The Physiological Role of Glucocorticoid and Mineralocorticoid Receptor Activation in Zebrafish

University of Calgary PRISM: University of Calgary's Digital Repository

Graduate Studies The Vault: Electronic Theses and Dissertations

2019-06-25 The Physiological Role of Glucocorticoid and Mineralocorticoid Receptor Activation in Zebrafish

Faught, Leslie Erin

Faught, L. E. (2019). The Physiological Role of Glucocorticoid and Mineralocorticoid Receptor Activation in Zebrafish (Unpublished doctoral thesis). University of Calgary, Calgary, AB. http://hdl.handle.net/1880/110575 doctoral thesis

University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca UNIVERSITY OF CALGARY

The Physiological Role of Glucocorticoid and Mineralocorticoid Receptor Activation in

Zebrafish

Leslie Erin Faught

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

GRADUATE PROGRAM IN BIOLOGICAL SCIENCES

CALGARY, ALBERTA

JUNE, 2019

Abstract

Glucocorticoids are key mediators of the vertebrate stress response. In teleosts, the primary glucocorticoid, cortisol, is a ligand for two corticosteroid receptors (CRs), the glucocorticoid receptor (GR) and the mineralocorticoid receptor (MR). The affinity of cortisol for these receptors is markedly different, with MR having an almost 10-fold higher affinity for the ligand compared to GR. This led to the hypothesis 30 years ago in mammals, that MR is responsible for basal cortisol function, while GR is active only when cortisol levels are high. In zebrafish (Danio rerio), the role of cortisol-GR signalling during stress is well characterized and is primarily involved in energy substrate mobilization to cope with stress. However, despite the persistence of MR in vertebrate evolution, there is no known physiological role for MR in ray-finned fishes. Ubiquitous knockout of MR (MRKO) and GR (GRKO) in mammals results in death, in utero or postnatally, due to dehydration or delayed lung maturation, respectively. This makes zebrafish an attractive model to study the physiological impacts of GR and MR activation at the systems level. Here we tested the hypothesis that both MR and GR activation are necessary for mediating the effects of cortisol in zebrafish. During embryogenesis (0-5 days post-fertilization [dpf]), the presence of both GR and MR are necessary for the activation of the hypothalamus-pituitary- interrenal axis, and stress-related behaviour. When treated with cortisol, postnatally (5-15 dpf) larvae at 15 dpf are significantly smaller compared to the wildype, and this response was abolished in both the GRKO and MRKO zebrafish larvae. MR activation is necessary for lipid accumulation, while GR activation is required for lipolysis. Both MR and GR activation are also required for growth hormone/insulin-like growth factor 1 axis by GR and MR. However, only GR activation results in an increase in transcript abundance of proteolytic genes. We then tested the hypothesis that under basal cortisol conditions, a loss of GR would have a substantial impact on muscle growth. Indeed, adult GRKO fish had larger body mass, more total muscle protein, increased phosphorylation of eIF4B (protein translation), and a decrease in genes involved in protein catabolism. Taken together this thesis highlights the dichotomy of GR and MR receptor activation; MR acting to promote anabolic processes, while GR is a potent initiator of catabolism during stress. Overall, this thesis establishes a physiological role for MR in ray-finned fishes and indicates that the mode of action may involve either MR activation alone and/or an interaction with GR in modulating metabolism during stress.

Preface

Chapter 1: Portions of the introductory text are reprinted with permission from: [Faught, E., Aluru, N. and Vijayan, M.M. 2016. The Molecular Stress Response. Fish Physiology vol 35: Biology of Stress in Fish. Edited by C.B. Schreck, L. Tort, Farrell, A. and Brauner, C.J. Elsevier, New York. Pp 113-166] [Faught E, Vijayan MM. 2018 Maternal stress and fish reproduction: The role of cortisol revisited. Fish Fisheries 19, 1016–1030] and [Faught, E., Hernandez-Perez, J., Wilson, J.M. and Vijayan, M.M. Stress in Response to Environmental Changes. Climate Change and Non-infectious Fish disorders. Edited by PTK Woo, and GK Iwama. CAB International – in press]

Chapter 3: This chapter is reprinted with permission from [Faught E., and Vijayan M.M. 2018. The mineralocorticoid receptor is essential for stress axis regulation. Scientific Reports 8, Article number: 18081]

Chapter 4: This chapter has been accepted in the Journal of Physiology and is currently in press [Faught E., and Vijayan M.M. 2019. Postnatal triglyceride accumulation is regulated by the mineralocorticoid receptor activation under basal and stress conditions. Journal of Physiology - 10.1113/JP278088]

Chapter 6: This chapter is reprinted with permission from [Faught, E., and Vijayan M.M. 2019. Loss of the glucocorticoid receptor in zebrafish improves muscle glucose availability and increases growth. American Journal of Physiology: Endocrinology and Metabolism, 316(6): E1093-1104].

Acknowledgements

First and foremost, I would like to thank my supervisor Dr. Matt Vijayan. After a decade in his lab, I never fail to be impressed with his creativity, enthusiasm, and drive to produce the best work possible. I will always be grateful for the time he took to teach me.

I would also like to thank my committee members, Dr. Joe Harrison and Dr. Doug Muench, for their patience and support. A further thank you to my external examiners, Dr. Dickmeis and Dr.

Chelikani for taking the time to critically review this thesis.

I would next like to acknowledge my lab members both past and present. In particular, Dr. Oana

Birceanu and Andrew Thompson who have always been a source of encouragement and friendship. Other members, Chinmayee Das, Marwa Thraya, Carol Best, Analisa Lazaro-Cote and Dr. Patrick Gauthier have also been an important source of support in the lab.

I would also like to thank Sophia George, Warren Fitch and Rob Hampton for their help with the many administrative tasks that made the lab set-up, function and fish care manageable.

Finally, I would like to thank my family for their patience and love, and to Penny, my constant companion.

Dedication

To my grandfather, Douglas James Gerrard.

Table of Contents

Abstract ...... 1 Preface ...... 2 Acknowledgements ...... 3 Dedication ...... 4 Table of Contents ...... 5 List of Tables ...... 8 List of Figures and Illustrations ...... 9 List of key Symbols, Abbreviations and Nomenclature ...... 11

CHAPTER 1: GENERAL INTRODUCTION ...... 13 Introduction ...... 14 The HPI axis: ...... 15 Cortisol signalling ...... 18 Glucocorticoid Receptor ...... 18 Mineralocorticoid Receptor ...... 20 Metabolic adjustments during stress ...... 22 Glucose as fuel: ...... 23 Energy substrates for oxidation and gluconeogenesis: ...... 26 Development of the HPI axis ...... 30 Glucocorticoid and Mineralocorticoid Morphants and Mutants: ...... 33 Hypothesis and Objectives: ...... 36

CHAPTER 2: GENERATING AND GENOTYPING GR AND MR KNOCKOUTS IN ZEBRAFISH WITH CRISPR/CAS9 ...... 37 Introduction: ...... 38 Materials and Methods: ...... 40 Zebrafish maintenance ...... 40 Generation of nr3c1 and nr3c2 mutants- Step-by-Step Methods: ...... 41 Design of CRISPR Target and Genotyping Primers: ...... 42 Synthesis of sgRNA ...... 43 Synthesis of Cas9 RNA ...... 43 Injections of sgRNA and Cas9: ...... 44 Checking Activity of your sgRNA: ...... 44 Founder screening: ...... 45 Genotyping of F1 ...... 45 Results: ...... 47 Glucocorticoid Receptor: ...... 47 Mineralocorticoid Receptor: ...... 51 Discussion: ...... 55

CHAPTER 3: THE MINERALOCORTICOID RECEPTOR IS ESSENTIAL FOR STRESS AXIS REGULATION IN ZEBRAFISH LARVAE ...... 57 Introduction ...... 58

Materials and Methods ...... 59 Zebrafish maintenance ...... 59 Generation of nr3c1 and nr3c2 null zebrafish ...... 60 nr3c1 functional experiment: ...... 62 Stress Experiment: ...... 62 Cortisol quantification ...... 63 Western blotting ...... 63 Transcript abundance ...... 64 Behavioural Analysis ...... 64 Statistics ...... 64 Results ...... 65 Generating MR and GR knockouts in zebrafish ...... 65 MR larvae have normal cortisol levels, whereas GR larvae are hypercortisolemic .67 MR regulates genes involved in HPA axis activity during development ...... 67 GR and MR knockouts alter the developmental profiles ...... 68 GR and MR differentially regulate the glucocorticoid stress response ...... 69 Both GR and MR are required for stress-related larval behaviour ...... 72 Discussion ...... 74

CHAPTER 4: POSTNATAL TRIGLYCERIDE ACCUMULATION IS REGULATED BY MINERALOCORTICOID RECEPTOR ACTIVATION UNDER BASAL AND STRESS CONDITIONS ...... 78 Introduction ...... 79 Materials and Methods ...... 81 Zebrafish care and husbandry: ...... 81 MR or GR Knockout zebrafish: ...... 81 Cortisol treatment: ...... 81 Cortisol quantification: ...... 82 RNA-Seq: ...... 82 Gene Expression by quantitative real-time PCR (qPCR): ...... 82 Metabolites: ...... 83 Enzyme Activity: ...... 83 Statistical analyses: ...... 84 Results ...... 84 Corticosteroid receptor activation: ...... 84 MR activation increases lipid accumulation under basal conditions: ...... 87 MR has little impact on genes involved in TG and cholesterol biosynthesis: ...... 89 MR activation restricts lipoprotein lipase (lpl) transcript abundance: ...... 91 Larval transcriptome changes specific to MR and/or GR activation: ...... 93 Discussion ...... 95

CHAPTER 5: CORTISOL-MEDIATED POSTNATAL GROWTH SUPPRESSION IN ZEBRAFISH LARVAE REQUIRES MINERALOCORTICOID RECEPTOR ACTIVATION ...... 101 Introduction ...... 102

Materials and Methods: ...... 104 Cortisol Treatment and Larval Rearing: ...... 104 Protein quantification and western blotting: ...... 105 Transcript abundance: ...... 105 Statistics: ...... 106 Results ...... 107 Growth regulation by MR and GR signalling: ...... 107 GR activation stimulates proteolysis: ...... 109 Myogenin (myog) but not myostatin (mystnb) is regulated by GR: ...... 111 MR and GR regulates GH/insulin/IGF1 expression: ...... 111 Regulation leptin by GR and MR: ...... 114 Discussion: ...... 117

CHAPTER 6: LOSS OF THE GLUCOCORTICOID RECEPTOR IN ZEBRAFISH IMPROVES MUSCLE GLUCOSE AVAILABILITY AND INCREASES GROWTH ...... 122 Introduction ...... 123 Materials and Methods ...... 125 Zebrafish maintenance ...... 125 Body composition measurements ...... 126 Feeding Trials ...... 126 Stress Experiment ...... 126 Cortisol and glucose determination ...... 127 In Vivo 2-NB-deoxy-D-glucose uptake ...... 127 Immunodetection ...... 127 Fasting study ...... 128 Transcript Abundance: ...... 128 Enzyme Activities ...... 129 Muscle Metabolomics ...... 129 Statistics ...... 130 Results ...... 130 Loss of GR increases body mass and alters body composition ...... 130 GRKO zebrafish have a functional HPI axis ...... 131 Loss of GR increased glucose uptake in the muscle ...... 133 Loss of GR promotes protein synthesis ...... 136 Loss of GR attenuates protein catabolism ...... 136 Muscle metabolomics ...... 139 Discussion ...... 143

CHAPTER 7: GENERAL CONCLUSION ...... 149 WORKS CITED: ...... 157 APPENDIX I: SUPPLEMENTAL METHODS ...... 186 APPENDIX II: SUPPLEMENTAL INFORMATION (CHAPTER 4) ...... 199 APPENDIX III: COPYRIGHT PERMISSIONS ...... 212 7

List of Tables

Chapter 4

Table 1: Gene specific primers: ...... 83

Chapter 5

Table 1: Gene specific primers ...... 106

Chapter 6

Table 1: Primers used for quantitative real-time PCR...... 128

Table 2: Enzyme activities...... 142

Appendix I

Table 1: GR and MR mutants and the associated phenotypes: ...... 187

Table 2: Exon regions of nr3c1 gene (Source: ENSEMBL) ...... 192

Table 3: Exon regions of nr3c2 gene (Source: ENSEMBL) ...... 194

Appendix II

Table 1: Possible Targets of MR in 5dpf zebrafish: ...... 200

Table 2: Targets of GR in 5dpf zebrafish: ...... 202

Table 3: DEG in 5dpf GRKO zebrafish (MR activated genes): ...... 206

Table 4: DEG in 5dpf MRKO zebrafish...... 209

Table 5: Possible Targets of GR:MR heterodimers in 5dpf zebrafish...... 210

Table 6: Cortisol regulated gene(s) independent of GR and MR: ...... 211

Table 7: Compensatory cortisol action by MR in the absence of GR, or a third corticosteroid receptor: ...... 211

List of Figures and Illustrations

Chapter 1

Figure 1: HPI Axis activation and biosynthesis of cortisol ...... 17

Figure 2: Corticosteroid Receptor Signalling in Target Cells ...... 22

Figure 3: Metabolic Regulation during stress ...... 29

Figure 4: Effect of Cortisol on Zebrafish Development ...... 32

Chapter 2

Figure 1: Overview of methods to generate maternal-zygotic zebrafish mutants...... 41

Figure 2: Sequence Verification of GR mutants: ...... 48

Figure 3: Sequence analysis: ...... 49

Figure 4: Representative results for fragment analysis of fluorescent PCR...... 50

Figure 5: Sequence Verification of MR mutants: ...... 52

Figure 6: Sequence analysis: ...... 53

Figure 7: Representative results for fragment analysis of fluorescent PCR of MR mutants...... 54

Chapter 3

Figure 1: GR and MR knockout in zebrafish: ...... 66

Figure 2: GR and MR regulate genes involved in HPA axis activity during development: ...... 70

Figure 3: GR and MR are necessary for the glucocorticoid profile post-stress: ...... 71

Figure 4: GR and MR are necessary to mediate stress-related behaviour: ...... 73

Chapter 4

Figure 1: Corticosteroid receptor activation: ...... 86

Figure 2: MR is essential for lipid accumulation under basal conditions...... 88

Figure 3: MR has little effect on genes involved in lipid biosynthesis...... 90

Figure 4: MR modulates GR signalling of by increasing lipolysis stressed conditions...... 92

Figure 5: Larval transcriptome in response to GR and/or MR activation...... 94

Chapter 5

Figure 1. Regulation of larvae growth by MR and GR signalling: ...... 108

Figure 2. GR activation stimulates proteolysis: ...... 110

Figure 3. Myogenin (myog) but not myostatin (mystnb) is regulated by GR: ...... 112

Figure 4. MR and GR regulates GH/insulin/IGF1 expression: ...... 113

Figure 5. Regulation leptin by GR and MR: ...... 115

Figure 6 CR activation paradigm: ...... 116

Chapter 6

Figure 1: GRKO zebrafish are larger than WT, with a different body composition...... 131

Figure 2: GRKO have an impaired glucose response to stress: ...... 134

Figure 3: GRKO increased glucose clearance due to increased muscle uptake: ...... 135

Figure 5: GRKO fish have greater protein synthetic capacity ...... 138

Figure 6: GRKO fish have higher body mass and protein content post-fasting...... 140

Figure 7: Muscle Metabolomics: ...... 141

Appendix I

Figure 1: pT3TS-nCas9n plasmid...... 198

List of key Symbols, Abbreviations and Nomenclature

Symbol Definition ACC acetyl-CoA carboxylase ACTH adrenocorticotropic hormone ALT alanine aminotransferase AST aspartate aminotransferase ATP adenosine-triphosphate CR corticosteroid receptor CRH corticotropin releasing hormone CRISPR clustered regularly interspaced short palindromic repeats DPF days post-fertilization DSB double-stranded break FAS fatty acid synthase G6P glucose-6-phosphate GC glucocorticoid GH growth hormone GHR growth hormone receptor GLUT2 glucose transporter 2 GLUT4 glucose transporter 4 GR glucocorticoid receptor GP glycogen phosphorylase HDR homology-directed repair HK hexokinase HOAD 3-hydroxylacyl-CoA HPF hours post-fertilization HPI hypothalamus-pituitary-interrenal axis IGF1 insulin-like growth factor 1 LDH lactate dehydrogenase MC2R melanocortin-2-receptor MR mineralocorticoid receptor

MRAP melanocortin-2-receptor accessory protein MURF1 muscle RING-finger protein 1 NHEJ non-homologous end joining PAM protospacer adjacent motif PEPCK/pck1 phosphoenolpyruvate carboxykinase POMC pro-opiomelanocortin Redd1 regulated in development and DNA damage responses 1 StAR steroidogenic acute regulatory protein TALENs transcription activator-like nucleases TCA tricarboxylic acid cycle ZFN zinc-finger nucleases

CHAPTER 1: GENERAL INTRODUCTION

Introduction

The endocrine stress response and the subsequent physiological changes are essential for adaptation to stress in vertebrates. In response to a perceived stressor, the sympathetic nervous system releases catecholamines, epinephrine and norepinephrine, into the blood stream as part of the ‘flight or fight’ response (Fabbri & Moon, 2015). Catecholamines are released in seconds to minutes and results in the rapid release of stored glucose, increased heart rate, and muscle contraction (Charmandari et al., 2005; Fabbri & Moon, 2015). In contrast, activation of the hypothalamus-pituitary-interrenal (HPI) axis results in the release of glucocorticoids (GCs) (Wendelaar Bonga, 1997; Mommsen et al., 1999), which are steroid hormones, released both constitutively and in response to stress (Mommsen et al., 1999; Charmandari et al., 2005). The effect of elevated GCs on the physiology of fish has been extensively studied (reviewed recently by Mommsen et al., 1999; Vijayan et al., 2010; Faught & Vijayan, 2016). The attendant rise in GCs post-stressor mobilizes energy resources to help restore homeostasis; in particular, upregulation of hepatic gluconeogenesis, and increased muscle proteolysis are hallmarks of post- stressor GC action (Mommsen et al., 1999; Charmandari et al., 2005; Kuo et al., 2013). The constitutive roles of GCs are much less well defined and their basal levels are approximately 10x less than stress levels (Mommsen et al., 1999). Evidence that basal circulating GCs are necessary to maintain specific, physiological functions was reported in zebrafish (Danio rerio), where sequestering of cortisol in the yolk of zebrafish using antibodies caused a dysregulation of developmental programming events (Nesan & Vijayan, 2016). Cortisol is the primary GC in teleosts, and it is a ligand for two corticosteroid receptors (CRs), the mineralocorticoid receptor (MR) and the glucocorticoid receptor (GR)(Bury, 2003; Charmandari et al., 2005). These two receptors are functionally-related ligand-bound transcription factors, which are ubiquitously expressed but have vastly different affinity for cortisol (Baker et al., 2013). MR, also called the type I CR, has a 10-fold lower Kd value compared to GR, the type II CR (Baker & Katsu, 2017). It was postulated 30 years ago that the co-expression of these receptors in the mammalian brain have a functional significance; MR to maintain tonic neuronal excitability, and GR to mediate the effects of stress (de Kloet & Reul, 1987). Here we expanded this hypothesis and applied it at the systems level. Specifically, we tested whether MR is involved in basal, anabolic functions and if GR, as a mediator of stress,

will regulate catabolic action to mobilize energy substrates. The effects of GR are well defined in fish and mammals, but while MR mediates ion and fluid balance in mammals (Baker & Katsu, 2017), there has been no identified physiological function of MR in lower vertebrates. Additionally, while cortisol is a well-known stress hormone, the studies examining the basal role of cortisol in affecting fish physiology are few (Nesan & Vijayan, 2016). This thesis focuses on the fundamental physiological role of GR and MR activation using zebrafish as a model.

The HPI axis:

The HPI stress response is driven primarily by neural mechanisms but involves the interaction between three distinct organ systems, each with its own control points for GC production (Herman et al., 2012; Faught et al., 2019) (Fig. 1). The perception of a stressor will increase circulating corticotropin-releasing hormone (CRH), which is produced in the paraventricular nucleus (PVN) of the hypothalamus in mammals and in the preoptic area of teleost fish (Backström & Winberg, 2013). The activation of CRH is stressor dependent, with reactive responses being mediated by noradrenergic stimulation of the PVN neurons, whereas anticipatory responses involve limbic stimulation (Herman et al., 2016). In fish, CRH stimulation is less well defined and several neurotransmitters and neuromodulators, including monoamines are thought to regulate the perception of stressor leading to hypothalamic release of CRH (Winberg & Nilsson, 1993). Once released, CRH binds G-protein coupled CRH receptors (CRHR1), which will activate adenylate cyclase, increasing ACTH production (Swift, 1982; Doyon et al., 2006; Herman et al., 2016). ACTH is a post-translational product of proopiomelanocortin (POMC) synthesis, which is cleaved by prohormone convertase 1 into ACTH and β-lipotropin (Herman et al., 2016). CRH promotes transcription of pomca via cAMP/protein kinase A (Herman et al., 2016); however, this gene is also regulated by other hormones, including corticosteroids (Bumaschny et al., 2007). Indeed, glucocorticoid response elements (GREs) are present in the promoter region of pomca, and excess cortisol stimulation causes a reduction in the transcript abundance of pomca, thereby regulating the steroid levels through a negative feedback loop to re-establish basal levels (Bumaschny et al., 2007). Knockout of GR in zebrafish (Danio rerio) results in hyperactivity of the HPI axis and results in hypercortisolemic fish (Ziv et al., 2013). CRH knockout mice have established the necessity of CRH in maintaining

basal or increasing stress-induced ACTH release (Venihaki & Majzoub, 2002), and fish lacking corticotropes (rx3 mutant zebrafish), have low pomca expression and decreased cortisol levels (Dickmeis et al., 2007). This emphasizes the conserved function of the HPI axis in maintain basal cortisol levels. Much of what is known about the regulation of CRH and POMC comes from mammalian studies as both genetic manipulation and pharmacological antagonism in fish has been limited. Additionally, the scarcity of antibodies against species-specific CRH often precludes measuring hormone levels in fishes; however, CRH transcript levels are upregulated in response to stressors and correlate with elevated cortisol levels during stress (Bernier et al., 2004; Doyon et al., 2006; Nesan & Vijayan, 2016). ACTH, released from the pituitary gland into the circulation, stimulates the melanocortin- 2-receptor (MC2R) on the steroidogenic cells of the interrenal tissue located in the head kidney region of teleosts (Wendelaar Bongs, 1997; Mommsen et al., 1999) (Fig. 1). This is analogous to the zona fasciculata of the adrenal cortex in mammals (Mommsen et al., 1999). MC2R is also a G-protein coupled receptor and activates adenylate cyclase and increases intracellular cAMP levels. The subsequent phosphorylation/activation of cAMP-dependant protein kinase (PKA) stimulates CREB (cAMP response element-binding) to transcribe steroidogenic acute regulatory protein (StAR) (Stocco, 2001; De Joussineau et al., 2012). Cholesterol is the precursor for steroids, and StAR facilitates the movement of cholesterol across the outer mitochondrial membrane to the inner mitochondrial membrane (Stocco, 2001), and is a rate-limiting step in steroid hormone biosynthesis (Herman et al., 2016). In humans, the promoter region of StAR has two cis-elements that are responsible for basal and cAMP-regulated gene expression by steroidogenic factor 1 (SF1) (Sugawara et al., 1997). In teleosts, higher mRNA levels of star are correlated with increased stress in fish (Geslin & Auperin, 2004). Expression of this gene transcript is also increased in GR knockout zebrafish, which are hypercortisolemic (Ziv et al., 2013; Faught & Vijayan, 2018a), suggesting that GR may have some negative regulatory control on star transcription. Indeed, there are 4 GREs located 1000 base pairs (bp) upstream of star in zebrafish, and in particular, the tandem elements at -136 and -165 may be indicative of a negative GRE (Dostert & Heinzel, 2004). Activation of MC2R on the steroidogenic cells is also dependent on the accessory proteins, the melanocortin receptor accessory protein 1 (MRAP1) and 2 (MRAP2) (Cooray & Clark, 2011). In particular, MRAP1 seems to be the most critical for

MC2R activation in teleosts and is involved in the recruitment of MC2R to the membrane (Dores et al., 2014). However, MRAP2 was also shown to be an important component for the functional expression of MC2R in sea bass (Dicentrarchus labrax) (Agulleiro et al., 2010; Cerdá-Reverter et al., 2012). The regulation of the HPI axis occurs through various mechanisms in addition to negative feedback by GCs, including modulation of adrenoceptors of the SNC (Reid et al., 1998), autoregulation of GR (Sathiyaa & Vijayan, 2003), and modulation of 11-β hydroxysteroid dehydrogenase 2 (11-βHSD2) (Alderman & Vijayan, 2012; Best & Vijayan, 2018; Faught & Vijayan, 2018a). Interestingly, fish that are deficient in GR are hypercortisolemic, which is thought to be not only due to a hyperactivity of the HPI axis (Ziv et al., 2013), but also through a lack of cortisol metabolism (Chapter 3, (Faught & Vijayan, 2018a)). This was also supported in fish lacking the MR, which had normal cortisol levels (Chapters 2 and 3), indicating that in addition to HPI activation, the metabolism of cortisol may also contribute to re-establishing plasma cortisol profile after an acute stress response (Chapter 3 (Faught & Vijayan, 2018a)).

Figure 1: HPI Axis activation and biosynthesis of cortisol Stress will cause release of CRH from the hypothalamus, which stimulates the corticotropes of the pituitary to release ACTH. Binding of ACTH to MC2R on the interrenal cells will stimulate production of cortisol from cholesterol. Cortisol will act via GR to reduce production of HPI-intermediates to restrict its own production (negative feedback loop). Abbreviations: glucocorticoid receptor (GR), paraventricular nucleus (PVN), corticotropin releasing hormone (CRH), CRH receptor 1 (CRHR1), cyclic adenosine monophosphate (cAMP), pro-opiomelanocortin A (pomca), adrenocorticotropic hormone (ACTH), cAMP response element (CREB), melanocortin-2-receptor (MC2R), melanocortin-2-receptor accessory protein (MRAP), steroidogenic acute regulatory protein (StAR). Adapted from Faught et al., 2019.

Cortisol signalling

Glucocorticoid Receptor

GR is the primary receptor for GCs action in teleosts and was first cloned and sequenced in ray-finned fishes nearly 20 years ago in rainbow trout (Oncorhynchus mykiss) (Ducouret et al., 1995). This gene is a homolog to the human GR and is part of a superfamily of ligand-bound transcription factors, which includes the MR, the progesterone receptor (PR), and the androgen receptors (Funder, 2017). GR is encoded by the gene nr3c1 (Baker et al., 2013). The resulting protein is modular consisting of three functional domains, an N-terminal domain (NTD), a central DNA binding domain (DBD) and a C-terminal ligand binding domain (LBD), which contains a ligand-dependent activation function (AF1,2) (Baker et al., 2013). In the absence of a ligand, GR is localized in the cytoplasm and is part of a heterocomplex, which includes the heat shock proteins 40, 70, 90, and p23 (Pratt et al., 2006) (Fig. 2). In steroid free cells, the CRs will shuttle continuously between the cytoplasm and the nucleus, and under basal conditions, unliganded GR resides in the cytoplasm, while MR is seen both in the cytoplasm and the nucleus (Madan & DeFranco, 1993). This is thought to be because MR is ~80% bound and likely transcriptionally active at basal cortisol concentrations in mammals (de Kloet & Reul, 1987). Once cortisol binds, GR will dissociate due to conformational changes and dimerize with another ligand-bound GR and translocate to the nucleus. Little information is available on the GR heterocomplex in fish; however, the GRE in the promoter region of target genes is conserved. The GRE is a specific 15 bp, imperfect, palindromic DNA motif (AGAACA nnnn TGTTCT), and is highly conserved (Esbaugh & Walsh, 2009; Kuo et al., 2013). The GRE acts as an allosteric activator by providing a scaffold to bind GR in the correct position (Schoneveld et al., 2004). GR will also exert repression of gene transcription by modulating the transactivating properties of other transcription factors, including nuclear factor k-B (NF-kB) and/or activator protein AP-1 (Schoneveld et al., 2004). Few studies have characterized GR-mediated responses of cortisol in fish (Vijayan & Leatherland, 1992; Sathiyaa & Vijayan, 2003; Aluru et al., 2007; Esbaugh & Walsh, 2009; Alderman et al., 2012), and GR action in fish has been classically studied by using the pharmacological antagonist, RU486. However, this antagonist is also a PR antagonist, making its use in developmental studies unreliable (Best & Vijayan, 2018). Despite

these limitations, combining transcriptomics and bioinformatic analyses revealed a large set of genes which contain GREs and are transcriptionally regulated by cortisol-GR signalling in fishes (Aluru et al., 2007; Alderman et al., 2012; Chatzopoulou et al., 2015; Philip & Vijayan, 2015). Multiple orthologues of GR have been cloned and sequenced in a number of fish species, (Greenwood et al., 2003; Terova et al., 2005; Acerete et al., 2007; Filby & Tyler, 2007; Stolte et al., 2008; Arterbery et al., 2010), including zebrafish (Alsop & Vijayan, 2008). All teleosts studied to date have two paralogues of GR (GR1 and GR2), which are thought to have risen from the whole genome duplication ~350 million years ago (Bury, 2003; Vijayan et al., 2005; Prunet et al., 2006), with zebrafish being the exception. Zebrafish are unique among ray-finned fishes in having only a single GR gene in the genome (Schaaf et al., 2008; Alsop & Vijayan, 2008, 2009a). A splice variant of GR2, α and β, have been characterized in Haplochromis burtoni (Greenwood et al., 2003), and zebrafish (Schaaf et al., 2008), and this is similar to the GRα and β that exist in humans (Schaaf et al., 2008). GRβ is the result of divergent transcription of the 3’most exon, resulting in a truncated LBD (Chatzopoulou et al., 2015). Using GRα and GRβ splice-blocking morpholinos and subsequent treatment with the synthetic GC dexamethasone, two distinct gene clusters were activated under either basal or stressed conditions (Chatzopoulou et al., 2015). However, further work determined that there was no transcriptional role for GR-β in zebrafish (Chatzopoulou et al., 2017). In teleosts, where there are two GR genes, GR2 is more sensitive to lower concentrations of both cortisol and dexamethasone compared to GR1 (Bury, 2003). Although the GR antagonist mifepristone inhibited transactivation activity of both GR1 and GR2, GR1 was more sensitive to this antagonist (Bury, 2003). In spite of these differences in receptor affinities and ligand specificity, most work was carried out by in vitro reporter assays (Bury, 2003), and a functionally distinct role for the two GRs during stress has not been clearly established in fish (Prunet et al., 2006). Transcription of GR is also directly affected by elevated cortisol levels suggesting autoregulation in both fish (Sathiyaa & Vijayan, 2003) and mammals (Burnstein et al., 1991). Studies have shown that while gr mRNA levels are elevated, GR protein expression is downregulated in response to cortisol stimulation in rainbow trout (Sathiyaa & Vijayan, 2003; Vijayan et al., 2003). This protein downregulation was abolished upon blockage of the proteasome. Therefore, it is clear that cortisol not only regulates protein levels by transcription,

but also by increasing protein breakdown. Furthermore, the mismatch between GR transcript levels and GR protein expression in response to cortisol treatment was seen both in vivo in trout liver and in vitro in trout hepatocytes, suggesting autoregulation of GR as a key cellular adaptation to stress in fish (Sathiyaa & Vijayan, 2003; Vijayan et al., 2003). This autoregulation of GR may be an adaptive mechanism to restrict continuous GR activation, which may have wide-reaching effects due to its ubiquitous expression (Vijayan et al., 2010). Given the ubiquitous expression of these receptors and the necessity for their activation in mobilizing energy resources, it has become essential to study GR and MR signalling at a systems-level. This is not well characterized in either mammalian or piscine models. For example, while recent work in mice has suggested a muscle-liver-fat signalling axis being regulated by GR (Shimizu et al., 2015), these were conditional knockouts (muscle), where GR was still present in other tissues. Cell lines studies also suggest that the uptake of glucose can be mediated by GR in the skeletal muscle (Bernal-Sore et al., 2018), but the physiological significance has not been shown at a systems level. Overall, the physiological implications of GR activation on energy partitioning at the systems level are unknown.

Mineralocorticoid Receptor

The MR is part of the same nuclear receptor subfamily as GR, with a similar modular structure, but is encoded by the nr3c2 gene (Hu & Funder, 2006). In mammals, the primary mineralocorticoid hormone is aldosterone, and aldosterone-MR signalling is responsible for hydro-mineral balance in tetrapods (Baker et al., 2013). This hormone is absent from ray-finned fishes where it seems that the GR, and not the MR, is responsible for osmoregulation (Cruz et al., 2013). However, GCs binds to MR with nearly 10-fold higher affinity compared to GR binding, and this persists in the vertebrate lineage, suggesting a functional role (Baker et al., 2013; Baker & Katsu, 2017). However, there is no known physiological role for MR in ray- finned fishes (Baker & Katsu, 2017; Funder, 2017). From an evolutionary perspective, GCs signalling in primitive vertebrates (agnathans), occurs through a single CR and separate orthologues only appear in elasmobranchs (Baker et al., 2013). MR is widely expressed in fish tissues, and studies suggest the possibility that 11-deoxycorticosterone (DOC) may act as an MR agonist (Stolte et al., 2006; Milla et al., 2009; Pippal et al., 2011). Additionally, MR has been

cloned and sequenced in a number of fish species, (Greenwood et al., 2003; Stolte et al., 2006; Alsop & Vijayan, 2008), and most species seem to have a single MR (Alsop & Vijayan, 2008). MR can also bind several 3-ketosteroids, including cortisol, 11-deoxycortisol, corticosterone, DOC, and progesterone, which further complicates the determination of physiological function (Baker & Katsu, 2017). If we look to mammals, extra-mineralocorticoid functions have recently been associated with MR activation, including modulation of behaviour (Berger et al., 2006; Rozeboom et al., 2007) and lipid differentiation (Marzolla et al., 2012). Increased levels of MR in the forebrain of mice decreased anxiety-like behaviour, and a suppression of rise in GCs post-restraint stress (Rozeboom et al., 2007), whereas knockout of MR in the limbic system impaired learning (Berger et al., 2006). Overall, neuronal MR is necessary to maintain neuronal integrity and basal excitability and involved in the assessment of novel situations (reviewed by (Joëls et al., 2008; Joëls & de Kloet, 2017)). These results were consistent with the observation of de Kloet and Reul (1987) that neuronal MR was necessary for basal neuronal excitability, but activation of GR mediates stress-related behaviour (de Kloet & Reul, 1987). Interestingly, this is conserved in fish, as medaka (Oryzias latipes) lacking MR had altered behaviour to novel stimuli (Sakamoto et al., 2016). Transcriptional control of target genes by MR occurs in much the same way as GR, in that, upon ligand binding, it will dimerize and translocate to the nucleus. It can both bind to specific mineralocorticoid response elements (MREs) as well as GREs (Mifsud & Reul, 2016). The binding of MR to GREs at baseline is restricted, indicating a duel role for MR activation during basal and stress conditions (Mifsud & Reul, 2016). To further complicate the transcriptional control of GCs, recent work in both fish and mammals has reported that MR and GR can heterodimerize to activate transcription. Work by Mifsund and Ruel (2018) in mice, showed that not only was the heterodimerization much more prevalent than initially thought, but that depending on cortisol levels, the dimerization and the subsequent effect on transcription was gene specific (Mifsud & Reul, 2016). For example, MR and GR will bind to the same GRE sites upstream of fkbp5 and per1, but not sgk1. Other genes known to be modulated by heterodimerization include the 5-HT1A receptor, which is consistent with the changes observed in anxiety phenotypes in response to GR and MR modulation (Ou et al., 2001). A recent study in Atlantic salmon also showed that in vitro GR and MR interact to modulate transcription of target

genes, suggesting that both these receptors are involved in modulating target tissue cortisol responsiveness during stress in fish (Kiilerich et al., 2015). This seems plausible in vivo given that both these receptors are expressed in most tissues in fish (Aruna et al., 2012). Other roles for MR in fish include the negative feedback regulation of cortisol during stress (Alderman et al., 2012), as well as developmental programming (Alsop & Vijayan, 2008). Given that the vast majority of work in fish involves raising cortisol levels through exogenous treatment or endogenously via stress, it is not surprising that the function of MR has been masked by activation of GR. This emphasizes the importance of examining corticosteroid signalling under both basal and stressed conditions.

Figure 2: Corticosteroid Receptor Signalling in Target Cells Cortisol will diffuse past the plasma membrane of target cells and bind to either GR or MR, which exist as heterocomplexes with chaperone proteins in the cytoplasm. Upon ligand binding a conformational change will cause GR or MR to disassociate from the associated proteins and dimerize with either itself (homodimerization [GR/GR], [MR/MR]) or the other CR (heterodimerization [GR/MR]).

Metabolic adjustments during stress

Due to the ubiquitous expression of the CRs, the effects of stress and the attendant rise in cortisol is wide-spread and affects the whole organism. From a metabolic perspective, GC-driven changes involve several major organ systems and exchanges of several key metabolites (Fig. 3). GR-induced transcription of enzymes involved in hepatic gluconeogenesis will increase glucose production while white muscle provides the amino acids and lactate as substrates for enhancing gluconeogenesis in the liver (Milligan, 1997, 2003; Mommsen et al., 1999; Frolow et al., 2004).

An increase in glucose production post-stressor can be abolished with the GR antagonist, RU486, confirming a direct role of GR activation (Aluru & Vijayan, 2009; Philip et al., 2012; Philip & Vijayan, 2015). Although studies point to a role for cortisol in glycogen mobilization to increase glucose production, this is thought to be more a rapid non-genomic response by epinephrine via glycogen phosphorylase (GP) (Fabbri et al., 1998; Mommsen et al., 1999; Fabbri & Moon, 2015). The interaction between GCs and epinephrine at the level of the liver is not well characterized (Reid et al., 1992), nor is the interaction of GCs with other hormones, including glucagon or insulin (Vijayan et al., 2010). Little is known about the interplay between glucoregulatory hormonal dynamics or the signalling mechanisms that modulate glycogen depletion during acute stress, and its repletion during recovery from stress.

Glucose as fuel:

Glucose is one of the most fundamental fuel for energy driving vertebrate metabolism. It can be absorbed through the gut during digestion or produced endogenously by the liver and kidney through either the breakdown of glycogen (glycogenolysis) or de novo synthesis (gluconeogenesis) from amino acids and/or glycerol (Mommsen et al., 1999). The regulation of glucose availability and storage is tissue-specific, and both catecholamines and GCs play key roles in glucose regulation during stress in fish (Mommsen et al., 1999; Fabbri & Moon, 2015; Faught & Vijayan, 2016). An increase in plasma GCs is often correlated with an increase in plasma glucose. Indeed, while glucose production 2h-post-confinement in tilapia (Oreochromis mossambicus) is thought to be primarily due to glycogen breakdown via catecholamines, high glucose 24h post-stressor is due to gluconeogenesis (Vijayan et al., 1997). Key enzymes involved in liver gluconeogenesis, including phosphoenolpyruvate carboxykinase (PEPCK/pck1), are under the transcriptional control of GR and can be abolished using GR antagonist (Aluru et al., 2007). While cortisol treatment has been associated with a decrease in hepatic glycogen in vitro, and an associated rise in GP in rainbow trout, there has been no evidence in teleosts of transcriptional activation by GR in glycogen synthesis/breakdown. This suggests that either we have not yet undertaken the necessary empirical studies, or that cortisol regulation of glycogen breakdown is non-genomic, similar to epinephrine. Indeed, cortisol is

capable of activating secondary signalling cascades, including PKA in trout hepatocytes (Dindia et al., 2012, 2013). In mammals, the transcription of pyruvate carboxylase (PC), fructose-1,6-bisphosphate 1 (FBP1), phosphofructokinase 1 (PFK1) and glucose-6-phosphate catalytic subunit (G6Pase) and G6P transporter are all under the transcriptional control of GCs (Kuo et al., 2015). In fish, the transcriptional control of these genes by GR is not well characterized. Cortisol activation was correlated with increased mRNA abundance of pck1 in rainbow trout (Vijayan et al., 2003), and in higher enzyme activity in the hypercortisolemic socially subordinate rainbow trout (DiBattista et al., 2006). Some evidence from the rx3 zebrafish mutants (glucocorticoid deficient) also suggests that pck1 can be rescued by dexamethasone injection and thus under transcriptional control of GR (Weger et al., 2016). Direct transcriptional control of pck1 was demonstrated in trout liver using the GR antagonist RU486 (Aluru et al., 2007), where RU486 abolished the cortisol-induced increase in pck1 transcript abundance. Cortisol implants in brook charr (Salvelinus fontinalis) correlated with an increase in PFK1, and G6Pase enzyme activity (Vijayan et al., 1991); however, GR transcriptional regulation of these gluconeogenic enzymes in fish is unclear. Interestingly, in the muscle of exercised fish, high circulating cortisol levels are correlated with lower rates of muscle glycogenesis (Milligan, 2003). However, the hormonal response may be dictated by the initial glycogen status of the muscle, as lower glycogen content favours glycogenesis by GCs and epinephrine (Frolow et al., 2004). The rapid glycogenesis in metyrapone treated fish (low cortisol) was associated with an increased stimulation of glycogen synthase and a reduction in glycogen phosphorylase a activity (Milligan, 2003). This suggests that not only does the skeletal muscle limit glucose but that it has also adapted to recover glycogen more slowly in the presence of cortisol. It is unknown whether cortisol will impact transcript abundance of glycogen synthase paralogs (gys1/gys2) or glycogen phosphorylase (pygma, pygmb and pygl) in fish. A working hypothesis at this point is that the restriction in muscle uptake of glucose during stress diverts glucose towards smaller, yet highly aerobic tissues, including the brain. This is supported, in part, by the work of Blasco et al. (Blasco et al., 1996), which clearly showed that the highest rate of glucose uptake was in the brain, while the majority of glucose was taken up by the muscle because if its larger mass (Blasco et al., 1996).

It is well known in mammals that GCs restrict glucose uptake in skeletal muscle, and the GR antagonist has been approved for use in type 2 diabetes (Bernal-Sore et al., 2018). The mechanism of action, in mammalian myotubules, is due to the increased recruitment of Glut4 to the membrane in response to a decrease in cytosolic ATP. A change in the ATP/ADP ratio, increased phosphorylation of the nutritional sensor, AMPK, and subsequently glut4 translocation (Bernal-Sore et al., 2018). The restriction of glucose uptake by the muscle was also shown recently in GR knockout zebrafish ((Faught & Vijayan, 2019)/Chapter 6), suggesting an evolutionarily conserved role for GR in glucose reallocation during stress, however the mechanisms is unknown in teleosts. This is in part due to the lack of characterization of an insulin-sensitive glucose transporter (i.e. glut4) in zebrafish (Tseng et al., 2009). With energy demands during stress being increased, the amount of energy that is allotted to other pathways, including growth and immune response may be downregulated at the transcriptional level (Tort, 2011; Reindl & Sheridan, 2012). In response to stress, growth hormone (GH) receptors and insulin-like growth factor 1 (IGF-1) transcript levels were downregulated, and this was also observed both in vivo (Small et al., 2006; Nakano et al., 2013) and in vitro in fish hepatocytes (Leung et al., 2008; Philip et al., 2014). While studies have examined the effects of chronic stress and cortisol treatment on growth (Bernier & Peter, 2001; Bernier et al., 2004), the mechanisms of action are far from clear. Myostatin is known to negatively regulate muscle growth in vertebrates. In mammals, myostatin has GRE in its promotors, suggesting a direct role of cortisol in regulating growth. However, this response of cortisol is not conserved in teleosts (Biga et al., 2004; Galt et al., 2014), and cortisol treatment downregulated mystnb (the gene encoding myostatin) transcript abundance in the tilapia (Rodgers et al., 2003). Cortisol treatment is also known to attenuate the GH-induced IGF1 mRNA abundance in the liver, suggesting a link between stress and growth suppression (Leung et al., 2008; Philip & Vijayan, 2015). However, IGF2 transcript level is actually stimulated by cortisol in Coho salmon (Oncorhynchus kisutch) (Pierce et al., 2010) and tilapia (Pierce et al., 2011). A recent study also showed that chronic cortisol treatment increased liver insulin-like growth factor binding protein 1 (igfbp1), suggesting a potential role in growth suppression by modifying IGFs action (Madison et al., 2015). Both IGF1 and IGF2 are the primary mediators of GH-promoting effects and elicit downstream effects on other tissues, including gonads and

heart. For instance, igf2b is upregulated in the heart during zebrafish regeneration (Huang et al., 2013), and a positive correlation between cardiosomatic index and cortisol levels were reported in adult European brown trout (Salmo trutta), suggesting a role for cortisol in myocardial remodelling (Johansen et al., 2011). The interaction between the somatotropic axis and the cortisol mediated effects on cardiac development requires further work, especially given the fact that excess cortisol during development is having a very pronounced effect on cardiac development and function in zebrafish (Nesan & Vijayan, 2012; Wilson et al., 2013).

Energy substrates for oxidation and gluconeogenesis:

Skeletal muscle in fish comprises a significant proportion of their final body mass (>50%)(Sadoul & Vijayan, 2016) and is, therefore, an excellent source of stored protein for energy (Mommsen, 2001). Regulation of protein stores in the muscle by GCs appears to be driven by both catabolic (protein breakdown) and anti-anabolic (inhibition of protein synthesis) processes. In mice and in fish with a GR knockout in the muscle, there was a reduced capacity for protein breakdown, even under high cortisol or fasting conditions (Shimizu et al., 2015; Faught & Vijayan, 2019/Chapter 6). There is strong evidence from mammalian studies that the primary changes in muscle mass in response to cortisol are due to changes in the protein synthetic rate, and that the protein degradation is secondary and adaptive in response to stress (Rennie, et al., 1983). In recent years, mammalian work has highlighted potential targets of GR involved in muscle catabolism (Kuo et al., 2013). These include muscle-specific E3 ubiquitin ligases, muscle ring finger 1 (MuRF1) and muscle atrophy F-box (MAFbx) (Kuo et al., 2015). Interestingly, both of these genes are important in myofibril integrity in zebrafish cardiomyocytes (Shimizu et al., 2017), but the effect of cortisol on these muscle-specific ligases in fish has not been determined. Classical negative regulators of muscle growth and differentiation, such as myostatin, are also targets of cortisol-GR signalling in mammals, but this may not be the case in salmonid fish. Not only are there no identifiable GREs in the promotor regions of the myostatin orthologs in rainbow trout, but exogenous cortisol treatment failed to induce their transcript abundance in trout myoblasts (Galt et al., 2014). Although the targets of GCs signalling are not clear in fish, it is clear that cortisol-induced proteolysis is an essential part

of the metabolic stress response, as amino acids released from the muscle are utilized by the liver for oxidation and gluconeogenesis (Mommsen et al., 1999). This glucose is either shunted into the circulation for use by target tissues to cope with the increased energy demand associated with stress and/or used for replenishing the liver glycogen stores that are depleted due to the initial stressor-induced catecholamine stimulation (Vijayan et al., 1997). In mammals, cortisol is a potent catabolic hormone and under chronically elevated conditions can lead to muscle wasting (Kettelhut et al., 1988). Several key targets of muscle breakdown are under transcriptional control of GR. For instance, the protein, regulated in development and DNA damage response 1 (REDD1) is induced in response to GR activation under various stressful conditions, including hypoxia, DNA damage and energy stress. Studies of REDD1 function in teleosts is limited to developmental studies in zebrafish (Feng et al., 2012). In mice, REDD1 is a glucocorticoid-responsive gene in the muscle and is thought to reduce muscle metabolism to enable adaptation under energetic stress (Britto et al., 2018). REDD1 protein is a negative regulator of muscle mass through inhibition of the Akt/mTORC1 signalling pathway and is induced following glucocorticoid secretion (Britto et al., 2018). Furthermore, REDD1 deletion will abolish glucocorticoid-induced skeletal muscle atrophy in mammals (Britto et al., 2014). In fish, proteolytic genes, including cathepsin D, are under the transcriptional control of GR in the liver, but this remains to be determined in the muscle (Aluru et al., 2007). In zebrafish lacking GR, there is a marked increase in muscle protein content, including higher expression of key translation factors, which supports the postulation that the anti-anabolic effects of cortisol may be mediated by GR in fishes (Faught and Vijayan, 2019/Chapter 6). Apart from glucose and amino acids, lipid and their constituent fatty acids are also major energy source in fish. Lipid in poikilothermic vertebrates are stored at several sites (mesenteric, liver, muscle) as triacylglycerols, and polyunsaturated fatty acids (PUFAs) (Sheridan, 1988, 1994). Lipid storage occurs during periods of feeding, and high insulin levels, while lipid depletion occurs during transitional and non-feeding periods, including during salmonid smoltification (Sheridan, 1988, 1994). While factors such as diet, age and reproductive cycle all play a role in fat deposition, GCs also play a key role in adipogenesis, but this is poorly understood in teleosts. In mammals, the effect of GCs on fat biosynthesis is well known. For instance, chronically high cortisol levels associated with Cushing’s syndrome often leads to fat

synthesis and targeted deposition (Peeke & Chrousos, 1995). Indications of GCs involvement in adiposity in fish arose when low responding and high responding trout were reared separately (Trenzado et al., 2006). High responding trout had lower lipid levels when subjected to a crowding stressor compared to either the control or low responding strains (Trenzado et al., 2006). While a higher degree of lipolysis could be the case in the high responding trout, genes involved in adipogenesis may also be under the transcriptional control of CRs with elevated cortisol levels. In the GR knockout zebrafish there is an increase in the amount of lipids reported in adults, suggesting that fat deposition may be independent of GR, but still governed by the high cortisol levels in these fish (Facchinello et al., 2017). In support of the lipolytic effect of cortisol, studies have shown that stress and/or cortisol increases the activity of lipolytic enzymes in fish (Sheridan, 1986; Vijayan et al., 1991; Sheridan, 1994; Mommsen et al., 1999). Also, adrenergic stimulation (norepinephrine but not epinephrine) showed a lipolytic effect in salmon liver leading to glycerol release in vitro (Sheridan, 1987). The glycerol released from the breakdown of triglycerides have the potential to be used gluconeogenic substrates in fish (Mommsen et al., 1999). Under acute stress, lipolysis can be activated to cope with increased energy demands. The catabolism of fatty acids, β-oxidation, involves the sequential cleavage with acetyl-CoA as a usable endpoint for the TCA cycle (Tocher, 2003). Various lipases including triglycerol (TG) lipase, lipoprotein lipase and lysosomal lipase are responsible for lipolysis in fish (Sheridan, 1988, 1994). While lipoprotein lipase was differentially expressed in rainbow trout liver in response to an acute stressor, a direct role for GR has not been determined (Wiseman et al., 2007). Gilthead seabream (Sparus aurata) held at a high stocking density resulted in high cortisol levels that was correlated with a reduction in liver lipid content and altered lipid composition (Montero et al., 2001). Interestingly high cortisol coupled with a high-fat diet led to liver steatosis, but reduced muscle lipid content (Montero et al., 2001). It also appears that the maintenance of energy hemostasis during food deprivation is directly related to the capacity of the liver to mobilize lipid reserves (Sheridan and Mommsen, 1991). Overall while there is a known link between GCs and lipid metabolism, the mechanism(s) of action are not well understood in teleosts.

Figure 3: Metabolic Regulation during stress During basal conditions, hepatocytes will take up glucose through GLUT2 transporters and store it as glycogen. Pyruvate can also be used to produce triglycerides, which contribute to adipose tissue formation. Growth hormone signalling in the liver causes an upregulation of IGF1, which signals the liver to increase protein synthesis through mTOR. Glucose and fatty acids can be taken up by adipocytes, which increase triglycerides. During stress, hepatocytes will produce glucose through gluconeogenesis, which can be taken up by the muscle through GLUT4. Cortisol will also stimulate the breakdown of proteins by upregulating E3 protein ligases, including Redd1 and Murf1. This liberates amino acids, which can be used by the hepatocyte to fuel gluconeogenesis. Cortisol will also breakdown triglycerides through beta-oxidation to produce energy (ATP). Enzymes/proteins in green are regulated by cortisol. Abbreviations: Glucose-6-phosphate (G6P), tricarboxylic acid cycle (TCA), adenosine-triphosphate (ATP), aspartate aminotransferase (AST), alanine aminotransferase (ALT), lactate dehydrogenase (LDH), acetyl-CoA carboxylase (ACC), fatty acid synthase (FAS), hexokinase (HK), Phosphoenolpyruvate carboxykinase (pck1), glucose transporter 2 (GLUT2), glucose transporter 4 (glut4), growth hormone receptor (GHR), insulin-like growth factor 1 (IGF1), IGF1 receptor (IGFR), 3- hydroxylacyl-CoA (HOAD), fatty acid transport protein (FATP), regulated in development and DNA damage responses 1 (Redd1), muscle RING-finger protein 1 (Murf1). Adapted from Faught et al., 2019

Development of the HPI axis

Much of the work on early development has used zebrafish as a model due to their rapid development, transparent embryos, and the utility and ease of this species for forward and reverse genetics manipulation (Nesan & Vijayan, 2013a; Phillips & Westerfield, 2014; Rossi et al., 2015; Prykhozhij & Berman, 2018). The chronology of HPI gene expression, including crh (Chandrasekar et al., 2007), pomc (Herzzog et al., 2003; Liu et al., 2003), star, mc2r (To et al., 2007), 11β-hydroxylase, mr, gr and 11βhsd2 (Alsop and Vijayan, 2008), during early development has been well characterized in zebrafish (Alsop & Vijayan, 2009a; Nesan & Vijayan, 2013a). Despite all the molecular machinery for HPI axis function in place by hatch (48 hpf) (Alsop & Vijayan, 2008). The earliest stress response occurs only at around 72-96 hpf, pointing to a delay in HPI axis activation, and the importance of a ‘hyporesponsive’ stress period during early development (Alsop & Vijayan, 2008; Alderman & Bernier, 2009; Wilson et al., 2013; Nesan & Vijayan, 2016). The maternal contribution of GR, and subsequent signalling is an important aspect of zebrafish developmental programming (Pikulkaew et al., 2011; Nesan et al., 2012; Nesan & Vijayan, 2012; Wilson et al., 2013; Nesan & Vijayan, 2016). Indeed, any perturbations in cortisol levels during this early period can also modify adult life behavioural and metabolic responses (Wilson et al., 2016) (Fig. 4). Much of the work outlining the importance of GR signalling was established by transiently knocking down GR using morpholino oligonucleotides (Appendix I: Table 1) and studying the associated changes in genotype and phenotype (Pikulkaew et al., 2011; Nesan et al., 2012; Nesan & Vijayan, 2013b). Knockdown of GR resulted in dysregulation of a number of key developmental genes involved in nervous system development, cellular movement, cell to cell signalling, cardiovascular development, organ morphology and skeletal/muscular system development at 36 hpf (Nesan & Vijayan, 2013b). Phenotypically, GR knockdown also caused delayed somitogenesis (12 hpf), defects in tail morphometrics and reduced embryo size (Nesan et al., 2012). Beyond the transient knockdown of morpholinos, GR knockout fish have also been established using forward genetics. GRs357 mutants have substitution of arginine to cysteine in the second zinc finger motif of the DNA binding domain of GR, which abolishes the ability of the mutant GR to bind to the GREs of target genes (Griffiths et al., 2012; Ziv et al., 2013). This mutant has chronically elevated

cortisol levels, and a hyperactive HPI axis, which was reflected by elevated crh and pomca expression (Ziv et al., 2013). These fish also exhibit behavioural phenotype akin to depression, including decreased exploratory behaviour and impaired habituation to repeated exposure to an anxiogenic environment (Ziv et al., 2013). Differences in phenotypes between knockouts and knockdowns is common and is thought to involve compensatory mechanisms, in the case of the knockouts (Rossi et al., 2015). Maternal stress and excess cortisol transfer is also detrimental in fish (Kumai et al., 2012; Nesan et al., 2012; Nesan & Vijayan, 2012). Cortisol microinjected into one-cell embryos to mimic maternal stress-related cortisol transfer cause heart deformities, and this was correlated to suppression of key cardiac genes, including nkx2.5, cardiac myosin light chain 1, cardiac troponin type T2A, and calcium-transporting ATPase (Nesan et al., 2012). These results suggest a role for cortisol in cardiac development; however, whether this is regulated by GR activation was not tested (Nesan et al., 2012). Recent work in mice reported that in cardiomyocyte-specific knockout of both GR and MR, the GRKO resulted in cardiac hypertrophy and premature death from heart failure, while MR knockout had no effect (Oakley et al., 2019). This supports the contention in fish that excess cortisol will activate GR to cause heart deformities (Nesan & Vijayan, 2012). Injection of cortisol at the one-cell stage in zebrafish also increases primary neurogenesis in the preoptic area and pallium (homologous to the hippocampus in mammals), and this corresponds to alterations in larval behaviour (Best et al., 2017). These cortisol-induced behavioural changes were abolished in larvae treated with RU486 (GR/PR antagonist), suggesting a role for either GR or PR in mediating behaviour. These results underscore the critical role for the tight regulation of maternal cortisol deposition in proper developmental programming, because stress and excess cortisol deposition affects brain function leading to abnormal behavioural and cardiac phenotypes (Fig. 4). From a metabolic standpoint, little work has been done on the effect of GR on the larval or postnatal metabolome. This is mostly because all the above-mentioned studies were carried out prior to feeding in zebrafish. However, using GRα and GRβ splice-blocking morpholinos, and subsequent treatment with the synthetic GC dexamethasone, two distinct gene clusters were activated under either basal or stressed conditions (Chatzopoulou et al., 2015). The increase in

glucose in the dexamethasone-treated group was linked to several key gluconeogenic genes, including pck1, pck2, pfkrb41, and g6pca (Chatzopoulou et al., 2015). However, further work determined that there was no transcriptional role for GRβ in zebrafish (Chatzopoulou et al., 2017). This clearly points to another receptor, such as MR, which may regulate basal function. Unfortunately, in all the above mentioned developmental studies, no work was carried out to elucidate the impact of MR signalling, with the exception of Cruz et al. (2013) that concluded GR and not MR mediates mineralocorticoid function in fish (Cruz et al., 2013). While most studies have examined the role of GR and MR function using pharmacological approaches, the introduction of gene knockdown and gene knockout technologies, especially in zebrafish, has resulted in mechanistic studies directly related to receptor activation. However, most of the

Figure 4: Effect of Cortisol on Zebrafish Development Alterations in zebrafish development have been noted in response to both excess cortisol and attenuated cortisol in the zygote, which persists to adulthood in zebrafish (Nesan and Vijayan, 2013, 2016; Best et al., 2017; Wilson et al., 2013; 2015). Excess cortisol (red arrows) is a scenario during maternal stress and affects progeny neurogenesis, larval stress response, and cardiac deformities. Basal cortisol levels (blue arrows) show the normal developmental profile of zebrafish embryos and larvae. Low cortisol, or attenuated GR signalling (green arrows) may occur due to altered cortisol dynamics during oogenesis and are also characterized by an altered stress response, and cardiac function. Homozygous GR knockout fish (-/-) are characterized by high cortisol levels, and no GR signalling and have therefore been included in both “excess” and “low” categories. Early life aberrations in cortisol signalling have not been examined on a multigenerational scale (denoted by ?). Manipulation of cortisol in the zygote is achieved by maternal stress, cortisol microinjection, bath exposure of this steroid, GR mutants [GR (-/-)], GR MO or cortisol-specific antibodies. Abbreviations: Glucocorticoid receptor (GR), morpholino (MO), cortisol (F), minutes post-fertilization (mpf). Adapted from Faught and Vijayan, 2018b.

studies so far have focussed on GR manipulation (Fig. 4), while only one study has tried to knockout MR in fish (Sakamoto et al., 2018).

Glucocorticoid and Mineralocorticoid Morphants and Mutants:

Prior to the adoption of nucleases to target genes for modification, genetic manipulation of the zebrafish genome was limited to RNA guided knockdowns. Morpholino-modified anti- sense oligonucleotides (MOs) bind to RNA to inhibit protein synthesis (Rossi et al., 2015). These were widely used for early developmental work and would silence both the maternal and zygotic de novo synthesis of proteins in zebrafish. While there were several papers on the function of GR by using morpholinos (Pikulkaew et al., 2010; Kumai et al., 2012; Cruz et al., 2013; Nesan & Vijayan, 2013a, 2013b; Chatzopoulou et al., 2015), there is a limited description of MR MOs. MOs were first used to knockdown GR protein in zebrafish to establish that GR was required for early life stage fin regeneration (Mathew et al., 2007). Also, studies reported increases in craniofacial, mesoderm and caudal deformities, with severe malformations of neural, vascular and visceral organs due to GR knockdown (Pikulkaew et al., 2011; Nesan et al., 2012). All these deformities could be rescued with an injection of gr mRNA, suggesting that GR signalling was essential for early embryogenesis. Other studies used GR MOs to determine the importance of GR, but not MR in the osmoregulation of fish (Kumai et al., 2012; Cruz et al., 2013; Lin et al., 2015). Analysis of the transcriptome (microarray) of GR morphants revealed that 1313 genes were differentially expressed by 24 hpf, with many of these genes being involved in the nervous system, cardiovascular and skeletal development (Nesan & Vijayan, 2013b). MOs were also used to antagonize alternative splicing, as in the case of GRα and GRβ, to establish that GRβ acts as a dominant-negative inhibitor of GRα in zebrafish (Chatzopoulou et al., 2015). However, the major disadvantage of MO use was the transient nature of the silencing. In zebrafish, GR protein synthesis was often restored by 5 dpf, which precluded any insights into the function of GR in fish post-feeding. Furthermore, the impact of early life perturbations of GR often had lifelong effects. Adult zebrafish, which had been injected with GR MOs in early life, exhibited decreased boldness compared to GC (dexamethasone) treated fish (Wilson et al., 2016). The use of the GR antagonist, RU486, to block this receptor action in vivo is limited, as

this drug is also a PR antagonist (Faught & Vijayan, 2018b). Therefore, it became paramount that stable GR and MR knockout strains be created. The GR knockout strain, created using a reverse genetics approach with ENU as a mutagen (GRs357), showed remarkably different phenotypes compared to the morphants described above (Griffiths et al., 2012; Ziv et al., 2013). This was a consistent observation, regardless of the gene, that the phenotypes obtained by knockdown or knockouts were different (Rossi et al., 2015). Through a series of comparative experiments, Rossi et al. (2015) established that there were changes in the transcriptome to compensate for the loss of a gene by upregulating genes with similar function (Rossi et al., 2015). It is not yet clear what gene could be compensating for GR loss in zebrafish. While MR is an obvious candidate, its transcription is often not upregulated in response to GR inactivation (Faught and Vijayan, 2018b). It may be that due to the tissue-specific role of GR (Appendix I: Table 1), different genes are compensating for the loss in a tissue-specific manner. Both the GRs357 mutant and the CRISPR/Cas9 generated mutants are characterized by hyperactive HPI axis and elevated cortisol levels (Ziv et al., 2013, Faught and Vijayan, 2018b/Chapter 3). The GR knockout mutants are hypercortisolemic beyond what can be induced by a stressor (Ziv et al., 2013, Faught and Vijayan, 2018b), and this appears to be due to a lack of regulation of 11βHSD2, which is responsible for the degradation of biologically active cortisol to cortisone (Faught and Vijayan, 2018a). The only described MR mutant in fish, to date, is in medaka (Sakamoto et al., 2016). The mutants are phenotypically similar to the wildtype, but as adults show altered behaviour (Sakamoto et al., 2016). Medaka are not an ideal fish to create an MR mutant, as they have two GR genes (GR1 and GR2), which makes the creation of double knockout fish more complicated. The duplication of GR is thought to be due to the teleost-specific genome duplication event 350 million years ago (Baker et al., 2013). Zebrafish only have a single GR gene (Alsop and Vijayan, 2008; Schaaf et al., 2008), which has two splice variants (GRα and GRβ), as is the case for higher vertebrates (Baker et al., 2013). A comparison between the GRs357 mutant and the CRISPR/Cas9 revealed that the GRs357 mutant may retain some of the protein-protein signalling ability (Facchinello et al., 2017). This is likely due to the point mutation which impacts transcriptional activation ability but still results in a full-length protein (Griffiths et al., 2012; Facchinello et al., 2017). Both GRKO mutants

created with CRISPR/Cas9 mutagenesis have a premature stop codon, resulting in a truncated protein (Facchinello et al., 2017/this thesis). Overall, these studies highlight the utility of zebrafish as a model to study the role of CR activation in developmental regulation and stress adaptation. However, from the above studies, a clear unknown is the role of MR activation in regulating GCs-mediated effects. Indeed, during a thematic review in the Journal of Endocrinology in 2018 entitled “30 years of the mineralocorticoid receptor”, it was explicitly stated in several reviews that the function of the MR was unknown in ray-finned fishes and lower vertebrates (Baker & Katsu, 2017; Funder, 2017).

Hypothesis and Objectives:

In fish, the role of stress, and the attendant rise in cortisol has been studied extensively. However, despite physiological processes that are known to be impacted by changes in cortisol levels, the mechanisms involved in bringing about these changes are still unclear. Indeed, much of the work to date has focused on GR signalling and its role in energy substrate mobilization to cope with stress. Almost all work has overlooked the other corticosteroid receptor, MR, as having any role in mediating the effects of cortisol. Despite the persistence of MR in vertebrate evolution, there is no known physiological role for MR in ray-finned fish. Furthermore, there has been little focus on the fundamental role of corticosteroid signalling and their physiological consequences, in the absence of stress. Loss of MR and GR in mammals results in death in utero, or postnatally due to dehydration or delayed lung maturation, respectively. However, loss of these two receptors is not lethal in zebrafish (Faught and Vijayan, 2018), making them an attractive model with which to study the physiological impacts of GR and MR signalling at the systems level. Therefore, the overarching hypothesis of this thesis is that both MR and GR activation either singly and/or in combination are necessary for mediating the physiological effects of cortisol in zebrafish. Specifically, the objectives were: 1. Generate GRKO or MRKO zebrafish using CRISPR/Cas9 mutagenesis as a model to study CR activation (Chapter 2). 2. The role of CRs on developmental programming of the HPI axis functioning (Chapter 3). 3. Determine transcriptional targets of MR and GR in postnatal adipose regulation (Chapter 4). 4. The role of GR and MR activation on postnatal growth and protein homeostasis (Chapter 5). 5. The role of GR signalling in glucose homeostasis and muscle protein synthesis in adults (Chapter 6).

CHAPTER 2: GENERATING AND GENOTYPING GR AND MR KNOCKOUTS IN

ZEBRAFISH WITH CRISPR/CAS9

Introduction:

Manipulation of the genome has been essential to our understanding of gene function. The CRISPR (clustered regularly interspaced short palindromic repeats)/Cas (CRISPR- associated protein) system is an RNA-guided platform which is a highly efficient and specific genetic editing technology (Sternberg et al., 2014; Wang et al., 2016; Komor et al., 2017). This system has greatly facilitated the targeted inactivation of genes in vivo and in vitro (Ablain et al., 2015). It was adapted from the bacterial ability to silence foreign nucleic acids. Like many genetic mutagenesis techniques, including Transcription activator-like effector nucleases (TALENS) and Zinc finger nucleases (ZFNs), this system involves a DNA-binding domain that recognizes a specific sequence of DNA and an effector domain that enables DNA cleavage or regulates transcription near the binding site (Wang et al., 2016). Double-stranded breaks (DSBs) in an early exonic sequence leads to insertions and deletions (indels), which frequently leads to gene disruption through frameshift or nonsense mutations, disrupting the proper transcription and/or translation causing a non-functional protein product (Sternberg et al., 2014; Komor et al., 2017). Once a Cas9-mediated DSB occurs the cell will use endogenous DNA repair mechanisms, either non-homologous end-joining (NHEJ) or homology-direct repair (HDR) to re-establish DNA integrity (Komor et al., 2017). DSBs are more likely to be repaired via NHEJ compared to HDR. Therefore in the absence of a template, NHEJ is often the default mechanism by which the DNA will repair itself after a DSB at target sites, which means that the resulting mutations will be indels (Sternberg et al., 2016). HDR is a targeted sequence replacement for precise mutations, gene corrections or gene-knock-ins (Sternberg et al., 2016). The CRISPR system is an adaptive immune mechanism present in many bacteria and most archaea (Sternberg et al., 2016). CRISPR-containing organisms acquire DNA fragments from invading bacteriophages and plasmids before transcribing them into CRISPR RNAs (crRNAs) to guide cleavage of invading RNA or DNA (Wang et al., 2016). CRISPR systems have been divided into two major classes based on differences in their components and mechanisms of action. In class I systems, RNA-guided target cleavage (Cas protein types I, III, IV) requires a large complex of several effector proteins, but in the class 2 systems only one RNA-guided endonuclease is required to mediated cleavage (Jinek et al., 2012). To elicit an immune response to foreign DNA using the CRISPR system, DNA fragments (protospacers) of

invading plasmids or phages are incorporated into the host genome (Wang et al., 2016). Protospacers are incorporated at the CRISPR locus as spacers between crRNA repeats (Sternberg et al., 2016). Cas proteins are then expressed, and the CRISPR array containing the protospacers are transcribed, cleaved and processed into mature crRNA (Sternberg et al., 2016; Wang et al., 2016). Finally, Cas proteins, with the guidance of the crRNA, will now recognize targets and mediate the cleavage of the invading genome (Wang et al., 2016). The action of CRISPR depends on the sequence-specific protospacer adjacent motif (PAM) that is adjacent to the crRNA target site (Wang et al., 2016). These PAM sites are absent in the host genome, which protects itself from self-cleavage (Sternberg et al., 2014). Cas9, the RNA guided endonuclease that cleaves target DNA in the class 2 type II CRISPR system, is the most widely used enzyme for genomic editing in eukaryotic cells (Komor et al., 2017). Each CRISPR nuclease differs in its size, PAM requirement and location of the introduced double-stranded break (reviewed by (Komor et al., 2017)). The most commonly used, and the one used herein, is Cas9 protein from Streptococcus pyogenes (SpCas9), which recognizes the ‘NGG’ PAM sequence (Jinek et al., 2012). Cas9 proteins targeted cleavage is guided by a duplex of two RNAs. The CRISPR-RNA (crRNA) that recognizes the invading DNA through a 20-bp region and the trans-activating RNA (tracrRNA) that hybridizes with the crRNA and is unique to the type II CRISPR system (Komor et al., 2017). Cas9 in conjunction with the crRNA-tracrRNA duplex can be repurposed for genomic editing. In work by Jinek et a., 2012, this RNA duplex was fused to a single guide RNA (sgRNA) (Jinek et al., 2012). The Cas9-sgRNA complex binds to the DNA that is adjacent to the PAM sequence (Sternberg et al., 2014) and is widely used in all zebrafish editing (Varshney et al., 2016). Prior to the adaptation of the CRISPR/Cas system to eukaryotes, targeted DSBs to induce mutagenesis in specific genes was still possible in the form of ZFNs and TALENS (reviewed by (Joung & Sander, 2013)). In both systems, the DNA-binding domains of transcription factors were used with the nuclease domain of the restriction enzyme FokI (Joung & Sander, 2013). When targeted to paired adjacent sequences, the FokI domains of these programmable, site- specific nucleases form a dimer that activates the nuclease activity, thus creating a DSB near their binding sites. Endogenous repair pathways can then be exploited, as described above, to create mutations at the desired DSB sites. However, because these tools function through

protein-DNA interactions, the efficiency of targeted DSB was much lower compared to Cas9, which is an RNA guided nuclease whose sequence specificity arises from Watson-Crick base- pairing between the guide RNA and the target DNA site (Komor et al., 2017). In addition, a direct interaction between Cas9 and a short protospacer adjacent motif (PAM) site on the DNA results in increased specificity of this system (Komor et al., 2017). One of the major concerns with any targeted gene editing technology is the possibility of off-target effects (Rossi et al., 2015). In CRISPR/Cas9, the potential that the 20-bp targeting sequence in the sgRNA plus the 3-bp PAM may potentially be present elsewhere in the genome is usually low. In zebrafish, the rate of off-target effects is estimated to be between 1- 2.5% (Hruscha et al., 2013). There are several ways to improve specificity, including optimizing sgRNA design, using paired nCas9s, paired Cas9-FokI nucleases, shorted sgRNAs or RNAs with two unpaired G on the 5’ end that are more sensitive to mismatches, and decreasing the concentration or length of time the Cas9-sgRNA complex exists within the cell (Wang et al., 2016). Overall, both the efficiency and the specificity of the CRISPR/Cas9 system made it an ideal method to generate mutants with targeted knockouts in the stress axis system. In order to study the role of CRs in stress adaptation, the focus of this thesis was to generate glucocorticoid receptor (GR) and mineralocorticoid receptor (MR) knockouts in zebrafish using the CRISPR/Cas9 mutagenesis. While the role of GR has been studied in zebrafish development and stress response, nothing is known about the role of MR, and a physiological role for this receptor has not yet been identified in teleosts.

Materials and Methods:

Zebrafish maintenance

Adult zebrafish (Tupfel long fin (TL) strain) were maintained on a recirculating system with a 14:10 light: dark cycle (Tecniplast, Italy) All experimental protocols were approved by the Animal Care and Use Committee at the University of Calgary Animal Care Committee (AC17- 0079), and were in accordance with the Canadian Council on Animal Care guidelines. Water was maintained at 28.5°C, ph7.6, and 750 µS conductivity, and 10% of the water was exchanged

daily. Animals were fed twice daily with Gemma micro 300 diet (Skretting, USA) in the morning and live Artemia (San Francisco Bay Brand, USA) in the afternoon. Zebrafish Rearing: Embryos/larvae were reared from 0-5 days post fertilization (dpf) in a 28.5°C incubator in 10 cm Petri dishes (Sarstedt, USA) at a density of 100 embryos/dish in E3 embryo media (5mM NaCl, 0.17 mM KCl, 0.33mM CaCl2, 0.33 mM MgSO4 + 0.1 ppm methylene blue antifungal agent (Nusslein-Volhard & Dahm, 2002). Embryos/larvae were raised on a 14h light: 10h dark cycle and 50% of the E3 embryo media was replaced daily. Euthanasia of embryos was done with MS222 (0.3g/L). Larvae that were being raised were fed starting at 5 dpf. Briefly 5dpf larvae were transferred to a 3L tank with 1.5 L of water and kept in a temperature-controlled room at 28.5°C on a 14h light:10h dark cycle. Larvae were fed Gemma micro 150 (Skretting, USA) in the morning, and live Artemia in the afternoon and 50% of the water was changed daily until 15 dpf. At 15 dpf larvae were transferred to the recirculating system (Tecniplast, Italy) and maintained as described above.

Generation of nr3c1 and nr3c2 mutants- Step-by-Step Methods:

F0: inject embryos Somatic mutation with sgRNA and Cas9 Outbreed to WT analysis RNA

F1: Indentify Breed heterzygous germline trasmitting Analysize sequences fish with the same founders mutation together

F2: Identify homozygous mutants F3: maternal-zygotic mutants (zygotic mutants)

Figure 1: Overview of methods to generate maternal-zygotic zebrafish mutants.

Design of CRISPR Target and Genotyping Primers:

1. Identify the target zebrafish orthologs. In some cases, you may identify two paralogs due to a duplication event in the zebrafish genome (https://www.ncbi.nlm.nih.gov/genbank/):

a. Search for your gene of interest: i. Specify the species (for zebrafish put “Danio”). ii. Ensure that you have selected the “Gene” database from the dropdown menu to the left of the search bar b. Results: c. Select Ensembl database from “See related” section d. Click on the protein-coding section of the graph and under “Transcript” Select “Exons” e. Select a coding region of an exon preferably close to the start (ATG) sight but downstream of in-frame ATG sites. f. Cut and paste the desired sequence into one of the design sites i. http://www.crisprscan.org/?page=sequence OR http://zifit.partners.org/ZiFiT/ ii. Ensure that you are selecting Danio rerio as the species and a T7 in vitro promoter

g. Select the best guide: i. Low off-target effects ii. Early in the exon/coding region iii. Guides that meet the requirements (i.e. beginning with GG if possible) iv. CHECK where your guide is using the Ensembl exon page

h. Design primers for fragment analysis (see section 2.5) i. Forward primer: Using the ENSEMBL sequence, select 20 bp sequence upstream of your target. Ensure that this forward primer is more than 25 bp away from your target for sequencing purposes. ii. Reverse primer. Select 20 bp downstream of your target and create the reverse complement. iii. Note the amplification size

i. Check that your primers are amplifying via In Silico PCR i. http://rohsdb.cmb.usc.edu/GBshape/cgi-bin/hgPcr i. Make sure you are selecting zebrafish as a species ii. Note the amplification size and add 25 bp to account for the amplification of your primer adaptors

j. Add the following adaptors to your designs for in vitro transcription and fragment analysis.

FORWARD PRIMER M13F TGTAAAACGACGGCCAGT REVERSE PRIMER PIG tail GTGTCTT TAATACGACTCACTATA GGNNNNNNNNNNNNNNNNNNNN TARGET GTTTTAGAGCTAGAAATAGC

Synthesis of sgRNA

1. Oligo Assembly: Reagent (final concentration) Amount (µl) nf Water 17 5x Phusion Buffer 5 dNTPs (10 mM) 0.5 Oligo 1 (10 uM) 1 Oligo 2 (10 uM) 1 Phusion DNA polymerase 0.5 TOTAL 25 µl

2. Run the program “Assemble Oligos” with the following conditions: 98°C 2 min; 50°C 10 min; and 72°C 10 min; hold at 4°C. 3. Run 4µl of product on a 2.5% gel*. You should see a band at 120 bp. *gel must be 2.5% to give you the resolution required to distinguish between your individual and assemble oligos). 4. Prepare the sgRNA using the NEB T7 high yield RNA synthesis kit. Assemble the following reaction:

Reagent (final concentration) Amount (µl) NTP buffer 5 Annealed oligo (fresh) 4 T7 RNA polymerase 1 TOTAL 10 µl

5. Run program “NEB RNA synthesis” with the following conditions: 37°C 240min (4h); hold at 4°C. 6. Add 1 ul of the provided DNase and run the program “DNase 1 - 37°C, 20 min” with the following conditions: 37°C, 20 min, hold at 4°C. 7. Clean up the resulting RNA using the Zymo RNA clean and concentrate kit. Elute with 20ul of nf water. 8. Quantify RNA and store at -80°C.

Synthesis of Cas9 RNA

1. Day 1: From the glycerol stock of the pT3TS-nCas9n plasmid in the -80°C streak a prepared ampicillin LB agar plate. Incubate at 37°C overnight. 2. Day 2: Check for growth in the morning and move the plate to 4°C. 3. At ~ 4pm set up superbroth media for colony growth by adding ampicillin to the appropriate amount of superbroth (~10 ml of superbroth/colony). Select at least two colonies. 4. Add 10 ml of the antibiotic inoculated superbroth to a 50 ml falcon tube. Select a colony with a pipette tip and eject into the superbroth. Incubate at 37°C overnight, shaking at 250 rpm. 5. Day 3: Check for growth (turbidity) in the morning and use the GeneJet mini prep kit to isolate the plasmids. 6. Quantify DNA 7. Digest ~5 µg of the plasmid with XbaI for 1h at 37°C and inactivate at 65°C for 20 minutes. Divide into 25 µl aliquots in the PCR cycler, total volume 50 µl. 8. Purify the digested DNA with the Wizard PCR clean-up kit 9. Using ~ 500 ng as a template for in vitro transcription using the T3 mMessage kit (no poly A tailing) 10. Precipitate RNA using LiCl (provided in mMessage kit) 11. Quantify and aliquot (300 pg in a final volume of 10 µl)

Injections of sgRNA and Cas9:

1. Microinject 50 pg sgRNA and 300 pg Cas9 per embryo (1.4 µl injection volume) – See “Zebrafish embryo microinjection protocol” 2. If pooling multiple targets use 25-50 pg of each sgRNA and 300 pg Cas9 3. Use the CRISPR Calculator available in (Carrington et al., 2015) to calculate injection volumes.

Checking Activity of your sgRNA:

Fragment Analysis separates fluorescent fragments of DNA based on size, using capillary electrophoresis, which are sized by comparison to a size standard (Carrington et al., 2015). Fluorescent fragments are generated using PCR amplification with the incorporation of a fluorescent tag. Capillary electrophoresis separates ionic fragments by size. The negatively charged DNA will travel based on its size in the capillary array. This technique can be used for multiplexing by using different fluorescent probes and can detect changes as small as 1 bp (Carrington et al., 2015). All fragment analyses of PCR products were performed by the University of Calgary DNA sequencing facility using a 3500/3500xL Genetic Analyzer (3500 Series instrument). Standards varied based on the project. All peaks were analysed using the Peak Scanner Software (Applied Biosystems, USA).

1. Collect 8 injected and 2 uninjected embryos in individual tubes. 2. DNA Extraction (Extract-N-Amp): a. Make a master mix by calculating for 25 µl/sample of extraction solution and 7ul/sample of tissue prep solution.

b. Add 32 µl of the above master mix to each sample, ensuring the sample is submerged. c. Incubate at room temperature for 10 min d. Incubate samples at 95°C for 5 minutes e. Add 25 µl/sample of Neutralization Solution and vortex f. Store DNA at -20°C 3. Fluorescent PCR: a. Dilute extracted DNA 10x b. Make a Primer ‘Pair’ 10 µM stock from 100 µM stocks (10 µl forward, 10 µl final, 10 µl M13-HEX/FAM +70 µl water) c. Set up PCR Master Mix:

Components (-30 freezer) [Final] 1x (ul) 10x Nuclease-free water to vol 7.75 77.5 10 x Buffer 1x 1 10 MgCl2 (50 mM) 1.5 mM 0.3 3 dNTP mix (10 mM) 0.2 mM 0.2 2 Primer Pair (10 µM) 0.2 uM 0.2 2 cDNA 1 ul/20 ul 0.5 X Taq 5U/ul 1 U/ml 0.05 0.5 Total 10 9.5ul/tube d. Run PCR under the following conditions: 95°C – 5 min, 40 cycles of 94°C – 1 min, 55°C-65°C –1 min, 72°C - 30 sec, extend at 72°C - 5 minutes, hold at 4°C. e. Send for Fragment Analysis at the U of C Sequencing lab. f. Analyze the .fsa files data using Peak Scanner from Applied BioSystems. g. Analyze peak volume according to (Carrington et al., 2015). h. sgRNA that displays high somatic activity levels has been directly correlated with germline transmission. Grow these fish to adulthood.

Founder screening:

1. Pair breed one adult injected fish with one WT fish. 2. Collect 10 embryos and extract DNA as described above. 3. Perform fluorescent PCR and send for fragment analysis. 4. Check if the F1 progeny are heterozygous and note the transmitting mutations. 5. Pair breed transmitting F0 fish with WT fish – aim to have at least 50-100 fish.

Genotyping of F1

While fragment analysis tells you the size of the mutation it does not tell you the specific indel. Furthermore, you may have two mutations that yield the same fragment size, but different mutations.

1. Collection of fin clips in adult fish: a. Anaesthetize fish in 0.15g/L. and remove a small slice of their caudal fin and place in individual tubes. 45

b. Place fish in genotyping trays until their tails have been sequenced and they can be sorted 2. DNA Extraction (Extract-N-Amp): a. Make a master mix by calculating for 50µl/sample of extraction solution and 14µl/sample of tissue prep solution. b. Add 64 µl of the above master mix to each sample, ensuring the sample is submerged. c. Incubate at room temperature for 10 min d. Incubate samples at 95°C for 5 minutes e. Add 50ul/sample of Neutralization Solution and vortex f. Store DNA at -20°C 3. Make a Primer pair of a 10µM stock from 100 µM stocks (10 ul forward, 10ul final, 80 ul water) 4. Set up PCR exactly as described above. 5. Exo-Sap Clean-up: a. Transfer 4 µl of the PCR mixture to a 0.2 ml PCR tube b. Add 1 µl of nuclease-free water and 1 µl of Exo Sap c. Run Exo-Sap protocol (37°C for 15 min/80°C for 15 min/4°C hold) d. Transfer to a tube for sequencing containing (4 ul PCR clean-up +7.36 µl water + 0.64 µl 5 uM primer). F2: Check the genotype (WT, heterozygous and homozygous) using fragment analysis (Figs 1 and 3).

Results:

Glucocorticoid Receptor:

The target for in exon 2 of the nr3c1 gene on chromosome 14 resulted in a complex mutation. Mosaic founder fish were outbred to wildtype fish, to limit off-target effects of inbreeding mosaic founders and sequence analysis was performed (Fig. 2, 3). The target for GR resulted in an 18 basepair (bp) deletion and 11bp pair insertion, resulting in a net change, -7 bp deletion (Fig. 2). Breeding similar mutations together (i.e. -7bp x-7 bp fish), resulted in progeny displaying typical mendelian rations (25% homozygous mutant, 25% WT and 50%, heterozygous). Genotype was confirmed with fragment analysis (Fig. 4). Lack of protein expression via western blots confirmed a lack of GR protein (Chapter 3).

NR3C1 - Exon #2:

TAATGCAAAATG|GAT|CAA|GGA|GGA|CTG|GAG|AAT|GGC|AAA|AAG|CGA|GAT|GAG|CGT|TTA|AAT|ACA|T TG|GAT|TAC|AAC|AAA|CGA|GCA|ACT|GAG|GGG|ATA|TTA|CCA|AGA|AGA|ATA|CAA|AGC|ACC|ATG|TCT GTG|GCC|CCT|ACA|TCT|ATG|GTT|CCT|CAA|GCC|GGT|CCA|ATG|ATG|CAG|CCA|GTT|TCT|GGG|GAC|AT T|CCC|AAT|GGC|CTG|AGC|AAT|TCG|CCC|ACT|CTG|GAG|GAG|CAC|ACC|AGC|TCG|GTG|TCC|TCC|ATC| TTT|GGC|GAC|GAT|TCC|GAA|CTC|AAA|CTG|CTT|GGG|AAG|GAG|CAG|AGG|GCC|CTG|CAA|CAG|CAG|AC C|TTG|GTC|CCC|TTC|ACT|TTG|GGT|GAC|AGT|CTT|TCA|GGT|CTG|GAG|GCC|AGC|ATT|GCA|GAC|CTT| AAC|AAC|CCC|TCT|CCC|TCA|ATG|GAT|TCC|CTG|ATT|GGT|GGA|GTC|GAT|CCC|AAT|CTT|TTC|CCC|TT A|AAA|ACA|GAG|GAC|TTT|TCT|CCC|ATG|ATT|AAA|GGC|GAT|ATG|GAC|CTT|GAC|CAA|GAT|TCCTTTGG ACACATTGGGAAAGATGTTGATGTTGGCAATCATAAGCTCTTTAGTGACAACACTCTGGACCTCCTGCAGGACTTTGAGCTG GATGGATCGCCATCAGACTTCTACGTTGCTGACGATGCGTTTCTCTCCACCATAGGTGAAGATGCTCTCCTTTCAGAGCTGC CGACAAATTTAGACAGGGACTCGAAGGCTGCGGTTTCCGGGAGCAACACGCTCAATGGCACAGCTTCTTCCAGCCTCAGCAC AGCCAACACCAGCATCTTGCCCAATATAAAGGTGGAGAAAGACTCTATAATCCAGCTGTGCACCCCAGGGGTCATCAAACAG GAGAACACCGGTGCGAGCTATTGTCAAGGAGGGCTCCACAGCACCCCCATTAACATATGCGGGGTCACCACTTCAAGCGGAC AGAGCTTCCTCTTTGGGAACAGCTCGCCCACAGCTGTCGTCGGTCTGCAGAAAGATCAGAAGCCGGACTTTAACATGTACAC CCCTCTGACCTCTTCAGGAGATGGTTGGAGCAGGAGCCAGGGCTTTGGGAATGTCAGTGGAATGCAGCAGAGGGCCAGTTTA TGCTTTTCCAAAAACTTCTCAAGCAGCCCCTATTCCAG

∆WT-7 mutant TAATGCAAAATG|GAT|CAA|GGA|GGA|CTG|GAG|AAT|GGC|AAA|AAG|CGA|GAT|GAG|CGT|TTA|AAT|ACA|T TG|GAT|TAC|AAC|AAA|CGA|GCA|ACT|GAG|GGG|ATA|TTA|CCA|AGA|AGA|ATA|CAA|AGC|ACC|ATG|TCT GTG|GCC|CCT|ACA|TCT|ATG|GTT|CCT|CAA|GCC|GGT|CCA|ATG|ATG|CAG|CCA|GTT|TCT|GGG|GAC|AT T|CCC|AAT|GGC|CTG|AGC|AAT|TCG|CCC|ACT|CTG|GAG|GAG|CAC|ACC|AGC|TCG|GTG|TCT|TTG|GCG| ACG|ATT|CCG|AAC|TCA|AAC|TGC|TTG|GGA|AGG|AGC|AGA|GGG|CCC|TGC|AAC|AGC|AGA|CCT|TGG|TC C|CCT|TCA|CTT|TGG|GTG|ACA|GTC|TTT|CAG|GTC|TGG|AGG|CCA|GCA|TTG|CAG|ACC|TTA|ACA|ACC| CCT|CTC|CCT|CAA|TGG|ATT|CCC|TGA|TTG|GTG|GAG|TCG|ATC|CCA|ATC|TTT|TCC|CCT|TAA|AAA|CA G|AGG|ACT|TTT|CTC|CCA|TGA|TTA|AAG|GCG|ATA|TGG|ACC|TTG|ACC|AAG|ATT|CCT|TTG|GAC|ACA| TTG|GGA|AAG|ATG|TTG|ATG|TTG|GCA|ATC|ATA|AGC|TCT|TTA|GTG|ACAACACTCTGGACCTCCTGCAGGAC TTTGAGCTGGATGGATCGCCATCAGACTTCTACGTTGCTGACGATGCGTTTCTCTCCACCATAGGTGAAGATGCTCTCCTTT CAGAGCTGCCGACAAATTTAGACAGGGACTCGAAGGCTGCGGTTTCCGGGAGCAACACGCTCAATGGCACAGCTTCTTCCAG CCTCAGCACAGCCAACACCAGCATCTTGCCCAATATAAAGGTGGAGAAAGACTCTATAATCCAGCTGTGCACCCCAGGGGTC ATCAAACAGGAGAACACCGGTGCGAGCTATTGTCAAGGAGGGCTCCACAGCACCCCCATTAACATATGCGGGGTCACCACTT CAAGCGGACAGAGCTTCCTCTTTGGGAACAGCTCGCCCACAGCTGTCGTCGGTCTGCAGAAAGATCAGAAGCCGGACTTTAA CATGTACACCCCTCTGACCTCTTCAGGAGATGGTTGGAGCAGGAGCCAGGGCTTTGGGAATGTCAGTGGAATGCAGCAGAGG GCCAGTTTATGCTTTTCCAAAAACTTCTCAAGCAGCCCCTATTCCAG

NR3C1: GGTGTCCTCCATCTTTGGCGACGATTCCGAACTCAAACTGCTTGGGAAGGAGCAGAGGGCCCTGCAACA ∆WT-7: GGTGTCTTTGGCGACGA------TTCCGAACTCAAACTGCTTGGGAAGGAGCAGAGGGCCCTG

Figure 2: Sequence Verification of GR mutants: The target for CRISPR/Cas9 mutagenesis was designed in exon 2 (first coding exon) of the nr3c1 (glucocorticoid receptor) gene on chromosome 14. Lime green denotes the start codon, and subsequent infame stop codons. Red denotes the forward and reverse primers for fragment analysis. Yellow denotes the sgRNA target sequence. Purple and light blue denote the insert and wildtype (WT) sequences, respectively. Dark green denotes the premature stop codon in the mutant sequence.

Figure 3: Sequence analysis: Representative results for genotyping the F1 generation of nr3c1 mutant fish. Fin clips are taken from each individual fish and the region surrounding the sgRNA target amplified with sequence-specific primers (Table 1). A) wildtype fish with two wildtype alleles and B) heterozygote mutant with a wildtype and mutant allele. Red peaks are thymine (T), black peaks denote guanosine (G), blue peaks denote cytosine (C), and green peaks denote adenosine (A).

Figure 4: Representative results for fragment analysis of fluorescent PCR. Fragment analysis of the F2 generation. Sequence-specific primers amplify the region of interest surrounding the sgRNA target for the nr3c1 (glucocorticoid receptor) gene. The primer amplified the region in the presence of a FAM conjugated primer. The wildtype fragment is 325 bp, whereas the mutant target has 7 base pair deletion (318 bp). A) wildtype B) heterozygote and C) homozygote mutant. This method of analysis is performed in both the F0 and F2 generations.

Mineralocorticoid Receptor:

The target for in exon 2 of the nr3c2 gene on chromosome 1 also resulted in a complex mutation. Mosaic founder fish were outbred to wildtype fish, to limit off-target effects of inbreeding mosaic founders and sequence analysis was performed (Fig. 5, 6). The target for MR resulted in a -5 bp deletion and a +13 bp insertion, resulting in a net change, +8 bp deletion (Fig 5). Breeding similar mutations together (i.e. +8bp x +8bp fish), resulted in progeny displaying typical mendelian rations (25% homozygous mutant, 25% WT and 50%, heterozygous). Genotype was confirmed with fragment analysis (Fig. 7). Lack of protein expression via western blots confirmed a lack of GR protein (Chapter 3).

NR3C2 - Exon #2:

GTTTATCAAATGACCCAGGTATG|GAG|ACT|AAA|AGA|TAC|CAA|AGT|TGC|TGC|GAA|GGG|GCG|AAA|GCA|GAA|AA C|AAA|TGG|GCA|CAG|ATG|CCA|AAC|ACC|GCG|GAC|TAT|ATC|TGC|TCT|GCA|GAA|GAG|AGT|TTG|ACA|AAC| AGT|GAT|GTG|CTC|ATG|GAT|ATT|GTT|AAC|TCG|AGC|AAT|ACC|CCT|AAT|GTG|CAA|TCT|ATC|TGC|AAG|GA C|AAC|AAC|TTT|AAG|AAT|ACA|GAG|GCA|ACA|ATG|ATT|AGA|GTT|AAC|CAG|AAT|CAA|CCA|TTA|CCT|TTC| CCC|ACC|TTC|AAC|AAC|TGC|TTC|AAC|AAC|CGC|AAG|TCA|GAA|ACA|GAC|TCA|AAG|GAA|CTT|TCC|AAA|AC A|GTT|GCA|GAG|TCA|ATG|GGC|TTG|TAC|ATG|AAT|GCT|GCC|AGA|GAG|GCA|GAC|TTT|GGC|TTT|TCC|CAA| CAA|GGT|GCT|GCG|GGA|GGG|CAG|GG|TAG|CCC|AGG|GAA|ACT|GTA|CCC|TCT|TTG|TGG|GAG|AAC|AAA|TGA |AGA|CAG|CCA|GGC|CAA|AAC|AAG|TGG|GAG|CCC|AAA|GAT|GAA|AG|CTC|CCC|CTG|CTT|CAT|TCC|CAC|CT G|GGG|CCC|AAC|TAC|CCA|ATG|GTG|GAC|CTC|AGG|AGT|GTG|CTG|TGG|TCT|CTG|ACT|CAG|TGC|CCT|CTG| CTC|TGG|TCA|CTG|CCC|TTT|CGT|CCA|GCACTGATGGGTCCTGTCCCATGTCTAGTCCTACCGGACACAACATGGTATCAT CCACCACCAGCCCCACTTACTTTGACTCTGATTGTCCGACGCTAGATTCTGCGACCAGCAGCTTGACCCATTGTCAACACACCAGC CCCAACATTTGCAGCCCAGTAAAGTCCAGCATTGTGGGGTCACCTCCACTGCCTAGTCCCCTCAGTGTAATGAAGTCTCCGGTTTC CAGTCCTCATAGTATAGGCAGCGTGAGGTCACCGCTTTCCTGTAACACCAACATGAGGTCATCTGTCTCAAGCCCAACAACAAATG GAGGCAATACCTGTAATATTAAACCATCCATTTCCAGCCCGCCCACTGCTGGCAGCATGTCCATGTCCAGTCCCAGGAACTCATCC AGGGGTTTCTCAGTCTCCAGCCCACCCAGTGGATTGGGCCTAGTGCAAAATGATGTCAACAGCCCTGAGAGCCGAGAGCACGACTT CAAGGGATTCGAGTTCCCGAAGGTAGAGAATGTGGATGGGGAGATCTTTAACATCGGCTTGGATGCTATGGGAGTGGCTAAATATA TCAAAAATGAACCTGGCACTGATTTCAGGAGTATGTGCCTAGGCAGCAGCAAGAGTGCCATGTCAAATTCACCTTTCGTAACTCAC ATTAAGACCGAGCCAAACAGAGAGGTGACTTGCTCGAACCTCCAGTTTGCTGAACCGCAGCACTCTCTGGGCTGCTTTCCTTCGAC AGAGACCACATACTTGTCTTTGAGGGATAATATTGACGAATATAGTCTTTCTGGGATCTTGGGACCTCCTGTTTCGTCTCTGAATG GCAACTATGAACCTGGTGTGTTTCCTAACAATGGCCTGCCCAAGGGGATTAAACAGGAAACCAATGATGGCAGCTATTACCAAGAG AATAATAATGTGCCCACTTCGGCTATTGTTGGCGTTAATTCAGGTGGACATTCATTTCATTACCAGATTGGAGCGCAAGGAACAAT GTCGTTTTCACGCCACAATTTGAGGGACCAGACAAACCCCTTGTTGAATCTAATTTCTCCAGTTACTGGATTAATGGAGACGTGGA AAACTCGCCCAGGCCTGTCACAGGGGCCCCTCACTGCTAGAGGGGACGGGTATCCAGGCGCAGTCTGCCTTACAGAAAACATGGAA AG

∆WT+8 mutant: GTTTATCAAATGACCCAGGTATG|GAG|ACT|AAA|AGA|TAC|CAA|AGT|TGC|TGC|GAA|GGG|GCG|AAA|GCA|GAA|AA C|AAA|TGG|GCA|CAG|ATG|CCA|AAC|ACC|GCG|GAC|TAT|ATC|TGC|TCT|GCA|GAA|GAG|AGT|TTG|ACA|AAC| AGT|GAT|GTG|CTC|ATG|GAT|ATT|GTT|AAC|TCG|AGC|AAT|ACC|CCT|AAT|GTG|CAA|TCT|ATC|TGC|AAG|GA C|AAC|AAC|TTT|AAG|AAT|ACA|GAG|GCA|ACA|ATG|ATT|AGA|GTT|AAC|CAG|AAT|CAA|CCA|TTA|CCT|TTC| CCC|ACC|TTC|AAC|AAC|TGC|TTC|AAC|AAC|CGC|AAG|TCA|GAA|ACA|GAC|TCA|AAG|GAA|CTT|TCC|AAA|AC A|GTT|GCA|GAG|TCA|ATG|GGC|TTG|TAC|ATG|AAT|GCT|GCC|AGA|GAG|GCA|GAC|TTT|GGC|TTT|TCC|CAA| CAA|GGT|GCT|GCG|GGA|GGG|CAG|GGT|AGC|CCA|GGG|AAA|CTG|TAC|CCT|GGA|AAC|TGT|ACC|CTG|GGA|GA A|CAA|ATG|AAG|ACA|GCC|AGG|CCA|AAA|CAA|GTG|GGA|GCC|CAA|AGA|TGA|AAG|CTC|CCC|CTG|CTT|CAT| TCC|CAC|CTG|GGG|CCC|AAC|TAC|CCA|ATG|GTG|GAC|CTC|AGG|AGT|GTG|CTG|TGG|TCT|CTG|ACT|CAG|TG C|CCT|CTG|CTC|TGG|TCA|CTG|CCC|TTT|CGT|CCA|GCA|CTG|ATG|GGT|CCT|GTC|CCA|TGT|CTA|GTC|CTA| CCG|GAC|ACA|ACA|TGG|TAT|CAT|CCA|CCA|CCA|GCC|CCA|CTT|ACT|TTG|ACTCTGATTGTCCGACGCTAGATTCT GCGACCAAGCTTGACCCATTGTCAACACACCAGCCCCAACATTTGCAGCCCAGTAAAGTCCAGCATTGTGGGGTCACCTCCACTGC CTAGTCCCCTCAGTGTAATGAAGTCTCCGGTTTCCAGTCCTCATAGTATAGGCAGCGTGAGGTCACCGCTTTCCTGTAACACCAAC ATGAGGTCATCTGTCTCAAGCCCAACAACAAATGGAGGCAATACCTGTAATATTAAACCATCCATTTCCAGCCCGCCCACTGCTGG CAGCATGTCCATGTCCAGTCCCAGGAACTCATCCAGGGGTTTCTCAGTCTCCAGCCCACCCAGTGGATTGGGCCTAGTGCAAAATG ATGTCAACAGCCCTGAGAGCCGAGAGCACGACTTCAAGGGATTCGAGTTCCCGAAGGTAGAGAATGTGGATGGGGAGATCTTTAAC ATCGGCTTGGATGCTATGGGAGTGGCTAAATATATCAAAAATGAACCTGGCACTGATTTCAGGAGTATGTGCCTAGGCAGCAGCAA GAGTGCCATGTCAAATTCACCTTTCGTAACTCACATTAAGACCGAGCCAAACAGAGAGGTGACTTGCTCGAACCTCCAGTTTGCTG AACCGCAGCACTCTCTGGGCTGCTTTCCTTCGACAGAGACCACATACTTGTCTTTGAGGGATAATATTGACGAATATAGTCTTTCT GGGATCTTGGGACCTCCTGTTTCGTCTCTGAATGGCAACTATGAACCTGGTGTGTTTCCTAACAATGGCCTGCCCAAGGGGATTAA ACAGGAAACCAATGATGGCAGCTATTACCAAGAGAATAATAATGTGCCCACTTCGGCTATTGTTGGCGTTAATTCAGGTGGACATT CATTTCATTACCAGATTGGAGCGCAAGGAACAATGTCGTTTTCACGCCACAATTTGAGGGACCAGACAAACCCCTTGTTGAATCTA ATTTCTCCAGTTACTGGATTAATGGAGACGTGGAAAACTCGCCCAGGCCTGTCACAGGGGCCCCTCACTGCTAGAGGGGACGGGTA TCCAGGCGCAGTCTGCCTTACAGAAAACATGGAAAG

MR WILDTYPE: GGGTAGCCCAGGGAAACTGTACCCTCTTTGTGGGAGAAC MR ∆+8 bp: GGGTAGCCCAG------GGAAACTGTACCCTGGGAGAAC

Figure 5: Sequence Verification of MR mutants: The target for CRISPR/Cas9 mutagenesis was designed in exon 2 (first coding exon) of the nr3c2 (mineralocorticoid receptor) gene on chromosome 14. Lime green denotes the start codon, and subsequent infame stop codons. Red denotes the forward and reverse primers for fragment analysis. Yellow denotes the sgRNA target sequence. Purple and light blue denote the insert and wildtype (WT) sequences, respectively. Dark green denotes the premature stop codon in the mutant sequence

Figure 6: Sequence analysis: Representative results for genotyping the F1 generation of nr3c2 mutant fish. Fin clips are taken from each individual fish and the region surrounding the sgRNA target amplified with sequence-specific primers (Table 1). A) wildtype fish with two wildtype alleles and B) heterozygote mutant with a wildtype and mutant allele. Red peaks are thymine (T), black peaks denote guanosine (G), blue peaks denote cytosine (C), and green peaks denote adenosine (A).

Figure 7: Representative results for fragment analysis of fluorescent PCR of MR mutants. Sequence-specific primers amplify the region of interest surrounding the sgRNA target for the nr3c2 (mineralocorticoid receptor) gene. The primer amplified the region in the presence of a HEX conjugated primer. Both the primers and the target can be found in Table 1. The wildtype fragment is 166 bp, whereas the mutant target has 8 base pair insertion (174 bp). A) wildtype B) heterozygote and C) homozygote mutant.

Discussion:

Mutant lines are an invaluable tool to understand gene function in both larvae and adults; however, the ubiquitous loss of some genes is embryonically lethal. This makes it difficult to assess gene function in adult organisms. Both ubiquitous GR and MR knockouts are lethal in mammals (Berger et al., 1998; Cole et al., 2001). Mice with a ubiquitous GR knockout die in utero due to delayed lung maturation (Cole et al., 2001). MR knockout mice have normal prenatal development, but during the first week of life, they develop symptoms of pseudohypoaldosteronism leading to mortality by 10d (Berger et al., 1998). Therefore, almost all the information reported on the function of GR and MR in mice has been in conditional knockouts (Appendix I: Table 1). The GRdim/dim mice are limited in their ability to dimerize and induce target gene transcription; however, they can still engage in protein-protein interactions with other transcription factors (Opherk et al., 2004; Tronche et al., 2004). While there are several existing ways to create conditional knockouts such as Cre/loxP, GAL4/UAS system, and Tol2 transposable elements, the use of targeted control of gene expression is limited in zebrafish (Halpern et al., 2008; Carney & Mosimann, 2018). To the best of our knowledge, there has been no conditional knockout of GR and MR created in zebrafish. The strength of the tissue-specific knockouts is to both isolate compensatory mechanisms of other tissues as well as to determine the tissue-specific importance of each receptor. In mice that had a liver-specific knockout of GR, there was compensation by the kidneys of genes involved in gluconeogenesis and glycogen metabolism (Bose et al., 2016). Furthermore, when muscle GR was inactivated, there was a decrease in adipose tissue (Shimizu et al., 2015), which is likely also a compensation, as liver GR knockout mice report increased adipose tissue (Bose et al., 2016). Tissue-specific knockouts are also important in understanding what the function of CR activation is in tissues where a physiological role is less obvious (Appendix I: Table 1). For example, despite the fact that GR is expressed in the distal nephron in mice, inactivation in this tissue resulted in no discernible difference in the renal function of mice (Goodwin et al., 2010). The use of CRISPR/Cas9 to create tissue-specific knockouts in zebrafish is accomplished by driving Cas9 under tissue- specific promoters, which was shown using the gata1 promoter to inactivate urod gene in erythrocytes (Ablain et al., 2015). Overall, the creation of our ubiquitous GR and MR mutants may only be the first step in understanding the functional relevance of CR activation.

In addition to tissue-specific knockouts, the knock-in of fluorescent proteins to create reporter lines may be equally informative of CR activation and the subsequent physiological role. Most frequently, the use of transposable elements, Tol2 technology, has been utilized to create transgenic reporter lines (Kawakami et al., 1998). In this technique, only 50-70% of the injected fish are germline founders (Kawakami, 2007). However, in combination with the Gal/UAS system, reliable gene expression under the control of a variety of regulatory elements can be achieved (Suster et al., 2009). Indeed, this technology was used to create stable transgenic zebrafish for monitoring GR activity by incorporation of eGFP (Benato et al., 2014; Krug et al., 2014) downstream of GRE sites. These fish showed strong fluorescence signals in many GC responsive organs (Benato et al., 2014; Krug et al., 2014). In zebrafish, incorporation of foreign DNA into the germline has been a challenge and it is here that CRISPR/Cas9 may be able to increase the efficiency (Kimura et al., 2014). Early studies used a donor plasmid which was digested concurrently with the genomic DNA and resulted in the incorporation of the donor plasmid into the genome through NHEJ (Auer et al., 2014). Follow-up work incorporated a reporter gene downstream of an hsp70 promoter which could be incorporated using the same concept as Auer et a., (2014), upstream of a target gene (Kimura et al., 2014). Of the injected founders with strong fluorescent expression, 60% had germline transmission. When employing HDR, there is a strong inverse relationship between the knock-in efficiency and the distance from the modification site as any modifications greater than 20 nt will result in only a 20-30% efficiency (Paquet et al., 2016). The design of the homology arms is also important for the success of HDR. Oligos of 36-90 nt which were asymmetric were superior to all other designs (Richardson et al., 2016). The oligos are anti-sense to the PAM containing (non-target) strand, which can separate from the Cas9-sgRNA ribonucleoprotein and become available to bind and facilitate HDR (Richardson et al., 2016). The use of CRISPR/Cas9 to create any transgenic strain of zebrafish related to stress axis function has not yet been done. However, tagging GR with a fluorescent protein such as GFP is an ideal next step in examining the intracellular location of the receptor to determine the real-time, non-genomic signalling dynamics of CRs (Das et al., 2018).

CHAPTER 3: THE MINERALOCORTICOID RECEPTOR IS ESSENTIAL FOR

STRESS AXIS REGULATION IN ZEBRAFISH LARVAE

Introduction

The primary role of the mineralocorticoid receptor (MR) in mammals is to mediate the effects of aldosterone in regulating fluid balance and potassium homeostasis. It is essential for survival in mammals as MR-null mice die 1-2 weeks postnatally from renal salt wasting and hyperkaliemia (Cole & Young, 2017). The physiological importance of MR is evident from its persistence in vertebrate evolution. MR exists in every major vertebrate clade, and even the agnathans have a CR, which is thought to be an ancestral form of MR, while the GR ortholog is first seen in the elasmobranch lineage (Baker et al., 2013). Despite its persistence, there is no known physiological role for MR in ray-finned fish (Baker & Katsu, 2017). Also, teleosts lack aldosterone, and glucocorticoid appears to mediate most of the changes in iono- and osmo- regulatory functions by activating glucocorticoid receptor (GR) signalling (Cruz et al., 2013). In mammals, in addition to aldosterone, the glucocorticoid is also a major ligand of MR; however, almost all the mineralocorticoid functions are mediated by aldosterone-MR signalling (Shibata, 2017). The high affinity of cortisol for MR means that aldosterone-mediated effects occur in tissues where 11β-HSD2, a key enzyme that breaks down cortisol, is prevalent. Tissues that lack 11β-HSD2, but still contain MR, suggests that this receptor also has an extra- mineralocorticoid role in mammals and activated by glucocorticoid (Berger et al., 2006; Rozeboom et al., 2007). Fish MR can be transcriptionally activated by several 3-ketosteroid hormones, including cortisol, 11-deoxycortisol, corticosterone and 11-deoxycorticosterone, suggesting its possible physiological significance(Baker & Katsu, 2017). Cortisol is the primary glucocorticoid in teleosts and this hormone is released during stress to promote energy substrate mobilization and metabolic recovery post-stressor (Mommsen et al., 1999; Charmandari et al., 2005). Vertebrates share a highly conserved corticosteroid stress response that is central to stress adaptation (Charmandari et al., 2005). As in mammals, activation of the hypothalamus-pituitary- interrenal (HPI; analogous to the HPA) axis in fish commences with the release of corticotropin- releasing hormone (CRH) from the hypothalamus, and this peptide stimulates the release of adrenocorticotropic hormone (ACTH) from the pituitary into the circulation (Charmandari et al., 2005)(Wendelaar Bonga, 1997). ACTH binds to the melanocortin 2 receptor (MC2R) on the interrenal cells (analogous to the adrenal cortex in mammals), distributed in the head kidney

region, stimulating the biosynthesis of cortisol in teleosts (Faught et al., 2016a). HPI axis activation by stressors elevates circulating cortisol levels, which facilitates glucose mobilization to increase energy availability in target tissues and restore homeostasis. The cortisol-driven molecular programming of HPI axis development in has focused only on GR signalling in fishes (Pikulkaew et al., 2010; Griffiths et al., 2012; Nesan et al., 2012; Ziv et al., 2013; Wilson et al., 2013; Facchinello et al., 2017). Although an MR knockout was generated in medaka fish (Oryzias latipes), a role for this receptor in HPI axis functioning was not addressed (Sakamoto et al., 2016). In mammals, the ratio of GR: MR signalling is thought to play a role in the stress- related behaviour (De Kloet, 2014), and the MR function may be brain region-specific (Joëls & de Kloet, 2017). Similarly, loss of MR in medaka also altered adult behaviour, suggesting a conserved role for MR in regulating stress-related behaviour (Sakamoto et al., 2016). To determine a physiological role for MR signalling in teleosts, and given the ancient origin of this receptor, we tested the hypothesis that MR signalling regulates the highly conserved stress axis function. To this end, we generated ubiquitous MR-/- and GR-/- knockouts in zebrafish (Danio rerio) using CRISPR/Cas9 mutagenesis. Zebrafish is an ideal model for loss- of-function studies because, unlike other teleosts with paralogs for GR (Alsop & Vijayan, 2008), it only has a single MR and GR in the genome. Our results for the first time highlight a key physiological role for MR signalling in fish.

Materials and Methods

Zebrafish maintenance

Adult zebrafish (Tupfel long fin (TL) strain) were maintained on a recirculating system with a 14:10 light: dark cycle (Pentair Aquatic Habitats, Florida, USA) All experimental protocols were approved by the Animal Care and Use Committee at the University of Calgary Animal Care Committee (AC17-0079) and were in accordance with the Canadian Council on Animal Care guidelines. Water was maintained at 28.5°C, pH 7.6, and 750 µS conductivity, and 10% of the water was exchanged daily. Animals were fed twice daily with Gemma micro 300 diet (Skretting, USA) in the morning and live Artemia (San Francisco Bay Brand, USA) in the afternoon.

Zebrafish embryos/larvae were reared for days 0-5 dpf in a 28.5°C incubator in 10 cm Petri dishes (Sarstedt, USA) at a density of 100 embryos/dish in E3 embryo media (5mM NaCl, 0.17 mM KCl, 0.33mM CaCl2, 0.33 mM MgSO4 + 0.1 ppm methylene blue antifungal agent (Nusslein-Volhard & Dahm, 2002). Embryos/larvae were raised on a 14h light: 10h dark cycle, and 50% of the embryo media was replaced daily. Embryos/larvae were euthanized with MS222 (0.4 g/L) at 2, 24, 48, and 96 hpf for cortisol and transcript analysis and stored at -80 until use. Larvae that were being raised were fed starting at 5 dpf. Briefly 5dpf larvae were transferred to a 3L tank with 1.5 L of water and kept in temperature-controlled room at 28.5°C on a 14h light:10h dark cycle. Larvae were fed AP100 larval food (Ziegler, USA) and Gemma micro 150 (Skretting, USA) in the morning, and live Artemia in the afternoon and 50% of the water was changed daily until 15 dpf. At 15 dpf larvae were transferred to the recirculating system (Aquatic Habitats, Pentair, USA or Tecniplast, Italy) and maintained as described above. Generation of nr3c1 and nr3c2 null zebrafish Nr3c1 and nr3c2 null fish were generated as exactly as described previously (Varshney et al., 2016). Briefly, a zebrafish specific Cas9 containing plasmid (Addgene plasmid #46757) was linearized by digestion with XbaI (New England Biolabs) for 2h at 37°C. Cas9 mRNA was immediately generated using in vitro transcription using the T3 message kit (Life Technologies) according to the manufacturer’s directions. sgRNA targets were designed using (http://zifit.partners.org/ZiFiT/, or http://www.crisprscan.org/) the second exon (first coding exon) of both nr3c1 and nr3c2 (Ensembl genome browser). Primers were also designed to create a 200-300 bp amplicon surrounding the target region. Target sequences (20 bp) were tagged with a 5’T7 sequence (TAATACGACTCACTATA) for in vitro transcription and a 3’ complementary sequence (GTTTTAGAGCTAGAAATAGC) to the universal ultramer. Both the target utltramer were ordered as DNA oligomers, annealed and extended using the following parameters: 98°C, 2 min; 50°C, 10 min; 72°C, 10 min; 4°C hold. The resulting product provided a template for sgRNA synthesis by in vitro transcription using the T7 high yield RNA synthesis kit (New England Biolabs) according the manufacturer’s directions. RNA was cleaned and concentrated (Zymo RNA Clean and Concentrator kit, ZYMO, USA) prior to injection. Primers were also tagged for fluorescent PCR and fragment analysis. Forward primers were tagged with an M13,

and reverse primers were tagged with a PIG tail for fragment analysis. All oligonucleotides used as described above are listed below: Universal Ultramer (Varshney et al., 2016): 5’ AAAAGCACCGACTCGGTGCCACTTTTTCAAGTTGATAACGGACTAGC CTTATTTTAACTTG CTATTTCTAGCTCTAAAAC-3′ GR: TAATACGACTCACTATAGGAATCGTCGCCAAAGATGGGTTTTAGAGCTAGAAATAGC GR-FW primer: tgtaaaacgacggccagtAACAAACGAGCAACTGAGGG GR-RV primer: gtgtcttTTAAGGTCTGCAATGCTGGC MR: TAATACGACTCACTATAGGAAACTGTACCCTCTTTGTGTTTTAGAGCTAGAAATAGC MR-FW primer: tgtaaaacgacggccagtGGCTTGTACATGAATGCTGCC MR-RV primer: gtgtcttGGGCTCCCACTTGTTTTGGCC M13 Primer (FAM or HEX conjugated): FAM/HEX-TGTAAAACGACGGCCAGT (100 mM) F0: Embryos at the 1-2 cell stage were microinjected with 300 pg of Cas9 protein and 50 pg of sgRNA into the yolk and raised as described above. To assess the somatic activity of our target guide we amplified the region using fluorescent PCR and used fragment analysis to separate the size of the fragments using capillary electrophoresis. Briefly, the fluorescent conjugated M13 primer will be incorporated during PCR amplification and when run on a genetic analyzer with known size standards, it will provide a signal that is reflective of the amplicon size(Carrington et al., 2015; Varshney et al., 2015, 2016). Samples were prepared from single embryos by first extracting genomic DNA using the Extract-N-Amp kit (Sigma, USA). The resulting gDNA was diluted 10x and 1 µl was added to our PCR master mix (Taq DNA polymerase [1U/ml; Life Technologies 10342020], 1x Taq Buffer [Life Technologies], 0.2 uM primer mix [FW, RV +

M13], 0.2 uM dNTPs, 1.5 mM MgCl2) to a final volume of 10 µl. The PCR reaction was run under the following conditions: 95°C, 5min; 35 cycles of 94°C, 1min, 57°C, 1min; 72°C, 30 sec; followed by 72°C, 5 min. Samples were then sent to the University of Calgary DNA Sequencing facility for fragment analysis on an Applied Biosystems 373XL genetic analyzer, where they were separated based on size using capillary electrophoresis. The size of each fragment was analyzed using Peak Scanner. Once somatic activity was confirmed we outbred the fish to wildtype fish to determine germ-line transmission and reduce enrichment of off-target effects.

Several mutant alleles were recovered for each gene; however, only one mutation for each gene was characterized and described here. F1: Juvenile (2-month-old) F1 fish were fin-clipped from the caudal fin and genomic DNA was extracted as described above. Amplification of the target region with PCR was performed as described above (without the fluorescent primer). The resulting PCR product was cleaned up using ExoSap (Thermo Fisher Scientific) as per the manufacturer's instructions. Samples were sequenced by Sanger sequencing at the University of Calgary’s DNA sequencing facility. These heterozygous fish were bred as described previously (Varshney et al., 2016). The resulting progeny were F2. F2: Juvenile (2-month-old) F2 fish were fin-clipped and gDNA was extracted. Fluorescent PCR and fragment analysis was performed to genotype and subsequently sort out WT, heterozygous fish and homozygous fish. F3: The progeny (maternal-zygotic homozygous larvae) from F2 homozygous (zygotic) fish were used in all experiments. The use of maternal- zygotic mutants in this paper was designed to eliminate any contribution of the maternal glucocorticoid system on the ontogeny of the stress response. nr3c1 functional experiment: To determine whether the mutant nr3c1 homozygous fish (GR-/-) had a functional GR, 72 hpf WT and GR-/- larvae were transferred to a 6-well plate (Sarstedt, USA). Each well contained 20 larvae and 4 ml of E2 embryo media. Larvae were exposed to either a vehicle (0.05% ethanol) or cortisol (5 µg/ml) for 24h. 96 hpf larvae were euthanized in MS222 (0.3g/L; Sigma), collected in pools of 10, all media removed and stored at -80°C for later analysis. As there is no known target gene of MR in fish, functional analysis via transcript abundance of a target gene of nr3c2 could not be completed. Stress Experiment: The zebrafish larval stress response was assessed in WT, GR-/- and MR-/- 96 hpf larvae by subjecting fish to an acute swirling stressor (Alsop & Vijayan, 2008). Setup: Fish at 80 hpf were transferred to 50 ml Falcon tubes (10-20 fish/tube) containing 20 ml of E3 media and kept in a temperature (28.5°C) and light controlled incubator overnight. There was a separate tube for each time point (4 tubes/genotype of fish) and 10 larvae were pooled for an n of 1. This experiment was repeated on at least three different days from different clutches to ensure an appropriate sample size. Stressor: Fish at 100 hpf (time 0) were subjected to a swirling stressor.

Briefly, time 0 fish were removed from the incubator and immediately euthanized with an overdose of MS222 (0.4 g/L). The remainder of the larvae were transferred to a temperature controlled (28.5°C) orbital shaker. Larvae were vortexed at 250 rpm for 1 min and allowed to recover. Sampling: Larvae were sampled at 0, 5, 10, and 30 min post-stressor, and euthanized as described above. For cortisol quantification pools of 10 larvae are required for n=1. Therefore, euthanized larvae were sampled as pools of 10 in a microcentrifuge tube, E3 media removed and stored at -80°C for later analysis. Cortisol quantification For cortisol analysis during development, pools of embryos and larvae were disrupted by sonication (3s pulse, ~5x) in10 µl/larvae in 50 mM Tris buffer (pH 7.5) with a protease inhibitor cocktail (Roche Diagnostics, USA). Samples were centrifuged at 13,000xg for 1 min. The supernatant was removed and stored at -80°C until use. Cortisol was quantified using an ELISA as previously described (Faught et al., 2016b). Western blotting SDS-PAGE and Western blotting was performed as previously described (Faught et al., 2017). Briefly, samples were homogenized in 50 mM Tris buffer (pH 7.5) with a protease inhibitor cocktail (Roche, 04693116001). The homogenate was sonicated (3s pulse, ~5x) and centrifuged (13,000 xg for 2 min). The supernatant was removed, and the protein concentration was determined using the bicinchoninic acid (BCA) method using bovine serum albumin (BSA) as the standard. The homogenate was then diluted with Laemmli’s buffer (156.25mM Tris, 50% glycerol, 5% SDS, 0.0625% bromophenol blue and 25% 2-mercaptoethanol). Samples were stored at -20°C. Equal amounts of protein (40 µg) were separated on an 8% polyacrylamide gel and transferred to nitrocellulose membrane using a SemiDry transfer unit (BioRad). After transfer, membranes were blocked with a solution of powdered skim milk (5% w/v in TTBS (20 mM Tris, 300 mM NaCl, pH7.5 with Tween 0.1%) containing 0.02% sodium azide). The primary antibodies included anti-trout GR antibody (Sathiyaa & Vijayan, 2003) used at a dilution of 1:1000, and anti-zebrafish MR (Jeffrey et al., 2012) at 1:500 dilution. Antibodies were prepared in the blocking solution and membranes were incubated overnight at 4C. Membranes were then washed with TTBS (5min,3x) and incubated for 1h with secondary antibody (1:3000 goat anti rabbit IgG; Bio-Rad, 170-6515). Bands were detected with Clarity

Western ECL substrate (BioRad, 170-5061). Molecular mass of the bands was confirmed using a low range molecular weight marker (FroggaBio, Canada), and the specificity of anti-GR was confirmed using rainbow trout liver homogenate. Equal loading was confirmed using a CY3 conjugated anti-β-actin (Sigma, C5838). Transcript abundance Transcript levels of specific genes were measured by quantitative real-time PCR (qPCR). Total RNA was extracted from larvae using Ribozol reagent (VWR, Canada) according to the manufacturer’s instructions and quantified using a SpectraDrop Micro-Volume microplate (VersaMax, Molecular Devices, CA, USA). One microgram of RNA was treated with DNase I (Thermo Scientific, USA) to remove genomic contamination prior to cDNA synthesis using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, USA) according to the manufacturer’s protocols. Transcript levels were measured by qPCR in duplicate using gene- specific primers exactly as described previously (Faught et al., 2016b). See (Alsop & Vijayan, 2008) for primer specific sequences and annealing temperatures. Behavioural Analysis Light/Dark zebrafish larval behaviour was performed exactly as described previously (Best et al., 2017). Briefly, 80 hpf larvae were transferred to clear 96-well plates with lids and allowed to acclimate overnight (96 hpf at time of measurement). Analyses were performed in a temperature- controlled room (28.5°C). The movement of 96 hpf larvae was video captured and quantified using the Zebrabox infrared camera setup and tracking extension of the ZebraLab software system (Viewpoint life sciences, Montreal, QC, CAN). In all behavioural protocols, the animal colour was set to black and the background-subtracted detection threshold was set to 20. The integration period for movement data was set to 30s. The light-dark protocol consisted of alternating periods of light and dark every 7.5 min (450s). This 15 min light/dark cycle was repeated four times for a total of 60 min. The total distance moved (mm) was calculated every 30s integration period and was a sum of distance covered during inactivity, small movement (2 mm/s-10 mm/s) and large movement (>10mm/s) Statistics Data are shown as mean +/- SEM and statistical comparison analysed using Sigma Plot 13 (Systat Software, Inc). Ontogeny data were analysed using a two-way ANOVA (Holm-Sidak

post hoc). All data were transformed to meet the assumptions of normality and equal variance. Untransformed data are shown in all figures. A significance level of p<0.05 was used in all cases. Results Generating MR and GR knockouts in zebrafish To determine the roles of GR and MR in HPA axis programming and the development of stress- related behaviour, we generated homozygous GR-/- and MR-/- zebrafish mutants (Varshney et al., 2016). The GR-/- mutants had an 18 bp deletion and an 11 bp insertion (net -7 bp change) in exon 2 (n-terminal domain of the GR protein), which was 232 bp downstream of the start codon in gene nr3c1 (Fig. 1A). In silico analysis predicted that a premature stop codon at ~500 bp downstream of the start codon would result in a severely truncated protein. This was confirmed by western blotting (Fig. 1C), and the loss of function was indicated by the abolishment of the glucocorticoid-induced elevation in 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2) mRNA levels (Fig. 1D). The MR-/- mutants had -5 bp deletion and a +13 bp insertion (net + 8bp change) in exon 2, which was 414 bp downstream of the start codon in gene nr3c2 (Fig. 1B). This corresponds to amino acid position 140, which is in the n-terminal domain of the protein. In silico analysis predicted that a premature stop codon should be encountered ~70 bp downstream of the NGG site, resulting in a truncated protein ~500 bp downstream of the start codon. MR-/- knockout was confirmed by Western blotting that showed the absence of MR protein expression (Fig. 1E), but an MR-specific gene target is yet to be reported in teleosts.

Figure 1: GR and MR knockout in zebrafish: A) Schematic representation of the zebrafish GR (nr3c1) gene. Exons are shown in boxes, introns are denoted by lines. The arrow shows the position of a -7 bp deletion in exon 2. Detailed below the schematic, the protospacer- adjacent motif (PAM) is denoted in red and the target site is bolded. B) Schematic representation of the zebrafish MR (nr3c2) gene. Exons are shown in boxes, introns are denoted by lines. The arrow shows the position of a +8 bp deletion in exon 2. Detailed below the schematic, the protospacer-adjacent motif (PAM) is denoted in red and the target site is bolded. C) Western blotting with anti-trout GR (1:1000). Lane 1) WT zebrafish whole-body; Lane 2) GR-/- zebrafish whole-body; Lane 3) WT trout liver Lane. Red box denotes area of interest corresponding to GR protein at ~90 kDa. D) Transcript abundance of 11βhsd2; bars with different letters are significantly different (One- way ANOVA, p<0.05, n=4, bars are mean ± SEM). E) Western blotting with anti-zebrafish MR (1:500). Lane 1) WT trout liver Lane; 2) WT zebrafish liver Lane; 3) MR-/- zebrafish liver; Lane 4) WT zebrafish head; Lane 5) MR- /- zebrafish head. Red box denotes area of interest corresponding to MR protein at ~110 kDa. Representative image below each blot shows β-actin expression as a loading control.

MR larvae have normal cortisol levels, whereas GR larvae are hypercortisolemic Here we report distinct changes to the HPA axis development as a result of a GR or MR knockout in zebrafish (Fig. 2). While MR-/- embryos showed cortisol levels comparable to that of the WT over the sampling period, the cortisol levels in the GR-/- mutants were exceptionally high at 2 and 96 hpf compared to the other two groups (Fig. 2A). High cortisol levels at 2hpf in GR-/- reflects the maternal cortisol levels that were transferred to the embryo (Faught et al., 2016b). There was a drastic drop (p<0.001) in the GR-/- embryo cortisol content (2.7 ± 1.5 pg/embryo) at 24 hpf compared to 2hpf, and this level was significantly lower than the WT at 24 (p=0.050) and 48 hpf (p=0.036). The MR-/- embryos showed no decrease in cortisol levels at 24 hpf (12.70 ± 0.82 pg/embryo), and they had 2-fold higher cortisol levels compared to both GR-/- (p<0.001) and WT (p=0.013) at this time-point (Fig. 2A).

MR regulates genes involved in HPA axis activity during development To determine the possible mechanisms behind the altered cortisol profile, we measured the transcript abundance of key HPI axis genes. There were distinct changes in the transcript abundances of key genes in response to either GR or MR loss (Fig. 2B-G). CRH: Both GR-/- and MR-/- mutants have increased crh transcript abundance at 2hpf compared to WT (p<0.001 and p=0.003, respectively), with GR-/- embryos having a greater abundance compared to MR-/- (p=0.031). At 24hpf, loss of MR resulted in a greater crh mRNA abundance compared to WT embryos (p<0.001) and GR-/- embryos (p=0.009). At 24hpf, the crh mRNA levels were similar in the GR-/- and WT (p=0.158). At 48 hpf, there was no difference in the transcript abundance between the groups, but at 96 hpf, crh transcript level was 4-fold higher in the MR-/- compared to WT (p<0.001) and GR-/- (p<0.005) (Fig. 2B). ACTH: Proopiomelanocortin (POMC) is encoded by the pomca gene in zebrafish and is the precursor protein for ACTH, which is released upon CRH stimulation (Alsop & Vijayan, 2008). WT pomca transcript abundance increases over the first 120 hpf in zebrafish (Fig. 2C). At 2hpf, both GR-/- and MR-/- mutants have higher pomca transcript abundance compared to WT (p<0.001), with MR-/- having a greater abundance compared to GR-/- (p=0.028). There was no difference at 24hpf between any of the three groups as pomca mRNA levels decreased from 2hpf in both the GR-/- (p<0.001) and MR-/- embryos (p<0.001). At 48 hpf, all embryos had similar

transcript levels. However, by 96 hpf, pomca abundance in GR-/- larvae is greater than WT larvae (p=0.024), while MR-/- larvae had increased significantly from its levels at 48 hpf (p<0.001) and the transcript abundance was greater than both WT (p<0.001) and GR-/- larvae (p<0.011) (Fig. 2C). StAR: A key rate-limiting step in corticosteroid biosynthesis is the StAR protein, which shuttles cholesterol to the inner mitochondrial membrane(Stocco, 2001). During early development, star transcript abundance increases after hatch in zebrafish (Alsop & Vijayan, 2008) (Fig. 2D). The maternal deposition of star transcripts was not affected by the loss of GR or MR compared to the WT. Similarly, there was no difference between the different treatments at 24hpf, while at 48 hpf all groups had increased star mRNA levels compared to 24hpf (p<0.001), with GR-/- larvae having greater abundance compared to WT (p=0.009) and MR-/- larvae (p=0.010) (Fig. 2D). 11β-HSD2: Another key player in modulating cortisol levels during early development is the enzyme 11β-HSD2 that break down cortisol to its inactive form cortisone for elimination (Seckl & Walker, 2014). As reported previously, 11β-hsd2 transcript levels are deposited in low amounts from the mother and increase by hatch (48 hpf; Fig. 2E) (Alsop & Vijayan, 2008). While there was no difference in 11β-hsd2 transcript abundance prior to hatch, in the mutants, there was a clear receptor-specific effect on this transcript profile after hatching in zebrafish (Fig. 2E). At 48 hpf, GR-/- have significantly higher 11β-hsd2 compared to MR (p<0.001), but not WT larvae (0.061). Transcript abundance of 11β-hsd2 in MR-/- mutants continue to increase at a greater rate compared to either WT or GR larvae, and by 96 hpf, MR-/- mutants have greater 11β-hsd2 transcript abundance compared to GR-/- larvae (p=0.005). GR and WT have similar levels of 11β-hsd2 mRNA at 96 hpf (p=0.082) (Fig 2E). GR and MR knockouts alter the developmental profiles of their respective receptors To test whether the dysregulation of cortisol levels during early development are intricately linked to GR and MR signalling, we also characterized the ontogeny of gr and mr transcript abundance in the mutants during early development (Fig. 2F-G). It should be noted that while the mr and gr are present in the respective knockouts, these are not indicative of protein levels (Fig. 1B, E). MR: In the WT, mr is deposited in low amounts and steadily increases over the first 4 days of zebrafish development (Alsop & Vijayan, 2008) (Fig. 2F). A loss of MR caused an increased

amount of mr mRNA at 24hpf compared to both WT (p<0.001) and GR embryos (p<0.001). GR and MR mutants had similar mRNA levels of mr at 48 hpf (p=0.672), which is significantly more than WT embryos (p<0.001). At 96hpf, mr mRNA levels were greatest in the MR mutant compared to the GR mutant (p<0.001) and WT (p<0.001) (Fig. 2F). GR: gr transcript abundance in WT embryos decreases until hatch when mRNA levels increase (Fig. 2G). A loss of GR caused substantially fewer gr transcripts to be deposited into the embryos compared to WT (p=0.002) and MR mutants (p=0.006). A loss of either receptor caused an increased in gr at 24hpf from 2 hpf (p<0.001), with MR having a greater amount of gr mRNA compared to WT and GR mutant embryos (p<0.001). At 48 hpf, there was no difference between GR mutants and WT in gr mRNA levels; however, the gr mRNA levels in MR was still greater than both GR (p=0.024) and WT (p=0.028). All larvae had similar gr transcript abundance at 96 hpf. GR and MR differentially regulate the glucocorticoid stress response To determine the integrity of the HPI axis and its capacity to elicit a cortisol stress response in the mutants, we subjected the larvae to an acute stressor and measured larval cortisol levels post- stressor exposure (Alsop & Vijayan, 2008). As expected, WT basal cortisol levels (16.6 ± 0.2 pg/larva) doubled to 35.5± 4.8 pg/larva at 5 min (p=0.007) and returned to resting levels at 30 min (p=0.468) post-stressor exposure (Fig. 3). The GR-/- mutants did not elicit a cortisol stress response and the steroid levels were consistently ~4-fold higher than the WT at all time points (p<0.001) post-stressor exposure (Fig. 3A). The MR-/- larvae were able to respond to an acute stressor by elevating cortisol levels, but there was a delay in HPI axis activation compared to the WT (Fig. 3B). There was no difference in the basal cortisol levels between WT and MR-/- prior to stress (p=0.554) or in the magnitude of the stress response post-stress compared to the WT. However, the loss of MR led to a delay (5 min) in the attendant rise in cortisol post-stress as MR- /- mutants experienced peak cortisol levels only at 10 min post-stressor exposure (p=0.003; Fig. 3B). Also, unlike the WT, the cortisol levels did not return to resting levels by 30 min (p=0.020) in the MR-/- larvae (Fig. 3B).

Figure 2: GR and MR regulate genes involved in HPA axis activity during development: (A) the cortisol ontogeny of WT, GR-/- and MR-/- embryos and transcript abundance of HPA axis intermediates at 2, 24, 48, 96 hpf. B) corticotropin-releasing hormone (CRH), C) Proopiomelanocortin (POMC), the gene for the precursor protein to ACTH, D) Steroidogenic acute regulatory protein (StAR), a rate-limiting step in steroid hormone biosynthesis, E) 11β-hydroxysteroid dehydrogenase which catalyzes the breakdown of cortisol to biologically inactive cortisone F) the mineralocorticoid receptor (MR) and the glucocorticoid receptor (GR). A significant interaction was detected in all two-way ANOVAs, p<0.05. Significant time effects within treatment groups are indicated by different letters (WT: A, GR-/-: a, MR-/-: a); significant treatment effects within each time points are indicated above each time point using legend symbols. All data points are mean ± SEM, n=4–6 (pools of 10).

Figure 3:GR and MR are necessary for the glucocorticoid profile post-stress: (A) Loss of GR results in hypercortisolemic larvae at 96 hpf which are unable to respond to a swirling (250 rpm) stressor. Their cortisol levels are beyond what could be induced in the WT larvae. B) Loss of MR causes a delay in the timing of peak cortisol levels compared to WT at 96 hpf but does not change the magnitude of the stress response. MR-/- larvae are unable to restore cortisol levels to resting. A significant interaction was detected in both two-way ANOVAs, p<0.05. Significant time effects within treatment groups are indicated by different letters; significant treatment effects are indicated by an asterisk. All data points are mean ± SEM, n=4-6 (pools of 10-15).

Both GR and MR are required for stress-related larval behaviour

Given the differences in the physiological stress response, we also evaluated whether the stress- related behaviour was altered in the CR mutants. Zebrafish have become an increasingly powerful model organism for translational neuroscience research (Stewart et al., 2014), and glucocorticoids are known to cause hyperactivity in larvae (Best et al., 2017; Best & Vijayan, 2018). To determine if CRs are involved in this stress-induced larval behaviour, we subjected the larvae to a light/dark stimuli (Best et al., 2017) (Fig. 4A). As shown previously, larvae treated with cortisol showed a higher activity in the light compared to the WT, but there was no difference in locomotor activity in the dark (Best et al., 2017) (Fig. 4A-C). This cortisol-induced behavioural response was completely abolished in GR-/-, despite being hypercortisolemic. Also, MR-/- larvae did not show this hyperactive response, however, as MR (-/-) mutants have cortisol levels similar to that of the WT, we treated MR-/- with cortisol to determine whether cortisol- induced hyperactivity was GR-mediated. Interestingly, even with cortisol, we were not able to recover the cortisol-induced hyperactivity (Fig. 4A-C). To determine whether this increase in activity was a reflection of increased boldness (decreased anxiety), we also looked at thigmotaxis, or the tendency to remain close to the periphery of the arena. This is used as an index of anxiety in both mice (Simon et al., 1994) and zebrafish (Schnörr et al., 2012). As previously reported fish treated with cortisol have decreased thigmotaxis (p=0.027), displaying an increased boldness (Best et al., 2017) (Fig. 4D). A loss of MR also caused the larvae to be bolder, compared to WT (p=0.045), suggesting that MR may be a primary mediator of anxiety behaviour. However, treatment with cortisol reduced the boldness of the MR-/- larvae similar to that of WT, suggesting GR may also be necessary for anxiety-related behaviour as a loss of GR caused a partial recovery of baseline WT anxiety (80% of total time in the periphery).

Figure 4: GR and MR are necessary to mediate stress-related behaviour: A) Behavioural profiles of 96 hpf zebrafish larvae. Mean activity in light and dark is represented, where the activity of 96hpf larvae is expressed as total distance moved during each 30 s recording bin. The total recording period was 1h with alternating light periods of 7.5min each (B) Total combined distance moved (mm) in the light and C) total combined distance moved in the dark period. Representative tracking data is included under each bar. Bars represent mean ± SEM (n=24). Different letters indicate significant differences (One-way ANOVA, p<0.05). (D) Thigmotaxis: Arenas (wells) were divided into inner and outer zones, and the propensity of 96 hpf larvae to stay close to the outer arena wall is expressed as % of total distance travelled (inner and outer arena) in 30min. Bars represent mean ± SEM (n=12). Different letters indicate significant differences. Representative paths for thigmotaxis are shown below each bar.

Discussion

While the role of GR is well defined in teleosts, nothing is known about the physiological role directly mediated by MR signalling. Here we report that MR is also a key player in affecting the stress response of lower vertebrates. By comparing two corticosteroid receptor (CR) knockout zebrafish lines, we reveal distinct and complementary roles for MR and GR signalling in the development of the stress axis. The results indicate that both GR and MR signalling are involved in the negative feedback regulation of cortisol during stress. GR is essential for maintaining the steady-state resting levels of glucocorticoids, whereas MR is involved in both the perception of stress and cortisol homeostasis after an acute stressor (Fig. 3). Our results also suggest a role for MR in the repression of key genes involved in the HPI axis (analogous to the HPA axis in higher vertebrates) function (Wendelaar Bonga, 1997), while the role of GR may include the regulation of cortisol metabolism (upregulation of 11βHSD2). The distinct roles of GR and MR in regulating HPI axis programming are particularly important given that dysregulated cortisol levels during embryogenesis lead to developmental defects in fish, including impaired HPI axis activity later in life (Harris & Seckl, 2011; Nesan & Vijayan, 2013a; Wilson et al., 2016; McGowan & Matthews, 2018). Furthermore, both receptors mediate cortisol-induced behaviour in larvae indicating that CRs play a central role in behavioural outcomes in fish (Fig. 4). Although this is well established in mammals (Berger et al., 2006; Rozeboom et al., 2007), a similar functional role of MR signalling in shaping stress-related behaviour in fish was lacking (Sakamoto et al., 2016). As behaviour is an essential component of the stress-coping mechanism (De Kloet, 2014), a loss of either receptor would have maladaptive consequences to an organism’s ability to adapt to stress, and this may explain the persistence of MR and GR in the vertebrate phylogeny. Together, these results provide a strong physiological underpinning for the evolutionarily conserved role for MR in the regulation of stress axis function in vertebrates. In fish, stress and the attendant rise in cortisol levels can be transferred to the developing oocytes, which can be detrimental to proper embryogenesis and development of the stress axis (Nesan & Vijayan, 2013a). As such, cortisol deposited into the oocyte is tightly regulated by cortisol-mediated upregulation of 11βHSD2, a key enzyme that inactivates cortisol to cortisone (Alderman & Vijayan, 2012; Faught et al., 2016b). The exceptionally high cortisol levels in the

GR-/- mutants at 2 hpf reflect the transfer of this steroid from the hypercortisolemic mother (Faught et al., 2016b). This may be due to a lack of oocyte buffering by 11β-hsd2 in the GR mutants (Fig. 1D), as GR signalling upregulates this gene expression in teleosts (Alderman & Vijayan, 2012; Faught et al., 2016b). The role of 11βHSD2 in regulating oocyte cortisol levels is similar in function to the protective role this enzyme has in the developing fetus in mammals (Seckl & Walker, 2014), indicating a critical role for GR, but not MR in maintaining the steady- state cortisol levels in the embryos. While the dynamics of maternal contribution of hormones and transcripts were outside the scope of this paper, it would be an interesting avenue for future research to examine GR-regulated maternal depositions and their contributions to progeny development. This can be achieved by comparing the maternal-zygotic mutants that we have generated to the zygotic mutants, which would have partial maternal deposition from their heterozygous mothers. Despite the high maternal cortisol deposition in GR-/- embryos, this steroid is cleared rapidly, and the low cortisol levels are maintained for 24h. While the mechanism is unclear as 11β-hsd2 transcript levels are low in GR-/- larvae (Fig.2A), the excretion of this excess steroid may involve either diffusion (Tagawa et al., 2000), activation of other biotransformation pathways (Li et al., 2012) and/or upregulation of efflux transporters (Paitz et al., 2016). Except for this brief period, the GR, but not MR, mutants were hypercortisolemic and the steroid levels were beyond what could be induced by a stressor in the WT larvae (Fig. 2A; (Ziv et al., 2013). This suggests either a non-functional negative feedback mechanism (Figs. 2B-2G) and/or a lack of cortisol breakdown in the GR mutant (Fig. 3A). In contrast to GR, a loss of MR resulted in basal cortisol levels similar to WT (Fig. 2A), supporting a key role for GR signalling in maintaining the steady-state basal glucocorticoid levels during early development (Nesan & Vijayan, 2013a), and this may involve a tight negative feedback regulation in fish (Alderman et al., 2012). Our results indicate a key role for MR signalling in tightly regulating the molecular components of the HPI axis during early development. This is clearly evident from the lack of regulation of crh and pomca transcript levels during HPI ontogeny (Fig. 2B and 2C). Despite elevated crh and pomca transcript abundance in the MR mutants at 96hpf, the larvae were not hypercortisolemic suggesting rapid cortisol clearance. Indeed, 11βHSD2 is upregulated at this

time point and may be a primary cause for maintaining basal cortisol levels (Fig. 2E). The control of HPI axis intermediates by MR is further supported by the lack of changes in these transcripts in the GR-/- mutants, which has a MR. These results clearly indicate that GR and MR have distinct roles in HPI axis function; MR repressing HPI axis activation during development, while GR regulates basal cortisol levels. Also, the delayed cortisol response, as well as a dysregulated cortisol level after an acute stressor in the MR-/- mutants (Fig. 3B), suggests a physiological role for this receptor in the regulating stressor-mediated HPI axis activity. As vertebrates share a highly conserved corticosteroid stress response that is central to stress adaptation (Charmandari et al., 2005), this study highlights the necessity of having both a functional MR and GR for HPI axis development and stress signalling. In addition to HPI axis activity, the corticosteroid receptors are also involved in stress- related behavioural changes (Ziv et al., 2013). The cortisol-mediated activity of the larvae in the light and dark (Figs. 4A-C) were affected by the loss of GR or MR supporting a complementary role for both these receptors in stress-related behavioural outcomes (Fig 4). In addition, the lower thigmotaxis seen with cortisol in the wildtype was also seen in the MR-/- mutant, suggesting that GR may be the primary mediator of this behaviour (Fig. 4D). However, the partial recovery to WT behaviour after exposure to cortisol in the MR-/- suggests that the response is not as simple as either the presence or absence of receptors but involves a complex interaction of both receptors in mediating specific behavioural outcomes. Indeed, an often-overlooked aspect of CR signalling is the MR: GR [24,27], which may be involved in regulating emotional responses, including fear and anxiety-related behaviour (Hartmann et al., 2017) and hyperactivity to novel stimuli (Berger et al., 2006). Our results suggest a key role for GR and MR signalling in the behavioural responses mediated by stressor-mediated elevated cortisol levels. While the mechanism for receptor interaction in mediating these responses is unclear, one possibility is GR and MR heterodimerization and the resultant transactivation or transrepression of genes controlling behaviour (Ou et al., 2001). In conclusion, we report that a key physiological role for MR signalling in fish is to regulate stress axis function during early development. While it is currently unknown how a lack of corticosteroid signalling might impact the physiology of adult fish, perturbations in the glucocorticoid system in early life have been correlated with adult behavioural changes(Wilson

et al., 2015). Furthermore, any changes in the cortisol stress response may cause fundamental alterations in intermediary metabolism as this is a primary target of corticosteroid signalling in adults(Mommsen et al., 1999). We also highlight that the presence of GR alone is clearly not sufficient to mediate the physiological role of stress or stress-related behaviour in fish. As the glucocorticoid stress response is highly conserved in vertebrates and essential for stress adaptation (Charmandari et al., 2005), we propose that stress axis regulation is a potential early role for MR signalling in vertebrates. While other physiological roles may exist, the MR regulation of the larval stress response provides a functional underpinning to the early origin of this receptor in vertebrates (Baker & Katsu, 2017).

Acknowledgements:

This study was supported by the Natural Sciences and Engineering Research Council of Canada Discovery Grant to MMV. EF was supported by an NSERC post-doctoral scholarship. We would like to thank Dr. Raman Sood and Blake Carrington for their assistance with the CRISPR/Cas9 mutagenesis.

CHAPTER 4: POSTNATAL TRIGLYCERIDE ACCUMULATION IS REGULATED BY

MINERALOCORTICOID RECEPTOR ACTIVATION UNDER BASAL AND STRESS

CONDITIONS

Introduction

Glucocorticoid hormones (GCs) play a pivotal role in metabolic adaptation to stress, but chronic elevation of GCs due to stress or exogenous treatment causes deleterious effects (Charmandari et al., 2005). GC signalling is tissue-specific, mediated by either the glucocorticoid receptor (GR) or the mineralocorticoid receptor (MR) (de Kloet & Reul, 1987; de Kloet et al., 2008). The GCs affinity for MR is 10-fold greater than GR (de Kloet & Reul, 1987) and, therefore, under basal conditions, almost 80% of MRs are occupied by GCs. This has been a central tenant of GCs research in the brain, where MR is thought to exert a tonic influence on hippocampal function, while GR is activated by elevated GCs level during stress (de Kloet & Reul, 1987). Similar to this, MR and GR are also thought to play a role in GC-mediated lipid metabolism (Hoppmann et al., 2010), but overall there is a lack of consensus as GCs have been shown to either enhance or reduce obesity (John et al., 2016). While there has been considerable interest in MR from the perspective of adipose tissue differentiation (Caprio & Fe, 2007; Guo et al., 2008; Hoppmann et al., 2010), little is known about the molecular mechanisms involved in MR regulation of early life adipogenesis. As adipose tissue lacks 11β-hydroxysteroid dehydrogenase 2 (11βHSD2), GCs appear to be the main ligand for MR-mediated lipogenesis (Viengchareun et al., 2007). Antagonism of MR with drospirenone led to an inhibition of TG accumulation in 3T3-l1 and 3T3-F442A cells, independent of the GR (Caprio et al., 2011). Also, high levels of MR mRNA were detected in the adipose tissue of obese mice (ob/ob), and treatment with eplerenone, a MR antagonist, attenuated obesity-related insulin resistance (Hirata et al., 2009, 2012). A role for MR in lipid metabolism in zebrafish (Danio rerio) was also evident from the higher lipid accumulation observed in ubiquitous GR knockout zebrafish (Facchinello et al., 2017; Faught & Vijayan, 2019), which lacks aldosterone, suggesting a conserved role for GCs. However, results contradictory to the above were also reported. For instance, knockdown of GR, but not MR, in human adipocytes blocked the pro-adipogenic actions of cortisol, and mice overexpressing MR by its proximal promoter were resistant to high-fat diet induced obesity (Kuhn et al., 2014). Importantly, all these studies were carried out with in vitro or adult animals, and evidence that MR may have a functional role in postnatal lipid accumulation is currently lacking (Marzolla et al., 2012).

Postnatal fat accumulation is essential for survival, as disruption of lipogenic enzymes, including acyl-CoA: diacylglycerol acyltransferase (DGAT2), result in severe lipopenia and death soon after birth (Stone et al., 2004). For many years it has been thought that the primary mediator of GC-driven metabolism is due to the activation of GR, with only a handful of studies implicating a role for MR (Marzolla et al., 2012). Moreover, there is a lack of consensus as to the role of GR, with GCs being capable of both enhancing and reducing obesity (John et al., 2016). Indeed, a long-standing question is what mediates this GC-driven catabolic/anabolic balance. While, the primary role for GCs during stress is to mobilize energy substrates, including fatty acids, work over the last decade has also outlined a critical basal role for GCs in fetal programming and tissue maturation during prenatal and postnatal periods (Cole et al., 1995). Zebrafish is an excellent translational model for early life stress disorders and early development (Phillips & Westerfield, 2014), particularly from the perspective of glucocorticoid physiology. The zebrafish and human corticosteroid systems are homologous, with cortisol biosynthesis by hypothalamus-pituitary-interrenal (HPI) axis activation similar to that of the hypothalamus-pituitary-adrenal (HPA) activity in humans. The zebrafish GR is also homologous to the human GR (Alsop & Vijayan, 2009b), and the GR signalling pathways in teleosts appear to be conserved (Mommsen et al., 1999). The role of MR in lower vertebrates is unclear as teleosts lack aldosterone, and GCs appear to be the primary endogenous ligand for MR (Baker & Katsu, 2017). This makes zebrafish an ideal model to study GC signalling as cortisol is thought to be the primary ligand for both receptors. Furthermore, MR may be important during zebrafish development because unlike GR, this receptor transcript abundance increases temporally post- fertilization (Alsop & Vijayan, 2008). The necessity for using zebrafish is also underscored by the fact that ubiquitous knockouts of both GR and MR in mammals are lethal in utero or shortly after birth due to delayed lung maturation and dehydration, respectively (Cole et al., 1995; Berger et al., 1998). Therefore, only conditional knockouts can be generated in mice, which may lead to compensatory mechanisms in other tissues (Shimizu et al., 2015). To interrogate the physiological role of corticosteroid receptors (CRs) in shaping lipid metabolism during postnatal growth at the systems level, we generated zebrafish with ubiquitous knockouts of MR (MRKO) or GR (GRKO) (Faught & Vijayan, 2018). Discriminating the contribution of MR and GR signalling in postnatal lipid regulation is paramount to our understanding of metabolic disorders

associated with childhood stress, as well as exogenous GC treatment. Our results underscore a critical role for MR either singly or in combination with GR on postnatal lipid metabolism.

Materials and Methods

Zebrafish care and husbandry:

Adult zebrafish (Tupfel long fin (TL) strain) were maintained exactly as described previously (Faught & Vijayan, 2018a) under standard conditions (14h:10h, light: dark photoperiod cycle at 28.5°C). Adult fish were fed Gemma micro 500 (Skretting, Norway) in the morning (10:30 am), and live artemia (Grade A, Brine Shrimp Direct, USA) in the afternoon (4pm). All animal procedures were approved by the University of Calgary animal care committee (protocol# AC17- 0079) and met the Canadian Council for Animal Care Guidelines.

MR or GR Knockout zebrafish:

GRKO fish and MRKO fish were generated as described previously (Faught & Vijayan, 2018a). Briefly, GRKO larvae have a net -7 bp deletion in exon 2 of the nr3c1 gene which results in a premature stop codon and a truncated protein ~500 bp downstream of the start codon. MRKO larvae have a net +8 bp deletion in exon 2 of the nr3c2 gene which results in a premature stop codon and a truncated protein ~400 bp downstream of the start codon (Chapter 2).

Cortisol treatment:

Embryos/larvae were reared for days 0-5 dpf in a 28.5C incubator in Petri dishes with 1x embryo (E3) media (Faught & Vijayan, 2018a). At 3 dpf, larvae were transferred to six-well plates and exposed to either a vehicle (0.05% ethanol) or cortisol (5 µg/ml). 5dpf: A sample size of 4-6 larvae (each sample was a pool of 10 larvae) were collected at 5 dpf by euthanasia in MS222 (0.4g/L, Sigma, USA), rinsed in 1xPBS (137mM, NaCl; 2.7 mM KCl,

10 mM Na2HP04, and 1.8 mM KH2PO4), and frozen at -80°C for later analysis. 15 dpf: At 5 dpf, fish were transferred to beakers and fed twice daily with Gemma micro 150 (Skretting, Norway). 50% of the water was changed daily and dosed appropriately. A sample

size of 4-8 larvae (each sample was a pool of 5 larvae) were collected at 15 dpf by euthanasia in MS222 (0.4g/L), rinsed in 1xPBS and frozen at -80°C for later analysis.

Cortisol quantification:

Cortisol ELISA was performed as described previously (Faught et al., 2016b).

RNA-Seq:

3 larval samples (each larval sample was a pool of 10 larvae) per treatment (4 x 3 treatment = 12 samples) were used for transcriptome analysis. Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Germany), DNase treated according to the manufacturers’ directions. The quantity and integrity of RNA was assessed using the Agilent Bioanalyzer (2100). Sequencing was carried out at the Alberta Children’s Hospital Research Institute at the University of Calgary. RNA-Seq libraries were prepared using the Illumina TruSeq Stranded mRNA kit. Next- generation sequencing (NGS) was performed on an Illumina NextSeq. 500 platform (Illumina Inc., USA) using paired-end reads v2 chemistry with 75 bp sequencing. Raw reads were trimmed and filtered using Illumina base-calling software bcl2fastq (2.14) according to published methods (Sadoul et al., 2017). Alignment of reads was done using TopHat to the zebrafish genome (GRCz10/danRer10) and abundance of the aligned counts were quantified using kallisto (Kim et al., 2013). Differentially expressed genes (DEG) and gene set enrichment were generated using the fpkm values with Blast2GO (BioBam Bioinformatics, Spain). Heat maps were produced using ClutVis, following established methods (Metsalu & Vilo, 2015).

Gene Expression by quantitative real-time PCR (qPCR):

Transcript levels of specific genes were measured by qPCR and consisted of a sample size of 4-6 larvae per group. Total RNA was extracted, and cDNA synthesis was performed using published methods (Faught & Vijayan, 2018a). qPCR was preformed in duplicate using Sso Advanced SYBR green (BioRad, Canada) on the Applied Biosystems Quant Studio 3. Analysis by qPCR was conducted using primer sets listed in Table 1. Melting curves for each amplicon were

analyzed to ensure specificity. Ct values for target genes were normalized against β-actin, which did not differ between the groups, and expression was calculated by the ∆∆Ct method (Livak & Schmittgen, 2001).

Table 1: Gene specific primers: Gene Forward primer (5’-3’) Reverse primer (5’-3’) Source

11βhsd2 TGCTGCTGGCTGTACTTCAC TGCATCCAACTTCTTTGCTG (Faught et al., 2016b) dgat2 CCATACTTGCTGCATATTCC ATGTCATGATAAACTGCAGC (Her et al., 2011) elovl2 CACTGGACGAAGTTGGTGAA TTGAGGACACACCACCAGA (Monroig et al., 2009) fas GGAGCAGGCTGCCTCTGTGC TTGCGGCCTGTCCCACTCCT (Her et al., 2011) lepb GATGAGCACTTCCAGATGTC TGTCTATGTTGAGGCAGAGC (Mania et al., 2017) lpl ACAATTGACCCAACTGCTGA GGTTCTTCGAGGGTCCGAAA (Gao et al., 2016) ppary TGCCGCATACACAAGAAGAG ATGTGGTTCACGTCACTGGA (Imrie & Sadler, 2011)

βactin TGTCCCTGTATGCCTCTGGT AAGTCCAGACGGAGGATG (Faught & Vijayan,

2018a)

Metabolites:

Triglyceride and glycerol (TG; Serum triglyceride kit, TR0100, Sigma), cholesterol (Amplex Red Cholesterol kit, A12216, Thermo Fisher Scientific) and protein concentrations (Pierce BCA kit 23223/23224, VWR) were measured using commercial kits.

Enzyme Activity:

3-hydroxyacyl-CoA dehydrogenase (HOAD; EC 1.1.1.35) was measured at room temperature (23°C) in a final volume of 250 µl by continuous spectrophotometry at 340 nm (VersaMax, Molecular Devices) as described previously (Vijayan et al., 1996). Briefly, samples were diluted in a 0.1mM NADH in imidazole buffer (50 mM, pH 7.4), and the reaction was started with 0.1mM acetoacetyl-CoA (Sigma, USA). 83

Statistical analyses:

Data are shown as mean ± SEM. Statistical comparisons were carried out using Sigma Plot 13 (Systat Software, USA). All data were analysed using a two-way ANOVA (Holm-Sidak post- hoc). Data on gene transcript and cortisol levels were transformed to meet the assumptions of normality and equal variance. Untransformed data are shown in all figures. Significance level (a) of p<0.05 was used in all cases.

Results

Corticosteroid receptor activation (Fig. 1):

To examine GC activation of specific CRs during postnatal growth, we monitored MRKO and GRKO larvae in combination with cortisol treatment during the rapid growth phase immediately post-exogenous feeding (5-15 dpf). In this model, the 5 dpf larvae represent a closed system (feeding on yolk), and we used the metabolite levels to represent endogenous energy stores (Fraher et al., 2016). Any increase above these levels at 15 dpf was due to the growth associated with exogenous feeding, while a decrease represents a lack of synthesis and/or enhanced utilization of energy substrates (Fig. 1a). In this way, we were able to assess the role of MR and GR under basal conditions, as well as in response to GR activation specifically by exogenous cortisol treatment (Fig 1a). To confirm GR activation at 15 dpf, we treated the fish with exogenous cortisol from 3 dpf onwards (Fig. 1a). The different corticosteroid activation groups were: i) No receptor activation (None) - MRKO larvae, that had basal cortisol level and therefore no GR activation (Fig 1c), ii) MR activation - WT larvae with basal cortisol levels, iii) MR activation (MR*) - GRKO larvae that also had high cortisol level, iv) MR + GR activation - WT larvae treated with exogenous cortisol, and v) GR activation - MRKO larvae treated with exogenous cortisol. Larvae with no receptor activation had normal cortisol levels (184.1±43.2 pg/larva), similar to WT larvae in which only MR was active (123±12 pg/larva; p=0.558) (Fig 1b). GRKO larvae (MR*) had 6-fold higher (727±200 pg/larva; p=0.001) cortisol levels (Fig. 1b). The

cortisol treated groups (GR activated larvae) had significantly higher cortisol levels both in the WT larvae (2236±89.9 pg/larva; p<0.001) and MRKO larvae (5031.8±1173.9 pg/larva p<0.001), respectively, compared to the vehicle treated wild type (Fig. 1b). We then quantified the transcript abundance of a GR-responsive gene 11βhsd2 (Alderman & Vijayan, 2012), to confirm GR-specific transcriptional activation. Cortisol treatment increased mRNA levels of 11βhsd2 ~2- fold in both the WT and MRKO larvae (p<0.001) compared to larvae with no CRs activation (MRKO) or only MR activation (WT and GRKO) (Fig. 1c). To confirm MR activation, we used leptin b (lepb) transcript abundance, as this gene mRNA levels were elevated only in the presence of MR (Fig. 1d). The lepb transcript abundance was not detectable in both the MRKO groups (either without (None) or with (GR activation) cortisol). There was a significantly higher transcript abundance of lepb in the MR+ GR activation group compared to all the other groups (Fig. 1d).

Figure 1: Corticosteroid receptor activation: a) Schematic diagram of methods used to modulate GR and/or MR activation. Fish lacking GR and MR were monitored over a 15 dpf developmental period. 5 dpf fish represent a closed system as the larvae feeds on yolk, while 15 dpf represents larvae 10 days post-feeding. b) Cortisol levels in 15 dpf larvae with no corticosteroid receptor activation (MRKO larvae), MR activation (WT larvae), MR activation in the presence of high cortisol (GRKO), GR and MR activation (WT treated with cortisol) and only GR activation (MRKO treated with cortisol). c) Confirmation of GR activation by measuring the transcript abundance of 11β hydroxysteroid dehydrogenase 2 (11βhsd2), a GR-specific gene. d) Confirmation of MR activation by measuring the transcript abundance of leptin b (lepb), which was expressed only in the presence of MR. Different letters denote significant differences; ND denotes no detection; one-way ANOVA (n = 4-8); Holm-Sidak post hoc text; p<0.05).

MR activation increases lipid accumulation under basal conditions (Fig. 2):

In zebrafish, visceral adipocytes first appear once exogenous feeding occurs (5-8dpf) (Minchen & Rawls, 2015), and this was confirmed by the elevated TG levels post-feeding in larvae with an active MR. In larvae with no corticosteroid receptor activation (MRKO larvae) TG levels were similar at 15 dpf (6.5±0.7 µg/larva) compared to 5 dpf (4.7±0.4 µg/larva, p=0.383) (Fig 2a). However, when MR was activated, as in the case of WT and GRKO larvae, TG levels rose to 15.5± 1.7 µg/larva (p<0.001) and 11.8±2.3 ug/larva (p<0.001), respectively. Cortisol treatment, which activates GR, caused TG levels to decrease 7-fold in the WT larvae from 5 dpf (6.7±0.8 µg/larva) to 15 dpf (1.0±0.1 ug/larva; p<0.001) and 2-fold in the MRKO larvae from 5 dpf (7.0 ±0.7 µg/larva) to 15 dpf (3.4±0.8 ug/larva; p<0.002) (Fig. 2a). Glycerol levels were below detection in the 5 dpf larvae, but increased with feeding by 15 dpf (Fig. 2b). MR activation significantly elevated glycerol levels [WT larvae (0.22 ±0.02 µg/larva) and GRKO larvae resulted in similar glycerol levels (0.4 ±0.07 µg/larva); p<0.683] (Fig. 2b) compared to larvae with no CR activation and the cortisol treated larvae (WT and MRKO) (Fig. 2b; p<0.001). There was no difference in cholesterol levels between the groups at 5 dpf (Fig. 2c). Cholesterol levels were higher in all groups at 15 dpf compared to the 5dpf larvae, except the GR+MR activation group (Fig. 2c). At 15 dpf, MR activation increased cholesterol levels (MR: 1189.7 ± 14.2 ng/larva, MR*: 1368.9 ± 145.4 ng/larva) compared to groups with no MR activation (None: 697.2 ± 31.1 ng/larva and GR: 733.7± 101.3 ng/larva, p<0.001). Also, the MR+GR activation group (WT + cortisol) had significantly lower cholesterol level at 15 dpf compared to the other groups, and this group also showed no increase from the 5dpf larvae (p=0.247; Fig. 2c).

Figure 2: MR is essential for lipid accumulation under basal conditions. The effect of MR and GR on lipid synthesis were were monitored over a 15 dpf developmental period. a) Triglyceride levels of 15 dpf larvae in larvae with no corticosteroid receptor activation ‘none’ (MRKO larvae), MR activation (WT larvae), MR* activation (GRKO larvae [hypercortisolemic]), GR and MR activation (WT + 5 µg/ml cortisol) and only GR activation (MRKO + 5 µg/ml cortisol). b) glycerol levels 15 dpf zebrafish larvae (5dpf glycerol was non-detectable), c) Cholesterol levels at 5 and 15 dpf. Two-way ANOVA (Holm-Sidak post hoc text; p<0.05). Different letters denote significant differences (5 dpf, lower case; 15 dpf upper case). Asterisk identify a significant difference between days.

MR has little impact on genes involved in TG and cholesterol biosynthesis (Fig. 3):

In our study, fatty acid synthase (fas), a key gene encoding lipogenic enzyme and a marker of lipogenesis (Marzolla et al., 2012), had higher mRNA transcript abundance in all groups at 15 dpf compared to 5 dpf, with the exception of WT larvae, in which only MR was activated (Fig. 3a, p=0.191). There was no effect of MR activation on fas mRNA abundance at 15 dpf (Fig. 3a), but GR activation increased fas mRNA levels above WT levels (Fig. 3a). Peroxisome proliferator-activated receptor gamma (ppary), a key transcription factor which promotes adipogenesis (Imrie & Sadler, 2011) was upregulated by MR activation at 15 dpf compared to 5 dpf (Fig. 3b). Within 15 dpf, there was no significant difference in ppary between the different groups; however, GR activation in the absence of MR lowers ppary transcript abundance compared to the other groups (Fig 3b). At 15 dpf, there was no effect of CRs activation on the transcript abundance of dgat2 (Fig. 3c), another key lipogenic gene essential for early developmental lipid synthesis and survival in mice (Stone et al., 2004). At 5 dpf, dgat2 transcript abundance was significantly lower with MR activation, but only in the presence of high cortisol (GRKO group), compared to all the other groups (Fig. 3c). The transcript abundance of elovl2 was significantly higher in all groups at 15 dpf compared to 5 dpf, except for the GR activation group that showed no difference between the two time periods (Fig. 3d). At 5 dpf, elovl2 transcript abundance was significantly higher in the MR activation group but only in the presence of higher cortisol levels (GRKO and WT + cortisol) (Fig. 3d). We also examined the transcript abundance of hmgcs1, the first step in cholesterol biosynthesis (Fig. 3e). The transcript abundance of hmgcs1 was significantly higher at 15 dpf compared to 5 dpf in all groups (p<0.001), except in response to GR activation (cortisol treatment) (Fig. 3e).

Figure 3: MR has little effect on genes involved in lipid biosynthesis. The effect of MR and GR on lipid synthesis were monitored over a 15 dpf developmental period by examining larvae with no corticosteroid receptor activation ‘none’ (MRKO larvae), MR activation (WT larvae), MR* activation (GRKO larvae [hypercortisolemic]), GR and MR activation (WT + 5 µg/ml cortisol) and only GR activation (MRKO + 5 µg/ml cortisol). Key enzymes/transcription factors involved in TG biosynthesis were measured via qPCR including: a) fatty acid synthase (fas), b) peroxisome proliferator-activated receptor y (ppary), c) diacylglycerol O-acyltransferase 2 (dgat2), d) elongation of very long chain fatty acid protein (elovl2), and e) 3- hydroxy-3-methylglutaryl-CoA synthase 1 (hmgcs1), measured for cholesterol biosynthesis. Two-way ANOVA (Holm-Sidak post hoc text; p<0.05). Different letters denote significant differences (5 dpf, lower case; 15 dpf upper case). Asterisk identify a significant difference between days.

MR activation restricts lipoprotein lipase (lpl) transcript abundance and increases β-oxidation capacity (Fig. 4):

Transcript abundance of lpl was inversely correlated with TG levels (Fig. 4a). MR activation significantly reduced lpl transcript abundance at 5 dpf compared to the lack of CRs activation (None group), and this was recovered in the presence of GR activation (Fig. 4a). Also, at 15 dpf, MR activation significantly decreased lpl mRNA levels below that of the 5 dpf levels (p<0.001) and this was brought back to the 5 dpf level in the presence of MR + GR activation. GR activation alone in the absence of MR resulted in significantly higher lpl transcript abundance at 15 dpf compared to all other groups (Fig. 4a). To further assess the lipolytic capacity of zebrafish larvae, we measured HOAD enzyme activity, an essential enzyme in the mitochondrial β-oxidation pathway in both fish (Mommsen et al., 1999) and mammals (Eaton et al., 1996). There was no difference in HOAD activity in any group in the 5dpf zebrafish larvae. MR activation (WT larvae) doubled HOAD activity from 5 dpf to 15 dpf (p<0.001), with smaller but still significant increases in HOAD activity in larvae with no receptor activation (none) or MR activation under high cortisol levels (GRKO; Fig. 4b; p<0.001). At 15 dpf, GR activation significantly lowered HOAD activity compared to MR only activated larvae (Fig. 4b; p<0.001).

Figure 4: MR modulates GR signalling of by increasing lipolysis stressed conditions. The effects of MR and GR activation on lipid metabolism were monitored over a 15 dpf developmental period by examining larvae with no corticosteroid receptor activation ‘none’ (MRKO larvae), MR activation (WT larvae), MR* activation (GRKO larvae [hypercortisolemic]), GR and MR activation (WT + 5 µg/ml cortisol) and only GR activation (MRKO + 5 µg/ml cortisol). a) lipoprotein lipase (lpl) mRNA abundance was measured as a proxy for TG metabolism. b) HOAD enzyme activity was measured as a proxy for β-oxidation. Two-way ANOVA (Holm-Sidak post hoc text; p<0.05). Different letters denote significant differences (5 dpf, lower case; 15 dpf upper case). Asterisk identify a significant difference between days.

Larval transcriptome changes specific to MR and/or GR activation (Fig. 5; SI Fig. 1 and SI

Tables 1-7):

To get a complete picture of lipid-related differentially expressed genes (DEG) that are regulated by GR and MR, we used next-generation sequencing (RNA-Seq). Treatment of larvae post-hatch (3 dpf) with cortisol caused the upregulation of 76 genes, and the downregulation of 13 in comparison with untreated controls. Of the 89 genes regulated by cortisol (F, Fig. 5a and 5b) in WT zebrafish (MR+GR activation), 60 of these were also regulated by cortisol in MRKO (GR activation). There were 29 genes that failed to be regulated in the MRKO in the presence of cortisol. Transcriptomics analysis of zebrafish larvae revealed that there are several genes involved in lipid metabolism under the transcriptional control of MR (Fig. 5c, d). Key genes involved in TG and cholesterol biosynthesis that were not upregulated in MRKO+F (GR activation), but upregulated in WT+F (GR+MR activation) include, 3-hydroxy-3-methylglutaryl- CoA synthase 1 (hmgcs1) UP (p=4.50E-08; fdr=6.87E-06), pyruvate dehydrogenase kinase, isozyme 4 (pdk4) (UP (p=1.38E-05; fdr=0.001149593), diacylglycerol O-acyltransferase 2 (dgat2) (UP (p=5.14E-06; fdr=4.86E-04), sterol-C5-desaturase (sc5d), UP (p=1.71E-05; fdr=0.001379792), ELOVL fatty acid elongase 2 (elovl2), UP (p=8.31E-06; fdr=7.23E-04) (Fig. 5c and 5d). Genes that are likely regulated by GR activation included carboxyl ester lipase, tandem duplicate 2 (cel.2) UP (p=2.48E-07; fdr=2.92E-05), fatty acid desaturase 2 (fads2) (UP (p=2.43E-15; fdr=1.16E-12), 3-hydroxy-3-methylglutaryl-CoA reductase a (hmgcra) UP (p=2.07E-10; fdr= 4.37E-08) and methylsterol monooxygenase 1 (msmo1) UP (p=9.88E-15; fdr=4.05E-12) (Fig. 5c and 5d).

Figure 5: Larval transcriptome in response to GR and/or MR activation. a) DEG in 5 dpf in zebrafish larvae treated with cortisol (F, 5 µg/ml), or lacking GR and MR and b) Venn diagrams show commonly regulated genes between the four groups (WT+ Cortisol, GRKO, MRKO, MRKO+ Cortisol) c) Glucocorticoid regulation of metabolic pathways involved in lipogenesis and lipolysis. Cortisol regulated genes under the transcriptional control of MR are denoted in green, GR-regulated genes in red, and possible targets of heterodimerization in purple. d) Heat map depicting changes in genes involved in TG and cholesterol biosynthesis, lipid metabolism and transport in MRKO and WT 5 dpf larvae, with and without cortisol, and GRKO larvae. Boxes denote corticosteroid receptor is activated. Abbreviations: 3-hydroxy-3-methylglutaryl-CoA synthase 1 (hmgcs1), pyruvate dehydrogenase kinase, isozyme 4 (pdk4), apolipoprotein A-IV a (apoa4a), apolipoprotein Bb, tandem duplicate 1 (apobb.1), diacylglycerol O-acyltransferase 2 (dgat2), sterol-C5-desaturase (sc5d), ELOVL fatty acid elongase 2 (elovl2), elastase 3 like (ela3l), fatty acid desaturase 2 (fads2), apolipoprotein Eb (apoeb), carboxypeptidase A4 (cpa4), elastase 2 like (ela2l), carboxypeptidase B1 (cpb1), somatostatin 2 (sst2), methylsterol monooxygenase 1 (msmo1), 3-hydroxy-3-methylglutaryl-CoA reductase a (hmgcra), carboxyl ester lipase, tandem duplicate 1 (cel.1), serine protease 59, tandem duplicate 1 (prss59.1), six-cysteine containing astacin protease 3 (c6ast3), apolipoprotein Ea (apoea), carboxypeptidase A5 (cpa5), serine protease 1 (prss1), F-box protein 32 (fbxo32), uncoupling protein 1 (ucp1), pyruvate dehydrogenase kinase, isozyme 2b (pdk2b).

Discussion

This study investigated the physiological significance of GR and MR activation on lipid metabolism during the postnatal growth in zebrafish. We demonstrate that MR is essential for the TG accumulation post-feeding in zebrafish larvae under basal conditions. Also, under elevated cortisol levels, mimicking a stress response, MR is essential for GR-mediated lipolysis and the associated transcriptional responses. From a functional standpoint, a lack of lipid accumulation during postnatal growth disrupts developmental programming and reduces survival (McLaren et al., 2018). Therefore, given the distinct and complementary roles for these receptors in lipid regulation during early development, our results suggest that GCs exposure prenatally or postnatally may lead to lipid dysfunction in children (Kelly et al., 2012). In adults, there is evidence that MR and GR modulate lipogenesis, but the molecular mechanisms leading to this phenotype are unclear (Hoppmann et al., 2010). Although aldosterone is the main MR ligand mediating water and salt homeostasis, this receptor activation also promotes inflammation, cardiovascular remodeling and endothelial dysfunction (Marzolla et al., 2012). In addition, MR signalling is evident in target tissue low in 11βHSD2, including adipocytes, suggests a key role for GCs in mediating MR effect in non-mineralocorticoid target tissues (Marzolla et al., 2012). Zebrafish are an excellent model to study the effect of GCs on MR action because teleosts lack the endogenous mineralocorticoid aldosterone (Baker & Katsu, 2017). Consequently, cortisol seems to be the primary ligand for GR and MR activation in teleosts, and more importantly the mineralocorticoid role appears to be GR-mediated (Cruz et al., 2013). This suggests that MR may have other extra-mineralocorticoid role in teleosts, especially given that phylogenetically MR preceded GR in vertebrate evolution (Baker et al., 2013). To activate GR, we treated fish with cortisol and confirmed transcriptional activation by measuring mRNA levels of a GR-responsive gene, 11βhsd2 (Alderman & Vijayan, 2012) (Fig. 1c). We also identified lepb as a marker of MR activation, as transcript abundance was only detectable in the presence of MR and absent in the MRKO groups (Fig. 1d). The 11βhsd2 and lepb transcript changes provided a readout for GR and MR activation, respectively, by GCs in the present study. Previous reports using knockout fish models indicate that MR plays an important role in regulating HPI axis function (Faught & Vijayan, 2018a), and stress-related behaviour (Faught & Vijayan, 2018a)(Sakamoto et al., 2016). In combination with mammalian work (Joëls & de

Kloet, 2017), this suggested that the primary role for this receptor may have been in stress sensing and appraisal prior to its mineralocorticoid role in the tetrapod lineage (Baker et al., 2013; Baker & Katsu, 2017). Here we demonstrate that in addition to stress sensing and behaviour, MR may also be involved in energy substrate partitioning during early-life adipogenesis. Lipid is a key energy substrate during stress adaption (Charmandari et al., 2005), but very little is known about adipogenesis under basal conditions when MR signalling would take precedence over GR (de Kloet et al., 2012). Indeed, given that lipid deposition during early life stage is essential for survival (Stone et al., 2004), it is necessary to understand the regulatory mechanisms surrounding GCs regulation of lipid stores by assessing both GR and MR signalling. In mammals, a regulatory role for MR has been described in both adipocyte tissue differentiation and adipose tissue inflammation (adipokine production); however, a role for MR-mediated adipogenesis during early development has not yet been described (Marzolla et al., 2012). While mice overexpressing MR were resistant to a high fat diet in vivo (Kuhn et al., 2014), there is clear evidence that MR has a primarily anabolic effect and promotes adipogenesis in mammals (reviewed by (Marzolla et al., 2012, 2014; Armani et al., 2014). This was also clearly evident in our study where MR activation led to the accumulation of TG, glycerol and cholesterol levels in zebrafish larvae postnatally (Fig. 2). Interestingly, this effect of MR was lost in the presence of GR activation, as the TG levels dropped below the 5 dpf levels, suggesting an increased capacity of lipid catabolism. This was also seen with GR activation alone at 15 dpf, but it was significantly less than in the presence of MR suggesting an interaction between the two receptors in promoting lipolysis. However, from the TG levels alone it was unclear whether MR activation was promoting lipogenesis (John et al., 2016) and/or reducing lipolysis. To address this, we first assessed the anabolic role of MR by measuring key genes involved in lipid biosynthesis, including fas, ppary, dgat2,elovl2, and hmgcs1 (Stone et al., 2004; Marzolla et al., 2012). The transcript abundance of fas was higher in all groups at 15 dpf compared to 5 dpf, with the exception of WT, and increased with GR activation in 15dpf larvae (Fig. 3a). This indicates that the lipogenic capacity is maintained over the postnatal growth period under basal and stress conditions. This was further confirmed by the similarity in the dgat2 and elovl2 mRNA profile between the genotypes (Fig. 3c and 3d). Fas, dgat2 and elovl2 genes are involved in adipogenesis and essential to lipid accumulation during early development

in mammals, but their regulation by MR is unknown (Fraher et al., 2016; Pauter et al., 2016; McLaren et al., 2018). Another key player in adipogenesis is the transcription factor ppary, and upregulation of this gene is associated with lipid accumulation in mammals (Kintscher, 2005). There is some evidence that MR activation modulates ppary transcript abundance, as MR blockade in the 3T3-L1 cell line reduced ppary transcript levels (Armani et al., 2014). This was also the case in our study, as we show that MR activation increased ppary transcript abundance at 15 dpf compared to 5 dpf in zebrafish larvae (Fig. 3b). The failure of the larvae lacking MR to upregulate ppary supports a conserved role for MR in ppary regulation (Armani et al., 2014), and may be a key mechanism by which MR signalling modulate lipid accumulation. Interestingly, the mammalian ppary variants lack MRE in the promoter regions, but have GREs, suggesting that the transcriptional control of this gene may involve the interaction of MR with GREs. Taken together, MR activation is essential for maintaining the lipid profile of developing zebrafish, and we propose that the promotion and maintenance of lipogenesis includes the upregulation of ppary. While the loss of MR indicated a direct role of MR activation in lipid accumulation under basal/low cortisol level conditions, the lack of regulation of lipogenic genes suggests that MR may also be anti-lipolytic. Indeed, under stress conditions, when GR is activated, there was no TG synthesis, and, in fact, the TG levels dropped below the endogenous (5 dpf) TG stores. This drop required both activation of GR and MR, as larvae in which only GR is active have similar levels to larvae with no CR activation (Fig. 2a). However, this is not the case with glycerol or cholesterol levels (Fig. 2b,c). When MR and GR are activated (WT+cortisol) in zebrafish larvae these metabolites are maintained at the endogenous 5 dpf cortisol levels, indicating the effect of MR on TG metabolism may be pathway specific. Therefore, we next tested the hypothesis that MR maintenance of TG levels during early development may involve suppression of lipolytic capacity. Classically GR has been thought to be the primary mediator of lipolysis in vertebrates, and in humans, physiological concentrations of cortisol were a potent stimulator of lipolysis in vivo (Djurhuus et al., 2002). Indeed, larvae with an active MR had the highest TG content at 15 dpf, but the lowest transcript abundance of lpl (Fig. 4a). Low lipolytic capacity in these groups may tilt the balance towards TG accumulation without having to regulate the lipogenic processes. This suggests that MR, may be a negative regulator of lipolysis under basal conditions. Adipocyte-specific GR is required for

lipolysis in vivo (Shen et al., 2017a), but to date, there has been no role for MR in the regulation of lipolysis. Our results indicate that lpl is a key target for MR signalling, and suppression of this gene’s transcription may lead to lipid accumulation (Fig.2a). This provides a new perspective on the lack of lipid synthesis in conditional MRKO mice, as the accumulation of lipids reported may also be due to a decrease in breakdown (Hoppmann et al., 2010). Whether MR is essential for lipid accumulation during early development in higher vertebrates remains to be determined. Larvae with no CR activation at 5 and 15 dpf have high lpl levels (Fig. 4a), which explain the low levels of TG in these groups (Fig.2a), and further supports the contention that MR negatively regulates lpl expression. Further evidence is provided when GR is activated in the absence of MR, which leads to a 7-fold increase in lpl mRNA levels. However, this is not the case when both MR and GR are activated (WT + cortisol group) (Fig. 4a), suggesting that MR has a negative effect on lpl transcription both during basal and stress conditions. Interestingly this gene only has a GRE and not an MRE in its promoter region, suggesting that, in addition to the classical homodimerization of these receptors, GR/MR heterodimerization may also play a role in transcriptional regulation (Mifsud & Reul, 2016). Indeed, stress in rats increased the binding of both GR and MR to GREs on target genes, and it is thought that the binding of MR may be restricted at basal levels, and facilitated by GR at high GC levels (Mifsud & Reul, 2016). Neither gene transcription nor HOAD activity completely explained the results of the TG profiles, and, the question remained how fish lacking MR were able to maintain TG levels when treated with cortisol, compared to larvae with no CR activation. Transcriptomics was used to assess novel targets of GR and MR regulation, which revealed additional targets of GR in lipolysis, including carboxyl ester lipase (cel), which increased in response to GR activation (Fig. 5c and 5d). While the regulation of genes by MR and GR was often distinct, possible targets of both GR and MR were also identified by identifying DEG in both the MRKO and GRKO larvae. These included F-box protein 32 (fbxo32), uncoupling protein 1 (ucp1) and pyruvate dehydrogenase kinase, isozyme 2b (pdh2b), all of which may play a role in lipid homeostasis (Fig. 5c and 5d), but warrants further study. While heterodimers may be one mode of GR-MR interaction, we also propose that tandem control of key metabolic pathways distinctly by GR and MR may be another level of regulation, and this is evident from the pathway-specific regulation seen from the transcriptome analysis (Fig. 5c and 5d). For instance, in the case of cholesterol

biosynthesis, cortisol upregulates several key genes, two of which are under the transcriptional control of GR (hmgcra and msmo1), while two others are under the transcriptional control of MR (hmgcs1 and sc5d) (Fig. 5c and 5d). This further reiterates that not only are both receptors required but altering the abundance of one can directly impact the other and may lead to abnormal TG and cholesterol profiles during postnatal development. This should particularly be taken into account in the case of exogenous GCs therapy, which may alter the signalling capacity of corticosteroid receptors, and may affect the metabolic programming (Nesan & Vijayan, 2012). To date, there is little information regarding the physiological role of MR in lower vertebrates or the molecular targets of MR. There is some evidence that MR complements the role of GR in the larval zebrafish stress response and mediates stress-related behaviour (Faught & Vijayan, 2018a). There is no evidence that MR plays any role in adipogenesis in lower vertebrates; however, in GRKO fish there was an increase in both total lipid and subcutaneous fat, despite being hypercortisolemic (Facchinello et al., 2017). Here we identify that not only does MR play a role in TG biosynthesis during zebrafish development, but that it may also modulate the transcriptional function of GR. The role of both splice variants of GR and 11βHSD2 can also not be discounted in this system, as both have a role in modulating corticosteroid receptor signalling. In mammals the role of 11βHSD2 is to reduce the amount of cortisol, to the biologically inactive cortisone, allowing for aldosterone signalling to occur. In fish, the function of 11βHSD2 appears only to maintain resting cortisol levels (Faught & Vijayan, 2018a). Whether 11βHSD2 acts in a tissue-specific manner to promote MR signalling by reducing cortisol levels remains to be ascertained. Overall, this work highlights the potential molecular mechanisms involved in GC-driven control of lipolysis and lipogenesis during critical periods of early development and energy substrate repartitioning in zebrafish. This work specifically underscores the importance of MR as a key lipogenic signal during this time, yet highlights how much is unknown about the interaction of GR and MR. As such, future studies may provide critical information about MR manipulation for diseases classically associated with GR signalling, including developmental stress-related disorders, diabetes and childhood obesity. Beyond childhood, it is also necessary to examine the impact that these early perturbations in TG and cholesterol biosynthesis may have on adults. It is already known that a lack of GR in

zebrafish promotes lipid accumulation (Facchinello et al., 2017), but whether MR plays a role in this phenotype remains to be determined.

Acknowledgements: We would like to thank Paul Gordon at the University of Calgary Children’s Hospital for assistance with the transcriptomics analysis and Dr. Alison Holloway for commenting on the manuscript.

Translational Perspective: The physiological underpinning of this study is that MR may have a more prominent role than GR in early developmental growth, as well as playing a key mediator of GR effect on stress coping mechanisms. We hypothesize that the balance of MR and GR in target tissues may be a key factor dictating the lipogenic or lipolytic properties of GCs, and this knowledge may have biomedical implications. To date, there has been little consideration of the role of GC-MR signalling to explain the increased risk of obesity and dyslipidemia following early life exposure to glucocorticoids. Here we identify that not only does MR play a role in TG biosynthesis during the period of normal adipose tissue development, but that it also modulates the transcriptional function of GR in the face of excess glucocorticoid exposure. Overall, this work highlights the importance of MR as a key lipogenic signal during postnatal development and underscores the need to understand the interactions of GR and MR in lipid regulation. Discriminating the contribution of MR and GR signalling is paramount to understanding the impact of glucocorticoid exposure during early life on the etiology of metabolic diseases.

100

CHAPTER 5: CORTISOL-MEDIATED POSTNATAL GROWTH SUPPRESSION IN

ZEBRAFISH LARVAE REQUIRES MINERALOCORTICOID RECEPTOR

ACTIVATION

101

Introduction

Both intrauterine and early postnatal environments program changes in the structure and function of various physiological systems in vertebrates. The two major factors affecting fetal growth are maternal nutrition and fetal stress. Exposure in utero to excess glucocorticoids (GCs), reduces skeletal muscle mass in rats, and this is independent of maternal nutrition (Gokulakrishnan et al., 2012). The increase in skeletal muscle mass depends on the balance between anabolic and catabolic processes, which are driven to a large extent by complex endocrine signals. Maintenance of skeletal muscle mass is an important component of postnatal growth because it comprises approximately 40% of adult body weight in mammals (Shimizu et al., 2011) and >50% in fish (Mommsen, 2001; Sadoul & Vijayan, 2016). Muscle growth is driven primarily by anabolic hormones, including the growth hormone (GH) /insulin-like growth factor 1 (IGF1) axis, and the associated activation of P13K/AKT/mTOR pathway (Shimizu et al., 2011). Muscle loss is associated with enhanced proteolysis, which involves the activation of the ubiquitin pathway, including transcriptional activation of E3 ligase-encoding genes, atrogin-1 and murf1, and the autophagic/lysosomal pathway (Kuo et al., 2013). GCs released in response to stress also stimulates proteolysis, and a key target for these steroids is the regulated in development and DNA damage response 1 (REDD1), which reduces skeletal muscle mass by inhibiting mTORC1/S6K pathway (Shimizu et al., 2011; Lipina & Hundal, 2016). At a systems level, postnatal stress, and the attendant rise in GCs cause dysregulation of normal growth trajectories (Charmandari et al., 2005; Kelly et al., 2012). This negative effect of GCs on growth is mediated by the glucocorticoid receptor (GR) activation of proteolytic pathways in animals (Shimizu et al., 2011; Lipina & Hundal, 2016; Britto et al., 2018). Indeed, both murf1 and redd1 are under the transcriptional control of GR in mammals (Kuo et al., 2013), and also in zebrafish (Danio rerio) (Faught & Vijayan, 2019). In addition, ubiquitous knockout of GR in zebrafish led to greater body mass, including higher protein and fat, supporting a role for GCs in regulating skeletal muscle mass (Facchinello et al., 2017; Faught & Vijayan, 2019). In teleosts, cortisol is the primary GC and the mode of action involves the activation of either GR and/or mineralocorticoid receptor (MR). Both these corticosteroid receptors (CRs) are ligand- bound transcription factors, which regulate gene expression by binding to glucocorticoid response elements (GREs) in the promoter regions of target genes (Kuo et al., 2013). The

102

primary role of cortisol during stress is to increase glucose availability by promoting muscle proteolysis to channel the C3 substrates for gluconeogenesis in the liver (Mommsen et al., 1999; Patel et al., 2014). Consequently, chronic cortisol stimulation leads to glucose intolerance and increased muscle proteolysis, thereby reducing skeletal muscle mass. The role of MR in regulating skeletal muscle mass and protein content is less well defined. In mammals, MRs are present in the skeletal muscle, and responsive to aldosterone (Chadwick et al., 2015), but a physiological role for MR in lower vertebrates has yet to be characterized. MR binds to glucocorticoid response elements (GREs) in the promoter regions of specific genes to mediate GR function (Mifsud & Reul, 2016). Recently, we demonstrated that MR activation either alone or in combination with GR regulates lipid accumulation in zebrafish larvae, suggesting a conserved role of CRs interactions (Chapter 4). In addition to regulation of proteolysis, GCs also exert effects on growth by modulating the functioning of the growth axis in fish (Mommsen, 2001; Madison et al., 2015; Sadoul & Vijayan, 2016) and mammals (Yuen et al., 2013). Much of the mechanistic work has been carried out in mammals, and GCs have been shown to either inhibit or stimulate growth hormone (GH) release (Shapiro et al., 1978; Sartin et al., 1994). The use of GCs-based therapeutics during childhood has also been associated with long-term dysregulation of growth, suggesting effects on developmental programming (Kelly et al., 2012).The lack of consensus in GCs-mediated GH regulation has also been reported in teleosts. For instance, chronic stress has been associated with an increase in GH and insulin-like growth factor 1 (IGF1) levels in Atlantic salmon parr (Salmo salar) (McCormick et al., 1998) and adult rainbow trout (Oncorhynchus mykiss) (Madison et al., 2015), whereas acute stress either results in no change or a decrease in circulating GH levels (Pickering, 1993; Shepherd et al., 2011). Stress and the attendant rise in GCs can also indirectly affect growth by reducing the food consumption and GH signalling in fish (Bernier, 2006; Ortega et al., 2013). Altogether, GCs effect on muscle growth may involve differential regulation of anabolic and catabolic processes by CRs, but the mechanisms are elusive. Here, using zebrafish larvae as a model of rapid growth, we tested the hypothesis that GCs effect on postnatal growth suppression is due to a dysregulation of the balance between the anabolic and catabolic processes, and mediated by both MR and GR activation. Zebrafish are often used as translational model for metabolic research (Seth et al., 2013), and they are the only

103

teleost lacking a GR paralog (Alsop & Vijayan, 2008, 2009a). Recent work has also highlighted the importance of cortisol during development (Nesan & Vijayan, 2016), and that early life stage cortisol exposure in zebrafish can lead to long term metabolic, cardiac and behavioural dysfunctions in adults (Wilson et al., 2016). Our results indicate that cortisol stimulation reduces postnatal growth, and this is due to the stimulation of proteolytic pathways due to GR activation. However, the MR/GR interaction upregulates anabolic hormones, including GH. IGF1 and insulin, and they may be playing a compensatory role in limiting protein breakdown in response to early-life stress and/or GCs treatments.

Materials and Methods:

Zebrafish maintenance: Adult zebrafish (Tupfel long-fin (TL) strain), nr3c1 knockouts (GRKO) and nr3c2 knockouts (MRKO) (Faught & Vijayan, 2018b) were maintained in 10L tanks on a recirculating system with a 14:10 light:dark cycle (Tecniplast, Italy) according to established animal care protocols approved by the University of Calgary Animal Care committee (#AC17-0079) in accordance with the Canadian Council for Animal Care. Water was maintained at 28.5°C, ph7.6, and 750 µS conductivity. Animals were fed twice daily with Gemma micro 300 diet (Skretting, USA) in the morning and live Artemia (San Francisco Bay Brand, USA) in the afternoon.

Cortisol Treatment and Larval Rearing:

Fish larvae were reared for days 0-5 days post fertilization (dpf) in a 28.5°C incubator in 20x100mm dishes with 1x embryo media. Day 3 larvae were transferred to a 6-well plate, each well contained 20 larvae and 4 ml of embryo media. Larvae were exposed to either a vehicle (0.05% ethanol) or cortisol (5 µg/ml). At 5 dpf, larvae were transferred to 3L tanks with 1.5 L of water. Larvae were fed Gemma micro 150 twice daily. 50% of the water, dosed appropriately, was exchanged daily until 15 dpf. 15 dpf larvae were euthanized MS-222 (0.4 mg/L, Sigma, USA). Each n represents a pool of 3-5 larvae. After sampling, larvae were rinsed with 1xPBS and stored at -80°C for later analysis.

104

Protein quantification and western blotting:

SDS-PAGE and Western blotting was performed as previously described (Faught & Vijayan, 2018a). Protein quantification was done by homogenizing samples in 50 mM Tris buffer (pH 7.5) with a protease inhibitor cocktail (Roche, 04693116001). The homogenate was sonicated (3s pulse, ~5x) and centrifuged (13,000 xg for 2 min). The supernatant was removed, and the protein concentration was determined using the bicinchoninic acid (BCA) method using bovine serum albumin (BSA) as the standard. The homogenate was then diluted with Laemmli’s buffer (156.25mM Tris, 50% glycerol, 5% SDS, 0.0625% bromophenol blue and 25% 2- mercaptoethanol). Samples were stored at -20°C. Equal amounts of protein (40 µg) were separated on an 8% polyacrylamide gel and transferred to nitrocellulose membrane using a SemiDry transfer unit (BioRad). After transfer, membranes were blocked with a solution of powdered skim milk (5% w/v in TTBS (20 mM Tris, 300 mM NaCl, pH7.5 with Tween 0.1%) containing 0.02% sodium azide). Primary: anti-human insulin (Sigma, USA) used at a dilution of 1:1000. Anti-insulin was prepared in 5% BSA+0.02% sodium azide and membranes were incubated overnight at 4°C. Membranes were then washed with TTBS (5min,3x) and incubated for 1h with secondary antibody (1:3000 goat anti-rabbit IgG; Bio-Rad, 170-6515). Bands were detected with Clarity Western ECL substrate (BioRad, 170-5061). The molecular mass of the bands was confirmed using a low range molecular weight marker (FroggaBio, Canada). Equal loading was confirmed using ponceau (Bio Shop, Canada)

Transcript abundance:

Transcript levels of specific genes were measured by quantitative real-time PCR (qPCR). Total RNA was extracted from larvae using Ribozol reagent (VWR, Canada) according to the manufacturer’s instructions and quantified using a SpectraDrop Micro-Volume microplate (VersaMax, Molecular Devices, CA, USA). One microgram of RNA was treated with DNase I (Thermo Scientific, USA) to remove genomic contamination prior to cDNA synthesis using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, USA) according to the manufacturer’s protocols. Transcript levels were measured by qPCR in duplicate using gene-

105

specific primers exactly as described previously (Faught & Vijayan, 2018a). See Table 1 for primer specific annealing temperatures.

Table 1: Gene-specific primers Gene Forward (5’-3’) Reverse (5’-3’) Source

leptinR GGTCTCACTGCCTGTCCATT AGATGGTGCTGCTCCACT (Liu et al., 2010)

leptina TCGTCAGAATCAGGGAACAC CCCAATGATGAGCGTTGGA (Liu et al., 2012)

mystnb AGACCGCTGTGGCTGCTCAT GCGGAAAGCACTGGTAATGT (Nesan et al., 2012)

myog AGTGGACAGCATAACGGGAACAG GCTGGTCTGAAGGTAACGGTGAG (Nesan et al., 2012)

igf1 CCACGATCTCTACGAGCACA TCGGCTCGAGTTCTTCTGAT (Nesan et al., 2012)

redd1 ATGCAAGATCAGTTGATTTCCAGCC TCAGCATTCTTCAATCAGGAGCTCT (Feng et al., 2012)

murf1a GGAAGAAAACTGCCAGGCACAG CTGGGTGATCTGCTCCAGAAGATG (Shimizu et al., 2017)

murf1b CAGGACAATGCTCAACGTGCC CTTGCTCTTTGCCAATACGCTCTAAGAG (Shimizu et al., 2017)

β-actin GTCCCTGTATGCCTCTGG AAGTCCAGACGGAGGATG (Nesan & Vijayan, 2016)

Statistics:

Data is shown as mean ± SEM and statistical comparison analysed using Sigma Plot 13 (Systat Software, Inc). All data was analysed using a one-way ANOVA (Holm-Sidak post hoc). Gene and cortisol levels were transformed to meet the assumptions of normality and equal variance. Untransformed data are shown in all figures. Significance level (a) of p<0.05 was used in all cases.

106

Results

Growth regulation by MR and GR signalling (Fig 1):

Using MRKO and GRKO zebrafish larvae, combined with exogenous cortisol treatment, we established a CR activation paradigm (Faught & Vijayan, 2019)( Fig. 1A). Briefly, CR activation (Fig. 1A) is represented as: i) None - MRKO larvae only have GR, but because they have basal cortisol levels there is no GR activation. ii) MR – Wildtype (WT) larvae have both MR and GR, but GR is inactive under basal cortisol levels. iii) MR* - GRKO larvae are hypercortisolemic, but lack GR, and, therefore, this group represents MR activation under high cortisol levels. iv) MR+GR - WT larvae treated with cortisol will activate both GR and MR. v) GR - MRKO larvae treated with cortisol represents a scenario in which only GR is activated. Zebrafish larvae with no CR activated (2.4±0.2 mg; p=0.350) were comparable in weight to MR activated larvae (2.2±0.09 mg) and MR* larvae (2.8±0.2 mg; p=0.142) (Fig. 1B). Activation of GR in WT (MR+GR activation) reduced the mass by half (1.3±0.1 mg) compared to the vehicle-treated WT larvae (MR activation; p<0.001). In fish lacking MR (GR activation; 2.9±0.4 mg), this effect of cortisol was abolished and the larvae were similar in size to MR activated under high cortisol (MR* activation; p=0.512) and no CR activation (p=0.522), but larger than MR activation under basal cortisol conditions (p=0.010). Total larval protein content showed a clear response to MR or GR activation (Fig. 1C). Fish lacking CR activation had 30.8±1.0 mg total protein/larva. When MR was activated under low cortisol conditions, total protein content increased (48.7±0.2 mg/larva, p<0.001), and this increase was even greater when MR was activated under high cortisol levels (MR*; 64.0±6.06 mg/larva) compard to no CR activation (Fig. 1C). Activation of GR in the presence of MR caused a significant loss of protein (21.1±0.3 mg) compared to all other groups. This cortisol-mediated protein breakdown was abolished in the GR group in the absence of MR (GR; 48.5±6.4 mg) and had protein content similar to that of the MR activated larvae (p=0.920; Fig. 1C).

107

Figure 1. Regulation of larvae growth by MR and GR signalling: A) Representative zebrafish larvae at 15 dpf which represent different CR activation. MRKO larvae have no CR activated, WT larvae have only MR activated, GRKO larvae have MR activated under high cortisol conditions (MR*), WT larvae treated with cortisol (5 µg/ml) have both MR and GR activated, and MRKO larvae treated with cortisol only have GR activated. B) Weight of larvae (n=10-20), and C) protein content (n=10-20). All bars are ± SEM, p<0.05, statistical differences were evaluated with a one-way ANOVA. Bars with different letters denote significance. Abbreviations: Wildtype (WT) and glucocorticoid receptor knockout (GRKO), mineralocorticoid receptor knockout (MRKO).

108

GR activation stimulates proteolysis (Fig. 2):

To determine the extent to which cortisol-stimulation of proteolytic genes were involved in the differences observed in body mass and protein content, the transcript abundance of redd1, murf1a and murf1b were quantified. MR activation, under high cortisol levels (MR*), decreased redd1 transcript abundance compared to the MR activated larvae (p=0.019). Cortisol treatment stimulated redd1 transcript abundance 7-fold in both WT (MR+GR; p<0.001) and only GR activated larvae (MRKO+cortisol; p<0.001) (Fig. 2A). Similarly, exogenous cortisol treatment increased mRNA levels of murf1a 4-fold in the MR+GR activated larvae, and 7-fold in the GR activated larvae compared to the MR activated larvae (p<0.001) (Fig. 2B). MR activated larvae had similar murf1a transcript abundance compared to larvae with no CR activation (MR*, p=0.515; MR, p=0.645). Also, murf1b transcript abundance RNA levels were significantly higher in GR activated larvae (WT+cortisol; MRKO+cortisol) compared to the WT (MR activation; p<0.001) (Fig. 2C). MR activation under high cortisol level had lower murf1b transcript abundance compared to MR activated larvae with basal cortisol levels (p=0.004; Fig. 1C).

109

Figure 2. GR activation stimulates proteolysis: Transcript abundance of proteolytic genes involved in skeletal muscle atrophy in zebrafish larvae (15 dpf) treated with cortisol. Transcript abundance of the A) redd1, B) murf1a and C) murf1b. All bars are ± SEM, p<0.05 (n=4-6), statistical differences were evaluated with a one-way ANOVA. Bars with different letters are significantly different (p<0.05). CR activation: None (MRKO), MR (WT), MR* (GRKO), MR+GR (WT + Cortisol), GR (MRKO +Cortisol).

110

Myogenin (myog) but not myostatin (mystnb) is regulated by GR (Fig. 3):

There was no effect of genotype or cortisol treatment on the transcript abundance of mystnb (p=0.410; Fig. 3A). MR activation did not significantly affect myog transcript abundance (p<0.001) compared to no CR activation (None; Fig 3B). GR activation either with MR (MR+GR activation) or in the absence of MR (GR activation) significantly reduced myog transcript abundance compared to the MR* activation larvae (p<0.001; Fig. 3B). There was no difference in myog transcript levels between MR activation larvae and MR+GR activation larvae (p=0.134; Fig. 3B).

MR and GR regulates GH/insulin/IGF1 expression (Fig. 4):

To determine the endocrine factors involved in the anabolic processes, we examined the transcript abundance of GH and IGF1, as well as the protein expression of insulin (Fig. 4). The transcript abundance of gh was lower in MR activated fish compared to larvae with no CR activation (Fig. 4A). GR activation in the presence of MR (MR+GR activation) increased gh transcript abundance nearly 3-fold in comparison to MR activated larvae (p<0.001), and larvae with no CR activation (Fig. 4A). Activation of GR alone, without the presence of MR, lowered gh transcript abundance to levels similar to larvae with no CR activation (p=0.133), but not different from the MR+ GR activated larvae (p=0.292; Fig. 4A). There is no difference in igf1 transcript abundance between larvae with no CR activation and larvae in which MR was activated (MR; p=0.930 and MR*; p=0.990). Igf1 transcript abundance is higher when GR is activated in the presence of MR (MR+GR; p=0.016), but not when GR alone is activated (Fig. 4B). GR activation without MR returns igf1 transcript abundance to levels seen in larvae that were no treated with cortisol (None, MR and MR* larvae; Fig. 4B). Insulin expression was similar between larvae with no CR activation and MR activation under low cortisol levels (None vs MR, p=0.230; Fig. 4C). Activation of MR under high cortisol levels (MR*), and in the presence of GR activation (MR+GR larvae) resulted in greater insulin expression compared to the MR activation larvae (p=0.016 and p=0.009, respectively; Fig. 4C). GR activation, in the absence of MR, did not increase insulin expression, and was similar to larvae with no CR activation (p=0.937) or MR activation (p=0.100) (Fig. 4C).

111

Figure 3. Myogenin (myog) but not myostatin (mystnb) is regulated by GR: A) Myostatin b (mystnb) and B) myogenin (myog) gene transcription in 15 dpf zebrafish. All bars are ± SEM, p<0.05 (n=4-6), statistical differences were evaluated with a one-way ANOVA. Bars with different letters are significantly different. CR activation: None (MRKO), MR (WT), MR* (GRKO), MR+GR (WT + Cortisol), GR (MRKO + Cortisol).

112

Figure 4. MR and GR regulates GH/insulin/IGF1 expression: To determine somatotropic axis potential transcript abundance of A) growth hormone (gh) and B) insulin-like growth factor 1 (igf1) transcript abundance was determined in 15 dpf zebrafish treated with cortisol for 12 days (3 dpf-15 dpf). C) Insulin, as a key player in anabolic processes, was also measured. The densometric expression is represented as % control (vehicle-treated WT) and a representative dot blot is below C. All bars are ± SEM (n=4-6), significance was determined using a one-way ANOVA, p<0.05. Bars with different letters are significantly different. CR activation: None (MRKO), MR (WT), MR* (GRKO), MR+GR (WT + Cortisol), GR (MRKO + Cortisol).

113

Regulation leptin by GR and MR (Fig. 5): lepa: Activation of MR increased lepa mRNA 10-fold compared to larvae that did not have MR activation (Fig. 1C). GR activation with exogenous cortisol treatment had no effect on lepa compared to MR activated larvae (WT, p=0.42). MR activation under high cortisol levels decreased the abundance of mRNA by half compared to MR activation under basal cortisol levels (p=0.014) or MR+GR activation (p= 0.026) (Fig. 5C). LepR: MR activation reduced lepr mRNA in all cases (Fig. 5D). Larvae with no CR activation had 3-fold higher lepr transcript abundance compared to larvae where MR was active. GR activation in the absence of MR increased lepr mRNA 6-fold compared to MR activated larvae, and 2-fold compared to larvae with no CR activation (p=0.045).

114

Figure 5. Regulation leptin by GR and MR: Transcript abundance of A) leptin A (lepA) and B) leptin receptor (lepr) were also assessed. All bars are ± SEM (n=4-6), a one-way ANOVA was used to determine significance, p<0.05. Bars with different letters are significantly different. CR activation: None (MRKO), MR (WT), MR* (GRKO), MR+GR (WT + Cortisol), GR (MRKO + Cortisol).

115

Figure 6 CR activation paradigm: A representative image of GR and MR activation with increasing cortisol levels and their roles in metabolism. Under basal cortisol conditions only MR is active, and this promotes protein accumulation. As the cortisol concentration increases MR signalling is amplified, as seen with increased protein, insulin signalling and a marked decrease in key proteolytic genes. This may be due to either the absence of GR, or differential binding of MR to different response elements. MR signalling in combination with GR causes an increase in anabolic processes (increase in transcript abundance of gh, igf1). The activation of GR alone, strongly upregulates catabolic processes such as proteolytic genes (redd1, murf1a and murf1b), but MR is required for this response.

116

Discussion:

Here we report a physiological role for MR activation in the regulation of postnatal protein status in response to GCs stimulation in zebrafish larvae. The negative effect of excess GCs on growth is conserved in vertebrates. From chronically stressed salmonid parr (McCormick et al., 1998), to therapeutics in children (Kelly et al., 2012), excess GCs during the postnatal periods are associated with reduced growth (Cianfarani et al., 2002; Greenwood & Bell, 2003), but the mechanisms are far from clear. Here we show that activation of both MR and GR by excess cortisol mediates the growth suppression, and the associated lower protein content, during postnatal development in zebrafish (Fig. 1). MR activation by itself, in the absence of GR activation, led to protein accumulation confirming an anabolic role for this receptor activation. Paradoxically, MR is required to promote GR-mediated protein breakdown due to cortisol treatment, as GR activation alone failed to reduce the larval protein content (Fig. 1C). To our knowledge, this is the first study to show that both GR and MR activation cooperatively mediates the cortisol-mediated growth suppression during postnatal development in zebrafish. Excess cortisol promotes proteolysis of skeletal muscle mass in mammals, and this is due to the upregulation of several ubiquitin ligases, and regulators of mTOR, which are GR-responsive (Kuo et al., 2013; Britto et al., 2014). This also is the case in zebrafish larvae, as GR activation upregulates murf1a, murf1b, and redd1 transcript levels, supporting a conserved role for GCs in muscle proteolysis ((Faught & Vijayan, 2019); Fig. 2). In all cases of the measured muscle proteolytic genes (Fig. 2), GR and not MR is the primary transcription factor regulating these genes. Also, myog, a transcription factor involved in skeletal muscle development and protein synthesis (Kuo et al., 2013), was downregulated in response to GR activation compared to larvae in which only MR was activated under high cortisol levels (Fig. 3B). These results support the conserved role of GR activation by excess cortisol as a key mediator of muscle wasting and negative protein balance during chronic stress (Kuo et al., 2013). Neither cortisol treatment nor genotype had an effect on mystnb; however, the role of mystnb in teleost is thought to be divergent from its role in mammals (Galt et al., 2014). Our results indicate that MR has a minimal role in cortisol-mediated proteolysis. This may have evolutionary significance, as MR is considered to be the type I CR, mediating actions under basal cortisol conditions, while GR, as the type II CR, is primarily responsible for mediating GCs

117

action post-stress, when cortisol levels are elevated (Joëls & de Kloet, 2017). The chronic stress- mediated muscle protein breakdown occurring during periods of elevated cortisol levels will increase available amino acids for hepatic gluconeogenesis (Mommsen et al., 1999). The increase in transcript abundance of proteolytic genes in zebrafish supports a conserved role for GR activation in promoting muscle wasting (Faught & Vijayan, 2019). Thus, MR activation under basal condition may be playing an important role in protein homeostasis, but this is disturbed in response to elevated cortisol levels. Interestingly, our results indicate that GR or MR activation by itself favours protein deposition, whereas the activation of both receptors as seen during excess cortisol stimulation is required for protein breakdown in zebrafish. This leads us to propose that the interaction of both receptors in response to elevated cortisol levels may be a key factor in the metabolic adaptation to stress, but that the control by CRs may be gene specific. It is generally accepted that cortisol signalling, especially in terms of metabolic regulation, occurs through GR in fish (Mommsen et al., 1999; Mommsen, 2001). There are evidences that MR is involved in lipid accumulation and adipogenesis (Marzolla et al., 2012), but the extent to which MR regulates key metabolic genes, either alone or in conjunction with GR, is unknown. The capacity for increased protein content in the MR activated larvae is supported by the lower expression of redd1 and murf1b in the MR* groups (Fig. 2A, C), suggesting also an anti- catabolic role for MR in promoting protein deposition. Under basal conditions ~80% of MR are bound by GCs (de Kloet & Reul, 1987), but there is little interaction of MR with GREs as indicated by the absence of an activation of GR-responsive gene (Fig. 2). MR interaction with GREs occurs under high cortisol levels (Mifsud & Reul, 2016), and this may be the reason for the reduction in transcript changes seen in the redd1 and murf1b in the MR* groups, which are hypercortisolemic (Ziv et al., 2013; Faught & Vijayan, 2018a, 2019). This indicates that the role of MR, even under stress conditions, is to limit catabolic processes and promote anabolic processes, and may be a compensatory mechanism to limit catabolic process to facilitate metabolic recovery post-stress. The mechanisms surrounding the interaction of GR and MR is unclear, but recent work suggests that either heterodimerization (Kiilerich et al., 2015; Mifsud & Reul, 2016), tethering (Rivers et al., 2019), or steric interference in the promoter regions of target genes may be involved (Mifsud & Reul, 2018).

118

We next examined whether MR effect on protein deposition may involve regulation of anabolic hormones, including insulin, gh and igf1 in zebrafish larvae. To date, the effects of cortisol on GH regulation in fish are associative at best, and the mechanisms involved are far from clear. For instance, during periods of rapid growth, as in the case of Atlantic salmon parr, repeated stress, and the attendant rise in cortisol, was associated with an increase in circulating GH levels (McCormick et al., 1998). However, in adult rainbow trout, there was no change in GH levels with acute stress (Shepherd et al., 2011), which suggests that GH stimulation during stress may be related to the developmental stage, and maturity of the animal. Also, in chronically stressed rainbow trout higher transcript abundance of the two GH paralogs was associated with lower circulating GH levels, but increased hepatic igf1 mRNA (Madison et al., 2015). In all of these studies, it is unknown whether either GR or MR or interaction of these receptors were involved in mediating the cortisol-induced changes in GH levels. Our results indicate that both GR and MR activation regulates the GH/IGF1 transcript abundance during postnatal growth in zebrafish. The transcript abundances of gh and igf1 are increased when both GR and MR are activated (MR+GR larvae), but neither GR nor MR activation alone affects these genes (Fig. 4B). This lack of regulation of gh and igf1 transcripts with MR activation alone suggests that these anabolic hormones may not be the MR targets promoting the protein accumulation observed in these larvae. However, during chronic stress and high cortisol levels, MR increases insulin expression and this may promote muscle glucose utilization and protein synthesis. This is supported by the recent observation that GRKO adult zebrafish, in which MR is activated under high cortisol levels, showed no difference in insulin expression, but there was an increase in glucose uptake, which was associated with increased phosphorylation of translation elongation factor eIF4B, and increased protein synthesis (Faught & Vijayan, 2019). Taken together, our results suggest that the promotion of protein deposition by MR under elevated cortisol levels, in the absence of GR (MR* activation) may involve a suppression of proteolytic genes (Fig. 2A, C), coupled with an increase in insulin responsiveness. Further studies are warranted to elucidate the mechanisms involved. It is clear from the work presented here that dysregulation of either catabolic or anabolic pathways by differentially activating MR or GR leads to marked differences in growth in zebrafish in response to GC stimulation.

119

Food consumption is also another contributing factor to growth and protein accretion, therefore, we evaluated whether leptin was differentially regulated based on CR activation (Michel et al., 2016). The role of leptin in the control of feeding in the lower vertebrates is not as clear as in mammals (Michel et al., 2016). For instance, leptin expression is more pronounced in the liver than in the adipose tissue, and regulates glucose levels in teleosts (Michel et al., 2016). Stress and the attendant rise in GCs modulate leptin expression in mammals (Zakrzewska et al., 1997) and fishes (Löhr et al., 2018); however, the effect of GCs on circulating leptin levels lacks consensus. In humans, prolonged treatment with GCs increases circulating levels of leptin, and inhibits food intake (Uddén et al., 2003), but reducing cortisol synthesis have also been shown to increase serum leptin concentrations in humans (Copeland et al., 2011). Leptin can also attenuate corticosterone-induced increases in plasma insulin levels (Solano & Jacobson, 1999), suggesting that GCs/leptin interaction may be important in maintaining energy substrate homeostasis. GCs can also counteract the effects of leptin, as exogenous GC treatment will reverse leptin-induced reductions in weight gain and body fat in mammals (Solano & Jacobson, 1999). Interestingly, we are showing for the first time that MR activation upregulate the transcript abundance of lepa, the primary paralog of leptin in zebrafish (Michel et al., 2016). However, MR activation also reduced the lepr transcript abundance leading to the proposal that MR activation may cause leptin resistance in zebrafish. The lack of leptin responsiveness may encourage the protein gain seen in larvae with MR activation, but this is lost in the presence of GR. Further work needs to be carried out to fully elucidate the interaction between GCs and leptin signalling in facilitating the metabolic repartitioning during postnatal growth in zebrafish. In conclusion, MR and GR activation plays a key physiological role in the regulation of postnatal growth in zebrafish. While adaptive in the context of starvation or acute infection, prolonged periods of elevated GCs during early development can result in muscle loss and long- term developmental dysfunction. Here we show that MR activation, under both basal and stressed conditions are critical for the regulation of protein turnover (Fig. 6). MR activation stimulates protein deposition, and this includes inhibiting proteolytic genes (anti-catabolic) and stimulating insulin responsiveness (anabolic processes). GR is activated under high cortisol levels and will stimulate proteolysis, but the protein breakdown seen with excess cortisol requires the activation of both GR and MR (Fig. 6). We propose that MR may be playing a

120

compensatory role in protein recovery post-stress and this requires the interaction with GR and involves upregulation of the gh/igf1 transcript levels. MR modulation of leptin resistance may also be another mechanism involved in energy repartitioning during cortisol stimulation, but this remains to be determined. Overall, both MR and GR activation have key roles to play in protein homeostasis in response to stress during postnatal development. Understanding the mechanisms of these receptor interactions, and the target genes involved, may have biomedical applications by reducing and managing stress related growth dysfunctions.

121

CHAPTER 6: LOSS OF THE GLUCOCORTICOID RECEPTOR IN ZEBRAFISH

IMPROVES MUSCLE GLUCOSE AVAILABILITY AND INCREASES GROWTH

122

Introduction

Glucocorticoids (GC) are one of the most primitive regulators of metabolic homeostasis. These steroid hormones, named for their ability to regulate glucose levels, are released during stress and are the hallmark of the conserved stress response in vertebrates. The glucose response is essential to fuel the increased energy demand during stress, and GCs facilitate this by enhancing liver gluconeogenesis (Charmandari et al., 2005; Vegiopoulos & Herzig, 2007; Kuo et al., 2013). GCs also increase the availability of other energy substrates, including free fatty acids and amino acids from endogenous stores, to offset the increased energy demand associated with stress coping (Charmandari et al., 2005; Vegiopoulos & Herzig, 2007; Kuo et al., 2013). The metabolic effect of glucocorticoids are primarily mediated by activation of the glucocorticoid receptor (GR), a ligand-bound transcription factor that is highly conserved in vertebrates (Charmandari et al., 2005; Stolte et al., 2006). In teleosts, the primary glucocorticoid is cortisol and is produced in response to the hypothalamus-pituitary-interrenal (HPI) axis activation (Mommsen et al., 1999), which is analogous to the HPA axis activation of higher vertebrates (Wendelaar Bonga, 1997). The adrenal cortex is not present as a distinct gland in teleosts, and the corticosteroid-producing interrenal cells are distributed throughout the head kidney region (Wendelaar Bonga, 1997; Mommsen et al., 1999). In both cases of HPA/HPI axis activation, corticotropin-releasing hormone (CRH) is secreted from the hypothalamus, stimulating the pituitary to release proopiomelanocortin (POMC), a precursor protein cleaved into adrenocorticotropic hormone (ACTH). ACTH binds to the melanocortin 2 receptor (MC2R) in the interrenal cells, stimulating steroidogenesis, and the associated increase cortisol production (Mommsen et al., 1999; Charmandari et al., 2005). Although an acute stressor-mediated cortisol increase is essential for stress adaption, a chronic increase of glucocorticoids has a negative impact on growth and weight maintenance that is highly conserved in vertebrates (Mommsen et al., 1999). The molecular mechanisms of glucocorticoid regulation of skeletal muscle protein metabolism have been well characterized in mammalian models. For instance, glucocorticoid stimulation causes muscle atrophy by regulating REDD1 (regulated in development and DNA damage response 1), leading to suppression of mTORC1 (mammalian target of rapamycin complex 1) signalling and an associated reduction in protein synthetic rates (Britto et al., 2014; Gordon et al., 2016). In

123

mammals it has been postulated that muscle mass is regulated primarily by alterations in protein synthetic rates, and that changes in muscle protein degradation are secondary and adaptive (Rennie et al., 1983). Other genes that are upregulated during muscle wasting in mammals include, two muscle-specific E3 ubiquitin ligases: muscle RING finer 1 (MuRF1) and muscle atrophy box (MAFbx) as well as myostatin (Kuo et al., 2013). While there are orthologs of these genes in fish, it is unknown whether these genes are under transcriptional control of GR. Indeed there is evidence to suggest that the function of these genes, in particular myostatin (mystnb), are not conserved in lower vertebrates (Galt et al., 2014). A direct role of GR signalling in modulating weight changes comes from conditional knockouts and pharmacological studies in mammals. Ubiquitous GR knockouts in mammals leads to lethality, due to delayed lung development (Tronche et al., 1998). Mice lacking muscle GR (GRmKO) had increased muscle mass, but smaller adipose tissue (Shimizu et al., 2015). Whereas mice lacking liver GR (GRlKO) had impaired growth hormone signalling, reduced growth (Tronche et al., 1998), and increased lipid stores (Mueller et al., 2011). These studies highlight the tissue-specific metabolic changes associated with GR signalling. However, the system level changes associated with GR signalling, including glucose regulation and its affect on muscle protein metabolism and growth is unknown. Muscle makes up over 50% of adult fish mass and, therefore, any change to this organ system, including enhanced energy substrates mobilization to offset the increased energy demand during stress, will be reflected in lower growth rate (Mommsen, 2001; Sadoul & Vijayan, 2016). As in mammals, GR signalling also increases glucose production via hepatic gluconeogenesis in teleosts (12, 30). However, whether glucocorticoids can modulate glucose uptake in target tissues during chronic stress, as in mammals, is unknown (Bernal-Sore et al., 2018). While many teleosts are considered to be glucose intolerant due to their limited capacity to clear a bolus of glucose, this does not appear to be the case in zebrafish (Milligan, 1997, 2003; Moon, 2001; Maddison et al., 2015; Zang et al., 2017). Recently mifepristone, a GR antagonist, was approved for use in type II diabetes; however, the physiological consequences of using this drug to modulate metabolism at the system level is unknown (Bernal-Sore et al., 2018). Indeed, the physiological impact of increased glucose uptake on muscle intermediary metabolism, and whether this is regulated by GR signalling is unclear in any animal model. A lack of muscle

124

glucose sensitivity has been shown to promote protein catabolism in mammals (Patel et al., 2001; Jeyapalan et al., 2007), but whether GR signalling is involved in this process is unclear. Against this backdrop, we tested the hypothesis that GR signalling mediates the glucocorticoid- induced reduction in muscle glucose uptake, causing changes to protein synthesis/breakdown, leading to lower body mass during hypercortisolemia. Using zebrafish (Danio rerio), we generated ubiquitous GR knockout (GRKO) using CRISPR/Cas9 mutagenesis. Zebrafish is the only teleost identified so far lacking a GR paralog (Alsop & Vijayan, 2008, 2009a), and unlike mammalian models (Tronche et al., 1998; Mueller et al., 2011; Shimizu et al., 2015), GRKO are not lethal in this species (Griffiths et al., 2012; Ziv et al., 2013; Facchinello et al., 2017). This makes zebrafish an ideal model to explore the metabolic changes associated with the absence of GR at the organismal level, and eliminates compensatory mechanisms evident in the conditional mammalian knockouts (Shimizu et al., 2015). Our results indicate that lack of a functional GR enhances muscle glucose utilization and reduces the anti-anabolic and catabolic capacities of this tissue, suggesting a direct role for GR signalling in the growth suppression with chronic stress. Given the conserved nature of glucocorticoid action, our results suggest that GRKO zebrafish is an excellent model for translational research to determine the fundamental roles of stress on energy substrate re-partitioning and nutrient homeostasis (Seth et al., 2013).

Materials and Methods:

Zebrafish maintenance Adult (9-month-old) zebrafish (Tupfel long fin (TL) strain) were maintained according to standard protocols. All procedures were performed according to a protocol approved by the University of Calgary Animal Care Committee (AC17-0079), and followed the guidelines set by the Canadian Council for Animal Care. All fish were maintained in 10L tanks on a recirculating system with a 14:10 light:dark cycle (Tecniplast, Italy). Water was maintained at 28.5°C, pH 7.6, and 750 µS conductivity. Animals were fed twice daily with Gemma micro 300 diet [Skretting, USA, 3% body weight; composition: protein (59%), oil (14%), ash (14%), fiber (0.2%) and phosphorus (1.3%)] in the morning and live Artemia (San Francisco Bay Brand, USA) in the afternoon. We used CRISPR/Cas9 mutagenesis to generate the ubiquitous GR knockout

125

zebrafish, and this was described recently (Faught & Vijayan, 2018a). Briefly, 7 base pairs have been deleted in exon 2 of the nr3c1 gene (corresponding to the N-terminal domain of the GR protein), resulting in a non-functional, truncated protein(Faught & Vijayan, 2018a). In all cases, fish were scarified before the first feed (10:30 am). Body composition measurements Moisture content was the difference in weight before and after drying the tissue for 48h at 70oC in an oven (Thermo Fisher Scientific, USA). Dehydrated tissue was used for total lipid content determination using the Folch method (Lazaro-Côte et al., 2018). Briefly, tissue was transferred to a 1.5 ml centrifuge tube and homogenized in 1 ml chloroform/methanol (2:1). The homogenate was incubated for 30 min after which the supernatant was removed to a clean glass tube and washed with 2x volume of 0.9% saline and vortexed for 30 s. The phases were allowed to separate for 30 min on an orbital shaker after which the lipid phase was transferred to a pre- weighed glass tube using a microcapillary pipette. The chloroform from this lipid/chloroform phase was allowed to evaporate overnight and the weight of lipids was determined gravimetrically. Lipid content was measured from female fish with gonads excised and, therefore, measured lipids in the somatic tissue. Total protein was measured using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Glucose and glycogen contents were analyzed as described previously (Lazaro-Côte et al., 2018). Feeding Trials Individual fish were placed in 1L water in a 2L clear tank and fed 10 pellets at a time. Additional pellets were added when the fish had eaten all previous pellets. Pellets were counted at 5-min intervals over a 20 min period and expressed as the total weight of pellets (average pellet weight 4.8 mg ± 0.1) consumed per fish. Trials were carried out with a minimum of 5 independent fish in a day and repeated on two different days. Stress Experiment

Fish were subjected to an acute handling disturbance as described previously (Alderman et al., 2012). Briefly, adult zebrafish were exposed to a 1 min air exposure. Fish were sampled by euthanasia in 0.5 mg/L of MS222 (buffered 1:2 with sodium bicarbonate) either prior to (0 h) or 0.5, 1 and 2 h post-stressor exposure. Fish were weighed, gonads and liver removed, and all tissues were frozen at -80oC for later analysis.

126

Cortisol and glucose determination

Whole body cortisol was extracted and quantified using an ELISA as previously described for zebrafish (Faught et al., 2016b). The intra-assay and inter-assay variations were <5% and <15%, respectively. Blood glucose level was determined by FreeStyle glucose strips and meter (Abbott, Mississauga, Canada) as described previously for zebrafish (Eames et al., 2010). In Vivo 2-NB-deoxy-D-glucose uptake

Glucose uptake in vivo followed the protocol of Itoh et al. (2004) with modifications. Briefly, fish were injected intraperitoneally with 25 umol/kg of 2-[N-7-nitrobenz-2-oxa-1,3-diazol-4-yl) amino-2-deoxy-D-glucose (2-NBDG), a fluorescent analogue of 2-deoxyglucose. Fish were allowed to recover for 1h and euthanized with 0.5mg/L MS222 (buffered 1:2 with sodium bicarbonate). White epaxial muscle and livers were removed and immediately homogenized in 50 mM Tris+ protease inhibitor using a sonicator (Fisher Scientific, 6x3s bursts, setting 3-4). Samples were centrifuged at 13,000 x g for 1min and the supernatant was added to a black 96 well plate and read by a Paradigm plate reader (excitation/emission 465/540 nm; Molecular Devices, USA). Saline-injected fish were used as background controls. Preliminary experiment revealed that zebrafish muscle and liver took up 2-NBDG in a dose-dependent manner. Immunodetection

SDS-PAGE and western blotting were performed as previously described (Faught & Vijayan, 2018a). For the dot blot, 2 µl of denatured protein samples in Laemmli’s buffer (normalized to 2mg/ml protein) were added to a nitrocellulose membrane and dried for 2h at 37°C. Immunodetection was carried out as described previously (Faught & Vijayan, 2018a). Primary antibodies: anti-phospho eIF4B (Cell Signalling Technologies #3591T, 1:1000; the antibody for total eIF4B (3592T) did cross react with zebrafish); anti-ubiquitin (Sigma U5379, 1:100); anti- insulin (Agilent Technologies #A056401-2, 1:500), Anti-Hsp70/Hsc70 (SPA-812, 1:1000) Equal loading for eIF4B and Hsc/Hsp70 western blots was assessed using Ponceau staining (0.1% w/v in 5% acetic acid; Bioshop, Canada). All densiometric quantification was done with ImageJ software (National Institutes of Health, USA).

127

Fasting study

Age-matched WT zebrafish were randomly assigned to two 3L tanks (12 fish/tank). Additionally, GRKO fish (age-matched with WT) were also randomly assigned to two 3L tanks. All fish were held on a recirculating system. One WT and one GRKO tank were fed 3% of their body weight/day and the other tank of WT and GRKO fish were fasted for 7 days. On day 7, fish were euthanized with MS222, muscle, and liver removed and stored at -80 for further analysis. Total muscle protein concentration was calculated as protein* total muscle weight. Muscle weight was taken as 50% of the total body mass (Mommsen, 2001). Transcript Abundance:

Transcript levels of specific genes were measured by quantitative real-time PCR (qPCR). Total RNA was extracted from larvae using Ribozol reagent (VWR, Canada) according to the manufacturer’s instructions and quantified using a SpectraDrop Micro-Volume microplate (VersaMax, Molecular Devices, USA). One microgram of RNA was treated with DNase I (Thermo Scientific, USA) to remove genomic contamination prior to cDNA synthesis using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, USA), according to the manufacturer’s protocols. Quantitative real-time PCR was carried out using a QuantStudio 3 Real-time PCR system (Applied Biosystems, USA), with an Sso Advanced SYBR green master mix (Bio-Rad, USA). Gene-specific primers are detailed below.

Table 1: Primers used for quantitative real-time PCR. Target Forward primer (5’-3’) Reverse primer (5’-3’) Source murf1a GGAAGAAAACTGCCAGGCACAG CTGGGTGATCTGCTCCAGAAGATG (Shimizu et al., 2017) murf1b CAGGACAATGCTCAACGTGCC CTTGCTCTTTGCCAATACGCTCTAAGAG (Shimizu et al., 2017) redd1 ATGCAAGATCAGTTGATTTCCAGCC TCAGCATTCTTCAATCAGGAGCTCT (Feng et al., 2012) mystnb AGACCGCTGTGGCTGCTCAT GCGGAAAGCACTGGTAATGT (Nesan et al., 2012)

Β-actin TGTCCCTGTATGCCTCTGGT AAGTCCAGACGGAGGATG (Nesan et al., 2012)

128

Enzyme Activities

Enzyme activities were measured as described previously (Ings et al., 2011). Whole body homogenate was stored frozen in 50% glycerol buffer (50% glycerol, 21mM Na2HPO4, 0.5mM EDTA, 0.2% BSA and 5mM β-mercaptoethanol, pH7.5). Enzyme kinetics were measured in 50mM imidazole (pH 7.4) at 340 nm using a microplate reader (VersaMax, Molecular Devices, California, USA).

Hexokinase (HK: EC 2.7.1.1): 1mM glucose, 5mM MgCl2, 10mM KCl, 0.25mM NADH, 2.5mM phosphoenolpyruvate (PEP), 5U/mL lactate dehydrogenase (LDH) and 2.5U/mL pyruvate kinase; reaction started with 1mM ATP. Pyruvate kinase (PK: EC 2.7.1.40): 3mM KCl, 10mM MgCl2, 0.12mM NADH, 2.5mM ADP, 20U/mL LDH; reaction started with 2.5mM PEP. Alanine aminotransferase (AlaAT: EC 2.6.1.2): 0.12mM NADH, 200mM L-alanine, 0.025mM pyridoxal 5-phosphate, and 12U/mL LDH; reaction started with 10.5mM a-ketoglutarate. Aspartate aminotransferase (AspAT: EC 2.6.1.1): 7mM a- ketoglutarate, 0.025mM pyridoxal 5- phosphate, 0.12mM NADH, and 8U/mL malate dehydrogenase; reaction started with 40mM aspartic acid.

Malic Enzyme (MS: EC 1.1.1.40): 1mM MgCl2, 0.4mM NADP; reaction started with 1mM L- malate ATP citrate lyase (EC 4.2.3.8): 20mM citrate, 0.2 mM coenzyme-A, 0.1mM NADH, 10mM

MgSO4, 10U/ml malate dehydrogenase; reaction started with 5mM ATP.

Glucose-6-phosphate dehydrogenase (G6PDH, EC 1.1.1.49): 7mM MgCl2, 0.4 mM NADP; reaction started with 1mM glucose-6-phosphate. Muscle Metabolomics

Epaxial muscle from the dorsal, posterior part of the fish was homogenized in 50mM Tris and extracted using methanol. Briefly, muscle homogenate was diluted 5x in 50% ultrapure methanol and incubated on ice for 15 min. Samples were then centrifuged for 10 min at max speed at 4°C. The supernatant was then analyzed using liquid chromatography (Vanquish UHPLC, Thermo Fisher Scientific)-mass spectrometry (LC-MS; Q-Exactive HF hybrid Quadruple-orbitrap, Thermo Fisher Scientific) at the University of Calgary’s Metabolomics

129

Research Facility (CMRF). Spectral intensity data was analyzed by the CMRF, using an in-house metabolite library in MAVEN (Metabolomic Analysis and Visualization Engine). Peak intensity data was normalized to muscle wet weight and analyzed using MetaboAnalyst (http://www.metaboanalyst.ca/). Statistics

Data are shown as mean ± SEM, and statistical comparisons were carried out using Sigma Plot 13 (Systat Software, Inc, USA). All data were analyzed using a t-test (p<0.05), unless otherwise stated. Muscle metabolomics was analyzed using one-way Anova (False discovery rate (FDR) corrected p<0.05; Fisher’s LSD post-hoc) in MetaboAnalyst. Weight data (Fig. 1B), stress response data (Fig. 2) and fasting data (Fig. 6) were analyzed with a two-way ANOVA (Holm-Sidak post-hoc). All data were transformed to meet the assumptions of normality and equal variance. Untransformed data are shown in all figures, and a significance level of p<0.05 was used in all cases.

Results

Loss of GR increases body mass and alters body composition Adult GRKO fish had higher body weight compared to WT fish, independent of sex (Fig. 1A, B; p=0.002). Female GRKO fish had 1.5-fold higher body mass (n=6, 870.76 ± 129.4 mg) compared to WT fish (n=12, 551.93 ± 89.1 mg; Fig. 1A, B) and male body size of the GRKO fish (n=6, 709.4± 27.9 mg) was also larger compared to WT (n=12, 589.2 ± 63.4 mg). As the gonadosomatic index was not different between the WT and GRKO female fish (Fig. 1C, n=6, p=0.479), the differences observed were due to the somatic tissues mass. The higher body mass was also not a reflection of increased appetite, as the GRKO fish ate only a third of the food consumed by WT fish (p=0.0191; Fig. 1D). The adult GRKO mutants differed in body composition, and the measurements were carried out with somatic tissue of female fish. GRKO had higher protein (Fig. 1E; p=0.0217) and total lipid content (Fig. 1F; p=0.007), while moisture content (Fig. 1G; p=0.00561) was lower compared to the WT. There were no changes in whole- body glucose or glycogen content between the two groups (Figs. H and I, respectively).

130

Figure 1: GRKO zebrafish are larger than WT, with a different body composition. A) A representative image of glucocorticoid receptor knockout (GRKO) adult female zebrafish and a wildtype (WT) female zebrafish. B) The body mass of GRKO zebrafish, both male and female, was higher than their WT counterparts, however, C) The gonadosomatic index was not significantly different between the two groups. D)The GRKO fish consumed less food, E) had greater total protein content, and F) higher total lipid content, and G) less moisture compared to the WT controls. There was no change in carbohydrate levels, including H) whole-body glucose and I) glycogen contents. All bars are mean ± SEM (n=4-6); Fig B inset shows main effect (two-way ANOVA; p<0.05); *significantly different from the WT (t-test; p<0.05).

131

GRKO zebrafish have a functional HPI axis, but an impaired glucose response to stress

We next tested whether glucose regulation was central to the GR-mediated metabolic programming. To determine whether target tissue glucose utilization was involved in modulating the nutrient stores during stress, we first measured whole body cortisol and blood glucose levels post-stressor exposure (Fig. 2A and 2B). GRKO fish were hypercortisolemic compared to the wildtype at every time point post-stressor exposure (p<0.001). There was a significant interaction between genotype and time for cortisol levels post-stressor (p=0.043; Fig. 2A). Within WT, cortisol levels increased 3-fold from 13.6 ± 7.9 ng/g ww to 39.6 ± 5.7 ng/g ww at 0.5h post- stress (p<0.001). This remained elevated for the rest of the 2h sampling (Fig. 2A). Within the GRKO, basal cortisol levels (99.7 ± 58.8 ng/g ww) increased 2-fold to 212.8 ± 68.0 ng/g ww at 0.5h post-stressor (p<0.001) and remained elevated for the duration of the 2-h sampling time point (Fig. 2A). There was only a treatment effect in blood glucose levels (Fig. 2B). WT had higher glucose levels compared to GRKO zebrafish (p=0.015; Fig. 2B), and there was no interaction between genotype and time post-stressor exposure for blood glucose levels.

132

Loss of GR increased glucose uptake in the muscle We then tested whether the lack of a glucose response in the GRKO was due to a faster glucose clearance from the circulation. Therefore, we tested glucose uptake in target tissues in vivo by injecting fish interperitoneally with 2-NBDG, a fluorescent analogue to 2-deoxyglucose commonly used to trace glucose utilization in neurons (Itoh et al., 2004). Injection of 2-NBDG raised blood glucose levels in the WT (n=8, 5.1 ± 0.7 mM), but not in the GRKO fish (n=6, 1.4 ± 0.2 mM) compared to the uninjected and saline-injected controls (Fig. 3A). There was no change in 2-NBDG fluorescence in the liver of the GRKO fish compared to the WT (Fig. 3B; p=0.0921). Unlike the liver, the GRKO fish muscle had double the amount of 2-NBDG (162,604 ± 10,087 RFU/mg tissue) compared to the WT muscle (79,105 ± 23,075 RFU/mg tissue), indicating an increased rate of glucose uptake (twice the rate; p=0.01) over the hour post-injection (Fig. 3C). We further confirmed an increased potential for glucose utilization by measuring hexokinase activity (pre-2NBDG injection). Adult GR-KO fish had >2-fold maximal hexokinase activity (1.1 ± 0.2 U/g) compared to the WT (0.414 ± 0.025 U/g; Fig. 4A). To test whether insulin may play role in the increased tissue glucose capacity in the GRKO fish, we determined insulin expression. Dot blot analysis of whole-body insulin showed that the expression was similar in the GRKO and WT fish at either the resting state or post-2NBDG injection (Fig. 4B).

133

Figure 2: GRKO have an impaired glucose response to stress: Adult zebrafish (mixed sex) were subjected to a 1-minute air exposure and sampled at different time-points post- stressor. A) GRKO zebrafish are hypercortisolemic compared to the WT, and both groups increase cortisol levels by 30 min post-stressor. B) GRKO fish had lower blood glucose levels compared to the WT regardless of sampling time post-stressor exposure. All bars are means ±SEM (n=4-6); Fig. A time-points with different letters are significantly different for the WT and the GRKO fish, while *represent significant difference from the WT at each time-point; Fig. B legend inset shows significant main effect (two-way ANOVA/Holm-Sidak; p<0.05). Abbreviations: Wildtype (WT) and glucocorticoid receptor knockout (GRKO).

134

Figure 3: GRKO increased glucose clearance due to increased muscle uptake: Glucose levels from adult zebrafish (mixed sex) were taken at resting (non-injected) and 1 h post-injection with a fluorescent glucose analogue (2-NBDG). A) Injection of 2-NBDG in WT fish (blue hatched bar) caused an increase in blood glucose levels at 1h post-injection (vertical dotted line) compared to the saline-injected controls (solid blue) and the GRKO fish (red hatched bar). Resting levels (non-injected fish) are depicted before the vertical dotted line. B) Relative fluorescence units of 2-NBDG in the liver, and C) the white muscle. All bars are means ±SEM (n=4-6); bars with different letters are significantly different (one-way ANOVA (A); p<0.05); *significantly different from the WT (t-test; p<0.05). Abbreviations: Wildtype (WT) and glucocorticoid receptor knockout (GRKO).

135

Loss of GR promotes protein synthesis As GRKO zebrafish are larger with increased total protein, we next assessed the protein synthetic capacity of GRKO zebrafish compared to WT by measuring the phosphorylation of eIF4B, a marker for enhanced translation capacity. GRKO fish had 2-fold higher phosphorylation of eIF4B compared to the WT (p=0.026; Fig. 5A). The mutants also showed a higher expression of polyubiquitinated proteins compared to the WT (Fig. 5B). Targets of glucocorticoid-induced muscle catabolism were also measured to assess the capacity of the fish to induce muscle wasting. There were no differences between the transcript abundance of redd1 (Fig. 5C; p=0.992), mystnb (Fig. 5F; p=0.0559) and murf1b (Fig. 5D; p=0.321). The transcript abundance of murf1a was upregulated in the GRKO fish (Fig. 5E, p=0.05). Loss of GR attenuates protein catabolism To further test our hypothesis that glucose regulation is central to GR-mediated metabolic changes to adult growth, we next restricted glucose availability by fasting. We hypothesized that if increased glucose uptake was contributing to increased protein synthesis, the GRKO fish would spare proteins under a fasting stressor. As expected, the fasting stressor reduced blood glucose levels in both GR and WT fish compared to fed fish (Fig. 6A, p=0.017). Fasted fish had significantly less protein compared to the fed fish (p=0.006; Fig. 6B); however, GRKO fish showed significantly less protein breakdown compared to the WT, regardless of the nutritional status (p=0.002; Fig. 6B). The 7-d fasting stressor also reduced the body mass of WT fish by ~32% (p=0.012) compared to the fed WT fish (Fig. 6B). However, the fasting regimen had a less pronounced effect on the body mass of GRKO fish, as they experienced only a 19% reduction in mass compared to the fed GRKO fish (p=0.54; Fig. 6C). The molecular chaperone, heat shock protein cognate 70 (Hsc70) was also evaluated as a measure of global protein synthesis as these are important for the folding of nascent polypeptides (Beckmann et al., 1990). Independent of fed status, the Hsc70 protein expression was greater in the GRKO fish (Fig. 6D; p=0.048).

136

Figure 4: GRKO fish have higher hexokinase activity but no change in insulin expression: A) Hexokinase activity (μmol/min/g tissue) was higher in the GRKO fish compared to the WT. B) Insulin expression was measured by dot-blot using an anti-insulin antibody (1:500) in the whole-body homogenates of WT and GRKO fish either pre- (basal) or post- 2-NBDG injection. All bars are means ±SEM (n=4-6); *significantly different from the WT (t-test; p<0.05). Abbreviations: Wildtype (WT) and glucocorticoid receptor knockout (GRKO).

137

Figure 5: GRKO fish have greater protein synthetic capacity A) Translational capacity was measured in whole body WT and GRKO zebrafish (mixed sex) injected with 2- NBDG by measuring the phosphorylation of eIF4B, which was normalized to β-actin. A representative western blot of anti-phospho-eIF4B (80kDa; 1:1000) and β-actin (42 kDa; 1:1000) are shown below the bar graph. B) Ubiquitin expression was assessed by dot blot of whole-body homogenate from WT and GRKO zebrafish using an anti- ubiquitin antibody (1:100). Targets of glucocorticoid-induced muscle catabolism were also measured, including C) redd1, D) murf1a, E) murf1b and F) mystnb. All bars are means ±SEM (n=5-6); asterisks (*) denotes significance from the WT (t-test; p<0.05). Abbreviations: Wildtype (WT) and glucocorticoid receptor knockout (GRKO).

138

Muscle metabolomics

To further analyze the physiological significance due to the absence of a functional GR, we examined the muscle metabolome under fed and fasting conditions. Principle component analysis shows a clear separation of GRKO and WT muscle under fasting conditions, with little difference between the two groups in the fed state (Fig. 7A). This is further reflected in the heatmap, which reveals distinct metabolite profiles between the fasted GRKO and WT fish (Fig. 7B). The majority of changes observed were in the amino acid profiles. For instance, the abundance of key gluconeogenic amino acids, including L-alanine, L-methionine, and L- tyrosine, was lower in the GRKO fish muscle compared to the WT (Figs. 7C, 7E and 7G). Other amino acids, including L-arginine, L-proline, and L-threonine were more abundant in fed fish, and were not affected by the genotype (Fig. 7D, 7F and 7H). The ketogenic L-lysine (Fig. 7I) was higher in the fed WT muscle compared to all other groups (Fig. 7I).

139

Figure 6: GRKO fish have higher body mass and protein content post-fasting. Adult WT and GRKO zebrafish (mixed sex) were subjected to a 7-day fasting stressor. A) Fasting reduced blood glucose levels in both groups. B) GRKO fish had higher muscle protein content compared to the WT in both the fed and fasted state. C) GRKO fish, regardless of nutritional status, had higher body mass compared to the WT. D) Hsc70/Hsp70 protein levels were also higher in the GRKO fish, independent of feeding status. All bars are means ±SEM (n=6). Bars with different letters are significantly different; *significantly different from the WT (two-way ANOVA/Holm-Sidak, p<0.05). Abbreviations: Wildtype (WT) and glucocorticoid receptor knockout (GRKO).

140

Figure 7: Muscle Metabolomics: The muscle metabolite profile was analyzed from fed and fasted glucocorticoid receptor knockout (GR) and wildtype (WT) zebrafish. Fed WT (light blue), fed GR (dark blue), fasted WT (green) and fasted GR (red) muscle was analyzed using LC-MS. A) A principal component analysis and B) heat map of all measured metabolites. Box plots of metabolites including gluconeogenic amino acids C) L-alanine, D) L-arginine, E) L- methionine, F) L-proline, G) L-tyrosine, H) L-threonine and ketogenic amino acids I) L-lysine. All bars are means ±SEM (n=4-6). Bars with different letters are significantly different (one-way ANOVA/Holm-Sidak, p<0.05).

141

Table 2: Enzyme activities. Activities of enzymes involved in intermediary metabolism, including glycolysis (pyruvate kinase), amino acid catabolism (alanine and aspartate aminotransferases) and lipid biosynthesis (glucose-6-phosphate dehydrogenase (G6PDH), citrate lyase, malic enzyme). Activity is shown as μmol/min/g wet weight; values shown are means ±SEM (n=6). A marker of gluconeogenic activity (Phosphoenolpyruvate carboxykinase (PEPCK) was measured by qPCR and mRNA transcript abundance of this enzyme is listed under activity.

Enzyme wildtype GRKO Significance

Activity Activity (P<0.05)

Malic Enzyme 0.992 ± 0.14 0.110 ± 0.040

G6PDH 0.415 ± 0.132 0.606 ± 0.099

Citrate Lyase 4.3 ± 0.5 3.1 ± 0.4

Pyruvate kinase 30.0 ± 2.1 27.6 ± 2.1

Alanine Aminotransferase 2.1 ± 0.4 1.1 ± 0.1 0.0598

Aspartate Aminotransferase 13.0 ± 3.0 15.2 ± 1.0

PEPCK (mRNA) 1.0± 0.2 0.7± 0.2 0.441

142

Discussion Here using a loss of function approach, our results reveal a key role for GR signalling in the cortisol-mediated growth suppression in fish. The larger body mass, including higher muscle protein content in the GRKO zebrafish, even with fasting compared to the WT, indicates a key role for GR signalling in affecting growth during stress in fish. A conserved response to chronic glucocorticoids stimulation involves enhanced protein catabolism, leading to muscle atrophy and suppressed growth (McCormick et al., 1998; Mommsen et al., 1999; Kelly et al., 2012). Our results indicate that GR-mediated reduction in muscle glucose uptake may be a key driver in promoting the protein breakdown, leading to the reduced growth with chronic stress in fish. Zebrafish lacking GR are hypercortisolemic (11, 19, 49, 15; Fig. 2A) and have a larger body mass which indicates that a functional GR is essential for the energy repartitioning during chronic stress. A hallmark of chronic glucocorticoid stimulation is hyperglycemia due to enhanced liver gluconeogenesis. This process is dependent on amino acids, produced by muscle proteolysis, to provide C3 substrates for gluconeogenesis (see reviews 3, 12, 34). The necessity of GR signalling in mediating this response is supported in the current study where hypercortisolemia did not lead to hyperglycemia in the GRKO fish. Although hyperglycemia may be a combination of altered glucose production and clearance, few studies have examined the role of stress and/or cortisol in delaying glucose clearance in fish (Vijayan & Moon, 1994; Blasco et al., 1996; Moon, 2001; Eames et al., 2010). In mammals, the delay in post-absorptive rates of systemic glucose clearance is a result of insulin resistance in peripheral tissues, including skeletal muscle and adipocytes (Saltiel & Kahn, 2001). While peripheral tissues, such as the muscle, are involved in glucose clearance in both mammals and fish (Blasco et al., 1996; Polakof et al., 2012a), the effect of elevated glucocorticoids is unknown. Similar to mammals, insulin- mediated glucose regulation occurs in fish, and the muscle upregulates insulin-sensitive glucose transporters (Polakof et al., 2012b). Poor glucose utilization of peripheral tissues is considered a factor for glucose intolerance in fishes (Moon, 2001). This is evident in the present study, as wildtype fish have a >2-fold increase in plasma glucose levels following a stressor (Fig. 2B), which was maintained over the 2 h sampling period, supporting slower clearance of this metabolite (Moon, 2001). Interestingly, GRKO fish did not have a glucose response to stress. Given that the stressor-induced increase in glucose levels is usually associated with epinephrine-

143

mediated glycogenolysis (Fabbri et al., 1998; Mommsen et al., 1999), we hypothesized that the lack of hyperglycemia in the GRKO fish is due to a faster glucose clearance from the circulation as a result of target tissue uptake (Bernal-Sore et al., 2018). To test whether lack of GR signalling leads to enhanced glucose clearance, we followed the uptake of a glucose tracer (2-NBDG) into the muscle and liver of zebrafish (Fig. 3). Injection of the tracer increased blood glucose in the wildtype, but not the GRKO fish, again suggesting faster clearance in the absence of GR signalling. Also, the absence of an increase in blood glucose levels in the saline control group pre- and post-injection argues against injection as a stressor, which could elicit a glucose response (Fig. 3A). Consequently, the lack of glucose response in the GRKO fish injected with 2NBDG suggests a faster clearance of glucose in the absence of GR signalling. This was further supported by an enhanced 2-NBDG uptake by the muscle, but not the liver (Fig. 3B, 3C). To our knowledge, this is the first study to show that GR signalling modulates glucose uptake by the muscle in fish. Higher hexokinase activity in the GRKO fish indicates a higher tissue capacity for glucose utilization. The increased muscle glucose uptake capacity did not involve changes in whole body insulin expression in the GRKO fish (Fig. 4B). However, we cannot rule out the possibility that GR deficiency will enhance insulin sensitivity. Mammalian studies have clearly shown that corticosteroids inhibit the insulin- mediated muscle glucose uptake and this involves target tissue insulin resistance (Geer et al., 2014). The lack of an antibody for zebrafish GLUT4 precluded us from testing whether the higher recruitment of this protein on muscle membrane may be responsible for the increased muscle glucose uptake in the GRKO fish. However, there was no change in whole body glycogen content (Fig. 1I) or in the activities of glycolytic enzymes (Table 2). As these are metabolic targets for insulin-mediated muscle glucose regulation, our results suggest possible metabolic reorganization specific to GR signalling and independent of insulin responsiveness. The reduced food consumption in the GRKO fish, despite higher body mass, may be due to the enhanced brain capacity for glucose uptake, which suppresses appetite (Fletcher, 1984), and similar to that seen in the muscle, but this remains to be determined. Together, these results underscore a direct role for GR signalling in reducing muscle glucose uptake in fish, while the mechanism of action remains to be elucidated. We propose this lack glucose uptake by the

144

muscle during stress may be a driver in the physiological reorganization of the tissue, favouring reduced protein synthesis and increased protein catabolism. Glucose is a known regulator of protein synthesis in mammalian skeletal muscle, and along with insulin has the ability to increase the phosphorylation of eukaryotic initiation factor 4E (eIF4E) (Patel et al., 2001; Jeyapalan et al., 2007). This initiation factor is a key player in protein translation and is used as a marker of the protein synthetic capacity of the cell (41). Glucose, and its hexose derivatives, interact with eIF4E binding proteins (eIF4E-BP), phosphorylating them and causing them to dissociate from eIF4E (Patel et al., 2001). The eIF4E will then form a heterocomplex with other proteins, including eIF4B, promoting translation. Phosphorylation of eIF4B is also important to the aforementioned translation heterocomplex, and a loss of this protein reduces translation efficiencies (Sen et al., 2016). In our study, GRKO fish showed a marked increase in phosphorylation of eIF4B suggesting an increase in muscle protein synthetic capacity due to the loss of GR function. This is further reflected at the system level by the higher body mass and whole-body protein content in the GRKO fish (Figs. 1E and 6C). Also, the higher protein ubiquitination seen in the GRKO fish is usually associated with increased protein synthetic capacity, as nearly 30% of newly synthesized proteins are degraded in a proteasome-dependent manner (Wang et al., 2015). However, we cannot rule out the possibility that there is also concurrent protein breakdown that may be mediated by mineralocorticoid receptor (MR) signalling. This seems plausible given that murf1a, a key player in protein degradation (Kuo et al., 2013) was upregulated in the GRKO, in which MR is the only functioning corticosteroid receptor. Consequently, lack of GR signalling exerts a predominantly net anabolic effect by increasing the protein synthetic capacity and this may be associated with the enhanced capacity for muscle glucose uptake in the GRKO fish. As stress increases the energy demand (Barton & Schreck, 1987; Schreck & Tort, 2016), there is a greater demand for tissue-specific energy substrate re-partitioning to fuel the metabolic processes (Mommsen et al., 1999). Indeed, a key role for glucocorticoid stimulation is enhanced muscle proteolysis, releasing amino acids for use as substrates for hepatic gluconeogenesis (Milligan, 1997; Mommsen et al., 1999). Stressor-mediated increases in glucose levels are necessary to fuel the aerobic tissues, including brain and gills, to cope with the increased energy demand. For instance, in brown trout (Salmo trutta) the uptake rates of glucose were highest in

145

the brain, kidney, spleen and gills, and lowest in red muscle, heart, and white muscle (Blasco et al., 1996). As muscle constitutes >50% of adult fish weight (Mommsen, 2001), even small changes in glucose utilization by this tissue may limit the availability for aerobic tissues, which are critical for coping with stress. Indeed, despite a low uptake rate, muscle in brown trout accumulated the most glucose because of the larger mass of this tissue (Blasco et al., 1996). Therefore, limiting muscle glucose uptake to facilitate re-allocation of this fuel to other tissues may be an evolutionarily conserved role of GR signalling in stress adaptation. While it is well known that key protein catabolic genes are under the transcriptional control of GR in mammals, this has yet to be shown in fish. While there is some evidence that myostatin does not have the same role in fish as it does in mammals (Galt et al., 2014), genes that are known to be essential to protein catabolism, including redd1 remain unchanged in the GRKO, despite the high cortisol levels. This further indicates that the cortisol-driven muscle wasting is abolished in the GRKO fish. Interestingly, the murf1a paralog was upregulated in the GRKO fish, which indicates that it may be under the transcriptional control of another transcription factor sensitive to the increased cortisol levels, such as the MR. To test whether GR signalling has a direct role in weight loss, we measured body mass and protein content in 7-day fasted zebrafish. Our prediction was that the body mass and protein content loss will be lower in the absence of GR signalling. Indeed, GRKO fish had larger body mass (Fig. 6B) and higher muscle protein content (Fig. 6D) relative to the WT during fasting. Also, the protein expression of Hsc70 was higher in the GRKO fish regardless of the nutritional state of the animal suggesting that the overall protein synthetic capacity was higher in the absence of GR in fish (Thulasiraman et al., 1999; Kim et al., 2008). Importantly, Hsc70 also interacts with CHIP (C-terminal of Hsp70-interacting protein), which is thought to coordinate the cellular balance between folding and degradation of proteins (Kim et al., 2008). While the physiological role for GR signalling may be primarily to enhance protein catabolism, a reduction in protein synthesis may also be an important part of the glucocorticoid-driven muscle atrophy. Mice with a conditional muscle GR knockout (GRmKO) also showed increased muscle mass and reduced muscle loss post-fasting (Shimizu et al., 2015), confirming a conserved GR-dependent phenotype. Cortisol-induced muscle wasting is a well-known effect of Cushing’s syndrome (Geer et al., 2014), and the maintenance of protein levels in the GRKO fish, despite

146

hypercortisolemia, suggests a key role for GR signalling in protein sparing. This was further evident from the overall change in the metabolome of the fasted GRKO and wildtype zebrafish (Fig. 7A and 7B), including lower alanine, methionine and tyrosine in the GRKO fish (Fig. 7C, 7E and 7G), suggesting a reduced capacity for proteolysis (Garvey et al., 2014). A limitation of this study was the lack of antibodies that were specific to fish proteins. For instance, mammalian antibodies for insulin receptor B and GLUT4 failed to cross-react with our zebrafish samples. Also, an ortholog of glut4 in zebrafish has not yet been identified, which limited our ability to establish whether glucose uptake was due to differences in transporter availability using transcript abundance (Maddison et al., 2015). We were also limited by small size of the animal, which limited the number of measurements we could make on a single tissue sample. For example, it was not possible to get enough blood to measure circulating glucose and insulin levels. Despite these limitations, the single copy of GR (compared to two paralogs in all other teleost fish) makes the zebrafish an ideal translational model to examine the conserved nature of glucocorticoid-driven glucose regulation and nutrient homeostasis. While our study has shed light on the whole-body metabolic changes evident in the absence of GR, we cannot differentiate the metabolic actions that are GR-isoform specific. For instance we know that GRα and GRβ have unique functions in animals (Kino et al., 2009), and the cross-talk between these two isoforms may also modulate target tissue metabolism (Chatzopoulou et al., 2015). While our CRISPR-Cas9 method deleted both the alpha and beta splice variant, generating knockout of the splice variant of GR in the future may help to elucidate the conserved functions of this receptor in the metabolic adaptation to stress in animals. In conclusion, we report that cortisol-GR signalling plays a key role in body mass regulation during stress in fish. GR knockout enhances muscle glucose uptake and protein content, leading to increased body mass in zebrafish. This study emphasizes the effect associated with a lack of functional GR on intermediary metabolism at the system level, and provides the physiological consequences to a global antagonism of GR. Our results indicate that GR plays a key physiological role in the regulation of muscle glucose availability during stress, and this has a potent effect on muscle protein homeostasis. Our findings have important implications in aquaculture, where enhancing growth rate with better feed-utilization is paramount, and manipulation of GR abundance may be one possible way to achieve this. Also, GR manipulation

147

may be an approach to increase target tissue glucose sensitivity (Bernal-Sore et al., 2018) in humans suffering from type 2 diabetes (Wang et al., 2006; Bernal-Sore et al., 2018), and our model has the potential to shed light on the physiological consequences of this treatment at the system level.

Acknowledgements

This work was supported by a NSERC discovery grant to MMV. EF was the recipient of a NSERC doctoral scholarship. We would like to thank Dr. Ramen Sood and Blake Carrington at NIH for assistance with establishing our CRISPR/Cas9 strains, and Dr. Ian Lewis and Ryan Groves at the University of Calgary metabolomics facility for assistance with metabolome analysis.

148

CHAPTER 7: GENERAL CONCLUSION

149

The overarching hypothesis of this thesis was that both MR and GR activation are necessary for mediating the physiological effects of cortisol in zebrafish. This was an extension of de Kloet and Reul’s hypothesis nearly 30 years ago, that MR will mediate basal neuronal excitability and GR will mediate stress effects (de Kloet & Reul, 1987). Indeed, cortisol binds to MR with 10x higher affinity than a mineralocorticoid hormone such as aldosterone (Baker et al., 2013), and at basal cortisol concentrations nearly 80% of the MR are bound to cortisol (de Kloet & Reul, 1987). This is in contrast to no cortisol bound by the type II corticosteroid receptor, GR, under basal hormone levels. This makes cortisol the likely primary ligand of MR in fish, and this was tested here using the CR knockout zebrafish (Chapters 3, 4 and 5). Using these fish in combination with cortisol treatment, we differentially activated GR and MR to determine the physiological consequences of CRs activation in a ray-finned fish. The zebrafish is a particularly nice model in which to study the activation of GR and MR. Firstly, ubiquitous knockouts of GR and MR in mammals are lethal due to delayed lung maturation (Cole et al., 1995) and renal failure (Berger et al., 1998), respectively. Conditional knockouts are possible, but this may result in compensation by other tissues to account for the loss of these receptors elsewhere (Shimizu et al., 2015), and may not be representative of the systems-level changes that occur when the receptor is absent. Ubiquitous knockouts of GR or MR are not lethal in zebrafish (Faught & Vijayan, 2018a/Chapter 3), allowing the use of this species for a systems-level approach at characterizing the role of CRs on metabolic adaptation to stress. Also, zebrafish is an excellent translational model and is being used extensively for metabolic research (Phillips & Westerfield, 2014). The results from this thesis, as well as the CR knockout developed here, have the potential to contribute a significant advance to our understanding of the functional roles of CRs in mediating stress/glucocorticoid-mediated metabolic diseases. The results of MR on metabolism were unexpected because for the past 30 years, the focus on MR activation was on either the mineralocorticoid function in fluid balance or on its central roles in neuronal excitability (de Kloet & Reul, 1987; Berger et al., 2006; Rozeboom et al., 2007; Joëls & de Kloet, 2017). Only a handful of studies in recent years have indicated that MR may have a metabolic role in peripheral tissues to regulate adipose tissue differentiation (Marzolla et al., 2012), but these studies were on adult tissues or cells in vitro. On the other hand, GR action is very well characterized in both mammals and fish, and it has a known

150

metabolic role, particularly during stress to increase energy availability (Mommsen et al., 1999; Charmandari et al., 2005; Vijayan et al., 2010). This thesis, for the first time, demonstrates a functional role of MR in response to basal and stressed levels of cortisol, particularly from the standpoint of energy repartitioning and growth. Despite the cloning and sequencing of MR in various ray-finned fishes there is no known physiological role for this receptor (Baker & Katsu, 2017), outside of the work presented here. From an evolutionary perspective this is an important receptor as it is found in every phylogenetic clade, and even the primitive vertebrates, the agnathans, have a primitive form of MR, the corticosteroid receptor (CR) (Baker et al., 2013). In contrast, MR’s role in tetrapods is very well characterized, and it has a ligand that is distinct for mineralocorticoid function. The primary ligand for MR in tetrapods is aldosterone, and this ligand binding to MR in Na+ transporting epithelia (kidney, colon) mediates the ion and fluid balance (Cole & Young, 2017). The extra-mineralocorticoid function of MR in nonepithelial cells became apparent when tritiated corticosterone could cross the blood-brain barrier, be retained in the hippocampus, and activate separate receptors (Coirini et al., 1983). Since then the focus of glucocorticoid-MR signalling has been in the modulation of neuronal function and behaviour (Funder, 2005). It is generally accepted that the balance of GR and MR in the brain is important in mediating the central roles of MR. However, recent work has admitted that the extent of GR and MR interaction has been underestimated (Mifsud & Reul, 2016, 2018), and indeed, this thesis confirms that GR activation can be modulated by MR activation with respect to metabolic function (Chapter 4, 5). In a similar way to the mammalian brain, MR appears to promote anabolic function, while GR stimulates catabolic processes during stress using the systems level approach with zebrafish. The balance between anabolic and catabolic processes is a central tenet to growth and metabolism (Fig 1). This thesis shows the metabolic consequence of GR activation under high cortisol levels. This has been characterized before, but the ability of MR to modulate the physiological response of GR was a novel finding. This suggests that not only is MR activation important in mediating basal cortisol function, but that when cortisol levels are high, MR also has an important functional role. Based on the results of chapters 4 and 5, it appears that MR is important in anabolic processes, both in terms of protein and lipid deposition. Interestingly,

151

while MR does promote anabolic outcomes (increased TG, cholesterol, protein), the research presented here indicates that this may be by limiting the catabolic processes. From the perspective of TG and cholesterol biosynthesis, MR acts to increase their accumulation in postnatal larvae (Chapter 4). The regulation of weight and protein is slightly more complex, but it is clear that MR only has a role in mediating anabolic processes (gh/igf gene transcription and insulin expression), while GR is responsible for the upregulation of proteolytic genes (redd1/murf1a/murf1b; Fig 1). It is unclear how GR and MR are able to interact, but there are several mechanisms suggested in the literature (Mifsud & Reul, 2018), and an additional level of control was also seen from the studies presented in this thesis. The most obvious and well-defined interactive model is heterodimerization (Mifsud & Reul, 2016, 2018). In Chapter 3, both GR and MR are required for stress-induced behaviour changes, and this was suggestive of heterodimerization. Indeed, of the few targets identified for GR/MR heterodimerization, one, the 5HT1A receptor, is known to modulate behaviour (Ou et al., 2001). Further work needs to be carried out using chromatin immunoprecipitation (ChiP) assays to further characterize the DNA binding dynamics of these two receptors. However, when it comes to regulation of lipoprotein lipase (Chapter 4), it was clear that MR activation has an inhibitory effect on GR action. When MR is removed, the ability of GR to increase the transcription of lpl increased 14-fold. The suppression of GR transcriptional ability by MR may be due to the interaction between GR and MR at the hormone response elements in the promoter region or due to the tethering of the receptors, thereby inhibiting GR function (Mifsud & Reul, 2018; Rivers et al., 2019). The transcriptomics also gave us an additional insight into MR and GR regulation; in particular, enzymes in the cholesterol biosynthetic pathway were modulated separately and in tandem by GR and MR activation (MR, hmgcs1 and GR, hmgcr). This suggests that when cortisol levels are high, GR activity will be the dominant signalling mechanism, as hmgcr is the rate-limiting step in the cholesterol biosynthesis pathway (Fraher et al., 2016). Indeed, in a situation where both GR and MR were activated, there was no increase in cholesterol accumulation, whereas when just MR was active there was a marked increase in cholesterol levels. While we cannot discount that GR may also be signalling at basal cortisol levels, our gene readout (11βhsd2) does not support this contention (Chapter 4).

152

The regulation of protein and weight was less straight-forward, but this work also brought forth an interesting concept, and an additional layer of complexity in CR regulation (Chapter 5). Here it was clear that MR, though gene-specific, is able to compensate for protein loss by upregulating anabolic processes. In the case of igf1 and pomca, cortisol-induced increases in mRNA levels were dependent on both GR and MR, but the mechanisms of interaction need to be further explored. In the case of insulin, there was only an upregulation when cortisol levels were high, which suggests that there can be differential regulation of MR, dependent on cortisol levels. It has been recently shown that MR is able to bind to the GREs when cortisol levels are high, indicating that MR activation, and the associated transcriptional regulation, are sensitive to changes in cortisol levels (Mifsud & Reul, 2016). This was further explored in the last chapter, which contrasted the differences in WT and GRKO fish, as they both have MR activation, but under different cortisol conditions. With the removal of GRKO, and with only MR active, these fish increase in weight, protein, lipid, and have fundamental shifts in glucose uptake by peripheral tissues, and protein translation (Chapter 6). The specific interactions of the GR and MR with the DNA regions were outside the scope of this thesis, but will be important in understanding the specific mechanisms of action. One drawback of this thesis was our inability to develop a double GR and MR knockout. While this is still a work in progress, stimulation of fish lacking both genomic receptors may be instrumental in determining the elusive “non- genomic” receptor-mediated response to cortisol (Das et al., 2018). The differences in affinity between GR and MR for cortisol, suggests an interesting avenue for future study. Specifically, the function of 11βHSD2 as a molecular switch to regulate the activation of GR and MR (Funder et al., 1988; Tomlinson & Stewart, 2005; Kiilerich et al., 2007; Alderman & Vijayan, 2012). While Chapter 3 suggests that this dehydrogenase is important for maintaining basal cortisol levels, what does this mean from a functional standpoint? The hypothesis that we would like to suggest is that 11βHSD2 activation may promote MR signalling by lowering cortisol levels below the GR activation threshold. This suggests that 11βHSD2 itself may act as a molecular switch between anabolic and catabolic function in fish. 11βHSD2 has a similar function in mammals, in which this enzyme will lower cortisol levels to the inactive cortisone in tissues, thereby allowing aldosterone signalling to take place (Chapman et al., 2013). This may also be the case in fish; however, 11βHSD2-mediated

153

lowering of cortisol levels in certain tissues may facilitate MR signalling. This enzyme has been shown to modulate cortisol deposition into the oocytes of zebrafish, indicating that activation of this enzyme modulates cortisol levels with physiological consequences (Faught & Vijayan, 2018b). The development of other knockout strains, including 11βHSD2 (-/-) may help answer this question. Central to the confusion surrounding the role of MR in ray-finned fishes was the lack of an endogenous mineralocorticoid hormone, such as aldosterone (McCormick et al., 2008; Cruz et al., 2013). There have been several studies which have tried to identify the ligand in ray- finned fish for MR (Sturm et al., 2005; McCormick et al., 2008), but MR is a promiscuous receptor and can bind to several other 3-ketosteroids, including 11-deoxycorticosterone (11- DOC), 11-deoxycortisol, progesterone, and corticosterone (Baker & Katsu, 2017; Katsu & Baker, 2018). While several groups have suggested that there is a distinct mineralocorticoid in fish, such as 11-DOC (Sturm et al., 2005; McCormick et al., 2008), this comes with a caveat – MR does not have mineralocorticoid function in fish (Cruz et al., 2013). In fact, GR and not MR is thought to modulate ionocyte and ion balance in zebrafish (Cruz et al., 2013). Activating MR using different ligands may be an interesting avenue of study to understand whether differentially activating MR has a functional relevance. Other concepts that were not covered in this thesis but are worthy of future study from the perspective of CR activation, is the role of GR activation in modulating the circadian rhythm (Joëls & de Kloet, 2017). Cortisol levels experience diurnal fluctuations throughout the day, and in mammals, it has been hypothesized that the concentration of cortisol at the peak of the diurnal fluctuations is enough to activate GR. This has yet to be proven in fish, and all the studies were carried out when the cortisol concentrations are at the lowest (11 am), as to avoid this compounding factor (data not shown). However, understanding the transient activation of GR may inform how both GR and MR can act together to maintain energy homeostasis, under basal conditions (Dickmeis & Foulkes, 2011). Beyond the fundamental aspects of this thesis, the implications of modulating CR signalling also have several broader implications. From an aquaculture perspective, we have designed an ideal fish. The GRKO fish will eat less, but weigh more due to increases in fat and protein. Further work needs to be done to assess whether this energy redistribution will

154

compromise other physiological systems, including immune responses. Indeed, due to the ubiquitous expression of both GR and MR in fish, this thesis only represents the one aspect of CR regulation. Recent work in mammals has shown distinct roles in heart development and cardiac function depending on whether GR or MR was activated (Oakley et al., 2019). Immune function is a known modulator by both GR and MR, but the extent to which these receptors may impact the immune response in ray-finned fishes remains to be determined. The inactivation of GR may also hold interest from a biomedical perspective (shown in chapter 4), as loss of this receptor improved muscle glucose utilization and promoted anabolic growth. We know this is a conserved effect, as mammalian skeletal muscle cell lines also show a similar enhanced glucose uptake and protein synthesis when treated with the GR antagonist mifepristone (Bernal-Sore et al., 2018). Overall, this thesis presents significant advances in the field of corticosteroid signalling. In particular, we have identified for the first time a physiological role for MR in ray-finned fishes. Using the unique systems-level approach afforded by using zebrafish, several key targets for MR have been identified, and this will also provide insights into the extra-mineralocorticoid functions of MR in mammals. The loss of GR at the systems level has also revealed a fundamental role that these two receptors play in maintaining energy homeostasis. To that end, the take-home message of this thesis is that MR is a key regulator of energy homeostasis under both basal and stress conditions in lower vertebrates.

155

Figure 1: Representative paradigm of the physiological role of GR and MR activation Under basal cortisol conditions MR will be active and this leads to increased energy substrate accumulation, such as protein, triglycerides and cholesterol. As the cortisol concentration increases, MR can differentially regulate nutrient stores, including protein accumulation, as well as insulin expression. At high levels of cortisol, both GR and MR will be activated, and this is necessary to mediate the stress-related behaviour, and gene-specific regulation, including igf1 and pomca. When the animal is stressed and GR is activated this initiates a cascade of catabolic processes, which results in increased lipolysis, proteolysis, reduced growth and altered glucose uptake in the muscle. Activation of 11βHSD2 may be an important switch that modulates MR and GR activation, but this remains to be tested.

156

WORKS CITED:

Ablain J, Durand EMM, Yang S, Zhou Y & Zon LII (2015). A CRISPR/Cas9 Vector System for Tissue-Specific Gene Disruption in Zebrafish. Dev Cell 32, 756–764. Acerete L, Balasch JC, Castellana B, Redruello B, Roher N, Canario AV, Planas JV., MacKenzie S & Tort L (2007). Cloning of the glucocorticoid receptor (GR) in gilthead seabream (Sparus aurata). Differential expression of GR and immune genes in gilthead seabream after an immune challenge. Comp Biochem Physiol - B Biochem Mol Biol 148, 32–43. Agulleiro MJ, Roy S, Sanchez E, Puchol S, Gallo-Payet N & Cerda-Reverter JM (2010). Role of melanocortin receptor accessory proteins in the function of zebrafish melanocortin receptor type 2. Mol Cell Endocrinol 320, 145–152. Alderman SL & Bernier NJ (2009). Ontogeny of the corticotropin-releasing factor system in zebrafish. Gen Comp Endocrinol 164, 61–69. Alderman SL, McGuire A, Bernier NJ & Vijayan MM (2012). Central and peripheral glucocorticoid receptors are involved in the plasma cortisol response to an acute stressor in rainbow trout. Gen Comp Endocrinol 176, 79–85. Alderman SL & Vijayan MM (2012). 11β-hydroxysteroid dehydrogenase type 2 in zebrafish brain: A functional role in hypothalamus-pituitary-interrenal axis regulation. J Endocrinol 215, 393–402. Alsop D & Vijayan MM (2008). Development of the corticosteroid stress axis and receptor expression in zebrafish. Am J Physiol Regul Integr Comp Physiol 294, R711-9. Alsop D & Vijayan MM (2009a). Molecular programming of the corticosteroid stress axis during zebra fish development. Comp Biochem Physiol Part A 153, 49–54. Alsop D & Vijayan MM (2009b). The zebrafish stress axis: Molecular fallout from the teleost- specific genome duplication event. Gen Comp Endocrinol 161, 62–66. Aluru N & Vijayan MM (2009). Stress transcriptomics in fish: a role for genomic cortisol signaling. Gen Comp Endocrinol 164, 142–150. Aluru N, Vijayan MM, Aluru N & Vijayan MM (2007). Hepatic transcriptome response to glucocorticoid receptor activation in rainbow trout. Physiol Genomics 31, 483–491. Armani A, Marzolla V, Rosano G & Caprio M (2014). Mineralocorticoid vs glucocorticoid

157

receptors: solo players or team mates in the control of adipogenesis? Int J Obes 38, 1580– 1581. Arterbery AS, Deitcher DL & Bass AH (2010). Corticosteroid receptor expression in a teleost fish that displays alternative male reproductive tactics. Gen Comp Endocrinol 165, 83–90. Aruna A., Nagarajan G & Chang CF (2012). Involvement of Corticotrophin-Releasing Hormone and Corticosteroid Receptors in the Brain-Pituitary-Gill of Tilapia During the Course of Seawater Acclimation. J Neuroendocrinol 24, 818–830. Auer TO, Duroure K, Cian A De, Concordet J & Bene F Del (2014). Highly efficient CRISPR / Cas9-mediated knock-in in zebrafish by homology-independent DNA repair. Genome Res 24, 142–153. Backström T & Winberg S (2013). Central corticotropin releasing factor and social stress. Front Neurosci 7, 1–10. Baker ME, Funder JW & Kattoula SR (2013). Evolution of hormone selectivity in glucocorticoid and mineralocorticoid receptors. J Steroid Biochem Mol Biol 137, 57–70. Baker ME & Katsu Y (2017). Evolution of the mineralocorticoid receptor: Sequence, structure and function. J Endocrinol 234, T1–T16. Barton BA & Schreck CB (1987). The Metabolic Cost of Acute Physical Stress in Juvenile Steelhead. Trans Am Fish Soc 116, 257–263. Beckmann RP, Mizzen LA & Welch WJ (1990). Interaction of Hsp 70 with newly synthesized proteins: implications for protein folding and assembly. Science 248, 850–854. Benato F, Colletti E, Skobo T, Moro E, Colombo L, Argenton F & Dalla Valle L (2014). A living biosensor model to dynamically trace glucocorticoid transcriptional activity during development and adult life in zebrafish. Mol Cell Endocrinol 392, 60–72. Berger S, Bleich M, Schmid W, Cole TJ, Peters J, Watanabe H, Kriz W, Warth R, Greger R, Schtz G & Evans RM (1998). Mineralocorticoid receptor knockout mice: Pathophysiology of Na+ metabolism. Genetics 95, 9424–9429. Berger S, Wolfer DP, Selbach O, Alter H, Erdmann G, Reichardt HM, Chepkova AN, Welzl H, Haas HL, Lipp H-P & Schutz G (2006). Loss of the limbic mineralocorticoid receptor impairs behavioral plasticity. Proc Natl Acad Sci 103, 195–200. Bernal-Sore I, Navarro-Marquez M, Osorio-Fuentealba C, Díaz-Castro F, del Campo A, Donoso-

158

Barraza C, Porras O, Lavandero S & Troncoso R (2018). Mifepristone enhances insulin- stimulated Akt phosphorylation and glucose uptake in skeletal muscle cells. Mol Cell Endocrinol 461, 277–283. Bernier NJ (2006). The corticotropin-releasing factor system as a mediator of the appetite- suppressing effects of stress in fish. Gen Comp Endocrinol 146, 45–55. Bernier NJ, Bedard N & Peter RE (2004). Effects of cortisol on food intake, growth, and forebrain neuropeptide Y and corticotropin-releasing factor gene expression in goldfish. Gen Comp Endocrinol 135, 230–240. Bernier NJ & Peter RE (2001). The hypothalamic-pituitary-interrenal axis and the control of food intake in teleost fish. Comp Biochem Physiol - B Biochem Mol Biol 129, 639–644. Best C, Kurrasch DM & Vijayan MM (2017). Maternal cortisol stimulates neurogenesis and affects larval behaviour in zebrafish. Sci Rep 7, 40905. Best C & Vijayan MM (2018). Cortisol elevation post-hatch affects behavioural performance in zebrafish larvae. Gen Comp Endocrinol 257, 220–226. Bhattacharyya S, Brown DE, Brewer JA, Vogt SK & Muglia LJ (2007). Macrophage glucocorticoid receptors regulate Toll-like receptor 4-mediated inflammatory responses by selective inhibition of p38 MAP kinase. Blood 109, 4313–4319. Bienvenu LA, Morgan J, Rickard AJ, Tesch GH, Cranston GA, Fletcher EK, Delbridge LMD & Young MJ (2012). Macrophage mineralocorticoid receptor signaling plays a key role in aldosterone-independent cardiac fibrosis. Endocrinology 153, 3416–3425. Biga PR, Cain KD, Hardy RW, Schelling GT, Overturf K, Roberts SB, Goetz FW & Ott TL (2004). Growth hormone differentially regulates muscle myostatin1 and -2 and increases circulating cortisol in rainbow trout (Oncorhynchus mykiss). Gen Comp Endocrinol 138, 32–41. Bigas J, Sevilla LM, Carceller E, Boix J & Pérez P (2018). Epidermal glucocorticoid and mineralocorticoid receptors act cooperatively to regulate epidermal development and counteract skin inflammation. Cell Death Dis 9, 588. Blasco J, Fernàndez-Borràs J, Marimon I & Requena A (1996). Plasma glucose kinetics and tissue uptake in brown trout in vivo: effect of an intravascular glucose load. J Comp Physiol B 165, 534–541.

159

Bose SK, Hutson I & Harris CA (2016). Hepatic glucocorticoid receptor plays a greater role than adipose gr in metabolic syndrome despite renal compensation. Endocrinology 157, 4943– 4960. Boyle MP, Brewer JA, Funatsu M, Wozniak DF, Tsien JF, Izumi Y & Muglia LJ (2005). Acquired deficit of forebrain glucocorticoid receptor produces depression-like changes in adrenal axis regulation and behavior. Proc Natl Acad Sci 102, 473–478. Britto FA, Begue G, Rossano B, Docquier A, Vernus B, Sar C, Ferry A, Bonnieu A, Ollendorff V & Favier FB (2014). REDD1 deletion prevents dexamethasone-induced skeletal muscle atrophy. AJP Endocrinol Metab 307, E983–E993. Britto FA, Cortade F, Belloum Y, Blaquière M, Gallot YS, Docquier A, Pagano AF, Jublanc E, Bendridi N, Koechlin-Ramonatxo C, Chabi B, Francaux M, Casas F, Freyssenet D, Rieusset J, Giorgetti-Peraldi S, Carnac G, Ollendorff V & Favier FB (2018). Glucocorticoid- dependent REDD1 expression reduces muscle metabolism to enable adaptation under energetic stress. BMC Biol 16, 1–17. Bumaschny VF, de Souza FSJ, López Leal RA, Santangelo AM, Baetscher M, Levi DH, Low MJ & Rubinstein M (2007). Transcriptional Regulation of Pituitary POMC Is Conserved at the Vertebrate Extremes Despite Great Promoter Sequence Divergence. Mol Endocrinol 21, 2738–2749. Burnstein KL, Bellingham DL, Jewell CM, Powell-Oliver FE & Cidlowski JA (1991). Autoregulation of glucocorticoid receptor gene expression. Steroids 56, 52–58. Bury N (2003). Evidence for two distinct functional glucocorticoid receptors in teleost fish. J Mol Endocrinol 31, 141–156. Canonica J, Sergi C, Maillard M, Klusonova P, Odermatt A, Koesters R, Loffing-Cueni D, Loffing J, Rossier B, Frateschi S & Hummler E (2016). Adult nephron-specific MR- deficient mice develop a severe renal PHA-1 phenotype. Pflugers Arch Eur J Physiol 468, 895–908. Caprio M, Antelmi A, Muscat A, Mammi C, Marzolla V, Fabbri A & Zennaro M (2011). Antiadipogenic Effects of the Mineralocorticoid Receptor Antagonist Drospirenone: Potential. Endocrinology 152, 113–125. Caprio M & Fe B (2007). Pivotal role of the mineralocorticoid receptor in corticosteroid-induced

160

adipogenesis. FASEB J 21, 2185–2194. Carney TJ & Mosimann C (2018). Switch and Trace: Recombinase Genetics in Zebrafish. Trends Genet 34, 362–378. Carrington B, Varshney GK, Burgess SM & Sood R (2015). CRISPR-STAT: an easy and reliable PCR-based method to evaluate target-specific sgRNA activity. Nucleic Acids Res 43, 1–8. Cerdá-Reverter JM, Agulleiro MJ, Cortés R, Sánchez E, Guillot R, Leal E, Fernández-Durán B, Puchol S & Eley M (2012). Involvement of melanocortin receptor accessory proteins (MRAPs) in the function of melanocortin receptors. Gen Comp Endocrinol 188, 133–136. Chadwick JA, Hauck JS, Lowe J, Shaw JJ, Guttridge DC, Gomez-Sanchez CE, Gomez-Sanchez EP & Rafael-Fortney JA (2015). Mineralocorticoid receptors are present in skeletal muscle and represent a potential therapeutic target. FASEB J 29, 4544–4554. Chapman K, Holmes M & Seckl J (2013). 11β-Hydroxysteroid Dehydrogenases: Intracellular Gate-Keepers of Tissue Glucocorticoid Action. Physiol Rev 93, 1139–1206. Charmandari E, Tsigos C & Chrousos G (2005). Endocrinology of the stress response. Annu Rev Physiol 67, 259–284. Chatzopoulou A, Roy U, Meijer AH, Alia A, Spaink HP & Schaaf MJM (2015). Transcriptional and metabolic effects of glucocorticoid receptor α and β signaling in zebrafish. Endocrinology1–13. Chatzopoulou A, Schoonheim PJ, Torraca V, Meijer AH, Spaink HP & Schaaf MJM (2017). Functional analysis reveals no transcriptional role for the glucocorticoid receptor β-isoform in zebrafish. Mol Cell Endocrinol. 447, 61-70 Chmielarz P, Kus̈ mierczyk J, Parlato R, Schütz G, Nalepa I & Kreiner G (2013). Inactivation of glucocorticoid receptor in noradrenergic system influences anxiety- and depressive-like behavior in mice. PLoS One 8, 1–6. Cianfarani S, Geremia C, Scott C & Germani D (2002). Growth, IGF system, and cortisol in children with intrauterine growth retardation: Is catch-up growth affected by reprogramming of the hypothalamic-pituitary-adrenal axis? Pediatr Res 51, 94–99. Coirini H, Marusic E, De Nicola A, TC R & McEwen B (1983). Identification of Mineralocorticoid Binding Sites in Rat Brain by Competition Studies and Density Gradient

161

Centrifugation. Neuroendocrinology 37, 354–360. Cole TJ, Blendy JA, Monaghan AP, Krieglstein K, Schmid W, Aguzzi A, Fantuzzi G, Hummler E, Unsicker K & Schiitz G (1995). Targeted disruption of the glucocorticoid receptor gene blocks adrenergic chromaffin cell development and severely retards lung maturation. Genes Dev 9, 1608–1621. Cole TJ, Myles K, Purton JF, Brereton PS, Solomon NM, Godfrey DI & Funder JW (2001). GRKO mice express an aberrant dexamethasone-binding glucocorticoid receptor, but are profoundly glucocorticoid resistant. Mol Cell Endocrinol 173, 193–202. Cole TJ & Young MJ (2017). Mineralocorticoid receptor null mice: Informing cell-type-specific roles. J Endocrinol 234, T83–T92. Cooray SN & Clark AJL (2011). Melanocortin receptors and their accessory proteins. Mol Cell Endocrinol 331, 215–221. Cruz SA, Lin C-H, Chao P-L & Hwang P-P (2013). Glucocorticoid receptor, but not mineralocorticoid receptor, mediates cortisol regulation of epidermal ionocyte development and ion transport in zebrafish (Danio rerio). PLoS One 8, e77997. Das C, Thraya M & Vijayan MM (2018). Nongenomic cortisol signaling in fish. Gen Comp Endocrinol 265, 121–127. DiBattista JD, Levesque HM, Moon TW & Gilmour KM (2006). Growth Depression in Socially Subordinate Rainbow Trout Oncorhynchus mykiss: More than a Fasting Effect. Physiol Biochem Zool 79, 675–687. Dickmeis T & Foulkes NS (2011). Glucocorticoids and circadian clock control of cell proliferation: At the interface between three dynamic systems. Mol Cell Endocrinol 331, 11–22. Dickmeis T, Lahiri K, Nica G, Vallone D, Santoriello C, Neumann CJ, Hammerschmidt M & Foulkes NS (2007). Glucocorticoids play a key role in circadian cell cycle rhythms. PLoS Biol 5, e78. Dindia L, Faught E, Leonenko Z, Thomas R, Vijayan MM, Mathilakath M, Dindia L, Faught E, Leonenko Z, Thomas R & Vijayan MM (2013). Rapid cortisol signaling in response to acute stress involves changes in plasma membrane order in rainbow trout liver. Am J Physiol Endocrinol Metab 304, E1157-66.

162

Dindia L, Murray J, Faught E, Davis TL, Leonenko Z & Vijayan MM (2012). Novel nongenomic signaling by glucocorticoid may involve changes to liver membrane order in rainbow trout. PLoS One 7, e46859. Djurhuus CB, Gravholt CH, Nielsen S, Mengel A, Christiansen JS, Schmitz OE & Møller N (2002). Effects of cortisol on lipolysis and regional interstitial glycerol levels in humans. Am J Physiol - Endocrinol Metab 283, E172–E177. Dores RM, Londraville RL, Prokop J, Davis P, Dewey N & Lesinski N (2014). Molecular Evolution of GPCRs: Melanocortin/melanocortin receptors. J Mol Endocrinol 52, T29–T42. Dostert a & Heinzel T (2004). Negative Glucocorticoid Receptor Response Elements and their Role in Glucocorticoid Action. Curr Pharm Des 10, 1–10. Doyon C, Leclair J, Trudeau VL & Moon TW (2006). Corticotropin-releasing factor and neuropeptide Y mRNA levels are modified by glucocorticoids in rainbow trout , Oncorhynchus mykiss. Gen Comp Endocrinol 146, 126–135. Ducouret B, Tujague M, Ashraf J, Mouchel N, Servel N, Valotaire Y & Thompson EB (1995). Cloning of a teleost fish glucocorticoid receptor shows that it contains a deoxyribonucleic acid-binding domain different from that of mammals. Endocrinology 136, 3774–3783. Eames SC, Philipson LH, Prince VE & Kinkel MD (2010). Blood Sugar Measurement in Zebrafish Reveals Dynamics of Glucose Homeostasis. Zebrafish 7, 205–213. Eaton S, Bartlett KB & Pourfarzam M (1996). Mammalian mitochondrial β -oxidation. Biochem J 320, 345–357. Esbaugh AJ & Walsh PJ (2009). Identification of two glucocorticoid response elements in the promoter region of the ubiquitous isoform of glutamine synthetase in gulf toadfish, Opsanus beta . Am J Physiol Integr Comp Physiol 297, R1075–R1081. Fabbri E, Capuzzo A & Moon TW (1998). The role of circulating catecholamines in the regulation of fish metabolism: An overview. Comp Biochem Physiol Part C 120, 177–192. Fabbri E & Moon TW (2015). Adrenergic signaling in teleost fish liver, a challenging path. Comp Biochem Physiol Part B Biochem Mol Biol 99, 114–124. Facchinello N, Skobo T, Meneghetti G, Colletti E, Dinarello A, Tiso N, Costa R, Gioacchini G, Carnevali O, Argenton F, Colombo L & Dalla Valle L (2017). nr3c1 null mutant zebrafish are viable and reveal DNA-binding-independent activities of the glucocorticoid receptor.

163

Sci Rep 7, 4371. Faught E, Aluru N & Vijayan MM (2016a). The Molecular Stress Responseed. Schreck CB, Tort L, Farrell AP & Brauner CJ. Elsevier Inc. Faught E, Best C & Vijayan MM (2016b). Maternal stress-associated cortisol stimulation may protect embryos from cortisol excess in zebrafish. R Soc Open Sci 3, 160032. Faught E, Henrickson L & Vijayan MM (2017). Plasma exosomes are enriched in Hsp70 and modulated by stress and cortisol in rainbow trout. J Endocrinol 232, 237–246. Faught E, Hernandez-Perez J, Wilson J & Vijayan MM (2019). Stress in Response to Environmental Changes. In Climate Change and Non-infectious Fish disorders, ed. Woo PTK & Iwama GK. CABI. Faught E & Vijayan MM (2016). Mechanisms of cortisol action in fish hepatocytes. Comp Biochem Physiol Part B 199B, 136. Faught E & Vijayan MM (2018a). The mineralocorticoid receptor is essential for stress axis regulation in zebrafish larvae. Sci Rep 8, 18081. Faught E & Vijayan MM (2018b). Maternal stress and fish reproduction: The role of cortisol revisited. Fish Fish1016–1030. Faught E & Vijayan MM (2019). Loss of the glucocorticoid receptor in zebrafish improves muscle glucose availability and increases growth. Am J Physiol Metab 316, E1093–E1104. Feng Q, Zou X, Lu L, Li Y, Liu Y, Zhou J & Duan C (2012). The Stress-Response Gene redd1Regulates Dorsoventral Patterning by Antagonizing Wnt/b-catenin Activity in Zebrafish. PLoS One 2012, e52674. Filby AL & Tyler CR (2007). Cloning and characterization of cDNAs for hormones and/or receptors of growth hormone, insulin-like growth factor-I, thyroid hormone, and corticosteroid and the gender-, tissue-, and developmental-specific expression of their mRNA transcripts in fathead m. Gen Comp Endocrinol 150, 151–163. Fletcher DJ (1984). The physiological control of appetite in fish. Comp Biochem Physiol -- Part A Physiol 78, 617–628. Fraccarollo D, Berger S, Galuppo P, Kneitz S, Hein L, Schütz G, Frantz S, Ertl G & Bauersachs J (2011). Deletion of cardiomyocyte mineralocorticoid receptor ameliorates adverse remodeling after myocardial infarction. Circulation 123, 400–408.

164

Fraher D, Sanigorski A, Mellett NA, Meikle PJ, Sinclair AJ & Gibert Y (2016). Zebrafish Embryonic Lipidomic Analysis Reveals that the Yolk Cell Is Metabolically Active in Processing Lipid. Cell Rep 14, 1317–1329. Frieler RA, Meng H, Duan SZ, Berger S, Schultz G, He Y, Xi G, Wang M & Mortensen RM (2011). Myeloid-Specific Deletion of the Mineralocorticoid Receptor Reduces Infarct Volume and Alters Inflammation During Cerebral Ischemia. Stroke 42, 179–185. Frolow J, Milligan CL & Anonymous (2004). Hormonal regulation of glycogen metabolism in white muscle slices from rainbow trout (Oncorhynchus mykiss Walbaum). Am J Physiol Regul Integr Comp Physiol 287, R1344-53. Funder J (2017). 30 YEARS OF THE MINERALOCORTICOID RECEPTOR: Mineralocorticoid receptor activation and specificity-conferring mechanisms: a brief history. J Endocrinol 234, T17–T21. Funder JW (2005). Mineralocorticoid Receptors: Distribution and Activation. Heart Fail Rev 10, 15–22. Funder JW, Pearce PT, Smith R & Smith AI (1988). Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science 242, 583–585. Galt NJ, Michael J, Remily EA, Romero SR & Biga PR (2014). The effects of exogenous cortisol on myostatin transcription in rainbow. Comp Biochem Physiol Part A 175, 57–63. Galuppo P, Vettorazzi S, Hövelmann J, Scholz CJ, Tuckermann JP, Bauersachs J & Fraccarollo D (2017). The glucocorticoid receptor in monocyte-derived macrophages is critical for cardiac infarct repair and remodeling. FASEB J 31, 5122–5132. Gao Y, Dai Z, Shi C, Zhai G, Jin X, He J, Lou Q & Yin Z (2016). Depletion of myostatin b promotes somatic growth and lipid metabolism in zebrafish. Front Endocrinol (Lausanne) 7, 1–10. Garvey SM, Dugle JE, Kennedy AD, McDunn JE, Kline W, Guo L, Guttridge DC, Pereira SL & Edens NK (2014). Metabolomic profiling reveals severe skeletal muscle group-specific perturbations of metabolism in aged FBN rats. Biogerontology 15, 217–232. Geer EB, Islam J & Buettner C (2014). Mechanisms of Glucocorticoid-Induced Insulin Resistance: Focus on Adipose Tissue Fuction and Lipid Metabolism. Endocrinol Metab Clin North Am 43, 75–102.

165

Geslin M & Auperin B (2004). Relationship between changes in mRNAs of the genes encoding steroidogenic acute regulatory protein and P450 cholesterol side chain cleavage in head kidney and plasma levels of cortisol in response to different kinds of acute stress in the rainbow trout. Gen Comp Endocrinol 135, 70–80. Gokulakrishnan G, Estrada IJ, Sosa HA & Fiorotto ML (2012). In utero glucocorticoid exposure reduces fetal skeletal muscle mass in rats independent of effects on maternal nutrition. Am J Physiol Integr Comp Physiol 302, R1143–R1152. Goodwin JE, Zhang J, Velazquez H & Geller DS (2010). The glucocorticoid receptor in the distal nephron is not necessary for the development or maintenance of dexamethasone- induced hypertension. Biochem Biophys Res Commun 394, 266–271. Gordon BS, Steiner JL, Williamson DL, Lang CH & Kimball SR (2016). Emerging role for regulated in development and DNA damage 1 (REDD1) in the regulation of skeletal muscle metabolism. Am J Physiol - Endocrinol Metab 311, E157–E174. Greenwood AK, Butler PC, White RB, Demarco U, Pearce D, Fernald RD, Greenwood AK, Butler PC, White RB, Marco UDE, Pearce D, Fernald RD & G PNAK (2003). Multiple Corticosteroid Receptors in a Teleost Fish: Distinct Sequences, Expression Patterns, and Transcriptional Activities. Endocrinology 144, 4226–4236. Greenwood P & Bell A (2003). Consequences of intra-uterine growth retardation for postnatal growth, metabolism and pathophysiology. Biosci Proc 61, 195–206. Griffiths BB, Schoonheim PJ, Ziv L, Voelker L, Baier H & Gahtan E (2012). A zebrafish model of glucocorticoid resistance shows serotonergic modulation of the stress response. Front Behav Neurosci 6, 1–10. Guo C, Ricchiuti V, Lian BQ, Yao TM, Coutinho P, Romero JR, Li J, Williams GH & Adler GK (2008). Mineralocorticoid Receptor Blockade Reverses Obesity-Related Changes in Expression of Adiponectin, Peroxisome Proliferator-Activated Receptor y, and Proinflammatory Adipokines. Circulation 177, 2253–2261. Halpern ME, Rhee J, Goll MG, Akitake CM, Parsons M & Leach SD (2008). Gal4/UAS Transgenic Tools and Their Application to Zebrafish Marnie. Zebrafish. Harris A & Seckl J (2011). Glucocorticoids, prenatal stress and the programming of disease. Horm Behav 59, 279–289.

166

Hartmann J, Dedic N, Pöhlmann ML, Häusl A, Karst H, Engelhardt C, Westerholz S, Wagner K V., Labermaier C, Hoeijmakers L, Kertokarijo M, Chen A, Joëls M, Deussing JM & Schmidt M V. (2017). Forebrain glutamatergic, but not GABAergic, neurons mediate anxiogenic effects of the glucocorticoid receptor. Mol Psychiatry 22, 466–475. He B, Cruz-Topete D, Oakley RH, Xiao X & Cidlowski JA (2016). Human Glucocorticoid Receptor β Regulates Gluconeogenesis and Inflammation in Mouse Liver. Mol Cell Biol 36, 714–730. Her GM, Hsu CC, Hong JR, Lai CY, Hsu MC, Pang HW, Chan SK & Pai WY (2011). Overexpression of gankyrin induces liver steatosis in zebrafish (Danio rerio). Biochim Biophys Acta - Mol Cell Biol Lipids 1811, 536–548. Herman JP, Mcklveen JM, Ghosal S, Kopp B, Wulsin A, Makinson R, Scheimann J & Myers B (2016). Regulation of the hypothalamic-pituitary-adrencocortical stress response. Compr Physiol 6, 603–621. Herman JP, Mcklveen JM, Solomon MB, Carvalho-Netto E & Myers B (2012). Neural regulation of the stress response: Glucocorticoid feedback mechanisms. Brazilian J Med Biol Res 45, 292–298. Hirata A, Maeda N, Hiuge A, Hibuse T, Fujita K, Okada T, Kihara S, Funahashi T & Shimomura I (2009). Blockade of mineralocorticoid receptor reverses adipocyte dysfunction and insulin resistance in obese mice. Cardiovasc Res 84, 164–172. Hirata A, Maeda N, Nakatsuji H, Hiuge-Shimizu A, Okada T, Funahashi T & Shimomura I (2012). Contribution of glucocorticoid-mineralocorticoid receptor pathway on the obesity- related adipocyte dysfunction. Biochem Biophys Res Commun 419, 182–187. Hoppmann J, Perwitz N, Meier B, Fasshauer M, Hadaschik D, Lehnert H & Klein J (2010). The balance between gluco- and mineralo-corticoid action critically determines inflammatory adipocyte responses. J Endocrinol 204, 153–164. Hruscha A, Krawitz P, Rechenberg A, Heinrich V, Hecht J, Haass C & Schmid B (2013). Efficient CRISPR/Cas9 genome editing with low off-target effects in zebrafish. Development 140, 4982–4987. Hu X & Funder JW (2006). The Evolution of Mineralocorticoid Receptors. Mol Endocrinol 20, 1471–1478.

167

Huang Y, Harrison MR, Osorio A, Kim J, Baugh A, Duan C, Sucov HM & Lien C-L (2013). Igf Signaling is Required for Cardiomyocyte Proliferation during Zebrafish Heart Development and Regeneration. PLoS One 8, e67266. Imrie D & Sadler KC (2011). White Adipose Tissue Development in Zebrafish Is Regulated by Both Developmental Time and Fish Size. Dev Dyn 239, 3013–3023. Ings JS, Servos MR & Vijayan MM (2011). Exposure to municipal wastewater effluent impacts stress performance in rainbow trout. Aquat Toxicol 103, 85–91. Itoh Y, Abe T, Takaoka R & Tanahashi N (2004). Fluorometric determination of glucose utilization in neurons in vitro and in vivo. J Cereb Blood Flow Metab 24, 993–1003. Jeffrey JD, Esbaugh AJ, Vijayan MM & Gilmour KM (2012). Modulation of hypothalamic- pituitary-interrenal axis function by social status in rainbow trout. Gen Comp Endocrinol 176, 201–210. Jeyapalan AS, Orellana RA, Suryawan A, O’Connor PMJ, Nguyen H V., Escobar J, Frank JW & Davis TA (2007). Glucose stimulates protein synthesis in skeletal muscle of neonatal pigs through an AMPK- and mTOR-independent process. AJP Endocrinol Metab 293, E595– E603. Jia G, Habibi J, Aroor AR, DeMarco VG, Ramirez-Perez FI, Sum Z, Hayden MR, Meininger GA, Mueller KB, Jaffe IZ & Sowers JR (2016). Endothelial Mineralocorticoid receptor mediates diet induced aortic stiffness in females. Circ Res 118, 935–943. Jinek M, Chylinski K, Fonfara I, Hauer M, Doudna JA & Charpentier E (2012). A Programmable Dual-RNA – Guided. Science 337, 816–822. Joëls M, Karst H, DeRijk R & de Kloet ER (2008). The coming out of the brain mineralocorticoid receptor. Trends Neurosci 31, 1–7. Joëls M & de Kloet ER (2017). The brain mineralocorticoid receptor: A saga in three episodes. J Endocrinol 234, T49–T66. Johansen IB, Lunde IG, Rosjo H, Christensen G, Nilsson GE, Bakken M & Overli O (2011). Cortisol response to stress is associated with myocardial remodeling in salmonid fishes. J Exp Biol 214, 1313–1321. John K, Marino JS, Sanchez ER & Hinds TD (2016). The glucocorticoid receptor: cause of or cure for obesity? Am J Endocrinol Metab 310, E249–E257.

168

Joung JK & Sander JD (2013). TALENs: a widely applicable technology for targeted genome editing. Nat Rev Mol Cell Biol 14, 49–55. De Joussineau C, Sahut-Barnola I, Levy I, Saloustros E, Val P, Stratakis CA & Martinez A (2012). The cAMP pathway and the control of adrenocortical development and growth. Mol Cell Endocrinol 351, 28–36. Katsu Y & Baker ME (2018). Progesterone activation of zebrafish mineralocorticoid receptor may influence growth of some transplanted tumors. Proc Natl Acad Sci 115, E2908–E2909. Kawakami K (2007). Tol2 : a versatile gene transfer vector in vertebrates. Genome Biol 8, 1–10. Kelly HW, Sternberg AL, Lescher R, Fuhlbrigge AL, Williams P, Zeiger RS, Raissy HH, Van Natta ML, Tonascia J & Strunk RC (2012). Effect of Inhaled Glucocorticoids in Childhood on Adult Height. N Engl J Med 367, 904–912. Kettelhut IC, Wing SS & Goldberg AL (1988). Endocrine regulation of protein breakdown in skeletal muscle. Diabetes Metab Rev 4, 751–772. Kiilerich P, Kristiansen K & Madsen SS (2007). Hormone receptors in gills of smolting Atlantic salmon, Salmo salar: expression of growth hormone, prolactin, mineralocorticoid and glucocorticoid receptors and 11beta-hydroxysteroid dehydrogenase type 2. Gen Comp Endocrinol 152, 295–303. Kiilerich P, Triqueneaux G, Christensen NM, Trayer V, Terrien X, Lombès M & Prunet P (2015). Interaction between the trout mineralocorticoid and glucocorticoid receptors in vitro. J Mol Endocrinol 55, 55–68. Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R & Salzberg SL (2013). TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol 14, R36. Kim J, Löwe T & Hoppe T (2008). Protein quality control gets muscle into shape. Trends Cell Biol 18, 264–272. Kimura Y, Hisano Y, Kawahara A & Higashijima S (2014). Transgenic Zebrafish Carrying Reporter/driver genes by CRISPR/Cas9-mediated genome engineering. Sci Rep 4, 1–7. Kino T, Su YA & Chrousos GP (2009). Human Glucocorticoid Receptor (GR) Isoform β: Recent Understanding of its Potential Implications in Physiology and Pathophysiology. Cell Mol Life Sci 66, 3435–3448.

169

Kintscher U (2005). PPAR -mediated insulin sensitization: the importance of fat versus muscle. AJP Endocrinol Metab 288, E287–E291. Kleiman A, Hubner S, Rodriguez Parkitna JM, Neumann A, Hofer S, Weigand MA, Bauer M, Schmid W, Schutz G, Libert C, Reichardt HM & Tuckermann JP (2012). Glucocorticoid receptor dimerization is required for survival in septic shock via suppression of interleukin- 1 in macrophages. FASEB J 26, 722–729. de Kloet E & Reul JMHM (1987). Feedback action and tonic influence of the corticosteroids on brain function: a concept arising from the heterogeneity of brain receptor systems. Psychoneuroendocrinology 12, 83–105. de Kloet ER (2014). From receptor balance to rational glucocorticoid therapy. Endocrinology 155, 2754–2769. de Kloet ER, Holmes MC, Harris AP, Seckl JR & Chapman KE (2012). Mineralocorticoid and glucocorticoid receptor balance in control of HPA axis and behaviour. Psychoneuroendocrinology 38, 648–658. de Kloet ER, Karst H & Joe M (2008). Corticosteroid hormones in the central stress response : Quick-and-slow. Neuroendocrinology 29, 268–272. Kolber BJ & Muglia LJ (2009). Defining brain region-specific glucocorticoid action during stress by conditional gene disruption in mice. Brain Res 1293, 85–90. Komor AC, Badran AH & Liu DR (2017). CRISPR-Based Technologies for the Manipulation of Eukaryotic Genomes. Cell 168, 20–36. Krug RG, Poshusta TL, Skuster KJ, Berg MR, Gardner SL & Clark KJ (2014). A transgenic zebrafish model for monitoring glucocorticoid receptor activity. Genes Brain Behav478– 487. Kuhn E, Bourgeois C, Keo V, Viengchareun S, Muscat A, Meduri G, Menuet D Le, Fève B & Lombès M (2014). Paradoxical resistance to high-fat diet-induced obesity and altered macrophage polarization in mineralocorticoid receptor-overexpressing mice. Am J Physiol Endocrinol Metab 306, E75–E90. Kumai Y, Nesan D, Vijayan MM & Perry SF (2012). Cortisol regulates Na + uptake in zebrafish, Danio rerio, larvae via the glucocorticoid receptor. Mol Cell Endocrinol 364, 113–125.

170

Kuo T, Harris CA & Wang J (2013). Metabolic functions of glucocorticoid receptor in skeletal muscle. Mol Cell Endocrinol 380, 79–88. Kuo T, Mcqueen A, Chen T & Wang J (2015). Glucocorticoid Signaling. In Advances in Experimental Medicine and Biology, ed. Wang J & Harris C, pp. 99–126. Springer Science. Lazaro-Côte A, Sadoul B, Jackson LJ & Vijayan MM (2018). Acute stress response of fathead minnows caged downstream of municipal wastewater treatment plants in the Bow River, Calgary. PLoS One 13, 1–15. Leung LY, Kwong AKY, Man AKY & Woo NYS (2008). Direct actions of cortisol , thyroxine and growth hormone on IGF-I mRNA expression in sea bream hepatocytes. Comp Biochem Physiol Part A 151, 705–710. Li M, Christie HL & Leatherland JF (2012). The in vitro metabolism of cortisol by ovarian follicles of rainbow trout (Oncorhynchus mykiss): comparison with ovulated oocytes and pre-hatch embryos. Reproduction 144, 713–722. Lin CH, Shih TH, Liu ST, Hsu HH & Hwang PP (2015). Cortisol regulates acid secretion of H+- ATPase-rich ionocytes in Zebrafish (Danio rerio) embryos. Front Physiol 6, 1–11. Lipina C & Hundal HS (2016). Is REDD1 a Metabolic Éminence Grise? Trends Endocrinol Metab 27, 868–880. Liu Q, Chen Y, Copeland D, Ball H, Duff RJ, Rockich B & Londraville RL (2010). Expression of leptin receptor gene in developing and adult zebrafish. Gen Comp Endocrinol 166, 346– 355. Liu Q, Dalman M, Chen Y, Akhter M, Brahmandam S, Patel Y, Lowe J, Thakkar M, Gregory AV, Phelps D, Riley C & Londraville RL (2012). Knockdown of leptin A expression dramatically alters zebrafish development. Gen Comp Endocrinol 178, 562–572. Livak KJ & Schmittgen TD (2001). Analysis of Relative Gene Expression Data Using Real- Time Quantitative PCR and the 2−ΔΔCT Method. Methods 25, 402–408. Löhr H, Hess S, Pereira MMA, Reinoß P, Leibold S, Schenkel C, Wunderlich CM, Kloppenburg P, Brüning JC & Hammerschmidt M (2018). Diet-Induced Growth Is Regulated via Acquired Leptin Resistance and Engages a Pomc-Somatostatin-Growth Hormone Circuit. Cell Rep 23, 1728–1741. Madan a P & DeFranco DB (1993). Bidirectional transport of glucocorticoid receptors across

171

the nuclear envelope. Proc Natl Acad Sci U S A 90, 3588–3592. Maddison LA, Joest KE, Kammeyer RM & Chen W (2015). Skeletal muscle insulin resistance in zebrafish induces alterations in β-cell number and glucose tolerance in an age- and diet- dependent manner. Am J Physiol - Endocrinol Metab 308, E662–E669. Madison BN, Tavakoli S, Kramer S & Bernier NJ (2015). Chronic cortisol and the regulation of food intake and the endocrine growth axis in rainbow trout. J Endocrinol 226, 103–119. Mania M, Maruccio L, Russo F, Abbate F, Castaldo L, D’Angelo L, de Girolamo P, Guerrera MC, Lucini C, Madrigrano M, Levanti M & Germanà A (2017). Expression and distribution of leptin and its receptors in the digestive tract of DIO (diet-induced obese) zebrafish. Ann Anat 212, 37–47. Marzolla V, Armani A, Feraco A, De Martino MU, Fabbri A, Rosano G & Caprio M (2014). Mineralocorticoid receptor in adipocytes and macrophages: A promising target to fight metabolic syndrome. Steroids 91, 46–53. Marzolla V, Armani A, Zennaro MC, Cinti F, Mammi C, Fabbri A, Rosano GMC & Caprio M (2012). The role of the mineralocorticoid receptor in adipocyte biology and fat metabolism. Mol Cell Endocrinol 350, 281–288. Mathew LK, Sengupta S, Kawakami A, Andreasen EA, Löhr C V., Loynes CA, Renshaw SA, Peterson RT & Tanguay RL (2007). Unraveling tissue regeneration pathways using chemical genetics. J Biol Chem 282, 35202–35210. McCormick SD, Regish A, O’Dea MF & Shrimpton JM (2008). Are we missing a mineralocorticoid in teleost fish? Effects of cortisol, deoxycorticosterone and aldosterone on osmoregulation, gill Na+,K+ -ATPase activity and isoform mRNA levels in Atlantic salmon. Gen Comp Endocrinol 157, 35–40. McCormick SD, Shrimpton JM, Carey JB, O’Dea MF, Sloan KE, Moriyama S & Björnsson BT (1998). Repeated acute stress reduces growth rate of Atlantic salmon parr and alters plasma levels of growth hormone, insulin-like growth factor I and cortisol. Aquaculture 168, 221– 235. McCurley A, Pires PW, Bender SB, Aronovitz M, Zhao MJ, Metzger D, Chambon P, Hill MA, Dorrance AM, Mendelsohn ME & Jaffe IZ (2012). Direct regulation of blood pressure by smooth muscle cell mineralocorticoid receptors. Nat Med 18, 1429–1433.

172

McGowan PO & Matthews SG (2018). Prenatal stress, glucocorticoids, and developmental programming of the stress response. Endocrinology 159, 69–82. McLaren DG et al. (2018). DGAT2 Inhibition Alters Aspects of Triglyceride Metabolism in Rodents but Not in Non-human Primates. Cell Metab 27, 1236-1248.e6. Metsalu T & Vilo J (2015). ClustVis: A web tool for visualizing clustering of multivariate data using Principal Component Analysis and heatmap. Nucleic Acids Res 43, W566–W570. Michel M, Page-McCaw PS, Chen W & Cone RD (2016). Leptin signaling regulates glucose homeostasis, but not adipostasis, in the zebrafish. Proc Natl Acad Sci 113, 3084–3089. Mifsud KR & Reul JMHM (2016). Acute stress enhances heterodimerization and binding of corticosteroid receptors at glucocorticoid target genes in the hippocampus. Proc Natl Acad Sci 113, 11336–11341. Mifsud KR & Reul JMHM (2018). Mineralocorticoid and glucocorticoid receptor-mediated control of genomic responses to stress in the brain. Stress 21, 389–402. Milla S, Wang N, Mandiki SNM & Kestemont P (2009). Corticosteroids: Friends or foes of teleost fish reproduction? Comp Biochem Physiol A Mol Integr Physiol 153, 242–251. Milligan CL (1997). The role of cortisol in amino acid mobilization and metabolism following exhaustive exercise in rainbow trout (Oncorhynchus mykiss Walbaum). Fish Physiol Biochem 16, 119–128. Milligan CL (2003). A regulatory role for cortisol in muscle glycogen metabolism in rainbow trout Oncorhynchus mykiss Walbaum. J Exp Biol 206, 3167–3173. Minchen J & Rawls JF (2015). In vivo Analysis of White Adipose Tissue in Zebrafish. Methods Cell Biol 14, 871–882. Mommsen TP (2001). Paradigms of growth in fish. Comp Biochem Physiol - B Biochem Mol Biol 129, 207–219. Mommsen TP, Vijayan MM & Moon TW (1999). Cortisol in teleosts: dynamics, mechanisms of action, and metabolic regulation. Rev Fish Biol Fish 9, 211–268. Monroig Ó, Rotllant J, Sánchez E, Cerdá-Reverter JM & Tocher DR (2009). Expression of long- chain polyunsaturated fatty acid (LC-PUFA) biosynthesis genes during zebrafish Danio rerio early embryogenesis. Biochim Biophys Acta - Mol Cell Biol Lipids 1791, 1093–1101. Moon TW (2001). Glucose intolerance in teleost fish: fact or fiction? Comp Biochem Physiol

173

Part B Biochem Mol Biol 129, 243–249. Mueller KM et al. (2011). Impairment of hepatic growth hormone and glucocorticoid receptor signaling causes steatosis and hepatocellular carcinoma in mice. Hepatology 54, 1398– 1409. Nakano T, Afonso LOB, Beckman BR, Iwama GK & Devlin RH (2013). Acute physiological stress down-regulates mRNA expressions of growth-related genes in coho salmon. PLoS One 8, e71421. Nesan D, Kamkar M, Burrows J, Scott IC, Marsden M & Vijayan MM (2012). Glucocorticoid receptor signaling is essential for mesoderm formation and muscle development in zebrafish. Endocrinology 153, 1288–1300. Nesan D & Vijayan MM (2012). Embryo exposure to elevated cortisol level leads to cardiac performance dysfunction in zebrafish. Mol Cell Endocrinol 363, 85–91. Nesan D & Vijayan MM (2013a). Role of glucocorticoid in developmental programming: Evidence from zebrafish. Gen Comp Endocrinol 181, 35–44. Nesan D & Vijayan MM (2013b). The transcriptomics of glucocorticoid receptor signaling in developing zebrafish. ed. Fuentes J. PLoS One 8, e80726. Nesan D & Vijayan MM (2016). Maternal Cortisol Mediates Hypothalamus-Pituitary-Interrenal Axis Development in Zebrafish. Sci Rep 6, 22582. Nusslein-Volhard C & Dahm R (2002). Zebrafish: a practical approach. Oxford University Press, New York. Oakley RH, Cruz-Topete D, He B, Foley JF, Myers PH, Xu X, Gomez-Sanchez CE, Chambon P, Willis MS & Cidlowski JA (2019). Cardiomyocyte glucocorticoid and mineralocorticoid receptors directly and antagonistically regulate heart disease in mice. Sci Signal 12, eaau9685. Oitzl MS, Reichardt HM, Joëls M & de Kloet ER (2001). Point mutation in the mouse glucocorticoid receptor preventing DNA binding impairs spatial memory. Proc Natl Acad Sci 98, 12790–12795. Opherk C, Tronche F, Kellendonk C, Kohlmüller D, Schulze A, Schmid W & Schütz G (2004). Inactivation of the Glucocorticoid Receptor in Hepatocytes Leads to Fasting Hypoglycemia and Ameliorates Hyperglycemia in Streptozotocin-Induced Diabetes Mellitus. Mol

174

Endocrinol 18, 1346–1353. Ortega VA, Lovejoy DA & Bernier NJ (2013). Appetite-suppressing effects and interactions of centrally administered corticotropin-releasing factor, urotensin I and serotonin in rainbow trout (Oncorhynchus mykiss). Front Neurosci 7, 196. Ou XM, Storring JM, Kushwaha N & Albert PR (2001). Heterodimerization of Mineralocorticoid and Glucocorticoid Receptors at a Novel Negative Response Element of the 5-HT1A Receptor Gene. J Biol Chem 276, 14299–14307. Paitz RT, Bukhari SA & Bell AM (2016). Stickleback embryos use ATP-binding cassette transporters as a buffer against exposure to maternally derived cortisol. Proc R Soc B Biol Sci 283, 1–7. Paquet D, Kwart D, Chen A, Sproul A, Jacob S, Teo S, Olsen KM, Gregg A, Noggle S & Tessier-Lavigne M (2016). Efficient introduction of specific homozygous and heterozygous mutations using CRISPR/Cas9. Nature 533, 125–129. Parnaudeau S, Dongelmans M, Turiault M, Ambroggi F, Delbes A-S, Cansell C, Luquet S, Piazza P-V, Tronche F & Barik J (2014). Glucocorticoid receptor gene inactivation in dopamine-innervated areas selectively decreases behavioral responses to amphetamine. Front Behav Neurosci 8, 1–12. Patel J, Wang X & Proud CG (2001). Glucose exerts a permissive effect on the regulation of the initiation factor 4E binding protein 4E-BP1. Biochem J 358, 497-503. Pauter AM, Trattner S, Gonzalez-Bengtsson A, Talamonti E, Asadi A, Dethlefsen O & Jacobsson A (2016). Both maternal and offspring Elovl2 genotypes determine systemic DHA levels in perinatal mice. J Lipid Res 58, 111–123. Peeke PM & Chrousos GP (1995). Hypercortisolism and Obesity. Ann New York Acad Sci 771, 665–676. Philip AM, Daniel Kim S & Vijayan MM (2012). Cortisol modulates the expression of cytokines and suppressors of cytokine signaling (SOCS) in rainbow trout hepatocytes. Dev Comp Immunol 38, 360–367. Philip AM, Jorgensen EH, Maule AG & Vijayan MM (2014). Tissue-specific molecular immune response to lipopolysaccharide challenge in emaciated anadromous Arctic charr. Dev Comp Immunol 45, 133–140.

175

Philip AM & Vijayan MM (2015). Stress-Immune-Growth Interactions: Cortisol Modulates Suppressors of Cytokine Signaling and JAK/STAT Pathway in Rainbow Trout Liver. PLoS One 10, e0129299. Phillips JB & Westerfield M (2014). Zebrafish models in translational research: tipping the scales toward advancements in human health. Dis Model Mech 7, 739–743. Pickering AD (1993). Growth and stress in fish production. Aquaculture 111, 51–63. Pierce AL, Breves JP, Moriyama S, Hirano T & Grau GE (2011). Differential regulation of Igf1 and Igf2 mRNA levels in tilapia hepatocytes: effects of insulin and cortisol on GH sensitivity. J Endocrinol 211, 187–200. Pierce AL, Dickey JT, Felli L, Swanson P & Dickhoff WW (2010). Metabolic hormones regulate basal and growth hormone-dependent igf 2 mRNA level in primary cultured coho salmon hepatocytes: Effects of insulin, glucagon, dexamethasone, and triiodothyronine. J Endocrinol 204, 331–339. Pikulkaew S, Benato F, Celeghin a, Zucal C, Skobo T, Colombo L & Dalla Valle L (2011). The knockdown of maternal glucocorticoid receptor mRNA alters embryo development in zebrafish. Dev Dyn 240, 874–889. Pikulkaew S, De Nadai A, Belvedere P, Colombo L & Dalla Valle L (2010). Expression analysis of steroid hormone receptor mRNAs during zebrafish embryogenesis. Gen Comp Endocrinol 165, 215–220. Pippal JB, Cheung CMI, Yao Y-Z, Brennan FE & Fuller PJ (2011). Characterization of the zebrafish (Danio rerio) mineralocorticoid receptor. Mol Cell Endocrinol 332, 58–66. Polakof S, Panserat S, Soengas JL & Moon TW (2012a). Glucose metabolism in fish: a review. J Comp Physiol B 182, 1015–1045. Polakof S, Panserat S, Soengas JL & Moon TW (2012b). Glucose metabolism in fish: A review. J Comp Physiol B Biochem Syst Environ Physiol 182, 1015–1045. Pratt WB, Morishima Y, Murphy M & Harrell M (2006). Chaperoning of glucocorticoid receptors. In Handbook of experimental pharmacology, pp. 111–138. Springer-Verlag Prunet P, Sturm A & Milla S (2006). Multiple corticosteroid receptors in fish: from old ideas to new concepts. Gen Comp Endocrinol 147, 17–23. Prykhozhij S V & Berman JN (2018). Zebrafish knock-ins swim into the mainstream. Dis Mod

176

Mech 11, dmm037515 Rapp AE, Hachemi Y, Kemmler J, Koenen M, Tuckermann J & Ignatius A (2018). Induced global deletion of glucocorticoid receptor impairs fracture healing. FASEB J 32, 2235– 2245. Reid SD, Moon TW & Perry SF (1992). Rainbow trout hepatocyte beta-adrenoceptors, catecholamine responsiveness, and effects of cortisol. Am J Physiol 262, R794-9. Reid SG, Bernier NJ & Perry SF (1998). The adrenergic stress response in fish: contraol of catecholamine storage and release. Comp Biochem Physiol Part C 120, 1–27. Reindl KM & Sheridan MA (2012). Peripheral regulation of the growth hormone-insulin-like growth factor system in fi sh and other vertebrates. Comp Biochem Physiol Part A 163, 231–245. Rennie MJ, Edwards RHT, Emery PW, Halliday D, Lundholm K & Millward DJ (1983). Depressed protein synthesis is the dominant characteristic of muscle wasting and cachexia. Clin Physiol 3, 387–398. Richardson CD, Ray GJ, DeWitt MA, Curie GL & Corn JE (2016). Enhancing homology- directed genome editing by catalytically active and inactive CRISPR-Cas9 using asymmetric donor DNA. Nat Biotechnol 34, 339–344. Rickard AJ, Morgan J, Bienvenu LA, Fletcher EK, Cranston GA, Shen JZ, Reichelt ME, Delbridge LM & Young MJ (2012). Cardiomyocyte mineralocorticoid receptors are essential for deoxycorticosterone/salt-mediated inflammation and cardiac fibrosis. Hypertension 60, 1443–1450. Rickard AJ, Morgan J, Chrissobolis S, Miller AA, Sobey CG & Young MJ (2014). Endothelial cell mineralocorticoid receptors regulate deoxycorticosterone/ salt-mediated cardiac remodeling and vascular reactivity but not blood pressure. Hypertension 63, 1033–1040. Rivers CA, Rogers MF, Stubbs FE, Conway-Campbell BL, Lightman SL & Pooley JR (2019). Glucocorticoid Receptor–Tethered Mineralocorticoid Receptors Increase Glucocorticoid- Induced Transcriptional Responses. Endocrinology 160, 1044–1056. Rodgers BD, Weber GM, Kelley KM & Levine MA (2003). Prolonged fasting and cortisol reduce myostatin mRNA levels in tilapia larvae; short-term fasting elevates. Am J Physiol Regul Integr Comp Physiol 284, R1277–R1286.

177

Ronzaud C, Loffing J, Bleich M, Gretz N, Grone HJ, Schutz G & Berger S (2007). Impairment of sodium balance in mice deficient in renal principal cell mineralocorticoid receptor. J Am Soc Nephrol 18, 1679–1687. Rossi A, Kontarakis Z, Gerri C, Nolte H, Hölper S, Krüger M & Stainier DYR (2015). Genetic compensation induced by deleterious mutations but not gene knockdowns. Nature Aug 13, 230–233. Rozeboom AM, Akil H & Seasholtz AF (2007). Mineralocorticoid receptor overexpression in forebrain decreases anxiety-like behavior and alters the stress response in mice. Proc Natl Acad Sci U S A 104, 4688–4693. Sadoul B, Birceanu O, Aluru N, Thomas JK & Vijayan MM (2017). Bisphenol A in eggs causes development-specific liver molecular reprogramming in two generations of rainbow trout. Sci Rep 7, 1–11. Sadoul B & Vijayan MM (2016). Stress and Growth. In Fish Physiology, ed. Schreck CB, Tort L, Farrell AP & Brauner CJ, pp. 167–205. Elsevier Inc. Sakamoto T, Yoshiki M, Takahashi H, Yoshida M, Ogino Y, Ikeuchi T, Nakamachi T, Konno N, Matsuda K & Sakamoto H (2016). Principal function of mineralocorticoid signaling suggested by constitutive knockout of the mineralocorticoid receptor in medaka fish. Sci Rep 6, 37991. Saltiel AR & Kahn CR (2001). Insulin signalling and the regulation of glucose and lipid metabolism. Nature 414, 799–806. Sartin J, Kemppainen R, ES C, Steele B & Williams J (1994). Cortisol inhibition of growth hormone-releasing hormone-stimulated growth hormone release from cultured sheep pituitary cells. J Endocrinol 141, 517–525. Sathiyaa R & Vijayan MM (2003). Autoregulation of glucocorticoid receptor by cortisol in rainbow trout hepatocytes. AJP Cell Physiol 284, C1508–C1515. Schaaf MJM, Champagne D, Van Laanen IHC, Van Wijk DCWA, Meijer AH, Meijer OC, Spaink HP & Richardson MK (2008). Discovery of a functional glucocorticoid receptor β- isoform in zebrafish. Endocrinology 149, 1591–1598. Schäfer N, Lohmann C, Winnik S, Van Tits LJ, Miranda MX, Vergopoulos A, Ruschitzka F, Nussberger J, Berger S, Lüscher TF, Verrey F & Matter CM (2013). Endothelial

178

mineralocorticoid receptor activation mediates endothelial dysfunction in diet-induced obesity. Eur Heart J 34, 3515–3524. Schnörr SJ, Steenbergen PJ, Richardson MK & Champagne DL (2012). Measuring thigmotaxis in larval zebrafish. Behav Brain Res 228, 367–374. Schoneveld OJLM, Gaemers IC & Lamers WH (2004). Mechanisms of glucocorticoid signalling. Biochim Biophys Acta - Gene Struct Expr 1680, 114–128. Schreck CB & Tort L (2016). The Concept of Stress in Fished. Schreck CB, Tort L, Farrell AP & Brauner CJ, pp 1-34. Elsevier Inc. Seckl JR & Walker BR (2014). Minireview: 11β-Hydroxysteroid Dehydrogenase Type1-A Tissue-Specific Amplifier of Glucocorticoid Action. Endocrinology 142, 1371–1376. Sen ND, Zhou F, Harris MS, Ingolia NT & Hinnebusch AG (2016). eIF4B stimulates translation of long mRNAs with structured 5′ UTRs and low closed-loop potential but weak dependence on eIF4G. Proc Natl Acad Sci 113, 10464–10472. Seth A, Stemple DL & Barroso I (2013). The emerging use of zebrafish to model metabolic disease. Dis Model Mech 6, 1080–1088. Shapiro LE, Samuels HH & Yaffe BM (1978). Thyroid and glucocorticoid hormones synergistically control growth hormone mRNA in cultured GH1 cells. Proc Natl Acad Sci 75, 45–49. Shen Y, Roh HC, Kumari M & Rosen ED (2017a). Adipocyte glucocorticoid receptor is important in lipolysis and insulin resistance due to exogenous steroids, but not insulin resistance caused by high fat feeding. Mol Metab 6, 1150–1160. Shen ZX, Chen XQ, Sun XN, Sun JY, Zhang WC, Zheng XJ, Zhang YY, Shi HJ, Zhang JW, Li C, Wang J, Liu X & Duan SZ (2017b). Mineralocorticoid receptor deficiency in macrophages inhibits atherosclerosis by affecting foam cell formation and efferocytosis. J Biol Chem 292, 925–935. Shepherd BS, Aluru N & Vijayan MM (2011). Acute handling disturbance modulates plasma insulin-like growth factor binding proteins in rainbow trout (Oncorhynchus mykiss). Domest Anim Endocrinol 40, 129–138. Sheridan MA (1988). Lipid dynamics in fish: aspects of absorption, transportation, deposition and mobilization. Comp Biochem Physiol -- Part B Biochem 90, 679–690.

179

Sheridan MA (1994). Regulation of lipid metabolism in poikilothermic vertebrates. Comp Biochem Physiol Part B Comp Biochem 107, 495–508. Shibata S (2017). Mineralocorticoid receptor and NaCl transport mechanisms in the renal distal nephron. J Endocrinol 234, T35–T47. Shimizu H, Langenbacher AD, Huang J, Wang K, Otto G, Geisler R, Wang Y & Chen JN (2017). The calcineurin-FoxO-MuRF1 signaling pathway regulates myofibril integrity in cardiomyocytes. elife 6, 1–19. Shimizu N, Maruyama T, Yoshikawa N, Matsumiya R, Ito N, Tasaka Y, Kuribara-souta A, Miyata K, Oike Y, Berger S & Tanaka H (2015). A muscle-liver-fat signalling axis is essential for central control of adaptive adipose remodelling. Nat Commun 6, 4–6. Shimizu N, Yoshikawa N, Ito N, Maruyama T, Suzuki Y, Takeda SI, Nakae J, Tagata Y, Nishitani S, Takehana K, Sano M, Fukuda K, Suematsu M, Morimoto C & Tanaka H (2011). Crosstalk between glucocorticoid receptor and nutritional sensor mTOR in skeletal muscle. Cell Metab 13, 170–182. Simon P, Dupuis R & Costentin J (1994). Thigmotaxis as an index of anxiety in mice. Influence of dopaminergic transmissions. Behav Brain Res 61, 59–64. Small BC, Murdock C a., Waldbieser GC & Peterson BC (2006). Reduction in channel catfish hepatic growth hormone receptor expression in response to food deprivation and exogenous cortisol. Domest Anim Endocrinol 31, 340–356. Solano JM & Jacobson L (1999). Glucocorticoids reverse leptin effects on food intake and body fat in mice without increasing NPY mRNA. Am J Physiol Endocrinol Metab 277, E708- 716. Sternberg SH, Redding S, Jinek M, Greene EC & Doudna JA (2014). DNA interrogation by the CRISPR RNA-guided endonuclease Cas9. Nature 507, 62–67. Sternberg SH, Richter H, Charpentier E & Qimron U (2016). Adaptation in CRISPR-Cas Systems. Mol Cell 61, 797–808. Stewart AM, Braubach O, Spitsbergen J, Gerlai R & Kalueff A V. (2014). Zebrafish models for translational neuroscience research: From tank to bedside. Trends Neurosci 37, 264–278. Stocco DM (2001). StAR Protein and Regulation of Steroid Hormone Biosynthesis. Annu Rev Physiol 63, 193–213.

180

Stolte EH, Nabuurs SB, Bury NR, Sturm A, Flik G, Savelkoul HFJ & Lidy Verburg-van Kemenade BM (2008). Stress and innate immunity in carp: Corticosteroid receptors and pro-inflammatory cytokines. Mol Immunol 46, 70–79. Stolte EH, Verburg van Kemenade BML, Savelkoul HFJ & Flik G (2006). Evolution of glucocorticoid receptors with different glucocorticoid sensitivity. J Endocrinol 190, 17–28. Stone SJ, Myers HM, Watkins SM, Brown BE, Feingold KR, Elias PM & Farese R V (2004). Lipopenia and Skin Barrier Abnormalities in DGAT2-deficient Mice. J Biol Chem 279, 11767–11776. Sturm A, Bury N, Dengreville L, Fagart J, Flouriot G, Rafestin-Oblin ME & Prunet P (2005). 11- Deoxycorticosterone is a potent agonist of the rainbow trout (Oncorhynchus mykiss) mineralocorticoid receptor. Endocrinology 146, 47–55. Sugawara T, Kiriakidou M, McAllister JM, Holt JA, Arakane F & Strauss 3rd JF (1997). Regulation of expression of the steroidogenic acute regulatory protein (StAR) gene: a central role for steroidogenic factor 1 [published erratum appears in Steroids 1997 Apr;62(4):395]. Steroids 62, 5–9. Suster ML, Kikuta H, Urasaki A, Asakawa K & Kawakami K (2009). Transgenesis in Zebrafish with the Tol2 Transposon System. In Transgenesis Techniques, Methods in Molecular Biology, ed. Cartwright EJ. Humana Press. Swift D. (1982). The measurement of ACTH in the plasma of rainbow trout (Salmo gairdneri Richardson) using two commercial radioimmunoassay kits. Comp Biochem Physiol Part A Physiol 72, 679–681. Tagawa M, Suzuki K & Specker J (2000). Incorporation and Metabolism of Cortisol in Oocytes of Tilapia. J Exp Biol 287, 485–492. Terker AS, Yarbrough B, Ferdaus MZ, Lazelle RA, Erspamer KJ, Meermeier NP, Park HJ, McCormick JA, Yang C-L & Ellison DH (2016). Direct and Indirect Mineralocorticoid Effects Determine Distal Salt Transport. J Am Soc Nephrol 27, 2436–2445. Terova G, Gornati R, Rimoldi S, Bernardini G & Saroglia M (2005). Quantification of a glucocorticoid receptor in sea bass (Dicentrarchus labrax, L.) reared at high stocking density. Gene 357, 144–151. De Theije CC, Schols AMWJ, Lamers WH, Ceelen JJM, Van Gorp RH, Rob Hermans JJ,

181

Elonore Köhler S & Langen RCJ (2018). Glucocorticoid receptor signaling impairs protein turnover regulation in hypoxia-Induced muscle atrophy in male mice. Endocrinology 159, 519–534. Thulasiraman V, Yang CF & Frydman J (1999). In vivo newly translated polypeptides are sequestered in a protected folding environment. EMBO J 18, 85–95. Tocher DR (2003). Metabolism and Functions of Lipids and Fatty Acids in Teleost Fish. Rev Fish Sci 11, 107–184. Tomlinson JW & Stewart PM (2005). Mechanisms of disease: Selective inhibition of 11beta- hydroxysteroid dehydrogenase type 1 as a novel treatment for the metabolic syndrome. Nat Clin Pract Endocrinol Metab 1, 92–99. Tort L (2011). Stress and immune modulation in fish. Dev Comp Immunol 35, 1366–1375. Trenzado CE, Morales AE & de la Higuera M (2006). Physiological effects of crowding in rainbow trout, Oncorhynchus mykiss, selected for low and high stress responsiveness. Aquaculture 258, 583–593. Tronche F, Kellendonk C, Reichardt HM & Schütz G (1998). Genetic dissection of glucocorticoid receptor function in mice. Curr Opin Genet Dev 8, 532–538. Tronche F, Opherk C, Moriggl R, Kellendonk C, Reimann A, Schwake L, Reichardt HM, Stangl K, Gau D, Hoeflich A, Beug H, Schmid W & Schütz G (2004). Glucocorticoid receptor function in hepatocytes is essential to promote postnatal body growth. Genes Dev 18, 492– 497. Tseng Y, Chen R, Lee J, Liu S, Lee S & Hwang P (2009). Specific expression and regulation of glucose transporters in zebrafish ionocytes. Am J Physiol-Reg I 297, 275–290. Uddén J, Björntorp P, Arner P, Barkeling B, Meurling L & Rössner S (2003). Effects of glucocorticoids on leptin levels and eating behaviour in women. J Intern Med 253, 225– 231. Usher MG, Duan SZ, Ivaschenko CY, Frieler RA, Berger S, Schütz G, Lumeng CN & Mortensen RM (2010). Myeloid mineralocorticoid receptor controls macrophage polarization and cardiovascular hypertrophy and remodeling in mice. J Clin Invest 120, 3350–3364. Varshney GK, Carrington B, Pei W, Bishop K, Chen Z, Fan C, Xu L, Jones M, LaFave MC,

182

Ledin J, Sood R & Burgess SM (2016). A high-throughput functional genomics workflow based on CRISPR/Cas9-mediated targeted mutagenesis in zebrafish. Nat Protoc 11, 2357– 2375. Varshney GK, Pei W, Lafave MC, Idol J, Xu L, Gallardo V, Carrington B, Bishop K, Jones M, Li M, Harper U, Huang SC, Prakash A, Chen W, Sood R, Ledin J & Burgess SM (2015). High-throughput gene targeting and phenotyping in zebrafish using CRISPR / Cas9. Genome Res 25, 1030–1042. Vegiopoulos A & Herzig S (2007). Glucocorticoids, metabolism and metabolic diseases. Mol Cell Endocrinol 275, 43–61. Venihaki M & Majzoub J (2002). Lessons from CRH knockout mice. Neuropeptides 36, 96–102. Viengchareun S, Le Menuet D, Martinerie L, Munier M, Pascual-Le Tallec L & Lombès M (2007). The mineralocorticoid receptor: insights into its molecular and (patho)physiological biology. Nucl Recept Signal 5, e012. Vijayan MM, Aluru N & Leatherland JF (2010). Stress Response and the Role of Cortisol. In Fish Diseases and Disorders Volume 2: Non-infectious disorders, ed. Leatherland JF & Woo P, pp. 182–201. CABI, Oxfordshire, UK. Vijayan MM, Ballantyne JS & Leatherland JF (1991). Cortisol-induced changes in some aspects of the intermediary metabolism of Salvelinus fontinalis. Gen Comp Endocrinol 82, 476– 486. Vijayan MM & Leatherland JF (1992). In vivo effects of the steroid analogue RU486 on some aspects of intermediary and thyroid metabolism of brook charr, Salvelinus fontinalis. J Exp Zool 263, 265–271. Vijayan MM, Mommsen TP, Glemet HC & Moon TW (1996). Metabolic Effects of cortisol treatment in marine teleost, the sea raven. J experimenal Biol 1514, 1509–1514. Vijayan MM & Moon TW (1994). The stress response and the plasma disappearance of corticosteroid and glucose in a marine teleost, the sea raven. Can J Zool 72, 379–386. Vijayan MM, Pereira C, Forsyth RB, Kennedy CJ & Iwama GK (1997). Handling stress does not affect the expression of hepatic heat shock protein 70 and conjugation enzymes in rainbow trout treated with β-naphthoflavone. Life Sci 61, 117–127. Vijayan MM, Prunet P & Boone AN (2005). Xenobiotic impact on corticosteroid signaling. In

183

Biochemistry and Molecular Biology of Fishes, ed. Mommsen TP & Moon TW, pp. 365– 394. Elsevier. Vijayan MM, Raptis S & Sathiyaa R (2003). Cortisol treatment affects glucocorticoid receptor and glucocorticoid-responsive genes in the liver of rainbow trout. Gen Comp Endocrinol 132, 256–263. Wang F, Canadeo LA & Huibregtse JM (2015). Ubiquitination of newly synthesized proteins at the ribosome. Biochimie 114, 127–133. Wang H, La Russa M & Qi LS (2016). CRISPR/Cas9 in Genome Editing and Beyond. Annu Rev Biochem 85, 227–264. Wang X, Hu Z, Hu J, Du J & Mitch WE (2006). Insulin resistance accelerates muscle protein degradation: Activation of the ubiquitin-proteasome pathway by defects in muscle cell signaling. Endocrinology 147, 4160–4168. Watson ML, Baehr LM, Reichardt HM, Tuckermann JP, Bodine SC & Furlow JD (2012). A cell- autonomous role for the glucocorticoid receptor in skeletal muscle atrophy induced by systemic glucocorticoid exposure. AJP Endocrinol Metab 302, E1210–E1220. Weger BD, Weger M, Görling B, Schink A, Gobet C, Keime C, Poschet G, Jost B, Krone N, Hell R, Gachon F, Luy B & Dickmeis T (2016). Extensive Regulation of Diurnal Transcription and Metabolism by Glucocorticoids. PLoS Genet 12, 1–24. Wendelaar Bonga SE (1997). The Stress Response in Fish. Physiol Rev 77, 591–625. Wilson KS, Baily J, Tucker CS, Matrone G, Vass S, Moran C, Chapman KE, Mullins JJ, Kenyon C, Hadoke PWF & Denvir MA (2015). Early-life perturbations in glucocorticoid activity impacts on the structure, function and molecular composition of the adult zebrafish (Danio rerio) heart. Mol Cell Endocrinol 414, 120–131. Wilson KS, Matrone G, Livingstone DEW, Al-Dujaili EAS, Mullins JJ, Tucker CS, Hadoke PWF, Kenyon CJ, Denvir MA, Al-Dujaili, Mullins JJ, Tucker PW, Kenyon CJ & Denvir MA (2013). Physiological roles of glucocorticoids during early embryonic development of the zebrafish (Danio rerio). J Physiol 591, 6209–6220. Wilson KS, Tucker CS, Al-Dujaili EAS, Holmes MC, Hadoke PWF, Kenyon CJ, Denvir MA, Al-Dujaili S, Holmes MC, Hadoke PWF, Kenyon CJ & Denvir MA (2016). Early-life glucocorticoids programme behaviour and metabolism in adulthood in zebrafish. J

184

Endocrinol 230, 125–142. Winberg S & Nilsson GE (1993). Roles of brain monoamine neurotransmitters in agonistic behaviour and stress reactions, with particular reference to fish. Comp Biochem Physiol Part C Comp 106, 597–614. Yuen KCJ, Chong LE & Riddle MC (2013). Influence of glucocorticoids and growth hormone on insulin sensitivity in humans. Diabet Med 30, 651–663. Zakrzewska KE, Cusin I, Sainsbury A, Rohner-Jeanrenaud F & Jeanrenaud B (1997). Glucocorticoids as Counterregulatory Hormones of Leptin: Toward an Understanding of Leptin Resistance. Diabetes 46, 717–719. Zang L, Shimada Y & Nishimura N (2017). Development of a Novel Zebrafish Model for Type 2 Diabetes Mellitus. Sci Rep 7, 1–11. Ziv L, Muto A, Schoonheim PJ, Meijsing SH, Strasser D, Ingraham HA, Schaaf MJM, Yamamoto KR & Baier H (2013). An affective disorder in zebrafish with mutation of the glucocorticoid receptor. Mol Psychiatry 18, 681–691.

185

APPENDIX I: SUPPLEMENTAL METHODS

186

Table 1: GR and MR mutants and the associated phenotypes: All known ubiquitous and conditional knockouts of both nr3c1 (glucocorticoid receptor) and nr3c2 (mineralocorticoid receptor) from mammals to fish. Details on the methodology are included in parenthesis below type of mutation in the mutation column. Conditional knockout of the nr3c2 gene have been recently reviewed by (Baker & Katsu, 2017). Gene/Target Species Mutation Phenotype Source Nr3c1 (glucocorticoid Mice (Mus Ubiquitous 90% of mice die at birth (Cole et al., 2001) receptor) musculus) due respiratory insufficiency. Increased ACTH and corticosterone Ubiquitious Impaired transcriptional (Oitzl et al., 2001; GRdim/dim activation, can still engage Tronche et al., 2004; in protein-protein Kleiman et al., 2012) interactions (i.e. STAT5) Necessary for suppression of interleukin1 in macrophages Impaired spatial memory Increased corticosterone Liver Double mutant mice (Mueller et al., 2011) Cre-loxP (driver: (GR/STAT5) developed albumin/α- liver steatosis, and cancer fetoprotein-Cre) Liver Reduced body size (Opherk et al., 2004; Cre-loxP (driver: Impaired GH signalling Tronche et al., 2004) albumin/α- through STAT5 fetoprotein-Cre) Liver Human GRBβ did not (He et al., 2016) Cre-loxP (driver: repress PEPCK albumin-Cre) Liver Increase fat mass, liver (Bose et al., 2016) Cre-loxP (driver: glycogen albumin-Cre) Increase kidney expression of

187

gluconeogenic and glycogen metabolism genes Muscle Increased muscle mass (Shimizu et al., 2015) Cre-loxP (driver: Reduced adipose tissue ACTA1-Cre) Impaired protein turnover, (De Theije et al., 2018) prevention of muscle atrophy during pair- feeding stress. Attenuation of the dex- (Watson et al., 2012) induced regulation of atrophy markers (MuRF1, MAFbx). Adipose No changes in glucose (Bose et al., 2016) Cre-loxP (driver: tolerance or insulin adiponectin-Cre) sensitivity Macrophage: Increased mortality due to (Galuppo et al., 2017) Cre-loxP cardiac rupture (driver: LysM-Cre) Macrophage Increased death with LPS (Bhattacharyya et al., Cre-loxP injection, impaired p38 2007) (driver: LysM-Cre) MAPK signalling Chondrocytes Impairment of fracture (Rapp et al., 2018) rtTA inactivation, healing. Initial increased Cre-loxP (driver: local inflammatory ESR1-Cre) response, but impaired endochondral ossification Forebrain: Alterations in despair-like (Kolber & Muglia, 2009) Cre-loxP behaviour (driver:CaMKII- Hyperactivity of the HPA (Boyle et al., 2005) Cre) axis, impaired feedback regulation of the HPA axis, Increased depressive behaviour

188

Noradrenergic no alterations in terms of (Chmielarz et al., 2013) neurons: normal cage behavior, Cre-loxP weight gain, spatial (driver: dopamine memory or spontaneous beta-hydroxylase locomotor activity (DBH)-Cre) Dopamine neurons: Essential to build (Parnaudeau et al., 2014) Cre-loxP amphetamine-induced (driver: D1 and conditioned place DAT-Cre) preference Distal Nephron Elevated blood pressure, (Goodwin et al., 2010) Cre-loxP but similar hypertensive (driver: Ksp response to dex cadherin-Cre). Zebrafish Ubiquitous: Larger size, decreased Faught and (Danio rerio-TL CRISPR/Cas9 muscle wasting and Vijayan,2018/ Chapter 2 strain) (-7bp, exon2) increased muscle glucose uptake Zebrafish Ubiquitous: Increased mortality during (Facchinello et al., 2017) (Danio rerio- CRISPR/Cas9 embryogenesis. Adults: AB strain) (-11bp, exon2) increased fat deposits, Ubiquitous: ENU Hypercortisolemic (Griffiths et al., 2012; (s357) hyperactive HPI axis and Ziv et al., 2013) altered behaviour Nr3c2 Mice (Mus Ubiquitous: HDR 8 days old: hyperkalemia, (Berger et al., 1998) (mineralocorticoid musculus) of exon 3 (DNA- hyponatremia, and a receptor) binding domain) strong increase in renin, angiotensin II, and aldosterone plasma concentrations. 10-days old: death Renal Tubule: Cre- Survival to adulthood, (Ronzaud et al., 2007) loxP (driver: Aqp2- elevated aldosterone Cre)

189

Renal Nephron: Adulthood: reduced (Terker et al., 2016) Cre-loxP (driver: weight, salt wasting, low Pax8-Cre) blood pressure, hyperkalemia Renal Nephron: Adulthood: reduced (Canonica et al., 2016) Cre-LoxP with weight, Symptoms of rtTA inactivation pseudohypoaldosteronism (driver: Pax8-Cre) Neurons: Cre-loxP Impair learning, altered (Berger et al., 2006) (driver: CamKII- memory and anxiety Cre) Vascular smooth Decreased blood pressure, (McCurley et al., 2012) muscle cells (SMC) altered vascular tone with Cre-loxP (driver: aging SMA-Cre) Endothelial Cells Nitric oxide function (Rickard et al., 2014) Cre-loxP (driver: impaired in the thoracic Tie2-Cre) aorta/mesenteric arteries Endothelial Cells Prevented aortic stiffening (Jia et al., 2016) Cre-loxP (driver: due to a western diet VECad-Cre) Prevented obesity induced (Schäfer et al., 2013) endothelial dysfunction Cardiomyocytes: Improved infract healing (Fraccarollo et al., 2011) Cre-loxP (driver: and prevented adverse MLC-Cre) cardiac remodelling Cardiomyocytes: Increased transforming (Rickard et al., 2012) Cre-loxP (driver: growth factorB/connective MLC-Cre) tissue growth factor

Macrophages Decreased expression of (Rickard et al., 2012) Cre-loxP (driver: inflammatory genes, LysM-Cre) protected against DOC macrophage recruitment MR is required for (Usher et al., 2010) macrophage polarization

190

Protects against cardiac hypertrophy, fibrosis and vascular damage Increase blood pressure (Bienvenu et al., 2012) due to nitric oxide deficiency Increased collagen (Shen et al., 2017b) coverage, reduced cholesterol-induced atherosclerosis Reduced activated (Frieler et al., 2011) macrophages at the ischemic core. Suppression of MR macrophage markers Medaka Ubiquitous: Changes in behaviour (Sakamoto et al., 2016) (Oryzias latipes TALENS (-2 bp) - Hd-rR strain) Double knockout Mice (Mus Epidermal Birth: epidermis displayed (Bigas et al., 2018) (MGR) musculus) defective differentiation and inflammation Adulthood: increased susceptibility to inflammation

191

Table 2: Exon regions of nr3c1 gene (Source: ENSEMBL)

Exon / Intron Start End Length Sequence 5' upstream sequence ...... agatcacactctctcacccaatgaagtcactgtaataggaagcgagatgc TAACTCTGTACTGCCAATAGGTCGCACTGCATGTTGTAA ACAGGGTCGCCGCGGACCTTTCTACTGTAATCTCCATTT TGAATGTTTTTATGGTTTAGGGTTTTGTGGCACGCCTCTT TTCTCATGGCCGAAAACGAGCAAATCGCCTGGATTACC CAGATGGTGTTATGAATTCGGACTAAAAACGCTGGACG CGATCTGAAGTGCAATCTACATTACGCAGTTTGCGATTT TTCTTGCTCGTCAGAAACGACCTCAGACTGCGGCGAAG GCTTTCATGTTGTTGCTAATCGCTCAGCAAAACAATGTC 23,801,389 23,801,036 AGATGATGCAACAAATCAAAACAGGAACGTAAATATCA Exon 1 354 AAAG

Intron 23,801,035 23,799,350 1-2 1,686 gtacgtattctctttcgaccttaaa...... aatgttgtgttgtgttttaccacag

TGCAAAATGGATCAAGGAGGACTGGAGAATGGCAAAA AGCGAGATGAGCGTTTAAATACATTGGATTACAACAAA CGAGCAACTGAGGGGATATTACCAAGAAGAATACAAA GCACCATGTCTGTGGCCCCTACATCTATGGTTCCTCAAG CCGGTCCAATGATGCAGCCAGTTTCTGGGGACATTCCCA ATGGCCTGAGCAATTCGCCCACTCTGGAGGAGCACACC AGCTCGGTGTCCTCCATCTTTGGCGACGATTCCGAACTC AAACTGCTTGGGAAGGAGCAGAGGGCCCTGCAACAGCA GACCTTGGTCCCCTTCACTTTGGGTGACAGTCTTTCAGG TCTGGAGGCCAGCATTGCAGACCTTAACAACCCCTCTCC CTCAATGGATTCCCTGATTGGTGGAGTCGATCCCAATCT TTTCCCCTTAAAAACAGAGGACTTTTCTCCCATGATTAA AGGCGATATGGACCTTGACCAAGATTCCTTTGGACACA TTGGGAAAGATGTTGATGTTGGCAATCATAAGCTCTTTA GTGACAACACTCTGGACCTCCTGCAGGACTTTGAGCTG GATGGATCGCCATCAGACTTCTACGTTGCTGACGATGCG TTTCTCTCCACCATAGGTGAAGATGCTCTCCTTTCAGAG CTGCCGACAAATTTAGACAGGGACTCGAAGGCTGCGGT TTCCGGGAGCAACACGCTCAATGGCACAGCTTCTTCCA GCCTCAGCACAGCCAACACCAGCATCTTGCCCAATATA AAGGTGGAGAAAGACTCTATAATCCAGCTGTGCACCCC AGGGGTCATCAAACAGGAGAACACCGGTGCGAGCTATT GTCAAGGAGGGCTCCACAGCACCCCCATTAACATATGC GGGGTCACCACTTCAAGCGGACAGAGCTTCCTCTTTGG GAACAGCTCGCCCACAGCTGTCGTCGGTCTGCAGAAAG ATCAGAAGCCGGACTTTAACATGTACACCCCTCTGACCT CTTCAGGAGATGGTTGGAGCAGGAGCCAGGGCTTTGGG AATGTCAGTGGAATGCAGCAGAGGGCCAGTTTATGCTT Exon 2 23,799,346 23,798,241 1,109 TTCCAAAAACTTCTCAAGCAGCCCCTATTCCAG Intron 2-3 23,798,240 23,765,884 32,357 gtaagccatcaagaccaccgaaatg...... agatagatatttctgattgtttcag GCCTGAAGACAGCACTGCCACATCGTCAGCTGGTGGAA Exon 3 23,765,883 23,765,735 149 AGACTGGCACACACAAGATCTGTCTGGTTTGCTCAGAT

192

GAGGCTTCTGGCTGCCATTATGGAGTTCTCACTTGCGGA AGCTGTAAAGTCTTCTTTAAAAGGGCTGTTGAAG Intron 3-4 23,765,734 23,763,413 2,322 gtattccatatatctctggcattaa...... cattacctttttttgtttttcatag

GGCAACACAATTACCTGTGTGCTGGGCGAAACGATTGC ATCATTGACAAAATCCGCCGGAAGAACTGCCCTGCCTG TCGTTTCCGCAAATGCCTCATGGCAGGCATGAACCTAG Exon 4 23,763,412 23,763,296 117 AGG Intron 4-5 23,763,295 23,756,865 6,431 gtaacatgaaatcttgcatttatta...... ctattttaataatttctctgctcag

CACGTAAAAGCAAGTCAAAGGCACGTCAAGCCGGAAA GGTAATCCAGCAGCAGTCCATTCCGGAGCGCAACCTTC CTCCACTGCCGGAGGCCCGCGCACTGGTGCCCAAACCA ATGCCTCAGCTGGTGCCCACTATGCTATCACTACTAAAG GCCATTGAGCCAGACACCCTCTATGCCGGATATGACAG CACTATACCAGACACATCCGTCCGCCTCATGACTACGTT GAACAGGCTGGGTGGCCGGCAGGTCATCTCTGCTGTCA Exon 4 23,756,864 23,756,577 288 AATGGGCTAAAGCTCTGCCAG Intron 5-6 23,756,576 23,756,490 87 gtacgaattgtgatgtctttaaaaa...... ataccctctgcttcatgtattttag GGTTCCGGAACCTTCATCTGGATGATCAGATGACCCTCC TGCAGTGTTCCTGGCTGTTCATCATGTCTTTTGGACTGG GCTGGAGGTCCTACCAGCACTGTAACGGAAACATGCTG Exon 5 23,756,489 23,756,345 145 TGTTTCGCTCCAGACCTGGTGATCAATGA Intron 6-7 23,756,344 23,752,704 3,641 gtatgtatctaattaaatgagttac...... aagggtcatgtgtgtctgtccacag AGAGAGGATGAAGTTGCCCTACATGAGTGATCAGTGTG AACAGATGTTGAAGATCTCCAATGAATTTGTGCGACTG CAAGTGTCCACTGAGGAGTATCTGTGCATGAAAGTCCTT Exon 6 23,752,703 23,752,573 131 CTGCTCCTGAACACAG Intron 7-8 23,752,572 23,752,472 101 gtgaacttctctagatacctcttta...... ttgcatgatttttgttgtttctcag TGCCGAAGGATGGGTTGAAGAGCCAGTCAGTGTTTGAT GAACTACGAATGTCTTACATTAAAGAATTGGGCAAGGC TATTGTGAAGAGAGAGGAGAACTCCAGCCAGAACTGGC AACGGTTCTATCAGCTCACTAAGCTACTGGACTCCATGC Exon 7 23,752,471 23,752,314 158 ACGAC 23,752,313 23,748,720 3,594 gtaatgatgcttctcctaaataatg...... ctaacacctctctttatctctgcag

TTGGTGGGTGGACTCCTGAACTTCTGCTTCTACACCTTT GTGAATAAATCTCTGAGCGTGGAGTTTCCCGAAATGCTT GCAGAAATCATCAGCAACCAGTTACCAAAATTCAAAGA TGGGAGTGTTAAACCGCTGCTCTTTCACCAGAAATGAAT Exon 8 23,748,719 23,748,552 168 CTCCTCCCAACAG

3' downstream sequence caggacacgatgccttaaatatcccctcactcccctccacccgacccgaa......

193

Table 3: Exon regions of nr3c2 gene (Source: ENSEMBL)

Exon / Intron Start End Length Sequence

5' upstream sequence ...... cacaatgacgtcgctacgcaagtctgtctgagtcctcggctgtttttatt

AGTCTTCAGTTGGGCGCCTTCAACAATCTAGTAGCAATGTGCACGTAAACAGAAAAGTGTCGACAGGACGT TTCGCTCAAGTCGCTGTCGACTGTTGTGTGCGCGTACAGTGCGCTGTCGTTAAGGAGCTCGCGTCATGCGC Exon GCGCCAGTTTGCGGTGCGTTCAGCTGTCACAACTGAAACTACATGCAAGCAGTCGCAACACTGGGAATACT 37,087,966 37,087,654 TCTTAACACTACTTAACCTGTCATTTTTCTCCGTTTTGTGCATTGGATTACTAGTGGCGAGCAGTAGAGGTAA 1 313 CTGTCAAAATCGGATGTCGTAATTACG Intron 37,087,653 37,084,422 1-2 3,232 gtaagatgtcttaatattatcatta...... tgtctctcatttattggtccattag

GTTTATCAAATGACCCAGGTATGGAGACTAAAAGATACCAAAGTTGCTGCGAAGGGGCGAAAGCAGAAAA CAAATGGGCACAGATGCCAAACACCGCGGACTATATCTGCTCTGCAGAAGAGAGTTTGACAAACAGTGAT GTGCTCATGGATATTGTTAACTCGAGCAATACCCCTAATGTGCAATCTATCTGCAAGGACAACAACTTTAAG AATACAGAGGCAACAATGATTAGAGTTAACCAGAATCAACCATTACCTTTCCCCACCTTCAACAACTGCTTC AACAACCGCAAGTCAGAAACAGACTCAAAGGAACTTTCCAAAACAGTTGCAGAGTCAATGGGCTTGTACAT GAATGCTGCCAGAGAGGCAGACTTTGGCTTTTCCCAACAAGGTGCTGCGGGAGGGCAGGGTAGCCCAGG GAAACTGTACCCTCTTTGTGGGAGAACAAATGAAGACAGCCAGGCCAAAACAAGTGGGAGCCCAAAGATG AAAGCTCCCCCTGCTTCATTCCCACCTGGGGCCCAACTACCCAATGGTGGACCTCAGGAGTGTGCTGTGGT CTCTGACTCAGTGCCCTCTGCTCTGGTCACTGCCCTTTCGTCCAGCACTGATGGGTCCTGTCCCATGTCTAGT CCTACCGGACACAACATGGTATCATCCACCACCAGCCCCACTTACTTTGACTCTGATTGTCCGACGCTAGAT TCTGCGACCAGCAGCTTGACCCATTGTCAACACACCAGCCCCAACATTTGCAGCCCAGTAAAGTCCAGCATT GTGGGGTCACCTCCACTGCCTAGTCCCCTCAGTGTAATGAAGTCTCCGGTTTCCAGTCCTCATAGTATAGGC AGCGTGAGGTCACCGCTTTCCTGTAACACCAACATGAGGTCATCTGTCTCAAGCCCAACAACAAATGGAGG CAATACCTGTAATATTAAACCATCCATTTCCAGCCCGCCCACTGCTGGCAGCATGTCCATGTCCAGTCCCAG GAACTCATCCAGGGGTTTCTCAGTCTCCAGCCCACCCAGTGGATTGGGCCTAGTGCAAAATGATGTCAACA GCCCTGAGAGCCGAGAGCACGACTTCAAGGGATTCGAGTTCCCGAAGGTAGAGAATGTGGATGGGGAGA TCTTTAACATCGGCTTGGATGCTATGGGAGTGGCTAAATATATCAAAAATGAACCTGGCACTGATTTCAGG AGTATGTGCCTAGGCAGCAGCAAGAGTGCCATGTCAAATTCACCTTTCGTAACTCACATTAAGACCGAGCC AAACAGAGAGGTGACTTGCTCGAACCTCCAGTTTGCTGAACCGCAGCACTCTCTGGGCTGCTTTCCTTCGAC AGAGACCACATACTTGTCTTTGAGGGATAATATTGACGAATATAGTCTTTCTGGGATCTTGGGACCTCCTGT TTCGTCTCTGAATGGCAACTATGAACCTGGTGTGTTTCCTAACAATGGCCTGCCCAAGGGGATTAAACAGG AAACCAATGATGGCAGCTATTACCAAGAGAATAATAATGTGCCCACTTCGGCTATTGTTGGCGTTAATTCA GGTGGACATTCATTTCATTACCAGATTGGAGCGCAAGGAACAATGTCGTTTTCACGCCACAATTTGAGGGA CCAGACAAACCCCTTGTTGAATCTAATTTCTCCAGTTACTGGATTAATGGAGACGTGGAAAACTCGCCCAG

37,084,421 37,082,639 GCCTGTCACAGGGGCCCCTCACTGCTAGAGGGGACGGGTATCCAGGCGCAGTCTGCCTTACAGAAAACAT Exon2 1,783 GGAAAG Intron 37,082,638 37,000,508 2-3 82,131 gtaagacaaaagaaggccattttta...... tggaatttatttgctcttatttcag Exon 37,000,507 37,000,380 TGCGTCGCTGAGGCACACGTCTTCGACAGCCAAAGTGTGTCTGGTGTGTGGAGACGAGGCGTCAGGATGC 3 128 CACTACGGGGTTGTTACATGTGGAAGTTGCAAAGTTTTCTTCAAAAGAGCGGTAGAAG Intron 37,000,379 36,969,183 3-4 31,197 gtaagactcatctgtttcctcactt...... tgtttttacactttctcttttgcag Exon 36,969,182 36,969,066 GTCAGCACAACTACCTTTGTGCTGGGAGGAATGACTGCATCATTGACAAGATTCGGCGGAAAAACTGTCCG 4 117 GCCTGCAGAGTACGCAAGTGTCTACAAGCTGGGATGAATCTTGGGG Intron 36,969,065 36,963,551 4-5 5,515 gtaagtgaatctcattactatgaga...... atgtacatttgctttttttactcag

CACGGAAGTCGAAGAAGTTGGGGAAAATAAAGAGCATTAGTGAAGATTCTTCCCTGCAGAGCTCAAAGGA TGGCCCGTTTTTGACATCCGAAAAGGAGTTGAGTTCCTCCACCGCTCTCGTACCACACACCCCGACGGTCGC Exon CCCATATCTGACGCCGTCCGTCTGCAGCGTTCTGGAGCTCATAGAGCCAGAAGTGGTGTTCGCTGGCTATG 36,963,550 36,963,236 ATAACACACAACCTGACACAACAGACCACCTGCTTACCAGCCTCAACCAGCTCGCCGGAAAACAGATGATC 5 315 CGAGTGGTCAAGTGGGCCAAAGTACTTCCAG Intron 36,963,235 36,957,446 5-6 5,790 gtatactcaaactttcttatgaaaa...... acttgccctttatatttttccccag Exon 36,957,445 36,957,301 GTTTCCGCAGTCTACCCATTGAGGACCAAATCACGCTGATCCAGTATTCCTGGATGTGTCTCTCCTCCTTCTC 6 145 CCTCAGCTGGAGATCCTACAAACACACCAACGCCCAAATGCTCTACTTCGCCCCGGATCTAGTCTTCAACGA

194

Intron 36,957,300 36,953,002 6-7 4,299 gtaagataattcacaaaacgctcaa...... gactctttctttctctcctatatag Exon 36,953,001 36,952,871 GGAGCGAATGCAGCAGTCTGCCATGTATGATCTGTGTGTGGGGATGCGGCAGGTGAGCCAGGAGTTTGTT 7 131 CGTCTTCAGCTCACATATGAGGAATATCTGGCCATGAAGGTTCTGCTGCTGCTCAGCACAG

Intron 36,952,870 36,938,362 7-8 14,509 gtgagactgtgtgcactgtgtccta...... taacatgcttttcttttgctaacag

Exon TGCCCAAAGATGGTCTGAAGAACCAAGGGGCATTTGAGGAGATGCGGGTGAATTATATTAAGGAGCTCCG 36,938,361 36,938,204 CCGCTCTGTTGGGAAAGCCACTAACAATTCTGGCCAAACATGGCAGCGCTTCTTCCAGCTGACAAAACTTCT 8 158 GGACGCCACGCATGAT Intron 36,938,203 36,935,604 8-9 2,600 gtaagtcaaaacgatcctctgtttt...... tctgactcctgacatctctttccag

CTTGTTCGGAGCCTGTTGGACTTTTGTTTCTACACTTTCCGCGAATCCCAGGCTCTGAAGGTTGAGTTTCCCG AGATGCTGGTGGAAATTATCAGCGACCAGATACCAAAGGTGGAGTCGGGTCAAACACACACTCTCTACTTT CATAAGAAATGACTGAAAACTCACGAAAGTAGGGGAAAACGAAGGATCAAATGAGAGGAAAAGCAGAG GACACACAGCCCTCGGAGAGCGAACAAGAGCAAGCCCGCTATGCACAAAGCAAGGCGTAATATGAGGCTT Exon GGTTTAAAGCCATATAAGGCTTGTGTAGTGAGGGAAGGCAAAGAGAGATCTCTGTACCCGGCCAACGGAG 36,935,603 36,935,154 AAGAACTCGAATTTCATGTGACAGTGCGTCCCTTACCCCCAAAGACAATGCTGACAATGTTCTATATGGATA 9 450 TGAAGCCATATTCTTTGGTAAAGCTA 3' downstream sequence tagtgcacagtcacatgtgaatacatcacacacagagaaatatgggccca......

195

REAGENTS AND BUFFERS All buffers must be labelled appropriately using i) the bolded titles below ii) the date made, and iii) your initials sgRNA: Phusion High Fidelity DNA polymerase (NEB #M0530S) dNTP (#NEB) T7 High Yield RNA synthesis kit (NEB #E2050S)

Cas9 mRNA: pT3TS-nCas9 plasmid (Addgene #46757) GeneJet Plasmid miniprep kit (#K0502) XbaI (NEB #R0145) Wizard Gel and PCR clean-up kit (Promega #A9281) T3 mMessage mMachine (Ambion Am1348) Zymo RNA clean and Concentrator kit (R1015 or R1016)

Superbroth – 500 ml Tryptone 16 g Yeast Extract 10 g NaCl 2.5 g dH20 up to 500 ml

Measure out 350 ml water to a beaker, add reagents, stir until dissolved. Bring to 500ml with analytical grade water in a graduated cylinder. Pour 250 ml into 2x500 ml bottles. Autoclave the same day. LB Agar Plates – 500 ml (~25 plates) Tryptone 5 g Yeast Extract 2.5 g NaCl 5 g dH20 up to 500 ml

NaCl 3.75 g/250 ml

Measure out 350 ml of water into a beaker, add reagents, stir until dissolved. Bring to 500ml with analytical grade water in a graduated cylinder. Pour 250 ml into 2x500 ml bottles. Add 3.75g of agar to each bottle. Autoclave the same day. Store at room temperature. If making plates immediately cool the agar bottle to 55°C in a water bath. If the agar has solidified, reheat the 500 ml bottle in the microwave (loosen the cap!) and cool to 55°C in a water bath. Working under the flame, add 250 ul of the 1000x ampicillin stock directly to the bottle. Gently poor enough agar broth to cover the bottom of a petri dish. Let cool at room temperature until the agar has solidified. Store at 4°C. Good for ~ 1 month.

196

Ampicillin stock (1000x – 100 mg/ml) Ampicillin 100 mg

Measure out ~100 mg in a 1.5 ml microcentrifuge tube. Adjust water precisely to a final concentration of 100 mg/ml. Add directly to LB Agar plate broth when warm (~55C) to a final concentration of 100 ug/ml.

IDT Oligos: Ultramar: 5’

AAAAGCACCGACTCGGTGCCACTTTTTCAAGTTGATAACGGACTAGCCTTATTTTAA

CTTG CTATTTCTAGCTCTAAAAC-3′

M13F-Hex OR FAM*: TGTAAAACGACGGCCAGT PIGtail: GTGTCTT *must be ordered as a 100 nm for the HEX or FAM option to be available

Table 4: Summary of Oligonucleotides

Gene Primer (5’-3’) Amplicon

Size (bp)

GR_Forward tgtaaaacgacggccagtAACAAACGAGCAACTGAGGG 325

GR_Reverse gtgtcttTTAAGGTCTGCAATGCTGGC

MR_Forward tgtaaaacgacggccagtGGCTTGTACATGAATGCTGCC 168

MR_Reverse gtgtcttGGGCTCCCACTTGTTTTGGCC

Sigma DNA Extraction: Extraction Solution (E7526 – 24 ml) Tissue Preparation solution (T3073- 30 ml) Neutralization solution (N3910- 24ml)

Fragment Analysis: Taq DNA polymerase (ThermoFisher #EP0402)

Sequencing: Exo Sap – IT PCR product cleanup (Affymetrix)

197

Figure 1: pT3TS-nCas9n plasmid. This plasmid was used to create Cas9 mRNA used in this thesis (Source: Addgene #46757)

198

APPENDIX II: SUPPLEMENTAL INFORMATION (CHAPTER 4)

199

Table 1: Possible Targets of MR in 5dpf zebrafish: Possible targets of 5dpf zebrafish that had been treated with 5 µg/ml for 48 hpf. These targets were calculated from the difference in DEG genes of the WT+F group (89 genes) from the MR+F/WT+F group (60 genes). Of the 89 genes regulated by cortisol (F) in WT zebrafish, 60 of these were also regulated by cortisol in MR fish (GR regulated genes – see table 2). There were 29 genes that failed to be regulated in the MR mutants in the presence of cortisol. DEG column denotes genes that were significantly up (UP) or down (DOWN) regulated in the WT+F cortisol group Gene Name DEG (UP/DOWN (p- Notes

value; FDR))

NM_201085 3-hydroxy-3-methylglutaryl-CoA UP (4.50E-08;

synthase 1 6.87E-06)

NM_001128704 pyruvate dehydrogenase kinase, isozyme UP (1.38E-05; MRE (H. sapiens) (-193, -

4 (pdk4) 0.001149593) 403, -426, -604, -817, -879, -

1349)

NM_001017909 apolipoprotein A-IV a (apoa4a) UP (5.63E-05;

0.003887214)

NM_001030062 apolipoprotein Bb, tandem duplicate 1 UP (2.97E-10;

(apobb.1) 5.93E-08)

NM_001309641 E3 ubiquitin-protein ligase RBBP6-like UP (7.65E-06;

6.70E-04)

NM_001013304 aerolysin-like protein (aep1) DOWN (9.13E-04;

0.042977993)

NM_001282005 si:ch211-251f6.6 (si:ch211-251f6.6) UP (8.52E-07;

9.00E-05)

NM_001128719 zgc:193725 (zgc:193725) UP (8.52E-07;

200

9.00E-05)

NM_001003450 zgc:92511 (zgc:92511) UP (1.96E-11;

4.84E-09)

NM_001030196 diacylglycerol O-acyltransferase 2 UP (5.14E-06; MRE (D. rerio): -531

(dgat2) 4.86E-04)

NR_023323 im:7160157 (im:7160157) UP (1.73E-07;

2.22E-05)

NM_200965 ATPase sarcoplasmic/endoplasmic UP (2.33E-11; MRE (D. rerio): -562, -454

reticulum Ca2+ transporting 2a (atp2a2a) 5.49E-09)

NM_001102618 Rh family, C glycoprotein, like 1 (rhcgl1) UP (4.70E-04;

0.02378436)

NM_001242770 glucagon b (gcgb) UP (7.22E-09;

1.25E-06

NM_001145236 secretory calcium-binding UP (6.74E-21;

phosphoprotein 5 (scpp5) 8.07E-18)

NM_001048055 leukocyte cell-derived chemotaxin 2 UP (4.43E-07;

like (lect2l) 4.93E-05)

NM_001145243 odontogenic, ameloblast associated UP (9.20E-06;

(odam) 7.96E-04)

NM_001128754 MIA SH3 domain containing (mia) UP (4.61E-08; MRE (D. rerio): -365

6.96E-06)

NM_001098396 N/A DOWN (3.14E-04; Ribosomal RNA gene

0.017018493)

NM_001004630 sterol-C5-desaturase (sc5d) UP (1.71E-05; MRE (H. sapiens): -1061

0.001379792) MRE (D. rerio): -883

201

NM_183348 sulfotransferase family 1, cytosolic UP (5.87E-06;

sulfotransferase 3 5.44E-04)

NM_001040362 ELOVL fatty acid elongase 2 (elovl2) UP (8.31E-06; MRE (H. sapiens): -131

7.23E-04)

NM_001007054 zgc:92590 (zgc:92590) UP (5.23E-04;

0.0257456)

NM_001082882 type I cytokeratin, enveloping layer, like DOWN (8.47E-15;

(cyt1l) 3.80E-12)

NM_001291498 protein phosphatase 1, regulatory UP(4.02E-05;

(inhibitor) subunit 7 (ppp1r7) 0.002972482)

NM_200894 zgc:77614 (zgc:77614) DOWN (8.26E-08;

1.17E-05)

NM_001128735 transcobalamin beta a (tcnba) UP (4.40E-25;

7.02E-22)

NR_030036 microRNA 29b-1 (mir29b-1) DOWN (2.69E-09;

5.12E-07)

NM_001044836 si:dkey-14d8.7 (si:dkey-14d8.7) UP (0.001068788;

0.049353278)

Table 2: Targets of GR in 5dpf zebrafish: Possible targets GR in 5dpf zebrafish that had been treated with 5ug/ml cortisol for 48 hpf. Targets were calculated as the DEG genes in MR-/- +F and WT+F larvae (MR+F). Total 60 genes Gene Name DEG (UP/DOWN (p-value; Notes

FDR))

NM_001195784 selenoprotein e (selenoe) UP (5.30E-11; 1.23E-08)

NM_001001730 cytochrome P450, family 51 (cyp51) UP (1.72E-11; 4.33E-09)

202

NM_207082 N/A UP (8.82E-04; 0.041808641)

NM_131108 type I cytokeratin (cki) UP (2.00E-14; 7.56E-12)

NM_001100958 heat shock protein, alpha-crystallin- UP (2.20E-11; 5.32E-09)

related, b6 (hspb6)

NM_001025173 carboxyl ester lipase, tandem UP (2.48E-07; 2.92E-05

duplicate 2 (cel.2)

NM_001017724 zgc:112160 (zgc:112160) UP (7.46E-13; 2.28E-10)

NM_001024408 elastase 3 like (ela3l) UP (2.48E-17;1.75E-14)

NM_001025180 six-cysteine containing astacin UP (1.68E-18; 1.51E-15)

protease 4 (c6ast4)

NR_034193 microRNA 24-2 (mir24-2), UP (3.68E-12; 9.98E-10)

microRNA

NM_131645 fatty acid desaturase 2 (fads2) UP (2.43E-15; 1.16E-12)

NM_131098 apolipoprotein Eb (apoeb) UP (1.18E-26; 2.42E-23)

NM_001002217 carboxypeptidase A4 (cpa4) UP (1.61E-26; 2.88E-23)

NM_001077607 immunoresponsive gene 1, like UP (1.36E-04;

(irg1l) 0.008279735)

NM_199886 elastase 2 like (ela2l) UP (1.48E-17; 1.12E-14)

NM_001110021 carboxypeptidase B1 (tissue) (cpb1) UP (3.32E-18; 2.81E-15)

NM_131727 somatostatin 2 (sst2) UP (5.21E-08; 7.63E-06)

NM_213353 methylsterol monooxygenase 1 UP (9.88E-15; 4.05E-12)

(msmo1)

NM_001079977 3-hydroxy-3-methylglutaryl-CoA UP (2.07E-10; 4.37E-08)

reductase a (hmgcra)

NM_001102388 ankyrin repeat domain 1b (cardiac UP (1.00E-24; 1.44E-21)

muscle) (ankrd1b)

203

NM_00104538 N/A UP (1.04E-06; 1.09E-04)

NM_001201351 apolipoprotein Da, duplicate 2 UP (1.31E-14; 5.24E-12)

(apoda.2)

NM_001201446 reticulon 4a (rtn4a) UP (1.86E-05; 0.00147765)

NR_048933 N/A UP (3.68E-12;

9.98E-10)

NM_001045174 hatching enzyme 1, tandem duplicate UP (0.001077534;

1 (he1.1) 0.049597659)

NM_199607 carboxyl ester lipase, tandem UP (1.31E-28; 3.75E-25)

duplicate 1 (cel.1)

NM_199605 serine protease 59, tandem duplicate UP (9.40E-15;

1 (prss59.1) 4.05E-12)

NM_001003737 chymotrypsin like elastase family UP (4.45E-16;

member 1, tandem duplicate 6 2.51E-13)

(cela1.6)

NM_001308608 creatine kinase, mitochondrial 2a UP (7.60E-11;

(sarcomeric) (ckmt2a) 1.73E-08)

NM_213011 amylase alpha 2A (pancreatic) UP (3.99E-18;

(amy2a) 3.18E-15)

NM_001013526 six-cysteine containing astacin UP (3.14E-04

protease 3 (c6ast3) 0.017018493)

NM_001080022 myosin regulatory light chain UP (2.82E-07;

interacting protein b (mylipb) 3.26E-05)

NM_001045282 zgc:136461 (zgc:136461) UP (1.32E-04; 0.008086772)

NM_001020565 apolipoprotein Ea (apoea) UP (8.54E-11;

204

1.89E-08)

NM_001082930 si:ch211-240l19.8 (si:ch211- UP (1.96E-08;

240l19.8) 3.28E-06)

NM_213149 FKBP prolyl isomerase 5 (fkbp5) UP (4.89E-13; 1.53E-10) heterodimer

target

NM_199271 carboxypeptidase A5 (cpa5) UP (2.30E-27;

5.51E-24)

NM_212618 chymotrypsinogen B1 (ctrb1) UP (2.02E-16;

1.21E-13)

NM_199826 N/A UP (5.18E-08; 7.63E-06)

NM_001082928 si:ch211-240l19.6 (si:ch211- UP (2.35E-15;

240l19.6) 1.16E-12)

NM_200509 cytochrome P450, family 2, UP (7.04E-34; 5.06E-30

subfamily K, polypeptide 6 (cyp2k6)

NM_201048 N/A UP (1.08E-16; 7.07E-14)

NR_111927 si:dkey-33c14.3 (si:dkey-33c14.3), UP (1.12E-05;

long non-coding RNA 9.44E-04)

NM_131708 serine protease 1 (prss1) UP (6.25E-16;

3.20E-13)

NR_015620 syncollin, tandem duplicate 2 UP (8.22E-14; 2.81E-11)

(sycn.2), long non-coding RNA

NM_001020482 carboxypeptidase A1 (pancreatic) UP (7.31E-09; 1.25E-06)

(cpa1)

NM_001114342 keratin type 1 c19e (krtt1c19e) UP (6.00E-06; 5.52E-04)

NM_001080090 ornithine decarboxylase antizyme 2b UP (7.07E-09; 1.24E-06)

(oaz2b)

205

NM_198914 sulfotransferase family 2, cytosolic UP (5.09E-09; 9.25E-07)

sulfotransferase 1 (sult2st1)

NM_001007132 interferon alpha inducible protein 45 DOWN (8.12E-05;

(ifi45) 0.005445947)

Table 3: DEG in 5dpf GRKO zebrafish (MR activated genes): Differentially expressed genes in the absence of GR. Targets were calculated as the DEG genes in GR-/- group. Total 51 genes Gene Name Notes

NM_131565 troponin T type 3a (skeletal, fast) (tnnt3a)

NM_001004555 reticulon 4a (rtn4a), transcript variant 1

NM_001002133 tripartite motif containing 63a (trim63a)

NM_001082832 si:dkey-7c18.24 (si:dkey-7c18.24)

NM_001008585 solute carrier family 43 member 2b (slc43a2b)

NM_200212 parvalbumin 3 (pvalb3)

NM_131884 CCAAT enhancer binding protein beta (cebpb)

NM_001002344 zgc:92184 (zgc:92184)

NM_213552 N/A

NM_001110126 ChaC, cation transport regulator homolog 1

NM_201323 serine and arginine rich splicing factor 3b (srsf3b),

transcript variant 1

NM_001045466 N/A

NM_214751 phosphoenolpyruvate carboxykinase 1 (soluble) (pck1)

NM_001045372 zgc:153759 (zgc:153759)

206

NM_001198751 serine peptidase inhibitor, Kazal type 2, tandem

duplicate 1 (spink2.1)

NM_213453 phosphogluconate dehydrogenase (pgd)

NM_001017803 zgc:111983 (zgc:111983)

NM_001102646 zgc:171310 (zgc:171310)

NM_199776 N/A

NM_001008651 prostaglandin reductase 1 (ptgr1)

NM_001115103 chemokine (C-C motif) ligand 39, duplicate 3 (ccl39.3)

NM_001013312 si:dkey-19a16.2 (si:dkey-19a16.2)

NM_001319158 transcription factor AP-2 alpha (tfap2a)

NM_001020532 complement factor D (adipsin) (cfd)

NM_001030101 retinol dehydrogenase 5 (11-cis/9-cis) (rdh5)

NM_001020668 troponin I4b, tandem duplicate 3 (tnni4b.3)

NR_120315 wu:fc47e11 (wu:fc47e11), long non-coding RNA

NM_176859 transcription factor AP-2 alpha (tfap2a), transcript

variant 1

NM_001135971 lecithin retinol acyltransferase b, tandem duplicate 2

(lratb.2)

NM_001114705 heat shock protein, alpha-crystallin-related, 9 (hspb9)

NM_001126474 membrane-spanning 4-domains, subfamily A, member

17A.8 (ms4a17a.8)

NM_213361 N/A

NM_001020683 zgc:112332 (zgc:112332)

NM_001144786 si:ch1073-126c3.2 (si:ch1073-126c3.2)

207

NM_001100056 liver-enriched gene 1, tandem duplicate 1 (leg1.1)

NM_001030146 si:ch211-238e22.2 (si:ch211-238e22.2)

NM_001278839 alkaline phosphatase 3 (alp3)

NM_001320090 RWD domain containing 4 (rwdd), transcript variant 1

NM_001114910 aquaporin 8b (aqp8b)

NM_001142583 immediate early response 2a (ier2a)

NM_001128721 si:dkey-63j12.4 (si:dkey-63j12.4)

NM_001127516 heme oxygenase 1a (hmox1a)

NM_001128729 Kruppel-like factor 9 (klf9)

NM_001077456 zgc:153968 (zgc:153968)

NM_001128297 si:dkey-122c11.1 (si:dkey-122c11.1)

NM_200893 adenosine monophosphate deaminase 1 (isoform M)

(ampd1)

NM_200899 myozenin 2a (myoz2a)

NM_001045409 zgc:153284 (zgc:153284), transcript variant 2

NM_212750 JunB proto-oncogene, AP-1 transcription factor subunit

b (junbb)

NM_001004575 phosphofructokinase, muscle a (pfkma)

208

Table 4: DEG in 5dpf MRKO zebrafish. Differentially expressed genes in the absence of GR. Targets were calculated as the DEG genes in GR-/- group.

Gene Name Notes

NM_001111147 N/A

NM_001109863 N/A

NM_001162851 glutathione S-transferase mu tandem duplicate 3 (gstm.3)

NM_001024438 ELOVL fatty acid elongase 8b (elovl8b), transcript variant

NM_131667 GTP cyclohydrolase 2 (gch2)

NM_001195246 si:dkey-205h13.1 (si:dkey-205h13.1), transcript variant 2

NM_213452 solute carrier family 13 member 2 (slc13a2)

NM_001006005 deoxyuridine triphosphatase (dut)

NM_131798 N/A

NM_001136523 slow myosin heavy chain 3 (smyhc3)

209

Table 5: Possible Targets of GR:MR heterodimers in 5dpf zebrafish. Possible targets MR:GR in 5dpf zebrafish that had been treated with 5ug/ml for 48 hpf. Type A targets were classified as DEG in MR-/- and GR-/- mutants. Type B targets were calculated as the DEG genes that were common to GR-/-, MR-/- and cortisol treated MR-/- larvae (MR+F). Gene Name Notes

NM_001115067 N/A A

NM_205569 v-fos FBJ murine osteosarcoma viral oncogene A

homolog Ab (fosab)

NM_130948 growth differentiation factor 3 (gdf3) A

NM_001024124 protocadherin 1 gamma 26 (pcdh1g26) A

NM_212658 protocadherin 1 gamma 22 (pcdh1g22) B

NM_001002480 myeloid-derived growth factor (mydgf) B

NM_001002580 prostate transmembrane protein, androgen B

induced 1 (pmepa1)

NM_200917 F-box protein 32 (fbxo32) B

NM_199523 uncoupling protein 1 (ucp1) B

NM_001105111 si:ch211-121a2.2 (si:ch211-121a2.2) B

NM_001128574 interleukin 21 (il21) B

NM_001104945 zinc finger protein 1156 (znf1156) B

NM_203460 alcohol dehydrogenase 8b (adh8b) B

NM_200996 pyruvate dehydrogenase kinase, isozyme 2b B

(pdk2b)

210

Table 6: Cortisol regulated gene(s) independent of GR and MR: Gene (s) regulated by cortisol in the absence of GR and MR in 5dpf zebrafish that had been treated with 5ug/ml for 48 hpf. (MR+F/WT+F/GR/MR) Gene Name Notes

NM_001281994 serine protease 59, tandem duplicate 2 (prss59.2)

Table 7: Compensatory cortisol action by MR in the absence of GR, or a third corticosteroid receptor: Gene (s) regulated by cortisol in the absence of GR and MR in 5dpf zebrafish that had been treated with 5ug/ml for 48 hpf (MR+F/WT+F/GR). Gene Name Notes

NM_001045488 antifreeze protein type IV (afp4)

NM_001081690 chymotrypsinogen 2-like (LOC562139)

NM_131056 preproinsulin (ins)

NM_001139464 elastase 2 (ela2)

NM_131176 uncoupling protein 2 (ucp2) MRE (D. rerio): -1683, -1654, -

530

NM_001128758 apolipoprotein A-IV b, tandem duplicate 2

(apoa4b.2)

NM_001004582 chymotrypsin-like (ctrl)

NM_001030152 si:ch211-240l19.5 (si:ch211-240l19.5)

NM_001145593 si:ch211-89o9.4 (si:ch211-89o9.4)

211

APPENDIX III: COPYRIGHT PERMISSIONS

212

213

214