Introduction

The X-linked lymphoproliferative syndrome (XLP) is an X-linked recessive genetic disorder characterized by fatal mononucleosis, hypogammaglobulinemia or malignant lymphoma following infection by Epstein-Barr virus (EBV) [1], Genetic linkage studies have localized the XLP locus to the Xq24–25 region. The DXS42 and DXS37 markers were shown to be respectively at about 1–2% (Z = 17.5) and 3% (Z = 11.8) recombination proximal to the XLP locus [2–5]. A third marker in the Xq25 region, DXS12, has also been reported to be linked to the XLP locus, but no recombination was detected with this marker (Z = 7.5) [4, 5]. An interstitial deletion involving a portion of the Xq25 region has recently been detected in an XLP family [6]. This deletion should be useful in physical mapping and the isolation of candidate genes in this region.

Establishing physical map around the DXS12 and DXS42 markers that are closely linked to the XLP locus is an important step towards isolating the susceptibility gene. In this paper, we report the assignment of several DNA markers to the Xq24–25 region that contains loci closely linked to the XLP locus, and the mapping of overlapping YACs covering a distance of about 850 kb around DXS42. We also show that DXS 12 and DXS42 are situated with a maximum distance of about 50 kb, and that DXS37 is separated from them by a maximum distance of about 3,700 kb.

Materials and Methods

Cell Lines

A human lymphoblastoid cell line GM 1202, which contains four X chromosomes per cell, was obtained from J. L. Mandel (Strasbourg, France). The lymphoblastoid cell line IARC745 was established in our laboratory from a female individual (XX).

Two human-mouse hybrid fibroblast cell lines were obtained from H. Ropers (Nijmegen, The Netherlands): 494 × 393, with translocation (X;13) (q22;q32), retaining human Xq22-qter; and 790 × 175, with translocation (X; 19) (q24;q13), retaining human Xq24-qter [7]. The rodent-human hybrid cell line CH63R, with translocation (X; 17) (q26;p12) retaining human Xq26-qter, was obtained from J. L. Mandel [8]. The human-mouse hybrid cell line CY2, with translocation (X; 16) (q26;q24) retaining human Xp-q26, was obtained from G. Sutherland (North Adelaide, Australia) [9].

C12D is a human-hamster cell hybrid containing a single human X chromosome and A23N is a parental hamster cell line of C12D; both of these cell lines were obtained from P. Goodfellow (ICRF, London, UK).

DNA Probes

The preparation of probes 36B-2 (DXS10), p22–33 (DXS11), pL2 (DXS12), 30RIB (DXS37), 7F1 (DXS42), St16 (DXS53), St1 (DXS86), pX58C (DXS99), pX45h (DXS100), and p2aB5 (DXS138), has been described elsewhere [10]. The left- and right-end probes of YAC vector were obtained by cutting Plasmid pBR322 with BamHI and PvuII: the 2.7-kb fragment was used as the left-end probe and the 1.6 kb fragment was used as the right-end probe.

We prepared the following probes: R3–5 was a 1.8 kb EcoRI/HindIII fragment in pUC18, isolated from YAC137. 63R5, 63L3 were subclones of YAC4563 and contained respectively 2- and 4.5-kb EcoRI/HindIII fragments in pUC18. 64–22 was a 1-kb EcoRI fragment in pUC18 and p64L8 was a EcoRI/HindIII fragment in pUC18; both of these were obtained from YAC4564. Y63.3 was a 1.8-kb BamHI fragment in pBluescript, obtained by the AluPCR technique [11] from YAC4563.

Screening of YAC Libraries

Two independently constructed YAC libraries were screened. One (from the Institute of Molecular Genetics and Howard Hughes Medical Institute) was constructed from the hybrid cell line X3000-11.1, which contains the human X chromosome region Xq24-qter [11]. The second (ICRF YAC library) was constructed from human lymphoblastoid cell line GM1416B which contains four X chromosomes [12]. Both libraries were made by partially digesting DNA with EcoRI and cloning it in pYAC4 vector. The YAC library made from X3000-11.1 was screened as described in Nelson et al. [11]. The ICRF YAC library was screened by hybridization as described in Larin et al. [12] using high-density filters from the ICRF reference library system [13].

DNA Preparation in Agarose Plugs and PFGE Analysis

For preparation of yeast DNA plugs, we used a modified protocol of McCormick et al. [14]. 50 ml of AHC medium cultured (OD600 = 2) yeast cells were harvested by centrifugation then suspended in 1.5 ml of 10 mM Tris pH 7.5/1 mM EDTA (TE) buffer and mixed with 2.5 ml of 1% LMP agarose (InCert) with the addition of 400 µg of Zymolyase 20T (ICN Biomedicals). Each plug was made from 85 µl of this mixture. The plugs were then incubated in ET solution (0.5 M EDTA pH 8.0/10 mM Tris) at 37°C for 8–10 h, then in ESP solution (0.5 M EDTA/1% N-lauroylsarcosine and 20 mg/ml proteinase K) at 50°C for about 12 h. After incubation, the plugs were washed with excess TE three times at room temperature, twice for 30 min with 4 µg/ml phenylmethylsulfonyl fluoride (PMSF) solution at 50°C, followed by storage in + 4°C. Preparation of genomic DNA plugs was performed as described by Nguyen et al. [15].

We used a Bio-Rad CHEF-II system for pulsed-field gel electrophoresis analysis. DNA fragments were routinely separated on 0.8% agarose gels in 0.5 × TBE buffer (1 ×: 90 mM Tris-borate/2 mM EDTA) at about 15°C. The migration conditions are described in the figure legends. After electrophoresis, DNA was blotted on to Hybond-N+ membrane in 0.5 M NaOH/1.5 M NaCl solution preceded by treatment with 0.25 N HCl solution for 15 min. The hybridization was carried out in CHURCH solution: 0.5 M phosphate buffer pH 7.2/7% sodium dodecyl sulfate (SDS)/1% bovine serum albumin (BSA) at 65°C and stringent washing conditions were used, i.e. 2 × SSC (1 ×: 0.15 M NaCl/1.5 mM sodium citrate)/0.1% SDS twice followed by 0.1 × SSC/0.1% SDS once at 55–65°C. Genomic DNA extraction and conventional Southern blotting were carried out as described elsewhere [16].

Results

Localization of Several Markers in the Xq24–25 Region

Table 1 presents the analysis of the localization of several DNA markers previously assigned to the Xq24–27 region [10], related to a set of somatic hybrids with different breakpoints (see Materials and Methods). On the basis of these results, DXS11, DXS12, DXS37, DXS42, DXS100 and DXS138 were assigned to the Xq24–25 region, DXS53 to Xq26, and DXS10, DXS86, DXS99 to Xq26-qter. These data are consistent with the observation of Reilly et al. [17], who localized several DNA markers from the Xq24–26 region using a different somatic hybrid panel and proposed the following order: Xcen-(DXS42, DXS37, DXS100)-DXS53-HPRT-(DXS10, DXS86, DXS177)-Xqter. DXS10 and DXS86 have been shown to be physically linked [17–19, Wang et al., unpubl. results].

PFGE Analysis and Physical Linkage between DXS12, DXS42 and DXS37

The results presented in table 1 suggested that the loci DXS11, DXS12, DXS37, DXS42, DXS100 and DXS138 are located in the same subregion of the X chromosome at Xq24–25. We therefore attempted to see if they are linked physically by using PFGE. DNA from the cell line GM1202 was digested completely with rare-cutter enzymes XhoI, SmaI, SalI, NotI, NruI and various combination of these, and hybridized with Xq24–25 probes (summarized in table 2). Figure 1 presents the hybridization results with probes for DXS12, DXS42 and DXS11. No evidence of physical linkage between DXS11, DXS100, DXS138 and either DXS12, or DXS37, or DXS42 was detected. However, we observed that the probes for DXS12 and DXS42 gave the same hybridization bands with all of the rare-cutter enzymes used (fig. 1 A, table 2), the smallest being 50 kb with ClaI (fig. 1B). Another observation was that the DXS37 probe hybridized a 3,700-kb NotI fragment, and this fragment was shared by DXS12 and DXS42 (fig. 2). Furthermore, DXS12, DXS42 and DXS37 revealed the same 4000 kb NruI fragment (table 2). The NotI band disappeared after double digestion with NotI and SalI (fig. 2). The cohybridization of these probes could not be an artifact due to incomplete probe stripping, as during the intervals between these hybridizations, the filter was also used with other probes which revealed different-sized fragments.

Fig. 1
figure 1

PFGE analysis of Xq24–25 region probes. GM1202 cell line DNA was digested with different rare-cutter enzymes and separated by a ramping pulse time of 1–10 s for 20 h migration at 200 V. After transfer, the same filters were used for subsequent hybridizations. A Hybridizations to DXS12, DXS42 and DXS11 with digestion of SmaI, XhoI, SalI or a combination of them. B Hybridizations to DXS42 (a) and DXS12 (b) with ClaI digestion. Lambda concatamers was used as size standard. LM = Limit of mobility.

Full size image
Fig. 2
figure 2

PFGE analysis of DXS42, DXS12 and DXS37 with NotI digestion and NotI/SalI. The gel was run at 60 V with a pulse time of 35 min for 108 h, in 1 × TAE buffer (0.04 M Tris-acetate/1 mM EDTA). The same filter was hybridized subsequently with three probes. S. pombe and S. cerevisiae YNN295 were used as size standard. The fragment of 3,700 kb is indicated.

Full size image
Table 1 Analysis from mapping Xq24–27 markers with somatic cell hybrids
Full size table
Table 2 Approximate sizes (kb) of rare-cutter restriction fragments detected with Xq24-25 probes
Full size table

Mapping of Overlapping YACs around DXS12 and DXS42

The fact that most PFGE fragments for DXS12, DXS42 were very small and that only a limited number of DNA markers were available made it difficult to establish a long-range restriction map directly with genomic DNA. However, the development of the YAC cloning technique provided us a very helpful approach to establishing an extended map of the region around DXS12 and DXS42.

The DXS42 probe was used to screen two independently constructed YAC libraries (see Materials and Methods), by colony hybridization. Three positive clones were isolated: YAC137 (200 kb) from the library made from cell line X3000-11.1; YAC4563 (ICRF code: ICRFy900F0981) and YAC4564 (ICRF code: ICRFy900A0816) from the library constructed form GM1416, 780 and 800 kb in size, respectively.

The restriction map of each YAC was constructed by indirect end-label mapping [20]. DNA plugs of these YACs were digested with various concentrations of rare-cutter enzymes. After electrophoresis and transfer, filters were hybridized successively with vector left- and right-end probes. On the basis of the individual restriction map of each YAC, we were able to orient and align these overlapping YACs to obtain a restriction map shown in figure 3B. As described above, DXS12 and DXS42 are closely linked, both being situated in the 90-kb XhoI fragment of YAC137 (fig. 3), but they could not be oriented further.

Fig. 3
figure 3

Physical map of overlapping YACs around DXS12 and DXS42. A Restriction map at the genomic DNA level, illustrating the localization of loci DXS 12, DXS42 and the probes generated from the YACs. The methylated restriction sites are indicated by*. B Three YACs: YAC137 (200 kb), YAC4563 (780 kb) and YAC4564 (800 kb) were mapped and aligned by partial rare-cutter enzyme digestion and PFGE analysis. X = XhoI; SI = SalI, Sc = SacII; BH = BssHII; Eg = EagI. The cocloned region at the left extremity of YAC4564 is indicated by a black box. L = Left end; R = right end.

Full size image

In theory, the restriction site pattern in an overlap region between YACs should be consistent. This was the case for the majority of the enzymes used. In order to verify the fidelity of this map, several probes were isolated by subcloning YAC DNA in lambda phage vectors or by the AluPCR technique [11] (fig. 3). Each probe was checked by hybridization on somatic hybrids, to confirm its localization on the same subregion in the X chromosome as DXS12 and DXS42. The restriction site patterns around the DXS42 probe, R3-5 and 64-22 are consistent in three YACs. For instance, R3-5 and 64-22 detected the same 350-kb SacII fragment in both YAC4563 and YAC4564. However, an important discrepancy between YAC4563 and YAC4564 was noticed in the left extremity. In fact, the probe 64L8 (left-end probe of YAC5464) was found to be located in the Xq22–23 region. Furthermore, the probe 63L3 (left-end probe of YAC4563), which overlaps with YAC4564, did not recognize any YAC4564 sequence (data not shown). Taken together these results indicated that a cocloning event had taken place at the left extremity of YAC4564 (fig. 3).

The other markers in the Xq24–25 region (DXS11, DXS37, DXS100 and DXS138) have been tested but none of them is contained in these overlapping YACs.

The Restriction Map from YACs Related to X Chromosome Genomic DNA Structure

The probes 63R5, R3-5, 64-22 and 63L3 were hybridized to genomic DNA, and all recognized the same 3,700 kb NotI fragment as DXS12, DXS42 and DXS37. This indicated that the 850-kb map is included in this 3,700 NotI fragment, but the relative position of DXS37 could not be determined at this stage.

To verify that the map from the YACs reflects the correct structure of restriction sites around DXS42 in the human X chromosome, DNA from YACs and cell lines GM1202, IARC745 and C12D were digested with various restriction enzymes and fractionated by classical or PFG electrophoresis. Using classical Southern blotting, all the probes tested hybridized to the same fragments in YACs and in genomic DNA. However, on PFGE, a small difference in fragment migration between YAC and genomic DNA was noticed with all enzymes and probes used, which may be due to the difference in DNA concentration between yeast and genomic DNA plugs. Sites for some rare-cutter enzymes such as SacII, BssHII, and EagI were not cleaved in genomic DNA (data not shown), in contrast to YAC DNA. Taking into account the fact that these enzymes are all sensitive to 5-methylcytosine in the CpG sequence, and that no methylation of cytosine has been observed in yeast [21], we assume that these sites in genomic DNA are methylated.

Discussion

We have used a combined approach to construct a physical map around DXS12 and DXS42 loci, which are genetically linked to the XLP gene. The results from localization of DNA markers with somatic hybrids are consistent with those obtained by other investigators [17, 18, 22]. We confirmed the localization of DXS11, DXS37, DXS42 and DXS100, previously assigned to the Xq24–25 region, and we localized two other loci (DXS12 and DXS138) to the same sub-region of the X chromosome. The results from PFGE and YACs indicate that DXS12 and DXS42 are tightly linked, within 50 kb, and DXS37 is separated from DXS12 and DXS42 by a maximum of 3,700 kb, but we do not know their positions relative to the centromere and telomere.

The YAC cloning technique proved its value for large-scale physical mapping, especially in this region where few DNA markers were available. However, caution must be exercised, as up to 20% of YACs could be chimeric [23], maybe particularly with larger inserts. In our case, YAC4564 was shown to be a cocloned one, with a large insert from Xq24–25 and a small insert probably from Xq23.

The methylation of CpG sequence in genomic DNA complicates the verification of the YAC contig map at genomic DNA level, as described by other authors [24]. However, the fidelity of our YAC map was largely confirmed in an overlapping region where the restriction patterns among three (or two) YACs are identical. Furthermore, internal probes were used and detected the same sized fragment in both YAC and genomic DNA.

In conclusion, although the X chromosome has been much studied, little was known hitherto about the Xq24–25 region, principally because of the lack of DNA markers in this region, which hampered PFGE mapping and isolation of YAC contigs. The work presented in this paper has provided us with insight into this region containing the XLP gene. The probes generated will contribute to the further long-range physical mapping, and the detection of polymorphism of these probes will be helpful in precisely localizing the XLP locus and isolating the gene.