The Feline SNP Array

Features and Utility

Hasan Alhaddad, Ph.D. Kuwait University

Barbara Gandolfi Hasan Alhaddad Mike Montague

Mona Abdi Erica K Creighton

Bianca Haase Maria Longeri Rashid Saif

Carlyn Peterson Brian Davis

William Murphy Ettore Randi

Shannon Joslin Grace Lan

Jeff Brockman Mike Hamilton Nick Dodman Richard Malik

Clare Rusbridge Nick Gustafson Diane Shelton

Robert A Grahn Jens Haggstrom

Serina Filler Hannes Lohi

James C Mullikin Chris Helps

Niels C Pedersen Wes Warren

Leslie A Lyons

A work team & a teamwork

AIMS

• Share information of the available dataset of genotyped cats

• Provide an updated version of Feline SNP array

• Analyze basic population genetics of available genotype dataset

• Initiate and encourage collaborations based on the available genotype dataset

Outline

• Introduction

• Genotype dataset

• Feline array features

• Array utility

Cat genome & SNPs

Mullikin et al. BMC Genomics 2010, 11:406http://www.biomedcentral.com/1471-2164/11/406

Open AccessDATABASE

© 2010 Mullikin et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction inany medium, provided the original work is properly cited.

DatabaseLight whole genome sequence for SNP discovery across domestic cat breedsJames C Mullikin*1, Nancy F Hansen1, Lei Shen2, Heather Ebling2, William F Donahue2, Wei Tao2, David J Saranga2, Adrianne Brand2, Marc J Rubenfield2, Alice C Young1, Pedro Cruz1 for NISC Comparative Sequencing Program1, Carlos Driscoll3, Victor David3, Samer WK Al-Murrani4, Mary F Locniskar4, Mitchell S Abrahamsen4, Stephen J O'Brien3, Douglas R Smith2 and Jeffrey A Brockman4

AbstractBackground: The domestic cat has offered enormous genomic potential in the veterinary description of over 250 hereditary disease models as well as the occurrence of several deadly feline viruses (feline leukemia virus -- FeLV, feline coronavirus -- FECV, feline immunodeficiency virus - FIV) that are homologues to human scourges (cancer, SARS, and AIDS respectively). However, to realize this bio-medical potential, a high density single nucleotide polymorphism (SNP) map is required in order to accomplish disease and phenotype association discovery.

Description: To remedy this, we generated 3,178,297 paired fosmid-end Sanger sequence reads from seven cats, and combined these data with the publicly available 2X cat whole genome sequence. All sequence reads were assembled together to form a 3X whole genome assembly allowing the discovery of over three million SNPs. To reduce potential false positive SNPs due to the low coverage assembly, a low upper-limit was placed on sequence coverage and a high lower-limit on the quality of the discrepant bases at a potential variant site. In all domestic cats of different breeds: female Abyssinian, female American shorthair, male Cornish Rex, female European Burmese, female Persian, female Siamese, a male Ragdoll and a female African wildcat were sequenced lightly. We report a total of 964 k common SNPs suitable for a domestic cat SNP genotyping array and an additional 900 k SNPs detected between African wildcat and domestic cats breeds. An empirical sampling of 94 discovered SNPs were tested in the sequenced cats resulting in a SNP validation rate of 99%.

Conclusions: These data provide a large collection of mapped feline SNPs across the cat genome that will allow for the development of SNP genotyping platforms for mapping feline diseases.

BackgroundAlong with dogs, the domestic cat enjoys extensive veter-inary surveillance, more than any other animal. A rich lit-erature of feline veterinary models reveals a uniqueopportunity to explore genetic determinants responsiblefor genetic diseases, infectious disease susceptibility,behavioral and neurological phenotypes, reproductionand physiology (see [1] and [2] for citations). As a highlyvenerated pet this extraordinarily successful domesticspecies comprises as many as one billion individualsworldwide. House cats have become a familiar compan-

ion to people since their original domestication from theAsian wildcat (Felis silvestris lybica), recently estimated atapproximately 10,000 years ago in the Middle East's Fer-tile Crescent[3]. In spite of our affection for cats,advances in clinical resolution of genetic maladies andcomplex diseases has been slower than for other specieslargely due to a delay in achieving a useful whole genomesequence of the cat. This has changed recently with thecompletion of a draft 1.9X genome sequence of a femaleAbyssinian cat named Cinnamon who gave us our firstglimpse and hope of developing the species as an activeplayer in the genomics era[1,4].

The availability of a sufficiently dense single-nucleotidepolymorphism (SNP) map for a species provides aresource which enables the power of automated high-

* Correspondence: mullikin@mail.nih.gov1 Genome Technology Branch and NIH Intramural Sequencing Center, National Human Genome Research Institute, National Institutes of Health, Bethesda, Maryland 20892, USAFull list of author information is available at the end of the article

10.1101/gr.6380007Access the most recent version at doi: 2007 17: 1675-1689 Genome Res.

Bourque, Glenn Tesler, NISC Comparative Sequencing Program and Stephen J. O’Brien Antunes, Marilyn Menotti-Raymond, Naoya Yuhki, Jill Pecon-Slattery, Warren E. Johnson, Guillaume A. Schäffer, Richa Agarwala, Kristina Narfström, William J. Murphy, Urs Giger, Alfred L. Roca, AgostinhoSante Gnerre, Michele Clamp, Jean Chang, Robert Stephens, Beena Neelam, Natalia Volfovsky, Alejandro Joan U. Pontius, James C. Mullikin, Douglas R. Smith, Agencourt Sequencing Team, Kerstin Lindblad-Toh,

Initial sequence and comparative analysis of the cat genome

dataSupplementary

http://www.genome.org/cgi/content/full/17/11/1675/DC1 "Supplemental Research Data"

References

http://www.genome.org/cgi/content/full/17/11/1675#ReferencesThis article cites 97 articles, 41 of which can be accessed free at:

serviceEmail alerting

click heretop right corner of the article or Receive free email alerts when new articles cite this article - sign up in the box at the

Notes

http://www.genome.org/subscriptions/ go to: Genome ResearchTo subscribe to

© 2007 Cold Spring Harbor Laboratory Press

on November 5, 2007 www.genome.orgDownloaded from

Comparative analysis of the domestic cat genomereveals genetic signatures underlying felinebiology and domesticationMichael J. Montaguea,1, Gang Lib,1, Barbara Gandolfic, Razib Khand, Bronwen L. Akene, Steven M. J. Searlee,Patrick Minxa, LaDeana W. Hilliera, Daniel C. Koboldta, Brian W. Davisb, Carlos A. Driscollf, Christina S. Barrf,Kevin Blackistonef, Javier Quilezg, Belen Lorente-Galdosg, Tomas Marques-Bonetg,h, Can Alkani, Gregg W. C. Thomasj,Matthew W. Hahnj, Marilyn Menotti-Raymondk, Stephen J. O’Brienl,m, Richard K. Wilsona, Leslie A. Lyonsc,2,William J. Murphyb,2, and Wesley C. Warrena,2

aThe Genome Institute, Washington University School of Medicine, St. Louis, MO 63108; bDepartment of Veterinary Integrative Biosciences, College ofVeterinary Medicine, Texas A&M University, College Station, TX 77843; cDepartment of Veterinary Medicine & Surgery, College of Veterinary Medicine,University of Missouri, Columbia, MO 65201; dPopulation Health & Reproduction, School of Veterinary Medicine, University of California, Davis, CA 95616;eWellcome Trust Sanger Institute, Hinxton CB10 1SA, United Kingdom; fNational Institute on Alcohol Abuse and Alcoholism, National Institutes of Health,Bethesda, MD 20886; gCatalan Institution for Research and Advanced Studies, Institute of Evolutionary Biology, Pompeu Fabra University, 08003Barcelona, Spain; hCentro de Analisis Genomico 08028, Barcelona, Spain; iDepartment of Computer Engineering, Bilkent University, Ankara 06800, Turkey;jDepartment of Biology, Indiana University, Bloomington, IN 47405; kLaboratory of Genomic Diversity, Center for Cancer Research, Frederick, MD 21702;lDobzhansky Center for Genome Bioinformatics, St. Petersburg State University, St. Petersburg 199178, Russia; and mOceanographic Center, NovaSoutheastern University, Fort Lauderdale, FL 33314

Edited by James E. Womack, Texas A&M University, College Station, TX, and approved October 3, 2014 (received for review June 2, 2014)

Little is known about the genetic changes that distinguishdomestic cat populations from their wild progenitors. Here wedescribe a high-quality domestic cat reference genome assemblyand comparative inferences made with other cat breeds, wildcats,and other mammals. Based upon these comparisons, we identifiedpositively selected genes enriched for genes involved in lipidmetabolism that underpin adaptations to a hypercarnivorous diet.We also found positive selection signals within genes underlyingsensory processes, especially those affecting vision and hearing in thecarnivore lineage. We observed an evolutionary tradeoff betweenfunctional olfactory and vomeronasal receptor gene repertoires in thecat and dog genomes, with an expansion of the feline chemosensorysystem for detecting pheromones at the expense of odorant de-tection. Genomic regions harboring signatures of natural selectionthat distinguish domestic cats from their wild congeners are enrichedin neural crest-related genes associated with behavior and reward inmouse models, as predicted by the domestication syndrome hypoth-esis. Our description of a previously unidentified allele for the glovingpigmentation pattern found in the Birman breed supports the hy-pothesis that cat breeds experienced strong selection on specificmutations drawn from random bred populations. Collectively, thesefindings provide insight into how the process of domestication alteredthe ancestral wildcat genome and build a resource for future diseasemapping and phylogenomic studies across all members of the Felidae.

Felis catus | domestication | genome

The domestic cat (Felis silvestris catus) is a popular pet species,with as many as 600 million individuals worldwide (1). Cats

and other members of Carnivora last shared a common ancestorwith humans ∼92 million years ago (2, 3). The cat family Felidaeincludes ∼38 species that are widely distributed across the world,inhabiting diverse ecological niches that have resulted in di-vergent morphological and behavioral adaptations (4). Theearliest archaeological evidence for human coexistence with catsdates to ∼9.5 kya in Cyprus and ∼5 kya in central China (5, 6),during periods when human populations adopted more agricul-tural lifestyles. Given their sustained beneficial role surroundingvermin control since the human transition to agriculture, anyselective forces acting on cats may have been minimal sub-sequent to their domestication. Unlike many other domesticatedmammals bred for food, herding, hunting, or security, most ofthe 30–40 cat breeds originated recently, within the past 150 y,largely due to selection for aesthetic rather than functional traits.

Previous studies have assessed breed differentiation (6, 7),phylogenetic origins of the domestic cat (8), and the extent ofrecent introgression between domestic cats and wildcats (9, 10).However, little is known regarding the impact of the domesti-cation process within the genomes of modern cats and how thiscompares with genetic changes accompanying selection identified inother domesticated companion animal species. Here we describe, toour knowledge, the first high-quality annotation of the complete

Significance

We present highlights of the first complete domestic cat referencegenome, to our knowledge. We provide evolutionary assessmentsof the feline protein-coding genome, population genetic discoveriessurrounding domestication, and a resource of domestic cat geneticvariants. These analyses span broadly, from carnivore adaptationsfor hunting behavior to comparative odorant and chemical de-tection abilities between cats and dogs. We describe how segre-gating genetic variation in pigmentation phenotypes has reachedfixation within a single breed, and also highlight the genomic dif-ferences between domestic cats and wildcats. Specifically, the sig-natures of selection in the domestic cat genome are linked to genesassociated with gene knockout models affecting memory, fear-conditioning behavior, and stimulus-reward learning, and poten-tially point to the processes by which cats became domesticated.

Author contributions: M.J.M., G.L., B.G., L.A.L., W.J.M., and W.C.W. designed research;M.J.M., G.L., B.G., P.M., L.W.H., D.C.K., B.W.D., C.A.D., C.S.B., K.B., G.W.C.T., M.W.H., M.M.-R.,S.J.O., L.A.L., W.J.M., and W.C.W. performed research; M.J.M., G.L., B.G., B.L.A., S.M.J.S.,D.C.K., B.W.D., C.A.D., J.Q., B.L.-G., T.M.-B., C.A., G.W.C.T., M.W.H., R.K.W., L.A.L., W.J.M.,and W.C.W. contributed new reagents/analytic tools; M.J.M., G.L., B.G., R.K., B.W.D., J.Q.,B.L.-G., T.M.-B., C.A., G.W.C.T., M.W.H., L.A.L., W.J.M., and W.C.W. analyzed data; andM.J.M., G.L., B.G., R.K., P.M., D.C.K., B.W.D., C.A.D., C.S.B., K.B., T.M.-B., M.W.H., L.A.L.,W.J.M., and W.C.W. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Data deposition: The sequences reported in this paper have been deposited in the Gen-Bank database (accession nos. GU270865.1, KJ923925–KJ924979, SRX026946, SRX026943,SRX026929, SRX027004, SRX026944, SRX026941, SRX026909, SRX026901, SRX026955,SRX026947, SRX026911, SRX026910, SRX026948, SRX026928, SRX026912, SRX026942,SRX026930, SRX026913, SRX019549, SRX019524, SRX026956, SRX026945, and SRX026960).1M.J.M. and G.L. contributed equally to this work.2To whom correspondence may be addressed. Email: wwarren@genome.wustl.edu,wmurphy@cvm.tamu.edu, or lyonsla@missouri.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1410083111/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1410083111 PNAS Early Edition | 1 of 6

GEN

ETICS

Cat SNPs

• Marker selected based on:

• Re-sequencing of Cinnamon

• Sequencing 7 discovery cats

• Sequencing 6 breeds + random bred population + wild cat (pooled individuals).

Array

• Illumina Infinium iSelect 63K Cat DNA array • Released 2011

http://felinegenetics.missouri.edu/feline-genome-project-2/cat-genomic-resources-strs-snps

• Introduction• Genotype dataset • Feline array features • Array’s utility

• Genotype dataset

Outline

Samples

Breed (1824)

Random bred (262)

Domestic

European wildcatsAsian Leopard catsBig cats

Non-Pedigree

Lyons Ped (139)

LxD Ped (81)

Toyger (34)Lykoi Ped (27)T.Rex Ped (21)

• > 2000 Domestic cats

• > 50 wild cats (different species)

• 5 pedigrees

• 262 ramdom bred cats

• > 1800 pedigreed breed cats

Samples

Sample size0 50 100 150 200 250 300

AbyssinianAmerican Curl

American ShorthairAmerican Wirehair

BengalBirmanBombay

British ShorthairBurmese

ChartreauxCornish RexDevon Rex

Egyptian MauHavana Brown

Japanese BobtailKhoa Manee

KoratKurillian Bobtail

La PermLykoi

Maine CoonManx

MunchkinNorweigian Forest Cat

OcicatOrientalPerisan

PeterbaldRagdoll

Russian BlueScottish Fold

Selkirk RexSiameseSiberian

SingapuraSomaliSphynx

Tennesse REXTurkish Angora

Turkish Van

40 breeds Average 39 cats/breed

22 breeds > 20 individuals

• Introduction • Genotype dataset • Feline array features • Array’s utility• Feline array features

Outline

Array’s features

Autosomal

X-linked

Unassigned(~ 11%)

(~ 4.5%)

Domestic

Wildcat (~ 6.5%)

Non-phenotypic Phenotypic(~ 0.05%)

Non-phylogenetic Phylogenetic(~ 0.15%)

• ~ 63K SNPs.

• Domestic cat SNPs

• Wildcat SNPs

• Autosomal

• X-linked

• Phenotypic

• Phylogentic

• ~11% unknown location !!!

Array’s features

~ 11% of SNPs unassigned to genomic location

lambs born, number born alive, number born dead, weight

of lambs born and weight of lambs weaned (Appendix S1).

Genotypes and analysis: The deletion variant was

g.29500068_29500069delAT relative to NC_019477.1

(rs397514112). Genotyping was performed as described2

(Table S1). Analysis employed the mixed model or glim-

mix procedure in SAS v9.2 (SAS Institute) as described.6,7

For each analysis, the production trait of interest was

included as the dependent variable. Other model terms

included breed, sire, year of birth, age in years, age at last

lambing and genotype as described (Appendix S1). Bonfer-

roni correction accounted for multiple testing.

Comments: The ZNF389 deletion variant was not consis-

tently associated with any tested production traits between

lifetime and partial-lifetime SRLV-negative groups (Tables

S1 and S4). Insertion homozygotes previously associated

with reduced SRLV proviral concentration were associated

with lower birth weight in the partial-lifetime group (Bon-

ferroni P = 0.033; Table S4), but the weight difference was

small (0.41 kg) and not replicated in the lifetime group

(nominal P = 0.41; Table S2). Further, there was no associ-

ation with weaning or later weights (Tables S2 and S4).

These results showed no consistent association of the

ZNF389 deletion variant with ewe lifetime production.

Other breeds and additional traits, such as wool diameter

and infectious disease traits besides control of SRLV, should

be examined to more fully assess this locus.

Acknowledgements: Thanks to James Reynolds, Caylee Birge,

Codie Durfee, Nic Durfee, Liam Broughton-Neiswanger,

Ralph Horn, James Allison, Tom Kellom, Natalie Pierce,

Mark Williams and USSES farm crew for technical assis-

tance. This work was supported by USDA-ARS Grant

5348-32000-031-00D.

Conflict of interest: The authors have no conflict of interest

to declare.

References1 White S.N. et al. (2012) PLoS One 7, e47829.2 White S.N. et al. (2014) Anim Genet 45, 297–300.3 Herrmann-Hoesing L.M. et al. (2009) Clin Vaccine Immunol

16, 551–7.4 White S.N. & Knowles D.P. (2013) Viruses 5, 1466–99.5 Herrmann-Hoesing L.M. et al. (2007) Clin Vaccine Immunol

14, 1274–8.6 Mousel M.R. et al. (2010) Anim Genet 41, 222–3.7 Gonzalez M.V. et al. (2013) PLoS One 8, e74700.

Correspondence: S. N. White (swhite@vetmed.wsu.edu)

Supporting information

Additional supporting information may be found in the

online version of this article.

Appendix S1 Additional Methods Description.

Table S1 ZNF389 deletion variant genotype counts among

lifetime SRLV-negative ewes.

Table S2 Association results between ZNF389 deletion

variant and production phenotypes in lifetime SRLV-nega-

tive ewes.

Table S3 ZNF389 deletion variant genotype counts among

partial-lifetime group SRLV-negative ewes.

Table S4 Association results between ZNF389 deletion

variant and production phenotypes among partial-lifetime

group SRLV-negative ewes.

doi: 10.1111/age.12169

An updated felCat5 SNP manifest for theIllumina Feline 63k SNP genotyping array

Cali E. Willet and Bianca Haase

Faculty of Veterinary Science, University of Sydney, Sydney,

NSW, 2006, Australia

Accepted for publication 01 April 2014

Background: The development of the first Illumina Infinium

iSelect 63k Cat DNA genotyping array has been a mile-

stone in feline research. Since its release in February

2011, the International Cat Genome Sequencing Consor-

tium released a new version of the feline genome assembly

(Felis_catus 6.2/felCat5; GenBank assembly ID

GCA_000181335.2). As inconsistencies between genome

assemblies can complicate genome-wide association stud-

ies, we compare SNP locations of the manifest provided

with the Illumina Infinium iSelect 63k Cat DNA genotyp-

ing array with the most recent feline genome assembly,

felCat5. We make the resultant updated SNP manifest

available to the cat research community.

Methods: A FASTA file was created using the SNP identifier

and the genomic sequence in top orientation from the cur-

rent feline SNP manifest. For each probe, the longest of

the two flanking sequences either side of the variant was

selected to create the BLAST input file, and a nucleotide

basic local alignment search (BLAST)1 against the Felis_ca-

tus 6.2/felCat5 whole-genome assembly, September 2011

build, downloaded from UCSC Genome Browser (http://

genome.ucsc.edu/) was performed. The single best hit for

each sequence was retained. A custom perl script was

used to extract the position of each array marker relative

to the felCat5 assembly from the BLAST output, rejecting

hits that failed to reach the base position immediately

adjacent to the SNP and those where the E value was

greater than 1e-05. Hits that were rejected as well as

© 2014 Stichting International Foundation for Animal Genetics 45, 614–615

Brief Notes614

Table 2 Ten most associated markers obtained by genome-wide analyses of Persian PRA

No. Chr. SNP ID Position TDT sib-TDT Case–control

Praw Pgenome Praw Pgenome Praw Pgenome

1 E1 chrUn5.6839723 1831172 4.32E-08 1.00E-05 1.00E-05 0.1632 7.64E-13 1.00E-05

2 E1 chrUn5.6133983 1106562 1.21E-07 0.00013 0.00013 0.4071 1.52E-10 1.00E-05

3 E1 chrUn5.6766609 1751066 5.73E-07 0.00282 0.00282 0.5451 2.97E-09 5.00E-05

4 E1 chrUn5.6481762 1452354 7.74E-06 0.06134 0.06134 0.9057 2.70E-07 0.00226

5 E1 chrUn5.6503134 1476010 2.21E-05 0.1905 0.1905 0.9644 1.15E-06 0.00841

6 E1 chrUn5.5846986 808916 3.74E-05 0.3135 0.3135 0.9487 4.65E-06 0.03121

7 E1 chrUn5.6185290 1154788 3.74E-05 0.3135 0.3135 0.9487 8.04E-06 0.05255

8 E1 chrUn5.6912692 1912858 3.74E-05 0.3135 0.3135 0.9793 – –

9 E1 chrUn5.6942249 1932982 3.74E-05 0.3135 0.3135 0.9793 5.95E-06 0.03955

10 E1 chrUn5.7293975 2293660 3.74E-05 0.3135 0.3135 0.9487 8.76E-07 0.00658

11 F1 chrUn5.7948667 2997440 – – – – 4.45E-06 0.02986

Genome-wide analyses of Persian PRA

Genome-wide association and linkage analyses localizea progressive retinal atrophy locus in Persian cats

Hasan Alhaddad • Barbara Gandolfi • Robert A. Grahn •

Hyung-Chul Rah • Carlyn B. Peterson • David J. Maggs •

Kathryn L. Good • Niels C. Pedersen • Leslie A. Lyons

Received: 19 February 2014 / Accepted: 3 April 2014! The Author(s) 2014. This article is published with open access at Springerlink.com

Abstract Hereditary eye diseases of animals serve asexcellent models of human ocular disorders and assist in

the development of gene and drug therapies for inherited

forms of blindness. Several primary hereditary eye condi-tions affecting various ocular tissues and having different

rates of progression have been documented in domestic

cats. Gene therapy for canine retinopathies has been suc-cessful, thus the cat could be a gene therapy candidate for

other forms of retinal degenerations. The current study

investigates a hereditary, autosomal recessive, retinaldegeneration specific to Persian cats. A multi-generational

pedigree segregating for this progressive retinal atrophy

was genotyped using a 63 K SNP array and analyzed viagenome-wide linkage and association methods. A multi-

point parametric linkage analysis localized the blindness

phenotype to a *1.75 Mb region with significant LODscores (Z & 14, h = 0.00) on cat chromosome E1. Gen-

ome-wide TDT, sib-TDT, and case–control analyses also

consistently supported significant association within thesame region on chromosome E1, which is homologous to

human chromosome 17. Using haplotype analysis, a

*1.3 Mb region was identified as highly associated forprogressive retinal atrophy in Persian cats. Several candi-

date genes within the region are reasonable candidates as a

potential causative gene and should be considered formolecular analyses.

Introduction

The eye is a highly complex organ comprised of several

highly specialized cells. The development, structure, and

function of the eye involves the interaction of thousands ofgenes. Genetic mutations in genes involving the eye are

likely to be detrimental to the fitness of cats, especially

random-bred cats. As of 2012, 232 genetic eye conditionshave been mapped to a genomic location in humans and

192 loci associated with vision abnormalities have beenElectronic supplementary material The online version of thisarticle (doi:10.1007/s00335-014-9517-z) contains supplementarymaterial, which is available to authorized users.

H. Alhaddad ! B. Gandolfi ! R. A. Grahn !C. B. Peterson ! L. A. LyonsDepartment of Population Health and Reproduction, School ofVeterinary Medicine, University of California - Davis, Davis,CA 95616, USA

H. AlhaddadCollege of Science, Kuwait University, 13060 Safat, Kuwait

B. Gandolfi ! L. A. Lyons (&)Department of Veterinary Medicine and Surgery, College ofVeterinary Medicine, University of Missouri-Columbia, E109Vet Med Building, 1600 E. Rollins St., Columbia, MO 65211,USAe-mail: lyonsla@missouri.edu

H.-C. RahCollege of Medicine, Chungbuk National University, Chongju,Chungbuk Province, South Korea

D. J. Maggs ! K. L. GoodDepartment of Surgical and Radiological Sciences, School ofVeterinary Medicine, University of California - Davis, Davis,CA 95616, USA

N. C. PedersenDepartment of Medicine and Epidemiology, School ofVeterinary Medicine, University of California - Davis, Davis,CA 95616, USA

123

Mamm Genome

DOI 10.1007/s00335-014-9517-z

(Fig. S3b). The most highly associated SNPs are shown in

Table 2. The genomic inflation of this modified dataset was

measured to be *1.3. Further reduction of genomicinflation was not possible due to the relatedness of

individuals in the pedigree. The average P̂ value as a

measure of relatedness of all individuals was *0.18 and,upon removal of the 20 genetically distant samples, the

average P̂ value was *0.24.

Table 2 Ten most associated markers obtained by genome-wide analyses of Persian PRA

No. Chr. SNP ID Position TDT sib-TDT Case–control

Praw Pgenome Praw Pgenome Praw Pgenome

1 E1 chrUn5.6839723 1831172 4.32E-08 1.00E-05 1.00E-05 0.1632 7.64E-13 1.00E-05

2 E1 chrUn5.6133983 1106562 1.21E-07 0.00013 0.00013 0.4071 1.52E-10 1.00E-05

3 E1 chrUn5.6766609 1751066 5.73E-07 0.00282 0.00282 0.5451 2.97E-09 5.00E-05

4 E1 chrUn5.6481762 1452354 7.74E-06 0.06134 0.06134 0.9057 2.70E-07 0.00226

5 E1 chrUn5.6503134 1476010 2.21E-05 0.1905 0.1905 0.9644 1.15E-06 0.00841

6 E1 chrUn5.5846986 808916 3.74E-05 0.3135 0.3135 0.9487 4.65E-06 0.03121

7 E1 chrUn5.6185290 1154788 3.74E-05 0.3135 0.3135 0.9487 8.04E-06 0.05255

8 E1 chrUn5.6912692 1912858 3.74E-05 0.3135 0.3135 0.9793 – –

9 E1 chrUn5.6942249 1932982 3.74E-05 0.3135 0.3135 0.9793 5.95E-06 0.03955

10 E1 chrUn5.7293975 2293660 3.74E-05 0.3135 0.3135 0.9487 8.76E-07 0.00658

11 F1 chrUn5.7948667 2997440 – – – – 4.45E-06 0.02986

Fig. 2 Genome-wide sib-TDT analysis of the Persian PRA. Upperplot represents the Praw values of the analysis, whereas the lower plotrepresents the genome-wide significant Pgenome values after 100,000

permutations. X-chromosome markers were removed in sib-TDTanalysis. Significant association is localized to cat chromosome E1

Genome-wide analyses of Persian PRA

123

Remapping and distances

• Remapping SNPs to genomic locations on genome assembly 8.0.

• Majority of SNPs are < 50Kb apart.

Mendelian errors

Percent Markers with Mendelian Errors

Num

ber o

f Ind

ivid

uals

0 2 4 6 8 10

020

4060

80100

a. Percent Mendelian Errors

Num

ber o

f SN

Ps

0 20 40 60 80

010000

20000

30000

40000

50000

60000

b.

A1 A2 A3 B1 B2 B3 B4 C1 C2 D1 D2 D3 D4 E1 E2 E3 F1 F2 UN X

Chromosome

No.

of S

NP

s

05

1015

2025

30

c. % Error (X-chromosome SNP Heterozygous Males)

No.

of S

NP

s

0 20 40 60 80 100

0200

400

600

800

1000

d.

Evaluating Mendelian inheritance in 83 Trios

Percent Markers with Mendelian Errors

Num

ber o

f Ind

ivid

uals

0 2 4 6 8 10

020

4060

80100

a. Percent Mendelian Errors

Num

ber o

f SN

Ps

0 20 40 60 80

010000

20000

30000

40000

50000

60000

b.

A1 A2 A3 B1 B2 B3 B4 C1 C2 D1 D2 D3 D4 E1 E2 E3 F1 F2 UN X

Chromosome

No.

of S

NP

s

05

1015

2025

30

c. % Error (X-chromosome SNP Heterozygous Males)

No.

of S

NP

s

0 20 40 60 80 100

0200

400

600

800

1000

d.

SNP information update

• New SNP locations based on 8.0 cat genome assembly

• Removal of SNPs with low genotyping rate

• Removal/marking SNPs with significant Mendelian errors.

Final SNP-set

A1 A2 A3 B1 B2 B3 B4 C1 C2 D1 D2 D3 D4 E1 E2 E3 F1 F2 X UN

Chromosome

Num

ber o

f SN

Ps

01000

2000

3000

4000

5000

6000

7000

AssignedUn-assigned

(~ 1%)

IncludedExluded(~ 1%)

Less than 1% SNPs excluded due to low genotyping rate Less than 1% SNPs remains unassigned

Final SNP number is 62272

• Introduction • Genotype dataset • Feline array features • Array’s utility

• Population genetics of cats • GWAS • Selection and breed history

• Array’s utility • Population genetics of cats • GWAS • Selection and breed history

Outline

Monomorphic

ABY

ACURL

ASH

Asian

BEN

BIR

BOM

BSH

BUR

CHR

Colony

CREX

DREX

DOM

EGY

HAV

HYD

JBOB

MANEE

KOR

KBOB

PERM

LYK

MCOON

MANX

MUNCH

NFC OCI

ORI

PER

PBALD

RAG

RBLUE

SFOLD

SREX

SIA

SIR

SIN

SOM

SPH

TREX

ANG

VAN

ALC

WIR

BIGW

XPED

FSI

0

10000

20000

30000

40000

50000

60000

No.

SN

Ps

A1 A2 A3 B1 B2 B3 B4 C1 C2 D1 D2 D3 D4 E1 E2 E3 F1 F2 X

Pretty graph with low information value

Population structure

-0.10 -0.05 0.00 0.05 0.10 0.15

-0.06

-0.04

-0.02

0.00

0.02

0.04

Dimesnion 1

Dim

ensi

on 2

Persian

Selkirk Rex

British Shorthair

Scotish Fold

Manx

Siberian

Sphynx

Abyssinian

Turkish Angora

Turkish Van

J. Bobtail

Cornish Rex

Bengal

Ocicat

Oriental cats

Siamese

Burmese

Korat

Birman

a.

-0.10 -0.05 0.00 0.05 0.10 0.15

-0.10

-0.05

0.00

0.05

0.10

Dimension 1

Dim

ensi

on 3

b.-0.06 -0.04 -0.02 0.00 0.02 0.04

-0.10

-0.05

0.00

0.05

0.10

Dimension 2

Dim

ensi

on 3

c.

East-West breed genetic structure

Linkage disequilibrium

0 1 2 3 4

0.0

0.1

0.2

0.3

0.4

0.5

Pairwise-SNPs Distace (MB)

Mean Squared

Correlation Coefficient (r2) Random Bred

Max Random Bred (r2)

a. Extent of LD (Kb)

0 500 1000 1500

b.

PersianSlekirk Rex

British shorthairScottish FoldMaine CoonAbyssinian

Turkish VanBengal

MunchikanLaPerm

Devon RexRagdoll

American CurlSiberianSphynx

PeterbaldOrientalSiameseBurmeseBirmanEU wild

• LD ranges (50Kb - 1.5MB)

• Eastern breeds higher LD

• Local LD vs. genome-wide LD

Alhaddad, H., et al. (2013). "Extent of linkage disequilibrium in the domestic cat, Felis silvestris catus, and its breeds." PloS one 8(1): e53537.

Runs of homozygosity

Birman

Burmese

Siamese

Oriental

Peterbald

Sphynx

Siberian

American Curl

Ragdoll

Devon Rex

LaPerm

Munchikan

Bengal

Turkish Van

Abyssinian

Maine Coon

Scottish Fold

British shorthair

Slekirk Rex

Persian 28

8

25

14

17

81

4

9

3

5

25

20

9

4

8

28

94

129

238

177

a. Total Number

21

5

14

8

11

48

3

5

3

2

14

14

7

3

3

16

63

82

156

93

b. < 500 Kb

3

2

9

5

5

20

1

3

0

2

9

3

1

0

3

9

18

34

50

40

c. 500 Kb - 1000 Kb

4

1

2

1

1

13

0

1

0

1

2

3

1

1

2

3

13

13

32

44

d. > 1000 Kb

GWAS

Trait: Dilute color Population: Random bred Cases: 33 Controls: 81 Haplotype: NA Causative SNP: on chip

GWAS

Sample size = 114 r2 = 0.26 Distance = 4.4 Kb Needed sample size = 427

-2 -1 0 1 2

0.0

0.2

0.4

0.6

0.8

1.0

Distance (Mb)

Squared

Correlation Coefficient (r2)

Pritchard and Przeworski: Linkage Disequilibrium in Humans 9

regions may not be meaningful unless the local recom-bination rates are taken into account.

Recombination rates may also vary at a scale that isnot detected by markers separated by megabases (see Yuet al. 2001). Dramatic changes in recombination ratehave been reported over distances as short as a few kilo-bases (e.g., see Chakravarti et al. 1984). The molecularbasis of hotspots for recombination remains unknown,but contributing factors might include high GC content(Eisenbarth et al. 2000; Yu et al. 2001) or whether theregion is transcribed (Nicolas 1998). Thus, genetic mapsmay provide limited information about rates at shortscales; more-refined estimates can be obtained by meansof single-sperm typing (e.g., see Lien et al. 2000).

One known aspect of recombination not taken intoaccount by most models is homologous gene conversion(here, gene conversion and crossing-over are thought ofas alternative outcomes of a common recombinationmechanism). For markers a megabase or so apart, thecontribution of gene conversion to the overall level ofgenetic exchange is negligible (Andolfatto and Nordborg1998). As a result, genetic map–based estimates of therecombination rate are essentially estimates of thecrossing-over rates alone. The latter should accuratelypredict the extent of pairwise LD between polymor-phisms far apart (given an adequate demographicmodel). For closely linked markers, however, LD mayalso be broken up by gene conversion. Indeed, in Dro-sophila and yeast, it appears that the rate of initiationof recombination events resolved as gene conversionsand as crossovers is similar. Thus, pedigree-based esti-mates may substantially underestimate the total rate ofrecombination at small scales.

Currently, very little is known about gene conversionin humans. However, as noted above, there appears tobe less LD at small scales than would be expected fromestimates of crossing-over rates and observed levels ofdiversity. An intriguing explanation for this pattern isthat gene conversion is quite frequent in humans. Insupport of this, Przeworski and Wall (2001) show thatthe data that they have analyzed are more likely undera model in which two-thirds of recombination eventsare gene-conversion events than under a model ofcrossing-over alone. An analysis of 10 anonymous in-tergenic regions also finds evidence for extensive geneconversion (A. Di Rienzo, L. Frisse, and R. R. Hudson,personal communication).

The relationship between LD and distance might alsobe shaped by the segregation of inversions. Inversionpolymorphisms are thought to suppress recombinationin heterozygotes throughout much of the length of theinverted segment (Roberts 1976; Martin 1999), al-though the precise details are unknown. Thus, the pres-ence of inversion polymorphisms in a given region willreduce the rate of recombination. For instance, if 50%

of the individuals used to construct a genetic map areheterozygous for an inversion, and if there is no recom-bination in heterozygotes, within the inverted region,then the average recombination rate in the region willbe halved.

Moreover, inversion polymorphism can potentiallyhave a second, much stronger effect on the extent of LD.Because recombination between the standard and in-verted types is rare or absent, strong LD can developbetween the two kinds of chromosomes. In the extremecase where there is complete suppression of recombi-nation, a mutation within the inverted region arises onone type of chromosome and cannot move to the othervia recombination. As mutations accumulate on bothgenetic backgrounds, the two arrangements diverge, po-tentially leading to a buildup of substantial LD. Theextent of the effect will depend on the history and fre-quency of the inversion—including what kind of naturalselection, if any, is acting on the inversion (see Andol-fatto et al. 2001).

Little is known about the number, size, and frequencyof inversions in the human genome. In particular, in-versions shorter than a few megabases are currently dif-ficult to detect by cytological methods but, if they wereto reach intermediate frequencies, could have a sub-stantial impact on the extent of LD. There are now anumber of findings of common inversion polymor-phisms: several studies of disease-associated inversionshave reported inversion-heterozygote frequencies of21%–33% in controls (Small et al. 1997; Saunier et al.2000; Giglio et al. 2001). The length of the invertedsegments varies from ∼50 kb (Small et al. 1997) to 3Mb (J. Weber, personal communication), long enoughfor these rearrangements to potentially contribute to is-lands of extensive LD.

It appears that many of these chromosome rearrange-ments are mediated by nonhomologous meioticexchange between inverted repeats (Small et al. 1997;Saunier et al. 2000; Giglio et al. 2001; Tilford et al.2001). Since much of the human genome consists ofrepetitive DNA, it is possible that chromosomal rear-rangements resulting in inversion polymorphisms arefairly common.

Implications for LD Mapping

The recent interest in LD in humans is due in large partto the prospect of large-scale association studies to locatecomplex disease genes. Risch and Merikangas (1996)showed that, under ideal circumstances, the power todetect disease mutations of small effect is much greaterwith association mapping than it is with linkage analysis.On the basis of this result, they argued that the futureof complex-disease genetics lies in the use of genomewidescreens of association.

10 Am. J. Hum. Genet. 69:1–14, 2001

r2 and the Power of Association Studies

Several recent articles have referred to the connection betweenvarious measures of LD and the power of association studies(Kruglyak 1999; Dunning et al. 2000; Abecasis et al. 2001).Here we clarify this relationship.

Suppose that we genotype a case-control sample of N1 in-dividuals at locus 1, a (true) disease-susceptibility locus, andthat we genotype N2 individuals at locus 2, a nearby markerlocus. We want to compare the power of association tests atthese two loci.

Assume that both loci are biallelic, with alleles A and a atlocus 1 and with alleles B and b at locus 2. Let pDA and pCA

be the frequencies of allele A in individuals with the diseaseand in controls, respectively, and let pDB and pCB be the anal-ogous frequencies at locus 2. Let qAB (respectively, qaB) be theprobability that a chromosome with allele A (respectively, allelea) at locus 1 has the B allele at locus 2. Then,

p ! p p (p ! p )(q ! q ) .DB CB DA CA AB aB

The standard x2 test statistic of association at locus 1 (call this“ ”) can be written as2X1

2ˆ ˆ(p ! p ) 2N f(1 ! f)DA CA 12X p ,1 ˆ ˆp (1 ! p )A A

where , , and are the sample frequencies of A inˆ ˆ ˆp p pDA CA A

affected individuals, in controls, and in the overall sample, re-spectively, and where f and are the fractions of the sample1 ! fthat are cases and controls, respectively. The test statistic forassociation at locus 2 ( ) is similar—but with B in place of2X2

A and with N2 in place of N1. The distributions of and2X1

are approximately the squares of normal random variables,2X2

with means

1

22N f(1 ! f)1(p ! p )DA CA [ ]p̄ (1 ! p̄ )A A

and

1

22N f(1 ! f)2(p ! p )(q ! q ) ,DA CA AB aB [ ]p̄ (1 ! p̄ )B B

where

p̄ p fp " (1 ! f)p ≈ p ,A DA CA A

and similarly for , and where the variances are ≈1 if thep̄B

difference, in frequency of A, between cases and controls issmall. Then, since

2 2 !1 !1r p (q ! q ) p (1 ! p )p (1 ! p ) ,AB aB A A B B

it follows that the distributions of and are approximately2 2X X1 2

the same if . In other words, to achieve (approxi-2N p N /r2 1

mately) the same power at the marker locus as is achieved atthe susceptibility locus, the sample size must be increased by afactor of 1/r2.

However, the practical aspects of genomewide asso-ciation mapping are currently daunting. It is clear thatthe number of markers that will be needed to scan thegenome for association is very large. Recent progresshas expanded the set of available SNPs across the ge-nome (International SNP Map Working Group 2001),but the costs of genotyping a large sample of cases andcontrols at sufficient marker density would still be ex-tremely high (although dropping). Clearly, we needgood estimates of the marker density that will be re-quired to achieve acceptable power in these studies.

The required density will depend on which statisticaltests are used to detect association. Currently, it is mostcommon to test for association at each marker in turn(or, sometimes, by combining pairs of nearby markers).In what follows, we consider the case in which only onemarker is used, while noting that there seems to be agreat need for the development of multilocus tests ofassociation that make use of haplotype information,since these might prove to be much more efficient (seethe Nordborg web site).

Density of Markers

Suppose that we test for association at a marker locusthat is near a disease-susceptibility mutation. It can beshown that, in order to achieve roughly the same powerat the marker locus as we would have if we could testthe disease mutation itself, we need to increase the sam-ple size by a factor of 1/r2, where r2 is the coefficient ofLD between the marker and the disease mutation (seethe sidebar; also see Kruglyak 1999). Hence, for smallvalues of r2, there is little power to detect association atthe marker locus.

In a highly influential paper, Kruglyak ran simulationsof the coalescent with recombination to predict the rateof decay of LD; he predicted that “a useful level of LDis unlikely to extend beyond an average distance ofroughly 3 kb in the general population” (Kruglyak 1999,p. 139). None of his models predicted “useful” LD over130 kb. His criterion for useful LD was that the samplesize necessary to detect association at the marker shouldnot be increased more than 10-fold (in our terms, thiscorresponds to ). (The formal criterion used by2r # .1Kruglyak (1999) is slightly different from ours: 2d #

.) The predictions of more-realistic models of growth.1are less drastic (see fig. 2).

As data on LD among SNPs become available, we canstart to get an empirical sense of the rate of decay of r2.Figure 3 shows plots of for SNPs in five regions. If wer̂make the assumption that the distribution of r2 betweentwo random SNPs is the same as that between a SNPand a disease mutation, then we can use plots such asthose in figure 3 to study the decay of r2 directly (seeDunning et al. 2000; Taillon-Miller et al. 2000; Abecasis

GWAS

Trait: Long hair Population: LaPerm breed Cases: 32 Controls: 22 Haplotype: ~ 150 Kb Causative SNP: on chip

GWAS

-3 -2 -1 0 1 2 3

0.0

0.2

0.4

0.6

0.8

1.0

Distance (Mb)

Squared

Correlation Coefficient (r2)

Sample size = 54 r2 = 0.81 Distance = 1 Kb Needed sample size = 66

GWAS

Trait: Point color Population: Persian breed Cases: 21 Controls: 28 Haplotype: ~ 1Mb Causative SNP: on chip

GWAS

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6

0.0

0.2

0.4

0.6

0.8

1.0

Distance (Mb)

Squared

Correlation Coefficient (r2)

Sample size = 49 r2 = 0.98 Distance = 55 Kb Needed sample size = 50

GWAS

Trait: Orange color Population: Random bred Cases: 24 Controls: 69 Haplotype: ~ 1.5 Mb Causative variant unknown

GWAS

First WNK4-Hypokalemia Animal Model Identified byGenome-Wide Association in Burmese CatsBarbara Gandolfi1 , Timothy J. Gruffydd-Jones2, Richard Malik3, Alejandro Cortes1, Boyd R. Jones4,

Chris R. Helps5, Eva M. Prinzenberg6, George Erhardt6, Leslie A. Lyons1

1 Department of Population Health and Reproduction, University of California Davis, Davis, California, United States of America, 2 The Feline Centre, University of Bristol,

Langford, Bristol, United Kingdom, 3 Centre for Veterinary Education, University of Sydney, Sydney, Australia, 4 Institute of Veterinary, Animal & Biomedical Sciences,

Massey University, Palmerston North, New Zealand, 5 Molecular Diagnostic Unit, University of Bristol, Langford, Bristol, United Kingdom, 6 Institute of Animal Breeding &

Genetics, Justus Liebig University, Giessen, Germany

Abstract

Burmese is an old and popular cat breed, however, several health concerns, such as hypokalemia and a craniofacial defect,are prevalent, endangering the general health of the breed. Hypokalemia, a subnormal serum potassium ion concentration([K+]), most often occurs as a secondary problem but can occur as a primary problem, such as hypokalaemic periodicparalysis in humans, and as feline hypokalaemic periodic polymyopathy primarily in Burmese. The most characteristic clinicalsign of hypokalemia in Burmese is a skeletal muscle weakness that is frequently episodic in nature, either generalized, orsometimes localized to the cervical and thoracic limb girdle muscles. Burmese hypokalemia is suspected to be a single locusautosomal recessive trait. A genome wide case-control study using the illumina Infinium Feline 63K iSelect DNA array wasperformed using 35 cases and 25 controls from the Burmese breed that identified a locus on chromosome E1 associatedwith hypokalemia. Within approximately 1.2 Mb of the highest associated SNP, two candidate genes were identified, KCNH4and WNK4. Direct sequencing of the genes revealed a nonsense mutation, producing a premature stop codon within WNK4(c.2899C.T), leading to a truncated protein that lacks the C-terminal coiled-coil domain and the highly conserved Akt1/SGKphosphorylation site. All cases were homozygous for the mutation. Although the exact mechanism causing hypokalemiahas not been determined, extrapolation from the homologous human and mouse genes suggests the mechanism mayinvolve a potassium-losing nephropathy. A genetic test to screen for the genetic defect within the active breedingpopulation has been developed, which should lead to eradication of the mutation and improved general health within thebreed. Moreover, the identified mutation may help clarify the role of the protein in K+ regulation and the cat represents thefirst animal model for WNK4-associated hypokalemia.

Citation: Gandolfi B, Gruffydd-Jones TJ, Malik R, Cortes A, Jones BR, et al. (2012) First WNK4-Hypokalemia Animal Model Identified by Genome-Wide Associationin Burmese Cats. PLoS ONE 7(12): e53173. doi:10.1371/journal.pone.0053173

Editor: Yann Herault, IGBMC/ICS, France

Received August 14, 2012; Accepted November 26, 2012; Published December 28, 2012

Copyright: ! 2012 Gandolfi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This work was supported by funding from the National Center for Research Resources R24 RR016094 and is currently supported by the Office ofResearch Infrastructure Programs/OD R24OD010928, the Cat Health Network grant D12FE-508, and the Center for Companion Animal Health, School of VeterinaryMedicine, University of California, Davis. Richard Malik is supported by the Valentine Charlton Bequest from the University of Sydney. Support for thedevelopment of the Illumina Infinium Feline 63K iSelect DNA array was provided by the Morris Animal Foundation (http://www.morrisanimalfoundation.org) via adonation from Hill’s Pet Food, Inc. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

* E-mail: lalyons@ucdavis.edu

Introduction

Potassium is the most abundant cation in mammals [1,2]. Theresting membrane potential of cells is affected by the relationshipbetween intracellular and extracellular potassium concentrationsand the resting potassium conductance [3]. Since the extracellularpotassium greatly affects the tendency of cells to fire actionpotentials, potassium plays a crucial role in the function of nervoustissue and muscle (skeletal, cardiac and smooth) throughout thebody [1,2], implying perturbations can be debilitating or even life-threatening. To maintain ideal body homeostasis, potassiumexcretion and dietary intake must be balanced [4–6]. Abnormal-ities of potassium homeostasis can occur as a primary condition, oras a secondary disorder [7–10]. Inherited hypokalemia has beendiscovered in a variety of mammals by genetic studies ofindividuals and families affected by clinical disease. A classicsyndrome of myopathic weakness, hyperkalemic periodic paralysis

(HYPP), has been defined genetically in humans [11–14] andhorses [15–18]. Genetic studies have shown that in humans, HYPPis attributable to a channelopathy associated with abnormalsodium conductance, usually inherited as an autosomal dominanttrait. Primary hypokalemic disorders have been documented inhumans, manifesting as episodic weakness associated with lowserum potassium [19].

Feline hypokalaemic periodic paralysis or Burmese hypokalaemic periodicpolymyopathy (BHP) has been recognized since the seminal casesdescribed by Blaxter and colleagues [20]. The disease ischaracterized by muscle weakness associated with intermittenthypokalemia [21]. Genetic studies suggest an autosomal recessivecondition in Burmese cats [20]. Blaxter [20] provided a fulldescription of the condition in the Burmese cat breed of theUnited Kingdom, Jones and collaborators recorded similarfindings within Burmese cats from New Zealand [22], whileMason and Lantinga documented the condition in cats in

PLOS ONE | www.plosone.org 1 December 2012 | Volume 7 | Issue 12 | e53173

*

Aristaless-Like Homeobox protein 1 (ALX1) variant associated withcraniofacial structure and frontonasal dysplasia in Burmese cats

Leslie A. Lyons a,f,n, Carolyn A. Erdman b,f, Robert A. Grahn c,f, Michael J. Hamilton d,f,Michael J. Carter e,f, Christopher R. Helps g, Hasan Alhaddad h, Barbara Gandolfi a,f

a Department of Veterinary Medicine & Surgery, College of Veterinary Medicine, University of Missouri-Columbia, Columbia, MO 65211, USAb Department of Psychiatry, University of California-San Francisco, San Francisco, CA 94143, USAc Veterinary Genetics Laboratory, School of Veterinary Medicine, University of California-Davis, Davis, CA 96516, USAd Department of Cell Biology and Neuroscience, Institute for Integrative Genome Biology, Center for Disease Vector Research, University of California-Riv-erside, Riverside, CA 92521, USAe MDxHealth Inc, 15279 Alton Parkway, Suite #100, Irvine, CA 92618, USAf Department of Population Health and Reproduction, School of Veterinary Medicine, University of California-Davis, Davis, CA 95776, USAg Langford Veterinary Services, University of Bristol, Bristol BS40 5DU, UKh College of Science, Kuwait University, Safat, Kuwait

a r t i c l e i n f o

Article history:Received 2 October 2015Received in revised form3 November 2015Accepted 20 November 2015

Keywords:Cartilage homeo protein 1CART1Domestic catFacial developmentFrontonasal dysplasiaFNDFelis silvestris catus

a b s t r a c t

Frontonasal dysplasia (FND) can have severe presentations that are medically and socially debilitating.Several genes are implicated in FND conditions, including Aristaless-Like Homeobox 1 (ALX1), which isassociated with FND3. Breeds of cats are selected and bred for extremes in craniofacial morphologies. Inparticular, a lineage of Burmese cats with severe brachycephyla is extremely popular and is termedContemporary Burmese. Genetic studies demonstrated that the brachycephyla of the ContemporaryBurmese is a simple co-dominant trait, however, the homozygous cats have a severe craniofacial defectthat is incompatible with life. The craniofacial defect of the Burmese was genetically analyzed over a 20year period, using various genetic analysis techniques. Family-based linkage analysis localized the trait tocat chromosome B4. Genome-wide association studies and other genetic analyses of SNP data refined acritical region. Sequence analysis identified a 12 bp in frame deletion in ALX1, c.496delCTCTCAGGACTG,which is 100% concordant with the craniofacial defect and not found in cats not related to the Con-temporary Burmese.

& 2015 Published by Elsevier Inc.

1. Introduction

Frontonasal dysplasia (FND) or median cleft syndrome is aheterogeneous group of disorders that describes an array of ab-normalities affecting development of the maxilla-facial structuresand the skull. The prevalence of FND is unknown and is considereda rare or “orphan” disease (ORPHA no.: ORPHA250), however af-fected children can have severe presentations that are life-longmedically and socially debilitating. Three genes have been im-plicated in FND conditions. Aristaless-Like Homeobox 1 (ALX1)(OMIM:601527) is associated with FND3, which was defined inthree Turkish sibs of consanguineous parents (Uz et al., 2010).ALX1 is also known as Cartilage homeoprotein-1 (CART1) (Zhaoet al., 1993), which has been demonstrated to cause neural tube

defects in mice (Zhao et al., 1996), presenting as acrania andmeroanencephaly in mice.

Domesticated animals are often selected for craniofacial var-iants that become breed defining traits. Conditions that would beconsidered abnormalities or severe craniofacial defects in humansare desired phenotypes in cats and dogs, thus companion animalsare excellent models for human facial development due to theirpopularity. Many dog and cat breeds are bred for brachycephaly,which is assumed to be preferred due to its neotenic effect on theanimal's face. In dogs, the definition of brachycephaly has beenquantified by morphological measurements (Huber and Lups,1968; Koch et al., 2012; Regodon et al., 1991; Schmidt et al., 2011)and two genes have been implicated for affecting head type(Haworth et al., 2001; Hunemeier et al., 2009; Schoenebeck et al.,2012). The health concerns associated with canine brachycephalyhave come under strong veterinary and public scrutiny (Kruijsenand Wayop, 2011; Oechtering et al., 2010; Roberts et al., 2010),suggesting severe modifications to breeding programs to alleviatethe extent of brachycephaly.

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/developmentalbiology

Developmental Biology

http://dx.doi.org/10.1016/j.ydbio.2015.11.0150012-1606/& 2015 Published by Elsevier Inc.

n Correspondence to: Department of Veterinary Medicine & Surgery, College ofVeterinary Medicine, University of Missouri-Columbia, E109 Vet Med Building,1600 E. Rollins St., Columbia, MO 65211, USA.

E-mail address: Lyonsla@Missouri.edu (L.A. Lyons).

Please cite this article as: Lyons, L.A., et al., Aristaless-Like Homeobox protein 1 (ALX1) variant associated with craniofacial structure andfrontonasal dysplasia in Burmese cats. Dev. Biol. (2015), http://dx.doi.org/10.1016/j.ydbio.2015.11.015i

Developmental Biology ∎ (∎∎∎∎) ∎∎∎–∎∎∎

SHORT COMMUNICATION

COLQ variant associated with Devon Rex and Sphynx felinehereditary myopathy

Barbara Gandolfi1, Robert A. Grahn2, Erica K. Creighton1, D. Colette Williams3, Peter J. Dickinson4,

Beverly K. Sturges4, Ling T. Guo4, G. Diane Shelton5, Peter A. J. Leegwater6, Maria Longeri7,

Richard Malik8 and Leslie A. Lyons1

1Department of Veterinary Medicine and Surgery, College of Veterinary Medicine, University of Missouri – Columbia, Columbia, MO

65211, USA. 2Veterinary Genetics Laboratory, School of Veterinary Medicine, University of California – Davis, Davis, CA 95616, USA. 3The

William R. Pritchard Veterinary Medical Teaching Hospital, School of Veterinary Medicine, University of California – Davis, Davis, CA 95616,

USA. 4Department of Surgical and Radiological Sciences, School of Veterinary Medicine, University of California – Davis, Davis, CA 95616,

USA. 5Department of Pathology, University of California – San Diego, La Jolla, CA 92093, USA. 6Department of Clinical Sciences of

Companion Animals, Faculty of Veterinary Medicine, Utrecht University, 3508 TD, Utrecht, The Netherlands. 7Dipartimento di Scienze

Veterinarie e Sanit!a Pubblica, University of Milan, Milan, Italy. 8Centre for Veterinary Education, University of Sydney, Sydney, NSW 2006,

Australia.

Summary Some Devon Rex and Sphynx cats have a variably progressive myopathy characterized by

appendicular and axial muscle weakness, megaesophagus, pharyngeal weakness and

fatigability with exercise. Muscle biopsies from affected cats demonstrated variable

pathological changes ranging from dystrophic features to minimal abnormalities. Affected

cats have exacerbation of weakness following anticholinesterase dosing, a clue that there is

an underlying congenital myasthenic syndrome (CMS). A genome-wide association study

and whole-genome sequencing suggested a causal variant for this entity was a c.1190G>Avariant causing a cysteine to tyrosine substitution (p.Cys397Tyr) within the C-terminal

domain of collagen-like tail subunit (single strand of homotrimer) of asymmetric acetyl-

cholinesterase (COLQ). Alpha-dystroglycan expression, which is associated with COLQ

anchorage at the motor end-plate, has been shown to be deficient in affected cats. Eighteen

affected cats were identified by genotyping, including cats from the original clinical

descriptions in 1993 and subsequent publications. Eight Devon Rex and one Sphynx not

associated with the study were identified as carriers, suggesting an allele frequency of ~2.0%in Devon Rex. Over 350 tested cats from other breeds did not have the variant.

Characteristic clinical features and variant presence in all affected cats suggest a model for

COLQ CMS. The association between the COLQ variant and this CMS affords clinicians the

opportunity to confirm diagnosis via genetic testing and permits owners and breeders to

identify carriers in the population. Moreover, accurate diagnosis increases available

therapeutic options for affected cats based on an understanding of the pathophysiology and

experience from human CMS associated with COLQ variants.

Keywords collagen-like tail subunit of asymmetric acetylcholinesterase, congenital myas-

thenic syndrome, domestic cat, Felis catus silvestris

Neuromuscular disorders encompass a variety of diseases

that impair normal function of skeletal muscle. A subset

of neuromuscular disorders, the congenital myasthenic

syndromes (CMSs), represent a heterogeneous group of

heritable diseases caused by abnormal signal transmission

at the motor endplate (EP), a synaptic connection between

motor axon nerve terminals and skeletal muscle fibers (see

review: Engel et al. 2015). Most CMSs are autosomal

recessive conditions characterized by functional or struc-

tural abnormalities of proteins localized to the presynaptic,

synaptic or postsynaptic regions of the motor EP. At least 18

genes are associated with genetically distinct forms of CMS

(http://www.ncbi.nlm.nih.gov/omim/?term=congenital+

Address for correspondence

L. A. Lyons, Department of Veterinary Medicine & Surgery, College ofVeterinary Medicine, E109 Vet Med Building, 1600 E. Rollins St.,University of Missouri – Columbia, Columbia, MO 65211, USA.E-mail: Lyonsla@Missouri.edu

Accepted for publication 30 July 2015

doi: 10.1111/age.12350

1© 2015 The Authors. Animal Genetics published byJohn Wiley & Sons Ltd on behalf of Stichting International Foundation for Animal Genetics.This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use anddistribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.

Genetic Susceptibility to Feline Infectious Peritonitis in Birman Cats

Lyudmila Golovkoa, Leslie A. Lyonsb, Hongwei Liua, Anne Sorensenc, Suzanne Wehnertc, and Niels C. Pedersena,1

aCenter for Companion Animal Health, School of Veterinary Medicine, University of California, One Shields Avenue, Davis, CA 95616, USAbDepartment of Population Health and Reproduction, University of California, One Shields Avenue, Davis, CA 95616, USAcFasanvejens Dyreklinik, Sondre Fasanvej 93, DK 2500 Valby Denmark

AbstractGenetic factors are presumed to influence the incidence of feline infectious peritonitis (FIP), especially among pedigreed cats. However, proof for the existence of such factors has been limited and mainly anecdotal. Therefore, we sought evidence for genetic susceptibility to FIP using feline high density single nucleotide polymorphism (SNP) arrays in a genome-wide association study (GWAS). Birman cats were chosen for GWAS because they are highly inbred and suffer a high incidence of FIP. DNA from 38 Birman cats that died of FIP and 161 healthy cats from breeders in Denmark and USA were selected for genotyping using 63K SNPs distributed across the feline genome. Danish and American Birman cats were closely related and the populations were therefore combined and analyzed in two manners: 1) all cases (FIP) vs. all controls (healthy) regardless of age, and 2) cases 1–1/2 years of age and younger (most susceptible) vs. controls 2 years of age and older (most resistant). GWAS of the second cohort was most productive in identifying significant genome-wide associations between case and control cats. Four peaks of association with FIP susceptibility were identified, with two being identified on both analyses. Five candidate genes ELMO1, RRAGA, TNFSF10, ERAP1 and ERAP2, all relevant to what is known about FIP virus pathogenesis, were identified but no single association was fully concordant with the disease phenotype. Difficulties in doing GWAS in cats and interrogating complex genetic traits were discussed.

KeywordsBirman cats; feline infectious peritonitis; genetics; susceptibility

© 2013 Elsevier B.V. All rights reserved.*Corresponding author: tel.: +1 530 752 7402; fax: +1 530 752 7701 ncpedersen@ucdavis.edu.1213 CCAH Bldg., University of California, One Shields Avenue, Davis, CA, USA 95616Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

NIH Public AccessAuthor ManuscriptVirus Res. Author manuscript; available in PMC 2015 February 27.

Published in final edited form as:Virus Res. 2013 July ; 175(1): 58–63. doi:10.1016/j.virusres.2013.04.006.

NIH-PA Author Manuscript

NIH-PA Author Manuscript

NIH-PA Author Manuscript

A splice variant in KRT71 is associatedwith curly coat phenotype of Selkirk RexcatsBarbara Gandolfi1, Hasan Alhaddad1, Shannon E. K. Joslin1, Razib Khan1, Serina Filler2, Gottfried Brem2

& Leslie A. Lyons1

1Department of Population Health and Reproduction, School of Veterinary Medicine, University of California - Davis, Davis, CA USA,2Institute of Animal Breeding and Genetics, Department for Biomedical Sciences, University of Veterinary Medicine - Vienna, Vienna,Austria.

One of the salient features of the domestic cat is the aesthetics of its fur. The Selkirk Rex breed is defined byan autosomal dominant woolly rexoid hair (ADWH) abnormality that is characterized by tightly curled hairshafts. A genome-wide case – control association study was conducted using 9 curly coated Selkirk Rex and29 controls, including straight-coated Selkirk Rex, British Shorthair and Persian, to localize the Selkirkautosomal dominant rexoid locus (SADRE). Although the control cats were from different breed lineages,they share recent breeding histories and were validated as controls by Bayesian clustering,multi-dimensional scaling and genomic inflation. A significant association was found on cat chromosomeB4 (Praw 5 2.87 3 10211), and a unique haplotype spanning ,600 Kb was found in all the curly coated cats.Direct sequencing of four candidate genes revealed a splice site variant within the KRT71 gene associatedwith the hair abnormality in Selkirk Rex.

Body homeostasis and tissue integration are supported by the hair, a highly keratinized tissue produced in thehair follicle (HF). Hair formation in mammalian HFs occurs during embryogenesis through a series ofreciprocal interactions between skin epithelial and underlying dermal cells1,2. The HF undergoes dynamic

cell kinetics composed of the anagen (active growth) phase, the catagen (transition) phase, and the telogen(resting) phase3. Among the skin appendages, the HF has a highly complex structure with eight distinct celllayers, where hundreds of gene products play key roles in hair cycle and maintenance. The hair shaft is sur-rounded and supported by the inner root sheath, the companion layer, and the outer root sheath4. The formationof a rigid structure during the HF differentiation is due to keratin proteins that are abundantly and differentiallyexpressed2,4,5.

A mammal’s pelage is generally one of its most noticeable attributes and is aesthetically pleasing. In cats, coatcolor and pelage types are often selected as a specific trait to develop a breed. The coat of a normal cat consiststhree hair types: long and straight guard hairs of uniform diameter, thinner awn hairs, and the fine undulatingdown hairs of uniform thickness6. Rexoid (curly / woolly) pelage is an easily recognized trans-species anomaly;detailed studies in various mammalian species, including mice7,8, chicken9, rat10, dog11 and human12–15, haveidentified causative genes and mutations. Nine rexoid-type pelage phenotypes are known within the domesticcat16–20 and recently, significant advances have been made toward the identification of these feline hair abnor-malities. One study revealed two alleles (KRT71re and KRT71hr) within KRT71, a crucial gene for keratinization inthe HF, which are responsible for recessive hypotricosis Hairless (Hr, hr) locus of the Sphynx breed and the Rexhair locus (Re, re) of the curly coated Devon Rex breed19. A second rexoid locus (R, r) with a mutation withinP2RY5 is responsible for the autosomal recessive woolly hair in the Cornish Rex breed (Gandolfi 2013, in press).For each of these autosomal recessive rexoid / woolly hair conditions, the identified mutations are responsible fora major change in the hair follicle, altering hair formation. Several dominant rexoid / woolly hair conditionsdefine other breeds, such as Selkirk Rex, LaPerm and American Wirehair, and these mutations await identifica-tion and characterization.

The Selkirk Rex breed harbors the most recently derived rexoid coat mutation that has successfully beendeveloped into an internationally recognized breed in the domestic cat fancy. The breed is proposed to haveoriginated from a de novo mutation in a domestic cat in 1987, and a previous study estimates that an average ofonly 8.4 generations elapsed since the occurrence of the Selkirk Rex rexoid mutation21. To maintain genetic

OPEN

SUBJECT AREAS:GENOME-WIDE

ASSOCIATION STUDIES

RNA SPLICING

SKIN MODELS

HAPLOTYPES

Received

17 December 2012

Accepted

22 May 2013

Published

17 June 2013

Correspondence andrequests for materials

should be addressed toB.G. (bgandolfi@

ucdavis.edu)

SCIENTIFIC REPORTS | 3 : 2000 | DOI: 10.1038/srep02000 1

GWAS

Please cite this article in press as: Pedersen, N.C., et al., The influence of age and genetics on natu-ral resistance to experimentally induced feline infectious peritonitis. Vet. Immunol. Immunopathol. (2014),http://dx.doi.org/10.1016/j.vetimm.2014.09.001

ARTICLE IN PRESSG ModelVETIMM-9250; No. of Pages 8

Veterinary Immunology and Immunopathology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Veterinary Immunology and Immunopathology

j ourna l h omepa ge: www.elsev ier .com/ locate /vet imm

Research paper

The influence of age and genetics on natural resistance toexperimentally induced feline infectious peritonitis

Niels C. Pedersena,∗, Hongwei Liua, Barbara Gandolfib,c, Leslie A. Lyonsb,c

a Center for Companion Animal Health, School of Veterinary Medicine, University of California—Davis, One Shields Avenue, Davis, CA95616, USAb Department of Population Health & Reproduction, School of Veterinary Medicine, University of California, Davis, One Shields Avenue,Davis, CA 95616, USAc Department of Veterinary Medicine and Surgery, College of Veterinary Medicine, University of Missouri, Columbia, Columbia, MO65211, USA

a r t i c l e i n f o

Article history:Received 20 May 2014Received in revised form 12 August 2014Accepted 8 September 2014

Keywords:Feline infectious peritonitisExperimentalNatural immunityAge resistanceGenetic resistanceGWAS

a b s t r a c t

Naturally occurring feline infectious peritonitis (FIP) is usually fatal, giving the impressionthat immunity to the FIP virus (FIPV) is extremely poor. This impression may be incor-rect, because not all cats experimentally exposed to FIPV develop FIP. There is also a beliefthat the incidence of FIP may be affected by a number of host, virus, and environmentalcofactors. However, the contribution of these cofactors to immunity and disease incidencehas not been determined. The present study followed 111 random-bred specific pathogenfree (SPF) cats that were obtained from a single research breeding colony and experimen-tally infected with FIPV. The cats were from several studies conducted over the past 5 years,and as a result, some of them had prior exposure to feline enteric coronavirus (FECV) oravirulent FIPVs. The cats were housed under optimized conditions of nutrition, husbandry,and quarantine to eliminate most of the cofactors implicated in FIPV infection outcome andwere uniformly challenge exposed to the same field strain of serotype 1 FIPV. Forty of the111 (36%) cats survived their initial challenge exposure to a Type I cat-passaged field strainsof FIPV. Six of these 40 survivors succumbed to FIP to a second or third challenge exposure,suggesting that immunity was not always sustained. Exposure to non-FIP-inducing felinecoronaviruses prior to challenge with virulent FIPV did not significantly affect FIP incidencebut did accelerate the disease course in some cats. There were no significant differences inFIP incidence between males and females, but resistance increased significantly between6 months and 1 or more years of age. Genetic testing was done on 107 of the 111 infectedcats. Multidimensional scaling (MDS) segregated the 107 cats into three distinct familiesbased primarily on a common sire(s), and resistant and susceptible cats were equally dis-tributed within each family. Genome-wide association studies (GWAS) on 73 cats that diedof FIP after one or more exposures (cases) and 34 cats that survived (controls) demonstratedfour significant associations after 100k permutations. When these same cats were analyzedusing a sib-pair transmission test, three of the four associations were confirmed althoughnot with genome-wide significance. GWAS was then done on three different age groups ofcases to take into account age-related resistance, and different associations were observed.

∗ Corresponding author. Tel.: +1 530 752 7402.E-mail address: ncpedersen@ucdavis.edu (N.C. Pedersen).

http://dx.doi.org/10.1016/j.vetimm.2014.09.0010165-2427/© 2014 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

Genome-wide association and linkage analyses localizea progressive retinal atrophy locus in Persian cats

Hasan Alhaddad • Barbara Gandolfi • Robert A. Grahn •

Hyung-Chul Rah • Carlyn B. Peterson • David J. Maggs •

Kathryn L. Good • Niels C. Pedersen • Leslie A. Lyons

Received: 19 February 2014 / Accepted: 3 April 2014 / Published online: 29 April 2014! The Author(s) 2014. This article is published with open access at Springerlink.com

Abstract Hereditary eye diseases of animals serve asexcellent models of human ocular disorders and assist in

the development of gene and drug therapies for inherited

forms of blindness. Several primary hereditary eye condi-tions affecting various ocular tissues and having different

rates of progression have been documented in domestic

cats. Gene therapy for canine retinopathies has been suc-cessful, thus the cat could be a gene therapy candidate for

other forms of retinal degenerations. The current study

investigates a hereditary, autosomal recessive, retinaldegeneration specific to Persian cats. A multi-generational

pedigree segregating for this progressive retinal atrophy

was genotyped using a 63 K SNP array and analyzed viagenome-wide linkage and association methods. A multi-

point parametric linkage analysis localized the blindness

phenotype to a *1.75 Mb region with significant LODscores (Z & 14, h = 0.00) on cat chromosome E1. Gen-

ome-wide TDT, sib-TDT, and case–control analyses also

consistently supported significant association within thesame region on chromosome E1, which is homologous to

human chromosome 17. Using haplotype analysis, a

*1.3 Mb region was identified as highly associated forprogressive retinal atrophy in Persian cats. Several candi-

date genes within the region are reasonable candidates as a

potential causative gene and should be considered formolecular analyses.

Introduction

The eye is a highly complex organ comprised of several

highly specialized cells. The development, structure, and

function of the eye involves the interaction of thousands ofgenes. Genetic mutations in genes involving the eye are

likely to be detrimental to the fitness of cats, especially

random-bred cats. As of 2012, 232 genetic eye conditionshave been mapped to a genomic location in humans and

192 loci associated with vision abnormalities have beenElectronic supplementary material The online version of thisarticle (doi:10.1007/s00335-014-9517-z) contains supplementarymaterial, which is available to authorized users.

H. Alhaddad ! B. Gandolfi ! R. A. Grahn !C. B. Peterson ! L. A. LyonsDepartment of Population Health and Reproduction, School ofVeterinary Medicine, University of California - Davis, Davis,CA 95616, USA

H. AlhaddadCollege of Science, Kuwait University, 13060 Safat, Kuwait

B. Gandolfi ! L. A. Lyons (&)Department of Veterinary Medicine and Surgery, College ofVeterinary Medicine, University of Missouri-Columbia, E109Vet Med Building, 1600 E. Rollins St., Columbia, MO 65211,USAe-mail: lyonsla@missouri.edu

H.-C. RahCollege of Medicine, Chungbuk National University, Chongju,Chungbuk Province, South Korea

D. J. Maggs ! K. L. GoodDepartment of Surgical and Radiological Sciences, School ofVeterinary Medicine, University of California - Davis, Davis,CA 95616, USA

N. C. PedersenDepartment of Medicine and Epidemiology, School ofVeterinary Medicine, University of California - Davis, Davis,CA 95616, USA

123

Mamm Genome (2014) 25:354–362

DOI 10.1007/s00335-014-9517-z

To the Root of the Curl: A Signature of a Recent SelectiveSweep Identifies a Mutation That Defines the CornishRex Cat BreedBarbara Gandolfi1*, Hasan Alhaddad1, Verena K. Affolter2, Jeffrey Brockman3, Jens Haggstrom4,

Shannon E. K. Joslin1, Amanda L. Koehne2, James C. Mullikin5, Catherine A. Outerbridge6,

Wesley C. Warren7, Leslie A. Lyons1

1 Department of Population Health and Reproduction, School of Veterinary Medicine, University of California - Davis, Davis, California, United States of America,

2 Department of Pathology, Microbiology, Immunology, School of Veterinary Medicine, University of California - Davis, Davis, California, United States of America, 3 Hill’s

Pet Nutrition Center, Topeka, Kansas, United States of America, 4 Department of Clinical Sciences, Faculty of Veterinary Medicine and Animal Science, Swedish University

of Agricultural Sciences, Uppsala, Sweden, 5 Comparative Genomics Unit, Genome Technology Branch, National Human Genome Research Institute, National Institutes of

Health, Bethesda, Maryland, United States of America, 6 Department of Veterinary Medicine & Epidemiology, School of Veterinary Medicine, University of California - Davis,

Davis, California, United States of America, 7 The Genome Institute, Washington University School of Medicine, St. Louis, Missouri, United States of America

Abstract

The cat (Felis silvestris catus) shows significant variation in pelage, morphological, and behavioral phenotypes amongst itsover 40 domesticated breeds. The majority of the breed specific phenotypic presentations originated through artificialselection, especially on desired novel phenotypic characteristics that arose only a few hundred years ago. Variations in coattexture and color of hair often delineate breeds amongst domestic animals. Although the genetic basis of several feline coatcolors and hair lengths are characterized, less is known about the genes influencing variation in coat growth and texture,especially rexoid – curly coated types. Cornish Rex is a cat breed defined by a fixed recessive curly coat trait. Genome-wideanalyses for selection (di, Tajima’s D and nucleotide diversity) were performed in the Cornish Rex breed and in 11phenotypically diverse breeds and two random bred populations. Approximately 63K SNPs were used in the analysis thataimed to localize the locus controlling the rexoid hair texture. A region with a strong signature of recent selective sweepwas identified in the Cornish Rex breed on chromosome A1, as well as a consensus block of homozygosity that spansapproximately 3 Mb. Inspection of the region for candidate genes led to the identification of the lysophosphatidic acidreceptor 6 (LPAR6). A 4 bp deletion in exon 5, c.250_253_delTTTG, which induces a premature stop codon in the receptor,was identified via Sanger sequencing. The mutation is fixed in Cornish Rex, absent in all straight haired cats analyzed, and isalso segregating in the German Rex breed. LPAR6 encodes a G protein-coupled receptor essential for maintaining thestructural integrity of the hair shaft; and has mutations resulting in a wooly hair phenotype in humans.

Citation: Gandolfi B, Alhaddad H, Affolter VK, Brockman J, Haggstrom J, et al. (2013) To the Root of the Curl: A Signature of a Recent Selective Sweep Identifies aMutation That Defines the Cornish Rex Cat Breed. PLoS ONE 8(6): e67105. doi:10.1371/journal.pone.0067105

Editor: Arnar Palsson, University of Iceland, Iceland

Received March 26, 2013; Accepted May 14, 2013; Published June 27, 2013

Copyright: ! 2013 Gandolfi et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: This project was supported by the National Center for Research Resources and the Office of Research Infrastructure Programs of the National Instituteof Health through Grant Number R24 RR016094, the Winn Feline Foundation (W10-14, W11-041), the Center for Companion Animal Health at University ofCalifornia Davis (2010-09-F) (http://www.vetmed.ucdavis.edu/ccah/index.cfm), and the George and Phyllis Miller Feline Health Fund of the San FranciscoFoundation (2008-36-F). Support for the development of the Illumina Infinium Feline 63K iSelect DNA array was provided by the Morris Animal Foundation (http://www.morrisanimalfoundation.org) via a donation from Hill’s Pet Food, Inc. The funders had no role in study design, data collection and analysis, decision topublish, or preparation of the manuscript.

Competing Interests: JB works for a private company (Hill’s Pet Food, Inc) that partially sponsored the development of the 63k feline SNP array. The funder hadno role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

* E-mail: bgandolfi@ucdavis.edu

Introduction

Phenotypic traits under strong artificial selection within catbreeds vary from body types, muzzle shape, tail length toaesthetically pleasant traits, such as hair color, length and texture.Hair represents one of the defining characteristic of mammals.Hair provides body temperature regulation, protection fromenvironmental elements, and adaptive advantages of camouflage,as well as often having aesthetic value to humans. The hair folliclehas a highly complex structure with eight distinct cell layers, inwhich hundreds of gene products play a key role in the hair cyclemaintenance [1,2]. In the past decade, numerous genes expressed

in the hair follicle have been identified and mutations in some ofthese genes have been shown to underlie hereditary hair diseasesin humans and other mammals [3]. Hereditary hair diseases inmammals show diverse hair phenotypes, such as sparse or shorthairs (hypotrichosis), excessive or elongated hairs (hypertrichosis),and hair shaft anomalies, creating rexoid/woolly hairs [3–12].

Causative genes for the diseases encode various proteins withdifferent functions, such as structural proteins, transcriptionfactors, and signaling molecules. Mutations within structuralproteins, such as epithelial and hair keratins, are often associatedwith hair disease. To date, mutations in several hair keratin genesunderlined two hereditary hair disorders: monilethrix, character-

PLOS ONE | www.plosone.org 1 June 2013 | Volume 8 | Issue 6 | e67105

Selection

• Breed defining traits (mutations fixed)

• No controls (can’t perform case-control GWAS)!

• Investigate the breed history to identify breed defining mutation(s).

Beyond traits/diseases

0 1 2 3 4 5 10 15 20 25

01.68753.375

6.75

12.5

25

50

100

Generation Crossed to Domestic

% B

enga

lnes

s

TheoriticalAsian Leopard CatsBxLLxDBxDRandom Bengal Breed cats

% B

enga

lnes

s0

2550

75100

ALC LxD BEN ACURL BOM BUR CREX EGY MANEE PERM MCOON MUNCH ORI PBALD RBLUE SREX SIR SPH VAN

BxL BxD ABY BIR BSH CHR DREX JBOB KOR LYK MANX NFC PER RAG SFOLD SIA SOM TREX WIR

Findings:

1)  Department of Biological Science, Kuwait University, Kuwait 2)  College of Veterinary Medicine, University of Missouri-Columbia, Columbia, MO

3)  Department of Population Health & Reproduction, University of California - Davis, Davis, CA

Degree of Bengalness: A measure of the measure of genomic contribution of Asian Leopard Cats into Bengal breed cats

Mona Abdi1, H. Alhaddad1, B. Gandolfi2 R. Grahn3, and L. A. Lyons2

Dataset &

Analysis:

The idea &

Significance

•  The Bengal breed cats result from hybridizing domestic cats (DOM) and Asian Leopard Cats (ALC). •  First generation hybrids can be further crossed to domestic cats and still be considered Bengal cats as long as phenotypic and

behavioral characteristics retained (see cat below). •  The ALC genomic contribution into registered Bengal cats is usually unknown and likely variable. •  Significance of the study:

1.  Develop a panel of diagnostic markers to measure ALC genomic contribution in Bengal cats (degree of Bengalness). 2.  The panel can be used to study the genetics of the Asian Leopard cat and the ALC-DOM introgression zones in the wild.

Dataset: •  A total of 2161 cats (from 48 breeds/populations) genotyped using the 63K Feline SNP array were used.

•  ALC (N = 9) and domestic cats (N = 1765). •  Calculate the allele frequency for each marker in each group

independently. •  Select the markers that are fixed with allele 1 (A1) in ALC

and absent in domestic cats or have a minor allele frequency < 0.05.

•  Evaluate the positions of the selected markers. •  Select a subset of the identified markers with near uniform

inter-marker distances.

Objective 1: Identify diagnostic markers that are fixed for different alleles in the two groups (ALC & DOM)

1.  674 markers were identified as diagnostic markers and they were distributed in all 19 cat chromosomes (Fig. 1a).

2.  To avoid markers being in linkage disequilibrium, 287 were selected with an inter-marker distance ~ 5 Mb (Fig. 1b).

Fig.1: Distribution of ALC specific markers along cat chromosomes. (a) Relative positions of all markers identified. Many markers are close to one another and may be in linkage disequilibrium. (b) Relative position of a subset of markers from (a) where markers are ~ 5Mb apart.

SNP Relative Position (Mb)

A1A2A3B1B2B3B4C1C2D1D2D3D4E1E2E3F1F2X

0 25 50 75 100 125 150 175 200 225 250

a. SNP Relative Position (Mb)b.0 25 50 75 100 125 150 175 200 225 250

Objective 2: Estimate the genomic contribution of ALC in: (1) known pedigree, (2) Bengal breed, (3) Other cat breed

Dataset &

Analysis:

•  Diagnostic panel (All: 287 markers, Auto: 262 markers). •  ALC-DOM pedigree (N = 98), random Bengal cat (N = 98),

and 33 cat breeds (N = 1452). •  Calculate % of ALC alleles (% bengalness) in each individual

using autosomal markers only. •  Using known pedigree samples, estimate range of %

bengalness and validate the estimates using pedigree information.

Fig. 2: Genomic contribution of ALC. (a). Known pedigree crosses of ALC-DOM. LxD: first generation hybrid. BxL: backcross to ALC. BxD: backcross to DOM. BEN: Bengal cat unknown pedigree (b). Percent bengalness in known pedigree, random bengal cats, and other cat breeds.

ALC DOM

LxD BxL BxD

BEN a.

b.

Findings:

1.  Known ALC-DOM pedigree (Fig. 2a) provides information about theoretical % bengalness. In pedigree samples, theoretical and observed are similar (Fig. 2b, Table 1).

2.  Random Bengal cats on average have ~ 7% benglness whereas other cat breeds, combined, exhibit ~ 1% with some variations (Table 1).

3.  Proportion of ALC into Bengal cats is unmatched by any of the other breeds with the exception of Turkish Van (Fig. 2b).

4.  The relatively low % bengalness of random Bengal cats indicates multiple crosses to DOM or breeding between Bengal cats.

5.  Using 262 (autosomal) diagnostic panel is sufficient to identify Bengal cats and estimate ALC contribution.

Groups Theoretical Mean (Observed) s.d.(Observed) ALC 100 100 0 BxL 75 73 10.75 LxD 50 47 11.5 BxD 25 24.5 3.5 BEN unknown 7 2 DOM ~ 0 1.25 0.5

Table 1: Theoretical and observed percent bengalness in different cat groups.

Objective 3: Estimate the number of generations that gives a particular % bengalness

Dataset &

Analysis:

•  ALC-DOM pedigree (N = 98), random Bengal cat (N = 98). •  Use estimated % bengalness in each group. •  Determine the number of generations crossed to DOM that results into random Bengal cats. •  Use the theoretical relationship between % bengalness and number of generations crossed to

domestic depicted as: % Bengalness = 100 x (½) # generations

•  Infer theoretical number of generations since the initial hybridization for each Bengal cat .

Findings:

1.  First generation hybrids (LxD) exhibit small variation in % bengalness while backcross to ALC (BxL) and DOM (BxD) show significant variation between individuals (Fig. 3).

2.  The % bengalness of random Bengal cats places them in the range of 3-5 generations of theoretical hybridization with concentration around generation 4 (6.75%) (Fig.3).

3.  Variation in % bengalness can be explained by breeding between dissimilar (ALC-DOM) and similar (BEN-BEN) individuals.

4.  The small variation of % bengalness among Bengal cats is a sign of state equilibrium that results from breeding between Bengal cats and no or infrequent introduction of ALC or DOM alleles into the breed.

Fig. 3: Percent bengalness as function of generation number.

Poster number P0501

Conclusion

• SNP information update

• Ready to use comprehensive dataset

• Genome-wide analysis of cat breeds

• GWAS-Selection-Beyond

Questions?

Disclaimer

Figures, photos, and graphs in my lectures are collected using google searches. I do not claim to have personally produced all the material (except for some). I do cite only articles or books used. I thank all owners of the visual aid that I use and apologize for not citing each individual item. If anybody finds the inclusion of their material into my lectures a violation of their copy

rights, please contact me via email.

hhalhaddad@gmail.com