Proc. Natl. Acad. Sci. USA Vol. 92, pp. 12352-12356, December 1995

Isolation and characterization of a pseudoautosomal region-specific genetic marker in C57BL/6 mice using genomic representational difference analysis (X-Y ) IVETA D. KALCHEVA*, YOICHI MATSUDAt, CHRISTOPH PLASS*, AND VERNE M. CHAPMAN** *Department of Molecular and Cellular Biology, Roswell Park Cancer Institute, Elm and Carlton Streets, Buffalo, NY 14263; and tNational Institute of Radiological Sciences, 9-1, Anagawa-4-chome Inage-ku, Chiba-shi 263, Japan Communicated by Stanley M. Gartler, University of Washington, Seattle, WA, August 14, 1995 (received for review April 11, 1995)

ABSTRACT Representational difference analysis was recombination in male meiosis, this region provides an ideal used to identify strain-specific differences in the pseudoau- opportunity to study chromosome pairing, crossing-over, and tosomal region (PAR) of mouse X and Y . One interference in a higher eukaryotic chromosome. second generation (C57BL/6 x Mus spretus) x Mus spretus In addition to these central issues in genetics and meiosis, an interspecific backcross male carrying the C57BL/6 (B6) PAR understanding of the physical structure of the PAR in mam- was used for tester DNA. DNA from five backcross males from mals is essential for understanding the molecular basis for the the same generation that were M. spretus-type for the PAR was failure of X and pairing that has been docu- pooled for the driver. A cloned probe designated B6-38 was mented in sterile, interspecific hybrid males (6-8). The failure recovered that is B6-specific in Southern analysis. Analysis of to pair is primarily a feature of the X and Y chromosomes in genomic DNA from several inbred strains of laboratory mice the laboratory mouse x Mus spretus hybrids in these studies, and diverse Mus and subspecies identified a charac- but an elevated level of autosomal dissociation is also observed teristic Pst I pattern of fragment sizes that is present only in (6, 7). However, the subsequent development of C57BL-M. the C57BL family of strains. Hybridization was observed with spretus congenic mice for the PAR has shown that the X-Y in and to a limited extent with Mus dissociation and male sterility remain after 12 backcross sequences DBA/2J generations while the rest of the is functionally musculus (PWK strain) and Mus castaneus DNA. No hybrid- C57BL and shows little autosomal dissociation (Y.M. and ization was observed in DNA of different Mus species, M. V.M.C., unpublished results). The occurrence of male sterility spretus, M. hortulanus, and M. caroli. Genetic analyses of B6-38 associated with a failure of X and Y chromosomes to pair was conducted using C57BL congenic males that carry M. during meiotic prophase provides a consistent mechanism that spretus alleles for distal X chromosome loci and the PAR and accounts for Haldane's rule: "When in the F1 offspring of two outcrosses of heterozygous congenic females with M. spretus. different animal races one sex is absent, rare, or sterile, that sex These analyses demonstrated that the B6-38 sequences were is the heterozygous sex" (9). inherited with both the X and Y chromosome. B6-38 sequences The X chromosome is highly conserved among eutherian were genetically mapped as a within the PAR using two mammals. However, there is little or no conservation of the interspecific backcrosses. The locus defined by B6-38 is des- that map within the PAR. For example, the CSF2RA ignated DXYRpl. Preliminary analyses of recombination be- locus of the PAR maps to the distal end of chromosome tween the distal X chromosome amelogenin (Amg) and 19 in the mouse and to additional chromosomes in M. spretus the PAR loci for either TelXY or sex chromosome association (10). In addition, none of the other characterized human PAR (Sxa) suggest that the locus DXYRpl maps to the distal portion genes is localized to the mouse PAR. Similarly, the steroid of the PAR. sulfatase (Sts) gene in the mouse maps to the PAR, whereas in it maps to distal Xp outside of the PAR (11, 12). The mouse X and Y chromosomes pair and recombine during This suggests that there is a divergence between the PAR of meiosis in a short region of the distal telomeric ends of each mouse and human and it is necessary to characterize the of the chromosomes. This region has been termed the pseudo- physical and genetic makeup of the PAR of each mammalian autosomal region (PAR) because of the autosomal-like mode species as a unique entity. At present, there are relatively few of segregation and inheritance of the loci from this part of the loci that have been defined and mapped in the PAR of the The PAR has been mapped and char- laboratory mouse and its interbreeding species. These include (1). physically May15 (3, 13, 14) and a telomere-related fragment identified acterized in the human genome, and several studies have in the C57BL strains (15). We have used the genomic repre- demonstrated that the region is -2.7 Mb and contains a sentational difference analysis (RDA) to recover C57BL- limited number of genes that are expressed on both the X and specific sequences from the PAR against the M. spretus Y chromosome and that typically escape X chromosome genomic background (16). This paper demonstrates that the inactivation in females (2). Genetic analysis of the PAR in both cloned probe for one of these sequences identifies species mice and humans indicates that there is an obligate crossing- variation that is inherited as a locus on both the X and the Y over that occurs in this region in males such that the estimated chromosomes, and preliminary results indicate where it is is 50 centimorgans (3-5). By contrast, the localized in the female genetic map. relative frequency of recombination within the PAR during female meiosis is similar to other regions on the X chromo- some (3). As a consequence, the relative genetic maps of PAR MATERIALS AND METHODS are about 7- to 10-fold greater in male meiosis as compared Mouse Strains and Genetic Crosses. The inbred mouse with female meiosis. Given the relatively high frequency of strains C57BL/6Ros (B6), C57BL/1OJ (BlO), and DBA/2J

The publication costs of this article were defrayed in part by page charge Abbreviations: PAR, pseudoautosomal region; RDA, representa- payment. This article must therefore be hereby marked "advertisement" in tional difference analysis. accordance with 18 U.S.C. §1734 solely to indicate this fact. 4To whom reprint requests should be addressed.

Downloaded by guest on September 30, 2021 12352 Genetics: Kalcheva et al. Proc. Natl. Acad. Sci. USA 92 (1995) 12353 (D2), which are representative of Mus domesticus, and PWK/ RESULTS Ros, which is representative of Mus musculus, were used. In Isolation of a C57BL/6-Specific Genetic Marker by Using addition, individuals from outbred colonies of M. spretus, Mus the RDA Method. Genomic DNA from a single second gen- castaneus, Mus hortulanus, Mus macedonicus, and Mus caroli eration BSS male that carried the C57BL/6 PAR (AmgB-SxaB) maintained at Roswell Park Cancer Institute were used in on its X chromosome and the M. spretus PAR on its Y these studies. A congenic C57BL strain (BSB) was produced chromosome was used as the RDA tester. DNA from five that carries the distal end of the M. spretus X chromosome and sibling males from the same backcross generation that carried the PAR on a C57BL background. The BSB strain was the M. spretus PAR on their X chromosomes and the M. spretus produced by selecting backcross females that were heterozy- PAR on the Y chromosome was pooled for the RDA driver. gous for the distal X gene amelogenin (Amg) and the distal The SxaB phenotype is characterized by a small testis weight telomere locus TelXY The selected females were mated with and a high frequency of X-Y dissociation that results from the either B6 or B10 in alternate generations using substrain allelic heterozygous combination of B6 X and M. spretus Y chromo- differences in TeiXY that distinguish the C57BL X chromo- somes (6). All of the males used for driver DNA wereAmgs and some PAR from the M. spretus X chromosome PAR. Congenic expressed the Sxas phenotype (all of these males had normal BSB females (generation N10-11) heterozygous for AmgB/S or large testis weight and low frequencies of X-Y dissociation). and TelXY B/S were mated with M. spretus to produce BSB x Three rounds of RDA subtraction and PCR amplification of S male progeny that were used to localize the B6-38 sequences the adaptor-ligated tester sequences were performed. The to the PAR of C57BL. The interspecific backcrosses from resulting PCR products were analyzed on agarose gels after Roswell Park Cancer Institute (BSS-RP; n = 59) (6) and from each round of RDA. After the third round of RDA, several The Jackson Laboratory (BSS-JAX) (17) used in this study distinctive bands from the subtracted tester DNA appeared on were produced by mating (B6 x M spretus) x M. spretus. a background smear upon ethidium bromide staining of the Probes. TeIXY probe was an oligonucleotide (TTATTG)5 agarose gel (data not shown). This DNA, ranging from 150 to containing the mammalian telomere consensus sequence (15). 400 bp, was digested with BamHI, gel purified, and cloned into Amg was characterized using the 170-bp Nco I-Xba I subfrag- the BamHI site of pBluescript II KS(+). The inserts from 15 ment excised from pMal6 (pMal6 is a pGEM-1 clone con- clones were independently used as hybridization probes on taining a 753-bp Pst I-Xba I fragment from the mouse Amg Southern blots containing C57BL/6 and M. spretus genomic gene). This probe identifies an EcoRI restriction site variation DNA digested with different restriction endonucleases. This of 6.2 kb (B6) and 3.1 kb (M. spretus) (18). analysis was performed to identify clones that detected dif- Recovery of Genomic Clones Using the RDA Method. ferences between C57BL/6 and M. spretus DNA. Many of the Second generation (B6 x M. spretus) x M. spretus (BSS) clones identified sequences homologous to long interspersed interspecific backcross (BC) male mice were used for the RDA repetitive DNA element 1 and satellite DNA that were present analyses (16). These were derived from BC1 females that were in both C57BL/6 and M. spretus genomic DNA. One of the heterozygous for the distal X chromosome region. These clones, designated B6-38, detected sequences only in the tester progeny carry an average of 12.5% of the B6 genome. Genomic and C57BL/6 DNA but not in the driver and M. spretus DNA. DNA from a single male that was B6 for the distal X We used the B6-38 clone as a probe in Southern analysis of chromosome was selected as the RDA tester. DNA from five genomic DNA from several inbred strains of laboratory mice males that were uniformly M. spretus for the same region was and several Mus species and subspecies to determine whether pooled to prepare the RDA driver. Five micrograms of DNA it would identify cross-reacting sequences across the Mus gene of both tester and driver was digested with BamHI. One pool and whether there was informative restriction fragment microgram of each digested sample was ligated to the R Bam length variation that could be used for genetic mapping. adaptor (24-mer, 5'-AGCACTCTCCAGCCTCTCACCGAG- Cross-hybridization of B6-38 was detected with sequences in tester D2, PWK, and M. castaneus, but no hybridization was observed 3'; 12-mer, 5'-GATCCTCGGTGA-3') to prepare the spretus, M. hortulanus, and and driver amplicon representations. Three rounds of RDA with DNA from the Mus species M. were performed. M. caroli. Strain variation between B6 and B10 DNA was subtractive hybridization/PCR amplification observed in the relative intensity and sizes of the Pst I The adaptorJBam (24-mer,5'-ACCGACGTCGACTATGCA- fragments detected by B6-38 (Fig. 1). Two Pst I fragments, 2.8 TGAACG-3'; 12-mer, 5'-GATCCGTTCATG-3') was ligated and 2.3 kb long, were detected in DNA from both male and to the tester DNA in rounds one and three, and the adaptor N female B6 mice. These 2.8- and 2.3-kb fragments and an Bam (24-mer, 5'-AGGCAACTGTGCTATCCGAGGGAG- additional one of 2.0 kb were detected in DNA from B10 3'; 12-mer, 5'-GATCCTCCCTCG-3') was used in round two. males, whereas only the 2.3- and 2.0-kb fragments were present After the third round of RDA, the resulting DNA was digested in DNA from B10 females. A pronounced difference in with BamHI to cleave away the adaptors, purified from an fragment lengths detected with B6-38 was observed between agarose gel, and ligated into the BamHI site of pBluescript II D2 and the C57BL family. Hybridization of B6-38 with KS(+) vector. Approximately 1000 clones were plated, and 50 BamHI-digested DNA from strains of the C57BL family clones were randomly selected for plasmid DNA isolation. The revealed the expected BamHI fragment (220 bp) predicted inserts were excised with BamHI, purified from an agarose gel, from the RDA. and used for further analysis. Localization of B6-38 Sequences to the PAR. The backcross Mouse Genomic DNA Preparation and Southern Blot Anal- progeny used to isolate genomic DNA for the RDA driver and ysis. High molecular weight DNA was isolated from mouse tester differed primarily for the distal X chromosome geno- kidney (19). Five to 10 ,ug of each DNA sample was digested types that would encompass the PAR. However, it was still with the corresponding restriction endonuclease under condi- possible that the locus detected by the B6-38 probe was located tions recommended by the manufacturer (GIBCO/BRL and to an autosomal region due to a not covered in Promega). The digested DNA samples were electrophoresed the driver DNA or to the X chromosome outside of the PAR. in 1% agarose gels and transferred onto a Zetabind membrane To test whether the clone identified an X and Y chromosome- (AMF Cuno). Hybridizations to the RDA subtracted clones specific sequence from B6, we examined the patterns of B6-38 and further washes were performed under high-stringency hybridization with a series of males from a C57BL congenic conditions (20). Probes were prepared by random primer strain whose female parents were selected to be heterozygous labeling (21), and -2 x 106 cpm/ml was used in a hybridization C57BL (B) x M. spretus (S) for the distal portion of the X reaction. chromosome and whose male parents were either B6 or B10. Downloaded by guest on September 30, 2021 12354 Genetics: Kalcheva et al. Proc. Natl. Acad. Sci. USA 92 (1995) chromosome, we examined the hybridization patterns in males that carried a C57BL PAR on their X chromosomes and the M. spretus PAR on their Y chromosomes by outcrossing the heterozygous, congenic BSB strain females with M. spretus males. In this case, the male progeny were either AmgB- TelXB6/TelYS orAmgs-TelXs/TelYs. A positive hybridization m pattern for B6-38 was observed in males that were AmgB_ r- 4r, TelXY B/S (PARB), but no hybridization signal was observed in males that were Amgs-TelXYs (PARS) (Fig. 3). These results kh clearly demonstrated that B6-38 sequences were transmitted with the distal portion of the C57BL X chromosome since the 21.1 4M proximal loci of both of the X chromosomes in the BSB females carry B6 alleles. Thus, these combined results are consistent with a Y and X chromosomal transmission of B6-38 sequences in the PAR. Genetic Mapping of B6-38. We examined two sets of inter- specific backcrosses to verify the segregation of B6-38 se- quences as a single locus and to map this locus relative to the 2. - _3 2. 4 distal X chromosome gene Amg and the distal loci of Sxa and TelXY i'n the mouse PAR. The first interspecific backcross was the BSt-RP (6), in which 3 of 59 males were recombinant between Amg and the X-Y pairing locus identified as Sxa. Southern analysis of the B6-38 sequences in these 3 recombi- nant males indicated that two had no cross-reacting sequences FIG. 1. Southern analysis of B6-38 with Pst I-digested genomic similar to M. spretus, while the third male showed a typical B6 DNA from inbred strains C57BL/6, C57BL/10, DBA/2J, and PWK pattern of fragments (Table 1). These results indicate that the (M. musculus) and outbred samples of M. castaneus, M. hortulanus, M. B6-38 sequences segregated as a single locus, which we named spretus, an F1 hybrid of C57BL/6 x M. macedonicus, and M. caroli. DXY Roswell Park 1 (DXYRpl). DXYRpl mapped more closely to Sxa than to Amg. The male progeny were previously analyzed as AmgB-TelXB6/ We examined the segregation and recombination of TelyBI and Amgs-TelXs/Tely B1 or AmgS-TelXS/TelY B6. DXYRpl in an independent, interspecific BSS-JAX cross that The hybridization patterns of B6-38 with DNA from the was segregating distal X chromosome markers. We used Amgs_TelXs/TelY Bi or AmgS-TelXs/TelYB6 progeny com- Southern analysis to determine the occurrence of B6-38 hy- pletely reflect the B6 or BlO allelic type of the male parent. bridization and the pattern of these sequences in the segre- These results strongly indicated that the B6-38 probe detected gating progeny. We observed 1 of 89 recombinants between sequences that were inherited on the Y chromosome. Auto- Amg and DXYRpl in this cross. As in our previous analysis, the somal linkage for B6-38 sequences is excluded since none of B6-38 sequences segregated as a single locus that mapped the progeny tested showed nonpaternal type patterns for B6-38 distal to Amg (Table 2). (Fig. 2). A positive hybridization for B6-38 was observed in Southern DISCUSSION analyses of Pst I-digested genomic DNA from AmgB-TelXB6/ Representational difference analysis was used to recover a TelyB10 a pattern of B6 and males showing combined B10 B6-specific probe that was localized to the mouse PAR. We fragments (data not shown). These results indicated that B6-38 were able to apply RDA to second generation BSS backcross sequence was transmitted on both the X and Y chromosomes. males that carried the distal X chromosome region from B6 To verify that B6-38 sequences are transmitted on the X in the case of the tester DNA and sibling males that were M. spretus-like for distal X chromosome markers in the case of kb 5 h - X ') a pooled source of driver DNA. A clone B6-38 was recovered that hybridized to B6 but not to M. spretus genomic DNA. Subsequent analysis of this clone in a variety of inbred strains 2.8 and divergent Mus species indicated that this probe strongly w -.x hybridized with sequences in the C57BL family and that it U's. produced divergent patterns in different strains and a di- minishing cross-hybridization in more distantly related X (hron1losonLc house mouse species M. musculus (strain PWK) and M. s s s s castaneus. No cross-hybridizing sequences were observed in the more related Mus of M. M. Tel X s s s distantly species spretus, hortulanus, and M. caroli. The sequences detected by B6-38 (Chromiosomne segregated as a single locus, DXYRpl, that was mapped to the distal part of the X chromosome in B6. Genetic variation Tel 13B 1 It' 13 B '' between B6 and B10 was used to confirm that DXYRpl can DXN'YRp B3 } IS 13 B' be inherited from both the X and Y chromosomes consistent with a PAR localization. This location was confirmed in FIG. 2. Southern hybridization of B6-38 with Pst I-digested segregating backcrosses from BSS, which indicated recom- genomic DNA from C57BL congenic male progeny that received the bination between Amg-one of the most distal X chromo- Amgs-TelXs X chromosome from heterozygous mothers and either some loci-and An X chromosome localization in the TeIYB6 (lanes 2-4) or TelYBlO (lanes 5-7) Y chromosome from DXYRpl. their fathers. The M. spretus allele of the segregating gene amelogenin the other B6-38-positive inbred strains was not verified in (Amgs) was separately ascertained for these progeny. Lanes 1, 8, and these studies. 9 are control DNAs from B6 male, B10 male, and BlO female, The variation observed for DXYRpJ between two closely respectively. related C57BL strains, B6 and B10, and the marked restriction Downloaded by guest on September 30, 2021 Genetics: Kalcheva et al. Proc. Natl. Acad. Sci. USA 92 (1995) 12355

1 2 kb 3 4 5 6 7 8 9 10

am - 6.2- i - * _

Amg

- 3.1 -

- 2.6 -

DXYYRp - 0.2 -

FIG. 3. Segregation of the X chromosome-linked B6-38 sequences among male progeny from BSB congenic mothers, heterozygous for AmgB-TelXB6/AmgS-TelX s, which were mated with M. spretus males. Amgs or AnigB alleles were identified in EcoRI-digested DNA, and B6-38 sequences were identified in BamHI-digested DNA. Lanes 1 and 2, C57BL/6 and M. spretus, respectively; lanes 3-6, AmgB-Te1XB6/TelY S males; lanes 7-10, Amgs-TelXs/TelY S males. The hybridization pattern with the TelXY probe was determined in a separate experiment and it is not shown here. variation between these strains and DBA/2J is consistent with not yet available for the mouse. However, if we use the size of a rapid evolutionary change for sequences in the PAR com- the human PAR as an approximation of the mouse PAR, we pared with other portions of the X chromosome. Moreover, can estimate that the PAR constitutes about 0.1% of the total the lack of cross-reaction between B6-38 and DNA from genome in single copy. If the PAR represents a highly diverged distantly related Mus species M. spretus, M. hortulanus, and M. genomic region between tester and driver , it is possible caroli is consistent with earlier suggestions that the sterility of that we can expect to find other strain-specific sequence differ- interspecific hybrid males is due in large measure to the lack ences in additional screening of the subtracted sample. of X-Y pairing during meiosis and may be a result of sequence The PAR localization of B6-38 as the DXYRpJ locus pro- divergence in the PAR between species. The identification of vides an important molecular entry into the physical analysis a single RDA-derived sequence that is diverged between M. of the mouse PAR. Moreover, the allelic difference between spretus and B6 does not establish a sufficiently large sampling B6 and BlO, as well as other inbred strains, provides an of PAR sequences to determine the relative similarity or important opportunity to genetically map this locus in male divergence of sequences in the PAR. The rate of success in our meiosis to better characterize the issues of genetic mapping initial screen of 15 clones is comparable to that from a similar and interference in this of the type of RDA application that highly recombinogenic region (22), suggesting the RDA genome (2, 3). method can be applied to congenic strains for the identifica- tion of sequence differences. In our experiments we used a relatively large range of fragment sizes for the cloning step, This work is dedicated to the memory of Verne M. Chapman. We thank Diane Poslinski and Darla Miller for their extending from 150 to 400 bp. This extended range of fragment technical assistance. sizes may have contributed to the high frequency of recovering clones that identified highly abundant sequence families that Table 2. Genetic analysis of DXYRpl variation in the BSS-JAX were present in both M. spretus and B6. A more limited size cross showing the segregation of maternal (Mat) and paternal (Pat) selection of the amplified RDA sequences may improve the types in progeny relative rates of recovering additional B6-specific fragments. Loci Mat Pat x n x/n SE Physical analysis of the PAR region of the human genome DXMit2O 43 51 indicates that it is about 2.3 Mb (2). Similar determinations are 2 93 2.15 1.50 Table 1. Segregation of B6-38 (DXYRpl) sequences in (B6 x M. DXMitlOO 44 49 spretus)Fj x M. spretius (BSS-RP) backcross males that were X 0 93 0.00 0.00 analyzed for amelogenin (Amg) and had an X-Y pairing phenotype DXMitl60 45 49 identified as Sxa 0 94 0.00 0.00 Amg 45 49 Amg DXYRpI Sxa n 0 94 0.00 0.00 B B B 17 DXMit3O 45 49 B x S S 2 1 89 1.12 1.12 S x B B 1 DXYRpl 42 47 S S S 39 x is the number of recombinants between adjacent loci, n is the Frequency of recombination:Amg-DXYRpl = 3/59 = 0.051 (5.1%); number of progeny typed, x/n is the percentage of recombination, and DXYRpl-Sxa = 0/59. n, Number of progeny typed. Total n = 59. SE is the standard error of the mean. Downloaded by guest on September 30, 2021 12356 Genetics: Kalcheva et al. Proc. Natl. Acad. Sci. USA 92 (1995)

This work was supported in part by U.S. Public Health Services Grant 11. Keitges, E. & Gartler, S. M. (1986)Am. J. Hum. Genet. 39,470-476. GM33160 to V.M.C. 12. Shapiro, L. J. (1985) Adv. Hum. Genet. 14, 331-381. 13. Harbers, K., Soriano, P., Muller, U. & Jaenisch, R. (1986) Nature 1. Burgoyne, P. S. (1982) Hum. Genet. 61, 85-90. (London) 324, 682-685. 2. Rappold, G. A. (1993) Hum. Genet. 92, 315-324. 14. Harbers, K., Francke, U., Soriano, P., Jaenisch, R. & Muller, U. 3. Soriano, P., Keitges, E. A., Schorderet, D. F., Harbers, K., (1990) Cytogenet. Cell Genet. 53, 129-133. Gartler, S. M. & Jaenisch, R. (1987) Proc. Natl. Acad. Sci. USA 15. Eicher, E. M., Lee, B. K., Washburn, L. L., Hale, D. W. & King, 84, 7218-7220. T. R. (1992) Proc. Natl. Acad. Sci. USA 89, 2160-2164. 4. Henke, A., Fischer, C. & Rappold, G. A. (1993) Genomics 18, 16. Lisitsyn, N., Lisitsyn, N. & Wigler, M. (1993) Science 259, 946-951. 478-485. 17. Rowe, L. B., Nadeau, J. H., Turner, R., Frankel, W. N., Letts, V., 5. Rouyer, F., Simmler, M.-C., Johnsson, C., Vergnaud, G., Cooke, Eppig, J. M. S. H. J. & Weissenbach, J. (1986) Nature (London) 319, 291-298. T., Ko, H., Thurston, S. J. & Birkenmeier, E. 6. Matsuda, Y., Hirobe, T. & Chapman, V. (1991) Proc. Natl. Acad. (1994) Mamm. Genome 5, 253-274. Sci. USA 88, 4850-4854. 18. Chapman, V. M., Keitz, B. T., Disteche, C. M., Lau, E. C. & 7. Matsuda, Y., Moens, P. B. & Chapman, V. M. (1992) Chromo- Snead, M. L. (1991) Genomics 10, 23-28. soma 101, 483-492. 19. Blin, N. & Stanford, D. W. (1976) NucleicAcids Res. 3, 2303-2308. 8. Hale, D. W., Washburn, L. L. & Eicher, E. M. (1993) Cytogenet. 20. Church, G. M. & Gilbert, W. (1984) Proc. Natl. Acad. Sci. USA Cell Genet. 63, 221-234. 81, 1991-1995. 9. Haldane, J. B. S. (1922) J. Genet. 12, 101-109. 21. Feinberg, A. P. & Vogelstein, B. (1983) Anal. Biochem. 132, 6-13. 10. Disteche, C. M., Brannan, C. I., Larsen, A., Adler, D. A., Schor- 22. Lisitsyn, N. A., Segre, J. A., Kusumi, K., Lisitsyn, N. M., Nadeau, deret, D. F., Gearing, D., Copeland, N. G., Jenkins, N. A. & Park, J. H., Frankel, W. N., Wigler, M. H. & Lander, E. S. (1994) Nat. L. S. (1992) Nat. Genet. 1, 333-336. Genet. 6, 57-63. Downloaded by guest on September 30, 2021