Proc. Natl. Acad. Sci. USA Vol. 88, pp. 4811-4815, June 1991 Genetics Mouse platelet-derived a is deleted in WJ9H and patch on 5 E. ANNE SMITH*, MICHAEL F. SELDINt, LISA MARTINEZ*, MARK L. WATSONt, GOUTAM GHOSH CHOUDHURY*, PETER A. LALLEYt, JACALYNE PIERCE§, STUART AARONSON§, JANE BARKER¶, SUSAN L. NAYLOR*, AND ALAN Y. SAKAGUCHI* II *Department of Cellular and Structural Biology, University of Texas Health Science Center, San Antonio, TX 78284-7762; tDuke University Medical Center, Durham, NC 27710; tCenter for Molecular Genetics, Wayne State University, Detroit, MI 48202; National Institute, Bethesda, MD 20892; and IThe Jackson Laboratory, Bar Harbor, ME 04609 Communicated by Elizabeth S. Russell, February 26, 1991

ABSTRACT The mouse W19H is an x-ray- nonviable, with mice dying before or shortly after birth. induced of more than 2 centimorgans on chromosome Homozygous Rw and Ph mice die prenatally (1, 14). 5 encompassing the white spotting mutation W (encoded by the Demonstration that W and Kit are allelic has provided an Kit protooncogene), patch (Ph), and recessive lethal (1) loci. The example of Mendelian inheritance of a mutant receptor platelet-derived a gene (PDGFRA) like linked to a developmental abnormality in Kit encodes a transmembrane receptor . By mouse. Molecular characterization of W mutants also pro- using mouse-Chinese hamster somatic cell hybrids and hap- vides a paradigm for determining whether other developmen- lotype analysis in interspecific backcross mice, mouse Pdgfra tal mutants in mouse, especially those affecting coat color was mapped to in tight linkage with Kit. pigmentation, are linked to receptor tyrosine or to Hybridization of a PDGFRA probe to DNAs from WI9H/+ growth factors. Recently, the human KIT and PDGFRA patch heterozygous mice, and their were mapped to the same region on the proximal long heterozygous mice and arm of human chromosome 4 (15, 16). Because of the general wild-type littermates, demonstrated deletion ofPdgf. Pulsed- interest in determining genes in the region ofthe W , we field gel electrophoresis indicated thatKitandPdgfra are linked tested the hypothesis that the mouse Pdgrfa gene is located on a 630-kilobaseMlu I DNA fragment. Thus the W19H deletion on chromosome 5 and is encompassed by the WI9H deletion, removes at least two receptor tyrosine kinases and the results as the human chromosome 4q region containing KIT and suggest Pdgfra as a candidate for the Ph locus. PDGFRA comprises a conserved linkage group found on mouse chromosome 5 (17). In the present report we show that A closely linked gene triplet on mouse chromosome 5, the mouse Pdgfra gene maps to chromosome 5, is deleted comprised of the rumpwhite (Rw), white-spotting (W), and both in WI9 and Ph , and is found on a patch (Ph) loci, yields a dominant white spotting phenotype 630-kilobase (kb) Mlu I fragment along with Kit. The Pdgfra arising from effects on developing melanoblasts (1-3). Mu- gene is, therefore, a candidate for the Ph locus. tations at the W locus generally lead to pleiotropic effects not only on coat color pigmentation but on hematopoiesis leading to macrocytic and on germ cell development as well, MATERIALS AND METHODS although the effects ofindividual alleles can vary significantly Mouse-Chinese Hamster Cell Hybrids. The EBS mouse- (2-4). The W locus was recently shown to encode the Kit Chinese hamster somatic cell hybrids have been extensively protooncogene (5, 6), a transmembrane tyrosine kinase re- characterized and described (18) and were formed by the ceptor for a encoded by the steel locus (e.g., refs. 7 and fusion of Chinese hamster fibroblasts (clone E36) with 8). Several W alleles encode receptors with reduced or mouse BALB/c spleen cells. undetectable in vitro tyrosine kinase activity (9-11). The Mice. C3H/HeJ-gld/gld and Mus spretus (Spain) mice and reduced ability of the mutant receptors to phosphorylate key [(C3H/HeJ-gld/gld X M. spretus)Fl x C3H/HeJ-gld/gld] cellular substrates might underlie their developmental effects interspecific backcross mice were bred and maintained as in affected tissues, as there is some suggestion that the described (19). M. spretus was chosen as the second parent severity of a particular W allele is reflected by the in vitro in this cross because of the relative ease of detection of kinase activity of the encoded Kit (10). informative restriction fragment length variants (RFLVs) in Lyon et al. (12) described a more extensive chromosome 5 comparison with crosses using conventional inbred strains. lesion, the WJ9H deletion, which encompasses W (Kit), Ph, C3HeB/FeJ-WI9H and C57/BL6J-Ph mice were obtained recessive lethal (I), and more than 2 centimorgans (cM) of from The Jackson Laboratory. DNA, but not the Rw locus, which is centromere-proximal, Mouse Gene Linkage Analyses. Maximum likelihood esti- or the a-casein gene (Csna), which is distal to Ph (13). mates of recombination probabilities and their standard er- Homozygous WI9H embryos die before implantation whereas rors among backcross progeny were calculated according to heterozygous WJ9H/+ mice are viable, with minimal white Green (20). The best gene order was determined according to spotting, and devoid of pigment dilution or anemia (14). The Bishop (21). Ph locus, which is also deleted in the WJ9H mutation, leads DNAs and Southern Blot Hybridization. For gene copy to a dominant white spotting phenotype similar to that of number analysis, DNA was obtained from the livers of piebald (s/s) and belted (bt/bt), in which areas of pigmented C3HeB/FeJ-WI9H heterozygous mice and their wild-type and nonpigmented fur are sharply demarcated (14). Homozy- littermates by the method of Blin and Stafford (22) and gotes for approximately half of the known W alleles are Abbreviations: RFLV, restriction fragment length variant; PFGE, pulsed-field gel electrophoresis; cM, centimorgan(s). The publication costs of this article were defrayed in part by page charge ITo whom reprint requests should-be addressed at: Department of payment. This article must therefore be hereby marked "advertisement" Cellular and Structural Biology, University of Texas Health Sci- in accordance with 18 U.S.C. §1734 solely to indicate this fact. ence Center, 7703 Floyd Curl Drive, San Antonio, TX 78284-7762. 4811 4812 Genetics: Smith et A Proc. Natl. Acad. Sci. USA 88 (1991) quantitated by fluorimetric assay with Hoechst 33258 using A -6.6 calf thymus DNA as a standard. For gene linkage analysis, was mouse organs. DNAs were cleaved DNA isolated from ks X -4.4 with restriction according to manufacturers speci- oo***~~~~~1A *- fications, and the fragments were fractionated on 0.8%'or 0.9%o agarose gels. DNAs were transferred to a Nytran - 2.3 membrane (Schleicher & Schuell) and hybridized as de- -2.0 scribed (23) using buffers containing 35%-40 (vol/vol) formamide and 10% (wt/vol) dextran sulfate at 42TC or buffers without formamide at 65TC (19). Probes included a 1 2 3 4 5 6 7 8 9 1011 12131415 B *qpr-ni- * - 1724-base-pair (bp) insert encompassing the cytoplasmic a domain of human PDGFRA cDNA (15), a 1.8-kb EcoRP -9.4 mouse Kit cDNA fragment (5), and a 900-bp Pst I mouse -6.6 a-fetoprotein (Afp) cDNA fragment (24). All probes were 4.44 as labeled by the hexanucleotide technique with [a-32PjdCTP A* .:40ao described (19). _,.: .. Scanning Densitometry. At least three different exposures I .. 44 so 4 -2.3-2.0 of each filter to x-ray film were made and analyzed. by soft -2 .0 laser 'scanning densitometry (model GS300; Hoefer). Areas under the peaks were measured (Photogrametric System Instruments, San Antonio, TX). A background measurement was determined and this value was subtracted from the peak reading. Wild-type signals for Afp and Pdgfra were assigned FIG. 1. Hybridization of PDGFRA and Kit probes to mouse- the value 100%. For Afp any deviation from 1 of the ratio of Chinese hamster somatic cell hybrid DNA. DNAs were cleaved with WJ9H/+ to wild type or of Ph/+ to wild type reflects Pvu II and blotted, and the filter was hybridized sequentially to differences in DNA loading. Therefore, to correct for vari- PDGFRA (A) and, after stripping, to Kit (B). DNAs are from cell ations in DNA loading, the W19H/+ Pdgfra and the Ph/+ hybrids (lanes 1-13), Chinese hamster cell line RJK36 (lanes 14), and Pdgfra mean values were divided by the mean values for mouse RAG cells (lanes 15) (these lanes refer to A and B). Cell W)9H/+ Afp and Ph/+ Afp, respectively. The corrected hybrids represented in lanes 1-13 are in the same order as those listed values were then used to calculate the ratios ofPdgfra to Afp in Table 1. Arrows indicate mouse Pdgfra gene fragments whose for W'9H/+ DNA and for Ph/+ DNA'shown in Fig. 4. presence was scored in cell hybrids. The PDGFRA probe encom- Pulsed-Field Gel Electrophoresis (PFGE). Single-donor passes the cytoplasmic domain (15). The K3A Kit probe encom- passes residues 136-791 and includes -60%6 of the C3H/HeJ-gld/gld mouse peripheral lymph node cells were cytoplasmic domain (5). Virtually all of the mouse DNA fragments prepared and suspended in'agarose blocks as described (25). hybridizing with PDGFRA had mobilities distinct from those hy- Digests of nuclei suspended in agarose blocks were carried bridizing with Kit. Cell hybrids in lanes 1, 3, and 8-10 are positive out using 0.5-20 unit(s) ofrestriction endonuclease per ,g of for mouse Pdgfra and Kit. Size markers (in kbp) in A and B are DNA (Boehringer Mannheim) in lx appropriate restriction selected bacteriophage A HindIll fragments. buffer for 4 hr at 100% activity temperature. Reactions were terminated by the addition of 0.5 M EDTA. PFGE was yeast Saccharomyces cerevisiae YNN295 (Bio-Rad), yeast performed as described (25). Large DNA fragments ranging Schizosaccharomyces pombe 972, and concatemers of intact in size from 1000 to 6000 kb were separated employing bacteriophage A DNA (FMC). After electrophoresis, gels ramped pulse times of 15 min to 90 min using a' LKB- were stained with ethidium bromide to visualize size stan- Pharmacia apparatus and a running time of 144 hr at 54 V. dards, which were marked with India ink prior to alkali DNA fragments 200 kb to 1200 kb in size were separated transfer onto nylon membranes (25). Hybridizations of ra- using pulse times of 70 sec to 145 sec at 145 V for 46 hr. High dioactively labeled PDGFRA and Kit probes were performed molecular weight size standards included chromosomes of as described elsewhere (17). Table 1. Segregation of mouse chromosomes with Pdgfra in somatic cell hybrids Mouse chromosome Cell hybrid Pdgfra 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 X EBS-1 + + + + + -* + + + + + - + + + + + + + + + EBS-2 - + + + - - + + + + + - + + - + + + - + + EBS-5 + + + + + + + + + - + - + + + + + + + + EBS-9 - - + + + - + + + + - - + + - + + + + + + EBS-10 - - + + - - - + - - + - + + + + + - + - + EBS- - + . . . . . + - - + - + - - + + - - - + EBS-15 - - + + + - + + + - + - + + - + - + - + + EBS-17 + + + - - + - + - + - - - + - + - + - - + EBS-51 + - + + - + + + + + - - + - - + + + + + + EBS-71 + + + + - + - + -* + - - + - - + + + + - + EBS-5CSt - + + + + - + + + - + - + + - + + + - + - EBS-9CSt + + + + + + + - + + - + + + + + EBS-13CSt - - + - - - - + - - - - + - + + + + - + - % discordant hybrids 31 54 54 54 0* 46 62 46 31 69 38 69 62 38 62 62 46 38 62 38 +, Gene or chromosome present; -, gene or chromosome absent. *EBS-1 contains both and gene markers for chromosome 5. EBS-71 contains Insr and Plat, but not Aprt, markers for chromosome 8. Both hybrids were considered "concordant" in the calculation of percent discordant hybrids. tEBS-CS hybrids were tested for marker enzymes only. Genetics: Smith et al. Proc. Natl. Acad. Sci. USA-88 (1991) 4813

Pdgfro Table 3. Linkage interval between genes studied 23.0- Linkage interval r E T Kit-Pdgfra 0 0 3.1 9.4- Pdgfra-Cas 0.9 0.0 4.8 6.6- Cas-Afp 0.9 0.0 4.8 4_3 Afp-Mgsa 1.8 0.2 6.1 4.3- Mgsa-Gus 19.3 12.47 27.83 r, Recombination frequency; r and F, 95% confidence intervals based on binomial distribution.

2.3- analyses for approximately 85% of the mouse genome (19, 2.0- FIG. 2. Southern blot identification of 28-32; M.F.S., unpublished results). An informative Pdgfra a Pdgfra RFLV. The informative Taq I RFLV that distinguished the two parental strains used to restriction digest is shown with the mo- generate the backcross is shown in Fig. 2. At each locus 1.3- lecular size standards (in kb) indicated to either the homozygous C3H (lane CC) pattern or heterozy- the left. The arrow signifies a band pre- 1. - sent in DNA from (C3H/HeJ-gld/gld x gous F1 pattern (lane SC) was observed in each of 114 meiotic events examined. The RFLV with 0.9- M. spretus)F, (lane SC) mice that are not Pdgfra segregated closely cc sc present in DNA from the C3H/HeJ-gld/ several mouse chromosome 5 genes previously examined in Taq gid parental mice (lane CC). this cross (32) (Tables 2 and 3). There were no crossovers in 114 meiotic events between Kit and Pdgfra (upper 95% RESULTS confidence limit = 3.1 cM). The best gene order (21) (+SD) (20) indicated the following Mapping Pdgfra in Somatic Ceil Hybrids. The human PDG- linkage relationships: (cen- tromere) cM ± 0.9 ± FRA probe detected major hybridizing mouse Pvu II DNA Kit-Pdgfra-(0.9 cM)-Csna-(0.9 cM 0.9 cM ± 1.2 cM ± fragments of 6.1, 2.6 cM)-Afp-(1.8 cM)-Mgsa-(19.3 3.7 cM)-Gus. 3.7, 3.1, and kb, plus at least three By using Bishop's formula (21), this gene order was greater additional minor fragments (Fig. 1A; control Chinese hamster than 3000-fold more likely than the best alternative order. and mouse DNAs are in lanes 14 and 15, respectively). The Deletion of Pdgfra in W19/+ and in Ph/+ DNAs. The 3.7-, 3.1-, and 2.6-kb fragments were sufficiently resolved mapping results raised the possibility that the mouse W19H from Chinese hamster DNA fragments to assess their segre- chromosome 5 deletion might also encompass Pdgfra, as this gation in mouse-Chinese hamster somatic cell hybrids. The deletion covers an estimated 2-4 cM ofDNA and includes the PDGFRA probe did not detect sequences on mouse chro- W locus (Kit) but neither Rw, which is more proximal, nor mosome 18 (where Pdgfrb and Csflr are located; refs. 26 and Csna, which is distal to W (Kit) and Ph (12, 13). Liver DNAs 27) (Fig. 1 and Table 1) and gave a pattern of hybridization to from heterozygous C3HeB/FeJ-Wl9H mice and their wild- mouse DNA that was distinct from that observed with a type littermates were cleaved with several restriction en- mouse Kit probe (Fig. 1B). However, segregation of mouse zymes and various amounts were fractionated by gel elec- Pdgfra gene fragments was completely concordant with Kit trophoresis followed by hybridization to mouse chromosome and chromosome 5, indicating synteny. 5 probes. Fig. 3A shows the results ofhybridization to the Afp Localization of the Pdgfra Gene on Mouse Chromosome 5. probe, which maps distal to the WI9H deletion and, therefore, To localize Pdgfra, we utilized DNAs isolated from [(C3H/ is present in diploid copy number in somatic tissues. The HeJ-gld/gld x M. spretus)F1 x C3H/HeJ-gld/gld] interspe- identical filter was then hybridized to the PDGFRA probe cific backcross mice. This panel of 114 DNAs has been (Fig. 3B). After each hybridization, resulting signals were characterized for more than 200 markers that allow linkage quantitated by scanning densitometry, and signals obtained with the PDGFRA probe were normalized to those obtained Table 2. Analysis of haplotypes in (C3H/HeJ-gld/gld x M. with the Afp probe. The results clearly indicated that Pdgfra spretus)F1 x C3H/HeJ-gld/gld backcross mice was in haploid copy number in Wl9H/+ liver DNA and was Gene order No event 1 recombination event A B Kit/Pdgfra CCSC CC SC CC SC CC SC CC SC 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 x x Cas CCSC SC cc cc SC cc SC cc SC x x -23.1 - Afp CCSC SC cc SC cc cc SC cc SC 410_ft x x -9.4 - Mgsa CCSC SC cc SC cc SC cc cc SC 6-6. - x x go Gus CCSC SC CC SC CC SC CC SC CC Me Mice, no. 32 56 1 0 0 1 1 1 9 13 - 4.4 - Ub Total M m events, no. 88 1 1 2 22 - 2.3 - -2.0- Each column represents a chromosome derived from the F1 parent that might be observed in the backcross mice. Alternative genotypes at each locus are designated CC, C3H/HeJ-gld/gld-like, or SC (C3H/HeJ-gld/gld x M. spretus)Fl-like, and were determined by FIG. 3. Hybridization ofAfp and PDGFRA probes to DNAs from RFLV typing (see Fig. 1 and ref. 32). The number of chromosomes WJ9H/+ mice and their wild-type littermates. DNA samples were with each crossover combination in the 114 meiosis examined is cleaved with EcoRI (lanes 1 and 2), BamHI (lanes 3 and 4), Pst I indicated at the bottom of each column. Presumed sites of recom- (lanes 5 and 6), or Pvu II (lanes 7 and 8) and sequentially hybridized bination are indicated with a x. The minor differences in haplotype to an Afp probe (A) and then to a PDGFRA probe (B). C3HeB/FeJ assignments in this table compared with those included in a previous DNA samples are in lanes 1, 3, 5, and 7. C3HeB/FeJ-W'9H DNA study (32) reflect corrections of previous assignments. samples are in lanes 2, 4, 6, and 8. 4814 Genetics: Smith et al. Proc. Natl. Acad. Sci. USA 88 (1991) observed in digests with four restriction enzymes (EcoRI, 630-kb segment of DNA on mouse chromosome 5 in the BamHI, Pst I, and Pvu II) (Fig. 4A). present study. This DNA segment normally must lie at least Gene copy number determination in DNA from Ph!+ partially within the region encompassed by the WJ9H muta- heterozygous mice also indicated a deletion of the Pdgfra tion as both Kit and Pdgfra are deleted. Although the gene (Fig. 4B) and suggests that the Ph mutation involves a intergenic distance may be different, this physical configura- chromosomal deletion. Pdgfra fragments of altered mobility tion is reminiscent ofhuman PDGFRB and CSFIR, which are compared to wild-type DNA were not detected in WI9H/+ within 500 bp in a head-to-tail arrangement on the distal long DNA (Fig. 3B) or in Ph/+ DNA (data not shown). arm of chromosome 5 (34). It has been speculated that the Analysis of Pdgfra by PFGE. To more precisely define the close proximity of PDGFRB and CSFIR might reflect an genomic organization of this chromosome 5 region and in unusual mode of gene regulation in the various tissues in particular the interval between Kit and Pdgfra, we initiated which the receptors are expressed (34). It is conceivable that efforts to develop a long-range restriction map by using human KIT and PDGFRA and mouse Pdgfrb and Csflr will PFGE. The mouse Pdgfra and Kit probes detected identical also be found in tandem arrays on their respective chromo- 630-kb and 740-kb Mlu I and 1200-kb Nru I bands on somes. Examination of the intergenic sequences separating sequential hybridization ofthe same membrane (Fig. 5). Thus the pairs of related receptors might reveal information con- these data indicate that the genes encoding Pdgfra and Kit are cerning their regulation and evolution. located within a 630-kb segment of mouse chromosome 5. Thus with previous studies (6, 12, 13), our findings show that the WI9H chromosome deletion removes at least two receptor tyrosine kinases important for development. DISCUSSION Whereas the role for Kit is known through studies of the W The human PDGFRA gene encodes a transmembrane recep- locus, a role for Pdgfra in early has tor with intrinsic tyrosine kinase activity (15, 33). PDGFRA been implied by the spatial distribution and temporal expres- exhibits homology with PDGFRB (the 83 receptor), Kit, sion of platelet-derived growth factor a (Pdgfa) and Pdgfra colony-stimulating factor 1 receptor (CSFIR), and fibroblast transcripts in mouse embryos (35). Expression ofPdgfra and growth factor receptor, and together these form a Pdgfa is detectable at 6.5 days of development, and it is subfamily of receptor kinases of the src superfamily (15, 33). known that Pdgfa message is maternally encoded in mouse Whereas Pdgfrb and Csflr are syntenic on mouse chromo- (35). The a isoform of platelet-derived growth factor has also some 18 (refs. 26 and 27), Pdgfra and Kit were localized to a been shown to drive the proliferation ofbipotential 0-2A glial

2 I I-IN "Q-, 4 "q 4. 2-A * .W19H/ ,.p", 1-1 x co 4. 4. 4 4 4. -.. . 0 WILD TYPE 40

.. I- LM-i _ w: -LM

Wi .* 125 - --1125 I- _U 020 - 945 - -1020 o 00 - 945 B50-_800- 850 770- * * * * -770 700- 700 630 - *-630 580 - - 580 0 460- -460 EcoRI Bam HI Pst I Pvu 11 370- 290 - - 370 245 - - 290 2 - 245 B U PATCH/+ 0 WILD TYPE Pdgfrc A< - FIG. 5. Sequential hybridization of Kit and Pdgfra probes on a PFGE blot separating DNA fragments. DNA from a C3H-HeJ-gld/ gid lymph node was separated by PFGE. Restriction endonucleases are indicated. Nt, Not I; Nt(P), Not I partial digest; Ml, Mlu I; Ml(P), Mlu I partial digest; Nt/Ml, Not I/Mlu I double digest; Nr, Nru l; 0 Nr/Ml; Nru I/Mlu I double restriction digest. Molecular size stan- LO) dards (in kb) are to the right and left. LM indicates limiting mobility. LU 1 Coincident Mlu I fragments of630 and 740 kb are seen for both Pdgfra and Kit probes in several lanes containing single or double digests, and a common 1200-kb Nru I fragment is visualized with both probes. Additional small Mlu I fragments (-200 kb) were detected with the Pdgfra probe but not the Kit probe. The Pdgfra and Kit probes Eco RI Bam Hi Pst I Hind III hybridized to unique Not I fragments of 580 kb and 650 kb, respec- tively. Pdgfra hybridized to the 580 kb, 630 kb, and 740 kb fragments RESTRICTION DIGESTS in the Not I/Mlu I double digest indistinguishable from those visualized with the Not I and Mlu I single digests. This result FIG. 4. Densitometric determination of Pdgfra gene copy num- suggested that the Not I fragment was internal to the Mlu I fragments ber. (A) Wild-type and WI9H/+ DNAs. The numbers over the bars since no new-sized fragments were detected. In contrast, Kit hy- represent the ratio of Pdgfra to Afp determined by scanning densi- bridized to 240-kb and 420-kb Not I/Mlu I fragments that were not tometry of the x-ray film shown in Fig. 3. (B) Wild-type and Ph/+ seen with Mlu I or Not I single digests, suggesting that at least one DNAs. Numbers over bars represent the ratio of Pdgfra to Afp. of the Mlu I sites is internal to the Not I fragment detected with the Restriction enzymes used to digest DNAs are indicated under the Kit probe. Thus these results suggest that Pdgfra and Kit are located bars. within 630 kb. Genetics: Smith et al. Proc. Natl. Acad. Sci. USA 88 (1991) 4815

progenitor cells in developing rat optic nerve and prevents gley, K. E., Smith, K. A., Takeishi, T., Cattanach, B. M., differentiation in culture (36-38). Of the two Galli, S. J. & Suggs, S. V. (1990) Cell 63, 213-224. their premature A., receptors for platelet-derived growth factor, only PDGFRA 9. Nocka, K., Majumder, S., Chabot, B., Ray, P., Cervone, Bernstein, A. & Besmer, P. (1989) Genes Dev. 3, 816-826. can B isoforms ofplatelet-derived growth bind both the A and 10. Tan, J. C., Nocka, K., Ray, R., Traktman, P. & Besmer, P. factor with high affinity, implying a role in the rat optic nerve (1990) Science 247, 209-212. for this receptor. 11. Reith, A. D., Rottabel, R., Giddens, E., Brady, C., Forrester, The close genetic linkage of W and Ph (14), along with the L. & Bernstein, A. (1990) Genes Dev. 4, 390-400. genetic and physical mapping described here, suggest a 12. Lyon, M. F., Glenister, P. H., Loutit, J. F., Evans, E. P. & possible link with Pdgfra. Both W and Ph loci yield a Peters, J. (1984) Genet. Res. (Cambridge) 44, 161-168. dominant white spotting phenotype and Ph homozygous 13. Geissler, E. N., Cheng, S. V., Gusella, J. F. & Housman, embryos die prenatally (14). Similarly, the timing and migra- D. E. (1988) Proc. Nati. Acad. Sci. USA 85, 9635-9639. tion of progenitor cells, from which melanoblasts 14. Gruneberg, H. & Truslove, G. M. (1960) Genet. Res. (Cam- are derived, appear to be affected in homozygous Ph mice bridge) 1, 69-90. (39). The Pdgfra transcript is expressed in tissues that arise 15. Matsui, T., Heidaran, M., Miki, T., Popescu, N., La Rochelle, from the neural crest, with high levels expressed in , W., Kraus, M., Pierce, J. & Aaronson, S. A. (1989) Science , and (15). The deletion of Pdgfra 243, 800-804. smooth muscle, 16. Yarden, Y., Kuang, W.-J., Yang-Feng, T., Coussens, L., in DNA from Ph/+ mice raises the possibility that Pdgfra Munemitsu, S., Dull, T. J., Chen, E., Schlessinger, J., and Ph are allelic, although some uncertainty remains be- Francke, U. & Ulirich, A. (1987) EMBO J. 6, 3341-3351. cause the extent of deletion has not yet been mapped pre- 17. Lalley, P. A., Davisson, M. T., Graves, J. A. M., O'Brien, cisely, and for example, might include the recessive spotting S. J., Womack, J. E., Roderick, T. H., Creau-Goldberg, N., (rs) locus. Doolittle, A. L. & Rogers, J. A. (1989) Cytogenet. Cell Genet. A mechanism whereby enzymatically defective receptor 51, 503-532. proteins encoded by certain W alleles "poison" the activity 18. Francke, U., Lalley, P. A., Moss, W., Ivy, J. & Minna, J. D. of the normal protein in cells through dimerization has been (1977) Cytogenet. Cell Genet. 19, 47-84. 11) and might also explain, for example, the 19. Seldin, M. F., Morse, H. C., III, Reeves, J. P., Scribner, proposed (10, C. L., LeBoeuf, R. C. & Steinberg, A. D. (1988) J. Exp. Med. autosomal dominant inheritance of resistance associ- 167, 668-693. ated with mutant insulin receptors (40). The white spotting 20. Green, E. L. (1981) in Genetics and Probability in Animal phenotype in Ph/+ mice is more pronounced compared with Breeding Experiments, ed. Green, E. (MacMillan, New York), W'9H/+ mice (12, 14) and is even more extensive in pp. 77-113. W/+:+/Ph compound heterozygotes. In contrast, WJ9H/+ 21. Bishop, D. T. (1985) Genet. Epidemiol. 2, 349-361. mutant mice exhibit minor white spotting, and no anemia 22. Blin, N. & Stafford, D. W. (1976) Nucleic Acids Res. 9, (12), despite the deletion of two receptor tyrosine kinases. It 2303-2308. is therefore unclear whether the phenotype ofPh/+ mice can 23. Sakaguchi, A. Y., Lalley, P. A., Ghosh-Choudhury, G., Mar- E.-S., Killary, A. M., Naylor, S. L. & Wang, be ascribed to a single-gene deletion. The possibility that the tinez, L., Han, to L.-M. (1989) Genomics 5, 629-632. Ph chromosome deletion affects other genes in addition 24. Tilghman, S. M., Kioussis, D., Gorin, M. B., GarciaRuis, J. P. Pdgfra should be considered. These questions can be exper- & Ingram, R. S. (1981) J. Biol. Chem. 254, 7393-7399. imentally addressed as the appropriate reagents become 25. Kingsmore, S. F., Watson, M. L., Howard, T. A. & Seldin, available. M. F. (1989) EMBO J. 8, 4073-4080. 26. Wang, L.-M., Killary, A. M., Fang, X.-E., Parriott, S. K., A.Y.S. dedicates this work to the memory of his brother Eckford, Lalley, P. A., Bell, G. I. & Sakaguchi, A. Y. (1988) Genomics a victim of cancer. We thank Thad Howard, Julie Rochelle, and Sean 3, 172-175. Todd for excellent technical assistance; Linda Howell for secretarial 27. Sundaresan, S. & Francke, U. (1989) Somat. Cell Mol. Genet. assistance; Ling-Mei Wang and Vic Sylvia for photography; Ellen 15, 367-371. Kraig and Nancy Smith for use of equipment; Shirley Tilghman for 28. Moseley, W. S. & Seldin, M. F. (1989) Genomics 5, 899-905. the Afp probe; Roger Fleischman for sending the Kit probe; and 29. Seldin, M. F., D'Hoostelaere, L. A., Huppi, K., Mock, B. A., David Housman for permission to use it. A.Y.S. is supported by the Steinberg, A. D., Parnes, J. D. & Morse, H. C., III (1988) National Institute on Aging (PO1-AG06872), the American Cancer Immunogenetics 27, 396-398. Society (CD358A), and The Meadows Foundation. M.F.S. is sup- 30. Seldin, M. F., Howard, T. A. & D'Eustachio, P. (1989) Ge- ported by National Institutes of Health Grant HGO0101. J.B. is nomics 5, 24-28. supported by National Institutes of Health Grants DK27726 and 31. Saunders, A. M. & Seldin, M. F. (1990) Genomics 6, 324-332. HL29305. 32. Seldin, M. F., Martinez, L., Howard, T. A., Naylor, S. L. & Sakaguchi, A. Y. (1990) Cytogenet. Cell Genet. 54, 68-70. 1. Searle, A. G. & Truslove, G. M. (1970) Genet. Res. (Cam- 33. Claesson-Welsh, L., Eriksson, A., Westermark, B. & Heldin, bridge) 15, 227-235. C. H. (1989) Proc. Natl. Acad. Sci. USA 86, 4917-4921. 2. Russell, E. S. (1979) Adv. Genet. 20, 357-459. 34. Roberts, W. M., Look, A. T., Roussel, M. F. & Sherr, C. J. 3. Silvers, W. K. (1979) The Coat Colors of Mice: A Model for (1988) Cell 55, 655-661. Gene Action and Interaction (Springer, New York), pp. 206- 35. Mercola, M., Wang, C., Kelly, J., Brownlee, C., Jackson- 241. Grusby, L., Stiles, C. & Bowen-Pope, D. (1990) Dev. Biol. 138, 4. Geissler, E. N., McFarland, E. C. & Russell, E. S. (1981) 114-122. Genetics 97, 337-361. 36. Pringle, J., Collarini, E. J., Mosley, B., Heldin, C. H., Wes- 5. Geissler, E. N., Ryan, M. A. & Housman, D. E. (1988) Cell 55, termark, B. & Richardson, W. D. (1989) EMBO J. 9, 1049- 185-192. 1056. 6. Chabot, B., Stephenson, D. A., Chapmen, V. M., Besmer, P. 37. Raff, M. C., Lillien, L. E., Richardson, W. D., Burne, J. F. & & Bernstein, A. (1988) Nature (London) 335, 88-89. Noble, M. D. (1988) Nature (London) 333, 562-565. 7. Williams, D. E., Eisenman, J., Baird, A., Rauch, C., Van Ness, 38. Richardson, W. D., Pringle, N., Moseley, M. J., Westermark, K., March, C. J., Park, L. S., Martin, U., Mochizuki, D. Y., B. & Dubois-Dalcq, M. (1988) Cell 53, 309-319. Boswell, H. S., Burgess, G. S., Cosman, D. & Lyman, S. D. 39. Morrison-Graham, K. & Weston, J. A. (1989) Trends Genet. 5, (1990) Cell 63, 167-174. 116-121. 8. Zsebo, K. M., Williams, D. A., Geissler, E. N., Broudy, 40. Kadowaki, T., Bevins, C. L., Cama, A., Ojamaa, K., Marcus- V. C., Martin, F. H., Atkins, H. L., Hsu, R.-Y., Birkett, Samuels, B., Kadowaki, L., Beitz, L., McKeon, C. & Taylor, N. C., Okino, K. H., Murdock, D. C., Jacobsen, F. W., Lan- S. I. (1988) Science 240, 787-790.