An in Vivo Study of GTF2IRD1 Function and Its Contribution to the Physical Features of Williams Beuren Syndrome

An in vivo study of GTF2IRD1 function and its contribution to the physical features of Williams Beuren Syndrome

Cesar P. Canales Martinez.

A thesis in fulfilment of the requirements for the degree of Doctor of Philosophy

School of Medical Sciences, Faculty of Medicine UNSW Australia Sydney Australia

August 2015

THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: CANALES MARTINEZ

First name: CESAR Other name/s: PATRICIO

Abbreviation for degree as given in the University calendar: PhD

School: SCHOOL OF MEDICAL SCIENCES Faculty: MEDICINE

Title: An in vivo study of GTF2IRD1 function and its contribution to the physical features of Williams Beuren Syndrome

Abstract 350 words maximum: (PLEASE TYPE)

Williams-Beuren Syndrome (WBS) is a complex neurodevelopmental genetic disorder caused by a hemizygous deletion involving up to 28 genes on chromosome 7q11.23. Amongst the spectrum of physical and neurological defects of WBS, it is common to find sensorineural hearing loss (SNHL) and a characteristic set of facial features. The gene GTF2IRD1, first discovered in our laboratory, and the adjacent, homologous gene GTF2I, both fall within the WBS deletion region. Haploinsufficiency of the transcriptional regulators they encode is thought to account for the major neurological and craniofacial aspects of the disease. The aim of the work presented here is to drill deeper into the function of GTF2IRD1 by investigating the impact of a targeted Gtf2ird1 mouse deletion on neurological and craniofacial phenotypes using direct physiological and cellular analysis methods. A detailed analysis of Gtf2ird1 expression in the inner ear and the skin was conducted. A comprehensive analysis of hearing capacity in Gtf2ird1 knockout mice was carried out by examining the auditory brainstem response (ABR) and the distortion product of otoacoustic emissions (DPOAE). The role of GTF2IRD1 in the development of facial skin was examined by a detailed investigation of the control of epidermal proliferation, differentiation and barrier function, followed by an RNA-Seq analysis, to identify the molecular mechanisms that underpin the facial abnormalities observed in these mice. The hearing analyses revealed that Gtf2ird1 knockout mice have hypoacusis (a higher hearing threshold) in both assessments, indicating that the principal hearing deficit in the mice can be traced to impairments in the cochlear amplifier. We suggest that similar mechanisms may underpin the SNHL experienced by WBS patients. Analysis of the facial skin in knockout mice revealed that keratinocyte proliferation was augmented, keratinized epidermal layers were disorganized and the barrier function was compromised in facial regions where Gtf2ird1 is normally expressed during development. The RNA-Seq analysis demonstrated that these abnormalities correlate with dysregulated expression of several key factors, including FGF7. These data suggest that GTF2IRD1 is a crucial component of the transcriptional regulation machinery controlling facial skin patterning and differentiation and supports a role for its involvement in the causation of the facial characteristics of WBS.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

30-08-16 …………………………………………………………… ……………………………………..……………… ……….……………………...…….… Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS

ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis. I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation and linguistic expression is acknowledged.’

SIGNED …………………………………………………… CESAR P. CANALES

DATE: 30-08-2016

ACKNOWLEDGEMENTS

I want to express my gratitude to my supervisors Dr Stephen Palmer and Prof Edna Hardeman for allowing me to join to their team and for providing with all the tools and resources to ensure the development of top quality research. Edna, being the head of the Cellular and Genetic Medicine Unit (CGMU) and head of the Anatomy department, makes you a very busy person, yet you were always available when I needed any type of guidance. You have been so supportive throughout my PhD and I cannot find enough words to express the level of gratitude and admiration I have toward you. Steve, I am fully indebted to you. Your advice on both research as well as on my career have been priceless, you have been a tremendous mentor for me and I am also thankful for your invaluable feedback and support during the intense writing process of thesis.

I was fortunate to be part of the CGMU. This is a wonderful place for young scientists to develop surrounded by great people. Thanks for the advice that I received from the CGMU supervisors Dr Anthony Kee, Dr Thomas Fath and Prof Peter Gunning, I really appreciate the time you all have taken to discuss and enrich my work at some stage. To my fellow HDR students Bassem, Cecilia, Nadia, Veronica, Melissa, Teresa, Bin and Iman; thanks for the friendship and for making the work so much enjoyable. Special thanks to Florence Tomasetig, I very much appreciated your enthusiasm for this project and your willingness to help in any possible way.

I would also like to thank to the Translational Neuroscience Facility (TNF), especially to Professor Gary Housley and Dr Ann Wong, who introduced me to the wonderful world of the hearing physiology. Thank you both for the time devoted to this project, for the long sessions of data analysis and manuscript preparation for publication, and also for helping me out in getting ready for conference presentations. Also thanks to Dr Jennie Cederholm and Dr Kristina Forud, who were always available in the TNF lab for providing assistance when I encountered technical difficulties utilising the equipment.

Thanks to our collaborators whose contributions made also possible to develop and interpret part of this research. To Marc Wilkins and Susan Corley from the NSW System Biology Initiative, for their help with the RNA-Seq analysis, to Anthony

Hannan and Monique Howard from the Florey Institute of Neuroscience, for performing the behavioural studies, to Andy Liang from Neuroscience Research Australia for providing assistance with the CLARITY technique in the skin and to Annemiek Beverdam from the Developmental and Regenerative Dermatology Unit for her assistance interpreting some of the skin findings from this project.

I would also like to thank to the physiology teaching department, especially to Dr Nicole Marden and Dr Lesley Ulman. Thanks for the opportunity of being part of the teaching team; I really enjoyed this experience that served as a great distraction from the hard work at the bench.

Last but not least, I want to thank and dedicate this thesis to Paulina, my wife and lab mate. Thanks for your understanding, patience, unconditional love and for being always there having the right answer to all my questions in life. To my family in Chile: my parents Carlos and Sonia and my brother Victor, for their endless support and encouragement to go through this journey. To my son Diego, who has been the light of my life for the past 3 years bringing the strength that I need to overcome any obstacle in life. I could not have completed my research without the support of all these wonderful people.

iii

PUBLICATIONS

PUBLISHED

Canales CP, Wong AC, Gunning PW, Housley GD, Hardeman EC, Palmer SJ. The role of GTF2IRD1 in the auditory pathology of Williams-Beuren Syndrome. Eur J Hum Genet. 2014 doi: 10.1038/ejhg.2014

IN PREPARATION

Canales CP, Corley SM, Kaur P, Smyth I, Carmona-Mora P, Beverdam A, Wilkins

MR, Hardeman EC, Palmer SJ. GTF2IRD1 controls epidermal cell proliferation and facial skin patterning in development.

Corley SM, Canales CP, Carmona-Mora P, Hardeman EC, Wilkins MR, Palmer SJ.

RNA-Seq analysis shows Gtf2ird1 knockout affects genes involved in epidermal and neurological development.

CONFERENCE ABSTRACTS

2012: Analysis of craniofacial and hearing defects in a mouse model of Williams- Beuren syndrome. (Oral Presentation) Cesar P. Canales, Ann C.Y. Wong, Gary D. Housley, Peter W. Gunning, Edna C. Hardeman and Stephen J. Palmer. Australian Society for Medical Research, Medical Research Week NSW Scientific Meeting, Sydney, NSW, Australia.

2012: The role of GTF2IRD1 in the Williams-Beuren syndrome cognitive profile. Hardeman EC, Howard M, Canales CP, Widagdo J, Taylor KM, Carmona-Mora P, Wong AC, Housley GD, Hannan AJ, Gunning PW and Palmer SJ. ComBio 2012, Adelaide, SA, Australia.

2012: The role of GTF2IRD1 in the auditory pathology of Williams-Beuren syndrome. Cesar P. Canales, Ann C.Y. Wong, Gary D. Housley, Peter W. Gunning, Edna C. Hardeman and Stephen J. Palmer. 62th Annual Meeting, The American Society of Human Genetics. San Francisco, CA, United States of America.

2012: Soft tissue aspects of the Williams-Beuren syndrome facial phenotype can be attributed to GTF2IRD1. Palmer SJ, Canales CP, Carmona-Mora P, Kaur P, Gunning PW and Hardeman EC. 62th Annual Meeting, American Society of Human Genetics. San Francisco, CA, United States of America.

2013: Lack of gtf2ird1 causes a hyperproliferative facial skin phenotype similar to facial abnormalities in williams-beuren syndrome. Cesar P. Canales, Pritinder Kaur, Ian Smyth, Peter W. Gunning, Edna C. Hardeman and Stephen J. Palmer. Australian Society for Medical Research, Medical Research Week NSW Scientific Meeting, Sydney, NSW, Australia.

2013: A mouse model of Williams-Beuren syndrome displays a similar hearing defect to humans. (Oral Presentation) Canales CP, Wong ACY, Housley GD, Gunning PW, Hardeman EC, Palmer SJ. Australasian Auditory Neuroscience Workshop, Melbourne, Vic, Australia.

2013: GTF2IRD1, a gene implicated in the auditory pathology of Williams-Beuren syndrome. Canales CP, Wong ACY, Housley GD, Gunning PW, Hardeman EC, Palmer SJ. Australian Neuroscience Society Meeting, Melbourne, Vic, Australia.

2013: The role of the transcriptional regulator GTF2IRD1 in craniofacial abnormalities. Palmer SJ, Canales CP, Carmona-Mora P, Tomasetig F, Kaur P, Smyth I, Gunning PW, Hardeman EC. 34th Lorne Genome Conference, Lorne, Vic, Australia.

2013: A novel epigenetic regulator of epidermal proliferation in specific craniofacial zones. Palmer SJ, Canales CP, Carmona-Mora P, Tomasetig F, Kaur P, Smyth I, Gunning PW and Hardeman EC. Australian Society of Dermatology, Sydney, NSW, Australia.

2013: GTF2IRD1 is an epigenetic regulator of gene expression important in developmental facial skin patterning. Cesar P. Canales, Susan Corley, Pritinder Kaur, Ian Smyth, Marc Wilkins, Edna C. Hardeman and Stephen J. Palmer. The Genetics Society of AustralAsia Conference. University of New South Wales, Sydney, NSW, Australia

2013: Lack of GTF2IRD1 causes a hyperproliferative facial skin phenotype similar to facial abnormalities in Williams-Beuren syndrome. Cesar P. Canales, Pritinder Kaur, Ian Smyth, Peter W. Gunning, Edna C. Hardeman and Stephen J. Palmer. The Genetics Society of AustralAsia Conference. University of New South Wales, Sydney, NSW, Australia.

2013: Studying the role of GTF2IRD1 in neuropathology of a mouse model of Williams-Beuren Syndrome. Cecilia CR. Chang, Cesar P. Canales, J. Power, Anthony J. Hannan, Peter W. Gunning, Edna C. Hardeman and Stephen J. Palmer. Brain Sciences UNSW Symposium 201, University of New South Wales, Sydney, NSW, Australia.

2013: GTF2IRD1 is a transcriptional regulator important in epidermal cell proliferation and developmental facial skin patterning. (Oral Presentation) Cesar P.

Canales, Susan Corley, Pritinder Kaur, Ian Smyth, Marc Wilkins, Edna C. Hardeman and Stephen J. Palmer. COMBIO 2013, Perth, WA, Australia.

2013: GTF2IRD1 is an epigenetic regulator involved in facial skin patterning that may underpin facial abnormalities of Williams-Beuren syndrome patients. Cesar P. Canales, Susan Corley, Pritinder Kaur, Ian Smyth, Marc Wilkins, Edna C. Hardeman and Stephen J. Palmer. American Society of Human Genetics ASHG, Boston, MA, United States of America.

2013: Analysis of GTF2IRD1’s role in the craniofacial and neurological features of Williams-Beuren syndrome indicates an epigenetic control function. S. J. Palmer, C. P. Canales, P. Carmona-Mora, F. Tomasetig, J. Widagdo, A. C. Y. Wong, G. D. Housley, P. Kaur, I. Smyth, A. J. Hannan, P. W. Gunning, E. C. Hardeman. American Society of Human Genetics ASHG, Boston, MA, United States of America.

2014: The molecular neuroscience of GTF2IRD1: a gene implicated in social interaction. Palmer SJ, Carmona-Mora P, Canales CP, Tomasetig F, Chang C, Cha AY, Widagdo J, Howard M, Hannan AJ, Gunning PWand Hardeman EC. Australian Neuroscience Society Meeting, Adelaide, SA, Australia.

2014: The role of GTF2IRD1 in developmental aspects of Williams-Beuren syndrome. (Oral Presentation) Canales CP, Carmona-Mora P, Chang C, Corley S, Kaur P, Smyth I, Wilkins M, Wong ACY, Housley G, Gunning PW, Hardeman EC and Palmer SJ. NSW Cell and Developmental Biology Meeting. UNSW Australia.

2014: Cellular neuropathology associated with cognitive and behavioural dysfunction in a mouse model of Williams-Beuren syndrome Chang CCR, Canales CP, Power J, Hannan AJ, Hardeman EC and Palmer SJ. ICON 2014 - The 12th International Cognitive Neuroscience Conference - Brisbane, QLD, Australia.

2014: The role of Gtf2ird1 in the facial abnormalities of Williams-Beuren syndrome patients. (Oral Presentation) Cesar P. Canales, Susan Corley, Annemiek Beverdam Pritinder Kaur, Ian Smyth, Marc Wilkins, Edna C. Hardeman and Stephen J. Palmer. COMBIO 2014, Canberra, ACT, Australia.

vii

2014: Unravelling epigenetic complexes associated with GTF2IRD1 to understand the cognitive features of Williams-Beuren syndrome. Paulina Carmona-Mora, Florence Tomasetig, Cesar P. Canales, Abdullah Alshawaf, Mirella Dottori, Edna C. Hardeman and Stephen J. Palmer. COMBIO 2014, Canberra, ACT, Australia.

2014: RNA-Seq analysis of lip tissue from a Mouse Model of Williams-Beuren Syndrome. Susan Corley, Cesar P. Canales, Edna C. Hardeman, Stephen J. Palmer, Marc Wilkins. AGTA - Australian Genomics Technologies Association Conference – Melbourne, VIC, Australia.

2014: Involvement of GTF2IRD1 in the complex hearing phenotype of Williams- Beuren Syndrome. Canales CP, Wong ACY, Housley GD, Gunning PW, Hardeman EC, Palmer SJ. American Society of Human Genetics ASHG, San Diego, CA, United States of America.

2014: Defining the presence of GTF2IRD1 in epigenetic complexes as a means to understand features of Williams-Beuren syndrome. Paulina Carmona-Mora, Florence Tomasetig, Cesar P. Canales, Abdullah Alshawaf, Mirella Dottori, Edna C. Hardeman1 and Stephen J. Palmer. American Society of Human Genetics ASHG, San Diego, CA, United States of America.

viii

ABSTRACT

WBS, it is common to find sensorineural hearing loss (SNHL) and a characteristic set of facial features. The gene GTF2IRD1, first discovered in our laboratory, and the adjacent, homologous gene GTF2I, both fall within the WBS deletion region.

Haploinsufficiency of the transcriptional regulators they encode is thought to account for the major neurological and craniofacial aspects of the disease.

The aim of the work presented here is to drill deeper into the function of

GTF2IRD1 by investigating the impact of a targeted Gtf2ird1 mouse deletion on neurological and craniofacial phenotypes using direct physiological and cellular analysis methods. A detailed analysis of Gtf2ird1 expression in the inner ear and the skin was conducted. A comprehensive analysis of hearing capacity in Gtf2ird1 knockout mice was carried out by examining the auditory brainstem response (ABR) and the distortion product of otoacoustic emissions (DPOAE). The role of

GTF2IRD1 in the development of facial skin was examined by a detailed investigation of the control of epidermal proliferation, differentiation and barrier function, followed by an RNA-Seq analysis, to identify the molecular mechanisms that underpin the facial abnormalities observed in these mice.

The hearing analyses revealed that Gtf2ird1 knockout mice have hypoacusis

(a higher hearing threshold) in both assessments, indicating that the principal hearing deficit in the mice can be traced to impairments in the cochlear amplifier. We suggest that similar mechanisms may underpin the SNHL experienced by WBS

patients. Analysis of the facial skin in knockout mice revealed that keratinocyte proliferation was augmented, keratinized epidermal layers were disorganized and the barrier function was compromised in facial regions where Gtf2ird1 is normally expressed during development. The RNA-Seq analysis demonstrated that these abnormalities correlate with dysregulated expression of several key factors, including FGF7. These data suggest that GTF2IRD1 is a crucial component of the transcriptional regulation machinery controlling facial skin patterning and differentiation and supports a role for its involvement in the causation of the facial characteristics of WBS.

ABBREVIATIONS

ABR auditory brainstem response AD autosomal dominant Ade adenine ADHD attention deficit hyperactivity disorder ASD autism spectrum disorder BAEPs brain stem auditory evoked potentials BAERs brain stem auditory evoked responses BCA bicinchoninic acid assay BSA bovine serum albumin CD complete deletion CNV copy number variation DBD disruptive behaviour disorder DEGs differentially expressed genes DD distal deletion DNA deoxyribonucleic acid DNAse deoxyribonuclease D/P distal/proximal deletion mice DPOAE distortion-product otoacoustic emission tests ECL enhanced chemiluminescence EMSA electrophoretic mobility shift assay FGF fibroblast growth factor FGFR fibroblast growth factor receptor GTF2I general transcription factor II i GTF2IRD1 GTF2I repeat domain containing 1 GUR GTF2IRD1 upstream region His Histidine IFE interfollicular epidermis Kap keratin associated proteins LCR low copy repeat Leu leucine Lys lysine

LZ leucine zipper MRI magnetic resonance imaging NAHR non-allelic homologous recombination NGS next generation sequencing NLS nuclear localisation signal ODD oppositional deﬁant disorder OMIM online mendelian inheritance in man ORF open reading frame PBS phosphate-buffered saline PCR polymerase chain reaction PD proximal deletion PFA paraformaldehyde RTqPCR reverse transcription quantitative real time PCR RD repeat domain RNA ribonucleic acid SDS-PAGE sodium dodecyl sulphate polyacrylamide gel electrophoresis SEM standard error of the mean SVAS supravalvular aortic stenosis TAE tris-acetate-EDTA buffer Trp tryptophan UV ultra violet WBS Williams-Beuren syndrome WSCP Williams syndrome cognitive profile WBSCR Williams-Beuren syndrome critical region Y2H yeast two-hybrid YL Y chromosome linked

xii

TABLE OF CONTENTS

ORIGINALITY STATEMENT ______i

ACKNOWLEDGEMENTS ______ii

PUBLICATIONS ______iv

CONFERENCE ABSTRACTS ______v

ABSTRACT ______ix

ABBREVIATIONS ______xi

TABLE OF CONTENTS ______xiii

LIST OF FIGURES ______xviii

LIST OF TABLES ______xx

CHAPTER 1 - INTRODUCTION ______1

1.1 Williams-Beuren Syndrome ______2

1.1.1 History ______2

1.1.2 WBS clinical presentation ______4

Cardiovascular disease and connective tissues abnormalities ______4

Growth, endocrine and metabolic abnormalities ______6

Craniofacial dysmorphology ______7

Cognitive and behavioural profile ______9

1.1.3 Genetics and molecular mechanisms of WBS ______13

1.1.4 Other chromosomal rearrangements in 7q11.23 ______20

WBS triplication syndrome ______24

Inversions ______24

1.1.5 Mapping individual gene contributions to WBS traits: genotype-phenotype correlations

and atypical deletions ______25

xiii

1.1.6 Animal models for WBS ______31

Multiple gene deletion models ______31

Single gene models ______33

1.2 The TFII-I protein family ______44

1.2.1 TFII-I ______45

1.2.2 GTF2IRD1 ______46

1.2.3 GTF2IRD2 ______50

1.3 Studying GTF2IRD1 function using in vivo models systems ______50

1.3.1 The auditory system ______53

Mechanical Transmission of Sound ______55

Spiral ganglion cells and the cochlear nerve ______57

The Auditory Periphery: Generation of Evoked Activity ______57

Hearing Loss ______60

The central auditory pathways and sound perception ______61

The mouse as an animal model for hearing research ______64

1.3.2 The skin ______65

The Epidermis ______67

Epidermal stem cells ______70

Barrier function and skin desquamation ______70

1.4 Aims of this thesis ______73

CHAPTER 2 - MATERIALS AND METHODS ______74

2.1 Materials ______75

2.2 Methods ______77

Mice ______77

Genomic DNA extraction from mouse tail samples and genotyping ______78

Agarose gel electrophoresis of PCR genotyping samples ______78

Transcardial perfusion ______79

xiv

Immunofluorescence analysis in the cochlea ______80

Acoustic startle response and prepulse inhibition of acoustic startle. ______81

Auditory brainstem response (ABR) ______81

Distortion product otoacoustic emissions ______83

Magnetic resonance imaging (MRI) analysis ______83

LacZ enzymatic detection ______84

Immunofluorescence in skin sections ______84

Whole mount skin clearing technique ______85

Skin barrier function assessment ______86

Total RNA extraction ______87

RNA-formaldehyde gel electrophoresis ______87

Reverse transcription ______88

Transcriptome sequencing ______88

RNA-Seq: differential expression and functional analyses ______89

Quantitative real time-PCR (qPCR) ______91

Total protein extraction from lip tissue. ______91

Protein quantification ______92

SDS-polyacrylamide gel electrophoresis (SDS-PAGE) ______92

Western blotting. ______93

Statistical data analysis ______94

CHAPTER 3 - THE ROLE OF GTF2IRD1 IN THE AUDITORY

PATHOLOGY OF WILLIAMS–BEUREN SYNDROME ______95

3.1 INTRODUCTION ______96

3.2 RESULTS ______99

3.2.1 Expression of Gtf2ird1 in the cochlea ______99

3.2.2 Structure of the Gtf2ird1-/- cochlea ______104

3.2.3 Behavioural analysis of hearing capacity in Gtf2ird1-/- mice ______108

3.2.4 Gtf2ird1-/- mice have a hearing deficit across a broad frequency range ______111

3.2.5 Recruitment and conduction of auditory nerve firing is not affected in Gtf2ird1-/- mice 114

3.2.6 Gtf2ird1 null mice exhibit moderate loss of outer hair cell function ______118

3.2.7 DPOAE input/output functions in Gtf2ird1-/- mice show amplitude reductions at higher

levels of sound ______120

3.2.8 Age-related assessment of Gtf2ird1-/- hearing capacity and genetic background effects on

hearing ______122

3.3 DISCUSSION ______124

CHAPTER 4 - GTF2IRD1 CONTROLS EPIDERMAL CELL

PROLIFERATION AND FACIAL SKIN PATTERNING DURING

DEVELOPMENT ______135

4.1 INTRODUCTION ______136

4.2 RESULTS ______139

4.2.1 Epidermal thickening in Gtf2ird1-/- mice is restricted to regions of the face and foot pads

______139

4.2.2 Gtf2ird1 is expressed in affected regions of the face ______144

4.2.3 Gtf2ird1-/- mice do not show skull abnormalities ______149

4.2.4 Gtf2ird1-/- mice have thickened lips as result epidermal hyperproliferation and enlarged

layers of differentiated keratinocytes ______153

4.2.5 Soft tissue facial abnormalities of Gtf2ird1-/- mice begin during foetal life ______160

4.2.6 Gtf2ird1-/- mice have a defect in skin barrier function ______164

4.3 DISCUSSION ______170

CHAPTER 5 - IN VIVO GENE REGULATION ASSOCIATED WITH

GTF2IRD1 IN LIP SKIN ______181

5.1 INTRODUCTION ______182

5.2 RESULTS ______184

5.2.1 RNA-Seq analysis ______184

xvi

5.2.2 Differential gene expression in lip tissue of Gtf2ird1-/- mice ______184

Keratin network dysregulation ______189

Dysregulation of epidermal markers involved in skin development and homeostasis ____ 194

Dysregulation of signalling pathways ______198

FGF signalling pathway dysregulation ______200

Differentially expressed transcription factors ______204

Genes highlighted in bold are involved in brain development. ______205

Genes annotated as associated with anatomical structure development and neural process 208

Direct interaction partners of Gtf2ird1 ______208

Gene targets of Gtf2ird1 ______210

5.2.3 Analysis of the differentially expressed genes by RTqPCR ______210

5.2.4 Dysregulation of FGF7 at the translational level ______215

5.3 DISCUSSION ______218

CHAPTER 6 – OVERVIEW AND CONCLUDING REMARKS ______231

6.1 Overview ______232

6.1.1 GTF2IRD1 has a role in the auditory system ______233

6.1.2 GTF2IRD1 has a role in skin control mechanisms ______235

6.1.3 Gene dysregulation in lip skin due to the lack of Gtf2ird1 ______238

6.3 Concluding remarks ______241

REFERENCES ______243

xvii

LIST OF FIGURES

Figure 1.1 Dysmorphic craniofacial features in WBS patients (page 8) Figure 1.2 The WBS deletion region in human chromosome 7q11.23 (page 13) Figure 1.3 The WBS rearranged region and flanking LCR blocks (page16) Figure 1.4 Typical and atypical deletions of the 7q11.23 region and correlations with key phenotypes (page 28) Figure 1.5 Model of GTF2IRD1 binding to its own promoter region (page 47) Figure 1.6 Overview of the external, middle and internal human ear (page 52) Figure 1.7 The cochlear duct and organ of Corti (page 54) Figure 1.8 Classification of AER waveforms as they occur in time (page 57) Figure 1.9 The central auditory pathways (page 60) Figure 1.10 Three-dimensional diagram of the skin (page 63) Figure 1.11 Structure of the epidermis (page 66) Figure 3.1 Gtf2ird1 expression in the adult Gtf2ird1tm2(LacZ)Hrd mouse cochlea (page 99) Figure 3.2 Histology of Gtf2ird1-/- cochleae confirms normal cytoarchitecture (page 103) Figure 3.3 Acoustic startle response and prepulse inhibition (PPI) of acoustic startle response of Gtf2ird1-/- mice (page 107) Figure 3.4 Auditory brainstem response (ABR) threshold analysis (page 110) Figure 3.5 ABR growth functions (page113) Figure 3.6 Analysis of the distortion products of otoacoustic emission (DPOAE) thresholds in Gtf2ird1 knockout mice (page 116) Figure 3.7 Analysis of DPOAE growth functions (page 118) Figure 3.8 Hearing function in aged mice (page 120) Figure 4.1 MRI analysis of Gtf2ird1-/- mice reveals no additional distinctive facial features (page 139) Figure 4.2 Epidermal thickening in Gtf2ird1-/- mice is restricted to regions of the face and footpads (page 140) Figure 4.3 Expression of Gtf2ird1 in mouse facial skin at different stages of development (page 144) Figure 4.4 Expression of Gtf2ird1 in mouse lip skin in the adult (page 145) Figure 4.5 Skull analysis in adult Gtf2ird1-/- and wild type mice (page 148) Figure 4.6 Increased cellular proliferation in lip skin sections from Gtf2ird1-/- mice (page 152)

xviii

Figure 4.7 The top lips of Gtf2ird1-/- mice have an abnormal number of differentiated keratinocytes (page 155) Figure 4.8 Changes in skin morphology start during embryonic development in Gtf2ird1-/- mice and persist after birth (page 159) Figure 4.9 Whole mount toluidine blue staining evaluates skin barrier function development (page 163) Figure 4.10 Trans-epidermal water loss (TEWL) assay as a means to quantify the skin barrier function in adult mice (page 166) Figure 5.1 Quality assessment of total RNA extracts from lip samples (page 183) Figure 5.2 Analysis of differential gene expression in lip tissue (page 184) Figure 5.3 Differentially expressed keratins in Gtf2ird1-/- mice (page 190) Figure 5.4 RTqPCR validation of DEGs identified by RNA-Seq analysis (page 210) Figure 5.5 FGF7 protein increase in Gtf2ird1-/- lip skin extracts (page 210)

xix

LIST OF TABLES

Table 1.1 A summary of the clinical features of WBS (page 11) Table 1.2 Some of the most common examples of CGS as genomic disorders (page 18) Table 1.3 Comparison of the clinical features of WBS and Dup7 syndrome (page 22) Table 1.4 Functions of the skin (page 69) Table 2.1 Commercially available reagents and kits (pages 72-73) Table 2.2 Primary antibodies (page 73) Table 2.3 Oligonucleotide primers utilised for genotyping (page 73) Table 2.4 Oligonucleotide primers utilised for qPRC (page 74) Table 5.1 The top 50 differentially expressed genes (DEGs) (page 185) Table 5.2 Differentially expressed keratins and keratin-associated protein-encoding genes (pages 188-189) Table 5.3 DEGs involved in epidermal processes (pages 192-194) Table 5.4 DEGs involved in signalling processes in skin (pages 199-200) Table 5.5 Differentially expressed transcription factors (page 202) Table 5.6 DEGs classified as Lhx2 downstream targets found to be dysregulated in Gtf2ird1-/- mice (pages 203-204) Appendix 1 DEGs involved in signalling pathways – complete list (pages 223-227)

CHAPTER 1 - INTRODUCTION

Chapter 1 - Introduction

1.1 Williams-Beuren Syndrome

Williams-Beuren syndrome (WBS, OMIM 194050) is a rare multisystem genetic disorder caused by a hemizygous microdeletion of a 1.55 to 1.84 Mb region in the long arm of chromosome 7, which contains up to 28 genes (Bayes et al., 2003;

Francke, 1999; Pober, 2010).

1.1.1 History

WBS involves an array of clinical manifestations that were originally described in the early 1960’s in two different reports by J.C.P. Williams, a cardiologist from New

Zealand (Williams et al., 1961), and by A.J. Beuren, a German physician (Beuren et al.,

1962). In the first report, John C.P. Williams, who had an interest in the study of supravalvular aortic stenosis (SVAS), described the case of four patients that presented with SVAS, combined with an intellectual disability and shared several physical characteristics, including fullness of the face, a broad forehead, well set-apart eyes, heavy cheeks, a wide mouth and pouting lips. The chin was described as pointed and some of them had prominent ears and malocclusions of the teeth. Although the report acknowledged that there were clearly differences from face to face, the described features seemed to be present in all four to varying degrees. In combination, the features gave the patients a similar appearance, which was not easily defined, but was evident to their parents and the subject of comment by casual observers (Williams et al., 1961). In the same report, Williams proposed that the association of SVAS with the physical and mental characteristics might constitute a previously unrecognized syndrome. One year after, A.J. Beuren and colleagues reported four new patients with SVAS that had very similar facial features as described above, and added that the children were

Chapter 1 - Introduction uncharacteristically friendly, loved and were loved by everyone, and possessed a very charming personality (Beuren et al., 1962). This behavioural trait later became a prominent feature of the condition and an additional means for diagnosis. In recognition of the contribution made by both authors, the condition is now described as Williams-

Beuren syndrome (WBS).

For the next 10-12 years, various studies broadened the clinical characterization of the WBS cardiovascular problems (Beuren, 1964; Beuren et al.,

1964; Eie et al., 1972; Farrehi et al., 1964; Fay et al., 1966; Guay et al., 1968; Hartel et al., 1968; Hasegawa et al., 1969; Hashiba et al., 1970; London and Blodgett, 1963;

Wyler et al., 1973). In 1975, Jones and Smith reported a large-scale study on WBS patients of different ages, in which the observable physical symptoms were described in greater detail and the behavioural profile was comprehensively assessed. This study, for the first time, allocated measures of genetic penetrance to the so-called common cardiac abnormality of the syndrome, which turned out to be present in no more than 32% of the analysed individuals, being a more consistent feature than the postnatal growth deficiency and the altered pattern of facial development, which also included short palpebral fissures, a stellate pattern in the iris, medial eyebrow flare, a depressed nasal bridge with anteverted nares and thick lips. However, by that time, the disorder was still thought to be just a sporadic occurrence with unknown aetiology (Jones and Smith,

1975; Lashkari et al., 1999). More recent studies have demonstrated that SVAS shows a particularly high incidence (70%) with a broad severity range (Collins et al., 2010;

Pober, 2010).

Chapter 1 - Introduction

1.1.2 WBS clinical presentation

The breadth of the symptoms and features presented by individuals with WBS, demonstrates that multiple organs and systems are affected. These include cardiovascular abnormalities, a distinctive facial gestalt, growth, endocrine, metabolic and neurological problems and a unique cognitive and behavioural profile with intellectual disability (Cherniske et al., 2004; Merla et al., 2010; Morris et al., 1988;

Pober, 2010). Descriptions of the main WBS clinical characteristics are presented below and a summary of WBS clinical findings is presented in Table 1.1.

Cardiovascular disease and connective tissues abnormalities

Cardiovascular abnormalities in WBS occur at a high frequency. The overall prevalence is 50-80% and, although in the past they used to account for most of the

WBS mortality, early medical attention and advances in surgical intervention have brought the mortality rates to relatively low levels (Collins et al., 2010). The most common cardiovascular defect of WBS is SVAS, which is caused by the narrowing of the ascending aorta. SVAS occurs in approximately 70% of WBS patients and can be variable in degree, ranging from minor to severe and can cause death in some cases when not treated on time (Pober et al., 2008). However, SVAS often develops in combination with other vascular stenoses (narrowing of the blood vessels) such as aortic arch hypoplasia and pulmonary artery, renal artery and coronary artery stenosis (Donnai and Karmiloff-Smith, 2000; Pober, 2010). Hypertension is also frequent, affecting ~50

% of the patients, as are electrocardiogram (ECG) abnormalities such as prolongation of the QT interval, which creates a greater risk of sudden death (Pober et al., 2008).

Chapter 1 - Introduction

Linkage analysis in family pedigrees has revealed that SVAS-affected individuals had a breakpoint in the elastin gene (ELN) (Curran et al., 1993). This study stablished the link between ELN and the cardiovascular stenosis phenotypes described above and attributes these defects in WBS patients to hemizygosity at the ELN locus.

ELN lies within the WBS deleted region and encodes the elastin protein (Curran et al.,

1993; Ewart et al., 1994; Tassabehji et al., 1997). Elastin is essential for providing strength, support and elasticity to tissues that contain elastic fibres such as the skin, heart, lungs and all major blood vessels such as the aorta and the pulmonary arteries.

Consequently, haploinsufficiency of ELN can lead to problems in all of these tissues, causing a range of vascular complications with variable severity. Moreover, haploinsufficiency of elastin can predispose WBS patients to congenital heart failure, hypertension and/or stroke, which is occasionally seen during childhood and becomes progressively more likely during adulthood (Broder et al., 1999).

Adult WBS patients also present with other prominent connective tissue defects, some of which have been attributed to the lack of elastin. These include a hoarse/deep voice, hernias, bladder and/or bowel diverticular disease, soft skin and musculoskeletal problems such as decreased bone mineral density, joint laxity, radioulnar synostosis, kyphosis, lordosis and scoliosis (Cherniske et al., 2004; Morris and Carey, 1990). In infants, muscle hypotonia is often present with the possible development of progressive joint limitations and contractures in adulthood (Beuren, 1972; Kaplan et al., 1989;

Morris et al., 1988).

Chapter 1 - Introduction

Growth, endocrine and metabolic abnormalities

WBS infants display symptoms that include pain from esophagitis, prolonged colic, feeding problems and vomiting due to gastroesophageal reflux, constipation and failure to thrive. Since the growth profile of WBS children is characterised by a delay in the first four years of life, the symptoms mentioned above have been associated with the growth retardation, which is found in 80% of WBS infants. It is also believed that this developmental delay could contribute to the immaturity of the central nervous system

(Morris et al., 1988). Although there is a catch-up of growth during the later stages of childhood, a shorter height in adulthood is commonly observed (Morris et al., 1988;

Perez Jurado et al., 1996). However, premature onset of puberty is also observed compared to the general population, and this is likely to be a significant contributor to the shorter stature (Cherniske et al., 1999; Partsch et al., 1999).

Endocrine and metabolic features in WBS include hypothyroidism, hypercalcaemia and an increased prevalence of diabetes (Palacios-Verdu et al., 2015;

Pober, 2010). Hypothyroidism is generally subclinical, but focused studies have shown increased levels of thyroid stimulating hormone (TSH) accompanied with mild thyroid hypoplasia. Hypercalcaemia is observed in early development in 6-15% of children with

WBS and it is commonly classified as a transient or mild phenotype that may or may not lead to temporary episodes of colic, vomiting, decreased appetite, and constipation

(Perez Jurado et al., 1996; Sforzini et al., 2002). The development of diabetes or impaired glucose tolerance, which is a pre-diabetic state of hyperglycaemia associated with insulin resistance, are the most prevalent endocrine features in WBS individuals, which are found in up to 75 % of adult WBS patients tested. Other features frequently found in children or adults with WBS include hypotriglyceridemia and increased blood levels of bilirubin, total protein and albumin (Palacios-Verdu et al., 2015; Pober, 2010).

Chapter 1 - Introduction

Craniofacial dysmorphology

Craniofacial abnormalities account for about one-third of all human congenital defects (Chacon et al., 2007; Tassabehji et al., 2005). The distinctive facial dysmorphology seen in WBS mainly affects soft tissues, although the underlying skull has also been shown to be affected in some cases (Mass and Belostoky, 1993; Urban et al., 2000). The extent of the characteristic features ranges from subtle to dramatic and, in some cases, the facial gestalt observations are sufficient to suggest a preliminary diagnosis to an experienced clinical geneticist even before clinical examination and genotyping studies are undertaken (Hammond et al., 2005; Hovis and Butler, 1997;

Winter, 1996). Three-dimensional imaging is a useful diagnostic tool in generating surface models that can discriminate these distinctive features from the background variability associated with ethnic and family-based genetic influences (Hammond et al.,

2005) (Figure 1.1).

The main features include a broad forehead, puffiness around the eyes, epicanthal folds, a short upturned nose, prominent earlobes, a long philtrum, full lips, full cheeks and a relatively small chin (Axelsson, 2005; Axelsson et al., 2005; Bellugi et al., 2000; Jones and Smith, 1975; Lashkari et al., 1999; Mass and Belostoky, 1993). The skull features have been defined as a short anterior cranial base, steep mandibular plane angle, a deficient chin button, shorter upper facial height and a longer lower facial height line (Mass and Belostoky, 1993). A recent report of 8 WBS patients with craniosynostosis, a birth defect in which one or more of the joints between the bones of the skull closes prematurely, suggests that this abnormality may occur more frequently in WBS than expected (Ueda et al., 2015).

Chapter 1 - Introduction

Figure 1.1 Dysmorphic craniofacial features in WBS patients

A) Top panels; images of an infant with WBS show the dysmorphic facial features from two different angles. The epicanthal fold, small chin and upturned nose can be seen from the side view and the periorbital fullness, bulbous nose, small chin, and long neck are better appreciated in the front view. Bottom panels show how WBS facial features evolve with age in the same individual at different stages of his life (15 months, 3 and 21 years old respectively). Images adapted from (Kaplan et al., 2001;

Pober et al., 2008). B) Front and side views from a tridimensional model analysis of facial morphology for WBS patients (labelled WS in the panel) and control individuals showing how the typical WBS features are expected to change as the face ages. Adapted from (Hammond et al., 2005; Tassabehji et al., 2005)

Chapter 1 - Introduction

Cognitive and behavioural profile

As with many of the human chromosomal abnormality syndromes, WBS displays specific cognitive and behavioural components. Unlike Down’s syndrome, which has a relatively global impact on neurological functions, the WBS cognitive profile (WBSCP) is marked by an uneven set of severe cognitive impairments and relative strengths, which is a distinctive feature of their personality (Bellugi et al.,

1999a; Bellugi et al., 1999b; Tassabehji, 2003). The WBSCP includes unique personality traits that indicate hypersociability by expressing overfriendliness and a lack of fear with strangers (Doyle et al., 2004; Riby et al., 2014), good verbal skills that incorporate charismatic speech patterns rich in vocabulary (Bellugi et al., 2000), musical abilities that are often a popular form of expression (Hopyan et al., 2001), together with deficient visuospatial abilities (Mervis et al., 1999) and motor control, anxiety, and simple phobias (Bellugi et al., 1990; Dilts et al., 1990). Measurements of intellectual ability show that the mean IQ of WBS individuals is considerably lower than normal age-matched controls. However, IQ scores range from severe intellectual disability (<40) to the low end of the normal range (>70), but the mean IQ indicates a mild intellectual disability (Mervis and Robinson, 2000).

A very striking characteristic of children with WBS is the reported over- sensitivity to sounds (hyperacusis), auditory fascinations and extreme fear of sounds

(phonophobia), which has been found in up to 95% of reported cases (Johnson et al.,

2001; Klein et al., 1990; Van Borsel et al., 1997). Although the cause of these auditory abnormalities remains unclear, some association has been suggested between hyperacusis and the hyperactivity of WBS and also between hyperacusis and the occurrence of otitis media (Klein et al., 1990; Van Borsel et al., 1997).

Chapter 1 - Introduction

However, we need to define sound tolerance and the term “hyperacusis”, which is commonly misused to describe a variety of forms of abnormal sound tolerance (e.g. auditory hyperesthesia, recruitment, dysacousis, auditory dysesthesia, odynacusis, and auditory allodynia), and sometimes confused with increased sensitivity or improved threshold to sound. Decreased sound tolerance is defined as present when a subject exhibits negative reactions when exposed to sound that would not evoke the same response in an average listener. From this perspective, hyperacusis have been proposed to describe decreased sound tolerance (Jastreboff and Jastreboff, 2004) and the classical definition for this phenomenon, which is currently accepted in the audiology field, describes hyperacusis as the abnormal acuteness of hearing due to increased irritability of the sensory neural mechanism causing discomfort (Celesia and Hickok, 2015;

Jastreboff and Jastreboff, 2015; Perlman, 1938). One approach even proposes to further divide the term hyperacusis, according to it cause, into loudness hyperacusis, annoyance hyperacusis, and fear hyperacusis (Tyler et al., 2009). However, to simplify the understanding behind this term, in this thesis we will understand the term hyperacusis as the discomfort to supra-threshold sounds, which summarises the classical definitions and avoids ambiguity sorroundy what hypercasis is (Jastreboff and Jastreboff, 2015).

Children with WBS also exhibit significantly greater acoustic startle eye-blinks, which is often viewed as an indication of heightened emotional state. Since the auditory perception phenotypes are marked by the presence of heightened reactivity to any stimuli, it has been postulated that the so-called ‘hyperacusis’ could be just another manifestation of the phobias and increased anxieties, stimulated by non-social cues that are common in the WBS population (Blomberg et al., 2006; Gallo et al., 2008; Levitin et al., 2005). Moreover, about a decade ago, several reports began to emerge demonstrating that WBS patients of all ages present with increased rates of mild to

Chapter 1 - Introduction moderate progressive sensorineural hearing loss (SNHL) with a prevalence that advances towards 80% in the older age categories (Barozzi et al., 2012; Barozzi et al.,

2013; Cherniske et al., 2004; Marler et al., 2005; Marler et al., 2010; Paglialonga et al.,

2011; Paglialonga et al., 2014).

Chapter 1 - Introduction

Table 1.1 A summary of the clinical features of WBS

Facial characteristics Broad forehead Low nasal root Long philtrum Full lips Dental malocclusion Growth and endocrine Growth retardation problems Short stature Premature ageing Glucose intolerance Hypercalcemia Cardiovascular and systemic SVAS Hypertension Renal abnormalities Gastrointestinal problems Urinary tract abnormalities Bladder diverticula Connective tissue Weakness of connective tissue abnormalities Joint laxity Hypotonia Neurological problems SNHL Brain MRI abnormalities (non-specific) Cognitive abnormalities Developmental delay Mild to moderate intellectual disability

Relative strength in expressive language Deficit of visuospatial skills Behaviour Excessively social Autism spectrum behaviours Anxiety disorder / Hyperacusis Auditory fascinations ADHD Weakness in daily living skills Deficit in motor abilities

SVAS: supravalvular aortic stenosis; SNHL: sensorineural hearing loss; MRI: magnetic resonance imaging; ADHD: attention deficit/hyperactivity disorder. (Fusco et al., 2014;

Merla et al., 2010)

Chapter 1 - Introduction

1.1.3 Genetics and molecular mechanisms of WBS

The first insights into the genetic causes of WBS were reported in the 1990’s.

Analysis of a patient with a chromosomal microdeletion in chromosome 7 containing the ELN gene, suggested that WBS was a contiguous gene disorder in which the vascular and connective tissue abnormalities were caused by the deletion of one of the

ELN alleles (Ewart et al., 1993). In the same year, evidence emerged that indicated a possible autosomal dominant inheritance pattern for this disorder (Morris et al., 1993b;

Pankau et al., 1993; Sadler et al., 1993). Putting these two elements together led to a partial understanding of the genetics and resulted in the conclusion that the pathogenesis of both familial and de novo WBS was caused by sub-microscopic deletions within chromosome 7 and hemizigosity at the ELN locus. It is now known that this hypothesis was correct and the microdeletion that causes WBS occurs in the chromosomal region,

7q11.23, and the typical deletion interval varies from 1.55 to 1.84 Mb including up to

28 genes (Bayes et al., 2003; Pober, 2010) (Figure 1.2).

Chapter 1 - Introduction

Figure 1.2 The WBS deletion region in human chromosome 7q11.23

Diagram showing the location of the common WBS deletion region in chromosome 7. Genes contained in the common WBS deletion region are shown with their chromosomal orientation. Gene approved symbols are up to date according to

HUGO Gene nomenclature Committee (HGNC database, www.genenames.org)

Chapter 1 - Introduction

It has been shown that the architectural changes incurred by chromosomal rearrangements, not only affect the CNV-affected genes within the region, but also produce changes in the expression of genes flanking the rearrangement. Thus, the phenotypic outcomes of a specific CNV region can be affected by the altered expression of the neighbouring genes with a normal copy-number (Merla et al., 2006; Reymond et al., 2007).

The molecular mechanisms that cause chromosomal rearrangements can be very complex and can cause copy number variations (CNVs) in the human genome, often resulting in pathological presentations (Ji et al., 2000; Lupski, 1998). Homologous low copy repeats (LCRs) mediate many of these rearrangements, predisposing chromosomes to recombination errors, which tend to occur during meiosis by non- allelic homologous recombination (NAHR) between these LCRs, (Stankiewicz et al.,

2003).

The chromosome 7 WBS contains LCRs that are positioned on the centromeric, medial and telomeric part of the WBS locus (Bayes et al., 2003). These

LCRs are made up of smaller blocks of homology, named A, B and C. These LCRs provide a substrate for NAHR during meiosis, which leads to the rearrangements responsible for WBS (Schubert, 2009; Stankiewicz and Lupski, 2002). The LCR sequences in 7q11.23 are approximately 320 Kb long and contain regions of very high homology; the highest being within block B. This explains why the 95 % of WBS patients that have the common ~1.5 Mb deletion have centromeric and medial LCR breakpoints that fall in block B (Bayes et al., 2003; Schubert, 2009). Block B in the medial LCR contains three genes (GTF2I, NCF1 and GTF2IRD2), while block B in the centromeric and telomeric LCRs contains their corresponding putative pseudogenes.

The single-copy gene region of ~1.8 Mb is located between blocks C and B in the

Chapter 1 - Introduction centromeric and medial LCRs. A larger deletion, only seen in 5 % of the cases, is mediated by recombination between the A blocks of the centromeric and medial LCRs

(Bayes et al., 2003) (Figure 1.3).

Chapter 1 - Introduction

Figure 1.3 The WBS rearranged region and flanking LCR blocks

A) Schematic representation of the rearranged WBS deletion region and its flanking LCR blocks with arrows showing orientation. Centromere (cen), telomere (tel) and the ELN gene are shown as reference landmarks. The position of the single copy gene region (~1.8Mb) and the A, B and C LCR blocks are shown. The diagram illustrates the arrangement of each block in the centromeric (CEN), middle (MID) and telomeric (TEL) LCR. Distances are not drawn to scale. B) Diagram showing the position and orientation of the genes within the LCR regions [adapted from (Schubert,

2009)].

Chapter 1 - Introduction

Genomic disorders that are produced by the rearrangement of a chromosomal interval that extends to multiple adjacent genes are classified as a contiguous gene syndrome (CGS) (Schmickel, 1986). Patients with CGSs typically exhibit multisystem phenotypes, which are generally due to the presence of one or more dosage sensitive gene in the chromosomal rearranged interval. WBS is therefore classified as a CGS.

Some other examples of this type of chromosomal rearrangement are listed in Table 1.2.

In general, WBS deletions occur sporadically, but a few cases of autosomal dominant inheritance have been reported (Morris et al., 1993b; Ounap et al., 1998;

Pankau et al., 1993; Pankau et al., 2001; Sadler et al., 1993). Estimation of the prevalence of the phenomenon has been made difficult by the considerable clinical variation among patients. Some of the transmitting parents that have been reported have very mild WBS features and, in some of the cases, the diagnosis has only been based on clinical observation without substantiating genetic or molecular analysis. The total prevalence of WBS is estimated to be approximately 1:7,500 to 1:10,000 live births, and according to demographic data analyses, WBS is found worldwide among all racial and ethnic groups (Pober, 2010; Stromme et al., 2002).

Chapter 1 - Introduction

Table 1.2 Some of the most common examples of CGS as genomic disorders.

Disorder OMIM Inheritance Chromosome Rearrangement pattern location Type

WBS 194 050 AD 7q11.23 del, inv

Dup7 609 757 AD 7q11.23 dup, trip

Prader–Willi 176 270 AD 15q11.2q13 del syndrome

Angelman syndrome 105 830 AD 15q11.2q13 del

15q11-q13duplication 608 636 AD 15q11.2q13 dup syndrome

Smith–Magenis 182 290 AD 17p11.2 del syndrome

Potocki-Lupski 610 883 AD 17p11.2 dup syndrome

DiGeorge/VCFS 188 400/ AD 22q11.2 del 192 430

Male infertility AZFa 415 000 YL Yq11.2 del microdeletion

Male infertility AZFc 400 024 YL Yq11.2 del microdeletion

OMIM: online mendelian inheritance in man; AD: autosomal dominant; YL: Y chromosome linked; del: deletion, dup: duplication, inv: inversion.

Chapter 1 - Introduction

1.1.4 Other chromosomal rearrangements in 7q11.23

In theory, NAHR should be capable of generating duplications at an equal frequency to deletions, but for many years these CNVs escaped clinical detection.

However, it has recently emerged that such duplications do exist, as well as triplication and inversion cases (Merla et al., 2010; Schubert, 2009). Patients with duplications

(OMIM#609757), also known as 7q11.23 duplication syndrome or Dup7, are predisposed to a range of clinical features. Genome-wide association studies (GWAS) have identified 7q11.23 duplications as a new risk factor for schizophrenia (Kirov et al.,

2012; Mulle et al., 2014), and similar studies in patients diagnosed with autism spectrum disorders (ASDs) have found an increased incidence of 7q11.23 recurrent de novo duplications amongst the sample group (Devlin and Scherer, 2012; Luo et al.,

2012; Sanders et al., 2011).

The first 7q11.23 duplication syndrome case was reported only 10 years ago, when a boy who was being tested for velocardiofacial syndrome using quantitative PCR

(qPCR), was found to have a duplication at the level of the elastin gene (Somerville et al., 2005). Further fluorescence in situ hybridization analysis (FISH) confirmed that the boy had a duplication of the classic WBS deletion region. To date, only a few dozen patients have been identified, and its frequency in the population has been estimated at

1:13,000-20,000 (Van der Aa et al., 2009). Similar to WBS, these cases are either familial or de novo, and share similar phenotypic characteristics (Dixit et al., 2013a;

Merla et al., 2010), which reveals insights into the role of several dosage sensitive genes in the WBS region and how their increased dosage impacts on neurodevelopmental and behavioural mechanisms.

Chapter 1 - Introduction

Dup7 clinical features

The most prominent feature of Dup7 is a severe expressive speech delay

(Osborne and Mervis, 2007), combined with developmental, craniofacial and neurocognitive abnormalities (Dixit et al., 2013b; Merla et al., 2010). Apart from speech delay, the key neurobehavioural and cognitive features that are present at a high frequency, are an intellectual disability and autism spectrum behaviours (Berg et al.,

2007; Dixit et al., 2013b; Merla et al., 2010; Van der Aa et al., 2009). Some children with Dup7 learn to develop non-verbal signs and gestures as a means of communication

(Berg et al., 2007). Dup7 individuals have relative strengths in visuospatial construction functions, which is in direct contrast to those of WBS individuals who have strengths in language capabilities but significant impairments in visuospatial cognition (Mervis and

Klein-Tasman, 2000).

The clinical profile of Dup7 appears to be more variable and not as well defined as WBS, as illustrated in a recent single study on the cognitive, behavioural and psychological profiles in the largest cohort analysed so far (Mervis et al., 2015). In this study, 63 children (27 girls, 36 boys), were systematically assessed for intellectual and vocabulary abilities, academic achievement, adaptive behaviour, autism spectrum disorder (ASD) and speech sound disorders. The results indicated that there is considerable variability in the overall level of intellectual abilities and in the patterns of relative strengths and weaknesses. The intellectual abilities, which ranged from severe intellectual disability to high average ability, were highly correlated with academic skills, and the adaptive behaviour was limited with social phobia and selective mutism.

Signs of attention deﬁcit hyperactivity disorder (ADHD), oppositional deﬁant disorder

(ODD), disruptive behaviour disorder (DBD), speech sound disorder and ASD were also present as isolated or combined features in these patients.

Chapter 1 - Introduction

A summary of the clinical features associated with the duplication and WBS is presented in Table 1.3 for comparison purposes. As expected in syndromes caused by reciprocal chromosomal rearrangements, most of the features from the duplication syndrome are opposite to WBS, suggesting that the genes impacting at the reduced dose also have an impact in the opposite direction when the dosage is increased. For example, the distinctive craniofacial abnormalities of the duplication include thin lips, a short philtrum and a high broad nose, which contrasts with the thick lips and long philtrum of WBS. However, there are a number of symptoms that are conserved in both disorders.

Despite the presence of a recognisable set of phenotypes for the 7q11.23 duplication, current observations suggest high variability and incomplete penetrance.

Furthermore, the duplication has been found in apparently healthy carriers and in parents of 7q11.23 duplication syndrome patients, displaying a much milder set of phenotypes (Berg et al., 2007; Dixit et al., 2013a; Van der Aa et al., 2009). This may explain why the occurrence of the 7q11.23 duplication via parental transmission has been observed more frequently than for WBS. One representative example of such variability includes a patient with speech delay, autistic features and behavioural problems, whose transmitting parent presented learning difficulties but normal speech

(Dixit et al., 2013a).

Chapter 1 - Introduction

Table 1.3 Comparison of the clinical features of WBS and Dup7 syndrome.

Finding WBS Dup7 syndrome Facial characteristics Broad forehead Broad forehead Low nasal root High, broad nose Long philtrum Short philtrum Full lips Thin lips Growth and endocrine Growth retardation Normal growth a problems Hypercalcemia Normocalcemia Cardiovascular SVAS Congenital heart abnormalities Hypertension defects Connective tissue Joint laxity Joint laxity abnormalities Hypotonia Hypotonia Neurological problems Seizures Brain MRI Brain MRI abnormalities abnormalities (non-specific) (non-specific) Cognitive abnormalities Developmental delay Developmental delay Mental retardation Mental retardation b

Relative strength in Speech and language expressive language delay Behavioural problems Deficit of visuospatial Visuospatial skills c skills spared Excessively social Deficits of social Autism spectrum interaction/aggressive behaviours behaviour Autism spectrum ADHD behaviours ADHD

(a) A few patients with growth retardation have been reported. (b) Transmitting parents with normal cognition have been reported. (c) Poor visuospatial skills reported in two patients with 7q11.23 duplication. Reprinted from (Merla et al., 2010).

Chapter 1 - Introduction

WBS triplication syndrome

To date, only a single case of triplication of the 7q11.23 WBS region has been reported (Beunders et al., 2010). This 1.25 Mb rearrangement was found to be de novo and shared the common distal WBS breakpoint. However the proximal breakpoint was located between the genes FZD9 and BAZ1B. The clinical features are similar to those seen in patients with 7q11.23 duplication syndrome, but features such as the mental disability, language and speech delay, autistic behaviour and dysmorphic facial features were even more severe than the typical presentation. The most prominent feature observed was the severe expressive language delay.

Inversions

7q11.23 inversions have been reported in 27 % of patients with atypical WBS

(Osborne et al., 2001). Interestingly, 33 % of these cases were found to have parental transmission. Compared to an estimated 5 % of the general population that carry the inversion, these findings suggest that this rearrangement increases the chance of further mutations in the region (Hobart et al., 2010; Osborne et al., 2001; Scherer et al., 2005).

A study that assessed families with the common inversion of the region (WBSinv-1), reported no symptoms that overlap with WBS in the inversion carriers, nor gene expression changes in the WBS common region or the surrounding genes, suggesting that WBSinv-1 does not cause clinical symptoms advising caution when diagnosing individuals with atypical presentation of rare syndromes (Tam et al., 2008).

Chapter 1 - Introduction

1.1.5 Mapping individual gene contributions to WBS traits: genotype-phenotype

correlations and atypical deletions

As mentioned above, the 1.5-1.8 Mb deletion comprises up to 28 annotated genes that have been assigned to the WBS critical region (WBSCR), and the typical deletion is relatively homogeneous due to the clustering of breakpoints within the LCRs surrounding WBSCR (Valero et al., 2000). Further dissection of the WBS phenotype relies mainly on evidences from functional studies of single genes, animal models, and analysis of individuals with atypical deletions in the WBS region. Studies of atypical deletion patients allows correlations to be made between haploinsufficiency of some of the WBSCR genes and the WBS phenotypic features, since some of these individuals lack elements of the typical WBS presentation. These analyses have permitted a relatively sophisticated level of genetic mapping to specific parts of the deletion region.

Detailed molecular characterizations of the deletion breakpoints, alongside well-defined cognitive profiling in WBS provides a unique opportunity to identify specific genes that play an important role in complex human behaviours and thus unlock some of the most important mechanisms linking genetics with neurocognitive functions (Tassabehji,

2003). These correlation analyses have pinpointed several candidate genes as contributors to specific aspects of the WBS phenotype, although the precise biological pathways affected are yet to be understood.

Although the deletion includes up to 28 genes that could potentially be associated with the development of WBS features, it is unlikely that the deletion of every single gene will play a role in the WBS pathogenesis, because not all of the proteins that they produce would be expected to be dosage sensitive. To date, ELN is the only gene to be unequivocally shown to contribute to a specific element of the disease; namely the SVAS cardiac phenotype (Curran et al., 1993) and mutations in ELN locus

Chapter 1 - Introduction have also been associated with non-syndromic conditions (Merla et al., 2012;

Tassabehji and Urban, 2006). Concurrently, association studies from patients carrying atypical deletions that were classified as normal blood pressure group, have concluded that the hemizygous deletion of NCF1, a gene located in the distal portion of the WBS rearranged region, has a protective role from such hypertension (Del Campo et al.,

2006). On the other hand, examination of several Dup7 patients revealed that there is evidence of aortic dilation, a feature that is opposite to the SVAS seen in WBS. These findings suggest that ELN could also be involved in the cardiovascular phenotype of the duplication syndrome through increased elastin abundance (Del Campo et al., 2006;

Parrott et al., 2015).

Genes in the telomeric portion of the WBS region have been implicated in playing important roles in some of the remaining clinical features (Antonell et al.,

2010). However, the ability to refine this map even further has been hampered by the small number of cases with breakpoints in informative locations and most of the clinical analyses are based on a very small pool of patients; sometimes single case reports.

A recent report described a patient carrying an 81.8 Kb deletion containing just the genes ELN and LIMK1. The clinical presentation of this patient showed vascular abnormalities, mild facial dysmorphology and a development delay in fine motor skills

(Euteneuer et al., 2014). Comparison with other partial deletion cases indicated that the cardiovascular abnormalities of this patient were more severe than those in larger deletion cases, suggesting the contribution of other factors that modify the outcome. A previous study reported a comparison between one patient with an atypical deletion encompassing FKBP6 to RFC2, and another patient with a typical WBS deletion. The patient presented with cardiovascular abnormalities but no signs of the typical WSCP.

The LIMK1 gene had previously been held to be a strong candidate for involvement in

Chapter 1 - Introduction the neurological phenotypes given its cellular role and location within the deletion, but this outcome severely undermines its possible role in the causation of the WSCP, since it falls within the FKBP6 to RFC2 region (see Figure 1.2). While this rules out LIMK1 as being sufficient to cause the WSCP on its own, it doesn’t necessarily rule it out as a contributor (Tassabehji et al., 1999b).

LIMK1 has also been implicated in the visuospatial construction deficits observed in two families with atypical 7q11.23 deletions (Frangiskakis et al., 1996).

However, in contrast with the previous study, individuals with breakpoints in similar regions did not display the typical WBS spatial impairments (Gray et al., 2006;

Tassabehji et al., 1999b). Therefore, the possibility that LIMK1 contributes to the WBS cognitive deficits if deleted in combination with other genes within the WBS deletion is consequently more accepted.

CLIP2, GTF2I and GTF2IRD1 have previously been suggested as the major genes involved in the cognitive and behavioural phenotype of WBS (Hoogenraad et al.,

2002; Tassabehji, 2003; van Hagen et al., 2007). The recent discovery of two siblings with a pure deletion of CLIP2 and none of the expected WBS neuropathology, dismissed the possibility of an exclusive contribution of this gene to the phenotype and highlighted the importance of GTF2I and GTF2IRD1 in this aspect of the syndrome

(Vandeweyer et al., 2012).

A number of atypical deletion case reports support the role of GTF2I and

GTF2IRD1 in the WSCP. In one study three cases were described that harbour smaller deletions and analysis of these patients indicated that these two genes contribute to the visuospatial processing defect (Hirota et al., 2003). Analyses of the clinical phenotypes in five families with deletions extending up to GTF2IRD1, attributed a role to GTF2I

Chapter 1 - Introduction for the intellectual disabilities and implicated GTF2IRD1 in the visuospatial defects

(Morris et al., 2003).

Investigation of a patient with a larger deletion of 2.4-3.1 Mb that extended towards the telomere from the first intron of GTF2IRD1, also implicated the deletion of

GTF2IRD1, GTF2I and GTF2IRD2 with the WBS cognitive and behavioural profile

(Edelmann et al., 2007). Further support for the part played by these genes came from two partial deletion cases; one with mild mental retardation, craniofacial dysmorphology, hypersociability and no visuospatial defects, carried a deletion that included GTF2IRD1 while the other had haploinsufficiency of GTF2IRD1 and GTF2I, but only presented mild craniofacial features and did not show the WSCP (Antonell et al., 2010). Thus, the additive dosage effect of GTF2IRD1 and GTF2I is also held to be responsible for the craniofacial features observed in WBS. Figure 1.4 illustrates how the genetic contribution of these genes to the craniofacial abnormalities and the WSCP was refined in analyses of patients with different size deletions.

The majority of these phenotype-genotype association studies have been substantiated by work in genetically modified mouse models, which play an important role in detailing the genetic basis of this syndrome and this work is discussed in the next section.

Chapter 1 - Introduction

This diagram has been removed due to copyright restrictions.

Figure legend in next page

Chapter 1 - Introduction

Figure 1.4 Typical and atypical deletions of the 7q11.23 region and correlations with key phenotypes

Schematic representation of a series of atypical deletions (black lines), showing the position of their breakpoints relative to the typical WBS deletion and the genes that they include. Some key clinical features are listed against each deletion. +: present; −: not present; +/−: ambiguous or borderline; NA: not available [reprinted from (Fusco et al., 2014)].

Chapter 1 - Introduction

1.1.6 Animal models for WBS

Although atypical deletions of the WBS region provide some insights into the role of specific genes in the WBS characteristics, genotype-phenotype correlations are difficult since there are based on only a few individuals and, in many cases, the exact breakpoint of each deletion has not been established. In an attempt to understand the molecular mechanisms that underlie WBS pathology, and to provide clues towards the function and contribution of specific genes to the complex WBS phenotype, a variety of mouse models have been generated. The WBS region shows a high conservation of synteny in the mouse, falling in chromosomal band 5G, although the LCRs are not present. Although, this provides a powerful tool to investigate the molecular basis underpinning the WBS phenotypes (DeSilva et al., 2002; Valero et al., 2000), haploinsufficiency of the genes in mice does not necessarily lead to phenotypes that are as strong as the human condition and the effect of the mouse genetic background plays an important role in the effect and the penetrance of mouse phenotypes. Therefore the information garnered from these studies must be interpreted carefully (Osborne, 2010).

A summary of mouse models carrying either multiple gene deletions or monogenic mutations of genes in the orthologous WBS region is presented below.

Multiple gene deletion models

Common WBS deletion mice

A mouse model that mimics the complete WBS deletion (CD), Gtf2i to Fkbp6, was generated using Cre-Lox recombination of LoxP sites inserted at the location of these genes. This deletion was viable in the hemizygous state and embryonic lethal in the homozygous state (Segura-Puimedon et al., 2014). Phenotypic characterization of

Chapter 1 - Introduction hemizygous mice showed reduced body weight, reduced brain weight, craniofacial abnormalities, and mild cardiovascular defects (such as mild hypertension and an increase of the arterial wall thickness). The neurobehavioural phenotypes included problems with motor coordination, an increased startle response and hypersociability.

Neuroanatomical findings included a decreased cell density in the basolateral amygdala and increased density of immature neurons in the dentate gyrus. This model provides the best representation of WBS in a mouse, as it recapitulates the complete deletion in the most accurate way and displays the majority of the features observed in the patients, although the severity of some of the features, such as the cardiovascular phenotype, is milder.

Partial deletion mice

Similar methods of genetic manipulation were used to generate a mouse model that split the deletion into two parts. Proximal deletion (PD) and distal deletion (DD) mouse lines were engineered with the LoxP sites targeted to Gtf2i, Limk1 and Fkbp6.

PD mice were generated by recombination of the LoxP sites in Gtf2i and Limk1 and the

DD mice were generated by recombination of LoxP sites in Limk1 and Fkbp6 (Li et al.,

2009). A third line of mice were generated (D/P) by crossing the PD and DD lines together, thus recapitulating the complete human deletion, although these mice are homozygous null for Limk1 because this gene is targeted in both PD and DD lines.

These mice also showed most of the WBS features and allowed a degree of genetic mapping to either the PD or DD region (Goergen et al., 2011; Li et al., 2009). DD mice were shown to have foreshortened skulls, with a similar trend in the D/P mice, suggesting that genes in the DD region (Limk1-Fkbp6) are primary contributors to the

Chapter 1 - Introduction hard tissue craniofacial defects of WBS (Li et al., 2009). Analyses of PD mice suggested that the genes in this region (Gtf2i to Limk1) are associated with reduced brain lateral ventricle volume, increased neuronal density in the somatosensory cortex, abnormal social behaviours, increased anxiety, motor coordination deficits and increased sensitivity to sound. (Li et al., 2009).

Single gene models

Gtf2i

GTF2I was the first member of the TFII-I family of genes to be identified. Over the last decade, three different Gtf2i mutant lines have been generated, either by gene- trap mutagenesis or by targeted deletion (Enkhmandakh et al., 2009; Lucena et al.,

2010; Sakurai et al., 2011). One of these gene-trap mutations (Gtf2iGt(XE029)Byg), generated through the insertion of a LacZ-neo cassette into the intron immediately after exon 3, caused embryonic lethality in the homozygous state (Enkhmandakh et al.,

2009). Analysis indicated that the cause of death was multiple developmental defects, suggesting that TFII-I plays an important role in embryonic development. No data were reported for the heterozygous mutant mice of this line, but in an effort to identify possible molecular changes caused by the lack of TFII-I protein, a microarray analysis performed on RNA extracted from E9.5 embryos, revealed down-regulation of many genes involved in core biological processes, such as oxidative metabolism, cell division, transcription, translation, the ubiquitin cycle, cytoskeleton regulation and cell motility

(Enkhmandakh et al., 2009).

Homozygous carriers of a different mutation of the same gene, Gtf2iGtBux, showed a similar but less pronounced embryonic lethality, with neural tube defects

Chapter 1 - Introduction being evident at E10.5 (Sakurai et al., 2011). Heterozygous mice from both lines were viable, with 10% of the mice from the Gtf2iGt(XE029)Byg line being smaller than wild type controls, whereas heterozygous mice from the Gtf2iGtBux line were indistinguishable from their wild type siblings. Behavioural analyses of these mice demonstrated measurable alterations in social interaction compared to wild type controls, showing increased interactions with stranger mice and a lack of habituation to familiar partners.

No difference was observed in levels of anxiety in response to non-social cues. The increased sociability in the Gtf2iGtBux heterozygous mice suggests that GTF2I could be the gene responsible for the hypersociability of WBS.

The third Gtf2i mutant mouse line, Gtf2itm(Δex2), was a model generated using a gene targeting strategy, in which exon 2 was replaced by a PGK-neo cassette (Lucena et al., 2010). This gene mutation resulted in the production of a truncated protein lacking the first 140 amino acids, due to the use of an alternative in-frame start codon in exon 5.

Nuclear localization of the truncated TFII-I was not affected and, contrary to the impact of the mutations in the previous two models, the modification of Gtf2i in these mice did not result in complete homozygous loss, although a reduction in the Mendelian ratio of homogozygous Gtf2itm(Δex2) mice was reported, indicating a higher probability of embryonic lethality. 8 % of the homozygous Gtf2itm(Δex2) mice survived to birth and were virtually indistinguishable from their wild type or heterozygous littermates, although a craniofacial abnormality was apparent in homozygous and heterozygous mice, showing a short symmetrical snout and midface hypoplasia.

Separation anxiety has also been attributed to Gtf2i dosage using mice that carry different copy numbers of the gene (Mervis et al., 2012). Anxiety induced by maternal separation was assessed in mice that had 1, 2, 3 or 4 copies of the Gtf2ird1 and Gtf2i genes and was found to be significantly increased in mice overexpressing Gtf2i. These

Chapter 1 - Introduction findings correlate with observations in humans, as separation anxiety is found at higher rates in patients with 7q11.23 duplication syndrome (Mervis et al., 2012).

Gtf2ird1

GTF2IRD1 is a protein related to TFII-I that has DNA binding properties and acts mainly as a transcriptional repressor (Masuda et al., 2014; O'Mahoney et al., 1998;

Palmer et al., 2010; Polly et al., 2003). However, activation of the goosecoid gene has been associated with the xenopus orthologue of GTF2IRD1 (Ring et al., 2002) and for both positive and negative regulation of an immunoglobulin heavy-chain promoter

(Tantin et al., 2004). Several mutant mouse models of Gtf2ird1 have been characterised.

The first Gtf2ird1 knockout mouse line described resulted from a fortuitous c-myc transgene insertion, which induced a ~40 Kb deletion that removed all of the upstream regulatory sequences and the first exon of Gtf2ird1 (Durkin et al., 2001). Homozygous knockout mice exhibited growth delay and craniofacial abnormalities, including misalignment of the jaw, a shorter snout and periorbital fullness (Tassabehji et al.,

2005). However, heterozygous mutant mice did not appear to have altered growth development and craniofacial abnormalities (Tassabehji et al., 2005). This suggested that hemizygosity of Gtf2ird1 in mice may not impact on the molecular pathways to the same extent, or that other genes in combination with Gtf2ird1 produce the typical WBS phenotype. One potential criticism of this mouse model, is that the 40 Kb deletion may have removed regulatory sequences that affect more than just Gtf2ird1 expression.

A different targeted mouse mutation generated by insertion of a LacZ cassette into exon 2 of Gtf2ird1 has served as a useful tool to map the expression of Gtf2ird1 through the LacZ activity, providing a comprehensive expression analysis from

Chapter 1 - Introduction development through to adulthood (Palmer et al., 2007). According to this study,

Gtf2ird1 is broadly expressed during development, however during adulthood, the highest levels of expression are found in the central and peripheral nervous system, including nerves of the retina, olfactory epithelium, and cochlea. A more recent study of a mutant mouse line that utilised a similar targeting strategy to generate Gtf2ird1 knockout mice without a LacZ reporter, revealed growth deficits, epidermal hyperplasia in the nose and lip region as well as decreased fat tissue weight in female animals

(Howard et al., 2012). Motor coordination deficits were observed using the accelerating rotarod and the inverted cage lid test, as well as differential exploratory activity; males were found to be significantly more active than wild type male siblings whereas females were significantly hypoactive compared to their wild type sisters. Gtf2ird1-/- males were found to produce increased numbers of ultrasonic vocalizations during a forced swim test (FST), compared to their wild type littermates and Gtf2ird1-/- female pups made shorter, fewer, and different types of vocalizations during maternal separation.

Following the FST, c-fos expression, which is indicative of neuronal activity, was increased in regions of the brain implicated in mammalian vocalization, such as the cingulate cortex and ventral region of the lateral septum. This suggested a possible dysfunction in these brain regions with altered neuronal activation (Howard et al.,

2012).

A forth knockout mouse targeting exons 2 to 5 of Gtf2ird1 showed decreased aggression and anxiety as well as increased social interactions (Young et al., 2008).

Both Gtf2ird1 heterozygous and homozygous mice exhibited decreased fear in an amygdala-dependent cued fear conditioning test; phenotypes that correlated with increased levels of the serotonin (5-HT) metabolite and 5-hydroxyindoleacetic acid in the amygdala as well as in the frontal and parietal cortices (Young et al., 2008). This

Chapter 1 - Introduction suggests altered serotonergic transmission in these mice, which was further confirmed by enhanced inhibitory 5-HT currents mediated by the 5-HT1A receptor in layer V pyramidal neurons of the prefrontal cortex (Proulx et al., 2010).

A mouse line carrying a Gtf2ird1 gene-trap mutation with a LacZ-neomycin insertion has also been described (Enkhmandakh et al., 2009). However, unlike the other mouse models, this Gtf2ird1 mutant mouse had very severe phenotypes, and homozygous mice were embryonic lethal. The phenotypic discrepancies could be explained by the technique used to generate these animals. The gene trap insertion was located in intron 22 of Gtf2ird1 and, therefore, most of the protein was left intact and potentially available for interactions with its targets. However, the Gtf2ird1-LacZ fusion protein would lack the nuclear localization signal, leading to an altered localization of any GTF2IRD1 protein expressed.

Cyln2

CYLN2, also known as CLIP-115 or CLIP2, belongs to a family of membrane– microtubule interacting proteins that is highly enriched in neurons of the hippocampus, piriform cortex, olfactory bulb, and inferior olive (Hoogenraad et al., 1998). It has also been implicated in the regulation of microtubule dynamics and facilitates binding of microtubules to other cellular structures (De Zeeuw et al., 1997; Hoogenraad et al.,

2002). Targeted mutation of Cyln2 has generated Cyln2-/- and Cyln2+/- mice

(Hoogenraad et al., 2002). These mice exhibited mild growth deficits, beginning during postnatal development. The brain ventricle volume of Cyln2-/- mice was significantly larger, although the size of the corpus callosum in both Cyln2+/- and Cyln2-/- was smaller than controls. Both heterozygous and homozygous knockout animals showed impaired

Chapter 1 - Introduction performance on the accelerating rotarod, a measure of motor coordination dependent upon cerebellar function. As evidenced by decreased fear in contextual fear conditioning and significantly smaller evoked post-synaptic potentials in hippocampal slices, hippocampal synaptic plasticity also appeared to be altered (Hoogenraad et al.,

2002).

Limk1

LIMK1 is a serine protein kinase which is prominently expressed in the developing brain and is localized in the neuromuscular synapse. It controls actin dynamics via phosphorylation of cofilin and its role in the control of growth cone motility in cultured neurons has been demonstrated (Hoogenraad et al., 2004). The role in actin remodelling could be crucial for dendritic spine modifications, which make the synaptic connections in the hippocampus and are associated with the formation and maintenance of memory and learning (Nimchinsky et al., 2002). Knockout mice models suggest a role of LIMK1 in synaptic structure and spine morphology in pyramidal neurons and functions related to the actin network. Limk1 knockout mice showed behavioural anomalies, including impaired fear conditioning and spatial learning. No changes in gross brain morphology were detected but neuronal growth cone size was significantly smaller with anomalous accumulation of cofilin and actin (Meng et al.,

2002).

Eln

Since heterozygous point mutations in the human ELN locus also lead to SVAS, this WBS cardiovascular phenotype is the only one that has unequivocally been

Chapter 1 - Introduction attributed to the haploinsufficiency of the ELN gene product (Li et al., 1997).

Cardiovascular abnormalities in mouse models lacking the orthologous Eln gene further corroborate the conclusion from the human mapping data and illustrate the importance of ELN in arterial pathology (Li et al., 1998a; Li et al., 1998b). Homozygous Eln mutations cause embryonic lethality, showing severe obstructive arterial disease, which supports a role for ELN in later arterial development (Li et al., 1998a). Heterozygous

Eln+/- mice present with hypertension, decreased aortic compliance and mild cardiac hypertrophy but no SVAS, suggesting that humans are more sensitive to ELN haploinsufficiency (Le et al., 2011; Li et al., 1998b).

Stx1a

STX1A encodes a plasma membrane protein that is abundantly expressed in neurons, playing an essential role in exocytosis of neuronal and neuroendocrine cells, forming a complex with the 25-kDa synaptosomal-associated protein (SNAP-25) and vesicle-associated membrane protein 2 (VAMP-2) (Fujiwara et al., 2006; Zhang et al.,

1995). It is involved in neurotransmitter release and vesicle fusion processes via presynaptic calcium-induced exocytosis and plays a role in modulating ion channels in exocrine and muscular cells (Yamakawa et al., 2007). Studies in mice have related this gene to the abnormal endocrinology found in WBS patients. A mouse model in which

Stx1a is overexpressed by 30%, displays fasting hyperglycemia and a more sustained elevation of plasma glucose levels after an intraperitoneal glucose tolerance test, with correspondingly reduced plasma insulin levels. They also show increased fear- conditioned memory (Lam et al., 2005). The knockout model, presents no hyperglycemia but impairments in the glucose tolerance test were found with

Chapter 1 - Introduction significantly higher glucose levels and reduced insulin levels (Ohara-Imaizumi et al.,

2007). Stx1a heterozygous and knockout mice do not appear to have altered hippocampal or cerebral cortical structures and have normal spatial memory (Fujiwara et al., 2006). However, these animals may have deficits in contextual and cued fear conditioning, which are indices of hippocampal and amygdala function, respectively.

These findings may be explained by the in vivo long-term potentiation impairments observed in hippocampal slices of the mutant mice. Spontaneous locomotor activity and anxiety in Stx1a+/- and Stx1a-/- mice are similar to that of wild type littermates.

Interestingly, a different knockout mouse model generated by deletion of exons 3 through 6 and a concomitant insertion of a stop codon in exon 2 showed that heterozygous knockout animals were viable with no apparent phenotype. In contrast, homozygous animals died in utero, with a drastic reduction in body size and development at E15.5. A few homozygous mice were born exhibiting reduced body weight and only mild behavioural deficiencies. The study concluded that STX1A is an important regulator of normal in utero development, but may not be essential for normal brain function later in life (McRory et al., 2008).

Mlxipl

MLXIPL, also known as WBSCR14, is a member of the basic-helix-loop-helix leucine zipper family of transcription factors, which binds to carbohydrate responsive element motifs in the promoter regions of some glucose-regulated and lipogenic genes, activating their expression. Therefore, it represents an important factor regulating de novo lipogenesis and glycolysis (Iizuka and Horikawa, 2008). Mlxipl is ubiquously expressed but higher levels are detected in the liver, brown and white adipose tissue,

Chapter 1 - Introduction and skeletal muscle. It is maintained in an inactive, phosphorylated state in the cytosol but high glucose levels result in dephosphorylation, nuclear translocation, and transcriptional activation of target genes (Merla et al., 2010). MLXIPL is involved in the regulation of genes encoding carbohydrate responsive enzymes in the liver, which in turn control glucose metabolism and synthesis of fatty acids and triglycerides (Iizuka and Horikawa, 2008). Both glycolysis and lipogenesis are affected in a Mlxipl knockout mouse. Mlxipl deficiency also led to intolerance to simple sugars such as fructose and glucose, and this dysregulation of glucose metabolism contributes to the accumulation of glycogen in the liver (Iizuka et al., 2004; Iizuka and Horikawa, 2008).

Baz1b

BAZ1B, is a subunit of two different chromatin remodelling complexes involved in

DNA repair and transcriptional regulation (Yoshimura et al., 2009). Expression analysis in mice has shown that Baz1b is abundant in the facial prominences during embryogenesis and mice carrying a chemically-induced ENU point mutation in this gene show cranial bone abnormalities that include shorter skulls, decreased parietal and nasal bone length and mandibular hypoplasia (Ashe et al., 2008). Cardiac defects and cardiovascular abnormalities have also been previously associated with Baz1b, through the analysis of a Baz1b null mouse. Baz1b-deficient mice died shortly after birth, both homozygous and heterozygous individuals exhibited atrial and ventricular septal defects, ventricular hypertrophy, and narrowing of the aorta. Although these abnormalities were observed at a similar frequency to those found in individuals with

WBS, this study was recently retracted (Yoshimura et al., 2009; Yoshimura et al.,

2014). Targeted disruption of Baz1b in another mouse line has shown that it is essential

Chapter 1 - Introduction for spermatogenesis through abnormalities of proper gene regulation. No other phenotypes were reported in this line (Dowdle et al., 2013). Baz1b is known to be involved in chromatin remodelling as the ENU mutagenesis screen was part of an unbiased screen for novel epigenetic modulators (Ashe et al., 2008).

Fzd9

FZD9 belongs to the Frizzled (FZD) family of proteins that function as surface receptors involved in WNT signalling pathways. WNT signalling pathways are important in the regulation of vertebrate development, where FZD proteins modulate the function of WNT proteins (Cadigan and Nusse, 1997; Rattner et al., 1997).

The FZD9 gene is widely expressed in testis, brain, skeletal muscle and kidney and is selectively expressed in the hippocampus (Wang et al., 1997). Fzd9 knockout mice have shown splenomegaly, atrophy of the thymus gland, enlarged lymph nodes, as well as a significant reduction in B-cell precursors (Ranheim et al., 2005). FZD9 was shown to induce osteoblast differentiation as homozygous Fzd9 knockout mice displayed low bone mass caused by impaired bone formation (Albers et al., 2011). However the knockout mouse model also shows some characteristics which point to a potential role for the gene in WBS; including increased apoptosis in the dentate gyrus, alterations to hippocampal structure and visuospatial-learning memory defects. These findings suggest that FZD9 has an important role in hippocampal development (Zhao et al.,

2005).

Fkbp6

Chapter 1 - Introduction

FKBP6 is a protein that contains a prolyl isomerase/FK506 binding domain and a tetratricopeptide protein-protein interaction domain/TPR motif (Meng et al., 1998). It is expressed during a specific stage of male and female meiosis and the disruption of the gene influences male reproduction; i.e. arrests spermatogenesis, but not female gametogenesis. A targeted disruption of the gene in mice demonstrated that Fkbp6 is involved in the homologous chromosome pairing process during male meiosis, since abnormal phosphorylation of the testis specific histone H2AX was discovered

(Crackower et al., 2003; Ikadai et al., 1992; Noguchi et al., 1993; Noguchi et al., 2008).

Lat2

LAT2 is a linker for activation of T cells that functions as a transmembrane adaptor protein expressed mostly in spleen and the hematopoietic B, NK and mast cells

(Janssen et al., 2004). This scaffolding molecule has been implicated in negatively regulating mast cell signalling, as the secretory response was partially inhibited in bone marrow-derived mast cells and augmented anaphylactic responses were observed in

Lat2 deficient mice (Volna et al., 2004; Zhu et al., 2004).

Eif4h

Eif4h, also known as WBSCR1, encodes a protein involved in regulating protein synthesis through initiating translation and RNA duplex unwinding (Richter et al.,

1999). It is expressed widely and although its role has not been studied in patients with partial deletions, a knockout mouse model was created (Capossela et al., 2012). The phenotypic characterisation of these gene-trap mutant mice showed reduced fertility, reduced body weight and length and craniofacial abnormalities, suggestive of

Chapter 1 - Introduction characteristics found in WBS. Brain volume was reduced, with a significant reduction in posterior areas and there was also a reduction in dendritic complexity, spine number and branching in cortical neurons. These mice also presented with behavioural impairments affecting fear related learning and associative memory formation.

1.2 The TFII-I protein family

The TFII-I protein family are encoded by the genes GTF2I, GTF2IRD1, and

GTF2IRD2, which are clustered in humans at the telomeric end of the critical WBS region (Bayarsaihan et al., 2002; Franke et al., 1999; Osborne et al., 1999; Perez Jurado et al., 1998; Tassabehji et al., 1999a; Tipney et al., 2004). The GTF2I gene overlaps the junction of the single-copy gene region and the LCR. Therefore, a truncated copy of

GTF2I is also present within the centromeric LCR block (GTF2IP), which is not deleted in WBS, but lacks the 5’ end and is thus a pseudogene with no protein product (Perez

Jurado et al., 1998).

The members of the TFII-I protein family are characterised by their unique repeat domains (RDs) and the presence of a leucine zipper (LZ) motif at the N-terminus.

The RDs were predicted to form a helix-loop-helix (HLH) structure due to broad similarities to basic HLH transcription factors (Roy, 1997) and the LZ is thought to mediate heterodimerisation or homodimerisation (Hinsley et al., 2004). The adjacent chromosomal locations and the high sequence homology shared by these three genes, in particular the 70% homology shared by the RDs (Franke et al., 1999), strongly indicates that the genes have arisen by duplication and divergence from a common ancestor

(Hinsley et al., 2004; Makeyev et al., 2004).

Chapter 1 - Introduction

1.2.1 TFII-I

TFII-I, encoded by GTF2I (general transcription factor 2I), is the first member identified of the TFII-I family (Roy et al., 1991). Although four alternatively spliced isoforms have been identified, all with subcellular localisation patterns that are within both the nucleus and the cytoplasm, TFII-I is predominantly observed with nuclear localization patterns in overexpression analyses (Cheriyath and Roy, 2000; Hakre et al.,

2006). In vivo expression analyses at the transcript and protein level at various developmental stages have suggested that it is maximally active during embryonic development, but remains detectable throughout postnatal life and adulthood (Danoff et al., 2004; Enkhmandakh et al., 2004; Fijalkowska et al., 2010). The expression of TFII-I during early development and the neuronal-specific expression in the brain are consistent with a proposed role in the developmental and neurocognitive aspects of

WBS.

TFII-I was initially isolated as a component of the basal transcriptional machinery that binds to the core promoter initiator (Inr) element and an E-box upstream element (Roy et al., 1991; Roy et al., 1997; Roy, 2007). However, TFII-I has been shown by numerous subsequent reports to play active roles in signal transduction pathways. The activity of TFII-I is known to be regulated by phosphorylation at serine and tyrosine residues, mediated by multiple kinases, in response to activation of distinct signalling pathways (Cheriyath and Roy, 2000; Cheriyath et al., 2002; Kim and

Cochran, 2001; Novina et al., 1998; Sacristan et al., 2009). TFII-I has also been proposed as a factor that translocates to the nucleus in response to activation, where it activates the transcription of c-fos and other immediate-early genes (Jiang et al., 2005).

Chapter 1 - Introduction

1.2.2 GTF2IRD1

The GTF2IRD1 protein, which is encoded by the GTF2IRD1 gene, was first isolated in a yeast one-hybrid screen for factors that could bind to the enhancer element of the human slow muscle fiber gene, Troponin-I slow (TNNI1) and was initially called

MusTRD1 (muscle TFII-I repeat domain-containing protein 1) (O'Mahoney et al.,

1998). GTF2IRD1 was originally identified as a repressor of this gene and others

(O'Mahoney et al., 1998; Polly et al., 2003). However, the Xenopus homolog was shown to be capable of goosecoid activation (Ring et al., 2002), and GTF2IRD1 can regulate immunoglobulin heavy-chain promoter activity positively or negatively, depending on the cell line studied (Tantin et al., 2004). Therefore, it is believed that

GTF2IRD1 can function as either an activator or repressor depending on the target gene and its context.

Expression of recombinant GTF2IRD1 in mammalian cells reveals a predominantly nuclear localisation and the functional role of the LZ domain in homodimerisation of GTF2IRD1 has been demonstrated biochemically (Vullhorst and

Buonanno, 2003). Furthermore, electrophoretic mobility shift assay (EMSA) experiments revealed a higher-order, protein-DNA complex which was abolished when the LZ domain of the protein was deleted (Palmer et al., 2010). The RDs have mainly been implicated in DNA binding, each with variable affinity for specific DNA sequences (Polly et al., 2003; Vullhorst and Buonanno, 2003) and the nuclear localisation signal (NLS) has been proved to be necessary for the nuclear localisation of the protein (Yan et al., 2000). A common core DNA binding sequence of GGATTA has been verified by a number of DNA binding studies (Bayarsaihan and Ruddle, 2000;

Calvo et al., 2001; O'Mahoney et al., 1998; Polly et al., 2003; Ring et al., 2002;

Thompson et al., 2007; Vullhorst and Buonanno, 2003).

Chapter 1 - Introduction

Several gene targets have been identified for GTF2IRD1 (Chimge et al., 2012;

Makeyev and Bayarsaihan, 2009; Masuda et al., 2014), but definition of a universal binding site has proved elusive, probably because binding may occur in a context specific manner. ChIP analyses have indicated that GTF2IRD1 binds to the fibroblast growth factor gene Fgf15 in C2C12 mouse myoblasts and an 8 bp core consensus sequence was proposed based on this study (Lazebnik et al., 2008). The DNA binding properties of GTF2IRD1 have also been studied in detail through the knowledge that this protein is capable of negatively autoregulating its own transcription both in vitro and in vivo (Figure 1.5) (Palmer et al., 2010). Transactivation experiments have shown that GTF2IRD1 binds to a highly-conserved GTF2IRD1 upstream region (GUR) and represses gene activity (Palmer et al., 2010). This negative autoregulatory mechanism leads to elevated production of the mutant transcript in knockout mice and a level of

GTF2IRD1 transcript in lymphoblastoid cell lines obtained from WBS patients that exceeds the expected level of 50 % of normal. However, measurements of GTF2RD1 transcript in fibroblasts, or iPS cells from WBS patients, indicate that levels are much closer to the expected 50 % than is found in lymphoblastoid cell lines (Adamo et al.,

2015; Merla et al., 2006), suggesting that the degree of autoregulation varies in a tissue specific manner.

Amino acidic sequence analysis indicates that GTF2IRD1 contains putative

PEST and SUMO motifs (Hinsley et al., 2004), which are related to reduced protein half-life, transcriptional regulation and protein interactions. However, little more can be learned from predictive sequence analysis as most of the conserved domains are unique to the TFII-I family.

Experimental evidence of GTF2IRD1 function is also limited. The E3 SUMO ligase, PIASxβ (PIAS2), and the histone deacetylase, HDAC3, have been found to be

Chapter 1 - Introduction interacting partners of GTF2IRD1, hence linking GTF2IRD1 with histone modification and the SUMO pathway (Tussie-Luna et al., 2002a; Tussie-Luna et al., 2002b). Given the presence of a putative SUMO motif in GTF2IRD1, it is possible that the interaction with PIASxβ results in GTF2IRD1 SUMOylation. This result was confirmed experimentally in assays that showed GTF2IRD1 can be SUMOylated by the E2 SUMO ligase, UBC9, and its SUMOylation level is enhanced by the presence of PIASxβ

(Widagdo et al., 2012). GTF2IRD1 SUMOylation was shown to have an effect over

GTF2IRD1 protein interactions with proteins that have SUMO interacting motifs

(SIMs), as SUMOylation increases the affinity of the interaction between GTF2IRD1 and PIASxβ or ZMYM5 (Widagdo et al., 2012).

In the mouse, Gtf2ird1 expression was found to be particularly high during development in a wide variety of tissues, including cartilage, skeletal muscle, the central nervous system, teeth and hair follicles (Palmer et al., 2007). In adult mice, the expression pattern becomes much more restricted to brown adipose tissue, peripheral ganglia and regions of the brain, such as the olfactory bulb, the Purkinje neurons of the cerebellum, layer II of the piriform cortex, the striatum and layer V of the cortex

(Howard et al., 2012).

Recently, in vitro evidence from our laboratory has shown that endogenous

GTF2IRD1 has a punctate pattern in the nuclei of cultured human cell lines and neurons, and using an unbiased yeast two-hybrid assay, 38 novel interaction partners were isolated and validated in mammalian cell lines. These relationships illustrate

GTF2IRD1 function, as the isolated partners are mostly involved in chromatin modification and transcriptional regulation (Carmona-Mora et al., 2015).

Chapter 1 - Introduction

Figure 1.5 Model of GTF2IRD1 binding to its own promoter region

Three conserved consensus sequences exist at the GTF2IRD1 upstream region

(GUR), at least two are required to achieve any level of GTF2IRD1 binding. EMSA evidence indicates that two of the five RDs bind to the GUR sites in monomeric interactions, while the formation of a dimer on the GUR is dependent on the presence of the leucine zipper domain [adapted from (Palmer et al., 2010)].

Chapter 1 - Introduction

1.2.3 GTF2IRD2

The third member of the TFII-I family, is the most recently described and the most recent from an evolutionary point of view also. It is structurally quite different as it only contains two I-repeats in the N-terminal half of the protein, while the remaining

C-terminal half is taken up by an in-frame CHARLIE8 transposon-like element (Tipney et al., 2004). Sequence analysis shows that the preserved elements of the N-terminal region are more closely related to GTF2I than to GTF2IRD1 (Makeyev et al., 2004).

The gene is located in the duplicated blocks of the LCRs and, therefore, two loci for

GTF2IRD2 have been identified at the telomeric end of the 7q11.23 region, which produce the variants GTF2IRD2α and GTF2IRD2β (Tipney et al., 2004).

RT-PCR analyses have indicated that human GTF2IRD2 is ubiquitously expressed, being detectable in developing embryonic organs as well as adult brain, bone, muscle, testis, and a variety of other tissues (Tipney et al., 2004). At the cellular level, recombinant GTF2IRD2 localises to the microtubules and also inside the nucleus, in a distinct punctate pattern that overlaps with the localisation of GTF2IRD1 and TFII-

Iβ generated from transfected expression constructs (Palmer et al., 2012). Functional studies also illustrate the potential for direct interaction between GTF2IRD2 and the other family members, although homodimerisation was the most favoured interaction, followed by interactions with TFII-Iβ (Palmer et al., 2012). Based on the antagonistic behaviour of GTF2IRD2 to TFII-I and GTF2IRD1 in transgenic mouse systems, it was proposed that GTF2IRD2 may inhibit the function of its other family members by such direct interactions (Palmer et al., 2012).

1.3 Studying GTF2IRD1 function using in vivo models systems

Chapter 1 - Introduction

Study of the function of GTF2IRD1 in vivo has relied on behavioural analyses in different mouse models where the Gtf2ird1 gene is generally absent. These type of analyses are good to assess the behavioural component of the WBS and aid to the molecular-phenotype correlations relating a particular neuropathology trait with abnormalities in a determined structure of the brain. However, the genotype-phenotype correlation analysis of patients carrying atypical deletions (Antonell et al., 2010;

Edelmann et al., 2007; Ferrero et al., 2010; Tassabehji et al., 1999b; Tassabehji et al.,

2005) and the phenotypic characterisation of the different Gtf2ird1knockout mice

(Enkhmandakh et al., 2009; Howard et al., 2012; Tassabehji et al., 2005; Young et al.,

2008), have suggested that GTF2IRD1 is involved in the development of other traits which are not solely linked to the WBS neuropathology features and that its function goes beyond the WBSCP.

However, there are disadvantages in mouse behaviour studies when trying to dissect protein function at the cellular and molecular level. While behavioural data can indicate abnormalities in specific neuronal populations for further analysis, most of the time, the heterogeneity of the cell populations within a brain region can become a problem when trying to drill down into specific defects. Moreover, behavioural analyses often rely on additional factors that are not always controllable and often lack an easy means of generating a quantifiable readout.

A number of studies have shown that the expression of Gtf2ird1 is relatively low in adult mice tissues (Bayarsaihan and Ruddle, 2000; Calvo et al., 2001; Palmer et al.,

2007; Tay et al., 2003). In the brain, Gtf2ird1 expression can be detected in virtually all neuronal cell types of Gtf2ird1tm2(LacZ)Hrd mice, and although it is found in much greater intensity in some neuronal subsets, including the mitral and granular cell layer in the olfactory bulb, the Purkinje neurons of the cerebellum and neurons in the piriform

Chapter 1 - Introduction cortex (Palmer et al., 2007), these subset populations are based on the anatomical location of these neurons within the brain, and sharing a particular location within a compartment of the brain does not necessarily mean that they are functionally homologous, indeed it is not unusual that neurons sharing common locations perform different neuronal functions.

Although Gtf2ird1 expression in sensory structures and the peripheral nervous system is less well defined, the analysis of its expression in the inner ear revealed a high

Gtf2ird1 content in the spiral ganglion of the cochlea, which contains the cell bodies of bipolar neurons interconnecting the hair cells of the cochlea with the cochlear nucleus

(Palmer et al., 2007). Conversely, spiral ganglion neurons form a homogeneous cell population not only at anatomical levels but also in terms of function. Being classified as type I and type II, the only functional difference is the type of hair cell they innervate in the organ of Corti (as described below in Section 1.3.1).

Analysis of Gtf2ird1tm1Hrd mice revealed a facial deformity, resulting from localised epidermal hyperplasia in specific sites (Howard et al., 2012). Expression analysis of Gtf2ird1 during development using the Gtf2ird1tm2(LacZ)Hrd reporter mice did not involve the analysis of the skin (Palmer et al., 2007). However, the published images in this study suggest that Gtf2ird1 is highly abundant in the stem-cell-containing basal layer of the epidermis in the developing face. The epidermal cellular phenotype resulting from the lack of Gtf2ird1 in the Gtf2ird1tm1Hrd mice and the expression pattern shown in the developing face of the Gtf2ird1tm2(LacZ)Hrd mice indicate that it would be possible to dissect the function of GTF2IRD1 in these cell types in vivo and thus provide a set of principles that may be applicable across a broader array of tissues.

Before considering how this analysis is approached, it is necessary to provide an introduction to both of these organs, the auditory system and the skin.

Chapter 1 - Introduction

1.3.1 The auditory system

The ear or vestibulocochlear organ is composed of external, middle and inner parts (Figure 1.6). The external ear consists of the auricle and the external acoustic meatus with the outer layer of the tympanic membrane. The middle ear is formed by the tympanic cavity, the auditory ossicles and the inner layer of the tympanic membrane.

The inner ear comprises the labyrinth, a series of fluid-filled spaces in the petrous part of the temporal bone. The auditory part of the inner ear consists of the cochlea with the organ of Corti, which contains hair cells as auditory receptors. The receptors that are sensitive to high frequencies are located near the cochlear base and those sensitive to low frequencies near the apex. The hair cells are innervated by the peripheral processes of bipolar ganglion cells in the spiral ganglion. At birth, humans have about 20,000 inner and outer hair cells in the organ of Corti, which often do not last a lifetime as they do not regenerate when lost (Stone et al., 1998). By the age of 65–75 years, many individuals have a bilateral, high-frequency progressive hearing loss known as presbycusis associated with hair cell attrition. Hair cell loss is the most common cochlear defect causing hearing impairment in presbycusis and noise-induced hearing loss.

Chapter 1 - Introduction

Figure 1.6 Overview of the external, middle and internal human ear

Cross sectional drawing taken from ‘Three Unpublished Drawings of the

Anatomy of the Human Ear’ [reprinted from (Max Brödel, 1946)]

Chapter 1 - Introduction

Mechanical Transmission of Sound

The cochlea is composed of three chambers: the scala vestibuli, scala media and scala tympani. The scalae are separated from each other by the vestibular Reissner’s membrane and the basilar membrane (Figure 1.7). The inner scala media is filled with intracellular fluid rich in potassium called endolymph. The outer cavities, scalae vestibuli and tympani, are filled with perilymph that have approximately the same composition as the cerebrospinal fluid.

The scala media or cochlear duct contains the organ of Corti, which rests on the basilar membrane (Figure 1.7). The superior wall of the cochlear duct (Reissner’ membrane) angles downwards from lateral to medial, making the cochlear duct wedge shaped. The lateral wall is called the stria vascularis and the thickened epithelium that constitutes the organ of Corti can be divided into hair cells and supporting cells.

The hair cells are the sensory receptor cells, which exist in a single row of inner hair cells and three rows of outer hair cells. The supporting cells include the inner and outer pillar cells, which are separated by the tunnel of Corti extending the length of the cochlea. Inner and outer hair cells have different roles in the transduction of energy within the cochlea. Inner hair cells provide direct input to almost all of the axons in the cochlear nerve. Their activity is modified by local amplification of the motion of the basilar membrane produced by the outer hair cells and hair-cell related supporting cells

(Flock et al., 1999). Several molecules have been identified as having a vital role in hair-cell transduction. They are specifically expressed in and around the stereocilia and mutations in their genes lead to deafness (Steel and Kros, 2001).

Chapter 1 - Introduction

Figure 1.7 The cochlear duct and organ of Corti

Schematic representation of a transverse section of the whole cochlea (above) and a cross section of the organ of Corti (below). The cochlear nerve (cn) fibres pass through a central pillar, the modiolus (mod), whereas their cells of origin form the spiral ganglia (spg) (above). Abbreviations: bm: basilar membrane; cd: cochlear duct; ihc and ohc: inner and outer hair cells; sv: stria vascularis; SM: scala media; ST: scala tympani;

SV: scala vestibuli; tC: tunnel of Corti; tm: tectorial membrane; vm: vestibular

(Reissner) membrane [adapted from (Pickles, 2012)].

Chapter 1 - Introduction

Spiral ganglion cells and the cochlear nerve

In the cochlea, the transition from hair cell activity to neural activity is mediated through the activation of the stereocilia, which results in changes in the intracellular potential that lead to the release of a neurotransmitter from synaptic vesicles located at the base of the hair cells. The afferent fibres, which convey auditory information from the cochlea to the central nervous system, have their cell bodies in the spiral ganglion in the modiolus. These cell bodies have one process projecting to the hair cells and the other to the cell of the cochlear nucleus in the brainstem. The axons project into the cochlear duct through openings in the bony shelf of the spiral lamina. Two types of ganglion cells are found: type I and type II. The majority (90–95%) are type I cells and contact inner hair cells. The unmyelinated peripheral processes of the remaining ganglion cells (5–10%), the type II cells, contact the outer hair cells. The central processes of both types of ganglion cells form the cochlear nerve (Spoendlin and

Schrott, 1989).

The Auditory Periphery: Generation of Evoked Activity

The neural activity is reflected as brain stem auditory evoked potentials or responses (BAEPs or BAERs), which correspond to an externally recordable series of small amplitude and short latency wave-like potentials induced by transient stimuli

(Jewett et al., 1970; Stockard et al., 1978). Auditory-evoked responses (AERs) can be subdivided according to their latency into auditory brainstem response (ABR), middle latency response (MLR) and auditory cortical response (ACR) (Pasman et al., 1997)

(Figure 1.8). In humans, the ABR is characterized by six or sometimes seven deflections

(I–VII) in the first 9 ms after the stimulus. Waves I, III and V are of greatest interest because they reflect volume conducted activity from the levels of the acoustic nerve,

Chapter 1 - Introduction pons and midbrain, respectively. The earliest ABR waves (waves I and II) are generated by activity in axons of the cochlear nerve, prior to their entrance into the brain stem

(Moller and Jannetta, 1982; Stockard et al., 1978).

The ABR can distinguish between pathologies of the middle and inner ear, the auditory nerve and the brain stem and the three major applications of ABR in humans are the detection of tumours in the region of the posterior cranial fossa, monitoring the evolution of a coma and the assessment of patients with suspected demyelinating diseases.

Chapter 1 - Introduction

Figure 1.8 Classification of AER waveforms as they occur in time

Trace showing the individual waves as they occur in time. Note that the ABR occurs within the first 9 ms after stimulation (Peaks I-VI), followed by the MLR for the next 40 ms until the signal reaches the cortex, where the ACR corresponds to the longest AERs in humans [reprinted from (Perez-Gonzalez and Malmierca, 2014)].

Chapter 1 - Introduction

Hearing Loss

The two main types of hearing loss are conductive hearing loss and SNHL.

Conductive hearing loss is manifested when the defects are in the conductive mechanisms in the middle ear and are mostly caused by conditions such as otitis media and otosclerosis. SNHL is caused by insults in the cochlea or its central connection; the cochlear nerve. Hearing loss of cochlear origin is common and can result from a variety of conditions, including tumours, infections, temporal bone fractures or from exposure to excessive noise or ototoxic drugs ((Merchant and Nadol, 2010). However, there are a large number of inherited diseases that can directly or indirectly affect hearing. Many of these, although rare in the general population, have been invaluable to researchers as a means of tracking down critical molecular components of the inner ear (Pickles, 2012).

In some cases, a mixture of conductive and sensorineural hearing loss can coexist. In presbycusis, which is the hearing loss triggered by the ageing process, the loss begins with degeneration of outer hair cells at the basal end of the cochlea, but does not seriously affect hearing until the upper range of speech frequencies, around 3000 Hz, is affected. Noise induced hearing loss tends to affect the anterior basal turn of the cochlea, the region that processes 3000–4000Hz (Urban and Caplan, 2011). Tinnitus is a common symptom in disorders of the inner ear, but it can also occur in disorders affecting the VIIIth nerve such as an acoustic neuroma or vertebrobasilar disease. A sudden onset of unilateral or bilateral deafness, which is usually accompanied by dizziness or vertigo, can be a sign of occlusion of the basilar artery (Celesia, 2013;

Huang et al., 1993).

Chapter 1 - Introduction

The central auditory pathways and sound perception

As mentioned above, the auditory nerve travels from the cochlea to the brainstem. The brainstem nuclei are groups of neurons that are mirrored in each hemisphere that perform analysis on the incoming information. The first processing station is the cochlear nucleus (Figure 1.9). The neural signal is then passed to the superior olivary complex, which also receives input from the contralateral ear. It is thought that the superior olivary complex is involved in sound localisation. Further up the pathway are the lateral lemniscus and the inferior colliculus. Each stage contains many different types of neurons with varying properties that decode and analyse the auditory information. The information is passed to the primary auditory cortex via the medial geniculate body of the thalamus. Different places on the auditory cortex respond to different frequencies. However, some neurons have complex tuning properties with peaks of sensitivity at multiple frequencies, and some show evidence of lateral inhibition, so that certain frequencies cause the firing rate of the neuron to decrease.

These neurons have different tuning properties to neurons in the auditory nerve. It is clear that they receive input from neurons with different characteristic frequencies, and combine the signals in various ways. These type of neurons may be involved in processing spectral information across frequency and time (Brodal, 1981).

In addition to the ascending pathway, there is also a descending auditory pathway, carrying signals from the cortex, through the brainstem nuclei, to the outer hair cells in the cochlea. It is possible that the cortex may use this pathway, not only to influence the processing in the brainstem, but also to influence the motion of the basilar membrane. When the efferent neurons that innervate the outer hair cells are stimulated, the basilar membrane becomes less sensitive to low sound levels and frequency selectivity is reduced (Guinan and Stankovic, 1996).

Chapter 1 - Introduction

Figure 1.9 The central auditory pathways

Diagram showing the central auditory pathways, which extend from the cochlear nucleus to the auditory cortex. Postsynaptic neurons in the cochlear nucleus send their

Chapter 1 - Introduction axons to other centers in the brain via three main pathways: the dorsal acoustic stria, the intermediate acoustic stria, and the trapezoid body. The first binaural interactions occur in the superior olivary nucleus, which receives input via the trapezoid body. In particular, the medial and lateral divisions of the superior olivary nucleus are involved in the localization of sounds in space. Postsynaptic axons from the superior olivary nucleus, along with axons from the cochlear nuclei, project to the inferior colliculus in the midbrain via the lateral lemniscus. Each lateral lemniscus contains axons relaying input from both ears. Cells in the colliculus send their axons to the medial geniculate nucleus of the thalamus. The geniculate axons terminate in the primary auditory cortex which is a part of the superior temporal gyrus [adapted from (Brodal, 1981)].

Chapter 1 - Introduction

The mouse as an animal model for hearing research

The mouse is an excellent animal model for the study of human genetic deafness

(Steel and Brown, 1996). The mouse cochlea is anatomically similar to that of humans

(Steel and Bock, 1983), and hereditary abnormalities of the inner ear have been shown to be similar in both humans and mice. As in humans, the mechano-sensory cells in mice are responsible for detecting sound in the cochlea and gravity and acceleration in the vestibular system.

The commonly stated range of human hearing is 20 Hz to 20 kHz. Under ideal laboratory conditions, humans can hear sound as low as 12 Hz, being most sensitive to frequencies between 2,000 and 5,000 Hz. Individual hearing range varies according to the general condition of a human's ears and nervous system (Rosen and Howell, 2011).

Mice, however, can hear higher frequencies than humans; their frequency range is 1 kHz to 70 kHz. They do not hear the lower frequencies that humans can and they communicate using high frequency noises some of which are inaudible by humans. The distress call of a young mouse can be produced at 40 kHz. The squeaks that humans can hear are lower in frequency and are used by the mouse to make longer distance calls, as low frequency sounds can travel farther than high frequency sounds (Willott, 2001).

Despite these differences in frequency range, there is a good physiological correspondence between mice and human in frequency sensitivity, and although humans hear in a significantly more narrow bandwidth compared with other mammals, human hearing capacities seem to be solidly ‘mammalian’. This sets the basis for AER comparisons as rodent models have largely been used to understand the way brain processes the sounds we hear.

Chapter 1 - Introduction

1.3.2 The skin

The skin is the largest organ of the body making up 16% of the total body weight. It has several functions, the most important being the formation of a physical barrier to the environment, allowing and limiting the inward and outward passage of water, electrolytes and various substances, while providing protection against micro- organisms, ultraviolet radiation, toxic agents and mechanical insults. The skin is derived from the embryonic ectoderm and it has three structural layers: the epidermis, the dermis and the subcutis (Figure 1.10). Hair, nails, sebaceous, sweat and apocrine glands are considered derivatives of skin.

The skin is in a constant state of change. It is maintained by self-renewing epithelial stem cells in the basal layer that produce progenitors that undergo terminal differentiation into various types of cells. The cells of the outer layers are continuously shed and replaced by inner cells moving up to the surface (Epstein and Maibach, 1965;

Watt et al., 2006; Weinstein et al., 1984). Although structurally the skin is consistent throughout the body, it varies in thickness according to anatomical site and age of the individual (Kakasheva-Mazhenkovska et al., 2011) (Figure 1.10).

Chapter 1 - Introduction

Figure 1.10 Three-dimensional diagram of the skin

Schematic three-dimensional representation of the skin structure in a cross- section of hairy and non-hairy skin. The boundaries of the three main layers of the skin

(subcutis, dermis and epidermis) are indicated in brackets. The differences observed in thickness between both types are mostly restricted to the epidermis.

Chapter 1 - Introduction

The Epidermis

The epidermis is a stratified squamous epithelium that is mostly composed of the main cell type, the keratinocytes, which synthesise the protein keratin. Protein bridges called desmosomes connect the keratinocytes, which are in a constant state of transition from the deeper layers to the superficial (Blanpain and Fuchs, 2009). The epidermis varies in thickness from 0.05 mm on the eyelids to 0.8±1.5 mm on the soles of the feet and palms of the hand (Kakasheva-Mazhenkovska et al., 2011). Moving from the lower layers upwards to the surface, the epidermis is formed by four layers: the stratum basale (basal or germinativum cell layer), the stratum spinosum (spinous or prickle cell layer), the stratum granulosum (granular cell layer) and the stratum corneum

(horny layer) (Figure 1.11). The four separate layers of the epidermis are formed by the differing stages of keratinocyte maturation.

The stratum basale is the innermost layer of the epidermis, which lies adjacent to the dermis and is composed mainly of dividing and non-dividing keratinocytes, which are attached to the basement membrane by hemidesmosomes. As keratinocytes divide and differentiate, they move from this deeper layer to the surface (Gandarillas and Freije, 2014). Making up a small proportion of the basal cell population is the pigment (melanin) producing melanocytes. These cells are characterised by dendritric processes, which stretch between relatively large numbers of neighbouring keratinocytes. Melanin accumulates in melanosomes that are transferred to the adjacent keratinocytes, where they remain as granules. Melanin pigment provides protection against ultraviolet (UV) radiation. Chronic exposure to light increases the ratio of melanocytes to keratinocytes, so more are found in facial skin compared to the lower back and a greater number on the outer arm compared to the inner arm. The number of

Chapter 1 - Introduction melanocytes is the same in equivalent body sites in white and black skin but the distribution and rate of production of melanin is different (Weiner et al., 2014).

As basal cells reproduce and mature, they move towards the outer layer of skin, initially forming the stratum spinosum. Intercellular bridges, the desmosomes, connect the cells. Langerhans cells are dendritic, immunologically active cells derived from the bone marrow, and are found in all epidermal regions, but are mainly located in the middle of this layer. They play a significant role in the immune response of the skin, acting as antigen-presenting cells (Igyarto and Kaplan, 2013).

Continuing their transition to the surface the cells continue to fatten, lose their nuclei and their cytoplasm appears granular at this level. They have now reached the stratum corneum, which is the place where keratinocytes reach their final stage of maturation. The stratum corneum is made up of layers of hexagonal-shaped, non-viable cornified cells known as corneocytes. In most areas of the skin, there are 10-30 layers of stacked corneocytes, with the palms and soles having the most. Each corneocyte is surrounded by a protein envelope and is filled with water-retaining keratin proteins. The cellular shape and orientation of the keratin proteins add strength to the stratum corneum. Surrounding the cells in the extracellular spaces are stacked layers of lipid bilayers. The resulting structure provides the natural physical and water-retaining barrier of the skin. The corneocyte layer can absorb three times its weight in water but if its water content drops below 10%, it no longer remains pliable and cracks. The movement of epidermal cells from the basal layer to the stratum corneum usually takes about 28 days, which is known as the epidermal transit time.

Chapter 1 - Introduction

Figure 1.11 Structure of the epidermis

Diagram showing the epidermis, which is divided into five layers. From outside to inside, these layers include; the stratum corneum, which is the outmost layer; the stratum lucidum; the stratum granulosum; stratum spinosum, Lagerhans cells are also located here and the stratum basale which also contains melanocytes (pigmented cells) and Merkel cells (mechanoreceptors) .

Chapter 1 - Introduction

Epidermal stem cells

Given that the terminally differentiated cells of the epidermis cannot divide, their replacement depends on maintaining a self-renewing population of stem cells. The best characterised stem cell population lies in a region of the hair follicle known as the bulge, but there are also stem cells in the basal layer of the interfollicular epidermis and, potentially, the sebaceous gland (Kretzschmar and Watt, 2014). Stem cells are not the only epidermal cells that are capable of proliferating. The stem cell progeny that are destined to terminally differentiate can first undergo a few rounds of division, during which time they are known as transit-amplifying cells (Niemann and Watt, 2002). At present, it is unclear whether transit-amplifying cells have multilineage differentiation potential or whether they are lineage-restricted.

Barrier function and skin desquamation

The skin is a complex metabolically active organ, which performs important physiological functions (summarised in Table 1.4). As the viable cells move towards the stratum corneum they begin to cluster proteins into granules in the granular layer. The granules are filled with fillagrin, a protein that becomes complexed with keratin to prevent its breakdown by proteolytic enzymes. As the degenerating cells move towards the outer layer, enzymes break down the keratin-fillagrin complex. Fillagrin forms on the outside of the corneocytes while the water-retaining keratin remains inside. When the moisture content of the skin reduces, fillagrin is further broken down into free amino acids by specific proteolytic enzymes in the stratum corneum. The breakdown of fillagrin only occurs when the skin is dry in order to control the osmotic pressure and the amount of water it holds. In healthy skin, the water content of the stratum corneum

Chapter 1 - Introduction is normally around 30%. The free amino acids, along with other components, such as lactic acid, urea and salts, are known as natural moisturising factors and are responsible for keeping the skin moist and pliable due to their ability to attract and hold water

(Jensen and Proksch, 2009; Nishifuji and Yoon, 2013; Proksch et al., 2008). However, the skin undergoes a natural degree of continuous water loss, which is made up of sweat loss, desorption, and diffusional water loss. Transepidermal water loss (TEWL) is the constitutive steady-state loss of water vapour from the skin, theoretically excluding desorption and sweat gland activity. It is considered to be the result of passive diffusion through the skin; the diffusional flux obeying the physicochemical laws of passive transport through a membrane. The stratum corneum is the most important rate limiting step for the transport of water to the exterior (Blank, 1965). However the stratum corneum is not an ideal membrane: it is perforated by skin appendages, and its properties as a barrier are dependant both on terminal differentiation and on environmental influences. The desquamating stratum corneum is nonliving and constantly renewing, and therefore the formation of a good layer may be hampered by skin diseases affecting epidermal differentiation, such as atopic dermatitis. External insults may also influence epidermal differentiation by damaging keratinocytes, and they may also have direct effects on corneocytes as well as on the intercellular domain, resulting in loss of barrier capacity (Elsner et al., 2001; van der Valk et al., 1985).

Chapter 1 - Introduction

Table 1.4 Functions of the skin

Function Structure/cell involved Protection against:

 chemicals, particles stratum corneum  ultraviolet radiation melanocytes  antigens, haptens Lagerhans cells  microbes Lagerhans cells

Preservation of a balanced stratum corneum internal environment

Prevents loss of water, stratum corneum electrolytes and macromolecules

Shock absorber dermis and subcutaneous fat Strong, yet elastic and compliant

Temperature regulation blood vessels eccrine sweat glands

Insulation subcutaneous fat

Sensation Specialised nerve endings

Lubrication Sebaceous glands

Calorie reserve subcutaneous fat

Vitamin D synthesis keratinocytes

Body odour/pheromones apocrine sweat glands

Psychosocial, display Facial skin, lips, hair and nails

Chapter 1 - Introduction

1.4 Aims of this thesis

Although there are some clues concerning the molecular role of GTF2IRD1 in vitro, and the analysis of mouse models has provided some insights regarding the processes that are affected by the lack of Gtf2ird1, the cellular and physiological events, of which this protein forms a part, still remain to be understood.

This work seeks to drill deeper into the function of GTF2IRD1 by investigating the impact of a targeted Gtf2ird1 deletion on the neurological and craniofacial phenotypes in mice, using direct physiological and cellular analysis methods. This work also seeks to contribute to a greater understanding of the genetic basis of WBS characteristics by relating the phenotypic consequences of the mouse knockout with the human pathology.

Specifically, by utilising Gtf2ird1-/- mice, this study aims to address the role of

GTF2IRD1 in:

I. the auditory system by assessing its expression in the ear, as wells as

determining the behavioural and physiological hearing responses (chapter 3)

II. the skin covering the face and its role in controlling cell proliferation,

differentiation and patterning (chapter 4)

III. the molecular mechanisms and biological pathways underlying the abnormalities

of the Gtf2ird1-/- lip tissue (chapter 5)

CHAPTER 2 - MATERIALS AND METHODS

Chapter 2 – Materials & Methods

2.1 Materials

A list of key reagents is presented in Table 2.1. All the primary antibodies utilised in this work are catalogued in Table 2.2 and a list of all the oligonucleotides needed for PCR are listed in Table 2.3.

Table 2.1 Commercially available reagents and kits.

Reagent Manufacturer 30 % Acrylamide/bis-acrylamide 29:1 solution Bio-Rad Laboratories Acepromazine Ceva Animal Health Bicinchoninic acid (BCA) protein assay kit Thermo Scientific Pierce Clarity bioluminescence detection reagent Bio-Rad Laboratories Goat Anti Mouse Alexa Fluor 568 2ry antibody Molecular Probes Goat Anti Mouse IgG HRP 2ry antibody DAKO Goat Anti Rabbit IgG Alexa Fluor 555 2ry antibody Molecular Probes Goat Anti Rabbit IgG Alexa Fluor 488 2ry antibody Molecular Probes Immun-Blot PVDF Membrane Bio-Rad Laboratories Ketamine Ceva Animal Health Luminata Forte, Western HRP chemiluminescence Merck Millipore substrate M-MLV Reverse Transcriptase, RNase H Minus, Point Promega corporation Mutant Oligo(dT) 15 Primer Promega Corporation OCT compound Tissue Tek ProLong Gold Antifade reagent with DAPI Molecular Probes Protease inhibitor cocktail (for use with mammalian Sigma cell and tissue extracts, DMSO solution)

Chapter 2 – Materials & Methods

Table 2.1 (continued) Commercially available reagents and kits.

Reagent Manufacturer Rhodamine- Phalloidin Thermo Fisher SB431542 Activin receptor-like kinase inhibitor Tocris SsoFast EvaGreen real-time PCR Supermix Bio-Rad Laboratories, Inc Taq DNA Polymerase with ThermoPol Buffer New England Biolabs, Inc

Tri-reagent Sigma

Xylazine Ceva Animal Health

Table 2.2 Primary antibodies.

All antibodies are listed according to the target protein (Antibody column).

Dilutions are shown according to the type of experiments carried out (WB: Western blot; IF: Immunofluorescence).

Dilution Antibody Source Product # Host IF / WB α6-Integrin Abcam 75737 rabbit 1:500 β-galactosidase Molecular Probes 63365 mouse 1:500 FGF7 (clone F-9) Santa Cruz sc-365440 mouse 1:1000 Keratin 10 Covance PRB-159P rabbit 1:1000 Keratin 14 Covance PRB-155P rabbit 1:1000 Ki67 Neomarkers RM-9106-S1 rabbit 1:200 Neurofilament-200 Sigma N4142 rabbit 1:1000

Table 2.3 Oligonucleotide primers utilised for genotyping

Mouse line reaction Primer ID Primer sequence (5’-3’) Gtf2ird1tm1Hrd KOF3 (forward) TGATGGTTGATGCCACTCAC MDEx2R (reverse 1) ATGGAGTCTAAGGCGGACAC KOSCREEN (reverse 2) GAATTCCAGGTGACCCTGGGAG Gtf2ird1tm2(LacZ)Hrd KOF3 (forward) TGATGGTTGATGCCACTCAC MDEx2R (reverse 1) ATGGAGTCTAAGGCGGACAC NLSR (reverse 2) GCGGGGGTCTTCTACCTTTC

Chapter 2 – Materials & Methods

Table 2.4 Oligonucleotide primers utilised for RTqPRC.

PCR Forward Primer (5’-3’) Reverse Primer (5’-3’) target Hprt TCAACTTGCGCTCATCTTAG AGCTTGCTGGTGAAAAGGAC Fgf7 GGAAATCAGGACCGTGGCAG GTGTCCATTTAGCTGATGCATAG Fgfbp1 GGTCCAAGAAGCTGTCTCCA GTCGCCTGTAACATGTTGAG Lhx2 CACCAGCTTCGGACAATGAA CTGCAGCGTGGCATCTGAC Tgfb2 CTTGGATGGAAATGGATCCAT GCTGTTCGATCTTGGGCGTAT Wnt11 CAGCCACGAAGGTGGTACAC CACATGAGGTCGCAGCTGTC

2.2 Methods

Mice

The Gtf2ird1tm2(LacZ)Hrd and Gtf2ird1tm1Hrd mouse lines were created using similar targeting strategies (Palmer et al., 2007; Palmer et al., 2010). In both cases, the mouse Gtf2ird1 allele was targeted in 129R1 embryonic stem cells (ES cells) using homologous arms flanking exon 2, inserted into a plasmid containing a PGK-neo selection cassette flanked by LoxP sites. In the case of the Gtf2ird1tm2(LacZ)Hrd mouse line, exon 2 was replaced with a LacZ cDNA fused to a nuclear localisation signal

(NLS) (Kalderon et al., 1984). Once germ line transmission had been established, the neomycin selection cassette was removed using cre/lox excision, by interbreeding with

C57BL/6J mice carrying the Tg(CMV-Cre)1Cgn transgene (Schwenk et al., 1995). The resulting progeny were continuously backcrossed to C57BL/6J mice for more than 10 generations and both lines were maintained on this genetic background.

Gtf2ird1tm2(LacZ)Hrd mice were used for expression analysis by detecting either

LacZ enzymatic activity or the β-galactosidase protein. Specimens used for these studies were obtained from mice produced by the mating of heterozygous Gtf2ird1tm2(LacZ)Hrd with wild type animals. The Gtf2ird1tm1Hrd line was

Chapter 2 – Materials & Methods used to perform all the functional analyses described in this thesis and homozygous and heterozygous knockouts are hereinafter referred to as Gtf2ird1-/- and Gtf2ird1+/- respectively. These mice were produced in the same litters from heterozygous parents and housed together in identical conditions. All experimental procedures were approved by the Animal Care and Ethics Committee (ACEC) at UNSW Australia under the protocols 09/46A, 12/45B and 15/53A.

Genomic DNA extraction from mouse tail samples and genotyping

DNA lysis buffer consisting of 100 mM Tris-HCl pH 8.5, 5 mM EDTA, 0.2%

SDS, 200 mM NaCl and 100 g/mL proteinase K (NEB) was added to a 2-3 mm tail biopsy and incubated at 65 oC overnight on a heating block. After incubation, any undigested tissues were spun down at 14,000 rpm for 10 min. The supernatant was mixed with 500 l of absolute ethanol and centrifuged at 14,000 rpm for 5 min to precipitate the genomic DNA. The DNA pellet was washed with 70% ethanol and centrifuged again at 12,000 rpm for 7 min. Ethanol was removed and the DNA pellet was air-dried for 5 to10 min and resuspended in 50-100 l of TE buffer (10 mM Tris-

HCl pH 7.5, 1 mM EDTA pH 8.0). Genotypes of the mice were determined utilizing 1l of each DNA sample by PCR reaction as previously described (Palmer et al., 2007;

Palmer et al., 2010) using the primers listed in Table 2.3.

Agarose gel electrophoresis of PCR genotyping samples

PCR products were analysed by agarose gel electrophoresis. 1% agarose gels were made by dissolving 1g of agarose powder in 100 ml of 1x Tris-borate-EDTA

Chapter 2 – Materials & Methods

(TBE) buffer (Green and Sambrook, 2012). The mixture was heated in a microwave oven until the agarose was completely dissolved. Once the solution was cooled down, ethidium bromide solution was added to a final concentration of 0.5 µg/ml and the solution poured into a sealed gel casting mould. DNA samples were mixed with loading dye (0.04% Bromophenol Blue and 10% glycerol in water) before loading into the wells, together with a 100 bp ladder as the size marker. Electrophoresis was carried out in 1x TBE buffer at 120 V (PowerPacTM Basic Power Supply, Biorad). Results were usually visualised after 30min by UV trans-illumination visualization and documented using a Gel Doc EZ System (BioRad).

Transcardial perfusion

Mice were anesthetised by IP injection with the ketamine-xylazine mixture

(dosage: 40 mg/Kg ketamine; 8 mg/Kg xylazine). Deep anaesthesia was confirmed by no withdrawal reflex response when pinching the foot with the tip of a forceps. The abdomen was swabbed with 70% ethanol and the mouse pinned on its back to a cork board placed on a deep plastic tray. Skin was removed to access the thoracic cavity by cutting through the ribs. A 23G butterfly needle (26G in young mice) was inserted through the apex of the heart into the cavity of the left ventricle. An incision in the right atrium was made at this stage to allow blood and perfused solutions to flow out of the body. Using a Minipulse 3 peristaltic pump set at 30-40 rpm, 20 mL of cold 1x PBS was flushed through, followed by 20 to 30 mL of cold 4% PFA fixative diluted in 1x PBS (1 mL per gram of mouse body mass). The body was then checked for stiffness and a pale appearance of the visceral organs, indicating successful systemic perfusion. Tissues were dissected as needed.

Chapter 2 – Materials & Methods

Immunofluorescence analysis in the cochlea

Mice were anaesthetised and transcardially perfused with cold 0.1 M phosphate- buffered saline (PBS) and cold 4% paraformaldehyde (PFA) as described above.

Tympanic bullae containing the cochleae were dissected under the microscope and locally perfused with 4% PFA through the round and oval windows. Samples were kept in 4% PFA at 4 °C overnight and then decalcified in 8% EDTA in 0.1M phosphate- buffer (PB, pH 7.4) for 2 weeks at 4 °C. Decalcified cochleae were rinsed in PB, transferred to a 30% sucrose solution for incubation overnight and then transferred to a

1:1 mixture of 30% sucrose and OCT compound (TissueTek) for 4 hours before being snap-frozen in OCT using liquid nitrogen.

Immunofluorescence analysis was performed using 30 µm floating cryo-sections in 24-well plates (Nalgen Nunc Int., Rochester, NY, USA) containing PB. Sections were permeabilised for 1 hour with 1% Triton X-100 in PB and blocked with 5% normal goat serum (NGS) and 5% bovine serum albumin (BSA) in PB. The primary anti-β-galactosidase antibody (Molecular Probes cat# 08633651) was diluted 1:500 in blocking solution and applied overnight at 4 °C. The primary anti-Neurofilament 200 antibody (Sigma cat# N4142) was diluted 1:1000. In control experiments, the primary antibodies were omitted. Sections were washed three times with PB containing 0.2%

Triton X-100, followed by 1 hour incubation in the secondary antibody solution containing goat anti-rabbit Alexa Fluor 488 (Molecular Probes) at 1:1000 in PB. Some of the sections were stained with Rhodamine-Phalloidin (Thermo Fisher cat# R415) at

1:250 for actin filament staining of the hair cells. Sections were washed 3 times with PB and mounted with ProLong Gold Antifade Reagent containing DAPI (Molecular Probes cat# P-36931). Images were captured using a confocal microscope (Zeiss T710).

Chapter 2 – Materials & Methods

Acoustic startle response and prepulse inhibition of acoustic startle.

Acoustic startle response (ASR) and prepulse inhibition (PPI) testing were performed by a collaborator at the Florey Neuroscience Institutes as described previously (Burrows et al., 2015; McOmish et al., 2008; van den Buuse et al., 2003; van den Buuse et al., 2005). Briefly, both behaviours were assessed using automated SR

Lab startle equipment (San Diego Instruments, San Diego, CA, USA). Each unit consisted of a plexiglas cylinder mounted on a platform, underneath which, a sensitive piezo-electric sensor measured movement. Startle response to pulse was measured on

SR Lab software. Mice (10 per genotype; 4 male, 6 female) were placed into the cylinder and the platform slotted into a sound-attenuating console with roof-mounted speakers. Background ‘white noise’ (70 dB) was transmitted through the speakers before, during and after prepulse and pulse stimuli. Each session consisted of 100 trials where the first and the last ten trials consisted of ‘pulse only’ startle-inducing stimuli of

115 dB, lasting 40 ms. The central 80 pulses were a pseudorandomized program of 20

P115 trials startle only, ten measured recordings with no stimulus to measure baseline movement, and 4, 8 and 115 dB pulses preceded by a prepulse of 5-45 dB above the baseline ‘white noise’ to measure the animals’ capacity for sensorimotor gating. In these instances, the prepulse preceded the pulse by 100ms and lasted for 20 ms. One day before testing, the mice were habituated to the chamber and the procedure by the same pulse alone and prepulse program.

Auditory brainstem response (ABR)

Mice were 8 to 10 weeks old at the time of hearing testing by ABR. Auditory brainstem response (ABR) was evaluated in Gtf2ird1-/- and wild type littermates at 8-10

Chapter 2 – Materials & Methods weeks old. All ABR measurements were performed in a sound attenuating chamber

(Sonora Technology, Japan). Mice were anaesthetised with a cocktail of ketamine (40 mg/kg), xylazine (8 mg/kg) and acepromazine (0.5 mg/kg). Supplemental doses of

K/X/A cocktail were given as necessary throughout the experiments (typically after 30 min). Body temperature was maintained using a heating pad regulated via a rectal probe and ophthalmic ointment was applied to the eyes of the animal to prevent corneal drying. ABRs were recorded by placing fine platinum subdermal electrodes at the mastoid region of the right ear (active electrode), scalp vertex (reference electrode) and on the lower back (ground electrode). The acoustic stimuli for ABR were produced and responses recorded using a Tucker-Davis Technologies auditory physiology System III workstation (TDT, FT Lauderdale, FL, USA), controlled by BIOSIG32 software (Ver.

4.1.1; TDT). ABR potentials were evoked with digitally-produced 5 ms tone pips (0.5 ms rise/fall time, delivered at 10/sec) delivered by an electrostatic speaker (EC1; TDT).

Signal recordings (10 ms duration) in response to broadband click (100 µs) or tone pip

(4, 8, 16, 24 and 32 kHz) sound stimuli were amplified, filtered and averaged up to 512 times. For each sound stimulus, sound pressure level (SPL) signals were decreased in 5 dB steps, starting from 70 dB, down to 10 dB below the threshold level. The intensity at which an ABR complex (waves I-V) was no longer distinguishable from the noise floor

(100 nV) was defined as the ABR threshold. All response values (thresholds, peak to peak amplitudes and latencies) were determined by off-line analysis. Speaker calibration was carried out using a calibrating microphone connected to the TDT system via its companion preamplifier (Aco Pacific, Belmont, CA, USA), with SIGCALRP and

RPVDSEX software (TDT).

Chapter 2 – Materials & Methods

Distortion product otoacoustic emissions

The cubic distortion product otoacoustic emissions (DPOAE at 2f1-f2) were recorded using a custom-made ear canal probe containing a microphone (ER-B10+,

Etymotic Research, IL, USA). Two EC1 speakers delivered the primary tones (f1 and f2; f2/f1 ratio = 1.25) of 8, 12, 16, 24 and 32 kHz (168 ms duration, 6/sec from 0 to 80 dB in 5 dB increments). The f1 and f2 amplitudes were equivalent. The auditory canal signals were analysed by fast Fourier transformation with 50 traces per sound level being averaged. The DPOAE threshold for a specific tonal level was defined as the lowest SPL at which a 2푓1 − 푓2 DPOAE was detectable above noise floor.

Magnetic resonance imaging (MRI) analysis

MRI was performed using a 9.4-T Bruker biospec 94/20 Avance III micro- imaging system (Bruker, Karlsruhe, Germany) at the Biological Resources Imaging

Laboratory (BRIL, UNSW). The system was equipped with a 15mm quadrature receive/transmit RF-coil and was running Paravision 5.1 system control software. For image acquisition, a 3D multi-gradient echo sequence (MGE) was used in coronal slice orientation with 15 echo times. Parameters used for anatomical MRI scans were: TR =

100ms, 16 echos TE= 3/60ms, first echo time = 2.7ms, echo spacing delta TE = 3.5ms, flip angle = 40-degree, FOV = 1.5 × 1.5 × 0.8 cm, matrix = 200 × 200× 106, and 2 averages, total acquisition time for this MRI sequence was 1 hour and 10 min per scan.

All data were analysed using 3D slicer 4.4.0 software.

Chapter 2 – Materials & Methods

LacZ enzymatic detection

Detection of LacZ enzymatic activity (or β-galactosidase activity) was used as a means to determine the expression of the Gtf2ird1 gene in the Gtf2ird1tm2(LacZ)Hrd mice.

The following procedure was performed for whole mount embryo staining as well as fresh snap-frozen cryo-sectioned tissues. After dissection of the embryo or cryo- sectioning of the snap-frozen tissues, the samples were briefly rinsed in cold 1xPBS and rapidly fixed by immersion in cold 0.5% glutaraldehyde for 5-30 minutes at room temperature. The fixation time depended on the size of the tissue to be stained. Fixed tissues were then rinsed with cold 1x PBS and immersed in fresh working solution (5 mM K4Fe(CN)6, 2 mM MgCl2, 5 mM K3Fe(CN)6, 1 mg/mL X-Gal in 1x PBS) at 37

°C. Blue colour staining was generally developed after 4 to 5 hour of incubation.

However, larger samples required overnight incubation at 37 °C to allow complete dye penetration. Samples were usually examined and photographed immediately after.

However, in some cases the stained tissues were kept in the fridge stored in 2% PFA solution for later analysis. The developed colour remained in the sample since the precipitated blue indole is not soluble in water.

Immunofluorescence in skin sections

10 µm skin cryo-sections were allow to dry for 5-10 min and rinsed twice with

PBS to remove excess OCT compound. Sections were then fixed and permeabilised for

15 min in 4% PFA, 0.25% Triton-X 100. After fixation, sections were incubated with blocking buffer (5% normal goat serum, 5% bovine serum, 1% BSA, 0.2% Triton in 1x

PBS) for 1 hour at room temperature. Primary antibodies were incubated in blocking solution overnight at 4 °C in a humid chamber. In order to reduce background

Chapter 2 – Materials & Methods fluorescence, sections were washed twice for 30 min each with 0.2% Triton in 1x PBS.

An extra 30 min wash with 1xPBS was then performed. Detection was carried out using secondary antibodies conjugated to Alexa Fluor Dyes (Molecular Probes) incubated for

45 min in a dark chamber. All samples were then incubated for 15 min with DAPI in

PBS and ProLong Gold Antifade (Molecular Probes) was used as the mounting media for all preparations.

Whole mount skin clearing technique

Skin was cleared for immunostaining with a modified CLARITY protocol, with the assistance of Dr. Andy Liang at Neuroscience Research Australia (NeuRA), who is an expert at these techniques. CLARITY refers to the transformation of intact biological tissue into a hybrid form in which specific components are replaced with exogenous elements that provide accessibility or functionality. Initially, the term was an acronym to describe the Clear Lipid-exchanged Acrylamide-hybridized Rigid

Imaging/Immunostaining/In situ hybridization-compatible Tissue-hYdrogel. However,

CLARITY is more inclusively defined because infused elements need not be exclusively hydrogel monomers or acrylamide-based, and the properties of infused elements may be adjusted for varying degrees of clarity, rigidity, macromolecule- permeability or other types of functionality. To clear the skin tissue, 12 week old mice were anaesthetized and transcardially perfused with 1x PBS and 4% PFA as described above. Skin patches were then dissected and incubated in monomer solution (4%

Acrylamide, 0.05% Bis-acrylamide, 0.25% VA-044 Initiator, 4% PFA in 1x PBS) at 4

°C for 3 days. Solution temperature was then increased to 37 °C to initiate polymerization. After 3h at 37 °C, hydrogel-embedded skin samples were placed in a

Chapter 2 – Materials & Methods custom-built organ-electrophoresis chamber. The clearing solution (200 mM Boric

Acid, 4% SDS, pH 8.5) was then circulated through the chamber using the temperature- controlled water circulator, with 10-60 V applied across the tissue at 37-50 °C for 7-8 weeks to clear the sample. After clearing, samples were incubated in PBS at 37 °C for 2 days to remove the excess SDS. Tissues were then used for immunostaining, which included a 2 day 1ry antibody incubation and longer washes between antibodies.

Skin barrier function assessment

The Toluidine Blue skin permeability dye exclusion assay was performed as described previously (DiTommaso et al., 2014; Hardman et al., 1998). The assay depends on barrier-dependant access of toluidine blue dye into untreated skin. Unfixed, untreated dead embryos or euthanased new born pups were collected and transferred through a methanol series with immersion for 1 minute each: 25% methanol in water,

50% methanol in water, 75% methanol in water, 100% methanol, 75% methanol in water, 50% methanol in water, 25% methanol in water and finally equilibrated in PBS.

All reagents were chilled. Specimens were then exposed to 0.1% Toluidine Blue solution in water for 2 minutes and distained in 1x PBS pH 7.4. After staining, the embryos/pups were photographed using a microscope with transmitted and surface illumination. Scanned images were processed with Adobe Photoshop to remove unwanted background.

The trans-epidermal water loss (TEWL) assay, that allowed quantification of the degree of barrier disruption, was also performed as described previously (DiTommaso et al., 2014; Smyth et al., 2008). 3 mm2 skin patches from lip and dorsal skin from 2 month old mice were harvested to evaluate permeability over a period of 300 minutes

Chapter 2 – Materials & Methods after dissection. Skin patches were stored in a dehumidifying chamber and their weight was measured every half an hour using an electronic analytical scale. Water loss was expressed as mg/cm2.

Total RNA extraction

Mice were sacrificed by cervical dislocation and the areas around the lips were carefully cleaned with 75% ethanol and DEPC water. The top lip was then dissected and immediately immersed in 2 mL of cold TRI-reagent. Tail biopsies were collected for genotyping purposes. Lip tissues were homogenised for approximately 30 sec (or until fully homogenised) using a T10 Ultra-Turrax homogeniser (Themo-Fisher Scientific).

RNA purification was carried out following the manufacturer’s instructions, which included a chloroform phase separation and ethanol precipitation. RNA was resuspended in 70 μL of RNase-free water. RNA samples were assessed for quantity and quality using a NanoDrop UV spectrophotometer (Thermo Fisher Scientific Inc), conventional RNA agarose gel electrophoresis and by using a Bioanalyzer (Agilent

Technology Inc).

RNA-formaldehyde gel electrophoresis

Gel apparatus and glassware were cleaned with RNase-Zap (Ambion, Austin,

TX) and DEPC water. Disposable RNase-free filtered tips and tubes were used to minimise RNA degradation. The agarose gel for RNA electrophoresis was prepared in a total volume of 100 mL as follows: 1% (w/v) agarose was dissolved and boiled in 70 mL of 1X MOPS buffer (40 mM MOPS, 10 mM sodium acetate, 1 mM EDTA disodium, pH 7 adjusted with glacial acetic acid) and 10 mL of filter-sterilised water. In

Chapter 2 – Materials & Methods the fume hood, 20 mL of formaldehyde was added into the molten agar solution, cooled to approximately 60 oC and poured into the casting tray. Electrophoresis was carried out in a tank filled with 1X MOPS running buffer. Typically, 2 to 5 g of RNA was needed to assess the quality of RNA. For each RNA sample, loading buffer containing 2.5 l of

10X MOPS, 3.5 l of formaldehyde, 10 l of deionised formamide, 1 l of ethidium bromide (1 mg/mL) and 1 l of Bromophenol Blue was added. The RNA/buffer mixtures were heated at 65 oC for 5 min to denature any secondary structure of the

RNA, cooled on ice for 2 min and then loaded onto the gel. The gel was run at 70-100 V until the dye had migrated 2/3 the length of the gel. RNA bands were visualised on a

UV transilluminator. Intact total RNA has sharp 28S and 18S ribosomal RNA bands with an approximate intensity ratio of 2:1.

Reverse transcription

First-strand cDNA synthesis was carried out using the RNase H Minus, Point

Mutant M-MLV Reverse Transcriptase enzyme system, with Oligo dT primers (both sourced from Promega Corporation), using 1 μg of total RNA as the template, according to the manufacturer’s instructions.

Transcriptome sequencing

Six mRNA (RNA-Seq) libraries were prepared from 3 wild type and 3 Gtf2ird1-

/- mice following the standard Illumina protocol. The six RNA-Seq libraries were sequenced on the Illumina HiSeq2000 platform at the Ramaciotti Center for Genomics

UNSW, to produce over 60 million, 100 nucleotide paired-end reads per sample. Read

Chapter 2 – Materials & Methods quality was reviewed with FASTQC (version 0.10.1). The base quality was good, however, there was some instability in the Per Base Sequence Content over the first 10 bases of all samples as previously reported in Illumina-generated sequences (Hansen et al., 2010).

RNA-Seq: differential expression and functional analyses

The RNA-Seq analysis was performed at the Systems Biology Initiative,

UNSW. The RNA-Seq reads were mapped to the Ensembl Mus musculus genome

(GRCm38) provided by Illumina iGenomes (downloaded from cufflinks.cbcb.umd.edu/igenomes.html). Mapping was performed with Tophat (v 2.0.8)

(Kim et al., 2013a) calling Bowtie2 (v 2.1.0) (Langmead and Salzberg, 2012) using the default settings. The HTSeq-count (Python package HTSeq, python v 2.7.3) was used to generate counts of reads uniquely mapped to annotated genes using the GRCm38 annotation gtf file. HTSeq-count was run with the options, mode=union, –t = exon, –I = gene_name.

The differential expression analysis was performed using two different count based methods, edgeR (v 3.8.6) (Robinson et al., 2010) and DESeq2 (v 1.6.3) (Anders and Huber, 2010), available as Bioconductor packages. Tables of raw counts generated using HTSeq-count (described above) were used as input in both analyses. In the edgeR analysis, low count transcripts were excluded and only those genes with at least 1 count per million (cpm) in at least 3 samples were used for analysis. A normalization factor was calculated using the trimmed mean of M values (TMM) method (Robinson and

Oshlack, 2010) and the dispersion parameter for each gene was estimated using the quantile-adjusted conditional maximum likelihood (qCML) method, appropriated for

Chapter 2 – Materials & Methods experiments with a single factor. The functions estimateCommonDisp() and estimateTagwiseDisp() were used to estimate dispersion. Following this, differential expression was using the exact test based on qCML methods. The Benjamini-Hochberg correction was used with a false discovery cut-off of 0.1. DESeq2 uses a generalized linear model (GLM) to assess differential expression. Dispersions were estimated using a Cox-Reid adjusted profile likelihood and the Wald test for significance of GLM.

Automatic filtering was incorporated to exclude low abundance genes in the testing process.

The BiNGO plug-in to Cytoscape (Saito et al., 2012) was used to investigate the functional associations of genes found to be either down-regulated or up-regulated. The set of 14,526 genes was used in our differential expression analysis (background list).

Of these, 13,376 genes were used in the gene ontology analysis. The Gene Ontology

(GO) terms in the category Biological Process was tested for overrepresentation using the hypergeometric test and P values were corrected using the Benjamini & Hochberg

FDR correction (Benjamini et al., 2001). Enriched GO terms were selected using a corrected P value of 0.01.

MetaCoreTM from Thomson Reuters was used to conduct functional analysis of the differentially expressed genes including pathway analysis. MetaCoreTM comprises a suite of software and an extensive database, which contains manually curated information on proteins, genes, complexes, metabolites, RNA and DNA and their interactions gleaned from the published literature.

Chapter 2 – Materials & Methods

Quantitative real time-PCR (qPCR)

10 or 20 ng of cDNA was used as a template for qPCR using the EvaGreen dye- based system SoFast EvaGreen Supermix (Bio-Rad Laboratories, Inc). Reactions were set up to a total volume of 20 μL according to the product protocol and performed on the Stratagene MX3005P qPCR system (Agilent Technologies). Each reaction was set up in triplicate for the target gene under test. Duplicate reactions were also set up with an identical amount of template using primers designed against mouse Hprt

(hypoxanthine phosphoribosyltransferase 1) as a housekeeping gene reference standard.

For all RTqPCR assays, the efficiency of the different primer sets were tested by establishing a standard curve using serial dilutions of a cDNA pool made by combining samples of all the templates used in each experiment. MxPro QPCR Software was used to analyse the dissociation and amplification curves of every experiment and to obtain the threshold cycle values (Ct). Data were then analysed in Microsoft Excel for quantitation of target gene transcript relative to the reference standard using the delta delta Ct (dCT) method (Livak and Schmittgen, 2001). Fold expression was calculated as

2-ddCt, and t-test analyses were performed on the dCt values for the different groups of samples.

Total protein extraction from lip tissue.

All procedures were performed on ice and all solutions were chilled to 4°C before use unless otherwise indicated. Total protein extracts from top lip tissue were prepared by homogenising tissues in 5mL ice-cold RIPA buffer (20 mM Tris pH 7.4,

1% Triton X-100, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA) per gram of tissue using a Precellys 24 homogeniser with a mounted Cryolys cooling apparatus. Protease

Chapter 2 – Materials & Methods inhibitors, consisting of EDTA-free complete protease inhibitor cocktail (1 tablet per 50 mL), 1 mM PMSF and 20 µg/mL leupeptin, were added just prior to homogenisation.

The homogenate was centrifuged at 14,000 rpm for 15 min at 4 °C and the cleared protein supernatant was collected. Lysates were stored at -20 °C and used for protein quantification and western blotting analysis.

Protein quantification

To determine the protein concentration of the cell lysates, a BCA protein assay was performed (Pierce, Thermo Fisher Scientific Inc.) following the standard manufacturer’s instructions for a 96-well format assay. A standard curve was prepared using different dilutions from a 2 mg/mL bovine serum albumin stock. Absorbance was measured at 562 nm in a SpectraMax microplate reader (Molecular Devices, LLC). The protein concentration of each sample was calculated by interpolating the sample A562 against the absorbance readings of the known standards.

SDS-polyacrylamide gel electrophoresis (SDS-PAGE)

Proteins were separated and analysed by SDS-PAGE using a BioRad mini-

Protean 3 system (Bio-Rad Laboratories). 8% or 10% Acrylamide/bisacrylamide gels were used according to the size of the proteins of interest. 0.1% SDS and 0.375 M Tris-

HCl pH 8.8 were incorporated into the solution. Polymerization of the gel was initiated by adding ammonium persulfate (APS) to a final concentration of 0.1% together with

0.1% TEMED. The gel was poured into the pre-washed and prepared mini-Protean plates mounted in the apparatus and covered with isopropanol to allow polymerisation.

Afterwards, the isopropanol was removed and the stacking gel solution (4.4%

Chapter 2 – Materials & Methods acrylamide/bisacrylamide, 0.125 M Tris-HCl pH 6.8 and 0.1% SDS, 0.1% APS and

0.1% TEMED) was poured over the separating gel. Protein samples containing 1x

Laemmli sample buffer and 0.1 M DTT were heated at 95° C for 5 min and then loaded into the wells of the gel. Electrophoresis was carried out in the mini-Protean electrophoresis tank containing running buffer (0.025 mM Tris-base, 0.19 M glycine and 0.1% SDS) using the advancing dye front of the Laemmli as a reference for the running distance. Samples were mixed with Laemmli 6x loading buffer (final concentration 1x) and 1M DDT (final concentration 0.1 M) and heated at 95°C for 5 min before loading into the gel. Electrophoresis was carried out in running buffer (0.025 mM Tris-base, 0.19 M glycine, 0.1% SDS) and a pre-stained protein (BIORAD) marker was loaded next to the first sample of the gel as a molecular weight reference.

Western blotting.

Proteins were transferred to Immun-Blot PVDF Membrane (Bio-Rad

Laboratories) for western blot analysis using standard methods. After transfer, membranes were blocked for 1 hour in blocking solution (TBS/Tween 20 and 5% non- fat milk powder), incubated with the primary antibody for 2 hours at room temperature in the same solution and washed three times for 10 min in TBS/Tween 20. The HRP- conjugated secondary antibody was incubated for 45 min at room temperature in blocking solution, washed as before and the signal was detected using the ECL substrate, Clarity (Bio-Rad Laboratories), and exposed to x-ray film.

Chapter 2 – Materials & Methods

Statistical data analysis

Results are presented as the mean ± S.E.M. as indicated in the figure legends.

Statistical analyses were performed using two-way repeated measures analysis of variance on ranks (Holm-Sidak method) (Sigmaplot®, Systat Software Inc) with confirmation of normal distribution of the data, or a two tailed Student’s T-test as indicated in the corresponding results. In all cases significance was determined as P <

0.05. P values are depicted in results figures as follows: * p≤ 0.05, ** p ≤ 0.01.

CHAPTER 3 - THE ROLE OF GTF2IRD1 IN THE AUDITORY PATHOLOGY OF WILLIAMS–BEUREN SYNDROME

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.1 INTRODUCTION

Given the genetic basis for WBS and the resulting neuroanatomical patterns, it is reasonable to expect a specific cognitive profile to be associated with the disorder.

The Williams-Beuren Syndrome Cognitive Profile (WSCP) encompasses a unique set of characteristic neurological features and personality traits (Mervis et al., 2000), including mild to moderate intellectual disability, a characteristic visuospatial construction deficit (Hirota et al., 2003) and abnormal motor coordination (Hocking et al., 2008). In addition, patients have a highly distinctive personality profile that includes

‘over-friendliness’ due to a lack of normal social inhibition and enhanced emotional empathy (Doyle et al., 2004), but this is combined with heightened anxiety in response to non-social cues, leading to high rates of phobias (Jarvinen et al., 2013).

Amongst the spectrum of physical and neurological defects in WBS, it is common to find distinctive responses to sound stimuli that include extreme adverse reactions to loud or sudden sounds and a fascination with other sounds that may correlate with the perceived strengths in musical ability commonly associated with

WBS. The strong reaction to loud sudden sounds has been classically described as

‘hyperacusis’. This term aims to describe the extreme adverse behavioural reactions of

WBS patients to sounds that are not normally regarded as loud or aversive and are generally acceptable to others (Blomberg et al., 2006; Levitin et al., 2005). However, the WBS hearing phenotype is more complex than this. Sounds stimulate a polarization of affect; i.e. an extreme aversion to sounds of one type whilst having strong affinities and fascinations for other sounds, which often includes music (Levitin, 2005) but can also include the sound of household objects such as vacuum cleaners and leaf blowers

(Levitin et al., 2005). The aversive reaction to sounds begins young, as a study of children aged 2-6 years showed that WBS children had increased startle responses, such

Chapter 3 - GTF2IRD1 in auditory pathology of WBS as putting their hands over their ears. However, these responses often began before the sound stimulus was triggered and may be a manifestation of anticipatory anxiety that forms part of the complex behavioural profile of WBS (Gallo et al., 2008).

The word hyperacusis implies that hearing sensitivity is heightened in WBS, whereas hearing tests indicate the opposite: sensorineural hearing loss (SNHL) is frequently found in WBS patients but the functional and genetic basis of this unusual auditory phenotype is currently unknown. Several studies have shown that WBS patients of all ages show increased rates of mild to moderate high-frequency SNHL

(Cherniske et al., 2004; Marler et al., 2005) but the level of incidence is debated, which may partially relate to the testing methodology employed and the age range of the patients (Barozzi et al., 2012). SNHL in WBS has been shown to be progressive and the prevalence may increase up to ~ 80% (Barozzi et al., 2013). While many younger WBS patients seem to have hearing in the normal range, it is widely agreed that analysis of otoacoustic emissions from the cochlea of these patients indicates high rates of abnormalities (Barozzi et al., 2012; Barozzi et al., 2013; Marler et al., 2010; Paglialonga et al., 2011), suggesting a degree of ‘cochlear fragility’(Barozzi et al., 2012) associated with altered sound transduction. Otoacoustic emissions are a by-product of the cochlear amplifier mechanism that modulates basilar membrane displacement and is mediated by the outer hair cells (Ashmore and Gale, 2004; Brownell, 1990). Measurement of the distortion product of otoacoustic emissions (DPOAE) is a non-invasive, highly sensitive, objective measure of cochlear function that allows early detection of hearing abnormality with good reliability (Lucertini et al., 2002). Thus rates of SNHL may be low in young WBS individuals but most show evidence of incipient cochlear dysfunction and will later go on to develop SNHL, particularly in the high-frequency range (Barozzi et al., 2013).

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

According to a previous report on Gtf2ird1 expression utilizing histochemical detection of β-galactosidase activity in a Gtf2ird1tm2(LacZ)Hrd knock-in reporter mouse

(Palmer et al., 2007), the spiral ganglion neurons (SGN) exhibit a high and homogeneous expression of Gtf2ird1. These neurons are located in the modiolus of the cochlea in the auditory system, which is a conical shaped central axis in the cochlea made of spongy bone. The abundance of Gtf2ird1 in this uniform neuronal population triggered the thought that it might be possible to measure the physiological impact of

GTF2IRD1 protein loss in a specific neuronal population by conducting a series of sophisticated hearing tests on Gtf2ird1-/- mice. By analysing these consequences in a relatively simple and self-contained system that has a defined physiological readout with no obvious behavioural complications, it may be possible to understand the general consequences of GTF2IRD1 loss on other neuronal cell types and thus gain insights into the cellular and molecular basis of GTF2IRD1 function.

This thesis chapter describes an analysis of the auditory system of the Gtf2ird1-

/- mice as a means to understand the neuronal function of the GTF2IRD1 protein and its possible role in the auditory pathology of WBS. This analysis includes a detailed investigation of Gtf2ird1 expression in the inner ear and an examination of the hearing capacity of Gtf2ird1-/- mice, by evaluating the auditory brainstem response (ABR) and the distortion product of otoacoustic emissions (DPOAE). The results of this experimental work provide a basis for understanding the contribution of GTF2IRD1 to the functional and genetic causes of the WBS hearing phenotype and lead to an evaluation of the practicality of using this model as a means to understand the cellular and molecular consequences of GTF2IRD1 loss in neurons.

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2 RESULTS

3.2.1 Expression of Gtf2ird1 in the cochlea

A number of studies have shown that the expression of Gtf2ird1 is relatively low in adult mice tissues (Bayarsaihan and Ruddle, 2000; Calvo et al., 2001; Palmer et al.,

2007; Tay et al., 2003). However, Gtf2ird1 expression can be detected in virtually all neuronal cell types of Gtf2ird1tm2(LacZ)Hrd mice, but is found in much greater intensity in specific neuronal subsets, including the mitral and granular cell layer in the olfactory bulb, the Purkinje neurons of the cerebellum and neurons in the piriform cortex (Palmer et al., 2007).

Gtf2ird1 expression in sensory structures and the peripheral nervous system is less well defined. However, Palmer et al. (2007) described the expression pattern in the eye of the Gtf2ird1tm2(LacZ)Hrd reporter mice. Using this system, Gtf2ird1 was shown to be highly expressed in the outer nuclear layer of the retina, which contains the nuclei of the rod and cone photoreceptor cell, and was more weakly expressed in the inner nuclear retinal layer. Low expression was also detected in the ganglion cell layer of the retina, the iris and some cells of the anterior epithelium and secondary fibres of the lens.

Analysis of expression in the inner ear of Gtf2ird1tm2(LacZ)Hrd mice was less detailed because it is technically much more demanding to achieve. The organ is encased in the bones of the skull and these preparations merely used sections though the whole adult head. However, based on this limited analysis, the authors described the expression of Gtf2ird1 as only detectable in the spiral ganglion of the cochlea, which contains the cell bodies of bipolar neurons interconnecting the hair cells of the cochlea with the cochlear nucleus. Gtf2ird1 was also detected in the neurons of the lateral

Chapter 3 - GTF2IRD1 in auditory pathology of WBS lemniscus, which connects the cochlear nucleus with brainstem nuclei (Palmer et al.,

2007).

Since the inner ear is a complex structure made up of several compartments and specialized cell types involved in different processes of auditory transduction and processing, imaging the specimen at a higher power and performing cochlear dissection to achieve different perspectives from the samples, would have been more informative to map in detail the expression of Gtf2ird1 in the inner ear. Performing this analysis properly involves the use of harsh decalcification methods to section the cochlea and explore its inner structure in an anatomically conserved specimen, which makes the use of the classical X-gal histochemical detection method impossible. Several factors are known to affect the sensitivity of this detection reaction including substrate availability, the level of enzyme activity, temperature and oxidation potential. In addition, the activity of endogenous β-galactosidase must be minimized through the use of neutral- alkaline buffers or inhibitory substrate analogs such as D-galactono-1,4-lactone (Fire,

1992). The decalcification method requires weeks of incubation in a phosphate buffered solution containing EDTA for calcium chelation, which breaks down the conditions required and destroys the activity of the enzyme. However, the enzyme retains sufficient integrity to be detectable by indirect-immunofluorescence (IIF) using an anti-β- galactosidase antibody, and thus, this becomes the method of choice for IIF analysis of cryo-sectioned inner ear after decalcification of the cochlear bones.

In order to determine the cell populations that express Gtf2ird1 in the inner ear, a detailed expression analysis is presented utilising IIF of nuclear-localized β- galactosidase in heterozygous Gtf2ird1tm2(LacZ)Hrd knock-in mice (Figure 3.1). β- galactosidase was clearly observed in the nuclei of several cell types, together with some background immunofluorescence, mainly associated with the remaining

100

Chapter 3 - GTF2IRD1 in auditory pathology of WBS extracellular matrix of the bone tissue. This could be easily distinguished from the real signal by the lack of nuclear localization and by comparison with control cochlea sections from wild type mice treated identically (Figure 3.1C). Expression was found in multiple cell types with varying levels of intensity; including the epithelial cell layer of

Reissner’s membrane, interdental cells, marginal cells of the stria vascularis and the spiral ganglion neurons of the cochlear nerve. Expression was also detected in the auditory sensory receptor inner and outer hair cells, inner and outer pillar cells, Deiters’ cells, Boettcher cells, and Hensen’s cells (Figure 3.1).

101

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.1 Gtf2ird1 expression in the adult Gtf2ird1tm2(LacZ)Hrd mouse cochlea

(A) β-galactosidase nuclear immunofluorescence (representing Gtf2ird1 expression) was present in the sound transducing inner hair cells (IHC) and outer hair cells (OHC). Expression of Gtf2ird1 was also detected in the inner and outer pillar cells

(I/O PC), Deiters’ cells (DC), cells of Boettcher (BC), interdental cells (IDC) and

Hensen’s cells (HC). A more intense level of immunofluorescence was seen in the spiral ganglion neurons (SGN). Inset shows expression of Gtf2ird1 in epithelial cells of

Reissner’s membrane (RM). (B) The same image overlaid with phalloidin staining (red) of actin filaments, which identifies the IHCs and OHCs, and DAPI (blue) which shows the location of all cell nuclei. (C) Control cochlea section from a wild type mouse showing phalloidin, DAPI and the extent of green fluorescence background with the β- galactosidase antibody, but no specific signal was observed in cell nuclei (SV: scala

102

Chapter 3 - GTF2IRD1 in auditory pathology of WBS vestibuli; SM: scala media; ST: scala tympani; TC: tunnel of Corti). (D) Low magnification immunofluorescence image of cochlear section from a Gtf2ird1tm2(LacZ)Hrd mouse showing Gtf2ird1 expression in the marginal cells of the stria vascularis (STV).

(E) Detailed Gtf2ird1 expression in the afferent auditory spiral ganglion neurons (SGN) that project to the hair cells of the organ of Corti via the osseus spiral lamina (OSL), identified using neurofilament 200 antibody (red). Panels B, C and E also include merged transmitted light images to highlight cochlear morphology. Scale bars: 50 μm.

103

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2.2 Structure of the Gtf2ird1-/- cochlea

The highly specialized and complex inner ear structure is generated through precise genetic controls that operate during development. Many of these genes have been identified and their role illustrated by targeted mutagenesis. Mutations of the transcription factors; Atoh1 (encoding a basic helix-loop-helix family transcription factor) (Bermingham et al., 1999), Sox2 (encoding a Sry-related HMG box transcription factor) (Kiernan et al., 2005), Pax2 (encoding a paired box homeotic transcription factor) (Torres et al., 1996), and Eya1 (encoding a transcriptional coactivator and phosphatase) (Xu et al., 1999) have resulted in severe developmental defects in the inner ear (Schimmang, 2013). Eya1 is a transcriptional regulator and phosphatase, which is required for early inner ear development. Mice lacking Eya1 arrest development at the otic vesicle stage. Moreover, neurogenesis is initiated but neuroblasts fail to form the otic ganglion due to increased apoptosis, demonstrating that it is also required for the maintenance of auditory neurons (Xu et al., 1999; Zou et al.,

2004). Eya1 haploinsufficiency in humans causes branchio-oto-renal syndrome, which corresponds to a series of congenital birth defects that involve malformation of the outer, middle and inner ears, and mild to profound hearing loss, that can be conductive, sensorineural, or both. This evidence indicates that it also plays a role for hair cell formation (Zou et al., 2008). Similar analyses of Sox2 knockouts indicate that it is required for specifying the sensory precursors in the organ of Corti (Kiernan et al.,

2005).

In order to assess the impact of the Gtf2ird1 loss in the gross anatomy of the inner ear, a comparison of the structure of the cochleae dissected from Gtf2ird1-/- and

C57BL/6 control mice was conducted. To evaluate whether Gtf2ird1 is part of the genetic control mechanism controlling normal cochlear development, the cyto-

104

Chapter 3 - GTF2IRD1 in auditory pathology of WBS architecture of the different cochlear components was examined closely to look for evidence of SGN depletion, OHC/IHC loss and any other signs of abnormal anatomy.

In general, Gtf2ird1-/- cochleae showed no obvious abnormalities from lack of the GTF2IRD1 protein (Figure 3.2 A-B). A more detailed morphological analysis was achieved by immunofluorescence analysis using a Neurofilament-200 antibody and fluorescent rhodamine phalloidin in cryo-sections and whole mount preparations. The neurofilament-200 antibody preferentially detects type II cell bodies whose innervations contact the majority of OHCs. Fluorescent rhodamine phalloidin labels the f-actin filaments present in the hair cells, which are easily identifiable within whole mounts of the organ of Corti. Normal structures were observed in low power images of Gtf2ird1-/- cochleae, including mid-modiolar sections showing that the scala tympani, scala media and vestibule were of normal proportions (Figure 3.2 A). Analysis of the Gtf2ird1-/- spiral ganglion indicated that the neuronal cell bodies were homogenously distributed throughout the spiral ganglion bundle, and fibres emerging from the spiral ganglion appeared to protrude normally through the tunnel of Corti to innervate the three planar

OHCs (Figure 3.2 B). Qualitative comparisons (2 WT versus 2 KO) of the fluorescent rhodamine phalloidin labelled hair cells within whole mounts of the organ of Corti, showed no discernible difference in hair cell density in the mid to basal turn region

(Figure 3.2 C).

105

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.2 Histology of Gtf2ird1-/- cochleae confirms normal cytoarchitecture

(A) Low power image of a mid-modiolar cryo-section of the cochlea from a

Gtf2ird1-/- mouse showing neurons detected using neurofilament 200 immunofluorescence (green) and the organ of Corti and structural tissues identified using Alexa Fluor 594 phalloidin fluorescence (red). White arrows indicate the organ of

Corti (O/C) in which the sensory hair cells and supporting cells show high phalloidin fluorescence due to the high abundance of f-actin. The positions of Reissner’s membrane (RM), scala media (SM), scala tympani (ST), scala vestibuli (SV), stria vascularis (STV) and spiral ganglion (SGN) are indicated (scale bar = 100 µm). (B)

Higher power image showing the structural integrity of the organ of Corti and the spiral ganglion. The arrowhead indicates the olivocochlear efferent innervation of the outer hair cells (OHC). The arrow indicates the afferent innervation of the inner hair cells.

The inner hair cells (IHC), outer hair cells (OHC) and spiral limbus (S) are indicated.

106

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

The fluorescent layers are merged with a trans-illumination image layer to allow visualisation of structural detail (scale bar = 50 µm). (C) Surface-mount image of the mid-basal turn region of the organ of Corti using AlexaFluor 594 phalloidin fluorescence, which detects the stereocilia of the three rows of outer hair cells (OHC 1-

3) and the single row of inner hair cells (IHC), which are supported by corresponding inner phalangeal cells (IPC) (scale bar = 50 µm). This analysis indicates that the hair cells are of normal appearance.

107

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2.3 Behavioural analysis of hearing capacity in Gtf2ird1-/- mice

There are two general approaches for assessing hearing function in mice: physiological and electrophysiological measurements and behavioural hearing response analysis. While physiological measurements, such as the distortion products of optoacoustic emissions (DPOAE), and electrophysiological measurements, such as the auditory brain response (ABR), can provide precise physiological parameters, they can only be used to analyse absolute hearing sensitivity and cannot provide information regarding an individual’s ability to experience and discriminate sounds (Hood, 1998).

Therefore, to perform a comprehensive evaluation of an animal’s hearing function and to look for potential hearing disorders, it is also necessary to include behavioural procedures that will help to assess overall auditory capacity.

A number of procedures have been described as valuable tools to assess hearing capacity in mice. The auditory behavioural tests are classified into two types: those that train an animal to respond to sound using conditioning procedures, and those that make use of unconditioned or reflexive responses to sound (Klump, 1995). Unconditioned procedures constitute the simplest methods for assessing hearing in mammals because they take advantage of an animal’s unconditioned responses to sudden loud sounds, without having to engage in lengthy training protocols.

Mice, like other mammals, show an unconditioned motor reaction to sudden loud sounds; a response referred to as the acoustic startle reflex (Hoffman and Ison,

1980). To evaluate hearing capacity in Gtf2ird1-/- mice, two behavioural unconditioned approaches were utilised; acoustic startle reflex and the prepulse inhibition test (PPI) of the startle reflex. The startle reflex is measured by placing an animal in a cage and presenting a startle sound at random intervals when the animal is not moving (Parham

108

Chapter 3 - GTF2IRD1 in auditory pathology of WBS and Willott, 1988). The response is detected and recorded using an accelerometer fitted to the cage. In the PPI test (otherwise known as the reflex modification test), the ability of an animal to detect a particular sound is investigated by determining if the presentation of a loud sound before the startle stimulus modifies the resulting reflex

(Rickye and Henry, 2001). These analyses were performed by a collaborator at the

Florey Institute of Neuroscience and Mental Health as a part of a large battery of behavioural tests that were conducted in order to discover the neurological features of the Gtf2ird1-/- but it was unpublished in the report that summarised those findings

(Howard et al., 2012). Results from both of these tests showed no significant differences between Gtf2ird1-/- and wild type mice. However, a trend indicative of higher auditory thresholds in the Gtf2ird1-/- mice was observed across the whole range of sound levels evaluated in the acoustic startle reflex test (Figure 3.3). This is suggestive of a mildly impaired hearing capacity in the Gtf2ird1-/- mice. However, due to the limitations of the behavioural hearing experiments mentioned before, subtle differences are hard to evaluate by these techniques and more sophisticated methods are called for.

109

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.3 Acoustic startle response and prepulse inhibition (PPI) of acoustic startle response of Gtf2ird1-/- mice

(A) Acoustic startle response of the Gtf2ird1-/- mice compared to wild type littermates when stimuli are 5 to 45 dB over a 70 dB baseline noise (i.e. 75 to 115 dB).

(B) PPI of acoustic startle when prepulses are 4, 8 and 115 dB over a 70 dB baseline noise. Data are presented as the mean of the acoustic startle amplitude response ± SEM

(n=4 male, 6 female per genotype).

110

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2.4 Gtf2ird1-/- mice have a hearing deficit across a broad frequency range

In general, acoustic tests in mice rely on behavioural responses such as freezing time or the analyses described above. However, because of the complex personality traits found in individuals with WBS (such as the extreme reaction to sudden loud sounds), it is possible that mice with WBS-associated mutations may also share these characteristics. The altered anxiety responses previously reported in Gtf2ird1-/- mice

(Howard et al., 2012; Young et al., 2008) support such a notion. These changed responses could confound the analysis of hearing capacity using behavioural measures.

Therefore, we chose to further investigate the consequences of the lack of GTF2IRD1 in the auditory system by utilising the objective physiological measures of auditory brainstem response (ABR) and distortion product of otoacoustic emissions (DPOAE), which are conducted under general anaesthesia and therefore have no, or minimal, behavioural components.

Hearing function was assessed firstly by ABR in a cohort of Gtf2ird1-/-,

Gtf2ird1-/+ and wild type littermates (n=15 for each genotype). Sound-evoked ABR recordings reflect synchronous activation of elements of the brainstem auditory pathway, from the cochlear nerve, through the cochlear nucleus, superior olivary complex, medial nucleus of the trapezoid body, lateral lemniscus, and inferior colliculus, manifested as four or five distinctive temporally synchronised peaks (Figure

3.4 A-B) (Wong et al., 2013). The ABR waveforms were similar for Gtf2ird1-/-,

Gtf2ird1-/+ and wild type mice. In response to click stimuli (broadband), Gtf2ird1-/- mice manifested a 6.7 dB increase in hearing threshold compared to their wild-type littermates (wild type = 25.5 ± 1.10 dB; Gtf2ird1-/- = 32.2 ± 3.03 dB; P = 0.047). In response to tone-pips, at 4 kHz the Gtf2ird1-/- mice showed a 10.2 dB increase in threshold (wild type = 26.7 ± 0.90; Gtf2ird1-/- = 36.8 ± 3.38; P = 0.007). The average

111

Chapter 3 - GTF2IRD1 in auditory pathology of WBS difference in hearing threshold between 8 to 32 kHz was 6.4 dB (average increase in

ABR threshold in Gtf2ird1-/- mice: 8 kHz = 7.53; 16 kHz = 4.00; 24 kHz = 6.67; 32 kHz

= 7.33, dB) revealing a mild hearing deficit across a broad range of test frequencies

(Figure 3.4 B-C) (P < 0.001 using 2 way ANOVA on ranks). No differences in threshold were detected in Gtf2ird1-/+ mice compared to wild type littermates (P > 0.05 using 2 way ANOVA on ranks).

112

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.4 Auditory brainstem response (ABR) threshold analysis

(A) ABR waveform response identified by peaks (P1–P5) and troughs (N1-N5).

Peaks reflect sound-evoked responses generated by successive regions of the auditory pathway; example from 80 dB, 4 kHz stimulus in a wild type mouse. (B) ABR waveform recordings over 10 ms to 4 kHz tone-pip stimuli from representative wild type and Gtf2ird1-/- mice. Dashed line shows ABR thresholds. Interval was increased in

5 dB steps. (C) ABR thresholds to click and tone-pips (4–32 kHz) in 8-10 week old

Gtf2ird1-/-, Gtf2ird1-/+ and wild type siblings (n=15/genotype) (*** P < 0.001, two way

ANOVA on ranks; * P < 0.05, Holm-Sidak multiple pairwise comparison between

Gtf2ird1-/- and wild type). Data are expressed as means ± SEM decibels sound pressure level (dB SPL).

113

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2.5 Recruitment and conduction of auditory nerve firing is not affected in

Gtf2ird1-/- mice

The term recruitment in the context of auditory physiology has been defined as an abnormally exaggerated loudness when hearing thresholds are elevated (Florentine et al., 2001). In simple terms, it refers to the phenomenon of a rapid growth in perceived loudness in the context of a sensorineural hearing loss, meaning that if a patient does not have SNHL, the patient cannot have recruitment. The hearing loss raises the threshold of auditory perception, but once the threshold is exceeded, sounds seem overly loud. It is thought to be caused by recruitment of working neighbouring hair cells into the response to frequencies that should be detected by the damaged hair cells. There are two other phenomena that are often confused with recruitment; hyperacusis

(discomfort to suprathreshold sounds) (Jastreboff and Jastreboff, 2015) and phonophobia (fear of normal sounds) (Gothelf et al., 2006).

By analysing the peak amplitudes and latencies of the ABR waveforms, effects on the recruitment and conduction elements of the auditory pathway can be assessed

(Legatt, 2002). The peak amplitudes and latencies of P1-N1, P2-N2, P3-N3 and P4-N4 elements of the ABR waveform recorded from the mice were analysed at 4 kHz (45 dB); a point at which the greatest threshold difference was observed across the range of frequencies measured (Figure 3.5 A). This analysis showed that there were no significant differences in the ABR between homozygous Gtf2ird1-/- and wild type mice in the latencies or growth functions of the first four peaks (Figure 3.4 A-B, wild type n

> 14; Gtf2ird1-/- n > 11; P = 0.12 via 2 way ANOVA on ranks for Figure 3.4 A; P =

0.41 for peak I, P = 0.48 for peak II, P = 0.15 for peak III and P = 0.08 for peak IV via

Holm-Sidak pairwise comparisons for Figure 3.4 B). The N1-P2 amplitudes at 4 KHz

114

Chapter 3 - GTF2IRD1 in auditory pathology of WBS also showed no significant differences (Figure 3.4 C, wild type min n=13; Gtf2ird1-/- min n=6; P = 0.36 via 2 way ANOVA on ranks).

115

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.5 ABR growth functions

(A) ABR input/output (I/O) functions for latency between peaks I to IV at 4 kHz in each genotype were equivalent. (B) Peak amplitudes at 4 kHz (45 dB, min n=11) measures auditory nerve conduction and recruitment. Despite genotype differences in

116

Chapter 3 - GTF2IRD1 in auditory pathology of WBS thresholds, ABR peak latency and amplitude were similar. (C) Amplitude measurements of N1-P2, which corresponds to the auditory relay between the cochlear nerve and the brainstem, at 4 kHz from 25 to 60 dB (minimum n=6). No significant difference was found between genotype, indicating a normal conduction and recruitment of auditory pathways in Gtf2ird1-/- mice (P>0.05, ANOVA on ranks; P>0.05

Holm-Sidak multiple pairwise comparison). Data are expressed as means ± SEM decibels sound pressure level (dB SPL)

117

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2.6 Gtf2ird1 null mice exhibit moderate loss of outer hair cell function

Distortion product otoacoustic emissions (DPOAEs) are generated by the reverse transduction cochlear micromechanics initiated by the outer hair cell, which changes its cell height and amplifies basilar membrane movement in response to the receptor potential generated by (forward) sound transduction. Two pure tones of close frequency range (1:1.25) and equal intensity are presented to the ear simultaneously and an ‘echo’ generated by the electromotility of the outer hair cells is detected by a sensitive microphone as DPOAEs. The cubic (2푓1 − 푓2) distortion products were analyzed for threshold screening in Gtf2ird1-/- and Gtf2ird1-/+ mice versus their wild type littermates. Thresholds were significantly higher across all tested frequencies in the

Gtf2ird1-/- mice (Figure 3.6) The ANOVA on ranks resulted in an overall P value <

0.01. The Holm-Sidak all pairwise comparisons, comparing Gtf2ird1-/- and wild type, resulted in statistical significant differences within the individual analysed frequencies:

8 kHz P = 0.0006; 12 kHz P = 0.0025; 16 kHz P = 0.015; 24 kHz P = 0.053; 32 kHz P =

0.015). The DPOAE threshold differences across all of the tested frequencies resembled the ABR results, where the strongest threshold difference was apparent at 8-16 kHz, being the average increase in the DPOAE threshold in Gtf2ird1-/- mice for each frequency as follows: 8 kHz = 19.02; 12 kHz = 15.52; 16 kHz = 13.67; 24 kHz = 8.59;

32 kHz = 11.17, dB).

118

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.6 Analysis of the distortion products of otoacoustic emission

(DPOAE) thresholds in Gtf2ird1-/- mice

-/- -/+ Cubic 2f1-f2 DPOAE in 8-10 week old Gtf2ird1 , Gtf2ird1 and wild type controls mice responding to 8 to 32 kHz puretone stimuli (n=15 for each genotype; ***

P < 0.001; ** P < 0.001; * P < 0.05). Data are expressed as means ± SEM.

119

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2.7 DPOAE input/output functions in Gtf2ird1-/- mice show amplitude reductions at higher levels of sound

The DPOAEs measured above at cubic (2푓1 − 푓2) have two components from different cochlear sources: a distortion component generated near f2 and a reflection component from the characteristic site of the distortion product. The interaction of these two sources may negatively affect the DPOAE input/output functions that are measurements usually used to predict either auditory thresholds or the compression characteristics of the basilar membrane (Mauermann and Kollmeier, 2004). The input/output growth functions determined by peak amplitude at 8 kHz, 12 kHz and 32 kHz were analysed to examine the strength of cochlear sound transduction contributed by the outer hair cells. Growth function analysis of the cubic DPOAEs showed a significant roll-off on the progressive increase in amplitude of the cubic (DPOAE) in

Gtf2ird1-/- mice compared to wild type controls (Figure 3.7).

120

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.7 Analysis of DPOAE growth functions

Input/output growth function analysis of cubic DPOAEs for wild type and

Gtf2ird1-/- littermates at (A) 8 kHz, (B)12 kHz and (C) 32 kHz characterises the gain of outer hair cell reverse transduction reflecting cochlear amplifier sensitivity (***

P<0.001; two way ANOVA on ranks; min n = 9 for wild type at 32 kHz; min n = 7 for

Gtf2ird1-/- at 8, 12 and 32 kHz; max n = 15 at 8 kHz for wild type; max n = 13 at 8kHz for Gtf2ird1-/-). Data are expressed as means ± SEM.

121

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.2.8 Age-related assessment of Gtf2ird1-/- hearing capacity and genetic background effects on hearing

Evidence shows that young individuals with WBS, with otherwise normal hearing

(<15 dB HL) and normal middle ear functionality, have subtle auditory dysfunctions that can be detected using clinical testing methods (Marler et al., 2010; Paglialonga et al., 2011; Paglialonga et al., 2014). Results presented here show a mild hearing loss in young 8-10 week old Gtf2ird1-/- mice, which would fall into the category of normal hearing, since for most of the analysed frequencies, the hearing loss is less than 15 dB.

Humans with WBS experience a progressive hearing loss with age (Paglialonga et al.,

2014), suggesting that the phenotype in the mice could also deteriorate with age.

In order to evaluate whether the threshold differences persist or become worse with age, the hearing capacity of different cohorts of mice, from 12-14 month old, was assessed. No significant differences in either ABR or DPOAE auditory thresholds were found between Gtf2ird1-/- mice and wild type age-matched controls. However, the recorded auditory thresholds for both ABR and DPOAE were, in general, very high at this age, with inconsistent standard deviations. This means that the analysis was an unreliable representation of hearing loss progression (Figure 3.8) (ABR; wild type controls: click =55 dB+14.7; 8 kHz =52.5+6.3; 16 kHz =52.5+11.4; 24 kHz =89.2+2.0;

Gtf2ird1-/-: click =48+17; 8 kHz =41.8+12.6; 16 kHz =70.3+10.0; 24 kHz =90+0 –

DPOAE; wild type controls: 8 kHz = 48.3+17.8; 12 kHz =66.7+12.5; 16 kHz =75+3.2;

24 kHz =79.2+2; 32 kHz =80+0; Gtf2ird1-/-: 8 kHz =38+16.9 ; 12 kHz =53.5+14.3 ; 16 kHz =67.5+14.4; 24 kHz =72+13.8; 32 kHz =75+12.7). We hypothesise that any hearing differences that are due to the genotype are confounded and overwhelmed by the intrinsic hearing loss associated with the C57BL/6 genetic background in aged mice

(Hequembourg and Liberman, 2001).

122

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Figure 3.8 Hearing function in aged mice

(A) Auditory Brainstem Response thresholds to click and tone-pips (8–24 kHz) in 12-14 month old Gtf2ird1-/- and wild type siblings. (B) Distortion products otoacoustic emissions (DPOAE) thresholds. Cubic 2f1-f2 DPOAE in 12-14 month old

Gtf2ird1-/-, wild type controls mice responding to 8 to 32 kHz puretone stimuli (n=6 for wild type and n=10 for Gtf2ird1 -/-). Data are expressed as means ± SEM.

123

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

3.3 DISCUSSION

Thanks to the technological advances in clinical assessment of auditory pathology, diagnosis of hearing disorders is now a routine procedure. As a consequence, over the last decade, a number of scientific reports have emerged describing some degree of hearing disability in individuals with WBS (Barozzi et al., 2012; Barozzi et al., 2013; Cherniske et al., 2004; Marler et al., 2005; Marler et al., 2010; Paglialonga et al., 2011; Paglialonga et al., 2014). This is also due to the way clinicians are assessing

WBS pathology; the increasing number of reports of this phenotype turns the underdiagnosed hearing pathology into a typical finding. Hearing disabilities, especially the pathology classified as sensory neural hearing loos (SNHL), should now be listed as part of the broad clinical presentation of the disease.

Recent reports have described new protocols to analyse young subjects affected by WBS, as a means to determine whether subtle auditory dysfunctions can be detected in advance (Marler et al., 2010). These studies have revealed that WBS subjects with normal hearing thresholds (≤15 dB HL), and normal middle ear functionality, have very subtle auditory dysfunctions that can only be detected using the clinically available methods of otoscopy, acoustic reflexes, tympanometry, pure-tone audiometry, and, especially, DPOAE. Therefore, the importance of using otoacoustic emissions as a complement to routine audiological examinations in individuals with WBS is to detect, before the onset of hearing loss, possible subtle auditory dysfunctions so that patients can be identified early, better monitored, and promptly treated (Paglialonga et al.,

2014).

The cause of the auditory pathology in WBS is currently unknown and has led some to speculate on a possible contribution from the observed effects of premature

124

Chapter 3 - GTF2IRD1 in auditory pathology of WBS ageing and/or a high predisposition for noise-induced cochlear damage that could in some way be connected to the so-called hyperacusis (Marler et al., 2010). Involvement of the haploinsufficiency of elastin due to hemizygous loss of the ELN gene is a popular hypothesis (Barozzi et al., 2012; Marler et al., 2010; Paglialonga et al., 2011) because of its profound impact on other organ systems and expression of the gene in several key hearing structures. Loss of the LIMK1 gene has also been implicated (Matsumoto et al.,

2011) as it encodes a kinase that regulates actin reorganization and plays a role in the control of outer hair cell motility (Matsumoto et al., 2010). However, there is currently no direct evidence for the specific involvement of either of these genes in WBS hearing deficits.

Since GTF2IRD1 is generally considered a low abundance protein in adult mice, the fact that the expression analysis performed by Palmer et al (2007) showed such a high expression in the cochlea is intriguing. The authors did not perform further expression analyses in the inner ear due to the lack of enzymatic β-galactosidase activity after the harsh procedure used for decalcification of the cochlear bones (Palmer et al, personal communication). Work presented in this chapter, using an antibody based β- galactosidase detection system, showed that the expression of Gtf2ird1 was strong in the spiral ganglion, which is mainly composed of the cell bodies of afferent bipolar neurons.

Spiral ganglion neurons are the first neurons in the auditory system to fire an action potential, and supply all the auditory input to the brain. Two populations of ganglion neurons have been defined by comparing cell body characteristics such as size, myelination, ultrastructural components, and staining properties. These somatic distinctions also correspond to differences in peripheral innervation. The larger type I neurons represent 90-95% of the population and contact inner hair cells (IHCs) by way

125

Chapter 3 - GTF2IRD1 in auditory pathology of WBS of myelinated radial fibers, whereas the smaller type II neurons constitute the remaining population and contact the more numerous outer hair cells (OHCs) by way of unmyelinated outer spiral fibers (Berglund and Ryugo, 1991). The expression analysis of the spiral ganglion neurons showed no distinction between these neuronal subtypes and Gtf2ird1 appeared to be expressed equally in all of them.

Our results also showed that Gtf2ird1 is expressed in a number of cell types within the cochlear compartment, including the epithelial cell layer of Reissner’s membrane, interdental cells and the marginal cells of the stria vascularis. Gtf2ird1 expression was also mapped to the outer hair cells, as well as a number of the closely associated supporting cell lineages within the organ of Corti and the interdental cells of the spiral limbus. The interdental cells play a role in the deposition of the tectorial membrane and deficits in the production of key structural proteins by these cells, such as CEACAM16, can lead to SNHL (Kammerer et al., 2012). Similarly, pillar cells and

Dieters’ cells create a rigid scaffold made of actin and microtubule bundles that support and surround the outer hair cells and facilitate an appropriate level of mechanical loading. These cells have been shown to be linked to the overall electromotility of the outer hair cell units, which can be impaired by the loss of the gap junction protein connexin 26, leading to SNHL (Zhu et al., 2013). The function of the outer hair cells is also dependent on a very high endocochlear potential (+95 mV) which is largely generated by the action of the marginal cells of the stria vascularis which pump K+ ions into the endolymph of the scala media and the passive outflow of K+ through the organ of Corti and Reissner’s membrane (Patuzzi, 2011b). A large and prolonged drop in endocochlear potential is known to impair the mechanical sensitivity of the cochlea

(Patuzzi, 2011a). Since the marginal cells of the stria vascularis, Reissner’s membrane and several cell types within the organ of Corti were shown to express Gtf2ird1, it is

126

Chapter 3 - GTF2IRD1 in auditory pathology of WBS possible that a dysfunction in these cells could contribute to the observed impairment of the cochlear amplifier through a disturbance to this strong ionic gradient.

It is intriguing that GTF2IRD1 is expressed in a variety of cell types in the inner ear. Insights from Palmer et al. (2007) suggested that the expression was abundant in the spiral ganglion, thus providing the basis for initiating this study on the function of

GTF2IRD1 in a neuronal cell type that has a physiological function that can be readily measured in a clear and direct way using hearing tests. According to our original plan, the finding of a potential hearing defect associated with these neurons could be generalised for malfunctions in other neurons that would account for the general function of GTF2IRD1 in the CNS and the behavioural abnormalities identified in

Gtf2ird1-/- mice. However, in this chapter the examination of the cochlear compartment showed a variety of cellular types that express Gtf2ird1, making it difficult to dissect cell-type-specific functional consequences. Furthermore, molecular analyses that could follow up on such functional abnormalities would also not be possible due to an inability to pinpoint and dissect the cells causing the majority of the defect.

The functional analysis of hearing capacity in Gtf2ird1-/- mice was initiated by evaluating the behavioural hearing response. No significant differences were observed.

However, the startle reflex is only stimulated by relatively loud sounds and cannot be used to determine absolute sensitivity. Nevertheless, it can provide information that may be used to supplement threshold measurements (Parham and Willott, 1988). For example, a normal startle to loud sounds in an animal with a hearing loss could indicate the occurrence of recruitment; a phenomenon in which absolute thresholds are elevated, but the apparent loudness of sounds at supra threshold levels is unchanged (Moore,

2012). While the behavioural hearing responses indicated no significant differences, a trend indicative of hearing abnormalities in the Gtf2ird1-/- mice could be distinguished

127

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

(Figure 3.3). However, as reported by Howard et. al. (2012), the Gtf2ird1-/- mice have shown evidence of increased responses to stressful stimuli, including increased vocalizations and increased CORT levels in response to the forced swim test and confinement, suggesting that these mice have abnormalities in the way they respond to stress-inducing stimuli. Nevertheless, these findings suggested that a more sensitive assessment of physiological hearing threshold measurements (such as ABR and

DPOAE) would reveal a more subtle phenotype.

As described earlier, physiological evaluation of the hearing response permitted a direct quantification of auditory capacity. These analyses are capable of identifying mild hearing differences between subjects when no significant differences are observed with conventional behavioural hearing tests. The reason behind these different outcomes resides in the ability of the physiological methods to detect smaller threshold differences (3-5 dB) with techniques such as DPOAE or ABR.

Whilst the Gtf2ird1 complete knockout mouse provides a means to identify any possible hearing phenotype, since WBS is a haploinsufficiency disorder, it would seem essential to examine auditory function in heterozygous animals. Although there is evidence that other phenotypes have not been identified in the heterozygous Gtf2ird1-/- mice (Howard et al., 2012), it does not mean that this is a general rule for all the phenotypes resulting from this genetic mutation, particularly if the sensitivity and accuracy of the physiological measurement techniques are considered. Therefore,

Gtf2ird1+/- heterozygous mice were included as an additional cohort to the analysis.

However, similar to what is observed for other phenotypes in these mice (Howard et al.,

2012), heterozygous mice did not display any type of hearing deficit. In the context of

WBS pathology, it is possible that a contribution from other genes within the deleted region could have an additive effect to create the totality of the hearing defect.

128

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

However, it is also likely that WBS is not merely the additive consequences of multiple individual genetic insults that can be separated into their component parts by the creation of monogenic mutations that will faithfully reproduce each phenotype in a hemizygous state. It is probable that, within the WBS condition, there are also interactions at the molecular level that create synergistic epistasis, exaggerating phenotypes due to the haploinsufficiency of multiple gene products encoded by genes in the WBS deletion.

ABR and DPOAE assays both showed that Gtf2ird1-/- mice have a significant loss in hearing sensitivity and the DPOAE data indicate that this can be attributed to a dysfunction of the cochlear amplifier.

This study demonstrates by ABR analysis that 8-10 week old Gtf2ird1-/- mice have a mild hypoacusis, which is evident in the elevated threshold. SGN dendrites make synaptic contact with the base of the hair cells, and their axons are bundled together to form the auditory portion of the eighth cranial nerve. Detection of an alteration in the conduction properties of these cells, or neurons comprising subsequent elements of the auditory pathway, may have provided a potential explanation for the auditory allodynia found in WBS. But if loss of the GTF2IRD1 protein impacts on these nerve cells, it has either had no effect on their conduction properties in this assay system, or functional compensation has developed.

The input/output functions of the ABR data was analysed in order to evaluate recruitment and conduction of auditory nerve firing. Although there is some disagreement as to the precise definition of recruitment, there is a general consensus that recruitment stems from a loss of the compressive non-linear characteristics of the basilar membrane, which is consistent with hair cell damage (Moore et al., 1999). Individuals

129

Chapter 3 - GTF2IRD1 in auditory pathology of WBS with normal hearing do not show recruitment, and individuals with SNHL exhibit recruitment as a reduced dynamic range, which suggests that these individuals have less

“headroom” for loudness growth. Once a louder intensity level is reached, it appears that individuals with normal hearing as well as with SNHL exhibit a similar perception of that loudness (Hellman and Hellman, 2001). While this could potentially explain the hyperacusis described for WBS patients, we were unable to detect any difference in recruitment and conduction of nerve auditory pathways. This would support the concept that the so-called hyperacusis is a more complex manifestation of the central hearing response that is bound up with increased anxiety, which is common in WBS children.

However, and as described in the introduction chapter of this thesis, the use of the term hyperacusis is inaccurate and auditory allodynia, meaning aversion to, or fear of certain sounds that are usually acceptable to others, is a more accurate description of this condition (Levitin et al., 2003). Nevertheless, analysis of the ABR waveforms suggests that there is no afferent conduction defect that might indicate a neural impairment of the cochlear nerve or subsequent elements of the brainstem auditory pathway.

Analysis of DPOAE revealed that Gtf2ird1-/- mice showed a significant increase in thresholds, averaging 14 dB across the 8-32 kHz spectrum, and with a greater difference at 8 – 16 kHz where hearing sensitivity is higher in the mouse. These data suggest that there is a defect in the cochlear amplifier associated with the outer hair cells and the extent of the difference is sufficient to explain the abnormality of the auditory threshold found in the ABR. These data indicate that the principal hearing deficit in the mice can be traced to impairments in the amplification process mediated by the outer hair cells and suggests that similar mechanisms may underpin the SNHL experienced by

WBS patients.

130

Chapter 3 - GTF2IRD1 in auditory pathology of WBS

Analysis of the DPOAE input/output curve of the Gtf2ird1-/- mice shows a progressive roll off, in comparison with the wild type curve, suggesting reduced outer hair cell amplification. This pattern is similar to what has been reported before in a mouse model of aging (Parham, 1997). However, mice were young at the time of testing

(8-10 weeks) and genotypes were caged together under high quality barrier conditions, thus making it unlikely that the hearing defect is due to secondary effects such as premature aging, infections or environmental differences. Although there is evidence that these mice have a craniofacial defect (Howard et al., 2012), analysis of skulls using micro-CT has failed to detect any changes in the dimensions or shape of the hard tissues and the defect appears to be restricted to the thickness of the epidermis overlying the nose and lip regions (Canales et al. unpublished, Chapter 5). In addition, visual inspection indicated normal external auditory canals in the Gtf2ird1-/- mice. Thus, it is also unlikely that these hearing defects can be attributed to a developmental anomaly of the hearing apparatus as part of the craniofacial gestalt. The extent of the SNHL is not large enough to suggest the likelihood of a detectable change in cochlear structure and our histological observations support this view. The balance of data suggests that the principal defect resides within the function of the outer hair cells and this is most likely caused by a molecular defect either within these cells or by a dysfunction within the other Gtf2ird1-expressing cell types within the cochlea that contribute to the driving force for sound transduction, or the regulation of cochlear mechanics.

A recent work describes the isolation of hair cells from P1 mice using a GFP expressing transgene, followed by a microarray analysis that analyses changes in the mRNA expression profile (Elkon et al., 2015). It is only possible to perform this experiment in neonates (P0-P1 mice), due to the limited accessibility of hair cells in adult mice. However, a potential future experiment could be contemplated in which

131

Chapter 3 - GTF2IRD1 in auditory pathology of WBS microarray analysis is performed on OHCs dissected from Gtf2ird1-/- and wild type neonates, in order to detect any developmental defects that are detectable at P1 that could go on to cause SNHL at 8 weeks. However, as mentioned before, OHCs are not autonomous systems as they depend on the functionality of other Gtf2ird1-expressing cells in their locality and it is not yet clear which of these cells are primarily responsible for the defect that manifests as a product of OHC function.

The defect detected in the Gtf2ird1-/- mice bears a strong similarity to that observed in WBS patients, in that the dysfunction can be most easily detected early in

DPOAE measurements (Barozzi et al., 2012; Barozzi et al., 2013; Marler et al., 2010;

Paglialonga et al., 2011). However, there is no evidence in our mice of a mechanism within the auditory signal processing pathway that might explain the auditory allodynia in WBS patients (Levitin et al., 2005). It is possible that this phenomenon results from events within the central nervous system beyond the capability of the ABR analysis to detect. It is also unclear to what extent the auditory allodynia is bound up with anticipatory anxiety and the polarization of affective response to different types of auditory stimulus (Levitin et al., 2005) that clearly has elements of emotional processing and anxiety control in addition to any mechanistic dysfunction of hearing.

While the expression analysis provides some candidate cell types and potential mechanistic causes, what molecular events would underpin such a change in cell phenotype? Some GTF2IRD1 candidate gene targets have been identified by direct binding to enhancer elements within the developmental genes Hoxc8 (Bayarsaihan and

Ruddle, 2000) and Goosecoid (Ring et al., 2002) and the skeletal muscle-specific

Troponin I slow (O'Mahoney et al., 1998) or by microarray detection of differential gene regulation in a Gtf2ird1 over-expression system (O'Leary and Osborne, 2011).

However, an analysis of mRNA from Gtf2ird1-/- brain samples failed to identify

132

Chapter 3 - GTF2IRD1 in auditory pathology of WBS evidence of any specific gene targets (O'Leary and Osborne, 2011). At this early stage, much of the basic information on how this protein operates is missing and suggests that the function is complex and cell context-dependent. Therefore, it is not possible to make realistic testable predictions of altered protein abundance that could be analysed in the cochlea of the Gtf2ird1-/- mice. Furthermore, as discussed above, there are multiple cell types within the cochlea that express Gtf2ird1 and loss of the protein in any of these cells could cause the defect in the amplifier function. Progress in the understanding of the specific nature of the cochlear defect will have to await either a clearer understanding of the cell type that is primarily responsible and/or the molecular function of the GTF2IRD1 protein.

There is evidence that young individuals with WBS, with otherwise normal hearing, exhibit mildly altered physiological hearing parameters in the clinic (Marler et al., 2010; Paglialonga et al., 2011; Paglialonga et al., 2014). The fact that our results showed an overall mild hearing loss in young 8 week old Gtf2ird1-/- mice, suggested that the phenotype would persist and perhaps aggravate with ageing. However, our analysis of 12-14 month old mice revealed no genotype-specific defects that were detectable beyond the age-related hearing losses that are known to occur as a result of the genetic background of the mice (Hequembourg and Liberman, 2001).

C57BL/6 mice are commonly used as animal models of progressive sensorineural hearing loss. The strain possesses recessive genes that cause an inevitable degeneration of outer hair cells (OHCs) beginning in the basal end of the cochlea, accompanied by a loss of sensitivity for high frequency sounds, followed by degeneration of spiral ganglion cells and inner hair cells (IHCs). Cochlear pathology progresses from the extreme cochlear base towards the middle cochlear turns, and also appears in the apical cochlea (Willott, 1996), resulting in hearing loss across the whole range of the

133

Chapter 3 - GTF2IRD1 in auditory pathology of WBS frequencies analysed. In C57BL/6 mice, high-frequency threshold elevation becomes severe between 6-12 months (Henry and Chole, 1980; Li and Borg, 1991; Mikaelian,

1979; Parham and Willott, 1988; Willott, 1986; Willott and Turner, 1999).

Histological examination of inner ear morphology at the later time point examined, would most likely have shown signs of cochlear abnormalities in both

Gtf2ird1-/- and wild type mice. To validate the hypothesis that the Gtf2ird1-/- mice hearing phenotype persists and is aggravate by ageing, it would have been necessary to transfer the mutation onto an alternative inbred background that does not show an intrinsic age-related augmentation of the auditory threshold. We evaluated the relative cost-benefit of this experiment and rejected it, since the breeding would take over 2 years.

In conclusion, the data presented in this chapter indicate a role for GTF2IRD1 in the hypoacusis of WBS patients and does not rule out a potential contribution to the auditory allodynia, which could be part of a central defect that cannot be assessed by the assays presented. No significant hearing defect was found in the Gtf2ird +/- animals suggesting that GTF2IRD1 insufficiency is not enough to explain all of the hypoacusis in WBS. Additional separate hearing defects resulting from the reduction in other WBS gene products (e.g. ELN, LIMK1, FZD9 and STX1A) may combine to form the whole.

Alternatively, interactions at the molecular level may cause synergistic epistasis that enhances the severity of GTF2IRD1 hemizygosity when set within the context of the entire deletion. For example, the combined loss of GTF2IRD1 and GTF2I may enhance the severity of the phenotypes they generate to more than just an additive degree since evidence suggests that they share overlapping molecular functions (O'Mahoney et al.,

1998). Addressing these more complex questions will necessitate the interbreeding of mice carrying separate monogenic mutations.

134

CHAPTER 4 - GTF2IRD1 CONTROLS EPIDERMAL CELL PROLIFERATION AND FACIAL SKIN PATTERNING DURING DEVELOPMENT

135

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.1 INTRODUCTION

For many years the skin features of WBS were exclusively linked to Elastin

(ELN); a gene within the common WBS deleted interval that was first to be causally associated with the cardiovascular abnormalities of the disease. Elastin is a highly elastic protein present in different amounts in all connective tissue (Parks et al., 1992;

Rosenbloom, 1987). Collagen is the most abundant protein in the skin and elastin makes up the 4% of the fat-free dry weight (Hussain et al., 2013). Together these proteins provide the strength and elasticity necessary to provide both a resilient physical barrier to the external environment and the flexibility necessary for movements of the body.

Since it plays such a crucial role in the skin, it is not surprising that hypotheses attribute the soft tissue facial features of WBS to the lack of this protein. However, although there is strong evidence to support the notion that hemizygous ELN mutations

(or deletions) that result in haploinsufficiency of the elastin protein, are sufficient to explain the cardiovascular abnormalities of WBS, relatively few reports describe abnormalities in other elastin-rich tissues (Ewart et al., 1994; Li et al., 1997; Metcalfe et al., 2000).

In general, these studies are on a small number of individuals with WBS, in whom the lack of the ELN gene has been confirmed, showing that they have softer skin that can be mobilised more easily and contained a reduced deposition of elastin when examined by electron microscopy (Dridi et al., 1999; Ghomrasseni et al., 2001; Urban et al., 2000). Accordingly, skin elasticity in individuals with typical WBS deletions was recently evaluated leading to the identification of detectable biomechanical differences in skin stiffness (Kozel et al., 2014). In addition, two related individuals, one carrying a small 7q11.23 deletion (83.6Kb) including the ELN and LIMK1 genes (Frangiskakis et

136

Chapter 4 – GTF2IRD1 in epidermal mechanisms al., 1996) and one carrying a translocation which disrupts the ELN gene (Morris et al.,

1993a), exhibited some facial features of WBS. However, other reports of patients with

ELN point mutations show that hemizygous inactivation of the gene by this mechanism is sufficient to induce SVAS and other connective tissue symptoms, but none of the additional common features of WBS are observed, including the craniofacial abnormalities (Li et al., 1997; Tassabehji et al., 1997). Altogether, these data suggest that ELN deletions may contribute to the soft tissue facial phenotypes observed in WBS patients but ELN point mutations do not. The mechanisms underpinning these observations remain unknown, but it is also possible that absence of the WBS facial gestalt has genes other than ELN as major contributors to the phenotype.

The analysis of features in patients carrying atypical deletions of the WBS region plays an important role in identifying which genes may be responsible for specific aspects of the WBS phenotype. These genotype-phenotype correlations indicate that haploinsufficiency of gene products from the two evolutionarily-related transcription factors GTF2IRD1 and GTF2I, account for the major aspects of the WBS neurocognitive profile and are also prime candidates for the cause of the facial dysmorphology (Antonell, Del Campo et al. 2010). Evidence indicates that these genes are not only structurally homologous but also share overlapping functions and thus the combined effect of reducing both GTF2IRD1 and GTF2I has also been proposed as playing an important part in the development of the facial features (Tassabehji et al.,

1999a).

Analyses of Gtf2ird1-/- mice have revealed a series of phenotypes, some of them sharing intriguing parallels with the features of WBS. For the Gtf2ird1-/- mice utilised in this thesis, these phenotypes currently include; reduced body weight, a motor

137

Chapter 4 – GTF2IRD1 in epidermal mechanisms coordination deficit, alterations in exploratory activity, novel audible vocalisations combined with increased serum corticosterone levels in response to specific stress- inducing stimuli, an abnormal hearing response and a facial deformity resulting from localised epidermal hyperplasia in specific sites (Canales et al., 2015; Howard et al.,

2012)

In this chapter, the epidermal hyperplasia phenotype is explored in depth to determine the cellular defect that underpins this feature. With the purpose of investigating the potential role of GTF2IRD1 in epidermal control mechanisms, and its possible involvement in the causation of the physical craniofacial features of WBS, the expression of Gtf2ird1 was mapped in the skin, beginning at embryonic stages. A detailed analysis of epidermal proliferation and differentiation rates was conducted and we have determined when the epidermal abnormalities are triggered and investigated whether the most important function of the skin, the skin barrier function, is compromised.

138

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.2 RESULTS

4.2.1 Epidermal thickening in Gtf2ird1-/- mice is restricted to regions of the face and foot pads

WBS facial features are often immediately recognisable and have been described in the past as elfin-like. The upturned nose, wide mouth and small chin has led to this condition being described as ‘the elfin face syndrome’ (Bhattacharjee, 2005). However, specialists of the syndrome have found the adjective ‘elfin’ inappropriate and suggested that clinicians should employ the currently accepted description of the WBS facial features. This description includes a broad forehead, a medial eyebrow flare, periorbital fullness, strabismus, a stellate iris pattern, a flat nasal bridge, an upturned nose (in some cases with anteverted nostrils), malar flattening, full cheeks and lips, a long smooth philtrum, a pointed chin, and a wide mouth in a face that becomes more coarse with age

(Burn, 1986; Morris et al., 1988).

Analysis of mice lacking the transcriptional regulator, GTF2IRD1, has revealed a soft tissue facial phenotype that shares intriguing parallels with WBS (Howard et al.,

2012). The authors described a facial deformity around the nose and lip area that results from localised epidermal hyperplasia. However, the skin of Gtf2ird1-/- mice was not exhaustively explored to test whether other zones presented any additional skin phenotypes.

Since the facial hyperplasia exhibited in Gtf2ird1-/- mice seems similar to the soft tissues enlargements observed in some facial regions of WBS patients (i.e. lips and nose), it was possible that some of the other soft tissue enlargements present in WBS patients may also be present in the Gtf2ird1-/- mice. However, due to anatomical differences between mouse and human and the small size of the mouse head, this

139

Chapter 4 – GTF2IRD1 in epidermal mechanisms assessment was hard to make with the naked eye. Therefore, magnetic resonance imaging (MRI) was used in order to try and address this question using more sophisticated imaging techniques.

This analysis was based on a small sample of adult heads from Gtf2ird1-/- mice and their wild type control littermates (n=3 mice per genotype). MRI is a non-invasive

3D imaging technique that relies on powerful magnetic fields, radiofrequency pulses, and computer analysis to produce detailed pictures of organs, soft tissues, bone, and virtually all other internal body structures. MRI is applicable for mouse phenotype analysis, while mice are still alive or ex vivo using fixed samples, and it has been extensively used for examining morphological changes in mice (Chen and Nieman,

2011; Ellegood et al., 2014; Gineste et al., 2013; Nieman et al., 2012). Although the

MRI computerised analysis of the face identified the visible abnormalities within the muzzle region of Gtf2ird1-/- mice, no further alterations were found, suggesting that they are either absent or too subtle to be detected using this method (Figure 4.1).

To address the question of whether other regions show alterations in skin structure, it was decided that samples from different regions of the mouse body would be examined. To do this, it is important to take into account that the overall conformational structure of the epidermis is similar in different skin types throughout the mouse body, but there are subtle differences; most of them to do with the thickness of the different differentiation layers (Kakasheva-Mazhenkovska et al., 2011). For example, the epidermis covering the ears is similar to the trunk epidermis but thicker, whereas that on the tail is significantly thicker. The footpads of mice, similar to the plantar and palmar surfaces in humans, are anatomically similar to the body skin in the arrangement of the strata, but due to the constant functional demands of contact with

140

Chapter 4 – GTF2IRD1 in epidermal mechanisms external surfaces, the number of layers in each stratum is enlarged, making these skin regions very thick compared to truncal skin (Kakasheva-Mazhenkovska et al., 2011).

Histological analysis of skin from different parts of the body was conducted using H&E staining and compared between genotypes. Analysis of skin covering the ears, back and anorectal junction revealed no detectable differences between Gtf2ird1-/- and wild type mice. However, the paws showed a visible thickening of the epidermis

(Figure 4.2), suggesting that the tissue enlargement of Gtf2ird1-/- mice is not only associated with the face, but also apparent in the footpads.

141

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.1 MRI analysis of Gtf2ird1-/- mice reveals no additional distinctive facial features

Representative images from the facial MRI analysis of a Gtf2ird1-/- mouse and wild type littermate. The top panels show 3D reconstructions of the mouse heads that indicate the position of the axial and coronal plane of the slices shown in the bottom panels.

142

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.2 Epidermal thickening in Gtf2ird1-/- mice is restricted to regions of the face and footpads

H&E stained sections of different skin regions from Gtf2ird1-/- mice show no differences to wild type littermates in the ear, back and anorectal junction. Paws however, show thickening of the skin in the footpad area. (sez) squamous epithelium zone, (rct) rectum, (cm) colorectal mucosa; scale bars represent 50µm.

143

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.2.2 Gtf2ird1 is expressed in affected regions of the face

Gtf2ird1 is expressed broadly during early to mid-embryonic development and becomes more restricted into specific tissues in late foetal and post-natal life (Palmer et al., 2007). A number of studies have also shown that the expression of Gtf2ird1 is relatively low in most adult tissues, other than brain tissue (Bayarsaihan and Ruddle,

2000; Calvo et al., 2001; Howard et al., 2012; Palmer et al., 2007; Tay et al., 2003). To date, there is almost no information regarding the expression of Gtf2ird1 in the skin.

Due to the unavailability of an antibody that is sensitive and specific enough to detect endogenous mouse GTF2IRD1 protein, the expression of Gtf2ird1 in the skin was analysed using the Gtf2ird1tm2(LacZ)Hrd reporter mice (Palmer et al., 2007). By determining the β-galatosidase activity in situ using a histochemical staining assay that detects LacZ produced under the control of the Gtf2ird1 promoter/enhancer, we observed that Gtf2ird1 activity can be detected in the developing facial skin from embryonic day 13.5. At this stage, expression appears to be restricted to the nose-snout area, developing ears and the sensory vibrissae, which sit above and below the eye.

After one further day of development (E14.5), the expression of Gtf2ird1 changes dynamically within the zones of the face, spreading to the developing lips, eyelids and hair follicles (Figure 4.3A).

Later in development (E15.5-18.5), analysis showed that Gtf2ird1 expression is found uniformly in the tongue, the hair follicles and the skin of the developing lips, where it is mainly restricted to cells of the basal stem cell-containing epidermal layer, although some random expression was also detected in keratinocytes from superficial layers of the stratified squamous epithelium in the interfollicular epidermis (Figure 4.3

144

Chapter 4 – GTF2IRD1 in epidermal mechanisms

B). LacZ activity was also clearly detected in the margin of the eyelids from E16.5, which is another mucocutaneous junction of the face (Figure 4.3 C).

After birth, Gtf2ird1 activity could not be detected using the conventional LacZ histochemical method, which is most likely due to the low expression levels in fully developed skin similar to other adult tissues (Bayarsaihan and Ruddle, 2000; Calvo et al., 2001; Howard et al., 2012; Palmer et al., 2007; Tay et al., 2003), The β- galactosidase protein could also not be detected in the skin using indirect immunofluorescence with antibodies targeted against it. Thus, no clear Gtf2ird1 expression could be identified in skin patches from different regions of the mouse body

(data not shown).

Analysis of sections through the top lip tissue, which typically shows the most extreme overgrowth, revealed high levels of background fluorescence associated with the adjacent dermal tissue. Nevertheless, weak nuclear localised signal was detectable sporadically in a subset of differentiating keratinocytes in the superficial layers of the epithelium, similar to what was observed at E18.5. However, differing from what was observed during development, no expression could be detected in the stem cell- containing basal layer of the epidermis (Figure 4.4 A).

Immunofluorescence detection techniques often encounter high background fluorescence in skin samples. The skin emits fluorescence when exposed to 420–600 nm wavelength excitation, and measurements of skin auto-fluorescence have been proposed as non-invasive methods to estimate the degree of accumulation for advanced glycation end-products (Ahmad et al., 2015; Meerwaldt et al., 2004). This creates difficulties for trying to detect low abundance proteins in samples fixed using organic reagents (Fig.

4.2D).

145

Chapter 4 – GTF2IRD1 in epidermal mechanisms

In order to assess whether the weak LacZ signal found in sporadic cells of the epidermis was real and reduce the auto-fluorescence observed in the previous analysis, a modified whole-mount clarifying method was used that was first developed for brain tissues as a means to interrogate structural and molecular information in intact biological systems; CLARiTY (Chung and Deisseroth, 2013; Chung et al., 2013; Kim et al., 2013b). The modified CLARiTY method replaces the lipids in a tissue sample with acrylamide-based hydrogels, leaving the tissue transparent and the proteins intact and allows the visualisation of low expression proteins by immunodetection in complex tissues like skin. This analysis revealed that -galactosidase is not detectable in the epidermis of whole-mount skin samples from Gtf2ird1tm2(LacZ)Hrd reporter mice by immunofluorescence. In contrast, Keratin 14 was clearly visualised by immunofluorescence, which is a cytoplasmic intermediate filament marker of the proliferating basal layer of the epidermis (Alam et al., 2011) (Figure 4.4 B).

146

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.3 Expression of Gtf2ird1 in mouse facial skin at different stages of development

Images from Gtf2ird1tm2(LacZ)Hrd embryos mapping gene expression to the developing lips, hair follicles, vibrissae, eyes and nasal region. (A) Whole mount preparations of heads from E13.5 and E14.5 embryos stained for LacZ activity. Arrows indicate sensory vibrissae which sit above and below the developing eye. (B) Sections of the mouth region of E15.5 and E18.5 (sagittal) and E17.5 (coronal) embryos show sharp boundaries of expression (arrows) in the basal layer of the epidermis. The box shows a magnified image of the lip epidermis, indicating that expression is restricted to cells of the basal layer and hair follicles. (C) Sagittal head section of E16.5 and E17.5 embryo shows Gtf2ird1 expression in the eye and at the mucocutaneous junction of the eyelids. Scale bars are 50 µm; (dtl) developing top lip, (dbl) developing bottom lip, (t) tongue, (n) nostrils, (de) developing eye, (del) developing eyelids, (dc) developing cornea.

147

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.4 Expression of Gtf2ird1 in mouse lip skin in the adult

(A) Immunofluorescence detection of β-galactosidase (green) in the adult lip skin shows nuclear-localised expression in the superficial layers of the stratified squamous epithelium but not in the stem cell-containing basal layer. Sections of top lip from wild type negative controls are included to demonstrate the level of background autofluorescence. DAPI (blue) staining is used to identify the position of the basal layer.

(B) Immunofluorescence analysis using the modified CLARiTY technique in a whole- mount facial skin patch shows no nuclear staining of β-galactosidase (red) in

Gtf2ird1tm2(LacZ)Hrd mice. Keratin 14, a marker of the stem cell-containing basal layer was included in the analysis (green) as a technical control and positional marker. Scale bars are 50 µm.

148

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.2.3 Gtf2ird1-/- mice do not show skull abnormalities

Individuals with WBS have characteristic cranial dysmorphologies (Axelsson,

2005; Axelsson et al., 2005; Mass and Belostoky, 1993; Morris et al., 1988; Morris et al., 1990; Perez Jurado, 2003). These cranial abnormalities have been quantified and classified by performing photo-anthropometric analyses, which correspond to measurements developed for understanding human physical variations. Thus, in 1997, a study of twenty nine WBS children under 10 years of age established that the craniofacial indices; nose length to midface height and palpebral fissure width to bizygomatic diameter, ranked outside the normal range when compared with standard photo-anthropometric indices from age-matched controls. The study also reported high incidence of abnormalities in midface height, interalar distance, length of back of nose, ears morphology, chin height, angle inclination of the ears and bizygomatic diameter.

The authors claimed that these craniofacial parameters, many of which had not previously been evaluated in WBS patients, were usually enlarged and they would be useful for early detection of the disease. However, others contend that the common skull features are not dominant enough to satisfactorily characterize the facial dysmorphology

(Mass and Belostoky, 1993). Nevertheless, analysis of the craniofacial features in individuals with atypical deletions of the WBS region have led to genotype-phenotype correlations that indicate the involvement of GTF2I and GTF2IRD1 in this aspect of the syndrome. However, in the majority of cases, no computerized analyses were performed and the presence or absence of cranial or soft tissue abnormalities was based on the judgements of the supervising clinician (Botta et al., 1999; Heller et al., 2003;

Tassabehji et al., 1999a; Tassabehji et al., 2005).

Although it has been reported that the Gtf2ird1+/- and Gtf2ird1-/- mice used in this study do not present with any obvious cranial dysmorphology (Howard et al.,

149

Chapter 4 – GTF2IRD1 in epidermal mechanisms

2012), it is possible that subtle differences that are not obvious to the naked eye could have been missed. Therefore, a computed tomography (CT)-scan analysis was performed in a small group of adult Gtf2ird1-/- and wild type control mice (n=3 and n=2 respectively) in order to address this question more thoroughly. The strategy for this analysis of skull morphology was based on what has been previously reported for craniofacial malformation in a mouse model of Down’s syndrome (Richtsmeier et al.,

2000) and includes distance measurements between key landmarks and analysis of the jaw bone angles. None of these measurements indicated any obvious differences in skull shape between genotypes (Figure 4.5).

150

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.5 Skull analysis in adult Gtf2ird1-/- and wild type mice

(A) Superior (top) and lateral (side) views of the mouse craniums from CT-scan analysis. Lateral view corresponds to a sagittal cross section in the middle of the cranium. (B) Mouse skull showing the position of landmarks (1-16) used for distance analysis to evaluate skull morphometery. The cranial landmarks (number, label) are as follows: 1, nasale; 2, nasion; 3 and 12, bregma; 4 intersection of interparietal and occipital bones at the midline; 5 and 6 joining of squasomal body to zygomatic process of the squasomal; 7 and 8 intersection of frontal process of maxilla with frontal and lacrimal bones; 9 and 10 arbitrary landmarks to measure wideness of the snout; 11 and

13, arbitrary landmarks to evaluate height and lateral length of the lower jaw; 14, anterior most point on mandibular condyle (Richtsmeier et al., 2000). Distal ends of the orange lines indicate the angles measured in upper and lower jaw bones from the lateral view. (C) Histogram showing the distances between the landmarks depicted in B as

151

Chapter 4 – GTF2IRD1 in epidermal mechanisms means to assess skull shape; (D) Histogram showing the jaw bone angles from the sample of Gtf2ird1-/- and wild type mice. For (C and D), n=2 for wild type and n=3 for

Gtf2ird1-/- and the error bars indicate standard error of the mean; (UJB) upper jaw bone and (LJB) lower jaw bone. P values above the bars indicate values from a Student’s T-

Test for the probability of a difference between genotypes.

152

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.2.4 Gtf2ird1-/- mice have thickened lips as result epidermal hyperproliferation and enlarged layers of differentiated keratinocytes

Gtf2ird1-/- mice have shown abnormal facial development that results from increased thickness of the skin in specific sites (Howard et al., 2012), and, as also confirmed above, the most obvious of these are the lips and the nose leading to extensive folding and overgrowth of the surrounding soft tissues (Figure 4.1).

To find out what the defect is at the cellular level, we have performed cell proliferation and differentiation analysis in sections of lip epidermis. The interfollicular epidermis can be divided into three basic compartments. The basal epidermis where the proliferation occurs, the suprabasal epidermis, containing post-mitotic keratinocytes committed to differentiation and the cornified layer, where differentiated squames are produced. Ki67 expression was analysed in lip tissue to assess the proliferative activity of the cells from the basal stem cell-containing layer of the epidermis. The expression of

Ki67 is strictly associated with cell proliferation and since it is present during active phases of the cell cycle (G1, S, G2 and mitosis), but is absent from resting cells (G0), its assessment constitutes an excellent marker for determining the so-called growth fraction of a given cell population (Gerdes et al., 1991; Scholzen and Gerdes, 2000).

10 µm cryo-sections from the top and lower lips of Gt2ird1-/- and wild type mice were analysed by immunofluorescence using an anti-Ki67 monoclonal antibody

(Figure 4.6 A). Observations showed that the increased thickness of specific sites of the facial skin can be explained by an approximate 4-fold increase in the number of dividing cells in the basal layer of the skin that is positive for Ki67 immunofluorescence

(Figure 4.6 B) (n=3 animals per genotype; 3 non-consecutive sections analysed per animal; P=0.015; 2 way ANOVA analysis followed by Holm-Sidak method for multiple

153

Chapter 4 – GTF2IRD1 in epidermal mechanisms pairwise comparisons). We conclude that there is increased cell proliferation at the basal layer of the lip epidermis in Gtf2ird1-/- mice. Quantification revealed that the increased cell proliferation in the top lip exceeded that found in the lower lip (Figure 4.6 B).

Lower lip samples showed the same trend towards increased Ki67 positive cells but not statistical significance with this sample size (P=0.035 for top lip and P=0.102 for lower lip; Holm-Sidak multiple comparison analysis).

Accurate angling of the cutting plane while sectioning skin tissue is crucial to evaluate the thickness of the layers and to draw conclusion from the immunohistology results. If skin sections are not cut at a 90 degree angle to the plane of the skin, the basal layer of the epidermis may appear thicker for technical reasons. In order to dismiss the notion that the observed phenotype of an enlarged Ki67 positive layer could be due to a technical artefact, consecutive sections obtained from the same tissue block were analysed by immunofluorescence using an anti-integrin alpha 6 antibody. A single thin layer of integrin alpha 6 positive cells demonstrates perpendicular sectioning of the samples, as the structure of the stem cell membrane appears to be adjacent to the basal lamina of the epidermis (Figure 4.6 C).

154

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure legend in next page

155

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.6 Increased cellular proliferation in lip skin sections from Gtf2ird1-

/- mice

(A) Representative Ki67 immunofluorescence analysis in sections of the top and bottom lips of Gtf2ird1-/- and wild type mice. Increased positive signal can be observed in both samples belonging to Gtf2ird1-/- mice. The nuclear stain DAPI helps to identify the position of the basal layer. Scale bars are 50µm. (B) Histogram showing quantification of the total area of Ki67 positive nuclei in the upper and lower lip samples (n=3 per genotype). Statistical analysis indicates that the probability of a difference between genotypes is significant (P=0.015; 2 way ANOVA followed by the

Holm-Sidak method of multiple pairwise comparisons). The majority of the difference occurs in the the top lip samples (P=0.035 for upper lip sections versus P=0.102 for lower lip sections using Holm-Sidak comparison analysis). (C) Adjacent sections to those shown in (A) showing single layers of positive alpha 6 integrin fluorescence throughout the wild type and Gtf2ird1-/- tissue samples indicates perpendicular sectioning and serves as a technical control for the samples assessed in the Ki67 proliferation assay. Scale bars are 50µm.

156

Chapter 4 – GTF2IRD1 in epidermal mechanisms

As previously described in the introductory chapter of this thesis, the epidermis of the skin is composed of keratinocytes that are organised into several layers. Basal cells divide and produce cells that are continually displaced towards the external surface of the epidermis while undergoing the process of terminal differentiation

(Vandenberghe et al., 2013; Watt et al., 2006). Keratins form the major cytoskeletal intermediate filament proteins of keratinocytes and have diverse cellular functions.

These proteins are accumulated progressively as the cells mature and provide a good means to delineate the different stages of cellular movement through the epidermal layers (Fuchs and Weber, 1994; Lane and McLean, 2004). Since increased cellular proliferation and epidermal thickening of the skin is generally associated with disorganisation of the differentiating layers of the epidermis, Keratin 10 (K10), a marker of the stratum spinosum and Keratin 14 (K10), the basal keratin marker, were evaluated.

Both keratin layers were found to be augmented in top lip sections of Gtf2ird1-/- mice compared to controls (Figure 4.7 A-B). This result confirms the presence of an excessive number of differentiated cells in the region associated with the phenotype.

To investigate whether this abnormal level of differentiated keratinised layers is only present in the skin of the external portion of the lip tissue, or whether it is also present in the oral mucosa, whole top lip samples were sectioned in a sagittal plane to generate a complete picture of the skin covering the lip structure. An enlarged K10 layer was observed all around the lip, including the oral mucosa facing the inside the mouth.

However, this phenotype is less exaggerated compared to the external skin layers

(Figure 4.7 C).

157

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure legend in next page

158

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.7 The top lips of Gtf2ird1-/- mice have an abnormal number of differentiated keratinocytes

(A-B) Sections of top lip tissue analysed by immunofluorescence with K10 and

K14 markers show augmented keratinocyte differentiation in Gt2ird1-/- mice. The spinous layer (K10 positive), as well as the basal layer (K14 positive), shows increased numbers of differentiated keratinocytes. Scale bars are 50µm. (C) Sagittal top lip sections from Gtf2ird1-/- and control mice analysed by K10 immunofluorescence. The yellow lines orientate the section, distinguishing the external-facing epidermis (EFE) from the oral-facing epidermis (OFE). K10 positive cells are augmented in the whole epidermis of the Gtf2ird1-/- mouse tissue, but the phenotype is stronger in the EFE zone.

Scale bars are 500µm.

159

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.2.5 Soft tissue facial abnormalities of Gtf2ird1-/- mice begin during foetal life

We have previously determined that Gtf2ird1 is expressed in the epidermis during development, and that its expression becomes almost undetectable in adult skin

(Figure 4.3 and 4.4). However, the lip skin hyperplasia has only been described in adult animals (Howard et al., 2012). At this stage, it is not clear whether there is a direct cause-effect relationship between the lack of Gtf2ird1 and the hyperplasia, or whether it is caused as an indirect protective response to some other abnormality of the epidermis.

The first approach to address this question was to determine how early the cellular hyperproliferation and excess of differentiated cells can be observed in the facial skin. Establishing the developmental time point at which the aberrant proliferation and differentiation processes appear would provide a strong clue as to whether the facial abnormalities are due to primary defects caused by the lack of

GTF2IRD1 in the affected cells.

The development of the skin starts from the single layer of embryonic ectoderm that can be visualised at E9.5. From E13.5 onwards, a single layer of keratinocytes develops into a stratified epithelium, progressing towards complete aqueous barrier formation that can be demonstrated by the ability of the embryos to exclude dye at around E18.5 (Hardman et al., 1998). Abnormal epidermal phenotypes are better understood when examined at a late stage of development, when the stratified epithelium has been established. Therefore, signs of facial dysmorphology were looked for in whole-mount preparations and by histological analysis just before birth (Figure

4.8 A-C). Examination of whole-mount preparations at E19.5 showed that overgrowth in the lip and nose areas, although subtle, is already observable at this stage (Figure 4.8

A). These macroscopic observations were confirmed in H&E-stained histology sections,

160

Chapter 4 – GTF2IRD1 in epidermal mechanisms where the epidermis of the developing lips and snout of the Gtf2ird1-/- mice displayed a disorganised cellular conformation compared to the regular and smooth skin of wild type littermates (Figure 4.8 B-C). These observations are almost identical to the adult skin phenotype described above (Howard et al., 2012) (Figure 4.1). Using the markers of cell proliferation (Ki67) and differentiation (K10 and K14) on sections of lip tissue from P10 mice, demonstrated that the same alterations found in the adult mice, were also observable at this early post-natal stage (Figure 4.8 D).

161

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.8 Changes in skin morphology start during embryonic development in Gtf2ird1-/- mice and persist after birth

(A) Angled views of heads from a Gtf2ird1-/- mouse and a wild type control littermate showing that facial hyperplasic patterns in the nose and lip area are subtle but

162

Chapter 4 – GTF2IRD1 in epidermal mechanisms still visible just before birth (E19.5). At the same embryonic age, analysis of H&E- stained sagittal (B) and coronal sections (C) of the developing lips and muzzle area respectively, show enlarged and disorganised epidermal formations. Scale bars are

200µm. (D) Immunofluorescence analysis in sections of top lip samples from Gtf2ird1-/- mice and wild type littermates at 10 days old, shows increased layers of cells expressing

Ki67, K10 and K14, confirming that aberrant cellular proliferation and differentiation processes are present in Gtf2ird1-/- mice in early life.

163

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.2.6 Gtf2ird1-/- mice have a defect in skin barrier function

A functioning water-resistant skin barrier is essential for the survival of land- based organisms. Its formation during development prepares the foetus to face the external environmental that begins at birth, preventing desiccation and providing natural protection against infection and poisoning. The skin barrier function is conferred by the stratum corneum, which is the outer layer of the epidermis. The stratum corneum consists of dead, keratin-filled cells embedded in a lipid matrix and is formed from the underlying keratinocytes of the granular layer that undergo terminal differentiation as they progress towards the surface (Elias and Menon, 1991).

In humans, the skin barrier function develops between weeks 30 and 33 of the

40 week human gestation (Hammarlund and Sedin, 1979; Wilson and Maibach, 1980).

In the mouse, development of the barrier is initiated at E16 in specific regions of the body from where it spreads around the embryo as a moving front. The barrier acquires almost full adult function by E18.5, accompanied by multiple changes in the outer, stratum corneum precursor cells (Hardman et al., 1998).

Using whole-mount permeability assays, it is possible to evaluate how well the cornified envelope has assembled, starting at the final stages of foetal life. The whole- mount toluidine blue staining assay is a qualitative method in which the dye penetrates and stains subdermal layers in sites where the skin barrier function is disrupted or not formed. This test was initially used to examine a litter of embryos at E15.5 to test the toluidine blue dye penetration before the skin barrier had formed. This experiment showed that the dye was able to penetrate the developing skin when the skin barrier function was not fully developed, validating its use for later stages of development

(Figure 4.9 A).

164

Chapter 4 – GTF2IRD1 in epidermal mechanisms

The same test was then performed on Gtf2ird1-/- and wild type newborn (P0) mice. The Gtf2ird1-/- P0 pups showed complete dye exclusion in all regions of the skin, except for an area around the snout, indicating a defect in the barrier function. By contrast, the wild type control littermates showed complete dye exclusion in all regions of the body including the face (Figure 4.9 B). The region of the face where dye penetrated in Gtf2ird1-/- mice corresponds to the areas of the skin where the excessive proliferation and abnormal layers of differentiated keratinocytes were found, as described above. This is indicative that there is a disruption or delay in the formation of the skin barrier function in the hyperplasic regions of the face of Gtf2ird1-/- mice.

165

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.9 Whole mount toluidine blue staining evaluates skin barrier function development

(A) Whole mount toluidine blue staining of E15.5 embryos shows that the dye is able to penetrate the developing skin when the barrier function is not fully developed, confirming that the test is effective. (B) Whole mount toluidine blue stained P0 pups.

Gtf2ird1-/- mice show dye exclusion in all regions of the body except for the developing lips and nose, indicating a barrier defect in the hyperplasic regions of the face. Complete dye exclusion is observed in the wild type control mice.

166

Chapter 4 – GTF2IRD1 in epidermal mechanisms

This result prompted further analysis of the barrier function using the transepidermal water loss (TEWL) assay, which allows quantification of the degree of disruption (DiTommaso et al., 2014; Hardman et al., 1998; Smyth et al., 2008). To determine whether the skin barrier formation is permanently abnormal or merely delayed in Gtf2ird1-/- mice, skin patches from the upper lip and the dorsal skin from 2 month old Gtf2ird1-/- mice were harvested to evaluate permeability using TEWL over a period of 300 minutes after dissection. The patches from the dorsal skin were taken in order to test whether subtle defect are present that were not detected by the toluidine blue assay. The weight of the skin patches from the different areas was measured every

30 minutes for 5 hours using an electronic analytical scale. Skin patches from wild type mice served as control and Gtf2ird1+/- heterozygous mice were also included in the analysis (Figure 4.8).

The TEWL assay showed that water loss via evaporation is faster in the skin samples derived from the lips of Gtf2ird1-/- mice compared to Gtf2ird1+/- and wild type littermates (Figure 4.10 A and C) (n=5 per genotype; p<0.001 for Gtf2ird1-/- vs wild type and Gtf2ird1-/- vs Gtf2ird1+/-; ANOVA – followed by Holm-Šídák multiple comparison test). However, the dorsal skin samples did not show any genotype-specific significant differences at any point across the experimental time frame (Figure 4.10 B and D), consistent with the observations of the dye exclusion assay described above.

Water loss in the dorsal skin of wild type animals showed a slight trend towards being increased over the Gtf2ird1-/- mice (Figure 4.10 D). However, the ANOVA showed that this difference was not significant. The skin samples from heterozygous mice were not significantly different from wild type in both the dorsal skin and lip skin assays.

167

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Altogether, these experiments demonstrate that there is a localized facial skin barrier function abnormality in Gtf2ird1-/- mice that begins during development and persists into adulthood. Furthermore, this abnormality overlaps with the location of the skin hyperplasia and corresponds to the site of Gtf2ird1 expression in the basal layer of the skin during foetal life. Analysis also indicates that the reduction of GTF2IRD1 due to heterozygous loss of Gtf2ird1 is not sufficient to cause skin defects in this model.

168

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Figure 4.10 Trans-epidermal water loss (TEWL) assay as a means to quantify the skin barrier function in adult mice

Skin patches from top lip (A and D) and dorsal skin (B and D) from 2 month old Gtf2ird1-/-, Gtf2ird1+/- and wild type mice (males 2 months old, n=5 per genotype) were harvested to evaluate water loss over a period of 300 minutes after dissection. Top panels (A and B) include data from Gtf2ird1+/-. (C and D) The same data as shown in

A and B excluding data from the heterozygous mice in order to depict the differences between Gtf2ird1-/- and wild type more clearly.

169

Chapter 4 – GTF2IRD1 in epidermal mechanisms

4.3 DISCUSSION

Our understanding of the genetic mechanisms governing craniofacial development in WBS is incomplete. Attempts to identify the genes underlying the craniofacial, cognitive, and behavioural features of the syndrome rely on the identification of rare individuals with smaller deletions in the WBS region, supplemented by studies of mouse models.

These studies have highlighted members of the TFII-I gene family as critical regulators of craniofacial and neurological development (Bayarsaihan and Ruddle,

2000; Chimge et al., 2008; Ohazama and Sharpe, 2007; Schneider et al., 2012;

Tassabehji et al., 2005). Furthermore, it has previously been shown that GTF2IRD1 can bind regulatory elements upstream of genes involved in tissue development and differentiation (Bayarsaihan and Ruddle, 2000; Ring et al., 2002), and the targeted mutation of Gtf2ird1-/- in mice resulted in a consistent overgrowth of the epidermis overlying the lips such that they become wrinkled and enlarged (Howard et al., 2012).

The facial deformity found in the Gtf2ird1-/- mice described here is dissimilar from the craniofacial abnormalities reported in another mouse line with a mutation of

Gtf2ird1 caused by the random integration of a transgene (Howard et al., 2012;

Tassabehji et al., 2005). The transgenic mutant mice used in Tassabehji et al. (2005), resulted from the random integration of a c-myc transgene, under the control of the albumin promoter, into the distal portion of chromosome 5, which also induced a deletion of approximately 40 kb (Durkin et al., 2001). The proximal deletion breakpoint falls downstream of Cyln2 and the distal breakpoint falls in the first intron of Gtf2ird1.

This deletes the promoter region of Gf2ird1 and exon 1, which contains the majority of the 5’ untranslated region, but retains exon 2, which contains the translation start site.

170

Chapter 4 – GTF2IRD1 in epidermal mechanisms

This mutation resulted in greatly reduced expression of Gtf2ird1 mRNA in mice homozygous for the transgene insertion.

Analysis of these mice indicated no defects in the heterozygotes and the authors proposed that haploinsufficiency of GTF2IRD1 alone cannot explain the craniofacial features of WBS. Instead, a cumulative dosage effect was proposed for the craniofacial abnormalities, in which the minor defects due to haploinsufficiency of GTF2IRD1 are enhanced by the additional loss of other genes in the WBS region. Whilst periorbital fullness was reported in the c-myc transgenic mutants, no further analysis supported this finding, making it difficult to know whether the skin abnormalities found in our

Gtf2ird1-/- mice are comparable to this transgenic deletion model. Another difference is that the c-myc transgenic mice presented with short and sometimes misaligned jaw formation, whereas our phenotypic observations using CT scan analysis detected no hard tissue defects and instead showed that our Gtf2ird1-/- mice only show soft tissue abnormalities of the epidermal layers surrounding the nose and mouth region.

The phenotypic differences between these two mouse models could be due to the different methods used for genetic manipulation in creating these mouse lines. Our mouse line resulted from a targeted deletion of the first coding exon in Gtf2ird1, which makes it unlikely that the phenotypes observed are due to the addition of exogenous sequences. However, it is known that trace levels of truncated GTF2IRD1 could survive in this line (Palmer et al., 2010). In the c-myc transgenic line, there is a possibility that additional effects are induced by the insertion of additional sequences. Mice carrying the c-myc transgene have served as a useful model for studying the progression of hepatocarcinomas because they are prone to develop liver-specific tumours (Durkin et al., 2001). Thus, it is not unreasonable to suggest that the phenotypes observed in these

171

Chapter 4 – GTF2IRD1 in epidermal mechanisms mice could result from additive effects that arise from the insertion of the c-myc transgene and are not necessarily the direct consequence of Gtf2ird1 inactivation.

As mentioned in the introduction section of this thesis, the CD mouse model recapitulates most of the WBS features, including craniofacial abnormalities such as a small mandible (Segura-Puimedon et al., 2014). However, the facial abnormalities of the CD mice were not explored in detail in this study and it is unclear if soft tissues are affected in these mice. Therefore a comparison of the facial features of the Gtf2ird1-/- mice with those of the CD mice is not possible.

Our preliminary MRI analysis revealed no additional distinctive soft tissue abnormalities in the faces of Gtf2ird1-/- mice. However, this technique presented some difficulties in visualising the overgrowth in the lips and the nose that are clearly visible with the naked eye. To our knowledge, this type of analysis has not been used in mice to detect subtle soft tissue changes in facial morphology before. The data analysis tools available for MRI examination of rodents are set up for the analysis of brain and other visceral organs, such as adipose tissue, heart, kidney, liver, pancreas and lungs

(Arumugam et al., 2014; Jiang et al., 2015; Ku et al., 2015; Metzinger et al., 2014;

Raabe et al., 2015; van Nierop et al., 2015).

While our preliminary MRI experiment involved the analysis of only 3 mice of each genotype, and it is possible that a larger sample group would better address this question, our data suggest that even if differences do exist in other regions of the face of

Gtf2ird1-/- mice, they are very subtle compared to the defects of the nose and lip regions.

In addition, the heads were fixed, stored and then imaged by MRI, opening the possibility that tissue preservation was affected by shrinkage or shape changes.

However, measuring volumes in a fixed specimen allows better delineation of the

172

Chapter 4 – GTF2IRD1 in epidermal mechanisms borders of individual structures because of the enhanced contrast and resolution

(Benveniste and Blackband, 2006). The data analysis of the MRI images was a limiting factor in the study due to the lack of specialised tools for rodent facial analysis. Manual segmentation of facial structures was employed, but 3D image sets contain very large amounts of data, and the development of advanced automated volumetric analysis and visualization software could have provided a more in-depth way to investigate this question. It is worthwhile noting that the toluidine blue dye penetration analysis did not detect any additional facial zones that had an impaired skin barrier, such as the periorbital zones. If the hyperplasia and skin barrier defects go together consistently, it would be reasonable to expect that a defect in other facial regions would have been detected by this method also.

The fact that Gtf2ird1-/- mice present facial soft tissue abnormalities that are restricted to regions where the WBS facial dysmorphology is most pronounced is quite striking. However, a thickened epidermis was also observed in the histological analysis of the footpads of Gtf2ird1-/- mice. It is conceivable that such an effect could be generated as a secondary consequence of behavioural activity as described previously

(Howard et al., 2012). Increased usage and load could induce variability in the degree of skin thickness of this area. However, the Gtf2ird1-/- mice were generally found to have reduced exploratory activity (Howard et al., 2012) which goes against this idea.

There are no reported homologous phenotypes in the plantar skin of WBS patients that could correspond to this mouse abnormality. However, as pointed out by

Osborne (2010), we have to be careful comparing and interpreting interspecies phenotypes. While mouse models can provide valuable insights into the pathogenic mechanisms underlying human disorders, the reality is that mice are not humans; they are physiologically different, and some proteins have divergent functions between

173

Chapter 4 – GTF2IRD1 in epidermal mechanisms species, meaning that some gene disruptions may result in quite different phenotypic consequences. Also metabolic rates, reproductive age, environmental adaptations, cognitive and behavioural traits, lifespan, DNA genetic composition and the essentiality of genes are only some of the characteristics that differ, and they must be considered when correlating phenotypes (Osborne, 2010).

In the same way, there are also interspecies differences between human and mouse skin. Within mammalian organisms, the skin structure is largely the same across the body and the specialized regions are also very similar. For example, the skin of the mouse footpad is identical to human plantar and palmar skin. However, it is necessary to take some points into consideration when making genotype-phenotype interpretations in humans based on phenotypes observed in mouse models. For example, the mouse muzzle skin has vibrissae, and while Gtf2ird1 was found abundantly expressed in the developing vibrissae, these are somatosensory organs that are not found in humans.

Humans have whiskers, a form of terminal hair found on the face of males but these are not found in mice and vibrissae are commonly mistaken to be homologous to human whiskers (Piper M. T, 2011). Also, with the exception of the footpads, mice are fur- bearing animals with specialised sebaceous glands, the skin of the tail and the muzzle has a thicker epidermis than that of the trunk, and, unlike humans, the footpads are the only anatomical region that contain eccrine glands (Hussain et al., 2013). Considering the melanocytes, which are the producers of the protective protein melanin, mice are unlike humans as they rarely have interfollicular epidermal pigmentation and although pigmented mice appear to have coloured skin, it is usually patchy, ranging from pink to various shades of grey (Piper M. T, 2011). It is also important to note that the shape of the mouse face is quite different to that of humans and it may be unreasonable to expect that phenotypes of face shape in the human would be identical in mice.

174

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Our Gtf2ird1-/- mice do not show any obvious abnormality of the hard tissues and our CT-scan skull analysis supports this statement. It is interesting to note that measurements of human craniofacial deformity in WBS has revealed relatively minor changes in bony tissues (Axelsson, 2005; Axelsson et al., 2005) and, although our CT- scan analysis does not have the statistical power to dismiss subtle differences in skull morphology in this line of Gtf2ird1-/- mice, it is well known that skull abnormalities in

WBS are not considered dominant enough to constitute a clinical parameter for diagnosis. According to a study that analysed lateral cephalometric radiographs from 8 children with WBS (Mass and Belostoky, 1993), the skeletal features that contribute to the facial appearance of the syndrome include; shortness of the anterior cranial base, although the cranial base angle was found to be normal; steepness of the angle of the mandibular plane, although total facial height was found to be normal; unusual proportions of upper to lower anterior facial height and posterior to anterior facial height; and the chin button (pogonion to N-B line) was deficient, although the mandible could not be classified as retrognathic (Mass and Belostoky, 1993).

It has been suggested that a large proportion of the WBS facial characteristics can be attributed to alterations in the soft tissue components (Mass and Belostoky,

1993). However, the assessment of craniofacial abnormalities are often subjective and, although there are three dimensional imaging tools available to generate dense surface models that facilitate the identification of the characteristic WBS face easily (Hammond et al., 2005), this type of analysis is generally inaccessible to the majority of WBS patients.

Individuals with smaller atypical deletions show varying degrees of dysmorphology, making the assessment of craniofacial abnormalities even harder. For example, a patient in which the GTF2IRD1 gene was only partially deleted, with the

175

Chapter 4 – GTF2IRD1 in epidermal mechanisms translation start codon in exon 2 still present, was analysed with a stereo photogrammetric device and her facial features were quantified using three-dimensional

(3D) face surface imaging. When compared to dense surface models of face shape, constructed from a collection of 185 control and 85 WBS individuals aged 2 weeks to

20 years, the patient’s profile fell at the interface between the WBS and control groups, and the classification was described as mild, but not classic, WBS features (Tassabehji et al., 2005).

At the cellular level, the phenotype of enlarged lips in Gtf2ird1-/- mice was shown to be caused by an extreme thickening of the epidermis, arising from an augmented keratinocyte proliferation ratio, evaluated by the Ki67 marker, and the disorganised keratinocyte layers, as illustrated by K10 and K14 immunofluorescence.

The thin sections of affected epidermis highlighted the striking hyperkeratosis with expansion of the stratum corneum and stratum granulosum, and the skin barrier formation was incomplete in the dysmorphic zones.

Consistent with interfollicular epidermal hyperproliferation, there was suprabasal expression of K14 and an increase in the number of suprabasal layers expressing differentiation markers such as K10. Keratin intermediate filament proteins protect epithelial cells from mechanical and non-mechanical stresses (Gu and

Coulombe, 2007), so an obvious question to ask is whether the onset of proliferation corresponds to a primary defect arising from the lack of Gtf2ird1 or whether the observed abnormalities in Gtf2ird1-/- mice are the consequence of altered barrier function. Many defects of barrier function have profound impacts on the epidermis as a whole, and experimental disruption of the barrier formation results in neonatal lethality in transgenic mice (Imakado et al., 1995). Thus, the barrier function is defective in a variety of diseases (Roop, 1995) and, as expected, also in premature infants

176

Chapter 4 – GTF2IRD1 in epidermal mechanisms

(Hammarlund and Sedin, 1979; Wilson and Maibach, 1980). To date, there is no suggestion that individuals with WBS are affected by skin barrier defects.

The results on skin dye permeability in newborn mice (P0) could have been interpreted in two ways. The finding is suggestive of abnormal skin barrier development, but it was also possible that the formation was merely delayed. However, the TEWL analysis demonstrated that adults also show increased water loss, suggesting that the defect is permanent. However, this does not provide a complete explanation of how this arises. Lipid content in the skin is important in maintaining barrier function.

Lipid-containing lamella bodies fuse with the plasma membrane and disperse their contents extracellularly (Hardman et al., 1998). Therefore potential defects in lipid homeostasis should be evaluated in order to understand the cellular basis of this abnormality. One way to do this is to image neutral lipids histologically by oil red O staining (Mehlem et al., 2013), or to measure mRNA levels of specific genes involved in the metabolism of lipids present in skin such as phosphatidylinositols, phosphatidylethanolamines, phosphatidylcholines, acyl- and lyso-phosphatidylcholines as well as different types of ceramides, sphingosine, by real time PCR (Smyth et al.,

2008). Regional variations in lipid metabolism could explain the alterations in barrier function in Gtf2ird1-/- mice. Although the skin structure is broadly similar in all body regions, the proportion of different lipid fractions varies considerably at individual locations. In fact, skin at different body sites shows distinct patterns of barrier recovery that are likely to be related to structural and physiological differences and lipid-rich skin areas are the most vulnerable to barrier disruption (Proksch et al., 2008).

However, the most intriguing question is whether the defects are triggered at the molecular level due to the lack of GTF2IRD1, or whether they are consequence of additional environmental insults that produce skin barrier defects. Expression analysis

177

Chapter 4 – GTF2IRD1 in epidermal mechanisms showed that Gtf2ird1 is highly expressed in the affected zones of the developing face.

Histological and immunofluorescence analyses showed that the facial skin hyperplasia and the skin barrier defects are both present before birth, before the pups become exposed to the external environment. These findings suggest that all the observed phenotypes are direct consequences of the lack of Gtf2ird1 in the Gtf2ird1-/- mice.

Expression analysis revealed that by E14.5, Gtf2ird1 is present in the developing lips, hair follicles, vibrissae, nose and the external ears. This dynamic pattern changes with time in late foetal life, but the lip expression persists, intensifies and becomes more localised to a region defining the future lips. Close examination reveals that this expression is localised in the basal layer. The fact that Gtf2ird1 is highly expressed in the specific region where the phenotype is observed in adult life suggests that

GTF2IRD1 plays an important part in the mechanism that regulates skin and facial patterning during development.

Gtf2ird1 was also found to be expressed in other mucocutaneous junctions; namely the eyelid and the ano-rectal junction. How specialised skin regions are patterned and maintained is a key question in skin biology, including the mucocutaneous junctions. These occur at sites where mucosa transitions to skin, such as the lips, eyelids, nostrils, vagina, foreskin and anus. At mucocutaneous zones, the internal epithelium becomes external epidermis and the underlying lamina propria becomes dermis. Although Gtf2ird1 is expressed in some of these junctions during development, none of these skin portions showed abnormal phenotypes later in life.

Nevertheless, we postulate that GTF2IRD1 is the first known skin marker restricted to these tissue boundaries and there are no reported knockout phenotypes of a similar nature. In future studies it would be of key significance to address how these distinct boundaries are regulated. This question is made more intriguing when one considers

178

Chapter 4 – GTF2IRD1 in epidermal mechanisms that epidermis is continuously renewing itself through the proliferative expansion of basal stem cells and therefore any differentiation-specific information must be continuously provided to the newborn cells throughout life. We believe that further analysis could reveal a potential role for GTF2IRD1 in these mechanisms.

We have recently provided some key clues to GTF2IRD1 function through the isolation and identification of multiple protein partners, which are mostly involved in chromatin modification and transcriptional regulation (Carmona-Mora et al., 2015).

These data lead to a hypothesis in which GTF2IRD1 is a novel regulator of gene expression in epidermal development that functions via chromatin modification and suggests that GTF2IRD1 lays down epigenetic marks in the basal stem cells of part of the face that persist within this proliferating population. This would explain why the overgrowth in the nose and lip areas continues and becomes more exaggerated in adult life, even though Gtf2ird1 expression is switched off in adult skin.

Although mammalian skin is one of the best-studied stem cell driven epithelial systems to date, the origin of most of the stem cell populations found in the adult epidermis is still largely unknown (Watt et al., 2006). However, it is well known that skin stem cells show profound differences in transcriptional regulation between development and adulthood (Benitah and Frye, 2012). This idea is consistent with our hypothesis that GTF2IRD1 acts at the stem cell level during development. In general, stems cells are defined by their ability to self-renew indefinitely, but also to produce daughter cells that have different, more restricted properties (Smith, 2006). One example of differential transcriptional regulation between development and adulthood is the well know epidermal stem cell marker P63 (Levy et al., 2005). P63 is one of the earliest-induced transcription factors associated with epidermal fate that plays a major role in embryonic development of ectodermal lineages (Koster and Roop, 2007;

179

Chapter 4 – GTF2IRD1 in epidermal mechanisms

Vanbokhoven et al., 2011). Enhancer binding proteins, such as C/EBPα and C/EBPβ as well as microRNAs such as miR-203, repress P63 expression at the suprabasal layer, thus triggering the onset of differentiation as the cells begin the process of stratification

(Lopez et al., 2009; Yi et al., 2008).

This example illustrates the complex interplay between regulatory mechanisms that underpin the transition between one epidermal state and another and suggests that it is unlikely that GTF2IRD1 would act alone. However, examples of key players of these complex mechanisms in the developing epidermis are the epigenetic regulators Setd8 and H3K27me3 (Driskell et al., 2012; Ezhkova et al., 2011; Morey and Helin, 2010;

Surface et al., 2010). Setd8 is a family member of a GTF2IRD1 protein partner recently identified (Setd6) and H3K27me3 is a silencing epigenetic mark found in close proximity with GTF2IRD1 in an in vitro proximity ligation assay (Carmona-Mora et al.,

2015). By combining information from different GTF2IRD1 studies, it is possible to build hypotheses for future investigations that may lead to an understanding of skin development in specialized body zones.

180

CHAPTER 5 - IN VIVO GENE REGULATION ASSOCIATED WITH GTF2IRD1 IN LIP SKIN

181

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

5.1 INTRODUCTION

Several studies have illustrated the DNA binding properties of GTF2IRD1

(Polly et al., 2003; Tay et al., 2003; Vullhorst and Buonanno, 2003) and its ability to act as a transcriptional regulator in transgenic mouse systems (Issa et al., 2006). Analysis has also demonstrated that the GTF2IRD1 protein auto-regulates its own transcription through direct binding to its own promoter region (Palmer et al., 2010). However, much of the information regarding GTF2IRD1 gene targets remains unknown. Analyses of

Gtf2ird1 knockout brain tissues, for example, has led to a disappointing lack of useful information (O'Leary and Osborne, 2011).

As shown in the previous chapter, evidence supports an important role for

GTF2IRD1 in the patterning and anatomical development of facial soft tissues. The inactivation of Gtf2ird1 impacts on processes such as cellular proliferation and differentiation of facial epidermis. Therefore, molecular analyses of these tissues could provide an important insight into the transcriptional control mechanisms that underpin these cellular changes.

High-throughput cDNA sequencing, or massively-parallel sequencing of mRNA

(RNA-Seq), is a relatively new approach to transcriptome analysis based on next- generation sequencing (NGS) (Morin et al., 2008). This technology has emerged as a comprehensive and cost-effective approach for transcriptome profiling due to continuing advancement in the technology and significant decreases in the cost of NGS

(Mortazavi et al., 2008). Compared to microarray techniques, RNA-Seq can be performed without prior knowledge of reference sequences and enables a wide range of applications, including transcriptome de novo assembly (Grabherr et al., 2011;

Robertson et al., 2010), abundance estimation (Oshlack et al., 2010; Robinson and

182

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Oshlack, 2010), and detection of alternative splicing events (Anders et al., 2012), all of which have revolutionized our understanding of the extent and complexity of eukaryotic transcriptomes (Wang et al., 2009).

In this chapter, an RNA-Seq analysis of lip tissue from Gtf2ird1-/- mice is described, which was chosen as a means to understand changes in the transcriptional landscape associated with absence of Gtf2ird1 in lip tissue. The analysis was performed as a means to establish the primary gene expression defect in the epidermal cells. It is expected that this study will provide a list of dysregulated genes that are either direct or indirect targets of GTF2IRD1 transcriptional control.

Changes in expression of Gtf2ird1 (absence or over-representation) results in widespread gene dysregulation, possibly attributable to dysregulation of several downstream transcription factors and signalling molecules (Chimge et al., 2007;

Chimge et al., 2008; O'Leary and Osborne, 2011). In the analysis described here, genes involved in growth factor signalling and cell cycle progression were seen to be particularly relevant to the skin dysmorphology observed in this mouse model. It is also noted that a number of the dysregulated genes have known roles in brain development as well as epidermal differentiation and maintenance. Therefore, this study provides potential clues to the underlying mechanisms involved in a broader context or

GTF2IRD1 function, and potentially, the WBS profile.

183

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

5.2 RESULTS

5.2.1 RNA-Seq analysis

Total RNA was extracted from mouse top lip tissue from three female Gtf2ird1-/- mice and three wild type littermates. The quality of the total RNA samples was assessed by visualising the 28S/18S ribosomal RNA ratio using agarose gel electrophoresis and by Bioanalyzer (Agilent Biotechnologist, Inc) analysis (Figure 5.1). The six RNA-Seq libraries were sequenced on the Illumina HiSeq2000 platform to produce over 60 million, 100 nucleotide, paired-end reads per sample.

RNA-Seq data are a set of short RNA reads that are often summarized as discrete counts. Five analysis packages are commonly used for RNA-Seq differential expression studies, including; DESeq, edgeR, DESeq2, sSeq and EBSeq. According to a study that compared the relative performance of the packages in analysing six different sets of publically-available data, it was determined that DESeq2 and edgeR generally gave the best performance (Ching et al., 2014). Therefore, the differential gene expression analysis in this study was performed using the Bioconductor packages, edgeR (Robinson et al., 2010) and DESeq2 (Anders and Huber, 2010), as detailed in

Methods (chapter 2).

5.2.2 Differential gene expression in lip tissue of Gtf2ird1-/- mice

The differential expression analysis produced a large number of genes with significantly increased or decreased expression in the Gtf2ird1-/- mice. A number of up and down-regulated genes, found by edgeR, were then compared with the number found by DESeq2, revealing a large overlap in genes that were called as being significantly differentially expressed (Figure 5.2 A-B). By looking at the collective expression of

184

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin these genes in the individual mice, it was clear that the two different genotypes fall into well-separated groupings, as shown by the multidimensional scaling (MDS) plot (Figure

5.2 C). This is also obvious in the heat map showing the expression level of the top 500 differentially expressed genes (Figure 5.2 D). Although a similar number of up- regulated and down-regulated genes were identified, the most statistically significant dysregulated genes were within the up-regulated group (Figure 5.2 D). The top 50 up- regulated and down-regulated genes, found using edgeR, are listed in Table 5.1.

As demonstrated previously (Palmer et al., 2010), the knockout mutation in this mouse model deletes exon 2 containing the start codon, which blocks normal protein synthesis (reducing it to trace levels – 3% of normal) but does not prevent the production of mutant transcripts lacking exon 2. Gtf2ird1 transcripts were detectable in the knockout tissues by RNA-seq analysis but all lacked exon 2 as expected.

185

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Figure 5.1 Quality assessment of total RNA extracts from lip samples

(A) Image of RNA separated by agarose gel electrophoresis, stained with ethidium bromide and photographed under UV light. The lanes show top lip RNA samples from four of the six samples used in RNA-Seq. The bands correspond to 28S

(top) and 18S (bottom) ribosomal RNA. (B) Before conducting the RNA-Seq assay, the samples were further assessed using the Bioanalyser (Agilent Biotechnologist, Inc). A representative result from one of the Gtf2ird1-/- RNA samples is shown. 18S and 28S peaks are clearly visible. All of the RNA integrity numbers (RINs) ranged from 9.0 to

9.5 for the six analysed samples.

186

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Figure 5.2 Analysis of differential gene expression in lip tissue

(A-B) Venn diagrams of the sets of differentially expressed genes (DEGs) identified using EdgeR and DEseq2. (A) Genes up-regulated in Gtf2ird1-/- mice, and (B) genes down-regulated in Gtf2ird1-/- mice. (C) Multidimensional scaling (MDS) plot generated using edgeR with the top 500 DEGs, shows clustering of the 3 wild type

(WT) replicates and the 3 Gtf2ird1-/- (KO) samples. (D) Heat map generated using heatmap.2 function of the gplots package in R (R is a software environment for statistical computing of data sets) for the top 500 differentially expressed genes found using edgeR.

187

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.1 The top 50 differentially expressed genes (DEGs)

Up-regulated genes Down-regulated genes Log2FC Log2FC Slc14a2 8.39 Krt2 -4.52 Gm16026 10.57 Acsm1 -3.35 Lrrn4 9.33 1100001G20Rik -2.54 Cml2 9.16 F3 -1.51 Sp110 7.82 Tac4 -1.79 Gm15753 8.36 Cyp2w1 -2.50 Prr18 6.13 Cacng1 -1.86 C130026I21Rik 9.14 Ces2f -1.99 Echdc3 4.48 Paqr5 -1.34 AI427809 5.10 Cyp2f2 -2.11 Dhtkd1 4.31 Gm1110 -2.87 Gm7029 10.86 Sebox -2.09 Gm5335 9.11 Wnt7b -1.36 Ccdc60 6.30 Gm26888 -2.27 Apol7b 7.45 Slc6a19 -1.26 0610040J01Rik 4.75 Ucp3 -1.67 Ccser1 5.02 Capn9 -2.16 Gm8674 9.03 Actc1 -1.74 Fbxo16 5.43 H19 -1.39 BB557941 8.95 Lypd6 -1.48 Gm12114 7.01 Slc46a1 -1.22 Gm13191 8.71 Hsd17b14 -1.35 Tpo 7.68 Csrp3 -1.21 Gm16239 7.43 Gm1078 -2.06 Kcnk15 7.19 Awat1 -1.09 Chrm3 4.67 Gm10228 -1.47 Mecom 4.37 Plin5 -1.57 Gm7592 3.94 Pdk4 -2.07 Tie1 3.54 Fam57b -1.32 C86695 10.24 Gfra2 -1.23 Dlgap2 7.01 Mybph -1.42 Colgalt2 4.66 Myoz2 -1.43 Ush2a 7.52 Fabp3 -1.13 A530040E14Rik 6.04 Tfr2 -1.74 B3gntl1 3.71 Zbtb16 -1.80 Gm10653 5.36 Skint10 -1.52 Gm12495 8.97 Mustn1 -1.18 Dcc 8.26 AU021092 -1.40 AU019990 6.31 Tspear -1.53 Gm16028 10.30 Krtap20-2 -1.67 Arhgap8 3.83 Alox15 -2.47 Oas1a 5.00 Krt36 -0.95 AI481877 6.11 Rarres1 -1.10 BC026585 3.29 P2ry4 -1.22 Sp140 4.27 Cyp17a1 -1.82 Aqp9 3.00 Ankrd2 -1.49 Abcb5 9.79 Foxq1 -1.19 Gm12724 6.50 Vgll2 -1.62 Defb13 7.65 Gm12551 -1.16 Ric3 4.41 Pdlim3 -1.09 Log2FC (fold change) positive values = up-regulation while negative values indicate down-regulation

188

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Of the genes listed in Table 5.1, the most statistically significant up and down- regulated genes have functions in transcriptional regulation (Foxq1, Sebox, Csrnp3,

Rorc, Vgll2, Lmx1a, Zbtb16, Sp110), cell signalling (Wnt7, Chrm3, P2yr4, Arhgap8,

Rho, Au021092, Epgn, Lypd6, F3, Skint10, Tac4, Gfra2, Dcc, Lrnrn4, Pglyrp1, Ric3,

Glt25D2, Defb13, Tie1, Ush2a), calcium binding properties (Pdlim3, Rorc, Acsm1,

Alox15, Cacng1, Capn9, Csrp3, Cyp17a1, Cyp2f2, Gm1110, Trim9, Zbtb16), membrane transport (Slc14a2, Slc6a19, Slc46a1, Aqp9), apoptosis (Csrnp3, Zptb16, Sp110, Dcc,

Actc1, Pglyrp1), neurological processes (Chrm3, Dcc, Gfra2, Lmx1a, Rho, Slc6a19,

Trim9, Ush2a) and development and maintenance of the stratified epidermis (Sc114a2,

Aqp9, Krt2).

Keratin network dysregulation

As described in the previous chapter, up-regulation of keratin differentiation markers (K10 and K14) was observed by immunofluorescence analysis in the interfollicular epidermis of Gtf2ird1-/- mice. The RNA-Seq analysis revealed additional insights into the abnormal differentiation of the epidermal keratin network in the absence of Gtf2ird1. A total of 51 genes, encoding keratins and keratin associated proteins, were found dysregulated in lip skin from Gtf2ird1-/- mice. The analysis showed that 13 keratin genes and 31 keratin-associated protein genes have decreased expression in the Gtf2ird1-/- mice, while only 7 keratin genes were found to be up-regulated (Table

5.3).

The gene showing the largest fold change is K2 (Table 5.2), the gene product of which is commonly found in the granular layer of the stratified epidermis and is known to be involved in the later stages of keratinocyte differentiation (Collin et al., 1992)

189

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

(Figure 5.3 A-B). Most of the other down-regulated keratin genes are expressed in different layers of the hair follicle (K83, K75, K33b, K31, K35, K82, K71 and K27)

(Figure 5.3 C). The K2 protein is restricted to the uppermost suprabasal layers of the epidermis (Herzog et al., 1994; Rentrop et al., 1987) where it colocalises with K10

(Smith et al., 1999). In humans, K2 is expressed throughout the IFE but is strongly increased in palmoplantar skin and in the epidermis of other mechanically-stressed body sites (Rothnagel et al., 1994; Swensson et al., 1998). This expression pattern has led to the suggestion that K2 provides additional mechanical resilience to the epidermis. In mice, the lack of K2 causes acanthosis, hyperkeratosis of the ear and the tail epidermis, corneocyte fragility, increased transepidermal water loss, and local inflammation in the ear skin (Fischer et al., 2014).

Also up-regulated were genes that encode proteins that form heterodimers; K1 and K10, which are active during suprabasal differentiation, and K16 and K6a, which are active in hyperproliferation and in response to environmental challenges (Lessard et al., 2013). Mutations in the K16 gene and its partner K6a, cause pachyonychia congenita (PC), a disorder typified by dystrophic nails, painful hyperkeratotic calluses in glabrous skin, and lesions involving other epithelial appendages (Leachman et al.,

2005). In vivo analysis of K16 null mice revealed a context-dependant co-regulation of these two genes, demonstrating that K16 and K6a are involved in skin barrier maintenance and innate immunity co-regulated through cooperative mechanisms

(Lessard et al., 2013).

190

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.2 Differentially expressed keratins and keratin-associated protein- encoding genes

Down-regulated Log2FC FDR K2 -4.51 1.06E-59 Kap20-2 -1.67 3.92E-05 K36 -0.93 9.49E-05 Kap15 -1.17 2.90E-04 Kap16-4 -1.37 3.95E-04 Kap16-8 -1.36 4.61E-04 Kap6-1 -1.28 5.82E-04 Kap8-1 -1.15 8.89E-04 Kap14 -1.29 9.25E-04 K84 -0.81 1.10E-03 Kap16-3 -1.2 1.54E-03 Kap22-2 -1.17 3.41E-03 K83 -1.01 3.54E-03 Kap10-4 -1.05 4.39E-03 Kap5-5 -1.06 4.61E-03 Kap17-1 -0.94 4.67E-03 Kap31-1 -1.5 5.47E-03 Kap9-5 -1.71 6.81E-03 Kap6-2 -1.16 7.01E-03 Kap7-1 -1.07 7.24E-03 Kap13-1 -0.95 8.07E-03 Kap16-7 -1.02 8.59E-03 Kap10-10 -1.03 9.35E-03 Kap16-10b -1.11 1.08E-02 K75 -0.69 1.39E-02 Kap5-4 -1.06 1.43E-02 Kap1-3 -1.22 1.44E-02 Kap5-1 -1.07 1.45E-02 Kap31-2 -1.46 1.46E-02 Kap11-1 -0.84 1.46E-02 Kap5-2 -0.93 1.57E-02 Kap12-1 -0.79 2.40E-02 K85 -0.91 2.65E-02 Kap9-3 -1.01 2.68E-02 Kap8-2 -1.08 2.75E-02 K33b -0.81 2.89E-02 Kap26-1 -0.64 3.46E-02

191

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.2 (continued) Differentially expressed keratins and keratin-associated protein-encoding genes

Down-regulated Log2FC FDR K31 -0.8 3.77E-02 K35 -0.61 5.26E-02 K82 -0.54 5.64E-02 K71 -0.54 6.20E-02 K27 -0.52 6.46E-02 K42 -0.76 8.44E-02 Up-regulated Log2FC FDR K7 1.5 2.76E-08 K16 2.08 9.36E-07 K1 1.19 2.88E-05 K12 1.17 1.86E-03 K6a 0.89 5.77E-03 K80 0.77 6.12E-03 K10 0.65 3.04E-02

K: keratin gene; Kap: keratin associated protein gene. FDR (false discovery rate)

192

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Figure 5.3 Differentially expressed keratins in Gtf2ird1-/- mice

(A) Histogram showing all the keratins that were found to be significantly dysregulated in skin from the top lip of Gtf2ird1-/- mice. (B) Schematic representation of a transverse section through interfollicular epidermis showing the normal layer position of the dysregulated keratins, K1, K2 and K10 (C) Diagram showing a cross section through a hair follicle and the normal position of the dysregulated keratins. Note that

K75 is present in the medulla as well as in the Henle’s layer.

193

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Dysregulation of epidermal markers involved in skin development and homeostasis

A transcriptomic analysis of the proteome in skin tissue has shown that 66% of all protein-encoding genes (n=20344) are expressed in the skin and about 2% of these genes show an elevated expression in skin compared to other tissue types (Uhlen et al.,

2010; Uhlen et al., 2015). These elevated genes encode proteins that relate to squamous differentiation and formation of the outermost cornified layer, and are therefore expressed in the different layers of the epidermis. The remaining genes are expressed in melanocytes, hair follicles and dermal cells with functions that include pigmentation, hair development and connective tissue structure (Gerber et al., 2013).

In order to investigate the functional associations of the large set of DEGs generated via RNA-Seq analysis of Gtf2ird1-/- lip tissue, we used the BiNGO plug-in of

Cytoscape software to analyse all up- and down-regulated genes for enrichment according to ontology. The DEGs were grouped according to the role of their gene products in skin biology such as differentiation/development, barrier function/immunity, cell-cell adhesion/barrier and fatty acid biogenesis (Table 5.3).

194

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.3 DEGs involved in epidermal processes

Group; Gene & Log2FC Gene product information with literature references DEG (FDR) status

Differentiation / Development K2 -4.51 Keratin 2: type II epithelial keratin expressed in granular layer, (1.06E-59) commonly used as a marker of late stage differentiation (Collin Down-reg et al., 1992) (Figure 5.3) K1 1.19 Keratin 1: type II epithelial keratin expressed in spinous layer (2.75E-08) with K10 concomitant with stratification (Fuchs, 1993) (Figure Up-reg. 5.3) K10 Up- 0.65 Keratin 10: type I epithelial keratin expressed in spinous layer reg. (3.04E-02) along with K1 (Fuchs, 1993) (Figure 5.3) K6a Up- 0.89 Keratin 6a: commonly found in hyperplasia but not normal reg. (5.77E-03) epidermis (Rorke et al., 2010) (Figure 5.3) K16 Up- 2.08 Keratin 16a: together with K6 - involved in skin barrier reg. (9.30E-07) maintenance and innate immunity (Leachman et al., 2005; Lessard et al., 2013) Ivl 1.56 Involucrin: found in cornified layer, commonly used as a marker (1.0E-05) of keratinocyte terminal differentiation (Marekov and Steinert, Up-reg. 1998; Michibata et al., 2004; Yaffe et al., 1992) Ptger1 Prostaglandin E receptor 1: expressed and functionally active in Up-reg. keratinocytes, its presence is pronounced in the granular layer of the epidermis and it has been suggested that it regulates keratin differentiation (Konger et al., 2005) Fgf7 Up- 1.57 Fibroblast growth factor 7: major factor in promoting skin cell reg. (4.7E-10) proliferation. Overexpression of FGF7 led to cells which prematurely reached the skin surface before completing their program of terminal differentiation. Inhibits hair follicle morphogenesis (Guo et al., 1993; Richardson et al., 2009) 1.65 Fibroblast Growth Factor Binding Protein: encodes a secreted (3.2E-11) fibroblast growth factor carrier protein, which plays a critical Fgfbp1 role in cell proliferation, differentiation and migration by Up-reg. binding to fibroblast growth factors and potentiating their biological effects on target cells. It interacts directly with FGF7 having effects in epithelial repair (Beer et al., 2005) Eda -0.65 Ectodysplasin A: interacts with bone morphogenetic proteins (BMPs) to establish hair follicle patterning (Narhi et al., 2008) Down- (4.9E-02) reg.

195

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.3 (continued) DEGs involved in epidermal processes

Group; Gene & Log2FC Gene product information with literature references DEG (FDR) status

Barrier function / Immunity

Slc14a2 Solute carrier family 14 (urea transporter), member 2: expressed Up-reg. 8.25 in human keratinocytes, impacts on keratinocyte differentiation (8.6E-119) promoting barrier function (Grether-Beck et al., 2012) Aqp9 Up- 3.02 Aquaporin transporter: expressed in human keratinocytes, reg. (1.4E-37) impacts on keratinocyte differentiation and barrier function (Grether-Beck et al., 2012) Dll1 -0.52 Delta-like protein 1: confined to the basal layer and participates in Notch signalling (Fuchs and Raghavan, 2002) Down- (7.02E-02) reg. Tgfbr3 -0.57 Transforming growth factor (TGF)-beta type III receptor: (5.0E-02) expressed in differentiating keratinocytes, it is considered a cell Down- membrane proteoglycan involved in ageing and often functions reg. as a co-receptor with other TGF-beta receptor superfamily members. Ectodomain shedding produces soluble TGFBR3, which may inhibit TGF signalling. Decreased expression of this receptor has been observed in various cancers (Farage et al., 2010) Hmmr 1.3 Hyaluronan mediated motility receptor: involved in cell Up-reg. (7.5E-04) locomotion and stimulates basal keratinocyte proliferation and epidermal thickness and promotes fibroblast migration and excisional wound repair (Symonette and Kaur Mann, 2014; Tolg et al., 2014) Has1 -0.78 Hyaluronan synthase 1: extrudes hyaluronan through cell (6.1E-03) membrane improving tissue remodelling (Chen and Abatangelo, Down-reg 1999). Cd44 Up- 0.78 Cd44 antigen: acts as a receptor for hyaluronan promoting reg. (2.5E-03) regulation of keratinocyte proliferation in response to extracellular stimuli and maintenance of hyaluronic acid homeostasis. (Chen and Abatangelo, 1999) Defb1 0.52 Up-reg. (5.4E-02) Defb13 7.21 Up-reg. (5.3E-34) Beta defensins are innate immune peptides. They act as Defb14 1.74 signalling molecules in the immune system and chemo- attractants for T-lymphocytes and immature dendritic cells. They Up-reg. (1.9E-04) are mostly found in oral and airway epithelia (Dommisch and Defb15 6.52 Jepsen, 2015; Hans and Madaan Hans, 2014; Liu et al., 1997) Up-reg. (1.1E-17)

196

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.3 (continued) DEGs involved in epidermal processes

Group; Gene & Log2FC Gene product information with literature references DEG (FDR) status

Cell-cell adhesion /barrier Dsg1b 1.23 Up-reg. (5.4E-05) Desmogleins: stabilize cell-cell adhesion and are important in Dsg3 Up- 0.72 maintaining skin barrier function properties (Proksch et al., 2008) reg. (1.5E-02) Cdhr1 0.92 Cadherin-related family member 1: calcium-dependent cell Up-reg. (4.8E-03) adhesion molecule; usually expressed in retinal epithelium. Mutations in Cdhr1 are related to retinal degeneration (Duncan et al., 2012)

Fatty acid biogenesis Abca12 8.43 ATP-binding cassette sub-family A, member 12: a membrane Up-reg. (1.0E-06) transporter responsible for cholesterol efflux. Epidermal cholesterol synthesis is increased during permeability barrier repair (Proksch et al., 2008) Acer3 Up- 0.71 Alkaline ceramidase 3: hydrolyzes phytoceramide to reg. (8.0E-03) phytosphingosine and free fatty acid; down-regulation is associated with inhibition of both cell proliferation and apoptosis (Hu et al., 2010)

197

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Dysregulation of signalling pathways

The gene ontology analysis also pointed to the involvement of several genes important in multiple signalling pathways vital to epidermal morphogenesis and homeostasis, including growth factor signalling, cytokine stimulation, Wnt signalling, calcium signalling and the sonic hedgehog pathway (Table 5.4).

Different functional themes emerged in the up-regulated and down-regulated gene sets. The up-regulated genes were enriched for terms involving the cell cycle, immunity and response to stimulus, while the down-regulated genes were enriched for structural development /morphogenesis and cell differentiation. The terms associated with the down-regulated genes were also enriched for signalling, ion transport/homeostasis, and cell adhesion. Combining the up-regulated and down- regulated gene sets, there was significant enrichment of terms associated with anatomical structure development, cell differentiation, signalling, immune processes and neurogenesis. Ngf (nerve growth factor) (log2FC=2.04, FDR=9.34E-11), Fgf7

(fibroblast growth factor 7) (also known as keratinocyte growth factor) (log2FC = 1.60,

FDR=1.04E-08) and Fgf1 (fibroblast growth factor 1) (log2FC= 0.79, FDR=3.89E-03) were significantly up-regulated in the Gtf2ird1-/- mice. We also observed significant upregulation of Fgfbp1 (FGF-binding protein) (log2FC= 1.63, FDR=6.11E-10), the product of which binds to FGF7 and enhances its activity (Beer et al., 2005). While Ngf, and Fgf1 have increased expression in the Gtf2ird1-/- mice, their respective receptors,

Ngfr and Fgfr1 have decreased expression.

With increased growth factor signalling, one might expect to find increased markers of cellular proliferation and this was found to be the case. Several of the up- regulated genes are involved in creating the cytoarchitecture of microtubules required

198

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin for cell division during the mitotic phase of the cell cycle. For example, several genes involved in cytokinesis or spindle formation and stabilization were up-regulated, such as; Prc1, Anln, Aspm, Cenpe, Kif11, Tacc3 and Spag5. Genes involved in chromatin condensation and the correct segregation of chromosomes during cell division were also found to be increased (Sgol1, Smc2, Smc3, Smc4). And whilst genes involved in the mitotic stage of cell division were most pronounced, genes involved in the earlier phases of cell division, such as DNA replication and associated processes including

Brca1, Clspn, FancI, Fancd2, Ki67 and Nbn, were also elevated (Table 5.4).

Genes involved in signalling that were down-regulated include members of the transforming growth factor family, Tgfb2 (log2FC=-0.86, FDR=2.16E-03) and the bone morphogenetic protein genes (Bmp2, Bmp3, Bmp4, Bmp6). Down-regulation was also observed for Tgfb2, the gene product of which acts as an inhibitor of cellular proliferation (Hu et al., 1998; Kim et al., 2001; Lu et al., 2006), which is consistent with the other indicia of increased cell division and proliferation seen in the absence of

Gtf2ird1.

Several Wnt pathway genes were significantly dysregulated in this analysis. The

Wnt proteins initiate signal transduction via three major pathways, the canonical Wnt/β- catenin pathway, the planar cell polarity pathway and the Wnt-calcium pathway (Croce and McClay, 2008). In each of the three major pathways, the Wnt ligand binds to its cognate receptor, Frizzled. In this study, we found that four of the Frizzled receptor genes (Fzd1, Fzd2, Fzd5 and Fzd9) were significantly down-regulated in the Gtf2ird1-/- mice. In addition, The Wnt genes, Wnt-3, Wnt-6, Wnt-7b, Wnt-9a, Wnt-10b, and Wnt-11 were all down-regulated. The largest fold change was found to be Wnt-7b (log2FC= -

1.36, FDR=1.17E-07), which is active in the canonical Wnt pathway, followed by Wnt-

11 (log2FC=-0.66, FDR=1.90E-02), which encodes a protein that activates the non-

199

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin canonical Wnt-calcium pathway that affects cytoskeletal dynamics and cell adhesion

(Croce and McClay, 2008; Doi et al., 2010) and, finally, Wnt-10b (log2FC=-0.69

FDR=3.19E-02), which is associated with the promotion of epidermal differentiation and the development of hair follicles (Ouji et al., 2006).

The gene ontology analysis also highlighted down-regulation of Wnt-calcium pathway genes, as well as calcium ion homeostasis genes. A number of genes, generally involved in calcium signalling, were found to be up and down-regulated in the Gtf2ird1-

/- lip skin tissue, as listed in Table 5.4. Additionally, while several growth factor signalling genes were found to be over-expressed, genes involved in other signalling pathways, such as the sonic hedgehog pathway, were found to be inhibited. A complete list of the differentially expressed genes involved in signalling pathways can be found in

Appendix 1, at the end of this chapter.

FGF signalling pathway dysregulation

The FGF (fibroblast growth factor) and FGFR (fibroblast growth factor receptor) families regulate a variety of developmental processes, including brain patterning, branching morphogenesis and limb development (Beenken and

Mohammadi, 2009). These proteins regulate development by orchestrating mesoderm patterning in the early embryo and then guide organ development by the regulation of cell proliferation, differentiation and survival. These factors are also involved in the regulation of normal skin development and the cranial vault (Fuchs, 2008; Fuchs and

Horsley, 2008; Hajihosseini, 2008; Katoh and Katoh, 2009; Ornitz and Marie, 2002) and studies of gene expression utilising genetically modified mice have revealed important roles for FGFs and FGFRs in keratinocytes during skin development and

200

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin homeostasis (Marchese et al., 1995; Petiot et al., 2003; Steiling and Werner, 2003;

Werner and Grose, 2003; Yang et al., 2010).

Transcripts from two members of the FGF family, Fgf1 and Fgf7, were significantly increased in the absence of Gtf2ird1. Concomitantly, the FGF binding protein gene, Fgfbp1, was also significantly up-regulated. Fgfbp1 is found in keratinocytes throughout normal epidermis, and its gene product is known to interact with FGF7 and stimulates its activity in the skin, including its role in the proliferation of keratinocytes (Beer et al., 2005).

The FGF7 protein activates the tyrosine kinase receptor FGFR2B, instigating activation of the Ras pathway, stimulating canonical mitogen-activated protein kinase

(MAPK) signalling or the phosphoinositide 3-kinase (PI3K)/AKT and PLC pathways.

These growth factor-activated pathways are involved in cell growth, cell division and cell differentiation. The PI3K pathway is implicated in cell survival and polarity control and the PLC pathway may be necessary for cell adhesion (Yun et al., 2010).

It has also been reported that the actions of Fgf7 and Fgfbp1 are relevant to wound healing and that these signals will overcome the usual anti-proliferative actions of members of the TGF family (Beer et al., 2005). It is noteworthy that there are similarities in the profile of gene expression associated with the skin hyperproliferation in the Gtf2ird1-/- mice and factors that are known to be involved in wound healing.

201

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.4 DEGs involved in signalling processes in skin

Gene product Log2FC FDR Signalling

Ngf Nerve growth factor 2.04 9.3E-11 Growth factor

Fgf1 Fibroblast growth factor 1 0.79 3.9E-03 Growth factor

Fgf7 Fibroblast growth factor 7/ 1.60 1.0E-08 Growth factor Keratinocyte growth factor Fgfrl1 Fibroblast growth factor receptor-like 1 -1.14 6.3E-04 Growth factor

Fgfbp1 fibroblast growth factor binding protein 1.63 6.1E-10 Growth factor 1 Bmp2 bone morphogenetic protein 2 -0.63 4.4E-02 Growth factor

Bmp3 Bone morphogenetic protein 3 -0.81 5.1E-02 Growth factor

Bmp4 Bone morphogenetic protein 4 -0.97 2.1E-03 Growth factor

Bmp6 Bone morphogenetic protein 6 -0.89 1.3E-03 Growth factor

Tgfb2 Transforming Growth Factor, Beta 2 -0.86 2.2E-03 Growth factor

Prc1 Protein regulator of cytokinesis 1 1.04 1.8E-04 Cytokinesis/spindle formation Anln Anillin, actin binding protein 1.12 1.0E-04 Cytokinesis/spindle formation Aspm Asp (abnormal spindle) homolog, 1.15 6.8E-03 Cytokinesis/spindle microcephaly associated (Drosophila) formation Cenpe Centromere protein E 1.45 7.2E-04 Cytokinesis/spindle formation Kif11 Kinesin Family Member 11 1.13 1.9E-03 Cytokinesis/spindle formation Tacc3 Transforming, Acidic Coiled-Coil 3 0.57 4.4E-02 Cytokinesis/spindle formation Spag5 Sperm Associated Antigen 5 0.84 3.2E-03 Cytokinesis/spindle formation Sgol1 Shugoshin-Like 1 0.90 1.1E-02 Chromatin/Chromosome structure Smc2 Structural maintenance of 1.21 1.6E-03 Chromatin/Chromosome chromosomes 2 structure Smc3 Structural maintenance of 0.95 0.8E-03 Chromatin/Chromosome chromosomes 3 structure Smc4 Structural maintenance of 1.17 1.6E-03 Chromatin/Chromosome chromosomes 4 structure Brca1 Breast Cancer 1, Early Onset 0.76 4.2 E-02 DNA replication / cell division Clspn Claspin 0.65 7.8E-02 DNA replication / cell division FancI Fanconi anemia, complementation 0.84 4.1E-02 DNA replication / cell Group I division

202

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.4 (continued) DEGs involved in signalling processes in skin

Gene product Log2FC FDR Signalling

Fancd2 Fanconi anemia, complementation 0.71 6.1E-02 DNA replication / cell Group D2 division Nbn Nibrin 0.64 6.4E-02 DNA replication / cell division Ki67 Protein MKi67 1.28 8.8E-04 DNA replication / cell division Wnt3 Wingless-type MMTV integration site -0.64 5.3E-02 Wnt family member 3 signalling Wnt6 Wingless-type MMTV integration site -0.74 3.8E-02 Wnt family member 6 signalling Wnt7b Wingless-type MMTV integration site -1.36 1.8E-07 Wnt family member 7B signalling Wnt9a Wingless-type MMTV integration site -0.74 2.8E-02 Wnt family member 9A signalling Wnt10b Wingless-type MMTV integration site -0.69 3.2E-02 Wnt family member 10B signalling Wnt11 Wingless-type MMTV integration site -0.68 1.5E-02 Wnt family member 11 signalling Fzd1 Frizzled class receptor 1 -0.68 1.2E-02 Wnt signalling Fzd2 Frizzled class receptor 2 -0.82 1.2E-02 Wnt signalling Fzd5 Frizzled class receptor 5 -0.57 4.0E-02 Wnt signalling Fzd9 Frizzled class receptor 9 -1.54 2.0E-03 Wnt signalling Shh Sonic hedgehog -1 2.5E-03 Sonic hedgehog

Ptch2 Patched2 -0.7 1.4E-02 Sonic hedgehog.

203

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Differentially expressed transcription factors

Twenty two of the DEGs were found to fall into the category of transcription factors (Table 5.5). Examination of the functional annotations associated with these transcription factors shows that 12 genes cluster into an association with tissue development, including embryonic development, and another 10 genes are associated with response to stimulus or immune processes. A subset of the differentially expressed transcription factors active in organ development are also specifically involved in brain development (Lh2, Gli2, Pou3f3 and Sox2), which may be particularly relevant to WBS.

As the dysregulated transcription factors are likely to affect the expression of their own set of target genes, this ripple effect may explain the large number of differentially expressed genes identified in this study. We have tested this assumption in the case of

Lhx2, which encodes a transcription factor that is important in epidermal differentiation and neurogenesis (Peukert et al., 2011; Rhee et al., 2006).

Lhx2 expression is decreased in lip skin in the absence of Gtf2ird1 (Log2FC=-

0.72; FDR=1.80E-02). The MetaCore database from Thomson Reuters was used to identify genes transcriptionally regulated by Lhx2. This analysis identified 429 genes with known or putative binding sites for Lhx2. Of these, 86 (20%) are differentially regulated in this study (Table 5.6). In the context of WBS, it is noteworthy that four of these genes, which are down-regulated in this study (Lhx8, Sobp, Ccl2, Adh1), are involved in behaviour control (GO access number: 0007610). Sobp (sine oculis binding protein), which has previously been shown to have strong expression in the limbic system and regulates learning, memory and affective behaviour, is reduced by 50% in the Gtf2ird1-/- mice. Moreover, mutations in Sobp are associated with intellectual disability (Birk et al., 2010).

204

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.5 Differentially expressed transcription factors

Gene symbol Log2FC FDR Role Hey1 -0.81 2.70E-02 Development Myf6 -1.27 2.50E-03 Development Myog -1.11 1.90E-02 Development Sebox -2.09 2.10E-07 Development Gli2 -0.77 5.50E-03 Development Lhx2 -0.72 1.80E-02 Development Mef2c -0.66 4.10E-02 Development Pou3f3 -1.59 5.70E-03 Development Sox2 -1.08 3.10E-02 Development Chd7 0.59 3.60E-02 Development Epas1 -0.64 1.20E-01 Development Foxp3 1.43 1.60E-03 Development Fosl1 1.38 Response to 6.20E-06 stimulus Rorc -1.26 Response to 1.00E-04 stimulus Ier3 0.54 Response to 5.70E-02 stimulus Irf7 1.98 Response to 1.40E-09 stimulus Irf9 1.84 Response to 6.60E-14 stimulus Rel 0.68 Response to 4.80E-02 stimulus Stat1 1.81 Response to 1.50E-08 stimulus Stat2 1.6 Response to 8.10E-08 stimulus Taf1 0.9 Response to 1.40E-02 stimulus Tgs1 0.87 Response to 2.20E-02 stimulus Genes highlighted in bold are involved in brain development.

205

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.6 DEGs classified as Lhx2 downstream targets found to be dysregulated in Gtf2ird1-/- mice

Down-regulated Log2FC FDR gene Acsm1 -3.35 8.1E-13 Ankrd2 -1.49 7.5E-05 Krtap14 -1.29 1.4E-03 Il22ra2 -1.19 2.1E-02 Krtap16-3 -1.19 1.7E-03 Ttn -1.17 4.0E-02 Krtap8-1 -1.16 6.5E-03 Gnb3 -1.09 4.0E-02 Pdlim3 -1.09 1.0E-04 Adh1 -1.05 7.6E-02 Ranbp17 -1.01 4.2E-02 Lhx8 -1.00 8.9E-03 Has1 -0.96 4.4E-02 Krtap13-1 -0.95 8.0E-03 Krtap17-1 -0.95 3.6E-03 Sobp -0.94 7.8E-03 Myoz3 -0.90 6.4E-02 Ndn -0.89 2.7E-02 C1qtnf9 -0.89 2.8E-02 Dact1 -0.87 3.3E-02 Dysf -0.87 4.3E-03 Cxcl14 -0.84 1.7E-02 Slc43a1 -0.84 1.7E-02 Htra3 -0.83 1.9E-02 Aldh1a2 -0.80 8.6E-03 Loxl2 -0.79 4.4E-02 Iqsec3 -0.79 1.1E-01 Tspan12 -0.77 1.6E-02 Mgll -0.77 7.1E-02 6330403K07Rik -0.75 2.5E-02 Phactr1 -0.73 2.7E-02 Hhip -0.72 8.9E-02 Thbs4 -0.71 3.7E-02 Cd34 -0.71 1.9E-02 Adcy2 -0.67 6.1E-02 Mef2c -0.66 4.1E-02 Myh4 -0.66 2.4E-01 Rnase1 -0.65 9.4E-02 Tmtc1 -0.63 5.0E-02 Tmem100 -0.63 5.5E-02 Cyp2d22 -0.61 3.6E-02 Procr -0.56 6.9E-02 Dtna -0.55 8.2E-02 Gabarapl1 -0.54 6.5E-02 Cryl1 -0.53 1.1E-01 Lgr5 -0.49 9.0E-02 Vamp1 -0.35 3.2E-01

Down-regulated genes highlighted in bold are involved in behaviour.

206

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Table 5.6 (continued) DEGs classified as Lhx2 downstream targets found dysregulated in Gtf2ird1-/- mice

Up-regulated Log2FC FDR gene Plxna2 0.55 4.4E-02 Stk4 0.59 5.7E-02 Il1f5 0.63 1.1E-01 Rab5a 0.63 6.5E-02 Gsta3 0.71 6.0E-02 Tfpi2 0.76 1.1E-01 Fgf1 0.79 3.9E-03 Elf3 0.79 1.2E-01 Lce1e 0.79 4.4E-02 Cast 0.83 1.2E-02 Casp1 0.85 4.9E-02 Lce1h 0.86 3.4E-02 Itsn2 0.90 1.5E-02 Parp8 0.94 1.3E-04 Tdrd7 0.97 5.7E-04 Camk4 1.00 6.2E-02 Esd 1.02 9.0E-05 Zfp677 1.03 5.0E-02 Mphosph10 1.03 8.7E-03 Ifi47 1.05 4.8E-02 Dsc1 1.19 4.3E-03 Scml2 1.28 5.9E-05 Serpinb6c 1.36 8.5E-03 Serpinb11 1.43 3.5E-03 Fam171b 1.49 2.2E-06 Ccl2 1.51 7.1E-07 Iigp1 1.69 1.6E-03 Robo1 1.70 2.8E-16 Gbp2 1.71 1.7E-05 Il1f9 1.76 5.3E-08 Gbp7 1.89 1.9E-07 Hsp90aa1 1.97 6.1E-08 Gucy1a3 2.02 1.7E-13 Tfec 2.29 6.6E-06 Samd12 2.56 5.9E-08 Ceacam1 3.00 4.9E-27 Sprr2d 3.47 2.2E-04 0610040J01Rik 4.75 6.9E-54 Ush2a 7.52 3.9E-46

Up-regulated genes highlighted in bold are involved in behaviour.

207

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Genes annotated as associated with anatomical structure development and neural process

The data were also analysed with respect to molecular mechanisms that could relate to the WBSCP. Although the analysed tissue is unrelated to the cognitive features of WBS, this analysis could provide some clues to pathways that overlap between these different cellular systems. The gene ontology analysis indicated that within the category of ‘anatomical structure development’ 135 genes (40%) were found to be dysregulated that also have annotations relating to the nervous system and brain development, and 64

(19%) have annotations relating to epithelial development. Some of these genes are reported to have functional roles in both types of tissue. Examples of genes with known roles in both epithelial development and brain development include the transcriptional regulators, Lhx2, Lmo4, Six1, Sox2, Tbx3, Tulp3 and Twist1. Genes involved in signalling pathways in both types of tissue include those that are involved variously in growth factor signalling, sonic hedgehog signalling and Wnt signalling, such as: Fgfr1,

Tgfb2, Vegf3, Bmp2, Bmp4, Bmp6, Nog, Ntf5, Ret, Shh, Gli2, Wnt7b and Sfrp2.

Direct interaction partners of Gtf2ird1

In order to evaluate the possibility of coordinated regulation of proteins involved in complexes or in related functions, the data were interrogated for the presence of genes that encode GTF2IRD1 protein partners. Firstly, a search for the small number of direct interaction partners that are currently annotated was initiated. These proteins include the E3 SUMO-protein ligase PIAS2 (Tussie-Luna et al., 2002a), the histone deacetylase HDAC3 (Tussie-Luna et al., 2002a), EXOSC4 (Lehner and Sanderson,

2004) and the retinoblastoma-associated protein (RB1) (Yan et al., 2000).

208

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

A recent study from our laboratory, utilising unbiased yeast 2-hybrid methods, has identified 38 novel proteins that are capable of direct binding to GTF2IRD1

(Carmona-Mora et al., 2015). These partners can be classified into different functional groups, such as chromatin modifying proteins, transcriptional regulators, proteins involved in signalling pathways or post-translational modifications (ubiquitination and

SUMOylation), and proteins associated with primary cilia.

Cross-referencing of these proteins with the genes found to be significantly dysregulated in the RNA-Seq analysis, revealed only five GTF2IRD1 partners: ATF7IP

(Log2FC=0.76), ALMS1(Log2FC=0.87), KPNA2 (Log2FC=1.07), PARPBP

(Log2FC=0.76) and TRIP11 (Log2FC=0.75); of these only ATF7IP was validated in mammalian cells, the others were confirmed only by yeast two-hybrid (Carmona-Mora et al., 2015). All of them are found to be up-regulated in this analysis. In addition, at least one of their respective interacting partners is also significantly dysregulated in the

RNA-Seq DEG list (interactions retrieved from IntAct database) (Kerrien et al., 2012).

This includes KPNA2, a nuclear importin (Kohler et al., 1999) that interacts with

MKI67, the transcript of which is also found to be up-regulated in the analysed samples

(Table 5.4). Another example is the down-regulation of HEYL (Log2FC=-0.66), a transcriptional repressor that binds to ATF7IP, and acts as a downstream effector of

Notch signalling (Weber et al., 2014).

Unpublished data from our laboratory has recently shown the direct interaction between GTF2IRD1 and the histone deacetylase HDAC2, which is involved in a range of chromatin modifier complexes (de Ruijter et al., 2003) and is down-regulated in the

RNA-Seq DEG list (Log2FC=052). An alteration in the levels of HDAC2 transcript may lead to a perturbation of gene regulation exerted by the complexes in which this

209

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin protein is involved, thus adding another means by which loss of GTF2IRD1 could cause differential expression in many downstream genes.

Gene targets of Gtf2ird1

The large number of differentially expressed genes identified in this study also contrasts with the small number of gene targets described for Gtf2ird1 that are currently annotated. Among the described gene targets currently described for GTF2IRD1 (a total of 24 retrieved from METACORE and BIOgrid databases), we found significantly altered expression in only two of these genes, Ccnd3 (Cyclin D3) and Tfgb2

(Transforming growth factor beta 2) (Chimge et al., 2007), which are reported to be transcriptionally activated by GTF2IRD1. CCND3 plays an important role in cell cycle progression and in phosphorylation of RB1. In addition to its interaction with the RB1 and GTF2IRD1, CCND3 also affects other transcription factors and may have a repressive effect on some genes; e.g. Runx1 (Peterson et al., 2005) and the androgen receptor (Zong et al., 2007), or an activating effect; e.g. PPARG (Sarruf et al., 2005),

RARA (Despouy et al., 2003) and VDR (Jian et al., 2005). TGFB2 plays a role in the process that leads to the phosphorylation of SMAD, its translocation to the nucleus and ultimately transcriptional activation.

5.2.3 Analysis of the differentially expressed genes by RTqPCR

Although the criteria for discovery of DEGs was based on parameters that ensure a high stringency and low FDR values, some of the DEGs were also measured using the reverse transcriptase quantitative polymerase chain reaction (RTqPCR) technique. Using this method, and by including samples from new animals that were not

210

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin in the original experimental cohort, it is possible to validate the RNA-Seq findings for these genes and ensure that the gene dysregulation is consistently associated with the

Gtf2ird1-/- genotype.

The DEGs analysed by this method were selected as being of primary interest due to their potential relevance to GTF2IRD1 function, and hypotheses of downstream functional consequences. Therefore, RTqPCR analysis was performed for two of the differentially expressed transcription factors identified in this study; Tfgb2 and Lhx2, and three of the differentially expressed genes involved in signalling pathways; Fgf7,

Fgfbp1and Wnt11.

The RTqPCR analysis was performed from a mix of samples composed of 2 samples per genotype that were previously utilised in the RNA-Seq analysis and 2 new cDNA samples per genotype. Levels of expression for the genes Fgf7, Fgfbp1, Lhx2,

Tgfb2 and Wnt11, were quantitated in comparison with the housekeeping gene HPRT

(hypoxanthine phosphoribosyltransferase 1) as an internal reference standard, using the delta Ct (ΔCt) method (Livak and Schmittgen, 2001). The results of this analysis are presented in Figure 5.4.

In agreement with the RNA-Seq results, ΔCt values for Fgf7, Fgfbp1 and Lhx2 were found to be significantly different in the Gtf2ird1-/- mice compared to wild type controls (Fgf7 P=0.011; Fgfbp1 P=0.0014; Lhx2 P=0.048). ΔCt values for Tgfb2 and

Wnt11 were not significantly different in this assay (Tgfb2 P=0.158; Wnt11 P=0.105)

(Figure 5.4 A). When the means of the ΔCt values per group were expressed as fold change relative to the control, which was set at 1, Fgf7 and Fgfbp1 showed an increase in expression and Lhx2 showed a decrease, which is in agreement with the RNA-Seq findings (Figure 5.4 B). A decrease in transcript abundance was also found for Tgfb2

211

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin and Wnt11, which agrees with the direction of the trend found in the RNA-Seq analysis, but perhaps due to the small sample size, the difference in the methodology or the variability within the additional samples, these genes did not achieve a statistically significant difference between genotypes.

212

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Figure legend in next page

213

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Figure 5.4 RTqPCR validation of DEGs identified by RNA-Seq analysis

(A) Scatter plots showing RTqPCR analysis using RNA extracts from lip tissue of Gtf2ird1-/- and wild type control mice (n=4 per genotype) with Hprt quantification as an internal reference. Genotype means of cycle threshold values (ΔCt) are indicated

(long horizontal bar) (* p≤ 0.05, ** p ≤ 0.01; two tailed Student’s T-test) (B) The same results are expressed as fold change in expression levels relative to the control, which is set at 1 in all cases. Error bars in A and B represent the standard error of the mean

(SEM).

214

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

5.2.4 Dysregulation of FGF7 at the translational level

As described above, a significant level of up-regulation was found for transcripts derived from the Fgfbp1 and Fgf7 genes, by two different methods (Fgfbp1, log2FC=1.63, FDR=6.11E-10 in RNA-Seq, FC=2.51, P=0.0014 in RTqPCR; Fgf7, log2FC=1.57, FDR=4.7E-10 in RNA-Seq, FC=1.63, P=0.011 in RTqPCR). Evidence has shown that FGFBP1 binds directly to FGF7 and enhances its activity; an event that plays an important role in epithelial repair processes (Beer et al., 2005). Because of its role in keratinocyte differentiation, the FGF7 protein is also called keratinocyte growth factor, and its targeted overexpression under the control of the K14 promoter has been shown to produce an epidermal phenotype that includes thickening of the epidermis, wrinkling of the skin, stunted hair growth with an inverse correlation between epidermal thickening and hair follicle density (Guo et al., 1993). FGF7 is also thought to promote epidermal differentiation at the expense of hair morphogenesis (Richardson et al.,

2009).

It is therefore striking to find up-regulation of Fgf7 and Fgfbp1 transcription in the Gtf2ird1-/- model, in association with epidermal thickening and skin wrinkles in the lip and nose areas (Howard et al., 2012) (chapter 4), together with down-regulation of most of the keratins associated with hair follicles (Figure 5.2 A and C). These findings are in agreement with what has been described for FGF7 function (Guo et al., 1993;

Richardson et al., 2009). A previous study of GTF2IRD1 molecular function found that when GTF2IRD1 was overexpressed in primary mouse embryonic fibroblasts (MEFs), a microarray analysis identified up-regulation of Fgf7, but this gene was not within the genes selected for further RTqPCR validation (Chimge et al., 2007).

215

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

These findings suggest that FGF7 could be playing an important role in the lip phenotype of Gtf2ird1-/- mice and led us to ask the question of whether FGF7 abundance could show differences at the protein level. A qualitative Western blot analysis was conducted using protein extracts isolated from lip skin of Gtf2ird1-/- and control wild type littermates, utilising β-actin as a loading control (Figure 5.5). β-actin levels show some variability among the samples, which could be a consequence of differential expression of this cytoskeletal protein in keratinocytes that are experiencing hyperproliferation and abnormal differentiation. Nevertheless, Ponceau red staining of total protein indicates reasonable loading equivalence. In spite of this complication, the analysis shows that the cognate band corresponding to the FGF7 protein is highly abundant in all of the Gtf2ird1-/- samples, but undetectable in all of the wild type controls (n=3 per genotype).

The results show that the cognate band corresponding to FGF7 protein is highly abundant in all of the Gtf2ird1-/- samples, but undetectable in all of the wild type controls (n=3 per genotype).

216

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Figure 5.5 FGF7 protein increase in Gtf2ird1-/- lip skin extracts

Western blot analysis detecting FGF7 in total protein extracts isolated from lip skin from 3 Gtf2ird1-/- and 3 wild type control littermates. 20µg of protein extract per sample was electrophoresed in a 10% SDS-Polyacrylamide gel. Immunoblotting using an anti-FGF7 monoclonal antibody detects a cognate band at 28KDa only in the Gtf2ird-

/- samples. Other bands at ~250KDa and 70KDa are of unknown origin and could be non-specific cross-reacting products. After FGF7 western blot analysis, the PVDF membrane was first stained with Ponceau red and then re-probed with an anti--Actin antibody (42KDa), in order to control for protein loading. Protein mass KDa was estimated relative to a molecular weight (MW) ladder.

217

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

5.3 DISCUSSION

This chapter describes the interrogation of the transcriptome using RNA-Seq analysis to understand the consequences of Gtf2ird1 loss in lip tissue. We have identified a large number of up-regulated and down-regulated genes, revealing that the absence of Gtf2ird1 has widespread effects. Gene ontology analysis indicates that these changes in gene expression impact on multiple gene classifications, including; structural/morphological development of tissues, cellular differentiation, cellular proliferation, the immune response, transcription factors, signalling molecules and cell adhesion proteins.

The findings described in the previous chapter indicated an increase in the cellular proliferation rate at the basal layer of lip skin epidermis and an increase in the number of differentiating keratinocytes, as manifested by thickened basal and suprabasal layers in the Gtf2ird1-/- mice. However, the abnormality was not just restricted to an enlargement of the usual layers. RNA-Seq analysis of the same tissue revealed an imbalance in the complement of keratin genes that are expressed. For example, the expression of genes, such as K2, that are normally active in the terminally differentiating keratinocytes in the granular layer of the epidermis, were found to be significantly reduced.

The differential expression of keratins and keratin associated proteins is consistent with the hyperproliferation of epithelial cells, accompanied by a failure of the keratinocytes to fully differentiate at the usual stage within the granular layer. The downregulation of K2, which was the most reduced of the keratin genes, contrasts with the upregulation of both K1 and K10, the gene products of which are known to form heterodimers (Virtanen et al., 2003). Studies on the K2 knockout mice, which have K2-

218

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin deficient suprabasal keratinocytes, showed that the loss of this gene was partially compensated by the upregulation of K1 expression. The developing keratinocytes of this mouse line were shown to lack a regular cytoskeleton and developed large aggregates of

K10 protein (Fischer et al., 2014).

The observed keratin network dysregulation in the Gtf2ird1-/- mice also points to defects in hair follicle development/maturation and barrier function disruption. Most of the downregulated keratin genes found in the analysis are components of the different layers of the hair follicle structure (i.e. K83, K75, K33b, K31, K35, K82, K71 and K27).

This result is somewhat surprising as the lip skin is not an expected site of hair follicle formation. 12 of the DEGS identified, and many of the dysregulated keratins, are known to be involved in processes important for immunity and barrier function of the skin.

These findings are consistent with the skin barrier function abnormalities described in the previous chapter that were identified using the toluidine blue permeability and

TEWL assays.

Gene ontology analysis of down-regulated and up-regulated genes revealed an enrichment of terms associated with tissue development and signalling in the down- regulated genes while the up-regulated genes were enriched for genes involved in the cell-cycle and the immune response. The up-regulation of genes involved in cell division is consistent with cellular hyperproliferation. Indeed, Ki67, the cellular proliferation marker utilised to evaluate keratinocyte proliferation in the previous chapter was also found upregulated in the RNA-Seq analysis (Log2FC=1.28;

FDR=8.8E-04).

Another strong signature identified within the cell-cycle and proliferation-related genes is the downregulation of Ccnd3, the gene product of which (Cyclin D3) is

219

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin involved in the phosphorylation of RB1 in association with CDK4 or CDK6 (Ren and

Rollins, 2004). The under-phosphorylated form of RB1 interacts with E2F1 and represses its transcriptional activity, leading to cell cycle arrest (Ren and Rollins,

2004). Therefore, a reduced level of Cyclin D3 might be expected to decrease cellular proliferation. This change in transcription could be part of a secondary compensatory mechanism that is triggered in order to control the increased level of proliferation in

Gtf2ird1-/- lip tissue.

GTF2IRD1 has been shown to complex with the RB1 transcriptional repressor

(Yan et al., 2000). RB1 controls cell cycle progression by interacting with the E2F family, preventing the cell from progressing from G1 to S phase and by attracting histone deacetylases (HDACs) to the chromatin, thus suppressing DNA synthesis. It has also been demonstrated that GTF2IRD1 (and TFII-I) associates with HDAC3 and to a lesser extent with HDAC1 (Tussie-Luna et al., 2002a). It is further proposed that this activity is refined by the activity of PIAS2 which associates with HDAC3 (Tussie-Luna et al., 2002b). Recent work in our laboratory also indicates that GTF2IRD1 can interact directly with HDAC2 (Carmona-Mora et al. unpublished). HDAC interactions are generally taken as evidence for involvement in transcriptional regulation via histone deacetylation. However, these associations also allow the possibility that GTF2IRD1 and TFII-I are somehow involved in HDAC functions during DNA replication and cell cycle control and the reported interaction with RB1 (Tussie-Luna et al., 2002a) would support such an idea.

However, if this hypothesis is correct, it must then be asked why loss of

GTF2IRD1 only impacts on cellular proliferation in facial skin zones and not in the many other sites where Gtf2ird1 is normally expressed. It is also difficult to understand how such a mechanism would work given that Gtf2ird1 expression is not detectable in

220

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin the adult basal skin progenitor cells and, therefore, the protein cannot be present to engage in such interactions with HDACs and RB1. It is much more likely, therefore, that the impact on cellular proliferation is not via direct regulatory control of the cell cycle machinery but via the epigenetic regulation of other genes that control such processes in adult life.

Several model systems indicate that GTF2IRD1 acts mainly as a transcriptional repressor (Palmer et al., 2010; Polly et al., 2003; Widagdo et al., 2012), which predicts that the primary targets of GTF2IRD1 would be increased by the loss of this protein in the Gtf2ird1-/- mice. It is noteworthy that the most statistically significant differences in gene expression fall within the upregulated group, in accordance with these expectations. GTF2IRD1 overexpression is also known to affect gene transcription

(Chimge et al., 2007; O'Mahoney et al., 1998; Tantin et al., 2004), but only a small number of direct targets have been identified to date.

The analysis pointed to the dysregulation of a number of transcription factors involved in tissue development. We have drawn attention to one of these factors, LHX2, which is active in tissue development and in neurological processes. LHX2 regulates hair follicle development (Folgueras et al., 2013) and skin repair (Mardaryev et al.,

2011). It has also been described as a central link in the genetic networks that coordinate multiple signalling pathways controlling organ development and cell fate determination

(Mardaryev et al., 2011), including playing an important role in the regulation of brain development (Peukert et al., 2011). LHX2 acts as an activator of the sine oculis binding protein (SOBP) (Mardaryev et al., 2011), which has a strikingly specific pattern of expression in the limbic system of the brain, with disruptions leading to abnormal cognition and intellectual disability (Birk et al., 2010). The RNA-Seq analysis also showed expression changes in many genes that are normally regulated by LHX2

221

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

(Mardaryev et al., 2011). These effects and correlations with LHX2 function seem relevant to both the skin abnormalities of the mice and the neurological phenotypes of

WBS and thus, the changes in expression of Lhx2 were validated by RTqPCR.

Altered expression levels were found for a number of genes involved in signalling pathways, including growth factor signalling, Wnt, BMP and the sonic hedgehog pathway. We found dysregulation in the FGF family members and the tissue phenotype observed in the Gtf2ird1-/- knockout seems a good fit with the role of these proteins in the regulation of organ development by controlling cell proliferation, differentiation and survival. In particular, Fgf7 was found to be significantly upregulated in the RNA-Seq analysis and we confirmed upregulation of the transcript by RTqPCR and FGF7 protein abundance by western blot analysis. Since FGF7 is only mitogenic for epithelial cells and not other cell types (because these cells express the appropriate receptor), it has been suggested that it is a unique stromal mediator of epithelial proliferation (Powers et al., 2000). Transgenic overexpression of FGF7 under the control of the keratin 14 (Krt14) promoter, which directs localization to the epidermal basal layer, resulted in an epidermis that was markedly thickened and wrinkled, and resulted in an inverse correlation between epidermal thickening and hair follicle density (Guo et al., 1993). This association between FGF7 and skin overgrowth fits with the observed phenotype in the lips of Gtf2ird1-/- mice and could also explain the downregulation of keratins and keratin-associated proteins that are involved in hair follicle development and maintenance.

The upregulation of FGF7 found in Gtf2ird1-/- mice disagrees with one of the previous studies on GTF2IRD1 function in MEFs, which was examined by overexpression of the recombinant protein (Chimge et al., 2007). However, this analysis has several technical and biological limitations. The upregulation of Fgf7 gene was

222

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin detected by microarray but not verified by RTqPCR and the control mRNA extracts were taken from untransfected MEFs instead of MEFs transfected with empty vector.

This means that any transcript profile differences found could have not only been induced by the increased level of GTF2IRD1 but also by the process of transfection.

MEFs are also a different cellular context and the immediate effects of GTF2IRD1 expression by cell transfection could be different to the behaviour of cells that have been exposed to GTF2IRD1 throughout their development.

FGF signalling also plays a critical role in brain development. Gene inactivation studies have shown that the receptors, FGFR1 and FGFR2, are necessary for brain development (Terwisscha van Scheltinga et al., 2013) and proper formation of the medial prefrontal cortex and its connections with the limbic circuits (Stevens et al.,

2010). Since prior studies have pointed to GTF2IRD1 having a role in craniofacial development and behaviour in mice (Enkhmandakh et al., 2009; Howard et al., 2012;

Tassabehji et al., 2005; Young et al., 2008), it is possible that the upregulation of Fgf7 seen in the lip skin tissue is not only of relevance to the epidermal hyperplasia, but could also be part of the abnormalities in the central nervous system that create such behavioural anomalies. However, it is necessary to be cautious with such conclusions as the interplay between FGF signalling and other signalling pathways is complex and the response of these systems to GTF2IRD1 loss is likely to differ from tissue to tissue.

The signalling pathways stimulated by the FGFs are also stimulated by TGF

(transforming growth factor beta), BMPs (the bone morphogenetic proteins) and Wnt ligands (Szebenyi and Fallon, 1999). Several members of these families show decreased expression as a consequence of the lack of Gtf2ird1 in the data described here. Overall, it appears that signalling through these ligands is decreased in the absence of Gtf2ird1.

The TGFβ proteins generally have an anti-proliferative effect in epithelial cells, as well

223

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin as regulating the immune response. Amelioration of this anti-proliferative action is consistent with the hyperproliferation of epithelial cells seen in the Gtf2ird1 knockout phenotype. The Wnt signalling pathway is also involved in the same key biological processes of cellular proliferation, differentiation, adhesion and survival in a range of tissues that are also regulated by the FGF family, and Wnt signalling and also plays an important role in the development of the nervous system (Godoy et al., 2014).

With such a diversity of processes in which these proteins are involved, it is difficult to draw specific conclusions based only on transcript profiles. Nevertheless, such work can provide preliminary clues for subsequent investigation. Our gene ontology analysis of down-regulated genes draws attention to the Wnt-calcium pathway, as well as calcium ion homeostasis. This is interesting as calcium signalling pathways play a key role in keratinocyte proliferation, migration and differentiation (Bikle et al.,

2012), inflammation (Kim et al., 2010) and the nervous system (Flavell and Greenberg,

2008). TFII-I has also been implicated in the regulation of calcium homeostasis as a consequence of a separate role in the cytoplasm (Caraveo et al., 2006). While endogenous GTF2IRD1 has a nuclear localisation pattern (Carmona-Mora et al., 2015) and has, so far, not been shown to have similar calcium regulatory properties, it is possible that this finding could indicate another functional overlap between GTF2IRD1 and TFII-I proteins.

In conclusion, this chapter describes a molecular analysis of lip tissue from

Gtf2ird1-/- mice, capturing effects that are most apparent in the epidermis. However, these data may also provide clues regarding GTF2IRD1 functional principles that are more generally applicable to all of the cell types that express Gtf2ird1. In this context, analysis of dysregulated genes that have roles in skin and brain development may provide insights into molecular factors and pathways that are central to the causality of

224

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

WBS. As a future direction, it would be interesting to perform RNA-seq analysis of affected vs unaffected skin in the Gtf2ird1-/- mice (i.e. lip versus dorsal skin) aiming to identify genes that are ‘universally dysregulated’ in the knockout skin, compared to those that are driving the proliferation and differentiation in the lips. This would also help with comparisons to previously generated data sets from elsewhere and may help to define whether there are a consistent set of genes that are ‘universally dysregulated’ in cells lacking GTF2IRD1. However, these hypotheses are speculative since we are still some way from an understanding of which gene sets are targeted by GTF2IRD1, and it is still unclear how it brings about changes in gene expression.”

225

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Appendix 1

Gene symbol Log2FC PValue FDR Classification Ngf 2.04244617 1.42E-12 9.34E-11 Growth factor Fgfbp1 1.63440504 9.90E-12 6.11E-10 Growth factor Fgf7 1.60151464 1.86E-10 1.04E-08 Growth factor Egfl6 -0.9927542 8.56E-06 2.44E-04 Growth factor Fgfrl1 -1.1457243 2.49E-05 6.25E-04 Growth factor Bmp6 -0.9034426 3.66E-05 8.71E-04 Growth factor Bmp4 -0.9666647 9.15E-05 1.88E-03 Growth factor Ngfr -0.9126221 1.06E-04 2.12E-03 Growth factor Fgf13 -1.1334501 1.15E-04 2.23E-03 Growth factor Fgf1 0.78884875 2.30E-04 3.89E-03 Growth factor Thpo -1.2140716 2.80E-04 4.54E-03 Growth factor Fgfr1 -0.6951985 4.91E-04 7.01E-03 Growth factor Igf2 -1.3125554 7.43E-04 9.66E-03 Growth factor Vegfb -0.8607325 2.11E-03 2.08E-02 Growth factor Gdf10 -0.6898116 7.02E-03 4.73E-02 Growth factor Bmp3 -0.8177133 7.12E-03 4.77E-02 Growth factor Vegfc -0.8269348 7.74E-03 5.06E-02 Growth factor Fgfbp3 -0.8980304 8.06E-03 5.20E-02 Growth factor Bmp2 -0.6059232 8.33E-03 5.32E-02 Growth factor Pdgfrl -0.7408365 1.01E-02 6.02E-02 Growth factor Epor -0.9775301 1.02E-02 6.05E-02 Growth factor Bmpr1a 0.65247964 1.06E-02 6.22E-02 Growth factor Pdgfb -0.6407244 1.07E-02 6.25E-02 Growth factor Pdgfc -0.5821044 1.43E-02 7.64E-02 Growth factor Cxcl10 4.23587605 1.22E-18 1.21E-16 Cytokine Ccl20 4.11384907 1.62E-17 1.53E-15 Cytokine Ccl1 3.36929687 3.56E-15 2.87E-13 Cytokine Il1f9 1.75727149 1.01E-09 5.31E-08 Cytokine Cxcl1 2.72374035 5.54E-09 2.65E-07 Cytokine Ccl2 1.5059188 1.55E-08 7.09E-07 Cytokine Tnfrsf23 1.4392636 3.77E-08 1.66E-06 Cytokine Il1b 2.03067905 6.67E-08 2.85E-06 Cytokine Cxcl9 4.13059328 5.50E-07 2.07E-05 Cytokine Tnfrsf9 1.96626447 2.06E-06 6.93E-05 Cytokine Tnfsf15 1.60756648 2.37E-06 7.87E-05 Cytokine Il7r 1.25825581 4.52E-06 1.41E-04 Cytokine Cxcl16 0.99298836 5.84E-06 1.77E-04 Cytokine Cxcr6 1.22943905 3.38E-05 8.11E-04 Cytokine Ccl6 -0.959006 4.93E-05 1.12E-03 Cytokine Il1f8 1.23857189 7.79E-05 1.64E-03 Cytokine Tgfb2 -0.8592351 1.10E-04 2.16E-03 Cytokine Il1rl2 0.80503792 5.28E-04 7.45E-03 Cytokine Il13ra1 0.76013419 1.37E-03 1.52E-02 Cytokine Cxcl14 -0.8431777 1.55E-03 1.67E-02 Cytokine Il6ra 0.58841808 1.97E-03 1.98E-02 Cytokine Il22ra2 -1.1937422 2.08E-03 2.05E-02 Cytokine Tnfsf18 1.17535404 3.01E-03 2.64E-02 Cytokine Il20rb 0.54429447 3.24E-03 2.78E-02 Cytokine Il17d -0.9598442 3.25E-03 2.78E-02 Cytokine Ccl17 -1.1824489 3.58E-03 2.99E-02 Cytokine

226

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Ccl24 -0.8598089 3.85E-03 3.13E-02 Cytokine Il34 0.61338404 4.01E-03 3.22E-02 Cytokine Tnfaip6 -0.8194121 4.15E-03 3.30E-02 Cytokine Tgfbi -0.5771503 4.33E-03 3.41E-02 Cytokine Tnfsf12 -0.871964 4.53E-03 3.50E-02 Cytokine Il1a 1.1857626 4.85E-03 3.67E-02 Cytokine Il17rd -0.6532555 5.07E-03 3.78E-02 Cytokine Tnfrsf19 0.5798386 6.10E-03 4.30E-02 Cytokine Il18r1 0.88705919 6.67E-03 4.58E-02 Cytokine Tgfbr3 -0.58106 6.90E-03 4.67E-02 Cytokine Tgfb1i1 -0.6240288 7.99E-03 5.18E-02 Cytokine Tnfrsf13c -0.8821486 8.68E-03 5.45E-02 Cytokine Il2rb 0.84003711 1.01E-02 6.01E-02 Cytokine Il20ra -0.8344081 1.17E-02 6.60E-02 Cytokine Il17rc 0.49141021 1.28E-02 7.02E-02 Cytokine Ccl7 0.73115294 1.49E-02 7.85E-02 Cytokine Cxcr2 0.7220673 2.16E-02 9.98E-02 Cytokine Csnk2a1 1.50560391 2.60E-12 1.69E-10 Wnt signalling Daam1 1.67646687 3.65E-12 2.32E-10 Wnt signalling Wnt7b -1.362668 2.34E-09 1.17E-07 Wnt signalling Fosl1 1.37504209 1.52E-07 6.20E-06 Wnt signalling Shh -1.2452433 2.19E-05 5.64E-04 Wnt signalling Cacybp 1.00154013 5.76E-05 1.29E-03 Wnt signalling Lef1 -0.9825974 6.50E-05 1.43E-03 Wnt signalling Wif1 -0.7924428 1.49E-04 2.75E-03 Wnt signalling Fzd9 -1.5096321 2.22E-04 3.79E-03 Wnt signalling Sfrp2 -0.9441217 3.76E-04 5.74E-03 Wnt signalling Fzd1 -0.6860401 5.20E-04 7.35E-03 Wnt signalling Ror2 -0.8089454 7.42E-04 9.66E-03 Wnt signalling Fzd2 -0.8291715 9.22E-04 1.14E-02 Wnt signalling Eda2r 0.91064916 1.23E-03 1.40E-02 Wnt signalling Wnt11 -0.6645598 1.84E-03 1.89E-02 Wnt signalling Lrp4 -0.6464435 1.90E-03 1.93E-02 Wnt signalling Wnt9a -0.7451474 3.11E-03 2.70E-02 Wnt signalling Ppap2b -0.7821088 3.55E-03 2.97E-02 Wnt signalling Sox2 -1.081059 3.82E-03 3.12E-02 Wnt signalling Wnt10b -0.6953804 3.95E-03 3.19E-02 Wnt signalling Prickle1 -0.6444138 4.21E-03 3.33E-02 Wnt signalling Nkd1 -0.7584446 4.37E-03 3.42E-02 Wnt signalling Grb10 -0.6459085 5.27E-03 3.89E-02 Wnt signalling Fzd5 -0.5765638 5.50E-03 4.00E-02 Wnt signalling Foxl1 -1.0258477 5.73E-03 4.13E-02 Wnt signalling Wnt6 -0.748321 5.83E-03 4.17E-02 Wnt signalling Bambi -0.5493759 6.07E-03 4.28E-02 Wnt signalling Bmp2 -0.6059232 8.33E-03 5.32E-02 Wnt signalling Prkcdbp -0.6551662 8.60E-03 5.42E-02 Wnt signalling Wnt3 -0.6383288 9.58E-03 5.82E-02 Wnt signalling Ccnd3 -0.5359914 9.93E-03 5.95E-02 Wnt signalling Prkaa2 -0.5241587 1.19E-02 6.72E-02 Wnt signalling Frzb -0.7766848 1.56E-02 8.12E-02 Wnt signalling Ppp3cc -0.546913 1.85E-02 9.01E-02 Wnt signalling Chrm3 4.66534091 2.84E-53 1.77E-50 Ca signalling Ryr3 3.20879302 1.82E-17 1.69E-15 Ca signalling Cacng1 -1.8645298 6.02E-12 3.75E-10 Ca signalling

227

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Csrp3 -1.2140997 1.17E-07 4.83E-06 Ca signalling Camk2a -1.2768084 1.47E-05 3.90E-04 Ca signalling P2rx7 1.05240501 2.26E-05 5.81E-04 Ca signalling Orai1 0.92088609 2.28E-05 5.84E-04 Ca signalling Ctgf -1.0091878 3.33E-05 8.01E-04 Ca signalling Cacnb1 -1.0611512 4.31E-05 1.01E-03 Ca signalling Bdkrb2 1.16407191 4.34E-05 1.01E-03 Ca signalling Camk2b -1.1237764 7.06E-05 1.52E-03 Ca signalling Kcna5 -1.5126091 7.32E-05 1.56E-03 Ca signalling Fxyd1 -1.0374407 1.15E-04 2.23E-03 Ca signalling Oxtr -1.2730261 1.29E-04 2.46E-03 Ca signalling Sypl2 -1.2282089 1.44E-04 2.68E-03 Ca signalling Cav3 -1.1022097 1.74E-04 3.11E-03 Ca signalling Bdkrb1 -1.3944395 2.12E-04 3.65E-03 Ca signalling Slc8a3 -1.0003288 2.33E-04 3.94E-03 Ca signalling Mylk2 -1.0168634 2.50E-04 4.16E-03 Ca signalling Cacng6 -1.1493415 2.72E-04 4.45E-03 Ca signalling Tmem37 -1.4512849 3.99E-04 5.99E-03 Ca signalling Cacna1s -0.9895246 4.90E-04 7.01E-03 Ca signalling Mylk4 -1.2861786 5.50E-04 7.67E-03 Ca signalling Nos1 -1.1880142 7.83E-04 1.01E-02 Ca signalling Lpar1 -0.7679122 8.07E-04 1.03E-02 Ca signalling Sphk1 -0.6684863 8.33E-04 1.06E-02 Ca signalling Pygm -1.098575 8.51E-04 1.07E-02 Ca signalling Fzd2 -0.8291715 9.22E-04 1.14E-02 Ca signalling Itpr2 0.74672362 1.15E-03 1.34E-02 Ca signalling Phkg1 -0.7893236 1.26E-03 1.43E-02 Ca signalling Agtr1a 1.17742936 1.27E-03 1.44E-02 Ca signalling Cacna1e 1.10804501 1.28E-03 1.44E-02 Ca signalling Adrb3 1.13310805 1.57E-03 1.69E-02 Ca signalling Jph2 -1.0775662 1.64E-03 1.75E-02 Ca signalling Slc25a4 -0.8732573 2.49E-03 2.32E-02 Ca signalling Ryr1 -1.010025 2.58E-03 2.38E-02 Ca signalling Pde1b -0.7690563 2.82E-03 2.54E-02 Ca signalling Ndn -0.8900525 3.09E-03 2.69E-02 Ca signalling Ednra -0.6375484 3.31E-03 2.81E-02 Ca signalling Phka1 -0.5810431 4.23E-03 3.35E-02 Ca signalling Adora2b -0.7242631 4.63E-03 3.55E-02 Ca signalling Trpv4 -0.6182011 4.65E-03 3.55E-02 Ca signalling Grin2d -0.7583487 5.11E-03 3.80E-02 Ca signalling Slc8a2 -0.9957636 5.15E-03 3.83E-02 Ca signalling Cacng7 -0.8805955 5.34E-03 3.93E-02 Ca signalling Itgb3 -0.7281304 5.36E-03 3.94E-02 Ca signalling Atp2a1 -1.2209544 5.63E-03 4.07E-02 Ca signalling Psen2 -0.6822826 6.55E-03 4.51E-02 Ca signalling Ptger1 -0.6929746 6.68E-03 4.58E-02 Ca signalling Prkcb 0.71925813 6.83E-03 4.65E-02 Ca signalling Stim1 -0.6417681 7.34E-03 4.89E-02 Ca signalling Pln -0.9243341 7.53E-03 4.98E-02 Ca signalling Cacna1h -0.8205379 7.75E-03 5.07E-02 Ca signalling Nos2 -1.0048721 7.79E-03 5.09E-02 Ca signalling Cacna1a -0.6041301 9.26E-03 5.68E-02 Ca signalling Phkb 0.63442835 9.35E-03 5.71E-02 Ca signalling Tpcn1 -0.5589493 9.65E-03 5.85E-02 Ca signalling

228

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Epor -0.9775301 1.02E-02 6.05E-02 Ca signalling Adcy2 -0.6749613 1.03E-02 6.11E-02 Ca signalling Chrna10 -0.8917698 1.05E-02 6.19E-02 Ca signalling Camk4 0.99701596 1.06E-02 6.22E-02 Ca signalling P2rx1 -0.8716937 1.40E-02 7.49E-02 Ca signalling Prkx 0.49201613 1.44E-02 7.66E-02 Ca signalling Ptk2b 0.46603931 1.50E-02 7.86E-02 Ca signalling Calm4 0.96928226 1.53E-02 7.98E-02 Ca signalling Pde1c 0.81990041 1.54E-02 8.03E-02 Ca signalling Myb -0.6053903 1.60E-02 8.24E-02 Ca signalling Ppp3cc -0.546913 1.85E-02 9.01E-02 Ca signalling Arhgap8 3.82640145 4.35E-42 1.55E-39 Cell cycle Ccnb1 1.16075618 6.61E-08 2.83E-06 Cell cycle Rbl1 0.95995552 1.63E-06 5.61E-05 Cell cycle Anln 1.11969571 3.15E-06 1.02E-04 Cell cycle Ckap2 1.04416898 4.66E-06 1.44E-04 Cell cycle Prc1 1.03556605 6.13E-06 1.84E-04 Cell cycle Dlgap5 1.12022521 6.93E-06 2.04E-04 Cell cycle Cep55 1.23635764 7.73E-06 2.23E-04 Cell cycle Avpi1 0.86865638 2.74E-05 6.79E-04 Cell cycle Cenpe 1.45480033 2.93E-05 7.20E-04 Cell cycle Mki67 1.28167016 2.95E-05 7.21E-04 Cell cycle Mapk6 0.91213077 4.45E-05 1.03E-03 Cell cycle Smc2 1.20676032 7.28E-05 1.56E-03 Cell cycle Dst 1.37170517 7.33E-05 1.56E-03 Cell cycle Tpx2 0.9334239 7.36E-05 1.57E-03 Cell cycle Smc4 1.17187024 7.45E-05 1.58E-03 Cell cycle Ereg 1.30428704 9.05E-05 1.86E-03 Cell cycle Kif11 1.13178944 9.39E-05 1.92E-03 Cell cycle Mis12 0.99954136 1.02E-04 2.06E-03 Cell cycle Mlh3 1.1201497 1.14E-04 2.22E-03 Cell cycle Spag5 0.84090787 1.78E-04 3.17E-03 Cell cycle Nuf2 1.04788845 2.25E-04 3.82E-03 Cell cycle Nusap1 0.95758546 2.87E-04 4.61E-03 Cell cycle Aspm 1.14968788 4.66E-04 6.76E-03 Cell cycle Xrn1 0.86861451 5.08E-04 7.21E-03 Cell cycle Ccar1 0.85217319 5.69E-04 7.88E-03 Cell cycle Smc3 0.945498 6.33E-04 8.53E-03 Cell cycle Ndc80 1.06517496 8.09E-04 1.03E-02 Cell cycle 4632434I11Rik 1.11740469 8.26E-04 1.05E-02 Cell cycle Dnajc2 0.8866995 8.39E-04 1.06E-02 Cell cycle Ckap5 0.87792394 8.77E-04 1.10E-02 Cell cycle Sgol1 0.89338107 8.85E-04 1.10E-02 Cell cycle Bub1 1.16346817 9.68E-04 1.17E-02 Cell cycle Nup37 0.77788561 1.06E-03 1.26E-02 Cell cycle Ncapg2 0.94579233 1.18E-03 1.36E-02 Cell cycle Taf1 0.90014096 1.18E-03 1.36E-02 Cell cycle Katna1 0.74269877 1.35E-03 1.51E-02 Cell cycle Kntc1 0.81126276 1.88E-03 1.91E-02 Cell cycle Top1 0.68864937 2.03E-03 2.03E-02 Cell cycle Dstn 0.82783607 2.25E-03 2.17E-02 Cell cycle Tgs1 0.87433021 2.28E-03 2.18E-02 Cell cycle Dusp1 0.78669849 2.80E-03 2.53E-02 Cell cycle Rala 0.5806418 2.85E-03 2.55E-02 Cell cycle

229

Chapter 5 – Gene regulation associated with GTF2IRD1 in lip skin

Sgol2 1.10087999 2.92E-03 2.58E-02 Cell cycle Cltc 0.79176252 3.27E-03 2.79E-02 Cell cycle Dbf4 0.6801419 3.33E-03 2.82E-02 Cell cycle Nek2 0.67150321 3.81E-03 3.12E-02 Cell cycle Appl1 0.78441619 4.12E-03 3.29E-02 Cell cycle 4922501C03Rik 0.77001278 4.45E-03 3.47E-02 Cell cycle Esco1 0.86867397 4.68E-03 3.57E-02 Cell cycle Rad50 0.7607609 4.95E-03 3.71E-02 Cell cycle Ahctf1 0.76667494 5.45E-03 3.97E-02 Cell cycle Fanci 0.84208649 5.69E-03 4.11E-02 Cell cycle Brca1 0.75542551 5.89E-03 4.19E-02 Cell cycle Esco2 0.85145725 5.92E-03 4.20E-02 Cell cycle Tacc3 0.56872875 6.28E-03 4.37E-02 Cell cycle Mnat1 0.65186098 6.71E-03 4.59E-02 Cell cycle Dynlt3 0.72442947 6.78E-03 4.63E-02 Cell cycle Kctd11 0.58346685 7.56E-03 4.99E-02 Cell cycle Nipbl 0.77706818 8.05E-03 5.20E-02 Cell cycle Topbp1 0.68734732 8.24E-03 5.29E-02 Cell cycle Zwint 0.65358873 8.62E-03 5.43E-02 Cell cycle Cenpj 0.85016049 8.71E-03 5.46E-02 Cell cycle Cep110 0.71610875 9.15E-03 5.64E-02 Cell cycle Fancd2 0.70719861 1.04E-02 6.14E-02 Cell cycle Ncapd3 0.59404221 1.06E-02 6.22E-02 Cell cycle Terf1 0.6075518 1.07E-02 6.26E-02 Cell cycle Anxa1 0.8648015 1.08E-02 6.30E-02 Cell cycle Nbn 0.63520162 1.10E-02 6.37E-02 Cell cycle Cdkn3 0.72305338 1.17E-02 6.61E-02 Cell cycle Smpd3 0.65313323 1.26E-02 6.97E-02 Cell cycle Clspn 0.65096007 1.48E-02 7.79E-02 Cell cycle Racgap1 0.55655298 1.53E-02 7.98E-02 Cell cycle Ccna2 0.48153356 1.77E-02 8.78E-02 Cell cycle Usp16 0.66309667 1.80E-02 8.86E-02 Cell cycle Pes1 0.47860617 1.84E-02 9.01E-02 Cell cycle Nudc 0.44643652 2.01E-02 9.51E-02 Cell cycle Lig4 0.61571292 2.04E-02 9.63E-02 Cell cycle Spc25 0.62014135 2.09E-02 9.79E-02 Cell cycle Macf1 0.69721263 2.33E-02 1.05E-01 Cell cycle

230

CHAPTER 6 – OVERVIEW AND CONCLUDING REMARKS

231

Chapter 6 – Overview and concluding remarks

6.1 Overview

Up to 28 genes are potentially associated with the development of the WBS features (Bayes et al., 2003; Pober, 2010) but not all of them will play a role in the pathology, because not all of the proteins they encode are expected to be dosage sensitive. In an effort to understand the causes of the disease, ongoing studies continue to narrow down the individual genetic contributions towards each specific feature of the syndrome. However, to date, only ELN has unequivocally been related to the causation of SVAS and other elastin-associated defects (Curran et al., 1993; Ewart et al., 1993;

Ewart et al., 1994).

Genotype-phenotype correlations in patients who carry atypical deletions of the

WBS critical region, have provided evidence affirming that GTF2I and GTF2IRD1 play a role in the features of WBS (Antonell et al., 2010; Dai et al., 2009; Edelmann et al.,

2007; Ferrero et al., 2010; Fusco et al., 2014; Hirota et al., 2003; Tassabehji et al.,

2005). The gene products share sequence and functional homology, they show all the hallmarks of being dosage sensitive and animal knockout models support a role in the pathology of the syndrome. However, there are several limitations that need to be taken into account when interpreting genotype-phenotype correlation data. In particular, the precise location of the breakpoints, the effect that the deletion has on the expression of non-deleted genes, non-genetic sources of variation that create inconsistent phenotypes and the effects of age and inaccurate phenotype assessment. These complexities can cloud the mapping of phenotypes to specific genes.

Setting these limitations aside, the human genetic mapping data suggest that

GTF2I and GTF2IRD1 both contribute to the craniofacial features, mental retardation, hypersociability and the visuospatial construction deficits of WBS (Antonell et al 2010).

232

Chapter 6 – Overview and concluding remarks

Despite the overlapping sequence homology and functional similarity, analyses of the phenotypes that are observed in mouse lines with monogenic mutations of the orthologous genes, suggest that the functions of these two related proteins are not redundant.

The main aim of the work described in this thesis was to drill deeper into the function of GTF2IRD1, by investigating the impact of a targeted Gtf2ird1 deletion on the neurological and craniofacial phenotypes in mice. It was hoped that this work would contribute to a greater understanding of the cellular and molecular basis of WBS causality. The key findings from this work and future directions are discussed below.

6.1.1 GTF2IRD1 has a role in the auditory system

The work described in chapter 3 is the first study that relates GTF2IRD1 with the auditory pathogenesis of WBS and resulted in a report published in the European

Journal of Human Genetics (Canales et al., 2015).

The idea of studying the hearing capacity of Gtf2ird1-/- mice was built on a hypothesis that aimed to dissect the in vivo function of GTF2IRD1 and was based on two preliminary findings. Firstly, analysis of Gtf2ird1tm2(LacZ)Hrd in mice showed that the gene is very active in the spiral ganglion neurons of the inner ear during development and adulthood (Palmer et al., 2007). Secondly, unpublished acoustic startle behavioural studies suggested that subtle hearing abnormalities may be present that sophisticated hearing tests would be capable of unpicking. The attractive feature of hearing as a model system to test GTF2IRD1 function is that it involves sensory and neuronal components that can be probed and measured very accurately without major intervention and without the potential confounds associated with behavioural studies.

233

Chapter 6 – Overview and concluding remarks

Results from behavioural acoustic startle tests did not show a statistically significant difference between genotypes, but a trend towards higher auditory thresholds in the Gtf2ird1-/- mice was observed across the range of sound levels tested (figure 3.3).

However, this analysis was done as part of a large battery of behavioural tests that were conducted in order to discover the neurological phenotypes associated with the inactivation of Gtf2ird1, which included evidence of increased responses to stressful stimuli, such as increased vocalizations and increased serum corticosterone levels in response to the FST. These findings also impact on the readout from the acoustic startle test; if Gtf2ird1-/- mice respond differently to stress-inducing tests, then behavioural hearing tests are an inadequate form of analysis because the measure is startle amplitude, which also has an emotional processing and anxiety control component.

Hence, the only way to evaluate hearing capacity properly was to perform physiological hearing tests such as DPOAE and ABR assays that are performed in anesthetised mice where no behavioural component could affect the interpretation of the results.

ABR analysis demonstrated that Gtf2ird1-/- mice have a mild hypoacusis, which is evident in the elevated auditory threshold (Canales et al., 2015). The analysis of the

ABR waveforms found no afferent conduction defect that might indicate a neural impairment of the cochlear nerve. The analysis of the DPOAE showed a significant increase in thresholds, suggesting that there is a defect in the cochlear amplifier associated with the OHC and the extent of the difference was sufficient to explain the elevated thresholds found in the ABR. However, expression of Gtf2ird1 was detected in a variety of cell types within the cochlear compartment (i.e. marginal cells of the stria vascularis, Reissner’s membrane, pillar cells, etc), suggesting that these cells could also contribute to the impairment of the cochlear amplifier through a disturbance of the ionic gradient controlled by some of these cell types.

234

Chapter 6 – Overview and concluding remarks

In summary, the auditory defect detected in these mice bears a strong similarity to that observed in the WBS patients, but there is no evidence in our mice of a mechanism that might explain the auditory allodynia in WBS patients. We postulate that this process forms part of a different mechanism in which the central auditory perception at the level of the auditory cortex may be compromised. Nevertheless, the principal deficit described in these mice can be traced to impairments in the amplification process mediated by the OHCs and similar mechanisms may underpin the

SNHL experienced by WBS patients.

Analysis of GTF2IRD1 function was not pursued further with this model system since the expression of the gene was found to be relatively broad within the inner ear and the auditory deficit was clearly mediated by a different cell type (OHCs) than originally envisaged (spiral ganglion neurons). Analysis of the OHCs could provide a system to further dissect the cellular and molecular basis of the hearing deficit, but, as discussed, it is not yet established whether it is the cells themselves that are impaired, or whether there are defects within the environment in which they operate. There are also major experimental difficulties to do with the inaccessibility of the cells and the small quantity of the material from which samples of RNA and protein would need to be extracted.

6.1.2 GTF2IRD1 has a role in skin control mechanisms

There are many diseases that affect the development and differentiation of the epidermis and its accessory structures; hair, teeth, nails and sweat glands (Lane and

McLean, 2004). Analysis of the genetic basis of these diseases is the primary source of insight into the molecular mechanisms of epidermal development and differentiation.

Mutations created by targeting specific genes in the mouse that lead to epidermal

235

Chapter 6 – Overview and concluding remarks

defects also provide a valuable source of such information (DiTommaso et al., 2014;

Smyth et al., 2008).

Previous reports have shown that mice lacking Gtf2ird1 present with craniofacial abnormalities (Durkin et al., 2001; Enkhmandakh et al., 2009; Tassabehji et al., 2005). There are some disadvantages of this mouse model that have been discussed previously (sections 1.1.6 and 4.3), but differences between specific mouse models could also be influenced by genetic background and the precise nature of the mutation.

Analysis of the Gtf2ird1-/- model used in this work indicates that processes such as cellular proliferation and differentiation of the facial epidermis are partly controlled by

GTF2IRD1. This role first became evident through the histological analysis of face sections taken from adult Gtf2ird1-/- mice (Howard et al., 2012). Here, this phenotype has been investigated to a deeper level and the cellular processes that are compromised are explored using cellular and molecular tools.

Gtf2ird1-/- mice showed increased cell proliferation in the basal layer of the skin, which is the stem cell-containing compartment. Expression analysis showed that

Gtf2ird1 is activated in the basal layer of the developing skin in restricted areas of the face that match the zones where the overgrowth phenotypes are observed in adult life.

The prime example of such a region is the lips, where we focussed most of our attention due to the presence of a very obvious phenotype. RNA-Seq analysis using RNA isolated from lip tissue identified a considerable number of differentially expressed genes that are associated with keratinocyte cell biology. However, the Gtfird1 transcript was almost undetectable in the adult skin. It is not clear why there is such a big difference in the pattern and level of expression of Gtf2ird1 at different stages of life. It is also unclear how the expression of Gtf2ird1 in the skin during embryonic life, creates a

236

Chapter 6 – Overview and concluding remarks

persistent defect in that epidermal zone for the rest of the animal’s life, even though

Gtf2ird1 is no longer detectable and the skin cells renew themselves at a very rapid rate.

Recent work from our laboratory has identified several GTF2IRD1 interacting partners in vitro; many of which are involved in chromatin regulation, suggesting that

GTF2IRD1 plays a role in chromatin modifications, leading to the conclusion that it acts as a regulator of epigenetic mechanisms (Carmona-Mora et al., 2015). In the context of the soft tissue overgrowth observed in the face of Gtf2ird1-/- mice, we hypothesise that GTF2IRD1 is mediating processes that create permanent epigenetic marks in the stem cell population of the basal epidermis during development, which continue to control the behaviour of the keratinocytes that are born from this progenitor population. In the absence of Gtf2ird1, these mechanisms fail to operate and this layer of control is lost for the remainder of the animal’s life. This hypothesis makes certain assumptions that need to be confirmed. Sections of lip skin showing increased proliferation should be analysed using a marker for stem cells, such as P63, in order to determine whether P63 positive cells are also more active.

Gtf2ird1-/- mice were also found to have a skin barrier defect in the regions of the face where the overgrowth occurs and the increased proliferation is observed. Skin thickening is a common response when the epidermis is exposed to environmental insults, but it also usually found in models where the skin is shown to have altered cell proliferation and differentiation processes (Brennan et al., 2007; De Fine Olivarius et al., 1993; DiTommaso et al., 2014; Elias and Menon, 1991; Grether-Beck et al., 2012;

Koster and Roop, 2007; Lessard et al., 2013; Smyth et al., 2008). Whether the lip skin thickening of Gtf2ird1-/- mice is a primary defect that is directly due to the lack of

Gtf2ird1, or a secondary response to the environmental insults that are incurred due to defects in the skin barrier, remains to be elucidated. However, analysis of early stages of

237

Chapter 6 – Overview and concluding remarks

mouse development indicate that the thickening is detectable from the very earliest stages and continues throughout life, which reduces the likelihood that this defect is caused by an environmental insult.

Other than the thickened skin, Gtf2ird1-/- mice do not exhibit any visible signs of skin disruption. We hypothesise that within the affected skin zones, terminal differentiation of the keratinocytes is incomplete, which leaves the properties of the skin at this site similar to the oral mucosa. This would explain the increased TEWL in the lips and would result in a lack of clear distinction between the lip epidermis and the oral mucosa that defines the mucocutaneous junction of the mouth. This idea is supported by the fact that Gtf2ird1 expression has restricted boundaries of expression in this and other mucocutaneous junctions (i.e. the developing eyelids and the anorectal-junction).

An analysis of the more external layers of the skin in affected zones, examining the levels of proteins that are important for the formation of the skin barrier (e.g. filaggrin and loricrin) would be an important next step for validation of this hypothesis.

6.1.3 Gene dysregulation in lip skin due to the lack of Gtf2ird1

The work presented in Chapter 5 described the consequences of Gtf2ird1 inactivation to the transcriptome of lip tissue using RNA-Seq analysis. A large number of up-regulated and down-regulated genes were found, revealing that the absence of

Gtf2ird1 has widespread effects. The functional analysis indicated that the changes in gene expression impacted on categories of genes including; structural/morphological development of tissue, cellular differentiation, cellular proliferation and the immune response. All of these processes are regulated by a variety of signalling pathways, some of which were also dysregulated. This molecular analysis correlated very well with the

238

Chapter 6 – Overview and concluding remarks

increased cell proliferation and the enlarged layers of differentiating keratinocytes discovered through the immunohistological methods. The differential expression of genes encoding keratins and keratin-associated proteins was consistent with the observed alterations in differentiation and the abnormality of the skin barrier function.

Both sets of findings support a model in which the final stages of keratinocyte maturation in the granular layer is impaired.

Fgf7 and Fgfbp1, were found to be significantly upregulated in Gtf2ird1-/- mice.

Fgfbp1 is found in keratinocytes throughout the normal epidermis, and its gene product is known to interact with FGF7, stimulating its activity in the skin, including its effect on the proliferation of keratinocytes. The presence of FGF7 also inhibits hair morphogenesis and epidermal differentiation is promoted at the expense of hair follicle fate (Richardson et al., 2009). FGF7 upregulation was also confirmed at the protein level. In fact, FGF7 was only detectable in the Gtf2ird1-/- mice and not in the wild type control littermates. This result suggests that GTF2IRD1 is important for the regulation of FGF7 signalling pathways in the skin. This hypothesis is in agreement with a concomitant downregulation of proteins associated with hair follicle development in the

Gtf2ird1-/- mice and is substantiated by a report in which transgenic FGF7, under the control of the K14 promoter, produced a similar phenotype to that found in Gtf2ird1-/- mice (Guo et al., 1993). However, part of this model is in disagreement with a previous report that found Fgf7 was up-regulated as a consequence of Gtf2ird1 overexpression in mouse embryonic fibroblasts (Chimge et al., 2007). The potential technical flaws of such in vitro studies were discussed previously (Section 5.3) and the inconsistency in gene regulation may also be indicative of cell context-dependant variables.

Our RNA-Seq analysis also identified a set of DEGs that may have an important impact on epithelial development and maintenance, but may also have a broader role in

239

Chapter 6 – Overview and concluding remarks

brain function. These data could provide important clues that may link GTF2IRD1 function with common cellular mechanisms in a variety of cell types that may help to explain the molecular role of GTF2IRD1 in the features of WBS. However, much of the basic information regarding GTF2IRD1 function remains unknown.

While the RNA-Seq identified a considerable set of potential gene targets in lip tissue, the next step would be to determine whether these targets are direct or indirect targets using chromatin immunoprecipitation (ChiP) in ChiP-PCR and ChIP-seq assays.

The use of cultured primary keratinocytes derived from lip tissue from wild type and

Gtf2ird1-/- foetal/post-natal pups could provide an excellent in vitro model system to address this question. However, an antibody that detects mouse GTF2IRD1 in a manner that is adequate to perform such a study is currently not available.

Based on the recent knowledge gained from in vitro studies in our laboratory regarding the protein partners of GTF2IRD1, it is predicted that this protein plays a role in chromatin regulation (Carmona-Mora et al., 2015). It is predicted, therefore, that the foetal/post-natal expression of GTF2IRD1 in the stem cells of the epidermis normally results in the establishment of epigenetic marks, such as DNA methylation, that regulate ongoing gene expression patterns in the daughter cells derived from these progenitors.

On this basis, analysis of DNA methylation patterns in cultured Gtf2ird1-/- and wild type keratinocytes and cross-correlations of these data with RNA-Seq and ChIP-seq data would provide profound new insights into the cellular function of GTF2IRD1 and its role in the pathogenesis of WBS.

It is known, however, that many epigenetic regulatory processes are active in the stem cell-containing basal layer of the epidermis, such as the marks left by the polycomb group 1 and 2 complexes, DNA methylation, and histone H4K20

240

Chapter 6 – Overview and concluding remarks

monomethylation (Driskell et al., 2012; Ezhkova et al., 2009; Ezhkova et al., 2011; Luis et al., 2011; Mejetta et al., 2011; Sen et al., 2008; Sen et al., 2010). Therefore, we must be careful with reductionist approaches in trying to pinpoint the function of individual components in cellular systems. Advances in genomics have enabled large-scale parallel interrogation of these elements and it will be important to take such approaches into the next phase of GTF2IRD1 research.

6.3 Concluding remarks

Mouse models are useful tools for the study of human disease. A mouse model that mimics the common deletion found in WBS shows a similar range of phenotypes to the human disease (Segura-Puimedon et al., 2014). The existence of atypical deletions in the WBS region and the analysis of the resulting phenotypes provides the means to dissect the role of individual genes. However, this type of analysis eventually reaches a limit of resolution due to the small numbers of atypical patients and the minimal number of discriminating deletion breakpoints. At some point it is necessary to use monogenic mouse mutations in order to progress to an understanding of gene function and genetic contribution to WBS. The difficulty with such work is that WBS is a contiguous gene deletion disease and the extent of each phenotype may be exaggerated by the haploinsufficiency of multiple gene products. Reducing the mutations to single genes may reduce the extent of such phenoytpes, thus making the analysis of hemizygous mutants inadequate. This seems to be true for several genes that fall within the WBS region, including GTF2IRD1. In the end, it is necessary to build an understanding based on multiple lines of evidence.

241

Chapter 6 – Overview and concluding remarks

In the setting of our current knowledge of GTF2IRD1 function, the findings presented in this work suggest that this protein operates in a cell context-dependant fashion. However, at some stage in the future, it may be possible to look back on such data sets and see the commonalities between the behaviour of GTF2IRD1 in each of these different tissues. The haploinsufficiency of GTF2IRD1 alone does not seem sufficient to explain all of the craniofacial and neurological phenotypes of WBS and evidence indicate that heterozygous Gtf2ird1+/- mice generally lack any major defects.

GTF2IRD1 and GTF2I (TFII-I) are believed to have very similar molecular functions, but whether one gene has more effect than the other or whether the effects are additive or synergistic remains to be determined. The combined hemizygous loss of GTF2IRD1 and GTF2I may enhance the severity of the phenotypes they generate to more than just an additive degree, and the reduction in the levels of other WBS gene products may also combine to form a much more exaggerated whole. Despite these potential drawbacks in analysis of gene function, work to date indicates that GTF2IRD1 plays an important role in the pathogenesis of WBS but does not rule out the contribution made by other genes within the deletion region.

242

References

REFERENCES

Adamo, A., et al., 2015. 7q11.23 dosage-dependent dysregulation in human pluripotent stem cells affects transcriptional programs in disease-relevant lineages. Nat Genet. 47, 132- 41. Ahmad, M.S., et al., 2015. A new gender-specific model for skin autofluorescence risk stratification. Sci Rep. 5, 10198. Alam, H., et al., 2011. Novel function of keratins 5 and 14 in proliferation and differentiation of stratified epithelial cells. Mol Biol Cell. 22, 4068-78. Albers, J., et al., 2011. Control of bone formation by the serpentine receptor Frizzled-9. J Cell Biol. 192, 1057-72. Anders, S., Huber, W., 2010. Differential expression analysis for sequence count data. Genome Biol. 11, R106. Anders, S., Reyes, A., Huber, W., 2012. Detecting differential usage of exons from RNA-seq data. Genome Res. 22, 2008-17. Antonell, A., et al., 2010. Partial 7q11.23 deletions further implicate GTF2I and GTF2IRD1 as the main genes responsible for the Williams-Beuren syndrome neurocognitive profile. J Med Genet. 47, 312-20. Arumugam, T., et al., 2014. Preliminary evaluation of 1'-[(18)F]fluoroethyl-beta-D-lactose ([(18)F]FEL) for detection of pancreatic cancer in nude mouse orthotopic xenografts. Nucl Med Biol. 41, 833-40. Ashe, A., et al., 2008. A genome-wide screen for modifiers of transgene variegation identifies genes with critical roles in development. Genome Biol. 9, R182. Ashmore, J., Gale, J., 2004. The cochlear amplifier. Curr Biol. 14, R403-4. Axelsson, S., 2005. Variability of the cranial and dental phenotype in Williams syndrome. Swed Dent J Suppl. 3-67. Axelsson, S., et al., 2005. Neurocranial morphology and growth in Williams syndrome. Eur J Orthod. 27, 32-47. Barozzi, S., et al., 2012. Audiological findings in Williams syndrome: a study of 69 patients. Am J Med Genet A. 158a, 759-71. Barozzi, S., et al., 2013. Audiological follow-up of 24 patients affected by Williams syndrome. Eur J Med Genet. 56, 490-6. Bayarsaihan, D., Ruddle, F.H., 2000. Isolation and characterization of BEN, a member of the TFII-I family of DNA-binding proteins containing distinct helix-loop-helix domains. Proc Natl Acad Sci U S A. 97, 7342-7. Bayarsaihan, D., et al., 2002. Genomic organization of the genes Gtf2ird1, Gtf2i, and Ncf1 at the mouse chromosome 5 region syntenic to the human chromosome 7q11.23 Williams syndrome critical region. Genomics. 79, 137-43. Bayes, M., et al., 2003. Mutational mechanisms of Williams-Beuren syndrome deletions. Am J Hum Genet. 73, 131-51. Beenken, A., Mohammadi, M., 2009. The FGF family: biology, pathophysiology and therapy. Nat Rev Drug Discov. 8, 235-53. Beer, H.D., et al., 2005. The fibroblast growth factor binding protein is a novel interaction partner of FGF-7, FGF-10 and FGF-22 and regulates FGF activity: implications for epithelial repair. Oncogene. 24, 5269-77. Bellugi, U., et al., 1990. Neuropsychological, neurological, and neuroanatomical profile of Williams syndrome. Am J Med Genet Suppl. 6, 115-25. Bellugi, U., et al., 1999a. Towards the neural basis for hypersociability in a genetic syndrome. Neuroreport. 10, 1653-7.

243

References

Bellugi, U., et al., 1999b. Bridging cognition, the brain and molecular genetics: evidence from Williams syndrome. Trends Neurosci. 22, 197-207. Bellugi, U., et al., 2000. I. The neurocognitive profile of Williams Syndrome: a complex pattern of strengths and weaknesses. J Cogn Neurosci. 12 Suppl 1, 7-29. Benitah, S.A., Frye, M., 2012. Stem cells in ectodermal development. J Mol Med (Berl). 90, 783- 90. Benjamini, Y., et al., 2001. Controlling the false discovery rate in behavior genetics research. Behav Brain Res. 125, 279-84. Benveniste, H., Blackband, S.J., 2006. Translational neuroscience and magnetic-resonance microscopy. Lancet Neurol. 5, 536-44. Berg, J.S., et al., 2007. Speech delay and autism spectrum behaviors are frequently associated with duplication of the 7q11.23 Williams-Beuren syndrome region. Genet Med. 9, 427- 41. Berglund, A.M., Ryugo, D.K., 1991. Neurofilament antibodies and spiral ganglion neurons of the mammalian cochlea. J Comp Neurol. 306, 393-408. Bermingham, N.A., et al., 1999. Math1: an essential gene for the generation of inner ear hair cells. Science. 284, 1837-41. Beunders, G., et al., 2010. A triplication of the Williams-Beuren syndrome region in a patient with mental retardation, a severe expressive language delay, behavioural problems and dysmorphisms. J Med Genet. 47, 271-5. Beuren, A.J., Apitz, J., Harmjanz, D., 1962. Supravalvular aortic stenosis in association with mental retardation and a certain facial appearance. Circulation. 26, 1235-40. Beuren, A.J., 1964. ["A NEW SYNDROME: SUPRAVALVULAR AORTIC STENOSIS, MULTIPLE PERIPHERAL PULMONARY STENOSIS, MENTAL RETARDATION, SIMILAR FACIAL FEATURES AND IDENTICAL DENTAL ABNORMALITIES"]. Monatsschr Kinderheilkd. 112, 218-21. Beuren, A.J., et al., 1964. THE SYNDROME OF SUPRAVALVULAR AORTIC STENOSIS, PERIPHERAL PULMONARY STENOSIS, MENTAL RETARDATION AND SIMILAR FACIAL APPEARANCE. Am J Cardiol. 13, 471-83. Beuren, A.J., 1972. [Assessment of pulmonary hypertension in congenital heart defects in childhood]. Klin Padiatr. 184, 1-3. Bhattacharjee, Y., 2005. Friendly faces and unusual minds. Science. 310, 802-4. Bikle, D.D., Xie, Z., Tu, C.L., 2012. Calcium regulation of keratinocyte differentiation. Expert Rev Endocrinol Metab. 7, 461-472. Birk, E., et al., 2010. SOBP is mutated in syndromic and nonsyndromic intellectual disability and is highly expressed in the brain limbic system. Am J Hum Genet. 87, 694-700. Blank, I.H., 1965. Cutaneous barriers. J Invest Dermatol. 45, 249-56. Blanpain, C., Fuchs, E., 2009. Epidermal homeostasis: a balancing act of stem cells in the skin. Nat Rev Mol Cell Biol. 10, 207-17. Blomberg, S., Rosander, M., Andersson, G., 2006. Fears, hyperacusis and musicality in Williams syndrome. Res Dev Disabil. 27, 668-80. Botta, A., et al., 1999. Detection of an atypical 7q11.23 deletion in Williams syndrome patients which does not include the STX1A and FZD3 genes. J Med Genet. 36, 478-80. Brennan, D., et al., 2007. Suprabasal Dsg2 expression in transgenic mouse skin confers a hyperproliferative and apoptosis-resistant phenotype to keratinocytes. J Cell Sci. 120, 758-71. Brodal, A., 1981. Neurological Anatomy in Relation to Clinical Medicine, Vol., Oxford University Press. Broder, K., et al., 1999. Elevated ambulatory blood pressure in 20 subjects with Williams syndrome. Am J Med Genet. 83, 356-60.

244

References

Brownell, W.E., 1990. Outer hair cell electromotility and otoacoustic emissions. Ear Hear. 11, 82-92. Burn, J., 1986. Williams syndrome. J Med Genet. 23, 389-95. Burrows, E.L., et al., 2015. Environmental Enrichment Ameliorates Behavioral Impairments Modeling Schizophrenia in Mice Lacking Metabotropic Glutamate Receptor 5. 40, 1947-56. Cadigan, K.M., Nusse, R., 1997. Wnt signaling: a common theme in animal development. Genes Dev. 11, 3286-305. Calvo, S., et al., 2001. Molecular dissection of DNA sequences and factors involved in slow muscle-specific transcription. Mol Cell Biol. 21, 8490-503. Canales, C.P., et al., 2015. The role of GTF2IRD1 in the auditory pathology of Williams-Beuren Syndrome. Eur J Hum Genet. 23, 774-80. Capossela, S., et al., 2012. Growth defects and impaired cognitive-behavioral abilities in mice with knockout for Eif4h, a gene located in the mouse homolog of the Williams-Beuren syndrome critical region. Am J Pathol. 180, 1121-35. Caraveo, G., et al., 2006. Action of TFII-I outside the nucleus as an inhibitor of agonist-induced calcium entry. Science. 314, 122-5. Carmona-Mora, P., et al., 2015. The nuclear localization pattern and interaction partners of GTF2IRD1 demonstrate a role in chromatin regulation. Hum Genet. Celesia, G.G., 2013. Disorders of Peripheral and Central Auditory Processing1: Disorders of Peripheral and Central Auditory Processing, Vol., Elsevier Science Health Science Division. Celesia, G.G., Hickok, G., 2015. The Human Auditory System: Fundamental Organization and Clinical Disorders, Vol., Elsevier Science. Chacon, G.E., Ugalde, C.M., Jabero, M.F., 2007. Genetic disorders and bone affecting the craniofacial skeleton. Oral Maxillofac Surg Clin North Am. 19, 467-74, v. Chen, W.Y., Abatangelo, G., 1999. Functions of hyaluronan in wound repair. Wound Repair Regen. 7, 79-89. Chen, X.J., Nieman, B.J., 2011. Mouse phenotyping with MRI. Methods Mol Biol. 771, 595-631. Cheriyath, V., Roy, A.L., 2000. Alternatively spliced isoforms of TFII-I. Complex formation, nuclear translocation, and differential gene regulation. J Biol Chem. 275, 26300-8. Cheriyath, V., Desgranges, Z.P., Roy, A.L., 2002. c-Src-dependent transcriptional activation of TFII-I. J Biol Chem. 277, 22798-805. Cherniske, E.M., et al., 1999. Early puberty in Williams syndrome. Clin Dysmorphol. 8, 117-21. Cherniske, E.M., et al., 2004. Multisystem study of 20 older adults with Williams syndrome. Am J Med Genet A. 131, 255-64. Chimge, N.O., et al., 2007. Expression profiling of BEN regulated genes in mouse embryonic fibroblasts. J Exp Zool B Mol Dev Evol. 308, 209-24. Chimge, N.O., et al., 2008. Identification of the TFII-I family target genes in the vertebrate genome. Proc Natl Acad Sci U S A. 105, 9006-10. Chimge, N.O., et al., 2012. PI3K/Akt-dependent functions of TFII-I transcription factors in mouse embryonic stem cells. J Cell Biochem. 113, 1122-31. Ching, T., Huang, S., Garmire, L.X., 2014. Power analysis and sample size estimation for RNA- Seq differential expression. Rna. 20, 1684-96. Chung, K., Deisseroth, K., 2013. CLARITY for mapping the nervous system. Nat Methods. 10, 508-13. Chung, K., et al., 2013. Structural and molecular interrogation of intact biological systems. Nature. 497, 332-7. Collin, C., et al., 1992. Characterization of human cytokeratin 2, an epidermal cytoskeletal protein synthesized late during differentiation. Exp Cell Res. 202, 132-41.

245

References

Collins, R.T., 2nd, et al., 2010. Cardiovascular abnormalities, interventions, and long-term outcomes in infantile Williams syndrome. J Pediatr. 156, 253-8.e1. Crackower, M.A., et al., 2003. Essential role of Fkbp6 in male fertility and homologous chromosome pairing in meiosis. Science. 300, 1291-5. Croce, J.C., McClay, D.R., 2008. Evolution of the Wnt pathways. Methods Mol Biol. 469, 3-18. Curran, M.E., et al., 1993. The elastin gene is disrupted by a translocation associated with supravalvular aortic stenosis. Cell. 73, 159-68. Dai, L., et al., 2009. Is it Williams syndrome? GTF2IRD1 implicated in visual-spatial construction and GTF2I in sociability revealed by high resolution arrays. Am J Med Genet A. 149a, 302-14. Danoff, S.K., et al., 2004. TFII-I, a candidate gene for Williams syndrome cognitive profile: parallels between regional expression in mouse brain and human phenotype. Neuroscience. 123, 931-8. De Fine Olivarius, F., Agner, T., Menne, T., 1993. Skin barrier function and dermal inflammation. An experimental study of transepidermal water loss after dermal tuberculin injection compared with SLS patch testing. Br J Dermatol. 129, 554-7. de Ruijter, A.J., et al., 2003. Histone deacetylases (HDACs): characterization of the classical HDAC family. Biochem J. 370, 737-49. De Zeeuw, C.I., et al., 1997. CLIP-115, a novel brain-specific cytoplasmic linker protein, mediates the localization of dendritic lamellar bodies. Neuron. 19, 1187-99. Del Campo, M., et al., 2006. Hemizygosity at the NCF1 gene in patients with Williams-Beuren syndrome decreases their risk of hypertension. Am J Hum Genet. 78, 533-42. DeSilva, U., et al., 2002. Generation and comparative analysis of approximately 3.3 Mb of mouse genomic sequence orthologous to the region of human chromosome 7q11.23 implicated in Williams syndrome. Genome Res. 12, 3-15. Despouy, G., et al., 2003. Cyclin D3 is a cofactor of retinoic acid receptors, modulating their activity in the presence of cellular retinoic acid-binding protein II. J Biol Chem. 278, 6355-62. Devlin, B., Scherer, S.W., 2012. Genetic architecture in autism spectrum disorder. Curr Opin Genet Dev. 22, 229-37. Dilts, C.V., Morris, C.A., Leonard, C.O., 1990. Hypothesis for development of a behavioral phenotype in Williams syndrome. Am J Med Genet Suppl. 6, 126-31. DiTommaso, T., et al., 2014. Keratin 76 is required for tight junction function and maintenance of the skin barrier. PLoS Genet. 10, e1004706. Dixit, A., et al., 2013a. 7q11.23 Microduplication: a recognizable phenotype. Clin Genet. 83, 155-61. Dixit, A., et al., 2013b. 7q11.23 Microduplication: a recognizable phenotype. Clinical Genetics. 83, 155-161. Doi, T., et al., 2010. Disruption of noncanonical Wnt/CA2+ pathway in the cadmium-induced omphalocele in the chick model. J Pediatr Surg. 45, 1645-9. Dommisch, H., Jepsen, S., 2015. Diverse functions of defensins and other antimicrobial peptides in periodontal tissues. Periodontol 2000. 69, 96-110. Donnai, D., Karmiloff-Smith, A., 2000. Williams syndrome: from genotype through to the cognitive phenotype. Am J Med Genet. 97, 164-71. Dowdle, J.A., et al., 2013. Mouse BAZ1A (ACF1) is dispensable for double-strand break repair but is essential for averting improper gene expression during spermatogenesis. PLoS Genet. 9, e1003945. Doyle, T.F., et al., 2004. "Everybody in the world is my friend" hypersociability in young children with Williams syndrome. Am J Med Genet A. 124a, 263-73. Dridi, S.M., et al., 1999. Skin elastic fibers in Williams syndrome. Am J Med Genet. 87, 134-8.

246

References

Driskell, I., et al., 2012. The histone methyltransferase Setd8 acts in concert with c-Myc and is required to maintain skin. Embo j. 31, 616-29. Duncan, J.L., et al., 2012. Identification of a novel mutation in the CDHR1 gene in a family with recessive retinal degeneration. Arch Ophthalmol. 130, 1301-8. Durkin, M.E., et al., 2001. Integration of a c-myc transgene results in disruption of the mouse Gtf2ird1 gene, the homologue of the human GTF2IRD1 gene hemizygously deleted in Williams-Beuren syndrome. Genomics. 73, 20-7. Edelmann, L., et al., 2007. An atypical deletion of the Williams-Beuren syndrome interval implicates genes associated with defective visuospatial processing and autism. J Med Genet. 44, 136-43. Eie, H., et al., 1972. Localized supravalvular aortic stenosis combined with mental retardation and peculiar facial appearance. Acta Med Scand. 191, 517-20. Elias, P.M., Menon, G.K., 1991. Structural and lipid biochemical correlates of the epidermal permeability barrier. Adv Lipid Res. 24, 1-26. Ellegood, J., et al., 2014. Neuroanatomical phenotypes in a mouse model of the 22q11.2 microdeletion. Mol Psychiatry. 19, 99-107. Elsner, P., Berardesca, E., Wilhelm, K.P., 2001. Bioengineering of the Skin: Skin Biomechanics, Vol., Taylor & Francis. Enkhmandakh, B., et al., 2004. The early embryonic expression of TFII-I during mouse preimplantation development. Gene Expr Patterns. 4, 25-8. Enkhmandakh, B., et al., 2009. Essential functions of the Williams-Beuren syndrome-associated TFII-I genes in embryonic development. Proc Natl Acad Sci U S A. 106, 181-6. Epstein, W.L., Maibach, H.I., 1965. Cell renewal in human epidermis. Arch Dermatol. 92, 462-8. Euteneuer, J., et al., 2014. Molecular and phenotypic characterization of atypical Williams- Beuren syndrome. Clin Genet. 86, 487-91. Ewart, A.K., et al., 1993. Hemizygosity at the elastin locus in a developmental disorder, Williams syndrome. Nat Genet. 5, 11-6. Ewart, A.K., et al., 1994. Supravalvular aortic stenosis associated with a deletion disrupting the elastin gene. J Clin Invest. 93, 1071-7. Ezhkova, E., et al., 2009. Ezh2 orchestrates gene expression for the stepwise differentiation of tissue-specific stem cells. Cell. 136, 1122-35. Ezhkova, E., et al., 2011. EZH1 and EZH2 cogovern histone H3K27 trimethylation and are essential for hair follicle homeostasis and wound repair. Genes Dev. 25, 485-98. Farage, M.A., Miller, K.W., Maibach, H.I., 2010. Textbook of Aging Skin, Vol., Springer. Farrehi, C., Dotter, C.T., Griswold, H.E., 1964. SUPRAVALVULAR AORTIC STENOSIS. UNDERDEVELOPMENT AND CHARACTERISTIC FACIES. Am J Dis Child. 108, 335-40. Fay, J.E., Lynn, R.B., Partington, M.W., 1966. Supravalvular aortic stenosis, mental retardation and a characteristic facies. Can Med Assoc J. 94, 295-7. Ferrero, G.B., et al., 2010. An atypical 7q11.23 deletion in a normal IQ Williams-Beuren syndrome patient. Eur J Hum Genet. 18, 33-8. Fijalkowska, I., et al., 2010. Expression of the transcription factor, TFII-I, during post- implantation mouse embryonic development. BMC Res Notes. 3, 203. Fire, A., 1992. Histochemical techniques for locating Escherichia coli beta-galactosidase activity in transgenic organisms. Genet Anal Tech Appl. 9, 151-8. Fischer, H., et al., 2014. Loss of keratin K2 expression causes aberrant aggregation of K10, hyperkeratosis, and inflammation. J Invest Dermatol. 134, 2579-88. Flavell, S.W., Greenberg, M.E., 2008. Signaling mechanisms linking neuronal activity to gene expression and plasticity of the nervous system. Annu Rev Neurosci. 31, 563-90. Flock, A., et al., 1999. Supporting cells contribute to control of hearing sensitivity. J Neurosci. 19, 4498-507.

247

References

Florentine, M., Marvit, P., Buus, S., 2001. Maximum-likelihood yes-no procedure for gap detection: effect of track length. J Am Acad Audiol. 12, 113-20. Folgueras, A.R., et al., 2013. Architectural niche organization by LHX2 is linked to hair follicle stem cell function. Cell Stem Cell. 13, 314-27. Francke, U., 1999. Williams-Beuren syndrome: genes and mechanisms. Hum Mol Genet. 8, 1947-54. Frangiskakis, J.M., et al., 1996. LIM-kinase1 hemizygosity implicated in impaired visuospatial constructive cognition. Cell. 86, 59-69. Franke, Y., Peoples, R.J., Francke, U., 1999. Identification of GTF2IRD1, a putative transcription factor within the Williams-Beuren syndrome deletion at 7q11.23. Cytogenet Cell Genet. 86, 296-304. Fuchs, E., 1993. Epidermal differentiation and keratin gene expression. J Cell Sci Suppl. 17, 197- 208. Fuchs, E., Weber, K., 1994. Intermediate filaments: structure, dynamics, function, and disease. Annu Rev Biochem. 63, 345-82. Fuchs, E., Raghavan, S., 2002. Getting under the skin of epidermal morphogenesis. Nat Rev Genet. 3, 199-209. Fuchs, E., 2008. Skin stem cells: rising to the surface. J Cell Biol. 180, 273-84. Fuchs, E., Horsley, V., 2008. More than one way to skin. Genes Dev. 22, 976-85. Fujiwara, T., et al., 2006. Analysis of knock-out mice to determine the role of HPC-1/syntaxin 1A in expressing synaptic plasticity. J Neurosci. 26, 5767-76. Fusco, C., et al., 2014. Smaller and larger deletions of the Williams Beuren syndrome region implicate genes involved in mild facial phenotype, epilepsy and autistic traits. Eur J Hum Genet. 22, 64-70. Gallo, F.J., et al., 2008. Expecting the worst: observations of reactivity to sound in young children with Williams syndrome. Res Dev Disabil. 29, 567-81. Gandarillas, A., Freije, A., 2014. Cycling up the epidermis: reconciling 100 years of debate. Exp Dermatol. 23, 87-91. Gerber, P.A., et al., 2013. Systematic identification and characterization of novel human skin- associated genes encoding membrane and secreted proteins. PLoS One. 8, e63949. Gerdes, J., et al., 1991. Immunobiochemical and molecular biologic characterization of the cell proliferation-associated nuclear antigen that is defined by monoclonal antibody Ki-67. Am J Pathol. 138, 867-73. Ghomrasseni, S., et al., 2001. Morphometric analysis of elastic skin fibres from patients with: cutis laxa, anetoderma, pseudoxanthoma elasticum, and Buschke-Ollendorff and Williams-Beuren syndromes. J Eur Acad Dermatol Venereol. 15, 305-11. Gineste, C., et al., 2013. In vivo and in vitro investigations of heterozygous nebulin knock-out mice disclose a mild skeletal muscle phenotype. Neuromuscul Disord. 23, 357-69. Godoy, J.A., et al., 2014. Signaling pathway cross talk in Alzheimer's disease. Cell Commun Signal. 12, 23. Goergen, C.J., et al., 2011. Induced chromosome deletion in a Williams-Beuren syndrome mouse model causes cardiovascular abnormalities. J Vasc Res. 48, 119-29. Gothelf, D., et al., 2006. Hyperacusis in Williams syndrome: characteristics and associated neuroaudiologic abnormalities. Neurology. 66, 390-5. Grabherr, M.G., et al., 2011. Full-length transcriptome assembly from RNA-Seq data without a reference genome. Nat Biotechnol. 29, 644-52. Gray, V., et al., 2006. In-depth analysis of spatial cognition in Williams syndrome: A critical assessment of the role of the LIMK1 gene. Neuropsychologia. 44, 679-85. Green, M.R., Sambrook, J., 2012. Molecular cloning : a laboratory manual, Vol., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

248

References

Grether-Beck, S., et al., 2012. Urea uptake enhances barrier function and antimicrobial defense in humans by regulating epidermal gene expression. J Invest Dermatol. 132, 1561-72. Gu, L.H., Coulombe, P.A., 2007. Keratin function in skin epithelia: a broadening palette with surprising shades. Curr Opin Cell Biol. 19, 13-23. Guay, J.M., Dorion, M., Roy, M., 1968. [Supravalvular aortic stenosis, mental retardation and peculiar facies]. Union Med Can. 97, 1811-6. Guinan, J.J., Jr., Stankovic, K.M., 1996. Medial efferent inhibition produces the largest equivalent attenuations at moderate to high sound levels in cat auditory-nerve fibers. J Acoust Soc Am. 100, 1680-90. Guo, L., Yu, Q.C., Fuchs, E., 1993. Targeting expression of keratinocyte growth factor to keratinocytes elicits striking changes in epithelial differentiation in transgenic mice. Embo j. 12, 973-86. Hajihosseini, M.K., 2008. Fibroblast growth factor signaling in cranial suture development and pathogenesis. Front Oral Biol. 12, 160-77. Hakre, S., et al., 2006. Opposing functions of TFII-I spliced isoforms in growth factor-induced gene expression. Mol Cell. 24, 301-8. Hammarlund, K., Sedin, G., 1979. Transepidermal water loss in newborn infants. III. Relation to gestational age. Acta Paediatr Scand. 68, 795-801. Hammond, P., et al., 2005. Discriminating power of localized three-dimensional facial morphology. Am J Hum Genet. 77, 999-1010. Hans, M., Madaan Hans, V., 2014. Epithelial antimicrobial peptides: guardian of the oral cavity. Int J Pept. 2014, 370297. Hansen, K.D., Brenner, S.E., Dudoit, S., 2010. Biases in Illumina transcriptome sequencing caused by random hexamer priming. Nucleic Acids Res. 38, e131. Hardman, M.J., et al., 1998. Patterned acquisition of skin barrier function during development. Development. 125, 1541-52. Hartel, G., Frick, M.H., Halonen, P.I., 1968. Supravalvular pulmonic stenosis, abnormal facial appearance, and mental retardation. Am Heart J. 75, 540-4. Hasegawa, T., et al., 1969. [Case report of pulmonary stenosis associated with atrial septal defect, abnormal facial appearance]. Kyobu Geka. 22, 905-9. Hashiba, K., et al., 1970. The syndrome of supravalvular aortic stenosis, peripheral pulmonary stenosis, mental retardation, and characteristic facies. Jpn Circ J. 34, 109-15. Heller, R., et al., 2003. Partial deletion of the critical 1.5 Mb interval in Williams-Beuren syndrome. J Med Genet. 40, e99. Hellman, W.S., Hellman, R.P., 2001. Revisiting relations between loudness and intensity discrimination. J Acoust Soc Am. 109, 2098-102. Henry, K.R., Chole, R.A., 1980. Genotypic differences in behavioral, physiological and anatomical expressions of age-related hearing loss in the laboratory mouse. Audiology. 19, 369-83. Hequembourg, S., Liberman, M.C., 2001. Spiral ligament pathology: a major aspect of age- related cochlear degeneration in C57BL/6 mice. J Assoc Res Otolaryngol. 2, 118-29. Herzog, F., Winter, H., Schweizer, J., 1994. The large type II 70-kDa keratin of mouse epidermis is the ortholog of human keratin K2e. J Invest Dermatol. 102, 165-70. Hinsley, T.A., et al., 2004. Comparison of TFII-I gene family members deleted in Williams- Beuren syndrome. Protein Sci. 13, 2588-99. Hirota, H., et al., 2003. Williams syndrome deficits in visual spatial processing linked to GTF2IRD1 and GTF2I on chromosome 7q11.23. Genet Med. 5, 311-21. Hobart, H.H., et al., 2010. Inversion of the Williams syndrome region is a common polymorphism found more frequently in parents of children with Williams syndrome. Am J Med Genet C Semin Med Genet. 154C, 220-8.

249

References

Hocking, D.R., Bradshaw, J.L., Rinehart, N.J., 2008. Fronto-parietal and cerebellar contributions to motor dysfunction in Williams syndrome: a review and future directions. Neurosci Biobehav Rev. 32, 497-507. Hoffman, H.S., Ison, J.R., 1980. Reflex modification in the domain of startle: I. Some empirical findings and their implications for how the nervous system processes sensory input. Psychol Rev. 87, 175-89. Hood, L.J., 1998. Clinical applications of the auditory brainstem response, Vol., Singular Publishing Group. Hoogenraad, C.C., et al., 1998. The murine CYLN2 gene: genomic organization, chromosome localization, and comparison to the human gene that is located within the 7q11.23 Williams syndrome critical region. Genomics. 53, 348-58. Hoogenraad, C.C., et al., 2002. Targeted mutation of Cyln2 in the Williams syndrome critical region links CLIP-115 haploinsufficiency to neurodevelopmental abnormalities in mice. Nat Genet. 32, 116-27. Hoogenraad, C.C., et al., 2004. LIMK1 and CLIP-115: linking cytoskeletal defects to Williams syndrome. Bioessays. 26, 141-50. Hopyan, T., et al., 2001. Music skills and the expressive interpretation of music in children with Williams-Beuren syndrome: pitch, rhythm, melodic imagery, phrasing, and musical affect. Child Neuropsychol. 7, 42-53. Hovis, C.L., Butler, M.G., 1997. Photoanthropometric study of craniofacial traits in individuals with Williams syndrome. Clin Genet. 51, 379-87. Howard, M.L., et al., 2012. Mutation of Gtf2ird1 from the Williams-Beuren syndrome critical region results in facial dysplasia, motor dysfunction, and altered vocalisations. Neurobiol Dis. 45, 913-22. Hu, D.N., et al., 1998. TGF-beta2 inhibits growth of uveal melanocytes at physiological concentrations. Exp Eye Res. 67, 143-50. Hu, W., et al., 2010. Alkaline ceramidase 3 (ACER3) hydrolyzes unsaturated long-chain ceramides, and its down-regulation inhibits both cell proliferation and apoptosis. J Biol Chem. 285, 7964-76. Huang, M.H., et al., 1993. Sudden bilateral hearing impairment in vertebrobasilar occlusive disease. Stroke. 24, 132-7. Hussain, S.H., Limthongkul, B., Humphreys, T.R., 2013. The biomechanical properties of the skin. Dermatol Surg. 39, 193-203. Igyarto, B.Z., Kaplan, D.H., 2013. Antigen presentation by Langerhans cells. Curr Opin Immunol. 25, 115-9. Iizuka, K., et al., 2004. Deficiency of carbohydrate response element-binding protein (ChREBP) reduces lipogenesis as well as glycolysis. Proc Natl Acad Sci U S A. 101, 7281-6. Iizuka, K., Horikawa, Y., 2008. ChREBP: a glucose-activated transcription factor involved in the development of metabolic syndrome. Endocr J. 55, 617-24. Ikadai, H., et al., 1992. An aspermia rat mutant (as/as) with spermatogenic failure at meiosis. J Vet Med Sci. 54, 745-9. Imakado, S., et al., 1995. Targeting expression of a dominant-negative retinoic acid receptor mutant in the epidermis of transgenic mice results in loss of barrier function. Genes Dev. 9, 317-29. Issa, L.L., et al., 2006. MusTRD can regulate postnatal fiber-specific expression. Dev Biol. 293, 104-15. Janssen, E., et al., 2004. Linker for activation of B cells: a functional equivalent of a mutant linker for activation of T cells deficient in phospholipase C-gamma1 binding. J Immunol. 172, 6810-9. Jarvinen, A., Korenberg, J.R., Bellugi, U., 2013. The social phenotype of Williams syndrome. Curr Opin Neurobiol. 23, 414-22.

250

References

Jastreboff, P.J., Jastreboff, M.M., 2015. Decreased sound tolerance: hyperacusis, misophonia, diplacousis, and polyacousis. Handb Clin Neurol. 129, 375-87. Jensen, J.M., Proksch, E., 2009. The skin's barrier. G Ital Dermatol Venereol. 144, 689-700. Jewett, D.L., Romano, M.N., Williston, J.S., 1970. Human auditory evoked potentials: possible brain stem components detected on the scalp. Science. 167, 1517-8. Ji, Y., et al., 2000. Structure of chromosomal duplicons and their role in mediating human genomic disorders. Genome Res. 10, 597-610. Jian, Y., et al., 2005. Cyclin D3 interacts with vitamin D receptor and regulates its transcription activity. Biochem Biophys Res Commun. 335, 739-48. Jiang, K., et al., 2015. The Impact of Cystic Fibrosis Transmembrane Regulator Disruption on Cardiac Function and Stress Response. J Cyst Fibros. Jiang, W., et al., 2005. An FF domain-dependent protein interaction mediates a signaling pathway for growth factor-induced gene expression. Mol Cell. 17, 23-35. Johnson, L.B., Comeau, M., Clarke, K.D., 2001. Hyperacusis in Williams syndrome. J Otolaryngol. 30, 90-2. Jones, K.L., Smith, D.W., 1975. The Williams elfin facies syndrome. A new perspective. J Pediatr. 86, 718-23. Kakasheva-Mazhenkovska, L., et al., 2011. Variations of the histomorphological characteristics of human skin of different body regions in subjects of different age. Prilozi. 32, 119-28. Kalderon, D., et al., 1984. A short amino acid sequence able to specify nuclear location. Cell. 39, 499-509. Kammerer, R., et al., 2012. Loss of mammal-specific tectorial membrane component carcinoembryonic antigen cell adhesion molecule 16 (CEACAM16) leads to hearing impairment at low and high frequencies. J Biol Chem. 287, 21584-98. Kaplan, P., et al., 1989. Contractures in patients with Williams syndrome. Pediatrics. 84, 895-9. Kaplan, P., Wang, P.P., Francke, U., 2001. Williams (Williams Beuren) syndrome: a distinct neurobehavioral disorder. J Child Neurol. 16, 177-90. Katoh, Y., Katoh, M., 2009. FGFR2-related pathogenesis and FGFR2-targeted therapeutics (Review). Int J Mol Med. 23, 307-11. Kerrien, S., et al., 2012. The IntAct molecular interaction database in 2012. Nucleic Acids Res. 40, D841-6. Kiernan, A.E., et al., 2005. Sox2 is required for sensory organ development in the mammalian inner ear. Nature. 434, 1031-5. Kim, D., et al., 2013a. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 14, R36. Kim, D.W., Cochran, B.H., 2001. JAK2 activates TFII-I and regulates its interaction with extracellular signal-regulated kinase. Mol Cell Biol. 21, 3387-97. Kim, J., et al., 2010. Wnt5a induces endothelial inflammation via beta-catenin-independent signaling. J Immunol. 185, 1274-82. Kim, S.Y., Chung, K., Deisseroth, K., 2013b. Light microscopy mapping of connections in the intact brain. Trends Cogn Sci. 17, 596-9. Kim, T.Y., et al., 2001. Role of p27(Kip1) in cAMP- and TGF-beta2-mediated antiproliferation in rabbit corneal endothelial cells. Invest Ophthalmol Vis Sci. 42, 3142-9. Kirov, G., et al., 2012. De novo CNV analysis implicates specific abnormalities of postsynaptic signalling complexes in the pathogenesis of schizophrenia. Mol Psychiatry. 17, 142-53. Klein, A.J., et al., 1990. Hyperacusis and otitis media in individuals with Williams syndrome. J Speech Hear Disord. 55, 339-44. Klump, G.M., 1995. Methods in comparative psychoacoustics, Vol., Birkhäuser Verlag. Kohler, M., et al., 1999. Evidence for distinct substrate specificities of importin alpha family members in nuclear protein import. Mol Cell Biol. 19, 7782-91.

251

References

Konger, R.L., et al., 2005. Immunolocalization of low-affinity prostaglandin E receptors, EP and EP, in adult human epidermis. J Invest Dermatol. 124, 965-70. Koster, M.I., Roop, D.R., 2007. Mechanisms regulating epithelial stratification. Annu Rev Cell Dev Biol. 23, 93-113. Kozel, B.A., et al., 2014. Skin findings in Williams syndrome. Am J Med Genet A. 164a, 2217-25. Kretzschmar, K., Watt, F.M., 2014. Markers of epidermal stem cell subpopulations in adult mammalian skin. Cold Spring Harb Perspect Med. 4. Ku, M.C., et al., 2015. "Reuse" study of low-tube-voltage CT arterial phase in the spoiled gadoxetic-acid liver MRI. Clin Imaging. Lam, P.P., et al., 2005. Transgenic mouse overexpressing syntaxin-1A as a diabetes model. Diabetes. 54, 2744-54. Lane, E.B., McLean, W.H., 2004. Keratins and skin disorders. J Pathol. 204, 355-66. Langmead, B., Salzberg, S.L., 2012. Fast gapped-read alignment with Bowtie 2. Nat Methods. 9, 357-9. Lashkari, A., Smith, A.K., Graham, J.M., Jr., 1999. Williams-Beuren syndrome: an update and review for the primary physician. Clin Pediatr (Phila). 38, 189-208. Lazebnik, M.B., Tussie-Luna, M.I., Roy, A.L., 2008. Determination and functional analysis of the consensus binding site for TFII-I family member BEN, implicated in Williams-Beuren syndrome. J Biol Chem. 283, 11078-82. Le, V.P., et al., 2011. Decreased aortic diameter and compliance precedes blood pressure increases in postnatal development of elastin-insufficient mice. Am J Physiol Heart Circ Physiol. 301, H221-9. Leachman, S.A., et al., 2005. Clinical and pathological features of pachyonychia congenita. J Investig Dermatol Symp Proc. 10, 3-17. Legatt, A.D., 2002. Mechanisms of intraoperative brainstem auditory evoked potential changes. J Clin Neurophysiol. 19, 396-408. Lehner, B., Sanderson, C.M., 2004. A protein interaction framework for human mRNA degradation. Genome Res. 14, 1315-23. Lessard, J.C., et al., 2013. Keratin 16 regulates innate immunity in response to epidermal barrier breach. Proc Natl Acad Sci U S A. 110, 19537-42. Levitin, D.J., et al., 2003. Neural correlates of auditory perception in Williams syndrome: an fMRI study. Neuroimage. 18, 74-82. Levitin, D.J., 2005. Musical behavior in a neurogenetic developmental disorder: evidence from Williams Syndrome. Ann N Y Acad Sci. 1060, 325-34. Levitin, D.J., et al., 2005. Aversion, awareness, and attraction: investigating claims of hyperacusis in the Williams syndrome phenotype. J Child Psychol Psychiatry. 46, 514- 23. Levy, V., et al., 2005. Distinct stem cell populations regenerate the follicle and interfollicular epidermis. Dev Cell. 9, 855-61. Li, D.Y., et al., 1997. Elastin point mutations cause an obstructive vascular disease, supravalvular aortic stenosis. Hum Mol Genet. 6, 1021-8. Li, D.Y., et al., 1998a. Elastin is an essential determinant of arterial morphogenesis. Nature. 393, 276-80. Li, D.Y., et al., 1998b. Novel arterial pathology in mice and humans hemizygous for elastin. J Clin Invest. 102, 1783-7. Li, H.H., et al., 2009. Induced chromosome deletions cause hypersociability and other features of Williams-Beuren syndrome in mice. EMBO Mol Med. 1, 50-65. Li, H.S., Borg, E., 1991. Age-related loss of auditory sensitivity in two mouse genotypes. Acta Otolaryngol. 111, 827-34.

252

References

Liu, L., et al., 1997. The human beta-defensin-1 and alpha-defensins are encoded by adjacent genes: two peptide families with differing disulfide topology share a common ancestry. Genomics. 43, 316-20. Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 25, 402-8. London, B., Blodgett, J., 1963. THE SYNDROME OF SUPRAVALVULAR AORTIC STENOSIS, MENTAL RETARDATION AND A CHARACTERISTIC FACIAL APPEARANCE. Bull Sinai Hosp Detroit. 11, 112-9. Lopez, R.G., et al., 2009. C/EBPalpha and beta couple interfollicular keratinocyte proliferation arrest to commitment and terminal differentiation. Nat Cell Biol. 11, 1181-90. Lu, J., et al., 2006. TGF-beta2 inhibits AKT activation and FGF-2-induced corneal endothelial cell proliferation. Exp Cell Res. 312, 3631-40. Lucena, J., et al., 2010. Essential role of the N-terminal region of TFII-I in viability and behavior. BMC Med Genet. 11, 61. Lucertini, M., Moleti, A., Sisto, R., 2002. On the detection of early cochlear damage by otoacoustic emission analysis. J Acoust Soc Am. 111, 972-8. Luis, N.M., et al., 2011. Regulation of human epidermal stem cell proliferation and senescence requires polycomb- dependent and -independent functions of Cbx4. Cell Stem Cell. 9, 233-46. Luo, R., et al., 2012. Genome-wide transcriptome profiling reveals the functional impact of rare de novo and recurrent CNVs in autism spectrum disorders. Am J Hum Genet. 91, 38-55. Lupski, J.R., 1998. Genomic disorders: structural features of the genome can lead to DNA rearrangements and human disease traits. Trends Genet. 14, 417-22. Makeyev, A.V., et al., 2004. GTF2IRD2 is located in the Williams-Beuren syndrome critical region 7q11.23 and encodes a protein with two TFII-I-like helix-loop-helix repeats. Proc Natl Acad Sci U S A. 101, 11052-7. Makeyev, A.V., Bayarsaihan, D., 2009. New TFII-I family target genes involved in embryonic development. Biochem Biophys Res Commun. 386, 554-8. Marchese, C., et al., 1995. Modulation of keratinocyte growth factor and its receptor in reepithelializing human skin. J Exp Med. 182, 1369-76. Mardaryev, A.N., et al., 2011. Lhx2 differentially regulates Sox9, Tcf4 and Lgr5 in hair follicle stem cells to promote epidermal regeneration after injury. Development. 138, 4843- 52. Marekov, L.N., Steinert, P.M., 1998. Ceramides are bound to structural proteins of the human foreskin epidermal cornified cell envelope. J Biol Chem. 273, 17763-70. Marler, J.A., et al., 2005. Sensorineural hearing loss in children and adults with Williams syndrome. Am J Med Genet A. 138, 318-27. Marler, J.A., et al., 2010. Auditory function and hearing loss in children and adults with Williams syndrome: cochlear impairment in individuals with otherwise normal hearing. Am J Med Genet C Semin Med Genet. 154c, 249-65. Mass, E., Belostoky, L., 1993. Craniofacial morphology of children with Williams syndrome. Cleft Palate Craniofac J. 30, 343-9. Masuda, T., et al., 2014. The transcription factor GTF2IRD1 regulates the topology and function of photoreceptors by modulating photoreceptor gene expression across the retina. J Neurosci. 34, 15356-68. Matsumoto, N., et al., 2010. Pivotal role of actin depolymerization in the regulation of cochlear outer hair cell motility. Biophys J. 99, 2067-76. Matsumoto, N., Kitani, R., Kalinec, F., 2011. Linking LIMK1 deficiency to hyperacusis and progressive hearing loss in individuals with Williams syndrome. Commun Integr Biol. 4, 208-10.

253

References

Mauermann, M., Kollmeier, B., 2004. Distortion product otoacoustic emission (DPOAE) input/output functions and the influence of the second DPOAE source. J Acoust Soc Am. 116, 2199-212. McOmish, C.E., et al., 2008. Phospholipase C-beta1 knockout mice exhibit endophenotypes modeling schizophrenia which are rescued by environmental enrichment and clozapine administration. Mol Psychiatry. 13, 661-72. McRory, J.E., et al., 2008. Syntaxin 1A is required for normal in utero development. Biochem Biophys Res Commun. 375, 372-7. Meerwaldt, R., et al., 2004. Simple non-invasive assessment of advanced glycation endproduct accumulation. Diabetologia. 47, 1324-30. Mehlem, A., et al., 2013. Imaging of neutral lipids by oil red O for analyzing the metabolic status in health and disease. Nat Protoc. 8, 1149-54. Mejetta, S., et al., 2011. Jarid2 regulates mouse epidermal stem cell activation and differentiation. Embo j. 30, 3635-46. Meng, X., et al., 1998. A novel human gene FKBP6 is deleted in Williams syndrome. Genomics. 52, 130-7. Meng, Y., et al., 2002. Abnormal spine morphology and enhanced LTP in LIMK-1 knockout mice. Neuron. 35, 121-33. Merchant, S.N., Nadol, J.B., 2010. Schuknecht's Pathology of the Ear, Vol., People's Medical Publishing House-USA. Merla, G., et al., 2006. Submicroscopic deletion in patients with Williams-Beuren syndrome influences expression levels of the nonhemizygous flanking genes. Am J Hum Genet. 79, 332-41. Merla, G., et al., 2010. Copy number variants at Williams-Beuren syndrome 7q11.23 region. Hum Genet. 128, 3-26. Merla, G., et al., 2012. Supravalvular aortic stenosis: elastin arteriopathy. Circ Cardiovasc Genet. 5, 692-6. Mervis, C.B., Robinson, B.F., Pani, J.R., 1999. Visuospatial construction. Am J Hum Genet. 65, 1222-9. Mervis, C.B., Klein-Tasman, B.P., 2000. Williams syndrome: cognition, personality, and adaptive behavior. Ment Retard Dev Disabil Res Rev. 6, 148-58. Mervis, C.B., Robinson, B.F., 2000. Expressive vocabulary ability of toddlers with Williams syndrome or Down syndrome: a comparison. Dev Neuropsychol. 17, 111-26. Mervis, C.B., et al., 2000. The Williams syndrome cognitive profile. Brain Cogn. 44, 604-28. Mervis, C.B., et al., 2012. Duplication of GTF2I results in separation anxiety in mice and humans. Am J Hum Genet. 90, 1064-70. Mervis, C.B., et al., 2015. Children with 7q11.23 duplication syndrome: Psychological characteristics. Am J Med Genet A. 167, 1436-50. Metcalfe, K., et al., 2000. Elastin: mutational spectrum in supravalvular aortic stenosis. Eur J Hum Genet. 8, 955-63. Metzinger, M.N., et al., 2014. Correlation of X-ray computed tomography with quantitative nuclear magnetic resonance methods for pre-clinical measurement of adipose and lean tissues in living mice. Sensors (Basel). 14, 18526-42. Michibata, H., et al., 2004. Identification and characterization of a novel component of the cornified envelope, cornifelin. Biochem Biophys Res Commun. 318, 803-13. Mikaelian, D.O., 1979. Development and degeneration of hearing in the C57/b16 mouse: relation of electrophysiologic responses from the round window and cochlear nucleus to cochlear anatomy and behavioral responses. Laryngoscope. 89, 1-15. Moller, A.R., Jannetta, P.J., 1982. Auditory evoked potentials recorded intracranially from the brain stem in man. Exp Neurol. 78, 144-57.

254

References

Moore, B.C., et al., 1999. Inter-relationship between different psychoacoustic measures assumed to be related to the cochlear active mechanism. J Acoust Soc Am. 106, 2761- 78. Moore, B.C.J., 2012. An Introduction to the Psychology of Hearing, Vol., Emerald. Morey, L., Helin, K., 2010. Polycomb group protein-mediated repression of transcription. Trends Biochem Sci. 35, 323-32. Morin, R., et al., 2008. Profiling the HeLa S3 transcriptome using randomly primed cDNA and massively parallel short-read sequencing. Biotechniques. 45, 81-94. Morris, C.A., et al., 1988. Natural history of Williams syndrome: physical characteristics. J Pediatr. 113, 318-26. Morris, C.A., Carey, J.C., 1990. Three diagnostic signs in Williams syndrome. Am J Med Genet Suppl. 6, 100-1. Morris, C.A., et al., 1990. Adults with Williams syndrome. Am J Med Genet Suppl. 6, 102-7. Morris, C.A., et al., 1993a. Supravalvular aortic stenosis cosegregates with a familial 6; 7 translocation which disrupts the elastin gene. Am J Med Genet. 46, 737-44. Morris, C.A., Thomas, I.T., Greenberg, F., 1993b. Williams syndrome: autosomal dominant inheritance. Am J Med Genet. 47, 478-81. Morris, C.A., et al., 2003. GTF2I hemizygosity implicated in mental retardation in Williams syndrome: genotype-phenotype analysis of five families with deletions in the Williams syndrome region. Am J Med Genet A. 123A, 45-59. Mortazavi, A., et al., 2008. Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods. 5, 621-8. Mulle, J.G., et al., 2014. Reciprocal duplication of the Williams-Beuren syndrome deletion on chromosome 7q11.23 is associated with schizophrenia. Biol Psychiatry. 75, 371-7. Narhi, K., et al., 2008. Sustained epithelial beta-catenin activity induces precocious hair development but disrupts hair follicle down-growth and hair shaft formation. Development. 135, 1019-28. Nieman, B.J., et al., 2012. If the skull fits: magnetic resonance imaging and microcomputed tomography for combined analysis of brain and skull phenotypes in the mouse. Physiol Genomics. 44, 992-1002. Niemann, C., Watt, F.M., 2002. Designer skin: lineage commitment in postnatal epidermis. Trends Cell Biol. 12, 185-92. Nimchinsky, E.A., Sabatini, B.L., Svoboda, K., 2002. Structure and function of dendritic spines. Annu Rev Physiol. 64, 313-53. Nishifuji, K., Yoon, J.S., 2013. The stratum corneum: the rampart of the mammalian body. Vet Dermatol. 24, 60-72.e15-6. Noguchi, J., et al., 1993. Age-related changes in blood concentrations of FSH, LH and testosterone and testicular morphology in a new rat sterile mutant with hereditary aspermia. J Reprod Fertil. 97, 433-9. Noguchi, J., et al., 2008. Affected homologous chromosome pairing and phosphorylation of testis specific histone, H2AX, in male meiosis under FKBP6 deficiency. J Reprod Dev. 54, 203-7. Novina, C.D., Cheriyath, V., Roy, A.L., 1998. Regulation of TFII-I activity by phosphorylation. J Biol Chem. 273, 33443-8. O'Leary, J., Osborne, L.R., 2011. Global analysis of gene expression in the developing brain of Gtf2ird1 knockout mice. PLoS One. 6, e23868. O'Mahoney, J.V., et al., 1998. Identification of a novel slow-muscle-fiber enhancer binding protein, MusTRD1. Mol Cell Biol. 18, 6641-52. Ohara-Imaizumi, M., et al., 2007. Imaging analysis reveals mechanistic differences between first- and second-phase insulin exocytosis. J Cell Biol. 177, 695-705.

255

References

Ohazama, A., Sharpe, P.T., 2007. TFII-I gene family during tooth development: candidate genes for tooth anomalies in Williams syndrome. Dev Dyn. 236, 2884-8. Ornitz, D.M., Marie, P.J., 2002. FGF signaling pathways in endochondral and intramembranous bone development and human genetic disease. Genes Dev. 16, 1446-65. Osborne, L.R., et al., 1999. Identification of a putative transcription factor gene (WBSCR11) that is commonly deleted in Williams-Beuren syndrome. Genomics. 57, 279-84. Osborne, L.R., et al., 2001. A 1.5 million-base pair inversion polymorphism in families with Williams-Beuren syndrome. Nat Genet. 29, 321-5. Osborne, L.R., Mervis, C.B., 2007. Rearrangements of the Williams-Beuren syndrome locus: molecular basis and implications for speech and language development. Expert Rev Mol Med. 9, 1-16. Osborne, L.R., 2010. Animal models of Williams syndrome. Am J Med Genet C Semin Med Genet. 154c, 209-19. Oshlack, A., Robinson, M.D., Young, M.D., 2010. From RNA-seq reads to differential expression results. Genome Biol. 11, 220. Ouji, Y., et al., 2006. Wnt-10b secreted from lymphocytes promotes differentiation of skin epithelial cells. Biochem Biophys Res Commun. 342, 1063-9. Ounap, K., et al., 1998. Familial Williams-Beuren syndrome. Am J Med Genet. 80, 491-3. Paglialonga, A., et al., 2011. Cochlear active mechanisms in young normal-hearing subjects affected by Williams syndrome: time-frequency analysis of otoacoustic emissions. Hear Res. 272, 157-67. Paglialonga, A., et al., 2014. Analysis of subtle auditory dysfunctions in young normal-hearing subjects affected by Williams syndrome. Int J Pediatr Otorhinolaryngol. 78, 1861-5. Palacios-Verdu, M.G., et al., 2015. Metabolic abnormalities in Williams-Beuren syndrome. J Med Genet. 52, 248-55. Palmer, S.J., et al., 2007. Expression of Gtf2ird1, the Williams syndrome-associated gene, during mouse development. Gene Expr Patterns. 7, 396-404. Palmer, S.J., et al., 2010. Negative autoregulation of GTF2IRD1 in Williams-Beuren syndrome via a novel DNA binding mechanism. J Biol Chem. 285, 4715-24. Palmer, S.J., et al., 2012. GTF2IRD2 from the Williams-Beuren critical region encodes a mobile- element-derived fusion protein that antagonizes the action of its related family members. J Cell Sci. 125, 5040-50. Pankau, R., et al., 1993. Williams-Beuren syndrome in monozygotic twins with variable expression. Am J Med Genet. 47, 475-7. Pankau, R., et al., 2001. Familial Williams-Beuren syndrome showing varying clinical expression. Am J Med Genet. 98, 324-9. Parham, K., Willott, J.F., 1988. Acoustic startle response in young and aging C57BL/6J and CBA/J mice. Behav Neurosci. 102, 881-6. Parham, K., 1997. Distortion product otoacoustic emissions in the C57BL/6J mouse model of age-related hearing loss. Hear Res. 112, 216-34. Parks, W.C., et al., 1992. Cellular expression of tropoelastin mRNA splice variants. Matrix. 12, 156-62. Parrott, A., et al., 2015. Aortopathy in the 7q11.23 microduplication syndrome. Am J Med Genet A. 167A, 363-70. Partsch, C.J., et al., 1999. Longitudinal evaluation of growth, puberty, and bone maturation in children with Williams syndrome. J Pediatr. 134, 82-9. Pasman, J.W., et al., 1997. Diagnostic and predictive value of auditory evoked responses in preterm infants: II. Auditory evoked responses. Pediatr Res. 42, 670-7. Patuzzi, R., 2011a. Ion flow in stria vascularis and the production and regulation of cochlear endolymph and the endolymphatic potential. Hear Res. 277, 4-19.

256

References

Patuzzi, R., 2011b. Ion flow in cochlear hair cells and the regulation of hearing sensitivity. Hear Res. 280, 3-20. Perez-Gonzalez, D., Malmierca, M.S., 2014. Adaptation in the auditory system: an overview. Front Integr Neurosci. 8, 19. Perez Jurado, A.L., 2003. Williams-Beuren syndrome: a model of recurrent genomic mutation. Horm Res. 59 Suppl 1, 106-13. Perez Jurado, L.A., et al., 1996. Molecular definition of the chromosome 7 deletion in Williams syndrome and parent-of-origin effects on growth. Am J Hum Genet. 59, 781-92. Perez Jurado, L.A., et al., 1998. A duplicated gene in the breakpoint regions of the 7q11.23 Williams-Beuren syndrome deletion encodes the initiator binding protein TFII-I and BAP-135, a phosphorylation target of BTK. Hum Mol Genet. 7, 325-34. Perlman, H.B., 1938. LXXIX Hyperacusis. Annals of Otology, Rhinology & Laryngology. 47, 947- 953. Peterson, L.F., et al., 2005. The hematopoietic transcription factor AML1 (RUNX1) is negatively regulated by the cell cycle protein cyclin D3. Mol Cell Biol. 25, 10205-19. Petiot, A., et al., 2003. A crucial role for Fgfr2-IIIb signalling in epidermal development and hair follicle patterning. Development. 130, 5493-501. Peukert, D., et al., 2011. Lhx2 and Lhx9 determine neuronal differentiation and compartition in the caudal forebrain by regulating Wnt signaling. PLoS Biol. 9, e1001218. Pickles, J.O., 2012. An Introduction to the Physiology of Hearing, Vol., Emerald. Piper M. T, S.M.D., 2011. Comparative Anatomy and Histology: A Mouse and Human Atlas, Vol., Academic Press. Pober, B.R., Johnson, M., Urban, Z., 2008. Mechanisms and treatment of cardiovascular disease in Williams-Beuren syndrome. J Clin Invest. 118, 1606-15. Pober, B.R., 2010. Williams-Beuren syndrome. N Engl J Med. 362, 239-52. Polly, P., et al., 2003. hMusTRD1alpha1 represses MEF2 activation of the troponin I slow enhancer. J Biol Chem. 278, 36603-10. Powers, C.J., McLeskey, S.W., Wellstein, A., 2000. Fibroblast growth factors, their receptors and signaling. Endocr Relat Cancer. 7, 165-97. Proksch, E., Brandner, J.M., Jensen, J.M., 2008. The skin: an indispensable barrier. Exp Dermatol. 17, 1063-72. Proulx, E., et al., 2010. Enhanced prefrontal serotonin 5-HT(1A) currents in a mouse model of Williams-Beuren syndrome with low innate anxiety. J Neurodev Disord. 2, 99-108. Raabe, N., et al., 2015. Determination of liver-specific r2 * of a highly monodisperse USPIO by (59) Fe iron core-labeling in mice at 3 T MRI. Contrast Media Mol Imaging. 10, 153-62. Ranheim, E.A., et al., 2005. Frizzled 9 knock-out mice have abnormal B-cell development. Blood. 105, 2487-94. Rattner, A., et al., 1997. A family of secreted proteins contains homology to the cysteine-rich ligand-binding domain of frizzled receptors. Proc Natl Acad Sci U S A. 94, 2859-63. Ren, S., Rollins, B.J., 2004. Cyclin C/cdk3 promotes Rb-dependent G0 exit. Cell. 117, 239-51. Rentrop, M., et al., 1987. An unusual type-II 70-kilodalton keratin protein of mouse epidermis exhibiting postnatal body-site specificity and sensitivity to hyperproliferation. Differentiation. 34, 189-200. Reymond, A., et al., 2007. Side effects of genome structural changes. Curr Opin Genet Dev. 17, 381-6. Rhee, H., Polak, L., Fuchs, E., 2006. Lhx2 maintains stem cell character in hair follicles. Science. 312, 1946-9. Riby, D.M., et al., 2014. Stranger danger awareness in Williams syndrome. J Intellect Disabil Res. 58, 572-82. Richardson, G.D., et al., 2009. KGF and EGF signalling block hair follicle induction and promote interfollicular epidermal fate in developing mouse skin. Development. 136, 2153-64.

257

References

Richter, N.J., et al., 1999. Further biochemical and kinetic characterization of human eukaryotic initiation factor 4H. J Biol Chem. 274, 35415-24. Richtsmeier, J.T., Baxter, L.L., Reeves, R.H., 2000. Parallels of craniofacial maldevelopment in Down syndrome and Ts65Dn mice. Dev Dyn. 217, 137-45. Rickye, S.H., Henry, E.H., 2001. Behavioral Assessment of Hearing in Mice. In: Handbook of Mouse Auditory Research. Vol., ed.^eds. CRC Press, pp. 19-29. Ring, C., et al., 2002. The role of a Williams-Beuren syndrome-associated helix-loop-helix domain-containing transcription factor in activin/nodal signaling. Genes Dev. 16, 820- 35. Robertson, G., et al., 2010. De novo assembly and analysis of RNA-seq data. Nat Methods. 7, 909-12. Robinson, M.D., McCarthy, D.J., Smyth, G.K., 2010. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 26, 139- 40. Robinson, M.D., Oshlack, A., 2010. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 11, R25. Roop, D., 1995. Defects in the barrier. Science. 267, 474-5. Rorke, E.A., et al., 2010. AP1 factor inactivation in the suprabasal epidermis causes increased epidermal hyperproliferation and hyperkeratosis but reduced carcinogen-dependent tumor formation. Oncogene. 29, 5873-82. Rosen, S., Howell, P., 2011. Signals and systems for speech and hearing, Vol. 29, Brill. Rosenbloom, J., 1987. Elastin: an overview. Methods Enzymol. 144, 172-96. Rothnagel, J.A., et al., 1994. Mutations in the rod domain of keratin 2e in patients with ichthyosis bullosa of Siemens. Nat Genet. 7, 485-90. Roy, A.L., et al., 1991. Cooperative interaction of an initiator-binding transcription initiation factor and the helix-loop-helix activator USF. Nature. 354, 245-8. Roy, A.L., 1997. Cloning of an Inr and E-box binding protein, TFII-I that interacts physically and functionally with USF1. EMBO J. 16, 7091-7104. Roy, A.L., et al., 1997. Cloning of an inr- and E-box-binding protein, TFII-I, that interacts physically and functionally with USF1. Embo j. 16, 7091-104. Roy, A.L., 2007. Signal-induced functions of the transcription factor TFII-I. Biochim Biophys Acta. 1769, 613-21. Sacristan, C., et al., 2009. Characterization of a novel interaction between transcription factor TFII-I and the inducible tyrosine kinase in T cells. Eur J Immunol. 39, 2584-95. Sadler, L.S., et al., 1993. The Williams syndrome: evidence for possible autosomal dominant inheritance. Am J Med Genet. 47, 468-70. Saito, R., et al., 2012. A travel guide to Cytoscape plugins. Nat Methods. 9, 1069-76. Sakurai, T., et al., 2011. Haploinsufficiency of Gtf2i, a gene deleted in Williams Syndrome, leads to increases in social interactions. Autism Res. 4, 28-39. Sanders, S.J., et al., 2011. Multiple recurrent de novo CNVs, including duplications of the 7q11.23 Williams syndrome region, are strongly associated with autism. Neuron. 70, 863-85. Sarruf, D.A., et al., 2005. Cyclin D3 promotes adipogenesis through activation of peroxisome proliferator-activated receptor gamma. Mol Cell Biol. 25, 9985-95. Scherer, S.W., et al., 2005. Observation of a parental inversion variant in a rare Williams- Beuren syndrome family with two affected children. Hum Genet. 117, 383-8. Schimmang, T., 2013. Transcription factors that control inner ear development and their potential for transdifferentiation and reprogramming. Hear Res. 297, 84-90. Schmickel, R.D., 1986. Contiguous gene syndromes: a component of recognizable syndromes. J Pediatr. 109, 231-41.

258

References

Schneider, T., et al., 2012. Anxious, hypoactive phenotype combined with motor deficits in Gtf2ird1 null mouse model relevant to Williams syndrome. Behav Brain Res. 233, 458- 73. Scholzen, T., Gerdes, J., 2000. The Ki-67 protein: from the known and the unknown. J Cell Physiol. 182, 311-22. Schubert, C., 2009. The genomic basis of the Williams-Beuren syndrome. Cell Mol Life Sci. 66, 1178-97. Schwenk, F., Baron, U., Rajewsky, K., 1995. A cre-transgenic mouse strain for the ubiquitous deletion of loxP-flanked gene segments including deletion in germ cells. Nucleic Acids Res. 23, 5080-1. Segura-Puimedon, M., et al., 2014. Heterozygous deletion of the Williams-Beuren syndrome critical interval in mice recapitulates most features of the human disorder. Hum Mol Genet. 23, 6481-94. Sen, G.L., et al., 2008. Control of differentiation in a self-renewing mammalian tissue by the histone demethylase JMJD3. Genes Dev. 22, 1865-70. Sen, G.L., et al., 2010. DNMT1 maintains progenitor function in self-renewing somatic tissue. Nature. 463, 563-7. Sforzini, C., et al., 2002. Renal tract ultrasonography and calcium homeostasis in Williams- Beuren syndrome. Pediatr Nephrol. 17, 899-902. Smith, A., 2006. A glossary for stem-cell biology. Nature. 441, 1060-1060. Smith, L.T., Underwood, R.A., McLean, W.H., 1999. Ontogeny and regional variability of keratin 2e (K2e) in developing human fetal skin: a unique spatial and temporal pattern of keratin expression in development. Br J Dermatol. 140, 582-91. Smyth, I., et al., 2008. A mouse model of harlequin ichthyosis delineates a key role for Abca12 in lipid homeostasis. PLoS Genet. 4, e1000192. Somerville, M.J., et al., 2005. Severe expressive-language delay related to duplication of the Williams-Beuren locus. N Engl J Med. 353, 1694-701. Spoendlin, H., Schrott, A., 1989. Analysis of the human auditory nerve. Hear Res. 43, 25-38. Stankiewicz, P., Lupski, J.R., 2002. Molecular-evolutionary mechanisms for genomic disorders. Curr Opin Genet Dev. 12, 312-9. Stankiewicz, P., et al., 2003. Genome architecture catalyzes nonrecurrent chromosomal rearrangements. Am J Hum Genet. 72, 1101-16. Steel, K.P., Bock, G.R., 1983. Hereditary inner-ear abnormalities in animals. Relationships with human abnormalities. Arch Otolaryngol. 109, 22-9. Steel, K.P., Brown, S.D., 1996. Genetics of deafness. Curr Opin Neurobiol. 6, 520-5. Steel, K.P., Kros, C.J., 2001. A genetic approach to understanding auditory function. Nat Genet. 27, 143-9. Steiling, H., Werner, S., 2003. Fibroblast growth factors: key players in epithelial morphogenesis, repair and cytoprotection. Curr Opin Biotechnol. 14, 533-7. Stevens, H.E., et al., 2010. Fgfr2 is required for the development of the medial prefrontal cortex and its connections with limbic circuits. J Neurosci. 30, 5590-602. Stockard, J.J., Sharbrough, F.W., Tinker, J.A., 1978. Effects of hypothermia on the human brainstem auditory response. Ann Neurol. 3, 368-70. Stone, J.S., Oesterle, E.C., Rubel, E.W., 1998. Recent insights into regeneration of auditory and vestibular hair cells. Curr Opin Neurol. 11, 17-24. Stromme, P., Bjornstad, P.G., Ramstad, K., 2002. Prevalence estimation of Williams syndrome. J Child Neurol. 17, 269-71. Surface, L.E., Thornton, S.R., Boyer, L.A., 2010. Polycomb group proteins set the stage for early lineage commitment. Cell Stem Cell. 7, 288-98. Swensson, O., et al., 1998. Specialized keratin expression pattern in human ridged skin as an adaptation to high physical stress. Br J Dermatol. 139, 767-75.

259

References

Symonette, C.J., Kaur Mann, A., 2014. Hyaluronan-phosphatidylethanolamine polymers form pericellular coats on keratinocytes and promote basal keratinocyte proliferation. 2014, 727459. Szebenyi, G., Fallon, J.F., 1999. Fibroblast growth factors as multifunctional signaling factors. Int Rev Cytol. 185, 45-106. Tam, E., et al., 2008. The common inversion of the Williams-Beuren syndrome region at 7q11.23 does not cause clinical symptoms. Am J Med Genet A. 146a, 1797-806. Tantin, D., et al., 2004. Regulation of immunoglobulin promoter activity by TFII-I class transcription factors. J Biol Chem. 279, 5460-9. Tassabehji, M., et al., 1997. Elastin: genomic structure and point mutations in patients with supravalvular aortic stenosis. Hum Mol Genet. 6, 1029-36. Tassabehji, M., et al., 1999a. A transcription factor involved in skeletal muscle gene expression is deleted in patients with Williams syndrome. Eur J Hum Genet. 7, 737-47. Tassabehji, M., et al., 1999b. Williams syndrome: use of chromosomal microdeletions as a tool to dissect cognitive and physical phenotypes. Am J Hum Genet. 64, 118-25. Tassabehji, M., 2003. Williams-Beuren syndrome: a challenge for genotype-phenotype correlations. Hum Mol Genet. 12 Spec No 2, R229-37. Tassabehji, M., et al., 2005. GTF2IRD1 in craniofacial development of humans and mice. Science. 310, 1184-7. Tassabehji, M., Urban, Z., 2006. Congenital heart disease: Molecular diagnostics of supravalvular aortic stenosis. Methods Mol Med. 126, 129-56. Tay, E.S., et al., 2003. Regulation of alternative splicing of Gtf2ird1 and its impact on slow muscle promoter activity. Biochem J. 374, 359-67. Terwisscha van Scheltinga, A.F., et al., 2013. Fibroblast growth factors in neurodevelopment and psychopathology. Neuroscientist. 19, 479-94. Thompson, P.D., et al., 2007. GTF2IRD1 regulates transcription by binding an evolutionarily conserved DNA motif 'GUCE'. FEBS Lett. 581, 1233-42. Tipney, H.J., et al., 2004. Isolation and characterisation of GTF2IRD2, a novel fusion gene and member of the TFII-I family of transcription factors, deleted in Williams-Beuren syndrome. Eur J Hum Genet. 12, 551-60. Tolg, C., Telmer, P., Turley, E., 2014. Specific sizes of hyaluronan oligosaccharides stimulate fibroblast migration and excisional wound repair. PLoS One. 9, e88479. Torres, M., Gomez-Pardo, E., Gruss, P., 1996. Pax2 contributes to inner ear patterning and optic nerve trajectory. Development. 122, 3381-91. Tussie-Luna, M.I., et al., 2002a. Physical and functional interactions of histone deacetylase 3 with TFII-I family proteins and PIASxbeta. Proc Natl Acad Sci U S A. 99, 12807-12. Tussie-Luna, M.I., et al., 2002b. The SUMO ubiquitin-protein isopeptide ligase family member Miz1/PIASxbeta /Siz2 is a transcriptional cofactor for TFII-I. J Biol Chem. 277, 43185-93. Tyler, R.S., et al., 2009. Tinnitus, hyperacusis, and music. Hearing loss in musicians Prevention and management. 41-52. Ueda, K., et al., 2015. Eight patients with Williams syndrome and craniosynostosis. Eur J Med Genet. 58, 355-7. Uhlen, M., et al., 2010. Towards a knowledge-based Human Protein Atlas. Nat Biotechnol. 28, 1248-50. Uhlen, M., et al., 2015. Proteomics. Tissue-based map of the human proteome. Science. 347, 1260419. Urban, P.P., Caplan, L.R., 2011. Brainstem Disorders, Vol., Springer. Urban, Z., et al., 2000. Elastin gene deletions in Williams syndrome patients result in altered deposition of elastic fibers in skin and a subclinical dermal phenotype. Pediatr Dermatol. 17, 12-20.

260

References

Valero, M.C., et al., 2000. Fine-scale comparative mapping of the human 7q11.23 region and the orthologous region on mouse chromosome 5G: the low-copy repeats that flank the Williams-Beuren syndrome deletion arose at breakpoint sites of an evolutionary inversion(s). Genomics. 69, 1-13. Van Borsel, J., Curfs, L.M., Fryns, J.P., 1997. Hyperacusis in Williams syndrome: a sample survey study. Genet Couns. 8, 121-6. van den Buuse, M., Simpson, E.R., Jones, M.E., 2003. Prepulse inhibition of acoustic startle in aromatase knock-out mice: effects of age and gender. Genes Brain Behav. 2, 93-102. van den Buuse, M., et al., 2005. Enhanced effect of dopaminergic stimulation on prepulse inhibition in mice deficient in the alpha subunit of G(z). Psychopharmacology (Berl). 183, 358-67. Van der Aa, N., et al., 2009. Fourteen new cases contribute to the characterization of the 7q11.23 microduplication syndrome. Eur J Med Genet. 52, 94-100. van der Valk, P.G., et al., 1985. Eczematous (irritant and allergic) reactions of the skin and barrier function as determined by water vapour loss. Clin Exp Dermatol. 10, 185-93. van Hagen, J.M., et al., 2007. Contribution of CYLN2 and GTF2IRD1 to neurological and cognitive symptoms in Williams Syndrome. Neurobiol Dis. 26, 112-24. van Nierop, B.J., et al., 2015. Assessment of Myocardial Fibrosis in Mice Using a T2*-Weighted 3D Radial Magnetic Resonance Imaging Sequence. PLoS One. 10, e0129899. Vanbokhoven, H., et al., 2011. p63, a story of mice and men. J Invest Dermatol. 131, 1196-207. Vandenberghe, M., et al., 2013. ORAI1 calcium channel orchestrates skin homeostasis. Proc Natl Acad Sci U S A. 110, E4839-48. Vandeweyer, G., et al., 2012. The contribution of CLIP2 haploinsufficiency to the clinical manifestations of the Williams-Beuren syndrome. Am J Hum Genet. 90, 1071-8. Virtanen, M., et al., 2003. Splice site and deletion mutations in keratin (KRT1 and KRT10) genes: unusual phenotypic alterations in Scandinavian patients with epidermolytic hyperkeratosis. J Invest Dermatol. 121, 1013-20. Volna, P., et al., 2004. Negative regulation of mast cell signaling and function by the adaptor LAB/NTAL. J Exp Med. 200, 1001-13. Vullhorst, D., Buonanno, A., 2003. Characterization of general transcription factor 3, a transcription factor involved in slow muscle-specific gene expression. J Biol Chem. 278, 8370-9. Wang, Y.K., et al., 1997. A novel human homologue of the Drosophila frizzled wnt receptor gene binds wingless protein and is in the Williams syndrome deletion at 7q11.23. Hum Mol Genet. 6, 465-72. Wang, Z., Gerstein, M., Snyder, M., 2009. RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet. 10, 57-63. Watt, F.M., Lo Celso, C., Silva-Vargas, V., 2006. Epidermal stem cells: an update. Curr Opin Genet Dev. 16, 518-24. Weber, D., Wiese, C., Gessler, M., 2014. Hey bHLH transcription factors. Curr Top Dev Biol. 110, 285-315. Weiner, L., et al., 2014. Skin as a living coloring book: how epithelial cells create patterns of pigmentation. Pigment Cell Melanoma Res. 27, 1014-31. Weinstein, G.D., McCullough, J.L., Ross, P., 1984. Cell proliferation in normal epidermis. J Invest Dermatol. 82, 623-8. Werner, S., Grose, R., 2003. Regulation of wound healing by growth factors and cytokines. Physiol Rev. 83, 835-70. Widagdo, J., et al., 2012. SUMOylation of GTF2IRD1 regulates protein partner interactions and ubiquitin-mediated degradation. PLoS One. 7, e49283. Williams, J.C., Barratt-Boyes, B.G., Lowe, J.B., 1961. Supravalvular aortic stenosis. Circulation. 24, 1311-8.

261

References

Willott, J.F., 1986. Effects of aging, hearing loss, and anatomical location on thresholds of inferior colliculus neurons in C57BL/6 and CBA mice. J Neurophysiol. 56, 391-408. Willott, J.F., 1996. Anatomic and physiologic aging: a behavioral neuroscience perspective. J Am Acad Audiol. 7, 141-51. Willott, J.F., Turner, J.G., 1999. Prolonged exposure to an augmented acoustic environment ameliorates age-related auditory changes in C57BL/6J and DBA/2J mice. Hear Res. 135, 78-88. Willott, J.F., 2001. Handbook of mouse auditory research: from behavior to molecular biology, Vol., CRC Press. Wilson, D.R., Maibach, H.I., 1980. Transepidermal water loss in vivo. Premature and term infants. Biol Neonate. 37, 180-5. Winter, R.M., 1996. What's in a face? Nat Genet. 12, 124-9. Wong, A.C., Birnbaumer, L., Housley, G.D., 2013. Canonical transient receptor potential channel subtype 3-mediated hair cell Ca(2+) entry regulates sound transduction and auditory neurotransmission. Eur J Neurosci. 37, 1478-86. Wyler, F., et al., 1973. Aortic aneurysms as a late complication in the syndrome of supravalvular aortic stenosis, peculiar facies and mental retardation. Am J Roentgenol Radium Ther Nucl Med. 119, 524-9. Xu, P.X., et al., 1999. Eya1-deficient mice lack ears and kidneys and show abnormal apoptosis of organ primordia. Nat Genet. 23, 113-7. Yaffe, M.B., Beegen, H., Eckert, R.L., 1992. Biophysical characterization of involucrin reveals a molecule ideally suited to function as an intermolecular cross-bridge of the keratinocyte cornified envelope. J Biol Chem. 267, 12233-8. Yamakawa, T., et al., 2007. Interaction of syntaxin 1A with the N-terminus of Kv4.2 modulates channel surface expression and gating. Biochemistry. 46, 10942-9. Yan, X., et al., 2000. Characterization and gene structure of a novel retinoblastoma-protein- associated protein similar to the transcription regulator TFII-I. Biochem J. 345 Pt 3, 749-57. Yang, J., et al., 2010. Fibroblast growth factor receptors 1 and 2 in keratinocytes control the epidermal barrier and cutaneous homeostasis. J Cell Biol. 188, 935-52. Yi, R., et al., 2008. A skin microRNA promotes differentiation by repressing 'stemness'. Nature. 452, 225-9. Yoshimura, K., et al., 2009. Distinct function of 2 chromatin remodeling complexes that share a common subunit, Williams syndrome transcription factor (WSTF). Proc Natl Acad Sci U S A. 106, 9280-5. Yoshimura, K., et al., 2014. Retraction for Yoshimura et al., Distinct function of 2 chromatin remodeling complexes that share a common subunit, Williams syndrome transcription factor (WSTF). Proc Natl Acad Sci U S A. 111, 2398. Young, E.J., et al., 2008. Reduced fear and aggression and altered serotonin metabolism in Gtf2ird1-targeted mice. Genes Brain Behav. 7, 224-34. Yun, Y.R., et al., 2010. Fibroblast growth factors: biology, function, and application for tissue regeneration. J Tissue Eng. 2010, 218142. Zhang, R., Maksymowych, A.B., Simpson, L.L., 1995. Cloning and sequence analysis of a cDNA encoding human syntaxin 1A, a polypeptide essential for exocytosis. Gene. 159, 293-4. Zhao, C., et al., 2005. Hippocampal and visuospatial learning defects in mice with a deletion of frizzled 9, a gene in the Williams syndrome deletion interval. Development. 132, 2917- 27. Zhu, M., et al., 2004. Positive and negative regulation of FcepsilonRI-mediated signaling by the adaptor protein LAB/NTAL. J Exp Med. 200, 991-1000. Zhu, Y., et al., 2013. Active cochlear amplification is dependent on supporting cell gap junctions. Nat Commun. 4, 1786.

262

References

Zong, H., et al., 2007. Cyclin D3/CDK11p58 complex is involved in the repression of androgen receptor. Mol Cell Biol. 27, 7125-42. Zou, D., et al., 2004. Eya1 and Six1 are essential for early steps of sensory neurogenesis in mammalian cranial placodes. Development. 131, 5561-72. Zou, D., et al., 2008. Eya1 gene dosage critically affects the development of sensory epithelia in the mammalian inner ear. Hum Mol Genet. 17, 3340-56.

263