MAPPING GAPING LIDS: A MUTATION CAUSING OPEN EYELIDS AT BIRTH IN MOUSE

by

KATHLEEN GRACE BANKS

B.Sc, The University of British Columbia in association with the University College of the Cariboo, 1996.

A THESIS SUMBITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

in

THE FACULTY OF GRADUATE STUDIES

(Faculty of Medicine; Department of Medical Genetics)

We accept this thesis as conforming to the required standard

THE UNIVERSITY OF BRITISH COLUMBIA

November 1999

© Kathleen Grace Banks, 1999 UBC Special Collections - Thesis Authorisation Form Page 1 of 1

In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.

Department of

The University of British Columbia Vancouver, Canada

http://www.library.ubc.ca/spcoll/thesauth.html 11/19/99 ABSTRACT

Gaping lids (gp) is an autosomal recessive mutation that arose spontaneously in the C57BL/6-ax strain of mice and is now maintained in the inbred strain GP/Bc. The purpose of this study included the mapping of the gp mutation, analysis of its segregation after two outcrosses to normal (non-open eyelid) strains and characterization of the mutant phenotype. The main objective, to map the gp locus, was undertaken based on the hypothesis that the loci with mutations that cause open eyelids with simple Mendelian

transmission patterns may also be loci involved in the open eyelids traits with more

complex inheritance. Since these mutations are viable, they are good models for studying

the interaction of multiple loci in a genetically complex birth defect.

The gaping lids mutation was mapped close to the centromere at the proximal end

of 11, employing PCR amplification of informative SSLP marker loci,

initially using 41 gaping lids F2 progeny from a cross of GP/Bc to the normal strain

CBA/J, followed by a refinement of the region using 23 gaping lids F2 progeny from a

second outcross of GP/Bc to ICR/Be, another normal inbred strain. Based on the

recombination breakpoints, gp is within 10 cM of Dl lMit80 and 2 cM of Dl lMit71.

The epidermal growth factor receptor (Egfr) was also mapped to mid-chromosome

11 between Dl lMit226 and Dl lMitl51, employing PCR amplification of SSLP markers,

using mice carrying a null allele at this locus. This map location supports the finding that

gp and Egfr are not allelic as determined by a complementation test between Egfr+I~ and

GP/Bc.

gp showed reduced penetrance, 82% and 33% in the outcrosses to CBA/J and

ICR/Be, respectively, but it was determined that this was likely due to suppressors-of

ii open eyelids loci introduced by the normal strains and not prenatal death of gp/gp progeny. The only phenotypic anomaly associated with this mutation appears to be eyelids at birth, due to failure of normal eyelid closure during late gestation.

iii TABLE OF CONTENTS

Abstract ii

Table of Contents iv

List of Tables vi

List of Figures viii

List of Appendices x

List of Abbreviations xii

Acknowledgments xiii

CHAPTER I: INTRODUCTION I: History of gaping lids 1 II: Review of eyelid development and open eyelids at birth mutations 2 A. Eyelid development 2 B. expressed in the developing eyelids 5 C. Open eyelids at birth mutants 7 III: Mouse mapping 14 A. Overview and history 14 B. Polymerase Chain Reaction and Simple Sequence Length 31 Polymorphisms C. Review of mouse maps 33 IV: Rationale and approach to this study 36

CHAPTER II: GENERAL METHODS AND MATERIALS I: gaping lids: Scientific progress before and during this study 38 A. The cross to CBA/J 38 B. The cross to ICR/Be 39 II: Mouse stocks and maintenance 40 III: Technical methods 42 A. GP/Bc study 42 B. Egfr study 46

CHAPTER III: PHENOTYPIC INVESTIGATIONS I: Introduction 49 II: Rationale, Materials and Approach 49 III: Results 50

CHAPTER IV: MAPPING GAPING LIDS I: Introduction 55

iv II: Rationale, Materials and Approach 55 A. Experimental design 55 B. Analysis of genetic transmission/penetrance 60 C. Molecular investigations 61 III: Results 65 A. Segregation studies 65 B. Mapping studies 67 I. GP/Bc x CBA/J crosses 67 II. GP/Bc x ICR/Be cross 77 C. Analysis of genetic transmission/penetrance after crosses to 84 CBA/J and ICR/Be D. Molecular investigations 85

CHAPTER V: MAPPING EGFR I: Introduction 90 II: Rationale, Materials and Approach 90 A. Experimental design 90 I. Egfr7BXA-2 x SWV/Bc cross 91 III: Results 93 A. Egfr7BXA-2 x SWV/Bc cross 93

CHAPTER VI: CORRECTING THE MGI MAP 101

CHAPTER VII: DISCUSSION I: Segregation studies 106 A. CBA/J cross 107 B. ICR/Be cross 107 C. Applications of the threshold model 110 II: Mapping studies 114 A. GP/Bc study 114 B. Egfr study 123 III: Phenotypic investigations 124 IV: Conclusions 126

Literature Cited 127

Appendices 141

v LIST OF TABLES

Table 1 Genes expressed in the developing eyelids. 6-7

Table 2 Open eyelids at birth mutations in mouse. 15-28 (a) nonsyndromic (b) syndromic (c) ectopic gene expression (d) strains with susceptibility to open eyelids (e) chromosomal

Table 3 Frequency of open eyelids in newborns from GP/Bc x CBA/J cross. 39

Table 4 Frequency of open eyelids in newborns from GP/Bc x ICR/Be cross. 40

Table 5 (a) Measurements of palpebral opening and eye in GP/Bc, CBA/J and 52 (GP/Bc x CBA/J) Fl autopsied animals, (b) Measurements of palpebral opening and eye in GP/Bc and AXB- 23/Pgn animals.

Table 6 Comparison ofSSLP marker loci map position between the Mouse 64 Genome Informatics Database (MGI), Massachusetts Institute of Technology/Research Genetics (MIT) and the European Collaborative Interspecific Backcross panel (EUCIB-BSB).

Table 7 (a) Data for GP/Bc x CBA/J F2 affected and normal progeny - 11 77 complete litters (b) Segregation at markers closest to gp, Dl lMit62 and Dl lMit226, 77 in F2 from GP/Bc x CBA/J in 11 complete litters.

Table 8 Segregation of alleles at D11 Mit74 in normal F2s in GP/Bc x ICR/Be 83 cross.

Table 9 (a) Data for GP/Bc x ICR/Be F2 affected and normal progeny - 6 83 complete litters (b) Segregation at marker closest to gp, Dl lMit74, in F2 from GP/Bc 84 x ICR/Be in 6 complete litters.

Table 10 Haplotype analysis of GP/Bc versus C57BL/6J DNA. 89

Table 11 Segregation of alleles at Dl 1 Mit 149 in open eyelie/pinhole F2 mice 101 from GP/Bc x CBA/J cross.

Table 12 Outline of modifier scenarios in GP/Bc x ICR/Be F2 109

vi Table 13 Loci in the region of gaping lids - between the centromere and 116-117 DllMit80 on Chr 11.

vii LIST OF FIGURES

Figure 1 (a) Scanning electron microscope picture of dl6 GP/Bc fetal eye. 54 (b) Close up view of inner canthus of dl6 GP/Bc fetal eye.

Figure 2 Comparison of informative SSLP marker loci locations used in 62 GP/Bc x CBA/J and ICR/Be crosses, where marker identification numbers follow the format Dl lMit##.

Figure 3 Comparison of SSLP marker loci map locations used in GP/Bc x 63 CBA/J and GP/Bc x ICR/Be crosses on MGI, MIT and EUCIB (BSB) maps.

Figure 4 Mapping matrix of open eyelid F2 mice in GP/BC x CBA/J cross. 69 Data does not include pinhole F2 mice.

Figure 5 Location of markers used in GP/Bc x CBA/J cross. 70

Figure 6 Pictures of representative gels of D11 Mit62; GP/Bc, CBA/J, F1 and 71 panel of open eyelid and pinhole F2 animals.

Figure 7 Pictures of representative gels of Dl lMit226; GP/Bc, CBA/J, Fl and 72 panel of open eyelid and pinhole F2 animals.

Figure 8 Map location of gaping lids and distances between markers as 74 determined in GP/Bc x CBA/J cross.

Figure 9 Mapping matrix of "pinhole" F2 mice from GP/Bc x CBA/J cross. 75

Figure 10 Mapping matrix of normal F2s in GP/Bc x CBA/J cross. 76

Figure 11 Mapping matrix of open eyelid F2 mice in GP/Bc x ICR/Be cross. 79

Figure 12 Locations of markers used in GP/Bc x ICR/Be cross. 80

Figure 13 Pictures of representative gels of D11 Mit74; GP/Bc, ICR/Be, F1 and 81 panel of open eyelid F2 animals.

Figure 14 Map location of gaping lids and distances between markers as 82 determined in GP/Bc x ICR/Be cross.

Figure 15 Comparison of SSLP marker map locations between CBA/J cross, 86 ICR/Be cross and MGI, MIT, and EUCIB.

viii Figure 16 Picture of representative agarose gel of primers which amplify the 94 Egfr null and wildtype alleles.

Figure 17 Locations of markers used in (Egfr7BXA-2)Fl x SWV/Bc special 95 testcross.

Figure 18 Outline of (Egfr7BXA-2)F 1 x SWV/Bc special testcross. 96

Figure 19 Haplotypes of Egfr +/+ and Egfr +/" mice in (Egfr7BXA-2)F 1 x 97 SWV/Bc special testcross.

Figure 20 Location of Egfr locus based on (Egfr7BXA-2)F 1 x SWV/Bc special 99 testcross.

Figure 21 Locations of gaping lids locus and Egfr locus based on GP/Bc x 100 CBA/J cross, GP/Bc x ICR/Be cross and (Egfr7BXA-2)Fl x SWV/Bc special testcross.

Figure 22 (a) Mapping matrix of open eyelid F2 mice in GP/Bc x CBA/J cross. 102 A comparison between Dl 1 Mit 149 and the six previously typed SSLP markers. Data does not include pinhole F2 mice, (b) Mapping matrix of "pinhole" F2 mice from GP/Bc x CBA/J cross. A comparison between Dl 1 Mit 149 and the three previously typed SSLP markers.

Figure 23 (a) Mapping matrix of open eyelid F2 mice in GP/Bc x CBA/J cross. 104 A comparison between Dl 1 Mit 149 and Dl 1 Mit 10. Data does not include pinhole F2 mice, (b) Mapping matrix of pinhole F2 mice in GP/Bc x CBA/J cross. A comparison between Dl lMitl49 and Dl IMitlO.

Figure 24 (a) Mapping matrix of open eyelid F2 mice in GP/Bc x CBA/J cross. 105 A comparison between D11 Mit 149 and D9Mit 191. Data does not include pinhole F2 mice, (b) Mapping matrix of pinhole F2 mice in GP/Bc x CBA/J cross. A comparison between Dl lMitl49 and D9Mitl91.

Figure 25 Graphical illustration of the hypothesized effect of the suppressor 111 locus/loci in the a) CBA/J and b) ICR/Be crosses.

Figure 26 Graphical illustration of the hypothesized two threshold model to 113 explain pinhole progeny in the F2 generations of both GP/Bc outcrosses.

ix LIST OF APPENIDICES

Appendix A GP/Bc x CBA/J F1 open eyelid data. 142

Appendix B Gp/Bc x CBA/J BC 1 open eyelid data. 143 (i) GP/Bc dam data

Appendix C GP/Bc x CBA/J BC 1 open eyelid data. 144 (ii) GP/Bc sire data

Appendix D GP/Bc x CBA/J F2 open eyelid data. 145

Appendix E Chromosome 13 and 6 markers screened in GP/Bc x CBA/J cross - 146 includes PCR conditions, and reported allele sizes between the strains.

Appendix F Chromsome 11 markers screened in GP/Bc x CBA/J cross - includes 147 PCR conditions, and reported allele sizes between the strains.

Appendix G Chromosome 11 markers screened in GP/Bc x ICR/Be cross- 148 includes PCR conditions, and reported allele sizes between strains.

Appendix H Chromosome 11 markers screened in (Egfr7BXA-2) Fix SWV/Bc 149 cross - includes PCR conditions, and reported allele sizes between the strains.

Appendix I BC1 pool data. 150

Appendix J Open eyelid and pinhole F2 animals typed in GP/Bc x CBA/J cross. 151-152

Appendix K Allele segregation data from GP/Bc x CBA/J cross for 11 complete 153-156 litters.

Appendix L Open eyelid F2 animals typed in GP/Bc x ICR/Be cross. 157

Appendix M Allele segregation data from GP/Bc x ICR/Be cross for 6 complete 158-159 litters.

Appendix N Open eyelid and pinhole F2 animals typed in GP/Bc x CBA/J cross at 160-161 DllMitl49,DllMitl0andD9Mitl91.

Appendix O Progeny typed in (Egfr7BXA-2)Fl x SWV/Bc special test cross. 162-165 Appendix P SSLP marker loci not tested for informativeness in theGP/Bc x 166 CBA/J or GP/Bc x ICR/Be crosses, between 1 - 15 cM on Research Genetics/MIT and 1999 Chromosome Committee Map (Mouse Genome Informatics).

xi LIST OF ABBREVIATIONS

BC Backcross

CI Confidence Interval

DNA Deoxyribonucleic Acid

EUCIB European Collaborative Interspecific Mouse Backcross

MGI Mouse Genome Informatics Database

MIT Massachusetts Institute of Technology

PCR Polymerase Chain Reaction

RFLP Restriction Fragment Length Polymorphism

SSLP Simple Sequence Length Polymorphism

SSR Simple Sequence Repeat

VNTR Variable Number of Tandem Repeats

YAC Yeast Artificial Chromosome

xii ACKNOWLEDGEMENTS

I would like to thank my supervisors, Drs. Diana Juriloff and Muriel Harris for their time, effort and dedication to me during my studies with them. Their vast knowledge is inspiring. I would also like to thank my colleagues and co-workers for their

support, insights, and discussions which helped me forge through. I also thank Dr. Fred

Dill for his thoughtful contributions and discussions during the course of my time here.

I would not have made it this far without the love and unwavering and

unquestioning support of my family. Thanks to my mom, Fran, for always telling me I

could do anything my heart desired and my mind craved. She is a beautiful, brilliant

woman, something I hope I can live up to. Thanks to my little sis, Marni, for calling me a

"scientist" - something only she knows the true meaning of. Thank you also to all of

those people who I consider my extended family - you know who you are - without all of

you, this would not have been possible.

I dedicate this work to the memory of my father, Arthur Banks, who succumbed

to throat cancer in 1985. I miss you and love you always.

xiii CHAPTER I: INTRODUCTION

I. History of gaping lids and the GP/Bc strain

Gaping lids (gp), an open eyelids at birth mutation, arose spontaneously in a mutant stock of mice (C57BL/6-ax) in 1961 (Kelton and Smith, 1964). This mutation was recognized as being phenotypically similar to another open eyelids at birth mutation

"open eyelids" (oe; see Table 2a), except that no scabs formed over the open eyes.

Kelton and Smith (1964) also reported the lens volume of gp and oe homozygotes to be two times and three times as large, respectively, as those of the C57BL/6 controls beginning at embryonic day 15. The gaping lids mutation was reported to be an autosomal recessive with 100% penetrance and bilateral expression in the initial crosses.

Complementation tests have shown that gp is not allelic with oe (Kelton and Smith,

1964) or the lidgaps (Juriloff et al., 1983; Stein et al., 1967; Ricardo and Miller, 1967).

Gaping lids came to UBC in 1971, was inbred to Fl 1, crossed to SM/M1 in 1973 and brother-sister inbred to F63. This strain, homozygous for gp, is now called GP/Bc, and

100% of newborn pups have bilateral open eyelids at birth (D.M. Juriloff and M.J. Harris, personal communications).

Linkage studies by Kelton and Smith (1964) indicated that gp is not linked to non- agouti (a; Chr 2), brown (b, now Tyrpl - tyrosinase related ; Chr 4), belted (bt;

Chr 15), dilute (d, now Myo5a - myosin Va; Chr 9), fuzzy (fz; Chr 1), leaden (In; Chr 1), pink-eyed dilution (p; Chr 7), rex (Re; Chr 11), piebald (s, now Ednrb - endothelin receptor type B; Chr 14), Danforth's short tail (Sd; Chr 2) or varitint-waddler (Va; Chr 3).

Apparent loose linkage was found to oe on Chr 11, however, with a recombination

1 fraction of (0.4065 ± .0312). These two recessive open eyelids at birth traits were mapped against each other using classical complementation tests, where zero open eyelid at birth progeny would be expected in the Fl generations of reciprocal crosses if these two mutations are not allelic. The recombination fraction found here is extremely large and since most open eyelids at birth traits are very complex, these results were initially discounted. A series of additional linkage studies by Juriloff and Harris during 1983 -

1992, using isozyme, coat colour, and morphological marker loci, also resulted in an

"exclusion map" that encompassed about 30% of the genome (unpublished data). At the beginning of this study the map location of gaping lids was not known.

II. Review of eyelid development and open eyelids at birth mutations

A. Eyelid development

In all mammals examined, the eyelids fuse closed and subsequently reopen during development. Whether or not the offspring are born with open (cows, guinea pigs, humans) or closed (dogs, cats, mice) eyes depends on the developmental stage at which they are born (Harris and McLeod, 1982). In mice, the formation of the eyelids begins around day 13 of gestation and from days 14 to 16 the eyelids grow across the eye and become tightly fused with each other, staying fused until approximately 12 to 14 days after birth. In humans, the eyelids close during the eighth week of development and reopen during the seventh month of pregnancy. Eyelid growth and fusion has been reviewed by Harris and McLeod (1982), Findlater et al. (1993), Hamming (1983),

Pearson (1980), and Michael et al. (1988).

2 a) Neural crest cells and eyelid development

The neural crest is a unique structure in the vertebrate embryo (LeDouarin et al.,

1993). It is both migratory and, initially, multipotent having the potential to form multiple neural crest derivatives (Bronner-Fraser, 1995). In mammals, shortly before closure of the neural tube, the cranial neural crest cells leave the neuroectoderm of forebrain, midbrain, and hindbrain regions and migrate ventrally into the branchial arches and rostrally around the forebrain and optic cups into the facial region, contributing to the formation of facial organs (Osumi-Yamashita et al., 1994, Osumi-Yamashita et al., 1997;

Sadler, 1995). In mouse, specifically, crest formation and migration in the head is underway well before the neural folds approach and fuse in the cranial region (Nichols,

1981; Serbedzija; 1992, Morriss-Kay et al., 1991) at the 4- to 5- somite stage (Osumi-

Yamashita et al., 1997). After migration, cranial neural crest cells supply almost all of the connective tissues of the frontonasal processes, maxillary processes, and first and second visceral arches, which ultimately are responsible for the definitive connective tissues of the face, lids, conjunctiva, orbit and much of the orbital bones (Jakobiec and

Iwamoto, 1982).

b) Facial prominences

At 9 - 10 days of embryonic development, four mesenchymal prominences can be recognized: the two mandibular prominences (first branchial arch), the two maxillary prominences (dorsal portion of the first branchial arch) which extend forward beneath the region of the eye (Sadler, 1995), as well as the region termed the frontonasal prominence

(Sulik and Schoenwolf, 1985). The development of the face is later completed by

3 formation of the nasal prominences (medial and lateral) when the nasal pits invaginate, due to cell proliferation in this region.

c) The eyelids

The first indication of the development of the eyelids in mouse is the formation of a small depression/invagination above the eye and another below it at approximately day

14 of gestation (Pearson, 1980; Pei and Rhodin, 1970). A small ridge/fold extends rostrally and another caudally from the eye and these are bounded above and below by the slight depressions. The upper lid folds most probably develop superiorly to the eye from a proliferation of mesenchymal cells from the medial and lateral prominences extending caudally from the frontonasal prominence (Duke-Elder and Cook, 1963; Sevel,

1988). The lower lid develops from a proliferation of mesenchymal cells arising from the dorsolateral area of the maxillary prominence, the anterior portion of the first branchial arch (Juriloff, 1985; Juriloff and Harris, 1993; Sevel, 1988)

The eyelids themselves are specialized motile skin folds covered on the outside by keratinizing surface epidermis and on the inside by nonkeratinizing conjunctival epithelium (Jakobiec and Iwamoto, 1982; Juriloff and Harris, 1989). Between days 14 and 16 of gestation in the mouse, the eyelids grow out across the eye from both inner and outer canthi as well as from the leading edges above and below the eye, progressing toward the middle of the gap (Harris and McLeod, 1982). As fusion proceeds, the shrinking gap fills with an abundance of rounded periderm cells that are extruded, flattened and sloughed off from the area of completed fusion (Harris and McLeod, 1982).

The periderm is an additional layer, one cell thick, also of ectodermal origin, that covers the outer surface of the eyelids during their growth and subsequent fusion (Harris

4 and McLeod, 1982). Fusion of the eyelids involves only the peridermal and epidermal layers, the mesenchymal layers of the lids remain separate. During the fusion of the lids, the periderm cells join and then are displaced to the outer surface as the epidermal layers fuse (Maconnachie, 1979). The periderm layer is shed when keratinization of the epidermis takes place on embryonic dl7 -18, but is never keratinized itself. The fusion of the epidermal layer is temporary, lasting only until approximately two weeks after birth.

The first signs of eyelid separation occur at around dl7.5 of gestation as a slight depression on the external surface opposite the epidermal plug which extends between the two eyelids (Findlater et al., 1993). As keratinization of the epidermal cells between the two lids progresses, the groove begins to deepen, and separation of the two lids moves from the epidermal surface towards the conjunctival surface. A corresponding depression to that mentioned above is first evident on the conjunctival side around 10 days after birth, and keratinization then appears to extend onto both the conjunctival and epidermal

sides until final separation occurs at around 12 days after birth (Findlater et al., 1993).

B. Genes expressed in the developing eyelids

A review of the literature found a number of papers which have both directly and indirectly investigated the expression of genes during eyelid development. Some of these results are indirect, since the eyelids were not the focus of investigation, while others looked specifically at the eyelids during their growth and extension across the eye. These studies are summarized in Table 1, which reports the genes, their map location (if known), and expression patterns.

5 Table 1: Genes expressed in the developing eyelids

Gene Gene Map Expression pattern/tissue (embryonic) References2 symbol location1 short stature Shox2 Chr 3, from day 8 onwards - craniofacial expression was observed 1,2,3 homeobox 2 31.6 cM in cells condensing around the developing eyelid, nasal cavity, and palate; also seen in heart and metanephric mesoderm; highest levels of expression were found in mesodermal tissues of the face involved in nose and palate formation, the developing eyelid and tissue surrounding the optic nerve days 9 - 16 - heart, otic region, maxillary and mandibular components of the first branchial arch, nasal processes, eyelid, midbrain, medulla oblongata, limbs, dorsal root ganglia and genital tubercle day 16 - inner layer of fused eyelids; highest level in region adjacent to the forming conjunctival sac Eph receptor Epha7 Chr 4, days 12.5 - 15.5 - most evident on the inner sides of the 1,4 A7 (a.k.a. 8.0 cM eyelids; also seen in inner ear, many areas of the embryo developing face, lungs and bronchi (from days 10.5-14.5), brain kinase) toes, genital tubercle, ears, many parts of the mouse and nose Ptch2 Chr 4, day 14.5 - lung, stomach, intestine (mesenchymal cells 1,5 homolog 2 Syntenic adjacent to the endodermal epithelium); epidermal derivatives: the developing nasal gland and eyelids sonic Shh Chr 5, day 14.5 - lung, stomach, intestine (mesenchymal cells 1,5 hedgehog 16.0 cM adjacent to the endodermal epithelium); epidermal homolog derivatives: the developing nasal gland and eyelids patched Ptch Chr 13, day 14.5 - lung, stomach, intestine (mesenchymal cells 1,5 homolog 36.0 cM adjacent to the endodermal epithelium); epidermal derivatives: the developing nasal gland and eyelids transforming Tgfa Chr 6, day 15.5 - concentrated at the advancing margins of the 1,6 growth 35.8 cM eyelid epithelium factor alpha day 16 - localized to the epidermal cells that bridge the growing eyelids day 16.5 - restricted to the cells at the outermost layer of lid fusion and cornea epidermal Egfr Chr 11, day 15.5 - throughout the entire epithelia of the eyelids 1,6 growth 9.0 cM days 16 and 16.5 - corneal epithelium and in all cells in factor between the eyelid margins receptor integrin Itga9 UN day 13.5 - detected in the developing epidermis at the site 1,7 alpha 9 of emergence of the eyelids day 15 - localized primarily to the non-basal cell layers making up the developing epidermis, including lid; also seen on a triangular cluster of cells forming the leading edge of the lids as they migrate over the surface of the cornea day 16 - epidermal cells which formed the epithelial bridge between the two eyelids tenascin-C tnc Chr 4, day 16 - detected between the epithelial cells comprising 1,7 32.2 cM the bridge joining the two lids early after fusion

6 Table 1 cont.

Gene Gene Map Expression pattern/tissue (embryonic) References2 symbol location' integrin Itgb8 UN day 13.5 - low levels in the epidermis of the developing 1,7 beta 8 eyelid day 15 - localized primarily to the non-basal cell layers making up the developing epidermis, including lid; also seen on a triangular cluster of cells forming the leading edge of the lids as they migrate over the surface of the cornea day 16 - epidermal cells which formed the epithelial bridge between the two eyelids integrin Itgb4 Chr 11, day 15 - detected underlying the epidermis of the lid 1,7 beta 4 76.0 cM integrin Itga6 Chr 2, day 15 - detected underlying the epidermis of the lid 1,7 alpha 6 38.0 cM laminin LamaS Chr 2, day 15 - epithelial cells at the very tip of the migrating 1,7 alpha 5 106.0 cM eyelid day 16 - detected in the bridge of epithelial cells between the fused lids 'cM position from Mouse Genome Informatics, 1999 Chromosome Committee Reports (MGI4, 1999); UN = unmapped 2 References: 1, MGI4, '99; 2, Blaschke et al., '98; 3, Semina et al., '98; 4, Ellis et al., '95; 5, Motoyama et al., '98; 6, Berkowitz et al., '96; 7, Stepp, '99.

C. Open eyelid at birth mutants

During my study, a literature review of known open eyelid genotypes in mice was undertaken to enable a comparison between gaping lids and other phenotypically similar mutations. This review included mutations where open eyelids at birth was the main defect, i.e. nonsyndromic, as well as those which had open eyelids associated with a syndrome, or with ectopic gene expression, or chromosomal abnormalities. It also included normal strains which have an apparent susceptibility to this defect. Tables 2a-2e summarize the literature survey of open eyelids at birth genotypes in mice.

Open eyelid mutations are known on approximately half of the in the mouse genome, with the exceptions being chromosomes 7, 8, 9, 10,14,17,18,19 and X. Chromosomes 2, 6, 11, 13, and 15 had at least three different open eyelid

7 mutation loci. There were also 11 currently unmapped mutations (including the 4 lidgap hypomorphs). The mutations that were considered to be most similar to gaping lids, the loci of which had not been excluded from allelism with gaping lids by previous linkage studies, were: transforming growth factor alpha/waved-1 (Tgfa/wal) on chromosome 6; lidgap Gates (lgGa), near integrin alpha 2 (Itga2) on chromosome 13; and the epidermal growth factor receptor/waved-2 (Egfr/wa2) on chromosome 11. It was thought that the waved hair that characterizes known mutations at Tgfa and Egfr could be due to loss of specific exons and that they might not be caused by all types of mutations at these loci

(D.M. Juriloff, personal communciations). The first arch (Far) locus was also a candidate because open eyelids occurs at high frequency in homozygotes, although other defects in the known Far mutation follow a dominant mode of inheritance. These loci are further discussed below.

a) the lgGa locus

The lgGa mutation is recognizable in homozygous newborn mice by their lack of eyelid fusion and wide open eyes (Gates and Bozarth, 1968). Originally called ophthalmatrophy (oa; Gates and Bozarth, 1968), it was later found by complementation tests to be allelic with the lidgap (Ig) mutation and was renamed lidgap-Gates (lgGa; Boyd et al., 1984). The only identified defect is open eyelids at birth and subsequent corneal opacity or degeneration of the eye. An inbred strain homozygous for lgGa (LGG/Bc) was created at the Juriloff/Harris Animal Unit (UBC, Vancouver, Canada).

After outcrosss, lgGa behaves as a fully penetrant Mendelian autosomal recessive in first backcross progeny but shows apparent reduced penetrance in the F2 (Juriloff et

8 al., 1996). An average of 19% (range of 15-23%) open eyelid F2 progeny were recovered from nine F2 generations obtained in a study involving this mutant. This pattern fits the frequency associated with a second unlinked recessive suppressor-of-lidgap locus introduced by some normal (i.e. non-lidgap) strains (Juriloff et al., 1996). In other words, a quarter of the 25% expected open eyelid progeny are suppressed and therefore appear

"normal" at birth. Even though there are fewer than expected open eyelid pups recovered after outcrosses, of all the lidgap mutants, lgGa is the "simplest" genetically. It was mapped to Chr 13 near the integrin alpha 1 (Itgal) and integrin alpha 2 (Itga2) loci

(Juriloff etal., 1996).

b) the Tgfa/wal locus

The transforming growth factor alpha (Tgfa), is one of the most extensively

studied growth factors which act on cells to regulate cell proliferation, differentiation, migration and adhesion (Reneker et al., 1995; Berkowitz et al., 1996; Luetteke et al.,

1993; Mann et al., 1993; Dunn et al., 1994; Luetteke et al., 1994) through interaction with

growth factor receptors, in this case the epidermal growth factor receptor (EGFR;

described below). Its production is most often associated with transformation and tumorgenesis, but it is also implicated to play an important role in cell migration and

control of cell differentiation (Reneker et al., 1995). Tgfa is expressed in a large number of tissues, including the eyelids, during development and adulthood. The gene was mapped to Chr 6 by Fowler et al. (1993).

Two independent studies have investigated the physiological role of Tgfa by

inactivating the gene. Mann et al. (1993) disrupted exon 3, which results in absence of

9 functional gene product. They found that 100% of homozygotes for this targeted mutation had waved coats and whiskers but only observed occasional open eyelids at birth. Luetteke et al. (1993) disrupted exon 4, which encodes both the third disulphide loop of the mature growth factor and the transmembrane domain of the precursor, which also inactivates the gene. They found that 100% of homozygotes for this mutation had waved coats and whiskers, but observed variable frequencies of open eyelids at birth.

Both of these studies recognized the similarities between their knockout phenotypes and that of the waved-1 (wal) hair mutation (Luetteke et al., 1993; Mann et al., 1993).

Luetteke et al. (1993) determined these two loci to be allelic by complementation tests between a test strain (ABP-Le) homozygous for wal and mice homozygous for their Tgfa null allele.

The initial waved-1 (wal) mutation arose spontaneously in a mixed mouse colony in 1930 (JAX, 1999), and this mutation has recurred independently five times on different genetic backgrounds (e.g. walu - wal53; MGI4,1999). This recessive mutation is characterized in homozygotes by curly vibrissae and a waved pelage, and frequent open eyelids at birth (Berkowitz et al., 1996). The open eyelids associated with this mutation were initially ascribed to the actions of a linked gene (Bennett and Gresham, 1956) but were later found to be part of the spectrum of defects associated with the wal mutation.

c) the Egfr/wa2 locus

The epidermal growth factor receptor (EGFR), a 170kd transmembrane glycoprotein cell surface receptor (Todderud and Carpenter, 1989; Merlino, 1990;

Adamson, 1990a, 1990b), is one of the most studied receptor tyrosine kinases (e.g. see

10 Brown, 1995; Hsuan et al., 1989; Wiley et al., 1995; Dunn et al., 1994; Mercola and

Stiles, 1988). This receptor is found in most mammalian tissues, both during development and in adulthood (Miettinen, 1997). The receptor contains three domains: an extracellular amino portion, a hydrophobic transmembrane region and an intracellular carboxy portion which contains the tyrosine kinase (Merlino, 1990). Binding of several structurally related ligands to EGFR activates the receptor through autophosphorylation on the kinase domain (Miettinen, 1997). These ligands include epidermal growth factor

(Egf), transforming growth factor alpha (Tgfa; described above), amphiregulin (AR); heparin-binding EGF-like growth factor (HB-EGF), betacellulin, vaccinia virus growth factor (VVGF) (Brown, 1995; Merlino, 1990; Luetteke et al., 1994), urogastrone

(Adamson, 1990a), and cripto (Dunn et al., 1994). Activation of the receptor begins a cascade of events which leads to changes in cell behaviour, cell proliferation rates, cell migration, cell adhesion and/or cell differentiation (Wiley et al., 1995). Studies of EGFR function in early mammalian development have indicated that it appears to play an important regulatory role in the pacing of epithelial development. Egfr, originally called c-erbB, the cellular homolog of the viral erythroblastosis transforming oncogene, was found to be linked to the alpha-globin locus (Hba) on Chr 11 (Silver et al., 1985) and placed at the proximal end of the chromosome in cytogenetic band Al by in situ hybridization (Munke and Franke, 1987).

The waved-2 (wa2) mutation, now considered to be at the Egfr locus, arose spontaneously in Dr. Clyde Keeler's laboratory and was described as being phenotypically similar to waved-1 (wal; described above; Keeler, 1935). The open

11 eyelids found with the curly vibrissae and waved pelage hairs of wa2 was not part of the initial description by Keeler (1935) but was later described by Butler and Robertson

(1953) in their Db A wa2 strain of mice. Luetteke et al. (1994) investigated whether wa2 resulted from a defect in either the expression or activity of Egfr since they both mapped to the same region on Chr 11. They found that wa2 EGFR contains a point mutation (a single nucleotide transversion resulting in the substitution of a glycine for a conserved valine residue) within the tyrosine kinase domain, and this mutation diminishes the activity of EGFR by 80-95% compared to that of control levels. Anywhere from 5-60% of Egfr/wa2 homozygotes have open eyelids at birth, depending on the genetic background on which the mutation occurs. Homozygotes also have curly vibrissae and guard hairs as well as curly pelage, which straightens with age.

Recently, three independent groups have inactivated different regions of the

EGFR gene. Miettinen et al. (1995) inactivated exon 2, which encodes the amino- terminal segment of the EGFR. Of the mice homozygous for the mutation, 100% were found to have open eyes at birth and short curly whiskers and died within 8 days. In a study by Siblia and Wagner (1995), part of the first exon was replaced by an Escherichia coli lacZ reporter gene, thereby inactivating the gene. At birth, mice homozygous for this mutation had absent or rudimentary whisker stubs and, after embryonic day 16.5, surviving mutant fetuses had open eyes. The third study by Threadgill et al. (1995) also inactivated the EGFR locus. This study is described in detail below, as mice with this null allele were used in my study.

Threadgill et al. (1995) created a null allele at the Egfr locus by homologous recombination in 129/Sv mice (129/Sv-derived DS embryonic stem (ES) cells, bred into

12 CF-1, CD-I, and 129/Sv stocks). The homologous recombination replaced 155 bp surrounding the splice acceptor site of exon 2 with a Neo cassette. This results in aberrant splicing around the targeted exon joining exon 1 to either exon 3 or exon 5. The first event would result in a nonsense protein, whereas the second retains the reading frame but removes domain 1 of the extracellular region essential for production of a mature EGFR. Intercrossing heterozygous mice in each line revealed that 100% of homozygous Egfr'1" mice on the CF-1 and 129/Sv backgrounds do not survive to birth

(CF-1: peri-implantation lethality due to degeneration of the inner cell mass; 129/Sv: mid-gestation lethality due to placental defects). However, on the random-bred CD-I stock background, some pups survive as long as postnatal day 18. Of those that survive,

100% have open eyelids at birth, due to failure of eyelid formation, and rudimentary waved whiskers that uniformly curl anteriorly and are fragile (Threadgill et al., 1995).

d) the Far locus

The first arch (Far) mutation arose spontaneously in the BALB/cGa (Gates) strain of mice (McLeod et al., 1980). The most readily observable defect in the Far mutants was open eyelids at birth, seen in approximately 70% of homozygous newborns. Most homozygous (affected) animals also have a cleft secondary palate and therefore do not often survive past the first postnatal day. Far causes a severe syndrome of craniofacial defects including abnormalities in the zygomatic, squamosal, sphenoid, and palatine bones, the stylohyal cartilage, the premaxilla, maxilla, malleus, and mandible (McLeod et al., 1980; Juriloff et al., 1992). All of the known defects are derived from the anterior first arch, and to a lesser extent, the dorsal second arch (Juriloff et al., 1992). Additional

13 defects noted in a population of Far homozygotes included skin tags on one or both sides of the face, lack of one or both infraorbital vibrissae below the medial margin of the eye, and bilateral deficiency and lack of linear organization of the maxillary vibrissae (Juriloff and Harris, 1983).

On the BALB/cGa background the Far mutation appears superficially to be autosomal recessive (McLeod et al., 1980), however, on all genetic backgrounds tested

(e.g. ICR/Be), Far is actually semi-dominant, expressed in all heterozygotes as an aberrant major bifurcation of the trunk of the maxillary nerve (Juriloff et al., 1992). This defect is the most consistently expressed dominant effect caused by Far, whereas the open eyelids associated with this syndrome mostly follows a recessive mode of inheritance. On the ICR/Be background, Juriloff et al. (1987) show that among heterozygotes approximately 1% have cleft palate and approximately 5% have open eyelids. The open eyelids associated with homozygosity in Far is a result of tissue deficiency in the lower eyelid in comparison to normals (Juriloff and Harris, 1983). The

Far locus has been mapped to Chr 2 (40.0 cM), closely linked to Ulnaless, near the Hox4 cluster (Juriloff and Harris, 1991).

III. Mouse Mapping

A. Overview and history

The mouse is the main mammal for genetic analysis, making it ideal as a model for studying human genetic diseases (Copeland et al., 1993). As such, genetic maps of the mouse genome, a "picture" of the locations of a set of loci within the genome (Silver,

1995), have been constructed. The resolution of the maps varies from the lowest level,

14 £ 3 B — „ N -5 c c <2 s — -S ° c"S " ° ^ s T3 ^ c « 8 — — cd cd C cd= 73 cj cj B - BO • - O >. ^ .a 3 '5i • > n CJ « -a -fi .ti S o _ o a Si s ° '5 2 TJ XI C *3 cd S o. « c « O K — a e S3 ES- 00 w O O 8 "= i- •I I £ N a. , *i C dj O •r- o t: 3 00 M cd "rt cj ^ O i T3 cr8. 1 E -a o c U

T E :i ac)j ; JO a •a .9 BE g a „ E El 1 S -a S .2 o .E 4= EM a .a E

c/> c O 00 c 00 E 3 OX uB w*5 00 cd B J3 hi, o -5 8. 1 c a c o CN JS T3 o ex .3 O V CJ cd j= -a oo J3

1?

-2 . a A eq PI'S I CQ cd » 3 3 « c eo E (2 5 S

_ o U S 2 i. O 5,

"S .E m •=

D. .5 a c s 3 cd —. ^ w

15 »> .5 18 J H Os " S rf t CU « 3 a S -i o ° 3 E S £ u IS I 1 8c n 1 ° s oo §to - o u 2 £••§••§ "i I u ^ c 15 J3 •° - <2 S A5 =~ £„ *.g § « cs « E HO, 3 "o N a s I s « n° ^5 O M ~ 5 g s^l * g> e 3S^ * I 3 V E 3 « g-E CgL U to C/1 — o -a E o o o « cn . ^ S >^" „ oo tu _r c cu co .S3o T _ SI •*-» —1 s'to to K o „ a CN ON 3 „ CO tu 00 ^ o tu tu CU 00 c a S a S "> •£ c cn * o ° g. •gout O M g &3 o E >•, E >>\ ^ „ o t3 £ o» gS uS S *wS o -a oo o oo rN 1 S 3 h 2 "= o E E2 5 = 2 § S ..S ias=> ? ofr o S ,_ l-S 8.2 CN O —' ^ O c oJ S x c 1E a ™ ° - °^ c o cu O O tr CN O l-S 8.2 coo) C8 oo J3 S u. c c o o ( t. ooO v - tu vo C/5 "o " M •>-> 5C vo 3 S Ov o Q TO +^ Ov j= _T cn" "ea CN o .« o tu "^1" o ._ oo o Q P U 2 o « c0 cd 4_> t/i .13 > a> .a or -g „ CO IS co o- a G w 5 r.'i I s tu cdv o - c - CN cu CN o ca r (N C/3 A a vcT c i o. 5 ca O cSd- c- ACo cd Co 00 CN. 3 fe" oo v 00 v 00 v Ov Ov •a oo 00 § V 12 -2? •o bp S 00 ^ 5 w a ici O _ o T3 = -2 VOOv C" 5 13 ii otu a u ^ 1(§ -? O CN CU cv> -C rn 3 ov s- >> O cu 16 •O „ B > E S E. « i IiII§ a s 3 • ,«> a 3 i ™ 3 tzi u « y 3 . .2 B ' w -O B a E K c fN> 3 • O « i " ^ * ' " -s w3 2 A g ' A s -a * •« E 2 u ~ " 8. 3 "2 cd M 3 S v. £?'» 2 3 mo ^o 5^ «> 1 s a — S Ha o. & 2 8 -S § 1 u ed < C O o ^ E « 551 i| 11 00 fi >» a u CJ P u 2 £ & u '5 O E o o gsi

4> y (- E M g SsS 00 CoM U to 5 -2 Sill >>£ B E IS ° C *o- -*a S-g5 O — 5. 3 .§• 5 -5 -5 x .5 ca .ti a. « u u 2 "3 Jj c s t- o '" - .S » •= B irt eo S C e cd IS s 3 " 11 '3 '£ S J3 o -y .ST «> u M £ 2-1 - _ o o t_ co cc O .= I a g .s g *"§| S «3 o *o o =5881

3 60 l*-gf III — O B 7.5.8

B3 EB

o u £ 2 2 Q. 1—1 o o.

17 _ S 73 (i3 .>>. 2 •S "3 = 8 S . o < _ O_ e*-« eu •sic E E •S cn o o S.f-3-i 5 B o r gE 5 ^ (0 £ *J 73 73 J3 c*3 -O o _ E -o o =-

eu « S)c2 g - cu

1 o E o I -S Z .2 I 3 eg 3 o S c00 c73 ^ C cd •;- cn o .3 os eu a is 2 S e .52 3 iB - _ ° S u S3 ^•o 2 c cu cu — EESo cu S °- g §• s I & E. 5g £ g I a ^ « tl CU eu Q. C o 45 o 00 frfr_g 2 -S« fr >> 2 o CL o E 2 B lU. O o E s eu E S c >- J3 C O ed C E-S I*- cu O P3 l._ tt U in 3 i j3 £ C

8 « o o

Is '1 ^ IU u u

ill

_ o uc -ac

S E o x>:> otu

18 J3 T3 oj IS .5 E oo w OS _g _o o •if a> o cj ft' c -a e .52 « o —a «5 c Jl e 5 T3 3 o e o u £ S o o a E g •O S u o u ° H ca aj S E 1.,5 o cj t3 o p D .O u "° O -H TJ a o s a ~| " cd on f 9- C O a. 5 2 o ca o o •o o.— " eo o on 3 c c/1 *"' •- s c.j £."S u B S - -o - S u u ffl • •- Jw3 CB xO -a g S c .o cj C 0. .3 si. •2" ^ > ~ 2 o s § i .5 g g g s £ it .2 £ g g .2 '5b a -i. § 3 o o -S 3 > O -3 C CJ X3 J- u w |- c -o STca Sf S i — CJ ~ o S S.5 S3 S•.3 '> 2 S X o 5 o

(N cn m rn o" rn m m a* CN ON CN CN m

© c oo CJ 5 t3 e -i. ™ x° e cj cjsp £ £ u E fe-T3 ^ C > * J> g t J= B u OO J= J= 2 I a. .=

= i

_ o B -C II B O P.

19 •S •aa

-EM

m oo —' ^- - — tN

00 c — — O. 00 c .5 & o 1 § a I J i- .2 5 o S s.1 Tf! OD- OX) § i I E-2 2 5 2 2^ 1 >> to .5 ! O 00 P C OB i

3 n ° C w 00 C o > u u Qn ° 00 g 3 C 3 oo > O. o o 0-0 = oo g 3•» * s ^ \D C ^ CU CU .id O3 r2s .t3? O cu 00 , c > cu a, c o j= -S XI — CU X> 1 o to >. cu cu cd ^ £ eu 73 oo xi

E £• •'8..S

I u -2

MS

* 7^ 1 S

I 73 > i v V V 63 A

S Xi eu s

20 tN OS —" oo"

•s 00 TS «§ CJ on ° C C a OJ .= g • 2 a o as o eo c > •= V o 'ob E cj •£ 1 ° cj .S S £ 8 8* « "5 « CL O IO 5 3 O. fc» X C j- 2 E

co (S o .5 v >N *J cn » e«

Mi

41 • S

•3 C3 Cl, >> V o cn 5 o

_ o B -=>

o 3 J

21 a •o" -a — t; ca g O 3 i J S -o E C &> 00 •3 1 o S: 3 '3 5. & ca xi - S la :« ^ a s op a cj fJ 5 cj CO on• a ' S-.H E o t" ^ E o_o "_S 2 _ tj M a 5 - a i co qj ' c a 3 S 00 >>.2 g 53 3 jl -a § £ .2, a 2 s 3 ~ *2 ° I-Si •—•op ca cj ca M 3 oo 3 o o xi P o CJ 3 2 u cn CJ O O .3 £P § a E -J .S o c o.

-3 3 £ " XI (S) rs\ Cd •O E § S E 5 S 5 E o .E .E 0 o -3 -S 3 « 3 a -o E o- ca O LS CO S 3 3 - Ti o O -a — w. g '3 P - - WHS •° is CJ . U < 3 « •- 3 E fe 1! o c H MS 3 O s <; < j CL

CJ .3 ^ oo S ca o >i Q. ,„ OO III"" « £ ¥ <^

c o o o V cj ca s I 2 ii 3 E .2

Sa -O5-2^; a 3 ai | I - .2 >. v cd cj ,a H E < b

•S 60 .2 — E 8 ? S S S S

22 2 o •8 I —cd cdo I* % s &| :E •§ j= 0 3 o cd <- e c 3 n =n o u S £ ob u- > o 2 •= ? XI 3i § £ .2 73 73 2 6 2 fj •o tu cd P c S3 o 3 § 1 S ~ Q, XI 3 = E •a E s S 9* "cd «5E I 'Si 111 0> 73 .O > .ti II o cd a. 00 § <§ XT' JS — ' CJ >. =j .. « 2* » Q. * ro cn tu E 1 00 £1" LT •2 42 fj - ». - E >, >• P o CJ CJ 2 -C 1/1 >, fj rt is si & u .2 •S E u 5? oo u x; x: fj il3 s ed cu2 S E >> o S3 E CO o £ 3 " C = IT

?0

4> > w BO tU -3 ti « c CeM a = a .2 2 00 73 < .a « r .2 s 2 XJ 73 < 3

< 8 ox o o 2 i 2

4* O co I t-J +•

u .2 cJ 5 « * o 5 &

a

a -

51

oo to cu cd

0*5 -- - •o D. 3 cd -w cd k« t- , oi]

23 OB C U T3 ' S?| *x1 C S Z -a 2 .2 C nl - y O ca £ cd

co o

xi .5 2 • i « o 8 2 -a •a "< 'cri £ 2 3 -o * p S .- -a 00 g E c _ oo (L> O 03 00 cu o o C N S - u ca S x— 5 E £ o J: s •» g >! 2 = 2 1 S a. 53 H ,1-1 E •c E •as! cd i—i — C*U ^Icg ',3 co * 1 •£ 8 B ° 8" cu eu 1 - •« 2

E .E g O cd 3 TD 0J)cC

'5. J> c3 2 E. o

"5 s lu S 5

CN CO CN) tu C

oo « 2 00

24 (o P.

c = 2 e 3 ~ 2 e •3 O- CJ o • °- .2 — c« -120 00 CJ Jt2

c = e-1 2- CQ S _ 3 o CJ T? cuts ic o "3 .= j= .is -o

S 1 "3 = SS.2 _00 *-' O •r- TJ r- cd » 3 .2 co. -s .E "o o c en -3 -3 2 * O .£P(N E o +•* .3 o o ca on in •B ~ a> a a « « .» Bn i. CCD £ c 8 e | o « 0> e cn •S le s 0 2s = •s c/i ra00 tu D. CD JD a — E a "I .E S o = -c .2 .If +•» VI O . „ ca c — — ^ CL T3 S.8S s .11 ..O g 1 !B a3 00 „ £. g . .1 £ 1 a E co erP aS — a 00 •— E C 3 O o t8 u 3 c c s a § £ g * o .2 a s -s ;§ a •O T3 C 2 oo • <2 O 3 O. D- a .gp .2 -c ° c ca I -s 1 » ••a a O• g — tj so o £ o op oo-? = a u c o £ 3 CJ CJ o - a u - — J3 S§c=.E j) r CL E o n CJ u o. , 2 C 3

5i

E .S g •o toca 3*

O CJ c C 5 u - ,2 ua j, ,, 8 m g o o CJ l) ° CJ — —i S in »n C CQ. CD. W ^3 i Cd O O - .2

o w

_ o e xi 4) c a

.5 m ob 2 cj 2

25 CD c i =3 ^ x: o ,73 ca 'X cn cn x> _ 2 '= 5 E = <3 o 2 So. S S 52 — §:§• is •22 " 8 73 8 « -G e _ ca c XI eu ^ 73 •- - ~ • 5 c 3 <*- 2• § £ g o i" E c = x. -5 ° -5 5 § o ca ,o e .ca".2 &! 5 •a is 'o. S o ca u o ag o ca« 73 o -° . -a c ca 0 B -a 8 -3 2 c3 73 £ 2 S 2^ ccu 2^ o. M • a. E « 3 b *s •= ca .S3 il S c tt E is xi a 1.2 8 cu ox > . ac o.XJ 3 3 2 • E cn oES e * u ca a. o.

a „ O (3 • i I = . — oo P £ .S x a § = C/5 — E .2 .i= cu 8 -a i 11 0§0 gM ca P 0 1 73 -S g nuiiiiliSo. i f:i I c S « I P ,i 3 « o 3 c2 a a °0 3 •= cu C 73 CJ o 73 2 "75 . 73 W cu G 73 Xi XI 73 g 73 _ •v o £ -o •= xi > S •S ° ».2 73 -2 X 73 E a 5 P eu E .2 73 oC >LoH K HE ?? 73 C G — 60 5 Mf O tn O 73 o -t; x £ cd S "O — E 73x . a cu eEu — S S .& o 2 c ° CU cn .3 cu X W = r- " _ eu — u 1, > Ox S g £ i •° 2 S I o out D. 73 > GI c S o >• o o g .£ X — o 5 Is S.EK eu -H *c3 ca

3 S? o ct» O. 00 ° •n 3 in C O X)

I "S .2 G o .a oE .5ca ga T3 00 CC E .3 g op ooecca a:

I - -> ~ cu ccnn O, CU 00 o. eu cn eu 2 O. S 2 a o. .E S E I E

2 t. xi U

,8 «N

B

26 CQ CJ _ « -r3 C3 S "3& PS O •3 a g E a | s T:" CJ a _ « S «i J3 2 a « O — J= o S oj aj >< c .«a o So 5« -3 o _-

.s-g £ - § 1? •a rt 7 a > g en CeM 60 E tP to O g 3 "2 S .2 oj o II f . aj T3 x: > ac .2 •§ c co "r; >, 2-.E c •2 u§ o * aCf>l ^« s 3 ° i— — •3 ! ^ E £ os CX 3 OS •° 13 — OS •o -a s S c s a

c E2 SS g•= t O £ 3. .3 to -3 O X: I*

§ OS E .3 g E oo o ca 3 O OOcC |3 IT tu cd CU 60 .2 oo & « o cn co o c a cn B S 13 S a u a 2 B.-3 • m o X <*; CJ CJ O U cn" os • - Os >£ is Os Os

o C a 3 IS 3 »0 cd , = -2 |«fcj - V s s « 0 o 1 i CL. ,

j= CQ

27 O.

•2 E

O 60 O 00

co -O

03 O E 73 S CO

"L»

; 2? e o o "ca M u E K E S S ! S co 00 co ; C :ie § s .5 j c o o to CU ~ ..— — "i* 8.caj i ° 5? — ii O CO E ; " Gc .s£

i Jc=o SC3 SO 1=X! a o

.SI

s -Cl

2 •5 42 s

Pi

28 e.g. placing a locus on a specific chromosome, with the levels ever increasing to the highest, e.g. mapping a locus to a specific DNA sequence. Genetic maps can be divided into linkage/recombination maps, chromosomal/cytogenetic maps, and physical maps.

Linkage maps are generated by examining progeny of classical breeding studies and identifying recombinants. The distances on these maps are measured in centimorgans

(cM), where 1 cM is equivalent to a 1% frequency of recombination between 2 loci. A chromosomal map is based upon the karyotype of the chromosomes of the mouse genome. Gene position, identified using band names delineated by Giemsa staining, is determined using cytogenetic analysis (e.g. somatic cell hybrid lines, karyotypic abnormalities which appear in conjunction with particular mutant phenotypes or in situ hybridization) or linkage to a locus previously mapped in this manner (Silver, 1995). The third type of genetic map, physical maps, are based upon analysis of DNA directly. The units of measurement are basepairs (bp), kilobasepairs (kbp) or megabasepairs (mb). Of these three maps, only physical maps can be used to describe the actual distances, i.e. length of DNA, that is between two loci. The distances determined by recombination maps can be influenced by recombination hot and cold spots which can give over- and under-representations, respectively, of the distances between two linked loci, and distances determined on cytogenetic maps can also be distorted since the DNA of the chromosomes being karyotyped is condensed, due to the nature of the procedures involved.

Generally, all three types of genetic maps are amalgamated in mouse studies, with classic linkage studies providing the major bulk of positional data (Silver, 1995). These linkage data include the classical studies of visible phenotype, isozymes, alloantigens,

29 cloned loci and polymorphisms of DNA sequences. Historically, much of the work done in mouse genetics was based on visible mutant phenotypes (Copeland et al., 1993;

Dietrich et al., 1992). This method involved two or three point crosses and was extremely slow since only one locus at a time could be examined (Silver, 1995). In the

1970's highly polymorphic enzyme isoforms were identified between different inbred strains, which facilitated mouse mapping, but was still hindered by a limited number of loci across the genome. In the 1980's, however, the identification of

DNA polymorphisms at "loci" that did not produce a visible phenotype and the introduction of the interspecific backcross allowed for an explosion in mouse mapping studies. The polymorphisms that were most readily used was restriction length fragment polymorphisms (RFLPs) identified by Southern (DNA) blots. The usefulness of RFLPs was limited, however, by the common ancestry of the traditional inbred strains, which reduced the identifiable polymorphisms. The second breakthrough in the 80's, the interspecific BC, involved crossing two distinct mouse species, an inbred lab strain (Mus musculus) and a distantly related species, generally Mus spretus. It was found that fertile female Fl hybrids resulted from this cross, which allowed investigators to follow the segregation of loci when the Fl female was backcrossed to one of the parental strains. In this cross most RFLPs were informative, and therefore thousands of loci potentially could be mapped in relation to each other in a single cross (Silver, 1995). This method allowed for the construction of the first complete linkage map (i.e. loci were identified on every chromosome except the Y) of the mouse genome based on DNA markers and provided mapping panels which would allow for mapping of new loci identified at the DNA level

(Silver, 1995). However, RFLPs did have limitations. Identifying them was tedious and

30 time consuming. Their polymorphic content can be limited, and the procedure can be quite labor-intensive.

B. Polymerase Chain Reaction and Simple Sequence Length Polymorphisms

a) Polymerase Chain Reaction (PCR)

In the late 1980's and early 1990's a major breakthrough came in the form of the polymerase chain reaction (PCR) and the identification of DNA polymorphisms, e.g. minisatellites, which could be rapidly amplified by this method (Copeland et al., 1993;

Silver, 1995). PCR is an in vitro method for the exponential amplification of a specific region of DNA (Saiki et al., 1985, 1988). Basic PCR involves the use of two oligonucleotide primers that hybridize, in reverse orientation to each other, to the opposite strands of DNA flanking the region of DNA to be amplified. Repeated cycles of heat denaturation of the target DNA, primer annealing, and extension of the primers by utilizing a thermostable (Taq) DNA polymerase, results in amplification of the target

DNA (the region 5' to the primers). With each round, the DNA synthesized in preceding rounds and the original target DNA are used as templates for further amplification of the region, resulting in exponential amplification after the first round. After 25 -30 cycles, the target DNA can be amplified by several millionfold. Both the annealing temperatures

and the concentration of the reagents can be adjusted to increase or decrease the

specificity of the amplification reaction.

PCR quickly became the preferred method for generating linkage data, since its

enzymatic amplification, gel electrophoresis, and ethidium bromide staining were quick and simple. PCR also allows for rapid and exponential amplification of DNA sequences

31 present in very low copy number, high resolution of polymorphisms ranging from a single basepair change to large rearrangements, and fairly inexpensive set-up and typing costs for large-scale projects.

It is possible to type a number of DNA polymorphisms by PCR. These include many of the RFLPs previously defined by Southern blots (Silver, 1995), minisatellites or variable number of tandem repeats (VNTRs), single strand conformation polymorphisms

(SSCPs), random amplification of polymorphic DNA (RAPD) and simple sequence length polymorphisms (SSLPs). RFLPs and SSCPs only have limited polymorphic content and tend to be di-alleleic, VNTRs occur in relatively low numbers, i.e. fewer than

1000 loci, and RAPD techniques are random, amplifying multiple segments across the genome, which prevents the assignment of linkage to a particular region. SSLPs on the other hand, are almost ideal as linkage mapping loci.

b) Simple Sequence Length Polymorphisms (SSLPs)

SSLPs, also known as microsatellites or simple sequence repeats (SSRs), are mono-, di-, tri-, or tetrameric sequences repeated multiple times in a tandem array (Silver,

1995; Copeland et al., 1993; Dietrich et al., 1992). In mouse the (CA)n(GT)n dimer is the most common class, most likely generated as a result of mispairing, or slippage, during recombination or replication within the tandem repeat sequence. They occur in high copy number and seem to be distributed randomly across the mouse genome. These repeats at each site are often highly polymorphic even among closely related strains

(where the difference between alleles is due to differences in the number of repeats or

"size" of the amplified copy between strains). SSRs in mouse have been identified by

32 screening both mouse genomic DNA and published gene sequences with probes for either

the (CA)n and/or (GT)n repeats. These methods identified an extremely large number of markers in a fairly short period of time (317 in 1992 to >6500 in 1996). SSLPs have

allowed a return to M. musculus strains for mapping studies, and hence, most current

maps are composed of SSLP positional data from interspecific crosses, placed on a

framework map of SSLPs taken from within known genes mapped by the methods

mentioned previously.

C. Review of Mouse Maps

There are several maps of the mouse genome available for reference, based both

on individual crosses and a consolidation of all available data. The consolidation of data

tends to be less accurate since the data are from different crosses which generate maps

with different distances. For this study, three maps of chromosome 11 were examined,

the Research Genetics/Whitehead Institute/MIT map, the European Collaborative

Interspecific Mouse Backcross (EUCIB) map, and the Mouse Genome Informatics

Database (MGI) map. All are available on the World Wide Web (WH/MIT, 1999;

EUCIB, 1999; MGI4,1999). Although all three maps are extremely useful, they all have

innate problems, and data are continually being modified.

The Research Genetics/Whitehead Institute/MIT map (Copeland et al., 1993;

Dietrich et al., 1992) was developed in the process of the identification of a large number

of SSLP loci in mice. With the identification of SSLPs/SSRs (Dietrich et al., 1992), a

project was undertaken to create a genetic linkage map of the mouse utilizing these loci.

To identify SSRs, an Ml 3 library of mouse genomic DNA was screened with (CA)15 and

(GT)15 probes and public sequence databases were searched for known genes containing

33 SSRs. A total of 455 primer pairs were tested first for polymorphism between C57BL/6J- oblob and CAST/Ei, and if a polymorphism was detected, allele sizes were determined in

12 additional inbred strains. The genetic linkage map was constructed using 46 OB x

CAST F2 progeny as well as 22 BXD recombinant inbred (RI) lines. The inheritance patterns were analyzed using the MAPMAKER computer program (Dietrich et al., 1992;

Copeland et al., 1993) to generate the linkage map. Markers were assigned into linkage groups and ordered based on pairwise LOD scores and minimizing the number of recombinants (Dietrich et al., 1992). Of the 455 primer pairs initially tested, a framework map of 317 SSLPs was developed which covers an estimated 99% of the mouse genome at an average spacing of 4.3 cM (Dietrich et al., 1992). By 1994 the number of SSLP loci had exceeded 1500 (Dietrich et al., 1994), and by the conclusion of this project in 1996

(Dietrich et al., 1996) 6580 SSLPs and 797 RFLPs were integrated with an average spacing of 0.2 cM or 400 kb. As the loci were mapped on the basis of only 96 meioses

(46 F2 animals) the smallest distance between markers is 1.1 cM (1 recombination event in 96 meioses) and the markers were placed in 1.1 cM groups. Since 1996, the map has not been modified very much so the positions of markers relative to each other within these groups have not been defined.

The EUCIB mapping project was begun as a means to develop a high resolution

(<1 cM) genetic map that would form the basis for the construction of a complete physical map of the mouse genome (Breen et al., 1994). 1000 backcross (BC) progeny were generated from a (C57BL/6J x SPR and/or SEG/Pas)Fl x C57BL/6J cross to allow for a map with genetic resolution of 0.1 cM on average (1 recombination event in 1000 meioses; Upper 95% confidence interval (CI) is 0.3 cM). These 1000 BC progeny were

34 analyzed for a variety of DNA markers, including genetic sequences and microsatellites, across the entire mouse genome. Three to four widely spaced markers per chromosome were scored in each animal, resulting in an anchor map of 70 loci, in which the markers chosen encompassed the mouse genome, allowing for detection of the largest number of recombination events on each chromosome. From these same 1000 animals, pools of animals recombinant in each chromosome were created (EUCIB, 1999; Breen et al.,

1994). Therefore, new markers could be tested against a panel of 40 - 50 mice to identify linkage to a chromosomal region followed by typing of a panel of animals with recombinants within a specific chromosomal region.

Using the previously mentioned methods of marker collection, EUCIB constructed a map which had fairly large distances between markers. This most likely resulted in a underestimation of chromosome length and linkage distances due to undetected recombination, i.e. double recombinants would not be detected. In 1996

EUCIB began a collaboration study with the MIT Genome Centre to map 6000 microsatellites on the European Backcross. Although they state that a number of chromosomes, including 11, are complete, an examination of the EUCIB chromosome 11 map shows a large number of SSLP markers known to be on Chr 11 remain unmapped by

EUCIB. Overall, EUCIBs primary goal was to create an international resource that would allow for high resolution mapping rather than to create a linkage map (EUCIB,

1999).

Compared to the other two maps of the mouse genome, the MGI map was created to capture, store and manage all publicly available data in composite form for the scientific community (MGI1, MGI2, 1998; MGI3, MGI4, 1999). Initially it was created

35 from the data of 356 crosses (MGI3,1999) plus data from scientific literature and various other mapping study submissions. The MGI resource is updated daily (MGI3, 1999), so this resource is constantly changing. Although this map is more extensive than the two previous maps, it also has innate problems because it is an amalgamation of multiple sources of data from numerous studies, that generated their own maps with widely differing map distances due to the nature of the crosses. This has been partly overcome by comparing data sets and maps that used a common group of anchor loci. This allows the maps to be oriented and compared relative to each other. Despite these problems, the

MGI map is considered to be the most up to date resource for mouse linkage data.

IV. Rationale and approach to this study

A number of strains of mice with open eyelids at birth are maintained in the

Juriloff/Harris (Be) animal unit at UBC, e.g. LM/Bc (see Harris et al., 1984), LGG/Bc

(see Juriloff et al., 1996), LST/Bc (see Juriloff et al., 1983) and GP/Bc. Open eyelids is a developmental threshold trait/birth defect that is often genetically complex in mice.

Unlike newborns affected with other developmental threshold traits such as neural tube defects or cleft palate, mice with open eyelids at birth are viable and can grow up and reproduce, making this trait a good model for examining the nature of gene interactions in complex birth defects in both mice and humans (Juriloff et al., 1996). It has been hypothesized that the major loci of the open eyelids mutations which follow expected

Mendelian segregation ratios may be involved in more complex open eyelid mutations, where the simpler traits have mutations at some of the same loci in more complex systems (D.M. Juriloff and M.J. Harris, personal communications; Juriloff et al., 1996).

36 Therefore, by mapping these simpler mutations (i.e. those with which expected segregation ratios are observed) first, the loci can later be tested to see if they are involved in the more complex traits. This was the approach taken for my study. Gaping lids tends to be less genetically complex than open eyelids at birth caused by other genotypes (e.g. LM/c, LST/Bc). Accordingly, once the gaping lids locus is mapped it can be investigated in other studies of more complex open eyelid mutations. With this in mind, several hypotheses on the possible location of the gaping lids locus were investigated. These loci have been outlined above. The experimental design, i.e. mouse strains and crosses, of my study is described in detail in Chapters II and III.

37 CHAPTER II: GENERAL METHODS AND MATERIALS

I. gaping lids: Scientific progress before and during this study

A. The cross to CBA/J

Prior to my project, preparatory work to facilitate use of simple sequence length polymorphism (SSLP) marker loci to map gaping lids had been started. Informative

SSLP markers between the GP/Bc and CBA/J strains at intervals across the first 11 chromosomes of the mouse genome had been identified by Diana Mah, a technician in the Juriloff/Harris lab. Identification of informative SSLP marker loci on Chr 12 to 19 and the X chromosome remained to be done, if necessary.

To generate the segregants necessary for mapping, two females and one male of the inbred strain homozygous for the gaping lids mutation, GP/Bc, had been outcrossed to two males and one female of the inbred strain CBA/J. These three mating pairs had produced six litters of heterozygous Fl animals (n = 43). Upon sexual maturity, 11 of these Fl mice (six males, five females) had been crossed back to the GP/Bc parental

strain to generate 20 first backcross litters (n = 132; 55 from GP/Bc dam, 77 from GP/Bc

sire). 14 of the Fl mice (seven females, seven males; not used in the generation of

backcross progeny) had been intercrossed to generate F2 mice (n = 249, 32 litters).

Phenotypic scoring of all animals was carried out within 24 hours of birth by

D.M. Juriloff and M.J. Harris based upon the state of eyelid closure. An animal was

classed as affected if it had unilateral (right or left) or bilateral open eyelids. Tissue

samples from liver and tail were collected individually from 61 open eyelid and 57

normal BC1 animals. The liver samples were then pooled in groups of 2 - 3. Individual

tail samples were taken from 41 open eyelid, 4 pinhole and 137 normal F2. All of these

38 samples were stored immediately at -20°C, and my participation in the study began with these frozen samples and the raw data from the pedigree cards. The crosses, numbers and phenotypes of all resulting progeny are summarized in Table 3 and are detailed in

Appendix A, B, C, and D.

Table 3: Frequency of open eyelids in newborns from GP/Bc x CBA/J cross

Generation No. litters No. No. with No. % open progeny open pinholes eyelids2 eyelids' Fl 6 43 0 0 0 F2 32 249 41 6 19 BC1 to GP dam 11 55 19 0 35 BC1 to GP sire 9 80 48 0 60

1 open eyelids includes newborns with unilateral (L or R) or bilateral defect 2 includes pinhole data

B. The cross to ICR/Be

A second outcross and observation of the segregation of the gaping lids genotype and phenotype was performed to facilitate finer mapping of the gaping lids locus. The normal strain chosen was ICR/Be because it was thought to have different informative

SSLP loci through the gaping lids candidate region identified by the first cross. GP/Bc mice (2 females, 1 male) were mated to ICR/Be (1 female, 2 males) mice. These 3 mating pairs produced four litters of Fl animals (n = 32). Upon sexual maturity, 14 of these Fl mice (7 females, 7 males) were intercrossed to generate F2 mice (n = 282, 23 litters). The crosses, numbers and phenotypes of all resulting progeny are summarized in

Table 4.

39 Table 4: Frequency of open eyelids in newborns from GP/Bc x ICR/Be cross

Generation No. litters No. progeny No. with No. pinholes % open open eyelids2 eyelids1 Fl 4 32 0 0 0 F2 23 282 25 6 11

1 open eyelids includes newborns with unilateral (L or R) or bilateral defect 2 includes pinhole data

Phenotypic scoring of all animals was carried out within 24 hours of birth by

D.M. Juriloff and M.J. Harris based upon the state of eyelid closure. An animal was classed as affected if it had unilateral (right or left) or bilateral open eyelids. Tissue samples from tail were collected individually from 23 open eyelid and 105 normal F2 animals. All of these samples were stored immediately at -20°C, and my participation in this part of the study began with assisting in the banking of these samples and with the raw data from the pedigree cards.

II. Mouse stocks and maintenance

a) Animal Maintenance

All mice were maintained in the Wesbrook Annex animal unit in the Department of Medical Genetics at the University of British Columbia (UBC), Vancouver, Canada, in windowless rooms on a 12 hour light (6 am - 6 pm), 12 hour dark cycle. The temperature was maintained at ~ 22°C (20-24°C). The mice were housed in standard polycarbonate

40 cages with dried corncob bedding and supplied with Purina Laboratory Rodent Diet

(#5001) and acidified water (pH 3.1, HC1) ad libitum,

b) Mouse Stocks

The GP/Bc mouse strain history has been described (see Chapter I). GP/Bc mice aged 3-9 months at F63 were used in this study. 100% of GP/Bc mice have open eyelids at birth.

The CBA/J mouse strain is a highly inbred strain developed from a cross of a

Bagg albino female and a DBA male (Festing, 1989). This strain is a general purpose normal strain and shows no unusual incidence of spontaneous open eyelids at birth.

CBA/J mice were obtained from the Jackson Lab (Bar Harbor, Maine, USA) for use in this project and were bred at age 3-8 months.

The ICR/Be mouse stock is a highly inbred normal strain of mice, developed and maintained in the Juriloff/Harris (Be) Animal Unit, UBC, Vancouver, Canada. This

strain was developed from the BLU:Ha(ICR) "random bred" mouse stock obtained from

Arbor Scientific in 1977 (Juriloff et al., 1989). ICR/Be mice show no unusual incidence of spontaneous open eyelids at birth (Juriloff et al., 1983, 1989; Macdonald et al., 1989;

Tom etal., 1991).

The SWV/Bc mouse stock is a highly inbred strain of mice, developed and maintained in the Juriloff/Harris (Be) Animal Unit, UBC, Vancouver, Canada. The strain was developed from stock obtained from a closed colony of mice at the Central Animal

Depot at UBC, which was derived from stock obtained from the Defense Research building, Suffield, Alberta, Canada in 1949 (Macdonald, 1988). Inbreeding began in

1959 (Staats, 1985). This strain is a general purpose normal strain and shows no unusual

41 incidence of spontaneous open eyelids at birth (Juriloff, et al., 1983, 1996). Mice from

Fl 13 of brother-sister inbreeding, aged three to three and a half months at breeding, were used in this study.

; The AXB-23/Pgn and BXA-2/Pgn mouse strains are recombinant inbred (RI) strains developed from a cross between a female A/3 and male C57BL/6J male and a male A/3 and a female C57BL/6J, respectively (Marshall, et al., 1992; Nesbitt and

Skamene, 1984). Originally created for use in the study of certain behaviours for which the progenitor strains differed, the family of AXB/BXA RI strains are also useful as general purpose normal strains, as is the case for my project. AXB-23/Pgn and BXA-

2/Pgn mice were obtained from the Jackson Lab (Bar Harbor, Maine) for other studies in the Juriloff/Harris (Be) Animal Unit, UBC, Vancouver, Canada.

Heterozygous epidermal growth factor receptor null (Egfr+/~) mice on a primarily

CD-I strain background were obtained from Dr. David Threadgill (Vanderbilt University,

Nashville, USA) for use in this and other studies in the Juriloff/Harris (Be) Animal Unit,

UBC, Vancouver, Canada. 100% of mice homozygous for the Egfr null allele that survive to birth have open eyelids, rudimentary waved whiskers that uniformly curl anteriorly and are fragile (Threadgill et al., 1995). The generation of the null allele and its introduction into mouse stocks is described in Chapter I.

III. Technical methods

A. GP/Bc Study

a) DNA Preparation

42 All mice were killed by carbon dioxide (C02) gas or decapitation (newborns) before tissues were collected. Individual liver and/or tail tip and pooled liver samples were collected for DNA preparation. The frozen tissue was either cut into 1 mm3 pieces

with scissors or crushed into small pieces with a plastic rod in a microcentrifuge tube.

300-500 of lysis buffer (100 mM NaCl, 10 mM Tris-HCl (pH 8), 25 mM EDTA (pH

8), 1% SDS) containing 65 ui proteinase K was added to the tissue in a 1.5 ml Eppendorf

tube. The tubes were then placed in a 60°C water bath for 6-20 hours to allow digestion

of the tissue samples. Once the samples were completely digested, phenol-chloroform

extractions were performed, followed by ethanol precipitation of the DNA (Sambrook et

al., 1989). The DNA pellet was rinsed with 70% ethanol, then resuspended in 100 - 200

ui Tris-EDTA (TE, pH 8; Sambrook et al., 1989). From these samples, a 1 in 100

dilution (in deionized filtered water) was made for optical density measurements in a

Pharmacia Biotech Ultrospec 2000 UV/Visible spectrophotometer. The absorbance at

260 nm (DNA absorbs irradiation maximally at this wavelength) and 280 nm (

absorb irradiation maximally at this wavelength) were taken for each sample. The ratio

between OD260/OD28o provides an estimate of the purity of the DNA where a ratio of 1.8

indicates a very pure preparation. Assuming fairly pure preparations, the following

formula was used to determine the concentration of each sample (ug/ul), where an optical

density (OD) of 1 corresponds to ~ 50 |ag/ml for double stranded DNA.

ug DNA = OD260 x 50 ug/ml x dilution factor x total volume (ml)

43 From the concentrations obtained a corresponding dilution of the stock to

1 OOng/ul (in autoclaved deionized filtered water) was made from each sample for use in the polymerase chain reaction (PCR; Saiki et al., 1985,1988; Sambrook et al, 1989).

b) PCR of SSLPs

Simple sequence length polymorphisms (SSLPs) were used as genetic markers to map the gaping lids trait. SSLPs were typed using PCR with mouse "MapPairs"™ primers obtained from Research Genetics Inc. (Huntsville, Alabama, USA). Each PCR reaction was carried out in a 25 ui volume overlaid with mineral oil in a 650 \xl reaction tube. Each reaction contained 100-200 ng of target DNA and 0.14 uM of each (forward and reverse) primer. The rest of the reaction mixture was provided by a "master mix", consisting of dATP, dGTP, dCTP, dTTP (final concentration, 50 uM each, Pharmacia

Biotech), Taq DNA polymerase (0.625 U per reaction, Gibco-BRL, Cat. No. 18038-018),

10 X PCR buffer (final concentration, 10 mM Tris-HCl, pH 8.3, 50 mM KC1; Gibco-Brl,

Part No. Y02028) and magnesium chloride (final concentration usually 1.5 mM, range

1.5-3.5 mM; Gibco-BRL, Part No. Y02016). PCR was performed in a Perkin-Elmer

4600 thermocycler, usually under the following conditions: 4.5 minutes at 94°C

(denaturation), followed by 30 cycles of 1 minute at 94°C (denaturation), 1 minute at

55°C (annealing) and 1 minute at 72°C (extension), followed by 7 minutes at 72°C

(extension). For optimal amplification, some primers required different annealing temperatures, ranging from 50-60°C. Additionally, some markers required a "hot start", where the DNA and primers were denatured at 94°C for several minutes before adding

44 the Taq polymerase (contained in the master mix) and continuing as above. These exceptions are listed in Appendix E, F, and G.

c) Visualization of PCR products for SSLPs

The marker dye bromophenol blue-xylene cyanol FF (5 ul) was added to the PCR product and 10 ul of this mixture was then run electrophoretically on 4% "NuSieve 3:1"

(3 parts NuSieve agarose: 1 part SeaKem LE agarose; FMC Bioproducts) horizontal gels containing 0.5 |u,g/ml of ethidium bromide. Gels were run in 1 x TAE (Sambrook et al.,

1989) usually at 140 V (130 - 145V) for VA to 2V2 hours, then observed and photographed (Polaroid 667 film) over UV light (302 nm). The resolution of small DNA fragments (under 300 bp) by NuSieve gels approaches that of polyacrylamide gels, but the agarose gels are generally faster and easier to run. However, for one marker, these allele products ran too closely together for a difference to be detected on NuSieve gels so these products were run on a 6% polyacrylamide vertical sequencing gel followed by

silver staining (this SSLP marker is listed in Appendix G). The procedures involving polyacrylamide were generously performed by Helen McDonald in Dr. Carolyn Brown's laboratory, Department of Medical Genetics, UBC.

d) Scanning Electron Microscopy

Pregnant females were killed by C02 on days 16 and 18 of gestation. The uterus was removed and the fetuses were dissected out under cold (4°C) Sorensen's phosphate buffer and their state of eyelid closure was noted. The five day 18 GP/Bc, three day 16

GP/Bc and one day 18 normal were immersed in cold 2.5% glutaraldehyde in Sorensen's phosphate buffer and stored at 4°C. Between 4 and 12 hours later, the fetuses were

45 decapitated, heads were returned to cold fixative and stored at 4°C. Between 4 and 12 hours later the heads were transferred to phosphate buffer for trimming, then placed in fresh buffer for processing the following day. The specimens were rinsed twice in buffer

(20 min. each), postfixed in 2% osmium tetroxide in phosphate buffer for 45 minutes, rinsed two times in buffer (10 min. each), washed once in water (10 min), fixed in 2% tannic acid (2% in water for 20 min.), washed twice in water (10 min. each) and postfixed in 2% osmium tetroxide for 30 minutes. The heads were then dehydrated in a graded series of ethanols (15 min. each) to 100% ethanol (3 x 20-25 min. each), during which

they returned to room temperature. They were critical point dried from liquid C02, mounted on stubs (one specimen per stub) using colloidal silver paste, and stored in a vacuum desiccator. The next day, they were sputter coated in vacuo with gold and examined within a few hours with a Cambridge Stereoscan 360 SEM (Cambridge

Instruments Ltd., Toronto, Canada), operated at lOkV. At least one photograph, a standardized view of the whole eye at approximately 100X, was taken of each specimen.

For this study, the crosses, collection of samples, preparation of samples, and fixing of samples were performed by M.J. Harris. The mounting of the specimens on the stubs was performed by myself. The processing for SEM was performed by Andre Wong in the

Faculty of Dentistry. I assisted Dr. M. J. Harris with the viewing and photography of the specimens.

B. Egfr study

a) DNA Preparation

46 All mice were killed by carbon dioxide (C02) gas or decapitation (newborns) before tissues were collected. Individual tail tip tissue was collected for DNA preparation by a method described by Drews, Drohan and Lubon (1994). The fresh or frozen tissue was mashed into 1 mm3 pieces using forceps in a microcentrifuge tube. 300ul of modified lysis buffer (50 mM Tris-HCl (pH 8), 10 mM EDTA (pH 8), 100 mM NaCl,

0.1% SDS) containing 30ul proteinase K was added to the tissue in a 1.5 ml Eppendorf tube. The tubes were placed in a 60°C water bath for 2-20 hours to digest the tissue.

Once completely digested the homogenate was briefly vortexed, then heated to 95-100°C for 10 minutes and a 1:20 to 1:50 dilution (in autoclaved deionized filtered water) was made for use in PCR. This modified method was used since it was reported to produce

DNA of sufficient quality for PCR allowing for rapid and effective screening of a large number of samples in a short period of time (S. Andrew, D. Mah personal communications).

b) PCR of SSLPs

Sequence for primer pairs directed at the null and wild type Egfr alleles was obtained from Dr. David Threadgill (Vanderbilt University, Nashville, USA), and primers were made at Nucleic Acid Protein Services (NAPS), UBC (Primer 1:5'-

GCCCTGCCTTTCCCACCATA-3'; Primer 2: 5'-TTGCAGCACATCCCCCTTTC-3';

Primer 3: 5'-ATCAACTTTGGGAGCCACAC-3', where Primers 1 and 2 amplify the null allele and Primers 1 and 3 amplify the wildtype allele). Each PCR reaction was carried out and contained the same reagents as described for the GP/Bc study. PCR was performed in a Perkin-Elmer 4600 thermocycler, initially under the following conditions

47 (modified from methods given by D. Threadgill): 5 minutes at 96°C (denaturation), followed by 35 cycles of 20 seconds at 96°C (denaturation), 30 seconds at 65°C

(annealing and extension), with no distinct extension step (i.e. 72°C) and 7 minutes at

72°C (final extension). These primers also required a "hot start", as previously described.

It was later determined that the standard PCR conditions described previously for the

GP/Bc study (4.5 minutes at 94°C (denaturation), followed by 30 cycles of 1 minute at

94°C (denaturation), 1 minute at 55°C (annealing) and 1 minute at 72°C (extension), followed by 7 minutes at 72°C (extension)), gave more consistent results, and did not require a hot start. These standard conditions were, therefore, used for the remainder of

PCR reactions with the Egfr primer pairs.

PCR of SSLP markers with primers obtained from Research Genetics Inc. also used in this study followed the conditions previously described in for the GP/Bc study.

The SSLP markers and conditions used in this study are listed in Appendix H.

c) Visualization of PCR products

Visualization of all PCR products followed the procedure previously described for the GP/Bc study.

48 CHAPTER III: PHENOTYPIC INVESTIGATIONS

I. Introduction

The goal of this part of the study was to characterize the phenotype of the gaping lids mutation which would thereby allow for comparison to other open eyelid mutations.

This was accomplished by observing the gross phenotype of gaping lids and normal mice and the specific phenotype of the surface of the fetal eyelids using scanning electron microscopy.

II. Rationale, Materials and Approach

a) Gross observation of phenotype

Rudimentary autopsies were performed on two each of GP/Bc (F - 7 mo; M - 9 mo.), CBA/J (2F - 11 mo.) and (GP/Bc x CBA/J)F1 ( F - 7 mo.; M - 7 mo.) mice to determine if there were other defects associated with the gaping lids mutation aside from the eyes. These autopsies consisted of gross examinations of thoracic and abdominal organs with regard to placement and size in phenotypically gaping lids animals in comparison to normal and first intercross animals. Further observations of two each of

GP/Bc (2 M - 5 mo.) and AXB-23/Pgn (2M - 5 mo.) mice were performed at a later time to compare the eyes of affected and normal mice in finer detail before and after death.

External examinations of these mice included close visual inspection of the eyes and measurements of the palpebral opening from canthus to canthus using a General MG

Ultratest Electronic Digital Caliper (No. MG8206). The belly area of the coat was also examined in the GP/Bc animals for the presence of a white belly spot which is known to appear commonly in this strain (D.M. Juriloff and M.J. Harris, personal

49 communications). Eyes were also removed and measurements at approximately the middle were taken using the digital caliper.

b) Scanning electron microscopy (SEM)

Scanning electron microscope photographs were taken of five dl8 GP/Bc, three dl6 GP/Bc and one dl8 normal F2 fetuses from the GP/Bc x CBA/J cross to find out what gaping lids eyelids look like at the time that eyelids grow across the eye and fuse in normal strains. The general SEM procedures have been outlined in Chapter II. At least one photograph, a standardized view of the whole eye at approximately 100X, was taken of each specimen. Higher magnification pictures were also taken of the inner and outer canthi of some of the specimens to enable further observation of the gaping lids phenotype.

III. Results

a) Gross observation of phenotype

As two sets of GP/Bc mice were examined at different times, they will be referred to as GP/Bc (1) and GP/Bc (2), where (1) indicates the first set and (2) indicates the

second set for this section of my thesis.

Rudimentary autopsies revealed no noticeable differences between GP/Bc (1),

CBA/J (normal) and (GP/Bc x CBA/J) Fl regarding the placement and size of internal organs. Neither of the first two GP/Bc mice examined had a white belly spot, whereas both of the second two GP/Bc mice did. Additionally, approximately the last half of the tail (to tip) was lacking pigmented hairs in all the GP/Bc mice examined. GP/Bc mice are

50 also known to have white knuckles on their paws (D.M. Juriloff and M.J. Harris, personal communications), but this was not noted in my examinations.

Examination of GP/Bc (2) mice prior to death found that they had "milky" corneas, the severity differing between the eyes and the mice. There was also an exudate present around the rim of the eyes. When a stimulus was presented (a puff of air) the

GP/Bc mice did have a blinking response, but it appeared that the "eyelids" did not come all the way together like those of the AXB-23/Pgn mice when they were presented with the same stimulus.

Close examination of the eye region of the GP/Bc (2) mice post mortem revealed that these mice do have upper and lower eyelids, as speculated above with the blinking response, as well as upper and lower eyelashes. However, compared to the AXB-23/Pgn mice, the eyelids appear to be shorter, i.e. they do not extend across the eyes as far, and the eyelashes seem to grow down from the eyelids over the eye instead of out from the eyelids. The corneas of the four eyes of the GP/Bc (2) mice all had varying degrees of bumpiness across the area that had been exposed between the eyelids, presumably giving the corneas the milky appearance seen before death. The lenses of the GP/Bc (2) and

AXB-23/Pgn mice were also removed and examined. The consistency, appearance, coloration and size of the GP/Bc (2) lenses appeared to be the same as those of the AXB-

23/Pgn when examined by light microscopy (dissection microscope).

The measurements of the eyes and palpebral opening indicated that the GP/Bc mice did not have distinctly smaller eyes when compared to the two normal strains and

(GP/Bc x CBA/J) Fl animals (see Tables 5a and 5b). Taking these two tables together, it appears that the hypothesis that the average eye size in GP/Bc may be smaller than

51 eu eu .a w> t» a oo in r~- o > r~ io o m w a © ^ rn on (N o\ CN ^ cn m rn rn rn cri (N cn cn cn 2 .a s- cu

(N t o en •-H --H rn >n cn rn rn rn o cn 7j- n CN W W ffl rn ri o- oo f 05 CN cn cn rn vo CN t ^ q h rn S3 U m Tt- m m TJ- o\ r--- oo IXI 4» 64 K CN a vo — in m m r-- 1*1 ; -rt; © vq r~; m I ©i1 m ^ rn m >t CN 0\ r-. O 3 © VO •S cn CN cn cn ca a a 2 2 6 6 6 6 Seu- ^ ^ ^ ^ ^ ^ o o o o cu S a a a a &4 3 CA in in m in a S3 eu ca a a oo £j m vo r-~ 00 m ca •«fr >n O N n m H O »n m u eu fl IS 02 c 22^ « 8 a T3 J3 ccS cu X! I-a a, a,

cn cn u CN CN § 2 05 PQ ffl a a i i l<8 S2 ^ m a a U U 60 < H-l 1 m « II o > u o o a; PL, 3< 3' • x s < o a < I— CN

52 the average eye size in both normal strains and the (GP/Bc x CBA/J) Fls, cannot be rejected. However, it should be noted that the measurements of the eye are subjective since they are dependent on the exact placement of the caliper, i.e. if it was not properly centered it would give different measurements for eyes from the same animal. A larger sample matched for age, sex and genetic background would be needed to obtain a definitive answer.

b) Scanning electron microscopy

Based on the observations of wide open eyes of five dl6 and three dl8 GP/Bc fetuses, no fusion between the eyelids of these mice has begun at either canthus, although both the inner and outer canthi are distinguishable. In these mice the eyelids have not advanced across the eye. There is the presence of rounded cells at the edges of the lids with large clumps at both canthi (see Figure la). Higher magnification of the inner canthus shows clumps of cells, of which a few appear to have flattened (see Figure lb).

Additionally, it is unclear if the flattened, step-like appearance of the cornea present in these SEM photographs is part of the gaping lids phenotype or if it is an artifact of the

SEM processing.

53 Figure 1(a): Scanning electron microscope picture of dl6 GP/Bc fetal eye; bar represents 500 fjm; ic = inner canthus, oc = outer canthus CHAPTER IV: MAPPING GAPING LIDS

I. Introduction

The goal of this part of the study was to map the location of the gaping lids mutation in the mouse genome and to test hypotheses that the gaping lids mutation is at one of various candidate loci. The long term aim is to identify the gene responsible for the gaping lids phenotype and its role in eyelid development. GP/Bc, homozygous for the gaping lids mutation, is one of several strains having the open eyelid defect with various degrees of genetic complexity maintained in the Juriloff/Harris lab as models of genetically complex developmental threshold traits. It is hypothesized that gaping lids may identify a locus that is also involved in other more complex open eyelid strains.

II. Rationale, Materials and Approach

A. Experimental design

a) Genetic crosses

As described previously, the genetic crosses were done prior to my involvement with the project. They were as follows. Homozygous gp/gp (GP/Bc) animals were outcrossed to the normal inbred strain CBA/J to generate Fl individuals. The Fl animals were crossed back to the affected parental strain to generate first backcross (BC1) animals, and intercrossed to generate the F2. A first backcross was initially made because, for most open eyelid mutants, the penetrance is higher in the backcross generation than in the F2 generations (see for example, Juriloff et al., 1983, 1996). F2 animals were created from the GP/Bc x CBA/J cross when it became clear that the BC1 penetrance was high.

55 Subsequently a second cross was made. Homozygous gplgp (GP/Bc) animals were outcrossed to the normal inbred ICR/Be strain to generate Fl individuals. These animals were then intercrossed to generate the F2 generation.

Mapping of the gaping lids locus was carried out in three stages. The first was to look for linkage to markers near candidate genes not previously excluded. This phase was done using pooled DNA from open eyelid BC1 samples from the CBA/J cross typed for SSLPs. Once linkage was indicated, the region was narrowed down by typing more closely spaced SSLP markers in individual gaping lids F2 animals from the CBA/J cross.

The F2 was used instead of the BC1 because each F2 animal represents two meiotic opportunities for recombination, whereas only one gamete leading to the BC1 contains this mapping information. The third stage was to further refine the map position of the gaping lids mutation by typing individual gaping lids F2s from the outcross to ICR/Be.

b) Screening for linkage to candidate regions

Mapping proceeded with the BC1 progeny initially using PCR amplification of

SSLPs at candidate regions of the genome that had not been tested by previous linkage studies (see Chapter I). Linkage to the gaping lids locus was investigated based on the

Mendelian principle that the BC1 to GP/Bc would have a 50:50 mix of homozygotes or heterozygotes at informative SSLP loci unlinked to the gaping lids locus, and therefore an allele ratio of 3 (GP/Bc): 1 (CBA/J) in the pooled BC1 animals, in contrast an SSLP locus that is physically very close to the gaping lids locus would be homozygous for the

GP/Bc allele in all, or nearly all, affected individuals, and therefore nearly all of the alleles would be GP/Bc in a pooled sample of BC1 animals.

56 The candidate regions initially tested in this study were chosen on the basis of containing candidate genes that are probably expressed in developing eyelids or that cause the open eyelids defect when mutated. The loci that were considered to be the best candidates were the lidgap Gates (lgGa) open eyelid mutation on distal chromosome 13 near integrin alpha2 (Itgal), the transforming growth factor alpha/waved-1 (Tgfalwal) locus on mid-chromosome 6, and the epidermal growth factor receptor/waved-2

(Egfrlwa2) locus on proximal chromosome 11. The general strategy was to use an informative marker that appeared to be near each candidate locus, according to the 1997

Chromosome Committee maps (Montgomery et al., 1997) plus flanking markers approximately 10 cM proximal and distal, respectively. Informative markers were identified by looking for discernible differences between GP/Bc, CBA/J, and (GP/Bc x

CBA/J)F1 DNA in the size of amplification products from primer pairs for SSLPs. The markers initially screened in each of these regions for informativeness were: D13Mit69,

D13Mitl47, D13Mit76, D6Mitl6, D6Mitl02, D6Mitl49, Dl lMit20, Dl lMitl52 and

Dl lMit38 (see Appendix E,F, and G).

The pools were used to estimate for each locus the approximate genotype of each individual without doing every individual separately, thereby reducing the number of individual PCR reactions. Empirically our laboratory has determined that we can detect a

single allele of one type, e.g. one "b" allele, among up to seven alleles of an alternate type, e.g. seven "a" alleles, in pooled samples, so the maximum number of individuals per pool is limited to four. For my study the maximum number per pool was three for simplicity during collection of tissue.

57 My work on this cross began with the banked frozen liver tissue. Liver samples from 23 bilaterally open eyelid BC1 individuals were combined in groups of two or three to create nine pools of DNA for screening candidate regions for the gaping lids locus.

There were four pools of two and five pools of three. Appendix I lists the pool sample identification numbers and the individual mouse identification numbers in each pool.

These pooled DNA samples underwent PCR amplification of SSLPs to type candidate regions for linkage with the gaping lids mutation. Analysis of the marker allele information for linkage involved several approaches. The first was to estimate at each marker locus the total number of GP/Bc and CBA/J alleles in the pooled samples, looking for significant deviation from the expected ratio of 3:1 alleles in the BC1. The percentage of GP/Bc alleles in each pool was estimated by comparing the intensity of the two bands on the gel (i.e. the allele band from GP/Bc and the allele band from CBA/J). For this procedure to produce useful results, it had to be ensured that the two bands of the control

Fl sample were equally amplified and were therefore of equal intensity on the gel. By comparing the intensity of the bands in the pooled samples to that of the Fl, an estimate of each number of alleles per pool was determined. For example, in a pool composed of two individuals, and therefore four alleles, if the intensity of the GP/Bc allele was greater than that of the CBA/J allele band, then the pool was scored as having three GP/Bc alleles and one CBA/J allele. For the overall sample of pools, when the percent of alleles was calculated it was taken into account that three quarters of the alleles would be from

GP/Bc due to the nature of the backcross.

c) GP/Bc x CBA/J - Refining the map position of gaping lids

58 Once linkage was detected using the pooled BC1 samples, DNA samples from 41 individual open eyelid F2s (GP/Bc x CBA/J) were used to construct recombination breakpoint maps around the gaping lids locus by determining the haplotypes of these animals at the informative SSLP markers in the region (see Appendix J). As with the

BC1 cross, informative markers in the region were determined by looking for allelic size differences in GP/Bc, CBA/J, and Fl DNA samples. All the markers screened are listed in Appendix F. As with the BC1 individuals, DNA was extracted from tail tissue and subjected to PCR amplification of SSLPs for mapping. Once the map region of the gaping lids locus was determined, 87 normal individuals (from 11 litters) were also screened at the closest SSLP markers (Dl lMit62 and Dl lMit226) to verify Mendelian segregation of marker alleles. This was done to ensure that the detected linkage was not an artifact due to segregation distortion or preferential amplification of the GP/Bc allele over the CBA/J allele. These animals and their open eyelid/pinhole litter mates are described in Appendix K.

A number of F2 individuals of ambiguous phenotype, identified as "pinholes" by the size of the gap between the lids, were not typed during the mapping phase. Four of these were subsequently typed for Dl lMit62, Dl lMit226, and Dl lMit80, the markers closest to the gaping lids locus. Three of these were chosen because all of their litter mates (open eyelid and normal) had already been typed at these markers, the fourth was typed to complete the sample analysis of "affected" progeny even though none of its normal litter mates had been collected (see Appendix J and K).

d) GP/Bc x ICR/Be - Refining the map position of gaping lids

59 To overcome a lack of informative markers and refine the map position of gaping lids, a second outcross F2 generation was created. This second outcross was to the

ICR/Be strain. It was chosen because it is a normal inbred strain maintained in the

Juriloff/Harris animal unit at UBC, and therefore it was readily available. Additionally, it was predicted to have different SSLP alleles from GP/Bc in the region surrounding and proximal to Dl lMit62/Dl lMit226. A screen of markers in the region identified nine that were informative between the GP/Bc and ICR/Be strains on NuSieve agarose and one that was informative on polyacrylamide. My work on the segregants from this cross began with the collection of the banked frozen tail tissue with Dr. M.J. Harris. The 23 gaping lids F2 animals were typed first for eight SSLP markers to generate a second recombination breakpoint map (see Appendix L). Typing of one marker, Dl lMcgl

(McGill University Chr 11 Marker 1; Primer 1: 5'-CTGGCTTGTTTGGGAACTCT-3';

Primer 2: 5'-CTCCCACAGCCGATTCTCAAT-3'; Claudio et al., 1994), was performed on identified recombinants only, as was typing of Dl lMit72, which required polyacrylamide gels.

58 normal individuals (from 6 litters) were also screened at Dl lMit74, the marker identified to be closest to the gaping lids locus, to verify Mendelian segregation of marker alleles for the same reasons as those of the CBA/J cross (see above). Thus, including their open eyelid litter mates who had already been typed at this marker, 6 complete litters were typed. Appendix M lists the details of these animals and their open eyelid litter mates.

B. Analysis of genetic transmission/penetrance

60 The clearly abnormal phenotype caused by the gaping lids mutation was known to segregate as a fully penetrant recessive mutation after some outcrosses (Kelton and

Smith, 1964; D.M. Juriloff and M.J. Harris, personal communications). However, in the outcrosses to normal strains in this project, a deficiency of phenotypically gaping lids animals was noted. This raised two hypotheses regarding the deficiency of open eyelid pups. The first was that gaping lids was showing reduced penetrance due to the presence of a modifier allele(s) introduced by the normal strain and that the modifier was suppressing the open eyelid phenotype in a proportion of the gplgp segregants, which resulted in a "normal" phenotype. The second was that the deficiency in open eyelid individuals was due to decreased viability in utero of gplgp progeny due to the introduction of a deleterious modifying factor(s) by the normal strain.

The first hypothesis was addressed in both outcrosses by screening a number of normal F2 progeny from complete litters (i.e. litters from which both the phenotypic affecteds and normals had been collected) at the informative SSLP marker found to be closest to the gaping lids locus to identify any phenotypically normal genetically gplgp homozygotes. The second question was not addressed in detail in this study but was partially addressed by looking for a deficiency of GP/Bc alleles at the closest SSLP markers in the F2 as a whole. This was done concurrently during the testing of the first hypothesis.

C. Molecular investigations

a) Chromosome 11 map refinement

There are three main sources from which current marker loci mapping information can be obtained. These are the Mouse Genome Informatics Database (MGI), Whitehead

61 Institute/ Massachusetts Institute of Technology (MIT), and the European Collaborative

Interspecific Backcross (EUCIB). The map produced from MIT is based on 92 meioses, which limits the resolution of the markers placed on the map to 1.1 cM. Many of these markers are therefore placed together at the same location, even though in reality they lie at different map positions. Many of the markers used to map gaping lids in one or both crosses are placed in such groups on the MIT map. These markers are detailed in Figure

2. To refine the locations of these markers with respect to each other in the gaping lids region, the other two main maps were reviewed. A comparison of the relative positions of these SSLP loci on the MGI map, EUCIB (BSB) map and the MIT map are illustrated in Figure 3 and detailed in Table 6. These three maps have been discussed in detail in

Chapter I.

CBA/J cross ICR/Be cross

74

149*, 62, 226 62, 72,71,226 2

162 80 80 152 151 19

Figure 2: Comparison of informative SSLP marker loci locations used in GP/Bc x CBA/J and ICR/Be crosses, where marker identification numbers follow the format DllMit##. These markers are at: 0.0, 2.2, 4.4, 6.6, 8.6, 9.8, and 10.9 cM, respectively as reported by MIT.

this marker has been mapped to Chr 9 (see Chapter IV)

62 MGI MIT EUCIB

D11MU74

DllMcgl (Nf2) DllMit72

DllMit71

DllMit62 DllMit226

DllMit2

DllMitl62 DllMit80 DllMit340

DllMitl51 DllMitl52

DllMitl9

Figure 3: Comparison of SSLP marker loci map locations used in GP/Bc x CBA/J and GP/Bc x ICR/Be crosses on MGI, MIT and EUCIB (BSB) maps

63 Table 6: Comparison of SSLP marker loci map position between the Mouse Genome Informatics Database (MGI), Massachusetts Institute of Technology/Research Genetics (MIT) and the European Collaborative Interspecific Backcross panel (EUCIB-BSB) - position is indicated in cM. Note: — indicates the marker was not mapped in this panel

Marker MGI MIT EUCIB DllMit74 0.0 2.2 — DllMcgl (Nf2) 0.25 ~ — DllMit72 0.25 2.2 0.0 DllMit71 1.1 0.0 — DllMit62 1.5 2.2 — DllMit226 1.55 2.2 0.9 DllMit2 2.4 4.4 3.9 DllMitl62 8.0 6.6 -- DllMit80 10.0 8.7 — DllMit340 11.0 — -- DllMitl51 13.0 9.8 10.4 DllMitl52 13.0 9.8 7.4 DllMitl9 14.0 10.9 8.4

Since these maps gave contradictory locations for some of the marker loci in the

gaping lids region, the composite genotypes of the F2 animals from both outcrosses in the

GP/Bc study were used to attempt to refine the map for the proximal end of Chr 11 (i.e.

determine the order of the markers). The genotypes of the 41 open eyelid, 4 pinhole and

87 normal F2 animals from the GP/Bc x CBA/J cross, and 23 open eyelid F2 from the

GP/Bc x ICR/Be cross were compared with these three maps (genotypes for these mice

are detailed in Appendix J, K, and L).

b) Haplotype analysis

Since the gaping lids mutation arose in the "B6" strain of mice (C57BL/6-ax;

Kelton and Smith, 1964), it was of interest to determine if the GP/Bc strain was "B6-like"

in the gaping lids region, and how far down chromosome 11 this haplotype might extend.

64 Therefore, a number of primers for SSLP loci that were used in the mapping of gaping lids and Egfr, plus a few in the mid- and distal regions of Chr 11, were typed in GP/Bc,

C57BL/6J and (GP/Bc x C57BL/6J) "Mock Fl" DNA (a mix of equal volumes of equally concentrated DNA from each parental strain, since a cross between these two strains was not conducted). This was done to determine if allelic size differences could be identified between these strains by PCR. The absence or presence of SSLP size differences is an indication that GP/Bc is or is not B6-like in the region.

III. Results

A. Segregation studies

a) Analysis of segregation after cross to CBA/J

As expected for a Mendelian recessive, the Fl generation contained no (0) open eyelid progeny among the 43 newborns from 6 litters. Open eyelids were observed in the first backcross to GP/Bc and in the F2. The reciprocal backcross data indicate that the genotype of the dam (i.e. GP/Bc dam or Fl dam) makes a significant (%2 = 7.86, p<0.005, df = 1) difference in the observed number of open eyelid pups. The cross of GP/Bc dams to (GP/Bc x CBA/J)F1 sires, gave an open eyelid frequency (35%, 19/55) significantly lower than the 50% expected (%2 = 5.26, pO.Ol, df = 1). The reciprocal cross of GP/Bc sires to (GP/Bc x CBA/J)F1 dams gave an open eyelid frequency (60%, 48/80) not significantly higher than the 50% expected (%2 = 3.20, p>0.10, df = 1). The F2 generation produced six animals of ambiguous phenotype, identified as "pinholes" due to the size of the gap between the lids. These animals were initially excluded from any analysis

65 because it was not known if they were truly gplgp. When they are excluded, the observed frequency of open eyelids in the F2 generation is approximately 16.5% (41/249), which is

significantly lower than the 25% expected for a simple Mendelian recessive (%2 = 8.499, p<0.001, df = 1). However, if the pinholes are included in the analysis, the frequency of

affected progeny is 19% (47/249), which is not statistically different (%2 = 0.0025,

p>0.25, df = 1) from the 25% homozygous affected progeny expected of a monogenic

trait. With the open eyelid frequency being 19% when the pinhole class of affected F2

individuals are included in the analysis, these F2 data also fit the hypothesis of a single

recessive causative locus (gp) plus a single modifier locus (recessive suppressor allele

from CBA/J), which gives an expected frequency of 19%. After the gaping lids locus

was mapped this hypothesis was addressed using SSLP markers linked to the gaping lids

locus (see below page 85-86) in this cross by screening for suppressed gplgp mice in the

normal F2 newborns. The sexes were statistically equally distributed in both the normal

and affected progeny in the BC1 and F2 generations (BC1 to GP/Bc dam: normal

17F:19M (x2 = 0.111, p>0.75, df = 1), affected 9F:10M (%2 = 0.053, p>0.90, df = 1); BC1

to GP/Bc sire: normal 18F: 14 (x2 = 0.50, p>0.50, df = 1), affected 21F:27M (%2 = 0.75,

p>0.50, df = 1); F2: normal 109F:93M (x2 = 1.267, p>0.25, df - 1), affected 24F:23M (x2

= 0.0213, p>0.90, df = 1)), where the F2 affected data includes the pinholes (4F:2M).

b) Analysis of segregation after cross to ICR/Be

As expected for a Mendelian recessive, the Fl generation contained no (0) open

eyelid progeny among the 32 newborns from four litters. The F2 generation contained

six animals identified as "pinholes". Due to the results from the CBA/J cross, these

66 animals were presumed to be affected (gplgp). However, if they are excluded from the affected class, the observed frequency of open eyelids in the F2 generation is approximately 9% (25/282) which is significantly lower than the 25% expected for a simple Mendelian recessive (%2 = 39.15, p<0.001, df = 1). Even when they are included in the analysis, the observed open eyelid frequency (11%, 31/282) is significantly lower than expected (%2 = 29.51, pO.OOl, df = 1). These F2 data also indicate the possibility of modifier loci introduced by the normal strain suppressing the open eyelid phenotype of some affected mice. This hypothesis was addressed, as in the first cross, (see below) by screening for gp/gp mice in the normal F2s. In addition, the sexes were statistically equally distributed in both the normal and affected progeny in the F2 generation (normal

130 F: 109 M (%2 = 1.845, p>0.25; data do not include 12 dl8 fetuses that were collected for mapping study), affected 16 F:15 M (%2 = 0.032, p>0.90)).

B. Mapping studies

/. GP/Bc x CBA/J crosses

a) BC1- pooled sample screen of candidate regions

Using the BC1 pools, I found evidence of linkage to Dl 1 Mit 152 on proximal Chr

11 in my screen of the first 3 candidate regions. At this point, as I had no data for the Chr

13 and Chr 6 regions, my next steps focused on the Chr 11 region as follows. At

Dl lMitl52, the pools appeared to be enriched for the GP/Bc strain SSLP allele

(approximately 87%, 20 of 23 GP/Bc alleles). To determine if the gaping lids locus lay proximal or distal to Dl lMitl52, a marker 30 cM distal, Dl lMit38, was then typed on the panel of pooled samples. This SSLP marker showed a lower proportion of GP alleles

67 (approximately 65%, 15 of 23 alleles) indicating that the gaping lids locus, if linked, was probably proximal to Dl 1 Mit 152. Once the probable general location of the gaping lids locus was identified in this way, the screening moved from the BC1 pools to individual

F2s because they had become available in large number, but more importantly because they are more efficient for gene mapping as each F2 represents two informative meioses, whereas each BC1 represents only one.

b) F2 - genotypic analysis of individuals

The chromosomal locations of the informative markers used to determine the composite genotypes of the open eyelid F2s are outlined in Figure 5. The genotypes and probable haplotypes seen at the six informative markers in the proximal region of chromosome 11 are shown in Figure 4 and detailed in Appendix J. Dl lMit62 and

Dl lMit226 were most highly associated with gaping lids (37/41 homozygous gg; %2 =

93.1, p<0.0005, df = 1). Pictures of representative agarose gels of Dl lMit62 and

Dl lMit226 are shown in Figures 6 and 7. None of the available markers proximal to

Dl lMit62 and Dl lMit226 that had been typed in this outcross were informative. This left the proximal breakpoint undefined in an approximately 1.5 cM region adjacent to the centromere. The distal breakpoint indicated that the gaping lids locus lay proximal to

Dl lMit80, in an approximately 10 cM region. However, the closest distance to the gaping lids locus that could be measured was 1.2 cM, due to the sample size of 41 open eyelid individuals (n = 82 meioses), where 1 recombination event in 82 meioses is equivalent to a recombination frequency of 1.2% or 1.2 cM (if the pinhole F2 data is included these number of affecteds increases to 45, n = 90 meioses, which results in a

68 (a) DllMit62 • DllMit226 • (b) DllMit80 • DllMit340 • • DllMitl52 • • DllMitl9 • • •

# of Individuals 30 7 3 1* (n=41)

• = GP/Bc homozygous • = GP/Bc x CBA/J heterozygous • = CBA/J homozygous

Figure 4: Mapping matrix of open eyelid F2 mice in GP/Bc x CBA/J cross. Data do not include pinhole F2 mice. (a) gp proximal to these markers if three heterozygotes at D11MU62 and Dl 1MU226 are recombinants (b) gp proximal to DllMit80 if three heterozygotes at D11MU62 and Dllmit226 are not recombinants

* Mouse U2775-B

69 gp'(a) gp'(b) 1.53 DllMit62,DllMit226'. 1 region of Egfr2

10.0 __ DllMit80 11.0 __ DllMit340

13.0 __ DllMitl52 14.0 __ DllMitl9

Figure 5: Location of markers used in GP/Bc x CBA/J cross

1 Location of gaping gene based on recombination breakpoints (see Figure 4) (a) gp proximal to these markers if three heterozygotes at D11MU62 and D11MU226 are recombinants (b) gp proximal to Dl 1MU80 if three heterozygotes at Dl 1MU62 and Dl lmit226 are not recombinants 2 Location of Egfr based on in situ hybridization of DNA probe (part of human c-ErbB proto- oncogene) to banded metaphase chromosomes of a rat x mouse hybrid clone (Munke and Francke, 1987) 3 cM position of SSLP markers as reported by Mouse Genome Informatics, 1999 Chromosome Committee Report (MGI3, 1999) 4 D11MU226 is 0.05 cM distal to D11MU62 (MGI3, 1999)

70 u uVH -a 1 •a o xi o> O ~* CN cn CN Tfr m VO r^ooo>vomvor^ooos ?9 ^ TJ- in in vo vo t> £9 ^ 0>00<—i in VI in in in 3 o t— oo o> 1 o OH CQ m^'^-^-inininmm'— o o fcCNCNCNCNCNCNCNCNCNUUcSCNCNCNCNCNCNCNCNfefcU 8 2

VH VH u CU tu 3 •a *a + o H n m m vo t— oo o o m vo oo o £9 ^ vo vo vo vo vo eg ^ vo vo vo vo vo o * £9 ^ r- r- oo « - in in vi vi in OH CQ OH CQ m h h h 1 11 in m m in in o *-» J *J > w r r i § « — —' >~f) *-f> " ) r ^ i > ^ i ; ' h-> \ Sou (N CN CN CN

200

100

OH o * OH O »-> CQ rn o CN cn CN in vo I-* CN cn o r*> * ?9 ^ o CN cn cn ^ O o OH «-i r- m >—| ooooooooooooooooooooo i i—' CN CN o CN CN [i-CNCNCNCNCN(N(N(NCNCN'—1 bp 2 o PH CN § 2 CQ

200 Sou

Figure 6: Pictures of representative Nusieve agarose gels of SSLP marker Dl 1MU62 in Gp/Bc x CBA/J cross; GP/Bc, CBA/J, Fl and panel of open eyelid and pinhole* F2 animals; + open eyelid mouse which is homozygous for CBA/J alleles. Sizes of relevant bands of 100 bp ladder are indicated.

71 -a

OH O ^ o a, ON O ^ N fl (N m v© o-soo^mmminm 5 ir> m vo vo i> 00 0\ cn-^-'^'^-ininininin ~ % « 8 fe fe o

(D

OH OH -O O CN CO in vo r- oo ov 69 ^ £5^ VO VO vo VO VO o O CH VO vO vO vO vO CH CQ CQ CQ OrcrCH C^^^^inininin^rrQ ,—I r-< in in in in vi ^vivjvjvjvir^r^L^L-^CQ wo <"'rrHrn S o bp iiOUpHfciNNNNiNOUrSNNNiNfcfeOUCN m in in in >n ^ tS O U

200

100

tH UJ I -a I ft o ^ + OH *° CQ a: rn r-- r- oo o 1 bp o — OO 00 00 OO OO 00 oo oo oo o fe CQ 1 -nOUfeCNCNCNCNCNCN CN CN (N '—

UM <-*« »,-* »*» mt «»i aan

-ua 1 * -_°o ffl£ ^< +m *c n *f - * o o Q~ rn r-» r- *-« r- CN o s R r ^ ^ * ^ O U fe CN fe ^ i—lOOfeCNCNCNCN

200

Figure 7: Pictures of representative Nusieve agarose gels of SSLP marker Dl 1MU226 in GP/Bc x CBA/J cross; GP/Bc, CBA/J, Fl and panel of open eyelid and pinhole* F2 animals; + open eyelid mouse which is homozygous for CBA/J alleles. Sizes of relevant bands of 100 bp ladder are indicated.

72 recombination frequency of 1.1% or 1.1 cM). These results indicate that the gaping lids locus is between Dl lMit80 and the centromere, very close to SSLP markers Dl lMit62 andDHMit226.

The lack of informative markers in the GP/Bc x CBA/J cross proximal to

Dl lMit62 and Dl lMit226, and therefore a lack of a marker proximal to gp, means that it is not possible to tell if the 3 gaping lids individuals that are heterozygous at Dl lMit62 and Dl lMit226 are recombinants with gp or are affected heterozygotes (+/gp).

Similarly, the one gaping lids individual homozygous for CBA/J alleles at Dl lMit62 and

Dl lMit226 (#2775-B) could be a recombinant on both chromosomes (unlikely in such an apparently short distance) or +/gp (as above) or a phenocopy (a normal mouse displaying the "open eyelid" phenotype due to maternal or in utero environmental effects). For my calculations of distance between the gaping lids locus and markers, I am assuming that the 3 heterozygotes are recombinants between Dl lMit62, Dl lMit226 and gp, and the 1 mouse homozygous for CBA/J alleles is a phenocopy. Therefore my data indicate that there are 10 recombinants in 80 meioses between gp and Dl lMit80, which converts to a map distance of 12.5 cM (95% CI from 6.2-22 cM). My data also indicate that if the 3 heterozygous animals are truly recombinants (i.e. are gplgp mice) above Dl lMit62 and

Dl 1MM226, the distance to the gaping lids locus is 3.8 cM (95% CI from 0.8-10.6 cM).

These results are illustrated in Figure 8.

The 4 F2s scored as "pinhole" were genotyped at Dl lMit62, Dl lMit226 and

Dl lMit80 and found to be homozygous for the GP/Bc allele product across these three markers. This indicates that they are likely to have the same genotype as the open eyelid

73 gp(a)

• DllMit62, 226

«p(b)

8.5 cM

DllMit80, 340, 152, 19

Figure 8: Map location of gaping lids and distances between markers as determined in GP/Bc x CBA/J cross (a) gp proximal to these markers if three heterozygotes at D11MU62 and DUMU226 are recombinants (b) gp proximal to D11MU80 if three heterozygotes at D11MU62 and Dl lmit226 are not recombinants

74 animals. The genotypes and composite haplotypes of these four mice are detailed in

Figure 9 (see also Appendix J).

DllMit62 •

DllMit226 •

DllMit80 •

# Indiv's 4

(n=4)

• = GP/Bc homozygous

Figure 9: Mapping matrix of "pinhole " F2 mice from GP/Bc x CBA/J cross.

i) Test for Mendelian segregation

To test for Mendelian segregation of the GP/Bc and CBA/J alleles in this region, an additional 87 normal F2 individuals from 11 litters were genotyped at Dl lMit62,

Dl lMit226, and Dl lMitl52 (30 of these 87 (litters 5 through 8) were also genotyped at

Dl lMit340, Dl lMitl9 and Dl lMit80). The animals and their genotypes used for this analysis are summarized in Figure 10 (see also Appendix K). When the open eyelid and pinhole F2 litter mates are included (i.e. whole litters, n = 106 progeny), the expected number of individuals with each genotype predicted by Mendelian segregation at a single locus (i.e. 1:2:1) would be 26.5 homozygous for GP/Bc alleles: 53 heterozygous for

GP/Bc and CBA/J alleles: 26.5 homozygous for CBA/J alleles. At the markers most highly associated with gaping lids, Dl lMit62 and Dl lMit226, it was found that 22 were homozygous for GP/Bc alleles (including 18 open eyelid/pinhole litter mates), 58 were heterozygous for GP/Bc and CBA/J alleles, and 26 were homozygous for CBA/J alleles

75 • • • • • • • • • • • • • •

S .s

a, * OH PH CQ m a "S o o u £ cu • •» cu • • • SH ft • tl • rn • cu • • • • • • • • un

CN rn

VD CN CN o o\ VD CN o o rn »—H 1—1 +-> -(-» +-> - .-H "> % s T3 00 c II P Q Q Q Q Q

76 (including one mouse that was identified as affected when tissue was collected at birth).

When compared to the expected values, the observed values are not statistically different than what is expected for random segregation of alleles at this locus (%2 = 1.245, p>0.50,

df=l). These data are summarized in Tables 7a and 7b.

Table 7a: Data for GP/Bc x CBA/J F2 affected and normal progeny -11 complete litters.

Dam Sire No. No. No. No. open No. litters progeny normal eyelids pinhole Fl 23 Fl 24 2 18 14 4 0 Fl 25 Fl 26 2 19 15 3 1 Fl 18 Fl 19 1 9 7 1 1 Fl 33 Fl 34 1 11 10 1 0 Fl 31 Fl 32 2 19 17 1 1 Fl 29 Fl 30 2 23 19 4 0 Fl 16 Fl 17 1 7 5 2 0 Totals 11 106 87 16 3

Table 7b: Segregation at markers closest to gp, D11MU62 and D11MU226 in 11 complete litters described in Table 7a.

No. progeny No. each genotype at genotyped D11 Mit62 and 226. gg' gc cc 106 22 58 26 1 g = GP/Bc allele c = CBA/J allele

II) GP/Bc x ICR/Be cross

a) F2 - genotypic analysis of individuals

Based on the region identified by the cross to CBA/J, markers were screened for

informativeness in the proximal 15 cM of Chr 11. The chromosomal locations of the

nine markers found to be informative using NuSieve gels, and the one marker found to be

77 informative using polyacrylamide gels are outlined in Figure 12. They surrounded and included SSLP markers Dl lMit62 and Dl lMit226, the two most associated with the gaping lids mutation in the cross to CBA/J. The genotypes and probable haplotypes seen at the 10 informative markers in the gaping lids region are shown in Figure 11 (see also

Appendix L). Dl lMit74, Dl lMit72 and Dl lMcgl were the markers most highly associated with gaping lids (Dl lMit74: 23/23 homozygotes; %2 = 69, p<0.0005, df = 1;

Dl lMit72 and Dl lMcgl: 2/2 homozygotes; %2 = 4, p<0.025, df=1). A representative picture of an agarose gel of Dl lMit74 is shown in Figure 13. These results indicate that the gaping lids locus is between Dl lMit71 and the centromere (based on one recombinant). In this cross the closest distance to the gaping lids locus that can be measured is 2.2 cM, again due to the sample size (n = 23 individuals, 46 meioses). Two recombinant mice indicate that D11 Mit 162 is 4.4 cM (95% CI from 0.1-15 cM) distal to the gaping lids locus. One recombinant mouse indicates that Dl lMit71 is 2.2 cM (95%

CI from 0.06-11.5 cM)) distal to the gaping lids locus, placing gaping lids at the most proximal end of Chr 11 near the centromere (see Figure 14), highly associated with

Dl lMit72, Dl lMcgl and Dl lMit74.

i) Test for Mendelian segregation

To test for Mendelian segregation of the GP/Bc and ICR/Be alleles in this region,

58 normal F2 individuals from 6 litters were genotyped at Dl lMit74. The animals and their genotypes used for this analysis are outlined in Table 8 (see also Appendix M).

When the open eyelid F2 litter mates are included (i.e. whole litters, n = 65), the expected

78 DllMit74

DllMcgl

) DllMit72

DllMit71

DllMit62

DllMit226

») D11MU2

DllMitl62

DllMit80

DllMitl51

# Indiv's (n=23)

' = GP/Bc homozygous • = GP/ICR heterozygous

Figure 11: Mapping matrix of open eyelid F2 mice in GP/Bc x lCPJBc cross (a) gp is proximal to D11MU71, 62, 226,2 based on one recombinant mouse (b) gp is proximal to D11MU162, 80, 151 based on one recombinant mouse

+ Mouse # 3345-B * Mouse #3329-B

79 0.03 DllMit74 gp'(a) gp'(b) DllMit72, DllMcgl4 1.1 DllMit71 DllMit62,DllMit2265 2.4 DllMit2 region of Egfr2

8.0 DllMitl62

10.0 DllMit80

13.0 DllMitl51

Figure 12: Locations of markers used in GP/Bc x ICR/Be cross

1 Location of gaping gene based on recombination breakpoints (see Figure 11) (a) gp is proximal to Dl 1MU71, 62, 226,2 based on one recombinant mouse (b) gp is proximal to D11MU162, 80, 151 based on one recombinant mouse 2 Location of Egfr based on in situ hybridization of DNA probe (part of human c-ErbB proto- oncogene) to banded metaphase chromosomes of a rat x mouse hybrid clone (Munke and Francke, 1987) 3 cM position of SSLP markers as reported by Mouse Genome Informatics, 1999 Chromosome Committee Report (MGI3, 1999) 4 D11MU72 and DllMcgl are at 0.25 cM (MGI3, 1999) 5 D11MH62 is at 1.5 cM and Dl 1MU226 is at 1.55 cM (MGI3, 1999)

80 Figure 13: Pictures of representative Nusieve agarose gels of SSLP marker DllMit 74 in GP/Bc x ICPJBc cross; GP/Bc, ICR/Be, Fl and panel of open eyelid F2 animals. Sizes of relevant bands of 100 bp ladder are indicated.

81 • Dl lMit74, 72, Mcgl

gp(a) 2.2 cM

gp(b) DllMit71,62, 226,2

2.2 cM

DllMitl62, 80, 151

/ 1

Figure 14: Map location of gaping lids and distances between markers as determined in GP/Bc x ICR/Be cross (a) gp is proximal to DllMitH, 62, 226, 2 based on one recombinant mouse (b) gp is proximal to D11MU162, 62, 226, 2 based on one recombinant mouse

82 number of individuals with each genotype predicted by Mendelian segregation at a single locus (i.e. 1:2:1) would be 16.25 homozygous for GP/Bc alleles, 32.5 heterozygous for

Table 8: Segregation of alleles at D11MU74 in normal F2s in GP/Bc x ICR/Be cross (n= 58)

Genotype1 # of each

2 gg 14 gi 31 ii 13

1 g = GP/Bc allele I = ICR/Be allele 2 these are the 14 "normal gplgp" referred to in the text

GP/Bc and ICR/Be alleles, and 16.25 homozygous for ICR/Be alleles. At a marker perfectly associated with gaping lids, Dl lMit74, it was found that 21 were homozygous for GP/Bc alleles (including the 7 with open eyes), 31 were heterozygous for GP/Bc and ICR/Be alleles and 13 were homozygous for ICR/Be alleles. The observed values are not statistically different than what is expected for random

segregation of alleles at this marker (%2 = 2.13, p>0.25, df=l). The data are summarized in Tables 9a and 9b.

Table 9a: Data for GP/Bc x ICR/Be F2 affected and normal progeny - 6 complete litters.

Dam Sire No. Litters No. No. No. open No. progeny normal eyelids pinhole Fl 101 Fl 102 1 12 10 2 0 Fl 103 Fl 104 2 22 18 4 0 Fl 105 Fl 106 1 8 8 0 0 Fl 107 Fl 108 2 23 22 1 0 Totals 6 65 58 7 0

83 Table 9b: Segregation at marker closest to gp, D11MU74 in 6 complete litters described in Table 9a.

No. progeny No. each genotype at genotyped DllMit74 gg1 gi ii 65 21 31 13

1 g = GP/Bc allele i = ICR/Be allele

C. Analysis of genetic transmission/penetrance after crosses to CBA/J and

ICR/Be

Segregation of marker alleles was used to investigate whether the significant deficiency of phenotypic gaping lids segregants was due to suppression of the mutant phenotype by modifiers introduced by the outcross. If so, genetic gp/gp mice are expected among the "normal" F2s. This method was also used to investigate whether there was a deficiency of GP/Bc alleles at the closest SSLP marker loci, indicating a possible viability problem in a proportion of gplgp progeny.

i) GP/Bc x CBA/J cross

In the GP/Bc x CBA/J cross, 87 normal F2s were screened at the informative

SSLP markers closest to gp, Dl lMit62 and Dl lMit226. Four of these segregants were found to be homozygous for GP/Bc alleles at these markers, outlined in Figure 10.

Thus, amongst the F2s which based on SSLP genotypes are presumed to be genetically gp/gp, obtained in litters where all individuals were typed, 19/23 were phenotypically affected (open eyelid or pinhole), indicating a penetrance of 82% in this cross and no support for a hypothesis of prenatal death of gp homozygotes was found. There was no

84 deficiency of GP/Bc alleles at these SSLP loci (see Table 8), supporting the hypothesis that modifier(s) suppressing penetrance were introduced by the outcross.

ii) GP/Bc x ICR/Be cross

In the GP/Bc x ICR/Be cross, 58 normal progeny were screened at SSLP marker

Dl lMit74. 14 of these progeny were found to be homozygous for GP/Bc alleles at this marker, summarized in Table 8. This number is slightly higher than expected (see above), but this may be due to sampling bias that occurred at the time of litter collection.

Not all litters were saved for DNA analysis. Since this cross produced few to zero (0) open eyelid animals per litter, the litters which did not contain any open eyelid pups were discarded more often. Since many of these were not available for analysis, it may have slightly skewed the data, although the values I obtained are not significant. Amongst the genetically gp/gp F2s obtained in litters where all individuals were typed, 7/21 were phenotypically gaping, indicating a penetrance of 33% in this cross. As in the CBA/J cross, there is no deficiency of GP/Bc alleles at this SSLP locus (see Table 9b), supporting the hypothesis that the deficiency of affected F2s is due to modifiers that suppress penetrance.

D. Molecular investigations

a) Chromosome 11 map refinement

Based on comparisons between the MGI, MIT and EUCIB maps and the data generated from the two GP/Bc outcrosses, map positions of marker loci were confirmed, refined and refuted. The maps generated by data from the two outcrosses are detailed in

Figures 5 and 12. These are compared to the MGI, MIT and EUCIB maps in Figure 15.

In the cross to CBA/J, the location of a previously syntenic marker on the MGI map,

85

Dl lMit340, was identified (July, 1997), placing it between Dl lMit80 and Dl lMitl52 based on the genotypes of the gplgp and normal F2. This position is confirmed by the mapping study at the Whitehead Institute/MIT (MGI3,1999). Additionally, in both outcrosses, the markers Dl lMit62 and Dl lMit226 did not recombine with each other, which did not allow for their fine mapping in either of these studies (see Figures 4 and

11). These two markers were found to be 0.05 cM apart in a large scale mapping project at the Whitehead Institute/MIT (Dl lMit226 is 0.05 cM distal to Dl lMit62; MGI3,

1999).

The locations of the markers on the CBA/J and ICR/Be maps agree with most of those on the three reference maps except for the following. Dl lMit74 and Dl lMit71 recombined in the ICR/Be cross (see Figure 11), supporting the map positions reported by MGI, (ICR/Be cross: Dl IMit 74 is 2.2 cM proximal to Dl lMit71; 1 recombinant in

46 meioses, 95% CI = from 0.06-11.5 cM; MGI: Dl lMit74 is 1.1 cM proximal to

Dl lMit71), refuting the positions reported by MIT. The EUCIB panel did not map either of these SSLPs.

b) Exploration of the strain origin of the haplotype surrounding gaping lids

To determine whether the haplotype surrounding gp is of C57BL/6 origin GP/Bc primer pairs amplifying SSLPs used in the mapping of gaping lids and Egfr, plus a few in the mid- and distal regions of Chr 11, were used. Of the 17 SSLP loci tested, one was unreadable, and two were not C57BL/6-like. These primer pairs are listed in Table 10.

Their relative locations are detailed in Figures 5 and 12 (see also Appendix F and G for

optimization information). The non-C57BL/6-like region is approximately 6 cM, between Dl lMit72 and Dl IMit 162 (MGI3, 1999). This region is distal to the location of

87 gaping lids mapped in the two crosses discussed previously but is within the region to which the Egfr locus was mapped in the third cross (see Chapter V). The observed allele sizes in the GP/Bc DNA are extremely close to those reported for the C57BL/6J (B6) strain by the Whitehead Institute/MIT. The observed and expected allele sizes are also reported in Table 10. This indicates that the region closely surrounding gp is compatible with the history that the gp mutation occurred on a C57BL/6 background.

88 Table 10: Haplotype analysis of GP/Bc versus C57BL/6J DNA (conducted using DNA from one GP/Bc individual (3060-B); plus C57BL/6J +/+ (97-04-027n) and GP/Bc x B6 Mock Fl DNA as controls).

Marker cM postion1 Is GP/Bc B6- Observed allele Expected allele like (Y/N)? size (bp) size (bp)3 GP/Bc B6 B6 DllMit74 0.0 Y = 210 = 214 DllMcgl 0.25 Y = 290 = 288 DllMit71 1.1 Y = 210 = 214 DllMit62 1.5 Y 145 148 DllMit226 1.55 Y 140 142 DllMit77 2.0 N 160 150 152 DllMit24 2.4 ? ? 120 ? 122 DllMitl62 8.0 N 150 130 123 Egfrwt5 9.0 Y = 350 = 350 DllMit80 10.0 Y = 180 = 172 DllMit340 11.0 Y = 140 = n/a6 DllMitl51 13.0 Y = 140 142 DllMitl52 13.0 Y = 140 137 DllMitl9 14.0 Y = 140 = 140 DllMit20 20.0 Y = 130 = 116 DllMit38 49.0 Y 110 = 76 DllMitlO 63.0 Y 100 = 100

1 cM position as reported in 1999 Chromosome Committee report (MGI2,1998) 2 Allele size is rounded to the closest 5bp 3 Expected allele sizes are those reported by Research Genetics 4 This marker gave unreadable results, where allele size could be roughly determined, but informativeness between these two strains could not 5 These primers were directed at the Egfr wildtype (wt) allele; the cM location is that of the Egfr locus reported by MGI (MGI2,1998); the expected allele size (bp) is that reported by D. Threadgill 6 This SSLP locus was not typed in the panel of standard strains by Research Genetics so the expected allele size is unavailable 7 DNA supplied by Diana Mah

89 CHAPTER V: MAPPING EGFR

I. Introduction

The goal of this part of the study was to map the location of the epidermal growth factor receptor (Egfr) gene against SSLPs that had been used in the mapping of the gaping lids locus with the aim of demonstrating that they map to different segments of proximal Chr 11 and are therefore separate loci. This was done to confirm at a molecular level the complementation, and therefore lack of allelism, between gaping lids and Egfr determined in a cross between GP/Bc and Egfr+I~ mice (all progeny had normal hair and vibrissae and closed eyes at birth; M.J. Harris, personal communications). Additionally, this was done to narrow down the map location of Egfr. Previous studies have not mapped Egfr against sets of SSLPs but have instead used the visible phenotype of the

Egfr mutation and/or RFLPs from Egfr detected by Southern blot to anchor their maps relative to other loci of interest. This study would thereby map the Egfr locus with relation to SSLPs, including those used to map gaping lids in the two GP/Bc crosses, and delineate its location on proximal Chr 11.

II. Rationale, Materials and Approach

A. Experimental design

To map the Egfr locus, mice carrying a null allele at the Egfr locus were utilized.

These mice were obtained from D. Threadgill (Vanderbilt University, Nashville, USA).

For my study, mice heterozygous for the null allele (Egfr+'~) induced in a 129/Sv-derived haplotype on a predominantly CD1 background (see Chapter I) were crossed with female

90 BXA-2/Pgn mice (see Chapter II). Male (Egfr~fBXA-2)F\ mice carrying the null allele

(the presence of the null allele was identified using PCR primers directed at the null and wild type alleles, see Chapter I) generated by this cross were then crossed to female

SWV/Bc mice in a "special testcross". SWV/Bc was chosen because it is an available inbred normal strain maintained by the Juriloff/Harris lab that had been determined to have detectable SSLP allelic differences from the Egfr"/CD-I or BXA-2/Pgn mice.

The mapping of Egfr was carried out in three phases. The first was to generate a large number of progeny from the (£,g/r7BXA-2)Fl x SWV/Bc cross. The second phase was to type these progeny at two informative markers that flanked the Egfr region. The third was to identify recombinants, determine their haplotypes at additional SSLP markers between the two flanking markers and type them for the presence or absence of the Egfr null allele. For this study, the crosses were set up by D.M. Juriloff and M.J.

Harris, and the banking of tissue was done by me.

/. Egfr /BXA-2 x SWV/Bc cross

Pregnant SWV/Bc mice were killed by carbon dioxide (C02) gas between day 13 and day 16 of gestation. To collect embryos, the uterus was immediately removed and placed in a petri dish on a black wax background (parafin plus Sudan Black). Isotonic saline (0.85% NaCl) was added to the petri dish to aid in the examination of the embryos.

The uterus was secured with pins, and the isotonic saline (0.85% NaCl) was changed to remove the presence of maternal contamination. The uterus was then cut open under a binocular dissection microscope to reveal the chorionic sacs of the embryos. Embryos were collected into individual 1cm deep wells in a 3"x 5" porcelain dish under isotonic

91 saline (0.85% NaCl), rinsed and placed in individually prelabelled vials and stored at

-20°C. The only exceptions were one day 19 and one newborn litter, which were killed by decapitation and from which only paws and tail were subsequently collected and individually stored at -20°C. In total 109 progeny (82 embryos, 13 day 19 fetuses, and 14 newborns) were collected.

In order to type progeny generated from this cross, SSLP markers between the centromere and approximately 20 cM distal to the centromere, including markers used in both the GP/Bc x CBA/J and GP/Bc x ICR/Be crosses, were typed in the BXA-2,

SWV/Bc and (isg/r7BXA-2)Fl parental genotypes to identify those that would be informative in the resultant progeny of the special test cross. Based on the reported location of Egfr (MGI2, 1998), SSLP loci within the proximal 20 cM of Chr 11 should surround the Egfr locus. In this special cross, the alleles at informative SSLP loci in the

(Egfr~/BXA-2)F1 that were received from the Egfr~ parent will be called the "Egfr haplotype". The markers tested for informativeness are listed in Appendix H.

a) Genotype analysis of individuals

The 109 progeny were initially typed for a pair of informative SSLP markers that flanked the region containing the gaping lids locus and the reported location of Egfr in order to identify recombinant progeny. These SSLP markers were Dl lMit74, the most proximal SSLP marker on Chr 11 and near the gaping lids locus, and Dl lMitl51, an

SSLP marker that should lie distal to the Egfr locus (based on its reported location). All identified recombinants were then typed at SSLP marker loci between Dl lMit74 and

Dl lMitl51 and for the Egfr null and wild type alleles by allele specific primers

92 (described in Chapter II) to construct haplotypes for these mice. A picture of a representative agarose gel of the primers which amplify the Egfr null allele is shown in

Figure 16.

III. Results

A. Esfr7BXA-2 x SWV/Bc cross

a) Genotype analysis of individuals

The chromosomal locations of the informative markers examined in the 109

progeny are outlined in Figure 17. The genotypes and probable haplotypes of the 109

mice at Dl lMit74 and Dl lMitl51 and of the 3 recombinants at all 7 informative markers

are shown in Figure 19 and outlined in Appendix O. The three recombinants were also

typed for the presence of the Egfr null allele. It was found that 2 of the 3 recombinants

(KB-1016 and KB-1061) carried the null allele, whereas 1 (KB-1008) did not. The two

recombinants that carry the null allele also have the Egfr haplotype at the 6 proximal

markers indicating that a recombination event had taken place distal to the Egfr locus.

This means that Egfr does not lie distal to Dl lMitl51. However, the one recombinant

that does not carry the Egfr null allele has the Egfr haplotype at only the first 4 proximal

markers indicating that a recombination event had taken place proximal to the Egfr locus.

This means that Egfr does not lie proximal to Dl lMit226. This cross and the results are

outlined in Figure 18. These results place the Egfr locus between two SSLP markers,

Dl lMit226 (proximal) and Dl IMit 151 (distal), a distance of approximately 12 cM

according to the 1999 Chromosome committee report (MGI3,1999) and 2.8 cM (95% CI

93 bp 300 400 200 Figure 16:PictureofrepresentativeNusieveagarosegelofprimerpairswhichamplifythe Egfrnull(01) HO O ,—i u CH lad der

and wildtype(02)alleles;Egfrparent(EgfrP)FlsamplesfromEgfr~/Cd-1xBXA-2cross. luU Sizes ofrelevantbandsin100bpladderareindicated;Nullalleleis~450bp,wildtype 1 02 <+-( W is -350bp;*indicatesabsenceofnullallele,i.e.homozygosityforwildtypeallele. • 01 2 m i—( 01 2 — Note: nullalleleinEgfrPisveryfaintbutpresent. 01 2 — cn. i—H m mm. 01 2 i-H rn *— 1 " 94 01 2 mm. 01 2 * cn. 0 "•**• 01 2 cn. Os 01 2 mm* O m |

10.0 DllMit805

13.0 DllMitl51

Figure 17: Locations of markers used in (Egfr /BXA-2)F1 x SWV/Bc special testcross

1 cM position of SSLP markers as reported by Mouse Genome Informatics, 1999 Chromosome Committee Report (MGI3, 1999) 2 location of Egfr based on in situ hybridization of DNA probe (part of human c-ErbB proto- oncogene) banded to metaphase chromosomes of a rat x mouse hybrid clone (Munke and Francke, 1987) 3 Dl 1MU106 is at 0.25 cM as reported by the 1999 Chromosome Committee Report (MGI3, 1999) 4 D11MU62 is at 1.5 cMand D11MU226 is at 1.55 cMas reported by the 1999 Chromosome Committee Report (MGI3, 1999) 5 marker location for reference only (used in CBA cross not in Egfr cross)

95 Egfr BXA-2/Pgn SWV/Bc Egfr _ + BXA-2 - + + + + +

b b X Fl b b b b b

Egfr SWV/Bc SWV/Bc + + +

Non recombinants

Egfr SWV/Bc Egfr SWV/Bc + + +

c e e Recombinants e • b b

Figure 18: Outline of (Egfr~/BXA-2)F1 x SWV/Bc special testcross: red = EgfrSCD-l and Egfr haplotype, blue = BXA-2 and BXA-2 allele, green = SWV/Bc and SWV/Bc allele. Presence of Egfr null allele is indicated by a (-), presence of Egfr wildtype allele is indicated by a ( + ). Letters indicate SSLP loci at proximal end of Chr 11 around the Egfr locus (e = Egfr haplotype, b = BXA-2 allele, s = SWV/Bc allele). Arrows indicate crossover points in recombinant progeny.

96 Egfr locus

DllMit74 • • • •

DllMitl06 • • DllMit62 • • DllMit226 • •

DllMit77 • • location of DllMit78 • °• } Egfr DllMitl51 • • • •

# each genotype 55 51 2 1 (n=109)

• = presence of SSLP allele from Egfr~ haplotype (= absence of BXA-2 allele) • = absence of SSLP allele from Egfr~ haplotype (= presence of BXA-2 allele)

Figure 19: Haplotypes of Egfr and Egfr mice in (Egfr'/ BXA-2)Fl x SWV/Bc special testcross

*carry the Egfr knockout allele A does not carry the Egfr knockout allele

97 from 0.6-7.8 cM; 3/109) according to my data (see Figure 20). My data indicate that gaping lids is proximal to Dl lMit62 and Dl lMit226 (assuming the three heterozygotes are recombinants with these markers) in the cross to CBA/J and proximal to Dl lMit72, near Dl lMit74, in the ICR/Be cross, as compared to Egfr, which has been mapped distal to Dl lMit226. These data support the interpretation that gaping lids and Egfr are at different loci and are not alleles. These data are summarized in Figure 21.

98 DllMit226

0.9 cM

-2.8 cM (Egfr) DllMit77,DllMit78

1.8 cM

DllMitl51

Figure 20: Location of Egfr locus based on (Egfr /BXA)F1 x SWV/Bc special test cross

99 1 6 o.o :DllMit74,DllMitl06 gp 9 1.5 ; Dl lMit71 , Dl 1MU62, Dl lMit226 • 2.0 :DllMit77, DllMit78

10.0 DllMit808

13.0 DllMitl51

Figure 21: Locations of gaping lids locus and Egfr locus based on GP/Bc x CBA/J cross, GP/Bc x ICR/Be cross and (Egfr~/BXA-2)F1 x SWV/Bc special test cross.

1 cM position of SSLP markers as reported by Mouse Genome Informatics, 1999 Chromosome Committee Report (MGI3, 1999) 2 location of gp based on GP/Bc x CBA/Jcross 3 location of gp based on GP/Bc x ICR/Be cross 4 location of Egfr~ based on (Egfr~/BXA-2)Fl x SWV cross 5 location of Egfr based on in situ hybridization of DNA probe (part of human c-ErbB proto- oncogene) to banded metaphase chromosomes of a rat x mouse hybrid clone (Munke and Francke, 1987) 6 Dl 1MU106 is at 0.25 cM as reported by the 1999 Chromosome Committee Report (MGI3, 1999) 7 D11MU62 is at 1.5 cM and D11MU226 is at 1.55 cMas reported by the 1999 Chromosome Committee Report (MGI3, 1999) 8 marker location for reference only (used in CBA cross not in Egfr cross) 9 marker location for reference only (used in ICR cross not in Egfr cross; marker is a 1.1 cM (MGI3, 1999)

100 CHAPTER VI: CORRECTING THE MGI MAP

An interesting finding in the CBA/J outcross (see Chapter IV) is the gross mismapping of Dl lMitl49 in the public databases. According to MIT and MGI,

Dl lMitl49 is 2.2 cM and 1.0 cM, respectively, from the centromere (it was not mapped by EUCIB). However, when this SSLP marker was typed in the panel of 41 open eyelid and 4 pinhole F2s from the CBA/J cross, the genotypes indicated it was segregating independently (i.e. 1:2:1, (n = 41) f = 1.63, p<0.05, df = 1; (n = 45) %2 = 2.20, p<0.05, df

= 1) when it should have shown strong linkage to Dl lMit62 and Dl lMit226 based on its reported map location. The genotypes of these 45 mice at Dl lMitl49 are outlined in

Table 11 (see also Appendix N) and the putative haplotypes, including the other 6 informative SSLPs typed in these 45 mice in Figure 22a and 22b. Since these data indicated Dl IMit 149 did not map to the proximal end of Chr 11, the question of where this SSLP marker actually mapped was therefore addressed.

Table 11: Segregation of alleles at D11MU149 in open eyeliaVpinhole F2 mice from GP/Bc x CBA/J cross (n = 41,4 respectively).

Open eyelid Pinhole Genotype1 # of each Genotype ' # of each gg 10 gg 2 gc 24 gc 2 cc 7 cc 0

1 g = GP/Bc allele, c = CBA/J allele

It was first hypothesized that Dl IMit 149 may map to the distal, rather than proximal, end of Chr 11. To address this, linkage between Dl lMitl49 and Dl IMitIO, a

101 DllMitl49 • • • D11MU62 • • DllMit226 • DllMit80 • DllMit340 • D11MU152 • DllMitl9 • # of Individuals 18 1 (n=41)

• = GP/Bc homozygous • = GP/Bc x CBA/J heterozygous • = CBA/J homozygous

Figure 22(a): Mapping matrix of open eyelid F2 mice in GP/Bc x CBA/J cross. A comparison between D11MU149 and the six previously typed SSLP markers. Data does not include pinhole F2 mice.

D11MU149 • •

DllMit62 • •

DllMit226 • •

DllMit80 • •

# Indiv's 2 2

(n=4)

• = GP/Bc homozygous • = GP/BC x CBA/J heterozygous

Figure 22(b): Mapping matrix of "pinhole " F2 mice from GP/Bc x CBA/J cross. A comparison between DllMit 149 and the three previously typed SSLP markers.

102 confirmed distal SSLP, was tested in the panel of open eyelid/pinhole F2 mice. The genotypes of the mice indicated a lack of linkage between these two markers denoting that Dl lMitl49 does not map to the distal end of Chr 11 (see Figure 23a and 23b, and

Appendix N). This then led to the hypothesis that Dl IMit 149 did not actually map to

Chr 11. A review of the YAC panels used to place this marker on the SSLP physical maps revealed a pattern which hinted to the location of Dl IMit 149 being on Chr 9. Of the 6 YACs reported (MGI4, 1999) Dl lMitl49 is associated with a Chr 9 SSLP marker on three of them (on the remaining three it is the single SSLP on two and is associated with a Chr 2 SSLP on the third). A search of the MGI, MIT and EUCIB databases with the standard genetic search engines was unproductive. However a generic web search engine located Dl lMitl49 in the JAX-BSB mapping panel on Chr 9 (D.M. Juriloff, personal communications) between D9Mit23 and D9Mitl63, data which had not been incorporated into the MGI data. To confirm this location, linkage between Dl IMit 149 and an informative Chr 9 marker, D9Mitl91, was investigated by PCR amplification of the 41 open eyelid and 4 pinhole F2 mice at this SSLP marker. The genotypes of these mice at D9Mitl91 exactly matched those at Dl IMit 149, indicating strong linkage between these markers, thereby confirming the map position of Dl IMit 149 on Chr 9 (see

Figure 24a and 24b and Appendix N).

103 DllMitlO

D11MU149

# of Individuals 13 6 5 5 4 3 3 1 1 (n = 41)

• = GP/Bc homozygous B = GP/Bc x CBA/J heterozygous • = CBA/J homozygous

Figure 23(a): Mapping matrix of open eyelid F2 mice in GP/Bc x CBA/J cross. A comparison between D11MU149 and DllMitlO. Data does not include pinhole F2 mice.

DllMitlO • U • U

DllMitl49 • E3 • •

# of Individuals 1111 (n-4)

* = GP/Bc homozygous • = GP/Bc x CBA/J heterozygous

Figure 23(b): Mapping matrix of pinhole F2 mice in GP/Bc x CBA/J cross. A comparison between Dl IMit 149 and DllMitlO.

104 DllMitl49 • • • D9Mitl91 • • •

# of Individuals 24 10 7 (n-41)

• = GP/Bc homozygous = GP/Bc x CBA/J heterozygous • = CBA/J homozygous

Figure 24(a): Mapping matrix of open eyelid F2 mice in GP/Bc x CBA/J cross. A comparison between Dl 1MU149 andD9Mitl91. Data does not include pinhole F2 mice.

DllMitl49 « •

D9Mitl91 • •

# of Individuals 2 2 (n = 4)

• = GP/Bc homozygous B = GP/Bc x CBA/J heterozygous • = CBA/J homozygous

Figure 24(b): Mapping matrix of pinhole F2 mice in GP/Bc x CBA/J cross. A comparison between Dl 1 Mit 149 and D9MU191.

105 CHAPTER VII: DISCUSSION

I. Segregation Studies

The gaping lids mutation was reported to be a fully penetrant, recessive, single locus mutation (Kelton and Smith, 1964) and is maintained as such in the GP/Bc strain in the Juriloff/Harris (Be) animal unit (UBC, Vancouver, Canada). During the course of my study, analysis of two different outcrosses to normal (non-open eyelid) strains confirmed the recessive, single major locus (gp) nature of gaping lids. However, the frequencies of affected (open eyelid) newborns recovered in the F2 generations of both outcrosses were lower than the 25% expected (i.e. 19% in the CBA/J cross and 11% in the ICR/Be cross).

From closely linked markers, it was estimated that gp had penetrance of 82% in the cross to CBA/J and 33% in the cross to ICR/Be based upon the observation of phenotypically normal mice homozygous for the GP/Bc allele at the SSLP loci closest to gp. There was no deficiency of the GP/Bc allele in the whole sample indicating there was no detectable loss of gplgp progeny in utero.

A hypothesis to explain the reduced penetrance seen in the F2 generations of both outcrosses and genotypically gplgp individuals in the normal F2 populations was proposed and investigated. We hypothesized the presence of a modifier locus or loci, i.e.

a suppressor-of-open eyelids locus, introduced by the normal strains and unlinked to gp, that is segregating in a Mendelian fashion within the gplgp F2 population. For example,

a single, fully penetrant, recessive unlinked modifier locus, suppressing the effect of the gp locus, would result in XA of the gplgp progeny appearing phenotypically normal at birth giving an open eyelid frequency of V* of 25% or 19% . As eyelid closure is a threshold trait, these proposed modifiers could act to "push" these progeny away from

106 the threshold, resulting in them appearing phenotypically normal at birth when they have the genotype associated with open eyelids, gplgp.

A. CBA/J cross

Although the 19% affected (open eyelid and pinhole) recovered in this cross is not significantly lower (x2 = 0.0025, p>0.25, df = 1) than the 25% expected, the pattern is consistent with there being an unlinked suppressor-of-open eyelids locus that is introduced by the normal (non-open eyelid) strain. Upon screening the normal F2 population with Dl lMit62 and Dl lMit226, 4 phenotypically normal mice apparently homozygous for the GP/Bc allele were identified. This is approximately 4% (4/106) of the total progeny screened, which is not statistically different than expected if the hypothesis of an unlinked suppressor locus is true (x2 = 0.707, p>0.5, df=l), where one would expect 6% or approximately 6/106 of the total progeny to be genetically gaping

lids but phenotypically normal.

B. ICR/Be cross

The 11% affected (open eyelid and pinhole) recovered in the F2 in this cross is

significantly (x2 = 29.51, pO.OOl, df = 1) lower than the 25% expected for a Mendelian

recessive. Based on the data from the first cross, it was hypothesized that there were

modifying loci introduced in this cross as well. Upon screening the normal F2 population

at Dl lMit74, 14 phenotypically normal mice homozygous for the GP/Bc allele were

identified. This is approximately 21% (14/65) of the total progeny screened. These data

do not fit the hypothesis of a single modifier locus proposed for the CBA/J cross. Instead

it appears that there is more than one suppressor locus segregating in this cross. For

107 example, if there are two recessive unlinked suppressors segregating in this cross, either of which could suppress open eyelids, I would expect 11% of the total progeny to be suppressed gplgp (7/65) which does not fit my data (14/65; %2 = 7.85, p<.005, df=1).

However, if two dominant unlinked suppressors were both required together to suppress open eyelids, then I would expect 14% of the total progeny to be suppressed gplgp (9/65) which is not statistically different from my data (14/65; x2 = 3.23, p>0.10, df=l). Table

12 outlines these hypothetical scenarios. However, these are only two of numerous hypothetical models that could explain the data.

For both of these crosses, a study to determine the number of modifying loci and then map them was not undertaken for the following reasons. To determine the number of modifier loci involved would require further study involving test crosses of individual suppressed F2 animals, which would identify the number of different genetic types of non-penetrant genetically open eyelids animals that there were. Additionally, a genome screen to map the modifier loci was not undertaken because the number of non-penetrant genetically open eyelids animals was too small, i.e. with a sample of 4 animals in the

CBA/J cross, there would be no statistical power, since those animals would all be homozygous for the GP/Bc allele QA)4 times by chance alone. In the ICR/Be cross 14 pups does not provide the necessary statistical power either, especially considering the number of modifying loci is unknown and it appears to be more than one.

108 Table 12: Outline of modifier scenarios in GP/Bc x ICR/Be F2

Scenario 1: two fully penetrant, recessive, unlinked suppressor loci Genotypes and genotypic Genotypes and genotypic Phenotype and phenotypic ratios at gp locus in F2 ratios at modifier loci1 ratio

9/16 Ml M2 14% open eyelids 1 gplgP 3/16 Ml_m2m2 >v 3/16 mlmlM2_ > 11% suppressed 1/16 mlmlm2m2

2+/gp as above normal closed eyelids

1 +/+ as above normal closed eyelids

Scenario 2: two fully penetrant, dominant, unlinked suppressor loci Genotypes and genotypic Genotypes and genotypic Phenotype and phenotypic ratios at gp locus in F2 ratios at modifier loci1 ratio

9/16 M1_M2_ 14% suppressed 1 gplgp 3/16 Ml_m2m2 3/16 mlmlM2_ > 11% open eyelids 1/16 mlmlm2m2

2+lgp as above normal closed eyelids

1 +/+ as above normal closed eyelids 1 Ml/ml = modifier locus 1, M2/m2 = modifier locus 2

109 C. Applications of Threshold Model

a) The normal F2 population

Assuming the presence of one or more modifiers of open eyelids at birth in the two normal strains and that open eyelids at birth is a threshold trait, a hypothesis to

explain these apparently gplgp "normal" progeny can be proposed. Whether the

threshold is time or tissue, these pups would appear to have some genotypic combination

of alleles at suppressor loci that decreases their liability to open eyelids at birth, and

hence, they are born phenotypically normal. As stated above, for the CBA/J cross it

appears as though it is one suppressor locus, whereas in the ICR/Be cross it appears to be

multiple loci, resulting in more phenotypically "normal" progeny in the second cross.

This threshold concept is illustrated in Figure 25.

b) The pinhole population

Additional phenotypic "anomalies" are the affected animals classed as pinholes.

They appear to have abundant eyelid tissue and have begun the process of eyelid fusion

but still have a small "pinhole" sized gap present at birth. These progeny occur at

extremely low frequency in both crosses (see Tables 3 and 4) and do not fit the one

threshold hypothesis proposed for the "normal" gplgp mice, i.e. they should be normal or

open eyelids not somewhere in the middle. This led to an alternative hypothesis to that

proposed above. If there were two thresholds occurring very close together, e.g. the right

amount of eyelid tissue and enough time in which to complete eyelid fusion, which both

had to be met for proper eyelid closure, then the pinholes could be explained as only

having met one of these thresholds. Which threshold occurs first, or has to be met first,

can not be determined without extensive study, but I speculate it could be either. If the

110 Threshold

b ICR/Be F2 GP/Bc /\ /

Figure 25: Graphic illustration of the hypothesized effect of the suppressor locus/loci in the a) CBA/J andb) ICR/Be crosses. The shaded area denotes open eyelids at birth. Genotypes of the parental strains are indicated. • indicates "normal" gp/gpprogeny.

Ill tissue threshold is met, i.e. the pinhole progeny have the correct amount of eyelid tissue to complete fusion, but do not meet the time threshold, i.e. they begin fusion later than normal progeny, they would be born with partially, but not completely, fused eyelids.

The converse could also be true, where these progeny begin eyelid closure at the proper

time, and fusion would be completed at the correct time, i.e. the time threshold is met, but

they do not have enough tissue to complete fusion, i.e. they do not meet the tissue

threshold. Such animals might also be born with partially but not completely fused

eyelids. However, this hypothesis does not explain the lower number of pinhole F2

progeny in the ICR/Be cross compared to the CBA/J cross. Since it appears there are

multiple suppressor-of-open eyelid loci in ICR/Be, which would move the ICR/Be

population curve further to the normal side of the thresholds, one would expect the gplgp

F2 population from this cross also to be moved towards the normal side of the thresholds.

This would increase the area under the curve between the two thresholds, thereby

increasing the relative frequency of pinholes in the ICR/Be F2 population, which is not

seen here. This threshold concept is illustrated in Figure 26.

The GP/Bc strain itself is homozygous at all loci, the gaping lids locus (gp) and

any modifier loci that could suppress or promote open eyelids in progeny. By selecting

for fully penetrant open eyelids at birth and breeding these gaping lids pups, the

suppressing loci have been selected against in progeny since inbreeding began in the

1960's. Therefore, in the GP/Bc strain loci which could affect open eyelids (i.e. suppress

its expression) are in a homozygous state (they could be dominant or recessive loci),

which either promote or do not affect open eyelids at birth.

112 Thresholds

SPlgP ICR/Be CBA/J F2 GP/Bc ICR CBA

' ' ** ** / \

Figure 26: Graphic illustration of the hypothesized two threshold model to explain pinhole progeny in the F2 generations of both GP/Bc outcrosses. The shaded area denotes open eyelids at birth. Genotypes of parental stains are indicated. | | indicates "normal" gp/gp progeny, indicates pinhole F2 progeny.

Normal strains, however, have loci which could potentially modify the state of open eyelids when crossed with an open eyelid strain. These loci will not affect eyelid closure in normal strains because they do not carry major loci for open eyelids at birth in a mutated state (i.e. CBA/J and ICR/Be would be wildtype (+/+) at the gaping lids locus).

In the Fl, the loci that affect open eyelids would become heterozygous (as does the gp locus). In the F2 a proportion of progeny (25%) will be homozygous gplgp with different combinations of modifying loci, which may or may not affect open eyelids. Some of the combinations of modifying loci will push a proportion of gplgp progeny towards the normal side of the threshold. Those gplgp progeny which receive the normal strain modifiers in a combination that they affects the state of eyelid closure could potentially be born with partially closed (i.e. pinhole) or completely closed (i.e. appear "normal" phenotypically) eyelids.

113 II. Mapping studies

A. GP/Bc study

The data accumulated during the course of my study allowed me to create recombination/linkage maps which placed gp at the proximal end of Chr 11, near the centromere. The map location of gp determined in the CBA/J cross places it at the proximal end of Chr 11 close to the SSLP markers Dl lMit62 and Dl lMit226. However, due to a lack of informative SSLP markers above Dl lMit62 in this cross, it could not be determined precisely whether gp lay proximal to these markers, making a second outcross necessary. In the ICR/Be cross, 2 mice were found to be recombinant with markers distal to Dl lMit62 and proximal to Dl lMit226. Mouse #3345-B is homozygous for the GP/Bc allele above Dl lMitl62 (from Dl lMit2 to Dl lMit74; see Figure 11) and mouse # 3329-

B is homozygous for the GP/Bc allele above Dl lMit71 (from Dl lMit72 to Dl lMit74; see Figure 11). Mouse # 3345-B supports the location identified in the CBA/J cross, i.e. gp is proximal to Dl lMit80 close to Dl lMit62 and Dl lMit226. Mouse #3329-B places the gp locus at the most proximal end of Chr 11 near Dl lMit74 and the centromere.

However, due to the lack of a flanking marker(s), it cannot be determined whether gp lies proximal or distal to this SSLP locus, a problem inherent with mapping genes close to the centromere (see below). Therefore the broadest region to which gp maps is between

Dl lMit80 and the centromere (approximately 12.3 cM), and the narrowest is between

Dl lMit71 and the centromere (approximately 2.2 cM), with the SSLP markers Dl lMit62 and Dl lMit226 being highly associated with gp in both crosses.

Overall, it appears that these estimates of distance around the gaping lids mutation are reliable, (3.8 - 12.3 cM in the CBA/J cross; see Figure 8; 2.2- 4.4 cM in the ICR/Be

114 cross; see Figure 14). These distances should be tempered with the knowledge that the sample sizes used to determine them were only 82 meioses (CBA/J) and 46 meioses

(ICR/Be), respectively. Frequently such estimates can be influenced by undetected double recombination events, which lead to an underestimate of the true distances.

However, based on human analyses, it was concluded that in experiments in which fewer than 1000 meiotic events are analyzed/typed, multiple recombination events within 10 cM intervals, are extremely unlikely (see Silver, 1995) - a phenomenon known as genetic interference. A similar degree of interference has also been observed in mice (Silver,

1995). This suggests that there are no double recombinants between the various SSLP loci and the gp locus and the number of recombinants observed is the actual number of recombinants in these 2 crosses. Additionally, it seems likely that the one open eyelid mouse (#2775-B) in the CBA/J cross that is homozygous for the CBA/J allele is most likely a phenocopy rather than recombinant on both chromosomes in such a short distance and such a small sample.

The region of Chr 11 to which gp maps has quite a few identified genes (see Table

13; MGI4, 1999). The original candidate in the region, Egfr, was eliminated by evidence gathered during my study, specifically the lack of allelism detected during the complementation test between gp and Egfr+I~, and the map location of Egfr relative to

SSLP loci used in my study (see Chapter V and below). Of the 14 loci mapped around the gp region, between Dl lMit80 and the centromere (see Table 12), the likeliest candidates are leukemia inhibitory factor (Lif) and neurofibromatosis 2 (NJ2), based on their reported gene function and/or expression patterns during development.

115 C3 O C3 B .3 3 o. cfl o •O -rH a S e5 -3 s„ C^3 cfl O B cfl v. eu cu 3 00 |T3 £ o cu ca cn ^H u "ca a. (4-c Q ca 3 T3 CB T3 . „ B Cfl CD B cfl a CO O C3 a —j cfl CU o £> |» Ml CCJ CO •3 cu .g , «> > 3 a.a C/l u tU M= -a CU

§ §1 n o Cfl '&§ T3 cu cu 00 cfl >- ' IX eu c B cu tfl a. a 3 x ea o u 3 lw « Si '•c5a cfl cfl ll cu cfl 0) +i +3 > ra o -3 o "a _ cu .3 fic ca o >> u 00 .c§u u o 2 a VH OH it ^ B o O. Cfl a .5

00 u OH B B 00 .caa > s cfl CU a cu 3 o00 o a 8 w o

Q. CS

cu o cn B -O OH 00 oo lo I

cn CN .3 o I? 2 a -a a 5 5 & 2 o ca 00 c C J-H -t-J 00 c i tu 5 •Hi 6 a •2 3 « H * s 2 .S I-H ca u -a oo O 3 2 o -O .id Cfl cj a ,ea a ,ca 116 3 TO - i- Q. P. •3 c e CS c to CD c CD > c TO CD C CD

"O T3 B p C CO CO O <; cO p J3 o 2 cn 3 O 13 cn .2 O O CD *H C cO 73 ^if ca Q P. CO CD ro CD TO M P .2io, o^ - c )_ J3 CO CD 75 p CO OJ -sCD. 3 CO .S CD CL) 1) 2 c cn CD TO CD § cn .3 c^ TO 'a P- o c/5 P .3 P CD g a. O 2 Q N .P 1.2 O Ii si 2 73 ° _C0 O O. CD to P. •s s 6J) ,CD CD cn CJ &. .3 s _ s p co 'S 73 S -3 ••H L- 2P ft 6 P S s s CO o §. .23 u 3 « S i I IS J- CD D. e u o T3 O *3 CD CD >TO c-.S ^ 3 S

CD O s -a

a CD s: o o o CD X c 1— CD CD >~ . o CD d 2 3 S P cn I o •3 TO £ ca ^ cl S CD P- TO > 3 co 2 § In CD N tn w CL, DC Xi o a. o TO CD O a ix 117 The leukemia inhibitory factor (LIF), encoded by Lif, is known to induce macrophage differentiation and suppress proliferation of murine myeloid leukemia cells as well as a number of other activities (MGI4, 1999). LIF is a member of a subfamily of cytokines which induce an acute-phase response to infection or inflammation. LIF is known to be expressed in a number of tissues during development, including thymus, skin, lung, intestine, and uterus (MGI4, 1999). There are two transcript forms of LIF, produced by use of alternate promoters. One transcript gives rise to a diffusible protein, while the other produces an immobilized form incorporated into the extracellular matrix.

Neurofibromatosis 2 is a disease of the nervous system characterized by tumors of the eighth cranial nerve, meningiomas of the brain and schwannomas of the dorsal roots

of the spinal cord (MGI4, 1999). Expression of Nf2 has been reported to be widespread, yet is limited to specific cell types such as lens epithelial and fiber cells. Mice homozygous for a null allele at this locus die between embryonic days 6.5 and 7 due to

collapse of the extraembryonic region (MGI4, 1999). Although no reports of open

eyelids at birth has yet been associated with this gene, a number of human studies have

found ocular anomalies, such as blindness and corneal opacity, in patients with this

disorder which could be hypothesized to be due to delayed eyelid development (OMIM,

1999).

Additionally, there are also any number of unmapped potential candidate genes,

such as members of the integrin family, the ADAM (a disintegrin and metalloproteinase)

family, or other genes involved in cell migration and proliferation.

118 Interestingly, the loose linkage (at 40 cM) initially found between gp and oe by

Kelton and Smith (1964), but which was discounted by the mouse genetics community, has been shown to be essentially correct using my data. Based on the distances reported by the 1999 Chromosome Committee map (MGI4, 1999), gp is 40.0 - 46.0 cM proximal to oe and these distances fall within the recombination fraction reported by Kelton and

Smith (1964).

a) SSLPs and the centromere

Historically, mapping and determining genetic distances of markers relative to the centromere has been problematic. The most commonly used method to identify the centromeric end of mouse chromosomes was by chromosomal aberrations such as

Robertsonian translocations (Ceci et al., 1994; Silver, 1995). However, many (but not all) Robertsonian translocations are known to suppress recombination, which results in an underestimate of genetic distances between the centromere and proximal markers

(Davisson and Akeson, 1993). Additional mapping techniques include ovarian teratomas

(Eppig and Eicher, 1983) and studying individual oocytes that have finished meiosis I but not meiosis II. The usefulness of ovarian teratomas is limited by their rarity in most

laboratory mouse strains, and examining oocytes is time consuming. C-band polymorphisms have also been used to map centromeres. This method is based on the major satellite DNA sequences located in the centromere of all but the Y chromosome, which are polymorphic between mouse species distantly related to Mus domesticus.

However, this method, although useful, is also time consuming and tedious, as it is done using cytogenetics.

119 Ceci et al. (1994) developed a method for mapping mouse centromeres which utilizes an interspecific backcross and a centromere specific major satellite sequence probe. The major satellite DNA is a 234 bp tandemly repeated family of DNA which comprises nearly 10% of the mouse genome. The relative copy number of these sequences differs between Mus species, e.g. there are 700,000 copies distributed among the centromeres of M. musculus compared to 25,000 copies spread among the centromeres in M. spretus (Silver, 1995). This differential in copy number is the basis of the method developed by Ceci et al. (1994). They determined, utilizing the centromeric repeats, that the segregation of chromosomes labeled by fluorescence in situ hybridization

(FISH) could be determined using karyotyped metaphase chromosomes of interspecific

BC progeny.

Using this approach, Ceci et al. (1994) mapped genetic distances between the pericentric heterochromatin (He) and proximal loci of all mouse chromosomes but the Y in C57BL/6Ros and M. spretus reciprocal BCs. Proximal markers were typed in the BC by Southern blot analysis using RFLPs. For Chr 11 specifically, they found that the

leukemia inhibitory factor (Lif) locus did not recombine with the centromere in 126 progeny. The upper 95% CI limit places Lif within 2.3 cM of the heterochromatic marker

at the centromere of Chr 11.

Since this study, three SSLP markers have been mapped against Hell, Dl lMitl,

Dl lMit71, and Dl lMit77 (MGI4, 1999). By using the map positions of these 3 SSLP loci relative to Hell, other SSLP loci have been mapped relative to the centromere.

Additionally, the SSLP markers Dl IMit 16 and Dl IMit 106 are within the Lif gene

(MGI3,1999). Mapping studies using these 2 SSLP loci have shown recombination

120 proximal to Lifwith the SSLP Dl lMit74, indicating this SSLP is most likely closest to the centromere (see Abdel-Majid et al., 1998).

b) The mouse maps

Inherent with using a map resource created from multiple sources of mapping data are the errors which come with it. These data sources tend to be in a state of flux, where data are continually being added and refined/revised, (this specifically relates to the MGI map). All three maps described in the Introduction used different crosses, i.e. different strains and numbers of progeny to generate their data. Therefore, when all of these data are integrated into a source such as MGI, the "best fit" is generally used but is not necessarily correct. In most cases, a framework map is used at the start of a project and is modified in accordance with the data generated, so the fit of the data and map is most parsimonious.

This point came to the forefront in my study with regard to the SSLP locus

Dl IMit 149. This marker was placed at the proximal end of Chr 11 by cytogenetic analysis (McCarthy et al., 1997) and the 1999 Chromosome Committee, (MGI3, 1999).

However, my data showed that it was highly improbable for this marker to map to Chr 11 based on the recombination patterns found in my sample of open eyelid F2 mice from the

CBA/J cross. A generic World Wide Web search engine located a reference site which placed this marker in the Jackson Laboratory BSB panel for Chr 9, data which had not been integrated into the 1999 Chromosome Committee Map (MGI3,1999).

The map positions of SSLP marker loci can also be influenced by preexisting map data. In a study by Fairchild et al. (1995), the SSLP markers Dl lMit71, Dl lMitl6 and

Dl lMit74 were mapped against each other. This study placed Dl lMit71 at the most

121 proximal end of Chr 11 in accordance with the order of SSLP loci reported by Research

Genetics/MIT. However, upon reviewing these data and placing the markers in an order that would reduce the apparent double recombinants, Dl lMit74 becomes the most proximal SSLP marker, followed by Dl lMitl6 and Dl lMit71. This order is supported by additional studies including my own (see Claudio et al, 1994; 1999 Chromosome

Committee report: MGI3,1999), pointing out, again, how these maps must be viewed with caution.

c) SSLPs not tested

Of the 44 SSLP marker loci listed on the Research Genetics map and 47 listed on the MGI map between 0 and approximately 15 cM, 18 and 20 SSLP loci, respectively, were not tested for informativeness in my two GP/Bc crosses. These loci are listed in

Appendix P. These loci were not tested for one or more of the following reasons: on both maps several of these markers occurred in clusters mapped to the same or a close by

(less than 0.5 cM) region; based on lack of polymorphism reported in the 12 standard laboratory strains tested by Dietrich et al. (1992) it did not appear likely that the marker(s) would be informative between either GP/Bc and CBA/J or GP/Bc and ICR/Be; the confidence level associated with the locations of some of the markers on the Research

Genetics map was not as high as with other markers; markers distal to the breakpoints determined in both crosses would not provide anymore information in such a small sample size (e.g. n = 41 in CBA/J cross; n = 23 in ICR/Be cross). The markers that occurred at the same map position that were tested were selected because they seemed likely to be informative, but if they were not another marker(s) at the same or close by location was then tested (e.g. Dl lMit226 and Dl lMit62)

122 B. Egfr study

The mapping of Egfr was undertaken to define the location of this locus relative to SSLP marker loci used to map the gaping lids locus in the first part of my study. As gp and Egfr appeared to be non-allelic based on the complementation test between Egfr+I~ and GP/Bc mice, it was of interest to determine if they actually mapped to unique locations in proximal Chr 11. The data accumulated during the course of my study allowed me to create haplotypes for mice from my special testcross, placing Egfr between

Dl lMit226 and Dl lMitl51, a distance of 2.8 cM (see Figure 19). As gp maps near

Dl lMit62 and Dl lMit226 and most likely is proximal to them, the map position of Egfr determined here supports the interpretation that gp and Egfr are not alleles. Further study to investigate the possible relationship and/or interaction of these loci is needed.

Considering the nature of the open eyelids associated with wa2, and that this feature was not described as part of the initial phenotypic description, it is possible that these two loci do interact in some way. It is also possible that these are simply two open eyelids at birth mutations that map relatively close to each other. There are now 6 different genotypes on

Chr 11 that have open eyelids at birth as a feature of their phenotype. They are: nonsyndromic - gp, oe; syndromic - Egfr/wa2, RARal (with RARyox RARa2+'~); ectopic gene expression - Hoxb6, Hoxbl.

III. Phenotypic Investigations

The only phenotypic anomaly associated with the gaping lids mutation is the open eyelids at birth. It is not known whether the white belly spot that occurs in the GP/Bc

123 strain is part of the mutant phenotype of the gp mutation or some other gene. This will not be discussed here in any detail. The gp mutation appears to affect the cells that make up the eyelids themselves. Newborns have wide open eyes which tend to scab over after birth (M.J. Harris, personal communications). Adult GP/Bc animals have damaged corneas, which seem to give them their milky/murky appearance later in life. Whether this damage occurs in utero, due to exposure of the developing eye to amniotic fluid, or after birth, due to mechanical injury obtained during the growth period, is not known. A timed observation of the phenotypic progression was not done in my study.

Close examination of GP/Bc eyes revealed that these mice do have "eyelids", or at least protrusions of tissue over the eyes which can come close together when the animal is forced to blink. Upper and lower eyelashes are also present indicating there is enough tissue for the hair follicles to form but not enough to close fully. A white milky exudate was also noted in all GP/Bc eyes examined, which I speculate is providing moisture and lubrication to the eye or is a reaction to corneal damage.

Kelton and Smith (1964) did not report scabbing of the eyes in newborn mice but did report corneal opacities a few days after birth, consistent with the hypothesis of damage to the eye after birth. They also reported that the lenses of d 14-18 embryos were approximately twice as large as those of their C57BL/6 controls. A comparison of the lenses of GP/Bc and AXB-23/Pgn (normal strain) mice did not reveal a larger size difference. I speculate that the C57BL/6 mice used as controls by Kelton and Smith

(1964) actually had small eyes, since these mice have a liability to microphthalmia, and that large lenses is not actually a feature of the gp phenotype. However, to determine this

124 with any certainty, further study is required, with histological sectioning of eyes of GP/Bc mice between embryonic dl4 and birth.

Stein et al. (1967) compared gp to oe and their open eyelids at birth mutant, which they called slit lids (now known as lidgap-Stein (lg^tn)). They reported a "hardened

serous exudate", i.e. a scab, over gplgp eyes in later developmental stages, which is fairly consistent with the observations of GP/Bc newborns in the Juriloff/Harris animal unit.

Stein et al. (1967) also reported that gplgp lenses were larger, but they were comparing them to their mutant, not normal controls, so this may again be an incorrect phenotypic

assignment since the lenses were expected to be larger because of Kelton and Smith's

(1964) report.

My scanning electron microscopy (SEM) results confirmed the lack of eyelid

fusion seen in these mutant mice (see Figure la and lb). When compared to SEM of

normal (SWV/Bc and CBA/J) mice (see Juriloff and Harris, 1989; Harris and Juriloff,

1986), it can be hypothesized that the gp mutation affects eyelid growth and/or cell

proliferation in the region, since GP/Bc mice have far fewer rounded cells at equivalent

stages of development. Normally by dl6 of development there are numerous rounded

cells around the perimeter of the eye, fusion has begun and is visible at both canthi, and

the gap between the eyelids is getting ever smaller. Further investigations may examine

the phenotypic progression of gaping lids from the time the eyelids normally begin to

grow until they would normally open.

IV. Conclusions

In summary, the main objective of this study was to map the gaping lids locus,

based on the hypothesis it was at one of several candidate loci. In the process I was able

125 to investigate the segregation of this recessive allele and investigate the phenotype in broad detail. Regarding the main objective, the work in this thesis has shown that the gaping lids mutation is a recessive allele with phenotypic effects that can be modified in outcrosses to normal (non-open eyelid) strains. These strains introduce apparent suppressor-of-open eyelid loci. I mapped gp to an approximately 2-14 cM, gene rich region on proximal mouse Chr 11.

The exact map location of gaping lids and its relationship, if any, to Egfr has yet to be determined. The apparent role of the gp locus in relation to other, more complex open eyelid mutations, e.g. lg^tn, IgMl, also remains to be determined. By mapping gp, however, the first of many steps in understanding the relationship between complex loci has been taken. This opens the door to multiple hypotheses regarding the genetic

interactions of seemingly "simple" loci with more complex ones.

126 LITERATURE CITED

Abdel-Majid, R.M., W.L. Leong, L.C. Schalkwyk, D.S. Smallman, S.T. Wong, D.R. Storm, A. Fine, M.J. Dobson, D.L. Guernsey, and P.E. Neumann. 1998. Loss of adenylyl cyclase I activity disrupts patterning of mouse somatosensory cortex. Nature Genetics 19(3), 289-291.

Adamson, E.D. 1990a. Developmental activities of the epidermal growth factor receptor. Current Topics in Developmental Biology 24,1-29.

Adamson, E.D. 1990b. EGF receptor activities in mammalian development. Molecular and Reproductive Development 27, 6-22.

Ball, S.T. and J. Peters. 1989. Koala, a dominant mutation. Mouse News Letter 83, 163-164.

Balling, R., G. Mutter, P. Grass, and M. Kessel. 1989. Craniofacial abnormalities induced by ectopic expression of homeobox gene Hox-1.1 in transgenic mice. Cell 58, 337-347.

Bennett, J.H. and G.A. Gresham. 1956. A gene for eyelids open at birth in the house mouse. Nature 178,272-273.

Berkowitz, E.A., K.B. Seroogy, J.A. Schroeder, W.E. Russell, E.P. Evans, R.F. Riedel, H.K. Phillips, CA. Harrison, D.C. Lee, and N.C. Luetteke. 1996. Characterization of the mouse transforming growth factor-a gene: its expression during eyelid development and in waved-1 tissues. Cell Growth and Differentiation 7, 1271-1282.

Blaschke, R.J., A.P. Monaghan, S. Schiller, B. Schechinger, E. Rqo, H. Padilla-Nash, T. Ried, and G.A. Rappold. 1998. SHOT, a SHOX-related homeobox gene, is implicated in craniofacial, brain, heart, and limb development. The Proceedings of the National Academy of Science 95, 2406-2411.

Boyd, J., M.J. Harris, and D.M. Juriloff. 1984. Research news. Mouse News Letter 70, 66-68.

Breen, M., L. Deakin, B. MacDonald, S. Miller, R. Sibson, E. Tarttelin, P. Avner, F. Bourgade, J.L. Guenet, X. Montagutelli, C. Poirier, D. Simon, D. Tailor, M. Bishop, M. Kelly, F. Rysavy, S. Rastan, D. Norris, D. Shepherd, C. Abbott, A. Pilz, S. Hodge, I. Jackson, Y. Boyd, H. Blair, G. Maslen, J.A. Todd, P.W. Reed, J. Stoye, A. Ashworth, L. McCarthy, R. Cox, L. Schalykwyk, H. Lehrach, J. Klose, U. Gangadharan, and S. Brown. 1994. Towards high-resolution maps of the

127 mouse and human genomes - A facility for ordering markers to 0.1 cM resolution. Human Molecular Genetics 3(4), 621-627.

Bronner-Fraser, M. 1995. Origins and developmental potential of the neural crest. Experimental Cell Research 218, 405-417.

Brown, K.D. 1995. The epidermal growth factor/transforming growth factor-a family and their receptors. The European Journal of Gastroenterology and Hepatology 7(10), 914-922.

Brown, K.S. and L.C. Home. 1973. Recessive anencephalus in the oel strain of mice (Abstract). Genetics 74(Suppl.), S31-S32.

Butler, L. and D.A. Robertson. 1953. A new eye abnormality in the house mouse. The Journal of Heredity 44, 13-16.

Carroll, J.M., M.R. Romero, and F.M. Watt. 1995. Suprabasal integrin expression in the epidermis of transgenic mice results in developmental defects and a phenotype resembling psoriasis. Cell 83, 957-968.

Carter, T.C. 1955. Genetics of the Little and Bagg X-rayed mouse stock. The Journal of Genetics 54,311-326.

Ceci, J.D., Y. Matsuda, J.M. Grubber, N.A. Jenkins, N.G. Copeland, and V.M. Chapman. 1994. Interspecific backcrosses provide an important tool for centromere mapping of mouse chromosomes. Genomics 19(3), 515-524.

Center, E.M. 1977. Genetical and embryological comparisons of two mutations which cause foetal blebs in mice. Genetical Research 29, 147-157.

Claudio, J.O., D. Malo, and G.A. Rouleau. 1994. The mouse neurofibromatosis type 2 gene maps to chromosome 11. Genomics 21(2),437-439.

Copeland, N.G., N.A. Jenkins, D.J. Gilbert, J.T. Eppig, L.J. Maltais, J.C. Miller, W.F. Dietrich, A. Weaver, S.E. Lincoln, R.G. Steen, L.D. Stein, J.H. Nadeau, and E.S. Lander. 1993. A genetic linkage map of the mouse: current applications and future prospects. Science 262, 57-66.

Davisson, M.T. and E.C. Akeson. 1993. Recombination suppression by heterozygous Robertsonian chromosomes in the mouse. Genetics 133, 649-667.

Dietrich, W.F., H. Katz, S.E. Lincoln, H.S. Shin, J. Friedman, N.C. Dracopoli, and E.S. Lander. 1992. A genetic map of the mouse suitable for typing intraspecific crosses. Genetics 131, 423-447.

128 Dietrich, W.F., J.C. Miller, R.G. Steen, M. Merchant, D. Damon, R. Nahf, A. Gross, D.C. Joyce, M. Wessel, R.D. Dredge, A. Marquis, L.D. Stein, N. Goodman, D.C. Page, and E.S. Lander. 1994. A genetic map of the mouse with 4.006 simple sequence length polymorphisms. Nature Genetics 7,220-225.

Dietrich, W.F., J. Miller, R. Steen, M.A. Merchant, D. Damon-Boles, Z. Husain, R. Dredge, M.J. Daly, K.A. Ingalls, T.J. O'Connor, CA. Evans, M.M. DeAngelis, D.M. Levinson, L. Kruglyak, N. Goodman, N.G. Copeland, N.A. Jenkins, T.L. Hawkins, L. Stein, D.C. Page, and E.S. Lander. 1996. A comprehensive genetic map of the mouse genome. Nature 380, 149-152.

Doolittle, D.P., M.T. Davisson, J.N. Guidi, and M.C. Green. 1996. Catalog of mutant genes and polymorphic loci. In: Genetic Variants and Strains of the Laboratory Mouse, 3rd Ed. M.F. Lyon, S. Rastan, and S.D.M. Brown, Eds., Oxford University Press, Oxford, pp. 17-854.

Dreger, R.T., M.J. Harris, CL. Peichel, T.G. Vogt, and D.M. Juriloff. 1995. The First arch (Far) mutation in mice recombines with Hoxd and Mdk. Mammalian Genome 6, 662-663.

Drews, R., W.N. Drohan and H. Lubon. 1994. Transgene detection in mouse tail digests. Biotechniques 17(5), 866-867.

Duke-Elder, S. and C. Cook. 1963. The lids, conjunctiva, and lacramal apparatus. In: System of Ophthalmology, Volume 3, Part 1: Normal and Abnormal Development -Embryology, Sir. S. Duke-Elder, Ed., CV. Mosby Company, pp. 231-246.

Dunn, A.R., G.B. Mann, K.J. Fowler, D. Grail, MX. Hibbs, W.S. Alexander, F. Walker, and A.W. Burgess. 1994. Insights into the physiology of Tgfa and signaling through the EGF receptor revealed by gene targeting and acts of nature. Princess Takamatsu Symposium 24, 276-289.

Ellis, J., Q. Liu, M. Breitman, N.A. Jenkins, D.J. Gilbert, N.G. Copeland, H.V. Tempest, S. Warren, E. Muir, H. Schilling, F.A. Fletcher, S.F. Ziegler, and J.H. Rogers. 1995. Embryo brain kinase: a novel gene of the eph/elk receptor tyrosine kinase family. Mechanisms of Development 52,319-341.

Eppig, J.T. and E.M. Eicher. 1983. Application of the ovarian teratoma mapping method in the mouse. Genetics 103(4), 797-812.

EUCIB: The European Collaborative Interspecific Mouse Backcross Database, UKMRC Mapping Project Resource Centre, Medical Research Council, UK. World Wide Web (http://www.hgmp.mrc.ac.uk/MBx/MBxHomepage.html),

129 May 1999.

Fairchild, S.P., O.A. Rosenwasser and K. Tomonari. 1995. Localization of the mouse mammary tumor provirus, Mtv44, on chromosome 11. Mammalian Genome 6(7), 472-473.

Falconer, D.S. 1989. Threshold characters. In: Introduction to Quantitative Genetics, 3rd Ed., Longman Scientific and Technical Publishers, pp. 300-312.

Feijen, A., M.J. Goumans, A.J. van den Eijnden-van Raaij. 1994. Expression of activin subunits, activin receptors and follistatin in postimplantation mouse embryos suggests specific developmental functions for different activins. Development 120(12),3621-3637.

Festing, M.F. 1989. Inbred strains of mice. In: Genetic Variants and Strains of the Laboratory Mouse, 2nd Ed. M.F. Lyon and A.G. Searle, Eds., Oxford University Press, Oxford, pp. 636-648.

Findlater, G.S., R.D. McDougall, and M.H. Kaufman. 1993. Eyelid development, fusion and subsequent reopening in the mouse. The Journal of Anatomy 183, 121-129.

Fowler, K.J., G.B. Mann, and A.R. Dunn. 1993. Linkage of the murine transforming growth factor a gene with Igk, Ly-2, and Fabpl on chromosome 6. Genomics 16, 782-784.

Franz, T. and A. Besecke. 1991. The development of the eye in homozygotes of the mouse mutant Extra-toes. Anatomical Embryology 84,355-361.

Fujii, S. and S. Teramoto. 1992. Evaluation of maternal and fetal factors in reduced postnatal survival of NC-eob mutant mice. Laboratory Animal Science 42(5), 478-481.

Fujii, S., N. Hatakenaka, M. Kaneda, and S. Teramoto. 1995. Morphological study of the eyelids in NC-eob mice fetuses with an open-eyelid malformation at birth. Laboratory Animal Science 45(2), 176-180.

Gates, A.H. and J.L. Bozarth. 1968. New mutations. Mouse News Letter 39,36-37.

Ghyselinck, N.B., V. Dupe, A. Dierich, N. Messaddeq, J.-M. Gamier, C. Rochette-Egly, P. Chambon and M. Mark. 1997. Role of the retinoic acid receptor beta (RARP) during mouse development. International Journal of Developmental Biology 41, 425-427.

Green, M.C. 1989. Catalog of mutant genes and polymorphic loci. In: Genetic

130 Variants and Strains of the Laboratory Mouse, 2nd Ed. M.F. Lyon and A.G. Searle, Eds., Oxford University Press, Oxford, pp. 12-403.

Gruneberg, H. 1952. The genetics ojthe mouse, 2nd'Ed. The Hague, pp. 147-167.

Gunn, T.M., D.M. Juriloff, and M.J. Harris. 1992. Further genetic studies of the cause of exencephaly in SELH mice. Teratology 45(6), 679-686.

Gunn, T.M., D.M. Juriloff, W. Vogl, M.J. Harris, and J.E. Miller. 1993. Histological study of the cranial neural folds of mice genetically liable to exencephaly. Teratology 48(5), 459-471.

Hall, J.L., M.J. Harris, and D.M. Juriloff. 1997. Effect of multifactorial genetic liability to exencephaly on the teratogenic effect of valproic acid in mice. Teratology 55(5), 306-313.

Hamming, N. 1983. Anatomy and embryology of the eyelids: A review with special reference to the development of divided nevi. Pediatric Dermatology 1(1), 51- 58.

Harris, M.J. Department of Medical Genetics, University of British Columbia, Vancouver, Canada.

Harris, M.J. 1989. Cortisone cure of the eyelid closure defect in lidgap-Stein fetal mice: a dose-response and time-response study as a test of the hypomorph hypothesis for the lidgap alleles. Teratology 39,601-609.

Harris, M.W. and F.C. Fraser. 1968. Lid gap in newborn mice: a study of its cause and prevention. Teratology 1(4), 417-424.

Harris, M.J. and D.M. Juriloff. 1986. Eyelid development and fusion induced by cortisone treatment in mutant, lidgap-Miller, foetal mice. A scanning electron microscope study. The Journal of Embryology and Experimental Morphology 91, 1-18.

Harris, M.J. and D.M. Juriloff. 1989. Test of the isoallele hypothesis at the mouse First arch (Far) locus. The Journal of Heredity 80(2), 127-131.

Harris, M.J. and D.M. Juriloff. 1998. Nonallelic noncomplementation models in mice: The first arch and lidgap-Gates mutations. Genome 41, 789-796.

Harris, M.J. and M.J. McLeod. 1982. Eyelid growth and fusion in fetal mice. Anatomy and Embryology. 164, 207-220.

Harris, M.J., D.M. Juriloff, and F.G. Biddle. 1984. Cortisone cure of the lidgap defect in

131 fetal mice: a dose-response and time-response study. Teratology 29(2), 287-295.

Harris, M.J., D.M. Juriloff, and C.E. Peters. 1990. Disruption of pattern formation in palatal rugae in fetal mice heterozygous for first arch (far). Journal of Craniofacial Genetics and Developmental Biology 10(4), 363-371.

Hayasaka, I., T. Nakatsuka, T. Fujii, I. Naruse, and S. Oda. 1980. Polydactyly nagoya, Pdn: a new mutant gene in the mouse. Experimental Animals 29(4), 391-395.

Hui, C-C. and A.L. Joyner. 1993. A mouse model of Greig cephalopolysyndactyly syndrome: the extra-toes-7 mutation contains an intragenic deletion of the GU3 gene. Nature Genetics 3, 241-246.

Hsuan, J.J., G. Panayotou, and M.D. Waterfield. 1989. Structural basis for epidermal growth factor receptor function. Progress in Growth Factor Research 1, 23-32.

Jakobiec, F.A. and T. Iwamoto. 1982. Ocular adnexa: Introduction to lids, conjunctiva, and orbit. In: Ocular Anatomy, Embryology and Teratology. F.A. Jakobiec, Ed., Harper and Row Publishing, Philadelphia, USA, pp. 677-731.

JAX: The Jackson Laboratory, Bar Harbor, Maine. World Wide Web (http://www.jax.org), February, 1999.

Juriloff, D.M. Department of Medical Genetics, University of British Columbia, Vancouver, Canada.

Juriloff, D.M. 1985. Prevention of the eye closure defect in lgMlllgMi fetal mice by thyroxine. Teratology 32, 73-86.

Juriloff, D.M. 1987. Maternal treatment with cortisone accelerates eyelid closure and other developmental fusion processes in fetal mice. Development 100, 611-618.

Juriloff, D.M. and M.J. Harris. 1983. Abnormal facial development in the mouse mutant first arch. Journal of Craniofacial Genetics and Developmental Biology 3, 317- 337.

Juriloff, D.M. and M.J. Harris. 1989. A scanning electron microscope study of fetal eyelid closure accelerated by cortisone in SWV/Bc mice. Teratology 40, 59-66.

Juriloff, D.M and M.J. Harris. 1991. Mapping the mouse craniofacial mutation first arch (Far) to chromosome 2. The Journal of Heredity 82,402-405.

Juriloff, D.M. and M.J. Harris. 1993. Retinoic acid, cortisone, or thyroxine suppresses the mutant phenotype of the eyelid development mutation, lgMl, in mice. The Journal of Experimental Zoology 265, 144-152.

132 Juriloff, D.M., M.J. Harris, and U. Froster-Iskenius. 1987. Hemifacial deficiency induced by a shift in dominance of the mouse mutation far: a possible genetic model for hemifacial microsomia. Journal of Craniofacial Genetics and Developmental Biology 7(1), 27-44.

Juriloff, D.M., M.J. Harris, and D.G. Mah. 1994. Mapping Far (First arch) in relation to molecular markers on mouse chromosome 2. Mammalian Genome 5, 450-452.

Juriloff, D.M., M.J. Harris, and J.R. Miller. 1983. The lidgap defect in mice: update and hypotheses. Canadian Journal of Genetics and Cytology 25(3), 246-254.

Juriloff, D.M., K.B. MacDonald, and M.J. Harris. 1989. Genetic analysis of the cause of exencephaly in the SELH/Bc mouse stock. Teratology 40(4), 395-405.

Juriloff, D.M., M.J. Harris, D.G. Mah, and A. Benson. 1996. The lidgap-Gates (lgGa) mutation for open eyelids at birth maps to mouse chromosome 13. Mammalian Genome 7(6), 403-407.

Juriloff, D.M., M.J. Harris, C. Tom, and K.B. MacDonald. 1991. Normal mouse strains differ in the site of initiation of closure of the cranial neural tube. Teratology 44(2), 225-233.

Juriloff, D.M., M.J. Harris, M.L. Harrod, T.M. Gunn, and J.E. Miller. 1993. Ataxia and a cerebellar defect in the exencephaly-prone SELH/Bc mouse stock. Teratology 47(4), 333-340.

Juriloff, D.M., M.J. Harris, J.E. Miller, D. Jacobson, and P. Martin. 1992. Is Far a Hox mutation? The Journal of Craniofacial Genetics and Developmental Biology 12, 119-129.

Kalter, H. 1975. Prenatal epidemiology of spontaneous cleft lip and palate, open eyelid, and embryonic death in A/J mice. Teratology 12, 245-258.

Kalter, H. 1979. Sex, side and severity in spontaneous malformations of fetal A/JKt mice, and their associations with each other and with fetal weight. Teratology 19,1-8.

Kaur, S., G. Singh, J.L. Stock, CM. Schreiner, A.B. Kier, K.L. Yager, M.L. Muchenski, W.J. Scott, Jr., and S.S. Potter. 1992. Dominant mutation of the murine Hox-2.2 gene results in developmental abnormalities. The Journal of Experimental Zoology 264, 323-336.

Keeler, CE. 1935. A second rexoid coat character in the house mouse. The Journal of Heredity 26, 189-191.

133 Kelton, D.E., and V. Smith. 1964. Gaping, a new open eyelid mutation in the house mouse (abstract). Genetics 50, 261-262.

Kelton, D.E. and H. Rauch. 1968. Linkage of open eyelids with linkage group VII of the mouse. The Journal of Heredity 59(1), 27-28.

Kume, T., K.Y. Deng, V. Winfrey, D.B. Gould, M.A. Walter, and B.L. Hogan. 1998. The forkhead/winged helix gene Mfl is disrupted in the pleiotropic mouse mutation congenital hydrocephalus. Cell 93, 985-996.

LeDourarin, N.M., C. Ziller, and G.F. Couly. 1993. Patterning of neural crest derivatives in the avian embryo: In vivo and in vitro studies. Developmental Biology 159,24-49.

Lipski, D.A. and E.T. Bersu. 1990. Examination of the eyelid closure defect in trisomy 16 mice. Teratology 42(3), 301-308.

Lohnes, D., M. Mark, C. Mendelsohn, P. Dolle, A. Dierich, P. Gorry, A. Gansmuller, and P. Chambon. 1994. Function of the retinoic acid receptors (RARs) during development. (I) Craniofacial and skeletal abnormalities in RAR double mutants. Development 120, 2723-2748.

Luetteke, N.C., T.H. Qiu, R.L. Peiffer, P. Oliver, O. Smithies, and D.C. Lee. 1993. TGFcc deficiency results in hair follicle and eye abnormalities in targeted and waved-1 mice. Cell 73, 263-278.

Luetteke, N.C., H.K. Phillips, T.H. Qiu, N.G. Copeland, H.S. Earp, N.A. Jenkins, and D.C. Lee. 1994. The mouse waved-2 phenotype results from a point mutation in the EGF receptor tyrosine kinase. Genes and Development 8,399-413.

Macdonald, K.B. 1988. M.Sc. Thesis: "The Genetics and Embryopathology of exencephaly in SELH/Bc mice," The University of British Columbia, Vancouver, Canada.

Macdonald, K.B., D.M. Juriloff, and M.J. Harris. 1989. Developmental study of neural tube closure in a mouse stock with a high incidence of exencephaly. Teratology 39(2), 195-213.

Mackensen, J.A. 1960. 'Open eyelids'in newborn mice. The Journal of Heredity SI, 188-190.

Maconnachie, E. 1979. A study of digit fusion in the mouse embryo. The Journal of Embryology and Experimental Morphology 49, 259-276.

134 Mann, G.B., K.J. Fowler, A. Gabriel, E.C. Nice, R.L. Williams, and A.R. Dunn. 1993. Mice with a null mutation of the TGFa gene have abnormal skin architecture, wavy hair, and curly whiskers and often develop corneal inflammation. Cell 73, 249-261.

Marshall, J.D., J.-L. Mu, Y.-C. Cheah, M.N. Nesbitt, W.N. Frankely, and B. Paigen. 1992. The AXB and BXA set of recombinant inbred mouse strains. Mammalian Genome 3, 669-680.

Matzuk, M.M., T.R. Kumar, A. Vassalli, J.R. Bickenbach, D.R. Roop, R. Jaenisch, and A. Bradley. 1995a. Functional analysis of activins during mammalian development. Nature 374, 354-356.

Matzuk, M.M., T.R. Kumar, and A. Bradley. 1995b. Different phenotypes for mice deficient in either activins or activin receptor type II. Nature 374, 356-360.

McCarthy, L.C., J. Terrett, M.E. Davis, C.J. Knights, A.L. Smith, R. Critcher, K.Schmitt, J. Hudson, N.K. Spurr and P.N. Goodfellow. 1997. A first-generation whole genome-radiation hybrid map spanning the mouse genome. Genome Research 7(12), 1153-1161.

McHenry, J.Z., A. Leon, K.I. Matthaei, and D.R. Cohen. 1998. Overexpression offra-2 in transgenic mice perturbs normal eye development. Oncogene 79(9), 1131- 1140.

McLain, K, C. Schreiner, K.L. Yager, J.L. Stock, and S.S. Potter. 1992. Ectopic expression of Hox-2.3 induces craniofacial and skeletal malformations in transgenic mice. Mechanisms of Development 39,3-16.

McLeod, M.J., M.J. Harris, G.F. Chernoff, and J.R. Miller. 1980. First arch malformation: a new craniofacial mutant in the mouse. The Journal of Heredity 71,331-335.

Mercola, M. and CD. Stiles. 1988. Growth factor superfamilies and mammalian embryogenesis. Development 102,451-460.

Merlino, G.T. 1990. Epidermal growth factor receptor regulation and function Cancer Biology 1, 277-284.

MGI1: Mouse Genome Informatics (MGI) Resource, Mouse Genome Informatics, The Jackson Laboratory, Bar Harbor, Maine. World Wide Web (http://www.informatics.jax.org) June 1998.

MGI2: Mouse Genome Informatics (MGI) Resource, Mouse Genome Informatics,

135 The Jackson Laboratory, Bar Harbor, Maine. World Wide Web (http://www.informatics.jax.org) November 1998.

MGI3: Mouse Genome Informatics (MGI) Resource, Mouse Genome Informatics, The Jackson Laboratory, Bar Harbor, Maine. World Wide Web (http://www.informatics.jax.org) January 1999.

MGI4: Mouse Genome Informatics (MGI) Resource, Mouse Genome Informatics, The Jackson Laboratory, Bar Harbor, Maine. World Wide Web (http://www.informatics.jax.org) May 1999.

Michael, M.I., S.H. Khalil, CA. Matta, and T.A. Rizk. 1988. Normal development of the eyelids in the mouse. Folia Morphologica 36(1), 53-58.

Miettinen, P.J. 1997. Epidermal growth factor receptor in mice and men - any applications to clinical practice? Annals of Medicine 29(6), 531-534.

Miettinen, P.J., J.E. Berger, J. Meneses, Y. Phung, R.A. Pedersen, Z. Werb, and R. Derynck. 1995. Epithelial immaturity and multiorgan failure in mice lacking epidermal growth factor receptor. Nature 376,337-341.

Montgomery, J.C, K.A. Silverman, and A.M. Buchberg. 1997. Chromosome 11. Mammalian Genome 7, S190-S208.

Morris-Kay, G., E. Ruberte, and Y. Fukiishi. 1993. Mammalian neural crest and neural crest derivatives. Annals of Anatomy 175,501-507.

Motoyama, J., H.Heng, M.A. Crackower, T. Takabatake, K. Takeshima, L-C Tsui, and C-C. Hui. 1998. Overlapping and non-overlapping Ptch2 expression with Shh during mouse embryogenesis. Mechanisms of Development 78,81-84.

Munke, M.J. and U. Francke. 1987. The physical map of Mus musculus chromosome 11 reveals evolutionary relationships with different syntenic groups of genes in Homo sapiens. Journal of Molecular Evolution 25(2), 134-140.

Nesbitt, M.N. and E. Skamene. 1984. Recombinant inbred mouse strains derived from A/J and C57B1/6J: a tool for the study of genetic mechanisms in host resistance to infection and malignancy. Journal of Leukocyte Biology 36,357-364.

Nichols, D.H. 1981. Neural crest formation in the head of the mouse embryo as observed using a new histological technique. The Journal of Embryology and Experimental Morphology 64, 105-120.

OMIM: Online Mendelian Inheritance in Man, OMIM™. Center for Medical Genetics,

136 Johns Hopkins University (Baltimore, MD) and National Center for Biotechnology Information, National Library of Medicine (Bethesda, MD). World Wide Web (http://wvvw.ncbi.nlm.nih.gov/omirn/), July, 1999.

Osumi-Yamashita, N., Y. Ninomiya, H. Doi, and K. Eto. 1994. The contribution of both forebrain and midbrain crest cells to the mesenchyme in the frontonasal mass of mouse embryos. Developmental Biology 164,409-419.

Osumi-Yamashita, N., Y. Ninomiya, and K. Eto. 1997. Mammalian craniofacial embryology in vitro. International Journal of Developmental Biology 41, 187- 194.

Pearson, A. A. 1980. The development of the eyelids - Part I. External features. The Journal of Anatomy 130(1), 33-42.

Pei, Y.F. and J.A.G. Rhodin. 1970. The prenatal development of the mouse eye. Anatomical Record 168(1), 105-126.

Peschon, J.J., J.L. Slack, P. Reddy, K.L. Stocking, S.W. Sunnarborg, D.C. Lee, W.E. Russell, B.J. Castner, R.S. Johnson, J.N. Fitzner, R.W. Boyce, N. Nelson, C.J. Koslozky, M.F. Wolfson, CT. Rauch, D.P. Cerretti, R.J. Paxton, CJ. March and R.A. Black. 1998. An essential role for ectodomain shedding in mammalian development. Science 282(5392), 1281-1284.

Price, J.V., E.D. Savenye, D. Lum, and A. Breitkreutz. 1997. Dominant enhancers of Egfr in Drosophila melanogaster: genetic links between the Notch and Egfr signaling pathways. Genetics 147,1149-1153.

Reneker, L.W., D.W. Silversides, K. Patel, and P.A. Overbeek. 1996. TGFoc can act as a chemoattractant to perioptic mesenchymal cells in developing mouse eyes. Development 121, 1669-1680.

Ricardo, N.S. and J.R. Miller. 1967. Further observations on lgMl (lid-gap Miller) and other open-eye mutants in the house mouse. The Canadian Journal of Genetics and Cytology 9, 596-605.

Rohlf, F.J. and R.R. Sokal. 1969. Statistical Tables. Freeman and Co., NY, USA.

Sadler, T.W. 1995. Langman's Medical Embryology, 7th Ed. Williams and Wilkins Publishers, pp. 132-230.

Saiki, R.K., S. Scharf, F. Faloona, K.B. Mullis, G.T. Horn, H.A. Erlich, and N. Arnheim. 1985. Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230,1350-

137 1354.

Saiki, R.K., D.H. Gelfand, S. Stoffel, S.J. Scharf, R. Higuchi, G.T. Horn, K.B. Mullis, and H.A. Erlich. 1988. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239, 487-494.

Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual, 2nd Ed. Cold Springs Harbour Laboratory Press, USA.

Schimmang, T., S.I. Oda, and U. Ruther. 1994. The mouse mutant Polydactyly Nagoya (Pdri) defines a novel allele of the zinc finger gene GU3. Mammalian Genome 5, 384-386.

Schrewe, H., M. Gendron-Maguire, M.L. Harbison, and T. Gridley. 1994. Mice

homozygous for a null mutation of activin (3B are viable and fertile. Mechanisms of Development 47,43-51.

Schwartzberg, P.L., E.J. Robertson, and S.P. Goff. 1990. Targeted gene disruption of the endogenous c-abl locus by homologous recombination with DNA encoding a selectable fusion protein. The Proceedings of the National Academy of Science 87,3210-3214.

Schwartzberg, P.L., A.M. Stall, J.D. Hardin, K.S. Bowdish, T. Humaran, S. Boast, M.L. Harbison, E.J. Robertson, and S.P. Goff. 1991. Mice homozygous for the ablml mutation show poor viability and depletion of selected B and T cell populations. Cell 65, 1165-1175.

Searle, A.G. 1965. Sightless. Mouse News Letter 33,29.

Semina, E.V., R.S. Reiter, and J.C Murray. 1998. A new human homeobox gene OG12X is a member of the most conserved homeobox gene family and is expressed during heart development in mouse. Human Molecular Genetics 7(3), 415-422.

Serbedzija, G.N., M. Bronner-Fraser, and S.E. Fraser. 1992. Vital dye analysis of cranial neural crest cell migration in the mouse embryo. Development 116, 297-307.

Sevel, D. 1988. A reappraisal of the development of the eyelids. Eye 2,123-129.

Siblia, M. and E.F. Wagner. 1995. Strain-dependent epithelial defects in mice lacking the EGF receptor. Science 269,234-238.

Silver, J., J.B. Whitney, C. Kozak, G. Hollis, and I. Kirsch. 1985. Erbb is linked to the , alpha-globin locus on mouse chromosome 11. Molecular and Cellular Biology

138 5(7), 1784-1796.

Silver, L.M. 1995. Mouse Genetics: Concepts and Applications. Oxford University Press, Oxford.

Sokal, R.R. and F.J. Rohlf. 1981. Biometry: The principles and practices of statistics in biological research, 2nd Ed. Freeman and Co., NY, USA.

Staats, J. 1985. Standardized listing for inbred strains of mice: eighth listing. Cancer Research 45(3), 945-977.

Stadler, T.W. 1995. Longman's Medical Embryology, 7th Ed. Williams and Wilkins Publishing, Baltimore, USA, pp. 312-346.

Stein, K.F., B.E. Norris, and J. Mason. 1967. Development of an open eyelid mutant in Mus musculus. Developmental Biology 16,315-330.

Stepp, M.A. 1999. a9 and 08 integrin expression correlates with the merger of the developing mouse eyelids. Developmental Dynamics 214,216-228.

Sulik, K.K. and G.C. Schoenwolf. 1985. Highlights of craniofacial morphogenesis in mammalian embryos, as revealed by scanning electron microscopy. Scanning Electron Microscopy IV, 1735-1752

Teicher, L.S. and E.W. Caspari. 1978. The genetics of blind - a lethal factor in mice. The Journal of Heredity 69, 86-90.

Teramoto, S., S. Fujii, A. Yoshida, and Y. Shirasu. 1988. Morphological and genetic characteristics of the open-eyelid mutant spontaneously occurring in NC-strain mice. Experimental Animals 37(4), 455-462.

Threadgill, D.W., A.A. Dlugosz, L.A. Hansen, T. Tennebaum, U. Lichti, D. Yee, C. LaMatia, T. Mourton, K. Herrup, R.C. Harris, J.A. Barnard, S.H. Yuspa, R.J. Coffey, and T. Magnuson. 1995. Targeted disruption of mouse EGF receptor: effect of genetic background on mutant phenotype. Science 269, 230-234.

Todderud, G. and G. Carpenter. 1989. Epidermal growth factor: the receptor and its function. Biofactors 2(1), 11-15.

Tom, C, D.M. Juriloff, and M.J. Harris. 1991. Studies of the effect of retinoic acid on anterior neural tube closure in mice genetically liable to exencephaly. Teratology 43(1), 27-40.

Tybulewicz, V.L.J., C.E. Crawford, P.K. Jackson, R.T. Bronson, and R.C. Mulligan.

139 1991. Neonatal lethality and lymphopenia in mice with a homozygous disruption of the c-abl proto-oncogene. Cell 65,1153-1163.

Varnum, D.S. and S.C. Fox. 1981. Head blebs: a new mutation on chromosome 4 of the mouse. The Journal of Heredity 72,293.

Vassalli, A., M.M. Matzuk, H.A.R. Gardner, K-F. Lee, and R. Jaenisch. 1994.

Activin/inhibin (3B subunit gene disruption leads to defects in eyelid development and female reproduction. Genes and Development 8,414-427.

Watney, M.J. and J.R. Miller. 1964. Prevention of a genetically determined congenital eye anomaly in the mouse by the administration of cortisone during pregnancy. Nature 202(4936), 1029-1031.

WH/MIT: Whitehead Institute for Biomedical Research/MIT Center for Genome Research. World Wide Web (http://www-genome.wi.mit.edu)

Whiting, J. 1997. Craniofacial abnormalities induced by the ectopic expression of homeobox genes. Mutation Research 396, 97-112.

Wiley, L.M., E.D. Adamson, and E.C. Tsark. 1995. Epidermal growth factor receptor function in early mammalian development. BioEssays 17(10), 839-846.

Zar, J.H. 1984. Biostatistical analysis, 2nd Ed. Prentice-Hall Inc., Englewood Cliffs, USA.

140 Appendices

141 Appendix A: GP/Bc x CBA/J Fl open eyelid data

Dam Sire D.O.B. # normal # open eyelid # pinhole CBA/J 06 GP/Bc 1349 29xii96 8 0 0 JAX

GP/Bc 1344 CBA/J 03 12xi96 7 0 0 JAX 22xii96 9 0 0

GP/Bc 1343 CBA/J 02 15xi96 7 0 0 16xii96 2 0 0 6i97 10 0 0

Totals 43 0 0

% open eyelids = 0/43 = 0%

142 Appendix B: GP/Bc x CBA/J BC1 open eyelid data (i) GP/Bc dam data

Dam Sire D.O.B. # normal # open eyelid # pinhole

GP/Bc 1364 Fl 01 lvi97* 4 2 0 22vi97 4 1 0

GP/Bc 1365 Fl 02 2vi97 3 3 0 24vi97 2 1 0

GP/Bc 1370 Fl 13 12vi97** 4 0 24vii97 3 0

GP/Bc 1371 Fl 14 llvi97 4 0 15vii97 2 0

GP/Bc 1375 Fl 04 3vi97 9 0 0 23vi97 2 2 0

GP/Bc 1372 Fl 15 12vi97

Total 36 19

* = 2 dead, unscoreable ** = 1 dead, unscoreable

% open eyelids = 19/55 = 35%

143 Appendix C: GP/Bc x CBA/J BC1 open eyelid data (ii) GP/Bc sire data

Dam Sire D.O.B. # normal # open eyelid # pinhole

Fl 10 GP/Bc 1381 8vi98 4 7 0 27vi98 7 1 0

Fill GP/Bc 1381 3vi97 3 7 0 22vi97 0 7 0

Fl 09 GP/Bc 1380 4vi97 0 23vi97 0

Fl 06 GP/Bc 1379 2vi97

Fl 05 GP/Bc 1361 3vi97 3 4 0 22vi97 3 4 0

Total 32 48

% open eyelids = 48/80 = 60%

144 Appendix D: GP/Bc x CBA/J F2 open eyelid data

Dam Sire D.O.B. # normal # open eyelid # pinhole F131 F132 22vi97* 10 1 0 llvii97** 8 1 0 31vii97** 4 3 0 llix97* 7 0 1 23x97* 7 3 0 Fl 33 Fl 34 22vi97* 10 1 0 llvii97** 6 2 0 31vii97+ 8 0 0 21ix97* 11 1 0 Fl 23 Fl 24 18vi97* 6 3 0 8vii97+ 5 0 0 18viii97** 5 1 0 7ix97* 8 1 0 1x97* 5 1 0 21x97* 4 1 1 10xi97* 5 0 0 Fl 18 Fl 19 20vi97* 7 1 1 10vii97** 4 2 0 Fl 16 Fl 17 6vii97* 5 2 0 26vii97** 2 2 0 15viii97+ 2 0 1 Fl 25 Fl 26 19vi97* 9 2 0 9vii97* 6 1 1 29vii97** 1 1 0 19viii97+ 4 0 1 1x97* 4 1 0 22x97* 4 1 0 Fl 29 Fl 30 23vi97* 9 2 0 14vii97** 9 1 0 lviii97** 7 2 0 13ix97* 11 2 0 29x97* 9 2 0

Total 202 41 6

* = whole litter used for mapping ** = only open eyelids used for mapping = not used for mapping

% open eyelids = 41/249 = 16.5% (if include pinholes = 18.9%)

145 Appendix E: Chromosome 13 and 6 markers screened in GP/Bc x CBA/J cross - includes PCR conditions, and reported allele sizes between the strains (= indicates that markers were not informative between strains, but allele size is still reported).

Alleles Size (bp) Marker fMgl (mM) Tanneal CO GP/Bc CBA/J

D13Mit31 3.5 55 no PCR product D13Mit53 1.5 55 no PCR product D13Mit69 1.5 55 210 200 D13MU76 1.5 hot start 55 = 98 = D13Mit77 1.5 55 280 = D13Mitl07 1.5 55 200 = D13Mitl47 1.5 55 94 104 D13Mitl96 1.5 55 = 140 =

D6Mitl6* 1.5 55 160 150 D6Mitl02* 2.5 55 150 125 D6Mitl32 1.5 55 200 D6MM49* 1.5 55 200 2 H) D6Mit261 1.5 hot start 58 130 110

* Markers also typed by Diana Mah

146 Appendix F: Chromosome 11 markers screened in GP/Bc x CBA/J cross - includes PCR conditions, and reported allele sizes between the strains (= indicates that markers were not informative between strains, but allele size is still reported). Alleles Size (bp) Marker IMel (mM) Tanneal CO GP/Bc CBA/J

DllMcgl 1.5 55 = 280 = DllMitl* 1.5 hot start 55 = 153 = DllMitl.l 1.5 55 = 130 = DllMit2* 2.5 60 = 110 = DllMit4* 2.5 62 275 250 DllMitlO* 1.5 55 100 132 DllMitl6 1.5 55 = 120 = DllMitl6.1 1.5 55 110 = DllMitl9 2.5 55 150 160 DllMit20* 1.5 55 120 150 DllMit23** 2.5 58 = 120 = DllMit26 1.5 55 190 180 DllMit38* 3.5 55 110 150 DllMit41** 2.5 55 No PCR product DllMit62 1.5 50 170 140 DllMit71 1.5 55 = 205 = DllMit72+ 1.5 55 160 DllMit73.1 1.5 60 = 120 = DllMit74 1.5 55 = 210 = DllMit75 2.5 55 = 130 DllMit77** 1.5 55 = 158 = DllMit78 1.5 55 = 85 = DllMit80 1.5 hot start 55 190 175 DllMitl04** 2.5 55 = 156 = DllMitl06 1.5 55 = 140 = DllMitl26**A 1.5 55 194 = 196 DllMitl29 1.5 55 = 140 = DllMitH^ 1.5 55 145 135 DllMitl50 1.5 55 = 190 = DllMitl52 1.5 55 140 150 DllMitl62 1.5 55 = 150 = DllMit226 1.5 55 150 130 DllMit278** 1.5 60 = 117 = DllMit340 1.5 55 130 140 DllMit341 1.5 55 90 80 DllMit370 2.5 55 = 50 =

* Markers also typed by Diana Mah ** Markers typed by Diana Mah only A Alleles very close together, reported as = AA Marker mapped to Chr 9

147 Appendix G: Chromosome 11 markers screened in GP/Bc x ICR/Be cross - includes PCR conditions, and reported allele sizes between strains (= indicates that markers were not informative between strains, but allele size is still reported

Marker [Mel (mM) Tanneal CO Allele Size (bp) Gp/Bc ICR/Be

DllMcgl 1.5 55 280 285 DllMitl 1.5 55 162 DllMitl.l 1.5 55 130 DllMit2 2.5 60 130 150 DllMitl6 2.5 55 = 120 = DllMitl6.1 1.5 55 = 110 DllMitl9 2.5 55 = 140 DllMit62 1.5 ' 50 170 ISO DllMit71 1.5 55 200 210 DllMit72* 1.5 55 160 162 DllMit73.1 1.5 60 120 DllMit74 1.5 55 210 230 DllMit80 1.5 hot start 55 190 180 DllMitl 06 1.5 55 = 140 = DllMitl29 1.5 55 = 140 = DllMitl49** 1.5 55 = 150 = DllMitl50 1.5 55 = 195 - DllMitl51 1.5 55 150 160 DllMitl52 1.5 55 150 152 DllMitl62 1.5 55 150 120 DllMit226 1.5 55 150 140

*bands are very close - marker was typed on polyacrylamide ** marker has been mapped to Chr 9

148 Appendix H: Chromosome 11 markers screened in (Egfr~/BXA-2) Fix SWV/Bc cross - includes PCR conditions, and reported allele sizes between the strains (= indicates that markers were not informative, but allele size is still reported)

Allele Size (bp) Marker fMel (m M) Tanneal('C) BXA-2 SWV/Bc (EGFR +/- x BXA-2 )

DllMitl 1.5 55 = 160 = 160 160 = DllMit2+ 2.5 55 110 105 & 115s = 110 = DllMitl6* 1.5 55 140 135 135 140 Dl lMitl9A 2.5 55 = 140 = 140 140 = DllMit62+*A 1.5 55 160 165 165 160 DllMit71+ 1.5 55 = 210 = 225 210 DllMit74+* 1.5 55 210 215 215 210 DllMit75 2.5 55 = 120 = 120 120 = DllMit77* 1.5 55 150 155 155 150 DllMit78* 1.5 55 105 85 85 105 DllMit80+A 1.5 55 = 170 = 170 170 = DllMit82 1.5 55 = 170 = 170 170 = DllMitl06* 1.5 55 135 130 130 135 DllMitl33 1.5 55 150 152 = 150 = DllMitl50 1.5 55 = 200 = 200 200 = DllMitl51+* 1.5 55 140 150 150 140 DllMitl52A 1.5 55 = 145 = 145 145 = DllMitl62+* 1.5 55 130 140 = 130 = DllMit226+*A 1.5 55 150 140 140 150 DllMit227 1.5 55 = 175 = 175 175 = DllMit228 1.5 55 150 155 = 150 = DllMit229 1.5 55 = 120 = 120 120 = DllMit306 1.5 55 = 110 = 110 110 = DllMit340A 1.5 55 150 = 150 150 =

* markers used in (Egfr'x BXA-2)F1 x SWV/Bc cross A markers typed in GP/Bc x CBA/J cross + markers typed in GP/Bc x ICR/Be cross > visualize two bands in SWV/Bc using this SSLP

149 Appendix I: BC1 pool data

BC1 pool sample # # individuals identification #s

1896 2 1894,1895 1912 3 1909,1910,1911 1913 3 1914,1915,1916 1919 2 1917,1918 1922 2 1920, 1921 1925 2 1923, 1924 1935 3 1932, 1933, 1934 1939 3 1936, 1937,1938 1948 3 1945, 1946, 1947 (n=23)

150 a CO co "o co co CU (U CU CU CU CU CO CU CU > co co o tu o o o cu co jo jo jo JO jo jo jo jo IB jo 3 jo 2 tu pi J

o o o o o o o o o o o o o ooooooooouooou o o o o o o o o o o o o o uoooouooouoouu

C/3 o a o o o o o o o o o o o o uoooooooouoooo o o o o o o o o o o o o o uoooouooouoouu CQ I-H o Q X o o CH o o o o o o o o o o o o o ooooooooouoooo o o o o o o o o o o o o o o uoooouooouoouu To3 ft SI OOOOOOOOOOOOOOOOOOOOOOOOUOOUU IT) ^1- ooooooooooooooouoooouooouoouu

vo CS ooooooooooooooooooooooooooooo oooooooooooooooooooouoooooooo

CN PH CSI ooooooooooooooooooooooooooooo oooooooooooooooooooouoooooooo ft 13 CQ 03 CQ ON O <—i tN m CN in vo ooc\vor>-mvDt^ooos © i—H CS m vo r> oo 1 00 OS Os © -H HH VO VO VO roror<->mmr<->r<-ir<">ro m cn t * t >n >n >n >n >n CS CS CS cs cs >» tu Cl CU ^TtTtvOVOO\Ov-cJ-CNOOt^t-~-VOVOC?\OS(N'ct'cJ- o r~ VO CS CS CN O O ft m i^u

ON VO VO CN X, ea li.ti.fcU,U.tt.UHU.U,lJ.fcU.U.^lLUHll.tt.fcIiLuti.LHlLli.U,li.[l.Ilu. . c a PH PH

r— t~~ t— t~~ t— t— r— r— t— r— i— t—r— t—- r— r~— r- t— c— t—- • r~- - *t—. t—* ft ffl Os Os , ff 0\ Os Os 0\ 0\ Os Os 0\ Os *^b\o\2lH22*£l-.H??'.Hl' * ' d > > > > > > > > :s :s > > ^> > •>= := := > > o > > > > m oo oo >>>>> O_O ^H0^HrSCSCSCSCSCSt^C-0 00 OS CS (N ro rn ~H Os RT RT OS OS O — — H M M CS m m ro

151 X) CO J3 CaO CU CD o o u o u o u u O O O O O U o o o u o u o o u o u o u o o o o u o a o o o o u o u u o o o u o u ooouououuoooouou ooouououuoooouou ooououooooooooou ooououooooooooou ooououooooooooou ooououooooooooou

• i i i i i i i i i i i i i i i sot^r~r~r~ooo©rH,^-HtNcNCNroco

"t VD CN CN tN O O CN CN ro co co ro ro

U.fcU.lJ.tl.U.fcUHtHfcU.Ii.li,(i,M.Ii. rororncjsOscoroinro ro co — —i — Os Os CNCNrOCNCNrOCNCNCN CN CN CO CO CO CN CN

Il,ll.UHU-ll,U.ll.Ll.tt,fcti.U.LLU.liIl.

t- r- r- r- r- ON os os os os t-t t- t-~ r- r- r- c- r~ o\ Os Os Os Os Os Os Os X X X X X X X > S .£ .£ x x * CN "t Os Os 00 CN CN CN CN

152 c CO PH O <3\ aCU > in o epppppcpoo cd ccj ccj cej "c3 e J ooooooooo Jc3u § § § § § § 3 * c c c c p c c a OOOO a. e a c a cO r-- oo vo a u a o u a a i i i i i i i i

ouuuooouo uouuooououu o u u u ouuuooouo uouuooououu a u o a

C/l 001 C/3 u

PQ O X •*H ouuuuouuu uouuooououu OOOO o ouuooouuu uouuooooouu OOOO 69 5 PH o «N| a ouuooouuu uouuooooouu OOOO ouuooouuu OOOO o uouuooooouu -a c o 03 03 03 CP 03 00 *r> vo r-~ oo 'ti co vo r~ -d-mvor-ooos©—• i i VO vo vo vo OO o — mcncncncncncnmcnvov>/nuni/n'/n'/->i/nvovotNcon cn cn m m CD E cn cn cn cn cn cn CNCNCNCNCNCNCNCNCNmm CN CN CN CN )H CNCNCNCNCNCNcnmcn CN CN tSO «3 CN CN CN

^ B cr, 00 cn CN CN a £ Q u. r- -3 O, Ov Ov c « b '>oo o> '> » o o IS CN ^ .OH CU

153 a ca cj a> Q cacacacacacacdcacaca ocflcpcepccci 8 8 g o o o "S. oooooooooo OOOOOOOOOO c e c ccsceeccccc cccccceccc 2 2 §

auuuouuuuuu o o u • 8 • aoaoaouoooo a a o

u o u a ouuuuuuuuua OUUUOUUUUOU o a u u o o a ouoououoaoa aooooououoo o o o

a Ouuuauuouuu a a u o ooaoaouauao a a o

ooouuouuuuo a u u aooaoououoa u o o

u u u o a uoouuouooou ououuuuuuuu a a u u u o o o uooououooou oaouoououoo a o o

u u u a a oooouuuuooo oouuouuuuuu a o u u u a a a ooaouauaoaa aoauoauauoo a o o

03 03 oo CN 03 03 Os © —i PQ os i I vo r- as © —< co «-> os vo r- r- CN CO t--r--t--t-~t--oooooooooo >/0 ©OOOOOOOOsOSOsOsOsOs vo _ co co co f- cocococococococococo oo cocococococococococo os os OcSo N cN M cn n co CNCNCNCNCNCNCNCNCNCN CO CNCNCNCNCNCNCNCNCNCN CO CO CN CN CN (N CN CN CN

CN © CO CO CO

PH PH

co OS CO CN

OS OS OS '> •> '> tN CN CO CN CN CN

154 CO CO CO ca CO CO CO CO CD i> CO CO CO CO C3 "o "w, "co "co 2> co co co co co co o a a a a a a O o o o o o o o ^ ^5 o o o o o •a o o o o o o no n C c a c c c c c •£> -° c a a a c no n no n no n no n -° a a a a c a

UUUUOOOO u u o u u u o U u u u u u o o o o o u u o u oouuoouo u u o o u u o uuuuuuuo U a u o u u u u u u o u u u u u u u u u u oouuoouo a u o o u u o u a a o o u a u o u a a a a

uouuuuuu u u u o u u u a u o o u o u u oouuoouo u u o o u u o a o o o o o u u oouuoouo u u u o u u u uuuuuoou ooouoooo a u o o u u o ooooooou u uuuuuuuu u u u u u u a uuuuuuuu u u u u u u u ououoooo a u a u a u a uuooooou u o u o a o u uuuuuuuo u u o u u u u uuuoooou u u u o u u u ououoaoo u u o u o u u uuooooou u o u o o o u

CO CO CO CO CQ © I—-1 tN cn VO • 1 1—1 fN m 1 1 00 Ov © CN m • CN rn CO © © © © O © © © 00 Os 1-H I—H rH vo r- rH rH CN CN CN CN © r~- r- r- r- © © 'T •o un tN CN fN tN tN tN tN tN tN tN tN fN tN tN fN fN tN fN tN i/n CN CN CN CN CN CN tN fN tN fN (N

VO CN CN

vo IT) CN CN

PH

C- Os Os Os X > > OS

155 cd cd cd cd ca ca1 1 O O o o o o o o o o o o e 2 § § a a a a c a a c a a a o S. -3

u o o u u u u u u u u uoououououuu a a o u u u a a o uaoaouauoouu

cj cj u a

u a o u a

uuuuuuou uauuuuououuu a a ouuuoaoo uouoauauoouu

u u uuuuouoa uouuuuuuuuuu o a ouuuooou uouoououooou

CQ 03 03 f- oo r^oooo—'CNcoi i i f- 00 ON © <— (N I<1 •* VI >C p-t— *ovovor--r— t-r-Tj- fN M CNCNeNCNCNCNCNr- t^r-~CNCNCNCNCNCNCNCN(N(N rN CN CN o CN ro P- Os CN PH

ON ON ,x CN

156 "o H T3 CD CD CD CD CD CD CD CD CD ID CD CD CD

i e (sm . h IS IS IS IS IS IS IS IS o S J J

Sssooooooaooaoooooooocjaaao QS-

SSoiOOODOOOODOOOOOOOOOOOOOO Q2°°IOOOOOOOOOOOOOOUOOOUOOOO

^.-SVOQOOOOOOOOOOOOOUOOO o o a a aSSoooooooooooooouooo a a a a

ooooooooooooooooooooooo Oin = 3 oooooooooooooouoooooooo oI-I a 2 o SlalSI^ OOOOOODOOOODOUOOOOOOOO olaSlo oooooooooooooooooooooo O HH

X HigN|0 OOOOOOOOOOOOOOUOOOOOOO o g Sr'o ooooooooooooouoooooooo £9 Hig-iO O OOOOOOOOODOOOOOOOOOOO o t l g SI~ oooooooooooooouoooooooo Tj tU CH •*i ooooooooooooooooooooooo Sis ^ooooooooooooooooooooooo

S ooo^invoh^m^o-^iovoaionrfvivohooaI I I I I I I I I I I I I I I I I I I I I I i I 1 ««MHHHrN|CAOOOHHHMrnttttttt Ooo^^HrtN^nrorocnnritnrf)t

ooooooooooooooooooooooo vA c MH M^H PH PH PH PH PH PH tin PH PH PH PH PH PH PH (D H•-(fl(flm((l(slnlnlnlo^^^HH|^^^^^HH O ooooooooooooooooooooooo

[XpH^^U-P-P-U-P-P-P-P-P-P-P-ti-UHClHtlHtHpHpHtl.

TJ uc CH : ; ; CH l l l I S § 5 5 5 5 § £ £ <

157 Appendix M: Allele segregation data from GP/Bc x ICR/Be cross for 6 complete litters (n=65: 58 normals and 7 open eyelid litter mates)

p.O.B Dam Sire sample DllMit74 phenotype

15xii97 Fl 101 Fl 102 3068-B GG bil. oe 3069- B GG L. oe 3070- B GG normal 3071- B GI normal 3072- B GI normal 3073- B GG normal 3074- B 11 normal 3075- B GG normal 3076- B GG normal 3077- B GI normal 3078- B 11 normal 3079- B GG normal

25i98 Fl 103 Fl 104 3084-B GG bil. oe 3085- B GI normal 3086- B 11 normal 3087- B GI normal 3088- B GI normal 3089- B GI normal 3090- B GI normal 3091- B GI normal 3092- B GI normal

25i98 F1105 Fl 106 3093-B GG normal 3094- B GG normal 3095- B GI normal 3096- B 11 normal 3097- B GI normal 3098- B 11 normal 3099- B GI normal 3100- B GG normal

25i98 Fl 107 Fl 108 3101-B GG normal 3102- B GI normal 3103- B GI normal 3104- B GI normal 3105- B GG normal 3106- B GI normal 3107- B GG normal 3108- B GI normal 3109- B II normal 3110- B II normal

158 14ii98 Fl 103 Fl 104 3111- B GI normal 3112- B II normal 3113- B GI normal 3114- B GG normal 3115- B GG bil. oe 3116- B GG bil.oe 3117- B GG bil.oe 3118- B GI normal 3119- B GI normal 3120- B GG normal 3121- B II normal 3122- B II normal 3123- B GI normal

12H98 Fl 107 Fl 108 3124- B GG bil.oe 3125- B II normal 3126- B GG normal 3127- B GI normal 3128- B GI normal 3129- B GI normal 3130- B GI normal 3131- B II normal 3132- B GI normal 3133- B II normal 3134- B GI normal 3135- B GI normal 3136- B GI normal

159 cu CU 2 CU CU o <3\ n o cu 1 e o o •S i-J os 15 IS OH B I ° -3 cu JS 3.-3 IS OH J

03 CO to OOUUUUUUUUUUOOOOUUUUUOUUUUOOOU o 2 OOOOUOOUDUOUOOOUOUOOUOOOOOOOOO o ON Q

PQ u X uouuouuuuuuuuauouuauuouuuuuuao cu 2 ouuoouuuouuuouoououooooouooouu S9 PH o -a

CO

a mcaoacacQmmmoammmcQcaaamcammmm • 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ^OHNmNmcfin^hecoMOf-iniOMio^OHncnt^nor-ooo 73 TfioviioiOhhhM(>o\ooHHinifiwiin»no>o>o^voxiio»o>o'o CN £

"o

E, cu MtNNNNHnnnnc'iHHtNtNHHtfinnnHHNmmmcfic'iN T3 In t^H hV. E^H PH PH PH PH PH PH PH PH PH PH PH tin tin C^. HH t^H PH PH tin tin tin HVH HVH HVH V^H VH

1) >N nnnin/iooMcioioifiHHHCji(j\rr) a> tNNNNNHHcnmNNHHNNHHnmmNHHNrfinnNcNN CS CD Q t^H HH VV. fi. pH E^H P^H E^H E^H fi. E^H E^H E^H E^H pH H^H E^H HH HH HH PH pH PH PH HH PH PH E^H h^H E^H OH o

03 0\0\O\O\0\Q^P^O\O\p\p\0\ ON ON ON ON ON ON ON ON S •> •> •> •> •> •> •> •> •> " O : : : OOOOOOONONOOCNCNCO : > > 'C 3 3 '> '> '> "> '> "> "> "> '> "> '> "> j 5 •> 00 1 Q ^Hrt^^HrtfNjfNlfNjfNjfSirNj f~ r~- ON ON cr, 00 00 ON -H ,C > > CD ^r.C C XXOO-HrHrHinOOOOON—1 > ,5 OH . OH ~

ON a

160 X> CD XI cd O. ID CD o CD CD CD CD CD £ CD S « o o O O O O O o inho l o ai q o x> X lo HJ IS "S HJ IXS> X) X) o X X bil . L . o e

UUUUUUUOOUUOOOO OOUODOOOOOOOOOO

ouuauoauucjuooou oouoooouoooouoo

ouuouuuaoououuu oouoooouuuocjooo

P3ffl(afflfflfflfflP3CPfflCPD3D3fflD3 I I I I I I I mr-t--r~t^oooooooooooooooooooo

t(NOOt'*>0't^lO(NNMOO

rO'HOvOicnmvifcicciiOHrHrHaiO

tt.U.U.U^tl.lLtl.&.fcll.U.U.U.U.U.

r~- ON ON ON ON ONONCNONONONONON X .Xs X X X0 \ ON X x x CN XXX —< CN

161 o

00

00 CO W W l/l M (« 00 (O W [/l M 1/1 7) 1/1 1/1 I/)

2 § "I I I I I I I I I I I I I I III

I I I I I I I I I I I I I I I I I I I I I I I 6C o ^ ^ j 3 N i i i • i i • i I I I I I I I I I I I I fi ^ N t/3 2 o I I I I I I I I I I I I I I I I I I I III -t-» 5 11 SI

• rH O ft r-H VO M O I I I I I I I I I I I I I I I I I I I Q « -H ?9

^2 £= ^1 oo oo c>o c>o cxi cyo CJO c>o oor/Dc/Dooczioo c/3 co oo oo ggr-IWWfflCQfflW(0W on w w w w c»o w fflffl CQ U U W CQCQW PJ CQ CD UJ CO UJ UJ

CD OHNcnt^l'OI-' ON O i—i CN co h OO ft O H CN -H -H -H m rf iri OOOOOOOO o o o o o o o CN CN CN oooooooo o o o o o o CN CN CN ctj I I I I I I I I o o o t/3 A HH MM HH HH HH HH HH

rf rf CN cu O O O 13 La P-) PJ Pd ft

o CN CD >r rf rf rr 53 130 00 00 o o rf rr rf

in O 3 rf' s ° 13 o T3 o OH •3 e o I OH o

162 moawunammumu cQDaDacnumcDcautu Dammpqwmupapgwmm

vot~~ooo.O«tNrn-^->r)vor-~oo oooooooooo oooooooooo ooooooooooooo CQ CQ CQ CQ CQ CQ CQ CQ CQ CQ CQ CQ CQ CQ CQ i4 r-^ ^ ^ ^ ^ ^ 3

o o W

oo 00

T3 •a U o n.

163 CO CD cQcowwmcQwww UMUUOIIIUISOOUUU COWPJWWCQWCQ

i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i

oo i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i

i i i i i i i i i i i i i i i i i i i i i i i i i i i

i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i

i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i

tU Ui CQCQWUCQD3WWW WWWWfflfflWCQDHCQWWW CQWWWWfflWffl

Os © in vo o o ooooooooo ooooooooooooo oooooooo i i i i i i i i i i i i i i COi i i i i i i i i i coi qi coi i i ^ r^H r^H r^H r^H r^H r^H ^ 2 ^

-3- tN o o O Ui u tQ

m tN o m IT) 00 oo oo

13 T3 C o O -O O. e

164 wwrnoaww WWPJOHOQWWWCQWW

T3 I I I I I I I I I I I I I I I I I o i

T3 Cl

I I I I I I I I I I I I I I I I I

*e-»-3» I I I I I I I I I I I I I I I I I 'S o -*-» C/o3 CM

I I I I I I I I I I I I I I I I I O a a PJ U CQ CQ U tU UtqWfflfflUWUfflBW o CD

CN m rf in vo r— ooftOHHCNrorfinvot--oo O » Oi 0\ ft O ftftOOOOOOOOO OOOOOO OO — — — — — — — I I I I I I I I I I I I I I I I CQ CQ CQ CQ CQ CQ CQ CQ CQ CQ S> >S> >S> K> K> HH H^ HH HH HH HH HH

Tf o W

-a u

165 Appendix P: SSLP marker loci not tested for informativeness in the GP/Bc x CBA/J or GP/Bc x ICR/Be crosses, between 1-15 cM on Research Genetics/MIT and 1999 Chromosome Committee Map (Mouse Genome Informatics). * not mapped on Research Genetics map

DllMit73 DllMitl48 DllMit304 DllMit63 DllMit76 DllMit204 DllMit305 DllMit259 DllMit79 DllMitl07 DllMitl70 DllMitl71 DllMit295 DllMit343 DllMitl85 DllMit294 DllMit81 DllMitl69

DllMit371*

166