Comparative mapping of the region of human chromosome 7 deleted in williams syndrome

Abstract

Williams syndrome (WS) is a complex developmental disorder resulting from the deletion of a large (approximately 1.5-2 Mb) segment of human chromosome 7q11.23. Physical mapping studies have revealed that this deleted region, which contains a number of known genes, is flanked by several large, nearly identical blocks of DNA. The presence of such highly related DNA segments in close physical proximity to one another has hampered efforts to elucidate the precise long-range organization of this segment of chromosome 7. To gain insight about the structure and evolutionary origins of this important and complex genomic region, we have constructed a fully contiguous bacterial artificial chromosome (BAC) and P1-derived artificial chromosome (PAC) contig map encompassing the corresponding region on mouse chromosome 5. In contrast to the difficulties encountered in constructing a clone-based physical map of the human WS region, the BAC/PAC-based map of the mouse WS region was straightforward to construct, with no evidence of large duplicated segments, such as those encountered in the human WS region. To confirm this difference, representative human and mouse BACs were used as probes for performing fluorescence in situ hybridization (FISH) to metaphase and interphase chromosomes. Human BACs derived from the nonunique portion of the WS region hybridized to multiple, closely spaced regions on human chromosome 7q11.23. In contrast, corresponding mouse BACs hybridized to a single site on mouse chromosome 5. Furthermore, FISH analysis revealed the presence of duplicated segments within the WS region of various nonhuman primates (chimpanzee, gorilla, orangutan, and gibbon). Hybridization was also noted at the genomic locations corresponding to human chromosome 7p22 and 7q22 in human, chimpanzee, and gorilla, but not in the other animal species examined. Together, these results indicate that the WS region is associated with large, duplicated blocks of DNA on human chromosome 7q11.23 as well as the corresponding genomic regions of other nonhuman primates. However, such duplications are not present in the mouse.

Figure 1

Schematic representation of the human WS region. A working model of the long-range physical organization of the human WS region is depicted based on data generated in numerous studies (Osborne et al. 1996, ,; Perez Jurado et al. 1996, ; Robinson et al. 1996; Wang et al. 1997; Lu et al. 1998; Meng et al. 1998a; E.D. Green and B.J. Trask, unpubl.). The relative positions (not to scale) of key gene/pseudogene sequences and genetic markers are indicated. A middle, single-copy region contains numerous known genes [FKBP6 (Meng et al. 1998b), FZD3 (Wang et al. 1997), WSTF (Lu et al. 1998), BCL7B (Jadayel et al. 1998; Meng et al. 1998a), WS–βTRP (Meng et al. 1998a), WS-bHLH (Meng et al. 1998a), STX1A (Osborne et al. 1997b; Nakayama et al. 1998), ELN (Ewart et al. 1993), LIMK1 (Frangiskakis et al. 1996; Tassabehji et al. 1996), WSCR1 (Osborne et al. 1996), RFC2 (Peoples et al. 1996), CYLN2 (Hoogenraad et al. 1998)]. This region is flanked by several large (estimated at ∼200–300 kb) genomic segments of nearly identical composition (represented by stippled boxes), each of which contains the indicated gene/pseudogene sequences [GTF2I/GTF2IP1 (Perez Jurado et al. 1998), PMS2L (Osborne et al. 1997a), p47–phox/p47–phox-P (Gorlach et al. 1997)]. Note that the relative order of the latter within the duplicated segments has not been established nor has the presence of GTF2IP1 in all the duplicated segments (reflected by the ? after GTF2IP1 in the far left duplicated segment). The genomic segment commonly deleted in WS that spans ∼1.5–2.0 Mb is indicated along the bottom (Osborne et al. 1996; Perez Jurado et al . 1996; Robinson et al. 1996; Urban et al. 1996; Wang et al. 1997; Meng et al. 1998a; Wu et al. 1998).

Figure 2

BAC/PAC-based STS-content map of the mouse WS region (oriented with centromere leftward and telomere rightward). The deduced positions of 44 STSs are depicted along the top, with the indicated clones shown as horizontal lines below. Relevant information about the STSs is available in GenBank (http://www.ncbi.nlm.nih.gov). The STSs were developed from clone insert ends (red), known genes (blue), or a conserved human DNA sequence (black). (●,█) The STS is confirmed to be present in that clone by PCR testing. When an STS corresponds to a clone insert end, a red square is present at the end of the clone from which it was derived. PAC clones are indicated with a P at the beginning of their names. Four of the BAC clones were isolated from the Genome Systems C57BL/6 mouse library (111E07, 285A04, 155F12, and 159L24), with the rest of the BACs coming from the Research Genetics CITB–CJ7-B mouse library. The size of each BAC/PAC, as assessed by pulsed-field gel analysis, is also provided. The indicated BAC/PAC overlaps were confirmed by restriction enzyme digest-based fingerprint analysis (Marra et al. 1997; data not shown). The depicted clones together span ∼680 kb of DNA from mouse chromosome 5. Preliminary genomic sequence analysis has revealed the presence of Eln, Limk1, and part of Wscr1 in BAC 42J20, Wscr1, Rfc2, and Cyln2 in BAC 303E12, and Gtf2i and p47–phox in BAC 391O16, as indicated along the bottom (U. DeSilva and E.D. Green, unpubl.).

Figure 3

Human–mouse comparative FISH analysis of the WS region. Representative BAC clones from the ELN/Eln- and p47–phox-containing regions were selected from the physical maps of the human and mouse WS region. FISH analyses were performed on both metaphase chromosomes (insets) and interphase nuclei (main panels). Clones containing ELN/Eln (RG030E19 and 42J20) show a hybridization pattern on human chromosome 7q11.23 and mouse chromosome 5G2 typical of a single-copy segment. In mouse, the p47–phox-containing BAC (391O16) also produces a single-copy signal on metaphase and interphase chromosomes. In contrast, the human p47–phox-containing BAC (RG350L10) produces a very large and intense signal on metaphase chromosomes at 7q11.23 and multiple fluorescent dots on interphase chromosomes (also see Table 1). Sequences that cross-hybridize to the p47–phox BAC are encountered in 7p22 and 7q22 in human, in addition to the cluster in 7q11.23. We note that there is overlap between these locations and the sites detected with probes for PMS2-related sequences (Nicolaides et al. 1995).

Figure 4

Comparative FISH analyses of WS-region segments on nonhuman primate chromosomes. (a) Idiograms of chromosome-7 orthologs in human (H), chimpanzee (C), gorilla (G), and orangutan (O) are taken from Yunis and Prakesh (1982) and used to indicate the positions of the breakpoints of inversions that have occurred during primate evolution. (b) Summary of the locations of FISH signals produced by a human ELN-containing BAC (RG030E19) on human, chimpanzee, gorilla, and orangutan chromosomes. Sequences in this ELN-containing segment appear single copy in all species and map to the expected locations, based on the known inversions. The ELN segment maps in gibbon near the centromere of a chromosome that has been shown by chromosome painting to contain sequences orthologous to human chromosome 7 (not shown; Jauch et al. 1992). (c) Summary of the locations of FISH signals produced by a human p47–phox-containing BAC on hominoid chromosomes (see Fig. 3 and d for corresponding FISH images). (d) Representative images of metaphase chromosomes (insets) and interphase nuclei (main panels) from chimpanzee, gorilla, orangutan, and gibbon after FISH with a human p47–phox-containing BAC (RG350L10). Local duplication of p47–phox sequences within the region corresponding to human chromosome 7q11.23 is evident in all species, both from the size of the signal in metaphase and the cluster of multiple dots in interphase. Additional cross-hybridizing sequences are detected in 7p22 and 7q22 in chimpanzee and in 7p22, 7p13, and 7q22 in gorilla. Bands are indicated using nomenclature from the human karyotype.