University of Alberta

Phenotypic Characterization of a Zygotic-Lethal unc45b~/~; unc45d/~ Mutant in Zebrafish, Danio rerio

by

Sophie Comyn

A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfillment of the requirements for the degree of

Master of Science in Molecular Biology and Genetics

Biological Sciences

©Sophie A. C. Comyn Fall 2011 Edmonton, Alberta

Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only Where the thesis is converted to, or otherwise made available m digital form, the University of Alberta will advise potential users of the thesis of these terms

The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission Library and Archives Bibliotheque et 1*1 Canada Archives Canada Published Heritage Direction du Branch Patrimoine de I'edition

395 Wellington Street 395, rue Wellington Ottawa ON K1A 0N4 Ottawa ON K1A 0N4 Canada Canada

Your Tile Votre reference ISBN: 978-0-494-81305-8 Our file Notre reference ISBN: 978-0-494-81305-8

NOTICE: AVIS:

The author has granted a non­ L'auteur a accorde une licence non exclusive exclusive license allowing Library and permettant a la Bibliotheque et Archives Archives Canada to reproduce, Canada de reproduire, publier, archiver, publish, archive, preserve, conserve, sauvegarder, conserver, transmettre au public communicate to the public by par telecommunication ou par I'lnternet, preter, telecommunication or on the Internet, distribuer et vendre des theses partout dans le loan, distribute and sell theses monde, a des fins commerciaies ou autres, sur worldwide, for commercial or non­ support microforme, papier, electronique et/ou commercial purposes, in microform, autres formats. paper, electronic and/or any other formats.

The author retains copyright L'auteur conserve la propriete du droit d'auteur ownership and moral rights in this et des droits moraux qui protege cette these. Ni thesis. Neither the thesis nor la these ni des extraits substantiels de celle-ci substantial extracts from it may be ne doivent etre imprimes ou autrement printed or otherwise reproduced reproduits sans son autorisation. without the author's permission.

In compliance with the Canadian Conformement a la loi canadienne sur la Privacy Act some supporting forms protection de la vie privee, quelques may have been removed from this formulaires secondaires ont ete enleves de thesis. cette these.

While these forms may be included Bien que ces formulaires aient inclus dans in the document page count, their la pagination, il n'y aura aucun contenu removal does not represent any loss manquant. of content from the thesis.

1+1 Canada Abstract

Members of the UCS (UNC-45/CR01/She4p) protein family perform essential myosin- and actin-dependent functions through direct interaction with the myosin motor domain. Two vertebrate UNC-45 genes, unc45a and unc45b, have been identified, and shown to be functionally redundant in vitro. The purpose of this project was to assess the phenotype of a zygotic-lethal unc45b"; unc45d~ mutant, and to determine whether Unc45a and Unc45b are functionally redundant in vivo. I performed a detailed phenotypic characterization of the zebrafish unc45b~'; unc45d' mutant focusing on morphology, gene compensation, and gene expression. I have determined that unc45b"; unc45d/~ and unc45b~/~ are phenotypically indistinguishable and that both mutants display cardiac, muscle, jaw, and eye defects. Therefore, unlike observations in vitro, my results do not support functional redundancy between Unc45a and Unc45b in vivo. Acknowledgements

I would like to thank my supervisor, Dr. David Pilgrim, and my committee members Dr. Ted Allison, Dr. Paul LaPointe, and Dr. Andrew Waskiewicz for their academic support and guidance. Thank you to members of the Pilgrim lab for providing valuable feedback and to Danielle French for teaching me about zebrafish and its methods. I would like to express my appreciation to Dr. Waskiewicz and his lab members, past and present, for their generosity in allowing me to use their equipment and reagents as well as for their assistance. Also, thank you to Aleah McCorry for maintaining the zebrafish facility. I am especially grateful to Tan Bao and Cheryl Han for their friendship, insightful discussions, and moral support. Finally, I would like to express my gratitude to my family for their encouragement, guidance, and patience. Table of Contents

CHAPTER ONE: INTRODUCTION

1.1 Why Study an unc45b~/~; unc45d/~ Mutant? 4

1.2 Why has the Zebrafish Become Such a Popular Model System? 7

1.3 Members of the UCS Protein Family are Present in Single-Celled Eukaryotes to Metazoans 8 1.3.1 Caenorhabditis elegans UNC-45 9 1.3.2 Vertebrate UNC-45 11 1.3.3 Biochemical Analyses Demonstrate Functional Redundancy between Unc45a and Unc45b in vitro 12 1.3.4 Unc45b and Hsp90 are Necessary for Myosin Motor Domain Folding and Assembly 13 1.3.5 Analysis of Unc45a Function in vivo 15 1.3.6 Do Vertebrate Unc45 Proteins have Functions Distinct from Myosin Chaperoning? 16 1.3.7 Fungal UCS Proteins 18 1.3.8 Crol in P. anserine 19 1.3.9 She4pinS. cerevisiae 19 1.3.10 Rng3p vaS.pombe 19

1.4 Folding of the Myosin Motor Domain into its Native Conformation is a Requirement for Thick Filament Assembly and Muscle Function 20 1.4.1 Myosins are a Large Superfamily with Diverse Cellular Functions . 20 1.4.2 Myosin Molecules Assemble into Thick Filaments 21 1.4.3 Myosin Folding, Thick Filament Assembly, and Sarcomere Organization Require Chaperones 22 1.4.4 Sarcomeres are the Basic Contractile Unit of Striated Muscle .... 23

1.5 Zebrafish have Three Striated Muscle Populations 24 1.5.1 Cardiac Muscle: Circulation 24 1.5.2 Skeletal Muscle of the Trunk and Tail: Locomotion 27 1.5.3 Craniofacial Muscle: Feeding, Respiration and Ocular Movement . . 28 1.6 Summary of Project Objectives 30

1.7 Table 32

1.8 Figures 33

1.9 References 41

CHAPTER TWO: MATERIALS AND METHODS

2.1 Animal Care and Zebrafish Lines 51 2.1.1 Zebrafish Maintenance 51 2.1.2 steif 51 2.1.3 kurzschluss 51 2 1 4 imr45hsb60/sb6°- unr45a'rI2/"'12 52

2.2 Genotyping 52 2.2.1 Genomic DNA Extraction 52 2.2.2 dCAPS Analysis 53

2.3 Whole Mount in situ Hybridization 54 2.3.1 Total RNA Extraction and cDNA Synthesis 54 2.3.2 Preparation ofPCR Based DIG-labelled RNA Probes 55 2.3.3 mRNA in situ Hybridization and Detection 56

2.4 Immunohistochemistry 57 2.4.1 MF-20 57 2.4.2 3,3'-Diaminobenzidine 58 2.4.3 Eye Cross Sections and Immunohistochemistry 59

2.5 Whole Mount Staining 59 2.5.1 Phalloidin 59 2.5.2 Alcian Blue 60 2.5.3 Alizarin Red 60 2.5.4 (9-dianisidine 61 2.6 Photography and Image Processing 61

2.7 Tables 62

2.8 Figures 66

2.9 References 67

CHAPTER THREE: RESULTS

3.1 Gross Morphology of unc45 Mutants 71 3.1.1 Circulation Defects 72 3.1.2 Skeletal Muscle Organization 73 3.1.3 Cranial Muscle Organization 74 3.1.4 Skeletal Defects 74 3.1.5 Gross Ocular Morphology 76

3.2 Compensation Pathways in unc45b"; unc45d~ Mutants 77 3.2.1 unc45 Expression 77 3.2.2 hsp90 Expression 78

3.3 Gene Expression and Determination 79 3.3.1 Cardiac Determination 79 3.3.2 Expression of the Myogenic Regulatory Factor myoD 80 3.3.3 Pharyngeal Arch Formation and Patterning 81

3.4 Conclusions 81

3.5 Figures 82

3.6 References 94

CHAPTER FOUR: DISCUSSION

4.1 Are Unc45a and Unc45b Functionally Redundant in vivol 97

4.2 unc45b'/' and unc45b'/~; unc45d'~ Mutants Share Similar Cardiovascular Defects 100 4.3 unc45b~/~ and unc45b~'~\ unc45d'~ Mutants Have Identical Muscle Phenotypes 102 4.4 Pharyngeal Arch Cartilage and Bone Mineralization is Disrupted in unc45 Mutants 103 4.5 Small Eyes and Disrupted Lens Fibre Cell Differentiation are Properties of unc45b'/~ Mutants 105 4.6 Different hsp90 Responses are Present in the unc45d' and unc45b~' Mutants 105

4.7 Future Directions 106 4.7.1 Do Unc45a or Unc45b Interact with a Non-Muscle Myosin? .... 106 4.7.2 What is the Function of Unc45a During Zebrafish Development? . 108 4.7.3 Does Unc45b Have a Role in Cardiac Muscle Maintenance? .... 109

4.8 General Conclusions 110

4.9 References Ill

APPENDIX: ELECTROCARDIOGRAPHY

A. 1 Introduction 116

A.2 Results and Discussion 117

A.3 Methods 122 A.3.1 Fish Maintenance 122

A.3.2. In situ Hybridization 122

A.4 Acknowledgements 122

A.5 Tables 123

A.6 Figures 124

A.7 References 128 List of Tables

Table 1-1. Molecular Conservation Amongst UCS Protein Family Members . . 32 Table 2-1. Primers and Restriction Enzymes Used in dCAPS Analysis .... 62 Table 2-2. Primer Sequences for Antisense PCR Based DIG-labelled RNA Probes 63 Table 2-3. Antisense Vector Based DIG-labelled RNA Probes 64 Table 2-4. Antibodies and Stains Used 65 Table A-l. Electrocardiogram measurements from wild type and kus embryos 123 List of Figures

Figure 1-1. Three potential fates of duplicate gene pairs 33 Figure 1-2. UNC-45 domain organization 34 Figure 1-3. Myosin Classes 35 Figure 1-4. Myosin and thick filament organization 36 Figure 1-5. Myosin folding pathway model 37 Figure 1-6. Sarcomere organization 38 Figure 1-7. Zebrafish cardiac development 39 Figure 1-8. Schematic of zebrafish pharyngeal arches at 5 dpf and 48 hpf ... 40 Figure 2-1. Representative dCAPS analysis genotyping gel electrophoresis . . 66 Figure 3-1. Morphology of 4 dpf unc45 mutants 82 Figure 3-2. Whole-mount o-dianisidine staining of hemoglobinized erythrocytes reveals circulation defects and blood pooling in unc45 mutants at 3 dpf 83 Figure 3-3. unc45b mutants have defective thick and thin filament organization in striated trunk muscles 84 Figure 3-4. Myosin expression in cranial muscles 85 Figure 3-5. Skeletal defects in unc45 mutants at 5 dpf 86 Figure 3-6. Analysis of eye phenotypes in unc45 mutants 87 Figure 3-7. Whole-mount in situ hybridization of unc45 and unc45b mRNA expression at 48 hpf 88 Figure 3-8. hsp90a mRNA is up-regulated in unc45b~'~ mutant embryos at 48 hpf 89 Figure 3-9. Whole-mount mRNA in situ hybridization of the cardiac myosin genes 90

Figure 3-10. Whole-mount in situ hybridization of the myogenic regulatory factor myoD in cranial and trunk muscle precursors at 48 hpf . . . 91 Figure 3-11. Formation of pharyngeal arches in unc45 mutants 92 Figure 3-12. Segmentation of pharyngeal arches of unc45 mutants 93 Figure A-l. Representative ECG recordings from 52-60 hpf zebrafish embryos 124 Figure A-2. Kus and steif mutants at 3 and 5 dpf 125 Figure A-3. Quantification of ECG recordings from wild-type and kus embryos 126 Figure A-4. Whole-mount mRNA in situ hybridization to detect expression of cmlc2, NCXlh, and SERCA2 127 List of Symbols, Nomenclature, or Abbreviations

Abbreviation / Symbol Definition

(NH4)2S04 IMttpi sulphate a atrium Altai activator of heat shock protein 90 ATPsfe I am adductor mandibulac amhe =tgc- ^atrM^^y^ieavy chain * ao adductor operculi ^aj^aline phosphatase Apo2 apobcc2 * 'kit:f? |jjJ|nosine triglfpsphate '•; ';" BCIP 5 -bromo-4-chloro-3 '-mdolyphosphate ftp * base pair „ ^ - br branchiostegal rays bovine_serum albumin BTS N-benzyl-p-toluene-sulphomi m ide C2C12 mouse myoblast cell line cb 1Vi.1iohu11Jii.il cDNA complementary DNA ch lVl.llo||\,|| chv constrictor humlcus \entralis cl Jeiiliiiiin cardiac myosin light chain 2 DAB 3,3 '-diaminobenzidinv.- 1>« 4'-6lffiamidind-2-phenyliiid(>lc dCAPS derived cleaved amplified polymorphic sequence den dentary DEPC dicthylpyrocarbonate %: DIG digoxygenin dbc2a distal-less homeobox gene 2a lit '.:»!• "•'«•'- dNTP deoxy nucleotide triphosphate \ dpf days post fertilization Abbreviation / Symbol Definition

dpw dorsal pharyngeal wall 1)1 T dithiothreitol i:c(; electrocardiogram 1 Dl \ ellr. k'lieili.iiiiiiicleiu.uclk' .k n\ en eiitopteiygoid I.M cih;. Iniiii'-i'iiu-.i F-aetin iilaiiicnlous actin / »/.\ lihiohl.isi ijiowih l.n.lor x KK 111*52 l;K50(>-himling protein 4 !iUlll4 < i \ 1 \ hnuhne piotein 1 giitu5 ti \T \-hinding protein 5 iiUlllfl < i \ 1 X-hiiuliiiL' pioiem '« G< l NC45 yciicral cell l;Nl"-45

ll;()2 lndiogen peroxide heart and iieura! crest derivatives exorv.rsSL'l l hand! transcript 2 hh h;.oli\oiilciis Mop llsp"0/llsp*)0 organizing piolein h|>i llOllls po^l k'lllll/.llll'll hs hyos\mplectic ,J llspvn lk-.il NII.IJ. piok-ni n 111*1-2 hea\> chain unconventional myosin 11 ih inii-ili\i>ulciis im iiilcriiiandibuliiris ima Mlk'llll.llulllMll.llls .lllk'Mi'l imp iiilerniandihularis posterior ir lllk'llol ICCllls K( 1 potassium chloride kl)a l.ilo D.ili.'h ROM potassium h\dro\ide

/l//S •.' • »• '.. •. * * hip lc\alor arcus palaliiii Abbreviation / Symbol Definition

1JC1 lithium chloride ll lateral recuis ma\ ma\illa me Meckel's C.IIIII.IJC MIIAI) pemapeptide moiif lor co-factor inleraelions \lcl2 i r.IH.\ le enh IIKVI l.k I

MRC12 magnesium chloride MIIC mvosin hea\\ chain mL millilitre uL micro! it re mr media! rectus

////•/-/ iir.oevmc le'jul.ili'i;. I.kioi '* ipie". M>u~l\ li mRNA messenger ribonucleic acid

HIS millisecond myfi myogenic factor 5 myolt in_\o:jemc ilil k'lenli.ilii'ii 1 NUT nilro-blue tetra/olium chloride ne IIOllK. huhl solute carrier I'amiK X (sodium calcium \CXIh exchanger) II» ii.iiii'L'uni nkx2.5 NK2 transcription factor related 5 N\H-2 i on mii-Je iii\n^in .1 op opercle pain ph.n;.ii'.v.il .neli iiiiNJe» PKSI phosphate buffered saline Tv\cen-2<) l>( K pi'l;. iiieuse Ji.i.n le.k'.iou PI" \ paraformaldehyde |iil PICK'.'UIII pilx2a paired-like transcription factor 2 Abbreviation / Symbol Definition

g>p\ pterygoiofpfe#§s palatoguadrate ;parasphe^ V •• \t RNA ribonucleic acid ^ revolufibls^er minute * „•$ |, ATPase, Ca2+ transporting, cardiac muscle, SERCA2 slow twitch \\ sh sternohyoideus. *". She4p Swi5p-dcpendent HO expression w\ SMUNC45 smooth muscle fcJNC-4 5 SNP single nucleotide polymorphism f \? SO superior oblique % ? sox9a SRY-box containing gene 9a : #v" ST Superior rectus If : h SSC saline sodium citrate y ; ft if tOOth ?t l' "• ,; Ibx5 tbox - rcA •? trichloroacetic acid %. TPR tetratncopeptide repeat # ' tRNA -: transfer' UCS UNC-45/CR01/She4p ri-D-2 uhiquitin fusion degradation uin micoordinated X \ •• entncle vmhc ventricular rmosin heavv chain ZL-I monoclonal antibody, stains the lens 1

CHAPTER ONE: INTRODUCTION 2

The purpose of this thesis was twofold: to assess the phenotype of a zygotic-lethal, unc45b"; unc45d~ mutant, and to determine whether Unc45a and Unc45b are functionally redundant in vivo. Both issues were addressed using the zebrafish (Danio rerio) vertebrate model system. The interest in creating and examining an unc45b~/~; unc45d/' mutant stems from the results of previous studies demonstrating an interaction between UNC-45 and non-muscle myosin, an association between UNC-45 and defective cytokinesis, and functional redundancy between vertebrate Unc45a and Unc45b, in vitro (Kachur et ah, 2004; Liu et ah, 2008). Therefore, we wished to determine if the above were also true for the vertebrate Unc45 proteins in vivo. UNC-45 is a member of the UCS (UNC-45/CR01/She4p) domain- containing protein family found in single-celled eukaryotes and metazoan animals (Hutagalung et ah, 2002). UCS proteins perform essential myosin- and actin- dependent functions by directly interacting with the myosin motor domain. C. elegans UNC-45 and the fungal UCS protein Rng3 have been shown to directly interact with non-muscle myosin II and are suspected to be involved in processes associated with cytokinesis (Wong et ah, 2002; Kachur et ah, 2004). This led us to infer that at least one of the vertebrate UNC-45 paralogues may have the potential to interact with non-muscle myosin. A cytokinesis phenotype, however, had not been reported in either of the zebrafish unc45 mutants (Wohlgemuth et ah, 2007a; Anderson et ah, 2008). Based on an in vitro assay demonstrating the capacity for both Unc45a and Unc45b to fold the myosin motor domain, we believed that functional redundancy between the two vertebrate Unc45 proteins could possibly be masking a cytokinesis phenotype in the zebrafish unc45 mutants (Liu et ah, 2008).

Genome duplications are thought to have played a major role in the evolution of metazoans, vertebrates, and mammals by providing the material necessary for the development of novel functions (Taylor et ah, 2003). The retention of duplicate genes in zebrafish makes them an ideal system with which to study aspects of genome evolution, such as redundancy within gene families (Taylor et ah, 2003). Furthermore, as a vertebrate model system, zebrafish are 3 ideal for studying early developmental events and for performing genetic analyses. In contrast to invertebrates, vertebrates have two UNC-45 homologues. We were, therefore, fortunate that mutants for both zebrafish unc45 genes were available as it facilitated the characterization of the double mutant and allowed for the examination of structures that appear later in development. I performed a thorough phenotypic characterization of the unc45b~~; unc45d/~ mutant paying particular attention to abnormalities identified previously in the single mutants. By focusing on traits that had already been reported, I was able to determine whether the unc45b~/~; unc45d'~ embryos have novel and/or more severe phenotypes than those of the single mutants, indicative of functional redundancy between the two vertebrate Unc45 proteins. The phenotypic characterization of the unc45b'/~; unc45d/' mutant centred on three areas: morphology, gene compensation, and gene expression and differentiation. I examined traits that were shown to be perturbed in the single mutants, namely: gross morphology; cardiac morphology and differentiation; muscle fibre determination and myofibril organization; jaw formation; and eye size. In the sections that follow I will address the value of characterizing a double unc45 mutant and the importance it may have on our understanding of functional redundancy in Unc45. This is achieved by presenting background information encompassing two topics: genome duplication and metazoan evolution, and the UCS protein family and their role in myosin-dependent processes. Finally, I will discuss the embryonic development of the traits perturbed in the zebrafish unc45 mutants by focusing on muscle fibre composition, architecture, and zebrafish striated muscle populations. 4

1.1 Why Study an unc45b"; unc45a ' Mutant?

Ohno (1970) saw genome duplication as a major driving force of evolution and that duplication provided the raw material from which new gene functions could arise. He postulated that two whole genome duplications took place at the beginning of the vertebrate lineage, around 600 million years ago (Wolfe, 2001). This was followed by a third whole genome duplication, specific to the fish lineage, which occurred approximately 350 million years ago around the time of the teleost radiation (Volff, 2005). The fate of duplicated genes and the mechanisms by which they arose are of interest to evolutionary scientists. While it can be problematic to assign molecular homology unambiguously, duplicate genes make up 15% percent of the human genome and 20% in teleosts, such as zebrafish (Prince and Pickett, 2002; Ravi and Venkatesh, 2008). Data gathered from the analysis of teleost genomes highlights two important issues regarding genome duplication. First, the rate of duplicate gene loss is not uniform across lineages as humans and teleosts retain a similar proportion of duplicate genes despite an elapsed time of 300 million years between the final round of whole genome duplication in their respective ancestral lineages (Ravi and Venkatesh, 2008). Second, Ohno's hypothesis that the early vertebrate lineage was subject to two rounds of whole genome duplication remains the subject of contentious debate amongst evolutionary scientists. Duplicate genes have at least three potential fates: nonfunctionalization, subfunctionalization, or neofunctionalization. Following duplication, one gene copy may accumulate sufficient degenerative mutations to become either a pseudogene or to be lost from the genome (Lynch and Force, 2000). This process, referred to as nonfunctionalization, is the most common fate for a duplicate gene pair (Figure 1-1) (Prince and Pickett, 2002). Nonfunctionalization can occur within a few million generations and the half-life of duplicated genes in eukaryotes has been estimated at 3-7 million years (Wolfe, 2001). The genome duplication model presented by Ohno cannot account for the large proportion of duplicate genes that have been retained over millions of years (Cresko et ah, 2003). As a consequence, a new model for the evolution of gene duplications was 5 created. The duplication/degeneration/complementation model (DDC), as it is called, was based on the observation that many genes required for development have a number independent functions based on their spatial and temporal expression patterns (Force et ah, 1999). One way for both gene duplicates to be preserved is through subfunctionalization, whereby both gene copies accumulate degenerative mutations resulting in the partitioning of the ancestral expression domains and/or functions (Lynch and Force, 2000). When duplicate gene pairs accrue loss-of-function mutations that affect separate sub-functions of the ancestral gene, both gene copies are required to produce the ancestral gene function, and therefore, both copies are retained in the genome (Prince and Pickett, 2002). A number of studies have documented subfunctionalization of gene duplicates in zebrafish and as the genomes of more teleost species are sequenced in the future, it is likely that many more cases will be discovered. Two examples from the zebrafish Sox transcription factor family may illustrate the process of subfunctionalization. Sox transcription factors are a family of DNA binding proteins that are involved in a number of developmental processes such as testis determination, neural crest formation, and chondrogenesis (Cresko et ah, 2003). Zebrafish have two Sox9 genes, sox9a and sox9b, whose combined and overlapping expression domains recapitulate that of the single Sox9 ancestral gene. At 3 dpf, both sox9a and sox9b mRNA transcripts are localized to the somites, pectoral fins, and neural crest cells of the pharyngeal arches (Cresko et ah, 2003). However, they have distinct expression domains in the developing brain where sox9a, but not sox9b, is expressed in the forebrain, while sox9b, but not sox9a, is expressed in the hindbrain (Cresko et ah, 2003). Their combined expression domains at 3 dpf recapitulate that of ancestral Sox9. Another example of subfunctionalization, this time of both spatial and temporal expression, is the sox21a and sox21b genes in zebrafish. During embryogenesis both sox21 genes are expressed in the hindbrain, spinal cord and ear (Lan et ah, 2001). Whereas sox21a is also expressed in the lateral line, olfactory placode, and at the midbrain hindbrain boundary, sox21b expression is restricted to the telencephalon, 6 hypothalamus, and lens (Lan et ah, 2001). While sox21a and sox21b retain their expression in the brain of adult zebrafish, sox21a is now expressed in the skin, ovary, and intestine whereas sox21b expression is found in the testis (Lan et ah, 2011). When assessing whether a duplicate gene pair has evolved by subfunctionalization it is important to have a comprehensive understanding of the functions and/or expression patterns of the single ancestral gene. Once this information is available, the spatial and temporal expression patters as well as the protein functions of each duplicate gene copy can be compared with that of the single ancestor. In the case of vertebrate UNC-45, if expression domains and/or apparent functions of unc45a and unc45b differ throughout development and during adulthood then this is an indication that UNC-45 has evolved by subfunctionalization in the vertebrate lineage. Finally, duplicate genes may be preserved if one copy accumulates mutations that lead to the development of a new and beneficial function (Lynch and Force, 2000). Neofunctionalization results from one gene being free from the constraints of positive selection and thereby able to accumulate mutations in the coding or regulatory sequences (Prince and Pickett, 2002). The other gene copy must retain the ancestral gene function. For example, teleosts have two copies of the human estrogen receptor gene: esr2a and esr2b. The two genes have diverged significantly in sequence from one another following their creation by whole genome duplication. While esr2b diverged rapidly from the ancestral sequence (as far as can be reconstructed) changes in the esr2a sequence have occurred much more slowly and so it retains the ancestral function (Hawkins et ah, 2000). Another example is the zebrafish duplicate aquaporin-1 genes, aqplaa and aqplab, that are important in the process of oocyte hydration (Zapater et ah, 2011). Although the role of aqplab has been conserved amongst teleosts, its carboxyl terminal domain that regulates the in vivo cytoplasmic trafficking of aquaporin-1 has degenerated such that a novel trafficking mechanism has evolved (Zapater et ah, 2011). It should be noted that neofunctionalization is rare and in no way does it represent the typical path of duplicate gene evolution (Prince and Pickett, 2002). 7

The protein products of duplicate genes are redundant when either can partially or fully substitute for the function of the other (Pickett and Meeks- Wagner, 1995). The study of redundancy between two proteins using genetics necessitates the creation of a double mutant and a phenotype comparison to that of its single loss-of-function mutants (Thomas, 1993). This in turn, however, requires at minimum, a detailed semi-quantitative characterization of the single mutant phenotypes (Perez-Perez et ah, 2009). Both the zebrafish unc45d/~ and unc45b~/~ mutants have been examined in detail and for most of the studies described in this thesis, published phenotypic information for at least one of the mutants was available (Wohlgemuth et ah, 2007a; Anderson et ah, 2008).

1.2 Why has the Zebrafish Become Such a Popular Model System?

The zebrafish (Danio rerio) is a teleost fish belonging to the cyprinid family, which is native to fresh waters of South Asia (Langeland and Kimmel, 1997; Dahm et ah, 2007). Other teleosts used as experimental models include the three-spined stickleback {Gasterosteus aculeatus), the pufferfish (Takifugu rubripes), and medaka {Oryzias latipes) (Ntisslein-Volhard et ah, 2002). Prior to the teleost radiation, genome tetraploidation arose as a result of a whole genome duplication event (Van de Peer et ah, 2002). The subsequent gene loss and functional specialization has led to approximately 20% of mammalian genes having two zebrafish orthologues, each with discrete expression domains and functions (Van de Peer et ah, 2002; Dahm et ah, 2007). Genome duplications are thought to have played a major role in the evolution of metazoans, vertebrates, and mammals by providing the material necessary for the development of novel gene functions (Taylor et ah, 2003). The retention of duplicate genes in zebrafish makes them an ideal system with which to study aspects of genome evolution such as redundancy within gene families and the conservation of regulatory elements associated with distinct expression domains (Taylor et ah, 2003). George Streisinger of the University of Oregon began working with the zebrafish during the 1960's with the aim of studying embryological development 8 using a vertebrate model system on which mutational analysis could be performed (Grunwald and Eisen, 2002). Since then, the zebrafish has become a popular model organism with which to study vertebrate development owing to a number of characteristics such as the production of large, transparent embryos that are fertilized externally, and undergo rapid development (Isogai et ah, 2001). These features facilitate genetic manipulation and real-time observation and imaging of developmental events and organ formation. Also, adults housed in an aquatics facility can be bred year-round, producing large clutches ideal for genetic analyses (Grunwald and Eisen, 2002). The virtue of using zebrafish for developmental studies in addition to other traditional model organisms such as Drosophila or Xenopus, is the availability of both genetic and embryological methods (Nusslein-Volhard et ah, 2002). Classic forward genetic screens have created thousands of zebrafish embryonic-lethal and zygotic-lethal mutants that cover an array of developmental processes (Grunwald and Eisen, 2002). Another advantage of the system for the study of cardiovascular, hematopoiesis, and vascular system development is that embryos can satisfy their oxygen needs through passive diffusion - because of their small size and external development - and thus can survive and continue development for several days without a functioning circulatory system (Chen et ah, 1996). This makes zebrafish an ideal model for cardiac research as it allows for the study of cardiovascular mutants to a much later stage of development than is possible with other vertebrate models that develop in utero.

1.3 Members of the UCS Protein Family are Present in Single- Celled Eukaryotes to Metazoans

Members of the UCS (UNC-45/CR01/She4p) domain-containing protein family are present in single-celled eukaryotes and metazoan animals (Hutagalung et ah, 2002). UCS proteins perform essential myosin- and actin-dependent functions by directly interacting with the myosin motor domain. This interaction is mediated by a conserved, 400 residue region (the UCS domain), located at the carboxyl-terminal of all UCS proteins (Hutagalung et ah, 2002; Yu and Bernstein, 9

2003). The four UCS family members are: UNC-45 (Caenorhabditis elegans and vertebrates), CROl (filamentous fungi, Podospora anserind), She4p (budding yeast, Saccharomyces cerevisiae), and Rng3p (fission yeast, Schizosaccharomyces pombe) (Yu and Bernstein, 2003).

1.3.1 Caenorhabditis elegans UNC-45

Epstein and Thomson (1974) were the first to describe the phenotype of an unc-45 temperature sensitive allele in the nematode, Caenorhabditis elegans. Worms raised at the restrictive temperature moved in an uncoordinated (Unc) manner, had disorganized myofibrils and fewer body wall muscle fibres compared to wild type worms. These observations indicated that UNC-45 is critical for both myofibril organization and muscle function and is essential for C. elegans development. It was not until later that recessive lethal alleles of C. elegans unc- 45 were isolated. Muscle contraction is reduced in these mutants and when bred to be genetically homozygous, embryos arrest during early development (Venolia and Waterston, 1990). To determine the spatial distribution of zygotic unc-45 transcripts, Venolia et ah (1999) placed a reporter gene under the control of the unc-45 promoter to drive reporter gene expression. They determined that zygotic unc-45 expression is restricted to all C. elegans muscle tissues: pharyngeal, body wall, vulval, and anal. In adult body wall muscles of wild type and temperature sensitive mutants, UNC-45 proteins co-localize with myosin heavy chain B (MHC B) (Ao and Pilgrim, 2000). The pattern is MHC B-dependent, as MHC B mutants do not exhibit UNC-45 localization, despite having increased levels of MHC A and normal thick filament organization (Ao and Pilgrim, 2000). If UNC-45 does not co-localize with MHC A, folding of the MHC A head domain must occur independently of UNC-45. In addition to zygotic UNC-45, maternally contributed protein, is present in all cells of the early embryo (Venolia and Waterston, 1990; Kachur et ah, 2004). Kachur et ah (2004) demonstrated that adults with reduced germline expression of UNC-45 produce offspring with cytokinesis defects, suggesting that maternally contributed UNC-45 has a critical role in early development. A direct 10 interaction between UNC-45 and NMY-2, a non-muscle type II myosin was subsequently identified from a yeast two-hybrid screen (Kachur et ah, 2004). Both colocalize in the early embryo at cellular boundaries, suggesting that UNC- 45 might be involved in myosin assembly and stability during cytokinesis (Kachur et ah, 2004). This discovery was significant as it was the first time a non-fungal myosin assembly protein was identified interacting with a class II non-muscle myosin. The C. elegans UNC-45 protein has a three-domain arrangement that is conserved in all animal homologues of UNC-45 (Venolia et ah, 1999; Barral et ah, 2002) (Figure 1-2). At the amino terminus are three tandem tetratricopeptide repeats (TPR) that facilitate the interaction between UNC-45 and Hsp90 (Venolia et ah, 1999; Barral et ah, 2002). TPR motifs serve to mediate protein-protein interactions and are composed of thirty-four amino acids arranged in three to sixteen consecutive repeats (Blatch and Lassie, 1999; Barral et ah, 2002). Proteins with TPR motifs can be found in single celled eukaryotes as well as plants and metazoan animals (Blatch and Lassie, 1999). TPR-containing proteins are not limited in distribution to specific regions within the cell and TPR motifs are present in numerous proteins, many of which share no functional relationship (Blatch and Lassie, 1999). UNC-45 TPR domains are similar to two other TPR- containing proteins: Hop (Hsp70/Hsp90-organizing protein) and protein phosphatase 5 (Barral et ah, 2002). At the carboxyl terminus of the protein is a domain that shares similarity with fungal proteins identified to function in the assembly of cytoplasmic myosin (Venolia et ah, 1999). The central region that joins the TPR and UCS domains acts in concert with the carboxyl terminal domain to bind and chaperone the myosin head domain (Venolia et ah, 1999; Barral et ah, 2002). Through the formation of a stoichiometric complex, UNC-45 mediates myosin folding by acting concurrently as a myosin chaperone and an Hsp90 co-chaperone (Barral et ah, 2002). Precise spatial and temporal regulation is necessary for thick filament assembly and the subsequent integration of sarcomere components to form myofibrils. Moreover, proteins associated with the numerous stages involved in 11 the process of muscle fibril organization are also under tight regulatory control. One way for cells to manage protein levels is via targeted protein degradation by the ubiquitin proteasome system. Proteins are marked for degradation by the covalent attachment of multiple ubiquitin moieties through the coordinated activities of three enzymes: El, ubiquitin-activating enzyme; E2, ubiquitin conjugating enzyme; and E3, ubiquitin ligase (Pickart, 2004). In some cases, an extra E4 enzyme is required to enhance the addition of ubiquitin onto a target substrate. Hoppe et ah (2004) identified two ubiquitin ligases responsible for the multiubiquitylation of UNC-45: CHN-1 and UFD-2. These proteins are the C. elegans homologues for the S. cerevisiae proteins CHIP (carboxyl terminus of Hsc70-interacting protein) and UFD2 (an E4 enzyme), respectively. On its own, CHN-1 has E3 activity and can link one to three ubiquitin polypeptides to UNC- 45 (Hoppe et ah, 2004). This level of ubiquitination, however, is not sufficient to target UNC-45 to the proteasome. Instead, UFD-2 and CHN-1 together form an E3/E4 complex that is capable of producing multiubiquitin conjugated UNC-45 (Hoppe et ah, 2004). A loss of chn-1 function in an unc-45 temperature sensitive mutant produces a mutant phenotype that is less severe than that of the unc-45 mutant alone. Conversely, if unc-45 is overexpressed in the same genetic system, disorganized thick filaments appear in body wall muscles (Hoppe et ah, 2004). Therefore, strict control of UNC-45 protein levels by the ubiquitin proteasome system results in both UNC-45 overexpression and loss of function to produce similar phenotypes characterized by muscle paralysis, myofibril disorganization, and decreased myosin quantities (Landsverk et ah, 2007). Consequently, in C. elegans at least, there is a connection between UNC-45 regulation, protein degradation, and myosin assembly.

1.3.2 Vertebrate UNC-45

All vertebrates examined have two UNC-45 homologues that are located on separate chromosomes and have distinct expression domains (Price et ah, 2002). Striated muscle UNC-45 (SMUNC-45 in mammals and Unc45b in zebrafish) is expressed exclusively in skeletal and cardiac muscles (Price et ah, 12

2002; Etard et ah, 2007; Wohlgemuth et ah, 2007), while general cell UNC-45 (GCUNC-45 in mammals and Unc45a in zebrafish) is ubiquitously expressed (Price et ah, 2002; Anderson et ah, 2008). Inconsistencies in the nomenclature of human Unc45 proteins are present in the literature. For reasons of clarity, throughout this text I will refer to all vertebrate UNC-45 proteins as Unc45a and Unc45b. UNC-45 proteins, from nematodes to humans, have common TPR, central, and UCS domain arrangements (Epping et ah, 2009). Zebrafish Unc45 isoforms are 55% identical to one another and approximately 30% identical to C. elegans UNC-45 (Table 1-1). Zebrafish Unc45a is 66% identical to human Unc45a and zebrafish Unc45b is 71% identical to human Unc45b. The whole genome duplication event that occurred prior to the teleost radiation is irrelevant to our study since mammals and zebrafish have single copies of the unc45a and unc45b genes.

1.3.3 Biochemical Analyses Demonstrate Functional Redundancy between Unc45a and Unc45b in vitro

Unc45a has a lower affinity for Hsp90 compared to Unc45b, but is more effective at folding smooth muscle myosin motor domains (Liu et ah, 2008). In an in vitro folding assay, Unc45a was more efficient than Unc45b at folding GFP- tagged smooth muscle myosin motor domains. Similar to the interaction between UNC-45 and NMY-2 in C. elegans, Unc45a function may be linked to non- muscle myosin as the smooth muscle myosin used by Liu et ah (2008) is closely related to non-muscle myosin type II. Vertebrates, unlike C. elegans and Drosophila melanogaster, have two cytosolic Hsp90 proteins: Hsp90a and Hsp90b. Studies designed to examine the mechanism and kinetics of Hsp90- and Unc45- mediated myosin folding are performed, for the most part, in vitro. The Hsp90 required in these experiments is obtained by protein purification or from rabbit reticulocyte lysates (Chadli et ah, 2008). One issue arising from the use of these protein sources is the indiscriminate and interchangeable manner with which 13 the two cytosolic Hsp90 proteins are used (Chadli et ah, 2008). The functional differences that exist between the two Hsp90 isoforms are not well understood. It is likely, however, that the isoform differences are dictated by their specific set of co-chaperones and client proteins (Chadli et ah, 2008). Unc45a is the first Hsp90 co-chaperone to demonstrate Hsp90a/b isoform selectivity (Chadli et ah, 2008). Unc45a specifically binds to Hsp90b and is required for its proper cellular distribution. Knockdown studies, performed using HeLa cells, resulted in morphological defects such as reduced cytoplasmic mass and cell shape changes. Hsp90 has been shown to interact with both actin and tubulin, leading Chadli et ah (2008) to speculate that Unc45a might interact with a subset of Hsp90 clients, and play a role in myosin cytoskeleton organization and cytoplasmic trafficking (Csermely et ah, 1998; Pratt et ah, 2004).

1.3.4 Unc45b and Hsp90 are Necessary for Myosin Motor Domain Folding and Assembly

Most studies examining vertebrate UNC-45 have been performed using the zebrafish model system. In zebrafish, unc45b is expressed in the skeletal, cardiac, and craniofacial muscles and contributes to myosin thick filament assembly (Wohlgemuth et ah, 2007). Prior to the identification of an unc45b~'~ mutant, Unc45b was characterized in embryos and early larvae by knocking down gene expression using antisense morpholino oligonucleotides (Etard et ah, 2007; Wohlgemuth et ah, 2007). Mutants and embryos injected with unc45b antisense MOs, have defective myofibrils containing disorganized thick and thin filaments as well as reduced levels of muscle myosin protein (Etard et ah, 2007; Wohlgemuth et ah, 2007). Consequently, fish are paralyzed, lack circulation and have extensive edema by 5 dpf. As in C. elegans, unc45b loss-of-function and overexpression produce similar phenotypes (Barral et ah, 2002; Bernick et ah, 2010). The mutant phenotype is believed to be the result of a null mutation as: wild type unc45b RNA injections rescue the mutant phenotype, translation blocking morpholinos recapitulate the mutant phenotype, and embryos that are heterozygous for unc45b have a wild type phenotype (Etard et ah, 2007; 14

Wohlgemuth et ah, 2007). An antibody against Unc45a could be used to determine whether the kurzschluss mutant has a null mutation, however, one is not available at the moment. Price et ah (2002) were the first to identify and map the Unc45a and Unc45b genes in humans. Unc45b expression is first detected during myoblast fusion with expression levels continuing to rise throughout muscle differentiation. Unc45a by contrast, is expressed in proliferating myoblasts prior to the onset of differentiation, at which time its expression begins to decrease (Price et ah, 2002). Antisense oligonucleotide treatments, directed towards Unc45b, were performed in vitro using C2C12 cells, a murine skeletal myogenic cell line. Treatments produced identical phenotypes to those that would later be observed in zebrafish, as myofibrils fail to assemble properly in the absence of Unc45 (Price et ah, 2002). Unc45b acts as a myosin chaperone and an Hsp90 co-chaperone in the same manner as C. elegans UNC-45 (Barral et ah, 2002; Srikakulam et ah, 2008). Srikakulam et ah (2008) demonstrated, in vitro, that Unc45b preferentially binds to partially folded myosin motor domains and promotes myosin folding by forming a stable complex with Hsp90. Effective motor domain folding was not contingent upon the presence of myosin rod domains, or the association of essential and regulatory light chains to the myosin heavy chain (Srikakulam et ah, 2008). unc45b and hsp90 alpha are thought to be co-regulated in vivo, as they both display increased levels of gene expression in the unc45b~/~ mutants (Etard et ah, 2007). Levels of hsp90ab.l remained constant in the mutants compared to wild type controls, as predicted, since hsp90ab.l expression is constitutive, not stress induced. Zebrafish have two Hsp90a genes: hsp90a.l and hsp90a.2 (Du et ah, 2008). In order to determine which Hsp90a protein interacts with Unc45b, Du et ah (2008) compared the knockdown phenotypes of both genes as well as their expression patterns. hsp90a.2 knockdown did not produce a muscle phenotype in zebrafish, indicating that Unc45b forms a complex with Hsp90a.l (Du et ah, 2008). Loss of hsp90a.l expression or protein function leads to the upregulation 15 of hsp90a.l and unc45b genes, with the simultaneous downregulation of genes encoding sarcomeric proteins (Du et ah, 2008; Hawkins et ah, 2010). Embryos injected with hsp90a.l antisense morpholinos develop into immotile larvae with disorganized thin and thick filaments and extensive myosin degradation (Du et ah, 2008). Surprisingly, hsp90a.l has no effect on cardiac muscle contraction suggesting that hsp90a.2 and hsp90ab.l may interact with Unc45b in the heart.

1.3.5 Analysis of Unc45a Function in vivo

The first study to examine the requirement of unc45a during vertebrate development began as a characterization of the vascular patterning defects found in the zebrafish unc45d/' mutant kurzschluss. It wasn't until later, however, when the kus mutant locus was mapped, that unc45a was identified as the gene responsible for the kus phenotype. The mutant phenotype is believed to be the result of a null mutation as: wild type unc45a RNA injections rescue the mutant phenotype, translation blocking morpholinos recapitulate the mutant phenotype, and embryos that are heterozygous for unc45a have a wild type phenotype (Anderson et ah, 2008). An antibody against Unc45a could be used to determine whether the kurzschluss mutant has a null mutation, however, one is not available at the moment. In wild type embryos at 48 hpf, unc45a transcripts are present in the brain, pharyngeal arches, and retina (Anderson et ah, 2008). Anderson et ah (2008) identified two phenotypes in the unc45d/~ mutants, both associated with the pharyngeal arch region. The most pronounced defect is an arteriovenous malformation involving aortic arches 5 and 6 that causes blood to shunt from the primary head sinus, back into the heart. The second phenotype is a spacing defect in the cartilages of the two most posterior pharyngeal arches (Anderson et ah, 2008). In contrast to unc45b'/~ mutants, myofibril organization and thick filament assembly appear normal. The nature of the mutant phenotypes is surprising given the chaperone activity of UNC-45 proteins, not to mention the broad distribution of unc45a transcripts. Anderson et ah (2008) hypothesized that Unc45a interacts with either a non-muscle or unconventional myosin in the pharyngeal arch region. 16

First, the lack of an observable muscle fibre phenotype suggested that Unc45a does not interact with a muscle myosin. Given that UNC-45 proteins at the time were known to act solely as myosin chaperones, the only plausible explanation was that Unc45a must chaperone other myosin isoforms. Second, unc45a develops an increasingly mosaic expression pattern as development proceeds and within the pharyngeal arches, unc45a expression is not localized to the muscle precursors. unc45a is expressed instead by the endoderm cells in the pharyngeal arch region (Beth Roman, pers comm.). Once again, Unc45a must interact with a myosin isoform, other than muscle myosin II, if its function is to chaperone myosin molecules, as is the case of Unc45b.

1.3.6 Do Vertebrate Unc45 Proteins have Functions Distinct from Myosin Chaperoning?

UNC-45 proteins were originally studied for their function in myosin assembly and myosin-related processes. Recently, mammalian UNC-45 proteins have been identified through various screening methods to have novel interaction partners. Apobec2, a putative cytidine deaminase, was identified in a yeast two- hybrid screen to interact with Unc45b. Unc45b and Apo2 proteins, acting independently from Hsp90, participate in the attachment of myofibrils to the myosepta and help to maintain structural stability (Etard et al., 2010). The mechanism through which Unc45b and Apo2 assist in myofibril attachment remains unknown. Morpholino knockdown of apo2a and apo2b mRNA, results in defective muscle organization and decreased cardiac function, phenotypes shared with the unc45b~/~ mutant. Despite the connection with muscle fibre organization, this newly ascribed role marks a departure of Unc45b function from myosin motor domain folding. The activities of Unc45a are also not limited to myosin related processes. Unc45a can interact with progesterone receptors A and B and is a novel regulator of progesterone receptor chaperoning by Hsp90 (Chadli et al., 2006). The progesterone receptor is released from its complex with Hsp90, following a 17 conformational change initiated by hormone-receptor binding (Chadli et ah, 2006). Once released, the progesterone receptor joins with transcription factors and coactivators to associate with the progestin response elements of target genes (Tsai et ah, 1994). Since Unc45a is a cytosolic protein, it does not co-localize with the activated progesterone receptor, which is translocated into the nucleus (Chadli et ah, 2006). The ATPase activity of Hsp90 is regulated by two proteins: Ahal and Hop. When ATP is present in the ATP binding pocket of Hsp90, Ahal enhances the naturally low level of Hsp90 ATPase activity. Hop can then oppose the effects of Ahal by either blocking ATP hydrolysis, or preventing future ATP binding by interacting with Hsp90 proteins that don't contain ATP in their binding pockets. Even though GCUNC-45 does not directly affect the ATPase activity of Hsp90, nor does it have any ATPase activity of its own, it can regulate Hsp90 ATPase activity almost as well as Hop (Chadli et ah, 2006). By influencing Hsp90 ATPase activity indirectly, Unc45 promotes progesterone function in the cell by reducing progesterone receptor chaperoning. Most Hsp90 co-chaperones interact directly with Hsp90 through the binding of their TPR motifs to the MEEVD sequence situated at the carboxyl terminal of Hsp90 (Figure 1-2). Unc45 and the Hop co-factor, FKBP52, act antagonistically to influence progesterone receptor activity by binding to a novel TPR site towards the amino-terminal of Hsp90 (Chadli et ah, 2008). The discovery of two TPR recognition sites situated at opposing ends of the Hsp90 protein might facilitate multiple simultaneous TPR protein interactions (Chadli et ah, 2008). In addition to its role in modulating progesterone receptor/Hsp90 chaperoning, Unc45 has been linked to cancer progression through its overexpression and ability to confer resistance to histone deacetylase inhibitors and retinoic acid. Overexpression of unc45a is associated with ovarian cancer cell proliferation and metastasis (Bazzaro et ah, 2009). In vitro overexpression of unc45a leads to increased cell proliferation and an accumulation of non-muscle myosin and Unc45a at the cleavage furrow during cytokinesis. Both proteins also 18 localize to the filopodia of motile cells. In humans, there is a positive correlation between levels of Unc45 and the stage and grade of ovarian cancer. This can be attributed, in part, to the elevated levels of Unc45a protein in ovarian carcinoma tumours compared to healthy ovarian epithelium (Bazzaro et ah, 2009). Cancer survival rates are linked to the efficacy of available treatments, but patients often respond differently to the same drug. Epping et ah (2009) set out to identify potential markers of therapy response by performing a genetic screen focusing on genes that confer resistance to histone deacetylase inhibitors. They were interested in histone deacetylase inhibitors, as they are potential anticancer drugs that specifically promote tumour cell cytotoxicity (Minucci et ah, 2006). Not only does Unc45a make cells resistant to HDACI, but it also appears to have numerous roles in retinoic acid signalling. For example, it was shown to inhibit signalling through the retinoic acid receptor alpha. Furthermore, Unc45a induces expression of endogenous retinoic acid receptor target genes while inhibiting the retinoic acid induced arrest in proliferation and differentiation of neuroblastoma cells.

1.3.7 Fungal UCS Proteins

The phenotypes of UNC-45 and fungal UCS protein mutants vary considerably, but they all have an effect on myosin assembly and/or function (Hutagalung et ah, 2002; Shi and Blobel, 2010). Consequently, structures that incorporate myosin molecules are also affected in these mutants. Phenotypic differences may be ascribed to the myosin classes with which fungal UCS proteins associate as well as their different domain structure. In contrast to UNC- 45, fungal UCS proteins interact with both conventional and non-conventional myosin, leading to their participation in numerous cellular functions (Hutagalung et ah, 2002). The amino terminal TPR domain is absent in fungi and the central domain, if present, shares little homology with UNC-45 or other fungal proteins (Shi and Blobel, 2010). The UCS domain, however, is highly conserved amongst species, with approximately 53% similarity between She4p and human UNC-45 proteins (Shi and Blobel, 2010). 19

1.3.8 Crol in P. anserina

CROl is the UCS protein present in the filamentous fungus, Podospora anserina. Essential for actin cytoskeleton organization and function, CROl is the least studied of all the UCS proteins. The mutant null allele cro-1 produces pleiotropic phenotypes such as a decrease in filamentous growth and the absence of septa between daughter nuclei following mitotic division (Berteaux-Lecellier et ah, 1998).

1.3.9 She4p in S. cerevisiae

She4p is involved in a number of processes requiring myosin function such as mRNA and myosin localization, actin polymerization, and endocytosis (Toi et ah, 2003; Wesche et ah, 2003). Interactions with non-conventional proteins from myosin classes I and V differentiate She4p from the other UCS proteins (Wesche et ah, 2003). For example, She4p is necessary for class I myosin-actin binding. Recently, the She4p UCS domain was shown to link two myosin heads at their motor domains, thus acting as a determinant of the step size of the myosin motor along the F-actin thin filaments (Shi and Blobel, 2010). A yeast two-hybrid screen conducted with the purpose of finding proteins that interact with the UCS domain of She4p resulted in the identification of fourteen non-muscle proteins (Wesche et ah, 2003). This suggests, that similar to the discovery of an Unc45b and Apobec2 interaction, UCS proteins may participate in a broader range of cellular functions than was thought previous to 2003.

1.3.10 Rng3p in S. pombe

Rng3p is necessary for the formation and function of the actomyosin cytokinetic ring (Wong et ah, 2000; Yu and Bernstein, 2003). It is thought to act as a template for ring formation by maintaining Myo2p, the essential myosin heavy chain of the actomyosin cytokinetic ring, in an assembly-competent state (Wong et ah, 2002). Myo2p function is regulated through the coordinated 20 activities of Rng3p and Swolp (Hsp90) (Mishra et ah, 2005). An inability to produce a functional cytokinetic ring, and the subsequent failure to carry out cytokinesis culminates in the production of multinucleated cells in rng3p null mutants (Wong et ah, 2000).

1.4 Folding of the Myosin Motor Domain into its Native Conformation is a Requirement for Thick Filament Assembly and Muscle Function

1.4.1 Myosins are a Large Superfamily with Diverse Cellular Functions

Myosins are actin based, ATP-dependent motor proteins implicated in numerous cellular processes such as muscle contraction, cytokinesis, and cellular trafficking (Landsverk and Epstein, 2005). Found in all eukaryotes, myosin proteins form a superfamily totalling at least 18 classes, each correlated with distinct functions (Vikstrom et ah, 1997). Two domains shared by all myosins are an amino-terminal head and a carboxyl-terminal tail (Figure 1-3). The head contains domains that hydrolyze ATP and bind F-actin, while the tail region performs class-specific functions dictated by differences in tail sequence and length (Landsverk and Epstein, 2005). The activities of these domains combined contribute to force generation and movement. Besides grouping by class, the myosin superfamily has traditionally been subdivided into two categories: conventional and non-conventional. Muscle and non-muscle myosin II, together referred to as conventional myosin, participate in activities such as muscle contraction, cell motility, cytokinesis, and maintenance of cell shape (Yu and Bernstein, 2003). Class II myosins are the most widely studied and form filamentous structures through the self-association of a-helical rod domains, a feature unique to this class (Landsverk and Epstein, 2005). Non- muscle myosin II is similar to smooth muscle myosin with regard to sequence homology, subunit composition, and regulation (Sellers and Knight, 2007). All remaining myosin classes are classified as unconventional. Unconventional myosins participate in a number of functions crucial to the cell such as 21 endocytosis, vesicle transport, cell migration, and actin filament formation (Yu and Bernstein, 2003). They are distinguished from the conventional myosins based on sequence, subunit composition, and kinetics (Sellers and Knight, 2007).

1.4.2 Myosin Molecules Assemble into Thick Filaments

Myosin II is a hexameric protein formed from two paired heavy chains, each with an associated essential and regulatory light chain (Landsverk and Epstein, 2005) (Figure 1-4). The myosin heavy chain has three domains: a head, neck and rod. Situated at the amino-terminal is the globular motor domain that contains the actin and ATP binding sites. Adjacent is the neck region that acts as a lever for the motor domain during contraction and associates with the regulatory and essential light chains. The remaining carboxyl-termini of the paired myosin heavy chains associate to form an a-helical coiled-coil rod that produces the characteristic filament structure of class II myosins (Lowey and Trybus, 2010). Myosin II molecules in striated muscle cells associate to form structured multiprotein complexes, or thick filaments. In vertebrates, these filaments are further organized into a hexagonal lattice (Agarkova and Perriard, 2005). Simple lattice structures such as those found in bony fish are constructed with adjacent myosin filaments aligned in the same orientation along the long axis of the cell (Agarkova and Perriard, 2005). Higher vertebrates have a much more intricate lattice, in part to maximize the tension generated in the thick filaments. The network of thick filaments is aligned into a bipolar array. At the centre, free from motor domains, rods interact in an antiparallel fashion whereas at the motor rich periphery, adjacent filaments are aligned in parallel (Craig and Woodhead, 2006) (Figure 1-4). Assembly of myosin into myofibrils is an intricate process requiring three stages: folding and assembly of the myosin molecule, subsequent assembly into filamentous structures and incorporation of thick filaments into myofibrils (Srikakulam and Winkelmann, 2003). The unique capacity of myosin II to form ordered filaments is essential for its molecular function and requires the entire myosin molecule (Vikstrom et ah, 1997; Landsverk and Epstein, 2005). Errors in 22 any of the above stages will result in disorganized myofibrils and a failure to form ordered sarcomeres.

1.4.3 Myosin Folding, Thick Filament Assembly, and Sarcomere Organization Require Chaperones

Molecular chaperones often assist in the folding of newly synthesized or misfolded proteins to prevent the formation of dangerous protein aggregates (Picard, 2002). All chaperones have two functional characteristics: they act to prevent the aggregation of partially unfolded proteins, and they conserve partially unfolded proteins in a folding competent state (Barral et ah, 2002). In striated muscle cells, the rate-limiting step of the myosin folding and assembly pathway is the chaperone-mediated folding of the catalytic head domain (Srikakulam and Winkelmann, 1999) (Figure 1-5). The hypothesis that additional factors might be required for complete motor domain folding originated from a failure to produce functionally active recombinant motor domain proteins in in vitro bacterial expression systems. Functional motor protein can only be produced when recombinant proteins are expressed in muscle systems or supplemented with muscle lysate (Srikakulam and Winkelmann, 1999). Other stages in the myosin assembly pathway, such as the dimerization of the a-helical coiled-coil rod domain and the association of the light chains with the neck region, are not affected in the bacterial expression system, as these events occur autonomously and precede motor domain folding (Srikakulam and Winkelmann, 1999). Therefore, muscle cell specific folding factors must be present to assist in the folding of the head domain. Srikakulam and Winkelmann (2003) identified two proteins, Hsp90 (heat shock protein, 90 kDa) and Hsc70 (constitutively expressed heat shock related protein) that form a transient complex with the partially folded myosin motor domain. Hsp90 is an abundant and highly conserved protein that interacts with a range of substrates, termed clients, in an ATP-dependent manner (Lai et ah, 1984; Picard, 2002). As a molecular chaperone, Hsp90 can prohibit the aggregation of 23 non-native proteins and promote their refolding (Wiech et al, 1992; Buchner, 1999). By forming stable associations with proteins, Hsp90 is also able to prevent unfolding or aggregation of native proteins (Picard, 2002). Co-chaperones work with Hsp90 to influence substrate recognition, chaperone function, and ATP binding and hydrolysis (Picard, 2002; Zuehlke and Johnson, 2009). The interaction of Hsp90 with its more than 100 mammalian co- chaperones has led Hsp90 to be involved in, and associated with, a diverse range of cellular processes such as: signal transduction, cell growth and differentiation, cellular trafficking, and chromosome remodelling (Picard, 2002; Zuehlke and Johnson, 2009). Most Hsp90 co-chaperones interact with Hsp90 through a conserved tetratricopeptide repeat (TPR) motif (Ramsey et ah, 2002; Chandli et ah, 2008). TPR motifs mediate protein-protein interactions and in the case of Hsp90 co-chaperones, the TPR domain interacts with the carboxyl terminal MEEVD domain of Hsp90 (Allan and Ratajczak, 2011). Although Hsp90 and Hsc70 associate with partially folded myosin motor domain intermediates, their activities alone are insufficient to produce a native myosin protein (Srikakulam and Winkelmann, 2003). This indicates that additional muscle-specific factors must be involved in the final stages of myosin folding and assembly.

1.4.4 Sarcomeres are the Basic Contractile Unit of Striated Muscle

Sarcomeres are the basic contractile unit of striated muscle. They have a characteristic structure of interdigitated thin (actin) and thick (myosin) filaments bordered by Z-disks (Figure 1-6). Sarcomeres measure approximately 2.5 um in length in vertebrate skeletal and cardiac muscle, but differ significantly in size between invertebrate species (Sanger et ah, 2010). Joined repeatedly end-to-end, sarcomeres form parallel myofibrils that span the length of the muscle cell and account for a large part of the muscle cytoplasm. It is the reiterated pattern of sarcomeres that gives myofibrils their striped, or striated, appearance. The sarcomere cytoskeleton is a protein scaffold that maintains the ordered structure of the contractile filaments and optimizes the transmission of the 24 forces they generate (Figure 1-6). Scaffold components indentified so far, include: Z-disks, a-actinin, myomesin, the M-band, and titin. Z-disks specify the lateral borders of the sarcomere and anchor titin and antiparallel thin filaments from adjacent sarcomeres (Agarkova and Perriard, 2005). Actin filaments are linked to the Z-disk by a-actinin and myomesin, an analogous protein, links thick filaments to the M-band. The M-band is a dynamic region at the centre of the sarcomere. It is responsible for monitoring stress development in the thick filament lattice during contraction. The large protein titin spans half the length of the sarcomere, from the Z-disk to the central M-band, and connects with both the thin and thick filaments. It may also be involved in defining the resting sarcomere length as well as maintaining the position of thick filaments at the centre of the sarcomere. Muscle contractions are produced as the result of a decrease in sarcomere length caused by the movement of opposing actin filaments along myosin filaments towards the centre of the sarcomere. The length of the contractile filaments, however, remains unchanged. Regulating contraction cycles in striated muscle cells is achieved through the strict control of cellular calcium levels and the conformational states of the calcium-sensitive regulatory proteins troponin and tropomyosin (Sellers and Knight, 2007). Smooth muscle cells employ a different mechanism to control contractile activity whereby the myosin regulatory light chain is phosphorylated by the calcium-calmodulin dependent enzyme myosin light chain kinase (Sellers and Knight, 2007).

1.5 Zebrafish have Three Striated Muscle Populations

1.5.1 Cardiac Muscle: Circulation

During early vertebrate development, the heart is the first organ to form and function (Boogerd et ah, 2009). For species that do not develop in an aquatic environment, and therefore cannot survive on passive diffusion of oxygen, heart function is vital for continued embryonic development (Stainier et ah, 1996). The zebrafish heart develops rapidly and by 24 hpf is comparable to the human heart at embryonic day 23 (Warren et ah, 2000). Cardiogenesis can be divided into four 25 stages: creation of a linear heart tube, patterning along the anterior-posterior axis, looping morphogenesis, and valve formation (Stainier et al, 1996) (Figure 1-7). Cardiac progenitor cells originate from the anterior lateral plate mesoderm and involute early during gastrulation (Stainier et ah, 1996; Stainier, 2001). Myogenic progenitor cells organize into two tubes bilateral to the midline, with the endocardial progenitors situated medially (Stainier et ah, 1996). The myocardial progenitors migrate towards the embryonic axis and fuse to form a concentric tube, with inner endocardial (endoderm) and outer myocardial (mesoderm) layers, separated by cardiac jelly (Chen et ah, 1996; Stainier, 2001; Boogerd et ah, 2009). Peristaltic waves of contraction begin once the transient heart tube is formed and circulation commences approximately two hours later at 24 hpf (Warren et ah, 2000). The initial peristaltic waves progress into rhythmic cardiac contractions as the embryonic heart develops (Stainier et ah, 1996; Warren et ah, 2000; Boogerd et ah, 2009). Cardiac chambers have yet to become morphologically distinct at 24 hpf, but they express different myosin heavy chain genes (Stainier et ah, 1996; Yelon et ah, 1999). Expression of vmhc is restricted to the ventricle, amhc to the atrium, while cmlc2, encoding a cardiac myosin light chain, is expressed in both. The heart tube undergoes looping morphogenesis moving the atrium to the right-hand side of the embryo and positioning it above the ventricle (Chen et ah, 1996). Now the chambers not only have differential gene expression, but also distinct morphologies (Chen et ah, 1996; Schoenebeck and Yelon, 2007). By 36 hpf looping is complete and the heart has four chambers: sinus venosus, the cardiac pacemaker; atrium; ventricle; and bulbus arteriosus, the outflow tract (Chen et ah, 1996; Stainier et ah, 1996; Lohr and Yost, 2000; Warren et ah, 2000; Boogerd et ah, 2009). Blood flow is directed from the trunk into the duct of Cuvier (future common cardinal vein) and into the sinus venosus. It then moves through the atrial and ventricular chambers passing through the bulbous arteriosus before exiting the heart and on to the ventral aorta (Isogai et ah, 2001). Heart valve formation is the final stage in the development of the zebrafish embryonic heart and occurs prior to hatching at around 48 hpf. Valves are 26 established at the cardiac chamber boundaries. The alternating contraction of the atrial and ventricular chambers, in conjunction with the rhythmic opening and closing of the valves, prevents retrograde flow and ensures the unidirectional movement of blood through the heart (Stainier et ah, 1996; Boogerd et ah, 2009). Cardiac and skeletal muscles have many structural features in common, yet they originate from spatially distinct mesoderm populations and differentiate using separate myogenic regulatory factors (Stainier, 2001). The homeodomain transcription factor nkx2.5 is the earliest known cardiac marker (Yelon et ah, 1999). Its expression is initially limited to the cardiac progenitor cells of the anterior lateral plate mesoderm, but later expands throughout the myocardium (Yelon et ah, 1999). Originally identified based on its homology with the Drosophila mutant tinman gene, nkx2.5 expression has been detected in the differentiating myocardium of all vertebrates that have been sampled (Stainier, 2001). Functional redundancy between Nkx2.5 and other Nkx family members may explain why nkx2.5 mutations in humans and mice result in incomplete or improper cardiac looping; a significantly milder phenotype compared to the Drosophila tinman mutant that has no heart (Stainier et ah, 1996). Expression of nkx2.5 is initiated at the end of gastrulation by the Gata5 transcription factor and then maintained through the activity of another transcription factor, Fgf8 (Stainier et ah, 1996). As gata5 is expressed in future myocardial progenitors before the onset of gastrulation, it is thought to function in regulating myocardial differentiation (Stainier et ah, 1996). While Gata5 has a role in myocardial differentiation, the basic helix-loop-helix transcription factor Hand2 regulates both cardiac differentiation and morphogenesis (Stainier, 2001). Myocardial cells begin to express terminal differentiation genes such as troponin T, myosin, and tropomyosin once the bilateral myocardial progenitors fuse at the midline, creating the embryonic heart tube. 27

1.5.2 Skeletal Muscle of the Trunk and Tail: Locomotion

Vertebrate skeletal muscle is derived from mesoderm precursor cells situated in somites (Hinitis et ah, 2009). In zebrafish, thirty to thirty-four pairs of somites are formed in a rostro-caudal sequence through the segmentation of paraxial mesoderm that flanks either side of the notochord (Molkentin and Olson, 1996; Schilling, 2002; Ochi and Westerfield, 2007). Muscle formation is a multi stage process involving the specification of precursor cells to a muscle fate, the fusion of myoblasts to form myotubes, and the transcriptional activation of structural genes for the assembly and function of myofibres (Molkentin and Olson, 1996; Lin et ah, 2006; Hinitis et ah, 2009). Four basic helix-loop-helix myogenic regulatory factors - MyoD, Myf5, myogenin, and MRF4 - function throughout muscle development to initiate myogenesis and promote muscle determination and differentiation (Molkentin and Olson, 1996; Ochi and Westerfield, 2007). Myogenic regulatory factors are critical for muscle development because when absent, muscle progenitors remain multipotent (Lin et ah, 2006). Myogenesis is initiated during convergence and extension by the myogenic regulatory factors MyoD and Myf5 whose expression is first detected at 70-75% and 80% epiboly, respectively (Ochi and Westerfield, 2007). Four muscle fibre groups, each having a specific morphology, are produced in every somite (Ochi and Westerfield, 2007; Hinitis et ah, 2009). This distinction arises from the division of each somite into medial and lateral domains that are both receptive and responsive to different induction signals (Molkentin and Olson, 1996; Ochi and Westerfield, 2007). The four fibre groups are: slow, fast, medial fast, and muscle pioneer (Hinitis et ah, 2009). Muscle pioneers are positioned at the dorsoventral midline and are distinct from the other three fibres. Cells of the future slow and fast muscle fibres express both the myogenic regulatory factors myogenin and mrf4 (Ochi and Westerfield, 2007). Slow muscle fibres originate from hedgehog- dependent adaxial cells that express both MyoD and Myf5 (Hinitis et ah, 2009) These cells are situated beside the notochord and will migrate laterally within the somite, localizing the slow muscle fibres superficially under the skin (Hinitis et ah, 2009). Paraxial cells from lateral positions of the somite give rise to two types 28 of fast muscle fibre. Adjacent and medial to the slow muscle fibres are the Fgf8- dependent fast fibres, and closest to the notochord are the Fgf8-independent medial fast fibres. By necessity, the process of myogenesis must progress quickly as zebrafish larvae must be able to sense and respond to their environment in order to survive when they hatch from their chorions around 48-72 hpf. Prior to hatching, embryos react to direct, tactile stimuli by producing a directional fast escape response (Granato et ah, 1996). Development to this point is rapid as the first muscle contractions in the trunk and tail occur spontaneously at 18 hpf and by six hours later, embryos are twitching their tails from side to side in response to general touch stimuli (Granato et ah, 1996).

1.5.3 Craniofacial Muscle: Feeding, Respiration and Ocular Movement

The pharyngeal arches of zebrafish embryos begin to develop in concert with their associated muscles around 2-3 dpf (Schilling et ah, 1996). Development progresses in an anterior to posterior direction and when complete, both segments represent a significant portion of the cranial musculoskeleton (Noden and Trainor, 2005). Cranial cartilage and muscle are believed to share a close connection with regards to development, function, and evolution. Disruptions in skeletal morphology result in abnormalities in their associated muscles, suggesting that the morphology of the two may be linked (Knight et ah, 2011). Similarly, differences in the configuration and function of the cranial musculature between mammals and fish are representative of the changes that have occurred in these structures following the divergence of these two lineages (Knight ef a/., 2011). Two components of the adult zebrafish cranial skeleton emanate from cartilage intermediates present by early larval stages. These are the seven pharyngeal arches and the neurocranium, which encases the brain, optic capsule, and surrounding anterior notochord (Schilling et ah, 1996). The first mandibular arch forms the jaw; the second hyoid, the jaw support (Yelick and Schilling, 2002; Javidan and Schilling, 2004). Four branchial arches (pharyngeal arches three to six) support the gills, and a fifth arch is associated with the pharyngeal teeth 29

(Yelick and Schilling, 2002; Javidan and Schilling, 2004) (Figure 1-8). The large size of the mandibular and hyoid arches, in contrast to the small posterior arches, is thought to reflect an evolutionary modification of a primitive pattern similar to that of the branchial arches (Schilling et ah, 1996). Shaped like inverted cones that curve and taper to a rounded tip, all seven pharyngeal arches are three-layered structures. The core is made of mesoderm and is enclosed by two concentric layers: an inner mesenchyme, derived from cranial neural crest, and an outer ectoderm (Talbot et ah, 2010) (Figure 1-7). Patterning of the pharyngeal arches into distinct segments (mandible, hyoid, and branchial) occurs as a result of three streams of cranial neural crest cells migrating into the region (Yelick and Schilling, 2002; Hernandez et ah, 2005). Each stream populates a specific segment as dictated by its origin along the anterior-posterior axis of the hindbrain (Hernandez et ah, 2005). Facial muscles are derived from cranial paraxial mesoderm and are involved in respiration, feeding, and ocular movement (Hernandez et ah, 2005; Chuang et ah, 2010). Together, six pairs of extraocular muscles and at least thirty pharyngeal arch muscles form the craniofacial musculature (Schilling, 2002). Muscles of the first and second pharyngeal arch open and close the pharyngeal cavity, those associated with the first four branchial arches move and support the gills, and muscles of the fifth branchial arch facilitate feeding (Schilling, 2002; Hernandez et ah, 2005; Knight et ah, 2011). Myogenesis progresses from the anterior to the posterior of the pharyngeal cavity in coordination with the adjacent cartilages (Schilling and Kimmel, 1997). Cartilages differentiate somewhat earlier than muscles within each pharyngeal region. Extraocular muscles are the first cranial muscles to begin myogenesis as by 72 hpf they are required for larvae to be able to track and catch prey (Easter and Nicola, 1996; Schilling, 2002). In addition to having separate origins, the regulation of craniofacial muscle development differs from that of skeletal muscle (Dong et ah, 2006). The myogenic regulatory factors Myod and Myf5 are expressed in the cranial paraxial mesoderm and demonstrate the same sequential temporal expression patterns as in skeletal muscle precursor cells. During murine 30 cranial myogenesis, Myod and Myf5 act redundantly, but only Myod is required for myogenesis in zebrafish (Hinitis et ah, 2009). Another myogenic regulatory factor, Mrf4, is critical for skeletal muscle development, but may not be essential during craniofacial myogenesis (Kassar-Duchossoy et ah, 2004).

1.6 Summary of Project Objectives

This project has two objectives: to assess the phenotype of a zygotic- lethal, unc45b~'~; unc45d'~ mutant, and to determine whether Unc45a and Unc45b are functionally redundant in vivo. The motive for creating and examining an unc45b~/~; unc45d/' mutant stems from the results of previous studies demonstrating an interaction between UNC-45 and non-muscle myosin, an association between UNC-45 and defective cytokinesis, and functional redundancy between vertebrate Unc45a and Unc45b in vitro. Therefore, we wanted to know if the same was also true for the vertebrate Unc45 proteins in vivo. Both objectives were addressed using the zebrafish (Danio rerio) vertebrate model. The retention of duplicate genes in zebrafish makes them an ideal system with which to study aspects of genome evolution, such as redundancy within gene families (Taylor et ah, 2003). Furthermore, as a vertebrate model system, zebrafish are ideal for studying early developmental events and for performing genetic analyses. I performed a thorough phenotypic characterization of the unc45b"; unc45dl~ mutant paying particular attention to abnormalities identified previously in the single mutants. By focusing on traits that had already been reported, I was able to determine whether the unc45b"; unc45d~ embryos have novel and/or more severe phenotypes than those of the single mutants, indicative of functional redundancy between the two vertebrate Unc45 proteins. If, however, the Unc45 proteins do not have identical molecular targets, then despite their functional redundancy in vitro, the unc45b~'; unc45d ' mutant would manifest a combination of the unc45d/' and unc45b~'~ phenotypes. 31

The phenotypic characterization of the unc45b"; unc45a" mutant centered on three areas: morphology, gene compensation, and gene expression and differentiation. I examined traits that were shown to be perturbed in the single mutants, namely: gross morphology; cardiac morphology and differentiation; muscle fibre determination and myofibril organization; jaw formation; and eye size. 32

1.7 Table

Table 1-1. Molecular Conservation Amongst UCS Protein Family Members

~ _ . i Protein % Protein % Gene Pair Identity Similarity2

C elegans UNC-45 and D. melanogaster UNC-45 31 29 C elegans UNC-45 and zebrafish Unc45a 33 26 C elegans UNC-45 and zebrafish Unc45b 32 26 Zebrafish Unc45a and zebrafish Unc45b 55 22 Zebrafish Unc45a and human Unc45a 66 17 Zebrafish Unc45b and human Unc45b 71 16

1. The RefSeq accession numbers for the sequences are: C. elegans UNC-45, NP_497205; D. melanogaster UNC-45, NP_524796; zebrafish Unc45a, NP001017671; zebrafish Unc45b, NP_705959; human Unc45a, NP_001034764; human Unc45b, NP_775259. 2. Similarity values were calculated using the following similarity scheme: (AVFPMILW), (DE), (RK), (STYHCNGQ). no B Duplication 8

\

\nJ\0 I L.

Nonfunctionalization Subfunctionalization Neofunctionalization Figure 1-1. Three potential fates of duplicate gene pairs. Circles denote regulatory elements and rectangles denote transcribed regions. On the left, one gene copy accumulates null mutations becoming nonfunctional. In the centre, both copies lose functions in non-overlapping protein domains. Both copies are essential to produce the ancestral gene function. On the right, one copy acquires a new function. If this new function differs significantly from the original, the other copy is maintained. Adapted from Hahn, 2009. Figure 1-2. UNC-45 domain organization. UNC-45 belongs to the UCS (UNC-45/CR01/She4p) family. The three-domain arrangement is conserved in all animal homologues of UNC-45. At the amino-terminus is the TPR domain that mediates a protein-protein interaction between UNC-45 and the MEEVD motif of Hsp90. At the carboxyl-terminus is the UCS domain that is homologus with all UCS proteins. The central region and the UCS domain bind to and chaperone the myosin motor domain (red and green circles). Light chains are depicted as grey and peach rectangles. 35

Class Function

Membrane Binding

Filament Sliding

II

Vesicle Transport

Figure 1-3. Myosin Classes. Myosins are actin based, ATP-dependent, motor proteins implicated in numerous cellular processes. All myosin proteins have an amino-terminal head domain and a carboxyl-terminal tail domain. The head domain binds actin and hydrolyses ATP while the tail domain performs class- specific functions dictated by differences in tail sequence and length. Adapted from Lodish et ah, 2000. 36

Figure 1-4. Myosin and thick filament organization. Hexameric myosin molecule: myosin heavy chains (red and green), essential light chains (blue), and regulatory light chains (yellow) (A). Arrangement of the bipolar thick filaments in the sarcomere (B). Thick filament lattice at the M-band. Myomesin (red), titin (yellow), and myosin (blue) (C). Adapted from Agarkova and Perriard, 2005; Craig and Woodhead, 2006. 37

•Q € <6

8- *6 I -=p-

-cA-

' 1-^.ii-L^.iJJ. •^^^^^B MM

-cP-

-cP-

-cP-

Figure 1-5. Myosin folding pathway model. Myosin folding and assembly is a three step process: myosin folding and dimerization; thick filament assembly; and incorporation of thick filaments into the sarcomere architecture. Nascent myosin molecules associate with essential and regulatory light chains (A). Alpha-helical rod domains dimerize and the chaperones Hsp90, Unc45b, Hsc70 (not shown) interact with the unfolded myosin head domain. The resulting myosin maturation complexes are transient structures and the pathway can be blocked at this step when chaperone activity is inhibited (B). Myosin molecules assemble into thick filaments (C). Thick filaments are incorporated into the sarcomere and assembled into a lattice structure (D). Z-disk M-band QQQQQQQQQQQQQQQ OOPDP ODDDDDDDDD -^ £3 •^A OOOOOOOOOO 0 0 000 t>t)t)t)t) ut)OOOm)DOu QQQQQQQQQQQQQQQ DDDDPDDDDDDDDDD •^7- T>A- QOQQQQOQQQQQQQQ OQDD^OuuuuOODD QQQQQQQQQQ QQQQQ PPPPP PPDDDDDPPP "^ ^p- o o oooo o ooo o o ooo QOOOOODMOOOOW QQQQQQQQQQQQQQQ PPDDDDDD •AIT- QQQQQQQQQQQQQQQ OOoWooo gOOoWD QQQQQQQQQQQQQQQ PPPPPDO •ACT -^A QQQQQQQQQQ0 0OOO tjygbt) uu QQQQQQQQQQ QQQQQ PPPPPpppppppppp -^7 •^ Actin FiiamenPoa^ao a ace a a aaa ODDoOODuOOODOuQ QQQQQQQQQQQQQQQ £>OD -^r OOOOOOOOOO0 0 000 Ot5D^ Titin Myosin Filament Figure 1-6. Sarcomere organization. The sarcomere is the basic contractile unit of striated muscle. Interdigitated actin (thin) and myosin (thick) filaments are bordered by Z-disks. The sarcomere cytoskeleton maintains the ordered structure of the contractile filaments and optimizes the transmission of the forces they generate. It consists of Z-disks, M-band, titin, a- actinin, and myomesin. 00 Cardiac Heart Tube Chamber Looping Valve Precursors Fusion Formation Morphogenesis Formation

Figure 1-7. Zebrafish cardiac development. By 12 hpf myogenic progenitor cells organize into two tubes bilateral to the midline with endocardial precursors situated medially. The myocardial progenitors migrate towards the embryonic axis and fuse to form a concentric tube, with inner endocardial (yellow) and outer myocardial (red) layers. Chambers develop around 24 hpf followed by looping morphogenesis at 36 hpf that moves the atrium to the right-hand side of the embryo and positions it above the ventricle. Valve formation is the final stage in cardiac development and occurs prior to hatching around 48 hpf. Adapted from Ackermann and Paw, 2003. w A B

Figure 1-8. Schematic of zebrafish pharyngeal arches at 5 dpf (A) and 48 hpf (B). The chondrocranium of zebrafish larvae comprises seven pharyngeal arches (A) and a neurocranium. The first mandibular arch forms the jaw (blue); the second hyoid, the jaw support (orange). The five branchial arches, pharyngeal arches 3-7, support the gills and pharyngeal teeth develop surrounding the seventh (green). All seven pharyngeal arches are three-layered structures (B). The core is made of mesoderm (purple) and is enclosed by two concentric layers: an inner mesenchyme, derived from cranial neural crest (yellow), and an outer ectoderm (green). 41

1.9 References

Agarkova, I. and Perriard, J.-C. (2005). The M-band: an elastic web that crosslinks thick filaments in the center of the sarcomere. Trends Cell Biol. 15, 477-485.

Allan, R. K. and Ratajczak, T. (2011). Versatile TPR domains accommodate different modes of target protein recognition and function. Cell Stress Chap. 16, 353-367.

Anderson, M. J., Pham, V. N., Vogel, A. M., Weinstein, B. M. and Roman, B. L. (2008). Loss of unc45a precipitates arteriovenous shunting in the aortic arches. Dev Biol. 318, 258-267.

Ao, W. and Pilgrim, D. (2000). Caenorhabditis elegans UNC-45 is a component of muscle thick filaments and colocalizes with myosin heavy chain B, but not myosin heavy chain A. J Cell Biol. 148, 375-384.

Barral, J. M., Hutagalung, A. H., Brinker, A., Hartl, F. U. and Epstein, H. F. (2002). Role of the myosin assembly protein UNC-45 as a molecular chaperone for myosin. Science 295, 669-671.

Bazzaro, M., Santillan, A., Lin, Z., Tang, T., Lee, M. K., Bristow, R. E., Shih, L.-M. and Roden, R. B. S. (2007). Myosin II co-chaperone general cell UNC- 45 overexpression is associated with ovarian cancer, rapid proliferation, and motility. Am J Pathol. 171, 1640-1649.

Bernick, E. P., Zhang, P.-J. and Du, S. (2010). Knockdown and overexpression of Unc45b result in defective myofibril organization in skeletal muscles of zebrafish embryos. BMC Cell Biol. 11, 70.

Berteaux-Lecellier, V., Zickler, D., Debuchy, R., Panvier-Adoutte, A., Thompson-Coffe, C. and Picard, M. (1998). A homologue of the yeast SHE4 gene is essential for the transition between the syncytial and cellular stages during sexual reproduction of the fungus. Podospora anserina. EMBO J. 17, 1248-1258.

Blatch, G. L. and Lassie, M. (1999). The tetratricopeptide repeat: a structural motif mediating protein-protein interactions. BioEssays 21, 932-939.

Buchner, J. (1999). Hsp90&Co. - a holding for folding. TIBS. 24, 136-141.

Chadli, A., Graham, J. D., Abel, M. G., Jackson, T. A., Gordon, D. F., Wood, W. M., Felts, S. J., Horwitz, K. B. and Toft, D. (2006). GCUNC-45 is a novel regulator for the progesterone receptor/hsp90 chaperoning pathway. Mol Cell Biol. 26, 1722-1730. 42

Chadli, A., Bruinsma, E. S., Stensgard, B. and Toft, D. (2008). Analysis of Hsp90 cochaperone interactions reveals a novel mechanism for TPR protein recognition. Biochemistry. 47, 2850-2857.

Chadli, A., Felts, S. J. and Toft, D. O. (2008). GCUNC45 is the first Hsp90 co- chaperone to show a/(3 isoform specificity. J Biol Chem. 283, 9509-9512.

Chen, J. N., Haffter, P., Odenthal, J., Vogelsang, E., Brand, M., van Eeden, F. J. M., Furutani-Seiki, M., Granato, M., Hammerschmidt, M., Heisenberg, C. P., Jiang, Y. J., Kane, D. A., Kelsh, R. N., Mullins, M. C. and Niisslein-Volhard, C. (1996). Mutations affecting the cardiovascular system and other internal organs in zebrafish. Development 123, 293-302.

Chuang, H.-N., Cheng, H.-Y., Hsiao, K.-M., Lin, C.-W., Lin, M.-L. and Pan, H. (2010). The zebrafish homeobox gene irxll is required for brain and pharyngeal arch morphogenesis. DevDyn. 239, 639-650.

Craig, R. and Woodhead, J. L. (2006). Structure and function of myosin filaments. Curr Opin Struct Biol. 16, 204-212.

Cresko, W. A., Yan, Y.-L., Baltrus, D. A., Amores, A., Singer, A., Rodriguez- Mari, A. and Postlethwait, J. H. (2003). Genome duplication, subfunction partitioning, and lineage divergence: sox9 in stickleback and zebrafish. Dev Dyn. 228, 480-489.

Csermely, P., Schnaider, T., Soti, C, Prohaszka, Z. and Nardai, G. (1998). The 90-kDa molecular chaperone family: structure, function, and clinical applications. A comprehensive review. Pharmacol Ther. 79, 129-168.

Dahm, R., Geisler, R. and Niisslein-Volhard, C. (2007). Zebrafish (Danio rerio) genome and genetics. In Genomics and genetics: from molecular details to analysis and techniques (ed. R. A. Meyers), pp. 469-502. Weinheim: Wiley- VCH GmbH & Co.

Dong, F., Sun, X., Liu, W., Ai, D., Klysik, E., Lu, M.-F., Hadley, J., Antoni, L., Chen, L., Baldini, A., Francis-West, P. and Martin, J. F. (2006). Pitx2 promotes development of splanchnic mesoderm-derived branchiomeric muscle. Development 133, 4891-4899.

Du, S. J., Li, H., Bian, Y. and Zhong, Y. (2008). Heat-shock protein 90a 1 is required for organized myofibril assembly in skeletal muscles of zebrafish embryos. Proc Natl Acad Sci USA. 105, 554-559.

Easter, S. S. Jr. and Nicola, G. N. (1996). The development of vision in the zebrafish (Danio rerio). Dev Biol. 180, 646-663. 43

Epping, M. T., Meijer, L. A., Bos, J. L. and Bernards, R. (2009). UNC45A confers resistance to histone deacetylase inhibitors and retinoic acid. Mol Cancer Res. 7, 1861-1870.

Etard, C, Behra, M., Fischer, N., Hutcheson, D., Geisler, R. and Strahle, U. (2007). The UCS factor Unc45b-/-/Unc-45b interacts with the heat shock protein Hsp90a during myofibrillogenesis. Dev Biol. 308, 133-143.

Etard, C, Roostalu, U. and Strahle, U. (2008). Shuttling of the chaperones Unc45b and Hsp90a between the A band and the Z line of the myofibril. J Cell Biol. 180,1163-1175.

Etard, C, Roostalu, U. and Strahle, U. (2010). Lack of Apobec2-related proteins causes a dystrophic muscle phenotype in zebrafish embryos. J Cell Biol. 189, 527-539.

Epstein, H. F. and Thomson, J. N. (1974). Temperature-sensitive mutation affecting myofilament assembly in Caenorhabditis elegans. Nature 250, 579- 580.

Etheridge, L., Diiorio, P. and Sagerstrom, C. G. (2002). A zebrafish unc-45- related gene expressed during muscle development. DevDyn. 224, 457-460.

Force, A., Lynch, M., Pickett, F. B., Amores, A., Yan, Y. L. and Postlethwait, J. (1999). Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151, 1531-1545.

Geach, T. J. and Zimmerman, L. B. (2010). Paralysis and delayed Z-disc formation in the Xenopus tropicalis unc45b mutant dicky ticker. BMC Dev Biol. 10, 75.

Granato, M., van Eeden, F. J. M., Schach, U., Trowe, T., Brand, M., Furutani-Seiki, M., Haffter, P., Hammerschmidt, M., Heisenberg, C.-P., Jiang, Y.-J., Kane, D. A., Kelsh, R. N., Mullins, M. C, Odenthal, J. and Niisslein-Volhard, C. (1996). Genes controlling and mediating locomotion behaviour of the zebrafish embryo and larvae. Development 123, 399-413.

Grunwald, D. J. and Eisen, J. S. (2002). Headwaters of the zebrafish — emergence of a new model vertebrate. Nat Rev Genet. 3, 717-724.

Hawkins, M. B., Thornton, J. W., Crews, D., Skipper, J. K., Dotte, A. and Thomas, P. (2000). Identification of a third distinct estrogen receptor and reclassification of estrogen receptors in teleosts. Proc. Natl. Acad. Sci. U. S. A. 97, 10751-10756. 44

Hawkins, T. A., Haramis, A.-P., Etard, C, Prodromou, C, Vaughan, C. K., Ashworth, R., Ray, S., Behra, M., Holder, N., Talbot, W. S., Pearl, L. H., Strahle, U. and Wilson, S. W. (2008). The ATPase-dependent chaperoning activity of Hsp90a regulates thick filament formation and integration during skeletal muscle myofibrillogenesis. Development 135, 1147-1156.

Hernandez, L. P., Patterson, S. E. and Devoto, S. H. (2005). The development of muscle fiber type identity in zebrafish cranial muscles. Anat Embryol. 209, 323-334.

Hinitis, Y., Osborn, D. P. S. and Hughes, S. M. (2009). Differential requirements for myogenic regulatory factors distinguish medial and lateral somitic, cranial and fin muscle fibre populations. Development 136, 403-414.

Hoppe, T., Cassata, G., Barral, J. M., Springer, W., Hutagalung, A. H., Epstein, H. F. and Baumeister, R. (2004). Regulation of the myosin-directed chaperone UNC-45 by a novel E3/E4-multiubiquitylation complex in C. elegans. Cell 118, 337-349.

Hutagalung, A. H., Landsverk, M. L., Price, M. G. and Epstein, H. F. (2002). The UCS family of myosin chaperones. J Cell Sci. 115, 3983-3990.

Isogai, S., Horiguchi, M. and Weinstein, B. (2001). The vascular anatomy of the developing zebrafish: an atlas of embryonic larval development. Dev Biol. 230, 278-301.

Javidan, Y. and Schilling, T.F. (2004). Development of cartilage and bone. Meth Cell Bio. 76, 415-436.

Kachur, T., Ao, W., Berger, J. and Pilgrim, D. (2004). Maternal UNC-45 is involved in cytokinesis and colocalizes with non-muscle myosin in the early Caenorhabditis elegans embryo. J Cell Sci. 117, 5313-5321.

Kassar-Duchossoy, L., Gayraud-Morel, B., Gomes, D., Rocancourt, D., Buckingham, M., Shinin, V. and Tajbakhsh, S. (2004). Mrf4 determines skeletal muscle identity in Myf5:Myod double-mutant mice. Nature 431, 466- 471.

Knight, R. D., Mebus, K., d'Angelo, A., Yokoya, K., Heanue, T., Tubingen 2000 Screen Consortium and Roehl, H. (2011). Ret signalling integrates a craniofacial muscle module during development. Development 138, 2015- 2024.

Lai, B.-T., Chin, N. W., Stanek, A. F., Keh, W. and Lanks, K. W. (1984). Quantitation and intracellular localization of the 85K heat shock protein by using monoclonal and polyclonal antibodies. Moh Cell Biol. 4, 2802-2810. 45

Lan, X., Wen, L., Li, K., Liu, X., Luo, B., Chen, F., Xie, D. and Kung, H. (2011). Comparative analysis of duplicated sox21 genes in zebrafish. Develop Growth Differ. 53, 347-356.

Landsverk, M. L. and Epstein, H. F. (2005). Genetic analysis of myosin II assembly and organization in model organisms. Cell Mol Life Sci. 62, 2270- 2282.

Landsverk, M. L., Li, S., Hutagalung, A. H., Najafov, A., Hoppe, T., Barral, J. M. and Epstein, H. F. (2007). The UNC-45 chaperone mediates sarcomere assembly through myosin degradation in Caenorhabditis elegans. J Cell Biol. Ill, 205-210. Langeland, J. A. and Kimmel, C. B. (1997). Fishes. In Embryology: constructing the organism (ed. S. F. Gilbert and A. M. Raunio), pp. 383-407. Sunderland: Sinauer Associates Inc.

Lin, C.-Y., Yung, R.-F., Lee, H.-C, Chen, W.-T., Chen, Y.-H. and Tsai, H.-J. (2006). Myogenic regulatory factors Myf5 and Myod function distinctly during craniofacial myogenesis in zebrafish. Dev Biol. 299, 594-608.

Liu, L., Srikakulam, R. and Winkelmann, D. A. (2008). Unc45 activates Hsp90-dependent folding of the myosin motor domain. J Biol Chem. 283, 13185-13193.

Lodish, H., Berk, A., Matsudaira, P., Kaiser, C. A., Krieger, M., Scott, M. P., Zipursky, S. L. and Darnell, J. (2000). Molecular Cell Biology, 5th edition. New York: W. H. Freeman and Company. Lohr, J. L. and Yost, H. J. (2000). Vertebrate model systems in the study of early heart development: Xenopus and zebrafish. Am J Med Genet. 91, 248- 257.

Lowey, S. and Trybus, K. M. (2010). Common structural motifs for the regulation of divergent class II myosins. J Biol Chem. 285, 16403-16407.

Lynch M. and Force, A. (2000). The probability of duplicate gene preservation by subfunctionalization. Genetics 154, 459-473.

McClintock, J. M., Kheirbek, M. A. and Prince, V. E. (2002). Knockdown of duplicated zebrafish hoxbl genes reveals distinct roles in hindbrain patterning and a novel mechanism of duplicate gene retention. Development 129, 2339- 2354.

Minucci, S. and Pelicci, P. G. (2006). Histone deacetylase inhibitors and the promise of epigenetic (and more) treatments for cancer. Nat Rev Cancer 6, 38- 51. 46

Mishra, M., D'souza, V. M., Chang, K. C, Huang, Y. and Balasubramanian, M. K. (2005). Hsp90 protein in fission yeast Swolp and UCS protein Rng3p facilitate myosin II assembly and function. Eukaryot Cell 4, 567-576.

Noden, D. M. and Trainor, P. A. (2005). Relations and interactions between cranial mesoderm and neural crest populations. JAnat. 207, 575-601.

Niisslein-Volhard, C, Gilmour, D. T. and Dahm, R. (2002). Zebrafish as a system to study development and organogenesis. In Zebrafish (ed. C. Niisslein- Volhard and R. Dahm), pp. 1-5. New York: Oxford University Press Inc.

Ochi, H. and Westerfield, M. (2007). Signaling networks that regulate muscle development: lessons from zebrafish. Develop. Growth Differ. 49, 1-11.

Ohno, S. (1970). Evolution by gene duplication. Springer-Verlag.

Perez-Perez, J. M., Candela, H. and Micol, J. L. (2009). Understanding synergy in genetic interactions. Trends Genet. 25, 368-376.

Picard, D. (2002). Heat-shock protein 90, a chaperone for folding and regulation. Cell Mol Life Sci. 59, 1640-1648.

Pickart, C. M. (2004). Back to the future with ubiquitin. Cell 116, 181-190.

Pratt, W. B., Galigniana, M. D., Harrell, J. M. and DeFranco, D. B. (2004). Role of hsp90 and the hsp90-binding immunophilins in signalling protein movement. Cell Signal. 16, 857-872.

Price, M. G., Landsverk, M. L., Barral, J. M. and Epstein, H. F. (2002). Two mammalian UNC-45 isoforms are related to distinct cytoskeletal and muscle- specific functions. J Cell Sci. 115, 4013-4023.

Prince, V. E. and Pickett, F. B. (2002). Splitting pairs: the diverging fates of duplicated genes. Nat Rev Genet. 3, 827-837.

Ramsey, A. J., Russell, L. C, Whitt, S. R. and Chinkers, M. (2000). Overlapping sites of tetratricopeptide repeat protein binding and chaperone activity in heat shock protein 90. J. Biol. Chem. 275, 17857-17862.

Ravi, V. and Venkatesh, B. (2008). Rapidly evolving fish genomes and teleost diversity. Curr. Opin. Genet. Dev. 18, 544-550.

Sanger, J. W., Wang, J., Holloway, B., Du, A. and Sanger, J. M. (2009). Myofibrillogenesis in skeletal muscle cells in zebrafish. Cell Motil Cytoskeleton 66, 556-566. 47

Sanger, J. W., Wang, J., Fan, Y., White, J. and Sanger, J. M. (2010). Assembly and dynamics of myofibrils. J Biomed Biotechnol. 858606.

Sellers, J. R. and Knight, P. J. (2007). Folding and regulation in myosins II and V. J Muscle Res Cell Motih 28, 363-370.

Schilling, T.F., Piotrowski, T., Grandel, H., Brand, M., Heisenberg, C.-P., Jiang, Y.-J., Beuchle, D., Hammerschmidt, M., Kane, D. A., Mullins, M. C, van Eeden, F. J. M., Kelsh, R. N., Furutani-Seiki, M., Granato, M., Haffter, P., Odenthal, J., Warga, R. M., Trowe, T. and Niisslein-Volhard, C. (1996). Jaw and branchial arch mutants in zebrafish I: branchial arches. Development 123, 329-344.

Schilling, T. F. and Kimmel, C. B. (1997). Musculoskeletal patterning in the pharyngeal segments of the zebrafish embryo. Development 124, 2945-2960.

Schilling, T. F. (2002). The morphology of larval and adult zebrafish. In Zebrafish (ed. C. Niisslein-Volhard and R. Dahm), pp. 59-94. New York: Oxford University Press Inc.

Schoenebeck, J. J. and Yelon, D. (2007). Illuminating cardiac development: Advances in imaging add new dimensions to the utility of zebrafish genetics. Semin Cell Dev Biol. 18, 27-35.

Srikakulam, R. and Winkelmann, D. A. (1999). Myosin II folding is mediated by a molecular chaperonin. J Biol Chem. 274, 27265-27273.

Srikakulam, R. and Winkelmann, D. A. (2003). Chaperone-mediated folding and assembly of myosin in striated muscle. J Cell Sci. Ill, 641-652.

Srikakulam, R., Liu, L. and Winkelmann, D. A. (2008). Unc45b forms a cytosolic complex with Hsp90 and targets the unfolded myosin motor domain. PLoSOneZ, e2137.

Stainier, D. Y. R., Fouquet, B., Chen, J.-N., Warren, K. S., Weinstein, B. M., Meiler, S. E., Mohideen, M. P. K., Neuhauss, S. C. F., Solnica-Krezel, L., Schier, A. F., Zwartkruis, F., Stemple, D. L., Malicki, J., Driever, W. and Fishman, M. C. (1996). Mutations affecting the formation and function of the cardiovascular system in the zebrafish embryo. Development 123, 285-292.

Stainier, D. Y. R. (2001). Zebrafish genetics and vertebrate heart formation. Nat Rev Genet. 2, 39-48.

Talbot, J. C, Johnson, S. L. and Kimmel, C. B. (2010). hand2 and Dlx genes specify dorsal, intermediate and ventral domains within zebrafish pharyngeal arches. Development 137, 2507-2517. 48

Taylor, J. S., Braasch, I., Frickey, T., Meyer, A. and Van de Peer, Y. (2003). Genome duplication, a trait shared by 22,000 species of ray-finned fish. Genome Res. 13,382-390.

Tsai, M.-J. and O'Malley, B. W. (1994). Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Annu Rev Biochem. 63, 451- 486.

Tzahor, E., Kempf, H., Mootoosamy, R. C, Poon, A. C, Abzhanov, A., Tabin, C. J., Dietrich, S. and Lassar, A. B. (2003). Antagonists of Wnt and BMP signaling promote the formation of vertebrate head muscle. Genes Dev. 17, 3087-3099.

Van de Peer, Y., Taylor, J. S., Joseph, J. and Meyer, A. (2002). Wanda: a database of duplicated fish genes. Nuclei Acids Res. 30, 109-112.

Venolia, L. and Waterston, R. H. (1990). The unc-45 gene of Caenorhabditis elegans is an essential muscle-affecting gene with maternal expression. Genetics 126, 345-353.

Venolia, L., Ao, W., Kim, S., Kim, C. and Pilgrim, D. (1999). unc-45 gene of Caenorhabditis elegans encodes a muscle-specific tetratricopeptide repeat- containing protein. Cell Motil Cytoskeleton 42, 163-177.

Vikstrom, K. L., Seiler, S. H., Sohn, R. L., Strauss, M., Weiss, A., Welikson, R. E. and Leinwand, L. A. (1997). The vertebrate myosin heavy chain: genetics and assembly properties. Cell Struct Fund. 22, 123-129.

Volff, J.-N. (2005). Genome evolution and biodiversity in teleost fish. Heredity 94, 280-294.

Warren, K. S., Wu, J. C, Pinet, F. and Fishman, M. C. (2000). The genetic basis of cardiac function: dissection by zebrafish (Danio rerio) screens. Phil. Trans. R. Soc. Lond. B. 355, 939-944.

Wiech, H., Buchner, J., Zimmermann, R. and Jakob, U. (1992). Hsp90 chaperones protein folding in vitro. Nature 358, 169-170.

Wohlgemuth, S. L., Crawford, B. D. and Pilgrim, D. B. (2007). The myosin co-chaperone UNC-45 is required for skeletal and cardiac muscle function in zebrafish. Dev Biol. 303, 483-492.

Wolfe, K. H. (2001). Yesterday's polyploids and the mystery of diploidization. Nat Rev Genet. 2, 333-341. 49

Wong, K. C, Naqvi, N. I., lino, Y., Yamamoto, M. and Balasubramanian, M. K. (2000). Fission yeast Rng3p: an UCS-domain that mediates myosin II assembly during cytokinesis. J Cell Sci. 113, 2421-2432.

Wong, K. C, D'Souza, V. M., Naqvi, N. L, Motegi, F., Mabuchi, I. and Balasubramanian, M. K. (2002). Importance of a myosin II-containing progenitor for actomyosin ring assembly in fission yeast. Curr Biol. 12, 724- 729.

Yelick, P. C. and Schilling, T. F. (2002). Molecular dissection of craniofacial development using zebrafish. Crit Rev Oral Biol Med. 13, 308-322.

Yelon, D., Home, S. A. and Stainier, D. Y. (1999). Restricted expression of cardiac myosin genes reveals regulated aspects of heart tube assembly in zebrafish. Development 214, 23-37.

Yu, Q. and Bernstein, S. I. (2003). UCS proteins: managing the myosin motor. Curr Biol 13, R525-R527.

Zapater, C, Chauvigne, F., Norberg, B., Finn, R. N. and Cerda, J. (2011). Dual neofunctionalization of a rapidly evolving aquaporin-1 paralog resulted in constrained and relaxed traits controlling channel function during meiosis resumption in teleosts. Mol Biol Evol. Advanced Access, June 2011.

Zuehlke, A. and Johnson, J. L. (2010). Hsp90 and co-chaperones twist the functions of diverse client proteins. Biopolymers. 93, 211-217. 50

CHAPTER TWO: MATERIALS AND METHODS 51

2.1 Animal Care and Zebrafish Lines

2.1.1 Zebrafish Maintenance

Zebrafish were housed in the Biological Sciences aquatic facility kept at 28.5°C under controlled light conditions (14 hours light, 10 hours dark) according to standard procedures (Westerfield, 2000). Adults were naturally spawned to obtain embryos, which were raised at 28.5°C and staged according to published morphological hallmarks (Kimmel et ah, 1995). Embryos analyzed past the 24 hpf stage were incubated in embryo medium supplemented with 0.003% phenylthiourea (Sigma, P-7629) to prevent melanin formation (Westerfield, 2000). All procedures were carried out in compliance with the guidelines stipulated by the Canadian Council for Animal Care and the University of Alberta.

2.1.2 steif(steifl unc45bsb6mb60)

The unc45b mutant steif (steif I unc45bsb60/sh6<)) was identified in an ethylnitrosourea (ENU) chemical screen for motility mutants (Behra et ah, 2002). A C-to-A transversion of base pair 2,473 replaces a cysteine residue with a premature stop codon at amino acid 788 (Etard et ah, 2007). This line was provided by the Max-Planck-Institute fur Entwicklungsbiologie (Tubingen, Germany). Throughout this thesis, the following abbreviated nomenclature will be used: unc45b~'~ (homozygous mutant unc45b and homozygous wild type unc45a: unc45b~"; unc45a+ +) and unc45b+' (heterozygous mutant unc45b).

2.1.3 kurzschluss (kus I unc45atrnitrU)

The unc45a mutant kurzschluss (kus I unc45axX trl2) was identified in a large-scale ENU mutant screen for zebrafish cardiovascular mutants (Chen et ah, 1996; Haffter et ah, 1996). A T-to-A transversion of base pair 2,146 changes leucine to a premature stop codon at amino acid 655 (Anderson et ah, 2008). This line was provided by Dr. Beth Roman, from the Department of Biological Sciences, University of Pittsburgh, (Pittsburgh, USA). Throughout this thesis the 52

following abbreviated nomenclature will be used: unc45d' (homozygous mutant unc45a and homozygous wild type unc45b: unc45b++; unc45a") and unc45a+~ (heterozygous mutant unc45d). For experiments performed on embryos aged 3 dpf or older, genotypes were scored by phenotype.

2.1.4 unc45bsb60hb60; unc45atrl2ltrU

This line was created through the crossing of adult male and female unc45b+' and unc45a+~ fish. Heterozygous (unc45b+~; unc45a+~) Fl fish were identified by derived cleaved amplified polymorphic sequence analysis, as described below. Genotyped homozygous, double unc45 mutants, will be referred to using the following short hand nomenclature: unc45b~/~; unc45a'A.

2.2 Genotyping

2.2.1 Genomic DNA Extraction

Genomic DNA was collected from fish aged 3 months or older by the fin clipping method, or from embryos following experimentation. For fin clipping, fish were anaesthetized using a 1.6% dilution of Tricaine (Sigma, A5040). One at a time, fish were rinsed in water and then a small portion of the tail fin was removed, using a scalpel, and placed in a labelled PCR tube. Fish were kept separately in labelled tanks to recover whilst genotyping was performed. Fin clippings and whole embryos were digested using 50 uL of ELVIS buffer (50 mM KC1; 10 mM Tris, pH 8.5; 5 mM EDTA; 0.01% gelatin; 0.45% IGEPAL; 0.45% Tween-20) and 5 uL of 10 mg/mL Proteinase K (Sigma, P2308). Using a PCR thermocycler, samples were extracted for 3 hours at 55°C, followed by Proteinase K inactivation by heating at 85°C for 30 minutes. Samples were stored at 4°C until genotyped using dCAPS analysis. 53

2.2.2 dCAPS Analysis

Genotyping was performed using derived cleaved amplified polymorphic sequence (dCAPS) analysis for all experiments that required the identification of either homozygous or heterozygous double mutants, or was conducted at a developmental stage prior to when mutants could be scored phenotypically. This method was chosen over other SNP assays, as it is simple, reliable, and relatively inexpensive (Neff et ah, 1998). dCAPS analysis is used in cases where the SNP of interest neither alters nor creates a restriction recognition site. For each embryo analyzed, the presence or absence of the SNP for the mutant haplotype was determined by the restriction pattern resulting from the enzymatic digestion of a PCR product generated through the use of special dCAPS primers (Neff et ah, 1998) (Table 2-1). Primers were designed using the free software program dCAPS Finder 2.0 available at http://helix.wustl.edu/dcaps/dcaps.html (Neff et ah, 2002). Genomic DNA was used as a template and primer pairs were designed so that either the forward or reverse primer contained a mismatch, thereby incorporating into the amplicon a restriction enzyme recognition sequence, which includes the mutant haplotype SNP of interest (Neff et ah, 1998; Neff et ah, 2002). PCR reactions were prepared using 17.5 pL water, 2.5 uL lOx PCR 2.0 buffer (500 mM Tris-HCl, pH 9.2; 160 mM (NH4)2S04; 22.5 mM MgCl2), 0.5 uL 10 mM dNTPs, 2 pL genomic DNA template, 1 pL 5 mM each of forward and reverse primers, and 0.5 pL Taq polymerase. Samples were amplified using an UNC45BGENO PCR program with the following conditions: 94°C, 3 minutes; (94°C, 30 seconds; 51.5°C, 40 seconds; 72°C, 1 minute) for 36 cycles; and a final extension at 72°C for 5 minutes. Samples were stored at 4°C until proceeding with enzymatic digestion. PCR amplification products were digested with restriction enzymes in a 15 pL reaction volume consisting of 5 uL PCR template, 8.25 pL water, 1.5 pL enzyme buffer, and 0.25 pL (2.5 U) restriction enzyme. Reactions were incubated in a 37°C incubator overnight. Enzymes used were EcoRI (Invitrogen, 15202-013) for unc45b products and Ddel (NEB, R0175L) for unc45a products. The following day, the entire digest was visualized on a 2% (wt/vol) agarose 54 gel in IX TAE buffer (40 mM Tris, 1 mM EDTA) run initially for 30 min at 100 V and then at 150 V until the lower loading dye had migrated to a point approximately 1 cm from the bottom of the gel. This allowed sufficient resolution of the two bands. Fragment sizes were determined using DNA ladder (Fermentas, SM1334) and samples were loaded with 6X upper loading buffer (0.25% xylene cyanol FF, 15% Ficol). Since samples across the gel do not migrate in a perfectly horizontal manner, once every 8 lanes an undigested PCR sample was added for sizing purposes. This sample was loaded with 6X lower loading buffer (0.25% bromophenol blue, 15% Ficol) and was also used to visualize the migration front to gauge when to stop the electrophoresis. A representative genotyping sample is shown in Figure 2-1.

2.3 Whole Mount in situ Hybridization

2.3.1 Total RNA Extraction and cDNA Synthesis

Total RNA was isolated from fifty embryos, staged at 24 hpf, by TRIzol (Invitrogen, 15595-026) extraction as follows. Embryos were collected in 1.7 mL microcentrifuge tubes, excess embryo media was removed, and the tubes were flash frozen in liquid nitrogen. Embryos were homogenized using a pestle and 500 pL of TRIzol for 5 minutes or until large chunks were no longer visible. An additional 500 pL of TRIzol was then added and embryos were homogenized for another minute. Next, 200 pL of chloroform was added and the samples were vortexed for 15 seconds and left on ice for 2 minutes before being centrifuged at 14,000 rpm for 30 minutes at 4°C. The resulting upper aqueous layer was transferred to a new, labelled microcentrifuge tube, taking care as to avoid the interphase. Once again, 200 pL of chloroform was added, the samples vortexed for 15 seconds, left on ice for 1 minute, and then spun at 14,000 rpm for 30 minutes at 4°C. The resulting aqueous layer was transferred to a new microcentrifuge tube, to which 500 pL of ice-cold isopropanol (stored at -20°C) was added. The tubes were inverted to mix and then left at room temperature for 10 minutes to allow the RNA to precipitate. To pellet the precipitated RNA, 55 samples were spun at 14,000 rpm for 30 minutes at 4°C. The supernatant was discarded and the pellet was washed with 70% RNase-free ethanol and spun once more for 15 minutes at 4°C. As much supernatant as possible was removed and the pellet was left to air dry for 10 minutes at room temperature. The pellet was then resuspended in 50 pL RNase-free water (DEPC). RNA was either used immediately for cDNA synthesis or stored at -80°C. To generate PCR based riboprobes for in situ hybridization, the SuperScript™III Reverse Transcriptase kit (Invitrogen, 18080-044) was used following the manufacturer's instructions. To a 200 pL PCR tube the following was added: 1 pg oligo (dT)2o primer, 10 pg - 5 pg RNA template, 1 pL 10 mM dNTP mix, and the appropriate amount of water to bring the reaction volume to 13 pL. The mixture was heated for 5 minutes at 65°C and then incubated on ice for 1 minute. Next, 4 pL of 5X first-strand buffer, 1 pL 0.1 M DTT, and 1 pL SuperScript™III reverse transcriptase were added to the tube and the contents were briefly vortexed to mix before being spun down. The reaction was performed in a thermocycler for 2 hours at 50°C and stopped by heating for 15 minutes at 70°C. To remove the RNA template, 1 pL (5 U) RNaseH (NEB, M0297S) was added to the samples and incubated for 20 minutes at 37°C. The resulting cDNA was diluted in 25 pL water and stored at -20°C.

2.3.2 Preparation of PCR Based DIG-labelled RNA Probes

Digoxigenin-labelled antisense probes, measuring between 800 and 1,000 bp in length, were made from a PCR template amplified using gene-specific primers. The 5' ends of the reverse primers were designed to have either a T3 or T7 RNA polymerase promoter sequence to enable mRNA production (Thisse and Thisse, 2008) (Table 2-2, 2-3). In order to minimize cross-reactivity, the coding sequence and 3' untranslated region of the cDNA were used as template. The probe sequence was amplified using 2 pL of cDNA as template and run on the following touchdown program: 95°C, 5 minutes; touchdown (94°C, 20 seconds; 65°C -1°C / cycle, 20 seconds; 72°C, 1 minute) for 10 cycles; amplification (94°C, 20 seconds; 55°C, 20 seconds; 72°C, 1 minute) for 30 cycles; and a final 56 elongation at 72°C for 10 minutes (Korbie and Mattick, 2008). Reactions were prepared in 1.7 mL microcentrifuge tubes on ice to which the following was added: 200^400 ng PCR template, 2 pL 10X transcription buffer (Ambion), 10X DIG RNA labelling mix (Roche, 11 277 073 910), 1 pL RNA polymerase (Ambion), and RNase-free water to bring the volume up to 20 pL. Reactions were incubated for 2 hours in a 37°C water bath. After one hour, an additional 1 pL of RNA polymerase was added. To remove the DNA template, 1 pL DNase was added and incubated for 5 minutes at 37°C. The reaction was then stopped, by adding 2 pL 0.2 M EDTA. Probes were left overnight at -20°C to precipitate in 30 pL water, 25 pL 4 M LiCl, and 200 pL 95% ethanol. The following day probes were collected by spinning at 14,000 rpm for 30 minutes at 4°C. The pellet was washed in 70% RNase-free ethanol and spun once more for 15 minutes at 4°C. The pellet was resuspended in 50 pL RNase-free water and then a portion of the probe was diluted 1:100 in hybridization solution and stored at -20°C, while the remainder of the probe solution was stored at -80°C.

2.3.3 mRNA in situ Hybridization and Detection

All stages of the protocol were performed in 1.7 mL microcentrifuge tubes at room temperature with gentle rocking unless stated otherwise. Dechorionated embryos were fixed overnight at 4°C and then stored in 100% methanol at -20°C. Prior to resuming the protocol, embryos were rehydrated for 5 minutes in a 75% MeOH / 25% PBST, 50% MeOH / 50 % PBST, 25% MeOH / 75% PBST series and then washed 5 times for 5 minutes each in PBS-Tw (PBS with 0.1 % Tween- 20). In order for the probe to penetrate into tissues, embryos were permeabilized with 10 pg/mL Proteinase K in PBS-Tw. Based on the developmental stage of the embryos used, the following permeation times were used: 27 hpf, 3 minutes; 48 hpf, 20 minutes; 3 dpf, 25 minutes. Following permeabilization, embryos were fixed in 4% PFA for 20 minutes and then washed 5 times for 5 minutes in PBS- Tw. Prehybridization, hybridization, and post-hybridization washes were all performed in a 65°C water bath. Embryos were prehybridized in hybridization 57 solution (50%) formamide; 5X SSC; 50 pg/mL heparin; 500 pg/mL yeast tRNA; 0.1%o Tween-20, pH 6.0) for 1 hour, after which the appropriate probe was added, and left to hybridize overnight. Single washes were performed for 5 minutes in each of the following solutions: 66% hybridization buffer (Hyb) / 33% 2X SSC; 33% Hyb / 66% 2X SSC; 2X SSC. High stringency washes were then performed once in 0.2X SSC / 0.1% Tween-20 for 20 minutes and twice in 0.1X SSC / 0.1% Tween-20, also for 20 minutes. The following washes were performed at room temperature for 5 minutes each: 66% 0.2X SSC / 33% PBS-Tw; 33% 0.2X SSC / 66% PBS-Tw; PBS-Tw. Once the washes were complete, embryos were incubated in blocking solution (PBS-Tw containing 2% sheep serum and 2 mg/mL BSA) for 1 hour at room temperature and then incubated overnight at 4°C in a 1:5,000 dilution of anti-DIG-AP antibody (Roche) in blocking solution. To visualize the DIG-labelled probes, embryos were given a quick rinse in PBS-Tw to remove excess antibody and then washed 5 times for 15 minutes in PBS-Tw. Prior to the colouration reaction, embryos were washed 4 times for 5 minutes in colouration buffer (100 mM Tris-HCl, pH 9.5; 50 mM MgCl2; 100 mM NaCl; 0.1 % Tween-20) and then developed with nitroblue tetrazolium (NBT) / bromo-chloro indoyl phosphate (BCIP). Samples were developed in the dark and those that developed colour slowly were placed in a 33°C incubator to increase the speed of the colouration reaction. To stop colour development, samples were rinsed twice with distilled water and then stored in 100%> methanol with 1% Tween-20 until imaged.

2.4 Immunohistochemistry

2.4.1 MF-20

Myosin organization was studied using the monoclonal antibody MF-20 (Developmental Studies Hybridoma Bank, Iowa City, I A) (Table 2-4). All steps were performed in 1.7 mL microcentrifuge tubes with gentle rocking. Embryos, 3 or 4 dpf, were fixed in 2% TCA for 3 hours at room temperature. Embryos were 58 first rinsed twice in PBS for 5 minutes and then washed four times with PBS containing 0.8% Triton X-100 (PBS-Tx) for 5 minutes at room temperature. To minimize non-specific binding, samples were incubated in blocking solution (5% BSA in PBS-Tx) for 1 hour at room temperature. Samples were incubated overnight at 4°C in a 1:10 dilution of MF-20 in blocking solution. The next day, embryos were washed five times with PBS-Tx for 5 minutes at room temperature. Subsequent to a single 5 minute PBS wash, embryos were incubated overnight at 4°C in a 1:1,000 dilution of anti-mouse Alexa 488 (Molecular Probes) secondary antibody. Prior to mounting embryos in 3% methylcellulose, excess secondary antibody was removed by three washes in PBS-Tx for 10 minutes each at room temperature.

2.4.2 3,3'-Diaminobenzidine

Cranial musculature was examined using the monoclonal antibody MF-20 (Developmental Studies Hybridoma Bank, Iowa City, IA) and visualized using 3,3'-diaminobenzidine (Sigma, D4293-5) (Table 2-4). All steps were performed in 1.7 mL microcentrifuge tubes with gentle rocking. At 3 dpf, embryos were fixed in 2% TCA for 2 hours at room temperature. Embryos were then rinsed three times in PBS containing 0.1 % Triton X-100 (PBS-Tx) for 5 minutes at room temperature. To minimize non-specific binding, samples were incubated in blocking solution (5% BSA in PBS-Tx) for 2 hours at room temperature. Samples were incubated overnight at 4°C in a 1:10 dilution of MF-20 in blocking solution. The next day, embryos were washed four times with PBS-Tx for 5 minutes at room temperature and subsequently incubated for 4 hours at room temperature in a 1:300 dilution of anti-mouse IgG horseradish peroxidase ECL (GE Healthcare, NA 931V) secondary antibody. Excess antibody was removed by washing 3 times in PBS-Tx for 5 minutes. Embryos were then incubated in 3,3'-diaminobenzidine solution in the fume hood for 30 minutes. To develop the stain, 2.5 pL 3% hydrogen peroxide was added and left until the colour had reached the desired level of saturation. To stop the reaction, embryos were washed 5 times in water for 5 minutes. Samples were stored in 100%) glycerol until imaged. 59

2.4.3 Eye Cross Sections and Immunohistochemistry

Cryosections of embryonic eyes were used to examine lens differentiation and eye morphology. This protocol is as described previously by Uribe and Gross (2007) and French et ah (2009). Embryos, ranging from 3 dpf to 5 dpf, were fixed overnight in 4%o PFA at 4°C. Following 3 washes in PBS for 5 minutes, embryos were soaked in 25%) sucrose in PBS for 2 hours at room temperature and then soaked in a 35% sucrose solution for another 2 hours or until the embryos sunk to the bottom of the tube. Embryos were then embedded in OCT in cryomolds and frozen at -20°C until cryosectioned at a thickness of 10 pm. Sections were left to adhere to slides at room temperature for 2 hours before being rehydrated in PBS- Tw for 10 minutes. Slides were then placed in a humid box and blocked (PBS-Tw containing 5% BSA and 1% sheep serum) for 1 hour while covered to prevent desiccation. The slides were incubated overnight at 4°C in the humid box with a 1:100 dilution of monoclonal Zl-1 antibody (ZIRC, Eugene, OR) in blocking solution (Table 2-4). The next day, slides were rinsed 3 times for 10 minutes in PBS-Tw before adding the following into the blocking solution: 1:1,000 dilution of anti-mouse Alexa-488 secondary antibody, 1:50 dilution of phalloidin, and 5 pg/mL DAPI. Slides were incubated in the dark at room temperature in a humid box for 2 hours. Excess antibody was removed by rinsing 3 times for 10 minutes with PBS-Tw before topping the slides with 70%> glycerol and a coverslip in preparation for imaging.

2.5 Whole Mount Staining

2.5.1 Phalloidin

F-actin structure was examined using Alexa-546 conjugated phalloidin, a phallotoxin from the fungus Amanita phalloides, commonly known as the death cap. All steps were performed in 1.7 mL microcentrifuge tubes with gentle rocking. Embryos, 3 or 4 dpf, were fixed in 4% PFA for one hour at room temperature, or overnight at 4°C. Embryos were then permeabilized by washing 60

three times for 10 minutes each with PBS containing 0.1% Triton X-100 (PBS- Tx) at room temperature. Staining was performed in the dark using a 1:50 dilution of phalloidin (Molecular Probes, A22283) in PBS for 30 minutes at room temperature (Table 2-4). Excess stain was removed by washing samples three times for 5 minutes in PBS-Tx. Embryos were then mounted in 3% methylcellulose for imaging, or in 100%) glycerol for storage.

2.5.2 Alcian Blue

Pharyngeal jaw cartilages were examined using Alcian Blue dye, which stains proteoglycan components of the extracellular matrix surrounding chondrocytes (Yelick and Schilling, 2002; Javidan et ah, 2004). The following method is adapted from Walker and Kimmel (2007). All experiments were performed in 1.7 mL microcentrifuge tubes at room temperature with gentle rocking unless stated otherwise. Larvae, 5 dpf, were fixed in 4%> PFA for 2 hours and then dehydrated in 50% ethanol for 10 minutes. Once dehydrated, larvae were left overnight in staining solution: 0.02%> Alcian Blue (Sigma, A5268), 60 mM

MgCl2, and 70% ethanol (Table 2-4). Excess dye was removed with a quick rinse in water and then embryos were bleached for 20 minutes in equal volumes of 3%

H202 and 2% KOH. While in bleaching solution, tubes were left open on the bench top. To visualize cartilage more readily, tissues were cleared for 20 minutes with gentle rocking using 1 mg/mL trypsin dissolved in saturated sodium tetraborate (Schilling et ah, 1996). Larvae were cleared in 20% glycerol with 0.25% KOH for 1 hour followed by 50% glycerol with 0.25% KOH for 2 hours. Larvae were then stored at 4°C in 100% glycerol.

2.5.3 Alizarin Red

Alizarin Red dye was used to visualize mineralized tissues such as pharyngeal teeth. All steps were performed in 1.7 mL microcentrifuge tubes with gentle rocking, while limiting exposure of Alizarin Red to light. This protocol is adapted from Engeman et al. (2009). Larvae staged at 5 dpf were fixed in 4% PFA overnight at 4°C. The next day they were given a quick rinse in 0.5% KOH 61 to remove residual fixative before being washed with 0.5% KOH for 10 minutes at room temperature. Larvae were then stained for 2 hours at room temperature with 0.5%) Alizarin Red (Sigma, A5533) in 0.5% KOH (Table 2-4). Excess stain was removed by rinsing in 0.5% KOH and then larvae were cleared for 3 hours in 0.5%o KOH. Following clearing, larvae were placed in a 50:50 solution of 50%) glycerol and 0.5% KOH before being stored in 100% glycerol at 4°C.

2.5.4 O-dianisidine

O-dianisidine staining was used to examine heme expression and as an indirect method to observe circulation defects and blood pooling. This method is adapted from Dietrich et al. (1995) and Ransom et ah (1996). Non-fixed dechorionated embryos were stained in the dark for 15 minutes at room temperature in 0.6 mg/mL O-dianisidine (Sigma, D9143), 0.01M sodium acetate pH 5.2, 0.65%o hydrogen peroxide, and 40%) ethanol (Table 2-4). Following staining, embryos 3 dpf or older were postfixed in 4% PFA for 1 hour at room temperature. Embryos were then rinsed 4 times for 5 minutes each in water to remove residual stain, followed by mounting in glycerol for imaging.

2.6 Photography and Image Processing

Whole-mount embryos were imaged using an Olympus stereoscope with a Qimaging micropublisher camera. Immunofluorescence images were captured using a Nikon Eclipse 80i confocal. All images were processed using Adobe Photoshop CS Version 8.0. 2.7 Tables

Table 2-1. Primers and Restriction Enzymes Used in dCAPS Analysis

Digested Digested dCAPS Product Product Gene Primer Name Sequence (5' - 3') Primer ^ Size (bp) Size (bp) wt mutant

Unc45a(Geno)MutForl CTT TTT TCC TCC TCT TCA CAG GGT GAC TT1 Yes unc45a Ddel 252 227 Unc45a(Geno)Revl TCA GAT TTG AGT TTG AGC AGT TGA TC No

Unc45bEcoRlMutF GTT CAT ACC TCC TTG CAG CAA ACT AGA AT1 Yes unc45b EcoRI 200 173 Unc45b(Geno)R3 GCT TAC CCC AAA ATC TTT AAA CAA ATA No 1. Forward primers contain a single mismatch (underlined) creating restriction enzyme sites in the mutant alleles.

62 Table 2-2. Primer Sequences for Antisense PCR Based DIG-labelled RNA Probes

RNA Gene Forward Primer (5'-3') T(m) Reverse Primer (5'-3') T(m) Size (bp) Polymerase

dlx2a CCT CCG GTG AGT CTC CAC CTC 61 2 GGC AGA GAT GTT CAT TCG GCT TT1 65 6 902 T7

dlx2b CAG GTT GCT TCC GOT GAC TCT C 59 9 AAA ATC GTT GAT GTC AGC TTA AAA GGA1 63 3 752 T7

gata4 CAT AAC TCG ACT TCT CCG GTG TAC G 58 6 GTT CCA CAC TTC ACT CTT GGA GCT G2 66 6 931 T3

gataS CTA CCG GGA AGG AGG TCC AGT ATA G 59 4 GCC ACC ATA AAT CAA GGA GGA AAA G1 64 5 814 T7

hand2 CCA TGA GTT TAG TTG GAG GGT TTC C 57 6 TCG GTT TGA TTC ATA ATG GGA CAA G2 64 3 849 T3

hsp90a. 1 TCT TTT GCG CTA CTA CAC TTC AGC TTC 58 9 ATA AAA TGC AAG AGC AGA CAC ACA AGG 64 4 956 T7

hsp90a.2 TGA AGA CTC GCA GAA CAG AAA GAA ACT 58 2 CTA AGC CCA GTT TGA TCA TCC TGT AGA1 65 3 740 T7

hsp90ab. 1 TCA TGA AGG AGA TTC TGG ACA AGA AAG 56.6 TTT TCC ATG CAG AAT AAA ATC ATG CTC1 63 0 775 T7

mef2a CAG AAT ACA ATG AAC CCC ACG AGA G 57 3 TTT AAT GTT GAC ATT CTG GCT GGT G2 64 6 905 T3

mef2ca AAT CTG ACT TAA TGC AAA TTT CAA GC 53 1 GAA TAA GGT CAA ATG ATT TTG GAC GAA1 63 3 981 T7

1 myf5 CAG GAG TGT CCA GTT TGC AGT GT 59 5 TTG TGG AGA TTT ATT ATT GGA AAA GCA 62 5 800 T7

1 ncxlh TCA AGC CAG AAA AGC ACT TGG TAT C 57 5 TTC AGC AAT ACG CCT CTC ATC TTT C 64 2 942 T7

2 pitx2 CCA GAC CAT GTT CTC TCC ACC TAA C 58 7 CAG AGC AA A AAC CAA GTT CAT TTC G 64 9 888 T3

1 serca2 AGA TAC TTG GCC ATC GGC TGT TAT G 58 9 GTA AAA GAC GGC CTG TTT CTC AAT G 64 5 933 T7

1 sox9a GCT CAG CAA AAC TCT GGG AAA AC 56 9 CAG CAT GGG TGT AAT AGG AGC TG 65 5 812 T7

1 unc4Sa ACT GGC AGA ACA GTC GAG AAC AG 58 9 TAG TCT AGG GGT GTC CAA ACT CG 64 7 816 T7

1. 5' end of reverse primer modified with T7 promoter sequence TAA TAC GAC TCA CTA TAG GG 2. 5' end of reverse primer modified with T3 promoter sequence CAT TAA CCC TCA CTA AAG GGA A

63 64

Table 2-3. Antisense Vector Based DIG-labelled RNA Probes

Restriction RNA Gene Vector Antibiotic Enzyme Polymerase

amhc pCRII-Topo AmpR BgUI SP6

cmlc2 pGEMT AmpR NotI T7

vmhc pGEMT AmpR Notl T7 Table 2-4. Antibodies and Stains Used

Antibody / Primary Antibody Secondary Antibody Source Stain Dilution Dilution Stain Reference

Barresi et ah, MF-20 DSHB 1:10 1:1,000* n/a 2001

Wohlgemuth DAB (MF-20) Sigma 1:10 1:300** n/a et ah, 2007

French et ah, 1:1,000* Zl-1 ZIRC 1:100 n/a 2009 Wohlgemuth Molecular et ah, 2007; Phalloidin 1:50 n/a n/a Probes French et ah, 2009 Walker and Alcian Blue Sigma n/a n/a 0.02% Kimmel, 2007 Engeman et Sigma n/a 0.5% Alizarin Red n/a ah, 2009 Dietrich et ah, n/a 0.6 mg/mL O-dianisidine Sigma n/a 1995 * anti-mouse Alexa 488 ** anti-mouse IgG horseradish peroxidase ECL

65 66

2.8 Figures

EcoRI

Figure 2-1. Representative dCAPS analysis genotyping gel electrophoresis. dCAPS analysis performed on genomic DNA extracted from embryos collected from the mating unc45b'A; unc45a~l~ parents. H = heterozygous, M = homozygous mutant, W = homozygous wild type. Restriction enzyme digestion of PCR products used to genotype unc45a (Ddel) and unc45b (EcoRI) mutant haplotypes. 67

2.9 References

Anderson, M. J., Pham, V. N., Vogel, A. M., Weinstein, B. M. and Roman, B. L. (2008). Loss of unc45a precipitates arteriovenous shunting in the aortic arches. Dev. Biol. 318, 258-267.

Barresi, M. J., D'Angelo, J. A., Hernandez, L. P. and Devoto, S. H. (2001). Distinct mechanisms regulate slow-muscle development. Curr Biol. 11, 1432- 1438.

Behra, M., Cousin, X., Bertrand, C, Vonesch, J. L., Biellmann, D., Chatonnet, A. and Strahle, U. (2002). Acetylcholinesterase is required for neuronal and muscular development in the zebrafish embryo. Nat. Neurosci. 5, 111-118.

Chen, J., Haffter, P., Odenthal, J., Vogelsang, E., Brand, M., van Eeden, F. J., Furutani-Seiki, M., Granato, M., Hammerschmidt, M., Heisenberg, C, Jiang, Y., Kane, D. A., Kelsh, R. N., Mullins, M. C. and Niisslein-Volhard, C. (1996). Mutations affecting the cardiovascular system and other internal organs in zebrafish. Development 123, 293-302.

Dietrich III, H. W., Kieran, M. W., Chan, F. Y., Barone, L. M., Yee, K., Rundstadler, J. A., Pratt, S., Ransom, D. and Zon, L. I. (1995). Intraembryonic hematopoietic cell migration during vertebrate development. Proc. Natl. Acad. Sci. USA 92, 10713-10717.

Engeman, J. M., Aspinwall, N. and Mabee, P. M. (2009). Development of the pharyngeal arch skeleton in Catostomus commersonii (Teleostei: Cypriniformes). JMorphol. 270, 291-305.

Etard, C. M., Fischer, N., Hutcheson, D., Geisler, R. and Strahle, U. (2007). The UCS factor Steif/Unc-45b interacts with the heat shock protein Hsp90a during myofibrillogenesis. Dev. Biol. 308, 133-143.

French, C. R., Erickson, T., French, D. V., Pilgrim, D. B. and Waskiewicz, A. J. (2009). Gdf6a is required for the initiation of dorsal-ventral retinal patterning and lens development. Dev. Biol. 333, 37-47.

Haffter, P., Granato, ML, Brand, M., Mullins, M. C, Hammerschmidt, M., Kane, D. A., Odenthal, J., van Eeden, F. J., Jiang, Y., Heisenberg, C, Kelsh, R. N., Furutani-Seiki, M., Vogelsang, E., Beuchle, D., Schach, U., Fabian, C. and Niisslein-Volhard, C. (1996). The identification of genes with unique and essential functions in the development of the zebrafish, Danio rerio. Development 123, 1-36. 68

Javidan, Y. and Schilling, T. F. (2004). Development of cartilage and bone. Methods Cell Biol. 76, 415-436.

Kimmel, C. B., Ballard, W. W., Kimmel, S. R., Ullmann, B. and Schilling, T. F. (1995). Stages of embryonic development of the zebrafish. Dev. Dyn. 203, 253-310.

Korbie, D. J. and Mattick, J. S. (2008). Touchdown PCR for increased specificity and sensitivity in PCR amplification. Nature Protocols 3, 1452- 1465.

Neff, M. M., Chory, J. and Pepper, A. E. (1998). dCAPS, a simple technique for the genetic analysis of single nucleotide polymorphisms: experimental applications in Arabidopsis thaliana genetics. Plant J. 14, 387-392.

Neff, M. M., Neff, E. T. and Kalishman, M. (2002). Web-based primer design for single nucleotide polymorphism analysis. Trends Genet. 18, 613-615.

Ransom, D. G., Haffter, P., Odenthal, J., Brownlie, A., Vogelsang, E., Kelsh, R. N., Brand, M., van Eeden, F. J., Furutani-Seiki, M., Granato, M., Hammerschmidt, M., Heisenberg, C, Jiang, Y., Kane, D. A., Mullins, M. C. and Niisslein-Volhard, C. (1996). Characterization of zebrafish mutants with defects in embryonic hematopoiesis. Development 123, 311-319.

Schilling, T. F., Piotrowski, T., Grandel, H., Brand, M., Heisenberg, C, Jiang, Y., Beuchle, D., Hammerschmidt, M., Kane, D. A., Mullins, M. C, van Eeden, F. J., Kelsh, R. N., Furutani-Seiki, M., Granato, M., Haffter, P., Odenthal, J., Warga, R. M., Trowe, T., and Niisslein-Volhard, C. (1996). Jaw and branchial arch mutants in zebrafish I: branchial arches. Development 123, 329-344.

Thisse C. and Thisse B. (2007). High-resolution in situ hybridization to whole- mount zebrafish embryos. Nat Protoc. 3, 59-69.

Uribe, R. A. and Gross, J. M. (2007). Immunohistochemistry on cryosections from embryonic and adult zebrafish eyes. Cold Spring Harbor Protocols 10.1101/pdb.prot4779.

Walker, M. B. and Kimmel, C. B. (2007). A two-color acid-free cartilage and bone stain for zebrafish larvae. Biotechnic & Histochemistry 82, 23-28.

Westerfield, M. (2000). The Zebrafish Book. A Guide for the Laboratory Use of Zebrafish (Danio rerio) (4th edn). Eugene: University of Oregon Press. 69

Wohlgemuth, S. L., Crawford, B. D. and Pilgrim, D. B. (2007). The myosin co-chaperone UNC-45 is required for skeletal and cardiac muscle formation in zebrafish. Dev. Biol. 303, 483-492. 70

CHAPTER THREE: RESULTS 71

Phenotypic assessment of the zebrafish unc45b''; unc45d' mutant is the requisite first step for determining whether the UNC-45 vertebrate paralogs Unc45a and Unc45b are functionally redundant in vivo. Results gathered from the characterization may have implications on our understanding of functional redundancy between the two vertebrate paralogs. I performed a detailed phenotypic characterization of the unc45b'A; unc45d/~ double mutant with a focus on the traits perturbed in the single mutants such as: gross morphology; cardiac morphology and function; muscle fibre determination and organization; and jaw formation. Detailed phenotypic characterizations of the single steif (unc45b~'~) and kus (unc45dl~) mutants have been reported previously (Etard et ah, 2007; Wohlgemuth et ah, 2007; Anderson et ah, 2008). Novel phenotypes not present in the single mutants, or a double mutant displaying phenotypes more severe than those of the single mutants, would be consistent with redundancy between the two vertebrate unc45 genes. If Unc45a and Unc45b are functionally redundant in vivo, the effects are likely to be found in regions where they are both expressed, such as in the pharyngeal arches and eyes. I would not however, expect to see changes in the heart, vasculature, or skeletal musculature, since these areas are severely affected in the individual unc45 mutants. My studies focused on three areas: morphology; gene compensation; and gene expression and differentiation.

3.1 Gross Morphology of unc45 Mutants

In regard to gross morphology, the unc45b'A; unc45dA and unc45b~A embryos appear identical (Figure 3-1). Blood circulates throughout the vasculature of wild type siblings and unc45dA embryos, as demonstrated by the presence of blood in the heart, whereas circulation is absent in the unc45b~' and unc45b"; unc45d~ mutants. The absence of both circulation and cardiac contraction in the unc45b" and unc45b"; unc45a" mutants, results in an accumulation of fluid in the pericardial space (cardiac edema) and by 5 dpf, yolk sac edema develops as well. To a lesser extent, the unc45d' mutants also develop 72 cardiac edema, but yolk sac edema does not advance to the degree seen in the other mutants. The swim bladder fails to inflate in unc45b" and unc45b"; unc45a" embryos due to the absence of circulation (Figure 3-1). Although the swim bladder inflates nominally in most of the unc45d' mutants, the level of inflation achieved is not enough to allow the embryos to remain in an upright position in the medium. unc45dA embryos are found to rest on their sides and therefore, from 4 dpf onwards, this is one of the criteria used to identify these mutants. Despite unc45b'A and unc45b~A; unc45dA embryos having the characteristic chevron shaped trunk muscle arrangement, somite birefringency is significantly reduced compared to wild type. Large, seemingly empty, regions develop in the centre of muscle segments, and rope like filaments accumulate at the myoseptum boundaries. Compared to wild type siblings, the muscles of unc45dA mutants appear identical with no loss in birefringency.

3.1.1 Circulation Defects

Circulation phenotypes have been reported for both the unc45d ' mutants and unc45b morphants (Wohlgemuth et ah, 2007; Anderson et ah, 2008). To test whether the unc45b"; unc45a" double mutant has circulation defects that are more severe than those of the single mutants — and therefore indicative of redundancy between unc45a and unc45b — I performed whole-mount o- dianisidine staining. This method permitted indirect examination of circulation defects and blood pooling through the observation of haemoglobin-containing erythrocytes. Staining confirms initial observations made in wild type siblings of erythrocytes circulating throughout the vessels of the head, heart, and trunk (Figure 3-2). As a result of an arteriovenous shunting in the aortic arches, unc45dA mutants demonstrate either a complete absence or a reduction in the degree of circulation in the trunk. The vascular malformation redirects blood from the heart, back into the head sinus, without first travelling throughout the trunk vasculature (Anderson et ah, 2008). unc45d ' mutants have the identical pattern of blood distribution as wild type embryos with the exception of an occasional 73 haemorrhage seen adjacent to the site of the arteriovenous malformation. unc45b~' and unc45b'A; unc45a" mutants lack cardiac contraction and so blood does not circulate throughout the head, heart, and trunk vasculature. Instead, blood pools in the dorsal aorta to the anterior of the embryo and in the caudal vein to the posterior. Staining then becomes concentrated in the regions of erythrocyte accumulation while absent in the head and heart (Figure 3-2). The circulation defects observed in the unc45b~A; unc45dA embryos are no more severe than those of the single unc45b'A mutants and therefore, these observations are inconsistent with functional redundancy between Unc45a and Unc45b in vivo.

3.1.2 Skeletal Muscle Organization

To determine if the muscle phenotype of the unc45b~A; unc45dA mutant was more severe than that reported in unc45b~A embryos, and therefore indicative of redundancy between unc45a and unc45b, I performed immunohistochemistry using the anti-myosin antibody MF-20. This antibody recognizes an epitope in the tail domain of the myosin heavy chain and labels both fast and slow muscles. At 4 dpf, both unc45b~' and unc45b'A; unc45d ' mutants have defective thick filament organization in the muscles of the trunk and lack the ordered striation patterning of the unc45dA and wild type embryos (Figure 3-3 A-D'). Mutants do not appear to have significantly lower myosin levels, but that which is present is arranged in a punctate pattern and distributed within wide myofibrils. Minor variations exist between the embryos of the unc45b mutants. Thin filaments, containing F-actin, were visualized using fluorescently conjugated phalloidin. Sarcomeres in the unc45dA and wild type embryos have a regular, striated pattern and are organized into uniformly spaced and sized myofibrils (Figure 3-3 E-H). The myofibrils of the unc45b~A and unc45b"; unc45dA mutants, however, varied in width, were wavy in appearance, and did not have a regular striated pattern. F-actin appears to accumulate at the myosepta boundaries in both the unc45b~ ' and unc45b~"; unc45d' embryos. 74

The nature and extent of the myofibril disorganization observed in the trunk musculature is similar between the unc45b~' and unc45b' ~; unc45d' mutants and therefore shows no redundancy between unc45a and unc45b.

3.1.3 Cranial Muscle Organization

Considering that unc45b~' and unc45b~'; unc45d ' mutants have defective myofibril organization in the trunk musculature, I examined the cranial musculature to see if this muscle population also contained gross morphological abnormalities or reduced myosin levels. I performed immunohistochemistry using the monoclonal antibody MF-20. Visualization was achieved in two ways by using: a fluorescent secondary antibody (Figure 3-4 A-D'), and a diaminobenzidine (DAB) colour reaction (Figure 3-4 E-H). No cranial muscles were missing in any of the mutants compared to wild type siblings (Figure 3-4). The sternohyoideus muscle was displaced laterally in unc45b" and unc45b"; unc45dA mutants and fibres appeared to be bowed in shape and packed in a less tight manner. Moreover, the adductor mandibulae had reduced DAB staining, even though the size of the muscle was unchanged. The severity of the unc45b'A; unc45dA mutant phenotype was similar to that of the unc45b'A embryos suggesting that the cranial musculature shares similar defects to those found in the trunk. Minor variations exist between the embryos of the unc45b mutants.

3.1.4 Skeletal Defects

Dysmorphic jaw structures have been reported in both the unc45d ' and unc45b" mutants, but those of unc45b" have not been examined in detail (Wohlgemuth et ah, 2007; Anderson et ah, 2008). The pharyngeal arches were of particular interest as both unc45a and unc45b are expressed in this region: unc45a within the arches and unc45b in the surrounding musculature (Wohlgemuth et ah, 2007; Anderson et ah, 2008). I was therefore interested to see if there were any additional jaw defects in the unc45b'~; unc45a" mutant and to determine whether the double mutant phenotype was similar to that of the unc45b~' mutants. 75

The majority of zebrafish skull bones develop indirectly through cartilaginous intermediates that can be visualized using Alcian Blue dye, which stains proteoglycan components of the extracellular matrix of chondrocytes (Schilling et ah, 1996). I performed Alcian Blue staining at 5 dpf and found that the unc45b'A and unc45b~'; unc45dA mutants have decreased staining in all of the pharyngeal arches (Figure 3-5). Compared to their wild type siblings, the unc45b'A and unc45b"; unc45d~ mutants — and to a lesser extent the unc45a" mutants — display improper angling of the ceratobranchial and ceratohyal cartilages and shortening of the palatoquadrates and Meckel's cartilage (Figure 3-5). The pectoral girdle is missing or reduced in the unc45b" and unc45b"; unc45d~ mutants and does not connect with the pelvic fins (Figure 3-5). Despite having a more severe phenotype than wild type embryos, no discernable differences were found between the unc45b~' and unc45b~'; unc45d' mutants. Minor variations exist between the embryos of the unc45b mutants. Zebrafish tooth development is initiated and sustained by retinoic acid and Fgf signaling pathways and occurs independently of pharyngeal arch formation (Jackman et ah, 2004; Gibert et ah, 2010). I was therefore interested to see if there were any dentition defects in the unc45 mutants. Zebrafish teeth are organized in three rostro-caudal rows, situated on each side of the 5th ceratobranchial arch and by 5 dpf, 2 or 3 teeth are present on both sides (Huyssenne et ah, 1998; Yelick and Schilling, 2002; Borday-Birraux et ah, 2006; Gibert et ah, 2010). To visualize pharyngeal teeth at 5 dpf I used Alizarin Red, a dye that binds free calcium ions and labels mineralized tissues such as tooth and bone (Javidan and Schilling, 2004). The unc45b" and unc45b"; unc45d' mutants do not stain as extensively as the unc45d' and wild type embryos, but no teeth are absent (Figure 3-5). The unc45d' mutants have no staining in the entopterygoid and maxilla and staining is reduced in the dentary, yet staining is similar to wild type for the rest of the cranial bones. The unc45b" and unc45b"; unc45a" mutants display an absence of staining in a number of the finer anterior structures namely, the dentary, entopterygoid, maxilla, and branchiostegal rays. Notably, staining is absent from the 5' ceratobranchial arch and notochord, which stain 76 intensely in wild type and unc45d' mutants. Staining is unchanged compared to wild type siblings in the cleithrum and opercle and is reduced in the parasphenoid. Although no teeth are absent in the unc45b'' and unc45b~'; unc45d ' mutants, the three teeth present at 5 dpf are either unstained or exhibit reduced staining compared to wild type. The extent of cranial skeleton mineralization in the unc45b'A and unc45b~A; unc45dA mutants is significantly less than that observed in unc45dA and wild type embryos. Although all teeth are present in the unc45b'A mutants, variation was observed amongst embryos with regards to the degree of mineralization. Minor variations exist between the embryos of the unc45b mutants.

3.1.5 Gross Ocular Morphology

unc45b'A mutants have been reported to have smaller eyes than their wild type siblings, yet this phenotype had not been quantified or examined in detail (Etard et ah, 2007; Wohlgemuth et ah, 2007). At 4 dpf, unc45b~ ' mutants have a visibly smaller ocular size compared to their wild type siblings and the lens does not protrude as extensively from the optic cup, suggesting that their field of view may be reduced (Figure 3-6 A-B'). The general shape of the eye and pupil, however, remain unchanged in the mutants. The difference in scale between the unc45b~' mutants and wild type embryos is evident in a camera lucida drawing created using the ventral view of 48 hpf embryos (Figure 3-6 C). The length and width of the wild type head is larger than that of the unc45b'A mutant and the eyes are much deeper. To quantify the differences in ocular size between wild type and unc45b'A embryos, I measured the pupil and eye diameters at 4 dpf. The average eye area, calculated using the whole eye diameter, is 58% smaller in unc45b'A mutants compared to their wild type siblings (P = 0.000, n = 23) (Figure 3-6 D). The average pupil diameter at 30% that of wild type, was also significantly smaller in the unc45b'A mutants (P = 0.000, n = 23) (Figure 3-6 E). I also examined the average eye area at 48 hpf to see if the differences in eye size were present from an earlier developmental stage as well as to compare sizes with those of the other unc45 mutants (Figure 3-6 F). There is no significant difference in 77 eye area between the wild type and unc45a" embryos (P = 0.998, n = 13) or the unc45b'A and unc45b'A; unc45dA mutants (P = 0.717, n = 14) using a Tukey ad hoc test (F(3,23) = 49.653). The eye area of the wild type and unc45dA embryos however, is significantly different to that of the unc45b'' and unc45b~'; unc45d ' mutants (P = 0.000). By quantifying the eye size of unc45 mutants, I was able to validate visual observations noting that the unc45b'A and unc45b'A; unc45dA mutants have significantly smaller eyes than those of unc45dA or wild type.

3.2 Gene Compensation in unc45b~/~; unc45a~/~ Mutants

3.2.1 unc45 Expression

In zebrafish, the expression patterns of unc45a and unc45b transcripts have been reported in both wild type and the respective mutants (Wohlgemuth et ah, 2007; Anderson et ah, 2008). The potential for compensation between the two, however has not been investigated. To determine whether there is compensation between the unc45 genes, I examined the mRNA expression patterns of unc45a and unc45b in the unc45dA, unc45b'A, and unc45b'A; unc45dA mutants in comparison to wild type siblings. If an unc45 gene had the capacity for compensation, then unc45b" mutants would be expected to have increased unc45a expression and conversely, unc45a" mutants would have increased unc45b expression compared to wild type. In the case of the double mutant, expression levels would be greater than wild type for both of the genes. As reported in Anderson et al. (2008), wild type embryos have a diffuse pattern of unc45a expression in the brain and pharyngeal arch region. I performed whole-mount in situ hybridization using an anti-sense unc45a mRNA probe and found my results concurred (Figure 3-7 A-D). Wild type siblings have robust expression throughout the head and pharyngeal arch region, whereas transcripts are almost absent in the unc45dA mutants. unc45b'A mutants have slightly lower expression levels compared to unc45b'A; unc45dA mutants, whose levels are similar to wild type. None of the mutants demonstrate an expansion of the expression domain of the unc45a transcript. 78

I next tested whether unc45b is also expressed differentially in the unc45 mutants (Figure 3-7 E-H"). Transcripts were detected in the extraocular, cranial, cardiac, pectoral fin, and trunk muscles of unc45d ' and wild type embryos. In contrast, expression was either present at low levels or absent in the extraocular, cranial, cardiac, and pectoral fin muscles of the unc45b'A and unc45b'A; unc45dA mutants. Expression in the trunk musculature however, is upregulated in these mutants as compared to wild type siblings. Expression levels of unc45 genes were diminished in their respective mutants except in the cases of the unc45b'A and unc45b'A; unc45dA mutants where unc45b expression was increased in the trunk musculature. Interestingly, unc45a transcripts were observed in the unc45b'A; unc45dA mutants (Figure 3-7 D) despite a lack of detectable expression in the unc45dA mutants (Figure 3-7 B). No evidence of compensation was seen in any of the mutants, as expression levels of the complementary genes did not exceed those of wild type for all genotypes.

3.2.2 hsp90 Expression

The chaperone hsp90a is co-expressed with unc45b in striated muscle and both proteins interact directly in vitro (Etard et ah, 2007). Both unc45b and hsp90a mRNA transcripts are upregulated in unc45b" mutants; however, expression of the hsp90 genes has not been examined in the unc45dA mutant (Etard et ah, 2007). I therefore examined the expression patterns of hsp90a.l, hsp90a.2, and hsp90ab.l genes in unc45 mutants to determine if a relationship, similar to the one seen in unc45b'A mutants, exists for unc45dA and unc45b'A; unc45dA mutants (Etard et ah, 2007). Whole-mount in situ hybridization performed at 48 hpf shows expression levels and patterning of the three genes to be identical for unc45dA mutants and wild type siblings (Figure 3-8). None of the hsp90 genes demonstrated an increase in transcript levels for these embryos. unc45b~A and unc45b~A; unc45dA mutants, in contrast, have increased expression of both hsp90a.l and hsp90a.2, but not of hsp90ab.l (Figure 3-8). Expression levels of both hsp90a.l and hps90a.2 not only increase, but expression expands into the trunk musculature and head region. 79

In accordance with a previous report, the expression levels and patterning of hsp90ab. 1 are analogous between wild type and unc45b" embryos (Etard et ah, 2007). Neither are any differences observed in wild type, unc45dA, or unc45b~A; unc45dA mutants. Therefore, hsp90ab.l expression remains unchanged in all of the unc45 mutants and hsp90a.l and hsp90a.2 expression increases in the unc45b '' and unc45b~"; unc45d' mutants, but remains constant in the unc45d ' and wild type embryos.

3.3 Gene Expression and Determination

3.3.1 Cardiac Determination

I examined the unc45 mutants for defective cardiac chamber formation, as the unc45b~' and unc45b~'; unc45d' mutants have neither circulation nor cardiac contraction. I did not expect to see an increase in the severity of cardiac phenotypes or any additional deformities in the unc45b'A; unc45d'~ mutants as compared to those previously reported in unc45b morphants (Wohlgemuth et ah, 2007). This is because, not only do the hearts of both mutants appear to be morphologically similar, but also the hearts of the unc45b'A mutants fail to complete cardiac looping, contributing to a significant cardiac malformation. I performed whole-mount in situ hybridization with the commonly used markers for cardiac chamber differentiation: vmhc, amhc, and cmlc2. Transcript levels and expression domains of vmhc and amhc are similar for wild type siblings, unc45d' , unc45b", and unc45b"; unc45a" mutants (Figure 3-9 A-H). cmlc2 expression domains are the same for wild type and unc45d' embryos, but atrial expression is increased in unc45b~' and unc45b~'; unc45d' mutants to levels similar to that of the ventricular chamber (Figure 3-9 I-L). These results indicate that not only is cardiac chamber fate and differentiation unchanged in the double mutant, but given that the phenotype is identical to that of the unc45b'' mutant, there is no evidence of redundancy between unc45a and unc45b. 80

3.3.2 Expression of the Myogenic Regulatory Factor myoD

In vitro data suggests that the two vertebrate UNC-45 homologues have distinct roles in myogenesis with unc45a being associated with myoblast proliferation and fusion and unc45b with fusion and sarcomere organization (Price et ah, 2002). Experiments in vivo, however, do not support this hypothesis, as myogenesis is unperturbed in unc45b morphants (Wohlgemuth et ah, 2007). This report only described myogenesis in the trunk musculature and did not examine the extraocular or pharyngeal arch musculature. The myogenic regulatory factor myoD is expressed in the branchiomeric and extraocular muscles and can be used to evaluate pharyngeal muscle formation (Lin et ah, 2006). I performed whole-mount in situ hybridization against myoD in 48 hpf embryos for two reasons. First, to determine if myogenesis is affected in the unc45a" and double unc45b"; unc45d~ mutants as they have not been examined previously and second, to establish whether myogenesis is disrupted in the head musculature of any of the mutants. Expression is unaltered between the wild type siblings, unc45dA, unc45b'A , and unc45b'A; unc45dA mutants in the cranial and extraocular muscle precursors to the: constrictor hyoideus, intermandibularis, inferior rectus, lateral rectus, medial rectus, superior rectus, pharyngeal arches, and sternohyoideus (Figure 3- 10 A-D). In the unc45b'A and unc45b'A; unc45dA mutants, expression is decreased in the precursors of the superior oblique (Figure 3-10 D). These embryos also exhibit a displacement of the intermandibularis, constrictor hyoideus, sternohyoideus, and pharyngeal muscle precursors to positions that are lateral and dorsal as compared to those of wild type siblings (Figure 3-10 A, C, D). In the trunk muscle precursors, unc45b~' and unc45b''; unc45d' mutants show elevated levels of myoD expression compared to unc45d' and wild type embryos with levels appearing to be similar between the two (Figure 3-10 A'-D'). The unc45dA and unc45b'A; unc45dA mutants display no disruption in myogenesis of the trunk musculature as demonstrated by unaltered myoD expression. The expression in the trunk musculature is the same for the unc45b" and unc45b"; unc45a" mutants. None of the mutants display aberrant myoD expression in the pharyngeal 81 and extraocular musculature; therefore, myogenesis is not disrupted in the head musculature of these embryos.

3.3.3 Pharyngeal Arch Formation and Patterning

I wished to determine whether the pharyngeal defects observed in the unc45b'A and unc45b~A; unc45dA mutants were linked to irregular pharyngeal arch formation and/or patterning. Using whole-mount in situ hybridization, there is no evidence for patterning defects in pharyngeal endodermal pouches (nkx2.3), pharyngeal mesenchyme (nkx3.2), pharyngeal neural crest (dlx2a, hand!), or tooth germ (Figures 3-11, 3-12). Moreover, the decreased Alcian Blue staining observed in the unc45b'A and unc45b'A; unc45dA mutants cannot be associated with defects in chondrogenesis as the levels of sox9a transcripts remain unchanged in the unc45dA, unc45b'A, and unc45b'A; unc45dA mutants, compared to wild type (Figure 3-12).

3.4 Conclusions

The goal of this thesis was to assess the phenotype of an unc45b~A; unc45a ~'~ double mutant. I have determined that the unc45b'A; unc45dA mutant is indistinguishable from that of unc45b'A. Both unc45b'A and unc45b'A; unc45dA mutants have cardiac, muscle, jaw, and eye defects. In contrast to unc45b, unc45a has no effect on hsp90 transcript levels, indicative of differential regulation of the two vertebrate unc45 genes. Based on the design of the studies presented in this thesis and the range of traits selected for observation, the data presented here are inconsistent with the hypothesis that Unc45a and Unc45b are functionally redundant in vivo. 82

3.5 Figures

Figure 3-1. Morphology of 4 dpf unc45 mutants, wild type siblings (A,A'); unc45a'- (B,B'); unc45b' (C,C); and unc45b'-, unc45cr/- (D,D') mutants. Blood circulates through the hearts of wild type siblings (A^V) and unc45cc'- mutants (B,B') (red) but not in those of the unc45b/- mutants (C,C»,D,D'). Cardiac edema is most pronounced in the unc45bA (C,C) and unc45b/-; unc45a' (D,D') mutants and absent in wild type siblings (AyV) (arrowheads). A fully inflated swim bladder is present only in wild type (AyV) embryos, absent in unc45b/ (C,C) and unc45bA; unc45or/' (D,D') mutants, and minimally inflated in unc45a'' (B,B') (arrows). Somite birefringency is reduced in unc45bA (C) and unc45b~A; unc45a~'' (D) mutants compared to wild type (A) and unc45a~/ (B) embryos. Apostrophies following a letter denote an increased magnification of the same embryo. 83

Figure 3-2. Whole-mount o-dianisidine staining of hemoglobinized erythrocytes reveals circulation defects and blood pooling in unc45 mutants at 3 dpf. Wild type siblings (AyV); unc45cr/' (B,B'); unc45b/~ (C,C); unc45b'\ unc45cr/ (D,D') mutants. Wild type siblings (A,A') and unc45a~/- mutants (B,B') have circulation throughout the vasculature of the trunk, heart, eyes, and head. unc45b~/~ (C,C) and unc45b~A; unc45a~/- (D,D') mutants do not have blood circulating the heart and head vessels (arrowheads). Instead, blood pools in the dorsal aorta towards the anterior and in the caudal vein at the ventral posterior of the embryo (arrows). Lateral (A-D) and ventral (A'-D') views allow for the examination of blood pooling and circulation through the heart, respectively and are alternative views of the same embryo. wt unc45ar/- unc45tr'- unc45brA; unc45ar-

Figure 3-3. unc45b mutants have defective thick and thin filament organization in striated trunk muscles. Wild type siblings (A,A',E); unc45cr/- (B,B',F); unc45bA (C,C',G); unc45bA; unc45a-A(D,D',t[) mutants. Myosin thick filaments, immunostained with the anti-myosin antibody MF-20 at 4 dpf, are disorganized in unc45b~A and unc45b~A; unc45a~'~mutants as seen at low 10X magnification (C,D) and high 40X magnification (C',D'), respectively. Wild type siblings (AyV) and unc45a~'' (B,B') mutants have normal thick filament organization. Staining of F-actin containing thin filaments with fluorescent phalloidin at 4 dpf (E-H). Wild type siblings (E) and unc45a~/~ (F) mutants have robust thin filament striatum patterning in contrast to unc45b~A (G) and unc45bA; unc45crA (H) mutants that lack a regular, organized striation. Apostrophies denote alternative views of the same embryo. wt unc45arA unc45brA unc45tr/; unc45aA

Figure 3-4. Myosin expression in cranial muscles. Wild type siblings (A^V',E); unc45crA (B,B',F); unc45bA (C,C',G); unc45bA; unc45aA (D,D',H) mutants. Lateral views at 4 dpf (A-D), ventral views at 4 dpf (A*- D') and 3 dpf (E-H) of embryos labelled with the MF-20 antibody. Expression is unaltered between wild type siblings (A,A',E) and unc45a~A (B,B*,F), unc45bA (C,C',G) or unc45bA; unc45cc'- (D,D',H) mutants. The sternohyoideus however, is displaced in unc45bA (C,C',G) and unc45b~A; unc45crA (D,D',H) embryos. Apostrophies denote alternative views of the same embryo, ao, adductor operculi; am, adductor mandibulae; do, dilator operculi; dpw 1-5, dorsal pharyngeal wall; hh, hyohyoideus; ih, interhyoideus; ima, intermandibularis anterior; imp, intermandibularis posterior; lap, levator arcus palatini; pp, pterygoid process; sh, sternohyoideus. 86

Figure 3-5. Skeletal defects in unc45 mutants at 5 dpf. Ventral views of Alcian Blue stained cartilages (A-D) and Alizarin Red stained mineralized bone (E-F). Wild type siblings (A,B); unc45a' (C,D); unc45bA (E,F); unc45bA; unc45ar/ (G,H) mutants. Wild type siblings and unc45a~A mutants, respectively, have robust cartilage (A,B) and bone staining (E,F) whereas unc45bA (C,G) and unc45b/; unc45aA (D,H) mutants exhibit decreased staining with shortening and improper angling of some cartilages. The pectoral girdle (arrows) is missing or reduced in unc45b' (C) and unc45b~/~; unc45a/ (D) mutants, br, branchiostegal rays; cb, ceratobranchial; ch, ceratohyal; cl, cleithrum; den, dentary; en, entopterygoid; hs, hyosymplectic; max, maxilla; mc, meckel's cartilage; nc, notochord; op, opercle; pq, palatoquadrate; ps, parasphenoid; t, tooth. 87

wt unc45a'- uncdStr^- uncdStr'-; unc45ar/- Figure 3-6. Analysis of eye phenotypes in unc45 mutants Lateral and dorsal views of representative 4 dpf wild type (A,A') and unc45b' embryos (B,B') unc45b mutants have visibly reduced ocular size and the lens does not protrude from the optic cup Camera lucida drawing depticting the difference in head size between wild type (thin line) and unc45b' (thick line) embryos at 48 hpf (C) Graphs showing the average eye area (D), and average pupil diameter (E) at 4 dpf and the average eye area (F) at 48 hpf unc45b' and unc45b', unc45a~' mutants show a significant reduction in eye area (P = 0 000) Error bars indicate Standard Error of the Mean 88

wt unc45arA unc45trA unc45tr'-; unc45aA

Figure 3-7. Whole-mount in situ hybridization of unc45a (A-D) and unc45b (E-H") mRNA expression at 48 hpf. Wild type siblings (A,E,E',E"); unc45crA (B,F,F',F"); unc45bA (C,G,G',G"); unc45b-A; unc45aA (D,H,H',H") mutants. unc45a expression is diffuse in the region of the brain and pharyngeal arches. Wild type siblings (A) have robust expression whereas detection of mRNA signal is almost absent in unc45a~A mutants (B). Surprisingly, the unc45bA; unc45a-A (D) mutants have slightly higher expression as compared to unc45bA mutants (C). Lateral (E-H), ventral (E'-H'), and dorsal (E"-H") views of unc45b mRNA expression. Wild type siblings (E,E',E") and unc45a-A (F,F',F") fish have expression in the extraocular, cardiac, trunk, and pectoral fin muscles. Expression is either absent or faint in the extraocular, cardiac, and pectoral fin muscles of the unc45bA (G,G',G") and unc45b'-\ unc45a-A (H,H',H") mutants whereas trunk muscle expression appears to up-regulated in these mutants. Apostrophies denote alternative views of the same embryo. 89

hsp90a.1 hsp90a.2 hsp90ab.1

Figure 3-8. hsp90a mRNA is up-regulated in unc45bA mutant embryos at 48 hpf. In situ hybridization with anti-sense probes for hsp90a.l (A,D,G,J), hsp90a.2 (B,E,H,K), and hsp90ab.l (C,F,I,L). Wild type siblings (A-C); unc45arA (D-F); unc45b-A (G-I); unc45bA; unc45crA (J-L) mutants. unc45bA and unc45bA; unc45arA mutants have increased expression of hsp90a.l (G,J) and hsp90a.2 (H,K) but not hsp90ab. 1 (I,L) as compared to wild type siblings (A,B,C) and unc45arA mutants (D,E,F). wt unc45aA unc45bA unc45tr/" unc45ar'-

Figure 3-9. Whole mount in situ hybridization of cardiac myosin genes. vmhc, ventricular myosin heavy chain (A-D) at 2.5 dpf; amhc, atrial myosin heavy chain (E-H) at 3 dpf; and cmlc2, cardiac myosin light chain 2 (I-L) at 2.5 dpf. Wild type siblings (A,E,I); unc45arA (B,F,J); unc45bA (C,G,K); unc45b/'\ unc45crA (D,H.L) mutants. Expression domains and levels of vmhc and amhc are the same for wild type siblings (A,E), unc45crA (B,F), unc45bA (C,G), and unc45bA; unc45aA (D,H) mutants. Levels and domains of cmlc2 expression are the same for wild type (I) and unc45aA (J) embryos but atrial expression is increased in unc45bA (K) and unc45b/-; unc45aA (L) embryos to levels similar to that of the ventricular chamber. Note the cardiac mislooping in unc45bA (K). v, ventricle; a, atrium. wt unc45arA unc45br'- unc45trf-; unc45arA

Figure 3-10. Whole-mount in situ hybridization of the myogenic regulatory factor myoD in cranial and trunk muscle precursors at 48 hpf. Wild type siblings (A,A'); unc45crA (B,B'); unc45b-A (C,C); unc45b-A; unc45crA (D,D') mutants. The expression pattern is unaltered in cranial muscle precursors of wild type siblings (A), unc45crA (B) or unc45b~A (C) mutants. Expression is decreased; however, in the precursors of the superior oblique in the unc45b~A; unc45aA mutant (D) (arrow). unc45b~A (C) and unc45b'A; unc45a~A (D') mutants show elevated levels of myoD in the trunk precursors in comparison to wild type siblings (A*) and unc45a~A mutants (B'). Apostrophies denote alternative views of the same embryo, chv, constrictor hyoideus ventralis; im, intermandibularis; ir, inferior rectus; lr, lateral rectus; mr, medial rectus; pam, pharyngeal arch muscles; sh, sternohyoideus; so, superior oblique; sr, superior rectus. wt unc45a-/- unc45bm/- unc45tr/-; unc45arA

Figure 3-11. Formation of pharyngeal arches in unc45 mutants. Localization of nkx3.2 transcripts by in situ hybridization is unaltered at 52 hpf. Wild type siblings (A,A'); unc45crA (B,B'); unc45bA (C,C); unc45bA; unc45aA (D,D') mutants. Lateral views (A-D) and ventral views (A'-D'). In situ hybridization of nkx2.3 mRNA at 36 hpf. Wild type siblings (E); unc45cr/- (F); unc45bA (G); unc45bA; unc45crA (H) mutants. Viewed dorsally, endodermal pouch patterning is unchanged in unc45b~A (G) and unc45b'~; unc45aA (H) mutants compared to wild type (E) and unc45aA (F) embryos. Apostrophies following a letter denote alternative views of the same embryo. wt unc45arA unc45br/- unc45tr'" unc45ar/-

Figure 3-12. Segmentation of pharyngeal arches in unc45 mutants. Wild type siblings (A,E,I,M); unc45cr/- (B,F,J,N); unc45bA (C,G,K,0); unc45bA; unc45aA (D,H?L?P) mutants. Localization of hand2 transcripts by in situ hybridization, dorsal views at 30 hpf of wild type siblings (A), unc45crA (B), unc45bA (C), and unc45bA; unc45aA (D) mutants. Expression of dlx2a and sox9a is unchaged between wild type siblings (E,I), unc45crA (F,J), unc45bA (G,K), and unc45bA; unc45a-A (H,L) embryos at 30 hpf and 48 hpf, respectively. Expression of pitx2a is unaltered in wild type (M); unc45crA (N); unc45b'/ (O); unc45bA; unc45crA (P) mutants. Arrowheads indicate pitx2a expression in the tooth germs, cb, ceratobranchial; p, pharyngeal arch. Numbers denote pharyngeal arch number. 94

3. 6 References

Anderson, M. J., Pham, V. N., Vogel, A. M., Weinstein, B. M. and Roman, B. L. (2008). Loss of unc45a precipitates arteriovenous shunting in the aortic arches. Dev. Biol. 318, 258-267.

Borday-Birraux, V., Van der heyden, C, Debiais-Thibaud, M., Verreijdt, L., Stock, D. W., Huysseune, A. and Sire, J. Y. (2006). Expression of dlx genes during the development of the zebrafish pharyngeal dentition: Evolutionary implications. Evol Dev. 8, 130-141.

Etard, C, Behra, M., Fischer, N., Hutcheson, D., Geisler, R. and Strahle, U. (2007). The UCS factor Unc45b-/-/Unc-45b interacts with the heat shock protein Hsp90a during myofibrillogenesis. Dev. Biol. 308, 133-143.

Francis, P. J., Berry, V., Moore, A. T. and Bhattacharya, S. (1999). Lens biology: Development and human cataractogenesis. Trends Genet. 15, 191- 196.

Gibert, Y., Bernard, L., Debiais-Thibaud, M., Bourrat, F., Joly, J., Pottin, K., Meyer, A., Retaux, S., Stock, D. W., Jackman, W. R., Seritrakul, P., Begemann, G. and Laudet, V. (2010). Formation of oral and pharyngeal dentition in teleosts depends on differential recruitment of retinoic acid signaling. FASEB J. 24, 3298-3309.

Greiling, T. M., Aose, M. and Clark, J. I. (2010). Cell fate and differentiation of the developing ocular lens. IOVS. 51, 1540-1547.

Hansen, L., Mikkelsen, A., Niirnberg, P., Niirnberg, G., Anjum, L, Eiberg, H. and Rosenberg, T. (2009). Comprehensive mutational screening in a cohort of Danish families with hereditary congenital cataract. IOVS. 50, 3291-3303.

Huyssenne, A., Van der heyden, C. and Sire, J. Y. (1998). Early development of the zebrafish (Danio rerio) pharyngeal dentition (Teleostei, Cyprinidae). AnatEmbryol. 198, 289-305.

Jackman, W. R., Draper, B. W. and Stock, D. W. (2004). Fgf signaling is required for zebrafish tooth development. Dev. Biol. 214, 139-157.

Javidan, Y. and Schilling, T. F. (2004). Development of cartilage and bone. Methods Cell Biol. 76, 415-436.

Lin, C.-Y., Yung, R.-F., Lee, H.-C, Chen, W., Chen, Y.-H. and Tsai, H. (2006). Myogenic regulatory factors Myf5 and Myod function distinctly during craniofacial myogenesis of zebrafish. Dev. Biol. 299, 594-608. 95

Price, M. G., Landsverk, M. L., Barral, J. M. and Epstein, H. F. (2002). Two mammalian UNC-45 isoforms are related to distinct cytoskeletal and muscle- specific functions. J Cell Sci. 115, 4013-4023.

Schilling, T. F., Piotrowski, T., Grandel, H., Brand, M., Heisenberg, C, Jiang, Y., Beuchle, D., Hammerschmidt, M., Kane, D. A., Mullins, M. C, van Eeden, F. J., Kelsh, R. N., Furutani-Seiki, M., Granato, M., Haffter, P., Odenthal, J., Warga, R. M., Trowe, T., and Niisslein-Volhard, C. (1996). Jaw and branchial arch mutants in zebrafish I: Branchial arches. Development 123, 329-344.

Soules, K. A. and Link, B. A. (2005). Morphogenesis of the anterior segment in the zebrafish eye. BMC Dev Biol. 5, 12.

Weber, G. F. and Menko, A. S. (2006). Actin filament organization regulates the induction of lens cell differentiation and survival. Dev. Biol. 295, 714-729.

Wohlgemuth, S. L., Crawford, B. D. and Pilgrim, D. B. (2007). The myosin co-chaperone UNC-45 is required for skeletal and cardiac muscle formation in zebrafish. Dev. Biol. 303, 483-492.

Yelick, P. C. and Schilling, T. F. (2002). Molecular dissection of craniofacial development using zebrafish. Crit Rev Oral Biol Med. 13, 308-322. 96

CHAPTER FOUR: DISCUSSION 97

In this thesis I have assessed the phenotype of a vertebrate unc45b"; unc45dA double mutant using the zebrafish (Danio rerio) model system. I performed a thorough phenotypic characterization of the unc45b'A; unc45dA mutant paying particular attention to abnormalities previously identified in the single mutants. By focusing on traits that had already been reported, I was able to determine whether the unc45b'A; unc45dA embryos have novel and/or more severe phenotypes than those of the single mutants. If either of the above were observed, it would be consistent with functional redundancy between the two vertebrate Unc45 proteins. In vertebrates, unc45a and unc45b are expressed in striated muscle and in vitro both are functionally redundant at the protein level. Previously functional redundancy of Unc45a and Unc45b had not been examined in vivo. In the phenotypic characterization of the unc45b''; unc45d' mutant I focused my attention on three areas: morphology, gene compensation, and gene expression and differentiation.

4.1 Are Unc45a and Unc45b Functionally Redundant in vivo?

Throughout this thesis I have presented experimental evidence that is inconsistent with the hypothesis that Unc45a and Unc45b are functionally redundant in vivo. This conclusion was reached after examining traits perturbed in the single mutants such as: gross morphology; cardiac morphology and function; muscle fibre determination and organization; and jaw formation. Of special interest were regions where the expression domains of unc45a and unc45b may overlap, such as in the pharyngeal arches during later stages of development. No phenotype was observed that was novel to the unc45b'A; unc45dA mutants. Nor did the mutants possess phenotypes that were more pronounced than those of either of the unc45dA or unc45b~A mutants alone. Since mammals and teleosts have two Unc45 genes, show no evidence of functional redundancy, and are involved in distinct cellular processes, one can ask the question: how did these different functions arise? The UNC-45 gene pair was most likely created by a whole genome duplication event sometime after the vertebrate lineage branched 98 from that of C. elegans and D. melanogaster. Both genes would have been redundant until mutations accumulated following the duplication event. As depicted in Figure 1-1, duplicate gene pairs have a number of potential fates: become a pseudogene, subdivide the ancestral function at the regulatory or protein level, develop a new function, or a combination of subfunctionalization and neofunctionalization. Both vertebrate Unc45b and C. elegans UNC-45 are required for myosin motor domain folding and thick filament assembly and are expressed in similar tissues. This suggests that few changes have occurred in the protein domains and regulatory elements responsible for producing the muscle myosin functions of UNC-45 and Unc45b. As the expression pattern and functions of Unc45a appear to have diverged significantly from those of Unc45b in vertebrates, it is likely that expression domain subfunctionalization and neofunctionalization occurred in the Unc45a copy prior to the branching of the bony fish lineage. Had the new Unc45 function not been established after the divergence of lobe and ray finned fish, either mammals or teleosts, but not both, would have the gene copy encoding a new Unc45 function. Functional redundancy between Unc45a and Unc45b may not have been observed in vivo because of maternal mRNA and protein having a role in early embryogenesis. I do not, however, believe this to be the case for the unc45b'A; unc45dA mutant. Maternal gene products that are synthesized and deposited into the egg during oogenesis play a role in early embryonic development and patterning (Ciruna et ah, 2002; Abrams and Mullins, 2009). For those genes that have both maternal and zygotic expression, the maternal contribution may partially compensate for a loss of the zygotic (Ciruna et ah, 2002). The presence of maternal expression in itself, however, is not a definitive indication of maternal effect or functional redundancy between the maternal and zygotic gene functions (Ciruna et ah, 2002). I argue that this is the case for unc45a and unc45b. In zebrafish, zygotic mRNA transcripts for both unc45a and unc45b have been identified by in situ hybridization yet only unc45a has detectable maternal expression (Wohlgemuth et ah, 2007a; Anderson et ah, 2008). Maternal transcripts of unc45b are, however seen from the 1-2 cell stage onwards when 99 tested for by using the more sensitive method of RT-PCR (Wohlgemuth, 2007b). Given that the levels of unc45b must be significantly lower than those of unc45a at early developmental stages, it is possible that Unc45b plays no role maternally and therefore, the unc45b" phenotype is representative of a loss of unc45b function. By this reasoning, it is equally unlikely that Unc45a has a role in early development as previous work has shown that the use of translation blocking morpholino oligonucleotides produced phenotypes identical to those of the zygotic unc45dA mutant. Once antibodies are available for both of the Unc45 proteins, it will be possible to determine whether maternal protein products are deposited into the embryo and help to resolve the issue of maternal-effect. In Caenorhabditis elegans, UNC-45 interacts with NMY-2, a non-muscle myosin type II known to play a role in embryonic cytokinesis and as a result, C elegans unc-45 non-sense alleles are embryonic lethal due to a failure to complete cytokinesis (Venolia and Waterston, 1990; Kachur et ah, 2004). The Drosophila homolog of UNC-45, dUNC-45 has also been shown to co-localize with a non- muscle myosin at a stage prior to the expression of muscle myosin, and a mutation results in zygotic lethality (Lee et ah, 2011). These observations, in conjunction with those from the UCS protein family member Rng3, which also interacts with a type II myosin and is required for cytokinesis, led us to hypothesize that the vertebrate paralogs of UNC-45 may have similar functions. A cytokinesis phenotype, however, had not been reported in either of the single unc45 mutants in zebrafish (Wohlgemuth et ah, 2007a; Anderson et ah, 2008). Our expectation was that the double mutant might present with a novel phenotype indicative of cytokinesis defects. No such phenotype was observed. One explanation may be that a protein other than Unc45 carries out the myosin chaperone activity for non-muscle myosin II in zebrafish. This is unlikely, however in light of the studies performed in vitro, which strongly support a role for Unc45a in relation with non-muscle myosin II. 100

4.2 unc45b" and unc45b"; unc45a" Mutants Share Similar Cardiovascular Defects

Identical patterns of blood pooling, non-discernable cardiac function, a global lack of circulation, and incomplete cardiac looping are observed in the unc45b'A and unc45b'A; unc45dA mutants (Figure 3-1, 3-2). This is in agreement with previous findings that established unc45b'A mutants to have a complete absence of cardiac contraction and circulation (Etard et ah, 2007). These phenotypes are not the result of an earlier downstream perturbation in cardiac chamber determination or specification as the expression levels and domains of cardiac myosin genes were comparable between all genotypes tested (Figure 3-9) (Wohlgemuth et ah, 2007a). Several genetic and epigenetic factors may account for the deficient cardiac looping observed in the unc45b mutants, namely: left-right asymmetry, cardiac contraction, conduction, and fluid flow. Correct cardiac morphogenesis and looping requires the establishment of distinct left and right identities in the embryo with heart looping being the first visible indication of left-right asymmetry (Kathiriya and Srivastava, 2000). Mutations in genes required for the pathway that creates left-right axis identity affect organ laterality and cardiac looping (Kathiriya and Srivastava, 2000). It is unlikely that defects in left-right patterning play a role in the cardiac phenotypes observed in the unc45b mutants. Since Unc45b functions as a myosin chaperone, and is expressed exclusively in the cardiac and skeletal musculature, it is unlikely that Unc45b could have a role in these patterning events. In addition, changes in mRNA expression patterns were only observed in limited cases in unc45dA, unc45b'A, or unc45b'A; unc45dA mutant embryos. The genes that did demonstrate an increase in expression levels are known to interact directly with Unc45b and their increase is thought to be due in part to a stress response. Therefore, it is unlikely that Unc45 has any effect on the expression of genes with which it is not co-regulated. The importance of epigenetic factors such as cardiomyocyte contraction and conduction as well as intracardiac haemodynamic flow on cardiac morphogenesis are beginning to be recognized as integral to the developmental 101 process (Hove et ah, 2003; Chi et ah, 2010). Using zebrafish as a model, Chi et ah (2010) found that electrical conduction in the absence of cardiac contraction and blood flow affected cardiomyocyte morphology and morphogenesis. This factor is not anticipated to account for the looping defects seen in the unc45 mutants as a recent report has demonstrated that the electrical gradient of the outer and inner curvature of unc45b'A hearts is comparable to that of wild type (Panakova et ah, 2010). Without the tools available to differentiate the effects of cardiac contraction from those of haemodynamic forces, I cannot be certain which are involved in the unc45dA, unc45b'A, and unc45b~A; unc45dA heart phenotypes. Circulation may play the larger role as the unc45dA mutants, that have no muscle defects, display improper cardiac looping, albeit not as severe as that seen in the unc45b'A and unc45b'A\ unc45dA mutants (Figure 3-9). It is probable that even though circulation is present in unc45dA embryos the fluid forces produced in the heart may be greater, or directed differently to those in wild type embryos. This is likely the result of the arteriovenous shunting in the aortic arches that redirects blood from the heart, back to the head sinus, without first travelling throughout the trunk vasculature (Anderson et ah, 2007). Circulation is also important in the formation of other organ systems in zebrafish, notably the swim bladder, which is affected in the unc45 mutants (Figure 3-1). The swim bladder is a buoyancy organ in fish and shares a common evolutionary origin with the tetrapod lung (Teoh et ah, 2010). Blood flow is required to sustain swim bladder growth, and although circulation is not necessary for organ budding, it is subsequently needed from the early growth stage onwards as well as during inflation (Teoh et ah, 2010; Winata et ah, 2010). Variation is observed in the magnitude of swim bladder inflation within the unc45dA mutants. A plausible explanation for the variation is that not all unc45dA embryos lack circulation in the trunk vasculature, although the quantity and velocity of blood flow is significantly reduced as compared to wild type siblings. The complete absence of circulation in the unc45b'' and unc45b~'; unc45d' embryos could also account for why the swim bladders fail to inflate in these mutants. 102

4.3 unc45b" and unc45b"; unc45a" Mutants Have Identical Muscle Phenotypes

Extraocular and pharnyngeal arch muscles comprise the cranial musculature and are derived from paraxial mesoderm (Schilling and Kimmel, 1997). The process of craniofacial myogenesis is influenced by the myogenic regulatory factors Myf5 and MyoD (Schilling and Kimmel, 1997; Lin et ah, 2006; Hinitis et ah, 2009). Previously, myogenesis had only been examined in the skeletal muscles of unc45b" mutants. Cranial myogenesis was normal in all the unc45 mutants as myoD expression levels were unchanged and no differences were observed between the unc45b" and unc45b~~; unc45a" mutants (Figure 3- 10). The observed displacement of four muscle precursor groups in the unc45b" and unc45b"; unc45d~ mutants can be attributed to the pericardial edema that begins to accumulate by 48 hpf and consequently leads to structures being shifted both laterally and ventrally compared to in wild type embryos. Surprisingly, myoD expression levels were increased in the trunk muscle precursors of the unc45b~' and unc45b~"; unc45d' mutants compared to wild type siblings and unc45a" mutants (Figure 3-10). The elevated levels of myoD may reflect the similar increase in hsp90a levels observed in the unc45b'A and unc45b~A ; unc45dA mutants (Figure 3-8). A subset of cells located in the somites and pectoral fin buds express both myoD and hsp90a and both genes are down regulated subsequent to the establishment of striated fibres (Krone et ah, 1997; Leleera/., 1999). The trunk muscles of unc45b'A; unc45dA mutants display identical myofibril organization defects to those observed in the unc45b'A mutants (Figure 3-3) (Wohlgemuth et ah, 2007a; Anderson et ah, 2008). Functional myosin-actin interactions are required for thin filament assembly. When myosin assembly or organization is defective, the result is a disruption in thin filament assembly and a decrease in actin protein levels. A similar observation has been made previously in zebrafish embryos treated with BTS (N-benzyl-p-toluene sulphonamide), a chemical that inhibits myosin ATPase activity and causes the dissociation of the myosin heads from actin filaments (Codina et ah, 2010). Primary cultures of chick 103 embryonic skeletal muscle treated with BTS displayed decreased thick and thin filament organization as well as disordered alpha-actinin and titin structures (Kagawa et ah, 2006). Together these results suggest that proper actin-myosin interaction is required for correct myofibrillogenesis. Previously, myofibril organization had only been examined in the trunk musculature of unc45b'A mutants. Although craniofacial myogenesis was normal in the unc45 mutants, I examined the muscle fibre arrangement in older embryos to see if any abnormalities were present (Figure 3-4). No differences were observed between the unc45 mutants and wild type except for the lateral displacement of the sternohyoideus in the unc45b'A and unc45b'A; unc45dA mutants. The modified placement of muscle groups is similar to the myoD expression observed at 48 hpf and is likely due to the pericardial edema that develops in the unc45b~A and unc45b~A; unc45dA mutants. The combination of muscle phenotypes observed in the unc45 mutants indicates that craniofacial myogenesis is unperturbed in the unc45 mutants and that the unc45b'A and unc45b'A; unc45dA mutants have an equal degree of myofibril disorganization.

4.4 Pharyngeal Arch Cartilage and Bone Mineralization is Disrupted in unc45 Mutants

The pharyngeal arches form through the close interaction of neural crest, mesoderm, and endoderm tissues (Hong et ah, 2005). All unc45 mutants have abnormal pharyngeal cartilages at 5 dpf, but the arches of unc45b'' and unc45b~'; unc45d' mutants are more severely disrupted than those of the unc45d' mutants (Figure 3-5). The observed phenotypes could be the result of segmentation or patterning defects or originate subsequent to the accumulation of fluid in the pericardial cavity. No abnormal pharyngeal patterning or segmentation was observed in the endodermal pouches, pharyngeal neural crest, or pharyngeal mesenchyme (Figure 3-11, 3-12). The decreased Alcian blue staining seen in the unc45b'A and unc45b"; unc45a" mutants could not be attributed to defects in chondrogenesis as sox9a expression was normal compared to wild type (Figure 3- 12). 104

Since no changes in gene expression were observed, it is likely that the cartilage defects of the unc45 mutants are associated with pericardial edema. Other zebrafish mutants (for example violet bauregarde and tnnt2 morphants) that develop edema exhibit the same pharyngeal arch defects. Evidence in support of edema as the origin of a set of universal abnormal pharyngeal arch phenotypes is compelling, yet it remains to be examined whether the interplay between muscle contraction and cartilage formation is a requirement for proper jaw organization and development. Although the unc45b" and unc45b"; unc45a" mutants accumulate fluid to a much larger extent than the unc45dA mutants, it seems unlikely that edema alone could account for the decreased cartilage staining observed in the unc45b mutants (Figure 3-5). Schilling and Kimmel (1997) support this idea by proposing that despite the craniofacial muscles differentiating slightly later than cartilages in the same region, the scaffold created through the interaction between the muscle and cartilage precursors might play a role in the patterning of the pharyngeal region. Tooth induction and pharyngeal arch patterning occur as two separate events with teeth developing independently of the pharyngeal arches (Gibert et ah, 2010). Having observed cartilage abnormalities in the unc45b" and unc45b"; unc45dA mutants, I decided to examine the mineralization of teeth and bones found in the cranial region of the unc45 mutants. The cranial skeleton of the unc45b~A and unc45b'A; unc45dA mutants did not stain as extensively with Alizarin Red as compared to unc45dA and wild type embryos (Figure 3-5). Three teeth were present in all unc45 mutants and wild type embryos, however those of the unc45b'A and unc45b'A; unc45dA mutants were either unstained or the staining was limited to the tip of the tooth. As tooth development proceeds from tip to base it might be that the teeth of unc45b'A and unc45b'A; unc45dA mutants are not as developmentally advanced as those in unc45dA and wild type embryos (Borday- Birraux et ah, 2006). No change in gene expression was seen for two tooth bud markers, dlx2a and pitx2a, suggesting that a delay in tooth initiation could not account for the observed reduced mineralization of the unc45b~' and unc45b~"; 105 unc45a" mutants (Figure 3-12) (Jackman et ah, 2004; Borday-Birraux et ah, 2006). Retinoic acid is an important inducer of pharyngeal tooth development in zebrafish, and changes in retinoic acid signaling can result in the loss of tooth induction (Gibert et ah, 2010). Interestingly, I found no dentition defects in the unc45a" mutants. This was unexpected as data from in vitro experiments demonstrate that GCUNC-45, the human homologue of Unc45a, inhibits both signaling through the retinoic acid receptor a and induction of endogenous retinoic acid receptor target genes (Epping et ah, 2009).

4.5 unc45b'/' Mutants have Small Eyes

unc45b" mutants have been reported to have much smaller eyes than their wild type siblings (Etard et ah, 2007; Wohlgemuth et ah, 2007). Since it was not obvious how a muscle defect might lead to smaller eyes, I felt this warranted more examination. At 4 dpf, the eyes and pupils of unc45b~' mutants are significantly smaller than the eyes of wild type embryos and the lens does not protrude as extensively from the optic cup (Figure 3-6). Additionally, at 48 hpf, the eye areas of unc45b'' and unc45b~'; unc45d' mutants are significantly smaller than those of unc45dA and wild type embryos (Figure 3-6). An explanation for the small head and eyes of the unc45b'A and unc45b'A; unc45dA mutants is a lack of circulation throughout development. Other zebrafish cardiovascular system mutants such as heart and soul and pandora, also have small eyes and lend support to this conclusion (Stainier et ah, 1996). Of note, no eye phenotypes were detected in the unc45d' mutants whose circulation is functional in the head.

4.6 Different hsp90 Responses are Present in the unc45a~/~ and unc45b~/~ Mutants

The up-regulation of Hsp90 related genes and the concomitant proteosome-mediated degradation of Hsp90 client proteins are a hallmark of diminished Hsp90 function (Proisy et ah, 2006). In zebrafish, Unc45b can interact 106 with all three Hsp90 proteins and the striking similarity of the hsp90a.V~ and unc45b'A mutant phenotypes implies that both play a role in myofibrillogenesis (Etard et ah, 2007; Du et ah, 2008; Hawkins et ah, 2008). The double unc45b'A; unc45dA mutants have increased expression of hsp90a.l and hsp90a.2, similar to what I observed in the unc45b'' mutants (Figure 3-8). This observation, together with the identical expression patterns of the hsp90a.l, hsp90a.2, and unc45b genes suggests that all three are co-regulated in skeletal muscle. Whether hsp90 expression is altered in the unc45d' mutants, however, had never been examined. Given that the expression domains of hsp90a.l, hsp90a.2, and unc45a do not coincide, the unchanged expression levels of unc45a" embryos was not unexpected. unc45a and hsp90ab.l have matching expression patterns, but hsp90ab. 1 expression remained constant in all genotypes tested. This is consistent with previous findings that hsp90ab.l is constitutively expressed and, unlike hsp90a, is unresponsive to cellular stress (Du et ah, 2008). My observations are the first to examine the expression patterns of hsp90 genes in the unc45d' mutant. I have demonstrated that in contrast to unc45b, unc45a has no effect on hsp90 transcript levels. This is indicative of not only differential regulation of the unc45 genes but also a divergence in function between Unc45a and Unc45b.

4.7 Future Directions

Developments made in the last two decades have increased our knowledge of the function and importance of Unc45 in myosin motor domain folding and myofibrillogenesis, yet many questions remain unanswered. I believe that the following three questions address critical issues that must be resolved in order for us to fully comprehend the roles played by Unc45 in vertebrates.

4.7.1 Do Unc45a or Unc45b Interact with a Non-Muscle Myosin?

Two lines of evidence suggest that either one or both of the vertebrate UNC-45 proteins directly interacts with non-muscle myosin II. First, the UCS proteins found in most non-vertebrates function in processes, such as cytokinesis, 107 which require a non-muscle myosin (Kachur et ah, 2004; Lee et ah, 2011). It is conceivable then that Unc45a is the vertebrate protein that performs this role. In an in vitro folding assay Liu et al. (2008) demonstrated that in comparison to Unc45b, Unc45a has a higher affinity for the smooth muscle myosin motor domain and also achieves a greater degree and efficiency of folding for this region. Importantly, smooth muscle myosin is closely related to non-muscle myosin II (Liu et ah, 2008). Second, numerous models have been proposed that try to encompass the multitude of proteins and apparent steps involved in myofibrillogenesis, but only the premyofibril model can account for the most recent data from in vivo studies (Sanger et ah, 2010). In this model, at an earlier stage before mature myofibril formation, premyofibrils are assembled containing non-muscle myosin II. Unc45b may be involved in the folding and incorporation of this myosin population before it is replaced permanently by muscle myosin. The first step in resolving the issue of whether Unc45a or Unc45b interacts with a non-muscle myosin would be to create an epitope-tagged Unc45a transgene. Co-localization studies could then be conducted to determine whether Unc45 proteins localize with non-muscle myosin in vivo, and if so, are they involved in the processes mentioned above. To test for transgene efficacy, rescue experiments should be conducted by injecting the Unc45a transgene into the unc45d ' mutants. If the transgene can rescue the wild type cellular function of Unc45a then the injected embryos will have a wild type phenotype. The transgene could also be used to determine if Unc45a is present in the early embryo. If so, the creation of a maternal-zygotic unc45a zebrafish line could be warranted to investigate the involvement of Unc45 in early developmental events such as cytokinesis. Previous in vitro studies have shown that Unc45b and Hsp90a.l can interact directly (Etard et ah, 2007). In order to see if the same is true for Unc45 and non-muscle myosin, in vitro pull-down assays could be performed using tagged proteins. Despite the possibility that the results from this test may not provide an accurate account of what occurs in vivo, it may offer some indication as to whether there is potential for an interaction to occur between Unc45 and non-muscle myosin. The data gathered from these critical experiments will not 108 only dictate if further examination of non-muscle myosin interactions is warranted in the unc45 mutants but will also provide insight into whether vertebrate Unc45 proteins have retained an important ancestral function.

4.7.2 What is the Function of Unc45a During Zebrafish Development?

The abnormal aortic arch development and arterioventricular malformations seen in the unc45d' mutants are puzzling as data from cell culture studies strongly suggest that Unc45a may have a role in cell division and proliferation mediated by an interaction with non-muscle myosin II. In humans, Unc45a inhibits retinoic acid signaling and its overexpression in tumors is associated with increased proliferation and metastasis (Bazzaro et ah, 2007; Epping et ah, 2009). This suggests that Unc45a is involved in important cellular functions that have yet to be confirmed in vivo. The expression domain of unc45a in zebrafish is significantly reduced compared to mouse, but the current resolution of unc45a gene expression by in situ hybridization is insufficient to determine a precise expression domain (Price et ah, 2002). An understanding of the localization of mRNA transcripts or protein distribution would therefore provide some indication as to the role of Unc45a. Creating a transgenic zebrafish line expressing green fluorescent protein under the control of the unc45a promoter may help to resolve this issue. It is also necessary to identify the proteins that interact directly with Unc45a to produce the highly specific phenotype of the unc45d ' mutants. To identify such partners a yeast-two hybrid screen could be performed. This analysis has the potential to produce an abundance of client proteins that need to be validated by biochemical and genetic tests (e.g., co-immunoprecipitation, morpholino knockdown, and co-localization) to eliminate those that are not biologically relevant. The remaining proteins would then have to be further characterized to determine how they function in association with Unc45a and in turn how this relates to aortic arch formation and/or other suspected roles of Unc45a. The yeast two-hybrid interaction screen approach has been successful in 109 finding proteins that interact with UNC-45. Examples include NMY-2 and HUM- 2 in C. elegans and Apobec2 in zebrafish (Kachur et ah, 2004; Etard et ah, 2010).

4.7.3 Does Unc45b Have a Role in Cardiac Muscle Maintenance?

Unc45b is necessary for proper cardiac myofibril organization and subsequent cardiac contraction (Etard et ah, 2007). A recent report suggests that in addition to myofibrillogenesis, Unc45b may play a role in cardiac muscle maintenance and regeneration. Stanley (2007) performed proteomic analysis on cardiac tissue from individuals with end-stage heart failure and found a 1.4 fold increase of UNC45B in ischemic cardiomyopathy patients and a 1.2 fold increase in patients with dilated cardiomyopathy. Since imaging of the heart using antibody staining is often obstructed by the yolk sac or severe edema, transgenic lines that produce a stronger fluorescent signal should be used. Transgenic lines commonly employed in cardiac research using the zebrafish model are Tg(cmlc2:eGFP) and Tg(cmlc2:dsRed Exp-nuc). Crossing these lines with the unc45b'A mutants would facilitate the visualization of cardiac morphology as well as the tracking of changes in the number of cardiac chamber cells. Information gathered from these experiments would add to our understanding of the affect Unc45b function has on cardiac morphogenesis and help in directing further projects. Both Unc45b, and Hsp90a.l are required for correct myofibril assembly and organization during trunk myofibrillogenesis, but hsp90a.l mutants have normal cardiac contraction (Du et ah, 2008; Hawkins et ah, 2008). This suggests that Unc45b interacts with a different Hsp90 protein in cardiac tissue, as to in trunk muscle, to achieve myofibril organization. It is likely that Hsp90a.2 and Hsp90ab.l interact with Unc45b in the heart, as both genes must be knocked down before cardiac defects develop (Du et ah, 2008). Since Unc45b cannot fold the myosin motor domain on its own, it is imperative that other proteins involved in the process are identified such that the mechanism of cardiac myosin assembly can be established. The proposed experiments would not only increase our 110 understanding of Unc45b function but could have serious implications for the management and care of serious cardiac diseases.

4.8 General Conclusions

In this thesis, I demonstrate that there is no evidence for redundancy between Unc45a and Unc45b in the zebrafish, Danio rerio. After examining traits perturbed in the single mutants such as: gross morphology; cardiac morphology and function; muscle fibre determination and organization; and jaw formation, I have determined that the unc45b~A; unc45dA mutants have neither unique or novel phenotypes nor are those observed more pronounced than those of either the unc45dA or unc45b'A mutants alone. Together, this data does not support functional redundancy between Unc45a and Unc45b in vivo. Ill

4.9 References

Abrams, E. W. and Mullins, M. C. (2009). Early zebrafish development: it's in the maternal genes. Curr Opin Genet Dev. 19, 396-403.

Anderson, M. J., Pham, V. N., Vogel, A. M., Weinstein, B. M. and Roman, B. L. (2008). Loss of unc45a precipitates arteriovenous shunting in the aortic arches. Dev. Biol. 318, 258-267.

Bazzaro, M., Santillan, A., Lin, Z., Tang, T., Lee, M. K., Bristow, R. E., Shih, L. and Roden, R. B. (2007). Myosin II co-chaperone general cell UNC-45 overexpression is associated with ovarian cancer, rapid proliferation, and motility. Am J Pathol. Ill, 1640-1649.

Borday-Birraux, V., Van der heyden, C, Debiais-Thibaud, M., Verreijdt, L., Stock, D. W., Huysseune, A. and Sire, J.-Y. (2006). Expression of Dlx genes during the development of the zebrafish pharyngeal dentition: evolutionary implications. EvolDev. 8, 130-141.

Brookfield, J. F. (1997). Genetic redundancy. An Advances in Genetics (ed. J. C. Hall, T. Friedmann, J. C. Dunlap and F. Giannelli), pp. 137-155. Academic Press.

Chi, N. C, Bussen, M., Brand-Arzamendi, K., Ding, C, Olgin, J. E., Shaw, R. M., Martin, G. R. and Stainier, D. Y. R. (2010). Cardiac conduction is required to preserve cardiac chamber morphology. Proc Natl Acad Sci USA. 107, 14662-14667.

Ciruna, B., Weidinger, G., Knaut, H., Thisse, B., Thisse, C, Raz, E. and Schier, A. F. (2002). Production of maternal-zygotic mutant zebrafish by germ-line replacement. Proc Natl Acad Sci USA. 99, 14919-14924.

Codina, M., Li, J., Gutierrez, J., Kao, J. P. Y. and Du, S. J. (2010). Loss of Smyhcl or Hsp90al function results in different effects on myofibril organization in skeletal muscles of zebrafish embryos. PLoS ONE 5, e8416.

Du, S. J., Li, H., Bian, Y. and Zhong, Y. (2008). Heat-shock protein 90a 1 is required for organized myofibril assembly in skeletal muscles of zebrafish embryos. Proc Natl Acad Sci USA. 105, 554-559.

Easter, S. S. Jr. and Nicola, G. N. (1996). The development of vision in the zebrafish (Danio rerio). Dev Biol. 180, 646-663.

Epping, M. T., Meijer, L. A. T., Bos, J. L. and Bernards, R. (2009). UNC45A confers resistance to histone deacetylase inhibitors and retinoic acid. Mol Cancer Res. 1, 1861-1870. 112

Etard, C, Behra, M., Fischer, N., Hutcheson, D., Geisler, R. and Strahle, U. (2007). The UCS factor Steif/Unc-45b interacts with the heat shock protein Hsp90a during myofibrillogenesis. Dev. Biol. 308, 133-143.

Etard, C, Roostalu, U. and Strahle, U. (2010). Lack of Apobec2-related proteins causes a dystrophic muscle phenotype in zebrafish embryos. J Cell Biol. 189, 527-539.

Gibert, Y., Bernard, L., Debiais-Thibaud, M., Bourrat, F., Joly, J.-S., Pottin, K., Meyer, A., Retaux, S., Stock, D. W., Jackman, W. R., Seritakul, P., Begemann, G. and Laudet, V. (2010). Formation of oral and pharyngeal dentition in teleosts depends on differential recruitment of retinoic acid signaling. FASEB J. 24, 3298-3309.

Hansen, L., Mikkelsen, A., Niirnberg, P., Niirnberg, G., Anjum, I., Eiberg, H. and Rosenberg, T. (2009). Comprehensive mutational screening in a cohort of Danish families with hereditary congenital cataract. Invest Ophthalmol Vis Sci. 50, 3291-3303.

Hawkins, T. A., Haramis, A.-P., Etard, C, Prodromou, C, Vaughan, C. K., Ashworth, R., Ray, S., Behra, M., Holder, N., Talbot, W. S., Pearl, L. H., Strahle, U. and Wilson, S. W. (2008). The ATPase-dependent chaperoning activity of Hsp90a regulates thick filament formation and integration during skeletal muscle myofibrillogenesis. Development 135, 1147-1156.

Hinitis, Y., Osborn, D. P. S. and Hughes, S. M. (2009). Differential requirements for myogenic regulatory factors distinguish medial and lateral somitic, cranial and fin muscle fibre populations. Development 136, 403-414.

Hong, S.-K., Haldin, C. E., Lawson, N. D., Weinstein, B. M., Dawid, I. B. and Hukriede, N. A. (2005). The zebrafish kohtalo/trap230 gene is required for the development of the brain, neural crest, and pronephric kidney. Proc Natl Acad Sci USA. 102, 18473-18478.

Hove, J. R., Koster, R. W., Forouhar, A. S., Acevedo-Bolton, G., Fraser, S. E. and Gharib, M. (2003). Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis. Nature 421, 172-177.

Jackman, W. R., Draper, B. W. and Stock, D.W. (2004). Fgf signaling is required for zebrafish tooth development. Dev. Biol. 214, 139-157.

Kachur, T., Ao, W., Berger, J. and Pilgrim, D. (2004). Maternal UNC-45 is involved in cytokinesis and colocalizes with non-muscle myosin in the early Caenorhabditis elegnas embryo. J Cell Sci. 117, 5313-5321. 113

Kachur, T., Ao, W., Berger, J. and Pilgrim, D. (2004). Maternal UNC-45 is involved in cytokinesis and colocalizes with non-muscle myosin in Caenorhabditis elegans embryo. J Cell Sci. Ill, 5313-5321.

Kathiriya, I. S. and Srivastava, D. (2000). Left-right asymmetry and cardiac looping: implications for cardiac development and congenital heart disease. Am. J. Med. Genet. (Semin. Med. Genet.) 97, 271-279.

Kagawa, M., Sato, N. and Obinata, T. (2006). Effects of BTS (N-benzyl-p- toluene sulphonamide), an inhibitor for myosin-actin interaction, on myofirbrillogenesis in skeletal muscle cells in culture. Zoolog Sci. 23, 969-975.

Krone, P. H., Lele, Z. and Sass, J. B. (1997). Heat shock genes and the heat shock response in zebrafish embryos. Biochem. Cell Biol. 75, 487-497.

Lee, C. F., Melkani, G. C, Yu, Q., Suggs, J. A., Kronert, W. A., Suzuki, Y., Hipolito, L., Price, M. G., Epstein, H. F. and Bernstein, S. I. (2011). Drosophila UNC-45 accumulates in embryonic blastoderm and in muscles, and is essential for muscle myosin stability. J Cell Sci. 124, 699-705.

Lele, Z., Hartson, S. D., Martin, C. C, Whitesell, L., Matts, R. L. and Krone, P. H. (1999). Disruption of zebrafish somite development by pharmacologic inhibition of Hsp90. Dev. Biol. 210, 56-70.

Lin, C-.Y., Yung, R-.F., Lee, H-.C, Chen, W-.T., Chen, Y-.H. and Tsai, H-.J. (2006). Myogenic regulatory factors Myf5 and MyoD function distinctly during craniofacial myogenesis of zebrafish. Dev. Biol. 299, 594-608.

Liu, L., Srikakulam, R. and Winkelmann, D. A. (2008). Unc45 activates Hsp90-dependent folding of the myosin motor domain. J Biol Chem. 283, 13185-13193.

Mathias, R. T., Kistler, J. and Donaldson, P. (2007). The lens circulation. J Membrane Biol. 216, 1-16.

Panakova, D., Werdich, A. A. and MacRae, C. A. (2010). Wntll patterns a myocardial electrical gradient through regulation of the L-type Ca2+ channel. Nature 466, 874-879.

Pickett, F. B. and Meeks-Wagner, D. R. (1995). Seeing double: appreciating genetic redundancy. Plant Cell 1, 1347-1356.

Perez-Perez, J. M., Candela, J. and Micol, J. L. (2009). Understanding Synergy in genetic interactions. Trends Genet. 25, 368-376. 114

Price, M. G., Landsverk, M. L., Barral, J. M. and Epstein, H. F. (2002). Two mammalian UNC-45 isoforms are related to distinct cytoskeletal and muscle- specific functions. J Cell Sci. 115, 4013-4023.

Proisy, N., Sharp, S. Y., Boxall, K., Connelly, S., Roe, S. M., Prodromou, C, Slawin, A. M., Pearl, L. M., Workman, P. and Moody, C. J. (2006). Inhibition of hsp90 with synthetic macrolactones: synthesis and structural and biological evaluation of ring and conformational analogs of radicicol. Chem. Biol. 13, 1203-1215.

Sanger, J. W., Wang, J., Fan, Y., White, J. and Sanger, J. M. (2010). Assembly and dynamics of myofibrils. JBiomed Biotechnol. 2010, 858606.

Schilling, T. F. and Kimmel, C. B. (1997). Musculoskeletal patterning in the pharyngeal segments of the zebrafish embryo. Development 124, 2945-2960.

Stainier, D. Y. R., Fouquet, B., Chen, J.-N., Warren, K. S., Weinstein, B. M., Meiler, S. E., Mohideen, M. P. K., Neuhauss, S. C. F., Solnica-Krezel, L., Schier, A. F., Zwartkruis, F., Stemple, D. L., Malicki, J., Driever, W. and Fishman, M. C. (1996). Mutations affecting the formation and function of the cardiovascular system in the zebrafish embryo. Development 123, 285-292.

Stanley, B. A. (2007). Proteomic analysis of heart failure: Insights into myofibril assembly and regeneration. PhD Thesis. Queen's University.

Teoh, P., Shu-Chien, A. C. and Chan, W. (2010). Pbxl Is essential for growth of zebrafish swim bladder. DevDyn. 239, 865-874.

Thomas, J. H. (1993). Thinking about genetic redundancy. Trends Genet. 9, 395- 398.

Venolia, L. and Waterston, R. H. (1990). The unc-45 gene of Caenorhabditis elegans is an essential muscle-affecting gene with maternal expression. Genetics 126, 345-353.

Winata, C. L., Korzh, S., Kondrychyn, I., Korzh, V. and Gong, Z. (2010). The role of vasculature and blood circulation in zebrafish swimbladder development. BMC Dev Biol. 10, 3.

Wohlgemuth, S. L., Crawford, B. D. and Pilgrim, D. B. (2007). The myosin co-chaperone UNC-45 is required for skeletal and cardiac muscle formation in zebarfish. Dev. Biol. 303, 483-492.

Wohlgemuth, S. L. (2007). Characterization of unc45b in zebrafish muscle development. PhD Thesis. University of Alberta. 115

APPENDIX: ELECTROCARDIOGRAPHY 116

A.l Introduction

The paper that follows was submitted as a Technical Report to genesis, The Journal of Genetics and Development on October 1, 2010. Regrettably, the manuscript was not accepted for publication based on concerns raised by the two anonymous reviewers. The first issue was related to the electrical signals we detected. Specifically, one reviewer had reservations about our ventricular signal and precisely what electrical events were being measured. We took our ventricular signal to represent ventricular depolarization, equivalent to the QRS complex on an ECG recorded from an adult. Ventricular repolarization was not detected in our recordings, as indicated by the absence of a T wave signal. It is possible that the T wave was not discernible because of an overlap with the much larger QRS complex that led to its obstruction. Data from a recent report supports this interpretation. Yu et al. (2010b) examined the evolution of cardiac conduction during zebrafish development spanning from 7-35 dpf and found that P waves and QRS complexes were present at 7 dpf but it wasn't until 14 dpf that distinct T waves started to develop. The second issue raised by the reviewers was whether the signals recorded were in fact mechanical and not electrical in origin. To address this concern we collected ECG recordings from zebrafish that had been bathed in blebbistatin prior to recording, so as to inhibit cardiac contractions. Blebbistatin is a small molecule inhibitor that acts as an excitation-contraction uncoupling agent by blocking myosin II in the actin-detached state (Kovacs et ah, 2004; Jou et ah, 2010). Treatment with blebbistatin does not alter ECG parameters nor interfere with myosin-actin binding or dissociation (Kovacs et ah, 2004; Jou et ah, 2010). Measurements performed on blebbistatin treated zebrafish would be expected to produce characteristic ECG signals if our ECG recordings were truly detecting electrical currents, as cardiac conduction would remain unaffected by exposure to the chemical. If the recordings, however, were due to a cardiac motion artifact, then the application of an uncoupling agent would lead to no signal being detected. We observed the latter. 117

Electrocardiography to assess cardiac function in embryonic zebrafish and genetic mutants

Sophie A. Comyn, Shunmoogum A. Patten, David B. Pilgrim, Declan W. Ali

The electrocardiogram (ECG) is a valuable tool for the analysis of cardiac function and we show it can be used in zebrafish embryos under 5 days post fertilization (dpf). We recorded ECGs in embryos 52-60 hours post fertilization (hpf) following the critical hatching period when the heart forms a two-chambered organ and applied it to the functional analysis of genetic mutants. These traces have distinct bimodal waves similar to those described at 5 dpf and permit the quantification of cardiac parameters such as frequency, total cycle time and ventricular depolarization. ECGs from a cardiovascular mutant revealed physiological differences that were not apparent from standard microscopy observations or in situ hybridization. This method thus adds a tool for detailed characterization of heart function in wild type or mutant zebrafish embryos following the formation of distinct atrial and ventricular chambers.

A.2 Results and Discussion

The electrocardiogram (ECG) has been used to analyze cardiac function in zebrafish (Forouhar et ah, 2004; Milan et ah, 2006; Yu et ah, 2010a). Other methods include: visual heart rate monitoring by stopwatch, laser-scanning velocimetry (Malone et ah, 2007) and digital motion analysis (Chan et ah, 2009). The benefits of the ECG method are that no expensive equipment, software, or materials are required beyond those already available in most departments. The ECG measures small potential differences on the surface of the skin from currents in the extracellular fluid created by the movement of action potentials through the cardiac myocytes (Boron and Boulpaep, 2009; Levick, 2010). The electrocardiogram has been used to assess drug-induced QT prolongation in adult zebrafish (Milan et ah, 2006) and to monitor cardiac function prior to and during heart regeneration (Yu et ah, 2010a). The earliest time at which the ECG recordings have been reported in zebrafish is 5 dpf; well after the hatching period 118 is complete (Forouhar et ah, 2004). Embryos hatch out of the chorion around 2 dpf marking a period of major physiological changes in preparation for life in the open environment (Patten and Ali, 2007; Patten et ah, 2010). During this time, the heart morphs from a primitive tube into a two-chambered organ demonstrating coordinated contractions. We tested whether the ECG could be used to assess cardiac function in embryos 52-60 hpf in the period immediately following hatching and compared these recordings to those detected at later larval stages. This method adds another facet to the characterization of fish cardiac phenotypes presenting early in development, detailed functional analysis of genetic mutants, as well as manipulation prior to the onset of pericardial edema that is common in zebrafish cardiovascular mutants. ECG recordings were collected from 52-60 hpf zebrafish. Zebrafish were anesthetized in 0.02% tricaine (MS-222) (Sigma-Aldrich, St. Louis, MO) and immobilized ventral side up on a Sylgard-lined dish. Preparations were moved to the recording set-up and the chamber was continuously perfused (1-2 mL/min) at room temperature (20-24°C) with an aerated 30% Danieau solution that contained 0.02%) tricaine. The pipette electrodes used for our extracellular recordings were pulled from borosilicate glass and filled with Danieau buffer. Electrode resistances ranged between 3.5-5 MQ.. The recording electrode was positioned on the body surface outside of the heart between the atrium and ventricle as described in Forouhar et al. (2004). Precise positioning of the electrodes is critical for a repeatable ECG signal. ECG signals were recorded using an Axopatch 200B amplifier (MDS Analytical Technologies, Sunnyvale, CA), low-pass filtered at 2 kHz and digitized at 50 kHz. ECGs were recorded for 5 minutes using pClamp 8.1 software (MDS Analytical Technologies, Sunnyvale, CA). ECG events were detected using the template function for events (>2.5 standard deviations above the basal noise) and analyzed with Axograph X (MDS Analytical Technologies, Sunnyvale, CA). The software detected all events that could be recognized visually and these events were then analyzed. 119

Electrocardiographic recordings were reproducible and had distinct atrial and ventricular depolarization waves similar to those described at 5 dpf by Forouhar et al. (2004) (Figure A-l). Recordings at 2 dpf or 5 dpf do not show a triplet QRS complex or a T wave seen in adult zebrafish (Forouhar et ah, 2004; Milan et ah, 2006; Yu et ah, 2010a). Our mean cardiac cycle time of 368.1 ± 9.1 ms is comparable to the 433 ± 36 ms measured at 5 dpf (Forouhar et ah, 2004) (Table A-l). The mean delay between atrial and ventricular depolarization was 96.5 ± 9.9 ms in our samples and 148.3 ± 30 ms for those at 5 dpf (Forouhar et ah, 2004). Although the mean values differ with longer delay and total cycle times in the 5 dpf larvae, the general shape of the bimodal traces are the same (Figure A- 1). Malone et al. (2007) analyzed cardiac function using laser-scanning velocimetry, a method used to measure parameters similar to those obtained with Doppler echocardiography and found that 2 and 5 dpf were modestly different in peak velocity, stroke volume, and cardiac output. These observations are not unexpected given that by 48 hpf the zebrafish heart has looped and forms two morphologically distinct chambers with functional valves and is therefore similar to its adult form (Chen et ah, 1996; Lohr and Yost, 2000; Warren et ah, 2000; Stainier, 2001). Our data show that use of the ECG as a tool to assess cardiac function in embryos as young as 52 hpf is both replicable and physiologically relevant. Forouhar et al. (2004) suggest that the ECG could be used as a tool to discern electrophysiological differences in fish with cardiac abnormalities linked to genetic lesions. Cardiovascular mutants can: die before 5 dpf; have severe and extensive edema; or present with defects secondary to the cardiac phenotype. Each of these phenotypes prevents an accurate assessment of cardiac functions in older embryos. To be effective, the ECG would have to allow a wider range and more accurate phenotypic analysis at a stage when qualitative assessment of heart function is currently poor. To evaluate the efficiency of ECG as a tool in 2 dpf embryos we tested two mutant strains of zebrafish Unc45 genes. 120

First identified in Caenorhabditis elegans, UNC-45 has two vertebrate homologues: Unc45a and Unc45b. In zebrafish, unc45a is ubiquitously expressed early in development and mutants (kurzschluss) have abnormal aortic arch formation resulting in defective circulation confined to the head and heart (Chen et ah, 1996; Anderson et ah, 2008). In contrast, unc45b expression is restricted to skeletal and cardiac muscle with mutants (steif) having disorganized skeletal and cardiac myofibrils leading to paralysis and cardiac dysfunction (Etard et ah, 2007; Wohlgemuth et ah, 2007). By 5 dpf both mutants have developed cardiac edema (Figure A-2). As a control, we recorded ECGs from steif embryos. As shown in Figure A-l, no bimodal waves or signal above background were detected in the steif embryos. Kurzschluss (kus) fish have bimodal traces visibly similar to those of wild type. We quantified various components of the ECG recordings to see if the two genotypes differed significantly. Frequency, defined as the number of events per second, was 20%> lower in kus embryos (Figure A-3 a) and the total cardiac cycle time was 31% longer compared to wild type (Figure A-3 b) (Table A-l). The prolonged cardiac cycle, or slower heart rate observed, could be the result of an increase in a combination of atrial and ventricular width or mean delay. We took atrial and ventricular width to be the measure of the duration of depolarization for each chamber. Both atrial (Figure A-3 c) and ventricular (Figure A-3 d) width were not statistically different between the kus and wild type genotypes (Table A-l). Consequently, an increase in delay time alone can account for the decreased frequency and increased total cycle observed in the kus fish. Indeed, the delay interval was 39%> longer in kus as compared with wild type (Figure A-3 e, Table A-l). We tested for obvious morphological or calcium handling defects in the kus mutants to exclude other potential contributing factors to the observed irregular cardiac rhythm. By 48 hpf, the two cardiac chambers are morphologically distinct and express discrete myosin heavy chain genes (Chen et ah, 1996; Schoenebeck and Yelon, 2007). Cardiac myosin light chain 2 (cmlc2) is expressed in both the atrium and ventricle permitting the observation of gross 121 cardiac morphology and heart looping. Consistent with previous reports, the kus heart does not appear to differ from wild type (Figure A-4 a, d). Strict control of intracellular calcium concentrations is essential for a coordinated and rhythmic heart contraction (Ebert et ah, 2005; Langenbacher et ah, 2005). Thus we investigated the expression patterns of mRNA for two calcium regulatory proteins, the sodium calcium exchanger (NCXlh) and the sarcoplasmic reticulum Ca2+-ATPase2 (SERCA2) (Ebert et ah, 2005; Langenbacher et ah, 2005). No apparent difference was seen in the distribution or abundance of either the expression of NCXlh (Figure A-4 b, e) and SERCA2I (Figure A-4 c, f). The differences in cardiac electrophysiology seen by ECG between the kus and wild type genotypes would have likely been overlooked using current methods. Over one hundred zebrafish mutants exhibiting abnormal cardiovascular development were identified in two large-scale forward genetic screens (Chen et ah, 1996; Stainier et ah, 1996). Since their publication in 1996, progress has been made towards understanding cardiac development using an approach that evaluates the function of genes based on mutant phenotypes without prejudice as to what role the genes may play in a given process (Chen et ah, 1996; Stainier et ah, 1996). However, Schoenebeck and Yelon (2007) argue that the classification of cardiac mutants by qualitative features alone has left the diversity of functional cardiac phenotypes unjustifiably under recognized. The ECG is one method that can be used to remedy this problem. Here, we have described the use of the ECG to examine cardiac function in embryonic zebrafish as early as 52 to 60 hpf. Our findings validate the use of electrocardiography to investigate cardiac function in these early embryos and demonstrate its use as a tool in the functional analysis of genetic mutants. 122

A.3 Methods

A.3.1 Fish Maintenance

Adult AB and kurzschluss zebrafish were maintained according to standard procedures (Westerfield, 2000) and were naturally spawned to obtain embryos. Embryos were raised at 28.5 °C and staged according to published morphological hallmarks (Kimmel et ah, 1995). All procedures were carried out in compliance with the guidelines stipulated by the Canadian Council for Animal Care and the University of Alberta.

A.3.2 In situ Hybridization

Digoxigenin-labeled antisense RNA probes were synthesized by in vitro transcription using T7 RNA polymerase (Ambion, Austin, TX) and digoxigenin- 11-UTP RNA labeling mix (Roche, Indianapolis, IN). The following markers were used: cmlc2 (Yelon et ah, 1999), NCXlh, SERCA2. Whole-mount in situ hybridization analysis was performed essentially as described (French et ah, 2009) with the following modifications: Proteinase K treatment (10 pg/mL) was performed for 75 min (3 dpf embryos) and 25 min (3 dpf embryos, cmlc2). Embryos were photographed using an Olympus stereoscope with a Qimaging micropublisher camera. Images were assembled using Adobe Photoshop CS Version 8.0.

A.4 Acknowledgments

This work was supported by grants from the Natural Sciences and Engineering Research Council of Canada (NSERC) to DBP and DWA and the Canadian Foundation for Innovation (CFI) to DWA. SAC is supported by a studentship from NSERC. 123

A.5 Tables

Table A-l. Electrocardiogram measurements from wild type and kus embryos. Measurements Wild type kus Significance

Frequency (Hz) 2.5 ±0.1 (8) 2.0 ±0.2 (10) P == 0.033

Total cycle (ms) 368.1 ±9.1 (8) 480.3 ±41.4 (10) P == 0.008

Atrial width (ms) 36.8 ± 4.5 (6) 44.1 ±9.1(8) P == 0.606 Ventricular width 174.6 ±20.7 (6) 209.4 ± 37.7 (8) == 0.606 (ms) P

Delay (ms) 96.5 ± 9.9 (6) 133.9 ±9.7 (8) P == 0.020

Values are means ± SE with sample sizes in parentheses. Differences were tested by Mann Whitney U test. P values of less than 0.05 are considered statistically significant. 124

A.6 Figures

WT

steif mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm

| 0.1 nV 1s

Figure A-l. Representative ECG recordings from 52-60 hpf zebrafish embryos. Wild-type (WT) and kus (kurzschluss) embryo traces have distinct atrial and ventricular depolarization waves. No electrical activity was detected in recordings from steif embryos. 125

Figure A-2. Gross morphology of kus and steif embryos as compared to wild type at 3 and 5 dpf. At 3 dpf, kus (c) and steif mutants (e) begin to develop cardiac edema not seen in wt siblings (a). By 5 dpf, wt fish (b) display no fluid accumulation in the pericardial space whereas steif (f) embryos have extensive edema and kus (d) to a lesser extent. Lateral views of live embryos with anterior towards the left. a b ^600 | 500

WT kus

J.300 200 S 250 I" 150 I 200 E « 150 ¥10(H M1 m iS

I 50 <§ 50 _J_J Lr^^^K_| v 0 WT km iVT *IIS WT Itos

Figure A-3. Quantification of ECG recordings from wild-type and kus embryos, (a) The frequency of cardiac events was reduced in kus (n = 10) mutants compared to wild-type (n = 8, P = 0.033). (b) Total cardiac cycle time was significantly shorter (P = 0.008) in wild-type (n = 8) to kus (n = 10) fish, (c) No difference in atrial depolarization time, measured by atrial width, is detected between kus (n = 8) and wild-type (n = 6, P = 0.606. (d) There was also no difference in ventricle width between kus (n = 8) and wild-type (n = 6, P = 0.606) (e) Delay was longer in kus (n = 8) than wild-type (n = 6, P = 0.020). This increase can account for both the decrease in frequency and increase in cycle time observed in kus mutants. Figure A-4. Whole-mount mRNA in situ hybridization to detect expression of cmlc2 (a, d), NCXlh (b, e) and SERCA2 (c, f) in 3 dpf zebrafish embryos. Cmlc2 expression is comparable in wild-type (a) and kus mutants (d) with no difference in cardiac looping. No variation in distribution or abundance of gene expression is discemable between wild-type (b, c) and kus mutants (e, f) for either NCXlh or SERCA2, respectively. Ventral views with anterior at top 128

A.7 References

Anderson, M. J., Pham, V. N., Vogel, A. M., Weinstein, B. M. and Roman, B. L. (2008). Loss of unc45a precipitates arteriovenous shunting in the aortic arches. Dev Biol. 318, 258-267

Boron, W. F. and Boulpaep, E. L. (2009). Medical Physiology: A cellular and molecular approach, 2nd ed., pp. 504-528. Philadelphia: Elsevier.

Chan, P. K., Lin, C. C. and Cheng, S. H. (2009). Noninvasive technique for measurement of heartbeat regularity in zebrafish (Danio rerio) embryos. BMC Biotechnol. 9, 11-21.

Chen, J. N., Haffter, P., Odenthal, J., Vogelsang, E., Brand, M., van Eeden, F. J. M., Furutani-Seiki, M., Granato, M., Hammerschmidt, M., Heisenberg, C. P., Jiang, Y. J., Kane, D. A., Kelsh, R. N., Mullins, M. C. and Niisslein-Volhard, C. (1996). Mutations affecting the cardiovascular system and other internal organs in zebrafish. Development 123, 293-302.

Ebert, A. M., Hume, G. L., Warren, K. S., Cook, N. P., Burns, C. G., Mohideen, M. A., Siegal, G., Yelon, D., Fishman, M. C. and Garrity, D. M. (2005). Calcium extrusion is critical for cardiac morphogenesis and rhythm in embryonic zebrafish hearts. Proc Natl Acad Sci USA 102, 17705-17710.

Etard, C, Behra, M., Fischer, N., Hutcheson, D., Geisler, R. and Strahle, U. (2007). The UCS factor Steif/Unc-45b interacts with the heat shock protein Hsp90a during myofibrillogenesis. Dev Biol. 308, 133-143.

Fishman, M. C. (1996). Mutations affecting the formation and function of the cardiovascular system in the zebrafish embryo. Development 123, 285-292.

Forouhar, A. S., Hove, J. R., Calvert, C, Flores, J., Jadvar, H. and Gharib, M. (2004). Electrocardiographic characterization of embryonic zebrafish. Conf Proc IEEE Eng Med Biol Soc. 5, 3615-3617.

French, C. R., Erickson, T., French, D. V., Pilgrim, D. B. and Waskiewicz, A. J. (2009). Gdf6a is required for the initiation of dorsal-ventral retinal patterning and lens development. Dev Biol. 333, 37-47.

Jou, C. J., Spitzer, K. W. and Tristani-Firouzi, M. (2010). Blebbistatin effectively uncouples the excitation-contraction process in zebrafish embryonic heart. Cell Physiol Biochem. 25,419-424.

Kimmel, C. B., Ballard, W. W., Kimmel, S. R., Ullmann, B. and Schilling, T. F. (1995). Stages of embryonic development of the zebrafish. Dev Dyn. 203, 253-310. 128

Kovacs, M., Toth, J., Hetenyis, C, Malnasi-Csizmadia, A. and Sellers, J. R. (2004). Mechanism of blebbistatin inhibition of myosin II. J Biol Chem. 279, 35557-35563.

Langenbacher, A. D., Dong, Y., Shu, X., Choi, J., Nicoll, D. A., Goldhaber, J. I., Philipson, K. D. and Chen, J. N. (2005). Mutation in sodium-calcium exchanger 1 (NCX1) causes cardiac fibrillation in zebrafish. Proc Natl Acad Sci USA 102, 17699-17704.

Levick, J. R. (2010). An Introduction to cardiovascular physiology, 5th ed., pp. 67-78. Hodder Arnold.

Lohr, J. L. and Yost, H. J. (2000). Vertebrate model systems in the study of early heart development: Xenopus and zebrafish. Am J Med Genet. 97, 248- 257.

Malone, M. H., Sciaky, N., Stalheim, L., Hahn, K. M., Linney, E. and Johnson, G. L. (2007). Laser-scanning velocimetry: A confocal microscopy method for quantitative measurement of cardiovascular performance in zebrafish embryos and larvae. BMC Biotechnol. 1, 40-51.

Milan, D. J., Jones, I. L., EUinor, P. T. and MacRae, C. A. (2006). In vivo recording of adult zebrafish electrocardiogram and assessment of drug-induced QT prolongation. Am J Physiol Heart Circ Physiol. 291, H269-H273.

Patten, S. A. and Ali, D. W. (2007). AMPA receptors associated with zebrafish Mauthner cells switch subunits during development. J Physiol. 581, 1043- 1056.

Patten, S. A., Roy, B., Cunningham, M. E., Stafford, J. L. and Ali, D. W. (2010). Protein kinase C gamma is a signaling molecule required for the developmental speeding of alpha-amino-3-hydroxyl-5-methyl-4-isoxazole- propionate receptor kinetics. Eur JNeurosci. 31, 1561-1573.

Schoenebeck, J. J. and Yelon, D. (2007). Illuminating cardiac development: Advances in imaging add new dimensions to the utility of zebrafish genetics. Semin Cell Dev Biol. 18, 27-35.

Stainier, D. Y. R., Fouquet, B., Chen, J. N., Warren, K. S., Weinstein, B. M., Meiler, S. E., Mohideen, M. P. K., Neuhauss, S. C. F., Solnica-Krezel, L., Schier, A. F., Zwartkruis, F., Stemple, D. L., Malicki, J., Driever, W. and Fishman, M. C. (1996). Mutations affecting the formation and function of the cardiovascular system in the zebrafish embryo. Development 123, 285-292. 128

Stainier, D. Y. (2001). Zebrafish genetics and vertebrate heart formation. Nat Rev Genet. 2, 39-48.

Warren, K. S., Wu, J. C, Pinet, F. and Fishman, M. C. (2000). The genetic basis of cardiac function: dissection by zebrafish (Danio rerio) screens. Philos Trans R Soc LondB Biol Sci. 355, 939-944.

Westerfield, M. (2000). The zebrafish book: A guide for the laboratory use of zebrafish (Danio rerio), pp. 1-27. Eugene: University of Oregon Press.

Wohlgemuth, S. L., Crawford, B. D. and Pilgrim, D. B. (2007). The myosin co-chaperone UNC-45 is required for skeletal and cardiac muscle function in zebrafish. Dev Biol. 303, 483-492.

Yelon, D., Home, S. A. and Stainier, D. Y. (1999). Restricted expression of cardiac myosin genes reveals regulated aspects of heart tube assembly in zebrafish. Dev Biol. 214, 23-37.

Yu, F., Li, R., Parks, E., Takabe, W. and Hsiai, T. K. (2010). Electrocardiogram signals to assess zebrafish heart regeneration: Implications of long QT intervals. Ann Biomed Eng. 38, 2346-2357.

Yu, F., Huang, J., Adlerz, K., Jadvar, H., Hamdan, M. H., Chi, N., Chen, J.- N. and Hsiai, T. K. (2010). Evolving cardiac conduction phenotypes in developing zebrafish larvae: implications to drug sensitivity. Zebrafish 1, 325- 331.