Anchoring of canine linkage group with chromosome ... - Springer Link

Mammalian Genome 10, 814–823 (1999).

Incorporating Mouse Genome

© Springer-Verlag New York Inc. 1999

Anchoring of canine linkage groups with chromosome-specific markers Petra Werner,1,* Cathryn S. Mellersh,2,* Michael G. Raducha,1 Susan DeRose,2 Gregory M. Acland,3 Ulana Prociuk,1 Neil Wiegand,2 Gustavo D. Aguirre,3 Paula S. Henthorn,1 Donald F. Patterson,1,* Elaine A. Ostrander2,* 1 Section of Medical Genetics and Center for Comparative Medical Genetics, University of Pennsylvania School of Veterinary Medicine, VHUP Room 4030, 3900 Delancey Street, Philadelphia, Pennsylvania 19104-6010, USA 2 Clinical Research Division, Fred Hutchinson Cancer Research Center, 1100 Fairview Ave. N., D4-100, Seattle, Washington 98109-1024, USA 3 Center for Canine Genetics and Reproduction, James A. Baker Institute for Animal Health, Cornell University, Ithaca, New York 14853, USA

Received: 23 February 1999 / Accepted: 28 April 1999

Abstract. A high-resolution genetic map with polymorphic markers spaced frequently throughout the genome is a key resource for identifying genes that control specific traits or diseases. The lack of rigorous selection against genetic disorders has resulted in many breeds of dog suffering from a very high frequency of genetic diseases, which tend to be breed-specific and usually inherited as autosomal recessive or apparently complex genetic traits. Many of these closely resemble human genetic disorders in their clinical and pathologic features and are likely to be caused by mutations in homologous genes. To identify loci important in canine disease genes, as well as traits associated with morphological and behavioral variation, we are developing a genetic map of the canine genome. Here we report on an updated version of the canine linkage map, which includes 341 mapped markers distributed over the X and 37 autosomal linkage groups. The average distance between markers on the map is 9.0 cM, and the linkage groups provide estimated coverage of over 95% of the genome. Fourteen linkage groups contain either gene-associated or anonymous markers localized to cosmids that have been assigned to specific canine chromosomes by FISH. These 14 linkage groups contain 150 microsatellite markers and allow us to assign 40% of the linkage groups to specific canine chromosomes. This new version of the map is of sufficient density and characterization to initiate mapping of traits of interest.

Introduction Purebred dogs offer a unique source of pedigrees to elucidate the molecular basis of simple and complex genetic diseases and traits. Selective breeding has produced more than 300 distinct breeds of domestic dogs worldwide, each defined by specific physical and behavioral characteristics. Each breed consists of a small, partially inbred genetic isolate, defined by a pedigree barrier (registration of an animal in a breed requires that the parents are registered members of the same breed). Owing to the lack of rigorous selection by breeders against genetic diseases, purebred dogs have a relatively high frequency of autosomal recessive and genetically complex disorders, many of which tend to be breed-specific. To date, over 350 canine diseases have been described (Patterson 1999). A large number of these have clinical and pathologic characteristics that are similar or identical to particular human genetic diseases and are probably due to mutations at homologous loci. The dog can thus serve as a useful * These authors and their associated laboratories contributed equally to this work. Correspondence to: E.A. Ostrander

model organism for unraveling the genetics of disease in humans. Indeed, the existence of so many partially inbred breeds, the large size of canine pedigrees, the relatively short generation time, and the possibility of planned matings in dogs suggest that some disease genes, particularly those underlying autosomal recessive and complex genetic disorders, may be more readily mappable in the dog than in most human populations. Previously, the rate-limiting step in the mapping of traits of interest in the canine genome has been the lack of a dense, informative genetic map. In the last two years, however, significant progress has been made. By utilizing growing collections of canine-specific microsatellite markers made available from a variety of laboratories (Dolf et al. 1997a, 1997b; Fischer et al. 1996; Francisco et al. 1996; Holmes et al. 1993, 1995; Ostrander et al. 1993, 1995; Shibuya et al. 1995), linkage groups defining the canine genome have been established. In 1997, Lingaas and collaborators described 16 linkage groups composed of 43 polymorphic markers, two of which, L16 and L13, were assigned to canine Chromosome (Chr) 18 (CFA18) and CFA20, respectively (Lingaas et al. 1997). Also in 1997, a first-generation framework map consisting of 150 markers, ordered into 30 linkage groups, was published (Mellersh et al. 1997). This map provided coverage of about 60% of the genome with markers spaced, on average, every 13 cM. Recently, the firstgeneration map has been expanded to a second-generation map of 276 markers ordered into 40 linkage groups (Neff et al. 1999). The primary limitations now facing the field are twofold. First, the map still lacks sufficient density of highly polymorphic markers to ensure that the entire genome can be successfully screened in pedigrees of purebred dogs. Second, the linkage groups of the first- and second-generation maps were composed almost exclusively of type II markers; thus, the map can not be anchored to the much denser human and mouse maps. This concern has been somewhat reduced by the recent development of a first-generation canine radiation hybrid (RH) map (Priat et al. 1998). The RH map is composed of 218 genes and 182 microsatellite markers, a subset of which had been previously placed on the linkage map, and was constructed with a panel of 126 canine rodent hybrid cell lines (Vignaux et al. 1999). There is partial conservation of synteny between the human and canine genomes, and the RH map enables portions of the canine genome to be aligned with the corresponding regions of human chromosomes. However, as with the linkage map, the majority of the RH groups, currently 57, remain unassigned to specific dog chromosomes, meaning that directed fine mapping and cytogenetic studies necessary for cloning genes of interest after initial mapping is still extremely difficult. Towards the development of a third-generation canine map, which resolves some of these issues, we have (i) expanded the number of microsatellite markers on the map to 341, thereby increasing the frequency of markers to about one every 9 cM, and (ii)

P. Werner et al.: Canine chromosome-specific markers

815

Table 1. Gene probes. Locus

Probe/PCR Condition/Primers

Reference

GLB1: Galactosidase, beta 1

dog-specific PCR product, 406 bp, Ta: 55°C TGACATCAACCCCAAACACCA TTCAAAGCTTCCATTCCAAACTG dog-specific PCR product, 380 bp; Ta: 55°C CGGCAACAGAGACGCTACA TTTGCACTGAGAGTTCCAG rat cDNA

This paper

NTF3: Neurotrophin 3

MYL2 (MLC-2v): Myosin light chain, regulatory ventricular PRKCD: Protein kinase C, delta RARG: Retinoic acid receptor gamma SLC3A1: solute carrier family 3, member 1 VIM: Vimentin

This paper

dog-specific PCR product

Henderson et al. 1988; O’Brien et al. 1993 ATCC 65978; Finkenzeller et al. 1990 TIGR/ATCC, 1158564 Lennon et al. 1996 P.S. Henthorn, in preparation

cHuVim1; human cDNA

ATCC, 59160; Perreau et al. 1988

human cDNA, hPKCalpha-GF29 mouse EST clone

Table 2. PCR primers for gene- or anonymous cosmid markers. Locus/ Cosmid

Ta (°C)

Repeat

Size (bp)

3

51

(GT)20

126

PRKCD

20

55

(CA)21

143

RARG SLC3A1

16 10

57 60

(CA)12 (AGCCCG)XNn (CTGGGG)y

231 437

VIM

2

58

(CA)21

307

COS6

6

55

(CA)19

161

COS8

8

63

(CA)14

246

COS15

15

61

(GGAT)30

207

COS18

18

60

(TC)19

360

MYL2

CFA

PCR Primers CAGACTGATAATGGATGGA CCTCTGCCTATGTCTCTGCCTCTC TGTTGCCTTCACTTGTAAT TTTGGAATGCTTGGAACTAA ACGGGCAGGGTGGGGCAGAG GGGGGCAGGGGGCAGTGTG GCCCGCCTGTCCACCGTTGTTATT GACACGGCGATGATGGCGATGGT TAGCCTAAGGGGAATACC CTTTAAGCTGATGAGTTTTG TCTGGAGTAGTGGGATTGAAGGAG AGGGCCACTGCTGAGATGTC CAACCATGGGATCAAGAGTCACAC TGCATTCGGCCCAGGACAT TGGTTTGATGGGTGAATAGTTAGA CATCCATCCATCCCTCCAGT CGTGGTGCCGGCCCTTTGAT TTTAGCGCCTGCCTTTGGAC

Table 3. Additional gene- or anonymous chromosome-mapped markers. Name

CFA

PCR Conditions/Primers/Product Size

Reference

AHTk20 ZuBeCa1 ZuBeCa2 ZuBeCa4 ZuBeCa6

20 10 1 3 5

as published 63°C 57°C 55°C 57°C, 395 bp TATTTGGGGCTGATTAGAACC GGATTTTACCAGCCTGCCTCGTC 184 bp CATAGCGACATCACATAAGT GGGCTGATAATTCAGTGAGA 59°C, 269 bp TTTCGTGATTCTCTTTGCGGTTGA GCTAGCGGGATCCCACTGTTCTGT

Fischer et al. 1996 Schelling et al. 1998b Schla¨pfer et al. 1998 Dolf et al. 1998 Unpublished, Genbank Accession Number AJ224121 Suter, 1998

59°C

Liu et al. 1998a

2A11

WT-1: Wilms tumor gene

SULT1A1/STP1: canine phenol sulfotransferase

6

Shibuya et al. 1996

These markers are mapped on the Keeshond-Beagle F1 backcross pedigree but not on the reference pedigrees, with the exception of 2A11, which is mapped on the reference pedigrees but not on the Keeshond-Beagle pedigree.

developed a set of genomic probes that allows us to assign 15 of the canine linkage groups, or 42% of the canine genome, to specific chromosomes. Materials and methods Isolation of genomic clones. Canine genomic cosmid or bacteriophage clones were identified for VIM, MYL2, RARG, PRKCD, and GLB1 by use

of previously described protocols (Werner et al. 1997). Probes used for screening of canine cosmid or bacteriophage genomic libraries are listed in Table 1. The screening with the probe for NTF3 did not reveal a positive clone. Isolated genomic clones were verified by sequencing and comparison with the relevant human, mouse, or rat sequence. In addition to the five specific clones described above, we also isolated four anonymous cosmids, termed COS6, COS8, COS15, and COS18. Each of these cosmids initially hybridized to a distinct human cDNA probe, but was found, upon sequencing, not to contain any sequences corresponding

816


Fig. 1. Assignment of linkage groups to canine chromosomes. Idiograms of canine chromosomes show the physical localization of gene and anonymous markers as determined by FISH. These markers were incorporated into existing linkage groups, which were thereby assigned to chromosomes as shown. *Assignment of LI to CFA1 is based on linkage to marker ZuBeCa2, the physical location of which was demonstrated on CFA1 by Schla¨pfer et al. (1998). The centromeric end of chromosomes is at the top of each idiogram. Orientation of linkage groups with respect to the cen-

tromere is definitive for CFA9, for which two or more FISH probes were used simultaneously, but tentative for other mapped chromosomes. Linkage distances between markers were based on lod scores of 3 or greater, except for the distance between ZuBeCa5 and C05.771 on CFA5, which was 33.2 cM, supported by a lod score of 2.5. HSA designations on the linkage maps indicate the chromosomal locations of the corresponding genes on human chromosomes. DGN8 on CFA10 was described by Mariat et al. (1998).

to the probe used for screening. Although there are no specific genes of interest in these cosmids, they could, nevertheless, be used in FISH analysis as anonymous cosmids to identify specific canine chromosomes and were, therefore, retained for polymorphism analysis as described below. The methods used for Southern analysis, library screening, and PCR analysis have been previously summarized (Werner et al. 1997).

Assignment of gene- and chromosome-associated markers to linkage groups. Initially, polymorphic, gene-associated, or chromosome-

Detection of polymorphisms. Initial Southern blot screening of canine genomic DNA from the founders of a Keeshond-Beagle F1-backcross pedigree digested with 27 different enzymes revealed informative restriction fragment length polymorphisms (RFLP) for GLB1 and NTF3. The dogspecific PCR products listed in Table 1 were used as hybridization probes. The RFLP for GLB1 was found after digestion with BglII, and the RFLP for NTF3 after digestion with HindIII. A PCR product for the canine SLC3A1 gene was isolated with primers designed from the canine cDNA (P. Henthorn, in preparation). The PCR product was found upon DNA sequencing to have a polymorphic hexamer compound repeat (Table 2). The marker was amplified with buffer M from the PCR Optimizer™ Kit (Invitrogen, San Diego, Calif.). For the isolated genomic cosmid or bacteriophage clones containing VIM, MYL2, RARG, PRKCD, and the anonymous cosmids COS6, COS8, COS15, and COS18, purified DNA was screened with 20-mer or 40-mer probes containing microsatellite probes according to previously described methods (Werner et al. 1997). Screening probes included (CA)10, (CAG)10, (GGAT)10, (GAAT)10, (GAAA)10, and (ATGA)10. Primers, annealing temperature, and PCR product size for each resulting marker are listed in Table 2. In addition to the above, several previously described gene- and chromosome-associated polymorphisms were utilized in this study. These were reported in Werner and coworkers (1997, 1998, 1999a) and are listed in Table 3.

mapped microsatellites and RFLPs were typed on a panel of 114 F1 backcross offspring from a Keeshond-Beagle cross, together with a set of microsatellite markers defining the canine framework map previously reported (Mellersh et al. 1997). This F1 backcross was originally made to define the mode transmission of inherited conotruncal heart defects in the Keeshond (Patterson et al. 1993) and has been used in the physical and linkage mapping of genes on CFA9 (Werner et al. 1997). Microsatellite typing in the Keeshond-Beagle backcross pedigree was part of an ongoing microsatellite genome scan for linkage to conotruncal defects in this pedigree. Following initial integration of gene and microsatellite markers in the Keeshond-Beagle backcross pedigree and physical mapping of the gene markers to canine chromosomes (Fig. 1), these markers were genotyped on standard reference pedigrees and integrated into the evolving canine linkage map (Fig. 2). Two-point lod scores used to assign markers to linkage groups were calculated using Map Manager (Manly 1993), and the most probable gene order was calculated by multipoint linkage analysis with Mapmaker Exp3.0 (Lander et al. 1987; Lincoln et al. 1992).

Physical mapping of markers to specific canine chromosomes by FISH. Canine genomic bacteriophage and cosmid clones containing the gene of interest and/or associated microsatellites were fluorescently labeled and hybridized to canine metaphase spreads by methods described previously (Werner et al. 1997). Chromosomal assignment of SLC3A1 utilized two PCR products (10 kb and 5 kb) that spanned several introns and exons from the canine gene isolated with primers designed from the canine SLC3A1 cDNA (P.S. Henthorn, in preparation). These were each confirmed by sequencing as having the SLC3A1 gene and then used as mixed


817

Fig. 2. Comprehensive canine linkage map. In total, 341 markers were assigned to linkage groups using a lod score of 5.0. Markers are ordered within linkage groups with a lod of 2.0 for the framework map (underlined on the figure) and 0.01 for the comprehensive map. Linkage groups that could not be assigned to chromosomes are ordered by relative size and named “L” followed by a number.

818

Fig. 2. Continued.



819

Fig. 2. Continued.

probes for FISH to canine chromosomes. Individual canine chromosomes were identified as in our previous studies (Werner et al. 1997, 1998).

Genotyping of gene- and chromosome-associated markers and addition of novel microsatellites. Most of the newly described gene and chromosome-associated markers, as well as 54 newly isolated microsatellite markers, were incorporated into the existing second-generation canine linkage map (Neff et al. 1999) by analysis on the reference pedigree panel described previously (Mellersh et al. 1997). The 54 new microsatellite markers were isolated from small insert plasmid genomic libraries after screening with (GAAA)10 and (CA)15 repeat probes as described previously (Francisco et al. 1996). The corresponding primer sequences, annealing temperature, and repeat types for each anonymous marker are summarized in Table 4. The reference panel is composed of 16 interrelated three-generation

pedigrees and contains 212 individuals, including 163 F2 offspring. Several distinct breeds of dog are represented in the panel including miniature and toy poodles, Norwegian elkhounds, Irish setters, and several genetically distinct lines of beagles. Markers were genotyped and double-scored as described previously (Mellersh et al. 1997). Each marker was checked for Mendelian segregation with the prepare option of MultiMap (Matise et al. 1994). Markers were assigned to linkage groups with the find-all-linkagegroups function of MultiMap; markers in each group were linked to at least one other marker with a recombination fraction ⱕ0.4 with a lod score of at least 5.0 (equivalent to odds of 100,000:1 in favor of linkage). Both a sex-averaged, framework map and a comprehensive map were constructed for each linkage group with MultiMap. For the framework map, markers were analyzed in decreasing order of informativeness; a marker was added to the map only when it could be localized to a unique interval with a lod score of ⱖ2.0. When each framework map was considered complete (no further markers fitting the criteria could be added to the map), the FLIPS

820


Table 4. Information on previously unmapped microsatellite markers.

Marker

Chromosome/ Linkage group

Approx PCR product size (bp)

Forward primer

Reverse primer

FH2598 FH2452 FH2593 FH2531 FH2541 FH2388 FH2594 FH2383 FH2525 FH2561 FH2576 C06.69 2A11 CPH3 FH2537 FH2422 FH2523 FH2407 FH2429 FH2528 FH2536 FH2566 FH2584 FH2548 FH2581 FH2396 FH2379 FH2534 FH2412 FH2457 FH2589 FH2574 FH2401 FH2547 FH2526 FH2394 FH2532 FH2587 FH2387 FH2540 FH2582 FH2535 FH2603 FH2441 FH2385 FH2585 FH2507 FH2380 FH2495 FH2600 Ren02C20 FH2516 FH2550 FH2377

CFA1 CFA1 CFA2 CFA3 CFA3 CFA3 CFA5 CFA5 CFA6 CFA6 CFA6 CFA6 CFA6 CFA6 CFA10 CFA10 CFA10 CFA16 CFA18 CFA20 CFA20 CFA26 X X L1 L1 L1 L2 L2 L2 L3 L4 L4 L5 L6 L8 L9 L9 L9 L10 L10 L12 L13 L13 L14 L15 L16 L17 L18 L21 L22 L23 Unlinked Unlinked

341 257 366 297 160 181 324 500 342 283 312 163 184 ND 169 238 485 473 173 345 375 159 308 175 358 474 590 318 160 360 294 407 220 328 320 585 322 198 405 410 285 145 366 129 520 306 191 174 134 339 315 190 448 355

CCTTGACTCAGCAGCCCTAC CGCTGTGAATATTGGTCCCT GTGAAGAGAGTGGGTGAGTCG GGAAGAGGAAGTCCTTGTATCC CGTATGAGTTGGTATAATCTCAGG CATGGAGTCTGCTTGAGATTCT TTTAAGGAGCTGCTCATGCA GACCTGTCTTCTCCTGAGTCTACC CTTCCAGCCTCTAGGGGC TGCTCAAGGTTGAATAAATATGC CCCCAAAATCTTATCATTTGC TGCAAATCCTGGTCGGTT CATAGCGACATCACATAAGT CAGGTTCAAATGATGTTTTCAG AAAAAGTGTAGAGCTTTCTTCAAA TTGCCCGTCCTATACTCCTG CAGAGTTTTAGGAGCTCTGTGC TGTCAGGCACAGTTTGAAGC GATCCACTTTGAATTGATTTTTG AGTTTTTCAGTTGACTTTGTGG TCCAATCTGCTATGCTGCC AGGGAGGAATTCCTCAGAGG GTTAGGTTCACAGTGGGCGT AAGGGAGGAAACAATGCTGA ACACTGGGTTGAGTGGTGG CCAGTGATGTATTGTTTAGAGGA CAGCCTTGGTTGCTCAGC AAGTCCCGGGATCGAGTC GCTGGGGATTTATTCTGACC AACCTGGTACTTTGAATTTGCA AGGCAAAGAATCTACACAATGA TTACAAATGGCTGAAAGAATGG CTGATTCTGCCCATTGGG TCACTATGCAGCTGATTTAGTTG TCCTTGCTGATCCCAGATG ATTCATGGAATCTCCTAATGAATG CACGCAGAAAGGCAGAAAG GGCATGAACAAATCAGTGGA TTGTTCACTCAGCTAGGAGACG GTGGTTGATTTCTAGTTTCATAGC TGGAGTGTGTTCCAAGGTCA GTCATTGACAGACTACAAATCTCC CCCAAACCATCTGATTCGTT TAGTTGTGTGCATGATCTCG AGTTCCCTTTATTCTTCCAAAGTG TCGATGTCTGCCTTTCTTGA GGCCCTTGAACATCATGG TTCTGGAATTCCCTATTCATGG ATTTCATATGTGAGGCTGAGATTG AGACCCTGGATCTCTGGGTT AGAAATTGCATCACTCACAT AATGGATGGAACTTAGGGCA CCCATCGTTCAGGCAAAG TCCCTTGGGGAAGTAGAGTG

CGACCTATCCATACATGCCC GATAGCCACTTCATGAAAGATCC GGAGTGGTCGTGAATATGCA CACCTGGCTGGCTCAGTC TGCTTTTCACCTCCCTCTTG CAGATTTTTCACACATTAGGGC CTGAAATTCCTGGCCCAGTA TACCAGAAATTACCTGCCCG TGCATCAGGCTCCACTCTC TTTATGGCCTGTGGGCTC CCTGCACTGGGAAAATTAGC CACAATTCTGGACCATGCG GGGCTGATAATTCAGTGAGA TTGACTGAAGGAGATGTGGTAA ATTGAGACCCAAGACTGTTAGTG CCACATGATTTCACTTGTATATGG CTAGCAGCCGGGAGTTTATG TGCCTTCAGCTCAGAACATG TTCAGCAAATGGTTCTGGAA AGTCCCAGGATCAAGCCC AGCAGGTCTCTTGGCTACCC TCTGAGCCTTAAGGCAGAAGC ACTCAAAGACCTGGAGGGGT GACATTCAGAGATTTCCGCC CCAGGGGCTAATATCCCTCT CTCAAACTCATACCTTGAGATCA AATGCCTCTGGTCCTGACTG ATCCATGTCTTTGCAAATTGC AAATTAACCAAATGTTTGCAACA TTTGGTCAACATACCTCTCTGG TGCCTGGGTAGGTCAGTTG TTAACACTATTTAGCCTTGGATGA ATGTAAGCTCTACTGGGGTACTGG AGTTTGGAGTTTGAACTCTACAGA TGGAAGCCTTAGTCTCAGCA AGGAGACTGTGAAAGAGAGAAAGG TTTCCATAGTGGCTGCATCA TTTGCTGTTTAATCCATCTGG TTTTTATTCAACAGCAGCTAGGC ACATTGCCTCAAAGAGCCTG GTTGTTCCCACAAAAGGCAG ACAGACTTGCAGTATTTTGTCTG ACCATGGATTGGCCAGTAAG TGGAGAAAGTTCCATGTGCA GATGAATCACTAATTCTATTCCCG GCCCCCTCACCTCATATTCT GATTCCAAGCTGGGTATGGA CATAGAAAACAAAATCACACAGGC CAGTGGGAGAAAGATGCCAT TTGTTAAATTTTGTGCCAGTACC GCTGCTCCGAAAACTAACTT CTGCATCTGGTAACCATCGA TGAACATGGAAATCATTCCG TAGCTAATGTGGTTAACGGTTACC

All of the above markers are previously unpublished, with the exception of CPH3 (Fredholm et al. 1995), 2A11 (Suter 1998), C06.69 (Ostrander et al. 1993), and Ren02C20 (Priat et al. 1998). ND, Not Determined.

function of CRI-MAP was used to compare the likelihoods of all orders of each block of three adjacent markers to ensure that the odds in favor of the order presented were at least 100:1 over the alternative orders checked. For the comprehensive map, markers were similarly assigned to linkage groups with a lod score of 5.0, and then the most likely order was determined with a lod score of 0.01. Linkage groups that could not yet be assigned to specific chromosomes (and thus assigned a CFA designation) were instead assigned an “L” number (1–23), based on the comparative size of the linkage group.

Results We have utilized a newly isolated and characterized series of cosmids and bacteriophage genomic clones, many of which contain known genes of interest, to identify canine chromosomes by FISH.

Because polymorphic markers are associated with each of these loci, they could also be placed on the canine linkage map. In addition, we have added 54 newly isolated microsatellite markers to the map to expand coverage of the map by 15%. Assignment of gene- and chromosome-associated markers to linkage groups. Fourteen of 38 previously identified linkage groups were assigned to specific canine chromosomes with 22 markers we developed for this purpose. These consisted of 11 newly isolated markers and 11 markers that we had previously mapped to canine chromosomes, but which had not been integrated with the evolving microsatellite map (Werner et al. 1997, 1998, 1999b). In addition, the marker ZuBeCa2 mapped to CFA1 by Schläpfer et al. (1998) was integrated with the microsatellite map, allowing a linkage group to be placed on CFA1. We have also incorporated data from


seven polymorphic markers previously reported by other laboratories (Table 3). Five of these markers had been FISH mapped and support our findings regarding the chromosomal localization of M-L3, M-L19, M-L14, M-L24, and M-L22 to CFA3, CFA5, CFA6, CFA10, and CFA20, respectively (Fig. 1 and 2). The nomenclature “M-L” refers to the linkage group assignments of Mellersh et al. (1997) and “N-L” to the assignments of Neff et al. (1999). To initially assign gene-linked or chromosome-mapped markers to their appropriate linkage group, we utilized data made available from an ongoing genome-wide screen of 114 KeeshondBeagle F1-backcross female meioses. As part of the genome screen, 108 informative microsatellites from the original framework map reported by Mellersh et al. (1997) had already been typed on the pedigree, thus allowing us to easily integrate these data with those generated by previous genotyping of microsatellite markers on the standard canine reference panel (Mellersh et al. 1997; Neff et al. 1999). Two-point analysis (lod score >3.0) revealed linkage between markers of 15 linkage groups and gene-associated or chromosomemapped markers assigned to 14 different chromosomes (Fig. 1). NTF3, linked to RARG, was linked to markers in two linkage groups, M-L20 and M-L26, which were integrated and assigned to CFA16. The most probable order of genes and the subset of microsatellites analyzed on the Keeshond-Beagle F1-backcross pedigree for the 14 chromosomally assigned linkage groups is shown in Fig. 1, along with the physical location of the FISH-mapped markers. Orientation of linkage groups with respect to the centromere is definitive for CFA9, for which two or more FISH probes were imaged simultaneously. Orientation is tentative for other mapped chromosomes. Chromosomal assignment of markers by FISH. Following integration of the gene and gene-associated markers into linkage groups, the assignment of a linkage group to a chromosome was undertaken by FISH analysis. These experiments utilized nine newly isolated canine genomic cosmid or bacteriophage clones, the dogspecific SLC3A1 PCR products, as well as previously mapped clones (Werner et al. 1997, 1998, 1999b). Specifically, six gene loci were newly assigned to chromosomes as follows: VIM-CFA2, MYL2-CFA3, SLC3A1-CFA10, RARG-CFA16, PRKCD-CFA20, and GLB1-CFA22. Four anonymous genomic cosmids that served as chromosome-specific markers were placed as follows: COS6CFA6, COS8-CFA8, COS15-CFA15, and COS18-CFA18. Our designations of Chrs 1 to 21 are consistent with the nomenclature recommended by the International Committee for the Standardization of the Canine Karyotype (Switonski et al. 1996). An international standard for identification of G-banded canine autosomes 22–38 has not yet been finalized, because these chromosomes are small, acrocentric, and difficult to identify by G-banding alone. We identified autosomes 22 and 26, using the G-banding pattern reported by Selden et al. (1975). Idiograms of the mapped chromosomes with the approximate physical localization of the loci by FISH are shown in Fig. 1. Integration of chromosome-based markers and additional anonymous markers to the map. Eleven of the 23 markers defining specific canine chromosomes were integrated into the canine linkage map by genotyping on the standardized reference panel of 16 three-generation pedigrees described above. It was not necessary to analyze all 23 markers on the reference panel, since some linkage groups contained multiple chromosome-specific markers, and only one needed to be analyzed to assign the entire linkage group to a chromosome. In addition to the above 11 markers, 54 additional new anonymous microsatellite markers, most containing tetranucleotide repeats, were also genotyped on the reference panel and placed on the map. Primers, product sizes, and PCR conditions, together with current versions of the map, are available

821

through the Fred Hutchinson Cancer Research Center web site (http://www.fhcrc.org/science/dog_genome/dog.html) and are listed in Table 4. The 341 markers that define this map are distributed among 37 autosomal linkage groups, plus one linkage group on the X Chr, and are assigned to each group with a lod score of ⱖ5.0 (Fig. 2). Only ten markers are informative in the reference families, but are unlinked to any other marker. Of the 341 markers, 224 can be ordered with a lod of 2.0 and thus comprise the framework map. For the comprehensive map (Fig. 2), markers are ordered within each linkage group with a lod score 0.01. Markers that comprise the framework map are underlined in Fig. 1. The comprehensive map spans 1822.40 cM, and the framework map spans 1658.7 cM. A total of 45.4% of the markers on the map are based upon tetranucleotide repeats. These were informative for an average of 64.9% of meioses in the reference families, compared with dinucleotide-based markers that were informative for about 50% of meioses. The average interval size on the lod 2.0 framework map is 9.0 cM. The average interval on the comprehensive (lod 0.01) map is 7.03 cM. There is an average of 7.6 markers/group on the comprehensive map, with a range of 2–20 markers. Seven microsatellites are located on the X Chr; heterozygous males were observed for all other markers, indicating that these are located on autosomes. All but two linkage groups contain three or more markers. Discussion Identification of disease-causing genes in the dog can be expedited by knowledge of the syntenic relationships among other mammalian genomes for which extensive genome mapping and sequencing information is available, such as the human or mouse. In these studies, we have begun to integrate the increasingly dense linkage map of canine microsatellites with a series of chromosomespecific markers that are informative for both linkage mapping and for physical mapping of the integrated linkage groups to canine chromosomes. As in our previous studies, most of the chromosome-specific markers used are gene-associated polymorphisms detected in canine genomic or phage clones (Werner et al. 1999a). Given the high degree of conservation of gene synteny between mammalian species, the gene-associated polymorphisms provide a means to align the canine linkage groups with orthologous regions of the genome in more completely mapped species. This is important in extending the canine gene map and in gaining access to the extensive local maps of expressed sequences in these species when selecting positional candidates after mapping of a canine disease to a region. The present study assigns 15 of the 37 existing autosomal microsatellite linkage groups to 14 canine chromosomes. In addition, we have added 54 new microsatellite markers to the map, thus expanding the coverage of the canine genome by 10% and the average number of markers/linkage group by 15% over that published previously (Neff et al. 1999). Because the majority of these markers are based upon tetranucleotide repeats, they should contribute significantly to the informativeness of the map (Francisco et al. 1996). We have reduced the average interval size between markers to 9.0 cM if we consider markers ordered on their respective chromosomes with a lod of 2.0. The average interval size for markers ordered with a lod of 0.01 is 7.03 cM. Linkage groups that can not yet be assigned to chromosomes are numbered L1, L2, etc., in descending order of size. Chromosomal assignments of gene loci to CFA5, CFA9, and CFA26 have been published earlier (Werner et al. 1997, 1998, 1999b) and are based on the localization of between two and eight different gene loci to the same chromosome (Fig. 1). The assignment of linkage groups M-L19, M-L18, and M-L23 to these chromosomes based on lod scores ⱖ3.0 can, therefore, be made with

822

high confidence. Further gene loci and their chromosomal assignments and linked microsatellite groups include VIM, assigned to CFA2 (M-L21), MYL2 (CFA3/M-L3), SLC3A1 (CFA10/M-L24), RARG (CFA16/M-L20/26), PRKCD (CFA20/M-L22), and GLB1 (CFA22/M-L2). Four polymorphic, chromosome-specific (COS) markers developed from annonymous cosmid clones and their chromosome assignments and microsatellite linkage groups are as follows: COS6 (CFA6/M-L14), COS8 (CFA8/M-L5), COS15 (CFA15/M-L34), and COS18 (CFA18/M-L16). The chromosomal assignment of M-L22 to CFA20 by linkage to PRKCD is supported by the linkage of this group to AHTk20 previously assigned to CFA20 by Fischer et al. (1996). CFA20 also likely incorporates L13 (Lingaas et al. 1997). Our assignment of M-L1 to CFA1 was based on incorporating the marker, ZuBeCa2, which had been previously assigned to CFA1 by Schla¨pfer et al. (1998) into our analysis. While the results reported here are in good agreement with those previously reported, we do note some differences. Our assignment of M-L5 to CFA8 by linkage to a polymorphic marker derived from the cosmid clone COS8 disagrees with the findings of Dutra et al. (1996). Dutra et al. FISH mapped the immunoglobulin heavy-chain locus (IGH) to CFA4, using a bacteriophage clone that was isolated by use of a human probe from the switch region of the IGH locus. Langston et al. (1997), using a PCR product from the IGH locus, observed several markers that were syntenic with IGH on a panel of canine rodent hybrid cell lines and termed these syntenic group D. Since IGH had been reported to be on CFA4, syntenic group D and the associated markers were similarly assigned. Our linkage data, however, suggest that the microsatellite markers in syntenic group D are linked to COS8. This would place the IGH locus also on CFA8. Discrepancies such as these can best be resolved by the placement of additional markers on the relevant chromosomes. Indeed, it would be optimal if each linkage group could be assigned to its relevant chromosome with several rather than a single marker. As the canine map evolves and multiple markers are placed on each chromosome, additional changes can likely be expected. The International Committee on Standardization of the Canine Karyotype (Switonski et al. 1996) recommended that autosomal chromosome pairs 22–38 be identified by a combination of Gbanding and molecular cytogenetic methods. In the studies reported here, we used the G-banding pattern of Selden et al. (1975) to identify the two smallest acrocentric autosomal pairs mapped as CFA22 and CFA26. The gene-associated markers that are FISH mapped to these chromosomes provide a more definitive means for identifying them (Fig. 1). In the map presented here, typing of new markers allowed for some smaller groups in the second and first generations of the linkage maps to be condensed. For instance, in the second generation map (Neff et al. 1999) CXX.733, FH2130, and AHTk200 formed a small linkage group 16 cM in length (linkage group N-L27). AHTk211 and CXX.420 were linked to each other in a separate linkage group (linkage group N-L33), 3 cM apart. In this analysis, we typed two additional markers, N41 and FH2566, which were linked to markers from both N-L27 and N-L33. This allowed us to conclude that these two small linkage groups are present on the same chromosome, CFA26. Similarly, in the second-generation map, AHT123 and FH2233 formed linkage group N-L29 (5 cM in length), while FH2161, CXX.889, and FH2312 formed linkage group N-L31 (6 cM). In this analysis, typing of two additional markers, FH2603 and FH2441, which were linked to markers in both of these smaller groups, allowed us to conclude that all seven markers reside on a single chromosome, forming a single linkage group (L13), which is 39.8 cM in length. Finally, markers AHT133 and CXX.172 were the sole markers in linkage group N-L28 in the second-generation map, and marker AHT004 was unlinked to any other marker. However, three additional markers typed in this study, FH2542, FH2587, and FH2387, are


linked to all three of these markers, enabling the construction of a single linkage group, L9, in this third-generation map, which is 45.2 cM in length. The existing map appears to cover a significant portion of the canine genome. It includes 37 autosomal linkage groups and the X Chr. There are currently ten informative, but unlinked, markers. These likely represent the ends of existing linkage groups for which intermediate markers have not yet been defined, or one or more of the smaller chromosomes. The projected size of the canine genome is 26.5 M ± 1.1 M (95% C.I. 24.3 M to 28.7 M) (Neff et al. 1999). If one assumes that linkage can be detected at least 10 cM on either side of each linkage group, and 10 cM on either side of each unlinked marker, it can be projected that the current map can detect linkage over 2676 cM. In addition to the work summarized here, additional markers have recently become available that can be integrated into the map that will allow further resolution. For instance, cosmid-derived polymorphic markers assigned to canine chromosomes have recently been reported for Chrs 13, 19, 21, as well as one of the smaller chromosomes 22–28 (Dolf et al. 1997a, 1997b; Fischer et al. 1996; Schelling et al. 1998a). In addition, several recently described, gene-associated polymorphic markers, often within the introns of genes, can now be incorporated (Ganjam et al. 1998; Liu et al. 1998b, 1998c; Shibuya et al. 1995, 1998; Zeiss et al. 1998; Zhou et al. 1998). While the work described here represents an important step forward in the development of a canine map suitable for disease gene mapping, significant work remains to be done. The most interesting pedigrees for disease gene mapping may be those with the highest levels of homozygosity. The current map may still lack a sufficient density of markers for mapping traits of interest in families with significant inbreeding, and the density of the map will need to at least triple to allow fine mapping of genes and positional cloning. In addition, the canine RH map and meiotic linkage map are not yet fully integrated, so only a small number of gene loci are currently mapped to specific chromosomes. However, further expansion of the canine RH map and integration of the linkage and RH maps will quickly resolve the syntenic relationship between the canine and human genomes, enabling primary linkages established with this map to be fully exploited. Acknowledgments. We thank James Kehler, Nicola Suter, Mark Gibbs, Melissa Fleming, and Kunal Ray together with other members of our laboratories for their interest and support. This work was funded by grants from the Canine Health Foundation of the American Kennel Club to E.A.Ostrander and D.F.Patterson; an American Cancer Society Research Grant to E.A.Ostrander; a Wellcome Prize Travelling Research Fellowship and Ralston Purina Postdoctoral Fellowship to C.S.Mellersh, NIH grants HL18848 from the National Heart, Lung and Blood Institute and RR02512 from the Division of Research Resources, as well as a grant from Mrs. Cheever Porter Foundation, to D.F.Patterson; P.Werner was supported by a fellowship from the Robert J. and Helen C. Kleberg Foundation. Development of reference pedigrees was supported by NIH grant EY-06855, support from the Foundation Fighting Blindness, Morris Animal Foundation, The Seeing Eye, Inc., and Baker Institute PRA Research Fund to G.D. Aguirre and G.M. Acland.

References Dolf G, Schla¨pfer J, Parfitt C, Schelling C, Zajac M et al. (1997a) Assignment of the canine microsatellite CanBern1 to canine chromosome 13q21. Anim Genet 28, 156–157 Dolf G, Schla¨pfer J, Schelling C, Jorg H, Ladon D et al. (1997b) Assignment of the canine microsatellite CanBern6 to the canine chromosome group 22 to 38. Anim Genet 28, 453–454 Dolf G, Schla¨pfer J, Switonski M, Stranzinger G, Gaillard C et al. (1998) The highly polymorphic canine microsatellite ZuBeCa4 is localized on canine chromosome 3q15-q18. Anim Genet 29, 403–404 Dutra AS, Mignot E, Puck JM (1996) Gene localization and syntenic mapping by FISH in the dog. Cytogenet Cell Genet 74, 113–117

P. Werner et al.: Canine chromosome-specific markers Finkenzeller G, Marme D, Hug H (1990) Sequence of human protein kinase C alpha. Nucleic Acids Res 18, 2183 Fischer PE, Holmes NG, Dickens HF, Thomas R, Binns MM et al. (1996) The application of FISH techniques for physical mapping in the dog (Canis familiaris). Mamm Genome 7, 37–41 Francisco LV, Langston AA, Mellersh CS, Neal CL, Ostrander EA (1996) A class of highly polymorphic tetranucleotide repeats for canine genetic mapping. Mamm Genome 7, 359–362 Fredholm M, Winterø AK (1995) Variation of short tandem repeats within and between species belonging to the Canidae family. Mamm Gemone 6, 11–18 Ganjam KK, Shibuya H, Stoy SJ, Liu P-C, Ganjam VK et al. (1998) A BstUI PCR/RFLP marker in an intron of the canine fibrillin I gene. Anim Genet 29, 240–241 Henderson SA, Xu YC, Chien KR (1988) Nucleotide sequence of full length cDNAs encoding rat cardiac myosin light chain-2. Nucleic Acids Res 16, 4722 Holmes NG, Mellersh CS, Humphreys SJ, Binns MM, Holliman A et al. (1993) Isolation and characterization of microsatellites from the canine genome. Anim Genet 24, 289–292 Holmes NG, Dickens HF, Parker HL, Binns MM, Mellersh CS et al. (1995) Eighteen canine microsatellites. Anim Genet 26, 132–133 Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ et al. (1987) MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of experimental and natural populations. Genomics 1, 174–181 Langston AA, Mellersh CS, Neal CL, Ray K, Acland GM et al. (1997) Construction of a panel of canine-rodent hybrid cell lines for use in partitioning of the canine genome. Genomics 46, 317–325 Lennon G, Auffray C, Polymeropoulos M, Soares MB (1996) The I.M.A.G.E. Consortium: an integrated molecular analysis of genomes and their expression. Genomics 33, 151–152 Lincoln SI, Daly M, Lander E (1992) Constructing Genetic Maps with MAPMAKER/EXP3.0, 3rd ed. (Cambridge, Mass. Whitehead Institute) Lingaas F, Sorensen A, Juneja RK, Johansson S, Fredholm M et al. (1997) Towards construction of a canine linkage map: establishment of 16 linkage groups. Mamm Genome 8, 218–221 Liu P.-C., Shibuya H, Nonneman D, Katz ML, Johnson GS (1998a) A BseRI PCR/RFLP in an intron of the canine phenol sulfotransferase gene. Anim Genet 29, 329 Liu PC, Shibuya H, Katz ML, Johnson GS (1998b) A BsmFI PCR/RFLP in the canine CD19 gene. Anim Genet 29, 64–65 Liu PC, Shibuya H, Lubahn DB, Johnson GS (1998c) Polymorphic (CT)(n) microsatellite in an intron of the canine SA-hypertension gene. Anim Genet 29, 72 Manly KF (1993) A Macintosh program for storage and analysis of experimental genetic mapping data from recombinant inbred strains. Mamm Genome 4, 303–313 Mariat D, Amigues Y, Boscher MY (1998) Eight canine tetranucleotide repeats. Anim Genet 29, 156–157 Matise TC, Perlin M, Chakravarti A (1994) Automated construction of genetic linkage maps using an expert system (Multimap): a human genome linkage map. Nat Genet 6, 384–390 Mellersh CS, Langston AA, Acland AF, Fleming MA, Ray K et al. (1997) A linkage map of the canine genome. Genomics 46, 326–336 Neff MW, Broman KW, Mellersh CS, Ray K, Acland GM et al. (1999) A second generation map of the domestic dog, Canis familiaris. Genetics 151, 803–820 O’Brien TX, Lee KJ, Chien KR (1993) Positional specification of ventricular myosin light chain 2 expression in the primitive murine heart tube. Proc Natl Acad Sci USA 90, 5157–5161 Ostrander EA, Sprague GF, Rine J (1993) Identification and characterization of dinucleotide repeat (CA)n markers for genetic mapping in dog. Genomics 16, 207–213

823 Ostrander EA, Mapa FA, Yee M, Rine J (1995) One hundred and one new simple sequence repeat-based markers for the canine genome. Mamm Genome 6, 192–195 Patterson DF (1999) Canine Genetic Disease Information System- A Computerized Knowledgebase of Genetic Diseases of the Dog. (St. Louis, Mo.) Mosby-Yearbook in press Patterson DF, Pexieder T, Schnarr WR, Navratil T, Alaili R (1993) A single major-gene defect underlying cardiac conotruncal malformations interferes with myocardial growth during embryonic development: studies in the CTD line of Keeshond dogs. Am J Hum Genet 52, 388–397 Perreau J, Lilienbaum A, Vasseur M, Paulin D (1988) Nucleotide sequence of the human vimentin gene and regulation of its transcription in tissues and cultured cells. Gene 62, 7–16 Priat C, Hitte C, Vignaux F, Renier C, Jiang Z et al (1998) A wholegenome radiation hybrid map of the dog genome. Genomics 54, 361– 378 Schelling C, Dolf G, Schla¨pfer J, Switonski M, Ladon D (1998a) Assignment of the canine microsatellite ZuBeCa5 to canine chromosome 19q7. Anim Genet 29, 409–410 Schelling C, Stranzinger G, Dolf G, Schla¨pfer J, Switonski M (1998b) Assignment of the canine microsatellite ZuBeCa1 to canine chromosome 10q22-q24. Anim Genet 29, 398 Schla¨pfer J, Switonski M, Schelling C, Stranzinger G, Gaillard C et al. (1998) The highly polymorphic canine microsatellite ZuBeCa2 is localized on canine chromosome 1q210-q211. Anim Genet 29, 329 Selden JR, Moorehead PS, Oehlert ML, Patterson DF (1975) The Giemsa banding pattern of the canine karyotype. Cytogenet Cell Genet 15, 380– 387 Shibuya H, Mrad DR, Collins BK, Stoy SJ, Nonneman D et al. (1995) Two polymorphism in the canine beta-A3/A1-crystallin gene, detectable by PCR-RFLP analysis. Anim Genet 26, 277–285 Shibuya H, Collins BK, Collier LL, Huang TH, Nonneman D et al. (1996) A polymorphic (GAAA)n microsatellite in a canine Wilms tumor 1 (WT1) gene intron. Anim Genet 27, 59–60 Shibuya H, Cassells MW, Huang TH-M, Johnson GS (1998) A polymorphic (TG)n microsatellite in an intron of the canine angiotensin I converting enzyme. Anim Genet 29, 66–67 Suter N (1998) Isolation and characterisation of canine microsatellites. Ph.D. Thesis, University of Leicester, Leicester, UK. Switonski M, Reimann N, Bosma AA, Long S, Bartnitzke S et al. (1996) Report on the progress of standardization of the G-banded canine (Canis familiaris) karyotype. Chromosome Res 4, 306–309 Vignaux F, Priat C, Jouquand S, Hitte C, Jiang Z et al. (1999) Towards a dog radiation hybrid map. J Hered 90, 62–67 Werner P, Raducha MG, Prociuk U, Henthorn PS, Patterson DF (1997) Physical and linkage mapping of human chromosome 17 loci to dog chromosome 9 and 5. Genomics 42, 74–82 Werner P, Raducha MG, Prociuk U, Lyons LA, Kehler JS et al. (1998) RXRA and HSPA5 map to the telomeric end of dog chromosome 9. Anim Genet 29, 220–223 Werner P, Raducha MG, Prociuk U, Henthorn PS, Patterson DF (1999a) A comparative approach to physical and linkage mapping of genes on canine chromosomes using gene-associated simple sequence repeat polymorphisms: illustrated by studies of dog chromosome. J Hered 90, 39–42 Werner P, Raducha MG, Prociuk U, Budarf M, Henthorn PS et al. (1999b) Comparative mapping of the DiGeorge region in the dog and exclusion of linkage to inherited canine conotruncal defects. J Hered in press Zeiss CJ, Trepanier LA, Aguirre GD, Ray K (1998) A highly conserved microsatellite in the dystrophin gene of diverse mammalian species. Anim Genet 29, 224–227 Zhou T, Nonneman D, Shibuya H, Khan S, Liu P-C et al. (1998) An AvaII PCR/RFLP in an exon of the canine gene for the beta subunit of the amiloride-sensitive sodium channel (SCNN1B). Anim Genet 29, 239