Determining Genetic Mechanisms of Vascular Stability: a Novel Role for Foxf2

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

Graduate Studies The Vault: Electronic Theses and Dissertations

2016 Determining Genetic Mechanisms of Vascular Stability: A Novel Role For FoxF2

Arnold, Corey

Arnold, C. (2016). Determining Genetic Mechanisms of Vascular Stability: A Novel Role For FoxF2 (Unpublished doctoral thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/25608 http://hdl.handle.net/11023/3223 doctoral thesis

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

Determining Genetic Mechanisms of Vascular Stability: A Novel Role For FoxF2

Corey R. Arnold

A THESIS

SUBMITTED TO THE FACULTY OF GRADUATE STUDIES

IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSPHY

GRADUATE PROGRAM IN BIOCHEMISTRY AND MOLECULAR BIOLOGY

CALGARY, ALBERTA

AUGUST, 2016

Endothelial cells of blood vessels interact with surrounding peri-endothelial cells (pericytes and vascular smooth muscle cells) to maintain integrity, modulate blood flow and promote homeostasis, collectively contributing to vascular stability. Both the initial establishment and subsequent maintenance of vascular stability is crucial to the integrity of the vasculature, and loss of this support can result in hemorrhage. Hemorrhagic events in brain vessels often lead to stroke, and accumulation of vascular insults originating from small brain vessels and capillaries, known as cerebral small vessel disease, can cause cognitive decline in addition to stroke. The identification of genetic factors promoting vascular stability is therefore fundamental to our understanding of cerebral vessel function in development and disease. Through microarray expression profiling of a genetic zebrafish model of cerebral hemorrhage, I found the transcription factor FoxF2 to be significantly downregulated. With further investigation, I discovered a novel role for FoxF2 in zebrafish cerebral vascular stability, both developmentally and in later stages of life. I show that transient knockdown and genetic knockout of the zebrafish FoxF2b paralog results in vascular stability defects during development. Furthermore, I provide evidence that FoxF2 is expressed in neural crest- and ventral mesoderm-derived head mesenchyme, and promotes differentiation of vascular smooth muscle and pericytes potentially through modulation of vascular stability-related signaling pathways. In the adult zebrafish brain, I find that FoxF2 is specifically expressed in pericytes and genetic knockout results in brain hemorrhages. These findings parallel recent observations of human FoxF2 mutations, which exhibit hallmarks of cerebral small vessel disease. Genetic variants in a potential regulatory region of human FoxF2 have been also recently been linked to stroke, further suggesting a role for FoxF2 as a human disease gene. Through this work, I have identified a potential mechanism by which FoxF2 promotes cerebral vascular stability and paved the way for future investigations of cerebral vascular development and disease.

ii Acknowledgements

I would like to acknowledge the members of my lab, both past and present, for their unending support and invaluable advice throughout my degree. In particular, I thank Nicole Munsie for her help with in situ hybridization experiments, Tom Whitesell for sharing his RNAseq data, and Jae-Ryeon Ryu for her help in Western blotting, generation of the TALEN mutants, and for her general expertise in the lab, without which this project would have gone much more slowly. I’d like to thank Dr. Ryan Lamont for initiating the microarray work that led to my project, and Dr. Chi-Yip Ho and Dr. Kin Chan for their expertise in bioinformatics. I thank the members of the BMB department and the ACHRI institute that have made this opportunity possible through financial, material and intellectual support, and provided an open and nurturing environment in which to pursue science. I’d like to thank my committee members, Jim and Savraj, for their advice and insights into my project throughout the years. I’d like to thank Dr. Ordan Lehmann, Dr. Curtis French, Dr. Ganesh Chauhan and Dr. Stephanie Debette for the opportunity to collaborate and expand our research into human studies. A big thank you to all the friends I have made in Calgary, both in and out of the lab, for making the Foothills campus and city of Calgary feel like a new home. I would also like to thank my family for their unquestioning support of my decision to pursue my interests, and a neverending thank you to my fiancée Heather who has been a source of constant inspiration and unwavering strength, especially whenever the going got tough. Finally, I’d like to acknowledge all the hard work, dedication, enthusiasm and relentless positivity that Dr. Sarah Childs has poured into not only my project, but all her students’ endeavours. None of this would have been possible without her obvious and infectious passion for scientific exploration.

iii Table of Contents

Abstract ...... ii Acknowledgements ...... iii Table of Contents ...... iv List of Tables ...... viii List of Figures and Illustrations ...... ix List of Symbols, Abbreviations and Nomenclature ...... xii

CHAPTER ONE: INTRODUCTION ...... 1 1.1 The Circulatory System ...... 2 1.1.1 Blood Vessel Structure ...... 2 1.1.2 Zebrafish Vascular Development ...... 4 1.1.3 Developmental Anatomy of the Zebrafish Cerebral Vasculature ...... 6 1.2 Mural Cells ...... 8 1.2.1 Vascular Smooth Muscle Cells ...... 8 1.2.2 Pericytes ...... 9 1.2.3 Mural Cell Origins ...... 10 1.2.4 Neural Crest ...... 11 1.2.5 Cranial Mesoderm ...... 11 1.3 Signaling Pathways in Development of Vascular Stability ...... 12 1.3.1 PDGF-B/PDGFRβ ...... 12 1.3.2 TGF-β ...... 14 1.3.3 Notch ...... 15 1.3.4 Other Vascular Stability Signaling Pathways ...... 16 1.4 Structural and Physical Interactions of Vascular Stability ...... 16 1.4.1 Basement Membrane ...... 17 1.4.2 Integrins ...... 17 1.4.3 Junctional Complexes ...... 18 1.5 Vascular Stability Disease ...... 19 1.5.1 Hemorrhagic Stroke ...... 20 1.5.2 Ischemic Stroke ...... 20 1.5.3 Cerebral Small Vessel Disease ...... 21 1.6 Iguana/Dzip1 mutants as a model of vascular stability defects ...... 23 1.6.1 Shh Signaling and Vascular Stability ...... 23 1.6.2 Primary Cilia are required for Shh Signaling ...... 24 1.7 Fox Transcription Factors ...... 26 1.7.1 Forkhead Winged Helix Motif Structure ...... 27 1.7.2 FoxF2 ...... 27 1.7.3 FoxF2 mediates mesenchyme in development ...... 29 1.7.4 FoxF2 is also involved in formation of the palate...... 29 1.7.5 FoxF2 inhibits EMT to repress cancer metastasis ...... 31 1.8 Unanswered questions in vascular stabilization ...... 31

CHAPTER TWO: MATERIALS AND METHODS ...... 34 2.1 Zebrafish Maintenance and Husbandry ...... 35 2.2 Morpholino Injections ...... 35

iv 2.3 Fin Clipping and HotSHOT Genomic DNA Isolation...... 37 2.4 TALEN Mutagenesis and Mutation Identification ...... 37 2.5 FoxF2b Mutant Sequencing and Genotyping ...... 38 2.6 Template Generation and In Situ Hybridization Probe Synthesis ...... 40 2.7 Wholemount In Situ Hybridization ...... 42 2.8 Immunostaining ...... 43 2.9 Staining, Embedding and Imaging for Transmission Electron Microscopy ...... 45 2.10 Confocal Microscopy ...... 45 2.11 DAPI-Dextran Injections ...... 46 2.12 Zebrafish Brain Dissections ...... 46 2.13 Brain Tissue Clearing ...... 47 2.14 Gateway BP and LR Cloning of full length FoxF2 ...... 47 2.15 Cell Culture and Transfection ...... 48 2.16 Western Blotting ...... 49 2.17 Drug Treatments ...... 49 2.18 Expression Profiling ...... 50 2.19 RNAseq expression profiling ...... 51

CHAPTER THREE: COMPARATIVE ANALYSIS OF GENES REGULATED BY DZIP- IGUANA AND HEDGEHOG IN ZEBRAFISH ...... 52 3.1 Introduction ...... 54 3.2 Results ...... 55 3.2.1 Gene Expression Profile of Iguana Mutants Partially Overlaps With Hh Deficient Embryos ...... 55 3.2.2 A Subset of Developmentally Expressed Genes is Regulated by Both Hh Signaling and Iguana...... 55 3.2.3 Independent Validation of Microarray Gene Expression Changes by In Situ Hybridization ...... 58 3.2.4 Microarray Validation by RNA Deep-Sequencing ...... 65 3.2.5 igu Mutants Have a Unique Gene Expression Signature ...... 67 3.2.6 Genes Uniquely Regulated by Dzip1 ...... 67 3.3 Discussion ...... 70 3.3.1 Potential vascular stability genes identified by the microarray ...... 70 3.3.2 Differentially regulated genes specific to the igu microarray...... 71 3.3.3 Additional Microarray Validation by RNAseq ...... 72 3.3.4 Conclusions ...... 72

CHAPTER FOUR: FOXF2B IS REQUIRED FOR EMBRYONIC VASCULAR MURAL CELL DEVELOPMENT ...... 73 4.1 Introduction ...... 75 4.2 Results ...... 77 4.2.1 foxf2 is expressed in head mesenchymal tissues adjacent to endothelial cells and distinct from smooth muscle cells...... 77 4.2.2 Transient knockdown of foxf2b results in embryonic cerebral hemorrhage and reduced mural cell coverage of vessels ...... 82 4.2.3 Generation of TALEN-mediated FoxF2b knockout mutant zebrafish ...... 87

v 4.2.4 Differential embryonic hemorrhage phenotypes between foxf2bCa22 and Ca23 mutant alleles ...... 91 4.2.5 FoxF2b is necessary for initial mural cell coverage of penetrating cerebral vessels during development ...... 93 4.2.6 Cerebral vessel permeability is unaffected in foxf2b mutants ...... 95 4.2.7 foxf2a expression pattern is expanded in foxf2b mutants ...... 95 4.2.8 Mural cell marker pdgfrβ is specifically reduced at 4dpf in foxf2b mutants ..99 4.2.9 The related foxc1 gene also promotes vessel coverage and mature muscle cell markers ...... 101 4.2.10 foxq1a/b expression is decreased in foxf2b mutants ...... 101 4.2.11 Examination of vascular mural cell markers and potential target gene expression in foxf2b mutants and morphants ...... 104 4.2.11.1 nkx3.2 ...... 104 4.2.11.2 Connective Tissues - col4a, ctgfa ...... 106 4.2.11.3 Neural crest and EMT – snai2, hand2 ...... 108 4.2.11.4 PDGF signaling – pdgfrα, pdgfaa, pdgfba ...... 108 4.2.11.5 Head mesenchyme and palate markers – tbx15, tbx18, hic1 ...... 111 4.2.11.6 Shh signaling feedback – gli2a, ptch2 ...... 115 4.3 Discussion ...... 120 4.3.1 foxf2a/b expression patterns reveal a potentially mesodermal source of head mural cells ...... 120 4.3.2 Knockdown and genetic knockout of foxf2b results in mural-cell associated vascular stability defects ...... 121 4.3.3 FoxF2b regulates markers of vSMC and pericyte differentiation...... 124 4.3.4 Vessel ultrastructure reveals a role for FoxF2b in promoting endothelial-mural cell interactions ...... 125 4.3.5 foxc1 knockdown produces vascular stability defects ...... 126 4.3.6 Knockdown and genetic knockout of foxf2b affects expression of genes involved in mural cell stabilization of endothelial cells ...... 126 4.3.7 Conclusions and Future Directions ...... 128

CHAPTER FIVE: FOXF2B MAINTAINS CEREBRAL VASCULAR INTEGRITY IN DEVELOPED ZEBRAFISH BRAINS THROUGH PERICYTES ...... 130 5.1 Introduction ...... 132 5.2 Results ...... 133 5.2.1 foxf2 is expressed in pericytes of the zebrafish brain ...... 133 5.2.2 foxf2b mutant brains exhibit hemorrhages more commonly than wildtype ..133 5.2.3 foxf2b mutant small cerebral vessels trend towards smaller diameter ...... 136 5.2.4 acta2 coverage of foxf2b mutant brain vessels is unchanged ...... 136 5.2.5 Human FoxF2 mutations are commonly associated with brain defects ...... 138 5.3 Discussion ...... 142 5.3.1 foxf2 expression in pericytes suggests a role in vascular stability ...... 142 5.3.2 foxf2b mutant brains hemorrhage in juvenile stages ...... 142 5.3.3 Vessel morphology and mural cell phenotype are not significantly altered in foxf2b mutant brains ...... 143 5.3.4 Human mutations show relevance for the investigation of zebrafish FoxF2 function in brain ...... 144 vi 5.3.5 Conclusions and Future Directions ...... 145

CHAPTER SIX: DISCUSSION ...... 146 6.1 FoxF2 and vascular stability ...... 147 6.1.1 FoxF2 in early mesenchyme ...... 147 6.1.2 FoxF2b potentially promotes differentiation of mural cells through inhibition of proliferative signals ...... 148 6.1.3 FoxF2b is required for cerebral vascular stability throughout life ...... 149 6.2 Allele-specific and paralog-specific phenotype differences ...... 150 6.2.1 FoxF2b mutant alleles exhibit slightly differing phenotypes ...... 150 6.2.2 FoxF2a genetic compensation ...... 151 6.3 Roles for other Fox genes in vascular stabilization ...... 153 6.3.1 FoxC1 ...... 153 6.3.2 FoxQ1 ...... 153 6.3.3 FoxF1 ...... 154 6.4 Comparison between mouse and fish FoxF2 knockouts ...... 155 6.5 Comparison of foxf2 mutants with other zebrafish genetic models of vascular stability ...... 156 6.6 FoxF2 in Human Stroke ...... 158 6.7 Future Directions ...... 161

CHAPTER SEVEN: REFERENCES ...... 164

CHAPTER EIGHT: APPENDIX ...... 188

vii List of Tables

Table 2.1 - Injection needle pulling programs for Flaminng/Brown Micropipette Puller ...... 36

Table 2.2 - Morpholino Oligonucleotide Sequences, Targets and Doses ...... 36

Table 2.3 - TALEN Target Sequences ...... 38

Table 2.4– TALEN-generated Mutant Genotyping Assays and Primers ...... 39

Table 2.5- Primers for ISH probes generated by PCR ...... 41

Table 2.6 - Probes generated from cloned sequences in vectors ...... 42

Table 2.7 - Proteinase K Incubation Concentrations and Times for ISH Permeabilization ...... 43

Table 2.8- List of antibodies used ...... 44

Table 2.9 – Fluorophore Excitation and Emission Wavelengths ...... 46

Table 3.1- 40 genes with significantly altered gene expression patterns in igu mutant and cyclopamine-treated zebrafish embryos as reported by microarray ...... 57

Table 3.2- Number of putative consensus and non-consensus Gli binding sites for all annotated double hits from igu and cyclopamine microarrays ...... 59

Table 3.3 – Selected set of differentially regulated genes for further verification, and fold changes reported by microarray ...... 60

Table 3.4– Illumina next-generation RNA sequencing hits of 12 genes with altered expression levels in igu mutants below 7.5% false discovery rate cutoff, as well as four genes examined by in situ hybridization from the array that did not pass the 7.5% false discovery rate ...... 66

Table 3.5 – Genes with altered expression in igu mutant zebrafish embryos only, as reported by microarray ...... 68

Table 4.1 – Summary of ISH experiments in foxf2b morphants and mutants ...... 119

Table 5.1 – Human FoxF2 mutation-related phenotypes as annotated in the DECIPHER database (v9.9) ...... 141

Table 6.1 – Top genes downregulated in foxf2bCa23/Ca23 mutant embryos ...... 159

viii List of Figures and Illustrations

Figure 1.1 – Vessel structure of capillaries, arteries and veins ...... 3

Figure 1.2 – Development of the zebrafish vasculature ...... 5

Figure 1.3 – Anatomy of the Developing Zebrafish Cerebral Vessels ...... 7

Figure 1.4 – Endothelial-Mural Cell Interactions ...... 13

Figure 1.5 – Cerebral Small Vessel Disease ...... 22

Figure 1.6 – Primary cilia formation and Hh signaling are dependent on dzip1 ...... 25

Figure 1.7 – Structure of FoxF2 and the winged-helix DNA-binding domain ...... 28

Figure 1.8 – Phylogenetic tree of Fox genes...... 30

Figure 3.1– Summary of genes with significantly altered expression in igu mutants and cyclopamine-treated embryos ...... 56

Figure 3.2 – In situ hybridization pattern changes of genes misregulated in igu mutants corresponds to direction of expression change on the microarray ...... 61

Figure 3.3 – In situ hybridization pattern changes of genes misregulated in cyclopamine- treated embryos corresponds to direction of expression change on the microarray ...... 62

Figure 4.1 – foxf2b is expressed in neural crest, ventral mesoderm and their head mesenchymal derivatives during early development ...... 78

Figure 4.2 – foxf2b expression surrounds endothelium at 48hpf ...... 79

Figure 4.3 - foxf2b does not co-localize with acta2 at 4dpf ...... 80

Figure 4.4 – foxf2a is expressed in head mesenchymal derivatives during early development ... 81

Figure 4.5 – FoxF antibodies specifically bind FoxF2 outside of zebrafish embryos ...... 83

Figure 4.6 – Morpholino knockdown of foxf2b results in embryonic hemorrhage ...... 85

Figure 4.7 – foxf2a morpholino does not result in embryonic hemorrhage or efficient mis- splicing of mRNA ...... 86

Figure 4.8 – foxf2b morphants have decreased acta2:GFP+ coverage of pharyngeal vessels ...... 88

Figure 4.9 – Decreased mural cell contacts with abluminal surface of head vessels in foxf2b morphant embryos ...... 89

ix Figure 4.10 – Sequences of TALEN-generated foxf2b mutations, genotyping and predicted resultant protein product ...... 90

Figure 4.11 – Allele-specific differences in hemorrhage in foxf2b mutant embryos ...... 92

Figure 4.12 – Initial decrease of cerebral vessel coverage by smooth muscle in foxf2bCa23/Ca23 mutants recovers by 1 week of age...... 94

Figure 4.13 – Cerebral vessel coverage by smooth muscle is unchanged in foxf2bCa22/Ca22 mutants at 102hpf...... 96

Figure 4.14 – Brain vessels do not show increased permeability to high or low molecular weight dyes in foxf2b mutants ...... 97

Figure 4.15 – foxf2a expression domain is expanded in foxf2bCa22/Ca22 mutant embryos ...... 98

Figure 4.16 – pdgfrβ expression is decreased at 4dpf in foxf2b mutants, other pericyte markers are unchanged...... 100

Figure 4.17 – Morpholino knockdown of foxc1 results in decreased acta2:GFP expression and poor vessel support by perivascular mural cells...... 102

Figure 4.18 – Expression of foxq1a/b is absent in foxf2b morphants ...... 103

Figure 4.19 – nkx3.2 expression is unchanged in foxf2b mutants and morphants ...... 105

Figure 4.20 – Expression of col4a2 is decreased in foxf2b mutants, but ECM-related growth factor ctgfa is unchanged ...... 107

Figure 4.21 – foxf2b morphant expression of neural crest gene snai2 is unchanged, but hand2 expression indicates loss of pharyngeal arch structures ...... 109

Figure 4.22 – Expression of PDGF signaling components in foxf2b morphant and mutant embryos ...... 110

Figure 4.23 – Allele-specific differences of tbx15 expression in foxf2b mutants ...... 112

Figure 4.24 – Allele-specific differences of tbx18 expression in foxf2b mutants ...... 113

Figure 4.25 – Allele-specific differences of hic1 mesenchyme expression patterns in foxf2b mutants ...... 116

Figure 4.26 – gli2a and ptch2 expression is increased in foxf2 mutant embryos ...... 117

Figure 5.1 – foxf2a and foxf2b are expressed in brain tissue in patterns reminiscent of pericyte marker expression ...... 134

Figure 5.2 – foxf2bCa22/Ca22 mutant brains exhibit hemorrhage at juvenile stages ...... 135

x Figure 5.3 – Small vessels of the brain show a non-significant trend towards decreased size, and constricted lumens relative to vessel wall size in foxf2bCa22/Ca22 mutants ...... 137

Figure 5.4 – acta2:GFP+ coverage of cerebral vessels is unchanged between wildtype and foxf2bCa23/Ca23 mutant brains ...... 139

Figure 5.5 – Human FoxF2 mutations are commonly associated with intellectual disability .... 140

Figure 6.1 – Phylogenetic analysis of Fox proteins between zebrafish, mouse and human ...... 152

Figure 6.2 – Human FoxF2 segmental mutations and stroke-associated SNPs ...... 160

Figure 6.3 – Human and mouse FoxF2 mutants exhibit cerebral insults...... 162

xi List of Symbols, Abbreviations and Nomenclature

Symbol Definition

Hpf Hours post fertilization

Dpf Days post fertilization

WT Wild type

UIC Uninjected control

MO Morpholino oligonucleotide (injected) vSMC Vascular smooth muscle cell

Shh Sonic hedgehog

Igu iguana mutant

ECM Extra-cellular matrix

EMT Epithelial-to-mesenchymal transition

PDGF Platelet-derived growth factor

PDFGFR Platelet-derived growth factor receptor

Acta2 Smooth muscle actin

TEM Transmission electron microscopy

xii

Chapter One: Introduction

1.1 The Circulatory System

The circulatory system is a fascinating hierarchal branching complex of vessels, acting as the delivery and drainage service for all tissues of the body. Arteries carry oxygen from the lungs to the most peripheral of tissues, and veins return deoxygenated blood cells to perform their function anew, as well as carbon dioxide to expel from the body. Nutrients are transferred from the intestine to the bloodstream for efficient distribution throughout the body, and waste is filtered through vascular interactions with the liver and kidneys. Hormones, immune cells and signaling molecules released from localized sources are carried to their target tissues by the vasculature. Systemic blood flow is largely controlled by heart rate, and vasoconstriction/ vasodilation by vascular smooth muscle cells in large vessels. Blood flow can also be locally modulated by smooth muscle and pericytes in smaller vessels. Transport of nutrients, molecules and cells from the vascular lumen to the surrounding tissues occurs through passive means (diffusion) or active transport (endo- and exocytosis, cell-cell contact reorganization), and occurs mostly in the microcirculation where flow rates are low.

1.1.1 Blood Vessel Structure

The typical blood vessel consists of an inner, one-cell thick layer of endothelial cells with an inner lumen at their apical surfaces, and basement membrane and vascular mural cells (vascular smooth muscle cells (vSMCs) and pericytes) on their basal surfaces (Fig. 1.1). Large vessel lumens are lined by multiple endothelial cells interacting through junctional complexes to form a larger vessel. Capillaries can be lined by multiple or single endothelial cells that form auto-cellular junctions to form smaller vessels (Blum et al., 2008; Lenard et al., 2015). The low pressure and flow rate in veins induces valve formation in venous endothelial cells to prevent backflow. Mural cells physically contact the endothelial lining to provide elastic resistance against fluid pressures, modulate blood flow through contractile function and provide signaling cues to mediate differentiation of endothelial cells. Collectively, endothelial-mural cell contact maintains the structural integrity of the vessel wall to prevent hemorrhage while maintaining selective permeability to allow bloodstream constituents access to the surrounding tissue. Organs such as 2

Figure 1.1 – Vessel structure of capillaries, arteries and veins

Capillaries are smaller vessels that interact with pericytes embedded in the basement membrane. Arteriole and venule diagrams show artery and vein structure. Arteries are continuously surrounded by vSMCs, venous vSMC coverage is more discontinuous. Adapted from (Jain, 2003), with permission of the publisher.

the brain and central nervous system are particularly sensitive to vascular insults, and require especially strict regulation of transport across the vessel wall. The type and size of vessel determine how much support and what type of mural cell they interact with. The largest arteries of the system, such as the aorta, are subject to high blood flow and pressure, and therefore have vessel walls consisting of several layers of smooth muscle and a thicker basement membrane. Veins do not encounter such large physical forces, and even the largest diameter veins only have, at maximum, a couple layers of vSMCs. Small capillaries are discontinuously covered by pericytes, embedded within the basement membrane. Pericytes extended processes along and around vessels to provide biochemical and contractile support.

1.1.2 Zebrafish Vascular Development

Vascular endothelial cells develop from mesoderm in all species. In the zebrafish, endothelial cell precursors (angioblasts) reside in the lateral mesoderm of the embryo proper, and migrate to the midline to coalesce and form the primary axial vessels in a process called vasculogenesis. Hedgehog (Hh) signalling from the notochord induces two waves of migration of angioblasts to the midline where they congregate and differentiate to form the axial vessels, the dorsal aorta (DA) and the posterior cardinal vein (PCV) (Fouquet et al., 1997; Sumoy et al., 1997; Kohli et al., 2013) (Fig. 1.2 A,B). Differentiation of the arterial and venous phenotypes of these two vessels is mediated largely by Hh induction of vascular endothelial growth factor a (VEGF-A) and subsequent activation of Notch signaling to promote arterial identity and block venous identity (Fig. 1.2 C) (Lawson et al., 2001; Lawson et al., 2002; Lamont & Childs, 2006). Through this process, an initial vascular circuit is formed, running from the heart, down the length of the trunk via the DA and back again via the PCV, through which circulation begins at about 24hpf. Angiogenesis, the sprouting of new vessels from existing ones, begins after the axial vessels have formed, giving rise to the intersegmental vessels of the trunk and vasculature of the head. Angiogenic sprouting is largely mediated by VEGF and Notch signaling. VEGF is a potent inducer for vascular sprouting during development, wound healing and tumour growth (Matsumoto & Ema, 2014) (Fig. 1.2 D). VEGF-A signaling through the VEGF-receptor 2 4

Figure 1.2 – Development of the zebrafish vasculature

A: Hemangioblasts undergo specification into angioblasts, expressing VEGFR2. Shh prompts VEGF secretion to induce migration to the midline. B: Angioblast coalesce at the midline ventral to the notochord. C: A gradient of secreted VEGF induces specification of artery and vein, and Notch inhibits artery identity in venous cells. D: VEGF acts as a chemoattractant during angiogenesis, signaling to tip cells which in turn inhibit tip cell identity in following stalk cells.

(VEGFR2, flk1 or kdr in zebrafish) to activate Notch expression, specifying the leading “tip” cell. Notch signals to the following “stalk” cells to suppress tip cell identity (Blanco & Gerhardt, 2013). The final maturation of vessels requires recruitment and interaction with perivascular mural cells. Physical interactions and reciprocal signaling create and maintain vascular stability (Gaengel et al., 2009).

1.1.3 Developmental Anatomy of the Zebrafish Cerebral Vasculature

Mapping of the developing zebrafish cerebral vasculature during development has been done using angiography and transgenic markers (Isogai et al., 2001; Ulrich et al., 2011) (Fig. 1.3). The zebrafish cerebral vasculature arises from anterior lateral mesoderm. Unlike trunk vasculature, angioblasts in these tissues do not migrate to the midline but undergo vasculogenesis to form two lateral dorsal aortae (LDA). The first cerebral vessels to form are actually the major extracerebral vessels, which are established by 36hpf. From these, new vessels sprout and penetrate the brain tissue, starting at around 48hpf. The basal artery (BA) forms beneath the hindbrain from medial migration of the primordial hindbrain channel (PHBC). Sprouts from both PHBCs and the BA migrate dorsally then meet at the medial plane to form the central arteries (CtAs). The anterior portion of the BA forks into bilateral posterior communicating segments (PCA), which reconnect through the basal communicating artery (BCA), forming the “Circle of Willis” in the ventral portion of the midbrain. From this loop, the main penetrating arteries of the midbrain and forebrain sprout. The anterior sprouts form the anterior mesencephalic central arteries (AMCtAs) and extend into the forebrain. The midbrain arteries, called the middle mesencephalic central arteries (MMCtAs) migrate dorsally and anteriorly, and branches into the arterioles and capillaries that will populate the midbrain. This artery and its branches are easily detected in a lateral view using fluorescent labelling of vascular tissues. By about 4dpf, the major vessels of the brain are established, and a series of sprouting, bifurcation and pruning events eventually give rise to a complex network of arterioles, capillaries and venules of the cerebral vasculature. Blood flow enters the brain primarily through the basal artery, travels anteriorly and dorsally, and exits the brain through veins on the dorsal and lateral periphery.

Figure 1.3 – Anatomy of the Developing Zebrafish Cerebral Vessels

A: Schematics based on microangiography of zebrafish vessels. Images at showing stages from 48- 84 hpf. 48 hpf is dorsolaterally positioned while 60 and 84hpf are dorsal images. 84hpf diagrams are split into more ventral and more dorsal regions. Adapted from (Isogai et al., 2001) with permission from the publisher. B: Dorsolateral image of acta2:GFP transgenic zebrafish embryo fluorescently labelled for mural cells (green) at 102hpf for context of angiography diagrams and showing mural cell coverage of basal and perforating arteries. Scale bar = 40µm. AMCtA – anterior mesencephalic central artery, BA – basilar artery, BCA – basal communication artery, CtA – central artery, LDA – lateral dorsal aorta, MMCtA – middle mesencephalic central artery, PHCB – primordial hindbrain channel, PICA – primitive internal carotid artery.

1.2 Mural Cells

Mural cells are peri-endothelial cells that interact with the abluminal surfaces of endothelial tubes to provide physical support and contractile function, deposit basement membrane proteins, and help regulate bloodstream constituents crossing to the tissue parenchyma. There are typically two types of mural cells: vascular smooth muscle cells and pericytes.

1.2.1 Vascular Smooth Muscle Cells

Vascular smooth muscle cells (vSMCs) are a subtype of smooth muscle that surround larger endothelial vessels to support against fluid pressures and regulate blood flow. Since the largest arteries are exposed to higher pressures, and thus more physical stress, they are typically ensheathed by multiple layers of vSMCs. Conversely, veins and smaller arteries require less physical support and generally have a single, sometimes discontinuous, layer of smooth muscle cells (Santoro et al., 2009). Where other muscle types like cardiac or skeletal muscle reach a state of terminal differentiation, vSMCs retain phenotypic plasticity in order to switch between a contractile and synthetic state (reviewed in (Zhang et al., 2016b)). In the contractile state, vSMCs are elongated, wrapping around vessels in a continuous sheath, and undergo little to no proliferation or migration. Contractile genes such as acta2, calmodulin, smooth muscle myosin heavy chain and SM22 are upregulated and serve as markers for “mature” vSMCs. Organelles are typically shunted towards the nucleus or cell periphery, and the majority of the cell cytoplasm is populated by contractile filaments. Conversely, synthetic vSMCs have a more fibroblast-like morphology, show increased proliferation and migration, and synthetic organelles such as the Golgi and endoplasmic reticulum become more prominent for the synthesis of extracellular matrix (ECM) proteins. Phenotypic switching has become a subject of immense study, given that a reversion to synthetic states is commonly seen in major vascular diseases like atherosclerosis, hypertension and restenosis (Gomez & Owens, 2012). Establishing markers for vSMCs is typically difficult because, although several cytoskeletal contractile proteins are found in vSMCs, they are often also expressed in related tissue types. Two commonly used markers for vSMC development 8

include smooth muscle actin alpha 2 (acta2/α-sma) and transgelin (tgln/sm22) (Li et al., 1996; Mack & Owens, 1999; Santoro et al., 2009; Whitesell et al., 2014). Acta2 is one of the earliest markers and most abundant filaments of specified vSMCs, but also exhibits some expression in non-vSMC populations such as myofibroblasts.

1.2.2 Pericytes

The pericyte is a monocytic cell type that associates with small vessels and capillaries. They were first described by French scientist Charles Rouget over 140 years ago, and were originally named Rouget cells (Rouget, 1873). The name was changed to pericytes not long after based on their biological location next to endothelial cells (Zimmerman, 1923). These researchers defined pericytes as singular cells located abluminally to small vessels and capillaries that are embedded within the basement membrane and extend processes along the vessels. Typical markers of pericytes include Ng2 and platelet-derived growth factor receptor beta (PDGFRβ), although these markers are not restricted solely to pericytes. With the discovery of more and more mural cell markers and the advancement of microscopy and imaging, the definition of what is a pericyte has become somewhat unclear in the last few years. Although many researchers adhere to this original definition, subpopulations of pericytes have shown morphologies and expression of markers typically associated with vSMCs. Without careful analysis of vessel ultrastructure, it is sometimes difficult to discern between mural cell types. Indeed, it seems that the stereotypical vSMC and pericyte represent extremes along a continuum of mural cell phenotypes as opposed to mutually exclusive, discrete cell types. However, there is some indication that the relative position of a pericyte (or pericyte-like cell) on a capillary can provide clues as to its phenotype and function. Some pericytes closer to the arteriole end of capillaries express markers like acta2, desmin, myosin and vimentin (Bandopadhyay et al., 2001), and will extend processes around vessels, suggesting that these pericytes may have some level of contractile function (Peppiatt et al., 2006; Hall et al., 2014). Pericytes on the middle of capillaries, however, may not express these markers and extend more processes along vessels rather than around them. In accordance with this range of phenotypes, pericytes have been suggested to play roles in capillary blood flow modulation, blood-brain barrier regulation, vessel

maturation, tissue regeneration and ECM synthesis (reviewed in (David E, 1986) (Armulik et al., 2010).

1.2.3 Mural Cell Origins

The origins and developmental programs of perivascular mural cells, particularly of the brain and head, have started to become clearer in the last few years. The development of mural cell markers and transgenic marker lines has provided useful tools to study perivascular mural cell morphology, pattern and timing of differentiation. However, determining their developmental origins and molecular programs has required more extensive research. vSMC origins are more diverse than one might at first think. As opposed to a single precursor tissue, vSMCs differentiate from several different sources depending on where along the vasculature they eventually end up (Majesky, 2007). The splanchnic mesoderm, sclerotome, somites, serosal mesothelium, proepicardium and secondary heart field all give rise to vSMCs which populate distinct domains and sets of vessels within the trunk (Hood & Rosenquist, 1992; Mikawa & Gourdie, 1996; Waldo et al., 2005; Wilm et al., 2005; Pouget et al., 2006; Wiegreffe et al., 2009). vSMCs and pericytes of the head arise in large part from neural crest. Chick-quail transplantation studies demonstrated that the neural crest gives rise to pericytes and vSMCs of the ventral and anterior structures (forebrain and face, including pharyngeal arches) (Etchevers et al., 2001), and that the neuroectoderm was able to differentiate in vSMCs on cerebral vessels. (Korn et al., 2002). Mammalian lineage tracing experiments identified a neural crest origin for retinal pericytes (Gage et al., 2005; Trost et al., 2013), as well as for mural cells in the brain (Heglind et al., 2005; Foster et al., 2008; Simon et al., 2012). Similar lineage tracing in zebrafish determined that neural crest cells give rise to the vSMCS of ventral head vessels (Cavanaugh et al., 2015), but origins of brain and eye mural cells have not been thoroughly examined. A recently developed PDGFRβ transgenic line shows fluorescent marker expression as early as 8 somites in the developing neural crest tissues, and shows late stage expression in perivascular mural cells of the head (Ando et al., 2016). However, there is also evidence for mesodermal origins of head pericytes and vSMCs (Etchevers et al., 2001; Tidhar et al., 2001). Furthermore, 10

compared to endothelial cells, the specific migratory and morphogenic processes through which head mural cells arise from their tissues of origin is poorly understood, hindering our understanding of mural cell development. Although it is generally accepted that mural cells are differentiated from peri-endothelial mesenchyme, where this mesenchyme comes from or how it is specified is not known.

1.2.4 Neural Crest

The neural crest arises from a population of ectodermal cells located on the border of the neural plate. Signals from the adjacent neural plate and non-neural ectoderm or underlying paraxial mesoderm, induce neural crest specification (Rogers et al., 2012). As the neural plate invaginates, the neural crest cells, located at the tip of the neural folds, follow suit and end up in the dorsal region of the neural tube after closure. The neural crest then delaminates and migrates ventrally through epithelial-to-mesenchymal transition (EMT), characterized by a downregulation of cell-cell adhesion molecules, cytoskeletal rearrangements and formation of filopodia. EMT is induced by several genes, particularly foxd3, snai1, sox9, sox10, and in the cranial neural crest, ets2 (Theveneau & Mayor, 2012 2007). The neural crest is segmented along the anterior-posterior axis, and these segments migrate in separate streams to give rise to different tissues and structures. In the head, the cranial neural crest (which is further divided into anterior head, posterior head and rhombomere segments) gives rise to structural tissues, such as bone, cartilage, connective tissue and neurons. Once neural crest has migrated to the ventral regions, it first forms the pharyngeal arches, from which bones and cartilage of the jaw and pharynx arise. These arches also contain tissues derived from mesoderm, part of which will eventually form the arch arteries. As mentioned above, the neural crest is also shown to give rise to vascular mural cells of the head.

1.2.5 Cranial Mesoderm

The mesoderm is one of the three germ layers, and gives rise to bone, muscle, mesenchyme and cardiovascular tissues. In zebrafish, the mesendoderm involutes during gastrulation, and segregates into distinct mesodermal and endodermal layers between the 11

ectoderm and yolk. Prior to neural tube formation, the mesoderm forms paraxial, intermediate and lateral plate populations, and the lateral plate further divides to form dorsal (somatic) and ventral (splanchnic) mesoderm. The paraxial mesoderm gives rise to the somites in the trunk, but in the head forms the cranial paraxial mesoderm which does not undergo segmentation. In zebrafish, there is also presence of an anterior lateral plate mesoderm which gives rise to secondary heart field and vasculature of the head (Guner-Ataman et al., 2013; Sorrell et al., 2013). The cranial mesoderm and neural crest both contribute to the mesenchyme, bones and muscles of the head and neck, but in a largely mutually exclusive manner as specific bone and cartilage structures are derived from either one or source or the other. Generally, more anterior structures are neural crest-derived and posterior structures come from mesoderm. However, formation of head structures is dependent on interactions at the interface of the two tissues (Rinon et al., 2007; Grenier et al., 2009; Liang et al., 2014). Thus, it appears as though the cranial mesoderm and neural crest are developmentally intertwined.

1.3 Signaling Pathways in Development of Vascular Stability

Although initially independent, the later stages of endothelial and mural cell development become heavily intertwined, each cell type relying on signals and cues from the other to properly establish a mature and functional phenotype (Fig. 1.4). Mural cell precursors are recruited to developing endothelial cells, where under the influence of autocrine and paracrine signaling cues, both cell types undergo further differentiation and maturation to form a functional and stable vessel wall. This dynamic and complex process requires input from several signalling pathways, and perturbations of any of these can lead to embryonic lethality due to ruptured or leaky vessels.

1.3.1 PDGF-B/PDGFRβ

Endothelial cells secrete platelet-derived growth factor B (PDGF-B) (Collins et al., 1985), which signals through PDGF receptor beta (PDGFRβ) expressed on pericytes and vSMCs (Lindahl et al., 1997; Hellstrom et al., 1999). PDGF-B activation of PDGFRβ induces mural cell proliferation and subsequent recruitment to the endothelium (Lindahl et al., 1997; Hellstrom et 12

Figure 1.4 – Endothelial-Mural Cell Interactions

Diagram outlining the major signaling pathways and their roles in coordinating development of endothelial cell, mural cells and vascular development, as well as junctional complexes contributing to vascular stability.

al., 1999; Hirschi et al., 1999; Abramsson et al., 2003). Tip-cells of angiogenic sprouts have higher PDGF-B expression, suggesting that mural cells are preferentially recruited to sites of actively developing endothelium (Gerhardt et al., 2003). Loss of PDGF-B, PDGFRβ, or downstream signaling components results in loss of pericytes causing hemorrhage and edema (Leveen et al., 1994; Soriano, 1994; Lindahl et al., 1997; Enge et al., 2002; Bjarnegard et al., 2004). Secreted PDGF-B interacts with heparin sulfate proteoglycans (HSPG) in the ECM and on the cell surface via a retention motif, and loss of this motif or proper HSPG expression on cell surfaces also produces vascular stability defects. Pericyte-specific ablation of HSPG causes several vascular stability defects, but localization to endothelial cells is unaffected (Stenzel et al., 2009), suggesting that endothelial cells primarily mediate initial recruitment, and subsequent stabilization requires pericyte-derived PDGF-B retention mechanisms.

1.3.2 TGF-β

The TGF-β/BMP superfamily of signaling factors control many different processes throughout development. TGF-β is important in development of vascular stability, and acts in both mural cells and endothelial cells. Two TGF-β receptors – Alk1 and Alk5 – induce opposing cellular processes: activation of Alk5 and subsequent phosphorylation of Smad2/3 promotes differentiation, whereas Alk1 signaling phosphorylates Smad1/5 to promote proliferation and migration. Generally, pericytes and vSMCs express the Alk5 receptor, and are induced to differentiate by TGF-β signaling (Chen et al., 2003; Nuessle et al., 2011; Tang et al., 2011) Endothelial cells, however, express both receptors and rely on a balance of activation for proper vessel growth and patterning (Oh et al., 2000; Goumans et al., 2002; Goumans et al., 2003). Alk1 mediates early endothelial migration and proliferation, but as mural cells are recruited, increased TGF-β secretion preferentially activates Alk5, inhibiting proliferation and sprouting and promoting mural cell-dependent vessel quiescence. Disruption of TGF-β or either Alk receptor results in vascular defects that commonly lead to gestational lethality (Dickson et al., 1995; Urness et al., 2000; Larsson et al., 2001).

1.3.3 Notch

In recent years, the importance of Notch signaling in vascular stability and mural cell development has come to light. Notch3 is the primary Notch receptor expressed in mural cells of the cerebral vasculature (Joutel et al., 2000; Wang et al., 2008), and binds Jagged1 and/or Dll4 ligands expressed on endothelium (High et al., 2008; Liu et al., 2009; Stewart et al., 2011). Induction of Notch3 signaling primarily promotes mural cell differentiation, since mutations in Notch3 decrease expression of mature vSMC markers (Domenga et al., 2004; Liu et al., 2010; Ghosh et al., 2011; Wang et al., 2014), as do mutations of Jagged1 (High et al., 2008; Xia et al., 2012) or Notch-related transcriptional coactivators (High et al., 2007; Manderfield et al., 2012). In some instances, Notch3 appears to promote mural cell proliferation, survival and investment on endothelial cells (Liu et al., 2010; Li et al., 2013; Yang et al., 2013; Wang et al., 2014; Baeten & Lilly, 2015), although some in vitro observations suggest Notch3 inhibits differentiation (Morrow et al., 2005). Initial Notch3 signaling leads to a feedback loop, whereby Jagged1 is induced in mural cells and allows for cell autonomous Notch signaling to propagate mural cell development (Domenga et al., 2004; Liu et al., 2009; Ghosh et al., 2011; Manderfield et al., 2012) Both mouse and zebrafish Notch3 mutants exhibit disrupted vascular wall integrity due to decreased mural cell coverage (Wang et al., 2014; Henshall et al., 2015). During development, Notch3 is more highly expressed in mural cells of neural crest origin than those of mesodermal origin, pointing to an increased requirement for Notch3 in mural cells that populate the cranial vasculature (Granata et al., 2015). Notch3 regulates expression of other signaling pathway constituents that promote mural cell-endothelial cell interactions, such as TGF-β and PDGFRβ. For instance Notch3 deficient vSMCs in the Circle of Willis show decreased PDGFRβ expression (Yang et al., 2013). Coculture models with activated Notch3 in pericytes induced expression of PDGFRβ, Ng2, TGF-β and other signaling pathways (Schulz et al., 2015). Furthermore, Notch3 functions cooperatively with Alk1-mediated TGF-β signaling to promote vessel quiescence (Larrivee et al., 2012). Endothelial expression of adherens junction component N-Cadherin is induced by physical interaction between Smad4 and the Notch intracellular domain on target gene promoters (Li et al., 2011). 15

1.3.4 Other Vascular Stability Signaling Pathways

In addition to the previously mentioned signaling pathways, many other molecular signals play a role in promoting endothelial-mural cell interactions. Angiopoetin1 (Ang1) and its target receptor TIE2 on endothelial cells, forms a reciprocal signaling loop in cooperation with PDGF-B/PDGFRβ. Ang1 is secreted from immature mesenchymal cells and signals through TIE2 receptors on endothelial cells to promote endothelial maturation (Dumont et al., 1993; Davis et al., 1996; Suri et al., 1996; Sundberg et al., 2002; Falcon et al., 2009) and pericyte recruitment (Cai et al., 2008b). Sphingosine-1-phosphate (S1P) is secreted from platelets and hemangioblasts to an endothelial receptor (S1P1) to promote endothelial maturation (Yatomi et al., 1997; Yang et al., 1999; Allende & Proia, 2002). Wnt signaling through canonical (β- catenin, TCF) pathways primarily affects vSMC proliferation, in that β-catenin and TCF promote cyclin D1 and downregulate p21 (Quasnichka et al., 2006; Bedel et al., 2008). Evidence of Wnt- driven VSMC proliferation in vivo is relatively scarce {Tsaousi, 2011, Ezan, 2004}, but in vitro studies show a role for β-catenin mediated proliferation of arterial and venous vSMCs (Wang et al., 2002b; Ezan et al., 2004; Slater et al., 2004; Wang et al., 2004).

1.4 Structural and Physical Interactions of Vascular Stability

In addition to signaling, vascular stability requires the physical interaction between mural cells and endothelial cells to enact support. vSMCs contacts with the endothelium are mediated by the basement membrane, whereas pericytes make direct cell-cell contact with the abluminal surface of endothelial cells. These contacts help to form the blood brain barrier (BBB), a central nervous system-specific structure comprised of endothelial cells, pericytes and astrocytes, both physically and biochemically mediates passage of bloodstream constituents to the brain parenchyma (Fig. 1.4). Pericytes of the blood brain barrier reside in the basement membrane of cerebral endothelial cells, and are located in between endothelium and astrocyte end-feet. Disruption of the BBB results in leaky cerebral vessels, cerebral edema and hemorrhage. These insults contribute to cerebral tissue degeneration, and can ultimately result in dementia and

stroke. In order to establish barrier function, pericytes must maintain highly regulated physical interactions with the endothelial cells.

1.4.1 Basement Membrane

The basement membrane is both a support structure and a means of signaling regulation. ECM proteins of the membrane provide elastic mechanical resistance and a substrate to which the cytoskeleton of cells can bind via integrins. Signaling molecules periodically require binding to ECM constituents to properly elicit a response, such as PDGF-B. Both pericytes and endothelial cells secrete vascular ECM proteins (Cohen et al., 1980; Mandarino et al., 1993). Major constituents of the cerebral vascular basement membrane include collagen IV (Khoshnoodi et al., 2008), laminins (Hallmann et al., 2005), nidogens (Bader et al., 2005), fibronectin (Caffo et al., 2008), the heparin-sulfate proteoglycan perlecan (Gustafsson et al., 2013) and agrin (Barber & Lieth, 1997). Loss of mural cells results in a decrease in basement membrane proteins, particularly collagen IV and laminin (Bell et al., 2010). Initial deposition of the basement membrane is protected from MMP degradation by pericyte secretion of tissue inhibitor of metalloproteinases 3 (TIMP3) (Lafleur et al., 2001). However, mural and endothelial cells also secrete various MMPs in response to angiogenic stimuli (Virgintino et al., 2007), when dynamic rearrangement of the basement membrane is required.

1.4.2 Integrins

Integrins interact with the extracellular matrix to form focal adhesions and mediate signaling. Integrins are widely expressed, but specific subunits are tissue-selective. Integrin subunits α5, αV, β1 and β8 are important for the vasculature of the central nervous system. In particular, αVβ8 integrin knockouts commonly result in mouse prenatal death, exhibiting excessive vascular sprouting and cerebral hemorrhage, which is thought to be due to disrupted TGF-β signaling (Weisiger & Zucker, 2002; Arnold et al., 2012; Arnold et al., 2014). This phenotype is observed in neural cell-specific knockout of αVβ8 as well as α5 integrins, implicating a role for neural cells in promoting vascular growth and patterning (McCarty et al., 2005a;

McCarty et al., 2005b; Arnold et al., 2014). β1, β3 and α4 subunits promote endothelial cell 17

growth and mural cell adhesion and coverage (Garmy-Susini et al., 2005; Grazioli et al., 2006;

Wang & Milner, 2006; Abraham et al., 2008), and β1 is required for assembly of endothelial tight and adherens junctions complexes (Osada et al., 2011; Yamamoto et al., 2015). Mutations affecting proteins involved in integrin and focal adhesion assembly can also result in vessel instability (Shen et al., 2005), such as zebrafish bbh and fl02k mutants (Liu et al., 2012; Wu et al., 2015).

1.4.3 Junctional Complexes

Junctional complexes play a large role in regulating the blood-brain barrier. Interactions between pericytes and endothelial cells are mediated by gap junctions, N-Cadherin, and cytoplasmic structures known as “peg-and-socket” junctions (Diaz-Flores et al., 2009), and tight junctions form connections between neighbouring endothelial cells. Gap junctions are comprised of connexion channels that create a direct avenue for pericyte-endothelial cell communication (Cuevas et al., 1984; Larson et al., 1987; Fujimoto, 1995). The primary microvascular connexins Cx43 and Cx45 (Cohen-Salmon et al., 2004; Figueroa & Duling, 2009) are required for mediating TGF-β induced mural cell differentiation and maintaining vascular integrity (Kruger et al., 2000; Hirschi et al., 2003). Adherens junctions connect cytoskeletons of neighbouring cells to provide support and mediate transduction of mechanical forces. N-Cadherin mediates pericyte-endothelial interactions (Gerhardt et al., 2000), whereas endothelial-endothelial adherens junction proteins include VE-Cadherin (VE-Cad), nectin and PECAM (Takai et al., 2003; Dejana & Vestweber, 2013; Privratsky & Newman, 2014). Inhibition of adherens junction formation disrupts cell-cell interactions, and can decrease vascular integrity (Gerhardt et al., 2000; Dejana & Vestweber, 2013; Privratsky & Newman, 2014). N-Cad expression in mural cells is coordinately regulated by TGF-β and Notch signaling, as mentioned earlier (Li et al., 2011), and VE-Cad function is dependent on the presence of endothelial N-Cad (Luo & Radice, 2005). Endothelial tight junctions are comprised of occludin (Furuse et al., 1993), claudins (primarily claudin-5 at the BBB) (Nitta et al., 2003; Daneman et al., 2010), and junctional adhesion molecules (Martin-Padura et al., 1998). Tight junctions are formed closest to the lumen, on the lateral membranes of endothelial cells, and provide a strong initial barrier against

vessel permeability. Pericyte interactions with endothelial cells, such as Ang1 signaling, are able to regulate tight junction protein expression and localization in endothelial cells (Hori et al., 2004; Kim et al., 2009; Bell et al., 2010). Peg-and-socket junctions are an interesting form of pericyte-endothelial interaction. Cytoplasmic protrusions from pericytes break through the basement membrane and invade pockets of the endothelial cytoplasm, maintaining cell membrane integrity (Matsusaka, 1975; Wakui et al., 1989; Rucker et al., 2000). Peg-and-socket junctions may create some level of physical anchorage between the two cell types, and potentially provide a site free of interfering basement membrane for other junctional complexes to assemble. They also act as an excellent morphological hallmark of pericytes on endothelial cells.

1.5 Vascular Stability Disease

Disruption of endothelial-mural cell interactions results in immediate or prolonged adverse effects on the health and functionality of an individual. One of the goals of my project is to connect our basic science knowledge of vascular development with human genetics to understand if any of the genes that we work on also play a role in human vascular disorders. Our fish models with developmental hemorrhage may provide insight on the genetics of vascular stabilization in fish, and potentially humans. In humans, loss of vascular stability can lead to a wide array of diseases, varying in terms of phenotype, severity, number and type of causative factors, and age of onset. Disruptions of endothelial-mural cell interactions can occur on several levels, from initial PDGF-mediated recruitment signaling to monogenic deletions of downstream basement membrane and junctional proteins (Gould et al., 2005; Tillet et al., 2005; Menezes et al., 2014; Tien et al., 2014). Vascular stability disease in the brain can arise internally due to defects in BBB integrity, or from external influences that place stress on endothelial-mural cell interactions. Regardless of source, the end result is almost always stroke or stroke-like symptoms. Stroke is defined by neuronal cell death in the brain due to lack of blood supply (ischemia) or by hemorrhage which increases pressure and inflammation, and both forms of stroke involve disruption of mural cells.

1.5.1 Hemorrhagic Stroke

Hemorrhagic stroke is caused by rupture of vessel walls, causing bleeding in the brain, and can be a direct result of poor mural-endothelial interactions. Developmentally, preterm infants have a high rate of intraventricular hemorrhage due to deficiencies in the basement membrane and reduced pericyte coverage of cerebral vessels (Ballabh, 2010). In adult brains, two forms of hemorrhagic stroke can occur: subarachnoid hemorrhage, where blood escapes into the meningeal space between the arachnoid and pia maters, and intracerebral hemorrhage that occurs in the brain itself. Hemorrhages can occur as a result of aneurysms, arterio-venous malformations (AVMs), cerebral cavernous malformations (CCMs), cerebral amyloid angiopathies (CAA) and hypertension, which also all involve disrupted mural cell function. Aneurysms can form specifically as a result of pericyte- or vSMC-deficient vasculature (Lindahl et al., 1997; Kondo et al., 1998), forming hyperdilated vessels prone to rupture. Cerebral amyloid angiopathy and hypertension, however, induce degeneration of mural cells secondary to other causative factors (Verbeek et al., 2000; Kruyer et al., 2015). AVMs and CCMs are erratic, poorly formed clusters of high- and low-flow vessels, respectively, lacking typical hierarchal structures and artery-vein identities. Both AVMs and CCMs can form due to disrupted Notch3 signaling between endothelial and mural cells (Krebs et al., 2004; Schulz et al., 2015). Additionally, human mutations in the Alk1 and Endoglin TGF-β receptors result in hereditary hemorrhagic telangiectasia, an AVM-related hemorrhagic disease, where hemorrhage is attributable to disrupted mural cell coverage and differentiation on AVMs and CCMs (Hoya et al., 2001; Meng & Okeda, 2001; Uranishi et al., 2001; Chen et al., 2013).

1.5.2 Ischemic Stroke

Ischemic stroke can be caused by any event that blocks or severely reduces blood flow to the brain, including prior cerebral hemorrhage. Although mural cells have flow-regulating contractile function, occlusion of vessels is generally not due to sustained or over-exaggerated contraction of vSMCs or pericytes. If mural cells are causative of ischemia, it generally occurs indirectly through formation of atherosclerotic plaques which thrombose and lodge in smaller vessels. Ischemic stroke does however induce various responses in cerebral pericytes. Following 20

an ischemic insult, reperfusion of the affected tissue is required to attenuate neuronal cell death. Surrounding vessels undergo angiogenesis in response to hypoxic signals, promoting neovascularization of the ischemic area, or infarct (Liu et al., 2014). These angiogenic sprouts increase PDGF-B secretion (Arimura et al., 2012) and pericytes are recruited to the peri-infarct region (Gonul et al., 2002; Melgar et al., 2005; Duz et al., 2007), and show upregulated PDGFRβ signaling to further promote reperfusion and neuroprotection (Krupinski et al., 1997; Arimura et al., 2012; Shen et al., 2012). Pericyte differentiation from bone marrow is also induced to populate the peri-infarct region (Kokovay et al., 2006). Interestingly, pericytes also develop a degree of multipotency, enabling differentiation into neuronal as well as vascular cell types (Nakagomi et al., 2015). Despite this neuroprotective function, activation of pericyte migration and multipotency also increases BBB permeability in ischemic stroke (Baumann et al., 2009; Israeli et al., 2010), due to detachment from the basement membrane and loss of normal vascular stability regulation (Al Ahmad et al., 2009). Another detrimental effect occurs in contractile mural cells, which constrict vessels in response to ischemic conditions (Yemisci et al., 2009; Hall et al., 2014). Often, this contraction is sustained for a period of time after the ischemic event, and paradoxically inhibits reperfusion of the atrophied tissue. Studies have suggested that this is due to cell death, resulting in a kind of rigor mortis that maintains mural cells in a contracted state (Hall et al., 2014). Sustained mural cell contraction has been a target of therapeutic studies in an attempt to increase reperfusion and neuronal cell survival in ischemic stroke.

1.5.3 Cerebral Small Vessel Disease

While much of stroke research focusses on strokes caused by single ischemic events, cerebral small vessel disease (CSVD) is also prevalent and risk increases with age. CVSD is a group of diseases caused by defects in capillaries and small vessels that result in neuronal damage (Fig. 1.5). These lead to increasing BBB permeability, causing cognitive decline, dementia and stroke. CSVD is identified by white matter hyperintensities, lacunar infarcts and microbleeds present in cerebral CT or MRI imaging.

Figure 1.5 – Cerebral Small Vessel Disease

Diagram depicting CSVD hallmarks and potential causes, with associated MRI/CT image examples of each. Diagram and microbleed images adapted from (Charidimou et al., 2012). Leukoaraiosis image adapted from (Debette & Markus, 2010). Lacunar infarct image adapted from (Norrving, 2008). Images reproduced with permission from the publishers.

Certain heritable factors can accelerate CSVD phenotypes. CADASIL is a dominant hemorrhagic disorder caused by mutations in Notch3 that disrupt normal distribution of cysteine residues in the extracellular domain (Ruchoux et al., 1995; Joutel et al., 1996; Federico et al., 2005). Notch3 is required for maintenance of vSMC contractile phenotype, and mutations in Notch3 cause defects in endothelial-mural cell interactions (Li et al., 2011; Henshall et al., 2015; Schulz et al., 2015) and mural cell survival (Wang et al., 2002a; Wang et al., 2003b; Sweeney et al., 2004; Arboleda-Velasquez et al., 2014) resulting in dysfunctional vessels. The effect of Notch3 mutations on the cerebral vessels in humans is the progressive loss of vSMC and pericyte coverage, and an accumulation of vascular-based neural insults. As a result, CADASIL culminates in stroke and dementia (Joutel et al., 2000; Dziewulska & Lewandowska, 2012; Gu et al., 2012; Ghosh et al., 2015). Notch3 mutations mainly affect vSMCS on arterioles, but some evidence suggests that capillary-associated pericytes also play a role in CADASIL pathology (Gu et al., 2012; Craggs et al., 2015; Ghosh et al., 2015). Mutations in collagens are another heritable CSVD factor, specifically Col4A, which represents the major collagen type in cerebral vascular basement membrane. Col4A1/2 mutations are autosomal dominant and can result in phenotypes as severe as ischemic stroke and intracerebral hemorrhage (Verbeek et al., 2012; Lemmens et al., 2013; Rannikmae et al., 2015), but consistently exhibit the MRI/CT imaging hallmarks of CSVD. Beyond these genetic risk factors we know very little about the causes of CVSD. However our recent collaborative projects on FoxC1 and FoxF2 highlights these transcription factors as potential genetic causes (French et al., 2014; Chauhan et al., 2016).

1.6 Iguana/Dzip1 mutants as a model of vascular stability defects

1.6.1 Shh Signaling and Vascular Stability

Hedgehog (Hh) signaling is an evolutionarily conserved pathway that regulates developmental processes ranging from early establishment of body axes to morphogenesis of tissues and organs (Weed et al., 1997).Of the three homologous subgroups of vertebrate Hh molecules (Sonic hedgehog (Shh), Indian hedgehog (Ihh) and Desert hedgehog (Dhh), Shh is the most broadly expressed. In vascular development, Shh signaling from the notochord induces 23

VEGF production to promote endothelial cell morphogenesis (Pola et al., 2001; Lawson et al., 2002). Shh also has roles in vSMC function. PDGF signaling induces Shh expression in mural cells, which mediates PDGF-induced vSMC migration (Yao et al., 2014), possibly through promoting phenotypic modulation to a synthetic state (Zeng et al., 2016). Shh has also been shown to promote vSMC proliferation in response to hypoxia (Wang et al., 2010) and autophagy for the purpose of cell survival (Li et al., 2012) However whether vessel integrity was compromised was not identified in these studies.

1.6.2 Primary Cilia are required for Shh Signaling

The primary cilia act as a focal point for Hh signaling in vertebrates. Hh acts by binding its receptor Patched (Ptch), relieving inhibition of Smoothened (Smo) and allowing Smo to translocate to the distal tip of primary cilia (Corbit et al., 2005),where it initiates activation of the Glioma (Gli) transcription factors. The presence of primary cilia is therefore indispensable for Hh signaling in vertebrates as inhibition of cilia formation blocks Hh signaling by disrupting localization of Hh pathway proteins (Caspary et al., 2007) (Fig. 1.6). Our lab found that the zebrafish iguana (igufo10) genetic mutant has a mutation in the zinc finger and coiled coil ciliary basal body protein Dzip1 (DAZ-like interacting protein) that causes the truncation of primary cilia, resulting in deficient Hh signalling (Sekimizu et al., 2004; Wolff et al., 2004; Glazer et al., 2010; Kim et al., 2010). One key phenotype of this mutant is an increased potential for cerebral hemorrhage. These mutants have decreased Ang1 signaling, and overexpression of ang1 partially rescues hemorrhage (Lamont et al., 2010), implicating Shh signaling in vascular stability. igu may influence Hh signalling through multiple mechanisms. igu mutants have stunted primary cilia, and as Hh signalling depends on cilia, loss of Dzip1 attenuates Hh signalling by this indirect mechanism (Glazer et al., 2010; Kim et al., 2010; Tay et al., 2010). Recently, Dzip1 has also been found to have an additional regulatory effect on Gli activity. Speckle-like POZ protein (Spop) is an adaptor protein for cullin 3 ubiquitin ligase and is involved in proteasome- mediated degradation of Gli proteins. Dzip1 functions to stabilize Spop, and slight downregulation of Dzip1 function maintains ciliary function, but causes a sensitization to Hh signaling due to reduced Gli turnover (Jin et al., 2011; Schwend et al., 2013; Wang et al., 2013). Evidence

Figure 1.6 – Primary cilia formation and Hh signaling are dependent on dzip1

A: Dzip1 localized to the basal body promotes ciliogenesis. B: Hh signaling occurs at the primary cilia, binding the Ptc receptor and allowing Smo to translocate to the distal portion of the cilia to activate Gli genes. Dzip is required for creation of the primary cilia to act as a focal point for signaling. Dzip1 also has a role in Gli turnover. Adapted from (Wilson & Stainier, 2010) with permission of the publisher.

that Dzip1 functions in a bipartite manner to regulate Hh signaling, may explain how loss of Dzip1 results in both inactivation of some, and over-activation of other, Hh target genes in a tissue-dependant manner. For example, igu mutants display a down-regulation of ptch2 and engrailed in the neural tube while these genes are up-regulated in the somites (Sekimizu et al., 2004; Wolff et al., 2004). However, the full extent of differential gene regulation in igu mutants is not known. In Chapter 3 I describe using the igu mutant to discover new genes involved in vascular stabilization downstream of Hh signaling.

1.7 Fox Transcription Factors

One of the genes that is activated by Hh signaling is a member of the Fox (Forkhead box) family of transcription factors. Fox proteins show evolutionary conservation across all metazoans and unicellular eukaryotes. They are named after the forkhead (fkh) gene in Drosophlia, discovered through a mutation that forms a spiked head phenotype (Weigel et al., 1989). The number of Fox genes within a species varies greatly; there are only 4 Fox genes in Saccharomyces cerevisiae, but 50 in humans (Jackson et al., 2010). Fox genes are grouped into 19 subfamilies, ranging from FoxA to FoxS, based on sequence homology (Kaestner et al., 2000; Jackson et al., 2010), with subfamilies having up to 4 members. All Fox proteins shared a highly conserved DNA-binding domain, called the forkhead domain, which is the basis of their classification. The forkhead domain forms a modified helix-turn-helix DNA-binding motif, featuring loops or “wing” structures, and as such is called a winged-helix motif (Brennan, 1993; Clark et al., 1993). Sequence conservation in the winged-helix motif of orthologous Fox genes of different species is surprisingly high. Human FoxA1 and Drosophila Forkhead show 90% sequence similarity between their respective forkhead domains (Hannenhalli & Kaestner, 2009). This conservation coincides with a relatively non-degenerate core DNA binding site, RTAAAYA (Pierrou et al., 1994), which is recognized by all Fox genes. Aside from the DNA- binding domain however, Fox protein sequences differ greatly, reflecting a wide range of biological functions in which they enact regulatory roles. Fox proteins bind DNA and regulate transcription as monomers, with the exception of the FoxP subfamily, which demonstrates some ability for dimerization (Stroud, 2006; Perumal et al., 2015). 26

1.7.1 Forkhead Winged Helix Motif Structure

The characteristic winged helix motif of Fox proteins combines the helix-turn-helix motif with winged protein loops (Brennan, 1993; Clark et al., 1993) (Fig. 1.7A,B). It is made up of three α-helices and three β-sheets, and for the purpose of conceptualization, can be roughly divided into two main structures: an N-terminal DNA-binding domain and the C-terminal wing structures. The DNA-binding side involves all three α-helices and resembles the helix-turn-helix motif: two α-helices separated by a β-strand, followed by a turn to the third α-helix, which is presented to the major groove of the DNA helix for binding. The wing structures follow sequentially, the first wing being a loop formed by antiparallel alignment of the second and third β-sheets, and the second wing being an unstructured protein chain at the C-terminal end of the third β-sheet that is thought to interact with the minor groove of the DNA helix. As such, the wing domains are set up such that the antiparallel β-sheets flank the first wing, and both wings flank the third β-sheet.

1.7.2 FoxF2

This thesis focuses on FoxF2 and its role in vascular stability, modeled in developing and juvenile zebrafish (Danio rerio). Originally identified as hepatocyte nuclear factor (Clark et al., 1993), the first study to distinctly characterize FoxF genes named them FREAC (Forkhead- RElated ACtivator), and described two related genes, FREAC-1 and FREAC2, in lung-specific gene activation in human cell lines (Hellqvist et al., 1996). With the discovery of additional related transcription factors, the family was re-categorized as Fox genes, and FREAC-1 and -2 became FoxF1 and FoxF2. The DNA binding site of FoxF2 is AAC(GTAAACA)A, (Pierrou et al., 1994; Pierrou et al., 1995), where the parentheses highlight the FoxF2 iteration of the core RTAAAYA conserved forkhead target sequence. Analysis of human FoxF2 protein has identified nuclear localization signals in the forkhead DNA binding domain, as well as at least two domains in the C-terminal side of the DNA binding domain that promote activation of FoxF2 by interaction with other proteins (Hellqvist et al., 1998) (Fig. 1.7C). In the genome of vertebrates, FoxF2 is located in a cluster of Fox genes flanked by FoxC1 and FoxQ1, and this 27

Figure 1.7 – Structure of FoxF2 and the winged-helix DNA-binding domain

A: Diagram of winged-helix DNA-binding domain showing 3 α-helices (H1-3) and 3 β-sheets (S1-3), and 2 wing domains (W1 and 2). N and C are the N-terminal and C-terminal ends, respectively. T’ indicates the turn sequence, characteristic of helix-turn-helix DNA-binding domains. B: Diagram depicting winged-helix DNA-binding domain interacting with the major and minor grooves of DNA via the third α-helix and second wing domain, respectively. A and B are adapted from (Clark et al., 1993) with permission of the publisher. C: Diagram of human FoxF2 protein showing relative position of forkhead DNA binding domain and putative activation domains. Nuclear localization signal was determined to be part of the conserved DNA binding domain. Adapted from (Hellqvist et al., 1996) with permission of the publisher.

cluster is evolutionarily conserved from zebrafish to humans. Phylogenetically, these genes are relatively closely related to one another, even across species (Fig. 1.8). These genes may share some similar functions, and selective pressures has maintained the cluster throughout evolution.

1.7.3 FoxF2 mediates mesenchyme in development

During development in mouse, FoxF2 is expressed in the splanchnic mesoderm (Ormestad et al., 2004) and mesenchyme of developing structures including the oral cavity, intestine, lung, and genitals (Aitola et al., 2000). Expression is most prominent in the oral cavity throughout development. FoxF2 mRNA is also detected in migrating cranial neural crest cells (Ormestad et al., 2004), and shows reduced expression in lung and foregut, but strong expression in hindgut. The drosophila FoxF2 homolog specifies visceral mesoderm and is required for gut differentiation, along with Nkx3.2 homolog Bagpipe (Perez-Sanchez et al., 2002; Jakobsen et al., 2007). Xenopus FoxF2 expression is also localized to the gut (McLin et al., 2010). A similar role for gut development has also been identified in mice, where FoxF2-/- null embryos showed decreased association between mesenchymal and epithelial layers and reduced ECM deposition, and a loss of acta2 (Ormestad et al., 2006). FoxF2-/- embryos also showed increased Wnt5a expression in epithelial layers due to decreased signaling from BMP4, implicating FoxF2 in regulation of BMP signaling (Ormestad et al., 2006). This FoxF2-mediated repression of Wnt translates to overproliferation of intestinal epithelium and adenoma in FoxF2+/- mice (Nik et al., 2013; van den Brink & Rubin, 2013). Smooth-muscle specific FoxF2 knockout mice show increased smooth muscle cell proliferation, and overexpression of PDGF and Shh signaling components, as well as the transcription factor myocardin (Bolte et al., 2015).FoxF2 therefore plays a strong role in visceral smooth muscle cell differentiation.

1.7.4 FoxF2 is also involved in formation of the palate.

It is expressed strongly in the oral cavity mesenchyme in mouse, and both mouse and human FoxF2 mutants show cleft palate (Wang et al., 2003a; Jochumsen et al., 2008). In the

Figure 1.8 – Phylogenetic tree of Fox genes

Phylogenetic tree depicting relationships of Fox genes from 5 different genomes: Hsa = Homo sapiens, Ga = Gallus gallus, Xl = Xenopus laevis, Tni = Tetraodon nigroviridis, Ce = Caenorhabditis elegans. FoxF (red box), FoxQ (blue box) and FoxC (purple boxes) show relatively low divergence from one another, even between species. Adapted from (Shen et al., 2011) with permission of the publisher.

palate, FoxF2 is induced by Shh signaling (Lan & Jiang, 2009), and both neural crest-specific and palatal mesenchyme specific knockout of FoxF2 leads to complete penetrance of cleft palate in mice (Xu et al., 2016). Palatal mesenchyme proliferation is decreased in FoxF2-/- mice (Nik et al., 2016; Xu et al., 2016), as well as TGFβ2 protein, due to the lack of ECM proteins required for TGFβ activation (Nik et al., 2016). A Shh-dependent role for FoxF2 is also seen in the secondary heart field, where FoxF2+/-;FoxF1+/- compound heterozygotes fail to undergo atrioventricular septation, and show overgrowth of mesenchyme at the tip of the septum (Hoffmann et al., 2014).

1.7.5 FoxF2 inhibits EMT to repress cancer metastasis

FoxF2 is downregulated in prostate, breast, colorectal and liver cancer (van der Heul- Nieuwenhuijsen et al., 2009b; Hirata et al., 2013; Kong et al., 2013; Zhang et al., 2015; Shi et al., 2016). FoxF2 deficiency is thought to enhance EMT and metastatic abilities of tumours. In the prostate, FoxF2 opposes TGFβ signaling, decreases matrix-metalloproteinase (MMP) expression and induces MMP inhibitors (van der Heul-Nieuwenhuijsen et al., 2009a). Hepatocellular carcinomas show increased proliferation with repressed FoxF2 expression (Shao et al., 2015; Shi et al., 2016), and FoxF2-mediated repression of EMT-promoting factors Twist1 and FoxC2 is relieved in basal-like breast cancer models due to FoxF2 inhibition (Cai et al., 2015; Wang et al., 2015). Interestingly, FoxF2 is found to promote metastasis of epithelial lung cancers (Kundu et al., 2016). In many cases of cancer, FoxF2 deficiency appears to be a result of miRNA targeting (Shi et al., 2011; Hirata et al., 2013; Shao et al., 2015; Zhang et al., 2015; Kundu et al., 2016), however the role for miRNA regulation of FoxF2 in development has not been investigated.

1.8 Unanswered questions in vascular stabilization

Although the general signaling pathways and how they contribute to vascular stabilization during development are largely known, the specific mechanisms through which these pathways act still remain a mystery. How is a pericyte specified to be different from a vascular smooth muscle cell? Are vascular smooth muscle cells and pericytes part of the same 31

cellular lineage, or do they differentiate separately? How are differences in mural cells established along the hierarchy of vessel branches? Fine-tuning of the developmental process by transcriptional regulation under the control of signaling pathways is thought to control these processes, but the specifics remain unclear. In the case of Shh signaling, it has been known for some time that Shh has a dual role in promoting both endothelial growth (through stimulation of VEGF signalling) and vascular stabilization (through an unknown mechanism). This question is one that I chose to address in my work as it is unclear what is regulated downstream of Gli transcription factors that could contribute to vascular mural cell development. FoxF2, a transcription factor identified as a target of Shh signaling, has potential for involvement in development of vascular mural cell development and establishment of vascular stability genes. However, few direct targets of FoxF2 are known, and investigating its mode of action would provide novel insights into a relatively unknown transcription factor. Studies across model organisms could also hint at the evolutionary conservation of FoxF2 function. Analysis can also be extended to the conserved cluster of Fox genes that includes foxc1 and foxq1. Are these genes also contributing to vascular stability, and does this indicate evolutionary conservation of a transcription factor cluster? Furthermore, since many vascular diseases involve a degeneration or malfunction of vascular stability components later in life, there must be a need for active maintenance of mural cells after its initial establishment. Do factors initially involved in mural cell development remain throughout life to play a role in maintaining vascular stability, or are new genes activated after mural cell differentiation to regulate stability in adults? In the case of genetic diseases, how is it that individuals are able to survive to a certain age before presenting symptoms of stroke and dementia? Using the zebrafish as a model system, I set out to investigate a role for FoxF2 in vascular mural cells during both development and adult stages.

Hypothesis: Shh-mediated FoxF2 plays a role in establishing and maintaining vascular stability, particularly in the head. In this thesis I present novel evidence that the transcription factor FoxF2 mediates the initial establishment of vascular stability in the zebrafish, and that FoxF2 expression is dependent on Shh signaling, indicating a novel role for Hh signaling in development. I analyze the

expression profile of the Shh signaling mutant iguana, identify FoxF2 as a downregulated gene in iguana mutants, determine the expression domain of FoxF2 in developing zebrafish, and investigate the effect of loss of FoxF2 on vascular stability in transient or genetic loss-of- function models of FoxF2 in developing and juvenile zebrafish tissues.

Chapter Two: Materials and Methods

2.1 Zebrafish Maintenance and Husbandry

Zebrafish were maintained in accordance with standard procedures under the approval of the Animal Care Committee at the University of Calgary (Permit Number: AC13-0249). Embryos were generated and maintained as outlined by Westerfield (Westerfield, 2007). Embryos were collected and maintained at 28.5⁰C in E3 buffer (5mM NaCl, 0.17mM KCl,

0.33mM CaCl2, 0.33mM MgSO4), and staged in hours post fertilization (hpf) as previously described (Kimmel et al., 1995). Where necessary, embryos were dechorionated with 20mg/ml pronase (Sigma, St. Louis MO) for 1.5 min. Also where necessary, 0.003% phenylthiourea (PTU; Sigma) was added to E3 at 24hpf to block pigment formation. Transgenic and mutant zebrafish lines have been previously described: Tg(6.5kdrl:mCherry)ci5 (Proulx et al., 2010), Tg(acta2:EGFP)ca7 (Whitesell et al., 2014), igufo10 (Lamont et al., 2010). For breeding, zebrafish mating sets were set up the afternoon prior in mating tanks with a sieve-bottom insert. Males and females were separated by a removable divider, and were preferentially placed in ratios of 3 females to 2 males. The following morning, water in the mating tank was changed and dividers were removed to allow males and females to interact. Eggs were laid, fertilized and collected within an hour of pulling dividers.

2.2 Morpholino Injections

Morpholinos were injected into the 1-8 cell stage of zebrafish embryos using 1mm microcapillary needles (World Precision Instruments, Sarasota, FL) pulled to a point using a P- 97 Flaming/Brown micropipette puller (Sutter Instruments Co., Novato, CA). Pulling programs were chosen based on injection technique (Table 2.1). Needles were attached to a Femtojet injector (Eppendorf AG, Hamburg, Germany), and maneuvered with a micromanipulator. Zebrafish embryos were collected in E3 solution and when necessary, dechorionated with pronase. Embryos were then placed in an agarose ramp and injected using the micromanipulator. The amount of solution injected was observed under a dissecting microscope (Leica Microsystems, Richmond Hill, ON) and volume was measured by diameter using an eyepiece graticule (Leica). After injection, embryos were transferred either to 100 mm glass dishes (VWR, Mississauga, ON) if dechorionated, or 100 mm plastic dishes (VWR) if non- dechorionated, in 35

E3 solution and allowed to develop at 28.5⁰C. Where necessary, embryos were treated with 0.003% PTU to block pigmentation.

Table 2.1 - Injection needle pulling programs for Flaminng/Brown Micropipette Puller

Sutter P-97 Flaming/Brown Micropipette Puller Heat Pull Velocity Time Injecting without chorions 780 60 80 100 Injecting with chorions 720 50 80 130

For morpholino injections, lyophilized morpholinos were obtained from Genetools, LLC (Corvallis, OR) and diluted to 2mM stocks in ddH2O. A list of morpholino sequences is given in Table 2.2. Working concentrations of morpholino were injected into yolk just below the embryo, up to but no later than the 8-cell stage. Hemorrhages were assayed in uninjected control and morpholino-injected (morphant) embryos anaesthetized with 0.4% Tricaine (ethyl 3- aminobenzoate methanesulfonate; Sigma) at 48hpf.

Table 2.2 - Morpholino Oligonucleotide Sequences, Targets and Doses

Name Target Type Sequence Doses Used 126 bp downstream FoxF2b i2e3 Translation Blocker GACCTGAGTTGGTCTTTTTACTCTT 5.4ng of second ATG Exon 1-intron 1 FoxF2b e2i2 Splice Blocker ATTTGCTTAATTATAGCTACCTGAC 0.5ng - 23.2ng splice site FoxF2b Translation Start Upstream First ATG ATAATACAAGCATCACGCATCCGAC 0.5ng - 23.2ng Blocker ATG Translation Start FoxF2b ATG Second ATG AGCTTTCAGTCGTCATCGGTCCAGC 0.5ng - 23.2ng Blocker Exon 1-intron 1 FoxF2a e1i1 Splice Blocker CGTGTAAAAGCGGCTTTACCTGAGA 1.7ng - 7.9ng splice site Translation Start FoxF2a ATG ATG TGGTCATCGCAAATGCTGGATGAAT 0.5ng - 23.2ng Blocker Translation Start FoxC1a ATG CCTGCATGACTGCTCTCCAAAACGG 4ng Blocker Translation Start FoxC1b ATG GCATCGTACCCCTTTCTTCGGTACA 4ng Blocker 36

2.3 Fin Clipping and HotSHOT Genomic DNA Isolation

HotSHOT genomic DNA isolation protocol was adapted from Truett et al. (Truett et al., 2000). For adult embryos, the tip portion of the tail fin was used. Adult fish were anesthetized using 0.4% Tricaine (Sigma) and laid on a cutting surface. Clips were made using a razor blade or scalpel, just anterior of tail fin tip junction. Fin clips were placed in 1X Base solution (see below) and fish were placed into system water to recover. To isolate genomic DNA (gDNA), embryonic or clipped adult tail fin tissue was placed in 50µL of 1X Base Solution (50X stock: 1.25M NaOH, 10mM EDTA

(ethylenediaminetetraacetic acid disodium salt dehydrate), pH12 in ddH2O) and incubated at

95⁰C for 30min. 50µL of 1X Neutralization Solution (50X stock: 2M Tris-HCl, pH5 in ddH2O) was added to neutralize the reaction. Samples were then centrifuged at 3000G for 5-10 min. gDNA was sampled from the top portion of the centrifuged solution to obtain DNA with low levels of undigested embryo contamination.

2.4 TALEN Mutagenesis and Mutation Identification

FoxF2b target sequence for transcription activator-like effector nuclease (TALEN) was chosen using web-based software at http://boglabx.plp.iastate.edu/TALENT/ (website redirects to https://tale-nt.cac.cornell.edu/. Table 2.3 lists the TALEN target sequence and left and right arm recognition sites used. FoxF2b-targeting TALENs were generated using the Golden Gate TALEN cloning kit (Cermak et al., 2011)(Addgene, Cambridge, MA), with the exception that repeat-containing pFUS_A and pFUS_B constructs for target sequences were cloned together into Fokl nuclease-encoding vectors, pCS2TAL3DD (Addgene) for upstream target sequence and pCS2TAL3RR (Addgene) for downstream target sequence;, which also contained the final half-repeat. Thus, only two constructs needed to be injected as opposed requiring additional injection of Fokl nuclease RNA or protein. TALEN constructs were linearized by NotI restriction enzyme (NEB) and TALEN RNA was transcribed from linearized DNA using SP6 reverse transcriptase (Promega).

Table 2.3 - TALEN Target Sequences

Name Sequence FoxF2b Tal1 GTCAAAGAGTAAAAA FoxF2b Tal2 AACCGCCCTACTCCT Target Spacer Sequence GACCAACTCAGGTCTGCGACGCCCGGAGA

TALEN RNA was injected into the cytoplasm of 1-cell stage embryos using methods described for morpholino injections. Approximately 510 pg of TALEN RNA was injected into each embryo. To detect targeted mutagenesis in founders, gDNA was isolated from embryonic or juvenile tissue and analyzed by high-resolution melt (HRM) analysis. A semi-quantitative PCR reaction was set up using 1X (University of Calgary Core DNA Services) and 1-1.5µL of gDNA template in a 10 µL reaction to amplify a 141 bp fragment of DNA. Primer sequences used for all mutant identification methods are listed in Table 2.4. The reaction was run using a DNA Engine Opticon™ System (PTC-200 DNA Engine™ Cycler + CFD-3200 Opticon™ Detector, Bio-Rad, Hercules, CA) to detect fluorescent dye signals incorporated into the product. At the end of the PCR cycles, a melt curve was performed to identify the melting point of the product. Melting temperatures were compared to determine differences due to mismatches introduced through mutagenesis. Potentially mutated founders were raised to sexual maturity and outcrossed to wildtypes, and gDNA of the progeny of these outcrosses were also analyzed by HRM to determine if the mutation was transmitted from the germline of the founders.

2.5 FoxF2b Mutant Sequencing and Genotyping

To determine the specific mutations introduced into the genomic DNA by TALEN mutagenesis, a fragment including the mutation target site was amplified from heterozygous mutant gDNA by a PCR reaction using 0.01 U/µL Q5 Hi-Fi Polymerase (NEB), 1X Q5 reaction buffer (NEB), 0.2mM of each dNTP (Invitrogen), 0.5µM forward and reverse primer oligonucleotides (University of Calgary Core DNA Services). Primer sequences for all mutant genotyping and sequencing reactions are listed in Table 2.4. PCR product was cloned into pCR- Blunt by Zero Blunt cloning (Invitrogen) and transformed into TOP10 OneShot bacterial cultures

(Invitrogen) according to manufacturer instructions. Plasmid was isolated from individual colonies using the GenElute HP Plasmid miniprep kit (Sigma), and the insert was extracted by EcoRI (NEB) digestion, and sequenced using gene-specific primers (see Table 2.4) at the University of Calgary Core Sequencing Facility. Since pCRII vector accepts only one DNA fragment, 50% of colonies contained wildtype alleles and 50% contained mutant alleles from the original heterozygous gDNA template.

Table 2.4 – TALEN-generated Mutant Genotyping Assays and Primers

Assay Forward Primer Tm (⁰C) Reverse Primer Tm Product Size (bp) HRM TGCATCCAGGACTGATGAAC 60 TCTGAATTGCCATCACAATGA 58 141 Sequencing/Cloning CGAGGCGTTTATCCTCAGTT 60 CCTAAATCCTCGGGGTCTTC 62 585 Ca22 Digest (HinfI) CGAGGCGTTTATCCTCAGTT 60 TGCCGAACAGAGTTCTTCCA 60 WT: 443, Ca22: 269 + 175 Ca23 Band Shift GCATAGCGCAGTGTCAAAGA 60 GGTTGGTGCGCTCTGAATTG 62 WT: 112, Ca23: 98

Mutant transcripts were also sequenced to determine whether genomic mutations were being transcribed. RNA was obtained from homozygous mutants, and synthesized into cDNA, as described above. Mutant sequence was amplified by PCR, purified, and sequenced using gene- specific primers. Sequencing showed that the FoxF2bCa22 mutant allele introduced a HinfI restriction enzyme recognition site into the sequence. To genotype for FoxF2bCa22, the target sequence was amplified from gDNA by PCR run with 0.1 U/µL Phusion® High Fidelity polymerase (NEB), 1X HF buffer (NEB), 200µM of each dNTP (Invitrogen), and 0.5µM forward and reverse primer oligonucleotides to create a 443 bp product (Table 2.4 lists mutant genotyping primers). PCR product was digested in a reaction of 1-2 U/µL HinfI (NEB) with 1X CutSmart buffer (NEB) at 37⁰C for a minimum of 2h. Digested PCR product fragments were separated by electrophoresis in 1% agarose (Invitrogen) in TBE buffer (90mM Tris, 90mM boric acid, 1.6mM EDTA). WT showed single band at 443 bp. Mutated product showed two bands of 269 and 175 bp. The FoxF2bCa23 mutant allele was genotyped by resolving a 14 bp difference between mutant and wildtype PCR fragments separated by electrophoresis in 2% agarose in TBE. Similar PCR conditions were used as for FoxF2bCa22 genotyping. 39

2.6 Template Generation and In Situ Hybridization Probe Synthesis

Templates for synthesis of in situ hybridization (ISH) probes were amplified from cDNA using PCR. A list of probes that were synthesized by this method and their primer sequences are shown in Table 2.5. Probe-specific primers were synthesized (University of Calgary Core DNA Services) based on annotated NCBI mRNA sequence entries or previously published probe primer sequences. PCR amplifications were performed using cDNA as template with 0.1U/µL Phusion® High Fidelity polymerase (NEB), 1x HF buffer (NEB), 200µM of each dNTP (Invitrogen), 0.5µM primer oligonucleotides (University of Calgary Core DNA Services) in a 50µL reaction. 5µL of product was run through gel electrophoresis to verify amplicon was the correct size. The remainder of PCR product was purified using QIAquick® PCR Purification kit (Qiagen) or Sigma GenElute PCR Clean-Up kit (Sigma) and used for template in reverse transcription probe synthesis. In a few instances, probes were synthesized from templates cloned into vectors already containing the T3 reverse transcription promotor (Table 2.6). Briefly, bacterial vector stocks were grown up and vector DNA was purified by GenElute HP Plasmid Miniprep kit (Sigma). Purified vector DNA was then linearized to expose template sequence by an appropriate restriction enzyme. Linearized vector DNA template size was verified by gel electrophoresis, and then used for template in reverse transcription probe synthesis (Thisse & Thisse, 2008). DNA template concentration needed for RNA probe synthesis was determined by template type: 0.1 µg was required for every 200 bp of PCR-generated template and 2 µg were required for every 200 bp of vector-generated template. Template was incubated with 20 U/µL T7 reverse transcriptase (Promega, Madison, WI) or 14U/µL SP6 reverse transcriptase (Promega), 1X transcription buffer (Promega), 20 mM DTT (dithiothriotol; Promega), 1X digoxigenin (DIG) RNA labelling mix (Roche, Laval, PQ) or 1X dinitrophenol (DNP) RNA labelling mix (DNP-11-UTP (Perkin-Elmer, Boston, MA) + NTPs (NEB)), and 20 U/µL RNAsin (Promega) at 37⁰C for 2-4h. 1 U/µL of RNAse-free DNAse (Promega) was added and incubated at 37⁰C for 15min to eliminate left-over DNA template. The reaction was stopped by adding 0.02M EDTA (pH8.0). Synthesized probe RNA was precipitated with 0.1 M LiCl and 100% 40

Table 2.5 - Primers for ISH probes generated by PCR

Product/Probe Gene Forward Primer Reverse Primer Size (bp) TAATACGACTCACTATAGGAGCATGGCCTTC col4a2 CAGGCTCGTGTCTTGAGGATT 532 TTTGGTTCA AATTTAATACGACTCACTATAGGTAGTGGTA Ctgfa CTCCCCAAGTAACCGTCGTA 595 CAGCCGGAAA AATTTAATACGACTCACTATAGGCCATCCGC fabp11b (Flynn et al., 2009) AACACTTTGTGCTATTATCTGTC 373 AAGGCTCATAG TGTAATACGACTCACTATACCCTTGACACACT foxf2a CGAATTGATAGGTGGTCGCG 1468 CCCTGGTGA AATTTAATACGACTCACTATAGGCGGCTGTG foxf2b TTCGGCACAACTTGTCTCTG 577 AATTGCTTGTAA AATTTAATACGACTCACTATAGGAGGCATCA foxq1a CCTCCTTCGAGCGGAAAGTT 588 TCGTTTGAGGCA AATTTAATACGACTCACTATAGGACTGGCAG foxq1b ATGGAGACTGCTCAGCCAAC 592 AAAGTGAGGAGC TAATACGACTCACTATAGGGAGATAGCCTTG netrin1b TACTCAACCCAGTGCAAGAAAAT 678 CTCATGTCTCTG TAATACGACTCACTATAGGGCGCCGTTGTTT netrin5 TTCCTCGCCAGACTTTGACT 463 CTTCTTCTC AATTTAATACGACTCACTATAGGAGAAAGG nkx2.2a (Yang et al., 2010) CTGCATCAAATGCTCCAGAA 958 GTCAAGCTGCAAA AATTTAATACGACTCACTATAGGTCCCAATT nkx2.2b (Yang et al., 2010) ACCTTCACTGTCGCAGTCCT 956 GTGACGTCATTG AATTTAATACGACTCACTATAGGTTGCGCTA nkx2.9 (Yang et al., 2010) CAGCCACCAAGTGCTGTTTA 894 AGTCCCGAAATA TAATACGACTCACTATAGGGGCTCCTCAAGT nkx3.2 AAGAAAGTGGCGGTGAAGGT 560 CCAGCAAAT TGTAATACGACTCACTATACCGTACGGGTCT pdgfaa CGTGGATGGTGCGTCAAATG 643 TACAGACGG TGTAATACGACTCACTATACCATGGTCCTCCA pdgfba TGGCACGGTGTTCAAAGTGA 643 CGGGTATG AATTTAATACGACTCACTATAGGCACAGGCT pdgfrα CCCCATTCCCTGAAGTGGAC 571 GAGAACGGCTTA AATTTAATACGACTCACTATAGGGAAGCTCT pdgfrβ (Wang et al., 2014) TCCAGACTAATGTCACCTACAACAG 1846 CCTCTACTTCTGGACTT TAATACGACTCACTATAGGGCAAGCAGGTCC pim1 (Yin et al., 2012) GCAGAGCCCATCAGTAGCTC 1418 CATCATCTG AATTTAATACGACTCACTATAGGTCCCAAAA ptch2 TCTGCAAGCCACTTTTGATG 867 GAGTGGAGATGG AATTTAATACGACTCACTATAGGTGACCGCT shha CTACGGCAGAAGAAGACAT 855 ATCATCAACAA TGTAATACGACTCACTATACCCATTGCATGTC snai2 CCTATTATGGAGGTCAATTCGGG 641 TGAGTGCTGG TGTAATACGACTCACTATACCCATGGCAATG tbx15 GACAAACAAGAAGCGGAAGC 427 TAATACTGCTG AATTTAATACGACTCACTATAGGTAGGCTGT tbx20 AGAGTGATACCATTTACCTTTCG 501 GCATTTGAT AATTTAATACGACTCACTATAGGAAGACGCT urp2 TGAATCTCCCGAAAACAAAGAAC 543 GCAAGATGGAAAG

Table 2.6 - Probes generated from cloned sequences in vectors

Gene Source Linearizing Enzyme Reverse Transcriptase hic1 Open Biosystems NotI T3 tbx18 Open Biosystems AfeI T3 gli2a Phil Ingham Lab BamHI T7

hand2 Gift from Deborah Yelon

EtOH at -80⁰C overnight, and once precipitated, centrifuged at 0⁰C at 13000 rpm for 30 min.

Supernatant was removed and resultant RNA pellet was dissolved in 20 µL of RNAse-free H2O. 1 µL was run on an agarose gel to check for presence of RNA while the remainder was mixed with hybridization buffer (50% formamide, 5x SSC (20x SSC stock is a 3M sodium chloride, 340mM sodium citrate solution), 5mg/mL torula yeast RNA, 50µg/mL heparin, 0.1% Tween-20,

H2O) for storage at -20⁰C.

2.7 Wholemount In Situ Hybridization

For wholemount in situ hybridization, samples (comprised of either embryos or dissected brains) were fixed in 4% (w/v) paraformaldehyde (PFA) for 2 hours and placed in 100% methanol for long-term storage at -20⁰C. Fixed samples stored in methanol were used for in situ hybridization using the method of Lauter (Lauter et al., 2011). Samples were permeabilized in Proteinase K (1mg/mL stock; Promega) at varying times and concentrations depending on age (Table 2.7). Embryos were then pre-hybridized for several hours in hybridization buffer at 60⁰C and then incubated overnight in hybridization buffer with a 1:100 dilution of DIG-labelled RNA probe (or DNP-labelled RNA probe for double fluorescent in situ hybridization) with 5% dextran sulfate overnight at 60⁰C. Embryos were then washed, in turn, in modified hybridization buffer (50% formaldehyde + 2X SSC + 0.1% Tween), 2X SSC and 0.2X SSC at 60⁰C before PBT (0.1% Tween-20 in 1X PBS

(phosphate buffered saline, 10X stock: 1.37 M NaCl, 26.8 mM KCl, 101.4 mM Na2HPO4, 17.6 mM KH2PO4)) washes at room temp. Samples were blocked for 1 hour with 10% normal sheep serum (NSS) in PBT and then incubated for 2h at room temperature or overnight at 4⁰C with anti-DIG mouse Fab antibodies conjugated with alkaline phosphatase at a 1:5000 dilution in 10% NSS to detect DIG-labelled probes. After extensive washes with PBT and equilibration in NTT 42

(100mM Tris-HCl (pH9.5), 100 mM NaCl, 0.1% Tween-20), location of bound RNA probe was detected using a substrate mix of 0.45mg/mL 4-nitro blue tetrazolium chloride (NBT; Roche) and 0.175 mg/mL 5-bromo-4-chloro-3-indolyl-phosphate (BCIP; Roche) in NTT , which precipitates as an indigo-blue colour in the presence of alkaline phosphatase. Reactions were left to occur in the dark until blue signal developed. Further alkaline phosphatase activity was stopped by 15 min incubation in 4% PFA.

Table 2.7 - Proteinase K Incubation Concentrations and Times for ISH Permeabilization

[Proteinase Incubation Stage K] Time Early Somitogenesis (1s-13s) 10µg/mL 1 min Late Somitogenesis (14s-22s) 10µg/mL 5 min 24hpf 10µg/mL 15-25 min 36hpf 15µg/mL 25 min 48hpf 40µg/mL 30 min 72hpf 75µg/mL 30 min 96hpf 120µg/mL 30 min 120hpf 165µg/mL 30 min

Whole-mount in situ hybridization samples were mounted either in 3% methylcellulose or on a 1% agar plate in PBT and imaged using a Stemi SV11 microscope and an Axiocam HRc digital camera (Carl Zeiss, Inc., Thornwood, NY), under brightfield or fluorescent illumination. Fluorescent samples were occasionally also mounted in 0.8% low-melt agarose and imaged on a Zeiss LSM 700 Scanning confocal microscope (Carl Zeiss, Inc.). For sectioning, embryos were embedded in JB4 medium (Polysciences, Warrington, PA) as per manufacturer’s instructions and sectioned at 7 µM using a Leica microtome.

2.8 Immunostaining

Antibodies used for experiments are listed in Table 2.8. Samples were fixed in 4% PFA, and depending on the primary antibody, different permeabilization methods were used. In the methanol/acetone method, samples were dehydrated by a series of methanol dilution washes and

stored in 100% methanol at -20⁰C for an overnight length at minimum. Samples were then treated with acetone at -20⁰C for 30 min. In the Triton method, samples were washed in PBT after fixation, and stored at maximum 1 week in PBT at 4⁰C. Samples were then treated with 0.5% Triton in 1X PBS for 30 min. Individual permeabilization steps varied slightly depending on the antibody and sample tissue composition. In both cases, after permeabilization, samples were blocked for a minimum of 1 hour at room temperature in 10% NSS in PBT. Samples were incubated with primary antibodies at appropriate dilutions for up to 2 days at 4⁰C. Several washes in 1% NSS in PBT removed remaining primary antibody, and samples were incubated in appropriate fluorescent secondary antibodies for up to 1 day at room temperature. Excess secondary antibody was removed in several PBT washes and samples were imaged on appropriate fluorescent microscopes.

Table 2.8 - List of antibodies used

Antibody Type Source Used For Dilution GFP Primary Clontech Immunostain 1:500 mCherry Primary Clontech Immunostain 1:500 FoxF2 Primary Abcam Western 1:1000 FoxF, epitope 1 Primary Abmart Immunostain 1:100 (fish), 1:1000 (cells) FoxF, epitope 2 Primary Abmart Immunostain 1:100 (fish), 1:1000 (cells) FoxF, epitope 3 Primary Abmart Immunostain 1:100 (fish), 1:1000 (cells) FoxF, epitope 4 Primary Abmart Immunostain and Western Immunostain: 1:100 (fish), 1:1000 (cells), WB: 1:1000 FoxF, epitope 5 Primary Abmart Immunostain 1:100 (fish), 1:1000 (cells) FoxF, epitope 6 Primary Abmart Immunostain 1:100 (fish), 1:1000 (cells) β-tubulin Primary DSHB Western 1:5000 Alexa 488 Donkey anti-mouse Secondary Molecular probes Immunostain 1:500 Alexa 555 Goat anti-rabbit Secondary Molecular probes Immunostain 1:500 Biotinylated anti-mouse Secondary Vectastain Immunostain and Western Immunostain: 1:200, Western: 1:5000 Biotinylated anti-rabbit Secondary Sigma Immunostain and Western Immunostain: 1:200, Western: 1:5000

Primary antibodies were detected using biotinylated secondary antibodies to facilitate colorometric reactions. 3,3′-diaminobenzidine (DAB; Vector Laboratories Inc., Burlingame, CA) was used as a substrate to produce a brown precipitate in the presence of peroxidase. After biotinylated secondary antibody incubation, samples were treated with a streptavidin/biotinylated peroxidase mixture to link peroxidase to secondary antibodies. Samples were then exposed to DAB substrate and hydrogen peroxide to facilitate deposition of precipitate. Stained samples were then embedded using the JB4 embedding kit (Polysciences, Inc.) and sectioned at 7 µm

using a Leica microtome. Slides of samples were visualized using a Leica DM R microscope and imaged with a MagnaFire S60800 CCD microscope camera (Optronics).

2.9 Staining, Embedding and Imaging for Transmission Electron Microscopy

Tissues for transmission electron microscopy (TEM) imaging were fixed in 2% glutaraldehyde (w/v)(Polysciences, Inc.) in 0.1M sodium cacodylate (pH7.4; Sigma) for 1 hour at room temperature, then rinsed three times for five minutes each in 0.1M sodium cacodylate. Tissue was then post-fixed in 1% osmium tetroxide (w/v) (Polysciences, Inc.) in 0.1M sodium cacodylate for 1 hour on ice, followed three rinses for five minutes each in 0.1M sodium cacodylate. Dehydration was performed by a series of ethanol (EtOH) washes, from 25% EtOH (v/v) to 50% EtOH to 70% EtOH for five minutes each, then two rinses in 95% EtOH for 5 minutes each and finally three 30 min washes in 100% EtOH. Tissues were then incubated three times in Propylene oxide (EM Sciences) washes for a minimum of 1 hour each. Tissues were infiltrated by incubation with Propylene oxide: EMBED 812 resin (EM Sciences) at 3:1 ratio for minimum 1 hour, then a 1:1 ratio for minimum 1 hour, and then a 1:3 ratio overnight at room temperature. Tissues were incubated 100% resin overnight at room temperature, and the following day, a polymerizing EMBED 812 resin was mixed as per manufacturer’s instructions. Samples were positioned as desired in the polymerizing resin in a rubber mold (Polysciences, Inc.) and blocks were left to polymerize at 60⁰C until blocks were hardened. Blocks were sectioned at 80 nm, mounted onto grids, stained with 3% Uranyl acetate (w/v) (EM Sciences) for 10 min, followed by staining in Reynolds lead citrate (0.08M lead nitrate, 0.12M sodium citrate, pH12.0) for 10 min. Samples were visualized using a Hitachi H-7650 Transmission Electron Microscope and photographed using a digital camera.

2.10 Confocal Microscopy

A Zeiss LSM 700 inverted laser scanning confocal microscope was used for imaging of live fluorescent transgene expression, fixed antibody stained embryos and fluorescent dye staining. In addition to EGFP and mCherry, other fluorophores used include Alexa 488, Alexa 555, rhodamine B and DAPI. Images were taken at 10x, 20x, and 40x magnifications. 40x 45

images were attained with an oil immersion lens. Excitation and emission wavelengths for fluorophores are outlined in Table 2.9. Samples were mounted in 0.8-2% low-melt agarose dissolved in E3 solution, and melted on glass bottom cell culture plates. Live samples could be extracted from set mounting agarose after imaging for further growth, fixation or genotyping depending on need.

Table 2.9 – Fluorophore Excitation and Emission Wavelengths

Fluorophore Excitation Wavelength Used (nm) Emission Wavelength (nm) EGFP 488 509 mCherry 555 610 Alexa 488 488 519 Alexa 555 555 565 DAPI 350 470 Rhodamine 555 625

2.11 DAPI-Dextran Injections

To assay vessel permeability, 4dpf embryos anaesthetized with 0.4% Tricaine (Sigma) were injected with 12nL of 1% dextran-rhodamine (10,000 MW, Molecular Probes, Eugene, OR) and 0.7 mg/mL DAPI (350 Da) into the common cardinal vein, as described previously (Tam et al., 2012). After 30 min of circulation, zebrafish were imaged dorsally on the confocal microscope.

2.12 Zebrafish Brain Dissections

Zebrafish to be dissected were anesthetized in Tricaine (Sigma) and were mounted on a gel plate with dissection pins. Scissors were used to sever the brain stem immediately posterior to the hindbrain. An incision was then made towards the anterior from the initial cut, slightly into the dorsal skull plate. Forceps were used to pull back the skull plate. The optic stalks were cut and eyes were removed. Orbital bones were pulled back with forceps, and ventral nerve connections were severed. Olfactory bulb bones were pulled forward, and brains were removed,

imaged on a Stemi SV11 microscope and an Axiocam HRc digital camera (Carl Zeiss, Inc.), and fixed in 4% PFA overnight at 4⁰C.

2.13 Brain Tissue Clearing

Dissected zebrafish brains were cleared by one of two methods. With ClearT2 (Kuwajima et al., 2013), brains were washed in a 25% formamide (v/v)/10% polyethylene glycol (PEG)(w/v) solution for 1 hour, then a 50% formamide/20% PEG solution for 1 hour, and again in the second solution overnight at room temperature. Brains were then imaged by confocal microscopy. Using the CLARITY method (Cronan et al., 2015), brains were incubated for 3 days at 4⁰C in a hydrogel solution (4% PFA, 1% acrylamide (w/v), 0.05% bisacrylamide (w/v), 0.25% photoinitiator 2,20-Azobis[2-(2-imidazolin-2-yl)propane]dihydrochloride (Wako Chemicals, Richmond, VA) in 1X PBS). Hydrogel solutions were overlaid with mineral oil and polymerized for 3 hour at 37⁰C. Brains were removed from excess hydrogel and incubated in 4% SDS (w/v) in 200mM boric acid (pH8.5) at 37⁰C with shaking for 10-14 days, changing the SDS solution every other day until tissue was cleared. After clearing, brains were washed twice over one day in PBS with 0.1% Triton at 37⁰C. For imaging on the confocal microscope, brains were mounted in 2,2'-thiodiethanol with 1% DABCO (1,4-diazabicyclo[2.2.2]octane)(w/v).

2.14 Gateway BP and LR Cloning of full length FoxF2

FoxF2b coding sequence was amplified from cDNA by PCR reaction using 0.1U Platinum Taq HiFi polymerase (Invitrogen), 1X High Fidelity buffer (Invitrogen), 1.6 mM dNTPs (Invitrogen), 1mM MgSO4 (Invitrogen), 0.4 µM gene-specific primers (University of Calgary Core DNA Services, forward primer plus attb1: GGGGACAAGTTTGTACAAAAAAGCAGGCTTCGAAGGAGATAGAACCAT GACGACTGAAAGCTCTCAC, reverse primer plus attb2: GGGGACCACTTTGTACAAGAAAGCTGGGTTCTACAT CACGCAAGGTTTGA) and PCR purified to remove remaining primers and dNTPs. BP cloning was performed by incubating 25 fmol of purified PCR product with 25 fmol of pDONR221 entry vector and 1 µL BP Clonase II 47

enzyme mix (Invitrogen) in a 5µL reaction overnight at 25⁰C. Reaction was terminated by a 10 min incubation with 0.17 mg/mL Proteinase K (Invitrogen) at 37⁰C. Constructs were transformed into TOP10 OneShot (Invitrogen) cells and miniprepped (Sigma GenElute kit). Isolated vectors were sent for sequencing with M13F and M13R(17) primers (University of Calgary Core DNA Services). Expression vectors for transfection were generated by incubating 300 ng of FoxF2b BP clones in pDONR221 with 300ng of pCSmCherryDest vector in an 8µL reaction containing 2 µL LR Clonase II Plus enzyme mix. Reactions were incubated overnight at 25⁰C and terminated by a 10 min incubation with 0.17 mg/mL Proteinase K at 37⁰C. Resulting miniprepped transformants were verified by restriction digest with PvuI and sequencing.

2.15 Cell Culture and Transfection

CRL-1651 Cos7 cells (gift from Cross lab, originally from ATCC, Manassas, VA) were thawed and maintained at 37⁰C with 5% CO2 in complete Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with L-Glutamine (Invitrogen), 0.1mM MEM Non-Essential Amino Acids Solution (Invitrogen), 1X Penicillin-Streptomycin (Invitrogen), and 10% Fetal Bovine Serum (Invitrogen). To subculture cells, culture medium was removed and cells were rinsed with PBS and incubated in 0.25% (w/v) Trypsin (Invitrogen)/1mM EDTA for 5-15 minutes until cell layer was dispersed. Trypsin reaction was terminated by addition of complete DMEM. Cells were gently suspended and aspirated in complete DMEM and subcultured in a 1:4 to 1:8 ratio from original confluency. Subculturing was performed 2-3 times a week. To express FoxF2b, Cos7 cells were grown in incomplete DMEM (no Pen/Strep) for a day prior to transfection to obtain 90-95% confluency. 5 µg of FoxF2b construct or empty vector was diluted in 0.5mL of Opti-MEM, and 12.5 µL of Lipofectamine® 2000 (Invitrogen) was diluted in a separate 0.5 mL of Opti-MEM to obtain a ratio of 1:3 DNA(µg):Lipofectamine (µL). After a 5 min incubation at room temperature, DNA and Lipofectamine solutions were mixed together and incubated for 20 min at room temperature. A negative control with no DNA construct was also prepared. Complexes were added to cells, and cultures were incubated at 37⁰C with CO2 for 18-48 hours prior to examining transgene expression. 48

2.16 Western Blotting

Embryos for protein were first deyolked by repetitive pipetting in deyolking buffer

(55mM NaCl, 1.8mM KCl, 1.25mM NaHCO3) with 0.3 mM phenylmethylsulfonylfluoride (PMSF). Yolk proteins were separated from embryos by repeated centrifugation at 1000 rpm and washes in Wash buffer (111mM NaCl, 3.5mM KCl, 2.7mMCaCl2, 10mM Tris-HCl pH 8.5). Protein was extracted from between 50 and 150 embryos, depending on stage and size, by homogenizing in sodium dodecyl sulfate (SDS) sample buffer (63mM Tris-HCl pH 6.8, 10% glycerol, 5% β-mercaptoethanol, 3.5% SDS) with a microfuge pestle and subsequent boiling for 5 min. Protein samples were mixed 1:1 with 2X SDS loading buffer (1X Stacking buffer (0.125M Tris-HCl pH 6.8, 2.5% SDS), 20% glycerol, 1% SDS (w/v), 2% 2-mercaptoethanol, 0.001% Bromophenol blue (w/v)) for a final concentration of 1X loading buffer. Samples were loaded into a 10 or 12% SDS-polyacrylamide gel and run at 160V for 1 hour. Proteins were then transferred to a PVDF membrane (Merck Millipore, Billerica, MA) at 100V for 1 hour. Membranes were blocked in 1X TBST (20mM Tris-HCl pH 7.6, 120mM NaCl, 0.1% Tween 20) with 5% skim milk powder for a minimum of 1 hour. Membranes were subsequently incubated in primary and secondary antibodies (Table 2.8) diluted in TBST with 5% skim milk for a minimum of 1 hour each, separated by TBST washes. Secondary antibodies are labelled with biotin, which were subsequently detected with HRP-labeled streptavidin. Transferred blots were detected using the Amersham ECL Prime Western Blotting Detection Kit (GE Healthcare, Buckinghamshire, UK) and imaged using a Fuji LAS 3000 imaging system.

2.17 Drug Treatments

Cyclopamine (Toronto Research Chemicals Inc., Toronto, ON) was diluted in dimethyl sulfoxide (DMSO) to a stock concentration of 10mM in 1% DMSO. Cyclopamine was applied at a working concentration of 100µM in E3 buffer. Controls were treated with an equal volume of 1% DMSO alone. Embryos collected for microarray analysis were treated from 4-30hpf. Hemorrhage was scored at 48hpf.

2.18 Expression Profiling

Differential gene expression between 30hpf igu mutants and their wild-type siblings or 30hpf cyclopamine-treated and their DMSO controls was compared. Microarray expression profiling was performed by NimbleGen Systems Inc. (Madison WI). cDNA was labeled with (single-colour) Cy3 fluorescent probes and hybridized to NimbleGen’s maskless photolithographic in situ synthesized DNA oligonucleotide microarrays (NimbleGen Gene Expression Danio Rerio 385k Array, 2006-05-16_Zv6), which analyzes approximately 38500 transcripts. Normalized gene expression values (or CALLS) were generated using the RMA (Robust Multichip Average) algorithm. Three biological replicates were performed for each of igu mutants, their wild type siblings, cyclopamine-treated embryos or DMSO-treated embryos. We reported gene expression changes as Z-ratios instead of globally normalized fold change values in order to more accurately compare gene expression changes, as described by Cheadle et al. (Cheadle et al., 2003). For reference, a range of Z-ratios from ±15.0 to ±2.0 corresponds to a range of fold-changes from ±27.0 to ±1.7. Data normalization and statistical analyses were performed on the Normalized_Calls data file. The Z_Ratios statistics for microarray experiments were performed using an EXCEL template as per Cheadle (Cheadle et al., 2003). Briefly, a Log base 10 transformation was applied to the “normalized call values” of the Roche/NimbleGen files. A grand average of all arrays was also obtained and used to normalize each array. Z-score was then calculated by standardization (with subtraction of the average and subsequently divided by the standard deviation (SD) of each array). Changes (Z- Ratio’s) in igu mutants or cyclopamine-treated embryos were calculated for each gene within a replicate according to the formula described in Cheadle (Cheadle et al., 2003), and subsequently averaged between all three replicates. Genes with absolute Z-Ratio values of greater than or equal to 1.96 were marked as significantly different (P < 0.05). The heat map was then generated using Gene-E software (http://www.broadinstitute.org/cancer/software/GENE-E/). This expression data has been deposited to GEO (GSE48335).

2.19 RNAseq expression profiling

RNAseq-based expression profiling was performed at the Lunenfeld-Tanenbaum Research Institute (Toronto, ON). Two cDNA libraries of pooled igu mutants (~75-100 embryos) and their wild-type siblings (~75-100 embryos) at 30hpf were prepared using 1µg of high quality total RNA and the Illumina TruSeq RNA Sample Prep Kit v2 (Illumina Inc., San Diego CA). The generated barcoded cDNA library had an average fragment size of 350-400bp. The quality of these barcoded libraries were checked with an Agilent Bioanalyzer, quantified by qPCR using KAPA SYBR FAST Universal 2X qPCR Master Mix (Kapa Biosystems, Wilmington MA), and detected using an Applied Biosystems 7900HT. The sequencing libraries were then loaded on a flow-cell for cluster generation using an Illumina c-Bot and TruSeq PE Cluster Kit v3 reagents (Illumina Inc., San Diego CA). Sequencing was carried out on a HiSeq2000 with TruSeq SBS Kit v3 (pair-ended 200 cycles, Illumina Inc., San Diego CA). The real-time base call (.bcl) files were converted to fastq files using CASAVA 1.8.2. Illumina reads in fastq format were aligned to the Ensembl Danio rerio reference genome (Zv9.75) by using the RNAseq analysis utilities tophat2 (version: tophat-2.0.11.Linux_x86_64), bowtie2 (version: bowtie2-2.2.0) and samtools (version: samtools- 0.1.19.0) to generate SAM alignment files, and by using the python script htseq-count (part of 'HTSeq' framework, version 0.6.0. http://www- huber.embl.de/users/anders/HTSeq/doc/count.html) to obtain the raw counts, as described previously by Anders (Anders et al., 2013). Read counts of 33737 gene models were used with the edgeR package (Robinson et al., 2010) to generate cpm (count per million) values for all sequencing libraries. Gene models with cpm values >= 1 were used to calculate the normalized cpm values, P- value and False Discovery Rate (or FDR) to identify differentially expressed genes. An estimated Biological Dispersion of value 0.2 (or Biological Coefficient of Variation BCV = ~0.4472) was used to account for the lack of sample replicates in our pooled samples comparison. A total of 17887 gene models that passed filtering were used, and 193 annotated differentially expressed genes passed an FDR cutoff of >=7.5%.

Chapter Three: Comparative Analysis of Genes Regulated by DZIP-Iguana and Hedgehog in Zebrafish

This chapter is adapted from the publication: “Arnold CR, Lamont RE, Walker JT, Spice PJ, Ho

C-Y, Chan C-K, Childs SJ. 2014. Comparative analysis of genes regulated by Dzip-iguana and

Hedgehog in zebrafish. Dev Dyn. 244(2):211-23”, with permission of the publisher.

Ryan Lamont prepared the RNA for the microarray. Chi-Yip Ho and Kin Chan performed the bioinformatics for the microarray and RNAseq analysis. 52

Abstract The zebrafish genetic mutant iguana (igu) has defects in the ciliary basal body protein Dzip1, causing improper cilia formation. Dzip1 also interacts with the downstream transcriptional activators of Hedgehog (Hh), the Gli proteins, and thus Hh signaling is disrupted in igu mutants. Hh governs a wide range of developmental processes, including stabilizing developing blood vessels to prevent hemorrhage. Using igu mutant embryos and embryos treated with the Hh pathway antagonist cyclopamine, we conducted a microarray to determine genes involved in Hh signaling mediating vascular stability. I identified 40 genes with significantly altered expression in both igu mutants and cyclopamine-treated embryos. For a subset of these, I used in situ hybridization to determine localization during embryonic development and confirm the expression changes seen on the array. Through comparing gene expression changes in a genetic model of vascular instability with a chemical inhibition of Hh signaling, I identified a novel set of differentially expressed genes with potential roles in vascular stabilization.

3.1 Introduction

Sonic hedgehog regulates VEGF and angiopoietin during vascular development (Pola et al., 2001), but additional effectors of vascular stabilization downstream of Shh remain unknown. In order to focus on genes that are specifically regulated by Hh during vascular stabilization, we compared changes in gene expression with two different methods of Shh disruption, igu mutants or embryos with chemical disruption of Hh signaling. The small molecule cyclopamine is a Hh pathway antagonist that functions by binding Smo, thus preventing its translocation to the cilia. Our lab showed that cyclopamine inhibition of Hh signaling results in a similar hemorrhage phenotype to the one observed in igu mutants (Lamont et al., 2010), suggesting a potentially shared molecular mechanism. Like many molecules, cyclopamine has off-target effects. On the other hand, igu mutants have a disruption in Dzip1, a ciliary basal body protein, and reduced to absent cilia. While this disrupts Shh signalling, igu mutants would also be expected to have disruptions in other signalling pathways dependent on cilia. However by cross-comparing datasets for igu mutants (vs. wild type siblings) and cyclopamine-treated embryos (vs. vehicle treated embryos), we hypothesized that we could obtain an enriched dataset of genes regulated by Shh that are involved in vascular stabilization. Here we report RNA expression profiles in order to identify novel potential downstream effectors of Dzip1 and Hh in vascular stabilization. We present analysis and verification of this microarray data through statistical, in vitro and in situ techniques.

3.2 Results

3.2.1 Gene Expression Profile of Iguana Mutants Partially Overlaps With Hh Deficient Embryos

igu mutants or embryos treated with cyclopamine demonstrate intracranial hemorrhage starting between 40 and 48hours post fertilization (hpf) (Fig. 3.1 A–C) (Lamont et al., 2010), thus we collected total RNA and sampled gene expression in igu mutant and cyclopamine-treated zebrafish embryos, and their respective sibling or vehicle controls at 30hpf, a window where vascular stabilization is being established (Lamont et al., 2010). Three biologically independent samples each of igu mutants, their wild type siblings (heterozygotes and wildtype), 100 µM cyclopamine treated (from 4 to 30hpf) and 1% DMSO-treated embryos each were hybridized and analyzed by Nimblegen using a gene expression microarray. The data showed 523 genes had statistically significant differential gene expression in igu mutants versus their wild-type siblings, while 429 genes had differential gene expression in cyclopamine-treated embryos versus DMSO- controls (Fig. 3.1D). In total, approximately 900 genes were differentially expressed, comprising approximately 3–4.5% of the estimated genome.

3.2.2 A Subset of Developmentally Expressed Genes is Regulated by Both Hh Signaling and Iguana

Comparing gene lists for igu mutants and cyclopamine-treated embryos, we identified 53 genes with differential expression in both conditions. Of these 53 gene entries, 13 have been removed and several have been updated to new transcript identifiers between the Zv6 genome release used on the microarray and the current Zv9 genome version, reducing the list to 40 genes (Table 3.1; Fig. 3.1E). Most of these genes were up- or downregulated in the same direction in igu mutants and cyclopamine-treated embryos, with the exception of netrin 5 and CPAMD8 (B8JL56_DANRE), which were up-regulated in igu embryos and down-regulated in cyclopamine-treated embryos. It is surprising that although each experimental condition had

Figure 3.1– Summary of genes with significantly altered expression in igu mutants and cyclopamine-treated embryos

A–C: Phenotypes of wild-type (A), igu mutant (B) and cyclopamine-treated wild-type (C) embryos at 48hpf. Arrows point to hemorrhages. Scale bars = 100 µm. D: Venn diagram of total genes significantly up- or down-regulated in treatment group as compared to their control. The purple circle represents the comparison of igu mutants and wild-type siblings and the blue circle represents the comparison of cyclopamine-treated to DMSO-treated siblings. The overlap represents 40 annotated genes that were changed in both groups. E: Heat map of 40 genes differently expressed in both igu mutant and cyclopamine-treated embryos. Green indicates downregulation, and red indicates up-regulation. Values on scale bar indicate range of fold changes from 6.2-fold down-regulated to 2.1-fold up-regulated. igu, iguana mutant; cyc, cyclopamine-treated. Adapted from (Arnold et al., 2015) with permission of the publisher.

Table 3.1- 40 genes with significantly altered gene expression patterns in igu mutant and cyclopamine-treated zebrafish embryos as reported by microarray

M/W indicates ratio of gene expression in mutant over wild type, while C/D indicates ratio of gene expression in cyclopamine-treated embryos over control. Negative sign (-) indicates downregulation by fold change. Adapted from (Arnold et al., 2015) with permission of the publisher.

hundreds of hits, only 40 of these were differentially expressed in both conditions when we limit our search to hits with Z-ratio magnitudes larger than 1.96. Although we have identified 40 genes with differential expression, it is not clear if these genes represent direct targets of the Gli effector transcription factors or are other indirect targets. Thus, we took the approach used by Xu et al. (Xu et al., 2006) to identify putative Gli transcription factor binding sites in conserved regions upstream and downstream of the 40 differentially expressed hits. We found the majority of genes in this list of 40 had at least one putative Gli binding site (Table 3.2). The genes nkx2.2a, nkx2.2b, nkx2.9, and ptch2 have known consensus Gli binding sites (Xu et al., 2006), and we found consensus sites in aqp8a.1, foxf2a, and wrd65. The FoxF2a consensus Gli site is located downstream of the gene, however as foxf2a is part of a cluster of Fox genes that include foxc1a and foxq1a, these may have a shared enhancer. Both putative Gli consensus sites in the wrd65 gene are found in intronic sequences. Although the functionality of these and other nonconsensus Gli binding sites is beyond the scope of this study, it provides a potential starting point for the investigation of several novel genes potentially regulated by Hh signaling.

3.2.3 Independent Validation of Microarray Gene Expression Changes by In Situ Hybridization

Many of the differentially expressed genes have no documented expression pattern in zebrafish. Thus, we selected a subset of genes to analyze based on known developmental roles, roles in vascular mural cells in other species, or genes with potentially interesting functions such as transcription factors, receptor ligands, and kinases. These genes included NK3 homeobox 2 (nkx3.2/bapx1), NK2 homeobox 2a, 2b and 9 (nkx2.2a, nkx2.2b, nkx2.9), forkhead box domain F2a and F2b (foxf2a, foxf2b), Tbox20 (tbx20), netrin 1b (ntn1b), netrin 5 (ntn5), fatty acid binding protein 11b (fabp11b), urotensin II-related peptide (urp2), and pim1 (Table 3.3). We included the genes sonic hedgehog a (shha) and patched 2 (ptch2) as internal controls. We assessed gene expression in wild-type, igu mutant, DMSO-treated and cyclopamine-treated embryos by in situ hybridization at 30 hpf (Figs. 3.2 and 3.3).

Non- Non- Consensus Consensus Consesus Gli Consesus Gene Gli Binding Gene Gli Binding Binding Gli Binding Sites Sites Sites Sites Apof 0 0 Nkx2.2b 1 1 Aqp8a.1 1 3 Nkx2.9 1 0 Aspn 0 0 Nkx3.2 0 1

Atf3 0 1 Npvf 0 2 CR558302 0 2 Ntn1b 0 16 Fabp11b 0 5 Ntn5 0 1 Fos 0 1 Pax1a 0 5 Foxa2 0 4 Pim1 0 3

Foxf2a 1 1 Ptc1 5 24 Foxf2b 0 0 Sim1a 0 3 Grap2b 0 2 Smpx 0 1 Heyl 0 1 Smhyc3 0 1 Iqgap2 0 10 Tbx20 0 7

Jdp2 0 0 Tcap 0 1 Lepa 0 1 Tm6sf2 0 11 MAP2K5 1 8 Urp2 0 1 Nkx2.2a 1 1 Wrd65 2 31

Table 3.2- Number of putative consensus and non-consensus Gli binding sites for all annotated double hits from igu and cyclopamine microarrays

Santa Cruz Genome Browser was used to access genomic DNA, and sequences spanning from ~280bp to ~46kb up- or downstream were examined depending on conservation sites, as well as spanning the corresponding gene of interest. Consensus Gli binding sites consisted of the sequence GACCACCCA. Non-consensus binding sites consisted of the same sequence, with one base pair substitutions at any point in the sequence excluding the initial G and terminal A (Xu et al., 2006). Adapted from (Arnold et al., 2015) with permission of the publisher.

Microarray Z-Ratios Gene Protein Type Function M/W C/D

homeobox domain transcription cranial bone and cartilage morphogenesis nkx3.2 -2.4 -4.9 factor (Miller et al., 2003) oligodendrocyte fate determination (Qi et al., homeobox domain transcription nkx2.2a 2001; Kucenas et al., 2008), floor plate -2.1 -7.7 factor determination (Holz et al., 2010) oligodendrocyte fate determination (Qi et al., homeobox domain transcription nkx2.2b 2001; Kucenas et al., 2008), floor plate -6.5 -10.2 factor determination (Holz et al., 2010) homeobox domain transcription floor plate determination (Holz et al., 2010), nkx2.9 -8.0 -14.2 factor motoneuron development (Pabst et al., 2003) palatal development, lung gene transcription, forkhead box protein, winged- epithelial-mesenchymal interactions* (Miura et foxf2a -2.3 -2.8 helix transcription factor al., 1998; Aitola et al., 2000; Wang et al., 2003a) palatal development, lung gene transcription, forkhead box protein, winged- epithelial-mesenchymal interactions* (Miura et foxf2b -3.4 -3.8 helix transcription factor al., 1998; Aitola et al., 2000; Wang et al., 2003a) heart morphogenesis, dorsal aorta formation tbx20 T-box transcription factor -3.8 -2.2 (Szeto et al., 2002) axon guidance (Strahle et al., 1997; Wilson et ntn1b secreted protein -2.0 -2.7 al., 2006; Castets et al., 2009) ntn5 secreted protein unknown 2.4 -3.5

fabp11b fatty acid binding protein unknown -3.2 -6.1

urp2 urotensin-related protein unknown -6.3 -3.2 cell cycle regulation, smooth muscle novel gene serine-threonine protein kinase proliferation* (Katakami et al., 2004; Willert 3.3 2.2 (pim1) et al., 2010)

Table 3.3 – Selected set of differentially regulated genes for further verification, and fold changes reported by microarray

Functions denoted by an asterisk are from other species and have not been studied in zebrafish. M/W indicates ratio of gene expression in mutant over wild type, while C/D indicates ratio of gene expression in cyclopamine-treated embryos over control. Negative sign (-) indicates downregulation. Adapted from (Arnold et al., 2014) with permission of the publisher. 60

Figure 3.2 – In situ hybridization pattern changes of genes misregulated in igu mutants corresponds to direction of expression change on the microarray

A,A’: Nkx3.2 is absent from the notochord (arrowheads) of igu mutants. B,B’: In floor plate (arrows) and brain (arrowheads), Nkx2.2a signal is decreased in igu mutants. C,C’: Nkx2.2b signal is absent from floor plate (arrows) in igu mutants and decreased in ventral forebrain (arrowheads). D,D’: Nkx2.9 is severely reduced in the brain (arrowheads) and floor plate (arrows) of igu mutants. E,E’: Foxf2a ventral head mesoderm (arrows) signal is reduced in igu mutants compared with wild-type siblings. F,F’: Foxf2b signal in ventral head mesoderm (arrows) of wild-type sibling is greatly reduced in igu mutant. G,G’: Tbx20 transcript is decreased in the branchiomotor neurons (arrows) and heart (arrowheads) in igu mutants. H,H’: Ntn1b is decreased in the head (arrowhead) and floor plate (arrows) of igu mutants. I,I’: Ntn5 is up-regulated in the trunk (arrowheads) of igu mutants. J,J’: Fabp11b is expressed in the eye (arrowhead) and lost in igu mutants. K,K’: Urp2 is expressed in the floor plate (arrowheads) and is absent in igu mutants. L,L’: Pim1 transcript is up- regulated in the brain and eye of igu mutants. M,M’: Shha shows no change between wild-type sibling and igu mutant. N,N’: ptch2 is strongly up-regulated in igu mutants. Scale bars = 100 µm. Adapted from (Arnold et al., 2014) with permission of the publisher.

Figure 3.3 – In situ hybridization pattern changes of genes misregulated in cyclopamine-treated embryos corresponds to direction of expression change on the microarray

A,A’: Nkx3.2 is absent from the jaw joint (black arrowheads), pharyngeal arches (white arrowheads) and notochord (arrow) of cyclopamine-treated embryos. B,B’: Nkx2.2a signal is absent from the floor plate (arrows) and reduced in the brain (arrowheads) of cyclopamine-treated embryos. C,C’,D,D’: Expression of Nkx2.2b (C,C’) and Nkx2.9 (D,D’) is completely lost from brain (arrowheads) and floor plate (arrows) of cyclopamine-treated embryos. E,E’,F,F’: Ventral head expression of Foxf2a (E) and Foxf2b (F) is absent in cyclopamine-treated embryos (E’,F’). G,G’: Tbx20 transcript is decreased in branchiomotor neurons (arrows), but maintains expression in the heart (arrowheads) cyclopamine-treated embryos. H,H’: Ntn1b is decreased in the head (arrowheads) and floor plate (arrows) of cyclopamine-treated embryos. I,I’: Ntn5 is down-regulated in the trunk (arrowheads) of cyclopamine-treated embryos. J,J’: Expression of Fabp11b in the eye (arrowheads) is absent in cyclopamine-treated embryos. K,K’: Urp2 in the floor plate (arrows) is unchanged in cyclopamine-treated embryos. L,L’: Pim1 transcript expression is increased in the brain and eye of cyclopamine-treated embryos. M,M’: Shha shows no change between DMSO and cyclopamine. N,N’: ptch2 is strongly downregulated in cyclopamine-treated embryos. Scale bars = 100 µm. Adapted from (Arnold et al., 2014) with permission of the publisher.

In wild-type and DMSO-treated embryos, nkx3.2 is expressed in the notochord and pharyngeal arches but is not present in igu mutants (Fig. 3.2 A,A’), nor in cyclopamine-treated embryos (Fig. 3.3 A,A’). nkx2.2a, 2.2b, and 2.9 are all expressed in similar regions of the brain and floor plate (Fig. 3.2 B–D) and have decreased levels and restricted expression patterns in igu mutants (Fig. 3.2 B’,C’,D’). In cyclopamine-treated embryos, nkx2.2a expression pattern is again severely restricted (Fig. 3.3 B,B’) and there is no expression of nkx2.2b or 2.9 (Fig. 3.3 C,C’,D,D’). FoxF2 is expressed in several mesenchymal tissues in mice, including neural crest- derived head mesenchyme, and is induced by Shh signaling (Aitola et al., 2000; Wang et al., 2003a; Ormestad et al., 2006). At 30hpf, foxf2a and b are expressed in similar regions near the pharynx (Fig. 3.2 E,F) and have strongly reduced expression, restricted to just behind the eye in igu mutants (Fig. 3.2 E’,F’), while they are almost absent in cyclopamine-treated compared with DMSO controls (Fig. 3.3 E,E’,F,F’). tbx20 is a transcription factor that promotes heart development and migration of cranial motor neurons (Takeuchi et al., 2005). At 30 hpf, tbx20 shows decreased expression in the heart and branchiomotor neurons in igu mutants (Fig. 3.2 G,G’). Although expression of tbx20 is lost in branchiomotor neurons of cyclopamine-treated embryos, robust expression is maintained in the heart (Fig. 3.3 G,G’). This suggests that tbx20 expression in the heart is not controlled by Hh signaling, but requires dzip1 function. ntn1b demonstrates a decrease of expression in the head and floor plate of igu as compared with wild-type siblings (Fig. 3.2 H,H’). Similar ntn1b expression changes are seen in cyclopamine-treated fish (Fig. 3.3 H,H’). ntn5, however, appears to have an expanded domain of expression in the trunk of igu mutants, (Fig 3.2 I,I’), but expression is lost with cyclopamine treatment (Fig. 3.3 I,I’). The ntn5 expression changes correlate with changes observed in the expression microarray. Netrins are important for vascular patterning (Lu et al., 2004; Larrieu- Lahargue et al., 2012). ntn1b promotes angiogenesis in zebrafish (Castets et al., 2009), and is ectopically induced by Shh (Strahle et al., 1997). ntn5 has no reported role as of yet. We note that the differential expression changes of ntn5 seen between igu and cyclopamine-treated embryos coincide with the expanded expression pattern of ptch2 specifically seen in igu mutants,

but not in other conditions of disrupted Hh signaling (Sekimizu et al., 2004; Wolff et al., 2004), suggesting it may be regulated by igu in a similar manner to ptch2. Fatty acid binding proteins (FABP) mediate transport of lipids and other insoluble molecules (Weisiger & Zucker, 2002). At 30hpf, zebrafish fabp11b is expressed in the retina and is lost both in igu mutants (Fig. 3.2 J,J’) and cyclopamine-treated embryos (Fig. 3.3 J,J0). urp2 is expressed in the anterior third of the spinal cord at 30hpf, and urp2 expression is lost in igu mutants (Fig. 3.2 K,K’). Of interest, urp2 expression appears relatively unchanged in cyclopamine-treated embryos compared with DMSO controls (Fig. 3.3 K,K’). This is unexpected, as urp2 was down-regulated on the cyclopamine microarray, and I hadexpected that in situ hybridization staining would show it down-regulated or absent in these embryos. urp2 is not known to be transcriptionally regulated by Shh.

The Pim proteins are family of serine/threonine kinases acting on several cellular processes and have known roles specifically in smooth muscle proliferation (Katakami et al., 2004; Willert et al., 2010). We note that a renaming of many of these genes between genome versions Zv6 and Zv9 has made it difficult to ascertain which genes were actually identified as differentially expressed on the array, although six of the eight pim-like genes in Zv6 are still annotated as pim-like genes in Zv9. As full length sequences are unavailable for these genes, we chose to examine the published pim1 gene (ENSDART00000082124, NM_131539). This full length gene was not annotated until after the Zv6 genome assembly. In situ hybridization for pim1 at 30 hpf showed a strong upregulation in igu mutants (Fig. 3.2 L,L’) and cyclopamine- treated embryos (Fig. 3.3 L,L’). As controls, we performed in situ hybridization for shha and ptch2. shha expression has been reported previously to be unaffected in Hh signaling mutants (Brand et al., 1996), while ptch2 is reduced in Hh signaling-deficient embryos, but also up-regulated in somites of igu mutants (Sekimizu et al., 2004; Wolff et al., 2004). We indeed find that at 30 hpf, shha expression is unchanged in igu mutants and cyclopamine-treated embryos with expression observed in the floor plate as well as ventral midbrain, ventral forebrain, and forebrain–midbrain boundary (Figs. 3.2 M,M’, 3.3 M,M’). Likewise ptch2 expression is up-regulated in the somites

of igu mutants, and down-regulated in cyclopamine-treated embryos (Fig. 3.2 N,N’, Fig. 3.3 N,N’). In summary, the genes we sampled have diverse expression patterns. All selected genes with reduced expression on the array showed similarly reduced expression levels in igu mutants and cyclopamine-treated embryos as compared to controls at 30hpf by in situ hybridization (with the exception of urp2). Conversely, pim1 showed increased levels in both comparisons, while ntn5 was up-regulated in igu mutants and down-regulated in cyclopamine-treated embryos, which also corresponds with the array data. shha expression remained unchanged, as was expected. Thus, it appears that Hh and igu normally activate most of the genes we have examined here but also have tissue-specific repression activity for some. Our in situ hybridization analysis support the microarray results that these transcripts are specifically and differentially regulated in igu mutants as a result of aberrant Hh signaling, as shown by expression levels in cyclopamine-treated embryos.

3.2.4 Microarray Validation by RNA Deep-Sequencing

To further verify the results obtained from our igu expression microarray by a third independent method, we performed RNA deep-sequencing on igu mutants and wild-type siblings to identify genes consistent through microarray, in situ hybridization and RNAseq. The 193 genes identified through deep sequencing passed a cut-off of a 7.5% false discovery rate. A list of 12 genes identified with differential expression on the microarray that also are found in the RNAseq dataset is presented in Table 3.4, while 136 genes were found to have two-fold or higher expression change on the RNAseq dataset alone. All 12 genes identified in the RNAseq dataset that had concordant expression changes between igu mutant and cyclopamine-treated embryo arrays showed the same direction of differential regulation by RNA sequencing. These hits represent a 37% overlap with annotated hits in the 40 gene list (Table 3.1).

Ensembl Identifier Gene Symbol Log2 Fold Change Log2 CPM p-value False Discovery Rate

ENSDARG00000052550 nkx2.2b -3.51 2.68 1.59E-19 0.0% ENSDARG00000020332 nkx2.9 -5.12 1.62 3.12E-16 0.0% ENSDARG00000053298 nkx2.2a -2.16 4.06 4.41E-16 0.0% ENSDARG00000002311 fabp11b -2.24 3.73 1.07E-15 0.0% ENSDARG00000017195 foxf2a -2.38 2.44 3.87E-11 0.0% ENSDARG00000067781 urp2 -8.01 0.38 2.27E-09 0.0% ENSDARG00000029057 tm6sf2 -1.97 2.18 9.53E-08 0.0% ENSDARG00000005150 tbx20 -1.58 2.91 2.68E-07 0.0% ENSDARG00000003411 foxa2 -1.21 3.38 1.36E-05 0.2% ENSDARG00000070389 foxf2b -2.45 0.75 3.27E-05 0.5% ENSDARG00000036227 npvf -4.05 -0.14 4.58E-05 0.7% ENSDARG00000045302 smpx -0.84 4.31 0.0005212 5.3% ENSDARG00000037639 nkx3.2 -0.96 2.67 0.0020729 14.5% ENSDARG00000036422 NTN5 1.40 1.21 0.0022549 14.8% ENSDARG00000022531 ntn1b -0.37 2.78 0.221009 99.2% ENSDARG00000059120 pim1 0.08 7.58 0.6918959 100.0%

12 gene hits from the next-generation RNA deep sequencing data are also found as significantly altered hits on the microarray. This represents a 37% overlap with named genes off the 40 gene list from the igu and cyclopamine arrays. The four genes below the double line were also significantly altered on the microarray, but did not pass the false discovery rate cutoff. Direction of expression changes for each gene is consistent with the microarray. CPM – counts per million. Adapted from (Arnold et al., 2014) with permission of the publisher.

3.2.5 igu Mutants Have a Unique Gene Expression Signature

One surprising finding from our data is how few genes are differentially regulated in the igu mutant dataset as compared to the cyclopamine-treated embryo given that the igu mutant has been studied as a genetic mutant model of disrupted Hh signaling (Sekimizu et al., 2004; Wolff et al., 2004; Glazer et al., 2010; Kim et al., 2010; Tay et al., 2010). If this was its only biological role, we would expect to see a much larger overlap in genes changed in both igu mutants and cyclopamine-treated embryos. Instead, we observe a far greater number of genes differentially in each condition individually, and only a small subset of genes regulated in common. We note that we sampled embryos at the 30hpf, a window important for mural cell differentiation (Lamont et al., 2010) and it is possible that the overlap in datasets between iguana and cyclopamine might be more similar during early development, especially because cyclopamine was applied after gastrulation, while igu mutants are deficient in Dzip1 from fertilization. However, genes that we identified as being regulated by Hh in our study overlap substantially with other Hh-deficient arrays (Xu et al., 2006; Bergeron et al., 2008; Buttner et al., 2012). We observe unanticipated differences in gene regulation with loss of Dzip1 as compared to reduction in Hh signaling, indicating Dzip1 plays other roles in development independent of Shh signaling, and our description of the genes regulated by igu is only a first step to understanding its broader functions.

3.2.6 Genes Uniquely Regulated by Dzip1

Potential Hh-independent roles for Dzip1 might be elucidated from gene expression changes from the igu microarray that do not overlap with the cyclopamine microarray (Table 3.5). Several down-regulated genes, including mych4, scn4ab, ttna, ttnb, and nfI/Xb, are potentially involved in skeletal muscle function and development. Also, the extracellular matrix component prelp, expressed in myotome borders in the trunk, is significantly up-regulated in igu mutants. ttna/b are cytoskeletal genes, and additional cytoskeletal components are also downregulated (tmod2, kif1a, myo1C). Other down-regulated hits are neuronal and brain-related genes (jhdm1da, hoxb3a), DNA and RNA-mediating genes (nudt15, snrnp70, rbm14a), the

Upregulated Hits Gene Symbol Gene Name Fold Change Genbank Ensembl ZDB Gene ifnphi1 interferon phi 1 4.976 AY135716 ENSDARG00000025607 ZDB-GENE-030721-3 prelp proline/arginine-rich end leucine-rich repeat protein 3.425 XM_001923555 ENSDART00000103846 ZDB-GENE-080327-24 mfap4 microfibril-associated glycoprotein 4-like* 3.197 XM_003197793 ------nitr5 novel immune-type receptor 5 2.714 BC163833 ENSDARG00000057900 ZDB-GENE-020225-40 N/A zgc:174680 2.475 NM_001114562 ENSDARG00000077110 ZDB-GENE-080204-24 N/A retinoblastoma binding protein (6 of 71) (si:ch211-198k9.6) 2.334 BX908780 ENSDARG00000069804 ZDB-GENE-090312-4

Downregulated Hits Gene Symbol Gene Name Fold Change Genbank Ensembl ZDB Gene myhc4 myosin heavy chain 4* -3.410 NM_001020485 ENSDARG00000035438 ZDB-GENE-030131-6206 tmod2 tropomodulin 2* -3.324 NM_001004608 ENSDARG00000002571 ZDB-GENE-040912-185 N/A si:dkey-195e19.4-001 -3.313 CV484246 --- ZDB-TSCRIPT-090929-6424 myo1c myosin 1C* -3.301 AW419653 ENSDARG00000061579 ZDB-GENE-090429-3 nudt15 nudix (nucleoside diphosphate linked moiety X)-type motif 15 -3.238 NM_001002107 ENSDARG00000022461 ZDB-GENE-040625-95 N/A protein kinase Npk (LOC405768) -3.236 NM_212842 ------ttnb titin b -3.012 DQ649453 ENSDARG00000000563 ZDB-GENE-030616-413 scn4ab sodium channel, voltage-gated, type IV, alpha, b -3.000 NM_001045065 ENSDARG00000034588 ZDB-GENE-051201-1 ttna titin a -2.995 AY033829 ENSDARG00000028213 ZDB-GENE-030113-2 snrnp70 small nuclear ribonucleoprotein 70 -2.952 NM_001003875 ENSDARG00000077126 ZDB-GENE-040825-2 rbm14a RNA binding motif protein 14a* -2.923 XM_005161334 --- ZDB-GENE-050522-496 jhdm1da jumonji C domain containing histone demethylase 1 homolog Da -2.901 XM_687822 ENSDARG00000018111 ZDB-GENE-030131-9829 tspan17 tetraspanin 17* -2.866 NM_001123045 --- ZDB-GENE-030131-3673 N/A wu:fb25b09* -2.863 XM_686268 --- ZDB-GENE-030131-379 kif1a kinesin family member 1A* -2.832 XM_002662438 ENSDARG00000061817 --- qdpra quinoid dihydropteridine reductase a* -2.789 NM_001110469 ENSDARG00000040190 ZDB-GENE-070705-197 N/A IMAGE:4886965 -2.743 BI842227 ------nfI/Xb nuclear factor I/Xb* -2.723 NM_001040248 ENSDARG00000061836 ZDB-GENE-060421-5000 N/A IMAGE:5465525 -2.547 BM186580 ------hoxb3a homeobox B3a, transcript variant X9 (Predicted)* -2.545 XM_005171562 ENSDARG00000029263 ZDB-GENE-990415-104 N/A FAM171A2-like (Predicted)* -2.472 XM_001337519 ------lmnb1 lamin B1 -2.438 NM_152972 ENSDARG00000044299 ZDB-GENE-020424-2 N/A IMAGE:5414138 -2.411 BM185738 ------

Table 3.5 – Genes with altered expression in igu mutant zebrafish embryos only, as reported by microarray

Genes denoted by an asterisk were obtained via top hits of BLAST analysis of associated protein sequences of retired identifiers. Adapted from (Arnold et al., 2014) with permission of the publisher.

nuclear lamina constituent lmnb1, and membrane scaffold tspan17. Of interest, three of the four highest up-regulated genes include the immune response-related genes ifnphi1, nitr5, and mfap4). Additionally, several retired identifiers of up-regulated hits BLAST to members of the pim family of kinases (pim1, 2, and 3-like).

3.3 Discussion

3.3.1 Potential vascular stability genes identified by the microarray

The hemorrhage phenotype seen in igu mutants is indicative of defects in establishing vascular stability. There are several genes in our dataset that may plausibly play a role in promoting vascular stability in development. nkx3.2 expression has been found in the medial layer of arterial smooth muscle cells where it co-activates transcription of the vascular smooth muscle-specific integrin α1 in concert with SRF (Nishida et al., 2002). In zebrafish, the neural crest gives rise to mesenchymal tissues in the head, and nkx3.2 interacts with the neural crest specifying gene sox9 (Zeng et al., 2002; Yamashita et al., 2009; Cairns et al., 2012). PAX1 in mouse is expressed in both sclerotome and the vascular smooth muscle cells and pericytes that derive from it (Pouget et al., 2008), and pax1a was down-regulated in both igu mutants and cyclopamine-treated embryos. heyl is down-regulated in igu mutants and cyclopamine-treated embryos, and the mouse homolog, HeyL, is expressed in vascular smooth muscle (Leimeister et al., 2000). HeyL is thought to be an effector for Notch3 signaling, which is disrupted in the genetic hemorrhagic disease CADASIL (Joutel et al., 1996). MAP2K5 is an activator of ERK5, which plays a role in promoting PDGF-mediated vascular smooth muscle cell migration (Izawa et al., 2007), and MAP2K5 was downregulated in both datasets. atf3 was one of the few significantly up-regulated genes in both igu mutants and cyclopamine-treated embryos, and is thought to regulate smooth muscle cell survival and migration in mice (Lv et al., 2011). The identification of multiple genes from igu mutants and cyclopamine-treated embryos that are potentially involved in vascular smooth muscle development lends confidence to the idea Hh signaling through Dzip1 promotes vascular stabilization. One of the surprises in the microarray dataset was that angpt1, previously shown by our lab to be down-regulated in igu mutants (Lamont et al., 2010), was not present in the list of genes with significantly altered expression. Angpt1 was initially identified as a potential downstream target of Shh signaling in other systems (Pola et al., 2001). While we showed that angpt1 expression is down-regulated in igu mutants by in situ hybridization, levels were not quantified in our previous study and may not have met the required threshold for significance.

One of the genes in the igu-cyclopamine comparative dataset for which there was little known about association with vascular mural cells is FoxF2, where both foxf2a and foxf2b were downregulated in both cyclopamine-treated and igu embryos. At the time, it was known that the Drosophila homologs of nkx3.2 (bagpipe) and foxf (biniou) act in concert to specify visceral muscle (Jakobsen et al., 2007), and mouse FoxF2 promotes mesenchymal interactions with epithelial cells to mediate gut muscle development (Ormestad et al., 2006). Given its similar expression to nkx3.2 in head tissues, and known roles in visceral muscle specification in fly and mouse led me to explore this gene further as a potential mediator of vascular stabilization.

3.3.2 Differentially regulated genes specific to the igu microarray.

Several genes are differentially expressed only on the igu array. Another phenotype of the zebrafish igu mutant is a curled body axis (Brand et al., 1996). The cause of this is not known, but skeletal muscle fibers in other curled body axis mutants are disrupted (Du & Dienhart, 2001), and may explain this enrichment for muscle-associated genes. In addition to being a Shh signaling center, primary cilia can act as mechanosensors and induce cytoskeletal reorganization in response to various cues (Jones et al., 2012). Defective cilia formation due to igu may explain the affected expression of cytoskeletal-related genes. The upregulated immune-related genes are particularly interesting. No previous studies have directly linked primary cilia to immune responses, and these genes may suggest a novel role for Dzip1/cilia function. Primary cilia are found ubiquitously and perform a wide array of cell-specific functions (Berbari et al., 2009; Zaghloul & Brugmann, 2011), which may account for the range of functions of the differentially expressed igu microarray genes. In fact, given that all primary cilia are defective in igu mutants (Tay et al., 2010), it is surprising there are not more genes with significantly altered expression. This could be due to the singular timepoint at which we have examined gene expression, and a comparison of igu mutants with wildtype siblings at earlier or later stages may produce other differentially regulated genes.

3.3.3 Additional Microarray Validation by RNAseq

The RNAseq dataset of igu mutants showed an overlap of several genes identified on the array, including nkx2.2a, 2.2b and 2.9, which have all been previously identified in other Hh gene expression analyses (Xu et al., 2006; Bergeron et al., 2008), as well as foxf2a/b. This RNAseq has one major caveat, as only one experimental replicate was run. At the time, this technique was still relatively in its infancy. We were advised that one sample each would be adequate, whereas three replicates is currently regarded as the standard minimum needed to draw reliable conclusions. Due to this, logCPM (counts per million) values and the depth of our RNAseq coverage are lower than optimal. However, many of these genes still show a very low false discovery rate, which partially supports the accuracy of the reported fold changes in the RNAseq data. In the future, RNAseq experiments should be run on an absolute minimum of three replicates of all biological conditions.

3.3.4 Conclusions

This gene expression analysis produced both expected and unexpected results and provides a novel dataset of genes potentially involved in vascular stabilization. From this, the transcription factor FoxF2 was chosen for further analysis in zebrafish of its potential for regulating vascular stability.

Chapter Four: FoxF2b is Required for Embryonic Vascular Mural Cell Development

Jae-Ryeon Ryu helped with Western blotting and TALEN mutagenesis. Nicole Munsie helped with PDGFRα, PDGFAA, PDGFBA, Gli2a and Ptch2a in situ hybridization experiments. Wei Dong and the Microscopy Imaging Facility helped with acquisition of TEM images. 73

Abstract In the previous chapter, the FoxF2 transcription factor was identified as a downregulated target of Shh signaling in iguana mutants with vascular instability manifesting as cerebral hemorrhage. Here I probed the role of FoxF2 in vascular stabilization. Interactions between endothelial cells and mural cells are required for the support and stability of the vasculature. In the head, mural cells (vascular smooth muscle cells and pericytes) largely originate from neural crest mesenchyme. FoxF2 has known roles in regulating mesenchymal tissues and is more strongly expressed in head tissues of the developing mouse than its FoxF1 counterpart. Here I asked whether the decreased FoxF2 in igu mutants was contributing to the hemorrhage, possibly by dysfunctional regulation of the neural crest-derived mesenchyme. By in situ hybridization, I found foxf2 is expressed in ventral mesoderm and neural crest, two sources of head mesenchyme tissues. Transient knockdown of the zebrafish foxf2b paralog results in embryonic hemorrhage due to poor endothelial-mural cell interactions. Directed mutagenesis of foxf2b also shows decreased endothelial-mural cell interactions and delayed mural cell coverage of cerebral vessels, but only a subtle difference in hemorrhage phenotypes. Investigation of potential FoxF2 target genes and pathways implicated it in regulating head mesenchyme and promoting differentiation of vascular mural cells. Thus I have identified a novel role for zebrafish FoxF2 in promoting the development of a functional vascular system.

4.1 Introduction

FoxF2 possesses several reported functions that suggest it is a candidate regulator of vascular stability. In mouse, FoxF2 is expressed in neural crest and neural crest-derived mesenchyme in the head (Ormestad et al., 2004; Xu et al., 2016), which is a source tissue for most mural cells of the head (Etchevers et al., 2001; Heglind et al., 2005). Zebrafish neural crest lineage tracing identified neural crest-derived vSMCs of the pharyngeal arteries (Mongera et al., 2013), verifying that mural cells can originate from neural crest. Proper neural crest migration requires epithelial to mesenchymal transition (EMT), a process that is co-opted in cancer cells to promote metastasis. FoxF2 is involved in EMT in cancer cells, however depending on the type of cancer it was either activating (Kundu et al., 2016) or inhibitory (Cai et al., 2015; Wang et al., 2015). In a breast cancer cell line, FoxF2 acts as a repressor of EMT-promoting Twist1 transcription factor (Wang et al., 2015). Known FoxF2 targets include genes involved in endothelial-mural cell interactions. Mouse FoxF2 mutants show altered expression of PDGF-B and PDGFRβ signaling components (Bolte et al., 2015; Reyahi et al., 2015), which can lead to impaired mural cell recruitment. As well, certain ECM components, particularly ColIV, are decreased in FoxF2 mutants (Ormestad et al., 2006; van der Heul-Nieuwenhuijsen et al., 2009a). Finally, TGF-β signaling pathway members, including TGFβR2 and Alk5, were reduced at both the transcript and protein level in mouse FoxF2 mutants, resulting in decreased phosphorylation of Smad2/3 (Reyahi et al., 2015). FoxF2’s regulatory control over these pathways demonstrates that it has the potential to mediate endothelial-mural cell interactions and vessel wall maturation. The FoxF2 chromosomal gene sequence is flanked by FoxQ1 and FoxC1, forming a Fox gene cluster that is highly conserved throughout evolution. FoxC1 and FoxQ1 have predominantly been examined in the context of cultured cancer cell growth and metastasis. FoxC1 is also strongly associated with human ocular defects such as anterior segment dysgenesis, particularly Axenfeld-Rieger syndrome (Reis et al., 2012). There is also evidence for FoxC1 involvement in cerebral small vessel disease. FoxC1 is expressed in pericytes of the mouse brain, and mural cell-specific knockout of FoxC1 leads to overproliferation of pericytes, altered vascular morphogenesis and focal hemorrhages of varying severity (Siegenthaler et al., 75

2013). The relatively small distance between FoxF2 and FoxC1 and the high level of evolutionary conservation of this cluster suggests there is potential for FoxF2 and C1 to share or compensate regulation of certain biological processes, including vascular stability. Nothing is known of FoxQ1 function in vascular stabilization. Finally, as shown in the previous chapter, FoxF2 expression is decreased downstream of Shh in a genetic model of cerebral hemorrhage, indicating a potential for FoxF2 promoting vascular stability. All of this evidence, taken together, place FoxF2 in good standing as a candidate for regulation of vascular stability. Through the following experiments, I have elucidated the expression domains of foxf2a/b in the developing zebrafish, investigated vascular stability phenotypes in transient knockdown and genetic knockout models of foxf2b, and identified potential genes and biological processes regulated by FoxF2.

4.2 Results

4.2.1 foxf2 is expressed in head mesenchymal tissues adjacent to endothelial cells and distinct from smooth muscle cells

To understand the function of FoxF2 in development I first looked at its expression pattern by in situ hybridization. I found that foxf2b is expressed during development as early as 8 somites (s). Expression is seen in ventral mesoderm and faintly in migrating neural crest at 8s and 12s stages (Fig. 4.1 A-B). From 26 to 36hpf, expression is in the head mesenchyme derivatives of neural crest and mesoderm. (Fig. 4.1 C,D). By 48hpf, foxf2b expression is restricted to distinct mesenchymal derivatives of the head mesenchyme, which include paraxial mesenchyme surrounding the anterior portion of the notochord, pharyngeal arch mesenchyme, periocular mesenchyme and oral mesenchyme which borders the upper and lower oral ectoderm layers (Fig. 4.1 E,E’). Pharyngeal arch mesenchyme expression at 48hpf is much stronger in posterior regions. Sectioning demonstrates foxf2b expression in oral and ocular mesenchyme at 48hpf (Fig. 4.2 A-C’), and some foxf2b expression can also be seen intercalating the oral ectoderm. foxf2b is clearly expressed in tissues surrounding endothelium at 48hpf in the head (Fig. 4.2 D,D’). By 4dpf, foxf2b expression is largely restricted to mesenchyme bordering the palate and dorsal notochord (Fig. 4.3). Interestingly, co-staining for acta2:GFP smooth muscle reporter shows foxf2b expression in these tissues is adjacent to, but distinctly separate from acta2- expressing cells (Fig. 4.3 A’B’). I next examined expression of the paralogous foxf2a gene and found it is expressed in similar patterns to ventral foxf2b up to 24hpf, but starts to show expression in head mesenchyme derivatives such as the paraxial head mesenchyme and pharyngeal arch mesenchyme beginning at 36hpf as opposed to 48hpf (Fig. 4.4). For instance, foxf2a staining in the pharyngeal arch mesenchyme at 36hpf is not seen at the same age for foxf2b (Fig. 4.4 D,E). foxf2a expression is seen in both anterior and posterior regions of head mesenchyme at 48hpf (Fig 4.4 F,G). Interestingly, no neural crest foxf2a expression is observed. To examine FoxF2b protein localization, I had several mouse monoclonal antibodies developed against six different peptide epitopes by a commercial supplier. Two of these epitopes

Figure 4.1 – foxf2b is expressed in neural crest, ventral mesoderm and their head mesenchymal derivatives during early development

A,A’: Lateral view (A) and dorsal-anterior view (A’) of foxf2b ventral mesoderm and neural crest expression at 8s. B,B’: Lateral view (B) and dorsal anterior view (B’) of foxf2b ventral mesoderm and neural crest expression in at 12s. C – Lateral view of foxf2b expression in head mesenchyme at 26hpf. D: Lateral view of foxf2b expression in head mesenchyme at 36hpf. E,E’: Lateral (E) and dorsal (E’) view of foxf2b expression in head mesenchyme derivatives at 48hpf. e = eye, nc = neural crest, hm = head mesenchyme, hpf = hours post fertilization, pam = pharyngeal arch mesenchyme, phm = paraxial head mesenchyme, pom = periocular mesenchyme, vm = ventral mesoderm. Scale bars = 100µm. 78

Figure 4.2 – foxf2b expression surrounds endothelium at 48hpf

A,A’: Sagittal section of foxf2b expression domains in head mesenchyme tissues. Boxed region magnified in A’. B,B’: Coronal section of foxf2b expression in paraxial head mesenchyme. Notochord, outlined by dashed line in magnified region (B’). C,C’: Transverse section of foxf2b expression domain in oral mesenchymes, adjacent to oral ectoderm. Magnified region shows oral ectoderm monolayers delineated by dashed line (C’). D,D’: Sagittal section showing foxf2b expression alongside staining for endothelial marker kdrl (brown). Magnified region shows foxf2b expression around endothelial cells in the head of the zebrafish embryo (D’). n = notochord, oe = oral ectoderm, om = oral mesenchyme, pam = pharyngeal arch mesenchyme, phm = paraxial head mesenchyme, pom = periocular mesenchyme. Scale bars = 100µm, 20µm in magnified regions.

Figure 4.3 - foxf2b does not co-localize with acta2 at 4dpf

A,A’,B,B’: Sagittal sections showing foxf2b expression (blue; arrows in magnified regions in A’ and B’) alongside staining for mural cell marker acta2 (brown; arrowheads in magnified regions in A’ and B’). acta2 signal marks penetrating cerebral vessel (A, A’), immediately dorsal to foxf2b expression domain. foxf2b and acta2 signals adjacent to the notochord but in mutually exclusive expression domains (B,B’). n = notochord. Scale bars = 100µm, 20µm in magnified regions.

Figure 4.4 – foxf2a is expressed in head mesenchymal derivatives during early development

A-C: Dorsal views of foxf2a expression in 8s (A), 20s (B) and 24hpf (C) embryos is strongly seen in ventral mesoderm and head mesenchyme. D,D’: foxf2a expression imaged dorsally (D) and laterally (D’) in 36hpf embryo in head mesenchyme derivatives. E,E’: foxf2a expression imaged dorsally (E) and laterally (E’) in 48hpf embryos is in head mesenchyme derivatives. hm = head mesenchyme, pam = pharyngeal arch mesenchyme, phm = paraxial head mesenchyme, pom = periocular mesenchyme, vm = ventral mesoderm. Scale bars = 100µm.

occur in sequence also conserved with FoxF2a and FoxF1 (pan-FoxF), and one is in sequence conserved with FoxF2a only (FoxF2 specific), while the remaining three were designed to target FoxF2b only (Fig 4.5 B). The company tested reactivity of the antibodies to peptides by ELISA but to further verify that they can detect FoxF2b protein, I tested antibodies on cos7 cells transfected with a vector containing zebrafish foxf2b sequence fused to an mCherry fluorescent reporter. I used untransfected cells and cells transfected with an empty vector as negative controls. For 5 of the 6 antibodies, signal was seen specifically in cells expressing the vector, as seen with the overlapping green and red fluorescence (Fig. 4.5 A). The 6th antibody (epitope 6) showed non-specific cytoplasmic staining in all cells. Signal did not appear in untransfected or empty vector-transfected cells with any antibody aside from epitope 6 (untransfected and empty vector images are representative). My data shows that five of the antibodies were able to recognize FoxF2b protein in transfected cells. I next tested whether the antibodies would recognize native FoxF2 protein in zebrafish by wholemount staining. Staining using the antibodies to epitopes 1-4 in 48hpf zebrafish embryos did not yield specific staining patterns, and some antibodies cross-reacted heavily with the skin (Fig 4.5 C). In contrast, the established Zn5 antibody, used as a positive control for testing staining specificity, yielded specific staining patterns, and negative controls showed no green signal, suggesting the non-specific fluorescence is due to the primary antibodies. These antibodies arrived as a fairly insoluble powder and I used an IgG purification protocol to clean the antibodies before use. It’s possible that additional non-antibody proteins were carried through the staining and interfere with specificity. Additionally, I did not affinity purify the antibodies which could further increase their binding efficiency. Preliminary experiments show that the epitope 4 antibody will be useful for Western blotting.

4.2.2 Transient knockdown of foxf2b results in embryonic cerebral hemorrhage and reduced mural cell coverage of vessels

foxf2b was knocked down using morpholino oligonucleotides and embryos were assayed for hemorrhage at 48hpf, based on the phenotype in the igu mutants. The foxf2b morpholino was designed as a translation-blocking oligonucleotide before complete genome sequence was

Figure 4.5 – FoxF antibodies specifically bind FoxF2 outside of zebrafish embryos

A: Antibody staining of mouse monoclonal antibodies targeting epitopes 1-6 in cos7 cells transfected with zebrafish foxf2b. Antibodies do not stain untransfected cells or cells transfected with an empty vector. B: FoxF2b protein sequence demonstrating location of the epitopes. Red sequence is conserved with FoxF2a and FoxF1, blue sequence is conserved with FoxF2a. C: Antibody staining of epitopes 1-4 in zebrafish embryos at 48hpf.

known. In the current foxf2b annotation, this morpholino (“i2e3”) targets foxf2b in coding sequence 288 base pairs downstream from a putative upstream translational start site or 141 base pairs upstream from a conserved downstream translation start site. The downstream translation start site is homologous among foxf2 genes in other species, while there is no evidence for a homologous upstream start site. For this reason, we think that this morpholino acts as a translation blocker upstream of the start site, although standard translation blocking morpholinos are targeted within 25 bp upstream or downstream of the translational start site. Injections of 5.4 ng of this morpholino resulted in hemorrhage occurring at a rate of 17.26 ±1.40% (n=2868 embryos, 35 experimental replicates) compared to uninjected control hemorrhage rate of 9.24 ±1.74% (n=3170 embryos, 31 experimental replicates, p<0.001 by the two-tailed t test; Fig. 4.6 A-C). Preliminary western blotting of morpholino-injected embryos (morphants) demonstrates a potential decrease in FoxF2 protein expression in both hemorrhaged and non-hemorrhaged embryos, while β-tubulin serves as a loading control. Further replicates are needed before quantitative knockdown can be assessed. A subset of injected embryos were examined to determine where hemorrhages were occurring (Fig. 4.6 D,E). Most hemorrhages occurred in the midbrain region of the developing embryos. It is also of note that of the ventral hemorrhages, 15 of the 16 recorded were in the region of the pharynx, between the gill arches and the eye, and of the eye hemorrhages, 14 of the 15 recorded were present in the space surrounding the lens. Knockdown experiments targeting foxf2a were performed with a morpholino targeted to the splice site between exon 1 and intron 1 to prevent proper splicing (“e1i1”). Injections did not produce significantly higher hemorrhage rates than uninjected controls at 48hpf (Fig. 4.7 A – uninjected siblings: 6.22 ±1.37%, n=1039 embryos, 12 experimental replicates; 1.7ng e1i1 MO: 5.98 ±1.87% n=227 embryos, 4 experimental replicates; 3.3ng e1i1 MO: 5.89 ±1.33% n=521 embryo, 7 experimental replicates; 7.9ng e1i1 MO: 7.22 ±2.04% n=266 embryos, 5 experimental replicates; p>0.05 in all unpaired t-tests against uninjected siblings). Morpholino efficiency was assayed by qualitative RT-PCR of foxf2a transcript and showed that, although some unspliced product was present, a large amount of spliced product remained (Fig. 4.7 B). Thus these results are derived from a weak partial knockdown and are inconclusive.

Figure 4.6 – Morpholino knockdown of foxf2b results in embryonic hemorrhage

A: Uninjected control (UIC) zebrafish embryo at 48hpf, exhibiting no hemorrhage. B: foxf2b MO injected embryo at 48hpf with massive cerebral hemorrhage in fore-, mid- and hindbrain regions. C: Graph depicting hemorrhage rates in uninjected control and foxf2b MO injected embryos (p=0.0006). D: Diagram showing regions for hemorrhage positions counts. Diagram of hemorrhage positions adapted from Haffter et al. 1996 (Haffter et al., 1996) with permission of publisher. E: Table of hemorrhage positions in foxf2b morphant embryos. F: Western blot verification of FoxF2 protein knockdown by morpholino in hemorrhaged and non-hemorrhaged embryos. β-tubulin was used as loading control. Scale bars = 100µm.

Figure 4.7 – foxf2a morpholino does not result in embryonic hemorrhage or efficient mis-splicing of mRNA

A: Graph depicting hemorrhage rates of zebrafish embryos injected with 1.7ng, 3.3ng and 7.9ng of foxf2a e1i1 splice morpholino. No significant differences in hemorrhage occur. B: RT-PCR gel electrophoresis of foxf2a and eef1α1l1 PCR products of cDNA reverse transcribed from mRNA of uninjected control or foxf2a morpholino-injected embryos. Spliced foxf2a product is 163bp, unspliced foxf2a product is 1338bp, eef1α1l1 product is 359bp.

One cause of vascular instability seen in foxf2b morphants could be that vascular mural cells are not properly supporting vessels. As such, I wanted to understand whether knockdown of foxf2b would affect coverage of vessels by smooth muscle actin (Acta2) positive mural cells.Acta2 is a marker of vascular smooth muscle cells that progressively associate with ventral head and large brain vessels starting at 3.5dpf (Whitesell et al., 2014). I assessed by GFP+ cells in foxf2b morphant and uninjected control Tg(acta2:EGFP)ca7 embryos that express GFP in acta2 positive mural cells at 4dpf (Fig. 4.8). Morphants showed a decreased number of GFP+ cells on the pharyngeal arch vessels (32.33 ±7.04, n=9) compared to uninjected controls (68.67 ±12.56, n=9, p<0.001 by two-tailed t test). Vessel coverage by mural cells and ultrastructure was further analysed by transmission electron microscopy of 48hpf morphants and wild type controls (Fig. 4.9). TEM images were quantified by measuring the fraction of abluminal vessel surface contacted by surrounding mural cells. Analysis of all vessels revealed a general decrease in the amount of abluminal surface in contact with mural cells (Fig. 4.9 C – uninjected siblings: 57.42 ±6.12%, n=12 vessels counted among 3 embryos; “i2e3” MO: 36.67 ±4.33%, n=15 vessels counted among 3 embryos; p<0.01 by two-tailed t test). Additionally, a comparison was made between two different vessel types: the primitive internal carotid artery (PICA), and the dorsal aorta, representing vessels of the head and trunk, respectively. The PICA shows a decrease in the amount of abluminal vessel surface contacted by mural cells in morphant embryos (33.00 ±7.90% n=4 vessels) compared to uninjected controls (60.20 ±7.81%, n=5 vessels; p<0.05 by two-tailed t test; Fig. 4.9 D) In contrast, no difference was seen between morphant) and control (“i2e3” MO: 57.67 ±9.40%, n=3 vessels; uninjected siblings: 52.00 ±4.00%, n=2 vessels; p>0.05; Fig. 4.9 D) dorsal aorta coverage by mural cells, suggesting FoxF2b knockdown affects vessels of the head, but not the trunk, consistent with its expression in head mesenchyme surrounding vessels.

4.2.3 Generation of TALEN-mediated FoxF2b knockout mutant zebrafish

To generate a genetic knockout model for FoxF2b, TALEN RNAs were constructed targeting a 20 bp region 300 bp from the putative upstream translational start site, at the beginning of the coding region for the DNA binding domain (Fig. 4.10). Injections of 87

Figure 4.8 – foxf2b morphants have decreased acta2:GFP+ coverage of pharyngeal vessels

A-B: acta2:GFP expression (mural cells; green) and kdrl:mCherry (endothelium; red) in 4 dpf control (A,A’) and foxf2b morphant (B,B’) embryos. A’ and B’ are enlargements of the boxes in A and B, respectively. GFP signal can also be seen in heart (h), around the notochord (n). Eye signal is autofluorescence. Numbers correspond to arch arteries 1-6 (2 is also called the opercular artery). C: Graph showing counts of acta2:GFP-expressing cells on pharyngeal vessels of control and foxf2b morphant embryos. p=0.0013. Scale bars = 100µm.

Figure 4.9 – Decreased mural cell contacts with abluminal surface of head vessels in foxf2b morphant embryos

A,B: Cross section TEM image of carotid arteries in control (A) and foxf2b morphant (B) embryos at 48hpf. Green pseudocolouring outlines endothelial cells, yellow indicates mural cells and red indicates erythrocytes. Control vessel shows peg-and-socket junction, characteristic of endothelial- mural cell interaction (arrowhead). Magnified regions depict tight junctions, typical of endothelial cells. C: Bar graph depicting percent of abluminal vessel surface in contact with mural cells of all uninjected control and foxf2b morphant vessels quantified (p=0.0087). D: Bar graph showing same measurement as C, but compared between the PICA (p=0.0462) and dorsal aorta (p>0.05). E: Diagram depicting region of 48hpf zebrafish in which imaged sections were sampled from. Scale bars = 2µm.

Figure 4.10 – Sequences of TALEN-generated foxf2b mutations, genotyping and predicted resultant protein product

A: TALEN target sequence alignment of wildtype, Ca22 and Ca23 alleles. B: Sequencing chromatograms depicting wildtype sequence compared to genomic DNA and transcript sequence of foxf2b Ca22 and Ca23 mutant alleles. Ca22 sequencing shows insertion of a guanine nucleotide in genomic and transcript sequence. Ca23 sequencing shows overall loss of 14 nucleotides in genomic and transcript sequence. Resultant frameshift is transcribed in both mutations. C: Gels depicting genotyping assay of Ca22 and Ca23 alleles. D: Western blot of FoxF2b in wildtype, FoxF2bCa22 and FoxF2bCa23 mutants, with differing loading amounts (5µL and 15µL). β-tubulin used as loading control. E: Diagram depicting predicted wildtype and foxf2b mutant protein sequence. Thick rectangles represent exons, white boxes represent 5’ and 3’ untranslated regions. Both alleles cause a frameshift protein sequence beginning before DNA binding domain (in yellow), and results in a premature STOP codon. 90

TALEN RNA yielded mutations in the founder generation as verified by high resolution melt

(HRM) analysis, and adult mosaic F0 mutants were identified. These mutants were outcrossed to wildtype lines to generate F1 fish, demonstrating mutations were present in the germline and heritable. I identified two alleles: a single base pair insertion, allele Ca22 and a 14 base pair deletion allele Ca23 (Fig. 4.10 A,B). Sequencing of genomic DNA and cDNA (generated from mutant mRNA) demonstrated that the mutation is heritable and the gene is still actively transcribed (Fig. 4.10 B). Genotyping assays were developed using digestion of a PCR product with HinfI (the restriction enzyme site is only present in the FoxF2bCa22 mutant allele) or by resolving a 14bp difference in the FoxF2bCa23 allele in a PCR product (Fig. 4.10 C). Both alleles can be clearly distinguished from wildtype. Western blotting using the antibody against epitope 4 shows a FoxF2b protein decrease in the FoxF2bCa22 allele, but slight to no difference in the FoxF2bCa23 allele (Fig. 4.10 D). β-tubulin levels do not appear to be largely different between alleles, suggesting protein amounts loaded were close to equal. Both mutations are predicted to cause a frameshift resulting in a short missense protein sequence ending in a premature stop, disrupting the forkhead DNA binding domain (Fig. 4.10 E).

4.2.4 Differential embryonic hemorrhage phenotypes between foxf2bCa22 and Ca23 mutant alleles

Mutant embryos for both alleles were screened at 48hpf for hemorrhage blind, followed by genotyping (Fig 4.11). For the Ca22 allele, embryos were offspring of foxf2b+/Ca22 x foxf2bCa22/Ca22 crosses. For the Ca23 allele, embryos were offspring of a foxf2b+/Ca23 incross. Samples of hemorrhaged and non-hemorrhaged embryos were taken for genotyping to determine if the mutants were over-represented in the hemorrhaged group, based on expected Mendelian distributions (Fig. 4.11). The Ca22 allele showed fewer homozygous mutants than heterozygous mutants in the non-hemorrhaged samples, which deviated significantly from an expected proportion of 1:1 (χ2=4.592, n=48, p<0.05) but equal distribution of heterozygotes and homozygotes in hemorrhaged samples, which exactly fits expected Mendelian proportions (χ2=0.024, n=41, p=0.876). Interestingly, the Ca23 allele showed relatively standard Mendelian distribution of alleles in both sample populations (No hemorrhage – χ2=0.806, n=62, p=0.668;

Figure 4.11 – Allele-specific differences in hemorrhage in foxf2b mutant embryos

A: Bar graph and table depicting contributions of each genotype to hemorrhaged or non- hemorrhaged samples of zebrafish progeny from foxf2b+/Ca22 x foxf2bCa22/Ca22 crosses. B: Bar graph depicting contributions of each genotype to hemorrhaged or non-hemorrhaged samples of zebrafish progeny from foxf2b+/Ca23 incrosses. Tables show observed vs. expected counts and whether samples deviate from Mendelian inheritance patterns.

hemorrhage – χ2=3.29, n=62, p=0.193) suggesting that hemorrhages do not occur more frequently in foxf2b+/Ca23 or foxf2bCa23/Ca23 genetic mutant embryos. One caveat to these experiments is that sample sizes were based on the number of hemorrhaged embryos, and were thus lower than desired, and replicates had to be pooled to perform statistical analysis.

4.2.5 FoxF2b is necessary for initial mural cell coverage of penetrating cerebral vessels during development

To determine if foxf2bCa23 mutants presented a vascular stability phenotype despite a lack of hemorrhage, embryos were examined for acta2:GFP-positive mural cell coverage of cerebral vessels at 102 and 174hpf (roughly 4 and 7dpf) (Fig. 4.12). I developed three quantitative analyses of acta2:GFP fluorescent signal: a) counts of how many vessel branch points were covered with at least one acta2-positive cell, starting from the base of the brain and working dorsally, b) counts of acta2-positive cells/nuclei, and c) measurements of the distance that acta2:GFP-positive cells extend along cerebral vessels starting at the base and measuring the longest linear distance. At 102hpf (Fig. 4.12 A-F), wildtypes show an average of 7.73 ±0.69 branching points while mutants have an average of 5.06 ±0.91 (p=0.028 by two-tailed t test), wildtypes show an average of 17.33 ±0.92 acta2:GFP+ cells while mutants have an average of 13.13 ±1.6 (p=0.034 by two-tailed t test), and wildtypes show an average acta2:GFP+ coverage distance of 191.7 ±7.62 µm from the base vessel while mutants show an average distance of 138.6 ±12.44 µm (p=0.001 by two-tailed t test, wildtype n=15, foxf2bCa23/Ca23 n=16, 3 experimental replicates for all). Thus, all measurements were decreased in foxf2bCa23/Ca23 mutants compared to wildtype siblings, demonstrating a general decrease in the number and coverage of acta2-expressing cells associating with cerebral vessels. However, when examined at 174hpf (Fig. 4.12 G-L), there is no significant difference between mutant and wildtype sibling acta2:GFP coverage of cerebral vessels (wildtype branching points: 6.82 ±1.04, mutant branching points 8.00 ±0.73, p>0.05; wildtype acta2:GFP+ cells: 12.18 ±1.34, mutant acta2:GFP+ cells: 14.80 ±1.01, p>0.05. wildtype acta2:GFP+ longest extension: 185.0 ±15.33µm, mutant acta2:GFP+ longest extension: 183.8 ±8.38, p>0.05. wildtype n=11, mutant n=20, 3 experimental replicates). Interestingly, similar measurements in foxf2bCa22/Ca22 mutants

Figure 4.12 – Initial decrease of cerebral vessel coverage by smooth muscle in foxf2bCa23/Ca23 mutants recovers by 1 week of age.

A-C: Representative images of acta2:GFP-expressing cells on cerebral vessels at 102hpf depicting various measurements of cerebral vessel coverage, including number of branching points (A), number of acta2+ cells (B), and length of longest acta2+ coverage distance along vessel (C). D-F: Bar graphs depicting number of acta2:GFP+ branching points (D, p=0.00281), number of acta2+ cells (E, p=0.0344), and longest acta2+ coverage distance (F, p=0.0012) compared between foxf2bCa23/Ca23 mutant and wildtype sibling embryos. G-I: Same as A-C, but at 174hpf. J-L: Same as D-F, but at 174hpf. 94

showed no significant difference in number of acta2:GFP+ branching points (wildtype: 2.12 ±0.441, foxf2bCa22/Ca22: 2.69 ±0.414, p>0.05), acta2:GFP+ cell bodies (wildtype: 9.12 ±1.043, foxf2bCa22/Ca22: 9.54 ±0.972, p>0.05) or coverage length from base vessel (wildtype: 109.3 ±21.14µm, foxf2bCa22/Ca22: 131.2 ± 11.67µm, p>0.05, wildtype n=8, foxf2bCa22/Ca22 n=13 for all, Fig. 4.13).

4.2.6 Cerebral vessel permeability is unaffected in foxf2b mutants

Given the decrease in acta2:GFP coverage in foxf2bCa23/Ca23 mutants, I next investigated whether this also corresponded to an increased general permeability of vessels in the head. I injected a mixture of a high molecular weight dye (dextran rhodamine, 10,000 MW) and low molecular weight dye (DAPI, 30 Da) to examine the extent of vessel permeability. Embryos were injected into the common cardinal vein and allowed 30 min to circulate dyes, then imaged dorsally using confocal microscopy (Fig. 4.14). No dextran rhodamine or DAPI was detected outside of cerebral vessels in either foxf2bCa22/Ca22 or foxf2bCa23/Ca23 mutants, but was readily visible within vessel lumen or endothelial nuclei. This suggests that foxf2b mutant vessel integrity is generally unaffected, and hemorrhage is stochastic but vessel integrity is restored after an event.

4.2.7 foxf2a expression pattern is expanded in foxf2b mutants

Phenotypes of the foxf2b genetic mutants are milder than those observed in morphants. Although one explanation could be toxicity or off targeting of the morpholino, it is also possible that there is genetic compensation in the foxf2b mutants. Given that foxf2a is paralogous to foxf2b as it arose from a whole-genome duplication event, and its expression pattern is largely identical, foxf2a may contribute to compensation in the foxf2b mutant zebrafish lines. foxf2a expression was compared between FoxF2bCa22/Ca22 mutants and wildtype embryos to determine if there was an increase in transcript (Fig. 4.15). At 48hpf, foxf2a-positive periocular mesenchyme staining is separated from foxf2a-positive paraxial mesenchyme in wildtype. Similarly anterior and posterior regions of pharyngeal arch mesenchyme have distinct and separated expression domains. I observed that foxf2a is expanded into this expression gap in FoxF2bCa22/Ca22 mutants 95

Figure 4.13 – Cerebral vessel coverage by smooth muscle is unchanged in foxf2bCa22/Ca22 mutants at 102hpf.

A-C: Representative images of acta2:GFP-expressing cells on cerebral vessels at 102hpf depicting various measurements of cerebral vessel coverage, including number of branching points (A), number of acta2+ cells (B), and length of longest acta2+ coverage distance along vessel (C). D-F: Bar graphs depicting number of acta2:GFP+ branching points (D), number of acta2+ cells (E), and longest acta2+ coverage distance (F) compared between foxf2bCa22/Ca22 mutant and wildtype sibling embryos.

Figure 4.14 – Brain vessels do not show increased permeability to high or low molecular weight dyes in foxf2b mutants

Injections of DAPI/rhodamine dextran mixtures circulated through 4dpf wildtype (A), foxf2bCa22/Ca22 (B) and foxf2bCa23/Ca23 (C) embryos for 30 min, then imaged dorsally. Arrowheads point to DAPI staining of nuclei in endothelial cells, but no DAPI is observed outside of the vessels in brain parenchyma. Scale bars = 50µm.

Figure 4.15 – foxf2a expression domain is expanded in foxf2bCa22/Ca22 mutant embryos

A,A’: Wildtype expression of Foxf2a imaged laterally shows separation of anterior and posterior domains (arrowheads in A), and dorsal images show a lack of expression in the midline region of the embryo (arrow in A’). B,B’: Expression domains of foxf2a expand into regions with no/low expression in foxf2bCa22/Ca22 mutants, disrupting discrete expression domains of periocular and paraxial head mesenchyme, as well as anterior/posterior separation of pharyngeal arch mesenchyme (B, arrowheads). Expression is also seen in the midline of foxf2b mutants (B’), as well as ectopic expression in the trunk (D’, white arrow).

(Fig. 4.15 A,B). The foxf2a expression pattern in mutant embryos is also expanded into the midline (Fig. 4.15 C). Additionally, presumptive foregut expression of foxf2a appears to encompass a larger region in foxf2bCa22/Ca22 mutants (Fig. 4.15 D). This suggests that tissues are responding to a lack of functional FoxF2b potentially by upregulating expression of FoxF2a.

4.2.8 Mural cell marker pdgfrβ is specifically reduced at 4dpf in foxf2b mutants

As pericytes are vascular mural cells I wanted to test whether they are affected by FoxF2b loss of function. Expression of pericyte markers pdgfrβ, ng2 and notch3 was examined by ISH in foxf2bCa22/Ca22 embryos at 48hpf and in foxf2bCa23/Ca23 mutant embryos 4dpf, with age- matched wildtype controls. In the case of pdgrβ at 4dpf, wildtype controls were siblings. ISH expression patterns of pdgfrβ, ng2 and notch3 are unchanged between foxf2bCa22/Ca22 mutants and wildtype embryos at 48hpf (Fig. 4.16, A-F). Expression of pdgfrβ can be seen in pericytes of wildtype embryos at 4dpf, however this pattern is lost in foxf2bCa23/Ca23 mutants (Fig. 4.16 G,J). Expression in pharyngeal arches and palatal region is also slightly decreased, but to a lesser extent than in the head. Expression patterns of ng2 and notch3 are unchanged between wildtype and mutant embryos (Fig. 4.16 H,I,K,L). I quantified pdgfrβ-expressing cells in wholemount heads by cell counts in dorsal and lateral images (Fig. 4.16 M). Both views of embryos showed a striking decrease in the number of cells expressing pdgfrβ (dorsal wildtype: 43.25 ±4.91 cells, dorsal foxf2bCa23/Ca23 mutants: 28.36 ±4.34 cells, p = 0.035; lateral wildtype: 47.33 ±3.9 cells, lateral foxf2bCa23/Ca23 mutants: 18.91 ±2.59 cells, p < 0.0001; wildtype n=12, foxf2bCa23/Ca23 mutant n=11). Loss of pericyte markers occurs at the same developmental stage as when I observe decreased smooth muscle coverage described earlier. Despite hemorrhage at 48hpf, pericyte expression of pdgfrβ is absent from the cerebral region of the zebrafish at 48hpf, suggesting that vascular mural cells do not yet express this marker. Regardless, pdgfrβ expression patterns remain unchanged between wildtype and mutants, suggesting that the expression differences seen at 4dpf are due to a specific effect of FoxF2b on pericytes sometime after 48hpf.

Figure 4.16 – pdgfrβ expression is decreased at 4dpf in foxf2b mutants, other pericyte markers are unchanged

A-F: Pericyte markers pdgfrβ (A,D), ng2 (B,E) and notch3 (C,F) are unchanged at 48hpf between wildtype (A-C) and foxf2bCa22/Ca22 embryos. G,J: pdgfrβ ISH marks pericytes in 4dpf embryos (arrows, G), but the pattern is lost in foxf2bCa23/Ca23 embryos (J). H,I,J,K: ng2 (H) and notch3 (I) ISH do not mark pericytes at 4dpf, and patterns are unchanged in foxf2bCa23/Ca23 (J,K). Scale bars = 100µm. M: Bar graph depicting lateral and dorsal counts of pdgfrβ-expressing cells in wildtype and foxf2bCa23/Ca23 4dpf embryos (dorsal p=0.035, ventral p<0.0001). 100

4.2.9 The related foxc1 gene also promotes vessel coverage and mature muscle cell markers

foxc1 is closely linked to foxf2 in the conserved cluster of Fox transcription factors, and collaborative work with Dr. Ordan Lehmann and Dr. Curtis French at the University of Alberta led to a published study investigating the role of FoxC1 in vascular stability (French et al., 2014). Our collaborators showed that foxc1a/b knockdown by morpholino causes hemorrhage in 48hpf zebrafish. To determine if loss of foxc1a/b resulted in similar mural cell phenotypes to foxf2b knockdown, I examined levels of acta2:GFP by confocal imaging in foxc1a/b double morphants. Tg(acta2:GFP)ca7 foxc1a/b morphant and control fish were imaged ventrally and the numbers of acta2:GFP+ cell bodies on pharyngeal arch vessels were counted (Fig. 4.17). At 4dpf, the number of acta2-expressing cells was decreased by approximately 15% in foxc1a/b morphants when compared to un-injected control embryos (control: 42.08 ±2.48 acta2:GFP+ cells, n=12; foxc1 morphants 31.00 ±3.94 acta2:GFP+ cells, n=11; p=0.0244; Fig. 4.17 A-C), similar to results seen in foxf2b morphants. I next examined vessel ultrastructure using TEM. I observed an increase in the numbers and sizes of perivascular spaces around the dorsal aorta and pharyngeal arch arteries, as well as a general decrease of mural cell number and contact with endothelial cells of the vessels in foxc1a/b morphants (Fig. 4.17 D-I). Both of these experiments indicate that FoxC1 is also important for the development of mural cells and their interactions with endothelial cells, and FoxC1 may have the potential to contribute to compensatory responses in foxf2b genetic mutant.

4.2.10 foxq1a/b expression is decreased in foxf2b mutants

foxq1 is a single-exon gene that is also part of the evolutionarily conserved cluster of Fox genes that includes foxf2 and foxc1. foxq1 is involved in EMT in normal development and in cancer (Qiao et al., 2011; Fan et al., 2014), and is a target of Wnt and PDGF signaling (Christensen et al., 2013; Meng et al., 2015). Little is known about foxq1 in zebrafish, other than it is potentially upregulated in jaw primordium in response to drug treatments (Planchart & Mattingly, 2010). Given that both foxf2 and foxc1 downregulation is associated with cerebral

101

Figure 4.17 – Morpholino knockdown of foxc1 results in decreased acta2:GFP expression and poor vessel support by perivascular mural cells

A,B: Ventral view of control (A) and foxc1 morphant (B) 4dpf embryos showing decreased acta2:GFP expression (arrowheads) on pharyngeal vessels of morphant embryos compared to control. Scale bars = 50µm. C: Bar graph depicting acta2:GFP+ cell counts between control and FoxC1 morphant embryos (p=0.0244). D-I: Transmission electron micrographs of cerebral vessels at 52hpf demonstrate increased perivascular spaces in cross sections at the level of the hindbrain in foxc1 morphants. Asterisks denote acellular areas adjacent to endothelial cells, suggesting detachment of vessels from surrounding cells. Vessels imaged are dorsal aorta (D,G) and pharyngeal arch artery (E,F,H,I). e = erythrocyte, ec = endothelial cell. Scale bar = 2µm.

102

Figure 4.18 – Expression of foxq1a/b is absent in foxf2b morphants

A,A’: ISH of foxq1a in wildtype can be seen in pharyngeal arch mesenchyme (arrowheads) and periocular mesenchyme (arrows) at 48hpf. Expression is absent in foxf2b morphants (A’). 48hpf ISH of foxq1b shows expression in the pharyngeal arch mesenchyme (arrowheads) in wildtype (B), and the pattern is absent in foxf2b morphants (B’). Scale bars = 200.

103

small vessel disease, foxq1 could also be involved in the same biological processes responsible for these diseases. foxq1a/b is expressed in mesenchymal tissues surrounding the oral cavity at 50hpf, similar to foxf2b (Fig. 4.18 A, B). I observed that expression is completely lost in FoxF2b morphant embryos (Fig. 4.18 A’,B’). My results suggest that FoxF2b may regulate expression of related Fox genes.

4.2.11 Examination of vascular mural cell markers and potential target gene expression in foxf2b mutants and morphants

To determine potential mechanisms that FoxF2 may be regulating, I qualitatively compared expression patterns and levels of genes involved in biological processes pertaining to vascular stability between morphants/mutants and age-matched controls. Three main developmental stages were examined: 24-30hpf, 48hpf and 4dpf (96hpf). 24-30hpf is an early window of developing vasculature, where initial vasculogenesis of the major axial vessels is complete and circulation has begun. Angiogenic sprouts are undergoing major morphogenic and patterning events, and are beginning to interact with developing mural cells. 48hpf is the stage at which hemorrhages are scored, the flow rate and blood pressure has increased. 4dpf is the stage at which expression of our acta2:GFP transgene is first seen on/around vessels, an indication that mural cells are entering final stages of maturation. The majority of major vessels have formed by this stage as well, and blood flow is actively regulating further vascularization of the brain (Bussmann et al., 2011). The genes described below were examined at one or more of these timepoints as appropriate to their roles and expression patterns, and I found several differences in expression patterns in foxf2b morphants and mutants.

4.2.11.1 nkx3.2

The transcription factor nkx3.2 is regulated similarly to foxf2 downstream of Sonic Hedgehog signalling as identified on the igu microarray (Arnold et al., 2015). It is an effector of chondrogenesis (Miller et al., 2003). The Drosophila nkx3.2 homolog, bagpipe, interacts with the FoxF homolog, biniou, to specify visceral muscle, a muscle type similar to smooth muscle (Jakobsen et al., 2007). Preliminary data in our lab suggests nkx3.2 also plays a role in vascular

104

Figure 4.19 – nkx3.2 expression is unchanged in foxf2b mutants and morphants

A,B: nkx3.2 expression is seen in the developing jaw joint (arrowhead) and anterior notochord (arrow) of 30hpf embryos. Expression is the same between control (A) and foxf2b morphant (B) embryos. C-E: nkx3.2 expression at 48hpf can be seen strongly in the jaw joint (arrowhead) and pharyngeal arches (arrows). These discrete expression patterns are unchanged in foxf2bCa22/Ca22 and foxf2bCa23/Ca23 mutants. Diffuse head staining is variable within control and morphant groups, and not representative of nkx3.2 expression patterns. Scale bars = 200µm.

105

stability (Spice, 2014). I examined expression of nkx3.2 in foxf2b knockdown/knockout models. At 30hpf, nkx3.2 is expressed in the notochord and head mesenchymal tissues. In foxf2b morphants, no difference is seen in the expression pattern (Fig. 4.19 A,B). At 48hpf, nkx3.2 expression is distinct in the pharyngeal arches, and the developing jaw joint (Fig. 4.19 C). Age- matched foxf2b Ca22 and Ca23 homozygous mutants show no difference in staining patterns in either area (Fig. 4.19 D,E). These results suggest that nkx3.2 is not transcriptionally regulated by foxf2b.

4.2.11.2 Connective Tissues - col4a, ctgfa

Extracellular matrix (ECM) is an important component of vascular stability. ECM proteins are found in the basement membrane surrounding vessels, and in mice, pericytes are embedded within the basement membrane on the abluminal surface of the vessels. ECM components of the mouse gut, particularly collagens, are downregulated in FoxF2 mutants (Ormestad et al., 2006). Type IV collagens form the basal lamina which surrounds blood vessels, and ColIV mutations are genetically linked to intracerebral hemorrhage and small vessel disease (Gould et al., 2005; Sibon et al., 2007; Kuo et al., 2012). I examined expression of col4a2 in FoxF2b mutants to determine if a similar lack of ECM component transcript occurred around the vasculature (Fig. 4.20). col4a2 at 48 hpf is expressed around major vessels of the trunk and intersegmental vessels, and in major vessels of the ventral head (Fig. 4.20 A). The general pattern of col4a2 remains the same between wildtype and foxf2bCa22/Ca22 embryos, but the intensity of staining around the major head vessels is decreased in mutant embryos (Fig. 4.20 B), suggesting that FoxF2 may regulate collagen IV expression. Connective tissue growth factor (CTGF) regulates ECM synthesis in mouse (Duncan et al., 1999), contributing to interactions between pericytes and endothelial cells. Given the apparent regulation of collagen by FoxF2, I inquired as to whether FoxF2 controls expression of upstream components such as CTGF to promote vascular stability. In zebrafish, ctgfa is expressed in the floor plate at 30hpf (Fig. 4.20 C). However, expression remains unchanged in foxf2b morphant embryos (Fig. 4.20 D). Based on these results, FoxF2b does not appear to control collagen expression via CTGF. 106

Figure 4.20 – Expression of col4a2 is decreased in foxf2b mutants, but ECM-related growth factor ctgfa is unchanged

A,B: col4a2 expression is present around large vessels of the trunk and head, and can be seen faintly on cerebral vessels as well (arrows). The col4a2 expression domain in 48hpf foxf2bCa22/Ca22 mutants (B) is unchanged compared to wildtype (A), but overall staining intensity is weaker than wildtype, particularly in the larger vessels of the head (arrowheads). C,D: ctgfa expression in the floor plate is unchanged between 30hpf control (C) and foxf2b morphant (D) embryos. Scale bars = 100µm.

107

4.2.11.3 Neural crest and EMT – snai2, hand2

Pericytes and smooth muscle of the cerebral vasculature in part arise from neural crest (Etchevers et al., 2001; Korn et al., 2002; Simon et al., 2012; Cavanaugh et al., 2015). Neural crest requires strict and efficient regulation of epithelial-to-mesenchymal transition (EMT) morphogenic events (and subsequent mesenchymal-to-epithelial reversion). Some studies have shown increased EMT activity in cancer models where FoxF2 function is decreased (Cai et al., 2015; Wang et al., 2015; Zhang et al., 2015; Shi et al., 2016). To determine if FoxF2 is bridging a gap between these processes by playing a similar EMT regulatory role with neural crest development, expression of neural crest-specific genes were investigated. snai2 is a member of the Snail transcription factor family, factors responsible for neural crest differentiation and migration (Thisse et al., 1995; Carl et al., 1999; LaBonne & Bronner- Fraser, 2000). In 30hpf zebrafish embryos, it is expressed in the developing pharyngeal arches. However the expression pattern and intensity is not altered in foxf2b morphant embryos (Fig. 4.21 A,B). hand2, a bHLH transcription factor, is active in late-stage neural crest to promote development of neural crest-derived tissues and cardiac mesoderm (Srivastava et al., 1997). In 30hpf zebrafish, it shows a very distinct expression pattern in the pharyngeal arches and developing heart field (Fig. 4.21 C). In foxf2b morphants hand2 is still expressed in the heart field and pharyngeal arches, but expression in the most posterior arch is absent (Fig. 4.21 D), suggesting that FoxF2b may regulate hand2 to promote neural crest EMT, potentially in the later stages of mesenchymal reversion to ordered or epithelial tissues.

4.2.11.4 PDGF signaling – pdgfrα, pdgfaa, pdgfba

Although pdgfrβ is required for initial pericyte recruitment and proliferation, it continues to be expressed after pericytes have established interactions with the endothelium and astrocytic endfeet, thus forming a functional blood brain barrier. Thus, active signaling to the receptor by PDGF-B may be a more accurate indicator of pericyte proliferation and migration. PDGF-A and PDGFRα have also been shown to affect mural cells, being induced in vSMCs undergoing phenotypic modulation to a synthetic and proliferative state (Resink et al., 1990; Sjolund et al., 108

Figure 4.21 – foxf2b morphant expression of neural crest gene snai2 is unchanged, but hand2 expression indicates loss of pharyngeal arch structures

A,B: snai2 is expressed in pharyngeal arch mesenchyme at 30hpf (arrowheads). Expression patterns are the same between wildtype (A) and foxf2b morphant (B). C,D: hand2 expression at 30hpf is in discrete domains corresponding to developing pharyngeal arches (arrows). In foxf2b morphants, the most posterior hand2 expression domain is absent (D). Scale bars = 100µm.

109

Figure 4.22 – Expression of PDGF signaling components in foxf2b morphant and mutant embryos

A,A’: pdgfrα expression in ventral mesenchyme (arrowheads) and pharyngeal arch tissues (arrows) in 30hpf control and foxf2b morphant embryos. B,B’: pdgfrα expression in pharyngeal arch mesenchyme and periocular mesenchyme at 48hpf (B). Expression domain is expanded in foxf2bCa22/Ca22 mutants (B’, white arrow). C,C’: pdgfrα expression in palatal mesenchyme, pharyngeal arch region (arrows) and larger cerebral vessels (white arrows) at 4dpf. D,D’: pdgfaa expression in pharyngeal arches and optic stalk at 48hpf. Diffuse signal is seen in the head of in foxf2bCa22/Ca22 mutants (D’). E,E’: pdgfaa expression at 4dpf is largely ubiquitous, with some specific signal in palatal mesenchyme and otic vesicle in wildtype (E, arrow). F,F’: pdgfba expression is diffuse in head and pharyngeal region at 48hpf. G,G’: pdgfba expression is largely ubiquitous at 4dpf, except for specific signal in palatal mesenchyme and posterior to the otic vesicle in wildtype (G, arrow). os = optic stalk, pam = pharyngeal arch mesenchyme, pm = palatal mesenchyme, pom = periocular mesenchyme. Scale bars = 100um. 110

1990; Jin et al., 2002; Kwon et al., 2015). A study by Bolte et al. identified the PDGF signaling components PDGFRα, PDGF-A and PDGF-B to be upregulated in the mouse gut after smooth muscle-specific ablation of FoxF2 (Bolte et al., 2015). Given the differences in pdgfrβ expression in foxf2b mutants, it seemed appropriate to investigate the expression patterns of other PDGF signaling molecules. In developing zebrafish, pdgfrα is expressed in the head mesenchyme and pharyngeal arches of zebrafish at 30 and 48hpf (Fig. 4.22 A,B). At 4dpf, specific staining can be seen in large vessels of the head, in the palate, and in the pharyngeal arches (Fig. 4.22 C). In 30hpf foxf2b morphants, no significant difference in pdgfrα expression is observed (Fig. 4.22 A’). However, at 48hpf, pdgfrα staining is potentially more intense and expansive in foxf2bCa22/Ca22 mutants than in wildtypes (Fig 4.22 B’). At 4dpf, pdgfrα expression in foxf2bCa23/Ca23 mutants is similarly upregulated in the pharyngeal arch area (Fig. 4.22 C,C’), suggesting FoxF2b normally represses pdgfrα in these tissues. Two PDGF ligand expression patterns were examined, pdgfaa and pdgfba. pdgfaa shows strong specific expression in the pharyngeal arches and the optic stalk at 48hpf (Fig. 4.22 D), but pdgfba shows a diffuse head expression without discrete expression domains (Fig. 4.22 F). Both genes show an increase in diffuse head expression in 48hpf FoxF2bCa22/Ca22 mutants (Fig. 4.22 D’,F’). At 4dpf, both genes still have diffuse staining in the head, except for some palate-specific signal, which remains unchanged in FoxF2bCa23/Ca23 mutants (Fig. 4.22 E,E’,G,G’). There is, however, a small region of staining just anterior to the otic vesicle which is more strongly expressed in wildtypes than in mutants. Overall, the expression of pdgfaa and pdgfba does not appear to be significantly changed between wildtype and mutant embryos at 4dpf.

4.2.11.5 Head mesenchyme and palate markers – tbx15, tbx18, hic1

Neural crest and lateral mesoderm are important for development of craniofacial structures. Both tissue types contribute to head mesenchyme tissues, which further differentiates to form bones, cartilage, connective tissue and muscle of the head. FoxF2 is expressed in the head mesenchyme, and may be playing a role in promoting development of craniofacial features. For instance, palatal mesenchyme is essential for formation of the palate, and FoxF2 mutations are linked to cleft palate defects (Wang et al., 2003a; Jochumsen et al., 2008). I also note that the 111

Figure 4.23 – Allele-specific differences of tbx15 expression in foxf2b mutants

A-A”: Dorsal view of tbx15 expression in wildtype and foxf2b mutant embryos at 24hpf in paraxial head mesenchyme. foxf2bCa23/Ca23 mutants show reduced expression in anterior (arrowheads) and posterior (arrows) domains. B-B”: Dorsal view of tbx15 expression in wildtype and foxf2b mutant embryos at 48hpf in paraxial head mesenchyme and periocular mesenchyme. foxf2bCa22/Ca22 mutants show decreased expression in periocular mesenchyme (arrowheads). C-C”: Lateral view of tbx15 expression in wildtype and foxf2b mutant embryos at 4dpf faintly in palatal mesenchyme. Scale bars = 200µm.

112

Figure 4.24 – Allele-specific differences of tbx18 expression in foxf2b mutants

A-A”: Lateral view of tbx18 expression in wildtype and foxf2b mutant embryos at 24hpf in head mesenchyme (arrowheads) and presumptive fin bud (arrows). B-B”: Lateral view of tbx18 expression in wildtype and foxf2b mutant embryos at 48hpf in periocular mesenchyme (arrowheads), paraxial mesenchyme (black arrows) and fin bud (white arrows). C-C”: Lateral view of tbx18 expression in wildtype and foxf2b mutant embryos at 4dpf in palatal mesenchyme (arrowheads). Scale bars = 200µm.

113

expression of foxf2b in zebrafish palatal mesenchyme is directly adjacent to acta2-expressing cells on the penetrating vessels of the brain. This suggests an alternative hypothesis for the role of FoxF2 in mural cell development may be that it can act non-autonomously in adjacent cells early in the migration/differentiation cascade. tbx18 has known roles in coronary smooth muscle differentiation and heart development (Cai et al., 2008a; Grieskamp et al., 2011; Greulich et al., 2012; Wu et al., 2013), whereas tbx15 has a lesser studied role in skeletal muscle formation (de Wilde et al., 2010; Lee et al., 2015). Both genes are required for mesodermal development, and are expressed in mesenchyme of the head (Kraus et al., 2001; Begemann et al., 2002; Singh et al., 2005; Farin et al., 2008), and mutations in tbx15 are linked to facial abnormalities (Lausch et al., 2008). Both genes were identified through RNAseq as overrepresented transcription factors in pericytes (Zhang et al., 2014). In zebrafish embryos, tbx15 is expressed in paraxial head mesenchyme at 24hpf (Fig. 4.23 A-A”), whereas tbx18 is also expressed in presumptive fin buds (Fig. 4.24 A-A”). I found allele-specific differences in the expression of these two genes. tbx15 expression is unchanged in foxf2bCa22/Ca22 mutants (Fig. 4.23 A’), but is decreased in head mesenchyme of foxf2bCa23/Ca23 mutants at 24hpf (Fig. 4.23 A”). Similarly, tbx18 expression is unchanged in foxf2bCa22/Ca22 mutants (Fig. 4.24 A’), but head mesenchyme in 24hpf foxf2bCa23/Ca23 mutants (Fig. 4.24 A”). At 48hpf, both tbx15 and tbx18 are expressed in paraxial mesenchyme and periocular mesenchyme, and tbx18 can also be seen in the fin bud (Fig. 4.23 & Fig. 4.24 B-B”). The pattern of mesenchymal tbx15 staining is slightly reduced in foxf2bCa22/Ca22 periocular mesenchyme, while tbx18 is significantly reduced in both mesenchymal tissues in the same mutants (Fig. 4.23 & 4.24 B’). foxf2bCa23/Ca23 mutant expression of both genes at 48hpf is unchanged (Fig. 4.23 & 4.24 B”). At 4dpf, tbx15 is faintly expressed in the palatal mesenchyme, while tbx18 is more strongly expressed in the mesenchyme and two small bilateral spots of expression slightly dorsal and posterior to the palate, potentially the remainder of the paraxial mesoderm (Fig. 4.23 & 4.24 C-C”). tbx15 remains unchanged in mutants (Fig. 4.23 C’,C”). Palatal expression of tbx18 is significantly reduced in foxf2bCa23/Ca23 mutants, but is unchanged in foxf2bCa22/Ca22 mutants (Fig.

114

4.24 C’,C”). It is interesting that when expression changes occurred in one mutant allele at each stage, the other allele showed expression patterns similar to that of wildtype, for both tbx15 and tbx18. The tumor suppressor gene hic1 (Wales et al., 1995; Rood & Leprince, 2013) is expressed in zebrafish head mesenchyme (Bertrand et al., 2004), and was also identified as an over-represented transcription factor in pericytes (Zhang et al., 2014). I observe hic1 expression in periocular mesenchyme and cardiac-associated regions of pharyngeal arch mesenchyme as well as in the pronephric duct at 48hpf, in addition to developing foregut, which was previously not reported (Fig. 4.25 A) (Bertrand et al., 2004). At 4dpf, hic1 is expressed in the palatal mesenchyme (Fig. 4.25 B). I also found allele specific differences in expression of hic1 at different ages. In 48hpf foxf2bCa22/Ca22 mutants, pharyngeal arch mesenchyme and foregut hic1 expression is decreased, but pronephric duct signal appears more intense (Fig. 4.25 A’). foxf2bCa23/Ca23 mutant embryos have stronger staining intensity of hic1 in all tissues at 48hpf, in addition to medial expansion of the foregut expression domain (Fig. 4.25 A”). In foxf2bCa22/Ca22 at 4dpf, expression can be seen strongly in the palate and is expressed ectopically in ventral tissues along the length of the trunk, corresponding to gut and ventral mesenchyme. hic1 in foxf2bCa23/Ca23 shows a more restricted expression domain in the palatal mesenchyme at 4dpf (Fig. 4.25 B”). These differences in expression patterns of tbx and hic1 genes in different mutant alleles between 48hpf and 4dpf stages is quite striking, and is some of the first evidence that both alleles are functionally disruptive but have differential effects on potential downstream targets. These differences may suggest that FoxF2b is required to regulate formation and maintenance of mesenchymal tissues of the head and possibly gut.

4.2.11.6 Shh signaling feedback – gli2a, ptch2

Shh signaling commonly utilizes feedback mechanisms to regulate its activity. For instance, the Shh receptor patched is also a direct target of Gli transcription factors, the signalling effectors of the pathway. However other factors may be mediating the feedback loop to obtain spatio-temporal control of Hh signaling including FoxF2. Bolte et al. describe upregulation of Gli1, Gli2 and Ptch1 in mice deficient in FoxF2 (Bolte et al., 2015). I next asked 115

Figure 4.25 – Allele-specific differences of hic1 mesenchyme expression patterns in foxf2b mutants

A-A”: 48hpf ISH of hic1 shows wildtype expression in pharyngeal arch mesenchyme (arrowheads) and periocular mesenchyme (arrows), as well as foregut (white arrows) and pronephric duct (white arrowhead). B-B”: 4dpf ISH of hic1 shows expression in palatal mesenchyme (arrowheads), which is expanded into pharyngeal arch mesenchyme and pronephric duct (arrows) in foxf2bCa22/Ca22 mutants (B’). Scale bars = 200µm.

116

Figure 4.26 – gli2a and ptch2 expression is increased in foxf2 mutant embryos

A-B: gli2a expression is seen at the midbrain-hindbrain and forebrain-midbrain boundaries (arrows), and in pharyngeal tissues (arrowhead) in 48hpf wildtype (A,B), foxf2bCa22/Ca22 (A’) and foxf2bCa23/Ca23 (B’) embryos. C,C’: gli2a expression is relatively ubiquitous at 4dpf, but some specific staining domains can be seen in the cerebral boundaries (arrowheads), pharyngeal arches (arrows) and palate region (white arrows). D,D’: ptch2 expression can be seen faintly in a pattern similar to gli2a in wildtype embryos at 48hpf, (D, brain boundaries, arrows; pharyngeal tissues, arrowhead), and shows increased expression intensity in foxf2bCa22/Ca22 embryos (D’). Expression of ptch2 at 4dpf is relatively unchanged between wildtype and foxf2bCa23/Ca23 mutants (E,E’). Expression can be seen in the pharyngeal arches at 4dpf (arrows) Scale bars = 100µm.

117

whether FoxF2 regulates Shh signalling as part of a feedback loop by examining expression of the gli2a and ptch2 genes. gli2a appears upregulated in the head of both foxf2bCa22/Ca22 and foxf2bCa23/Ca23 mutants at 48hpf and 4dpf (Fig. 4.26 A-C). Increased expression appears most strongly in the pharyngeal arch region of both alleles at both stages, but the effect is subtle in the foxf2bCa23 allele at 48hpf. There is also increased expression in the palate of 4dpf embryos. ptch2 expression occurs in a pattern similar to gli2a at 48hpf (Fig. 4.26 D), yet appears more strongly in pharyngeal arches at 4dpf (Fig. 4.26 E). foxf2bCa22/Ca22 mutant embryos have increased intensity of ptch2 expression in all head tissues at 48hpf, particularly the brain boundaries and pharyngeal region (Fig. 4.26 D’), whereas foxf2bCa23/Ca23 4dpf mutant expression of ptch2 appears relatively unchanged (Fig. 4.26 E’). Together, this data suggests that FoxF2b normally acts to repress Shh signaling, particularly at earlier stages of vessel development.

118

Morphant FoxF2b Ca22 FoxF2b Ca23 Potential for Target ISH 24hpf-30hpf 48hpf 24hpf 48hpf 4dpf 24hpf 48hpf 4dpf Appeared on microarray Nkx3.2 Unchanged N/A N/A Unchanged N/A N/A Unchanged N/A Col4a2 N/A N/A N/A Slight decrease N/A N/A N/A N/A Basement Membrane CTGFa Unchanged N/A N/A N/A N/A N/A N/A N/A Snai2 Unchanged N/A N/A N/A N/A N/A N/A N/A Neural crest markers Decreased in Hand2 N/A N/A N/A N/A N/A N/A N/A pharyngeal arches Decreased in Pdgfrβ N/A N/A N/A N/A N/A N/A N/A pericytes Pdgfrα Unchanged N/A N/A Increased N/A N/A N/A Increased PDGF signaling Increased diffuse Pdgfaa N/A N/A N/A N/A N/A N/A Unchanged head signal Increased diffuse Pdgfba N/A N/A N/A N/A N/A N/A Unchanged head signal Tbx15 N/A N/A Unchanged Decreased Unchanged Decreased Unchanged Unchanged Head Mesenchyme Tbx18 N/A N/A Unchanged Decreased Unchanged Decreased Unchanged Decreased Markers Hic1 N/A N/A N/A Decreased Increased N/A Increased Decreased FoxQ1a N/A Decreased N/A N/A N/A N/A N/A N/A Fox Gene Cluster FoxQ1b N/A Absent N/A N/A N/A N/A N/A N/A Increased Increased Gli2a N/A N/A N/A Increased N/A N/A Shh signaling ventrally ventrally Ptch2 N/A N/A N/A Increased N/A N/A N/A Unchanged

Table 4.1 – Summary of ISH experiments in foxf2b morphants and mutants

119

4.3 Discussion

4.3.1 foxf2a/b expression patterns reveal a potentially mesodermal source of head mural cells

foxf2 expression first appears in head mesenchyme. Although the origins of head mesenchyme have not been identified in fish, based on other vertebrates, it is likely that head mesenchyme arises from both ventral mesoderm and neural crest. foxf2b is later expressed in tissues surrounding endothelial cells in the head, potentially in developing mural cells. foxf2b expression first wraps the mid-mesencephalic central artery (MMCtA). Noting the continuous domain of foxf2b expression from the MMCtA to the ventral mesenchyme, and that expression is initially seen in ventral tissues of the fish more proximal to the heart, I postulate that foxf2b marks mesenchymal cells that originate ventrally and migrate dorsally and anteriorly along vessels. However, as seen by foxf2 expression in the palate that is adjacent to sites of acta2 expression (a more definitive marker of smooth muscle) I cannot rule out that foxf2b-expressing cells of non-mural origin may interact to promote differentiation of mural cells or endothelium non-autonomously. Lineage tracing experiments are needed to identify tissue origin of head mesenchyme and tissue-specific knockout experiments are needed to understand cell autonomy of FoxF2 in vascular stabilization. Interestingly, foxf2b expression is not detected around ventral head vessels in later stages, when mature vSMC markers are expressed. Indeed, foxf2b and acta2:GFP expression appear to be mutually exclusive, suggesting that foxf2b is perhaps necessary for initiating differentiation of head mural cells but is turned off once differentiation has progressed past a certain point. My data provide identification of foxf2 expression around head vessels in an early developmental context, placing foxf2 in an ideal location for promoting mural cell-endothelial cell interactions.. A whole genome duplication event in teleosts has led to an increase in gene number in the zebrafish over mammals (26,206 genes in zebrafish vs. ~20,500 in human). While some duplicated genes have been lost from the genome, others have been retained in two copies located on different chromosomes, as have foxf2a and foxf2b. Slight differences are seen between the expression patterns of the paralogous foxf2 genes, starting around 36hpf. foxf2a can be seen in distinct mesenchymal tissue compartments at 36hpf, whereas foxf2b appears to be broadly expressed throughout all head mesenchyme. At 48hpf, foxf2b expression is restricted to the posterior regions of the pharyngeal arch mesenchyme, whereas foxf2a expression is seen throughout the pharyngeal arch mesenchyme. This suggests slightly different developmental roles for foxf2a and foxf2b, and may reflect subfunctionalization whereby the two copies have evolved different expression patterns and potentially

120

functional roles since they arose from the whole genome duplication event that occurred in fish. This is thought to be achieved by modification of regulatory elements. However, some regulatory elements may still be shared, given that the whole gene cluster (FoxF2, FoxC1, FoxQ1) was duplicated. I observed ectopic expression of foxf2a in foxf2b mutant embryos. The foxf2a expression domain expands into regions where neither gene is expressed at that stage. This may suggest that cells detect a loss of foxf2b expression and compensate by upregulation of foxf2a, however the mechanism of ectopic expansion of foxf2a expression is unknown. Not all compartmentalization is lost, as the normal foxf2a expression pattern is also observed. Further analysis by RT-PCR is required to quantify any increase in foxf2a expression. The opposite experiment of foxf2b expression in foxf2a genetic mutants would also prove useful in identifying the relationships between the two paralogs in specific mesenchymal subregions. Another interesting piece of evidence for foxf2b regulation of other genes in the Fox gene cluster is seen in the foxq1 ISH experiments. Expression of both foxq1 paralogs is completely lost from their normal pharyngeal expression domains, suggesting they are regulated by FoxF2b. The role of FoxQ1 in zebrafish has not been thoroughly examined, and there is no known vascular stability role for FoxQ1 in any other organism. Going forward, it would be interesting to also examine the expression patterns of foxc1 in foxf2 knockdown or knockout models to determine if there is perhaps a role for FoxF2 in regulation of the entire conserved Fox gene cluster.

4.3.2 Knockdown and genetic knockout of foxf2b results in mural-cell associated vascular stability defects

My project initiated from the observation that foxf2 was downregulated in the vascular stability mutant iguana. Using the hypothesis that foxf2 mediates vascular stability downstream of Shh signalling, I quantitated hemorrhage phenotypes at 48hpf in foxf2b morphants and foxf2bCa22/Ca22 mutants. Early experiments with morphants showed that their hemorrhage rate at 48hpf is significantly higher than uninjected controls, suggesting foxf2b is important for vascular stability. I note that this hemorrhage rate (average of 17%) is lower than what has been found in igufo10 mutants (25-45%) (Lamont, 2008). This may be due to incomplete knockdown of foxf2b, since Western blot shows diminished but not a complete absence of FoxF2b protein in injected embryos. Embryos that did not hemorrhage also show a decrease in FoxF2b protein, suggesting that the hemorrhage phenotype is not fully penetrant. In addition, in Chapter 3 I showed that multiple genes are downregulated in igu mutants that might play overlapping or independent roles in vascular stabilization. I hypothesize that

121

parallel knockdown of another vascular stability component in addition to FoxF2 would be needed to fully destabilize vessels. Knockdown of foxf2b and foxf2a was attempted with several morpholinos, and the “foxf2b i2e3” consistently gave a hemorrhage phenotype. With morpholino experiments there is always concern about potential off-target effects and non-specific toxicity. For this reason it is essential to replicate results using an independent morpholino. However I was unable to identify a second foxf2b targeting morpholino that gave a phenotype (although there is some indication that some of the foxf2b morpholinos may produce slightly higher hemorrhage rates than wildtype, phenotypes were not as robust). One caveat to this early work is that I only looked at hemorrhage and not at other markers of mural cell differentiation. There may have been a phenotype with these reagents that I missed. None of the foxf2a morpholinos gave hemorrhage, but for at least one of these I showed that effective knockdown of the gene was not occurring. I did not pursue this further as the development of targeted genetic mutation techniques such as TALEN and CRISPR shifted the focus in the field away from transient knockdown experiments. I generated two genetic mutant alleles of foxf2 with similar mutation site but slightly different sequences. One allele is a 1 bp insertion, while the second allele is a 14bp complex indel. Since the two alleles would be predicted to translate similar proteins (if they were translated), I was somewhat surprised to observe differences in phenotypes and gene expression patterns between the two alleles. These are outlined in the discussion below. Both alleles are homozygous viable and appear to survive in proportions similar to wild type (i.e. Mendelian ratios). However it is possible that the most strongly affected mutants die early and that the remaining fish have found some way to compensate for loss of FoxF2b and are therefore have milder phenotypes. Some of the differences between the two alleles may therefore reflect compensation mechanisms. To examine vascular stability in mutants, I was therefore very careful about the origin of the parents. For instance, I differentiated between homozygous parents that were incrossed (with homozygous wild types as controls) and heterozygous incrosses. For the latter experiments, all embryos were genotyped by PCR after analysis and experiments were therefore blinded. 48hpf mutant hemorrhage experiments were conducted in the progeny of either heterozygous-homozygous (FoxF2bCa22) or heterozygous-heterozygous (FoxF2bCa23) crosses. There is always a baseline level of hemorrhage in wildtype fish that is often enhanced in our vascular transgenics (fli1a and kdrl), and therefore all experiments were compared to fish of a similar genetic background. I looked at hemorrhage by isolating samples of embryos with or without hemorrhage in the population and 122

genotyped, hypothesizing that if hemorrhage is the result of FoxF2b mutation, samples of hemorrhaged and non-hemorrhaged embryos should show a non-Mendelian distribution of genotypes. Not all mutant animals hemorrhage but there is a lower proportion of mutants in the non-hemorrhaged sample for FoxF2bCa22 embryos. Interestingly, approximately equal proportions of FoxF2bCa22 heterozygote and homozygote embryos were present in hemorrhaged samples, which is the expected 1:1 distribution of a heterozygote-homozygote cross and would suggest that loss of FoxF2 does not confer a higher risk of hemorrhage. However in the non-hemorrhage group, homozygous mutants were strongly underrepresented. This suggests that haploinsufficiency of FoxF2b allele confers an elevated risk of hemorrhage. The caveats to this experiment are the relatively small number of fish and the lack of multiple replicates. Also, this cross excluded fully wildtype embryos, and repetition of this experiment with foxf2b+/Ca22 incrosses may shed more light on penetration of the FoxF2bCa22 allele. We know that human mutations of FoxC1 are haploinsufficient, and comparison of heterozygous and homozygous embryos may have shown phenotypes in both groups, hence the lack of difference. On the other hand, heterozygous crosses were used with the FoxF2bCa23 allele, and I found no increased risk of hemorrhage as they showed a generally Mendelian distribution in both hemorrhaged and non-hemorrhaged samples. There is an apparent slight decrease of foxf2bCa23/Ca23 mutants in hemorrhaged samples, however it does not reach statistical significance. Excessive embryonic death before 48hpf was not apparent, and as previously mentioned, progeny of heterozygous crosses showed normal Mendelian distribution of genotypes, suggesting this is not due to embryonic lethality of the homozygous genotype. This data suggests that the FoxF2bCa23 allele does not confer elevated risk of hemorrhage. The lack (or mild increase in) hemorrhage mutant data contrasts with morphant data where hemorrhage is observed at a statistically elevated frequency. The lack of concordance of mutant and morphant phenotypes has led to the suggestion that mutant phenotypes should be preferred over morphant phenotypes (Kok et al., 2015). However, observations of compensation in genetic mutants makes the interpretation of mutant and morphant phenotypes less clear (Rossi et al., 2015). Indeed, I have noted an increase in expression of the paralog foxf2a in my foxf2b mutants as well as loss of the related gene foxq1. I also show that loss of foxc1 (the third linked gene in this cluster) has a similar phenotype to loss of foxf2. Taken together, the difference between mutant and morphant phenotypes that I observe could be due to morpholino off target effects as is known to happen (Robu et al., 2007; Kok et al., 2015) or could be due to genetic compensation. To address this, I am making a foxf2a

123

genetic mutant on a foxf2b mutant background to achieve a full foxf2 null. Genetic interactions with foxc1 and foxq1 also remain to be addressed.

4.3.3 FoxF2b regulates markers of vSMC and pericyte differentiation.

acta2 is a key gene expressed early in the differentiation of the smooth muscle lineage. Our lab developed an acta2:GFP transgenic that is very useful for understanding in vivo mural cell differentiation. Transgene expression was analyzed on pharyngeal arch arteries in foxf2b morphants and on the mid-mesencephalic central artery and associated branches in mutants as these are the earliest sites of acta2 expression. I found a significant decrease in the number of Acta2-expressing cells on the pharyngeal arch arteries in foxf2b morphants compared to control embryos. At this point I do not know whether this decrease is due to a lack of cells, or if the cells are present in an undifferentiated state, delayed or unable to express Acta2. Given that some cells are present and expressing Acta2, I hypothesize that this occurs due to a lack of interactions between endothelial cells and mural cells normally required for the expression of differentiated markers, due to lack of FoxF2 function. In terms of acta2 expression I found concordance between genetic mutants and morphants that both had a similar reduction in acta2 expression. For this work I developed a new assay. Instead of looking at the ventral head vessels, I became interested in looking at cerebral vascular stability. We found that the mid-mesencephalic central artery (MMCtA) and its associated branches develop acta2- expressing coverage at the same timepoints that the ventral head vessels show expansion of acta2:GFP coverage. Expression of acta2 in foxf2bCa23/Ca23 mutants and wildtype siblings was examined on the cerebral vessels at 102hpf, just after initial onset of acta2 expression around vessels. I found there is a decrease in acta2 expression on vessels, quantified by three parameters: number of acta2:GFP+ cell- associated branching points, number of acta2:GFP+ cells, and length of longest acta2:GFP+ extension from the base of the MMCtA (the basal communicating artery, BCA). Surprisingly, by 174hpf (approximately 1 week), embryos have no significantly different coverage of vessels by acta2:GFP+ mural cells, showing that delayed acta2 expression eventually recovers and reaches wild type levels. However, I observed that the absolute numbers of acta2- expressing cells appears to decrease between 102 and 174hpf, possibly because of increasing coverage of the vessel by acta2 positive cells, making it more difficult to discern between individual cells than at 102hpf. Furthermore, no difference is apparent in the other two measurements, which seems contradictory to expansion of mural cell coverage observed by eye. One explanation is that mural cells 124

are already in final position around vessels by 4dpf but loss of FoxF2b delays the onset of acta2 expression. Mechanisms of compensation may be initiated during this phase so that by 7dpf the difference between wild type and mutant is no longer seen. I chose to examine pericyte markers pdgfrβ, notch3 and ng2, in wildtype and foxf2b mutant embryos. foxf2bCa23/Ca23 mutants had a decreased number of pdgfrβ-expressing cells at 4dpf, suggestive of a loss or delay of pericyte maturation. It is unclear whether FoxF2 is acting directly or indirectly on pdgfrβ, but the decrease in pdgfrβ expression suggests that FoxF2b is required for proper pericyte recruitment and maturation. Interestingly, this difference is not seen at 48hpf, as distinct pericytes are not identifiable in the head at this stage by any marker, including ng2 or notch3. In summary, some, but not all smooth muscle and pericyte markers are decreased in the head vasculature of 4dpf foxf2b genetic mutants suggesting a defect in specification, migration or differentiation of vascular mural cells in these mutants.

4.3.4 Vessel ultrastructure reveals a role for FoxF2b in promoting endothelial-mural cell interactions

Transmission electron microscopy is a powerful technique that allows us to see interactions between mural cells and endothelial cells at sub-cellular resolution without the need for transgenic markers or antibodies. I analyzed ultrastructure of vessels at 48hpf by TEM in morphant and control vessels, focussing on vessels that are easily recognizable in TEM, such as the primitive internal carotid artery (PICA) and the dorsal aorta. The dorsal aorta runs below the hindbrain and is the major vessel to the trunk while the PICA runs anteriorly. Both are large, early developing vessels. Similar to previous observations in our lab (Liu et al., 2007; Lamont et al., 2010), TEM revealed that peri-endothelial cells that are pericyte-like in morphology are present at the abluminal surface of vessels at 48hpf prior to acta2-expression onset. I quantitated the contacts between mural and endothelial cells and found less abluminal endothelial cell surface in close contact with adjacent peri-endothelial cells. This suggests that vessels are poorly supported in foxf2b morphants. Additionally, this appears to be specific to cells of the head, as dorsal aorta contacts are not affected. This finding is consistent with the notion that the decreased expression of the mural cell differentiation marker acta2 is due to reduced interactions between endothelial cells and mural cells in foxf2b morphants and mutants.

125

4.3.5 foxc1 knockdown produces vascular stability defects

Work by our collaborators, the Lehmann lab, linked FoxC1 mutations in humans and cerebral small vessel disease, using a combination of human imaging and modelling of foxc1 mutations in a developing zebrafish model. MRI brain scans of human patients with FoxC1 mutations (duplications, deletions and mutations) reveal white matter hyperintensities and lacunar infarcts, indicative of CSVD. Transient knockdown of zebrafish foxc1 results in cerebral hemorrhage, suggestive of vascular stability defects (French et al., 2014). With my expertise in TEM and the Tg(acta2:GFP)ca7 line, I analyzed foxc1 morphants, and I demonstrated an increased amount of perivascular space, and a decrease of acta2 expression on pharyngeal arch arteries, similar to knockdown of foxf2b. The number of acta2-expressing cells in foxc1 morphants is decreased by about 25%, which is less severe than the 50% decrease I observe in foxf2b morphants. This may suggest that foxf2b is the more important gene for mural cell differentiation of the two, however the differences could also be attributable to differences in morpholino knockdown efficiency, the different imaging positions (lateral vs ventral). Ventral images show acta2-expressing cell on the ventral aorta, which is not visible in lateral views. Either way, this data points to a role for foxc1 promoting mural cell-endothelial cell interactions, and suggests some overlapping function of foxf2 and foxc1.

4.3.6 Knockdown and genetic knockout of foxf2b affects expression of genes involved in mural cell stabilization of endothelial cells

Proper migration and specification of origin tissues (mural and endothelial), reciprocal recruitment and differentiation cues between mural cells and endothelium, and formation of an elastic basement membrane are all crucial for promoting vessel wall integrity. Examination of the expression of components of these aspects of vascular stability in foxf2b morphants and mutants reveals FoxF2 may regulate PDGF signaling, neural crest-derived structure formation, Shh feedback signaling and deposition of basement membrane components. I have performed ISH of several key genes in morphant and mutant embryos to identify potential roles for FoxF2 during three stages of development (see Table 4.1 for summary). There is a caveat that not all three stages were examined in all models, or for each gene. Morphant embryos were largely examined at 24-30hpf to identify gene expression differences that may be present prior to hemorrhage phenotype at 48hpf. Morpholinos dilute over developmental time with cell divisions, and are effective for 2-5 days. However, given the reduction in acta2:GFP in foxf2b morphants at 4dpf the knockdown phenotype appears to be present at 4dpf. Therefore I could examine expression patterns of important genes at later stages as well. Certain genes 126

were not expected to yield additional informative results at later stages, such as neural crest markers snai2 or hand2, since the migration of neural crest mesenchyme is largely complete by 24hpf. In addition to the figures presented, many ISH experiments were attempted at other stages and on other genes that, for technical reasons produced variable results and are not presented here. Interestingly, pdgfrβ appears to be the only PDGF signaling component with decreased expression in foxf2b mutants. The other receptor, pdgfrα, shows an increase in pharyngeal expression at 48hpf and 4dpf, and both pdgfaa and pdgfba ligands also have increased intensity of ISH signal at 48hpf, but no apparent difference at 4dpf. The increased expression of the ligands, however, corresponds to diffuse head signal that may be ubiquitous or non-specific. Furthermore, pdgfba is the endothelial-expressed ligand that is secreted to mural cells, and the pattern does not recapitulate that of vessels, suggesting this staining may not be accurately representing pdgfba expression. Previous work examining gene expression differences in mouse gut smooth muscle also saw upregulation of PDGF- A, PDGF-B and PDGFRα in FoxF2 mutants (Bolte et al., 2015). Here, however, I have shown decreased pdgfrβ expression for the first time in FoxF2 mutants, and potentially identified a primary mechanism through which FoxF2 regulates endothelial-mural cell interactions. The decreased expression of col4a2 is in line with decreased endothelial-mural cell interactions, as deposition of basement membrane proteins is induced by PDGF signaling (Mima et al., 2011) and ColIV can induce smooth muscle cell differentiation (Xiao et al., 2007). The neural crest EMT marker snai2 has unchanged expression in pharyngeal arch mesenchyme of foxf2b morphants at 30hpf, but hand2, which is expressed in more discrete domains, shows a loss of expression in the 4th pharyngeal arch, These arches arise from migrating neural crest cells, and a lack of this last arch implicates FoxF2b in proper neural crest cell migration. Decreased FoxF2 in cancer tissues is associated with increased EMT and metastasis, suggesting FoxF2 may normally inhibit migration of cells following EMT to promote genes once they have migrated to the proper region. Given the apparent lack of change in snai1b, it seems FoxF2b may promote differentiation of neural- crest derived mural cells through other means than general repression of EMT-promoting genes. nkx3.2 expression was unchanged in foxf2b mutant and morphant embryos, showing that Nkx3.2 is not regulated by FoxF2b. The Drosophila homologs of FoxF2 and Nkx3.2 (Biniou and Bapx) specify visceral muscle of the gut, and this is thought to be highly similar to vascular smooth muscle. The FoxF2b homolog in fly is actually a homolog for the whole FoxF family, and subfunctionalization of the family members through evolution has likely altered their specific roles and cofactors in vertebrates. 127

The increase of gli2a and ptch2 expression in foxf2bCa22/Ca22 and foxf2bCa23/Ca23 mutants suggests that FoxF2 is normally repressing Shh signaling. Gli2 is able to promote vSMC proliferation (Li et al., 2010), and is an activator of Shh signaling, two processes are inhibited in order to promote differentiation of vSMCs. The specific upregulation of gli2a and ptch2 in the region of neural crest- derived vSMCs indicates that inhibition of proliferative Shh signals may be another mechanism by which foxf2b promotes the maturation of vascular mural cells of the pharyngeal arch arteries. Shh signalling often has negative feedback loops that attenuate signalling (i.e. binding of Shh to the Patched releases inhibition on Smoothened, resulting in the upregulation of Patched and more inhibition of Smoothened). Since FoxF2 is positively regulated by Shh, Shh signalling would increase FoxF2 which would then potentially repress Ptch2 and Gli2 by FoxF2 and modulate further signalling of the pathway. Quantitative PCR of the levels of gene changes or RNAseq might help us understand how this feedback occurs, although as I note regional changes in expression, techniques that rely on measurements in bulk embryos may not detect subtle regional changes in expression. ISH expression patterns of the tbx15, tbx18 and hic1 genes mark populations of the head mesenchyme, and can act as markers of whether these mesenchymal tissues are properly specified and localized. Interestingly, these markers show a large variability of staining pattern and intensity between stages and mutant alleles. This is rather puzzling, and may reflect variability in the ISH staining, but suggests that FoxF2b is required for establishing head mesenchyme populations in the head. Further work to examine expression at both stages and in both foxf2b alleles will provide a more consistent picture than is available currently.

4.3.7 Conclusions and Future Directions

First identified as a differentially regulated gene in a model of genetic hemorrhage, these experiments are the first to exhibit a role for zebrafish FoxF2 in the development of vascular stability. Reduction in expression of a pericyte (pdgfrβ) and a contractile mural cell marker (acta2) show a role for FoxF2b in promoting development of mural cells, potentially through involvement with PDGF and Shh signaling. I have also discovered a role in the formation of neural crest cell derivatives (hand2, tbx15, tbx18). Moving forward, work should focus on the role of FoxF2 in PDGF and Shh signaling pathways to identify whether and how FoxF2 is specifically modulating these signals. This may shed additional light on how FoxF2 acts as a transcription factor, whether it is inhibitory or activating, and whether it coordinates with co-factors to confer spatio-temporal function. In addition the Notch and TGF-β signaling pathways have not been examined here, and determining whether FoxF2 also has an 128

effect on these key vascular stability signals would further elucidate its specific role in vascular development. Mutagenesis using CRISPR technology is currently underway to create a foxf2a/b double mutant zebrafish line to obtain a full FoxF2 null and provide insight into the ability of these paralogous copies to compensate for one another. Creation of a foxf2a mutant will also tell us whether the two paralogs have obtained additional divergent roles in development. Importantly, the allele- specific differences which are detected in hemorrhage rates, smooth muscle coverage and marker expression may be the results of individual lines of foxf2b mutant fish ‘compensating’ in their own way for loss of FoxF2b. Elimination of the second allele of foxf2 may decrease the compensation allowing the phenotypes to become more consistent. However there is also the risk that other Fox cluster members might then be upregulated to compensate (FoxC1 for instance), and that knocking out the entire cluster may be necessary to see a strong effect on mural cell development.

129

Chapter Five: FoxF2b Maintains Cerebral Vascular Integrity in Developed Zebrafish Brains through Pericytes

Wei Dong and the Microscopy Imaging Facility helped with acquisition of TEM images. Sarah Childs helped with imaging of dissected brains.

130

Abstract Zebrafish foxf2b mutants are homozygous viable. Human patients with homozygous mutations/deletions in the FoxC1/F2/Q1 cluster have not been observed. However zebrafish foxf2b mutants retain FoxF2a, and therefore only have a 50% reduced dosage of FoxF2. Segmental deletions of one allele or duplications of FoxC1 are observed in patients with Axenfeld-Rieger Syndrome and glaucoma. To understand if FoxF2 loss might also be associated with human disease, I interrogated the DECIPHER database, a database of human genetic variants coupled with phenotypes. I found that deletions and/or duplications encompassing FoxF2 are commonly associated with intellectual disability and craniofacial abnormalities. To determine if zebrafish FoxF2 continues to play a role in vascular stability in the mature organism, I chose to examine zebrafish at juvenile to young adult stages. Expression of foxf2 in zebrafish is in brain pericytes, and foxf2b mutants exhibit increased cerebral hemorrhage over wildtype siblings from 1 to 5 months of age. No significant differences have been found in the vascular architecture of the brain, but ultrastructure of vessels may show a general decrease in size. My data implicate a role for FoxF2 in maintenance of vascular stability into adulthood, even though the changes in vessels appear subtle. Further analysis of FoxF2 in pericytes may yield greater insight into the role these cells are playing in maintaining cerebral vascular integrity, thus safeguarding the health and function of the brain.

131

5.1 Introduction

Cerebral small vessel disease (CSVD) pathologies often manifest in late stages of life. CVSD encompasses a heterogenous group of diseases including genetic diseases such as CADASIL, CARASIL, amyloid angiopathies, and many less well characterized disorders. Symptoms of the small vessel disease CADASIL generally present between 30-60 years of age (Choi et al., 2015) (although some reports show juvenile onset (Hartley et al., 2010; Benabu et al., 2013)). CSVD is often progressive, suggesting accumulation of vascular insults. NOTCH3 is the gene most commonly associated with CADASIL, but some cases are not linked with NOTCH3 mutations (Santa et al., 2003), and not all genetic contributions of CSVD have been determined. Given the association of FoxC1 mutations in humans with a small number of patients with Axenfeld-Rieger syndrome and CVSD, we were curious to understand whether there are adult phenotypes in FoxF2 mutants. In mice with conditionally deleted FoxF2 in the adult, mice exhibit BBB defects and leaky vasculature, specifically in the brain (Reyahi et al., 2015), although the defects are more subtle than embryonic knockouts. Furthermore, human segmental deletions that include FoxF2 exhibit hallmarks of CSVD, and SNPs linked to FoxF2 are associated with ischemic stroke in aged Europeans (Chauhan et al., 2016). This led me to investigate whether FoxF2 mediates continued endothelial-mural cell interactions in mature brain vessels of the zebrafish. The FoxF2b mutant alleles are homozygous viable. I predict that with aging, vessels will show some amount of hemorrhage or general leakiness, but that molecular and cellular phenotypes may be subtle. Here I use brightfield, confocal and TEM imaging techniques on dissected 1-5 month old homozygous mutant and wild type brains to examine the state of cerebral vessels in FoxF2b mutants.

132

5.2 Results

5.2.1 foxf2 is expressed in pericytes of the zebrafish brain

Since FoxF2b affects the mural cell coverage of cerebral vessels during development, and it has a maintenance role in gut smooth muscle (Ormestad et al., 2006), the next step involved examination of juvenile (1-5 month) brains to investigate any potential role for FoxF2 in cerebral vascular mural cells. I first wanted to examine whether foxf2 is expressed in mural cells of the juvenile/adult. ISH of foxf2a/b in 1 month old zebrafish brains revealed a punctate pattern throughout sectioned tissue (Fig. 5.1 A,B). When co-stained with markers for endothelium, it was found that foxf2a/b staining was directly adjacent to endothelial tubes. FoxF2 expression was compared to pericyte markers pdgfrβ and notch3 in sectioned 1 month brain tissues (Fig. 5.1 C,D), and shows that pericyte markers are expressed in the same pattern as foxf2a/b, suggesting foxf2 is expressed by pericytes.

5.2.2 foxf2b mutant brains exhibit hemorrhages more commonly than wildtype

foxf2b mutants were assayed for superficial cerebral hemorrhage between 1-5 months of age through imaging of dissected brains. Hemorrhages just below the surface of the brain could be visualized and scored (Fig. 5.2 A,B). FoxF2bCa22 brains were taken from progeny of foxf2b+/Ca22 incrosses and genotyped at the time of brain dissection. 5% of wildtype sibling brains hemorrhage (n=22) compared to 23% hemorrhage in heterozygotes (n=40) and 31% in homozygotes (n=36; Fig. 5.2 C). Hemorrhage rates are significantly different between wildtypes and homozygous mutants (p=0.009 by N-1 two proportions test), but not significant between wildtype and heterozygous mutants (p=0.068 by N-1 two proportions test). This striking difference in hemorrhage occurrences suggests that FoxF2b is important for the maintenance of cerebral vascular stability in juvenile and adult stages and is similar to the phenotype of the embryonic FoxF2bCa22 mutants. Interestingly, in the progeny of a foxf2bCa23/Ca23 homozygous incrosses, mutant brains show no significant hemorrhage over age matched wildtypes (Fig. 5.2 D). I note that in these experiments both wildtypes and foxf2bCa23/Ca23 mutants hemorrhage at rates of just under 40% (39.2% in Ca23, 37.5% in WT), a surprisingly high rate for wildtypes.

133

Figure 5.1 – foxf2a and foxf2b are expressed in brain tissue in patterns reminiscent of pericyte marker expression

A,B: foxf2a (A) and foxf2b (B) expression is punctate (arrowheads) and found adjacent to and surrounding endothelium (brown). C,D: Established pericytes marker genes pdgfrβ (C) and notch3 (D) indicate pericytes (arrowheads), adjacent to endothelium (brown) in a similar pattern to foxf2, indicating that foxf2-expressing cells are likely pericytes. Scale bars = 10µm.

134

Figure 5.2 – foxf2bCa22/Ca22 mutant brains exhibit hemorrhage at juvenile stages

A,B: Dissected wildtype and foxf2bCa22/Ca22 brains. Hemorrhage can be seen in mutant brain tissues (arrowheads). C: Bar graph depicting percent hemorrhage in wildtype, foxf2b+/Ca22 and foxf2bCa22/Ca22 mutant brain tissue between 1-5 months of age (p=0.0187). No significant difference is seen between heterozygotes and wildtype (p=0.0681). D: Bar graph depicting hemorrhage rates in foxf2bCa23/Ca23 mutants compared to wildtype. No statistical significant difference of hemorrhage rates is seen between foxf2bCa23/Ca23 mutant and wildtype embryos. Scale bars = 100 µm.

135

5.2.3 foxf2b mutant small cerebral vessels trend towards smaller diameter

To determine whether the ultrastructure of juvenile brain vessels is changed in foxf2 mutants, and to understand the source of hemorrhage, small vessels of the brain (arterioles, capillaries and venules) were imaged using TEM to examine vessel ultrastructure and interactions between endothelial cells and pericytes where FoxF2b is expressed (Fig. 5.3). Vessel wall area (including any associated mural cells), luminal area, and the ratio of vessel wall area:lumen was measured using ImageJ. Brains were taken from progeny of foxf2b+/Ca22 crosses. Although no significant differences were observed between wildtypes (n=27 vessels across 9 brains) and foxf2bCa22/Ca22 mutants (n=20 vessels across 9 brains), there is a trend towards overall decreased vessel size in mutants (wildtype vessel wall area: 27.93 ±3.43 µm2, foxf2bCa22/Ca22 mutant vessel wall area: 22.85 ±4.24 µm2, p=0.352 by two-tailed t test; wildtype lumen area: 50.37 ±9.10 µm2, foxf2bCa22/Ca22 mutant lumen area: 40.60 ±8.06 µm2, p=0.223 by two-tailed t test, wildtype endothelial cell area: 17.67 ±2.615 µm2, foxf2bCa22/Ca22 mutant endothelial cell area: 15.39 ±2.331 µm2, p=0.535 by two-tailed t test) as well as a decreased ratio of vessel wall:lumen area (wildtype ratio: 0.88 ±0.16, foxf2bCa22/Ca22 mutant ratio: 0.63 ±0.06, p=0.148 by two-tailed t test), similar to what is seen in mouse FoxF2 mutants (Reyahi et al., 2015). In this analysis I did not separate vessels by different size (i.e capillary where pericytes would be expected to associate with vessels vs. larger vessels where smooth muscle would be present).

5.2.4 acta2 coverage of foxf2b mutant brain vessels is unchanged

In the previous chapter I found that vessel coverage by mural cells is decreased in foxf2bCa23/Ca23 mutant embryos at 4dpf. Although it is recovered by 7dpf, maintenance or further development of mural cells into juvenile and adult stages may also be disrupted in FoxFbCa23 mutants. Our acta2:GFP transgenic offers a unique method to observe smooth muscle in embryos, and I expanded this use to adult brains. By pairing confocal microscopy with ClearT2 and CLARITY tissue clearing methods (Kuwajima et al., 2013; Cronan et al., 2015), I developed a technique to image acta2:GFP fluorescent signal in dissected whole brains of 4 month-old zebrafish. Although I expected to see more elaborate smooth muscle coverage of brain vessels in adult fish, I found that the pattern of acta2:GFP on cerebral vessels was qualitatively similar to that of embryonic zebrafish and consisted of one or two major vessels entering the brain ventrally and branching a limited number of times before the acta2:GFP signal ended (Fig 5.4). I used a measurement system similar to that I had previously used for embryos, examining total extent of acta2:GFP+ coverage of vessels (additive length of all

136

Figure 5.3 – Small vessels of the brain show a non-significant trend towards decreased size, and constricted lumens relative to vessel wall size in foxf2bCa22/Ca22 mutants

A,B: TEM cross-section images of cerebral vessels, showing endothelium (green), pericytes (yellow) and erythrocytes (red) in wildtype (A) and FoxF2bCa22/Ca22 mutant (B) tissues. C: Bar graph depicting average cerebral vessel luminal area in wildtype and mutant fish. D: Bar graph depicting cell wall size, including any associated mural cells, of wildtype and mutant cerebral vessels. E: Bar graph depicting ratio of vessel luminal area:cell wall area. F: Bar graph depicting endothelial cell area, excluding pericytes. None of the graphs show statistically significant differences. Scale bars = 2µm.

137

acta2:GFP+ vessels), and normalized to the size of the brain to determine if there was any change to acta2-expressing cell coverage. No significant difference of total vessel coverage was observed between wildtype and FoxF2bCa23/Ca23 mutant siblings (wildtype average ratio: 1.65 ±0.18, n=9, foxf2bCa23/Ca23 average ratio: 2.01 ±0.31 n=9, p=0.335 by two-tailed t-test, Fig 5.4).

5.2.5 Human FoxF2 mutations are commonly associated with brain defects

I utilized the DECIPHER database of human mutations to identify phenotypes commonly associated with FoxF2 mutations (Bragin et al., 2014). Each entry represents a single patient with mutation encompassing FoxF2 entered by clinicians, describing a spectrum of symptoms that have been SNP genotyped for copy number variants indicating duplications and deletions. These mutations overlap the FoxF2 locus and likely also encompass neighbouring genes. However, since all patients have deletion/duplication of FoxF2 in common, I hypothesized that I might find common sets of phenotypes corresponding to FoxF2 mutations. I found that 27.3% of the FoxF2 mutation entries specifically listed intellectual disability as a phenotype, and 2 additional entries exhibited white matter abnormalities, suggestive of CSVD (Fig. 5.5 A). Of the ~170 phenotypes listed, 23% were associated with some form of brain defect, although several of these overlapped in the same patient (total entries = 44, although 12 have no reported phenotype). The majority of FoxF2 mutations were loss or deletion mutations (29 of 44, Fig. 5.5 B), suggesting that FoxF2 function is more commonly lost than gained in humans. Of the 44 entries, half were male, 17 were female and 5 were not disclosed (Fig. 5.5 C). The higher number of male patients may be due to the fact that several entries listed male genital defects, and suggests that FoxF2 also regulates male genital formation. A list of the affected regions and processes annotated in the DECIPHER FoxF2 page can be found in Table 5.1.

138

Figure 5.4 – acta2:GFP+ coverage of cerebral vessels is unchanged between wildtype and foxf2bCa23/Ca23 mutant brains

A,B: Representative fluorescent confocal microscope image of dissected 4 month wildtype (A) and foxf2bCa23/Ca23 mutant (B) zebrafish brains, viewed from midline axis. acta2:GFP+ cells are labelled with green fluorescence, anterior to the left. Magnified section of A depicts quantification method (A’). Scale bars = 200µm. C: Graph depicting total vessel length covered by acta2:GFP+ cells, normalized to total area of brain (ratio of additive length of all acta2:GFP+ extensions along vessels (µm) to total brain tissue area (µm2)). D: Comparative image of 7dpf embryonic head vasculature and acta2:GFP expression. Scale bar = 40µm.

139

Figure 5.5 – Human FoxF2 mutations are commonly associated with intellectual disability

A: Pie chart depicting DECIPHER database annotation of human FoxF2 mutations and the prevalence of intellectual disability and CSVD phenotypes in all entries. B: Pie chart depicting the number of FoxF2 mutation types in all entries. C: Pie chart depicting the numbers of male and female patients with FoxF2 in all database entries.

This study makes use of data generated by the DECIPHER community. A full list of centres who contributed to the generation of the data is available from http://decipher.sanger.ac.uk and via email from [email protected]. Funding for the project was provided by the Wellcome Trust.

140 Affected Organ/Process Subcategories # of Reported Phenotypes Examples Brain 39 Intellectual Disability, Delayed speech Cognitive Function 24 development, Learning disability Micro/Macrocephaly, Hydocephalus, Structure 11 Hypoplasia of the corpus callosum White matter lesions, Demyelination of White Matter 4 subcortical white matter Face and Neck 24 General Face and Neck 9 Neck abnormalities, Facial abnormalities Nose 5 Prominent or short nose, Anteverted nares Low-set ears, Preauricular skin tag, Pinna Ears (External) 4 abnormalities Eyes (External) 6 Eyelid abnormalities, Palpebral abnormalities Nasal bridge abnormalities, Dental Cranial Skeleton 18 abnormalities, Skull shape abnormalities Axenfeld/Rieger/Peters anomalies, Eyes 30 Hypertelorism, Strabismus, Hypermetropia, Iris and pupil abnormalities Hearing impairment, Abnormalities of the Ears 13 cochlea and ossicles Mouth and Foregut 7 Mouth 5 Uvula abnormalities, Polydipsia, Dysarthria Foregut 2 Esophageal atresia, Dysphagia Cardiovascular 11 Heart 7 Septal defects, Arrhythmia Dorsal Aorta 3 Patent ductus arteriosis, Coarction of aorta Other 1 Hemangioma Distal Limb 9 Brachydactyly syndrome, Clinodactyly, Nail Digit 6 abnormalities Metacarpal aplasia/hypoplasia, Abnormalities Hand/Foot 3 of the foot/palm

Growth 12 Short/tall stature, Global developmental delay

Glands/Organs 10 Genitals 6 Cryptorchidism, Hypospadias, Micropenis Hydronephrosis, Renal agenesis, Horseshoe Kidney 3 kidney Thyroid 3 Hypothyroidism, Hypoparathyroidism Skin 1 Eczema Other 7 Motor delay, Abnormality of motor neurons Motor/Muscle Function 4 and musculature Bone 2 Osteopetrosis, Arthritis Diabetes 1 Maternal Diabetes

Table 5.1 – Human FoxF2 mutation-related phenotypes as annotated in the DECIPHER database (v9.9)

DECIPHER entries of FoxF2 mutation-related phenotypes show a large number of brain, eye and craniofacial defects and abnormalities.

141

5.3 Discussion

5.3.1 foxf2 expression in pericytes suggests a role in vascular stability

Expression of foxf2 in brain pericytes has previously been observed in mouse (Reyahi et al., 2015), and here I show evidence for similar expression in juvenile and adult zebrafish, suggesting a conserved role for FoxF2 in cerebral vascular stability. It is interesting that foxf2b is not expressed in pericytes in embryonic zebrafish, but is in the mesenchyme around ventral head vessels. Whether expression in mesenchyme or pericytes represents a switch in roles at different life stages or whether foxf2 is expressed in mural cells on vessels of a particular caliber/flow rate remains to be determined. Pericyte expression shows that foxf2 is expressed in a cell type directly involved in vascular stability and in position to regulate the blood-brain barrier. The pericyte markers, pdgfrβ and notch3, have known associations with stroke and small vessel disease, showing that proper regulation of brain pericyte function is important to avoid vascular-based neuronal cell death. Thus, the finding of foxf2 expression in brain pericytes suggests it is a good candidate for a novel vascular stability factor. Coupled with my data in the previous chapter to suggest that perictye pdgfrβ expression is regulated by FoxF2, this suggests a possible mechanism by which FoxF2 regulates cerebral vascular stability in the adult, although regulation of pdgfrβ in adult pericytes by FoxF2 has not yet been confirmed.

5.3.2 foxf2b mutant brains hemorrhage in juvenile stages

At juvenile stages (1-5 months), foxf2bCa22/Ca22 brains exhibit significantly more hemorrhage than wildtype siblings, suggesting a role for FoxF2 in maintenance as well as development of vascular stability. Interestingly, foxf2b+/Ca22 sibling brains show a non-significant increase in hemorrhage rates over wildtype siblings, but not as strongly as foxf2bCa22/Ca22, further suggesting that there is dosage sensitivity for FoxF2. Homozygotes do not show complete penetrance of hemorrhage, but hemorrhage is a stochastic event that occurs after vascular stability has broken down in a localized area and thus I would not expect to see full penetrance at a single timepoint. One caveat is that brains were examined for superficial hemorrhages. It is entirely possible that hemorrhage events could occur before or after the time of dissection, and even in dissected brains, small hemorrhages located more internally may not be as easily visible. I did undertake preliminary histological staining of brains looking for areas of potential cell death (hematoxylin), hypoxia (hypoxyprobe) and ischemia (cresyl violet). As we have not looked at these parameters before in adult zebrafish and I had low numbers of adult brains to

142

compare, I could not make any strong conclusions and this work remains to be completed in the future when more brains are available. Vascular stability is the results of multiple factors: strong endothelial junctions, paracrine signaling pathways and extracellular matrix and basement membrane proteins. FoxF2 is likely not a master regulator of the vascular stability, but rather one of many transcriptional regulators required for fine-tuning of some of the specific processes and interactions in vascular stability. Interestingly, I once again found evidence for allele-specific differences in hemorrhage rates. foxf2bCa23/Ca23 mutant brains, when compared to wildtype siblings, do not have significantly higher hemorrhage rates, consistent with the normal distribution of alleles in hemorrhaged and non- hemorrhaged samples at embryonic stages. Wildtype siblings hemorrhage rates are atypically high in this experiment, and it is possible that the genetic background of this particular line of fish is prone to hemorrhage. Hemorrhage rates in FoxF2bCa23 mutants, therefore, cannot be reliably attributed to the mutation. I tried as much as possible to control for dissection quality to avoid injury to brains, but some “hemorrhage” could be due to injury during dissection. One potential caveat with this particular experiment is that controls were unrelated wildtype (transgenic) embryos and not wildtype siblings of mutants, and may have had a higher background hemorrhage rate. Repeating the experiment with related controls may yield a more significant difference.

5.3.3 Vessel morphology and mural cell phenotype are not significantly altered in foxf2b mutant brains

Vessel ultrastructure in cerebral vessels was measured by vessel wall and lumen area as previously described (Reyahi et al., 2015). Analysis of these measurements shows no significant difference between wildtype and foxf2bCa22/Ca22 mutant vessels, but did suggest a general trend towards decreased vessel size. More brain samples would likely allow this result to reach significance. Interestingly, a decrease in the ratio of vessel wall area:lumen area may show a relative thinning of the cell wall and dilation of the lumen. I found the endothelial cell layer is thinner, further suggesting an overall vessel decrease in size. We have previously observed smaller vessel diameters in iguana mutants that also have poor mural-endothelial cell interactions (Lamont et al., 2010). Pericytes are known to acquire a sustained contractile state in response to ischemia (Hall et al., 2014). Thus, functional mural cells are normally required for preventing vessel lumen occlusion in addition to regulating blood flow (i.e. need the ability to contract and relax). The relative dilation of the lumen

143

may reflect an increase in blood flow, due to similar volumes of blood attempting to pass through overall smaller vessels. It is very possible that the lumen diameter changes in foxf2 mutants would result in altered microvascular circulation in mutant brains and contribute to further cerebrovascular anomalies. Over a lifetime, this may result in cerebral small vessel disease. Although this hypothesis is not easily testable in the adult zebrafish where the skull occludes imaging of the brain, it would be very interesting to look at the functional circulation of foxf2 mutants. Pericytes and smooth muscle cells are not completely distinct cell types, but rather exist along a continuum of mural cell phenotypes (Attwell et al., 2016), where some mural cells exist in a “transition zone” between multilamellar vSMCs and monocytic pericytes, displaying characteristics of both. The expression domain of acta2:GFP in juvenile brains shows a slight but non-significant increase in foxf2bCa23/Ca23 mutants compared to wildtypes. While this is not significant, the notion of increased acta2 expression would suggest that vessels in the transition zone (i.e. medium diameter) have increased contractile ability as compared to normal vessels of the same size due to increased Acta2 expression. It is difficult to separate cause and effect in this case, as this may either be a direct consequence of loss of FoxF2, or as a secondary effect requiring increased investment of acta2- expressing cells to compensate for loss of stabilization or dysregulated blood flow. Compensation may arise from smooth muscle cells on the feeding vessels altering their distribution or from increasing contractility in pericytes on the capillary side of the transitional zone. Thus, smaller microvessels and constricted lumens might lead to alterations in smooth muscle on feeding vessels. Whether there is an effect in pericytes or smooth muscle would require careful study of the architecture, vessel size, mural cell phenotype and relationship to foxf2 to understand which vascular defects are primary and which are secondary.

5.3.4 Human mutations show relevance for the investigation of zebrafish FoxF2 function in brain

A search of the DECIPHER database (Firth et al., 2009; Bragin et al., 2014) suggests that large genomic deletions or duplications overlapping the human FoxF2 locus cause a range of phenotypes, many of which are associated with defects in intellectual disability (of the 44 patients in the database, 32 have phenotype descriptions and 11 of these 32 patients have a description of intellectual disability). In addition to intellectual disability, other cognitive phenotypes of the brain have been identified, including learning disabilities, autism and delay of speech and language development. White matter abnormalities were also detected, although not as commonly, likely due to not all

144

patients being assessed through brain imaging. How much change in dosage of FoxF2 contributes to intellectual disability is speculative, as other genes are often mutated in the same patient. However, the relatively high prevalence of intellectual disability suggests that FoxF2 is plausibly contributing to brain development and maintenance. For instance, cerebral insults originating in the small vessels could contribute to intellectual disability or cognitive decline as seen in CSVD. Brain imaging of patients exhibiting intellectual disability and loss or gain of FoxF2 would prove useful for investigating the presence of CSVD hallmarks and provide us a clearer answer for whether this phenotype is due to cerebral vascular defects. Our collaborator notes increased white matter hyperintensities in young children with FoxC1/and or FoxF2 mutations (French et al., 2014; Chauhan et al., 2016). In addition, behavioural studies of older foxf2b mutant zebrafish would be interesting to conduct to see if similar cognitive defects occur in fish as well. The fact that cerebral defects are seen in both human and zebrafish FoxF2 mutants affirms the usefulness for this zebrafish mutant model beyond development.

5.3.5 Conclusions and Future Directions

These experiments show there is disrupted vascular integrity at older stages in foxf2b mutant zebrafish, although it is subtle, as would be expected for homozygous viable mutant animals. The particular source of instability appears to be alterations at the pericyte level and brain microvasculature, but the mechanism still needs to be elucidated, as the differences seen in vessel structure and mural cell markers showed non-significant changes in juvenile brains. The next step will be to understand the role of FoxF2b in establishment of cerebral vascular stability vs. maintenance in the juvenile/adult. Further analysis of other vascular stability constituents, such as basement membrane, endothelial junctions and signaling pathways, may reveal further defects. It is likely that the small additive changes in several parameters will lead to weakened vessels prone to rupture.

145

Chapter Six: Discussion

146

I have identified a role for FoxF2b in establishing vascular stability in developing zebrafish embryos downstream of Shh signaling. I also provide evidence for FoxF2b in the maintenance of cerebral vascular stability into juvenile and adult stages. I identified expression of both zebrafish foxf2 paralogs in head mesenchymal tissues, surrounding endothelial cells, and later in brain vessel pericytes, but exclusive from Acta2-expressing mural cells. I found that transient knockdown and genetic mutation of foxf2b in zebrafish results in hemorrhage in the head and brain, an indication of vascular stability. foxf2b knockdown and knockout studies also revealed a decrease in the expression of the endothelial-associated acta2-expressing mural cells just after the initial onset of Acta2 expression. Knockdown studies also revealed reduction in contact between head endothelial vessels and peri-endothelial mural cells. The expression patterns of genes potentially involved in different aspects of vascular stability were examined in FoxF2-deficient embryos, revealing a potential role for FoxF2 in ECM synthesis, PDGF signaling, neural crest migration, as well as potentially mediating Shh signaling and formation of mesenchymal tissues. Knockdown of related gene foxc1 results in decreases in Acta2-coverage and mural cell interaction phenotypes similar to those seen in foxf2b knockdown, and decreased foxf2b causes a loss of foxq1, the other member of the Fox gene cluster. Taken together, these results implicate FoxF2 (and potentially the entire FoxQ-FoxF-FoxC cluster) in the establishment and maintenance of cerebral vascular stability by mediating mural cell-endothelial cell interactions to promote differentiation during development and maintenance of the vasculature.

6.1 FoxF2 and vascular stability

6.1.1 FoxF2 in early mesenchyme

Mesenchymal tissues of the head are derived from both neural crest and mesoderm. These tissues interact with one another to coordinate development of bone, cartilage and muscle of the head and face (Trainor et al., 1994; Rinon et al., 2007). Mesenchyme is also likely the initial source of cerebral mural cells, as evidence exists for both neural crest and mesodermal origins of vSMCs and pericytes. foxf2 is expressed in head mesenchymal tissues, and early expression patterns suggest it is in mesenchyme deriving from both neural crest and mesodermal populations. How Sonic Hedgehog signalling interacts with neural crest cells to induce foxf2 expression in neural crest mesenchyme remains to be determined, but what is interesting is that the formation of pharyngeal arch placodes is disrupted in foxf2b mutants, as evidenced by the marker hand2, implicating FoxF2 in the formation of neural-crest

147

derived tissues. I also observed aberrant expression of mesenchymal markers tbx15, tbx18 and hic1 suggesting foxf2 control in the early stages of patterning the head mesenchyme. I focussed most of my work on the cells that express foxf2 around endothelium of the head, both in the brain (later stages) and ventral head (early stages). This places foxf2 expression in the time and region of the majority of hemorrhages observed at 48hpf in morphant embryos. Hemorrhages were also recorded in ventral pharyngeal regions, where foxf2 expression is strongest, suggesting that FoxF2 is promoting vascular stability in both cerebral and non-cerebral regions of the head. Interestingly, foxf2b expression is not seen around all cerebral vessels at 48hpf, which may suggest that mural cell precursors are either not yet present around these vessels or are not yet expressing foxf2b. In the developing ventral head, I showed that acta2 and foxf2b expression patterns are mutually exclusive, yet directly adjacent, suggesting FoxF2 is expressed in smooth muscle precursors. One hypothesis is that foxf2 expression in precursor tissues is promoting differentiation. As we do not see expression of foxf2b in mature smooth muscle, it is likely turned off in differentiated contractile vSMCs. This is in contrast to pericytes, which initially do not express foxf2, but later express it strongly when associated with mature brain capillaries. These different expression patterns at different developmental stages in related but different cell types suggest different early and late roles in vascular mural cells.

6.1.2 FoxF2b potentially promotes differentiation of mural cells through inhibition of proliferative signals

The regulation of PDGF and Shh signaling pathways are two mechanisms by which FoxF2 may be promoting vascular stability through differentiation of mural cells. PDGF signaling promotes proliferation of pericytes, primarily through PDGF-B signaling to PDGFRβ, but there are also minor roles for PDGF-A and PDGFRα promoting vascular smooth muscle cell proliferation. ISH of both ligands suggest increased expression in zebrafish FoxF2b mutants, as well as increased pdgfrα in the pharyngeal mesenchyme. This may translate to increased proliferation of pericytes and vSMCs at the expense of differentiation, and suggests that FoxF2b normally functions to repress these signaling pathways. Interestingly, there is a marked decrease of pdgfrβ-expressing cells at 4dpf in the heads of FoxF2b mutants. This may seem to contradict the published evidence that expression of PDGFRβ promotes pericyte proliferation. Since pdgfrβ continues to be expressed after pericyte-endothelial cell interactions are established, this decrease may instead indicate a loss of functional pericytes. A similar situation is observed in a conditional FoxF2 mouse knockout model, where Pdgfrβ is decreased but a striking increase in the investment of neural crest-derived perivascular cells on endothelial vessels is

148

observed (Reyahi et al., 2015). This study makes the argument that residual Pdgfrβ expression is sufficient for recruitment of pericytes but not for proliferation, which is instead induced by other factors. Assaying the numbers of pericytes using other markers would determine if this is similar in the zebrafish FoxF2b mutants. However the other classical pericyte markers Ng2 and Notch3 do not appear to specifically demarcate pericytes at 4dpf, and we do not have the necessary zebrafish lines to examine neural-crest derived cells through lineage tracing. Increased Shh signaling has been observed in vSMCs undergoing a phenotypic switch to a synthetic state, mediated by the Gli2 effector (Li et al., 2010). Interestingly, this appears to be in response to PDGF signaling (Zeng et al., 2016). Gli2 expression can be seen in the pharyngeal arch region of 48hpf and 4dpf zebrafish embryos, and shows increased expression in FoxF2b mutants. Both this and the increased expression of PDGF signaling components suggest that mural cells are in a sustained proliferative or synthetic state at the expense of differentiation, and the FoxF2b normally functions to repress these signaling pathways to allow for mural cell differentiation. This is potentially reflected in the decrease of acta2-expressing cells seen around pharyngeal vessels in FoxF2b morphants, and as well as on cerebral vessels in mutants.

6.1.3 FoxF2b is required for cerebral vascular stability throughout life

The maintenance of vascular stability via the blood-brain barrier is crucial for longevity. Defects in the blood-brain barrier of small vessels can cause microaneurysms and microbleeds, the accumulation of which is associated with progressive dementias and increased risk of stroke. Adult PDGFRβ+/- mice are deficient in cerebral pericytes and show a progressive neurodegeneration and breakdown of the BBB (Bell et al., 2010). The hemorrhage phenotype observed in FoxF2b mutant zebrafish brains suggests a potential role for FoxF2 maintaining interactions between mural cells and endothelial cells after development, and given foxf2 expression in pericytes, these are likely the source of vascular instability. Two subtle differences are seen in FoxF2b mutant brains: decreased vessel size and increased acta2+-cell vessel coverage. Taken together, these observations potentially suggest an increase in mural cell contraction. However, a mechanism for this is purely speculative, as these changes were not statistically significant. Hemorrhagic events can occur in over-constricted vessels, where increased resistance to blood flow results in higher and unevenly distributed fluid pressure, and cause ruptures at weaker points of the vasculature. Given that the increase in Acta2 reflects an expanded expression domain along vessels, it is possible that pericytes at the transition zones between arterioles and capillaries gain ectopic contractile function and confer abnormal vasoconstriction. 149

Alternatively, though more unlikely, the increase in Acta2 may indicate vSMCs switching off the synthetic state, thus no longer depositing basement membrane proteins required for elastic resistance to blood flow. This can lead to stiffer vessel walls that are prone to rupture, as with arteriosclerosis. Indeed, stiffening of major arteries outside the brain is often linked to CSVD (van Sloten et al., 2015; Saji et al., 2016), although cerebral vessel-specific occurrences of this have not been previously investigated. Alternatively, the hemorrhagic events could be preceding vessel constriction. Degradation of oxygenated hemoglobin and release of iron produces a number of reactive superoxides that are thought to induce vasoconstriction (Macdonald & Weir, 1991; Rey et al., 2002). Hemorrhagic exposure of erythrocytes to the perivascular tissues could feasibly induce this process. Further analysis of cerebral vessel structure and signaling needs to be performed before we entirely understand the role of FoxF2 in maintaining vascular stability, although it will be difficult to experimentally tease out as the foxf2b mutation only has slight effects on cerebral vessels over a long period of aging.

6.2 Allele-specific and paralog-specific phenotype differences

6.2.1 FoxF2b mutant alleles exhibit slightly differing phenotypes

The FoxF2b Ca22 and Ca23 mutant alleles were generated from the same targeting event, and the resulting amino acid changes occur at a similar position, both resulting in several amino acids of nonsense sequence before terminating. However I found the two alleles display differing phenotypes. Both embryonically and in the juvenile brain, the FoxF2bCa22 mutant allele exhibits hemorrhage, but the FoxF2bCa23 allele does not. Yet the foxf2bCa23/Ca23 mutants exhibit a phenotype in decreased acta2:GFP and pdgfrβ expression at 4dpf, suggesting this allele is having functional effects on the developing embryo in processes where the FoxF2bCa22 allele appears to have no effect. Additionally, tbx15, tbx18 and hic1 ISH showed variable staining patterns between alleles at different stages. The two foxf2 mutant alleles differ on the genomic level in that FoxF2bCa22 is a single base pair insertion, while FoxF2bCa23 is a 14bp complex deletion. I found that both mutations are transcribed into RNA (i.e neither is degraded by nonsense-mediated decay and both are transcribed in mutant embryos). Whether protein levels were affected, however, is unclear. Preliminary Western blotting using a FoxF2 antibody developed in mouse suggested that FoxF2bCa23 protein is still produced, whereas the FoxF2bCa22 protein is greatly reduced compared to wildtype. The epitope recognized by this antibody shares 30% similarity of amino acids with FoxF2a, and may be cross-reacting with FoxF2a. If FoxF2bCa22 mutants have decreased FoxF2 protein (a and b), this potentially explains the differences in hemorrhage rates

150

between Ca22 and Ca23. The Ca23 allele may increase foxf2a expression and therefore not seemingly change protein amounts. The predicted protein products for the FoxF2b Ca22 and Ca23 alleles are disrupted by a frameshift at the same amino acid location, leading to a premature STOP codon after a chain of missense amino acid sequence. Both frameshifts begin early in the DNA binding domain, theoretically disrupting any ability of the protein to regulate transcription even if it was translated. Furthermore, human FoxF2 appears to be unstructured in this region suggesting that the N-terminus is not a functionally important part of the protein. The only difference between the two alleles is the missense sequence before the premature STOP. The FoxF2bCa22 allele has 21 missense amino acids (RLRSATPCDGNSERTNQKTDT), and the FoxF2bCa23 allele has 27 missense amino acids (RSGETALLLHCSHCDGNSERTNQKTDT). Analysis of the missense mutant peptide sequences does not show any similarity to known protein motifs and thus these domains are unlikely to be neomorphic (have acquired new functions). When I analyzed predicted mRNA secondary structures, I found large differences between Ca23 mRNA folding and wildtype, but not between the Ca22 allele and wildtype. Whether mRNA secondary structure blocks or inhibits translation of the missense peptide, whether the peptides are toxic if translated, or whether genetic background plays a role in the allele-specific differences remains to be determined. Differences in phenotypes could be attributed to background modifiers, as some lines of transgenic zebrafish (fli1:GFP and kdrl:GFP) are more prone to hemorrhage than others (acta2:GFP). Both mutant alleles were generated by crossing to the same transgenic line, Tg(acta2:GFP;kdrl:mCherry), but zebrafish are highly polymorphic (estimated 1 polymorphism per 47-84 bp in non-coding regions (Guryev et al., 2006)). Differential phenotypes arising from a modifier effect from the polymorphic background of the fish used for the initial outcross cannot be ruled out.

6.2.2 FoxF2a genetic compensation

As a result of the whole-genome duplication event that occurred in the teleost lineage of ray- finned fishes, zebrafish have developed two copies of many genes that only exist as single genes in the mammalian orthologs. While there is only one FoxF2 gene in humans and mice, zebrafish have a FoxF2a and FoxF2b paralogous pair. These genes have acquired differences in sequence, but remain more closely related to each other than similar Fox genes (Fig. 6.1). Their expression patterns are nearly identical in the developing zebrafish, save a few small differences, discussed previously. Genetic compensation is the upregulation of related or paralogous genes in response to deleterious 151

Figure 6.1 – Phylogenetic analysis of Fox proteins between zebrafish, mouse and human

Protein sequence comparison of zebrafish Fox gene cluster proteins (FoxC1a/b, FoxF2a/b, FoxQ1a/b) and mammalian FoxF family members. Zebrafish FoxF proteins are more closely related to mammalian FoxF family members than other zebrafish Fox genes (FoxC1 and FoxQ1). Paralogs of genes are most closely related to one another. Multiple sequence alignment and tree construction done with ClustalW.

152

genetic mutations, often resulting in rescue of the mutated gene’s function, and has been observed in zebrafish (Rossi et al., 2015). Indeed, a slight expansion of the foxf2a expression domain is seen in foxf2b mutants. A null FoxF2 knockout model, including genetic mutations in both FoxF2a and FoxF2b would prove useful in further identifying the role of FoxF2 in vascular stability and for comparison to FoxF2 mutant models in other organisms. I have initiated CRISPR mutagenesis to make this mutation in the FoxF2Ca22 background.

6.3 Roles for other Fox genes in vascular stabilization

6.3.1 FoxC1

FoxF2 is linked to FoxC1 in the genome. Given that FoxF2 and FoxC1 appear to mediate similar processes, FoxC1 is also a potential source of genetic compensation, although mouse studies did not find altered FoxC1 expression in FoxF2 mutants (Reyahi et al., 2015). In a phylogenetic tree FoxF2 and FoxC1 are not the most closely related pair, however their functions may have some overlap. FoxC1 knockdown and overexpression produces hemorrhage in zebrafish (French et al., 2014), as well as decreased vSMC maturation and increased perivascular space. FoxC1 mutations are associated with anterior segment dysgenesis (ASD) in the eye (Weisschuh et al., 2008; Reis et al., 2012), as neural crest-specific ablation of FoxC1 causes hypervascularization of the cornea (Seo et al., 2012). FoxF2 has also been linked to eye development, as one study connects a particular FoxF2 mutation to ASD (McKeone et al., 2011). Preliminary transcriptional profiling of FoxF2b zebrafish mutants in our lab shows a downregulation of vsx2 which is required in retinal progenitors (Zou & Levine, 2012), suggesting FoxF2 and C1 have the capacity to regulate similar development processes in the eye. The similar phenotypes of FoxC1 and F2 knockdowns in zebrafish (French et al., 2014), and the apparent exacerbation of CSVD phenotypes when both genes are segmentally deleted in humans (Chauhan et al., 2016) further support a notion that the evolutionarily conserved neighbouring of these genes corresponds to overlapping or shared regulatory roles. I did not test foxc1a/b upregulation in my FoxF2b mutants, and increases in FoxC1 protein would also be a potential mechanism of compensation for loss of FoxF2.

6.3.2 FoxQ1

FoxQ1 is also part of the Fox gene cluster along with FoxF2 and FoxC1, and is slightly more related to FoxF2 than FoxC1 (Fig. 6.1) Studies of FoxQ1 have been extensively focused on its role in

153

cancer, where it promotes metastasis through regulation of EMT genes such as Snail, Twist1 and E- Cadherin (Qiao et al., 2011; Huang et al., 2015; Li et al., 2016; Zhang et al., 2016a). Curiously, this appears to be opposite to the function of FoxF2, as mutations of FoxF2 in cancer promote EMT and metastasis. FoxQ1 also represses transcription of smooth muscle-specific genes in smooth muscle of the gut (Hoggatt et al., 2000). Although a role for FoxQ1 has not been identified in development of the vasculature, these EMT-promoting and smooth muscle-repressive functions suggest an opposite function to that of FoxF2. Zebrafish foxq1 expression can be seen in neural-crest derived tissues (Planchart & Mattingly, 2010), and I found expression is lost in foxf2b morphants, suggesting a functional relationship between these transcription factors. Further analysis of foxq1 expression and function in the head and vasculature during development may potentially show a novel role for FoxQ1 in promoting neural crest migration through EMT. Moreover, given the expression of FoxF2 and FoxC1 in neural crest and potential regulation of neural crest-derived tissue by FoxF2, this cluster of genes may be instrumental in the development of neural crest-derived tissues. Deletion of this conserved cluster may provide insight into why this tightly linked cluster of transcription factors has been maintained throughout vertebrate evolution. Zebrafish are excellent choices to explore functions of this cluster of genes, given the relatively small genomic size of the cluster compared to higher order organisms. Additionally, the genomic duplication of this cluster may protect against total lethality and offer the unique opportunity to assay dosage of a cluster of genes.

6.3.3 FoxF1

FoxF1 is the most closely related gene to FoxF2 and was identified alongside FoxF2 in the mouse lung (Hellqvist et al., 1996). These genes share expression domains, but differ in their levels of expression. Whereas FoxF2 shows a more dominant expression pattern in the neural crest and head mesenchyme, FoxF1 is predominant in gut and lung (Costa et al., 2003; Ormestad et al., 2004). FoxF1 is also necessary for lateral plate and extra-embryonic mesoderm (Mahlapuu et al., 2001b). In zebrafish, FoxF1 expression is also seen in gut and swim bladder (homologous to the lung) (Zheng et al., 2011). Despite these differences, there are several interesting parallels between the two family members. FoxF1 is activated by Shh signaling to promote development of lung and foregut in mouse (Mahlapuu et al., 2001a). Several phenotypes of FoxF1 heterozygous null mice are reminiscent of the mural cell phenotype in FoxF2b mutants. In the lung of FoxF1+/- mice, interaction between epithelial and mesenchymal tissues are disrupted and result in hemorrhage (Costa et al., 2003), and FoxF1 mutations in humans are associated with alveolar capillary dysplasia (Stankiewicz et al., 2009). 154

Interestingly, only about half of FoxF1+/- mice exhibit lung hemorrhage, while the other half expressed wildtype levels of FoxF1, suggesting compensation. These are similar to my hemorrhage results in FoxF2bCa22 and Ca23 mutant zebrafish (Kalinichenko et al., 2004). FoxF1+/- gall bladder loses its external smooth muscle layer (Kalinichenko et al., 2002), while smooth muscle-specific deletion of FoxF1 disrupts the expression of contractile genes of colonic smooth muscle (Hoggatt et al., 2013). Knockdown of Xenopus FoxF1 also impairs gut formation and reduces expression of Acta2 (Tseng et al., 2004). Many of these functions are reminiscent of FoxF2 promoting vSMC differentiation and interactions between mural cells and endothelium in zebrafish. In contrast to this, there also appears to be a role for FoxF1 in endothelium that is not evident in FoxF2b. FoxF1 is important for vasculogenesis (Astorga & Carlsson, 2007), as may be expected by its expression in, and regulation of, developing mesoderm. Endothelial-specific knockout of FoxF1 disrupts several aspects of endothelial development (Ren et al., 2014). Aside from this, however, it appears FoxF2 and FoxF1 are acting in different tissues to promote similar processes of mesenchymal cell differentiation for the purpose of supporting epithelial or endothelial cells. Evolutionary divergence of these genes may explain their differences in active domains but similarities in function. The ‘New Head’ theory suggests that, with the advent of forebrain and facial structures, the cranial neural crest has developed the ability to give rise to tissues in the head that are normally mesoderm-derived (Gans & Northcutt, 1983). The spatially divided expression and function of the FoxF family member between neural crest-derived and mesodermally-derived tissues is potentially a result of this evolutionary event.

6.4 Comparison between mouse and fish FoxF2 knockouts

A FoxF2 conditional knockout mouse model has been established and examined in the context of cerebral vessel integrity (Reyahi et al., 2015). These conditional mutants exhibit hemorrhage and leaky brain vessels at both embryonic and adult stages, in agreement with my observations of hemorrhage in zebrafish, however we have never been able to find evidence for vessel leakage in fish. Mouse FoxF2 is expressed in head mesenchyme associated with developing large vessels during development, and later is expressed in pericytes of the brain, again similar to FoxF2 expression in zebrafish. Specific expression of FoxF2 in migrating neural crest mesenchyme was observed in mouse, consistent with previous mouse studies (Ormestad et al., 2004), and I potentially observe expression of FoxF2 in this tissue in zebrafish as well. FoxF2 mutants have a thinning of the basal lamina, likely due to decreased Collagen IV. col4a2 expression appeared to be slightly decreased in my foxf2b morphant zebrafish, but thickness of basement membrane could not be easily assessed using my available TEM images. 155

PDGFRβ was decreased in FoxF2 mouse mutant embryos, and in contrast to the associated loss in pericytes that has been previously reported in PDGF-B and PDGFRβ mutants, there was a significant increase in the investment of neural crest-derived perivascular cells on endothelium. Furthermore, there was increasedproliferation of pericytes in mouse FoxF2 mutants, suggesting that pericytes were in a hyperproliferative state. pdgfrβ showed decreased expression in zebrafish FoxF2b mutants, but I did not assay proliferation. We have recently obtained germline PDGFRβ:GFP transgenic fish that could be used to mark this lineage in future. I observed decreased Acta2 expression at 4dpf, which would be consistent with mural cells preferentially proliferating at the expense of differentiation. However, examination of cerebral vessels by TEM presents contrasting findings between mice and zebrafish. The mouse vessels have thicker vessel walls and decreased lumen size, likely as a result of the increased investment of neural crest, whereas zebrafish mutant vessels suggest dilated lumens relative to vessel wall thickness, albeit an overall decrease in vessel size as well. Comparison between these is imperfect, since FoxF2 mutant mice cerebral vessels were examined embryonically, whereas zebrafish mutant brain vessels were examined at juvenile stages. Immunogold staining of endothelium, pericytes or smooth muscle cells would help direct TEM imaging of cerebral vessels in developing zebrafish. The mouse FoxF2 mutants also show decreased differentiation of the pericytes, as well as a delayed onset of desmin expression, similar to the delay in acta2 expression I observe in FoxF2bCa23 mutants. TGF-β signaling pathway constituents are decreased in FoxF2 mouse mutants, particularly those involved in mural cell differentiation, further suggesting loss of differentiation. My experiments do not include analysis the TGF-β pathway, yet I observe an increase in signaling pathways involved in proliferation, particularly the Shh effector Gli2, and PDGFRα/PDGFa. Additionally, while the FoxF2 mouse mutant studies have focused solely on pericytes in the brain, my work with the zebrafish FoxF2 knockdown and knockout models have also examined cells expressing the typical vSMC marker Acta2. In this way, these studies support each other in the concept of FoxF2-driven mural cell differentiation, but offer unique insights into complimentary aspects of vascular stability and BBB integrity through the examination of different mural cell types and signaling pathways.

6.5 Comparison of foxf2 mutants with other zebrafish genetic models of vascular stability

The zebrafish has become an attractive model in which to study development of vascular stability. The relative ease of gene knockdown by morpholino or knockout by targeted CRISPR mutagenesis, paired with the ex utero and translucent nature of zebrafish embryonic growth, facilitates manipulation and detection of vessel integrity. A common readout of vascular stability is hemorrhage, 156

and there are several zebrafish genetic mutants which exhibit this phenotype. Here I highlight a few that have identified or modeled different components for vascular stability and CSVD. A large part of my work has been initiated by mutagenesis screens aimed at identifying defects of the cardiovascular system (Stainier et al., 1996; Chen et al., 2001; Jin et al., 2007). Out of these came several hemorrhagic mutants, including bubblehead (bbh) which our lab determined to be a mutation in the Rho GEF arhgef7b (βPix) (Liu et al., 2007). Shortly after, mutants for the arhgef7b- interacting gene pak2a (redhead) were also found to hemorrhage (Buchner et al., 2007), and this work lead to the discovery that these genes interact with integrin αVβ8 to promote interactions between mural cells and endothelial cells, as well as promote cerebral angiogenesis (Liu et al., 2012). These mutants are the first zebrafish genetic models of integrin-mediated vascular instability. There is no evidence that FoxF2 interacts with this pathway, although we can determine this in future experiments. Endothelial cell survival is also, as one might expect, necessary for vascular integrity. Novel roles for ubiad1a and birc2 in endothelial cell survival were identified by zebrafish hemorrhagic mutants. Birc2 is involved in assembly of the survival-promoting TNFR complex I, and suppresses assembly of the TNFR complex II which promotes caspase activation (Santoro et al., 2007). Ubiad1a synthesizes vitamin K, and implicated vitamin K in the survival of cranial vasculature (Hegarty et al., 2013). In my TEM, I never observed a loss of integrity of the endothelium and therefore conclude that these pathways are unlikely to interact with FoxF2. Zebrafish mutants in TGF-β, Shh and Notch signaling pathways can also have vascular stability defects. The violet beauregarde (vbg) mutant affects activin receptor like kinase 1 (Acvrl1, otherwise known as Alk1), the gene which causes HHT in humans (Roman et al., 2002). Alk1 is a receptor for TGF-β, and the zebrafish acvrl1 mutant displays dilation in specific cranial vessels, disrupting systemic blood flow, although it does not hemorrhage. Several Shh signaling pathway mutants with defective cerebral vessel integrity have been isolated, and all appear to affect ciliary formation. I have already mentioned igu as a hemorrhagic mutant with defective ciliary basal bodies that disrupt cilia formation. However, intraflagellar transport (IFT) is important for transduction of signal as well. Mutations in three separate IFT proteins were identified as hemorrhagic, and exhibited decreased Shh signaling (Kallakuri et al., 2015). Interestingly, although igu mutants have previously showed disrupted Ang1 signaling, neither my FoxF2 mutants nor the IFT mutants exhibited changes in Ang1 or the Tie2 receptor. Zebrafish mutants in Notch and its downstream target genes (Hey, Hes, Her etc.) have given us much insight into the role of Notch signaling during development. Zebrafish Notch signaling 157

mediates fate decisions in the vasculature, as the zebrafish gridlock mutant, affecting hey2, has been instrumental in determining the mechanisms of artery-vein specification (Zhong et al., 2001), and mutation of mind bomb (an E3 ligase necessary for Notch signaling) shows decreased arterial markers (Lawson et al., 2001). Notch3 is expressed in mural cells, and zebrafish mutants for notch3 exhibit hemorrhage (Zaucker et al., 2013; Wang et al., 2014). Further analysis of this mutant revealed a decrease in the number of cerebral pericytes due to Notch3-mediated proliferation defects (Wang et al., 2014). Interestingly, a possible link between FoxF2 and Notch signaling has been recently identified by preliminary RNAseq transcriptional profiling of zebrafish foxf2b mutants (Whitesell and Childs unpublished). Several Notch signaling targets were downregulated in the foxf2b mutants, including hes5/her4.2, her4.5, hey1 and maml3 (Table 6.1). Given the similar decreased expression of pdgfrβ in both notch3 and foxf2b zebrafish mutants, FoxF2 may regulate PDGFRβ signaling via Notch. Experiments to elucidate the hierarchy of signaling pathways in regulating pericyte development and the role of FoxF2 in regulating these pathways should be a top priority going forward.

6.6 FoxF2 in Human Stroke

I was a co-first author on a large-scale study that found a link between ischemic stroke and the Fox gene cluster using genome-wide association study (GWAS). My role was to provide supporting functional data in zebrafish to show plausibility for SNPs near FoxF2 and stroke (Chauhan et al., 2016). This study analyzed genomes of 84 961 participants, 4348 of which had a stroke. This analysis identified several single nucleotide polymorphisms (SNPs) in linkage disequilibrium that were associated with stroke. These SNPs were located in the intergenic region between FoxF2 and FoxQ1 in the human genome, a locus that exhibits hallmarks of a regulatory region, including DNase hypersensitivity and an enrichment for transcription factor binding sites as reported by ENCODE (Fig. 6.2). These associations strongly suggest a role for FoxF2 and/or the conserved Fox gene cluster in stroke in a European population sample. Furthermore, these SNPs were highly associated with ischemic stroke events that were non-cardioembolic and/or occurred in small vessels, whereas associations with hemorrhagic, cardioembolic or large vessel stroke were generally non-significant. The analysis of the patient symptoms is however indicative of CSVD. Patients were commonly young (<32 years old) and had high levels of white matter hyperintensities indicative of CSVD. This agrees with a previous GWAS study paired with MRI data (Verhaaren et al., 2015) that showed a high association of these same SNPs with white matter hyperintensities.

158

Table 6.1 – Top genes downregulated in foxf2bCa23/Ca23 mutant embryos

159

Figure 6.2 – Human FoxF2 segmental mutations and stroke-associated SNPs

Association of SNPs in the FoxF2 intergenic region with incident all-stroke. rs12204590 indicates the SNP with highest association. Colour variation shows linkage disequilibrium between SNPs as calculated in 1000 Genomes Project (phase 1, version 3). Blue lines show estimated recombination rates. Coloured tracks seen at bottom of figure were added using the University of California, Santa Cruz genome browser and the RegulomeDB database. SNP track=SNPs encompassing the selected region, red dashed line shows position of top SNP. Regulome track=RegulomeDB scores, variants with lower scores have higher probability of acting as regulatory variants. DNase track=DNase hypersensitive regions assayed in 125 cell types (ENCODE project, Release 3, 2014). TFbs track=regions where transcription factors bind to DNA as assayed by ChIP-seq assay (ENCODE project, Release 3, 2013). Adapted from Chauhan et al. (Chauhan et al., 2016).

160

To extend my work to humans, I asked the Lehmann lab to test whether patients with deletions that encompassed FoxC1 and the adjacent FoxF2 locus demonstrated more severe white matter hyperintensities. They found a 10-fold increase in white matter hyperintensities in these patients although there are only two scans to date, suggesting that FoxF2 may also play a role in CSVD, and may be partially redundant with FoxC1 (Chauhan et al., 2016). My data along with this data from the Lehmann lab (human MRI imaging) and the Carlsson lab (conditional knockout in adult mice) support that reduction in FoxF2 is causative of cerebral vascular defects. The zebrafish data have been described previously in this thesis, (Chapter 4.1 and 4.2, respectively), and the human data has been mentioned, but the mouse data additionally demonstrated neural cell death as a result of ischemic and hemorrhagic insults (Fig. 6.3). This 2016 study is not the only GWAS study to investigate genetic determinants of stroke; previous stroke studies have identified SNPs in or near other genes (reviewed in (Falcone et al., 2014)). These other studies, however, have typically examined one subtype of stroke, such as intracerebral hemorrhage or large vessel ischemic stroke, whereas this GWAS incorporated all stroke incidents. Furthermore, it did so in a population predominantly of European- descent, which has previously not been examined as thoroughly (Falcone et al., 2014). Thus, these SNPs in a potential regulatory region of FOXF2 are a likely genetic risk factor for small vessel stroke in this particular population. This makes my work very relevant for understanding the mechanism of this type of stroke as it cannot be mechanistically studied in humans. In addition, a search of the DECIPHER database (Firth et al., 2009) suggests that large genomic deletions or duplications overlapping the human FoxF2 locus cause a range of phenotypes, many of which are associated with defects in intellectual disability (of the 32 patients in the database with phenotype descriptions, 11 of these 32 patients have a description of intellectual disability). Similarly, eye defects are common (8/32 with phenotypes including Axenfeld, Coloboma, RPE deficiency). This database is curated from patients who have copy number variants detected in this region, coupled with a description of their symptoms, but most patients do not have a definitive diagnosis. Together with our clinical collaborator (Ordan Lehmann, University of Alberta) we may be able to shed light on the consequences of reduction in FoxF2 levels on human disease.

6.7 Future Directions

My work has brought to light a potential role for FoxF2 in regulating vascular stability via differentiation of mural cells, but much more is still to be discovered. For FoxF2 in particular, establishing a full knockout of FoxF2 by targeted mutagenesis of the foxf2a paralog in a foxf2b mutant

161

Figure 6.3 – Human and mouse FoxF2 mutants exhibit cerebral insults

A-D: White matter hyperintensities are noted in periventricular region (arrowhead in A) and subcortical regions (arrowhead in B) in two patients with segmental deletion encompassing FOXC1. Mean white matter hyperintensities in two patients with segmental deletion of both FOXC1 and FOXF2 (arrowheads in C,D), show increased volume by more than ten times (E), in subcortical and periventricular regions. F: Foxf2 conditional knockout mouse cerebral cortex with condensed eosinophilic cytoplasm and pyknotic nuclei (to the right dashed line), indicative of recent ischaemic infarction. Normal tissue (F’) and tissue with ischaemic infarction (F’’) at higher magnification. G: Glial fibrillary acidic protein immunofluorescence of area with reactive astrogliosis in cerebral cortex of Foxf2 conditional knockout mouse. H,I: Cerebral cortex from showing normal neuronal tissue and intact capillaries in control mouse (H) and hemorrhage in Foxf2 conditional knockout mouse (I). Extravascular erythrocytes seen both as intact cells (pink arrows in I) and lysed cells (pink arrowheads in I). Scale bars = 20µm (50µm in G). Adapted from Chauhan et al. (Chauhan et al., 2016).

162

background may induce a more robust phenotype of hemorrhagic and vascular instability. I have initiated the mutagenesis, but it will be up to others to characterize the mutants. Additionally, knockout of the entire Fox gene cluster (FoxF2, C1 and Q1) may provide insight into shared regulatory roles and whether evolutionary conservation of the cluster was through necessity or chance. CRISPR mutagenesis would be ideal for this experiment. I predict that removing one locus containing all three genes will show a similar phenotype to loss of both alleles of FoxF2b. The recent discovery of stroke-associated SNPs in a potential regulatory region of human FoxF2 prompts the questions of what this domain is, what trans-activating factors does it interact with, and how are these SNPs conferring stroke. Zebrafish can provide an excellent model in which to investigate any regulatory ability of this enhancer and the effect of these SNPs on transcription. Our lab has initiated experiments to clone this region and identify whether it has regulatory potential, and whether this region containing the 7 European vs. 7 non-European SNP haplotypes confers different regulatory potential. Several pathways and processes appear to be affected by FoxF2, and further investigation of FoxF2 transcriptional targets will reveal its role in vascular stability. Of note, the notch pathway and PDGF pathway appear to be potential targets (direct or indirect) of FoxF2. A full understanding of the role of these pathways in vascular stabilization is needed. Finally, there are larger questions that remain to be answered. Does FoxF2 function in a context- specific fashion activated by the major signaling pathways of vascular stability, or is FoxF2 an integrator between upstream signalling in vascular stability (i.e. Shh), mediating complex interplay between downstream signaling pathways (PDGF, Notch)? Regardless, FoxF2 has emerged as an important regulator of endothelial-mural cell interactions, both during development and throughout life, and a further understanding of how this gene functions will benefit both basic and health research.

163

Chapter Seven: References

164

Abraham S, Kogata N, Fassler R, Adams RH. 2008. Integrin beta1 subunit controls mural cell adhesion, spreading, and blood vessel wall stability. Circ Res 102:562-570. Abramsson A, Lindblom P, Betsholtz C. 2003. Endothelial and nonendothelial sources of PDGF-B regulate pericyte recruitment and influence vascular pattern formation in tumors. J Clin Invest 112:1142-1151. Aitola M, Carlsson P, Mahlapuu M, Enerbäck S, Pelto-Huikko M. 2000. Forkhead transcription factor FoxF2 is expressed in mesodermal tissues involved in epithelio-mesenchymal interactions. Developmental Dynamics 218:136-149. Al Ahmad A, Gassmann M, Ogunshola OO. 2009. Maintaining blood-brain barrier integrity: pericytes perform better than astrocytes during prolonged oxygen deprivation. J Cell Physiol 218:612- 622. Allende ML, Proia RL. 2002. Sphingosine-1-phosphate receptors and the development of the vascular system. Biochim Biophys Acta 1582:222-227. Anders S, McCarthy DJ, Chen Y, Okoniewski M, Smyth GK, Huber W, Robinson MD. 2013. Count- based differential expression analysis of RNA sequencing data using R and Bioconductor. Nat Protoc 8:1765-1786. Ando K, Fukuhara S, Izumi N, Nakajima H, Fukui H, Kelsh RN, Mochizuki N. 2016. Clarification of mural cell coverage of vascular endothelial cells by live imaging of zebrafish. Development 143:1328-1339. Arboleda-Velasquez JF, Primo V, Graham M, James A, Manent J, D'Amore PA. 2014. Notch signaling functions in retinal pericyte survival. Invest Ophthalmol Vis Sci 55:5191-5199. Arimura K, Ago T, Kamouchi M, Nakamura K, Ishitsuka K, Kuroda J, Sugimori H, Ooboshi H, Sasaki T, Kitazono T. 2012. PDGF receptor beta signaling in pericytes following ischemic brain injury. Curr Neurovasc Res 9:1-9. Armulik A, Genove G, Mae M, Nisancioglu MH, Wallgard E, Niaudet C, He L, Norlin J, Lindblom P, Strittmatter K, Johansson BR, Betsholtz C. 2010. Pericytes regulate the blood-brain barrier. Nature 468:557-561. Arnold CR, Lamont RE, Walker JT, Spice PJ, Chan CK, Ho CY, Childs SJ. 2015. Comparative analysis of genes regulated by Dzip1/iguana and hedgehog in zebrafish. Dev Dyn 244:211-223. Arnold TD, Ferrero GM, Qiu H, Phan IT, Akhurst RJ, Huang EJ, Reichardt LF. 2012. Defective retinal vascular endothelial cell development as a consequence of impaired integrin alphaVbeta8- mediated activation of transforming growth factor-beta. J Neurosci 32:1197-1206. Arnold TD, Niaudet C, Pang MF, Siegenthaler J, Gaengel K, Jung B, Ferrero GM, Mukouyama YS, Fuxe J, Akhurst R, Betsholtz C, Sheppard D, Reichardt LF. 2014. Excessive vascular sprouting underlies cerebral hemorrhage in mice lacking alphaVbeta8-TGFbeta signaling in the brain. Development 141:4489-4499. Astorga J, Carlsson P. 2007. Hedgehog induction of murine vasculogenesis is mediated by Foxf1 and Bmp4. Development 134:3753-3761. Attwell D, Mishra A, Hall CN, O'Farrell FM, Dalkara T. 2016. What is a pericyte? J Cereb Blood Flow Metab 36:451-455. Bader BL, Smyth N, Nedbal S, Miosge N, Baranowsky A, Mokkapati S, Murshed M, Nischt R. 2005. Compound genetic ablation of nidogen 1 and 2 causes basement membrane defects and perinatal lethality in mice. Mol Cell Biol 25:6846-6856. Baeten JT, Lilly B. 2015. Differential Regulation of NOTCH2 and NOTCH3 Contribute to Their Unique Functions in Vascular Smooth Muscle Cells. J Biol Chem 290:16226-16237. Ballabh P. 2010. Intraventricular hemorrhage in premature infants: mechanism of disease. Pediatr Res 67:1-8.

165

Bandopadhyay R, Orte C, Lawrenson JG, Reid AR, De Silva S, Allt G. 2001. Contractile proteins in pericytes at the blood-brain and blood-retinal barriers. J Neurocytol 30:35-44. Barber AJ, Lieth E. 1997. Agrin accumulates in the brain microvascular basal lamina during development of the blood-brain barrier. Dev Dyn 208:62-74. Baumann E, Preston E, Slinn J, Stanimirovic D. 2009. Post-ischemic hypothermia attenuates loss of the vascular basement membrane proteins, agrin and SPARC, and the blood-brain barrier disruption after global cerebral ischemia. Brain Res 1269:185-197. Bedel A, Negre-Salvayre A, Heeneman S, Grazide MH, Thiers JC, Salvayre R, Maupas-Schwalm F. 2008. E-cadherin/beta-catenin/T-cell factor pathway is involved in smooth muscle cell proliferation elicited by oxidized low-density lipoprotein. Circ Res 103:694-701. Begemann G, Gibert Y, Meyer A, Ingham PW. 2002. Cloning of zebrafish T-box genes tbx15 and tbx18 and their expression during embryonic development. Mech Dev 114:137-141. Bell RD, Winkler EA, Sagare AP, Singh I, LaRue B, Deane R, Zlokovic BV. 2010. Pericytes control key neurovascular functions and neuronal phenotype in the adult brain and during brain aging. Neuron 68:409-427. Benabu Y, Beland M, Ferguson N, Maranda B, Boucher RM. 2013. Genetically proven cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL) in a 3-year-old. Pediatr Radiol 43:1227-1230. Berbari NF, O'Connor AK, Haycraft CJ, Yoder BK. 2009. The primary cilium as a complex signaling center. Curr Biol 19:R526-535. Bergeron SA, Milla LA, Villegas R, Shen MC, Burgess SM, Allende ML, Karlstrom RO, Palma V. 2008. Expression profiling identifies novel Hh/Gli-regulated genes in developing zebrafish embryos. Genomics 91:165-177. Bertrand S, Pinte S, Stankovic-Valentin N, Deltour-Balerdi S, Guerardel C, Begue A, Laudet V, Leprince D. 2004. Identification and developmental expression of the zebrafish orthologue of the tumor suppressor gene HIC1. Biochim Biophys Acta 1678:57-66. Bjarnegard M, Enge M, Norlin J, Gustafsdottir S, Fredriksson S, Abramsson A, Takemoto M, Gustafsson E, Fassler R, Betsholtz C. 2004. Endothelium-specific ablation of PDGFB leads to pericyte loss and glomerular, cardiac and placental abnormalities. Development 131:1847- 1857. Blanco R, Gerhardt H. 2013. VEGF and Notch in tip and stalk cell selection. Cold Spring Harb Perspect Med 3:a006569. Blum Y, Belting HG, Ellertsdottir E, Herwig L, Luders F, Affolter M. 2008. Complex cell rearrangements during intersegmental vessel sprouting and vessel fusion in the zebrafish embryo. Dev Biol 316:312-322. Bolte C, Ren X, Tomley T, Ustiyan V, Pradhan A, Hoggatt A, Kalin TV, Herring BP, Kalinichenko VV. 2015. Forkhead box F2 regulation of platelet-derived growth factor and myocardin/serum response factor signaling is essential for intestinal development. J Biol Chem 290:7563-7575. Bragin E, Chatzimichali EA, Wright CF, Hurles ME, Firth HV, Bevan AP, Swaminathan GJ. 2014. DECIPHER: database for the interpretation of phenotype-linked plausibly pathogenic sequence and copy-number variation. Nucleic Acids Res 42:D993-D1000. Brand M, Heisenberg CP, Warga RM, Pelegri F, Karlstrom RO, Beuchle D, Picker A, Jiang YJ, Furutani-Seiki M, van Eeden FJ, Granato M, Haffter P, Hammerschmidt M, Kane DA, Kelsh RN, Mullins MC, Odenthal J, Nusslein-Volhard C. 1996. Mutations affecting development of the midline and general body shape during zebrafish embryogenesis. Development 123:129- 142. Brennan RG. 1993. The winged-helix DNA-binding motif: another helix-turn-helix takeoff. Cell 74:773-776. 166

Buchner DA, Su F, Yamaoka JS, Kamei M, Shavit JA, Barthel LK, McGee B, Amigo JD, Kim S, Hanosh AW, Jagadeeswaran P, Goldman D, Lawson ND, Raymond PA, Weinstein BM, Ginsburg D, Lyons SE. 2007. pak2a mutations cause cerebral hemorrhage in redhead zebrafish. Proc Natl Acad Sci U S A 104:13996-14001. Bussmann J, Wolfe SA, Siekmann AF. 2011. Arterial-venous network formation during brain vascularization involves hemodynamic regulation of chemokine signaling. Development 138:1717-1726. Buttner A, Busch W, Kluver N, Giannis A, Scholz S. 2012. Transcriptional responses of zebrafish embryos exposed to potential sonic hedgehog pathway interfering compounds deviate from expression profiles of cyclopamine. Reprod Toxicol 33:254-263. Caffo M, Caruso G, Galatioto S, Meli F, Cacciola F, Germano A, Alafaci C, Tomasello F. 2008. Immunohistochemical study of the extracellular matrix proteins laminin, fibronectin and type IV collagen in secretory meningiomas. J Clin Neurosci 15:806-811. Cai CL, Martin JC, Sun Y, Cui L, Wang L, Ouyang K, Yang L, Bu L, Liang X, Zhang X, Stallcup WB, Denton CP, McCulloch A, Chen J, Evans SM. 2008a. A myocardial lineage derives from Tbx18 epicardial cells. Nature 454:104-108. Cai J, Kehoe O, Smith GM, Hykin P, Boulton ME. 2008b. The angiopoietin/Tie-2 system regulates pericyte survival and recruitment in diabetic retinopathy. Invest Ophthalmol Vis Sci 49:2163- 2171. Cai J, Tian AX, Wang QS, Kong PZ, Du X, Li XQ, Feng YM. 2015. FOXF2 suppresses the FOXC2- mediated epithelial-mesenchymal transition and multidrug resistance of basal-like breast cancer. Cancer Lett 367:129-137. Cairns DM, Liu R, Sen M, Canner JP, Schindeler A, Little DG, Zeng L. 2012. Interplay of Nkx3.2, Sox9 and Pax3 regulates chondrogenic differentiation of muscle progenitor cells. PLoS One 7:e39642. Carl TF, Dufton C, Hanken J, Klymkowsky MW. 1999. Inhibition of neural crest migration in Xenopus using antisense slug RNA. Dev Biol 213:101-115. Caspary T, Larkins CE, Anderson KV. 2007. The graded response to Sonic Hedgehog depends on cilia architecture. Dev Cell 12:767-778. Castets M, Coissieux MM, Delloye-Bourgeois C, Bernard L, Delcros JG, Bernet A, Laudet V, Mehlen P. 2009. Inhibition of endothelial cell apoptosis by netrin-1 during angiogenesis. Dev Cell 16:614-620. Cavanaugh AM, Huang J, Chen JN. 2015. Two developmentally distinct populations of neural crest cells contribute to the zebrafish heart. Dev Biol 404:103-112. Cermak T, Doyle EL, Christian M, Wang L, Zhang Y, Schmidt C, Baller JA, Somia NV, Bogdanove AJ, Voytas DF. 2011. Efficient design and assembly of custom TALEN and other TAL effector-based constructs for DNA targeting. Nucleic Acids Res 39:e82. Charidimou A, Shakeshaft C, Werring DJ. 2012. Cerebral microbleeds on magnetic resonance imaging and anticoagulant-associated intracerebral hemorrhage risk. Front Neurol 3:133. Chauhan G, Arnold CR, Chu AY, Fornage M, Reyahi A, Bis JC, Havulinna AS, Sargurupremraj M, Smith AV, Adams HHH, Choi SH, Pulit SL, Trompet S, Garcia ME, Manichaikul A, Teumer A, Gustafsson S, Bartz TM, Céline Bellenguez, Vidal JS, Jian X, Kjartansson O, Wiggins KL, Satizabal CL, Xue F, Ripatti S, Liu Y, Deelen J, Hoed Md, Bevan S, Hopewell JC, Malik R, Heckbert SR, Rice K, Smith NL, Levi C, Sharma P, Sudlow CL, Nik AM, Cole JW, Schmidt R, Meschia J, Thijs V, Lindgren A, Melander O, Grewal RP, Sacco RL, Rundek T, Rothwell PM, Arnett DK, Jern C, Johnson JA, Benavente OR, Wassertheil-Smoller S, Lee J-M, Wong Q, Aparicio HJ, Engelter ST, Kloss M, Leys D, Pezzini A, Buring JE, Ridker PM, Berr C, Dartigues J-F, Hamsten A, Magnusson PK, Traylor M, Pedersen NL, Lannfelt L, Lind L, 167

Lindgren CM, Morris AP, Jimenez-Conde J, Montaner J, Radmanesh F, Slowik A, Woo D, Hofman A, Koudstaal PJ, Portegies MLP, Uitterlinden AG, Craen AJMd, Ford I, Jukema JW, Stott DJ, Allen NB, Sale MM, Johnson AD, Bennett DA, Jager PLD, White CC, Grabe HJ, Markus MRP, Schminke U, Boncoraglio GB, Clarke R, Kamatani Y, Dallongeville J, Lopez OL, Rotter JI, Nalls MA, Gottesman RF, Griswold ME, Knopman DS, Windham BG, Beiser A, Markus HS, Vartiainen E, French CR, Dichgans M, Pastinen T, Lathrop M, Gudnason V, Kurth T, Psaty BM, Harris TB, Rich SS, deStefano AL, Schmidt CO, Worrall BB, Rosand J, Salomaa V, Mosley TH, Ingelsson E, Duijn CMv, Tzourio C, Rexrode KM, Lehmann OJ, Launer LJ, Ikram MA, Carlsson P, Chasman DI, Childs SJ, Junior WTL, Seshadri S, Debett S. 2016. Identification of additional risk loci for stroke and small vessel disease: a meta-analysis of genome-wide association studies. Lancet Neurol. Cheadle C, Vawter MP, Freed WJ, Becker KG. 2003. Analysis of microarray data using Z score transformation. J Mol Diagn 5:73-81. Chen JN, van Bebber F, Goldstein AM, Serluca FC, Jackson D, Childs S, Serbedzija G, Warren KS, Mably JD, Lindahl P, Mayer A, Haffter P, Fishman MC. 2001. Genetic steps to organ laterality in zebrafish. Comp Funct Genomics 2:60-68. Chen S, Kulik M, Lechleider RJ. 2003. Smad proteins regulate transcriptional induction of the SM22alpha gene by TGF-beta. Nucleic Acids Res 31:1302-1310. Chen W, Guo Y, Walker EJ, Shen F, Jun K, Oh SP, Degos V, Lawton MT, Tihan T, Davalos D, Akassoglou K, Nelson J, Pile-Spellman J, Su H, Young WL. 2013. Reduced mural cell coverage and impaired vessel integrity after angiogenic stimulation in the Alk1-deficient brain. Arterioscler Thromb Vasc Biol 33:305-310. Choi JC, Song SK, Lee JS, Kang SY, Kang JH. 2015. Diversity of stroke presentation in CADASIL: study from patients harboring the predominant NOTCH3 mutation R544C. J Stroke Cerebrovasc Dis 22:126-131. Christensen J, Bentz S, Sengstag T, Shastri VP, Anderle P. 2013. FOXQ1, a novel target of the Wnt pathway and a new marker for activation of Wnt signaling in solid tumors. PLoS One 8:e60051. Clark KL, Halay ED, Lai E, Burley SK. 1993. Co-crystal structure of the HNF-3/fork head DNA- recognition motif resembles histone H5. Nature 364:412-420. Cohen-Salmon M, Maxeiner S, Kruger O, Theis M, Willecke K, Petit C. 2004. Expression of the connexin43- and connexin45-encoding genes in the developing and mature mouse inner ear. Cell Tissue Res 316:15-22. Cohen MP, Frank RN, Khalifa AA. 1980. Collagen production by cultured retinal capillary pericytes. Invest Ophthalmol Vis Sci 19:90-94. Collins T, Ginsburg D, Boss JM, Orkin SH, Pober JS. 1985. Cultured human endothelial cells express platelet-derived growth factor B chain: cDNA cloning and structural analysis. Nature 316:748- 750. Corbit KC, Aanstad P, Singla V, Norman AR, Stainier DY, Reiter JF. 2005. Vertebrate Smoothened functions at the primary cilium. Nature 437:1018-1021. Costa RH, Kalinichenko VV, Holterman AX, Wang X. 2003. Transcription factors in liver development, differentiation, and regeneration. Hepatology 38:1331-1347. Craggs LJ, Fenwick R, Oakley AE, Ihara M, Kalaria RN. 2015. Immunolocalization of platelet- derived growth factor receptor-beta (PDGFR-beta) and pericytes in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL). Neuropathol Appl Neurobiol 41:557-570.

168

Cronan MR, Rosenberg AF, Oehlers SH, Saelens JW, Sisk DM, Jurcic Smith KL, Lee S, Tobin DM. 2015. CLARITY and PACT-based imaging of adult zebrafish and mouse for whole-animal analysis of infections. Dis Model Mech 8:1643-1650. Cuevas P, Gutierrez-Diaz JA, Reimers D, Dujovny M, Diaz FG, Ausman JI. 1984. Pericyte endothelial gap junctions in human cerebral capillaries. Anat Embryol (Berl) 170:155-159. Daneman R, Zhou L, Agalliu D, Cahoy JD, Kaushal A, Barres BA. 2010. The mouse blood-brain barrier transcriptome: a new resource for understanding the development and function of brain endothelial cells. PLoS One 5:e13741. David E S. 1986. The pericyteâ€”A review. Tissue and Cell 18:153-174. Davis S, Aldrich TH, Jones PF, Acheson A, Compton DL, Jain V, Ryan TE, Bruno J, Radziejewski C, Maisonpierre PC, Yancopoulos GD. 1996. Isolation of Angiopoietin-1, a Ligand for the TIE2 Receptor, by Secretion-Trap Expression Cloning. Cell 87:1161-1169. de Wilde J, Hulshof MF, Boekschoten MV, de Groot P, Smit E, Mariman EC. 2010. The embryonic genes Dkk3, Hoxd8, Hoxd9 and Tbx1 identify muscle types in a diet-independent and fiber- type unrelated way. BMC Genomics 11:176. Debette S, Markus HS. 2010. The clinical importance of white matter hyperintensities on brain magnetic resonance imaging: systematic review and meta-analysis. BMJ 341:c3666. Dejana E, Vestweber D. 2013. The role of VE-cadherin in vascular morphogenesis and permeability control. Prog Mol Biol Transl Sci 116:119-144. Diaz-Flores L, Gutierrez R, Madrid JF, Varela H, Valladares F, Acosta E, Martin-Vasallo P, Diaz- Flores L, Jr. 2009. Pericytes. Morphofunction, interactions and pathology in a quiescent and activated mesenchymal cell niche. Histol Histopathol 24:909-969. Dickson MC, Martin JS, Cousins FM, Kulkarni AB, Karlsson S, Akhurst RJ. 1995. Defective haematopoiesis and vasculogenesis in transforming growth factor-beta 1 knock out mice. Development 121:1845-1854. Domenga Vr, Fardoux P, Lacombe P, Monet M, Maciazek J, Krebs LT, Klonjkowski B, Berrou E, Mericskay M, Li Z, Tournier-Lasserve E, Gridley T, Joutel A. 2004. Notch3 is required for arterial identity and maturation of vascular smooth muscle cells. Genes & Development 18:2730-2735. Du SJ, Dienhart M. 2001. Gli2 mediation of hedgehog signals in slow muscle induction in zebrafish. Differentiation 67:84-91. Dumont DJ, Gradwohl GJ, Fong GH, Auerbach R, Breitman ML. 1993. The endothelial-specific receptor tyrosine kinase, tek, is a member of a new subfamily of receptors. Oncogene 8:1293- 1301. Duncan MR, Frazier KS, Abramson S, Williams S, Klapper H, Huang X, Grotendorst GR. 1999. Connective tissue growth factor mediates transforming growth factor beta-induced collagen synthesis: down-regulation by cAMP. Faseb J 13:1774-1786. Duz B, Oztas E, Erginay T, Erdogan E, Gonul E. 2007. The effect of moderate hypothermia in acute ischemic stroke on pericyte migration: an ultrastructural study. Cryobiology 55:279-284. Dziewulska D, Lewandowska E. 2012. Pericytes as a new target for pathological processes in CADASIL. Neuropathology 32:515-521. Enge M, Bjarnegard M, Gerhardt H, Gustafsson E, Kalen M, Asker N, Hammes HP, Shani M, Fassler R, Betsholtz C. 2002. Endothelium-specific platelet-derived growth factor-B ablation mimics diabetic retinopathy. Embo J 21:4307-4316. Etchevers HC, Vincent C, Le Douarin NM, Couly GF. 2001. The cephalic neural crest provides pericytes and smooth muscle cells to all blood vessels of the face and forebrain. Development 128:1059-1068.

169

Ezan J, Leroux L, Barandon L, Dufourcq P, Jaspard B, Moreau C, Allieres C, Daret D, Couffinhal T, Duplaa C. 2004. FrzA/sFRP-1, a secreted antagonist of the Wnt-Frizzled pathway, controls vascular cell proliferation in vitro and in vivo. Cardiovasc Res 63:731-738. Falcon BL, Hashizume H, Koumoutsakos P, Chou J, Bready JV, Coxon A, Oliner JD, McDonald DM. 2009. Contrasting actions of selective inhibitors of angiopoietin-1 and angiopoietin-2 on the normalization of tumor blood vessels. Am J Pathol 175:2159-2170. Falcone GJ, Malik R, Dichgans M, Rosand J. 2014. Current concepts and clinical applications of stroke genetics. Lancet Neurol 13:405-418. Fan DM, Feng XS, Qi PW, Chen YW. 2014. Forkhead factor FOXQ1 promotes TGF-beta1 expression and induces epithelial-mesenchymal transition. Mol Cell Biochem 397:179-186. Farin HF, Mansouri A, Petry M, Kispert A. 2008. T-box protein Tbx18 interacts with the paired box protein Pax3 in the development of the paraxial mesoderm. J Biol Chem 283:25372-25380. Federico A, Bianchi S, Dotti MT. 2005. The spectrum of mutations for CADASIL diagnosis. Neurol Sci 26:117-124. Figueroa XF, Duling BR. 2009. Gap junctions in the control of vascular function. Antioxid Redox Signal 11:251-266. Firth HV, Richards SM, Bevan AP, Clayton S, Corpas M, Rajan D, Van Vooren S, Moreau Y, Pettett RM, Carter NP. 2009. DECIPHER: Database of Chromosomal Imbalance and Phenotype in Humans Using Ensembl Resources. Am J Hum Genet 84:524-533. Flynn EJ, 3rd, Trent CM, Rawls JF. 2009. Ontogeny and nutritional control of adipogenesis in zebrafish (Danio rerio). J Lipid Res 50:1641-1652. Foster K, Sheridan J, Veiga-Fernandes H, Roderick K, Pachnis V, Adams R, Blackburn C, Kioussis D, Coles M. 2008. Contribution of neural crest-derived cells in the embryonic and adult thymus. J Immunol 180:3183-3189. Fouquet B, Weinstein BM, Serluca FC, Fishman MC. 1997. Vessel Patterning in the Embryo of the Zebrafish: Guidance by Notochord. Developmental Biology 183:37-48. French CR, Seshadri S, Destefano AL, Fornage M, Arnold CR, Gage PJ, Skarie JM, Dobyns WB, Millen KJ, Liu T, Dietz W, Kume T, Hofker M, Emery DJ, Childs SJ, Waskiewicz AJ, Lehmann OJ. 2014. Mutation of FOXC1 and PITX2 induces cerebral small-vessel disease. J Clin Invest 124:4877-4881. Fujimoto K. 1995. Pericyte-endothelial gap junctions in developing rat cerebral capillaries: a fine structural study. Anat Rec 242:562-565. Furuse M, Hirase T, Itoh M, Nagafuchi A, Yonemura S, Tsukita S, Tsukita S. 1993. Occludin: a novel integral membrane protein localizing at tight junctions. J Cell Biol 123:1777-1788. Gaengel K, GenovÃ© G, Armulik A, Betsholtz C. 2009. Endothelial-Mural Cell Signaling in Vascular Development and Angiogenesis. Arteriosclerosis, Thrombosis, and Vascular Biology 29:630- 638. Gage PJ, Rhoades W, Prucka SK, Hjalt T. 2005. Fate maps of neural crest and mesoderm in the mammalian eye. Invest Ophthalmol Vis Sci 46:4200-4208. Gans C, Northcutt RG. 1983. Neural crest and the origin of vertebrates: a new head. Science 220:268- 273. Garmy-Susini B, Jin H, Zhu Y, Sung RJ, Hwang R, Varner J. 2005. Integrin alpha4beta1-VCAM-1- mediated adhesion between endothelial and mural cells is required for blood vessel maturation. J Clin Invest 115:1542-1551. Gerhardt H, Golding M, Fruttiger M, Ruhrberg C, Lundkvist A, Abramsson A, Jeltsch M, Mitchell C, Alitalo K, Shima D, Betsholtz C. 2003. VEGF guides angiogenic sprouting utilizing endothelial tip cell filopodia. J Cell Biol 161:1163-1177.

170

Gerhardt H, Wolburg H, Redies C. 2000. N-cadherin mediates pericytic-endothelial interaction during brain angiogenesis in the chicken. Dev Dyn 218:472-479. Ghosh M, Balbi M, Hellal F, Dichgans M, Lindauer U, Plesnila N. 2015. Pericytes are involved in the pathogenesis of cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Ann Neurol 78:887-900. Ghosh S, Paez-Cortez JR, Boppidi K, Vasconcelos M, Roy M, Cardoso W, Ai X, Fine A. 2011. Activation dynamics and signaling properties of Notch3 receptor in the developing pulmonary artery. J Biol Chem 286:22678-22687. Glazer AM, Wilkinson AW, Backer CB, Lapan SW, Gutzman JH, Cheeseman IM, Reddien PW. 2010. The Zn Finger protein Iguana impacts Hedgehog signaling by promoting ciliogenesis. Developmental Biology 337:148-156. Gomez D, Owens GK. 2012. Smooth muscle cell phenotypic switching in atherosclerosis. Cardiovasc Res 95:156-164. Gonul E, Duz B, Kahraman S, Kayali H, Kubar A, Timurkaynak E. 2002. Early pericyte response to brain hypoxia in cats: an ultrastructural study. Microvasc Res 64:116-119. Gould DB, Phalan FC, Breedveld GJ, van Mil SE, Smith RS, Schimenti JC, Aguglia U, van der Knaap MS, Heutink P, John SW. 2005. Mutations in Col4a1 cause perinatal cerebral hemorrhage and porencephaly. Science 308:1167-1171. Goumans MJ, Lebrin F, Valdimarsdottir G. 2003. Controlling the angiogenic switch: a balance between two distinct TGF-b receptor signaling pathways. Trends Cardiovasc Med 13:301-307. Goumans MJ, Valdimarsdottir G, Itoh S, Rosendahl A, Sideras P, ten Dijke P. 2002. Balancing the activation state of the endothelium via two distinct TGF-beta type I receptors. Embo J 21:1743- 1753. Granata A, Bernard WG, Zhao N, McCafferty J, Lilly B, Sinha S. 2015. Temporal and embryonic lineage-dependent regulation of human vascular SMC development by NOTCH3. Stem Cells Dev 24:846-856. Grazioli A, Alves CS, Konstantopoulos K, Yang JT. 2006. Defective blood vessel development and pericyte/pvSMC distribution in alpha 4 integrin-deficient mouse embryos. Dev Biol 293:165- 177. Grenier J, Teillet MA, Grifone R, Kelly RG, Duprez D. 2009. Relationship between neural crest cells and cranial mesoderm during head muscle development. PLoS One 4:e4381. Greulich F, Farin HF, Schuster-Gossler K, Kispert A. 2012. Tbx18 function in epicardial development. Cardiovasc Res 96:476-483. Grieskamp T, Rudat C, Ludtke TH, Norden J, Kispert A. 2011. Notch signaling regulates smooth muscle differentiation of epicardium-derived cells. Circ Res 108:813-823. Gu X, Liu XY, Fagan A, Gonzalez-Toledo ME, Zhao LR. 2012. Ultrastructural changes in cerebral capillary pericytes in aged Notch3 mutant transgenic mice. Ultrastruct Pathol 36:48-55. Guner-Ataman B, Paffett-Lugassy N, Adams MS, Nevis KR, Jahangiri L, Obregon P, Kikuchi K, Poss KD, Burns CE, Burns CG. 2013. Zebrafish second heart field development relies on progenitor specification in anterior lateral plate mesoderm and nkx2.5 function. Development 140:1353- 1363. Guryev V, Koudijs MJ, Berezikov E, Johnson SL, Plasterk RH, van Eeden FJ, Cuppen E. 2006. Genetic variation in the zebrafish. Genome Res 16:491-497. Gustafsson E, Almonte-Becerril M, Bloch W, Costell M. 2013. Perlecan maintains microvessel integrity in vivo and modulates their formation in vitro. PLoS One 8:e53715. Haffter P, Granato M, Brand M, Mullins MC, Hammerschmidt M, Kane DA, Odenthal J, van Eeden FJ, Jiang YJ, Heisenberg CP, Kelsh RN, Furutani-Seiki M, Vogelsang E, Beuchle D, Schach U,

171

Fabian C, Nusslein-Volhard C. 1996. The identification of genes with unique and essential functions in the development of the zebrafish, Danio rerio. Development 123:1-36. Hall CN, Reynell C, Gesslein B, Hamilton NB, Mishra A, Sutherland BA, O'Farrell FM, Buchan AM, Lauritzen M, Attwell D. 2014. Capillary pericytes regulate cerebral blood flow in health and disease. Nature 508:55-60. Hallmann R, Horn N, Selg M, Wendler O, Pausch F, Sorokin LM. 2005. Expression and function of laminins in the embryonic and mature vasculature. Physiol Rev 85:979-1000. Hannenhalli S, Kaestner KH. 2009. The evolution of Fox genes and their role in development and disease. Nat Rev Genet 10:233-240. Hartley J, Westmacott R, Decker J, Shroff M, Yoon G. 2010. Childhood-onset CADASIL: clinical, imaging, and neurocognitive features. J Child Neurol 25:623-627. Hegarty JM, Yang H, Chi NC. 2013. UBIAD1-mediated vitamin K2 synthesis is required for vascular endothelial cell survival and development. Development 140:1713-1719. Heglind M, Cederberg A, Aquino J, Lucas G, Ernfors P, Enerback S. 2005. Lack of the central nervous system- and neural crest-expressed forkhead gene Foxs1 affects motor function and body weight. Mol Cell Biol 25:5616-5625. Hellqvist M, Mahlapuu M, Blixt Ãs, EnerbÃ¤ck S, Carlsson P. 1998. The Human Forkhead Protein FREAC-2 Contains Two Functionally Redundant Activation Domains and Interacts with TBP and TFIIB. Journal of Biological Chemistry 273:23335-23343. Hellqvist M, Mahlapuu M, Samuelsson L, EnerbÃ¤ck S, Carlsson P. 1996. Differential Activation of Lung-specific Genes by Two Forkhead Proteins, FREAC-1 and FREAC-2. Journal of Biological Chemistry 271:4482-4490. Hellstrom M, Kalen M, Lindahl P, Abramsson A, Betsholtz C. 1999. Role of PDGF-B and PDGFR- beta in recruitment of vascular smooth muscle cells and pericytes during embryonic blood vessel formation in the mouse. Development 126:3047-3055. Henshall TL, Keller A, He L, Johansson BR, Wallgard E, Raschperger E, Mae MA, Jin S, Betsholtz C, Lendahl U. 2015. Notch3 is necessary for blood vessel integrity in the central nervous system. Arterioscler Thromb Vasc Biol 35:409-420. High FA, Lu MM, Pear WS, Loomes KM, Kaestner KH, Epstein JA. 2008. Endothelial expression of the Notch ligand Jagged1 is required for vascular smooth muscle development. Proc Natl Acad Sci U S A 105:1955-1959. High FA, Zhang M, Proweller A, Tu L, Parmacek MS, Pear WS, Epstein JA. 2007. An essential role for Notch in neural crest during cardiovascular development and smooth muscle differentiation. J Clin Invest 117:353-363. Hirata H, Ueno K, Shahryari V, Deng G, Tanaka Y, Tabatabai ZL, Hinoda Y, Dahiya R. 2013. MicroRNA-182-5p promotes cell invasion and proliferation by down regulating FOXF2, RECK and MTSS1 genes in human prostate cancer. PLoS One 8:e55502. Hirschi KK, Burt JM, Hirschi KD, Dai C. 2003. Gap junction communication mediates transforming growth factor-beta activation and endothelial-induced mural cell differentiation. Circ Res 93:429-437. Hirschi KK, Rohovsky SA, Beck LH, Smith SR, D'Amore PA. 1999. Endothelial cells modulate the proliferation of mural cell precursors via platelet-derived growth factor-BB and heterotypic cell contact. Circ Res 84:298-305. Hoffmann AD, Yang XH, Burnicka-Turek O, Bosman JD, Ren X, Steimle JD, Vokes SA, McMahon AP, Kalinichenko VV, Moskowitz IP. 2014. Foxf genes integrate tbx5 and hedgehog pathways in the second heart field for cardiac septation. PLoS Genet 10:e1004604.

172

Hoggatt AM, Kim JR, Ustiyan V, Ren X, Kalin TV, Kalinichenko VV, Herring BP. 2013. The transcription factor Foxf1 binds to serum response factor and myocardin to regulate gene transcription in visceral smooth muscle cells. J Biol Chem 288:28477-28487. Hoggatt AM, Kriegel AM, Smith AF, Herring BP. 2000. Hepatocyte nuclear factor-3 homologue 1 (HFH-1) represses transcription of smooth muscle-specific genes. J Biol Chem 275:31162- 31170. Holz A, Kollmus H, Ryge J, Niederkofler V, Dias J, Ericson J, Stoeckli ET, Kiehn O, Arnold HH. 2010. The transcription factors Nkx2.2 and Nkx2.9 play a novel role in floor plate development and commissural axon guidance. Development 137:4249-4260. Hood LC, Rosenquist TH. 1992. Coronary artery development in the chick: origin and deployment of smooth muscle cells, and the effects of neural crest ablation. Anat Rec 234:291-300. Hori S, Ohtsuki S, Hosoya K, Nakashima E, Terasaki T. 2004. A pericyte-derived angiopoietin-1 multimeric complex induces occludin gene expression in brain capillary endothelial cells through Tie-2 activation in vitro. J Neurochem 89:503-513. Hoya K, Asai A, Sasaki T, Kimura K, Kirino T. 2001. Expression of smooth muscle proteins in cavernous and arteriovenous malformations. Acta Neuropathol 102:257-263. Huang W, Chen Z, Shang X, Tian D, Wang D, Wu K, Fan D, Xia L. 2015. Sox12, a direct target of FoxQ1, promotes hepatocellular carcinoma metastasis through up-regulating Twist1 and FGFBP1. Hepatology 61:1920-1933. Isogai S, Horiguchi M, Weinstein BM. 2001. The vascular anatomy of the developing zebrafish: an atlas of embryonic and early larval development. Dev Biol 230:278-301. Israeli D, Tanne D, Daniels D, Last D, Shneor R, Guez D, Landau E, Roth Y, Ocherashvilli A, Bakon M, Hoffman C, Weinberg A, Volk T, Mardor Y. 2010. The application of MRI for depiction of subtle blood brain barrier disruption in stroke. Int J Biol Sci 7:1-8. Izawa Y, Yoshizumi M, Ishizawa K, Fujita Y, Kondo S, Kagami S, Kawazoe K, Tsuchiya K, Tomita S, Tamaki T. 2007. Big mitogen-activated protein kinase 1 (BMK1)/extracellular signal regulated kinase 5 (ERK5) is involved in platelet-derived growth factor (PDGF)-induced vascular smooth muscle cell migration. Hypertens Res 30:1107-1117. Jackson BC, Carpenter C, Nebert DW, Vasiliou V. 2010. Update of human and mouse forkhead box (FOX) gene families. Hum Genomics 4:345-352. Jain RK. 2003. Molecular regulation of vessel maturation. Nat Med 9:685-693. Jakobsen JS, Braun M, Astorga J, Gustafson EH, Sandmann T, Karzynski M, Carlsson P, Furlong EEM. 2007. Temporal ChIP-on-chip reveals Biniou as a universal regulator of the visceral muscle transcriptional network. Genes & Development 21:2448-2460. Jin J, Zhu SJ, Zhu ZM, Yang YJ, Ding G. 2002. [Relationship between proliferation of vascular smooth muscle cells and PDGF-AA and PDGFR-alpha expression in SHRs]. Sheng Li Xue Bao 54:145-148. Jin SW, Herzog W, Santoro MM, Mitchell TS, Frantsve J, Jungblut B, Beis D, Scott IC, D'Amico LA, Ober EA, Verkade H, Field HA, Chi NC, Wehman AM, Baier H, Stainier DY. 2007. A transgene-assisted genetic screen identifies essential regulators of vascular development in vertebrate embryos. Dev Biol 307:29-42. Jin Z, Mei W, Strack S, Jia J, Yang J. 2011. The antagonistic action of B56-containing protein phosphatase 2As and casein kinase 2 controls the phosphorylation and Gli turnover function of Daz interacting protein 1. J Biol Chem 286:36171-36179. Jochumsen U, Werner R, Miura N, Richter-Unruh A, Hiort O, Holterhus PM. 2008. Mutation analysis of FOXF2 in patients with disorders of sex development (DSD) in combination with cleft palate. Sex Dev 2:302-308.

173

Jones TJ, Adapala RK, Geldenhuys WJ, Bursley C, AbouAlaiwi WA, Nauli SM, Thodeti CK. 2012. Primary cilia regulates the directional migration and barrier integrity of endothelial cells through the modulation of hsp27 dependent actin cytoskeletal organization. J Cell Physiol 227:70-76. Joutel A, Andreux F, Gaulis S, Domenga V, Cecillon M, Battail N, Piga N, Chapon F, Godfrain C, Tournier-Lasserve E. 2000. The ectodomain of the Notch3 receptor accumulates within the cerebrovasculature of CADASIL patients. J Clin Invest 105:597-605. Joutel A, Corpechot C, Ducros A, Vahedi K, Chabriat H, Mouton P, Alamowitch S, Domenga V, Cecillion M, Marechal E, Maciazek J, Vayssiere C, Cruaud C, Cabanis EA, Ruchoux MM, Weissenbach J, Bach JF, Bousser MG, Tournier-Lasserve E. 1996. Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia. Nature 383:707- 710. Kaestner KH, Knochel W, Martinez DE. 2000. Unified nomenclature for the winged helix/forkhead transcription factors. Genes Dev. 14:142-146. Kalinichenko VV, Gusarova GA, Kim IM, Shin B, Yoder HM, Clark J, Sapozhnikov AM, Whitsett JA, Costa RH. 2004. Foxf1 haploinsufficiency reduces Notch-2 signaling during mouse lung development. Am J Physiol Lung Cell Mol Physiol 286:L521-530. Kalinichenko VV, Zhou Y, Bhattacharyya D, Kim W, Shin B, Bambal K, Costa RH. 2002. Haploinsufficiency of the mouse Forkhead Box f1 gene causes defects in gall bladder development. J Biol Chem 277:12369-12374. Kallakuri S, Yu JA, Li J, Li Y, Weinstein BM, Nicoli S, Sun Z. 2015. Endothelial cilia are essential for developmental vascular integrity in zebrafish. J Am Soc Nephrol 26:864-875. Katakami N, Kaneto H, Hao H, Umayahara Y, Fujitani Y, Sakamoto K, Gorogawa S, Yasuda T, Kawamori D, Kajimoto Y, Matsuhisa M, Yutani C, Hori M, Yamasaki Y. 2004. Role of pim-1 in smooth muscle cell proliferation. J Biol Chem 279:54742-54749. Khoshnoodi J, Pedchenko V, Hudson BG. 2008. Mammalian collagen IV. Microsc Res Tech 71:357- 370. Kim HR, Richardson J, van Eeden F, Ingham PW. 2010. Gli2a protein localization reveals a role for Iguana/DZIP1 in primary ciliogenesis and a dependence of Hedgehog signal transduction on primary cilia in the zebrafish. BMC Biol 8:65. Kim JH, Kim JH, Yu YS, Kim DH, Kim KW. 2009. Recruitment of pericytes and astrocytes is closely related to the formation of tight junction in developing retinal vessels. J Neurosci Res 87:653- 659. Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling TF. 1995. Stages of embryonic development of the zebrafish. Dev Dyn 203:253-310. Kohli V, Schumacher JA, Desai SP, Rehn K, Sumanas S. 2013. Arterial and venous progenitors of the major axial vessels originate at distinct locations. Dev Cell 25:196-206. Kok FO, Shin M, Ni CW, Gupta A, Grosse AS, van Impel A, Kirchmaier BC, Peterson-Maduro J, Kourkoulis G, Male I, DeSantis DF, Sheppard-Tindell S, Ebarasi L, Betsholtz C, Schulte- Merker S, Wolfe SA, Lawson ND. 2015. Reverse genetic screening reveals poor correlation between morpholino-induced and mutant phenotypes in zebrafish. Dev Cell 32:97-108. Kokovay E, Li L, Cunningham LA. 2006. Angiogenic recruitment of pericytes from bone marrow after stroke. J Cereb Blood Flow Metab 26:545-555. Kondo S, Hashimoto N, Kikuchi H, Hazama F, Nagata I, Kataoka H. 1998. Apoptosis of medial smooth muscle cells in the development of saccular cerebral aneurysms in rats. Stroke 29:181- 188; discussion 189.

174

Kong PZ, Yang F, Li L, Li XQ, Feng YM. 2013. Decreased FOXF2 mRNA expression indicates early- onset metastasis and poor prognosis for breast cancer patients with histological grade II tumor. PLoS One 8:e61591. Korn J, Christ B, Kurz H. 2002. Neuroectodermal origin of brain pericytes and vascular smooth muscle cells. J Comp Neurol 442:78-88. Kraus F, Haenig B, Kispert A. 2001. Cloning and expression analysis of the mouse T-box gene Tbx18. Mech Dev 100:83-86. Krebs LT, Shutter JR, Tanigaki K, Honjo T, Stark KL, Gridley T. 2004. Haploinsufficient lethality and formation of arteriovenous malformations in Notch pathway mutants. Genes Dev 18:2469- 2473. Kruger O, Plum A, Kim JS, Winterhager E, Maxeiner S, Hallas G, Kirchhoff S, Traub O, Lamers WH, Willecke K. 2000. Defective vascular development in connexin 45-deficient mice. Development 127:4179-4193. Krupinski J, Issa R, Bujny T, Slevin M, Kumar P, Kumar S, Kaluza J. 1997. A putative role for platelet-derived growth factor in angiogenesis and neuroprotection after ischemic stroke in humans. Stroke 28:564-573. Kruyer A, Soplop N, Strickland S, Norris EH. 2015. Chronic Hypertension Leads to Neurodegeneration in the TgSwDI Mouse Model of Alzheimer's Disease. Hypertension 66:175-182. Kucenas S, Snell H, Appel B. 2008. nkx2.2a promotes specification and differentiation of a myelinating subset of oligodendrocyte lineage cells in zebrafish. Neuron Glia Biol 4:71-81. Kundu ST, Byers LA, Peng DH, Roybal JD, Diao L, Wang J, Tong P, Creighton CJ, Gibbons DL. 2016. The miR-200 family and the miR-183~96~182 cluster target Foxf2 to inhibit invasion and metastasis in lung cancers. Oncogene 35:173-186. Kuo DS, Labelle-Dumais C, Gould DB. 2012. COL4A1 and COL4A2 mutations and disease: insights into pathogenic mechanisms and potential therapeutic targets. Hum Mol Genet 21:R97-110. Kuwajima T, Sitko AA, Bhansali P, Jurgens C, Guido W, Mason C. 2013. ClearT: a detergent- and solvent-free clearing method for neuronal and non-neuronal tissue. Development 140:1364- 1368. Kwon SH, Li L, He Y, Tey JC, Li H, Zhuplatov I, Kim SJ, Terry CM, Blumenthal DK, Shiu YT, Cheung AK. 2015. Prevention of Venous Neointimal Hyperplasia by a Multitarget Receptor Tyrosine Kinase Inhibitor. J Vasc Res 52:244-256. LaBonne C, Bronner-Fraser M. 2000. Snail-related transcriptional repressors are required in Xenopus for both the induction of the neural crest and its subsequent migration. Dev Biol 221:195-205. Lafleur MA, Forsyth PA, Atkinson SJ, Murphy G, Edwards DR. 2001. Perivascular cells regulate endothelial membrane type-1 matrix metalloproteinase activity. Biochem Biophys Res Commun 282:463-473. Lamont R. 2008. Molecular Pathways in Zebrafish Vascular Development. In: Biochemistry and Molecular Biology: University of Calgary. Lamont RE, Childs S. 2006. MAPping out arteries and veins. Sci STKE 2006:pe39. Lamont RE, Vu W, Carter AD, Serluca FC, MacRae CA, Childs SJ. 2010. Hedgehog signaling via angiopoietin1 is required for developmental vascular stability. Mechanisms of Development 127:159-168. Lan Y, Jiang R. 2009. Sonic hedgehog signaling regulates reciprocal epithelial-mesenchymal interactions controlling palatal outgrowth. Development 136:1387-1396. Larrieu-Lahargue F, Thomas KR, Li DY. 2012. Netrin ligands and receptors: lessons from neurons to the endothelium. Trends Cardiovasc Med 22:44-47.

175

Larrivee B, Prahst C, Gordon E, del Toro R, Mathivet T, Duarte A, Simons M, Eichmann A. 2012. ALK1 signaling inhibits angiogenesis by cooperating with the Notch pathway. Dev Cell 22:489-500. Larson DM, Carson MP, Haudenschild CC. 1987. Junctional transfer of small molecules in cultured bovine brain microvascular endothelial cells and pericytes. Microvasc Res 34:184-199. Larsson J, Goumans MJ, Sjostrand LJ, van Rooijen MA, Ward D, Leveen P, Xu X, ten Dijke P, Mummery CL, Karlsson S. 2001. Abnormal angiogenesis but intact hematopoietic potential in TGF-beta type I receptor-deficient mice. Embo J 20:1663-1673. Lausch E, Hermanns P, Farin HF, Alanay Y, Unger S, Nikkel S, Steinwender C, Scherer G, Spranger J, Zabel B, Kispert A, Superti-Furga A. 2008. TBX15 mutations cause craniofacial dysmorphism, hypoplasia of scapula and pelvis, and short stature in Cousin syndrome. Am J Hum Genet 83:649-655. Lauter G, Soll I, Hauptmann G. 2011. Multicolor fluorescent in situ hybridization to define abutting and overlapping gene expression in the embryonic zebrafish brain. Neural Dev 6:10. Lawson ND, Scheer N, Pham VN, Kim CH, Chitnis AB, Campos-Ortega JA, Weinstein BM. 2001. Notch signaling is required for arterial-venous differentiation during embryonic vascular development. Development 128:3675-3683. Lawson ND, Vogel AM, Weinstein BM. 2002. sonic hedgehog and vascular endothelial growth factor Act Upstream of the Notch Pathway during Arterial Endothelial Differentiation. Developmental Cell 3:127-136. Lee KY, Singh MK, Ussar S, Wetzel P, Hirshman MF, Goodyear LJ, Kispert A, Kahn CR. 2015. Tbx15 controls skeletal muscle fibre-type determination and muscle metabolism. Nat Commun 6:8054. Leimeister C, Schumacher N, Steidl C, Gessler M. 2000. Analysis of HeyL expression in wild-type and Notch pathway mutant mouse embryos. Mech Dev 98:175-178. Lemmens R, Maugeri A, Niessen HW, Goris A, Tousseyn T, Demaerel P, Corveleyn A, Robberecht W, van der Knaap MS, Thijs VN, Zwijnenburg PJ. 2013. Novel COL4A1 mutations cause cerebral small vessel disease by haploinsufficiency. Hum Mol Genet 22:391-397. Lenard A, Daetwyler S, Betz C, Ellertsdottir E, Belting HG, Huisken J, Affolter M. 2015. Endothelial cell self-fusion during vascular pruning. PLoS Biol 13:e1002126. Leveen P, Pekny M, Gebre-Medhin S, Swolin B, Larsson E, Betsholtz C. 1994. Mice deficient for PDGF B show renal, cardiovascular, and hematological abnormalities. Genes Dev 8:1875- 1887. Li F, Duman-Scheel M, Yang D, Du W, Zhang J, Zhao C, Qin L, Xin S. 2010. Sonic hedgehog signaling induces vascular smooth muscle cell proliferation via induction of the G1 cyclin- retinoblastoma axis. Arterioscler Thromb Vasc Biol 30:1787-1794. Li F, Lan Y, Wang Y, Wang J, Yang G, Meng F, Han H, Meng A, Wang Y, Yang X. 2011. Endothelial Smad4 maintains cerebrovascular integrity by activating N-cadherin through cooperation with Notch. Dev Cell 20:291-302. Li H, Li J, Li Y, Singh P, Cao L, Xu LJ, Li D, Wang Y, Xie Z, Gui Y, Zheng XL. 2012. Sonic hedgehog promotes autophagy of vascular smooth muscle cells. Am J Physiol Heart Circ Physiol 303:H1319-1331. Li L, Li L, Xie F, Zhang Z, Guo Y, Tang G, Lv D, Lu Q, Chen L, Li J. 2013. Jagged-1/Notch3 signaling transduction pathway is involved in apelin-13-induced vascular smooth muscle cells proliferation. Acta Biochim Biophys Sin (Shanghai) 45:875-881. Li L, Miano JM, Cserjesi P, Olson EN. 1996. SM22 alpha, a marker of adult smooth muscle, is expressed in multiple myogenic lineages during embryogenesis. Circ Res 78:188-195.

176

Li Y, Zhang Y, Yao Z, Li S, Yin Z, Xu M. 2016. Forkhead box Q1: A key player in the pathogenesis of tumors (Review). Int J Oncol 49:51-58. Liang D, Wang X, Mittal A, Dhiman S, Hou SY, Degenhardt K, Astrof S. 2014. Mesodermal expression of integrin alpha5beta1 regulates neural crest development and cardiovascular morphogenesis. Dev Biol 395:232-244. Lindahl P, Johansson BR, Leveen P, Betsholtz C. 1997. Pericyte loss and microaneurysm formation in PDGF-B-deficient mice. Science 277:242-245. Liu H, Kennard S, Lilly B. 2009. NOTCH3 expression is induced in mural cells through an autoregulatory loop that requires endothelial-expressed JAGGED1. Circ Res 104:466-475. Liu H, Zhang W, Kennard S, Caldwell RB, Lilly B. 2010. Notch3 is critical for proper angiogenesis and mural cell investment. Circ Res 107:860-870. Liu J, Fraser SD, Faloon PW, Rollins EL, Vom Berg J, Starovic-Subota O, Laliberte AL, Chen JN, Serluca FC, Childs SJ. 2007. A betaPix Pak2a signaling pathway regulates cerebral vascular stability in zebrafish. Proc Natl Acad Sci U S A 104:13990-13995. Liu J, Wang Y, Akamatsu Y, Lee CC, Stetler RA, Lawton MT, Yang GY. 2014. Vascular remodeling after ischemic stroke: mechanisms and therapeutic potentials. Prog Neurobiol 115:138-156. Liu J, Zeng L, Kennedy RM, Gruenig NM, Childs SJ. 2012. betaPix plays a dual role in cerebral vascular stability and angiogenesis, and interacts with integrin alphavbeta8. Dev Biol 363:95- 105. Lu X, Le Noble F, Yuan L, Jiang Q, De Lafarge B, Sugiyama D, Breant C, Claes F, De Smet F, Thomas JL, Autiero M, Carmeliet P, Tessier-Lavigne M, Eichmann A. 2004. The netrin receptor UNC5B mediates guidance events controlling morphogenesis of the vascular system. Nature 432:179-186. Luo Y, Radice GL. 2005. N-cadherin acts upstream of VE-cadherin in controlling vascular morphogenesis. J Cell Biol 169:29-34. Lv D, Meng D, Zou FF, Fan L, Zhang P, Yu Y, Fang J. 2011. Activating transcription factor 3 regulates survivability and migration of vascular smooth muscle cells. IUBMB Life 63:62-69. Macdonald RL, Weir BK. 1991. A review of hemoglobin and the pathogenesis of cerebral vasospasm. Stroke 22:971-982. Mack CP, Owens GK. 1999. Regulation of smooth muscle alpha-actin expression in vivo is dependent on CArG elements within the 5' and first intron promoter regions. Circ Res 84:852-861. Mahlapuu M, Enerback S, Carlsson P. 2001a. Haploinsufficiency of the forkhead gene Foxf1, a target for sonic hedgehog signaling, causes lung and foregut malformations. Development 128:2397- 2406. Mahlapuu M, Ormestad M, Enerback S, Carlsson P. 2001b. The forkhead transcription factor Foxf1 is required for differentiation of extra-embryonic and lateral plate mesoderm. Development 128:155-166. Majesky MW. 2007. Developmental Basis of Vascular Smooth Muscle Diversity. Arteriosclerosis, Thrombosis, and Vascular Biology 27:1248-1258. Mandarino LJ, Sundarraj N, Finlayson J, Hassell HR. 1993. Regulation of fibronectin and laminin synthesis by retinal capillary endothelial cells and pericytes in vitro. Exp Eye Res 57:609-621. Manderfield LJ, High FA, Engleka KA, Liu F, Li L, Rentschler S, Epstein JA. 2012. Notch activation of Jagged1 contributes to the assembly of the arterial wall. Circulation 125:314-323. Martin-Padura I, Lostaglio S, Schneemann M, Williams L, Romano M, Fruscella P, Panzeri C, Stoppacciaro A, Ruco L, Villa A, Simmons D, Dejana E. 1998. Junctional adhesion molecule, a novel member of the immunoglobulin superfamily that distributes at intercellular junctions and modulates monocyte transmigration. J Cell Biol 142:117-127.

177

Matsumoto K, Ema M. 2014. Roles of VEGF-A signalling in development, regeneration, and tumours. J Biochem 156:1-10. Matsusaka T. 1975. Tridimensional views of the relationship of pericytes to endothelial cells of capillaries in the human choroid and retina. J Electron Microsc (Tokyo) 24:13-18. McCarty JH, Cook AA, Hynes RO. 2005a. An interaction between {alpha}v{beta}8 integrin and Band 4.1B via a highly conserved region of the Band 4.1 C-terminal domain. Proc Natl Acad Sci U S A 102:13479-13483. McCarty JH, Lacy-Hulbert A, Charest A, Bronson RT, Crowley D, Housman D, Savill J, Roes J, Hynes RO. 2005b. Selective ablation of alphav integrins in the central nervous system leads to cerebral hemorrhage, seizures, axonal degeneration and premature death. Development 132:165-176. McKeone R, Vieira H, Gregory-Evans K, Gregory-Evans CY, Denny P. 2011. Foxf2: a novel locus for anterior segment dysgenesis adjacent to the Foxc1 gene. PLoS One 6:e25489. McLin VA, Shah R, Desai NP, Jamrich M. 2010. Identification and gastrointestinal expression of Xenopus laevis FoxF2. Int J Dev Biol 54:919-924. Melgar MA, Rafols J, Gloss D, Diaz FG. 2005. Postischemic reperfusion: ultrastructural blood-brain barrier and hemodynamic correlative changes in an awake model of transient forebrain ischemia. Neurosurgery 56:571-581. Menezes MJ, McClenahan FK, Leiton CV, Aranmolate A, Shan X, Colognato H. 2014. The extracellular matrix protein laminin alpha2 regulates the maturation and function of the blood- brain barrier. J Neurosci 34:15260-15280. Meng F, Speyer CL, Zhang B, Zhao Y, Chen W, Gorski DH, Miller FR, Wu G. 2015. PDGFRalpha and beta play critical roles in mediating Foxq1-driven breast cancer stemness and chemoresistance. Cancer Res 75:584-593. Meng JS, Okeda R. 2001. Histopathological structure of the pial arteriovenous malformation in adults: observation by reconstruction of serial sections of four surgical specimens. Acta Neuropathol 102:63-68. Mikawa T, Gourdie RG. 1996. Pericardial mesoderm generates a population of coronary smooth muscle cells migrating into the heart along with ingrowth of the epicardial organ. Dev Biol 174:221-232. Miller CT, Yelon D, Stainier DY, Kimmel CB. 2003. Two endothelin 1 effectors, hand2 and bapx1, pattern ventral pharyngeal cartilage and the jaw joint. Development 130:1353-1365. Mima A, Abe H, Nagai K, Arai H, Matsubara T, Araki M, Torikoshi K, Tominaga T, Iehara N, Fukatsu A, Kita T, Doi T. 2011. Activation of Src mediates PDGF-induced Smad1 phosphorylation and contributes to the progression of glomerulosclerosis in glomerulonephritis. PLoS One 6:e17929. Miura N, Kakinuma H, Sato M, Aiba N, Terada K, Sugiyama T. 1998. Mouse forkhead (winged helix) gene LUN encodes a transactivator that acts in the lung. Genomics 50:346-356. Mongera A, Singh AP, Levesque MP, Chen YY, Konstantinidis P, Nusslein-Volhard C. 2013. Genetic lineage labeling in zebrafish uncovers novel neural crest contributions to the head, including gill pillar cells. Development 140:916-925. Morrow D, Scheller A, Birney YA, Sweeney C, Guha S, Cummins PM, Murphy R, Walls D, Redmond EM, Cahill PA. 2005. Notch-mediated CBF-1/RBP-J{kappa}-dependent regulation of human vascular smooth muscle cell phenotype in vitro. Am J Physiol Cell Physiol 289:C1188-1196. Nakagomi T, Kubo S, Nakano-Doi A, Sakuma R, Lu S, Narita A, Kawahara M, Taguchi A, Matsuyama T. 2015. Brain vascular pericytes following ischemia have multipotential stem cell activity to differentiate into neural and vascular lineage cells. Stem Cells 33:1962-1974.

178

Nik AM, Johansson JA, Ghiami M, Reyahi A, Carlsson P. 2016. Foxf2 is required for secondary palate development and Tgfbeta signaling in palatal shelf mesenchyme. Dev Biol 415:14-23. Nik AM, Reyahi A, Ponten F, Carlsson P. 2013. Foxf2 in intestinal fibroblasts reduces numbers of Lgr5(+) stem cells and adenoma formation by inhibiting Wnt signaling. Gastroenterology 144:1001-1011. Nishida W, Nakamura M, Mori S, Takahashi M, Ohkawa Y, Tadokoro S, Yoshida K, Hiwada K, Hayashi K, Sobue K. 2002. A triad of serum response factor and the GATA and NK families governs the transcription of smooth and cardiac muscle genes. J Biol Chem 277:7308-7317. Nitta T, Hata M, Gotoh S, Seo Y, Sasaki H, Hashimoto N, Furuse M, Tsukita S. 2003. Size-selective loosening of the blood-brain barrier in claudin-5-deficient mice. J Cell Biol 161:653-660. Norrving B. 2008. Lacunar infarcts: no black holes in the brain are benign. Pract Neurol 8:222-228. Nuessle JM, Giehl K, Herzog R, Stracke S, Menke A. 2011. TGFbeta1 suppresses vascular smooth muscle cell motility by expression of N-cadherin. Biol Chem 392:461-474. Oh SP, Seki T, Goss KA, Imamura T, Yi Y, Donahoe PK, Li L, Miyazono K, ten Dijke P, Kim S, Li E. 2000. Activin receptor-like kinase 1 modulates transforming growth factor-beta 1 signaling in the regulation of angiogenesis. Proc Natl Acad Sci U S A 97:2626-2631. Ormestad M, Astorga J, Carlsson P. 2004. Differences in the embryonic expression patterns of mouse Foxf1 and -2 match their distinct mutant phenotypes. Dev Dyn 229:328-333. Ormestad M, Astorga J, Landgren H, Wang T, Johansson BR, Miura N, Carlsson P. 2006. Foxf1 and Foxf2 control murine gut development by limiting mesenchymal Wnt signaling and promoting extracellular matrix production. Development 133:833-843. Osada T, Gu YH, Kanazawa M, Tsubota Y, Hawkins BT, Spatz M, Milner R, del Zoppo GJ. 2011. Interendothelial claudin-5 expression depends on cerebral endothelial cell-matrix adhesion by beta(1)-integrins. J Cereb Blood Flow Metab 31:1972-1985. PÃ©rez SÃ¡nchez C, Casas-TintÃ³ S, SÃ¡nchez L, Rey-Campos J, Granadino Ba. 2002. DmFoxF, a novel Drosophila fork head factor expressed in visceral mesoderm. Mechanisms of Development 111:163-166. Pabst O, Rummelies J, Winter B, Arnold HH. 2003. Targeted disruption of the homeobox gene Nkx2.9 reveals a role in development of the spinal accessory nerve. Development 130:1193-1202. Peppiatt CM, Howarth C, Mobbs P, Attwell D. 2006. Bidirectional control of CNS capillary diameter by pericytes. Nature 443:700-704. Perumal K, Dirr HW, Fanucchi S. 2015. A Single Amino Acid in the Hinge Loop Region of the FOXP Forkhead Domain is Significant for Dimerisation. Protein J 34:111-121. Pierrou S, Enerback S, Carlsson P. 1995. Selection of high-affinity binding sites for sequence-specific, DNA binding proteins from random sequence oligonucleotides. Anal Biochem 229:99-105. Pierrou S, Hellqvist M, Samuelsson L, Enerback S, Carlsson P. 1994. Cloning and characterization of seven human forkhead proteins: binding site specificity and DNA bending. Embo J 13:5002- 5012. Planchart A, Mattingly CJ. 2010. 2,3,7,8-Tetrachlorodibenzo-p-dioxin upregulates FoxQ1b in zebrafish jaw primordium. Chem Res Toxicol 23:480-487. Pola R, Ling LE, Silver M, Corbley MJ, Kearney M, Blake Pepinsky R, Shapiro R, Taylor FR, Baker DP, Asahara T, Isner JM. 2001. The morphogen Sonic hedgehog is an indirect angiogenic agent upregulating two families of angiogenic growth factors. Nat Med 7:706-711. Pouget C, Gautier R, Teillet MA, Jaffredo T. 2006. Somite-derived cells replace ventral aortic hemangioblasts and provide aortic smooth muscle cells of the trunk. Development 133:1013- 1022. Pouget C, Pottin K, Jaffredo T. 2008. Sclerotomal origin of vascular smooth muscle cells and pericytes in the embryo. Dev Biol 315:437-447. 179

Privratsky JR, Newman PJ. 2014. PECAM-1: regulator of endothelial junctional integrity. Cell Tissue Res 355:607-619. Proulx K, Lu A, Sumanas S. 2010. Cranial vasculature in zebrafish forms by angioblast cluster-derived angiogenesis. Developmental Biology 348:34-46. Qi Y, Cai J, Wu Y, Wu R, Lee J, Fu H, Rao M, Sussel L, Rubenstein J, Qiu M. 2001. Control of oligodendrocyte differentiation by the Nkx2.2 homeodomain transcription factor. Development 128:2723-2733. Qiao Y, Jiang X, Lee ST, Karuturi RK, Hooi SC, Yu Q. 2011. FOXQ1 regulates epithelial- mesenchymal transition in human cancers. Cancer Res 71:3076-3086. Quasnichka H, Slater SC, Beeching CA, Boehm M, Sala-Newby GB, George SJ. 2006. Regulation of smooth muscle cell proliferation by beta-catenin/T-cell factor signaling involves modulation of cyclin D1 and p21 expression. Circ Res 99:1329-1337. Rannikmae K, Davies G, Thomson PA, Bevan S, Devan WJ, Falcone GJ, Traylor M, Anderson CD, Battey TW, Radmanesh F, Deka R, Woo JG, Martin LJ, Jimenez-Conde J, Selim M, Brown DL, Silliman SL, Kidwell CS, Montaner J, Langefeld CD, Slowik A, Hansen BM, Lindgren AG, Meschia JF, Fornage M, Bis JC, Debette S, Ikram MA, Longstreth WT, Schmidt R, Zhang CR, Yang Q, Sharma P, Kittner SJ, Mitchell BD, Holliday EG, Levi CR, Attia J, Rothwell PM, Poole DL, Boncoraglio GB, Psaty BM, Malik R, Rost N, Worrall BB, Dichgans M, Van Agtmael T, Woo D, Markus HS, Seshadri S, Rosand J, Sudlow CL. 2015. Common variation in COL4A1/COL4A2 is associated with sporadic cerebral small vessel disease. Neurology 84:918-926. Reis LM, Tyler RC, Volkmann Kloss BA, Schilter KF, Levin AV, Lowry RB, Zwijnenburg PJ, Stroh E, Broeckel U, Murray JC, Semina EV. 2012. PITX2 and FOXC1 spectrum of mutations in ocular syndromes. Eur J Hum Genet 20:1224-1233. Ren X, Ustiyan V, Pradhan A, Cai Y, Havrilak JA, Bolte CS, Shannon JM, Kalin TV, Kalinichenko VV. 2014. FOXF1 transcription factor is required for formation of embryonic vasculature by regulating VEGF signaling in endothelial cells. Circ Res 115:709-720. Resink TJ, Scott-Burden T, Hahn AW, Rouge M, Hosang M, Powell JS, Buhler FR. 1990. Specific growth stimulation of cultured smooth muscle cells from spontaneously hypertensive rats by platelet-derived growth factor A-chain homodimer. Cell Regul 1:821-831. Rey FE, Li XC, Carretero OA, Garvin JL, Pagano PJ. 2002. Perivascular superoxide anion contributes to impairment of endothelium-dependent relaxation: role of gp91(phox). Circulation 106:2497- 2502. Reyahi A, Nik AM, Ghiami M, Gritli-Linde A, Ponten F, Johansson BR, Carlsson P. 2015. Foxf2 Is Required for Brain Pericyte Differentiation and Development and Maintenance of the Blood- Brain Barrier. Dev Cell 34:19-32. Rinon A, Lazar S, Marshall H, Buchmann-Moller S, Neufeld A, Elhanany-Tamir H, Taketo MM, Sommer L, Krumlauf R, Tzahor E. 2007. Cranial neural crest cells regulate head muscle patterning and differentiation during vertebrate embryogenesis. Development 134:3065-3075. Robinson MD, McCarthy DJ, Smyth GK. 2010. edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics 26:139-140. Robu ME, Larson JD, Nasevicius A, Beiraghi S, Brenner C, Farber SA, Ekker SC. 2007. p53 activation by knockdown technologies. PLoS Genet 3:e78. Rogers CD, Jayasena CS, Nie S, Bronner ME. 2012. Neural crest specification: tissues, signals, and transcription factors. Wiley Interdiscip Rev Dev Biol 1:52-68. Roman BL, Pham VN, Lawson ND, Kulik M, Childs S, Lekven AC, Garrity DM, Moon RT, Fishman MC, Lechleider RJ, Weinstein BM. 2002. Disruption of acvrl1 increases endothelial cell number in zebrafish cranial vessels. Development 129:3009-3019. 180

Rood BR, Leprince D. 2013. Deciphering HIC1 control pathways to reveal new avenues in cancer therapeutics. Expert Opin Ther Targets 17:811-827. Rossi A, Kontarakis Z, Gerri C, Nolte H, Holper S, Kruger M, Stainier DY. 2015. Genetic compensation induced by deleterious mutations but not gene knockdowns. Nature 524:230- 233. Rouget C. 1873. Mémoire sur le développement, la structure et les propriétés physiologiques des capillaires sanguins et lymphatiques. Archives de Physologie Normale et Pathologique 5:601- 631. Ruchoux MM, Guerouaou D, Vandenhaute B, Pruvo JP, Vermersch P, Leys D. 1995. Systemic vascular smooth muscle cell impairment in cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy. Acta Neuropathol 89:500-512. Rucker HK, Wynder HJ, Thomas WE. 2000. Cellular mechanisms of CNS pericytes. Brain Res Bull 51:363-369. Saji N, Toba K, Sakurai T. 2016. Cerebral Small Vessel Disease and Arterial Stiffness: Tsunami Effect in the Brain? Pulse (Basel) 3:182-189. Santa Y, Uyama E, Chui DH, Arima M, Kotorii S, Takahashi K, Tabira T. 2003. Genetic, clinical and pathological studies of CADASIL in Japan: a partial contribution of Notch3 mutations and implications of smooth muscle cell degeneration for the pathogenesis. J Neurol Sci 212:79-84. Santoro MM, Pesce G, Stainier DY. 2009. Characterization of vascular mural cells during zebrafish development. Mechanisms of Development 126:638-649. Santoro MM, Samuel T, Mitchell T, Reed JC, Stainier DY. 2007. Birc2 (cIap1) regulates endothelial cell integrity and blood vessel homeostasis. Nat Genet 39:1397-1402. Schulz GB, Wieland E, Wustehube-Lausch J, Boulday G, Moll I, Tournier-Lasserve E, Fischer A. 2015. Cerebral Cavernous Malformation-1 Protein Controls DLL4-Notch3 Signaling Between the Endothelium and Pericytes. Stroke 46:1337-1343. Schwend T, Jin Z, Jiang K, Mitchell BJ, Jia J, Yang J. 2013. Stabilization of speckle-type POZ protein (Spop) by Daz interacting protein 1 (Dzip1) is essential for Gli turnover and the proper output of Hedgehog signaling. J Biol Chem 288:32809-32820. Sekimizu K, Nishioka N, Sasaki H, Takeda H, Karlstrom RO, Kawakami A. 2004. The zebrafish iguana locus encodes Dzip1, a novel zinc-finger protein required for proper regulation of Hedgehog signaling. Development 131:2521-2532. Seo S, Singh HP, Lacal PM, Sasman A, Fatima A, Liu T, Schultz KM, Losordo DW, Lehmann OJ, Kume T. 2012. Forkhead box transcription factor FoxC1 preserves corneal transparency by regulating vascular growth. Proc Natl Acad Sci U S A 109:2015-2020. Shao J, Cao J, Liu Y, Mei H, Zhang Y, Xu W. 2015. MicroRNA-519a promotes proliferation and inhibits apoptosis of hepatocellular carcinoma cells by targeting FOXF2. FEBS Open Bio 5:893-899. Shen J, Ishii Y, Xu G, Dang TC, Hamashima T, Matsushima T, Yamamoto S, Hattori Y, Takatsuru Y, Nabekura J, Sasahara M. 2012. PDGFR-beta as a positive regulator of tissue repair in a mouse model of focal cerebral ischemia. J Cereb Blood Flow Metab 32:353-367. Shen TL, Park AY, Alcaraz A, Peng X, Jang I, Koni P, Flavell RA, Gu H, Guan JL. 2005. Conditional knockout of focal adhesion kinase in endothelial cells reveals its role in angiogenesis and vascular development in late embryogenesis. J Cell Biol 169:941-952. Shen X, Cui J, Gong Q. 2011. Fox gene loci in Takifugu rubripes and Tetraodon nigroviridis genomes and comparison with those of medaka and zebrafish genomes. Genome 54:965-972. Shi W, Gerster K, Alajez NM, Tsang J, Waldron L, Pintilie M, Hui AB, Sykes J, P'ng C, Miller N, McCready D, Fyles A, Liu FF. 2011. MicroRNA-301 mediates proliferation and invasion in human breast cancer. Cancer Res 71:2926-2937. 181

Shi Z, Liu J, Yu X, Huang J, Shen S, Zhang Y, Han R, Ge N, Yang Y. 2016. Loss of FOXF2 Expression Predicts Poor Prognosis in Hepatocellular Carcinoma Patients. Ann Surg Oncol 23:211-217. Sibon I, Coupry I, Menegon P, Bouchet JP, Gorry P, Burgelin I, Calvas P, Orignac I, Dousset V, Lacombe D, Orgogozo JM, Arveiler B, Goizet C. 2007. COL4A1 mutation in Axenfeld-Rieger anomaly with leukoencephalopathy and stroke. Ann Neurol 62:177-184. Siegenthaler JA, Choe Y, Patterson KP, Hsieh I, Li D, Jaminet SC, Daneman R, Kume T, Huang EJ, Pleasure SJ. 2013. Foxc1 is required by pericytes during fetal brain angiogenesis. Biol Open 2:647-659. Simon C, Lickert H, Gotz M, Dimou L. 2012. Sox10-iCreERT2 : a mouse line to inducibly trace the neural crest and oligodendrocyte lineage. Genesis 50:506-515. Singh MK, Petry M, Haenig B, Lescher B, Leitges M, Kispert A. 2005. The T-box transcription factor Tbx15 is required for skeletal development. Mech Dev 122:131-144. Sjolund M, Rahm M, Claesson-Welsh L, Sejersen T, Heldin CH, Thyberg J. 1990. Expression of PDGF alpha- and beta-receptors in rat arterial smooth muscle cells is phenotype and growth state dependent. Growth Factors 3:191-203. Slater SC, Koutsouki E, Jackson CL, Bush RC, Angelini GD, Newby AC, George SJ. 2004. R- cadherin:beta-catenin complex and its association with vascular smooth muscle cell proliferation. Arterioscler Thromb Vasc Biol 24:1204-1210. Soriano P. 1994. Abnormal kidney development and hematological disorders in PDGF beta-receptor mutant mice. Genes Dev 8:1888-1896. Sorrell MR, Dohn TE, D'Aniello E, Waxman JS. 2013. Tcf7l1 proteins cell autonomously restrict cardiomyocyte and promote endothelial specification in zebrafish. Dev Biol 380:199-210. Spice P. 2014. Nkx3.2 regulates zebrafish vascular stabilization through recruitment of mural cells. In: Biochemistry and Molecular Biology: University of Calgary. Srivastava D, Thomas T, Lin Q, Kirby ML, Brown D, Olson EN. 1997. Regulation of cardiac mesodermal and neural crest development by the bHLH transcription factor, dHAND. Nat Genet 16:154-160. Stainier DY, Fouquet B, Chen JN, Warren KS, Weinstein BM, Meiler SE, Mohideen MA, Neuhauss SC, Solnica-Krezel L, Schier AF, Zwartkruis F, Stemple DL, Malicki J, Driever W, Fishman MC. 1996. Mutations affecting the formation and function of the cardiovascular system in the zebrafish embryo. Development 123:285-292. Stankiewicz P, Sen P, Bhatt SS, Storer M, Xia Z, Bejjani BA, Ou Z, Wiszniewska J, Driscoll DJ, Maisenbacher MK, Bolivar J, Bauer M, Zackai EH, McDonald-McGinn D, Nowaczyk MM, Murray M, Hustead V, Mascotti K, Schultz R, Hallam L, McRae D, Nicholson AG, Newbury R, Durham-O'Donnell J, Knight G, Kini U, Shaikh TH, Martin V, Tyreman M, Simonic I, Willatt L, Paterson J, Mehta S, Rajan D, Fitzgerald T, Gribble S, Prigmore E, Patel A, Shaffer LG, Carter NP, Cheung SW, Langston C, Shaw-Smith C. 2009. Genomic and genic deletions of the FOX gene cluster on 16q24.1 and inactivating mutations of FOXF1 cause alveolar capillary dysplasia and other malformations. Am J Hum Genet 84:780-791. Stenzel D, Nye E, Nisancioglu M, Adams RH, Yamaguchi Y, Gerhardt H. 2009. Peripheral mural cell recruitment requires cell-autonomous heparan sulfate. Blood 114:915-924. Stewart KS, Zhou Z, Zweidler-McKay P, Kleinerman ES. 2011. Delta-like ligand 4â€“Notch signaling regulates bone marrowâ€“derived pericyte/vascular smooth muscle cell formation. Blood 117:719-726. Strahle U, Fischer N, Blader P. 1997. Expression and regulation of a netrin homologue in the zebrafish embryo. Mech Dev 62:147-160. Stroud JC. 2006. Structure of the forkhead domain of FOXP2 bound to DNA. Structure 14:159-166. 182

Sumoy L, Keasey JB, Dittman TD, Kimelman D. 1997. A role for notochord in axial vascular development revealed by analysis of phenotype and the expression of VEGR-2 in zebrafish flh and ntl mutant embryos. Mech Dev 63:15-27. Sundberg C, Kowanetz M, Brown LF, Detmar M, Dvorak HF. 2002. Stable expression of angiopoietin-1 and other markers by cultured pericytes: phenotypic similarities to a subpopulation of cells in maturing vessels during later stages of angiogenesis in vivo. Lab Invest 82:387-401. Suri C, Jones PF, Patan S, Bartunkova S, Maisonpierre PC, Davis S, Sato TN, Yancopoulos GD. 1996. Requisite role of angiopoietin-1, a ligand for the TIE2 receptor, during embryonic angiogenesis. Cell 87:1171-1180. Sweeney C, Morrow D, Birney YA, Coyle S, Hennessy C, Scheller A, Cummins PM, Walls D, Redmond EM, Cahill PA. 2004. Notch 1 and 3 receptor signaling modulates vascular smooth muscle cell growth, apoptosis, and migration via a CBF-1/RBP-Jk dependent pathway. Faseb J 18:1421-1423. Szeto DP, Griffin KJ, Kimelman D. 2002. HrT is required for cardiovascular development in zebrafish. Development 129:5093-5101. Takai Y, Irie K, Shimizu K, Sakisaka T, Ikeda W. 2003. Nectins and nectin-like molecules: roles in cell adhesion, migration, and polarization. Cancer Sci 94:655-667. Takeuchi JK, Mileikovskaia M, Koshiba-Takeuchi K, Heidt AB, Mori AD, Arruda EP, Gertsenstein M, Georges R, Davidson L, Mo R, Hui CC, Henkelman RM, Nemer M, Black BL, Nagy A, Bruneau BG. 2005. Tbx20 dose-dependently regulates transcription factor networks required for mouse heart and motoneuron development. Development 132:2463-2474. Tam SJ, Richmond DL, Kaminker JS, Modrusan Z, Martin-McNulty B, Cao TC, Weimer RM, Carano RA, van Bruggen N, Watts RJ. 2012. Death receptors DR6 and TROY regulate brain vascular development. Dev Cell 22:403-417. Tang Y, Yang X, Friesel RE, Vary CP, Liaw L. 2011. Mechanisms of TGF-beta-induced differentiation in human vascular smooth muscle cells. J Vasc Res 48:485-494. Tay SY, Yu X, Wong KN, Panse P, Ng CP, Roy S. 2010. The iguana/DZIP1 protein is a novel component of the ciliogenic pathway essential for axonemal biogenesis. Dev Dyn 239:527-534. Theveneau E, Mayor R. 2012. Neural crest delamination and migration: from epithelium-to- mesenchyme transition to collective cell migration. Dev Biol 366:34-54. Thisse C, Thisse B. 2008. High-resolution in situ hybridization to whole-mount zebrafish embryos. Nat Protoc 3:59-69. Thisse C, Thisse B, Postlethwait JH. 1995. Expression of snail2, a second member of the zebrafish snail family, in cephalic mesendoderm and presumptive neural crest of wild-type and spadetail mutant embryos. Dev Biol 172:86-99. Tidhar A, Reichenstein M, Cohen D, Faerman A, Copeland NG, Gilbert DJ, Jenkins NA, Shani M. 2001. A novel transgenic marker for migrating limb muscle precursors and for vascular smooth muscle cells. Dev Dyn 220:60-73. Tien T, Muto T, Barrette K, Challyandra L, Roy S. 2014. Downregulation of Connexin 43 promotes vascular cell loss and excess permeability associated with the development of vascular lesions in the diabetic retina. Mol Vis 20:732-741. Tillet E, Vittet D, Feraud O, Moore R, Kemler R, Huber P. 2005. N-cadherin deficiency impairs pericyte recruitment, and not endothelial differentiation or sprouting, in embryonic stem cell- derived angiogenesis. Exp Cell Res 310:392-400. Trainor PA, Tan SS, Tam PP. 1994. Cranial paraxial mesoderm: regionalisation of cell fate and impact on craniofacial development in mouse embryos. Development 120:2397-2408.

183

Trost A, Schroedl F, Lange S, Rivera FJ, Tempfer H, Korntner S, Stolt CC, Wegner M, Bogner B, Kaser-Eichberger A, Krefft K, Runge C, Aigner L, Reitsamer HA. 2013. Neural crest origin of retinal and choroidal pericytes. Invest Ophthalmol Vis Sci 54:7910-7921. Truett GE, Heeger P, Mynatt RL, Truett AA, Walker JA, Warman ML. 2000. Preparation of PCR- quality mouse genomic DNA with hot sodium hydroxide and tris (HotSHOT). Biotechniques 29:52, 54. Tseng HT, Shah R, Jamrich M. 2004. Function and regulation of FoxF1 during Xenopus gut development. Development 131:3637-3647. Ulrich F, Ma LH, Baker RG, Torres-Vazquez J. 2011. Neurovascular development in the embryonic zebrafish hindbrain. Dev Biol 357:134-151. Uranishi R, Baev NI, Kim JH, Awad IA. 2001. Vascular smooth muscle cell differentiation in human cerebral vascular malformations. Neurosurgery 49:671-679; discussion 679-680. Urness LD, Sorensen LK, Li DY. 2000. Arteriovenous malformations in mice lacking activin receptor- like kinase-1. Nat Genet 26:328-331. van den Brink GR, Rubin DC. 2013. Foxf2: a mesenchymal regulator of intestinal adenoma development. Gastroenterology 144:873-876. van der Heul-Nieuwenhuijsen L, Dits N, Van Ijcken W, de Lange D, Jenster G. 2009a. The FOXF2 pathway in the human prostate stroma. The Prostate 69:1538-1547. van der Heul-Nieuwenhuijsen L, Dits NF, Jenster G. 2009b. Gene expression of forkhead transcription factors in the normal and diseased human prostate. BJU International 103:1574-1580. van Sloten TT, Protogerou AD, Henry RM, Schram MT, Launer LJ, Stehouwer CD. 2015. Association between arterial stiffness, cerebral small vessel disease and cognitive impairment: A systematic review and meta-analysis. Neurosci Biobehav Rev 53:121-130. Verbeek E, Meuwissen ME, Verheijen FW, Govaert PP, Licht DJ, Kuo DS, Poulton CJ, Schot R, Lequin MH, Dudink J, Halley DJ, de Coo RI, den Hollander JC, Oegema R, Gould DB, Mancini GM. 2012. COL4A2 mutation associated with familial porencephaly and small-vessel disease. Eur J Hum Genet 20:844-851. Verbeek MM, Van Nostrand WE, Otte-Holler I, Wesseling P, De Waal RM. 2000. Amyloid-beta- induced degeneration of human brain pericytes is dependent on the apolipoprotein E genotype. Ann N Y Acad Sci 903:187-199. Verhaaren BF, Debette S, Bis JC, Smith JA, Ikram MK, Adams HH, Beecham AH, Rajan KB, Lopez LM, Barral S, van Buchem MA, van der Grond J, Smith AV, Hegenscheid K, Aggarwal NT, de Andrade M, Atkinson EJ, Beekman M, Beiser AS, Blanton SH, Boerwinkle E, Brickman AM, Bryan RN, Chauhan G, Chen CP, Chouraki V, de Craen AJ, Crivello F, Deary IJ, Deelen J, De Jager PL, Dufouil C, Elkind MS, Evans DA, Freudenberger P, Gottesman RF, Guethnason V, Habes M, Heckbert SR, Heiss G, Hilal S, Hofer E, Hofman A, Ibrahim-Verbaas CA, Knopman DS, Lewis CE, Liao J, Liewald DC, Luciano M, van der Lugt A, Martinez OO, Mayeux R, Mazoyer B, Nalls M, Nauck M, Niessen WJ, Oostra BA, Psaty BM, Rice KM, Rotter JI, von Sarnowski B, Schmidt H, Schreiner PJ, Schuur M, Sidney SS, Sigurdsson S, Slagboom PE, Stott DJ, van Swieten JC, Teumer A, Toglhofer AM, Traylor M, Trompet S, Turner ST, Tzourio C, Uh HW, Uitterlinden AG, Vernooij MW, Wang JJ, Wong TY, Wardlaw JM, Windham BG, Wittfeld K, Wolf C, Wright CB, Yang Q, Zhao W, Zijdenbos A, Jukema JW, Sacco RL, Kardia SL, Amouyel P, Mosley TH, Longstreth WT, Jr., DeCarli CC, van Duijn CM, Schmidt R, Launer LJ, Grabe HJ, Seshadri SS, Ikram MA, Fornage M. 2015. Multiethnic genome-wide association study of cerebral white matter hyperintensities on MRI. Circ Cardiovasc Genet 8:398-409.

184

Virgintino D, Girolamo F, Errede M, Capobianco C, Robertson D, Stallcup WB, Perris R, Roncali L. 2007. An intimate interplay between precocious, migrating pericytes and endothelial cells governs human fetal brain angiogenesis. Angiogenesis 10:35-45. Wakui S, Furusato M, Hasumura M, Hori M, Takahashi H, Kano Y, Ushigome S. 1989. Two- and three-dimensional ultrastructure of endothelium and pericyte interdigitations in capillary of human granulation tissue. J Electron Microsc (Tokyo) 38:136-142. Waldo KL, Hutson MR, Ward CC, Zdanowicz M, Stadt HA, Kumiski D, Abu-Issa R, Kirby ML. 2005. Secondary heart field contributes myocardium and smooth muscle to the arterial pole of the developing heart. Dev Biol 281:78-90. Wales MM, Biel MA, el Deiry W, Nelkin BD, Issa JP, Cavenee WK, Kuerbitz SJ, Baylin SB. 1995. p53 activates expression of HIC-1, a new candidate tumour suppressor gene on 17p13.3. Nat Med 1:570-577. Wang C, Low WC, Liu A, Wang B. 2013. Centrosomal protein DZIP1 regulates Hedgehog signaling by promoting cytoplasmic retention of transcription factor GLI3 and affecting ciliogenesis. J Biol Chem 288:29518-29529. Wang G, Zhang Z, Xu Z, Yin H, Bai L, Ma Z, Decoster MA, Qian G, Wu G. 2010. Activation of the sonic hedgehog signaling controls human pulmonary arterial smooth muscle cell proliferation in response to hypoxia. Biochim Biophys Acta 1803:1359-1367. Wang J, Milner R. 2006. Fibronectin promotes brain capillary endothelial cell survival and proliferation through alpha5beta1 and alphavbeta3 integrins via MAP kinase signalling. J Neurochem 96:148-159. Wang QS, Kong PZ, Li XQ, Yang F, Feng YM. 2015. FOXF2 deficiency promotes epithelial- mesenchymal transition and metastasis of basal-like breast cancer. Breast Cancer Res 17:30. Wang T, Baron M, Trump D. 2008. An overview of Notch3 function in vascular smooth muscle cells. Prog Biophys Mol Biol 96:499-509. Wang T, Tamakoshi T, Uezato T, Shu F, Kanzaki-Kato N, Fu Y, Koseki H, Yoshida N, Sugiyama T, Miura N. 2003a. Forkhead transcription factor Foxf2 (LUN)-deficient mice exhibit abnormal development of secondary palate. Dev Biol 259:83-94. Wang W, Prince CZ, Hu X, Pollman MJ. 2003b. HRT1 modulates vascular smooth muscle cell proliferation and apoptosis. Biochem Biophys Res Commun 308:596-601. Wang W, Prince CZ, Mou Y, Pollman MJ. 2002a. Notch3 signaling in vascular smooth muscle cells induces c-FLIP expression via ERK/MAPK activation. Resistance to Fas ligand-induced apoptosis. J Biol Chem 277:21723-21729. Wang X, Adhikari N, Li Q, Hall JL. 2004. LDL receptor-related protein LRP6 regulates proliferation and survival through the Wnt cascade in vascular smooth muscle cells. Am J Physiol Heart Circ Physiol 287:H2376-2383. Wang X, Xiao Y, Mou Y, Zhao Y, Blankesteijn WM, Hall JL. 2002b. A role for the beta-catenin/T- cell factor signaling cascade in vascular remodeling. Circ Res 90:340-347. Wang Y, Pan L, Moens CB, Appel B. 2014. Notch3 establishes brain vascular integrity by regulating pericyte number. Development 141:307-317. Weed M, Mundlos S, Olsen BR. 1997. The role of sonic hedgehog in vertebrate development. Matrix Biol 16:53-58. Weigel D, Jurgens G, Kuttner F, Seifert E, Jackle H. 1989. The homeotic gene fork head encodes a nuclear protein and is expressed in the terminal regions of the Drosophila embryo. Cell 57:645- 658. Weisiger RA, Zucker SD. 2002. Transfer of fatty acids between intracellular membranes: roles of soluble binding proteins, distance, and time. Am J Physiol Gastrointest Liver Physiol 282:G105-115. 185

Weisschuh N, Wolf C, Wissinger B, Gramer E. 2008. A novel mutation in the FOXC1 gene in a family with Axenfeld-Rieger syndrome and Peters' anomaly. Clin Genet 74:476-480. Westerfield M. 2007. The Zebrafish Book. A Guide for the Laboratory Use of Zebrafish (Danio rerio), 5th Edition. Eugene: University of Oregon Press. Whitesell TR, Kennedy RM, Carter AD, Rollins EL, Georgijevic S, Santoro MM, Childs SJ. 2014. An alpha-smooth muscle actin (acta2/alphasma) zebrafish transgenic line marking vascular mural cells and visceral smooth muscle cells. PLoS One 9:e90590. Wiegreffe C, Christ B, Huang R, Scaal M. 2009. Remodeling of aortic smooth muscle during avian embryonic development. Dev Dyn 238:624-631. Willert M, Augstein A, Poitz DM, Schmeisser A, Strasser RH, Braun-Dullaeus RC. 2010. Transcriptional regulation of Pim-1 kinase in vascular smooth muscle cells and its role for proliferation. Basic Res Cardiol 105:267-277. Wilm B, Ipenberg A, Hastie ND, Burch JB, Bader DM. 2005. The serosal mesothelium is a major source of smooth muscle cells of the gut vasculature. Development 132:5317-5328. Wilson BD, Ii M, Park KW, Suli A, Sorensen LK, Larrieu-Lahargue F, Urness LD, Suh W, Asai J, Kock GA, Thorne T, Silver M, Thomas KR, Chien CB, Losordo DW, Li DY. 2006. Netrins promote developmental and therapeutic angiogenesis. Science 313:640-644. Wilson CW, Stainier DY. 2010. Vertebrate Hedgehog signaling: cilia rule. BMC Biol 8:102. Wolff C, Roy S, Lewis KE, Schauerte H, Joerg-Rauch G, Kirn A, Weiler C, Geisler R, Haffter P, Ingham PW. 2004. iguana encodes a novel zinc-finger protein with coiled-coil domains essential for Hedgehog signal transduction in the zebrafish embryo. Genes Dev 18:1565-1576. Wu Q, Zhang J, Koh W, Yu Q, Zhu X, Amsterdam A, Davis GE, Arnaout MA, Xiong JW. 2015. Talin1 is required for cardiac Z-disk stabilization and endothelial integrity in zebrafish. Faseb J 29:4989-5005. Wu SP, Dong XR, Regan JN, Su C, Majesky MW. 2013. Tbx18 regulates development of the epicardium and coronary vessels. Dev Biol 383:307-320. Xia Y, Bhattacharyya A, Roszell EE, Sandig M, Mequanint K. 2012. The role of endothelial cell- bound Jagged1 in Notch3-induced human coronary artery smooth muscle cell differentiation. Biomaterials 33:2462-2472. Xiao Q, Zeng L, Zhang Z, Hu Y, Xu Q. 2007. Stem cell-derived Sca-1+ progenitors differentiate into smooth muscle cells, which is mediated by collagen IV-integrin alpha1/beta1/alphav and PDGF receptor pathways. Am J Physiol Cell Physiol 292:C342-352. Xu J, Liu H, Lan Y, Aronow BJ, Kalinichenko VV, Jiang R. 2016. A Shh-Foxf-Fgf18-Shh Molecular Circuit Regulating Palate Development. PLoS Genet 12:e1005769. Xu J, Srinivas BP, Tay SY, Mak A, Yu X, Lee SG, Yang H, Govindarajan KR, Leong B, Bourque G, Mathavan S, Roy S. 2006. Genomewide expression profiling in the zebrafish embryo identifies target genes regulated by Hedgehog signaling during vertebrate development. Genetics 174:735-752. Yamamoto H, Ehling M, Kato K, Kanai K, van Lessen M, Frye M, Zeuschner D, Nakayama M, Vestweber D, Adams RH. 2015. Integrin beta1 controls VE-cadherin localization and blood vessel stability. Nat Commun 6:6429. Yamashita S, Andoh M, Ueno-Kudoh H, Sato T, Miyaki S, Asahara H. 2009. Sox9 directly promotes Bapx1 gene expression to repress Runx2 in chondrocytes. Exp Cell Res 315:2231-2240. Yang K, Banerjee S, Proweller A. 2013. Regulation of pre-natal circle of Willis assembly by vascular smooth muscle Notch signaling. Dev Biol 381:107-120. Yang L, Rastegar S, Strahle U. 2010. Regulatory interactions specifying Kolmer-Agduhr interneurons. Development 137:2713-2722.

186

Yang L, Yatomi Y, Miura Y, Satoh K, Ozaki Y. 1999. Metabolism and functional effects of sphingolipids in blood cells. Br J Haematol 107:282-293. Yao Q, Renault MA, Chapouly C, Vandierdonck S, Belloc I, Jaspard-Vinassa B, Daniel-Lamaziere JM, Laffargue M, Merched A, Desgranges C, Gadeau AP. 2014. Sonic hedgehog mediates a novel pathway of PDGF-BB-dependent vessel maturation. Blood 123:2429-2437. Yatomi Y, Igarashi Y, Yang L, Hisano N, Qi R, Asazuma N, Satoh K, Ozaki Y, Kume S. 1997. Sphingosine 1-phosphate, a bioactive sphingolipid abundantly stored in platelets, is a normal constituent of human plasma and serum. J Biochem 121:969-973. Yemisci M, Gursoy-Ozdemir Y, Vural A, Can A, Topalkara K, Dalkara T. 2009. Pericyte contraction induced by oxidative-nitrative stress impairs capillary reflow despite successful opening of an occluded cerebral artery. Nat Med 15:1031-1037. Yin J, Shine L, Raycroft F, Deeti S, Reynolds A, Ackerman KM, Glaviano A, O'Farrell S, O'Leary O, Kilty C, Kennedy C, McLoughlin S, Rice M, Russell E, Higgins DG, Hyde DR, Kennedy BN. 2012. Inhibition of the Pim1 oncogene results in diminished visual function. PLoS One 7:e52177. Zaghloul NA, Brugmann SA. 2011. The emerging face of primary cilia. Genesis 49:231-246. Zaucker A, Mercurio S, Sternheim N, Talbot WS, Marlow FL. 2013. notch3 is essential for oligodendrocyte development and vascular integrity in zebrafish. Dis Model Mech 6:1246- 1259. Zeng L, Kempf H, Murtaugh LC, Sato ME, Lassar AB. 2002. Shh establishes an Nkx3.2/Sox9 autoregulatory loop that is maintained by BMP signals to induce somitic chondrogenesis. Genes Dev 16:1990-2005. Zeng Q, Wei B, Zhao Y, Wang X, Fu Q, Liu H, Li F. 2016. Shh mediates PDGF-induced contractile- to-synthetic phenotypic modulation in vascular smooth muscle cells through regulation of KLF4. Exp Cell Res 345:82-92. Zhang J, Liu Y, Zhang J, Cui X, Li G, Wang J, Ren H, Zhang Y. 2016a. FOXQ1 promotes gastric cancer metastasis through upregulation of Snail. Oncol Rep 35:3607-3613. Zhang Y, Chen K, Sloan SA, Bennett ML, Scholze AR, O'Keeffe S, Phatnani HP, Guarnieri P, Caneda C, Ruderisch N, Deng S, Liddelow SA, Zhang C, Daneman R, Maniatis T, Barres BA, Wu JQ. 2014. An RNA-sequencing transcriptome and splicing database of glia, neurons, and vascular cells of the cerebral cortex. J Neurosci 34:11929-11947. Zhang Y, Wang X, Wang Z, Tang H, Fan H, Guo Q. 2015. miR-182 promotes cell growth and invasion by targeting forkhead box F2 transcription factor in colorectal cancer. Oncol Rep 33:2592-2598. Zhang YN, Xie BD, Sun L, Chen W, Jiang SL, Liu W, Bian F, Tian H, Li RK. 2016b. Phenotypic switching of vascular smooth muscle cells in the 'normal region' of aorta from atherosclerosis patients is regulated by miR-145. J Cell Mol Med 20:1049-1061. Zheng W, Wang Z, Collins JE, Andrews RM, Stemple D, Gong Z. 2011. Comparative transcriptome analyses indicate molecular homology of zebrafish swimbladder and mammalian lung. PLoS One 6:e24019. Zhong TP, Childs S, Leu JP, Fishman MC. 2001. Gridlock signalling pathway fashions the first embryonic artery. Nature 414:216-220. Zimmerman K. 1923. Der Peinere Bau der Blutcapillaren. Zeitschrift fur Anatomie und Entwicklungsgeschichte 68:29-102. Zou C, Levine EM. 2012. Vsx2 controls eye organogenesis and retinal progenitor identity via homeodomain and non-homeodomain residues required for high affinity DNA binding. PLoS Genet 8:e1002924.

187

Chapter Eight: Appendix

188

JOHN WILEY AND SONS LICENSE TERMS AND CONDITIONS

This Agreement between Corey Arnold ("You") and John Wiley and Sons ("John Wiley and Sons") consists of your license details and the terms and conditions provided by John Wiley and Sons and Copyright Clearance Center.

License Number 3912110705268

License date Jul 18, 2016

Licensed Content Publisher John Wiley and Sons

Licensed Content Publication Developmental Dynamics

Licensed Content Title Comparative analysis of genes regulated by Dzip1/iguana and hedgehog in zebrafish

Licensed Content Author Corey R. Arnold,Ryan E. Lamont,John T. Walker,Peter J. Spice,Chi•Kin Chan,Chi•Yip Ho,Sarah J. Childs

Licensed Content Date Jan 8, 2015

Licensed Content Pages 13

Type of use Dissertation/Thesis

Requestor type Author of this Wiley article

Format Print and electronic

Portion Full article

Will you be translating? No

Title of your thesis / Determining genetic mechanisms of vascular stability: A novel role dissertation for FoxF2

Expected completion date Jul 2016

Expected size (number of 150 pages)

Requestor Location Corey Arnold Canada Attn: Corey Arnold

Publisher Tax ID EU826007151

Billing Type Invoice

Billing Address Corey Arnold Canada Attn: Corey Arnold

Total 0.00 CAD

189

Terms and Conditions

TERMS AND CONDITIONS

This copyrighted material is owned by or exclusively licensed to John Wiley & Sons, Inc. or one of its group companies (each a"Wiley Company") or handled on behalf of a society with which a Wiley Company has exclusive publishing rights in relation to a particular work (collectively "WILEY"). By clicking "accept" in connection with completing this licensing transaction, you agree that the following terms and conditions apply to this transaction (along with the billing and payment terms and conditions established by the Copyright Clearance Center Inc., ("CCC's Billing and Payment terms and conditions"), at the time that you opened your RightsLink account (these are available at any time at http://myaccount.copyright.com).

Terms and Conditions

 The materials you have requested permission to reproduce or reuse (the "Wiley Materials") are protected by copyright.

 You are hereby granted a personal, non•exclusive, non•sub licensable (on a stand• alone basis), non•transferable, worldwide, limited license to reproduce the Wiley Materials for the purpose specified in the licensing process. This license, and any CONTENT (PDF or image file) purchased as part of your order, is for a one•time use only and limited to any maximum distribution number specified in the license. The first instance of republication or reuse granted by this license must be completed within two years of the date of the grant of this license (although copies prepared before the end date may be distributed thereafter). The Wiley Materials shall not be used in any other manner or for any other purpose, beyond what is granted in the license. Permission is granted subject to an appropriate acknowledgement given to the author, title of the material/book/journal and the publisher. You shall also duplicate the copyright notice that appears in the Wiley publication in your use of the Wiley Material. Permission is also granted on the understanding that nowhere in the text is a previously published source acknowledged for all or part of this Wiley Material. Any third party content is expressly excluded from this permission.

 With respect to the Wiley Materials, all rights are reserved. Except as expressly granted by the terms of the license, no part of the Wiley Materials may be copied, modified, adapted (except for minor reformatting required by the new Publication), translated, reproduced, transferred or distributed, in any form or by any means, and no derivative works may be made based on the Wiley Materials without the prior permission of the respective copyright owner. For STM Signatory Publishers clearing permission under the terms of the STM Permissions Guidelines only, the terms of the license are extended to include subsequent editions and for editions in other languages, provided such editions are for the work as a whole in situ and does not involve the separate exploitation of the permitted figures or extracts, You may not alter, remove or suppress in any manner any copyright, trademark or other notices displayed by the Wiley Materials. You may not license, rent, sell, loan, lease, pledge, offer as security, transfer or assign the Wiley Materials on a stand•alone basis, or any of the rights granted to you hereunder to any other person.

 The Wiley Materials and all of the intellectual property rights therein shall at all times remain the exclusive property of John Wiley & Sons Inc, the Wiley Companies, or their respective licensors, and your interest therein is only that of having possession of and the right to reproduce the Wiley Materials pursuant to Section 2 herein during the continuance of this Agreement. You agree that you 190

own no right, title or interest in or to the Wiley Materials or any of the intellectual property rights therein. You shall have no rights hereunder other than the license as provided for above in Section 2. No right, license or interest to any trademark, trade name, service mark or other branding

("Marks") of WILEY or its licensors is granted hereunder, and you agree that you shall not assert any such right, license or interest with respect thereto

 NEITHER WILEY NOR ITS LICENSORS MAKES ANY WARRANTY OR REPRESENTATION OF ANY KIND TO YOU OR ANY THIRD PARTY, EXPRESS, IMPLIED OR STATUTORY, WITH RESPECT TO THE MATERIALS OR THE ACCURACY OF ANY INFORMATION CONTAINED IN THE MATERIALS, INCLUDING, WITHOUT LIMITATION, ANY IMPLIED WARRANTY OF MERCHANTABILITY, ACCURACY, SATISFACTORY QUALITY, FITNESS FOR A PARTICULAR PURPOSE, USABILITY, INTEGRATION OR NON•INFRINGEMENT AND ALL SUCH WARRANTIES ARE HEREBY EXCLUDED BY WILEY AND ITS LICENSORS AND WAIVED BY YOU.

 WILEY shall have the right to terminate this Agreement immediately upon breach of this Agreement by you.

 You shall indemnify, defend and hold harmless WILEY, its Licensors and their respective directors, officers, agents and employees, from and against any actual or threatened claims, demands, causes of action or proceedings arising from any breach of this Agreement by you.

 IN NO EVENT SHALL WILEY OR ITS LICENSORS BE LIABLE TO YOU OR ANY OTHER PARTY OR ANY OTHER PERSON OR ENTITY FOR ANY SPECIAL, CONSEQUENTIAL, INCIDENTAL, INDIRECT, EXEMPLARY OR PUNITIVE DAMAGES, HOWEVER CAUSED, ARISING OUT OF OR IN CONNECTION WITH THE DOWNLOADING, PROVISIONING, VIEWING OR USE OF THE MATERIALS REGARDLESS OF THE FORM OF ACTION, WHETHER FOR BREACH OF CONTRACT, BREACH OF WARRANTY, TORT, NEGLIGENCE, INFRINGEMENT OR OTHERWISE (INCLUDING, WITHOUT LIMITATION, DAMAGES BASED ON LOSS OF PROFITS, DATA, FILES, USE, BUSINESS OPPORTUNITY OR CLAIMS OF THIRD PARTIES), AND WHETHER OR NOT THE PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. THIS LIMITATION SHALL APPLY NOTWITHSTANDING ANY FAILURE OF ESSENTIAL PURPOSE OF ANY LIMITED REMEDY PROVIDED HEREIN.

 Should any provision of this Agreement be held by a court of competent jurisdiction to be illegal, invalid, or unenforceable, that provision shall be deemed amended to achieve as nearly as possible the same economic effect as the original provision, and the legality, validity and enforceability of the remaining provisions of this Agreement shall not be affected or impaired thereby.

 The failure of either party to enforce any term or condition of this Agreement shall not constitute a waiver of either party's right to enforce each and every term and condition of this Agreement. No breach under this agreement shall be deemed waived or excused by either party unless such waiver or consent is in writing signed by the party granting such waiver or consent. The waiver by or consent of a party to a breach of any provision of this Agreement shall not operate or be construed as a waiver of or consent to any other or subsequent breach by such other party.

191

 This Agreement may not be assigned (including by operation of law or otherwise) by you without WILEY's prior written consent.

 Any fee required for this permission shall be non•refundable after thirty (30) days from receipt by the CCC.

 These terms and conditions together with CCC's Billing and Payment terms and conditions (which are incorporated herein) form the entire agreement between you and WILEY concerning this licensing transaction and (in the absence of fraud) supersedes all prior agreements and representations of the parties, oral or written. This Agreement may not be amended except in writing signed by both parties. This Agreement shall be binding upon and inure to the benefit of the parties' successors, legal representatives, and authorized assigns.

 In the event of any conflict between your obligations established by these terms and conditions and those established by CCC's Billing and Payment terms and conditions, these terms and conditions shall prevail.

 WILEY expressly reserves all rights not specifically granted in the combination of (i) the license details provided by you and accepted in the course of this licensing transaction, (ii) these terms and conditions and (iii) CCC's Billing and Payment terms and conditions.

 This Agreement will be void if the Type of Use, Format, Circulation, or Requestor Type was misrepresented during the licensing process.

 This Agreement shall be governed by and construed in accordance with the laws of the State of New York, USA, without regards to such state's conflict of law rules. Any legal action, suit or proceeding arising out of or relating to these Terms and Conditions or the breach thereof shall be instituted in a court of competent jurisdiction in New York County in the State of New York in the United States of America and each party hereby consents and submits to the personal jurisdiction of such court, waives any objection to venue in such court and consents to service of process by registered or certified mail, return receipt requested, at the last known address of such party.

WILEY OPEN ACCESS TERMS AND CONDITIONS Wiley Publishes Open Access Articles in fully Open Access Journals and in Subscription journals offering Online Open. Although most of the fully Open Access journals publish open access articles under the terms of the Creative Commons Attribution (CC BY) License only, the subscription journals and a few of the Open Access Journals offer a choice of Creative Commons Licenses. The license type is clearly identified on the article. The Creative Commons Attribution License The Creative Commons Attribution License (CC•BY) allows users to copy, distribute and transmit an article, adapt the article and make commercial use of the article. The CC•BY license permits commercial and non• Creative Commons Attribution Non•Commercial License The Creative Commons Attribution Non•Commercial (CC•BY•NC) License permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes.(see below)

Creative Commons Attribution•Non•Commercial•NoDerivs License 192

The Creative Commons Attribution Non•Commercial•NoDerivs License (CC•BY•NC•ND) permits use, distribution and reproduction in any medium, provided the original work is properly cited, is not used for commercial purposes and no modifications or adaptations are made. (see below) Use by commercial "for•profit" organizations Use of Wiley Open Access articles for commercial, promotional, or marketing purposes requires further explicit permission from Wiley and will be subject to a fee. Further details can be found on Wiley Online Library http://olabout.wiley.com/WileyCDA/Section/id•410895.html

Other Terms and Conditions:

v1.10 Last updated September 2015

193

NATURE PUBLISHING GROUP LICENSE TERMS AND CONDITIONS

This Agreement between Corey Arnold ("You") and Nature Publishing Group ("Nature Publishing Group") consists of your license details and the terms and conditions provided by Nature Publishing Group and Copyright Clearance Center.

License Number 3886141296283

License date Jun 11, 2016

Licensed Content Publisher Nature Publishing Group

Licensed Content Publication Nature Medicine

Licensed Content Title Molecular regulation of vessel maturation

Licensed Content Author Rakesh K Jain

Licensed Content Date Jun 1, 2003

Licensed Content Volume 9 Number

Licensed Content Issue 6 Number

Type of Use reuse in a dissertation / thesis

Requestor type academic/educational

Format print and electronic

Portion figures/tables/illustrations

Number of 1 figures/tables/illustrations

High•res required no

Figures Figure 1

Author of this NPG article no

Your reference number

Title of your thesis / Determining genetic mechanisms of vascular stability: A novel role dissertation for FoxF2

Expected completion date Jul 2016

Estimated size (number of 150 pages)

Requestor Location Corey Arnold Canada Attn: Corey Arnold

Billing Type Invoice

194

Billing Address Corey Arnold Canada Attn: Corey Arnold

Total 0.00 USD

Terms and Conditions

Terms and Conditions for Permissions Nature Publishing Group hereby grants you a non•exclusive license to reproduce this material for this purpose, and for no other use, subject to the conditions below:

1. NPG warrants that it has, to the best of its knowledge, the rights to license reuse of this material. However, you should ensure that the material you are requesting is original to Nature Publishing Group and does not carry the copyright of another entity (as credited in the published version). If the credit line on any part of the material you have requested indicates that it was reprinted or adapted by NPG with permission from another source, then you should also seek permission from that source to reuse the material.

2. Permission granted free of charge for material in print is also usually granted for any electronic version of that work, provided that the material is incidental to the work as a whole and that the electronic version is essentially equivalent to, or substitutes for, the print version. Where print permission has been granted for a fee, separate permission must be obtained for any additional, electronic re•use (unless, as in the case of a full paper, this has already been accounted for during your initial request in the calculation of a print run).NB: In all cases, web•based use of full•text articles must be authorized separately through the 'Use on a Web Site' option when requesting permission.

3. Permission granted for a first edition does not apply to second and subsequent editions and for editions in other languages (except for signatories to the STM Permissions Guidelines, or where the first edition permission was granted for free).

4. Nature Publishing Group's permission must be acknowledged next to the figure, table or abstract in print. In electronic form, this acknowledgement must be visible at the same time as the figure/table/abstract, and must be hyperlinked to the journal's homepage.

5. The credit line should read: Reprinted by permission from Macmillan Publishers Ltd: [JOURNAL NAME] (reference citation), copyright (year of publication) For AOP papers, the credit line should read: Reprinted by permission from Macmillan Publishers Ltd: [JOURNAL NAME], advance online publication, day month year (doi: 10.1038/sj.[JOURNAL ACRONYM].XXXXX)

Note: For republication from the British Journal of Cancer, the following credit lines apply. Reprinted by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL NAME] (reference citation), copyright (year of publication) For AOP papers, the credit line should read: Reprinted by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL NAME], advance online publication, day month year (doi: 10.1038/sj.[JOURNAL ACRONYM].XXXXX)

6. Adaptations of single figures do not require NPG approval. However, the adaptation should be credited as follows:

Adapted by permission from Macmillan Publishers Ltd: [JOURNAL NAME] (reference citation), 195

Note: For adaptation from the British Journal of Cancer, the following credit line applies.

Adapted by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL NAME] (reference citation), copyright (year of publication)

7. Translations of 401 words up to a whole article require NPG approval. Please visit http://www.macmillanmedicalcommunications.com for more information. Translations of up to a 400 words do not require NPG approval. The translation should be credited as follows:

Translated by permission from Macmillan Publishers Ltd: [JOURNAL NAME] (reference citation), copyright (year of publication).

Note: For translation from the British Journal of Cancer, the following credit line applies.

Translated by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL NAME] (reference citation), copyright (year of publication)

We are certain that all parties will benefit from this agreement and wish you the best in the use of this material. Thank you. Special Terms: v1.1

196

ELSEVIER LICENSE TERMS AND CONDITIONS

This Agreement between Corey Arnold ("You") and Elsevier ("Elsevier") consists of your license details and the terms and conditions provided by Elsevier and Copyright Clearance Center.

License Number 3886180646595

License date Jun 11, 2016

Licensed Content Publisher Elsevier

Licensed Content Publication Developmental Biology

Licensed Content Title The Vascular Anatomy of the Developing Zebrafish: An Atlas of Embryonic and Early Larval Development

Licensed Content Author Sumio Isogai,Masaharu Horiguchi,Brant M. Weinstein

Licensed Content Date 15 February 2001

Licensed Content Volume 230 Number

Licensed Content Issue 2 Number

Licensed Content Pages 24

Start Page 278

End Page 301

Type of Use reuse in a thesis/dissertation

Intended publisher of new other work

Portion figures/tables/illustrations

Number of 5 figures/tables/illustrations

Format both print and electronic

Are you the author of this No Elsevier article?

Will you be translating? No

Order reference number

Original figure numbers Figure 1, 2, 3, 5, 7

Title of your Determining genetic mechanisms of vascular stability: A novel thesis/dissertation role for FoxF2

Expected completion date Jul 2016

Estimated size (number of 150 pages)

197

Elsevier VAT number GB 494 6272 12

Requestor Location Corey Arnold Canada Attn: Corey Arnold

Total 0.00 CAD

Terms and Conditions INTRODUCTION 1. The publisher for this copyrighted material is Elsevier. By clicking "accept" in connection with completing this licensing transaction, you agree that the following terms and conditions apply to this transaction (along with the Billing and Payment terms and conditions established by Copyright Clearance Center, Inc. ("CCC"), at the time that you opened your Rightslink account and that are available at any time at http://myaccount.copyright.com). GENERAL TERMS 2. Elsevier hereby grants you permission to reproduce the aforementioned material subject to the terms and conditions indicated. 3. Acknowledgement: If any part of the material to be used (for example, figures) has appeared in our publication with credit or acknowledgement to another source, permission must also be sought from that source. If such permission is not obtained then that material may not be included in your publication/copies. Suitable acknowledgement to the source must be made, either as a footnote or in a reference list at the end of your publication, as follows: "Reprinted from Publication title, Vol /edition number, Author(s), Title of article / title of chapter, Pages No., Copyright (Year), with permission from Elsevier [OR APPLICABLE SOCIETY COPYRIGHT OWNER]." Also Lancet special credit • "Reprinted from The Lancet, Vol. number, Author(s), Title of article, Pages No., Copyright (Year), with permission from Elsevier." 4. Reproduction of this material is confined to the purpose and/or media for which permission is hereby given. 5. Altering/Modifying Material: Not Permitted. However figures and illustrations may be altered/adapted minimally to serve your work. Any other abbreviations, additions, deletions and/or any other alterations shall be made only with prior written authorization of Elsevier Ltd. (Please contact Elsevier at [email protected]) 6. If the permission fee for the requested use of our material is waived in this instance, please be advised that your future requests for Elsevier materials may attract a fee. 7. Reservation of Rights: Publisher reserves all rights not specifically granted in the combination of (i) the license details provided by you and accepted in the course of this licensing transaction, (ii) these terms and conditions and (iii) CCC's Billing and Payment terms and conditions. 8. License Contingent Upon Payment: While you may exercise the rights licensed immediately upon issuance of the license at the end of the licensing process for the transaction, provided that you have disclosed complete and accurate details of your proposed use, no license is finally effective unless and until full payment is received from you (either by publisher or by CCC) as provided in CCC's Billing and Payment terms and conditions. If full payment is not received on a timely basis, then any license preliminarily granted shall be deemed automatically revoked and shall be void as if never granted. Further, in the event that you breach any of these terms and conditions or any of CCC's Billing and Payment terms and conditions, the license is automatically revoked and shall be void as if never granted. Use of materials as described in a revoked license, as well as any use of the materials beyond the scope of an unrevoked license, may constitute copyright infringement and publisher reserves the right to take any and all action to protect its copyright in the materials. 198

9. Warranties: Publisher makes no representations or warranties with respect to the licensed material. 10. Indemnity: You hereby indemnify and agree to hold harmless publisher and CCC, and their respective officers, directors, employees and agents, from and against any and all claims arising out of your use of the licensed material other than as specifically authorized pursuant to this license. 11. No Transfer of License: This license is personal to you and may not be sublicensed, assigned, or transferred by you to any other person without publisher's written permission. 12. No Amendment Except in Writing: This license may not be amended except in a writing signed by both parties (or, in the case of publisher, by CCC on publisher's behalf). 13. Objection to Contrary Terms: Publisher hereby objects to any terms contained in any purchase order, acknowledgment, check endorsement or other writing prepared by you, which terms are inconsistent with these terms and conditions or CCC's Billing and Payment terms and conditions. These terms and conditions, together with CCC's Billing and Payment terms and conditions (which are incorporated herein), comprise the entire agreement between you and publisher (and CCC) concerning this licensing transaction. In the event of any conflict between your obligations established by these terms and conditions and those established by CCC's Billing and Payment terms and conditions, these terms and conditions shall control. 14. Revocation: Elsevier or Copyright Clearance Center may deny the permissions described in this License at their sole discretion, for any reason or no reason, with a full refund payable to you. Notice of such denial will be made using the contact information provided by you. Failure to receive such notice will not alter or invalidate the denial. In no event will Elsevier or Copyright Clearance Center be responsible or liable for any costs, expenses or damage incurred by you as a result of a denial of your permission request, other than a refund of the amount(s) paid by you to Elsevier and/or Copyright Clearance Center for denied permissions. LIMITED LICENSE The following terms and conditions apply only to specific license types: 15. Translation: This permission is granted for non•exclusive world English rights only unless your license was granted for translation rights. If you licensed translation rights you may only translate this content into the languages you requested. A professional translator must perform all translations and reproduce the content word for word preserving the integrity of the article. 16. Posting licensed content on any Website: The following terms and conditions apply as follows: Licensing material from an Elsevier journal: All content posted to the web site must maintain the copyright information line on the bottom of each image; A hypertext must be included to the Homepage of the journal from which you are licensing at http://www.sciencedirect.com/science/journal/xxxxx or the Elsevier homepage for books at http://www.elsevier.com; Central Storage: This license does not include permission for a scanned version of the material to be stored in a central repository such as that provided by Heron/XanEdu. Licensing material from an Elsevier book: A hyper•text link must be included to the Elsevier homepage at http://www.elsevier.com. All content posted to the web site must maintain the copyright information line on the bottom of each image.

reviewed, nor has it had any other value added to it by a publisher (such as formatting, copyright, technical enhancement etc.). Authors can share their preprints anywhere at any time. Preprints should not be added to or enhanced in any way in order to appear more like, or to substitute for, the final versions of articles however authors can update their preprints on arXiv or RePEc with their Accepted Author Manuscript (see below). If accepted for publication, we encourage authors to link from the preprint to their formal publication via its DOI. Millions of researchers have access to the formal publications on ScienceDirect, and so links will help users to find, access, cite and use the best available version. Please note that Cell Press, The Lancet and some society•owned have different preprint policies. Information on these policies is available on the journal homepage. Accepted Author Manuscripts: An accepted author manuscript is the manuscript of an article that has been accepted for publication and which typically includes author• incorporated changes suggested during submission, peer review and editor•author communications. Authors can share their accepted author manuscript:

 immediately  via their non•commercial person homepage or blog by updating a preprint in arXiv or RePEc with the accepted manuscript via their research institute or institutional repository for internal institutional uses or as part of an invitation•only research collaboration work•group directly by providing copies to their students or to research collaborators for their personal use for private scholarly sharing as part of an invitation•only work group on commercial sites with which Elsevier has an agreement  after the embargo period  via non•commercial hosting platforms such as their institutional repository via commercial sites with which Elsevier has an agreement

In all cases accepted manuscripts should:

 link to the formal publication via its DOI   bear a CC•BY•NC•ND license • this is easy to do   if aggregated with other manuscripts, for example in a repository or other site, be shared in alignment with our hosting policy not be added to or enhanced in any way to appear more like, or to substitute for, the published journal article. 

Subscription Articles: If you are an author, please share a link to your article rather than the full•text. Millions of researchers have access to the formal publications on ScienceDirect, and so links will help your users to find, access, cite, and use the best available version. Theses and dissertations which contain embedded PJAs as part of the formal submission can be posted publicly by the awarding institution with DOI links back to the formal publications on ScienceDirect. If you are affiliated with a library that subscribes to ScienceDirect you have additional private 200

sharing rights for others' research accessed under that agreement. This includes use for classroom teaching and internal training at the institution (including use in course packs and courseware programs), and inclusion of the article for grant funding purposes. Gold Open Access Articles: May be shared according to the author•selected end•user license and should contain a CrossMark logo, the end user license, and a DOI link to the formal publication on ScienceDirect. Please refer to Elsevier's posting policy for further information. 18. For book authors the following clauses are applicable in addition to the above: Authors are permitted to place a brief summary of their work online only. You are not allowed to download and post the published electronic version of your chapter, nor may you scan the printed edition to create an electronic version. Posting to a repository: Authors are permitted to post a summary of their chapter only in their institution's repository. 19. Thesis/Dissertation: If your license is for use in a thesis/dissertation your thesis may be submitted to your institution in either print or electronic form. Should your thesis be published commercially, please reapply for permission. These requirements include permission for the Library and Archives of Canada to supply single copies, on demand, of the complete thesis and include permission for Proquest/UMI to supply single copies, on demand, of the complete thesis. Should your thesis be published commercially, please reapply for permission. Theses and dissertations which contain embedded PJAs as part of the formal submission can be posted publicly by the awarding institution with DOI links back to the formal publications on ScienceDirect.

Elsevier Open Access Terms and Conditions You can publish open access with Elsevier in hundreds of open access journals or in nearly 2000 established subscription journals that support open access publishing. Permitted third party re•use of these open access articles is defined by the author's choice of Creative Commons user license. See our open access license policy for more information. Terms & Conditions applicable to all Open Access articles published with Elsevier: Any reuse of the article must not represent the author as endorsing the adaptation of the article nor should the article be modified in such a way as to damage the author's honour or reputation. If any changes have been made, such changes must be clearly indicated. The author(s) must be appropriately credited and we ask that you include the end user license and a DOI link to the formal publication on ScienceDirect. If any part of the material to be used (for example, figures) has appeared in our publication with credit or acknowledgement to another source it is the responsibility of the user to ensure their reuse complies with the terms and conditions determined by the rights holder. Additional Terms & Conditions applicable to each Creative Commons user license: CC BY: The CC•BY license allows users to copy, to create extracts, abstracts and new works from the Article, to alter and revise the Article and to make commercial use of the Article (including reuse and/or resale of the Article by commercial entities), provided the user gives appropriate credit (with a link to the formal publication through the relevant DOI), provides a link to the license, indicates if changes were made and the licensor is not represented as endorsing the use made of the work. The full details of the license are available at http://creativecommons.org/licenses/by/4.0. CC BY NC SA: The CC BY•NC•SA license allows users to copy, to create extracts, abstracts and new works from the Article, to alter and revise the Article, provided this is not done for commercial purposes, and that the user gives appropriate credit (with a link to the formal publication through the relevant DOI), provides a link to the license, indicates if changes were made and the licensor is not represented as endorsing the use made of the work. Further, any new works must be made available on the same conditions. The full details of the license are available at

201

http://creativecommons.org/licenses/by•nc•sa/4.0. CC BY NC ND: The CC BY•NC•ND license allows users to copy and distribute the Article, provided this is not done for commercial purposes and further does not permit distribution of the Article if it is changed or edited in any way, and provided the user gives appropriate credit (with a link to the formal publication through the relevant DOI), provides a link to the license, and that the licensor is not represented as endorsing the use made of the work. The full details of the license are available at http://creativecommons.org/licenses/by•nc•nd/4.0. Any commercial reuse of Open Access articles published with a CC BY NC SA or CC BY NC ND license requires permission from Elsevier and will be subject to a fee. Commercial reuse includes:

 Associating advertising with the full text of the Article   Charging fees for document delivery or access   Article aggregation   Systematic distribution via e•mail lists or share buttons 

Posting or linking by commercial companies for use by customers of those companies.

20. Other Conditions: v1.8

202

BMJ PUBLISHING GROUP LTD. LICENSE TERMS AND CONDITIONS

This Agreement between Corey Arnold ("You") and BMJ Publishing Group Ltd. ("BMJ Publishing Group Ltd.") consists of your license details and the terms and conditions provided by BMJ Publishing Group Ltd. and Copyright Clearance Center.

License Number 3886191251028

License date Jun 11, 2016

Licensed Content Publisher BMJ Publishing Group Ltd.

Licensed Content Publication The BMJ

Licensed Content Title The clinical importance of white matter hyperintensities on brain magnetic resonance imaging: systematic review and meta•analysis

Licensed Content Author Stéphanie Debette, H S Markus

Licensed Content Date Jul 26, 2010

Licensed Content Volume 341 Number

Volume number 341

Type of Use Dissertation/Thesis

Requestor type Individual

Format Print and electronic

Portion Figure/table/extract

Number of 1 figure/table/extracts

Descriptionof Figure 1 figure/table/extracts

Will you be translating? No

Circulation/distribution 3666

Title of your thesis / Determining genetic mechanisms of vascular stability: A novel dissertation role for FoxF2

Expected completion date Jul 2016

Estimated size(pages) 150

Requestor Location Corey Arnold Canada Attn: Corey Arnold

Publisher Tax ID GB674738491

Billing Type Invoice

Billing Address Corey Arnold Canada Attn: Corey Arnold 203

Total 0.00 CAD

Terms and Conditions

BMJ Group Terms and Conditions for Permissions

When you submit your order you are subject to the terms and conditions set out below. You will also have agreed to the Copyright Clearance Center's ("CCC") terms and conditions regarding billing and payment https://s100.copyright.com/App/PaymentTermsAndConditions.jsp. CCC are acting as the BMJ Publishing Group Limited's ("BMJ Group's") agent. Subject to the terms set outherein, the BMJ Group hereby grants to you (the Licensee) a non• exclusive, non•transferable licence to re•use material as detailed in your request for this/those purpose(s) only and in accordance with the following conditions:

1) Scope of Licence: Use of the Licensed Material(s) is restricted to the ways specified by you duringthe order process and any additional use(s) outside of those specified in that request, require a further grant of permission.

2) Acknowledgement: In all cases, due acknowledgement to the original publication with permission from the BMJ Group should be stated adjacent to the reproduced Licensed Material. The format of such acknowledgement should read as follows: "Reproduced from [publication title, author(s), volume number, page numbers, copyright notice year] with permission from BMJ Publishing Group Ltd."

3) Third Party Material: BMJ Group acknowledges to the best of its knowledge, it has the rights to licence your reuse of the Licensed Material, subject always to the caveat that images/diagrams, tables and other illustrative material included within, which have a separate copyright notice, are presumed as excluded from the licence. Therefore, you should ensure that the Licensed Material you are requesting is original to BMJ Group and does not carry the copyright of another entity (as credited in the published version). If the credit line on any part of the material you have requested in any way indicates that it was reprinted or adapted by BMJ Group with permission from another source, then you should seek permission from that source directly to re•use the Licensed Material, as this is outside of the licence granted herein.

4) Altering/Modifying Material: The text of any material for which a licence is granted may not be altered in any way without the prior express permission of the BMJ Group. Subject to Clause 3 above however, single figure adaptations do not require BMJ Group's approval; however, the adaptation should be credited as follows: "Adapted by permission from BMJ Publishing Group Limited. [publication title, author, volume number, page numbers, copyright notice year]

5) Reservation of Rights: The BMJ Group reserves all rights not specifically granted in the combination of (i) the licence details provided by you and accepted in the course of this licensing transaction, (ii) these terms and conditions and (iii) CCC's Billing and Payment Terms and Conditions.

6) Timing of Use: First use of the Licensed Material must take place within 12 months of the grant of permission. 7). Creation of Contract and Termination: Once you have submitted an order via Rightslink and this is received by CCC, and subject to you completing accurate details of your proposed use, this is 204

when a binding contract is in effect and our acceptance occurs. As you are ordering rights from a periodical, to the fullest extent permitted by law, you will have no right to cancel the contract from this point other than for BMJ Group's material breach or fraudulent misrepresentation or as otherwise permitted under a statutory right. Payment must be made in accordance with CCC's Billing and Payment Terms and conditions. In the event that you breach any material condition of these terms and condition or any of CCC's Billing and Payment Terms and Conditions, the license is automatically terminated upon written notice from the BMJ Group or CCC or as otherwise provided for in CCC's Billing and Payment Terms and Conditions, where these apply.. Continued use of materials where a licence has been terminated, as well as any use of the Licensed Materials beyond the scope of an unrevoked licence, may constitute intellectual property rights infringement and BMJ Group reserves the right to take any and all action to protect its intellectual property rights in the Licensed Materials.

8. Warranties: BMJ Group makes no express or implied representations or warranties with respect to the Licensed Material and to the fullest extent permitted by law this is provided on an "as is" basis. For the avoidance of doubt BMJ Group does not warrant that the Licensed Material is accurate or fit for any particular purpose. 9. Limitation of Liability: To the fullest extent permitted by law, the BMJ Group disclaims all liability for any indirect, consequential or incidental damages (including without limitation, damages for loss of profits, information or interruption)arising out of the use or inability to use the Licensed Material or the inability to obtain additional rights to use the Licensed Material. To the fullest extent permitted by law, the maximum aggregate liability of the BMJ Group for any claims, costs, proceedings and demands for direct losses caused by BMJ Group's breaches of its obligations herein shall be limited to twice the amount paid by you to CCC for the licence granted herein. 10. Indemnity: You hereby indemnify and hold harmless the BMJ Group and their respective officers, directors, employees and agents, from and against any and all claims, costs, proceeding or demands arising out of your unauthorised use of the Licensed Material.

13. Objection to Contraryterms: BMJ Group hereby objects to any terms contained in any purchase order, acknowledgment, check endorsement or other writing prepared by you, which terms are inconsistent with these terms and conditions or CCC's Billing and Payment Terms and Conditions. These terms and conditions, together with CCC's Billing and Payment Terms and Conditions (which to the extent they are consistent are incorporated herein), comprise the entire agreement between you and BMJ Group (and CCC) and the Licensee concerning this licensing transaction. In the event of any conflict between your obligations established by these terms and conditions and those established by CCC's Billing and Payment Terms and Conditions, these terms and conditions shall control.

14. Revocation: BMJ Group or CCC may, within 30 days of issuance of this licence, deny the permissions described in this licence at their sole discretion, for any reason or no reason, with a full refund payable to you should you have not been able to exercise your rights in full. Notice of such 205

denial will be made using the contact information provided by you. Failure to receive such notice from BMJ Group or CCC will not, to the fullest extent permitted by law alter or invalidate the denial. For the fullest extent permitted by law in no event will BMJ Group or CCC be responsible or liable for any costs, expenses or damage incurred by you as a result of a denial of your permission request, other than a refund of the amount(s) paid by you to BMJ Group and/or CCC for denied permissions.

15. Restrictions to the license:

15.1 Promotion: BMJ Group will not give permission to reproduce in full or in part any Licensed Material for use in the promotion of the following: a) non•medical products that are harmful or potentially harmful to health: alcohol, baby milks and/or, sunbeds b) medical products that do not have a product license granted by the Medicines and Healthcare products Regulatory Agency (MHRA) or its international equivalents. Marketing of the product may start only after data sheets have been released to members of the medical profession and must conform to the marketing authorization contained in the product license. 16. Translation: This permission is granted for non•exclusive world English language rights only unless explicitly stated in your licence. If translation rights are granted, a professional translator should be employed and the content should be reproduced word for word preserving the integrity of the content. 17. General: Neither party shall be liable for failure, default or delay in performing its obligations under this Licence, caused by a Force Majeure event which shall include any act of God, war, or threatened war, act or threatened act of terrorism, riot, strike, lockout, individual action, fire, flood, drought, tempest or other event beyond the reasonable control of either party.

17.1 In the event that any provision of this Agreement is held to be invalid, the remainder of the provisions shall continue in full force and effect.

17.2 There shall be no right whatsoever for any third party to enforce the terms and conditions of this Agreement. The Parties hereby expressly wish to exclude the operation of the Contracts (Rights of Third Parties) Act 1999 and any other legislation which has this effect and is binding on this agreement.

17.3 To the fullest extent permitted by law, this Licence will be governed by the laws of England and shall be governed and construed in accordance with the laws of England. Any action arising out of or relating to this agreement shall be brought in courts situated in England save where it is necessary for BMJ Group for enforcement to bring proceedings to bring an action in an alternative jurisdiction.

206

BMJ PUBLISHING GROUP LTD. LICENSE TERMS AND CONDITIONS

License Number 3886200167019

License date Jun 11, 2016

Licensed Content Publisher BMJ Publishing Group Ltd.

Licensed Content Publication Practical Neurology

Licensed Content Title Lacunar infarcts: no black holes in the brain are benign

Licensed Content Author B Norrving

Licensed Content Date Aug 1, 2008

Licensed Content Volume 8 Number

Licensed Content Issue 4 Number

Volume number 8

Issue number 4

Type of Use Dissertation/Thesis

Requestor type Individual

Format Print and electronic

Portion Figure/table/extract

Number of 1 figure/table/extracts

Descriptionof Figure 2 figure/table/extracts

Will you be translating? No

Circulation/distribution 6

Title of your thesis / Determining genetic mechanisms of vascular stability: A novel dissertation role for FoxF2

Expected completion date Jul 2016

Estimated size(pages) 150

Requestor Location Corey Arnold Canada Attn: Corey Arnold

Publisher Tax ID GB674738491 Billing Type Invoice

207

Billing Address Corey Arnold Canada Attn: Corey Arnold

Total 0.00 CAD

Terms and Conditions

BMJ Group Terms and Conditions for Permissions

1) Scope of Licence: Use of the Licensed Material(s) is restricted to the ways specified by you during the order process and any additional use(s) outside of those specified in that request, require a further grant of permission.

4) Altering/Modifying Material: The text of any material for which a licence is granted may not be altered in any way without the prior express permission of the BMJ Group. Subject to Clause 3 above however, single figure adaptations do not require BMJGroup's approval; however, the adaptation should be credited as follows: "Adapted by permission from BMJ Publishing Group Limited. [publication title, author, volume number, page numbers, copyright notice year]

6) Timing of Use: First use of the Licensed Material must take place within 12 months of the grant of permission. 208

7) Creation of Contract and Termination: Once you have submitted an order via Rightslink and this is received by CCC, and subject to you completing accurate details of your proposed use, this is when a binding contract is in effect and our acceptance occurs. As you are ordering rights from a periodical, to the fullest extent permitted by law, you will have no right to cancel the contract from this point other than for BMJ Group's material breach or fraudulent misrepresentation or as otherwise permitted under a statutory right. Payment must be made in accordance with CCC's Billing and Payment Terms and conditions. In the event that you breach any material condition of these terms and condition or any of CCC's Billing and Payment Terms and Conditions, the license is automatically terminated upon written notice from the BMJ Group or CCC or as otherwise provided for in CCC's Billing and Payment Terms and Conditions, where these apply. Continued use of materials where a licence has been terminated, as well as any use of the Licensed Materials beyond the scope of an unrevoked licence, may constitute intellectual property rights infringement and BMJ Group reserves the right to take any and all action to protect its intellectual property rights in the Licensed Materials.

11. No Transfer of License: This licence is personal to you, and may not be assigned or transferred by you without prior written consent from the BMJ Group or its authorised agent(s). BMJ Group may assign or transfer any of its rights and obligations under this Agreement, upon written notice to you. 12. No Amendment Except in Writing: This licence may not be amended except in a writing signed by both parties (or, in the case of BMJ Group, by CCC on the BMJ Group's behalf). 13. Objection to Contraryterms: BMJ Group hereby objects to any terms contained in any purchase order, acknowledgment, check endorsement or other writing prepared by you, which terms are inconsistent with these terms and conditions or CCC's Billing and Payment Terms and Conditions. These terms and conditions, together with CCC's Billing and Payment Terms and Conditions (which to the extent they are consistent are incorporated herein), comprise the entire agreement between you and BMJ Group (and CCC) and the Licensee concerning this licensing transaction. In the event of any conflict between your obligations established by these terms and conditions and those established by CCC's Billing and Payment Terms and Conditions, these terms and conditions shall control.

14. Revocation: BMJ Group or CCC may, within 30 days of issuance of this licence, deny the 209

permissions described in this licence at their sole discretion, for any reason or no reason, with a full refund payable to you should you have not been able to exercise your rights in full. Notice of such denial will be made using the contact information provided by you. Failure to receive such notice from BMJ Group or CCC will not, to the fullest extent permitted by law alter or invalidate the denial. For the fullest extent permitted by law in no event will BMJ Group or CCC be responsible or liable for any costs, expenses or damage incurred by you as a result of a denial of your permission request, other than a refund of the amount(s) paid by you to BMJ Group and/or CCC for denied permissions.

15. Restrictions to the license:

16. Translation: This permission is granted for non•exclusive world English language rights only unless explicitly stated in your licence. If translation rights are granted, a professional translator should be employed and the content should be reproduced word for word preserving the integrity of the content. 17. General: Neither party shall be liable for failure, default or delay in performing its obligations under this Licence, caused by a Force Majeure event which shall include any act of God, war, or threatened war, act or threatened act of terrorism, riot, strike, lockout, individual action, fire, flood, drought, tempest or other event beyond the reasonable control of either party.

17.1 In the event that any provision of this Agreement is held to be invalid, the remainder of the provisions shall continue in full force and effect.

210

NATURE PUBLISHING GROUP LICENSE TERMS AND CONDITIONS

License Number 3886200975410

License date Jun 11, 2016

Licensed Content Publisher Nature Publishing Group

Licensed Content Publication Nature

Licensed Content Title Co•crystal structure of the HNF•3/fork head DNA•recognition motif resembles histone H5

Licensed Content Author Kirk L. Clark, Elaine D Halay, Eseng Lai, Stephen K Burley

Licensed Content Date Jul 29, 1993

Licensed Content Volume 364 Number

Licensed Content Issue 6436 Number

Type of Use reuse in a dissertation / thesis

Requestor type academic/educational

Format print and electronic

Portion figures/tables/illustrations

Number of 1 figures/tables/illustrations

Figures Figure 3

Author of this NPG article no

Your reference number

Title of your thesis / Determining genetic mechanisms of vascular stability: A novel dissertation role for FoxF2

Expected completion date Jul 2016

Estimated size (number of 150 pages)

Requestor Location Corey Arnold Canada Attn: Corey Arnold

Billing Type Invoice

Billing Address Corey Arnold Canada Attn: Corey Arnold

211

Total 0.00 CAD Terms and Conditions

Note: For republication from the British Journal of Cancer, the following credit lines apply. Reprinted by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL NAME] (reference citation), copyright (year of publication)For AOP papers, the credit line should read: Reprinted by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL NAME], advance online publication, day month year (doi: 10.1038/sj.[JOURNAL ACRONYM].XXXXX)

6. Adaptations of single figures do not require NPG approval. However, the adaptation should be credited as follows:

Adapted by permission from Macmillan Publishers Ltd: [JOURNAL NAME] (reference citation), copyright (year of publication)

Note: For adaptation from the British Journal of Cancer, the following credit line applies. Adapted by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL 212

NAME] (reference citation), copyright (year of publication)

Translated by permission from Macmillan Publishers Ltd: [JOURNAL NAME] (reference citation), copyright (year of publication).

Note: For translation from the British Journal of Cancer, the following credit line applies.

Translated by permission from Macmillan Publishers Ltd on behalf of Cancer Research UK: [JOURNAL NAME] (reference citation), copyright (year of publication)

We are certain that all parties will benefit from this agreement and wish you the best in the use of this material. Thank you. Special Terms: v1.1

213

American Society for Biochemistry and Molecular Biology

To whom it may concern,

It is the policy of the American Society for Biochemistry and Molecular Biology to allow reuse of any material published in its journals (the Journal of Biological Chemistry, Molecular & Cellular Proteomics and the Journal of Lipid Research) in a thesis or dissertation at no cost and with no explicit permission needed. Please see our copyright permissions page on the journal site for more information.

Best wishes,

Sarah Crespi

American Society for Biochemistry and Molecular Biology

214

NRC RESEARCH PRESS LICENSE TERMS AND CONDITIONS

This Agreement between Corey Arnold ("You") and NRC Research Press ("NRC Research Press") consists of your license details and the terms and conditions provided by NRC Research Press and Copyright Clearance Center.

License Number 3902750880966

License date Jul 05, 2016

Licensed Content Publisher NRC Research Press

Licensed Content Publication Genome

Licensed Content Title Fox gene loci in Takifugu rubripes and Tetraodon nigroviridis genomes and comparison with those of medaka and zebrafish genomes

Licensed Content Author Xueyan Shen, Jianzhou Cui, Qingli Gong

Licensed Content Date Dec 1, 2011

Licensed Content Volume 54 Number

Licensed Content Issue 12 Number

Type of Use Thesis/Dissertation

Requestor type Author (original work)

Format Print and electronic

Portion Figure/table

Number of figures/tables 1

Order reference number

Title of your thesis / Determining genetic mechanisms of vascular stability: A dissertation novel role for FoxF2

Expected completion date Jul 2016

Estimated size(pages) 150

Requestor Location Corey Arnold Canada Attn: Corey Arnold

Billing Type Invoice

Billing Address Corey Arnold Canada Attn: Corey Arnold

Total 0.00 CAD

Terms and Conditions General Terms & Conditions 215

Permission is granted upon the requester's compliance with the following terms and conditions:

1. A credit line will be prominently placed in your product(s) and include: for books the author, book title, editor, copyright holder, year of publication; for journals the author, title of article, title of journal, volume number, issue number, and the inclusive pages. The credit line must include the following wording: "© 2008 Canadian Science Publishing or its licensors. Reproduced with permission," except when an author of an original article published in 2009 or later is reproducing his/her own work.

2. The requester warrants that the material shall not be used in any manner that may be derogatory to the title, content, or authors of the material or to Canadian Science Publishing, including but not limited to an association with conduct that is fraudulent or otherwise illegal.

3. Permission is granted for the term (for Books/CDsShelf Life; for Internet/IntranetIn perpetuity; for all other forms of printthe life of the title) and purpose specified in your request. Once term has expired, permission to renew must be made in writing.

4. Permission granted is nonexclusive, and is valid throughout the world in English and the languages specified in your original request. A new permission must be requested for revisions of the publication under current consideration.

5. Canadian Science Publishing cannot supply the requester with the original artwork or a "clean copy."

6. If the Canadian Science Publishing material is to be translated, the following lines must be included: The authors, editors, and Canadian Science Publishing are not responsible for errors or omissions in translations. v1.4

216

The Company of Biologists Ltd LICENSE TERMS AND CONDITIONS

This is a License Agreement between Corey Arnold ("You") and The Company of Biologists Ltd ("The Company of Biologists Ltd") provided by Copyright Clearance Center ("CCC"). The license consists of your order details, the terms and conditions provided by The Company of Biologists Ltd, and the payment terms and conditions.

All payments must be made in full to CCC. For payment instructions, please see information listed at the bottom of this form.

License Number 3886210891705

License date Jun 11, 2016

Licensed content publisher The Company of Biologists Ltd

Licensed content title Development

Licensed content date Jan 1, 1987

Type of Use Thesis/Dissertation

Requestor type Academic institution

Format Print, Electronic

Portion image/photo

Number of images/photos 1 requested

Title or numeric reference of Figure 2B the portion(s)

Title of the article or chapter The identification of genes with unique and essential functions in the portion is from the development of the zebrafish, Danio rerio.

Editor of portion(s) N/A

Author of portion(s) N/A

Volume of serial or 123 monograph.

Page range of the portion 1•36

Publication date of portion December 1996

Rights for Main product

Duration of use Current edition and up to 5 years

Creation of copies for the no disabled

With minor editing privileges yes 217

For distribution to Canada

In the following language(s) Original language of publication

With incidental promotional no use

The lifetime unit quantity of Up to 499 new product

Made available in the Education following markets

The requesting Corey Arnold person/organization is:

Order reference number

Author/Editor Corey Arnold

The standard identifier of CRA001 New Work

Title of New Work Determining genetic mechanisms of vascular stability: A novel role for FoxF2

Publisher of New Work University of Calgary

Expected publication date Jul 2016

Estimated size (pages) 150

Terms and Conditions TERMS AND CONDITIONS The following terms are individual to this publisher: The acknowledgement should state "Reproduced / adapted with permission" and give the source journal name. The acknowledgement should either provide full citation details or refer to the relevant citation in the article reference list. The full citation details should include authors, journal, year, volume, issue and page citation. Where appearing online or in other electronic media, a link should be provided to the original article (e.g. via DOI): Development: dev.biologists.org Disease Models & Mechanisms: dmm.biologists.org Journal of Cell Science: jcs.biologists.org The Journal of Experimental Biology: jeb.biologists.org Other Terms and Conditions: STANDARD TERMS AND CONDITIONS 1. Description of Service; Defined Terms. This Republication License enables the User to obtain licenses for republication of one or more copyrighted works as described in detail on the relevant Order Confirmation (the “Work(s)”). Copyright Clearance Center, Inc. (“CCC”) grants licenses through the Service on behalf of the rightsholder identified on the Order Confirmation (the “Rightsholder”). “Republication”, as used herein, generally means the inclusion of a Work, in whole or in part, in a new work or works, also as described on the Order Confirmation. “User”, as used herein, means the person or entity making such republication.

218

2. The terms set forth in the relevant Order Confirmation, and any terms set by the Rightsholder with respect to a particular Work, govern the terms of use of Works in connection with the Service. By using the Service, the person transacting for a republication license on behalf of the User represents and warrants that he/she/it (a) has been duly authorized by the User to accept, and hereby does accept, all such terms and conditions on behalf of User, and (b) shall inform User of all such terms and conditions. In the event such person is a “freelancer” or other third party independent of User and CCC, such party shall be deemed jointly a “User” for purposes of these terms and conditions. In any event, User shall be deemed to have accepted and agreed to all such terms and conditions if User republishes the Work in any fashion. 3. Scope of License; Limitations and Obligations. 3.1 All Works and all rights therein, including copyright rights, remain the sole and exclusive property of the Rightsholder. The license created by the exchange of an Order Confirmation (and/or any invoice) and payment by User of the full amount set forth on that document includes only those rights expressly set forth in the Order Confirmation and in these terms and conditions, and conveys no other rights in the Work(s) to User. All rights not expressly granted are hereby reserved. 3.2 General Payment Terms: You may pay by credit card or through an account with us payable at the end of the month. If you and we agree that you may establish a standing account with CCC, then the following terms apply: Remit Payment to: Copyright Clearance Center, Dept 001, P.O. Box 843006, Boston, MA 02284•3006. Payments Due: Invoices are payable upon their delivery to you (or upon our notice to you that they are available to you for downloading). After 30 days, outstanding amounts will be subject to a service charge of 1•1/2% per month or, if less, the maximum rate allowed by applicable law. Unless otherwise specifically set forth in the Order Confirmation or in a separate written agreement signed by CCC, invoices are due and payable on “net 30” terms. While User may exercise the rights licensed immediately upon issuance of the Order Confirmation, the license is automatically revoked and is null and void, as if it had never been issued, if complete payment for the license is not received on a timely basis either from User directly or through a payment agent, such as a credit card company. 3.3 Unless otherwise provided in the Order Confirmation, any grant of rights to User (i) is “onetime” (including the editions and product family specified in the license), (ii) is non exclusive and non•transferable and (iii) is subject to any and all limitations and restrictions (such as, but not limited to, limitations on duration of use or circulation) included in the Order Confirmation or invoice and/or in these terms and conditions. Upon completion of the licensed use, User shall either secure a new permission for further use of the Work(s) or immediately cease any new use of the Work(s) and shall render inaccessible (such as by deleting or by removing or severing links or other locators) any further copies of the Work (except for copies printed on paper in accordance with this license and still in User's stock at the end of such period). 3.4 In the event that the material for which a republication license is sought includes third party materials (such as photographs, illustrations, graphs, inserts and similar materials) which are identified in such material as having been used by permission, User is responsible for identifying, and seeking separate licenses (under this Service or otherwise) for, any of such third party materials; without a separate license, such third party materials may not be used. 3.5 Use of proper copyright notice for a Work is required as a condition of any license granted under the Service. Unless otherwise provided in the Order Confirmation, a proper copyright notice will read substantially as follows: “Republished with permission of [Rightsholder’s 219

name], from [Work's title, author, volume, edition number and year of copyright]; permission conveyed through Copyright Clearance Center, Inc. ” Such notice must be provided in a reasonably legible font size and must be placed either immediately adjacent to the Work as used (for example, as part of a by•line or footnote but not as a separate electronic link) or in the place where substantially all other credits or notices for the new work containing the republished Work are located. Failure to include the required notice results in loss to the Rightsholder and CCC, and the User shall be liable to pay liquidated damages for each such failure equal to twice the use fee specified in the Order Confirmation, in addition to the use fee itself and any other fees and charges specified. 3.6 User may only make alterations to the Work if and as expressly set forth in the Order Confirmation. No Work may be used in any way that is defamatory, violates the rights of third parties (including such third parties' rights of copyright, privacy, publicity, or other tangible or intangible property), or is otherwise illegal, sexually explicit or obscene. In addition, User may not conjoin a Work with any other material that may result in damage to the reputation of the Rightsholder. User agrees to inform CCC if it becomes aware of any infringement of any rights in a Work and to cooperate with any reasonable request of CCC or the Rightsholder in connection therewith. 4. Indemnity. User hereby indemnifies and agrees to defend the Rightsholder and CCC, and their respective employees and directors, against all claims, liability, damages, costs and expenses, including legal fees and expenses, arising out of any use of a Work beyond the scope of the rights granted herein, or any use of a Work which has been altered in any unauthorized way by User, including claims of defamation or infringement of rights of copyright, publicity, privacy or other tangible or intangible property. 5. Limitation of Liability. UNDER NO CIRCUMSTANCES WILL CCC OR THE RIGHTSHOLDER BE LIABLE FOR ANY DIRECT, INDIRECT, CONSEQUENTIAL OR INCIDENTAL DAMAGES (INCLUDING WITHOUT LIMITATION DAMAGES FOR LOSS OF BUSINESS PROFITS OR INFORMATION, OR FOR BUSINESS INTERRUPTION) ARISING OUT OF THE USE OR INABILITY TO USE A WORK, EVEN IF ONE OF THEM HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. In any event, the total liability of the Rightsholder and CCC (including their respective employees and directors) shall not exceed the total amount actually paid by User for this license. User assumes full liability for the actions and omissions of its principals, employees, agents, affiliates, successors and assigns. 6. Limited Warranties. THE WORK(S) AND RIGHT(S) ARE PROVIDED “AS IS”. CCC HAS THE RIGHT TO GRANT TO USER THE RIGHTS GRANTED IN THE ORDER CONFIRMATION DOCUMENT. CCC AND THE RIGHTSHOLDER DISCLAIM ALL OTHER WARRANTIES RELATING TO THE WORK(S) AND RIGHT(S), EITHER EXPRESS OR IMPLIED, INCLUDING WITHOUT LIMITATION IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. ADDITIONAL RIGHTS MAY BE REQUIRED TO USE ILLUSTRATIONS, GRAPHS, PHOTOGRAPHS, ABSTRACTS, INSERTS OR OTHER PORTIONS OF THE WORK (AS OPPOSED TO THE ENTIRE WORK) IN A MANNER CONTEMPLATED BY USER; USER UNDERSTANDS AND AGREES THAT NEITHER CCC NOR THE RIGHTSHOLDER MAY HAVE SUCH ADDITIONAL RIGHTS TO GRANT. 7. Effect of Breach. Any failure by User to pay any amount when due, or any use by User of a Work beyond the scope of the license set forth in the Order Confirmation and/or these terms and conditions, shall be a material breach of the license created by the Order Confirmation and these

220

terms and conditions. Any breach not cured within 30 days of written notice thereof shall result in immediate termination of such license without further notice. Any unauthorized (but licensable) use of a Work that is terminated immediately upon notice thereof may be liquidated by payment of the Rightsholder's ordinary license price therefor; any unauthorized (and unlicensable) use that is not terminated immediately for any reason (including, for example, because materials containing the Work cannot reasonably be recalled) will be subject to all remedies available at law or in equity, but in no event to a payment of less than three times the Rightsholder's ordinary license price for the most closely analogous licensable use plus Rightsholder's and/or CCC's costs and expenses incurred in collecting such payment. 8. Miscellaneous. 8.1 User acknowledges that CCC may, from time to time, make changes or additions to the Service or to these terms and conditions, and CCC reserves the right to send notice to the User by electronic mail or otherwise for the purposes of notifying User of such changes or additions; provided that any such changes or additions shall not apply to permissions already secured and paid for. 8.2 Use of Userrelated information collected through the Service is governed by CCC’s privacy policy, available online here: http://www.copyright.com/content/cc3/en/tools/footer/privacypolicy.html. 8.3 The licensing transaction described in the Order Confirmation is personal to User. Therefore, User may not assign or transfer to any other person (whether a natural person or an organization of any kind) the license created by the Order Confirmation and these terms and conditions or any rights granted hereunder; provided, however, that User may assign such license in its entirety on written notice to CCC in the event of a transfer of all or substantially all of User’s rights in the new material which includes the Work(s) licensed under this Service. 8.4 No amendment or waiver of any terms is binding unless set forth in writing and signed by the parties. The Rightsholder and CCC hereby object to any terms contained in any writing prepared by the User or its principals, employees, agents or affiliates and purporting to govern or otherwise relate to the licensing transaction described in the Order Confirmation, which terms are in any way inconsistent with any terms set forth in the Order Confirmation and/or in these terms and conditions or CCC's standard operating procedures, whether such writing is prepared prior to, simultaneously with or subsequent to the Order Confirmation, and whether such writing appears on a copy of the Order Confirmation or in a separate instrument. 8.5 The licensing transaction described in the Order Confirmation document shall be governed by and construed under the law of the State of New York, USA, without regard to the principles thereof of conflicts of law. Any case, controversy, suit, action, or proceeding arising out of, in connection with, or related to such licensing transaction shall be brought, at CCC's sole discretion, in any federal or state court located in the County of New York, State of New York, USA, or in any federal or state court whose geographical jurisdiction covers the location of the Rightsholder set forth in the Order Confirmation. The parties expressly submit to the personal jurisdiction and venue of each such federal or state court.If you have any comments or questions about the Service or Copyright Clearance Center, please contact us at 978•750•8400 or send an e•mail to [email protected]. v 1.1

221

ELSEVIER LICENSE TERMS AND CONDITIONS

This Agreement between Corey Arnold ("You") and Elsevier ("Elsevier") consists of your license details and the terms and conditions provided by Elsevier and Copyright Clearance Center.

License Number 3912111096405

License date Jul 18, 2016

Licensed Content Publisher Elsevier

Licensed Content Publication The Lancet Neurology

Licensed Content Title Identification of additional risk loci for stroke and small vessel disease:a meta•analysis of genome•wide association studies

Licensed Content Author

Licensed Content Date June 2016

Licensed Content Volume 15 Number

Licensed Content Issue 7 Number

Licensed Content Pages 13

Start Page 695

End Page 707

Type of Use reuse in a thesis/dissertation

Intended publisher of new other work

Portion figures/tables/illustrations

Number of 2 figures/tables/illustrations

Format both print and electronic

Are you the author of this Yes Elsevier article?

Will you be translating? No

Order reference number

Original figure numbers Figures 1 and 2

Title of your Determining genetic mechanisms of vascular stability: A novel thesis/dissertation role for FoxF2

Expected completion date Jul 2016

222

Estimated size (number of 150 pages)

Elsevier VAT number GB 494 6272 12

Requestor Location Corey Arnold Canada

Total 0.00 CAD

223

revoked and shall be void as if never granted. Further, in the event that you breach any of these terms and conditions or any of CCC's Billing and Payment terms and conditions, the license is automatically revoked and shall be void as if never granted. Use of materials as described in a revoked license, as well as any use of the materials beyond the scope of an unrevoked license, may constitute copyright infringement and publisher reserves the right to take any and all action to protect its copyright in the materials. 9. Warranties: Publisher makes no representations or warranties with respect to the licensed material. 10. Indemnity: You hereby indemnify and agree to hold harmless publisher and CCC, and their respective officers, directors, employees and agents, from and against any and all claims arising out of your use of the licensed material other than as specifically authorized pursuant to this license. 11. No Transfer of License: This license is personal to you and may not be sublicensed, assigned, or transferred by you to any other person without publisher's written permission. 12. No Amendment Except in Writing: This license may not be amended except in a writing signed by both parties (or, in the case of publisher, by CCC on publisher's behalf). 13. Objection to Contrary Terms: Publisher hereby objects to any terms contained in any purchase order, acknowledgment, check endorsement or other writing prepared by you, which terms are inconsistent with these terms and conditions or CCC's Billing and Payment terms and conditions. These terms and conditions, together with CCC's Billing and Payment terms and conditions (which are incorporated herein), comprise the entire agreement between you and publisher (and CCC) concerning this licensing transaction. In the event of any conflict between your obligations established by these terms and conditions and those established by CCC's Billing and Payment terms and conditions, these terms and conditions shall control. 14. Revocation: Elsevier or Copyright Clearance Center may deny the permissions described in this License at their sole discretion, for any reason or no reason, with a full refund payable to you. Notice of such denial will be made using the contact information provided by you. Failure to receive such notice will not alter or invalidate the denial. In no event will Elsevier or Copyright Clearance Center be responsible or liable for any costs, expenses or damage incurred by you as a result of a denial of your permission request, other than a refund of the amount(s) paid by you to Elsevier and/or Copyright Clearance Center for denied permissions. LIMITED LICENSE The following terms and conditions apply only to specific license types: 15. Translation: This permission is granted for non•exclusive world English rights only unless your license was granted for translation rights. If you licensed translation rights you may only translate this content into the languages you requested. A professional translator must perform all translations and reproduce the content word for word preserving the integrity of the article. 16. Posting licensed content on any Website: The following terms and conditions apply as follows: Licensing material from an Elsevier journal: All content posted to the web site must maintain the copyright information line on the bottom of each image; A hypertext must be included to the Homepage of the journal from which you are licensing at http://www.sciencedirect.com/science/journal/xxxxx or the Elsevier homepage for books at http://www.elsevier.com; Central Storage: This license does not include permission for a scanned version of the material to be stored in a central repository such as that provided by Heron/XanEdu. 224

Licensing material from an Elsevier book: A hyper•text link must be included to the Elsevier homepage at http://www.elsevier.com. All content posted to the web site must maintain the copyright information line on the bottom of each image.

Posting licensed content on Electronic reserve: In addition to the above the following clauses are applicable: The web site must be password•protected and made available only to bona fide students registered on a relevant course. This permission is granted for 1 year only. You may obtain a new license for future website posting. 17. For journal authors: the following clauses are applicable in addition to the above: Preprints: A preprint is an author's own write•up of research results and analysis, it has not been peer• reviewed, nor has it had any other value added to it by a publisher (such as formatting, copyright, technical enhancement etc.). Authors can share their preprints anywhere at any time. Preprints should not be added to or enhanced in any way in order to appear more like, or to substitute for, the final versions of articles however authors can update their preprints on arXiv or RePEc with their Accepted Author Manuscript (see below). If accepted for publication, we encourage authors to link from the preprint to their formal publication via its DOI. Millions of researchers have access to the formal publications on ScienceDirect, and so links will help users to find, access, cite and use the best available version. Please note that Cell Press, The Lancet and some society•owned have different preprint policies. Information on these policies is available on the journal homepage. Accepted Author Manuscripts: An accepted author manuscript is the manuscript of an article that has been accepted for publication and which typically includes author• incorporated changes suggested during submission, peer review and editor•author communications. Authors can share their accepted author manuscript:

In all cases accepted manuscripts should:

 link to the formal publication via its DOI   bear a CC•BY•NC•ND license • this is easy to do 

225

 if aggregated with other manuscripts, for example in a repository or other site, be shared in alignment with our hosting policy not be added to or enhanced in any way to appear more like, or to substitute for, the published journal article. 

Published journal article (JPA): A published journal article (PJA) is the definitive final record of published research that appears or will appear in the journal and embodies all value•adding publishing activities including peer review co•ordination, copy•editing, formatting, (if relevant) pagination and online enrichment. Policies for sharing publishing journal articles differ for subscription and gold open access articles: Subscription Articles: If you are an author, please share a link to your article rather than the full•text. Millions of researchers have access to the formal publications on ScienceDirect, and so links will help your users to find, access, cite, and use the best available version. Theses and dissertations which contain embedded PJAs as part of the formal submission can be posted publicly by the awarding institution with DOI links back to the formal publications on ScienceDirect. If you are affiliated with a library that subscribes to ScienceDirect you have additional private sharing rights for others' research accessed under that agreement. This includes use for classroom teaching and internal training at the institution (including use in course packs and courseware programs), and inclusion of the article for grant funding purposes. Gold Open Access Articles: May be shared according to the author•selected end•user license and should contain a CrossMark logo, the end user license, and a DOI link to the formal publication on ScienceDirect. Please refer to Elsevier's posting policy for further information. 18. For book authors the following clauses are applicable in addition to the above: Authors are permitted to place a brief summary of their work online only. You are not allowed to download and post the published electronic version of your chapter, nor may you scan the printed edition to create an electronic version. Posting to a repository: Authors are permitted to post a summary of their chapter only in their institution's repository. 19. Thesis/Dissertation: If your license is for use in a thesis/dissertation your thesis may be submitted to your institution in either print or electronic form. Should your thesis be published commercially, please reapply for permission. These requirements include permission for the Library and Archives of Canada to supply single copies, on demand, of the complete thesis and include permission for Proquest/UMI to supply single copies, on demand, of the complete thesis. Should your thesis be published commercially, please reapply for permission. Theses and dissertations which contain embedded PJAs as part of the formal submission can be posted publicly by the awarding institution with DOI links back to the formal publications on ScienceDirect.

226

nor should the article be modified in such a way as to damage the author's honour or reputation. If any changes have been made, such changes must be clearly indicated. The author(s) must be appropriately credited and we ask that you include the end user license and a DOI link to the formal publication on ScienceDirect. If any part of the material to be used (for example, figures) has appeared in our publication with credit or acknowledgement to another source it is the responsibility of the user to ensure their reuse complies with the terms and conditions determined by the rights holder. Additional Terms & Conditions applicable to each Creative Commons user license: CC BY: The CC•BY license allows users to copy, to create extracts, abstracts and new works from the Article, to alter and revise the Article and to make commercial use of the Article (including reuse and/or resale of the Article by commercial entities), provided the user gives appropriate credit (with a link to the formal publication through the relevant DOI), provides a link to the license, indicates if changes were made and the licensor is not represented as endorsing the use made of the work. The full details of the license are available at http://creativecommons.org/licenses/by/4.0. CC BY NC SA: The CC BY•NC•SA license allows users to copy, to create extracts, abstracts and new works from the Article, to alter and revise the Article, provided this is not done for commercial purposes, and that the user gives appropriate credit (with a link to the formal publication through the relevant DOI), provides a link to the license, indicates if changes were made and the licensor is not represented as endorsing the use made of the work. Further, any new works must be made available on the same conditions. The full details of the license are available at http://creativecommons.org/licenses/by•nc•sa/4.0. CC BY NC ND: The CC BY•NC•ND license allows users to copy and distribute the Article, provided this is not done for commercial purposes and further does not permit distribution of the Article if it is changed or edited in any way, and provided the user gives appropriate credit (with a link to the formal publication through the relevant DOI), provides a link to the license, and that the licensor is not represented as endorsing the use made of the work. The full details of the license are available at http://creativecommons.org/licenses/by•nc•nd/4.0. Any commercial reuse of Open Access articles published with a CC BY NC SA or CC BY NC ND license requires permission from Elsevier and will be subject to a fee. Commercial reuse includes:

 Associating advertising with the full text of the Article   Charging fees for document delivery or access   Article aggregation   Systematic distribution via e•mail lists or share buttons 

Posting or linking by commercial companies for use by customers of those companies.

20. Other Conditions: v1.8

227