UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Analysis of transcriptional targets of SOX9 during embryonic heart valve development reveals a critical… Garside, Victoria C. 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2016_february_garside_victoria.pdf [ 7.87MB ]
Metadata
JSON: 24-1.0220645.json
JSON-LD: 24-1.0220645-ld.json
RDF/XML (Pretty): 24-1.0220645-rdf.xml
RDF/JSON: 24-1.0220645-rdf.json
Turtle: 24-1.0220645-turtle.txt
N-Triples: 24-1.0220645-rdf-ntriples.txt
Original Record: 24-1.0220645-source.json
Full Text
24-1.0220645-fulltext.txt
Citation
24-1.0220645.ris

Full Text

ANALYSIS OF TRANSCRIPTIONAL TARGETS OF SOX9 DURING EMBRYONIC HEART VALVE DEVELOPMENT REVEALS A CRITICAL NETWORK OF TRANSCRIPTION FACTORS  by Victoria C. Garside  M.Sc., The University of Western Ontario, 2008 B.Sc., The University of Western Ontario, 2006  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in The Faculty of Graduate and Postdoctoral Studies  (Cell and Developmental Biology)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  November 2015  © Victoria C. Garside, 2015 ii  ABSTRACT Cardiac malformations affect approximately 1% of human newborns and a large number of these are due to defects in the heart valves and septum. It has been suggested that cardiac valve diseases, which make up about one third of all cardiovascular defects, arise from underlying developmental malformations that occur during embryogenesis. Interestingly, the development of the heart valves (cardiac cushions) and tissues that form cartilage templates (such as the limb) share a number of key TFs, such as TWIST1, SOX9, and NFATC1 suggesting that they have similar transcriptional programs. It has been proposed that regulatory networks involved in cartilage formation, are also active during valve development and disease. The transcription factor SOX9 has an essential role in heart valve and cartilage formation and its loss leads to major congenital abnormalities in the embryo. Regardless of this critical role, little is known about how SOX9 regulates heart valve development or its transcriptional targets. Therefore, to identify transcriptional targets of SOX9 and elucidate the role of SOX9 in the developing valves, we have used ChIP-Seq on the E12.5 atrioventricular canal (heart valves) and limb buds. Comparisons of SOX9 DNA-binding regions among tissues revealed both context-dependent and context–independent SOX9 interacting regions. Context-independent SOX9 binding suggests that SOX9 may play a role in regulating proliferation-associated genes across many tissues. Generation of two endothelial specific Sox9 mutants uncovers two potential roles for SOX9 in heart valve formation: first in the initial formation of valve mesenchyme and later in the survival and differentiation of valve mesenchyme. Analysis of tissue-specific SOX9-DNA binding regions with gene expression profiles from Sox9 mutant heart valves indicates that SOX9 directly regulates a collection of transcription factors known to be important for heart development. Taken together, this study identified that SOX9 controls transcriptional iii  hierarchies involved in proliferation across tissues and heart valve differentiation. SOX9 transcriptional targets identified in this data could be used as predictive factors of heart valve disease, or as targets for new therapeutic strategies for disease and congenital defects. iv  PREFACE  The embryonic material for the SOX9 Chromatin Immunoprecipitation coupled with deep sequencing (ChIP-Seq) libraries on the Embryonic day (E) 12.5 atrioventricular canal (AVC) and E12.5 limb buds was collected by R. Cullum and the ChIPs were also performed by R.Cullum. Library construction, sequencing, and initial bioinformatics for peak generation were performed by the Michael Smith Genome Sciences Centre. SOX9 peak to gene associations were carried out by R. Cullum (Appendix IID, Appendix III). Genomic locations of SOX9 peaks and SOX9 motifs were assigned by bioinformatic analyses completed by R. Cullum (Figure 3-3C, D, Table 3-1). ChIP-qPCR validation of SOX9 ChIP-Seq peaks was executed by R. Cullum (Figure 6-1C, D). Z-scores for the SOX9 ChIP-Seq libraries were calculated by M. Bilenky (Appendix IA, B). The LacZ staining on VE-Cre:LacZ mouse E10.5 and E11.5 hearts was performed by ACY Chang (Figure 2-2B).  All animal protocols were approved by the UBC Animal Care and Ethics Committee (protocols: A12-0305 and A12-0297). The UBC Biosafety Committee approved the use of any biohazardous chemicals and material.  Portions of this thesis are modified from: Victoria C. Garside, Alex C. Chang, Aly Karsan, and Pamela A. Hoodless (2013) Co-ordinating Notch, BMP, and TGF-β signaling during heart valve development. Cellular and Molecular Life Sciences 70(16):2899-917. Chapter 1 Introduction text and Figure 1-1 and 1-2.  Victoria C. Garside, Rebecca Cullum, Olivia Alder, Daphne Y. Lu, Ryan Vander Werff, Mikhail Bilenky, Yongjun Zhao, Steven J. M. Jones, Marco A. Marra, T. Michael Underhill, Pamela A. Hoodless (2015) SOX9 directly modulates the expression of key transcription factors required for heart valve development. Submitted. Chapter 3,5,&6 including Figures 3-3, 5-2A, B, 5-8A, B, 6-1, 6-5A, B, and 6-6.  v  TABLE OF CONTENTS  Abstract....................................................................................................................................ii Preface.....................................................................................................................................iv Table of contents.....................................................................................................................v List of tables.........................................................................................................................viii List of figures.........................................................................................................................ix List of abbreviations..............................................................................................................xi Acknowledgements...............................................................................................................xiii Dedication...............................................................................................................................xiiv Chapter one: Introduction.........................................................................................................1 1.1 General introduction.......................................................................................................1 1.2 Heart valve formation.....................................................................................................1 1.2.1 Embryonic development of the heart valves...................................................1 1.2.2 Composition of the adult heart valve...............................................................5 1.2.3 Major signalling events during heart valve development................................6 1.3 Embryonic limb development and shared features of limb and heart valve formation.............................................................................................................................11 1.3.1 Development of the limbs.............................................................................11 1.3.2 Functional similarities between the developing heart valves and limbs............12 1.4 The role of the Sex Related Y (SRY) box like 9 (SOX9) during embryonic development........................................................................................................................13 1.4.1 SOX transcription factors..............................................................................13 1.4.2 The many roles of SOX9 during development..............................................15 1.4.3 The role of SOX9 in proliferation, EMT, and ECM.....................................16 1.5 SOX9 is essential for the development of the heart valves and limb...........................18 1.6 Campomelic dysplasia..................................................................................................21 1.7 Congenital heart defects and heart valve disease.........................................................22 1.7.1 Congenital heart valve abnormalities............................................................22 1.7.2 Adult heart valve disease...............................................................................23 1.7.3 The involvement of SOX9 and signaling pathways in heart valve disease...24 1.8 Hypothesis and aims of the thesis project.....................................................................25 Chapter two: Materials and methods......................................................................................27 2.1 Mice strains and tissue dissection.................................................................................27 2.2 Immunofluorescence, cell counts, in situ hybridization, and H&E staining................31 2.3 RNA isolation...............................................................................................................33 2.4 RT-PCR, qRT-PCR, and ChIP-qPCR..........................................................................34 2.5 Chromatin immunoprecipitation coupled with high-throughput sequencing...............34 2.6 Bioinformatic analysis of ChIP-Seq.............................................................................38 2.7 Genotyping Sox9 mutant embryos..................................................................................38 vi  2.8 RNA-Seq and bioinformatic analysis..............................................................................39 2.9 Cell culture, transfection, cloning and luciferase assays.................................................41 2.10 Site directed mutagenesis..............................................................................................42 Chapter three: Identification and characterization of SOX9 binding sites in the developing heart valve and limb genome.....................................................................................................43 3.1 Sox9 mRNA and protein is enriched in the mouse heart valves throughout development..........................................................................................................................43 3.2 SOX9 directly binds thousands of DNA regions in the developing heart and limb......................................................................................................................................48 3.3 SOX9 binds active promoter regions within the genome...............................................51 3.4 Identification of potential co-factors of SOX9................................................................53 3.4.1 DNA motif analysis identifies numerous potential co-factors of SOX9............53 3.4.2 Comparison of SOX9 ChIP-Seq with published ChIP-Seq data sets reveals new insights into potential co-factors of SOX9..................................................................58 Chapter four: Characterization of the Sox9fl/fl;VE-Cre mice.................................................62 4.1 SOX9 negative AV mesenchyme cells are absent in Sox9 mutant valves......................62 4.2 SOX9 is maintained at later stages of heart valve development in Sox9 mutants..........67 4.3 Adult Sox9 mutants have ventricular hypertrophy and thickened valves leaflets...........70 Chapter five: Characterization and analysis of the Sox9fl/fl;Tie2-Cre mice...........................75 5.1 SOX9 negative mesenchyme was detected as early as E10.5 in the AVC.....................75 5.2 Deletion of Sox9 was variable in the Sox9 cKO heart valves.........................................79 5.3 Proliferation defects in the Sox9 cKO valves..................................................................83 5.4 Global transcriptional alterations in the Sox9 cKO heart valves.....................................87 Chapter six: SOX9 has functions involved in regulation of proliferation, transcriptional networks, and ECM formation during heart valve development...............................................94 6.1 SOX9 occupies regulatory regions of genes associated with proliferation.....................94 6.2 Context independent SOX9 binding regions in the AVC, limb, and HF-SCs................96 6.3 Proliferation associated target genes are both activated and repressed by SOX9...........99 6.4 SOX9 targets a network of TFs known to be involved in heart development..............103 6.5 Additional known roles for SOX9 in EMT and ECM organization.............................111 Chapter seven: Discussion.....................................................................................................115 7.1 Thesis overview............................................................................................................115 7.2 SOX9 occupies the promoter and upstream regulatory regions associated with thousands of target genes including its future co-factors....................................................116 7.3 SOX9 negative mesenchyme was absent in the Sox9fl/fl;VEC heart valves................122 7.4 Confirmation that Sox9 cKO (Sox9fl/fl;Tie2-Cre) valves have decreased size and proliferation that leads to major heart defects and embryonic death..................................125 7.5 Identification and regulation of SOX9 target genes in the developing heart valves.....128 7.6 SOX9 directly controls genes associated with proliferation across cell types..............129 7.7 SOX9 is a master regulator of a core network of TFs in heart valve development......133 7.8 Examination of other defined roles of SOX9 in EMT and ECM generation................135 vii  7.9 Concluding remarks......................................................................................................137 References...............................................................................................................................140 Appendices..............................................................................................................................153 Appendix I SOX9 immunostaining on E12.5 whole embryo and E10.5 heart...................153 Appendix II Quality verification of the SOX9 ChIP-Seq libraries and gene associations.........................................................................................................................154 Appendix III Characteristics of SOX9 ChIP-Seq libraries.................................................155 Appendix IV GO analysis on E12.5 AVC and limb overlapping SOX9 peaks..................156 Appendix V GO analysis on E12.5 AVC SOX9 peaks......................................................160 Appendix VI GO analysis on E12.5 limb SOX9 peaks......................................................164 Appendix VII Co-factor analysis on E12.5 AVC SOX9 peaks using oPPOSSUM...........167 Appendix VIII Co-factor analysis on E12.5 limb SOX9 peaks using oPPOSSUM...........170 Appendix IX Co-factor analysis on E12.5 AVC and limb overlapping peaks using oPPOSSUM.........................................................................................................................173 Appendix X Comparisons of SOX9 peaks in AVC and limb with SMAD3 and TWIST1 peaks....................................................................................................................................176 Appendix XI Top 100 differentially expressed genes in the Sox9 cKO AVC RNA-Seq library..................................................................................................................................177 Appendix XII Genes with altered expression (>1.5FC down) with a SOX9 peak in the AVC....................................................................................................................................180 Appendix XIII Genes with altered expression (>1.5FC up) with a SOX9 peak in the AVC....................................................................................................................................184 viii  LIST OF TABLES  Table 2-1 Primer Sequences......................................................................................................28 Table 2-2 Antibodies for immunostaining.................................................................................32 Table 3-1 SOX9 monomer and dimer binding sites under the SOX9 peaks in heart and limb............................................................................................................................................52 Table 4-1 Genotypes from Sox9fl/fl;+/+ and Sox9fl/+;VE-Cre/+ mouse crosses to generate SOX9 mutant embryos...............................................................................................................71 Table 5-1 Genotypes from Sox9fl/fl;+/+ and Sox9fl/+;Tie2-Cre/+ mouse crosses suggest that Sox9 cKO embryos die after E13.5............................................................................................80 Table 5-2 The top ten SOX9 targets with altered gene expression in the Sox9 cKO............................................................................................................................................91 Table 6-1 The top 15 Biofunctions identified by IPA on the genes with overlapping SOX9 peaks in the AVC, limb, and HF-SCs........................................................................................97 Table 6-2 The top 20 Biofunctions identified by IPA on SOX9 target genes with differential expression in the Sox9 cKO AVC...........................................................................................104  ix  LIST OF FIGURES  Figure 1-1 Mouse heart valve development.................................................................................3 Figure 1-2 Major signalling pathways during heart valve development......................................8 Figure 2-1 Generation of SOX9 ChIP-Seq libraries from E12.5 AVC and E12.5 limb buds............................................................................................................................................36 Figure 2-2 Generation of the Sox9fl/fl;VE-Cre and/or Sox9fl/fl;Tie2-Cre mice........................40 Figure 3-1 Sox9 mRNA is enriched in the developing heart valves..........................................44 Figure 3-2 SOX9 protein is enriched in the heart valves during heart development.................47 Figure 3-3 Comparison of SOX9 initiated transcriptional programs in developing limb and heart and genomic peak locations..............................................................................................50 Figure 3-4 GO analysis reveals tissue specific functions for genes associated with SOX9 binding sites...............................................................................................................................54 Figure 3-5 UCSC Genome browser screen shots of shared SOX9 binding locations in the developing heart and limb that are associated with cell proliferation and/or cell cycle............55 Figure 3-6 SOX9 peaks are enriched for numerous potential co-factor binding sites...............57 Figure 3-7 Comparison of SOX9 peaks in the AVC and limb with other SOX9 ChIP-Seq libraries and other potential co-factor TFs.................................................................................59 Figure 4-1 SOX9 is not lost in the developing heart valves of the Sox9fl/fl;VE-Cre mice despite valve abnormalities........................................................................................................64 Figure 4-2 Sox9fl/fl;VE-Cre mice have reduced numbers of SOX9 + valve cells and decreased total valve cell numbers.............................................................................................................66 Figure 4-3 Additional sources of SOX9+ mesenchyme found in the developing heart at E12.5..........................................................................................................................................68 Figure 4-4 SOX9 positive cells can still be detected at later stages of valve development in Sox9fl/fl;VE-Cre mice................................................................................................................69 Figure 4-5 Ventricular abnormalities in the Sox9fl/fl;VE-Cre postnatal mice hearts when compared to WT (Sox9fl/fl;+/+)................................................................................................72 Figure 4-6 Hematoxylin and eosin staining of 10 month old adult WT and Sox9fl/fl;VE-Cre AV heart valves..........................................................................................................................73 Figure 5-1 SOX9 deletion occurs in the Sox9fl/fl;Tie2-Cre  heart valves as early as E10.5 in the mouse...................................................................................................................................77 Figure 5-2 The loss of SOX9 in the Sox9fl/fl;Tie2-Cre E12.5 heart valves leads to major valve abnormalities and reduced valve cell numbers..........................................................................78 Figure 5-3 The absence of SOX9 in the Sox9fl/fl;Tie2-Cre heart valves causes embryonic lethality between E13.5-14.5.....................................................................................................81 Figure 5-4 Sox9 deletion by Tie2-Cre is extremely variable in the Sox9 cKO heart valves.........................................................................................................................................82 Figure 5-5 Verification of Sox9 deletion by Tie2-Cre in the Sox9 cKO heart valves................84 Figure 5-6 Proliferation is reduced in the Sox9 cKO valves compared to WT..........................85 x  Figure 5-7 Cyclin D1 immunostaining suggests that SOX9 is required to exit S phase during cell cycle....................................................................................................................................86 Figure 5-8 Comparison of differential transcripts in the Sox9 cKO AVC identified by RNA-Seq and SOX9 ChIP-Seq reveals a key subset of genes important for heart valve formation....................................................................................................................................88 Figure 5-9 SOX9 selectively targets a subset of AV-enriched genes in the AVC.....................90 Figure 5-10 Sox9 cKO valves have major changes in gene expression when compared to WT valves.........................................................................................................................................93 Figure 6-1 Common SOX9 targets among developing tissues provide evidence for a role in proliferation................................................................................................................................95 Figure 6-2 SOX9 targets genes involved in proliferation of cells.............................................98 Figure 6-3 SOX9 regulates genes associated with cycle and proliferation..............................101 Figure 6-4 SOX9 ChIP-Seq and differential transcripts in the Sox9 cKO AVC comparison reveals critical transcriptional networks involved in valve formation.....................................105 Figure 6-5 Sox9 activates transcription factors that are known to be essential for heart valve development.............................................................................................................................107 Figure 6-6 The critical EMT regulator Twist1 is reduced in the Sox9 cKO heart valves........109 Figure 6-7 EVI1 is enriched in the AVC specifically in condensing mesenchyme and lost upon deletion of Sox9........................................................................................................................110 Figure 6-8 EMT is unaffected by the loss of Sox9 in the developing heart valves…...….......112 Figure 6-9 ECM molecules mRNA levels are reduced in the Sox9 cKO heart valves............114   xi  LIST OF ABBREVIATIONS  A- Atria AV- Atrioventricular AVC- Atrioventricular canal BAV- Bicuspid aortic valve BMP-Bone morphogenetic protein BSA- Bovine serum albumin CD- Campomelic Dysplasia ChIP-Seq- Chromatin immunoprecipitation coupled with high-throughput sequencing CPC- Cardiac progenitor cells Cre- Cre recombinase DABCO- 1,4-diazabicyclo[2.2.2]octane DAPI- 4',6-Diamidino-2-Phenylindole, Dihydrochloride Dll- Delta-like ligand E- Embryonic day ECM- Extracellular matrix EMT- Epithelial-to-mesenchmyal transition EndMT- Endothelial-to-mesenchymal transition Epi- Epicardium FDR- False discovery rate FPKMs- Fragments per kilobase of exon per million reads GO- Gene ontology HF-SCs- Hair follicle stem cells HMG- High mobility group IF- Immunofluorescence ISH- In situ hybridization MSCs- Mesenchymal stem cells MVP- Mitral valve prolappse NC- Neural crest NICD- Notch intracellular domain OC- Ovarian cancer OFT- Outflow tract PBS-Phosphate buffered saline PCR-Polymerase chain reaction PFA- Paraformaldehyde PHF- Primary heart field pHH3- phospho histone H3 PWM- Positional weight matrix qPCR- Quantitative PCR xii  RT-PCR- Reverse transcriptase PCR TF- Transcription factor TGFβ- Transforming growth factor beta TSS- Transcriptional start site TTS- Transcription termination site SHF- Secondary heart field SL- Semilunar SOX9- SRY (sex determining region Y) related box 9 Sox9 cKO- Sox9fl/fl;Tie2-Cre Sox9 mutant- Sox9fl/fl;VE-Cre V- Ventricles VC- Vertebral column VE- Vascular endothelial cadherin VEC- Valvular endocardial cells VIC- Valvular interstitial cells WT- Wildtype xiii  ACKNOWLEDGEMENTS  Foremost, I would really like to thank my family especially my parents who always supported and encouraged me throughout my entire career both emotionally and financially. I would not be where I am if it was not for you! I would really like to thank my dad for opening my eyes to science at a young age and helping me think critically. Thank you to my mum who enlightened my creative side and always dared me to think outside the box. In addition, I would like to thank my husband Colin Prior for always being there for me through the good times and the bad times in research and in life. He always knows how to make me smile even when I find it hard to. Next, I would like to thank my supervisor Dr. Pamela Hoodless for being an excellent supervisor and guiding me through the world of research. She taught me a lot about critical thinking, experimental design, and what is involved in pursuing an academic career. I am forever grateful for all of her help. Additonally, I would like to thank my fellow work collegues and lab mates. The lab environment has greatly enriched my scientific experience during my PhD. I would especially like to thank Rebecca Cullum and Olivia Alder for all of their help, advice and support during my PhD. I am thankful for all of the scientific discussions and collaborations that we have had over the years and they have really helped me to grow as a scientist and as a person. I am thankful for all of the help that the people of the Terry Fox Labs have given to me over the years and thank you to all of you that have made this a very fun experience as well. I would like to thank to my supervisory committee for their excellent advice and comments throughout my PhD. I am sure that our meetings and discussions have made me think much more critically about my research. A final thanks to everyone I have met and worked with along the way. Every little bit brought me here today. xiv  DEDICATION I would like to dedicate this dissertation to my family. I love you all very much.   1  CHAPTER ONE: Introduction 1.1 General introduction Transcription factors (TFs) are essential for the precise coordination of lineage specification and differentiation, and ultimately the control of cell identity. Developing an understanding how TFs regulate and integrate many cellular processes during development has been a major focus of embryology. Recent advances in sequencing technologies have prompted the use chromatin immunoprecipitation coupled with deep sequencing (ChIP-Seq) to determine the DNA binding sites of TFs on a genome-wide level in numerous tissues. With the discovery of technologies like ChIP-Seq and/or transcriptome analysis (RNA-Seq) researchers are one step closer to uncovering the regulatory networks that govern the cellular processes in embryonic development. However with each technological advance forward in the field, the complexity of these relationships and networks deepens. Transcriptional networks play a major role in guiding the correct development of the embryonic heart valves and many TFs are essential to this process during development. Aberrant expression of TFs has been linked to a number of congenital heart valve defects and adult heart valve diseases and therefore, a more complete picture of how these TFs function during normal heart valve development and adult valve maintenance would be highly valuable for the innovation of novel therapeutics and establishment of additional biomarkers of disease.  1.2 Heart valve formation 1.2.1 Embryonic development of the heart valves The heart is one of the first organs to form and function within the developing embryo (1). It delivers sufficient oxygen and nutrients throughout and establishes proper blood flow in the 2  fetus. The first step of heart formation is specification of cardiac progenitor cells (CPCs), which takes place prior to their ingression through the primitive streak during gastrulation (embryonic day (E) 6-7.0 in mouse). Subsequently, CPCs undergo an epithelial-to-mesenchymal transition (EMT) and migrate out from the primitive streak to create the left and right heart fields (E7.5). The heart fields will move laterally and fuse on the midline of the anterior side of the embryo. Fusion of the two heart fields forms the cardiac crescent at E8.0 (reviewed in (2)). The differentiation of the cells within the cardiac crescent produces the endocardial and myocardial progenitor cells and together they make up the primary heart field (PHF). The PHF will give rise to the left ventricle, atrioventricular canal (AVC), parts of the right ventricle and atria. The secondary heart field (SHF), a secondary population of progenitor cells originating from the splanchnic mesoderm which are located anterior to the PHF, will contribute to parts of the right ventricle and atria and the outflow tract (OFT). The two arms of the cardiac crescent fuse along the embryonic midline forming the linear heart tube (E8.5) (2). The linear heart tube consists of a monolayer of endothelium and several layers of myocardium at E9.0 (Figure 1-1) and loops rightward to form the chambers of the heart. The process of rightward looping brings the chambers into their final positions in the mature heart. Two constrictions appear in the looped heart, the AVC, between the atria and ventricle, and the OFT, between the ventricle and great arteries (Figure 1-1) and these regions will eventually form the mature valves and contribute to the septa of the heart (additional information on early heart formation in (2-4)).  The septum divides the heart into four functional chambers and valves function to ensure uni-directional blood flow. During heart valve development, the AVC and the OFT will form four sets of heart valves: two sets of atrioventricular (AV) valves and two sets of  3     Figure 1-1 Mouse heart valve development. Valve development begins at E9.5 as the atrioventricular canal (AVC) endocardial cells undergo endothelial-to-mesenchymal transformation (EMT) to create the valve mesenchyme. Following EMT, the valve mesenchyme undergoes remodeling and differentiation to generate the adult heart valve leaflets. Cells depicted in green are endocardial cells, and cells in orange are valve mesenchyme. AVC – atrioventricular canal, RA – right atria, LA – left atria, RV – right ventricle, LV – left ventricle, PA – pulmonary artery. 4  semilunar (SL) valves, respectively. The AV valves are made up of the mitral valve, which regulates blood flow from left atrium to left ventricle; and the tricuspid valve, which prevents blood backflow between right atrium and right ventricle. The two SL valves are the aortic valve, which regulates blood flow from the left ventricle into the aorta; and the pulmonary valve, which regulates blood flow between the right ventricle and pulmonary artery (5).  Heart valve formation is initiated through locally increased production of cardiac jelly in the AVC and OFT. The cardiac jelly is comprised of extracellular matrix (ECM) secreted by the myocardium into the interstitial space between the endocardium and myocardium and creates swellings known as the cardiac cushions. Although the cushions are initially acellular, endocardial cells overlying the cushion undergo EMT (also referred to as EndMT (endothelial-to-mesenchymal transition)) to indicate the endothelial origin of the cells) to form mesenchymal cells (6). In the mouse, EMT begins at approximately E9.5 in the AVC and E10.5 in the OFT (Figure 1-1). During EMT, endocardial cells lose cell-cell junctions and cell polarity, transition into mesenchymal cells, and acquire a migratory phenotype (7). The mesenchymal cells invade the ECM and populate the cardiac cushions (8,9). The OFT, which will form the SL valves, develops similarly with the exception that neural crest cells migrate and contribute to the OFT cardiac cushions (reviewed in (10)).  Following invasion of the mesenchyme into the cardiac cushion, these cells will proliferate, differentiate and remodel to form thin delicate valve leaflets and septal structures of the mature heart (11,12). The cardiac cushions form the valves and septa through two major steps: remodelling and maturation (E10.5-adult), and elongation (E14.5-adult). Valve remodelling can be divided into a number of overlapping steps: proliferation and expansion of mesenchyme cells (E10.5-E12.5), differentiation of mesenchyme cells (E12.5-E16.5), and valve 5  condensation and maturation (E15.5-adult) (11,13-15). To date, the majority of research in the field has concentrated on the initial stages of the EMT process since AVC explant cultures allows for robust measurement of EMT (7). In contrast, our understanding of valve differentiation, maturation and condensation are currently lacking. This may be due to the absence of an established culture model system to examine the later stages of valve development.  1.2.2 Composition of the adult heart valve The adult heart valve leaflets are highly organized structures composed of three stratified layers: the atrialis in AV valves or ventricularis in SL valves, spongiosa and fibrosa which are mainly composed of elastin, proteoglycans, and collagens, respectively (12). In AV valves, the fibrosa layer is located on the ventricular side of the valve whereas in SL valves, it is located away from the ventricle. The fibrosa layer maintains strength and integrity of the valve (16). The atrialis and ventricularis face the blood flow and provide the flexibility of valve (17). The spongiosa, the middle layer of the valve, acts as a sponge and allows for valve compression to absorb the pressures from blood flow. The heart valve leaflets are enclosed in a sheath of valvular endocardial cells (VECs) with valvular interstitial cells (VICs) dispersed throughout the valve leaflet. VICs are descendants of the mesenchymal cells found in the cardiac cushions during embryogenesis. Lineage tracing studies in mice using Tie2, which is expressed in endocardial cells prior to EMT, shows that the bulk of cells present in the valves after birth are derived from endocardium with the exception of the AV parietal leaflets (18-20). It was shown that epicardial-derived cells start to migrate into the lateral AV cushions at E12.5 and selectively contribute to parietal leaflets of the mouse AV valves (20). VICs play an important role in maintaining proper 6  valve homeostasis but can be aberrantly activated during valve disease (21). More recent research is now focused on how VICs are generated during development, how they maintain adult homeostasis, and how they become activated during disease progression.  A number of studies have suggested that crucial developmental signalling pathways, involved in normal embryonic valve formation, such as transforming growth factor beta (TGFβ), bone morphogenetic protein (BMP), and Notch, are also activated during heart valve disease (3,22). Thus, advancing our understanding of how these signalling pathways function and interact during heart valve development will provide key insights into mechanisms of adult heart valve disease.  1.2.3 Major signalling events during heart valve development Many signalling pathways are implicated in the formation of the cardiac valves however three critical signalling pathways are required for early specification and initiation of EMT in the cardiac cushions. To create the appropriate environment for EMT in the cushions, BMPs from the myocardium signal to the overlying endocardium. For the initiation of EMT, Notch signalling is also required and together with BMP and TGFβ signalling pathways as they synergize to transform the endothelial cells into mesenchyme and promote mesenchymal cell invasiveness. Together, these three crucial signalling pathways create the cardiac cushions and populate them with mesenchyme cells, setting off the cascade of events required to form mature heart valves and septa.  BMPs and TGFβs are part of the TGFβ superfamily. This family comprises over 30 ligands that can be categorized into several subgroups: activins/inhibins, nodals, BMPs, growth and differentiation factors, Müllerian inhibiting substance and TGFβs. To activate signalling, the 7  ligands bind to a tetrameric, transmembrane receptor complex that contains two type I and two type II receptors. In mammals, there are five distinct type II receptors and seven type I receptors, which form specific combinations that dictate ligand binding specificity. The receptors will phosphorylate and activate an intracellular canonical signalling pathway that is mediated by receptor-regulated Smad proteins (R-Smads). Following phosphorylation, R-Smads interact with the binding partner, Smad4, and move into the nucleus where they interact with DNA-binding proteins to regulate transcription of TGFβ superfamily responsive genes (23,24) (Figure 1-2A).  There are two major phases where BMP signalling is essential for valve formation. First, BMP signalling sets up a permissive environment that allows endocardium to become activated, and secondly, together with both Notch and TGFβ signalling promotes EMT and mesenchyme invasion. Together, these two roles of BMP signalling ensure mesenchyme growth, survival, and eventually leading to valve remodelling. The majority of the mouse models for BMP signalling result in defects during early cardiac cushion formation (reviewed in (22)), and therefore, the essential role of BMP signalling in the early stages of cardiac cushion formation are well documented. However, due to early lethality in mouse models the role of the BMP signalling molecules during valve remodelling, differentiation, and adult homeostasis is less understood. Emerging data suggests that BMP signalling may be abnormally activated during valve disease. To corroborate this, BMP2 is increased in calcified regions in diseased valve leaflets (25,26) and there are higher levels of BMP signalling in fibrosa endothelium of human diseased aortic valves (27). This suggests that BMP signalling may be involved in the process of calcification during adult valve disease and that there may be additional roles for BMP signalling in the later stages 8    Figure 1-2 Major signalling pathways during heart valve development. A. TGFβ superfamily signalling. TGFβ signalling initiates as aTGFβ ligand binds to the heterodimeric receptor (containing twoTGF βR1 and twoTGF βR2) leading to activation of SMAD2/3. SMAD2/3 then binds to its co-factor SMAD4 and enters the nucleus to activate TGFβ responsive genes. BMP signaling functions similarly where a BMP ligand binds to the receptor (BMPRII/ALK2/3/6) and activates SMAD1/5/8. SMAD1/5/8 binds to SMAD4 and moves to the nucleus to activate BMP responsive genes B. Notch signalling pathway begins as a signal sending cell expressing Jag binds to Notch on a signal receiving cells which leads to a cleavage event that releases Notch intracellular domain (NICD). Following this, NICD moves to the nucleus and together with its co-factors activates Notch responsive genes. 9  of embryonic valve formation and in adult valve maintenance.  TGFβ signalling plays an essential role in the initial promotion and cessation of EMT, and in cushion mesenchyme proliferation and differentiation during heart valve development. Tissue-specific knockout mouse models suggest that the TGFβ receptors have very diverse yet specific roles dependent on the tissue and time point in which they are expressed (reviewed in (22)). This makes it difficult to determine the exact role of TGFβ signalling molecules in developing heart valves. Moreover, early lethality of many of the mouse models precludes our understanding of the potential roles of these signalling components in later stages of valve remodelling and differentiation. The use of inducible knockout systems will be highly valuable in teasing out the exact roles of TGFβ signalling components during heart development. Furthermore, TGFβ signalling has a significant role in maintaining adult heart health by regulating cardiac fibrosis and hypertrophy after injury and in hypertension (23,28). In a normal adult valve, VICs, the main cellular component, are quiescent and maintain the integrity of the valve leaflet. Following injury, TGFβ signalling activates VICs (29), sustains VIC activation and regulates in vitro valve repair via activated VICs (30). Persistent activation of VICs can alter the mechanical properties of the valve through changes in ECM composition and thus increasing susceptibility to disease. TGFβ signalling has been associated with a number of valve diseases (31-33) and suggests that aberrant activation/inhibition of this pathway during embryonic development may increase the probability of valve disease later in life. To aid in the discovery and design of new potential therapeutics for congenital heart defects and valve disease additional studies are required examining the role of TGFβ signalling in later stages of heart valve development and adult VIC activation. 10   Notch signalling is involved in numerous developmental events and processes including formation of the heart valves. Notch signalling is an essential driver of EMT and its components are widely expressed throughout heart valve development (22). Activation of Notch requires the binding of a transmembrane Notch ligand on a signalling cell to a transmembrane Notch receptor on a signal-receiving cell. In mammals, there are four Notch receptors, Notch 1-4, and five Notch ligands, Delta-like (Dll) 1, 3, 4 and Jagged (Jag) 1, 2. Activation of Notch signalling has three major steps: ligand binding, release of the Notch intracellular domain (NICD) via two proteolytic cleavages of the Notch receptor, and finally translocation of NICD into the nucleus to function as a transcription factor along with its binding partners RBPJ and MAML (Figure 1-2B).  To examine the functions of Notch signalling during valve development, transgenic mouse models have been generated for downstream, intracellular effectors, such as NICD, RBPJ and MAML. Of note, not all of these Notch intracellular effectors are specific to only Notch signalling. The loss of Notch intracellular effectors can lead to a complete block in Notch signalling, while overexpression of NICD leads to constitutive activation of Notch signalling. This allows examination of effects in the presence or absence of Notch signalling. Gain of function experiments with constitutive endocardial Notch activation using NICD leads to the activation of a mesenchymal gene program in the ventricular endocardium and ventricular explants have the ability to undergo a non-invasive EMT and upon addition of BMP2, ventricular explants can undergo a full invasive EMT (34). This data indicates that Notch signalling plays an important role in endocardial patterning of the AVC and chambers of the heart and that BMP2 has a role in inducing invasive EMT. However, constitutive activation of NICD may not recapitulate normal in vivo activities of Notch signalling and it should be noted 11  that some phenoptypes may be an artifact. Conversely, absence of RBPJ causes a loss of cushion mesenchyme in valve regions, EMT defects, and collapsed endocardium in the developing heart (35). Overall, the data suggests that Notch activation and signalling via RBPJ is essential for the endothelial cell lineage and EMT in the cardiac cushions. Mouse Notch receptor knockout studies reveal that Notch1 is essential for cardiac valve formation. Notch ligand mutant mice do not show abnormalities in cardiac valves, suggesting that redundancy may play a role for Notch ligands during valve formation. Additionally, the loss of Notch target genes has further emphasized the significance of Notch signalling during cardiac development. A number of the complete and endothelial specific mouse knockout studies for Notch signalling result in embryonic lethality at E10.5 (reviewed in (22)) and suggests that Notch signalling has a critical role in early phases of cardiac valve formation following EMT. However, the roles of Notch signalling in later valve development and in the adult valve disease are not fully understood. Notch signalling has also been implicated in adult heart valve calcification (36) and valve diseases share many similarities with cartilage and bone formation in the developing limbs (37,38).  1.3 Embryonic limb development and shared features of limb and heart valve formation 1.3.1 Development of the cartilage template in the developing limb The development of the bones of the vertebrate skeleton is derived through two different processes: intramembranous ossification, which forms flat bones; and endochondral ossification, which forms long bones (39). Endochondral ossification requires the formation of a cartilage template prior to bone formation whereas intramembranous ossification does not form a cartilage template and forms directly from mesenchymal condensations (reviewed in (40)). During the 12  process of endochondral ossification, the undifferentiated mesenchyme cells will migrate into regions that are fated to become bone. These cells will become packed in this locale (without proliferating) triggering condensation, which signifies the start of cartilage formation. Undifferentiated mesenchyme secretes a primitive ECM enriched with collagen type I, tenascin, hylauronan, and fibronectin (41) but as condensation commences, the mesenchyme in the core differentiate and alter the composition of the ECM. To date, little is known about what drives the process of mesenchymal condensation, however, it is known that BMPs are critical for the formation of chondrogenic condensations (39). The differentiated mesenchyme in the core of the condensation develop into chondrocytes (cartilage cells) and secrete ECM rich in collagen type II, IX, and XI, aggrecan and link protein (41). Chondrocytes start proliferating, then undergo further differentiation, and turn into hypertrophic chondrocytes. During this time, these cells reduce production of collagen type II and start to secrete collagen type X (41). Hypertrophic chondrocytes enlarge, terminally differentiate, and exit cell cycle to become terminally differentiated chondrocytes. Simultaneously, blood vessels start to invade the cartilage template and bring in the osteoblast cells (bone forming cells). Following this, the terminally differentiated chondrocytes undergo apoptosis to allow the osteoblasts to fill this area and replace cartilage with bone (reviewed in (42)). Any alterations in the process of endochondral ossification lead to major abnormalities of the skeleton such as achondroplasia and osteochondrosis.  1.3.2 Functional similarities between the developing heart valves and limbs The composition of the developing limb cartilage is similar to that of developing heart valves, where both tissues are largely composed of highly organized ECM and mesenchymal cells that 13  undergo proliferation, condensation and remodeling to form heart valves or limb buds. Moreover, heart valve and tissues that form cartilage templates (such as the limb) share a number of key TFs, such as TWIST1, SOX9, and NFATC1 (43) suggesting that their transcriptional programs are comparable (37,38,44). Furthermore, heart valves and tissues made up of cartilage have functionally similar tissue composition in that they both differentiate from SOX9 positive mesenchyme and produce comparable ECM structures. To support the similarity seen in the transcriptional programs of developing heart and limbs, loss of the transcription factor, SOX9, prior to mesenchymal condensation during limb formation leads to agenesis of the cartilage and bone (45) and loss of SOX9 in the heart valve endothelium results in major abnormalities in the heart valves cauding embryonic death (46). This suggests that SOX9 is essential for heart valve and limb development and that SOX9 plays a key role in their formation, differentiation and organization.  1.4 The role of the sex determining region Y (SRY) box 9 (SOX9) during embryonic development Signalling pathways play a crucial role during development and initiate a cascade of events that can lead to the activation of TFs that coordinate gene expression. The SRY (sex determining region Y) box (SOX) TFs play a key role in the formation of numerous organs during development and receive signals from many signalling pathways such as TGFβ, BMP and Notch. 1.4.1 SOX transcription factors The SOX TF family is subdivided into groups (A-H) based on their sequence similarity in the high mobility group (HMG) domain that is required for DNA binding (reviewed in (47)). The 14  SOXA subgroup only has one member, SRY, while the SOXB group is further divided into SOXB1 containing SOX1, 2, and 3 and SOXB2 containing SOX14 and SOX21. The SOXC subgroup contains SOX4, 11, and 12 and the SOXD group is comprised of SOX5, 6, and 13. SOXE includes SOX8, 9, 10 and SOXF has SOX7, 17, 18. Lastly, SOXG and SOXH only have one member each, SOX15 and SOX30 respectively. SOXs have been shown to bind to the DNA sequence ATTGTT or similar motifs via the HMG domain (reviewed in (47)) and it has been suggested that the HMG domain’s interaction with the DNA itself can result in DNA bending. This DNA bending capability may be another regulatory role for SOX factors during development although this exact function remains to be shown in vivo.  SOX factors require the binding of other additional co-factor TFs on the nearby DNA and form TF complexes to efficiently regulate their target genes (47). Of note, the SOXD subgroup contains a coiled-coil dimerization domain that allows these factors to bind with other SOXD proteins. The SOXE TFs consist of a self-dimerization domain that could aid in the formation of these TF complexes with itself and other partner proteins. In addition, different SOX subgroups can work together to form TF complexes to regulate their target genes. For example, during chondrogenesis SOX9 can dimerize with itself to regulate genes (Col2a1) but it also works together with SOXD factors, SOX5 and SOX6, as a SOX trio to regulate chondogenic genes like Aggrecan (48). During successive developmental events, several SOX TFs along with their binding partners are known to bind and regulate their future co-factors. Two examples of these types of interactions are found during melanocyte development. For example, SOX10 binds together with PAX3 to activate the Mitf gene, following this activation MITF then binds with SOX10 to activate other melanocyte-specific genes (49,50). The next example illustrates when a SOX factor activates another SOX factor and then both SOX factors work together to regulate 15  additional downstream genes. In sex determination, SRY and its binding partner SF1 activate SOX9. Following that, SRY and SOX9 make a co-factor complex that regulates the Amh gene which is essential for male gonad development (51). Additionally, SOX proteins within the same subgroup often have similar functions and are expressed in the same developing tissues, although exact expression patterns of each individual SOX factor can vary within the tissue (47). For example, all of the SOXE TFs have roles involved in neural crest cell development and SOXF factors have important functions in the developing vasculature (47). This suggests that there may be some level of redundancy between SOX factors within a subgroup. However this is not always the case, as it has been demonstrated that one member of a SOX factor subgroup may be more critical than the others for a specific developmental process or function (reviewed in (47)).  1.4.2 The many roles of SOX9 during development The SOXE TFs have important roles in gene regulation during the development of numerous organ systems, in sex determination, and in the neural crest (reviewed in (47,51,52). In particular, SOX9 plays a critical role in the development of testis, pancreas, limb, heart, intestine, liver, and many others (reviewed in (53)). In humans, mutations in the SOX9 locus that generate haploinsufficency cause a disorder called campomelic dysplasia (CD) which includes skeletal abnormalities, sex reversal, and in some cases heart defects (54) for further details on CD see Section 1.6). Loss of one allele of Sox9 in mice is lethal at perinatal stages and mutant embryos have defects such as cleft palate and bending and hypoplasia of cartilage-derived skeletal structures resembling the phenotype of CD in humans. To bypass this, Sox9 and its functions during development have been examined using the Cre/Lox system to delete specifically within different subsets of cells. It has been described that SOX9 has three general 16  roles during development: proliferation of progenitor cells, EMT, and in ECM differentiation including ECM organization and deposition (53).  1.4.3 The role of SOX9 in proliferation, EMT, and ECM The role of SOX9 in progenitor cell proliferation has been described in the developing pancreas where SOX9 is expressed in all early pancreatic epithelial cells and is subsequently down regulated upon commitment (reviewed in (53)). Pancreata from CD patients are hypoplastic, consistent with a potential role in proliferation (55). Mouse studies where SOX9 was specifically deleted in the developing pancreas also produced severe hypoplasticity (56), confirming the importance of SOX9 in the maintenance of pancreatic progenitor fate and proliferation. Intriguingly, numerous SOX9 mutant organs are hypoplastic and have defects in proliferation (46,57-59) and yet to date, no direct SOX9 target genes involved in regulating proliferation or cell cycle have been identified in any of these systems. The direct relationship between SOX9 and proliferation has been somewhat tenuous but several studies have attempted to draw a more mechanistic link between the two. For instance, in rat mesenchymal stem cells (MSCs), a stable knockdown of SOX9 caused reduced proliferation, increased levels of cyclin D1 and p21, delayed S-phase progression, and increased stability of the cyclin D1 protein (60). This suggests that SOX9 plays an essential role in the progression of cell cycle through S-phase via degradation of cyclin D1 in rat MSC. Not only does SOX9 have a role in proliferation in progenitor/stem cells but it has been shown to have important functions in stem cell maintenance (61,62). In adult hair follicle stem cells, genes bound by SOX9 are required to maintain stemness via secreted factors in the niche (62) and regulate chromatin dynamics at super enhancers (63). 17   Moreover, altered levels of SOX9 have been associated with many cancers (53,64), and altered SOX9 levels are linked to poor prognosis, increased proliferation and invasiveness in many cancers. In breast cancer, SOX9 is associated with more aggressive cancer subtypes (65) and in colorectal cancer, SOX9 promotes tumour growth and progression (64). The majority of studies demonstrating a link between SOX9 and proliferation have suggested that SOX9 positively affects proliferation. On the contrary, SOX9 has also been shown to have a suppressive role on proliferation in the intestinal epithelium and SOX9 deficient crypts have increased levels of proliferation (66). These differences in the proliferative role of SOX9 in the intestinal epithelium may be due todifferences in expression levels in different subsets of cells within the crypts. Overall, this shows that SOX9 is an extremely context-dependent TF and therefore this complexity must be taken into account when trying to understand the function of SOX9 in a given tissue. Overall, SOX9 plays a critical but context-specific role in maintaining appropriate levels of proliferation and sustaining cell cycle progression during development, in stem/progenitor cells, and in cancer.  SOX9 has also been implicated in the process of EMT in neural crest (NC) cells (reviewed in (53)). During development, NC cells undergo EMT and migrate out from the dorsal neural tube to populate multiple regions within the embryo. NC cells contribute to a diverse number of tissues such as the cartilage, skin and the heart. Premigratory and migratory NC cells express SOX9 and loss of function studies demonstrate that SOX9 is important for neural crest cell formation (53). SOX9 has been shown to activate transcription factors like Snai2 and together they can induce EMT (67). Several other SOX9-expressing cells undergo EMT-like processes such as astrocytes, pancreatic progenitors, cartilage, and cardiac valves and further support the notion that SOX9 may be important for the process of EMT during development. The process of 18  EMT is central to cancer metastasis and SOX9 may be playing an important role in driving cell invasiveness via EMT.  The regulation of ECM organization and differentiation in chondrogenesis is one of the well known functions of SOX9 (68). SOX9 is highly expressed as the condensing mesenchyme is becoming committed to chondroprogenitors and the expression of SOX9 is not down regulated until they become terminally differentiated chondrocytes (68). The ECM environment is constantly changing during cartilage formation and SOX9 contributes to this process by regulating genes like the collagens (Col2a1, Col9a1, Col11a2, Col27a1) and key matrix proteins (Aggrecan, Matrilin-1 and COMP) in chondrocytes (53). The composition of the ECM is also important for the formation of the developing heart valves where SOX9 is highly expressed in the valve mesenchyme (following EMT) throughout their development and suggests that SOX9 has a necessary role in regulating ECM components during valve formation.  1.5 SOX9 is essential for the development of the heart valves and limb. SOX9 has been shown to have a crucial role during heart valve development (46,69). SOX9 is highly expressed in the developing mouse endocardial cushions (E9.5-E16.5) within the mesenchyme cells (69-71) and in the remodelling and maturing heart valve leaflets (E16.5-after birth)(46,72). Sox9 null embryos could be generated by conditionally deleting one allele of Sox9 in oocytes (using Zona pellucid 3 (Zp3)-Cre) and one allele in spermatids (using Protamine 1 (Prm1)-Cre) followed by crossing these mice together to obtain completely null embryos (69). Embryos that completely lack Sox9 die at E11.5-12 due to congestive heart failure and have severely hypoplastic endocardial cushions and authors suggest that SOX9 may play a role in the proliferation and differentiation of the endocardial cushions during EMT (69). Using a LacZ 19  reporter mouse to examine Sox9 expression, Akiyama et al. found that Sox9 turns on in newly transformed mesenchymal cells that have migrated into the cardiac cushion (69) and it has been shown that its expression is directly downstream of Notch signaling in endothelial cells transduced with NICD (73). Endothelial Sox9 deletion using the Tie2-Cre driver strain (which is a receptor tyrosine kinase that is expressed in all endothelial cells and its descendents including the valve mesenchyme) with the Sox9 flox mouse was embryonic lethal just prior to E14.5 with hypoplastic endocardial cushions, reduced mesenchymal cell proliferation and alterations in ECM composition (46). These data indicate that SOX9 expression is turned on in the mesenchyme (the valve precursor cells) and is involved in proliferation of these cells during early valve formation. Additionally, SOX9 has been demonstrated to have an essential role in late heart valve formation, where the loss of SOX9 using Col2a1-Cre, which deletes Sox9 in the fibrosa layer of the valve, causes abnormal ECM patterning, loss of cartilage-associated proteins and thickened valve leaflets (46). These findings demonstrate that SOX9 has multiple roles during cardiac valve formation, initially during the expansion of the valve precursors and later for the proper expression and distribution of the ECM in the valve leaflets.  The loss of one allele of Sox9 in mice using Col2a1-Cre promotes calcification of the heart valve leaflets, increased matrix remodelling and inflammation of the heart valves suggesting that SOX9 may also play a role in heart valve disease (46,72). In addition, SOX9 has been associated with valve calcification in human aortic valve disease (36,74,75) and in mouse valve disease models (76,77). Therefore, understanding the role of SOX9 and its downstream target genes during heart valve development may give us better insights into the progression of cardiac valve disease and may lead to the development of novel therapeutics. 20   SOX9 is essential for chondrogenesis (78) and is highly expressed in mesenchmyal condensations, maintained in differentiated chondrocytes but is eventually silenced at terminal differentiation in hypertrophic chondrocytes in order for proper bone formation to occur (68). SOX9 is known to bind to the DNA sequence WWCAAWG in promoters and enhancers of cartilage-specific genes (79-84) and can bind and regulate DNA as monomers and dimers (85,86). The loss of SOX9 before mesenchymal condensation (using Prx1-Cre) in the developing limb buds leads to cartilage agenesis and subsequent bone formation although patterning of the limb axes remains intact (45). The loss of SOX9 prior to mesenchmyal condensations produces an expanded apoptotic domain in the limb and indicates a potential role for SOX9 in suppression of apoptosis. In addition, SOX5 and SOX6 (known transcriptional targets and co-factors of SOX9 in the limb) expression was lost in the mutant limbs (45). Furthermore, deletion of SOX9 after mesenchymal condensation formation (using Col2a1-Cre, which allows the initial formation of chondrocytes) leads to chondrodysplasia. Cells failed to fully differentiate into chondrocytes, and had decreased proliferation resulting in abnormal joint formation (45). SOX9 limb mutant mouse models demonstrate that SOX9 is essential for the formation of cartilage and subsequent development of bone in the developing limb buds.  Surprisingly, there are only a few genes known to be directly transcriptionally regulated by SOX9 and most of these “target” genes have been identified in the developing limb. Some of the identified cartilage-specific targets of SOX9 are Col2a1, Col9a1, Col11a2, Acan, Hapln1 (CLP), Comp, and Mia1 (Cd-rap) (79-84,87,88) with Col2a1 being the most well characterized to date. Although a few genes have been speculated to be SOX9 targets in the developing heart valves, none have been shown to be direct. One study demonstrated that SOX9 represses Spp1 in maturing heart valves and chondrocytes to inhibit matrix mineralization (89). Moreover, SOX9 21  knockdown in primary valve explants lead to an increase in Spp1 transcripts and a decrease in Col2a1 and Hapln1 (cartilage-associated genes) whereas an over-expression of SOX9 leads to the inverse. This suggests that Spp1 expression is required for matrix mineralization following the loss of SOX9 in primary heart valve explants (89). This may indicate why the loss of SOX9 can lead to calcification as seen in mice with heterozygous loss of SOX9 using Col2a1-Cre. In summary, SOX9 plays an essential role in the formation of the heart valves and limb buds and identification of the global downstream transcriptional targets of SOX9 during heart valve and limb development will provide a better understanding of the exact role of SOX9 during the expansion, differentiation, remodelling and maturation of the valves and limbs and in congenital defects and disease.  1.6 Campomelic dysplasia Campomelic dysplasia (CD) is characterized by the bowing of long bones, facial abnormalities, pelvic and vertebral defects, hypoplastic scapulae, clubbed feet, absence of a set of rib bones, and sex determination defects (54). Overall, CD is usually lethal but there is a small subset of patients that survive. Acampomelic CD, a variant of CD is also due to mutations in the SOX9 locus and has all of the same clinical phenotypes except the bowing of the long bones. Interestingly, mild versions of CD tend to be associated with rearrangements within the chromatin, potentially disrupting regulatory elements, rather than alterations in the Sox9 coding region (85). This indicates that mild forms of CD are likely due a decreased dosage of SOX9 during development. A mutation that disrupts the dimerization domain of SOX9 was identified in a CD patient that did not have sex reversal and suggests that the dimerization capability of SOX9 is necessary for chondrogenesis and not for sex determination (85). Furthermore, loss of 22  one allele of Sox9 in mice results in perinatal lethality with defects resembling the phenotype of CD (90). The mouse model of CD indicates that dosage of SOX9 is critical at two major steps during cartilage formation: mesenchymal condensation and the transition of differentiated chondrocytes to hypertrophic chondrocytes (90). Since SOX9 takes a central role in multiple abnormalities found in CD these mice can be used as a model to understand how the loss of one TF affects the formation numerous organs during development.  1.7 Congenital heart defects and heart valve disease 1.7.1 Congenital heart valve abnormalities Congenital abnormalities of the heart can affect 1-5% of human newborns and approximately a third of these cardiac defects are due to heart valve malformations (91,92). The most prevalent heart valve malformations are bicuspid aortic valve (BAV), in which patients exhibit two aortic valve cusps instead of three; and mitral valve prolapse (MVP), in which patients suffer from floppy mitral valve leaflets that can slip past their normal position into the left atrium. Reports on the prevalence of congenital heart defects vary widely (ie. 1-5% of newborns) since BAV and MVP are not usually included. BAV affects 2% of the general population but the consequences are rarely seen until adulthood (91,92) whereas MVP affects up to 5% of the general population but is rarely detected in newborns as symptoms are frequently not severe (91). However, these congenital anomalies can lead to an increased pressure on the heart and predispose these patients to valve disease later in life. Tetralogy of Fallot, hypoplastic left heart syndrome, and Ebstein's anomaly are other examples of congenital heart defects that include heart valve and septal abnormalities. There are a number of syndromes that also include heart valve and septal defects such as Williams, Marfan, Trisomy 21, Alagille, Turner, and Noonan syndrome (92). 23  1.7.2 Adult heart valve disease According to the World Health Organization, cardiovascular diseases are the number one cause of death worldwide; in 2008 approximately 17.3 million people died from cardiovascular disease, which accounts for 30% of global deaths. In the US, approximately 3-5% of cardiovascular deaths are due to valve disease (93). Adult valve disease can become evident as stenosis, a narrowing of the valve opening resulting in less blood flow; or as regurgitation, an incomplete closure of the valve causes backflow of blood in the heart. If valve disease goes undiagnosed it can lead to secondary effects, such as improper ventricular function and eventually heart failure. Initial stages of heart valve disease involve activation of VICs, which leads to abnormal ECM deposition and disorganization. There are two types of ECM changes that can occur in the heart valves during valve disease: myxomatous disease involves increased deposition of proteoglycans, loss of collagen, and destruction of elastin fibrils leading to “floppy” valves and regurgitation, while fibrotic disease involves degradation of proteoglycans along with increased levels of collagen and elastin fibre fragmentation resulting in stiffening of the valve leaflets known as valve stenosis (11,94,95). Valve fibrosis can often progress further, leading to valve calcification (3). Accumulation of calcium in the valve causes them to stiffen, which impacts the valve’s ability to open and close and can eventually lead to heart failure. To date, little is known about the progression of valve calcification, myxomatous and fibrotic valve disease.  Many cases of valve disease in adults involve pre-existing defects in the heart valves and suggest that abnormalities that occur during embryonic valve development may lead to susceptibility to valve disease later in life (3,96-99). The literature suggests that key pathways and factors regulating valve development are also implicated in valve disease and congenital 24  heart defects (37,74). Unfortunately, there are limited options for treatment of valve disease and the current treatment option is valve replacement using either mechanical or prosthetic valves (100); however, a major fault with replacement surgery is that additional surgeries are often required. Consequently, the search for alternate treatments of heart valve disease drives the need for further investigation of heart valve development and disease and their underlying biology.  1.7.3 The involvement of SOX9 and signaling pathways in heart valve disease Sox9 expression has been shown to be up-regulated in human myxomatous mitral valve disease (31,76). Scleraxis mutant mouse AV valves displayed characteristics of human valve disease and it was found that both SOX9 and ECM proteins were up-regulated in the mutant valves (101). Heart valves from mice with a mutation in Col1a2 (Osteogenesis imperfecta murine mice) share a number of characteristics with myxomatous valve disease including increased expression of Sox9 (77). In addition, SOX9 has been linked to human aortic valve stenosis and pediatric and adult aortic valve disease (74,75). In mouse, heterozygous loss of Sox9 using a flox allele with Col2a1-Cre resulted in thickened heart valve leaflets and deposits of calcium on the valves similar to the calcification seen in heart valve disease (46,72). Generally, the literature suggests that the loss of SOX9 in the heart valves results in calcification, however in human and mouse diseased aortic valves, SOX9 was found to be up-regulated along with RUNX2, a known osteogenic marker that is indicative of calcification (75,77). To date, up-regulation of Sox9 has been implicated in both myxomatous and calcific valve diseases which have very different disease processes and reveals conflicting roles for SOX9 in the progression heart valve disease and consequently additional studies are required to tease out the role of SOX9 in valve disease. 25   Signalling pathways also play a key role in the progression of heart valve disease. For example, Notch has been implicated in valve calcification, a common form of heart valve disease (36), and mutations in NOTCH1 have been linked with familial, non-syndromic, autosomal-dominant calcific aortic valve disease (CAVD) and associated with BAV (102). To further support a role for Notch signalling during valve disease, heterozygous Notch or Rbpj mice have a higher risk of developing calcification of the aortic valve (103,104). Inhibition of Notch in rat aortic valve interstitial cells causes a significant decrease in SOX9 (already known to be involved in aortic valve calcification) and cartilage-associated factors (36). Using an in vitro valve calcification model, loss of Notch lead to high levels of calcification and conversely, increased levels of Notch signalling reduced calcification (36). Of note, the addition of SOX9 could rescue the accelerated calcification triggered by the loss of Notch (36) and demonstrates that Notch signalling may be regulating SOX9 in VICs. This data suggests that the loss of both Notch and SOX9 in the cardiac valves could lead to adult valve disease and subsequent calcification of the valves. Therefore, understanding the role of Notch signalling and the role of SOX9 during valve formation may provide key insights into their involvement during valve disease.  1.8 Hypothesis and aims of this study My hypothesis is that SOX9 regulates the same critical genes in both heart and limb development and by identifying unique and shared SOX9 target genes I will provide novel insights into similarities and differences between valve and limb regulatory networks and will determine genes essential for heart valve formation. To explore the in vivo functional role of SOX9 in valve and limb development, I have analyzed and characterized the transcriptional 26  targets of SOX9 identified by chromatin immunoprecipitation coupled with deep sequencing (ChIP-Seq) on E12.5 mouse embryonic AVC and limb. Briefly, ChIP-Seq is a technique that will identify TF DNA binding sites on a genomic level within a given tissue (105,106). The ChIP pulls down (enriches) DNA sequences that SOX9 antibodies bind to and these are sequenced and mapped back to the genome. Regions of the genome that are enriched with DNA sequences are called ‘peaks’ and these represent SOX9 DNA binding regions within the genome. This method allows for the identification of in vivo transcriptional targets of SOX9 in developing valve and limb on a genome wide level.  To identify the critical of role of SOX9 in the heart valves, I have generated and characterized the phenotype of a novel endothelial specific (Vascular Endothelial cadherin (VE) Cre crossed with the Sox9 flox mice) SOX9 mutant mouse and the previously described endothelial specific SOX9 mutant mouse (Tie2-Cre crossed with the Sox9 flox mice). After the proper characterization of the SOX9 mutant mice, I examined gene expression alterations in the absence of SOX9 by comparing WT and SOX9 mutant heart valves using RNA-Seq. To determine the critical targets of SOX9, I have employed various bioinformatic analyses to compare the SOX9 ChIP-Seq libraries to the RNA-Seq libraries generated from WT and SOX9 mutant heart valves. Lastly, I have selected several candidate genes from the list of genes discovered by comparing ChIP-Seq and RNA-Seq libraries based on differential expression and gene function for further examination in the developing valves. Collectively, this work will help to elucidate novel and key factors involved in the regulatory networks required for heart valve formation, and improve our understanding of heart development and heart-related disease processes. 27  CHAPTER TWO: Material and methods  2.1 Mice strains and tissue dissection All animal protocols were approved by the Animal Care Committee at the University of British Columbia. C57BL/6J mice were used for all ChIP-Seq libraries. ICR outbred mice were used for ChIP-qPCR validation. To generate Sox9fl/fl (WT) and Sox9fl/fl;VE-Cre or Sox9fl/fl;Tie2-Cre/+ (Sox9 cKO) embryos, Sox9fl/fl female mice were crossed with VE-Cre/Tie2-Cre male mice to generate Sox9fl/+;VE/Tie2-Cre male mice and then these males were mated with Sox9fl/fl female mice. All mice were backcrossed more than seven generations. Sox9fl/fl (B6.129S7-Sox9tm2Crm/J) and Tie2-Cre (B6.Cg-Tg(Tek-cre) 12Flv/J) mice were obtained from Jackson Laboratories (Sacramento, USA). The VE-Cre mice were a kind gift from the Karsan lab. The Sox9fl allele was genotyped using primers specific to the flox region and VE/Tie2-Cre was detected using primers specific to Cre recombinase (Table 2-1). Embryos from timed matings were considered embryonic day (E) 0.5 at noon of the day a vaginal plug was observed. For the SOX9 ChIP-Seq libraries, the atrioventricular canal (AVC) and limb buds were manually dissected in cold phosphate buffered saline (PBS) using forceps and a dissecting microscope. Whole hearts and heart regions including the AVC, atria, ventricles, and outflow tract were manually dissected from Sox9fl/fl and Sox9fl/fl;VE-Cre or Sox9fl/fl;Tie2-Cre/+ embryos for immunofluorescence, qPCR, in situ hybridization and/or RNA-Seq libraries. Littermates were used for experiments comparing WT and Sox9 cKO and at the least an n of 3 was used for all experiments. Sox9fl/fl, Sox9fl/fl;VE-Cre and Sox9fl/+;Tie2-Cre mice were maintained on a C57BL6 background.   28  Table 2-1 Primer sequences  Primers for Genotyping   Name Forward 5'to 3' Reverse 5' to 3' Sox9fl/fl AGACTCTGGGCAAGCTCTGG GTCATATTCACGCCCCCATT Cre CGTACTGACGGTGGGAGAAT CCCGGCAAAACAGGTAGTTA Primers for qRT-PCR Name Forward 5'to 3' Reverse 5' to 3' Sox9 CAGCAAGAACAAGCCACACGTCAA TTGTGCAGATGCGGGTACTGGTCT Fgfr2 TAGTCATGGCTGAAGCAGTGGGAA TCTGATACCAGATCAGACAGGTCC Trp53 ACAAGAAGTCACAGCACATGACGG TTCCTTCCACCCGGATAAGATGCT Akt2 GGAGGTCATGGAGCATAGATTC AAGTACCTTGTGTCCACTTCTG Prkaca ATGTAGCTGGTGGCGAGATGTTCT TGAGAAGATTCTCGGGCTTCAGGT Cdkn1b AAACTCTGAGGACCGGCATTTGGT TCTTCTGTTCTGTTGGCCCT Rfwd3 AAAGTGCCATCTTCCAAAGCCCAG AGGTGGCCAGATAGCTGTTCTGAT Junb CTTTAAAGAGGAACCGCAGACCGT TTTGATGCGCTCCTGGTCTTCCAT Hdac2 TACAACAGATCGCGTGATGACCGT TCCCTTTCCAGCACCAATATCCCT Sox4 GCCCGACTTCACCTTCTTT AAGGACAGCGACAAGATTCC Mecom/Evi1 TCCTCCTCATCCAACAACACCTCA ATCCGCAATTTCATCGGGAACAGC Pitx2 GGAAGCCACTTTCCAGAGAA CGGCGATTCTTGAACCAAAC Hand2 CACCAGATACATCGCCTACCTCAT GGTCTTCTTGATCTCCGCCTTGAA Nfia GTCACAAACACCAATAGCTGC AGACTTGAGGCGCTTTGTAG Vim GACCTTGAACGGAAAGTGGA AGCCACGCTTTCATACTGCT Cdh2  AGGGTGGACGTCATTGTAGC CTGTTGGGGTCTGTCAGGAT Cdh1  ACGACACAGCCAATGGACCAAGAT TCGGGCATATACTCCTGCAGTGTT Cdh5  ACCGAGAGAAACAGGCTGAA AGACGGGGAAGTTGTCATTG Postn TGTGTATCGGACGGCTATCT CTCTGCTGGTTGGATGATTTCT Eln CTCATCCATCCATCCATCCATC GACAGGTGAACCAGGTTGATAG Fbn1 ATGAATGCAACCAGGCTCCCAAAC AGCTCCTTCCATCCTCTTGCAGAA Mgp ACCCTGTGCTACGAATCTCACGAA TGTTGATCTCGTAGGCAGGCTTGT β-actin CCAGAGCAAGAGAGGTATCCTGAC CATTGTAGAAGGTGTGGTGCCAG    29  Primers for luciferase vector cloning Name Forward 5'to 3' Reverse 5' to 3' Mecom CAGAAGGTTCTAGAAGCAAGGCAC CCCAGGTATCCAGTACAAAGTAAATTATCA    Prkaca TGTCTCCGCCAATCGACGGCTT TGCTTCCCGCGTCTCTCT Rfwd3 AGAAGAAGCAGGGAGCAGCCTCA CCTGAGATTAAAGGCGTGAGCCAT Trp53 ACTAGCGGTGCTAGCCAGAAGTAT ACGATGTCTGGTGCGCGATAAGAA Junb AGAAACAGGCTAGGGAAAGAGAGC TTCCTGTGCCCTAATATGGGTGCT Nfia TGCAAACGATGAGGGTACTGGACT CCGTCCAGCTCCAGTGAATAATGGAA Fgfr2 TCTCTGCCCTTTGTCCTTTG TCTGTGTGCTTCCATGTTCC Cby1/Fam227a CATTTTTAACACAACAAAGC GCCTGCAGCTTTCTGCAAGG Wasf1/Cdc40 TCATCTTCCCTCATTCCCGAGCC ACCCGCGCCTGCAGCAGGGGGA Primers for site directed mutagenesis Name Forward 5'to 3' Reverse 5' to 3' Hdac2 1st M GGGGCTAAAGTCCGCTTGTGCGCACCTCCG GCGGACTTTAGCCCCGCGCTCAGAGACCCG Hdac2 2nd M GGGGCTAAAGTCCGCCGGTGCGCACCTCCG GCGGACTTTAGCCCCTTGCTCAGAGACCCG Hdac2 DM GGGGCTAAAGTCCGCCGGTGCGCACCTCCG GCGGACTTTAGCCCCGCGCTCAGAGACCCG Flanking Primers CGAGCTCTAGACTGCCCGGGATTCG GGCTAGCGACGGCCGGTGCTGCAGC Primers for Taqman qRT-PCR Name Company, Catalog #  Sox9 ABI, Mm00448840_m1  Gapdh ABI, Mm99999915_g1  Lef1 ABI, Mm00550265_m1  Twist1 ABI, Mm00442036_m1  Tbx20 ABI, Mm00451515_m1  Ccnd1 IDT, Mm.PT.58.28503828  Primers for ChIP-qPCR Cops5 AAACACTTCCTTAGGGTTGGCTCG CTCGCAGTTCACACGAACGGATTT Eed GGGAGGAAAGAGAAGTCACCT TAACTCGAAGTTGTTCTCCCGAGC Fos ATCTCCGAATCCTACACGCGGAA CCGTCTTGGCATACATCTTTCACC Hdac1 TGGGCCTGTACCAAAGTCCG TGTGAAGCGGGCTGCAGAGTTTA Hdac2 AACCAGTGCGCGTAAGACCGA TTCTACGGGTAGTCACACACAGTC Primers for ChIP-qPCR Name Forward 5'to 3' Reverse 5' to 3' Junb CCAGCTACAGACGCTTCTAGTCAT AGGCTTATTAGTCGCCGATGGTTG 30     Primers for ChIP-qPCR Name Forward 5'to 3' Reverse 5' to 3' Srpk2 GCTACAACACAAACACGAAATGCT GGACAAATCAAATAGAAGCAGCCAGGG Timp2 CCATTCACGTGCCGCTGAATCATTTG TGGCTTCTTGGCATAGAAACTGCG Ctgf TAAGCATGCACTGCTCACTCCAGA AAACACATGGCAGCTCCCTAAAGC Mecom/Evil TGTCCAGGGCATGCTTCTGACTAA CTGGCTGCTAAACCTGCTTACACA Fgfr2 ACCTCTGTGCGACGCAGGAAATAA TGGCAGGAGACACGGAACTAAACA Id3 TTCCCACACATTCGCCATCAAAGC CCACAACAGATTAAAGACCAGGAGGG Trp53 GCAGGAGCATTTCCGGTTTCTTGT TCCTCAAACCACAGAACCAGCCTA Prkaca TCACTCATCCAGGAGCCTATGGT ATTGGTTCAGGCCTCCCTGACTGA Rfwd3 ACGGCGACCAATCTCTTCTCTTCT CTCGCCCAGAAATGTATCAAAGGC Cul4a TACTGGTTAATGGTGATGTGCGCC AACCAACGTGCAGAGGTTACCGT Fgf11 GGGACTCCCTAACTGTCGT AACAGGGAAATCGGCAGAGAGCAA Col2a1 AGAGCTGTGAATCGGGCTCTGTAT AGGCTGTGCATTGTGGGAGA Apoc3 CGTGAAAAGCATGGGCAAATC AGGGATAAAACTGAGCAGGC Hnf1 CATGAGGCCTGCACTTGCAA GGGAAATTCTCCAAGGTTCA Tat GAGTCAGGCTTCAAATCTCTGGTC GGGAAATTCTCCAAGGTTCA Wasf1 AAAGTGCGGATCGGGCAATACC AGAATGAGCATCCACCCACTTACC Wasf2 AAGTTCATAGGCTCGGCCTGTTCT ACAGGGAAAGACCTCGGCTAACAT   31  2.2 Immunofluorescence, cell counts, in situ hybridization and H&E staining Hearts were fixed in 4% paraformaldehyde (PFA, Sigma) overnight and were subjected to a sucrose gradient (15%, 30%, 65%) prior to embedding. Hearts were embedded in TissueTek O.C.T. (Sakura) in cryomolds and were cryosectioned using the Leica CM3050S cryostat at 6-8μm thickness. For in situ hybridization (ISH) slides were sectioned at 10 μm thickness. Slides were maintained at -80˚C until immunostaining. Following removal of slides from the -80˚C freezer, slides were dried for 10 minutes at room temperature. Borders were drawn around sections with a wax pen (Diagnostic Biosystems) and were re-fixed with 4% PFA for 10-20 minutes at room temperature. Slides were washed with 1X PBS three times for 5 minutes each. A blocking solution of 5% Bovine Serum Albumin (BSA, Roche) with 0.1% Triton X-100 (Sigma) diluted in PBS was added to the sections for 1 hour at room temperature. Primary antibodies were diluted in blocking solution (see antibody dilutions in Table 2-2) and placed into a humidified chamber overnight at 4˚C. The following day, the primary antibody was removed and washed with 1XPBS three times for 5 minutes at room temperature. Secondary antibodies were diluted in blocking solution at 1:500 (see Table 2-2 for secondary antibodies) and incubated for one hour at room temperature. Sections were washed with 1X PBS three times for 5 minutes each. 4',6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI, 1 µg/mL, Sigma) was added to the last PBS wash for 10 minutes at room temperature to label nuclei. DAPI was washed off with PBS. Slides were mounted with 50-100 µL of 20 mg/mL DABCO (1,4-diazabicyclo[2.2.2]octane, Sigma) and coverslips were added and sealed with nail polish for visualization. Images were captured with OpenLab v5.0 (PerkinElmer) on a Zeiss Axioplan 2 compound microscope or TCS SP5 Leica confocal microscope with the Leica Application suite software. For immunofluorescence images, cell counts for SOX9 positive nuclei or phospho 32  Table 2-2 Antibodies for immunostaining  Primary antibodies NAME HOST SPECIES COMPANY & CATALOG DILUTION SOX9 rabbit Millipore  1/600 PHOSPHO HISTONE H3 (PHH3) rabbit AbCam 1/100 CYCLIND1 mouse Santa Cruz 1/100 EVI1 rabbit Santa Cruz 1/100 PERIOSTIN rabbit Abnova 1/100 CD31 rat BD Biosciences 1/100 Desmin Mouse Medicorp 1/100 Secondary antibodies NAME HOST SPECIES COMPANY & CATALOG DILUTION ALEXA FLUOR 488 rabbit ThermoFisher Scientific,  1/500 ALEXA FLUOR 594 rabbit ThermoFisher Scientific, 1/500 ALEXA FLUOR 488 mouse ThermoFisher Scientific,  1/500 ALEXA FLUOR 594 mouse ThermoFisher Scientific, 1/500     33  histone H3 (pHH3) positive nuclei in the AVC and/or the eipcardium (epi) were performed on three to five sections and averaged between at least three WTs and three Sox9 cKOs hearts (N=3) using ImageJ software or by manually counting. ISH was performed as described previously(107). Hemotoxylin and Eosin (H&E) staining (Sigma) was performed according to the manufacturer’s protocols.  2.3 RNA isolation Atrioventricular canals (AVCs), atria, ventricles and outflow tracts (OFT) from ICR or C57BL6 or single AVCs, atria, ventricles and OFT (generated from crossing Sox9fl/fl females and Sox9fl/+;Tie2-Cre or Sox9fl/+;VE-Cre male mice) were dissected out from the heart of embryos. Tissues from ICR and C57BL6 were pooled together and for Sox9/Cre crosses only single tissues were used for RNA isolation. Tissues were directly placed into Trizol (ThermoFisher Scientific). For tissues from Sox9/Cre crosses, the samples were stored at -80˚C until embryos were genotyped (See section 2.7). RNA was isolated with Trizol using the manufacturer’s protocol. Briefly, tissues were homogenized in Trizol and incubated for 5 minutes at room temperature. Chloroform (EMD Biosciences) was added at 1/5 the volume of Trizol and vigorously agitated for 20 seconds. Samples were ccentrifuged at high speed (>13000rpm) for 10 minutes at 4˚C. The aqueous layer was taken and RNA was precipitated with an equal volume of isopropanol (EMD Biosciences). Samples were centrifuged down at high speed (>13000 rpm) at 4˚C to collect precipitates for at least 10 minutes. Isopropanol was removed and pellets were washed with 70% ethanol (EMD Biosciences). The ethanol was removed and samples were air dried for 10 minutes at room temperature. Pellets were re-suspended in RNase and DNAse free water (ThermoFisher Scientific). The quality of the RNA was assessed using the Nanodrop 1000 34  (RNA with approximately OD260/280= 1.8-2.0, OD260/230= 2.0-2.2 was considered good quality) and RNA was stored at -80˚C.  2.4 RT-PCR, qRT-PCR, and ChIP-qPCR RT-PCR was performed as previously described (108) with a several modifications. cDNA was synthesized using the Transcriptor First Strand cDNA Synthesis Kit (Roche) and HIFI Taq polymerase (ThermoFisher Scientific) was used for RT-PCR and for initial steps of cloning. qPCR was performed with FastStart Universal SYBR Master (sRox, Roche) according to the manufacturer’s protocol on the ABI 7900HT Fast Real-Time PCR System. Conditions for qPCR were as described (109). Primers used in RT-PCR, qRT-PCR and ChIP-qPCR are found in Table 2-2. Relative quantification was used for qRT-PCR. Housekeeping genes were used as endogenous controls for all Taqman assays (Gapdh) and qRT-PCR (Actb). For single heart part specific qRT-PCR analysis on AVC, ventricles and atria, mouse Sox9, Lef1, Twist1, Tbx20 and Gapdh Taqman primers and probes (ThermoFisher Scientific) were used on WT and Sox9 cKO embryonic heart parts. For ChIP-qPCR, fold enrichment was calculated by 2^(Ct difference between SOX9 and IgG ChIP). Regions tested for enrichment in ChIP-qPCR were within SOX9 peak regions (see Table 2-1 for primer sequences. If multiple peaks were present in heart or limb for shared target genes then the most likely candidate was chosen based on Positional Weight Matrix (PWM) score and proximity to the gene.  2.5 Chromatin immunoprecipitation coupled with high-throughput sequencing (ChIP-Seq) Three independent ChIPs generated from pooled embryos were combined for SOX9 ChIP-Seq libraries (total: 164 E12.5 AVCs and 39 E12.5 limb buds). For a simplistic illustration of the 35  ChIP-Seq protocol see Figure 2-1). For the ChIP, tissues were homogenized in 1% formaldehyde and incubated at room temperature for 10 minutes. Samples were incubated with 0.125 M glycine at room temperature for 5 minutes. Following pelleting and washing, samples were re-suspended in 5 volumes of ChIP cell lysis buffer (10 mM Tris–Cl, pH 8.0, 10 mM NaCl, 3 mM MgCl2, 0.5% NP-40) with a protease inhibitor cocktail (Roche). Cells were re-homogenized and put on ice for 5 min. Cells were pelleted again and re-suspended in 3 volumes of ChIP nuclear lysis buffer (1% SDS, 5 mM EDTA, 50 mM Tris–Cl, pH 8.1) containing a protease inhibitor cocktail (Roche). For sonication (Sonicator 3000, Misonix), samples were placed in an ice water bath and sonicated for 20 cycles of 30 seconds on, 40 seconds off. Sonicated chromatin was diluted with ChIP dilution buffer (to 250 μL, 0.01% SDS, 1.1% Triton X-100, 167 mM NaCl, 16.7 mM Tris–Cl, pH 8.1). 3μg of rabbit polyclonal anti-mouse SOX9 antibody (Millipore AB5535) or negative control of 3μg of rabbit polyclonal IgG antibody (Santa Cruz) was added to samples and incubated overnight while rocking at 4˚C. Simultaneously, Protein A/G beads (ThermoFisher Scientific) were blocked with 1 mg/mL BSA and 0.1 mg/mL herring sperm DNA in ChIP dilution buffer overnight. The following day, samples were incubated with blocked Protein A/G beads for four hours rocking at 4˚C. Beads were precipitated and washed with several buffers. The washes are as follows: low salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris–Cl, pH 8.1, 150 mM NaCl), high salt buffer (low salt buffer with 500 mM NaCl), lithium chloride buffer (0.25 M LiCl, 1% NP-40, 1% deoxycholate, 1 mM EDTA, 10 mM Tris–Cl, pH 8.1) and two final washes with TE buffer. For the elution of the bead complexes 125 μL elution buffer (1% SDS, 0.1 M NHCO3) was added and samples were rotated at room temperature for 15 minutes, twice. Proteinase K with RNAse A (ThermoFisher Scientific) and 0.192 M NaCl (reverse crosslinking) were added to samples and incubated 36    Figure 2-1 Generation of SOX9 ChIP-Seq libraries from E12.5 AVC and E12.5 limb buds. Approximately 5-15ug of total DNA was collected from the E12.5 AVC and limb buds. Tissues were homogeneized and subjected to crosslinking (protein:DNA). Crosslinked protein:DNA was sonicated and immunoprecipitated using the SOX9 antibody. Protein:DNA complexes (shown in different colours:red, green, yellow) were reverse crosslinked and DNA was isolated. DNA was sent for sequencing and bioinformatic analyses were performed to determine SOX9 peaks. 37  at 65˚C overnight. Purification of DNA was performed with two rounds of phenol-chloroform extraction and ethanol precipitation. Samples were spun down and pellets were re-suspended in 50 μL dH2O. A sample of genomic DNA was extracted as input for sequencing. Limb ChIP was performed alongside the heart ChIP to confirm enrichment of known target genes via ChIP-qPCR. ChIP DNA was sent to the Genome Sciences Centre (Vancouver, BC) for library generation. DNA was purified using 8-12% PAGE to isolate 100-300 bp fragments for short read (50 bp) sequencing on an Illumina Genome Analyzer 2 as described previously (110). The Burrows-Wheeler Aligner (111) aligned reads to the mouse genome (mm9) and unmapped reads were removed. FindPeaks3.1 (112) created virtual fragments by directionally extending uniquely mapped reads to a constant length (200 bp). Virtual fragments were profiled across the genome to identify regions of enrichment or “peaks”. A False Discovery Rate (FDR) was applied to threshold ChIP-Seq data as described (113) and false positives were limited using sequenced control data (input DNA). For each ChIP-Seq peak that passed the FDR ~0.01 threshold, we found the maximal coverage of the control sample in the region +/-300bp. A local z-score was calculated between the peak height and control coverage and peaks below the threshold were filtered out. Peaks that passed filtering, z-score, peak height, FDR based peak height cut-offs and did not overlap with control peaks (height over 30) were retained for analysis. Peak edges were refined by using a percentage of the maximum peak height to determine the peak edge cut off point. For each dataset, different criteria were used due to differences in sequencing depth and background noise. For screenshots and visual comparisons, we used unthresholded big wig or bed files. The data has been deposited to the GEO database (GSE73225).  38  2.6 Bioinformatic analysis of ChIP-Seq Wig and BED files generated by FindPeaks3.1 were analyzed using tools in UCSC Genome browser (114) (http://genome.ucsc.edu/), Galaxy (115) (http://main.g2.bx.psu.edu/), and Cistrome (116) (http://cistrome.dfci.harvard.edu/). SOX9 bound regions (peaks) were associated with genes through a “yes-no” process that ensured transcriptional start site (TSS) proximity was weighted prior to distance from either end of a gene (Appendix IID). This mapping system ensured that most peaks were only associated with one gene. The distribution of SOX9 peaks was analyzed by dividing the genome into categories: greater than 10 kb from a TSS; distal promoter (5-10 kb from a TSS); proximal promoter (less than 5 kb from a TSS); at a TSS (+/-100 bp); within the first 1 kb of a gene, within the first 1-5 kb of a gene; mid-gene (greater than 5 kb from either end of a gene); within the last 5 kb of a gene; at a transcription termination site (TTS); within 5 kb of the 3’ end of a gene; and greater than 5 kb from the 3’ end of a gene. SOX9 peaks were assigned to one of these regions using Cistrome. Cistrome’s SeqPos tool (116) and Screen Motif tool were used for motif analysis of SOX9 peaks. PWMs for the SOX monomer and dimer motifs were generated by SeqPos. PWMs used in Screen Motif were based on highest enrichment, p-value and top z-score. Gene Ontology (GO) analysis was performed using GOrilla (117). Lists were ranked using p-value and top categories were filtered for redundancy. Co-factor analysis on SOX9 peaks was performed using oPOSSUM (http://opossum.cisreg.ca/oPOSSUM3/) (118).  2.7 Genotyping Sox9 mutant embryos Sox9 mutant/cKO crosses were setup as mentioned above (for further information see Figure 2-2). Genomic DNA was collected from mouse ear punches or embryonic tissue (limb and hind 39  torso) and extracted using the KAPA mouse genotyping kit (Kapa Biosystems) according to the manufacturer’s protocol. For the PCR reaction, 0.5 µL (for embryonic tissue) or 1.5 µL (for ear punches) was used as template for the genotyping PCR reaction. HIFI Taq polymerase (ThermoFisher Scientific) was used for the PCR reaction according to the manufacturer’s protocol. The conditions used for the Sox9fl/fl PCR are: one step of 94˚C for 3 minutes, followed by 94˚C for 30 seconds, 60˚C for 30 seconds, 72˚C for 30 seconds repeated for 35 cycles and one final extension step of 72˚C for 2 minutes. The PCR program for Cre primers is similar with the exception of the 35 cycles which have 94˚C for 30 seconds, 55˚C for 1 minute, and 72˚C for 1 minute. The expected sizes of the bands for the Sox9fl/fl primers for WT were 250 base pairs (bp), heterozygous 300 bp and 250 bp and mutant 300bp. See genotyping primers Table 2-1.   2.8 RNA-Seq and bioinformatic analysis RNA from single AVCs for each genotype (WT and Sox9 cKO hearts) were used to synthesize cDNA (see RT-PCR method section) for Taqman assays to confirm efficient loss of Sox9 in the Sox9 cKO (VE/Tie2-Cre) AVCs. For the RNA-Seq libraries: RNA from two or three littermate AVCs with enrichment (WT, Sox9fl/fl) or loss (Sox9 cKO) of Sox9 in the AVC were pooled together. Pooled AVC RNA samples were further purified by GeneJet clean up and concentration micro kit (ThermoFisher Scientific). RNA quality was assessed on the BioAnalyzer (Agilent Technologies) and all samples had scores over 8.7 with a required passing score of 7. Duplicate RNA-Seq libraries for each genotype were generated and sequenced using Illumina Mi-Seq. Sequence reads were aligned with the Tophat2 tool (119) on Galaxy using the mouse reference genome mm9 (NCBI build 37) to generate BAM files. Aligned data from all  40    Figure 2-2 Generation of the Sox9fl/fl;VE-Cre and/or Sox9fl/fl;Tie2-Cre mice. A. The crosses needed to generate Sox9 mutant mice. B. Lineage tracing with the VE-cadherin-tTA (VEtTA) strain (a gift from L. Benjamin, Harvard Medical School) crossed with TetOS-lacZ (a gift from D. Dumont, Samuel Lunenfeld Research Institute) illustrating the cell types in the E10.5 and E11.5 heart that VE will delete in. These images are shown with permission from Dr. A. Chang during his time at Dr. Aly Karsan's laboratory (unpublished). 41  four libraries were analyzed and Fragments per Kilobase of exon per Million reads (FPKMs) were calculated using Cufflinks (120). Differential expression between WT and Sox9 cKO RNA-Seq libraries was determined using Cuffdiff (120). Gene FPKMs from Cufflinks were an average of the duplicate libraries for each genotype and only included genes represented in both duplicate libraries. Separate from Cuffdiff outputs, fold change was determined between WT and Sox9 cKO gene FPKMs to assess differential expression. For genes to be included for downstream analyses several criteria needed to be met: greater than or equal to 1 FPKM in either WT or Sox9 cKO and greater than or equal to 1.5 fold change between Sox9 cKO and WT AVC gene expression. Cistrome and Galaxy were used to determine differentially expressed genes that were associated with SOX9 peaks. GO analysis was performed using GOrilla (117). Mouse genes from RNA-Seq were used as background. Lists were ranked using p-value and top categories were filtered for redundancy. Ingenuity Pathway Analysis (Qiagen) biofunctions was used for GO analysis on SOX9 target genes overlapping in all libraries and to generate an interaction network of the biofunctions: transcription, cardiogenesis, and abnormal morphology of the heart on down-regulated SOX9 targets in the heart. Heat maps of gene expression data was generated using TM4 meV.   2.9 Cell culture, transfection, cloning and luciferase assays Cells were maintained in an incubator at 37˚C with 20% oxygen and 5% carbon dioxide. HEK 293T cells were maintained in normal growth medium (DMEM (Stemcell) + 10%FBS + Penstrep, ThermoFisher Scientific). Growth medium for the cells was changed every two to three days. HEK 293T cells were transfected at 60-70% confluency using polyethylenimine (PEI, Polysciences Inc) following a published protocol(121). Luciferase assays were performed 42  two days following transfection using the Dual-Luciferase Reporter Assay System (Promega). SOX9 peak regions of interest were cloned into pGL3 vector with a minimal promoter (E1B) or pGL4b promoter-less vector to test luciferase activity for individual enhancers/promoters in the presence or absence of SOX9 (pcDNA3-SOX9 over-expression vector).  2.10 Site directed mutagenesis Primers were designed to mutate the putative SOX9 dimer site in the SOX9 peak associated with the Hdac2 promoter region. The first SOX9 site the CAA portion was mutated to ACC and the second site of TGG to GGT similar to other studies to generate mutant constructs. Smaller fragments with homologous regions were generated and melted together. See primers in Table 2-1. The PCR conditions used to generate mutant sequences were 94˚C for 3 minutes for one cycle, 94˚C for 30 seconds, 60˚C for 30 seconds, and 72˚C for 1 minute for all for 10 cycles. There was a 20 minute hold at room temperature to allow for the addition of flanking primers followed by 35 cycles of 94˚C for 30 seconds, 60˚C for 30 seconds, and 72˚C for 1 minute. There was one extension cycle of 72˚C for 10 minutes. HIFI Taq was used to maintain high fidelity and decrease random mutation. Mutant sequences were cloned into the pGL4b vector and sent for sequencing to confirm the mutations in the SOX9 dimer site. Mutant vectors were used in luciferase assays to assess the effect on the activity of SOX9.  43  CHAPTER THREE: Identification and characterization of SOX9 binding sites in the developing heart valve and limb genome.  The goal of this analysis was to determine where SOX9 binds within the genome of the E12.5 AVC and limb buds to identify potential transcriptional targets of SOX9 and ultimately understand the function (through its gene targets) of SOX9 within these tissues. To do this, ChIP-Seq was performed on E12.5 AVC (primitive heart valves) and limb buds with an antibody that recognizes the SOX9 protein. This analysis identified the genomic locations of SOX9 binding in these two tissues and based on the genomic location of SOX9 bound regions, we determined the potential SOX9 target genes. SOX9 bound regions and transcriptional targets were compared among different tissues to determine tissue-specific and context-independent functions of SOX9. Furthermore, bioinformatic analyses of SOX9 bound regions can reveal tissue-specific co-factors of SOX9. However, this analysis is entirely dependent of the specificity of the SOX9 antibody and therefore I decided first to characterize SOX9 expression using this antibody during heart valve development using this antibody.  3.1 Sox9 mRNA and protein is enriched in the mouse heart valves throughout development At E12.5, SOX9 has been reported to be widely expressed in mesenchyme throughout the developing mouse heart valves (69) and condensing mesenchyme in the developing limb (122). To confirm the enrichment of Sox9 mRNA in the developing heart valves, qRT-PCR and in situ hybridization were used (Figure 3-1). Sox9 mRNA levels were enriched in the AVC region at E10.5, E12.5, and E14.5 from 3-12 fold over the atria (Figure 3-1 part A-C). Sox9 mRNA levels were also enriched in the OFT region at E10.5 and E12.5 which will form the aortic and  44    Figure 3-1 Sox9 mRNA is enriched in the developing heart valves. A-C. A representative qRT-PCR for Sox9 on E10.5 (A.), E12.5 (B.) and E14.5 (C.) valves (AVC), ventricles (V), atria (A), and outflow tract (OFT) relative to β-actin. D-F. Whole mount in situ hybridization with a probe specific for Sox9 on E12.5 hearts. D-E. Highlight valve (AVC) specific expression in cross-section (D.) and saggital section (E.) F. Shows both AVC and OFT Sox9 expression in the cardiac cushions. 45  pulmonary valves (Figure 3-1A, B). Similar to the qRT-PCR results, Sox9 transcripts were found in the valve forming regions of the E12.5 heart by in situ hybridization (Figure 3-1D). Since mRNA transcripts and protein levels do not necessarily correlate, I wanted to examine SOX9 protein levels in the embryo. To ensure that the SOX9 antibody (Millipore, AB5535) binding was specific to locations in the embryo known to express SOX9 protein, immunofluorescence was performed on E12.5 whole embryo sections (Appendix IA, B, E) and E10.5 heart sections (Appendix IC, D, F). In the E12.5 embryo SOX9 protein was found in locations previously demonstrated to express SOX9 such as the limb buds (arrowhead Supplemental Figure 1 part A) and somites. SOX9 protein was highly expressed in the E10.5 heart valves as expected (Appendix I).  To determine how SOX9 protein is expressed throughout heart valve development, embryonic hearts from E9.5 (when the valves first form via EMT) through to E16.5 (during remodeling and differentiation stages) were examined using immunofluorescence (Figure 3-2). SOX9 protein was found to be enriched in the heart valves at all stages examined. Overall, this data suggests that SOX9 plays an important role in heart valve development during initial valve formation, remodeling and differentiation. SOX9 was also expressed in another subset of cells within the developing heart called the epicardium (Figure 3-2, see arrowheads). The epicardial progenitor cells are a thin layer of cells found on the outside of the heart (E10.5-E14.5) and a subset of these cells will undergo an EMT at around E14.5 and invade the myocardium to give eventually give rise to the coronary vasculature and aid in myocardial growth (reviewed in (123)). However, epicaridal progenitor cells are not derived from endocardial EMT (124) like the AV valves.  47   Figure 3-2 SOX9 protein is enriched in the heart valves during heart development. Immunofluorescence on embryonic hearts at E9.5 (A, B), E10.5 (C, D), E12.5 (E, F), E13.5 (G, H), E13.5-14.5 (I, J), and E16.5 (K, L) with an antibody specific to SOX9 (red). Panels A, C, E, G, I, and K illustrate SOX9 staining only and panels B, D, F, H, J, L show the merged image of SOX9 and DAPI (blue). For later stage hearts the image is focused in on the AV valve region of the heart.  48  3.2 SOX9 directly binds thousands of DNA regions in the developing heart and limb One way to determine the function of a TF is to analyze the genes that this TF controls in a given context. Therefore to investigate the role of SOX9 in heart valve development, genome-wide profiles of SOX9 DNA-binding sites were generated for E12.5 heart valves (AVC) and limb buds using ChIP-Seq. The SOX9 genome-wide profiles for the E12.5 limb buds were generated to illustrate the similarities of the SOX9 initiated transcriptional programs in the developing heart valves and limbs. It has been suggested that developing valves and limbs share many similarities to one another and SOX9, and its transcriptional target genes, have been most well characterized in the developing limb. Thus, the limb provides a method of validating the SOX9 ChIP-Seq data by detecting previously identified SOX9 transcriptional target genes and an excellent tool to identify similarities between heart valve and limb SOX9-initiated transcriptional programs. Embryonic tissues were manually dissected out of the E12.5 embryo. Limb buds were pinched off the embryo. The E12.5 hearts were taken out of the embryo and further dissected to obtain only the AVC region by removing the ventricles, atria and OFT. A total of 164 E12.5 AVCs and 39 E12.5 limb buds were pooled in separate groups and used for three independent SOX9 (Millipore AB5535) ChIPs. The SOX9 antibody used for immunofluorescence analysis was also used for ChIP. Dissections and ChIPs were performed by R. Cullum in the Hoodless lab. The DNA from the three ChIPs was sent to the Genome Sciences Center for sequencing. Sequenced reads for each library were mapped to the mouse genome (mm9, NCBI build 37) and SOX9 peaks (regions of the genome where SOX9 is binding) were identified using a false discovery rate (FDR) of 0.01 followed by subtraction of an input DNA control. A local z-score was calculated between SOX9 peak height and control coverage for each library and SOX9 peaks below the threshold were filtered out (Appendix IIA, B). Peaks that passed filtering, z-49  score, FDR based peak height cut-offs and did not overlap with control peaks were retained for analysis. This identified a total of 2607 and 9092 SOX9 peaks in the E12.5 AVC and limb, respectively (Figure 3-3A, Appendix III). To validate the specificity of the SOX9 peaks in the limb, SOX9 binding was confirmed at the exact same regions as previously identified to be regulated by SOX9 in Col2a1, Acan, and Col11a2 (79,80,82) target genes (Appendix IIC).  To determine the degree of similarity in SOX9 binding between the AVC and limb, ChIP-Seq libraries were compared using a Venn diagram and identified 782 SOX9 peaks that were shared (29.8% of E12.5 AVC peaks) (Figure 3-3A). This data supports that there are similarities between the SOX9 initiated transcriptional programs in the valves and limb. Although there are similarities between the libraries, there were 1825 and 8310 unique SOX9 peaks in the E12.5 AVC and limb, respectively and indicate that SOX9 also has many tissue-specific binding sites. The increased number of unique SOX9 peaks in the limb is likely due to the heterogeneity of the limb buds at this time point and may reflect the diverse transcriptional programs in each different type of chondorcyte. To determine the potential target genes of SOX9, SOX9 peaks for each library were associated with genes through a “yes-no” process that ensured transcriptional start site (TSS) proximity was weighted prior to distance from either end of a gene (Appendix ID). This mapping system ensured that most peaks were only associated with one gene. From peak-to-gene associations, 2453 and 5750 potential gene targets of SOX9 were identified in the E12.5 AVC and limb respectively. Notably, 1605 genes were targeted by SOX9 in both tissues; this was more than double the shared SOX9 binding sites (Figure 3-3B) and suggests SOX9 targets similar genes in the valves and limb by using tissue-specific regulatory elements.  To determine how many SOX9 peaks contain a consensus SOX motif, de novo motif analysis was performed on SOX9 peaks with SeqPos (116). SeqPos is a program that identifies 50    Figure 3-3 Comparison of SOX9-initiated transcriptional programs in developing limb and heart and genomic peak locations. A. Venn diagram of peaks and B. targeted gene overlap between the E12.5 valve (AVC) and limb ChIP-Seq libraries. Note: SOX9 peaks overlap and SOX9 target gene overlap are completely separate from one another. ie. 782 peaks do not correspond to 1605 genes in the Venn diagram overlap. C. Positional weight models for the SOX monomer and dimer binding sites as identified from the ChIP-Seq data. Letter height indicates importance of the base for SOX9 binding. D. Distribution of SOX9 peaks across the genome in E12.5 AVC, E12.5 limb and shared binding sites in both tissues. 51  TF DNA binding motifs found in the SOX9 peak DNA sequences. This analysis generated both a SOX monomer (aACAAa/tg/t) and a SOX dimer (ACAaaGnnnnt/at/cTGT) position weight matrix (PWM) that is similar to the SOX9 JASPAR motif (Figure 3-3C). Of note, several other SOX monomer motifs were also identified in the SOX9 peaks but for simplicity the top ranked SOX motif based on highest enrichment, p-value and top z-score is shown. In sex determination, SOX9 is known to bind as a monomer to a single DNA binding site in the regulatory regions of Steroidogenic Factor 1 and Anti-Müllerian Hormone, whereas proteins required for chondrogenesis, like Collagens XI and IX, feature a SOX9 homodimer binding motif (85). To establish how many SOX9 peaks contain a SOX monomer and/or dimer site, all SOX9 peaks were scanned with Screen Motif using the top SeqPos PWMs for the monomer and dimer motif. Approximately, 77% of SOX9 limb peaks and 58% of SOX9 E12.5 AVC peaks contained at least one monomer or dimer motif (Table 3-1). The SOX dimer sequence was identified in 34% of SOX9 limb peaks compared to 13.5% of SOX9 AVC peaks, suggesting that SOX9 primarily binds as a monomer in E12.5 heart valves. Consensus SOX binding motifs are highly degenerate and thus this analysis may not capture all versions of the SOX9 binding motifs.  3.3 SOX9 binds active promoter regions within the genome To determine where SOX9 binds within the genome in relation to genes, SOX9 peaks were associated to putative target genes by dividing the genome into categories (Appendix IID & Methods). Approximately 22% and 31% of SOX9 binding sites in the limb and AVC, respectively, were found either directly over the TSS or in the 5kb upstream promoter regions. Furthermore, 64% of the overlapping SOX9 peaks in the AVC and limb were bound to the 52  Table 3-1: SOX9 monomer and dimer binding sites under the SOX9 peaks in heart and limb  Library # Peaks Monomer sites Peaks with sites Dimer Sites Peaks with sites Peaks with either monomer or dimer E12.5 AVC SOX9 2607 2589 1409 (54.0%) 386 353 (13.5%) 1516 (58.2%) E12.5 Limb SOX9 9092 10095 5678 (62.5%) 3491 3118 (34.3%) 6975 (76.7%) Random Average 10 of 2607 2607 1917 +/-56 1101 +/-20 (42.2 +/- 0.76%) 131 +/- 10 123 +/- 8 (4.71 +/-0.29%) 1147 +/- 39 (44.02 +/-1.50%) Random Average 10 of 9092 9092 6793 +/-117 3852 +/- 50 (42.38 +/-0.55%) 493 +/-33 459 +/-30 (5.05 +/-0.33%) 3946 +/- 42 (43.43% +/-0.46%) Shared peaks 777 957 498 (64.1%) 175 155 (19.9%) 510 (65.6%) Statistics on SOX9 motifs  SOX9 motifs Peaks w sites peaks w/o sites expected w/ sites expected w/o sites two-tailed p-value Chi-squared AVC monomer 1409 1198 1101 1506 <0.0001 149.152 AVC dimer 353 2254 123 2484 <0.0001 451.378 Limb monomer 5678 3414 3852 5240 <0.0001 1501.908 Limb dimer 3118 5974 459 8633 <0.0001 16222.645  53  TSS/5kb upstream regions (Figure 3-3D) and this suggests that SOX9 binding is biased to sites at or near promoter regions. Interestingly, the genes associated with overlapping SOX9 peaks in AVC and limb were enriched by Gene Ontology (GO) analysis for genes associated with metabolic processes and cell cycle regulation (Figure 3-4A, Appendix IV). Several examples of overlapping SOX9 peaks in the AVC and limb that are associated with cell cycle and proliferation and chromatin modifier genes such as Trp53 (p53), Junb, Rfwd3, Hdac1/2, and Eed are shown using the UCSC genome browser (Figure 3-5). A role for SOX9 in proliferation has been suggested on numerous occasions since many tissue-specific SOX9 mutant mice have decreased cell numbers and hypoplastic organ structures (46,57-59).  To determine the underlying functions of SOX9 target genes in the AVC and limb only, GO analysis was performed on E12.5 AVC SOX9 peaks and E12.5 limb SOX9 peaks (Figure 3-4B, C, Appendix V). In the AVC, SOX9 target genes were enriched for functions like adhesion, transcription, motility, WNT and TGFβ signaling and heart development (Figure 3-4B). SOX9 limb specific targets were involved in processes like mesenchyme development, chondrocyte differentiation, WNT and BMP signaling, and limb development (Figure 3-4C, Appendix VI). Interestingly, GO analysis on limb target genes also highlighted functions implicated in heart valve morphogenesis, EMT, and endocardial cushion development (Figure 3-4C, Appendix VI) and further support that the AVC and limb SOX9 transcriptional programs share many similarities.  3.4 Identification of potential co-factors of SOX9 3.4.1 DNA motif analysis identifies numerous potential co-factors of SOX9   Given that many SOX factors require the binding of additional co-factor TFs on the nearby DNA to efficiently regulate their target genes (47), SOX9 peak DNA sequences were analyzed 54   Figure 3-4 GO analysis reveals tissue specific functions for genes associated with SOX9 binding sites. GO analysis on the shared E12.5 AVC and limb (A.), AVC (B.), and limb (C.) target genes associated with SOX9 peaks. 55    Figure 3-5 UCSC genome browser screen shots of shared SOX9 binding locations in the developing heart and limb that are associated with cell proliferation/cell cycle and chromatin modifier genes. From top left down: Trp53, Junb, and Rfwd3. From top right down: Hdac2, Eed, and Hdac1. The top library is the E12.5 AVC SOX9 ChIP-Seq (pink) and the bottom library is the E12.5 limb SOX9 ChIP-Seq (blue). Below the libraries is the gene structure.56  for potential co-factor binding motifs using an online program called oPOSSUM (http://opossum.cisreg.ca/oPOSSUM3/) (118) (Figure 3-6, Appendices VII, VIII, IX). oPOSSUM uses a background set of DNA sequences of equivalent peak sizes and G:C content as the SOX9 peaks (generated by another tool in oPOSSUM) and searches the DNA sequences found under the SOX9 peaks for enriched TF DNA binding motifs. Based on a number of criteria, such as the number of TF motif sites found in SOX9 peaks versus background etc, were taken into account and TF motifs were ranked by z-score. The top co-factor binding site found in the AVC, limb and shared SOX9 binding regions is NFY which is known to bind at promoters to the DNA sequence 5’ CCAAT 3’ and regulate transcription. It is not surprising to find a promoter biased TF like NFY to be enriched in the SOX9 binding regions since we have already demonstrated that SOX9 binding is biased to promoter regions (Figure 3-3D). The SOX9 motif was enriched in all three sets as expected (Figure 3-6) and durther supports that SOX9 was bound to these regions. Additional SOX motifs were also identified such as SOX17 and SOX5. Many of the TF motifs identified are shared between libraries however the level of enrichment within SOX9 peak DNA sequences varies (Appendix VII, VIII, IX). Interestingly, the Arnt:Ahr binding motif was within the top five enriched motifs for each dataset (Figure 3-6) and suggests that HIF1a (where ARNT is a subunit) and SOX9 potentially regulate hypoxic genes together. However, additional experiments would be required to demonstrate this relationship.  Another way to examine the targets of SOX9 and to identify potential co-factors of SOX9 is to compare AVC and limb SOX9 ChIP-Seq libraries to other publicly available ChIP-Seq libraries. To look at the gene targets of SOX9 in other SOX9 expressing cell populations, SOX9 AVC and limb ChIP-Seq libraries were compared to SOX9 ChIP-Seq on hair follicle stem cells   57     Figure 3-6 SOX9 peaks are enriched for numerous potential co-factor binding sites. Co-factor analysis using oPOSSUM (http://opossum.cisreg.ca/oPOSSUM3/) on SOX9 E12.5 AVC (A.), limb (B.) and shared (C.) SOX9 peaks.58  (HFSCs) (62) and on the vertebral column (VC) (125) using a Venn diagram. This identified overlapping and non-overlapping SOX9 peaks (and their respective gene targets) indicative of common and tissue specific functions in different cell types for SOX9 (Figure 3-6A-B). The degree of overlap between the HFSCs SOX9 peaks and SOX9 AVC and limb peaks was fairly high considering the diversity of these cell types from one another and identifies potential common functions between these tissues (Figure 3-7A). An additional reason for the higher degree of similarity between these SOX9 ChIP-Seq libraries could be due to the use of the same SOX9 antibody (Millipore, AB5535) for the ChIP. Surprisingly, the overlap between the SOX9 VC peaks and that AVC and limb was much lower than expected given that the VC cells should have a more comparable function to the AVC and limb (Figure 3-7B). However, this study used a different SOX9 antibody (R&D Systems AF3075) and this may account for lower degree of overlap between libraries or that these cells types do not share similarity in function. A third study published SOX9 ChIP-Seq on rib chondrocytes (129) and would be ideal to compare the SOX9 AVC and limb peaks with but unfortunately the ChIP-Seq data was not fully released. These comparisons highlight the extremely context dependent role of SOX9 within different tissues and further supports that SOX9’s regulation of its target genes is likely to be dependent on the binding of specific co-factors in different tissues.  3.4.2 Comparison of SOX9 ChIP-Seq with published ChIP-Seq data sets reveals new insights into potential co-factors of SOX9 Different SOX factors are known to work together to regulate their target genes. SOX4 is expressed highly in the developing heart valves at a level as high as SOX9 (based on RNA-Seq data from the lab/data not shown) and known to be important for valve development (130). 59    Figure 3-7 Comparison of SOX9 peaks the AVC and limb with other TF ChIP-Seq libraries using Venn diagrams: SOX9 on hair follicle stem cells (HF-SCs(126)) and SOX9 on the vertebral column (VC) (125) (A, B), and other potential co-factor TFs: SOX4 on B cells (127) and EVI1 on ovarian cancer cells (OC) (128) (C, D). SOX9 H represents the SOX9 AVC ChIP-Seq library. 60  Therefore, it could be possible that SOX9 and SOX4 function together as co-factors to regulate valve-specific genes. The only publicly available SOX4 ChIP-Seq data set was generated on B-cells, and therefore SOX9 AVC and limb peaks were compared this SOX4 ChIP-Seq data set (127) (Figure 3-7C). Although it is not a functionally relevant cell type almost 1000 peaks were shared between the libraries (Figure 3-7C) similar to the SOX9 HFSC overlap suggesting that there are common binding regions for SOX4 and SOX9. This data supports the notion that SOX9 and SOX4 may work together as co-factors in the developing heart valve. Another transcription factor of interest was EVI1 (Mecom) since the EVI1 motif was identified by oPOSSUM in the SOX9 peaks. In the oPOSSUM analysis, the EVI1 motif was not found in a high number of SOX9 peaks and therefore was not highly enriched in this analysis (Appendices VII, VIII, IX). However, identifying the EVI1 motif was still intriguing as it has been shown to have a role in heart valve development (131). Currently, the only EVI1 ChIP-Seq data set available was performed on human ovarian cancer cells (OC) (128). Although the EVI1 ChIP-Seq was performed on an unrelated cell type, a follow up study by the same group that generated the ChIP-Seq demonstrated that a number of targets of EVI1 identified in OC cells were implicated in congenital heart defects (131). Upon comparison of SOX9 and EVI1 binding regions, the overlap identified by the Venn diagram was not very high (Figure 3-7D) and this may be due the highly dissimilar cellular contexts of the ChIP-Seq libraries. Additional co-factor analysis on SOX9 peaks in the E12.5 AVC using SeqPos revealed 431 target genes with EVI1 motifs that have interesting implications in heart valve development. Overall, this suggests that SOX4 and EVI1 may function as potential co-factors together with SOX9 to regulate different subsets of genes during heart valve development.  61  In the interest of identifying potential co-factors of SOX9 in a relevant cellular context SOX9 AVC and limb peaks were also compared to TWIST1 and SMAD3 AVC ChIP-Seq libraries that were readily available in the lab (Appendix X). TWIST1 and SMAD3 were not identified as potential co-factors of SOX9 by oPOSSUM but this may be due to the short highly degenerate nature of TWIST1 and SMAD3 binding motifs. Previous work suggests that SMAD3 associates with SOX9, recruits P300 and activates SOX9-dependent transcription in chondrocytes (132,133). Several hundred peaks overlap between SOX9 and SMAD3 and this supports that there may be some type of cooperation together in the developing heart valves similar to chondrocytes. Further analysis of co-targeted SOX9 and SMAD3 target genes would necessary to determine the nature of their relationship in gene regulation. Interestingly, TWIST1 is targeted by SOX9 in the AVC data set and it is known that SOX factors often regulate their co-factors. Comparison of the TWIST1 and SOX9 AVC ChIP-Seq libraries demonstrates that the degree of overlap between the TWIST1 and SOX9 AVC peaks was very low (Appendix X) and indicates that they likely do not function as co-factors. To support this, previous work in chondrocytes suggests that TWIST1 functions to suppress cartilage formation by directly inhibiting SOX9 during chondrogenesis (134,135). However, there may be a feedback relationship between TWIST1 and SOX9 in the heart valves. Future studies will be directed at understanding the critical co-factors of SOX9 and how they regulate their target genes during heart valve and limb development.  62  CHAPTER FOUR: Characterization of the Sox9fl/fl;VE-Cre mice  To gain a better understanding of which transcriptional targets of SOX9 are critical for heart valve development, I wanted to generate a mouse line with an endothelial-specific deletion of Sox9 (specifically deleting SOX9 in the AVC). Subsequently, I wanted to use this endothelial-specific Sox9 mutant to analyze the transcriptional changes in the AVC following the loss of SOX9 with the ultimate goal of comparing altered genes identified in the SOX9 mutant with the transcriptional target genes identified in the AVC. These comparisions would help to narrow down the potential SOX9 transcriptional targets that are essential to heart valve development.  4.1 SOX9 negative AV mesenchyme cells are absent in Sox9 mutant valves To generate an endothelial-specific Sox9 mutant, I obtained the genetically modified Sox9fl/fl mouse from Jax Laboratories (USA) and crossed to the Vascular Endothelial cadherin (VE) Cre mouse (a kind gift from Dr. Aly Karsan). In this system, the presence of Cre recombinase (Cre) in the endothelium will induce the excision of a portion of the Sox9 coding region by binding flox sites contained in the mutant flox allele. Sox9 will be deleted by VE-Cre in the AVC endocardium and in subsequent AVC mesenchymal cells of heart valves as early as E9.5. Although Tie2-Cre is the most-commonly used mouse line for deletion in endothelium and an endothelial-specific knockout of Sox9 using Tie2-Cre has already been generated (46). The VE-Cre mouse line was chosen since it has been shown to be more specific to the endothelium than Tie2-Cre (136) and was readily available in house. To generate Sox9flox/flox;VE-Cre mice, it is important that the female is Sox9fl/fl without Cre to generate Sox9 mutants embryos as it has been shown that in some cases Cre can be inappropriately activated in the mother causing 63  excision in all progeny (137). Inappropriate excision of Sox9 by Cre in the developing embryos could lead to abnormalities that are not dependent on the genotype of the embryo. Of note, the VE-Cre mouse was made in house by the Karsan lab and is not the commercially available mouse. Dr. A. Chang (during his time in the Karsan lab) evaluated the efficiency of VE as a promoter by crossing the VE-cadherin-tTA (VEtTA) strain (a gift from L. Benjamin, Harvard Medical School) with TetOS-lacZ mouse (a gift from D. Dumont, Samuel Lunenfeld Research Institute) and performing LacZ staining on E10.5 and E11.5 hearts following removal of doxycycline the previous day (see Figure 2-2, with permission granted by Dr. A Chang). When doxycycline is removed it will activate β-galasctosidase and allows it to be expressed. Therefore, any cell that has VE will express β-galasctosidase and can be visualized by LacZ staining, which turns cells blue. LacZ staining indicates that there is efficient recombination using the VE promoter in the E10.5 and E11.5 AVC mesenchyme cells however some unmarked cells were still present. Although it was not the exact same VE-Cre mouse, it demonstrates that VE as promoter can drive efficient recombination within the developing heart valves. However, it is essential to test Cre efficiency in every mouse line as there can be variability in efficiency between lines.  To evaluate the efficiency of Sox9 deletion in Sox9fl/fl;VE-Cre/+ heart valves, embryonic hearts were collected at E9.5, E10.5 and E12.5 and examined using immunofluorescence with an antibody specific for SOX9 (Figure 4-1). As a note, the SOX9 antibody used for immunofluorescence is exactly the same one used for ChIP-Seq experiments. Very little differences between Sox9fl/fl;+/+ (WT) and Sox9fl/fl;VE-Cre/+ (Sox9 mutant) were detected at E9.5 and E10.5 (Figure 4-1A-H). At E12.5, Sox9 mutant AVC cushions were reduced in size  64    Figure 4-1 SOX9 is not lost in the developing heart valves of the Sox9fl/fl;VE-Cre mice despite valve abnormalities. Immunofluorescence for SOX9 on WT (A, B, E, F, I, J, Sox9fl/fl;+/+) and Sox9 cKO (C, D, G, H, K, L, Sox9fl/fl;VE-Cre) on E9.5 (A-D), E10.5 (E-H) and E12.5 (I-L) hearts.The AVC is highlighted by the white box and magnified in the following panel. 65  when compared to WT (Figure 4-1I-L). Additionally, the atrial septum of the Sox9 mutant had not fused with the AVC whereas the WT septum had already fused (Figure 4-1I-L). At all of the embryonic stages examined SOX9 negative mesenchyme could not be detected suggesting that either Sox9 was not being deleted properly or that Sox9 negative mesenchymal cells are lost. Cell counts showed 50% fewer SOX9 positive cells in E12.5 Sox9fl/fl;VE-Cre/+ AVC cushions compared to WT confirming the reduced number in overall SOX9 positive cells (Figure 4-2A) Since the Sox9 mutant AVC is smaller in size, the numbers of SOX9 positive cells were compared to DAPI (marks nuclei) positive cells to adjust for decrease cell numbers and demonstrated that Sox9 mutant AVCs had 28% less SOX9 positive cells over DAPI (Figure 4-2B).  To confirm that Sox9 transcript was also still present in the Sox9 mutant valves, I examined the Sox9 transcript levels in the AVC relative to other regions of the heart using qRT-PCR on WT, double heterozygous (Sox9fl/+;VEC/+) and Sox9 mutant AVC, atria (A) and ventricles (V) using primers specific to the flox region of Sox9 (Figure 4-2C). Sox9 mRNA levels are enriched in the WT AVC relative to other regions of the heart. Thus for this analysis, WT atria was set to 1 to show Sox9 enrichment in the AVC. As expected Sox9 was 9 fold higher in the WT AVC compared to the WT atria (Figure 4-2C). The double heterozygous and Sox9 mutant AVCs also had highly enriched levels of Sox9 transcript (Figure 4-2C). This data indicates that Sox9 mRNA was not significantly reduced in the Sox9 mutant AVC. The increased levels of Sox9 in the double heterozygous and Sox9 mutant atria are likely due to AVC or epicardial contamination that occurred during manual dissections. However, it could indicate that Sox9 transcript levels have reduced enrichment in the double heterozygous and Sox9 mutant AVC (Figure 4-2C) although immunofluorescence data does not support this. 66    Figure 4-2 Sox9fl/fl;VE-Cre mice have reduced numbers of SOX9+ valve cells and decreased  total valve cell numbers. Cell counts of SOX9+ cells (A.) and SOX9+ cells relative to DAPI (B.) on WT (Sox9fl/fl;+/+) and Sox9 mutant (Sox9fl/fl;VE-Cre) valves. C. A representative Taqman qRT-PCR for Sox9 relative to Gapdh on AVC (valves), A (atria), V (ventricles) for WT, Heterozygous (Sox9fl/+;VE-Cre) and Sox9 mutant hearts.   67   It is intriguing that SOX9 negative mesenchyme cells could not be identified in Sox9 mutant AVCs and there are a number of possible reasons for this phenomenon. One potential explanation for the absence of SOX9 negative cells in the AVC could be that other SOX9 positive mesenchymal cells are migrating into the mutant AVC to assist recovery from the loss of the normal AVC mesenchyme cells. In the developing heart there are several other SOX9 positive populations of cells such as the mesenchymal caps of the atrial septum and ventricular septum and the epicardial cells. Of note, the neural crest cells that migrate into the OFT valves are also SOX9 positive. Immunofluorescence for SOX9 on the WT E12.5 heart easily identified the other SOX9 positive cell populations of the septal caps and epicardium (Figure 4-3, arrowheads). Another potential explanantion for the lack of SOX9 negative mesenchyme may be due to the inefficiency of the VE-Cre system and future work should be focussed on performing lineage tracing with this mouse line.   4.2 SOX9 is maintained at later stages of heart valve development in Sox9 mutants SOX9 protein has been demonstrated to be a long lived protein in limb development (138). For example, it has been shown that SOX9 protein is maintained in hypertrophic chondrocytes even though there is no Sox9 transcript in these cells. SOX9 protein is lost once cells become terminally differentiated chondrocytes (138). To determine if SOX9 protein is eventually lost in the Sox9 mutant heart valves, later stages (E13.5 and E16.5) of heart valve development were examined (Figure 4-4). In E13.5 and E16.5 hearts SOX9 protein was still present in the AV valves but there were noticeable changes in the valve phenotype at E16.5 such as changes in size and structure in valve leaflets. This data indicates that Sox9 deletion by VE-Cre is having an effect on the heart valve development (Figure 4-4E-H). However in the majority of cases the  68    Figure 4-3 Additional sources of SOX9+ mesenchyme found in the developing heart at E12.5. (A, B) The mesenchymal cap of the atrial septum (AS), (C,D) the mesenchymal cells on the surface of the ventricular septum (VS), and (E,F) the epicardial layer on the outside of the heart. SOX9 is labelled in red and DAPI in blue. 69    Figure 4-4 SOX9 positive cells can still be detected at later stages of valve development in Sox9fl/fl;VE-Cre mice. The majority of Sox9fl/fl;VE-Cre mice are not embryonic lethal. Immunofluorescence for SOX9 on WT (A, B, E, F) and Sox9 cKO (C, D, G, H) hearts at E13.5(A-D) and E16.5 (E-H).70  valve phenotype was not severe enough to be embryonic lethal and Sox9 mutant mice survive well into adulthood (Table 4-1). In addition to the valve phenotype, the myocardium of the ventricles appear to be thickened in the Sox9 mutants compared to WT. To see if the observed phenotypes were transient in the E16.5 Sox9 mutant hearts, Sox9 mutant mice were taken at birth to examine their hearts for physical differences (Figure 4-5). Similar to the E16.5 hearts, there was increased thickness in the ventricular myocardium and decreased chamber size in the Sox9 mutant hearts compared to WT (Figure 4-5, arrowheads). Additionally, the Sox9 mutant ventricular myocardium looks more compact with less open spaces within the heart and less trabeculated than its WT counterparts. Interestingly, the phenotype in the ventricular myocardium of Sox9 mutant hearts is similar to cardiac hypertrophy. Cardiac hypertrophy can occur in patients with hypertension or heart valve stenosis and these conditions lead to a thickening of the ventricular myocardium and reduced ventricular chamber size (reviewed in (139)). It is possible that the increased ventricular myocardial thickness in Sox9 mutant hearts could be indicative of cardiac hypertrophy due to improper heart valve function.  4.3 Adult Sox9 mutants have slightly thickened valves leaflets Since the ventricular phenotypes in the Sox9 mutants at birth were similar to hypertrophy, mice were aged until approximately to one year and taken for histological analyses to examine the heart. Hematoxylin and Eosin staining of the adult Sox9 mutant AV heart valves reveals Sox9 mutants have a slight thickening at the base of the leaflets (Figure 4-6, arrows). Although this thickening is slight, even minor changes in valve structure and composition can lead to the improper function of the heart valves, such as floppy valves that regurgitate or stiff valves that   71  Table 4-1 Genotypes from Sox9fl/fl;+/+ and Sox9fl/+;VE-Cre/+ mouse crosses to generate SOX9 mutant embryos  Time point WT (Sox9fl/+;+/+ & Sox9fl/fl;+/+) HET (Sox9fl/+;VE-Cre/+) SOX9 cKO (Sox9fl/fl;VE-Cre/+) # of litters E9.5 4 3 2 1 E10.5 5 1 3 1 E12.5 19 6 10 4 E13.5 5 1 2 1 E16.5 2 1 4 1 Adult 42 24 13 14  72    Figure 4-5 Ventricular abnormalities in the Sox9fl/fl;VE-Cre postnatal mice hearts when compared to WT (Sox9fl/fl;+/+). (A, B) Highlights the differences in chamber size in Sox9fl/fl;VE-Cre versus WT (arrowheads). (C, D) Highlights differences in ventricular thickness between Sox9fl/fl;VE-Cre/+ and WT (line). N=2.  73    Figure 4-6 Hematoxylin and Eosin staining of 10 month old adult WT and Sox9fl/fl;VE-Cre AV heart valves. Arrows indicate AV heart valves. A=atria, V=ventricles. N=2.   74  cause stenosis. Currently, this analysis was performed on two sets of hearts that gives an n of 2 for each genotype and to obtain a significant difference in valve thickness further animals will need to be examined. Additional mice are being aged to one year for analysis. Surprisingly, there were no overt differences in the ventricular myocardium  thickness or in the architecture of the ventricular myocardium of the adult Sox9 mutant hearts examined to date. However when the hearts were taken for dissection, the Sox9 mutant female heart was noticeably bigger in size compared to her WT littermate (data not shown). Hypertrophic hearts often have an increase in total size of the heart and it is possible that there are increases in myocardial thickness not seen in the images taken of the Sox9 mutant female heart. Further analyses are required to establish if there are similar changes in the Sox9 mutant adult ventricular myocardium as seen in the postnatal mutant hearts. Overall, Sox9 mutant hearts have a very mild phenotype on heart valve development and may be a useful model to study heart valve disease. However, it will be necessary to determine if Sox9 mutant mice are susceptible to or already valve disease by testing heart function and blood flow as well as additional staining procedures to examine differences in heart valve composition and structure. 75  CHAPTER FIVE: Characterization and analysis of the Sox9fl/fl;Tie2-Cre mice  Given that the Sox9fl/fl;VE-Cre mutant mice had no detectable SOX9 negative mesenchyme in the AVC it was not an efficient system to examine the differences in SOX9 transcriptional changes upon loss of Sox9. Since our goal is to parse out critical targets of SOX9 in the developing heart by comparing the transcriptional changes identified between WT and Sox9-deficient heart valves to the SOX9 AVC ChIP-Seq data, it was necessary to find another system to delete SOX9 in the embryonic heart valves. Previous work by Lincoln et al had shown that endothelial-specific deletion of Sox9 using Tie2-Cre where both endothelial cells and mesenchymal cells of the AVC cushions fail to express SOX9 are embryonic lethal (46). Tie2-Cre has been previously shown to specifically label epithelial cells and not cardiomyocytes, epicardium or distal OFT mesenchyme, however it does have some non-specific labeling in mesoderm (124,140). Sox9fl/fl;Tie2-Cre mice die during embryogenesis as a result of heart valve defects including hypoplastic cardiac cushions, reduced mesenchyme proliferation and altered ECM composition (46). Therefore, I decided to take advantage of this previously published mouse model with the intent of analyzing transcriptional changes that occur upon loss of Sox9 in the developing valves. In order to reduce confusion between mutant mouse models Sox9fl/fl;Tie2-Cre mice will always be referred to as the Sox9 cKO whereas the Sox9fl/fl;VE-Cre mice will be called Sox9 mutants.   5.1 SOX9 negative mesenchyme was detected as early as E10.5 in the AVC To ensure efficient deletion of Sox9 by Tie2-Cre and to determine if SOX9 negative mesenchyme can be detected in the AVC, immunofluorescence using an antibody specific for 76  SOX9 was performed on Sox9fl/fl (WT) and Sox9fl/fl;Tie2-Cre (Sox9 cKO) hearts starting at E9.5 (Figure 5-1). At E9.5, SOX9 was still detected in AVC mesenchyme of the Sox9 cKO (Figure 5-1 C, D) but by E10.5 and E11.5 SOX9 negative AVC mesenchyme was identified in Sox9 cKO valves (Figure 5-1 G, H, K, L, arrowheads). However, some SOX9 positive mesenchyme cells were observed in the E11.5 AVC (Figure 5-1, asterisk) suggesting that Sox9 deletion by Tie2-Cre was not complete. No overt morphological differences between the E9.5 and E10.5 WT and Sox9 cKO hearts were observed but there was a slight decrease in AVC size between WT and Sox9 cKO by E11.5 (Figure 5-1).  Since we wanted to match the embryonic time point chosen for SOX9 E12.5 AVC ChIP-Seq to the transcriptome analysis of Sox9-deficient valves, I wanted to fully characterize the E12.5 Sox9 cKO AVC prior to library generation (Figure 5-2). Immunofluorescence for SOX9 coupled with confocal microscopy revealed that the E12.5 Sox9 cKO valves were significantly reduced in size and overall SOX9 positive mesenchyme in the AVC was also reduced when compared to WT (Figure 5-2A, B). Sox9 cKO hearts had several additional defects including decreased thickness of the ventricular myocardium and delayed septal fusion with the AVC (Figure 5-2A). Illustrating the specificity of Sox9 deletion in the Tie2-Cre lineage, SOX9 expression was lost in the AVC, whereas the epicardial cells that descend from a separate lineage still express SOX9 (Figure 5-2A, B). To confirm the decreased numbers of SOX9 positive mesenchyme cells in the AVC of the Sox9 cKO hearts, the total numbers of SOX9 positive mesenchyme cells were counted in the AVC for WT and Sox9 cKO (Figure 5-2C). Sox9 cKO valves had approximately 65% less SOX9 positive cells than the WT (Figure 5-2C). Due to the extreme compaction of the nuclei in the Sox9 cKO AVC (Figure5-2A, B), DAPI nuclei counts could not be performed since it was extremely difficult to separate nuclei apart in the 77    Figure 5-1 SOX9 deletion occurs in the Sox9fl/fl;Tie2-Cre  heart valves as early as E10.5 in the mouse. Immunofluoresence on the WT (A, B, E, F, I, J) and Sox9 cKO (C, D, G, H, K, L) hearts for SOX9 at E9.5 (A-D), E10.5 (E-H), and E11.5 (I-L). Panels A, C, E, G, I, K are SOX9 (red) staining only. Panels B, D, F, H, J, L are merged images with SOX9 (red)  and DAPI (blue) staining. Images were captured on a Zeiss Axioplan 2 microscope. 78    Figure 5-2 The loss of SOX9 in the Sox9fl/fl;Tie2-Cre E12.5 heart valves leads to major valve abnormalities and reduced valve cell numbers. A. Immunofluorescence for SOX9 on E12.5 WT (top panel) and Sox9 cKO (bottom panel) hearts. B. An inset of part A highlighting the valve region of the heart. Average valve cell counts (C.) and epicardial cell counts at the AV junction (D.) on WT and Sox9 cKO hearts. 79  Sox9 cKO AVC. There appeared to be more epicardial cells in the AVC region of the Sox9 cKO than in WT based on immunofluorescence for SOX9 and suggests that eipcardial cells could be migrating into the Sox9 cKO AVC or that these cells are not migrating out over the surface of the heart as they should be at this time point. However, epicardial lineage tracing would be necessary to confirm that epicardial cells are present within the AVC. To examine changes in epicardial cell numbers in the Sox9 cKO AVC, SOX9 positive cells were counted on each side of the heart and totaled up (Figure 5-2D). Overall, there were no significant differences but there was a slight trend towards an increase in total epicardial cell numbers in the Sox9 cKO AVC region particularly on the right side of the heart (Figure 5-2D). This may indicate that there is a slight delay in epicardial cell migration on the right side of the heart of the Sox9 cKO however further studies are required to show this more definitively.  5.2 Deletion of Sox9 was variable in the Sox9 cKO heart valves As expected from the Lincoln et al study, deletion of Sox9 by Tie2-Cre leads to death in the majority of embryos by or just after E13.5 due to severely hypoplastic and malformed AV valves (Figure 5-3, Table 5-1). The Sox9 cKO valves are severely reduced in size compared to WT and the blood begins pooling and clotting within the chambers of the heart (Figure 5-3). Of note, the efficiency of Sox9 deletion in the AVC by Tie2-Cre is variable. Immunofluorescence for SOX9 was performed on three separate Sox9 cKO hearts to demonstrate the variability of Sox9 deletion by Tie2-Cre (Figure 5-4). Some Sox9 cKO valves had 90-95% Sox9 deletion (Figure 5-4A, B), others had 45-50% Sox9 deletion (Figure 5-4C, D) and some even had only 10-15% Sox9 deletion (Figure 5-4E, F). This illustrates that complete deletion of Sox9 was not always achieved in the Sox9 cKO AVC and therefore demonstrates a major caveat in pooling 80  Table 5-1 Genotypes from Sox9fl/fl;+/+ and Sox9fl/+;Tie2-Cre/+ mouse crosses suggest that Sox9 cKO embryos die after E13.5.  Time point WT (Sox9fl/+;+/+ & Sox9fl/fl;+/+) HET (Sox9fl/+;Tie2-Cre/+) SOX9 cKO (Sox9fl/fl;Tie2-Cre/+) # of litters E9.5 6 2 2 1 E10.5 5 2 1 1 E11.5 1 3 3 1 E12.5 96 65 55 30 E13.5 7 6 2 2 E16.5 5 1 0 1 Adult 19 14 1* 7  81    Figure 5-3 The absence of SOX9 in the Sox9fl/fl;Tie2-Cre heart valves is embryonic lethal between E13.5-14.5. Immunofluorescence for SOX9 on E12.5 WT (A, C) and Sox9 cKO (B, D) hearts. Panels C and D are an inset of A and B highlighting the valve region of the heart.  82    Figure 5-4 Sox9 deletion by Tie2-Cre is extremely variable in the Sox9 cKO heart valves. Immunofluorescence for SOX9 on three different E12.5 Sox9 cKO hearts illustrating 90-95% deletion (A, B), 45-50% deletion (C, D) and 10-15% deletion (E, F). Panels A, C and E are SOX9 staining only (red) and panels B, D, and F are the merged images with DAPI (blue).  83   Sox9 cKO hearts for any type of future analysis. This variability in Sox9 deletion was further confirmed at the transcript level using Taqman assays for Sox9 relative to Gapdh on WT and Sox9 cKO AVCs with more or less deletion of Sox9, compared to atria and ventricles for each genotype (Figure 5-5). Atria were set to 1 for each genotype to demonstrate the lack of Sox9 enrichment in the Sox9 cKO AVC compared to WT. This method could be used to identify Sox9 cKO AVCs with more or less efficient deletion of Sox9 and separate them out for future analyses.  5.3 Proliferation defects in the Sox9 cKO valves Similar to Lincoln et al.(46), a decrease in proliferation was observed in Sox9 cKO AVCs when compared to WT using immunofluorescence with an antibody specific to phospho histone H3 (pHH3, a marker of mitosis) (Figure 5-6). Interestingly, numerous SOX9 conditional mutants have defects in proliferation and decreased cell numbers in the mutant tissue (46,57-59). Additionally, in rat mesenchymal stem cells, knockdown of SOX9 caused reduced proliferation and increased levels and stability of cyclin D1 (60). Consequently, cyclin D1 protein was investigated in the WT and Sox9 cKO valves by immunofluorescence (Figure 5-7A). Similar to rat mesenchymal stem cells, cyclin D1 protein was maintained in the cytoplasm in a subset of the Sox9 cKO AVC mesenchyme (Figure 5-7A, arrows). Several cyclin D1 positive cells were also observed in the WT (Figure 5-7A, arrows) however fewer cells were detected and not found in a specific pattern as seen in the Sox9 cKO AVC. Ccdn1 (cyclin D1) transcript levels were measured between WT and Sox9 cKO AVC using Taqman assays (Figure 5-7B). Surprisingly, Ccdn1 mRNA levels were slightly reduced in the Sox9 cKO AVC when compared to WT although the levels were not significant (Figure 5-7B). Although cyclin D1 was not a direct  84    Figure 5-5 Verification of Sox9 deletion, by Tie2-Cre, in the Sox9 cKO heart valves. Sox9 cKOs were grouped by the level of Sox9 deletion (more complete deletion (g) versus incomplete deletion (b)). Taqman assays for Sox9 and Gapdh were used for this analysis. Atria were set to 1 to determine the level of SOX9 enrichment (WT) or loss (Sox9 cKO). An N of 12 or greater for AVC and A and an n of 8 for the ventricles (V). 85    Figure 5-6 Proliferation is reduced in the Sox9 cKO valves compared to WT. A. Immunofluorescence on WT (top panels) and Sox9 cKO (bottom panels) hearts for phospho histone H3 (a marker of mitosis, green). B. Counts of pHH3 positive cells on WT and Sox9 cKO valve regions that are highlighted by a boxed in region in part A. 86    Figure 5-7 Cyclin D1 immunostaining suggests that SOX9 is required to exit S phase during cell cycle. A. Immunofluorescence for cyclin D1 on WT (top panel, n=2) and Sox9 cKO (bottom panel, n=3) E12.5 hearts sections. Arrows indicate cells with cyclin D1 staining. Representative images taken on a confocal microscope. B. Taqman qRT-PCR assays on WT and Sox9 cKO for cyclin D1 (Ccnd1) relative to Gapdh. WT levels are set to 1. N= 5 or greater. 87  target of SOX9 in the AVC by ChIP-Seq analysis this data suggests that SOX9 may target other factors that are involved in regulating cyclin D1 at the protein level to maintain its stability such as the F-box and WD repeat domain containing proteins that degrade cyclinD1 (141).  5.4 Global transcriptional alterations in the Sox9 cKO heart valves To identify critical gene targets of SOX9 that may be responsible for the valve abnormalities resulting in embryonic death in the Sox9 cKO, the transcriptome of the E12.5 AVC was compared in the presence (WT) and absence of Sox9 (Sox9 cKO). Since the efficiency of Sox9 deletion in the AVC by Tie2-Cre is variable, RNA was isolated from individual AVCs from E12.5 WT and Sox9 cKO and qRT-PCR verified the loss of Sox9 for each sample. RNA-Seq libraries were generated with RNA from 2-3 AVCs pooled for WT and confirmed Sox9 cKO and performed in duplicate for each genotype (Figure 5-8A). Differential expression between WT and Sox9 cKO RNA-Seq libraries was determined using Cuffdiff (120) and fold change was calculated between WT and Sox9 cKO gene FPKMs. Gene FPKMs from Cufflinks were an average of the duplicate libraries for each genotype. This analysis identified 657 genes that were at least 1.5 fold down-regulated in the E12.5 Sox9 cKO AVC and 352 genes that were at least 1.5 fold up-regulated in the E12.5 Sox9 cKO AVC (Appendix XI). Interestingly, down-regulated genes in the Sox9 cKO AVC are involved in cartilage and heart valve development, and EMT based on GO analysis (Figure 5-8B). While up-regulated genes in the Sox9 cKO AVC included functions like response to hypoxia and stress and regulation of cardiac muscle hypertrophy (Figure 5-8C).  To rule out a general reduction in AVC expressed genes due to the reduced size of the 88    Figure 5-8 Comparison of differential transcripts in the Sox9 cKO AVC identified by RNA-Seq and SOX9 ChIP-Seq reveals a key subset of genes important for heart valve formation. A. Taqman assays for Sox9 relative to Gapdh for each library on the WT and Sox9 cKO AVC and atria. Significance is determined using Student’s T-test. N=3 for each genotype in each library. GO analysis on genes reduced (B.) or increased (C.) by greater than 1.5 fold change in the Sox9 cKO versus WT valves. D. The workflow for comparing RNA-Seq data with the ChIP-Seq data to identify critical SOX9 targets required for valve formation.  89   Sox9 cKO AVC cushion, differentially expressed  transcripts identified in the Sox9 cKO AVC were compared to a previously generated list of 206 AV-enriched genes identified in mouse heart (142). I found that 59% of AV-enriched genes have no change or changes less than 1.5 fold between WT and Sox9 cKO. This suggests that the reduced expression of a subset of AV-enriched genes in the Sox9 cKO valves is likely due to the specific loss of SOX9 and not from the decreased valve size (Figure 5-9A). To illustrate this, subsets of AV-enriched genes that have no change in expression in WT versus Sox9 cKO AVC are shown (Figure 5-9B). Additionally, to demonstrate that housekeeping genes were not heavily changed by Sox9 deletion in the AVC, housekeeping gene expression levels from the RNA-Seq libraries are shown (Figure 5-9C, D).  To further investigate genes that are critical to heart valve development and are regulated by SOX9, genes with altered expression in Sox9 cKO AVCs were associated with gene targets of SOX9 in the E12.5 AVC using Cistrome (116) (Figure 5-8D). This analysis identified 139 genes that are both SOX9 gene targets in the E12.5 AVC and have altered expression (greater than 1.5 fold change) in Sox9 cKO AVC. Of the 139 identified genes, approximately three quarters were down-regulated and one quarter were up-regulated (Figure 5-8D, Table 5-2, Appendix XII, XIII). To illustrate the overall gene expression changes of SOX9 targets with a 1.5 fold change (or greater) in expression level heat maps were generated for the down and up-regulated genes (Figure 5-10). Overall, this data suggests that SOX9 likely fine tunes gene expression levels rather than acting as an on/off switch as many traditional transcription factors. On the other hand, SOX9 may have another unknown role other than regulating transcriptional activity of its target genes. 90    Figure 5-9 SOX9 selectively targets a subset of AV-enriched genes in the AVC. A. A pie chart of the AV enriched genes identified in the Delaughter paper separated by genes that are up and downregulated and with no change B. Selected AV-enriched genes that show no change in Sox9 cKO valves. Actin (C.) and housekeeping genes (D.) expression levels between WT and Sox9 cKO illustrating no or little change in expression.   91  Table 5-2 The top ten SOX9 targets with altered gene expression in the Sox9 cKO  Genes with ≥1.5 fold change down in the Sox9 cKO with restrictions (≥ 1FPKM in either WT or Sox9 cKO) that have a SOX9 peak Gene  symbol Full Gene Name Gene ID WT FPKM cKO FPKM p-value from Cuffdiff Fold change down in cKO Btn1a1 butyrophilin, subfamily 1, member A1 NM_013483 6.92 0.29 0.0001 18.197 Prelp proline/arginine-rich end leucine-rich repeat protein NM_054077 3.97 0.32 0.0002 9.701 Tnrc6b trinucleotide repeat containing 6B NM_144812 66.09 13.91 0.0004 4.724 Ctnna2 catenin (cadherin-associated protein), alpha 2 NM_001109764 3.30 0.56 0.0006 5.141 Col6a6 collagen, type VI, alpha 6 NM_001102607 4.10 1.19 0.0010 3.246 Ntn4 netrin 4 NM_021320 9.84 3.24 0.0026 2.975 Pitx2 paired-like homeodomain 2 NM_001042504 19.54 7.02 0.0028 2.758 Twist1 twist family bHLH transcription factor 1 NM_011658 37.16 13.38 0.0035 2.765 Fgfr2 fibroblast growth factor receptor 2 NM_010207 19.88 7.65 0.0035 2.577 Clmp/ 9030425E11Rik CXADR-like membrane protein NM_133733 17.83 7.00 0.0058 2.527 Genes with ≥1.5 fold change up in the Sox9 cKO with restrictions (≥ 1FPKM in WT or Sox9 cKO) that have a SOX9 peak Gene symbol Full Gene Name Gene ID WT FPKM cKO FPKM p-value from Cuffdiff Fold change up in cKO Bhlhe40 Basic Helix-Loop-Helix Family, Member E40 NM_011498 12.17 56.04 5.00E-05 4.575 Fos FBJ murine osteosarcoma viral oncogene homolog NM_010234 2.45 13.97 0.0002 5.522 Dusp4 Dual specificity phosphatase 4 NM_176933 9.16 22.42 0.0072 2.431 92  Gene symbol Full Gene Name Gene ID WT FPKM cKO FPKM p-value from Cuffdiff Fold change up in cKO        Stc1 Stanniocalcin 1 NM_009285 4.32 9.66 0.0227 2.206 Ddit3 DNA-damage-inducible transcript 3 NM_001290183 11.93 72.50 0.0461 6.036 Nrg1 Neuregulin 1 NM_178591 2.98 6.50 0.0466 2.140 Junb Jun B proto-oncogene NM_008416 2.40 5.41 0.0510 2.201 Fhdc1 FH2 domain containing 1 NM_001033301 1.06 2.05 0.0801 1.859 Gramd1b GRAM domain containing 1B NM_172768 10.72 18.23 0.1035 1.694 Gm14005 Gm14005 predicted gene 14005 NR_028589 1.07 3.18 0.1101 2.803  93    Figure 5-10 Sox9 cKO valves have major changes in gene expression when compared to WT valves. A heat map generated using meV of the expression changes occurring between WT and Sox9 cKO valves. 94  CHAPTER SIX: SOX9 has functions involved in regulation of proliferation, transcriptional networks, and ECM formation during heart valve development  To determine the context-independent functions of SOX9, regions of DNA bound by SOX9 were compared amoung different tissues using publicly available data sets. Identification of regions that are commonly bound by SOX9 in multiple tissues suggests that these are potential context-independent transcriptional targets of SOX9. Consequently, the group of genes associated with commonly bound regions may represent a general function(s) of SOX9 across tissues. Furthermore, to identify tissue-specific (context-dependent) functions of SOX9 in the developing heart valves, genes associated with SOX9 bound regions in the AVC were compared with genes that have altered expression in the Sox9 cKO AVC. This analysis will help to reveal the specific roles of SOX9 in the developing heart valves.  6.1 SOX9 occupies regulatory regions of genes associated with proliferation As previously mentioned, Sox9 conditional mutant mice in different organ systems often have defects in proliferation and it seems that regulation of proliferation associated genes may be a common role for SOX9 across tissues. Although it is known that Sox9 mutants have proliferation defects, to date no direct transcriptional targets have been associated with this role in proliferation. To further delineate common functions of SOX9, SOX9 peaks in the AVC and limb were compared with a publicly available SOX9 ChIP-Seq dataset that was generated in HF-SCs (62). A Venn diagram of the three SOX9 ChIP-Seq libraries demonstrated that SOX9 bound at 293 identical genomic locations in all three libraries (Figure 6-1A), suggesting that SOX9 has common targets in AVC, limb and HF-SCs. GO analysis using the biofunctions category 95   Figure 6-1 Common SOX9 targets among developing tissues provide evidence for a role in proliferation. A. Venn diagram of SOX9 peaks in E12.5 AVC, E12.5 limb, and hair follicle stem cells (HF-SCs). B. The top 8 terms from GO analysis on SOX9 target genes overlapping in all and unique GO terms for each tissue. C. ChIP-qPCR (n=3) on SOX9 target genes bound in all associated with proliferation for SOX9 binding in the E12.5 AVC and limb or D. E12.5 lung, gut and liver (n=3 for Wasf1/2, Hdac1/2, n=2 for Junb, p53, Fgf11,Col2a1, Apoc, Tat, n=1 for Fos and Eed). Apoc3, Hnf1, Tat are negative controls regions not bound by SOX9.96  (Ingenuity) on the 293 common SOX9 peaks associated with target genes featured proliferation of cells as one of the top GO categories (Figure 6-1B, Table 6-1). In fact, 106 of the overlapping SOX9 target genes are included the in "proliferation of cells" GO category (Figure 6-2). Several examples of genes associated with “proliferation of cells” that were bound by SOX9 in all three libraries include Junb, Cops5, Fosl1, Fosl2 and Fos. Of note, several genes involved in cell proliferation had SOX9 binding sites in heart valve and limb only, such as Trp53 and Fgfr2. To validate the genes identified by the SOX9 ChIP-Seq libraries, ChIP-qPCR was performed on E12.5 AVC and limb to confirm that shared binding sites were occupied by SOX9 (Figure 6-1C). For the first time, it has been demonstrated that SOX9 occupies the regulatory regions of genes that are associated with proliferation and that this binding was shared among different cell types suggesting that these sites are context independent.  Given that SOX9 is required for the development of many organs and is known to be involved in progenitor proliferation (53), we anticipated that shared SOX9 binding sites regulating proliferation genes would also be occupied by SOX9 in other SOX9-expressing tissues. ChIP-qPCR for SOX9 examining the regulatory regions of proliferative genes was performed on the E12.5 lung, gut, and liver (Figure 6-1D). The ChIP-qPCR demonstrated that many of the proliferation associated shared binding regions were also occupied by SOX9 (Figure 6-1D) and suggests that SOX9 may have a common role in the regulation of cell proliferation across developing tissues.  6.2 Context-independent SOX9 binding regions in the AVC, limb, and HF-SCs To identify context independent activities of SOX9, GO analysis was performed on genes targeted by SOX9 from non-overlapping SOX9 peaks from each of the AVC, limb, and HF-SC  97  Table 6-1 The top 15 biofunctions identified by IPA on the genes with overlapping SOX9 peaks in the AVC, limb, and HF-SCs  Diseases or Functions Annotation p-Value # of Molecules minus 10*(LOG10 (p-value)) expression of RNA 1.12E-11 81 109.5078 expression of DNA 7.56E-09 61 81.21478 transcription 1.03E-08 68 79.87163 transcription of RNA 1.11E-08 67 79.54677 transcription of DNA 2.03E-08 58 76.92504 activation of DNA endogenous promoter 9.22E-08 46 70.35269 proliferation of cells 2.89E-07 106 65.39102 organismal death 1.42E-06 84 58.47712 infection by RNA virus 1.43E-05 37 48.44664 proliferation of tumor cell lines 2.33E-05 52 46.32644 Viral Infection 2.43E-05 56 46.14394 cell transformation 3.17E-05 19 44.98941 HIV infection 4.79E-05 31 43.19664  98    Figure 6-2 SOX9 targets genes involved in proliferation of cells. Ingenuity Pathway Analysis using biofunctions on genes targeted by SOX9 in all three ChIP-Seq libraries (E12.5 AVC, E12.5 limb, HF-SCs) that are associated with proliferation of cells (106 genes). Genes are subdivided by their gene products location within the cell or outside the cell.   99  libraries. Redundant terms were filtered out between categories (Figure 6-1B). HF-SC-specific SOX9 targets contained unique GO terms like stem cell division; hair follicle morphogenesis and cell fate commitment (Figure 6-1B) as previously reported. Limb-specific SOX9 targets revealed unique GO terms implicated in mesenchyme development, ECM organization and forelimb morphogenesis (Figure 6-1B). SOX9 AVC-specific target genes identified unique GO categories involved in DNA binding, cardiac neural crest cells (NCC) development, and ascending aorta morphogenesis (Figure 6-1B). Identifying genes involved in cardiac NCC in the AVC specific GO categories supports the model that many genes are shared in AVC and OFT development, such as Twist1, Hand2, and Pitx2. Overall, the genes associated with tissue-specific SOX9 bound regions strongly reflect the unique characteristics of each tissue and indicate that SOX9 plays important and context specific roles in different tissues which are linked with tissue identity.  6.3 Proliferation associated target genes are both activated and repressed by SOX9 Currently, little is known about how SOX9 regulates its target genes. Col2a1 is the most well characterized SOX9 target gene and its regulation by SOX9 is very complex (reviewed in (132)). Overall, SOX9 has been shown to be a transcriptional activator but recent evidence suggests that it can also act as a repressor. SOX9 can bind to gene regulatory regions as a monomer, in sex determination, or as a dimer during chondrogenesis (85,86). Furthermore, the SOX9 DNA binding site tends to be variable and degenerate with a consensus site of WWCAAWG. Given that the understanding of SOX9’s transcriptional regulation of its target genes is lacking, I examined SOX9 peak regions of a set of proliferation associated target genes (Hdac2, Prkaca, Rfwd3, Trp53, and Junb) to identify whether SOX9 acts as an activator or repressor on these 100  target genes and to determine the exact SOX9 DNA binding site. Target genes of interest were selected based on the GO category of proliferation of cells or regulation of cell cycle identified from common targets of SOX9. Although SOX9 targets have been identified by ChIP-Seq analysis, it is not known whether these sites are functional and whether it is activating or repressing the target gene.  Unfortunately, relevant cell lines for the heart valve and limb buds are limited and primary cultures or explants derived from these tissues are not easily transfectable. Therefore, HEK 293T cells were selected as they are easily transfectable and ideal for over-expression studies. To examine how SOX9 regulates select target genes, luciferase assays, a measurement of transcriptional activity, were coupled with and without SOX9 (via pDNA3-SOX9) over-expression in HEK 293T cells (Figure 6-3A). All of the SOX9 bound regions associated with Hdac2, Prkaca, Rfwd3, Trp53, and Junb were located at promoter regions. The average size of the SOX9 peak size for the selected targets was 200-300 base pairs. Reporter vectors were generated by cloning in the overlapping SOX9 peak regions (see primers in Table 2-1) identified in the AVC and limb SOX9 ChIP-Seq libraries into the pGL4b (promoter-less) luciferase vector. Titrations for SOX9 over-expression were performed on the pGL4-Hdac2 promoter vector and the optimal SOX9 concentration was determined to be 0.1 μg per well in a 24 well plate (data not shown). In the presence of SOX9, the Hdac2, Prkaca, Rfwd3, and Trp53 regulatory regions were activated by SOX9 by at least 1.5-1.7 fold whereas SOX9 inhibited Junb by 1.7 fold (Figure 6-3A).  Interestingly, the SOX9 peak region associated with the Hdac2 promoter contains a potential SOX9 dimer site. To determine if the dimer site in the Hdac2 promoter is in fact the SOX9 binding site and secondly if SOX9 is binding as a dimer, I mutated the SOX9 dimer site in  101    Figure 6-3 SOX9 regulates genes associated with cycle and proliferation. A. Luciferase assays on SOX9 binding regions (promoter bound) of cell cycle associated genes in HEK 293T cells with /without overexpression of SOX9. B. Site directed mutagenesis coupled with luciferase assays in HEK293T with/without overexpression of SOX9 of the SOX9 dimer site within the SOX9 peak located in the Hdac2 promoter region. 1st M indicates mutation of the first SOX9 site in the dimer. 2nd M indicates mutation of the second SOX9 site in the SOX9 dimer and DM indicates the mutation of both SOX9 sites in the dimer. C. A representative qRT-PCR for cell cycle and proliferation associated genes on WT and Sox9 cKO heart valves.  β-actin was used as a control for qRT-PCR. Ccdn1 is not a SOX9 target gene.  102  the Hdac2 promoter using site directed mutagenesis and generated three luciferase constructs. The first construct contained mutations in both SOX9 sites (double mutant, DM), the second contained a mutation in the first SOX9 site (1st site mutant, 1st M) and the last contained a mutation in the second SOX9 site (2nd site mutant, 2nd M). Based on previous literature, the CAA portion of the SOX9 consensus site (WWCAAWG) was mutated to ACC using primers containing the mutated sequence (Table 2-1). Using these vectors in luciferase assays, the induction of the Hdac2 promoter by SOX9 over-expression was eliminated upon mutation of both SOX9 sites (Figure 6-3B). Additionally, based on single site mutations of SOX9 sites in the Hdac2 promoter it appeared that the second site in the SOX9 dimer was the more critical SOX9 binding site for activation of Hdac2 (Figure 6-3B). Since the SOX9 DNA binding motif is highly degenerate, identifying the exact binding motifs of SOX9 will help to clarify the exact positions required for SOX9 binding and may aid in the discovery of additional SOX9 binding sites in other organ systems.  To further substantiate SOX9’s role in activation or inhibition of its target genes, I examined the expression level changes that occur following the loss of SOX9 in vivo. RNA isolated from the WT and Sox9 cKO AVC was used for qRT-PCR to examine the expression levels of the genes identified by GO to impact proliferation: Hdac2, Prkaca, Rfwd3, Trp53, Junb, Fgfr2, Akt2, Cdkn1b, and Ccdn1 (Figure 6-3C). Overall, no significant changes in the expression of proliferation associated genes were observed between WT and Sox9 cKO AVCs with the exception of Fgfr2. However, a trend towards a decrease for Hdac2, Prkaca, Rfwd3, and Trp53 mRNA expression levels was noted (Figure 6-3C). The SOX9 bound regions for Hdac2, Prkaca, Rfwd3, and Trp53 were activated by SOX9 (Figure 6-3A) in the luciferase assays except for Fgfr 2(data not shown) and matched the slight decreased in transcript levels in the RNA-Seq. 103  qRT-PCR revealed that Akt2 and Junb had no change in mRNA expression levels between WT and Sox9 cKO whereas Cdkn1b (also known as p27, a cell cycle inhibitor) had increased expression in the Sox9 cKO AVC (Figure 6-3C). Generally, the changes in transcript expression levels between WT and Sox9 cKO AVCs in the RNA-Seq and qRT-PCR matched for the SOX9 target genes associated with cell cycle/proliferation. However, there were a few exceptions such as Junb which was up-regulated in the Sox9 cKO AVC RNA-Seq library and Cdkn1b which had no change in expression level in the RNA-Seq (Appendix XI). Although no drastic changes were observed in gene expression of the proliferation associated genes between WT and Sox9 cKO AVC, it could be that even small to moderate changes in cell cycle and proliferation could have detrimental effects on organ formation during development.  6.4 SOX9 targets a network of TFs known to be involved in heart development Given that a major goal of this study was to discover SOX9 target genes that are critical for heart valve formation, up- and down-regulated SOX9 target genes identified by ChIP-Seq and RNA-Seq analysis were further analyzed by Ingenuity Pathway Analysis (IPA, Qiagen) for GO terms and functions specific to heart development. IPA highlighted enrichment in functions associated with transcription, cardiogenesis, and abnormal heart morphogenesis (Table 6-2). IPA was used to generate a relationship network between these three GO terms to determine if there were any commonalities between them (Figure 6-4). Lines indicate the type of relationship between the gene and GO term. For example, loss of Pitx2 (green colour of the molecule=loss, red=increase) leads to an activation of abnormal morphology of the heart (orange line) and inhibition of transcription and cardiogenesis (blue line) (Figure 6-4). Yellow lines indicate that data supports both activation and inhibition relationships for the gene and GO term. Notable TFs associated  104  Table 6-2 The top 20 biofunctions identified by IPA on SOX9 target genes with differential expression in the Sox9 cKO AVC.  Diseases or Functions Annotation p-Value Activation z-score # Molecules morphology of digestive system 3.00E-10 1.398 22 perinatal death 1.11E-09 2.453 22 abnormal morphology of skull 4.26E-09   13 expression of DNA 4.33E-09 -0.484 33 abnormal morphology of digestive system 4.51E-09 1.169 20 transcription of DNA 5.15E-09 -0.223 32 neonatal death 5.66E-09 1.997 18 organismal death 1.77E-08 3.71 45 cell death 2.37E-08 0.846 51 activation of DNA endogenous promoter 2.85E-08 -0.487 26 seizures 3.16E-08 0.308 16 morphology of head 3.29E-08 1.499 24 apoptosis 4.02E-08 1.548 43 seizure disorder 4.44E-08 -0.047 17 abnormal morphology of head 5.89E-08 2.219 23 tooth development 1.10E-07   8 morphology of body cavity 1.18E-07 1.513 27 cardiogenesis 1.43E-07   14 differentiation of connective tissue 1.51E-07 0.486 18 abnormal morphology of heart 1.73E-07   15   105    Figure 6-4 SOX9 ChIP-Seq and differential transcripts in the Sox9 cKO AVC comparison reveals critical transcriptional networks involved in valve formation. A network of transcription, cardiogenesis and abnormal heart morphology biofunctions identified by Ingenuity. Lines indicate the type of relationship between the gene and GO term. Ie. Loss of Pitx2 leads to an activation of abnormal morphology of the heart and inhibition of transcription and cardiogenesis. Yellow lines indicate that there is data support both activation and inhibition. Libraries performed in duplicate from pooled valves.  106  with heart development that are down-regulated in the Sox9 cKO AVC and are SOX9 targets in the AVC include Sox4, Hand2, Twist1, Foxp4, Mecom, and Pitx2. Interestingly, almost 20% of the top differentially expressed genes that are targeted by SOX9 are transcription factors and suggests that SOX9 may be regulating a network of transcription factors to drive heart valve development. In addition, several ECM components known to be important for heart valve formation such as Periostin and Elastin were reduced (Figure 6-4, Appendix XII). Up-regulated genes that were SOX9 targets also contain several TFs, for example, Bhlhe40, Ddit3, and Junb (Figure 6-4, Appendix XIII). Thus, indicating that SOX9 can likely act as both a transcriptional activator and repressor of critical factors based on ChIP-Seq and RNA-Seq analysis.  Our data indicates that SOX9 activates a network of critical TFs and ECM components that may have crucial functions that are required during heart valve development. To confirm that SOX9 acts as a transcriptional activator of the TFs, Mecom and Nfia, luciferase assays coupled with SOX9 over-expression were employed as described above (Figure 6-5A). SOX9 activated the Mecom and Nfia enhancers by 2.44 and 2.17 fold respectively (Figure 6-5A). This data supports that direct binding of SOX9 activates transcription of important TFs during heart valve development. Since SOX9 lies upstream of a group of known TFs involved in heart development, I determined whether the TFs, Sox4, Mecom, Twist1, Pitx2, Hand2, and Nfia transcript levels were reduced in the Sox9 cKO AVC by qRT-PCR (Figure 6-5B). Of note, Sox4, Mecom, Twist1, Lef1 and Tbx20 levels were significantly reduced in the Sox9 cKO AVC (Student’s T-test between WT and Sox9 cKO levels for each transcript, p>0.05). Two additional SOX9 targeted TFs important in heart valve development, Lef1 and Tbx20, were also reduced in Sox9 cKO AVC but were below the 1.5 fold change cut-off used for the RNA-Seq analysis (Figure 6-5B). In all cases, expression of these TFs were reduced in Sox9 cKO AVCs compared  107    Figure 6-5 Sox9 activates transcription factors that are known to be essential for heart valve development. A. Luciferase assays for SOX9 enhancer sites associated with the transcription factors Mecom and Nfia  on HEK293T with SOX9 overexpression. N=3.  B. qRT-PCR for Sox9 and selected target TFs on WT and SOX9 cKO AVCs. Sox9, Twist1, Lef1, Tbx20 have an n of 8 or greater and Sox4, Mecom/Evi1, Pitx2, Hand2, and Nfia have an n of 3 or greater. C. FPKMs from the RNA-Seq analysis on WT and Sox9 cKO heart valves illustrating similar expression patterns to those identified in the qRT-PCR.  *significance by a two tailed Student’s T-test. 108  to WT. Alterations in transcript levels identified by qRT-PCR correlated to the expression changes seen in the RNA-Seq libraries on WT and Sox9 cKO AVC (Figure 6-5C).  To verify that loss of SOX9 in heart valves causes downstream effects that disrupt valve development, we focused on the known and essential TF, Twist1. Quantitative RT-PCR confirmed that the Twist1 transcript is reduced at E12.5 and also at an earlier time point (E10.5) in the Sox9 cKO AVC (Figure 6-6A, B). Using in situ hybridization, Twist1 mRNA was specifically expressed in the valve mesenchyme in the E12.5 AVC and was reduced to about half of the WT levels in the SOX9 mutant valves (Figure 6-6C). This was similar to the level of reduction in Twist1 mRNA seen in mutant valves at E12.5 via qRT-PCR (Figure 6-6B). This data suggests that SOX9 may fine tune the levels of Twist1 in the developing heart valves. Mecom/Evi1was another TF of interest, as generation of a hypomorphic allele of Mecom/Evi1 demonstrated that it has an important role in OFT valve formation (131). Mecom/Evi1 is expressed in AVC but no role in AVC valve development was described. Quantitative RT-PCR for Mecom/Evi1 confirmed that the transcript is enriched in the AVC (Figure 6-7A). Since I had already confirmed that the Mecom/Evi1 transcript was reduced in the Sox9 cKO valves, I decided to examine the protein levels of EVI1 using immunofluorescence with an antibody specific to EVI1 in the developing WT heart valves from E12.5-E14.5 (Figure 6-7B). EVI1 protein was specifically expressed in regions of the valve mesenchyme that was condensing at E12.5 and remained expressed there until E14.5 (Figure 6-7B, arrowheads). Of note, EVI1 protein was reduced in the Sox9 cKO hearts except within regions where SOX9 had escaped Cre excision based on the location in serial sections (Figure 6-7C, asterisk). This data demonstrates that SOX9 directly regulates Mecom/Evi1, and that EVI1 is reduced in the absence of SOX9.  109    Figure 6-6 The critical EMT regulator Twist1 is reduced in the Sox9 cKO heart valves. Taqman assays for A. Sox9 and B. Twist1 relative to Gapdh on E10.5 (n=2) and E12.5 (n>6) Sox9fl/fl (WT) and Sox9fl/fl;Tie2-Cre (Sox9 cKO) AV-enriched regions. C. In situ hybridization for Sox9 and Twist1 on WT (Sox9fl/fl;+/+) and Sox9 cKO (Sox9fl/fl;Tie2-Cre/+) hearts. AS=antisense DIG-labelled probe. S=sense DIG-labelled probe. N=3.  110    Figure 6-7 EVI1 is enriched in the AVC specifically in condensing mesenchyme and reduced upon deletion of Sox9. A. A representative qRT-PCR on WT E12.5-E14.5 AVC, ventricle (V), and atria (A) for Mecom/Evi1 transcript illustrates AVC enrichment over both V and A during embryonic valve development. B. Immunofluorescence (IF) for EVI1 (green) on E12.5-E14.5 hearts. DAPI (blue) was used to stain nuclei. Arrowheads indicate EVI1 expression in the condensed mesenchyme of the heart valves. C. IF for EVI1 on E12.5 WT and Sox9 cKO hearts. *EVI1 positive regions. Arrowhead =SOX9 negative and EVI1 negative regions. 111  6.5 Additional known roles for SOX9 in EMT and ECM organization SOX9 has been shown to induce EMT along with the help of other factors in cell types such as the neural crest (67) and since the embryonic heart valves require EMT to populate them with mesenchyme, I decided to analyze EMT markers in the WT and Sox9 cKO heart valves. As a first step, qRT-PCR was performed on E12.5 AVC from WT and Sox9 cKO for the epithelial markers, Cdh1 (E-cadherin) and Cdh5 (VE-cadherin) and mesenchymal markers, Vim (Vimentin) and Cdh2 (N-cadherin) (Figure 6-8A). There were no significant differences in epithelial or mesenchymal markers mRNA expression levels between WT and Sox9 cKO AVC. N-cadherin mRNA levels via qRT-PCR had a slight decrease, however when examining the transcript levels in the WT and Sox9 cKO RNA-Seq libraries there were no differences in any of the markers (Figure 6-8B). The E-cadherin transcript was below the level of detection in the RNA-Seq libraries and consequently was not included (Figure 6-8B). Although the transcript levels did not change between WT and Sox9 cKO valves, it is possible that epithelial and mesechymal markers could be changing at the protein level. To examine epithelial markers at the protein level, immunofluorescence was performed for CD31 (Pecam1, epithelial marker, BD Biosciences) on E10.5 WT and Sox9 cKO hearts (Figure 6-8C). Overall, epithelial markers were not changed at the transcript or protein level in the Sox9 cKO AVCs and suggest that SOX9 was not essential for maintaining the expression these markers in the heart valves. This was not entirely surprising as SOX9 negative mesenchyme was found in Sox9 cKO valves and suggests that SOX9 is not essential for the initial stages of EMT or for the epithelium.  Another critical function of SOX9 in the limb is in ECM generation and organization and the ECM is very important for heart valve development, and therefore I decided to investigate if there are any differences in ECM molecules in the Sox9 cKO hearts. Quantitative RT-PCR for  112    Figure 6-8 EMT markers are unaffected by the loss of Sox9 in the developing heart valves. A. qRT-PCR on WT and Sox9 cKO E12.5 heart valves for Vimentin, N-cadherin, E-cadherin and Vascular endothelial cadherin transcripts. β-actin was used as a control. N=3. B. FPKMs from the RNA-Seq analysis on WT and Sox9 cKO heart valves illustrating similar expression patterns to those identified in the qRT-PCR. C. Immunofluorescence for CD31 (green) on E12.5 WT and Sox9 cKO hearts. DAPI (blue) was used to stain nuclei.   113  Postn (Periostin), Eln (Elastin), Fbn1 (Fibrillin 1), and Mgp (Matrix Gla Protien) on WT and Sox9 cKO AVC revealed that ECM molecules were reduced in the Sox9 cKO valves with the exception of Fbn1 (Figure 6-9A). These alterations in transcript levels of the selected ECM molecules matched the RNA-Seq libraries expression levels perfectly (Figure 6-9B). Additionally, the RNA-Seq libraries on WT and Sox9 cKO AVC demonstrated that Col9a1, Col9a3, Prelp, Col6a6, Mfap4, Matn4, etc. were also greater than 1.5 fold down-regulated in Sox9 cKO valves when compared to WT (Appendix XI). Immunofluorescence was performed for PERIOSTIN on WT and Sox9 cKO hearts to determine if the protein level was also reduced in the Sox9 cKO AVC (Figure 6-9C). PERIOSTIN levels did not appear to be reduced in the Sox9 cKO AVC and seemed to be up-regulated compared to overall valve area (Figure 6-9C). Additional studies are required in order to determine the exact role of SOX9 in the regulation of ECM molecules as many of these molecules are carefully controlled at the protein level as well. Overall, the data suggests that SOX9 plays a role in regulating the transcript level of key ECM genes during heart valve development. 114    Figure 6-9 ECM molecules mRNA levels are reduced in the Sox9 cKO heart valves. A. qRT-PCR on WT and Sox9 cKO E12.5 heart valves for Periostin (Postn), Elastin (Eln), Fibrillin 1 (Fbn1) and Matrix Gla protein (Mgp) transcripts. β-actin was used as a control. N=3. B. FPKMs from the RNA-Seq analysis on WT and Sox9 cKO heart valves illustrating similar expression patterns to those identified in the qRT-PCR. C. Immunofluorescence (IF) for PERIOSTIN and SOX9 on E12.5 WT and Sox9 cKO hearts. DAPI (blue) was used to stain nuclei.115  CHAPTER SEVEN: Discussion  7.1 Thesis overview Taken together, we have shown that shared SOX9 binding sites identifies a common function of SOX9 in regulating proliferation in a diverse number of tissues; SOX9 binding was biased toward promoter regions, SOX9 tissue-specific binding sites with transcriptome data from tissues lacking SOX9 supports discovery of critical SOX9 targets; and SOX9 activates a core network of TFs required for heart valve development.  In this thesis, I have investigated the global SOX9 binding sites in the E12.5 heart valves and limb and found that the transcriptional programs initiated by SOX9 in the developing heart valves and limb share numerous similarities in SOX9 binding locations and downstream target genes. Analysis of SOX9 binding in the valves and limb also revealed that SOX9 preferentially binds at or near promoters, can bind as both monomers and dimers, and that SOX9 binding regions contain motifs for numerous potential co-factors such as HIF1α, FOXs, and RUNX proteins. Generation of two endothelial specific conditional Sox9 knockout mouse models using VE-Cre and Tie2-Cre crossed with Sox9fl/fl revealed two potentially different roles for SOX9 during heart valve development. The VE-Cre crossed with Sox9fl/fl suggests a role for SOX9 in the early stages of EMT and generation of mesenchyme while Tie2-Cre crossed with Sox9fl/fl indicates a role for SOX9 in maintenance of the mesenchyme and mesenchyme differentiation. Analysis of the Sox9fl/fl;Tie2-Cre (Sox9 cKO) mutant valves demonstrated that Sox9 mutants have reduced AVC cushion size, proliferation defects and defects were ultimately embryonic lethal. Analysis of shared SOX9 binding sites among different tissues revealed a role for SOX9 in proliferation. Furthermore, transcriptome analysis on WT and Sox9 cKO valves compared 116  with SOX9 transcriptional targets in the heart valve highlighted critical roles for SOX9 in the regulation of heart valve specific transcription factor networks and ECM molecules. Overall, I have identified critical characteristics of SOX9 binding and identified its transcriptional targets, identified roles for SOX9 in proliferation, transcription factor network regulation and maintenance of the ECM in the developing heart valves.  7.2 SOX9 occupies the promoter and upstream regulatory regions associated with thousands of target genes including its future co-factors SOX9 is known to be essential for heart valve development as the loss of SOX9 specifically in the valves is embryonic lethal (46). Even though SOX9 has a critical role in the heart valves, the protein expression pattern has not been well characterized throughout the literature with the majority of studies focussed on the early stages of valve formation. For this reason, I analyzed SOX9 protein expression throughout embryonic valve development (E9.5-E16.5) using immunofluorescence. SOX9 was highly enriched in cardiac cushions and the entire AV valve regions at all stages examined and suggests that SOX9 likely has multiple roles in heart valve formation as the AV valves are undergoing diverse cellular processes including proliferation, differentiation and remodeling during this time. Immunofluorescence for SOX9 on the WT and Sox9 cKO hearts also helped to support the specificity of the SOX9 antibody for the SOX9 ChIP-Seq libraries. In addition, SOX9 was found specifically in the limb, lung, and somites in the E12.5 whole embryo using immunofluorescence as expected from previous publications.  To gain a better understanding of the role of a specific transcription factor, one can examine its downstream transcriptional targets. To analyze SOX9 binding in the embryonic heart valves and limb, two genome-wide SOX9 ChIP-Seq libraries on the E12.5 AVC and limb were 117  generated. To our knowledge, we are the first group to perform ChIP-Seq for SOX9 on the embryonic heart valves and limb buds. To date, studies that have performed ChIP-Seq on mouse embryonic tissues are limited (143-145). Only three groups within the last year has analyzed SOX9 binding sites using ChIP-Seq on HF-SCs (126), on the vertebral column (125), and on postnatal rib chondrocytes (129). Of note, the data was not released for the rib chondrocyte study. Additionally, in the last month, two other studies have published global SOX9 binding sites in a chondrocyte cell line (146) and in postnatal rib chondrocytes (147) using the same SOX9 antibody. Global analysis of SOX9 binding in similar but different developing tissues has revealed novel common- (context-independent) and tissue-specific (context-dependent) mechanisms of gene regulation.  For the first time, we have identified over 2400 novel SOX9 target genes in the mouse embryonic heart valves. In addition to the handful of known SOX9 target genes in the limb, we have identified 5700 SOX9 target genes during chondrogenesis. The SOX9 ChIP-Seq limb library has identified Col2a1, Col9a1, Col11a2, Acan, Hapln1 (CLP), Comp, and Mia1 (Cd-rap) (79-84,87,88), known cartilage-specific targets of SOX9, and thus verifies that the SOX9 ChIP-Seq libraries are of high-quality and highly specific for SOX9 binding. Interestingly, SOX9 targets a number of its known co-factors during chondrogenesis, including Sox5, Sox6, Barx2, and Zfp219 (48,148,149), and suggests that SOX9 has the ability to enhance its own specificity to its cartilage-specific target genes. In the limb, SOX9 may regulate factors that are known to control its own expression. For instance, in the limb SOX9 ChIP-Seq data, SOX9 targets Bmpr1b and Notch1 which have been shown to have a role in the regulation SOX9 expression during chondrogenesis (150,151). 118   A number of novel characteristics about SOX9 DNA-binding behavior within the genome were discovered in the developing heart and limb. For example, SOX9 binding sites are widely distributed throughout the genome but are biased to promoter regions of its target genes particularly when a SOX9 binding site is shared in multiple tissues. Additionally, this work demonstrated that SOX9 preferentially binds as monomers in the heart valves and that the SOX9 dimer motifs were more prevalent in the limb than the AVC. Dimer motifs were expected to be more abundant in the limb as dimer sites have been previously reported for chondrogenic genes required for limb development (86). Although a SOX9 dimer motif has been described in chondrogenic genes, we are the first to present a positional weight matrix for the SOX9 dimer motif in the developing limb and heart valves.  Several studies have suggested that heart valve development shares many similarities with cartilage and bone formation in the developing limb (37,38). Comparisons of the SOX9 DNA bound regions (peaks) in the AVC and limb by using Venn diagrams revealed that 782 SOX9 binding sites directly overlap in the AVC and limb and supports that these developmental programs are similar. Furthermore, when SOX9 DNA binding regions are assigned to genes and these target genes were compared between the AVC and limb using a Venn diagram, there were 1605 target genes shared between the two tissues, further highlighting the similarities in these two tissues. SOX9 DNA binding regions that directly overlap in the AVC and limb suggests a common (context independent) function in these two tissues. GO analysis on the genes in the AVC and limb with directly overlapping DNA bound regions identified numerous metabolic functions, transcription, and regulation of cell cycle and further supports the notion of context-independent functions of SOX9. Regulation of cell cycle was of particular interest as these two tissues are rapidly growing during development and several genes found in this category include 119  Trp53 (p53), Junb, Rfwd3, Hdac1/2 and Eed. Based on the higher number of commonly targeted genes versus shared SOX9 binding sites in the valves and limbs, SOX9 likely has tissue-specific binding sites for the heart valves and limb for the same target genes. GO analysis on the AVC and limb libraries revealed many context dependent functions such as regulation of WNT signaling, atrioventricular valve development, and atrial septum morphogenesis in the AVC and mesenchyme development, chondrocyte differentiation, and connective tissue development in the limb.   SOX transcription factors are known to bind in complexes with other SOXs or other transcription factors in order to efficiently regulate their target genes (47). To get an idea of the potential co-factors of SOX9, the DNA sequences under SOX9 peaks were analyzed using oPOSSUM to search for enriched transcription factor motifs. For AVC, limb, or shared SOX9 peaks the top motifs by z-score that were found in all three analyses were motifs for NFYA, SOX9 and ARNT. It was reassuring to have SOX9 as one of the top enriched motifs as these are SOX9 DNA binding sites identified by ChIP-Seq. Enrichment of the NFYA motif is expected since SOX9 binding is biased to the TSS/promoter regions and NFY is known to bind to promoters to activate transcription (reviewed in (152)). Interestingly, the ARNT motif was enriched in SOX9 DNA binding regions for all three analyses. HIF1 is a heterodimer made up of an oxygen regulated α subunit and a β subunit called ARNT (153). Both hypoxia and HIF1a are known to play a role in the development of the limb buds and heart valves (154,155). The inactivation of HIF1a in limb bud mesenchyme leads to decreased levels of SOX9 and HIF1a can bind to the SOX9 promoter, suggesting SOX9 is directly regulated by HIF1a (154). It is possible that HIF1a activates SOX9 and then together they regulate genes involved in hypoxia. Many of the motifs identified in the SOX9 DNA binding regions were found in all three libraries 120  however the degree of enrichment of these motifs changes. Other motifs of note were the other SOX, FOX, RUNX, and NFI factors. Further analyses will be aimed at understanding the relationship between these co-factors and SOX9 and their co-targeted genes.  Analysis of transcription factors using ChIP-Seq is becoming more prevalent. This data can be used to compare with the SOX9 ChIP-Seq data sets as a way to identify common functions for SOX9 or to identify co-factors and their downstream co-targeted genes. To examine common functions for SOX9 among different cell types, I compared our SOX9 ChIP-Seq data sets with two publicly available ChIP-Seq data sets for SOX9 on HF-SCs (126) and the E12.5 vertebral column (125). Comparison of SOX9 DNA binding regions in the AVC and limb with the HF-SCs revealed a high level of similarity between these cells, which was more than expected based on the functions of these different cell types. This may indicate that SOX9 positive mesenchyme progenitor cells in the AVC and limb have stem cell-like functions during development. Of note, the SOX9 antibody (Millipore, AB5535) used for ChIP-Seq on HF-SCs was the same antibody used for the SOX9 ChIP-Seq on the E12.5 AVC and limb. The E12.5 vertebral column undergoes endochondral ossification of the cartilage template comparable to the development of the embryonic limb. Due to this, it was expected that the SOX9 limb DNA binding sites would share a higher degree of overlap. However, SOX9 DNA binding sites in the HF-SCs and the limb had more overlap than SOX9 DNA binding sites in the E12.5 vertebral column and limb. On the other hand, the vertebral column study used a different SOX9 antibody (R&D Systems AF3075), which may account for the lower degree of overlap between libraries or it may be that vertebral column cells have a very different SOX9-initiated transcriptional program from the limb. Comparisons of SOX9 ChIP-Seq data sets in different cell types 121  highlight the extremely context dependent role of SOX9 and supports that regulation of SOX9 and of its target genes is likely dependent on the binding of its co-factors in different tissues.  In the embryonic limb SOX9 and SOX5/6 are known to work together to regulate chondrogenic genes like Aggrecan as the SOX trio (48) and therefore I compared our SOX9 ChIP-Seq data to SOX4 ChIP-Seq data as it is expressed in the developing heart valves (130) and SOXC factors such as SOX4 are also known to be important in limb development (156). The only SOX4 ChIP-Seq data available used B-cells as a source of tissue (127). Remarkably, the degree of overlap between the SOX9 DNA binding sites in the AVC and limb and SOX4 DNA binding in B cells was as high as the comparisons with SOX9 HF-SC DNA binding sites. This data suggests that SOX9 and SOX4 could be functioning together to regulate their targets and may have similar roles as they are bound in same regions of the genome even within different cell types. If the SOX4 ChIP-Seq was performed on a more relevant cell type, the degree of overlap between SOX9 and SOX4 DNA binding sites would likely be higher. Since SOX4 is a direct target of SOX9, the ability of SOX4 to act as a co-factor of SOX9 would likely be dependent on SOX9-mediated activation of SOX4 and consequently SOX4 and SOX9 may work together on genes necessary for later stages of heart valve development.  An EVI1 (Mecom) motif was identified by oPOSSUM as a potential co-factor of SOX9 and has been shown to have a role in heart valve development (131). The only publicly available EVI1 ChIP-Seq data set was on human OC cells (128). The comparison of SOX9 and EVI1 DNA binding sites did not show a high degree of overlap between these tissues. However, a follow up study from the same group that generated the EVI1 ChIP-Seq, demonstrated that genes known to be implicated in congenital heart defects were targets of EVI1 in OC cells (131). SOX9 peaks that contain EVI1 motif suggest that SOX9 and EVI1 may function together to 122  regulate valve-specific genes during heart valve development. Overall, this analysis suggests that SOX4 and EVI1 may function as co-factors together with SOX9 to regulate different subsets of genes during development. Future studies will be directed at understanding the critical co-factors of SOX9 and how they regulate their target genes together during heart valve and limb development.  7.3 SOX9 negative mesenchyme was absent in the Sox9fl/fl;VEC heart valves The Sox9 mutant (Sox9fl/fl;VEC) was generated to determine how SOX9 transcriptional target genes are affected by the loss of SOX9 in the developing heart valves. To my surprise, no SOX9 negative mesenchyme cells could be identified in the developing heart valves at any of the time points that were examined. Initially one would assume that the VE-Cre deletion was inefficient and was not deleting in the AVC mesenchyme cells, however the Sox9 mutants had significantly smaller cardiac cushions suggesting that Cre was in fact deleting in at least some portion of the cells. Although SOX9 should be deleted as early as E9.5 in the AVC, SOX9 has been shown to be a long lived protein and perhaps explaining the presence of SOX9 positive mesenchyme in Sox9 mutants. Additionally, examination of Sox9 transcript levels in the Sox9 mutant AVCs revealed that Sox9 transcript levels were not significantly affected when compared to WT.  So, why are the Sox9 mutant valves so much smaller than the WT? One possibility is that when Sox9 is deleted by VE-Cre in the endothelium, the cells undergo apoptosis immediately and never become mesenchyme and as a result fewer mesenchymal cells are migrating into the AVC. SOX9 positive mesenchyme present in the Sox9 mutant AVCs could be due to cells escaping complete VE-Cre deletion and proliferating. In order to completely delete Sox9 both alleles have to be deleted and it is possible that Cre is only efficiently deleting one allele. Due to 123  the lower numbers of mesenchymal cells in the valves, the cushion size was reduced compared to WT. However, data from other SOX9 mutant mice suggests it is possible to detect apoptotic SOX9 negative cells. For example, when Sox9 was deleted in the limb, they found that apoptotic domains were expanded in SOX9 mutants within a specific zone of chondrocytes (45). On the other hand, deletion of Sox9 in the heart valves using another epithelial-specific Cre (Tie2-Cre) did not identify any differences in the levels of apoptosis between WT and Sox9 mutants but rather they found decreased levels of proliferation in cells with Sox9 deletion (46,69). We also observed this in our work on the Sox9 cKO using Tie2-Cre. Although apoptosis of SOX9 negative cells may not be a likely explanation for the defect seen in the Sox9fl/fl;VEC mutant valves as these studies examined the entire AVC cushions. It may be that increased apoptosis only occurs in the AV endocardium or specific regions of the AVC ie. proliferating versus differentiating mesenchyme. It is difficult to distinguish these regions within the AVC as there are currently no known markers for proliferating and differentiating zones of the valves.  Another possibility for the less severe phenotype and presence of SOX9 positive mesenchyme in the Sox9 mutants is that there may be contribution of SOX9 positive cells from other mesenchymal cell sources in the heart such as the epicardium and the mesenchymal caps on the atrial and ventricular septums. These different cell types could migrate into the AVC region as a result of the mutant heart being under stress from insufficient blood circulation and increased pressure/demand on the heart. One famous example of other cell types migrating into the heart from other regions of the embryo is in OFT development where the neural crest cells migrate into the OFT and give rise parts of the valves and great arteries (157). The contribution of SOX9 positive mesenchyme from the septal mesenchymal caps to Sox9 mutant valves is less likely as the mesenchymal caps of the septums are derived through EMT (124) and would be 124  affected by the deletion of Sox9 by VE-Cre. The involvement of the epicardial cells migrating into the Sox9 mutant valve region is much more likely as epicardial cells are derived from a completely separate lineage (158,159) and would not be affected by VE-Cre. Epicardial cells have already been shown to migrate in the AV valve region in WT valves as another group demonstrated by lineage tracing that they contribute to the parietal leaflet of the AV valves at later stages of heart valve development (160). However, further studies using lineage tracing would be required to demonstrate whether epicardial cells can migrate into Sox9 mutant AV valves. Another alternative is that the VE-Cre is less efficient at driving recombination than Tie2-Cre and to rule out this possibility further experiments must be performed.  Although SOX9 protein is maintained during later stages of heart valve development in the Sox9 mutant valves, it appears that the hearts were undergoing some forms of stress due to the decreased valve mesenchyme cell numbers. In postnatal Sox9 mutant heart valves, the chamber size was decreased and ventricular thickness appeared to be greatly increased. The increased ventricular thickness may be indicative of cardiac hypertrophy and therefore increased and prolonged stress in the Sox9 mutant hearts when compared to WT. The cardiac hypertrophy seen in the postnatal Sox9 mutant myocardium might suggest that the Sox9 mutant valves are not functioning properly and leading to the increased stress on the heart. Given that and the fact that Sox9 mutants survived to adulthood, I wanted to see if older adult Sox9 mutants of approximately one year show signs of heart valve disease such as altered structure/composition of the ECM of the adult heart valves, calcification, or increased thickness of the developing heart valves. Hematoxylin and eosin staining on the adult Sox9 mutant valves demonstrated that the mutant valves had a slight thickening at the base of the valves and may be indicative of floppy valves that can lead to regurgitation of blood flow. However, the differences in valve thickness 125  were minimal and animal numbers were not high enough to provide significance. Further analysis of adult Sox9 mutant valves is required to determine if there is any indication of valve disease. Surprisingly, the adult Sox9 mutant ventricular myocardium did not show the overtly increased thickness as seen in the postnatal Sox9 mutant hearts. It is possible that Sox9 mutant mice that survive up to a year may not develop cardiac hypertrophy and Sox9 mutant mice that do develop signs of cardiac hypertrophy may not survive until this time point. In addition, the variation seen in adult Sox9 mutant hearts could also be due to inefficient deletion by VE-Cre or the variability in timing of Cre deletion. Further analysis of the Sox9 mutant mice is required to determine if there are any signs of heart valve disease or if these mice develop a way to recover from the loss of SOX9 positive mesenchyme during heart valve development.  7.4 Confirmation that Sox9 cKO (Sox9fl/f;Tie2-Cre) valves have decreased size and proliferation that leads to major heart defects and embryonic death Given that the Sox9 mutants derived from VE-Cre deletion did not result in Sox9 negative mesenchymal cells in the AVC, I decided to look into other options for Sox9 deletion in the valves. Previous work demonstrated that Sox9 deletion using Tie2-Cre leads to embryonic lethality between E11.5-14.5 with hypoplastic cardiac cushions, decreased mesenchyme proliferation and altered ECM (46). A similar phenotype was observed in this study and I noted several additional phenotypes not mentioned by the previous work such as a delay in septation and decreased myocardial thickness in the Sox9 cKO. The delay in septation may occur due to the decreased AVC cushion size and therefore making it more difficult for septal fusion with the AVC. Alternatively, since Tie2-Cre deletes SOX9 in the mesenchymal cap of the atrial septum as it is derived through EMT (124), the migration/growth of the septum guided by the 126  mesenchymal cap towards the AVC cushions could be affected. The size of the atrial mesenchymal cap was greatly reduced in size and almost absent in some Sox9 cKO hearts while their WT counterparts’ septums had already fused. Further analyses would be required to determine the exact cause of the delayed septation in Sox9 cKO hearts.  The decreased myocardial thickness was a bit perplexing as SOX9 is not expressed in the myocardium. The epicardial population surrounding the myocardium express SOX9 which is not affected by Tie2-Cre deletion and these cells do not migrate into the myocardium until later stages of heart development to form the coronary vasculature, smooth muscle cells, and cardiac fibroblasts (124). Since there are no cells within the myocardium that would be affected by Tie2-Cre deletion, it would be more likely that the thinning of the myocardial walls may be a defect secondary to valve abnormalities that lead to increased stress or pressure on the heart due insufficient regulation of blood flow. Overall, the data supports that SOX9 is a critical regulator of heart valve formation as found previously by the Yutzey laboratory. To understand the septal and myocardial defect, further studies of the embryonic Sox9 cKO hearts using echocardiography to determine how the hearts chambers and valves are functioning and to detect areas of damage or Doppler ultrasound to examine blood flow.  Unlike the VE-Cre, the Sox9 cKO heart valves generated using Tie2-Cre has SOX9 negative mesenchyme that can be easily detected. Although both VE-Cre and Tie2-Cre are presumably both endothelial specific Cre driver strains, they have very different outcomes in the heart where Sox9fl/fl;VE-Cre (Sox9 mutants) are not embryonic lethal and survive to adulthood with smaller valves that contain SOX9 positive mesenchyme whereas the Sox9fl/fl;Tie2-Cre (Sox9 cKO) are embryonic lethal with SOX9 negative mesenchyme in the heart valves. However, Sox9 deletion by VE-Cre also suggests that SOX9 may have a very critical role in the 127  endothelium as endothelial cells with Sox9 deletion were never detected as SOX9 negative mesenchyme in the heart valves. This indicates that EMT may have never occurred in the VE-Cre or the Sox9 deleted cells perished immediately following EMT. This leads to the question: why do the Sox9 cKO valves have such a different phenotype from the Sox9 mutant valves? One possibility is that in the Tie2-Cre driver strain, Cre is still expressed in the endothelium and early mesenchyme cells that have just undergone EMT whereas the VE-Cre expression is more restricted to the endothelium. Thus, Tie2-Cre allows for a longer time for Sox9 to be deleted in the AVC mesenchyme thus giving a more severe embryonic phenotype.  The Tie2-Cre driver strains was not completely efficient in Sox9 deletion as SOX9 positive mesenchyme was detected in the Sox9 cKO heart valves and suggests some cells escaped deletion by Cre and were able to populate the heart valves. This is a major caveat in using these Sox9 cKO mice to understand the role of SOX9 in the developing heart valves and therefore careful analyses of the extent of Sox9 deletion in the heart valves is required. For downstream analyses, the Sox9 cKO valves were evaluated for SOX9 protein or Sox9 transcript levels to determine which Sox9 cKOs had efficient loss of SOX9.  Before proceeding onto additional analyses, I confirmed that the Sox9 cKO heart valves had a decrease in cell proliferation as was seen in the previous publication (46). I observed that the number of phospho histone H3 positive cells was reduced in the AVC region of the Sox9 cKO hearts when compared to WT. I next looked at a cell cycle marker, cyclin D1. To progress through cell cycle appropriately and to continue to proliferate quickly, cyclin D1 must be rapidly moved between the nucleus and cytoplasm and subsequently degraded (reviewed in (161)). Interestingly, cyclin D1 was found in a subset of SOX9 negative mesenchymal cells and cyclin D1 was localized within the cytoplasm in the Sox9 cKO heart valve. Similar to our findings, 128  another study published that knockdown of SOX9 in rat mesenchymal stem cells caused reduced proliferation and increased levels and stability of cyclin D1 (60). Cyclin D1 is not a target of SOX9 in the AVC however it could be that one of its downstream regulators are affected by loss of SOX9, such as the F-box proteins (such as Fbxw8 that is targeted by SOX9) that are known to degrade cyclin D1 in the cytoplasm. Also, the lack of cyclin D1 degradation could account in part for the slowed/reduced proliferation in the Sox9 cKO heart valves.  7.5 Identification and regulation of SOX9 target genes in the developing heart valves Although many studies have alluded to the role of SOX9 in proliferation and other cellular processes, there has been little progress in understanding the downstream targets of SOX9 and how gene expression is altered in the absence of SOX9. Therefore, I generated transcriptome profiles from the WT and Sox9 cKO heart valves to determine which transcripts were altered upon loss of Sox9 in the E12.5 heart valves. Transcriptome analysis on E12.5 WT and Sox9 cKO valves revealed that down-regulated genes in the Sox9 cKO valves were heavily involved in processes critical for heart development such as mesenchymal cell proliferation, EMT, and endocardial cushion morphogenesis whereas up-regulated genes were mostly involved in stress related functions. The down-regulated category of genes had the most heart valve specific functions and correlates nicely with the fact that SOX9 is known to be an activator more than a repressor of transcription. To rule out any bias in the RNA-Seq libraries, I compared the RNA-Seq data to a published list of AV-enriched genes (142) and found that approximately 60% of these genes did not substantially change in expression levels and supports that there is not an overall loss of valve specific genes in the Sox9 cKO valves. Housekeeping genes are also not drastically altered in the Sox9 cKO valves. 129   To discover critical targets of SOX9 in the heart valves, I compared genes with altered expression in the Sox9 cKO (>1.5 fold change) with the gene targets identified in the SOX9 AVC ChIP-Seq library. This analysis revealed 139 genes that were both a SOX9 target in the AVC and have altered expression in the Sox9 cKO valves. Of these, three quarters were down-regulated and one quarter up-regulated in gene expression. In general, the changes in gene expression of SOX9 targets in the Sox9 cKO AVC were not substantial. This suggests that SOX9 is involved in fine tuning gene expression rather than acting as an on/off switch. It is also possible that SOX9 may have another unknown role other than regulating transcriptional activity of its targets. For example, SOX9 has been suggested in other systems to alter chromatin dynamics near regions densely populated with enhancers termed super enhancers (63). Additionally, SOXE factors like SOX10 have been shown to have DNA bending capabilities (162) and it could be that SOX9 also bends DNA to bring complexes of transcription factors together to regulate gene expression.  7.6 SOX9 directly controls genes associated with proliferation across cell types To better understand SOX9’s context independent roles in multiple cells types, I compared three SOX9 ChIP-Seq data sets (AVC, Limb, and HF-SCs) and showed that nearly 300 SOX9 binding sites were located at the exact same genomic location across these three different tissues. Nearly a third of these shared SOX9 binding sites are associated with genes that are involved in proliferation, which supports the potential role for SOX9 in maintaining a proliferative state during embryonic development. Numerous publications over the years have drawn links between SOX9 and proliferation. However, a direct mechanistic connection via its transcriptional targets has remained elusive. In AVC, limb and HF-SCs, SOX9 directly targets the TFs: Fos, FosL1, 130  and FosL2 suggesting that SOX9 can regulate their mRNA expression levels. FOS and JUN family members heterodimerize, to form the AP-1 complex, which is known to regulate cell proliferation and cell survival, in part via cyclin D1 expression (163). In mesenchymal stem cells, a stable SOX9 knockdown caused reduced proliferation, delayed S-phase progression, and increased cyclin D1 protein stability (60). I also observed that cyclin D1 protein was increased in SOX9-deficient heart valves, suggesting that SOX9 may play a role in the progression through S-phase via regulation of cyclin D1. Of note; Junb is occupied by SOX9 in its regulatory regions. JUNB is best known to inhibit cell growth by antagonizing c-JUN activity although this is highly context dependent (163). In this study, Fos and Junb transcripts were up-regulated in SOX9-deficient AVCs, suggesting that their uncontrolled and altered mRNA expression levels may in part contribute to organ hypoplasia seen in Sox9 mutants.  In addition to AP-1 factors, several other genes known to have roles in cell proliferation have SOX9 binding within their regulatory regions, including Cops5, Srpk2, Akt2, Eed, Hdac1, Hdac2, p53 (Trp53) and protein kinase A (Prkaca, PKA) in all three tissues examined. COPS5 associates with JUN proteins to increase binding specificity and can degrade the cell cycle inhibitor p27Kip1 (164). Loss of Cops5 in embryonic limb results in shortened limbs due to impaired chondrogenesis and Sox9 levels were decreased in mutant long bones (165) suggesting a potential feedback loop between SOX9 and COPS5. Additionally the kinase, SRPK2 can promote proliferation and cell cycle progression by enhancing Cyclin D1 levels (166). While AKT2, another kinase targeted by SOX9, regulates progression of cell cycle via phosphorylation of its targets. These phosphorylation targets of AKT2 include the cyclin-dependent kinase inhibitors. Additionally, AKT2 maintains protein stability of important cell cycle regulators such as c-MYC and D-type cyclins via Gsk3β during cell cycle (167). 131   Interestingly, several epigenetic regulators, EED, HDAC1 and HDAC2 were identified as commonly targeted by SOX9 and are associated with cell proliferation (168,169). Although HDACs are classically known for their roles in deaceylation and transcriptional repression, there is evidence that points to a more general role in proliferation. In both Hdac1/2 knockout and knockdown experiments, proliferation was decreased and inhibitors of HDACs exhibit similar anti-proliferative effects (reviewed in (169)). Altered proliferation levels associated with Hdac1/2 loss or inhibition may be attributable to the requirement of a fine balance between transcriptional activation and repression within cells. Eed is part of the Polycomb repressive complex (PRC) 2 complex that is involved in trimethylation of histone H3 lysine 27 (H3K27me3) leading to transcriptional repression. PRC2 members, EZH2 and EED, are controlled by E2F transcription factors and are required for cell proliferation (168). Together, E2F and the PRC2 members, EZH2 and EED, mediate the functions of the pRB-E2F growth control pathway. Of interest, inactivation of EZH2 in the developing heart causes major congenital heart defects such as defects of the ventricular septum, hypertrabeculation, and hypoplasia of the myocardium (170).  Although not identified as a common target in the HF-SCs, SOX9 occupied the regulatory regions associated with p53 (TRP53) and protein kinase A (Prkaca) in the AVC and limb, of which both of these factors have known roles in cell proliferation. p53 activates DNA repair and arrests proliferation by pausing cell cycle to fix DNA damage and if the damage is severe, p53 initiates cell death (171). PKA is induced by cyclic adenosine monophosphate (cAMP) and regulates cellular growth and proliferation through a variety of mechanisms (172). Of note, PKA phosphorylates SOX9 and increases its activity during chondrogenesis (173). The activities of 132  p53 and PKA in cell proliferation may be specific to mesenchyme as SOX9 binding sites are only found in AVC and limb.  A major caveat of my work demonstrating that SOX9 activates and promotes cell cycle as a common role of SOX9 across many cell types is that SOX9 has also previously been shown to be involved in suppression of proliferation and cell cycle in other cell types such as the intestinal epithelial cells of the crypt (174). This could be related to the fact that SOX9 may play an important role in the switch between proliferation and differentiation. In general cells must exit cell cycle and stop proliferating to differentiate, however there are always exceptions to the rule. In the intestinal epithelial stem/progenitor cells it has been shown that different levels of SOX9 promote (low level) or suppress proliferation (high levels) (174). This highlights the importance of understanding the level of SOX9 expression and in which context you are examining for the function of SOX9. Another situation where SOX9 has been shown opposing roles in proliferation is in cancers. For example, in some colorectal cancers high levels of SOX9 have been associated with more aggressive and proliferative subtypes while at the same time it has been shown to decrease the tumorigenicity in other colorectal cancer cells (64). This discrepancy for the role of SOX9 in colorectal cancers is likely due to the origins of the cancer initiating cells in these cancer subtypes and emphasizes that it is crucial to understand the context in which the functions of SOX9 are studied.  Taken together, our data suggests that SOX9 promotes proliferation across heart valve and limb mesenchyme cell types during development via binding to promoters and enhancers of critical regulators like AP1 proteins, kinases, and histone modifiers.   133  7.7 SOX9 is a master regulator of a core network of TFs in heart valve development By comparing transcriptome analysis in Sox9 mutant mice with genome wide SOX9 binding sites in the AVC, I have identified many potential SOX9 target genes that may be critical for valve formation, including ECM related genes and transcription factors. Interestingly, SOX9 targets that were down-regulated in the Sox9 cKO valves included numerous critical transcription factors known to be involved in heart development, such as Lef1, Pitx2 and Hand2. Loss of Lef1 (via TBX20 deletion), Pitx2 and Hand2 are all known to be associated with heart valve defects (175-177). The most highly down-regulated, SOX9-targetted TFs in Sox9 cKO valves were Twist1, Sox4, and Mecom/Evi1. Similar to SOX9, TWIST1, SOX4 and Mecom/EVI1 are highly expressed in the cardiac cushions during development and mutation of these factors leads to major valve abnormalities which are embryonic lethal (130,131,178).  SOX proteins are known to regulate other transcription factors that will function as their future co-factors (47). For instance, SOX9 regulates and cooperates with SOX5/6 to regulate target genes in the developing limb (48) and corroborating this, both SOX5/6 were targeted by SOX9 in the SOX9 limb ChIP-Seq library. Similarly, SOX9 may activate SOX4 in the heart to help co-regulate heart valve-specific genes. Motif analysis on SOX9 bound regions in the AVC revealed EVI1 as another potential co-factor for SOX9 and comparison of EVI1 bound regions in cancer cells (128) with SOX9 AVC bound regions identified hundreds of potentially overlapping target genes (Hoodless lab, unpublished). This supports a model in which EVI1 is a co-factor of SOX9 in the developing heart valves.  TWIST1 is well known to be associated with EMT in many different systems (reviewed in (179)). Here, I have shown that SOX9 can modulate the level of Twist1 expression in the developing valves via qRT-PCR. In the SOX9-deficient heart valves, Twist1 transcript levels 134  were reduced by approximately 3 fold in the valve mesenchyme by qRT-PCR. TWIST1 has been previously shown to have important roles in the developing heart valves. For example, TWIST1 can induce proliferation and migration of valve mesenchyme during early valve formation (180,181) and following EMT, TWIST1 plays a role in regulating differentiation of the AVC mesenchyme (182). Additionally, when TWIST1 persists at later stages of valve development, it leads to increased mesenchyme proliferation, increased TBX20 expression, and more primitive ECM composition (181) resembling a more early embryonic valve phenotype. Of interest, it has been shown that TWIST1 directly regulates Tbx20 (183). Our study found that TBX20 was a SOX9 target with down-regulated transcript levels in the Sox9 cKO valves. This suggests that TWIST1 and SOX9 may cooperate to regulate Tbx20 in developing heart valves.  Taken together, the correlation of transcriptome analysis and ChIP-Seq analysis reveals that SOX9 regulates a core regulatory network of transcription factors that are required at various steps of heart valve development. It should be noted that loss of SOX9 in the heart valves alters transcript expression levels modestly. Thus the role of SOX9 may be to modulate the level of gene regulation rather than to absolutely activate or repress gene expression. Nevertheless, the number of critical TFs that are known to be involved in heart valve development targeted by SOX9 and have reduced mRNA expression in the Sox9 cKO heart valves is remarkable and suggests that their combined loss of expression may explain the lethal phenotype observed in the Sox9 cKO embryos. Intriguingly, many of the TFs regulated by SOX9 also have been suggested to regulate each other’s expression such as TWIST1, which has been shown to regulate Tbx20 (183) and TBX20 in turn has been shown to regulate LEF1 (177). EVI1 can regulate SOX4 and alters its expression (131) and EVI1 and SOX4 can collaborate together 135  to regulate gene expression in myeloid leukemia (184). This illustrates how networks of transcription factors are working together to generate the heart valves during development.  Recently, SOX9 has been show to regulate chromatin dynamics near super enhancers in HF-SCs (63). Super enhancers are a cluster of enhancers that have been shown to be connected to genes that are associated with tissue identity. It is possible that binding of SOX9 to locations near many essential heart transcription factors to allows for activation of these regions and promote further valve cell differentiation. This illustrates the complex interactions that occur at multiple levels in heart valve development and suggests that these essential TFs are regulated in numerous ways to ensure proper valve formation. This work establishes that SOX9 and its transcriptional target TFs form a critical gene regulatory network to drive valve morphogenesis. In humans, aberrant expression of SOX9 and its transcriptional targets have been associated with adult heart valve disease (31). Thus, understanding the SOX9 initiated transcriptional networks in heart valve development may provide additional insights into adult heart valve disease.  7.8 Examination of other defined roles of SOX9 in EMT and ECM generation Although SOX9 has been implicated in the process of EMT in neural crest cells, this study did not find any overt differences in EMT markers at the transcript level or at the protein level in the E12.5 Sox9 cKO heart valves. This suggests that perhaps in the developing heart valves that SOX9 is not essential for EMT but may necessary for the maintenance and survival of the AVC mesenchyme cells. In contrast, the phenotype of the Sox9fl/fl;VE-Cre mice indicates there may be an early role for SOX9 in EMT and that Sox9-deficient endocardial cells may never undergo EMT. Another study where Sox9 null embryos were generated by crossing mice where they conditionally deleted one allele of Sox9 in oocytes and one Sox9 allele in spermatids lead to 136  embryonic death at E11.5-12 due to congestive heart failure and embryos had severely hypoplastic cardiac cushions (69). Further analysis of the Sox9-deficient heart valves revealed that SOX9 inhibits EMT following the delamination and preliminary migration of the cells away from the endocardium before they become definitive mesenchyme (69). Of note, I did not detect any SOX9 negative mesenchyme even at the earliest stages of EMT starting at E9.5. Since there is conflicting evidence on the role of SOX9 in EMT, more in depth studies are required to discover the exact role of SOX9 in the heart valves.  SOX9 is well known for its role in regulating ECM production in the successive steps of cartilage formation in the developing limb (reviewed in (41)). Since the limb and heart valves share many similarities in composition and structure, expression of ECM components were analyzed in Sox9 cKO heart valves. In the absence of Sox9, Periostin, Elastin, and Matrix Gla protein transcripts were reduced as identified by qRT-PCR and RNA-Seq analysis. Additional ECM components such as Col9a1, Col9a3, Prelp, Col6a6, Mfap4, Matn4, etc were also reduced in the Sox9 cKO AVC RNA-Seq library. All of the aforementioned ECM components are all targets of SOX9 in the AVC ChIP-Seq library. PERIOSTIN has been shown to have a critical role in AV valve development by binding integrins and linking intracellular kinase dependent signalling (185) and is required for AVC mesenchyme maturation and ECM compaction (186). However, analysis of PERIOSTIN protein levels in WT and Sox9 cKO valves demonstrated that PERIOSTIN was not reduced in the Sox9 cKO AVC and almost seemed up-regulated compared to overall valve area. However, a small number of SOX9 positive mesenchyme remained in the Sox9 cKO valves that could be secreting higher levels of PERIOSTIN to compensate for decreased transcript levels in the SOX9 negative mesenchyme. Elastin is another interesting molecule as it is required to maintain the elasticity of the adult heart valves and would be 137  appealing to investigate further. Future analyses of the relationship between SOX9 and ECM would be focused on developing a better understanding of heart valve ECM composition during development and how it is altered upon loss of Sox9 in the heart valves.  7.9 Concluding remarks In this work, I have analyzed the global in vivo transcriptional targets of SOX9 in the embryonic heart valves and limb generated from ChIP-Seq libraries previously made in our lab (R. Cullum for ChIP and GSC for sequencing). For the first time, thousands of SOX9 binding sites have been revealed in the heart valves and limb and will provide an excellent resource for future research by our lab and others upon publication (Garside et al., accepted). Understanding of the downstream transcriptional targets of SOX9 in embryonic tissues is extremely limited and to my current knowledge the only embryonic SOX9 ChIP-Seq library was performed on the vertebral column (125) as was discussed earlier in this study. Surprisingly, the minority of studies are composed of ChIP-Seq from embryonic tissues while the majority of ChIP-Seq is performed on cell lines or embryonic stem cells and thus making this data even more valuable to the field. Future analyses should be focused on further understanding the transcriptional targets of SOX9 in a given tissue and how SOX9 transcriptional targets change throughout the development of a tissue. In this study, I have shown that SOX9 is expressed throughout heart valve development. It has been proposed by us and others that SOX9 likely changes its roles during tissue formation and this change in roles is highly dependent on its co-factors. Therefore, SOX9 binding should be analyzed dynamically over time within an embryonic tissue with a specific focus on identifying its binding partners as they are likely guiding the change in the role of SOX9 over time. 138   To gain a better understanding of the function of SOX9 it is extremely valuable to generate a Sox9-deficient mouse model to analyze the changes that occur upon the loss of Sox9 in the heart valves. Therefore, in this study I had generated two Sox9 mutant models using the Cre/Lox system: Sox9fl/fl;VE-Cre and Sox9fl/f;Tie2-Cre mutant mice which deletes Sox9 specifically in endothelial cells and subsequently the heart valve mesenchyme. Since the goal of this investigation was to analyze the heart valves in absence of Sox9, the Sox9fl/fl;VE-Cre mice were not suitable for this purpose as SOX9 negative mesenchyme could not be detected in the embryonic heart valves. Examination of the Sox9fl/fl;VE-Cre mice revealed an unexpected phenotype and future analysis should be focused on teasing out the role of SOX9 in the endothelium itself or within the early EMT stages before becoming fully mesenchymal. Fortunately, SOX9 negative mesenchyme could be detected in the Sox9fl/fl;Tie2-Cre (Sox9 cKO) mutant heart valves and could be used for our downstream analyses. Furthermore, proliferation defects were confirmed in the Sox9 cKO heart valves, which is a common feature of many organs in which Sox9 is deleted. This suggests that a common role of SOX9 may be involved in proliferation. To examine this further, I compared SOX9 binding sites from other tissues with our data and identified a subset of commonly targeted genes that were involved in proliferation. Taken together this data suggests that SOX9 has both context independent and context dependent roles. Future work should focus on how exactly SOX9 regulates each of the different processes that occur during proliferation and cell cycle and focus in on critical regulators like the AP1 proteins like Junb and Fos and as well as critical factors like p53. Additionally, although SOX9 has been generally shown to be an activator of transcription, my work has also shown that SOX9 can act as repressor for factors such as Junb. Understanding how SOX9 represses its targets could also be an interesting avenue to follow up in the future. 139   To uncover the context dependent and critical targets of SOX9 in the heart valves, I generated transcriptome profiles from WT and Sox9fl/fl;Tie2-Cre mutant heart valves using RNA-Seq and compared them with the SOX9 binding sites in the heart valves. This analysis generated a short list of 139 critical genes were both SOX9 targets in the heart valves and have altered expression in the Sox9 cKO heart valves. Strikingly, a high number of these were transcription factors that had known roles in the developing heart and suggest that SOX9 regulates a heart specific transcriptional network to guide heart valve formation. Further studies should focus on how all of these critical transcription factors work together to form the heart valves. Taken together, there are many avenues to be followed up to get a better handle on the function of SOX9 and this is only just touching on the vast interactions and processes that SOX9 is involved in during development. This data has identified a number of key factors involved in early valve formation that are regulated by SOX9 and hopefully in turn, these could be used as predictive factors of heart disease, or as targets for new therapeutic strategies for heart valve disease and congenital heart valve defects. As SOX9 is involved in development of many different organ systems as well as implicated in a variety of diseases, this data may provide new insights on the role of SOX9 and its transcriptional network within a diverse group of organs and diseases.  140  REFERENCES  1. Miquerol, L. and Kelly, R.G. (2013) Organogenesis of the vertebrate heart. Wiley Interdiscip Rev Dev Biol, 2, 17-29. 2. Abu-Issa, R. and Kirby, M.L. (2007) Heart field: from mesoderm to heart tube. Annu Rev Cell Dev Biol, 23, 45-68. 3. Hinton, R.B. and Yutzey, K.E. (2011) Heart valve structure and function in development and disease. Annu Rev Physiol, 73, 29-46. 4. Combs, M.D. and Yutzey, K.E. (2009) Heart valve development: regulatory networks in development and disease. Circ Res, 105, 408-421. 5. Snarr, B.S., Kern, C.B. and Wessels, A. (2008) Origin and fate of cardiac mesenchyme. Dev. Dyn., 237, 2804-2819. 6. Person, A.D., Klewer, S.E. and Runyan, R.B. (2005) Cell biology of cardiac cushion development. Int Rev Cytol, 243, 287-335. 7. Camenisch, T.D., Molin, D.G., Person, A., Runyan, R.B., Gittenberger-de Groot, A.C., McDonald, J.A. and Klewer, S.E. (2002) Temporal and distinct TGFbeta ligand requirements during mouse and avian endocardial cushion morphogenesis. Dev Biol, 248, 170-181. 8. Schroeder, J.A., Jackson, L.F., Lee, D.C. and Camenisch, T.D. (2003) Form and function of developing heart valves: coordination by extracellular matrix and growth factor signaling. J Mol Med (Berl), 81, 392-403. 9. Markwald, R.R., Fitzharris, T.P. and Manasek, F.J. (1977) Structural development of endocardial cushions. Am J Anat, 148, 85-119. 10. Restivo, A., Piacentini, G., Placidi, S., Saffirio, C. and Marino, B. (2006) Cardiac outflow tract: a review of some embryogenetic aspects of the conotruncal region of the heart. Anat Rec A Discov Mol Cell Evol Biol, 288, 936-943. 11. Hinton, R.B., Jr., Lincoln, J., Deutsch, G.H., Osinska, H., Manning, P.B., Benson, D.W. and Yutzey, K.E. (2006) Extracellular matrix remodeling and organization in developing and diseased aortic valves. Circ Res, 98, 1431-1438. 12. Hinton, R.B., Jr., Alfieri, C.M., Witt, S.A., Glascock, B.J., Khoury, P.R., Benson, D.W. and Yutzey, K.E. (2008) Mouse heart valve structure and function: echocardiographic and morphometric analyses from the fetus through the aged adult. Am J Physiol Heart Circ Physiol, 294, H2480-2488. 13. Aikawa, E., Whittaker, P., Farber, M., Mendelson, K., Padera, R.F., Aikawa, M. and Schoen, F.J. (2006) Human semilunar cardiac valve remodeling by activated cells from fetus to adult: implications for postnatal adaptation, pathology, and tissue engineering. Circulation, 113, 1344-1352. 14. Butcher, J.T. and Markwald, R.R. (2007) Valvulogenesis: the moving target. Philos Trans R Soc Lond B Biol Sci, 362, 1489-1503. 15. Kruithof, B.P., Krawitz, S.A. and Gaussin, V. (2007) Atrioventricular valve development during late embryonic and postnatal stages involves condensation and extracellular matrix remodeling. Dev Biol, 302, 208-217. 16. Sacks, M.S. and Yoganathan, A.P. (2007) Heart valve function: a biomechanical perspective. Philos Trans R Soc Lond B Biol Sci, 362, 1369-1391. 17. Vesely, I. (1998) The role of elastin in aortic valve mechanics. J Biomech, 31, 115-123. 141  18. de Lange, F.J., Moorman, A.F., Anderson, R.H., Manner, J., Soufan, A.T., de Gier-de Vries, C., Schneider, M.D., Webb, S., van den Hoff, M.J. and Christoffels, V.M. (2004) Lineage and morphogenetic analysis of the cardiac valves. Circ Res, 95, 645-654. 19. Lincoln, J., Alfieri, C.M. and Yutzey, K.E. (2004) Development of heart valve leaflets and supporting apparatus in chicken and mouse embryos. Dev Dyn, 230, 239-250. 20. Wessels, A., van den Hoff, M.J., Adamo, R.F., Phelps, A.L., Lockhart, M.M., Sauls, K., Briggs, L.E., Norris, R.A., van Wijk, B., Perez-Pomares, J.M. et al. (2012) Epicardially derived fibroblasts preferentially contribute to the parietal leaflets of the atrioventricular valves in the murine heart. Developmental biology, 366, 111-124. 21. Liu, A.C., Joag, V.R. and Gotlieb, A.I. (2007) The emerging role of valve interstitial cell phenotypes in regulating heart valve pathobiology. Am J Pathol, 171, 1407-1418. 22. Garside, V.C., Chang, A.C., Karsan, A. and Hoodless, P.A. (2013) Co-ordinating Notch, BMP, and TGF-beta signaling during heart valve development. Cellular and molecular life sciences : CMLS, 70, 2899-2917. 23. Xiao, H. and Zhang, Y.Y. (2008) Understanding the role of transforming growth factor-beta signalling in the heart: overview of studies using genetic mouse models. Clin Exp Pharmacol Physiol, 35, 335-341. 24. Conway, S.J., Doetschman, T. and Azhar, M. (2011) The inter-relationship of periostin, TGF beta, and BMP in heart valve development and valvular heart diseases. ScientificWorldJournal, 11, 1509-1524. 25. Mohler, E.R., 3rd, Gannon, F., Reynolds, C., Zimmerman, R., Keane, M.G. and Kaplan, F.S. (2001) Bone formation and inflammation in cardiac valves. Circulation, 103, 1522-1528. 26. Kaden, J.J., Bickelhaupt, S., Grobholz, R., Vahl, C.F., Hagl, S., Brueckmann, M., Haase, K.K., Dempfle, C.E. and Borggrefe, M. (2004) Expression of bone sialoprotein and bone morphogenetic protein-2 in calcific aortic stenosis. J Heart Valve Dis, 13, 560-566. 27. Ankeny, R.F., Thourani, V.H., Weiss, D., Vega, J.D., Taylor, W.R., Nerem, R.M. and Jo, H. (2011) Preferential activation of SMAD1/5/8 on the fibrosa endothelium in calcified human aortic valves--association with low BMP antagonists and SMAD6. PLoS One, 6, e20969. 28. Rosenkranz, S. (2004) TGF-beta1 and angiotensin networking in cardiac remodeling. Cardiovasc Res, 63, 423-432. 29. Walker, G.A., Masters, K.S., Shah, D.N., Anseth, K.S. and Leinwand, L.A. (2004) Valvular myofibroblast activation by transforming growth factor-beta: implications for pathological extracellular matrix remodeling in heart valve disease. Circ Res, 95, 253-260. 30. Liu, A.C. and Gotlieb, A.I. (2008) Transforming growth factor-beta regulates in vitro heart valve repair by activated valve interstitial cells. Am J Pathol, 173, 1275-1285. 31. Hulin, A., Deroanne, C.F., Lambert, C.A., Dumont, B., Castronovo, V., Defraigne, J.O., Nusgens, B.V., Radermecker, M.A. and Colige, A.C. (2012) Metallothionein-dependent up-regulation of TGF-beta2 participates in the remodelling of the myxomatous mitral valve. Cardiovasc Res, 93, 480-489. 32. Girdauskas, E., Schulz, S., Borger, M.A., Mierzwa, M. and Kuntze, T. (2011) Transforming growth factor-beta receptor type II mutation in a patient with bicuspid aortic valve disease and intraoperative aortic dissection. Ann Thorac Surg, 91, e70-71. 142  33. Attaran, S., Sherwood, R., Dastidar, M.G. and El-Gamel, A. (2010) Identification of low circulatory transforming growth factor beta-1 in patients with degenerative heart valve disease. Interact Cardiovasc Thorac Surg, 11, 791-793. 34. Luna-Zurita, L., Prados, B., Grego-Bessa, J., Luxan, G., del Monte, G., Benguria, A., Adams, R.H., Perez-Pomares, J.M. and de la Pompa, J.L. (2010) Integration of a Notch-dependent mesenchymal gene program and Bmp2-driven cell invasiveness regulates murine cardiac valve formation. J Clin Invest, 120, 3493-3507. 35. Timmerman, L.A., Grego-Bessa, J., Raya, A., Bertran, E., Perez-Pomares, J.M., Diez, J., Aranda, S., Palomo, S., McCormick, F., Izpisua-Belmonte, J.C. et al. (2004) Notch promotes epithelial-mesenchymal transition during cardiac development and oncogenic transformation. Genes Dev, 18, 99-115. 36. Acharya, A., Hans, C.P., Koenig, S.N., Nichols, H.A., Galindo, C.L., Garner, H.R., Merrill, W.H., Hinton, R.B. and Garg, V. (2011) Inhibitory role of Notch1 in calcific aortic valve disease. PLoS One, 6, e27743. 37. Lincoln, J., Lange, A.W. and Yutzey, K.E. (2006) Hearts and bones: shared regulatory mechanisms in heart valve, cartilage, tendon, and bone development. Dev Biol, 294, 292-302. 38. Chakraborty, S., Cheek, J., Sakthivel, B., Aronow, B.J. and Yutzey, K.E. (2008) Shared gene expression profiles in developing heart valves and osteoblast progenitor cells. Physiol Genomics, 35, 75-85. 39. Long, F. and Ornitz, D.M. (2013) Development of the endochondral skeleton. Cold Spring Harb Perspect Biol, 5, a008334. 40. Mackie, E.J., Tatarczuch, L. and Mirams, M. (2011) The skeleton: a multi-functional complex organ. The growth plate chondrocyte and endochondral ossification. J Endocrinol, 211, 109-121. 41. DeLise, A.M., Fischer, L. and Tuan, R.S. (2000) Cellular interactions and signaling in cartilage development. Osteoarthritis Cartilage, 8, 309-334. 42. Zuscik, M.J., Hilton, M.J., Zhang, X., Chen, D. and O'Keefe, R.J. (2008) Regulation of chondrogenesis and chondrocyte differentiation by stress. J Clin Invest, 118, 429-438. 43. Chakraborty, S., Combs, M.D. and Yutzey, K.E. (2010) Transcriptional regulation of heart valve progenitor cells. Pediatr Cardiol, 31, 414-421. 44. Lange, A.W. and Yutzey, K.E. (2006) NFATc1 expression in the developing heart valves is responsive to the RANKL pathway and is required for endocardial expression of cathepsin K. Dev Biol, 292, 407-417. 45. Akiyama, H., Chaboissier, M.C., Martin, J.F., Schedl, A. and de Crombrugghe, B. (2002) The transcription factor Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required for expression of Sox5 and Sox6. Genes Dev, 16, 2813-2828. 46. Lincoln, J., Kist, R., Scherer, G. and Yutzey, K.E. (2007) Sox9 is required for precursor cell expansion and extracellular matrix organization during mouse heart valve development. Dev Biol, 305, 120-132. 47. Kamachi, Y. and Kondoh, H. (2013) Sox proteins: regulators of cell fate specification and differentiation. Development, 140, 4129-4144. 143  48. Han, Y. and Lefebvre, V. (2008) L-Sox5 and Sox6 drive expression of the aggrecan gene in cartilage by securing binding of Sox9 to a far-upstream enhancer. Mol Cell Biol, 28, 4999-5013. 49. Bondurand, N., Pingault, V., Goerich, D.E., Lemort, N., Sock, E., Le Caignec, C., Wegner, M. and Goossens, M. (2000) Interaction among SOX10, PAX3 and MITF, three genes altered in Waardenburg syndrome. Hum Mol Genet, 9, 1907-1917. 50. Ludwig, A., Rehberg, S. and Wegner, M. (2004) Melanocyte-specific expression of dopachrome tautomerase is dependent on synergistic gene activation by the Sox10 and Mitf transcription factors. FEBS Lett, 556, 236-244. 51. Barrionuevo, F. and Scherer, G. (2010) SOX E genes: SOX9 and SOX8 in mammalian testis development. Int J Biochem Cell Biol, 42, 433-436. 52. Haldin, C.E. and LaBonne, C. (2010) SoxE factors as multifunctional neural crest regulatory factors. Int J Biochem Cell Biol, 42, 441-444. 53. Pritchett, J., Athwal, V., Roberts, N., Hanley, N.A. and Hanley, K.P. (2011) Understanding the role of SOX9 in acquired diseases: lessons from development. Trends Mol Med, 17, 166-174. 54. Wagner, T., Wirth, J., Meyer, J., Zabel, B., Held, M., Zimmer, J., Pasantes, J., Bricarelli, F.D., Keutel, J., Hustert, E. et al. (1994) Autosomal sex reversal and campomelic dysplasia are caused by mutations in and around the SRY-related gene SOX9. Cell, 79, 1111-1120. 55. Piper, K., Ball, S.G., Keeling, J.W., Mansoor, S., Wilson, D.I. and Hanley, N.A. (2002) Novel SOX9 expression during human pancreas development correlates to abnormalities in Campomelic dysplasia. Mech Dev, 116, 223-226. 56. Seymour, P.A., Freude, K.K., Tran, M.N., Mayes, E.E., Jensen, J., Kist, R., Scherer, G. and Sander, M. (2007) SOX9 is required for maintenance of the pancreatic progenitor cell pool. Proc Natl Acad Sci U S A, 104, 1865-1870. 57. Rockich, B.E., Hrycaj, S.M., Shih, H.P., Nagy, M.S., Ferguson, M.A., Kopp, J.L., Sander, M., Wellik, D.M. and Spence, J.R. (2013) Sox9 plays multiple roles in the lung epithelium during branching morphogenesis. Proc Natl Acad Sci U S A, 110, E4456-4464. 58. Thomsen, M.K., Butler, C.M., Shen, M.M. and Swain, A. (2008) Sox9 is required for prostate development. Dev Biol, 316, 302-311. 59. Trowe, M.O., Shah, S., Petry, M., Airik, R., Schuster-Gossler, K., Kist, R. and Kispert, A. (2010) Loss of Sox9 in the periotic mesenchyme affects mesenchymal expansion and differentiation, and epithelial morphogenesis during cochlea development in the mouse. Dev Biol, 342, 51-62. 60. Stockl, S., Bauer, R.J., Bosserhoff, A.K., Gottl, C., Grifka, J. and Grassel, S. (2013) Sox9 modulates cell survival and adipogenic differentiation of multipotent adult rat mesenchymal stem cells. J Cell Sci, 126, 2890-2902. 61. Guo, W., Keckesova, Z., Donaher, J.L., Shibue, T., Tischler, V., Reinhardt, F., Itzkovitz, S., Noske, A., Zurrer-Hardi, U., Bell, G. et al. (2012) Slug and Sox9 cooperatively determine the mammary stem cell state. Cell, 148, 1015-1028. 62. Kadaja, M., Keyes, B.E., Lin, M., Pasolli, H.A., Genander, M., Polak, L., Stokes, N., Zheng, D. and Fuchs, E. (2014) SOX9: a stem cell transcriptional regulator of secreted niche signaling factors. Genes & development, 28, 328-341. 144  63. Adam, R.C., Yang, H., Rockowitz, S., Larsen, S.B., Nikolova, M., Oristian, D.S., Polak, L., Kadaja, M., Asare, A., Zheng, D. et al. (2015) Pioneer factors govern super-enhancer dynamics in stem cell plasticity and lineage choice. Nature. 64. Matheu, A., Collado, M., Wise, C., Manterola, L., Cekaite, L., Tye, A.J., Canamero, M., Bujanda, L., Schedl, A., Cheah, K.S. et al. (2012) Oncogenicity of the developmental transcription factor Sox9. Cancer Res, 72, 1301-1315. 65. Wang, H., He, L., Ma, F., Regan, M.M., Balk, S.P., Richardson, A.L. and Yuan, X. (2013) SOX9 regulates low density lipoprotein receptor-related protein 6 (LRP6) and T-cell factor 4 (TCF4) expression and Wnt/beta-catenin activation in breast cancer. J Biol Chem, 288, 6478-6487. 66. Shi, Z., Chiang, C.I., Mistretta, T.A., Major, A. and Mori-Akiyama, Y. (2013) SOX9 directly regulates IGFBP-4 in the intestinal epithelium. Am J Physiol Gastrointest Liver Physiol, 305, G74-83. 67. Sakai, D., Suzuki, T., Osumi, N. and Wakamatsu, Y. (2006) Cooperative action of Sox9, Snail2 and PKA signaling in early neural crest development. Development, 133, 1323-1333. 68. Akiyama, H. (2008) Control of chondrogenesis by the transcription factor Sox9. Mod Rheumatol, 18, 213-219. 69. Akiyama, H., Chaboissier, M.C., Behringer, R.R., Rowitch, D.H., Schedl, A., Epstein, J.A. and de Crombrugghe, B. (2004) Essential role of Sox9 in the pathway that controls formation of cardiac valves and septa. Proc Natl Acad Sci U S A, 101, 6502-6507. 70. Ng, L.J., Wheatley, S., Muscat, G.E., Conway-Campbell, J., Bowles, J., Wright, E., Bell, D.M., Tam, P.P., Cheah, K.S. and Koopman, P. (1997) SOX9 binds DNA, activates transcription, and coexpresses with type II collagen during chondrogenesis in the mouse. Dev Biol, 183, 108-121. 71. Rahkonen, O., Savontaus, M., Abdelwahid, E., Vuorio, E. and Jokinen, E. (2003) Expression patterns of cartilage collagens and Sox9 during mouse heart development. Histochem Cell Biol, 120, 103-110. 72. Peacock, J.D., Levay, A.K., Gillaspie, D.B., Tao, G. and Lincoln, J. (2010) Reduced sox9 function promotes heart valve calcification phenotypes in vivo. Circ Res, 106, 712-719. 73. Chang, A.C., Garside, V.C., Fournier, M., Smrz, J., Vrljicak, P., Umlandt, P., Fuller, M., Robertson, G., Zhao, Y., Tam, A. et al. (2014) A Notch-dependent transcriptional hierarchy promotes mesenchymal transdifferentiation in the cardiac cushion. Dev Dyn, 243, 894-905. 74. Wirrig, E.E., Hinton, R.B. and Yutzey, K.E. (2011) Differential expression of cartilage and bone-related proteins in pediatric and adult diseased aortic valves. J Mol Cell Cardiol, 50, 561-569. 75. Alexopoulos, A., Bravou, V., Peroukides, S., Kaklamanis, L., Varakis, J., Alexopoulos, D. and Papadaki, H. (2010) Bone regulatory factors NFATc1 and Osterix in human calcific aortic valves. Int J Cardiol, 139, 142-149. 76. Caira, F.C., Stock, S.R., Gleason, T.G., McGee, E.C., Huang, J., Bonow, R.O., Spelsberg, T.C., McCarthy, P.M., Rahimtoola, S.H. and Rajamannan, N.M. (2006) Human degenerative valve disease is associated with up-regulation of low-density 145  lipoprotein receptor-related protein 5 receptor-mediated bone formation. J Am Coll Cardiol, 47, 1707-1712. 77. Cheek, J.D., Wirrig, E.E., Alfieri, C.M., James, J.F. and Yutzey, K.E. (2012) Differential activation of valvulogenic, chondrogenic, and osteogenic pathways in mouse models of myxomatous and calcific aortic valve disease. J Mol Cell Cardiol, 52, 689-700. 78. Bi, W., Deng, J.M., Zhang, Z., Behringer, R.R. and de Crombrugghe, B. (1999) Sox9 is required for cartilage formation. Nat Genet, 22, 85-89. 79. Bell, D.M., Leung, K.K., Wheatley, S.C., Ng, L.J., Zhou, S., Ling, K.W., Sham, M.H., Koopman, P., Tam, P.P. and Cheah, K.S. (1997) SOX9 directly regulates the type-II collagen gene. Nat Genet, 16, 174-178. 80. Bridgewater, L.C., Lefebvre, V. and de Crombrugghe, B. (1998) Chondrocyte-specific enhancer elements in the Col11a2 gene resemble the Col2a1 tissue-specific enhancer. J Biol Chem, 273, 14998-15006. 81. Xie, W.F., Zhang, X., Sakano, S., Lefebvre, V. and Sandell, L.J. (1999) Trans-activation of the mouse cartilage-derived retinoic acid-sensitive protein gene by Sox9. J Bone Miner Res, 14, 757-763. 82. Sekiya, I., Tsuji, K., Koopman, P., Watanabe, H., Yamada, Y., Shinomiya, K., Nifuji, A. and Noda, M. (2000) SOX9 enhances aggrecan gene promoter/enhancer activity and is up-regulated by retinoic acid in a cartilage-derived cell line, TC6. J Biol Chem, 275, 10738-10744. 83. Kou, I. and Ikegawa, S. (2004) SOX9-dependent and -independent transcriptional regulation of human cartilage link protein. J Biol Chem, 279, 50942-50948. 84. Liu, C.J., Zhang, Y., Xu, K., Parsons, D., Alfonso, D. and Di Cesare, P.E. (2007) Transcriptional activation of cartilage oligomeric matrix protein by Sox9, Sox5, and Sox6 transcription factors and CBP/p300 coactivators. Front Biosci, 12, 3899-3910. 85. Bernard, P., Tang, P., Liu, S., Dewing, P., Harley, V.R. and Vilain, E. (2003) Dimerization of SOX9 is required for chondrogenesis, but not for sex determination. Hum Mol Genet, 12, 1755-1765. 86. Coustry, F., Oh, C.D., Hattori, T., Maity, S.N., de Crombrugghe, B. and Yasuda, H. (2010) The dimerization domain of SOX9 is required for transcription activation of a chondrocyte-specific chromatin DNA template. Nucleic Acids Res, 38, 6018-6028. 87. Genzer, M.A. and Bridgewater, L.C. (2007) A Col9a1 enhancer element activated by two interdependent SOX9 dimers. Nucleic Acids Res, 35, 1178-1186. 88. Zhang, P., Jimenez, S.A. and Stokes, D.G. (2003) Regulation of human COL9A1 gene expression. Activation of the proximal promoter region by SOX9. J Biol Chem, 278, 117-123. 89. Peacock, J.D., Huk, D.J., Ediriweera, H.N. and Lincoln, J. (2011) Sox9 transcriptionally represses Spp1 to prevent matrix mineralization in maturing heart valves and chondrocytes. PLoS One, 6, e26769. 90. Bi, W., Huang, W., Whitworth, D.J., Deng, J.M., Zhang, Z., Behringer, R.R. and de Crombrugghe, B. (2001) Haploinsufficiency of Sox9 results in defective cartilage primordia and premature skeletal mineralization. Proc Natl Acad Sci U S A, 98, 6698-6703. 91. Hoffman, J.I. and Kaplan, S. (2002) The incidence of congenital heart disease. J Am Coll Cardiol, 39, 1890-1900. 146  92. Pierpont, M.E., Basson, C.T., Benson, D.W., Jr., Gelb, B.D., Giglia, T.M., Goldmuntz, E., McGee, G., Sable, C.A., Srivastava, D. and Webb, C.L. (2007) Genetic basis for congenital heart defects: current knowledge: a scientific statement from the American Heart Association Congenital Cardiac Defects Committee, Council on Cardiovascular Disease in the Young: endorsed by the American Academy of Pediatrics. Circulation, 115, 3015-3038. 93. Roger, V.L., Go, A.S., Lloyd-Jones, D.M., Adams, R.J., Berry, J.D., Brown, T.M., Carnethon, M.R., Dai, S., de Simone, G., Ford, E.S. et al. (2011) Heart disease and stroke statistics--2011 update: a report from the American Heart Association. Circulation, 123, e18-e209. 94. Rabkin, E., Aikawa, M., Stone, J.R., Fukumoto, Y., Libby, P. and Schoen, F.J. (2001) Activated interstitial myofibroblasts express catabolic enzymes and mediate matrix remodeling in myxomatous heart valves. Circulation, 104, 2525-2532. 95. Rabkin-Aikawa, E., Farber, M., Aikawa, M. and Schoen, F.J. (2004) Dynamic and reversible changes of interstitial cell phenotype during remodeling of cardiac valves. J Heart Valve Dis, 13, 841-847. 96. Pomerance, A. (1972) Pathogenesis of aortic stenosis and its relation to age. Br Heart J, 34, 569-574. 97. Passik, C.S., Ackermann, D.M., Pluth, J.R. and Edwards, W.D. (1987) Temporal changes in the causes of aortic stenosis: a surgical pathologic study of 646 cases. Mayo Clin Proc, 62, 119-123. 98. Peterson, M.D., Roach, R.M. and Edwards, J.E. (1985) Types of aortic stenosis in surgically removed valves. Arch Pathol Lab Med, 109, 829-832. 99. Stephan, P.J., Henry, A.C., 3rd, Hebeler, R.F., Jr., Whiddon, L. and Roberts, W.C. (1997) Comparison of age, gender, number of aortic valve cusps, concomitant coronary artery bypass grafting, and magnitude of left ventricular-systemic arterial peak systolic gradient in adults having aortic valve replacement for isolated aortic valve stenosis. Am J Cardiol, 79, 166-172. 100. Sheikh, A.M. and Livesey, S.A. (2010) Surgical management of valve disease in the early 21st century. Clin Med, 10, 177-181. 101. Levay, A.K., Peacock, J.D., Lu, Y., Koch, M., Hinton, R.B., Jr., Kadler, K.E. and Lincoln, J. (2008) Scleraxis is required for cell lineage differentiation and extracellular matrix remodeling during murine heart valve formation in vivo. Circ Res, 103, 948-956. 102. Garg, V., Muth, A.N., Ransom, J.F., Schluterman, M.K., Barnes, R., King, I.N., Grossfeld, P.D. and Srivastava, D. (2005) Mutations in NOTCH1 cause aortic valve disease. Nature, 437, 270-274. 103. Nigam, V. and Srivastava, D. (2009) Notch1 represses osteogenic pathways in aortic valve cells. J Mol Cell Cardiol, 47, 828-834. 104. Nus, M., MacGrogan, D., Martinez-Poveda, B., Benito, Y., Casanova, J.C., Fernandez-Aviles, F., Bermejo, J. and de la Pompa, J.L. (2011) Diet-induced aortic valve disease in mice haploinsufficient for the Notch pathway effector RBPJK/CSL. Arterioscler Thromb Vasc Biol, 31, 1580-1588. 105. Impey, S., McCorkle, S.R., Cha-Molstad, H., Dwyer, J.M., Yochum, G.S., Boss, J.M., McWeeney, S., Dunn, J.J., Mandel, G. and Goodman, R.H. (2004) Defining the CREB 147  regulon: a genome-wide analysis of transcription factor regulatory regions. Cell, 119, 1041-1054. 106. Kim, J., Bhinge, A.A., Morgan, X.C. and Iyer, V.R. (2005) Mapping DNA-protein interactions in large genomes by sequence tag analysis of genomic enrichment. Nat Methods, 2, 47-53. 107. Hou, J., Charters, A.M., Lee, S.C., Zhao, Y., Wu, M.K., Jones, S.J., Marra, M.A. and Hoodless, P.A. (2007) A systematic screen for genes expressed in definitive endoderm by Serial Analysis of Gene Expression (SAGE). BMC Dev Biol, 7, 92. 108. McKnight, K.D., Hou, J. and Hoodless, P.A. (2007) Dynamic expression of thyrotropin-releasing hormone in the mouse definitive endoderm. Dev Dyn, 236, 2909-2917. 109. Hassan, A.S., Hou, J., Wei, W. and Hoodless, P.A. (2010) Expression of two novel transcripts in the mouse definitive endoderm. Gene Expr Patterns, 10, 127-134. 110. Robertson, G., Hirst, M., Bainbridge, M., Bilenky, M., Zhao, Y., Zeng, T., Euskirchen, G., Bernier, B., Varhol, R., Delaney, A. et al. (2007) Genome-wide profiles of STAT1 DNA association using chromatin immunoprecipitation and massively parallel sequencing. Nat Methods, 4, 651-657. 111. Li, H. and Durbin, R. (2009) Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics (Oxford, England), 25, 1754-1760. 112. Fejes, A.P., Robertson, G., Bilenky, M., Varhol, R., Bainbridge, M. and Jones, S.J. (2008) FindPeaks 3.1: a tool for identifying areas of enrichment from massively parallel short-read sequencing technology. Bioinformatics, 24, 1729-1730. 113. Robertson, A.G., Bilenky, M., Tam, A., Zhao, Y., Zeng, T., Thiessen, N., Cezard, T., Fejes, A.P., Wederell, E.D., Cullum, R. et al. (2008) Genome-wide relationship between histone H3 lysine 4 mono- and tri-methylation and transcription factor binding. Genome Res, 18, 1906-1917. 114. Kent, W.J., Sugnet, C.W., Furey, T.S., Roskin, K.M., Pringle, T.H., Zahler, A.M. and Haussler, D. (2002) The human genome browser at UCSC. Genome Res, 12, 996-1006. 115. Blankenberg, D., Von Kuster, G., Coraor, N., Ananda, G., Lazarus, R., Mangan, M., Nekrutenko, A. and Taylor, J. (2010) Galaxy: a web-based genome analysis tool for experimentalists. Curr Protoc Mol Biol, Chapter 19, Unit 19 10 11-21. 116. Liu, T., Ortiz, J.A., Taing, L., Meyer, C.A., Lee, B., Zhang, Y., Shin, H., Wong, S.S., Ma, J., Lei, Y. et al. (2011) Cistrome: an integrative platform for transcriptional regulation studies. Genome Biol, 12, R83. 117. Eden, E., Navon, R., Steinfeld, I., Lipson, D. and Yakhini, Z. (2009) GOrilla: a tool for discovery and visualization of enriched GO terms in ranked gene lists. BMC Bioinformatics, 10, 48. 118. Kwon, A.T., Arenillas, D.J., Worsley Hunt, R. and Wasserman, W.W. (2012) oPOSSUM-3: advanced analysis of regulatory motif over-representation across genes or ChIP-Seq datasets. G3 (Bethesda), 2, 987-1002. 119. Kim, D., Pertea, G., Trapnell, C., Pimentel, H., Kelley, R. and Salzberg, S.L. (2013) TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol, 14, R36. 120. Trapnell, C., Williams, B.A., Pertea, G., Mortazavi, A., Kwan, G., van Baren, M.J., Salzberg, S.L., Wold, B.J. and Pachter, L. (2010) Transcript assembly and quantification 148  by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation. Nat Biotechnol, 28, 511-515. 121. Baker, A., Saltik, M., Lehrmann, H., Killisch, I., Mautner, V., Lamm, G., Christofori, G. and Cotten, M. (1997) Polyethylenimine (PEI) is a simple, inexpensive and effective reagent for condensing and linking plasmid DNA to adenovirus for gene delivery. Gene Ther, 4, 773-782. 122. Wheatley, S., Wright, E., Jeske, Y., A, M.C., Bowles, J. and Koopman, P. (1996) Aetiology of the skeletal dysmorphology syndrome campomelic dysplasia: expression of the Sox9 gene during chondrogenesis in mouse embryos. Ann N Y Acad Sci, 785, 350-352. 123. Winter, E.M. and Gittenberger-de Groot, A.C. (2007) Epicardium-derived cells in cardiogenesis and cardiac regeneration. Cell Mol Life Sci, 64, 692-703. 124. Snarr, B.S., Kern, C.B. and Wessels, A. (2008) Origin and fate of cardiac mesenchyme. Dev Dyn, 237, 2804-2819. 125. Chatterjee, S., Sivakamasundari, V., Yap, S.P., Kraus, P., Kumar, V., Xing, X., Lim, S.L., Sng, J., Prabhakar, S. and Lufkin, T. (2014) In vivo genome-wide analysis of multiple tissues identifies gene regulatory networks, novel functions and downstream regulatory genes for Bapx1 and its co-regulation with Sox9 in the mammalian vertebral column. BMC Genomics, 15, 1072. 126. Kadaja, M., Keyes, B.E., Lin, M., Pasolli, H.A., Genander, M., Polak, L., Stokes, N., Zheng, D. and Fuchs, E. (2014) SOX9: a stem cell transcriptional regulator of secreted niche signaling factors. Genes & development, 28, 328-341. 127. Ramezani-Rad, P., Geng, H., Hurtz, C., Chan, L.N., Chen, Z., Jumaa, H., Melnick, A., Paietta, E., Carroll, W.L., Willman, C.L. et al. (2013) SOX4 enables oncogenic survival signals in acute lymphoblastic leukemia. Blood, 121, 148-155. 128. Bard-Chapeau, E.A., Jeyakani, J., Kok, C.H., Muller, J., Chua, B.Q., Gunaratne, J., Batagov, A., Jenjaroenpun, P., Kuznetsov, V.A., Wei, C.L. et al. (2012) Ecotopic viral integration site 1 (EVI1) regulates multiple cellular processes important for cancer and is a synergistic partner for FOS protein in invasive tumors. Proc Natl Acad Sci U S A, 109, 2168-2173. 129. Oh, C.D., Lu, Y., Liang, S., Mori-Akiyama, Y., Chen, D., de Crombrugghe, B. and Yasuda, H. (2014) SOX9 Regulates Multiple Genes in Chondrocytes, Including Genes Encoding ECM Proteins, ECM Modification Enzymes, Receptors, and Transporters. PLoS One, 9, e107577. 130. Ya, J., Schilham, M.W., de Boer, P.A., Moorman, A.F., Clevers, H. and Lamers, W.H. (1998) Sox4-deficiency syndrome in mice is an animal model for common trunk. Circ Res, 83, 986-994. 131. Bard-Chapeau, E.A., Szumska, D., Jacob, B., Chua, B.Q., Chatterjee, G.C., Zhang, Y., Ward, J.M., Urun, F., Kinameri, E., Vincent, S.D. et al. (2014) Mice carrying a hypomorphic Evi1 allele are embryonic viable but exhibit severe congenital heart defects. PLoS One, 9, e89397. 132. Furumatsu, T. and Asahara, H. (2010) Histone acetylation influences the activity of Sox9-related transcriptional complex. Acta Med Okayama, 64, 351-357. 133. Furumatsu, T., Ozaki, T. and Asahara, H. (2009) Smad3 activates the Sox9-dependent transcription on chromatin. Int J Biochem Cell Biol, 41, 1198-1204. 149  134. Gu, S., Boyer, T.G. and Naski, M.C. (2012) Basic helix-loop-helix transcription factor Twist1 inhibits transactivator function of master chondrogenic regulator Sox9. J Biol Chem, 287, 21082-21092. 135. Goodnough, L.H., Chang, A.T., Treloar, C., Yang, J., Scacheri, P.C. and Atit, R.P. (2012) Twist1 mediates repression of chondrogenesis by beta-catenin to promote cranial bone progenitor specification. Development, 139, 4428-4438. 136. Alva, J.A., Zovein, A.C., Monvoisin, A., Murphy, T., Salazar, A., Harvey, N.L., Carmeliet, P. and Iruela-Arispe, M.L. (2006) VE-Cadherin-Cre-recombinase transgenic mouse: a tool for lineage analysis and gene deletion in endothelial cells. Dev Dyn, 235, 759-767. 137. Vincent, S.D. and Robertson, E.J. (2003) Highly efficient transgene-independent recombination directed by a maternally derived SOX2CRE transgene. Genesis, 37, 54-56. 138. Dy, P., Wang, W., Bhattaram, P., Wang, Q., Wang, L., Ballock, R.T. and Lefebvre, V. (2012) Sox9 directs hypertrophic maturation and blocks osteoblast differentiation of growth plate chondrocytes. Dev Cell, 22, 597-609. 139. Maganti, K., Rigolin, V.H., Sarano, M.E. and Bonow, R.O. (2010) Valvular heart disease: diagnosis and management. Mayo Clin Proc, 85, 483-500. 140. Kisanuki, Y.Y., Hammer, R.E., Miyazaki, J., Williams, S.C., Richardson, J.A. and Yanagisawa, M. (2001) Tie2-Cre transgenic mice: a new model for endothelial cell-lineage analysis in vivo. Developmental biology, 230, 230-242. 141. Sarikas, A., Xu, X., Field, L.J. and Pan, Z.Q. (2008) The cullin7 E3 ubiquitin ligase: a novel player in growth control. Cell Cycle, 7, 3154-3161. 142. DeLaughter, D.M., Christodoulou, D.C., Robinson, J.Y., Seidman, C.E., Baldwin, H.S., Seidman, J.G. and Barnett, J.V. (2013) Spatial transcriptional profile of the chick and mouse endocardial cushions identify novel regulators of endocardial EMT in vitro. J Mol Cell Cardiol, 59, 196-204. 143. Vokes, S.A., Ji, H., Wong, W.H. and McMahon, A.P. (2008) A genome-scale analysis of the cis-regulatory circuitry underlying sonic hedgehog-mediated patterning of the mammalian limb. Genes Dev, 22, 2651-2663. 144. Visel, A., Blow, M.J., Li, Z., Zhang, T., Akiyama, J.A., Holt, A., Plajzer-Frick, I., Shoukry, M., Wright, C., Chen, F. et al. (2009) ChIP-seq accurately predicts tissue-specific activity of enhancers. Nature, 457, 854-858. 145. Blow, M.J., McCulley, D.J., Li, Z., Zhang, T., Akiyama, J.A., Holt, A., Plajzer-Frick, I., Shoukry, M., Wright, C., Chen, F. et al. (2010) ChIP-Seq identification of weakly conserved heart enhancers. Nat Genet, 42, 806-810. 146. Liu, C.F. and Lefebvre, V. (2015) The transcription factors SOX9 and SOX5/SOX6 cooperate genome-wide through super-enhancers to drive chondrogenesis. Nucleic acids research. 147. Ohba, S., He, X., Hojo, H. and McMahon, A.P. (2015) Distinct Transcriptional Programs Underlie Sox9 Regulation of the Mammalian Chondrocyte. Cell Rep, 12, 229-243. 148. Meech, R., Edelman, D.B., Jones, F.S. and Makarenkova, H.P. (2005) The homeobox transcription factor Barx2 regulates chondrogenesis during limb development. Development, 132, 2135-2146. 150  149. Takigawa, Y., Hata, K., Muramatsu, S., Amano, K., Ono, K., Wakabayashi, M., Matsuda, A., Takada, K., Nishimura, R. and Yoneda, T. (2010) The transcription factor Znf219 regulates chondrocyte differentiation by assembling a transcription factory with Sox9. J Cell Sci, 123, 3780-3788. 150. Yoon, B.S., Ovchinnikov, D.A., Yoshii, I., Mishina, Y., Behringer, R.R. and Lyons, K.M. (2005) Bmpr1a and Bmpr1b have overlapping functions and are essential for chondrogenesis in vivo. Proc Natl Acad Sci U S A, 102, 5062-5067. 151. Haller, R., Schwanbeck, R., Martini, S., Bernoth, K., Kramer, J., Just, U. and Rohwedel, J. (2011) Notch1 signaling regulates chondrogenic lineage determination through Sox9 activation. Cell Death Differ. 152. Dolfini, D., Gatta, R. and Mantovani, R. (2012) NF-Y and the transcriptional activation of CCAAT promoters. Crit Rev Biochem Mol Biol, 47, 29-49. 153. Semenza, G.L. (1998) Hypoxia-inducible factor 1: master regulator of O2 homeostasis. Curr Opin Genet Dev, 8, 588-594. 154. Amarilio, R., Viukov, S.V., Sharir, A., Eshkar-Oren, I., Johnson, R.S. and Zelzer, E. (2007) HIF1alpha regulation of Sox9 is necessary to maintain differentiation of hypoxic prechondrogenic cells during early skeletogenesis. Development, 134, 3917-3928. 155. Liu, H. and Fisher, S.A. (2008) Hypoxia-inducible transcription factor-1alpha triggers an autocrine survival pathway during embryonic cardiac outflow tract remodeling. Circ Res, 102, 1331-1339. 156. Kato, K., Bhattaram, P., Penzo-Mendez, A., Gadi, A. and Lefebvre, V. (2015) SOXC Transcription Factors Induce Cartilage Growth Plate Formation in Mouse Embryos by Promoting Noncanonical WNT Signaling. Journal of bone and mineral research : the official journal of the American Society for Bone and Mineral Research. 157. Neeb, Z., Lajiness, J.D., Bolanis, E. and Conway, S.J. (2013) Cardiac outflow tract anomalies. Wiley Interdiscip Rev Dev Biol, 2, 499-530. 158. Zhou, B., von Gise, A., Ma, Q., Rivera-Feliciano, J. and Pu, W.T. (2008) Nkx2-5- and Isl1-expressing cardiac progenitors contribute to proepicardium. Biochemical and biophysical research communications, 375, 450-453. 159. Cai, C.L., Martin, J.C., Sun, Y., Cui, L., Wang, L., Ouyang, K., Yang, L., Bu, L., Liang, X., Zhang, X. et al. (2008) A myocardial lineage derives from Tbx18 epicardial cells. Nature, 454, 104-108. 160. Wessels, A., van den Hoff, M.J., Adamo, R.F., Phelps, A.L., Lockhart, M.M., Sauls, K., Briggs, L.E., Norris, R.A., van Wijk, B., Perez-Pomares, J.M. et al. (2012) Epicardially derived fibroblasts preferentially contribute to the parietal leaflets of the atrioventricular valves in the murine heart. Developmental biology, 366, 111-124. 161. Alao, J.P. (2007) The regulation of cyclin D1 degradation: roles in cancer development and the potential for therapeutic invention. Mol Cancer, 6, 24. 162. Peirano, R.I. and Wegner, M. (2000) The glial transcription factor Sox10 binds to DNA both as monomer and dimer with different functional consequences. Nucleic Acids Res, 28, 3047-3055. 163. Shaulian, E. (2010) AP-1--The Jun proteins: Oncogenes or tumor suppressors in disguise? Cell Signal, 22, 894-899. 151  164. Claret, F.X., Hibi, M., Dhut, S., Toda, T. and Karin, M. (1996) A new group of conserved coactivators that increase the specificity of AP-1 transcription factors. Nature, 383, 453-457. 165. Bashur, L.A., Chen, D., Chen, Z., Liang, B., Pardi, R., Murakami, S. and Zhou, G. (2014) Loss of jab1 in osteochondral progenitor cells severely impairs embryonic limb development in mice. J Cell Physiol, 229, 1607-1617. 166. Jang, S.W., Liu, X., Fu, H., Rees, H., Yepes, M., Levey, A. and Ye, K. (2009) Interaction of Akt-phosphorylated SRPK2 with 14-3-3 mediates cell cycle and cell death in neurons. J Biol Chem, 284, 24512-24525. 167. Xu, N., Lao, Y., Zhang, Y. and Gillespie, D.A. (2012) Akt: a double-edged sword in cell proliferation and genome stability. J Oncol, 2012, 951724. 168. Bracken, A.P., Pasini, D., Capra, M., Prosperini, E., Colli, E. and Helin, K. (2003) EZH2 is downstream of the pRB-E2F pathway, essential for proliferation and amplified in cancer. EMBO J, 22, 5323-5335. 169. Kelly, R.D. and Cowley, S.M. (2013) The physiological roles of histone deacetylase (HDAC) 1 and 2: complex co-stars with multiple leading parts. Biochem Soc Trans, 41, 741-749. 170. He, A., Ma, Q., Cao, J., von Gise, A., Zhou, P., Xie, H., Zhang, B., Hsing, M., Christodoulou, D.C., Cahan, P. et al. (2012) Polycomb repressive complex 2 regulates normal development of the mouse heart. Circ Res, 110, 406-415. 171. Giono, L.E. and Manfredi, J.J. (2006) The p53 tumor suppressor participates in multiple cell cycle checkpoints. J Cell Physiol, 209, 13-20. 172. Stork, P.J. and Schmitt, J.M. (2002) Crosstalk between cAMP and MAP kinase signaling in the regulation of cell proliferation. Trends Cell Biol, 12, 258-266. 173. Huang, W., Zhou, X., Lefebvre, V. and de Crombrugghe, B. (2000) Phosphorylation of SOX9 by cyclic AMP-dependent protein kinase A enhances SOX9's ability to transactivate a Col2a1 chondrocyte-specific enhancer. Mol Cell Biol, 20, 4149-4158. 174. Formeister, E.J., Sionas, A.L., Lorance, D.K., Barkley, C.L., Lee, G.H. and Magness, S.T. (2009) Distinct SOX9 levels differentially mark stem/progenitor populations and enteroendocrine cells of the small intestine epithelium. Am J Physiol Gastrointest Liver Physiol, 296, G1108-1118. 175. Liu, C., Liu, W., Palie, J., Lu, M.F., Brown, N.A. and Martin, J.F. (2002) Pitx2c patterns anterior myocardium and aortic arch vessels and is required for local cell movement into atrioventricular cushions. Development, 129, 5081-5091. 176. Holler, K.L., Hendershot, T.J., Troy, S.E., Vincentz, J.W., Firulli, A.B. and Howard, M.J. (2010) Targeted deletion of Hand2 in cardiac neural crest-derived cells influences cardiac gene expression and outflow tract development. Dev Biol, 341, 291-304. 177. Cai, X., Zhang, W., Hu, J., Zhang, L., Sultana, N., Wu, B., Cai, W., Zhou, B. and Cai, C.L. (2013) Tbx20 acts upstream of Wnt signaling to regulate endocardial cushion formation and valve remodeling during mouse cardiogenesis. Development, 140, 3176-3187. 178. Vincentz, J.W., Barnes, R.M., Rodgers, R., Firulli, B.A., Conway, S.J. and Firulli, A.B. (2008) An absence of Twist1 results in aberrant cardiac neural crest morphogenesis. Dev Biol, 320, 131-139. 152  179. Teng, Y. and Li, X. (2014) The roles of HLH transcription factors in epithelial mesenchymal transition and multiple molecular mechanisms. Clin Exp Metastasis, 31, 367-377. 180. Shelton, E.L. and Yutzey, K.E. (2008) Twist1 function in endocardial cushion cell proliferation, migration, and differentiation during heart valve development. Dev Biol, 317, 282-295. 181. Chakraborty, S., Wirrig, E.E., Hinton, R.B., Merrill, W.H., Spicer, D.B. and Yutzey, K.E. (2010) Twist1 promotes heart valve cell proliferation and extracellular matrix gene expression during development in vivo and is expressed in human diseased aortic valves. Dev Biol, 347, 167-179. 182. Vrljicak, P., Cullum, R., Xu, E., Chang, A.C., Wederell, E.D., Bilenky, M., Jones, S.J., Marra, M.A., Karsan, A. and Hoodless, P.A. (2012) Twist1 transcriptional targets in the developing atrio-ventricular canal of the mouse. PLoS One, 7, e40815. 183. Lee, M.P. and Yutzey, K.E. (2011) Twist1 directly regulates genes that promote cell proliferation and migration in developing heart valves. PLoS One, 6, e29758. 184. Boyd, K.E., Xiao, Y.Y., Fan, K., Poholek, A., Copeland, N.G., Jenkins, N.A. and Perkins, A.S. (2006) Sox4 cooperates with Evi1 in AKXD-23 myeloid tumors via transactivation of proviral LTR. Blood, 107, 733-741. 185. Ghatak, S., Misra, S., Norris, R.A., Moreno-Rodriguez, R.A., Hoffman, S., Levine, R.A., Hascall, V.C. and Markwald, R.R. (2014) Periostin induces intracellular cross-talk between kinases and hyaluronan in atrioventricular valvulogenesis. J Biol Chem, 289, 8545-8561. 186. Snider, P., Hinton, R.B., Moreno-Rodriguez, R.A., Wang, J., Rogers, R., Lindsley, A., Li, F., Ingram, D.A., Menick, D., Field, L. et al. (2008) Periostin is required for maturation and extracellular matrix stabilization of noncardiomyocyte lineages of the heart. Circ Res, 102, 752-760.   153  APPENDIX I SOX9 immunostaining on E12.5 whole embryo (A, B, E) and E10.5 heart (C, D, F).    SOX9 is labeled in green and DAPI (to label nuclei) is labeled in blue. E. and F. Show nuclear staining in tissues know to express SOX9. Arrowheads indicate SOX9 positive regions.  154  APPENDIX II Quality verification of the SOX9 ChIP-Seq libraries and gene associations. Part A and B were generated by M. Bilenky and Part C was generated by R. Cullum.   A. Comparison of control coverage peak height versus SOX9 E12.5 AVC or B. E12.5 limb peak height to determine a local z-score threshold for peak inclusion in downstream analysis. C. UCSC genome browser screenshots of known SOX9 target genes in the limb: Col2a1, Acan, Col11a2 in the SOX9 E12.5 limb ChIP-Seq library. D. A schematic of how SOX9 peaks were associated with different genomic regions at or near the target gene. 155  APPENDIX III Characteristics of the SOX9 ChIP-Seq libraries ChIP-Seq Library Mapped Reads (M) Reads into Peaks Height Threshold Width of peaks Peak Criteria Number of Peaks Number of Genes SOX9 E12.5 AVC 6.84 39,847 8 307.3 bp (+/- 86.0 bp) Height 8+, control 30 subtracted, z-score 2.3, sub0.3 2602 2453  SOX9 E12.5 Limb 76.4 181,294 28 321.5 bp (+/- 90.3 bp) Height 28+, control 30 subtracted, z-score 2.5, sub0.5 9087 5750   156  APPENDIX IV Gene Ontology (GO) analysis on the E12.5 AVC and limb overlapping SOX9 peaks using GOrilla GO term description P-value Enrichment -10*(LOG10 (p-value)) regulation of Rab protein signal transduction 8.50E-05 22.72 40.70581 immunoglobulin V(D)J recombination 1.80E-05 18.18 47.44727 B cell lineage commitment 7.95E-04 13.63 30.99633 pyramidal neuron development 7.95E-04 13.63 30.99633 mRNA splice site selection 6.49E-04 6.68 31.87755 demethylation 8.04E-04 3.95 30.94744 negative regulation of protein localization to nucleus 9.11E-04 3.53 30.40482 regulation of organelle assembly 5.84E-04 3.44 32.33587 cellular response to nutrient levels 6.98E-04 2.96 31.56145 cellular response to extracellular stimulus 3.67E-04 2.87 34.35334 cofactor biosynthetic process 7.04E-04 2.81 31.52427 positive regulation of cytoskeleton organization 1.58E-04 2.74 38.01343 positive regulation of protein complex assembly 6.04E-04 2.62 32.18963 positive regulation of cellular component biogenesis 1.23E-05 2.45 49.10095 response to extracellular stimulus 7.83E-04 2.32 31.06238 mRNA processing 1.59E-05 2.26 47.98603 regulation of cell cycle process 2.73E-06 2.25 55.63837 mitotic nuclear division 1.52E-04 2.24 38.18156 positive regulation of organelle organization 2.76E-05 2.2 45.59091 chordate embryonic development 1.71E-04 2.18 37.67004 response to oxidative stress 4.28E-04 2.17 33.68556 in utero embryonic development 3.11E-04 2.14 35.0724 embryo development ending in birth or egg hatching 2.43E-04 2.13 36.14394 modification-dependent protein catabolic process 2.95E-04 2.07 35.30178 DNA metabolic process 4.32E-07 2.06 63.64516 ubiquitin-dependent protein catabolic process 4.21E-04 2.06 33.75718 modification-dependent macromolecule catabolic process 3.84E-04 2.04 34.15669 mRNA metabolic process 7.95E-05 2.03 40.99633 DNA repair 2.62E-04 2 35.81699 histone modification 8.17E-04 2 30.87778 RNA processing 3.24E-06 1.99 54.89455 regulation of nervous system development 5.41E-07 1.98 62.66803 covalent chromatin modification 9.49E-04 1.98 30.22734 positive regulation of nervous system development 1.38E-04 1.97 38.60121 chromatin modification 4.01E-05 1.96 43.96856 chromatin organization 1.80E-05 1.94 47.44727 regulation of neurogenesis 5.45E-06 1.93 52.63603 regulation of cell cycle 1.54E-06 1.92 58.12479 157  GO term description P-value Enrichment -10*(LOG10 (p-value)) nervous system development 9.22E-04 1.9 30.35269 regulation of neuron differentiation 1.06E-04 1.88 39.74694 positive regulation of apoptotic process 1.18E-04 1.85 39.28118 regulation of cellular component biogenesis 1.18E-04 1.85 39.28118 cell division 9.63E-04 1.85 30.16374 regulation of organelle organization 7.48E-06 1.83 51.26098 positive regulation of programmed cell death 1.45E-04 1.83 38.38632 mitotic cell cycle process 3.09E-04 1.83 35.10042 negative regulation of transcription from RNA polymerase II promoter 1.96E-05 1.81 47.07744 regulation of cell development 1.74E-05 1.77 47.59451 positive regulation of cell death 2.68E-04 1.77 35.71865 cellular macromolecule catabolic process 9.41E-04 1.74 30.2641 nucleic acid metabolic process 2.58E-17 1.72 165.8838 negative regulation of RNA metabolic process 3.84E-06 1.72 54.15669 cellular response to DNA damage stimulus 6.48E-04 1.72 31.88425 positive regulation of cellular component organization 6.18E-05 1.7 42.09012 RNA metabolic process 1.43E-13 1.69 128.4466 negative regulation of nitrogen compound metabolic process 1.52E-06 1.69 58.18156 negative regulation of cellular macromolecule biosynthetic process 6.75E-06 1.68 51.70696 negative regulation of RNA biosynthetic process 1.44E-05 1.68 48.41638 negative regulation of nucleic acid-templated transcription 1.63E-05 1.68 47.87812 negative regulation of transcription, DNA-templated 1.85E-05 1.68 47.32828 negative regulation of nucleobase-containing compound metabolic process 4.61E-06 1.67 53.36299 negative regulation of macromolecule biosynthetic process 6.71E-06 1.66 51.73277 RNA biosynthetic process 2.04E-09 1.65 86.9037 regulation of cell differentiation 4.20E-07 1.65 63.76751 transcription from RNA polymerase II promoter 3.79E-04 1.65 34.21361 negative regulation of cell differentiation 8.53E-04 1.65 30.69051 nucleobase-containing compound metabolic process 8.34E-18 1.64 170.7883 transcription, DNA-templated 3.91E-09 1.64 84.07823 nucleic acid-templated transcription 3.91E-09 1.64 84.07823 cellular response to stress 3.02E-05 1.64 45.19993 system development 6.74E-04 1.63 31.7134 heterocycle metabolic process 3.15E-17 1.62 165.0169 macromolecule biosynthetic process 8.35E-11 1.62 100.7831 heterocycle biosynthetic process 6.07E-10 1.62 92.16811 nucleobase-containing compound biosynthetic process 1.66E-09 1.62 87.79892 macromolecular complex subunit organization 1.65E-07 1.62 67.82516 158  GO term description P-value Enrichment -10*(LOG10 (p-value)) positive regulation of cell differentiation 4.25E-04 1.61 33.71611 cellular aromatic compound metabolic process 1.69E-16 1.6 157.7211 cellular macromolecule biosynthetic process 6.06E-10 1.6 92.17527 aromatic compound biosynthetic process 2.44E-09 1.6 86.1261 cellular nitrogen compound biosynthetic process 2.60E-09 1.59 85.85027 negative regulation of gene expression 2.06E-05 1.59 46.86133 cellular nitrogen compound metabolic process 2.05E-16 1.58 156.8825 organic cyclic compound biosynthetic process 2.99E-09 1.58 85.24329 regulation of transcription from RNA polymerase II promoter 1.06E-06 1.58 59.74694 cellular biosynthetic process 5.65E-12 1.57 112.4795 negative regulation of cellular biosynthetic process 4.62E-05 1.57 43.35358 negative regulation of biosynthetic process 4.86E-05 1.57 43.13364 positive regulation of developmental process 1.11E-04 1.57 39.54677 organic cyclic compound metabolic process 1.77E-15 1.56 147.5203 cellular macromolecule metabolic process 4.60E-20 1.55 193.3724 nitrogen compound metabolic process 5.68E-16 1.54 152.4565 biosynthetic process 4.61E-11 1.52 103.363 positive regulation of transcription from RNA polymerase II promoter 6.11E-04 1.52 32.13959 organic substance biosynthetic process 1.92E-10 1.51 97.16699 cellular component organization or biogenesis 3.39E-12 1.5 114.698 cellular component organization 7.67E-12 1.5 111.152 regulation of multicellular organismal development 5.81E-05 1.49 42.35824 positive regulation of RNA metabolic process 2.12E-04 1.49 36.73664 macromolecule modification 3.46E-06 1.48 54.60924 regulation of developmental process 6.07E-06 1.47 52.16811 protein localization 3.42E-04 1.47 34.65974 regulation of RNA metabolic process 2.99E-08 1.46 75.24329 regulation of transcription, DNA-templated 7.24E-08 1.46 71.40261 organelle organization 1.01E-05 1.46 49.95679 regulation of cellular component organization 3.93E-05 1.46 44.05607 cellular component assembly 1.96E-04 1.46 37.07744 positive regulation of transcription, DNA-templated 5.28E-04 1.46 32.77366 positive regulation of nucleic acid-templated transcription 5.28E-04 1.46 32.77366 positive regulation of RNA biosynthetic process 5.54E-04 1.46 32.5649 macromolecule metabolic process 2.45E-16 1.45 156.1083 regulation of nucleobase-containing compound metabolic process 1.97E-08 1.45 77.05534 regulation of cellular macromolecule biosynthetic process 2.10E-08 1.45 76.77781 regulation of nucleic acid-templated transcription 1.03E-07 1.45 69.87163     159  GO term description P-value Enrichment -10*(LOG10 (p-value)) positive regulation of gene expression 2.10E-04 1.45 36.77781 cellular catabolic process 2.31E-04 1.45 36.36388 macromolecule localization 4.57E-04 1.45 33.40084 cellular protein modification process 3.03E-05 1.44 45.18557 protein modification process 3.03E-05 1.44 45.18557 negative regulation of macromolecule metabolic process 4.54E-05 1.44 43.42944 single-organism organelle organization 2.42E-04 1.44 36.16185 establishment of localization in cell 7.33E-04 1.44 31.34896 regulation of macromolecule biosynthetic process 6.46E-08 1.43 71.89767 positive regulation of nucleobase-containing compound metabolic process 4.57E-04 1.43 33.40084 cellular metabolic process 6.76E-19 1.42 181.7005 regulation of nitrogen compound metabolic process 3.18E-08 1.42 74.97573 cellular protein metabolic process 9.94E-06 1.42 50.02614 regulation of gene expression 5.39E-08 1.41 72.68411 regulation of cellular biosynthetic process 1.82E-07 1.4 67.39929 regulation of biosynthetic process 1.92E-07 1.4 67.16699 negative regulation of cellular metabolic process 1.64E-04 1.4 37.85156 positive regulation of cellular process 1.42E-07 1.37 68.47712 negative regulation of metabolic process 1.38E-04 1.37 38.60121 positive regulation of macromolecule metabolic process 1.25E-04 1.36 39.0309 cellular response to stimulus 5.91E-04 1.36 32.28413 anatomical structure development 8.37E-05 1.35 40.77275 primary metabolic process 6.43E-14 1.34 131.9179 regulation of macromolecule metabolic process 5.23E-08 1.34 72.81498 organic substance metabolic process 3.03E-14 1.33 135.1856 regulation of primary metabolic process 1.07E-07 1.33 69.70616 metabolic process 6.49E-14 1.31 131.8776 regulation of cellular metabolic process 9.69E-07 1.3 60.13676 regulation of metabolic process 1.03E-06 1.27 59.87163 positive regulation of biological process 1.29E-05 1.27 48.8941 developmental process 2.53E-05 1.27 45.96879 single-organism metabolic process 1.91E-04 1.26 37.18967 single-organism developmental process 1.74E-04 1.25 37.59451 cellular process 9.73E-14 1.23 130.1189 negative regulation of biological process 5.73E-04 1.23 32.41845 single-organism cellular process 1.16E-05 1.18 49.35542  160  APPENDIX V GO analysis on the E12.5 AVC SOX9 peaks (>2 fold enrichment) GO term description P-value Enrichment -10*(LOG10 (p-value)) atrioventricular valve development 1.95E-04 8.46 37.09965 fat pad development 1.24E-04 7.05 39.06578 neutrophil homeostasis 8.82E-04 6.77 30.54531 immunoglobulin V(D)J recombination 8.82E-04 6.77 30.54531 pyramidal neuron development 8.82E-04 6.77 30.54531 branching morphogenesis of a nerve 8.82E-04 6.77 30.54531 histone H4-K20 methylation 3.93E-04 6.04 34.05607 presynaptic membrane assembly 3.93E-04 6.04 34.05607 radial glia guided migration of Purkinje cell 3.93E-04 6.04 34.05607 presynaptic membrane organization 1.66E-04 5.64 37.79892 receptor localization to synapse 9.44E-04 5.29 30.25028 chorio-allantoic fusion 9.44E-04 5.29 30.25028 cell migration in hindbrain 2.86E-04 4.56 35.43634 regulation of cardiac muscle cell contraction 1.12E-04 4.51 39.50782 positive regulation of oligodendrocyte differentiation 2.01E-04 4.23 36.96804 mitotic metaphase plate congression 3.41E-04 3.98 34.67246 atrial septum morphogenesis 8.66E-04 3.95 30.62482 regulation of actin filament-based movement 5.50E-04 3.76 32.59637 metaphase plate congression 3.44E-04 3.63 34.63442 membrane assembly 8.52E-04 3.56 30.6956 positive regulation of glial cell differentiation 1.38E-05 3.48 48.60121 regulation of oligodendrocyte differentiation 1.38E-05 3.48 48.60121 positive regulation of transforming growth factor beta receptor signaling pathway 5.21E-04 3.46 32.83162 regulation of neuron migration 1.31E-04 3.45 38.82729 adipose tissue development 1.93E-04 3.32 37.14443 heterophilic cell-cell adhesion via plasma membrane cell adhesion molecules 7.10E-05 3.23 41.48742 positive regulation of gliogenesis 1.12E-05 2.99 49.50782 regulation of myelination 5.90E-04 2.82 32.29148 regulation of cardiac muscle contraction 7.80E-04 2.74 31.07905 regulation of glial cell differentiation 1.04E-05 2.73 49.82967 negative regulation of Ras protein signal transduction 6.08E-04 2.68 32.16096 regulation of synaptic transmission, glutamatergic 2.83E-04 2.64 35.48214 cell-cell adhesion via plasma-membrane adhesion molecules 9.55E-08 2.63 70.19997 negative regulation of small GTPase mediated signal transduction 7.87E-04 2.62 31.04025 cell-cell adhesion 1.23E-07 2.61 69.10095 regulation of gliogenesis 1.87E-06 2.6 57.28158 negative regulation of osteoblast differentiation 7.69E-04 2.52 31.14074     161  GO term description P-value Enrichment -10*(LOG10 (p-value)) homophilic cell adhesion via plasma membrane adhesion molecules 9.09E-05 2.4 40.41436 regulation of cation channel activity 4.21E-04 2.4 33.75718 regulation of DNA binding 8.51E-05 2.36 40.7007 regulation of transforming growth factor beta receptor signaling pathway 4.82E-04 2.2 33.16953 regulation of dendrite development 8.30E-05 2.12 40.80922 negative regulation of cytoplasmic transport 8.17E-04 2.12 30.87778 negative regulation of cell migration 1.69E-06 2.09 57.72113 negative regulation of neuron projection development 4.08E-04 2.09 33.8934 striated muscle tissue development 7.25E-04 2.09 31.39662 muscle tissue development 2.71E-04 2.07 35.67031 negative regulation of neuron differentiation 7.17E-06 2.06 51.44481 positive regulation of cell cycle process 3.62E-06 2.05 54.41291 central nervous system neuron differentiation 4.24E-04 2.05 33.72634 negative regulation of canonical Wnt signaling pathway 7.52E-04 2.05 31.23782 negative regulation of cell motility 3.74E-06 2.03 54.27128 regulation of cellular response to growth factor stimulus 2.28E-05 2.03 46.42065 positive regulation of transmembrane receptor protein serine/threonine kinase signaling pathway 6.60E-04 2.03 31.80456 nervous system development 4.20E-10 2.02 93.76751 regulation of muscle tissue development 4.36E-04 2.01 33.60514 negative regulation of cell projection organization 4.36E-04 2.01 33.60514 negative regulation of neurogenesis 9.74E-07 2 60.11441 blood vessel development 5.07E-04 2 32.94992  162  APPENDIX VI GO analysis on the E12.5 Limb SOX9 ChIP-Seq peaks (>2fold enrichment) GO term description P-value Enrichment  -10*(LOG10 (p-value)) endocardial cushion development 3.53E-05 3.6 44.52225 positive regulation of Wnt signaling pathway, planar cell polarity pathway 1.27E-04 3.6 38.96196 regulation of Wnt signaling pathway, planar cell polarity pathway 6.75E-06 3.3 51.70696 chondrocyte development 3.19E-08 3.22 74.96209 positive regulation of non-canonical Wnt signaling pathway 2.39E-04 3.2 36.21602 negative regulation of transcription by competitive promoter binding 7.71E-04 3.15 31.12946 axon regeneration 7.71E-04 3.15 31.12946 chorio-allantoic fusion 7.71E-04 3.15 31.12946 negative regulation of retinoic acid receptor signaling pathway 7.71E-04 3.15 31.12946 spinal cord dorsal/ventral patterning 7.71E-04 3.15 31.12946 cardiac chamber formation 3.04E-04 2.95 35.17126 face development 3.04E-04 2.95 35.17126 negative regulation of fibroblast growth factor receptor signaling pathway 3.93E-05 2.88 44.05607 regulation of non-canonical Wnt signaling pathway 3.93E-05 2.88 44.05607 cardiac right ventricle morphogenesis 9.03E-04 2.88 30.44312 cardiac ventricle formation 9.03E-04 2.88 30.44312 cerebellar Purkinje cell layer development 3.26E-04 2.77 34.86782 heart valve morphogenesis 7.49E-07 2.74 61.25518 chondrocyte differentiation 5.28E-09 2.7 82.77366 lens morphogenesis in camera-type eye 1.17E-04 2.7 39.31814 embryonic digestive tract development 1.17E-04 2.7 39.31814 body morphogenesis 9.13E-04 2.7 30.39529 embryonic camera-type eye development 9.13E-04 2.7 30.39529 positive regulation of myelination 9.13E-04 2.7 30.39529 mesenchyme development 2.76E-10 2.68 95.59091 ventricular septum development 4.21E-05 2.65 43.75718 outflow tract septum morphogenesis 3.20E-04 2.64 34.9485 regulation of cardiac muscle cell contraction 3.20E-04 2.64 34.9485 negative regulation of smoothened signaling pathway 2.05E-06 2.63 56.88246 axonal fasciculation 1.13E-04 2.6 39.46922 sialylation 1.13E-04 2.6 39.46922 negative chemotaxis 4.01E-05 2.57 43.96856 negative regulation of anoikis 8.56E-04 2.57 30.67526 regulation of p38MAPK cascade 8.56E-04 2.57 30.67526 atrioventricular valve morphogenesis 8.56E-04 2.57 30.67526 mitotic G2/M transition checkpoint 8.56E-04 2.57 30.67526 negative regulation of oligodendrocyte differentiation 8.56E-04 2.57 30.67526     163  GO term description P-value Enrichment -10*(LOG10 (p-value)) dendrite morphogenesis 3.64E-08 2.56 74.38899 cartilage development involved in endochondral bone morphogenesis 2.97E-04 2.54 35.27244 cardiac septum morphogenesis 9.29E-11 2.53 100.3198 negative regulation of embryonic development 5.12E-06 2.53 52.9073 regulation of fibroblast growth factor receptor signaling pathway 5.12E-06 2.53 52.9073 ureteric bud development 1.84E-06 2.52 57.35182 neural crest cell development 1.04E-04 2.52 39.82967 negative regulation of canonical Wnt signaling pathway 2.23E-16 2.49 156.517 protein hydroxylation 7.66E-04 2.48 31.15771 regulation of cell proliferation involved in heart morphogenesis 7.66E-04 2.48 31.15771 mesenchyme morphogenesis 7.66E-04 2.48 31.15771 establishment of epithelial cell polarity 7.66E-04 2.48 31.15771 regulation of axon extension involved in axon guidance 7.66E-04 2.48 31.15771 neural tube development 2.18E-07 2.46 66.61544 cranial nerve development 2.66E-04 2.46 35.75118 negative regulation of cartilage development 9.35E-05 2.45 40.29188 positive regulation of epithelial to mesenchymal transition 1.18E-05 2.44 49.28118 positive regulation of mesenchymal cell proliferation 1.34E-06 2.4 58.72895 cardiac ventricle morphogenesis 2.91E-05 2.4 45.36107 tight junction assembly 6.66E-04 2.4 31.76526 regulation of extracellular matrix organization 6.66E-04 2.4 31.76526 regulation of actin filament-based movement 6.66E-04 2.4 31.76526 positive regulation of neurological system process 6.66E-04 2.4 31.76526 regulation of smoothened signaling pathway 3.54E-10 2.38 94.50997 neuron recognition 2.52E-05 2.36 45.98599 ventricular septum morphogenesis 2.52E-05 2.36 45.98599 mesenchymal cell development 7.08E-05 2.35 41.49967 cardiac septum development 2.00E-04 2.35 36.9897 positive regulation of stem cell differentiation 3.66E-07 2.34 64.36519 positive regulation of cardiac muscle hypertrophy 5.67E-04 2.34 32.46417 melanocyte differentiation 5.67E-04 2.34 32.46417 cell differentiation involved in kidney development 5.67E-04 2.34 32.46417 camera-type eye morphogenesis 5.67E-04 2.34 32.46417 embryonic digestive tract morphogenesis 5.67E-04 2.34 32.46417 positive regulation of muscle hypertrophy 5.67E-04 2.34 32.46417 positive regulation of cardiac muscle cell proliferation 5.67E-04 2.34 32.46417 regulation of cartilage development 1.52E-08 2.33 78.18156 mesonephric epithelium development 7.77E-06 2.33 51.09579 negative regulation of Wnt signaling pathway 4.58E-17 2.33 163.3913 regulation of mesenchymal cell proliferation 8.69E-07 2.31 60.6098 164  GO term description P-value Enrichment -10*(LOG10 (p-value)) ephrin receptor signaling pathway 1.83E-05 2.29 47.37549 mesonephric tubule development 1.83E-05 2.29 47.37549 pigment cell differentiation 4.76E-04 2.29 33.22393 mesenchymal cell differentiation 4.76E-04 2.29 33.22393 regulation of chondrocyte differentiation 2.03E-06 2.28 56.92504 collagen fibril organization 5.60E-06 2.27 52.51812 connective tissue development 1.94E-07 2.26 67.12198 cardiac muscle cell differentiation 1.54E-05 2.26 48.12479 embryonic digit morphogenesis 6.00E-08 2.25 72.21849 digestive tract development 4.70E-06 2.25 53.27902 appendage development 4.26E-05 2.25 43.7059 regulation of cardiac muscle cell proliferation 4.26E-05 2.25 43.7059 limb development 4.26E-05 2.25 43.7059 embryonic eye morphogenesis 3.96E-04 2.25 34.02305 positive regulation of cartilage development 3.96E-04 2.25 34.02305 cell differentiation in hindbrain 3.96E-04 2.25 34.02305 cellular response to transforming growth factor beta stimulus 1.18E-04 2.23 39.28118 somatic stem cell maintenance 1.20E-06 2.22 59.20819 regulation of oligodendrocyte differentiation 3.54E-05 2.22 44.50997 positive regulation of cardiac muscle tissue growth 3.26E-04 2.22 34.86782 negative regulation of glial cell differentiation 9.72E-05 2.21 40.12334 cardiac chamber morphogenesis 9.72E-05 2.21 40.12334 regulation of myelination 2.92E-05 2.2 45.34617 regulation of catenin import into nucleus 9.06E-04 2.19 30.42872 mesoderm development 2.41E-05 2.18 46.17983 positive regulation of BMP signaling pathway 7.98E-05 2.18 40.97997 cartilage development 4.96E-10 2.16 93.04518 central nervous system neuron axonogenesis 2.18E-04 2.16 36.61544 lung cell differentiation 2.18E-04 2.16 36.61544 establishment or maintenance of bipolar cell polarity 7.36E-04 2.16 31.33122 oligodendrocyte differentiation 7.36E-04 2.16 31.33122 establishment or maintenance of apical/basal cell polarity 7.36E-04 2.16 31.33122 regulation of epithelial to mesenchymal transition 1.51E-06 2.15 58.21023 eye morphogenesis 1.51E-06 2.15 58.21023 epithelial to mesenchymal transition 1.61E-05 2.14 47.93174 response to transforming growth factor beta 1.77E-04 2.14 37.52027 mammary gland development 1.77E-04 2.14 37.52027 regulation of organ formation 1.77E-04 2.14 37.52027 regulation of keratinocyte proliferation 5.95E-04 2.13 32.25483 positive regulation of smoothened signaling pathway 5.95E-04 2.13 32.25483 165  GO term description P-value Enrichment -10*(LOG10 (p-value)) regulation of neuron migration 5.95E-04 2.13 32.25483 synapse assembly 1.43E-04 2.12 38.44664 outflow tract morphogenesis 1.07E-05 2.11 49.70616 positive regulation of heart growth 4.79E-04 2.11 33.19664 lung epithelial cell differentiation 4.79E-04 2.11 33.19664 adipose tissue development 4.79E-04 2.11 33.19664 regulation of canonical Wnt signaling pathway 1.95E-15 2.11 147.0997 neural tube closure 1.08E-08 2.09 79.66576 negative regulation of cellular carbohydrate metabolic process 3.85E-04 2.09 34.14539 nerve development 3.85E-04 2.09 34.14539 proteoglycan metabolic process 9.33E-05 2.08 40.30118 negative regulation of carbohydrate metabolic process 9.33E-05 2.08 40.30118 sensory organ morphogenesis 1.15E-06 2.07 59.39302 kidney epithelium development 7.51E-05 2.07 41.2436 regulation of cardiac muscle hypertrophy 3.09E-04 2.07 35.10042 positive regulation of cardiac muscle tissue development 3.09E-04 2.07 35.10042 embryonic hindlimb morphogenesis 3.09E-04 2.07 35.10042 stem cell development 3.09E-04 2.07 35.10042 tube closure 1.84E-08 2.06 77.35182 regulation of cardiac muscle tissue growth 6.04E-05 2.06 42.18963 regulation of stem cell differentiation 2.48E-09 2.05 86.05548 glial cell differentiation 7.49E-07 2.05 61.25518 hindlimb morphogenesis 4.84E-05 2.05 43.15155 embryonic skeletal system development 4.84E-05 2.05 43.15155 regulation of astrocyte differentiation 8.21E-04 2.04 30.85657 developmental growth involved in morphogenesis 2.51E-08 2.03 76.00326 negative regulation of cellular response to growth factor stimulus 7.97E-08 2.03 70.98542 regulation of heart growth 3.11E-05 2.03 45.0724 neuroepithelial cell differentiation 6.53E-04 2.03 31.85087 appendage morphogenesis 8.48E-11 2.03 100.716 limb morphogenesis 8.48E-11 2.03 100.716 peptidyl-tyrosine dephosphorylation 2.49E-05 2.02 46.03801 actin filament bundle organization 1.26E-04 2.02 38.99629 tube formation 3.59E-11 2.02 104.4491 regulation of BMP signaling pathway 1.63E-07 2.01 67.87812 regulation of morphogenesis of a branching structure 4.98E-06 2.01 53.02771 cardiocyte differentiation 1.99E-05 2.01 47.01147 positive regulation of axon extension 5.19E-04 2.01 32.84833 heterophilic cell-cell adhesion via plasma membrane cell adhesion molecules 5.19E-04 2.01 32.84833 regulation of muscle hypertrophy 5.19E-04 2.01 32.84833 166  GO term description P-value Enrichment -10*(LOG10 (p-value)) regulation of embryonic development 5.48E-09 2 82.61219 establishment or maintenance of cell polarity 2.68E-08 2 75.71865 negative regulation of stem cell differentiation 4.12E-04 2 33.85103 regulation of Wnt signaling pathway 2.55E-17 2 165.9346  167  APPENDIX VII Co-factor analysis on E12.5 AVC SOX9 peaks using oPOSSUM TF Name Family Z-score Fisher score KS score NFYA NFY CCAAT-binding 198.475 39.548 1.06 Foxd3 Forkhead 38.57 0 10.733 ELK4 Ets 31.851 0.092 1.841 GABPA Ets 31.781 0 9.582 Arnt::Ahr Helix-Loop-Helix 28.338 0 12.047 ELK1 Ets 27.371 0 3.325 E2F1 E2F 25.698 0.038 2.588 MZF1_5-13 BetaBetaAlpha-zinc finger 22.501 0 2.384 MIZF BetaBetaAlpha-zinc finger 20.69 1.115 0.857 RREB1 BetaBetaAlpha-zinc finger 19.293 0.077 1.279 Sox17 High Mobility Group 18.322 0 2.194 znf143 BetaBetaAlpha-zinc finger 16.135 0.394 3.373 Gfi BetaBetaAlpha-zinc finger 14.88 0 3.395 Pax4 Homeo 12.267 0.91 0.337 Egr1 BetaBetaAlpha-zinc finger 10.597 0 5.781 Myb Myb 10.562 0 1.161 Arnt Helix-Loop-Helix 9.454 0 3.251 SOX9 High Mobility Group 8.339 0 3.356 Pax5 Homeo 6.826 0.101 1.02 Pax6 Homeo 6.498 0.241 0.382 HIF1A::ARNT Helix-Loop-Helix 5.448 0 3.784 SRF MADS 5.162 0.164 1.222 Myc Helix-Loop-Helix 5.006 0 7.057 STAT1 Stat 4.411 0 3.3 USF1 Helix-Loop-Helix 4.239 0 7.264 Stat3 Stat 4.154 0 2.299 SP1 BetaBetaAlpha-zinc finger 3.791 0 16.503 Gata1 GATA 3.617 0 2.903 TAL1::TCF3 Helix-Loop-Helix 3.081 0 1.921 Tal1::Gata1 Helix-Loop-Helix 2.982 0 2.236 TLX1::NFIC Homeo::Nuclear Factor I-CCAAT-binding 2.977 0.037 0.269 Klf4 BetaBetaAlpha-zinc finger 2.619 0 16.978 Mycn Helix-Loop-Helix 2.369 0 6.218 CREB1 Leucine Zipper 1.899 0 1.753 Ar Hormone-nuclear Receptor 1.759 0.097 0.377 Hand1::Tcfe2a Helix-Loop-Helix 1.163 0 0.764 YY1 BetaBetaAlpha-zinc finger 0.86 0 1.79 MYC::MAX Helix-Loop-Helix 0.061 0 6.456      168  TF Name Family Z-score Fisher score KS score Nr2e3 Hormone-nuclear Receptor -0.291 0 0.336 SPIB Ets -0.765 0 1.333 PBX1 Homeo -2.132 0 5.21 Spz1 Other -2.207 0 1.678 T T -2.275 0 1.145 NFATC2 Rel -2.319 0 8.571 IRF2 IRF -2.424 0.013 0.924 TBP TATA-binding -2.454 0 1.693 RXRA::VDR Hormone-nuclear Receptor -3.107 0.01 2.188 NR3C1/glucocorticoid receptor Hormone-nuclear Receptor -3.139 0 1.29 FOXI1 Forkhead -3.324 0 1.81 EWSR1-FLI1 Ets -3.539 0.026 3.024 TP53 Loop-Sheet-Helix -3.658 0 NA NR2F1 Hormone-nuclear Receptor -4.251 0 2.986 RUNX1 Runt -4.789 0 4.038 Evi1 BetaBetaAlpha-zinc finger -5.29 0 0 SRY High Mobility Group -5.933 0 6.734 IRF1 IRF -6.826 0 1.044 RELA Rel -6.829 0 0.848 TEAD1 Homeo -7.2 0 0.945 PPARG Hormone-nuclear Receptor -7.925 0 NA REL Rel -8.643 0 1.707 NR1H2::RXRA Hormone-nuclear Receptor -9.215 0 0.833 MAX Helix-Loop-Helix -10.207 0 3.379 ESR1 estrogen receptor Hormone-nuclear Receptor -10.391 0 0.298 REST BetaBetaAlpha-zinc finger -10.395 0 4.182 PLAG1 BetaBetaAlpha-zinc finger -10.405 0 1.142 FOXA1 Forkhead -10.416 0 1.502 SPI1 Ets -10.973 0 5.005 Zfp423 BetaBetaAlpha-zinc finger -11.668 0 2.002 Sox2 High Mobility Group -11.798 0 0.358 Tcfcp2l1 CP2 -12.09 0 0.711 FEV Ets -12.355 0 1.916 RORA_2 Hormone-nuclear Receptor -12.449 0 3.055 RXR::RAR_DR5 Hormone-nuclear Receptor -12.614 0 0.679 ELF5 Ets -12.781 0 0.445 FOXF2 Forkhead -13.08 0 0.705 FOXO3 Forkhead -13.195 0 1.016 FOXD1 Forkhead -13.33 0 0.815      169  TF Name Family Z-score Fisher score KS score HNF4A Hormone-nuclear Receptor -14.244 0 6.243 NFKB1 Rel -14.302 0 0.607 NF-kappaB Rel -14.45 0 0.731 ZNF354C BetaBetaAlpha-zinc finger -15.125 0 2.976 HLF Leucine Zipper -15.155 0 1.383 ESR2 Hormone-nuclear Receptor -15.313 0 4.212 Myf Helix-Loop-Helix -15.41 0 4.787 Nobox Homeo -15.741 0 3.037 Nkx3-2 Homeo -16.487 0 2.24 Foxq1 Forkhead -16.879 0 1.355 Ddit3::Cebpa Leucine Zipper -17.045 0 5.092 NR4A2 Hormone-nuclear Receptor -17.765 0 1.421 NHLH1 Helix-Loop-Helix -17.806 0 0.814 Sox5 High Mobility Group -18.321 0 6.668 NFE2L2 Leucine Zipper -18.347 0 2.082 EBF1 Helix-Loop-Helix -18.387 0 2.684 RORA_1 Hormone-nuclear Receptor -18.965 0 0.495 CEBPA Leucine Zipper -19.414 0 1.749 NKX3-1 Homeo -19.869 0 6.957 MEF2A MADS -20.129 0 0.448 ARID3A Arid -20.208 0 13.33 HNF1B Homeo -20.221 0 0.311 HNF1A Homeo -20.293 0 0.018 INSM1 BetaBetaAlpha-zinc finger -20.383 0 0.431 AP1 Leucine Zipper -20.659 0 1.066 NFIL3 Leucine Zipper -20.784 0 0.842 MZF1_1-4 BetaBetaAlpha-zinc finger -21.086 0 4.195 Pou5f1 Homeo -21.101 0 7.828 Esrrb Hormone-nuclear Receptor -21.905 0 0.565 PPARG::RXRA Hormone-nuclear Receptor -24.086 0 0.976 Foxa2 Forkhead -24.433 0 0.62 Nkx2-5 Homeo -25.019 0 15.005 CTCF BetaBetaAlpha-zinc finger -25.447 0 7.839 Lhx3 Homeo -26.743 0 2.374 Pdx1 Homeo -31.159 0 13.441 ZEB1 BetaBetaAlpha-zinc finger -32.138 0 4.17 Prrx2 Homeo -32.235 0 10.028 Zfx BetaBetaAlpha-zinc finger -36.939 0 2.924 HOXA5 Homeo -38.02 0 8.106  170  APPENDIX VIII Co-factor analysis on E12.5 Limb SOX9 ChIP-Seq peaks using oPOSSUM TF Name Family Z-score Fisher score KS score NFYA NFY CCAAT-binding 158.069 12.692 9.07 SOX9 High Mobility Group 126.862 23.474 inf Arnt::Ahr Helix-Loop-Helix 107.402 0 10.71 Sox17 High Mobility Group 84.886 0 inf SRY High Mobility Group 79.23 0 inf Nobox Homeo 51.933 0 26.905 FOXO3 Forkhead 48.618 0 13.324 Sox5 High Mobility Group 42.458 0 inf NFATC2 Rel 40.316 0 19.607 TBP TATA-binding 39.536 0 22.72 FOXD1 Forkhead 37.307 0 13.622 ELK4 Ets 33.496 0 3.971 FOXA1 Forkhead 31.86 0 6.198 ELK1 Ets 29.737 0 7.474 znf143 BetaBetaAlpha-zinc finger 24.473 0.255 1.102 E2F1 E2F 24.222 0 2.799 TAL1::TCF3 Helix-Loop-Helix 21.905 0 2.354 RUNX1 Runt 21.887 0 11.594 Myb Myb 20.083 0 4.638 RREB1 BetaBetaAlpha-zinc finger 20.006 0.006 1.512 Sox2 High Mobility Group 17.821 0.001 5.323 Nr2e3 Hormone-nuclear Receptor 16.391 0 1.054 ARID3A Arid 16.313 0 inf REST BetaBetaAlpha-zinc finger 15.684 0 17.27 Foxa2 Forkhead 15.307 0 2.252 GABPA Ets 13.726 0 17.468 MIZF BetaBetaAlpha-zinc finger 13.56 0 2.097 Foxq1 Forkhead 11.655 0 14.274 TLX1::NFIC Homeo::Nuclear Factor I-CCAAT-binding 11.432 0.022 0.503 Arnt Helix-Loop-Helix 9.015 0 15.301 Pdx1 Homeo 8.98 0 29.36 Myf Helix-Loop-Helix 8.732 0 12.504 NR3C1 Hormone-nuclear Receptor 8.428 0 4.618 NFIL3 Leucine Zipper 8.306 0 1.583 TP53 Loop-Sheet-Helix 7.438 0.693 0.405 Ar Hormone-nuclear Receptor 6.944 0.027 5.734 Stat3 Stat 6.376 0 0.723 T T 5.47 0 0.413 Hand1::Tcfe2a Helix-Loop-Helix 5.015 0 2.47 Egr1 BetaBetaAlpha-zinc finger 4.831 0 2.551 171  TF Name Family Z-score Fisher score KS score TEAD1 Homeo 4.751 0 0.474 Pax4 Homeo 4.579 0.191 0.722 FOXI1 Forkhead 4.299 0 20.282 STAT1 Stat 2.854 0 1.349 PBX1 Homeo 2.478 0 1.979 SPIB Ets 1.959 0 1.37 Nkx2-5 Homeo 1.69 0 inf HIF1A::ARNT Helix-Loop-Helix 0.688 0 8.938 FOXF2 Forkhead 0.519 0 2.212 CEBPA Leucine Zipper 0.24 0 14.912 HNF1B Homeo 0.223 0 2.913 ELF5 Ets 0.131 0 3.574 CREB1 Leucine Zipper -0.404 0 0.737 NHLH1 Helix-Loop-Helix -1.791 0 3.673 Pax6 Homeo -2.146 0 0.317 Pax5 Homeo -2.214 0 0.628 Gfi BetaBetaAlpha-zinc finger -2.869 0 5.289 MYC::MAX Helix-Loop-Helix -2.941 0 6.58 FEV Ets -3.018 0 16.956 HNF1A Homeo -3.02 0 2.607 Myc Helix-Loop-Helix -3.796 0 8.464 NR1H2::RXRA Hormone-nuclear Receptor -5.658 0.005 2.565 REL Rel -6.14 0 2.427 HOXA5 Homeo -6.463 0 26.411 Tcfcp2l1 CP2 -6.666 0 11.518 USF1 Helix-Loop-Helix -7.045 0 22.988 RELA Rel -7.601 0 5.184 PLAG1 BetaBetaAlpha-zinc finger -8.018 0 0.219 Mycn Helix-Loop-Helix -8.085 0 11.659 RXRA::VDR Hormone-nuclear Receptor -9.307 0 0.55 CTCF BetaBetaAlpha-zinc finger -10.641 0 inf RORA_2 Hormone-nuclear Receptor -10.729 0 0.183 MZF1_5-13 BetaBetaAlpha-zinc finger -10.77 0 3.058 PPARG Hormone-nuclear Receptor -11.813 0 0.485 SPI1 Ets -11.905 0 18.301 RXR::RAR_DR5 Hormone-nuclear Receptor -12.632 0 0.151 SP1 BetaBetaAlpha-zinc finger -13.081 0 15.687 SRF MADS -13.32 0 0.514 Lhx3 Homeo -14.774 0 3.375 NKX3-1 Homeo -16.276 0 4.105 Spz1 Other -16.922 0 1.57 172  TF Name Family Z-score Fisher score KS score Foxd3 Forkhead -17.417 0 24.447 HLF Leucine Zipper -17.471 0 1.01 EWSR1-FLI1 Ets -19.461 0 1.946 Pou5f1 Homeo -19.626 0 6.752 ESR1 Hormone-nuclear Receptor -19.944 0 2.899 YY1 BetaBetaAlpha-zinc finger -20.346 0 11.262 IRF1 IRF -20.689 0 0.406 IRF2 IRF -21.023 0 1.204 INSM1 BetaBetaAlpha-zinc finger -24.527 0 2.253 MEF2A MADS -24.844 0 14.004 RORA_1 Hormone-nuclear Receptor -24.911 0 16.866 Zfp423 BetaBetaAlpha-zinc finger -24.935 0 2.117 NF-kappaB Rel -26.889 0 2.164 ESR2 Hormone-nuclear Receptor -27.728 0 5.333 Klf4 BetaBetaAlpha-zinc finger -28.154 0 25.153 EBF1 Helix-Loop-Helix -28.941 0 3.851 MAX Helix-Loop-Helix -29.2 0 8.283 NR2F1 Hormone-nuclear Receptor -29.349 0 13.531 Evi1 BetaBetaAlpha-zinc finger -30.224 0 2.405 NFE2L2 Leucine Zipper -30.379 0 13.965 Prrx2 Homeo -30.551 0 14.077 NFKB1 Rel -31.456 0 2.732 PPARG::RXRA Hormone-nuclear Receptor -33.146 0 5.169 Ddit3::Cebpa Leucine Zipper -34.676 0 0.776 AP1 Leucine Zipper -36.486 0 3.356 HNF4A Hormone-nuclear Receptor -37.596 0 7.214 ZNF354C BetaBetaAlpha-zinc finger -37.804 0 15.248 NR4A2 Hormone-nuclear Receptor -41.454 0 8.67 Nkx3-2 Homeo -43.266 0 3.417 Esrrb Hormone-nuclear Receptor -47.361 0 21.754 MZF1_1-4 BetaBetaAlpha-zinc finger -50.975 0 5.293 ZEB1 BetaBetaAlpha-zinc finger -57.591 0 15.323 Zfx BetaBetaAlpha-zinc finger -65.478 0 2.103 Tal1::Gata1 Helix-Loop-Helix -86.39 0 18.036 Gata1 GATA -90.866 0 12.197  173  APPENDIX IX Co-factor analysis on E12.5 AVC and Limb overlapping SOX9 peaks using oPOSSUM TF Name Family Z-score Fisher score KS score NFYA NFY CCAAT-binding 230.391 36.368 7.472 SOX9 High Mobility Group 114.989 2.762 inf Arnt::Ahr Helix-Loop-Helix 108.149 0 4.151 Sox17 High Mobility Group 83.192 0 inf SRY High Mobility Group 64.902 0 inf ELK4 Ets 44.534 0 6.128 ELK1 Ets 39.113 0 9.663 Nobox Homeo 36.639 0 28.151 FOXO3 Forkhead 36.064 0 14.221 NFATC2 Rel 34.236 0 28.366 E2F1 E2F 33.365 0 3.659 TBP TATA-binding 32.724 0 22.547 znf143 BetaBetaAlpha-zinc finger 29.187 0.225 3.196 RREB1 BetaBetaAlpha-zinc finger 26.91 0.002 0.598 Sox5 High Mobility Group 26.815 0 inf GABPA Ets 26.637 0 26.61 FOXD1 Forkhead 25.686 0 13.686 Myb Myb 22.756 0 5.986 FOXA1 Forkhead 22.445 0 4.374 MIZF BetaBetaAlpha-zinc finger 21.525 0.003 3.054 TAL1::TCF3 Helix-Loop-Helix 20.773 0 1.596 RUNX1 Runt 17.368 0 14.43 Nr2e3 Hormone-nuclear Receptor 14.197 0 0.49 Arnt Helix-Loop-Helix 12.404 0 17.215 TLX1::NFIC Homeo::Nuclear Factor I-CCAAT-binding 11.639 0.005 0.276 Pax4 Homeo 9.88 0.354 0.44 REST BetaBetaAlpha-zinc finger 9.53 0 18.768 Sox2 High Mobility Group 9.324 0 4.348 Egr1 BetaBetaAlpha-zinc finger 9.259 0 5.81 Stat3 Stat 7.655 0 0.62 Ar Hormone-nuclear Receptor 7.121 0.013 4.855 NR3C1 Hormone-nuclear Receptor 5.884 0 4.918 Hand1::Tcfe2a Helix-Loop-Helix 4.962 0 2.597 STAT1 Stat 4.622 0 2.174 Gfi BetaBetaAlpha-zinc finger 4.53 0 7.787 Foxd3 Forkhead 4.306 0 28.679 T T 3.902 0 0.067 HIF1A::ARNT Helix-Loop-Helix 3.179 0 11.859 ARID3A Arid 3.027 0 inf 174  TF Name Family Z-score Fisher score KS score TP53 Loop-Sheet-Helix 2.781 0.375 1.099 FOXI1 Forkhead 1.618 0 20.268 SPIB Ets 1.402 0 2.973 Pax5 Homeo 1.219 0 0.762 TEAD1 Homeo 1.039 0 0.568 PBX1 Homeo 0.907 0 4.661 MZF1_5-13 BetaBetaAlpha-zinc finger 0.82 0 3.638 Pax6 Homeo 0.81 0 0.052 Foxq1 Forkhead 0.694 0 13.617 Foxa2 Forkhead 0.628 0 1.024 Myf Helix-Loop-Helix 0.607 0 16.752 CREB1 Leucine Zipper 0.593 0 2.212 Myc Helix-Loop-Helix -1.015 0 11.258 MYC::MAX Helix-Loop-Helix -2.531 0 11.206 NFIL3 Leucine Zipper -3.987 0 0.687 USF1 Helix-Loop-Helix -4.261 0 23.441 FOXF2 Forkhead -5.805 0 2.39 ELF5 Ets -5.994 0 3.914 Mycn Helix-Loop-Helix -6.03 0 14.956 Pdx1 Homeo -8.166 0 inf FEV Ets -8.469 0 16.07 REL Rel -9.294 0 0.618 SRF MADS -9.549 0 1.638 RXRA::VDR Hormone-nuclear Receptor -9.667 0 0.751 CEBPA Leucine Zipper -9.667 0 15.168 SP1 BetaBetaAlpha-zinc finger -9.756 0 25.403 RELA Rel -9.849 0 2.747 NHLH1 Helix-Loop-Helix -9.874 0 2.696 NR1H2::RXRA Hormone-nuclear Receptor -9.931 0 4.079 HNF1B Homeo -10.556 0 3.164 Tcfcp2l1 CP2 -11.49 0 10.385 Nkx2-5 Homeo -11.654 0 inf PLAG1 BetaBetaAlpha-zinc finger -12.044 0 0.011 HNF1A Homeo -13.587 0 1.837 PPARG Hormone-nuclear Receptor -14.079 0 0.835 RORA_2 Hormone-nuclear Receptor -15.548 0 0.366 SPI1 Ets -15.597 0 23.064 Spz1 Other -15.998 0 2.814 RXR::RAR_DR5 Hormone-nuclear Receptor -17.09 0 0.107 YY1 BetaBetaAlpha-zinc finger -17.53 0 10.936 EWSR1-FLI1 Ets -19.215 0 1.774 175  TF Name Family Z-score Fisher score KS score IRF2 IRF -20.07 0 1.118 CTCF BetaBetaAlpha-zinc finger -21.33 0 inf IRF1 IRF -21.469 0 1.341 ESR1 Hormone-nuclear Receptor -22.396 0 1.753 HLF Leucine Zipper -22.662 0 0.854 Klf4 BetaBetaAlpha-zinc finger -23.679 0 inf HOXA5 Homeo -24.581 0 34.252 NKX3-1 Homeo -24.614 0 8.469 Lhx3 Homeo -27.367 0 4.691 Zfp423 BetaBetaAlpha-zinc finger -27.394 0 1.259 Pou5f1 Homeo -27.74 0 9.828 NR2F1 Hormone-nuclear Receptor -27.889 0 13.979 Evi1 BetaBetaAlpha-zinc finger -28.83 0 1.495 NF-kappaB Rel -30.428 0 1.466 MAX Helix-Loop-Helix -30.549 0 7.222 RORA_1 Hormone-nuclear Receptor -30.998 0 12.518 INSM1 BetaBetaAlpha-zinc finger -31.084 0 2.862 ESR2 Hormone-nuclear Receptor -31.593 0 6.775 MEF2A MADS -31.916 0 11.244 EBF1 Helix-Loop-Helix -34.008 0 3.41 NFKB1 Rel -34.469 0 1.314 NFE2L2 Leucine Zipper -35.357 0 15.804 Ddit3::Cebpa Leucine Zipper -38.586 0 2.868 HNF4A Hormone-nuclear Receptor -39.885 0 9.767 PPARG::RXRA Hormone-nuclear Receptor -40.293 0 3.076 ZNF354C BetaBetaAlpha-zinc finger -40.302 0 17.371 AP1 Leucine Zipper -41.801 0 4.691 Prrx2 Homeo -42.856 0 21.505 NR4A2 Hormone-nuclear Receptor -44.91 0 5.274 Nkx3-2 Homeo -45.861 0 3.077 Esrrb Hormone-nuclear Receptor -52.043 0 17.442 MZF1_1-4 BetaBetaAlpha-zinc finger -54.772 0 8.3 ZEB1 BetaBetaAlpha-zinc finger -65.784 0 18.683 Zfx BetaBetaAlpha-zinc finger -75.099 0 3.942 Tal1::Gata1 Helix-Loop-Helix -75.525 0 9.895 Gata1 GATA -78.193 0 5.479  176  APPENDIX X Comparisons of SOX9 peaks in the AVC and limb with SMAD3 (A.) and TWIST1 (B.) peaks in the E11.5 and E10.5 AVC respectively.    177  APPENDIX XI Top 100 differentially expressed genes in the Sox9 cKO AVC RNA-Seq Gene symbol WT FPKM Sox9cKO FPKM p-value  Fold change down in cKO Fold change up in cKO Egr1 10.22 47.26 5.00E-05 0.218 4.590 Bhlhe40 12.17 56.04 5.00E-05 0.219 4.575 Shisa2 30.94 8.33 5.00E-05 3.681 0.272 Lum 75.59 18.94 5.00E-05 3.976 0.252 Col9a1 12.62 1.44 5.00E-05 8.239 0.121 Col9a3 12.10 1.35 5.00E-05 8.406 0.119 Papss2 29.53 3.01 5.00E-05 9.533 0.105 Mstn 9.79 0.84 5.00E-05 10.465 0.096 Btn1a1 6.92 0.29 5.00E-05 18.197 0.055 Fos 2.45 13.97 0.00015 0.181 5.522 Prelp 3.97 0.32 0.00015 9.701 0.103 Hmgcs2 9.68 2.52 0.0002 3.726 0.268 Ndrg1 5.81 19.15 0.00025 0.307 3.256 Ier3 20.74 76.65 0.00035 0.272 3.682 Tnrc6b 66.09 13.91 0.00035 4.724 0.212 Mir208a 1248.83 0.00 0.00045 1248.830 0.000 Dct 3.90 0.47 0.00055 6.970 0.143 Ctnna2 3.30 0.56 0.0006 5.141 0.195 Kdm5d 0.62 4.73 0.00065 0.148 6.742 Col6a6 4.10 1.19 0.001 3.246 0.308 Iigp1 5.39 1.40 0.001 3.665 0.273 Ddx3y 1.81 9.76 0.0012 0.194 5.160 Uty 0.72 4.90 0.0015 0.164 6.111 Shroom1 3.70 0.92 0.0015 3.718 0.269 Stc2 2.58 9.35 0.0017 0.284 3.523 Dpysl4 3.17 0.60 0.0018 4.664 0.214 Nppb 121.27 346.05 0.0021 0.351 2.852 Slc17a7 3.49 0.77 0.0021 4.138 0.242 Serpine1 2.91 8.96 0.0023 0.332 3.013 Ntsr1 2.59 0.61 0.0023 3.803 0.263 Hbb-bh1 556.69 1682.09 0.0024 0.331 3.021 2610203C20Rik 25.89 8.92 0.0025 2.881 0.347 Car3 11.84 2.90 0.0025 3.983 0.251 Ntn4 9.84 3.24 0.0026 2.975 0.336 Adcy2 0.96 3.65 0.0026 0.284 3.527 Pitx2 19.54 7.02 0.0028 2.758 0.363 Sox5 4.33 1.43 0.0028 2.902 0.345 Derl3 0.77 5.65 0.0029 0.152 6.597 Col9a2 6.08 1.80 0.0030 3.259 0.307 178  Gene symbol WT FPKM Sox9cKO FPKM p-value  Fold change down in cKO Fold change up in cKO Fdx1l 185.14 48.86 0.0030 3.783 0.264 AI646023 1.74 0.41 0.0031 3.580 0.279 Mfap4 60.79 22.22 0.0032 2.729 0.366 Vsnl1 14.24 4.45 0.0034 3.148 0.318 Twist1 37.16 13.38 0.0035 2.765 0.362 Matn4 4.16 0.26 0.0035 11.711 0.085 Fgfr2 19.88 7.65 0.0035 2.577 0.388 Rtn1 8.40 2.47 0.0041 3.309 0.302 Serpinb9b 0.72 3.42 0.0044 0.233 4.285 Shox2 8.56 2.38 0.0044 3.496 0.286 Sfrp2 24.36 8.74 0.0047 2.767 0.361 Eif2s3y 2.32 11.09 0.0048 0.217 4.617 Mt1 32.99 93.26 0.0048 0.354 2.821 9030425E11Rik 17.83 7.00 0.0058 2.527 0.396 Rn45s 2519.92 3336.26 0.0060 0.755 1.324 Oprl1 6.71 2.40 0.0063 2.728 0.367 Btn2a2 1.51 0.37 0.0065 3.441 0.291 Slc2a3 16.24 39.24 0.0065 0.415 2.408 Nr4a1 8.16 20.72 0.0067 0.397 2.519 Dlg2 6.99 2.68 0.0067 2.550 0.392 Dusp4 9.16 22.42 0.0072 0.411 2.431 Baalc 2.89 0.94 0.0073 2.877 0.348 Nr2f1 9.44 3.08 0.0075 3.000 0.333 Islr 5.72 1.87 0.0080 2.957 0.338 Hsd3b6 29.81 67.14 0.0087 0.445 2.248 Figf 6.52 2.08 0.0087 3.042 0.329 Eln 28.80 12.92 0.0093 2.219 0.451 Crabp1 12.24 3.97 0.0094 3.036 0.329 Rd3 1.50 0.46 0.0104 2.865 0.349 Lims2 1.43 5.66 0.0112 0.265 3.773 Meox1 23.20 9.76 0.0112 2.364 0.423 Aqp3 0.46 2.42 0.0113 0.223 4.476 Hemgn 1.52 4.52 0.0116 0.351 2.850 Gpr50 4.64 1.42 0.0120 3.109 0.322 Dnm1 7.66 3.13 0.0124 2.400 0.417 Snai3 0.51 2.63 0.0125 0.222 4.502 Eya4 2.50 0.88 0.0125 2.641 0.379 Shisa6 3.05 1.04 0.0125 2.758 0.363 Irx6 1.54 4.11 0.0129 0.390 2.567 Foxp2 6.99 2.97 0.0130 2.306 0.434 179  Gene symbol WT FPKM Sox9cKO FPKM p-value  Fold change down in cKO Fold change up in cKO Scube1 10.11 4.47 0.0133 2.233 0.448 Zfp503 4.16 1.47 0.0134 2.719 0.368 Thsd7b 4.73 1.98 0.0135 2.318 0.431 Epha3 12.45 5.51 0.0137 2.238 0.447 Chst3 1.68 0.61 0.0139 2.515 0.398 Rassf2 6.12 2.58 0.0142 2.320 0.431 Krt19 22.93 52.14 0.0144 0.441 2.268 Sox9 18.23 7.34 0.0144 2.464 0.406 Bmp10 29.69 65.56 0.0149 0.454 2.204 Guca2b 18.79 4.99 0.0151 3.715 0.269 Prss35 43.84 20.39 0.0154 2.145 0.466 Hba-a2 689.57 1585.85 0.0155 0.435 2.300 Prap1 46.11 14.51 0.0156 3.162 0.316 Per1 6.32 14.58 0.0159 0.438 2.285 Aplnr 14.16 6.22 0.0161 2.256 0.443 Prl8a2 0.44 6.03 0.0163 0.088 11.315 Lmod1 0.63 1.79 0.0183 0.387 2.583 Prrx1 9.20 3.94 0.0191 2.300 0.435 Rarg 7.64 3.18 0.0194 2.362 0.423 Nfatc4 13.78 6.36 0.0199 2.150 0.465  *≧1FPKM in the WT or Sox9 cKO library. ** Combined set of differentially expressed genesgenerated from duplicate libraries for each genotype as determined by the Cufflinks program. 180  APPENDIX XII Genes with altered expression (>1.5FC down) with a SOX9 peak in the AVC Gene symbol Full Gene Name WT FPKM cKO FPKM p-value  FC down  Btn1a1 butyrophilin, subfamily 1, member A1 6.92 0.29 0.0001 18.197 Prelp proline/arginine-rich end leucine-rich repeat protein 3.97 0.32 0.0002 9.701 Tnrc6b trinucleotide repeat containing 6B 66.09 13.91 0.0004 4.724 Ctnna2 catenin (cadherin-associated protein), alpha 2 3.30 0.56 0.0006 5.141 Col6a6 collagen, type VI, alpha 6 4.10 1.19 0.0010 3.246 Ntn4 netrin 4 9.84 3.24 0.0026 2.975 Pitx2 paired-like homeodomain 2 19.54 7.02 0.0028 2.758 Twist1 twist family bHLH transcription factor 1 37.16 13.38 0.0035 2.765 Fgfr2 fibroblast growth factor receptor 2 19.88 7.65 0.0035 2.577 Clmp/ 9030425E11Rik CXADR-like membrane protein 17.83 7.00 0.0058 2.527 Dlg2 discs, large homolog 2 (Drosophila) 6.99 2.68 0.0067 2.550 Nr2f1 nuclear receptor subfamily 2, group F, member 1 9.44 3.08 0.0075 3.000 Eln elastin 28.80 12.92 0.0093 2.219 Eya4 eyes absent homolog 4 (Drosophila) 2.50 0.88 0.0125 2.641 Foxp2 forkhead box P2 6.99 2.97 0.0130 2.306 Scube1 signal peptide, CUB domain, EGF-like 1 10.11 4.47 0.0133 2.233 Thsd7b thrombospondin, type I, domain containing 7B 4.73 1.98 0.0135 2.318 Chst3 carbohydrate (chondroitin 6) sulfotransferase 3 1.68 0.61 0.0139 2.515 Nfatc4 nuclear factor of activated T-cells, cytoplasmic, calcineurin-dependent 4 13.78 6.36 0.0199 2.150 Mecom MDS1 and EVI1 complex locus 21.24 10.72 0.0280 1.973 Osr1 odd-skipped related transciption factor 1 4.16 1.58 0.0311 2.532 Negr1 neuronal growth regulator 1 3.79 1.69 0.0316 2.172 Plcb1 phospholipase C, beta 1 (phosphoinositide-specific) 4.47 2.12 0.0329 2.055 Gria1 glutamate receptor, ionotropic, AMPA 1 2.32 1.03 0.0368 2.132 Ptx3 pentraxin 3, long 7.57 2.89 0.0386 2.568 Acvr1c activin A receptor, type IC 1.95 0.91 0.0462 2.023 Antxr1 anthrax toxin receptor 1 13.24 7.02 0.0496 1.873 Garnl3 GTPase activating Rap/RanGAP domain-like 3 7.35 3.67 0.0496 1.975 Msx1 msh homeobox 1 10.41 4.84 0.0559 2.130 Spint2 serine peptidase inhibitor, Kunitz type, 2 17.41 8.65 0.0561 2.002        181  Gene symbol Full Gene Name WT FPKM cKO FPKM p-value  FC down  Cpne5 copine V 14.08 7.43 0.0580 1.884 Ly86 lymphocyte antigen 86 3.50 1.33 0.0597 2.515 Postn periostin, osteoblast specific factor 402.15 200.48 0.0625 2.005 Bmf Bcl2 modifying factor 10.88 5.91 0.0650 1.829 Klhl29 kelch-like family member 29 1.41 0.67 0.0676 1.951 Tmem26 transmembrane protein 26 3.68 1.88 0.0685 1.914 Phactr1 phosphatase and actin regulator 1 2.80 1.40 0.0706 1.937 Plce1 phospholipase C, epsilon 1 10.50 6.00 0.0711 1.739 Nfix nuclear factor I/X (CCAAT-binding transcription factor) 15.16 7.16 0.0741 2.103 1700084E18Rik 1700084E18Rik 2.82 0.49 0.0827 4.916 Inhba inhibin, beta A 10.28 5.29 0.0842 1.925 Gfra1 GDNF family receptor alpha 1 4.79 2.57 0.0853 1.829 Grm7 glutamate receptor, metabotropic 7 1.39 0.60 0.0964 2.122 Mapk8ip1 mitogen-activated protein kinase 8 interacting protein 1 5.26 2.83 0.0976 1.825 Rasl11a RAS-like, family 11, member A 3.40 1.46 0.0977 2.243 Sox4 SRY (sex determining region Y)-box 4 74.78 44.61 0.0980 1.675 Hexim1 hexamethylene bis-acetamide inducible 1 13.00 7.62 0.1005 1.696 Olfm1 olfactomedin 1 6.68 3.55 0.1069 1.860 Ociad2 OCIA domain containing 2 10.40 5.83 0.1103 1.772 Odz4/Tenm4 teneurin transmembrane protein 4 6.78 4.07 0.1107 1.652 Slc24a3 solute carrier family 24 (sodium/potassium/calcium exchanger), member 3 5.33 2.97 0.1140 1.769 Sncaip synuclein, alpha interacting protein 10.65 6.39 0.1203 1.657 Axin2 axin 2 9.08 5.43 0.1238 1.660 Clstn2 calsyntenin 2 3.30 1.87 0.1241 1.731 Id4 inhibitor of DNA binding 4, dominant negative helix-loop-helix protein 4.68 2.43 0.1251 1.891 Id2 inhibitor of DNA binding 2, dominant negative helix-loop-helix protein 118.92 73.04 0.1254 1.627 Efna4 ephrin-A4 6.90 3.75 0.1307 1.820 Kcnab2 potassium voltage-gated channel, shaker-related subfamily, beta member 2 2.46 1.36 0.1327 1.754 Rbm34 RNA binding motif protein 34 4.68 2.76 0.1446 1.673 Asb13 ankyrin repeat and SOCS box containing 13 3.77 2.14 0.1489 1.727 Kitl KIT ligand 19.78 12.60 0.1526 1.565 Tpbg Trophoblast glycoprotein 5.53 3.33 0.1557 1.645 Odz3 Teneurin transmembrane protein 3 (Tenm3) 16.66 10.77 0.1571 1.543       182  Gene symbol Full Gene Name WT FPKM cKO FPKM p-value  FC down  3000002C10Rik 3000002C10Rik 3.01 1.58 0.1580 1.850 Clip3 CAP-GLY Domain containing linker protein 20.10 11.85 0.1612 1.690 Rftn2 raftlin family member 2 4.71 2.89 0.1699 1.606 Abcb1b ATP-binding cassette, sub-family B (MDR/TAP), member 1 1.62 0.86 0.1707 1.783 Mylip myosin regulatory light chain interacting protein 5.56 3.37 0.1713 1.633 Hand2 heart and neural crest derivatives expressed 2 46.27 30.34 0.1741 1.523 Epha7 EPH receptor A7 30.86 19.95 0.1752 1.544 Cbx3 chromobox homolog 3 2.06 1.09 0.1774 1.812 Nap1l3 nucleosome assembly protein 1-like 3 1.68 0.93 0.1807 1.716 Fbln2 fibulin 2 91.25 59.80 0.1835 1.525 Zdhhc14 zinc finger, DHHC-type containing 14 1.68 0.94 0.1855 1.718 Arrdc4 arrestin domain containing 4 10.13 6.35 0.1868 1.588 Rbfox1 RNA binding protein, fox-1 homolog (C. elegans) 1 1.47 0.84 0.1872 1.673 2610035D17Rik 2610035D17Rik 4.56 2.67 0.1885 1.682 Msrb2 methionine sulfoxide reductase B2 14.29 8.67 0.1936 1.640 Tusc1 tumor suppressor candidate 1 2.68 1.39 0.1947 1.863 Txnip thioredoxin interacting protein 115.03 75.79 0.2031 1.517 Angptl4 angiopoietin-like 4 7.53 4.70 0.2032 1.590 Tbcc tubulin folding cofactor C 8.76 5.54 0.2073 1.570 Rab7l1 RAB7, member RAS oncogene family-like 1 4.29 2.46 0.2082 1.711 Mecom MDS1 and EVI1 complex locus 2.85 1.52 0.2284 1.824 Foxp4 forkhead box P4 28.89 18.54 0.2343 1.555 Gpr173 G protein-coupled receptor 173 3.05 1.88 0.2373 1.590 Efcab1 EF-hand calcium binding domain 1 1.47 0.75 0.2393 1.843 Fam174a family with sequence similarity 174, member A 6.29 4.04 0.2444 1.545 Ehd3 EH-domain containing 3 3.70 2.43 0.2535 1.500 Prox2 prospero homeobox 2 1.35 0.80 0.2630 1.599 Mafb v-maf avian musculoaponeurotic fibrosarcoma oncogene homolog B 3.08 1.96 0.2637 1.540 Zfp185 zinc finger protein 185 (LIM domain) 2.64 1.71 0.2647 1.512 Xrcc4 X-ray repair complementing defective repair in Chinese hamster cells 4 10.83 7.17 0.2674 1.503 Acrbp acrosin binding protein 3.90 2.42 0.2740 1.589 Tiam2 T-cell lymphoma invasion and metastasis 2 4.43 2.74 0.2741 1.593 Nr4a2 nuclear receptor subfamily 4, group A, member 2 2.27 1.44 0.2759 1.535 B230217C12Rik B230217C12Rik 2.99 1.93 0.3526 1.527 183  Gene symbol Full Gene Name WT FPKM cKO FPKM p-value  FC down  Rbp7 retinol binding protein 7, cellular 2.01 1.12 0.3740 1.730 C230052I12Rik C230052I12Rik 5.38 3.35 0.4240 1.588 Mpv17l2 MPV17 mitochondrial membrane protein-like 2 12.63 7.90 0.4455 1.592 Psd pleckstrin and Sec7 domain containing 2.18 1.29 0.4741 1.635 Dnajb11 DnaJ (Hsp40) homolog, subfamily B, member 11 1.37 0.55 0.6987 2.254 Stat4 signal transducer and activator of transcription 4 1.16 0.62 1.0000 1.757 Htr1b 5-hydroxytryptamine (serotonin) receptor 1B, G protein-coupled 1.15 0.34 1.0000 2.861 Col4a4 collagen, type IV, alpha 4 1.14 0.55 1.0000 1.923 4930572J05Rik/Them6 thioesterase superfamily member 6 1.05 0.65 1.0000 1.533 Lrrn1 leucine rich repeat neuronal 1 1.05 0.21 1.0000 3.752 *FC = Fold change 184  APPENDIX XIII Genes with altered expression (>1.5FC up) with a SOX9 peak in the AVC Gene symbol Full Gene Name WT FPKM cKO FPKM p-value  FC up in cKO Bhlhe40 Basic Helix-Loop-Helix Family, Member E40 12.17 56.04 5.00E-05 4.575 Fos FBJ murine osteosarcoma viral oncogene homolog 2.45 13.97 0.0002 5.522 Dusp4 Dual specificity phosphatase 4 9.16 22.42 0.0072 2.431 Stc1 Stanniocalcin 1 4.32 9.66 0.0227 2.206 Ddit3 DNA-damage-inducible transcript 3 11.93 72.50 0.0461 6.036 Nrg1 Neuregulin 1 2.98 6.50 0.0466 2.140 Junb Jun B proto-oncogene 2.40 5.41 0.0510 2.201 Fhdc1 FH2 domain containing 1 1.06 2.05 0.0801 1.859 Gramd1b GRAM domain containing 1B 10.72 18.23 0.1035 1.694 Gm14005 Gm14005 predicted gene 14005 1.07 3.18 0.1101 2.803 1700026D08Rik 1700026D08Rik 0.92 2.12 0.1105 2.178 Fam46c Family with sequence similarity 46, member C 22.75 35.72 0.1494 1.568 Edn1 Endothelin 1 7.25 11.75 0.1629 1.613 Lama1 Laminin, alpha 1 1.79 2.87 0.1685 1.576 Snora28 Small nucleolar RNA, H/ACA box 28 0.00 78.38 0.1711 78.484 Nrn1 Neuritin 1 3.43 5.90 0.1774 1.698 Map3k13 Mitogen-activated protein kinase kinase kinase 13 0.78 1.32 0.1836 1.610 St8sia1 ST8 alpha-N-acetyl-neuraminide alpha-2,8-sialyltransferase 1 2.14 3.44 0.1989 1.581 Zfp97 Zinc finger protein 97 1.18 2.04 0.2002 1.673 Nampt Nicotinamide phosphoribosyltransferase 10.04 15.21 0.2161 1.509 Sgca Sarcoglycan, alpha (50kDa dystrophin-associated glycoprotein) 6.56 10.15 0.2277 1.539 Abra Actin-binding Rho activating protein 6.49 10.00 0.2441 1.533 Cox7c Cytochrome c oxidase subunit VIIc 25.46 39.16 0.2538 1.536 Snord49a Small nucleolar RNA, C/D box 49A 0.00 12066.80 0.2543 12066.900 Fah Fumarylacetoacetate hydrolase (fumarylacetoacetase) 2.07 3.34 0.2577 1.587 Dgat2 Diacylglycerol O-acyltransferase 2 3.12 4.80 0.2664 1.525 Dbp albumin D-box binding protein 2.63 4.13 0.6176 1.549 6430411K18Rik 6430411K18Rik 2.47 4.32 0.6789 1.722 Steap1 Six transmembrane epithelial antigen of the prostate 1 0.58 1.46 1.0000 2.294 1700092M07Rik 1700092M07Rik 0.58 1.11 1.0000 1.769 Hspa1a Heat shock 70kDa protein 1A 0.57 1.04 1.0000 1.706 Adamtsl3 ADAMTS-like 3 0.66 1.19 1.0000 1.694 *FC = Fold change  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0220645/manifest

Comment

Related Items