Li 3 RA RY

> Michigan State
2, 0 If) ‘ U. .r-sersxy i

 

 

-"f1—w

This is to certify that the
dissertation entitled

IDENTIFICATION OF GENE-SPECIFIC SINGLE
NUCLEOTIDE POLYMORPHISMS WITHIN THE CANINE
GENOME AND THEIR USE TO DETERMINE NUCLEOTIDE
DIVERSITY AND INBREEDING COEFFICIENTS WITHIN
THE CANINE GENOME

presented by
JAMES A. BROUILLETTE, MD

has been accepted towards fulfillment
of the requirements for the

Ph.D. degree in Genetics

 

 

Major Professor’ idnatfiFe
/ /3 / H?
I I

Date

MSU is an Affirmative Action/Equal Opportunity Employer

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5/08 KzlProleocaPresICIRCIDateDue.indd

 

IDENTIFICATION OF GENE-SPECIFIC SINGLE NUCLEOTIDE
POLYMORPHISMS WITHIN THE CANINE GENOME AND THEIR USE TO
DETERMINE NUCLEOTIDE DIVERSITY AND INBREEDING COEFFICIENTS
WITHIN THE CANINE GENOME

By

James A. Brouillette, MD

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Genetics

2010

ABSTRACT
IDENTIFICATION OF GENE-SPECIFIC SINGLE NUCLEOTIDE
POLYMORPHISMS WITHIN THE CANINE GENOME AND THEIR USE TO
DETERMINE NUCLEOTIDE DIVERSITY AND IN BREEDING COEFFICIENTS
WITHIN THE CANINE GENOME
By
James A. Brouillette, MD

The domestic dog, Canisfamiliaris, has lived in close relationship with man for
thousands of years, working as a hunter, herder, and loyal companion. Through selective
breeding, the various dog breeds have a high prevalence of various genetic diseases.
Genetic studies are ongoing to elucidate the nature of disease-causing mutations within
the various dog breeds. In the work presented here, I elucidate a method to identify
single nucleotide polymorphisms (SNPs) within canine genes of interest by pooling and
sequencing DNA from across ten breeds of dog. The SNP markers generated are used to
estimate heterozygosity within the canine genome, to demonstrate that the markers

generated by across breed SNP identification will be heterozygous within breeds, and to

estimate the coefficient of inbreeding within three breeds of dog.

DEDICATION

This manuscript is dedicated to my sons Jacob and Andrew, who made as many sacrifices

as I did in order to complete this degree. I love you and appreciate your patience during

the long course of my graduate education.

I also dedicate this work to my wife Tammie, without whom this work would not have

been completed. Thank you for your tireless support.

iii

ACKNOWLEDGEMENTS

I would like to thank my advisors Patrick Venta and Vilma Yuzbasiyan-Gurkan whose
enthusiasm and support made this work possible. Your knowledge, inquisitiveness, and

energy were greatly appreciated.

I would also like to thank Donna Housley, David Entz, Tracy Hammer, Sarah Colombini,
Margo Machen, Tiffy Zachos, Susan Ewart, Martha Mulks, Betty Werner, Cheryl
Swensen, John Kruger, and Simon Peterson-Jones for your many discussions and

problem solving sessions along the way.

Finally, I would like to thank Kathy Lovell for shepherding this manuscript to

completion. This project would not have been completed without you.

TABLE OF CONTENTS

LIST OF TABLES ..................................................................................... vii
LIST OF FIGURES ................................................................................... viii
CHAPTER 1

INTRODUCTION ....................................................................................... 1

Finding Canine Disease Genes3
Evolution ofCanine Genetic MarkersS

The Canine Genetic Map ...................................................................... 6
The SNP Markers .............................................................................. 10
Searching for SNPs in the Human Genome12
SNPs and Gene Mapping .................................................................... 17
Methods of SNP Identification .............................................................. 22
Methods of SNP Detection .................................................................. 26
Canine Breed History ........................................................................ 28
Appendix 1-2 Breed History ............................................................... 31

CHAPTER 2

GENE-SPECIFIC UNIVERSAL MAMMALIAN SEQUENCE TAGGED SITES:

APPLICATION TO THE CANINE GENOME ................................................... 34
Abstract ........................................................................................ 36
Introduction ................................................................................... 37
Materials and Methods ....................................................................... 39
Results ......................................................................................... 44
Discussion ..................................................................................... 48
Acknowledgements .......................................................................... 5 1

CHAPTER 3

ESTIMATE OF NUCLEOTIDE DIVERSITY IN DOGS USING A POOL-AND-

SEQUENCE METHOD .............................................................................. 68
Abstract ........................................................................................ 70
Introduction ................................................................................... 71
Materials and Methods ....................................................................... 73
Results ......................................................................................... 78
Discussion ..................................................................................... 84
Acknowledgements .......................................................................... 88

CHAPTER 4

WITHIN-BREED HETEROZYGOSITY OF CANINE SINGLE NUCLEOTIDE

POLYMORPHISMS IDENTIFIED BY ACROSS-BREED COMPARISON.............101
Summary .................................................................................... 103
Introduction ................................................................................. 1 04
Materials and Methods ..................................................................... 106

Results and Discussion .................................................................... 108

Acknowledgements ........................................................................ I l 1
CHAPTER 5
SUMMARY .......................................................................................... 114
Appendix S-l Brief Explanation of the Wahlund Effect ...................................... 123
BIBLIOGRAPHY ................................................................................... 125

vi

LIST OF TABLES

Appendix 1-1 Summary of physical characteristics of dog

breeds .................................................................................................... 30
Appendix 2-5 Amplification Conditions for Canine UM-STSs ....................................... 58
Appendix 2-6 Summary of Amplification Results for UM-STSs for

Several Mammalian DNAs ............................................................................................... 59
Appendix 2-7 Primer Sets for 11 Universal Mammalian Sequence-Tagged Sites ........... 60
Appendix 2- 8 Eighty- Six Universal Mamammalian Sequence Tagged Sites-

Human Chromosomal Locations and Names... ....62
Appendix 2-9 Eighty-Six Universal Mammalian Sequence Tagged Sites —

Sequences and Sizes .................................................................................... 65
Appendix 3-1 Gene Segments With Amount of DNA Sequenced................................89
Appendix 3-2 Location of SNPs, Types of Base Changes, and Diagnostic

Tests ....................................................................................................... 90
Appendix 3- 3 Allele Counts and Heterozygosity for SNPs Found Among
DogsUsedintheSequencingPool... . ....93

Appendix 3-4 Genotypes for Four SNPs Within Canine Breeds

Appendix 3-5 Summary of SNPs Identified vs. Nucleotide and Amino

Acid Conservation in Loc1

Appendix 3-9 UM-STSS Primers Used for PCR Amplification

and Sequencing .................................................................................

Appendix 3-10 Diagnostic Primers for Testing SNPs

Appendix 4-1 Diagnostic Primers, Restriction Enzymes, and SNP

Locations .......................................................................................

Appendix 4-2 Genotypes for Dogs from Three Breeds Using Four

SNPs .............................................................................................

vii

..94

95

.......... 99

..100

......... 112

......... 113

LIST OF FIGURES
Appendix 2-1 Amplification of Several Canine Genes using UMSTSs. . . . . . . . . . .............52

Appendix 2-2 Lineups of Several Canine Gene Sequences With Homologous
Mammalian Genes ....................................................................................... 53

Appendix 2-3 Sequence of a Portion of the F ES Proto-Oncogene from Several

Mammalian DNAs ..................................................................................... 56
Appendix 2-4 Amplification of a Portion of the F ES Protooncogene From

Several Mammalian DNAs Using UM-STS Primer857
Appendix 3-6 Identification of Four SNPs in the Canine TS Gene by
Pool-And-Sequence......................................................................................96
Appendix 3-7 Identification of Four SNPs in the Canine CFTR Gene. ......................... 97
Appendix 3-8 Location of Four SNPs Identified in the Canine TS Gene. ...................... 98

viii

KEY TO ABBREVIATIONS]

 

Term Abbreviation
Base(s) pairs bp
centiMorgans cM
Deoxyribonucleic acid DNA
Kilobase(s) (pairs) kb
Mega base(s) pairs Mb
Micrograms pg
Microliters ul
Micromolar uM
Milliliters ml
Millimolar li
Nanograms ng
Picograms Pg
Picomoles pmol
Polymerase Chain Reaction PCR
Ribonucleic acid RNA
Units U

1. Abbreviations for specific gene names are found in the text.

ix

Chapter 1

Introduction

Introduction

Domestic dogs and humans have lived in close association for centuries. During that
time, man has selectively bred dogs to the point where today there are over three hundred

different dog breeds in existence worldwide (Ostrander et al., 2000).

Man’s understanding of the genetic diseases that occur in domestic dogs has steadily
grown over the last few decades. As recently as 1979, there were only 13 canine
disorders that were established as congenital or inherited in dogs (Pearson, 1979). That
number has rapidly increased. In 1988, there were more than 200 genetic disorders that
had been identified in domestic dogs, more than 70% of which were inherited as

autosomal recessive disorders (Patterson et al., 1988).

With the widespread use of antibiotics, antihelmintics, vaccination against viral diseases,
and improved and standardized diets, genetic disease has grown in clinical importance
among veterinarians, dog breeders, and dog owners (Patterson, 2000). At latest count,
there are 370 different canine genetic disorders, with 50% of these having breed-specific
aggregations (Patterson, 2000; Ostrander et al., 2000). Of these 370 diseases, 215 (58%)
have clinical and laboratory abnormalities that resemble a human genetic disease and
more than 70% of these are inherited as autosomal recessive traits, X-linked recessive
traits, or have a complex pattern of inheritance (Ostrander et al., 2000). An additional 5-
10 diseases are added to the growing list each year. The price of this genetic disease

manifests itself as pain and suffering for the affected dogs, emotional suffering to owners

and breeders, and an estimated $500 million per year in costs to diagnose and treat
affected animals (Padgett, 1998). Currently, 50 of the 370 canine genetic diseases have

been defined at the molecular level (Giger et al., 2006).

Finding Canine Disease Genes

The disease genes that have been identified to date fall into three different categories.

The first category is those diseases that have an identifiable protein that is absent or not
functioning. The search for the mutation in these cases becomes a matter of searching the
gene encoding the faulty protein for mutations in the coding or control regions. Factor IX
deficiency causing hemophilia B in Cairn Terriers (Evans et al., 1989), von Willebrand
factor deficiency causing von Willebrand’s disease in Scottish Terriers (V enta et al.,
2000), and C3 deficiency causing Complement third component deficiency in Brittany

Spaniels (Ameratunga et al., 1998) are examples of diseases in this category.

A second category of disease genes are those that have been identified by examining
candidate genes for the presence of a mutation. A candidate gene is any gene that is
suspected to harbor the mutation causing the disease phenotype by virtue of its role in a
similar disease in another species or based on some prior knowledge of its biochemical
properties. Mutations in the dystrophin gene causing Duchenne’s muscular dystrophy in
Golden Retrievers (Sharp et al., 1992), the phosphodiesterase 6 beta gene causing rod-

cone dysplasia in Irish Setters (Suber et al., 1993), and the phosphodiesterase alpha gene

causing rod-cone dysplasia in Cardigan Welsh Corgis (Petersen-J ones et al., 1999) fall

into this category.

The final category of genetic diseases contains those diseases that have no obvious
defective protein or candidate genes to test as causative agents. In these cases, purely
genetic analysis can be performed using linkage analysis in families in which the disease
is present or by association analysis in affected individuals. Yuzbasiyan-Gurkan et al.
(1997) were able to identify a marker closely linked to the mutation causing copper
toxicosis in Bedlington Terriers using linkage analysis. The COMMDl gene was
subsequently] cloned, leading to new insights in copper metabolism in mammals and the
discovery of a whole new family of related proteins (van de Sluis et al., 2002; Burstein et
al., 2005). Similarly, Lin et a1. (1999) were able to identify a mutation in the hypocretin
receptor which causes narcolepsy in Doberman Pinschers and Labrador Retrievers using
linkage analysis, opening a whole new field of investigation for sleep and sleep-related
disorders (Zeiter et al., 2006). Ostrander and Kruglyak have demonstrated the feasibility
of whole genome association analysis in dogs using computer modeling (Ostrander and
Kruglyak, 2000). More recent work has moved association mapping in dogs fiom a
theoretical possibility to a future certainty, as all the necessary groundwork has been done
to enable geneticists to undertake association studies in dogs (Kirkness et al., 2003;
Lindblad-Toh et al., 2005; Sutter et al., 2004; Ostrander and Kruglyak, 2000; Clark et al.,
2004). A recent use of association analysis in dogs resulted in the discovery that the

merle coat color is due to a mutation in the SILV gene (Clarke et al., 2006).

Evolution of Canine Genetic Markers

The first genetic markers in canines were identified as protein polymorphisms observed
as variants in electrophoretic mobility. The vast majority of these were structural
proteins or enzymes of the various components of the blood such as isocitrate
dehydrogenase, albumin, and hemoglobin (Meera Khan et al., 1973; Weiden et al., 1974;
Simonsen, 1976). By today’s standards, these markers exhibited little polymorphism and

were limited in scope. These markers were soon replaced by RF LP markers.

During the 19805, several RFLP markers were discovered in genes such as the DLA-D
and DLA-A (Sarmiento and Storb, 1988, 1989). These markers, while more abundant
and polymorphic than the protein polymorphisms, were labor and time intensive and

required relatively large quantities of DNA.

A major leap forward in canine genetics occurred in the 19903. Microsatellite (simple
sequence length polymorphisms or SSLPs) markers, simple sequence motifs of two to six
nucleotides repeated in arrays of varying lengths, were found to be highly abundant
throughout mammalian genomes, highly polymorphic, and rapidly typable using PCR.
They have the advantage of being present at about 20 kb intervals throughout the canine
genome (Yuzbasiyan-Gurkan and Venta, Pers. Comm). The one drawback is that they
are only occasionally found associated with coding regions of the canine genome. Our
collaborator, Dr. Vilma Yuzbasiyan-Gurkan, previously developed several hundred

anonymous SSLP markers throughout the canine genome and used this resource to

identify a marker that is tightly linked to the mutation causing copper toxicosis in
Bedlington Terriers (Yuzbasiyan-Gurkan et al., 1997). Many other SSLP type markers
have been discovered and developed in other labs (Francisco et al., 1996; Ostrander et al.,

1993, 1995).

This work centers on the identification and development of single nucleotide
polymorphisms as genetic markers. At the time this work was undertaken, nothing was
known about the frequency of occurrence of SNPs in the canine genome (Chapter 3).
Based on data from the human genome, it was likely that the SNPs would occur more
frequently in the canine genome than SSLP markers (Collins et al., 1997, Nickerson et
al., 1998). In addition, the introduction of DNA arrays made high throughput genotyping
of SNP markers feasible (Wang et al., 1998; Chee et al., 1996; Landegren et al., 1998, see

below, ), which would allow for more rapid whole-genome scans of SNP markers.
The Canine Genetic Map

The first maps of the canine genome were published in 1997 (Lingaas et al., 1997;
Mellersh et al., 1997; Langston et al., 1997). Lingaas et al. established 16 linkage groups

and assigned a total of 43 markers to those 16 groups.

Using 17 three generation families, Mellersh et al. (1997) assigned 139 microsatellite
markers to 30 linkage groups containing at least one other linked marker (with a lod score

of 3 or greater). The linkage groups ranged in size from 2.3 to 106.1 cM. This map

covered an estimated 884.2 cM of the canine genome with an average marker spacing of
14.03 cM. An additional 11 polymorphic markers were not linked to any other marker.
Of the 150 markers, 47 were dinucleotide repeats, 102 were tetranucleotide repeats, and

one was a hexanucleotide repeat.

In a companion paper, Langston et al. (1997) developed the first canine-rodent somatic
cell hybrid panel. This panel contains a total of 43 microcell hybrid clones that each
display unique canine chromosome retention patterns and three whole cell hybrids that
contained the X chromosome. They assigned 181 microsatellite markers and 27 canine
genes to 31 syntenic groups consisting of two or more markers and/or genes. Many of
these markers were also used by Mellersh et al. (1997). Each of the syntenic groups had
between 2 and 11 markers. Since the canine karyotype consists of 38 pairs of autosomes
plus the X and Y chromosomes (Langston et al., 1997), this does not represent full
coverage of the canine genome. It does, however, represent a substantial portion of the

canine genome.

Priat et al. (1998) followed up these mapping projects with a radiation hybrid map cf the
canine genome. This map contains a total of 400 markers consisting of 2 1 8 gene markers
and 182 microsatellite markers. The map contains 347 markers assigned to 57 groups
with an additional 53 markers being unlinked in the current map. The groups contain
between 2 and 11 linked markers. The radiation hybrid panel consists of 126 cell lines,

and the map is thought to cover about 80% of the canine genome.

The work by Priat et al. began the process of integrating the linkage maps of Lingaas et
al. (1997) and Mellersh et al. (1997) into a radiation hybrid (RH) map. It also indicates

areas of synteny shared by the dog, human, and pig genomes.

The next generation of linkage map was produced by Neff et al. (1999). This map
extends the original map of Mellersh et al. (1997) from 150 microsatellite markers to 276
markers divided into 40 linkage groups. Average marker spacing in this map dropped
from 14 cM to 9.3 cM. This map is estimated to cover 90% of the canine genome (Neff

et al., 1999).

The canine genetic map was integrated into a single map by the assignment of the linkage
groups from the RH map (Priat et al., 1998) and linkage map (Lingaas et al., 1997;
Mellersh et al., 1997; Neff et al., 1999) to specific canine chromosomes using
chromosome painting (Yang et al., 1999). Of the 44 published RH groups and 40
published linkage groups, 39 and 33 groups were assigned to specific chromosomes,
respectively. In addition, Yang et al. were able to align chromosomal regions of the
canine karyotype with syntenic regions of both the human and red fox karyotypes. The
syntenic alignments should enable researchers to identify genes in the human
comparative map that will be candidates for canine genetic diseases by cross-species
comparison, once linkage is established to a canine marker (Yang et al., 1999). A
combined 3MB resolution RH map of CF A1 that incorporated SNP markers was

published in 2004 ( Housley et al., 2004).

Several refinements have been made in the canine genome map since the above reports
were published (Werner et al., 1999; Sargan et al., 2000; Mellersh et al., 2000; Richman
et al., 2001; Lingaas et al., 2001). At the time, the expanded and integrated RH and
linkage map consisted of approximately 800 markers, more than 300 of which were genes
(Mellersh et al., 2000; Lingaas et al., 2001; Sargan et al., 2000). The average spacing
between markers was 9 cM. In addition, each of the synteny groups had been assigned to
specific canine chromosomes (Sargan et al., 2001). One additional refinement was the
characterization of a set of 172 markers for genome-wide screens of the canine genome
(Minimal screening set—1 [MSS-l] Richman et al., 2001). This set of markers, all of
which were microsatellites, were chosen because they provided as complete coverage of
the canine genome as was possible, they were highly informative, and they had been
ordered in linkage groups with high statistical certainty (Richman et al., 2001). It had
been estimated that 42% of the canine genome is within 5 cM of at least one of these
markers, and 77% of the genome was within 10 cM (Richman et al., 2001). While there
were some gaps within the canine genetic map, the map taken in total was thought to be
sufiicient for the whole-genome linkage analysis (Richman et al., 2001; Ostrander and
Kruglyak, 2000). An extended and improved marker set (MSS-2) was published in 2004

(Clark et al., 2004).

By 2004, Breen et al. created an integrated FISH and Radiation hybrid map of the canine
genome (Breen et al., 2004). That map contained a total of 4250 markers, 4100 of which

were assigned to linkage groups and to canine chromosomes. The genes were assigned to

60 different linkage groups that could be assigned to the 38 canine autosomes and two

sex chromosomes (Breen et al., 2004).

In 2005, the entire canine genome was sequenced from a female Boxer (Lindblad-Toh et
a1, 2005). This map represents 7.5-fold redundancy and is thought to cover 99% of the
canine genome. In addition, the sequence data revealed 2.5 million SNP markers within
the canine genome. Commercially available SNP arrays (Lindblad-Toh et al., 2006) have

been developed.

The SNP Markers

The canine genetic markers that have been developed in this series of studies are single
nucleotide polymorphisms, or SNPs. These markers are even more abundant than SSLP
markers and are also amenable to typing by PCR amplification. In addition, it has been
possible to develop SNPs as type I, or gene-associated, markers (Werner et al., 1999;
Sargan et al., 2000). This has the effect of anchoring them in the canine genome and
furthers comparative genetics across mammalian species. Early reports of SNP markers
in the canine genome began appearing in the mid-19903 in such genes as erythroid
aminolevulinate synthase, y-D-crystallin and opsin (Boyer et al., 1995; Shibuya et al.,

1995; Ray et al., 1996).

10

SNPs are defined as single base pair substitutions in genomic DNA at which different
sequence alternatives exist in normal individuals in some population. The allele

frequency of the most common allele must also be 99% or less (Brookes, 1999).

The frequencies of transitions, transversions, and indels are not equal. Two thirds of all
SNPs are C —+ T transitions while the other third is made up of all the other possible
changes (Wang et al., 1998; Brookes, 1999). It has been speculated (Halliday and Grigg,
1993) that the reason for the propensity of C -> T changes is that 3-5% of cytosine
residues in mammalian genomes are presumed to be methylated. These residues can
undergo spontaneous deamination to yield thymine (Halliday and Grigg, 1993). Thus, a
methylated cytosine residue gives rise to a thymidine residue. The result is the
conversion from a C-G base pair to a T-A base pair. An unmethylated C residue that
undergoes deamination will be recognized as a uracil residue, and readily repaired back

to the original C residue.

Before this work, the frequency of occurrence of SNPs in the canine genome was
unknown. It had been established that in humans, if one randomly analyzes two
chromosomes, a SNP will typically be observed to occur once per 1000 bp of DNA
(Brookes, 1999). This means that there is a 0.1% chance that any base will be
heterozygous in a given individual. Within gene coding regions, the fi'equency of
occurrence of a SNP drops to around 1 in 4000 bp, with half of these changes resulting in
non-synonymous changes (Brookes, 1999). These numbers indicated that there would be

several million nucleotide differences between any two individuals and around 100,000

11

differences in their proteomes (Brookes, 1999). This estimate would later be borne out

by sequence analysis of the canine genome (Lindblad-Toh et al., 2005).

Before this work, there had not been any publications on a systematic search for SNPs in
the canine genome. The human genome project had resulted in a huge amount of human
DNA sequence data being available to the scientific community. Several groups have
taken advantage of this resource to locate SNPs in the human genome. Two groups
(Buetow et al., 1999, Picoult-Newberg et al. 1999) searched the expressed sequence tag
(EST) database for SNPs. They scanned the database for multiply sequenced ESTs and
examined them for sequence differences. They then went back to the DNA and

confirmed that the SNPs did indeed exist in the DNA among the population.

Taillon-Miller et al. (1998) used a similar strategy on genomic DNA. They analyzed
overlapping clones of genomic DNA and looked for sequence differences. They then
analyzed sequence differences to see if they represented sequencing errors or whether

they indeed represented SNPs.

Searching for SNPs in the Human Genome

Several large-scale SNP identification projects have been undertaken to establish

estimates of the frequency of occurrence of SNPs and to develop methods of

identification and genotyping of SNPs once they had been located. Wang et al. (1998)

examined 2.3 Mb of DNA from three individuals and a pool of 10 individuals using gel-

12

based sequencing and high-density variation-detection DNA chips. For this study, they
selected 1,139 STS sequences for analysis by both sequencing and DNA chip
hybridization. They found a total of 279 candidate SNPs distributed across 239 of the
STS sequences, yielding a SNP in roughly every 1000 bp of DNA screened. Among the
SNPs identified, the ratio of transitions to transversions was 2:1. In addition, 25% of
changes occurred in CpG dinucleotides even though they made up only 2% of the

sequence surveyed. Almost all of the changes were C ——r T transitions.

This project also involved using DNA arrays to survey STS sequences for SNPs. This
was done by establishing 25-bp oligomers in groups of 4 with position 13 of each
oligomer representing one of the four bases. By knowing the nucleotide occurring at this
position based on the known reference sequence, variability could be detected as a
change in the expected hybridization pattern. They identified 2748 SNPs in this manner,
with a SNP occuring once every 721 nucleotides. Among these SNPs, the mean

heterozygosity was 33% and the mean frequency of the minor allele was 25%.

In addition to the identification of SNPs, Wang et al. used chip hybridization to genotype
individuals for the collection of SNPs they had previously identified. They were easily
able to simultaneously test for 558 SNPs on one chip. They established two tiles for each
SNP, one for each allele. The oligonucleotide arrays again consisted of 25-mers that
were complementary to one of the two alleles at position 13. The individual’s DNA to be
hybridized was synthesized using specific PCR primers with uniform sequence on each

end to allow batch labeling of all PCR products. They were able to perform multiplex

l3

PCR on all 558 loci in a single PCR reaction and make allele determinations for each of 3
individuals tested at 50% of the loci tested. When dividing the loci into 24 sets of 23
primer pairs each, they were able to make allele determinations for all three individuals
tested at 92% of the loci tested. They have thus demonstrated the feasibility of using chip
hybridization to perform large-scale genotyping of hundreds of SNPs simultaneously.
Another group (Lai et al., 1998), examined the region around the human APO B gene for
the presence of SNPs with results similar to those of the other studies listed. However,
they analyzed a contiguous stretch of DNA and confirmed that the development of a
high-density SNP map (with SNP markers spaced every 30 kb) was feasible given current

technology.

Similarly, Cargill et al. (1999) used DNA chip hybridization along with denaturing HPLC
to identify SNPs occuring in the coding regions and adjacent sequences of 106 candidate
genes for caridovascular disease, endocrine disease, and neuropsychiatric disease. They
searched a total of 196.2 kb of DNA and identified 392 cSNPs and an additional 168
SNPs in the adjacent noncoding sequence. They found a SNP at a fiequency of one SNP
per 346 bp in the coding region and one SNP per 354 bp in the noncoding region. They
calculated nucleotide diversity to be 0.0005 in coding regions and 0.00052 in noncoding

regions.

In addition, they were able to examine the cSNPs for occurrence of synonymous vs. non-

synonymous nucleotide changes. They found that roughly half of the cSNPs were of

each type with 207 cSNPs being synonymous and 185 cSNPs being non-synonymous.

14

Since roughly two thirds of all random nucleotide mutations would be expected to alter
the amino acid sequence of the encoded protein, they argue that there is strong selection
against non-conservative DNA mutations. In fact, they calculate that non-synonymous
nucleotide changes survive at only 38% of the rate of synonymous nucleotide changes
(Cargill et al., 1999). Based on their data, they conclude that the average gene contains
approximately 4 SNPs in their coding regions, each of which occur at a fiequency of at
least a few percent in the human population. By extrapolating these data, they would
estimate the number of cSNPs in the human genome to be between 240,000 and 400,000.
More recent estimates of the number of genes in the human genome would push this

number down to between 120,000 and 160,000 (Venter et al., 2001).

In a companion study, Halushka et al. (1999) examined the coding sequences and
adjacent sequences for SNPs in 75 candidate genes for essential hypertension by chip
hybridization and gel-based sequencing. They surveyed a total of 28 Mb of DNA, 190 kb
in 148 alleles. They identified a total of 874 SNPs, of which 387 were cSNPs. The
nucleotide diversity from the data of Halushka et al. are very close to those for Cargill et
a1. (1999), with the nucleotide diversities reported by Halushka et a]. being 0.00045 for

coding regions and 0.00054 for noncoding regions.

In another series of experiments, an area of either 9.7 kb or 24 kb of contiguous DNA
was sequenced around the lipoprotein lipase or angiotensin converting enzyme genes,
respectively (Nickerson et al., 1998; Clark et al., 1998; Rieder et al., 1999). In the first

set of experiments (N ickerson et al., 1998 and Clark et al., 1998), researchers sequenced

15

9.7 kb of DNA within the lipoprotein lipase gene in a total of 71 individuals. The
individuals were Afi'ican-American (24 individuals) Eurpoean (24 individuals) and
European-American (23 individuals). They found a total of 79 SNPs, of which, 47 were
transversions. They also found 9 insertion/deletion variations. There were 7 variable
sites in the coding region, a stretch of 998 bp of DNA, with the remaining 81 variable
sites in the 8,736 bp of noncoding DNA. This gave a nucleotide diversity of 0.002 in the

entire sample and 0.0005 in the coding region.

In the second study (Rieder et al., 1999), the investigators sequenced 24 kb of DNA
around the DCPl gene, which encodes angiotensin converting enzyme. They did this in
six individuals of European descent and 5 individuals of African descent. They
identified a total of 78 varying sites on 22 chromosomes. They found the nucleotide
diversity to be 0.00093 overall. Using a combination of techniques, they were able to

determine that there were 13 distinct haplotypes among the individuals tested.

Taken together, these studies support one another and likely provide a reasonable
estimate for the nucleotide diversity across the human genome. They also validate the

chip hybridization approach as both a method of SNP screening and genotyping.

The work below follows these projects in several respects. A method was developed to
systematically scan coding and noncoding regions of various canine genes for the
presence of SNPs in a pool of ten dogs of different breeds. From these results, an

estimate for nucleotide diversity was calculated for the canine genome (Chapter 3). At

16

the time of this work, a limited amount of canine nucleotide sequence data was available,
and the sequencing data that resulted from the SNP search also represented newly
cataloged sequence data for the canine genome (Genbank accession numbers in Chapter
3). Since that time, the complete nucleotide sequence of the canine genome has become
available (Kirkness et al., 2003, Lindblad-Toh et al., 2005). The sequencing of the canine
genome led to the identification of 2.5 million SNPs within the canine genome (Lindblad-

Toh et al., 2005).

SNPs and Gene Mapping

SNPs are the most abundant form of polymorphism known to exist in the genome. Like
any type of genetic marker, family-based linkage studies can be performed using SNP-
based markers. One disadvantage of SNP markers compared to the more commonly used
SSLP markers, is that the informativeness of the markers is less than that of SSLP
markers, due to the fact that SNP markers are biallelic, whereas SSLPs generally have
several alleles. With only two alleles, the maximum heterozygosity is 0.50. In contrast,
SSLP markers have a heterozygosity that typically ranges from 0.65-0.80 (Kruglylak,
1997). However, the greater abundance of SNP markers easily makes up for this shortfall

in heterozygosity because several can be combined to increase informativeness.

Kruglyak (1997) set out to test the feasibility of performing whole-genome linkage

searches using SNP markers. In his computer modeling, he reached several key

conclusions. First, a map of biallelic markers with a density of 2.25-2.5 times that of a

17

microsatellite map provides comparable information content. Thus, a 4 cM map of

biallelic markers is comparable to a 10 cM map of microsatellites.

Next, the frequencies of the two alleles do not have a great effect on the information
content of the map of biallelics as long as the fiequency of the rare allele is 0.2 or greater.

Thus, perfect “50/ 50” alleles are not required for an informative map.

Finally, the abundance of the SNPs in the genome makes development of large numbers
of markers to create a very dense map of the genome (1 cM or less) theoretically and
technically feasible. In fact, if current estimates hold, there should be on the order of 10
million SNPs in the human genome. While testing such large numbers of markers in
family based linkage studies is technically daunting, methods are being developed to

increase throughput to make such genotypings feasible (Wang et al., 1998; see above).

It has been suggested that one of the true breakthroughs in genetics that SNPs will allow
to come to pass is the mapping of genes conferring risk for complex diseases (Risch and
Merikangas, 1996; Collins et al., 1997; Kruglyak, 1999). Risch and Merikangas
examined the possibility of detecting genes conferring a genome relative risk (GRR)
between 1.5 and 4. (GR is defined as the increased chance that an individual with a
particular genotype has the disease.) They conclude that disease susceptibility alleles
with moderate frequency in the population (p is 0.1 to 0.5) that confer a GRR of 4 or

greater will be detectable by family-based linkage analysis. However, for disease

18

susceptibility loci with GRR of 2 or less, the number of families needed to detect linkage

would exceed 2500 and thus be practically unachievable.

They suggest that association analysis is a much better approach in this case. Instead of
family-based linkage analysis, association analysis would be performed using affected
sib-pairs or single affected individuals and their parents. Association analysis would then
be performed based on inheritance of a given allele or associated marker in affected
individuals as compared to appropriately selected controls. A significant deviation from
random inheritance based on allele frequencies would be suggestive of association
between the marker under consideration and the disease susceptibility allele. Similar
calculations could be performed based on inheritance of a given allele or market fi‘om
unaffected parents to affected offspring. An inheritance of a given allele or marker that
was significantly greater than 50% would be suggestive of association between the

marker and the disease susceptibility allele.

Two approaches to whole-genome association analysis have been suggested (Collins et
al., 1997). The direct method involves characterizing the approximately 25,000 genes in
the human genome to identify SNPs in the coding regions (cSNPs) of these genes. It is‘
assumed that the SNPs resulting in an amino acid change in the encoded protein will be
directly responsible for disease susceptibility. The tests would directly examine these
coding changes for association with disease susceptibility (Collins et al., 1997). In fact,
many investigators have begun identifiying these cSNPs within the human genome

(Picoult-Newburg et al., 1999; Cargill et al., 1999; Halushka et al., 1999).

19

Kruglyak (1999) has done computer modeling to assess the feasibility of whole-genome
association analysis using an indirect approach. The indirect approach would rely on
linkage disequilibrium (LD) between the variable site which confers the disease
susceptibility and tighly linked markers. However, it has not been established what
levels of linkage disequilibrium can be generally expected across the human genome.
Based on his modeling, Kruglyak suggests that useful levels of LD are only on the order
of a few kilobases in the outbred human population. This implies that it would take
500,000 SNPs to undertake whole-genome association studies in outbred human
populations. He also suggests that similar numbers of SNPs would be required in
isolated populations unless the founding population is very small (effective size of 10-
100 unrelated individuals). The assertions of Kruglyak have been controversial. Collins
et al. (1999) examined linkage disequilibria between 1000 pairs of loci and found that LD
was on the order of 300 kb throughhout the human genome. They assert that unlike the
computer models of Kruglyak which simulated the human population as steadily
expanding to its current size, the human population has gone through a series of
expansions and contractions over its existence. The contractions, due to events such as
epidemics, famines, massacres and pressure from technologically more advanced or more
aggressive neighbors would result in greater LD than the model suggested by Kruglyak.
Collins et al. conclude that as few as 30,000 SNP markers, 1 per 100 kb of DNA, may be

sufficient to perform whole-genome association analysis in the human genome.

20

 

More recent work by The International HapMap Consortium (2005) indicates that there is
much more linkage disequilibrium in the human genome than simple modeling studies
would indicate. The HapMap Consortium obtained complete DNA sequences from 269
individuals froom four different human populations, including ten 500kb regions in
which essentially all common DNA variation was determined. This study, in addition to

identifying more than 1 million SNPs, found that

 

Ostrander and Kruglyak (2000) performed computer modeling to evaluate the feasibility
of association analysis in the various dog breeds. They concluded that LD mapping is
practical given the current state of the canine linkage map, with microsatellite markers
spaced an average of 8.86 cM apart (Ostrander and Kruglyak, 2000; Werner et al., 1999).
Indeed, they herald some characteristics of purebred dogs that make them intriguing for
LD mapping. First, gene flow between breeds is limited by the pedigree structure.
(Registration of a dog as a member of a given breed requires that both his parents be
registered members of the same breed.) The modern dog breeds are relatively young,
with most being developed in the last 300 years (Wilcox and Walkowicz, 1995;Wayne
and Ostrander, 1999; Ostrander and Kruglyak, 2000). Many breeds have a small
founding population. Popular sires have decreased the effective population size of the
breeds. Finally, for many breeds, the breed’s natural history has been such that severe
population bottlenecks have occurred in the recent past (Ostrander and Kruglyak, 2000).
All of these factors combine to increase the area of linkage disequilibrium in the various
dog breeds. For example, Ostrander and Kruglyak (2000) performed computer modeling

on the Rottweiler breed. Based on pedigree data provided by the American Kennel Club

21

and breed history (Wilcox and Walkowicz, 1995), they estimated that there will be high
levels of LD extending 5-10 cM around a disease mutation (Ostrander and Kruglyak,
2000). They firrther propose that screening a sample of 40 affected dogs for identity by
descent will be sufficient for gene localization. While the above analysis is specific to
Rottweilers, further modeling indicates that similar areas of LD will exist even in breeds
that haven’t suffered the types of severe population bottlenecks as those of the Rottweiler
(Ostrander and Kruglyak, 2000). Similar results have been demonstrated by Lindblad-

Toh et a1. (2005) for the Boxer and in five different breeds by Sutter et al. (2004).

Methods of SNP Identification

Since the development of RFLP markers (Botstein et al., 1980) it has been known that
there was nucleotide variation within mammalian genomes. When the search for markers
was first undertaken, the only method available was to isolate a cloned gene fragment for
use as a probe and perform restriction digestion with as many different restriction
enzymes as were necessary to locate an RF LP marker. One of the drawbacks of this
method is that even performing endonuclease digestion with all the restriction enzymes
available today, only about 50% of the SNPs would be identified as RFLPs. In fact,
Nickerson et al. (1998) report that if they had performed restriction digestion on their
target DNA with all of the restriction enzymes with either five- or six-base specificities
(Roberts and Macelis, 1997), only 34 of their 88 variable sites would have been

discovered.

22

With the advent of high throughput DNA sequencing techniques, methods of SNP
identification have been developed using DNA sequencing. Direct sequencing has the
advantage of examining all nucleotides in a sequencing run for the presence of SNPs. Its
other advantage is that only DNA sequencing will precisely define both the location and
the exact nature of the DNA variation detected (Kwok et al., 1994). The major
disadvantage has been the high cost associated with sequencing the DNA of several
individuals within a population under study in order to locate variable nucleotide

SCQUCIICCS.

We and others, most notably Kwok’s research group at Washington University, have
developed a method of identifying SNPs by pooling DNA for sequencing (Chapter 3,
Taillon-Miller et al., 1999). This has the advantage of simultaneously surveying several
copies of DNA sequence for the presence of nucleotide variability while reducing the

cost to that of just two sequencing reactions.

One early effort to identify SNPs in the human genome by Kwok et al. sought to utilize
the large overlapping clones already available from the human genome project and
inspect these sequences for nucleotide variability (Taillon-Miller et al., 1998). Where no
nucleotide sequence information is available, one must develop STSs and then sequence
the DNA from several individuals in order to identify SNPs found in that area of the
genome. Kwok et al.. (1 994, 1996) were performing automated sequencing of the DNA

fiom 4 individuals plus a pooled DNA sample for allele frequency estimates. This

23

strategy enabled them to identify with > 85% probability all the SNPs that occurred in the

regions sequenced at a frequency of greater than 20% (Kwok et al., 1994).

Kwok et al. (1996) then applied this technique on a larger scale by scanning a series of
STS markers for the presence of SNPs. They obtained primers for 194 STSs from the
Whitehead Institute’s collection of 838 STSs (as of July, 1994). They were able to
amplify DNAs from 154 of the primer sets in four individuals and a pool of 80
individuals, and examine the amplified DNA for SNPs as given above (Kwok et al.,

1994, 1996). They found 39 SNPs among the 154 STSs tested and estimated that a
polymorphism occurred at a frequency of once per 791 bp, similar to the SNP fi'equencies

reported above.

Taillon-Miller et al., (1997, 1999) further refined this method. First, they used a
complete hydatiform mole (CHM) to serve as a sequencing control (Taillon-Miller et al.,
1997). A CHM is the product of an abnormal conception. It is generally the product of
the union of an enucleated ovum with a single sperm cell that later duplicates its genome
to give a diploid tumor (Taillon-Miller et al., 1997; Grimes, 1984; Kajii and Ohama,
1977). Since the genome of the mole is from a single haploid sperm cell that has
undergone a duplication event, every nucleotide position should be homozygous in the
CHM. This serves as a control reaction in that it allows false positive SNPs resulting
from amplification of duplicated sequences in the genome to be distinguished from true

SNPs. It is estimated that the worldwide incidence of hydatiform moles in humans is one

24

per one thousand pregnancies (Taillon-Miller et al., 1999; Grimes, 1984). Thus, they

argue that sample material should be available for all populations of interest.

With the improvement in dye-labeled dideoxy chain terminators, Taillon-Miller et al.
(1999) now recommend sequencing only two DNA samples in parallel. These are the
CHM DNA as a control and a pool of 80 individuals. They found that they could cut the
number of sequencing reactions by 60%, from five parallel sequencing reactions to just
two and still identify SNPs with the same sensitivity as separately sequencing the four

individuals’ DNA as was done previously.

At the time of this work, several other methods of SNP identification had been
developed. These methods have been reviewed by Kwok and Chen (1998) and are
briefly outlined below. From the time of completion of this research to present time, the
availability of automated sequencing has virtually eliminated the use of these techniques
to identify SNPs. They are included for the purpose of placing the work completed here

in the context of the time it was completed.

SSCP: SSCP is single strand conformational polymorphism. The technique is based on
the fact that single stranded DNA will form a unique tertiary structure based on its DNA
sequence (Kwok and Chen, 1998). Any changes in nucleotide sequence will change the
tertiary structure of the molecule. When these single stranded molecules are
electrophoresed on a native gel, molecules with sufficient differences in conformation

will migrate at different rates and can be distinguished on the gel. The advantage of this

25

technique is its technical simplicity. The disadvantages are that target molecules in
which a polymorphism are to be identified must be smaller than 300 bp for differences in
single nucleotides to sufficiently influence conformation so as to be resolvable on the gel

and the need for multiple buffer conditions to achieve 90% sensitivity.

DGGE: DGGE is denaturing gradient gel electrophoresis. It is based on the fact that
denaturation of double stranded DNA is sequence dependent. A difference in a single
nucleotide between two DNA molecules often causes a great enough difference in their
denaturation temperatures to distinguish between the two molecules. When a partially-
open DNA molecule is migrating through a gel, it is for all intents and purposes
irnmobilzed at the site where one end first denatures. Thus, DNA molecules with
different low-melting domains will have different final positions in the gel. When
heteroduplex DNA is run on a denaturing gradient gel, the heteroduplex DNA will
denature at a concentration of denaturant that is much lower than its homoduplex
counterpart. This forms the basis of the detection procedure. The advantage of this
procedure is its ability, with some modification of the basic procedure above, to locate
polymorphisms in DNA fragments as large as 1000 bp. Its disadvantage is the need to

use specialized equipment to perform the analysis.

Methods of SNP Detection

Once a SNP has been identified, the next step is to develop a means to genotype

individuals at the marker. The “gold standar ” of SNP detection is to use allele-specific

26

restriction digestion and gel electrophoresis to genotype individuals for a given SNP.
This method has the advantage of being highly accurate, technically reliable, and
inexpensive. The disadvantages are that the method is labor intensive and has a low
throughput. Several other methods to speed throughput have been developed (Landegren
et al., 1998). The current methods all use amplification by PCR followed by allele
determination by allele-specific hybridization or allele-specific restriction digestion,
determination of mismatched DNA substrates by polymerases or li gases, or by template

specific incorporation of nucleotides by polymerases.

There is a great deal of overlap between methods of SNP identification and detection.
Certainly, given unlimited budgets, DNA sequencing could be used for SNP detection.
Other methods, such as SSCP, DDGE, and heteroduplex analysis could be used to
determine if polymorphism existed in a given individual. They may even be used to
detennine which alleles were present in an individual, with the inclusion of appropriate

control reactions.

DNA chip hybridization is best suited for high throughput SNP detection. It has the
advantage of being able to genotype an individual at thousands of polymorphic sites
simultaneously (Wang et al., 1998; Chee et al., 1996; Landegren et al., 1998). Its main
disadvantage at the time this work was undertaken was its high cost. Since the
completion of this work, the cost and availability of DNA chip hybridization technology
had decreased and become more reliable, putting it within the budget of most

laboratories.

27

Canine Breed History

In the series of experiments detailed below, DNA from ten different dog breeds were
used to form a working pool of DNA. The breeds making up this pool were chosen
because they differ in size, behavior, and temperment. Presumably, genetic variation
among this pool of DNA will reflect such variation. A summary of breed characteristics
is given in Appendix l-l following this chapter. Breed histories are also included in

Appendix 1-2 at the end of this chapter.

I have published the following papers during the course of this work. Chapter 2 was
previously published as Venta, et al., “Gene-specific Universal Mammalian Sequence
Tagged Sites: Application to the Canine Genome”, Biochemical Genetics 34: 321-341
(1996). In this work, I designed approximately 20% of the primer pairs, performed all of

the DNA amplifications, and performed all of the sequencing reactions within the paper.

Chapter 3 was previously published as Brouillette et al., “Estimate of Nucleotide

Diversity in Dogs with a Pool-and-Sequence Method”, Mammalian Genome 11: 1079-

1086 (2000). In this work, I performed approximately 90% of the experiments.

Chapter 4 was previously published as Brouillette and Venta, “Within-breed

Heterozygosity of Canine Single Nucleotide Polymorphisms Identified by Across-Breed

28

Comparison”, Animal Genetics 33: 464-467 (2002). In this work, I performed all of the

experiments.

In addition, I have coauthored four other papers. They have been previously published as
follows: 1. Brouillette et al., “le I PCR/RFLP Marker in the Canine Connexin 40
Gene”, Animal Genetics 30: 229 (1999), in which I performed about 75% of the
experiments; 2. Brouillette and Venta, “T th 1 PCR/ RFLP Marker in the Canine Rod
Transducin Alpha Gene”, Animal Genetics 31: 68 (2000), in which I performed all of the
experiments; 3. Lingaas et al., “A Canine Linkage Map: 39 Linkage Groups”, J. Animal
Breeding and Genetics 118: 3-19 (2001), in which I performed segregation analysis for 7
of the 222 markers, as part of the DogMap consortium; and 4. Ernst et al., “Mapping of
PBS and FURIN Genes to Porcine Chromosome 7”, Animal Genetics 35: 142-167
(2004), for which I provided PCR primers and amplification conditions for the mapping

of the PBS gene.

29

Appendix 1-1

a
Summary of physical characteristics of dog breeds

 

 

Breed Height in Weight in Coat color Fur Style Classb
inches pounds
Am. Cooker 15 24-28 Black, tan, Silky, long Sporting
Spaniel chocolate,
cream, tricolor
Greyhound 26-28 65-70 Cinnamon, _ Short, smooth Hound
chestnut, red,
black, brindle
Doberman 26-28 66-88 Black, red, Short, smooth Working
Pinscher blue, fawn
Siberian Husky 21-23 45-60 Gray, black, Thick, dense Working
red
Labrador 2 l -24 55-75 Black, Moderately Sporting
Retriever chocolate, short
yellow
Collie 24-26 60-75 Sable and Short, smooth, Herding
white, tricolor, double
blue merle
Scottish Terrier 10-1 1 19-23 Black, brindle, Wiry Terrier
wheat, gray
German 22-26 7 5-95 Black and tan, Short, dense Herding
Shepherd black, sable
Beagle 13-15 55-75 Any color Short, dense, Hound
smooth
Pointer 25-28 55-75 Liver, lemon, Short, dense Sporting
orange, white smooth

a. Information is from The Complete Dog Book, 1997 and Wilcox and Walkowicz, 1995. Height and
weight are for male dogs of each breed. Where differences existed between the references cited, data
was from the first reference above. In all cases, the female dog was slightly shorter and lighter than
the male dog.

b. “Class” refers to the grouping used by the American Kennel Club in The Complete Dog Book, 1997.

30

Appendix 1-2

Breed History
The history of each of the breeds used in this study is outlined briefly below.

American Cocker Spaniel: The American Cocker Spaniel can trace its roots back as far as the 14th century.
In 1368, the Spanyell was first mentioned in the literature. Through the years the spaniel family was
divided into two groups, the land spaniels and the water spaniels. As time passed, the land spaniels were
divided into the smaller cooker spaniels and the larger varieties. Later, the toy spaniels were divided from
the cooker spaniels.

The first registry of the Cocker Spaniel breed was in England in 1892. It was brought to the United States
in the 18805 and went through a change in breed standard such that by the 19305 it came to be considered a
separate breed from the English Cocker Spaniel from which it originated. It is considered to be a sporting
dog, and is reputed to be an excellent hunter. The breed is known for being handsome, happy, eager to
please, trusting, and intelligent. These traits have made it one of the most popular dog breeds in the United
States (American Kennel Club, 1997; Wilcox and Walkowicz, 1995).

Greyhound: The greyhound can trace its lineage back to ancient times. The first known record of the
greyhound dates back to the hieroglyphs of ancient Egypt, around 3000 B.C. The greyhound has long been
a favorite of the aristocracy. Documents from 9th century England indicate that it was a favorite hunting
dog of the Duke of Mercia.

The earliest accounts of the greyhound in America date back to Spanish explorers in the 15005. Known as
hunters, there are reports of greyhounds running down deer, stags, and foxes. Yet, it is probably best
known for its hunting ability for rabbits and hares. Known today for being gentle, well—behaved, and
graceful pets, greyhounds are elegant show dogs and thrilling competitors (American Kennel Club, 1997;
Wilcox and Walkowicz, 1995).

Doberman Pinscher: The origin of this breed is well established. The breed began in Thueringen, Germany
in 1890 by Louis Doberrnann. Doberrnann, a tax collector by trade, needed a dog to protect him from
bandits. The breed mixed the hardiness and intelligence of the German Shepherd, the reaction and fire of
the German Pinscher, and the hunting ability of the pointer. Further outbreeding added the Rottewiler’s
strength, courage, and guarding instinct and the Greyhound’s foot speed. In only ten years, the breed
standard had been established.

The breed is known today for its intelligence, its ability to absorb and retain training, and its loyalty. It is
these qualities that put the breed in demand as a police and military dog American Kennel Club, 1997;
Wilcox and Walkowicz, 1995).

Siberian Husky: The Siberian Husky traces its roots to the dogs of the ancient Chukchi people of
northeastern Asia. The dog was bred to be a sled dog. Its primary mission was as a dog that would travel
great distances at moderate speed while carrying a light load and wouldn’t flinch at the subzero
temperatures of the Arctic region.

The reputation of this breed of sled dog was made in the United States in 1925. A diphtheria epidemic was
sweeping through Nome, Alaska and dogsleds were used to take antiserum from Anchorage to Nome. This
serum run was the forerunner of the famous lditarod dogsled race, and it focused the spotlight on the
Siberian Husky.

The Siberian Husky is naturally friendly and gentle. He is an exceptional family dog, and is still the
favorite of dog mushers across the United States (American Kennel Club, 1997; Wilcox and Walkowicz,
1995).

31

Labrador Retriever: The Labrador Retriever was originally seen in the early 18005 in Canada as a hunting
dog and was particularly useful at retrieving water fowl. The breed was transported to England on fishing
boats and soon became a popular sport dog there as well. Later in that century, the breed was all but
eliminated in Canada, due to a heavy dog tax. The breed’s development into its current form occurred
largely in England.

The breed was first recognized in England in 1903 and in the United States in 1917. This breed is known
for its eagerness to please its master and still possesses its sensibility, even-temper, intelligence, and strong
marking and retrieving skills. The breed is renowned as a bird flusher, companion, drug-detector, and as a
guide dog for the blind. These traits consistently put the breed in the top five in popularity in both England
and the United States (American Kennel Club, 1997; Wilcox and Walkowicz, 1995).

Collie: The breed known as the collie is thought to have its origin in the dogs that were brought to Scotland
with the Roman invaders of 50 BC. These ancient dogs interbred with other Scottish herding dogs to yield
the breed known today. This breed of dogs has been used to herd sheep for centuries in Scotland.

In 1860, Queen Victoria became a fancier of the breed after a trip to Scotland. With her blessing, the breed
became a favorite of the aristocracy and affluent, as well as maintaining its traditional role as a herding
dog. The two types, rough and smooth (referring to the length of the coat), were fixed enough in
characteristics by 1886 that little has changed with regard to the breed standard since that time. By 1877,
the breed had become established in the United States, though the breed was first introduced in this country
with the early settlers over a century earlier.

The breed is consistently popular today as a family pet. It is known for its loyalty and affection and as a
self-appointed guardian of the entire family, but particularly of small children. In recent years, the dog has
maintained its popularity, due in part, to the Lad stories of writer Albert Payson Terhune and the “Lassie”
movies and television series (American Kennel Club, 1997; Wilcox and Walkowicz, 1995).

Scottish Terrier: The Scottish Terrier has been in existence for centuries. There are those that will argue
that descriptions of the Skye Terrier written in the 15705 are not the Skye Tenier that is known today but
the Scottish Tenier of antiquity. At the very least, the modern breed can trace its lineage to Scotland in the
18605. The first standard was established in England in 1880. It has remained the standard, With pillsI
minor changes, up to the present day.

The Scotty was first introduced into the United States in 1883. Since this time, there have been thousands
of Scotties imported. The terrier temperament is taken to the extreme in Scotties. He is alert, quick, and
feisty. These qualities make the breed well suited to being a watchdog and varmint killer. The breed of
dog requires discipline to prevent him from becoming a bully (American Kennel Club, 1997; Wilcox and
Walkowicz, 1995).

German Shepherd Dog: This breed was founded in 1899 by Max von Stephanitz. It has always been a
working dog, originally as a herder, and in a variety of roles today. This breed grew steadily in popularity
around the world up to World War I, but the popularity of the breed suffered due to the anti-German
backlash in Europe and America following World War I.

The breed is known today for its loyalty, courage, and ability to assimilate and retain training for a number
of specific purposes. German Shepherds are often used as guide dogs for the blind, and as police dogs,
military dogs, and as a key component of search-and-rescue units. Considered by some to be aloof, the
German Shepherd Dog doesn’t give affection freely. However, once the dog warms to a person, he is loyal
and dedicated, even to the point of giving his life for his master (American Kennel Club, 1997; Wilcox and
Walkowicz, 1995).

Beagle: The history of the Beagle breed is cloudy. Some reports indicate that the origin of the Beagle dates
as far back as ancient Rome. Other accounts note that the Beagle has been used to hunt hares in Wales for
centuries. Modern records of the Beagle date at least to the middle 17005. Their keen sense of smell and
compact size has made them a favorite to hunt rabbits, hares, and foxes, either individually or in packs.

.32

In the United States, the Beagle has been in existence since colonial times. However, these dogs had the
look of a Basset Hound rather than that of the Beagle of today. Imports of Beagles from the kennels of
Great Britain in the 18805 and 18905 gave rise to the Beagle that is recognized today.

The Beagle is known as a capable hunting dog as well as a playmate for adults and children alike. The
breed’s inquisitiveness and happy-go-lucky nature have made it a consistent member of the top ten dog
breeds in the United States (American Kennel Club, 1997; Wilcox and Walkowicz, 1995).

Pointer: The Pointer breed got its start in England around 1650. These are excellent hunting dogs and were
considered to be the first true pointing dogs. They were originally used to hunt hares. The Pointer was sent
out to locate a hare, at which time, greyhounds were brought in to chase the hare. During the early 17005,
the pointer’s hunting ability was more thoroughly exploited due to the increased popularity of wing-
hunting. Legends abound about this breed’s pointing ability. One example is the story of a sportsman who
lost his dog in the moors of England. He returned a year later to find the skeleton of the dog still pointing
at the skeleton of a bird.

The pointer of today is a hunting specialist. He is muscular, courageous, speedy and has great endurance.

His ability to concentrate on his job and ability to work with people other than his master keep him as a
favorite among hunting dogs (American Kennel Club, 1997; Wilcox and Walkowicz, 1995).

33

Chapter 2
Gene-Specific Universal Mammalian Sequence-Tagged Sites:

Application to the Canine Genome

34

Gene-Specific Universal Mammalian Sequence—Tagged Sites: Application to the Canine

Genome

1 2,3

Patrick J. Venta ’ 1’3

, James A. Brouillette , Vilma Yuzbasiyan-Gurkanz, and
George J. Brewer4
Departments of Microbiology1 and Small Animal Clinical Sciencesz, College of

Veterinary Medicine, and the Genetics Program3, Michigan State University, East

Lansing, MI 48824-1314 and the Department of Human Genetics4, The University of

Michigan Medical School, Ann Arbor, MI 48109-0618

Key Words: Genome mapping; evolution; homology; polymerase chain reaction

Corresponding Author:
Patrick J. Venta, Ph.D.

Phone: 517-432-2515
FAX: 517-432-2514

e-mail: venta@cvm.msu.edu

35

Abstract

We are developing a genetic map of the dog based partly upon markers contained within
known genes. In order to facilitate the development of these markers, we have used PCR
primers designed to conserved regions of genes that have been sequenced in at least two
species. We have refined the method for designing primers to maximize the number that
produce successful amplifications across as many mammalian species as possible. We
report the development of primer sets for eleven loci in detail: CF TR, COL10A1, CSFIR,
CYP1A1, DCNI, FES, GHR. GLBl, PKLR, PVALB, and RBI. We also report an
additional 75 primer sets in the appendices. The PCR products were sequenced to show
that the primers amplify the expected canine genes. These primer sets thus define a class
of gene-specific sequence-tagged sites (STSs). There are a number of uses for these
STSs, including the rapid development of various linkage tools and the rapid testing of
genomic and cDNA libraries for the presence of their corresponding genes. Six of the
eleven gene targets reported in detail have been proposed to serve as “anchored reference
loci” for the development of mammalian genetic maps [O’Brien et al., Nat. Genet. 3:103-
112, 1993]. The primer sets should cover a significant portion of the canine genome for
the development of a linkage map. In order to determine how useful these primer sets
would be for other genome projects, we tested the eleven primer sets on the DNA from
species representing five mammalian orders. Eighty-four percent of the gene-species
combinations amplified successfully. We have named these primer sets “universal
mammalian sequence-tagged sites” (U M-STSs) because they should be useful for many

mammalian genome projects.

36

Introduction

Efforts have intensified in recent years to develop comprehensive genomic maps for
many eukaryotio species using molecular techniques. Many of these efforts have focused
on mammalian species, including human, mouse, rat, ox, sheep, pig, horse, cat, and dog
(e.g., Buchanan et al., 1994; Dietrich et al., 1992; Ellegren et al., 1992; O’Brien, 1986;
Serikawa et al., 1992; Weissbach et al., 1992; WinterO et al., 1991; Barendse et al., 1994;
and the present report). For the non-human species, these projects should lead to more
successful breeding strategies, both for selecting desirable characteristics and for
removing genes that lead to various genetic diseases. Comparisons made between these
genome maps should also lead to new insights on the mechanisms of chromosomal

evolution (e.g., see O’Brien et al., 1993).

We are developing a comprehensive map of the canine genome, with our ultimate aim
being to reduce the incidence of canine genetic diseases. In addition to developing
random, highly polymorphic genetic markers (Type 2 markers; Yuzbasiyan-Gurkan et al.,
submitted), we are also developing markers for specific genes (Type 1 markers). An
appropriate mix of these two types of markers should maximize our ability to map

disease genes.

The traditional method for developing gene-specific markers, Southern blotting and

cross-species hybridization, is very time consuming, labor intensive and limited in

flexibility. This method has been the mainstay for developing gene-specific markers in

37

most animal genome projects. There is a need to develop more efficient methods. This is
particularly important for animal genome projects where scientific resources are more
limited. One method that has excellent potential is cross-species polymerase chain
reaction (PCR). This method has been used successfully for the study of a number of
individual genes but has not been applied on a genome-wide basis for the purpose of map
development. To study a single gene, the cost associated with the failure of a few
primers sets to amplify the correct target is negligible and new primer sets can be easily
redesigned and synthesized. However, when primer sets are being designed for many
genes, the cost for failed primers can become substantial, both in terms of time and other

resources, so we have refined the design method to minimize this problem.

We describe here, in detail, eleven primer sets that can amplify gene-specific targets of
dogs and other mammalian species. Seventy-five additional primer sets are listed in the
Appendices. Because markers based on PCR primers are called sequence-tagged sites
(STSs; Olsen et al., 1989), we call these primer sets universal mammalian STSs (UM-

STSs) because they should be useful for many mammalian genome projects.

38

Materials and methods

DNA Isolation

DNA from dog, human, pigtail macaque, horse, pig, rat and mouse were isolated from
various tissues by standard phenol-chloroform extraction methods (Sambrook et al.,
1989). Goat DNA was kindly supplied by Dr. Karen F riderici, Michigan State

University. DNA was purified by standard methods from a canine liver cDNA library

(Clontech) and from a canine genomic DNA library (Clontech) after growing 1 x 106
phage in E. coli strain LE392 (Murray et al., 1977) in liquid culture (Sambrook et al.,

1989)

Design of PCR Primers

The method of primer design detailed here was used throughout this series of
experiments for primer design, unless the primer was designed based on available
sequence data for the species being studied. Primers were designed to genes where the
intron-exon structure was known in at least one species and where the nucleotide
sequence was known in at least two species (the “index species”) that are not closely
related. Tandemly duplicated genes known to have undergone gene conversion in any
species were avoided. Primers were generally designed so that the amplified product
contained an intron. Since the canine sequence was unknown in most cases, the sizes of

the canine introns were not known prior to amplification. We have since determined that

39

the vast majority of canine introns will be between 50% and 150% of the size of the
corresponding human intron (V enta, unpublished observations). We have followed the
human gene nomenclature system (ISGN, 1987) for naming the canine genes. The
eleven loci described in detail in this chapter, and their protein products, are: CF_"LR,
cystic fibrosis transmembrane regulator; COL10A1, type X collagen, alpha 1 chain;

CSFIR colony stimulating factor 1 receptor; CYP1A1 cytochrome P-450 1, alpha 1;

 

 

D_CN1, decorin; EE_S, c-fes (feline sarcoma) proto-oncogene; GHR, grth hormone
receptor; Glam, beta galactosidase; LKfl, pyruvate kinase - liver, RBC form; w,
parvalbumin; and RA], retinoblastoma protein. The Genbank Accession numbers or
reference for the sequence of the two index species for each locus are as follows: CF__TR,
M55129, M60493; COLIOAl, X65120, X65121; $13, X14720, K01643; CYP1A1,

Uchida et al., 1990, X04300; DCNl, L01125, Z12298; FES, X06292, J02088; GHR,

 

211802, J0481]; GLBl, 859584, M57734; PVALB, X63578, M15452; PKLR, $59798,

M17088; and RB], L11910, M26391.

Primers were designed to highly conserved nucleotide sequences contained within coding
regions. It is presumed that, in the absence of parallel evolution, regions that are
conserved among distantly related mammalian species represent the nucleotide sequence
of the most recent common ancestor of the two modern mammals. Thus, any mismatches
to the primers should result only from evolution of the canine nucleotide sequence since

the divergence from the most recent common ancestor.

40

Within the areas that were conserved among the two index species, an attempt was made
to place the primers in the nucleotide sequence corresponding to the least mutable amino
acids (Collins and Jukes, 1994; Jones et al., 1992). While there are slight differences
among studies, there is a consensus that Gly, Phe, Tyr, Trp, and Cys are among the least

mutable amino acids (Collins and Jukes, 1994; Jones et al., 1992; Dayhoff et al., 1978).

In addition, an attempt was made to choose primers overlying codons with the fewest
degenerate codon positions (Li and Grauer, 1987). With only a single codon, Met and
Trp are excellent amino acids over which to design a primer. Others that are both rarely

mutable and have few codons are Phe, Cys, Lys, and Gln.

The final step was to attempt to place the 3' end of the primer in the second position of
the codon for all codons except glycine, which has greater conservation at the first

position rather than the second position (Venta, unpublished observations). An attempt
was made to use general principles such as the avoidance of primer-dimer formation as

well.

Conservation of amino acids within multigene families was also taken into account, when
possible. Where unavoidable nucleotide mismatches occurred between the two index
species, the primer sequence was designed to exactly match one of the two which we then
call the “primary” index species. GC-rich genes were generally avoided due to the
amplification difficulties that can occur, even with exactly matching primers. Primers

were twenty bp in length on average. Each primer in a pair was adjusted to be of

41

approximately the same annealing temperature (Breslauer et al., 1986). All fits of primer
pairs were designed to have approximately the same annealing temperature as well, in
anticipation of performing multiplex amplifications. It was not always possible to follow
every rule for every gene, given the actual circumstances; however, the majority of the
rules were generally applicable. Primers were synthesized by either the Michigan State
University Macromolecular Structure Facility or the University of Michigan DNA

Synthesis Facility.

PCR Amplifications

Correct design and syntheses of the primers were examined by amplifying the DNA from
the primary index species. Standard buffer, nucleotide, and primer concentrations were

50 mM Tris-HCI (pH 8.3 at room temperature), 50 mM KCl, 1.5 mM MgClz, 200 pM

dNTPs, 0.1 pg of each primer, and 0.5 - 1.0 pg of target DNA in a 25 pl reaction.
Reactions were routinely boiled for three min prior to the addition of 2.0 U of Taq DNA
polymerase. Optimal cycling conditions for the amplification of canine genomic DNA
were usually found by testing one of several sets of conditions in general use in the lab
(see Appendix 2-5). Occasionally it was necessary to use “hot-start” conditions (Bassam
and Caetano-Anolles, 1993) in order to get stronger, cleaner amplifications. The
presence of an amplification product was determined by electrophoresis of a portion of
the reaction on a 1% agarose TBE gel (TBE = 90 mM Tris, pH 8.3, 90 mM sodium

borate, 2.5 mM EDTA) followed by staining with ethidium bromide.

42

DNA Sequence Analysis

The identity of each amplified canine gene was confirmed by “single pass” direct
sequencing of PCR products using Sequenase or Taq cycle sequencing kits (United States
Biochemical Corp., Cleveland). The PCR products were gel purified with Qiaex (Qiagen
Corp., Chatsworth, CA) or by elution from polyaorylamide gel slices (Bergenhem et al.,
1992) prior to their use in the sequencing reactions. The canine sequences were visually
aligned with the sequences of the other species used to design the PCR primers in order

to verify the degree of sequence identity.

43

Results

The primer sets for the various UM-STSs reported here are given in Appendix 2-7 and
efficient amplification conditions for the canine genes are given in Appendix 2-5. It is
probable that these conditions could be optimized further (e.g., reduction in the time in
each cycle). However, the conditions reported here were found to work effectively while
minimizing the number of conditions that had to be examined. A representative gel
showing amplification of the canine target DNA along with the human target DNA is
shown in Appendix 2-1. The human target serves as a positive control for the
amplification system because these primers were designed to exactly match the human
sequence. The ability to quickly screen genomic and cDNA libraries for the presence of
sequences is also demonstrated in Appendix 2-1. The genomic clones for GHR,
COLIOAI , and m (a very faint signal, stronger on other gels [data not shown]) are
present in this particular canine genomic library. The presence of a decorin cDNA clone
(encoded by the M locus) in the canine liver cDNA library is shown by the presence
of the 122 bp band; cDNA clones for GE and COL10A1 are not present. The Q9111
PCR product from the cDNA library was sequenced and its identity confirmed (see
Appendix 2-2). The human and canine genomic bands have different sizes for Gfl and
9% because of the intron size differences. The size for the COLl 0A1 PCR product is
the same between the species because an intron was not spanned for this is the UM-STS.
Although the PCR product bands in Appendix 2-1 are unique, a few UM-STS-species
combinations sometimes contained one to several non-specific amplification products.

This is a minor problem with unique sequence primers, because it is almost always

44

possible to deduce the correct band based upon staining intensity, and on the similarity in

size compared to the band of the primary index species.

The amplified products for all of the canine loci were sequenced to confirm their identity
and the results are shown in Appendix 2-2. The degree of identity between the canine
and index species sequences for each locus is within the range generally accepted
(roughly 70 to 100%) as demonstrating homology between the genes of mammalian
species (Li and Grauer, 1987). These results support the hypotheses that the canine PCR
products are homologous to the respective index species’ genes. The canine COL10A1
sequence matched the human and mouse sequences to a similar extent (data not shown).
The sequences for _PK_LR and CYP1A1 exactly matched previously published canine
coding sequences (Whitney etal., 1994; Uchida et al., 1990); the sequence for canine
EES is given in Appendix 2-3. Although the majority of the canine sequence for P_V_A_L_B_
is from an intron, we believe the degree of sequence identity from this region is sufficient
evidence to confirm that the PCR product is from the correct canine locus. As expected,
the canine sequences tend to show greater identity with the human sequences than with
the rodent sequences because of the faster evolutionary rate of the rodent genome (Gu
and Li, 1993). A microsatellite repeat was found. within the amplified product itself for

RBI. Preliminary results show that the RBI repeat, (GA)12(avg)a has moderate genetic

variability within several canine breeds.

We hypothesized that each primer set should work for many different mammals, given

the evolutionary rate at which nucleotide substitutions occur (Li and Graur, 1987) and the

45

number of primer nucleotide mismatches that can be tolerated by PCR. We tested the
‘universal’ utility of these primers on the DNAs from mammals representing several
different orders. We used the same reaction conditions that were found to amplify the
canine sequences. We have termed these reactions “Zoo PCRs.” Appendix 2-4 shows a
representative experiment. The F_ES_ proto-oncogene was amplified from all of the DNAs
examined. These PCR products were purified and sequenced directly without subcloning
(see Methods and Materials). The sequences are tabulated in Appendix 2-3. The degree
of sequence identity makes it highly likely that the canine PCR products are all
homologous to the corresponding index species’ genes. The pattern of nucleotide
interchange is also what would be expected for homologous genes; members of the same
mammalian order share more sequence similarity with one another than with those of

other orders.

The data for the Zoo PCRs for the other UM-STS primer sets reported in this paper are
given in Appendix 2-6. Greater than eighty-four percent of the targets, excluding the
index and canine species, amplified under the single condition used to amplify the canine
sequence. These species represent five different mammalian orders; primates (human
and macaque), carnivores (dog), artiodactyls (goat and pig), perissodactyls (horse), and
rodents (mouse and rat). Limited experiments on other members of these orders (e.g., cat
and ox) produced similar results (data not shown). Lack of amplification for D_CN__1_ for
one of the artiodactyls (goat) would be predicted because there are four mismatches
between the UM-STS primers and the sequence of the closely related bovine m (Day

et al., 1987). We have found it difficult (although not impossible) to amplify DNA using

46

primers that contain more than two mismatches with the target, when using 20-mers
(P.V., unpublished results). It is likely that the homologous gene from at least some of
the non-amplifying species would appear using these primer sets if other PCR conditions

were examined.

47

Discussion

This study has shown the feasibility of generating a series of UM-STSs, useful for studies
of many genomes, and addressed methodological considerations for their development.
UM-STSs should serve as useful tools both for amplifying regions of interest from
genomes as well as for isolation of clones from genomic and cDNA libraries and for
cross-species comparisons. The data reported in this paper indicate that approximately
85% of all carefully designed UM-STSs will be useful for any given mammalian species.
We believe that this method is far more efficient, less costly and considerably less labor
intensive than traditional hybridization and Southern blotting-based methods. An
additional important benefit is that the information for the necessary reagents (i.e., the
primer sequences) is transmitted much more easily and quickly than the clones that are

necessary for Southern blotting.

UM-STSs will also be useful for developing genetic markers within various genomes.

We have found a microsatellite within one of the eleven loci reported here (Q) and
have found other microsatellite repeats associated with genomic clones isolated through
the use of UM-STSs (unpublished results). Single site variability should also be found
directly in at least some of the amplified products by using one of a number of techniques
developed for scanning for variability, such as the single-strand conformation
polymorphism technique. For example, this method has been used to find two
polymorphic sites in a study of the canine A_L_A_§2 gene in a PCR product of a size similar

to those reported here (Boyer et al., 1995). If the frequency of single site polymorphic

48

variability for other mammals is as high as that estimated for humans (roughly one in 200
to 400 nucleotides), then a significant portion of UM-STSs will have these sites. We are
currently screening for this variability in the canine genome to estimate the frequency of
such variation in the dog. It will be necessary to screen each species individually for
genetic variability. However, the availability of previously designed UM-STS primer
sets, such as those reported here, should make this work proceed more rapidly compared

to the traditional method.

An example of the utility of cross-species comparisons is given by the case of
Waardenburg syndrome. The clue to the location of one of the human Waardenburg
syndrome genes--well known for causing a syndromic hearing loss--was first gleaned
from comparative mapping with the mouse (Asher and Friedman, 1990). The map
locations in the mouse suggested possible locations of the human disease gene, one of
which eventually was proven correct (e.g., Morell et al., 1992). Because the identity of
the gene in the mouse was not known at the time, this approach might more properly be
called a 'positional candidate' approach. UM-STSs will be useful for rapidly producing
mammalian genetic maps so that the positional candidate approach can be applied to

more species.

Very little is known about location of genes within the canine genome. Indeed, except
for genes located on the X-chromosome (Meera-Khan, 1984; Deschenes et al., 1994) and
a few small unassigned linkage groups (Meera-Khan, 1984), the rest of the genome has

remained unexplored. New linkage groups are being developed by us and others

49

(Holmes et al., 1992; Ostrander et al., 1993; Rothuizen et al., 1994; Yuzbasiyan-Gurkan
et al., submitted) based primarily on simple sequence repeats. The development of UM-
STSs should help to rapidly identify the location of linkage groups on specific canine
chromosomes. The identification of conserved syntenies will allow candidate linkages to
be tested in the canine genome. The assignment of the proposed anchor loci (O’Brien et
al., 1993) as defined by UM-STSs to specific chromosomes can be accomplished by the
somatic cell hybrid, flow sorted chromosome, and fluorescent in situ hybridization
(FISH) methodologies. Other methods, such as assignment by use of linkage to
previously mapped loci, are also possible. We have already assigned several genes by
FISH to canine chromosomes using cosmids isolated with UM-STSs (F ujita et al., in
press). Using the methods described in this paper, we have developed a much greater
number of UM-STSs that should cover, for linkage mapping purposes, a substantial

portion of the canine and other mammalian genomes (see Appendices 2-8 and 2-9).

50

Acknowledgements

We thank Ya Shiou Yu and Murat Gurkan for their valuable technical assistance. We
also thank Tracy Hammer, Neal Dittmer, Jessica Nadler, Elizabeth Tullett, Marc
Crotteau, and Kristen Penner for their contributions to the development of a number of
the primer sets. This work was supported by the American Kennel Club, the Orthopedic
Foundation for Animals, and the Morris Animal Foundation. We also thank the
Washington Regional Primate Facility for supplying the pigtail macaque tissues to Dr.

Richard E. Tashian, with whom one of us (P.V.) originally isolated the DNA.

51

Appendix 2-1

123456789101112M

Size
lkbl
r —2.32

 

Amplification of several canine gene segments using UM—STSs. The following lanes were amplified with
the gene-specific primer sets (see table 1): lanes l-4, GHR; lanes 5-8, COLIOAI; and lanes 9-12, DCNI.
Lane 13 contains a mixture of DNA size markers; 1 bacteriophage cut with the restriction endonuclease
BstEII and the plasmid pSK- (Stratagene) cut with Mspl. Lanes 1, 5, and 9 contain PCR products
amplified from human genomic DNA. Lanes 2, 6, and 10 contain PCR products amplified from canine
genomic DNA. Lanes 3, 7, and l 1 contain PCR product amplified fi'om DNA purified fi'om a canine
genomic library contained in a it bacteriophage vector. Lanes 4, 8, and 12 contain PCR products amplified
from a canine liver cDNA library.

52

Appendix 2-2

CFTR
A.A. 1346
l intron 22
Dog - - - - - - -| I - - - -
Mouse - - - - - — — -| I - - - V - — - V — — - —
Human E P S A H L D PI V T Y Q I I R R T L K Q A
F A
Human
GAACCCAGTGCTCATTTGGATCClAGTAACATACCAAATAATTAGAAGAACTCTAAAACAAGCAT
TTGCT
Mouse ..Gm.....CmC.A..C..ICAm........G.CmC..C..GTm.........C..Cm
Dog G .................... |.A- ............ C

COLIOAl

Dog — - - — - - - - K - - - - — — — - - - - -
- H

Mouse - I Y E — — — - — - - — — — - — S - - — -
- K

Human P F D K I L Y N R Q Q H Y D P R T G I F T
C Q

Human
CCATTTGATAAAATTTTGTATAACAGGCAACAGCATTATGACCCAAGGACTGGAATCTTTACTTGTCAG
Mouse ..CA..T..G.G C .C..Tm..Gm.....Cm.....ATm.Tm.....CmA..

Dog m........G..Cm.......A ..................... Am........C..C..C..C
dog
coatttgataagatcttgtataacaagcaacagcattatgacccaagaactggaatcttcacctgccag

CSFIR

intron 3 |
Dog|-—v———Q--———-——V—G———-

FeLV l- - A — — — Q — — - — — — T _ L _ G _ _ _ _

Human ID P A R P W N V L A Q E V V V F E D Q D A L
L

Human
lACCCTGCCCGGCCCTGGAACGTGCTAGCACAGGAGGTGGTCGTGTTCGAGGACCAGGACGCACT
ACTGC
FeLV lm....Tm..T ..G ..G..Cm..AMACGM..G..A.GTW..T..GT.Gm.
Dog Im...TTm..Tm..Gm..G..Gm.....Cm...Gm...GGm..T..G..Gm.
dog

accctgttcggccttggaaggtgctggcgcaggaggtcgtcgtggtcgaggggcaggatgcgctgctgc

53

Appendix 2-2 (cont’d).

DCNl

| intron 6
Dog - — - — - - |- — — - — - - -

Human V D A A S L K G L N N L A [K L G L S F N S

 

I 8
Human

GTTGATGCAGCTAGCCTGAAAGGACTGAATAATTTGGCTAlAGTTGGGATTGAGTTTCAACAGCA
TCTCT
Rat m........Cm.........A..TCm....Tml..Cm.Tm..Cm..Tm...A.C
Dog ................... lm.....Cm....Tm........N
dog ggactgaataatttggcta

agttgggactgagttttaacagcatctc

GHR
A A. 333
l
Dog - D L - - - - - G - - - - — — — — N _ _ _
Rat - D A - - — - — — — — — — — — — _ D _ Q _

Human D E P D E K T E E S D T D R L L S S D H E

K S
Human

TGATGAGCCAGATGAAAAGACTGAGGAATCAGACACAGACAGACTTCTAAGCAGTGACCATGAGA
AATCA
Rat m...TG.G ..G .....A..G .....C ............... GAM...Gm......
Dog m...C.Tm.........C..A.G .......................... AC ...............
dog

tgatgacctagatgaaaagaccgaaggatcagacacagacagacttctaagcaacgaccatgagaaatca

GLBl
A.A. 268
|
Dog — — — — — — — — - v - — — v -
Mouse — - — - - - - — - — — K - — — — v — - K T L
— T

Human E F Y T G W L D H W G Q P H S T I K T E A V
A S

 

Human

GAATTCTATACTGGCTGGCTAGATCACTGGGGCCAACCTCACTCCACAATCAAGACCGAAGCAGT
GGCTTCC
Mouse ..G .................... C .....TAm.C..T ..GG.G..A..TA..A..C ..A..
Dog m..T .....G..A ..A 6.6 ..T .TC ...
dogglbl

gatcattggggccagccacactcaacagtgaagactgaagtcgtggcttcc

54

Appendix 2-2 (cont’d).

PVALB
A.A. 59
I
Dog — intron 2 — -
Rat — — — — - — - (bp) S -
Human I E E D E L G F I
L K
Human ATCGAGGAGGATGAGCTGGthaagctggagg — 1300 -
tttctcctccagATTCATCCTAAAAG
Rat ..T ..................... a. — 1500 — m....- .G.C T..G..G.
Dog m....agactcc. - 1300 - m....-m........
dogpvalb ggggtaaagactccg tttctcc
ccagattcatcc
R81
A.A. 890
|
intron 22
Dog A - - - - - - - - - - L - - - - - - - - - -
--l
Mouse - G - - - - - - - N V — — - - — - A - - - -

Human G S N P P K P L K K L R F D I E G S D E A D

GGAAGCAACCCTCCTAAACCACTGAAAAAACTACGCTTTGATATTGAAGGATCAGATGAAGCAGA
TGGAAGI

Mouse ..CG ....c..c .............. CG.G .....C..C..GmG.C .............. G..I
Dog .c ................... Tm.........TGm.....C .......................... I
dogrbl

gcaagcaaccctcctaaaccattgaaaaaactactgtttgatatcgaaggatcagatgaagcagatggaag

Lineups of several canine gene sequences with homologous mammalian genes. The nucleotide and amino
acid sequences are compared for each of several anchor loci between dog and two other species. The
locations of PCR primers are underlined, although not all PCR primer sites are shown. Some of the lineups
show intron sequence whereas others simply identify the location of the introns. Genbank accession
numbers for canine sequences are as follows: CFTR, L77683 and L77689;COL10A1, L77672; CSFI R,
L77670; DCNl, L77648; GHR, L77673; GLBl, L7767l; PVALB, L77685 and L77686; and RBI, L77669.

55

Appendix 2-3

Sequence of a portion of the PBS proto-oncogene from several mammalian DNAs. Sequences are from
exon 15 and intron 15. Notations for the sequence lineups are as follows: HUM, human; MAC, macaque;
CAT, domestic cat; F ES, feline sarcoma virus; DOG, dog; COW, ox; GOA, goat; HOR, horse; PIG, pig;
RAT, rat; and MOU, mouse. The upper two lines for each block of text represent amino acid sequences
and the lower lines represent nucleotide sequences. Dots indicate nucleotides in the various species that are
identical to those of the human sequence. The human and cat sequences determined here exactly match the
published sequences (Alcalay et al, 1990p; Roebroek et al, 1987). The feline sarcoma virus sequence was
not determined in this study but is included for comparative purposes. Only a single amino acid
interchange was found among these sequences; isoleucine (I) for macaque, cat and feline sarcoma virus and
leucine (L) in all others. Sequence alignments for the intron were done visually and may not be optimal.
Genbank accession numbers for these sequences are as follows: AMACFES, L77678; DOGFES, L77674,
CATFES, L77675, COWFES, L77677; GOAFES, L77681; PIGFES, L77679; HORFES, L77676;
RATFES, L77680, and MOUFES, L77682.

MAC,CAT,FES I

HUM A D N T L V A V K S C R E T L P P D L K
HUM GCCGACAACACCCTGGTGGCGGTGAAGTCTTGTAGAGAGACGCTCCCACCTGACCTCAAG

MAC m.........Tm....A ................................. Am..
CAT m.........Tm....Cm..Am...C.Cm..A. ........... Am..

FES m........Tm.....Cm..Am...C.Cm..Am..Am...Am..

DOG m.....T..T .............. Am..CCm....C ..................
COW ..A ....................... Am...C.Cm..A ..................
GOA ..A ....................... Am...Cm....A..G..Cm.........
PIG ..A..T .................... Am..CCm.A .....................
HOR ..T ....................... Am..CCm..........C..Gm......
RAT ..Am.......Cm...Tm.........C ................... NW...
MOU ................. Tm.........Cm...NNN .................

HUM A K F L Q E A R
HUM GCCAAGTTTCTACAGGAAGCGAG GTGGGTGATAAACTAATGATCACCACGGGTCCCGCAT

DOG ....................... m....Cm.GmCm--..CA.A.CT..Am

CAT T ..A.A ..Am.ACmAG..Cm--..CATAA.TW..C

FES m........Tm.....A.A

COW m........Cm........ m.......G.AC.CCCMA.TGTA..C CATA
GOA G .A.. m.......G.AC.CCmA.TGTA..CmT.C.C
PIG m........Gm........ m.........AG..CCM.TGTGATAAAAGA.CC
HOR ................. G..A.. m.....CmAmCCm.TGGTAT.CTAA.G..
RAT m........G..NNNNm.. m...Cm..A.GGGA.CAGT..A..T TTGTG
MOU .................... A.. m.........Am.AT

56

Appendix 2-4

Size
1ka .
1.5- ”‘9

   
 

0.6- - "

0.1-

Amplification of a portion of the PBS protooncogene fi’om several mammalian DNAs using UM-STS
primers. Target DNAs for each lane are as follows: 1, human; 2, pigtailed macaque; 3, dog; 4, goat; 5, pig;
6, horse; 7, mouse; and 8, rat. The mouse DNA here was degraded; strong amplification was with another
lot (sequence shown in Appendix 2-3). The DNA marker lane, (M) contains a 100-bp ladder.

57

Appendix 2-5

Amplification Conditions for Canine UM—STSs

 

 

Size of PCR Product (bp)

Locus Temperatures (C) Times (min) Human Dog
CFTR 95, 57, 72 0.5, 1.5, 4 700 1000
COLIOAI 94, 57, 72 (hs)a 1, 2, 3 384 384
CSFIR 94, 59, 72 1, 2, 3 730 730
CYP1A1 95, 57, 72 0.5, 1.5, 4 700 600
DCNI 94, 57, 72 1, 2, 3 1422 2000
PBS 94, 57, 72 0.5, 1, 1.5 484 500
GHR 94, 57, 72 1, 2, 3 765 800
GLBI 94, 57, 72 1, 2, 3 238 240
PKLR 94, 59, 72 l, 2, 3 600 630
PVALB 94, 57, 72 (hs) 0.5, 1.5, 4 1400 1300
RB] 94, 59, 72 l 2 3 695 1300

O
9

 

a. hs indicates “hot start” used.

58

Appendix 2-6

a
Summary of Amplification Results for UM-STSs for Several Mammalian DNAs
Locus Human Macaque Dog Goat Pig Horse Mouse Rat

 

 

CFTR +1) + + - + + + +
COLIOAI + + + - + + + +
CSFI R + + + + - - + +
CYP1A1 + + + + + + + +
DCNI + + + — + + + +
FES + + + + + + + +
GHR + + + + + + + +
GLBI + + + + + + + +
PKLR + + + + + - + +
PVALB + + + + + + - -
RBI + + + - + + + +

 

a. +, Amplification; -, no amplification.
b. Boldface symbols indicate index species.

 

§§§i****§§§§§§***§**

§§**i**§*§§**§§**§§*

and

 

38AM Dva
33 m2 Nb Jaw—Mm: Imus—mm <HOH<<UHOUHOUUUO<OH< <<UH<OH<O<OO<<UHUUOU 58:2 MAMA—
cacaaaaaaaaaaaaaasaa: aaaaaaaaaaaaaaaaaO: 3:02
_Na :38 SE Ho
...—Sam 2 m won -mAOI -mAOI Ot<<<<UOO<H<<UUt<U FUOOHUDOHU<H<HUE<<O 583E :30
eaaacacaaaeaaaaaaaa< :E:::::<a:: and
Na 23a can.
.. _ .m Em va _om -MIOI -MEOI <UOO<<UPOOEUEUE<OH H<O<<<UUHHO<UUEO<UU 55:: MID
aaaaacaeaaaaaaaaaa: aaaaaaaa<acaaaaaaaa= >wom
.55 DQXm Owe—Km ,
-mNUm— 3% mm -mmn—I -mmmI OOOH<O<HOH<OU<OH<UUF ECHO<<OUOOEU<<OOOO 525.: mm“—
aaaaaaaaaaaaaaOaaa: acacaacaoaaaaaaaaaaa and
me a o
.. _ NUS 54m bON Sam _ ZOO: cXN _ZUQI OH<O<UHUUUHUO<<OPO<< <<OHOUO<HUO<UOH<OZO 52:3.— _ ZOO
acaaaaaaeaacaaaaca: aaaaaaaaaaaaaaaaaa: 58:3
Beam Qme
_ m3. N. _ v a _ m - _ < _ FWUQ .. _ < _ m>UD :OOHU<UUOHUH<OEOOH OOPUO<OOEUHUU<OOE on _< _ m>U
cacaaaaaaeaaaaaaaO. aaasaacaiaaaaaaaa: >mom
was 2me omxm .
.mme _E on -M Emu: -m _mmUI <OO<<O<OUOOO<UUOH<U <HUU<OOOOU<U<<<<UUE 383$ m Emu
aaaaaaaaaaaaaaaaea: aaaaaaaaaaahaaaaaa: 3:32
mac :83 BE _
A NU© moo mow - _ <o _ 400: .. _ <0 _ AOUI <<O<OHUUH<<OO<HU<UUO UUU<HHUO<<<UUHUHUE< 52:3.— <o _ AOU
aaaaaaaaaaaaaaacaaoa: asaaahaaaaaaaaaaaaaa 3:02
NS 288 Sea
A mam no: 03: -mHmUI -MHLUI UEOHOHUUH<HUEUOHH<U <UO<<U<UUOOH<UUO<<HU 5:51 MPLU
<< c<< a
885 E E 0:32 E oosaz E an: N 35F. 9:5 _ 35E 88% 253
g: gnu—Va:

 

.mozm vowwwgbocoawom cam—«6:82 Resin: 2 .5.“ flow Spin

2 5883

60

3:5

2:5
-NENN

moo

ma

Saw

me

3038
- _ mm:

Dva
-m.._<>n=._

.8: SEE 02820:: we use . m £033 .53 A<<v Eon oEE< d
d: . D E38 .0 £563: :88 £383: + Xm $32 .9532 :8: 338% 59: SEE .832 5.5 .o
32822:. 30:52 8865 AL macaw .c
.65 was: 86on 59: $55 .a
aaaaaaaaaaaaaeaaaaa: aaaaaaaaoaaaaaaaaaaacaaa 3:32
AHan HO
-_ mam OHOEO<OOOHOE<UHO< OH<O<OH<<<H<<<<O<UOEO 535m _ mm

aaaaaaaaaaaaaaaaaU: asaaaUaaacaaaaaaaaaa: Huy—

omxm
-miZ: e<ooo<oo<ooeoeoEoe e<o<oEoeoo<<o<<oeoe< 55:: 522
.5588 3 522.3.

61

Appendix 2-8

7 . . a . . . .
Eighty-SIX universal mammalian sequence tagged SlteS - human chromosomal locations and names

 

 

Locus Gene Product Human Band Primer 1 Primer 2
Name
Chromosome Name Name
PND Pronatriodilatin 1 p36 DPNDEXlD DPNDEXZU
PKLR Pyruvate Kinase — RBC 1 q21 HPKLREX4D HPKLREX6U
AT3 Antithrombin III I q23-q25 HAT3EX3D HAT3EX4U
REN Renin l q32 HRENEX8D HRENEX9U
SFTP3 Pulmonary Surfactant 2 pl 1.2 DSFPT3EX4D DSFTP3EX5U
Protein 3
SPTBNl Beta Spectrin (Non-RBC) 2 p21 HSPTBNlEXl3D HSPTBNIEXMU
APOB Apolipoprotein B 2 p24-p23 HAPOBEX26D HAPOBEX26U
[LIA Interleukin 1 Alpha 2 q13 HILIAEXZD SILIAEX3U
COL3A1 Collagen 11] Alpha 1 2 q3l-q32.3 HCOL3AIEX24D HCOL3AIEX25U
ELN Elastin 2 q3l-qter HELNEX32D HELNEX33U
PAX3 Human Paired Domain 2 2 q34-qB6 HHUP2EX2D HHUP2EX3U
GCG Glucagon 2 q36-q37 HGCGEX4D HGCGEXSU
PITl Pituitary-Specific 3 pl] HPlTlEX4D HPITIEXSU
Transcription Factor 1
GLBl B-Galactosidase 3 pter-p21 HGLBIEX8D HGLBIEX9U
GPXl Glutathione Peroxidase l 3 ql l-qu HGPXlEXlD HGPXIEXZU
TF Transferrin 3 q21 HTFEX7D HTFEXSU
RHOl Rhodopsin 3 q21-qter HRHOEX3D HRHOEX4U
GLUT2 Glucose Transport-like 2 3 q26. l- HGLUUEX9D HGLUT‘ZEX lOU
26.3
SST Somatostatin 3 228 HSSTEXID HSSTEX2U
HOX7 Homeobox 7 4 pl6.l HHOX7EX2D HHOX7EX2U
PDEB cGMP Phosphodiesterase 4 pter HPDEBEXMD HPDEBEXISU
Beta
ALB Albumin 4 q] l-ql3 HALBEX4D HALBEXSU
KIT c-KIT Protooncogene 4 q12-ql3 HKITEXISD HKITEXZOU
F GG Fibrinogen Gamma 4 q28 HFGGEXSD HFGGEX9U
GHR Growth Hormone 5 pl3.l-p12 HGHREX9D HGHREXIOU
Receptor
HEXB Beta Hexosaminidase 5 ql3 HHEXBEXIZD HHEXBEX13U
1L4 Interleukin 4 5 q23-q3l HIL4EXlD HIL4EX2U
ADRBZ Adrenergic Receptor Beta 5 q3 l-q32 HADRBZEXID HADRBZEXIU
2
CSFIR CSF-l Receptor 5 q33-q35 HCSFIREX3D HCSFlEX4U
TNFA Tumor Necrosis Factor 6 p213 HTNFAEXID HTNFAEX4U
Al ha
EDNl Enidothelin l 6 p24-p23 HEDNIEX3D HEDNIEX4U
COL9AI Collagen IX Alpha l 6 q12-ql4 HCOL9A1EX3D HCOL9AIEX4U

62

Appendix 2-8 (cont’d).

COLIOAI Collagen Type X Alpha I

PLG
EPO
CFT R

TCRB
SFTP2

CA2
TG
ALDOB

C5

ABL
RET
TDT
OAT

WT]
LDHA
INS
CD20
ROM]

APOC3

LDHB

1L6

TPI
COL2AI
DCNI
IGF l

PLA2
RBI

F7

CHY
CKBB
TCRA
BZM
CYP1A1

PKM

F ES
HGBA

Plasminogen

Erythropoeitin

Cystic Fibrosis Trans.
Regulator

T-Cell Receptor Beta

Pulmonary Suractant
Protein 2
Carbonic Anhydrase II

Thyroglobulin
Aldolase B

Complement Factor 5
ABL Proto-oncogene
RET Proto-Oncogene
Terminal Transferase
Omithine
Aminotranferase

Wilms Tumor 1

Lactate Dehydrogenase A
Insulin

CD20

Rod Outer Segment
Protein-1

Apolipoprotein C3

von Willebrand's Factor
Lactate Dehydrogenase B

Interleukin 6
Triosphosphate Isomerase
Collagen 11 Alpha 1
Decorin

Insulin-Like Growth
Factor 1
Phospholipase A2
Retinoblastoma 1
Clotting Factor VII
Chymase (Mast Cell)
Creatine Kinase Brain
T-Cell Receptor Alpha
Beta-2 Microglobulin
Cytochrome P-450
(AHH)

Pyruvate Kinase —

Muscle
FES Prom-Oncogene

Alpha Hemoglobin

6 q21-q22 HCOLIOAIEXZD
6 q25-q27 HPLGEXI8D
7 q21 HEPOEXZD
7 q31-q32 HCFEX22D
7 q35 DTCRBEXZD
8 p21 HSFTP2EX2D
8 q22 CAUNIVEX3D
8 q24 HTGEX9D
9 q21.3- HALDOBEX7D
q22.2
9 q22-q34 HC5EX36D
9 q34 HABLEXIOD
10 q11.2 HRETEX19D
10 q23-qZ4 HTDTEX9D
IO q26 HOATEX7D
11 p13 HWTIEX8D
11 p14-15.5 HLDHAEX3D
11 p15.5 DINSEXZD
11 q12-q13.l HCD20EX6D
11q13 HROMIEXID
ll q23-qter DAPOC3EX2D
12 p HVWFEX46D
12 p12.1- HLDHBEX3D
12.2
12 p12.2-p12 HIL6EX3D
12 p13 HTPIEXZD
12 q14.3 HCOL2AIEX2D
12 q21-q23 HDCNEX6D
12 q22 HIGFIEX3D
12 q23-qter DPLA2EX2D
13 q14.2 HRBIEXZSD
13 q34 HF7EX7D
14 ql 1.2 DCHYEX4D
l4 q32.3 DCKBEX6D
l4 q34 DTCRAEX3D
15 q21-q22.2 HBZMEXZD
15 q22-q24 DCYPlAlEX3D
15 q22-qter HPKMEXZD
15 q25-qter HFESEXMD
l6 pl3.3 HHGBAEXZD

63

HCOLIOAI EX2U
HPLGEX 19U
HEPOEX3U
HCFEX23U

DTCRBEX3U
HSFTPZEX4U

HCAIIEX4U
HTGEX IOU
HALDOBEXSU

HC5EX37U
HABLEX l 1U
HRETEXZOU
HTDTEX l 0U
HOATEX8U

HWTI EX9U
HLDHAEX4U
DINSEXBU
HCDZOEX7U
HROM 1 EX 1 U

DAPOCBEX3U
HVWFEX47U
HLDHBEX4U

DIL6EX4U
HTPIEXSU
HCOL2A l EX3U
HDCNEX7U
HIGFI EX4U

DPLAZEX3U
HRBl EX26U
HF7EX8U
DCHYEXSU
DCKBEXBU
DTCRAEX4U
HBZMEX3U
DCYPI A l EXSU

HPKMEX3U

HFESEXISU
HHGBAEX3U

Appendix 2-8 (cont’d).

GOT2

CTRB

APRT

TP53
NF 1
SCN4A

TS

CKMM

PVALB
DYS
MNK
HPRT

F9

F8
SRY

Glutamate Oxaloacetate
Transaminase 2
Chymotrypsinogen

Adenosine PR
Transferase
Tumor Protein 53

Neurofibromatosis 1

Skeletal Muscle Sodium
Channel

Thymidylate Synthetase
Apolipoprotein C2
Creatine Kinase Muscle

Parvalbumin
Dystrophin
Menkes Protein

Hypoxanthine PR
Transferase
Clotting Factor IX

Clotting Factor VIII
Sex Determining Region
- Y

16 q21-q22 HGOTZEXSD

16 q22.3- DCTRBEXSD
q23.2

16 q24 HAPRTEX3D

17 pl3.1 HTP53EX5D

l7 qll.2 HNFIEX6D

l7 q23.1- HSCN4AEX23D
q25.3

l8 pter-q12 HTSEXSD

l9 ql3.2 DAPOCZEX3D

l9 ql3.2— DCKMEXZD
q13.3

22 q12-ql3.1 HPVALBEX3D

X p21 DDYSEX7D

X q12-ql3.3 HMNKEX

X q26 HHPRTEX7D

X q26.3- DF9EX7D
q27.1

X q28 HF8EX24D

Y p11.3 HSRYEXID

HGOTZEX7U

DCTRBEX6U

HAPRTEXSU

HTP53EX7U
HNFI EX7U
HSCN4AEX24U

HTSEX6U
DAPOCZEX4U
DCKMEX3U

HPVALBEX4U
DDYSEX7U
HMNKEX
HHPRTEX8U

DF9EX8U

HF8EX25U
HSRYEX 1 U

 

a- For convenience, the eleven loci described in detail are included in the Appendices

64

Appendix 2-9

Eighty-six universal mammalian sequence tagged sites - sequences and sizes

 

 

PCR Product Size
Locus Primer 1 Primer 2 Human Dog
Name
Sequence Sequence Genomic Genomica

PND GCAGACCTGCTGGA'I‘I‘TCAAG CAGTCCGCTCTGGGCTCCAAT 360 360
PKLR CGCCTCAAGGAGATGATCAA ATGAGCCCGTCGTCAATGTA 660 500
AT3 CTTCTITGCCAAACTGAACTG GGGCTGAACT'I'TGACTTCCA 658 660
REN ACACTCCCCGACATCTCTIT CGCCGATCAAACTCTGTGTA 137 137
SFTP3 GGAAGTTCCTGGAGCATGAG CACAGGCCCAGGTGCTTACA 308 3 10
SPTBNl TCTCAAGACTATGGCAAACA CTGCCATCTCCCAGAAGAA 640 800
APOB GTAAAAGCTCAGTATAAGAAA GTGCCCTCTAATTTGTACTG 460 460

AAC
ILIA AGAAGTCAAGATGGCCAAAGT TGATI‘CAGAGACAGATGGTC 1900 1900
COL3A1 GGACCAGGAAGTGATGGGAA ACT'I'TCTCC'ITGAC'I'I‘CCCT 752 1400
ELN GCTGCAGCCGCTAAAGCAG AGGACACCTCCAAGGCCAG 600 1300
PAX3 GCCACAAGATCGTGGAGATG GGTTCTCTC'ITI'I‘GTATTCCTC 1020 1 170
GCG TTCATTGC'I'I‘GGCTGGTGAA GTGTTCATCTCATCAGAGAA 700 600
PITI TTCAGTCAAACAACAATCTG GCTCCCACTTTTTCATTGTA 700 1000
GLB 1 GAATTCTATACTGGCTGGCT CATTCCAATAGGCAAAATTGG 700 1 000

T
GPX l GACTACACCCAGATGAACGA CAGGAACTTCTCAAAG'I'TCC 633 633
TF GCTGACAGGGACCAGTATGA AACAGCAGGTCCTTCCCATG l 700 585
RHOI TACATGTTCGTGGTCCAC'IT TGGTGGGTGAAGATGTAGAA 1479 553
GLUT2 TGGATGAGT'TATGTGAGCAT GACT’ITCCT'ITGGTTTCTGG 364 364
SST GACTCCCGAGGCTI‘CCTCTTTG ATACTGCAGGAGAGAGAAGA 1200 1200
A

HOX7 AAGTTCCGCCAGAAGCAGTA ATCT'I‘CAGCTTCTCCAGCTC 400 400
PDEB CTGAAGAGCTACTACACGGA TGACACTTGTTCATCCACCA 300 300
ALB GGCTGACTGCTGTGCAAAACA AAGTAAGGATGTCTTCTGGC 730 730
KIT CCTGTGAAGTGGATGGCACC GCATCCCAGCAAGTCTTCAT 1000 1000
FGG CAATATAAAGAAGGATTTGGA TGACAC'I'I‘GTTCATCCACCA 1422 3000

CA
GHR CCAGT’I‘CCAGT’I‘CCAAAGAT TGATI‘CTTCTGGTCAAGGCA 238 200
HEXB TTCATTGGTGGAGAAGCTTG ATCT'ITGGAACTCCAGAGTC 1400 1000
1L4 CTATTAATGGGTCTCACCTCCC TCAACTCGGTGCACAGAGTCT 469 450

AACT TGG
ADRBZ CCCATTCAGATGCACTGGTA GCAGCCAGCAGAGGGTGAA 381 281
CSF 1 R TTCCAAAACACGGGGACCTA CATGCCAGGGCGAGAAGGA 1200 800
TNFA CTCAGCCTCTTCTCCTTCCT ATGGGCTCATACCAGGGCTT 1 198 1200
EDNI CCAAAAAGACAAGAAGTGCTG TGGAACAGTCTITTCCTITCTT I400 800
COL9A1 ATCAGGATTGGCCAAGATGA GGAATCCTGAAGTCTACATT 484 500
COLI 0A1 ATTCTCTCCAAAGCTI‘ACCC GCCACTAGGAATCCTGAGAA 340 340
PLG CAGCTCCCTGTGA'I'I‘GAGAA TAGACACCAGGCTI‘A'I'I‘GGG 1 100 1 100

65

Appendix 2-9 (cont’d).

EPO

CFTR
TCRB
SF“?
CA2
TG

ALDOB
C5

CD20

RON“
APOC3
VWW'
LDHB
1L6
TPI
COL2A1
DCNI
IGF I
PLA2
RBI

F7
CHY
CKBB
TCRA

82M
CYPLAI
PKNI

FES
HGBA
GOTZ
CTRB
APRT

TP53

CTCCCTCTGGGCCTCCCAGT

CTAAGCCATGGCCACAAGCA
GACTGTGGCTTCACCTCGG
CAGAAACACACGGAGATGGT
CAGTTCCATTTTCAGTGGGG
TTCACCTCAGAGTGCTACTG

GTGACTGCTGGACATGCCTG
TGTGTACGATTCCGGATAIIIG
A
TCAGACGAAGTGGAAAAGGA

CCCTTCCACATGGATTGAAA
ACCTGGAAGGCCATCCGTGT
CGTGCTCTTCAGGATCCAAA
GAGAAACCATACCAGTGTGA
AACTCCAAGCTGGTCATTAT
GAGCGCGGCTTCTTCTACAC

CTCI I IGCTGCCAI I ICTGGAA
T
CAGAGGACGGGCCACAGAA

CAGGAACAGAGGTGCCATGC
CCAGAGCGCATGGAGGCCTG
TTCCTCAGATCGTCAAGTACA
GCACTGGCAGAAAACAACCT
TATATCGACTTCGCCCGGCA
CTCTGCGACGACATAATCTG
GTTGATGCAGCTAGCCTGAA
GGCATCGTGGATGAGTGCTG
GACTACGGCTGCTACTGTG

CHTCCAGAAAATAAATCAGAT
GGT
AATGGAGCTCAGTTGTGTGG

GTCCCACCTGGGAGAATGTG
TGGATCAACGAGGAAGACCA

ACTGTCTGCCTGTTCACCGATT
T
TTCAGCAAGGACTGGTCIII

TTGGACCTC I I I GGAGCTGG
GCCTTCATTCAGACCCAGCA

GGGGAACTTTGGCGAAGTGTT
CCCACCACCAAGACCTACTT
TTTAAGTTCAGCCGAGATGT
AACGACATCACCCTGCTGTT
GACTCCCGAGGCTTCCTCTTNS

TACAAGCAGTCACAGCACAT

CCATCCTCTTCCAGGCATAGA
A
CATTGCTTCTATCCTGTGTT

GATCTCATAGAGGATGGTGG
GCCATCTTCATGATGTAGCA
GGCCAGTCCATCAGGTTGCT

GCTTCTCTGTAGCTCATGATC
TT
TTTGCAGCCTTGCCACCCC

GCTCCTTCACAGAC I I I CTG

AGAAGGCGCTCATC'ITCATT
CATCCAGTTAGCATATACAC
CGCCGGAGGTCTCTCTCAAA
GCCAGCCATCTACCAGTTCT
GTTTTACCTGTATGAGTCCT

GAATCCAGATTGCAACCACT

GGTAGAGGGAGCAGATGCTG
G
TGGAAGAAGGCAAAGATCAG
CAT
GTTAAACACCACAGAGGCCTT

TGCGCCACCTGGGACTCCTG
CTGCACTCCAGCTTGAATCC
CTGCTGGGATGAATGCCAAG
ATCTGAAACTCCACAAGACC
ATGGCCCACACAGGCTCATA
TCTCCAGGTTCTCCTTTCTG
AAGTGAAGCTCCCTCAGATG
CTCCTTCTGTTCCCCTCCTG
TTACAGCTGGCCAGTTTCTT
ACTCATTTCTGCCAGTTTCTG

CGATGTCGTGGTTGGTGGT
TGGGAGATTCGGGTGAAGAC
TTCACACCATCCACCACCAT

GTAACAACTTGGCATCACAGG
AAT
CTGCTTACATGTCTCGATCT

TGGTTGATCTGCCACTGGTT

ATTCCAGACTTAATCATCTCC
TT
TCCATGACGATGTAGATGGG

CGGTA I I I GGAGGTCAGCAC
CTTGGTAGGCCATGTCAAA
TGCAGGAGGAGACGCCACT

ATACTGCAGGAGAGAGAAGA
A
TCTTCCAGTGTGATGATGGT

66

478

Wfi
238
SN)
3X)
6%)

MB
300

151
I8M)
1200
153
8%)
786
6%)

WE

9M)
1mm
7M)
6fl)
I72
3M)
6M)
IZM)
9X)
420
I600

1600
9“)
2M)
mm

12¢4
8m)
1200

467
548
I400
592
an

1600

6W)

8W)
238
6M)
15m)
6“)

4fl)
300

151
raw
1250
153
Bfl)
264
6fl)

8M)

mm
18“)
8fl)
6fl)
I72
500
1100
IBM)
9M)
3m)
1600

7N)

9fl)
9M)

1100
8a)
1200

436
4m)
1300
6M)
1300

I600

Appendix 2-9 (cont’d).

NFI
SCN4A

TS
APOCZ
CKMM
PVALB
DYS
MN K

HPRT
F9
F8

SRY

ATTCACTCTCTGTGTACTTG CAAAGCTTCTGTGACTGTTT

CTCAAGGTGGACATCCTGTACA AGCAGCGTCCGGATGCCCTT
A
TGCCAGTTCTATGTGGTGAA AGGTAAATATGTGCATCTCC

GAATCACTCTACAGTTACTGG AGCTGCTGTGCI I I IGCTGTA
AAGAAGCTGCGGGACAAGGA CAGCCCACGGTCATGATGAAA
ATGTGAAGAAGGTGI I ICACAT TCI I IGTCTCCAGCAGCCAT

GI I ICAGGCCAGACCTCI I I TACCGACCTTCAGGATCAAG
GGCATGACTTGTAATTCCTG CATCAAATCCCATGTCTTCTA
T

AGCTTGCTGGTGAAAAGGAC TTATAGTCAAGGGCATATCC
TGGGTGGTAACTGCAGCCCACT CTACGCACACTCTTCACCCCA

GATGCACAGATTACTGCTTC GTAAGCAGAGAI I I IACTCCC
TG
AAGCGACCCATGAACGCATT TTCGGGTAI I ICTCTCTGTG

440
1 177

915
201
I I I9
420
500
669

766
650
779

2500

350
1100

750
201
I 100
830
500
750

650
650
800

2500

 

a - All PCR products have been sequenced or, in a few cases, are derived from primers made to
the published canine sequence.

67

Chapter 3

Estimate of Nucleotide Diversity in Dogs Using a Pool-and-
Sequence Method

68

 

Estimate of Nucleotide Diversity in Dogs Using a Pool-and-Sequence Method

Running Head: Dog Nucleotide Diversity

James A. Brouillette,l’3’4 Jennifer R. Andrew,1 Patrick J. Venta";3

1Department of Microbiology, College of Veterinary Medicine, Michigan State
University, East Lansing, MI 48824, USA

2Department of Small Animal Clinical Sciences, College of Veterinary Medicine,
Michigan State University, East Lansing, MI 48824-1314, USA

3Genetics Program, Michigan State University, East Lansing, MI 48824, USA

dColIege of Human Medicine, Michigan State University, East Lansing, MI 48824, USA

Corresponding Author:
Patrick J. Venta
Telephone: 517—432-251 5

FAX: 517-432-2524

E-mail: venta@gvmmsuedu

 

69

Abstract

Nucleotide diversity (7t), the average number of base differences per site for two
homologous sequences randomly selected from a population, is an important parameter
used to understand the structure and history of populations. It is also important for
determining the feasibility of developing a genetic map for a species based upon single
nucleotide polymorphisms (SNPs). Nucleotide diversity has never been estimated for
dogs. Segments of twelve canine genes from ten diverse dog breeds were examined for
nucleotide variation by using a pool-and-sequence method. We identified three SNPs in
the coding regions (2,501 bp) and eleven SNPs in the introns (2,953 bp). Each of these
putative SNPs was tested by restriction enzyme analysis and all were verified. Six
additional SNPs were identified in a single SINE contained in one gene. Using these
data, canine sequence diversity across breeds was estimated to be 0.001 and 0.0004 in
intronic and coding regions, respectively, with SNPs spaced every 400 bp on average.
Discovery of useful SNPs in seven of the twelve genes suggests that construction of
canine SNP-based map can be accomplished using current technology. Thirteen
polymorphic SNPs were also found in 5,847 bp in the cat, horse, ox, and pig, using four
of the same genes from which canine nucleotide diversity was estimated. These results

suggest that these species may have similar amounts of nucleotide diversity.

70

Introduction

Although canine isozyme studies have been used to estimate genetic heterozygosity in
dogs, nucleotide diversity has never been examined (Fisher et al., 1976; Simonsen, 1976;
Weiden et al., 1974). Estimating the nucleotide diversity, 1t, of a species is important for
answering questions about population structures and for understanding how genetic
variation is maintained in populations (Nei and Li, 1979). Although most nucleotide
diversity studies have been conducted using mitochondrial DNAs, several studies have
been done in the past few years on human nuclear genes (e. g., Harding et al., 1997; Clark
et al., 1998; Nickerson et al., 1998, Cargill et al., 1999, Halushka et al., 1999). These
studies are already yielding information on the histories and relationships of populations

around the world.

From the standpoint of mapping medically important genes, the studies of human
nucleotide diversity have shown that it will be possible to build genetic maps with a very
high density of single nucleotide polymorphisms (SNPs). It is anticipated that tens or
hundreds of thousands of SNPs will be used in linkage- and disequilibrium-based
mapping because of the potential to automate the typing of SNPs (Brookes, 1999; Collins
et al., 1999; Kruglyak, 1999). In domestic animals, similar mapping strategies could be
used to locate genes of interest. Linkage disequilibrium methods may be even easier to
use, given the possible greater amount of linkage disequilibrium in domestic animal
species (cf. Farnir et al., 2000). However, estimates of nucleotide diversity for most
domesticated animals do not exist and these would be useful for determining the amount

of effort needed to develop SNP-based maps.

71

Dogs are important work and companion animals in many human societies. We are
interested in identifying the mutations that cause inherited diseases in dogs so that
breeders can avoid producing affected animals (Padgett, 1998; Willis, 1989; Clark and
Stainer, 1994). We would also like to gain insights on the genetic history and relatedness
of the various breeds. Although many SNPs have been found in various genes in the dog,
the amount of nucleotide diversity cannot be calculated from these studies because

negative results for polymorphism searches are rarely reported.

This study provides an estimate of canine nucleotide diversity by ascertaining all variant
and non-variant nucleotides in portions of twelve canine genes using a pool-and-sequence
(PAS) method. Four of the genes studied in the dog were also examined for SNPs in four

additional species (pigs, horses, cattle, and cats) and the results are reported here.

72

Materials and methods

Target DNA

Equimolar DNA pools were made from the following breeds for each species: dogs;
Cocker spaniel, Greyhound, Doberman pinscher, Siberian husky, Labrador retriever,
Collie, Scottish terrier, German shepherd dog, Beagle, and Pointer; cattle; Red-and-White
Holstein, Holstein, Aryshire, Brown Swiss, Hereford, and Angus; horses; Appaloosa,
Arabian, Belgian, Percheron, Miniature, Thoroughbred, Quarterhorse, and Rocky
Mountain; and pigs; Chesterwhite, Duroc, Hampshire, Landrace, and Yorkshire. A DNA
pool was also made from the DNA of 10 unrelated cats of unspecified breed. The
unmixed purified DNA samples were kindly provided by each of the following
individuals; dog, Dr. Vilma Yuzbasiyan-Gurkan; horse, Dr. Susan Ewart; pig, Dr. Cathy
Ernst; and cat, Dr. John Kruger. Dr. Robert Holland kindly provided peripheral blood
samples from cattle for DNA isolation. DNA was isolated fiom white blood cells using a
previously published phenol-chloroform extraction method, except that the proeinase K
digestion step was omitted (Sambrook et al., 1989). DNA was precipitated using
ammonium chloride / ethanol precipitation. A DNA sample was used from an individual
animal to serve as a sequencing control. Cats, cattle, horses and pigs were chosen as
variation 'controls (see Results) because they are domesticated animal species
representing three different mammalian orders. In addition, mapping projects are

underway in each of these species as well as in dogs.

73

..1

Target genes

Six genes were chosen for the study because they are candidate genes for other studies by
our lab (RHOI, GNA T1, NCAD, RDS, AC T C, and RYRZ). Six additional genes (C5,

CFT R, CYP1A1, FES, TS, and WT I ) were chosen because the primer pairs had already
shown to amplify reliably, the product size was between 500 bp and 1000 bp, the primer
pairs spanned an intron, and some of them had been shown to amplify target regions in

other mammalian species of interest.

Primer Design

Primer pairs have been previously described for seven of the genes examined, C5, CFT R,
CYP1A1, FES, RHOI, TS, and WT I (Venta et al., 1996). For the other five genes (ACTC,
GNA T1 , NCAD, RDS, and RYRZ), primers were designed by comparing nucleotide
sequences of two known mammalian species (usually human with mouse or rat), for

which at least one exon-intron gene structure was available (Appendix 3-1).
New primers were designed as before, except that the 3’ ends of the primers always

overlie the most conserved codon position for a given amino acid (Venta et al., 1996;

Fitch, 1976). Pairs of primers were designed to have similar annealing temperatures, and

74

the potential for primer-dimer formation was avoided. Primer sequences are given in

Appendix 39.

PCR Amplification of Gene Segments

PCR was performed using 50 ng of genomic DNA from either the pooled sample or from
the individual control sample, in 25 pl vols containing the following: 20 mM Tris-HCI
(pH 8.3), 50 mM KCI, 1.5 mM MgC12, 100 M dNTPs, 10 pmol each primer, and 1.25 U
of Taq polymerase. Cycling conditions for each primer pair are given in Appendix 3-9.
PCR products were run on 1% agarose, 1 X TBE gels at 125 volts, stained in ethidium

bromide, and photographed on a 360 nm ultraviolet transilluminator.

Purification of PCR Products

PCR products were run on agarose gels, gel slices containing PCR products were taken,
and the DNA was extracted from the agarose using the QIAEX 11 Gel Extraction Kit
(QIAGEN Corp., Valencia, CA) as recommended by the manufacturer. Aliquots of the
extraction products were visualized afier electrophoresis on a 1% agarose, l X TBE gel

by staining with ethidium bromide.

75

Pool-and-Sequence Method

Approximately 100 ng of each purified PCR product was used as template for cycle
sequencing. Cycle sequencing was performed using the Thermo Sequenase 33P-
radiolabeled terminator cycle sequencing kit (USB Corp., Cleveland, OH), using the
dGTP termination mix, as directed by the manufacturer. Products of the sequencing
reaction were run at 50 W for 1.5 to 4.5 hr on 6% acrylamide gels (19:1
acrylamidezbisacrylamide) containing 7M urea. Gels were dried and exposed to x-ray
film (Kodak X-omat) at room temperature for 24-48 hr. The identity of the sequence was
examined by performing a BLAST search (http://wwwncbi.nlm.nih.gov/BLAST)
(Altschul et al., 1997). Possible polymorphic sites were identified by visual inspection of
the sequencing ladders for bands in two lanes at identical base positions or by the band-
doubling effect seen for insertion/deletion polymorphisms. Presrunptive false positive

bands were identified by comparison of the single animal and pooled animal sequencing

ladders.
Diagnostic Tests for Identified SNPs

Once a putative single nucleotide polymorphism was located on the sequencing gel, a
restriction site was created by PCR primer mutagenesis adjacent to the polymorphic site,
unless a naturally occurring site was present (Venta et al., 1991). New primers were

designed to complement the species-specific sequence as determined by the sequencing

76

reactions, except that one or two nucleotides were substituted to create a restriction site in
one of the two alleles at the polymorphic site (Appendix 3-10). Nested amplifications
were required to produce consistent results for cat WTOI , cat WT02, dog CFTROI , horse
WTOI, and ox TSOl using the sequencing primers first, followed by the diagnostic

primers.

Restriction Digestion and Estimation of Nucleotide Diversity

All restriction enzymes were purchased from one supplier (New England Biolabs,
Berverly, MA) except for Tail (F errnentas, Amherst, NJ) and MaeIII (Boehringer
Mannheim, Indianapolis, IN). Restriction digestion was performed directly in the PCR
buffer with the addition of 50 mM MgCl 2 to reach a final concentration of approximately
10 mM. Restriction digestion reactions were carried out at the temperature recommended
by the manufacturer. All of the enzymes used in this study worked well in PCR bufier
supplemented with extra MgCIz except TM] 1 1 I and NIaIII. For these enzymes, PCR
products were precipitated with ethanol and ammonium acetate, resuspended in a suitable
volume of TE (10 mM Tris, 1 mM EDTA, pH 7.4-8.0) and digested following the

manufacturer’s instructions.

Nucleotide diversity was calculated according to a previously published method (N ei and
Li, 1979). Briefly, the average number of nucleotide differences across the total length of
DNA sequenced in the study was calculated taking the allele frequency for each variant

site into account.

77

Results

Identity of the Genes

In order to confirm that we were amplifying the homologue of the gene of interest, a
BLAST search of the Genbank database (release 113) was performed using the
nucleotide sequence data we obtained (Genbank accession numbers are listed in
Appendix 3-1). The highest BLAST score always matched one or more mammalian
homologs of the gene for which the primers had been designed (data not shown). Most
of the nucleotide sequence matches were contained within the coding region of each

gene. The intronic sequence generally had a relatively minor impact on the score.

Use of the Pool-and-Sequence Method for Identification of SNPs

Several SNPs can be seen as bands that occur at identical positions in the sequencing
ladders of Appendix 3-6 and Appendix 3-7. The use of a sequencing ladder from a single
animal served as an important control to recognize false positive SNPs for several of the
genes examined (data not shown). The allele frequencies for any given SNP could be
estimated from the relative intensities of the bands. For example, in Appendix 3-6A,
SNP 1 (TSOI) shows a very weak band in the C lane (Pool) and a much stronger one at
the same position in the A lane. The alleles of this SNP can be distinguished by digestion
of the PCR product with Mon (Appendix 3-2). The allele counts for the Pool are 2 (the

C allele) and 18 (the A allele) (Appendix 3-3). The allele counts for SNP 4 (T804) are 10

78

and 10 which are reflected in the banding intensities of the sequencing ladder (Appendix

3-6, Pool).

Indels (insertion/deletions) were recognized as the occurrence of doublet bands above the
site of variation (Appendix 3-6A, SNP 2 [T802] and Appendix 3-7B, SNP 3 [CFTR03]).
Banding intensities for indels also give an indication of the relative allele frequencies as
shown by CF TR03 (Appendix 3-7B, Pool and Appendix 3-3). A SNP caused by a
transition was recognized as a single band among the doublets in the sequencing ladder

above the indel site (Appendix 3-6A, SNP 3 [T803], Pool).

A total of twenty SNPs were identified in 2,501 bp of coding and 2,953 bp of intronic
sequence. Six of the SNPs were contained within one SINE element in the T8 gene.
This SINE is 88% identical to the previously reported consensus sequence (Minnick et
al., 1992, Coltrnan and Wright, 1994, Bentolila et al., 1999). The fourteen non-SINE
SNPs were confirmed by restriction enzyme digestion of the DNA from the individual
pool members, either using naturally occurring restriction enzyme sites or ones that were
created by PCR mutagenesis (Appendix 3-2). The minor allele frequencies for each SNP
for the animals in the pool varied from 0.05 to 0.5 (Appendix 3-3). The rarest allele
discovered (0.05) was found in the heterozygous state in the single mixed breed animal,
and so it is not certain that it would have been detected by itself in the pool. We were able
to easily detect a minor allele which occurred at a frequency of 0.15 in the pool that was
not present in the single animal (R YR2, Appendix 3-3). We have also been able to detect

a rare allele at a frequency of 0.062 in horses (TS, see below).

79

Four SNPs were found in the TS gene in addition to the six SNPs seen in the SINE
(Appendix 3-2). Four SNPs were also found in the CFT R gene, three in the ACTC gene,
and one each in the GNA TI , NCAD, and RYR2 genes. Three of the fourteen SNPs were
found in the coding regions of the CFT R and GNA T1 genes. These three changes are
silent. Three of the SNPs (ACTCOI , CFTR02, and GNATOl) had a CpG dinucleotide
overlying one of the alleles. Cst are thought to be over-represented in SNPs due to
their higher mutation rates compared to other dinucleotides (Krawczak et al., 1998).
Three of the SNPs overlie indels. Of the remaining eleven non-indel SNPs, ten are
transitions and one is a transversion (Appendix 3-2). One simple tandem repeat (STR)
was also contained in one gene (WTI ), although we knew of its existence before the SNP

search was started (Shibuya et al., 1996; Venta et al., 1996).

Estimate of Canine Sequence Diversity

Canine sequence diversity was calculated with and without the six SNPs found in the TS
SIN E. The allele frequency of the TS SIN E SNPs were estimated by the relative banding
intensity from the sequencing film. The estimate for the rarer allele in each SNP in the
SINE is 0.4. Sequence diversity was calculated from the number of bases sequenced
(Appendix 3-1) and the amount of variability observed (Appendix 3-3). The estimate of
sequence diversity for intronic sequences with the SINE SNPs included was 0.0013 and
without the SINE SNPs was 0.0010. Coding region sequence diversity was 0.0004. If
we had missed all of the alleles whose frequency is less that 0.15, the sequence diversity

estimates would be biased downward about 16% from their true value, assuming that the

80

i .

infinite site model would produce a reasonable approximation for the distribution of

nucleotide variation in dog populations (cf. Clark et al., 1998).

Another useful parameter to estimate sequence diversity is the average frequency of
segregating polymorphic sites. One SNP was found every 834 bp of coding region

sequenced and every 268 bp of intronic region sequenced for an average of one SNP

every 390 bp of the total sequence.

We have not determined the haplotypes for the genes in which more than one SNP was
found. In many cases, however, it is possible to deduce that three or more haplotypes
exist by inspection of the SNP pattern across breeds. For example, at least four
haplotypes can be inferred for the SNPs found in the ACT C gene (Appendix 3-3). For
SNPs ACTCOI , ACTC02, and ACTC03, the Cocker spaniel must have a 2,1,2 haplotype,
the Doberman pinscher must have a 2,1,1 haplotype, the Siberian husky must have a
2,2,2 haplotype, and the Scottish terrier must have a 1,1,2 haplotype. For those genes
containing SNPs in the region sequenced (about 500 bp), the average heterozygosity is
0.33 (that is, one third of dogs examined are heterozygous for at least one SNP in a given

polymorphic gene). This value is about half of that seen for the average STR (Weber,

1990)

81

 

Variability of SNPs Within Dog Breeds

The variability of the SNPs within breeds is of interest because most linkage studies in
dogs will be conducted in single breeds. In order to examine the within-breed variability
of SNPs discovered by pooling across breeds, four SNPs were examined in which the
minor allele fiequency in the initial pool were typed. The dogs were unrelated for at least
three generations. The number of breeds in which heterozygotes are seen for each SNP is
correlated with the allele frequencies seen in the original pool (Appendices 3-3 and 3-4).
Only four gene/breed combinations (ACTC02, Scottish terriers; RYR201, Pointers; and
T804, Greyhounds and Collies) appear to show deviations from Hardy-Weinberg (HW)
expectations (as might be expected from inbreeding), although the sample sizes are
relatively small. For the data summed across breeds, there is a significant decrease in the
observed number of heterozygotes relative to the number expected under HW

equilibrium for ACTC02 (Chi-square = 24.49, P < 0.001), as would be predicted for
subdivisions of a base population (Wahlund, 1928; Wright, 1951). However, no
significant deviation was seen for T804 (Chi-square = 1.70, P = 0.43). It is uncertain if
the other two SNPs tested (RYR201 and T802) show a significant deviation because of
the low number of animals carrying the rarer allele. The important practical observation
is that a reasonable number of heterozygotes are seen within breeds for the SNPs that

were discovered by pooling breeds (Appendix 3-4).

82

Identification of SNPs in Other Mammalian Species

Examination of the same gene regions in other mammalian species would be expected to
show a correlation in sequence diversity if certain genes had a greater-than-average
tendency to maintain genetic variation. No correlation would suggest that our estimate of
nucleotide diversity was not biased by genes that were more, or less, prone to mutation.
Four genes (CFT R, FES, TS, and WT 1 ) used in the canine SNP search were also used to
search for SNPs in cats, horses, pigs, and cattle. Of the 16 gene-species combinations, all
targets amplified except for two (cat F ES and pig T8). For the 14 gene-species
combinations that amplified, putative SNPs were found in seven of them, including four

with two or more SNPs (Appendix 3-1).

Ten of the non-canine SNPs were tested by restriction enzyme analysis. All of the ten
tested SNPs were confirmed. Allele frequencies for the ten SNPs examined were
(species, SNP, allele 1, 2): horse; CFTROI, l3, 3; T801, 1, 15;WT101, 11, 5: cattle,
CFTROI, 6, 8; T801, 1, 13: cat; WTOI, 6, l4; WT02, 14, 6; WT03, 12, 6 (one sample
failed to amplify): and pig, FESOI, 4, 10. A simple tandem repeat was also discovered in
the ox TS gene, although it was not examined for genetic variability (GenBank Accession
No. AF 203030). Each of these SNPs is unique to the species in which it was located.
There is no obvious correlation across species for the presence of SNPs within a given
gene (Appendix 3-5), suggesting that the nucleotide diversity estimate has not been

biased by the inclusion of unusually polymorphic or non-polymorphic genes .

83

Discussion

The DNA sequence diversity estimate (N ei’s 1:) of 0.001 suggests that the
construction of a SNP based map of the canine genome is a feasible task with current
technology. In the present study, seven of the twelve genes contained useful SNPs within
an average of only 500 bp sequenced, suggesting that many of the available canine STSs
will also contain SNPs (e.g., Venta et al., 1996; Lyons et al., 1997). Many of the
universal mammalian STS primer sets tested on the canine genome will also be directly

applicable to the search for SNPs in other mammalian genomes (Appendix 3-5).

The estimated frequency of SNPs in the human genome is about 1 SNP every 500-1000
bp of DNA (Collins et al., 1997; Nickerson et al., 1998; Wang et al., 1998). The unusual
number of SNPs found in the canine TS SINE may represent a feature of some repetitive
elements and would not seem to reflect the average nucleotide diversity in a species
(Harumi et al., 1995). However, their inclusion in the diversity calculation had only a
minor impact on the result. The rest of the data suggest that the SNP frequency in the
canine genome is about one SNP every 400 bp. The frequency of SNPs (and the within
breed 1:) for a given breed will be lower (Appendix 3-4), but still adequate for most
mapping purposes. The data for the feline, bovine, porcine, and equine genomes suggest
that the levels of sequence diversity in these genomes may also be similar to that of dog

and human.

Pool-and-sequence allows a large number of alleles to be scanned simultaneously for

84

 

variability. The current detection limit for the minor allele is between five and ten
percent, although work is underway to determine if this detection limit can be lowered.
The manual method reported here is similar to recently reported automated PAS methods,
but may be more sensitive because of the more uniform banding intensities seen with
manual Thermo Sequenase with labeled terminators (Appendices 3-6 and 3-7). By
pooling ten animals (20 alleles) at once, it is possible to be at least 95% certain to detect
any SNP that occurs with an allele frequency of at least 20% in a given population (Lai et
al., 1998).

The method of PCR primer mutagenesis followed by restriction digestion to make allele
determinations has the advantage of being relatively inexpensive and simple to use.
Many other techniques are also available (e.g., allele-specific PCR, ASO hybridization,
or single-base extension) that can be used in combination with automated systems to

increase throughput.

It has been suggested (Kruglyak, 1999, Collins et. al., 1999) that a whole genome search
for disease genes could be performed utilizing linkage disequilibrium. Based on
computer modeling, it had been suggested that the area of linkage disequilibrium in
humans will be only a few kilobases (Kruglyak, 1999). Recent empirical data, however,
suggest that the area of linkage disequilibrium may be closer to 300 kb (Collins et al.,
1999). To our knowledge, no data exists on the range of linkage disequilibrium in dogs.
Due to the similarity in breeding patterns between dogs and cows (that is, a significant

influence of the founder effect), we suspect that the extent of linkage disequilibrium in

85

 

the canine populations will be similar to that found in cattle (Farnir et al., 2000). In
cattle, D’ values were around 50% at distances of 5 cM, falling off rapidly at distances

greater than 5 cM. Further work will be needed to judge the extent of linkage

disequilibrium in dogs.

A significant number of SNPs were found within a canine SINE element in the
thymidylate synthetase (TS) gene. There are an estimated 400,000 SINE elements in the
haploid canine genome, although the distribution of the SINES is not known (Bentolila et
al., 1999). It may be necessary to use nested primers with at least one of the primers in

unique flanking sequence to obtain usable data when a polymorphism is contained within

a SINE (e.g., Venta et al., 1999).

We expected to see a decrease in heterozygosity within breeds because some inbreeding
is typically used to fix desirable characteristics in lines. Surprisingly, there did not
appear to be much deviation from Hardy-Weinberg expectations. However, the number
of animals in each data set is small, and larger samples may show an effect of inbreeding
like that seen for other breeds using STR markers (Zajc et al., 1997; Morera et al., 1999).
A decrease in heterozygosity was also expected when the data was summed across breeds
(Wahlund, 1928; Wright, 1951). For the two most polymorphic SNPs tested, one showed

the expected decrease (ACTC02), but the other did not (T804). The reason for this

difference is presently unknown.

86

 

The SNP markers described in this report are all Type I (gene-specific) STS markers.
These markers could be used in concert with high-density DNA probe arrays or other
methods of automating SNP typing to make a rapid whole-genome search for linkage
(e.g., Wang et al., 1998). It has been calculated that about 1000 evenly spaced SNPs with

minor allele frequencies greater than or equal to 0.2 would produce the linkage power of
a BOO-marker STR map (Kruglyak, 1997). It should be feasible to produce this many
Type I SNPs in the near future. SNPs and STRs have complementary advantages for
linkage analysis (simpler automation vs. greater single marker genetic power [Brookes,
1999; Weber, 1990]) and therefore it seems likely that they will be used synergistically to
locate genes of interest. Finally, because the mutation rate is lower for single nucleotides

compared to tandem repeats, SNPs may help to produce a clearer picture of the history

and genetic relationships of the many dog breeds.

87

Acknowledgments

We thank Dr. Jerry Dodgson, Dr. Vilma Yuzbasiyan-Gurkan, Dr. Susan Ewart, and the
anonymous reviewers for a number of helpful comments on this manuscript. We also
thank Dr. Jill Murtha, Ellen Bishop, Maika Symkowiak, Pam Edwards, and Michael
Mienaltowksi for performing some of the typings done in this study. This work was
supported by grants from the American Kennel Club Canine Health Foundation, The
Morris Animal Foundation, and the Doberman Pinscher Foundation of America, and

from private donations from individuals interested in canine health and welfare.

88

Appendix 3-1

Gene segments with amount of DNA sequenced.

 

 

Gene Species GenBank 1a GenBank 23b Coding Seq. Intron Seq.
FES Dog L77674 157 239
Horse L77676 141 221

Pig L77679 99 118

Ox L77677 150 272

WTl Dog U00687 87 373
Cat AF201739 AF201740 89 345

Horse AF201736 AF201737 82 342

Pig AF202067 AF202068 81 377

Ox AF202074 AF201738 85 349

CFTR Dog L77683 L77689 139 248
Cat AF203024 AF203023 142 304

Horse AF202072 AF202071 146 305

Pig AF202075 AF203018 I47 327

Ox AF201741 AF201742 I44 294

T8 Dog AF201743 AF202073 136 183
Cat AF203027 AF203028 134 264

Horse AF202065 AF202066 134 313

Ox AF203029 AF203030 135 307

C5 Dog AF202069 AF202070 99 295
CYP1A1 Dog AF203025 AF203026 147 238
ACTC Dog AF203020 AF203019 202 87
GNATI Dog AF153706 186 208
NCAD Dog AF203017 284 107
RDS Dog U27349 416 0
RHO Dog X71380 319 307
RYR2 Dog AF203022 AF203021 93 436

 

a. Bold entries in these columns indicate sequences previously submitted by us or by others.
b. Blank cells in this column indicate that the complete sequence is contained in Genbank entry in the
column to the left.

89

Appendix 3-2

Location of SNPs, types of base changes, and diagnostic tests.
Type OI;

Species SNP

Dog ACTCOI
ACTC02
ACTC03

CFT'RO 1

CFTR02
CFTR03
CFTR04
GNATOI
NCADOI
ICYRZOI
RYR201i
T801
T802
T803
T804
T805

T806

Genbank
Accession
No.

5

Nucleotide

. . a
Posrtron

289
38
166

52

73
64
106
123
47
I 74
174

133

180
215
221
42

46

Change

A/T

C/T

C/T

A/G

NC

NC

l/D(1)

C/T

A/G

I/D (2)

C/G

90

Restriction
c
Enzyme

Tai I

Cac8 l

Tsp509 I

Ban II

Taq I

351 I

Hinf I

Tthllll

1le I

Mae III

Hinf l

Mwo I

Mse I

Pst I

Mse I

NT

NT

Invariagt
Bands

4 10,(290),

230
500

430,72,
68,50

117

450,082),
132,70,
22,14
225,186,
(182Ll4

225,186,
(182),“

Diagnosctlic
Bands
331,31 1,
20

178,157,
18

450,310,
144

500,430,
78

400,370,
27

161,141,
20

120,100,
20

400,380,
20

145,128,
17

1010,800,
200

1 01,75,
26

75,66,
9

246,200,
46

770,550,
219

92,87,

Appendix 3-2 (cont’d).

T807
T808
T809
T8 1 0
Horse CF TRO 1
CF TR02
T801
T802
WTl 01
Ox CF TRO l

T801

54

68

76

92

I36

144

I69

157

70

233

127

MG

C/T

A/G

A/C

C/T

C/G

A/T

A/C

C/T

A/G

NC

91

NT
NT
NT
NT
Cac8 1
NT
Hinfl
Tsp509 Ig
Mnl I

Mwo 1

BS] 1

200,166

219,161,
15

161,137,
24

229,192,
37

1 46,88,
58

103,68,
35

274,258,
1 6

390,370,
1 5

Appendix 3-2 (cont’d).

goo-m

5‘09 {-150

i.

Cat WTOI 5 l 12 NC 851 l --- 92,76,
16
WT02 5 162 NC Sau3A l --- 92,73,
- l9
WT03 3 119 A/G Nla III 530 139,104,
19
. . -_- h
Pig FESOI 5 196 C/T Hrnf l [340] ,249,
[l90],215,
[140],34
FESOZ 5 233 C/T NT --- --
F E803 5 290 CH“ NT --- «-
. Position of base change in Genbank entry.
. Base change seen on coding strand.
. NT = not tested. Bold entries in this column show naturally occurrm' g sites.
Band sizes in parentheses indicate confirmed or potential processed pseudogene bands that can

occasionally appear if the annealing temperature is not high enough. Bold numbers indicate sizes
determined by actual sequence analysis. Other sizes are estimated by the local Southern method
(Southern, 1979).

. Underlined nucleotides are cSNPs.

Number represents the number of bases involved in the indel.

. Hind III was also needed to separate two co-migrating bands that interfered with allele calling.
. A second segregating Hinfl site was also observed in a non-sequenced region that is in complete

linkage disequilibrium with the first site for all samples tested so far (this report and Ernst and Raney,
personal communication). The band sizes for this site are shown in brackets.

Alternative test for RYR201 that uses a more powerful restriction enzyme and a site

created by PCR mutagenesis.

92

.260. o... 55.3 8% 05 ..o 28 7.8. an 8.. 303388.. 0.3 8... .oon o... 5 2.8.5 ..o .8982. o... ..o «.63 2.. no .88328 w. 9838.82. 881. ..G

.3863 mzombocoS: m. x18. .0 So 28w 28 40.5% 28.80 2.. .8. 6...wa3 .955 some 8.. 38.3.8 m. £838.58: 280. mmob< .o

60.8% 0.8 m3...” gomeocouo... ...

6.92.5.

.8522. o... .8. 038 2.. ..o .8.— o... a. com: 28.3.3.8 :. 3.5.0:. 8: Pa .958 ...... 8.. 3.3.2 2.. 68.5 .8838 a on. 9 3839:. m. .955 2?: 2.... d

 

 

 

 

ebfiofieos:

m.mm E. o. ON ON on on 260..

ov om o. ON o. ON ON ON ow om ON on o o $838.82. new

e. .. a. m. m o. o. a e. h w m. w e. N o.o..<

o. m . N n. o. o. .. v m. N. m c. v .o.o..<
...e. .. NN NN NN NN .. .. NN N. NN NN NN .. .. .859.
new N. N. NN NN NN N. N. N. N. .. N. NN .. NN 038m.
m.mm N. N. NN NN NN .. .. NN N. N. NN NN .. NN Poznanmfigoc
...e. NN NN N. NN NN NN .. NN NN NN NN NN .. .. cow—285.30%
2mm .. NN NN N. .. .. NN N. N. N. .. .. NN NN 3:00
m.mm N. N. N N N N N N N N N N . . N N . . . . N. . . N N 85.52 8.853
m.mm NN .. NN NN N. NN NN .. NN .. .. N. NN NN Emsngtonw
m.mm .. NN NN N. NN .. .. .. NN .. .. N. .. NN honomfigqmaconon.
odm NN .. NN NN NN N. N. N. NN N. N. NN .. NN .uqzonmoco
he. N. .. NN NN NN NN NN .. NN .. .. NN .. NN 3.33.380
. N _ .255, 295
new NN .. N. N. .. N. .. .. N. N. N. N. n NN a
.. £50.. 3 no No we no No .o no No S

o m... m... m... .o m... Nm>m Q<UZ ....<20 #50 #50 M85 y.......O U...U< UHU< O...U<

$83., teem.

 

.86 3.28308 05 a. com: mwoe wcofia 950.. £75 8.. $833282. use $5.8 o.o..<

E 588?.

93

Appendix 3-4

Genotypes for four SNPs within Canine Breeds.

 

 

 

Breed No. of ACTC02 RYR201 T802 T804
Dogs Genotype Genotype Genotype Genotype
l/l, l/2,2/2 1/1, 1/2, 2/2 l/l, l/2,2/2 ll], “2,
2/2
Cocker Spaniel 7 7, 0, 0 0, 0, 63 0, 0, 6a 1, 3, 23
Greyhound 8 8, 0, 0 0, 0, 8 0, 0, 8 2, 2, 4
Doberman Pinscher 25 3, l3, 9 0, 0, 25 0, 0, 5 4, l, 0
Siberian Husky 9 3, 5, l 0, l, 7" 0, 0, 9 0, l, 8
Labrador Retriever l l 2, 5, 4 0, 2, 9 0, 0, ll 1, 9, l
Collie 9 8, l, 0 4, 5, 0 0, 3, 6 2, 3, 4
Scottish Terrier l l 2, 0, 9 0, 0, ll 2, 5, 4 l, 2, 8
German Shepherd 9 0, 3, 6 0, 0, 9 0, 0, 9 l, 6, 2
Beagle ll 11,0,0 0,1,10 0,0,1] 0,6,5
Pointer 8 8, 0, 0 l, 0, 7 0, 0, 8 5, 3, 0
Total 108 52, 27, 29 5, 9, 92 2, 8, 77 I7, 36, 34

 

a. One DNA sample failed to amplify in this group.

94

Appendix 35

a
Summary of SNPs identified vs. nucleotide and amino acid conservation in loci studied .

 

 

Gene % Identity % Identity SNPs Identified
Amino acid nucleotide Dog Horse Cat Ox Pig

ACTC 100 I 95 3 - - - -
RBC 95 92 0 - - - -
RYR2 99 92 l - - -
WT] 96 91 00 l 3 0 O
GNATl 99 89 l - - - -
TS 89 89 10 2 0 l -
NCAD 88 88 l - - -

FES 80 87 0 0 - 0 3
CYP1A1 80 83 O - - -
C5 79 80 0 - - - -
CFTR 8] 80 4 2 0 l“ 0
RDS 91 76 0 - - - -

 

a. All SNPs were confirmed by restriction digestion except for six contained in a STNE in the canine TS
gene, one horse CF TR SNP, and two pig FES SNPs.

b. The table is sorted by descending percentage of nucleotide sequence identity for human compared to
rat or mouse.

c. Contains a microsatellite.

95

Appendix 3-6

 

Identification of four SNPs in the canine TS gene by pool-and-sequence. Arrowheads indicate the identical
location of four SNPs in a single animal and the pool of ten animals from different breeds as seen in
sequencing ladders using the downstream (panel A) and upstream amplification primers (1 = TSOl , 2 =
T802, 3 = TS03, 4 = TS04). Approximate allele frequencies can be visually estimated fi'om the relative
intensities of the two bands seen in the pool. Arrowhead I shows the position of A and C bands in the
heterozygous single animal. The C overlies a naturally occurring Mwol restriction enzyme site. The C
band also appears weakly in the pool. Arrowheads labeled as 2 show the position of an insertion or
deletion (indel) of a single base in the pool and heterozygous single animal. Note the doubling effect of
bands above this point. Shadow bands could be faintly seen in the pool on the original film. One indel
allele overlies a naturally occurring Msel site. Arrowhead 3 shows two alleles in the pool (C and T) which
is homozygous (T) in the single animal. The C allele overlies a naturally occurring Pstl site. Arrowhead 4
shows the location of a SNP found using the downstream primer. The relative location of these SNPs is
shown in Fig. 3.

96

Appendix 3-7

B
Single Pool Single Pool

GATCVGATC GATC GATC

m.

    

<—3

 

Identification of four SNPs in the canine CFTR gene. The arrowheads in panel A indicate SNPs CFTROI
(arrowhead I) and CFTR02 (2). These are two of the three coding region SNPs found in this study. The
arrowheads in panel B indicate SNPs CFTR03 (3) and CFTR04 (4). For CFTR04, note the characteristic
fall off in intensity in the A lane at the position of the SNP relative to the other bands in the A lane.
CFTR03 is a second example of an indel SNP.

97

Appendix 3-8

B L S C Q L Y Q R S G D
TGCCAGTTCTATGTGGTGAANNGTGAGCTGTCCTGCCAGCTATACCAGAGGTCAGGAGAC 6O

 

MGLGVPFNIASYALLTYMIA
ATGGGGCIGQQCGIGCCCTTCAACATCGCCAGCTACGCCCTGCTCACCTACATGATCGCA 120

H I T G L K V
CACATCACAGGCCTGAAthgggctgctctcaggcaagaacggctgctgccagccgaaag 180

A
GCNNNNNNNGC MWO I
V V

CACTGAGCGCTTAGATTATTTAATGTGGAAAGACTGACTGCAGAGAAGTCAGTGTCTGAT 240

TTAA MSE I CTGCAG PST I
tagttttaaaatgtgacgttagaaatctagttggtcacctggcaagcttccaggaagcta 300

tcctgtctttgtcgtttgtataacaaaca(100 bp)gtttttattttacattaaaagat 460

tRNA SINE V V V V
tttatttatttattcatgagagacacagaaagaggcagagacataggcagagggagaagc 520
-- c g c a
V

AGGCTCCATGCAGGGAGCCTGATGCGGGATTCCATCCTAGGACCCCAGGATCATGACCTG 580
C

AGCCAAAGGCAGAgggjgggggggTGAQQQAQQQAQQQGCCCCCATAATGTGCTngaAA 640
v

TGATCTGTCTTAQATTAAAACTAAAACATTTACCATTTTGCTTGTTCTGGTGGGATCAAG 720

g
TTAA MSE I

P G D F I H T L

CCAGGTGACTTCATACATACTTTGGGAGATGCACATATTTACCT 764

 

Location of four SNPs identified in the canine TS gene. The PCR primers are underlined. The deduced
amino acid sequence is shown above the nucleotide sequence. The invariant GT and AG splice signals are
shown in bold italics. SNP alleles are identified with a triangle (V) above the SNP, with the diagnostic
restriction site shown underneath each SNP. A SINE is shown in bold. Double-underlined sequences are
invariant Msel sites and dash-underlined sequences are invariant Mwol sites.

98

 

.555 mm 5.. ...... 553 8555.. MD.— ..< .o

5.55 .555 5.. 28-03. 5... 5. 55> 58m .555... 5555 o... 5. 53m 5.... 58m 5
.. .55...
5:5. FEES SE. ...... 5555...? 55.5.8»... ..E 95.55.. 5555b 68.56 .353. 5.55.5... .Q5 ”.55.... Be... 55555.. .55.. we .5555
.85.. .QVUZ .. 8...... 52.35... ...N 3,5 55525-05... 55> 859.8 55.5.. o... .5 2505.5... 5.2.8 .33 5.5.0.5... we... omv ... 55955.5

438.9 55.35.. 555855.58 85.5... 6.55 .255 mm .55.... 525.585 RU .55.. 685.56 8...... ....ka 5.... .5. o... 5 5.505 855... ...

 

 

05. ...... 5 85> Nb .mm .mm m .N .. cow HUUHO<OH<HOHUU<HE9m <OHOHO<UU<H<UU<<<O<pm .55
05. ....w 5 85> Nb .3 .3 m .N .. can UUHUH<U©HOH<H<<<EOO<R <<OHOOHOH<EUEO<UUO5W m...
.55. ......w Nb ...m .3 . .. .. c8. :UH<UEOH0:OBOHOH% O<H<OOOHOHUH<UEUOH5m N55.
v
.55.. 8.... N.. ...m .3 .m.. .m... omm <<O<HOP<O<<OHOOOHOOH5 EU<UUHOOHOUEOH<U<HE GEM
.55. 8.... Nb ...m .3 m .N .. cmm OHUHEOHU<UUH<09005~5 HOOOUO<<O<<O<UU<OE% may.
.55.. 8.? Nb Km .3 m .. .. coo. <OO<<UO<<O<UHOE<UPW <UE<<OOO<<UUO<E<UPW Q<OZ
oooN 85>
c5 85:55.5 N.. ..e .3 . .. .. m3. UZUUHO<OUUU<E<OOHU$ <O<UO<<UHOUH<UU<UO<5 3.4.20
v
55. ....w 5 85> N» Km .3 .m... m... owv OOOH<O<HOH<OU<OH<UUH5 ECHO<<OUOOEU<<OOOPW mm...
v
55. ...a 5 85> Nb Km .3 .m... m... coo :OOHO<UUOHU..<O.E.OO..% OOHUO<OOEUHUU<OOEk .556
v
05. ...e 5 85> Nb .5 .3 .m. .2. oNo. ECHOHUUH<HUEUOHH<Pm <UO<<U<UUOOH<UUO<<H05 FED
. v .
55. ...e 5 85> N.. ...m .3 .m... m... o2. 0HOEU<O<U<UEUUHUPn <OE<H<OOUUE<OU<EOHO5m mo
.55.. 8.... Nu Km .3 m .N .. co: <0<HOOO<UU<E<<UO<UH<OU<5 P<O<OH<<O<OEH<OOHUUUPm 0H0...
9...... ......
6 av 2...... 55 2.5..
855.5... 5.88585... 0 . 5 . N 5.5.... . 5.5.... a

 

$55.68 55 5565.95 won. 5.. .55 858cm 5.5-35

a. 888.5

99

.82.»... 5.. .5. o>.............< ...
655.50.... 5.5 5.5:“ 25 5,. 5.... 5.5....5. 255 o. .55.... 5.53 5... 52.5.2.2
..N 5.5.... 8.55.2.5 50.558-05-52. ......w.5 o... 55....2. 0. .55 55... .5... 255...... o... 5... 353 55.5 2.558.. 55... $5 255 5.5... ...

 

 

3.3.3 . . . 23 3200500555553 555505505053 8...... 30
3.3.3 5.. 3 35050003550053 03003550525553 8.... E5 ..0
3.3.3 3.. 8. 53005003055553 5305505505053 8...? 38..
3.3.3 3.. 5. 55555053552553 05500525030553 .350 83..
3.3.3 ..... 3. 0.005550050309053 00550550505053 85 .6
3.3.3 2.. .3 05505505050503 <00<<0<0000H<000<§03 8.55 .6
3.3.3 .... 3. 005550555553 205505505505. 8.85 3..
3.33 E. m . m 05000555550053 05500555050053 .8905. won.
3.3.3.3 ... . . .. 93 5050050555503 305505053355... 8052 won
3.5.3 .... o8 03050005005253 <00<<0e00h<00<00<0<3 8.5.20 30
3.5.3 .... 8. E055025E00E<03 5.5050050335030903 3.30 woo
3.3.3 .... 5. 50905055500553 5005550555053 855 3..
3.3.3 5.. o3 50.05025500553 05505505055003 855 30
3.3.3 ..... w: <0<e000<00§<<00<0~<00<3 55005050505053 805... won
333 .... .3 05055203055553 20555555050003 S05...~ 30
Go. .....5 a...
.55... 2...... 5% N .55.... . .55.... 500.. 5.0on

 

mmZm w....5. 5.. 355.... 0.32.55
5

o.-. 55......

100

Chapter 4
Within-Breed Heterozygosity of Canine Single Nucleotide
Polymorphisms Identified

By Across-Breed Comparison

10]

Within-Breed Heterozygosity of Canine Single Nucleotide Polymorphisms Identified

By Across-Breed Comparison

J A Brouillettel’z’4, P J Vental'2’3’5

Microbiology and Molecular Genetics‘, Genetics Programz, Small Animal Clinical
Sciences3, College of Human Medicine‘, and College of Veterinary Medicines, Michigan

State University, East Lansing, MI 48824-1314

Correspondence:

Patrick J. Venta, Ph.D.

Small Animal Clinical Sciences
College of Veterinary Medicine
Michigan State University

East Lansing, MI 48824-1314
Phone: 1-517-432-2515

Email: venta@cvm.msu.edu

 

102

Summary

Identification of single nucleotide polymorphisms (SNPs) by DNA sequence comparison
across breeds is a strategy for developing genetic markers that are useful for many
breeds. However, the heterozygosity of SNPs identified in this way might be severely
reduced within breeds by inbreeding or population subdivision. The effect of inbreeding
and population subdivision on heterozygosity of SNPs in dog breeds has never been
investigated in a systematic way. We determined the genotypes of dogs from three
divergent breeds for SNPs in four canine genes (ACTC, LMNA, SCGB, and TYMS)
identified by across-breed DNA sequence comparison and compared the genotype
frequencies to those expected under Hardy-Weinberg equilibrium. Although population
subdivision significantly skewed allele frequencies across breeds for two of the SNPs, the
deviations of observed heterozygosities compared to those expected within breeds were
minimal. These results indicate that across-breed DNA sequence comparison is a

reasonable strategy for identifying SNPs that are useful within many canine breeds.

Keywords: SNPs, Beagle, Doberman Pinscher, Scottish terrier, DNA pooling, inbreeding

coefficient

103

Introduction

Purebred dog populations are affected by hundreds of different genetic diseases, many of
which are homologous with human genetic diseases (Patterson 2000). The canine
diseases are the bane of breeders and owners alike, while at the same time they present
useful models of human genetic diseases. The same methods used to find human disease
genes and mutations can be used to find canine disease genes and mutations (e. g.,
Yuzbasiyan-Gurkan et al, 1996). The most comprehensive method is linkage analysis,

which requires the availability of genetic markers spread throughout the genome.

Single nucleotide polymorphisms (SNPs) have become more readily accessible for use in
human mapping projects, and are thought to have good potential for very high throughput
automation (Brookes 1999; Shi, 2001). A nucleotide diversity estimate in dogs suggests
that the number of SNPs present in the dog genome is similar to the number found in the
human genome and could be used for automated canine genotying (Brouillette et al,

2000). Unfortunately, few canine SNPs have been identified to date.

We recently developed a simple method to identify SNPs in any mammalian genome that
can help to increase the number of SNPs for use in canine genetic studies (Brouillette et
al. 2000). A pool of DNAs from different dogs is manually sequenced, after which SNPs
are identified as bands occurring at identical positions in the sequencing ladder. In order
to find SNPs that are useful for as many breeds as possible, we pool DNA from single

representatives of ten diverse breeds. However, there is concern that the SNPs found

104

using a diverse pool will be of little use within dog breeds because variability will have

been greatly reduced or lost due to inbreeding or population subdivision.

In order to test the hypothesis that these factors prevent useful SNPS from being found by
pooling DNA samples across breeds, we determined over 500 genotypes using four of
these SNPs in 42 to 49 individual dogs from three different breeds, and tested for
significant departures from Hardy-Weinberg equilibrium predictions. We also calculated
the inbreeding coefficients (a measure of the reduction of heterozygosity) for each breed-

SNP combination.

105

Materials and Methods

The dogs genotyped were Beagles, Doberman pinschers, and Scottish terriers. They were
unrelated for at least three generations, and should be a reasonably random sample of
each breed. Some of the Beagles used were part of the DoGMap REfference Panel.

Other dog DNA samples came from other projects in our lab.

DNA was isolated after collection of whole blood from dogs or from buccal cells
(Brouillette et al. 2000; Richards et al. 1992). SNPs were selected to give a broad
spectrum of allele frequencies as identified from a pool of single dog DNAs (Brouillette
et al. 2000). Samples were genotyped for SNPs in the following genes: TYMS, SGCB,
ACTC, and LMNA. Primers, amplification conditions, and restriction enzymes used for
genotyping are given in Appendix 4-1. Two of the four SNPs were previously reported,
including small genotype surveys (TYMS SNP T804, and ACTC SNP ACTC02 in
Brouillette et al. 2000). Allele frequencies in the ten-breed pool for these two SNPs were

0.50 and 0.20 (allele 2), respectively.

The cross-species primers for LMNA were 5’-ATCGCATCGACCTCCTCTC and 5’-
AGGTCCTGGGACATGGCTGG. The LMNA SMP is a G/A transition located in intron
5 (position 174 in Genbank accession no. AF 427092) and , in a survey of ten single dogs
of different breeds (the same dogs used in the pool to identify the SNPs), allele 1

occurred at a frequency of 0.10. The cross-species primers for SGCB were 5’-

ATTGGACCAAATGGCTGTG and 5’-GTCCTCGGGTCAAAAAACT. The SGCB

106

SNP is a T/C transition located in intron 4 (position 265 in Genbank accession number
AF 427093) and, in a survey of the same ten single dogs, allele 1 occurred with a

frequency of 0.45.

107

Results and Discussion

The inbreeding coefficient was calculated using the equation F = (He — Ho)/Ho, where He
is the expected number of heterozygotes based on Hardy-Weinberg equilibrium and H0 is
the observed number of heterozygotes (Juneja et al. 1981). Among breed comparisons
was calculated in the same manner combining data for all three breeds for each SNP.

Chi-square values were used to determine significance.

Genotypes are given in Appendix 4-2. A few amplifications failed for some of the breed-
SNP combinations, so that the number of dogs typed per breed is not always the same for
the different SNPs. There were no cases in which there was a statistically significant
difference between observed and expected genotypes based on Hardy-Weinberg

equilibrium for any SNP within a given breed (Appendix 4-2).

Under the assumption that inbreeding is actually present but not statistically detectable,
we calculated inbreeding coefficients. They ranged from a high of 0.143 for the SGCB
SNP in Beagles to a low of -0.190 for the same SNP in Doberman pinschers. The
average inbreeding coefficient for Beagles was 0.047 for the two markers for which
inbreeding coefficients could be calculated, for Doberman pinschers was -0.044, for the
four SNPs studied, and for Scottish terriers was 0.019 for the three SNPs that were

variable in this breed.

108

The LMN A SNP showed variation only in the Doberman pinscher (allele 1 at 0.10). The
ACTC SNP had allele 2 frequencies of 1.00, 0.38, and 0.67, and the SGCB SNP had
allele 2 frequencies of 0.56, 0.84, and 0.44 in Beagles, Doberman pinschers, and Scottish
terriers, respectively. The TYMS SNP had allele 1 frequencies of 0.39, 0.81, and 0.30 for
the above breeds respectively. For the three breeds combined into one population,
significant deficiencies of heterozygotes were seen for two SNPs (ACTC F = 0.298, P =
0.002; TYMS F = 0.223, P= 0.032), but not for the other two (SGCB F = 0.099, P = 0.51;

LMNA F = -0.018, P = 0.98).

If the average inbreeding rates were high within breeds, correspondingly greater numbers
of markers or dogs would be needed to obtain statistically significant linkages in genome
scans. The inbreeding coefficients for the breeds found in this study suggest that I
inbreeding has only a very minor impact on the observed heterozygosity for SNPs.
Population subdivision appears to have a greater impact on the variability of SNPs across
breeds, but reasonable variability was still found within breeds in many cases (Appendix
4-2). This loss in heterozygosity across breeds is not unique to SNPs, but is shared with
other commonly used marker types as well (Koskinen and Bredbacka, 2000; J uneja et al.
1981). Although fiirther studies should be performed, we conclude from the present data
that population subdivision and inbreeding will probably not preclude the use of pooling
across breeds to find SNPs that are useful for studies within breeds.

An important implication of this study is that it should be possible to predict the number
of heterozygotes within a purebred canine population with reasonable accuracy directly

from allele frequencies under the assumption of Hardy-Weinberg equilibrium. Because

109

allele frequencies for SNPs can be determined within a few percent in a pooled DNA
sample by various methods, the utility of a given SNP can be deduced by typing one
pooled s\arnple before a decision is made to type that marker in numerous individual dogs
(Shi 2001). The identification of more canine gene-specific SNPs in the future by across-
breed comparisons will make it possible to take advantage of the automated genotyping
systems for mapping and ultimately identifying the genes that cause canine genetic

diseases.

110

 

Acknowledgements

We thank Mr. Mark Stinson for help in obtaining some of the samples used in this study
and Dr. Donna Housley for helpful suggestions. This work was supported by the Morris
Animal Foundation, the Doberman Pinscher Foundation of America, Inc., and the
American Kennel Club Canine Health Foundation with co-sponsorship by the Westie
Foundation of America, the Health Trust Fund of the Scottish Terrier Club of America,

and the Cairn Terrier Club of America.

111

gov—EB £5 .5.“ :08 83 203 3:3 “5:33: _So>om .0
gene as a N _Baaz . 28$ .3 e 32:35 a _ “35:2 .m

.on :fimcoo wfiueommotoo 5 528: 02820:: 380:2: 35:32 .v

.3 N _ -v 8: 239—an2 Buss—E82
{238358 2: «a and .ES 3 mo 5:828:00 3:: a 8 N53: mo 8:63 2: 5:: than mo: 5 So 3E8 953 maoumowE 25.20 5:850: .m

.55 m 33 2:: :owmeoio 2: :5 as N 33 2:: wéaoecm 2: wit. 5.: Ba .0 owe me; <22: 8.: 232383 3:325 05 E85 A8353 <75 38:3
335 cm cc @2958 <20 323 335 mm .5: 5:: m .0 ONN .EE _ .0 arm .58 _ .0 ova 303 £53280 3:96 dogma.— .1 N .5: omEofimfim <ZQ 3H
3 no 23 SE “035 w: an; 5.5a 58 .85 2 EH2“. :1 2: .8»: 26 2 OSN a 3:3 6:5 2:. 2 :8. 25 ow 22» 22:28 mom .N

.395 .5: m8: .853me Snob 8 won: woman cocoa—m 38:2: 8%: 98 Ben E 3884 .—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

uoeoa<uo€a§<<§oo<sm
_ NBSN: _NN .30 m .3 0N0 _az fiepooeozeutofiuofia WE:
uooEe<<B<oEuEB<um
N 82?: SN .30 ON .NN_ N: e5: S<Eo~<<ou§<uo$<<sm 80m
ooSooEuéooSBoofm
N 8332 E .30 ON .2: E :38 BBBBEoquooBfih <23:
563208305058;
_ 2°8an NM Eu 2 .5 N: :38 Cucoe<<<o<uutguu<um u5<
. omsfiu thcm
0: EommmDQON V M
00:20.3”: muucozaow BEEN:
m €350 “28282 as N 22? ea _ 222 8:25; N; . 32:

 

 

 

20:32 me can defiance 3:058.— .EoEtm osmocwaa

E. 5283

112

Appendix 4-2

Genotypes for dogs from three breeds using four SNPs.

 

 

Locus Breed Genotype [Expectedl Observed p
ACTC Beagle 1/1 0 0
1/2 0 0 NA3
2/2 42 42
Doberman Pinscher 1/1 19.0 19
1/2 23.6 24 0.99
2/2 7.2 7
Scottish Terrier l/ l 5.3 6
1/2 21.3 20 0.86
2/2 21.3 23
LMNA Beagle 1/1 0 0
1/2 0 0 NA
2/2 42 42
Doberman Pinscher U] 0.1 0
1/2 4.8 5 0.95
2/2 45.1 45
Scottish Terrier 1/ l 0 O
1/2 0 0 NA
2/2 47 47
SGCB Beagle 1/1 8.5 10
1/2 21.0 18 0.66
2/2 12.7 14
Doberman Pinscher l/ l 1.2 0
1/2 12.6 15 0.41
2/2 33.2 31
Scottish Terrier 1/1 14.7 14
1/2 23.1 25 0.85
2/2 9.1 8
TYMS Beagle 1/1 6.4 6
1/2 20.0 21 0.95
2/2 15.6 15
Doberman Pinscher U] 32.] 32
1/2 15.1 14 0.94
2/2 1.8 2
Scottish Terrier 1/1 4.4 4
1/2 20.6 19 0.88
2/2 24.0 25
1. Under Hardy-Weinberg Equilibrium.
2. Chi-square probability.
3. NA = not applicable for chi-square analysis because expected cells are 0.

113

Chapter 5

Summary

114

Summary

In this thesis I have; (1) developed a method to produce genetic markers to test
hypotheses regarding specific candidate genes as underlying traits and diseases in dogs
and other domestic animals, and (2) shown that the markers developed across canine
breeds are useful for linkage studies within breeds. Each of the methods will be briefly

discussed in turn.

This process began with the design of canine gene-specific primer sets and establishing
sequence tagged sites for several genes within the canine genome. For the work found in
chapter two, cross-species PCR primer sets, referred to as UMSTSs, were developed for a
total of 86 genes. Through additional work, UMSTSs have been developed for over 1100
genes (Chapter 4, Chapter 3, Venta and Vidal, 1999, Brouillette and Venta, 2000;
Brouillette et al., 2000; Housley et al., 2006). The UMSTS were designed to optimize
their usefulness across a wide spectrum of mammalian genomes. DNA was amplified
from a set of ten different mammalian species with a subset of 11 primer pairs, and the
amplified PCR products were sequenced to confirm that the PCR products were the
species-specific orthologs of canine gene of interest. This verified the usefulness of these
primer sets across mammalian species. Others have shown the utility of these primer
pairs in the identification of genes in the equine and porcine genomes (Shubitowski et al,

2001; Farber et al., 2003).

Gene-specific sequence tagged sites could be used in a variety of ways. One way was to

use the gene-specific amplification product to screen canine DNA libraries for the

115

 

presence of clones containing the gene of interest. The genes of interest could then be
isolated from the library and a search made for SSLP markers within or near the gene of
interest. The cloned gene could also be directly sequenced to yield nucleotide

information about the particular mammalian gene.

With the development of a radiation hybrid map in the canine genome (Vignaux et al.,
1999, Guyon et al, 2003, Housley et a1, 2004, Breen et al., 2004), the primer pairs could

be used to determine physical linkage groups and chromosome locations of specific genes

 

within the canine genome.

At the time of this work, a limited amount of DNA sequence data was available for the
canine genome. With the availability of the complete canine genome sequence
(Lindblad-Toh et al., 2005), there is no need to design cross-species PCR primers for
canine genes. However, the primers are still useful in other mammalian species when the
species—specific nucleotide sequence of a gene of interest is unknown (eg. Brinkmeyer-

Langford et al., 2005; Aitken etal., 2004).

The next step was to identify and isolate markers within specific genes within the canine
genome. Our collaborator, Dr. Vilma Yuzbasiyan-Gurkan, has developed several
hundred randomly occurring SSLP markers within the canine genome, which would be
even more useful if they were linked to gene-specific markers, so that a comparative
positional candidate gene approach could be realized. Because of the time and labor

involved in isolating clones for the identification of SSLPs, it seemed that identification

116

of SNPs within the amplified intervals would be more efficient, provided that the canine

SNPs were dense enough that many of the amplicons would contain a SNP.

The pool-and-sequence technique was developed to identify SNPs within genes of
interest (Chapter 3). It was observed that individuals that were heterozygotes at a given
nucleotide location in a gene being sequenced had two bands at a given position on the
sequencing ladder that were about half of the intensity of the adjacent sites with single
bands. I began experimenting with this observation to see if other ratios of nucleotide
frequencies at a given position in the sequencing ladder would be indicated by
corresponding changes in the intensities between bands in the sequencing ladder. Both a
qualitative and a quantitative method to determine allele frequencies were developed, if

indeed a polymorphism existed in a given DNA sample being sequenced.

This technique was used to locate polymorphisms within the canine genome that would
be heterozygous across multiple breeds. A pool of DNA was created from a panel of ten
dog breeds. These dogs were chosen based on the assumption that if the breeds were
phenotypically different enough to be placed into different breeds by The American
Kennel Club, they would also be different at the nucleotide level. Objective evidence by
another lab has confirmed that this group of dog breeds is broadly representative of all
other dog breeds (Parker et al., 2005). Taking the ability to detect alleles using the pool-
and-sequence method and the ability to design primers for most canine genes based on
the nucleotide sequence of known index species (Chapters 3 and 2), a search for SNPs

was made in 12 canine genes. From this work, an estimate of nucleotide diversity was

117

 

made for the canine genome (Chapter 3). The nucleotide diversity was high enough that
a SNP with a minor allele frequency of 0.2 was identified in all the genes studied.
Similar frequencies were found in the other mammalian species investigated, suggesting
that it is feasible to use this method to identify a SNP marker in virtually any candidate
gene and any mammalian species of interest. It may, however, be necessary to sequence
additional areas of a gene of interest if a SNP was not found within the first area
amplified. These markers would then be used to test the hypothesis that a given
candidate gene was associated with a given disease phenotype in a breed. This work also
showed that the frequency of occurrence of SNPs in the canine genome is likely to be
sufficient to perform genome-wide scans for linkage in the canine genome. This
conclusion was borne out by the work of Lindblad- Toh et al. (2005). Finally, the work
in chapter 3 provided the first estimate of nucleotide diversity and frequency of SNP

occurrence in the canine genome.

Currently, a set of SNPs is being established throughout the canine genome (Lindblad-
Toh et al., 2005), and a rapid scan of the entire canine genome will soon be possible for
linkage or association studies using DNA arrays. It is estimated that approximately
10,000 evenly-spaced SNP markers would be necessary for association studies in the
canine genome (Ostrander and Kruglyak, 2000; Sutter et al., 2004; Lindblad-Toh et al.,
2005). The recent development of commercial arrays of 20,000 SNPs may make scans

for association in the dog a practical reality (Lindblad-Toh, 2006).

118

The conserved synteny between canine, human, primate, and murine genomes is
approximately 94% (Lindblad-Toh et al., 2005). This synteny immediately leads to
comparative positional candidates for disease genes if a mutation causing a disease
phenotype in humans or mice is mapped to a known location within the genome. The
area of interest in the canine genome will correspond to the syntenic region of the human
or murine genome where a candidate gene has already been mapped for a given disease
phenotype of interest. Markers mapping to genes in that region could be immediately

tested for association or linkage.

 

The pool-and sequence-technique allowed us to explore how much diversity would be
found in the canine genome. Based on sequence data that incorporated fragments of 12
canine genes and 5.4 kbp of DNA, an estimate was made of how frequently a SNP would
be found in the canine genome, approximately once every 400 nucleotides. It was
estimated that the nucleotide diversity for exons and introns is 0.001 and 0.004,
respectively in the canine genome (Chapter 3). These values are similar to what is found
in the human genome (Collins et al., 1997; Wang et al., 1998), and are similar to the
estimate of Lindblad-Toh et al. (2005). Lindblad-Toh et. al. compared nucleotide
sequence data between a single Boxer and a single individual from various other dog
breeds and found that the rate of SNP occurrence was 1 in 800 nucleotides (Lindblad-Toh
et al., 2005). Preliminary estimates on a smaller data set in the feline, bovine, porcine,
and equine genomes suggest similar nucleotide diversity in these species (data not
shown). For the previously reported work (Chapter 3), the limit of detection for the
minor allele is about 5% allele frequency. Other studies indicate that a detection limit of

around 1% allele frequency for the minor allele is possible in our hands (Data not

119

shown). The manual pool-and-sequence method was used because it was considerably
less expensive than automated methods at the time these experiments were performed.
The relative cost of the two methods has changed and automated methods are now less
expensive. Automated sequencing is recommended for SNP discovery in non-sequenced

genomes.

A method of PCR mutagenesis was used to create restriction sites for endonuclease

digestion that will preferentially cleave DNA with one or the other allele at the SNP site.

 

This became a rapid and technically reliable means to type the SNPs that had been
previously identified. Although primer mutagenesis remains an option for small scale

studies, it would not be appropriate for large studies.

A final study was undertaken to determine if SNPs found across multiple breeds would
be useful as markers for within-breed linkage or association analysis. This information
would also be used to estimate coefficient of inbreeding by means of a reduction in

heterozygosity within breeds.

It was known that population subdivision would decrease the heterozygosity of the
individual population divisions within a population taken as a whole. This is known as
the Wahlund Effect. If a population is divided into smaller units and breeding within
those units are restricted, there will be a decrease of heterozygotes among the entire
population due to the subdivisions, assuming that the alleles under study have different
frequencies in the various subpopulations. An example of this effect is given in

Appendix 5-1.

120

In this study, the genotypes were determined at four different SNP markers using 42-49
individuals fi'om three different dog breeds. It was determined that population
subdivision does significantly skew allele frequencies, at least in the breeds studied, and
the coefficient of inbreeding ranged from 0.297 to -0.018. These values are in close
agreement with other canine breeds. For example, Dom and Schneider (1976) calculated
inbreeding coefficients for 14 breeds of dog which varied from a low of 0.009 in Boxers
to a high of 0.1 17 in German Shepherds. The median value across all 14 breeds tested

was 0.18.

As a practical matter, there is sufficient heterozygosity of SNP markers within breeds for
genetic studies to be undertaken without a significant increase in the number of animals
or markers required for detection of linkage or association. The coefficient of inbreeding
will be affected by a number of factors including popular sire effects, recent population
bottlenecks, the age of the breed, and the population size of a given breed (Dom and

Schneider, 1976, Kathmann et al., 1999, Mandigers et al., 1993).

If these results can be generalized across breeds, the major implication is that SNP
markers developed across breeds will be useful within breeds to a large extent, though the
utility of breed-marker combinations will vary among the population of interest. This
data supports the feasibility of the development of a panel of SNP markers across the
canine genome to be used as part of a DNA array. The resulting array will have utility
within many breeds for traditional linkage studies as well as candidate gene and allele
association studies, with varying levels of usefulness depending on the breed under study
and the markers used in the whole-genome analysis (Sutter et al., 2004; Parker et al.,

2005)

121

 

 

The future of canine genomics and genetics is bright. The identification of SNPs within
genes will allow for an increasingly dense canine linkage map (Mellersh et al., 2000),
with more and more disease causing mutations being mapped in the coming years (Giger
et al., 2006). Coupling this with the physical map (Mellersh et al., 2000; Guyon et al.,
2003, Housley et al., 2004) will allow for increasingly detailed mapping of genes. DNA
array technology will allow for higher throughput of linkage and association studies and
further increase our knowledge of the canine genome as well as our understanding of
mammalian evolution. The structure of the canine genome and pattern of breeding will
make genome-wide linkage disequilibrium studies feasible in the near future, given the
large number of SNP markers available today (Sutter et al., 2004; Ostrander and

Kruglyak, 2000; Lindblad-Toh et al., 2005).

The first draft of the canine genome is now available with the identification of over 2.5
million SNP markers (Lindblad-Toh et al., 2005). The influx of data from this project
will expand our knowledge of canine genetics as well as comparative genetics. There
will likely be an explosion of information as disease-causing mutations are identified at
an increasingly rapid rate. Breeders can use this information to improve overall breed
health by selective breeding practices. This will lead to a better life for dogs and their

owners alike.

122

Appendix 5-1
Brief Explanation of the Wahlund Effect

The Wahlund Effect is an observation that when a population is divided into subpopulations, there is an
increase in the frequency of homozygotes relative to what would be expected by Hardy-Weinberg

Equilibrium, even if the two subpopulations are each in Hardy-Weinberg Equilibriuma’ . This is a

generalizable rule among populations which are stratified into subpopulations as long as the allele
frequencies for a given gene are not equal to one another in the subpopulations.

As an example of this, consider a population of two hundred individuals divided into two subpopulations,
each with 100 individuals. Assume that the subpopulations are each in Hardy-Equilibrium. For a gene, A,
there are two alleles, “A” and “a”. In subpopulation l, the allele frequency of A is 0.7. In subpopulation 2,
the allele frequency of A is 0.4.

 

By assuming the subpopulations are each in Hardy-Weinberg equilibrium, we can conclude that the allele
frequency of a is 0.3 and 0.6, in subpopulations 1 and 2 respectively. We can use this information to
calculate the occurrence of the various genotypes for gene A in each subpopulation.

Based on Hardy-Weinberg Equilibrium, the frequency of occurrence of the three genotypes in the two
subpopulations is given below.

Subpopulation l Subpopulation 2

Genotype No. of Individuals Genotype No. of Individuals Total
AA 49 AA 16 65
Aa 42 Aa 36 78
aa 9 aa 48 57

The allele frequency of allele A for the entire population is calculated as 2(65) + 78/ 2(65 + 78 + 57) =
0.52

Given the frequency of allele A in the population is 0.52, the expected allele frequencies in the total
population is: '

(Expected): AA = 54.08 Aa = 99.84 a = 46.08

(Observed): AA = 65 Aa = 78 a = 57

123

This depression of heterozygotes is a mathematical phenomenon that can be generalized to any allele
frequencies pl and p2 in subpopulations l and 2, respectively, as long as the allele frequency in
subpopulation 1 is not the same as the allele frequency of subpopulation 2, that is, pl 96 p2.

If these subpopulations are in Hardy-Weinberg Equilibrium, the second allele in each subpopulation, q]
and q2, respectively, are related to one another by the following equation.

pl + ql = 1
p2 + q2 = 1
The frequency of the occurrence of heterozygotes in each subpopulation is:

2pqu and 2p2q2 respectively.

The Heterozygosity, H, for the entire population is simply the mean of the heterozygosities of each of the
subpopulations. Stated mathematically, this is:

H = 2p1q1 + 2p2q2
2

 

H=p1q1+p2q2
H=p1(1-p1)+p2(1—p2)

This value is always smaller than

p(l — p), which equals 2pq, unless p = q.

a. http://blackwellpublishingcom/ridley/a-z/Wahlund_effect.asp
b. http://statgenxbs.unipi.it/courses_file/documents/pdf/Holsinger-Wahlundpdf

 

124

BIBLIOGRAPHY

Aitken, N., Smith, S., Schwarz, C. and Morin, P. “Single Nucleotide Polymorphism
(SNP) Discovery in Mammals: A targeted-gene Approach” Molecular Ecology
13: 1423-1431 (2004).

Alcalay, M., Antolini, F., Van de Ven, W.J., Lanfrancone, L., Grignani, F., and Pelicci,
P.G. “Characterization of human and mouse c-fes cDNA clones and identification
of the 5' end of the gene”, Oncogene 5: 267-275 (1990).

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J ., Ahang, Z., Miller, W., and
Lipman, D.J. “Gapped BLAST and PSI-BLAST: a new generation of protein
database search programs”, Nucleic Acids Research 25: 33 89-3402 (1997).

Ameratunga, R., Winkelstein, J .A., Brody, L., Binns, M., Cork, L.C., Colombani, P., and
Valle, D. “Molecular analysis of the Third Component of canine Complement
(C3) and Identification of the Mutation responsible for Hereditary Canine C3
Deficiency”, Jounal of Immunology 160(6): 2824-2830 (1998).

American Kennel Club, “The Complete Dog Book”, 19th ed., Howell Book House, New
York, NY. 1997.

Asher, J .HJ . and Friedman, T.B. “Mouse and hamster mutants as models for
Waardenburg syndromes in humans”, Journal of Medical Genetics 27: 618-626
(1990)

Barendse, W., Armitage, S.M., Kossarek, L.M., Shalom, A., Kirkpatrick, B.W., Ryan,
A.M., Clayton, D., Li, L., Neibergs, H.L., Zhang, N., Grosse, W.M., Weiss, J .,
Creighton, P., McCarthy, F., Ron, M., Teale, A.J., Fries, R., McGraw, R.A.,
Moore, S.S., Georges, M., Soller, M., Womack, J .E., and Hetzel, D.J.S. “ A
genetic linkage map of the bovine genome”, Nature Genetics 6: 227-235 (1994).

Bassam, B.J., and Caetano-Anolles, G. “Automated “hot start” PCR using mineral oil and
paraffin wax”, Biotechniques 14: 30-34 (1993).

Bentolila, S., Bach, J .M., Kessler, J .L., Bordelais, I., Cruaud, C., Weissenbach, J ., and
Panthier, J.J. “Analysis of major repetitive DNA sequences in the dog (Canis
familiaris) genome”, Mammalian Genome 10: 699-705 (1999).

Bergenhem, N.C.H., Venta, P.J., Hopkins, P.J., and Tashian, R.E. “Variation in coding
exons of two electrophoretic alleles at the pigtail macaque carbonic anhydrase I
locus as determined by direct, double-stranded sequencing of polymerase chain
reaction (PCR) products”, Biochemical Genetics 30: 279-287 (1992).

125

 

 

Botstein, D., White, R.L., Skolnick, M., and Davis, R.W. “Construction of a Genetic
Linkage Map in Man Using Restriction Fragment Length Polymorphisms”,
American Journal of Human Genetics 32: 314-331 (1980).

Boyer, G., Nonneman, D.J., Shibuya, H., Stoy, S.J., O’Brien, D., and Johnson, G.S. “A
PCR-RFLP marker of the Erythroid Aminolevulinate Synthase Gene (ALASZ) on
Canine Chromosome X”, Animal Genetics 26: 206-207 (1995).

Breen, M., Hitte, C., Lorentzen, T., Thomas, R., Cadieu, E., Sabacan, L., Scott, A.,
Evanno, G., Parker, H., Kirkness, E., Hudson, R., Guyon, R., Mahairas, G.,
Gelfenbeyn, B., Fraser, C., Andre, C., Galibert, F., and Ostrander, E. “An
Integrated 4249 Marker FISH/fRH Map of the Canine Genome”, BMC Genomics
5: 65 (2004).

Breslauer, K., Frank, R., Blocker, H., and Markey, L.A. “Predicting DNA duplex stability
from the base sequence”, Proceedings of the National Academy of Science of the
United States of America 83: 3746-3 750 (1986).

Brinkmeyer-Langfore, C., Raudsepp, T., Lee, E-J., Goh, G., Schaffer, A., Agarwala, R.,
Wagner, M., Tozaki, T., Slow, L., Womack, J ., Mickelson, J ., and Chowdhary, B.
“A High-Resolution Physical Map of the Equine Homologs of HSA19 Shows
Divergent Evolution Compared to Other Mammals” Mammalian Genome 16:
631-649 (2005).

Brookes, A. J. “The Essence of SNPs”, Gene 234: 177-186 (1999).

Brouillette, J .A. and Venta, PJ. “T thl I I I PCR/RFLP marker in the canine rod transducin
alpha (GNATl) gene” Animal Genetics 31: 68 (2000).

Brouillette, J .A., Andrew, J .R., and Venta, P.J. “Estimate of Nucleotide Diversity in Dogs
using a Pool-and-Sequence Method”, Mammalian Genome 11: 1079-1086 (2000).

Buchanan, F.C., Adams, L.J., Littlejohn, R.P., Maddox, J .F ., and Crawford, A.M.
“Determination of evolutionary relationships among sheep breeds using
microsatellites”, Genomics 22: 397-403 (1994).

Buetow, K.H., Edmonson, M.N., and Cassidy, A.B. “Reliable identification of large
numbers of candidate SNPs from public EST data”, Nature Genetics 21: 323-325
(1999)

Burstein, E., Hogberg, J .E., Wilkinson, A.S., Rumble, J .M., Csomos, R.A., Komarck,
C.M., Maine, G.N., Wilkinson, J .C., Mayo, M.W., and Duckett, C.S. “COMMD
Proteins, A Novel Family of Structural and Functional Homologs of MURRl”,
Journal of Biological Chemistry 280 (23): 22222-22232 (2005).

126

Cargill, M., Altculer, D., Ireland, J ., Sklar, P., Ardlie, K., Patil, N., Lane, C.R., Lim, E.P.,
Kalyanaraman, N., Nemesh, J ., Ziagura, L., Friedland, L., Rolfe, A., Warrington,
J ., Lipshutz, R., Daley G.Q., and Lander, E. S. “Characterization of single-
nucleotide polymorphisms in coding regions of human genes”, Nature Genetics
22: 231-238 (1999).

Chee, M., Yang, R., Hubell, E., Berno, A., Huang, X.C., Stern, D., Winkler J ., Lockhart,
D.J., Morris, M.S., and Fodor, S.P.A. “Accessing Genetic Information with High-
Density DNA Arrays”, Science 274: 610-614 (1996).

Clark, A.G., Weiss, K.M., Nickerson, D.A., Taylor, S.L., Buchanan, A., Stengard, J .,
Salomaa, V., Vartianien, E., Perola, M., Boerwinkle, E., and Sing, C.F.
“Haplotype Structure and Population Genetic Inferences from Nucleotide-

Sequence Variation in Human Lipoprotein Lipase”, American Journal of Human
Genetics 63: 595-612 (1998).

Clark, L.A., Wahl, J .M., Rees, C.A., and Murphy, K.E. “Retrotransposon Insertion in
SILV is Responsible for Merle Patterning of the Domestic Dog” Proceedings of
the National Academy of Sciences USA 103(5): 1376-1381 (2006).

Clark, L., Tsai, K., Steiner, J ., Williams, D., Guerra, T., Ostrander, E., Galibert, F., and
Murphy, K. “Chromosome-specific Microsatellite Multiplex Sets for Linkage
Studies in the Domestic Dog”, Genomics 85: 550-554 (2004).

Clark, RD. and Stainer, J .R. “Medical & Genetic Aspects of Purebred Dogs”, Fairway,
KS: Forum Publications, Inc (1994).

Collins, A., Lonjou, C., and Morton, N.E. “Genetic epidemiology of single-nucleotide
polymorphisms”, Proceedings of the National Academy of Science USA 96 (26):
15173-15177 (1999).

Collins, D.W. and Jukes, T.H. “Rates of transition and transversion in coding sequences
since the human-rodent divergence”, Genomics 20: 386-396 (1994).

Collins, F.S., Guyer, M.S., and Chakavarti, A. “Variations on a theme: cataloging human
DNA sequence variation”, Science 278: 1580-1581 (1997).

Coltman, D.W. and Wright, J .M. “Can SINES: a family of tRNA-derived retroposons
specific to the superfarnily Canoidea”, Nucleic Acids Research 22: 2726-2730
(1994)

Day, A.A., McQuillan, C.I., Termine, J .D., and Young, M.R. “Molecular cloning and

sequence analysis of the cDNA for small proteoglycan II of bovine bone”
Biochemical Journal 248: 801-805 (1987).

127

Deschenes, S.M., Puck, J .M., Dutras, A.S., Somberg, R.L., Felsburg, P.J., and Henthom,
P.S. “Comparative mapping of canine and human proximal Xq and genetic
analysis of canine X-linked severe combined immunodeficiency Genomics 23:
62-68 (1994).

Dietrich, W., Katz, H., Lincoln, S.E., Shin, H.-P., Friedman, J ., Dracopoli, NC, and
Lander, E.S. “A genetic map of the mouse suitable for typing intraspecific
crosses”, Genetics 131: 423-447 (1992).

Dom, C. and Schneider, R. “Inbreeding and canine mammary cancer: A retrospective
study”, Journal of the National Cancer Institute 57:545-548 (1976).

Ellegren, H., Johansson, M., Sandberg, K., and Andersson, L. “Cloning of highly
polymorphic microsatellites in the horse”, Animal Genetics 23: 133-142 (1992).

Evans, J .P., brinkhous, K.M., Brayer, G.D., Reisner, H.M., and High, K.A. “canine
Hemophilia B Resulting from a Point Mutation with Unusual Consequences”,
Proceedings of the National Academy of Sciences of the United States of America

90(9): 3968-3972 (1989).

Farber, C.R., Raney, N.E., Rilington, D.E., Venta, P.J., and Ernst, C.W. “Comparative
Mapping of Genes Flanking the Human Chromosome 12 Evolutionary Break
Points in the Pig”, C ytogenetics and Genome Research 102(1-4): 139-144 (2003).

Farnir, F ., Coppieters, W., Arranz, J .J., Berzi, P., Cambisano, N., Grisart, B., Karim, L.,
Marcq, F., Moreau, L., Mni, M., Nezer, C., Simon, P., Vanmanshoven, P.,
Wagenaar, D., and Georges, M. “Extensive genome-wide linkage disequilibrium
in cattle”, Genome Research 10: 220-227 (2000).

Fisher, R.A., Putt, W., and Hackel, E. “An investigation of the product of 53 gene loci in
three species of wild canidae: canis lupus, canis latrans, and canis familiaris”,
Biochemical Genetics 14: 963-974 (1976).

Fitch, W.M. (1976) “The molecular evolution of cytochrome c in eukaryotes”, Journal of
Molecular Evolution 8: 13-40 (1976).

Francisco, L.V., Langston, A.A., Mellersh, C.S., Neal, C.L., and Ostrander, E.A. “A
Class of Highly Polymorphic Tetranucleotide Repeats for Canine Genetic
Mapping”, Mammalian Genome 7: 359-362 (1996).

F ujita, M., Loechel, R., McFarlin, K., Brewer, G.J., Venta, P., and Yuzbasiyan-Gurkan,

V. “Assignment and localization of a set of genes to canine chromosomes using
fluorescent in situ hybridization”, Mammalian Genome (in press).

128

 

 

Giger, U., Sargan, D., and McNiel, E. “Breed-Specific Hereditary Diseases and Genetic
Screening” in The Dog and Its Genome, Ostrander, E., Giger, U., and Lindblad-
Toh, K. Eds. Cold Spring Harbor Laboratory Press, Cold Spring Harbor NY,
2006.

Grimes, D.A., “Epidemiology of Gestational Trophoblastic Disease”, American Journal
of Obstetrics and Gynecology 150: 309-318 (1984).

Gu, X. and Li, W.-H. “Higher rates of amino acid substitution in rodents than in
humans”, Molecular Phylogeny and evolution 1: 211-214 (1993).

Halliday, R. and Grigg, G. W. “ DNA methylation and mutation”, Mutation Research
285: 61-67 (1993).

Halushka MK, Fan J -B, Bentley K, Hsie L, Shen N, Weder A, Cooper R, Lipshutz R,
Chakavarti, A. “Patterns of single-nucleotide polymorphisms in candidate genes
for blood-pressure homeostasis”, Nature Genetics 22: 239-247 (1999).

Harding, R.M., Fullerton, S.M., Griffiths, R.C., Bond, J ., Cox, M.J., Schneider, J .A.,
Moulin, D.S., and Clegg, J .B. “Archaic Afiican and Asian lineages in the genetic
ancestry of modern humans”, American Journal of Human Genetics 60: 772-789
(1997)

Harumi, T., Kimura, M., and Yasue, H. “Survey on swine SIN Es (PRE 1) as candidates
for SSCP markers in genetic linkage analysis” Animal Genetics 26: 403-406
(1995)

Holmes, N.G., Mellersh, C.S., Humphreys, S.J., Holliman, A., Curtis, R., and Sampson,
J. “Isolation and characterization of microsatellites from the canine genome”,
Animal Genetics 24: 289-292 (1992).

Holsinger, K.E. “Wahlund Effect, Wright’s F Statistics”,
http://statgendpsunipi.it/coursesjle/documents/pdf/Holsinger-Wahlundmdf,
(2001).

Housley, D., Zalewski, Z., Beckett, S., and Venta, P. “ Design Factors that Influence PCR
Amplification Success of Cross-species Primers among 1147 Mammalian Primer
Pairs” BMC Genomics 7: 253 (2006).

Housley, D.J.E., Ritzert, E., and Venta, P.J. “Comparative Radiation Hybrid Map of
Canine Chromosome 1 Incorporating SNP and Indel Polymorphisms” Genomics
84(2): 248-264 (2004).

ISGN. “Guidelines for human gene nomenclature: an international system for human

gene nomenclature (ISGN, 1987)” Cytogenetics and Cell Genetics 46: 11-28
(1987)

129

Juneja, R., Christensen, K., Andersen, E., and Gahne, B. “Frequencies of transferring
types in various breeds of domestic dog”, Animal Blood Groups and Biochemical
Genetics 12: 79-88 (1981).

Kajii, T. and Ohama, K. “Androgenetic origin of the hydatiform mole”, Nature 268: 633-
634 (1977).

Kathmann, 1., Jaggy, A., Busato,A., Bartschi, M., and Gaillard, C. “Clinical and genetic
investigations of idiopathic epilepsy in the Bernese Mountain Dog”, Journal of
Small Animal Practice 40: 319-325 (1999).

Kirkness, E., Bafna, V., Halpem, A., Levy, 8., Remington, K., Rusch, D., Delcher, A.,
Pop, M., Wang, W., and Fraser, C. “The Dog Genome: Survey Sequencing and
Comparative Analysis”, Science 301: 1898-1903 (2003).

Kirov, G., Nikolov, I., georgieva, L., Moskvina, V., Owen, M., and O’Donovan, M.
“Pooled DNA Genotyping on Affymetrix SNP Genotyping Arrays”, BMC
Genomics 7: 27 (2006).

Koskinen, M. and Bredbacka, P. “Assessment of the population structure of five Finnish
dog breeds with microsatellites”, Animal Genetics 31: 310-317 (2000).

Krawczak, M., Ball, E.V., and Cooper, D.N. “Neighboring-nucleotide effects on the rates
of germ- line single-base-pair substitution in human genes American Journal of
Human Genetics 63: 474-488 (1998).

Kruglyak, L. “Prospects for whole-genome linkage disequilibrium mapping of common
disease genes”, Nature Genetics 22: 139-144 (1999).

Kruglyak, L. “The use of a genetic map of biallelic markers in linkage studies”, Nature
Genetics 17: 21-24 (1997).

Kwok, P-Y. and Chen, X. “Detection of Single Nucleotide Variations” Genetic
Engineering 20: 125-134 (1998).

Kwok, P-Y., Deng, Q., Zakeri, H., and Nickerson D.A. “Increasing the Information
Content of STS-Based Genome Maps: Identifying Polymorphisms in Mapped
STSs”, Genomics 31: 123-126 (1996).

Kwok, P-Y., Carlson, C., Yager, T., Ankener, W., and Nickerson, D.A. “Comparative
Analysis of human DNA Variations by Fluorescence-based Sequencing of PCR
Products”, Genomics 23: 138-144 (1994).

Lai, E., Riley, J ., Purvis, 1., and Roses, A. “A 4-MB High-Density Single Nucleotide
Polymorphism-Based Map around Human APOE”, Genomics 54: 31 -38 (1998).

130

 

Landegren, U., Nilsson, M., and Kwok, P-Y. “Reading Bits of Genetic Information:
Methods for Single-Nucleotide Polymorphism Analysis”, Genome Research 8:
769-776 (1998).

Langston, A.A., Mellersh, C.S., neal, C.L., Ray, K., Acland, G.M., Gibbs, M., Aguirre,
G.D., Foumier, R.E.K., and Ostrander, E.A. “Construction of a Panel of Canine-
Rodent Hybrid Cell Lines for Use in partitioning of the Canine Genome”,
Genomics 46: 3 17-325 (1997).

Li, W.-H. and Graur, D. “Fundamentals of Molecular Evolution”, Sunderland, MA:
Sinauer Associates, Inc (1987).

Lin, L., Faraco, J ., Li, RKadotani, H., Rogers, W., Lin, X., Qiu, X, deJong, P.J., nishino,
S., and mignot, E. “The Sleep Disorder Canine Narcolepsy is Caused by a
Mutation in the Hypocretin (Orexin) Receptor 2”, Cell 98(3): 365-3 76 (1999).

Lingaas, F., Aarskaug, T., Gerlach, J .A., Juneja, R.K., Fredolrn, M., Sampson, J, Suter,
N., Holmes, N.G., Binns, M.M., Ryder, E.J., Van Haeringen, W.A., Venta, P.J.,
Brouillette, J .A., Yuzbasiyan-Gurkan, V., Wilton, A.N., Bredbacka, P., Koskinen,
M., Dunner, S., Parra, D., Schmutz, S., Schelling, C., Schlapfer, J ., and Dolf, G.
“A Canine Linkage Map: 39 Linkage Groups”, Journal of Animal Breeding and
Genetics 118: 3-19 (2001).

Lindblad-Toh, K., Wade, C., Mikkelsen, T., Karlsson, E., Jaffe, D., Kamal, M., Clamp,
M., Chang, J ., Kulbokas III, E., Zody, M., Mauceli, E., Xie, X., Breen, M.,
Wayne, R., Ostrander, E., Ponting, C., Galibert, F., Smith, D., deJong, P.,
Kirkness, E., Alvarez, P., Biagi, T., Brockman, W., Butler, J ., Chin, C-W., Cook,
A., Cuff, J ., Daly,M., DeCaprio, D., Gnerre, S., Grabherr, M., Kellis, M., Kleber,
M., Bardelben, C., Goodstadt, L., Heger, A., Hitte, C., Kim, L., Koepfli, K-P.,
Parker, H., Pollinger, J ., Searle, S., Sutter, N., Thomas, R., Webber, C., Broad
Institute Genome Sequencing Platform, and Lander, E. “Genome Sequence,
Comparative Analysis, and Haplotype Structure of the Domestic Dog”, Nature
438 (8): 803-819 (2005).

Lindblad-Toh, K. “Trait Mapping Using a Canine SNP Array Third International
Conference: Advances in Canine and Feline Genomics University of California
Davis, August 2-5, 2006.

Lingaas, F ., Sorensen, A., Juneja, R.K., Johansson, S., Fredholm, M., Wintero, A.K.,
Sampson, J ., Mellersh, C., Curzon, A., Holmes, N.G., Binns, M.M., Dickens,
H.F., Ryder, E.J., Gerlach, J ., Baumle, E., and Dolf, G. “Towards Construction of
a Canine Linkage Map: Establishment of 16 Linkage Groups”, Mammalian
Genome 8: 218-221 (1997).

131

Lyons, L.A., Laughlin, T.F., Copeland, N.G., Jenkins, N.A., Womack, J .E., and O’Brien,
SJ. “Comparative anchor tagged sequences (CATS) for integrative mapping of
mammalian genomes”, Nature Genetics 15: 47-56 (1997).

Mandiger, P., Knol, B., Ubbink, G., and Gruys, E. “Hereditary Necrotising Myelopathy
in Kooiker Dogs”, Research in Veterinary Science 54: 305-316 (1993).

Meera Khan, P., Brahe, C., and Wijnen, L.M.M. “Gene map of the dog: six conserved
and three disrupted syntenies”, C ytogenetics and Cell Genetics 37: 537-538
(1984)

Meera Khan, P., Los, W.R.T., van den Does, J .A., and Epstein, R.B. “Isoenzyrne markers
in Dog Blood Cells”, Transplantation 15: 624-628 (1973).

Mellersh, C.S., Hitte, C., Richman, M., Vignaux, F., Priat, C., Jouquant, S., Werner, P.,
Andre, C., DeRose, S., Patterson, D.F., Ostrander, E.A., and Galibert, F. “An
integrated Linkage-Radiation Hybrid Map of the Canine Genome”, Mammalian
Genome 11: 120-130 (2000).

Mellersh, C.S., Langston, A.A., Acland, G.M., Fleming, M.A., Ray, K., Weigand, N.A.,
Francisco, L.V., Gibbs, M., Aguirre, G.D., and Ostrander, E.A. “A Linkage Map
of the Canine Genome”, Genomics 46: 326-336 (1997).

Minnick, M.F., Stillwell, L.C., Heineman, J .M., and Stiegler, CL. “A highly repetitive
DNA sequence possibly unique to canids”, Gene 110: 23 5-238 (1992).

Morell, R., Friedman, T.B., Moeljopawiro, S., Hartono, Soewito, and Asher, J .H., Jr. “A
frameshifi mutation in the HuP2 paired domain of the probable human homolog
of murine Pax-3 is responsible for Waardenburg syndrome type 1 in an
Indonesian family”, Human Molecular Genetics 1: 243-247 (1992).

Morera, L., Barba, C.J., Garrido, J .J ., Barbancho, M., and de Andres, D.F. “Genetic
variation detected by microsatellites in five Spanish breeds”, Journal of Heredity
90: 654-656 (1999).

MurIay, N.E., Brammer, W.J., and Murray, K. “Lambdoid phages that simplify the
recovery of in vitro recombinants”, Molecular and General Genetics 156: 53-59
(1977)

Neff, M.W., Broman, K.W., Mellersh, C.S., Ray, K., Acland, G.M., Aguirre, G.D.,
Ziegle, J .S., Ostrander, E.A., and Rine, J. “A Second-Generation Genetic Linkage
Map of the Domestic Dog, Canis Familiaris”, Genetics 151: 803-820 (1999).

Nei, M. and Li W.H. “Mathematical model for studying genetic variation in terms of
restriction endonucleases”, Proceedings of the National Academy of Science of
the United States of America 76: 5269-5273 (1979).

132

Nickerson, D.A., Taylor, S.L., Weiss, K.M., Clark, A.G., Hutchinson, R.G., Stengard, J .,
Salomaa, V., Vartianen, E., Boerwinkle, E., and Sing, C.F. “DNA Sequence

diversity in a 9.7 KB Region of the Human Lipoprotein Lipase Gene”, Nature
Genetics 19: 233-240 (1998).

O'Brien, SJ. “Molecular genetics in the domestic cat and its relatives”, Trends in
Genetics 2: 137-142 (1986).

O'Brien, S.J., Womack, J .E., Lyons, L.A., Moore, K.J., Jenkins, N.A., and Copeland,
N.G. “Anchored reference loci for comparative genome mapping in mammals”,
Nature Genetics 3: 103-112 (1993).

Olsen, M., Hood, L., Cantor, C., and Botstein, D. “A common language for physical
mapping of the human genome”, Science 245: 1434-1435 (1989).

Ostrander, EA. and Kruglyak, L. “Unleashing the Canine Genome”, Genome Research
10: 1271-1274 (2000).

Ostrander, E.A., Galibert, F., and Patterson, D.F. “Canine Genetics Comes of Age”,
Trends in Genetics 16(3): 117-124 (2000).

Ostrander, E.A., Mapa, F.A., Yee, M., and Rine, J. “One Hundred and One Simple
Sequence Repeat-Based Markers for the Canine Genome”, Mammalian Genome
6: 192-195 (1995).

Ostrander, E.A., Sprage, Jr., GR, and Rine, J. “Identification and Characterization of

Dinucleotide Repeat (CA)n markers for for Genetic Mapping in Dog”, Genomics
16: 207-213 (1993).

Padgett, G.A. Control of canine genetic Diseases, Howell Book House, New York, NY,
(1998), pg. 4.

Parker, H., and Ostrander, E. “Canine genomics and genetics: running with the pack”,
PLOS Genetics 1(5): 508-513 (2005).

Patterson, D.F. “Companion Animal Medicine in the Age of medical Genetics”, journal
of Veterinary Internal medicine 14: 1-9 (2000).

Patterson, D.F., Haskins, M.E., Jezyk, P.F., Giger, U., meyers-Wallen, V.N., Aguirre, G.,
Fyfe, J .C., and Wolfe, J .H. “Research on Genetic Diseases: Reciprocal Benefits
to Animals and Man”, Journal of the American Veterinary Medical Association
193(9): 1131-1144 (1988).

Pearson, H. “Changing Attitudes to Congenital and Inherited Diseases”, Veterinary
Record 105: 318-323 (1979).

133

Petersen-Jones, S.M., Entz, DD, and Sargan, D.R. “cGMP Phosphodiesterase-alpha
Mutation Causes Progressive Retinal Atrophy in the Cardigan Welsh Corgi Dog”,
investigative Ophthalmology and Visual Science 40: 1637-1644 (1999).

Picoult-Newberg, L, Ideker, T.B., Poh., M.G., Taylor, S.L., Donaldson, M.A., nickerson,
D.A., and Boyce-Jacino, M. “Mining SNPs From EST Databases”, Genome
Research 9: 167-174 (1999).

Priat, C., Hitte, C., Vignaux, F., Renier, C., Jiang, Z., jouquand, S., Cheron, A., Andre,
C., and Galibert, F. “A Whole-Genome Radiation Hybrid Map of the Dog
Genome”, Genomics 54: 361-378 (1998).

Ray, K., Trepanier, L.A., Acland, G.M., and Aguirre, G.D. “PCR/RFLP Marker in the
Canine Opsin Gene”, Animal Genetics 27: 293-294 (1996).

Richards, B., Skoletsky, J ., and Shuber, A. “Multiplex PCR amplification from the CF TR
gene using DNA prepared from buccal brushes/swabs”, Human Molecular
Genetics 2: 159-163 (1992).

Richman, M., Mellersh, C.S., Andre, C., Galibert, F., and Ostrander, E.A.
“Characterization of a Minimal Screening Set of 172 microsatellite markers for

Genome-Wide Screens of the Canine Genome”, Journal of Biochemical and
Biophysical Methods”, 47: 137-149 (2001).

Ridley, M “Wahlund Effect” in Evolution, 3rd Edition, .
http://blackwellpublishingcom/ridley/a-z/Wahlund_effect.asp, November, 2009.

Rieder, M.J., Taylor, S.L., Clark, A.G., and Nickerson, D.A. “Sequence Variation in the
Human Angiotensin Converting Enzyme”, Nature Genetics 22: 59-62 (1999).

Risch, N. and Merikangas, K. “The Future of Genetic Studies of Complex Human
Diseases”, Science 273: 1516-1517 (1996).

Roberts, R]. and Macelis, D. “REBASE-Restriction Enzymes and Methylases”, Nucleic
Acid Research 25: 248-262 (1997).

Robinson, R. “Genetic anomalies in dogs”, Canine Practice 16: 29-34 (1991).

Roebroek, A.J.M., Schalken, J .A., Onnekink, C., Bloemers, H.P.J., and Van de Ven, W.J.
“Structure of the feline c-fes/ fps proto-oncogene: Genesis of a retroviral
oncogene”, Journal of Virology 61: 2009-2016 (1987).

Rothuizen, J ., Wolfswinkel, J ., Lenstra, J .A., and F rants, R.R. “The incidence of mini-

and micro-satellite repetitive DNA in the canine genome”, Theoretical and
Applied Genetics 89: 403 -406 (1994).

134

Sambrook, J ., Fritsch, BF. and Maniatis, T. “Molecular Cloning. A Laboratory Manual”,
(2nd ed.) Cold Springs Harbor: Cold Springs Harbor Laboratory Press (1989).

Sargan, D.R., Yang, F ., Squire, M., Milne, B.S., O’Brien, P.C.M., and Ferguson-Smith,
M.A. “Use of Flow-Sorted Canine Chromosomes in the Assignment of Canine
Linkage, Radiation Hybrid, and Syntenic Groups to Chromosomes: Refinement
and Verification of the Comparative Chromosome Map for Dog and Human”,
Genomics 69: 182-195 (2000).

Sarmiento, U.M. and Storb, R.F. “RFLP Analysis of DLA Class I Genes in the Dog”,
Tissue Antigens 34(3): 158-163 (1989).

Sarmiento, U.M. and Storb, R.F. “Characterization of Class 11 Alpha Genes and DLA-D
Region Allelic Asociations in the Dog”, Tissue Antigens 32(4): 224-234 (1988).

Serikawa, T., Kurarnoto, T., Hilbert, P., Mori, M., Yamada, J ., Dubay, C.J., Lindpainter,
K., Ganten, D., Guenet, J .-L., Lathrop, G.M., and Beckmann, J .S. “Rat gene
mapping using PCR-analyzed microsatellites”, Genetics 131: 701-721 (1992).

Sharp, N.J.H., Komegay, J .N., VanCamp, S.D., Herbstreith, M.H., Secore, S.L., Kettle,
S., Hung, W.Y., Dykestra, M.J., et al. “An Error in Dystrophin mRNA Processing

in Golden Retriever Muscular Dystrophy, an Animal Homologue of Duchenne
Muscular Dystrophy”, Genomics 13(1): 115-121 (1992).

Shi, M. “Enabling large-scale pharmacogenomic studies by high-throughput mutation
detection and genotyping technologies”, Clinical Chemistry 47: 164—172 (2001).

Shibuya, H., Collins, B.K., Collier, L.L., Huang, T.H-M., Nonneman, D., and Johnson,
G.S. “A polymorphic (GAAA)n microsatellite in a canine Wilms tumor 1 (WTl)
gene intron”, Animal Genetics 27 : 59-60 (1996).

Shibuya, H., Collins, B.K., Stoy, S.J., nonneman, D.J., and Johnson, G.S. “PCR/RF LP
Markers in the Canine Garnma-D-Crystallin Gene”, Animal Genetics 26: 445-446
(1995).

Shubitowski, D. M., Venta, P.J., Douglass, C.L., Zhou, R.X., and Ewart, S.L.
“Polymorphism Identification Within 50 Equine Gene-specific Sequence Tagged
Sites”, Animal Genetics 32(2): 78-88 (2001).

Simonsen, V. “Electrophoretic Studies on the Blood proteins of Domestic Dogs and other
Canidae”, Hereditas 82: 7-18 (1976).

Southem, E.M. “Measurement of DNA length by gel electrophoresis”, Analytical
Biochemistry 100: 319-323 (1979).

135

Suber, M.L., Pittler, S.J., Qin, N., Wright, G.C., Holcombe, V., Lee, R.H., Craft, C.M.,
Lolley, R.N., Baehr, W., and Hurwitz, R.L. “Irish Settr Dogs Affected with
Rod/Cone Dysplasia Contain a nonsense Mutation in the Rod cGMP
Phosphodiesterase Beta-subunit Gene”, Proceedings of the National Academy of
Sciences of the United States of America 90(9): 3968-3972 (1993).

Sutter, N., Eberle, M., Parker, H., Pullar, B., Kirkness, E., Kruglyak, L., and Ostrander,
E. “Extensive and BAreed-Specific Linkage Disequilibrium in Canis familiaris”,
Genome Research 14: 23 88-2396 (2004).

Taillon-Miller, P., Piemot, BE, and Kwok, P-Y., “Efficient Approach to Unique Single-
Nucleotide Polymorphism Discovery”, Genome Research 9: 499-505 (1999).

Taillon-Miller, P., Gu, Z., Li, Q., Hillier, L., and Kwok, P-Y. “Overlapping Genomic
Sequences: A Treasure Trove of Single-Nucleotide Polymorphisms”, Genome
Research 8: 748-754 (1998).

Taillon-Miller, P., Bauer-Sardina, 1., Zakeri, H., Hillier, L., Mutch, D.G., and Kwok, P-Y.
“The Homozygous Complete Hydatiform Mole: A unique Resource for Genome
Studies”, Genomics 46: 307-310 (1997).

Uchida, T., Komori, M., Kitada, K., and Karnataki, T. “Isolation of cDNAs coding for
three different forms of liver microsomal cytochrome P—450 form polychlorinated
biphenyl-treated Beagle dogs”, Molecular Pharmacology 38: 644-651 (1990).

Van de Sluis, B., Rothuizen, J ., Pearson, P.L., van Oost, B., and Wijmenga, C.
“Identification of a New Copper Metabolism Gene by Positional Cloning in a
Purebred Dog Population”, Human Molecular Genetics 11(2): 165-173 (2002).

Venta, P.J. and Vidal, M. “MspI RSP in Exon 42 of the Canine von Willebrand Factor
(VWF) Gene”, Animal Genetics 30: 229 (1999).

Venta, P.J., Yuzbasiyan—Gurkan, V., Brewer, G.J., and Schall, W.D. “Mutation Causing
vonWillebrand’s Disease in Scottish Terriers”, Jounal of Veterinary Internal
Medicine 14: 10-19 (2000).

Venta, P.J., Cao, Y., Alexander, L., and Yuzbasiyan-Gurkan, V. “A polymorphic
microsatellite in the canine retinoblastoma (RBI) gene”, Animal Genetics 30:
470-472 (1999).

Venta, P.J., Brouillette, J .A., Yuzbasiyan-Gurkan, V., and Brewer, G.J. “Gene-specific

universal mammalian sequence-tagged sites: application to the canine genome”,
Biochemical Genetics 34: 321-341 (1996).

136

Venta, P.J., Hewett-Emmett, D., and Tashian, R.E. “Simple method to convert DNA
sequence variation into sites cut by restriction endonucleases: utility shown by
typing the human CA3 and mouse strain Car-2 polymorphisms”, American
Journal of Human Genetics 49: a445 (1991).

Venter, J .C., Adams, m.D., myers, B.W., Li, P.W., Mural, R.J., et al. “The Sequence of
the Human Genome”, Science 291: 1304-1351 (2001).

Vignaux, F., Hitte, C., Priat, C., Chuat, J. C., Andre, C., and Galibert,F. “Construction
and Optimization of a Dog Whole-genome Radiation Hybrid Panel”, Mammalian
Genome 10: 888-894 (1999).

Wahlund, S. “Zusammensetzung von populationen und korrelation-sercheinungen von
standpunkt der vererbungslehre aus betrachtet”, Hereditas 11: 65-106 (1928).

Wang, D.G., Fan, J-B., Siao, C-J., Berno, A., Young, P., Sapolsky, R., Ghandour, G.,
Perkins, N., Winchester, E., Spencer, J ., Kruglyak, L., Stein, L., Hsie, L.,
Topaloglou, T., Hubbell, E., Robinson, E., Mittrnann, M., Morris, M.S., Shen, N.,
Kilbum, D., Rioux, J ., Nubaum, C., Rozen, 8., Hudson, T.J., Lipshutz, R., Chee,
M., and Lander, E.S. “Large-Scale Identification, Mapping, and Genotyping of
Single-Nucleotide Polymorphisms in the Human Genome”, Science 280: 1077-
1082 (1998).

Weber, J. “Informativeness of human (dC-dA)n.(dG-dT)n polymorphisms”, Genomics 7:
524-30 (1990).

I Weiden, P., Storb, R., Kolb, H.J., Graham, T., Anderson, J ., and Giblett, E. “Genetic
Variation of the Red Blood Cell Enzymes in the Dog”, Transplantation 17(1):
115-120 (1974).

Weissbach, J ., Gyapay, G., Dib, C., Vignal, A., Morissette, J ., Millasseau, P., Vaysseix,
G., and Lathrop, M. “A second-generation linkage map of the human genome”
Nature 359: 794-801 (1992).

Werner, P., Mellersh, C.S., Raducha, M.G., DeRose, S., Acland, G.M., prociuk, U.,
Wiegand, N., Aguirre, G.D., Henthom, P.S., Patterson, DR, and Ostrander, E.A.
“Anchoring of Canine Linkage Groups with Chromosome-Specific Markers”,
Mammalian Genome 10: 814-823 (1999).

Whitney, K.M., Goodman, S.A., Bailey, E.M., and Lothrop, C.D., Jr. “The molecular
basis of canine pyruvate kinase deficiency”, Experimental Hematology 22: 866-
874 (1994). ' '

Wilcox, B. and Walkowicz, C. “The Atlas of Dog Breeds of the World”, 5th ed., T.F.H.
Publications, Neptune City, NJ. 1995.

137

 

Willis, M.B. “Genetics of the Dog”, New York, NY: Howell Book House (1989).

WinterO, A.K., Fredholm, M., and Thomsen, P.D. “Variable (dG-dT),,-(dC-dA)n
sequences in the porcine genome”, Genomics 12: 281-288 (1991).

Wright, S. “The genetical structure of populations”, Annals of Eugenics 15: 323-321
(1951)

Wu, D.Y., Ugozzoli, L., Pal, B.K., and Wallace, B.R. “Allele-specific enzymatic
amplification of beta-globin genomic DNA for diagnosis of sickle cell anemia”,
Proceedings of the National Academy of Science of the United States of America
86: 2757-2760 (1989).

Yang, F., O’Brien, P.C.M., Milne, B.S., Graphodatsky, A.S., Solanky, N., Trifonov, V.,
Rens, W., Sargan, D., and F erguson-Smith, MA. “A Complete Comparative
Chromosome Map for the Dog, Red Fox, and Human and its Integration with
Canine Genetic Maps”, Genomics 62: 189-202 (1999).

Yuzbasiyan-Gurkan, V., Blanton, S.H., Cao, Y., Ferguson, P., Li, J ., Venta, P.J., and
brewer, G.J. “Linkage of a Microsatellite Marker to the Canine Copper Toxicosis
Locus in Bedlington Terriers”, American Journal of Veterinary Research 58(1):

23-27 (1997).

Yuzbasiyan-Gurkan, V., Wagnitz, S., Blanton, SH, and Brewer, G.J. “Linkage studies of
the esterase D and retinoblastoma genes to canine copper toxicosis: a model for
Wilson disease”, Genomics 15: 86-90 (1993).

Zajc, 1., Mellersh, C.S., and Sampson, J. “Variability of canine microsatellites within and
between different dog breeds”, Mammalian Genome 8: 182-185 (1997).

Zeitzer, J .M., Nishino, S., and Mignot, E. “The Neurobiology of Hypocretins (Orexins),

Narcolepsy and Related Therapeutic Interventions”, Trends in Pharmacological
Sciences 27(7): 368-374 (2006).

138

 

   

mllllllllllllll llllll llllllllllllllll Ill'lllllllles
3 1293 03063 5886