARC neurons, leading to increased expression of NPY and decreased expression of TRH in the PVN (Fekete and Lechan, 2007). Repressed TRH expression serves to reduce thermogenic output to BAT (Rogers et al., 2009), and thus increased expression of Dio2 and reduced expression of TRH could be expected during torpor.

Alongside the ependymal layer, GPR50 is highly expressed in the DMH (Appendix A), which has a well-established role in the regulation of thermogenesis, with a polysynaptic connection existing between neurons of the DMH and interscapular BAT (Bamshad et al., 1999; Voss-Andreae et al., 2007), and the DMH also projecting heavily to the PVN (Dimicco and Zaretsky, 2007; Lechan and Fekete, 2006). Intense expression of GPR50 in the ependymal layer of the third ventricle and the DMH means it is ideally placed to influence hypothalamic circuits that gate metabolic and thermogenic responses to changing energy status. The following studies therefore aimed to examine gene expression changes within the hypothalamus of ad libitum fed and fasted (torpid) Gpr50-/- mice compared to fed and fasted (non-torpid) WT mice, with the aim to identify key pathways altered in Gpr50-/- mice, which perhaps underlie the metabolic and torpor phenotype demonstrated by these animals.

5.2 RESULTS

5.2.1 Acute hypometabolism in Gpr50-/- mice upon fasting and selection of animals for analysis of gene expression in the hypothalamus

Gene expression within the hypothalamus of ad libitum fed and fasted (torpid) Gpr50-/- mice compared to fed and fasted (non-torpid) WT mice was examined using a microarray study. Metabolic responses to fasting were assessed in WT and Gpr50-/- mice using indirect calorimetry (Figure 5.1). Fasting caused a decrease in metabolic

120 rate in both genotypes. However, fasted Gpr50-/- mice readily entered a state of deep torpor, demonstrated by an approximate 65% drop in metabolic rate (VO2), at the time of tissue collection (ZT: 2) (Figure 5.1). Hypothalamic blocks were isolated from animals, with fasted Gpr50-/- mice exhibiting stable torpor responses with little evidence of arousal being specifically selected for inclusion in the study (Figure 5.1). 1.5mm coronal slices of brain were cut using a vibrotome, beginning rostrally at the level of the SCN, so that the whole hypothalamus would be included in the slice, finishing caudally at the level of the DMH. Blocks were trimmed to remove tissue dorsal to the hypothalamus and laterally adjacent to the optic tract. Total RNAs were extracted, before conversion of mRNA to cDNA, which were subsequently analysed for transcript expressions using Affymetrix mouse exon arrays. Such recently developed arrays have multiple probes per exon and thus enable two complementary levels of analysis: total gene expression and individual exon expression levels.

5.2.2 Microarray results

The experimental design for the microarray allowed analysis of differentially expressed genes between fed and fasted WT mice, fed and fasted Gpr50-/- mice, fed WT and fed Gpr50-/- mice, and fasted WT and fasted Gpr50-/- mice. Genes with a fold change >1 and statistical significance of q<0.05 (q is a false discovery rate-adjusted P-value) only were included in the resulting gene lists (Appendices B-E).

The largest gene alterations were demonstrated between fed and fasted Gpr50-/- mice, with 354 transcripts significantly upregulated and 428 transcripts downregulated with fasting (Appendix E). Fasting of WT mice revealed the upregulation of 81 genes and downregulation of 64 genes (Appendix D). Comparing gene changes between fasted Gpr50-/- and WT mice allowed assessment of transcript changes during torpor. 73 probes were significantly upregulated during torpor and 97 probes were downregulated during torpor (Appendix B). Interestingly, when ad libitum fed mice were compared, the array results revealed downregulation of only 2 genes in Gpr50-/- mice compared to WT mice (Appendix C). A summary of overall numbers of transcripts altered is given in Figure 5.2.

121

Figure 5.1 Metabolic rates of individual mice used in the microarray study.

Recordings of oxygen consumption (VO2) by indirect calorimetry in the individual WT and Gpr50-/- mice housed in a 12h light:12h dark lighting schedule (as indicated by black and white boxes on the first WT fed graph) fed ad libitum or fasted (time of food removal indicated by the dotted line on the first fasted WT graph) that were used in the microarray study (Chapter 5). Recordings were continued until the time of sacrifice. -/- With fasting, Gpr50 mice enter torpor late at night, as shown by the rapid drop in VO2 below 1 L/hr/Kg, indicated by the blue line.

122

Figure 5.2 Transcript changes in ad libitum fed and fasted WT and Gpr50-/- mice. Affymetrix mouse exon microarrays were used to screen differentially expressed genes in the hypothalamus of fed and fasted WT and Gpr50-/- mice, with tissue collected at a time when the fasted Gpr50-/- mice had entered torpor. Numbers of transcripts with significantly altered expression (q<0.05) are denoted on the histogram.

123

Figure 5.3 Functional clustering of genes with mRNA differences between fasted WT and Gpr50-/- mice. mRNAs extracted from hypothalamic blocks isolated from fed and fasted WT and Gpr50-/- mice were subject to microarray analysis. Significantly altered transcripts (q<0.05) between genotypes with fasting were analysed for GO and KEGG pathway functional enrichment. Groupings with fold enrichment >2 , n>3 and significance p<0.01 are shown, groups of similar functional clustering were excluded; line, fold enrichment of 1. Genes included in the functional clustering groups are listed in Appendix G.

124

Functional clustering analysis was carried out using The Database for Annotation, Visualization and Integrated Discovery (DAVID) version 6.7 resource. Significantly altered transcripts (q<0.05) between genotypes with fasting were analysed for GO and KEGG pathway functional enrichment. Groupings with fold enrichment >2, n>3 and significance p<0.01 are shown in Figure 5.3. Functional clustering analysis revealed an enrichment of “negative transcriptional regulator” genes that were altered during torpor, along with an enrichment of “transcriptional regulation” genes, which were mostly decreased during torpor. This suggests that transcription in the hypothalamus is decreased in torpid mice. An enrichment of “” genes was also demonstrated, indicating alterations of biological rhythms during torpor.

5.2.3 Gpr50 expression in the hypothalamus of Gpr50-/- mice

Analysis of the microarray data, assessing gene changes in the hypothalamus between fasted Gpr50-/- and fasted WT mice (Appendix B), revealed the transcript for Gpr50 to have 2.06-fold increased expression in the Gpr50-/- compared to WT animals, which was obviously unexpected. This ‘upregulation’ is surmised from average expression of four probes along the length of the Gpr50 gene. However, examination of the individual probe data clarifies this apparent contradictory expression of Gpr50 (Figure 5.4, A). In the WT mice, expressions of the four probes were comparable, however in the Gpr50-/- mice, expression of the first probe in exon 1 was considerably downregulated compared to the WT level of expression, whereas the following three probes in exon 2 were upregulated. When averaged, this manifests as an overall increase in gene expression, despite impaired transcription of exon 1. The gene-targeting technique used to create these Gpr50-/- mice obtained from DeltaGen (further explained in the discussion section of this chapter) involved genetic manipulation such that exon 1 and the beginning of exon 2 of Gpr50 would be lost in these animals. This is in line with the data generated from the microarray probe expressions, and upregulation of the downstream probes in exon 2 could be a result of inappropriate transcriptional regulation. There is no evidence of GPR50 protein expression in the Gpr50-/- mice (Figure 5.4, B), using an antibody directed against the C-terminal tail of the receptor (which reveals GPR50 immunoreactivity in the DMH and ependymal layer of the third ventricle in brain sections from WT mice, Figure 5.4, B).

125

Figure 5.4 GPR50 expression in WT and Gpr50-/- mice.

A. Exon expression of Gpr50 in ad libitum fed WT and Gpr50-/- mice. Exons comprising Gpr50 are represented by two grey boxes. Plotted points represent the 4 probes targeting murine Gpr50 on the Affymetrix mouse exon microarray. Purple line is expression in WT and blue line represents expression in Gpr50-/- mice. Error bars are ± SEM, n=4/group. B. GPR50 immunoreactivity in WT murine hypothalamus was observed in the DMH and ependymal cells lining the third ventricle. Positive staining was not observed in Gpr50-/- mice.

126

5.2.4 Hprt1 expression in the hypothalamus of Gpr50-/- mice

The microarray data revealed that under both ad libitum fed and fasted conditions, there was virtually no expression of the transcript encoding Hprt1 in the Gpr50-/- mice, with an approximate 30-fold decreased expression in Gpr50-/- compared to WT animals (appendices B and C). This was confirmed by qPCR analysis of total RNA extracted from hypothalamic tissue blocks, which revealed almost negligible levels of the transcript encoding Hprt1 in Gpr50-/- mice (Figure 5.5, A).

To assess whether this vastly reduced expression of Hprt1 was a consequence of loss of Gpr50, levels of expression were measured in a second line of Gpr50-/- mice, obtained from Organon Pharmaceuticals. Like the mice obtained from DeltaGen and used in the microarray study, the Organon Gpr50-/- mice demonstrate torpor upon fasting, with a comparable depth and temporal profile as the DeltaGen animals (Appendix H). qPCR analysis of Hprt1 expression was repeated on total RNA isolated from hypothalamic tissue blocks from this second line of Gpr50-/- mice (Figure 5.5, B). Unlike the DeltaGen Gpr50-/- mice, the Organon Gpr50-/- mice show no difference in the level of Hprt1 expression when compared to WT animals. This indicates that reduced expression of Hprt1 in the knockouts obtained from DeltaGen was not a consequence of the loss of Gpr50 function in these animals. To ensure reduced expression of Hprt1 in the DeltaGen Gpr50-/- mice does not contribute to the torpor demonstrated by these animals, a collaborative study with Hyder Jinnah at Emory University showed that Hprt1-/- mice do not exhibit altered thermogenic responses to fasting (Figure 5.6).

The added complexity in the genotype of the DeltaGen Gpr50-/- mice meant that further qPCR validations of the microarray results were confirmed in both the DeltaGen and Organon lines of Gpr50-/- mice, to allow the assessment of true gene changes during torpor.

127

Figure 5.5 Hprt1 expression in two lines of Gpr50-/- mice. A-B. Expression of Hprt1 was quantified by qPCR of RNA isolated from hypothalamic tissue blocks from WT and Gpr50-/- mice (A Gpr50-/- mice obtained from DeltaGen (n= 8/group), B Gpr50-/- mice obtained from Organon Pharmaceuticals (n=5/group)). The results from the microarray were confirmed in the DeltaGen mice, with a near total loss (99.2% reduction) of Hprt1 expression in the Gpr50-/- mice compared to WT mice (A). In the Gpr50-/- mice from Organon Pharmaceuticals, there was no significant difference in expression of Hprt1 between Gpr50-/- and WT (B). Data shown are mean ± SEM and normalised to the mouse housekeeping 18S rRNA internal control; ###P< 0.001 significance between genotypes, two-way ANOVA with Bonferroni’s post hoc test.

128

Figure 5.6 Body temperature recordings in fasted WT and Hprt1-/- mice. -/- Core Tb was monitored by remote telemetry in WT and Hprt1 mice subject to a 24 hour fast (period of fasting between grey dotted lines; lighting schedule indicated above with black and white boxes). No large drop in Tb indicative of torpor was observed in the Hprt1-/- mice. Representative recordings of three independent experiments; data kindly provided by Hyder Jinnah at Emory University.

129

5.2.5 Validation of the microarray results by qPCR in the DeltaGen mice qPCR was used to test 7 genes differentially expressed during torpor that were identified by the microarray study (Figure 5.7). The genes selected for validation were chosen based on perceived potential functional importance in hypothalamic signalling during torpor i.e. potential influence on energy sensing or metabolism, along with the genes showing the largest increased or decreased expressions during torpor. All 7 genes showed significant changes of expression levels between fed and fasted WT and DeltaGen Gpr50-/- mice (Figure 5.7, A). In each case, the direction of the response was similar to that observed on the array. Specifically, expressions of Txnip, Per1, Rev- erb alpha and AdipoR1 were all significantly elevated in the DeltaGen Gpr50-/- mice during fasting compared to WT mice, whereas Gabrq, Fabp7, and Osbpl11 all demonstrated significantly decreased expression in the DeltaGen Gpr50-/- mice during fasting compared to WT mice.

5.2.6 Validation of the microarray results by qPCR in the Organon mice

To test if the gene expression changes during torpor indicated by the microarray and validated by qPCR on hypothalamic RNA isolated from the DeltaGen mice were consistent in the second line of Gpr50-/- mice from Organon, further qPCRs were executed on RNA extracted from hypothalami of these animals (Figure 5.7, B). Expressions of Txnip, Per1, Rev-erb alpha and Fabp7 confirmed in these animals; in each case, the direction of response was similar to that seen in the DeltaGen mice. However, no significant differences with fasting between Organon Gpr50-/- and WT mice for AdipoR1, Gabrq, or Osbpl11 were observed.

Per1, Fabp7 (see below), and Txnip (Chapter 6) were therefore selected for further study.

5.2.5 Modulation of period genes in torpor

Alterations in the phase and amplitude of rhythms of core clock machinery have previously been demonstrated in the SCN during torpor (Herwig et al., 2006), and in the present study, Per1 was revealed by the microarray to be upregulated in torpid Gpr50-/- mice compared to normothermic WT animals. Therefore, expressions of Per1,

130

Figure 5.7 Validation of microarray results.

131

Figure 5.7 Validation of microarray results. A-B. Expression of genes selected from the microarray results were quantified by qPCR of RNA isolated from hypothalamic tissue blocks from WT and Gpr50-/- mice (A Gpr50-/- mice obtained from DeltaGen (n= 8/group), B Gpr50-/- mice obtained from Organon Pharmaceuticals (n=5/group)). The results from the microarray were confirmed for all genes selected in the DeltaGen mice. Specifically, Txnip, Per1, Rev- erb alpha, and AdipoR1 were all significantly elevated in the Gpr50-/- mice during fasting compared to the WT mice, whereas Gabrq, Fabp7, and Osbpl11 were all significantly downregulated in the Gpr50-/- mice during fasting compared to the WT mice (A). In the Gpr50-/- mice from Organon Pharmaceuticals, gene expression changes for Txnip, Per1, Rev-erb alpha and Fabp7 also confirmed the results from the microarray, however there were no significant differences with fasting between Gpr50-/- and WT mice for AdipoR1, Gabrq, or Osbpl11 (B). Data shown are mean ± SEM and normalised to the mouse housekeeping 18S rRNA internal control; *P< 0.05, **P< 0.01, ***P< 0.001, significance with fasting compared to ad libitum fed in WT or Gpr50-/- mice; ##P< 0.01, ###P< 0.001 significance between genotypes, two-way ANOVA with Bonferroni’s post hoc test.

132 and its relative Per2, were examined in the brains of fed and fasted WT and Gpr50-/- mice by in situ hybridisation. As in the microarray study, brains were collected from animals kept on a 12:12 light:dark cycle at ZT:2, at a time when fasted Gpr50-/- mice are in deep torpor. In line with previous work (Shearman et al., 1997), within the hypothalamus, intense expression of Per1 was demonstrated in the SCN (Figure 5.8), and further demonstrated considerable expression within several hypothalamic nuclei, namely the ARC, VMH and DMH. Quantification of the hybridisation signal showed Per1 expression in the SCN was significantly increased by fasting in both genotypes (Figure 5.8, A), but expression was significantly higher in the fasted Gpr50-/- mice compared to fasted WTs. In the ARC, VMH and DMH, fasting did not alter levels of Per1 expression in WT mice, however in the Gpr50-/- mice, Per1 expression was significantly upregulated in all areas by fasting (Figure 5.8, B-D). Per2 was strongly expressed in the SCN and was not observed in other areas of the hypothalamus (Figure 5.9). Quantification of the hybridisation signal showed Per2 expression levels were similar between ad libitum fed WT and Gpr50-/- mice, and fasting led to similar increases in gene expression in the two genotypes (Figure 5.9). Probing with sense riboprobes produced no detectable hybridisation signal (data not shown).

5.2.6 Hypothalamic Fabp7 expression during torpor

Another gene selected from the microarray results that showed alteration during torpor in both lines of Gpr50-/- mice by qPCR, was Fabp7. Again, in situ hybridisation was employed to assess where in the hypothalamus this gene is expressed and the areas showing altered expression during torpor. Fabp7 demonstrated widespread expression throughout the brain, but did exhibit an enhanced expression in the SCN (Figure 5.10). Quantification of the hybridisation signal showed Fabp7 expression was similar between ad libitum fed WT and Gpr50-/- mice in the SCN (Figure 5.10) and that fasting caused a significant decrease in Fabp7 expression in the SCN of Gpr50-/- mice compared to fasted WTs, further confirming the microarray and qPCR results. Probing with sense riboprobe produced no detectable hybridisation signal (data not shown).

133

Figure 5.8 Per1 expression in the hypothalamus. Per1 expression in the brain of ad libitum fed and fasted WT and Gpr50-/- mice was examined by in situ hybridisation (n = 5/group), tissue was collected at ZT:2 at a time when fasted Gpr50-/- mice were in deep torpor. Within the hypothalamus Per1 is strongly localised in the SCN and further demonstrates considerable expression within the ARC, VMH and DMH. Quantification of the hybridisation signal showed Per1 expression in the SCN was significantly increased by fasting in both genotypes (A), with expression significantly higher in the fasted Gpr50-/- mice compared to fasted WTs. In the other hypothalamic nuclei, fasting did not alter levels of Per1 expression in WT mice, however in the Gpr50-/- mice, Per1 was significantly upregulated by fasting (B-D). Data shown as mean ± SEM; *P< 0.05, **P< 0.01, ***P< 0.001, significance with fasting compared to ad libitum fed in WT or Gpr50-/- mice, #P< 0.05, ##P< 0.01 significance between genotypes, two-way ANOVA with Bonferroni’s post hoc test.

134

Figure 5.9 Per2 expression in the SCN. Per2 expression in the brain of ad libitum fed and fasted WT and Gpr50-/- mice was examined by in situ hybridisation (n = 5/group), tissue was collected at ZT:2 at a time when fasted Gpr50-/- mice were in deep torpor. Within the hypothalamus Per2 is strongly localised in the SCN. Quantification of the hybridisation signal showed that Per2 expression was similar between ad libitum fed WT and Gpr50-/- mice and fasting caused similar changes in gene expression in both genotypes. Data shown as mean ± SEM; ***P< 0.001, **P<0.01, two-way ANOVA with Bonferroni’s post-hoc test.

135

Figure 5.10 Fabp7 expression in the SCN. Fabp7 expression in the brain of ad libitum fed and fasted WT and Gpr50-/- mice was examined by in situ hybridisation (n = 5/group). Fabp7 demonstrates a widespread general expression throughout the brain, but within the hypothalamus Fabp7 is strongly localised in the SCN. Quantification of the hybridisation signal showed Fabp7 expression was similar between ad libitum fed WT and Gpr50-/- mice in the SCN. Fasting caused a significant decrease in Fabp7 in the SCN of Gpr50-/- mice compared to fasted WTs. Data shown as mean ± SEM; ***P< 0.001, significance with fasting compared to ad libitum fed in WT or Gpr50-/- mice, #P< 0.05 significance between genotypes, two-way ANOVA with Bonferroni’s post hoc test.

136

5.3 DISCUSSION

This work aimed to identify gene expression changes within the hypothalamus during torpor, specifically to determine key physiological pathways altered in Gpr50-/- mice. A relatively small proportion of the transcriptome was altered in the hypothalamus during a torpor bout, and the magnitudes of expression changes in response to torpor were also relatively modest. This observation is consistent with earlier publications using the hibernating golden-mantled ground squirrel, which showed few transcripts in the heart and liver were altered in association with hibernation, and that these changes were generally less than twofold (O'Hara et al., 1999; Williams et al., 2005), and are also congruent with a study of cardiac gene expression changes during torpor in the Siberian hamster, which showed that less than 5% of the transcriptome exhibited altered expression over a torpor bout (Crawford et al., 2007).

The mechanism by which thermogenic drive is altered in the Gpr50-/- mice during torpor remains unclear. Surprisingly, genes involved in known torpor pathways were not significantly altered in the Gpr50-/- mice compared to WT animals (see Chapter 7, Table 1). Although Npy expression increased with fasting in the WT and Gpr50-/- mice, there was no difference in levels of expression between the two genotypes as a consequence of torpor. Likewise, no differences in transcript levels encoding the NPY- Y1 receptor were observed. Furthermore, there were no alterations in expression of components involved in thyroid hormone signalling, including Trh and Dio2, demonstrated between WT and Gpr50-/- mice with fasting. This could be due to the lack of specificity of the hypothalamic areas analysed during the microarray study, whereby opposing changes in different hypothalamic nuclei would nullify a demonstration of altered transcript level, or indeed, a low magnitude change in gene expression in one particular hypothalamic area would be lost during analysis. Thus, a more restricted analysis may have proved useful in the determination of gene expression changes during torpor, or perhaps a focus on sites of GPR50 expression, such as the ependymal layer of cells lining the third ventricle or the DMH, would allow assessment of transcript levels altered specifically as a result of loss of Gpr50.

Curiously, Gpr50 itself appeared in the microarray data as being upregulated in the Gpr50-/- mice compared to the WT mice. This ‘upregulation’ is deduced from an average expression generated from four probes along the length of the Gpr50 gene.

137

However, closer inspection of the individual probe data explains this apparent contradictory expression of Gpr50 (Figure 5.4, A). In the WT mice, expressions of the four probes are comparable, however, for the Gpr50-/- mice, expression of the first probe in exon 1 is considerably downregulated compared to the WT level of expression, whereas the following three probes in exon 2 are upregulated. When averaged, this gives an overall apparent increase in gene expression, despite impaired transcription of exon 1. A gene trapping technique was used by DeltaGen Incorporated to create the Gpr50-/- mice, whereby a construct containing the Escherichia coli LacZ gene encoding β-galactosidase and the neomycin-resistance sequence was inserted into the gene encoding Gpr50, targeting bases 45 to 411, resulting in the loss of exon 1 and the beginning of exon 2. This would be in line with the data generated from the microarray probe expressions, and perhaps the upregulation of expression of exon 2 is a result of inappropriate transcriptional regulation. Southern blotting substantiated the appropriate targeting of the Gpr50 gene (unpublished data, Loudon lab), and confirmation of expression of LacZ via detection of β-galactosidase activity within the ependymal layer of cells lining the third ventricle in knockout mice further validates that the insert did specifically target the Gpr50 gene (Ivanova et al., 2008). Finally, there is no evidence of GPR50 protein expression in the Gpr50-/- mice, when using an antibody directed against the C-terminal tail of the receptor (Figure 5.4, B). Taken together with the demonstration of the same metabolic phenotype by a second line of Gpr50-/- mice (Appendix A), the weight of evidence demonstrates that these animals truly lack a functional protein for GPR50.

Transcripts encoded by Hprt1 and Gabrq were markedly downregulated in the hypothalamus of the DeltaGen Gpr50-/- mice under both ad libitum fed and fasted conditions, but showed no difference in expression in a second line of Gpr50-/- mice from Organon Pharmaceuticals. Hprt1, Gabrq and Gpr50 are all located at the extreme of the q arm of the X- (Figure 5.11); Hprt1 is localised at Xq26.2 and both Gabrq and Gpr50 are localised at Xq28. Therefore, it seems likely that errors in targeting the gene encoding GPR50 during engineering of this knockout mouse are responsible for altered expression of Hprt1 and Gabrq. Although the nature of this complication is unknown, along with Gpr50, it has resulted in an essential knockout of Hprt1 gene expression and marked downregulation of Gabrq. Influence of this genetic abnormality on the phenotype of the Gpr50-/- mouse and knock-on effects on the microarray results cannot be discounted; however, as the second line of Gpr50-/- mice have normal expression of both genes but display the same phenotype as the original 138

Figure 5.11 Chromosomal locations of Gpr50, Hprt1 and Gabrq. Gpr50, Hprt1 and Gabrq are all located at the extreme of the q arm of the X- chromosome. Hprt1 is localised at Xq26.2, and both Gpr50 and Gabrq are localised at Xq28. The start and end of each gene is listed in base pairs (bps). Bands and locations are according to Ensembl.

139 line, it can be concluded that loss of GPR50 alone is responsible for the metabolic and behavioural changes demonstrated by these animals. Further, as Hprt1 and Gabrq do not show a difference with fasting in the Organon and DeltaGen Gpr50-/- mice, it can be assumed that they are not altered during torpor.

It remains possible however, that reduced expression of Hprt1 and Gabrq contribute to the metabolic phenotype of the DeltaGen Gpr50-/- mice. Hprt1 encodes hypoxanthine- guanine phosphoribosyltransferase, an enzyme involved in the generation of purine nucleotides through the salvage synthesis pathway of nucleic acids. Mice with null mutations in this gene exhibit an increase in de novo purine synthesis, but otherwise are phenotypically normal, with only slight defects in striatal dopamine uptake and excess striatal seratonin reported (Bertelli et al., 2009; Jinnah et al., 1994). In humans, mutation of Hprt1 leads to severe neurological defects and the mechanisms underlying why mice deficient in this enzyme do not replicate the human phenotype, remain to be determined. As expression of Hprt1 is almost absent in the DeltaGen Gpr50-/- mice, to ensure loss of its expression is not responsible for the torpor demonstrated by these animals, I collaborated with Hyder Jinnah at Emory University to conduct a fasting study on Hprt1-/- mice. No differences were observed in thermogenic response to fasting between these mice and WT littermates (Figure 5.6). Given that the mice obtained from Organon exhibit normal expression of Hprt1, and the same phenotype as the DeltaGen Gpr50-/- mice, it must be loss of GPR50 and not HPRT1 that is responsible for both the metabolic alterations and susceptibility to torpor demonstrated by these animals.

Gabrq encodes a θ subunit of the GABA-A receptor (Bonnert et al., 1999), which functions as a major mediator of inhibitory neurotransmission in the mammalian CNS. There are numerous subunit isoforms for the GABA-A receptor, which in distinct combinations can alter the ligand-binding and conductance of the holoreceptor (Bollan et al., 2003). The role of the θ subunit remains understudied at this time. Available knockout animals examined with a number of tests, only appear to differ from their WT counterparts by presenting with a significant decrease in prepulse inhibition (PPI), as evaluated by a series of behavioural tasks (MGI, 2011). PPI is the normal suppression of a startle response when an intense stimulus is preceded by a weak non-startling prestimulus, which could mean that Gabrq has some function in sensorimotor gating within the CNS. Despite substantial downregulation in the Gpr50-/- mice obtained from 140

DeltaGen, as the second line of Gpr50-/- mice show normal expression of Gabrq, and exhibit the same phenotype as the DeltaGen Gpr50-/- mice, it can again be inferred that loss of GPR50 and not GABRQ is responsible for the altered phenotype seen in these animals.

To establish gene expression changes within the hypothalamus during torpor, I selected six genes from the data generated by the microarray analysis for further validation, including Txnip, which demonstrated the most increased expression during torpor, along with the clock genes Per1 and Reverb-alpha, and three genes potentially involved in lipid metabolism: AdipoR1, Fabp7 and Osbpl11. qPCRs on total hypothalamic RNA collected from the DeltaGen and Organon Gpr50-/- mice, confirmed that in the DeltaGen mice, in each case, the direction of response was similar to that observed on the array. In the Organon mice, Txnip, Per1, Reverb-alpha and Fabp7 were altered by torpor in a similar manner to the DeltaGen animals; however although showing a trend towards the same changes, AdipoR1 and Osbpl11 alterations in the Organon mice failed to reach significance. This is perhaps not surprising, considering the relatively subtle differences in their expressions demonstrated during torpor in the DeltaGen mice, with only fold changes of 1.21 for AdipoR1 and -1.27 for Osbpl11 revealed by the microarray data. Possibly, a larger sample size may have seen these genes significantly altered in the Organon mice also, but nonetheless they were not considered for further investigation. On the other hand, as Txnip, Fabp7, Per1 and Reverb-alpha were altered during torpor in the two lines of Gpr50-/- mice; they can be credited as true transcriptional variances during hypometabolism. The altered expression of Txnip shall be further discussed in Chapter 6, and changes in Fabp7, Per1 and Reverb-alpha expressions will be addressed below.

5.3.1 Fabp7 expression in the hypothalamus is decreased during torpor

In both lines of Gpr50-/- mice, Fabp7 was significantly downregulated in the hypothalamus during fasting-induced torpor. This gene encodes the brain-type fatty acid binding protein, which belongs to a conserved multigene family of the intracellular lipid-binding proteins (Storch and Thumser, 2000). Fabp7 is capable of binding long- chain polyunsaturated fatty acids (PUFAs) and therefore may be involved in the uptake, transport, and solubilisation of these hydrophobic ligands, targeting them to specific metabolic pathways (Balendiran et al., 2000).

141

The spatio-temporal expression pattern of Fabp7 has been extensively described in the developing rodent brain, showing specific localisation to radial glial cells and immature astrocytes (Kurtz et al., 1994). Increasing mRNA levels correlate with specific developmental processes including establishment of the radial glial fibre system and neurogenesis (Owada et al., 1996). However, the function of this protein in the adult brain is not well understood, and thus renders speculation of the precise role of its decreased expression during torpor difficult. Expression of Fabp7 in the adult primate hippocampus has been shown to initially decrease and then increase steadily following ischemic insult, where it is proposed to influence astrocyte proliferation (Ma et al., 2010). Interestingly, the transcript encoding glial fibrillary acidic protein (GFAP) was shown by the microarray to be increased during torpor in Gpr50-/- mice (Appendix B), which could imply increased astrocyte proliferation in the hypothalamus, concomitant with decreased Fabp7 expression.

In mice, Fabp7 has been shown to be diurnally regulated within the hypothalamus (Gerstner et al., 2006), and subsequently been shown to exhibit synchronised cycling throughout the brain (Gerstner et al., 2008). mRNA levels peak at the beginning of the light phase, followed by a reduction of relative expression throughout the light period, with lowest levels following lights-off. Whether Fabp7 expression is tied to known clock mechanisms is unknown, but bioinformatic analysis detected the presence of peroxisome proliferator activated receptor (PPRE) and E-box elements within the proximal Fabp7 promoter (Gerstner et al., 2008), which have known roles in clock- related transcriptional regulation. In this work, similar to previous reports (Gerstner et al., 2008; Gerstner et al., 2006), in situ hybridisation showed low levels of Fabp7 throughout the murine brain, and within the hypothalamus, Fabp7 strongly localised in the SCN. Gpr50-/- mice express similar levels of Fabp7 in the SCN as WT mice when fed ad libitum, however, upon fasting when these animals are induced to enter torpor, levels of Fabp7 within the SCN were significantly decreased compared to WT mice. As Fabp7 mRNA is known to be diurnally regulated and torpor is known to alter the phase and amplitude of circadian rhythms (discussed further in section 5.3.2), Fabp7 may represent a diurnally regulated component whose phase or amplitude is likewise altered during torpor.

Suppression of Fabp7 expression during torpor occurs not only in the SCN, but generally throughout the brain (although this was not quantified); perhaps indicating 142 that Fabp7 is intimately related to the maintenance of the torpor process. Regulation of Fabps during hibernation has previously been examined. Study of hibernating ground squirrels and the little brown bat demonstrate increased expression of the heart- and adipose-type Fabps in BAT and increased expression of heart-type Fabp in skeletal muscle during hibernation (Eddy and Storey, 2004; Hittel and Storey, 2001). PPAR transcription factors, along with the PPAR coactivator PGC-1α promote the expression of Fabps, and concurrent increases in both PPARγ and PGC-1α expression during hibernation in BAT and muscle of the little brown bat were also demonstrated (Eddy and Storey, 2003). In contrast to other organs studied, expressions of PPARγ and PGC-1α in the brain were suppressed during hibernation in the bats, which may lead to suppression of Fabps within this tissue, namely Fabp7, although this was not assessed. In the present study, PGC-1α was increased in the hypothalamus of Gpr50-/- mice during torpor, although was not significantly different when compared to fasted WT mice. PPARγ expression was low in all groups, suggesting that different mechanisms regulate Fabp7 in the mouse brain.

During torpor, a gradual switch from carbohydrate to lipid metabolism is described (Heldmaier et al., 1999). However, in order to conserve energy, overall ATP-consuming processes are downregulated. Liver mitochondrial respiration rate in Siberian hamsters is reduced by 30-70% during torpor (Brown et al., 2007), and in mice, mitochondrial respiration is similarly suppressed during fasting-induced torpor (Brown and Staples, 2010). These reductions were due to decreased capacity for substrate oxidation. It could be hypothesised that although becoming the primary fuel source during torpor, net oxidation of lipids within mitochondria is reduced. It is thus possible that reduced Fabp7 expression plays a role in reducing mitochondrial fatty acid oxidation during torpor, via decreased transport of lipids to mitochondria. Previous studies also point to a functional link between body lipid composition and torpor. Acclimation to cold is generally accompanied by increased unsaturated fatty acids and decreased saturated fatty acids in tissues and cell membranes (Munro and Thomas, 2004) in order to maintain cell membrane function at varying Tb. Speculatively, reduction of Fabp7 expression could have a function in the regulation of membrane phospholipid composition during the hypothermia of torpor.

Despite such generalised theories, the pronounced expression of Fabp7 in the SCN implies some specific role for this protein in this area of the hypothalamus. 143

Interestingly, ectopic overexpression of murine or Drosophila Fabp7 in Drosophila resulted in increased net sleep in these animals, implicating it as a novel molecular player in sleep consolidation (Gerstner et al., 2011). The SCN controls circadian rhythms of sleep and wakefulness via projections to the DMH (Saper et al., 2005). Originally thought to be an extension of the normal rest phase, it has subsequently been shown that sleep debt actually accumulates during torpor as it does during prolonged waking (Palchykova et al., 2002), and thus perhaps a low level of Fabp7 during torpor has some role to play in this process.

5.3.2 Clock gene expression in the hypothalamus during torpor

Among the genes significantly altered during torpor were components of the circadian clock; Per1 and Rev-erb alpha. The clock’s molecular machinery still oscillates during torpor, but alterations in phase and amplitude of these rhythms have previously been shown during photoperiod-induced torpor in Siberian hamsters (Herwig et al., 2006; Herwig et al., 2007). These studies reported increased expression of Per1 in the SCN during torpor, which reached significance during arousal from torpor. The present work similarly demonstrated increased levels of Per1 in the SCN; however, this significantly elevated expression was observed during the deep hypometabolism of the torpor bout itself. This temporal difference may be species-related, as mice were used instead of Siberian hamsters, or may be due to the timing of the torpor bout, with the fasted mice entering torpor during the dark phase and arousing in the middle of the light phase, whereas the Siberian hamsters showed entry into torpor soon after lights on, arousing just prior to lights off. Finally, the temporal difference in increased Per1 expression demonstrated between mice and hamsters may be a factor of fasting-induced as opposed to spontaneous torpor. Indeed, fasting WT mice increased expression of Per1 and Per2 in the SCN, indicating that altered energy status influences the expression of these clock genes. Previous reports demonstrate that short-term fasting leads to increased Per1 expression in peripheral tissues, without affecting the phase of the circadian rhythm (Kawamoto et al., 2006; Kobayashi et al., 2004), suggesting that this circadian regulator is readily influenced by feeding and perhaps has a role in some aspect of metabolism to maintain energy balance. Calorie restriction has been shown to entrain the central clock and a restricted feeding schedule has been shown to phase shift Per2 expression in the SCN (Froy, 2007); however the effects of fasting per se on the expression of central clock genes has not previously been reported.

144

Examination of Per1 expression in the peripheral organs of Siberian hamsters during torpor demonstrated its elevated expression during arousal in the heart and liver (Crawford et al., 2007). Altered expression of Per1 therefore occurred concurrently in the SCN and periphery, in contrast to normothermic animals that show a lag in rhythmic gene expression in peripheral tissues behind that of the SCN by approximately 4-6 h (Balsalobre et al., 1998). Similarly, in the present work using torpid Gpr50-/- mice, simultaneous increases in expression of Per1 during torpor appeared in a number of hypothalamic nuclei outside the SCN, with higher levels in the ARC, VMH and DMH also demonstrated.

The work using torpid Siberian hamsters (Herwig et al., 2006) further identified an increase in Bmal1 expression in the SCN during torpor; however this was not confirmed by the present microarray study. This is perhaps due to reduced specificity of the area analysed, with the inclusion of extra-SCN hypothalamic tissue perhaps negating the change shown in previous studies. The work using Siberian hamsters further demonstrated downregulation of Bmal1 expression in the SCN on the day following a torpor bout (Herwig et al., 2006). In the current study, increased expression of Reverb-alpha, a repressor of Bmal1 expression (Preitner et al., 2002), was observed in the hypothalamus of torpid mice, and a similar upregulation in Siberian hamsters may precede the downregulation of Bmal1 expression on the day following torpor. Further consideration of clock genes in peripheral tissues of Siberian hamsters (Crawford et al., 2007) revealed a tendency for increased Reverb-alpha expression during torpor in the heart and liver, which reached significance in the lung, and thus, increased expression in the SCN of these animals seems likely.

The increases in Per1 and Reverb-alpha expressions in the hypothalamus during torpor may be due to directed upregulation of their gene expression. As previous work revealed that simultaneous increases in expression of these genes occur in peripheral tissues (Crawford et al., 2007), this could suggest a regulation by either low Tb or by some common neurotransmitter/neuroendocrine-mediated pathway. Feedback of hypothermia on the clock has been proposed, by the prediction that the period of circadian rhythms shortens as temperature decreases (Aschoff, 1979) and that torpor shortens period in Siberian hamsters in constant conditions (Thomas et al., 1993).

Moreover, in other species that show longer periods of hypothermia and lower Tb, the shortening of the freerunning rhythm is of much stronger magnitude (Lee et al., 1990), 145 implying a potential feedback on the clock that is directly temperature-dependent, increasing with decreasing Tb.

Interestingly, a more recently identified clock gene, Dec2 (BHLHE41) (Honma et al., 2002), which is rhythmically expressed in the SCN and peripheral tissues (Noshiro et al., 2004), is also upregulated in the Gpr50-/- mice during torpor, as indicated by the microarray. Dec2 may thus represent a further clock component whose phase or amplitude is altered during torpor. Interestingly, during hypoxic conditions, Dec2 expression is induced by Hypoxia-Inducible Factor-1α (HIF-1α), and represses transcription of fatty acid synthase (FAS) by inhibiting the activity of its upstream transcription factor sterol regulatory element binding protein-1c (SREBP-1c) (Choi et al., 2008; Guillaumond et al., 2008). Dec2 inhibits SREBP-1c induced transcription via interaction with SREBP-1c and competitively binding E boxes in the FAS promoter. Therefore, increased Dec2 during torpor could reflect a response to perceived hypoxia as a result of decreased respiratory rate during this state of hypometabolism, and may have a role in reducing ATP-consuming processes such as FA synthesis.

146

CHAPTER 6

TXNIP: A NOVEL TORPOR-INDUCED GENE

147

6.1 INTRODUCTION

The ability to enter a hypometabolic state such as that achieved during torpor is pivotal to the survival of some animals; balancing the energy budget when food is not available and/or demand for heat production to maintain homeothermy becomes excessive. The hypothalamus is a major centre of convergence and integration of multiple nutrient-related signals important in the regulation of energy homeostasis, with subsets of specialised neurons regulating behavioural and metabolic effectors of energy expenditure, glycaemic control and lipid metabolism (Blouet and Schwartz, 2010). Gene expression microarray experiments (Chapter 5), revealed that Txnip showed the most increased expression in the hypothalamus of fasted Gpr50-/- mice, which entered torpor, compared to fasted WT mice, which remained normothermic, suggesting that this protein plays an important role during torpor.

Thioredoxin-interacting protein (Txnip) is an endogenous negative regulator of thioredoxin, a ubiquitously expressed antioxidant (Nishiyama et al., 1999). Emerging data marks Txnip as an important regulator of cellular glucose and fatty acid metabolism. Both naturally occurring and engineered Txnip-null mice present with a metabolic profile similar to that of a fasted animal; mice are hypoglycaemic, ketotic and have increased levels of plasma FFAs, LDL and VLDL cholesterol (Oka et al., 2006; Sheth et al., 2005). However, they are also hyperinsulinaemic, and exhibit a 40% increase in their fat to muscle ratio, and gain more weight when fed a high-fat diet due to increased food intake and increased adipogenesis (Chutkow et al., 2010; Sheth et al., 2005). Txnip-null mice further exhibit increased hepatic lipogenesis along with elevated triglyceride and cholesterol ester stores (Donnelly et al., 2004).

During fasting, mice lacking Txnip demonstrate a deterioration of FA utilisation, and do not survive a 48 h fast. Autopsy revealed significant bleeding in the lower gastro- intestinal tract in these fasted animals, along with marked liver steatosis and renal failure (Oka et al., 2006). Further investigation indicated that acetyl-CoA consumption is reduced, suggesting inefficient function of the Krebs cycle and/or the electron transport chain in Txnip-deficient mice. Studies have since demonstrated impaired mitochondrial oxidation of all major fuels (glucose, FAs and ketone bodies) in the soleus muscle and heart of Txnip-null mice alongside increased glucose uptake by these tissues (Hui et al., 2008). Impaired mitochondrial fuel oxidation means the

148 surplus glucose influx is channelled into glycogen storage, and glycogen accumulates in heart and soleus of fasting Txnip-null mice (Andres et al., 2011). Although mitochondrial glucose oxidation is impaired in mice lacking Txnip, the energy state of cells is paradoxically elevated, with the finding that fasted Txnip-null mice have diminished cellular AMP levels in the heart and soleus muscles, along with blunted activation of AMPK (Andres et al., 2011). An elevated energy state in these fasted animals may be attributable to an increase in glycolysis, which has additionally been reported in mice lacking Txnip (Hui et al., 2008).

Txnip is intimately linked to glycolysis and mitochondrial oxidative phosphorylation through the glucose sensing transcriptional complex MondoA:Mlx. In response to high intracellular glucose, in particular its metabolites glucose-6-phosphate (G6P) and glyceraldehyde-3-phosphate (GADP), MondoA:Mlx accumulate in the nucleus and drives the transcription of Txnip through binding carbohydrate response elements (ChoREs) in its promoter (Stoltzman et al., 2008; Yu et al., 2010). Increasing glycolytic flux by blocking oxidative phosphorylation within cells results in downregulated Txnip expression due to exhaustion of glycolytic intermediates (Yu et al., 2010). Along with glucose, Txnip expression can also be induced by lactic acidosis, which presumably acts as a feedback mechanism to reduce glucose uptake and lactate production (Chen et al., 2010). Txnip expression is responsive to of a number of physiological stimuli and regulators; it is induced by glucocorticoid (Chen et al., 2011; Wang et al., 2006), vitamin D (Chen and DeLuca, 1994), PPARα and PPARγ ligands (Billiet et al., 2008; Qi et al., 2009; Rakhshandehroo et al., 2007), as well as stress-related stimuli such as

H2O2, UV, reactive-oxygen species (ROS), heat-shock and hypoxia (Junn et al., 2000; Karar et al., 2007; Kim et al., 2004; Le Jan et al., 2006). Expression is also induced by fasting, caloric restriction (Swindell, 2009), and high-fat diet (Dreja et al., 2010); which are energetic challenges that result in decreased Gpr50 expression (Ivanova et al., 2008). Conversely, Txnip expression can be inhibited by insulin (Parikh et al., 2007; Shaked et al., 2009), leptin (Blouet and Schwartz, 2011) and glutamine (Kaadige et al., 2009).

The regulation of Txnip expression by such diverse stimuli and its proposed role in directing glucose and FA utilisation, places this molecule centrally in a mechanism whereby cells can integrate different signalling pathways to control fuel partitioning and utilisation. Based on a potential functional importance of the observed regulation of 149

Txnip expression during torpor, this chapter aims to determine its localisation within the hypothalamus and explore its expression in other tissues, as well as its relationship with GPR50 and the response of Txnip to other thermogenic/metabolic drives (cold and glucopenia).

6.2 RESULTS

6.2.1 Txnip expression in the hypothalamus

Txnip protein in the murine hypothalamus was limited to the ependymal cells lining the third ventricle (Figure 6.1, A-C), and was not observed in sections incubated with secondary antibody alone (Figure 6.1, D). The morphology and anatomical location of the staining is suggestive of ciliated ependymal cells. Immunoreactivity observed in blood vessels and the median eminence (Figure 6.1, A) was due to non-specific labelling by the α-mouse IgG secondary antibody (Figure 6.1, D).

6.2.2 Txnip expression in the hypothalamus in response to fasting and torpor

A gene expression microarray experiment (Chapter 5) revealed Txnip as the most increased transcript in the hypothalamus of Gpr50-/- mice during torpor. This was confirmed in both lines of Gpr50-/- mice by qPCR. In situ hybridisation on mouse brain sections was employed to determine where altered expression of Txnip occurred (Figure 6.2). Txnip localised to the lateral ventricles and the choroid plexus. Within the hypothalamus, Txnip expression was limited to the ependymal cells lining the third ventricle. Probing with sense riboprobe produced no detectable hybridisation signal (data not shown). Txnip expression was quantified by measuring optical density of autoradiographic films within the ependymal region of the third ventricle and the parenchyma of the hypothalamus (whole hypothalamic area minus the ependymal region). Quantification of the hybridisation signal showed Txnip expression was similar in ad libitum fed WT and Gpr50-/- mice in both the ependyma of the third ventricle (Figure 6.2, A) and the parenchyma of the hypothalamus (Figure 6.2, B). Within ependymal cells, fasting caused a significant increase in Txnip expression in both genotypes, but induction of Txnip in the Gpr50-/- mice was profoundly higher than in fasted WT mice (Figure 6.2, A; 3.7 fold increase in Txnip expression in Gpr50-/- mice vs 1.6 fold increase in WT mice). Txnip showed a similar pattern of expression with fasting in the parenchyma of the hypothalamus (Figure 6.2, B).

150

Figure 6.1 Txnip expression in the hypothalamus.

A-D. Txnip immunoreactivity in the murine hypothalamus was limited to the ependymal cells lining the third ventricle (A-C) and the median eminence (A). Positive staining of the ependymal layer was not seen when no primary antibody was included (D), but blood vessels were labelled by the secondary α-mouse antibody, including those of the median eminence . Blue = DAPI counterstaining.

151

Figure 6.2 Txnip expression in the hypothalamus.

Txnip expression in the brain of ad libitum fed and fasted WT and Gpr50-/- mice was examined by in situ hybridisation (n = 5/group). Txnip was expressed within the ependyma of the lateral ventricles and the choroid plexus. Within the hypothalamus, Txnip expression was pronounced in the ependymal cells lining the third ventricle. Quantification of the hybridisation signal showed Txnip expression was similar between fed WT and Gpr50-/- mice in both the ependyma of the third ventricle (A) and in the whole hypothalamic area minus the ependymal area (B). Fasting caused a significant increase in Txnip expression in both genotypes within the ependyma of the third ventricle (A), with expression significantly higher in the fasted Gpr50-/- mice. Txnip levels showed a similar pattern of expression with fasting in the whole hypothalamus minus the ependyma of the third ventricle (B). Data shown are mean ± SEM; *P< 0.05, ***P< 0.001, significance with fasting compared to ad libitum fed in WT or Gpr50-/- mice, #P< 0.05 significance between genotypes, two-way ANOVA with Bonferroni’s post hoc test.

152

6.2.3 Txnip expression in peripheral tissues in response to fasting and torpor

As Txnip demonstrated widespread peripheral expression (Figure 6.3, A), and expression appears to be responsive to energy status within the hypothalamus, qPCR was used to assess transcript levels in some metabolically important tissues in WT and Gpr50-/- mice during fasting. Txnip expression was similar in ad libitum fed WT and Gpr50-/- mice in the liver, WAT and BAT (Figure 6.3, B-D). Fasting did not alter Txnip expression in the liver or WAT of WT mice, but significantly increased its expression in BAT. In contrast, Txnip expression was significantly increased by fasting in Gpr50-/- mice in all tissues studied. Interestingly, Txnip demonstrated a trend to correlate with -/- the Tb of Gpr50 animals at the time of tissue collection, showing higher expression the lower the body temperature (Figure 6.3, E-G). Thus, Txnip expression may be responsive to fluctuations in Tb. Furthermore, altered expression of Txnip could be directly influenced by the torpor phenotype, leading us to analyse expression in WT mice driven into torpor (section 6.2.6 and 6.2.7).

Changes in protein expression of Txnip in the hypothalamus and BAT from ad libitum fed and fasted WT and Gpr50-/- mice, were assessed by Western blot analysis (Figure 6.4). Like Txnip mRNA, protein levels were comparable between genotypes in the fed state in both the hypothalamus and BAT. In response to fasting, levels of Txnip were greatly increased in the hypothalami of Gpr50-/- mice with a trend to an increase in the WT mice. In BAT, fasting increased levels of Txnip protein in both genotypes, however expression was higher still in the fasted Gpr50-/- mice.

6.2.4 Gpr50 expression in peripheral tissues in response to fasting

Like Txnip, expression of GPR50 has been demonstrated in the ependymal layer of the third ventricle, and is responsive to energy status; specifically, food restriction causes a significant decrease in Gpr50 expression, which recovers upon refeeding (Ivanova et al., 2008). Again, similarly to Txnip, Gpr50 demonstrated a widespread expression profile in mouse tissues (Figure 6.5, A), indicating the potential for interplay of the proteins in vivo. Consequently, expression of Gpr50 in response to fasting in WT mice was analysed by qPCR. Confirming previous reports, expression of Gpr50 was significantly reduced in the hypothalamus by fasting (Figure 6.5, B). Gpr50 expression

153

Figure 6.3 Peripheral Txnip expression in fed and fasted WT and Gpr50-/- mice. A. RT-PCR profiling of Txnip mRNA expression in mouse tissues. B-D. Txnip expression was quantified by qPCR in WT and Gpr50-/- mice (n = 4/group). Txnip expression was similar between ad libitum fed WT and Gpr50-/- mice in the liver (B), WAT (C) and BAT (D). Fasting did not alter Txnip expression in WT mice in the liver or WAT (B, C), but significantly increased expression in BAT (D). Txnip expression was significantly increased by fasting in all tissues in Gpr50-/- mice (B-D). Individual fold- expressions of Txnip are plotted against Tb at the time of tissue collection for the fed and fasted Gpr50-/- (knockout; KO) mice (E-G). Data shown as mean ± SEM, normalised to the mouse housekeeping 18S rRNA internal control, and as fold change vs WT fed; *P< 0.05, **P< 0.01, ***P< 0.001, significance with fasting compared to ad libitum fed in WT or Gpr50-/- mice, ##P< 0.01, ###P< 0.001 significance between genotypes, two-way ANOVA with Bonferroni’s post hoc test.

154

Figure 6.4 Txnip expression in the hypothalamus and BAT. Total protein extracted from hypothalamic tissue blocks and interscapular BAT from ad libitum fed and fasted WT and Gpr50-/- mice, was analysed by western blotting using α- Txnip antibody. Densitometric analysis showed Txnip expression in the hypothalamus was comparable between genotypes when fed ad libitum, however, with fasting, expression of Txnip was significantly increased in Gpr50-/- mice with a trend to an increase in the WT mice that failed to reach significance. Similarly, Txnip expression in BAT was comparable between genotypes in the fed state, and increased in both with fasting, however Txnip expression increased significantly more in Gpr50-/- mice with fasting compared to WTs. *P< 0.05, **P< 0.01, ***P< 0.001, significance with fasting compared to ad libitum fed in WT or Gpr50-/- mice, #P< 0.05, ##P< 0.01 significance between genotypes, two-way ANOVA with Bonferroni’s post hoc test.

155

Figure 6.5 Gpr50 expression in fasted WT mice. A. RT-PCR profiling of Gpr50 mRNA expression in mouse tissues. B-E. Gpr50 expression was quantified by qPCR in WT mice (n = 4/group). Fasting caused a significant reduction in Gpr50 expression in the hypothalamus (B), liver (C), and WAT (D), and significantly increased expression in BAT (E). Expression of Dio2 (F) and UCP1 (G) were also analysed in BAT of ad libitum fed and fasted WT and Gpr50-/- mice. Data shown are mean ± SEM as fold change relative to WT fed; *P< 0.05, **P< 0.01, ***P< 0.001 students’ t-test (B-E) two-way ANOVA with Bonferroni’s post-hoc test (F-G).

156 was likewise markedly reduced in the liver and WAT of fasted mice (Figure 6.5, C-D). Conversely, expression of Gpr50 was approximately doubled in BAT in response to fasting (Figure 6.5, E). Therefore, it seems that global alteration in Gpr50 expression occurs in response to fasting, albeit in a tissue-dependent manner. As changes in Txnip and Gpr50 expression do not parallel one another (i.e. fasting increased Txnip and decreased Gpr50 expression in the hypothalamus, but increased levels of both transcripts in BAT), it is unlikely that Txnip expression is directly responsive to GPR50 signalling. This also suggests that the rise in Txnip expression in torpid Gpr50-/- mice is reflective of the torpid state. This will be further tested below by modulating ambient temperature to induce torpor in WT mice.

6.2.5 Gpr50 and Txnip expression following acute cold exposure

The literature suggests that WT mice can be driven into torpor by fasting at lower ambient temperature (Hudson and Scott, 1979). Before testing whether similar changes in Txnip expression are demonstrated in torpid WT mice as in torpid Gpr50-/- mice, the influence of cold ambient temperature on expression of Txnip (and Gpr50) was first evaluated. WT and Gpr50-/- mice were acutely exposed to low temperature (4°C) for 2 h. Cold challenge did not alter the expression of Gpr50 in the hypothalamus, liver or WAT (Figure 6.6, A-C); however expression was significantly reduced in BAT (Figure 6.6, D). As Gpr50 expression is oppositely regulated by fasting in BAT compared to all other tissues studied, and again is distinctly altered in BAT following cold challenge, some discrete regulation in this tissue may be inferred.

Cold exposure did not alter Txnip expression in the hypothalamus or liver (Figure 6.6, E-F), but significantly increased expression in WAT in both WT and GPR50-/- mice (Figure 6.6, G) and significantly decreased expression in BAT in both genotypes (Figure 6.6, H). The induction of Txnip in WAT may be reflective of lipolytic involvement of this tissue during both a fast and cold exposure. Further, opposing regulation in BAT in response to fasting and cold exposure, may point to altered metabolic activity of this tissue in response to these challenges, whereby thermogenesis in BAT is reduced during a fast and induced in response to cold.

157

Figure 6.6 Gpr50 and Txnip expression following acute cold exposure. WT and Gpr50-/- mice were exposed to 4°C or maintained at 20-22°C (RT; room temperature) for 2 h followed by qPCR analysis of Gpr50 (A-D) and Txnip (E-H) expression. Cold challenge did not alter the expression of Gpr50 in the hypothalamus, liver or WAT (A-C), but was significantly reduced levels in BAT (D). Txnip showed a significant increase in expression in WAT (G) and decreased expression in BAT (H) of both WT and Gpr50-/- mice, but no change in hypothalamus or liver (E, F). Expressions of Dio2 and UCP1 were significantly induced in both genotypes following cold exposure (I-J). Data shown are mean ± SEM and normalised to mouse housekeeping 18S rRNA internal control, fold change relative to RT-housed WT animals; a. *P< 0.05, students’ t-test; b. **P<0.01, ***P<0.001, two-way ANOVA with Bonferroni’s post hoc test.

158

Finally, comparable levels of Ucp1 and Dio2 in BAT of ad libitum fed WT and Gpr50-/- animals were observed, alongside typical changes (Sivitz et al., 1999) in expression of these transcripts in response to fasting (Figure 6.5, G-E). Further, expected changes (de Jesus et al., 2001) in BAT expression of UCP1 and Dio2 were observed in cold challenged WT and Gpr50-/- mice (Figure 6.6, I-J). This indicates that a modified thermogenic output of BAT is not responsible for the altered metabolic responses demonstrated by mutant animals when fed ad libitum and fasted.

6.2.6 Gpr50 and Txnip expression following glucoprivation

As both Gpr50 and Txnip expressions are responsive to altered nutrient availability, transcript levels in tissues of WT mice treated with 2-deoxyglucose (2DG), a glucose mimetic which induces a state of perceived hypoglycaemia, were assessed by qPCR. 2DG treatment significantly increased the expression of Gpr50 in the hypothalamus, liver, WAT and BAT (Figure 6.7, A-D). This indicates that some difference between fasting and 2DG ‘hypoglycaemia’ results in augmented Gpr50 expression in response to these challenges. 2DG treatment similarly increased Txnip expression in all tissues studied, apart from the liver, which showed no change in expression (Figure 6.7, E-H). The lack of response in the liver here suggests strongly that changes observed in Txnip expression during fasting do not reflect a passive response to Tb reduction (Figure 6.7, I-J).

6.2.7 Txnip expression is increased by fasting-induced torpor

Changes in Txnip expression were assessed in a further model of torpor; food- restricted female WT mice can be induced to enter torpor by housing at a marginally reduced ambient temperature (16 °C). To ensure there were no sex differences in the regulation of Txnip expression, qPCR was used to measure expression in tissues of ad libitum fed and fasted female Gpr50-/- mice. As in male Gpr50-/- mice, fasting significantly increased Txnip expression in the hypothalamus, liver, WAT and BAT of torpid female Gpr50-/- mice (Figure 6.8, A-D). Within the group of fasted animals, one mouse did not enter, or was aroused from torpor (Tb Figure 6.8, E), and Txnip levels in the tissues of this individual remained comparable to those that were ad libitum fed. This suggests that it is a condition of the torpid state per se which drives Txnip expression.

159

Figure 6.7 Gpr50 and Txnip expression following 2-deoxyglucose treatment. WT mice were treated with x mg/kg of 2-deoxyglucose (2DG) for one hour. qPCR analysis of Gpr50 expression (A-D) and Txnip expression (E-H) showed significant increases in expression of both genes in all tissues studied, apart from Txnip in the liver (F) which showed no change in expression. Txnip expressions of individual mice plotted against Tb at the time of tissue collection (I-L). Data shown are mean ± SEM, normalised to the mouse housekeeping 18S rRNA internal control, and as fold change relative to vehicle treated animals; *P< 0.05, students’ t-test.

160

Figure 6.8 Txnip expression in female Gpr50-/- mice. A-D. Txnip expression was quantified by qPCR in tissues of female Gpr50-/- mice fed ad libitum or fasted (n= 4/group). As in male Gpr50-/- mice, fasting induced a state of torpor, as assessed by body temperature (E), except in one animal, which did not enter, or was aroused from torpor. Torpor caused significant increases in Txnip expression in all tissues studied, whereas in the fasted non-torpid (NT) animal, levels of Txnip were comparable to those in the fed group. Data shown are mean ± SEM, normalised to the mouse housekeeping 18S rRNA internal control, and as fold change relative to fed animals; *P< 0.05, **P< 0.01, ***<0.001 one-way ANOVA with Bonferroni’s post-hoc test.

161

Because no sex differences in the regulation of Txnip expression were apparent, expression was examined by in situ hybridisation in the brain of ad libitum fed or fasted WT female mice acclimated to 16°C ambient temperature. Of the 8 fasted females, 6 entered torpor, as assessed by indirect calorimetry (VO2 < 1000ml/h/Kg) and Tb at time of sacrifice. Induction of torpor caused a marked elevation in Txnip expression within the brain, compared to ad libitum fed and non-torpid fasted mice (Figure 6.9). Quantification of the hybridisation signal revealed no significant increase in Txnip expression in the non-torpid mice within the ependyma of the third ventricle or the entire hypothalamic area minus the ependymal area (Figure 6.9, A-B), unlike the pronounced increase in Txnip expression seen in the torpid animals. Individual Tb and fold-expression of Txnip are shown in Figure 6.9, C-D. Examination of Txnip expression in liver, WAT and BAT revealed substantial increases in torpid mice (Figure 6.10, A-C), with no significant increases in non-torpid fasted animals. Together, this data strongly demonstrates upregulation of Txnip expression in tissues during torpor.

6.2.8 Txnip expression in the hypothalamus of the Siberian hamster

In the previous experiments, torpor was induced in mice by cold exposure and/or fasting, with Txnip expression showing similar changes in WT and Gpr50-/- mice driven into torpor by these mechanisms. To assess whether Txnip is directly linked to torpor, irrespective of the mechanism used to induce this hypometabolic response, quite a different model, the seasonal Siberian hamster was selected for investigation of Txnip expression during torpor. Riboprobe specific to Siberian hamster Txnip was cloned and used for in situ hybridisation to measure expression in the brains of animals housed on long photoperiod and short photoperiod (a time when Siberian hamsters spontaneously utilise torpor; torpor in short photoperiod-housed hamsters was assessed by inactivity and Tb, shown in Figure 6.11, B). As in mice, Txnip localised to the lateral ventricles and the choroid plexus, and within the hypothalamus, localised to the ependymal layer of cells lining the third ventricle (Figure 6.11). Txnip expression showed no difference in expression with photoperiod, with similar levels in the ependymal layer of the third ventricle of hamsters exposed to long-days (LD) and non-torpid animals exposed to short-days (SD). However, in torpid SD hamsters, significant elevation in Txnip expression within the ependyma lining the third ventricle was demonstrated (Figure, 6.11, A). Although the induction of Txnip expression during torpor was not as pronounced as in the torpid mouse, it is compelling that a similar mechanism is utilised by different species.

162

Figure 6.9 Txnip expression in the hypothalamus during torpor. Txnip expression in the brain of ad libitum fed or fasted WT female mice housed at 16°C was examined by in situ hybridisation (n = 6/group, 2 fasted non-torpid mice). Induction of torpor by fasting caused a significant elevation in Txnip expression within the brain, compared to ad libitum fed and non-torpid (NT) fasted mice. Quantification of the hybridisation signal showed that although there was a trend to an increase in Txnip expression in the non-torpid mice within the ependyma of the third ventricle (A) and the entire hypothalamic area minus the ependymal area (B) this failed to reach significance, unlike the profound induction of Txnip expression seen in torpid animals.

Individual Txnip expressions plotted against Tb at the time of tissue collection in the ependymal layer lining the third ventricle (C) and the whole hypothalamus minus the ependymal layer (D). Data shown are mean ± SEM; *P< 0.05, ***P< 0.001, significance with fasting compared to ad libitum fed, #P< 0.05 significance between torpid and non- torpid mice, one-way ANOVA with Bonferroni’s post-hoc test.

163

Figure 6.10 Peripheral Txnip expression in WT mice during torpor. A-C. Txnip expression was quantified by qPCR in ad libitum fed or fasted WT female mice housed at 16°C (n = 6/group, 2 fasted non-torpid mice). In all tissues studied, the induction of torpor by fasting caused a significant elevation in Txnip expression, compared to ad libitum fed and non-torpid (NT) fasted mice (although there was a trend to an increase in Txnip expression with fasting, this failed to reach significance).

Individual Txnip expressions plotted against Tb of the mice at the time of tissue collection (D-F). Data shown are mean ± SEM, normalised to the mouse housekeeping 18S rRNA internal control, and as fold change relative to fed animals; ***P< 0.001, **P< 0.01 significance with fasting compared to ad libitum fed, ###P< 0.001, ##P<0.01 significance between torpid and non-torpid mice, one-way ANOVA with Bonferroni’s post-hoc test.

164

Figure 6.11 Txnip expression in the hypothalamus of Siberian hamsters. Txnip expression in the brain of Siberian hamsters housed in different photoperiods; long-days (LD) and short-days (SD) upon which some hamsters became torpid (T) and some remained non-torpid (NT), was examined by in situ hybridisation (n = 4/group). Induction of torpor by short-photoperiod caused a significant elevation in Txnip expression within the ependymal layer of cells lining the third ventricle, compared to animals kept on long-photoperiod and NT animals in short-photoperiod (A). Tb of animals at the time of tissue collection is shown in B. Data shown are mean ± SEM; **P< 0.001, significance between LD and SD torpid, ###P< 0.001 significance between SD torpid and non-torpid mice, one-way ANOVA with Bonferroni’s post-hoc test.

165

6.3 DISCUSSION

The principal finding of the present study is the revelation of consistently increased Txnip expression during torpor. Txnip is dramatically upregulated in the tissues of fasted Gpr50-/- mice, which have entered torpor, with levels in tissues of fasted normothermic animals remaining comparable to those under ad libitum fed conditions. Txnip mRNA is also increased in the tissues of WT mice driven into torpor, again remaining at lower levels in animals that could not be induced into torpor. Finally, Txnip expression was increased during torpor in the hypothalamus of an additional torpid species, the Siberian hamster.

Within the hypothalamus, Txnip is strongly expressed in the ependymal layer of cells lining the third ventricle. The proposed function of Txnip as a molecular nutrient sensor, important in the regulation of energy metabolism, along with pronounced localisation in the ependyma, implicates a possible role for this protein in the detection of metabolic alterations of nutrients in the CSF, which can then be communicated to hypothalamic neurons as to the condition of whole-body energy status. Txnip expression has recently been identified in the hypothalamus by another group (Blouet and Schwartz, 2011), with reported expression in the arcuate, ventromedial, lateral, and paraventricular nuclei. Although not addressed by the authors, positive staining in the ependymal layer is evident in the images shown. The lack of discrete hypothalamic nuclei staining with anti-Txnip antibody in our hands may have been due to masking by the high level of background staining observed with secondary antibody alone. In the present work, expression of Txnip within the ependymal layer of cells lining the third ventricle, is nutritionally regulated, and is activated by fasting in WT mice. This data is in line with other studies showing that Txnip expression is suppressed in the mediobasal hypothalamus in mice refed after a 24 h fast (Blouet and Schwartz, 2011). Unlike the in situ hybridisation data, qPCR analysis of Txnip mRNA did not show a significant increase in expression in WT mice with fasting. This may be due to reduced specificity of the area analysed, with more of the surrounding brain tissue included in the RNA- extraction for qPCR, negating the change in Txnip expression demonstrated by the more restricted in situ analysis.

The morphology and anatomical location of the Txnip third ventricular staining is suggestive of ciliated ependymal cells rather than tanycytes (Mathew, 2008; Rodriguez

166 et al., 2005). These cells are metabolically active, and express much of the machinery required for sensing energy status, such as the long-form of the leptin receptor (Baskin et al., 1999), the glucose transporters GLUT1, GLUT2, GLUT4 (Farrell et al., 1992; Jetton et al., 1994; Leloup et al., 1994; Ngarmukos et al., 2001), and the sodium/glucose cotransporter, SGLT1 (Briski and Marshall, 2001). Expression of key enzymes involved in energy metabolism have further been shown, including endothelial lipase (Paradis et al., 2004), glucokinase (Jetton et al., 1994; Maekawa et al., 2000), and glycogen phosphorylase (Cataldo and Broadwell, 1986; Pfeiffer et al., 1990). As Txnip expression is responsive to diverse stimuli, including alterations in levels of glycolytic intermediates and leptin, it is possible that alterations in these signals are responsible for the changes in Txnip expression demonstrated in the ependyma lining the third ventricle.

Txnip expression in the mediobasal hypothalamus was decreased by acute application of both insulin and leptin, which do not alter circulating glucose (Blouet and Schwartz, 2011), suggesting that Txnip expression is responsive to both negative and positive energy balance.

In peripheral tissues, fasting increased expression of Txnip mRNA and protein in BAT of WT mice, but no significant increases in Txnip expression were detected in the liver or WAT of these animals. As Txnip expression is known to be upregulated by high glucose levels (see Figure 6.12), it appears paradoxical that it should be increased during fasting when circulating glucose levels would be low. However, Txnip expression is not induced by glucose per se, but in response to metabolic intermediates, with expression strongly affected by glycolytic rate. This idea is substantiated by this study, in that treatment of mice with 2DG increased the expression of Txnip in all tissues studied apart from the liver. 2DG is a stable glucose analogue, which upon transport into cells, is phosphorylated by hexokinase, the enzyme responsible for the first step in glycolysis. However, unlike glucose-6- phosphate, phosphorylated 2DG cannot be further metabolised by the next enzyme in the glycolytic pathway, phosphoglucose isomerase. 2DG-6-phosphate therefore accumulates in cells, leading to the inhibition of glycolysis through inhibition of glycolytic enzymes; inhibiting phosphoglucose isomerase in a competitive and hexokinase in a noncompetitive manner (Ralser et al., 2008). As MondoA:Mlx are extremely sensitive monitors of glucose-6-phosphate concentration and translocate to 167

Figure 6.12 Model linking Txnip expression, glycolysis and oxidative phosphorylation. Accumulation of the glycolytic intermediates glucose-6-phosphate (G6P) and glyceraldehyde-3-phosphate (GADP) cause the transcription factor complex MondoA:Mlx to translocate to the nucleus and drive the transcription of Txnip through binding carbohydrate response elements (ChoREs) in its promoter. The glucose mimetic 2-deoxyglucose (2DG) is similarly transported in to and phosphorylated within cells to 2DG6P, but unlike G6P, this cannot be further metabolised and causes the robust nuclear translocation of MondoA:Mlx and increased Txnip expression. Factors that reduce glycolytic rate, such as during periods of fasting, when fatty acids (FAs) become the major fuel source of mitochondrial oxidative phosphorylation allow the accumulation of glycolytic intermediates and induction of Txnip expression. Txnip has an inhibitory role on glucose transport; hence, unnecessary glucose uptake in cells is repressed when Txnip expression is induced. In a complimentary fashion, increased glycolytic flux which dynamically depletes glycolytic intermediate metabolites in turn represses Txnip expression and allows increased uptake of glucose by cells. Diagram by the author.

168 the nucleus when levels increase, the accumulation of 2DG-6-phosphate causes their robust nuclear translocation, and coupled with their very potent transcription activation domain, results in an effective transcriptional response and increased Txnip expression (Stoltzman et al., 2008)(see Figure 6.12). However, liver proved to be an exception to this mechanism, and levels of Txnip mRNA remained similar between 2DG and vehicle treated mice. Liver (and kidney) possesses significant glucose-6-phosphatase activity and in this tissue trapped 2DG-6-phosphate can be dephosphorylated and returned to the blood (Young et al., 1984), relieving the inhibition of glycolytic enzymes. Further, glucokinase, a variant of hexokinase found in the liver, is not subject to product inhibition by glucose-6-phosphate (or 2DG-6-phosphate), so will continue to metabolise glucose even if there is no immediate use (Massa et al., 2011). Therefore, during a fast, glycolysis within tissues (such as brain and BAT) may be reduced as circulating glucose levels become depleted and alternative fuels begin to be metabolised, resulting in accumulation of certain glycolytic intermediates and activation of Txnip expression (Figure 6.12).

It is interesting that although showing a trend towards increased Txnip expression with fasting in the liver and WAT of WT mice, expression in BAT was markedly increased. Further tissue-specific differences in Txnip regulation were demonstrated in mice acutely exposed to cold (4°C). This challenge did not alter Txnip expression in the hypothalamus or liver, but significantly increased expression in WAT and decreased expression in BAT. This is perhaps a consequence of different metabolic activities within the adipose tissues. During cold exposure, sympathetic nervous system (SNS) drive to both WAT and BAT is increased, leading to elevated lipolysis in WAT and nonshivering thermogenesis (NST) in BAT. In essence, during cold exposure, WAT responds as though the body is in a fasted state and it is perhaps not surprising that levels of Txnip are therefore increased within this tissue, considering its previously discussed induction during fasting. On the other hand, in BAT, as energy production in the mitochondria is uncoupled from fuel oxidation during NST, cells rely on increased glycolytic rates to generate sufficient ATP to meet the cell’s energy requirements (Inokuma et al., 2005). Reflective of this, is the demonstration of increased glucose uptake by BAT during cold exposure, with increased expression of the glucose transporters GLUT1, GLUT3 and GLUT4 and increased transcription of hexokinase (Daikoku et al., 2000; Olichon-Berthe et al., 1992; Yu et al., 2002). Increased glycolytic flux therefore decreases levels of glycolytic intermediates important for the function of MondoA:Mlx, and is presumably responsible for decreased Txnip expression 169 demonstrated in this tissue (Figure 6.12). Similarly, the induction of Txnip expression in BAT during fasting may again be reflective of metabolic activity within this tissue. Fasting sees decreased SNS drive to BAT, to reduce energy wasteful heat production. Decreased SNS outflow results in decreased glucose uptake by BAT (Shimizu et al., 1991) and reduced rates of glycolysis, which would lead to increased Txnip expression. Alternatively, it may be imperative to induce Txnip expression in BAT early during a fast as a means to limit glucose uptake, glycolysis and thermogenic activities of this tissue. In prolonged fasting, as demonstrated with increased Txnip in WAT during cold exposure, it could be expected that Txnip expression would also increase in tissues other than BAT.

This study also revealed apparent global alterations in Gpr50 expression in response to food restriction in WT mice. Confirming previous reports, fasting decreased Gpr50 expression in the hypothalamus (Ivanova et al., 2008), with similar downregulation of expression further demonstrated in the liver and WAT. Conversely, Gpr50 expression is upregulated in BAT of fasted mice. The means by which GPR50 mRNA levels are regulated and the requirement to drive expression in the opposite direction in BAT with fasting are intriguing. However, the interesting observation that treatment with 2DG, a glucose mimetic, increased expression of Gpr50 in all tissues examined, might indicate that Gpr50 expression is responsive to glucose availability. During fasting, when Gpr50 expression is generally reduced (with the exception of BAT), circulating levels of glucose are low, whereas although 2DG induces a state of perceived hypoglycaemia in cells by blocking glycolysis, circulating levels of glucose are actually increased (Dreau et al., 1998). Gpr50-/- mice are themselves hypersensitive to 2DG, exhibiting a torpor- like drop in metabolic rate and Tb in response to doses of 2DG which have little effect on WT mice (Appendix A). Both genotypes exhibit comparable increases in blood glucose in response to 2DG, indicating the exaggerated metabolic response of Gpr50-/- mice to 2DG-induced glucopenia is not due to an aberrant counter-regulatory glucose response, and may instead be reflective of altered glucose sensing. Indeed, mice lacking the GLUT2 glucose transporter also show increased susceptibility to fasting- induced torpor and abnormal hypothermic response to ICV administration of 2DG (Mounien et al., 2010). Within the hypothalamus, GPR50 localises to tanycytes, which express various components of glucose-sensing machinery, including the glucose transporters GLUT1 and GLUT2, and ATP-sensitive K+ channels (Garcia et al., 2003), and selective destruction of tanycytes using alloxan attenuates 2DG-induced feeding in

170 rats (Sanders et al., 2004). Therefore, GPR50 is ideally placed for a potential role in the perception of glucose availability.

Regulation of Gpr50 in BAT contrasted with other tissues examined during both fasting (showed increased versus decreased expression) and cold exposure (showed decreased versus no change in expression). This alteration in Gpr50 expression in response to fasting and cold, suggests that sympathetic activity may modulate its gene expression. During fasting, SNS drive to BAT is decreased, in order to save energy by reducing thermogenesis. On the other hand, during cold exposure, high sympathetic outflow is fundamental for heat production by BAT. Therefore increased SNS activity parallels decreased Gpr50 expression and reduced activity corresponds to an increase in expression. Treatment with 2DG has been shown to decrease norepinephrine turnover in both WAT and BAT (Brito et al., 2008), and mice treated with 2DG exhibit increased Gpr50 expression in these tissues. Norepinephrine has known stimulatory effects on glucose uptake (Liu et al., 1994), therefore, GPR50 could play a role in modulating tissue responses to changes in glucose supply and/or metabolic rate during fasting and hypothermia.

During torpor, it is known that animals switch from utilising glucose as a fuel source and instead rely on the metabolism of FAs (Heldmaier et al., 1999). Txnip functions to inhibit cellular glucose uptake and metabolism, and promotes FA utilisation. Therefore, its elevated expression may be crucial in ‘resetting’ cellular metabolism during torpor. Indeed, mice lacking Txnip demonstrate a deterioration of FA utilisation, and do not survive a 48 h fast (Oka et al., 2006), suggesting some protective role of Txnip during fasting, which is likely to be of heightened necessity during the hypometabolism of torpor. A tight link between Txnip expression and glycolytic rate has been proposed (Yu et al., 2010), where decreased glycolysis and thus accumulation of glycolytic intermediates drives increased Txnip expression, which feeds back to reduce cellular glucose uptake, enabling cells to efficiently regulate glucose homeostasis. During torpor, levels of metabolic activity are globally suppressed (Staples and Brown, 2008), which implies a decrease in cellular glycolytic rate. In such a state, Txnip induction would feed back to inhibit unnecessary glucose influx.

171

More recent work points to a pivotal role for Txnip within the hypothalamus, in the regulation of whole-body energy homeostasis (Blouet and Schwartz, 2011). Here, selective overexpression of Txnip in the medio-basal hypothalamus (MBH) of mice caused increased hepatic glucose production and lower peripheral glucose uptake in these animals. Therefore, increased expression of Txnip during torpor might be a means by which more glucose is made available as an energy source for the brain. Mice overexpressing Txnip in the MBH further demonstrated a decreased response of adipose tissue to sympathetic stimuli, with WAT showing reduced lipolysis in response to fasting and treatment with the selective β3 agonist, CL316243, and BAT demonstrating reduced thermogenesis in response to cold challenge and CL316243 administration. Therefore, during torpor, increased expression of Txnip could help conserve fat fuels and reduce Tb. Indeed, this could be the true role of Txnip during torpor, such that a large decrease in BAT response is necessary to drive mice into torpor and maintain the torpid state.

In summary, previous reports indicating that regulation of hypothalamic Txnip expression is an effective strategy to alter whole-body energy expenditure and fuel utilisation (increased levels downregulate energy expenditure; decreased levels promote energy expenditure), taken with the present work, imply that Txnip may act as a graduated signal, with increasing levels of expression during fasting and even higher levels of expression during torpor. Production of the appropriate level of Txnip could be a means by which the hypothalamus directs WAT to preserve its stored triglyceride, BAT to reduce thermogenesis, and the liver to increase production of glucose and ketones as an alternative fuel source. Altered Txnip expression in peripheral tissues may allow them to adapt to altered energy status. Overall, these adjustments culminate in reduced energy expenditure, and promote survival in the face of energetic challenge.

172

CHAPTER 7

GENERAL DISCUSSION

173

7.1 GPR50 SIGNALLING

As an orphan receptor, different routes can be taken in an attempt to understand GPR50’s function. Ideally, the identification of potential ligands or activating factors would provide strong functional information, and provide a means by which the receptor can be manipulated in order to assess putative downstream signalling pathways mediated by GPR50. A high-throughput ligand-screen has been carried out by Astra-Zeneca, but failed to identify a ligand. Unfortunately, further screening was beyond the scope of this study, but future work aimed at discovering GPR50’s ligand would undoubtedly be valuable. The rate at which transmitter GPCRs have been deorphanised has drastically decreased, with very few being paired with a ligand since 2000. Moreover, the pool of known transmitters has been exhausted, with each matched to a GPCR (Chung et al., 2008), and consequently, the remaining orphan GPCRs must bind unknown transmitters. In this respect, it can be expected that few remaining orphan GPCRs will be neuropeptide receptors, and the rest will bind nucleotides, neurotransmitters, or lipid mediators. The implication that GPR50 has a role in energy sensing and as a regulator of metabolism would suggest a focus on screening for a fatty acid derived ligand or similar compound could form the basis of ligand searches.

Immediately prior to commencing this project, the hypothesis that GPCRs formed and functioned as homo- and herterodimeric units, or indeed higher oligomers, was gaining support in the literature, challenging the dogma that GPCRs generally acted as monomers (Hansen and Sheikh, 2004; Milligan, 2004). It is now believed that interactions between distinct GPCRs can lead to the formation of novel pharmacological receptors and can diversify the function of GPCRs, and a number of novel GPCR dimers have been implicated in the development of pathophysiological conditions (Dalrymple et al., 2008; Park and Palczewski, 2005). It is therefore possible that some of the remaining orphan GPCRs do not act alone, but in conjunction with other GPCRs. The suggestion that GPR50 does not induce its own second messenger pathway, but instead acts as a negative modulator of the MT1 receptor (Levoye et al., 2006), provides a novel concept for the elucidation of the function of orphan GPCRs. However, sites of GPR50 expression, such as the ependymal layer lining the third ventricle, do not demonstrate coincident melatonin receptor expression, rendering this type of association of little significance, at least in this region of the hypothalamus.

174

The present work shows for the first time that GPR50 does indeed possess implicit signalling capability. Novel co-immunoprecipitation experiments demonstrated a specific association of GPR50 with Gαi1 subunits, indicating that like the related melatonin receptors, GPR50 can signal through Gi proteins. Further confirmation of this comes from in vitro second messenger assays, showing that overexpression of GPR50 can inhibit forskolin-stimulated cAMP accumulation within cells, the classic pathway mediated by Gi proteins. These data imply that GPR50 provides an inhibitory signal, and that this requires some previous stimulatory input, as basal levels of cAMP were not affected by GPR50 overexpression. GPR50’s capacity to interact with further Gα subunits and its effects on other second messengers such as intracellular Ca2+, cGMP, PDE or PKC should be explored, as like most GPCRs, it is likely to couple more than one family of G proteins.

Western blotting of an epitope-tagged construct encoding GPR50, demonstrated an approximate whole-C-terminal tail cleavage product of the receptor. Further work conducted in our lab has demonstrated an interaction of the C-terminal tail of GPR50 with the transcriptional co-activator TIP60, and a role for GPR50 as a signalling partner and modulator of TIP60 activity (Li et al., 2011). Interestingly, when co-expressed, the C-terminal tail of GPR50 translocates to the nucleus and co-localises with TIP60. The possibility that the intracellular tail of GPR50 is specifically cleaved in response to an extracellular (or intracellular) signal, liberating it as a functional signalling protein, is intriguing and merits further investigation. Characterisation of the proteolytic site and enzyme responsible for this cleavage is required.

7.2 GPR50 IN THE HYPOTHALAMUS

There is considerable evidence that cAMP signalling in the hypothalamus plays an important role in the regulation of energy balance in mammals. Central injection of cAMP analogues can stimulate feeding and increase activity of NPY-expressing neurons (Akabayashi et al., 1994; Gillard et al., 1997). Activation of CREB controls the expression of numerous genes implicated in the regulation of food intake and energy expenditure, including TRH, CRH, oxytocin and vasopressin (Itoi et al., 2004; Lechan and Fekete, 2006). Deletion of CREB1 specifically in the PVN causes mice to develop obesity as a result of decreased energy expenditure and an impairment of BAT- mediated thermogenesis (Chiappini et al., 2011). Mice that lack Gαs in the brain

175 develop a similar phenotype, displaying obesity associated with reduced SNS activity and energy expenditure (Chen et al., 2009). It is interesting that Gpr50-/- mice (lacking the inhibitory signal provided by the receptor) have the opposing phenotype, exhibiting resistance to weight gain and an increased metabolic rate (Ivanova et al., 2008).

During fasting, GPR50 expression within the hypothalamus is decreased (Ivanova et al., 2008), and so relief of GPR50-mediated inhibition of cAMP would be expected, allowing increased levels of cAMP to accumulate in cells expressing the receptor. Indeed, during fasting, increased activation of CRE-dependent gene transcription, indicating increased cAMP signalling, in cells of the DMH has been demonstrated (Shimizu-Albergine et al., 2001). It is tempting therefore, to speculate that relief of GPR50’s inhibitory effect on cAMP signalling may contribute to increased CRE- mediated gene induction during a fast, both in the DMH and in the ependymal tanycyte cells lining the third ventricle. Expression of Dio2, which is responsible for the local conversion of T3, is highly concentrated in tanycytes (Diano et al., 1998; Tu et al., 1997), and is strongly induced by cAMP-dependent pathways due to the presence of a CRE element in its promoter (Bartha et al., 2000). During fasting, expression of Dio2 in the hypothalamus is upregulated (Diano et al., 1998), concomitant with decreased expression of GPR50, resulting in increased T3 feedback and continued repression of TRH during a fast, which has established roles in the control of energy balance and BAT activity (Griffiths et al., 1988; Rogers et al., 2009). Therefore, it could be that GPR50-mediated cAMP inhibition influences Dio2 expression, and reduced levels of GPR50 in tanycytes during fasting allows intracellular levels of cAMP to increase, driving Dio2 expression.

Previous work has demonstrated elevated constitutive expression of Dio2 in tanycytes of Gpr50-/- mice (Appendix A) and reduced levels of TRH in the PVN (perhaps a result of increased feedback of T3 due to greater rates of its production by augmented levels of Dio2). However, in the current study, microarray data revealed no difference in basal expression of Dio2 or TRH in the hypothalamus of Gpr50-/- mice. Further, fasting did not significantly alter TRH expression in either genotypes, and a significant increase in Dio2 expression with fasting was only demonstrated in the Gpr50-/- mice. These differences may be attributable to reduced specificity of the hypothalamic areas analysed by the microarray, whereby opposing gene expression changes in different hypothalamic nuclei could negate an overall detectable change in transcript expression 176 in such small regional tissue, or low magnitude alterations in gene expression, with only four animals per group could result in non-significant differences following statistical testing.

7.3 GPR50 IN WHITE ADIPOSE TISSUE

The present studies investigating GPR50 in adipocytes provide evidence for the first time that this receptor can directly influence WAT biology. Before commencing these experiments, it was known that Gpr50-/- mice have reduced WAT stores and a partial resistance to diet-induced obesity, and in humans, polymorphisms in Gpr50 have been linked to elevated circulating triglycerides and reduced high-density lipoprotein cholesterol levels (Bhattacharyya et al., 2006). The localisation of GPR50 in murine WAT and demonstration of increasing levels of Gpr50 mRNA expression during differentiation of the adipocytic 3T3-L1 cell line, suggests some function of the receptor in mature adipocytes.

Knockdown of endogenously expressed Gpr50 in 3T3-L1s profoundly altered the metabolic phenotype of the cells, with a drastic reduction in stored triglyceride and basal lipid mobilisation. In a complementary fashion, overexpression of GPR50 led to increased triglyceride accumulation in cells and basal lipolysis. On the other hand, both overexpression and knockdown of Gpr50 led to impaired stimulated lipolysis in differentiated 3T3-L1 cells. Thus, expression levels of GPR50 impact on the ability of WAT to mobilise lipid stores, as well as the inherent lipogenic activity of the tissue. Too high levels of GPR50 may be associated with an obese phenotype, whereby a new steady state is achieved in WAT leading to increased accumulation of triglyceride, and conversely, loss of the receptor is associated with a lean phenotype, demonstrated by the Gpr50-/- mice, which have decreased fat stores. In vivo studies (Bechtold, unpublished data), demonstrated that circulating levels of triglyceride are increased in Gpr50-/- mice, which either implies increased hepatic production, or decreased uptake by WAT. The latter would be in line with the present study, showing decreased triglyceride accumulation in differentiated 3T3-L1s with targeted knockdown of Gpr50, which also demonstrated decreased expression of Lpl, responsible for the cellular uptake of triglycerides by hydrolysing them into FAs and glycerol (Gonzales and Orlando, 2007).

177

Overexpression of Gpr50 impairs acute lipid mobilisation from differentiated 3T3-L1s, thus during fasting, reduction of Gpr50 expression in WAT may be necessary to improve lipid mobilisation from this tissue in response to SNS stimulation. The SNS generates signals to increase lipolytic rate by producing adrenergic stimuli that activate β-adrenergic receptors, which then signal through Gs proteins to increase intracellular cAMP. Thus, as GPR50 inhibits forskolin-stimulated cAMP accumulation (demonstrated in HEK293 cells), reduction of this inhibitory signal during fasting might be necessary to allow cAMP-mediated induction of lipolysis. However, too low levels of the receptor also impairs lipolysis, as demonstrated by the in vitro knockdown studies and also in vivo work showing Gpr50-/- mice demonstrate reduced weight loss during a fast when compared to WT mice (Ivanova et al., 2008). Therefore, expression of the appropriate level of GPR50 in WAT appears to impact greatly on lipolytic rate.

Interestingly, the microarray data shows that during a fast, WT mice have increased transcript levels of hormone sensitive lipase (HSL) within the hypothalamus, whereas Gpr50-/- mice fail to upregulate levels of this mRNA. If a similar effect occurs in WAT, a reduction in lipolysis could underlie the reduced weight loss demonstrated by Gpr50-/- mice during fasting. The role of HSL within the hypothalamus has not been widely investigated; however a novel role for this enzyme in the control of hypothalamic neuropeptide expression and feeding behaviour has been suggested. Essentially, double knockout ob/ob/Hsl-/- mice exhibit reduced food intake and rescued hypothalamic expression of NPY and AgRP, compared to the hyperphagia demonstrated by leptin-deficient ob/ob mice, mediated by hypothalamic upregulation of NPY and AgRP (Sekiya et al., 2004). Along with a demonstrated hypothalamic expression of HSL, the authors suggest that the defective generation of FFAs in the hypothalamus due to the absence of HSL mediates the altered expression of orexigenic neuropeptides. The crucial role of hypothalamic metabolism of FAs in the control of feeding behaviour and energy expenditure is generally appreciated. It is noteworthy that inhibition of hypothalamic FA synthesis suppresses food intake (Loftus et al., 2000) and fasting-induced induction of NPY and AgRP (Shimokawa et al., 2002), and increases peripheral energy expenditure (Cha et al., 2005; Cha et al., 2006). Likewise, inhibition of hypothalamic carnitine palmitoyltransferase-1, which is responsible for FA transport into mitochondria for oxidation, decreases food intake (Obici et al., 2003), whereas inhibition of hypothalamic acetyl-coA-carboxylase (catalyses first step in FA synthesis) stimulates food intake (Lane et al., 2008). Thus it is possible that altered levels of hypothalamic HSL expression and subsequent 178 generation of FFAs may impact on hypothalamic circuits controlling food intake and energy expenditure. This could underlie the observation of increased HSL expression during fasting in WT mice, but not during torpor in the Gpr50-/- mice, as demonstrated by the microarray data.

7.4 GENE CHANGES DURING TORPOR

Most studies describing torpor in laboratory mice employ females, which are more prone to torpor, and induce this response with a combination of food restriction and reduced ambient temperature, or chronic food restriction. In the two lines of Gpr50-/- mice used in this study, male mice housed at 20-22°C enter torpor within as little as 6 hr of fasting. It is unlikely that diminished energy reserves of Gpr50-/- mice underlie the heightened torpor response, as torpor could not be induced in WT mice during a 48 hr fast (Appendix A), when energy reserves are severely depleted. In the current study, torpor could only be induced in female WTs when fasting was combined with reduced ambient temperature. Therefore, GPR50 may play some role in regulating metabolic responses with loss of the receptor permissive for entering the torpid state.

The distinct alterations of the metabolic phenotype of Gpr50-/- mice, and the impressive displays of fasting-induced torpor by these animals led to the implementation of a microarray study in pursuit of gene changes underlying their altered metabolism and transcript changes induced by torpor. As the hypothalamus is a major centre of convergence and integration of multiple nutrient-related signals important in the regulation of energy homeostasis, exhibits unequivocal control of torpor, and is an intense site of GPR50 expression, it seemed obvious to focus attention here. Curiously, apart from the apparent genomic abnormalities discussed in Chapter 3, there were no significant basal differences in gene expression in the fed state between Gpr50-/- and WT mice. This was surprising, considering the marked metabolic alterations of the - Gpr50-/- mice, and as discussed previously, may be due to the lack of specificity of the hypothalamic areas analysed, with RNA extracted from whole hypothalamic blocks for use in the microarray. Thus, opposing changes in different hypothalamic nuclei could negate a demonstration of altered transcript level, or indeed, very subtle changes in gene expression in a particular area would be lost during analysis. In this scenario, it may have been more prudent to employ laser microdissection to analyse gene changes in specific areas, such as the ependyma of the third ventricle or the DMH.

179

However, gene expression changes in areas that do not demonstrate GPR50 expression were already known (PVN), so the rationale to use hypothalamic blocks was not considered unadvisable when commencing the microarray.

Despite the impressive phenotype of torpor, relatively few gene changes were apparent within the hypothalamus, implying that gene regulation is generally well protected against the effect of temperature change. However, an enrichment of “negative transcriptional regulator” genes were altered during torpor and transcripts in the “transcriptional regulation” group were mostly downregulated (Figure 5.3), suggesting that transcription is decreased in the hypothalamus of hypometabolic, torpid mice, which is in line with previous studies (Berriel Diaz et al., 2004; van Breukelen and Martin, 2002). An enrichment of “circadian rhythm” genes was also demonstrated, which again, is consistent with previous studies (Crawford et al., 2007; Herwig et al., 2006; Herwig et al., 2007), and might be expected due to the compelling evidence for the involvement of the circadian system in the maintenance of correctly timed torpor.

The mechanism underlying torpor in the Gpr50-/- mice and how thermogenic drive in these animals is modulated remains unclear. No differences in the expression of NPY, AgRP, Pomc, Cart, or their respective receptors, within the hypothalamus were revealed by the microarray data during fasting compared with WT mice (Table 7.1), suggesting that altered hypothalamic signalling in the Gpr50-/- mice occurs downstream of these well-studied ARC neuropeptides. Alterations in thyroid hormone signalling were anticipated due to the strong expression of GPR50 in tanycytes, the site of Dio2 expression, and the well-established role for TRH in regulating energy expenditure and BAT thermogenesis. However, again, no differences in the expression of Dio or TRH or its receptor were demonstrated.

7.4.1 Txnip, a novel torpor-induced gene

Of the genes selected from the microarray for further validation, Txnip was the most upregulated in the hypothalamus during torpor. Expression levels of Txnip were the most increased in the WT mice with fasting, showing a 2.39 fold increase; however the Gpr50-/- mice showed an impressive 7.67 fold induction with fasting. Txnip expression showed a similar pattern in peripheral tissues also, with consistently higher levels of

180

Gene Average Probe Intensity KO Fast v WT Fast Fed Fasted q-value Fold- WT KO WT KO change Npy 101.508 118.521 178.983 154.768 0.687 -1.16 Npy1r 121.416 108.075 140.383 150.945 0.795 1.08 Npy2r 25.336 25.456 22.062 25.888 0.595 1.17 Npy3r 14.081 15.107 13.989 12.424 0.648 -1.13 Npy4r 11.749 11.668 11.640 11.343 0.838 -1.03 Npy5r 19.100 19.508 17.113 18.671 0.799 1.09 Trh 152.680 149.198 159.567 132.858 0.708 -1.20 Trhr 43.379 41.941 42.699 42.836 0.866 1.00 Trhr2 32.720 36.220 48.446 46.812 0.830 -1.03 Tshr 15.525 14.212 13.776 13.242 0.770 -1.04 Tshb 9.968 9.650 10.172 9.855 0.765 -1.03 Dio1 11.102 11.002 10.883 10.453 0.781 -1.04 Dio2 52.449 52.369 72.590 85.417 0.647 1.18 Dio3 9.656 9.421 9.715 9.606 0.814 -1.01 Crh 31.384 30.533 36.341 30.452 0.553 -1.19 Crhr1 43.907 40.687 44.722 49.230 0.729 1.10 Crhr2 19.805 17.767 18.229 19.712 0.688 1.08 Crhbp 28.599 28.856 38.580 41.362 0.725 1.07 Agrp 11.582 12.770 16.827 12.315 0.114 -1.37 Cartpt 165.368 214.404 183.505 203.569 0.794 1.11 Pomc 31.544 36.184 33.441 37.224 0.683 1.11 Mc1r 36.828 32.911 33.625 40.224 0.638 1.20 Mc2r 12.584 12.636 12.498 12.248 0.831 -1.02 Mc3r 54.050 56.482 50.600 47.294 0.768 -1.07 Mc4r 77.308 71.370 61.993 76.514 0.679 1.23 Mc5r 17.002 17.192 17.233 17.501 0.835 1.02

Table 7.1 Hypothalamic transcripts with known/predicted changes in expression with altered energy status/torpor. Affymetrix mouse arrays were used to screen the differentially expressed genes in the hypothalamus between fed and fasted WT and Gpr50-/- mice. Average expression of transcripts is listed (n=4/group). No significant differences in any of the listed genes were observed either during the normothermic or torpid state between Gpr50-/- and WT mice. Npy: neuropeptide Y; R: receptor; Trh: thyrotropin-releasing hormone; Dio: deiodinase; Crh: corticotropin-releasing hormone; Agrp: agouti-related peptide; Cartpt: cocaine and amphetamine-related transcript prepropeptide; Pomc: pro- opiomelanocortin; Mc: melanocortin.

181 expression in the torpid Gpr50-/- mice. Fascinatingly, Txnip expression is also increased in the liver, BAT and WAT of female WT mice driven into torpor by fasting at lower ambient temperature. The upregulation of Txnip occurs only with torpor itself, with the pronounced increase lost in normothermic fasted Gpr50-/- mice and also the fasted female WT mice that failed to enter a torpid state. As mice lacking Txnip demonstrate lethality during a prolonged fast (Oka et al., 2006), it could be that Txnip offers some protective function during hypometabolism that is further induced during the more extreme metabolic stress of torpor. Txnip may be essential during the ‘starvation response’ to inhibit energy-consuming processes, which would be fundamental to the hypometabolism demonstrated during torpor. It is highly interesting that a further torpid species, the Siberian hamster, also demonstrates increased expression of Txnip within the hypothalamus upon entrance into torpor. However, the induction of Txnip in torpid Siberian hamsters is not as exaggerated as in the Gpr50-/- or WT mice driven into torpor. This could be reflective of the role of fasting itself in influencing Txnip expression, as although Siberian hamsters reduce their food intake preceding the demonstration of spontaneous daily torpor; it is perhaps not exactly comparable to the fasting-induced torpor in laboratory mice strains.

Feasibly, induction of Txnip expression is a required and conserved mechanism for the maintenance of the torpor in mammals. Chronic food restriction is a known means of inducing torpor in mice, and so an inability of mice lacking Txnip to upregulate its expression could therefore mean these animals fail to utilise torpor as an energy- saving strategy, which contributes to the increased mortality during a prolonged fast. If the increase of hypothalamic Txnip during torpor was perturbed in Gpr50-/- mice, it would be interesting to observe whether these animals can enter torpor with fasting, and indeed whether they would survive the fast.

Presumably, increased levels of Txnip during torpor leads to increased inhibition and decreased activity of thioredoxin. On the other hand, although the modulation of glucose uptake by Txnip remains to be precisely elucidated, in vitro studies demonstrated that it is independent of its ability to bind thioredoxin (Patwari et al., 2009), thus the role of elevated Txnip during torpor may not involve its ability to regulate this enzyme. Nevertheless, with the emerging view that the redox state within hypothalamic neurons is implicated in the response to nutritional signals and the regulation of energy metabolism, it seems possible that the thioredoxin redox buffer 182 could be involved in this regulation. Further, the inclusion in experiments of the C247S- Txnip mutant, which is unable to bind thioredoxin, demonstrated that this activity is required for the effects of hypothalamic Txnip on energy homeostasis (Blouet and Schwartz, 2011).

Mitochondrial production of ROS by electron leakage during intracellular glucose and fatty acid metabolism has been shown to contribute to hypothalamic sensing of both types of fuel source. Quenching ROS production in the hypothalamus prevents glucose-induced electrical activation of arcuate neurons (Leloup et al., 2006), and reduces glucose-induced hyperpolarisation of glucose-inhibited neurons in the VMH and oleic acid-induced inhibition or activation of VMH neurons (Le Foll et al., 2009). Mitochondrial ROS production is increased in the hypothalamus of obese rats, associated with a constitutive oxidised environment and lower antioxidant enzyme activity (Colombani et al., 2009). These alterations led to hypersensitivity to low glucose load which could be fully reversed by restoration of the hypothalamic redox state. Speculatively, altered ROS levels could influence redox signalling involving ROS- sensitive voltage-dependent channels which will then modulate neuronal electrical activity (Avshalumov et al., 2005; Hudasek et al., 2004).

Differential subsets of hypothalamic neurons have contrasting responses to ROS in a substrate-dependent manner. For example, whereas POMC neurons have been shown to utilise glucose as their main fuel, NPY/AgRP neurons do not, and instead use FFAs as substrates (Horvath et al., 2009). Thus, via competitive mechanisms of fuel utilisation (when glycolysis is elevated, β-oxidation is inhibited, and vice versa ( et al., 1985)) high levels of glucose enhances POMC and reduces NPY/AgRP neuronal activity, whereas fatty acid β-oxidation has the opposite effect (Horvath et al., 2009). The by-product of either substrate oxidation is ROS; however, the responsiveness of these neuronal populations to increasing levels differs. When NPY/AgRP neurons are activated during negative energy balance by increased levels of free fatty acids, ROS levels are not raised in these cells despite increased firing and substrate utilisation. By contrast, when glucose-utilising POMC neurons are firing at high levels, ROS accumulates in these cells. (Andrews et al., 2008). Thus, sustained ROS levels in POMC neurons seem to favour satiety, whereas for the appropriate responses to negative energy balance, ROS levels are buffered in the hypothalamus, and failure to do so leads to impaired neuronal firing (Andrews et al., 2008). In NPY/AgRP neurons, 183 the endogenous buffering of ROS occurs via the activation of UCP2, which demonstrates increased expression in the hypothalamus with fasting, and induces ROS scavenging (Andrews et al., 2008). By contrast, glucose-induced ROS generation in POMC neurons during positive energy balance does not trigger such a buffering mechanism (Andrews et al., 2008). As well as being essential for ROS scavenging and thus allowing continuous FA oxidation during increased NPY/AgRP neuronal firing, UCP2 also supports the bioenergetic needs to maintain electrical activation of NPY/AgRP cells by increasing mitochondrial respiration rate and biogenesis in these neurons. Dio2-expressing tanycytes are in direct apposition to NPY/AgRP neurons in the ARC, and fasting-induced increases in Dio2 activity and local T3 production mediates the activation of UCP2 and mitochondrial proliferation in NPY/AgRP neurons (Coppola et al., 2007). Therefore, it seems that local thyroid hormone production is critical for increased excitability of orexigenic neurons during food deprivation.

Increased Txnip in the hypothalamus would lead to impairment of intracellular redox status, and Txnip expression can be activated by ROS via the MEK/MAPK pathway (Fang et al., 2011), which would further serve to increase Txnip levels. Increased levels of ROS would be expected to decrease cell survival. However, an increase of ROS in NPY/AgRP neurons during fasting is buffered via the induction of UCP2 (Andrews et al., 2008), and this could be a mechanism employed by other cell types, as UCP2 demonstrates ubiquitous expression (Boss et al., 1997; Fleury et al., 1997). Furthermore, during torpor, a general depression in metabolic activity would presumably result in minimal ROS production due to reduced substrate oxidation, and so the large increases in Txnip and potential increased inhibition of thioredoxin may not prove deleterious to cells under these conditions.

This work provides the first evidence in favour for a role of increased Txnip expression in torpor and further evidence for acute regulation during fasting. As with fasting, Txnip is similarly induced in the hypothalamus by acute nutrient excess and in mouse models of obesity and diabetes, and alteration of hypothalamic Txnip expression alone, impacts on whole energy metabolism (Blouet and Schwartz, 2011). Along with the present data, this suggests that hypothalamic Txnip plays a critical and previously unevaluated role in nutrient sensing and the regulation of energy balance. It is interesting that overexpression of MBH Txnip does not significantly affect food intake, body weight gain or body composition on a normal chow diet (Blouet and Schwartz, 184

2011), with the authors conducting all subsequent studies on animals under high-fat feeding. Therefore, it may be that augmented expression of hypothalamic Txnip only affects energy homeostasis under conditions of nutrient imbalance. Like Txnip, Gpr50 expression is influenced by negative and positive energy status, being suppressed by both fasting and high-fat feeding (Ivanova et al., 2008), and although Gpr50-/- mice exhibit significantly lower body weights than WT littermates in the first 10 weeks following weaning, weight gain over the following 5 weeks is similar in both genotypes when maintained on a standard chow diet, and it is only when they are placed on a high energy diet that their resistance to weight gain is observed (Ivanova et al., 2008). Altered expression of Txnip and Gpr50 may therefore be responsive to similar physiological signals and may be required to allow animals to respond appropriately to nutrient excess or depletion.

7.4.2 Gpr50 in brown adipose tissue

Interestingly, Gpr50 is regulated in a novel way in BAT in response to fasting and cold exposure, and as in the hypothalamus, may have a role in influencing Dio2 expression here. Dio2 is induced in BAT in response to SNS outflow which increases intracellular cAMP via the action of NE at β-adrenergic receptors. The increase in Dio2 activity and of local Dio2-generated T3 is critical for increasing the responsiveness of BAT to NE signalling and directly mediates the increase in total levels of UCP1 (Bianco et al., 1992; Carvalho et al., 1991), such that Dio2 is essential during adaptive thermogenesis, and animals that lack Dio2 are cold intolerant due to deficient BAT function and develop severe obesity when kept on a HFD at thermoneutrality (Castillo et al., 2011; de Jesus et al., 2001). During cold exposure, regulation of Gpr50 expression in BAT contrasted with other tissues, as levels were down-regulated, with no significant changes in expression in other tissues examined. As expected, Dio2 levels in BAT were strongly upregulated with cold exposure, and so, as in the hypothalamus during fasting, perhaps relief of GPR50-mediated inhibition of cAMP signalling is necessary to allow cAMP-mediated induction of Dio2. A trend to higher levels of Dio2 upon cold exposure is seen in Gpr50-/- mice; however there is no difference in basal levels.

During fasting, in all other tissues examined (hypothalamus, liver and WAT), Gpr50 expression is decreased, however in BAT, expression is induced. Unexpectedly, levels

185 of Dio2 were elevated in BAT with fasting in Gpr50-/- mice, which seems counterintuitive as during fasting there is decreased SNS drive to BAT to direct reduced energy expenditure in this tissue. Induction of Dio2 was not as dramatic as seen during cold exposure (in Gpr50-/- mice 4-fold versus 18-fold increase, respectively), but there was no significant increase of expression in WT mice. This would suggest that an elevation of Gpr50 expression in WT BAT during fasting might impact on intracellular cAMP levels sufficiently to control expression of Dio2, whereas this is lost in the Gpr50-/- animals. However, it could also be that GPR50 signalling does not have a major impact on cAMP in this tissue, and increased expression during fasting is required to activate other signalling pathways.

Gpr50-/- mice exhibit an exaggerated thermal and hypometabolic response during fasting. Despite this, these animals show the appropriate changes in Dio2 and Ucp1 expression in response to cold exposure, suggesting neither sympathetic tone nor BAT response and non-shivering thermogenesis are impaired in these mice. It might be assumed that sympathetic output to BAT and thus levels of Ucp1 would be lower in Gpr50-/- mice, underlying the demonstrated hypometabolism in these animals in response to fasting. However, levels of Ucp1 are similar to that of WT mice in both the fed and fasted state, although previous work in this lab suggested reduced Ucp1 expression in Gpr50-/- mice (Appendix A). Slight differences may be attributable to the time of day or arousal status of the mice at the time of sacrifice.

SNS activity is essential for the regulation of torpor, controlling both entry and arousal. Blockade of peripheral SNS signalling in Siberian hamsters eliminates the display of torpor showing that its expression is dependent on intact sympathetic signalling (Braulke and Heldmaier, 2010). An increase in metabolic rate is often displayed in Gpr50-/- mice immediately prior to entry into torpor, a phenomenon also demonstrated by other species such as the dormouse, the Gould’s long-eared bat, the eastern pygmy possum, the golden-mantled ground squirrel, and the alpine marmot, which frequently show similar peaks of metabolic rate anticipating the entrance into torpor (Braulke and Heldmaier, 2010). This high metabolic rate indicates increased SNS activity, which may be necessary to increase outflow to WAT and reduce leptin secretion, a signal required for the entrance into torpor (Freeman et al., 2004; Gavrilova et al., 1999; Swoap and Weinshenker, 2008). Further, SNS activity is essential for heat production and increased heart rate during arousal from torpor (Swoap and Weinshenker, 2008). 186

Although metabolic rate during torpor may be a fraction of that in normothermic individuals, regulation of Tb during torpor is not abandoned, but is instead regulated around a lower set point (Geiser, 2004). This is demonstrated by Gpr50-/- mice, which if fasted at lower ambient temperature (~15°C), do not decrease their Tb during torpor any lower than when fasted at 20-22°C (Appendix F), and so metabolic heat production -/- must be employed to maintain Tb. Therefore, it is perhaps not surprising that Gpr50 mice can regulate levels of Ucp1 in BAT to a similar extent as WT mice in response to fasting, despite the demonstration of obvious hypothermia.

7.5 THESIS OVERVIEW

Many GPCRs, including GPR50 were discovered on the basis of , which suffers from one obvious problem; these receptors lack their pharmacological and physiological identities. As the vast majority of these receptors are conserved in different species, it is assumed that they are functionally active; however the search for their endogenous ligands and physiological roles is a formidable task. Despite the vast and longstanding efforts of academic and industrial researchers, approximately 100 nonchemosensory/transmitter GPCRs remain orphan receptors, for which the cognate ligands have not yet been identified (Chung et al., 2008). This group includes GPR50.

Expression of GPR50 in response to changes in nutritional status appears to be an essential feature of the biological function of this receptor. Specific regulation in response to fasting, high-fat feeding (Ivanova et al., 2008), 2DG glucoprivation and cold-exposure imply some critical role for GPR50 in the regulation of energy metabolism. It would be fascinating to investigate whether using viral strategies to alter levels of hypothalamic GPR50 expression alone, would impact on the metabolic phenotype of mice. Preventing fasting-induced decreases in expression might adjust energy conservation mechanisms during this time of nutrient depletion. Further, knockdown of GPR50 expression in the hypothalamus of WT mice might cause them to display metabolic phenotypes similar to that of the whole-animal knockouts. Finally, it would be intriguing to examine whether reinstatement of GPR50 expression in the hypothalamus of mice lacking the receptor can rescue their altered metabolic phenotype and prevent fasting-induced torpor.

187

REFERENCES

188

Ahima, R.S. (2006). Adipose tissue as an endocrine organ. Obesity (Silver Spring) 14 Suppl 5, 242S-249S.

Ahima, R.S., and Lazar, M.A. (2008). Adipokines and the peripheral and neural control of energy balance. Mol Endocrinol 22, 1023-1031.

Ahmadian, M., Wang, Y., and Sul, H.S. (2010). Lipolysis in adipocytes. Int J Biochem Cell Biol 42, 555-559.

Akabayashi, A., Zaia, C.T., Gabriel, S.M., Silva, I., Cheung, W.K., and Leibowitz, S.F. (1994). Intracerebroventricular injection of dibutyryl cyclic adenosine 3',5'- monophosphate increases hypothalamic levels of neuropeptide Y. Brain Res 660, 323- 328.

Alaerts, M., Venken, T., Lenaerts, A.S., De Zutter, S., Norrback, K.F., Adolfsson, R., and Del-Favero, J. (2006). Lack of association of an insertion/deletion polymorphism in the G protein-coupled receptor 50 with in a Northern Swedish population. Psychiatr Genet 16, 235-236.

Alewijnse, A.E., Timmerman, H., Jacobs, E.H., Smit, M.J., Roovers, E., Cotecchia, S., and Leurs, R. (2000). The effect of mutations in the DRY motif on the constitutive activity and structural instability of the histamine H(2) receptor. Mol Pharmacol 57, 890- 898.

Andres, A.M., Ratliff, E.P., Sachithanantham, S., and Hui, S.T. (2011). Diminished AMPK signaling response to fasting in thioredoxin-interacting protein knockout mice. FEBS Lett 585, 1223-1230.

Andrews, Z.B., Liu, Z.W., Walllingford, N., Erion, D.M., Borok, E., Friedman, J.M., Tschop, M.H., Shanabrough, M., Cline, G., Shulman, G.I., et al. (2008). UCP2 mediates ghrelin's action on NPY/AgRP neurons by lowering free radicals. Nature 454, 846-851.

Aschoff, J. (1979). Circadian rhythms: influences of internal and external factors on the period measured in constant conditions. Z Tierpsychol 49, 225-249.

Ataman, B., Ashley, J., Gorczyca, D., Gorczyca, M., Mathew, D., Wichmann, C., Sigrist, S.J., and Budnik, V. (2006). Nuclear trafficking of Drosophila Frizzled-2 during synapse development requires the PDZ protein dGRIP. Proc Natl Acad Sci U S A 103, 7841-7846.

189

Avshalumov, M.V., Chen, B.T., Koos, T., Tepper, J.M., and Rice, M.E. (2005). Endogenous hydrogen peroxide regulates the excitability of midbrain dopamine neurons via ATP-sensitive potassium channels. J Neurosci 25, 4222-4231.

Ayoub, M.A., Levoye, A., Delagrange, P., and Jockers, R. (2004). Preferential formation of MT1/MT2 melatonin receptor heterodimers with distinct ligand interaction properties compared with MT2 homodimers. Mol Pharmacol 66, 312-321.

Bakker, R.A., Lozada, A.F., van Marle, A., Shenton, F.C., Drutel, G., Karlstedt, K., Hoffmann, M., Lintunen, M., Yamamoto, Y., van Rijn, R.M., et al. (2006). Discovery of naturally occurring splice variants of the rat that act as dominant-negative isoforms. Mol Pharmacol 69, 1194-1206.

Balendiran, G.K., Schnutgen, F., Scapin, G., Borchers, T., Xhong, N., Lim, K., Godbout, R., Spener, F., and Sacchettini, J.C. (2000). Crystal structure and thermodynamic analysis of human brain fatty acid-binding protein. J Biol Chem 275, 27045-27054.

Ballesteros, J.A., Jensen, A.D., Liapakis, G., Rasmussen, S.G., Shi, L., Gether, U., and Javitch, J.A. (2001). Activation of the beta 2-adrenergic receptor involves disruption of an ionic lock between the cytoplasmic ends of transmembrane segments 3 and 6. J Biol Chem 276, 29171-29177.

Balsalobre, A., Damiola, F., and Schibler, U. (1998). A serum shock induces circadian gene expression in mammalian tissue culture cells. Cell 93, 929-937.

Bamshad, M., Song, C.K., and Bartness, T.J. (1999). CNS origins of the sympathetic nervous system outflow to brown adipose tissue. Am J Physiol 276, R1569-1578.

Barrett, P., Conway, S., and Morgan, P.J. (2003). Digging deep--structure-function relationships in the melatonin receptor family. J Pineal Res 35, 221-230.

Barrett, P., Ebling, F.J., Schuhler, S., Wilson, D., Ross, A.W., Warner, A., Jethwa, P., Boelen, A., Visser, T.J., Ozanne, D.M., et al. (2007). Hypothalamic thyroid hormone catabolism acts as a gatekeeper for the seasonal control of body weight and reproduction. Endocrinology 148, 3608-3617.

Barrett, P., Ivanova, E., Graham, E.S., Ross, A.W., Wilson, D., Ple, H., Mercer, J.G., Ebling, F.J., Schuhler, S., Dupre, S.M., et al. (2006). Photoperiodic regulation of

190 cellular retinol binding protein, CRBP1 and nestin in tanycytes of the third ventricle ependymal layer of the Siberian hamster. J Endocrinol 191, 687-698.

Barrett, P., MacLean, A., and Morgan, P.J. (1994). Evidence for multiple forms of melatonin receptor-G-protein complexes by solubilization and gel electrophoresis. J Neuroendocrinol 6, 509-515.

Bartha, T., Kim, S.W., Salvatore, D., Gereben, B., Tu, H.M., Harney, J.W., Rudas, P., and Larsen, P.R. (2000). Characterization of the 5'-flanking and 5'-untranslated regions of the cyclic adenosine 3',5'-monophosphate-responsive human type 2 iodothyronine deiodinase gene. Endocrinology 141, 229-237.

Baskin, D.G., Schwartz, M.W., Seeley, R.J., Woods, S.C., Porte, D., Jr., Breininger, J.F., Jonak, Z., Schaefer, J., Krouse, M., Burghardt, C., et al. (1999). Leptin receptor long-form splice-variant protein expression in neuron cell bodies of the brain and co- localization with neuropeptide Y mRNA in the arcuate nucleus. J Histochem Cytochem 47, 353-362.

Bechtold, D.A., and Loudon, A.S. (2007). Hypothalamic thyroid hormones: mediators of seasonal physiology. Endocrinology 148, 3605-3607.

Belfrage, P., Fredrikson, G., Nilsson, N.O., and Stralfors, P. (1981). Regulation of adipose-tissue lipolysis by phosphorylation of hormone-sensitive lipase. Int J Obes 5, 635-641.

Bellinger, L.L., and Bernardis, L.L. (2002). The dorsomedial hypothalamic nucleus and its role in ingestive behavior and body weight regulation: lessons learned from lesioning studies. Physiol Behav 76, 431-442.

Bernardis, L.L., and Bellinger, L.L. (1998). The dorsomedial hypothalamic nucleus revisited: 1998 update. Proc Soc Exp Biol Med 218, 284-306.

Berriel Diaz, M., Lange, M., Heldmaier, G., and Klingenspor, M. (2004). Depression of transcription and translation during daily torpor in the Djungarian hamster (Phodopus sungorus). J Comp Physiol B 174, 495-502.

Bertelli, M., Alushi, B., Veicsteinas, A., Jinnah, H.A., and Micheli, V. (2009). Gene expression and mRNA editing of serotonin receptor 2C in brains of HPRT gene knock- out mice, an animal model of Lesch-Nyhan disease. J Clin Neurosci 16, 1061-1063.

191

Bezaire, V., Mairal, A., Ribet, C., Lefort, C., Girousse, A., Jocken, J., Laurencikiene, J., Anesia, R., Rodriguez, A.M., Ryden, M., et al. (2009). Contribution of adipose triglyceride lipase and hormone-sensitive lipase to lipolysis in hMADS adipocytes. J Biol Chem 284, 18282-18291.

Bhattacharyya, S., Luan, J., Challis, B., Keogh, J., Montague, C., Brennand, J., Morten, J., Lowenbeim, S., Jenkins, S., Farooqi, I.S., et al. (2006). Sequence variants in the melatonin-related receptor gene (GPR50) associate with circulating triglyceride and HDL levels. J Lipid Res 47, 761-766.

Bianco, A.C., Kieffer, J.D., and Silva, J.E. (1992). Adenosine 3',5'-monophosphate and thyroid hormone control of uncoupling protein messenger ribonucleic acid in freshly dispersed brown adipocytes. Endocrinology 130, 2625-2633.

Bianco, A.C., Salvatore, D., Gereben, B., Berry, M.J., and Larsen, P.R. (2002). Biochemistry, cellular and molecular biology, and physiological roles of the iodothyronine selenodeiodinases. Endocr Rev 23, 38-89.

Billiet, L., Furman, C., Larigauderie, G., Copin, C., Page, S., Fruchart, J.C., Brand, K., and Rouis, M. (2008). Enhanced VDUP-1 gene expression by PPARgamma agonist induces apoptosis in human macrophage. J Cell Physiol 214, 183-191.

Bird, M.I., Munday, L.A., Saggerson, E.D., and Clark, J.B. (1985). Carnitine acyltransferase activities in rat brain mitochondria. Bimodal distribution, kinetic constants, regulation by malonyl-CoA and developmental pattern. Biochem J 226, 323- 330.

Blouet, C., and Schwartz, G.J. (2010). Hypothalamic nutrient sensing in the control of energy homeostasis. Behav Brain Res 209, 1-12.

Blouet, C., and Schwartz, G.J. (2011). Nutrient-sensing hypothalamic TXNIP links nutrient excess to energy imbalance in mice. J Neurosci 31, 6019-6027.

Bockaert, J., Claeysen, S., Becamel, C., Pinloche, S., and Dumuis, A. (2002). G protein-coupled receptors: dominant players in cell-cell communication. Int Rev Cytol 212, 63-132.

Bockaert, J., and Pin, J.P. (1999). Molecular tinkering of G protein-coupled receptors: an evolutionary success. EMBO J 18, 1723-1729.

192

Boden, G. (2011). Obesity, insulin resistance and free fatty acids. Curr Opin Endocrinol Diabetes Obes 18, 139-143.

Bollan, K., King, D., Robertson, L.A., Brown, K., Taylor, P.M., Moss, S.J., and Connolly, C.N. (2003). GABA(A) receptor composition is determined by distinct assembly signals within alpha and beta subunits. J Biol Chem 278, 4747-4755.

Bonnert, T.P., McKernan, R.M., Farrar, S., le Bourdelles, B., Heavens, R.P., Smith, D.W., Hewson, L., Rigby, M.R., Sirinathsinghji, D.J., Brown, N., et al. (1999). theta, a novel gamma-aminobutyric acid type A receptor subunit. Proc Natl Acad Sci U S A 96, 9891-9896.

Boss, O., Samec, S., Dulloo, A., Seydoux, J., Muzzin, P., and Giacobino, J.P. (1997). Tissue-dependent upregulation of rat uncoupling protein-2 expression in response to fasting or cold. FEBS Lett 412, 111-114.

Bougneres, P., Stunff, C.L., Pecqueur, C., Pinglier, E., Adnot, P., and Ricquier, D. (1997). In vivo resistance of lipolysis to epinephrine. A new feature of childhood onset obesity. J Clin Invest 99, 2568-2573.

Brasaemle, D.L. (2007). Thematic review series: adipocyte biology. The perilipin family of structural lipid droplet proteins: stabilization of lipid droplets and control of lipolysis. J Lipid Res 48, 2547-2559.

Braulke, L.J., and Heldmaier, G. (2010). Torpor and ultradian rhythms require an intact signalling of the sympathetic nervous system. Cryobiology 60, 198-203.

Braulke, L.J., Klingenspor, M., DeBarber, A., Tobias, S.C., Grandy, D.K., Scanlan, T.S., and Heldmaier, G. (2008). 3-Iodothyronamine: a novel hormone controlling the balance between glucose and lipid utilisation. J Comp Physiol B 178, 167-177.

Bridges, T.M., and Lindsley, C.W. (2008). G-protein-coupled receptors: from classical modes of modulation to allosteric mechanisms. ACS Chem Biol 3, 530-541.

Briski, K.P., and Marshall, E.S. (2001). Induction of ependymal, glial, and neuronal transactivation by intraventricular administration of the SGLT1 Na+-D-glucose cotransporter inhibitor phlorizin. Neurochem Res 26, 783-792.

193

Brito, N.A., Brito, M.N., and Bartness, T.J. (2008). Differential sympathetic drive to adipose tissues after food deprivation, cold exposure or glucoprivation. Am J Physiol Regul Integr Comp Physiol 294, R1445-1452.

Brown, J.C., Gerson, A.R., and Staples, J.F. (2007). Mitochondrial metabolism during daily torpor in the dwarf Siberian hamster: role of active regulated changes and passive thermal effects. Am J Physiol Regul Integr Comp Physiol 293, R1833-1845.

Brown, J.C., and Staples, J.F. (2010). Mitochondrial metabolism during fasting-induced daily torpor in mice. Biochim Biophys Acta 1797, 476-486.

Brown, P.D., Davies, S.L., Speake, T., and Millar, I.D. (2004). Molecular mechanisms of cerebrospinal fluid production. Neuroscience 129, 957-970.

Bruni, J.E. (1998). Ependymal development, proliferation, and functions: a review. Microsc Res Tech 41, 2-13.

Cabrera-Vera, T.M., Vanhauwe, J., Thomas, T.O., Medkova, M., Preininger, A., Mazzoni, M.R., and Hamm, H.E. (2003). Insights into G protein structure, function, and regulation. Endocr Rev 24, 765-781.

Capsoni, S., Viswanathan, M., De Oliveira, A.M., and Saavedra, J.M. (1994). Characterization of melatonin receptors and signal transduction system in rat arteries forming the circle of Willis. Endocrinology 135, 373-378.

Carrillo-Vico, A., Reiter, R.J., Lardone, P.J., Herrera, J.L., Fernandez-Montesinos, R., Guerrero, J.M., and Pozo, D. (2006). The modulatory role of melatonin on immune responsiveness. Curr Opin Investig Drugs 7, 423-431.

Carvalho, S.D., Kimura, E.T., Bianco, A.C., and Silva, J.E. (1991). Central role of brown adipose tissue thyroxine 5'-deiodinase on thyroid hormone-dependent thermogenic response to cold. Endocrinology 128, 2149-2159.

Casey, P.J. (1994). Lipid modifications of G proteins. Curr Opin Cell Biol 6, 219-225.

Castillo, M., Hall, J.A., Correa-Medina, M., Ueta, C., Won Kang, H., Cohen, D.E., and Bianco, A.C. (2011). Disruption of thyroid hormone activation in type 2 deiodinase knockout mice causes obesity with glucose intolerance and liver steatosis only at thermoneutrality. Diabetes 60, 1082-1089.

194

Castro-Fernandez, C., Maya-Nunez, G., and Conn, P.M. (2005). Beyond the signal sequence: protein routing in health and disease. Endocr Rev 26, 479-503.

Cataldo, A.M., and Broadwell, R.D. (1986). Cytochemical identification of cerebral glycogen and glucose-6-phosphatase activity under normal and experimental conditions. II. Choroid plexus and ependymal epithelia, endothelia and pericytes. J Neurocytol 15, 511-524.

Cha, S.H., Hu, Z., Chohnan, S., and Lane, M.D. (2005). Inhibition of hypothalamic fatty acid synthase triggers rapid activation of fatty acid oxidation in skeletal muscle. Proc Natl Acad Sci U S A 102, 14557-14562.

Cha, S.H., Rodgers, J.T., Puigserver, P., Chohnan, S., and Lane, M.D. (2006). Hypothalamic malonyl-CoA triggers mitochondrial biogenesis and oxidative gene expression in skeletal muscle: Role of PGC-1alpha. Proc Natl Acad Sci U S A 103, 15410-15415.

Chao, L., Marcus-Samuels, B., Mason, M.M., Moitra, J., Vinson, C., Arioglu, E., Gavrilova, O., and Reitman, M.L. (2000). Adipose tissue is required for the antidiabetic, but not for the hypolipidemic, effect of thiazolidinediones. J Clin Invest 106, 1221-1228.

Chaves, V.E., Frasson, D., and Kawashita, N.H. (2011). Several agents and pathways regulate lipolysis in adipocytes. Biochimie 93 1631-1640.

Chen, J.L., Merl, D., Peterson, C.W., Wu, J., Liu, P.Y., Yin, H., Muoio, D.M., Ayer, D.E., West, M., and Chi, J.T. (2010). Lactic acidosis triggers starvation response with paradoxical induction of TXNIP through MondoA. PLoS Genet 6.

Chen, K.S., and DeLuca, H.F. (1994). Isolation and characterization of a novel cDNA from HL-60 cells treated with 1,25-dihydroxyvitamin D-3. Biochim Biophys Acta 1219, 26-32.

Chen, M., Wang, J., Dickerson, K.E., Kelleher, J., Xie, T., Gupta, D., Lai, E.W., Pacak, K., Gavrilova, O., and Weinstein, L.S. (2009). Central nervous system imprinting of the G protein G(s)alpha and its role in metabolic regulation. Cell Metab 9, 548-555.

Chen, Z., Lopez-Ramos, D.A., Yoshihara, E., Maeda, Y., Masutani, H., Sugie, K., Maeda, M., and Yodoi, J. (2011). Thioredoxin-binding protein-2 (TBP-2/VDUP1/TXNIP) regulates T-cell sensitivity to glucocorticoid during HTLV-I-induced transformation. Leukemia 25, 440-448.

195

Chiappini, F., Cunha, L.L., Harris, J.C., and Hollenberg, A.N. (2011). Lack of cAMP- response element-binding protein 1 in the hypothalamus causes obesity. J Biol Chem 286, 8094-8105.

Choi, S.M., Cho, H.J., Cho, H., Kim, K.H., Kim, J.B., and Park, H. (2008). Stra13/DEC1 and DEC2 inhibit sterol regulatory element binding protein-1c in a hypoxia-inducible factor-dependent mechanism. Nucleic Acids Res 36, 6372-6385.

Choi, S.M., Tucker, D.F., Gross, D.N., Easton, R.M., DiPilato, L.M., Dean, A.S., Monks, B.R., and Birnbaum, M.J. (2010). Insulin regulates adipocyte lipolysis via an Akt- independent signaling pathway. Mol Cell Biol 30, 5009-5020.

Choi, Y.H., Hartzell, D., Azain, M.J., and Baile, C.A. (2002). TRH decreases food intake and increases water intake and body temperature in rats. Physiol Behav 77, 1-4.

Chung, S., Funakoshi, T., and Civelli, O. (2008). Orphan GPCR research. Br J Pharmacol 153 Suppl 1, S339-346.

Chutkow, W.A., Birkenfeld, A.L., Brown, J.D., Lee, H.Y., Frederick, D.W., Yoshioka, J., Patwari, P., Kursawe, R., Cushman, S.W., Plutzky, J., et al. (2010). Deletion of the alpha-arrestin protein Txnip in mice promotes adiposity and adipogenesis while preserving insulin sensitivity. Diabetes 59, 1424-1434.

Colombani, A.L., Carneiro, L., Benani, A., Galinier, A., Jaillard, T., Duparc, T., Offer, G., Lorsignol, A., Magnan, C., Casteilla, L., et al. (2009). Enhanced hypothalamic glucose sensing in obesity: alteration of redox signaling. Diabetes 58, 2189-2197.

Confente, F., Rendon, M.C., Besseau, L., Falcon, J., and Munoz-Cueto, J.A. (2010). Melatonin receptors in a pleuronectiform species, Solea senegalensis: Cloning, tissue expression, day-night and seasonal variations. Gen Comp Endocrinol 167, 202-214.

Conway, S., Canning, S.J., Barrett, P., Guardiola-Lemaitre, B., Delagrange, P., and Morgan, P.J. (1997). The roles of valine 208 and histidine 211 in ligand binding and receptor function of the ovine Mel1a beta melatonin receptor. Biochem Biophys Res Commun 239, 418-423.

Conway, S., Drew, J.E., Mowat, E.S., Barrett, P., Delagrange, P., and Morgan, P.J. (2000). Chimeric melatonin mt1 and melatonin-related receptors. Identification of domains and residues participating in ligand binding and receptor activation of the melatonin mt1 receptor. J Biol Chem 275, 20602-20609.

196

Cook, J.L., Mills, S.J., Naquin, R.T., Alam, J., and Re, R.N. (2007). Cleavage of the angiotensin II type 1 receptor and nuclear accumulation of the cytoplasmic carboxy- terminal fragment. Am J Physiol Cell Physiol 292, C1313-1322.

Coppack, S.W., Persson, M., Judd, R.L., and Miles, J.M. (1999). Glycerol and nonesterified fatty acid metabolism in human muscle and adipose tissue in vivo. Am J Physiol 276, E233-240.

Coppola, A., Liu, Z.W., Andrews, Z.B., Paradis, E., Roy, M.C., Friedman, J.M., Ricquier, D., Richard, D., Horvath, T.L., Gao, X.B., et al. (2007). A central thermogenic- like mechanism in feeding regulation: an interplay between arcuate nucleus T3 and UCP2. Cell Metab 5, 21-33.

Cordoba-Chacon, J., Gahete, M.D., Duran-Prado, M., Luque, R.M., and Castano, J.P. (2011). Truncated somatostatin receptors as new players in somatostatin-cortistatin pathophysiology. Ann N Y Acad Sci 1220, 6-15.

Coupar, I.M., Desmond, P.V., and Irving, H.R. (2007). Human 5-HT(4) and 5-HT(7) receptor splice variants: are they important? Curr Neuropharmacol 5, 224-231.

Cowley, M.A., Pronchuk, N., Fan, W., Dinulescu, D.M., Colmers, W.F., and Cone, R.D. (1999). Integration of NPY, AGRP, and melanocortin signals in the hypothalamic paraventricular nucleus: evidence of a cellular basis for the adipostat. Neuron 24, 155- 163.

Cowley, M.A., Smart, J.L., Rubinstein, M., Cerdan, M.G., Diano, S., Horvath, T.L., Cone, R.D., and Low, M.J. (2001). Leptin activates anorexigenic POMC neurons through a neural network in the arcuate nucleus. Nature 411, 480-484.

Crantz, F.R., and Larsen, P.R. (1980). Rapid thyroxine to 3,5,3'-triiodothyronine conversion and nuclear 3,5,3'-triiodothyronine binding in rat cerebral cortex and cerebellum. J Clin Invest 65, 935-938.

Crawford, F.I., Hodgkinson, C.L., Ivanova, E., Logunova, L.B., Evans, G.J., Steinlechner, S., and Loudon, A.S. (2007). Influence of torpor on cardiac expression of genes involved in the circadian clock and protein turnover in the Siberian hamster (Phodopus sungorus). Physiol Genomics 31, 521-530.

197

Daikoku, T., Shinohara, Y., Shima, A., Yamazaki, N., and Terada, H. (2000). Specific elevation of transcript levels of particular protein subtypes induced in brown adipose tissue by cold exposure. Biochim Biophys Acta 1457, 263-272.

Dalrymple, M.B., Pfleger, K.D., and Eidne, K.A. (2008). G protein-coupled receptor dimers: functional consequences, disease states and drug targets. Pharmacol Ther 118, 359-371.

Dark, J., and Pelz, K.M. (2008). NPY Y1 receptor antagonist prevents NPY-induced torpor-like hypothermia in cold-acclimated Siberian hamsters. Am J Physiol Regul Integr Comp Physiol 294, R236-245.

Date, Y., Ueta, Y., Yamashita, H., Yamaguchi, H., Matsukura, S., Kangawa, K., Sakurai, T., Yanagisawa, M., and Nakazato, M. (1999). Orexins, orexigenic hypothalamic peptides, interact with autonomic, neuroendocrine and neuroregulatory systems. Proc Natl Acad Sci U S A 96, 748-753.

Daulat, A.M., Maurice, P., Froment, C., Guillaume, J.L., Broussard, C., Monsarrat, B., Delagrange, P., and Jockers, R. (2007). Purification and identification of G protein- coupled receptor protein complexes under native conditions. Mol Cell Proteomics 6, 835-844. de Jesus, L.A., Carvalho, S.D., Ribeiro, M.O., Schneider, M., Kim, S.W., Harney, J.W., Larsen, P.R., and Bianco, A.C. (2001). The type 2 iodothyronine deiodinase is essential for adaptive thermogenesis in brown adipose tissue. J Clin Invest 108, 1379- 1385.

De Vries, L., Zheng, B., Fischer, T., Elenko, E., and Farquhar, M.G. (2000). The regulator of G protein signaling family. Annu Rev Pharmacol Toxicol 40, 235-271.

Del Bigio, M.R. (1995). The ependyma: a protective barrier between brain and cerebrospinal fluid. Glia 14, 1-13.

Delavest, M., Even, C., Benjemaa, N., Poirier, M.F., Jockers, R., and Krebs, M.O. (2011). Association of the intronic rs2072621 polymorphism of the X-linked GPR50 gene with affective disorder with seasonal pattern. Eur Psychiatry.

DeWire, S.M., Ahn, S., Lefkowitz, R.J., and Shenoy, S.K. (2007). Beta-arrestins and cell signaling. Annu Rev Physiol 69, 483-510.

198

Diano, S., Naftolin, F., Goglia, F., and Horvath, T.L. (1998). Fasting-induced increase in type II iodothyronine deiodinase activity and messenger ribonucleic acid levels is not reversed by thyroxine in the rat hypothalamus. Endocrinology 139, 2879-2884.

Dimicco, J.A., and Zaretsky, D.V. (2007). The dorsomedial hypothalamus: a new player in thermoregulation. Am J Physiol Regul Integr Comp Physiol 292, R47-63.

Donnelly, K.L., Margosian, M.R., Sheth, S.S., Lusis, A.J., and Parks, E.J. (2004). Increased lipogenesis and fatty acid reesterification contribute to hepatic triacylglycerol stores in hyperlipidemic Txnip-/- mice. J Nutr 134, 1475-1480.

Drake, M.T., Shenoy, S.K., and Lefkowitz, R.J. (2006). Trafficking of G protein-coupled receptors. Circ Res 99, 570-582.

Dreau, D., Morton, D.S., Foster, M., Fowler, N., and Sonnenfeld, G. (1998). Effects of 2-deoxy-D-glucose administration on immune parameters in mice. Immunopharmacology 39, 201-213.

Dreja, T., Jovanovic, Z., Rasche, A., Kluge, R., Herwig, R., Tung, Y.C., Joost, H.G., Yeo, G.S., and Al-Hasani, H. (2010). Diet-induced gene expression of isolated from a polygenic mouse model of the metabolic syndrome. Diabetologia 53, 309-320.

Drew, J.E., Barrett, P., Mercer, J.G., Moar, K.M., Canet, E., Delagrange, P., and Morgan, P.J. (2001). Localization of the melatonin-related receptor in the rodent brain and peripheral tissues. J Neuroendocrinol 13, 453-458.

Drew, J.E., Barrett, P., Williams, L.M., Conway, S., and Morgan, P.J. (1998). The ovine melatonin-related receptor: cloning and preliminary distribution and binding studies. J Neuroendocrinol 10, 651-661.

Dubocovich, M.L., Delagrange, P., Krause, D.N., Sugden, D., Cardinali, D.P., and Olcese, J. (2010). International Union of Basic and Clinical Pharmacology. LXXV. Nomenclature, classification, and pharmacology of G protein-coupled melatonin receptors. Pharmacol Rev 62, 343-380.

Dufourny, L., Levasseur, A., Migaud, M., Callebaut, I., Pontarotti, P., Malpaux, B., and Monget, P. (2008). GPR50 is the mammalian ortholog of Mel1c: evidence of rapid evolution in mammals. BMC Evol Biol 8, 105.

199

Ebling, F.J., Arthurs, O.J., Turney, B.W., and Cronin, A.S. (1998). Seasonal neuroendocrine rhythms in the male Siberian hamster persist after monosodium glutamate-induced lesions of the arcuate nucleus in the neonatal period. J Neuroendocrinol 10, 701-712.

Eddy, S.F., and Storey, K.B. (2003). Differential expression of Akt, PPARgamma, and PGC-1 during hibernation in bats. Biochem Cell Biol 81, 269-274.

Eddy, S.F., and Storey, K.B. (2004). Up-regulation of fatty acid-binding proteins during hibernation in the little brown bat, Myotis lucifugus. Biochim Biophys Acta 1676, 63-70.

Edwards, C.M., Abusnana, S., Sunter, D., Murphy, K.G., Ghatei, M.A., and Bloom, S.R. (1999). The effect of the orexins on food intake: comparison with neuropeptide Y, melanin-concentrating hormone and galanin. J Endocrinol 160, R7-12.

Eglen, R.M. (2005). An Overview of High Throughput Screening at G Protein Coupled Receptors. Frontiers in Drug Design & Discovery 1, 97-111.

Evans, S.B., Wilkinson, C.W., Gronbeck, P., Bennett, J.L., Zavosh, A., Taborsky, G.J., Jr., and Figlewicz, D.P. (2004). Inactivation of the DMH selectively inhibits the ACTH and corticosterone responses to hypoglycemia. Am J Physiol Regul Integr Comp Physiol 286, R123-128.

Fang, S., Jin, Y., Zheng, H., Yan, J., Cui, Y., Bi, H., Jia, H., Zhang, H., Wang, Y., Na, L., et al. (2011). High glucose condition upregulated Txnip expression level in rat mesangial cells through ROS/MEK/MAPK pathway. Mol Cell Biochem 347, 175-182.

Farrell, C.L., Yang, J., and Pardridge, W.M. (1992). GLUT-1 glucose transporter is present within apical and basolateral membranes of brain epithelial interfaces and in microvascular endothelia with and without tight junctions. J Histochem Cytochem 40, 193-199.

Farrens, D.L., Altenbach, C., Yang, K., Hubbell, W.L., and Khorana, H.G. (1996). Requirement of rigid-body motion of transmembrane helices for light activation of rhodopsin. Science 274, 768-770.

Fekete, C., and Lechan, R.M. (2007). Negative feedback regulation of hypophysiotropic thyrotropin-releasing hormone (TRH) synthesizing neurons: role of neuronal afferents and type 2 deiodinase. Front Neuroendocrinol 28, 97-114.

200

Fekete, C., Legradi, G., Mihaly, E., Tatro, J.B., Rand, W.M., and Lechan, R.M. (2000). alpha-Melanocyte stimulating hormone prevents fasting-induced suppression of corticotropin-releasing hormone gene expression in the rat hypothalamic paraventricular nucleus. Neurosci Lett 289, 152-156.

Fleury, C., Neverova, M., Collins, S., Raimbault, S., Champigny, O., Levi-Meyrueis, C., Bouillaud, F., Seldin, M.F., Surwit, R.S., Ricquier, D., et al. (1997). Uncoupling protein- 2: a novel gene linked to obesity and hyperinsulinemia. Nat Genet 15, 269-272.

Flynn, F.W., and Grill, H.J. (1985). Fourth ventricular phlorizin dissociates feeding from hyperglycemia in rats. Brain Res 341, 331-336.

Follett, B.K., and Nicholls, T.J. (1985). Influences of thyroidectomy and thyroxine replacement on photoperiodically controlled reproduction in quail. J Endocrinol 107, 211-221.

Forrester, M.T., Seth, D., Hausladen, A., Eyler, C.E., Foster, M.W., Matsumoto, A., Benhar, M., Marshall, H.E., and Stamler, J.S. (2009). Thioredoxin-interacting protein (Txnip) is a feedback regulator of S-nitrosylation. J Biol Chem 284, 36160-36166.

Foster, L.A., Ames, N.K., and Emery, R.S. (1991). Food intake and serum insulin responses to intraventricular infusions of insulin and IGF-I. Physiol Behav 50, 745-749.

Frayling, C., Britton, R., and Dale, N. (2011). ATP-mediated glucosensing by hypothalamic tanycytes. J Physiol 589, 2275-2286.

Frederich, R.C., Hamann, A., Anderson, S., Lollmann, B., Lowell, B.B., and Flier, J.S. (1995). Leptin levels reflect body lipid content in mice: evidence for diet-induced resistance to leptin action. Nat Med 1, 1311-1314.

Freeman, D.A., Lewis, D.A., Kauffman, A.S., Blum, R.M., and Dark, J. (2004). Reduced leptin concentrations are permissive for display of torpor in Siberian hamsters. Am J Physiol Regul Integr Comp Physiol 287, R97-R103.

Froy, O. (2007). The relationship between nutrition and circadian rhythms in mammals. Front Neuroendocrinol 28, 61-71.

Fruhbeck, G., Aguado, M., and Martinez, J.A. (1997). In vitro lipolytic effect of leptin on mouse adipocytes: evidence for a possible autocrine/paracrine role of leptin. Biochem Biophys Res Commun 240, 590-594.

201

Fuchtenbusch, M., Standl, E., and Schatz, H. (2000). Clinical efficacy of new thiazolidinediones and glinides in the treatment of mellitus. Exp Clin Endocrinol Diabetes 108, 151-163.

Fuzesi, T., Wittmann, G., Liposits, Z., Lechan, R.M., and Fekete, C. (2007). Contribution of noradrenergic and adrenergic cell groups of the brainstem and agouti- related protein-synthesizing neurons of the arcuate nucleus to neuropeptide-y innervation of corticotropin-releasing hormone neurons in hypothalamic paraventricular nucleus of the rat. Endocrinology 148, 5442-5450.

Gangal, R., and Kumar, K.K. (2007). Reduced alphabet motif methodology for GPCR annotation. J Biomol Struct Dyn 25, 299-310.

Ganguly, S., Coon, S.L., and Klein, D.C. (2002). Control of melatonin synthesis in the mammalian pineal gland: the critical role of serotonin acetylation. Cell Tissue Res 309, 127-137.

Gao, Q., and Horvath, T.L. (2008). Neuronal control of energy homeostasis. FEBS Lett 582, 132-141.

Garcia, M.A., Millan, C., Balmaceda-Aguilera, C., Castro, T., Pastor, P., Montecinos, H., Reinicke, K., Zuniga, F., Vera, J.C., Onate, S.A., et al. (2003). Hypothalamic ependymal-glial cells express the glucose transporter GLUT2, a protein involved in glucose sensing. J Neurochem 86, 709-724.

Gavrilova, O., Leon, L.R., Marcus-Samuels, B., Mason, M.M., Castle, A.L., Refetoff, S., Vinson, C., and Reitman, M.L. (1999). Torpor in mice is induced by both leptin- dependent and -independent mechanisms. Proc Natl Acad Sci U S A 96, 14623-14628.

Geiser, F. (2004). Metabolic rate and body temperature reduction during hibernation and daily torpor. Annu Rev Physiol 66, 239-274.

Geiser, F., and Baudinette, R. (1985). The influence of temperature and photophase on daily torpor in Sminthopsis macroura (Dasyuridae: Marsupialia). Journal of Comparative Physiology B: Biochemical, Systemic, and Environmental Physiology 156, 129-134.

Geiser, F., and Ruf, T. (1995). Hibernation versus Daily Torpor in Mammals and Birds: Physiological Variables and Classification of Torpor Patterns. Physiological Zoology 68, 935-966.

202

Gerstner, J.R., Bremer, Q.Z., Vander Heyden, W.M., Lavaute, T.M., Yin, J.C., and Landry, C.F. (2008). Brain fatty acid binding protein (Fabp7) is diurnally regulated in astrocytes and hippocampal granule cell precursors in adult rodent brain. PLoS One 3, e1631.

Gerstner, J.R., Vander Heyden, W.M., Lavaute, T.M., and Landry, C.F. (2006). Profiles of novel diurnally regulated genes in mouse hypothalamus: expression analysis of the cysteine and histidine-rich domain-containing, zinc-binding protein 1, the fatty acid- binding protein 7 and the GTPase, ras-like family member 11b. Neuroscience 139, 1435-1448.

Gerstner, J.R., Vanderheyden, W.M., Shaw, P.J., Landry, C.F., and Yin, J.C. (2011). Fatty-acid binding proteins modulate sleep and enhance long-term memory consolidation in Drosophila. PLoS One 6, e15890.

Gillard, E.R., Khan, A.M., Ahsan ul, H., Grewal, R.S., Mouradi, B., and Stanley, B.G. (1997). Stimulation of eating by the second messenger cAMP in the perifornical and lateral hypothalamus. Am J Physiol 273, R107-112.

Gluck, E.F., Stephens, N., and Swoap, S.J. (2006). Peripheral ghrelin deepens torpor bouts in mice through the arcuate nucleus neuropeptide Y signaling pathway. Am J Physiol Regul Integr Comp Physiol 291, R1303-1309.

Godson, C., and Reppert, S.M. (1997). The Mel1a melatonin receptor is coupled to parallel signal transduction pathways. Endocrinology 138, 397-404.

Gonzales, A.M., and Orlando, R.A. (2007). Role of adipocyte-derived lipoprotein lipase in adipocyte hypertrophy. Nutr Metab (Lond) 4, 22.

Green, H., and Kehinde, O. (1979). Formation of normally differentiated subcutaneous fat pads by an established preadipose cell line. J Cell Physiol 101, 169-171.

Griffiths, E.C., Rothwell, N.J., and Stock, M.J. (1988). Thermogenic effects of thyrotrophin-releasing hormone and its analogues in the rat. Experientia 44, 40-42.

Grosse, R., Schoneberg, T., Schultz, G., and Gudermann, T. (1997). Inhibition of gonadotropin-releasing hormone receptor signaling by expression of a splice variant of the human receptor. Mol Endocrinol 11, 1305-1318.

203

Grossman, E., Laudon, M., Yalcin, R., Zengil, H., Peleg, E., Sharabi, Y., Kamari, Y., Shen-Orr, Z., and Zisapel, N. (2006). Melatonin reduces night blood pressure in patients with nocturnal hypertension. Am J Med 119, 898-902.

Grunewald, E., Kinnell, H.L., Porteous, D.J., and Thomson, P.A. (2009). GPR50 interacts with neuronal NOGO-A and affects neurite outgrowth. Mol Cell Neurosci 42, 363-371.

Gubitz, A.K., and Reppert, S.M. (2000). Chimeric and point-mutated receptors reveal that a single glycine residue in transmembrane domain 6 is critical for high affinity melatonin binding. Endocrinology 141, 1236-1244.

Guillaumond, F., Lacoche, S., Dulong, S., Grechez-Cassiau, A., Filipski, E., Li, X.M., Levi, F., Berra, E., Delaunay, F., and Teboul, M. (2008). Altered Stra13 and Dec2 circadian gene expression in hypoxic cells. Biochem Biophys Res Commun 369, 1184- 1189.

Haemmerle, G., Zimmermann, R., and Zechner, R. (2003). Letting lipids go: hormone- sensitive lipase. Curr Opin Lipidol 14, 289-297.

Hamm, H.E. (1998). The many faces of G protein signaling. J Biol Chem 273, 669-672.

Hansen, J.L., and Sheikh, S.P. (2004). Functional consequences of 7TM receptor dimerization. Eur J Pharm Sci 23, 301-317.

Harada, N., Yamada, Y., Tsukiyama, K., Yamada, C., Nakamura, Y., Mukai, E., Hamasaki, A., Liu, X., Toyoda, K., Seino, Y., et al. (2008). A novel GIP receptor splice variant influences GIP sensitivity of pancreatic beta-cells in obese mice. Am J Physiol Endocrinol Metab 294, E61-68.

Hardeland, R., Cardinali, D.P., Srinivasan, V., Spence, D.W., Brown, G.M., and Pandi- Perumal, S.R. (2011). Melatonin--a pleiotropic, orchestrating regulator molecule. Prog Neurobiol 93, 350-384.

Haynes, W.G., Morgan, D.A., Walsh, S.A., Mark, A.L., and Sivitz, W.I. (1997). Receptor-mediated regional sympathetic nerve activation by leptin. J Clin Invest 100, 270-278.

Head, B.P., Patel, H.H., Roth, D.M., Lai, N.C., Niesman, I.R., Farquhar, M.G., and Insel, P.A. (2005). G-protein-coupled receptor signaling components localize in both

204 sarcolemmal and intracellular caveolin-3-associated microdomains in adult cardiac myocytes. J Biol Chem 280, 31036-31044.

Heinrichs, S.C., Menzaghi, F., Pich, E.M., Hauger, R.L., and Koob, G.F. (1993). Corticotropin-releasing factor in the paraventricular nucleus modulates feeding induced by neuropeptide Y. Brain Res 611, 18-24.

Heldmaier, G., Klingenspor, M., Werneyer, M., Lampi, B.J., Brooks, S.P., and Storey, K.B. (1999). Metabolic adjustments during daily torpor in the Djungarian hamster. Am J Physiol 276, E896-906.

Heldmaier, G., Ortmann, S., and Elvert, R. (2004). Natural hypometabolism during hibernation and daily torpor in mammals. Respir Physiol Neurobiol 141, 317-329.

Herman, J.P., Figueiredo, H., Mueller, N.K., Ulrich-Lai, Y., Ostrander, M.M., Choi, D.C., and Cullinan, W.E. (2003). Central mechanisms of stress integration: hierarchical circuitry controlling hypothalamo-pituitary-adrenocortical responsiveness. Front Neuroendocrinol 24, 151-180.

Herwig, A., Revel, F., Saboureau, M., Pevet, P., and Steinlechner, S. (2006). Daily torpor alters multiple gene expression in the suprachiasmatic nucleus and pineal gland of the Djungarian hamster (Phodopus sungorus). Chronobiol Int 23, 269-276.

Herwig, A., Saboureau, M., Pevet, P., and Steinlechner, S. (2007). Daily torpor affects the molecular machinery of the circadian clock in Djungarian hamsters (Phodopus sungorus). Eur J Neurosci 26, 2739-2746.

Herzig, S., Hedrick, S., Morantte, I., Koo, S.H., Galimi, F., and Montminy, M. (2003). CREB controls hepatic lipid metabolism through nuclear hormone receptor PPAR- gamma. Nature 426, 190-193.

Hiebert, S.M., Fulkerson, E.K., Lindermayer, K.T., McClure, S.D., (2000). Effect of temperature on preference for dietary unsaturated fatty acids in the Djungarian hamster (Phodopus sungorus). Canadian Journal of Zoology 78, 1361-1368.

Hillebrand, J.J., de Wied, D., and Adan, R.A. (2002). Neuropeptides, food intake and body weight regulation: a hypothalamic focus. Peptides 23, 2283-2306.

Himms-Hagen, J. (1985). Defective brown adipose tissue thermogenesis in obese mice. Int J Obes 9 Suppl 2, 17-24.

205

Hittel, D., and Storey, K.B. (2001). Differential expression of adipose- and heart-type fatty acid binding proteins in hibernating ground squirrels. Biochim Biophys Acta 1522, 238-243.

Honma, S., Kawamoto, T., Takagi, Y., Fujimoto, K., Sato, F., Noshiro, M., Kato, Y., and Honma, K. (2002). Dec1 and Dec2 are regulators of the mammalian molecular clock. Nature 419, 841-844.

Horvath, T.L., Andrews, Z.B., and Diano, S. (2009). Fuel utilization by hypothalamic neurons: roles for ROS. Trends Endocrinol Metab 20, 78-87.

Hua, X., Yokoyama, C., Wu, J., Briggs, M.R., Brown, M.S., Goldstein, J.L., and Wang, X. (1993). SREBP-2, a second basic-helix-loop-helix-leucine zipper protein that stimulates transcription by binding to a sterol regulatory element. Proc Natl Acad Sci U S A 90, 11603-11607.

Hudasek, K., Brown, S.T., and Fearon, I.M. (2004). H2O2 regulates recombinant Ca2+ channel alpha1C subunits but does not mediate their sensitivity to acute hypoxia. Biochem Biophys Res Commun 318, 135-141.

Hudson, J.W., and Scott, I.M. (1979). Daily Torpor in the , Mus musculus Var. Albino. Physiological Zoology 52, 205-218.

Hui, S.T., Andres, A.M., Miller, A.K., Spann, N.J., Potter, D.W., Post, N.M., Chen, A.Z., Sachithanantham, S., Jung, D.Y., Kim, J.K., et al. (2008). Txnip balances metabolic and growth signaling via PTEN disulfide reduction. Proc Natl Acad Sci U S A 105, 3921-3926.

Hunt, A.E., Al-Ghoul, W.M., Gillette, M.U., and Dubocovich, M.L. (2001). Activation of MT(2) melatonin receptors in rat suprachiasmatic nucleus phase advances the circadian clock. Am J Physiol Cell Physiol 280, C110-118.

Inagaki, T., Dutchak, P., Zhao, G., Ding, X., Gautron, L., Parameswara, V., Li, Y., Goetz, R., Mohammadi, M., Esser, V., et al. (2007). Endocrine regulation of the fasting response by PPARalpha-mediated induction of fibroblast growth factor 21. Cell Metab 5, 415-425.

Inokuma, K., Ogura-Okamatsu, Y., Toda, C., Kimura, K., Yamashita, H., and Saito, M. (2005). Uncoupling protein 1 is necessary for norepinephrine-induced glucose utilization in brown adipose tissue. Diabetes 54, 1385-1391.

206

Itoi, K., Jiang, Y.Q., Iwasaki, Y., and Watson, S.J. (2004). Regulatory mechanisms of corticotropin-releasing hormone and vasopressin gene expression in the hypothalamus. J Neuroendocrinol 16, 348-355.

Ivanova, E.A., Bechtold, D.A., Dupre, S.M., Brennand, J., Barrett, P., Luckman, S.M., and Loudon, A.S. (2008). Altered metabolism in the melatonin-related receptor (GPR50) knockout mouse. Am J Physiol Endocrinol Metab 294, E176-182.

Jacoby, E., Bouhelal, R., Gerspacher, M., and Seuwen, K. (2006). The 7 TM G-protein- coupled receptor target family. ChemMedChem 1, 761-782.

Jensen, M.D., Haymond, M.W., Rizza, R.A., Cryer, P.E., and Miles, J.M. (1989). Influence of body fat distribution on free fatty acid metabolism in obesity. J Clin Invest 83, 1168-1173.

Jetton, T.L., Liang, Y., Pettepher, C.C., Zimmerman, E.C., Cox, F.G., Horvath, K., Matschinsky, F.M., and Magnuson, M.A. (1994). Analysis of upstream glucokinase promoter activity in transgenic mice and identification of glucokinase in rare neuroendocrine cells in the brain and gut. J Biol Chem 269, 3641-3654.

Jinnah, H.A., Wojcik, B.E., Hunt, M., Narang, N., Lee, K.Y., Goldstein, M., Wamsley, J.K., Langlais, P.J., and Friedmann, T. (1994). Dopamine deficiency in a genetic mouse model of Lesch-Nyhan disease. J Neurosci 14, 1164-1175.

Jocken, J.W., and Blaak, E.E. (2008). Catecholamine-induced lipolysis in adipose tissue and skeletal muscle in obesity. Physiol Behav 94, 219-230.

Jockers, R., Maurice, P., Boutin, J.A., and Delagrange, P. (2008). Melatonin receptors, heterodimerization, signal transduction and binding sites: what's new? Br J Pharmacol 154, 1182-1195.

Johnston, C.A., and Siderovski, D.P. (2007). Receptor-mediated activation of heterotrimeric G-proteins: current structural insights. Mol Pharmacol 72, 219-230.

Jordan, S.D., Konner, A.C., and Bruning, J.C. (2010). Sensing the fuels: glucose and lipid signaling in the CNS controlling energy homeostasis. Cell Mol Life Sci 67, 3255- 3273.

207

Junn, E., Han, S.H., Im, J.Y., Yang, Y., Cho, E.W., Um, H.D., Kim, D.K., Lee, K.W., Han, P.L., Rhee, S.G., et al. (2000). Vitamin D3 up-regulated protein 1 mediates oxidative stress via suppressing the thioredoxin function. J Immunol 164, 6287-6295.

Kaadige, M.R., Looper, R.E., Kamalanaadhan, S., and Ayer, D.E. (2009). Glutamine- dependent anapleurosis dictates glucose uptake and cell growth by regulating MondoA transcriptional activity. Proc Natl Acad Sci U S A 106, 14878-14883.

Karar, J., Dolt, K.S., Mishra, M.K., Arif, E., Javed, S., and Pasha, M.A. (2007). Expression and functional activity of pro-oxidants and in murine heart exposed to acute hypobaric hypoxia. FEBS Lett 581, 4577-4582.

Karasek, M. (2004). Melatonin, human aging, and age-related diseases. Exp Gerontol 39, 1723-1729.

Karpa, K.D., Lin, R., Kabbani, N., and Levenson, R. (2000). The dopamine D3 receptor interacts with itself and the truncated D3 splice variant d3nf: D3-D3nf interaction causes mislocalization of D3 receptors. Mol Pharmacol 58, 677-683.

Kawamoto, T., Noshiro, M., Furukawa, M., Honda, K.K., Nakashima, A., Ueshima, T., Usui, E., Katsura, Y., Fujimoto, K., Honma, S., et al. (2006). Effects of fasting and re- feeding on the expression of Dec1, Per1, and other clock-related genes. J Biochem 140, 401-408.

Kennaway, D.J., and Wright, H. (2002). Melatonin and circadian rhythms. Curr Top Med Chem 2, 199-209.

Kim, J.M., Altenbach, C., Thurmond, R.L., Khorana, H.G., and Hubbell, W.L. (1997). Structure and function in rhodopsin: rhodopsin mutants with a neutral amino acid at E134 have a partially activated conformation in the dark state. Proc Natl Acad Sci U S A 94, 14273-14278.

Kim, K.Y., Shin, S.M., Kim, J.K., Paik, S.G., Yang, Y., and Choi, I. (2004). Heat shock factor regulates VDUP1 gene expression. Biochem Biophys Res Commun 315, 369- 375.

Kitahara, T., Li-Korotky, H.S., and Balaban, C.D. (2005). Regulation of mitochondrial uncoupling proteins in mouse inner ear ganglion cells in response to systemic kanamycin challenge. Neuroscience 135, 639-653.

208

Klein, D.C., Coon, S.L., Roseboom, P.H., Weller, J.L., Bernard, M., Gastel, J.A., Zatz, M., Iuvone, P.M., Rodriguez, I.R., Begay, V., et al. (1997). The melatonin rhythm- generating enzyme: molecular regulation of serotonin N-acetyltransferase in the pineal gland. Recent Prog Horm Res 52, 307-357; discussion 357-308.

Kobayashi, H., Oishi, K., Hanai, S., and Ishida, N. (2004). Effect of feeding on peripheral circadian rhythms and behaviour in mammals. Genes Cells 9, 857-864.

Kokoeva, M.V., Yin, H., and Flier, J.S. (2007). Evidence for constitutive neural cell proliferation in the adult murine hypothalamus. J Comp Neurol 505, 209-220.

Kortner, G., and Geiser, F. (2000). The temporal organization of daily torpor and hibernation: circadian and circannual rhythms. Chronobiol Int 17, 103-128.

Kostenis, E., Waelbroeck, M., and Milligan, G. (2005). Techniques: promiscuous Galpha proteins in basic research and drug discovery. Trends Pharmacol Sci 26, 595- 602.

Kroeze, W.K., Sheffler, D.J., and Roth, B.L. (2003). G-protein-coupled receptors at a glance. J Cell Sci 116, 4867-4869.

Krysiak, R., Obuchowicz, E., and Herman, Z.S. (1999). Interactions between the neuropeptide Y system and the hypothalamic-pituitary-adrenal axis. Eur J Endocrinol 140, 130-136.

Kubota, N., Terauchi, Y., Miki, H., Tamemoto, H., Yamauchi, T., Komeda, K., Satoh, S., Nakano, R., Ishii, C., Sugiyama, T., et al. (1999). PPAR gamma mediates high-fat diet- induced adipocyte hypertrophy and insulin resistance. Mol Cell 4, 597-609.

Kurtz, A., Zimmer, A., Schnutgen, F., Bruning, G., Spener, F., and Muller, T. (1994). The expression pattern of a novel gene encoding brain-fatty acid binding protein correlates with neuronal and glial cell development. Development 120, 2637-2649.

Lane, M.D., Wolfgang, M., Cha, S.H., and Dai, Y. (2008). Regulation of food intake and energy expenditure by hypothalamic malonyl-CoA. Int J Obes (Lond) 32 Suppl 4, S49- 54.

Larkin, J.E., Yellon, S.M., and Zucker, I. (2003). Melatonin production accompanies arousal from daily torpor in Siberian hamsters. Physiol Biochem Zool 76, 577-585.

209

Le Foll, C., Irani, B.G., Magnan, C., Dunn-Meynell, A.A., and Levin, B.E. (2009). Characteristics and mechanisms of hypothalamic neuronal fatty acid sensing. Am J Physiol Regul Integr Comp Physiol 297, R655-664.

Le Jan, S., Le Meur, N., Cazes, A., Philippe, J., Le Cunff, M., Leger, J., Corvol, P., and Germain, S. (2006). Characterization of the expression of the hypoxia-induced genes neuritin, TXNIP and IGFBP3 in cancer. FEBS Lett 580, 3395-3400.

Le Lay, S., Krief, S., Farnier, C., Lefrere, I., Le Liepvre, X., Bazin, R., Ferre, P., and Dugail, I. (2001). Cholesterol, a cell size-dependent signal that regulates glucose metabolism and gene expression in adipocytes. J Biol Chem 276, 16904-16910.

Lechan, R.M., and Fekete, C. (2006). The TRH neuron: a hypothalamic integrator of energy metabolism. Prog Brain Res 153, 209-235.

Lee, S.P., O'Dowd, B.F., Ng, G.Y., Varghese, G., Akil, H., Mansour, A., Nguyen, T., and George, S.R. (2000). Inhibition of cell surface expression by mutant receptors demonstrates that D2 dopamine receptors exist as oligomers in the cell. Mol Pharmacol 58, 120-128.

Lee, T.M., Holmes, W.G., and Zucker, I. (1990). Temperature dependence of circadian rhythms in golden-mantled ground squirrels. J Biol Rhythms 5, 25-34.

Leloup, C., Arluison, M., Lepetit, N., Cartier, N., Marfaing-Jallat, P., Ferre, P., and Penicaud, L. (1994). Glucose transporter 2 (GLUT 2): expression in specific brain nuclei. Brain Res 638, 221-226.

Leloup, C., Magnan, C., Benani, A., Bonnet, E., Alquier, T., Offer, G., Carriere, A., Periquet, A., Fernandez, Y., Ktorza, A., et al. (2006). Mitochondrial reactive oxygen species are required for hypothalamic glucose sensing. Diabetes 55, 2084-2090.

Levine, A.S., Jewett, D.C., Cleary, J.P., Kotz, C.M., and Billington, C.J. (2004). Our journey with neuropeptide Y: effects on ingestive behaviors and energy expenditure. Peptides 25, 505-510.

Levoye, A., Dam, J., Ayoub, M.A., Guillaume, J.L., Couturier, C., Delagrange, P., and Jockers, R. (2006). The orphan GPR50 receptor specifically inhibits MT1 melatonin receptor function through heterodimerization. EMBO J 25, 3012-3023.

210

Li, J., Hand, L.E., Meng, Q.J., Loudon, A.S., and Bechtold, D.A. (2011). GPR50 Interacts with TIP60 to Modulate Glucocorticoid Receptor Signalling. PLoS One 6, e23725.

Liu, X., Perusse, F., and Bukowiecki, L.J. (1994). Chronic norepinephrine infusion stimulates glucose uptake in white and brown adipose tissues. Am J Physiol 266, R914-920.

Livak, K.J., and Schmittgen, T.D. (2001). Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25, 402-408.

Loftus, T.M., Jaworsky, D.E., Frehywot, G.L., Townsend, C.A., Ronnett, G.V., Lane, M.D., and Kuhajda, F.P. (2000). Reduced food intake and body weight in mice treated with fatty acid synthase inhibitors. Science 288, 2379-2381.

Luchetti, F., Canonico, B., Betti, M., Arcangeletti, M., Pilolli, F., Piroddi, M., Canesi, L., Papa, S., and Galli, F. (2010). Melatonin signaling and cell protection function. FASEB J 24, 3603-3624.

Ma, D., Zhang, M., Mori, Y., Yao, C., Larsen, C.P., Yamashima, T., and Zhou, L. (2010). Cellular localization of epidermal-type and brain-type fatty acid-binding proteins in adult hippocampus and their response to cerebral ischemia. Hippocampus 20, 811- 819.

Ma, X., Zubcevic, L., Bruning, J.C., Ashcroft, F.M., and Burdakov, D. (2007). Electrical inhibition of identified anorexigenic POMC neurons by orexin/hypocretin. J Neurosci 27, 1529-1533.

Macintyre, D.J., McGhee, K.A., Maclean, A.W., Afzal, M., Briffa, K., Henry, B., Thomson, P.A., Muir, W.J., and Blackwood, D.H. (2010). Association of GPR50, an X- linked orphan G protein-coupled receptor, and affective disorder in an independent sample of the Scottish population. Neurosci Lett 475, 169-173.

Maekawa, F., Toyoda, Y., Torii, N., Miwa, I., Thompson, R.C., Foster, D.L., Tsukahara, S., Tsukamura, H., and Maeda, K. (2000). Localization of glucokinase-like immunoreactivity in the rat lower brain stem: for possible location of brain glucose- sensing mechanisms. Endocrinology 141, 375-384.

211

Maggio, R., Novi, F., Scarselli, M., and Corsini, G.U. (2005). The impact of G-protein- coupled receptor hetero-oligomerization on function and pharmacology. FEBS J 272, 2939-2946.

Malpaux, B., Migaud, M., Tricoire, H., and Chemineau, P. (2001). Biology of mammalian photoperiodism and the critical role of the pineal gland and melatonin. J Biol Rhythms 16, 336-347.

Markovic, D., and Challiss, R.A. (2009). Alternative splicing of G protein-coupled receptors: physiology and pathophysiology. Cell Mol Life Sci 66, 3337-3352.

Marrache, A.M., Gobeil, F., Zhu, T., and Chemtob, S. (2005). Intracellular signaling of lipid mediators via cognate nuclear G protein-coupled receptors. Endothelium 12, 63- 72.

Massa, M.L., Gagliardino, J.J., and Francini, F. (2011). Liver glucokinase: An overview on the regulatory mechanisms of its activity. IUBMB Life 63, 1-6.

Mastorakos, G., and Zapanti, E. (2004). The hypothalamic-pituitary-adrenal axis in the neuroendocrine regulation of food intake and obesity: the role of corticotropin releasing hormone. Nutr Neurosci 7, 271-280.

Mathew, T.C. (2008). Regional analysis of the ependyma of the third ventricle of rat by light and electron microscopy. Anat Histol Embryol 37, 9-18.

Mayr, B., and Montminy, M. (2001). Transcriptional regulation by the phosphorylation- dependent factor CREB. Nat Rev Mol Cell Biol 2, 599-609.

McArthur, A.J., Hunt, A.E., and Gillette, M.U. (1997). Melatonin action and signal transduction in the rat suprachiasmatic circadian clock: activation of protein kinase C at dusk and dawn. Endocrinology 138, 627-634.

Melvin, R.G., and Andrews, M.T. (2009). Torpor induction in mammals: recent discoveries fueling new ideas. Trends Endocrinol Metab 20, 490-498.

MGI (2011). Deltagen Knockout Mice Phenotypic Data Summary. [online] (updated 31 May 2011) Available at: http://www.informatics.jax.org/external/ko/deltagen/1631.html MGI, ed.

Miaczynska, M., Pelkmans, L., and Zerial, M. (2004). Not just a sink: endosomes in control of signal transduction. Curr Opin Cell Biol 16, 400-406.

212

Mihaly, E., Fekete, C., Legradi, G., and Lechan, R.M. (2001). Hypothalamic dorsomedial nucleus neurons innervate thyrotropin-releasing hormone-synthesizing neurons in the paraventricular nucleus. Brain Res 891, 20-31.

Milligan, G. (2004). G protein-coupled receptor dimerization: function and ligand pharmacology. Mol Pharmacol 66, 1-7.

Minneman, K.P. (2001). Splice variants of G protein-coupled receptors. Mol Interv 1, 108-116.

Miselis, R.R., and Epstein, A.N. (1975). Feeding induced by intracerebroventricular 2- deoxy-D-glucose in the rat. Am J Physiol 229, 1438-1447.

Miyoshi, H., Perfield, J.W., 2nd, Obin, M.S., and Greenberg, A.S. (2008). Adipose triglyceride lipase regulates basal lipolysis and lipid droplet size in adipocytes. J Cell Biochem 105, 1430-1436.

Moller, S., Vilo, J., and Croning, M.D. (2001). Prediction of the coupling specificity of G protein coupled receptors to their G proteins. Bioinformatics 17 Suppl 1, S174-181.

Morgan, P.J., Barrett, P., Howell, H.E., and Helliwell, R. (1994). Melatonin receptors: localization, molecular pharmacology and physiological significance. Neurochem Int 24, 101-146.

Morgan, P.J., and Mercer, J.G. (2001). The regulation of body weight: lessons from the seasonal animal. Proc Nutr Soc 60, 127-134.

Motomura, T., Hashimoto, K., Koga, M., Arita, N., Hayakawa, T., Kishimoto, T., and Kasayama, S. (1998). Inhibition of signal transduction by a splice variant of the growth hormone-releasing hormone receptor expressed in human pituitary adenomas. Metabolism 47, 804-808.

Mounien, L., Marty, N., Tarussio, D., Metref, S., Genoux, D., Preitner, F., Foretz, M., and Thorens, B. (2010). Glut2-dependent glucose-sensing controls thermoregulation by enhancing the leptin sensitivity of NPY and POMC neurons. FASEB J 24, 1747- 1758.

Munro, D., and Thomas, D.W. (2004). The role of polyunsaturated fatty acids in the expression of torpor by mammals: a review. Zoology (Jena) 107, 29-48.

213

Murphy, M., and Ebling, F.J. (2011). The role of hypothalamic tri-iodothyronine availability in seasonal regulation of energy balance and body weight. J Thyroid Res 2011, 387562.

Ngarmukos, C., Baur, E.L., and Kumagai, A.K. (2001). Co-localization of GLUT1 and GLUT4 in the blood-brain barrier of the rat ventromedial hypothalamus. Brain Res 900, 1-8.

Nieuwenhuizen, A.G., and Rutters, F. (2008). The hypothalamic-pituitary-adrenal-axis in the regulation of energy balance. Physiol Behav 94, 169-177.

Nillni, E.A. (2010). Regulation of the hypothalamic thyrotropin releasing hormone (TRH) neuron by neuronal and peripheral inputs. Front Neuroendocrinol 31, 134-156.

Nishiyama, A., Matsui, M., Iwata, S., Hirota, K., Masutani, H., Nakamura, H., Takagi, Y., Sono, H., Gon, Y., and Yodoi, J. (1999). Identification of thioredoxin-binding protein- 2/vitamin D(3) up-regulated protein 1 as a negative regulator of thioredoxin function and expression. J Biol Chem 274, 21645-21650.

Noshiro, M., Kawamoto, T., Furukawa, M., Fujimoto, K., Yoshida, Y., Sasabe, E., Tsutsumi, S., Hamada, T., Honma, S., Honma, K., et al. (2004). Rhythmic expression of DEC1 and DEC2 in peripheral tissues: DEC2 is a potent suppressor for hepatic cytochrome P450s opposing DBP. Genes Cells 9, 317-329.

Novikoff, A.B., Novikoff, P.M., Rosen, O.M., and Rubin, C.S. (1980). Organelle relationships in cultured 3T3-L1 preadipocytes. J Cell Biol 87, 180-196.

Ntambi, J.M., and Young-Cheul, K. (2000). Adipocyte differentiation and gene expression. J Nutr 130, 3122S-3126S.

O'Hara, B.F., Watson, F.L., Srere, H.K., Kumar, H., Wiler, S.W., Welch, S.K., Bitting, L., Heller, H.C., and Kilduff, T.S. (1999). Gene expression in the brain across the hibernation cycle. J Neurosci 19, 3781-3790.

O'Malley, K.L., Jong, Y.J., Gonchar, Y., Burkhalter, A., and Romano, C. (2003). Activation of metabotropic mGlu5 on nuclear membranes mediates intranuclear Ca2+ changes in heterologous cell types and neurons. J Biol Chem 278, 28210-28219.

214

O'Shea, P.J., and Williams, G.R. (2002). Insight into the physiological actions of thyroid hormone receptors from genetically modified mice. J Endocrinol 175, 553-570.

Obici, S., Feng, Z., Arduini, A., Conti, R., and Rossetti, L. (2003). Inhibition of hypothalamic carnitine palmitoyltransferase-1 decreases food intake and glucose production. Nat Med 9, 756-761.

Obici, S., Zhang, B.B., Karkanias, G., and Rossetti, L. (2002). Hypothalamic insulin signaling is required for inhibition of glucose production. Nat Med 8, 1376-1382.

Oka, S., Liu, W., Masutani, H., Hirata, H., Shinkai, Y., Yamada, S., Yoshida, T., Nakamura, H., and Yodoi, J. (2006). Impaired fatty acid utilization in thioredoxin binding protein-2 (TBP-2)-deficient mice: a unique animal model of Reye syndrome. FASEB J 20, 121-123.

Olichon-Berthe, C., Van Obberghen, E., and Le Marchand-Brustel, Y. (1992). Effect of cold acclimation on the expression of glucose transporter Glut 4. Mol Cell Endocrinol 89, 11-18.

Ollmann, M.M., Wilson, B.D., Yang, Y.K., Kerns, J.A., Chen, Y., Gantz, I., and Barsh, G.S. (1997). Antagonism of central melanocortin receptors in vitro and in vivo by agouti-related protein. Science 278, 135-138.

Owada, Y., Yoshimoto, T., and Kondo, H. (1996). Spatio-temporally differential expression of genes for three members of fatty acid binding proteins in developing and mature rat brains. J Chem Neuroanat 12, 113-122.

Palchykova, S., Deboer, T., and Tobler, I. (2002). Selective sleep deprivation after daily torpor in the Djungarian hamster. J Sleep Res 11, 313-319.

Pandi-Perumal, S.R., Trakht, I., Srinivasan, V., Spence, D.W., Maestroni, G.J., Zisapel, N., and Cardinali, D.P. (2008). Physiological effects of melatonin: role of melatonin receptors and signal transduction pathways. Prog Neurobiol 85, 335-353.

Paradis, E., Clavel, S., Julien, P., Murthy, M.R., de Bilbao, F., Arsenijevic, D., Giannakopoulos, P., Vallet, P., and Richard, D. (2004). Lipoprotein lipase and endothelial lipase expression in mouse brain: regional distribution and selective induction following kainic acid-induced lesion and focal cerebral ischemia. Neurobiol Dis 15, 312-325.

215

Parikh, H., Carlsson, E., Chutkow, W.A., Johansson, L.E., Storgaard, H., Poulsen, P., Saxena, R., Ladd, C., Schulze, P.C., Mazzini, M.J., et al. (2007). TXNIP regulates peripheral glucose metabolism in humans. PLoS Med 4, e158.

Park, P.S., and Palczewski, K. (2005). Diversifying the repertoire of G protein-coupled receptors through oligomerization. Proc Natl Acad Sci U S A 102, 8793-8794.

Park, S., Kim da, S., and Daily, J.W. (2011). Central infusion of ketone bodies modulates body weight and hepatic insulin sensitivity by modifying hypothalamic leptin and insulin signaling pathways in type 2 diabetic rats. Brain Res 1401, 95-103.

Patwari, P., Chutkow, W.A., Cummings, K., Verstraeten, V.L., Lammerding, J., Schreiter, E.R., and Lee, R.T. (2009). Thioredoxin-independent regulation of metabolism by the alpha-arrestin proteins. J Biol Chem 284, 24996-25003.

Paul, M.J., Freeman, D.A., Park, J.H., and Dark, J. (2005). Neuropeptide Y induces torpor-like hypothermia in Siberian hamsters. Brain Res 1055, 83-92.

Paul, M.J., Kauffman, A.S., and Zucker, I. (2004). Feeding schedule controls circadian timing of daily torpor in SCN-ablated Siberian hamsters. J Biol Rhythms 19, 226-237.

Pelz, K.M., and Dark, J. (2007). ICV NPY Y1 receptor agonist but not Y5 agonist induces torpor-like hypothermia in cold-acclimated Siberian hamsters. Am J Physiol Regul Integr Comp Physiol 292, R2299-2311.

Pelz, K.M., Routman, D., Driscoll, J.R., Kriegsfeld, L.J., and Dark, J. (2008). Monosodium glutamate-induced arcuate nucleus damage affects both natural torpor and 2DG-induced torpor-like hypothermia in Siberian hamsters. Am J Physiol Regul Integr Comp Physiol 294, R255-265.

Peschke, E., and Muhlbauer, E. (2010). New evidence for a role of melatonin in glucose regulation. Best Pract Res Clin Endocrinol Metab 24, 829-841.

Petit, L., Lacroix, I., de Coppet, P., Strosberg, A.D., and Jockers, R. (1999). Differential signaling of human Mel1a and Mel1b melatonin receptors through the cyclic guanosine 3'-5'-monophosphate pathway. Biochem Pharmacol 58, 633-639.

Pevet, P., Agez, L., Bothorel, B., Saboureau, M., Gauer, F., Laurent, V., and Masson- Pevet, M. (2006). Melatonin in the multi-oscillatory mammalian circadian world. Chronobiol Int 23, 39-51.

216

Pevet, P., Bothorel, B., Slotten, H., and Saboureau, M. (2002). The chronobiotic properties of melatonin. Cell Tissue Res 309, 183-191.

Pfeiffer, B., Elmer, K., Roggendorf, W., Reinhart, P.H., and Hamprecht, B. (1990). Immunohistochemical demonstration of glycogen phosphorylase in rat brain slices. Histochemistry 94, 73-80.

Phillips, L.J., McGorry, P.D., Garner, B., Thompson, K.N., Pantelis, C., Wood, S.J., and Berger, G. (2006). Stress, the hippocampus and the hypothalamic-pituitary-adrenal axis: implications for the development of psychotic disorders. Aust N Z J Psychiatry 40, 725-741.

Pitcher, J.A., Freedman, N.J., and Lefkowitz, R.J. (1998). G protein-coupled receptor kinases. Annu Rev Biochem 67, 653-692.

Poulos, S.P., Hausman, D.B., and Hausman, G.J. (2010). The development and endocrine functions of adipose tissue. Mol Cell Endocrinol 323, 20-34.

Preitner, N., Damiola, F., Lopez-Molina, L., Zakany, J., Duboule, D., Albrecht, U., and Schibler, U. (2002). The orphan nuclear receptor REV-ERBalpha controls circadian transcription within the positive limb of the mammalian circadian oscillator. Cell 110, 251-260.

Prevot, V., Bellefontaine, N., Baroncini, M., Sharif, A., Hanchate, N.K., Parkash, J., Campagne, C., and de Seranno, S. (2010). Gonadotrophin-releasing hormone nerve terminals, tanycytes and neurohaemal junction remodelling in the adult median eminence: functional consequences for reproduction and dynamic role of vascular endothelial cells. J Neuroendocrinol 22, 639-649.

Qi, W., Chen, X., Holian, J., Tan, C.Y., Kelly, D.J., and Pollock, C.A. (2009). Transcription factors Kruppel-like factor 6 and peroxisome proliferator-activated receptor-{gamma} mediate high glucose-induced thioredoxin-interacting protein. Am J Pathol 175, 1858-1867.

Qi, Y., Takahashi, N., Hileman, S.M., Patel, H.R., Berg, A.H., Pajvani, U.B., Scherer, P.E., and Ahima, R.S. (2004). Adiponectin acts in the brain to decrease body weight. Nat Med 10, 524-529.

Rakhshandehroo, M., Sanderson, L.M., Matilainen, M., Stienstra, R., Carlberg, C., de Groot, P.J., Muller, M., and Kersten, S. (2007). Comprehensive analysis of

217

PPARalpha-dependent regulation of hepatic lipid metabolism by expression profiling. PPAR Res 2007, 26839.

Ralser, M., Wamelink, M.M., Struys, E.A., Joppich, C., Krobitsch, S., Jakobs, C., and Lehrach, H. (2008). A catabolic block does not sufficiently explain how 2-deoxy-D- glucose inhibits cell growth. Proc Natl Acad Sci U S A 105, 17807-17811.

Reiter, R.J. (1991). Pineal melatonin: cell biology of its synthesis and of its physiological interactions. Endocr Rev 12, 151-180.

Reppert, S.M., and Weaver, D.R. (2002). Coordination of circadian timing in mammals. Nature 418, 935-941.

Reppert, S.M., Weaver, D.R., Cassone, V.M., Godson, C., and Kolakowski, L.F., Jr. (1995). Melatonin receptors are for the birds: molecular analysis of two receptor subtypes differentially expressed in chick brain. Neuron 15, 1003-1015.

Reppert, S.M., Weaver, D.R., Ebisawa, T., Mahle, C.D., and Kolakowski, L.F., Jr. (1996). Cloning of a melatonin-related receptor from human pituitary. FEBS Lett 386, 219-224.

Revankar, C.M., Cimino, D.F., Sklar, L.A., Arterburn, J.B., and Prossnitz, E.R. (2005). A transmembrane intracellular estrogen receptor mediates rapid cell signaling. Science 307, 1625-1630.

Revankar, C.M., Mitchell, H.D., Field, A.S., Burai, R., Corona, C., Ramesh, C., Sklar, L.A., Arterburn, J.B., and Prossnitz, E.R. (2007). Synthetic estrogen derivatives demonstrate the functionality of intracellular GPR30. ACS Chem Biol 2, 536-544.

Rodriguez, E.M., Blazquez, J.L., Pastor, F.E., Pelaez, B., Pena, P., Peruzzo, B., and Amat, P. (2005). Hypothalamic tanycytes: a key component of brain-endocrine interaction. Int Rev Cytol 247, 89-164.

Rogers, R.C., Barnes, M.J., and Hermann, G.E. (2009). Leptin "gates" thermogenic action of thyrotropin-releasing hormone in the hindbrain. Brain Res 1295, 135-141.

Rohleder, N., and Kirschbaum, C. (2007). Effects of nutrition on neuro-endocrine stress responses. Curr Opin Clin Nutr Metab Care 10, 504-510.

218

Roka, F., Brydon, L., Waldhoer, M., Strosberg, A.D., Freissmuth, M., Jockers, R., and Nanoff, C. (1999). Tight association of the human Mel(1a)-melatonin receptor and G(i): precoupling and constitutive activity. Mol Pharmacol 56, 1014-1024.

Rosen, E.D., and MacDougald, O.A. (2006). Adipocyte differentiation from the inside out. Nat Rev Mol Cell Biol 7, 885-896.

Rosen, E.D., and Spiegelman, B.M. (2001). PPARgamma : a nuclear regulator of metabolism, differentiation, and cell growth. J Biol Chem 276, 37731-37734.

Rosen, E.D., Walkey, C.J., Puigserver, P., and Spiegelman, B.M. (2000). Transcriptional regulation of adipogenesis. Genes Dev 14, 1293-1307.

Rousseau, K., Atcha, Z., and Loudon, A.S. (2003). Leptin and seasonal mammals. J Neuroendocrinol 15, 409-414.

Rozenfeld, R., and Devi, L.A. (2008). Regulation of CB1 trafficking by the adaptor protein AP-3. FASEB J 22, 2311-2322.

Ruby, N.F. (1995). Paraventricular nucleus ablation disrupts daily torpor in Siberian hamsters. Brain Res Bull 37, 193-198.

Ruby, N.F. (2003). Hibernation: when good clocks go cold. J Biol Rhythms 18, 275- 286.

Ruby, N.F., Ibuka, N., Barnes, B.M., and Zucker, I. (1989). Suprachiasmatic nuclei influence torpor and circadian temperature rhythms in hamsters. Am J Physiol 257, R210-215.

Ruby, N.F., and Zucker, I. (1992). Daily torpor in the absence of the suprachiasmatic nucleus in Siberian hamsters. Am J Physiol 263, R353-362.

Sadowski, H.B., Wheeler, T.T., and Young, D.A. (1992). Gene expression during 3T3- L1 adipocyte differentiation. Characterization of initial responses to the inducing agents and changes during commitment to differentiation. J Biol Chem 267, 4722-4731.

Salti, R., Galluzzi, F., Bindi, G., Perfetto, F., Tarquini, R., Halberg, F., and Cornelissen, G. (2000). Nocturnal melatonin patterns in children. J Clin Endocrinol Metab 85, 2137- 2144.

219

Sanders, N.M., Dunn-Meynell, A.A., and Levin, B.E. (2004). Third ventricular alloxan reversibly impairs glucose counterregulatory responses. Diabetes 53, 1230-1236.

Saper, C.B., Scammell, T.E., and Lu, J. (2005). Hypothalamic regulation of sleep and circadian rhythms. Nature 437, 1257-1263.

Sauzet, S., Besseau, L., Herrera Perez, P., Coves, D., Chatain, B., Peyric, E., Boeuf, G., Munoz-Cueto, J.A., and Falcon, J. (2008). Cloning and retinal expression of melatonin receptors in the European sea bass, Dicentrarchus labrax. Gen Comp Endocrinol 157, 186-195.

Scanlan, T.S., Suchland, K.L., Hart, M.E., Chiellini, G., Huang, Y., Kruzich, P.J., Frascarelli, S., Crossley, D.A., Bunzow, J.R., Ronca-Testoni, S., et al. (2004). 3- Iodothyronamine is an endogenous and rapid-acting derivative of thyroid hormone. Nat Med 10, 638-642.

Scheer, A., Fanelli, F., Costa, T., De Benedetti, P.G., and Cotecchia, S. (1996). Constitutively active mutants of the alpha 1B-adrenergic receptor: role of highly conserved polar amino acids in receptor activation. EMBO J 15, 3566-3578.

Scheer, F.A., Van Montfrans, G.A., van Someren, E.J., Mairuhu, G., and Buijs, R.M. (2004). Daily nighttime melatonin reduces blood pressure in male patients with essential hypertension. Hypertension 43, 192-197.

Schomerus, C., and Korf, H.W. (2005). Mechanisms regulating melatonin synthesis in the mammalian pineal organ. Ann N Y Acad Sci 1057, 372-383.

Schubert, K.A., Boerema, A.S., Vaanholt, L.M., de Boer, S.F., Strijkstra, A.M., and Daan, S. (2010). Daily torpor in mice: high foraging costs trigger energy-saving hypothermia. Biol Lett 6, 132-135.

Schuhler, S., Warner, A., Finney, N., Bennett, G.W., Ebling, F.J., and Brameld, J.M. (2007). Thyrotrophin-releasing hormone decreases feeding and increases body temperature, activity and oxygen consumption in Siberian hamsters. J Neuroendocrinol 19, 239-249.

Schwartz, M.W., Woods, S.C., Porte, D., Jr., Seeley, R.J., and Baskin, D.G. (2000). Central nervous system control of food intake. Nature 404, 661-671.

220

Schweiger, M., Schreiber, R., Haemmerle, G., Lass, A., Fledelius, C., Jacobsen, P., Tornqvist, H., Zechner, R., and Zimmermann, R. (2006). Adipose triglyceride lipase and hormone-sensitive lipase are the major enzymes in adipose tissue triacylglycerol catabolism. J Biol Chem 281, 40236-40241.

Seifert, R., and Wenzel-Seifert, K. (2002). Constitutive activity of G-protein-coupled receptors: cause of disease and common property of wild-type receptors. Naunyn Schmiedebergs Arch Pharmacol 366, 381-416.

Sekiya, M., Osuga, J., Okazaki, H., Yahagi, N., Harada, K., Shen, W.J., Tamura, Y., Tomita, S., Iizuka, Y., Ohashi, K., et al. (2004). Absence of hormone-sensitive lipase inhibits obesity and adipogenesis in Lep ob/ob mice. J Biol Chem 279, 15084-15090.

Shaked, M., Ketzinel-Gilad, M., Ariav, Y., Cerasi, E., Kaiser, N., and Leibowitz, G. (2009). Insulin counteracts glucotoxic effects by suppressing thioredoxin-interacting protein production in INS-1E beta cells and in Psammomys obesus pancreatic islets. Diabetologia 52, 636-644.

Shearman, L.P., Zylka, M.J., Weaver, D.R., Kolakowski, L.F., Jr., and Reppert, S.M. (1997). Two period homologs: circadian expression and photic regulation in the suprachiasmatic nuclei. Neuron 19, 1261-1269.

Sheth, S.S., Castellani, L.W., Chari, S., Wagg, C., Thipphavong, C.K., Bodnar, J.S., Tontonoz, P., Attie, A.D., Lopaschuk, G.D., and Lusis, A.J. (2005). Thioredoxin- interacting protein deficiency disrupts the fasting-feeding metabolic transition. J Lipid Res 46, 123-134.

Shimizu-Albergine, M., Ippolito, D.L., and Beavo, J.A. (2001). Downregulation of fasting-induced cAMP response element-mediated gene induction by leptin in neuropeptide Y neurons of the arcuate nucleus. J Neurosci 21, 1238-1246.

Shimizu, Y., Nikami, H., and Saito, M. (1991). Sympathetic activation of glucose utilization in brown adipose tissue in rats. J Biochem 110, 688-692.

Shimokawa, T., Kumar, M.V., and Lane, M.D. (2002). Effect of a fatty acid synthase inhibitor on food intake and expression of hypothalamic neuropeptides. Proc Natl Acad Sci U S A 99, 66-71.

Sidibe, A., Mullier, A., Chen, P., Baroncini, M., Boutin, J.A., Delagrange, P., Prevot, V., and Jockers, R. (2010). Expression of the orphan GPR50 protein in rodent and human

221 dorsomedial hypothalamus, tanycytes and median eminence. J Pineal Res 48, 263- 269.

Simonneaux, V., and Ribelayga, C. (2003). Generation of the melatonin endocrine message in mammals: a review of the complex regulation of melatonin synthesis by norepinephrine, peptides, and other pineal transmitters. Pharmacol Rev 55, 325-395.

Sivitz, W.I., Fink, B.D., and Donohoue, P.A. (1999). Fasting and leptin modulate adipose and muscle uncoupling protein: divergent effects between messenger ribonucleic acid and protein expression. Endocrinology 140, 1511-1519.

Slominski, A., Pisarchik, A., Zbytek, B., Tobin, D.J., Kauser, S., and Wortsman, J. (2003). Functional activity of serotoninergic and melatoninergic systems expressed in the skin. J Cell Physiol 196, 144-153.

Smit, M.J., Vischer, H.F., Bakker, R.A., Jongejan, A., Timmerman, H., Pardo, L., and Leurs, R. (2007). Pharmacogenomic and structural analysis of constitutive g protein- coupled receptor activity. Annu Rev Pharmacol Toxicol 47, 53-87.

Stanley, B.G., Lanthier, D., Chin, A.S., and Leibowitz, S.F. (1989). Suppression of neuropeptide Y-elicited eating by adrenalectomy or hypophysectomy: reversal with corticosterone. Brain Res 501, 32-36.

Staples, J.F., and Brown, J.C. (2008). Mitochondrial metabolism in hibernation and daily torpor: a review. J Comp Physiol B 178, 811-827.

Stoltzman, C.A., Peterson, C.W., Breen, K.T., Muoio, D.M., Billin, A.N., and Ayer, D.E. (2008). Glucose sensing by MondoA:Mlx complexes: a role for hexokinases and direct regulation of thioredoxin-interacting protein expression. Proc Natl Acad Sci U S A 105, 6912-6917.

Storch, J., and Thumser, A.E. (2000). The fatty acid transport function of fatty acid- binding proteins. Biochim Biophys Acta 1486, 28-44.

Sun, L., Xie, H., Mori, M.A., Alexander, R., Yuan, B., Hattangadi, S.M., Liu, Q., Kahn, C.R., and Lodish, H.F. (2011). Mir193b-365 is essential for brown fat differentiation. Nat Cell Biol 13, 958-965.

Swindell, W.R. (2009). Genes and gene expression modules associated with caloric restriction and aging in the laboratory mouse. BMC Genomics 10, 585.

222

Swoap, S.J., and Weinshenker, D. (2008). Norepinephrine controls both torpor initiation and emergence via distinct mechanisms in the mouse. PLoS One 3, e4038.

Szele, F.G., and Szuchet, S. (2003). Cells lining the ventricular system: evolving concepts underlying developmental eventsin the embryo and adult. In Advances in Molecular and Cell Biology, H. Leif, ed. (Elsevier), pp. 127-146.

Tanioka, M., Yamada, H., Doi, M., Bando, H., Yamaguchi, Y., Nishigori, C., and Okamura, H. (2009). Molecular clocks in mouse skin. J Invest Dermatol 129, 1225- 1231.

Tasker, J.G., and Dudek, F.E. (1993). Local inhibitory synaptic inputs to neurones of the paraventricular nucleus in slices of rat hypothalamus. J Physiol 469, 179-192.

Thomas, E.M., Jewett, M.E., and Zucker, I. (1993). Torpor shortens the period of Siberian hamster circadian rhythms. Am J Physiol 265, R951-956.

Thomson, P.A., Wray, N.R., Thomson, A.M., Dunbar, D.R., Grassie, M.A., Condie, A., Walker, M.T., Smith, D.J., Pulford, D.J., Muir, W., et al. (2005). Sex-specific association between bipolar affective disorder in women and GPR50, an X-linked orphan G protein-coupled receptor. Mol Psychiatry 10, 470-478.

Tobin, A.B., Butcher, A.J., and Kong, K.C. (2008). Location, location, location...site- specific GPCR phosphorylation offers a mechanism for cell-type-specific signalling. Trends Pharmacol Sci 29, 413-420.

Tontonoz, P., Kim, J.B., Graves, R.A., and Spiegelman, B.M. (1993). ADD1: a novel helix-loop-helix transcription factor associated with adipocyte determination and differentiation. Mol Cell Biol 13, 4753-4759.

Tordjman, J., Chauvet, G., Quette, J., Beale, E.G., Forest, C., and Antoine, B. (2003). Thiazolidinediones block fatty acid release by inducing glyceroneogenesis in fat cells. J Biol Chem 278, 18785-18790.

Tordoff, M.G., Flynn, F.W., Grill, H.J., and Friedman, M.I. (1988). Contribution of fat metabolism to 'glucoprivic' feeding produced by fourth ventricular 5-thio-D-glucose. Brain Res 445, 216-221.

Tu, H.M., Kim, S.W., Salvatore, D., Bartha, T., Legradi, G., Larsen, P.R., and Lechan, R.M. (1997). Regional distribution of type 2 thyroxine deiodinase messenger

223 ribonucleic acid in rat hypothalamus and pituitary and its regulation by thyroid hormone. Endocrinology 138, 3359-3368.

Ulloa-Aguirre, A., Janovick, J.A., Miranda, A.L., and Conn, P.M. (2006). G-protein- coupled receptor trafficking: understanding the chemical basis of health and disease. ACS Chem Biol 1, 631-638. van Beekum, O., Brenkman, A.B., Grontved, L., Hamers, N., van den Broek, N.J., Berger, R., Mandrup, S., and Kalkhoven, E. (2008). The adipogenic acetyltransferase Tip60 targets activation function 1 of peroxisome proliferator-activated receptor gamma. Endocrinology 149, 1840-1849. van Breukelen, F., and Martin, S.L. (2002). Reversible depression of transcription during hibernation. J Comp Physiol B 172, 355-361. van den Top, M., Lee, K., Whyment, A.D., Blanks, A.M., and Spanswick, D. (2004). Orexigen-sensitive NPY/AgRP pacemaker neurons in the hypothalamic arcuate nucleus. Nat Neurosci 7, 493-494.

Vanecek, J. (1998). Cellular mechanisms of melatonin action. Physiol Rev 78, 687-721. von Gall, C., Stehle, J.H., and Weaver, D.R. (2002). Mammalian melatonin receptors: molecular biology and signal transduction. Cell Tissue Res 309, 151-162.

Voss-Andreae, A., Murphy, J.G., Ellacott, K.L., Stuart, R.C., Nillni, E.A., Cone, R.D., and Fan, W. (2007). Role of the central melanocortin circuitry in adaptive thermogenesis of brown adipose tissue. Endocrinology 148, 1550-1560.

Wang, H., Hu, L., Dalen, K., Dorward, H., Marcinkiewicz, A., Russell, D., Gong, D., Londos, C., Yamaguchi, T., Holm, C., et al. (2009). Activation of hormone-sensitive lipase requires two steps, protein phosphorylation and binding to the PAT-1 domain of lipid droplet coat proteins. J Biol Chem 284, 32116-32125.

Wang, Z., Rong, Y.P., Malone, M.H., Davis, M.C., Zhong, F., and Distelhorst, C.W. (2006). Thioredoxin-interacting protein (txnip) is a glucocorticoid-regulated primary response gene involved in mediating glucocorticoid-induced apoptosis. Oncogene 25, 1903-1913.

Watanabe, M., Horai, Y., Houten, S.M., Morimoto, K., Sugizaki, T., Arita, E., Mataki, C., Sato, H., Tanigawara, Y., Schoonjans, K., et al. (2011). Lowering Bile Acid Pool Size

224 with a Synthetic Farnesoid X Receptor (FXR) Agonist Induces Obesity and Diabetes through Reduced Energy Expenditure. J Biol Chem 286, 26913-26920.

Watts, A.G. (2005). Glucocorticoid regulation of peptide genes in neuroendocrine CRH neurons: a complexity beyond negative feedback. Front Neuroendocrinol 26, 109-130.

Webb, G.P., Jagot, S.A., and Jakobson, M.E. (1982). Fasting-induced torpor in Mus musculus and its implications in the use of murine models for human obesity studies. Comp Biochem Physiol A Comp Physiol 72, 211-219.

Williams, D.R., Epperson, L.E., Li, W., Hughes, M.A., Taylor, R., Rogers, J., Martin, S.L., Cossins, A.R., and Gracey, A.Y. (2005). Seasonally hibernating phenotype assessed through transcript screening. Physiol Genomics 24, 13-22.

Worthington, W.C., Jr., and Cathcart, R.S., 3rd (1963). Ependymal cilia: distribution and activity in the adult human brain. Science 139, 221-222.

Wu, Y.H., Zhou, J.N., Balesar, R., Unmehopa, U., Bao, A., Jockers, R., Van Heerikhuize, J., and Swaab, D.F. (2006). Distribution of MT1 melatonin receptor immunoreactivity in the human hypothalamus and pituitary gland: colocalization of MT1 with vasopressin, oxytocin, and corticotropin-releasing hormone. J Comp Neurol 499, 897-910.

Wueest, S., Rapold, R.A., Rytka, J.M., Schoenle, E.J., and Konrad, D. (2009). Basal lipolysis, not the degree of insulin resistance, differentiates large from small isolated adipocytes in high-fat fed mice. Diabetologia 52, 541-546.

Yamaguchi, T. (2010). Crucial role of CGI-58/alpha/beta hydrolase domain-containing protein 5 in lipid metabolism. Biol Pharm Bull 33, 342-345.

Yamauchi, T., Kamon, J., Ito, Y., Tsuchida, A., Yokomizo, T., Kita, S., Sugiyama, T., Miyagishi, M., Hara, K., Tsunoda, M., et al. (2003). Cloning of adiponectin receptors that mediate antidiabetic metabolic effects. Nature 423, 762-769.

Yasuo, S., Nakao, N., Ohkura, S., Iigo, M., Hagiwara, S., Goto, A., Ando, H., Yamamura, T., Watanabe, M., Watanabe, T., et al. (2006). Long-day suppressed expression of type 2 deiodinase gene in the mediobasal hypothalamus of the Saanen goat, a short-day breeder: implication for seasonal window of thyroid hormone action on reproductive neuroendocrine axis. Endocrinology 147, 432-440.

225

Yokoyama, C., Wang, X., Briggs, M.R., Admon, A., Wu, J., Hua, X., Goldstein, J.L., and Brown, M.S. (1993). SREBP-1, a basic-helix-loop-helix-leucine zipper protein that controls transcription of the low density lipoprotein receptor gene. Cell 75, 187-197.

Young, P., Cawthorne, M.A., Levy, A.L., and Wilson, K. (1984). Reduced maximum capacity of glycolysis in brown adipose tissue of genetically obese, diabetic (db/db) mice and its restoration following treatment with a thermogenic beta-adrenoceptor agonist. FEBS Lett 176, 16-20.

Yu, F.X., Chai, T.F., He, H., Hagen, T., and Luo, Y. (2010). Thioredoxin-interacting protein (Txnip) gene expression: sensing oxidative phosphorylation status and glycolytic rate. J Biol Chem 285, 25822-25830.

Yu, X.X., Lewin, D.A., Forrest, W., and Adams, S.H. (2002). Cold elicits the simultaneous induction of fatty acid synthesis and beta-oxidation in murine brown adipose tissue: prediction from differential gene expression and confirmation in vivo. FASEB J 16, 155-168.

Zhu, X., and Wess, J. (1998). Truncated V2 vasopressin receptors as negative regulators of wild-type V2 receptor function. Biochemistry 37, 15773-15784.

Zisapel, N. (2001). Melatonin-dopamine interactions: from basic neurochemistry to a clinical setting. Cell Mol Neurobiol 21, 605-616.

Zoeller, R.T., Tan, S.W., and Tyl, R.W. (2007). General background on the hypothalamic-pituitary-thyroid (HPT) axis. Crit Rev Toxicol 37, 11-53.

226

APPENDIX A

Publications

Personal contribution sees me co-authored on two publications: (Li et al., 2011) and the following paper, entitled A Role for the Melatonin-Related Receptor GPR50 in Leptin Signaling, Adaptive Thermogenesis, and Torpor.

227

Current Biology 22, 70–77, January 10, 2012 ª2012 Elsevier Ltd All rights reserved DOI 10.1016/j.cub.2011.11.043 Report A Role for the Melatonin-Related Receptor GPR50 in Leptin Signaling, Adaptive Thermogenesis, and Torpor

1, 2,3,4 1 David A. Bechtold, * Anissa Sidibe, Ben R.C. Saer, Long-term recording of body temperature (Tb) revealed 1 1 1 2/2 Jian Li, Laura E. Hand, Elena A. Ivanova, Gpr50 mice to have a lower nighttime Tb compared with Veerle M. Darras,5 Julie Dam,2,3,4 Ralf Jockers,2,3,4 WT mice, despite showing 25% higher locomotor activity Simon M. Luckman,1 and Andrew S.I. Loudon1,* (Figures 1E and 1F). Remarkably, upon fasting, Gpr502/2 1Faculty of Life Sciences, University of Manchester, mice consistently entered a state of torpor, characterized by

Manchester M13 9PT, UK severe drops in metabolic rate (VO2)(Figures 1G and 1I) and 2 Inserm, U1016, Institut Cochin, Paris, France Tb (25.8C60.7; Figures 1H and 1I). The depth of hypothermia 3 CNRS UMR 8104, Paris, France was not dependent on ambient temperatures (Ta)of23Cor 4 Universite´ Paris Descartes, Sorbonne Paris Cite, 16C(Figure 1; Figure S2). The temporal profile of torpor indi- 75006 Paris, France cated that the process was gated by the circadian clock; 5Laboratory of Comparative Endocrinology, Katholieke this was confirmed by housing mice under constant light Universiteit Leuven, 3000 Leuven, Belgium (removing exogenous timing cues) and fasting for 48 hr (Fig- ure 1J). Two distinct torpor bouts with defined circadian profiles were observed in the Gpr502/2 mice. Torpor was not Summary observed in WT mice during 48 hr of food deprivation, demon- strating that torpor is not driven simply by depletion of energy The ability of mammals to maintain a constant body temper- stores. Further, the expression of torpor in knockout (KO) mice ature has proven to be a profound evolutionary advantage, was not due to inherent differences in body weight between allowing members of this class to thrive in most environ- the genotypes (Figure S2C). Torpor has been reported previ- ments on earth. Intriguingly, some mammals employ ously in laboratory mice. However, profound hypothermia bouts of deep hypothermia (torpor) to cope with reduced (<30C) has typically been observed only following prolonged 2/2 food supply and harsh climates [1, 2]. During torpor, physio- food restriction and reduced Ta [7, 13–16]. Gpr50 mice logical processes such as respiration, cardiac function, enter a state of torpor (Tb w25C) within as little as 6 hr of fast- and metabolic rate are severely depressed, yet the neural ing when housed at constant Ta of w23C, which is character- mechanisms that regulate torpor remain unclear [3]. ized by a stable level of hypothermia. Hypothalamic responses to energy signals, such as leptin, In line with reduced Tb, expression of uncoupling protein 1 influence the expression of torpor [4–7]. We show that the (ucp1) in brown adipose tissue (BAT) was lower in Gpr502/2 orphan receptor GPR50 plays an important role in adaptive mice than WT animals (Figure 1K). Nonetheless, KO mice 2/2 thermogenesis and torpor. Unlike wild-type mice, Gpr50 exhibited a normal elevation in Tb and BAT expression of mice readily enter torpor in response to fasting and 2-deox- ucp1 and deiodinase 2 (dio2) in response to the b3-adrenergic yglucose administration. Decreased thermogenesis in receptor agonist CL-316243 (Figures 1L and 1M; Figure S2D). 2/2 2/2 Gpr50 mice is not due to a deficit in brown adipose The robust response of the Gpr50 mice to b3-adrenergic tissue, the principal site of nonshivering thermogenesis in receptor stimulation demonstrates that the low nocturnal Tb mice [8]. GPR50 is highly expressed in the hypothalamus and expression of torpor in these mice are not due to dimin- of several species, including man [9, 10]. In line with this, ished thermogenic capacity or responsiveness of BAT, but altered thermoregulation in Gpr502/2 mice is associated rather likely reflect altered sympathetic drive to this tissue. It with attenuated responses to leptin and a suppression is also possible that heat loss may be accentuated in the of thyrotropin-releasing hormone. Thus, our findings iden- Gpr502/2 mice, for example through alterations in sympathetic tify hypothalamic circuits involved in torpor and reveal control of blood flow. GPR50 to be a novel component of adaptive thermogenesis Fasted Gpr502/2 mice exhibited significantly lower circu- in mammals. lating glucose in comparison with WT mice (Gpr502/2, 3.9 mmol/l 6 0.4; WT, 5.4 6 0.3; p < 0.05) suggesting that hypo- Results and Discussion glycemia could contribute to the torpor response. Consistent with this, Gpr502/2 mice were found to be hypersensitive to Severe Hypometabolism in Gpr502/2 Mice upon Fasting deoxyglucose (2-DG), a glucose mimetic that induces a state We demonstrated previously that in comparison to wild-type of perceived hypoglycemia. KO mice exhibited torpor-like 2/2 (WT) littermates, Gpr50 mice show resistance to diet- drops in VO2 and Tb following doses of 2-DG (250–500 mg/kg) induced obesity, yet lose less weight when fasted [9]. Intense that had little effect on WT mice (Figure S3). However, both GPR50 immunoreactivity (IR) is observed in the neurons of genotypes exhibited comparable rises in blood glucose in the dorsomedial nucleus (DMN) of the hypothalamus and tany- response to 2-DG (Figure S3), indicating that the exaggerated cytes that line the third ventricle (Figures 1A–1D; see also Fig- metabolic response of the Gpr502/2 mice is not due to dimin- ure S1 available online), both implicated in nutrient sensing ished glucose storage or ineffective counter-regulatory and energy balance [11, 12]. This prompted us to further glucose release. Similarly, no genotype differences in glucose characterize the metabolic phenotype of the Gpr502/2 mice. clearance were observed during a glucose tolerance test (p > 0.05, two-way analysis of variance; Figure S3E). Tanycytes express much of the glucose-sensing machinery employed by *Correspondence: [email protected] (D.A.B.), andrew. pancreatic b-cells [17], and selective destruction of these cells [email protected] (A.S.I.L.) attenuates 2-DG-induced feeding in rats [18]. This suggests GPR50 Modulates Leptin Signaling 71

Figure 1. Altered Thermogenesis in Gpr502/2 Mice (A–D) GPR50 immunoreactivity in the brain is limited to the ventral portion of the third ventricle ependymal layer (A) and neurons of the DMN. (B–D) The identity of GPR50 expressing cells as tanycytes was confirmed by colocalization of the receptor (red) with the tanycyte markers GLUT1 (C; green) and vimentin (D; green). Lowercase letters in (A) represent the approximate positions of (B), (C), and (D). Nonmerged images are available in Figure S1. GPR50 expression was undetectable in Gpr50 knockout (KO) mice used in the current studies (Figure S1). Scale bar represents 200 mm in (A) and 15 mm in (B–D). 2/2 (E and F) Mean body temperature (Tb) and locomotor activity in wild-type (WT) and Gpr50 mice collected over 30 days of monitoring (n = 8/group). Night- 2/2 time Tb was reduced in Gpr50 mice (E), despite a 25% increase in locomotor activity (F). 2/2 2/2 (G–J) Representative recordings of O2 consumption (VO2; G) and Tb (H) records in Gpr50 and WT mice subjected to a 24 hr fast in which Gpr50 mice entered a state of torpor (n = 8–10/group, group data shown in I). Fasting period is represented by the gray bar above trace, and ambient room temperatureis represented by the hashed line in (H). The following abbreviation is used: ZT, zeitgeber time (ZT0 indicates lights on). (J) To remove exogenous timing cues, we housed mice in constant light and fasted for 48 hr (n = 4/group). Gpr502/2 mice exhibited two distinct and precisely timed torpor bouts. (K) Lower ucp1 expression was observed in brown adipose tissue (BAT) from Gpr502/2 mice compared with WT mice in either a fed or fasted state (n = 6–12/group).

(L and M) Administration of the b3-adrenergic receptor agonist CL-316243 (1 mg/kg intraperitoneal [ip]) elicited comparable increases in ucp1 (L) and dio2 (M) expression in BAT in WT and Gpr502/2 mice. Thermogenic responses to CL-316243 are shown in Figure S2. (E–M) Data is shown as mean 6 SE with *p < 0.05 and **p < 0.01 Student’s t test (E–I); *p < 0.05 versus fed or vehicle and #p < 0.05 versus WT using a two-way analysis of variance (ANOVA) and Bonferonni’s post hoc test (K–M).

that GPR50 signaling in tanycytes modulates the response of colocalization of GPR50 (red) and the T3-transporter MCT8 these cells to changes in glucose. (green) was observed in tanycytes (Figure 2A). Expression of mct8 was significantly induced in the ependymal layer and Altered Hypothalamic Thyroid Hormones Are Not paraventricular nucleus (PVN) of WT mice in response to fast- Sufficient to Drive Torpor ing (Figures 2C and 2D). This induction was observed in the 2/2 Tanycytes play a central role in dictating T3 availability within PVN, but not the ependymal layer of Gpr50 mice (Figures the hypothalamus [19–21], and local T3 signaling is impli- 2F and 2G). The T3-converting enzyme dio2 was also strongly cated in hypothalamic responses to fasting [20]. Extensive induced in the ependymal layer of fasted WT mice (Figures 2B Current Biology Vol 22 No 1 72

Figure 2. Loss of Gpr50 Alters Thyroid Hormone Availability in the Hypothalamus (A) Extensive colocalization of GPR50 (red) and MCT8 (green) was observed in tanycyte cell bodies and processes. Blue represents DAPI staining. Scale bar represents 100 mm in the three left panels and 10 mm in the right panel. (B–F) Hypothalamic expressions of dio2 and mct8 and were determined by in situ hybridization in fed and fasted WT and Gpr502/2 mice (n = 6–8/group). Autoradiographs are representative of dio2 (B) and mct8 (C and D) expression, in fasted WT (B and C) or Gpr502/2 (D) mice. Fasting induced the expression of dio2 (E) and mct8 (F) in the ependymal layer of WT mice. Constitutive expression of dio2 was elevated in the knockouts. Fasting also led to a significant 2/2 increase in mct8 expression in the paraventricular nucleus (PVN) of WT and Gpr50 mice. Increased dio2 expression was reflected in hypothalamic T3 in Gpr502/2 mice (Gpr502/2, 3.9 6 0.3 pmol/g; WT, 3.3 6 0.2). (G) Data is shown as mean 6 SE with *p < 0.05 versus fed and #p < 0.05 versus WT using two-way ANOVA and Bonferonni’s post hoc test. Scale bar repre- sents 1.6 mm in (B)–(D).

(H) Central administration of T3 had no acute effects on Tb in WT mice (8 ng, intracerebroventricular [icv], n = 5/group). and 2E). Interestingly, dio2 expression was constitutively GPR50 and leptin receptor (Ob-Rb; LepR) are expressed at elevated in fed Gpr502/2 mice, achieving a similar level as high levels in the DMN (Figure 3E; [24]). Thus, attenuated that of fasted WT mice. These findings demonstrate that Gpr50 expression may facilitate torpor in ob/ob mice, and GPR50 modulates T3 handling in tanycytes and that altered the induction of GPR50 may be an underlying mechanism by T3 availability may therefore influence thermogenesis. A role which leptin blocks torpor in ob/ob mice and other species for hypothalamic T3 in seasonal torpor has been suggested [4, 5]. Leptin-responsive Gpr50 expression was confirmed [22]. However, neither acute (Figure 2H) nor chronic (Figure S3) in vitro, where leptin administration to Ob-Rb-expressing cells administration of T3 altered Tb in WT mice, suggesting that significantly enhanced Gpr50 promoter-driven luciferase changes in T3 are not sufficient to drive torpor and that activity (Figure 3F; Figure S4). a secondary signal of reduced energy status must be required. To test the ability of leptin to block torpor in Gpr502/2 mice, we treated fasted WT and KO mice with leptin 2 to 3 hr prior to GPR50 Modulates Leptin Signaling In Vivo and In Vitro the onset of torpor. Leptin administration blocked fasting-

Leptin is a peripherally derived indicator of energy status that induced drops in VO2 in WT mice (Figures 3G–3I; Figure S4). acts at multiple sites within the hypothalamus, including the In contrast, central or peripheral administration of leptin was DMN [23, 24]. Leptin is capable of blocking torpor in seasonal unable to attenuate torpor in the Gpr502/2 mice, despite the species [4–6], and mice deficient in leptin (ob/ob) are prone to ability of leptin to block torpor in other mouse and hamster torpor despite their massive fat reserves [5, 7]. Interestingly, models [5, 25]. Furthermore, in comparison to WT mice, leptin- Gpr50 expression in the DMN is significantly reduced in induced thermogenesis was also attenuated in KO mice in ob/ob mice but can be increased to levels seen in control a fed state (Figures 3J and 3K; Figure S4). The reduced efficacy mice (ob/wt) by leptin treatment (Figures 3A–3D). Both of leptin on Gpr502/2 mice was specific to thermogenesis, GPR50 Modulates Leptin Signaling 73

Figure 3. Interaction of GPR50 and Leptin (A–D) Gpr50 expression in torpor-prone ob/ob mice was examined by in situ hybridization (n = 8/group). In comparison with control ob/wt mice (A), Gpr50 expression was lower in ob/ob mice (B) in both the DMN and the ependymal layer of the third ventricle (D). Gpr50 expression could be rescued by admin- istration of exogenous leptin (2 mg/kg/day, ip) for 5 days (C and D). Data is shown as mean 6 SE with *p < 0.05 and **p < 0.01 using a one-way ANOVA and Bonferonni’s post hoc test. Scale bar represents 1.5 mm. (E) Dual immunolocalization of GPR50 (red) and LepR-eGFP (green) in the DMN. Scale bar represents 80 mm. (F) Leptin-responsive Gpr50 promoter activity. Leptin (100 nM) elicited a significant increase in Gpr50-luciferase reporter activity in HEK293 cells stably ex- pressing OB-Rb, but not in WT or OB-Ra-expressing cells. Induction of Gpr50 by leptin was also dependent on Jak2 and PI3K (Figure S4). Results are normalized first to an internal renilla-luciferase control, then to the activity of vehicle treated cells. No alteration in Gpr50 reporter activity was observed following leptin treatment of cells expressing the short (nonsignaling) form of the leptin receptor (Ob-Ra). Data is shown as mean 6 SE with *p < 0.05 versus control using a Student’s t test. (G–I) Administration of leptin at midnight (ZT17) prevented the fasting-induced drop in metabolic rate in WT mice (G), yet did not alter the expression or depth 2/2 of torpor in the Gpr50 mice (H; 200 pmol, icv, n = 6/group). Dashed lines in (I) reflect VO2 recorded the previous day during ad libitum feeding. (J and K) Leptin administration to fed WT mice increased Tb for approximately 6 hr (J; n = 6–8/group). The thermogenic response to leptin was significantly attenuated in Gpr502/2 mice (J and K). (L) To assess the effects of leptin on feeding, we administered leptin at the onset of the dark phase of the light cycle and ad libitum nocturnal food intake monitored (200 pmol, icv, n = 8/group). The anorexic actions of leptin were maintained in Gpr502/2 mice. Data is shown as mean 6 SE with *p < 0.05 versus vehicle and #p < 0.05 versus WT using a two-way ANOVA and Bonferonni’s post hoc test. because leptin administration remained equally effective at segregation of the anorexic and thermogenic actions of leptin reducing nocturnal food intake in both genotypes (Figure 3L). has been demonstrated in mice in which melanocortin These results demonstrate that the feeding-related and signaling has been targeted [26–28]. Specifically, loss of the thermogenic actions of leptin can be clearly differentiated MCR4 receptor blocks the effects of leptin on sympathetic and that only the latter is modulated by GPR50. A similar outflow to BAT, adipose gene expression, and fat mass, Current Biology Vol 22 No 1 74

Table 1. Impact of GPR50 on Transcriptional Response to Leptin

Condition Increased expression Decreased expression Total Ob-Rb vs Ob-Rb + Gpr50 299 307 606 Ob-Rb vs Ob-Rb + Leptin 827 490 1,327 Ob-Rb vs Ob-Rb + Gpr50 + leptin 1,605 1,100 2,705 Ob-Rb + leptin vs Ob-Rb + Gpr50 + leptin 1,022 1,016 2,038 p < 0.05 considered a change in gene expression. Functional pathway analysis revealed that genes exhibiting a significant change in expression, following leptin administration when cells express both Ob-Rb and Gpr50, are implicated in range of cellular processes, including amino acid transport, lipid meta- bolism, and metabolic disease (Figure S5). whereas the anorexic actions of leptin are maintained [27, 28]. mice (Figures 4F–4H) indicating that attenuated TRH is indeed The ability of GPR50 to influence cellular responses to leptin a causal mechanism in the thermogenic responses of the was assessed in vitro by gene microarray. This study revealed Gpr502/2 mice. that the proportion of genes showing a significant alteration in GPR50, Ob-Rb, and MCR4 receptors are all highly ex- expression following leptin treatment was substantially higher pressed in the DMN (here and [37–40]), suggesting that loss (2,705 genes) in cells expressing Ob-Rb in combination with of GPR50 in the DMN contributes to decreased thermogenesis Gpr50, compared with cells expressing Ob-Rb alone (1,327 under normal and fasted conditions, as well as the attenuated genes) (Table 1; Figure S5). response to leptin and MTII. That we did not observe any Our work reveals a critical role for GPR50 in leptin signaling. differences in the expression of npy, agrp, pomc, or cart within We show that Gpr50 expression is both responsive to leptin the ARC of WT and Gpr502/2 mice is in accord with a central and essential for the full impact of leptin signaling on energy role for the DMN in altered thermogenic response of the KO expenditure in vivo. mice. We envision that GPR50 neurons in the DMN modulate thermogenesis via brainstem sites that innervate BAT [41–43] Depressed Expression of Hypothalamic TRH Underlies and/or by modulating TRH-expressing neurons in the hypo- Altered Thermogenesis thalamus. The absence of GPR50-IR in the PVN (Figure S1) GPR50 exhibits an intense and restricted expression pattern necessitates that the influence of GPR50 on TRH expression within the hypothalamus, a structure that is a major regulator in the PVN is indirect; the most likely route being via DMN of BAT activity. Within the hypothalamus, the involvement of neurons known to contact CRH and TRH neurons in the PVN arcuate nucleus (ARC) neurons in fasting and leptin responsive [34]. Through this pathway, GPR50 would serve to modulate thermoregulation is well established [29], and the ARC projects DMN relay of fasting- and leptin-related signals to other hypo- heavily to both the PVN and DMN. Further, neuropeptide Y thalamic and brainstem nuclei. (NPY) and melanocortin signaling have been directly impli- cated in torpor [30, 31], and administration of the melanocortin Conclusions receptor agonist MTII elevates BAT thermogenesis and Tb In seasonal species, rhythmic torpor bouts are not controlled / [32]. Yet surprisingly, Gpr502 2 and WT mice exhibit a similar by proximal changes in food supply but require long-term expression of ARC neuropeptides npy, agrp, pomc, and cart metabolic adaption, which accompanies seasonal changes (Figure S4). Further, administration of MTII did not dampen in daylength (photoperiod). Physiological responses to photo- 2/2 torpor in the Gpr50 mice (Figures 4A–4C), despite blocking period cycles are driven by melatonin. Our studies now reveal torpor in other models [31]. Metabolic rate was increased by that GPR50, an ortholog of the avian melatonin Mel1C receptor 2/2 MTII in fed Gpr50 mice (Figure S4), demonstrating that the [44], has evolved a new role: as a critical modulator of adaptive mice are not deficient in melanocortin sensitivity per se but thermogenesis and torpor. That we have now linked reduced that GPR50 modulates its impact during fasting. expression of Gpr50 with heightened torpor response in three ARC neurons project heavily to corticotropin-releasing models, Siberian hamsters [45], leptin deficient (ob/ob) mice, hormone (CRH) and thyrotropin-releasing hormone (TRH)-ex- and Gpr50 knockout mice, implies that GPR50 normally serves pressing neurons of the PVN both directly and via the DMN to repress entry into a hypometabolic state by modulating [33, 34]. These populations are important in regulating energy thermal responses to energy signals such as glucose and balance and drive neuroendocrine and autonomic outputs of leptin. the hypothalamus. Quantification of CRH and TRH expression 2/2 revealed clear differences between WT and Gpr50 mice Experimental Procedures (Figures 4D and 4E). In contrast to WT mice, fasting did not elicit a rise in CRH expression in the PVN of Gpr502/2 mice. Animals and Surgical Procedures 2/2 Constitutive expression of TRH was also significantly lower Gpr50 mice were generated by DeltaGen (CA, USA) and obtained via in Gpr502/2 mice compared with WT mice. Fasting further AstraZeneca (Alderley Park, Cheshire, UK) [9]. Ob/ob mice were purchased from Charles River (UK). Age-matched or littermate male mice were used in inhibited TRH expression in both genotypes. TRH was also all experiments and maintained under a 12-12 hr light-dark schedule, unless significantly reduced in the anterior hypothalamus of KO stated otherwise. Studies were licensed under the Animals Act of 1986 and mice when compared with WT mice (Figure S3), but transcrip- local animal welfare committee. Implantation of intracerebroventricular (icv) tional changes were not universal, because CRH expression in guide cannulae and remote telemetry probes (DataScience International, the amygdala and TRH expression in the DMN did not differ The Netherlands) was as previously described [46]. significantly between genotypes. Central TRH administration increases BAT activity [35, 36], suggesting that depressed Indirect Calorimetry, 2-DG, and Blood Glucose 2/2 Metabolic gases (O2 and CO2) were measured using indirect calorimetric expression of TRH in the PVN and AH of Gpr50 mice under- cages (Columbus Instruments, Columbus, OH, USA). Icv injection of mouse pins their hypometabolic phenotype. Central administration of leptin (200 pmol; Sigma), RX77368 (2 ng, gift from Professors Geoffrey 2/2 RX77368 (a stable analog of TRH) blocked torpor in Gpr50 Bennett and Fran Ebling, University of Nottingham), T3 (8 ng, Sigma) or GPR50 Modulates Leptin Signaling 75

Figure 4. Association of Torpor with Depressed TRH Expression in the Hypothalamus (A–C) Administration of MTII at midnight (ZT17) increased metabolic rate in fasted WT mice (A and C) but did not alter the expression or depth of torpor in the Gpr502/2 mice (B and C; 1 nmol, icv, n = 4–6/group). Data is shown as mean 6 SE with *p < 0.05 versus vehicle and #p < 0.05 versus WT using a two-way ANOVA and Bonferonni’s post hoc test. (D and E) Expression of CRH (D) and thyrotropin releasing hormone (TRH) (E) were quantified in the PVN of WT and Gpr502/2 mice under fed (top) and fasted (bottom) conditions by in situ hybridization histology (n = 8/group). WT mice exhibited a significant induction of CRH expression in response to fasting, whereas no such change was observed in Gpr502/2 mice (D). A significant decrease in TRH expression was observed in the PVN of WT mice upon fasting (E). TRH expression in ad libitum fed Gpr502/2 mice was significantly lower than that of WT mice and was reduced further upon fasting (F). Data is shown as mean 6 SE with **p < 0.01 versus fed and #p < 0.05 versus WT using a two-way ANOVA and Bonferonni’s post hoc test. Scale bar represents 800 mm. (F–H) Central administration of the TRH agonist RX77368 (2 mg, icv) to fasted WT (n = 4/group; F) and Gpr502/2 mice (n = 8/group; G) prevented the fasting- 2/2 induced drop in VO2 in Gpr50 mice (H). Data is shown as mean 6 SE with **p < 0.01 versus vehicle and #p < 0.05 versus WT using a two-way ANOVA and Bonferonni’s post hoc test. vehicle (0.9% NaCl) were delivered in 1 ml over a period of 30 s. During 2-DG Omission of primary or secondary antibody resulted in no positive (Sigma) administration, food was removed and circulating glucose immunoreaction. measured by tail tip bleed with an Optimum Plus glucose meter (Abbott Laboratories, UK). Mice were acclimated to calorimetric cages, handling, Reporter Gene Activation Assay and blood collection for 2 days prior to study. HEK293 stable cell lines expressing human OB-Rb or OB-Ra were cotrans- fected with a Gpr50 luciferase reporter plasmid (257 bp upstream of Gpr50 In Situ Hybridization and Real-Time Quantitative PCR start codon; 250 ng) and a Renilla luciferase plasmid (25 ng). Cells were then Tissues were processed as previously described [1, 47]. Primers used for treated with or without leptin (100 nM) for 6 hr after 1 hr pretreatment with or generating riboprobes (in situ) and qPCR probes are listed in the Supple- without 50 mM AG490 or 50 mM Wortmanin (Sigma). Luciferase activity was mental Information. TRH, CRH, and Dio2 probes were kindly provided by normalized to renilla activity. Dr. Perry Barrett (University of Aberdeen, UK). Products were cloned into p-GEMT Easy Vector (Promega, Madison, USA), riboprobes synthesized m-Array Experiments with 33P-UTP (MP Biomedical, USA), and hybridization visualized by film HEK293 cells stably expressing OB-Rb were transfected with a Gpr50 autoradiography (Kodak BioMax MR film, Kodak, USA). Optical density expression plasmid or mock transfected and treated with or without leptin was determined using 3–4 sections/mouse/area. QPCR was performed (100 nM) for 6 hr. RNA was purified and validated (using Agilent RNA6000 using the Platinum SyBR Green Kit (Invitrogen). Housekeeping genes 18S nano chip kit; Bioanalyzer 2100). Complementary DNA was produced, end rRNA or cyclophilin were used as controls. labeled with biotin, hybridized to GeneChipÒ human Gene (Affymetrix), and scanned using the GCS3000 7G. Images were analyzed with Expression Immunohistochemistry Console software (Affymetrix). Robust multichip average (RMA) was Brains were removed and fixed for 48 hr in Bouin’s fixative (Sigma). normalized using the RMA option of Expression Console and subjected to Frozen sections (20–30 mm) were blocked with serum followed by primary statistical analysis. Pathway analysis was performed with the Ingenuity and secondary antibodies (see Supplemental Experimental Procedures). program. Current Biology Vol 22 No 1 76

Supplemental Information 15. Zhang, J., Kaasik, K., Blackburn, M.R., and Lee, C.C. (2006). Constant darkness is a circadian metabolic signal in mammals. Nature 439, Supplemental Information includes five figures and Supplemental 340–343. Experimental Procedures and can be found with this article online at 16. Swoap, S.J., Gutilla, M.J., Liles, L.C., Smith, R.O., and Weinshenker, D. doi:10.1016/j.cub.2011.11.043. (2006). The full expression of fasting-induced torpor requires beta 3-adrenergic receptor signaling. J. Neurosci. 26, 241–245. Acknowledgments 17. Garcı´a, M.A., Milla´ n, C., Balmaceda-Aguilera, C., Castro, T., Pastor, P., Montecinos, H., Reinicke, K., Zu´ n˜ iga, F., Vera, J.C., On˜ ate, S.A., and 2/2 We thank John Brennand (AstraZeneca, UK) for providing Gpr50 mice, Nualart, F. (2003). Hypothalamic ependymal-glial cells express the Thomas Bourgeron (Institut Pasteur, Paris) for kindly providing the Gpr50- glucose transporter GLUT2, a protein involved in glucose sensing. luc reporter, and Martin Myers Jr. (University of Michigan) for kindly J. Neurochem. 86, 709–724. providing LepR-eGFP tissue. Microarray experiments and data analysis 18. Sanders, N.M., Dunn-Meynell, A.A., and Levin, B.E. (2004). Third ventric- were performed by the Genomic Facility of the Cochin Institute. This work ular alloxan reversibly impairs glucose counterregulatory responses. was supported by grants from the Biotechnology and Biological Sciences Diabetes 53, 1230–1236. Research Council (S.M.L., A.S.I.L.), Equipe Fondation Recherche Me´ dicale 19. Rodrı´guez, E.M., Bla´ zquez, J.L., Pastor, F.E., Pela´ ez, B., Pen˜ a, P., (J.D., R.J.), Institut National de la Sante´ et de la Recherche Me´ dicale Peruzzo, B., and Amat, P. (2005). Hypothalamic tanycytes: a key (J.D., R.J.), and Centre National de la Recherche Scientifique (J.D., R.J.). component of brain-endocrine interaction. Int. Rev. Cytol. 247, 89–164. 20. Coppola, A., Liu, Z.W., Andrews, Z.B., Paradis, E., Roy, M.C., Friedman, Received: July 18, 2011 J.M., Ricquier, D., Richard, D., Horvath, T.L., Gao, X.B., and Diano, S. Revised: November 3, 2011 (2007). A central thermogenic-like mechanism in feeding regulation: Accepted: November 21, 2011 an interplay between arcuate nucleus T3 and UCP2. Cell Metab. 5, Published online: December 22, 2011 21–33. 21. Lo´ pez, M., Varela, L., Va´ zquez, M.J., Rodrı´guez-Cuenca, S., References Gonza´ lez, C.R., Velagapudi, V.R., Morgan, D.A., Schoenmakers, E., Agassandian, K., Lage, R., et al. (2010). Hypothalamic AMPK and fatty 1. Heldmaier, G., Ortmann, S., and Elvert, R. (2004). Natural hypometabo- acid metabolism mediate thyroid regulation of energy balance. Nat. lism during hibernation and daily torpor in mammals. Respir. Physiol. Med. 16, 1001–1008. Neurobiol. 141, 317–329. 22. Murphy, M., Jethwa, P.H., Warner, A., Barrett, P., Nilaweera, K.N., 2. Geiser, F. (2004). Metabolic rate and body temperature reduction during Brameld, J.M., and Ebling, F.J. (2011). Effects of Manipulating hibernation and daily torpor. Annu. Rev. Physiol. 66, 239–274. Hypothalamic Triiodothyronine Concentrations on Seasonal Body 3. Melvin, R.G., and Andrews, M.T. (2009). Torpor induction in mammals: Weight and Torpor Cycles in Siberian Hamsters. Endocrinology, in recent discoveries fueling new ideas. Trends Endocrinol. Metab. 20, press. Published online October 25, 2011. 490–498. 4. Webb, G.P., Jagot, S.A., and Jakobson, M.E. (1982). Fasting-induced 23. Gautron, L., Lazarus, M., Scott, M.M., Saper, C.B., and Elmquist, J.K. torpor in Mus musculus and its implications in the use of murine models (2010). Identifying the efferent projections of leptin-responsive neurons for human obesity studies. Comp. Biochem. Physiol. Comp. Physiol. 72, in the dorsomedial hypothalamus using a novel conditional tracing 211–219. approach. J. Comp. Neurol. 518, 2090–2108. 5. Gavrilova, O., Leon, L.R., Marcus-Samuels, B., Mason, M.M., Castle, 24. Patterson, C.M., Leshan, R.L., Jones, J.C., and Myers, M.G., Jr. (2011). A.L., Refetoff, S., Vinson, C., and Reitman, M.L. (1999). Torpor in mice Molecular mapping of mouse brain regions innervated by leptin is induced by both leptin-dependent and -independent mechanisms. receptor-expressing cells. Brain Res. 1378, 18–28. Proc. Natl. Acad. Sci. USA 96, 14623–14628. 25. Do¨ ring, H., Schwarzer, K., Nuesslein-Hildesheim, B., and Schmidt, I. 6. Freeman, D.A., Lewis, D.A., Kauffman, A.S., Blum, R.M., and Dark, J. (1998). Leptin selectively increases energy expenditure of food- (2004). Reduced leptin concentrations are permissive for display of restricted lean mice. Int. J. Obes. Relat. Metab. Disord. 22, 83–88. torpor in Siberian hamsters. Am. J. Physiol. Regul. Integr. Comp. 26. Satoh, N., Ogawa, Y., Katsuura, G., Numata, Y., Masuzaki, H., Physiol. 287, R97–R103. Yoshimasa, Y., and Nakao, K. (1998). Satiety effect and sympathetic 7. Gluck, E.F., Stephens, N., and Swoap, S.J. (2006). Peripheral ghrelin activation of leptin are mediated by hypothalamic melanocortin system. deepens torpor bouts in mice through the arcuate nucleus neuropeptide Neurosci. Lett. 249, 107–110. Y signaling pathway. Am. J. Physiol. Regul. Integr. Comp. Physiol. 291, 27. Ste Marie, L., Miura, G.I., Marsh, D.J., Yagaloff, K., and Palmiter, R.D. R1303–R1309. (2000). A metabolic defect promotes obesity in mice lacking melanocor- 8. Cannon, B., and Nedergaard, J. (2004). Brown adipose tissue: function tin-4 receptors. Proc. Natl. Acad. Sci. USA 97, 12339–12344. and physiological significance. Physiol. Rev. 84, 277–359. 28. Zhang, Y., Kilroy, G.E., Henagan, T.M., Prpic-Uhing, V., Richards, W.G., 9. Ivanova, E.A., Bechtold, D.A., Dupre´ , S.M., Brennand, J., Barrett, P., Bannon, A.W., Mynatt, R.L., and Gettys, T.W. (2005). Targeted deletion Luckman, S.M., and Loudon, A.S. (2008). Altered metabolism in the of subtypes 3 and 4, but not CART, alters nutrient melatonin-related receptor (GPR50) knockout mouse. Am. J. Physiol. partitioning and compromises behavioral and metabolic responses to Endocrinol. Metab. 294, E176–E182. leptin. FASEB J. 19, 1482–1491. 10. Sidibe, A., Mullier, A., Chen, P., Baroncini, M., Boutin, J.A., Delagrange, 29. Elmquist, J.K., Coppari, R., Balthasar, N., Ichinose, M., and Lowell, B.B. P., Prevot, V., and Jockers, R. (2010). Expression of the orphan GPR50 (2005). Identifying hypothalamic pathways controlling food intake, body protein in rodent and human dorsomedial hypothalamus, tanycytes and weight, and glucose homeostasis. J. Comp. Neurol. 493, 63–71. median eminence. J. Pineal Res. 48, 263–269. 30. Dark, J., and Pelz, K.M. (2008). NPY Y1 receptor antagonist prevents 11. Nilaweera, K., Herwig, A., Bolborea, M., Campbell, G., Mayer, C.D., NPY-induced torpor-like hypothermia in cold-acclimated Siberian Morgan, P.J., Ebling, F.J., and Barrett, P. (2011). Photoperiodic regula- hamsters. Am. J. Physiol. Regul. Integr. Comp. Physiol. 294, R236– tion of glycogen metabolism, glycolysis, and glutamine synthesis in R245. tanycytes of the Siberian hamster suggests novel roles of tanycytes in hypothalamic function. Glia 59, 1695–1705. 31. Mounien, L., Marty, N., Tarussio, D., Metref, S., Genoux, D., Preitner, F., 12. Dimicco, J.A., and Zaretsky, D.V. (2007). The dorsomedial hypothal- Foretz, M., and Thorens, B. (2010). Glut2-dependent glucose-sensing amus: a new player in thermoregulation. Am. J. Physiol. Regul. Integr. controls thermoregulation by enhancing the leptin sensitivity of NPY Comp. Physiol. 292, R47–R63. and POMC neurons. FASEB J. 24, 1747–1758. 13. Tsukada, M., Hamade, N., Wada, Y., Mogi, T., Kamiyama, S., and 32. Brito, M.N., Brito, N.A., Baro, D.J., Song, C.K., and Bartness, T.J. (2007). Koizumi, A. (1993). Energy restriction suppresses microsomal Ca(2+)- Differential activation of the sympathetic innervation of adipose tissues ATPase activities in various organs in C57BL/6 female mice in both by melanocortin receptor stimulation. Endocrinology 148, 5339–5347. euthermic and torpor states. Mech. Ageing Dev. 68, 183–189. 33. Cone, R.D. (2005). Anatomy and regulation of the central melanocortin 14. Rikke, B.A., Yerg, J.E., 3rd, Battaglia, M.E., Nagy, T.R., Allison, D.B., and system. Nat. Neurosci. 8, 571–578. Johnson, T.E. (2003). Strain variation in the response of body tempera- 34. Lechan, R.M., and Fekete, C. (2006). The TRH neuron: a hypothalamic ture to dietary restriction. Mech. Ageing Dev. 124, 663–678. integrator of energy metabolism. Prog. Brain Res. 153, 209–235. GPR50 Modulates Leptin Signaling 77

35. Griffiths, E.C., Rothwell, N.J., and Stock, M.J. (1988). Thermogenic effects of thyrotrophin-releasing hormone and its analogues in the rat. Experientia 44, 40–42. 36. Rogers, R.C., Barnes, M.J., and Hermann, G.E. (2009). Leptin ‘‘gates’’ thermogenic action of thyrotropin-releasing hormone in the hindbrain. Brain Res. 1295, 135–141. 37. Elmquist, J.K., Ahima, R.S., Elias, C.F., Flier, J.S., and Saper, C.B. (1998). Leptin activates distinct projections from the dorsomedial and ventromedial hypothalamic nuclei. Proc. Natl. Acad. Sci. USA 95, 741–746. 38. Kishi, T., Aschkenasi, C.J., Lee, C.E., Mountjoy, K.G., Saper, C.B., and Elmquist, J.K. (2003). Expression of mRNA in the central nervous system of the rat. J. Comp. Neurol. 457, 213–235. 39. Bi, S., Robinson, B.M., and Moran, T.H. (2003). Acute food deprivation and chronic food restriction differentially affect hypothalamic NPY mRNA expression. Am. J. Physiol. Regul. Integr. Comp. Physiol. 285, R1030–R1036. 40. Caron, E., Sachot, C., Prevot, V., and Bouret, S.G. (2010). Distribution of leptin-sensitive cells in the postnatal and adult mouse brain. J. Comp. Neurol. 518, 459–476. 41. Cano, G., Passerin, A.M., Schiltz, J.C., Card, J.P., Morrison, S.F., and Sved, A.F. (2003). Anatomical substrates for the central control of sympathetic outflow to interscapular adipose tissue during cold expo- sure. J. Comp. Neurol. 460, 303–326. 42. Oldfield, B.J., Giles, M.E., Watson, A., Anderson, C., Colvill, L.M., and McKinley, M.J. (2002). The neurochemical characterisation of hypotha- lamic pathways projecting polysynaptically to brown adipose tissue in the rat. Neuroscience 110, 515–526. 43. Morrison, S.F. (2004). Activation of 5-HT1A receptors in raphe pallidus inhibits leptin-evoked increases in brown adipose tissue thermogene- sis. Am. J. Physiol. Regul. Integr. Comp. Physiol. 286, R832–R837. 44. Dufourny, L., Levasseur, A., Migaud, M., Callebaut, I., Pontarotti, P., Malpaux, B., and Monget, P. (2008). GPR50 is the mammalian ortholog of Mel1c: evidence of rapid evolution in mammals. BMC Evol. Biol. 8, 105. 45. Barrett, P., Ebling, F.J., Schuhler, S., Wilson, D., Ross, A.W., Warner, A., Jethwa, P., Boelen, A., Visser, T.J., Ozanne, D.M., et al. (2007). Hypothalamic thyroid hormone catabolism acts as a gatekeeper for the seasonal control of body weight and reproduction. Endocrinology 148, 3608–3617. 46. Bechtold, D.A., Ivanov, T.R., and Luckman, S.M. (2009). Appetite-modi- fying actions of pro-neuromedin U-derived peptides. Am. J. Physiol. Endocrinol. Metab. 297, E545–E551. 47. Li, J., Hand, L.E., Meng, Q.J., Loudon, A.S., and Bechtold, D.A. (2011). GPR50 interacts with TIP60 to modulate glucocorticoid receptor signal- ling. PLoS ONE 6, e23725. APPENDIX B

Significantly altered genes in Gpr50-/- versus WT mice when fasted.

Affymetrix mouse arrays were used to screen the differentially expressed genes in the hypothalamus between fed and fasted WT and Gpr50-/- mice. Significantly altered genes with a q<0.05 and fold change>1 only are included.

Gene Assignment Gene Symbol q value Fold Change ENSMUST00000079045 // B230325K18Rik // RIKEN B230325K18Rik 0.014 2.58 cDNA B230325K18 gene // 7 F3 // 319 NM_001009935 // Txnip // thioredoxin interacting Txnip 0.011 2.55 protein // 3 47.1 cM // 56338 / NM_175232 // 5830427D03Rik // RIKEN cDNA 5830427D03Rik 0.009 2.53 5830427D03 gene // 15 F1 // 76061 /// B NM_015786 // Hist1h1c // histone cluster 1, H1c // 13 Hist1h1c 0.009 2.49 A2-A3 // 50708 /// ENSMUST NM_010340 // Gpr50 // G-protein-coupled receptor 50 Gpr50 0.028 2.06 // X A7.2|X 26.0 cM // 14765 NM_010426 // Foxf1a // forkhead box F1a // 8 67.0 cM Foxf1a 0.011 2.02 // 15227 /// ENSMUST0000005 NM_027453 // Btf3l4 // basic transcription factor 3-like 4 Btf3l4 0.014 1.96 // 4 C7 // 70533 /// NM_001131020 // Gfap // glial fibrillary acidic protein // Gfap 0.035 1.88 11 D|11 62.0 cM // 14 NM_001081430 // Nat12 // N-acetyltransferase 12 // 14 Nat12 0.011 1.78 C1 // 70646 /// ENSMUST000 NM_133237 // Apcdd1 // adenomatosis polyposis coli Apcdd1 0.049 1.74 down-regulated 1 // 18 E1 // NM_013642 // Dusp1 // dual specificity phosphatase 1 Dusp1 0.043 1.67 // 17 A2-C|17 13.0 cM // 19 NM_010831 // Sik1 // salt inducible kinase 1 // 17 B1|17 Sik1 0.026 1.65 18.18 cM // 17691 /// E NM_153287 // Csrnp1 // cysteine-serine-rich nuclear Csrnp1 0.009 1.63 protein 1 // 9 F4 // 215418 NM_001159367 // Per1 // period homolog 1 Per1 0.016 1.62 (Drosophila) // 11 B // 18626 /// NM_01 NM_008654 // Myd116 // myeloid differentiation Myd116 0.002 1.59 primary response gene 116 // 7 B4 NM_008211 // H3f3b // H3 histone, family 3B // 11 E2 // H3f3b 0.014 1.58 15081 /// BC037730 // H3 NM_001159496 // Ppm1b // protein phosphatase 1B, Ppm1b 0.009 1.50 magnesium dependent, beta isofo NM_178691 // Yod1 // YOD1 OTU deubiquitinating Yod1 0.009 1.50 enzyme 1 homologue (S. cerevisiae NM_023799 // Mgea5 // meningioma expressed antigen Mgea5 0.018 1.49 5 (hyaluronidase) // 19 D1|19 NM_010495 // Id1 // inhibitor of DNA binding 1 // 2 Id1 0.009 1.49 H1|2 84.0 cM // 15901 /// EN NM_008803 // Pde8a // phosphodiesterase 8A // 7 D3 // Pde8a 0.030 1.48 18584 /// ENSMUST000000266 NM_024469 // Bhlhe41 // basic helix-loop-helix family, Bhlhe41 0.037 1.48

236 member e41 // 6 G2-G3|6 7 AK049490 // C230029D21Rik // RIKEN cDNA C230029D21Rik 0.021 1.46 C230029D21 gene // 17 A1 // 320255 /// A NM_029804 // Hnrnpm // heterogeneous nuclear Hnrnpm 0.019 1.45 ribonucleoprotein M // 17 B1 // 769 NM_178194 // Hist1h2be // histone cluster 1, H2be // 13 Hist1h2be 0.007 1.45 A2-A3 // 319179 /// NM_0 NM_175548 // Lsamp // limbic system-associated Lsamp 0.000 1.44 membrane protein // 16 B5 // 2688 NM_027304 // H1fnt // H1 histone family, member N, H1fnt 0.031 1.44 testis-specific // 15 F1 // 7 NM_145556 // Tardbp // TAR DNA binding protein // 4 Tardbp 0.010 1.44 E2 // 230908 /// NM_00100854 NM_029248 // Taf1d // TATA box binding protein (Tbp)- Taf1d 0.034 1.42 associated factor, RNA poly NM_020518 // Vsig2 // V-set and immunoglobulin Vsig2 0.049 1.41 domain containing 2 // 9 B // 572 NM_175286 // C430004E15Rik // RIKEN cDNA C430004E15Rik 0.014 1.41 C430004E15 gene // 2 A3 // 97031 /// EN NM_028287 // Zufsp // zinc finger with UFM1-specific Zufsp 0.029 1.40 peptidase domain // 10 B1 / AK021026 // B430319H21Rik // RIKEN cDNA B430319H21Rik 0.048 1.39 B430319H21 gene // 11 B3 // 77849 NM_016967 // Olig2 // oligodendrocyte transcription Olig2 0.020 1.39 factor 2 // 16 C3.3|16 63.0 NM_008842 // Pim1 // proviral integration site 1 // 17 Pim1 0.029 1.35 A3.3|17 16.4 cM // 18712 NM_026072 // Sdccag10 // serologically defined colon Sdccag10 0.019 1.35 cancer antigen 10 // 13 D1 NM_027427 // Taf15 // TAF15 RNA polymerase II, TATA Taf15 0.032 1.35 box binding protein (TBP)-as NM_177089 // Tacc1 // transforming, acidic coiled-coil Tacc1 0.049 1.33 containing protein 1 // 8 NM_016669 // Crym // crystallin, mu // 7 F2|7 55.0 cM Crym 0.023 1.32 // 12971 /// ENSMUST000000 NM_013712 // Itgb1bp2 // integrin beta 1 binding Itgb1bp2 0.020 1.32 protein 2 // X D // 26549 /// N NM_011659 // Tnfrsf4 // tumor necrosis factor receptor Tnfrsf4 0.023 1.32 superfamily, member 4 // NM_008235 // Hes1 // hairy and enhancer of split 1 Hes1 0.031 1.31 (Drosophila) // 16 B2|16 27.0 NM_145434 // Nr1d1 // nuclear receptor subfamily 1, Nr1d1 0.032 1.31 group D, member 1 // 11 D // NM_007455 // Ap1g2 // adaptor protein complex AP-1, Ap1g2 0.009 1.30 gamma 2 subunit // 14 C3 // NM_146041 // Gmds // GDP-mannose 4, 6-dehydratase Gmds 0.037 1.30 // 13 A3.2 // 218138 /// ENSMU NM_001039522 // Leo1 // Leo1, Paf1/RNA polymerase II Leo1 0.035 1.30 complex component, homolog NM_029239 // Prkd3 // protein kinase D3 // 17 E3 // Prkd3 0.015 1.29 75292 /// ENSMUST00000003191 NM_001004436 // Wapal // wings apart-like homolog Wapal 0.047 1.28 (Drosophila) // 14 B // 218914 NM_007842 // Dhx9 // DEAH (Asp-Glu-Ala-His) box Dhx9 0.042 1.27 polypeptide 9 // 1 G3|1 77.0 cM

237

NM_009703 // Araf // v-raf murine sarcoma 3611 viral Araf 0.017 1.27 oncogene homolog // X A2-A3 NM_027226 // Fyttd1 // forty-two-three domain Fyttd1 0.049 1.26 containing 1 // 16 B3 // 69823 /// NM_172265 // Eif2b5 // eukaryotic translation initiation Eif2b5 0.015 1.25 factor 2B, subunit 5 ep NM_181345 // Npm2 // nucleophosmin/nucleoplasmin Npm2 0.021 1.25 2 // 14 D2 // 328440 /// NM_023 NM_023651 // Pex13 // peroxisomal biogenesis factor Pex13 0.024 1.24 13 // 11 A3.3 // 72129 /// E NM_023644 // Mccc1 // methylcrotonoyl-Coenzyme A Mccc1 0.037 1.24 carboxylase 1 (alpha) // 3 B // NM_011875 // Psmd13 // proteasome (prosome, Psmd13 0.024 1.24 macropain) 26S subunit, non-ATPase, NM_009102 // Rrh // retinal pigment epithelium derived Rrh 0.010 1.24 rhodopsin homolog // 3 G3 NM_145985 // Arcn1 // archain 1 // 9 A5.2 // 213827 /// Arcn1 0.034 1.23 ENSMUST00000034607 // Ar NM_028298 // Zfp655 // zinc finger protein 655 // 5 G2 Zfp655 0.021 1.23 // 72611 /// NM_001083958 NM_183000 // Accn3 // amiloride-sensitive cation Accn3 0.009 1.22 channel 3 // 5 A3 // 171209 /// NM_013862 // Rabgap1l // RAB GTPase activating Rabgap1l 0.045 1.22 protein 1-like // 1 H2.1 // 29809 NM_001013380 // Dync1li2 // dynein, cytoplasmic 1 Dync1li2 0.049 1.21 light intermediate chain 2 // NM_028320 // Adipor1 // adiponectin receptor 1 // 1 E4 Adipor1 0.035 1.21 // 72674 /// ENSMUST00000 NM_009640 // Angpt1 // angiopoietin 1 // 15 B3.1|15 Angpt1 0.020 1.20 14.3 cM // 11600 /// ENSMUST NM_023493 // Cml5 // camello-like 5 // 6 C3 // 69049 Cml5 0.013 1.20 /// NM_053097 // Cml3 // ca NM_015823 // Magi2 // membrane associated guanylate Magi2 0.037 1.20 kinase, WW and PDZ domain co NM_009828 // Ccna2 // cyclin A2 // 3 B|3 19.2 cM // Ccna2 0.019 1.20 12428 /// ENSMUST00000029270 NM_021450 // Trpm7 // transient receptor potential Trpm7 0.026 1.19 cation channel, subfamily M, NM_028108 // Nat13 // N-acetyltransferase 13 // 16 B4 Nat13 0.018 1.19 // 72117 /// ENSMUST000000 NM_144892 // Ncoa5 // nuclear receptor coactivator 5 Ncoa5 0.029 1.19 // 2 H3 // 228869 /// ENSMU NM_199080 // Ddx17 // DEAD (Asp-Glu-Ala-Asp) box Ddx17 0.024 1.15 polypeptide 17 // 15 E1 // 6704 NM_008942 // Npepps // aminopeptidase puromycin Npepps 0.032 1.09 sensitive // 11 D|11 56.0 cM // NM_078478 // Ghitm // growth hormone inducible Ghitm 0.045 1.08 transmembrane protein // 14 B|14 NM_152894 // Pop1 // processing of precursor 1, Pop1 0.022 -1.10 ribonuclease P/MRP family, (S. c NM_029291 // Ascc2 // activating signal cointegrator 1 Ascc2 0.048 -1.11 complex subunit 2 // 11 A NM_019651 // Ptpn9 // protein tyrosine phosphatase, Ptpn9 0.042 -1.15 non-receptor type 9 // --- / NM_144835 // Heatr1 // HEAT repeat containing 1 // 13 Heatr1 0.024 -1.15

238

A1 // 217995 /// ENSMUST00 NM_145140 // Abcc10 // ATP-binding cassette, sub- Abcc10 0.033 -1.16 family C (CFTR/MRP), member 10 NM_013925 // Adat1 // adenosine deaminase, tRNA- Adat1 0.036 -1.16 specific 1 // 8 D3 // 30947 /// NM_172768 // Gramd1b // GRAM domain containing 1B Gramd1b 0.023 -1.16 // 9 B // 235283 /// ENSMUST00 NM_015752 // Sufu // suppressor of fused homolog Sufu 0.034 -1.17 (Drosophila) // 19 C3|19 47.0 c NM_172694 // Megf9 // multiple EGF-like-domains 9 // Megf9 0.035 -1.18 4 C2 // 230316 /// ENSMUST0 NM_025617 // 2210012G02Rik // RIKEN cDNA 2210012G02Rik 0.009 -1.18 2210012G02 gene // 4 C7 // 66526 /// EN NM_010808 // Mmp24 // matrix metallopeptidase 24 // Mmp24 0.043 -1.19 2 H1|2 87.5 cM // 17391 /// NM_008992 // Abcd4 // ATP-binding cassette, sub- Abcd4 0.005 -1.20 family D (ALD), member 4 // 12 D NM_001025156 // Ccdc93 // coiled-coil domain Ccdc93 0.033 -1.21 containing 93 // 1 E2 // 70829 /// NM_010719 // Lipe // lipase, hormone sensitive // 7 Lipe 0.029 -1.22 A3|7 5.5 cM // 16890 /// NM_ NM_025864 // Tmem206 // transmembrane protein 206 Tmem206 0.025 -1.22 // 1 H6 // 66950 /// ENSMUST00 NM_010247 // Xrcc6 // X-ray repair complementing Xrcc6 0.037 -1.23 defective repair in Chinese ham NM_026277 // Nob1 // NIN1/RPN12 binding protein 1 Nob1 0.049 -1.23 homolog (S. cerevisiae) // 8 D NM_172151 // Zdhhc8 // zinc finger, DHHC domain Zdhhc8 0.045 -1.23 containing 8 // 16 A3|16 10.83 c NM_172581 // Fam161b // family with sequence Fam161b 0.009 -1.25 similarity 161, member B // 12 D1 / NM_146185 // Zfp790 // zinc finger protein 790 // 7 B1 Zfp790 0.049 -1.26 // 233056 /// NM_00114588 NM_194334 // Tbc1d2b // TBC1 domain family, member Tbc1d2b 0.045 -1.27 2B // 9 E3.1 // 67016 /// NM_ BC029621 // Prpsap1 // phosphoribosyl pyrophosphate Prpsap1 0.050 -1.27 synthetase-associated protei NM_001005247 // Hps5 // Hermansky-Pudlak syndrome Hps5 0.011 -1.27 5 homolog (human) // 7 B4|7 25 NM_176840 // Osbpl11 // oxysterol binding protein-like Osbpl11 0.011 -1.27 11 // 16 B3 // 106326 /// ENSMUST00000019572 // 9630025I21Rik // RIKEN cDNA 9630025I21Rik 0.047 -1.27 9630025I21 gene // 13 B3 // 40 NM_015795 // Fbxo16 // F-box protein 16 // 14 D1 // Fbxo16 0.049 -1.27 50759 /// ENSMUST00000043554 NM_177460 // Parp16 // poly (ADP-ribose) polymerase Parp16 0.045 -1.28 family, member 16 // 9 C // NM_017407 // Spag5 // sperm associated antigen 5 // Spag5 0.024 -1.28 11 B5|11 44.95 cM // 54141 / NM_029581 // Mtif3 // mitochondrial translational Mtif3 0.035 -1.28 initiation factor 3 // 5 G3 // NM_012017 // Zfp346 // zinc finger protein 346 // --- // Zfp346 0.034 -1.28 26919 /// ENSMUST000000 NM_178782 // Bcorl1 // BCL6 co-repressor-like 1 // X A4 Bcorl1 0.019 -1.29 // 320376 /// ENSMUST000

239

NM_174847 // C2cd2 // C2 calcium-dependent domain C2cd2 0.045 -1.30 containing 2 // 16 C4 // 20778 NM_019756 // Tubd1 // tubulin, delta 1 // 11 C // 56427 Tubd1 0.027 -1.31 /// ENSMUST00000069503 / NM_028099 // Dusp11 // dual specificity phosphatase 11 Dusp11 0.002 -1.31 (RNA/RNP complex 1-intera NM_053081 // Fancg // Fanconi anemia, Fancg 0.000 -1.32 complementation group G // 4 B1 // 60534 / BC008558 // 1700030K09Rik // RIKEN cDNA 1700030K09 1700030K09Rik 0.009 -1.32 gene // 8 B3.3 // 72254 /// E NM_172256 // Dync2li1 // dynein cytoplasmic 2 light Dync2li1 0.048 -1.32 intermediate chain 1 // 17 E NM_011945 // Map3k1 // mitogen-activated protein Map3k1 0.024 -1.32 kinase kinase kinase 1 // 13 D2 NM_001142647 // Tmem194b // transmembrane Tmem194b 0.032 -1.33 protein 194B // 1 C1.1 // 227094 /// N NM_153802 // Zfp128 // zinc finger protein 128 // 7 A1 Zfp128 0.009 -1.34 // 243833 /// ENSMUST0000 NM_001122768 // Lrrc8d // leucine rich repeat Lrrc8d 0.034 -1.34 containing 8D // 5 E5 // 231549 // NM_175247 // Zfp28 // zinc finger protein 28 // 7 A1 // Zfp28 0.035 -1.35 22690 /// ENSMUST0000008 BC157917 // Slc38a6 // solute carrier family 38, member Slc38a6 0.044 -1.35 6 // 12 C3 // 625098 /// NM_001037707 // Zfp27 // zinc finger protein 27 // 7 B1 Zfp27 0.011 -1.36 // 22689 /// NM_011754 / NM_010509 // Ifnar2 // interferon (alpha and beta) Ifnar2 0.043 -1.36 receptor 2 // 16 C3.3|16 63.1 NM_175490 // Gpr75 // G protein-coupled receptor 75 Gpr75 0.010 -1.36 // 11 A4 // 237716 /// ENSMU NM_133931 // Pot1a // protection of telomeres 1A // 6 Pot1a 0.049 -1.36 A3.1 // 101185 /// ENSMUST NM_001142918 // Tcf7l2 // transcription factor 7-like 2, Tcf7l2 0.028 -1.37 T-cell specific, HMG-bo BC002181 // 2310033P09Rik // RIKEN cDNA 2310033P09 2310033P09Rik 0.022 -1.38 gene // 11 B2 // 67862 /// EN NM_198176 // Fastkd5 // FAST kinase domains 5 // 2 F1 Fastkd5 0.037 -1.40 // 380601 /// NM_001146084 NM_026574 // Ino80 // INO80 homolog (S. cerevisiae) // Ino80 0.015 -1.40 2 E5 // 68142 /// ENSMUST NM_001113354 // Phf8 // PHD finger protein 8 // X F3 // Phf8 0.043 -1.41 320595 /// NM_177201 // NM_178732 // Zfp324 // zinc finger protein 324 // 7 A1 Zfp324 0.009 -1.41 // 243834 /// ENSMUST0000 NM_172794 // Zfp454 // zinc finger protein 454 // 11 Zfp454 0.007 -1.41 B1.3 // 237758 /// ENSMUST0 NM_019747 // Zfp113 // zinc finger protein 113 // 5 G1 Zfp113 0.025 -1.43 // 56314 /// ENSMUST00000 NM_007435 // Abcd1 // ATP-binding cassette, sub- Abcd1 0.021 -1.43 family D (ALD), member 1 // X B| NM_026962 // Kbtbd3 // kelch repeat and BTB (POZ) Kbtbd3 0.036 -1.43 domain containing 3 // 9 A1 // NM_183275 // 1110002N22Rik // RIKEN cDNA 1110002N22Rik 0.032 -1.45 1110002N22 gene // 11 B5|11 47.23 cM // NM_145627 // Rbm10 // RNA binding motif protein 10 Rbm10 0.033 -1.46

240

// X A1.3 // 236732 /// ENSMU NM_144853 // Cyyr1 // cysteine and tyrosine-rich Cyyr1 0.048 -1.47 protein 1 // 16 C3.3|16 56.2 cM NM_023348 // Snap29 // synaptosomal-associated Snap29 0.019 -1.48 protein 29 // 16 A3 // 67474 /// NM_019940 // Zfp111 // zinc finger protein 111 // 7 A3 Zfp111 0.014 -1.48 // 56707 /// NM_018791 // NM_198657 // EG381438 // predicted gene, EG381438 EG381438 0.026 -1.48 // 3 B|3 // 381438 /// ENSMUST NM_145618 // Narg2 // NMDA receptor-regulated gene Narg2 0.038 -1.49 2 // 9 C // 93697 /// ENSMUST NM_170588 // Cpne1 // copine I // 2 H2 // 266692 /// Cpne1 0.012 -1.49 NM_170590 // Cpne1 // copin NM_001007575 // Zfp58 // zinc finger protein 58 // 13 Zfp58 0.019 -1.50 B3 // 238693 /// ENSMUST00 NM_172656 // Stradb // STE20-related kinase adaptor Stradb 0.011 -1.51 beta // 1 C1.3|1 34.0 cM // NM_144820 // Ccdc28a // coiled-coil domain containing Ccdc28a 0.019 -1.53 28A // 10 A3 // 215814 /// NM_009779 // C3ar1 // complement component 3a C3ar1 0.040 -1.53 receptor 1 // 6 F1 // 12267 /// EN NM_175446 // Zmat1 // zinc finger, matrin type 1 // X E3 Zmat1 0.046 -1.54 // 215693 /// ENSMUST00 NM_176953 // Lig4 // ligase IV, DNA, ATP-dependent // Lig4 0.017 -1.55 8 A1.1 // 319583 /// ENSMU NM_175116 // P2ry5 // P2Y, G- P2ry5 0.024 -1.56 protein coupled, 5 // 14 D3 // NM_017476 // Akap8l // A kinase (PRKA) anchor protein Akap8l 0.016 -1.56 8-like // 17 B2 // 54194 / NM_144515 // Zfp52 // zinc finger protein 52 // 17 Zfp52 0.049 -1.57 A3.2|17 9.9 cM // 22710 /// E NM_021383 // Rqcd1 // rcd1 (required for cell Rqcd1 0.009 -1.58 differentiation) homolog 1 (S. pom NM_011657 // Tulp3 // tubby-like protein 3 // 6 F3|6 Tulp3 0.015 -1.58 62.5 cM // 22158 /// ENSMUS AK083919 // D130060J02Rik // RIKEN cDNA D130060J02 D130060J02Rik 0.020 -1.59 gene // 4 A1 // 320266 /// AK NM_001024539 // Shc2 // SHC (Src homology 2 domain Shc2 0.009 -1.59 containing) transforming prot NM_009442 // Ttf1 // transcription termination factor, Ttf1 0.013 -1.59 RNA polymerase I // 2 A3 BC107288 // Phxr4 // per-hexamer repeat gene 4 // 9 A1 Phxr4 0.034 -1.60 // 18689 /// X12806 // Ph NM_026734 // Tmem126b // transmembrane protein Tmem126b 0.012 -1.60 126B // 7 E1 // 68472 /// ENSMUST NM_001001447 // Zscan22 // zinc finger and SCAN Zscan22 0.014 -1.62 domain containing 22 // 7 A1 // AK019926 // 5330431K02Rik // RIKEN cDNA 5330431K02 5330431K02Rik 0.009 -1.67 gene // 13 D1 // 68189 NM_183174 // Homez // homeodomain leucine zipper- Homez 0.040 -1.68 encoding gene // 14 C2-C3 // 23 NM_178375 // Zswim3 // zinc finger, SWIM domain Zswim3 0.017 -1.69 containing 3 // 2 H3 // 67538 // BC049349 // BC049349 // cDNA sequence BC049349 // 8 BC049349 0.010 -1.73 B3.3 // 234413 /// BC065788

241

NM_021272 // Fabp7 // fatty acid binding protein 7, Fabp7 0.032 -1.77 brain // 10 B4 // 12140 /// NM_025280 // Kin // antigenic determinant of rec-A Kin 0.020 -1.78 protein // 2 A1|2 A1-A3 // 16 NM_009620 // Adam4 // a disintegrin and Adam4 0.016 -1.79 metallopeptidase domain 4 // 12 D1 // 11 ENSMUST00000056274 // C230013L11Rik // RIKEN cDNA C230013L11Rik 0.031 -1.79 C230013L11 gene // 17 A3.3 // NM_176987 // 4732471D19Rik // RIKEN cDNA 4732471D19Rik 0.015 -1.83 4732471D19 gene // 13 B1 // 319719 /// --- 0.029 -1.91 NM_011812 // Fbln5 // fibulin 5 // 12 F1 // 23876 /// Fbln5 0.047 -1.97 ENSMUST00000021603 // Fbln NM_020488 // Gabrq // gamma-aminobutyric acid Gabrq 0.007 -2.22 (GABA) A receptor, subunit theta / NM_001079695 // Sfrs5 // splicing factor, Sfrs5 0.009 -2.26 arginine/serine-rich 5 (SRp40, HRS) // AK032243 // 9530019H20Rik // RIKEN cDNA 9530019H20 9530019H20Rik 0.020 -2.40 gene // 8 B3.3 // 320999 NM_013556 // Hprt1 // hypoxanthine guanine Hprt1 0.000 -33.66 phosphoribosyl transferase 1 // X A6|

242

APPENDIX C

Significantly altered genes in Gpr50-/- versus WT mice when fed ad libitum.

Affymetrix mouse arrays were used to screen the differentially expressed genes in the hypothalamus between fed and fasted WT and Gpr50-/- mice. Significantly altered genes with a q<0.05 and fold change>1 only are included.

Fold Gene Assignment Gene Symbol q value Change NM_013556 // Hprt1 // hypoxanthine guanine phosphoribosyl transferase 1 // X A6| Hprt1 0.000 -30.84 NM_020488 // Gabrq // gamma-aminobutyric acid (GABA) A receptor, subunit theta / Gabrq 0.029 -2.24

243

APPENDIX D

Significantly altered genes in WT mice when fasted versus fed ad libitum.

Affymetrix mouse arrays were used to screen differentially expressed genes in the hypothalamus in fed and fasted WT and Gpr50-/- mice. Significantly altered genes with a q<0.05 and fold change>1 only are included.

Fold Gene Assignment Gene Symbol q value Change NM_001009935 // Txnip // thioredoxin interacting protein // 3 47.1 cM // 56338 / Txnip 0.002 2.39 NM_015786 // Hist1h1c // histone cluster 1, H1c // 13 A2-A3 // 50708 /// ENSMUST Hist1h1c 0.017 1.49 NM_001039365 // Mobp // myelin-associated oligodendrocytic basic protein // 9 F4 Mobp 0.046 1.43 NM_177449 // Lrrc29 // leucine rich repeat containing 29 // 8 D3 // 234684 /// E Lrrc29 0.045 1.37 NM_011404 // Slc7a5 // solute carrier family 7 (cationic amino acid transporter, Slc7a5 0.019 1.31 AK086466 // D930030O05Rik // RIKEN cDNA D930030O05 gene // 7 C // 320387 D930030O05Rik 0.044 1.24 NM_011657 // Tulp3 // tubby-like protein 3 // 6 F3|6 62.5 cM // 22158 /// ENSMUS Tulp3 0.012 1.23 NM_172694 // Megf9 // multiple EGF-like-domains 9 // 4 C2 // 230316 /// ENSMUST0 Megf9 0.018 1.23 NM_030697 // Kank3 // KN motif and ankyrin repeat domains 3 // 17 B1 // 80880 // Kank3 0.042 1.20 NM_133898 // N4bp2l1 // NEDD4 binding protein 2-like 1 // 5 G3 // 100637 /// ENS N4bp2l1 0.019 1.20 NM_144731 // Galnt7 // UDP-N-acetyl-alpha-D- galactosamine: polypeptide N-acetylg Galnt7 0.049 1.20 NM_146039 // Wdr60 // WD repeat domain 60 // 12 F2 // 217935 /// ENSMUST00000039 Wdr60 0.016 1.20 NM_176987 // 4732471D19Rik // RIKEN cDNA 4732471D19 gene // 13 B1 // 319719 /// 4732471D19Rik 0.002 1.19 NM_175247 // Zfp28 // zinc finger protein 28 // 7 A1 // 22690 /// ENSMUST0000008 Zfp28 0.011 1.19 NM_001113354 // Phf8 // PHD finger protein 8 // X F3 // 320595 /// NM_177201 // Phf8 0.048 1.18 NM_019964 // Dnajb8 // DnaJ (Hsp40) homolog, subfamily B, member 8 // 6 D2|6 38. Dnajb8 0.037 1.17 NM_023538 // Agk // acylglycerol kinase // 6 B1 // 69923 /// ENSMUST00000031977 Agk 0.019 1.16 NM_153599 // Cdk8 // cyclin-dependent kinase 8 // 5 G3 // 264064 /// ENSMUST0000 Cdk8 0.043 1.16 NM_021383 // Rqcd1 // rcd1 (required for cell differentiation) homolog 1 (S. pom Rqcd1 0.002 1.15 NM_008825 // Pfkfb2 // 6-phosphofructo-2- kinase/fructose-2,6-biphosphatase 2 // Pfkfb2 0.014 1.15 NM_199157 // Ifnk // interferon kappa // 4 A5 // 387510 /// ENSMUST00000058595 / Ifnk 0.030 1.13 NM_183174 // Homez // homeodomain leucine zipper- encoding gene // 14 C2-C3 // 23 Homez 0.039 1.12 NM_011057 // Pdgfb // platelet derived growth factor, B polypeptide // 15 E|15 4 Pdgfb 0.040 1.12

244

NM_172557 // Rufy1 // RUN and FYVE domain containing 1 // 11 B1.3 // 216724 /// Rufy1 0.036 1.12 NM_001037707 // Zfp27 // zinc finger protein 27 // 7 B1 // 22689 /// NM_011754 / Zfp27 0.001 1.12 NM_026734 // Tmem126b // transmembrane protein 126B // 7 E1 // 68472 /// ENSMUST Tmem126b 0.002 1.12 NM_144851 // Senp1 // SUMO1/sentrin specific peptidase 1 // 15 F1|15 55.8 cM // Senp1 0.032 1.12 NM_172992 // Phtf2 // putative homeodomain transcription factor 2 // 5 A3 // 687 Phtf2 0.022 1.12 NM_153176 // Spg7 // spastic paraplegia 7 homolog (human) // 8 E1 // 234847 /// Spg7 0.019 1.11 BC059896 // 2410002F23Rik // RIKEN cDNA 2410002F23 gene // 7 B3 // 66976 /// BC0 2410002F23Rik 0.044 1.11 NM_175548 // Lsamp // limbic system-associated membrane protein // 16 B5 // 2688 Lsamp 0.000 1.11 NM_007397 // Acvr2b // activin receptor IIB // 9 F3 // 11481 /// ENSMUST00000035 Acvr2b 0.046 1.11 NM_026574 // Ino80 // INO80 homolog (S. cerevisiae) // 2 E5 // 68142 /// ENSMUST Ino80 0.008 1.10 NM_146185 // Zfp790 // zinc finger protein 790 // 7 B1 // 233056 /// NM_00114588 Zfp790 0.024 1.10 NM_134011 // Tbrg4 // transforming growth factor beta regulated gene 4 // 11 A1 Tbrg4 0.008 1.09 NM_198657 // EG381438 // predicted gene, EG381438 // 3 B|3 // 381438 /// ENSMUST EG381438 0.012 1.09 NM_009525 // Wnt5b // wingless-related MMTV integration site 5B // 6 F1|6 56.2 c Wnt5b 0.031 1.09 NM_001007575 // Zfp58 // zinc finger protein 58 // 13 B3 // 238693 /// ENSMUST00 Zfp58 0.012 1.09 NM_009647 // Ak3l1 // adenylate kinase 3-like 1 // 4 C6|4 47.6 cM // 11639 /// D Ak3l1 0.035 1.08 NM_001114098 // 1110012J17Rik // RIKEN cDNA 1110012J17 gene // 17 E1.1 // 68617 1110012J17Rik 0.018 1.08 NM_133756 // Gpn1 // GPN-loop GTPase 1 // 5 B1 // 74254 /// ENSMUST00000076949 / Gpn1 0.022 1.08 NM_176840 // Osbpl11 // oxysterol binding protein-like 11 // 16 B3 // 106326 /// Osbpl11 0.010 1.07 NM_001083897 // Mpzl1 // myelin protein zero-like 1 // 1 H2.3 // 68481 /// NM_00 Mpzl1 0.036 1.07 NM_025794 // Etfdh // electron transferring flavoprotein, dehydrogenase // 3 E3| Etfdh 0.027 1.07 NM_027453 // Btf3l4 // basic transcription factor 3-like 4 // 4 C7 // 70533 /// Btf3l4 0.019 1.07 NM_029581 // Mtif3 // mitochondrial translational initiation factor 3 // 5 G3 // Mtif3 0.008 1.07 NM_145140 // Abcc10 // ATP-binding cassette, sub- family C (CFTR/MRP), member 10 Abcc10 0.029 1.07 NM_153287 // Csrnp1 // cysteine-serine-rich nuclear protein 1 // 9 F4 // 215418 Csrnp1 0.003 1.07 NM_172743 // Plekha7 // pleckstrin homology domain containing, family A member 7 Plekha7 0.032 1.07 NM_008654 // Myd116 // myeloid differentiation primary response gene 116 // 7 B4 Myd116 0.000 1.06 NM_001005247 // Hps5 // Hermansky-Pudlak syndrome 5 homolog (human) // 7 B4|7 25 Hps5 0.015 1.06 NM_177755 // Klhl38 // kelch-like 38 (Drosophila) // 15 D1 // 268807 /// ENSMUST Klhl38 0.039 1.06 NM_144908 // Galnt11 // UDP-N-acetyl-alpha-D- Galnt11 0.019 1.06 245 galactosamine:polypeptide N-acetylg NM_144820 // Ccdc28a // coiled-coil domain containing 28A // 10 A3 // 215814 /// Ccdc28a 0.008 1.05 NM_145591 // BC003267 // cDNA sequence BC003267 // 8 A1.1 // 233987 /// ENSMUST0 BC003267 0.009 1.05 NM_178375 // Zswim3 // zinc finger, SWIM domain containing 3 // 2 H3 // 67538 // Zswim3 0.012 1.05 NM_009442 // Ttf1 // transcription termination factor, RNA polymerase I // 2 A3 Ttf1 0.008 1.05 NM_172768 // Gramd1b // GRAM domain containing 1B // 9 B // 235283 /// ENSMUST00 Gramd1b 0.009 1.04 NM_027448 // Lca5 // Leber congenital amaurosis 5 (human) // 9 E2 // 75782 /// N Lca5 0.027 1.04 NM_029291 // Ascc2 // activating signal cointegrator 1 complex subunit 2 // 11 A Ascc2 0.013 1.04 NM_025617 // 2210012G02Rik // RIKEN cDNA 2210012G02 gene // 4 C7 // 66526 /// EN 2210012G02Rik 0.031 1.03 NM_023162 // Znrd1 // zinc ribbon domain containing, 1 // 17 B3 // 66136 /// ENS Znrd1 0.027 1.03 NM_025864 // Tmem206 // transmembrane protein 206 // 1 H6 // 66950 /// ENSMUST00 Tmem206 0.038 1.03 NM_177089 // Tacc1 // transforming, acidic coiled-coil containing protein 1 // 8 Tacc1 0.044 1.03 NM_012031 // Spag1 // sperm associated antigen 1 // 15 C // 26942 /// ENSMUST000 Spag1 0.030 1.03 BC008558 // 1700030K09Rik // RIKEN cDNA 1700030K09 gene // 8 B3.3 // 72254 /// E 1700030K09Rik 0.000 1.03 NM_007455 // Ap1g2 // adaptor protein complex AP-1, gamma 2 subunit // 14 C3 // Ap1g2 0.017 1.03 NM_025280 // Kin // antigenic determinant of rec-A protein // 2 A1|2 A1-A3 // 16 Kin 0.029 1.03 NM_019940 // Zfp111 // zinc finger protein 111 // 7 A3 // 56707 /// NM_018791 // Zfp111 0.037 1.02 NM_053247 // Lyve1 // lymphatic vessel endothelial hyaluronan receptor 1 // 7 F2 Lyve1 0.025 1.02 NM_010509 // Ifnar2 // interferon (alpha and beta) receptor 2 // 16 C3.3|16 63.1 Ifnar2 0.029 1.02 NM_001024539 // Shc2 // SHC (Src homology 2 domain containing) transforming prot Shc2 0.013 1.02 NM_009102 // Rrh // retinal pigment epithelium derived rhodopsin homolog // 3 G3 Rrh 0.001 1.02 NM_008913 // Ppp3ca // protein phosphatase 3, catalytic subunit, alpha isoform / Ppp3ca 0.015 1.02 NM_172794 // Zfp454 // zinc finger protein 454 // 11 B1.3 // 237758 /// ENSMUST0 Zfp454 0.000 1.01 NM_009648 // Akap1 // A kinase (PRKA) anchor protein 1 // 11 C // 11640 /// NM_0 Akap1 0.014 1.01 NM_011145 // Ppard // peroxisome proliferator activator receptor delta // 17 A3. Ppard 0.042 1.01 NM_017476 // Akap8l // A kinase (PRKA) anchor protein 8-like // 17 B2 // 54194 / Akap8l 0.017 1.01 NM_013712 // Itgb1bp2 // integrin beta 1 binding protein 2 // X D // 26549 /// N Itgb1bp2 0.021 1.01 NM_001001447 // Zscan22 // zinc finger and SCAN domain containing 22 // 7 A1 // Zscan22 0.008 1.01 NM_010247 // Xrcc6 // X-ray repair complementing defective repair in Chinese ham Xrcc6 0.039 1.01 NM_178732 // Zfp324 // zinc finger protein 324 // 7 A1 // 243834 /// ENSMUST0000 Zfp324 0.002 -1.01 246

NM_175490 // Gpr75 // G protein-coupled receptor 75 // 11 A4 // 237716 /// ENSMU Gpr75 0.027 -1.01 NM_183000 // Accn3 // amiloride-sensitive cation channel 3 // 5 A3 // 171209 /// Accn3 0.000 -1.01 NM_008211 // H3f3b // H3 histone, family 3B // 11 E2 // 15081 /// BC037730 // H3 H3f3b 0.030 -1.02 NM_027226 // Fyttd1 // forty-two-three domain containing 1 // 16 B3 // 69823 /// Fyttd1 0.017 -1.02 NM_009703 // Araf // v-raf murine sarcoma 3611 viral oncogene homolog // X A2-A3 Araf 0.003 -1.02 NM_145618 // Narg2 // NMDA receptor-regulated gene 2 // 9 C // 93697 /// ENSMUST Narg2 0.027 -1.03 NM_023799 // Mgea5 // meningioma expressed antigen 5 (hyaluronidase) // 19 D1|19 Mgea5 0.042 -1.03 NM_009828 // Ccna2 // cyclin A2 // 3 B|3 19.2 cM // 12428 /// ENSMUST00000029270 Ccna2 0.013 -1.03 NM_016669 // Crym // crystallin, mu // 7 F2|7 55.0 cM // 12971 /// ENSMUST000000 Crym 0.037 -1.03 NM_170588 // Cpne1 // copine I // 2 H2 // 266692 /// NM_170590 // Cpne1 // copin Cpne1 0.031 -1.03 NM_175258 // Rapgef6 // Rap guanine nucleotide exchange factor (GEF) 6 // 11 B1. Rapgef6 0.016 -1.03 NM_030174 // Mctp1 // multiple C2 domains, transmembrane 1 // 13 C1 // 78771 /// Mctp1 0.029 -1.03 NM_053081 // Fancg // Fanconi anemia, complementation group G // 4 B1 // 60534 / Fancg 0.000 -1.04 NM_021450 // Trpm7 // transient receptor potential cation channel, subfamily M, Trpm7 0.042 -1.04 NM_027427 // Taf15 // TAF15 RNA polymerase II, TATA box binding protein (TBP)-as Taf15 0.038 -1.04 AK019926 // 5330431K02Rik // RIKEN cDNA 5330431K02 gene // 13 D1 // 68189 5330431K02Rik 0.008 -1.04 NM_026812 // Hddc3 // HD domain containing 3 // 7 D3 // 68695 /// ENSMUST0000003 Hddc3 0.030 -1.04 NM_008235 // Hes1 // hairy and enhancer of split 1 (Drosophila) // 16 B2|16 27.0 Hes1 0.040 -1.04 NM_134114 // Sft2d1 // SFT2 domain containing 1 // 17 A1 // 106489 /// BC091770 Sft2d1 0.027 -1.04 NM_020033 // Ankrd2 // ankyrin repeat domain 2 (stretch responsive muscle) // 19 Ankrd2 0.042 -1.05 NM_023716 // Tubb2b // tubulin, beta 2B // 13 A4 // 73710 /// NM_009450 // Tubb2 Tubb2b 0.025 -1.05 NM_178691 // Yod1 // YOD1 OTU deubiquitinating enzyme 1 homologue (S. cerevisiae Yod1 0.014 -1.05 NM_175482 // Usp28 // ubiquitin specific peptidase 28 // 9 A5.3 // 235323 /// EN Usp28 0.046 -1.06 NM_008992 // Abcd4 // ATP-binding cassette, sub- family D (ALD), member 4 // 12 D Abcd4 0.000 -1.06 BC017643 // BC017643 // cDNA sequence BC017643 // 11 E2 // 217370 /// BC026616 / BC017643 0.039 -1.06 NM_153802 // Zfp128 // zinc finger protein 128 // 7 A1 // 243833 /// ENSMUST0000 Zfp128 0.019 -1.06 NM_172581 // Fam161b // family with sequence similarity 161, member B // 12 D1 / Fam161b 0.009 -1.06 ENSMUST00000056274 // C230013L11Rik // RIKEN cDNA C230013L11 gene // 17 A3.3 // C230013L11Rik 0.031 -1.07 NM_178782 // Bcorl1 // BCL6 co-repressor-like 1 // X A4 // 320376 /// ENSMUST000 Bcorl1 0.008 -1.08 NM_008942 // Npepps // aminopeptidase puromycin Npepps 0.027 -1.08 247 sensitive // 11 D|11 56.0 cM // NM_026072 // Sdccag10 // serologically defined colon cancer antigen 10 // 13 D1 Sdccag10 0.004 -1.09 NM_145985 // Arcn1 // archain 1 // 9 A5.2 // 213827 /// ENSMUST00000034607 // Ar Arcn1 0.025 -1.09 NM_011945 // Map3k1 // mitogen-activated protein kinase kinase kinase 1 // 13 D2 Map3k1 0.017 -1.09 NM_199080 // Ddx17 // DEAD (Asp-Glu-Ala-Asp) box polypeptide 17 // 15 E1 // 6704 Ddx17 0.017 -1.10 BC049349 // BC049349 // cDNA sequence BC049349 // 8 B3.3 // 234413 /// BC065788 BC049349 0.027 -1.10 NM_053109 // Clec2d // C-type lectin domain family 2, member d // 6 F3 // 93694 Clec2d 0.022 -1.11 NM_144853 // Cyyr1 // cysteine and tyrosine-rich protein 1 // 16 C3.3|16 56.2 cM Cyyr1 0.042 -1.11 NM_001142918 // Tcf7l2 // transcription factor 7-like 2, T-cell specific, HMG-bo Tcf7l2 0.022 -1.11 NM_175334 // Maml1 // mastermind like 1 (Drosophila) // 11 B1.3 // 103806 /// EN Maml1 0.035 -1.12 AK032243 // 9530019H20Rik // RIKEN cDNA 9530019H20 gene // 8 B3.3 // 320999 9530019H20Rik 0.019 -1.13 NM_016767 // Batf // basic leucine zipper transcription factor, ATF-like // 12 D Batf 0.036 -1.14 NM_001131020 // Gfap // glial fibrillary acidic protein // 11 D|11 62.0 cM // 14 Gfap 0.038 -1.14 NM_144892 // Ncoa5 // nuclear receptor coactivator 5 // 2 H3 // 228869 /// ENSMU Ncoa5 0.006 -1.14 NM_023651 // Pex13 // peroxisomal biogenesis factor 13 // 11 A3.3 // 72129 /// E Pex13 0.009 -1.14 AK049490 // C230029D21Rik // RIKEN cDNA C230029D21 gene // 17 A1 // 320255 /// A C230029D21Rik 0.006 -1.14 NM_028099 // Dusp11 // dual specificity phosphatase 11 (RNA/RNP complex 1-intera Dusp11 0.002 -1.15 NM_023493 // Cml5 // camello-like 5 // 6 C3 // 69049 /// NM_053097 // Cml3 // ca Cml5 0.019 -1.15 NM_021343 // Spata5 // spermatogenesis associated 5 // 3 B // 57815 /// ENSMUST0 Spata5 0.012 -1.15 NM_001029842 // Slc16a6 // solute carrier family 16 (monocarboxylic acid transpo Slc16a6 0.015 -1.15 NM_145556 // Tardbp // TAR DNA binding protein // 4 E2 // 230908 /// NM_00100854 Tardbp 0.008 -1.15 NM_001081430 // Nat12 // N-acetyltransferase 12 // 14 C1 // 70646 /// ENSMUST000 Nat12 0.005 -1.15 NM_133931 // Pot1a // protection of telomeres 1A // 6 A3.1 // 101185 /// ENSMUST Pot1a 0.043 -1.16 NM_008078 // Gad2 // glutamic acid decarboxylase 2 // 2 A3|2 9.0 cM // 14417 /// Gad2 0.015 -1.16 NM_019517 // Bace2 // beta-site APP-cleaving enzyme 2 // 16 C4 // 56175 /// ENSM Bace2 0.014 -1.16 NM_010426 // Foxf1a // forkhead box F1a // 8 67.0 cM // 15227 /// ENSMUST0000005 Foxf1a 0.002 -1.17 NM_029239 // Prkd3 // protein kinase D3 // 17 E3 // 75292 /// ENSMUST00000003191 Prkd3 0.034 -1.18 NM_178194 // Hist1h2be // histone cluster 1, H2be // 13 A2-A3 // 319179 /// NM_0 Hist1h2be 0.000 -1.18 NM_027304 // H1fnt // H1 histone family, member N, testis-specific // 15 F1 // 7 H1fnt 0.014 -1.21 NM_028287 // Zufsp // zinc finger with UFM1-specific peptidase domain // 10 B1 / Zufsp 0.016 -1.22 248

NM_028320 // Adipor1 // adiponectin receptor 1 // 1 E4 // 72674 /// ENSMUST00000 Adipor1 0.003 -1.22 NM_001079695 // Sfrs5 // splicing factor, arginine/serine-rich 5 (SRp40, HRS) // Sfrs5 0.003 -1.24 NM_001009819 // A3galt2 // alpha 1,3- galactosyltransferase 2 (isoglobotriaosylce A3galt2 0.048 -1.24 NM_028667 // D3Ertd751e // DNA segment, Chr 3, ERATO Doi 751, expressed // 3 C / D3Ertd751e 0.018 -1.37

249

APPENDIX E

Significantly altered genes in Gpr50-/- mice when fasted versus fed ad libitum.

Affymetrix mouse arrays were used to screen differentially expressed genes in the hypothalamus in fed and fasted WT and Gpr50-/- mice. Significantly altered genes with a q<0.05 and fold change>1 only are included.

Gene Assignment Gene Symbol q value Fold Change NM_007669 // Cdkn1a // cyclin-dependent kinase Cdkn1a 0.000 8.32 inhibitor 1A (P21) // 17 A3.3|17 NM_001009935 // Txnip // thioredoxin interacting protein Txnip 0.000 7.67 // 3 47.1 cM // 56338 / NM_020568 // S3-12 // plasma membrane associated S3-12 0.001 4.43 protein, S3-12 // 17 D // 57435 NM_015786 // Hist1h1c // histone cluster 1, H1c // 13 A2- Hist1h1c 0.001 3.28 A3 // 50708 /// ENSMUST ENSMUST00000079045 // B230325K18Rik // RIKEN cDNA B230325K18Rik 0.003 2.79 B230325K18 gene // 7 F3 // 319 NM_029662 // Mfsd2 // major facilitator superfamily Mfsd2 0.000 2.78 domain containing 2 // 4 D1 NM_029083 // Ddit4 // DNA-damage-inducible transcript 4 Ddit4 0.007 2.76 // 10 B3 // 74747 /// EN NM_010217 // Ctgf // connective tissue growth factor // Ctgf 0.020 2.74 10 A3-B1|10 17.0 cM // 1 NM_010220 // Fkbp5 // FK506 binding protein 5 // 17 Fkbp5 0.004 2.60 A3.3|17 13.0 cM // 14229 /// NM_025802 // Pnpla2 // patatin-like phospholipase Pnpla2 0.001 2.48 domain containing 2 // 7 F5 // NM_011361 // Sgk1 // serum/glucocorticoid regulated Sgk1 0.001 2.44 kinase 1 // 10 A3 // 20393 / NM_133670 // Sult1a1 // sulfotransferase family 1A, Sult1a1 0.007 2.44 phenol-preferring, member 1 NM_175232 // 5830427D03Rik // RIKEN cDNA 5830427D03Rik 0.002 2.43 5830427D03 gene // 15 F1 // 76061 /// B NM_019564 // Htra1 // HtrA serine peptidase 1 // 7 F3 // Htra1 0.003 2.27 56213 /// ENSMUST000000 NM_010907 // Nfkbia // nuclear factor of kappa light Nfkbia 0.012 2.24 polypeptide gene enhancer i NM_001159367 // Per1 // period homolog 1 (Drosophila) Per1 0.000 2.20 // 11 B // 18626 /// NM_01 NM_013642 // Dusp1 // dual specificity phosphatase 1 // Dusp1 0.003 2.10 17 A2-C|17 13.0 cM // 19 NM_013602 // Mt1 // metallothionein 1 // 8 C5|8 45.0 cM Mt1 0.010 2.10 // 17748 /// ENSMUST0000 NM_172267 // Phyhd1 // phytanoyl-CoA dioxygenase Phyhd1 0.001 2.09 domain containing 1 // 2 B // 2 NM_010426 // Foxf1a // forkhead box F1a // 8 67.0 cM // Foxf1a 0.003 2.08 15227 /// ENSMUST0000005 NM_027453 // Btf3l4 // basic transcription factor 3-like 4 Btf3l4 0.003 2.08 // 4 C7 // 70533 /// NM_010271 // Gpd1 // glycerol-3-phosphate Gpd1 0.022 2.07

250 dehydrogenase 1 (soluble) // 15 56.8 c NM_011723 // Xdh // xanthine dehydrogenase // 17 E2|17 Xdh 0.008 2.03 45.3 cM // 22436 /// ENSM NM_019671 // Net1 // neuroepithelial cell transforming Net1 0.003 1.93 gene 1 // 13 A1 // 56349 NM_139306 // Acer2 // alkaline ceramidase 2 // 4 C4 // Acer2 0.001 1.92 230379 /// ENSMUST0000004 NM_016868 // Hif3a // hypoxia inducible factor 3, alpha Hif3a 0.010 1.91 subunit // 7 A2 // 53417 NM_008587 // Mertk // c-mer proto-oncogene tyrosine Mertk 0.001 1.90 kinase // 2 F1 // 17289 /// NM_011172 // Prodh // proline dehydrogenase // 16 Prodh 0.001 1.90 A3|16 10.73 cM // 19125 /// EN NM_021877 // Hr // hairless // 14 D2 // 15460 /// Hr 0.006 1.88 ENSMUST00000022691 // Hr // ha NM_025279 // Hnrnpk // heterogeneous nuclear Hnrnpk 0.015 1.87 ribonucleoprotein K // 13 B1 // 153 NM_010831 // Sik1 // salt inducible kinase 1 // 17 B1|17 Sik1 0.004 1.86 18.18 cM // 17691 /// E NM_008076 // Gabrr2 // gamma-aminobutyric acid Gabrr2 0.002 1.85 (GABA) C receptor, subunit rho 2 NM_172791 // Pla2g3 // phospholipase A2, group III // 11 Pla2g3 0.005 1.85 A1 // 237625 /// ENSMUS NM_029035 // Spsb1 // splA/ryanodine receptor domain Spsb1 0.020 1.84 and SOCS box containing 1 / NM_009402 // Pglyrp1 // peptidoglycan recognition Pglyrp1 0.010 1.83 protein 1 // 7 A3 // 21946 /// AK033936 // 9330119M13Rik // RIKEN cDNA 9330119M13 9330119M13Rik 0.013 1.82 gene // X E3 // 319963 /// AK NM_025404 // Arl4d // ADP-ribosylation factor-like 4D // Arl4d 0.029 1.81 11 D // 80981 /// BC016 NM_008872 // Plat // plasminogen activator, tissue // 8 Plat 0.004 1.79 A2|8 9.0 cM // 18791 /// NM_172152 // Slc24a4 // solute carrier family 24 Slc24a4 0.010 1.79 (sodium/potassium/calcium excha NM_146787 // Olfr920 // 920 // --- // Olfr920 0.014 1.79 258783 /// ENSMUST00000 NM_153287 // Csrnp1 // cysteine-serine-rich nuclear Csrnp1 0.001 1.78 protein 1 // 9 F4 // 215418 NM_146042 // Rnf144b // ring finger protein 144B // 13 Rnf144b 0.035 1.78 A5 // 218215 /// ENSMUST0 NM_134133 // 2010002N04Rik // RIKEN cDNA 2010002N04Rik 0.043 1.76 2010002N04 gene // 18 D3 // 106878 /// NM_016693 // Map3k6 // mitogen-activated protein Map3k6 0.001 1.76 kinase kinase kinase 6 // 4 D2. NM_010171 // F3 // coagulation factor III // 3 G1|3 50.0 F3 0.001 1.74 cM // 14066 /// ENSMUST NM_001131020 // Gfap // glial fibrillary acidic protein // Gfap 0.026 1.73 11 D|11 62.0 cM // 14 NM_007514 // Slc7a2 // solute carrier family 7 (cationic Slc7a2 0.037 1.73 amino acid transporter, NM_008055 // Fzd4 // frizzled homolog 4 (Drosophila) // 7 Fzd4 0.001 1.73 E1|7 44.5 cM // 14366 NM_001110197 // Rnf146 // ring finger protein 146 // 10 Rnf146 0.005 1.72 A4 // 68031 /// NM_00111

251

NM_019588 // Plce1 // phospholipase C, epsilon 1 // 19 D1 Plce1 0.001 1.71 // 74055 /// ENSMUST00 NM_027560 // Arrdc2 // arrestin domain containing 2 // 8 Arrdc2 0.010 1.70 C1 // 70807 /// ENSMUST NM_011756 // Zfp36 // zinc finger protein 36 // 7 A3|7 Zfp36 0.005 1.70 10.2 cM // 22695 /// ENSM NM_026718 // Ankrd13a // ankyrin repeat domain 13a // Ankrd13a 0.000 1.70 5 F // 68420 /// ENSMUST00 NM_009743 // Bcl2l1 // BCL2-like 1 // 2 H1|2 87.5 cM // Bcl2l1 0.004 1.70 12048 /// ENSMUST0000010 NM_175286 // C430004E15Rik // RIKEN cDNA C430004E15 C430004E15Rik 0.000 1.69 gene // 2 A3 // 97031 /// EN NM_008654 // Myd116 // myeloid differentiation primary Myd116 0.000 1.68 response gene 116 // 7 B4 NM_001077364 // Tsc22d3 // TSC22 domain family, Tsc22d3 0.004 1.68 member 3 // X F1 // 14605 /// NM NM_008010 // Fgfr3 // fibroblast growth factor receptor 3 Fgfr3 0.000 1.68 // 5 B|5 20.0 cM // 14 NM_029620 // Pcolce2 // procollagen C-endopeptidase Pcolce2 0.019 1.68 enhancer 2 // 9 E3.3 // 7647 NM_011521 // Sdc4 // syndecan 4 // 2 H3|2 94.0 cM // Sdc4 0.001 1.66 20971 /// ENSMUST0000001715 NM_026797 // Dbndd2 // dysbindin (dystrobrevin binding Dbndd2 0.012 1.65 protein 1) domain contain NM_030257 // Lysmd3 // LysM, putative peptidoglycan- Lysmd3 0.016 1.65 binding, domain containing 3 NM_011404 // Slc7a5 // solute carrier family 7 (cationic Slc7a5 0.000 1.64 amino acid transporter, NM_010104 // Edn1 // endothelin 1 // 13 A4|13 26.0 cM Edn1 0.034 1.64 // 13614 /// ENSMUST000000 NM_027907 // Agxt2l1 // alanine-glyoxylate Agxt2l1 0.002 1.63 aminotransferase 2-like 1 // 3 G3 // NM_010050 // Dio2 // deiodinase, iodothyronine, type II Dio2 0.037 1.63 // 12 D3 // 13371 /// AF NM_011400 // Slc2a1 // solute carrier family 2 (facilitated Slc2a1 0.002 1.63 glucose transporter) NM_194064 // Nanos2 // nanos homolog 2 (Drosophila) // Nanos2 0.040 1.62 7 A3 // 378430 /// ENSMUS NM_001081430 // Nat12 // N-acetyltransferase 12 // 14 Nat12 0.009 1.61 C1 // 70646 /// ENSMUST000 BC068129 // Fam83d // family with sequence similarity 83, Fam83d 0.001 1.61 member D // 2 H1 // 71 NM_145570 // Fam176a // family with sequence similarity Fam176a 0.018 1.61 176, member A // 6 C3 // NM_177420 // Psat1 // phosphoserine aminotransferase 1 Psat1 0.035 1.60 // 19 A|19 32.5 cM // 107 NM_181547 // Nostrin // nitric oxide synthase trafficker // Nostrin 0.004 1.59 2 C2 // 329416 /// E NM_145478 // Pim3 // proviral integration site 3 // 15 E3 Pim3 0.034 1.59 // 223775 /// ENSMUST0 NM_001010937 // Gjb6 // gap junction protein, beta 6 // Gjb6 0.002 1.59 14 C3|14 22.5 cM // 1462 NM_133237 // Apcdd1 // adenomatosis polyposis coli Apcdd1 0.041 1.59 down-regulated 1 // 18 E1 // NM_026414 // Asprv1 // aspartic peptidase, retroviral-like Asprv1 0.043 1.59

252

1 // 6 D1 // 67855 // NM_001142804 // Acss3 // acyl-CoA synthetase short- Acss3 0.010 1.58 chain family member 3 // 10 D NM_027687 // Cabyr // calcium-binding tyrosine-(Y)- Cabyr 0.005 1.58 phosphorylation regulated (fi NM_008494 // Lfng // LFNG O-fucosylpeptide 3-beta-N- Lfng 0.005 1.57 acetylglucosaminyltransferas NM_206822 // Olfr10 // olfactory receptor 10 // 11 B1.2 // Olfr10 0.035 1.56 18307 /// ENSMUST0000 NM_146251 // Pnpla7 // patatin-like phospholipase Pnpla7 0.001 1.56 domain containing 7 // 2 A3 // NM_001042591 // Arrdc3 // arrestin domain containing 3 Arrdc3 0.008 1.55 // 13 C3 // 105171 /// EN BC151170 // 3222402P14Rik // RIKEN cDNA 3222402P14 3222402P14Rik 0.037 1.55 gene // 9 E4 // 235542 /// BC NM_153782 // Fam20a // family with sequence similarity Fam20a 0.013 1.55 20, member A // 11 E1 // NM_175548 // Lsamp // limbic system-associated Lsamp 0.000 1.55 membrane protein // 16 B5 // 2688 NM_133955 // Rhou // ras homolog gene family, member Rhou 0.038 1.55 U // 8 E2 // 69581 /// ENSM NM_019518 // Grasp // GRP1 (general receptor for Grasp 0.015 1.54 phosphoinositides 1)-associated NM_010638 // Klf9 // Kruppel-like factor 9 // 19 C1 // Klf9 0.015 1.54 16601 /// BC138724 // Klf NM_020518 // Vsig2 // V-set and immunoglobulin domain Vsig2 0.007 1.53 containing 2 // 9 B // 572 NM_027782 // Kctd6 // potassium channel tetramerisation Kctd6 0.045 1.51 domain containing 6 // 1 NM_026514 // Cdc42ep3 // CDC42 effector protein (Rho Cdc42ep3 0.036 1.51 GTPase binding) 3 // 17 E3 NM_025543 // Mcts2 // malignant T cell amplified Mcts2 0.031 1.50 sequence 2 // 2 H1 // 66405 /// NM_133753 // Errfi1 // ERBB receptor feedback inhibitor 1 Errfi1 0.023 1.49 // 4 E1 // 74155 /// E NM_130452 // Bbox1 // butyrobetaine (gamma), 2- Bbox1 0.038 1.49 oxoglutarate dioxygenase 1 (gamma NM_008211 // H3f3b // H3 histone, family 3B // 11 E2 // H3f3b 0.010 1.49 15081 /// BC037730 // H3 NM_021897 // Trp53inp1 // transformation related Trp53inp1 0.011 1.49 protein 53 inducible nuclear pr NM_009186 // Tra2b // transformer 2 beta homolog Tra2b 0.022 1.48 (Drosophila) // 16 B1 // 20462 NM_018881 // Fmo2 // flavin containing monooxygenase Fmo2 0.007 1.48 2 // 1 H1 // 55990 /// ENSM NM_010222 // Fkbp7 // FK506 binding protein 7 // 2 C3|2 Fkbp7 0.047 1.48 60.5 cM // 14231 /// ENS NM_028390 // Anln // anillin, actin binding protein // 9 A4 Anln 0.002 1.47 // 68743 /// ENSMUST NM_007771 // Cry1 // cryptochrome 1 (photolyase-like) // Cry1 0.010 1.47 10 C|10 46.0 cM // 1295 NM_021366 // Klf13 // Kruppel-like factor 13 // 7 C // Klf13 0.003 1.47 50794 /// ENSMUST00000063 NM_011066 // Per2 // period homolog 2 (Drosophila) // 1 Per2 0.010 1.46 C5 // 18627 /// ENSMUST0

253

NM_001008231 // Daam2 // dishevelled associated Daam2 0.012 1.46 activator of morphogenesis 2 // NM_054087 // Slc19a2 // solute carrier family 19 Slc19a2 0.002 1.46 (thiamine transporter), member NM_008037 // Fosl2 // fos-like antigen 2 // 5 B1|5 18.1 cM Fosl2 0.019 1.46 // 14284 /// ENSMUST0 NM_183187 // Fam107a // family with sequence similarity Fam107a 0.019 1.45 107, member A // 14 A1 / NM_023065 // Ifi30 // interferon gamma inducible protein Ifi30 0.046 1.45 30 // 8 B3.3 // 65972 / NM_007819 // Cyp3a13 // cytochrome P450, family 3, Cyp3a13 0.048 1.45 subfamily a, polypeptide 13 / NM_010495 // Id1 // inhibitor of DNA binding 1 // 2 H1|2 Id1 0.004 1.44 84.0 cM // 15901 /// EN NM_023317 // Nde1 // nuclear distribution gene E Nde1 0.014 1.44 homolog 1 (A nidulans) // 16 A1 BC057071 // 1810026J23Rik // RIKEN cDNA 1810026J23 1810026J23Rik 0.014 1.44 gene // 9 A3 // 69773 /// ENS NM_029763 // Polr3f // polymerase (RNA) III (DNA Polr3f 0.012 1.44 directed) polypeptide F // 2 H1 NM_008090 // Gata2 // GATA binding protein 2 // 6 D1|6 Gata2 0.003 1.43 38.5 cM // 14461 /// ENSM NM_198408 // Crhbp // corticotropin releasing hormone Crhbp 0.016 1.43 binding protein // 13 D1|1 NM_008630 // Mt2 // metallothionein 2 // 8 C5|8 45.0 cM Mt2 0.002 1.43 // 17750 /// ENSMUST0000 NM_177089 // Tacc1 // transforming, acidic coiled-coil Tacc1 0.007 1.43 containing protein 1 // 8 NM_001159496 // Ppm1b // protein phosphatase 1B, Ppm1b 0.005 1.43 magnesium dependent, beta isofo NM_197985 // Adipor2 // adiponectin receptor 2 // 6 F1 // Adipor2 0.004 1.43 68465 /// ENSMUST00000 NM_008655 // Gadd45b // growth arrest and DNA- Gadd45b 0.011 1.43 damage-inducible 45 beta // 10 C1| NM_173378 // Trp53bp2 // transformation related protein Trp53bp2 0.002 1.43 53 binding protein 2 // AK049490 // C230029D21Rik // RIKEN cDNA C230029D21 C230029D21Rik 0.011 1.43 gene // 17 A1 // 320255 /// A NM_023476 // Tinagl1 // tubulointerstitial nephritis Tinagl1 0.010 1.42 antigen-like 1 // 4 D3 // 9 NM_011372 // St6galnac3 // ST6 (alpha-N-acetyl- St6galnac3 0.009 1.42 neuraminyl-2,3-beta-galactosyl-1, NM_133966 // Taf5l // TAF5-like RNA polymerase II, Taf5l 0.049 1.42 p300/CBP-associated factor (P BC039988 // Gatsl3 // GATS protein-like 3 // 11 A1 // Gatsl3 0.002 1.42 71962 /// ENSMUST000000206 NM_016787 // Bnip2 // BCL2/adenovirus E1B interacting Bnip2 0.026 1.42 protein 2 // 9 D // 12175 NM_172584 // Itpk1 // inositol 1,3,4-triphosphate 5/6 Itpk1 0.024 1.42 kinase // 12 E // 217837 / NM_172751 // Arhgef10 // Rho guanine nucleotide Arhgef10 0.001 1.41 exchange factor (GEF) 10 // 8 A1 NM_010595 // Kcna1 // potassium voltage-gated channel, Kcna1 0.029 1.41 shaker-related subfamily, NM_025590 // Acot11 // acyl-CoA thioesterase 11 // 4 C7 Acot11 0.039 1.41

254

// 329910 /// ENSMUST000 NM_133222 // Eltd1 // EGF, seven Eltd1 0.018 1.41 transmembrane domain containing 1 / NM_145151 // Crebzf // CREB/ATF bZIP transcription Crebzf 0.016 1.41 factor // 7 E1 // 233490 /// NM_174852 // Phf12 // PHD finger protein 12 // 11 B5 // Phf12 0.010 1.40 268448 /// ENSMUST000000 NM_001080798 // Aff1 // AF4/FMR2 family, member 1 // Aff1 0.048 1.40 5 E|5 56.0 cM // 17355 /// NM_023799 // Mgea5 // meningioma expressed antigen 5 Mgea5 0.015 1.40 (hyaluronidase) // 19 D1|19 NM_172941 // Zkscan17 // zinc finger with KRAB and Zkscan17 0.010 1.40 SCAN domains 17 // 11 B1.3 // NM_175121 // Slc38a2 // solute carrier family 38, member Slc38a2 0.023 1.40 2 // 15 F1 // 67760 /// NM_028298 // Zfp655 // zinc finger protein 655 // 5 G2 // Zfp655 0.001 1.39 72611 /// NM_001083958 NM_010496 // Id2 // inhibitor of DNA binding 2 // 12 B|12 Id2 0.038 1.39 7.0 cM // 15902 /// EN NM_029804 // Hnrnpm // heterogeneous nuclear Hnrnpm 0.012 1.39 ribonucleoprotein M // 17 B1 // 769 NM_138741 // Sdpr // serum deprivation response // 1 Sdpr 0.010 1.39 C1.1 // 20324 /// ENSMUST00 NM_194342 // Unc84b // unc-84 homolog B (C. elegans) // Unc84b 0.046 1.39 15 E1 // 223697 /// ENSM AK021026 // B430319H21Rik // RIKEN cDNA B430319H21 B430319H21Rik 0.021 1.39 gene // 11 B3 // 77849 AK041495 // 9330169L03Rik // RIKEN cDNA 9330169L03 9330169L03Rik 0.047 1.38 gene // --- // 77226 NM_011076 // Abcb1a // ATP-binding cassette, sub-family Abcb1a 0.015 1.38 B (MDR/TAP), member 1A / NM_007839 // Dhx15 // DEAH (Asp-Glu-Ala-His) box Dhx15 0.038 1.38 polypeptide 15 // 5 C1 // 13204 NM_020510 // Fzd2 // frizzled homolog 2 (Drosophila) // Fzd2 0.022 1.38 11 E1 // 57265 /// ENSMU NM_175628 // A2m // alpha-2-macroglobulin // 6 F1|6 A2m 0.023 1.37 61.7 cM // 232345 /// ENSMUS NM_008929 // Dnajc3 // DnaJ (Hsp40) homolog, subfamily Dnajc3 0.017 1.37 C, member 3 // 14 E4 // 1 NM_027460 // Slc25a33 // solute carrier family 25, Slc25a33 0.016 1.37 member 33 // 4 E1 // 70556 // NM_008043 // Frat1 // frequently rearranged in advanced Frat1 0.038 1.37 T-cell lymphomas // 19 C NM_172632 // Mapk4 // mitogen-activated protein kinase Mapk4 0.003 1.37 4 // 18 E2 // 225724 /// NM_033269 // Chrm3 // cholinergic receptor, muscarinic Chrm3 0.019 1.37 3, cardiac // 13 A1|13 7. NM_028059 // Zfp654 // zinc finger protein 654 // 16 C1.3 Zfp654 0.019 1.37 // 72020 /// ENSMUST00 NM_144855 // Cbs // cystathionine beta-synthase // 17 A- Cbs 0.019 1.37 C|17 17.4 cM // 12411 // NM_011817 // Gadd45g // growth arrest and DNA- Gadd45g 0.039 1.37 damage-inducible 45 gamma // 13 A5 NM_178691 // Yod1 // YOD1 OTU deubiquitinating Yod1 0.010 1.36 enzyme 1 homologue (S. cerevisiae

255

NM_030174 // Mctp1 // multiple C2 domains, Mctp1 0.012 1.36 transmembrane 1 // 13 C1 // 78771 /// NM_023275 // Rhoj // ras homolog gene family, member J Rhoj 0.010 1.36 // 12 C3 // 80837 /// ENS NM_144828 // Ppp1r1b // protein phosphatase 1, Ppp1r1b 0.018 1.36 regulatory (inhibitor) subunit 1B NM_013455 // Acr // acrosin prepropeptide // 15 E-F|15 Acr 0.017 1.36 48.6 cM // 11434 /// ENSM NM_013712 // Itgb1bp2 // integrin beta 1 binding protein Itgb1bp2 0.004 1.36 2 // X D // 26549 /// N NM_011055 // Pde3b // phosphodiesterase 3B, cGMP- Pde3b 0.028 1.36 inhibited // 7 F1|7 53.3 cM // NM_009204 // Slc2a4 // solute carrier family 2 (facilitated Slc2a4 0.004 1.35 glucose transporter) NM_008131 // Glul // glutamate-ammonia ligase Glul 0.015 1.35 (glutamine synthetase) // --- // 1 NM_008402 // Itgav // integrin alpha V // 2 D|2 46.0 cM // Itgav 0.029 1.35 16410 /// ENSMUST0000 NM_026072 // Sdccag10 // serologically defined colon Sdccag10 0.007 1.35 cancer antigen 10 // 13 D1 NM_008983 // Ptprk // protein tyrosine phosphatase, Ptprk 0.029 1.35 receptor type, K // 10 A4|10 NM_029248 // Taf1d // TATA box binding protein (Tbp)- Taf1d 0.028 1.35 associated factor, RNA poly NM_009703 // Araf // v-raf murine sarcoma 3611 viral Araf 0.002 1.34 oncogene homolog // X A2-A3 NM_019739 // Foxo1 // forkhead box O1 // 3 C|3 22.5 cM Foxo1 0.006 1.34 // 56458 /// ENSMUST00000 NM_011851 // Nt5e // 5' nucleotidase, ecto // 9 E3.2 // Nt5e 0.014 1.34 23959 /// ENSMUST0000003 BC067068 // BC067068 // cDNA sequence BC067068 // 10 BC067068 0.025 1.34 D1 // 216292 /// ENSMUST000 NM_025836 // M6prbp1 // mannose-6-phosphate M6prbp1 0.016 1.34 receptor binding protein 1 // 17 D / NM_022019 // Dusp10 // dual specificity phosphatase 10 Dusp10 0.036 1.34 // 1 H5 // 63953 /// ENSM NM_007661 // Cdc2l1 // cell division cycle 2-like 1 // 4 Cdc2l1 0.013 1.34 E2|4 79.4 cM // 12537 / NM_020560 // Mrps31 // mitochondrial ribosomal protein Mrps31 0.040 1.34 S31 // 8 A3 // 57312 /// NM_173763 // Ccbl2 // cysteine conjugate-beta lyase 2 // Ccbl2 0.009 1.34 3 H1 // 229905 /// ENSM NM_053247 // Lyve1 // lymphatic vessel endothelial Lyve1 0.004 1.34 hyaluronan receptor 1 // 7 F2 NM_024258 // Usp16 // ubiquitin specific peptidase 16 // Usp16 0.010 1.34 16 C3.3 // 74112 /// EN NM_010417 // Heph // hephaestin // X C3|X 36.4 cM // Heph 0.023 1.34 15203 /// NM_001159628 // H NM_027924 // Pdgfd // platelet-derived growth factor, D Pdgfd 0.031 1.33 polypeptide // 9 A1 // 7 NM_007470 // Apod // apolipoprotein D // 16 B2|16 21.2 Apod 0.028 1.33 cM // 11815 /// ENSMUST00 NM_001111324 // Nedd9 // neural precursor cell Nedd9 0.039 1.33 expressed, developmentally down-r NM_008163 // Grb2 // growth factor receptor bound Grb2 0.044 1.33

256 protein 2 // 11 E2|11 75.0 cM NM_019677 // Plcb1 // phospholipase C, beta 1 // 2 F3|2 Plcb1 0.023 1.33 76.7 cM // 18795 /// NM_ NM_023868 // Ryr2 // ryanodine receptor 2, cardiac // 13 Ryr2 0.006 1.33 A1-A2|13 7.0 cM // 2019 NM_027226 // Fyttd1 // forty-two-three domain Fyttd1 0.008 1.33 containing 1 // 16 B3 // 69823 /// NM_153509 // AF529169 // cDNA sequence AF529169 // 9 AF529169 0.050 1.32 E3.1 // 209743 /// AF529169 NM_181345 // Npm2 // nucleophosmin/nucleoplasmin 2 Npm2 0.002 1.32 // 14 D2 // 328440 /// NM_023 NM_025873 // Trit1 // tRNA isopentenyltransferase 1 // 4 Trit1 0.017 1.32 D2.2 // 66966 /// ENSMU NM_013454 // Abca1 // ATP-binding cassette, sub-family A Abca1 0.008 1.32 (ABC1), member 1 // 4 A NM_009102 // Rrh // retinal pigment epithelium derived Rrh 0.001 1.32 rhodopsin homolog // 3 G3 NM_027427 // Taf15 // TAF15 RNA polymerase II, TATA Taf15 0.017 1.32 box binding protein (TBP)-as NM_144549 // Trib1 // tribbles homolog 1 (Drosophila) // Trib1 0.025 1.32 15 D1 // 211770 /// ENS NM_176993 // Mpzl3 // myelin protein zero-like 3 // 9 Mpzl3 0.031 1.32 A5.2 // 319742 /// NM_0010 NM_009676 // Aox1 // aldehyde oxidase 1 // 1 C1-C2|1 Aox1 0.004 1.31 23.2 cM // 11761 /// ENSMUS NM_028190 // Luc7l // Luc7 homolog (S. cerevisiae)-like // Luc7l 0.031 1.31 17 B1|17 11.8 cM // 6 NM_025950 // Cdc37l1 // cell division cycle 37 homolog (S. Cdc37l1 0.010 1.31 cerevisiae)-like 1 // NM_010513 // Igf1r // insulin-like growth factor I receptor Igf1r 0.008 1.31 // 7 D1|7 33.0 cM // NM_025811 // Nhlrc2 // NHL repeat containing 2 // 19 D2 Nhlrc2 0.013 1.31 // 66866 /// ENSMUST0000 NM_133750 // Fam118a // family with sequence similarity Fam118a 0.022 1.31 118, member A // 15 E3 / NM_028783 // Robo4 // roundabout homolog 4 Robo4 0.005 1.31 (Drosophila) // 9 A4 // 74144 /// ENS NM_001039522 // Leo1 // Leo1, Paf1/RNA polymerase II Leo1 0.013 1.31 complex component, homolog NM_172745 // Tufm // Tu translation elongation factor, Tufm 0.030 1.30 mitochondrial // 7 F3 // NM_001013780 // Slc25a34 // solute carrier family 25, Slc25a34 0.028 1.30 member 34 // 4 E1 // 38407 NM_148673 // Slu7 // SLU7 splicing factor homolog (S. Slu7 0.026 1.30 cerevisiae) // 11 A5-B1.1 NM_009287 // Stim1 // stromal interaction molecule 1 // Stim1 0.020 1.30 7 E3|7 50.0 cM // 20866 NM_010601 // Kcnh3 // potassium voltage-gated channel, Kcnh3 0.017 1.30 subfamily H (eag-related) NM_028748 // Paqr5 // progestin and adipoQ receptor Paqr5 0.041 1.30 family member V // 9 C // 74 NR_003960 // EG432987 // predicted gene, EG432987 // EG432987 0.019 1.30 15 F2|15 // 432987 /// NM_0 NM_011101 // Prkca // protein kinase C, alpha // 11 E1|11 Prkca 0.021 1.30 68.0 cM // 18750 /// E

257

NM_177259 // Dab1 // disabled homolog 1 (Drosophila) // Dab1 0.017 1.30 4 C6|4 52.7 cM // 13131 NM_001159582 // Pde1a // phosphodiesterase 1A, Pde1a 0.030 1.30 calmodulin-dependent // 2 D // 18 NM_007455 // Ap1g2 // adaptor protein complex AP-1, Ap1g2 0.003 1.29 gamma 2 subunit // 14 C3 // NM_016967 // Olig2 // oligodendrocyte transcription Olig2 0.025 1.29 factor 2 // 16 C3.3|16 63.0 NM_010706 // Lgals4 // lectin, galactose binding, soluble 4 Lgals4 0.031 1.29 // 7 A3|7 4.0 cM // NM_028072 // Sulf2 // sulfatase 2 // 2 H3 // 72043 /// Sulf2 0.017 1.29 NM_008679 // Ncoa3 // nuc NM_011896 // Spry1 // sprouty homolog 1 (Drosophila) // Spry1 0.017 1.29 3 B // 24063 /// AF17690 NM_183000 // Accn3 // amiloride-sensitive cation channel Accn3 0.001 1.29 3 // 5 A3 // 171209 /// NM_175258 // Rapgef6 // Rap guanine nucleotide Rapgef6 0.010 1.29 exchange factor (GEF) 6 // 11 B1. NM_173001 // Kdm3a // lysine (K)-specific demethylase Kdm3a 0.026 1.29 3A // 6 C1 // 104263 /// N NM_015823 // Magi2 // membrane associated guanylate Magi2 0.003 1.29 kinase, WW and PDZ domain co NM_011202 // Ptpn11 // protein tyrosine phosphatase, Ptpn11 0.026 1.28 non-receptor type 11 // 5 F NM_145556 // Tardbp // TAR DNA binding protein // 4 E2 Tardbp 0.020 1.28 // 230908 /// NM_00100854 NM_024194 // Lrrc40 // leucine rich repeat containing 40 Lrrc40 0.034 1.28 // 3 H4 // 67144 /// EN NM_001081232 // D5Ertd579e // DNA segment, Chr 5, D5Ertd579e 0.028 1.28 ERATO Doi 579, expressed // 5 NM_016669 // Crym // crystallin, mu // 7 F2|7 55.0 cM // Crym 0.016 1.28 12971 /// ENSMUST000000 NM_008635 // Mtap7 // microtubule-associated protein 7 Mtap7 0.040 1.28 // 10 A3|10 11.0 cM // 17 NM_020588 // Tmem183a // transmembrane protein Tmem183a 0.019 1.28 183A // 1 E4 // 57439 /// NM_0010 NM_029492 // Zdhhc20 // zinc finger, DHHC domain Zdhhc20 0.044 1.28 containing 20 // 14 C3 // 75965 NM_181407 // Me3 // malic enzyme 3, NADP(+)- Me3 0.006 1.28 dependent, mitochondrial // 7 E1 // NM_178912 // Fancm // Fanconi anemia, Fancm 0.019 1.28 complementation group M // 12 C1|12 29.0 c NM_013495 // Cpt1a // carnitine palmitoyltransferase 1a, Cpt1a 0.044 1.27 liver // 19 A|19 2.0 cM NM_007842 // Dhx9 // DEAH (Asp-Glu-Ala-His) box Dhx9 0.017 1.27 polypeptide 9 // 1 G3|1 77.0 cM NM_013659 // Sema4b // sema domain, immunoglobulin Sema4b 0.016 1.27 domain (Ig), transmembrane do NM_009031 // Rbbp7 // retinoblastoma binding protein 7 Rbbp7 0.026 1.27 // X F4 // 245688 /// ENS NM_007696 // Ovgp1 // oviductal glycoprotein 1 // 3 F3 // Ovgp1 0.027 1.27 12659 /// ENSMUST00000 NM_133815 // Lbr // lamin B receptor // 1 H5|1 97.3 cM Lbr 0.041 1.27 // 98386 /// ENSMUST00000 NM_008235 // Hes1 // hairy and enhancer of split 1 Hes1 0.022 1.27

258

(Drosophila) // 16 B2|16 27.0 NM_019923 // Itpr2 // inositol 1,4,5-triphosphate receptor Itpr2 0.022 1.27 2 // 6 G3|6 73.0 cM / NM_001110197 // Rnf146 // ring finger protein 146 // 10 Rnf146 0.009 1.27 A4 // 68031 /// NM_00111 NM_134164 // Syt12 // synaptotagmin XII // 19 A // Syt12 0.009 1.27 171180 /// ENSMUST00000059295 NM_021398 // Slc43a3 // solute carrier family 43, member Slc43a3 0.028 1.27 3 // --- // 58207 /// E NM_009180 // St6galnac2 // ST6 (alpha-N-acetyl- St6galnac2 0.034 1.27 neuraminyl-2,3-beta-galactosyl-1, NM_148933 // Slco4a1 // solute carrier organic anion Slco4a1 0.037 1.27 transporter family, member NM_183308 // Pon2 // paraoxonase 2 // 6 A1 // 330260 Pon2 0.002 1.27 /// ENSMUST00000057792 // P NM_001130187 // Smarcd2 // SWI/SNF related, matrix Smarcd2 0.027 1.27 associated, actin dependent r NM_023644 // Mccc1 // methylcrotonoyl-Coenzyme A Mccc1 0.010 1.26 carboxylase 1 (alpha) // 3 B // NM_018864 // Impa1 // inositol (myo)-1(or 4)- Impa1 0.017 1.26 monophosphatase 1 // 3 A1 // 55980 NM_030249 // Cttnbp2nl // CTTNBP2 N-terminal like // 3 Cttnbp2nl 0.010 1.26 F2.2 // 80281 /// ENSMUST NM_026185 // 1300007F04Rik // RIKEN cDNA 1300007F04 1300007F04Rik 0.023 1.26 gene // 11 B5 // 67477 /// F NM_145705 // Tinf2 // Terf1 (TRF1)-interacting nuclear Tinf2 0.010 1.26 factor 2 // 14 C3|14 22.5 NM_026243 // Mgat4c // mannosyl (alpha-1,3-)- Mgat4c 0.023 1.26 glycoprotein beta-1,4-N-acetylgluco NM_207624 // Ace // angiotensin I converting enzyme Ace 0.041 1.26 (peptidyl-dipeptidase A) 1 / NM_178194 // Hist1h2be // histone cluster 1, H2be // 13 Hist1h2be 0.015 1.26 A2-A3 // 319179 /// NM_0 NM_153136 // Nudt18 // nudix (nucleoside diphosphate Nudt18 0.019 1.26 linked moiety X)-type motif NM_133641 // Rtkn // rhotekin // 6 C3|6 34.81 cM // Rtkn 0.035 1.26 20166 /// NM_009106 // Rtkn NM_011600 // Tle4 // transducin-like enhancer of split 4, Tle4 0.039 1.26 homolog of Drosophila NM_146041 // Gmds // GDP-mannose 4, 6-dehydratase // Gmds 0.029 1.26 13 A3.2 // 218138 /// ENSMU NM_001040435 // Tacc3 // transforming, acidic coiled-coil Tacc3 0.021 1.25 containing protein 3 / NM_028227 // Brap // BRCA1 associated protein // 5 F // Brap 0.012 1.25 72399 /// ENSMUST0000003 NM_145533 // Smox // spermine oxidase // 2 F1 // Smox 0.049 1.25 228608 /// ENSMUST00000028806 / NM_008102 // Gch1 // GTP cyclohydrolase 1 // 14 C2-3|14 Gch1 0.044 1.25 19.5 cM // 14528 /// ENS NM_147072 // Olfr641 // olfactory receptor 641 // --- // Olfr641 0.021 1.25 259075 /// ENSMUST00000 NM_008842 // Pim1 // proviral integration site 1 // 17 Pim1 0.043 1.25 A3.3|17 16.4 cM // 18712 NM_138744 // Ssx2ip // synovial sarcoma, X breakpoint 2 Ssx2ip 0.023 1.25 interacting protein // 3

259

BC056942 // D6Wsu116e // DNA segment, Chr 6, Wayne D6Wsu116e 0.031 1.25 State University 116, express NM_007859 // Dffb // DNA fragmentation factor, beta Dffb 0.043 1.25 subunit // 4 E2 // 13368 /// NM_001077495 // Pik3r1 // phosphatidylinositol 3-kinase, Pik3r1 0.036 1.25 regulatory subunit, pol NM_145987 // Tmem82 // transmembrane protein 82 // 4 Tmem82 0.029 1.24 E1 // 213989 /// ENSMUST000 NM_016926 // Sart3 // squamous cell carcinoma antigen Sart3 0.016 1.24 recognized by T-cells 3 // NM_015790 // Icosl // icos ligand // 10 C1 // 50723 /// Icosl 0.031 1.24 ENSMUST00000105393 // Ic NM_016813 // Nxf1 // nuclear RNA export factor 1 Nxf1 0.030 1.23 homolog (S. cerevisiae) // 19 A NM_177028 // 5330437I02Rik // RIKEN cDNA 5330437I02 5330437I02Rik 0.045 1.23 gene // 18 E1 // 319888 /// NM_021343 // Spata5 // spermatogenesis associated 5 // Spata5 0.037 1.23 3 B // 57815 /// ENSMUST0 NM_177059 // Fstl4 // follistatin-like 4 // 11 B1.3 // Fstl4 0.039 1.23 320027 /// ENSMUST0000003 NM_146118 // Slc25a25 // solute carrier family 25 Slc25a25 0.023 1.22 (mitochondrial carrier, phosph NM_133672 // Vps26a // vacuolar protein sorting 26 Vps26a 0.036 1.22 homolog A (yeast) // 10 B3-B5 NM_033524 // Spred1 // sprouty protein with EVH-1 Spred1 0.024 1.22 domain 1, related sequence // NM_133904 // Acacb // acetyl-Coenzyme A carboxylase Acacb 0.019 1.22 beta // 5 F // 100705 /// EN NM_029095 // Hhatl // hedgehog acyltransferase-like // 9 Hhatl 0.046 1.22 F4 // 74770 /// ENSMUST NM_026509 // Murc // muscle-related coiled-coil protein Murc 0.022 1.22 // 4 B2 // 68016 /// ENS NM_008974 // Ptp4a2 // protein tyrosine phosphatase 4a2 Ptp4a2 0.031 1.21 // 4 D2.3 // 19244 /// E NM_025298 // Polr3e // polymerase (RNA) III (DNA Polr3e 0.002 1.21 directed) polypeptide E // 7 F3 NM_026004 // Nt5c3 // 5'-nucleotidase, cytosolic III // 6 Nt5c3 0.006 1.21 B3 // 107569 /// ENSMU NM_145216 // Rasl10a // RAS-like, family 10, member A // Rasl10a 0.037 1.21 11 A1 // 75668 /// ENSM NM_021299 // Ak3 // adenylate kinase 3 // 19 C1 // 56248 Ak3 0.050 1.21 /// NR_003968 // ENSMUS NM_028315 // Dis3 // DIS3 mitotic control homolog (S. Dis3 0.011 1.21 cerevisiae) // 14 E2.2 // NM_026391 // Ppp2r2d // protein phosphatase 2, Ppp2r2d 0.023 1.21 regulatory subunit B, delta isofo NM_080288 // Elmo1 // engulfment and cell motility 1, Elmo1 0.039 1.21 ced-12 homolog (C. elegans NM_144824 // Wrap53 // WD repeat containing, antisense Wrap53 0.026 1.21 to TP53 // 11 B3 // 21685 NM_144792 // Sgms1 // sphingomyelin synthase 1 // 19 Sgms1 0.039 1.21 C1 // 208449 /// ENSMUST000 NM_153550 // Dirc2 // disrupted in renal carcinoma 2 Dirc2 0.023 1.21 (human) // 16 B3 // 224132 BC116260 // 9130227C08Rik // RIKEN cDNA 9130227C08Rik 0.019 1.21

260

9130227C08Rik gene // 14 C1 // 219094 // NM_028882 // Sema3d // sema domain, immunoglobulin Sema3d 0.044 1.21 domain (Ig), short basic doma NM_178591 // Nrg1 // neuregulin 1 // 8 A3 // 211323 /// Nrg1 0.028 1.20 ENSMUST00000073884 // Nr NM_019696 // Cpxm1 // carboxypeptidase X 1 (M14 Cpxm1 0.023 1.20 family) // 2 F1 // 56264 /// ENS NM_028829 // Paqr8 // progestin and adipoQ receptor Paqr8 0.005 1.20 family member VIII // 1 A5 / NM_009640 // Angpt1 // angiopoietin 1 // 15 B3.1|15 14.3 Angpt1 0.008 1.20 cM // 11600 /// ENSMUST NM_011728 // Xpa // xeroderma pigmentosum, Xpa 0.023 1.20 complementation group A // 4 C2|4 21. NM_001080129 // Tmpo // thymopoietin // 10 C2|10 49.0 Tmpo 0.050 1.20 cM // 21917 /// NM_0010801 NM_153142 // Slc35e4 // solute carrier family 35, member Slc35e4 0.018 1.20 E4 // 11 A1 // 103710 / NM_009201 // Slc1a5 // solute carrier family 1 (neutral Slc1a5 0.008 1.19 amino acid transporter), NM_013540 // Gria2 // glutamate receptor, ionotropic, Gria2 0.035 1.19 AMPA2 (alpha 2) // 3 E3|3 NM_015804 // Atp11a // ATPase, class VI, type 11A // 8 A2 Atp11a 0.037 1.19 // 50770 /// ENSMUST00 NM_033370 // Copb1 // coatomer protein complex, Copb1 0.028 1.19 subunit beta 1 // 7 F1|7 53.3 cM NM_016752 // Slc35b1 // solute carrier family 35, member Slc35b1 0.031 1.19 B1 // 11 D|11 55.5 cM / NM_009828 // Ccna2 // cyclin A2 // 3 B|3 19.2 cM // Ccna2 0.009 1.19 12428 /// ENSMUST00000029270 NM_009171 // Shmt1 // serine hydroxymethyltransferase Shmt1 0.024 1.19 1 (soluble) // 11 B2 // 20 NM_013862 // Rabgap1l // RAB GTPase activating protein Rabgap1l 0.033 1.19 1-like // 1 H2.1 // 29809 NM_026186 // Ccdc49 // coiled-coil domain containing 49 Ccdc49 0.026 1.18 // 11 D // 67480 /// ENS NM_153144 // Ggnbp2 // gametogenetin binding protein 2 Ggnbp2 0.043 1.18 // 11 C // 217039 /// ENS NM_016711 // Tmod2 // tropomodulin 2 // 9 D|9 38.0 cM Tmod2 0.034 1.18 // 50876 /// NM_001038710 NM_194355 // Spire1 // spire homolog 1 (Drosophila) // Spire1 0.029 1.18 18 E1 // 68166 /// NM_176 ENSMUST00000038890 // Dennd4a // DENN/MADD Dennd4a 0.033 1.18 domain containing 4A // 9 C // 102442 NM_013930 // Aass // aminoadipate-semialdehyde Aass 0.026 1.18 synthase // 6 A3.1|6 4.5 cM // 30 NM_145605 // Klhdc4 // kelch domain containing 4 // 8 E1 Klhdc4 0.024 1.18 // 234825 /// ENSMUST00 NM_010028 // Ddx3x // DEAD/H (Asp-Glu-Ala-Asp/His) Ddx3x 0.041 1.18 box polypeptide 3, X-linked / NM_001077202 // Hs6st2 // heparan sulfate 6-O- Hs6st2 0.017 1.17 sulfotransferase 2 // X A3.3 // 50 NM_008156 // Gpld1 // glycosylphosphatidylinositol Gpld1 0.028 1.17 specific phospholipase D1 // NM_011250 // Rbl2 // retinoblastoma-like 2 // 8 C5|8 Rbl2 0.022 1.17 40.99 cM // 19651 /// ENSMU

261

BC017643 // BC017643 // cDNA sequence BC017643 // 11 BC017643 0.039 1.17 E2 // 217370 /// BC026616 / NM_025926 // Dnajb4 // DnaJ (Hsp40) homolog, subfamily Dnajb4 0.031 1.17 B, member 4 // 3 H3 // 67 NM_028071 // Cotl1 // coactosin-like 1 (Dictyostelium) // Cotl1 0.045 1.17 8 E1 // 72042 /// ENSM NM_145985 // Arcn1 // archain 1 // 9 A5.2 // 213827 /// Arcn1 0.043 1.17 ENSMUST00000034607 // Ar NM_001013414 // Neurl4 // neuralized homolog 4 Neurl4 0.045 1.17 (Drosophila) // 11 B3 // 216860 / NM_011075 // Abcb1b // ATP-binding cassette, sub-family Abcb1b 0.028 1.17 B (MDR/TAP), member 1B / NM_130880 // Otud7a // OTU domain containing 7A // 7 C Otud7a 0.047 1.17 // 170711 /// NM_146153 / NM_021041 // Abcc9 // ATP-binding cassette, sub-family C Abcc9 0.043 1.17 (CFTR/MRP), member 9 // NM_172265 // Eif2b5 // eukaryotic translation initiation Eif2b5 0.028 1.17 factor 2B, subunit 5 ep NM_011875 // Psmd13 // proteasome (prosome, Psmd13 0.045 1.16 macropain) 26S subunit, non-ATPase, NM_028784 // F13a1 // coagulation factor XIII, A1 subunit F13a1 0.038 1.16 // 13 A3.3 // 74145 // NM_007993 // Fbn1 // fibrillin 1 // 2 F|2 71.0 cM // 14118 Fbn1 0.028 1.16 /// ENSMUST0000010323 NM_178703 // Slc6a1 // solute carrier family 6 Slc6a1 0.050 1.15 (neurotransmitter transporter, GA NM_021450 // Trpm7 // transient receptor potential Trpm7 0.028 1.15 cation channel, subfamily M, NM_023887 // Gcnt2 // glucosaminyl (N-acetyl) Gcnt2 0.032 1.15 transferase 2, I-branching enzyme NM_030612 // Nfkbiz // nuclear factor of kappa light Nfkbiz 0.040 1.15 polypeptide gene enhancer i NM_175341 // Mbnl2 // muscleblind-like 2 // 14 E4 // Mbnl2 0.010 1.15 105559 /// NM_207515 // Mbn NM_170778 // Dpyd // dihydropyrimidine dehydrogenase Dpyd 0.046 1.14 // 3 G1 // 99586 /// ENSMUS NM_008913 // Ppp3ca // protein phosphatase 3, catalytic Ppp3ca 0.003 1.14 subunit, alpha isoform / NM_015799 // Trfr2 // transferrin receptor 2 // 5 G2 // Trfr2 0.038 1.14 50765 /// ENSMUST0000003 NM_011100 // Prkacb // protein kinase, cAMP dependent, Prkacb 0.043 1.14 catalytic, beta // 3 H3 / NM_183428 // Epb4.1 // erythrocyte protein band 4.1 // 4 Epb4.1 0.026 1.14 D2.3 // 269587 /// NM_0 NM_001004156 // Plekhg5 // pleckstrin homology domain Plekhg5 0.022 1.14 containing, family G (with NM_016770 // Folh1 // folate hydrolase // 7 D1-D2 // Folh1 0.037 1.14 53320 /// NM_001159706 // F NM_008905 // Ppfibp2 // protein tyrosine phosphatase, Ppfibp2 0.043 1.14 receptor-type, F interacti NM_145598 // Nxnl1 // nucleoredoxin-like 1 // 8 B3.3 // Nxnl1 0.015 1.13 234404 /// BC021911 // N NM_175750 // Plxna4 // plexin A4 // 6 B1 // 243743 /// Plxna4 0.023 1.12 ENSMUST00000115096 // Plx NM_028460 // Pear1 // platelet endothelial aggregation Pear1 0.017 1.11

262 receptor 1 // 3 F1 // 731 NM_022319 // Clstn2 // calsyntenin 2 // --- // 64085 /// Clstn2 0.049 1.11 ENSMUST00000035027 // C NM_144787 // Kdm4c // lysine (K)-specific demethylase 4C Kdm4c 0.035 1.08 // 4 C3 // 76804 /// EN NM_153127 // Mmrn2 // multimerin 2 // 14 B // 105450 Mmrn2 0.028 -1.10 /// ENSMUST00000111908 // M NM_025617 // 2210012G02Rik // RIKEN cDNA 2210012G02Rik 0.029 -1.11 2210012G02 gene // 4 C7 // 66526 /// EN NM_178087 // Pml // promyelocytic leukemia // 9 B|9 Pml 0.029 -1.11 32.0 cM // 18854 /// NM_0088 NM_029291 // Ascc2 // activating signal cointegrator 1 Ascc2 0.016 -1.12 complex subunit 2 // 11 A NM_010472 // Agfg1 // ArfGAP with FG repeats 1 // 1 C5 Agfg1 0.039 -1.12 // 15463 /// ENSMUST00000 NM_011711 // Fmnl3 // formin-like 3 // 15 F3 // 22379 /// Fmnl3 0.031 -1.12 ENSMUST00000088233 // NM_172752 // Sorbs2 // sorbin and SH3 domain Sorbs2 0.028 -1.12 containing 2 // 8 B1.1 // 234214 // NM_019651 // Ptpn9 // protein tyrosine phosphatase, Ptpn9 0.033 -1.12 non-receptor type 9 // --- / NM_008444 // Kif3b // kinesin family member 3B // 2 86.0 Kif3b 0.041 -1.12 cM // 16569 /// ENSMUST NM_178688 // Ablim1 // actin-binding LIM protein 1 // 19 Ablim1 0.018 -1.12 D2|19 53.0 cM // 226251 NM_001039484 // Kcnj10 // potassium inwardly-rectifying Kcnj10 0.039 -1.13 channel, subfamily J, me NM_173012 // Letm2 // leucine zipper-EF-hand containing Letm2 0.031 -1.13 transmembrane protein 2 NM_007578 // Cacna1a // calcium channel, voltage- Cacna1a 0.017 -1.13 dependent, P/Q type, alpha 1A s NM_009553 // Zscan2 // zinc finger and SCAN domain Zscan2 0.050 -1.13 containing 2 // 7 D3|7 42.0 c NM_130454 // Recql5 // RecQ protein-like 5 // 11 E2 // Recql5 0.033 -1.13 170472 /// ENSMUST0000002 NM_026444 // Cs // citrate synthase // 10 D3 // 12974 /// Cs 0.047 -1.13 NM_027945 // Csl // ci BC050895 // 9130401M01Rik // RIKEN cDNA 9130401M01 9130401M01Rik 0.023 -1.13 gene // 15 D1 // 75758 /// EN NM_133756 // Gpn1 // GPN-loop GTPase 1 // 5 B1 // Gpn1 0.048 -1.13 74254 /// ENSMUST00000076949 / NM_172748 // Fbxl19 // F-box and leucine-rich repeat Fbxl19 0.039 -1.14 protein 19 // 7 F3 // 23390 NM_198298 // Helz // helicase with zinc finger domain // Helz 0.018 -1.14 11 E1 // 78455 /// ENSM NM_144835 // Heatr1 // HEAT repeat containing 1 // 13 Heatr1 0.014 -1.14 A1 // 217995 /// ENSMUST00 NM_001004721 // Pigu // phosphatidylinositol glycan Pigu 0.020 -1.14 anchor biosynthesis, class U NM_027448 // Lca5 // Leber congenital amaurosis 5 Lca5 0.023 -1.14 (human) // 9 E2 // 75782 /// N NM_175381 // 2700081O15Rik // RIKEN cDNA 2700081O15Rik 0.022 -1.14 2700081O15 gene // 19 A // 108899 /// E NM_145415 // AA408296 // expressed sequence AA408296 0.038 -1.15 AA408296 // 1 H6 // 215193 /// ENSMU

263

NM_172283 // Fuk // fucokinase // 8 E1 // 234730 /// Fuk 0.004 -1.15 NM_181666 // Fuk // fucokin NM_025812 // Hmg20a // high mobility group 20A // 9 C Hmg20a 0.046 -1.15 // 66867 /// ENSMUST000000 NM_172768 // Gramd1b // GRAM domain containing 1B // Gramd1b 0.010 -1.16 9 B // 235283 /// ENSMUST00 NM_172280 // 2210018M11Rik // RIKEN cDNA 2210018M11Rik 0.020 -1.16 2210018M11 gene // 7 E2 // 233545 /// E NM_013850 // Abca7 // ATP-binding cassette, sub-family A Abca7 0.008 -1.16 (ABC1), member 7 // 10 NM_199145 // 3110062M04Rik // RIKEN cDNA 3110062M04Rik 0.012 -1.16 3110062M04 gene // 6 B1 // 78412 /// NM NM_133191 // Eps8l2 // EPS8-like 2 // 7 F5 // 98845 /// Eps8l2 0.028 -1.16 NM_177280 // B230206H07R NM_029081 // 5730419I09Rik // RIKEN cDNA 5730419I09 5730419I09Rik 0.023 -1.16 gene // 6 G2 // 74741 /// NM NM_152894 // Pop1 // processing of precursor 1, Pop1 0.001 -1.16 ribonuclease P/MRP family, (S. c NM_030075 // Klhdc8b // kelch domain containing 8B // 9 Klhdc8b 0.045 -1.16 F2 // 78267 /// ENSMUST0 NM_013729 // Mixl1 // Mix1 homeobox-like 1 (Xenopus Mixl1 0.047 -1.16 laevis) // 1 H4 // 27217 /// NM_177755 // Klhl38 // kelch-like 38 (Drosophila) // 15 D1 Klhl38 0.039 -1.17 // 268807 /// ENSMUST NM_178704 // Dpy19l3 // dpy-19-like 3 (C. elegans) // 7 B2 Dpy19l3 0.048 -1.17 // 233115 /// ENSMUST NM_026872 // Ubap2 // ubiquitin-associated protein 2 // Ubap2 0.028 -1.17 4 A5 // 68926 /// ENSMUS NM_010110 // Efnb1 // ephrin B1 // X D|X 37.0 cM // Efnb1 0.044 -1.18 13641 /// ENSMUST00000052839 NM_145426 // Mfap3 // microfibrillar-associated protein 3 Mfap3 0.039 -1.18 // 11 B1.3 // 216760 / NM_025864 // Tmem206 // transmembrane protein 206 Tmem206 0.023 -1.18 // 1 H6 // 66950 /// ENSMUST00 NM_008172 // Grin2d // glutamate receptor, ionotropic, Grin2d 0.020 -1.18 NMDA2D (epsilon 4) // 7 B NM_028812 // Gtf2e1 // general transcription factor II E, Gtf2e1 0.022 -1.18 polypeptide 1 (alpha s NM_134011 // Tbrg4 // transforming growth factor beta Tbrg4 0.021 -1.18 regulated gene 4 // 11 A1 NR_003544 // Nnt // nicotinamide nucleotide Nnt 0.024 -1.18 transhydrogenase // 13 D2|13 64.0 cM NM_008535 // Lyl1 // lymphoblastomic leukemia 1 // 8 Lyl1 0.013 -1.18 C3|8 38.5 cM // 17095 /// E NM_178738 // Prss35 // protease, serine, 35 // 9 E3.1 // Prss35 0.046 -1.18 244954 /// ENSMUST00000 NM_153402 // Eif2c3 // eukaryotic translation initiation Eif2c3 0.038 -1.18 factor 2C, 3 // 4 D2.2 NM_018854 // Ift20 // intraflagellar transport 20 homolog Ift20 0.026 -1.19 (Chlamydomonas) // 11 NM_001005247 // Hps5 // Hermansky-Pudlak syndrome 5 Hps5 0.020 -1.19 homolog (human) // 7 B4|7 25 NM_025473 // Fam3a // family with sequence similarity 3, Fam3a 0.045 -1.19 member A // X A6 // 662 NM_178848 // Sirt5 // sirtuin 5 (silent mating type Sirt5 0.041 -1.19

264 information regulation 2 hom NM_011829 // Impdh1 // inosine 5'-phosphate Impdh1 0.035 -1.19 dehydrogenase 1 // 6 A3 // 23917 /// NM_013784 // Pign // phosphatidylinositol glycan anchor Pign 0.023 -1.19 biosynthesis, class N // NM_019942 // Sept6 // septin 6 // X A2 // 56526 /// Sep-06 0.027 -1.19 ENSMUST00000115241 // Sept6 NM_011026 // P2rx4 // purinergic receptor P2X, ligand- P2rx4 0.039 -1.19 gated ion channel 4 // 5 F NM_176840 // Osbpl11 // oxysterol binding protein-like 11 Osbpl11 0.017 -1.19 // 16 B3 // 106326 /// NM_001005863 // Mtus1 // mitochondrial tumor Mtus1 0.007 -1.19 suppressor 1 // 8 A4 // 102103 /// NM_001083937 // Slc35a2 // solute carrier family 35 (UDP- Slc35a2 0.041 -1.20 galactose transporter), NM_013520 // Flt3l // FMS-like tyrosine kinase 3 ligand // Flt3l 0.019 -1.20 7 B2-C|7 23.0 cM // 1 NM_030245 // Tada1l // transcriptional adaptor 1 (HFI1 Tada1l 0.023 -1.20 homolog, yeast) like // 1 NM_053250 // Crip3 // cysteine-rich protein 3 // 17 C|17 Crip3 0.027 -1.20 // 114570 /// NM_181664 NM_183178 // Fsd1 // fibronectin type 3 and SPRY Fsd1 0.031 -1.20 domain-containing protein // 17 NM_009568 // Zfp94 // zinc finger protein 94 // 7 A3|7 Zfp94 0.041 -1.20 5.25 cM // 22756 /// NM_0 NM_009648 // Akap1 // A kinase (PRKA) anchor protein 1 Akap1 0.006 -1.20 // 11 C // 11640 /// NM_0 NM_009367 // Tgfb2 // transforming growth factor, beta 2 Tgfb2 0.022 -1.21 // 1 H5|1 101.5 cM // 2 NM_019426 // Atf7ip // activating transcription factor 7 Atf7ip 0.026 -1.21 interacting protein // NM_012031 // Spag1 // sperm associated antigen 1 // 15 C Spag1 0.016 -1.21 // 26942 /// ENSMUST000 NM_001015876 // Tyw1 // tRNA-yW synthesizing protein Tyw1 0.041 -1.21 1 homolog (S. cerevisiae) / NM_009616 // Adam19 // a disintegrin and Adam19 0.020 -1.21 metallopeptidase domain 19 (meltrin bet NM_023750 // Zfp84 // zinc finger protein 84 // 7 B1|7 9.0 Zfp84 0.045 -1.21 cM // 74352 /// ENSMU NM_001042719 // Ddhd1 // DDHD domain containing 1 // Ddhd1 0.010 -1.21 14 C1 // 114874 /// NM_1768 NM_009746 // Bcl7c // B-cell CLL/lymphoma 7C // 7 F4 // Bcl7c 0.034 -1.21 12055 /// NR_027380 // 1 NM_144550 // Ccdc52 // coiled-coil domain containing 52 Ccdc52 0.010 -1.21 // 16 B4|16 28.9 cM // 2 NM_019835 // B4galt5 // UDP-Gal:betaGlcNAc beta 1,4- B4galt5 0.047 -1.21 galactosyltransferase, polyp NM_026307 // Cuta // cutA divalent cation tolerance Cuta 0.011 -1.21 homolog (E. coli) // 17 B1 / NM_146142 // Tdrd7 // tudor domain containing 7 // 4 B1 Tdrd7 0.038 -1.21 // 100121 /// ENSMUST000 NM_198625 // Mtss1l // metastasis suppressor 1-like // 8 Mtss1l 0.030 -1.21 E1 // 244654 /// ENSMUS NM_027799 // Ankrd40 // ankyrin repeat domain 40 // 11 Ankrd40 0.026 -1.21 C // 71452 /// NM_146024

265

NM_011755 // Zfp35 // zinc finger protein 35 // 18 A2 // Zfp35 0.026 -1.21 22694 /// ENSMUST000000 NM_011585 // Tia1 // cytotoxic granule-associated RNA Tia1 0.012 -1.21 binding protein 1 // 6 D2 NM_027376 // Fuz // fuzzy homolog (Drosophila) // 7 B2 Fuz 0.044 -1.21 // 70300 /// ENSMUST00000 NM_026275 // Ube2r2 // ubiquitin-conjugating enzyme Ube2r2 0.043 -1.21 E2R 2 // 4 B1 // 67615 /// E NM_145974 // C330016O10Rik // RIKEN cDNA C330016O10Rik 0.043 -1.21 C330016O10 gene // 11 B1.3 // 212706 // BC064462 // Cep192 // centrosomal protein 192 // 18 E1 Cep192 0.025 -1.21 // 70799 /// ENSMUST00000 NM_144908 // Galnt11 // UDP-N-acetyl-alpha-D- Galnt11 0.020 -1.22 galactosamine:polypeptide N-acetylg NM_026793 // Myct1 // myc target 1 // 10 A1 // 68632 /// Myct1 0.037 -1.22 ENSMUST00000051809 // M NM_016745 // Atp2a3 // ATPase, Ca++ transporting, Atp2a3 0.024 -1.22 ubiquitous // 11 B4 // 53313 / BC007156 // Myo19 // myosin XIX // 11 B5 // 66196 /// Myo19 0.015 -1.22 BC060115 // Myo19 // myosi NM_011587 // Tie1 // tyrosine kinase with Tie1 0.028 -1.22 immunoglobulin-like and EGF-like domai NM_011841 // Mapk7 // mitogen-activated protein kinase Mapk7 0.034 -1.22 7 // 11 B2 // 23939 /// E NM_024203 // Fam120b // family with sequence similarity Fam120b 0.041 -1.22 120, member B // 17 A2 / NM_175382 // 2700049P18Rik // RIKEN cDNA 2700049P18 2700049P18Rik 0.035 -1.22 gene // 1 E4 // 108900 /// B NM_001010833 // Mdc1 // mediator of DNA damage Mdc1 0.047 -1.22 checkpoint 1 // 17 B1 // 240087 / NM_026304 // l7Rn6 // lethal, Chr 7, Rinchik 6 // 7 E1 // l7Rn6 0.045 -1.22 67669 /// ENSMUST00000 NM_026210 // Fam18b // family with sequence similarity Fam18b 0.044 -1.22 18, member B // 11 B2 // NM_027007 // Zfp397 // zinc finger protein 397 // 18 A2 // Zfp397 0.018 -1.22 69256 /// ENSMUST0000 NM_146185 // Zfp790 // zinc finger protein 790 // 7 B1 // Zfp790 0.035 -1.22 233056 /// NM_00114588 NM_009863 // Cdc7 // cell division cycle 7 (S. cerevisiae) // Cdc7 0.023 -1.23 5 E // 12545 /// E NM_013703 // Vldlr // very low density lipoprotein Vldlr 0.044 -1.23 receptor // 19 C1|19 20.0 cM NM_175031 // Stk36 // serine/threonine kinase 36 (fused Stk36 0.005 -1.23 homolog, Drosophila) // NM_144844 // Pcca // propionyl-Coenzyme A carboxylase, Pcca 0.023 -1.23 alpha polypeptide // 14 E NM_175316 // Slco2b1 // solute carrier organic anion Slco2b1 0.043 -1.23 transporter family, member NM_178877 // Nhedc2 // Na+/H+ exchanger domain Nhedc2 0.029 -1.23 containing 2 // 3 G3 // 97086 /// NM_138668 // Ufsp2 // UFM1-specific peptidase 2 // 8 Ufsp2 0.022 -1.23 B1.1 // 192169 /// ENSMUST0 NM_008453 // Klf3 // Kruppel-like factor 3 (basic) // 5 C3.1 Klf3 0.017 -1.23 // 16599 /// DQ9818 NM_009632 // Parp2 // poly (ADP-ribose) polymerase Parp2 0.017 -1.23

266 family, member 2 // 14 C1|14 NM_026031 // Utp11l // UTP11-like, U3 small nucleolar Utp11l 0.016 -1.23 ribonucleoprotein, (yeast) NM_001081154 // 4921513D23Rik // RIKEN cDNA 4921513D23Rik 0.036 -1.23 4921513D23 gene // 16 A1 // 223989 / NM_008709 // Mycn // v-myc myelocytomatosis viral Mycn 0.012 -1.23 related oncogene, neuroblastom NM_015810 // Polg2 // polymerase (DNA directed), Polg2 0.032 -1.23 gamma 2, accessory subunit // 1 NM_173416 // BC068281 // cDNA sequence BC068281 // BC068281 0.018 -1.23 12 A1.1 // 238037 /// BC06828 NM_010508 // Ifnar1 // interferon (alpha and beta) Ifnar1 0.011 -1.23 receptor 1 // 16 C3.3|16 63.2 NM_053078 // D0H4S114 // DNA segment, human D4S114 D0H4S114 0.017 -1.23 // 18 B1 // 27528 /// NM_0011 NM_008541 // Smad5 // MAD homolog 5 (Drosophila) // Smad5 0.014 -1.23 13 B1|13 35.0 cM // 17129 // NM_016900 // Cav2 // caveolin 2 // 6 A2 // 12390 /// Cav2 0.040 -1.24 ENSMUST00000000058 // Cav2 NM_015795 // Fbxo16 // F-box protein 16 // 14 D1 // Fbxo16 0.033 -1.24 50759 /// ENSMUST00000043554 NM_010247 // Xrcc6 // X-ray repair complementing Xrcc6 0.012 -1.24 defective repair in Chinese ham NM_016703 // Preb // regulatory element Preb 0.024 -1.24 binding // 5 B1 // 50907 /// N NM_133828 // Creb1 // cAMP responsive element binding Creb1 0.022 -1.24 protein 1 // 1 C2|1 31.0 c NM_026765 // Uckl1 // uridine-cytidine kinase 1-like 1 // 2 Uckl1 0.016 -1.24 H4 // 68556 /// ENSM NM_028334 // Nup37 // nucleoporin 37 // 10 C2 // 69736 Nup37 0.033 -1.24 /// NM_027191 // Nup37 // NM_152818 // Osbp2 // oxysterol binding protein 2 // 11 Osbp2 0.031 -1.24 A1 // 74309 /// ENSMUST0 NM_001146180 // Mtss1 // metastasis suppressor 1 // 15 Mtss1 0.024 -1.24 D1 // 211401 /// NM_14480 NM_008546 // Mfap2 // microfibrillar-associated protein 2 Mfap2 0.012 -1.24 // 4 D3-E1 // 17150 // NR_024097 // OTTMUSG00000015563 // predicted gene, OTTMUSG00000 0.009 -1.24 OTTMUSG00000015563 // 2 F1|2 015563 NM_175437 // Pion // pigeon homolog (Drosophila) // 5 Pion 0.006 -1.25 A3 // 212167 /// ENSMUST00 NM_028195 // Cyth4 // cytohesin 4 // 15 E1 // 72318 /// Cyth4 0.045 -1.25 ENSMUST00000043069 // Cy NM_001014974 // Ttll4 // tubulin tyrosine ligase-like Ttll4 0.037 -1.25 family, member 4 // 1 C3 / NM_030675 // Krit1 // KRIT1, ankyrin repeat containing // Krit1 0.010 -1.25 5 A1|5 1.1 cM // 79264 NM_024289 // Osbpl5 // oxysterol binding protein-like 5 Osbpl5 0.049 -1.25 // 7 F5|7 69.59 cM // 79 NM_178617 // Necab1 // N-terminal EF-hand calcium Necab1 0.047 -1.25 binding protein 1 // 4 A2 // 6 NM_001113545 // Lima1 // LIM domain and actin binding Lima1 0.044 -1.25 1 // 15 F1|15 60.4 cM // 6 NM_182839 // Tppp // tubulin polymerization promoting Tppp 0.038 -1.25 protein // 13 C1 // 72948

267

NM_001081214 // Pprc1 // peroxisome proliferative Pprc1 0.031 -1.25 activated receptor, gamma, coa NM_011849 // Nek4 // NIMA (never in mitosis gene a)- Nek4 0.037 -1.25 related expressed kinase 4 / BC048942 // D11Wsu47e // DNA segment, Chr 11, Wayne D11Wsu47e 0.004 -1.25 State University 47, express NM_001038700 // Fnbp1 // formin binding protein 1 // 2 B Fnbp1 0.050 -1.25 // 14269 /// NM_019406 NM_008952 // Pipox // pipecolic acid oxidase // 11 B5|11 Pipox 0.030 -1.25 44.83 cM // 19193 /// E NM_175482 // Usp28 // ubiquitin specific peptidase 28 // Usp28 0.002 -1.25 9 A5.3 // 235323 /// EN NM_009783 // Cacna1g // calcium channel, voltage- Cacna1g 0.031 -1.25 dependent, T type, alpha 1G sub NM_175247 // Zfp28 // zinc finger protein 28 // 7 A1 // Zfp28 0.045 -1.25 22690 /// ENSMUST0000008 NM_175556 // Plch2 // phospholipase C, eta 2 // 4 E2 // Plch2 0.025 -1.25 269615 /// NM_001113360 NM_146188 // Kctd15 // potassium channel Kctd15 0.031 -1.25 tetramerisation domain containing 15 // NM_001038699 // Fn3k // fructosamine 3 kinase // 11 E2 Fn3k 0.022 -1.26 // 63828 /// NM_022014 // NM_177460 // Parp16 // poly (ADP-ribose) polymerase Parp16 0.023 -1.26 family, member 16 // 9 C // NM_001082960 // Itgam // integrin alpha M // 7 F4 // Itgam 0.028 -1.26 16409 /// NM_008401 // Itga NM_172621 // Clic5 // chloride intracellular channel 5 // Clic5 0.010 -1.26 17 C // 224796 /// ENS NM_029756 // Sdccag8 // serologically defined colon Sdccag8 0.017 -1.26 cancer antigen 8 // 1 H3 // NM_207706 // Elmo2 // engulfment and cell motility 2, Elmo2 0.028 -1.26 ced-12 homolog (C. elegans NM_029949 // Snapc3 // small nuclear RNA activating Snapc3 0.027 -1.26 complex, polypeptide 3 // 4 NM_172743 // Plekha7 // pleckstrin homology domain Plekha7 0.025 -1.26 containing, family A member 7 NM_030210 // Aacs // acetoacetyl-CoA synthetase // 5 F Aacs 0.020 -1.26 // 78894 /// ENSMUST00000 NM_012017 // Zfp346 // zinc finger protein 346 // --- // Zfp346 0.019 -1.26 26919 /// ENSMUST000000 NM_007714 // Clk4 // CDC like kinase 4 // 11 B1.3 // Clk4 0.040 -1.27 12750 /// ENSMUST0000009313 NM_023162 // Znrd1 // zinc ribbon domain containing, 1 Znrd1 0.013 -1.27 // 17 B3 // 66136 /// ENS NM_013768 // Prmt5 // protein arginine N- Prmt5 0.031 -1.27 methyltransferase 5 // --- // 27374 /// NM_174847 // C2cd2 // C2 calcium-dependent domain C2cd2 0.028 -1.27 containing 2 // 16 C4 // 20778 NM_178735 // A730069N07Rik // RIKEN cDNA A730069N07Rik 0.045 -1.27 A730069N07 gene // 8 A4 // 244425 /// E NM_172920 // Dpy19l1 // dpy-19-like 1 (C. elegans) // 9 Dpy19l1 0.017 -1.27 A4 // 244745 /// BC11678 NM_010851 // Myd88 // myeloid differentiation primary Myd88 0.038 -1.27 response gene 88 // 9 F3|9 NM_028815 // Cep97 // centrosomal protein 97 // 16 C1.1 Cep97 0.002 -1.27

268

// 74201 /// NM_00115936 NM_013925 // Adat1 // adenosine deaminase, tRNA- Adat1 0.001 -1.28 specific 1 // 8 D3 // 30947 /// NM_001037841 // Cklf // chemokine-like factor // 8 D3 // Cklf 0.044 -1.28 75458 /// NM_029295 // NM_010023 // Dci // dodecenoyl-Coenzyme A delta Dci 0.024 -1.28 isomerase (3,2 trans-enoyl-Coeny NM_025664 // Snx9 // sorting nexin 9 // 17 A1|17 3.3 cM Snx9 0.012 -1.28 // 66616 /// ENSMUST0000 NM_019963 // Stat2 // signal transducer and activator of Stat2 0.031 -1.28 transcription 2 // 10 D NM_030678 // Gys1 // glycogen synthase 1, muscle // 7 Gys1 0.023 -1.28 B4|7 23.0 cM // 14936 /// NM_001024918 // Rfx4 // regulatory factor X, 4 (influences Rfx4 0.046 -1.28 HLA class II expressi NM_153537 // Phldb1 // pleckstrin homology-like domain, Phldb1 0.045 -1.28 family B, member 1 // 9 NM_010808 // Mmp24 // matrix metallopeptidase 24 // 2 Mmp24 0.003 -1.29 H1|2 87.5 cM // 17391 /// NM_027106 // Avpi1 // arginine vasopressin-induced 1 // Avpi1 0.041 -1.29 19 C3 // 69534 /// ENSMU NM_027093 // 2310003L22Rik // RIKEN cDNA 2310003L22 2310003L22Rik 0.032 -1.29 gene // 2 G3 // 69487 /// EN NM_011981 // Zfp260 // zinc finger protein 260 // 7 B1 // Zfp260 0.029 -1.29 26466 /// ENSMUST00000 NM_172151 // Zdhhc8 // zinc finger, DHHC domain Zdhhc8 0.006 -1.29 containing 8 // 16 A3|16 10.83 c NM_011629 // Nr2c1 // nuclear receptor subfamily 2, Nr2c1 0.022 -1.29 group C, member 1 // 10 C2|1 NM_019456 // Apbb1ip // amyloid beta (A4) precursor Apbb1ip 0.038 -1.29 protein-binding, family B, m NM_001037756 // Brms1l // breast cancer metastasis- Brms1l 0.038 -1.29 suppressor 1-like // 12 C1|12 NM_028048 // Slc25a35 // solute carrier family 25, Slc25a35 0.040 -1.29 member 35 // 11 B3 // 71998 / NM_021527 // Mkks // McKusick-Kaufman syndrome Mkks 0.027 -1.30 protein // 2 F3 // 59030 /// NM_0 NM_025755 // Pacrgl // PARK2 co-regulated-like // 5 B3 // Pacrgl 0.046 -1.30 66768 /// ENSMUST00000 NM_027984 // Epn3 // epsin 3 // 11 C // 71889 /// Epn3 0.029 -1.30 ENSMUST00000010224 // Epn3 // BC066002 // BC024479 // cDNA sequence BC024479 // 9 BC024479 0.036 -1.30 A4 // 235184 /// BC024479 // NM_007941 // Stx2 // syntaxin 2 // 5 G1.3|5 70.0 cM // Stx2 0.010 -1.30 13852 /// ENSMUST00000031 NM_009582 // Map3k12 // mitogen-activated protein Map3k12 0.019 -1.30 kinase kinase kinase 12 // 15 NM_172845 // Adamts4 // a disintegrin-like and Adamts4 0.026 -1.30 metallopeptidase (reprolysin type NM_010688 // Lasp1 // LIM and SH3 protein 1 // 11 C- Lasp1 0.022 -1.30 D|11 56.0 cM // 16796 /// EN NM_001008785 // Kbtbd8 // kelch repeat and BTB (POZ) Kbtbd8 0.024 -1.31 domain containing 8 // 6 D2 NM_133804 // Tmem132a // transmembrane protein Tmem132a 0.044 -1.31 132A // 19 A // 98170 /// ENSMUST

269

NM_012019 // Aifm1 // apoptosis-inducing factor, Aifm1 0.038 -1.31 mitochondrion-associated 1 // X NM_177049 // Jph4 // junctophilin 4 // 14 C3 // 319984 Jph4 0.019 -1.31 /// NM_001003829 // Jph4 NM_178253 // Klhdc1 // kelch domain containing 1 // 12 Klhdc1 0.022 -1.31 C2 // 271005 /// ENSMUST0 NM_029512 // Ttpal // tocopherol (alpha) transfer Ttpal 0.049 -1.31 protein-like // 2 H3 // 76080 NM_001143777 // Fam13c // family with sequence Fam13c 0.009 -1.31 similarity 13, member C // 10 B5. NM_011057 // Pdgfb // platelet derived growth factor, B Pdgfb 0.050 -1.31 polypeptide // 15 E|15 4 NM_030139 // Zfp449 // zinc finger protein 449 // X A4 // Zfp449 0.050 -1.31 78619 /// ENSMUST00000 NM_025368 // Josd2 // Josephin domain containing 2 // 7 Josd2 0.043 -1.31 B2 // 66124 /// ENSMUST0 NM_175490 // Gpr75 // G protein-coupled receptor 75 // Gpr75 0.006 -1.31 11 A4 // 237716 /// ENSMU NM_026551 // Dcakd // dephospho-CoA kinase domain Dcakd 0.010 -1.31 containing // 11 E1 // 68087 / NM_172581 // Fam161b // family with sequence similarity Fam161b 0.001 -1.31 161, member B // 12 D1 / NM_008902 // Pp11r // placental protein 11 related // 15 Pp11r 0.049 -1.31 F1|15 56.8 cM // 19011 NM_026962 // Kbtbd3 // kelch repeat and BTB (POZ) Kbtbd3 0.048 -1.31 domain containing 3 // 9 A1 // NM_025690 // Sltm // SAFB-like, transcription modulator Sltm 0.021 -1.31 // 9 D // 66660 /// NM_0 NM_025429 // Serpinb1a // serine (or cysteine) peptidase Serpinb1a 0.031 -1.31 inhibitor, clade B, mem NM_172465 // Zdhhc9 // zinc finger, DHHC domain Zdhhc9 0.005 -1.32 containing 9 // X A4 // 208884 / NM_008883 // Plxna3 // plexin A3 // X B-C1|X 29.86 cM // Plxna3 0.028 -1.32 18846 /// ENSMUST000000 NM_023608 // Gdpd2 // glycerophosphodiester Gdpd2 0.010 -1.32 phosphodiesterase domain containing NM_007783 // Csk // c-src tyrosine kinase // 9 B|9 32.0 cM Csk 0.010 -1.32 // 12988 /// ENSMUST0 NM_054071 // Fgfrl1 // fibroblast growth factor receptor- Fgfrl1 0.004 -1.32 like 1 // --- // 116701 NM_011803 // Klf6 // Kruppel-like factor 6 // 13 A1 // Klf6 0.016 -1.32 23849 /// ENSMUST00000000 NM_018791 // Zfp108 // zinc finger protein 108 // 7 A3 // Zfp108 0.026 -1.32 54678 /// NM_019941 // NM_173868 // St18 // suppression of tumorigenicity 18 // St18 0.012 -1.32 1 A1 // 240690 /// ENSM NM_028036 // Tmco6 // transmembrane and coiled-coil Tmco6 0.004 -1.32 domains 6 // 18 B3 // 71983 NM_009866 // Cdh11 // cadherin 11 // 8 D2|8 46.5 cM // Cdh11 0.022 -1.32 12552 /// ENSMUST00000075 NM_026844 // 2310061C15Rik // RIKEN cDNA 2310061C15 2310061C15Rik 0.039 -1.33 gene // 8 E1 // 66531 /// EN NM_029581 // Mtif3 // mitochondrial translational Mtif3 0.008 -1.33 initiation factor 3 // 5 G3 // NM_001042634 // Clk1 // CDC-like kinase 1 // 1 C1.3|1 Clk1 0.039 -1.33

270

30.1 cM // 12747 /// NM_00 NM_175128 // 4930430F08Rik // RIKEN cDNA 4930430F08 4930430F08Rik 0.043 -1.33 gene // 10 D1 // 68281 /// E NM_145853 // Tpcn1 // two pore channel 1 // 5 F // Tpcn1 0.007 -1.33 252972 /// ENSMUST00000046426 NM_028457 // Dem1 // defects in morphology 1 homolog Dem1 0.031 -1.33 (S. cerevisiae) // 4 D2.2 / NM_001037707 // Zfp27 // zinc finger protein 27 // 7 B1 // Zfp27 0.005 -1.33 22689 /// NM_011754 / NM_172795 // Sarm1 // sterile alpha and HEAT/Armadillo Sarm1 0.012 -1.34 motif containing 1 // 11 NM_173749 // Pamr1 // peptidase domain containing Pamr1 0.038 -1.34 associated with muscle regener NM_199025 // Zbtb26 // zinc finger and BTB domain Zbtb26 0.035 -1.34 containing 26 // 2 B // 320633 NM_198649 // Ablim3 // actin binding LIM protein family, Ablim3 0.019 -1.34 member 3 // 18 E1 // 31 NM_026433 // Tmem100 // transmembrane protein 100 Tmem100 0.033 -1.34 // 11 C // 67888 /// ENSMUST00 NM_011145 // Ppard // peroxisome proliferator activator Ppard 0.011 -1.34 receptor delta // 17 A3. NM_018738 // Igtp // interferon gamma induced GTPase Igtp 0.047 -1.34 // 11 B1.3|11 32.0 cM // 16 NM_011657 // Tulp3 // tubby-like protein 3 // 6 F3|6 62.5 Tulp3 0.041 -1.34 cM // 22158 /// ENSMUS NM_026574 // Ino80 // INO80 homolog (S. cerevisiae) // 2 Ino80 0.010 -1.34 E5 // 68142 /// ENSMUST NM_001008549 // BC043301 // cDNA sequence BC043301 BC043301 0.005 -1.34 // 7 B4 // 210104 /// ENSMUST NM_007435 // Abcd1 // ATP-binding cassette, sub-family Abcd1 0.021 -1.34 D (ALD), member 1 // X B| NM_139141 // Zfp192 // zinc finger protein 192 // 13 A3.1 Zfp192 0.044 -1.34 // 93681 /// ENSMUST00 NM_016845 // Acrbp // proacrosin binding protein // 6 F2 Acrbp 0.049 -1.35 // 54137 /// NM_0011273 NM_029357 // Pcdh1 // protocadherin 1 // 18 B3 // 75599 Pcdh1 0.046 -1.35 /// ENSMUST00000057185 / NM_026670 // Zmym1 // zinc finger, MYM domain Zmym1 0.027 -1.35 containing 1 // 4 D2.2 // 68310 // NM_010778 // Cd46 // CD46 antigen, complement Cd46 0.033 -1.35 regulatory protein // 1 H6|1 106.6 NM_001114312 // 4930506M07Rik // RIKEN cDNA 4930506M07Rik 0.028 -1.35 4930506M07 gene // 19 D3 // 71653 // BC016248 // Zfp71-rs1 // zinc finger protein 71, releated Zfp71-rs1 0.046 -1.35 sequence // 13 37.0 cM NM_028141 // Zfp661 // zinc finger protein 661 // 2 F1 // Zfp661 0.022 -1.35 72180 /// NM_001111029 NM_144873 // Uhrf2 // ubiquitin-like, containing PHD and Uhrf2 0.026 -1.35 RING finger domains 2 / NM_027797 // Tmem80 // transmembrane protein 80 // 7 Tmem80 0.010 -1.35 F5 // 71448 /// NM_00114195 NM_026812 // Hddc3 // HD domain containing 3 // 7 D3 // Hddc3 0.004 -1.35 68695 /// ENSMUST0000003 NM_181540 // Tm6sf2 // transmembrane 6 superfamily Tm6sf2 0.043 -1.35 member 2 // 8 B3.3 // 107770

271

NM_019747 // Zfp113 // zinc finger protein 113 // 5 G1 // Zfp113 0.022 -1.35 56314 /// ENSMUST00000 NM_183024 // Raver2 // ribonucleoprotein, PTB-binding 2 Raver2 0.017 -1.35 // 4 C6 // 242570 /// EN NM_001001792 // Zfp239 // zinc finger protein 239 // 6 Zfp239 0.048 -1.35 F1|6 52.8 cM // 22685 /// NM_013690 // Tek // endothelial-specific receptor Tek 0.026 -1.36 tyrosine kinase // 4 C5|4 43.6 NM_025794 // Etfdh // electron transferring flavoprotein, Etfdh 0.019 -1.36 dehydrogenase // 3 E3| ENSMUST00000064254 // Atf7 // activating transcription Atf7 0.022 -1.36 factor 7 // 15 F3 // 2239 NM_198176 // Fastkd5 // FAST kinase domains 5 // 2 F1 // Fastkd5 0.023 -1.36 380601 /// NM_001146084 NM_001004173 // Sgpp2 // sphingosine-1-phosphate Sgpp2 0.030 -1.36 phosphotase 2 // 1 C4 // 433323 NM_011808 // Ets1 // E26 avian leukemia oncogene 1, 5' Ets1 0.016 -1.36 domain // 9 A4|9 15.0 cM NM_021534 // Pxmp4 // peroxisomal membrane protein 4 Pxmp4 0.021 -1.36 // 2 H2 // 59038 /// ENSMUS NM_172538 // Vezt // vezatin, adherens junctions Vezt 0.043 -1.36 transmembrane protein // 10 C2 NM_178906 // AI593442 // expressed sequence AI593442 AI593442 0.029 -1.36 // 9 A5.3 // 330941 /// BC1 NM_201518 // Flrt2 // fibronectin leucine rich Flrt2 0.022 -1.36 transmembrane protein 2 // 12 E / NM_175106 // Tmem177 // transmembrane protein 177 Tmem177 0.026 -1.36 // 1 E2.3 // 66343 /// ENSMUST NM_153802 // Zfp128 // zinc finger protein 128 // 7 A1 // Zfp128 0.002 -1.36 243833 /// ENSMUST0000 NM_009170 // Shh // sonic hedgehog // 5 B1|5 16.0 cM // Shh 0.028 -1.36 20423 /// ENSMUST0000000 NM_010840 // Mthfr // 5,10-methylenetetrahydrofolate Mthfr 0.030 -1.37 reductase // 4 E2|4 76.4 cM BC002181 // 2310033P09Rik // RIKEN cDNA 2310033P09 2310033P09Rik 0.010 -1.37 gene // 11 B2 // 67862 /// EN NM_027030 // Dcps // decapping enzyme, scavenger // 9 Dcps 0.024 -1.37 A4 // 69305 /// ENSMUST000 NM_178577 // Tmem205 // transmembrane protein 205 Tmem205 0.011 -1.37 // 9 A3 // 235043 /// ENSMUST0 NM_001012325 // Zfp708 // zinc finger protein 708 // 13 Zfp708 0.010 -1.37 B3 // 432769 /// NM_0010 NM_008487 // Arhgef2 // rho/rac guanine nucleotide Arhgef2 0.028 -1.37 exchange factor (GEF) 2 // 3 NM_177716 // AI836003 // expressed sequence AI836003 AI836003 0.037 -1.37 // 15 F1 // 239650 /// ENSM NM_029331 // 1700019G17Rik // RIKEN cDNA 1700019G17Rik 0.036 -1.37 1700019G17 gene // 6 C3 // 75541 /// NM NM_008992 // Abcd4 // ATP-binding cassette, sub-family Abcd4 0.000 -1.38 D (ALD), member 4 // 12 D NM_013673 // Sp100 // nuclear antigen Sp100 // 1 C5|1 Sp100 0.019 -1.38 50.0 cM // 20684 /// ENSMU NM_019940 // Zfp111 // zinc finger protein 111 // 7 A3 // Zfp111 0.012 -1.38 56707 /// NM_018791 // NM_028387 // Macrod2 // MACRO domain containing 2 // Macrod2 0.047 -1.38

272

2 F3 // 72899 /// NM_001013 NM_009829 // Ccnd2 // cyclin D2 // 6 F3|6 61.1 cM // Ccnd2 0.007 -1.38 12444 /// ENSMUST0000000018 NM_053081 // Fancg // Fanconi anemia, complementation Fancg 0.000 -1.38 group G // 4 B1 // 60534 / NM_023061 // Mcam // melanoma cell adhesion molecule Mcam 0.049 -1.38 // 9 A5.1 // 84004 /// ENSM NM_133345 // Ing4 // inhibitor of growth family, member Ing4 0.014 -1.38 4 // 6 F2|6 59.3 cM // 2 NM_008715 // Ints6 // integrator complex subunit 6 // 14 Ints6 0.026 -1.38 C3 // 18130 /// ENSMUST NM_025739 // Rnf220 // ring finger protein 220 // 4 D1 // Rnf220 0.024 -1.38 66743 /// ENSMUST00000 NM_025910 // Mina // myc induced nuclear antigen // 16 Mina 0.036 -1.38 C1.3 // 67014 /// NM_1748 BC062127 // BC062127 // cDNA sequence BC062127 // 10 BC062127 0.041 -1.39 C1 // 331188 /// ENSMUST000 NM_011243 // Rarb // retinoic acid receptor, beta // 14 Rarb 0.015 -1.39 A1-A3 // 218772 /// NM_2 NM_007430 // Nr0b1 // nuclear receptor subfamily 0, Nr0b1 0.019 -1.39 group B, member 1 // X C1|X NM_145580 // Tmem149 // transmembrane protein 149 Tmem149 0.034 -1.39 // 7 B1 // 101883 /// ENSMUST0 NM_023543 // Chn2 // chimerin (chimaerin) 2 // 6 B3 // Chn2 0.007 -1.39 69993 /// ENSMUST00000067 NM_172371 // Slc16a13 // solute carrier family 16 Slc16a13 0.004 -1.39 (monocarboxylic acid transport NM_013598 // Kitl // kit ligand // 10 D1|10 57.0 cM // Kitl 0.024 -1.39 17311 /// ENSMUST00000105 NM_019511 // Ramp3 // receptor (calcitonin) activity Ramp3 0.036 -1.40 modifying protein 3 // 11 A NM_009062 // Rgs4 // regulator of G-protein signaling 4 // Rgs4 0.030 -1.40 1 H3|1 86.5 cM // 197 --- 0.038 -1.40 BC095996 // Tbrg3 // transforming growth factor beta Tbrg3 0.023 -1.40 regulated gene 3 // 15 E1 / NM_007789 // Ncan // neurocan // 8 B3.3|8 33.5 cM // Ncan 0.023 -1.40 13004 /// ENSMUST0000000241 ENSMUST00000019572 // 9630025I21Rik // RIKEN cDNA 9630025I21Rik 0.004 -1.40 9630025I21 gene // 13 B3 // 40 NM_172256 // Dync2li1 // dynein cytoplasmic 2 light Dync2li1 0.008 -1.40 intermediate chain 1 // 17 E NM_134006 // Rdh5 // retinol dehydrogenase 5 // 10 Rdh5 0.017 -1.40 D3|10 72.0 cM // 19682 /// EN NM_026546 // Fam173b // family with sequence similarity Fam173b 0.003 -1.40 173, member B // 15 B2 / NM_001007465 // Rffl // ring finger and FYVE like domain Rffl 0.012 -1.41 containing protein // 1 NM_183248 // Nkx6-2 // NK6 homeobox 2 // 7 F4|7 68.0 Nkx6-2 0.023 -1.41 cM // 14912 /// NM_00112736 NM_026837 // Tmem53 // transmembrane protein 53 // 4 Tmem53 0.029 -1.41 D1 // 68777 /// ENSMUST0000 NM_172479 // Slc38a5 // solute carrier family 38, member Slc38a5 0.002 -1.41 5 // X A1.1 // 209837 / NM_025718 // Dnase1l2 // deoxyribonuclease 1-like 2 // Dnase1l2 0.004 -1.41

273

17 A3.3 // 66705 /// ENSM NM_010509 // Ifnar2 // interferon (alpha and beta) Ifnar2 0.010 -1.41 receptor 2 // 16 C3.3|16 63.1 NM_178728 // Napepld // N-acyl Napepld 0.030 -1.42 phosphatidylethanolamine phospholipase D // 5 A3 NM_175127 // Fbxo28 // F-box protein 28 // 1 H5|1 98.7 Fbxo28 0.021 -1.42 cM // 67948 /// ENSMUST00 NM_021383 // Rqcd1 // rcd1 (required for cell Rqcd1 0.007 -1.42 differentiation) homolog 1 (S. pom NM_018757 // Nme6 // non-metastatic cells 6, protein Nme6 0.031 -1.42 expressed in (nucleoside-di BC059896 // 2410002F23Rik // RIKEN cDNA 2410002F23 2410002F23Rik 0.032 -1.42 gene // 7 B3 // 66976 /// BC0 NM_134138 // Psmg2 // proteasome (prosome, Psmg2 0.010 -1.43 macropain) assembly chaperone 2 // 18 NM_172410 // Nup93 // nucleoporin 93 // 8 C5 // 71805 Nup93 0.039 -1.43 /// ENSMUST00000079961 // NM_001005341 // Ypel2 // yippee-like 2 (Drosophila) // 11 Ypel2 0.031 -1.43 C // 77864 /// ENSMUST NM_001114333 // Grm1 // glutamate receptor, Grm1 0.026 -1.43 metabotropic 1 // 10 A2 // 14816 /// NM_019964 // Dnajb8 // DnaJ (Hsp40) homolog, subfamily Dnajb8 0.045 -1.43 B, member 8 // 6 D2|6 38. NM_172434 // Tnrc4 // trinucleotide repeat containing 4 Tnrc4 0.021 -1.43 // 3 F2.1 // 78784 /// E NM_008058 // Fzd8 // frizzled homolog 8 (Drosophila) // Fzd8 0.048 -1.43 18 A1|18 2.0 cM // 14370 NM_019721 // Mettl3 // methyltransferase like 3 // 14 C2 Mettl3 0.010 -1.43 // 56335 /// NM_023434 NM_016922 // Gal3st1 // galactose-3-O-sulfotransferase 1 Gal3st1 0.027 -1.43 // 11 A1 // 53897 /// E NM_023348 // Snap29 // synaptosomal-associated protein Snap29 0.010 -1.43 29 // 16 A3 // 67474 /// NM_026897 // Haghl // hydroxyacylglutathione hydrolase- Haghl 0.050 -1.43 like // 17 A3.3 // 68977 NM_001025163 // Zfp78 // zinc finger protein 78 // 7 A1|7 Zfp78 0.043 -1.44 8.0 cM // 330463 /// N NM_028099 // Dusp11 // dual specificity phosphatase 11 Dusp11 0.000 -1.44 (RNA/RNP complex 1-intera NM_052993 // C1galt1 // core 1 synthase, glycoprotein-N- C1galt1 0.037 -1.44 acetylgalactosamine 3-be NM_175446 // Zmat1 // zinc finger, matrin type 1 // X E3 Zmat1 0.037 -1.44 // 215693 /// ENSMUST00 NM_183261 // Nr2f2 // nuclear receptor subfamily 2, Nr2f2 0.037 -1.45 group F, member 2 // 7 D1|7 BC028319 // Ap3s2 // adaptor-related protein complex 3, Ap3s2 0.043 -1.45 sigma 2 subunit // 7 D3 BC008558 // 1700030K09Rik // RIKEN cDNA 1700030K09 1700030K09Rik 0.000 -1.45 gene // 8 B3.3 // 72254 /// E NM_001007575 // Zfp58 // zinc finger protein 58 // 13 B3 Zfp58 0.010 -1.45 // 238693 /// ENSMUST00 NM_172656 // Stradb // STE20-related kinase adaptor Stradb 0.006 -1.45 beta // 1 C1.3|1 34.0 cM // NM_144888 // Mavs // mitochondrial antiviral signaling Mavs 0.038 -1.46 protein // 2 F1 // 228607

274

NM_134114 // Sft2d1 // SFT2 domain containing 1 // 17 Sft2d1 0.004 -1.46 A1 // 106489 /// BC091770 NM_012035 // Trpc7 // transient receptor potential cation Trpc7 0.026 -1.46 channel, subfamily C, NM_170588 // Cpne1 // copine I // 2 H2 // 266692 /// Cpne1 0.005 -1.46 NM_170590 // Cpne1 // copin NM_053163 // Mrpl36 // mitochondrial ribosomal protein Mrpl36 0.003 -1.46 L36 // 13 C1 // 94066 /// NM_001024504 // Dcun1d2 // DCN1, defective in cullin Dcun1d2 0.031 -1.46 neddylation 1, domain conta NM_153546 // Mboat1 // membrane bound O- Mboat1 0.002 -1.47 acyltransferase domain containing 1 // 1 NM_016863 // Fkbp1b // FK506 binding protein 1b // 12 Fkbp1b 0.022 -1.47 A1.1 // 14226 /// ENSMUST0 NM_015732 // Axin2 // axin2 // 11 E1|11 70.0 cM // Axin2 0.010 -1.48 12006 /// ENSMUST00000052915 NM_145824 // Ranbp10 // RAN binding protein 10 // 8 D3 Ranbp10 0.017 -1.48 // 74334 /// ENSMUST00000 NM_178782 // Bcorl1 // BCL6 co-repressor-like 1 // X A4 // Bcorl1 0.001 -1.48 320376 /// ENSMUST000 NM_008609 // Mmp15 // matrix metallopeptidase 15 // 8 Mmp15 0.024 -1.48 D1|8 45.5 cM // 17388 /// NM_145568 // Krcc1 // lysine-rich coiled-coil 1 // 6 C1 // Krcc1 0.038 -1.48 57896 /// ENSMUST0000 NM_198657 // EG381438 // predicted gene, EG381438 // 3 EG381438 0.010 -1.48 B|3 // 381438 /// ENSMUST NM_025636 // 2310079N02Rik // RIKEN cDNA 2310079N02Rik 0.042 -1.48 2310079N02 gene // 8 E2 // 66566 /// BC NM_011945 // Map3k1 // mitogen-activated protein Map3k1 0.002 -1.49 kinase kinase kinase 1 // 13 D2 NM_172763 // Zfp809 // zinc finger protein 809 // 9 A3 // Zfp809 0.031 -1.49 235047 /// ENSMUST0000 NM_026362 // 5033414D02Rik // RIKEN cDNA 5033414D02Rik 0.031 -1.49 5033414D02 gene // 19 C1 // 67759 /// B NM_009987 // Cx3cr1 // chemokine (C-X3-C) receptor 1 // Cx3cr1 0.031 -1.49 9 F4 // 13051 /// BC0126 NM_145627 // Rbm10 // RNA binding motif protein 10 // Rbm10 0.010 -1.50 X A1.3 // 236732 /// ENSMU NM_001024539 // Shc2 // SHC (Src homology 2 domain Shc2 0.005 -1.50 containing) transforming prot NM_130448 // Pcdh18 // protocadherin 18 // 3 C // 73173 Pcdh18 0.002 -1.50 /// ENSMUST00000035931 / NM_011925 // Cd97 // CD97 antigen // 8 C2|8 38.0 cM // Cd97 0.010 -1.50 26364 /// ENSMUST00000075 NM_145633 // Ankrd27 // ankyrin repeat domain 27 (VPS9 Ankrd27 0.022 -1.50 domain) // 7 B2 // 245886 NM_011125 // Pltp // phospholipid transfer protein // 2 Pltp 0.015 -1.50 H3|2 93.0 cM // 18830 // NM_001142744 // 2610110G12Rik // RIKEN cDNA 2610110G12Rik 0.002 -1.50 2610110G12 gene // 17 B3 // 73242 // NM_178732 // Zfp324 // zinc finger protein 324 // 7 A1 // Zfp324 0.001 -1.51 243834 /// ENSMUST0000 NM_178061 // Mobkl2b // MOB1, Mps One Binder kinase Mobkl2b 0.022 -1.51 activator-like 2B (yeast) // NM_029280 // Mettl5 // methyltransferase like 5 // 2 C3 Mettl5 0.031 -1.52

275

// 75422 /// NM_009278 / NM_019756 // Tubd1 // tubulin, delta 1 // 11 C // 56427 Tubd1 0.001 -1.52 /// ENSMUST00000069503 / NM_030733 // Gpr63 // G protein-coupled receptor 63 // 4 Gpr63 0.007 -1.53 A3 // 81006 /// ENSMUST NM_026385 // Pllp // plasma membrane proteolipid // 8 Pllp 0.023 -1.54 C5 // 67801 /// ENSMUST000 NM_172392 // Zfp759 // zinc finger protein 759 // 13 Zfp759 0.006 -1.54 B3|13 39.9 cM // 268670 /// NM_026497 // Nudt12 // nudix (nucleoside diphosphate Nudt12 0.010 -1.55 linked moiety X)-type motif NM_148930 // Rbm5 // RNA binding motif protein 5 // 9 Rbm5 0.011 -1.55 F1 // 83486 /// ENSMUST000 NM_172794 // Zfp454 // zinc finger protein 454 // 11 B1.3 Zfp454 0.000 -1.55 // 237758 /// ENSMUST0 NM_017476 // Akap8l // A kinase (PRKA) anchor protein 8- Akap8l 0.006 -1.55 like // 17 B2 // 54194 / NM_023716 // Tubb2b // tubulin, beta 2B // 13 A4 // Tubb2b 0.004 -1.56 73710 /// NM_009450 // Tubb2 NM_028002 // Dus4l // dihydrouridine synthase 4-like (S. Dus4l 0.010 -1.56 cerevisiae) // 12 A3 // NM_009437 // Tst // thiosulfate sulfurtransferase, Tst 0.037 -1.57 mitochondrial // 15 E1|15 45. NM_011267 // Rgs16 // regulator of G-protein signaling 16 Rgs16 0.020 -1.57 // 1 G3|1 78.0 cM // 1 NM_144820 // Ccdc28a // coiled-coil domain containing Ccdc28a 0.005 -1.57 28A // 10 A3 // 215814 /// NM_172260 // Cep68 // centrosomal protein 68 // 11 Cep68 0.028 -1.57 A3.1|11 10.91 cM // 216543 // NM_183174 // Homez // homeodomain leucine zipper- Homez 0.029 -1.57 encoding gene // 14 C2-C3 // 23 NM_001142918 // Tcf7l2 // transcription factor 7-like 2, T- Tcf7l2 0.002 -1.58 cell specific, HMG-bo NM_001010941 // Gpr12 // G-protein coupled receptor 12 Gpr12 0.022 -1.58 // 5 G3 // 14738 /// NM_0 NM_016923 // Ly96 // lymphocyte antigen 96 // 1 A3 // Ly96 0.049 -1.58 17087 /// NM_001159711 // NM_011765 // Zfp97 // zinc finger protein 97 // 17 A3.1 // Zfp97 0.022 -1.58 22759 /// BC002059 // NM_009442 // Ttf1 // transcription termination factor, Ttf1 0.004 -1.59 RNA polymerase I // 2 A3 NM_175116 // P2ry5 // purinergic receptor P2Y, G-protein P2ry5 0.008 -1.59 coupled, 5 // 14 D3 // NM_026734 // Tmem126b // transmembrane protein Tmem126b 0.004 -1.59 126B // 7 E1 // 68472 /// ENSMUST NM_013761 // Srr // serine racemase // 11 B4 // 27364 /// Srr 0.008 -1.59 NM_177325 // Tsr1 // T NM_029770 // Unc5b // unc-5 homolog B (C. elegans) // Unc5b 0.004 -1.59 10 B4|10 32.0 cM // 107449 BC076634 // BC049807 // cDNA sequence BC049807 // 17 BC049807 0.008 -1.60 A3.2 // 381066 /// ENSMUST0 AK087889 // E330037G11Rik // RIKEN cDNA E330037G11 E330037G11Rik 0.039 -1.60 gene // 8 A1.1 // 102220 AK086466 // D930030O05Rik // RIKEN cDNA D930030O05 D930030O05Rik 0.050 -1.60 gene // 7 C // 320387

276

NM_010098 // Opn3 // 3 // 1 H3 // 13603 /// Opn3 0.017 -1.61 NM_021350 // Chml // choroider NM_026772 // Cdc42ep2 // CDC42 effector protein (Rho Cdc42ep2 0.020 -1.61 GTPase binding) 2 // 19 A|1 NM_146072 // Grik1 // glutamate receptor, ionotropic, Grik1 0.000 -1.61 kainate 1 // 16 C3.3|16 58 NM_178086 // Fa2h // fatty acid 2-hydroxylase // 8 E1 // Fa2h 0.047 -1.62 338521 /// ENSMUST00000 NM_008409 // Itm2a // integral membrane protein 2A // X Itm2a 0.016 -1.64 A2-A3 // 16431 /// ENSMU NM_145618 // Narg2 // NMDA receptor-regulated gene 2 Narg2 0.005 -1.64 // 9 C // 93697 /// ENSMUST NM_133931 // Pot1a // protection of telomeres 1A // 6 Pot1a 0.002 -1.65 A3.1 // 101185 /// ENSMUST NM_009620 // Adam4 // a disintegrin and Adam4 0.012 -1.65 metallopeptidase domain 4 // 12 D1 // 11 NM_011607 // Tnc // tenascin C // 4 C1|4 32.2 cM // Tnc 0.037 -1.65 21923 /// ENSMUST00000107379 NM_011449 // Spa17 // sperm autoantigenic protein 17 // Spa17 0.048 -1.65 9 B // 20686 /// ENSMUST NM_008608 // Mmp14 // matrix metallopeptidase 14 Mmp14 0.004 -1.66 (membrane-inserted) // 14 C2|14 NM_011844 // Mgll // monoglyceride lipase // 6 D1 // Mgll 0.008 -1.67 23945 /// ENSMUST0000008944 NM_133228 // Zfp87 // zinc finger protein 87 // 13 B3|13 Zfp87 0.031 -1.67 43.0 cM // 170763 /// N NM_146074 // Tfb1m // transcription factor B1, Tfb1m 0.028 -1.68 mitochondrial // 17 A1 // 224481 NM_025780 // Thap2 // THAP domain containing, Thap2 0.005 -1.68 apoptosis associated protein 2 // NM_176953 // Lig4 // ligase IV, DNA, ATP-dependent // 8 Lig4 0.003 -1.68 A1.1 // 319583 /// ENSMU NM_028808 // P2ry13 // purinergic receptor P2Y, G- P2ry13 0.022 -1.68 protein coupled 13 // 3 D // 7 NM_026527 // Chac2 // ChaC, cation transport regulator Chac2 0.022 -1.70 homolog 2 (E. coli) // 11 NM_178375 // Zswim3 // zinc finger, SWIM domain Zswim3 0.006 -1.70 containing 3 // 2 H3 // 67538 // BC107288 // Phxr4 // per-hexamer repeat gene 4 // 9 A1 Phxr4 0.007 -1.72 // 18689 /// X12806 // Ph NM_144853 // Cyyr1 // cysteine and tyrosine-rich protein Cyyr1 0.004 -1.72 1 // 16 C3.3|16 56.2 cM NM_183140 // Zfp691 // zinc finger protein 691 // 4 D2.1 Zfp691 0.014 -1.72 // 195522 /// NM_001145 NM_001001447 // Zscan22 // zinc finger and SCAN domain Zscan22 0.003 -1.72 containing 22 // 7 A1 // NM_025280 // Kin // antigenic determinant of rec-A Kin 0.010 -1.73 protein // 2 A1|2 A1-A3 // 16 NM_011674 // Ugt8a // UDP galactosyltransferase 8A // 3 Ugt8a 0.035 -1.73 E3-F1 // 22239 /// ENSMU AK019926 // 5330431K02Rik // RIKEN cDNA 5330431K02 5330431K02Rik 0.002 -1.73 gene // 13 D1 // 68189 BC049349 // BC049349 // cDNA sequence BC049349 // 8 BC049349 0.003 -1.75 B3.3 // 234413 /// BC065788 NM_016718 // Ninj2 // ninjurin 2 // 6 F1 // 29862 /// Ninj2 0.008 -1.75

277

ENSMUST00000112711 // Ninj AK051152 // D130007C19Rik // RIKEN cDNA D130007C19 D130007C19Rik 0.008 -1.82 gene // 4 D2.2 // 442805 NM_176987 // 4732471D19Rik // RIKEN cDNA 4732471D19Rik 0.005 -1.82 4732471D19 gene // 13 B1 // 319719 /// NM_146179 // Zfp418 // zinc finger protein 418 // 7 A1 // Zfp418 0.033 -1.87 232854 /// NM_011860 / NM_001004190 // Zfp560 // zinc finger protein 560 // 9 A3 Zfp560 0.041 -1.94 // 434377 /// ENSMUST0 --- 0.010 -1.95 NM_145591 // BC003267 // cDNA sequence BC003267 // 8 BC003267 0.004 -1.96 A1.1 // 233987 /// ENSMUST0 AK037594 // A130014H13Rik // RIKEN cDNA A130014H13 A130014H13Rik 0.010 -1.98 gene // 12 F1 // 319630 /// A NM_001025381 // Gpr17 // G protein-coupled receptor 17 Gpr17 0.010 -1.98 // 18 B1|18 // 574402 /// ENSMUST00000056274 // C230013L11Rik // RIKEN cDNA C230013L11Rik 0.005 -2.01 C230013L11 gene // 17 A3.3 // NM_178706 // Siglech // sialic acid binding Ig-like lectin H Siglech 0.029 -2.07 // 7 B5 // 233274 / NM_009647 // Ak3l1 // adenylate kinase 3-like 1 // 4 C6|4 Ak3l1 0.014 -2.11 47.6 cM // 11639 /// D NM_026332 // Dnajc19 // DnaJ (Hsp40) homolog, Dnajc19 0.024 -2.15 subfamily C, member 19 // 3 F1|3 / NM_021272 // Fabp7 // fatty acid binding protein 7, brain Fabp7 0.002 -2.31 // 10 B4 // 12140 /// NM_053109 // Clec2d // C-type lectin domain family 2, Clec2d 0.003 -2.42 member d // 6 F3 // 93694 NM_178875 // 8430426H19Rik // RIKEN cDNA 8430426H19Rik 0.044 -2.69 8430426H19 gene // 13 B3 // 71508 /// N NM_001079695 // Sfrs5 // splicing factor, arginine/serine- Sfrs5 0.001 -2.87 rich 5 (SRp40, HRS) // AK032243 // 9530019H20Rik // RIKEN cDNA 9530019H20 9530019H20Rik 0.003 -2.90 gene // 8 B3.3 // 320999

278

APPENDIX F

Thermogenesis in Gpr50-/- and WT mice fasted at 15 °C.

Representative recordings of body temperature in Gpr50-/- (red line) and WT (blue line) mice subjected to a 24 h fast in which Gpr50-/- mice entered a state of torpor (n=4/group).

279

APPENDIX G

Functional clustering of genes with mRNA differences between fasted WT and Gpr50-/- mice. mRNAs extracted from hypothalamic blocks isolated from fed and fasted WT and Gpr50-/- mice were subject to microarray analysis. Significantly altered transcripts (p<0.05) between genotypes with fasting were analysed for GO and KEGG pathway functional enrichment. Groupings with fold enrichment >2, n>3 and significance (FDR=False Discovery Rate) p<0.01 are shown, groups of similar functional clustering were excluded. Category Genes Fold FDR p Value Enrichment Transcriptional TAF1D, GM13886, ZFP113, TNFRSF4, TCF7L2, SUFU, ZFP111, ZFP454, 2.03 0.03 1.71E-05 Regulation GM12966, SFRS5, ZSCAN22, NR1D1, TARDBP, HOMEZ, RQCD1, LEO1, PER1, OLIG2, PDE8A, BHLHE41, CCNA2, TXNIP, LOC100044274, ZFP52, ZFP58, ZFP324, TTF1, GM3203, 2210012G02RIK, ZFP28, ZFP128, ZFP27, HES1, GM13487, IFNAR2, FOXF1A, ZFP655, ASCC2, BCORL1, ID1, ZFP790, CSRNP1, NCOA5, LOC100047749 Transcriptional ZFP113, TNFRSF4, TCF7L2, SUFU, ZFP111, ZFP454, ZSCAN22, NR1D1, 2.34 0.03 2.18E-05 Regulation, HOMEZ, PER1, OLIG2, PDE8A, BHLHE41, TXNIP, ZFP52, DNA-dependent LOC100044274, ZFP324, ZFP58, TTF1, GM3203, 2210012G02RIK, ZFP28, ZFP128, ZFP27, HES1, IFNAR2, FOXF1A, ZFP655, ID1, CSRNP1, ZFP790, LOC100047749 Circadian NR1D1, PER1, BHLHE41 31.53 3.68 0.0037 Rhythm DNA Binding ZMAT1, TAF1D, GM13886, GM10257, XRCC6, POT1A, GM6485, 1.78 3.95 0.0032 GM7227, INO80, GM8095, LOC677395, GM6749, TARDBP, HOMEZ, OLIG2, BHLHE41, GM7179, GM6186, LOC100047823, HIST1H1C, LOC676337, LOC100048476, TTF1, LOC100045490, LIG4, ZFP28, GM4938, HES1, GM6421, GM13487, HIST2H2BB, LOC640877, LOC100045290, GM14383, LOC674507, GM6132, GM4028, GM1986, H1FNT, GM3835, GM13529, KIN, TCF7L2, GM9511, ZSCAN22, GM12657, NR1D1, GM6128, GM7100, LOC630737, GM8029, GM7194, DHX9, HIST1H2BC, GM6817, ZFP58, HIST1H2BE, HIST1H2BG, LOC100044525, AKAP8L, GM12950, 2210012G02RIK, GM3203, GM7900, GM9014, GM2099, FOXF1A, CSRNP1, H3F3A, H3F3B, H3F3C, GM2198, GM12271 Negative TXNIP, HES1, ID1, PER1, TTF1, OLIG2, BHLHE41, TNFRSF4, ZFP128, 3.11 4.09 0.0027 Regulation of SUFU, EIF2B5 Macromolecule Biosynthetic Process Chromosomal GM10257, POT1A, GM7227, GM6485, GM8095, LOC677395, 3.93 4.56 0.0039 Part GM6749, GM6186, GM7179, LOC100047823, HIST1H1C, LOC676337, LOC100045490, GM4938, GM6421, HIST2H2BB, LOC640877, LOC100045290, SPAG5, GM14383, LOC674507, GM6132, GM4028, GM1986, H1FNT, GM13529, GM3835, GM9511, GM12657, GM7100, GM6128, NPM2, LOC630737, GM8029, GM7194, GM6817, HIST1H2BC, HIST1H2BE, HIST1H2BG, LOC100044525, WAPAL, GM12950, GM7900, GM9014, GM2099, H3F3A, H3F3B, H3F3C, GM2198, GM12271 Negative TXNIP, HES1, ID1, PER1, OLIG2, BHLHE41, ZFP128, SUFU 4.09 5.01 0.0033 Transcriptional Regulator From RNA Polymerase II Promoter Negative TXNIP, HES1, ID1, PER1, OLIG2, BHLHE41, TNFRSF4, ZFP128, SUFU 3.45 6.60 0.0044 Regulation of Transcription, DNA-dependent Negative TXNIP, HES1, ID1, PER1, OLIG2, BHLHE41, TNFRSF4, ZFP128, SUFU 3.43 6.85 0.0046 Regulation of RNA Metabolic Process

280

APPENDIX G

Torpor expression in second line of Gpr50-/- mice.

Gpr50-/- mice obtained from Organon Pharmaceuticals exhibit torpor upon fasting with similar temporal profile and depth as Gpr50-/- obtained from DeltaGen. Congeneic WT mice do not enter torpor upon 24hr fasting. Grey bar indicates period of fasting.

281