Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Curly : a new hair defect mutation in the SELH/bc mouse strain Taylor, Lydia Anne 1999

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

Item Metadata

Download

Media
831-ubc_1999-0295.pdf [ 21.4MB ]
Metadata
JSON: 831-1.0089024.json
JSON-LD: 831-1.0089024-ld.json
RDF/XML (Pretty): 831-1.0089024-rdf.xml
RDF/JSON: 831-1.0089024-rdf.json
Turtle: 831-1.0089024-turtle.txt
N-Triples: 831-1.0089024-rdf-ntriples.txt
Original Record: 831-1.0089024-source.json
Full Text
831-1.0089024-fulltext.txt
Citation
831-1.0089024.ris

Full Text

CURLY: A NEW HAIR DEFECT MUTATION IN THE SELH/Bc MOUSE STRAIN by LYDIA ANNE TAYLOR B.Sc, The University of British Columbia, 1996 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDD2S (Genetics Graduate Program) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA April 1999 © Lydia Anne Taylor, 1999 U B C Special Collections - Thesis Authorisation Form Page 1 of 1 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y a v a i l a b l e f o r reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. I t i s understood that copying or p u b l i c a t i o n of t h i s thesis f o r f i n a n c i a l gain s h a l l not be allowed without my written permission. of The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada http://www.library.ubc.ca/spcoll/thesauth.html 4/27/99 11 A B S T R A C T The curly mutation (cur) is an autosomal recessive mutation that arose spontaneously on the inbred strain SELH/Bc. This study included the characterization of the curly phenotype and the mapping of the curly mutation in order to report this mutation, evaluate its importance as an animal model for human genetic disorders and possibly contribute towards understanding the mechanism of mutation in the SELH strain. This mutation is one of the eight single locus, spontaneous mutations that have arisen on the SELH strain indicating that this strain has an elevated mutation rate. This mutant has a waved and ruffled coat, crimped and broken vibrissae, and in later life may develop curled toe nails and hair loss at the eyes, nose, ears and tail tips. The curly mutation was mapped to the distal end of chromosome 11 via PCR amplification of informative SSLP's using 32 homozygous curly F2's from curly S E L H stock outcrossed to the normal inbred strain LM/Bc . Based on recombination breakpoints in individual affected F2 genotypes, the region containing the curly mutation was narrowed to between DI l M i t l 4 and DI IMitIO, an approximately 6 c M interval. This placed the curly mutation in a region of the genome known to contain several dominant hair defect mutations, epidermal regulatory genes and keratin (the main structural protein of epidermal, nail and hair cells) genes. A second outcross of curly SELH to the AXB-10/Pgn normal inbred strain was undertaken to further refine this region containing the curly mutation and to map curly relative to the keratin genes. Based on recombination breakpoints in individual affected F2s (n = 71) in this cross, the curly mutation was mapped between DI l M i t l 4 and DI lMit360, an approximately 4 c M region. No recombinants were found between the curly mutation and two markers from within candidate keratin genes, KHA1 and Krtl9. I l l TABLE OF CONTENTS Abstract ii Table of Contents i i i List of Tables v List of Figures vi List of Appendix Tables and Figures viii Acknowledgments x CHAPTER I: INTRODUCTION Historical Introduction 1 S E L H and curly origin and history 1 Review of Hair and Hair Defect Mutations Hair 4 Hair defect mutations 9 Rex 11 Keratins 16 Mouse Mapping Overview and history 24 PCR and SSLPs 26 Review of the state of the mouse map 28 Rationale and approach to this study 31 CHAPTER II: GENERAL MATERIALS AND METHODS SELH curly: scientific progress before this study 35 Mouse stocks and maintenance 37 Technical Methods 37 a) D N A preparation b) PCR of SSLPs c) Source of SSLPs d) Visualization of PCR products e) Histology CHAPTER IH: PHENOTYPE AND GENETIC TRANSMISION OF CURLY Introduction 40 Rationale, Materials and Approach 40 Gross observations of phenotype 40 Genetic transmission 41 a) SELH background b) Segregating background Segregation and sex effects 46 Hair 47 Histological analysis 48 Litter raising capability 48 Results I V a) The normal phenotype (+/+ and +lcur) 49 b) The curly phenotype (cur/cur) 53 c) Comparisons of curly vs. normal 69 d) The AXB-10 normal phenotype 69 Genetic transmission 71 a) SELH background b) Segregating background Segregation and sex effects 72 Hair 73 Histological analysis 75 a) Cheek/vibrissae follicle in cross-section b) Cheek/vibrissae follicle in transverse-section (at the surface) c) Ear pinnae in cross-section d) Tail in transverse-section e) Dorsal skin in cross-section Litter raising capability 81 CHAPTER IV: MAPPING CURLY Introduction 82 Rationale, Materials and Approach 82 Experimental design 82 L M x S E L H cross 86 a) Pooled sample genome screen b) Genotype analysis of individuals A X B - l O x S E L H cross 89 a) Scoring F2s b) Genotype analysis of individuals Molecular Investigations 90 a) SSLP deletion survey b) D l l M i t l 2 3 double band investigation c) Chromosome 11 map improvement d) Amplification of keratin fragment Results L M x S E L H cross 96 a) Pooled sample genome screen b) Genotype analysis of individuals A X B - l O x S E L H cross 99 a) Scoring F2s b) Genotype analysis of individuals Molecular Investigations 102 a) SSLP deletion survey b) DI lMit l23 double band investigation c) Chromosome 11 map improvement d) Amplification of keratin fragment CHAPTER V: DISCUSSION 108 References 126 Appendix 136 V L I S T O F T A B L E S Table 1 Literature survey of known mouse hair defect mutations without 10 alopecia (baldness): their chromosomal location, mode of inheritance and main phenotypic difference in comparison to the curly phenotype. Table 2 Type I keratin intermediate filaments and keratin associated protein 19-20 genes linked to the type I keratin gene cluster on mouse chromosome 11: their type II partners, areas of expression and associated disease phenotype in mouse and human. Table 3 Data summary of F2 L M x S E L H progeny: parents, phenotypic 36 proportions, sex ratios and animal numbers on the basis of individual litters. . Table 4 Normal animals examined for description of the normal external 50 phenotype. Table 5 Summary of 32 SELH cur/cur animals older than one month of age 63 used to define the curly phenotype: proportions of animals affected with each of the various traits of the curly phenotype. Table 6.1 SELH cur/cur animals with ages ranging from 1 to 3 months, used 66 to define the curly phenotype: a detailed description of the traits affecting individual animals, their age, sex and animal numbers. Table 6.2 S E L H cur/cur animals with ages ranging from 4 to 10 months, used 67 to define the curly phenotype: a detailed description of the traits affecting individual animals, their age, sex and animal numbers. Table 7 The primer pairs and chromosomes scanned, in chronological order, 97 in the L M x S E L H F2 pooled samples. Table 8 SSLP deletion survey. 103 Table 9 Strain survey for DI lMit!23 double band investigation. 104 vi LIST OF FIGURES Figure 1 A diagrammatic representation of (a) the organization of a normal 5 mouse hair demonstrating the relative positions of the cuticle, cortex and medulla and (b) the four main hair types of the normal mouse pelage. Figure 2 A Rex pink dilution adult female demonstrating the rough coat and 12 curled vibrissae typical to the Rex phenotype (Crew and Auerbach, 1939). Figure 3 The adult Rex coat seen in profile demonstrating the abnormal, short 12 and irregular character of the guard hair (Crew and Auerbach, 1939). Figure 4 Photographs of SELH normal coat and guard hair at a) 3 weeks (+/cur), 51 b) 2.5 months (+/cur) and c) 3 weeks i+lcur), magnified by a light microscope (x63). Figure 5 SELH normal whisker phenotype at a) 8.5 months in a normal 52 homozygote (+/+), b) 4 months in a curly heterozygote (+lcur) and c) 3 weeks in a curly heterozygote (+lcur). Figure 6 Photographic comparison of whisker phenotypes of 3 day old (a) cur/cur 54 and (b) +lcur littermates. (c) An overexposed blow-up of the cur/cur whiskers for better visualization of the abnormal whisker phenotype. Figure 7 Photographs of cur/cur guard hair at three weeks of age in comparison 55 to a normal littermate. Figure 8 Photographs of cur/cur and +lcur SELH guard hair magnified by a light 56 microscope (x63). Figure 9 Whisker comparison of 3 week old SELH cur/cur and +lcur animals. 57 Figure 10 Coat character of SELH normal and curly littermates at (a) three weeks 58 and (b) 2.5 months of age. Figure 11 SELH curly and normal ear pinnae demonstrating the tissue distortion 59 and curled hair on the cur/cur individual. Figure 12 S E L H curly and normal tails demonstrating (a) the uneven hair lengths 60 and distribution on the cur/cur tail and (b) loss of hair and tip of a cur/cur tail. Figure 13 SELH curly and normal hind feet demonstrating the elongated and 61 curled fifth digit nail of the cur/cur individual. Figure 14 SELH cur/cur adult whisker phenotype in comparison to (a) normal 62 and (b) cur/cur; (c) blow-up of (b) whiskers and (d) normal. vu Figure 15 Photographs of AXB-10 normal animals at 1 month of age demonstrating 70 (a) the non-straight vibrissae and (b) the appearance of the coat and guard hair. Figure 16 Photos demonstrating the vibrissae medullary cells of cur/cur and +/cur 74 littermates. Figure 17 Forepaw of SELH cur/cur animal demonstrating the non-straight 76 character of the foot hair. Figure 18a Histology slides of the whisker follicles of normal and curly littermates 78 in transverse section. Figure 18b Histology slides of dorsal backskin from normal and curly animals 79 demonstrating the epidermis and hair follicles. Figure 19 Research Genetics map of the cur mutation region of distal chromosome 92 11. Figure 20 A diagram comparing three maps of chromosome 11 in the region 95 containing the cur mutation including the M G D , EUCIB and Research Genetics maps. Figure 21 Illustration of the chromosome 11 marker order and distances as 97 outlined by Research Genetics. Figure 22 Diagrammatic representation of heterozygotic breakpoints: fine mapping 99 of cur mutation using genotypes of L M x S E L H F2 cur/cur animals. Figure 23 Diagrammatic representation of heterozygotic breakpoints: fine mapping 101 of the cur locus using X10.S F2 cur/cur animals. Figure 24 Diagram comparing maps of the cur mutation region of chromosome 11 105 including the M G D , EUCIB and Research Genetics maps for comparison with the maps generated by the two crosses for mapping curly. Figure 25 Photographic comparison of homozygous SELH curly guard hair at (a) 110 2 months and (b) 3 weeks with (c) SELH normal (+/cur) 2.5 month old. Figure 26 Comparison of (a; Crew and Auerbach, 1939) the adult coat of Rex and 111 (b) a 3 week old SELH curly demonstrating the differing guard hair phenotype. Vlll LIST OF A P P E N D I X T A B L E S A N D F I G U R E S Table A SSLP primer pairs whose optimum conditions differed from Tanneai = 55°C and/or [Mg 2 +] = 1.5 mM. Table B SELH/Bc mice used for gross observations to determine the homozygous normal (+/+), heterozygous curly (+/cur) and homozygous curly (cur/cur) phenotypes. Table C AXB-10 animals used to determined the normal (wildtype) AXB-10 phenotype. Table D Research Genetic PrimerPairs SSLP loci chosen by D.Mah for initial scan of genome in search of curly locus. Loci that were actually used are marked with an asterisk. 136 136 138 138 Table E Table F Table G Table H Table I Table J Table K Primer pairs tested for informativeness between SELH and L M strains 139 for both pooled sample and individual F2 mapping of the curly mutation. L M x S E L H cur/cur F2 pooled samples: mouse numbers and information. 141 Three litters of L M x S E L H F2 litters including both cur/cur and normal 142 littermates. Genotyped at DI I M i t H to confirm equal segregation of alleles in this region. The genotypes of individual L M x S E L H cur/cur and normal F2 animals 143 used for genotype analysis of heterozygote breakpoints to map the curly mutation and to refine the chromosome 11 'Mit ' SSLP map. Primer pairs tested for informativeness in L M x S E L H cross for the possibility of finer mapping between DI I M i t H and DI IMitlO. X10.S F2 litters: mating pairs and phenotyping data. 145 146 Primer pairs that were tested for informative SSLP alleles between S E L H 147 and AXB-10 strains. Table L Information regarding PCR primer pair designed to amplify exon 1 of 147 KrtlO on chromosome 11. Table M The genotypes of individual X10.S cur/cur and normal F2 animals used 148 for genotype analysis of heterozygote breakpoints to map the curly mutation and to refine the chromosome 11 primer pair map. Table N Three litters of X10.S F2 litters including both cur/cur and normal littermates. Genotyped at D l l M i t l 2 3 to confirm equal segregation of alleles in this region. 150 ix Figure A PCR product obtained from PCR amplification of DI l M i t l 4 with 151 cur/cur X10.S F2 animals' (n = 71) DNA. A C K N O W L E D G E M E N T S I would like to thank my supervisors, Drs Diana Juriloff and Muriel Harris, whose vast knowledge and passion for science helped to fuel me during this work. I would also like to thank my colleagues and co-workers for the precious discussions and insights which helped to contribute to both this work and my continued quest for knowledge. In addition, I thank the members of my thesis committee for their thoughtful contributions. This work would not have been possible without the love and support of my family. Thanks to mom for paving the way towards all scholastic achievements and inspiring me to follow in her footsteps as a perpetual student both in academia and life. Thanks to dad for constantly supporting me in countless ways and for being the kind of person whose own achievements are the benchmark by which I measure my own successes. Thanks to my sister, Jillian, for constantly raising the standard of our mutual development. You are a remarkable woman who I hold in tremendous esteem. Thanks for being my antithesis and my friend. Lastly and most importantly, I would like to thank my husband, Juan, without whose counsel I would not be the person I am today. Your love, understanding, patience and wisdom are a constant beacon in my life, especially during this work. I love and admire you immeasurably. I couldn't have done this without you. 1 CHAPTER I: INTRODUCTION Historical Introduction SELH and curly origin and history The SELH/Bc stock was created by D . M . Juriloff and M.J . Harris and is unique to the JurilofrTHarris lab in the Department of Medical Genetics at the University of British Columbia. It was created from brother-sister breeding pairs with selection for the production of exencephaly in offspring (Juriloff et al., 1989) in the hopes of creating an animal model for human multifactorial anencephaly. The SELH/Bc strain originated with a pair of genetically heterogeneous mice that produced spontaneous exencephaly (Gunn et al., 1993). The resulting SELH/Bc strain is presently highly inbred (D.M. Juriloff and M.J. Harris, personal communication) based on continued inbreeding following an estimation that the inbreeding coefficient was higher than 90% in the early 1990's (Gunn et al., 1993). A detailed account of the genetic history of the SELH/Bc is outlined in Juriloff et al. (1989). While inbreeding, several other factors, when they arose, were also selected for in conjunction with exencephaly. These were ataxia, ovarian teratomas and the occurrence of spontaneous mutations (D.M. Juriloff and M.J. Harris, personal communication). Presently 10-20% of SELH/Bc newborns are exencephalic, 5% have ataxia (Harris et al., 1994), and up to 3% of females have ovarian teratomas (D.M. Juriloff and M.J. Harris, personal communication). In addition to these various defects, eight spontaneous mendelian recessive mutations have been isolated from the SELH/Bc strain. All of these mutations arose separately and homozygotes occur independently from the exencephaly, ataxia and ovarian teratoma traits. This occurrence of spontaneous mutations appears to be more frequent than normal. Only one mutation, cmdBc (Bell et al., 1986) arose in other concurrently maintained mouse stocks over the 2 same time period that the eight mutations arose in the SELH/Bc strain, which occupied only 25% of the animal unit (D.M. Juriloff and M.J. Harris, personal communication). Three of these mutations, which arose in not more than 20 000 mice (Juriloff et al., 1994), were at the albino (c) locus on chromosome 7 which translates into 3 mutations at the c locus in 40 000 gametes (7.5 x 10-5). This can be contrasted directly to the spontaneous mutation rate, determined at the c locus among 4 other coat colour loci, which appeared to be 1 x 10"5 per locus per gamete on average (Schlager and Dickie, 1966). It appears therefore, that SELH/Bc may have a higher than normal mutation rate. The eight mendelian recessive mutations that have arisen on the SELH/Bc strain occurred at different times during the creation of the SELH/Bc strain and are either on different chromosomes and loci or are entirely different lesions. The first of these spontaneous mutations, spherocytosis-British Columbia (sph 2Bc\ chromosome 1), arose in the second intercross generation (N3xN3) in the lineage of SELH/Bc. At F14, a second mutation arose which was a new recessive allele, nuBc, at the nude locus (whn; chromosome 11) (Juriloff et al., 1989). The lesion in this allele has been recently elucidated to be a 5537 bp insertion of early transposon (ETn) sequence located between exons lb and 2 in the whn locus (Hofmann et al., 1998). The third mutation was a recessive lethal mutation at the albino (tyrosinase; c; chromosome 7) locus, cBc, which is a deletion of at least 2 c M and was first observed in a somatic and germline mosaic male (Juriloff et al., 1992). Albino mutations are semi-dominant to the chinchilla, cch, allele with heterozygotes having a pale brown coat colour. This made the discovery of albino mutations obvious in SELH/Bc mice, which are uniformly homozygous for the cch allele. In particular, the cBc allele would otherwise not have been detected without looking for post-implantation lethality. The fourth and fifth mutations were also at the albino locus, c2Bc and c3Bc, and in contrast to cBc, are viable when homozygous. All three of these albino mutations occurred in a period of 10 months. c2Bc was first noticed in a somatic and germline mosaic female while c3Bc was transmitted by single breeding pair, one of which must have been a germline mosaic. The c2Bc allele has the first three exons of the tyrosinase gene deleted (Juriloff et al., 1994). The c3Bc allele has an early transposon (ETn) sequence of 5542 bp inserted into exon 1 of the tyrosinase gene (Hofmann et al., 1998). Prior to 1994, the sixth spontaneous mutation, causing a recessive lens rupture phenotype, arose and to date remains unmapped and uncharacterized. Between 1994 and 1996, the last two mutations arose, these being a recessive curly hair mutation and a recessive white belly spot mutation. During the summer of 1996, the recessive white belly spot was mapped to the kit locus on chromosome 5 (D.M. Juriloff and M.J Harris, personal communication). In the fall of 1996, the recessive curly hair mutation still remained unmapped and the phenotype of the affected animals remained uncharacterized, the completion of which became the focus of this study. The curly hair mutation was first observed in a single litter where several two-week-old animals were noted to have abnormal whiskers and a ruffled appearance to their coat. The parents of these curly-haired animals were then used to initiate a substrain segregating for the curly hair mutation called SELH/Bc curly. This substrain was maintained by brother-sister breeding between homozygous curly male animals and their obligate heterozygote female littermates (D.M. Juriloff and M.J . Harris, personal communication). Based on the segregation of the curly phenotype within the SELH/Bc curly colony, it was hypothesized that the curly trait was due to a fully penetrant mutation that was recessive to the wildtype allele at the curly locus (D.M. Juriloff and M.J. Harris, personal communication). 4 Review of Hair and Hair Defect Mutations Hair There are eight major hair types in the mouse based on location and morphology. These include the pelage or trunk hairs, vibrissae, cilia, tail hairs, ear hairs, and hairs around the feet, nipple and genital regions (Dry, 1926). Each of these 'islands' of hair types have their own characteristic ensemble of hair forms, the majority of which are overhairs except the hairs of the nipple regions which more closely resemble under- or zigzag hairs. Hair subtypes are generally classified by their length, by the amount of medullary region they contain and by the presence or absence of bends in the hair shaft (Dry, 1926). The medulla is composed of cells separated by air-filled spaces arranged in rectangular rows. The presence or absence of the medulla is dependent on the diameter of the hair shaft and generally increases with increasing fibre diameter. The medulla can either be absent, as is the case with thin hairs or at the tip of the hair shaft, or present with a number of cells of thickness (Gruneberg, 1969). Mouse hairs all have a similar morphology with the medulla, if present, surrounded by a cortex, which is composed of hard cornified material, surrounded by a cuticle which is made up of a series of thin overlapping scales whose edges point distally (Sundberg and Hogan, 1994; Trigg, 1972). See figure 1(a) for a representation of the arrangement of the hair shaft. In thin hairs lacking a medulla, the cortex can then occupy as much as 90% of the shaft volume (Powell and Rogers, 1990). An additional similarity of all mouse hairs is a concave trough that runs the length of the hair shaft (Sundberg and Hogan, 1994). The hair follicle consists of several cylindrical cell layers. The most exterior cell layer is the outer root sheath (ORS) which is composed of several cell layers and is continuous with the epidermis. Inside the ORS is the inner root sheath (IRS) which itself is composed of three layers. 5 a ed cd 8 00 he I— ts c of o C/5 a pe de t - i '3 '3 <U C OUS mai a Id e fou noi ed o on an JS N 3 gani med u o he i an X CU cd ort e « o O dT C o *o '3 3 o sen he eu -*-» C. c+-ft O CD ons CD 00 o ons ed '3 cd '•3 pel C O e CL, —• agra tive lOUI Cd 13 cd < ii C CD M noi CU &• S 6 These IRS cell layers are the Henle layer, Huxley layer and the IRS cuticle, which are oriented in that order moving inward. The IRS encircles the hair shaft and grows out with it (Rogers and Powell, 1994). The hair shaft composition, as described above, includes a medulla, cortex and cuticle. The cells of the cortex and cuticle contain a regular arrangement of filaments in an interfilamentous matrix, the proteins of which represent the main structural components of the hair shaft. These filaments are composed of pairs of type I and type II intermediate filament keratin proteins (KIFs) which are embedded in a cysteine-rich matrix of keratin associated proteins (KAPs). The main structural proteins of the medullary cells, in contrast, are trichohyalin and KIFs (Powell and Rogers, 1997). Trichohyalin is also found in combination with KIFs to make up the major structural proteins of the ERS while it is not found in the ORS whose main structural protein are KIFs (Rogers and Powell, 1994). Throughout the main dorsal pelage, the hairs emerge from the skin at a small angle pointing more or less caudally. The overhairs appear to have a small curvature running the length of the shaft and, at least for some mouse strains, the frequent presence of a much slighter bend at the tip (Dry, 1926). The hairs of the pelage are further divided into four subgroups: 3 of which are overhairs, the guard, awl, and auchene hairs, and one of which, the zigzag hair, is an underhair (Dry, 1926). Guard hairs and awls have shafts with no bends while auchene hairs have a single bend and zigzag hairs have more than one bend. Guard and awl hairs can be distinguished based on length with the awl hairs about half as long as the approximately 1 cm long guard hairs (Sundberg and Hogan, 1994). Zigzags have a medulla composed of only one row of cells while guard hairs have two and awl hairs have two or more rows of air spaces. Auchene hairs are similar to awl hairs with the exception of a single constriction at the distal two fifths of the shaft. It should be noted however, that some authors question that auchene hairs are 7 a distinct type and not just a form of awl hairs (Fraser, 1951). See figure 1(b) for a diagrammatic representation of these four pelage hair types. Each of these pelage subgroups are produced by follicles which are evenly distributed and usually produce a similar type in successive hair generations (Dry, 1926). Guard hair follicles however, while evenly distributed, are most highly concentrated on the caudal dorsum and least on the rostral dorsum (Mann, 1962). The follicles of the different hair subgroups are indistinguishable from each other by hematoxylin and eosin staining and can only be determined by examination of the hair shaft itself (Sundberg and Hogan, 1994). Al l vibrissae and pelage hair follicles are initiated in development prior to birth with vibrissae follicles initiating at El2.0-12.5 followed by the initiation of guard hair follicles at E l 3.25 which progresses caudally from the shoulder region and finishes at El6.25 when the awl follicles initiate. Awl follicle development is then complete at the initiation of the development of the auchene and zigzag hairs that are formed from E18.25-19.25. Hair in each of these follicles then subsequently erupts between 8-9 days after the onset of initiation. In contrast to the pelage and vibrissae hair follicles, the hair follicles of the remaining hair types, i.e. the cilia, tail hairs, ear hairs, and hairs around the feet, nipple and genital regions, are first observed between four to five days after birth on the caudal dorsum but their time of initiation is unknown. By ten days after birth these follicles are actively producing hair (Mann, 1962). All of the hairs of the coat are thought to be fully initiated by 10-13 days of age and fully grown by about 20 days of age (Fraser, 1951). Following the initiation times of the vibrissae and pelage hair types, their order of eruption from the epidermis is vibrissae followed by guard hairs, awls, and then auchenes and zigzags. This is most likely due both to the order of initiation and to the rates of formation of the various hair pelage types. Guard hairs appear to have the fastest rate of formation followed by 8 the awl and auchene hair types, which have an intermediate rate, and the zigzag hair type that have the slowest rate of formation (Fraser, 1951). Hair growth occurs in cycles consisting of three phases. The first is the actively growing phase, or anagen phase, which is the longest part of the cycle where the hair follicle grows downwards and the epithelial matrix cells at the bottom of the hair follicle encompasses specialized mesenchymal cells named the dermal papilla. These matrix cells constitute the stem cells, or keratinocytes, which will give rise to the cells of the emerging hair fiber. At the end of the anagen phase the dermal papilla ceases trigger matrix cell proliferation and the follicle begins to rapidly decrease in length, returning towards the epidermis. This signals the beginning of the catagen phase. During this phase the IRS disappears and the terminus of the hair fiber becomes 'clubbed' and lacks medullary cells. Following catagen is the telegen phase which is the resting or dormant phase of the hair cycle (Dry, 1926; Lavker et al., 1994). The anagen phase appears to last 10-14 days followed by 2-3 days in catagen. The telegen or dormant phase of the hair cycle then continues for 2-3 months before the retained "club" hair is shed. At approximately one month, the next hair cycle is initiated with the new hair growing alongside the first. In mouse, the hair cycle is synchronized and waves of nascent hair occur in regular cycles for at least the first several coats. After the first few cycles this whole animal rostral to caudal wave of new hair growth breaks down and hair growth occurs in large islands of synchronous growth (Dry, 1926; reviewed by Powell and Rogers, 1990). Transgenic mice (Powell and Rogers, 1990), whose hair fibres were extremely fragile and broke off soon after catagen was initiated, confirmed the pattern of cyclic hair growth in mice by displaying cyclic hair loss. Their hairs were found to break below the surface of the skin after attaining full length. The first signs of this breakage were noticeable at approximately 19 days and by one month the mice were 9 completely naked with the exception of a band of hair at the base of the tail. Then approximately two days following this naked condition new hair growth was observed. The timing of this cyclic hair loss corresponds to the end of the first anagen phase (the beginning of hair loss at roughly 2 weeks) and the estimated start of the second anagen cycle (the initiation of the second coat at one month) described above. Hair defect mutations At the initiation of this study, a literature survey was undertaken of known hair defect mutations in mouse to look for phenotypes and modes of inheritance similar to the curly mutation. It was thought that this approach would indicate potential candidate mutations which could be alleles of the curly mutation. Table 1 summarizes the literature survey of mouse hair defect mutations. Mutations without alopecia (baldness) are listed in order based on chromosomal location and for each mutation the major phenotypic difference from curly is outlined. Hair defect mutations were found scattered on two thirds of the chromosomes in the genome with the exception of chromosomes 4, 10, 12, 16, 17, 18, and 19. There were also three currently unmapped mutations. Most of these mutations had phenotypes that either differed greatly from the curly phenotype or if their phenotypes were similar, like rex on chromosome 11 for example, they had different modes of inheritance than the curly mutation. The mutations that were considered to be closest to curly with regard to similar phenotype based on their literary phenotypic description and mode of inheritance were rough (ro) and wellhaarig (we) on chromosome 2 and curly whiskers (cw) on chromosome 9. The ro mutation confers a wavy vibrissae and a rough coat phenotype. The pelage hairs appear to stick together in bundles as if wet or greasy causing the overall rough appearance to the coat (Green, 1989). we mutation homozygotes have curly whiskers and a wavy first coat which is most severe between 10-21 10 Table 1: Literature survey of known mouse hair defect mutations without alopecia (baldness): their chromosomal location, mode of inheritance and main phenotypic difference in comparison to the curly phenotype (Green, 1989). (chr = chromosome mutation is located on; rec = recessive mode of inheritance; dom = dominant mode of inheritance) mutation locus chr rec/dom main phenotvoic difference with curlv fuzzy fz 1 rec uneven coat; adult coat thin and wavy/curly ragged Ra 2 dom first coat develops slowly; normal whiskers rough ro 2 rec very similar to curly mutation phenotype wellhaarig we 2 rec very similar to curly mutation phenotype flaky tail ft 3 rec hair types not affected; tail and ears similar to curly mutation phenotype matted ma 3 rec hairs erect not curled; normal whiskers; all hairs break easily soft coat soc 3 rec first coat delayed; sparse whiskers angora go 5 rec guard hairs and whiskers twice normal length; not curled marcel mc 5 rec female sterile; strong waves in coat scruffy Scr 6 dom shiny coat; normal whiskers waved-1 wa-1 6 rec stronger curls in whiskers; some open eyes at birth frizzy fr 7 rec coat short and rough until 6-7 weeks; adult normal long hair lgh 8 rec zig zag hairs are missing; coat and whiskers normal curly whiskers cw 9 rec strongly curled whiskers; coats not waved, but slightly abnormal rough coat rc 9 rec hairs erect not curled; normal whiskers; all hairs break easily rough fur ruf 9 rec normal whiskers lustrous It 11 rec short curled whiskers at birth; glossy coat rex Re 11 dom very similar to curly mutation phenotype waved-2 wa-2 11 rec stronger curls in whiskers; some open eyes at birth satin sa 13 rec silky coat with high sheen; normal whiskers waved coat Wc 14 dom short fuzzy hair on the head and upper back caracul Ca 15 dom adult hair has plush-like look crimpy cpy 15 rec plushy-appearing coat koala Koa 15 dom hairy ears and muzzle; coat and whiskers normal velvet coat Ve 15 dom velvety sheen to coat; normal whiskers greasy Gs X dom shiny fur; whiskers not affected mottled Mo X dom pleiotropic effects incl. skeletal abnormalities fuzzy tail fzt - rec early hair with slight waving but normal in adults; whiskers unaffected hair interior hid - rec homozygotes have normal appearance defect retarded hair growth rhg - rec hair development is delayed but normal 11 days. As in curly homozygotes, the waviness of the coat is lost in the subsequent coats of we homozygotes. The hairs of these mice were found to have a lower average diameter than normal (Green, 1989). cw homozygotes have strongly curled whiskers and their coats are slightly abnormal in appearance but not waved. There is also an allele of cw, cw'hd (tail hair depletion) whose homozygotes have abnormal or depleted whiskers and tail hair beginning at 2 weeks of age. Hair loss was also found on the feet and genitalia and around the nipples. All four of these mutations arose spontaneously (Green, 1989). The ro, we and cw mutations conferred phenotypes most closely resembling what was initially known regarding the curly mutation phenotype. It should also be noted that the rex (Re), on chromosome 11, phenotype was very similar to the curly phenotype but it had a dominant mode of inheritance. These similar mutations were considered as possible candidate regions at the initiation of mapping and chromosome 2 and 9 were first looked at for linkage to the curly mutation. Rex As described in Chapter IV, linkage was detected between the curly mutation and chromosome 11 in the same region to which the rex mutation had been linked (Nadeau et al., 1989). Therefore rex became a candidate allele for the curly gene even though it was previously not a candidate due to its dominant mode of inheritance. The rex (Re) mutation arose spontaneously in an Essex mouse farmer's stock. It was termed rex after a similar coat character mutant in rabbit named Rex. The mutation was first described in 1939 (Crew and Auerbach, 1939) as being detectable as early as 2 to 3 days after birth by curliness of their vibrissae which are strongly bent forward like those of a walrus (Gruneberg, 1952). The first coat of the mutant animals is markedly wavy which disappears in the adult coat. The adult coat has curly vibrissae accompanied by an unkempt appearance of its coat with short curly and irregularly distributed Figure 3: The adult Rex coat seen in profile demonstrating the abnormal, short and irregular character of the guard hairs (Crew and Auerbach 1939) 13 guard hairs. See Figure 2 and 3 for photographs depicting the rex mutation's overall coat appearance and guard hairs. Breeding experiments showed that the Re allele was an autosomal monogenic dominant, with the heterozygote resembling the homozygote (Crew and Auerbach, 1939). In contrast, Carter in 1951 determined that Re heterozygotes could be distinguished from the Re homozygotes with a high degree of accuracy in young animals and therefore the Re allele was incompletely dominant to the wildtype allele at the rex locus. It should be noted however, that the stock that was used for these Re segregation investigations was only mildly inbred (F4) (Carter, 1951). From this it can be surmised that on at least some genetic backgrounds the dominance of the Re allele is incomplete. Several other alleles of rex have arisen to date. Carter (1951) found an allele of rex which he termed rexoid (Red) and was found to be incompletely dominant in the heterozygote. It was found that this mutation was either an allele of, or very closely linked to, rex by compound rexoid/rex (RedIRe) heterozygotes testcrossed to normal wildtype animals producing 211 curly-haired offspring (Carter, 1951). In addition to the rexoid allele (Red), two other alleles of Re are known. The first is the wavy coat (Rewc) allele which is a radiation induced dominant mutation whose phenotype resembles rex phenotypically but has a more extreme phenotype where the waviness of the first coat is visible earlier and the adult coat is rougher (Green, 1989). Again the inheritance of this mutation is incompletely dominant (Green, 1989) with the homozygotes most affected, then the Rewc/+ and RewcIRe heterozygotes less so and the Re/+ heterozygotes least affected. This mutation was considered to be an allele of Re based on the wavy-coat phenotype of all 134 offspring from Re/ Rewc outcrosses to wildtype (Searle, 1968). The second additional 14 allele of Re is denuded (Reden) which is a semidominant mutation that arose spontaneously. The phenotype conferred by this allele is normal coat growth until 4 weeks of age followed by thinning of the hair on the head and neck and progressive hair loss until, as adults, the animals are / almost naked, and usually blind, by 6 months. The whiskers of both homozygotes and heterozygotes, in contrast to the other "alleles" of rex, are normal (Green, 1989). Allelism ofite and Rede" was determined by the backcrossing of F l RelReden mice where neither the phenotype of the F l parent nor the normal phenotype was observed in the offspring (n = 504). The F l parent phenotype was almost naked with curly whiskers (Eicher and Varnum, 1986). Al l hair types appear to be present in Re/Re and Rewc/Rewc homozygotes although classification was made difficult due to the marked caliber irregularities in the hair shafts and septation of the medulla (Trigg, 1972). The Rewc/Rewc hair samples are particularly abnormal where the whole shaft is irregular with either local medullary defects or defects that span long distances within the shaft. The tips of all hair types are usually bent but are not broken. Re/Re homozygotes show very similar abnormalities but they are neither as extensive nor as severe, with the medulla rarely affected. These defects exist in the adult coat despite the loss of waviness of the adult coat. Histology from Re/Re samples revealed that irregularity of the pelage hair appeared to arise from an abnormal or wavering internal root sheath (IRS) affecting the developing hair shaft (Trigg, 1972). Unfortunately no pictures of this histology were published by Trigg (1972) and no other rex histology was found. The IRS is thought to be essential to maintain the even caliber of the hair shaft during its development. In the case of the rex phenotype therefore, it appears to be an IRS defect which produces an uneven caliber along the hair shaft in all hair types (reviewed by Trigg, 1972). 15 Although this defect is still present in adult animals, the rex coat waving phenotype has reduced severity in the adult animal in comparison to three week old animals. A hypothesis regarding why this difference is seen between adults and three week old coats (reviewed by Trigg, 1972), was developed which stipulated that space was a limiting factor during development. As all the hair follicles that are present in adulthood are also present at birth, the limited size of the epidermis in youth forces the hairs to grow curved together in groups. Therefore as the skin area is increased in adulthood, the waviness of coat the disappears. It is also interesting to note that a number of other hair defect mutations are closely linked to the rex mutation on chromosome 11. These mutations were excluded from Table 1 above due to an element of alopecia (baldness) in their phenotype. The mutation bald-arthritic (Bda) is a semidominant mutation, with baldness occurring in both homozygotes and heterozygotes, that arose spontaneously (Green, 1989). Bda was determined to be closely linked to Re with an estimated recombinant frequency of 0.032 (Wallace and Ferguson, 1984). Another dominant mutation with alopecia is bareskin (Bsk) which was found to be closely linked to Re with one recombinant found among 811 backcross offspring (Lyon and Zenthon, 1987). One further spontaneous dominant hair defect mutation that has been mapped to this same region is called recombinantion-induced mutation 3 (RimS). This mutation has a phenotype similar to Bsk and Rede" with corneal opacity and a sparse first coat followed by nearly bare adults with hair left solely on the nose (Green, 1989; Sato et al., 1998). While Rim3 was not mapped directly against Re, it was mapped with two mapping panels to within 0.9 and 1.05 c M (13 recombinants out of 1545, and 6 recombinants out of 667 respectively) of epidermal type I keratin intermediate filament gene 10 (Krtl-10) and 12 (Krtl-12) (Sato et al., 1998). These two keratin genes are thought to be part of a type I keratin supercluster (Krtl), which may contain upwards of 60 16 keratin and keratin related genes (Powell and Rogers, 1997), that maps to the distal region of chromosome 11 (Nadeau et al., 1989). In comparison, Re was mapped with regard to this same keratin cluster, in particular the Krtl-10 and Krtl-14 genes, and no recombinants in 239 progeny were found (Nadeau et al., 1989). Keratins Keratin proteins are the main structural proteins of the cytoskeleton of epidermis and epidermal cell derivatives like hair and nail. For example in the epidermis, keratins can make up to 70% of the total cellular dry weight mass (Steinert et al., 1985). This fact combined with the close linkage of the curly mutation to the type I keratin gene cluster (see Chapter IV), lead type I keratin genes into the spotlight as possible candidate genes for the curly mutation. Therefore a thorough investigation of keratins and keratin genes was done and the following is a summary. Keratin proteins are structural proteins belonging to the intermediate filament (IF) superfamily of proteins. The members of this superfamily all are thought to have originated from a primordial nuclear lamin gene (Fuchs, 1995). In contrast to other types of structural proteins like actins and tubulins, which are highly evolutionarily conserved, IF proteins share as little as 20% sequence identity (Fuchs, 1994). Features that JTs do share however are a similar secondary structure, the ability to self-assemble into lOnm filaments, and similar biochemical and immunological properties (Fuchs, 1994). The secondary structure of all IFs ( reviewed by Fuchs, 1994; McCormick et al., 1991) consists of a central a helical, or rod, domain with heptad repeats of hydrophobic amino acids and flanking non-helical carboxy and amino terminal domains. These hydrophobic repeats are thought to enable the coiled-coil association of two of these polypeptides into dimers which then, about 20,000 of them, are further associated to make a single intermediate filament (Fuchs, 1995). The central a helical domain, which for all non-17 neuronal IFs is 310 amino acids in length, consists of four helical segments, helices 1A, IB, 2A and 2B, which are interrupted by three short non-helical linker regions designated L l , L l - 2 , and L2. The start of helix 1A and near the end of helix 2B share the most sequence identity among all IF proteins. The flanking amino and carboxy terminal domains are highly variable in length and share the least similarity both between and within IF types. The superfamily of intermediate filaments is divided into five types of proteins that are all non-nuclear with the exception of the type V IFs, the nuclear lamins. These five types are classified by the type of cells in which they are expressed and by their sequence identities within their a helical domains (Fuchs, 1994; Lazarides, 1980). Type I and type II IFs are both keratin intermediate filaments (KIFs), although they are not more related to one another than other types of IFs (Darmon and Blumenberg, 1993), and are expressed in epidermal and epithelial cell types. Type III IFs are expressed in mesenchymal cell types, type IV IFs are expressed in axons, dendrites and perikarya and again type V IFs are nuclear lamins which are the structural proteins of the nuclear lamina on the inner surface of the nuclear membrane (Fuchs, 1994). Keratin intermediate filaments are divided into two classes termed type I and type II. KIFs initially were classified into these two groups based on their molecular sizes and the alternate hybridization of their cDNA sequences with two distinct human keratin mRNAs (Fuchs and Coppock, 1981). Later they were classified based on their isoelectric points (Moll et al., 1982), and the amino acid sequence of their a helices, with type I keratins being classified as acidic and type II as basic as a result of having a higher percentage of basic amino acids (Steinert et al., 1985). Keratins of one of these types can share as much as 50-99% sequence identity in their a helical domains while these regions share only 30% sequence identity with keratins of the opposite type (Hanukoglu and Fuchs, 1982; Hanukoglu and Fuchs, 1983; Steinert et al., 1983). 18 Classification of these proteins and genes have remained based on isoelectric points and molecular weights as no site in the sequences of either type I or type II keratins can be particularly designated as characteristic for classifying a keratin into these two types (Darmon and Blumenberg, 1993). Specific type I keratins form obligate heterodimers with specific type II keratins and do not dimerize with other types of IFs (Hatzfeld et al., 1987). In contrast all other mammalian IFs can exist as either homopolymers or heteropolymers with a variety of IF types (reviewed by Fuchs, 1994). The sequences that appear to be most critical for keratin type I and type II heterodimer formation are either helix IB or helix 2B and these sequences may influence higher ordered associations of these polypeptides into intermediate filaments (Hatzfeld et al., 1987; McCormick et al., 1991). These keratin pairs, of specific type I and type II proteins, are coordinately expressed in specific tissues, stages of differentiation or times during development. A survey of these various keratin pairs was conducted and their patterns of expression in mouse and human, mouse transgenic or knockout phenotypes and human disease phenotypes were investigated. This survey was carried out to elucidate which of these keratin pairs could be implicated, and which could be ruled out, as candidates for the curly mutation. The human disease phenotypes were surveyed for the purpose of looking for homologous phenotypes which could indicate whether curly could be an animal model for any human genetic diseases. Table 2 reviews these keratin pairs with regard to where they are expressed based on data from both human and mouse, and their associated transgene, knockout or human disease phenotype. Al l of the genes listed are grouped under the larger class of keratins which is then broken down into epithelial type I keratin intermediate filaments (i.e. K9-K22), epithelial type II keratin intermediate filaments (i.e. K1-K8), hair keratin intermediate filaments (i.e. HKA1-5) and keratin 19 es 4> >? 2 *e3 S" in es v. S> es on •» i -3 S .3 '-3 c3 in .2J XI cn 1) C3 2 «* 2 5 o (/3 o PS S O. £ S a .1 =) em es XI & ej CJ 5 CO o •b" a> •B .3 a. em cl es _«J | 5 o •a -a ' .a i ^ 1 l^ 1 ej II u w =1 o •a 00 •S CL, W — i—i 1 I U u CJ a ha CB .s u .3 •K oo •3 ll M o w w bli noi B o •a oo g 1 2 S o < a. CJ 20 cu •3 & cu T3 .3 it o u I .3 8 TJ a ° > o 3 § ,cu w op I ^  T3 g CU •3 & 3 o 00 oo 00 00 5 CN 5 .fei J 5 CL. X cu s o cu 1 I cu .fci 5 .fc!" CB 1 * O 3 Ti Q "S 2 -S T3 13 73 u eu u eu Cu X tu ON Os oo T3 oo o ON ON — ON +j CB <N ON ON 3 o f—I T3 eu 00 ON o P3 . r~ ON VI ON ON ON ON ON ~^ 18 « •3 8 3 > ON ON ON ON r-; ON ca ON • CB eu > CU cu 4-* >,jU —^» <u aul B •5 Oi oo 03 NO NO oo ON . . O N ON i — 1 ON ON « ON ON _ ; ca cd n n -4-. * J _; _^  eu ca cu "S Cu 8" r-ON ON .a cu w w T3 cu NO r~ 1-1 ON ON Q ON ON r—l r— a vT hristi rh rh cN hristi r- CN ON =1 =) OO hristi ON  tt, ON o ON >, >. T~~l >-> x> , ; x> , ; T3 •a 73 CB CB y -«-> tj & a CU CU w eu CU cu O revi revi Mol revi End Fuel CN 21 associated protein genes (i.e. the KAPs). Based on the literature survey, it appears that K17, K22, and the K H A s and KAPs represent the best candidates for the curly mutation. All of their expression patterns indicate their possible importance in the regulation, formation or structure of the hair shaft and/or follicle. The other keratins were excluded as candidates based either on their zone of expression, as was the case with K12, K13, K15, K18 or K19, or on their transgenic or human disease phenotype differing significantly from the curly phenotype, as was the case with K9, K10, K14, and K16. Al l keratin intermediate filaments appear to be encoded by separate genes (Fuchs, 1995) which are transcriptionally regulated (Stellmach et al., 1991). These genes, like all non-neuronal mammalian IF genes, have five to seven nearly identically positioned exons in their rod domain sequences, a further one to four variably positioned exons located in their carboxy-tail segments (Fuchs, 1994) and are 4-5 kb in size (Kaytes et al., 1991; Wilson et al., 1988). Presently K10, K12, K13, K14, K15, K19, keratin hair acidic 1 gene (KHA1) and K H A 2 of the type I keratin intermediate filament genes have all been mapped to mouse chromosome 11 while most of the remaining type I genes remain unmapped (Compton et al., 1991; Correll et al., 1992; Filion et al., 1994; Liu et al., 1994). The linkage of these type I epidermal and hair keratin genes to a single region of the genome has lead to the suggestion that these genes exist in a gene cluster, with genes most likely spaced only 5-10 kb apart (reviewed by Filion et al., 1994; Powell and Rogers, 1997). In support of this cluster hypothesis, K10, 12, 13, 14, 15, 19, 20, K H A 1 , H K A 2 , and H K A 3 have all been mapped to the homologous region of human chromosome 17 (GeneMap98, 1999). In addition, the linkage of the keratin associated protein 1 (KAP1) gene family (Parsons et al., 1994) and KAP2 gene family to the region of the type I keratin gene cluster on sheep chromosome 11 has raised the possibility of a supercluster of keratin genes at the type I keratin 22 locus which probably totals in excess of 60 genes (reviewed by Powell and Rogers, 1997). These findings indicate that this keratin type I gene supercluster exists in mouse, human, sheep and cow. Moreover, it appears that whole chromosomal regions are homologous in these four species based on the presence and retained order of many homologs surrounding and including the type I keratin locus. These regions are distal chromosome 11 in mouse, chromosome 17(q) in human, chromosome 11 in sheep and chromosome 19(q) in cow (Powell and Rogers, 1990). The exception to type I keratins mapping to the type I keratin locus is K18, a type I keratin, which has been mapped to distal mouse chromosome 15 with the type II keratin genes K I , K2, K3, K4, K5, K6, K6a, K7 and K8 (Powell and Rogers, 1997; MGD2, 1998). With the addition of type II hair keratin genes mapping to this locus, the type II epidermal and hair keratin genes also appear to be clustered into a type II keratin locus. This locus is predicted to contain 20-30 genes and span between 500-600 kb (reviewed by Powell and Rogers, 1997). It is interesting to note that several dominant hair loss and several dominant hair waving mutations are both located near two keratin gene clusters on chromosome 11 and chromosome 15 in the mouse (Lyon and Zenthon, 1986; Nadeau et al., 1992). Bda, Bsk, Re and Rim3 are closely linked to the type I keratin cluster (Krtl) on chromosome 11 and naked (N), shaven (Sha), velvet coat (Ve) and caracul (Ca) are closely linked to the type II keratin cluster (Krt2) on chromosome 15 (Lyon and Zenthon, 1986; Nadeau et al., 1989; Nadeau et al., 1992). The phenotype of N appears to be like Rim3, Bsk and Rede" and the Ca phenotype is similar to Re with regard to its complete dominance, markedly curved vibrissae, and pelage hair which lies in waves in the first coat and later straightens in adulthood (Gruneberg, 1952). In addition to the mouse hair defect mutations being linked to the keratin clusters on chromosome 11 and 15 in mouse, retinoic acid receptor genes and homeo box (Hox) genes are 23 also linked to these chromosomes close to the two keratin clusters. Retinoic acid receptor a (Rard) and Hoxb are closely linked to the type I keratin cluster (Krtl) on chromosome 11 and retinoic acid receptor y (Rary) and Hoxc are closely linked to the type II keratin cluster (Krtl) on chromosome 15 (Hart et al., 1988; Lyon and Zenthon, 1986; Nadeau et al., 1989; Nadeau et al., 1992; Sato et al., 1998). Keratin and retinoic acid receptor genes being linked in two regions of the genome is interesting because they are both involved in skin and hair formation. The former is the major structural protein of epidermal cells (Powell and Rogers, 1997; Rothnagel and Roop, 1994) and the latter is involved in the regulation of epidermal cell proliferation and differentiation (Fuchs and Green, 1981; Saitou et al., 1995). The vertebrate Hox genes (reviewed by Fuchs, 1995) encode transcription factors that are mainly involved in specifying segmental positional information in early development. While the epidermis itself is not a segmented structure, it overlies and is influenced by segmented mesodermal structures whose Hox protein gradients have been implicated in providing the positional information directing skin appendage (hair and nail) patterning of some vertebrates. The transient expression patterns of the Hox genes during later development also support a role for the Hox gene products in the regulation of the differentiation status of the epidermis and epidermal derivatives. Endorsing this hypothesis is the action of a homeobox protein, a member of the Distal-less family (Dlx-3), which is thought to be involved more in the cellular differentiation of epidermal cells than in providing positional information. Dlx-3 has been found to be extensively expressed in the skin and hair follicles (Robinson and Mahon, 1994) and differentiating keratinocytes in vitro (Morasso et al., 1993). In addition, a mutation in Dlx-3 has been credited in a human disease phenotype affecting hair and skin (Price et al., 1998) and the gene has been mapped to the region of the type I keratin cluster and Hoxb cluster in humans (Nakamura et al., 1996; Scherer et al., 1994). Interestingly it has been 24 established that keratin genes (Saitou et al., 1995; Stellmach et al., 1991) and Hox genes alike, are regulated by retinoids (reviewed by Fuchs, 1995). The fact that these three types of genes, keratins, Hox, and RARs, are linked on two apparently paralogous chromosomes in mice and humans suggests some sort of regulatory or functional importance for these genes to remain linked through evolution (Nadeau et al., 1992). Mouse Mapping Overview and history A genetic map is a graphic portrayal of the relative locations of a subset of loci within the genome (reviewed by Silver, 1995). Maps can vary greatly in their degree of resolution from the most primary level, where loci are given a chromosomal assignment, to ever increasing levels of resolution where the relative order and approximate distance separating individual loci are elucidated. This can be continued until the most complete level of mapping is obtained where loci are mapped relative to the D N A sequence itself. Genetic maps are divided into three types based on both the information generating them and their base units of measure. One of these types is the linkage map whose map positions are delineated by recombination found through classic breeding analyses. These maps are measured in centimorgans (cM) with 1 c M equaling the distance between 2 loci amid which recombinants are observed with a frequency of 1%. The second type of genetic map is the chromosomal map which is based on chromosomal karyotype. Gene positions, which are enumerated with the chromosomal band name, are elucidated by direct cytogenetic analysis or by linkage to a locus that has been previously mapped in this manner. The third type of map is the physical map which is directly based on the analysis of the D N A sequence and is measure in basepairs (bp), kilobasepairs (kb) or megabasepairs (mb). The relative distances of all three of these maps differ, with the physical map representing the only truly accurate 25 measure of the distance between loci. This is because the D N A packing of the condensed chromosome used for karyotyping distorts the distances represented by chromosomal maps and the non-randomly distributed recombination sites affects the distances defined by linkage maps. This non-random distribution of crossover sites has been repeatedly observed and it appears that there are variable and unequal rates of recombination among subchromosomal regions, the telomere in particular, between laboratory strains and between the sexes (reviewed by Silver, 1995). In practice, data from all three maps: linkage, chromosomal and physical, generate genetic maps, with linkage data currently providing the greatest bulk of the positional information (reviewed by Copeland et al., 1993; Silver, 1995; MGD2, 1998). This linkage data encompasses classic recombination studies using mutations with detectable phenotypes, isoenzyme loci, cloned loci and D N A sequence polymorphisms. For most of this century (reviewed by Copeland et al., 1993; Silver, 1995), two or three point crosses using visible phenotypic markers detected genetic linkage. Progress using this method was slow and expansion of the map took place one locus at a time. Possibly the biggest hindrance with this type of mapping, however, was the problems inherent in creating composite maps with this data because only a handful of markers could be examined in each cross. One of the first mapping improvements occurred in the 1970's when polymorphisms were identified between enzyme isoforms. These loci were highly polymorphic among different inbred laboratory stains facilitating mapping studies using easily accessible strains but had a limited number of loci across the genome. Later, in the 1980's, two scientific advances revolutionized the mouse map. One of these was the discovery of polymorphisms in the D N A sequence itself and the realization that these sequences were "loci" even though they generated no detectable phenotype. One type of 26 polymorphism that was quickly embraced and utilized was restriction fragment length polymorphisms (RFLPs) scored using Southern (DNA) blots. Unfortunately, due to the common ancestry of most inbred laboratory strains, RFLPs had a low rate of polymorphism between laboratory inbred strains thus making their immediate usefulness for linkage mapping limited. The second advance was the development of the interspecific cross, which involved crossing an inbred laboratory strain (Mm musculus) with a distantly related species of mice, typically Mm spretus. This cross was discovered to produce fertile female F l hybrids, which were then backcrossed to one of the parental strains, allowing the segregation of loci to be traced to determine linkage. Due to the genetic diversity of these two species most loci are polymorphic; thus in a single cross potentially thousands of loci could be mapped relative to each other. These two advances together, with RFLP loci being polymophic in the interspecific backcross, added thousands of markers to linkage maps and solved the problem of data integration by allowing the analysis of a multitude of markers with a single cross. New problems did arise, however, with these techniques. Obtaining and scoring RFLP sequences was time consuming, dissemination of clones for use as Southern blot probes was difficult and the mapping of loci through interspecific backcrosses posed limitations on certain types of analyses. For instance, mapping using an interspecific backcross would greatly increase the difficulty of studying mutations whose phenotypes were affected by genetic background or maternal genotype. P C R and SSLPs In the late 1980's and early 1990's, RFLP typing of interspecific backcross data was overtaken by polymerase chain reaction (PCR) amplification of microsatellites, another D N A polymorphism, as the primary technique generating linkage data (reviewed by Copeland et al., 1993; Silver, 1995). The PCR technique, whose enzymatic amplification plus gel electrophoresis 27 and ethidium bromide staining was expeditious and simple, quickly became preferred to the time consuming blotting, probe labeling, hybridization and autoradiography of the Southern blot. Other advantages conferred by PCR were rapid and exponential amplification of D N A sequences present in extremely low copy number, high resolution of polymorphisms from single basepair changes to large rearrangements, easy dissemination of methodology via published primer sequences, automatable typing, and relatively inexpensive set-up and typing costs for sophisticated mapping analyses. Several types of D N A polymorphisms can be typed by PCR including the typical point mutation polymorphisms of RFLP sites (reviewed by Silver, 1995). These include variable number tandem repeat (VNTR), single-strand conformation polymorphisms (SSCPs), random amplification of polymorphic D N A (RAPD) and simple sequence length polymorphisms (SSLPs). However, these loci are not equal with regard to their usefulness for linkage analysis. RFLPs, as mentioned previously, and SSCPs display only limited polymorphisms and most are di-allelic, V N T R loci are limited in number with fewer than 1000 loci, and R A P D techniques are directed towards amplifying multiple random segments across the genome, making their linkage assignment to any region in particular impossible. Fortunately, SSLPs fulfill many of the requirements of the perfect linkage mapping loci. Simple sequence length polymorphisms (SSLPs) are simple, mono-, di-, tri- or tetrameric, sequence tandem repeats (SSRs), the most frequent of which in mouse are (CA) n /(GT) n repeats, which are highly polymorphic with regard to the number of tandem repeats in a given allele (reviewed by Copeland et al., 1993; Dietrich et al., 1992; Silver, 1995). These alleles are most likely created de novo through unequal recombination or slippage during D N A replication and are very polymorphic even between inbred laboratory strains. In addition they are present in a 28 high copy number and appear to be randomly distributed across the mouse genome. SSRs in mouse can be found by screening both published D N A sequence databases and mouse genome libraries with probes for (CA) n and (GT) n repeats. Once found, these clones can then be isolated and sequences obtained for the SSR flanking sequences. These methods for searching for SSR loci generated an enormous number of markers in a short time span. SSLPs quickly became the preferred method for genetic mapping in the mouse and made it possible to return to Mus musculus to map genetic variants, including those affected by genetic background or maternal genotype. Most modern genetic maps therefore are mostly composed of SSLP positional data but it should be noted that the framework or anchors for these maps were laid down by SSLPs taken from genes mapped by the previous methods. Review of the state of the mouse map Several maps of the mouse genome exist to date based on both individual crosses and composites of all available data. One would assume that the latter would be more accurate but it must be recognized that there are problems inherent with integrating maps and map distances arising from widely differing crosses, mapping panels and markers examined. For this study, three maps of chromosome 11 were examined, the Research Genetics map, the European Collaborative Interspecific Mouse Backcross (EUCIB) map, and the Mouse Genome Database (MGD) map, and are all available on the World Wide Web (EUCIB, 1998; MGD1, 1998). Al l of these maps have their own inherent pitfalls and it should be noted that they are considered by the mouse scientific community to be largely works-in-progress. The Research Genetics map (Copeland et al., 1993; Dietrich et al., 1992; Dietrich et al., 1996; Dietrich et al., 1994) was generated in conjunction with the discovery of a large number of SSLP loci in mice. With the discovery of SSRs, a project (reviewed by Dietrich et al., 1992) was 29 undertaken to create a genetic map of the mouse using SSLPs loci. These SSRs were found by two methods: screening a M l 3 library of mouse genomic D N A with (CA)i 5 and (GT)i 5 probes and screening public computer sequence databanks for genes containing SSRs. Using these methods, primers for 455 SSLPs were tested, initially for polymorphism between C57BL/6J-oblob and CAST/Ei , and then, if polymorphism was detected, the alleles (product sizes) of 12 inbred strains were determined. These SSLPs were mapped using 46 OBxCAST F2s and several B X D recombinant inbred (RI) lines of mice. These data were then computer analyzed using the M A P M A K E R computer package (reviewed by Dietrich, W.F. et al, 1992) to generate the Research Genetics linkage map. Markers were assigned into linkage groups and ordered based on L O D scores and minimizing the number of recombinants. This initial project resulted in a map of 317 SSLPs, of which about 50% were polymorphic among inbred strains, placed across 99% of the genome with an average spacing of 4.3 c M (Dietrich et al., 1992). By the time the project was completed in 1996 (Dietrich et al., 1996), a genetic map of 6,580 SSLPs and 797 RFLPs was integrated with the average spacing between markers of 0.2 c M or 400 kb. This map remains mostly unchanged from the end of the project in 1996 and many loci remain with their map positions unresolved relative to each other. This is because these loci were mapped using only 92 meioses, i.e. the 46 OBxCAST F2s animals mentioned above; therefore the smallest increment that could be mapped was 1.1 c M (1 recombinant/92 total). Markers were therefore mapped to 1.1 c M groups. The EUCIB project was undertaken to develop a high resolution genetic map that would form the backbone for the construction of the completed genetic and physical maps of the mouse genome. 1000 interspecific backcross (BC1) animals were generated by a (C57BL/6J x SPR or SEG/Pas) F l x C57BL/6J cross in order to create a mapping panel with a genetic resolution of 30 0.3 c M with 95% confidence. These 1000 animals were then typed for 70 loci, 3-4 per chromosome, across the genome to create a primary anchor map with the order determined based on minimizing the number of recombinants. From these same 1000 animals, pools of recombinant animals for each chromosome were created. Thus, by using this backbone map and pools of recombinant animals, EUCIB then could provide a high resolution (>lcM) mapping service to the scientific community. A new marker could be tested against a panel of 40-50 animals for chromosomal linkage and then could be mapped against the pools of recombinant animals for that chromosome. Using these two ways of collecting markers for the EUCIB map, a map was created that had rather large map distances between markers and therefore an underestimation of chromosome length and linkage distances, most likely due to errors in recombinant frequency due to undetected recombination. In other words the actual distance between loci will be underestimated because the number of recombinants observed would not include those which are a product of a double crossover event. Then in 1996 EUCIB, in collaboration with MIT, started a project to map the 6000 micro satellites on the EUCIB mapping panel. While this increased the number of markers significantly per chromosome, many of the markers remain unmapped, at least on chromosome 11 despite claims that it is completed. The bottom line is that EUCIB was created as an international resource for high resolution mapping rather than to create the map itself (EUCIB, 1998). In contrast to the other two maps of the mouse genome, the M G D map was created to integrate all publicly available data to create a composite map as a resource to the scientific community. It was created from the data of 356 crosses (MGD2, 1998) in addition to data from journal submissions and various other mapping submissions. M G D is updated daily (Shaw, 1998) and, thus, is in a state of constant flux. While this map is more comprehensive, it is also more 31 error prone because it integrates data from widely differing sources with widely differing map distances. A partial solution to this problem has been reached by comparing data sets and map distances that include a common group of anchor loci for orientation, comparison and integration. As a result, despite the above mentioned short comings, the M G D map is considered to be the most up to date, detailed and comprehensive 'work-in-progress' map of the mouse genome. Rationale and approach to this study The S E L H curly substrain was maintained for study, like the seven other mutations that arose on the SELH/Bc strain, in the hopes of elucidating the mechanism of mutation causing the apparently elevated mutation rate in the SELH strain. If any of these mutations mapped to cloned regions of the genome, the molecular nature of the mutations could be determined which in turn could give insight into the mechanism creating the mutations. Therefore it was desirable to map and characterize the molecular nature of as many of these eight mutations as possible. It should be noted that at the initiation of the curly mapping project, the molecular nature of the nuBc and the c3Bc alleles remained to be determined. Therefore it was unknown that the insertion of early transposon (ETn) sequences were responsible for at least two of the above eight mutations (Hofmann et al., 1998). While this information was not useful in mapping the curly mutation, in the future this information could facilitate the molecular characterization of additional mutants and gives an important molecular explanation for the enhanced mutation rate in the SELH/Bc strain. The study of the SELH curly mutation was undertaken for three main purposes. The first, mentioned above, was to possibly contribute towards an understanding of the mechanism of mutation causing the elevated rate of mutation in the SELH/Bc strain by trying to characterize 32 the lesion responsible for the curly allele. The second was to characterize the phenotype, genetic transmission and map position of the curly mutation both to report this new hair defect mutation to the scientific community and, thirdly, to evaluate curly as a possible model for human clinical disorders. The determination of the phenotype of the curly mutation was done to contribute to the second and third objectives while the mapping of the curly mutation was done to satisfy all three. Facilitating both the reporting of the curly mutation to the scientific community and the evaluation of curly as a possible model for a human genetic disorder was the characterization of the phenotype of the curly mutation. In addition, characterizing the phenotype of the curly mutation aided in the search for the map position of the curly mutation by indicating candidate genes based on similar phenotype properties. Towards all of these aims, the phenotype of the curly hair animals was determined and carefully documented by external gross examinations, by histological analysis of affected regions and by examination of the hair shaft itself. This fulfilled the objective of reporting this mutation and advanced the assessment of both candidate genes and candidate human disorders based on phenotypic comparisons with known mutations. This was done by comparing the phenotype of the curly mutation to known hair defect mutation phenotypes, based on literature surveys, in human and mouse to look for similar phenotypes and inheritances to evaluate potential candidate genes or candidate human disorders of the curly gene. These candidate regions then became important starting points for the mapping of the curly mutation. Accompanying the characterization of the curly mutation for report to the scientific community and to confirm assumptions regarding the mode of inheritance of the curly allele made prior to this study, the mode of inheritance of the curly allele was examined. This was 33 undertaken through careful observation and breeding of both curly homozygotes and heterozygotes on both the SELH/Bc background and outcross backgrounds. The phenotype of the curly heterozygote was closely examined to confirm that this phenotype did not differ from the normal phenotype and therefore that the curly allele was completely recessive. The segregation patterns of these crosses were analyzed to confirm that the curly mutation fit a single locus recessive mode of inheritance. In order to map the curly mutation, an experimental approach needed to be utilized that would allow a single cross to indicate the region to which the curly mutation was linked, based on the hypothesis that the curly mutation was located somewhere in the mouse genome. This approach was to take phenotypically curly animals which, based on the proposed recessive mode of inheritance of the curly allele, were homozygous for the curly allele at the curly locus and outcross them to another strain that would be homozygous for the normal allele at the curly locus. This would produce F l animals that were heterozygous for the curly mutation at the curly locus. By intercrossing these Fls , an F2 generation would be created that would have a 1:2:1 segregation of genotypes at all loci made heterozygous in the F l , including the curly locus. This 1:2:1 segregation of genotypes at a single locus from a heterozygote intercross is known to occur due to equal segregation and independent assortment in meiosis. Therefore, if all of the animals displaying the curly phenotype were selected from the F2 population, they would have a 1 (homozygous SELH):2 (heterozygous SELH/other strain): 1 (homozygous other strain) segregation of all of their alleles across the genome except for markers near or at the curly locus where they would all be homozygous SELH. If a difference could be detected between the alleles inherited from each of the two strains, the curly locus could then be located by searching the genome for the one place where all curly F2 animals are homozygous SELH. 34 With forethought to the above strategy, the curly mutation was outcrossed to the L M / B c inbred mouse strain prior to the initiation of this study (D.M. Juriloff and M.J . Harris; see Chapter II below). This strain was chosen because informative SSLP alleles between S E L H and L M were known across the genome (Gunn, 1995) and hence gave a method for searching for the location that was homozygous SELH in the curly L M x S E L H F2 population. This was first approached by pooling curly L M x S E L H F2 D N A samples and typing SSLP loci in regions thought to contain candidate loci. These candidate loci were found by comparing the phenotype of the curly mutation to known hair defect mutations. Next random regions of the genome were scanned for linkage with the curly mutation. Once linkage was found, then individual curly L M x S E L H F2 D N A samples were used to analyze heterozygous breakpoints to narrow the region containing the curly mutation. It was hoped, at this point, that the curly mutation would map to a cloned region of the genome so that the exact locus and lesion responsible for the curly phenotype could be elucidated to address the issue of the mechanism of mutation in the S E L H strain. This was not the case however. Therefore, this study continued to further resolve the map position of the curly mutation. This was a second outcross to the inbred strain AXB-10/Pgn, which was known have different allele sizes than SELH at SSLP loci in the region to which the curly mutation had been mapped. This further mapping again utilized the heterozygous breakpoints of curly AXB- lOxSELH (X10.S) F2 individual D N A samples to narrow the region containing the curly mutation. 35 CHAPTER H: GENERAL MATERIALS AND METHODS SELH curly: scientific progress before this study The inbred lab strain LM/Bc was chosen for this outcross because a method of detecting alleles based on their strain of origin was known between SELH/Bc and LM/Bc . This method was polymerase chain reaction (PCR) amplification of simple sequence length polymorphisms (SSLPs). Based on research carried out by a previous graduate student in the JurilofrTHarris lab, Teresa Gunn, informative SSLPs were elucidated between SELH/Bc and L M / B c with spacing of approximately 20 cM across the genome (Gunn, 1995). Therefore the mating of a single L M / B c female ( ? 7576) to a single SELH/Bc homozygous curly male (0*7148) was carried out for the outcross of the curly mutation. These parents produced one litter of heterozygous (LM/SELH) F l animals, 3 males (0*02;d04;0*06) and 3 females (?01; ?03"; 905). Upon sexual maturity these six animals were then intercrossed. The crosses and the numbers and phenotypes of the resulting F2 progeny are summarized in Table 3.33 animals out of the 174 F2 L M x S E L H progeny were deemed homozygous curly which equaled 19% of the total progeny. This amount of homozygous curly animals is not statistically different (%2= 3.379) from the expected number of homozygous curly animals (25%) from the F2 of a monogenic trait. In addition the sexes were statistically equally distributed in both the normal and the curly hair phenotypic classes with 72 9:69 c* ( ( x 2 = 0.064) and 17 9:16c? ((x 2= 0.030) respectively. Phenotypic classifying of the F2 animals occurred at two to three weeks of age based upon the whiskers and femur hair of the animals and was carried out D . M . Juriloff and M.J. Harris. This task was somewhat complicated by the whisker-nibbling social grooming behavior of the LM/Bc strain that was apparently passed on to the L M x S E L H F2 progeny of this cross. One animal had to be scored as "probably curly" due to the distinct lack of scorable whiskers ( D M . Juriloff and M.J. Harris, personal communication) 36 and was omitted from future mapping analysis. Tissue was collected individually from 32 curly F2 L M x S E L H animals plus individuals from three complete F2 litters (i.e. including animals with the "normal" phenotype; n = 31). All tissue samples were taken from the liver and for the 33 (32 "curly" + 1 "probably curly") homozygous curly animals, equal amounts of these samples were also pooled in groups of 2-4. In addition, tail tissue samples were taken from the 33 homozygous curly F2 animals and stored individually. All of these samples were stored immediately in -20°C, and my participation in the study began with these frozen samples. Table 3: Data summary of F2 L M x S E L H progeny: parents, phenotypic proportions, sex ratios and animal numbers on the basis of individual litters (each horizontal row represents data from a single litter). F1 Normal curly id number parent ? d1 ? ? c? normal curly 05 06 5 4 2 1 1-9 10-12 3 7 1 1 53-62 63; 64 3 4 1 4 65-69 8 6 1 2 85-90; 93-96; 91; 92; 97 98-101 01 02 4 6 1 2 13-22 23-25 7 5 1 1 — 50-51 9 3 0 0 — — 2 2 1 0 102; 103; 104 105; 106 5 4 3 1 121; 122; 123; 124; 128; 125-127; 133 129-132 03 04 7 5 0 1 26-37 38 1 5 3 0 39; 43-47 40-42 6 5 1 1 — 48; 49 9 5 0 1 70-74; 76-84 75 3 8 2 1 107-114; 115; 119; 120 116-118 37 Mouse stocks and maintenance All mice were maintained in the animal unit of the Department of Medical Genetics at the University of British Columbia. They were housed in windowless rooms under the conditions of a 20-24°C ambient temperature and a 12 hour light (6am to 6pm), 12 hour dark cycle. The mice were housed in standard polycarbonate cages with dried corncob bedding and provided with Purina Laboratory Rodent Diet (#5001) and acidified water (pH 3.1, HC1) ad libitum. Mice were weaned at 3 weeks of age and mated at approximately 6 weeks of age. Technical methods a) DNA preparation Liver and tail tissue was collected from mice freshly killed by C 0 2 gas. The collected liver and tail tissue was stored in -20°C. Tissue samples used for D N A extraction procedures were washed with a 2% PBS solution (150 mM NaCl, 2 mM N a / K P 0 4 pH 7.3, 5 mM KC1), minced with a fresh scapel into approximately 1mm pieces and digested in a 60°C water bath for 12-16 hours in 1 ml lysis buffer ( 100 mM NaCl, 10 mM Tris-HCl pH 8.0, 25 m M E D T A pH 8.0, 1% SDS) containing 1.3 mg/ml of proteinase K. After digestion the mixture was subject to a standard phenol/chloroform extraction procedure followed by ethanol precipitation of the D N A (Sambrook et al., 1989). The D N A pellet was then rinsed with 70% ethanol and resuspended in tris-EDTA, pH 8.0 (Sambrook et al., 1989). The concentration of the D N A solutions were measured on 1/100 dilutions (in water) of each sample in a UV-Spectrophotometer. The concentration of D N A (ug/ul) was calculated by multiplying the absorbance (A26o) by the standard converting absorbance at A260 into D N A concentration (50(ig/ml), by the dilution factor and by the volume diluted and used. This is abridged in the following formula: 38 [DNA] (u,g/ul) = (A26o)(50uJg/ml)(dilution factor)(volume of diluted )(l/volume of stock) sample used to solution (pi) obtain absorbance reading (ml) b) PCR o f SSLPs For each of the DNA samples, lOOng/uJ D N A stock solutions were prepared for PCR by the addition of autoclaved distilled water (ddH 20) to fractions of these samples. The diluted samples were then subject to polymerase chain reaction (PCR; Innis et al., 1990; Saiki et al., 1985; Saiki et al., 1988; Sambrook et al., 1989) amplification of simple sequence length polymorphisms (SSLPs) for use as genetic markers for the mapping of the curly mutation. 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 pi volume overlaid with mineral oil. The reaction contained 200ng of DNA; 0.14 p M of each forward and reverse primer; and the rest of the volume of the reaction was completed by a "master mix". The "master mix" in each reaction consisted of 50pM of each of dATP, dGTP, dTTP and dCTP (Pharmacia); 0.625 units of Taq D N A polymerase (Gibco-Brl); lOmM Tris-HCl pH 8.3 and 50 mM KC1 given by 10X PCR buffer (Gibco-Brl); and 1.5 mM magnesium chloride (Gibco-Brl). PCR was performed in a Perkin-Elmer 4600 thermocycler programmed for 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. For optimal amplification some primers required different conditions than listed above. These exceptions are listed in appendix table A. 39 c) Visualization of PCR products for SSLPs PCR products were visualized by horizontal agarose gel electrophoresis. The marker dye bromophenol blue-xylene cyanol FF (5(0.1) was added to the PCR product and 10u.l of the mixture was run usually on 4% 3:1 NuSieve agarose gel ( 3 parts NuSieve agarose to one part SeaKem L E agarose (FMC Bioproducts)) containing approximately 0.5 u.g/ml of ethidium bromide. Gels were run in l x T A E (Sambrook et al., 1989) at about 130V for 1 Vz to 2 1/2 hours, then photographed (Polaroid 667 film) over U V light. A few of the allele products ran too closely together to be discriminated on NuSieve gels and were better visualized on 4% Metaphor (FMC Bioproducts) gels, run in 0.5xTBE (Sambrook et al., 1989) and visualized as above. d) Histology Harvested epidermal tissue, approximately 5 mm2 in size, was fixed in either Bouin's solution for 4-7 days, or 10% formalin buffered in sodium acetate for 1 day, and sent to Wax-it Histology Services (Vancouver, B.C.). These tissues were then embedded in paraffin, sectioned at 5 microns thickness and stained with hematoxylin and eosin. Hematoxylin is a basic or cationic dye that stains nuclei and heterochromatin (Sheehan and Hrapchak, 1980). While this dye brings out the nuclear detail, a counter-stain is needed to stain the rest of the cell. This counter-stain is most often eosin which is known to stain blood and muscle cells particularly well (Sheehan and Hrapchak, 1980). Each paraffin embedded block of tissue was used for at least 4 consecutive sections. The blocks of dorsal back skin were also used for 5 micron sections taken every 100 microns, with at least eight such sections taken from each block. 40 CHAPTER HI: PHENOTYPE AND GENETIC TRANSMISSION OF CURLY Introduction The aim of this part of the study was to characterize the phenotype of the curly mutation in detail. This was done for the purpose of reporting this new mutation to the scientific community, to aid in the evaluation of candidate genes and to assess the curly mutation's value as an animal model for known human genetic disorders. Rationale, Materials and Approach Gross observations of phenotype Close examination of external phenotype was carried out using SELH/Bc animals, with homozygous normal (+/+; n = 17), heterozygous curly (+/cur; n = 16) and homozygous curly (cur/cur; n = 38) genotypes, of various ages for the purpose of documenting the curly phenotype. This was done in order to carefully document the cur/cur phenotype to assess its expressivity, to compare the curly phenotype with other hair defect mutations that might be candidate genes and for the purpose of reporting this novel mutation to the scientific community. The phenotype of normal S E L H (+/+) animals needed to be characterized, in addition, to judge which aspects of the mouse phenotype were altered by the curly mutation. A list of animal numbers, ages and sexes is located in appendix Table B. The external examinations included close inspection of the coat and coat hairs, whiskers, ears, eyes, tail, feet and nails mostly with the naked eye. Examinations also included ruffling of the pelage by petting the coat in a caudal to rostral direction or against the grain. Frequency differences in the phenotypic data between age and sex groups outlined above were tested with 2x2 contingency tables using the null hypothesis that the relative frequencies of the affected traits were the same for both sexes and for animals grouped by age. These tests used 41 numbers of animals and Chi-squared tests with Yates correction for continuity (%2 = £ ( I expected - observed I - 0.5)2/expected) to determine statistical significance. These tests had degrees of freedom = 1 and , thus, rejected the null hypothesis if yl values exceeded 3.841 (a = 0.05). Rudimentary autopsies were also performed on four of the above animals, two eight month old siblings, 6*7765 +/(+ or cur) and 97764 cur I cur, and two six month old siblings, ? 7802 +/cur and 6*7803 cur/cur. These autopsies consisted of gross examinations of thorasic and abdominal organs of phenotypically curly animals in comparison to their normal littermates with regard to placement and size. In addition, the inside of their pelts were examined and compared. Photographs were made of most of the traits examined in the various genotypes and strains using Kodak 35 mm film and a Pentax camera with a 200 mm zoom lens. An additional four of the above animals, two five and one half month old siblings, o*7855 +/cur, 9 7854 cur/cur, and two five and three quarter month old siblings, 6*7889 +/+ and 97888 +/+, were used for skull preparations to examine and compare the teeth of phenotypically curly and normal animals. Skulls were prepared by removing excess tissue, boiling in water for one minute with a subsequent cooling period. Heads were then placed in 20 mis of saline (0.85% NaCl) solution with lOmg of papain for 12-16 hours, rinsed repeatedly with tap water, debrided physically with forceps to remove remaining tissue, and defatted in acetone for 24 hours (Luther, 1949). Genetic transmission The clearly abnormal phenotype caused by the curly mutation appeared to segregate as a fully penetrant recessive mutation on the SELH background (D.M. Juriloff and M.J . Harris, personal communication). However, a detailed study had not been done, and a milder abnormal phenotype in heterozygotes (semi-dominance) on the SELH background had not been ruled out. 42 The potential effects of genetic background on the curly mutation phenotype had also not been studied, apart from the recovery of the 'typical' whisker and hair traits in the affected F2's from the L M x S E L H cross. Therefore, three questions arose regarding the genetic transmission of the curly allele. The first was whether, upon examining homozygous normal and heterozygous curly animals in detail, there were any detectable phenotypic effects due to the single copy of the curly mutation. The second question was whether the previously recognized curly phenotype was inherited as a recessive single locus trait. This was addressed by comparing the phenotypic ratios of segregants from curly homozygote intercrosses, curly heterozygote intercrosses and crosses between curly homozygotes and curly heterozygotes to the expected phenotypic ratios for these crosses. The third question that needed to be answered regarding the genetic transmission of the curly mutation was whether the genetic background affects the penetrance of the curly mutation either in one or two copies. The first two questions were addressed on the S E L H genetic background while the third was done on a segregating X10.S background. The crosses on the X10.S segregating background also provided segregation data to address the second question above, that of mode of inheritance of the curly allele. It should be noted reciprocal crosses were not done using cur/cur dams because it was thought that they did not raise their young. This was investigated and is reported below (see Chapter II: Litter raising capability), a) S E L H background In order to address the question of the complete dominance of the wildtype allele at the curly locus, two types of crosses within the SELH strain, curly homozygotes (cur/cur) crossed with curly heterozygotes (+/cur) and intercrosses of curly heterozygotes (+/cur), were examined at 2-4 weeks of age for a difference in phenotype between normal homozygotes and curly heterozygotes. Both of these crosses created pools of'normal' animals containing both the normal 43 homozygous and curly heterozygous genotypes. These 'normal' animals were closely examined for any indication that they differed from the normal homozygote (+/+) S E L H phenotype that was determined. For the (cur/cur) x (+/cur) crosses, 2 litters from 2 mating pairs (both had cur I cur dams) produced 21 animals, 10 of which had a 'normal' phenotype and were examined. Two litters (n = 9) were examined from SELH (+/cur) intercrosses, all of which had a 'normal' phenotype. In addition, obligate curly heterozygous (+/cur; n = 6; 6*9081, 6*9082, 6*9083, 97802, 6*7855 and 6*8736) and normal homozygous (+/+; n = 4; 98683, 6*7889, 97888 and 6*8198) adult S E L H animals provided by D . M . Juriloff and M.J . Harris plus obligate normal homozygous animals (+/+; n = 13; 6*L41-44, 9L45-48, 6*L49-51, 9L52 and 9L53) generated in this study, were used for the same purpose. To characterize the phenotype of the above animals, close examination of these animals included detailed, naked eye inspection of the coat, whiskers, ears, eyes, feet and nails plus dissecting scope (8X to 40X magnification) inspection of the whiskers and guard hairs. These animals had been freshly killed at the time of examination. Examinations were carried out in order to detect abnormalities in comparison to the homozygous normal (+/+) S E L H phenotype. In this manner it was determined whether or not the heterozygote (+lcur) phenotype differed from the wildtype homozygote (+/+) phenotype. In order to determine whether the curly phenotype was due to a recessive mutation at a single locus, phenotypic ratios of segregants from curly homozygote intercrosses, curly heterozygote intercrosses and crosses between curly homozygotes and curly heterozygotes were compared to the expected phenotypic ratios for these crosses. Alternatively the curly phenotype could have been attributable to an allele which had dominant effects in the heterozygote or to mutations in more than one locus. For a simple Mendelian recessive mutation, the expected phenotypic ratios for the above crosses would be 100% curly, 1 curly: 3 normal, and 1 curly: 1 44 normal, respectively. For the curly homozygote intercrosses, one mating pair of homozygous curly (cur/cur) animals produced one litter of 8 animals. Segregation data for crosses between heterozygous curly animals arose from two litters (n = 9) from a single mating pair. Crosses between homozygous curly (cur/cur) and heterozygous curly (+/cur) mating pairs yielded 17 litters (n = 140) from 11 mating pairs, 8 of which had homozygous curly dams. A Chi-square test was done to see if the observed phenotypic proportions met with the expected proportions. The critical value for a Chi-square distribution with df = 1 and a = 0.05 is %2 = 3.841 and for df = 2 and a = 0.05 is x 2 = 5.991. b) Segregating background To determine whether the genetic background affects the penetrance of the curly mutation, both the phenotype of obligate heterozygotes and the proportions of the F2 progeny with the curly phenotype were examined. The former served the purpose of determining whether the wildtype allele is completely dominant on the segregating background and the latter determined whether the proportion of curly animals in the F2 population represent all of the expected cur/cur animals present. If the number of phenotypically curly animals were not to meet the expected 25% in the F2 generation this would indicate incomplete penetrance of the cur/cur genotype and modification of the phenotype by the genetic background. Crosses between three mating pairs of homozygous curly (cur/cur) SELH males and homozygous normal (+/+) A X B -10/Pgn females (n = 6 litters) created obligate curly A X B - l O x S E L H (X10.S) F l heterozygote (+/cur) animals (n = 40). The first 10 of these animals were examined while alive by the naked eye in the animal unit as they were being weaned at 3 weeks of age. The phenotype of these animals was then compared to the phenotypes of both the normal homozygous AXB-10 and S E L H phenotypes to determine whether there were some mild effects in heterozygous animals 45 (+lcur) attributable to the presence of the curly allele. It should be noted that neither the AXB-10 normal nor the SELH normal phenotype alone could be directly compared to the X10.S F l phenotype but if the wildtype allele truly is completely dominant then the phenotype of the X10.S F l curly heterozygotes should be similar to either one phenotype or the other or a combination of the two normal phenotypes. Crosses between X10.S F l s yielded 265 animals, 263 of which were grouped into curly and normal classes based on the scoring protocol outlined below. An additional two animals (9155 and 6* 158) were termed 'probably curly' as their phenotypic classification was uncertain. A Chi-square test was done to see if the observed phenotypic proportions met with the expected proportions. An X10.S F2 animal was considered to be cur/cur if it displayed any two of the three following criteria: • Affected guard hairs on dorsal coat near the base of the tail. Affected guard hairs appear thin and malleable. They may point in several directions as opposed to the appearance of normal guard hairs that only point in a rostral to caudal direction. In addition, the ends of the guard hairs frequently display hooks or "c"s. • A ruffled or messy overall appearance to the coat. Pelage hair that seems to clump or part and becomes easily ruffled upon handling. The coat does not snap back into a sleek normal appearance upon back-stroking. The clumping or parting aspect of the coat is easily visualized on the posterior aspect of the femur. • Whiskers with uneven fanning and spread. Ends may or may not be kinked. Non-straight whiskers that display irregular curves with many that have swooping bends like a winding highway. 46 The X10.S cross was also used to test the single locus recessive hypothesis regarding the inheritance of the curly mutation on this segregating background. Two separate crosses generated data for this test, namely the parental ((+/+) x (curI cur)) cross and the F l ((+/cur) x (+/cur)) cross. The parental cross of three mating pairs, all with cur/cur sires, yielded 40 animals in 6 litters. The F l cross of 14 mating pairs yielded 265 animals in 25 litters which were scored between 2-4 weeks of age. The phenotypic proportions expected for these crosses based on a recessive single locus hypothesis for the curly allele would be 100% normal for the parental cross and 1 curly:3 normal for the F l intercross. The affected, curly, phenotype was defined as described below. A Chi-square test was done to see if the observed phenotypic proportions met with the expected proportions. In order to determine the normal AXB-10 phenotype to aid in the inevitable classification of curly X10.S F2 animals, the exterior phenotype of 7 homozygous normal AXB-10 animals at one month of age were examined as outlined above for the SELH animals. See appendix Table C for a list of these animals. Segregation and sex effects To test for a sex effect on the segregation and penetrance of the curly mutation the independence of the curly phenotype and sex were tested. If the curly allele segregated equally and independently with regard to the sex chromosome it would be expected that equal numbers of males and females would be curly homozygotes. With crosses generating curly homozygous (cur/cur) animals, therefore, the sexes of the phenotypically curly animals were tallied and tested for a 1:1 segregation ratio. Sex was scored based on anal-genital distance with the males having a longer distance than females and on the presence of obvious hairless patches surrounding the nipples in the female. This scoring was done on 2 week old and older animals. In addition, the sexes of the phenotypically normal animals generated in the same crosses were tested for a 1:1 47 ratio of the sexes. Litters from crosses of SELH curly homozygotes (cur/cur) with curly heterozygotes (+/cur) (n = 16 litters; 11 mating pairs, 8 with cur/cur dams; n = 136 animals) and intercrosses of X10.S curly heterozygotes (+/cur) (n = 265 minus 'probably curly' animals 9 155 and c* 158 in 25 litters) were used for this analysis. In addition, the data, obtained from D . M . Juriloff and M.J. Harris, for the L M x S E L H F2 population (n = 33 progeny minus "probably curly' animal 6* 104 in 14 litters) was used. Chi-square tests were done to see if the observed phenotypic proportions agreed with the expected proportions. H a i r To determine which hair types were affected by the curly mutation, 37 curly S E L H homozygotes (cur/cur) with ages ranging from 3 weeks to 9 months were examined. The eight hair types: pelage hairs, vibrissae, cilia (eyelashes), hair of ears, tail, and hair around the genitals, nipples and feet, were examined under a dissecting microscope in these animals and compared to the hairs observed in normal homozygotes (+/+; n = 17) and heterozygotes (+/cur; n = 15). Photos were taken of as many of these various hair types as possible using Kodak 35 mm film and a Pentax camera with a 200 mm zoom lens. Photographs of individual vibrissae and guard hairs were taken,by D . M . Juriloff, through a light microscope at a magnification of 63X and 500X. Mystacial vibrissae (Dun, 1958) and guard hair samples, from the anterior dorsal pelage between the shoulder blades, were taken from homozygous curly (cur/cur) animals (d ,L139; 6*9076) and heterozygous curly (+/cur) animals ({j*L138; 6"9081). Four or five hairs of each type were examined from each individual. These samples were affixed to slides in a pool of liquid Permount and covered with a glass cover slip. These slides were then examined with a light microscope at 25, 100 and 400X magnification. 48 Histological analysis Histological sections were examined from adult skin samples to search for visible morphological differences between cur/cur and normal skin and hair follicles. Ear pinnae and base, mid and tip of the tail samples were taken from o*7803 (cur/cur) at six months of age, 9 7854 (cur/cur) at five months and 9 7888 (+/+) at five months. Four serial sections were obtained from each of these samples. Mid-tail, pinnae, dorsal backskin and snout/vibrissae samples were collected from o"8259 (cur/cur) and c?8198 (+/+) at approximately nine months of age. Four serial sections were also obtained from these samples with the exception of the dorsal skin, for which 2x4 serial sections were obtained. Further snout/vibrissae and dorsal backskin sections were also obtained from cj*L138 (+lcur) and cfL139 (cur/cur) that were two and a half month old littermates. Two snout tissue samples were taken from both animals and embedded in different orientations. Four serial sections from each were obtained four in transverse-section and four in cross-section for each. A single dorsal backskin section from each animal was embedded and 2x4 serial sections and 12 sections taken 100 microns apart were obtained for each sample. Further pinnae samples and sections (n = 4) were obtained from c?L132 (cur/cur) and o*L76 (+/cur). Al l histological sections were examined using a dissecting scope with 8 to 40 times magnification and a light microscope with 25 times magnification. Litter raising capability It was originally thought that homozygous curly females were incapable of raising their litters based on observations by D . M . Juriloff and M.J. Harris (personal communication) that their litters died shortly after birth. Thus, attempts to breed homozygous curly females were undertaken to formally test their 'mothering' ability. 10 cur/cur SELH female animals were 49 continuously housed in pairs with +/cur and cur/cur SELH males and were closely observed for ability to bear and raise litters. Results Gross observations of phenotype a) The normal phenotype (+/+ and +lcur) The phenotypes of the homozygous normal (+/+) and heterozygous curly (+/cur) genotypes were indistinguishable from each other at the ages examined (3 weeks to 8.5 months). Their phenotype was a sleek coat with guard hairs pointing all in one direction. See figure 4 for a photograph of SELH normal guard hair. These hairs snap back into normal sleek appearance after ruffling (see materials and methods above). Guard hairs of the dorsal pelage themselves may have some slight curvature either upwards or downwards while others are perfectly straight. Normal guard hairs emerge from the skin at a small angle relative to the skin and all point in a single direction, away from the head and towards the tail. These guard hairs contribute to the overall sleek and smooth appearance of the normal coat. All normal S E L H animals have whiskers that are mostly straight with extremely few curves, even lengths and a well-fanned out spatial arrangement. Normal animals may have one or two vibrissae that have a slight overall curvature to their length that disrupts the overall evenly and widely fanned whisker arrangement but this is the exception rather than the rule. See Figure 5 for photos of SELH normal (+/+ and +lcur) whiskers at 8.5 months, 4 months and 3 weeks. This figure demonstrates the straight evenly fanned whiskers typical of the normal SELH animal. The tail, ears, eyelashes, and nose all appear to have straight, evenly sized hairs on and around them with no hair loss. The genitals and feet, however, have hairs that closely resemble zigzag hair and , therefore, they are somewhat kinked in appearance. This observation has also been noted by Dry (1926) who states that the special 1 s o 6 <4-l CN a CO O <L> ^ S CO CN t CN cn T3 CO o a tot) ed 00 o >< 3 H O H 52 c o E + CD o >> b N S O O E + O *~ ' CD 1 £ § 2 Cd £ .S x C 3 O O £ * in .2 00 j g ID m CL, ^ o O -rt C 3 CD 3 P; si + CD J 3 cd CD O 60 & O L -CD O •5 1 hJ >> W "C on 3 .. o *fi cd CD l a 3 • 53 hairs near the nipples are more like zigzag hairs as opposed to most of the hairs in the regions above which more closely resemble overhair. Table 4 summarized the numbers of normal animals examined to define and describe the normal phenotype. b) The curly phenotype (cur/cur) The first sign of abnormality in homozygous curly (cur/cur) animals was in the whiskers and was noticeable at 3 days of age. See Figure 6 for a photographic comparison of +lcur and cur/cur whiskers at three days of age. Normal whiskers at this age had a wide, regularly spaced fanning of the whiskers, while a curly animal's whiskers did not have this nice even fanning. Curly whiskers appeared to be cramped against the head and non-straight rather than fanned outwards, straight and evenly spaced. These whiskers did not appear broken or stunted. At this stage the first coat was only just becoming visible and the animals, normal and curly, appeared nude. It was not until 1 week of age that there was a noticeable difference between normal and curly guard hairs upon close inspection and it was not until two weeks of age that there was a difference in their coats from a distance. At one week the guard hairs appeared spiked and un-orderly compared to the sleek and even lengthed appearance of normal guard hairs. At 2 weeks of age, curly guard hairs were long with an unruly appearance much like the photo of the cur/cur guard hairs at three weeks in Figure 7. They appeared malleable in comparison to normal and often were non-straight or waved and many had slight hooks at their tips (see figure 8). These abnormal guard hairs appeared to be the major contributor to the overall messy, unruly and ruffled appearance of the curly animal's pelt at this time. At three weeks of age these guard hairs were straighter but were extremely malleable. In other words, upon ruffling of the coat (described in the materials and methods) the hairs of the pelage remained upright and messy and did not snap back into the sleek pattern of normal guard hairs. In comparison the normal guard 54 55 cur/cur +/cur Figure 8: Photographs of cur I cur and +/cur SELH guard hair under a microscope (x63). 57 cd C s + -o a cd k a Q JS u PC -J w CO cu o a o th 'C cd Cu O o u 3 M 58 62 63 S3 tC to <u -a &, 0 * o TS a QJ <L) cn 43 3 a, £ .8 1 s S3 cn S3 CO 3 s 2 O 43 -2 £ ccj O B -g S t cj CD »< 43 a .-a ^ I CO OJ <N *rt O S3 c3 O w u Pi! w W5 o H cu a I * o s , *» '•S. es a T3 <U S3 1 X cu T3 <U c<3 £ cu <+H o -a co T3 .3 co <-™ -2 cu o o - s cu + J c c£ <1=! <3 =tfc -a CU c e ccj X cu T3 CU -*-> at O ^ O ON * n r 1 P O = vo = r ^ c j = i / - > _ r o 0 0 ^ -O 2 cN ~ ° © o o o o o o § o o o S o o r- >/-> r- •«t r-» o T m oo vo ON oo oo ~ o o o r - - r - 4 0 o o o c n r - r ^ ' - < >nONONr~-r--> i^ONONTf r- CN o o r- r-- r-- r- r- CO ON - g CO oo 2 ON O 00 00 o o vo oo t-- >/o ~ ON 00 O O >T) cN fN ON in tN vo £ 2 2 2 § 8 vo >—i ON m m r-- i — i •H m N N tN ^ n r - I CO — 00 O CN CN •—I i — I r - H n - H m - H N r o m r n m c n r - H CO S ^ N r N r H r H N ^ S ^ ^ r H S M ^ * ^ ^ ^ ro T3 cu T3 cu T3 CU s s o c+i CU CO CU 40 3 CCj CO O CO CO '^ 3 T3 JU T< O 13 S3 T3 cci CD 60 url Ion CJ CO '3 '08 43 S3 T3 <U 60 S3 V H T3 a 43 43 CO 1> .3 jO .S £ 3 T3 Xi T3 , 0 f l B f j W o JO co 43 to CO 1 43 ccS 3 60 cu ccj S3 S3 cu ,1) CO O co 3 o <u co O S3 CU >, CU 45 "S <U T3 3 U S ftrO O CO 4^ r-3 M & CU S3 >, <U <U 60 64 hairs from three weeks onward were all slicked backwards away from the head. The overall appearance of the curly coat at this time continued to be ruffled and messy. The whiskers at three weeks had great sweeping curves and single curls, which contrasted with the straight vibrissae seen in normal animals. After three weeks the whiskers appeared to break off and had singed or frizzy looking ends (see figure 9). In other words these hairs were severely curved with a marked ripple superimposed on the general curvature at the tip of the hair shaft. Figure 14 demonstrates this whisker phenotype. As the mouse reached adulthood several other phenotypic features arose with variable frequency (these are summarized by Table 5 which is the source of the per trait percentage of animals affected of total examined). A total of 32 adult cur/cur S E L H animals with ages ranging from one to 10 months were used to determine the curly adult phenotype. These were the 38 cur/cur animals mentioned in the materials and methods minus 6 animals with ages under one month of age. Note that not all animals were examined for all traits and entries under the heading "% examined of 32 animals" indicates the percentage of animals that were examined for that particular trait out of the total 32 animals. This was because as the various aspects of the phenotype became apparent they were then included in subsequent characterizations. As the curly animal aged, its overall coat appearance progressively resembled that of normal in that it became less ruffled and sleeker in appearance. The coat appeared to get straighter after three weeks of age (see figure 10) but the guard hairs maintained their abnormal malleability and inability to snap back into position upon ruffling (100%). Figure 10 demonstrates this difference between the curly coat at three weeks in comparison to normal and the curly coat as an adult in comparison to normal. This smoothing of the coat appearance is reflected in Table 65 5 with only 81% of animals observed with a ruffled coat. While curly animals maintained the traits previously observed from birth to 3 weeks, in adulthood several new traits arose. These new traits began to appear in animals as young as one month old and most appeared with variable frequency. In a proportion of curly animals their ears tended to accumulate scabs (71%) and progressively lose hair, culminating in tissue distortion (58%; see figure 11). Curly animals' tails may have also undergone a similar process whereby hair loss (50%) and scabbiness appeared and in some cases the tip became necrotic (25%; see figure 12). In addition, curled hair was very frequently observed (88% for the pinnae and 90% for the tail) at these locations. In the ears this meant that the hairs were observed running together into clumps or the shaft actually forming a single, tight curve. In figure 11 some of these hairs can be visualized upon close inspection around the periphery of the cur/cur pinnae. Regarding affected hair on the tail of cur/cur animals, it was not that they were curled but rather they had uneven lengths and were not evenly distributed along the length of the tail. Figure 12 demonstrates this in comparison to the hairs on the normal SELH tail. Some curly animals also displayed hair loss (10%), scabbing (19%) and swollen tissue (16%) around the eyes and hair loss (6%) and scabbing (13%) around the nose. Interestingly the hind first and fifth nail also appeared to be affected by growing long and curled in some animals (29%; see Figure 13), either unilaterally or bilaterally. Curled and uneven lengths of hair were also observed on the feet (100%) but it was not clear whether or not this phenomenon was restricted to cur/cur animals as uneven lengths of hair were observed on a few normal animals as well. Some curly animals also appeared to develop calluses or small, yellow, skin hardenings on their footpads (28%). Other traits included eyelashes that appeared bent or non-normal (100%) and curled hair around the genitals (100%). These two traits, however, were only searched for in an extremely small number of animals and 66 o o c o JS "3 o '•S cd f i td JS Cu ~ CU CJ cd X cu +-> o e o c/) CD t 3 T 3 JH "I CU cd CD c« OH § O c cu JS Cu cd E CD C cd <u O £ .2 CD es c J2 O ctf -o CQ <D v —' „ CD § § e -cd o I E c O cd 3° S . U- l_ £ 'S op -5 J 3 J2 2 fi CD E § a .2 ffi 2 a 8 CD 4 S ° I 2 2 ° § U 5 '2 £ g m o rn g 1 o s E f N o £ £ ( S o s s s 2 s 2 e ^ —< a a o o 5 a a SB « O 0 0 0 0 V ~ t « O W ^ 0 0 0 0 V £ > — l O ^ O O O O O O O O M M O C o o o o o o o o o o o g 2 ° S o o o o t - t ->> c c c e c >. c c c c e c c c c s C B C >> ^ ;>>>,>.>> C G c c > . c c c c c o c >^ j^ i G ,^ ,^ c c c c c c ^ ^ >, c c >> a c e c c c c c C > > > > > . > . C C C ^ > C 5 C > . C C C C C C C > < > , > , c c > . c c > . c c c c c c c fi 3 3 o ° 8 g " « 1 1 1 1 1 1 3 .2 3 b M ^ 60 oo "5 3 U —i cU o .5 i> .5 X) .5 -o >. t _ T 3 3 T ) - O -2 SC ' 67 *2 -tj -+-» tU 43 -4-» C + H O c o 3 o C H CCJ .3. rj C O CD T3 tU a 3 S -13 CCJ tU 8. § § * tu 43 C H c3 s 3 O c ccj tU 43 43 <U cd O T3 S3 4*i S3 ctj cu S3 IS tU 1 3 O T3 C U -"H io pa 3 O cn « « -2 2 S3 £ 3 S3 O „ CTS O S3 a -3 c2 § S3° S .3 co g> cu C O V H ccj "f3 co" ' 3 ccj * E u 'a S .S J .S .. c£ ri c3 vo cu CS H total 1 u total 1 <ri r- ON ON o <r> ON O o ON ON ON ON ON o o o o NO NO NO NO \D oo vO m m NO VI r- NO NO VI VI V I VI VI lales c T 3 iffecte o o o O O o o o ^ o o O O O O O O r- <n CN CN o o NO CN o o O CN O CO O fN o o o er 3 •o u W O .a ro >r. m —i •—< o m t-H © ON O v» ON ON r-H ,_ o O o i ro m m ro m rN m * m CO VI CN CO CO OJ w PERC o o o o o 2 2 2 ° o o o ^ O 2 » 2 vo § oo 2 o o 0 0 CN Vt oo CN OO fN o o o PERC affected 1 of total examined | o o O O O 2 2 2 ^  o o 2^2 O CN V> NO o o V> TTj" fN CN OO ON fN CN o o o o T 3 u NO f- co ro in ^ ^ CN 0 0 0 0 0 0 r- O - ro <H a O •o i> % affect o^ ^  — 2 ^  <n ^ ^ rN ON Vi oo ^  0 0 fN ro m CO NO o © in E J ^ ^ ^ >, c e c d d d 0 0 -< ON o O in < H £ ^ ^ ^ c >-. c c s >. d >. oo ON ro r- <HI o I c c c c c d d d 0 0 ON o o v> s JB ^ d ^ >. c c cs c d d c 0 0 ON -* o NO E >, >^ >> c c d d d r-~ OO o o ro >• ^ c c d d d oo OO o £ B >; >; >: >; c >. >. c s d d c oo OO ON o fN S £ >~, >~> >. >• c >> >> OO oo oo o in (*« a >> >> c >> ^ >^ >> >> oo •*t o r- E a C £ ^ c C5 d d d 0 0 o co i— g a d c c c « c d >, >> 0 0 NO CO ro f- E mo >, ^ c ^ >. c = c d d c d oo NO fN o cn (—. E a >, >: >, C >^ ^ c c c c d d d oo NO ro o o & a X >^  c d d d d NO o 0 0 <4H a >. >> d d d r- vi r- O o E a ^ ^ ^ >> >, >^ a c ^ c c d >> >, >> ON vi NO O r-o S a ?^  ;>> ->» ^ ->» c c s d ->> d >» ON v> VI 1 d % M V M SP £ 1 curlec ani D. T J u *4H fl> curlec •o u •s henoty 1 urled izzy er talleabl urled air curl ;abs ssue de air curl )enails air curl aiding ssue lo; aiding tabbed aiding ;abbed uffy 1 <u d u urled h a. xi x. -a X) on x> OH Xi tS Ibody pa [coat whiskers guard ha 1 pinnae feet nails calluses Itail Inose leyes eyelashe Igenitals 68 cannot be added to the list of cur/cur traits with great certainty. Of all of the traits examined, curled whiskers (100%) with or without frizzy ends (93% with) and malleable guard hairs (100%) appeared in all cur/cur animals. While the whiskers of curly animals were all determined to be affected in comparison to normal, the severity of their curl appeared to worsen with age. Table 6.1 and 6.2 describes the traits observed for each individual animal of the 32 adult curly animals and outlines the proportion of animals displaying each trait of the curly phenotype. While these tables display the same type of data, they are separated so that the traits observed in 1 to 3 month old animals (Table 6.1) could be compared to the traits observed in 4 to 10 month old animals (Table 6.2). This was done to determine whether the frequency of any of these traits increased with age. Chi-square tests were performed to test whether any of the curly traits were observed in a greater proportion of animals 4 to 10 months old than animals 1 to 3 months old. Chi-square tests were performed as outlined in the materials and methods above. The only traits that appeared to change with age were the pinnae scabbing and tissue deformations, hind toenails, calluses and tail tissue loss. With regard to the pinnae, the proportion of animals observed with scabs increased from 36 to 100 % between the two groups (x 2 - 12.46; df = 1; p<0.05) and the proportion with tissue distortion increased from 14 to 94% (%2 = 16.95; df = 1; p<0.05). In addition, regarding the palms and toenails of the older group, 62% were found with calluses (%2 = 5.70; df = 1; p<0.05) and the toenails were found to be elongated and curled in 50% of the animals (x 2 = 6.88; df = 1; p<0.05) while these areas were found to be normal in the younger group. A large increase was also seen with age in the percentage of animals with tissue loss (from 0 to 44%; x 2 = 6.10; df = 1; p<0.05) of the tail. 69 A Chi-square analysis was performed on the data from Table 5 to see whether the proportion of traits differed between the sexes. The null hypothesis for these tests was that the relative frequencies of affected animals were the same for both sexes. No apparent difference was seen between the sexes for most traits. Where a difference was indicated, Chi-square tests were performed. In particular these were performed on the data for the coat, curled guard hairs, curled pinnae hair, nails, calluses, tissue loss of the tail and nose balding and scabbing. The hypothesis of no difference was only rejected for one trait; the appearance of scabs on the nose of curly animals with 36 % affected in females and 0% affected in males (%2 = 5.42; df = 1; p<0.05). c) Comparisons of curly vs. normal Other features of the phenotype were also examined and judged in comparison to normal. Rudimentary autopsies revealed no noticeable deviations from normal with regard to placement and size of internal organs. Similarly skull preparations revealed no noticeable deviations of the cranial skeleton and teeth from normal. Two interesting observations were made regarding the coat, however. One was that the coats of older (+6 months) cur/cur animals tended to take on an orange colour to their fur. This observation supported similar observations made by D M . Juriloff and M.J . Harris in the process of maintaining the SELH curly stock. The second interesting observation regarding the coat was that the interior surface to all of the S E L H pelts, normal or curly, shows large patches of pigmentation. This is remarkable only because the skin beneath the coat is normally unpigmented (reviewed by Budd et al., 1997). d) The AXB-10 normal phenotype Surprisingly, the SELH normal phenotype is not similar to the AXB-10 phenotype with regard to whisker and guard hair characteristics. The AXB-10 whiskers more often than not have several curves in a handful of their whiskers. These curves resemble those of a winding highway 70 03 O o c u uo c o E CU -o OJ 00 cci c/5 0 "es - C • = 1 "§ O ccj B ^ C cd cd -H cd O O 'E ^ cd - f i CD e3 • - H 71 or sine wave. See figure 15(a) for a photo demonstrating these non-straight whiskers. In addition, the guard hairs are different from those of SELH normal animals. AXB-10 normal guard hairs have more swoop or curve to their shape. This is demonstrated in Figure 15(b). Genetic transmission a) S E L H background To address the issue of the complete dominance of the wildtype allele at the curly locus, +/+ and +/cur animals (from cur/cur x +lcur crosses (n = 21 animals) and +lcur intercrosses (n = 9 animals)) were closely examined to judge if there was a visible difference between the +/+ and +lcur phenotypes. No phenotypic sub-populations could be distinguished in this group. In addition, the phenotypes of 19 (13 + 6 from D . M . Juriloff) heterozygous curly (+/cur) animals with ages ranging from 3 weeks to 7 months and 17 (4 +13) homozygous normal (+/+) animals were closely characterized. The phenotypes of these animals indicated that the wildtype allele was fully dominant with respect to the coat, guard hairs, whiskers, eyes, nose, ear pinnae, and tail hair. The various crosses between SELH curly homozygotes and curly heterozygotes were used to test the hypothesis of the recessive, single locus, 100% penetrant nature of the curly mutation. As expected based on this hypothesis, crosses (n = 8 animals from 1 litter) between curly homozygotes (curI cur) yielded 100% curly progeny. Crosses (n = 9 animals from 2 litters) between curly heterozygotes (+/cur) yielded 0 curly animals and 9 normal animals which is not statistically different from the expected 3:1 phenotypic ratio for a single locus lesion with the curly allele being recessive (%2 = 2.81; df = 1; p>0.05). Crosses (n = 140 animals from 17 litters) between curly homozygotes (cur/cur) and curly heterozygotes (+lcur) yielded 65 curly and 75 normal progeny which fits the expected 1:1 phenotypic ratio for segregation of a fully penetrant 72 recessive mutation (yj2 = 1.06; df = 1; p>0.05). The results of these crosses suggest that the curly trait is due to a fully recessive allele at a single locus (i.e. there are no major modifiers of the penetrance of the curly allele), b) Segregating background The phenotype of X10.S F ls and proportions of X10.S F2s were examined to examine whether genetic background affects the penetrance of the curly allele. The X10.S F l animals (+lcur; n= 10) all retained the whisker and guard hair phenotypes of the AXB-10 strain (see figure 14) mentioned above. In addition, the phenotypic ratios of the F2 population resulting from the X10.S F l intercross (+/cur; n = 263 animals from 25 litters) were examined to test for reduced penetrance of the cur/cur phenotype. 192 normal and 71 curly animals were scored in the F2 progeny, a phenotypic ratio which fits with the expected 3:1 phenotypic ratio for a fully penetrant recessive mutation (%2 = 0.559; df = 1; p>0.05). Thus, it appears that the curly mutation is a fully penetrant, recessive allele on both the SELH and X I OS segregating backgrounds. Segregation and sex effects For the S E L H crosses between curly homozygotes (cur/cur) and curly heterozygotes (+/cur), there were equal numbers of male (n = 39) and female (n = 34) phenotypically normal animals (which fits the expected 1:1 phenotypic ratio with %2 = 0.34; df = 1; p>0.05). Equal segregation of the sexes was not seen in the phenotypic curly group, however, with 17 females and 45 males (%2 = 12.6; df = 1; p«0 .05 ) . For the SELH (n = 9 progeny) crosses between curly heterozygotes (+/cur), a 1:1 segregation of females to males was seen with 5 normal female and 4 normal males (%2 = 0.111; df = 1; p>0.05). The X10.S F l crosses (n = 265 progeny minus ? 155 and cTl58) resulted in a 1:1 segregation of females to males in both the phenotypically 73 curly and normal populations with 35 9:36c* in the curly population (x 2 = 0.014; df = 1; p>0.05) and 95 9 :97 c* (%2 = 0.021; df - 1; p>0.05) in the normal population. In addition, when S E L H curly homozygotes (cur/cur) were crossed to AXB-10 normal homozygotes (+/+), the resulting progeny were equally distributed among the sexes with 23 females and 16 males ( x 2 = 1-26; df= 1; p>0.05)r The data obtained by D . M . Juriloff and M.J. Harris in the creation of the L M x S E L H F2 population was also equally distributed between the sexes (see table 1). The normal population was divided into 72 females and 69 males (x 2 = 0.064; df = 1; p>0.05) and curly population was divided into 17 females and 15 males (minus c*104 termed 'probably curly'; x 2 = 0.125; df = 1; p>0.05). Overall, the normal population is statistically equally distributed between the sexes and both the curly and the normal populations are equally distributed when the background has been outcrossed to either L M or AXB-10. When the curly population is on the S E L H background, however, it appears that the curly group has significantly fewer females than males. Hair When visualized with a light microscope at 100X, vibrissae and guard hair from both phenotypes, normal and curly, have obvious medullary cells (see figure 16). At the tips of the hairs there is no medulla at all. There were no gross morphological differences observable between the medullary cells of the hairs of the two genotypes (cur/cur and +lcur) other than the thin wispy or rippled tip and overall curvature of the curly hairs. This observable wispy, hook or curve, which is common to the tips of cur I cur guard hairs, can be seen in figure 8. 74 Figure 16: Photos demonstrating the vibrissae medullary cells of cur/cur and +/cur littermates magnified X500 original size. 75 Regarding the various types of mouse hairs, most of them were affected, in comparison to normal, by the curly mutation. Of the pelage hair types, the guard hairs appeared to be the most different from normal (see figures 7 and 8). However, the guard hairs were also the most easily visualized pelage hair type, which may have brought them more attention than the other pelage types. These hairs appeared to be affected in 100% of curly animals examined. Their appearance is described above in the gross observations section of the results. Zigzag hairs were also easily distinguished from the other pelage types and appeared have a similar morphology to normal zigzag hairs. The other two pelage hair types were difficult to distinguish from each other and, therefore, were observed as a group. They again did not appear to be different from normal. As mentioned above, the vibrissae were noticeably different in 100% of curly animals in comparison to normal. In one curly animal examined ( 9L75), the cilia (eyelashes) also appeared to be non-straight in comparison to normal. Overall, the cilia of the curly animals were difficult to visualize and quantify. The hair of the ears (see figure 11), tail (see figure 12), and genitals, however, were easy to visualize and appeared, as mentioned previously, to be non-straight or positively curled in comparison with normal. Hair types for which a difference was not detected between curly and normal animals, were the nipple and foot hair. The nipple hairs were difficult to visualize as separate entities from the pelage hair surrounding them due to their zigzag-like nature (Dry, 1926). Regarding the foot hair, while it did appear to be non-straight, particularly in the region surrounding the toes (see figure 17), a qualitative difference could not be found with respect to normal. Histological analysis On the whole, no gross morphological differences could be detected between normal (+/+ and +lcuf) and curly (curlcur) sections. Differences in the thicknesses of the epidermal layers 77 and/or shape, distribution, size, orientation of the hair follicles were sought. While some variation was apparent, the normal and curly samples appeared to have the same range of morphological variation and, therefore, a definitive difference between them could not be detected. It should be noted that some variation might have been imposed by the normal cycling of the thickness of the epidermis with the hair cycle for both curly and normal samples (Trigg, 1972). This may account for some of the differences seen in these histological sections (see figure 18a and 18b). a) Cheek/vibrissae follicles in cross-section It appeared that curly (cur/cur) follicles are not situated in the normal upright orientation, perpendicular to the surface of the tissue from which they were erupting. This may have been an artifact of the sectioning angle for sections because this non-perpendicular angle was not seen for all of the four blocks of curly tissue from which sections were cut. In contrast, normal (+/cur) follicles did appear to be perpendicular to the surface. Sectioning was able to show the entire follicle, from root to where the hair breaks the surface of the epidermis. There did not appear to be any gross morphological differences between the curly and normal follicles in vertical cross section. b) Cheek/vibrissae follicles in transverse-section (at the surface) In these sections, two blocks of tissue, one curly and one normal, were sectioned and compared. Both curly (cur/cur) and normal (+/cur) sections had ordered whisker patterns and round follicles. Both follicles appeared to erupt from the skin perpendicular to the surface, countering the above observation that curly vibrissae follicles may not erupt perpendicular to the surface of the epidermis. While the gross morphology appeared to differ for almost every follicle in these sections, every curly follicle appeared to have a counterpart normal follicle with much 79 • \ » * * " t 5 -^r a a Q a cd Xi T3 C cd CD T3 'B< CD 00 cd w C O s CD T3 ^ c« CD 1 | 5 Cd Q § a C - M cd >> O cd O £ S c '-3 § 2 C CD £ o ll c+3 O - * — O C3 cd o ° II <+- X i O . » uo CD T J „ SB cd E i -o c ° E uo cd l - M UO 00 cd »—< *—' CD U 3 M) 80 the same characteristics. Thus, it appeared in these sections that there are no gross morphological differences between curly and normal. See figure 18a for photos of histology sections. c) Ear pinnae in cross-section In the ear pinnae sections from normal (+/cur; n = 2 samples) samples, the hair follicles appeared to be confined to the epidermis and all lay on the same angle. This was also seen for the curly (cur/cur) samples. Unfortunately no two sections appeared to be cut on the same angle between normal and curly samples. Again, no gross morphological differences could be detected between curly and normal sections of the ear tissue when, layers of tissue, hair follicle abundance and hair follicle placement were examined. d) Tail in transverse-section In these sections, normal (+/+) and curly (cur/cur) samples appeared to resemble each other. The tail hair follicles were not visible in any of the sections. The epidermis in both sample types appeared to have the same layers and thickness. e) Dorsal skin in cross-section: In these sections normal (+/+ and +/cur) and curly (cur/cur) samples appeared to have the same range of morphological variation. Although, the sample types, seen in figure 18b, appeared to differ greatly, with the curly sample having an apparently thicker stratum corneum, different texture to the spinous and granular layers, less well defined basal layer, thinner lipid layer and hair follicles with an inverse orientation in comparison to the normal, these kinds of differences appeared to be within the realm of normal variation. Other sections depicting the opposite situations, with the normal having a thicker stratum corneum etc., were found. Therefore, no gross morphological difference could be detected. Again sections were not found with the same 81 angle of section through the hair follicles but the follicles were confined to the epidermis and leave at similar angles to the surface. Litter-raising capability When the litter-raising capabilities of curly homozygous S E L H females were more closely examined, it was found that, although a few animals neglected to raise their litters, the majority of curly females did raise their litters. In total, 10 homozygous curly (curI cur) dams, had only three out of nineteen litters that were abandoned. Seventeen of these litters had heterozygous curly (+/cur) sires and two litters, both of which were raised, had homozygous curly (cur/cur) sires. This indicates that the genotype of the pups was not likely a contributing factor in the probability that a litter would not be raised. These three abandoned litters were by three different females, one of which went on to raise two subsequent litters while the other two had no subsequent litters. For the litters that were neglected, females would give birth normally, but would leave the pups scattered about the cage and unfed. Normal maternal behavior at this stage is to clean, pile and nurse newborn pups. The presence or absence of a milk-patch (the stomach containing milk) showing through the thin skin of the newborns indicated whether or not a newborn was being fed. Neglected pups were not fed and died within a day. 82 CHAPTER IV: MAPPING CURLY Introduction The aim of this part of the study was to map the location of the curly mutation in the mouse genome. This was done to facilitate the reporting of this new mutation and the evaluation of potential candidate genes, with the ultimate goal being to find the gene responsible for the curly phenotype. This would both complete the characterization of this mutant and, if the lesion responsible for this mutation was identified, possibly contribute towards an understanding of the mechanism of mutation in the SELH strain. Rationale, Materials and Approach: Experimental design In order to map the curly mutation, homozygous cur/cur S E L H animals were outcrossed to other inbred strains to facilitate the generation of F2 animals. In the F2 generation, curly animals were then selected based on their phenotype, which was presumed to be the result of a cur I cur genotype based on the presumed recessive mode of inheritance of the cur allele. Mapping then proceeded using PCR amplification of previously mapped, informative simple sequence length polymorphisms (section on SSLPs in Chapterl) distributed at various locations in the genome to scan for linkage to the cur allele. Linkage was searched for using the theory that these F2 curly animals would have a mix of homozygous and heterozygous alleles from each of the two strains at most loci except for those closely linked to and at the curly locus. In other words, loci not linked to the cur allele would segregate independently of the curly phenotype. Near and at the curly locus, though, most or all of the curly animals would have two SELH-like alleles. This is because the curly mutation was introduced to the cross on the S E L H strain and, 83 therefore, the S E L H alleles would co-segregate with the cur mutation at loci closely linked to the cur locus. The task of mapping the curly mutation against SSLPs was carried out in four phases. The first phase was to scan a few candidate regions identified by a literature survey of hair defect mutations for evidence of linkage. This was followed by a random genome scan to find linked SSLPs. Both of these first two phases were carried out using pooled F2 curly D N A samples typed for SSLPs. The region identified as linked to the curly mutation was then refined in the third phase when individual curly F2's from the first outcross were used to narrow the linkage region. The fourth phase was the further refinement of the map position of the curly mutation by typing individual curly F2's from a second outcross. S E L H curly (cur/cur) was outcrossed initially to the inbred strain LM/Bc . The L M / B c strain was chosen for this outcross because SSLPs that were informative between the S E L H and L M strain were already known at intervals of approximately 20 c M across the entire genome (Gunn, 1995). In order to identify the candidate regions to be searched initially for linkage to the curly mutation, a literature survey of known mouse hair defect mutations was undertaken. Candidates were indicated by mutations with a similar inheritance and phenotype to the curly mutation. Table 2 summarizes the information found in this survey. The mutations that were considered to be closest to curly with regard to similar phenotype and mode of inheritance were rough (ro) and wellhaarig (we) on chromosome 2 and curly whiskers (cw) on chromosome 9. After negative linkage results were found using SSLPs linked to chromosome 2 and 9, the candidate gene approach to searching for linkage to the curly mutation was abandoned. This was because no other candidates resembled curly very closely among a large number of possible candidate sites. 84 Thus, a systematic genome scan approach was adopted thereafter. This genome scan for linkage was carried out using pooled samples of cur/cur F2 animals as well. SSLP sites from across the genome were chosen for this scan by D.Mah based on her experience in mapping another mutation using a L M x S E L H cross. The criteria she used for choosing the SSLP loci were: loci that were easily amplified by their respective primer pairs and gave ample product; loci that gave easily detected differences in product sizes between SELH and L M strains; and loci that were positioned in the middle of chromosomes so that each SSLP locus could sweep both proximal and distal adjacent regions and, therefore, produce the maximum amount of information possible. Appendix Table D lists these SSLP loci. Had this first scan produced negative results, a second scan covering the proximal and distal regions of each chromosome would have been done. The rationale behind using pooled D N A samples was to reduce the number of individual PCR reactions to be done. However, it was also desired that the relative frequency of heterozygotes (recombinants) be estimable, so that the linkage of curly to marker loci at some distance could be detected. Our lab has determined that under our conditions, a single allele of one type can be detected in a pool of up to 7 alleles of an alternate type. Thus, the maximum number of individual D N A samples added to a given pool was four. Once linkage was detected using the pooled samples, 32 individual L M x S E L H F2 D N A samples were used to characterize individual genotypes in the linked region and to construct heterozygote breakpoint maps surrounding the cur locus , thus, narrowing the region containing the cur mutation. Informative markers were found either from Gunn (1995) or by looking for differences in the amplification products of other primer pairs, (Research Genetics SSLP loci mapped to the candidate region) from L M , SELH and L M x S E L H F l DNA. Primer pairs tested for informativeness between the SELH and L M strains are listed in appendix Table E. When 85 mapping was complete with the L M x S E L H cross, there was still an approximately 6 c M region, as defined by the number of recombinants in this cross, remaining containing the cur mutation. This 6 c M region was the smallest region to which the curly mutation could be mapped with the L M x S E L H cross. Continued narrowing of this region was not possible because no further SSLPs, at least none possessed by the Juriloff/Harris lab (see appendix Table E), existed with allele sizes differing between the SELH and L M strains. This, combined with the fact that certain SSLP loci located within candidate genes had uninformative alleles in the L M x S E L H cross, spurred the creation of a second outcross F2 generation. The SSLPs referred to above are DI lMit59 and DI lMit l23 which are located within the type I keratin intermediate filament 19 (K19) gene and the hair keratin acidic 1 (HKA1) respectively. These SSLPs were of interest not only because the genes they were in were good candidates themselves but also because they belong to the type I keratin cluster which may contain numerable other candidates for the curly locus. Therefore, the mapping of the curly mutation to this keratin cluster was of keen interest, in addition to the aim of narrowing the region containing the curly gene. A second cross was set up to serve both purposes. The second outcross was to the inbred strain AXB-10/Pgn. This strain was chosen both based on its immediate availability at the time of the cross and on its predicted allele sizes in the region containing the curly mutation. This strain was determined to have different alleles than S E L H in the region of interest both by comparison of allele sizes recorded by Research Genetics and by a PCR test using D l l M i t 123, DI I M i t H , D l l M i t l 9 9 andDHMit99. The logic for choosing AXB-10 was that it was known to be like strain C57BL/6J in this region where L M is known to be like strain A/J. Therefore, a marker that is not informative in S E L H versus L M should be informative against AXB-10 (D.M. Juriloff and M.J. Harris, personal communication). 86 The second cross was carried out and F2 animals scored for further PCR based mapping of individual heterozygote breakpoints in the genotypes of curly A X B - l O x S E L H (X10.S) F2 animals. L M x S E L H cross This cross was outlined in detail in Chapter I and II. A curly (cur/cur) male on the SELH background (animal #7148) was outcrossed to a homozygous normal female (animal #7576) from the inbred strain LM/Bc . Six F l ' s resulting from this cross were then intercrossed (three mating pairs) to produce F2's. F2's were classified at three to eight weeks of age based on their whisker and thigh hair phenotype and individual samples of liver and tail tissue were taken and stored at -20°C. This first cross, the phenotype scoring, and the banking and pooling of tissue were done by D . M . Juriloff and M.J. Harris, a) Pooled sample genome screen My work on this cross began with the banked frozen liver tissue. The liver tissue from 33 L M x S E L H curly F2s had been halved and one of the halves had been pooled into groups of two to four curly F2 individuals. There were nine pools of 3, one pool of 2 and one pool of 4 individuals. Appendix Table F lists mouse identification numbers for each pooled sample. These pooled D N A samples were subject to PCR amplification of SSLPs for use as genetic markers to, initially, type candidate regions and, later, to screen the genome for linkage with the curly mutation. Linkage was sought by looking for an enriched proportion of S E L H alleles at a given marker site. The percentage of SELH alleles in each pool was estimated by comparing the brightness of the two bands on the gel, the L M allele band and the S E L H allele band. If for the marker in question, the control F l sample had two bands of equal brightness then the brightnesses of the samples bands were compared to each other. When they were equal then the 87 number of L M and SELH alleles was said to be equal. If one band appeared brighter than the other then the number of alleles was divided unequally for that sample. For example if the L M band was slightly brighter than the SELH band for a sample that had D N A from three individuals, then the pool was scored as having 4 alleles that were L M and 2 that were SELH. For a locus linked to the curly mutation, it was expected that there would be a significant deviation from the unlinked expected ratio of 50% SELH: 50% L M alleles. The closer a given marker locus is to the cur locus, the less chance there is for a recombination event and, therefore, the greater the proportion of SELH alleles. An estimate of 50 or more SELH-like alleles among the 66 alleles was taken to suggest linkage. This number was derived from a x 2 test with the null hypothesis being that the alleles are segregating equally at a given marker. The lowest number of SELH-like alleles possible to reject this null hypothesis is 47 SELH-like alleles out of 66, based on a critical %2 value corresponding to p > 0.001 to correct for the multiple tests of a genome scan. With genome scans, the 1,500 c M genome is divided up into 20 to 30 c M chunks which are tested independently (totaling 50-75 tests), therefore, the ordinary standards for x2 tests are no longer sufficient. For example, if the normal x2 critical value of p = 0.05 was used to provide the cut-off for linkage, it will only provide a probability of linkage of 44% (according to the Bayesian analysis; reviewed by Silver, 1995). This is because if the critical value of 0.05 provides evidence for linkage then 5% of all of the unlinked loci will be falsely considered linked. Using ap = 0.001 increases the probability of linkage, with our sample size, to greater than 95% (Silver, 1995). This 47 SELH-like alleles, out of a possible 66, assigned to detect linkage with a probability of 95%, corresponds to a recombinant frequency of 29% (19 recombinants out of 66). Thus, a radius of 18 c M either side of a test marker would be swept for detection of linkage between the curly locus and the test marker. In other words, there are frontiers that separate 88 significant from non-significant rates of observed recombination outside of which linkage between a marker and the test locus cannot be detected. In this example, the distance between the two outside boundaries surrounding marker locus spans 29 c M in total. Instead of 47 SELH-like alleles, however, the critical number of 50 SELH-like alleles was assigned to make up for possible human error or clarity of the gels. Assigning the number of 50 SELH-like alleles out of 66 possible alleles to indicate linkage meant that the genome was to be scanned in approximately 24 c M chunks. If 50 out of 66 alleles were SELH-like that meant that for a marker linked to the curly locus, 16 recombination events have taken place between the marker and the curly gene equalling a recombinant frequency of 24% and corresponding to a distance of 24 cM. This set the swept radius at 12 cM, which was the measure of the distance over which linkage could be detected between the curly locus and the marker being tested for linkage. With the size of the genome estimated at 1,500 cM, this meant that the entire genome could be scanned using approximately 70 markers (Silver, 1995). While the statistical approach outlined above generated guidelines with which to look for linkage to the curly mutation, it should be noted that the pooled sample technique was not intended to map the curly mutation directly. Thus the guidelines for detecting linkage were derived from the above statistical approach in order to indicate regions intended for further analysis via the resolution of individual curly F2 genotypes, b) Genotype analysis of individuals Once the chromosomal location of the curly mutation was indicated by the limited candidate approach and genome scan, 32 cur/cur L M x S E L H F2 animals (the same animals used for the pooled samples minus d1104 termed 'probably curly') were used for obtaining individual genotypes. Heterozygote breakpoints constructed from the genotypes of these individuals were 89 used to narrow the chromosomal region containing the curly gene. D N A was isolated from tail specimens taken from the same animals used for the pooled samples. One individual (#104) was omitted from genotype mapping due to a lack of confidence in the phenotyping of this individual (D. Juriloff, personal communication). Again PCR amplifications of SSLPs were used for mapping. In addition, the genotypes of 3 complete L M x S E L H F2 litters (n=38) were typed to test whether the S E L H and L M alleles were segregating in Mendelian proportions (equally) at DI I M i t H . This test was to ensure that the apparent linkage was not an artifact of segregation distortion or of preferential amplification of one allele over the other. The L M x S E L H F2 litters used for this segregation analysis are described in appendix Table G. A X B - l O x S E L H (X10.S) cross Genotype analysis of F2 cur/cur individuals from the X10.S cross was carried out in the hopes of further narrowing the region containing the curly mutation and to map the curly mutation with regard to two SSLPs known to lie within keratin genes. These SSLP are DI lMit l23 and DI lMit59, and are located within the keratin hair acidic 1 (KHA1) gene and the keratin 19 (Krt-19) gene respectively. For the second cross, two homozygous curly males on the S E L H background (c?8506 and c?8510) were outcrossed to three normal AXB-10/Pgn females (purchased from the Jackson Laboratory; 991, 9 93 and 9 111) to produce the heterozygous curly F l generation (n = 40 in 6 litters). F2s (n = 265; 25 litters from 14 breeding pairs) were obtained by intercrossing Fls . a) Scoring F2s The F2 litters were scored at approximately 3 weeks of age and curly homozygotes (n = 71+2 'probably curly' animals) were selected based on the appearance of their whiskers, coat, and individual guard hairs in their dorsal coat near the base of their tail. F2 animals were scored 90 using the definition of curly phenotype on the AXB- lOxSELH (X10.S) background described previously (page 45). Males and females were phenotyped by comparison based on anal-genital distance, with females having the smaller distance. In addition, females were categorized in young animals (around 1-2 weeks) by scoring for the visibility of nipples. No nipples were visible on male animals at this age. Tail samples from all 73 curly (+ probably curly) X10.S F2's, and normal littermates from 3 litters, were collected into cryovials on ice and immediately stored at -20°C. See appendix Table J for details of these X10.S F2 litters, b) Genotype analysis of individuals 71 cur I cur F2 animals were used to further map the cur allele within the 6 c M region to which it was mapped in the L M x S E L H cross. The two probably curly F2 animals (9155 and d 158) that were excluded from mapping analysis were later genotyped at DI I M i t H , DI lMi t l23 , and DI lMit59 to determine whether phenotyping was accurate. Again PCR amplifications of SSLPs were used for genotype analysis. Heterozygous breakpoints of individual genotypes were used to narrow the region containing the curly mutation. Molecular investigations a) SSLP and KrtlO deletion survey The lesion of the cur mutation may be a large insertion or deletion. This is based on the observation that 2 other mutations that have occurred on the SELH/Bc background are large deletions (Juriloff et al., 1994) and two separate SELH/Bc mutations are insertions of an early transposon (Hofmann et al., 1998). If curly is a sizable deletion, there is a possibility that SSLPs in the region containing the curly mutation could be deleted. Previously this has been demonstrated for markers near the tyr locus for the c B c mutation (D.M. Juriloff, personal communication). Therefore, all SSLP loci for which our lab has primers, in and surrounding the 91 DI l M i t l 4 to DI IMitIO interval, were tested for the possible deletion of an SSLP. SSLPs loci on either side of the interval of interest were also included in this survey because often SSLP loci can be incorrectly located on genetic maps (see comments in (c) Chromosome 11 map refinement below). This survey was carried out by amplifying D N A from normal SELH, normal L M , L M x S E L H F l and cur/cur L M x S E L H F2. Also, primers (F 5 ' - C C A T G T C T G T T C T A T A C A G C -3'; R 5 ' -CCTGCCCCTTAAGGTCCTCG-3 ' ; Sato et al., 1998) for a region of the Krt-10 gene were obtained from NAPS (Nucleic Acid Protein Sequencing Unit, UBC, Vancouver, BC.) and similarly tested. b) DllMitl23 double band investigation When looking at DI lMit l23 with the L M x S E L H cross it was noted that two products (bands) of approximately 320 and 380 bp, were amplified from an S E L H +/+ animal, while only one band was amplified from a L M x S E L H F2 cur/cur animal. It was thought that this unusual double band might be somehow related to the curly mutation and, hence, investigated. This was initially investigated by doing a strain survey to test whether the single band was unique to the S E L H curly mice. D N A from 6 different strains (SELH, SELHA, L G G , ICR, S E L H opaque eyes, and LM/Bc) plus SELH curly D N A were amplified with the primers for DI lMit l23 and their product sizes compared. Then a panel of 4 SELH heterozygous normal (+lcur) and 4 S E L H homozygous curly (curI cur) D N A samples were examined for DI lMit l23 products to look for possible co-segregation of the single band with the curly phenotype on the S E L H background. 92 c) Chromosome 11 map refinement The location of the 'Mit ' SSLP's on the current Research Genetics map of chromosome 11 was not very finely determined because only 92 meioses were used to map their locations. Therefore, the smallest increment that could be mapped was 1.1 c M and SSLP loci were therefore mapped to 1.1 c M groups. These groupings were found in the area of the cur mutation on chromosome 11 too, with several markers assigned the same map position. Figure 19 demonstrates the Research Genetics map in the curly region and outlines which SSLP loci had not been mapped relative to each other. Figure 19: Research Genetics map of the cur mutation region of distal chromosome 11. (Numbers represent SSLP loci that are designated by DI lMit##.) Chr 11 CP 58, 67, 98, 99, 160, 222, 328 14, 145, 197, 264 59, 123, 124, 132, 198, 249, 329, 330 52, 125, 146, 199, 200, 250, 265, 331, 332, 359, Due to the unresolved, relative map positions of markers in the region to which the curly mutation had mapped, attention was turned to several other maps of the region to obtain a better 93 idea of the positions of SSLPs and genes in the region. Figure 20 shows the comparison between three maps of chromosome 11 in the region of the curly mutation, the Mouse Genome Database (MGD) map, the European Collaborative Interspecific Mouse Backcross (EUCIB) map, and the Research Genetics map. Genetic distance is indicated in cM. Each map has different values for marker positions due to the different data sets used to generate their map positions, and, therefore, their measures of the genetic length of chromosome 11 differ as well. These three maps were compared to create a better understanding of the actual map of the cur mutation region of chromosome 11. For reviews of the generation of all three maps see Chapter I: Review of the state of the mouse map. The Research Genetics mapping project was initiated in 1991 and ended in 1996 (Dietrich, et al, 1996). Since then this map has changed little and the relative positions of many of the 7377 markers they identified remain relatively roughly mapped to islands of 1.1 c M intervals (mentioned above). The EUCIB map was last modified November 9, 1998 but localizes very few of the informative markers used in our cross. The M G D map, in contrast to the previous two maps, is a composite map derived from a wide scope of mapping data and is possibly the most comprehensive genetic map. Even on the M G D map, though, there are multiple markers mapped to single locations and the map distances delineated by this map should be tempered with the realization that the assignment of accurate genetic distances based on multiple different crosses is difficult. As all three maps fell short of giving the actual marker order of loci in the region to which the curly mutation was mapped, the genotypes of the F2 animals from both outcrosses were used to try to refine the map of the distal end of chromosome 11. The genotypes of the 33 cur I cur F2 and 31 normal F2 animals from the L M x S E L H cross and 71 cur/cur X10.S F2 animals were used to create a linkage map in the region of the cur mutation for comparison to the M G D , EUCIB 94 and Research Genetics maps. These genotypes, again resolved by PCR amplification of SSLPs, are listed in Table H and M of the appendix. 95 MGD DllMit38 Dllmit341 DllMit28 DllMit28' D l l M i t H Rar; Rim3 Re; Bda; Bsk Dl lMit l DllMit222 DllMitl24 DllMitl98 DllMit33C DllMit59 DllMitl Krtl Krtl-10 Krtl-12 Krtl-13 Krtl-14 Krtl-15 Krtl-19 Dl lMit l DllMitl2 Dl lMit l DllMit36i DllMitl61 Research Genetics c M 44.8 EUCIB cM 46.8 57.3 59.5 61.0 65.0 68.61 69.05 Figure 20: A diagram comparing three maps of chromosome 11 in the region containing the cur mutation including the M G D (black lines), EUCIB (green lines) and Research Genetics (red lines) maps. (MGD was last updated 11/16/98; EUCIB 11/09/98 and R.G. 02/5/98 (MGD1, 1998).) 96 Results LMxSELH cross a) Pooled sample genome screen 8 of the 19 mouse autosomal chromosomes were sampled before linkage of the curly mutation was indicated to chromosome 11 by the pooled samples of L M x S E L H cur I cur F2 animals. Table 7 lists in chronological order the primer pairs and chromosomes scanned with the pooled samples before finding evidence of linkage to chromosome 11. Linkage was first detected with DI lMit4 where an estimated number of SELH-like alleles were 54 (approximately 80%) out of the total number of 66 alleles. These pooled samples were then used to further narrow the candidate region containing curly before individual F2 genotype analysis was applied to confirm linkage. This was done by assessing the percentage of S E L H alleles for SSLPs both proximal and distal to DI lMit4. Proximal to DI lMit4 the estimated number of SELH-like alleles decreased while distal to DI lMit4 they increased to a maximum at DI lMi t l4 . This information is summarized in Table 7 and Figure 21 illustrates the marker order and map distances as outlined by Research Genetics. The most closely linked area was determined to be near DI I M i t H where it appeared that 95% of the alleles were SELH-like. b) Genotype analysis of individuals Once the region thought to contain the curly allele was found by pooled L M x S E L H cur/cur F2 samples, individual L M x S E L H cur/cur F2 animals were typed in the region both to confirm linkage and further narrow the region of chromosome 11 containing the curly mutation. Due to the sample size (n=64 meioses) the closest distance to the curly gene that could be measured was 1.6 c M (1 recombination event in 64 meioses = recombination frequency of 1.6). Table 7: The primer pairs and chromosomes scanned, in chronological order, in the L M x S E L H F2 pooled samples. primer chromosome estimated number of SELH alleles % S E L H alleles* D2Mit7 2 16/66 25% D 9 M i t l l 9 31/66 45% DlMit83 1 22/66 35% D7Mit62 7 25/66 40% D l M i t l 4 9 1 33/66 50% D4Mit58 4 27/66 40% D17MitlO 17 31/66 45% D12Mit5 12 32/66 50% D l l M i t 4 11 54/66 80% D l l M i t 2 0 11 48/66 75% D l l M i t 3 8 11 60/66 90% D l l M i t H 11 62/66 95% *rounded to the nearest 5% Chr 11 O c M - - D l l M i t 2 0 15.7 - D l l M i t 4 * 1 0 - -Dl lMi t320 5 2 - - D l l M i t 3 8 16.0 -D11MH14 3 2 - D l l M i t l O Figure 21: Illustration of the chromosome 11 marker order and distances as outlined by Research Genetics. 98 Figure 22 outlines the primer pairs used for this task and the groups of genotypes that were resolved from the typing of individual L M x S E L H cur I cur F2 animals. For an example of the PCR products visualized via electrophoresis on agarose gels and ethidium bromide staining see appendix Figure A. From this analysis of individual cur/cur F2 animals, the curly mutation was mapped to an approximately 6 c M region (4 recombinants in 64 meioses) between DI l M i t l 4 and DI IMitlO. The standard error (sp) of this estimate is approximately 3%; the 95% confidence interval is from 1.7 to 16.1 c M (for all statistical data, Sokal and Rohlf, 1995; Zar, 1974). The number of recombinants suggested that the curly locus was located closer to DI I M i t H (1 recombinant) than to DI IMitlO (3 recombinants). Further narrowing down of this region using the L M x S E L H cross was prevented by lack of informative markers mapping between DI I M i t H and DI IMitlO. Primer pairs that were assessed for informativeness in the L M x S E L H cross are listed in appendix Table I. The L M x S E L H F2 individual (d 104) that was excluded from the genotype mapping analysis of the curly mutation was of uncertain phenotype and categorized as "probably curly". This individual turned out to be heterozygous for the SELH allele at DI I M i t H and homozygous for the S E L H allele at DI IMitlO. As the curly mutation mapped between these two markers, the heterozygous breakpoints of d 104 could not be deduced from its genotype and, therefore, it could not be determined whether or not this animal was homozygous for the curly mutation. The genotypes of individual L M x S E L H F2 curly animals used are outlined in appendix Table H. To test for equal segregation of the SELH and L M alleles in this region, three litters of L M x S E L H F2 animals (n= 38), including both cur/cur (#10, 11, 12, 23, 24, 25, and 38) and normal littermates, were genotyped at DI lMi t l4 . Table G in the appendix outlines the animals, and their genotypes, used for this segregation analysis. The expected ratios from 38 genotypes, 99 D l l M i t 3 8 m • • • • • • DllMit341 m • • • • • • DllMi t288 m • • • • • • D11MH289 • • D l l M i t 6 7 • • Dl lMi t222 • • D l l M i t l 6 0 • • DllMi t264 • D l l M i t H n • D l l M i t l O i L _ l D l l M i t l 2 6 • • D l l M i t l 6 6 • • D l l M i t l 6 1 • • • • • • • # of animals: 23 3 1 2 1 1 1 mouse i.d.#s 124 25+66 42 12 119 Figure 22: Diagramatic representation of heterozygotic breakpoints: Fine mapping of cur allele using genotypes of L M x S E L H F2 cur/cur animals, (n = 32 cur/cur F2 animals; • = homozygous SELH; • = heterozygous S E L H / L M ; the arrows indicate the region containing the curly mutation) predicted by equal segregation of alleles at a single locus (i.e. a 1:2:1 segregation ratio), were 9.5 homozygous L M : 19 heterozygous SELH/LM: 9.5 homozygous S E L H animals. At DI lMi t l4 , 10 animals were determined to be homozygous L M , 21 animals heterozygous S E L H / L M , and 7 animals homozygous SELH. When these numbers were compared to the expected numbers that were predicted by equal segregation, they were not statistically different from 1:2:1 segregation in this region (x 2 = 0.298; df = 2; p>0.05). A X B - l O x S E L H (X10.S) cross Two of the three AXB-10 females were tested via PCR amplification at three SSLP loci in the region of the curly gene to confirm homozygosity in this region and different SSLP alleles in comparison with SELH. All were found to be homozygous at all DI I M i t H , D l l M i t l 2 3 , 100 DI 1MU99 and DI lMit l99. The AXB-10 alleles at DI lMitl4/123/199 gave different product sizes than S E L H alleles and were, therefore, informative in the X10.S cross. a) Scoring F2s In total, 265 X10.S F2 animals were collected and examined from 25 litters produced by 14 mating pairs of X10.S F l animals. From these F2's, 71 animals were identified with the curly phenotype. These animals were assumed to be homozygous cur/cur at the curly locus and were used for the X10.S mapping study. Two F2 animals (9 155 and 158) of the 265 X10.S F2's were identified as "probably curly" due to doubts regarding their phenotype. These animals were left out of the mapping study. The rest of the 192 X10.S F2 animals were identified as having normal phenotypes and were postulated to be composed of a 2:1 ratio of animals with +/cur and +/+ genotypes. See appendix Table J for the details regarding these 25 litters including their proportions of curly and normal phenotypes and animal numbers. b) Genotype analysis of individuals SSLP markers assayed for informativeness between the S E L H and AXB-10 strains are listed in appendix Table K. Figure 23 lists the SSLP loci and the groups of genotypes that were resolved from the typing of individual X10.S cur/cur F2 animals. For an example of the PCR products obtained from amplifying DI lMi t l4 to determining the genotype of the cur/cur X10.S F2 animals (n = 71) see appendix Figure A. This second analysis revealed one genotype, which occurred three times and was found to be recombinant between the curly mutation and DI I M i t H , DI lMit l98, DI lMit330, and DI lMitl24. Another genotype, that occurred three times, was found between the curly mutation and DI lMit360. These recombinants indicated that the curly locus is located 2.1 c M distal to DI IMit 14/124/198/330 and 2.1 c M proximal to DI lMit360 (Sp = 1% and 95% confidence intervals between 0.4 and 6.1 c M for both genetic 101 distances). Thus, the region containing the curly mutation was narrowed to approximately 4 c M (6 recombinants/142 meioses x 100% = 4.2 cM; standard error (sp) = 1.7 % and 95% confidence interval spanning 1.5 to 9.0 cM). In addition, no recombinants between the curly mutation and either DI lMit l23 or DI lmit59 were found in the 142 meioses examined. This indicated that the curly mutation is closer to the keratin cluster than to either DI lMitl4/124/198/330 or DI lMit360, the two closest SSLP sites identified in the X10.S cross. The genotypes of individual X10.S F2 animals used in this cross are outlined in appendix Table M . D l l M i t l 4 D l l M i t l 9 8 Dl lMi t330 D l l M i t l 2 4 D l l M i t l 2 3 (KHA1) D l l M i t 5 9 (Krtl9) Dl lMi t360 # of individuals: 65 3 3 Figure 23: Diagrammatic representation of heterozygotic breakpoints: Fine mapping of cur allele using X10.S F2 cur/cur animals, (n = 71 cur/cur F2 animals; • = homozygous SELH; • = heterozygous SELH/AXB-10; the arrow indicates the region containing the curly mutation.) Three litters of X10.S F2 animals (n= 34), including both cur/cur (#025, 026, 130, 131, 132, 133, 142, 143, and 144) and normal littermates, were genotyped at D l l M i t l 2 3 to test for Mendelian segregation of the SELH and AXB-10 alleles in the curly mutation region in the X10.S cross. The expected ratio, for 34 genotypes, predicted by equal segregation of alleles at a single locus (i.e. a 1:2:1 segregation ratio) was 8.5 homozygous AXB-10: 17 heterozygous SELH/AXB-10: 8.5 homozygous SELH. At DI lMit l23, 9 animals were determined to be homozygous AXB-10, 16 heterozygous SELH/AXB-10, and 9 homozygous SELH. When these • • • • 102 numbers were compared to the expected numbers, predicted by equal segregation, they were not statistically different from 1:2:1 segregation of alleles in this region (%2 = 0.118; the critical value for a Chi-square distribution with df = 2 and a = 0.05 being %2= 5.991). Table N in the appendix outlines the animals used for segregation analysis in this cross. The two animals (9155 and d" 158) that were excluded from the mapping studies, were later determined by SSLP markers to be homozygous for SELH alleles at DI lMit l23 and DI lMit59. These are the same two loci at which the 71 X10.S curl cur F2 animals were also homozygous SELH. This implies that these two animals were in fact homozygous cur/cur. Molecular investigations a) SSLP and KrtlO deletion survey Primer pairs amplifying SSLPs between DI I M i t H and DI IMitlO and part of the Krtl-10 locus in the region were tested using cur/cur L M x S E L H D N A to see whether any SSLPs, or any part of exon 1 of the KrtlO gene, had been deleted by the curly mutation. None of the SSLPs tested were deleted by the curly mutation. These primer pairs are listed below in Table 8 and see Figure 19 for a map of the relative positions of these markers. The sequence and optimization information regarding amplification of the KrtlO fragment is available in appendix table L . No other primer pairs amplifying SSRs within keratin genes were found except for an SSLP within Ki t 12. This keratin was not considered a candidate for the curly gene due to its corneal gene expression in mouse (see table 2), therefore, this SSLP was not tested for a deletion. In addition, to the SSLPs outlined in table 8, two other SSLPs were found without deletions. DI lMit330 and 360 were used to map the curly mutation using cur/cur X10.S F2 D N A and, therefore, were also found lacking deletions. 103 Table 8: SSLP deletion survey. (Conducted using D N A from one cur I cur individual (1054) with homozygous SELH alleles at all primer sites tested; used L M / B c +/+ (93.04.014), SELH +/+ (96.009), and L M x S E L H F l +/cur (1079 or 1080) as controls) D11MU59 1 5 mM 55 + D l l M i t l 9 9 1 5 mM 55 + D l l M i t l 9 8 1 5 mM 55 + D11MH99 1 5 mM 55 + D l l M i t 5 8 1 5 mM 55 + D l l M i t l 2 3 1 5 mM 55 + D l l M i t l 2 4 1 5 mM 55 + DllMi t200 1 5 mM 55 + D11MU265 1 5 mM 55 + DllMi t332 1 5 mM 55 + DllMi t359 1 5 mM 55 + Dl lNds7 1 5 mM 55 + D l l M i t 5 2 1 5 mM 55 + DllMi t329 1 5 mM 55 + D l l M i t l 3 2 1 5 mM 55 + D l l M i t l 4 6 1 5 mM 55 + KrtlO 1 5 mM 55 + b) D11MH123 double band investigation When the PCR products from DI lMitl23 alleles of 6 different strains plus the DI 1MH123 allele inherited with the cur/cur genotype in L M x S E L H F2s were examined, it was found that the L M x S E L H F2 homozygous curly animals were not the only genotype to have a single band at 320 bp. Table 9 summarizes the strains and the product sizes amplified from their DI lMi t l23 alleles. SELH, SELHA, L G G and L M all had two bands, at 320 bp and 380 bp respectively, while L M x S E L H curlcur and SELH opaque eyes had a single band at 320 bp. The ICR strain had a band at 310 bp which differed from the products of all of the other strains surveyed. 4 +/cur and 4 cur/cur samples of SELH D N A were also tested to examine the product sizes of their DI lMit l23 alleles to determine if the double band allele segregated with the wildtype allele. The result was that all of the SELH +lcur and cur/cur animals had only a single band of product at 320 bp. It was, therefore, concluded that the absence of the band at 380 bp was not involved in the alteration caused by the mutation in the curly allele. Table 9: Strain survey for DI lMit l23 double band investigation. Banding pattern Strain double band 320 bp (main) SELH (96.07.035); SELHA (96.10.005); L G G (93.05.010) 380 bp (faint) LM/Bc (93.04.014) single band 320 bp cur/cur L M x S E L H F2 (1054); SELH opaque (96.09.005) single band 310 bp ICR (93.05.017) single band 305 bp AXB-10 (92) b) Chromosome 11 map refinement Regarding the region surrounding the curly mutation on chromosome 11, both map refinements and confirmations of loci positioning in previous maps were made using data from the two mapping outcrosses. Figure 24 diagrams the M G D , EUCIB and Research Genetics maps in comparison to the two maps generated by the crosses in this project. Refinements were made to the map of this region of chromosome 11 by mapping the previously syntenic marker DI lMit341 (September, 1996). With the L M x S E L H cross, this marker was not found to recombine (n = 66 meioses) with DI 1MH38. While this marker was not mapped at the initiation of this study, M G D also mapped this marker to the same location as DI lMit38 (MGD1, 1998). The L M x S E L H and X I OS maps agreed with most of the various marker locations and orders on the three mouse maps, M G D , EUCIB and Research Genetics, with the following exceptions. DI lMit288 and DI lMit289 recombined in both the L M x S E L H cross in this study and in the Research Genetics panel, indicating a genetic distance of 1.5 c M and 2.1 c M respectively (1 recombinant in 66 meioses; standard error (sp) = 1.5% and 95% confidence 105 This study c M L.S X10.S M G D Research Genetics EUCIB c M c M c M 1.5 1.5 1.6 4.7 4.0 D11MU38 Dllmit341 DllMit288' DllMit28' DllMitl4 Rar Rim3 Re; Bda; Bsk. DllMitieu 5 DllMit222 DllMitl24 DllMitl98 DllMit33C DllMit59 DllMitl Krtl Krtl-10 Krtl-12 Krtl-13 Krtl-14 Krtl-15 Krtl-19 Dl lMit l DllMitl26' DllMitl6 DllMit3 DllMitl61 Figure 24: Diagram comparing maps of the cur mutation region of chromosome 11 including the M G D (black lines), EUCIB (green lines) and Research Genetics (red lines) maps for comparison with the maps generated by the two crosses for mapping curly. (L.S = L M x S E L H cross; X10.S = A X B - l O x S E L H cross; M G D was last updated 11/16/98; EUCIB 11/09/98 and R.G. 02/5/98 (MGD1, 1998).) 106 interval spanning 0 to 8.2 cM). The M G D map, in contrast, places them at the same location. The EUCIB panel did not map these markers. The X10.S cross mapped DI lMitl23 andDHMit59 apart from DI IMit 124, D l l M i t l 9 8 and DI lMit330, which the M G D and Research Genetics maps did not accomplish. The X I OS cross indicated that Dl lMi t l23 /59 were 2.1 c M (3 recombinants in 142 meioses; sp = 1.2% and 95% confidence interval spanning 0.4 to 6.1 cM) distal to DI lMitl24/198/330. For DI IMitlO and DI IMit 126, the L M x S E L H (n = 118 meioses), M G D and Research Genetics maps did not find recombinants between them. EUCIB, on the other hand, found recombinants between them with DI IMit 126 mapping 3 c M proximal to DI IMitlO. The L M x S E L H cross found these two markers together, 5.6 c M (7 recombinants in 126 meioses; sp = 2.0% and 95% confidence interval spanning 2.3 to 11.1 cM) apart from DI I M i t H . Recombinants were not found in this study between a number of SSLP loci that were found to recombine in the other maps. This difference between the maps is most likely due to chance and/or sampling sizes, as recombinants have previously been found between these loci and, thus, reflect the actual distances between SSLP loci more accurately. In the L M x S E L H cross, recombinants were not found between DI lMit289, DI lMit222 and DI lMi t l60 (n = 66 meioses) while the M G D and Research Genetics maps were able to locate DI lMit289 3.0 c M away from DI lMit222 and 6.5 c M away from DI lMit l60. None of the maps were able to resolve the relative positions of DI lMit222/160, however. Recombinants were also not found between DI l M i t l 4 and DI lMit264 in the L M x S E L H cross (n = 126 meioses) which, although it agreed with the Research Genetics map, were resolved to be 2.0 c M apart by M G D . The second cross, X10.S, also placed DI lMitl24/198/330 at the same location as DI l M i t l 4 (n = 142 107 meioses) which differs from both the M G D and Research Genetics maps that place these three markers 3.0 c M and 1.2 c M distal to D l l M i t l 4 respectively. 108 C H A P T E R V : DISCUSSION The phenotype attributable to the curly mutation appears to be novel among known hair defect mutations. This mutation appears to affect the hair shaft of the homozygous recessive mouse resulting in a grossly bent or curved overall nature and a thinned curl or ripple at the tip. Al l other aspects, particularly the medullary cells, appeared to be similar to wildtype SELH. Almost all of the hair types in the cur I cur mouse, the guard hair of the pelage, the vibrissae, eyelashes, ear hair, genital hair and tail hair, were found to be affected with the exception of the awl, auchene and zigzag hair of the pelage, foot and nipple hair. For this latter group, the nature of the hair shaft was hard to ascertain both with regard to identity and shaft quality. Based on the observations from the rest of the hair types, however, they too are likely to be affected. It appears , therefore, that the curly mutation somehow affects the quality of the hair, apparently biochemically rather than with respect to the morphological features of the medulla, cortex, cuticle, IRS or ORS, to create non-straight hairs whose structural integrity is compromised. This compromised integrity was especially noticeable at the tips of the guard hairs which became wispy and bent (see figures 8) and in the vibrissae which became greatly shortened, markedly curved and had thin, rippled ends (see figure 14c). The reason for the breaking-off of the vibrissae and loss of integrity at the tips of hairs which is not seen at other locations in the hair shaft may be due to the lack of medullary cells in these regions. Vibrissae in particular have a greater apical region that lacks medullary cells (Dry, 1926), than other cell types. This may be why the vibrissae are more strongly affected than other hair types with their tips being the only ones becoming rippled or crimped. Other hairs were found to have a reduced structural integrity moreover, with hair loss being apparent at the tail tip and surrounding the eyes and nose of some curly homozygotes. One other major change that occurs in the hair with age is the orange hue 109 that the pelage takes on in the mature animal (>6 months). This also supports a biochemical defect in the hair shaft whose phenotypic consequence is increased with age and possibly with exposure to a light source over time. Opposing a hypothesis of the hair shaft having reduced integrity however, is the disappearance of the ruffled quality of the coat with age. This contradicts the reduced integrity hypothesis because one would expect that the ruffled nature of the coat would increase with age if the pelage hairs were breaking off. The loss of severity in the coat may be explained independently of the hair-breakage hypothesis though. Trigg (1972) explained, for Rex, that the lost waviness in the adult coat was due to the limited surface area in the juvenile mouse forcing hairs to grow curved together in groups. Later, as the surface area increases in the adult, the space issue was resolved and the hairs were able to lie independently, forming an overall smoother coat. An alternate hypothesis, one that is more probable based on observations made regarding the curly phenotype, is that the guard hairs are responsible for the difference in waviness observed between the adult and three week old coat. The guard hairs are the first to emerge and reach full length (see introduction re: hair) and , therefore, are twice as long as the rest of the pelage at 3 weeks of age. Thus, in the three week old curly homozygote the guard hairs, which are markedly affected in the homozygote even from the first coat, make up the largest part of the coat giving a messy and disheveled overall appearance to the coat. This corresponds to the age at which the coat is the most ruffled and in a state of upheaval in curly homozygotes. Shortly after that time, around 4 to 5 weeks of age, the waviness of the coat is lost, corresponding to the rest of the pelage catching up to the guard hair as the entire pelage becomes fully grown. See Figure 25 for homozygous curly SELH guard hair at 3 weeks and 2 months. This hypothesis also fits the rex phenotype loss of waviness in the adult. From the I l l ed X ui in -a "o CD CU a CD e X cu cd O o 3 T3 CU +-> o ST* c ON CU 4 3 ON a. xi S O JB s i < 00 T 3 C 5 'C TO CU CU - T 3 U _g 5 2 O *i 6 £ C CU s "° C J _>> . . t-1 fN o CU l a 3 OX) • mm 112 photos of the rex pelage, it appears that the guard hairs break off in the adult coat and do not extend above the surface of the rest of the coat. Therefore, the pelage of the rex mutant can lie more normally once the affected hairs have disappeared. See Figure 26(a) for a photograph of the rex pelage. Other major abnormalities of the curly phenotype, which differ from the phenotypes of known hair defect mutations, is tail tip tissue loss, and scabbing and tissue distortion in the ear pinnae. This may be due to the length of retention of the hair by the hair follicle in these areas. Retention times of the club hair are the longest in the ear and second longest in the tail (Dry, 1926). Why hair retention, scabbing and damage to tissue may be related remains to be seen but they may be correlated. Certainly in curly homozygotes, the timing of the formation of the club hair, from 3 weeks to one month, corresponds approximately to the initiation of some of the adult onset traits, like the ear and tail traits, which increase with age. Possibly their relationship could be that an affected club hair may break off with a greater probability in areas where it is retained for a longer period of time and result in the irritation of the area. This hypothesis would fit with the observed balding followed by scabbing, tissue deformation and tissue loss in some adults homozygous for the curly mutation. By analysis of various crosses throughout this study, the curly mutation was found to be a fully penetrant, recessive, single locus mutation. While this was based on three types of crosses within the S E L H strain, only one cross was done with a large enough sample size to detect decreased penetrance. This was the cross of curly homozygotes with curly heterozygotes (n = 140 animals) yielding 1 curly: 1 normal ratios (65:75) in their progeny. The other two crosses, between curly homozygotes (n = 8) and between curly heterozygotes (n = 9), had sample sizes which were too small to detect anything but very large levels of decreased penetrance. The 113 smallest amount of decreased penetrance that would be statistically significant with a sample size of eight would be approximately 75%, corresponding to a penetrance value of 25% for the crosses between curly homozygotes. In other words six out of eight homozygous curly animals would have to not express the curly phenotype (%2 = 4.5; df = 1; p<0.05) in order to reject the hypothesis that the homozygous curly genotype is fully penetrant. For crosses between curly heterozygotes, a sample size of nine would not be able to detect any reduced penetrance. This is because if a maximum number of individuals had reduced penetrance, in other words if all nine animals did not express the curly phenotype, a hypothesis of the curly homozygous genotype being fully penetrant could still not be rejected (%2 = 3.0; df = 1; p>0.05). For crosses between curly homozygotes and curly heterozygotes however, the smallest decrease in penetrance detectable would be 17%, corresponding to a penetrance value of 83%. This would correspond to 82 animals out of 140 not expressing the curly phenotype (%2 = 4.11; df = 1; p<0.05) in order to reject the hypothesis that the homozygous curly genotype is fully penetrant. In addition, the expressivity of the curly homozygous genotype was concurrently examined in these crosses, as animal phenotypes were closely examined while they were scored for the generation of these segregation ratios. Overall, it appears that the segregation of the curly allele is not affected by the segregation of the sex chromosomes. One part of the study however, the crosses between S E L H curly homozygotes and curly heterozygotes, yielded a lower than expected number of female curly animals. This effect was not seen with the L M x S E L H and X10.S segregation data. This lower than expected number of female curly animals could be due to an overall reduced number of females in the progeny population as would be expected by human error in scoring. This is unlikely, however, as there is a Llsegregation ratio of females to males in the normal population 114 (34:39). Therefore, the reduced number of females appears to be in the curly population which, if true, would predict an overall deficiency in the number of curly animals relative to the 50% expected in this cross. The observed proportion of curly animals was 62:74 and fit with a 1:1 segregation (%2 = 1.06; df = 1; p> 0.05) of the cur/cur and +lcur genotypes and, therefore, did not appear to be deficient in curlcur animals. A deficiency of cur/cur females may be attributable, however, to the increased proportion of females that are exencephalic and die at birth in the S E L H strain (MacDonald et al., 1989). If this deficiency was due solely to curlcur females affected with exencephaly, though, this would imply an interaction in the S E L H strain between the genes for exencephaly and the curly locus. This phenomenon is, however, explainable without evoking an interaction hypothesis between these genes. With a hypothesis of no interaction, and based on documented frequencies of exencephaly in the SELH strain (MacDonald et al., 1989), then the original sample size would be 170 with 20% of those (n = 34) dying with exencephaly, leaving 136 to be scored. If then 66% of these exencephalics were female (n = 23), the documented proportion of females in a given exencephalic population (MacDonald et al., 1989), and they were divided into 14 curlcur and 9 +/cur genotypes (equally divided statistically; yl = 1.09; df = 1; p>0.05) and the male exencephalics were divided into 3 cur/cur and 8 +/cur genotypes (equally divided statistically; yl = 2.27; df = 1; p>0.05) then the sex ratios of +lcur animals would be 43 9:47 c? (equally divided statistically; yl = 0.18; df = 1; p>0.05) and the sex ratios of curlcur animals would be 31 9:48 c? (equally divided statistically; yl = 3.66; df = 1; p>0.05). These ratios would then all fit the expected 1:1 segregation of sex chromosomes. In keeping with hypothesis that it is the increased proportion of females in the exencephalic population in the SELH strain that is causing the decreased number of curlcur females resulting from S E L H crosses, is the sex chromosome segregation ratios from crosses which 'dilute' the 115 S E L H genotype. The progeny from the L M x S E L H cross and the X10.S cross did not show this decreased proportion of curly females. Mapping of new mutations is carried out primarily to provide a means for moving from diseases and phenotypes to genes, and the understanding of gene function. In addition, obtaining clones of the causative genes of human diseases can then in turn provide important tools for diagnosis, understanding and treatment. This mapping is the process of moving from a phenotype to the gene responsible and is generally referred to as positional cloning. This type of mapping occurs in two phases. The first phase is to utilize classic linkage analysis, in this case typing of SSLP loci, to find the region containing the gene of interest. The main goal of this phase is to narrow the region enough so that the D N A markers flanking the locus of interest can be determined. The identification of these flanking markers is dependent, however, on the presence and abundance of polymorphic D N A markers in the area of the gene of interest. Ideally this set of markers would be spaced an average of a few hundred kilobases apart, the goal being to find markers close enough to the gene of interest so the second phase of positional cloning can take place. This second phase is then to clone across the region, defined by the markers flanking the gene of interest, to sequence and identify the gene itself (reviewed by Silver, L . M . , 1995). With regard to the mapping of the mutation responsible for the curly phenotype, the first phase of positional cloning, outlined above, was undertaken with the hopes that the mutation would be mapped to a previously sequenced region of the genome. If it didn't map to a sequenced region of the genome then neither the curly gene nor the lesion responsible for the curly mutation would be elucidated in this study. This is because neither the resolution of the first cross (LMxSELH; smallest distance =1.6 cM) nor the second cross (XI OS; smallest distance = 0.7 cM) would be able to resolve close enough flanking markers to carry out the second phase of 116 positional cloning to elucidate the curly gene. These distances correspond to physical distances of 3200 kb and 1400 kb respectively. With these sorts of physical distances to span, no physical mapping of the curly mutation was planned even if close enough flanking markers were found. Mapping was undertaken using linkage data generated from an intercross rather than a backcross due to the increased number of meiotic events per animal. In other words there are two possibly recombinant gametes that form an F2 animal but only one possibly recombinant gamete produced from a backcross, or testcross. While F2 data increases the number of meiotic events analyzed per animal, it is more complicated to analyze because individual haplotypes can not be determined. For this study, however, haplotypes were irrelevant to determining linkage. Difficulties, however, do arise from ordering markers using F2 data because interference is negated as nearby crossover sites can be assembled from two F l parents in the F2 as stated above. Interference refers to a phenomenon where one recombination event interferes with the initiation of other recombination events in its proximity (reviewed by Silver, L . M . , 1995). This aids in the ordering of markers on genetic maps because the correct order can most often be obtained from the order indicated by the least number of recombination events. This approach cannot be used in F2 mapping panels, however, as two single recombination events cannot be differentiated from a double recombinant. As this wasn't a large scale mapping effort, however, and the distance mapped was small, the benefits of using an F2 were considered to outweigh the disadvantages. The curly mutation was mapped to an approximately 4 c M region, between DI IMit 124/198/3 30 and D l l M i t l 0/3 60, based on the X10.S cross. If this region is compared to the same region of the Mouse Genome Database (MGD) map, the smallest distance to which the curly mutation was mapped is 2 cM, between 61.0 c M (DllMitl24/198/330) and 63.0 c M 117 (DI IMitIO). It should be noted, however, that the Rim3 mutation mapping project (Sato et al., 1998) was not able to resolve DI lMit l24 and DI I M i t H in 1545 meioses. This indicates that the DI lMitl24/198/330 may have been misplaced on the M G D map and actually maps closer to DI I M i t H . This would concur with this study's data where 0 recombinants in 142 meioses were found between DI lMi t l 4 and DI lMitl24/198/330 and make the region containing the curly mutation estimated to be 6 c M by the M G D map (57-63 cM). Exactly in the middle of both of these regions on both maps are the DI lMitl23 and DI lMit59 loci, which were found not to recombine with the curly mutation in 142 meioses. This could mean that the curly locus is in one of these keratin genes (KHA1 and K19) or that it is closely linked to them. For D N A markers, not from within the locus of interest, that show no recombination with the locus of interest in 140 meioses, the distance between them is estimated to be 0.5 cM with a 95% confidence interval extending to 2.6 c M (Silver, L . M . , 1995). This still leaves the region containing the curly mutation to be an estimated 1 to 5 c M region surrounding DI lMi t l23 , DI lMit59 and the type I keratin cluster. Comparing the distances determined for the region containing the curly mutation, 4 c M for this study, 6 c M for the M G D map (see logic above) and 6 c M for the EUCIB map, it appears that recombinant frequency found in this study may underestimate the actual map distance of the region. Contrasting this, however, is the distance estimate of the region by the Research Genetics map, which estimates the region at 3 cM. This distance should be tempered with the knowledge that the sample size determining this distance is composed of only 92 meioses. An underestimate of the map distance in animals containing the curly mutation, however, could support the hypothesis of a large deletion or some kind of inversion being the lesion responsible for the curly mutation. These types of lesions would result in a decreased 118 amount of recombination because, in the case of a deletion, crossovers would be reduced by a decreased amount of D N A with which to crossover, and in the case of inversions, crossovers would produce lethal recombinants. It appears that this estimate of the genetic distance surrounding the curly mutation is a reliable estimate. Often such estimates are modulated by undetected double recombination events that lead to an underestimate of the true distance. Based on analyses in humans, it was concluded that in experiments in which fewer than 1000 meiotic events are typed, multiple recombination events within 10 c M intervals will be extremely unlikely. This phenomenon is called genetic interference (also mentioned above) and is thought to be due to a crossover on a chromosome interfering with the initiation of another crossover in its proximity. A similar degree of interference has also been observed in mice (reviewed by Silver, L . M . , 1995). This suggests that there are no double crossover recombinants between the various SSLP loci and the curly locus, and the number of recombinants observed is the actual number of recombinants in the L M x S E L H and X10.S crosses. The genotype analysis of both crosses concurs with this no double crossover premise, as there were only two individuals (i.d.# 42 and 119) whose genotypes, although they could be construed as containing double recombinants, were most likely the genotypes of individuals who had inherited a separate recombination event from each parent. Figure 22 and 23 illustrate the genotypes of the individual curly F2 animals from each cross including the above individuals. Since these are genotypes, not haplotypes, it is impossible to determine whether these genotypes are due to a double recombination event or the joining of two recombinant haplotypes. It also appears that the non-random distribution of crossover sites, mentioned above in the mouse mapping introduction, can modulate the distance estimated by linkage data. It appears, however, that these "recombination hotspots" would not effect the estimate of the region 119 containing the curly mutation. This is because the effects of hotspots on the relative distances designated by linkage data only begin to appear when the resolution of the study goes below 0.2 cM. For studies whose resolution is above this level, the location of recombination sites approaches a continuous distribution (Silver, L . M . , 1995). The region of chromosome 11 containing the curly mutation, which appears to be from 4 (data from this study) to 6 cM (MGD) in length, is a gene rich region encompassing many possible candidates for the curly locus. A 1 c M region has been estimated to be equal to 2000 kb on average (Silver, L . M . , 1995) which indicates that the region containing the curly mutation could be as large as 12,000 kb. With average genes sizes reaching 30 kb (Silver, L . M . , 1995), the number of genes in the smallest estimate of the curly region, 4 cM, could exceed 250. M G D , as outlined above, estimates this same region is 6 c M in size and it presently states that this region contains 60 known genes (MGD2, 1998), only 7 of which are from the type I keratin cluster. This means that, in addition to the 50 or so genes estimated to be in the type I keratin cluster, there could be up to 150 genes left to be found in the region to which the curly mutation has been narrowed. The loci that were considered as candidates for the curly mutation were genes affecting hair and/or skin differentiation, proliferation or structure. From this group, genes were further excluded if their mouse or human gene expression and/or disease phenotype differed greatly from the affected regions of the curly phenotype. One of the first candidates considered for the curly locus was the rex locus because it is responsible for one hair defect mutation whose phenotype is similar to the curly phenotype. This mutation again became a candidate for the curly mutation as the region responsible for the curly phenotype was mapped to the proximity of rex. If rex had been a recessive mutation, a classic complementation test could have determined whether or not rex and curly were alleles of the 120 same gene. However, curly was not tested for allelism with rex due to the dominant nature of the rex allele. Thus, the gene responsible for the rex mutation still has not been ruled out as a candidate for the curly mutation. In addition, there are the other hair defect mutations plus the other so-called alleles of rex that could also be candidates for the curly allele. It is impossible to determine by complementation test whether or not two dominant alleles are alleles of a single gene. Therefore, while they all appear to be closely linked (see page 13), it has not been determined whether or not Re, Red, Rede" and Rewc are all alleles of the rex gene. In addition, no data was found indicating that Red was an allele of the rex gene or even linked to the rex locus (Carter, 1951). The phenotype associated with the rex mutation is thought to be due to an abnormality of the internal root sheath (IRS; Trigg, 1972). This defect appears to disrupt the normal appearance of the hair and leads to a shortened appearance of the guard hairs. Unfortunately the examination of the curly hair could not determine whether or not the IRS was affected by the curly mutation. The apparently shortened hair in rex appears to contradict the candidacy of this locus for the curly mutation, however, as rex and curly guard hairs are completely dissimilar with respect to this trait (see Figure 26). While both produce a ruffled coat, the rex guard hairs are short and do not appear to be elevated above the level of the rest of the pelage, and the curly guard hairs are long and extend considerably past the rest of the pelage hairs. In light of this, it seems unlikely that rex and curly are alleles of the same gene as it is improbable that two alleles of the same gene could produce ruffled coats by totally different mechanisms. While all of the keratin genes (at the type I gene cluster) were initially candidates for the curly gene, many were later excluded due to their expression patterns and disease phenotypes (see table 3). Remaining were keratin 17 (Krtl7), keratin 22 (Krt22), the hair keratin acidic 121 (HKA) genes and the keratin associate protein (KAP) genes. Krt l7 was included due to its human disease phenotype, including kinked hair and affected nails and patterns of expression in mouse, particularly in the hair follicles and tail epidermis. Krt 22 was included due to its protein product's resemblance to hair keratin intermediate filaments and its pattern of expression in mouse, especially in the hair follicles, basal epidermis and tail. The other keratins, K H A s and KAPs remained candidates due to their role in the structural integrity of the hair fiber. It should be noted, however, that these types of keratin genes might make up to 40 of an estimated 60 possible keratin genes in the type I cluster (Powell and Rogers, 1997). K H A s encode the structural proteins that make up the 10 nm filament networks in the hair shaft while KAPs code for the matrix proteins of the hair shaft which bind and stabilize the K H A network. Defects in K A P genes or in the regulation of K A P expression have been implicated in both the mouse hair defect mutation, naked (AO, and the human hair defect disorder, trichothiodystrophy (reviewed by Rogers and Powell, 1994). The mouse mutation naked on chromosome 15 causes gross distortion of the cellular structure of hairs. The human disorder trichothiodystrophy causes abnormally brittle hair accompanying abnormal histological changes in the hair cuticle and cortex. While presently no K A P genes have been mapped to chromosome 15 in mouse, the fact that the naked locus maps to the type II keratin gene cluster on chromosome 15, and that the KAP1 and KAP2 gene families map to the type I keratin cluster on chromosome 11, indicates that other K A P genes may also be located in these regions. This increases the number of candidates for the curly locus as, between the naked mutation and trichothiodystrophy, disruption of hair structure is implied in K A P gene disruption. If the curly mutation is in a keratin gene, the lesion in the gene would most likely create a completely null allele given the recessive inheritance of the curly mutation. This lesion is implied 122 because most point mutations in keratin genes are dominant in nature due to a dominant negative action of the mutant protein. This mode of inheritance for keratin mutations is thought to be caused by mutant keratin proteins disrupting normal dimerization and network structure of the wildtype keratin proteins. This mechanism of action of the proteins encoded by these mutant alleles results in a disease phenotype with just one dose or allele of the mutation. This dominant negative hypothesis is supported by the many human and mouse keratin mutations which are inherited as dominants and the rare mutations that are complete knockouts (or null mutations) of the coding region of keratin genes that are recessive in nature. An example of this phenomenon is the wealth of dominant point mutations in Krtl-14 which cause epidermolysis bullosa simplex (EBS) (reviewed by Coulombe et al., 1991; Fuchs, 1994) and the two known recessive mutations in this gene which are null alleles (Chan et al., 1994; Rugg et al., 1994). Therefore, if the curly mutation is in a keratin gene, the mutation itself is likely a null mutation, which could take the form of a large deletion of the coding region or any type of deletion or insertion which interrupts the translation or transcription of the coding region of the gene. It is very possible that the nature of the curly mutation lesion could be a large deletion or an ETn insertion based on the S E L H mutations which are known to be large deletions (n = 2; Juriloff et al., 1994) or ETn insertions (n = 2; Hofmann et al., 1998). Both of these types of lesions could result in the creation of a null allele and be the lesion responsible for the curly mutation. Interestingly, all of the hair defect mutations in the curly region of chromosome 11, the region of the type I keratin gene cluster, namely Re, Red, Rede", Rewc Bda, Bsk and Rim3, and in the region of the type II keratin gene cluster on chromosome 15, namely N, Sha, Ve and Ca, are dominant mutations (Green, 1989). This leads to speculation that a portion of these mutations 123 may be in keratin genes or some other type of epidermal or epidermal derivative structural dimerizing protein for which dominant negative mechanisms are evoked by their mutant alleles. While keratin genes are excellent candidates for the curly locus, the region to which curly has mapped represents an extremely large physical distance and encompasses many other candidate genes presently identified plus a large number of unknown genes. Other known candidates in the region, including the retinoic acid receptor a (Rara) and Hox genes described in the introduction, map to the curly region and are involved in the regulation of epidermal cells (reviewed by Fuchs, 1995; Fuchs and Green, 1981; Saitou et al., 1995). In addition, to these genes, two other genes are known in the region that could be candidates for the curly mutation. These are the granulin (Grn) and junctional plakoglobin (Jup) genes. Granulin encodes a protein precursor which is cleaved into epithelin 1 and 2 which act to regulate proliferation of keratinocytes (Plowman et al., 1992; Shoyab et al., 1990). Jup, on the other hand, encodes a structural protein of epithelial desmosomes which are specialized cell-cell adheren junctions and are intimately involved with keratin IF networks (reviewed by Fuchs, 1995; Lazarides, 1980). A third type of gene that has also been implicated in epithelial development, the extracellular matrix receptor protein integrin alpha 3 (Itga3), has been mapped to 56.0 c M on chromosome 11, just 1 c M proximal to DI I M i t H . Itga3 plus its cousins, integrin alpha 2 and 5 which are expressed in the epidermis, all share the same beta subunit integrin beta 1. ct2, p i , ct5/pi and a2/pi transgenic mice have hair and whisker abnormalities indicating their involvement with epidermal processes (Carroll et al., 1995). While none of these genes are remarkably good candidates for curly, both because of their expression and chromosomal position, their positioning along with the Rara, Krt and Hox genes may indicate that this region of chromosome 11 has a cluster of genes controlling the structure, growth and differentiation of epithelial cells (suggested in part by 124 Sato et al., 1998). Towards this end all of the genes within 1 c M distal and 5 c M proximal to the Krt cluster on chromosome 11 were entered into a knockout database to look for phenotypes with skin or hair defects. While knockouts were found for many of these genes, none were associated with skin or hair defects (BioMedNet, 1999). Regarding future approaches to the curly mutation, further mapping of the curly locus should most likely wait until clones of this region of the genome are available. Alternatively, the fact that the curly mutation may either be an insertion or a deletion of ETn sequences could provide an approach, as yet undeveloped, for identifying the curly locus. Presently, ETn sequences are present in the genome in too high copy number (Shell et al., 1990; D. Mager, personal communication) to reveal novel insertions or deletions in the S E L H curly genome. Other possible approaches towards finding the curly locus assume, most likely incorrectly based on the number of other candidate loci in the region, that the curly mutation is in a keratin locus. The feasibility of one such approach was investigated by searching the literature for a probe that would detect type I keratin genes on a Southern Blot of genomic D N A . While probes were found to individual keratin loci, no "keratin probe" was found that would detect all type I genes. This could be because keratins are highly variable, with keratins of a single type displaying only 50 to 99% sequence identity in their rod domain, the region that is the most conserved. Furthermore, the sequences of this region that are the most highly conserved, the start of helix 1A and end of helix 2B, are conserved among all intermediate filament proteins. Based on these findings it would be extremely difficult to generate a probe that would detect the type I keratin genes solely and no such probe has been developed to date. This complication may also be why extremely large classical linkage analyses (n = 1545 and 667) were adopted over a molecular or keratin probe approach to characterize the latest hair defect mutation, Rim3, suspected to be a 125 defect in a keratin gene (Sato, H. et al., 1998). This may also be why none of the genes of the hair defect mutations mapping to the proximity of the type I and type II keratin loci on chromosome 11 and 15 in mice have been identified or evaluated for their potential to be mutations in keratin genes. In summary, the objectives of this study were to characterize the phenotype and map position of the curly mutation in order to report to the scientific community, to evaluate the mutant as a potential animal model for human genetic disorders and to, if possible, characterize the lesion responsible for the curly phenotype. Regarding these objectives, the work in this thesis has shown that the curly mutation is a novel recessive allele, with a unique phenotype caused by a defect in the hair shaft, that maps to a 4-6 cM, candidate gene-rich region of distal mouse chromosome 11. No recombinants were found between the curly mutation and two markers from within candidate keratin genes, KHA1 and Krtl9. The curly mutation's usefulness, as both a contributor to the elucidation the S E L H mechanism of mutation and an animal model for human disease, is less clear, however. Investigations into the mechanism of mutation in the SELH strain were curtailed by the curly mutation mapping to a presently un-sequenced region of the mouse genome. Therefore, the identification of the curly gene and lesion, and, thus, a further understanding of the mechanism of mutation responsible for the elevated SELH mutation rate, remains for future studies of the curly mutation. This future potential is also present regarding the curly mutation as a possible animal model for human genetic disorders. Although presently, the curly phenotype does not mimic any known human disorder, it is attributable to a single locus mutation and, therefore, may be an animal model for future or presently undocumented mutations in the human genome. 126 R E F E R E N C E S : Baribault, H. , Penner, J., Iozzo, R. V., and Wilson-Heiner, M . (1994). Colorectal hyperplasia and inflammation in keratin 8-deficient F V B / N mice. Genes and Development, 8: 2964-2973. Bell, L . , Juriloff, D. M . , and Harris, M . J. (1986). A new mutation at the cmd locus in the mouse. Journal of Heredity, 77: 205-206. Bertolino, A. P., Checkla, D. M . , Notterman, R., Sklaver, I., Schiff, T. A., Freedberg, I. M . , and DiDona, G. J. (1988). Cloning and characterization of a mouse type I hair keratin cDNA. Journal of Investigative Dermatology, 91: 541-546. BioMedNet: Mouse Knockout and Mutation Database, BioMedNet, GenomeSystemsInc. World Wide Web (http://www.biomednet.com/). February 18 (1999). Budd, P. S., Antoniou, J., Mellor, A. L. , and Jackson, I. J. (1997). Sooty foot, a novel mouse mutation that affects the pigmentation of exposed skin, but not hair, maps to chromosome 2. Mammalian Genome, 8: 631-635. Carroll, J. M . , Romero, M . R., and Watt, F. M . (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. (1951). Wavy-coated mice, phenotypic interactions and linkage tests between Rex and (a) waved-1, (b) waved-2. Journal of Genetics, 50(2): 268-277. Chan, Y.-m., Anton-Lamrecht, I., Yu, Q.-C, Jackel, A. Zabel, B. , Ernst, J.-P., and Fuchs, E. (1994). A human keratin 14 "knockout": the absence of K14 leads to severe epidermolysis bullosa simplex and a function for an intermediate filament protein. Genes and Development, 8: 2574-2587. Christiano, A. M . (1997). Frontiers in keratodermas: pushing the envelope. Trends in Genetics, 13(6): 227-233. Couloumbe, P.A., Hutton, M.E. , Letai, A., Hebert, A., Paller, A.S., and Fuchs, E. (1991) Point mutations in human keratin 14 genes of epidermolysis bullosa simplex patients: generic and functional analyses. Cell, 66: 1301-1311. Compton, J. G., Ferrara, D. M . , Yu, D.-W., Recca, V., Freedberg, I. M . , and Bertolino, A. P. (1991). Chromosomal localization of mouse hair keratin genes. Annals of the New York Acadamy of Sciences, 642: 32-43. Copeland, N . G , Jenkins, N . A., Gilbert, D. J., Eppig, J. T., Maltais, L . J., Miller, J. C , Dietrich, W. F., Weaver, A., Lincoln, S. E., Steen, R. G., Stein, L. D., Nadeau, J. 127 H. , and Lander, E. S. (1993). A genetic linkage map of the mouse: current applications and future prospects. Science, 262: 57-66. Correll, A. T., Webb, G. C , Ford, J. H. , Rogers, G. E., and Powell, B. C. (1992). Localization by in situ hybridization of a type I keratin intermediate filament gene (Krt-1.14) to band D of mouse chromosome 11. Cytogenet Cell Genet, 60: 245-246. Crew, F. A. E., and Auerbach, C. (1939). Rex: a dominant autosomal monogenic coat texture character in the mouse. Journal of Genetics, 38: 341-345. Darmon, M . , and Blumenberg, M . (1993). Molecular Biology of the Skin. The Keratinocyte., Academic Press, California, ch.l & 2. Dietrich, W., Katz, H. , Lincoln, S. E., Shin, H.-S., Friedman, J., Dracopoli, N . C , and Lander, E. S. (1992). A genetic map of the mouse suitable for typing intraspecific backcrosses. Genetics, 131: 423-447. Dietrich, W. F., Miller, J., Steen, R., Merchant, M . A., Damron-Boles, D., Husain, Z., Dredge, R., Daly, M . J., Ingalls, K. A., O'Connor, T. J., Evans, C. A., DeAngelis, M . M . , Levinson, D. M . , Kruglyak, L. , Goodman, N . , Copeland, N . G., Jenkins, N . A., Hawkins, T. L., Stein, L., Page, D. C , and Lander, E. S. (1996). A comprehensive genetic map of the mouse genome. Nature, 380: 149-152. Dietrich, W. T., Miller, J. C , Steen, R. G., Merchant, M . , Damron, D., Hahf, R , Gross, A., Joyce, D. C , Wessel, M . , Dredge, R. D., Marquis, A., Stein, L . D., Goodman, N . , Page, D. C , and Lander, E. S. (1994). A genetic map of the mouse with 4006 simple sequence length polymorphisms. Nature Genetics, 7: 220-225. Dry, F. W. (1926). The coat of the mouse (mus musculus). Journal of Genetics, 16: 287-340. Dun, R. B. (1958). Growth of the mouse coat. Australian Journal of Biological Sciences, 11: 95-105. Eicher, E. M . , and Varnum, D. (1986). Allelism of Den and Re. Mouse News Letters, 75: 29-30. Endo, H , Hatamochi, A. , and Shinkai, H. (1997). A novel mutation of a leucine residue in coil 1A of keratin 9 in epidermolytic palmoplantar keratoderma. Journal of Investigative Dermatology, 109: 113-115. Epstein, E. H . (1996). The genetics of human skin diseases. Current Opinion in Genetics and Development, 6: 295-300. 128 EUCIB: The European Collaborative Interspecific Mouse Backcross, UKMRCHuman Genome Mapping Project Resource Centre, Medical Research Council, U K . World Wide Web (http://www.hgmp.mrc.ac.uk/MBx/MBxHomepage.html). November 7 (1998). Filion, M . , Sarafian, V. , Lussier, M . , Belanger, C , Lapointe, L. , and Royal, A. (1994). Arrangement of a cluster of three mouse type I keratin genes expressed sequentially during esophageal-type epithelial cell differentiation. Genomics, 24: 303-310. Fraser, A. S. (1951). Growth of the mouse coat. Journal of Experimental Zoology, 117: 15-29. Fuchs, E. (1994). Intermediate filaments: structure, dynamics, function, and disease. Annual Review of Biochemistry, 63: 345-382. Fuchs, E. (1995). Keratins and the skin. Annual Review of Cellular Developmental Biology, 11: 123-153. Fuchs, E. (1996). Genetic disorders of keratins and their associated proteins. Journal of Dermatological Science, 13: 181-192. Fuchs, E. (1997). Of mice and men: genetic disorders of the cytoskeleton. Molecular Biology of the Cell, 8: 189-203. Fuchs, E., and Coppock, S. M . (1981). Two distinct classes of keratin genes and their evolutionary significance. Cell, 27: 75-84. Fuchs, E., and Coulombe, P. A. (1992). Of mice and men: genetic skin diseases of keratin. Cell, 69: 899-902. Fuchs, E., and Green, H . (1981). Regulation of terminal differentiation of cultured human keratinocytes by vitamin A. Cell, 25: 617-625. GeneMap98: GeneMap98: A new gene map of the human genome: The international R H mapping consortium, National Center for Biotechnology Information, National Library of Medicine. World Wide Web (http://www.ncbi.nlm.nih.gov/genemap98/). February 18 (1999). Green, M . C. (1989). Catalog of mutant genes and polymorphic loci. Oxford University Press, Oxford. Gruneberg, H . (1952). Skin and Hair Structure. Genetics of the Mouse. Martinus Nijhoff, The Hague, Netherland, 87-121. 129 Gruneberg, H. (1969). Threshold phenomena versus cell heredity in the manifestation of sex-linked genes in mammals. Journal of Embryology and Experimental Morphology, 22(2): 145-179. Gunn, T. (1995). Genetic and developmental studies of abnormal neural tube closure in SELH/Bc mice, Ph.D., University of British Columbia, Vancouver, British Columbia. Gunn, T. M . , Juriloff, D. M . , Vogl, W., Harris, M . J., and Miller, J. E. (1993). Histological study of the cranial neural folds of mice genetically liable to exencephaly. Teratology, 48: 459-471. Hanukoglu, I., and Fuchs, E. (1982). The cDNA sequence of a human epidermal keratin: divergence of sequence but conservation of structure among intermediate filament proteins. Cell, 31: 243-252. Hanukoglu, I., and Fuchs, E. (1983). The cDNA sequence of a type II cytoskeletal keratin reveals constand and variable structural domains among keratins. Cell, 33: 915-924. Harris, M . J., Juriloff, D. M . , Gunn, T. M . , and Miller, J. E. (1994). Development of the cerebellar defect in ataxic SELH/Bc mice. Teratology, 50. 63-73. Hart, C. P., Dalton, D. K., Nichols, L. , Hunihan, L., Thomas H, R., Langley, S. H , Taylor, B. A., and Ruddle, F. H. (1988). The Hox-2 homeo box gene complex on mouse chromosome 11 is closely linked to Re. Genetics, 118: 319-327. Hatzfeld, M . , Maier, G., and Franke, W. W. (1987). Cytokeratin domains involved in heterotypic complex formation determined by in-vitro binding assays. Journal of Molecular Biology, 197: 237-255. Hofmann, M . , Harris, M . , Juriloff, D., and Boehm, T. (1998). Spontaneous mutations in SELH/Bc mice due to insertions of early transposons: molecular characterization of null alleles at the nude and albino loci. Genomics, 52: 107-109. Ichinose, Y. , Hashido, K. , Miyamoto, H , Nagata, T., Nozaki, M . , Morita, T., and Matsushiro, A. (1989). Molecular cloning and characterization of cDNA encoding mouse cytokeratin no. 19. Gene, 80: 315-323. Innis, M . A. , Gelfand, D. H , and Sninsky, J. J. (1990). PCR protocols: a guide to methods and applications. , Academic Press, San Diego, 3-20. Joh, G.-Y., Traupe, H , Metze, D., Nashan, D., Huber, M . , Hohl, D , Longley, M . A., Rothnagel, J. A., and Roop, D. R. (1997). A novel dinucleotide mutation in keratin 10 in the annular epidermolytic ichthyosis variant of bullous congenital ichthyosiform erythoderma. Journal of Investigative Dermatology, 108: 357-361. 130 Juriloff, D. M , Harris, M . J., Wong, V., and Miller, J. E. (1992). Studies of a spontaneous lethal mutation at the albino locus in SELH/Bc mice. Genome, 35: 342-346. Juriloff, D. M . , MacDonald, K. B., and Harris, M . J. (1989). Genetic analysis of the cause of exencephaly in the SELH/Bc mouse stock. Teratology, 40: 395-405. Juriloff, D. M . , Porter, S. D., and Harris, M . J. (1994). Three spontaneous mutations at the albino locus in SELH/Bc mice. Genome, 37: 190-197. Kaytes, P. S., McNab, A. R., Rea, T. J., Groppi, V., Kawabe, T. T., Buhl, A. E., Bertolino, A. P., Hatzenbuhler, N . T., and Vogeli, G. (1991). Hair-specific keratins: characterization and expression of a mouse type I keratin gene. Journal of Investigative Dermatology, 97: 835-842. Knapp, B. , Rentrop, M . , Schweizer, J., and Winter, H . (1986). Nonepidermal members of the keratin multigene family: cDNA sequences and in situ localization of the mRNAs. Nucleic Acids Research, 14(2): 751-763. Lavker, R. M . , Sun, T.-T., and Sundberg, J. P. (1994). Hair follicle stem cells. Handbook of Mouse Mutations with Skin and Hair Abnormalities: Animals Models and Biomedical Tools. J. P. Sundberg, ed., CRC Press, Inc., U.S.A., 29-38. Lazarides, E. (1980). Intermediate filaments as mechanical integrators of cellular space. Nature, 283: 249-256. Liu, C.-Y. , Zhu, G , Converse, R., Kao, C. W . - C , Nakamura, H. , Tseng, S. C . - G , Mui, M . - M . , Seyer, J., Justice, M . J., Stech, M . E., Hansen, G. M . , and Kao, W. W.-Y. (1994) . Characterization and chromosomal localization of the cornea-specific murine keratin gene K r t l . 12. Journal of Biological Chemistry, 269(40): 24627-24636. Lloyd, C , Yu, Q. -C, Cheng, J., Turksen, K , Degenstein, L. , Hutton, E., and Fuchs, E. (1995) . Defining K15 function in the absence of K14. Journal of Cell Biology, 129(5): 1329-1344. Lussier, M . , Filion, M . , Compton, J. G., Nadeau, J. H , Lapointe, L. , and Royal, A. (1990). The mouse keratin 19-encoding gene: sequence, structure and chromosomal assignment. Gene, 95: 203-213. Luther, P. G. (1949). Enzymic maceration of skeletons. Proc. Linn. Soc, 161: 146-147. Lyon, M . F., and Zenthon, J. F. (1986). Close linkage of bareskin (Bsk) and rex (Re). Mouse News Letter, 74: 96. 131 Lyon, M . F., and Zenthon, J. F. (1987). Relationship between bareskin, denuded and rex. Mouse News Letter, 77: 125. MacDonald, K B . , Juriloff, D M . , and Harris, M.J. (1994) Developmental study of neural tube closure in a mouse stock with a high incidence of exencephaly. Teratology, 19:195-213. Mann, S. J. (1962). Prenatal formation of hair follicle types. Anatomical Record, 144: 135-142. McCormick, M . B., Coulombe, P. A., and Fuchs, E. (1991). Sorting out IF networks: consequences of domain swapping on IF recognition and assembly. Journal of Cell Biology, 113(5): 1111-1124. McLean, W. H . I., Rugg, E. L., Lunny, D. P., Morley, S. M . , Lane, E. B., Swensson, O., Dopping-Hepenstal, P. J. C , Griffiths, W. A. D., Eady, R. A. J., Higgins, C , Navsaria, H . A., Leigh, I. M . , Strachan, T., Kunkeler, L. , and Munro, C. S. (1995). Keratin 16 and keratin 17 mutations cause pachyonychia congenita. Nature Genetics, 9: 273-278. M G D 1 : Mouse Genome Database, Mouse Genome Informatics, The Jackson Laboratory, Bar Harbour, Maine. World Wide Web (http://www.informatics.jax.org/). November 11 (1998). MGD2: Mouse Genome Database, Mouse Genome Informatics, The Jackson Laboratory, Bar Harbor, Maine, (http://www.informatics.jax.org/). November 16 (1998). Moll , R., Franke, W. W., and Schiller, D. L. (1982). The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell, 31: 11-24. Morasso, M . I., Jamrich, M . , and Sargent, T. D. (1993). The homeodomain gene xenopus distal-less-like-2 (Xdll-2) is regulated by a conserved mechanism in amphibian and mammalian epidermis. Developmental Biology, 162: 267-276. Nadeau, J. H. , Berger, F. G., Cox, D. R., Crosby, J. L. , Davisson, M . T., Ferrara, D., Fuchs, E., Hart, C , Hunihan, L. , Lalley, R. A., Langley, S. H , Martin, G. R., Nichols, L. , Phillips, S. J., Roderick, T. H , Roop, D. R., Ruddle, F. H . , Skow, L. C , and Compton, J. G. (1989). A family of type I keratin genes and the homeobox-2 gene complex are closely linked to the rex locus on mouse chromosome 11. Genomics, 5: 454-462. Nadeau, J. H. , Compton, J. G., Giguere, V., Rossant, J., and Varmuza, S. (1992). Close linkage of retinoic acid receptor genes with homeobox and keratin encoding genes on paralogous segments of mouse chromosomes 11 and 15. Mammalian Genome, 3: 202-208. 132 Nakamura, S., Stock, D. W., Wydner, K. L., Bollekens, J. A., Takeshita, K. , Nagai, B. M . , Chiba, S., Kitamura, T., Freeland, T. M . , Zhao, Z., Minowada, J., Lawrence, J. B. , Weiss, K. M . , and Ruddle, F. H. (1996). Genomic analysis of a new mammalian distal-less gene: DIx7. Genomics, 38: 314-324. Panteleyev, A. A., Paus, R., Wanner, R., Nurnberg, W., Eichmuller, S., Zhang, R. T. J., Henz, B. M . , and Rosenbach, T. (1997). Keratin 17 gene expression during the murine hair cycle. Journal of Investigative Dermatology, 108: 324-329. Parsons, Y . M . , Piper, L . R., and Cooper, D. W. (1994). Linkage relationships between keratin-associated protein (KRTAP) genes and growth hormome in sheep. Genomics, 20: 500-502. Plowman, G.D., Green, J.M., Neubauer, M.G. , Buckley, S.D., McDonald, V . L . , Todaro, G.J., and Shoyab, M . (1992). The epithelin precursor encodes two proteins with opposing activities on epithelial cell growth. Journal of Biological Chemistry, 267(18): 13073-13078. Porter, R. M . , Leitgeb, S., Melton, D. W., Swensson, O., Eady, R. A. J., and Magin, T. M . (1996). Gene targeting at the mouse cytokeratin 10 locus: severe skin fragility and changes of cytokeratin expression in the epidermis. Journal of Cell Biology, 132(5): 925-936. Powell, B. C , and Rogers, G. E. (1990). Cyclic hair-loss and regrowth in transgenic mice overexpressing an intermediate filament gene. E M B O Journal, 9(5): 1485-1493. Powell, B. C , and Rogers, G. E. (1997). The role of keratin proteins and their genes in the growth, structure and properties of hair. EXS, 78: 59-148. Price, J. A., Bowden, D. W., Wright, J. T., Pettenati, M . J., and Hart, T. C. (1998). Identification of a mutation in DLX3 associated with tricho-dento-osseous (TDO) syndrome. Human Molecular Genetics, 7(3): 563-569. Robinson, G. W., and Mahon, K. A. (1994). Differential and overlapping expression domains of Dlx-2 and Dlx-3 suggest distinct roles for Distal-less homeobox genes in craniofacial development. Mechanisms of Development, 48: 199-215. Rogers, G. E., and Powell, B. C. (1994). Hair follicle keratins. Handbook of Mouse Mutations with Skin and Hair Abnormalities: Animals Models and Biomedical Tools. J. P. Sundberg, ed., CRC Press, Inc., U.S.A., 105-116. Rothnagel, J. A., and Roop, D. R. (1994). Epidermal keratins. Handbook of Mouse Mutations with Skin and Hair Abnormalities: Animals Models and Biomedical Tools. J. P. Sundberg, ed., CRC Press, Inc., U.S.A., 95-104. 133 Rugg, E.L. , McLean, W.H.I., Lane, E.B., Pitera, R. McMillan, JR. , Dopping-Hepenstal, P.J .C, Navsaria, H.A., Leigh, I.M., and Eady, R.A.J. (1994). A functional "knockout" of human keratin 14. Genes and Development, 8: 2563-2573. Saiki, R. K. , Gelfand, D. H , Stoffel, S., Scharf, S. J., Higuchi, R., Horn, G. T., Mullis, K. B., and Erlich, H. A. (1988). Primer-directed enzymatic amplification of D N A with a thermostable D N A polymerase. Science, 239: 487-491. Saiki, R. K. , Scharf, S., Faloona, F., Mullis, K. B., Horn, G. T., Erlich, H . A. , and Arnheim, N . (1985). Enzymatic amplification of -globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science, 230: 1348-1354. Saitou, M . , Sugai, S., Tanaka, T., Shimouchi, K., Fuchs, E., Narumiya, S., and Kakizuka, A. (1995). Inhibition of skin development by targeted expression of a dominant-negitive retinoic acid receptor. Nature, 374: 159-162. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N Y . Sato, H. , Koide, T., Masuya, H. , Wakana, S., Sagai, T., Umezawa, A., Ishiguro, S.-i., Tama, M . , and Shiroishi, T. (1998). A new mutation Rim3 resembling Re(den) is mapped close to retinoic acid receptor alpha (Rara) gene on mouse Chromosome 11. Mammalian Genome, 9: 20-25. Scherer, S. W., Heng, H . H. Q., Robinson, G. W., Mahon, K. A., Evans, J. P., and Tsui, L . -C . (1994). Assignment of the human homolog of mouse Dlx3 to chromosome 17q21.3-q22 by analysis of somatic cell hybrids and fluorescence in situ hybridization. Mammalian Genome, 6(4): 310-311. Schlager, G., and Dickie, M . M . (1966). Spontaneous mutation rates at five coat-color loci in mice. Science, 151: 205-206. Searle, A .G. 1968. New mutant. Mouse News Letter 39: 25. Shaw, D. (November 16, 1998). Personal communication. Mouse Genome Informatics User Support, The Jackson Laboratory, Bar Harbor, Maine. Sheehan, D. C , and Hrapchak, B. B. (1980). Theory and practice of histotechnology. C V Mosby Co, Toronto. Shell, B E . , Collins, J.T., Elenich, L .A. , Szurek, P.F., and Dunnick, W.A. (1990). Two subfamilies of murine retrotransposon ETn sequences. Gene 86: 269-274. Shoyab, M . , McDonald, V .L . , Byles, C , Todaro, G.L., and Plowman, G.D. (1990). Epithelins 1 and 2: Isolation and characterization of two cysteine-rich growth-134 modulating proteins. Proceeds of the National Academy of Sciences, 87(20): 7912-7916. Silver, L . M . (1995). Mouse Genetics: Concepts and Applications, Oxford University Press, New York. Smith, F. J. D., Corden, L. D., Rugg, E. L. , Ratnavel, R., Leigh, I. M . , Moss, C , Tidman, M . J., Hohl, D., Huber, M . , Kunkeler, L. , Munro, C. S., Lane, E. B., and McLean, W. H . I. (1997). Missence mutations in keratin 17 cause either pachyonychia congenita type 2 or a phenotype resembling steatocystoma muliplex. Journal of Investigative Dermatology, 108: 220-223. Sokal, R. R., and Rohlf, F. J. (1995). Biometry: the principles and practice of statistics in biological research, W.H. Freeman, San Francisco. Steinert, P. M . , Idler, W. W., Zhou, X . - M . , Johnson, L. D., Parry, D. A. D., Steven, A. C , and Roop, D. R. (1985). Structural and functional implications of amino acid sequences of keratin intermediate filament subunits. Annals of the New York Academy of Sciences, 455: 451-461. Steinert, P. M . , Rice, R. H., Roop, D. R., Trus, B. L. , and Steven, A. C. (1983). Complete amino acid sequence of a mouse epidermal keratin subunit and implications for the structure of intermediate filaments. Nature, 302(28): 794-800. Stellmach, V. , Leask, A., and Fuchs, E. (1991). Retinoid-mediated transcriptional regulation of keratin genes in human epidermal and squamous cell carcinoma cells. Proceeds of the National Academy of Science, 88: 4582-4586. Sundberg, J. P., and Hogan, M . E. (1994). Hair types and subtypes in the laboratory mouse. Handbook of Mouse Mutations with Skin and Hair Abnormalities: Animals Models and Biomedical Tools, J. P. Sundberg, ed., CRC Press, Inc., U.S.A., 57-68. Tobiasch, E., Winter, H. , and Schweizer, J. (1992). Structural features and sites of expression of a new murine 65 kD and 48 kD hair-related keratin pair, associated with a special type of parakeratotic epithelial differentiation. Differentiation, 50: 163-178. Trigg, M . J. (1972). Hair growth in mouse mutants affecting coat texture. Journal of Zoology, 168: 165-198. Vassar, R., Coulombe, P. A., Degenstein, L. , Albers, K., and Fuchs, E. (1991). Mutant keratin expression in transgenic mice causes marked abnormalities resembling a human genetic skin disease. Cell, 64: 365-380. 135 Wallace, M . E., and Ferguson, J. M . (1984). Linkage for bald-arthritic, Bda. Mouse News Letter, 71: 19. Wilson, B. W., Edwards, K. J., Sleigh, M . J., Byrne, C. R., and Ward, K. A. (1988). Complete sequence of a type I microfibrillar keratin gene. Gene, 73: 21-31. Winter, H. , Siry, P., Tobiasch, E., and Schweizer, J. (1994). Sequence and expression of murine type I hair keratins mHa2 and mHa3. Experimental Cell Research, 212: 190-200. Zar, J. H . (1974). Biostatistical Analysis. Prentice Hall, New Jersey. 136 A P P E N D I X Table A: SSLP primer pairs whose optimum conditions differed from T a n n e a | = 55°C and/or [Mg 2 +] = 1.5 mM. SSLP \Mg2+](mM) T anneal. D7MM62 2.5 55 D1MKT49 3.5 55 D l l M i t l 7 3.5 55 D l l M i t 4 2.5 60 D11MU38 3.5 55 D l l M i t l 4 3.5 55 DllMit263 3.5 50 D11MH222 1.5 60 D l l M i t l 6 6 1.5 60 D l l M i t l 6 0 1.5 60 Dl lMi t98 1.5 60 Dl lMit264 1.5 60 D l l M i t l 2 7 1.5 60 Table B: SELH/Bc mice used for gross observations to determine the homozygous normal (+/+), heterozygous curly (+/cur) and homozygous curly (cur/cur) phenotypes. (d.o.b. = date of birth; d.o.d. = date of death) sex animal # genotype d.o.b. d.o.d. approx. age d L41 +/+ 27xii97. 19i98 3 weeks e L42 +/+ 27xii97 19i98 3 weeks d1 L43 +/+ 27xii97 19i98 3 weeks d1 L44 +/+ 27xii97 19i98 3 weeks 9 L45 +/+ 27xii97 19i98 3 weeks 9 L46 +/+ 27xii97 19i98 3 weeks 9 L47 +/+ 27xii97 19i98 3 weeks 9 L48 +/+ 27xii97 19i98 3 weeks d1 L49 +/+ 29xii97 19i98 3 weeks d1 L50 +/+ 29xii97 19i98 3 weeks cf L51 +/+ 29xii97 19i98 3 weeks 9 L52 +/+ 29xii97 19i98 3 weeks 9 L53 +/+ 29xii97 19i98 3 weeks 9 8683 +/+ 1x97 3iii98 5 months e 7889 +/+ 6v96 22x96 6 months 9 7888 +/+ 6v96 22x96 6 months d1 8198 +/+ 19xii96 8ix97 8.5 months ? L103 +/cur 20iv98 23iv98 3 days 137 ? L31 -T-1 cur 4xii97 24xii97 3 weeks 9 L32 -T-1 cur 4xii97 24xii97 3 weeks 9 L33 +lcur 4xii97 24xii97 3 weeks 9 L34 +lcur 4xii97 24xii97 3 weeks e L35 +lcur 4xii97 24xii97 3 weeks d L38 +lcur 4xii97 24xii97 3 weeks 9 L55 +lcur 8i98 9iv98 3 months d L76 +lcur 8i98 15v98 4 months d 9081 -T-1 cur 30iii98 l l ix98 5 months d 9082 +/cur 30iii98 l l ix98 5 months d 9083 -T-1 cur 30iii98 l l ix98 5 months 9 7802 -T-1 cur 21iii96 27ix96 6 months d 7855 +lcur 13v96 21x96 6 months d 8736 +lcur 18viii97 23ii98 7 months d 7765 +/(+ or cur) 27i96 27ix96 8 months ? L102 cur/cur 20iv98 23iv98 3 days d L36 curlcur 4xii97 24xii97 3 weeks d L37 curl cur 4xii97 24xii97 3 weeks d L39 curl cur 4xii97 24xii97 3 weeks d L40 curlcur 4xii97 24xii97 3 weeks d L99 curl cur 3iv98 23iv98 3 weeks d L I 16 curlcur 26iii98 23iv98 1 month d L70 curlcur 2ii98 lli i i98 1 months 9 L129 curlcur 2iv98 15v98 1.5 months 9 L130 curlcur 2iv98 15v98 1.5 months 9 L131 curlcur 2iv98 15v98 1.5 months d L132 curlcur 2iv98 15v98 1.5 months d L133 curlcur 2iv98 15v98 1.5 months d L134 curlcur 2iv98 15v98 1.5 months d 8740 curlcur 28xii97 26H98 2 months d L78 curlcur 8i98 ll i i i98 2 months d L67 curlcur 2ii98 23iv98 2 months d L57 curlcur 8i98 9iv98 3 months 9 L74 curlcur 2ii98 15v98 3 months d L68 curlcur 2ii98 15v98 3 months 9 L75 curlcur 8i98 15v98 4 months d 9076 curlcur liv98 ll ix98 5 months d 9077 curlcur liv98 ll ix98 5 months 9 7854 curlcur 13v96 22x96 5.5 months d 7803 curlcur 21iii96 27ix96 6 months d 8732 curlcur 18vii97 23ii98 6 months d 8733 curlcur 18viii97 23ii98 6 months d 8734 curlcur 18viii97 23ii98 6 months 138 cf 8574 curlcur 14vii97 3iii98 7 months 9 8738 curlcur 18viii97 ll i i i98 7 months 9 7764 curlcur 27i96 27ix96 8 months cf 8259 curlcur 31xii96 8ix97 8 months cf 8331 curlcur 12iv97 8xii97 8 months 9 8330 curlcur 12iv97 8xii97 8 months cf 8506 curlcur 16vi97 23ii98 8 months cf 8510 curlcur 29v97 23ii98 9 months 9 8509 curlcur 29v97 3iii98 9 months 9 8573 curlcur 13vi97 lli i i98 9 months Table C: AXB-10 animals used to determined the normal (wildtype) AXB-10 phenotype. (d.o.b. = date of birth; d.o.d. = date of death) sex animal # genotype d.o.b. d.o.d. age 9 148 +/+ 30x97 3xii97 1 month cf 149 +/+ 30x97 3xii97 1 month cf 150 +/+ 30x97 3xii97 1 month 9 151 +/+ 30x97 3xii97 1 month cf 152 +/+ 30x97 3xii97 1 month 9 153 +/+ 30x97 3xii97 1 month cf 154 +/+ 30x97 3xii97 1 month Table D: Research Genetic PrimerPairs SSLP loci chosen by D.Mah for initial scan of genome in search of curly locus. Loci that were actually used are marked with an asterisk. Chr Name Chr Name Chr Name 1 D1MU231 6 D6MH102 13 D13Mit39 1 D lMi t l 70 7 D7Mit62* 14 D14Mit5 1 DlMit83* 8 D8Mit30 15 D15Mit29 1 DlMi t l49* 9 D9Mit89 16 D16Mitl lO 2 D2Mit7* 9 D 9 M i t l l * 17 D17MitlO* 2 D2Mit92 9 D9Mit81 18 D18Mit36 3 D3Mit28 10 D10Mitl58 19 D19Mit5 4 D4Mit58* 11 DI lMit4* 5 D5Mit24 12 D12Mit5* 139 Table E: Primer pairs tested for informativeness between SELH and L M strains for both pooled sample and individual F2 mapping of the curly mutation. SSLP locus TMe2+l status DI lmi t l l7 1.5 mM 55 uninformative DI lmit7 2.5 mM 55 abandon DI lmitl94 1.5 mM 55 uninformative DI lmit280 2.5 mM 55 haven't got working DI lmit322 1.5 mM 55 uninformative DI lmitl23 1.5 mM 55 uninformative DI lmitl80 3.5 mM 55 abandon DI lmit40 1.5 mM 55 uninformative DI lmit54 1.5 mM 55 informative DI lmit67 1 .5P&EmM 55 h/s .25ul DNA; somewhat readable; informative DI lmitl32 1.5 mM 55 uninformative DI lmit59 1.5 mM 55 uninformative DI lmitl99 1.5 mM 55 uninformative DI lmit224 1.5 mM 55 uninformative DI lmi t l l 9 1.5 mM 55 uninformative DI Llmitl45 all mM 55 & 60 abandon D] Umitl98 1.5 mM 55 uninformative DI Umit222 1.5 mM 60 informative DI lmit263 3.5 mM 50 informative DI lmit330 all mM 55 uninformative DI Llmit285 1.5 mM 55 uninformative D] Llmit286 1.5 mM 55 uninformative DI Llmit288 1.5 mM 55 informative DI Umit289 1.5 mM 55 informative DI Umit58 1.5 mM 55 uninformative DI Umit99 1.5 mM 55 uninformative DI llmitl98 1.5 mM 55 uninformative DI Llmit360 1.5 mM 55 uninformative DI lmitl24 1.5 mM 55 uninformative DI Umitl46 1.5 mM 55 h/s uninformative DI Umit200 1.5 mM 55 uninformative DI Umit265 1.5 mM 55 uninformative D] llmit332 1.5 mM 55 uninformative D] llmit359 1.5 mM 55 uninformative D] Llmitl26 1.5 mM 55 informative DI 1 lNds7 1.5 mM 55 uninformative D] Umit52 1.5 mM 55 uninformative D1 limit 166 1.5 mM 60 hard to read; informative D] Umitl60 1.5 mM 60 informative D] llmit329 1.5 mM 55 uninformative D l l K y o l 1.5 mM 55 uninformative Dllmit98 all mM 60 informative needs opt Dl lmi t l97 all mM 60 haven't got working Dllmit264 1.5 mM 60 informative Dllmit258 1.5 mM 55 uninformative Dl lmi t l61 1.5 mM 55 informative Dllmit223 1.5 mM 55 uninformative D l l m i t l 3 1.5 mM 55 uninformative Dl lmi t l47 1.5 mM 55 uninformative Dllmit201 1.5 mM 55 uninformative Dllmit341 1.5 mM 55 informative Dl lmi t l80 1.5 mM 55 informative needs opt Dllmit224 1.5 mM 55 uninformative Dl lmi t l46 1.5 mM 55 uninformative; shd h/s Dl lmi t l27 1.5 mM 60 uninformative 141 Table F: L M x S E L H cur/cur F2 pooled samples: mouse numbers and information. Sample # mouse #s parents generation phenotype 1068 10 05x06 F2 curly 11 05x06 F2 curly 12 05x06 F2 curly 1069 23 01x02 F2 curly 24 01x02 F2 curly 25 01x02 F2 curly 1070 38 03x04 F2 curly 50 01x02 F2 curly 51 01x02 F2 curly 1071 40 03x04 F2 curly 41 03x04 F2 curly 42 03x04 F2 curly 1072 48 03x04 F2 curly 49 03x04 F2 curly 75 03x04 F2 curly 1073 63 05x06 F2 curly 64 05x06 F2 curly 115 03x04 F2 curly 1074 65 05x06 F2 mild curly 66 05x06 F2 mild curly 67 05x06 F2 mild curly 68 05x06 F2 mild curly 1075 69 05x06 F2 mild curly 104 01x02 F2 probably curly* 120 03x04 F2 curly 1076 91 05x06 F2 curly 92 05x06 F2 curly 97 05x06 F2 curly 1077 119 03x04 F2 curly 123 01x02 F2 curly 124 01x02 F2 curly 1078 128 01x02 F2 curly 133 01x02 F2 curly 1079 01 F l 1080 02 F l * This animal not used for individual haplotype analysis for narrowing the cur region. Table G: Three litters of L M x S E L H F2 litters including both cur/cur and normal littermates. Genotyped at DI 1MU14 to confirm equal segregation of alleles in region (S= SELH-like allele; L= LM-like allele). Parents j? o* d.o.b. animal # sex phenotype genotype at D l l M i t l 4 05 06 30i96 01 02 03 04 9 10 11 12 31196 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 2ii96 sex cf cf d d 9 9 9 9 9 9 d 9 9 9 9 9 d d d d d d 9 d d d d d d d 9 9 9 9 9 9 9 cf normal normal normal normal normal normal normal normal normal curly curly curly normal normal normal normal normal normal normal normal normal normal curly curly curly normal normal normal normal normal normal normal normal normal normal normal normal curly SL L L SL SL SL L L SL SL SL SS SS SS L L SL SL L L SL L L SL SL L L L L SS SS SS SL L L SL SL L L SL SL SL SL L L SL SL SS 143 Table H: The genotypes of individual L M x S E L H cur/cur and normal F2 animals used for genotype analysis of heterozygote breakpoints to map the curly mutation and to refine the chromosome 11 'Mit ' SSLP map (S= SELH-like allele; L= LM-like allele). Primer pair D11 Mit Animal # 38 341 288 289 67 222 160 14 264 126 10 166 161 10 SS SS SS SS SS SS SS SS SS SS SS SS SS 11 SS SS SS SS SS SS SS SS SS SS SS SS SS 12 SL SL SL SL SL SL SL SS SS SS SS SS SS 23 SS SS SS SS SS SS SS SS SS SS SS SS SS 24 SS SS SS SS SS SS SS SS SS SS SS SS SS 25 SS SS SS SS SS SS SS SS SS SS SS SS SL 38 SS SS SS SS SS SS SS SS SS SS SS SS SS 40 SS SS SS SS SS SS SS SS SS SS SS SS SS 41 SS SS SS SS SS SS SS SS SS SS SS SS SS 42 SL SL SS SS SS SS SS SS SS SS SS SS SL 48 SS SS SS SS SS SS SS SS SS SS SS SS SS 49 SS SS SS SS SS SS SS SS SS SS SS SS SS 50 SS SS SS SS SS SS SS SS SS SS SS SS SS 51 SS SS SS SS SS SS SS SS SS SL SL SL SL 63 SS SS SS SS SS SS SS SS SS SS SS SS SS 64 SS SS SS SS SS SS SS SS SS SL SL SL SL 65 SS SS SS SS SS SS SS SS SS SS SS SS SS 66 SS SS SS SS SS SS SS SS SS SS SS SS SL 67 SS SS SS SS SS SS SS SS SS SS SS SS SS 68 SS SS SS SS SS SS SS SS SS SS SS SS SS 69 SS SS SS SS SS SS SS SS SS SS SS SS SS 75 SS SS SS SS SS SS SS SS SS SS SS SS SS 91 SS SS SS SS SS SS SS SS SS SS SS SS SS 92 SS SS SS SS SS SS SS SS SS SS SS SS SS 97 SS SS SS SS SS SS SS SS SS SS SS SS SS 104 SL SL SL SL SL SL SL SL SL SS SS SL SL 115 SS SS SS SS SS SS SS SS SS SS SS SS SS 119 SL SL SL SL SL SL SL SL SL SS SS SS SL 120 SS SS SS SS SS SS SS SS SS SL SL SL SL 123 SS SS SS SS SS SS SS SS SS SS SS SS SS 124 SL SL SL SS SS SS SS SS SS. SS SS SS SS 128 SS SS SS SS SS SS SS SS SS SS SS SS SS 133 SS SS SS SS SS SS SS SS SS SS SS SS SS normal 1 SL SL SL SL SL normal 2 LL LL LL LL LL normal 3 SL SL SL SL SL 144 normal 4 normal 5 normal 6 normal 7 normal 8 normal 9 normal 13 normal 14 normal 15 normal 16 normal 17 normal 18 normal 19 normal 20 normal 21 normal 22 normal 26 normal 27 normal 28 normal 29 normal 30 normal 31 normal 32 normal 33 normal 34 normal 35 normal 36 normal 37 SL SL SL SL SL SL SL SL SL SL SL LL LL LL LL LL SL SL SL SL SL SL SL SL SL SL SL SL SL SL LL LL LL LL SL SL SL SL SL SL SL SL SL LL LL LL LL LL SL SL SS SS SS LL LL LL LL LL SL SL SL SL SL SL SL SL SL SL LL LL LL LL LL LL LL LL LL LL SL SL SL SL SL LL LL LL LL LL SL SL SL SL SL SL SL SL SL SL LL LL LL LL LL SL SL SL SL SL SL SL SL SL SL SL SL SL SL SL SL SL SL SL LL LL SL SL SL SL SL SL SL SL SL 145 Table I: Primer pairs tested for informativeness in L M x S E L H cross for the possibility of finer mapping between DI IMit 14 and DI IMitlO. (Excludes primer pairs used to generate above recombinant break point map. Bars to the right of the primer pairs indicate primer pairs that have been mapped to the same location but have not been mapped in relation to each other.) Primer pair [Mg2+1 Tanneal status D l l M i t 9 8 1.5 mM 60 informative but needs optimization D l l M i t l 4 D l l M i t l 45 all mM 55 & 60 abandoned D l l M i t l 9 7 all mM 55&60 abandoned D l l M i t 5 8 1.5 mM 55 uninformative D l l M i t 9 9 1.5 mM 55 uninformative D l l M i t 5 9 1.5 mM 55 uninformative D l l M i t l 2 3 1.5 mM 55 uninformative D l l M i t l 2 4 1.5 mM 55 uninformative D l l M i t l 3 2 1.5 mM 55 uninformative D l l M i t l 9 8 1.5 mM 55 uninformative Dl lMi t249 not recommended* Dl lMi t329 1.5 mM 55 uninformative Dl lMi t330 1.5 mM 55 uninformative D l l M i t l 2 5 not recommended* D l l M i t l 4 6 1.5 mM 55 (hotstart) uninformative D l l M i t l 9 9 1.5 mM 55 uninformative Dl lMi t200 1.5 mM 55 uninformative Dl lMi t250 not recommended* Dl lMi t265 1.5 mM 55 uninformative Dl lMit331 not recommended* Dl lMi t332 1.5 mM 55 uninformative Dl lMi t359 1.5 mM 55 uninformative D l lNds7 1.5 mM 55 uninformative D l l M i t 5 2 1.5 mM 55 uninformative D l l M i t l O 1.5 mM 55 uninformative Dl lMi t360 1.5 mM 55 uninformative * = not recommended; allele size not polymorphic across inbred strains as outlined in Research Genetics MapPairs literature. 146 Table J: X10.S F2 litters: mating pairs and phenotyping data. Parents born scored # pups cur/cur normal animal (3wks/born) female male female male numbers 005x009 6i98 27i98 10/12 0 1 6 3 036-045 006x008 6i98 27i98 11/1 2 0 4 5 000; 025-035 002x010 7i98 28i98 9/1 1 1 4 3 046-054 004x009 7i98 28i98 6/0 1 0 1 4 055-060 003x010 7i98 28i98 11/1 1 2 5 3 061-071 001x010 13i98 3ii98 9/0 2 1 4 2 072-080 007x008 18i98 9ii98 9/0 1 0 4 4 081-088 017x024 19i98 9ii98 10/2 1 0 3 6 089-098 015x022 20i98 9ii98 12/1 2 0 6 4 099-110 014x021 20i98 10H98 8/2 0 0 3 5 111-118 013x020 20i98 10ii98 11/2 1 2 6 2 119-129 016x023 23i98 13U98 11/0 2 2 4 3 130-140 012x019 21i98 13H98 11/0 1 2 6 2 142-152 011x018 22i98 13H98 9/0 2(+l) 2(+l) 1 2 153-161 004x009 13ii98 4iii98 11/4 1 3 2 5 190-200 011x018 14H98 4iii98 13/1 3 2 3 5 177-189 012x019 13ii98 4iii98 3/0 1 1 0 1 201-203 003x022 18ii98 lli i i98 13/l 2 1 7 3 204-216 007x008 23H98 16iii98 14/0 2 2 5 5 217-230 016x023 24ii98 17iii98 15/0 2 3 6 4 231-245 012x019 4iii98 18iii98 9/0 2 0 4 3 246-254 017x024 liii98 18iii98 12/0 2 2 0 8 255-266 002x010 3iii98 18iii98 14/0 1 3 3 6 267-279 014x021 liii98 18iii98 13/0 2 3 5 3 280-292 004x009 ll i i i98 25iii98 12/3 0 3 3 6 314-325 TOTALS: 36 37 95 97 73(711 265 192 147 Table K: Primer pairs that were tested for informative SSLP alleles between S E L H and AXB-10 strains. Primer rM g2+i Tanneal Status D l l m i t l 4 3.5 mM 55 informative Dl lmi t l23 1.5 mM 55 informative Dllmit99 1.5 mM 55 maybe informative D l lmi t l 99 1.5 mM 55 informative Dllmit59 1.5 mM 55 informative Dllmit360 1.5 mM 55 informative D l lmi t l 24 1.5 mM 55 informative Dl lmi t l 32 1.5 mM 55 uninformative Dl lmi t l98 1.5 mM 55 informative Dllmit330 1.5 mm 55 informative Table L: Information regarding PCR primer pair designed to amplify exon 1 of Krt 10 on chromosome 11. Sequence [Mg2+] T a n n ^ Status ~ product size F = 5 ' -CCATGTCTGTTCTATACAGC-3 ' 1.5 Mm 55 uninformative 600 bp R = 5 ' -CCTGCCCCTTAAGGTCCTCG-3 ' 148 Table M: The genotypes of individual X10.S cur/cur and normal F2 animals used for genotype analysis of heterozygote breakpoints to map the curly mutation and to refine the chromosome 11 primer pair map (S= SELH-like allele; A= AXB-10-like allele). D11Mit 14 198 330 124 123 59 360 Animal# 025 SS SS ss ss 026 SS SS ss ss 036 SS SS ss ss 046 SS ss ss ss 047 SS ss ss ss 055 SS ss ss ss 061 SS ss ss ss 062 SS ss ss ss 063 SS ss ss ss 072 SA SA SA SA ss ss ss 073 SS ss ss ss 074 SS ss ss ss 081 SA SA SA ss ss ss 089 SS ss ss ss 099 SS ss ss ss 100 SS ss ss ss 119 SS ss SS SS ss ss ISA | 120 ss ss ss ss 121 ss ss ss ss 130 ss ss ss ss 131 ss ss ss ss 132 ss ss ss ss 133 ss ss ss ss 142 ss ss ss ss 143 ss ss ss ss 144 ss ss ss ss 153 ss ss ss ss 154 ss ss ss ss 155* ss ss 156 ss ss ss ss 157 ss ss ss ss 158* ss ss 177 ss ss ss ss 178 ss ss ss ss 179 ss ss ss ss 180 ss ss ss ss 181 ss ss ss ss 190 SS SS SS SS 191 SS SS SS SS 192 SS SS SS SS 193 SS SS SS SS 201 SS SS SS SS 202 SS SS SS SS 204 SS SS SS SS 205 SS SS SS SS 206 SS SS SS SS 217 SS SS SS SS 218 SS SS SS SS 219 SS SS SS SS 220 SS SS SS SS 231 SS SS SS SS 232 SS SS SS SS 233 SS SS SS SS 234 SS SS SS SS 235 SS SS SS SS 246 SA SA SA SA SS SS SS 247 SS SS SS SS 255 SS SS SS SS 256 SS SS SS SS 257 SS SS SS SS 258 SS SS SS SS 267 SS SS SS SS 268 SS SS SS SS SS SS S A | 269 SS SS SS SS 270 SS SS SS SS 280 SS SS SS SS 281 SS SS SS SS 282 SS SS SS SS 283 SS SS SS SS SS SS SA 284 SS SS SS SS 314 SS SS SS SS 315 SS SS SS SS 316 SS SS SS SS 150 Table N: Three litters of X10.S F2 litters including both cur/cur and normal littermates. Genotyped at DI lMit l23 to confirm equal segregation of alleles in this region (S= SELH-like allele; A= AXB-10-like allele). Parents j? cf d.o.b. animal # sex phenotype genotype at D l l M i t l 2 3 006 008 016 023 012 019 6i98 000 025 026 027 028 029 030 031 032 033 034 035 23i98 130 131 132 133 134 135 136 137 138 139 140 21i98 142 143 144 145 146 147 148 149 150 151 152 sex 9 9 9 9 9 9 9 C ? cf cf cf cf 9 9 cf cf 9 9 9 9 cf cf cf 9 cf cf 9 9 9 9 9 9 cf cf normal curly curly normal normal normal normal normal normal normal normal normal curly curly curly curly normal normal normal normal normal normal normal curly curly curly normal normal normal normal normal normal normal normal A A SS SS A A A A A A SA SA SA SA A A SA SS SS SS SS SA SA SA SA A A SA A A SS SS SS SA A A SA SA SA SA A A A A 151 jappei dq ooi i * LVZ 9PZ II III i 91 £ i sie H£ g . 1 Id mas i oi-axv i 111 i 182 082 i OLZ 69Z 892 Id HTHS oi-axv jappei dq ooi § $ $ 1 * jappEi dqooi 0Z,2 LfZi Id HIHS oi-axv jappej dq ooi jappqdq 001 1702 i Id 202 1 HIHS 102 i oi-axv £61 i L9Z 261 i 852 I6T i Z.S2 061 I 9S2 181 i gs2 081 I LVZ 6LI i 9XZ Id 5 £2 HIHS :| fr£2 oi-axv HIHS 8Z.I " i oi-axv LL\ 1 ££2 LSI i 2£2 95\ i I £2 V9\ i 022 £SI i 612 w i 812 m i LIZ zn i 902 ££I 1 £02 Id Id HIHS 1 HIHS oi-axv • : $ : . oi-axv jappei dq 001 jappBidqooi 11 in § S CD Cvj PH CD <*> ,. C/) 3- ° ^ 1 If I •fl 'S £ § '3 * 5 / 3 "e 2 fl ' 3 H^ O X Jo Is E | .2 c -o T 3 2 5 o S3 O '•{3 ca o 3> c 2 CD IH CD >> CD CO *3 CD fl 60 g CD ± j 8 * § O S s • O - a ai e & » o u £ E c3 ^ g PH £ o *o CD u . -fl Its s E <+H -o CD fl •8 CH ^ P O <D <u J> o - E * -O N fl J> 2 '-3 -2 £ S s -O fl- 3 ccj fl c« cj O O > 5§ "fl . . CD S .SP < £ 3 E V •-3 Ml 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0089024/manifest

Comment

Related Items