Fine Mapping of Canine XLPRA Establishes Homology of the Human and Canine RP3 Intervals

Abstract

purpose. Canine X-linked progressive retinal atrophy (XLPRA) is a hereditary, progressive retinal degeneration that has been mapped previously to the canine X chromosome in a region flanked by the dystrophin (DMD) and tissue inhibitor of metalloproteinase 1 (TIMP1) genes, and is tightly linked to the gene RPGR. The comparable region of the human X chromosome includes the disease locus for RP3, an X-linked form of retinitis pigmentosa, although the current canine disease interval is much larger.

methods. To refine the map of the canine XLPRA disease interval, 11 X-linked markers were mapped, both meiotically, in two extensive canine pedigrees informative for XLPRA, and on a 3000-rad canine-hamster radiation hybrid (RH) panel. A 12th marker was mapped on the RH panel alone.

results. The integrated map of this region of CFAX now covers approximately 47.3 centimorgans (cM) and 194 centirays (cR)₃₀₀₀, and demonstrates strong conservation of synteny between humans and dogs. Genes defining the human RP3 zero-recombination interval (human homologue of mouse t complex[ TCTE1L], sushi repeat-containing protein, X chromosome [SRPX], and retinitis pigmentosa guanosine triphosphatase [GTPase] regulator [RPGR]) are tightly linked to each other, to the XLPRA locus, and to the gene ornithine transcarbamylase (OTC) in dogs.

conclusions. Strong conservation of gene order was demonstrated in the short arm of the X chromosome between dogs and humans as was homology of the canine XLPRA and human RP3 intervals. These results create a valuable tool for investigating canine XLPRA and other X-linked eye diseases in dogs.

X-linked progressive retinal atrophy (XLPRA) in the Siberian husky ¹ is a naturally occurring disease model for human X-linked retinitis pigmentosa (XLRP). XLRP accounts for approximately 10% to 25% of all classes of retinitis pigmentosa (RP) and is one of the most severe forms of the disease in clinical onset and progression. ² One form of XLRP, termed RP3, has been localized by linkage analysis to an interval of approximately 500 kb on HSAXp21.1, between markers DXS1110 and DXS6679, which are telomeric to the genes human homolog of mouse t complex (TCTE1L) and ornithine transcarbamylase (OTC), respectively. ³ Mutations in the retinitis pigmentosa guanosine triphosphatase (GTPase) regulator (RPGR) gene, which maps within the RP3 interval, have been causally associated with the disease in many, but not all, human families in whom disease maps to this interval. ^{3 4 5}  

The XLPRA disease locus maps to the canine X chromosome (CFAX) in a region flanked by the genes dystrophin (DMD) and tissue inhibitor of metalloproteinase 1 (TIMP1) and is tightly linked to the RPGR gene. ⁶ Although the RPGR gene is located within the human RP3 interval, ³ the distance between the disease locus and the nearest flanking markers is great. 

To refine the locus homology between XLPRA and RP3 and prepare for positional cloning of the canine disease gene, a higher resolution map of this region of CFAX is needed. Although significant progress has been made in the development of a canine map in recent years, ^{7 8 9 10} the current integrated meiotic and radiation hybrid (RH) map includes few ordered markers on CFAX. Therefore, we developed both RH and meiotic maps of CFAX using five anonymous microsatellites, three gene-associated microsatellites, and three intragenic sequence variants. An additional intragenic marker that was not informative in the study pedigrees was also placed on the RH map alone. The genes tested are all conserved between human and dog, and the interval on CFAX from TCTE1L to OTC defines a zero-recombination region that is homologous to the RP3 interval in humans. ³  

Methods

Animals, Pedigrees, and Samples

Two different outcrossed, multigenerational canine pedigrees informative for XLPRA were used to develop a meiotic map of CFAX and to refine the XLPRA interval. The dogs were part of a research colony maintained under sponsorship of the National Eye Institute, Bethesda, MD. Methods for phenotypic ascertainment of disease status have been described. ^{1 6 11} Genomic DNA was isolated from citrated blood samples, either by standard phenol-chloroform extraction or by a rapid protocol, as described previously. ¹² All procedures were in full compliance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Canine Bacterial Artificial Chromosome Library and Microsatellite Markers

The RPCI81 canine bacterial artificial chromosome (BAC) library was purchased from the Roswell Park Cancer Institute (Buffalo, NY; available at http://www.chori.org/bacpac/mcanine81.htm). Details have been published previously, ¹³ and a description of the library can be found on the Bacpac Web site. 

Two microsatellite markers, CUX20001 and CUX40002, were obtained from canine BAC clones (132E5 and 119B21) positive for orthologous genes in the human RP3 interval identified as TCTE1L and OTC, respectively (Zangerl et al., manuscript submitted). Briefly, PCR primers were designed from human and other mammalian sequences for TCTE1L, sushi repeat-containing protein, X chromosome (SRPX), RPGR, and OTC, a cluster of genes covering approximately 500 to 600 kb on human Xp21.1. Fragments amplified from canine retinal cDNA or genomic DNA using these primers were then used as probes to identify positive BAC library clones. The BAC clone DNA was purified with a large construction kit (Qiagen, Valencia, CA), digested with HaeIII, precipitated, and cloned into the M13mp19 phage vector. Positive clones were isolated by hybridizing with pooled oligo probes (di- and tetranucleotide repeat) and verified by direct sequencing. 

The tetranucleotide repeat marker of CUX40001 was obtained from a BAC clone (255O20) positive for the SRPX gene by end sequencing (Table 1) . Microsatellite markers FH2548 and FH2985 have been published and localized to CFAX. ¹⁰ FH2916, FH2997, and FH3027 were isolated from canine genomic DNA, and details are posted in the Fred Hutchinson Cancer Research Center (Seattle, WA) dog genome Web page (available at http://www.fhcrc.org/science/dog_genome/markers/all600.html#CFAX). The microsatellite nucleotide sequence data reported in this article have been submitted to EMBL (European Molecular Biology Laboratory, Heidelberg, Germany; available at http://www.embl-heidelberg.de) and have been assigned accession numbers AJ279083, AJ279085, and AJ298320. Intragenic polymorphic markers derived from the canine genes RPGR, androgen receptor (AR), TIMP1, and δ-aminolevulinate synthase 2 (ALAS2) have been described. ^{6 14 15 16} Table 1 provides further information on the primers used to amplify these 12 markers. 

RH Map Construction

The canine-hamster RH panel (T27) purchased from Research Genetics, Inc. (Huntsville, AL; http://www.resgen.com) was used. This whole-genome RH panel was constructed by fusing hamster (host) cells with donor cells from a canine (mongrel male) cell culture that had received 3000 rads of X-irradiation. The 12 markers were scored on the RH panel based on their unique retention patterns. Each marker was simultaneously amplified on genomic DNA from 92 cell lines. PCR was performed with 25 ng DNA in a final volume of 15 μl under the conditions of 94°C for 2 minutes; 30 cycles of 94°C (20 seconds), 58°C (20 seconds), and 72°C (20 seconds); and a final extension at 72°C for 5 minutes. The PCR products were separated on 1.8% agarose gel. 

Genotyping

Eleven markers were typed in 150 different dogs from two extended XLPRA-informative canine pedigrees, one with 93 members and the second with 58; one dog was common to both. Marker CUX20001 was amplified with one primer end labeled withγ ³²P-adenosine triphosphate (ATP) for 30 cycles at 94°C (30 seconds), 54°C (30 seconds), and 72°C (30 seconds) and a final extension at 72°C for 15 minutes. PCR products were separated in a 7% denatured polyacrylamide gel (5.6 M urea, 32% formamide) and visualized by autoradiography. The remaining microsatellite markers were typed with PCR under the conditions of 94°C (2 minutes); 30 cycles of 94°C (20 seconds), 58°C (20 seconds), and 72°C (20 seconds); and a final extension at 72°C for 5 minutes. PCR products were size fractionated by 8% PAGE. For intragenic makers of RPGR, TIMP1, and AR, the PCR products were digested with NlaIII, MspI, and Eco109I, respectively, and separated with 6% polyacrylamide gel, as described previously. ^{6 14 15} To unambiguously identify alleles in the related dogs, in all cases the PCR products from parents and offspring were separated and scored on the same gel. 

Map Construction

A total of 11 and 12 makers were used to construct the meiotic linkage and RH maps, respectively; 11 markers were common to both. To place the XLPRA locus on the meiotic linkage map, the disease trait was scored as X-linked with full penetrance and no phenocopy. RH mapping results and genotypic data from pedigrees were analyzed with Multimap (http://compgen.rutgers.edu/multimap/). ¹⁷ Two-point linkage analyses were performed for all pairs of markers. Framework maps were then ordered for both the meiotic and RH maps at a lod score of 2.0, allowing the program to select starting markers with a criterion lod score of 8.0. The framework map was then extended at a lod score of 0.5 to include those markers that could not be placed in the lod score 2.0 map. Meiotic map distances were expressed using the Kosambi map function. Intermarker distance (D) for RH mapping was calculated as D = −ln (1 − θ), where θ is the frequency of breakage in the panel expressed in centirays 3000 (cR₃₀₀₀), reflecting the radiation dose used for generating the RH panel. 

Results

One hundred fifty dogs in two canine pedigrees were used for mapping the XLPRA interval, constructing a meiotic linkage map of CFAX, and establishing homology to the human RP3 interval. Figure 1 shows three representative informative families selected to demonstrate recombinations in the segregating haplotypes. Recombination events are indicated in Figures 1A (II; FH2985, FH2997, TIMP1), 1B (III; FH2985, FH2548, FH2916), and 1C (III; FH3027) by horizontal bars. The number of phase-known informative meioses ranged from 58 to 110 for the markers used in the analysis. Two-point analyses, summarized in Table 2 , showed that loci RPGR, XLPRA, CUX40002, CUX40001, and CUX20001 link tightly to each other with no observed recombinations (recombination fraction θ = 0; lod score, Z = 38.23–49.67; Table 2 ). Flanking this zero-recombination region are markers FH2548 and FH2916. More distantly, the intragenic markers TIMP1 and AR, are closely linked to each other (θ = 0.023, Z = 27.43). 

Multipoint linkage analyses for the meiotic and RH data are illustrated in Figure 2 . For this analysis, starting markers were chosen by Multimap with a criterion lod score of 8.0. A threshold lod score of 2.0 was used to build the meiotic and RH framework maps. Subsequently, the framework maps were expanded to include all remaining markers at a threshold lod score of 0.5. 

The genetic distance covered between the distal and proximal markers on the meiotic map is approximately 47.3 cM from FH2985 (telomeric) to FH3027 (centromeric). Consistent with two-point linkage data, markers RPGR, CUX40002, CUX40001, and CUX20001 (genes RPGR, OTC, SRPX, and TCTE1L) and the XLPRA locus cluster in an interval of less than 1 cM and define a region homologous to the human RP3 interval. Multipoint analysis cannot order the markers within the disease locus. Flanking this region are markers FH2548 (telomeric) and FH2916 (centromeric), with recombination fractions of 0.03 from each marker to the interval and spanning a total distance of approximate 6 cM (Fig. 2B) . In turn, markers FH2985 (telomeric) and FH2997 (centromeric) flank the region defined by FH2548 and FH2916 and are located approximately 30 cM apart (Figs. 2A 2B) . 

The 3000-rad RH map (Figs. 2C 2D) includes the same 11 markers used for the meiotic linkage map, plus an additional intragenic marker that was not informative in the pedigrees used for linkage analysis. These 12 markers span approximately 194 cR₃₀₀₀ from FH2548 to AR. From previously published RH mapping data using a 5000-rad panel, ¹⁰ we also know that DMD is placed between markers FH2985 and FH2548 (Fig. 2E) . 

Marker order is in very good agreement between the meiotic and RH maps. The only exception was FH2985, which was placed on the genetic map but was not assigned to the same linkage group on RH mapping. This could be an anomaly of the RH3000 panel, because this marker was placed in the same RH linkage group with FH2548 and DMD by use of the 5000-rad canine RH panel. ¹⁰  

Discussion

In a previous study, we used five genes (DMD, RPGR, TIMP1, AR, and canine factor IX[ FIX]) to map the XLPRA disease locus on CFAX and establish the relative gene order and linkage distances among these six loci. ⁶ From these studies we know that XLPRA is flanked by DMD and TIMP1, a genetic distance greater than 35 cM, and is tightly linked to RPGR. FIX has been mapped by fluorescence in situ hybridization (FISH) to the distal end of the long arm of CFAX. ¹⁸ AR, the canine androgen receptor gene, has been shown to be tightly linked to the genes phosphoglycerate kinase (PGK) and choroideremia (CHM), both of which were localized by FISH to the proximal end of the long arm of CFAX. ¹⁹ Based on two-point linkage distances from AR, the inferred chromosomal location of genes DMD, RPGR, and TIMP1 and of the XLPRA interval, is therefore on the short arm of CFAX (CFAXp). Taking the previous and current linkage data together with the well-established cytogenetic similarity of the canine and human X chromosomes, ^{20 21 22} it is clear that the chromosomal organization and gene order of CFAX is thus broadly very similar to that of the human X, and unlike, for example, the rearrangement seen in cattle, ²³ sheep and goats, ^{24 25} pigs, ²⁶ and rodents ^{27 28} (Fig. 3) . 

Our present studies have further narrowed the XLPRA disease interval. Eleven polymorphic gene-associated type II markers were developed by identifying microsatellites in canine BAC clones containing genes that map in human to Xp. These include microsatellites associated with the TCTE1L (CUX2001), SRPX (CUX4001), and OTC (CUX4002) genes. In humans, the genes TCTE1L, SRPX, and RPGR, are located within the zero-recombination region of the defined RP3 interval. ³ Our linkage studies in dogs demonstrate the same results and place XLPRA in a zero-recombination interval on CFAXp which is less than 1 cM and contains the RPGR, OTC, SRPX, and TCTE1L genes. The nearest flanking markers, FH2548 (telomeric) and FH2916 (centromeric), span a total distance of approximate 6 cM. 

The order of these four genes cannot be determined from the meiotic map and can only be assigned from the RH data with relatively low confidence. Because the RH mapping panel used in this study was constructed at 3000 rads, its power to resolve gene order among tightly linked genes is relatively limited. The interval from markers FH2548 to FH3027 (D = 29.10 cM and 141.80 cR₃₀₀₀) yields a ratio of 4.87 cR₃₀₀₀/cM, similar to the previously published average of 6.71 cR/cM obtained with the canine cR₅₀₀₀ panel. ¹⁰ At this level of resolution, the best-supported order for these four genes, on the extended lod 0.5 map, is RPGR-OTC-SRPX-TCTE1L. 

In evaluating this order, and the likelihood of a microrearrangement in the dog compared with the human, two other observations are relevant. First, exonic sequences from both RPGR and OTC have been identified within the same canine BAC clone, thus these two genes are contiguous, as in humans. Second, stretches of both exonic and noncoding sequence from canine BAC clones positive for TCTE1L and SRPX match the sequence of a human cosmid contig spanning the interval from TCTE1L to SRPX (GenBank accession no. AL121578). The sequence match for noncoding regions presumably represents evolutionarily conserved regulatory sequences. We expect this indicates that TCTE1L and SRPX are also likely to be contiguous in the dog, as in humans. Thus, taking all available information together, although the placement of these four genes is highly conserved between dog and human, it is slightly more likely that the canine interval is ordered (RPGR-OTC)-(SRPX-TCTEIL), compared with TCTEIL-SRPX-RPGR-OTC on HSAXP. Resolution of this question depends on completion of current sequencing studies of the canine XLPRA interval. 

On HSAX, the relationship between physical and meiotic linkage distance varies markedly along the chromosome. A large region of low recombination, 0.16 cM/Mb compared with an average rate of 1.3 cM/Mb, ²⁹ is present just distal to the centromere in the region of Xq13. This region encompasses the X-inactivation center and several genes including PGK, PHKA1, ATP7A, and CHM. This relationship is maintained in mouse and sheep. ²⁴ Similarly, a region of low recombination on CFAX between AR, PGK, and CHM ¹⁹ apparently extends, from our analyses, to include the TIMP1 gene. Although we have observed recombination between the TIMP1 and AR genes in the larger and more informative pedigrees used for the meiotic linkage map, these two genes could not be integrated well in the multipoint linkage map. Because they map on either side of the centromere in humans, a location that is likely to be maintained in dogs, we assume that in the dog, suppression of recombination extends across this region as well. 

This map should be useful for positional cloning and refinement of linkage results for canine X-linked retinal diseases and other genetic traits. In particular, they will prove useful in our ongoing efforts to clone XLPRA, the locus homologue of human RP3. 

 Supported by National Eye Institute Grants EY06855 and EY13132; the Foundation Fighting Blindness; Morris Animal Foundation/The Seeing Eye, Inc. EAO was funded by the American Kennel Club Canine Health Foundation, the American Cancer Society, and an award from the Burroughs Wellcome Fund.

 Submitted for publication April 13, 2001; accepted June 5, 2001.

 Commercial relationships policy: P (GDA, GMA); N (all others).

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked“ advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Marker and PCR Primer Information Used for Typing

Marker	Forward Primer (5′–3′)	Reverse Primer (5′–3′)	Gene Associated with Marker in BAC Clone
FH2985	AGGGGCAACTCAAAGGTTAC	ATGTGTGGAACTGAGCTTGG	Unknown
FH2548	AAGGGAGGAAACAATGCTGA	GACATTCAGAGATTTCCGCC	Unknown
FH2916	CTGGCATACCTCAACAAAGC	CAAAATTGTGGACCTAAAATCG	Unknown
FH2997	TCTCTCATCTCTCCCTCTGC	AAAAACTGACATCAACAAATGC	Unknown
FH3027	GTTTCCTCACATGCAAAAGC	GCTGGAGGTCAAGGATAAGG	Unknown
CUX40001	TGGCACGCGCGGATCCTTGG	TGTGCGCGCTGCCTGGGTAGG	SRPX
CUX40002	GCATGGAGTTTCCTTGCTCCTC	TATTCAAGGTGCTGAATGGGGA	OTC
CUX20001	GGGTCTGAGCATGGCTTTGA	TTGATGCCTCGGGCTTGGG	TCTE1L
RPGR/NlaIII	GACATAGGTAATGACTCAGGCCAG	AATTTGGACAGTATGTGTTCGGTC	RPGR 15
TIMP1/MspI	TAGATGTCGGGGTTCCAAGGAGTG	TAGATGTCGGGGTTCCAAGGAGTG	TIMP1 14
AR/Eco109I	CCGTGAGCGCAGCACCTCCCGGTG	TGCTCTCCCGCTGCTGCTACCTTCTG	AR 6 11
ALAS2	GATAAGCAGAGGTCTGGGAAAGGAACC	TAGGAGCGCACCATGTCCACCAAGTC	ALAS 6 11

 Markers of FH2985, FH2548, FH2916, FH2997, FH3027, CUX40001, CUX40002 are tetranucleotide repeats. CUX20001 is a dinucleotide repeat. RPGR/NlaIII, TIMP1/MspI and AR/Eco109I are intragenic RFLP markers for genes RPGR, TIMP1, and AR. ALAS2 is an intragenic marker that is not informative in the pedigree but was used for RH mapping.

Figure 1.

 

Three subsets of representative pedigrees (A, B, C) from informative and related dog families illustrate the critical recombinants for the markers typed. The order of loci is arranged according to multipoint linkage analysis (see Fig. 2 ). Horizontal bars: obligate recombination of markers in the haplotype of some representative dog families. Phenotypic status of male (squares) and female (circles) animals is homozygous normal (open), heterozygous female (open circle with dot), hemizygous affected male (filled square), and homozygous affected female (filled circle). The restriction fragment length polymorphic (RFLP) marker of AR and the XLPRA locus are not presented in the haplotype.

Table 2.

 

Best Two-Point Recombination Fractions and Lod Scores for Markers Used for Mapping

Loci	2548	RPGR	XLPRA	40002	40001	20001	2916	2997	TIMP1	3027	AR
2985
θ	0.235	0.225	0.228	0.227	0.197	0.235	0.234	0.405	0.300	0.333	0.315
Z	22.01	24.45	24.34	23.06	25.20	21.48	22.38	18.35	17.90	21.24	20.48
2548
θ		0.048	0.048	0.048	0.033	0.058	0.05	0.138	0.146	0.231	0.152
Z		32.69	32.69	31.21	32.92	28.73	30.35	21.74	22.91	23.37	22.81
RPGR
θ			0.000	0.000	0.000	0.000	0.029	0.112	0.136	0.226	0.175
Z			49.67	48.47	48.77	39.13	41.67	25.54	24.92	25.96	25.26
XLPRA
θ				0.000	0.000	0.000	0.030	0.115	0.139	0.229	0.164
Z				48.17	48.47	38.83	41.40	25.34	24.73	25.85	25.51
40002
θ					0.000	0.000	0.029	0.114	0.136	0.226	0.175
Z					47.56	39.13	41.67	25.29	24.92	24.76	24.06
40001
θ						0.000	0.029	0.118	0.129	0.233	0.175
Z						38.23	41.67	24.50	25.57	25.25	24.96
20001
θ							0.014	0.050	0.119	0.171	0.154
Z							35.96	22.36	24.39	25.15	23.72
2916
θ								0.059	0.109	0.200	0.159
Z								27.56	26.63	25.77	24.46
2997
θ									0.065	0.112	0.154
Z									21.77	25.24	19.55
TIMP1
θ										0.116	0.023
Z										26.43	27.43
3027
θ											0.207
Z											24.01

 Loci 2985, 2548, 2916, 2997, 3027, 20001, 40001, and 40002 refer to microsatellite markers FH2985, FH2548, FH2916, FH2997, FH3027, CUX20001, CUX40001, and CUX40002, respectively (Table 1) . Loci RPGR, AR, and TIMP1 are all intragenic markers. Order of markers is the one supported by the 0.5 lod score RH map; XLPRA is placed arbitrarily between RPGR and OTC.θ , maximum likelihood estimate of recombination fraction between loci; Z, lod score. Data are shown in bold face where θ is zero.

Figure 2.

 

Integrated meiotic and radiation hybrid map of CFAXp. Twelve markers are shown: seven gene-associated markers (RPGR, TIMP1, AR, ALAS, CUX20001, CUX40001, CUX40002) and five anonymous type II markers (FH2985, FH2548, FH2916, FH2997, FH3027). For both the meiotic (A, B) and RH (C–E) components, a lod score 2 framework map (B, C) is flanked by a lod score 0.5 map (A, D) that includes markers not incorporated into the framework map. Markers positioned on both the meiotic and RH maps are connected with dotted lines. Markers CUX20001, CUX40001, and CUX40002, derived from BACs containing the TCET1L, SRPX, and OTC genes, respectively, are identified by the corresponding gene designation and, together with XLPRA, form a zero-recombination interval with RPGR. Distances between markers are in centimorgans (cM) and centirays (cR₃₀₀₀) for genetic and RH maps, respectively. The DMD marker was placed between FH2985 and FH2548 (E) based on RH mapping data, ¹⁰ and added manually.

Figure 3.

 

Schematic representation, not drawn to scale, of gene order for selected genes in the X chromosome of six mammalian species. In those species in which the genes have been mapped, the gene cluster of AR, IL2RG, PGK, and CHM remains together. In mouse, rat, cow, and goat there is a rearrangement of gene order in comparison to the human X chromosome. Data for this figure were obtained from different sources: goat, ²³ cow, ^{23 30} rat, ^{27 28} mouse and human, ^{27 28} (and also see http://www.informatics.jax.org/menus/homology_menu.shtml), dog, ^{6 18 19} and the present study.

The authors thank James Kijas for helpful discussions, and Keith Watamura for figures.