. 4 x;
3.1.1.:
x

 

 

was

This is to certify that the
thesis entitled

ANALYSIS OF CANDIDATE GENES, PITX2, FOXC1 AND
FOXE3, FOR ANTERIOR SEGMENT DYSGENESIS IN
HORSES

presented by
Jessica A. Eason-Butler

has been accepted towards fulfillment
of the requirements for the

MS. degree in Animal Science

 

 

gtmim 4

V—

Maiqggofessiar’s Signature

12-13-02

 

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cJClRC/DateDuopGS-sz

 

ANALYSIS OF CANDIDATE GENES, PITXZ, FOXC1 AND FOXE3, FOR
ANTERIOR SEGMENT DYSGENESIS IN HORSES

By

Jessica A. Eason-Butler

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Animal Science

2002

ABSTRACT

ANALYSIS OF CANDIDATE GENES, PITX2, FOXC1 AND FOXE3, FOR
ANTERIOR SEGMENT DYSGENESIS IN HORSES

By

Jessica Ann Eason-Butler

Anterior Segment Dysgenesis (ASD) syndrome is a congenital eye defect that
occurs frequently in Rocky Mountain Horses. Anterior Segment Dysgenesis in horses
manifests itself as two phenotypes: horses affected with cysts, and horses fully affected
with the A80 syndrome. The objective of this study was to investigate the molecular
mechanisms that underlie A80 in horses. Candidate genes, PITX2, FOXC1 and FOXE3,
were chosen based on their roles in similar diseases in other species. The study
population was taken from 516 horses that were previously characterized for A80. The
study population was used in three experiments: 1) association study; 2) linkage study;
and 3) sequence variability analysis. One marker (UM11) for FOXC1 showed a
significant association between its alleles and the phenotypes when painivise
comparisons of each phenotype were analyzed. However, the linkage study showed no
linkage between this marker and the disease phenotypes. An additional marker for
FOXC1 and a marker for PITX2 were not linked or associated with the disease
phenotypes. Partial genomic sequence analysis revealed no variability in the three
genes in pools of horses (n = 10 per pool) that represented the three phenotypes.
Results of these studies do not support PITX2 and FOXE3 as candidate genes for

Anterior Segment Dysgenesis in horses, but FOXC1 warrants further investigation.

This thesis is dedimted to all of the Rocky Mountain Horse owners who were

willing to acknowledge a problem in their breed and proactively pursue it.

iii

ACKNOWLEDGEMENTS

I acknowledge a few people who have helped me along the way. I first
thank Dr. Susan Ewart for pushing and teaching. I’m grateful to her for the
knowledge she has given me. I also thank her for having a horse project that was ,
so wonderful to work with. I thank my committee members for all of their
expertise and time: Drs. David Ramsey, Simon Petersen-Jones, Cathy Ernst and
Dennis Banks.

I am also extremely grateful to Dennis Shubitowski, not only for his help in
the lab but for his friendship outside of the lab as well. He has become a good
friend and like family, too. Thank you for helping me through some of the rough
times, Dennis. I’ll never forget it.

I thank the numerous undergraduate students who have assisted me in
some way or another: Natalie Kent, Emily Susott and David Newton. Nancy
Raney from Dr. Emst’s lab was also helpful and I thank her for taking the time to
train me on the ABI 373 sequencer.

I am very grateful for my family and friends who stood by me and
supported me. I’m especially thankfulto my husband, who sacrificed some of his
own dreams so that I could pursue mine. I love you. Without you, this dream of
mine would never have happened.

I’m especially grateful to the Rocky Mountain Horse Association who has
been so willing to help us. Lots of Rocky owners have acknowledged this disease

and have been wonderful enough to let us examine their horses and collect blood

iv

samples for our DNA analysis. I now understand Rocky Moutain owners a little
better and I am a Rocky owner now because of it. I thank the association and I
thank the Rocky Mountain Horse, especially a certain one that has given me love

and a lot of trail miles... .Sadie.

TABLE OF CONTENTS

List of Tables
List of Figures

CHAPTER 1
LITERATURE REVIEW
Introduction
Ocular Anatomy
Anterior Segment Dysgenesis in Horses
Similar Diseases in Humans and Other Species
Candidate Genes
PITX2
FOXC1
FOXE3
Equine Genomics
Summary
Chapter 1 Tables and Figures

CHAPTER 2

MATERIALS AND METHODS
Study Population
Genomic DNA Isolation
Sample Organization
Comparative Mapping
DNA Microsatellite Markers
PCR Amplification
Marker Genotyping
Statistical Analysis
Linkage Analysis
Primer Design
Gene Sequencing
Radiation Hybrid Panel
Chapter 2 Tables and Figures

CHAPTER 3
MARKER ANALYSIS - GENOTYPING
Introduction
Results
Discussion
Chapter 3 Tables and Figures

vi

35
35
35
37
38
39
39
42
42
43
43

45

61
63
69
73

CHAPTER 4
GENE ANALYSIS - SEQUENCE
Introduction
Results
Discussion
Chapter 4 Tables and Figures

CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS FOR
FUTURE RESEARCH

APPENDIX A
APPENDIX 8
APPENDIX C
APPENDIX D
APPENDIX E
REFERENCES

vii

83

88
91

101

103
105
107
109
112
117

Tables

Table 1. Pools and Plate 1 Study Population

Table 2. Multibreed Pool DNA ata

Table 3. Plate 2 Study Population Data

Table 4. Marker Primer Data

Table 5. Sequence Primer Data

Table 6. SAS Results for the Association Study - HMS42

Table 7. SAS Results for the Association Study - UM11
Unaffected Horses versus Cyst Affected Horses

Table 8. SAS Results for the Association Study — UM11
Unaffected Horses versus ASD Affected Horses

Table 9. SAS Results for the Association Study - UM11
Cyst Affected Horses versus ASD Affected Horses

Table 10. SAS Results for the Association Study - A14

Table 11. Dye Wavelengths

Table 12. Dye Combinations

Table 13. Dye Availability and Filter

Table 14. Plate 1 Genotype Results for HMS42, UM11 and A14

Table 15. Plate 2 Genotype Results for HMS42 and UM11

viii

46-47
48
49-50
51
52

74

75

76

77

78

110

110

111
113-114

115-116

Figures

Figure 1. Diagram of the Eye

Figure 2. Segments and Chambers of the Eye

Figure 3. Pedigree Showing Semidominant Inheritance

Figure 4. Photograph of a Cyst of the Ciliary Body

Figure 5 and 6.’ Photographs of ASD Syndrome

Figure 7. Pedigree of the Founder Stallion

Figure 8. Horse Karyotype

Figure 9. Pedigree of Foundation Sire (V-30)

Figure 10. Pedigree of a Common Sire (VI-7)

Figure 11a-c. Comparative Map for FOXC1, PITX2, and FOXE3
Figure 12. Example of Genotyping using a 33F Radioactive Label
Figure 13. Radiographic Film of Marker Sequence

Figure 14. Electropherogram of a Fluorescent Genotype

Figure 15a and b. Pedigree of Plate 2 Showing Genotypes
for HMS42 and UM11

Figure 16. Human PITX2 Gene Structure

Figure 17. Horse PITX2 lntron 2 Sequence Compared to Human

Figure 18. Horse PITX2 Exon 6 Sequence Compared to Human
Figure 19. Human FOXC1 Gene Structure

Figure 20. Horse FOXC1 Sequence Compared to Human
Figure 21. Human FOXE3 Gene Structure

Figure 22. Horse FOXE3 Sequence Compared to Human

ix

25
26
27
28
29-30
31
32-34
53

55-57
58
59
60

79-82
92
93
94-95
96
97
98

99

Figure 23. Agarose Gel of Hamster and Horse Sequence -
FOXC1 and FOXE3 100

Chapter 1

Literature Review

Introduction

Anterior Segment Dysgenesis (ASD) syndrome is a congenital eye defect
that occurs frequently in horses with the chocolate coat color, including the
Rocky Mountain Horse. Anterior Segment Dysgenesis in horses manifests itself
as two phenotypes: horses affected with ciliary or retinal cysts, and horses fully
affected with the ASD syndrome, which is a collection of clinical signs that
include ciliary and retinal cysts, iris hypoplasia, retinal dysplasia, retinal
detachment, flattened and circumferential granula iridica, cornea globosa,
iridocomeal angle abnormalities, goniosynechiae, lens subluxations, cataracts,
microphthalmia, and other extra-ocular deformities such as telewnthus
(abnormally increased distance between the eyes) and excessive bony
protrusion of the supraorbital space (Ramsey et al. 1999a). The objective of this
study was to investigate the molecular mechanisms that underlie ASD in horses
using the Rocky Mountain Horse as a model.

Rocky Mountain Horses have gained enormous popularity as pleasure
horses because of their smooth ambling gait and gentle‘ personality. The Rocky
Mountain Horse breed registry was established in 1986. Many of these horses
descend from a single foundation sire. The small foundation stock appears to

contribute to the high prevalence of this disease.

Similar ocular diseases are well documented in humans, mice and
Drosophila such as Axenfeld-Rieger Syndrome and Peters Anomaly (human),
small eye and dysgenetic lens (mouse), and eyes absent (Drosophila). Anterior
Segment Dysgenesis in humans is a genetic defect that involves the cornea,
iridocomeal angle, iris, ciliary body, lens and retina. It occurs because of a
malformation of neural-crest-derived mesenchymal tissues, which form the
cornea, iris, iridocomeal angle, and ciliary body (Lines et al. 2002).

Anterior Segment Dysgenesis in horses was reported as a familial
abnormality with codominant inheritance (Ewart et al. 2000). To investigate the
cause of ASD in horses, we began by identifying genes that logically relate to the
disease. Candidate genes for ASD in horses were chosen based on genes that
have been researched extensively in other species, specifically mice and
humans. Mutations in FOXC1 and PITX2 have been shown to cause a disease in
humans called AxenfeId-Rieger Syndrome (Espinoza et al. 2002; Saleem et al.
2001; Semina et al. 1996). Because of the similarities between AxenfeId-Rieger
Syndrome and ASD in horses, PITX2 and FOXC1 were identified as candidate
genes. A third gene, FOXE3, was chosen basedon reports that mutations in
FOXE3 are one cause of Peters Anomaly, which is also similar to ASD in horses
(Blixt et al. 2000; Semina et al. 2001).

In a previous investigation, an eye development gene, PAX6, was
evaluated as a candidate gene for ASD. The homeobox region of PAX6 in
humans (PAX6 exon 8) was sequenced in horses to determine variability since

mutations in this region in humans can be detrimental to the binding specificity of

PAX6. The regions surrounding exon 8 (PAX6 intron 8, exon 9, intron 9, and
exon 10) were also investigated (Shubitowski et al. 2001). From this region, 852
base pairs (excluding primers) were sequenced. One microsatellite (a poly-T
repeat) and a single nucleotide polymorphism (a guanine-adenine transition)
were located in PAX6 intron 8, however there was no variability found in exon 8,
exon 9, intron 9 and exon 10 (Shubitowski et al. 2001). When the single
nucleotide polymorphism was analyzed in 76 Rocky Mountain Horses, 73 of the
76 were homozygous, guanine-guanine, and none were homozygous, adenine-
adenine. Three of the 76 were heterozygous (guanine-adenine). These results
suggested that the adenine allele occurred at relatively low frequency in the
population.

The purpose of this literature review is to describe Anterior Segment
Dysgenesis (ASD) in horses and other species, describe the candidate genes,
and provide a perspective on the status of the horse genome database. The goal
for the present study was to begin evaluating genes that control eye
development. This was accomplished by identifying known variable
microsatellites that flanked the comparable regions for the candidate genes for
ASD in horses, testing these markers using association and linkage analysis for
association or linkage to ASD, and identifying sequence differences between the

unaffected, cyst affected and fully affected horses.

Ocular Anatomy

The eye is composed of two segments, the anterior segment which
includes the cornea, iris, and ciliary body, and the posterior segment which
includes the lens, vitreous body, and the ocular fundus (Reviewed by Forrester et
al. 2002). The fundus consists of the retina, choroid, and sclera. Vitreous humor
is a gelatinous fluid that fills the vitreous cavity. Anterior Segment Dysgenesis
may affect all structures in the anterior segment (Figure 1) and also ocular
structures located in the anterior part of the posterior segment (Ramsey et al.
1999a). The two ocular segments of the eye are divided into three chambers: the
anterior chamber (in front of the iris), the posterior chamber (between the iris and
the lens) and the vitreous chamber (behind the lens) (Diesem 1968) (Figure 2).
The anterior and posterior chambers are joined through the pupil and are filled
with a clear fluid (aqueous humor), which provides nutrition to anterior segment
ocular tissues and the lens.

The most anterior portion of the eye is called the cornea. The central
equine cornea is 600-900 mm thick and the peripheral cornea is 850-1200 mm
thick in horses (Andrew at al. 2001; Ramsey et al. 1999b). The cornea has a I
surface epithelium, a middle stromal layer containing regularly arranged collagen
fibers, and an inner endothelial layer containing a basement membrane
(Descemet’s Membrane). The cornea also forms the anterior fibrous portion of

the globe.

The iris is a pigmented muscular structure that is located in front of the
lens and is visible through the cornea. Its central border defines the pupil which
contracts or dilates to control the amount of light that passes to the retina
(Reviewed by Crispin et al. 1990).

The lens is a transparent structure that is suspended in the anterior
portion of the posterior segment between the aqueous and vitreous chambers.
Aqueous and vitreous humor provide nourishment for the lens. Opacification of
the lens (cataract) may affect vision, if centrally located. Cataracts located in the
anterior or peripheral parts of the lens may also impair vision. The cornea and
lens are necessary for focusing and transmitting light to the retina (Reviewed by
Crispin et al. 1990).

The ciliary body is continuous with the iris anteriorly and the choroid
posteriorly. One of the primary functions of the ciliary body is to produce aqueous
humor, which circulates through the posterior and anterior chambers and
egresses through the trabecular meshwork, angular aqueous plexus, and scleral
venous plexus before entering the general venous circulation (Reviewed by Gum
et al. 1999). If obstruction to the outflow of aqueous humor occurs, intraocular
pressure increases and may result in glaucoma. An additional role of the ciliary
body is accommodation of the lens. Accommodation is caused by the contraction
and relaxation of the musculature in the ciliary body, which in turn alters the
shape of the lens by changing the tension on suspensory zonule fibers that
attach to the lens (Reviewed by Gum et al. 1999). The ciliary body thus allows

the lens to alter its refractive power allowing the animal to focus on objects at

different distances (Harman et al. 1999)). A poorly developed ciliary muscle in
the horse results in a poorly developed level of accommodation (Harman et al.
1999)

The retina is a light sensitive neurosensory tissue that overlies the inner
surface of the choroid. Histologically, the retina consists of ten layers: 1) retinal
pigment epithelium, 2) photoreceptor layer, 3) external limiting membrane, 4)
outer nuclear layer, 5) outer plexifonn layer, 6) inner nuclear layer, 7) inner
plexiforrn layer, 8) ganglion cell layer, 9) nerve fiber layer, and 10) internal
limiting membrane (Ehrenhofer et al. 2002). Layers two through ten comprise the
neurosensory retina and layer 1 (retinal pigment epithelium) is non-sensory. The
inner surface of the neurosensory refine is in contact with the vitreous humor
(Ehrenhofer et al. 2002). The retina contains a complex system of nervous
tissue that is responsible for phototransduction (conversion of light energy to
chemical energy). Light rays enter the globe and are refracted to form an image
on the retina. Nerve impulses from the absorption of the light energy are initiated
in the rods and cones, which are located in the outer retina (Ehrenhofer et al.
2002). Rods and cones absorb the light, and then transform it into a nerve
impulse. The nerve impulse is then sent via nerve fibers on to the optic nerve.
The optic nerve, which is located at the back of the globe, then passes the
messages onto the brain where they are translated into the visual image
(Ehrenhofer et al. 2002).

The eye is formed early in development of the fetus, but at this time the

timing of development of the normal equine eye is unknown (Crispin 2000). The

process of development is divided into three stages: embryogenesis,
organogenesis, and differentiation. Embryogenesis is the segregation of the
primary layers of the developing embryo, and begins just after fertilization.
Organogenesis is the separation of the cells into the pattern of the various
organs. The characteristic structure of each organ develops during differentiation
(Reviewed by Forrester et al. 2002). The eye is derived from three neural crest
cell types: surface ectoderm, neuroectoderrn, and mesenchyme (Reviewed by
Crispin et al. 1990). These cells migrate into the developing eye in three waves
(Churchill and Booth 1996). The first wave differentiates to form the trabecular
meshwork and corneal endothelium. The second wave gives rise to the corneal
stroma, which is the tissue that forms the foundation for the cornea. The third
wave of cells leads to the development of the iris (Churchill and Booth 1996).
The corneal stroma derives from neural crest mesoderrn that surrounds
the anterior portion of the invaginating optic cup. The neural crest mesoderrn
then encloses the ectoderrnal lens vesicle and the mesenchyme, which causes
the anterior chamber to be formed (Reviewed by Crispin et al. 1990). The
development of the anterior chamber is initiated by invagination of the lens
vesicle. It involves cleavage, differentiation and atrophy of the mesenchyme that
lies between the embryonic cornea and iridopupillary membrane (Reviewed by
Crispin et al. 1990). The trabecular meshwork derives from undifferentiated
neural crest mesodenn (Reviewed by Crispin et al. 1990). Minor anomalies that

occur during anterior chamber development, such as colobomas and corneal

opacities, often are due to abnormal chamber cleavage (Reviewed by Matthews
et al. 1990).

The lens develops in four stages: 1) formation of the lens plate, 2)
formation of the lens vesicle, 3) formation of primary lens fibers, and 4) formation
of secondary lens fibers (Reviewed by Crispin et al. 1990). The lens plate is
formed due to proliferation of the surface ectoderrn. A hollow lens vesicle forms
when the lens plate invaginates (Reviewed by Crispin et al. 1990). The lens
vesicle becomes filled as the posterior lens epithelial cells elongate. After
elongation, these cells make up the primary lens fibers and eventually are
positioned in the center of the nucleus of the lens (Reviewed by Crispin et al.
1990). The anterior layer of epthelial cells of the lens vesicle persist throughout
life and actively divide. The anterior lens capsule is the basement membrane of
the anterior epithelial cells. The epithelial cells at the equator of the lens elongate
to form the secondary lens fibers (Reviewed by Crispin et al. 1990). These fibers
grow and encircle the primary lens fibers. Secondary lens fibers continue to be
produced throughout life (Reviewed by Crispin et al. 1990).

The iris, ciliary body, and choroid, collectively referred to as the uveal
tract, are derived from early mesoderrn and neuroectoderrn cells. The rim of the
optic cup gives rise to the formation of the iris (Crispin 2000). Abnormal
development of the uveal tract can cause aniridia, persistent pupillary membrane,
cysts, tumors, iris hypoplasia and colobomas (Crispin 2000).

Defects that arise during induction, differentiation, proliferation or

migration of any of the above structures can result in dysgenesis of the anterior

segment (Semina et al. 2001). Abnormalities are classified based on the stage of
development in which they occur. Defects that occur during embryogenesis are
generally lethal. Those that occur during organogenesis typically result in gross
deformities of the eye (Reviewed by Peiffer et al. 1999). If developmental defects
occur during the fetal period when the most active stage of development has
passed, minor defects are often the result (Reviewed by Cook 1999). Defects
such as colobomas (portion of the structure is absent) may result from
incomplete closure of the embryonic fissure (Reviewed by Matthews et al. 1990).
Abnormalities that are associated with anterior segment defects, such as
Axenfeld-Rieger syndrome in humans and small eye and dysgenic lens in mice,
are due to a defect in the control of neural crest cell migration into the developing

eye (Churchill and Booth 1996).

Anterior Segment Dysgenesis in Horses _

Anterior Segment Dysgenesis in horses has a semidominant mode of
inheritance (Figure 3). It is characterized by two distinct phenotypes that consist
of: 1) large fluid filled cysts that originate from the posterior surface of the iris,
ciliary body or peripheral retina (Figure 4), and 2) ciliary cysts accompanied by
multiple anterior segment anomalies including iris hypoplasia, iridocomeal
adhesions and peripheral corneal opacification, nuclear cataracts, and cornea
globosa (Figure 5 and 6) (Ramsey et al. 1999a). Some of the horses affected
with the syndrome also have microphthalmia, lens subluxation, retinal

detachment and craniofacial abnormalities (Ramsey et al. 1999a). Horses with

ASD had abnormal pupillary light responses and the pupil did not dilate with the
use of mydriatic drugs (Ramsey et al. 1999a). Anterior Segment Dysgenesis in
horses shows phenotypic variation with some eyes being more severely affected
than others. Horses with cysts can have different size cysts and fully affected
horses can have only a few clinical signs or the full range of developmental
abnormalities. Typically, the defects were bilateral (observed in both eyes). The
cyst phenotype is represented in the heterozygous state and the complex ASD
syndrome is represented in the homozygous state. Of those horses studied, only
one was blind and not as a result of the ASD syndrome (Ramsey et al. 1999a).
Unlike the position in humans with Axenfeld-Rieger syndrome where glaucoma is
a common occurrence, none of the affected horses examined were diagnosed
with glaucoma. This could be due to the fact that in horses the majority of
aqueous drainage is by the alternative drainage pathway, also known as the
uveoscleral outflow, and therefore abnormalities involving the conventional
drainage pathway rarely cause glaucoma in horses (Ramsey et al. 1999a). The
ophthalmic abnormalities, including nuclear cataracts, appeared to be
nonprogressive.

Anterior Segment Dysgenesis was first identified in a breed of horse called
the Rocky Mountain Horse. The Rocky Mountain Horse breed is believed to have
started with a single horse that was from the Rocky Mountain region of the
United States. Most pedigrees collected on this breed trace back to a single
stallion, lV-7, thus this stallion is considered the foundation sire of the Rocky

Mountain Horse breed (Figure 7). Horse IV-7 was not examined but seven

10

offspring (five stallions (VI-38, V-12, V-30, V-20, and V-34) and two mares (V-9
and V-29)) were examined and are included in this study (Figure 7). The cyst
phenotype occurred in five of the seven offspring that were examined (Ewart et
al. 2000). Because of the high occurrence of cyst-affected offspring, lV-7 is
believed to have been heterozygous for ASD. Eighty-six percent of the horses in
the research population of 516 horses derive from those original five half-sib
stallions.

A segregation analysis of 516 Rocky Mountain Horses showed that a
large number were affected: 250 horses had ciliary body or retinal cysts, and 72
horses had ASD (Ewart et al. 2000). Nonpenetrance of the ASD phenotypes
occurred in a small percentage of Rocky Mountain Horses and 12 of the 13

studied were traced to a shared ancestry with horse Vl-38 (Ewart et al. 2000).

Similar Diseases in Humans and Other Species
AxenfeId-Rieger syndrome

Several diseases in humans have been identified that are similar to ASD
in horses, including iris hypoplasia, Axenfeld anomaly, Rieger anomaly and
syndrome, and iridogoniodysgenesis. More commonly, these diseases have
been lumped together and given the umbrella name of AxenfeId-Rieger
syndrome. Axenfeld-Rieger syndrome is characterized by malformations of the
eyes, teeth and umbilicus (Espinoza et al. 2002). Other abnormalities detected in

humans include hydrocephalus, hearing defects, cardiac and renal abnormalities

11

and congenital hip anomalies (Phillips et al. 1996). Typically, the embryos that
are homozygous for Axenfeld-Rieger syndrome are not viable.

Each of the individual diseases that comprise Axenfeld-Rieger syndrome
have slight phenotypic differences. Axenfeld anomaly is usually characterized by
posterior embryotoxon, an abnormal mesenchymal tissue that crosses the
comeoscleral limbus and attaches to the posterior cornea. Rieger anomaly has
the same abnormalities as Axenfeld anomaly with the addition of stromal
hypoplasia, corectopia (a displaced pupil) and polycoria (supranumery pupils).
Rieger syndrome has the ocular defects as described in Axenfeld-Rieger
anomaly, but other body abnormalities can occur such as facial, dental and
umbilical anomalies (Espinoza et al. 2002). Iris hypoplasia is the mildest form of
Axenfeld-Rieger syndrome and solely affects the ocular structures. It is
characterized by abnormal development of the iris stroma resulting in partial or
complete absence of the iris and early onset glaucoma (Heon et al. 1995).
lridogonodysgenesis is characterized by iris hypoplasia, goniodysgenesis, and
juvenile glaucoma. It involves a poorly developed irido-comeal angle through
which the aqueous humor must pass before exiting the eye (Smith et al. 2000).

There are several possible explanations for the variability in expression of
Axenfeld-Rieger syndrome. The variability could be caused by different mutations
in the same gene or an interaction between the mutation and the genetic
background. Another possibility is that multiple mutations in genes with related

functions could cause variability in expression (Smith et al. 2000).

12

Axenfeld-Rieger syndrome is an autosomal dominant trait and has
varying levels of expression. Three genetic loci have been associated with
Axenfeld-Rieger syndrome and are located on human chromosomes 4q25, 6p25
and 13q14 (Nishimura et al. 1998; Phillips et al. 1996). The Axenfeld-Rieger
syndrome genes on 4q25 and 6p25 were identified as PITX2 and FOXC1,
respectively. No gene has been identified for the 13q14 locus (Kawase et al.

2001).

Peters anomaly

Peters anomaly is a form of anterior ocular dysgenesis in humans. It is
characterized by central corneal opacification, which is usually bilateral, adhesion
between the anterior surface of the lens and the posterior surface of the cornea,
and adhesions between the iris and anterior surface of the lens (Hanson et al.
1994). The condition most often results in glaucoma (Doward et al. 1999). Three
forms of this disease have been identified: sporadic, autosomal dominant and
autosomal recessive (Doward et al. 1999). The majority of cases have been
attributed to PAX6 mutations (Hansen et al. 1994). However, in a study by
Doward et al. (1999), a splice site mutation in the third intron of PITX2 was
identified in a patient with Peters anomaly with cataract'and iris hypoplasia. This
study was the first to identify that mutations in PITX2 could cause both Peters

anomaly and Axenfeld-Rieger syndrome.

13

Dysgenetic Lens

Dysgenetic lens (dyl) is an autosomal recessive disease in mice that is
very similar to other forms of anterior segment dysgenesis, specifically Peters
anomaly (Blixt et al. 2000). It is characterized by small cataractous lens and other
anterior segment anomalies such as: fusion of the lens, iris and cornea; corneal
dysplasia; and lens and iris hypoplasia. Defects in mice are first seen at
embryonic day 10.5-11 when the lens vesicle fails to separate from the ectoden'n
(Semina et al. 2001). The failure of separation results in a fusion between the
lens and cornea, which leads to an arrest of the development of the anterior lens

epithelium (Semina et al. 2001).

Small eye

Small eye (Sey) is an autosomal dominant disease that affects the eyes of
mice. Small eye is the murine counterpart of a human ocular anomaly, aniridia
(absence of iris), and is characterized by iris hypoplasia, corneal opacification,
cataracts, and microphthalmia (reduced eye size) (Presser and Van Heyningen
1998). In the homozygous state, Sey/Sey, mice have a complete lack of eyes
and nasal primordium (Hill et al. 1991). Homozygous mice (Say/Soy) die
perinatally due to the absence of nasal structures and pancreas, severe brain
defects and lack eyes (Van Heyningen and Williamson 2002). Heterozygotes,
Sey/+, have varying degrees of malformations including reduced lens size and

cataracts (Dellovade et al. 1998).

14

The gene that has been predominantly associated with this disease is
Pax6. It was shown by Schedl et al. (1996) that over expression of Pax6 caused
reduced eye size in mice. Hansen et al. (1994) identified that a proportion of
Seyl+ mice had a heterozygous Pax6 point mutation with ocular phenotypes that

are similar to the human ocular disease Peters anomaly.

Eyes Absent

The Drosophila eye, although structurally very different from the vertebrate
eye, has very similar molecular structure to the vertebrate eye. Therefore, it is
often a good starting point for looking for possible causes of vertebrate ocular
malformations. One such disease is called Eyes absent, which is an ocular
abnormality that occurs in Drosophila melanogaster. Studies have attributed this
disease to mutations in the Drosophila eye gene (Bonini et al. 1997) and a
second gene that is similar to human PAX6, ey (Czemy et al. 1999). The eye and
ey genes are expressed in the eye and other tissues such as: 1) the adult head,
2) the embryonic Bolwig’s organ, central nervous system, and eye-antennal
discs, and 3) the larval eye-antennal discs. Mutations in these genes have been
shown to result in flies with no eyes (Quiring et al. 1994). Mutations in ey affect

the optic lobe, the eye, and the eye-antennal disc (Flybase, 2002).

Candidate Genes

There are more than 40 genes that are involved in eye development.

Several of these genes have been identified as underlying the similar diseases in

15

humans and mice. Three of the genes involved in anterior segment anomalies
are PITX2, FOXC1 and FOXE3. These genes are transcription factors that are
highly conserved across species. Transcription factors are by definition any
protein other than RNA polymerase that is required for transcription. They usually
have one of three functions: 1) bind to RNA polymerase, 2) bind another

transcription factor, or 3) bind to cis-acting DNA sequences.

PITX2

Paired-like homeodomain transcription factor 2, PITX2, is a member of
the bicoid-like transcription factor family. This gene family is involved in the
genetic control of development, which includes pattern formation and cell fate
determination. In several studies, it has been shown to control left-right
asymmetry (Hamada et al. 2001; Kathiriya and Srivastava 2000; Liu et al. 2001).
The PITX2 gene is located on chromosome 4q25 in humans. It consists of 20 kb
of genomic sequence, and contains 6 exons. The initiation codon is located in
exon 2 and the homeobox region is located in exon 5. Most mutations occur in
the homeodomain (Espinoza et al. 2002). Several isoforms‘have been identified
in human, mouse, and zebrafish. The isofon'ns are not well documented.

Mutations of PITX2 have been found in Rieger syndrome patients (Semina
et al. 1996). In a study by Flomen et al. (1998), it was suggested that regions
upstream of the gene might contain sequences with a regulatory role in the
control of homeobox gene expression. They showed that Rieger syndrome could

arise from a whole gene deletion of PITX2 or from a translocation breakpoint 90

16

kb upstream from the gene (Flomen et al. 1998). PITX2 has also been identified
with mutations that cause iris hypoplasia, iridogoniodysgenesis syndrome, and
Axenfeld-Rieger syndrome. It has been hypothesized that mutant PITX2 proteins
that retain partial function result in milder anterior segment abnormalities
(Kozlowski and Walter 2000). One child presented with Peters anomaly was also

found to have a mutation in PITX2 exon 3 (Doward et al. 1999).

FOXC1

Forkhead box C1, FOXC1, is a member of the forkheadlwinged helix
transcription factor family. It is characterized by a forkhead-binding domain that is
identified by a 110-amino acid motif. The forkhead binding motif contains three
alpha helices and two large loops that form the “wing” structure (Saleem et al.
2001). The forkhead domains are highly conserved across species (Kaufmann
and Knochel 1996). Research has shown that FOXC1 is expressed in humans in
the developing brain, skeletal system and the eye (Davies et al. 1999). Foxc1 in
mice is expressed in the mesenchyme from which the ocular drainage angle is
formed (Smith et al. 2000). FOXC1 is located on human chromosome 6p25. It is
encoded by a single exon that is 1,662 bp in humans. In mice, Foxc1 is
expressed in prechondrogenic mesenchyme, periocular mesenchyme, meninges,
endothelial cells, and kidney.

Mutations in FOXC1 have been identified in patients with primary
congenital glaucoma, Rieger anomaly, Axenfeld anomaly, iridogoniodysgenesis

syndrome, Axenfeld-Rieger syndrome and iris hypoplasia. Nishimura et al.

17

(2001) predicted that both FOXC1 haploinsufficiency and increased gene dosage
might cause the anterior defects in the eye. In a study by Saleem et al. (2001), it
was reported that mutations of FOXC1 in human patients with Axenfeld-Rieger
malformations had a reduced FOXC1 transactivation ability. They indicated that
reduced stability, DNA binding, or transactivation could cause a decrease in the
ability of FOXC1 to transactivate genes that can underlie Axenfeld-Rieger
malformations (Saleem et al. 2001).

In studies by Nishimura et al. (1998) and Mears et al. (1998), mutations
most often occurred in one of two places, nonsense mutations between the
initiation codon and the forkhead domain, and missense mutations in the
forkhead domain. Of the six mutations that have been identified by these studies,
those located around the first helix cause more severe ocular defects (Mears et

al. 1998; Nishimura et al. 1998).

FOXE3

Forkhead box E3, FOXE3, is also a transcription factor that has the
identifiable forkhead domain. FOXE3 is located on human chromosome 1p32.
FOXE3 has a single exon with no introns. The genomic DNA in humans is 2,011
bp and the coding sequence is 957 bp. The predicted human protein is 319
amino acids in length. The forkhead domain is located between 466 bp and 732
bp, and tends to be conserved among the forkhead family. In mice, the open
reading frame is 864 nucleotides encoding a 288 amino-acid protein (Blixt et al.

2000).

18

Foxe3 in mice is expressed in the lens epithelium around embryonic day
9.5 at the start of lens placode induction. Expression increases as the lens
placode is formed (Blixt et al. 2000). Expression is limited to the anterior
proliferating cells when lens fiber differentiation begins (Blixt et al. 2000). This
study also found that sequence from a homozygous dyl mouse contained five
single nucleotide substitutions, including three that altered the amino acid
sequence (Blixt et al. 2000). Foxe3 mutations altered the DNA binding domain,
thus causing the recessive phenotype (Semina et al. 2001). Blixt et al. (2001)
concluded that Foxe3 is responsible for regulation of the closure of the lens
vesicle and its separation from the ectoderm. ln dyl mice, this fails to occur.
Foxe3 may also be important for the growth and viability of the lens epithelium
(Blixt et al. 2000).

A study by Semina et al. (2001) identified a mutation in the coding region
of FOXE3 in a family identified to have anterior segment ocular dysgenesis and
cataracts. This mutation caused a frameshift that resulted in abnormal sequence
in the last 5 amino acids and added an additional 111 amino acids to the
predicted protein (Semina et al. 2001).

FOXE3 has also been implicated as a candidate gene for some forms of
Peters anomaly (Blixt et al. 2000). A mutation in FOXES was identified in a family
afflicted with Peters anomaly with corneal opacities and glaucoma. The G to T
mutation was located in the binding domain and led to a nonconservative amino

acid substitution (Ormestad et al. 2002).

19

Equine Genomics

There are currently three horse genome databases: INRA (Institut
National De La Recherche Agronomique) horsemap in France, ARK database in
the United Kingdom, and horsemap in Japan. The horse mapping effort is
focused on identifying landmarks on chromosomes and creating a framework for
studying horse genes. A meeting of the mapping consortium resulted in the
standard for the horse karyotypic assignment, which was established in 1997
(Figure 8). Another goal of the mapping consortium is the development of a linear
linkage map for the horse, which is primarily comprised of microsatellite DNA
markers and other types of polymorphic markers. Guerin et al. (1999) have a
half-sibling family, based on 13 sires and 480 offspring, which is available for
linkage mapping. The family is large and diverse, which provides for many
opportunities to identify linkage relationships between genetic markers. The most
recently published dataset for this map was produced using data from 20
laboratories in addition to the previously published data (Guerin et al. 1999). The
Newmarket map is a full sibling map based on a stallion bred to two sets of
identical twins to produce 60 offspring embryos. Mares were artificially
inseminated and the embryos were retrieved from the uterus in order to produce
the high number of offspring in a short amount of time. This design allowed
researchers to make a linkage map that includes the X-chromosome (Swinburne
et al. 2000).

According to the INRA horsmap database as of November 30, 2002, the

horse genome map consists of 427 genes and 735 microsatellites. Four hundred

20

five of those genes and 507 of the microsatellites have been mapped via linkage
and cytogenetic techniques.

Two mapping technologies are commonly used to map loci including
genes and microsatellites; a somatic cell panel and a radiation hybrid panel. An
equine somatic cell hybrid map is available through the University of California at
Davis. The somatic cell hybrid panel for the horse consists of 108 horse-mouse
heterohybridoma cell lines. The targeted horse DNA from each cell line is
amplified using the primers for a targeted gene, and then scored for the presence
or absence of horse-specific fragments after electrophoresis. This information
then is utilized to identify where the gene is located in the horse genome
(Caetano et al. 1999).

The other mapping panel is a radiation hybrid panel. Radiation hybrid
maps were first designed to aid in human gene mapping (Goss and Harris 1975).
Goss and Harris pioneered this technology by using a single human
chromosome, but in 1996, the idea of a whole genome radiation hybrid panel that
is more efficient was developed (McCarthy 1996). The panel was created by first
irradiating the genome of interest with a high dose of x-ray. The donor DNA was
broken into many fragments due to the irradiation. The donor DNA was then
incorporated into a recipient hamster cell line using either Sendai virus or
polyethylene glycol (McCarthy 1996). The nonrecombinant donor cells died due
to irradiation. The recipient cell line was labeled with a selection marker that
causes any cells that are labeled with it to not grow on a hypoxanthine,

aminopterin, and thymidine (HAT) medium. Recipient cells that contain the donor

21

cells were the only ones that would grow colonies on HAT medium (McCarthy
1996). The colonies were then selected and expanded for DNA extraction, and
placed into a 96-well format. The panel was then screened for genetic markers
and the retention pattern for each hybrid was compared to determine linkage and
map distances between markers (McCarthy 1996).

Radiation hybrid mapping is an efficient mapping tool and allows
researchers to map both polymorphic and nonpolymorphic markers (McCarthy
1996). Radiation hybrid panels are used for ordering markers in the region of
interest and for establishing the distance between these markers. The radiation
hybrid panel is extremely useful and efficient for mapping in horses since there is
minimal characterization of the horse map and horses have long gestation
periods and single births, which makes it difficult to produce a large mapping
family (McCarthy 1996).

There are currently two groups that have a horse/hamster radiation hybrid
panel, Dr. Bhanu Chowdhary at Texas A&M University and a privately owned
company, ResGen (A subsidiary of lnvitrogen Corp, California). Dr. Chowdhary’s
radiation hybrid panel was created using 5000..., of x-rayto irradiate horse DNA
from a normal diploid Arabian male horse (Chowdhary et al. 2002). The recipient
cell line is a non-irradiated thymidine-kinase (T K-) deficient hamster cell line
(A23). The average retention rate for seven of the chromosomes (3, 4, 10, 11,
14, 20 and X) tested in the study was approximately 26%, which is predicted to

be representative of the whole genome (Chowdhary et al. 2002). Retention rate

22

is the amount of donor DNA retained by the hybrids in the panel and is usually
estimated as an average over all of the chromosomes.

The ResGen radiation hybrid panel consists of 90 radiation hybrid clones
covering the entire equine genome (Kiguwa et al. 2000). An equine cell line
(donor: EEL/SO+, male horse embryonic endothelial primary lung cells) was
exposed to 3,000 rad of x-rays and then fused with non-irradiated thymidine-
kinase (T K-) deficient hamster recipient cells (A23). The average retention is

estimated at 28% (Kiguwa et al. 2000).

Literature Review Summary

By comparing the phenotypes of equine ASD with known, well-researched
diseases in humans and other species, we have identified candidate genes that
may underlie this disease in horses. Anterior Segment Dysgenesis is very similar
to Axenfeld-Rieger syndrome, Peters anomaly, Small eye and dysgenic lens.
Defects of all five syndromes include iris hypoplasia, iridocomeal adhesion,
corneal opacification, and nuclear cataracts. A major difference between
humans, mice and horses is that horses with ASD have a homozygous
phenotype that is viable, whereas mice and humans do not. While glaucoma is
documented in more than 50% of the diagnosed cases 'in mice and humans
(Espinoza et al. 2002), horses with ASD do not have glaucoma (Ramsey et al.
1999a). By determining the similarities of the diseases, researchers may
investigate the genes known to cause the diseases in other species and apply

this knowledge to investigating the mechanisms that underlie ASD in horses.

23

Chapter 1

Tables and Figures

24

   

Canal or Schlemm

    
 

Vitreous

Retinal Pigment
Epithelium

Figure 1. Diagram of the eye. Anterior Segment Dysgenesis syndrome can
affect all of the structures of the anterior segment (front portion) of the eye
including the cornea, iris, pupil, ciliary body and lens. The retina and exterior
ocular structures can also be involved (Figure from Prince, 1960).

25

Posterior Vitreous

Chamber Chamber

Ciliary
Body

  

Anterior
Chamber

 

Pupil

,il

Optic Nerve

   

Lens

I ]

Anterior Segment Posterior Segment

Figure 2. Segments and Chambers of the Eye. The eye is divided into two
segments: the anterior segment which includes the cornea, iris, lens, ciliary body
and anterior parts of the retina; and the posterior segment which starts at the
posterior surface of the lens and goes to the optic cup. The segments are divided
into three chambers: 1) anterior chamber (spotted black); 2) posterior chamber
(solid black); and 3) vitreous chamber (white) (Figure adapted from Gelatt 1999).

26

 

 

 

1 1

Elwin? {limo

(
Ill-121 one 01-120 mos Ill-30 tic-a

 

 

Figure 3. Anterior Segment Dysgenesis is inherited in a codominant
manner. One segment of the population exhibits incomplete penetrance.
This is displayed in the offspring of VIII-124 and Vl-9, where one of the
offspring is phenotypically normal. The roman numeral indicates the
generation number and the following number is the animal identification
number. The squares represent males and the circles represent females.
A blank symbol indicates an unaffected horse, half filled indicates a cyst
affected horse and a solid symbol indicates a fully affected horse. A
symbol with a question mark indicates an animal that has not been
examined. (Adapted from Ewart et al. 2000)

27

 

Figure 4. Photograph of a horse’s left eye containing a cyst
(arrows) of the ciliary body.

28

A)

3)

 

Figure 5. Photographs of the right cornea of two horses. (A) Corneal
curvature of an unaffected horse. (B) Cornea globosa is characterized
by an excessively large corneal optical diameter and excessive
protrusion. This is a deformity that commonly occurs in horses with ASD
syndrome.

29

 

Figure 6. (A) Photograph of an unaffected horse eye. (B) Photograph of
a horse with ASD syndrome. The arrows point to iridocomeal adhesion
and the arrowheads point to an encircling and flattened granula iridica.
This eye also has cornea globosa and iris stromal hypoplasia.

30

 

 

 

0
11-1 114
I] ———— 0
111-3 111-2
a) —— ------ e
/ 1v-7 111-1
0 Form 0
v.21
0 0 0 0
I I O I I O I
was v-12 v 11 V40 v 20 v.29 v 34

Figure 7. Rocky Mountain Horse Pedigree. Horse IV-7 is believed to be the
founder stallion. Eighty-six percent of the horses in the study population derive
from the five half-sibling sons (VI-38, V-12, V-30, V-20, and V-34). (Adapted

from Ewart et al. 2000).

31

SIR-Ella.
III: I... I..2

37

9.0.9.
‘5.
II gun—I-Iar
§~I.-I

:I‘E.

octave

0
‘ I
, if '
r“, .

is
.III
4

‘8‘:
is!

w
.d a.
In. u.

53......

5: Eu.

.I It
a I.
e .5

'9»
I.

In...

. 6
iii!

a...
at
a...

In.
C o e
.35

a... :50
:52...

.8.
.2:

as. :u. 7
93:3.

ufilr.
no.9.

:3

:6
III

‘5

_- ~I_I~O
.3 .I I1

no...

. I.
2!

‘ eh .,

' a“
gnu
5...: 5.9
pea-l...
I!
II

h 9.

Figure 8. Karyotype of the horse. Horse chromosomes 1-10. The standard
horse karyotype was identified at the Second lntemational Conference for
Standardization of Domestic Animal Karyotypes (Richer et al. 1990).

32

on.
3...!
may:
3.!15
Ruin-t...
a. as.
$5.
36v

Run-aw

In.

null-m.

0-.- I.

Figure 8 (Cont). Horse Karyotype. Chromosomes 11-21 (Richer et al. 1990).

33

i as: mitt“
24 25
5""EE 3%.- "”3515;
25 27
~11le ‘38 ’T' @3925:
23 29
3° 31

Figure 8 (Cont). Horse Karyotype. Chromosomes 22-31, X and Y
(Richer et al. 1990).

CHAPTER 2
MATERIALS AND METHODS

Study Population

Ophthalmic examinations, blood samples and pedigree information were
collected prior to the current study on a population of 516 Rocky Mountain horses
that represented nine generations of an extended family. Another 232 horses that
extend back another four generations had just pedigree information, for a
completed pedigree of 748 horses in 13 generations. The horses were from 30
different farms in six states: Michigan, Indiana, Illinois, Kentucky, Missouri and
Ohio (Ramsey et al. 1999a). The horses ranged in age from 10 hours to 29.2
years. Eighty-six percent of the research population can be traced back to five

stallions that derive from a founder stallion.

Genomic DNA Isolation

Blood samples (n = 516) were previously collected from the study
population, and genomic DNA was isolated from peripheral white blood cells by a
modification of the sodium perchlorate extraction technique (Johns and Paulus-

Thomas 1989) and stored for the current study.

Sample Organization

Equimolar amounts of DNA from individual animals were combined, or

“pooled” to create polyallelic samples of DNA (hereafter referred to as pools).

35

Pools were developed to increase the efficiency of analyzing markers and
sequence data within groups of animals that had the same phenotype. Eleven
pools of DNA from ten horses each were designed based on the horses’
phenotypes (Table 1). There were seven pools that collectively contained all of
the 70 fully affected horses in the population, two pools of horses with cysts and
two pools of unaffected horses. If a difference was found in a pool, then the
variability was individually typed in all of the horses in that pool. There were three
DNA control samples, a single Rocky Mountain Horse, a single Morgan horse
and a pool of ten different breeds of horses, which included American Paso Fino,
Belgian, Morgan, Paint, Pinto, Pony of America, Quarter Horse, Standardbred,
Thoroughbred, and Arabian (Table 2). .1
Two approaches, association and linkage analysis, were taken to initially
test markers. The samples were organized into a 96—well format commonly
referred to as a plate. For the association study, the first plate (RMH plate 1)
consisted of horses taken from the pools: all 70 fully affected horses, 11 horses
with cysts, and 13 unaffected horses (Table 1). For linkage analysis, a second
plate was designed (RMH Plate 2) that consisted of horses that were a segment
of the extended family (I able 3). One of the five original stallions, V-30, and his
offspring were chosen for plate 2 based on the availability of parent and offspring
DNA (Figure 9). Also included on this plate is a son of V—30, VI-7, and all of VI-7’s
offspring and their dams (Figure 10). Horses for both plates were chosen based

on their phenotype and quality of DNA.

36

Comparative Mapping

Due to the incomplete characterization of the horse genome, a
comparative map had to be established comparing the locations of horse and
human genes (Figure Ila-c). The candidate genes (PITX2, FOXC1 and FOXE3)
have been mapped in humans, but not horses. By comparing genes that map
near a candidate gene in humans to those already mapped in horses, one can
infer where the genes most likely reside in horses.

FOXC1 has been mapped to human chromosome 6 between F13A
(6p25.3—p24.3) and PIM1 (6p21.2) at 6p25 (Figure 11a). F13A maps to horse
20q13 and PIM1 maps to horse 20q24. Other genes that are mapped near
FOXC1 in humans are ITPR3 and MUT. These genes map to horse 20q21.
Since this large segment of human chromosome 6 appears to be conserved on
chromosome 20 in horses, it is hypothesized that FOXC1 is located between
F13A and PIM1 on horse chromosome 20. Genes on the q arm of human
chromosome 6, ME 1 and PEX7, have been mapped to horse chromosome 10.
Because this region is physically distant from FOXC1, horse chromosome 10 is
not as likely to contain FOXC1.

PITX2 is located on human chromosome 4q25 between genes MTP and
SMARCA5, located at 4q24 and 4q31.1-31.2, respectively. Other- genes that
have been mapped cytogenetically near PITX2 on human chromosome 4 are
ALB, KIT and UCHL1, located at 4q11-13, 4q11-12, and 4p14, respectively;
these genes map to horse chromosome 3q. MTP and SMARCA5 have been

mapped to horse chromosome 2q (Figure 11b). It is likely that PITX2 in horses is

37

located on horse chromosome 2q between MTP (2q13) and SMARCA5 (2q21)
due to the high degree of conservation between humans and horses.

FOXE3 is located on the p arm of human chromosome 1 at 1p32. Genes
that are mapped in both humans and horses that surround FOXE3 are PGD,
MATN1, FUCA1, and VCAM1 (Figure 11c). PGD, MATN1, and FUCA1 are
mapped to horse chromosome 2p. However, VCAM1 has been mapped to horse
chromosome 5. Therefore, FOXE3 in horses may be located on either horse
chromosomes 2p or 5q. Horse chromosome 2 was chosen to pursue due to the
availability of markers, but if results exclude chromosome 2, then the location on
chromosome 5 can be pursued. Mapping of these genes would aid in marker

selection.

DNA Microsatellite Markers

DNA markers were identified that were located near the hypothesized
locations of the candidate genes on the comparative maps (Figure 11). Since the
genomic map of the horse is in the beginning stages of development, few
markers were available, so the markers chosen were based on reported location,
heterozygosity, and number of alleles. Oligonucleotide primer sequences were
obtained from published protocols for each DNA marker. A number of markers
were tested and at least one marker was successfully optimized for each

candidate gene to test across the 2 plates (Table 4).

38

PCR Amplification

DNA samples were amplified by the polymerase chain reaction (PCR) for
either genotyping or sequencing analysis. The PCR mixture contained 50 ng of
genomic DNA from either the single horse sample or pooled samples combined
in a 25 pl volume with ZOmM Tris-HCI pH 8.3, 50 mM KCI, 1.5 mM MgCl2, 200 mM
each dNTP, 20 pmol of each primer, 1.25 U Taq polymerase, unless otherwise
noted in Table 4. The reactions were performed on a Perkin Elmer 9600
Geneamp PCR System (Norwalk, CT), MJ Research FTC-200 DNA Engine
(Watertown, MA) or a Stratagene Robocycler (La Jolla, CA) using a 1 minute 95
degree denaturation, 1 min 55 degree annealing and 1 min 72 degree extension
unless otherwise noted in Tables 4 and 5. The PCR products were optimized by
adjusting annealing temperatures, magnesium concentrations and PCR cycling

conditions as needed (Tables 4 and 5).

Marker Genotyping

Three detection methods were used for genotyping: ethidium bromide
staining of DNA in agarose gels, randomly incorporated radioactively labeled
(dCTP) PCR products, and fluorescently labeled primer PCR products. Each
marker was initially tested on agarose gels to detect if the primers yielded a
product of the expected size. Markers were then tested with either a radioactive
or fluorescent label. Radioactively labeled products were separated on
polyacrylamide gels rather then agarose gels to allow for resolution of products

that differed by only a few base pairs in length. The disadvantage to this method

39

is that if markers are not fully optimized, the true alleles can often be masked by
strand slippage artifacts (peaks with sizes in multiples of two base pairs smaller
than the actual allele size). Fluorescently labeled products were separated on
polyacylamide gels using either an ABI Prism 373 or 377 DNA sequencer
(Applied Biosystems, CA) that excites the fluorescent label and produces
electropherograms that allow visualization of the product and its size. This
technique also has disadvantages in that if the product is not optimized, it can
produce strand slippage artifacts or a pattern that does not allow for a distinct
size allele to be called, such as a bell-shape.

All of the markers were initially tested on the two single horse controls and
the pool of multiple breeds and were detected on 4.5% agarose gels (3 grams of
agarose, 1.5 grams of NuSieve combined with 100 ml of TAE buffer) that were
stained with ethidium bromide to see if product was obtained with the given
conditions. If no product was obtained, the conditions were altered until
optimization was achieved. The markers were then tested in the pools of horses
using a 33p radioactive label (dCTP), which was randomly incorporated in the
PCR at C nucleotides. The PCR products were separated on 5.75% Long
Ranger Singel (BioWhittaker Molecular Applications, ME) acrylamide gels. The
gels were dried and exposed to Biomax MR radiographic film (Eastman Kodak
Corp., Rochester, NY) for 24 to 72 hours. After developing, the films were
analyzed to determine if there was variation between the alleles of the fully

affected, cyst affected and normal individuals. To confirm variability seen in the

40

pools, the individual horses were radioactively genotyped following the procedure
above (Figure 12).

The DNA sequence of each marker was confirmed as follows: the primer
pairs were optimized and products from the optimized PCR reactions of each
primer pair for two single horses (one cyst affected Rocky Mountain Horse and
one Morgan) and the pools of horses (multi breed pool, unaffected pool, cystme
and ASD pool) were separated on 0.8% agarose gels. The DNA bands of interest
were excised from the gels and the DNA was isolated using the Qiaquick ll Gel
Extraction Kit (Qiagen Inc., Valencia, CA) (See Appendix A). Purified DNA
templates were sequenced using the Therrno Sequenase radiolabeled terminator
cycle sequencing kit (USB, OH) (See Appendix B). The products from the
sequencing reactions were separated on 5.75% acrylamide gels (See Appendix
C). The gels were dried and exposed to film. Sequences were read manually
(Figure 13) and sequence identities were confirmed by performing Basic Local
Alignment Search Tool (BLAST) searches of the GenBank database (NCBI,
2002)

Individual genotypes were verified using fluorescently labeled primers in
the PCR reactions. Primers for each marker were labeled a different color (6-
FAM, HEX, NED or TET (lDTDNA, Iowa)) since each dye fluoresces at a
different wavelength (See Appendix D). The different colored dyes allow for
similar sized products to be multiloaded for each sample, which allowed for more
efficient use of each gel. The labeled PCR products from horses from both plates

were analyzed on an ABI 377 PRISM sequencer (Figure 14). Once samples

41

were loaded, voltage was applied so that the negatively charged DNA migrates
toward the anode, which causes the fragments to move through the gel and
separate according to size. At the lower portion of the gel, the fragments pass
through an area called the “read region”, where a laser beam continuously scans
across the gel. The laser excites the fluorescent dyes that are attached to the
fragments and when this occurs the fragment emits light at a wavelength specific
for each dye.

At the end of data collection, analysis software (ABI Prism Genescan
Analysis) was used to process, analyze and translate the collected data into

fragment size lnfonnation.

Statistical Analysis

To determine if there was an association between the alleles identified for
the samples on plate 1 (plate of disease phenotypes) and the disease
phenotypes, the chi square option of SAS Version 8.02 (SAS Institute 1999-
2001) was used. To accommodate for the low sample size within some of the
allele sizes, the Fishers Exact test was computed. A 95% confidence level was

used so that p-values of less than 0.05 would be considered significant.

Linkage Analysis
Two point analysis was performed on the genotype data from the plate
containing the half-sib family using the CRIMAP software version 2.4 (Green et

al. 1990). Two point analysis allows each marker to be tested for pairwise linkage

42

against the disease phenotype. The lod score threshold was first set at 0.0 to
determine if there was any suggestion of linkage. For Mendelian traits, a Iod

score threshold of 3.0 is accepted as significant evidence for linkage.

Primer Design

Primer pairs for amplifying the candidate gene(s) were designed from
known conserved human and mouse DNA sequences that were obtained from
Genbank entries (Table 5). The computer program Oligo, Version 6.65 (Rychlik
and Rychlik 2002), was used to design the primers. When choosing the primers
from those that Oligo designed, several rules of design were followed. Since
variability tends to be more common in the introns of sequence, primers were
designed in exon sequences that flanked intron sequence wherever possible.
They were also designed over less mutable amino acids and in most cases the 3’
end of each primer was designed to match the second codon position of an
amino acid with low mutability, which minimizes the chance of a mismatch at this
position (Collins and Jukes 1994). Primer pairs were also matched based on
similar annealing temperatures, and the potential for primer-dimer formation was
avoided. Primers were no shorter than 15 bp and generally not longer than 25

bp.
Gene Sequencing

Controls for sequence reactions included one Rocky Mountain Horse of

each phenotype and one human sample, since the primers were designed based

43

on human sequence. The phenotype pools used for genotyping were used to
sequence the candidate gene to look for variability among the three phenotypes.
Pools allowed for a greater efficiency of detecting variability. When variability was
identified, individual horses from each phenotype were tested to verify the

sequence difference.

Sequence for Radiation Hybrid Mapping

For radiation hybrid panel mapping, short segments of less than 400 bp of
horse DNA are needed that are not conserved in hamster in order to map a
gene. If equine specific amplification is successful, a binary code is generated
and when compared to a database, the sequence is localized to a certain region
of a particular chromosome, thus giving the location for the gene in the horse
genome. For genes with more than one exon such as PITX2, the DNA segments
that contain introns are preferred since intronic sequences are not as conserved
as exonic sequence. FOXC1 and FOXE3 have only one exon, so primers were
designed in highly conserved areas, such as binding domains that flank areas
that were not as conserved across species. Identification of sequence that was
not conserved with hamster was attempted, so that the hamster/horse radiation
hybrid panel (Chowdhary et al 2002) could be screened to determine which of

the cell lines contained the equine specific DNA sequence.

44

Chapter 2

Tables and Figures

45

 

Tablet.PoolsandPlatelStudyPopulation

 

 

 

 

 

 

 

 

80015202112 W W W
[ML

Gen’ Animal Farm Patient Sire Darn A280 4200 m
A301 9 93 828 3 2 2 0.083 0.033 315 1.91
8 128 837 3 4 4 0.038 0.027 190 1.41
0 112 045 3 2 4 0.070 0.042 300 1.01
10 131 848 3 2 3 0.107 0.071 535 1.51
10 17 852 3 4 4 0.107 0.057 535 1.88
10 107 o 13 3 4 2 0.024 0.007 120 3.43
10 124 098 3 2 2 0.171 0.1 855 ' 1.71
10 30 N 0 3 4 4 0.050 0.020 200 2.00
0 100 w0 3 4 4 0.042 0.010 210 2.21
10 100 z 2 3 2 2 0.110 0.054 500 2.15
A502 10 11 8 4 3 2 2 0.11 0.085 550 1.89
7 03 012 3 2 2 0.104 0.100 070 1.70
0 07 020 3 2 2 0.040 0.021 240 2.20
10 14 050 3 2 2 0.144 0.003 720 1.73
11 11 053 3 2 3 0.144 0.070 720 1.02
11 40 050 3 2 3 0.112 0.000 500 1.05
10 00 c 7 3 2 2 0.10 0.000 000 1.03
10 117 o 5 3 2 3 0.000 0.001 400 1.57
0 10 o 37 3 2 2 0.222 0.102 1110 1.37

10 All 0 04 3 2 3 0.100 0.000 005 3.02 ,
ASD3 10 105 o 10 3 4 4 0.044 0.010 220 2.44
1 1 20 F 18 3 4 4 0.047 0.025 235 1.88
10 113 G 9 3 4 4 0.057 0.02 285 2.85
10 111 027 3 4 4 0.270 0.151 1300 1.04
10 100 K18 3 4 2 0.028 0.013 130 2.00
11 39 031 3 2 2 0:101 0.058 505 1.74
8 58 U1 3 2 2 0.041 0.014 205 2.93
0 57 us 3 2 2 0.054 0.03 270 1.00
0 22 W7 3 2 4 0.054 0.022 270 2.45
10 20 we 3 2 3 0.140 0.003 730 1.57
A804 11 10 040 3 2 3 0.157 0.007 705 1.00
10 120 o 12 3 2 2 0.074 0.045 370 1.04
10 00 r 0 3 3 2 0.077 0.040 305 1.07
0 7 K 5 3 3 4 0.000 0.040 445 1.02
11 10 K 0 3 2 2 0.004 0.044 320 1.45
1 1 23 K11 3 2 2 0.098 0.077 490 1.27
10 125 KKKKK13 3 2 2 0.05 0.027 250 1.85
8 98 R 8 3 2 4 0.089 0.033 345 2.09
0 07 R 0 3 2 4 0.074 0.020 370 2.55
10 51 we 3 2 3 0.000 0.035 330 1.00
ASD5 0 10 010 3 2 2 0.107 0.00 535 1.70
0 11 015 3 2 2 0.103 0.001 515 1.00
10 10 023 3 2 2 0.025 0.010 125 1.50
10 110 025 3 2 2 0.04 0.015 200 2.07
10 110 020 3 2 1 0.050 0.03 205 1.07
0 12 030 3 2 4 0.054 0.027 270 2.00
10 5 B34 3 2 2 0.048 0.022 230 2.09
7 85 B39 3 2 2 0.208 0.115 1030 1.79
11 2 D 28 3 4 0 0.041 0.014 205 2.93
11 31 I 0 3 4 1 0.001 0.031 305 1.07
A806 8 124 A1 3 3 0 0.088 0.034 880 1.94
10 10 822 3 4 3 0.042 0.018 210 2.33
8 105 880 3 3 0 0.118 0.073 580 1.59
10 67 C19 3 3 2 0.097 0.063 485 1.54
9 75 C21 3 3 2 0.112 0.073 580 1.53
1 1 25 C29 3 3 2 0.058 0.04 290 1.45
11 17 C33 3 3 2 0.084 0.048 420 1.75
10 42 F15 3 4 1 0.142 0.107 710 1.33
unk F19 3 0 0 0.122 0.085 810 1.44
8 89 M 4 3 4 0 0.06 0.038 300 1.87

 

 

 

 

A
O)

 

Table 1 (Cent). Pools and Plate 1 Study Population

 

 

 

 

 

 

 

201620112 10000231100 2110130111201 0862013030101
101000
Gen: Animal Farm Patient Sire Dem A200 A200 mm: M

A007 10 10 047 3 2 2 0.051 0.020 255 1.02
10 13 013 3 2 1 0.002 0.057 410 1.44
7 50 c 1 3 2 4 0.100 0.055 545 1.00
10 43 F18 3 4 0 0.070 0.034 300 2.20
10 30 628 3 2 0 0.003 0.052 405 1.70
10 03 L3 3 2 1 0.003 0.053 415 1.57
0 10 o 0 3 2 0 0.000 0.042 330 1.57
11 42 024 3 2 1 0.002 0.033 310 1.00
11 10 029 3 2 2 0.035 0.011 175 3.10
mm 3 4 4 0.005 cross 425 1.31
Cyst A 0 7 021 2 2 1 0.077 0.042 305 1.03
7 01 0 22- 2 2 0 0.073 0.030 305 2.03
0 117 0 0 2 2 4 0.115 0.07 575 1.04
10 70 010 2 2 0 0.075 0.045 375 1.07
0 04 R12' 2 2 1 0.053 0.010 205 2.04
7 00 R3' 2 2 4 0.005 0.042 425 2.02
7 20 U4 2 2 1 0.004 0.033 320 1.04
0 10 us 2 2 0 0.054 0.03 270 1.00
10 130 W? 2 2 3 0.051 ' 0.032 255 1.50
10 21 Q0 2 2 2 0.075 0.000 375 1.02
Cyst 0 0 0 0 3- 2 2 0 0.000 0.047 345 1.47
7 40 wide 2 2 0 0.000 0.047 400 2.00
0 77 1:2- 2 2 0 0.050 0.025 200 2.24
0 02 010' 2 2 0 0.030 0.017 105 2.20
11 30 020 2 2 0 0.047 0.020 235 1.01
7 00 r 3 2 2 o 0.040 0.020 245 1.00
5 12 Tim 2 - 0 0.004 0.042 420 2.00
0 17 c s 2 2 0 0.074 0.030 370 1.05
10 01 c 0' 2 2 0 0.051 0.023 255 2.22
0 14 c 1 2 4 0 0.001 0.044 405 1.04
Unaffected A 5 10 0 00 1 0 0 0.007 0.040 435 1.00
0 15 0 04' 1 0 0 0.00 0.020 300 2.07
0 01 0114 1 0 0 0.005 0.051 .. 475 1.00
s 14 0121 1 0 0 0.000 0.052 445 1.71
0 111 0125 1 0 0 0.075 0.04 375 1.00
0 55 0130 1 0 0 0.000 0.037 345 1.00
0 00 020 1 0 0 0.043 0.020 215 1.05
0 43 L4' 1 0 0 0.011 0.005 55 2.20
0 05 z 0 1 0 0 0.11 0.007 550 1.04

5 20 1111- 1 - 0 0.125 0.050 025 2g
Unaffected 0 7 53 020- 1 2 0 0.000 0.004 440 1.30
10 7 020- 1 2 1 0.00 0.000 450 1.30
7 0 .110- 1 2 0 0.001 0.050 455 1.57
7 1 KKKKK11 1 2 1 0.070 0.040 300 1.50
10 01 KKKKK14 1 2 1 0.003 0.057 405 1.03
7 00 037 1 2 0 0.005 0.051 , - 475 1.00
5 0 U7 1 - 0 0.052 0.03 200 1.73
10 114 w 1- 1 2 4 0.000 0.050 340 1.21
7 07 211 1 2 1 0.150 0.1 700 1.50
7 05 z 5 1 2 1 0.123 0.000 015 1.01

 

 

 

 

‘ Phenotypes: 0-Unexamined, 1-Unal'fected. 2-Cyst Aliected, 3-ASD Affected. 4-Unexamined but 03005 to founder stallion
'-Founder Stallion (unexamined)

2 Gen= Generation of horse
3 The stock DNA concentration is calculated by (A260'(50'Dilution)). In all samples dilution = 100x.

47

 

Table 2. Multibreed Pool DNA Data

 

 

Breed Sample ID A260 A200 :75? A260/A280
Pasofino AP CP-1 0.056 0.028 280 2.00
Belgian Bel CP-l 0.098 0.062 490 1.58
Morgan Mor M-1 0.173 0.101 865 1.71
Paint Paint CP-7 0.147 0.097 735 1.52
Pinto Pinto CP-4 0.068 0.042 340 1.62
Pony of America POA CP-1 0.082 0.053 410 1.55
Quarter Horse QH M-5 0.017 0.013 85 1.31
Standardbred STD CP-7 0.043 0.023 215 1.87
Throughbred TB CP-6 0.107 0.055 535 1.95
Arabian Arab M-5 0.129 0.067 645 1.93

 

1The Stock DNA concentration is calculated by (A260*(50*Dilution)). In all
samples dilution = 100x.

48

Table 3. Plate 2 Study Population

 

 

 

 

identification Phenogpe‘ 000mm
Genz Animal Farm Patient Sire Dam A280 A280 [stock] A260IA280
jig/ml3
5 29 W 1 * 0 0.137 0.007 005 2.04
10 10 047 3 2 2 0.002 0.047 410 1.74
10 110 00 2 2 3 o. 070 0.04 390 1.95
7 13 030 1 2 2 0.09 0.044 450 2.05
10 0 017 1 2 1 0. 050 0.027 295 2.10
9 110 .10 2 2 4 011 0.005 550 1.09
0 13 03 1 4 0 0. 073 0.030 305 1.92
0 10 014 2 3 2 0. 414 0.340 2070 1.10
0 130 G4 2 4 o o. 047 0.021 235 2.24
10 10 023 3 2 2 o. 057 0.029 205 1.97
10 123 R4 2 2 3 0.102 0.007 010 1.00
9 33 I5 1 4 0 0.045 0.019 225 2.37
10 115 R10 2 2 4 0.100 0.054 530 1.00
10 05 00 2 2 2 0.050 0.020 200 2.07
9 4 033 2 2 4 0.059 0.035 295 1.09
10 7 020 1 2 1 0.090 0.054 490 1.01
0 30 E1 2 * 4 0.044 0.024 220 1.03
10 40 10 1 2 1 0.000 0.034 330 1.94
0 20 05 2 2 4 0.071 0. 041 355 1.73
7 09 R3 2 2 4 0.132 0. 073 000 1.01
7 10 R5 2 2 0 0.097 0.045 405 2.10
0 04 R12 2 2 1 0.114 0. 001 570 1.07
9 100 we 3 4 4 0.171 0. 157 055 1.09
0 24 07 2 2 0 0.152 0.109 700 1.39
9 29 00 1 2 2 0.002 0. 04 410 2.05
10 12 031 2 2 2 0.049 0. 021 245 2.33
10 5 034 3 2 2 0.055 0. 029 275 1.90
0 51 004 1 2 0 0.11 0.001 550 1.00
7 04 002 2 2 1 0. 009 0. 034 345 2.03
9 09 I3 1 4 4 0.107 0 055 535 1.95
10 90 I4 1 2 1 0. 004 o. 031 320 2.00
10 01 KKKKK14 1 2 1 0.110 0. 00 590 1.97
10 21 00 2 2 2 0. 075 0. 039 375 1.92
10 70 019 2 2 o 0. 052 0. 025 200 2.00
10 114 W1 1 2 4 0. 004 0.035 320 1.03
0 33 W5 1 4 0 0. 072 0.044 300 1.04
9 22 W7 3 2 4 0.110 0.00 500 1.45
10 29 W8 3 2 3 0.102 0.009 510 1.40
10 109 W10 2 2 3 0.030 0.035 100 1.03
0 9 A7 2 0 0 0.032 0. 012 100 2.07
9 9 03 2 2 4 0.052 0.023 200 2.20
9 5 010 1 4 4 0.101 0.057 505 1.77
0 7 021 2 2 1 0.097 0.05 405 1.94
10 110 020 3 2 1 0.059 0.03 295 1.97
9 90 027 2 2 3 0.000 0. 034 340 2.00
9 12 030 3 2 4 0.042 0. 021 210 2.00
11 9 032 2 2 2 0.07 0. 037 350 1.09
0 0 035 1 4 0 0.04 0.017 200 2.35
9 2 030 2 2 1 0. 040 0.023 240 2.09
9 1 1 1 030 2 2 3 0.125 0.004 025 1.95

 

 

.h.
(D

 

Table 3 (Cent). Plate 2 Study Population

 

 

 

 

Identification Phengym1 D P 0
Gen2 Animal Farm Patient Sire Dam A260 A280 [stock] A260IA280
ttg/ml3
7 3 B42 1 2 1 0.086 0.047 430 1.83
6 12 843 1 2 0 0.052 0.026 260 2.00
9 10 849 2 4 4 0.07 0.034 350 2.06
8 41 857 2 2 0 0.073 0.041 365 1.78
9 45 858 1 2 2 0.112 0.059 560 1.90
9 95 D11 2 2 3 0.236 0.187 1180 1.41
9 81 D33 2 2 2 0.078 0.04 390 1.95
8 119 D95 2 4 4 0.029 0.014 145 2.07
10 124 D96 3 2 2 0.176 0.095 880 1.85
9 96 J5 2 2 2 0.083 0.046 415 1.80
10 139 W2 2 2 3 0.047 0.022 235 2.14
10 56 W3 2 2 4 0.124 0.064 620 1.94
10 11 B4 3 2 2 0.11 0.065 550 1.69
9 68 B5 2 4 0 0.083 0.041 415 2.02
8 17 B9 2 0 0 0.058 0.025 290 2.32
9 126 811 2 4 4 0.087 0.043 435 2.02
7 63 812 3 2 2 0.135 0.069 675 1.96
9 11 815 3 2 2 0.087 0.045 435 1.93
9 97 820 3 2 2 0.083 0.042 415 1.98
10 1 19 825 3 2 2 0.081 0.042 405 1.93
9 12 830 3 2 4 0.042 0.021 210 2.00
8 128 837 3 4 4 0.054 0.026 270 2.08
7 65 839 3 2 2 0.161 0.089 805 1.81
9 112 845 3 2 4 0.076 0.042 380 1.81
10 14 850 3 2 2 0.122 0.064 610 1.91
10 17 852 3 4 4 0.058 0.03 290 1.93
11 11 853 3 2 3 0.116 0.06 580 1.93
1 1 48 856 3 2 3 0.078 0.041 390 1.90
8 105 860 3 3 0 0.083 0.046 415 1.80
8 49 862 2 4 0 0.069 0.033 345 2.09
7 50 C1 3 2 4 0.109 0.055 545 1.98
10 68 C7 3 2 2 0.161 0.086 805 1.87
10 117 DS 3 2 3 0.055 0.028 275 1.96
9 117 06 2 2 4 0.08 0.04 400 2.00
8 54 D34 2 4 4 0.055 0.027 275 2.04
9 54 D36 1 2 2 0.086 0.048 430 1.79
9 19 D37 3 2 2 0.036 0.016 180 2.25
7 47 065 1 2 1 0.068 0.038 340 1 .79
10 120 D84 3 2 3 0.146 0.077 730 1.90
6 55 D136 1 0 0 0.101 0.06 505 1.68
5 12 Tim 2 * 0 0.097 0.065 485 1 .49
6 16 U6 2 2 0 0.071 0.03_2 355 2.22

 

 

' Phenotypes: 0-Unexamined 1-Unaffected 2-Cyst Affected 3-ASD Affected
4-Unexamined but traces to founder stallion

*-Founder Stallion (unexamined)

2 Gen = Generation of horse
3 The Stock DNA concentration is equated by (A260*(50*dilution)). In all samples dilution =

1 00X.

50

 

Table 4. m Primer Dltl

 

 

 

 

 

 

 

 

 

 

 

 

 

Marker Location Aleles (hp) afiefis Primer name 8 sequence (F) Primer name 8. sequeme (R) Optimizad' Tm
FOXEJ:
A8811 2p17-18 132-146 6 A8811 Forward A8811 Reverse No
CCACCTATGTGTI’CAGTTCACC GCACCMTGTTTATAGACTCCC
AH735 2p15.1-16 125-141 5 M735 Forward M735 Reverse Yes 67°C
TGACTTAGAGCTITTGCTCCC CCAGAAGTCCAGGCATTTGT
TKY24 2p14-16 1 W4 Forward 77(Y24 Reverse Yes 600
GATCCMCAGCAGCMCAGCAG AGTGGCATGGCATGAGCTTC
A8817 2p14-15 92-116 10 A8817 Forward A8817 Reverse No
GAGGGCGGTACCTWGTACC ACCAGTCAGGATCTCCACCG
TKYO3 2p13—14 162-168 4 ma Forward TKY03 Reverse No
GGTTCACACAGGAGTCAGGGA CCTTCTGGTTTGCCTCGTCTC
PITX2:
A14 2 q14-21 213-235 8 A14 Forward A14 Reverse Yes 63'C
CAGCTGGGTGACACAGAGAG GTCATCACTACTCCCTACAC
Not fine
mapped:
COR065 2 264-262 6 COR065 Forward COR065 Reverse No
CMMGCACACACMAGTGC TCCGGMAGTGCMAGTTAG
UMOO7 2 115-150 7 UMOO7 Forward UMOO7 Reverse No
GGGAATAGAGAMGGTGAAG TI'AGAGTTCCTGCTCCTCC
UCDEQ3GO 2 129-133 3 UCDEQ380 Forward ucoeoaao Reverse No
6W _<;G2A_A___ACT GTGCTGCMA
FOXC1:
HMS42 20q24 126-136 4 HMS42 Forward HMS42 Reverse Yes 55‘C
TAGATI'I’CTTAGTGCCMTCGTGG GAACTGCTATAGATATACCTAACTC
TKYOO 20q21.1 7 5 TKYOG Forwad TKY08 Reverse No
TTQCTI’GTGCATMIQ QTI’QQCMTGTGTGAQGMT
0R8 20q14 13 W Forward 0R8 Reverse No
CTCTGCAGCACATTTCCTGGAG CGCCGCTGCACCAGGAA
TKY21 20q13 117-132 4 TKY21 Forward TKY21 Reverse No
AGGTGMcgch
Not fine
mapped:
H765 20 79-95 7 H765 Forward H765 Reverse No
TGCTMGCCTCAGCACATACA TGGAAATMGGTTAGCAGGGATGC
@052 20 18225 7 m2 Forwad LEXDSZ Reverse Yes 55°C
GGAACGGAAGAGTGTAGTI’TT CATTTATI’CATCAGCGATTG
LBtO71 20 185-204 7 LB(O71 Forwcd LEXO71 Reverse No
CTTTATTCTACTCTTI’GGTCC CCGATATTTCACTGATTATT
UMO11 20 161-179 9 UM11 Forwa'd UM011 Reverse Yes 63‘C
TGAAAGTAGAAAGGGATGTGG TCTCAGAGCAGAAGTCCCTG
WAS-F164 20 141-165 8 WAS-H64 Forward WAS-HM Reverse No

_ CTAQTITATGAGCTGAGCCACTC QTGGAGAACTTTQATTCTCCTCT
‘AllmemenweretestedusinganhitialdenatmumeonMesaWS'Cfollowedby35cyclesofa95°cdenatweataoseconds.ananneelingstepet
the‘rmoivmhuntabhfuaommanextmiond72‘c1u30smw1aHM842whichwesdenahndtor20seoondsandLEX052
wnidmadaSOseomddensunASsecmdmthodeOsecmdextmsbntinu.Nurealimtemperantesnotlistedforpr‘lnerpaisthatwerenot
successfullyopt‘mized.

51

88:58 or .2 0.3- 80 3.2688 in: a 53 new .8258 _. .8 cows 8 868 c.2336 cm can 65:? P .2 058 05 E :96 E ._-
05 “a no.8 3825 cm .2358 e .2 2333 .3 a 3 3.08 mm .3 332.2 228.3 0.3 .55 cm 5? 3.08 2o; 83 BEE .2.N

888.2 «XE 8.. 885122 mmxom 8:81.22 Exon ”<20
2523 :95; .2 2383: c2330.." xcmmcow 88:982. 2o; >85. :02; 63 35:0 €3.80 9.: 555, 808.5 Co 20:80.?

 

 

 

 

 

 

 

 

 

 

 

.8 8> 8 8. .2082 888 .8 8-5 .9 288m 8 8 9.8 88-8.3 8 m 88 82E
.8 mo> 3 3.. 88898333888398 m-oommomommmmmmmmmmmmmm-m mmmow-vwnow 8.. o coxm. «CC-E
.8 8> 8 88 8888698908298 888898888288 «88.58 o :98 «XE
.8 oz 8 N: 8-8...8898888888 8880888888 omen-888 888 88cm. «ER
.8 oz 8 E: 888388889898 8888898088 858-888 8:98 88m
.8 oz 8 at 8-882883888898- 8-28858892588 88882-8 888-825 «ER
.8 oz 3 8 8888388888809...- 8-53868868898-8 888-888 888-892. 88¢
.88 oz 8 88 8682880888888 .8-088888m808988 888-283 808-888 NEE
.mm mo> an mom 8688038333383 8888889906898 comm-R3 mcoxw-Ncoxm Nx-tm
.mm mo> an new 8-835308838880398 .m-mmzmmo999@omm-.m gag-two N :92. -N :96 Nth
.8 oz 8 B.- .n- s a as 8 .8 .m- as N 883893 8.18 88.. mmxou
.8 oz 8 88 888588988888 883888888888 8:88 Ev“. mmxom
.8 oz 8 88 8838888889088 8-886868288898 «wt-«8. 88.. Exon
.8 oz 8 88 8-008808808888878 8888988888....- 8. 788 88 Exon
.8 8> 8 at 8888888888....- 8-8082888032898 838 Er. mmxom
.8 Ba .8 oz 8 Sb .m-oommoosmommammome-.m 88808888092898 8. 788 NE“. Exon
.8 8> 8 a: 888858898888 8888888888098 838 EN“. Exon.
.58 oz 8 «8 88988880229...- 8§§8mouudgo8888 «855 88$ .88..
.838 oz 8 84 8989888888 8.088880988898898 82-88 0&8“. Exon
.8..8 oz 8 o8 8888898988 8-080880660808898 88-88 888 .oxou‘
.mm 02 an new 8508890330366maum .n-mmmom8803909888 moo 73¢ 2&5“. Foxou‘
.8 8> 8 8v 888588288288 .8. 8.9820308888- 2-1 888 fixed
.8 oz 8 v8 .m-somsmszummmgmmmoom-m 888888888898 83-3..- 88 Exon
.8 8> 8 88 88880828809098 88568898889...- 8 08... Emmi Exon
.8 8> 8 N8 8.898.882.8898 889888888898 81.8 Er. Exon
«Eh $0398 mum SEE 3.93... .955 2920... 20:80.. 9:82 050

 

Ema .oEto mocoaumw .m was.

52

. .2. 9:9”. 5 2:28 9.: .0 0mm: 05 um 8:88 85 0. 82.8 30.8
mi N Ema __m.s-cm .05 .0 :3 85500 3.9qu £5 E 88.0; 0:... .om-> .c0___m.w cos—80:30“. .0 8.903 ..m 0.39....

nch vch 3. 7x 925

 

 

 

‘v 9 I 9
2.8 2" a; 8.x
v v
2.5 02>
‘8 O
23. 8.x 87x 2.x
1- C v
2.x.
8.x. 8.x 8.: 3.x
0 C v .-
2.3
O
8.: fix
1- O
2...: 87x 2.:
9 v 0
23. 83.
8.> Ju— % 83.. 2.x
1- v
2.x 8.x
.
«.2 .2: 9...:
a 9 0

'8

53

.8 -
N > 96E ucm om.> COEEmv

mt . . u n . .
n —
u

«3. an;

3’ .u -
O i
3v; “3
i
F‘s
0
3.5
O
«:3 5.!
uni <94»
I
no... 88 21!, ouao
. I v I

1“;
v

1%

913
v
n’
.
5’

hr’ 0.,
v .
«vi
'
Or! 3: Ni .8
' v v
0‘ vri
I
8:
v
3: a:
I
not, not,
2, ,
.I
on.) on}
. ‘I
a: P1
9 a.

3’
1
3! tea :1
3X 9;
v
23. 0.x
I
0.3
.
.2... 3%
v U
a?!)
58
at;
v
31: mi
«8
U
38 38
. ‘-
:0)
v

54

APOCNV Eatomcoo 02.060300w GEO-500 CNEDI _NCO.~NEQ~C_ E055 05w; mgr—N m.
606%... new 520m mEocmO 58:... 502 E9: :39 9.03 20:80. 235m 3sz .mmmamfiuxmx‘ 9: Eat Bear”

295 30:80. mcom mas—um 60552: 96 >25 m5 5 new: $9.58
. . . new Eon c_ m_ mcmm 336:3 m . 3.0. a
comhhucofioo. o>w8ma£oo mEeEEou 2 Bus 859 ms... .8 van cm .3 mmEomoEoEo mean 320:. o Mfiomoioficw
cm 2 m>=§m E8 9m 5...: mac; .9: E «.me .35; m5 c_ 339: :39 a?» 8: mm... new 0 mEoonoEo :mEan
v9.80. mm; NOXOK Exon 9.00 93550 .8 wear new 5631 c8502 ans. 3:83:80 6: 2:9...

 

 

 

 

 

 

 

 

 

om 38m ON 38: 2 as: o §§m
903$ 3 N82: . fl m £2
.53.?38 «NgN "
IIINNE ENgN Em NNE: "
_ . N358 _
an. .3: .235 _NgN $289 an? as: Em " m2
.
3 3d :gN I I as? NE: 5:? N NN- N3 " :_
REV 29% I - .III E: N_3_ " g
H28 NE SE SEN [Fl :3 NE: 338 E- -23 fit "2:
§ 58:5 89% N32 II
on? 22: ll ,5: " ow
§ 5.55 BENENES II S: N._N% " 2
N2 5&3: shuns» 352% ll 2:3 3% H mm
W 2&2 N.¢NQ-N.mN% m
NNNU Magma: ll NUXQBN mNao . g

 

55

.3003 E20350 050:0:000 050:00 :08:

. . I .0:0.:0::0 : Eo
0%30MHMMW wafiwoummmwhww MHBM00 :0E:I 502 ”:0: :0:0: 0:03 050000. 20:00 :0:::I .000:0:00v”m._< 0.“
. . .3 m. .00:__:00:= 0:0 00:0 0:: :_ 000: 0:00:08 0:0 0.0: :_ 0. 0:00 0 0 _

.00~_0_.0:_ 0:0 0:2:000. 0>_:0:0a::00 05.0000 0: 000: 00:00 0:... .0. 0:0 . . : 00:00 0:...

. . N 00E000::0::0 00:0: :0 0:000 0:: 0020:.

v 0E000E0::0 :08:: 0: 02:0:00E00 0:0 :0:: 00:0: 0:: :_ 000:< 00:0 .
. . . . : 0:: :_ 00000:: :0.» :0: 00: 0:0 0 0E000Eo: 0

:0E:: :0 00:000. 003 NEE NEE 0:00 0:020:00 :0: 00:01 0:0 :0:::... :00500 00.2 02:0:00800 .02. 050...“.

 

 

 

 

 

 

 

 

m 008$ N 008$ v 008:...
. :2
Saw 8% I I 2,: 23 I I u a:
a 983:: W50 VNX I I 30:05 :NaN I I “00:05 02-23. I I 2.
3:00 000 NNcN I I II _
E :Ncm II :0 :NicN II NEE «N3 II 2::
. 2000 N.:N-:.:0N I E: :ch . 8:
NNV 2K 001% II I MQV 2-200 " mm
.. . on
3500 Nam I I an: NGN 00 2 N3. .
:0: NZ :3 . a
. 3:00 3:: IWI . 3
50: ESE Nam 00: NTNEN II 09: 00:: I I . 2
500 .050: 2N3: IIIIBc: EARN .II : c m
IItmww. 2-23 MEI? 2.3.x II .
0:000 3-2% II 2.23 0000.: WEN II .
:IIlINE 8% .II 00:00 02.2% [I

 

 

56

::00N:

E2:00:00 05050000 0E0:00 c080... 0500532 :5: 0:03 000:0:0_0 .00.0>:0 0:0 :0>:0m 0E0:00 :0::0... .002
E0:: :0:.0: 0:0; 05:03. 0:0:00 :0E:I .00000:00v.m< 0:: E2: 00:.0:00 0:03 050000. 0:00 0505 0020005
0:0 >020 0:: :_ 000: 0:00:08 0:0 0.00 :_ 0. 0:00 0:020:00 0:... 0002.0: 0:0 05:08. 02:0:00E00 05:00:00

0: 000: 00:00 0:... .0 050008020 00:0: :0 0:00 ::00 0:0 N 050008050 00:0: :0 :::0.0 0:: 0:0 : 0E000::0::0
:0E0: 0: 02:93:80 0:0 :0:: 00:0: 0:: :_ 0002. 00:0: 0::. :_ 00000:: :00: :0> :0: 00: 0:0 _‘ 0::000E0::0

:0::0: :0 00:000. 002, mmxon. .9008": 0:00 0:020:00 :0: 00:0: 0:0 :0::0... :0050: 00.2 0>_:0:00E00 .0: 050:“.

 

 

 

 

 

 

 

 

 

 

m 00:0:: N 00:0:: . 000:3:
. :2
.
ANN 5:050: 30:00 II 0:000:00 :NwN II
00:00:.0zcm .I.I III .
:80 2-030 II 25300 n 0003 N03 II .ch
Ill . I . U .
ANN aesmvNGfiES: :00 II 2003 :N :3 N .
«SEEN II N023 SANS II . EN
. : a .
00:2-NEN II N00 NN-_N.: I .
8:0: EARN II .
I 0 ll .0:
Sc: 2-23 30.0: :0 N0 : .
E00: N30 II 20000:-ng 0.68.0 N0: II . 2
00:30:00 II 2:02 0000.023 .
00000200 II 00 002.2% E00,: 3:: II n .0
80:32.03“ II '2003723 LI 2:00:03: II . :N
0002.30.02. II .2

57

A) 59°C

CNCAAH

MI!

E m°c
c N c A H

 

Figure 12. Examples of genotyping using a 3 3 P radioactive label. A randomly
incorporated radio labeled dCTP can be incorporated into the PCR to identify allele
sizes in a genotype. A) Marker UM11 at 59° C with PCR conditions with a denature
time of 60s, an annealing time of 60s and an extension time of 60s. B) Marker A14 at
63° C with PCR conditions with a denature time of 303, an annealing time of 30s and
an extension time of 305. The “C” indicates a cyst horse, an “A” indicates a fully
affected horse and an “N” indicates an unaffected horse. The lane marked with a (-) is
a blank lane.

58

 

M-
5*...-
W .0
4‘ J" ‘9: v
“‘11-. 0::?
1‘37 1;:
(GT)16 m 1::
4o
7 w
h.‘ «nu-
." “I
” v
~ ~
‘ «up
w ”
> I ‘ _: ::
1': :
(AT)3 ‘:_ «up
“ a”
* w
\. “ e
u a»
I.
I . uh "'
:3 a.» ‘
.‘m
a ’
‘ «It-
‘-
‘ -
—‘> -

Figure 13. Radiographic Example of Marker, A14 Sequence. The
microsatellite in this sample is a diallelic repeat of (AT)3(GT)15.

59

1 500 1800

800
600
400
200

 

 

1800 2100 ‘ 2400 2700 3000
r‘ "-_‘ fl

1 800
1 200

800
400 [C
o -

Figure 14. Electropherogram Example. An electropherogram is the visual
results from a fluorescently labeled marker product. A) A heterozygous
genotype for marker HMS42. B) A homozygous genotype for marker UM11.
The arrowheads point to the fluorescent peaks, for which allele sizes can be
determined using the size standards (denoted by arrows). The y-axis is the
intensity and the x-axis is the points where the laser recorded the data.

 

 

60

CHAPTER 3
MARKER ANALYSIS - GENOTYPING

Introduction

Markers are segments of DNA that can be unique from one individual to
the next. Examples of markers are single nucleotide polymorphisms (SNPs) and
simple sequence repeats also known as microsatellites. For this study,
microsatellite markers were evaluated. Two types of analysis were used to study
marker genotype data, linkage analysis and association mapping.

Association studies test whether individual differences in genes or
markers are statistically associated with a phenotype (Reviewed by Gibson and
Muse 2002). The phenotype can be disease status such as affected or not;
constantly varying measures such as intraocular pressures; or response to
environmental stimuli such as drug stimuli (Cardon and Bell 2001). The statistical
test for association studies is used to determine whether the distribution of
genotypes in the unaffected animals is different from the distribution in affected
animals. The association study has two main advantages over the linkage study:
1) statistical power, and 2) knowledge (Reviewed by Gibson and Muse 2002).
The design of the association study allows for smaller numbers of samples to be
used to detect an association, which leads to the increased power. The theory is
that markers that are close to the candidate gene will tend to be homozygous for

a particular allele and those further away will have a greater tendency to

61

recombine. The power for an association study is the probability of correctly
identifying a true association (Cardon and Bell 2001).

An example of an association is linkage disequilibrium. lf most affected
animals in a population share the same allele, then it is possible to narrow the
interval around the disease locus by detecting the disequilibrium between close
markers and the disease locus (Jorde 1995). A linkage disequilibrium design that
is used to identify the defect that underlies common diseases is called test of
association. Test of association is a study that searches for a population
association between a phenotype and a certain allele. This type of design uses a
case versus control test. For this test an allele in the affected or case population
is looked for that an unaffected or control population does not have (Ardlie et al.
2002). The downfalls to association studies are: 1) replication is often not
possible, 2) small sample size can affect results, 3) overinterpretation of results,
and 4) failure to detect linkage disequilibrium with nearby markers (Cardon and
Bell 2001).

Linkage analysis gives the location of loci or phenotypes in relation to
previously mapped loci (Reviewed by Gibson and Muse 2002). The approach for
linkage analysis isto identify markers that are located throughout the genome or
near candidate genes and track their alleles through large family pedigrees to
identify a marker that is linked to the gene or disease of interest (Reviewed by
Hartl 2000). The lod score is a statistical term that indicates degree of linkage
between two loci. A lod score is computed by dividing the pedigree likelihood

when disease is linked by the pedigree likelihood when disease is unlinked and

62

taking the log of base 10 of the ratio (Ott 1992). A lod score of greater than 3.0
for Mendelian loci is considered to be significant evidence for linkage (Reviewed
by Hartl 2000).

Linkage studies can be used in tandem with the association study. The
large families typically used in linkage studies are suitable for association
studies, especially for within-family segregation designs (Reviewed by Gibson

and Muse 2002).

Results

Markers were chosen for the three candidate genes based on their
locations and variability. Based on comparative mapping, FOXC1 is
hypothesized to be located on horse chromosome 20. Markers that are on
equine chromosome 20 near the predicted location of FOXC1 are DRB, HMS42,
TKY21, TKY22, and TKY08. Five other markers, LEXOSZ, LEXOZI, UM11,
HTG5, and VIAS-H64, have been mapped to horse chromosome 20, but have
not been fine mapped. PCR amplification was attempted on all markers,
however only HMS42, UM1 1, and LEX052 were successfully optimized.

Marker HMS42 was prioritized over the others due to its location (20q24)
and its reported variability (4 alleles, sizes 126-136 bp).‘ HMS42 was sequenced
and the single cyst horse had a repeat of (GA)I4 and the single Morgan had a
repeat of (GA)15. This marker was genotyped using a fluorescently labeled PCR
product on Rocky Mountain Horse Plates 1 and 2. Alleles for plate 1 ranged from

122-132 bp (See Table 6 and Appendix E). Seventy-six point three percent of the

63

fully affected horses had allele size 130 bp. By comparison, 75.0% of the
unaffected and 70.0% of the cyst affected horses also had allele size 130 bp
(Table 6). Another common allele size seen in the fully affected horses was 122
bp (21.2%).

The chi square analysis for HMS42 yielded a p-value of 0.1158 with 4
degrees of freedom (Table 6). Chi-square results can be invalid when greater
than 20% of the cells contain values less than five; forty-four percent of the cells
in the HMS42 table had counts less than five. To accommodate for the low cell
numbers a Fishers Exact test was computed. The Fishers Exact test resulted in a
p-value of 0.1173 with 4 degrees of freedom. The chi-square and Fishers Exact
test gave similar p-values; both tests provide no evidence for an association
between the alleles of HMS42 and the disease phenotypes.

Plate 2 had alleles that ranged from 122-132 bp (See Appendix E). The
lod scores for the two family sets on plate 2 were both 0.0, indicating no
significant linkage between the marker and disease (Figure 15a and b). The two
families also had a recombination fraction of 0.5. A recombination fraction is the
relative distance between two loci at the computed lod score. A recombination
fraction of 0.5 is the maximum score and indicates no linkage.

The other two microsatellites hypothesized to mark FOXC1, although not
fine mapped, were chosen based on ease of optimization and variability. Marker
UM11 was investigated using ”P labeled PCR products and polyacrylamide gels
using the control horses, but it had strand slippage artifacts, which obscured

allele determination. To help alleviate this problem, both plates were analyzed

fluorescently. Strand slippage artifacts were still visible, but there was a pattern
across samples that was used to identify a true allele from an artifact. Nine
alleles were detected in animals of the three phenotypes on plate 1 (See Tables
7-9 and Appendix E). The fully affected horses had the majority of their alleles at
sizes 164 bp and 170 bp with allele frequencies of 29.5% and 33.6%,
respectively. The cyst affected horses also had a preponderance of allele sizes
164 and 170 bp. which occurred at allele frequencies of 44.4% and 33.3%,
respectively. The unaffected horses had a variety of allele sizes including 158
(9.1%), 160 (18.2%), 162 (4.6%), 164 (31.8%), 166 (13.6%), 168 (13.6%) 170
(4.6%) and 176 hp (4.6%) (Tables 7 and 8). The 170 bp allele occurred
frequently in ASD and cyst affected horses (33.6% and 33.3%, respectively), but
occurred rarely in unaffected horses (4.6%). The SAS computer program was
unable to compare the three phenotypes simultaneously for the numerous alleles
(n = 10) of marker UM11, therefore pair-wise comparisons between each of the
phenotypes were analyzed. A chi-square analysis of unaffected horses as
compared to horses with cysts resulted in a p-value of 0.0636 with 7 degrees of
freedom (Table 7). This p-value would suggest that the allele frequency
differences between unaffected horses and horses with cysts approached
statistical significance. To confirm the chi-square value since the Cell counts had
88% that were below five, the Fishers exact test was calculated and resulted in a
p-value of 0.0399, which would be indicative of an association between at least
one of the alleles and one of the disease phenotypes at a 95% confidence level.

The second pair-wise comparison of unaffected horses to horses with ASD

65

resulted in a p-value of 0.0013 with 8 degrees of freedom (Table 8). With 61% of
the table's cells having less than five samples, a Fishers exact test was
performed. The Fishers exact test confirmed the p-value from the chi-square test
at 0.0012. Both analyses provide a high likelihood that there is an association
between at least one of the alleles of UM11 and the ocular phenotypes. The last
pairwise test compared horses with cysts to horses with ASD. The resulting p-
value for the chi-square test with 6 degrees of freedom was 0.6649 (Table 9).
The chi-square test was followed up by the Fishers exact test since 50% of the
table’s cells had less than five samples. The resulting p-value (0.7525) confirmed
the chi-square analysis and both were suggestive of no difference between allele
frequencies in cyst and ASD affected horses. Overall, the results suggest that
there is a high probability that at least one of the alleles of UM11 is associated
with the disease phenotypes.

The association study was followed by a linkage study since it has been
reported that association studies can often result in false positives. The lod
scores for the two family sets on Plate 2 were 0.12 for the V-30 family and 0.0 for
the Vl-7 family (Figure 15a and b). The recombination fractions were 0.30 and
0.47, respectively. These values are not supportive of linkage to the disease.

LEX052 was optimized using the horse controls and sequenced. The
single cyst horse had a repeat of (TG)13. The single Morgan was heterozygous
with repeats of (TG)13 and (T G)15. Sequence variability was detected in the multi-
breed pool and the pool of horses with ASD. The unaffected pooled sequence

was not variable. Individual genotypes have yet to be determined.

66

The predicted location of PITX2 is on horse chromosome 2q. One marker
has been identified on 2q. A14 (2q14-21). A14 was genotyped using a
radioactively labeled PCR product in pools of horses: two unaffected pools, two
cysts pools and seven ASD pools. This data showed that there were multiple
alleles in each of the pools. To clarify which alleles were segregating in this
population, A14 was initially genotyped using a radioactively labeled PCR
product across all of the Rocky Mountain Horse plate 1. This method did not
allow for definitive sizing of bands but it was evident that there were at least three
different sized alleles among the horses fully affected with ASD and many of
these same alleles were found in the unaffected and cyst affected horses.
Subsequently, A14 was genotyped using fluorescently labeled PCR product in
ten unaffected horses, eight horses affected with cysts, and forty-one horses with
ASD. Allele sizes ranged from 213-235 bp. with the majority of all three
phenotypes having allele sizes 223 bp and 225 bp (Table 10). The chi square
analysis yielded a p-value of 0.3029 with 14 degrees of freedom (Table 10).
Once again, the validity of the chi square test was questionable, since seventy-
one percent of the cells in the table had counts less than five. To accommodate
for the low cell numbers, a Fishers exact test was computed. The Fishers exact
test resulted in a p-value of 0.2110 with 14 degrees of freedom, which suggests
that there is no association between A14 and the disease phenotypes. A14 was
sequenced and the single cyst horse had a repeat of (AT)3(GT)15. Variability was
found in the sequence of the multi-breed pool, the pool of horses with ASD and

the unaffected pool.

67

We were not able to fully optimize other markers located on horse
chromosome 2. These markers, UM007, UCDEQ380, and COR065 have not
been fine mapped. UCDEQ380 and COR065 produced doublet bands when
amplified and detected using a radioactively labeled PCR product. UM007
produced no product at varying temperatures and cycling conditions when
radioactively labeled.

FOXE3 is hypothesized to be located on horse chromosome 2p or
possibly on horse chromosome 5. Optimization was attempted for five markers
on horse chromosome 2: ASB11, AHT35, A8817, TKY24, and TKY03. ASB11,
ASB17, and TKY03 had strand slippage artifacts that hindered accurate
determination of allele sizes. AHT35 was prioritized due to its location and
extensive effort was expended to optimize this marker. Unfortunately, definitive
data were not obtained from this marker. Multiple alleles were present within
each phenotype, however specific allele sizes were not readily determined using
either radiolabeled or fluorescent techniques. AHT35 was sequenced in two
horses: the single cyst horse had a repeat of (TG)14, and the single Morgan had a
repeat of (T G)16,

Genotype results of two horses on plate 2 were not in keeping with their
parental genotypes. The first horse, lX-100, was located in the V-30 family.
Horse lX-100 was gendtyped for HMS42 on four separate occasions. This horse
had a homozygous allele size of 130 bp on three occasions, but had a
homozygous allele size of 132 bp on the final genotyping. Based on offspring

genotypes and repetition of results, the 130 bp allele was considered to be

68

accurate, therefore it was used in the analysis of the V-30 family. The second
horse, lX-5, yielded inconsistent allele sizes for HMS42 (136 bpl136 bp, 134
bp/134 bp, and 129 bp/129 bp) on repeated analysis. Therefore, data for horse
lX-5 was not used in the current study. These two horses may be interesting to

further research to confirm their genotypes.

Discussion

Results from our association study indicate that the gene that underlies
ASD in horses does not seem to be located near HMS42 or A14. The association
study did identify an association between UM11 and the disease phenotypes.
The statistical analysis for this marker was run as a series of pair-wise

comparisons instead of the comparison of the three phenotypes, which was used

with HMS42 and A14, because the SAS computer software was unable to
process the data. This was likely the result of numerous allele sizes in
combination with multiple phenotypes. To provide for accuracy ofithe data, both a
chi-square test and a Fishers exact test were used since low sample numbers in
the table’s cells can often lead to an invalid chi-square result. The Fishers exact
test takes into account the low numbers when computing the p-value, thus giving
a more accurate representation of the data.

For marker UM11, no alleles predominated in the unaffected horses,
whereas the majority of cyst affected and ASD affected horses had three allele
sizes: 160 bp. 164 bp and 170 bp. The high p-values between the pairwise

comparisons of the phenotype data suggest that there is a high degree of

69

association between UM11 and the disease phenotypes. The association
between the alleles and phenotypes might be caused by the cyst and ASD
affected horses having only three sizes of alleles and the unaffected horses
having a wide range of alleles. Another interesting aspect of the number of
alleles is thattlfe'iinaffected horses had a broad range of allele sizes, whereas
the ASD affected horses seemed to have a much smaller range of sizes. With a
large number of ASD affected horses (n = 70) versus the much lower number of
unaffected horses (n = 11) one may expect to have a much broader range for the
sample size. This can also be a problem in association studies if the control
population is not equal to the disease population. The design was initially set up
for a homozygosity mapping test. The homozygosity mapping test is used to look
for a common genotype among all of the ASD affected horses. Unfortunately, this
type of pattern was not seen, so an association study was done. Adding
additional unaffected horses would help to alleviate the unbalanced control
population.

To follow up the association study, a linkage study was performed on both
HMS42 and UM11 since HMS42 is located near the hypothesized location of
FOXC1 and UM11 is located on the same chromosome. The results from the
linkage study indicate that there was no evidence for linkage between UM11 and
the disease phenotype with lod scores of 0.12 for the V-30 family and 0.0 for the
Vl-7 family. These results suggest that the area surrounding UM11 can be
excluded as containing the disease locus. This is confirmed by analyzing the

recombination fractions, which were 0.30 for the V-30 family and 0.47 for the VI-7

70

family. These recombination fractions suggest that the marker may be too far
from the disease gene to detect linkage. These results contradict the results from
the association study which indicated that UM11 was associated with the disease
locus. The contradictory results do not mean that FOXC1 is not the underling
cause of ASD, but that further research needs to be done, such as the typing of
additional markers that map closer to FOXC1 and adding more unaffected
horses to the association study.

Marker HMS42 was not highly variable in the population having only three
allele sizes. Over 70% of the horses in all three phenotypes had the 130 bp
allele. Linkage study results confirmed the association study results for marker
HMS42 with both families having a Iod score of 0.0 at the maximum
recombination fraction of 0.5. HMS42 is too far from the disease gene to detect
linkage. This recombination fraction can also mean that the disease locus is on
an entirely different chromosome.

Marker A14 was informative in the Rocky Mountain Horse. population, but
there was no association between any of its alleles and the disease phenotypes.
Lack of association does not mean that the gene can be excluded but that the
marker is not associated. Different markers that are located near the candidate
gene may show an association.

Other reasons for the lack of association and linkage could be due to the
fact that the markers may be physically distant from the candidate genes.
Mapping the candidate genes would help alleviate this problem as it would

resolve any discrepancies between the comparative maps by identifying exactly

71

what chromosome the genes are on and their precise locations. Markers that are
known to be in proximity to the candidate gene can then be chosen, which would
give a higher likelihood of finding an association or linkage.

It is also possible that the candidate genes pursued in this study are not
the genes that underlie ASD in horses, since there are over forty known genes
that are involved in eye development. Most of the mechanisms and functions of
these genes are not well documented. Furthermore, because ASD in horses is
slightly different from the ocular diseases in other species, there might be better
diseases and candidate genes that have not been identified. Further studies on
the expression and function of the candidate genes should be pursued.

Further research should also be done on the discrepancies in the two
horses” genotypes. A technical error possibly caused the genotype of horse lX-
100 to be inaccurate. This could be confirmed by reanalyzing the genotype and
testing other markers. Horse lX-5 should be regenotyped at HMS42 to help
identify a consistent genotype, and verifying parentage may help eliminate any

discrepancies.

72

Chapter 3

Tables and Figures

73

Table 6. Results for the Association Study - HMS42

Table of population by allele
population allele

Frequency
Percent
Row Pct
Col Pct 122 130 132 Total

 

Unaff 2 15 3 20
1.27 9.49 1.90 12.66

6.25 12.61 42.86

 

Cyst s 14 1 20
3.16 8.86 0.63 12.66
25.00 70.00 5.00
15.63 11.76 14.29

 

A80 25 90 3 118
15.82 56.96 .90 74.68
21.19 76.27 2.54
78.13 75.63 42.86

A

 

 

 

 

 

Total 32 119 7 158
20.25 75.32 4.43 100.00

Statistics for Table of population by allele

 

Statistic 0F Value Prob
Chi-Square 4 7.4083 0.1158
Likelihood Ratio Chi-Square 4 5.8536 0.2104
Mantel-Haenszel Chi-Square 1 3.2581 0.0711
Phi Coefficient 0.2165
Contingency Coefficient 0.2116
Craner's v 0.1531

WARNING: 44% of the cells have expected counts less
than 5. Chi-Square may not be a valid test.

Fisher's Exact Test

 

Table Probability (P) 4.873E-04
Pr <= P 0.1173

Sample Size = 158

711

Table 7. Results for the Association Study - UM11 Unaffected Horses versus
Cyst Affected Horses

Table of population by allele

 

 

 

 

 

 

 

 

 

 

 

 

 

population allele
Frequency
Percent
Row Pct
Col Pct 158 160 162 164 166 168 170 176 Total
Unaff 2 4 1 7 3 3 1 1 22
5.00 10.00 2.50 17.50 7.50 7.50 2.50 2.50 55.00
9.09 18.18 4.55 31.82 13.64 13.64 4.55 4.55
100.00 50.00 100.00 46.67 100.00 100.00 14.29 100.00
Cyst 0 4 0 8 0 0 6 0 18
0.00 10.00 0.00 20.00 0.00 0.00 15.00 0.00 45.00
0.00 22.22 0.00 44.44 0.00 0.00 33.33 0.00
0.00 50.00 0.00 53.33 0.00 0.00 85.71 0.00
Total 2 8 1 15 3 3 7 1 40
5.00 20.00 2.50 37.50 7.50 7.50 17.50 2.50 100.00
Statistics for Table of population by allele
Statistic 0F Value Prob
Chi-Square 7 13.3718 0.0636
Likelihood Ratio Chi-Square 7 17.4914 0.0145
Mantel-Haenszel Chi-Square 1 0.7076 0.4002
Phi Coefficient 0.5782
Contingency Coefficient 0.5005
Craaer's V 0.5782

WARNING: 88% 0f the cells have expected counts less
than 5. Chi-Square aay not be a valid test.

Fisher's Exact Test

 

2.781E-05
0.0399

Table Probability (P)
Pr <= P

Saaple Size = 40

75

00: u 00:0 can-a0

0:00.: : .v ::
00.0000.0 .a. >::::naao:: .:n-:

 

HOOP Poaxw 0.L020dm

.000: 0:H0> 0 on :0: >00 000000.:zo .m :0::
000a 0:::00 00:000x0 0>0: 0::00 0:: :0 arc “oznz:<§

 

oomv.o > 0.:00000
:500.0 :c.:o:::ooo >0:.0:::co°
0000.0 «:0:0:::000 «:0
0000.0 «000.: : o:0:00.::0 Houaceuz-aoucaa
:000.0 :o:0.:~ o 0:0:00.::0 0::0: 000::H0x:4
0:00.: oo:v.m« o 0:0:00.::0
:0:: 0::.> no 0::0::au0

eased: a: :0::0a0000 :0 0:00: :0: 00:00::aum

0L30000Lm Owcm och

 

 

 

 

 

 

 

 

 

 

 

 

 

00.00: 00.0 00.0 5:.ou 5v.» «5.0 00.0“ 00.0 v:.o« on.»
_ ev: : : «v n 0: av : an a :0:0:
AII
00.00: 00.0 «0.50 00.00 «0.05 «5.00 00.0 :N.oo 00.0:
«0.0 00.0 :o.0n 00.: «0.0 :m.0« 00.0 00.00 «0.0
«5.00 00.0 00.0 50.0“ 00.: #0.: 00.0“ 00.0 00.5: h:.v
: um: : o :0 a :: on 0 mu 0 ow<
:II
00.0 00.00: 00.0 00.00 nv.:« o«.0: 00.00: 0n.o: 00.00
00.0 00.0 00.1 v0.0: 00.0: «0.:0 no.0 0:.0: 00.0
00.0: 00.0 00.0 00.0 no.0 00.0 00.0 00.0 on.“ 00.:
_ «N o : : o 0 h : c N ::0::
:II
H000: an: on: as: no: no: 00: «o: 00: on: :0: H00
:0: 30:
«:0000:
>0:0000:m
odoaaa :0::0H0000

0:0:H0 >0 :0::0::000 :0 0:00:

086: 860:2 00¢. gems 806: 00:00:05 :35 - 520 800802 05 :o: 0:380 .0 030:

76

Table 9. Results for Association Study - UM11 Cyst Affected Horses versus ASD
Affected Horses

Table of population by allele

 

population allele

Frequency

Percent

Row Pct

Col Pct 160 164 170 158 166 168 172 Total

Cyst 4 8 6 O 0 0 O 18
2.86 5.71 4.29 0.00 0.00 0.00 0.00 12.86
22.22 44.44 33.33 0.00 0.00 0.00 0.00

13.79 18.18 12.77 0.00 0.00 0.00 0.00

 

A80 25 36 41 6 11 2 1 122
17.86 25.71 29.29 4.29 7.86 1.43 0.71 87.14
20.49 29.51 33.61 4.92 9.02 1.64 0.82
66.21 81.82 87.23 100.00 100.00 100.00 100.00

 

 

 

 

 

 

 

 

 

Total 29 44 47 6 11 2 1 140
20.71 31.43 33.57 4.29 7.86 1.43 0.71 100.00

Statistics for Table of population by allele

 

Statistic 0F Value Prob
Chi-Square 6 4.0873 0.6649
Likelihood Ratio Chi-Square 6 6.5322 0.3663
Mantel-Haenszel Chi-Square 1 2.4923 0.1144
Phi Coefficient 0.1709
Contingency Coefficient 0.1684
Craaer's V 0.1709

WARNING: 50% of the cells have expected counts less
than 5. Chi-Square may not be a valid test.

Fisher's Exact Test

 

Table Probability (P) 0.0021
Pr <= P 0.7525

Sample Size = 140

77

Table 10. Results for Association Study - A 14

Table of population by allele

 

 

 

 

 

 

 

 

 

 

 

 

 

population allele
Frequency
Percent
Row Pct
Col Pct 213 223 225 233 235 215 217 229
Unaff 1 10 7 2 0 0 0 0
0.85 8.47 5.93 1.69 0.00 0.00 0.00 0.00
5.00 50.00 35.00 10.00 0.00 0.00 0.00 0.00
33.33 25.64 12.07 50.00 0.00 0.00 0.00 0.00
Cyst 0 2 12 1 1 0 0 0
0.00 1.69 10.17 0.85 0.85 0.00 0.00 0.00
0.00 12.50 75.00 6.25 6.25 0.00 0.00 0.00
0.00 5.13 20.69 25.00 12.50 0.00 0.00 0.00
A80 2 27 39 1 7 1 2 3
1.69 22.88 33.05 0.85 5.93 0.85 1.69 2.54
2.44 32.93 47.56 1.22 8.54 1.22 2.44 3.66
66.67 69.23 67.24 25.00 87.50 100.00 100.00 100.00
Total 3 39 58 4 8 1 2 3
2.54 33.05 49.15 3.39 6.78 0.85 1.69 2.54
Statistics for Table of population by allele
Statistic DF Value Prob
Chi-Square 14 16.1742 0.3029
Likelihood Ratio Chi-Square 14 19.1309 0.1600
Mantel-Haenszel ChioSquare 1 3.1487 0.0760
Phi Coefficient 0.3702
Contingency Coefficient 0.3472
Craner's V 0.2618

WARNING: 71% of the cells have expected counts less

than 5. Chi-Square may not be a valid test.

Fisher's Exact Test

 

Table Probability (P) 1.8335-07
Pr <= P 0.2110

Sample Size = 118

78

 

Total

20
16.95

16
13.56

82
69.49

118
100.00

Figure 15a. Marker Data for Plate 2. The family derives from the founder
stallion lV-7. Stallion V-30 (a son of IV-7) is the main sire on this pedigree.
Below the identifiers are the genotypes for HMS42 (top set of numbers) and
UM11 (bottom set of numbers). The allele sizes were assigned a number
based on numerical order of the sizes. The arrow points to the stallion, Vl-7, at
the head of the family in Figure 15b.

79

 

 

 

v £83
5.... a. $35 ..
.. e 35::
O Q35
g :00
e 333
i 33$ g £00
v 333-
3 EM
.1. ‘33:
; £00
4.3225
e ESE\DK
fl iss 5:55
a 8°“:
—E5 :33 ' $.-
8
'33
8 0 ”gas
V g0: * 3‘73
e is:
v §“3
.9 £00
- £93 I is:
9 3°“ . 38S ' £8: 9 200
.. $8:
0 3%:
Figure 153. §8$

80

Figure 15b. Marker Data for Plate 2. The family derives from the stallion V-30. A
grandson of lV-7, Vl-7 (a son of V-30) is the main sire on this pedigree. Below
the identifiers are the genotypes for HMS42 (top set of numbers) and UM1 1
(bottom set of numbers). The allele sizes were assigned a number based on
numerical order of the sizes. A question mark in the genotype indicates no data
reported.

81

v I”

Figure 15b.

l! x".

36
on

10
a!

.. is:

82

a is:

1‘
7”

“20 “12
‘D
71’

CHAPTER 4

SEQUENCE ANALYSIS

Introduction

The aim in sequencing the candidate genes of interest was three-fold.
First, was to determine the equine sequences for PITX2, FOXC1 and FOXE3,
since horse sequence has not been reported. Second. was to look for sequence
variability among the phenotypes, and finally to identify short segments of DNA
that are not conserved in hamsters in order to place each of the candidate genes
on the horse radiation hybrid map.

Eukaryotic genes are made up of two types of DNA segments, exons and
introns. Exons are active sequences of DNA that are spliced together to form
mature or processed messenger RNA (mRNA). Introns make up the rest of
genes that contain more than one exon and are excised from mRNA after
transcription. Introns are non-coding sequences that are cut out before the gene
can be expressed. The exact function of introns is not known.

Mutations are alleles that deviate from the majority type. especially if the
change in alleles causes an effect on some disease or phenotype. Mutations can
alter protein function, which can produce changes in the phenotype. Mutations
are rare and any that affect a gene are typically selected against through
evolution. A polymorphism is a mutation that has been maintained in the
population as an allele. These may occur in coding or non-coding regions of the

DNA and may or may not alter gene function. The most common polymorphism

83

is the single nucleotide polymorphisms which are naturally occurring variants that
affect single nucleotides.

The pool-and-sequenoe method (as described in the materials and
methods section) allows us to efficiently screen sequence for genomic variability.
Shubitowski et al. (2001) found that when pooling samples from ten horses, the
lowest allele frequency detected was 10% or two alleles out of 20. Variability is
seen as bands that occur at identical positions in a sequence. Single animals are
used for comparison purposes (Brouillette et al. 2000). Allele frequencies in the
pools can also be estimated by the intensity of the identically positioned bands. A
stronger band represents relatively more copies of the given allele and a lighter
band represents a rarer allele (Brouillette et al. 2000). Once variability is
identified, the individual animals in the pool are then tested to confirm their
genotype at the variable locus.

Sequencing of genes can aid in mapping the gene if usinga radiation
hybrid panel. Knowing the location of a candidate gene better aids researchers in
identifying markers that may be linked to the putative disease gene. Mapping of
candidate genes also adds to the framework of the horse genome map thus

giving a better idea of where other genes may be located.

Results
Sequencing was attempted on candidate genes PITX2, FOXC1 and
FOXE3 to look for variability between phenotype pools and to identify short

segments of DNA for radiation hybrid mapping. Primers were designed in

84

conserved regions between mouse and human that flanked areas that were less

conserved between the two species.

PITX2

PITX2 in humans has 20 kb of genomic sequence, which contains 6
exons. The initiation codon is located in exon 2 and the homeobox region is
located in exon 5 (Figure 16). Most known mutations occur in the homeodomain
(Espinoza et al. 2002). Primers were designed to amplify intron 2 and portions of
flanking exons. lntron 3 was not sequenced in horses since it is approximately 9
kb in length in humans. Primers flanking intron 5, which is about 2.5 kb in length
in humans, were also designed. Several primers were designed in exon 6 since it
contains the OAR (Otp and aristaless) domain which is believed to inhibit
transactivation (Cox et al. 2002).

The intron 2 primer set amplified 467 bp that contained only primer
sequence in exon 2, started at the first nucleotide of intron 2 and spanned 37 bp
into exon 3 (Figure 17). A BLAST search of this sequence produced a match with
human PITX2 (AF238048) with a bit score of 370 and an e—value of 1e-102. A bit
score above 200 and an e-value close to zero signifies a high probability that the
sequence would not occur at random. No variability was detected in the
sequence obtained between unaffected horses and ASD affected horses, when a
pool and sequence method was used.

Three sets of PITX2 primers within exon 6 amplified exon 6 sequence.

The first primer set (exon 6 forward and reverse) amplified 507 bp of horse

85

 

PITX2 (Figure 18). The other two sets of primers overlapped this region and
confirmed the initial sequence. The sequence obtained spans the entire OAR
domain. A BLAST search of this sequence resulted in a high match to human
PITX2 (AF238048) with a bit score of 722 and an e-value of 0.0. There was no
variability in the sequence obtained between unaffected horses and ASD

affected horses, when a pool and sequence method was used.

FOXCf

FOXC1 is a single exon gene that has a coding sequence of 1,662 bp in
humans. The forkhead region lies between 232 bp and 504 bp. Primers (eight
sets) were designed that covered the coding sequence of the entire gene (Figure
19). Two of the eight primer sets tested amplified horse sequence that matched
human FOXC1.

The first set of primers, F1 R1, amplified a segment of DNA that was 282
bp in length. This primer set amplified all but eight nucleotides of the forkhead
domain (Figure 20). When the sequence of the F1 R1 fragment was used in a
BLAST search, a match to human FOXC1 (NM_001453.1) was identified at a bit
score of 400 and an e-value of 1e-109. Primer set F1aR1a amplified 315 bp but
when analyzed using a BLAST search, only about 100 base pairs of the forkhead
domain region matched FOXC1 sequence. The majority of the sequence from
the primer set F1aR1a matched highest to FOXCZ, another forkhead gene, at a
BLAST score of 287 and an e-value of 4e-75. Mutations in FOXC2 have been

attributed to ocular malformations (Lehmann et al. 2002). No variability was

86

found in the FOXC1 sequence obtained between unaffected and ASD affected
horses using the pool and sequence method.
FOXE3

FOXE3 is 2,011 bp of mRNA in humans. The coding sequence is from
256 bp to 1215 bp for a total of 959 bp. The forkhead domain lies between 466
bp‘and 732 bp. Seven primer pairs were designed to span from the 5’
untranslated regions of FOXE3 to the 3' untranslated regions with emphasis on
the forkhead domain (Figure 21). Two of the seven primer pairs amplified horse
FOXE3 product.

Forward primer F1 and reverse primer R1 amplified horse product for a
total of 173 bp (Figure 22). All but three nucleotides sequenced were the
forkhead domain. When this sequence was analyzed by BLAST, it matched
human FOXE3 (NM_O12186) at a bit score of 307 and an e-value of 2081.
Forward primer F2 when combined with R1 amplified 162 bp that confirmed the
primer set F1 R1 sequence. No variability was seen in the sequence obtained
between unaffected and ASD affected horses using a pool and sequence

method.

Radiation Hybrid Panel Mapping

Radiation Hybrid mapping of PITX2, FOXC1 and FOXE3 was not
successful. The primer set (exon 2-exon 3) for PITX2 amplified bands of similar
size in horse, human and hamster negating one's ability to distinguish horse

versus hamster origin of product in a horse/hamster radiation hybrid panel.

87

 

Likewise, FOXC1 and FOXE3 products were each similarly sized in hamsters
and horses (Figure 23). This was likely due to the high sequence conservation of

these genes.

Discussion

No variability was detected in the sequenced portions of PITX2, FOXC1
and FOXE3 between the phenotypes of ASD in horses. Polymorphisms may
exist in regions that remain to be sequenced. One of the major problems in
sequencing was the forkhead genes are very GC rich in nucleotide content
(FOXC1 = 71.84% and FOXE3 = 67.67%). Regions that are GC rich are difficult
to amplify since the melting temperatures of the products can be extremely high.
F urtherrnore, the paired primer temperatures frequently did not match. This
problem was addressed by using Taq polymerases that withstood higher
temperatures for denaturation, but this did not resolve the issue. Other factors
that contributed to sequencing problems were large introns and small exons in
PITX2. These factors made designing primers difficult since the regions of
conservation between species were small. A possible solution for this would be
to sequence cDNA, which consists of only coding sequence, however not all
candidate genes are expressed in accessible tissues (blood). Another inherent
problem is the conservation of the forkhead genes. FOXC1 matches FOXC2 at a
92% similarity. This confounded our ability to selectively amplify FOXC1.

PITX2, FOXC1 and FOXE3 are highly conserved across species. Similar

sized PCR product bands were identified in horses and hamsters. Therefore,

88

 

horse/hamster radiation hybrid mapping could not be performed since
researchers would not be able to distinguish between horse and hamster genes.
A way to overcome this obstacle would be to sequence intronic regions where
available, such as in PITX2, which would possible give a more horse specific
sequence so that hamster would not be amplified. This was attempted in horse
PITX2 intron 2, and redesigning the reverse primer in PITX2 intron 2 may help to
alleviate this problem. Another confounding factor with PI'D(2 is the structure.
When the gene was first annotated, it was reported as having four exons. Since
then, several references have identified it as having six .exons with the first exon
being noncoding (Cox et al. 2002). Also, four isoforms have been identified. All
four isoforms have exons 5 and 6 where the homeodomain and OAR domain are
located. PITXZa and PITX2b are produced by alternative splicing mechanisms.
PITX2a contains exons 1, 2, 5, and 6; whereas PITX2b contains exons 1, 2, 3, 5,
and 6. A third isoform, PITXZC uses an alternative promoter upstream of exon 4
so that it contains a part of exon 4, and exons 5 and 6. (Cox et al. 2002). The
function of each isoform is currently being studied, though it has been found that
some isoforms are not seen in all species, such as PITX2d which has so far only
been found in human craniofacial libraries and consists of the 5’ part of exon 4
and exons 5 and 6 (Cox et al. 2002). Better understanding of the structure of
PITX2 and redesigning primers in the newly identified exons may help with the
sequencing problems.

Identifying sequence in the 5’ or 3’ untranslated regions of the forkhead

genes could be beneficial for radiation hybrid mapping, because these genes

89

contain no introns. Unfortunately primers in the untranslated regions of FOXE3
did not amplify horse product. Investigating other techniques for sequencing
these conserved forkhead genes, such as using a horse BAC library or using

plasmids to clone the genes, may aid in obtaining horse gene sequence.

90

Chapter 4

Tables and Figures

91

0:00 .0800 0.5000005 00.0: :0 0:00:00 :00 .0600 .0 :90 0::. 8000.0 0.03

:00 20 :00: 0:00 .0800 2000.00. 039.0 00: 000 6000.0 000:1 00.00 005050 :00: 0.0.5.0 00: 2000.00. 000002600
00... 6.2 00 02000. 0_ 0:0 0 00x0 5 00:32 0_ 0008 t0:0 0::. .0 :96 E 0_ 0000280000.: 050905 c_ 00202:

00 0: 00>0=00 0_ 0203 20:50 «20 00: 0:0 0 :98 c. 00:82 0_ 000:0 :0 v60900.00 000560.000 0::. .000: 0:00
0.0 009:5 00.: .00x00 00 090000.00. 0.0 :00: 0:98 0 000 000:0 0.3250 0000 009.0. 000.00 :0 0.0.005 .0: 0.09“.

m 09:90ch Ar“. 093.0...
m moxm-0c_..v Ann. moxmtmfi

 

 

 

 

    

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0 00.0. i A. n. 00.0
m 0.0me . Au“. 0.0xm
0..me A. .000 1 0.00.0 V A“. 0.05
00 000.: 00 000 00 000 00 :00 00 00: 00 00 00 000
e_ 0.0 v. 0.0 e. 0.0 e. 0.0.. 9. 0.0 e. 0.0..

 

£0,000 m<O 02000. 8000801 0.5..

92

0000.00 000 000.00

0002000 00200000 0.0 000002000 00:00:00_0 00... .00000000 500.:5 00: .5090 0.000. 0000 .032 00: 000 00000000
50008 00: .5090 0.000. .00000 00 0. .m 0-0090 000 0 0-000%”: 0.00.00 0:05 005050 00000000 00.00 0_ 00: 0.0000
00: 00300000 0.000000 0.9. 000.050 0.0.00 .0 800080 000.00 0. 0:: 0o: 0.5 3000000 00.0: 0.30 .0: 0.00..

m000 UHBw<<ow400¢006<uw€0<¢¢¢00dew<wBUDBUBU¢0w0¢UBBw<GUBOw¢€¢¢ mach C0650

03 00,0
000: 0000000000000..n:

 

UWMHBUPU

0a;.1fi

0000000000. _ 0- 000 00000
.00.;00.0.00 00 0000 00000

\L.

0......000... .0. 80:0 0 800

Hmv to U 00 0 O U
Pmmm H 0 ll 0 m U 0

00m u llllllll m 0 I u o 0 00m 00000
H000 0 ouuoouou u u 0 o u u 0000 cmEs:

mmm mmuom
Nomw cmfizm

 

 

mvm u I o u o 0 o 0 1: omm 00000
M000 n u u 0 0 00 0 0000 00800
000 0 0 00 0 0 00 0 00 I 0 0 0mm 00000
0000 0 o 00 0 00 00 00 o 0 0 0000 00500
000 0 0 u 0 o o 00 u u 00 000 00000
0000 00 u 00 u 00 u no u 0 000 0000 00800

 

   

00:
0000

u I: o o o 0000 u o u 0:: 00000
0 00 u u u u 0 0000 0 u 0 0 0:00 00600
00: 0 0 - 0 o 00 000cm
0:00 0 0 0 0 0 0 0000 00000

 

 

   

H0 0 0 . .s u 00» 000 o 0 00000
I... F...._...I.|
m H 00000
0000 ———W0000000000000000000a0000004000000000000000000000000000 0000 00000
N conucH mnmxm m 008000
0000 edomoouuuuououuououuoo00000000000000000000000000uoonUOuoo H000 00600

N coxm H couucH

93

Figure 18. PITX2 Horse Sequence. The top line is human PITX2 sequence
obtained from GenBank (AF238048). The bottom line is horse sequence
obtained with primers Exon 6 F and Exon 6 R. The capital letters signify the
exonic sequence and the lower case letters signify the intronic sequence. The
highlighted nucleotides are conserved between horses and humans. The OAR
domain, which is believed to inhibit transactivation, is underlined.

94

mHVHN ¢<0¢<49U¢0w009w0800¢<0¢90w9¢9B¢¢ow090<0<¢<0BHUDU®UUHU<¢UO¢OG

mom

dodwwhod

mmmHN 9400UUOBaUUflUUflHoueduww00UUOflUfiUUUfiUUUUUflUBOBDUUUOOUD<UUHO<

mmv UQHdHUdUUOHH
oomHN UQE<HU<UUUBB

me
OvNHN

mum
omHHN

mam
ONHHN

mmN
omoam

mmH
oooHN

mmH
ovmow

mp
ommom

ma
ommow

owbom

 

oxm m uwefl

um

U
U B U

 

a e
B
be
m <<Lur<-trt-(.x.c.<<sm<<oaooauwooe¢¢o¢¢0990

 

w xm m Hosaum

oHHHummooooomoommoooovuuoouuowuuumuoumomououomomuoummummuuuu

o coxm

m couucH

omeN

mmv
Homam

va
HvNHN

mhm
HmHHN

mam
HNHHN

mmN
HonN

mmH
HOOHN

mma
vaom

mp
Hmwom

om
ammom

H
Hmhom

Horom

cmfism

wmuom
swan:

omuom
amen:

mwuom
cmfism

wwuom
cuss:

mmuom
cues:

wmuom
Cass:

wmuom
amen:

mono:
amen:

mmuo:
cmEDm

omno:
cmES:

cmEs:

Hgme18.

95

.8385 22> 6: 2c 65 £3 BEE

mom E3659 959.6 9: was 8285 Exon mac; umEEEm $5 Emeta o5 Bowman... mumocsotm 9:. .55 05

co msccaon 9: “a m_ 6.2 638 :8». 9F .vmm new mom $25203: c0053 now—moo. m_ EmEou name—.2 25:98.20
och .xon m mm noucmmoao. m_ 52;; 696 F was axon 93025 ocom Exon cmEac co E9920 .3 2:9“.

 

oEV 40E
03* Low“.
9E 1v AnE
gum 1v mum Aha“. Ann“.
IV
3* AN“. Let...
E v «E Y Lu Aim
8 3m 3 new
_ _

 

 

KJ|K
/ tom”. EmEoD umocvtot

3 News 02

 

96

.mcmEsc new mmflo: comics @95ch 95 $28.20;: 82929: 9:
.EmEou “gamete; 05 m_ 85.68 cmEzc cocEmnc: on... .E 53> E 29:3 £3, 8:650 8533 8.6; m_ cc: Eozon
9: .Ammioolzzv xcmmcmo Ea: 8598 8533 SEE Exon m_ 05. a9 9: .853me 85: Exon .8 9:3

man muddumoomdwHOOUOU<00¢¢0¢¢O<¢UBH000000000000600080UHHUDwaUOU< wmm cmfidm

 

 

wbv Odoooomdoflooehoe<o<<o o mmv wwuo:
hmm DO¢<O<<0¢<UHHoouooooouomoowUOHoUHBUO<UG00<¢G¢GUHH®E<U<<U U hbv cmEd:
va Nmm wmuom
mhv oHv Gmfism
Hmm H Hmm wmuom
mHv U U mmm CMESm
omm B 0 Ohm wmuom
vmm U va :mEdm
mom mom wmuom
mmm mmm CMES:

 

mom 0 o o oowuwuo o U< o B mVH mmuo:
NMN 0 llllll 0 U U 0 U mbH cmESm
wVH H 0 we 0 mm 0 <0 mm wmuo:
th 0 no owoueo o 0 U omH Goad:
vm u U I H 9 III mm wmuo:
mHH U 0 U 0 I ow woo Hm cmfisz
0 B o 000 H wwuox

0 o H U UUHwHUUUHUmHUwawa<UGH¢ H cmESm

Hm HwEHHm

 

     

mm

97

.8335 29> 8: Rt .55 flow .953 «commie wBotm 05 can 6:005 mean 38:

853.5 55 2955 m5 “comma“: mono—Echo on... .53 05 co mcEEmon o5 an m_ 6.5. .covoo tEm och .Nmn new
oov 82820:: c858 2382 2 59:8 townie ozmtmfiflmco ozh .3059. 85.2555 .m can .m 05 3:55 we: a
53> macaw”. .xon m mm 82.8059 m_ :02; .53 H mm... mmXOl 932:3 ocom mmxom :mEac cc 8835 ._.N 939“.

 

3*
EV A F
F; .2...
E: .m 8 8“ 35 , .m 138 . E: .m 3 mmm
_ . . , _ _

 

K coaom EmEoo ummcvto“.

j\

an mmm

 

 

O._.<

98

.EmEou 585:8 05 “.o c0580. 05 m_ @2533 cmEsc 85:05::

on... .mcmEsc can $20; 5053 823:8 9m $5500.03: 5959.59: of. .8533 oonm 05 3:55 Ego.
3:98 0:: .E BEE 0999 53> noEnEoo :0:? E ESE Edie 53> 8:650 8:033 020; m_ 0:: :5on on:
.6290:sz xcmmcoo so: 8598 8838 mmxom $52 2 9.: a9 9: .8538 35: mmxom .NN 9%:

 

mmm U00000000009:.000wUwodfiuoooehoUUUHDwouwoowwwdoowOwwowuuwoowo owb cmfidm
MPH UHHoo/Noowogomou mmH mmuom
m E. 40000UUUOHUG<OUU$UGUU<¢UBeowuwéuooowoooowHUDHBUU<UGQU§U<®U omb cmfism

Hm umEHum

 

 

 

bmH .. , _ . w,, : ;.M mm omuom
ma: Hawk. . . , .-, ,59; any. ; a, ».w._ Hem amasm
:m mm mmuom
com How cmesm
:m o a mmuom
com .rouomoaeueaowoomaoheooo Hem sags:
Hm “meflum
oem Jaoaauou<daJJouqaoaoaaouuuaodowoeuaoaucua<uoooa¢oeooomoa¢o<e awe amen:

 

omv OUBU<HUUUOUUGEOUUUUUU¢JJ._.JJJJ9DJJDJDDJDJJDDJJDDJJJJDJJJJJJJ HNV CMESI

99

 

Figure 23. Agarose Gel of FOXC1 and FOXE3 in hamster, human and horse.
Products of similar size were obtained in all three species. This could be due to
the high conservation of the forkhead genes. Although the sequence may be
different between the three species, the radiation hybrid panel will not
differentiate between horse and hamster gene location. The “L" qunifies a 100 bp
ladder where the bright band is at 600 bp. Strong bands are seen at ~350 bp for
FOXC1 and ~150 bp for FOXE3 in all species.

100

CHAPTER 5

Conclusions and Future Research

Although results show that the three candidate genes for this studyodo not
seem to be linked to ASD in horses, further research should be done to look at
other aspects of the genes. The sequence analysis of the genes needs to be
completed both within the genes and in the surrounding upstream and
downstream areas because these areas can also affect gene expression. lf
subsequent linkage and association results give reason for the genes to be
studied further, gene expression should also be investigated since defects could
be manifested during transcription or translation. Identifying the equine
chromosomal locations of the candidate genes would also aid in the study
greatly. Map locations would allow for greater efficiency in marker selection,
because markers could be chosen that were known to be linked to the candidate
genes, thus giving a higher likelihood of identifying linkage with the phenotype. If
the horse genome was better characterized, a whole genome scan would give a
better idea of the location of the disease locus, so genes known to be in that area
could be prioritized.

Another option for furthering this study would be to look at other candidate
genes. One such candidate is FOXCZ. FOXCZ is part of a cluster of forkheads
(FOXCZ/FOXF1IFOXL1), which is located on human 16q24. This location lies

within the interval for a second Rieger syndrome locus (Lehmann et al. 2002).

101

Human FOXC2 is also 92% similar in nucleotide sequence to human FOXC1, so
it has been inferred that they may have similar function.

Other horse breeds have also been identified as having a cyst phenotype.
Typically, these are breeds that have a chocolate coat color, such as the Belgian
and Miniature horse (Ramsey, Personal Communication. 2002). It may be
interesting to include these breeds in the study to identify differences in the
candidate genes between breeds. More research in the areas of the horse
genome characterization and in equine ocular development is needed in order to
answer questions regarding the mechanisms that underlie Anterior Segment

Dysgenesis in horses.

102

Appendix A

103

QlAquick Gel Extraction Kit Protocol (Taken froleAquick Spin Handbook)

This protocol is used to extract and purify DNA of 70 bp to 10 kb from
standard or low-melt agarose gels in TAE or TBE buffer.

1.

Excise DNA fragment from the agarose gel with a clean, sharp scalpel.
Minimize the gel slice size by removing extra agarose. Place in a 1.5
ml tube.

2. For average size gel pieces add 500 pl of Buffer 06 to the gel slice.

3. Incubate at 50° C for about 10 minutes or until the gel is completely

9.

dissolved. Mix by vortexing every 2-3 minutes.

Make sure the color stays a yellowish similar to the Buffer 06 color. If
it is orange or violet add 10 pl of 3 M sodium acetate and mix.

Add about 100 pl of isopropanol for DNA fragments <500 bp and >4
kb.

Place a QlAquick spin column in a provided 2-ml collection tube.

To bind DNA, apply the sample to QlAquick column and centrifuge for
1 min.

Discard flow-through and place QlAquick column back .in the same
collection tube.

Add .5 ml of Buffer 00 to QlAquick column and centrifuge for 1 min.

10. To wash, add .75 ml of Buffer PE to column and centrifuge for 1 min.

Let column stand for 2-5 minutes.

11. Discard the flow—through and centrifuge for an additional minute at

210,000 x g.

12. Place QlAquick column into a clean 1.5 ml microfuge tube.
13.To elute DNA, add 35 pl of dd Water to the center of- the column, let sit

for 1 minute and centrifuge for 1 minute at maximum speed.

14. Store the purified DNA in the freezer.

104

Appendix B

105

Thermo Sequenase Radiolabelled Terminator Cycle Sequencing Kit

Taken directly from the Brief Protocol:

1.

Choose the termination master mix, dGTP or leP, appropriate for
your particular sequencing application. leP is often used for
sequences that are 60 rich or have compression bands. dGTP is used
for standard sequences.
Prepare the 4 termination mixes using 2.0 pl of termination master mix
and .5 pl of labeled ddNTP for each sequence.
Dispense termination mixes: Label 4 tubes (‘6', ‘A’, ‘T', ‘C') for each
sequence. Fill tubes with 2.5 pl of the appropriate termination mix
prepared in step 2 and cap.
Prepare the reaction mixture. Combine the following:

a. Reaction Buffer - 2 pl

b. DNA - 5 pl

c. Primer - 1pl

d. H20 - 10 pl

e. Thermo Sequenase DNA polymerase — 2 pl
Cycling termination reactions: Transfer 4.5 pl of the template and
polymerase mixture from step 4 to each termination tube (‘G’, ‘A’, ‘T',
‘C) from step 3. Mix well and place the tube in a thermal cycler.
Cycle 30—60 times as follows:

dGTP leP

95°C 303, 95°C 308,
55°C 308, 50°C 308,
72°C 1min 60°C 5-10 min

Heat samples to 70°C for 2-10 mins immediately before loading onto
sequencing gel.

106

Appendix C

107

Procedure for Gel Preparation with AccuGel 19:1 (BioWhittaker Molecular
Applications, ME) Taken directly from acrylamide protocol.

The acrylamide percentage to be used depends on the size of the nucleic
acid acid fragments to be separated. The greater the number of base pairs to be
separated, the larger the pore size required and therefore the lower the
acrylamide percentage to be used. A 6% gel can separate a size range of 60—250
bp. The bromophenol blue will run at 25 nucleotides and xylene cyanol runs at
110 nucleotides in the denaturing AccuGel 19:1 gels.

To make a 5.75% gel:

Combine:

15 ml of 40% AccuGel

42 g of 7 M Urea

10 ml of 10X TBE

Fill to 100 ml with Distilled Water

Add 1 ml of 10% (freshly-made) Ammonium Persulfate for every 100ml of
gel casting solution. Swirl gently to mix and add 20 pl of TEMED for every 100ml
of gel casting solution. Pour the solution into the plates. The gel sets up 10-20
minutes after. Allow gel to set up for 1 ‘/z to 2 hours after pouring." After 2 hours
of setting up, the gel and plates can be wrapped in plastic and stored for up to 48
hours.

108

Appendix D

109

ABI Genescan lnforrnation (Taken from GeneScan manual)

Rules for Dye Choice:
1. Each filter set should be used with a combination of four or fewer dyes
(including the dye reserved for the internal lane size standard) that:
a. Display in four different colors.
b. Are available in compatible chemical forms.
2. You can combine phosphoramidite labels with NHS-ester labels. You
should not combine [F]dNTP-labeling with any other labeling method.

Table 11. Dye Wavelengths

Emission Excitation

 

Dye (nm) (nm)
6-FAM 517 494
5—FAM 522 493
R110 525 501
TET 538 522
R66 549 529
HEX 553 535
JOE 554 528
[F]dNTP

TAMRA 572 555
NED 575 553
TAMRA 583 560
ROX 607 587

Table 12. Dye Combinations

Dye Combinations Filter Set
6-Fam, HEX, TAMRA, ROX (Std)
R110, R66, TAMRA, ROX (Std)

 

A
5-FAM, JOE, TAMRA, ROX (Std)
6-FAM, TET, HEX, TAMRA (Std) C
6-FAM, HEX, NED, ROX (Std) D
5-FAM, JOE, NED, ROX (Std) F

110

Table 13. Dye Availability and Filter Sets

Color Display (In virtual

 

Dye Available as: set)

5-FAM Labeled primer in kits Blue (A,F)

6-FAM Phosphoramidite Blue (A,C,D)

R1 10 [F]dNTP Blue (A)

TET Phosphoramidite Green (C )

JOE Labeled primer in kits Green (A,F)

R66 [F]dNTP Green (A)

HEX Phosphoramidite Green (A,D), Yellow (C )

NED NHS-ester, Phosphoramidite Yellow (D,F)
NHS-ester, [F]dNTP, or GeneScan internal

TAMRA lane size standard Yellow (A), Red (c )
NHS-ester or GeneScan internal lane size

ROX standard Red (A,D,F)

111

APPENDIX E

112

Table 14. Plate 1 Genoty

s for HMS42, UM1 1 and A14

 

 

 

199031696 LLFSAZZ f i
__F_arm PM Allele 1‘ Allele 2‘ Allele 1‘ Allele 2" Allele 1‘ Allele 2‘
626 ASD 130.26 160.57 160.57 7 7
645 ASD 123.69 131.56 171.1 171.1 225.67 225.67
846 ASD 122.79 130.67 160.57 171.03 7 7
B52 ASD 131.92 131.92 160.49 164.12 7 7
096 ASD 7 7 7 7 234.92 234.92
N6 ASD 130.83 130.63 170.16 170.16 7 7
64 ASD 122.79 130.92 160.41 164.02 225.74 234.36
612 ASD 130.24 130.24 166.28 170.29 7 7
K8 ASD 122.69 132.64 156.55 172.09 223.59 223.59
650 ASD 130.77 130.77 167.29 170.29 225.73 225.73
653 ASD 130.2 130.2 164.24 170.27 225.69 229.97
656 ASD 122.66 122.66 160.49 164.09 225.66 226.96
621 Cyst 123.36 130.46 164.36 170.3 225.74 234.4
C7 ASD 131.75 131.75 170.3 170.3 223.56 225.66
05 ASD 130.73 130.73 164.19 171.13 225.74 225.74
K16 ASD 124.04 131.67 168.51 170.4 212.93 225.74
627 A60 130.76 130.76 163.92 167.13 223.53 225.74
F16 ASD 122.91 130.75 160.4 164.26 223.23 233.65
Q31 ASD 130.99 130.99 170.33 170.33 217.16 225.63
u3 ASD 122.96 131.29 164.3 170.29 223.52 234.26
W8 Aso 130.76 130.76 170.29 170.29 7 7
646 ASD 131.07 131.07 164.57 170.37 223.52 225.62
012 ASD 122.93 130.6 156.66 160.69 214.96 225.68
F8 ASD 130.8 130.6 160.47 166.54 7 7
K5 ASD 130.66 130.66 170.16 170.16 7 7
KKKKK13 ASD 122.93 122.93 164.11 164.11 7 7
we ASD 130.72 130.72 160.5 164.05 7 7
Blank
B10 ASD 124.73 131.14 163.99 163.99 7 7
B15 ASD 130.56 130.56 164.16 170.16 7 7
623 A50 130.67 130.67 157.91 169.41 223.52 223.52
626 ASD 131.72 131.72 164.06 170.1 7 7
630 A50 122.74 130.69 170.14 170.14 225.62 225.62
639 ASD 132.22 132.22 7 7 225.67 234.21
813 ASD 122.74 130.69 164.26 166.35 223.67 225.66
l9 ASD 130.77 130.77 164.36 170.25 223.56 225.76
A1 ASD 131.63 131.63 160.65 164.46 7 7
660 ASD 123.21 131.01 164.16 171.04 223.62 223.62
019 ASD 122.63 130.74 164.45 171.01 223.62 229.97
c21 ASD 131.69 131.69 160.56 160.56 7 7
C33 ASD 130.61 130.61 170.3 170.3 212.96 223.62
M4 ASD 122.65 130.79 7 7 7 7
647 ASD 130.11 130.11 160 177.93 223.63 223.63
c1 ASD 131.49 131.49 160.64 171.03 223.57 225.63
628 ASD 130.73 130.73 160.31 160.31 223.63 225.66
L3 ASD 122.93 130.76 164.5 171.09 223.53 223.53
09 ASD 130.76 130.76 166.48 166.46 7 7
024 ASD 130.72 130.72 164.37 166.48 7 7
637 ASD 130.73 122.9 160.56 164.29 7 7
013 ASD 7 7 164.22 166.3 7 7
W9 ASD 131.58 131.56 164.05 164.05 223.57 223.57
2 ASD 7 7 164.11 170.16 7 7
620 ASD 123.52 131.37 159.74 170.21 7 7
037 ASD 7 7 164.1 164.1 7 7

 

 

 

 

113

Table 14 (Cont). Plate 1 Genoypesjor HMS42, UM11 and A14

 

 

 

 

 

1666696 M912- 2121.1: 6113
Farm Phenotype Allele 1‘ Allele 2‘ Allele 1‘ Allele 2‘ Allele 1‘ Allele 2‘
Blank
064 ASD 7 7 169.92 160.17 7 7
D16 ASD 7 7 160.5 170.2 7 7
69 ASD 131.56 131.56 156.29 164.13 7 7
u1 ASD 7 7 156.55 164.15 225.63 234.14
W7 ASD 7 7 7 7 225.63 225.63
Tim Cyst 130.57 130.57 160.65 164.29 225.64 225.64
K11 ASD 122.96 130.61 156.57 166.24 223.57 225.64
R8 ASD 132.19 132.19 160.56 170.14 223.61 225.59
R9 ASD 130.44 130.44 166.29 170.25 225.63 225.63
625 Aso 122.66 130.69 164.16 164.16 7 7
634 ASD 131.59 131.59 164.34 164.34 7 7
026 A50 131.6 131.6 166.36 171.09 217.23 225.63
622 ASD 122.63 130.7 170.66 170.68 225.56 225.56
C29 ASD 130.73 130.73 166.4 171.05 225.67 225.67
F15 ASD 125.76 130.91 164.26 170.39 223.5 225.57
F19 Aso 130.64 130.64 160.63 166.49 225.72 234.32

KKKKK14 Normal 130.67 130.67 160.4 164.26 223.59 7225.67
F16 Aso 122.77 130.62 164.22 164.22 223.26 225.36
029 ASD 7 7 7 7 7 7
VMC1 Aso 122.77 130.71 160.17 164.06 225.44 225.44
211 Normal 131.64 131.64 166.16 156.99 7 7
06 Cyst 130.33 130.33 164.01 170.95 7 7
Q19 Cyst 122.64 130.66 164.06 164.08 7 7
U4 Cyst 132.69 123.66 7 7 7 7
us Cyst 122.66 130.65 170.2 170.2 225.36 233.9
06 Cyst 122.64 130.59 163.98 170.6 225.32 225.32

KKKKK11 Normal 130.37 130.37 7 7 212.77 233.96
020 Cyst 130.94 130.94 160.49 164.12 225.37 225.37
73 Cyst 130.66 130.66 164.12 170.06 223.28 225.36
65 Cyst 122 130 7 7 225.32 225.32
G1 Cyst 131.1 131.1 160.65 162.5 223.36 225.38
0114 Normal 132.73 13273 162.52 162.52 223.26 223.26
0121 Normal 131.03 131.03 164.14 164.14 223.26 223.26
066 Normal 131.6 131.6 164.42 166.6 223.37 225.38
0125 Normal 122.75 130.69 160.64 160.64 223.32 225.32
D136 Normal 123.49 133.66 166.45 166.45 223.37 223.37
629 Normal 131.01 164.25 167.29 7 7
28 Normal 7 ? 168.36 168.36 225.38 225.38
037 Normal 130.81 130.61 164.29 170.29 225.43 225.43
U7 Normal 130.72 130.72 156.61 176.05 223.32 233.93
25 Normal 300.11 300.11 365.67 365.87 7 7

 

' Alleles for HMS42 were binned as follows: 1-122 and 123 bp. 2- 128 bp. 3» 130 bp. 4- 132 bp.

2 Alleles for UM11 were binned as follows:1- 156 bp. 2158 bp. 3- 160 bp. 4- 164 bp. 5- 166 and 167 bp. 6-
168 bp. 7- 170 and 171 bp. 8-172 bp

3 Alleles for A14 were binned as follows:1 - 212 and 213 bp. 2 - 214 and 215 bp. 3 - 217 bp. 4 - 223 bp. 5 -
224 and 225 bp. 6 ~ 229 bp. 7- 233 bp. 8- 234 and 235 bp.

‘ Allele sizes are given in raw data format All markers were dinucleotide repeats so alleles

were binned by 2 bp units for data analysis.

?- Unknown genotype

114

Table 15. Plate 2 Genotypes for HMS42 and UM11

 

 

 

lgeng'llers H7754; ([5411
Farm 17 Gen-ID#‘ Allele j Allele 3’ Genotype° Allele 12 Allele 22 Genotype‘
A7 Vl-9 130.74 130.74 373 164.2 164.2 474
611 Vl-60 131.04 131.04 373 166.37 172.26 5/8
612 VII-63 130.79 130.79 373 166.27 170.35 577
614 VIII-10 131.65 131.65 373 164.05 170.94 477
615 lX-11 353.47 353.47 777 405.64 413.94 777
B16 lx-5 7 7 777 164.06 164.06 474
617 X-6 130.79 130.79 373 164.17 164.17 474
620 lX-97 122.92 130.44 173 170.4 170.4 777
621 Vl-7 122.76 130.71 173 164.04 170.09 477
B23 X-18 332.62 332.62 777 363.67 369.69 777
625 x-119 122.62 130.62 173 164.13 164.13 474
626 X-118 131.56 131.56 373 164.16 170.12 477
627 lX-98 130.72 130.72 373 164.35 170.13 477
63 lX-9 122.66 130.76 173 160.64 170.29 377
630 lX-12 123.01 130.88 173 170.4 170.4 777
631 x-12 131.69 131.69 373 164.13 171.01 477
632 Xl-9 122.77 130.71 173 164.06 164.06 474
833 lx-4 130.62 130.62 373 164.16 164.16 474
634 x-s 130.66 130.66 373 164.36 164.36 474
635 vm-s 122.66 130.73 173 166.44 170.3 577
636 lX-2 122.65 130.79 173 7 7 777
637 VIII-128 7 7 777 160.64 164.46 374
636 lX-111 7 7 777 160.41 164.14 374
639 VII-65 131 131 373 167.03 170.42 577
64 x-11 122.77 130.67 173 160.42 164.07 374
642 VII-3 122.96 130.66 173 156.66 170.24 277
643 Vl-12 130.62 130.62 373 156.59 164.21 274
645 lX-112 130.66 130.66 373 164.07 164.07 474
647 X—16 131.56 131.56 373 164.06 170.1 477
649 lX-10 122.96 122.96 173 164.64 167.31 475
65 lX-68 122.99 130.82 173 156.55 166.34 275
650 x-14 130.65 130.65 373 166.66 170.12 577
652 M7 130.92 130.92 373 160.63 164.27 374
653 Xl-17 130.65 130.65 373 164.05 170.1 477
656 Xl-48 122.79 130.76 173 160.42 164.07 374
657 VIII-41 130.32 130.32 373 7 7 777
656 lX-45 123.92 130.66 173 164.16 164.16 474
66 X-85 123.9 131.06 173 156.61 164.25 274
B60 VIII-105 123.37 131.69 173 164.19 171.12 477
862 vm-49 130.22 130.22 373 160.31 170.22 377
67 VIII-24 7 7 777 160.41 170.91 377
66 lX-29 130.76 130.76 373 170.33 170.33 777
69 Vlll-17 124.7 132.64 174 163.99 172 476
C1 VII-50 131.56 131.56 373 160.33 170.96 377
C26 x-7 130.76 130.76 373 164.55 171.1 477
C7 X-68 130.75 130.75 373 156.49 170.21 277
D11 lX-95 130.87 130.67 373 164.26 170.37 477

 

 

 

115

 

 

Table 15 (gm!) Plate 2 Genotypes for HMS42 and UM1 1
0m HMS42 UM11
Farm 77 Gen-ID#‘ Allele 12 Allele 22 Genotypes Allele 12 Allele 22 Genotype“

0136 Vl-55 123.56 132.75 174 166.21 166.21 575
033 mm 123.06 131.02 173 160.32 166.37 375
034 VIII-54 131.66 131.66 373 164.49 166.5 475
036 lX-54 131.59 133 374 164.2 164.2 474
037 IX-19 131.02 131.02 373 164.32 166 475
05 x-117 131.91 131.91 373 164.12 170.97 477
06 IX-117 130.48 130.46 373 164.06 170.94 477
064 Vl-51 131.91 131.91 373 164.03 164.03 474
D65 VII-47 130.53 130.53 373 164.16 170.1 477
D8 X-116 130.66 130.68 373 164.2 170.22 477
062 VII-64 7 7 777 166.59 170.36 577
064 x-120 123.7 130.67 173 170.33 170.33 777
095 VIII-119 130.81 130.61 373 164.37 170.39 477
096 x-124 122.77 130.66 173 7 7 777
E1 Vl-38 131.69 131.69 373 160.7 170.35 377
64 VIII-136 131.16 131.16 373 164.24 164.24 474
I3 was 130.78 130.76 373 164.44 164.44 474
I4 X90 7 7 777 164.29 164.29 474
IS lX-33 130.72 130.72 373 164.21 164.21 474
I6 x-46 130.79 130.79 373 164.51 170.31 477
J5 lx-96 122.96 130.69 173 164.16 171.04 477
J6 lx-11o 130.77 130.77 373 156.7 164.23 174
KKKKK14 X-61 131.79 131.79 373 160.17 164.01 374
03 Vl-13 130.78 130.76 373 164.42 164.42 474
019 X-76 7 7 '77? 164.12 164.12 474
036 Vll-13 130.78 130.76 373 170.42 170.42 777
05 Vl-2o 122.71 130.67 173 164.12 171.06 477
Q6 x-21 7 7 777 164.13 171 477
R10 x-115 130.79 130.79 373 164.21 170.27 477
R11 v.29 122.71 122.71 171 170.37 170.37 777
R12 Vl-64 122.65 130.66 173 164.12 170.98 477
R2 v-3o 131.56 131.56 373 164.04 170.15 477
R3 Vll-69 130.72 130.72 373 170.26 170.26 777
R4 x-123 130.75 130.75 373 164.03 170.05 477

R5 VII-10 130.76 130.76 373 7 7 7
Tim v.12 131.66 131.66 373 160.5 164.13 374
U6 Vl-16 122.98 130.86 373 171 171 777
w1 x-114 131.63 131.63 373 164.22 164.22 474
w1o X-169 7 7 777 164.13 164.13 474
w2 x-139 130.62 130.62 373 164.07 164.07 474
we X56 7 7 777 164.13 172.09 476
W5 Vl-33 130.36 130.36 373 160.17 168.21 376
w7 IX-22 131.46 131.48 373 164.17 164.17 474
we x-29 130.62 130.62 373 164.1 164.1 474
W9 lx-1oo 130.69 130.69 373 166.62 168.62 474

 

 

 

' Gen-lD# is the generation number of the animal and the number of the animal in that generation.

2 Allele sizes are given in raw data format. All markers were dinucleotide repeats so alleles were binned
by 2 bp units for data analysis.

3 Alleles for HMS42 were binned as follows: 1-122 and 123 bp. 2- 128 bp. 3- 130 bp. 4- 132 bp.

‘ Alleles for UM11 were binned as follows: 1- 156 bp. 2- 158 bp. 3- 160 bp. 4- 164 bp. 5- 166 and 167 bp.
6- 168 bp. 7- 170 and 171 bp. 8-172 bp

?- Unknown genotype

116

REFERENCES

Andrew SE, Ramsey DT, Hauptman JG, Brooks DE (2001) Density of
corneal endothelial cells and corneal thickness in eyes of euthanatized
horses. Am J Vet Res 62, 479-482.

Ardlie KG, Kruglyak L, Seielstad M (2002) Patterns of linkage
disequilibrium in the human genome. Nat Rev Genet 3, 299-309.

Blixt A, Mahlapuu M, Aitola M, Pelto-Huikko M, Enerback S. (2000) A
forkhead gene, FOXE3, is essential for lens epithelial proliferation and
closure of the lens vesicle. Genes Dev 14, 245-254.

Bonini NM, Bui QT, Gray-Board GL, Warrick JM (1997) The Drosophila
eyes absent gene directs ectopic eye formation in a pathway conserved
between flies and vertebrates. Development 124, 4819-4826.

Brouillette JA, Andrew JR, Venta PJ (2000) Estimate of nucleotide
diversity in dogs with a pool-and-sequence method. Mamm Genome 11,
1079-1086.

Caetano AR, Lyons LA, Laughlin TF, O'Brien SJ, Murray JD. (1999)
Equine synteny mapping of comparative anchor tagged sequences
(CATS) from human Chromosome 5. Mamm Genome 10, 1082-1084.

Cardon LR, and Bell Jl (2001) Association study designs for complex
diseases. Nat Rev Genet 2, 91-99.

Chowdhary BP, Raudsepp T, Honeycutt D, Owens EK, Piumi F. (2002)
Construction of a 5000(rad) whole-genome radiation hybrid panel

in the horse and generation of a comprehensive and comparative map
for ECA11. Mamm Genome 13, 89-94.

Churchill A, and Booth A (1996) Genetics of aniridia and anterior segment
dysgenesis. Br J Ophthalmol 80, 669-673.

Collins DW, and Jukes TH (1994) Rates of transition and transversion in

coding sequences since the human-rodent divergence. Genomics 20,
386-396.

Cook C. (1999) Chapter 1. Ocular Embryology and Congenital
Malformations. Veterinary Ophthalmology (3rd edn). Philadelphia:
Lippincott Williams & Wilkins.

117

Cox CJ, Espinoza HM, McWilliams B, Chappell K, Morton L, Hjalt TA,
Semina EV and Amendt BA. (2002) Differential regulation of gene
expression by PITX2 isoforms. The Journal of Biological Chemistry
277228, 25001-25010.

Crispin SM (2000) Developmental anomalies and abnormalities of the
equine iris. Vet Ophthalmol 3, 93-98.

Crispin SM, Matthews AG, Parker J (1990) The equine fundus. l:
Examination, embryology, structure and function. Equine Vet J Suppl,
42-49.

Czerny T, Halder G, Kloter U, Souabni A, Gehring WJ. (1999) twin of
eye/ass, a second Fax-6 gene of Drosophila, acts upstream of eye/ass
in the control of eye development. Mol Cell 3, 297-307.

Davies AF, Mirza G, Flinter F, Ragoussis J (1999) An interstitial deletion of
6p24-p25 proximal to the FKHL7 locus and including AP-26Ipha that
affects anterior eye Chamber development. J Med Genet 36, 708-710.

Dellovade TL, Pfaff DW, Schwanzel-Fukuda M (1998) The gonadotropin-
releasing hormone system does not develop in Small-Eye (Sey) mouse
phenotype. Brain Res Dev Brain Res 107, 233-240.

Diesem C (1968) Gross anatomic structure of equine and bovine orbit and
its contents. Am J Vet Res 29, 1769-1781.

Doward W, Perveen R, Lloyd IC, Ridgway AE, Wilson L. (1999) A
mutation in the RIEG1 gene associated with Peters' anomaly. J Med
Genet 36, 1 52-1 55.

Ehrenhofer MC, Deeg CA, Reese S, Liebich HG, Stangassinger M.
(2002) Normal structure and age-related changes of the equine retina.
Vet Ophthalmol 5, 39-47.

Espinoza HM, Cox CJ, Semina EV, Amendt BA (2002) A molecular basis
for differential developmental anomalies in Axenfeld- Rieger syndrome.
Hum Mol Genet 11, 743-753.

Ewart SL, Ramsey DT, Xu J, Meyers D (2000) The horse homolog of
congenital aniridia conforms to codominant inheritance. J Hered 91, 93-
98.

Flomen RH, Vatcheva R, Gorman PA, Baptista PR, Groet J. (1998)

Construction and analysis of a sequence-ready map in 4q25: Rieger
syndrome can be caused by haploinsufficiency of RIEG, but also by

118

chromosome breaks approximately 90 kb upstream of this gene.
Genomics 47, 409-413.

FLYBASE Drosophila Database. (August 2002). flybase.bio.indiana.edu/

Forrester J, Dick A, McMenamon P, Lee W (2002) The eye. Basic
Sciences in Practice. Edinburgh: WB Saunders.

Gelatt K ( 1999) Veterinary Ophthalmology (3rd edn). Philadelphia:
Lippincott Williams & Wilkins.

Gibson G, and Muse S (2002) A primer of genome science. Sunderland,
Mass:Sinauer.

Goss SJ, and Harris H (1975) New method for mapping genes in human
Chromosomes. Nature 255, 680-684.

Green P, Falls K, Crooks S (1990) CRIMAP 2.4 Computer Software.

Guerin G, Bailey E, Bernoco D, Anderson I, Antczak DF. (1999)
Report of the lntemational Equine Gene Mapping Workshop: male
linkage map. Anim Genet 30, 341-354. *

Gum G, Gelatt K, Ofri R. (1999) Chapter 3. Physiology of the Eye.
Veterinary Ophthalmology (3rd edn) Philadelphia: Lippincott Williams &
Wilkins.

Hamada H, Meno C, Saijoh Y, Adachi H, Yashiro K. (2001) Role of
asymmetric signals in left-right patterning in the mouse. Am J Med
Genet 101, 324-327.

Hanson IM, Fletcher JM, Jordan T, Brown A, Taylor D. (1994)
Mutations at the PAX6 locus are found in heterogeneous anterior

segment malformations including Peters' anomaly. Nat Genet 6, 168-
173.

Harman AM, Moore 8, Hoskins R, Keller P (1999) Horse vision and an
explanation for the visual behaviour originally explained by the 'ramp
retina'. Equine Vet J 31, 384-390.

Hartl D (2000) A Primer of Population Genetics. Massachusetts: Sinauer
Associates.

Heon E, Sheth BP, Kalenak JW, Sunden SL, Streb LM. (1995)

Linkage of autosomal dominant iris hypoplasia to the region of the
Rieger syndrome locus (4q25). Hum Mol Genet 4, 1435-1439.

119

Hill RE, Favor J, Hogan BL, Ton CC, Saunders GF. (1991) Mouse
small eye results from mutations in a paired-like homeobox- containing
gene. Nature 354, 522-525.

International Human Genome Sequencing Consortium. (2001) Initial
sequencing and analysis of the human genome. Nature 409, 860-920.

Johns MB, Jr., Paulus-Thomas JE (1989) Purification of human genomic
DNA from whole blood using sodium perchlorate in place of phenol.
Anal Biochem 180, 276-278.

Jorde LB (1995) Linkage disequilibrium as a gene-mapping tool. Am J
Hum Genet 56, 11-14.

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

Kaufrnann E, and Knochel W (1996) Five years on the wings of fork head.
Mech Dev 57, 3-20.

Kawase C, Kawase K, Taniguchi T, Sugiyama K, Yamamoto T. (2001)
Screening for mutations of Axenfeld-Rieger syndrome caused by
FOXC1 gene in Japanese patients. J Glaucoma 10, 477-482.

Kiguwa SL, Hextall P, Smith AL, Critcher R, Swinburne J. (2000) A
horse whole-genome-radiation hybrid panel: chromosome 1 and 10
preliminary maps. Mamm Genome 11, 803-805.

Kozlowski K, and Walter MA (2000) Variation in residual PITX2 activity
underlies the phenotypic spectrum of anterior segment developmental
disorders. Hum Mol Genet 9, 2131-2139.

Lehmann OJ, Ebenezer ND, Ekong R, Ocaka L, Mungall AJ. (2002)
Ocular developmental abnormalities and glaucoma associated with
interstitial 6p25 duplications and deletions. Invest Ophthalmol Vis Sci 43,
1843-1849. - -

Lewin B (2000) Genes VII. North Yorkshire: Oxford University Press.

Lines MA, Kozlowski K, Walter MA (2002) Molecular genetics of Axenfeld-
Rieger malformations. Hum Mol Genet 11, 1177-1 187.

Liu C, Liu W, Lu MF, Brown NA, Martin JF (2001) Regulation of left-right

asymmetry by thresholds of Pitx20 activity. Development 128, 2039-
2048.

120

Matthews AG, Crispin SM, Parker J (1990) The equine fundus. ll: Normal
anatomical variants and colobomata. Equine Vet J Suppl, 50-54.

McCarthy LC (1996) Whole genome radiation hybrid mapping. Trends
Genet 12, 491-493.

Mears AJ, Jordan T, Mirzayans F, Dubois S, Kume T. (1998)
Mutations of the forkheadlwinged-helix gene, FKHL 7, in patients with
Axenfeld-Rieger anomaly. Am J Hum Genet 63, 1316-1328.

NCBl National Center for Biotechnology Information. (September 2002).
www.ncbi.nlm.nih.gov/

Nishimura DY, Searby CC, Alward WL, Walton D, Craig JE et al. (2001) A
spectrum of FOXC1 mutations suggests gene dosage as a mechanism
for developmental defects of the anterior chamber of the eye. Am J Hum
Genet 68, 364-372.

Nishimura DY, Swiderski RE, Alward WL, Searby CC, Patil SR. (1998)
The forkhead transcription factor gene FKHL7 is responsible for
glaucoma phenotypes which map to 6p25. Nat Genet 19, 140-147.

Orrnestad M, Blixt A, Churchill A, Martinsson T, Enerback S. (2002)
Foxe3 haploinsufficiency in mice: a model for Peters' anomaly. Invest
Ophthalmol Vis Sci 43, 1350-1357.

Ott J (1992) Analysis of Human Linkage. Boston: The Johns Hopkins
University Press.

Peiffer Jr R, Wilcock B, Dubielzig R, Render J, Whiteley H. (1999) Chapter
9. Fundamentals of Veterinary Ophthalmic Pathology. Veterinary
Ophthalmology (3ml edn) Philadelphia: Lippincott Williams 6.

Wilkins.

Phillips JC, del Bono EA, Haines JL, Pralea AM, Cohen JS. (1996) A
second locus for Rieger syndrome maps to chromosome 13q14. Am J
Hum Genet 59, 613-619.

Prince J, Diesem C, Egilitis l, Ruskell G. (1960) Anatomy and Histology of
the Eye and Orbit in Domestic Animals. Illinois: CC Thomas.

Prosser J, van Heyningen V (1998) PAX6 mutations reviewed. Hum Mutat
1 1 , 93-108.

Quiring R, Walldorf U, Kloter U, Gehring WJ (1994) Homology of the
eyeless gene of Drosophila to the Small eye gene in mice and Aniridia in

121

humans. Science 265, 785-789.
Ramsey DT (2002) Personal Communication.

Ramsey DT, Ewart SL, Render JA, Cook CS, Latimer CA (1999a)
Congenital ocular abnormalities of Rocky Mountain Horses. Vet
Ophthalmol 2, 47-59.

Ramsey DT, Hauptman JG, Petersen-Jones SM (1999b) Corneal
thickness, intraocular pressure, and optical corneal diameter in Rocky
Mountain Horses with cornea globosa or clinically normal corneas. Am J
Vet Res 60, 1317-1321.

Richer CL, Power MM, Klunder LR, McFeely RA, Kent MG (1990)
Standard karyotype of the domestic horse (Equus caballus). Committee
for standardized karyotype of Equus caballus. The Second lntemational
Conference for Standardization of Domestic Animal Karyotypes, INRA,
Jouy-en Josas, France, 22nd-26th May 1989. Hereditas 112, 289-293.

Rychlik P, and Rychlik W (2002) Oligo. Primer Design Software 6.
Computer Software.

Saleem RA, Banerjee-Basu S, Berry FB, Baxevanis AD, Walter MA (2001)
Analyses of the effects that disease-causing missense mutations have
on the structure and function of the winged-helix protein FOXC1. Am J
Hum Genet 68, 627-641.

SAS Institute I. (1999-2001) The SAS System for Windows 8.02.
Computer Software.

Schedl A, Ross A, Lee M, Engelkamp D, Rashbass P et al. (1996)
Influence of PAX6 gene dosage on development: overexpression
causes severe eye abnormalities. Cell 86, 71-82.

Semina EV, Brownell I, Mintz-Hittner HA, Murray JC, Jamrich M (2001)
Mutations in the human forkhead transcription factor FOXE3 associated

with anterior segment ocular dysgenesis and cataracts. Hum Mol Genet
1 0, 231 -236.

Semina EV, Reiter R, Leysens NJ, Alward WL, Small KW et al. (1996)
Cloning and characterization of a novel bicoid-related homeobox
transcription factor gene, RIEG, involved in Rieger syndrome. Nat Genet
14, 392-399.

Shubitowski DM, Venta PJ, Douglass CL, Zhou RX, Ewart SL (2001)
Polymorphism identification within 50 equine gene-specific sequence

122

tagged sites. Anim Genet 32, 78-88.

Smith RS, Zabaleta A, Kume T, Savinova OV, Kidson SH et al. (2000)
Haploinsufficiency of the transcription factors FOXC1 and FOXCZ
results in aberrant ocular development. Hum Mol Genet 9, 1021-1032.

Swinburne J, Gerstenberg C, Breen M, Aldridge V, Lockhart L et al. (2000)
First comprehensive low-density horse linkage map based on two 3-

generation, full-sibling, cross-bred horse reference families. Genomics
66, 123-134.

Van Heyningen V, Williamson KA (2002) PAX6 in sensory development.
Hum Mol Genet 11, 1161-1167.

123