Mouse Platelet-Derived Growth Factor Receptor a Gene Is Deleted in WJ9H and Patch Mutations on Chromosome 5 E
Total Page:16
File Type:pdf, Size:1020Kb
Proc. Natl. Acad. Sci. USA Vol. 88, pp. 4811-4815, June 1991 Genetics Mouse platelet-derived growth factor receptor a gene is deleted in WJ9H and patch mutations on chromosome 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 Cancer Institute, Bethesda, MD 20892; and IThe Jackson Laboratory, Bar Harbor, ME 04609 Communicated by Elizabeth S. Russell, February 26, 1991 ABSTRACT The mouse W19H mutation is an x-ray- nonviable, with mice dying before or shortly after birth. induced deletion 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 growth factor receptor a gene (PDGFRA) like tyrosine kinase linked to a developmental abnormality in Kit encodes a transmembrane receptor tyrosine kinase. 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 chromosome 5 in tight linkage with Kit. pigmentation, are linked to receptor tyrosine kinases or to Hybridization of a PDGFRA probe to DNAs from WI9H/+ growth factors. Recently, the human KIT and PDGFRA patch heterozygous mice, and their genes 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 locus, 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 chromosomes, 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 anemia 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 ligand encoded by the steel locus (e.g., refs. 7 and fusion of Chinese hamster lung 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 protein (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 enzymes 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 amino acid 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.