GENOMICS

12,26-34

(1992)

A Structural Analysis of the Sxr Region of the Mouse Y Chromosome MICHAEL

J. MITCHELL* AND COLIN E. BISHOP*st

*Molecular

Genetics Research Laboratory, Department of Ob/Gyn, Division of Reproductive Genetics, University of Tennessee, 77 1 Jefferson Avenue, Memphis, Tennessee 38105; and t Unit& d’lmmunog&&ique Humaine, INSERM U276, lnstitut Pasteur, Paris IS, France ReceivedJune

tigen (H-Y), mapping Hyu, the gene controlling H-Y expression, into the Sxr region (Simpson et al., 1984). An Sxr variant termed Sxf that, although still capable of directing male determination, lacks H-Y expression was reported (McLaren et al., 1984). The original Sxr mutation has been renamed Sxr” and the Sxr’ mutation has been renamed Sxrb (McLaren et al., 1988). A comparative study of XSxr”0 and XSxrbO mice revealed that, while the former had near normal spermatogenesis, the latter had no spermatogenic cells beyond the spermatogonial stages (Burgoyne et at., 1986; Sutcliffe and Burgoyne, 1989). This mapped a factor involved in spermatogenesis (Spy) to the Sxr” region. It has been shown that the mutation that gave rise to the Sxrb region involved a deletion, as XXSxrb DNA is missing an Sxr locus defined by anonymous DNA probe pY291 (Roberts et al., 1988) and lacks part of the Sxr gene, Zfy-2 (Mardon et al., 1989). Thus, Hya and Spy can be mapped to the Sxr DNA deleted from the Sxr’ region, termed here the ASxr’ DNA. In this study all Sxr loci described can be said to be either Sxrb loci (present in the SxP and Sxrb regions) or ASxrb loci (present in the Sxr’l region but absent from the Sxrb region). To identify and clone Hyu and Spy it now becomes important to analyze the ASxrb region at the molecular level. With this aim in mind we have initiated chromosome walking in the ASxrb DNA and have begun to characterize this unique region. We present here a preliminary structural map of the region, which can be used as a basis for future cloning strategies. In addition, our data show that this region is evolving at a very rapid rate through a variety of complex mechanisms.

Three genetic functions have been mapped to the minute Sxr (sex-reversed) region of the mouse Y chromosome. These are Tdy, the primary testis determinant; Hya, the locus (either structural or regulatory) controlling the expression of the male-specific minor histocompatibility antigen H-Y; and Spy, a spermatogenic gene. Hya and Spy map to DNA deleted from the Sxr region in the deletion variant Sxfi (the ASxs DNA). With the object of cloning Hya and Spy, we initiated chromosome walking in the ASxrb DNA. From three independent loci-Sxl, Zf2, and T5-we have isolated approximately 270 kb of ASx4 DNA lying in three contigs of 145, 60, and 65 kb, respectively. Within 17 kb of the 3’ end of the Zfy-2 gene, lowcopy repeat elements were found in a region that extends for -35 kb. Probes isolated from this region detect multiple Sxr loci, some of which map to the ASxrb DNA present in the T5 contig DNA. Three of these multicopy probes detect ASxrb loci not represented in our three contigs, which means that six distinct ASxrb loci have now been identified. Here we present a preliminary model of the molecular structure of the DNA in this unique region. o 1992 Academic Press, Inc.

INTRODUCTION

The Sxr region of the mouse Y chromosome is defined as the portion of the Y short arm that generates the Sxr (sex-reversed) phenotype in XXSxr or XSxrO mice (Cattenach et al., 1971; Singh and Jones, 1982; Evans et al., 1982; McLaren et al., 1988; Roberts et al., 1988). The Sxr mutation is propagated through carrier males who carry the 5% region on the distal end of the Y pseudoautosomal region. During meiosis it is transferred to the X chromosome by pseudoautosoma1 recombination, causing sex reversal in XXSxr or XSxrO offspring. XXSxr and XSxrO mice develop as sterile males, mapping the testis-determining gene Tdy into the Sxr region. In addition, these mice express the male-specific minor histocompatibility an0888-7543/92 $3.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form

MATERIALS

AND

METHODS

Library Construction Chromosome walking was undertaken using an XYSxr” cosmid library in pWE15 (Wahl et al., 1987) 26

Inc. reserved.

17, 1991

STRUCTURE

OF

Srr

REGION

and an XXSxr” phage library in hDash (Stratagene). High-molecular-weight DNA was prepared from the XYSxr” carrier male and a liver of a C57BL/6 C57BL/6 XX&-” male (kindly supplied by Dr. A. McLaren) using standard techniques. For cosmids the DNA was partially digested with Mb01 and size fractionated on a 0.3% agarose gel. The 35 to 50-kb fraction was excised from the gel and pushed through a 25-gauge needle. After mixing with an equal volume of buffer-saturated phenol, the DNA was isolated from the agarose by freeze-fracture in liquid nitrogen. The DNA was then ligated to BarnHI-restricted, pWE15 vector arms (dephosphorylated at their noncloning end), packaged in. vitro using Gigapack gold (Stratagene), and used to transfect Escherichia coli NM554. Approximately 2 X lo6 primary colonies were directly plated onto Hybond-N filters (Amersham). Replica filters were made and the original filters were stored as a “glycerol sandwich” at -80°C. A similar procedure was followed for the construction of phage libraries except that DNA between 18 and 25 kb was ligated into hDash arms, packaged, and plated on E. 2 X lo6 coli strain Sure (Stratagene). Approximately primary plaques were screened. Chromosome

Walking

All probes used to screen the libraries were first hybridized to Southern blots of, at least, EcoRI-digested DNA from a C57BL/6 male and female, a male, and a C57BL/6 XXSxrb C57BL/6 XXSxr” male according to our previously published procedures (Roberts et al., 1988; Mitchell et al., 1989). This revealed the size of the EcoRI fragment to which a probe should hybridize in a cosmid or phage from the same region. Positive recombinants were checked for the correct size fragment and those that carried it, or a shorter fragment (the Mb01 site in the insert can lie in the middle of an EcoRI restriction fragment), were further characterized. Recombinants were mapped with four restriction enzymes BamHI, EcoRI, SacI, and XbaI, as in Wahl et al. (1987) except that the products were indirectly end labeled using 32P-labeled T3 and T7 primers, following gel electrophoresis and Southern transfer. The hybridization of the primers was carried out in Church’s buffer (Church and Gilbert, 1984) at 25°C overnight and the filters were washed at 42°C in 5X SSC, 0.1% SDS. The maps of the cosmid recombinants generated in this way were most accurate up to 12-13 kb from the ends of the insert. For the central lo-18 kb only an approximate order of sites was achieved. The phage recombinants could be accurately mapped throughout the length of the insert. After mapping, a restriction fragment near the end of the insert was isolated either directly from the recombinant, by slicing the fragment from an agarose gel, or via the construction of a mini cosmid

OF

MOUSE Sxl

27

Y CHROMOSOME

Walk

cSx.2.6 csx.1.1 csx.1.10 csx.1.3

ISx.D

cSx.2.7 1 ___

2.9-

li

FIG. 1. Southern blot analysis of genomic DNA hybridized with probes derived from the Sxl contig. A, B, and C show the probes hybridized to (1) C57BL/6 male, (2) C57BL/6 female, (3) C57BL/6 XXSxr” male, and (4) C57BL/6 XXSxrb male DNAs. The probes (A) pSx.D.l.B2, (B) pY291, and (C)OOpSx.S.B.C were hybridized to EcoRI (A and C) or BamHI (B) digested DNA. The sizes of the hybridizing Sxr restriction fragments are indicated in kilobases. The Sxl locus was initially defined by the 10.5kb BamHI restriction fragment, detected by pY291, which is absent from the XXSrrb DNA but present in the XX&P DNA. At the top of the figure the Sxl contig is represented; its size is indicated beneath by a scale in kilobases. The positions of all probes isolated are indicated by the arrowed lines. Only pY291 detects multiple Sxr loci.

(Heilig et al., 1987). This fragment was screened for repeats with total male DNA and if necessary was further subcloned to obtain a single-copy probe. Prior to rescreening the library this probe was tested by Southern blot analysis of genomic DNA. This provided us with a check of its repeat-free nature and also revealed the size and nature of the fragments detected. The library was then screened and the procedure outlined above repeated. In this way we were unambiguously able to identify those cosmids or phage that truly overlapped at each step, even in regions where the probes detected multiple Sxr loci. All probes are shown washed at low stringency (1X SSC, 0.2% SDS, 65°C) with the exception of pSx1 and pT5.2.12.A2, which are shown washed at high stringency (0.1X SSC, 0.2% SDS, 65°C). RESULTS The Origins of the ASxrb Walks We have isolated 270 kb of ASxrb DNA from three

separate regions by chromosome walking. Our largest contig is from the locus termed Sxl and is 145 kb in length. The Sxl contig is diagramed in Fig. 1 (top).

28

MITCHELL

212 Walk

AND

/Zfv-2) hZf.F.6 XZf.F.2 CZf.2.1 cZf.1.8

Zfy-2

5’4

d =3 0 I

A.

kb

1

/ 2

3

20 I

I

B. 1

4

40 I

I

/ 2

3

C. 4

1

7

I

kb

/ 2

3

4

kb

2.1-m

#m

FIG. 2. Southern blot analysis of genomic DNA hybridized with probes derived from the Zt2 contig. A, B, and C show the probes hybridized to (1) C57BL/6 male, (2) C57BL/6 female, (3) C57BL/6 XX&r” male, and (4) C57BL/6 XXSxr* male DNAs. The probes (A) pZf.1.8.B2, (B) pZf.Z.l.D, and (C) pZf.F.8.A2 were hybridized to EcoRI-digested DNA. The sizes of the hybridizing Sxr restriction fragments are indicated in kilobases. At the top of the figure the Zf2 contig is represented; its size is indicated beneath by a scale in kilobases. The positions of pY8 (A) and pT5.1.1.PH (0) are indicated. The position of the zinc finger exon of Zfy-2 is represented by the rectangle and the orientation of the Zfy-2 gene (dotted line) is indicated.

This contig was initiated using pY291 (Roberts et al., 1988). This probe detects multiple loci within the Sxr region, one of which, defined by the 10.5-kb BumHI fragment, can be seen to be deleted from the Sxrb region (Fig. 1B). The exact extent of the ASxrb lowlevel repeat element defined by pY291 is not known but it does not extend beyond the region covered by &xl.1 and cSxl.3, as the probes pSxl.1.A and pSxl.3.A detect single-copy Sxr loci. The positions of these cosmids and subclones within the contig DNA are shown in Fig. 1 (top). The rest of the contig appears essentially locus specific, as the end probes pSx.2.6.C and pSx.D.l.B2 also detect single loci, as shown in Figs. 1A and lC, respectively. Both probes hybridize to single restriction fragments in male and XXSxr” DNA but fail to hybridize to female or XXSxrb DNA. Thus, they define a single ASxrb locus. By contrast, the other two contigs are composed in large part of DNA that is present in two or more copies within the Sxr region. One of these contigs (Zf2) is diagramed in Fig. 2 (top). It has been extended 60 kb 3’ from the zinc finger exon of Z&2, which maps to the ASxrb DNA (Mardon et al., 1989). As expected, the DNA around the zinc finger exon was found to be

BISHOP

duplicated, but within 17 kb 3’ of this exon the DNA had been further amplified within the Sxr region. The third contig, from the T5 locus, is diagramed in Fig. 3 (top) and originates from sequences detected by a 0.3-kb DNA probe, pT5.1.1.PH. This probe was subcloned from a 17-kb recombinant phage insert that was isolated from a C3H/HeJ/Pas male genomic Sxr loci library using pY8, which detects multiple (Bishop et al., 1987; Roberts et al., 1988). The subclone pT5.1.1.PH was isolated from a 5-kb EcoRI fragment and can be seen to detect several Sxr loci in Figs. 4A and 4B. The T5 contig was initiated from the locus defined by the 9-kb EcoRI fragment detected by pT5.1.1.PH. This fragment is present in the male and XXSxr” DNA but absent from female or XXSxrb DNA (Fig. 4B), which maps it to the ASxrb region. The T5 contig is 65 kb in extent. By hybridizing T5.1.1.PH to our Zf2 contig DNA, it was discovered that the 5-kb EcoRI fragment in Fig. 4B is a doublet in the male and XXSxr” tracks, one copy of which maps to Zf2 and is, therefore, absent from the XXSxrb track. The positions of pY8 (A) and pTS.l.l.PH (0) homologues are indicated in Figs. 2 and 3 (top). None T5

Zfy-1

5

A0

cT5.2.12

-3

.

AT5.G.9 )iT5.G.?

cT5.2.16 cT5.2.13

1

2

3

4

1

2

3

4

12

3

4

kb

kb 11 9.3 -

-ASxrb -sub

6.6-

-T5

4-

--Y

6.3-

FIG. 3. Southern blot analysis of genomic DNA hybridized with probes derived from the T5 contig. A, B, and C show the probes hybridized to (1) C57BL/6 male, (2) C57BL/6 female, (3) C57BL/6 XX&r” male, and (4) C57BL/6 XXSrr* male DNAs. The probes (A) pT5.2.13.D3, (B) pT5.2.12.A2, and (C) pT5.G.9.AZ were hybridized to EcoRI-digested DNA. The sizes of the hybridizing Srr restriction fragments, and the non&r Y fragment detected by pT6.2.13.D3, are indicated in kilobases. At the top of the figure the T6 contig is represented and its size is indicated beneath by a scale in kilobases. The positions of pY8 (A.) and pT5.1.1.PH (0) are indicated as are the approximate position and orientation of the 5’ end of the Zfy-1 gene. The assignment of the fragments of pT5.2.13.D3 are indicated to the right of A. The ll-kb (Szr*) and 9.3-kb (A&r*) fragments define two novel Sxr loci.

STRUCTURE 1

A.

2

3

4

0.

1

2

OF 3

Sxr

REGION

4

kb

- Y and ASxrb -T5 - 83 - 212 and

ZfB

- sxrb

FIG. 4. Southern blot analysis of genomic DNA hybridized with pT5.1.1.PH. A and B show the probes hybridized to (1) C57BL/6 male, (2) C57BL/6 female, (3) C57BL/6 XXSxr” male, and (4) C57BL/6 XXSrr’ male DNAs. The probe was hybridized to BgZII (A) and EcoRI (B)-digested DNA. The sizes of the hybridizing restriction fragments are indicated in kilobases. The loci to which the EcoRI restriction fragments map are indicated in B, where known. The “Y” indicates a non-&v Y fragment. The 16-kb EcoRI fragment has been deduced to define a non-&v Y locus and a ASxr* locus because the hybridization is more intense in the male track than in the XXSxr” track, whereas the opposite is the case for the other fragments. This suggests that the male has a copy of this fragment not present in XXSrr”. Additionally, a nonSW Y BglII restriction fragment of 11.5 kb can be seen in A. The 5 and 7-kb EcoRI fragments are of a similar intensity in the XXSrrb track but the 5 kb is more intense in the XXSxr” track. This is because pT5.1.1.PH detects a 5-kb fragment in the Zf2 and ZfB loci and so one copy is deleted from XXSd, causing the drop in intensity. The same can be seen for BglII, where the common Zf2/ ZfB fragment is 18 kb.

of the three contigs overlaps with any other or crosses either of the deletion breakpoints. DNA

Structure

of the Zf2 Locus

The DNA of the Zf2 contig is diagramed in Fig. 2 (top) and the characteristic hybridization patterns of the different probe types isolated are shown below. The DNA at the 3’ end of Zfy-2 is known to be duplicated, the other copy being at the locus defined by the Zfy-I gene (Mardon et aZ., 1989). This is demonstrated in Fig. 2A by pZf.1.8.B2, which lies 2 kb 5’ of the zinc finger exon. It detects two EcoRI restriction fragments from the Sxr region: one of 11 kb that is present in the Sxrb region and the other of 2.8 kb that maps into the ASxrb region, as it is present in the male and XX&r” DNA but absent from the female and XXSxrb DNA. However, further 3’ of Zfy-2 a region where the DNA cross-hybridizes with more than two homologous Sxr regions was encountered. This region can be split in two on the basis of the Southern blot hybridization patterns of probes subcloned from along the contig DNA. The first region is defined by the probes pY8 (A) and pT5.1.1.PH (0) (see Fig. 2, top), which are essentially Sxr specific although pT5.1.1.PH can be seen to detect a faint non-&r Y-specific 11.5kb

OF

MOUSE

Y CHROMOSOME

29

BglII fragment (Fig. 4A). At least four Sxr loci are detected by pY8 and six by pT5.1.1.PH. Two of the loci detected by pY8 and three of those detected by pT5.1.1.PH are deleted from the Sxr region in the Sxrb mutation. Roberts et al. (1988) have previously reported that pY8 does not detect any ASxrb loci on Southern blot analysis of genomic DNA. Here, using the more sensitive approach of hybridizing pY8 to Southern blots of cloned Sxr DNA, we have detected homologues at the ASxrb loci, Zf2 and T5. The previous failure to observe these ASxr’ homologues was because of divergence between the Zf2 homologue and pY8 and because the T5 homologue is defined by restriction fragments of the same size as one of the undeleted Sxrb loci detected by pY8. Similarly, Zf2 and T5 are two of the three deleted loci detected by pT5.1.1.PH. This probe is presented in more detail below. In the second region the probes also detect more than two Sxr loci but, in addition, they define many loci on the Y chromosome outside the Sxr region and on the X or autosomes. These probes are pZf2.1.B and pZf2.1.D and the hybridization pattern of pZf2.1.D is shown in Fig. 2B. The probe pZf.2.1.D is a 2.1-kb EcoRI restriction fragment derived from cZf.2.1 and, therefore, the hybridizing 2.1-kb EcoRI fragment in Fig. 2B derives from the Zf2 locus. It does not appear to be deleted from XXSxrb because this probe hybridizes to a 2.1-kb EcoRI fragment in a highly homologous nondeleted locus. The drop in intensity of the 2.1-kb fragment between the XXSxF and XXSxrb tracks however, can be seen to be greater than that for the other hybridizing fragments, suggesting that in the XXSxr” track the 2.1-kb fragment is a doublet, one copy of which is absent from the XXSxrb DNA. Between these two regions lies the probe pZf.1.8A, which can be seen to detect only two Sxr fragments, the Zf2 fragment and a nondeleted Sxrb fragment (data not shown). The possibility that this probe has homologues at other Sxr loci cannot be ruled out because of cross-hybridizing X or autosomal bands that may obscure faint Sxr bands. At the opposite end of the Zf2 contig from Zfy-2 there is a region that is unamplified as defined by the probe pZf.F.8.A2, which hybridizes to a single Sxr fragment absent from XXSxrb. In Fig. 2C it is shown hybridized to an 8-kb EcoRI ASxrb fragment. A faint X or autosomal fragment of slightly higher molecular weight can also be detected. DNA

Structure

of the T5 Locus

A diagram of the T5 contig is presented in Fig. 3 (top)andthepositionsofthepY8(~)andpT5.1.1.PH (0) homologies are indicated. Below are shown hybridization patterns characteristic of the different probe types isolated from the contig DNA. Southern

30

MITCHELL

analysis of the T5 contig reveals that, like Zf2, the locus consists in part of DNA that detects multiple Sxr loci. At one end the probe pT5.2.13.D3 together with pY8 and pT5.1.1.PH potentially defines a block of sequences detecting multiple Sxr loci. Figure 3A shows the hybridization pattern of pT5.2.13.D3 to EcoRI-digested DNA. The 2.4-kb ASxrb fragment of pT5.2.13.D3 maps to the Zf2 contig/locus and the 6.6kb ASxrb fragment is the cognate fragment mapping to T5. It also detects a 9.3-kb fragment present in both the SxP and the Sub regions, and a strong Y fragment at 4 kb. The 11-kb ASxrb fragment defines a new ASxrb locus, as it is not present in either the T5 or the Zf2 contig. The other end of the contig lies in a duplicated region as defined by the hybridization pattern of pT5.G.9.A2 (Fig. 3C), which detects a 14-kb fragment deleted from the Sub region and a nondeleted 9-kb fragment. However, between these two blocks lies pT5.2.12.A2, which detects only a single ASxrb locus defined by the 6.3-kb ASxrb fragment in Fig. 3B. Thus, in the T5 locus, regions containing sequences with homologues elsewhere in the Sxr region are separated by DNA that defines a single Sxr locus. The T5 Contig Maps to the 5’ End of Zfy-1 The finding that pT5.G.9.A2 (Fig. 3B) defines a duplicated segment of Sxr DNA prompted us to hybridize the T5 contig DNA with probes derived from the Zfy-2 full-length cDNA, pDP1122 (Mardon et al., 1989), and the results are shown in Fig. 5. The 5’ end of the Zfy-2 cDNA, pDP1122.B (a 320-bp EcoRINcoI fragment of pDP1122), was found to hybridize strongly to XT5.G.9 (Fig. 5ii), implying that the “XT5.G.9 end” of the T5 contig overlaps the 5’ end of one of the Zfy genes. As the NcoI site cuts at the ATG start codon of the gene, pDP1122.B essentially contains only the 5’ noncoding portion of the gene. The exact position of the smallest pDP1122.B +ve fragment, a 5.7-kb SacI/SalI end fragment, is indicated in Fig. 5i as is the origin of pT5.G.9.A2. The TS/Zfy overlap is further argued by the finding that pT5.G.9.A2, which originates from the same Sac1 fragment as the pDP1122.B homologue, detects a 14and a 9-kb band on Southern analysis of genomic DNA. These bands are identical in size to two of the three Sxr bands detected by pDP1122.B when hybridized to the same Southern blot (Fig. 5iii). We have recently characterized a yeast artificial chromosome (YAC) recombinant that contains both Zfy-1 and the T5 locus but not Zfy-2 (C. E. Bishop and Z. Larin, unpublished). This places the T5 contig at the 5’ end of Zfy-1. It also puts one of the Sxrb deletion breakpoints within the Zfy-1 gene or close to its 5’ end. The Organization of the Amplified DNA within the Sxr Region Using anonymous probe pT5.1.1.PH and DNA fragments 3’ of Zfy-2, we have defined a group of se-

AND

BISHOP

(i) EXB

\I

E

E

sxx

s

I

1

I

1

I

*---

0

1

I pDPl122.8

cT5.2.12

(ii)

%I

I pT5.G.9.A2HI

hT5.G.9

+“e

,2kb

ES

16-

(ii)

1234

1

2

3

4

9.4 -14 5.7 -

-

a~.

*

c)

FIG. 5. The end of the T5 contig overlaps the 5’ end of the Zfy-1 gene. (i) A restriction map of the 1%kb insert of XT5.G.9, which is at the end of the T5 contig (B, BarnHI; E, EcoRI; S, &cl; Sal, SalI; and X, XbaI). The hatched boxes represent the vector and the line between represents the insert. The position of pDP1122.B homology, as deduced from (ii), is indicated. Also detailed is the source of pT5.G.9.A2, a HindIII-Sal1 fragment of XT5.G.9, whose hybridization pattern is shown in (iii) left panel. The overlap of XT5.G.9 with cT5.2.12 is indicated by the arrowed line. (ii) Southern blot analysis of XT5.G.9 hybridized with pDP1122.B, a 250-bp EcoRINcoI fragment of pDP1122 (Ref. (9)). As the NcoI site lies at the translation-start AUG of the message, it essentially contains only the 5’ noncoding part of the gene. The XT5.G.9 DNA is cut with BamHI/SalI (B), EcoRI (E), and SacI/SalI (S). (iii) Southern blot analysis of EcoRI-digested genomic DNA hybridized with pT5.G.9.A2 (left) and pDP1122.B (right). The DNAs are (1) C57BL/6 male, (2) C57BL/6 female, (3) C57BL/6 XXSxr” male, and (4) C57BL/6 XX&r* male. In (ii) and (iii) the sizes of the hybridizing restriction fragments are indicated in kilobases.

quences that are found in multiple copies within the Sxr region. This is exemplified in Fig. 4 by pT5.1.1.PH, where the loci are defined by the EcoRI fragments of 16 kb (Y and ASxr’), 9 kb (T5), 7 kb (B3), 5 kb (Zf2 andZfB), and 4.2 kb (Sxrb). To characterize further the nature of this amplification we used pT5.1.1.PH to screen our cosmid library for the loci detected by Southern blot analysis. In addition to the two ASxrb loci, Zf2 and T5, we have isolated DNA from the nondeleted loci, defined by the 7-kb EcoRI fragment (the B3 locus) and 5-kb fragment (the ZfB locus). No DNA that contained a 16- or 4.2-kb, pT5.1.1.PH-positive, EcoRI band was isolated, which indicates that these fragments define another two Sxr loci, one of which is deleted from the Sxrb region. These latter contigs are detailed in Fig. 6. It should be noted that the ZfB contig has a gap, between the ends of cZf2.10 and XZfB.21, but the contig has been constructed on the basis of the observed restriction site conservation with Zf2 on either side of the gap. To analyze how these loci relate to each other, probes isolated from each region were hybridized directly to

STRUCTURE

OF

Sxr

REGION

OF

MOUSE

31

Y CHROMOSOME

z!2 s cZf.‘J3

-

I

I

3 cZf.2.1 ,a

; I.Zf.F.2

I

XZf.F.8

I

m

I

I

Zf.B cZf.2.lcl

I

d

; hZf.B.21

I I 4

pf.B.6w Increased conservation at the restriction site level

l-l

cT5.2.13

nA,

A

0

cT5.2.16 cT5.2.12 hT5.G.7 -

5’h3

hT5.G.9

Uf.B.3

-

cT5.2.11

hii+

-

A

- pZf.2.1 .B

0

- pT5.1.1.PH

o

- pZf.2.1 .D

.

- pZf.l.8.A

e 0

I

- 3’ end of i’&-2

A

-PM

m - pZf.B.21 .A2

- pT5.2.12.A2 - pT5.2.13.03

&I -5’endofZ&-1 FIG. T5, and to these exhibit

6. Schematic representation of DNA isolated from four Srr loci containing sequences homologous to pTB.l.l.PH. The loci Zf’2, ZfB, B3 are shown here. The positions of sequences homologous to probes isolated from these loci are indicated by various symbols. A key symbols is provided at the foot of the diagram. The orientations of the Zfy gene exons are indicated. The areas at which Zf2 and ZfB increased conservation at the restriction enzyme level are marked.

the cosmid and phage DNA of each contig. All the probes used were ones that gave clear results on Southern blots and, in general, the fragment sizes, detected strongly in the DNA of the contigs, were approximately the same size as those detected in genomic DNA. Exceptions were observed either when the cross-reacting area of the contig lay at the end of an insert or, specifically in the case of pY8, when 1.8 and 0.7-kb EcoRI fragments detected in both cZf.l.8 and cZf2.1 were not detected on Southern analysis of genomic DNA. These data are diagramed in Fig. 6. It can be seen that only pT5.1,l.PH and pY8 are present in all four regions, while probes pZf2.1.B and pZf2.l.D are present in all but the T5 contig. It can also be seen that the distribution of the probes suggests greater conservation between B3 and T5 or Zf2 and ZfB than

between these two pairs of loci. This pairing of loci is particularly reflected at the restriction enzyme site level for Zf2 and ZfB, which share many restriction sites in the DNA that lies between pY8 and pZfZ.l.B, as is shown by the 5kb EcoRI and Wkb BglII fragments of pT5.1.1.PH (Figs. 4A and 4B), which define both ZfB and Zf2. The intensity of hybridization to these fragments between the XXSxr” and the XX&r’ tracks can be seen to drop when compared to the relative intensity of other fragments in these two tracks. This results from the absence of the Zf2 locus from the XXSxrb DNA. This is similarly the case for the 2.1-kb EcoRI band of pZf2.1.D (Fig. 2B) as previously described. Furthermore, limitedsequence analysis of ZfB and Zf2 revealed 100% homology in an 87bp overlap.

32 The Occurrence

MITCHELL

of CpG-Rich

Restriction

Sites

In an attempt to identify any CpG-rich islands, which may be a part of the promoter of a gene (Lindsay and Bird, 1987), we analyzed the contig DNA with the restriction endonucleases BssHII, EagI, NotI, and Sac11 because each contains two CpG dinucleotides in its cutting site. In the 270 kb of ASxrb DNA tested, only one BssHII, two EagI, two SacII, and one Not1 site were found. None of these sites were clustered, with the closest pair of sites being 4 kb apart. The Sac11 and Not1 sites were shown to be methylated in genomic DNA from liver using probes from the EcoRI restriction fragments containing these rare-cutting sites. The methylation of these sites together with the lack of clustering suggests that there are no CpG islands in our contig DNA.

AND

BISHOP

1

2

3

4

-A

Sxrb

-ZfB - Sxrb - Zf2

Novel ASxrb Loci The probes presented here that detect multiple Sxr loci can be used to clone further ASxrb DNA from new regions. The probe pT5.1.1 .PH clearly detects a novel ASxr’ locus defined by the 16-kb EcoRI fragment present in the genomic DNA of XXSxr” but not XXSxrb (Fig. 4B) and as described earlier an ll-kb EcoRI fragment detected by pT5.2.13.D3 also defines a previously uncloned locus (Fig. 3A). So far we have been unable to isolate DNA from the regions defined by these fragments. Furthermore, another probe, pZf.B.21.A2, hybridizes to a distinct ASxrb fragment at 6.7 kb on Sac1 digestion of genomic DNA (Fig. 7). This probe was subcloned from XZf.B.21, which is part of the ZfB contig. Analysis of all the contig DNA with this probe did not reveal the presence of a 6.7-kb Sac1 fragment and no hybridization was seen to end fragments of inserts. Therefore, the 6.7-kb Sac1 fragment must define a further ASxrb locus. A novel locus, present in Sun and Sxrb, is defined by the 4.5kb hybridizing fragment (Fig. 7). A 4-kb fragment mapping to the Zf2 contig/locus can barely be seen, but the probe failed to hybridize to the DNA of the T5 contig. This means that we now have six distinct loci within the ASxrb region. Nothing can be said about the positions of these new loci within the ASxrb region except that they lie outside the 270 kb of DNA already cloned. DISCUSSION

We have isolated 270 kb of ASxrb DNA by chromosome walking in cosmids and phage. Molecular analysis of the Zfy genes by others indicates that the Sxrb deletion arose by an unequal crossover event between Zfy-1 and Zfy-2 in the homozygous XSxPYSxr” father (E. M. Simpson and D. C. Page, unpublished data). This has resulted in the formation of a chimeric Zfy gene (“Zfy-3” ?), in the Sxrb region, carrying the

blot analysis of SacI-digested DNA hybridFIG. 7. Southern ized with pZf.B.21.A2. The probe is shown hybridized to (1) C57BL/6 male, (2) C57BL/6 female, (3) C57BL/6 XXSnr” male, and (4) C57BL/6 XXSxrb male DNAs. The sizes of the hybridizing restriction fragments are indicated in kilobases. The Srr loci to which the hybridizing restriction fragments map are indicated where known. The 6.7-kb (ASxrb) and 4.5kb (Snrb) fragments define new Sxr loci. The 4.0.kb Zf2 fragment is very faint but pZf.B.Zl.A2 does hybridize strongly to a 4.0.kb Sac1 fragment in the Zf2 contig.

promoter and start codon of Zfy-2, fused to the structural exons and 3’ end of Z&-l. Thus, the deletion breakpoints lie within Zfy-1 and Zfy-2, the deleted DNA being flanked by the duplicated Zfy genes. These data confirm our own findings obtained from the preliminary analysis of a YAC clone from this region. This YAC contains both Z&l and the T5 locus, but not Zfy-2. This places the T5 locus at the 5’ end of Zfy-1 and localizes one of the deletion breakpoints to within the Zfy-1 gene, or close to its 5’ end (C. E. Bishop and Z. Larin, unpublished). As the Zf2 contig maps to the 3’ end of Zfy-2, the unpublished data of Simpson and Page would place the Zf2 locus at the other extremity of the deletion. The Sxl contig is, in large part, composed of nonduplicated DNA, lies entirely within the deletion, and does not overlap with either the T5 or the Zf2 contig. The Sxl locus, therefore, must lie in the central portion of the deletion. A diagram of the deleted DNA, showing the position of the contigs, is presented in Fig. 8. A minimum size estimate can be calculated for the ASxrb DNA based on the size of our contigs and the sizes of the Zfy genes. The unequal crossover between the Zfy genes will, in total, have deleted one whole Zfy gene. By comparison with the human ZFY (Page et

STRUCTURE Sxl(145kb)

Zf2 (60kb)

\’

‘I 1 -85kb I

I

I -35kb

I I

OF

‘\

15(65kb)

\

/ ' >145kb >320kb

SW REGION

'

,

/

I :3Okb

II I>

/’ .