Ordered shotgun sequencing, a strategy for integrated mapping and sequencing of YAC clones

Ordered shotgun sequencing proposes to organize the mapping and sequencing of YACs with a hierarchical strategy that incorporates a feedback loop. Building on current protocols, a YAC is subcloned into plasmids, plasmid insert ends are sequenced, and the sequences are overlapped to create a partial map. Complete sequencing then starts with plasmids whose end-sequence tracts have overlapped, but to a minimal extent. The next plasmids to be sequenced are again selected for least overlap, striking out progressively to span the YAC with minimal directed gap-filling. Simulations support its feasibility and indicate that during the generation of the complete sequence, the approach facilitates the early choice of regions for selective sequencing, for example, for coding units. The sequencing of plasmids would also require less redundancy, and discriminate repetitive sequences more easily, than random sequencing across larger clones. The overall effort scales with YAC size and can be further reduced by additional mapping information.

Isochores and CpG islands in YAC contigs in human Xq26.1-qter

GC levels were assessed at 37 loci across 30 Mb of Xq26.1-qter, a region physically mapped in overlapping yeast artificial chromosome clones. In 8 Mb of R band Xq26, GC is relatively high (up to 44%) in the proximal 4 Mb and relatively low (40-41%) in the distal 4 Mb. Consistently low GC values (38-41%) are observed in G band Xq27. In contrast, further toward the telomere in Xq28, the GC level rises progressively to reach 52% at 2 to 4 Mb from the end of the chromosome; this region is delimited by low GC loci. Across these regions of Xq, the content of rare-cutter restriction enzyme sites containing CpG, including "CpG islands" in the most completely mapped Xq26-27.1 region, is correlated with GC level. Isochore mapping can thus provide one index of putative gene content across mapped regions.

High-resolution mapping of probes near the X-linked lymphoproliferative disease (XLP) locus

Fluorescence in situ hybridization (FISH) was employed in high-resolution mapping of probes near the X-linked lymphoproliferative disease (XLP) locus. The map includes the DXS42, DXS12, DXS6, DXS982, DXS739, DXS75, DXS100, DXS10, and DXS177 loci. Metaphase analysis showed that DXS12 and DXS42 mapped to proximal Xq25, while DXS10 and DXS177 mapped to proximal Xq26.1. DXS6, DXS982, DXS739, DXS75, and DXS100 were in Xq25. The order of probes deduced from interphase FISH was: Xq24-(DXS12, DXS42)-DXS6-DXS982-DXS739-DXS75-DXS100+ ++-DXS10-DXS177-Xq26.2. We estimate that the entire region between DXS12 and DXS177 is about 7 Mb. Our previous study indicated that all three XLP deletions (63-3, 66-1, and 43-4) lacked DXS739. We now report that DXS75 and DXS982 are also missing in these deletions. Using interphase FISH measurements, we estimate that 2 Mb are absent in 63-3, and 4 Mb are absent in 66-1 and 43-4. This FISH map confines the XLP candidate gene region to a 2-Mb interval between DXS6 and DXS100 and places DXS100 distal to the XLP locus. This study also demonstrates that small probes (0.6 to 3.6 kb) can be utilized in FISH.

1.5-Mb YAC contig in Xq28 formatted with sequence-tagged sites and including a region unstable in the clones

A contig of 20 yeast artificial clones (YACs) has been assembled across 1.5 Mb of Xq28 and formatted with nine previously reported probes and nine STSs developed from the sequence of probes and end fragments of YACs. YAC end fragments were obtained by subcloning, Alu-vector PCR, or primer-ligation PCR methods. Eighteen of the YACs were recovered from a library specific for Xq24-q28; two that fill a gap were obtained from a second library made from total human DNA. One region, containing probes pX78c and 2A1.1, was unstable in YACs, but it was possible to generate a self-consistent map of DNA over the entire contig. Overlaps were confirmed by Southern blot analyses of YAC DNAs, and pulsed-field gel electrophoresis confirmed the extent of the contig and identified at least four CpG islands in the region.

Alu and L1 sequence distributions in Xq24-q28 and their comparative utility in YAC contig assembly and verification

The contents of Alu- and L1-containing TaqI restriction fragments were assessed by Southern blot analyses across YAC contigs already assembled by other means and localized within Xq24-q28. Fingerprinting patterns of YACs in contigs were concordant, and using software based on that of M. V. Olson et al. (1986, Proc. Natl. Acad. Sci. USA 83: 7826) to analyze digitized data on fragment sizes, fingerprinting itself could establish matches among about 40% of a test group of 435 YACs. At 100-kb resolution, both repetitive elements were found throughout the region, with no apparent enrichment of Alu or L1 in DNA of G compared to that found in R bands. However, consistent with a random overall distribution, delimited regions of up to 100 kb contained clusters of repetitive elements. The local concentrations may help to account for the reported differential hybridization of Alu and L1 probes to segments of metaphase chromosomes.

Chromosomal bar codes produced by multicolor fluorescence in situ hybridization with multiple YAC clones and whole chromosome painting probes

Colored chromosome staining patterns, termed chromosomal 'bar codes' (CBCs), were obtained on human chromosomes by fluorescence in situ hybridization (FISH) with pools of Alu-PCR products from YAC clones containing human DNA inserts ranging from 100 kbp to 1 Mbp. In contrast to conventional G- or R-bands, the chromosomal position, extent, individual color and relative signal intensity of each 'bar' could be modified depending on probe selection and labeling procedures. Alu-PCR amplification products were generated from 31 YAC clones which mapped to 37 different chromosome bands. For multiple color FISH, Alu-PCR amplification products from various clones were either biotinylated or labeled with digoxigenin. Probes from up to twenty YAC clones were used simultaneously to produce CBCs on selected human chromosomes. Evaluation using a cooled CCD camera and digital image analysis confirmed the high reproducibility of the bars from one metaphase spread to another. Combinatorial FISH with mixtures of whole chromosomes paint probes was applied to paint seven chromosomes simultaneously in different colors along with a set of YAC clones which map to these chromosomes. We discuss the potential to construct analytical chromosomal bar codes adapted to particular needs of cytogenetic investigations and automated image analysis.

Anhidrotic ectodermal dysplasia gene region cloned in yeast artificial chromosomes

Anhidrotic ectodermal dysplasia (EDA), an X-chromosomal recessive disorder, is expressed in a few females with chromosomal translocations involving bands Xq12-q13. Using available DNA markers from the region and somatic cell hybrids we mapped the X-chromosomal breakpoints in two such translocations. The breakpoints were further mapped within a yeast artificial chromosome contig constructed by chromosome walking techniques. Genomic DNA markers that map between the two translocation breakpoints were recovered representing putative portions of the EDA gene.

Sequence of mouse glucose-6-phosphate dehydrogenase cDNA

A full-length mouse glucose-6-phosphate dehydrogenase (G6PD) cDNA has been isolated and sequenced, and the evolutionary conservation of many portions of the sequence has been verified by comparison with that of human and other sources.

Ribosomal DNA clusters in pulsed-field gel electrophoretic analysis of human acrocentric chromosomes

For determination of the extent to which ribosomal DNA (rDNA) is organized in tandemly repeated arrays, cellular DNA was digested with a restriction enzyme (EcoRV) that does not cut within the single 44-kb rDNA unit, and fragments separated by PFGE were hybridized to specific rDNA probes. A series of bands large enough to contain 15 to more than 30 rDNA repeat units was observed. In YACs containing cloned rDNA, however, such clusters were not observed, presumably because, as shown here for a clone starting with 1.5 tandem repeat units, there is a tendency for repeat units to delete out of the insert. By comparative gel electrophoretic analyses of DNAs from rodent hybrid cells containing singly isolated human chromosomes, most of the bands seen in total human DNA were assigned to at least one of the acrocentric chromosomes. Thus, large characteristic assemblies of DNA containing rDNA and lacking EcoRV sites were stable enough to be conserved in some human/rodent hybrid lines. When further digested with HindIII, which cuts rDNA at several points, the rDNA in each band yielded the expected fragments. If the large species consist completely of clusters of tandemly repeated rDNA units, they account for about half of the total cellular rDNA content estimated by saturation hybridization measurements.

Genome analysis and the human X chromosome

A unified genetic, physical, and functional map of the human X chromosome is being built through a concerted, international effort. About 40 percent of the 160 million base pairs of the X chromosome DNA have been cloned in overlapping, ordered contigs derived from yeast artificial chromosomes. This rapid progress toward a physical map is accelerating the identification of inherited disease genes, 26 of which are already cloned and more than 50 others regionally localized by linkage analysis. This article summarizes the mapping strategies now used and the impact of genome research on the understanding of X chromosome inactivation and X-linked diseases.

Mapping human chromosomes by walking with sequence-tagged sites from end fragments of yeast artificial chromosome inserts

Sequence-tagged sites (STSs) derived from end fragments of chromosome-specific yeast artificial chromosomes (YACs) can facilitate the assembly of an overlapping YAC/STS map. Contigs form rapidly by iteratively screening YAC collections with end-fragment STSs from YACs that have not yet been detected by any previous STS. The map is rendered rapidly useful during its assembly by incorporating supplementary STSs from genes and genetic linkage probes with known locations. Methods for the systematic development and testing of the end-fragments STSs are given here, and a group of 100 STSs is presented for the X chromosome. The mapping strategy is shown to be successful in simulations with portions of the X chromosome already largely mapped into overlapping YACs by other means.

A 530kb YAC contig tightly linked to the Friedreich ataxia locus contains five CpG clusters and a new highly polymorphic microsatellite

Friedreich ataxia (FA) is a severe autosomal recessive neurodegenerative disease. The defective gene has been previously assigned to chromosome 9q13-q21 by demonstration of tight linkage to the two independent loci D9S15 and D9S5. Linkage data indicate that FRDA is at less than 1 cM from both markers. Previous physical mapping has shown that probes defining D9S15 (MCT112) and D9S5 (26P) are less than 260 kb apart and are surrounded by at least six CpG clusters within 450 kb, which might indicate the presence of "candidate" genes for FA. We isolated and characterized a 530 kb YAC (yeast artificial chromosome) contig that contains five of the CpG clusters. The YACs were used to search for new polymorphic markers needed to map FRDA precisely with respect to the cloned segment. In particular, we found a (CA)n microsatellite polymorphism, GS4, that detects 13 alleles with a PIC value of 0.83 and allows the definition of haplotypes extending over 310 kb when used in combination with polymorphic markers at D9S5 and D9S15.

A 1.6-Mb contig of yeast artificial chromosomes around the human factor VIII gene reveals three regions homologous to probes for the DXS115 locus and two for the DXYS64 locus

Two yeast artificial chromosome (YAC) libraries were screened for probes in Xq28, around the gene for coagulation factor VIII (F8). A set of 30 YACs were recovered and assembled into a contig spanning at least 1.6 Mb from the DXYS64 locus to the glucose 6-phosphate dehydrogenase gene (G6PD). Overlaps among the YACs were determined by several fingerprinting techniques and by additional probes generated from YAC inserts by using Alu-vector or ligation-mediated PCR. Analysis of more than 30 probes and sequence-tagged sites (STSs) made from the region revealed the presence of several homologous genomic segments. For example, a probe for the DXYS64 locus, which maps less than 500 kb 5' of F8, detects a similar but not identical locus between F8 and G6PD. Also, a probe for the DXS115 locus detects at least three identical copies in this region, one in intron 22 of F8 and at least two more, which are upstream of the 5' end of the gene. Comparisons of genomic and YAC DNA suggest that the multiple loci are not created artifactually during cloning but reflect the structure of uncloned human DNA. On the basis of these data, the most likely order for the loci analyzed is tel-DXYS61-DXYS64-(DXS115-3-DXS115-2)-5'F8-(D XS115-1)-3'F8-G6PD.

Conserved sequence-tagged sites: a phylogenetic approach to genome mapping

Cognate sites in genomes that diverged approximately 100 million years ago can be detected by PCR assays based on primer pairs from unique sequences. The great majority of such syntenically equivalent sequence-tagged sites (STSs) from human DNA can be used to assemble and format corresponding maps for other primates, and some based on gene sequences are shown to be useful for mouse and rat as well. Universal genomic mapping strategies may be possible by using sets of STSs common to many mammalian species.

Distribution of moderately repetitive sequences pTR5 and LF1 in Xq24-q28 human DNA and their use in assembling YAC contigs

Xq24-q28 DNA, from a hamster/human hybrid cell containing only that portion of the human X chromosome, was found to contain 56 TaqI restriction fragments that hybridized to the moderately repetitive sequence pTR5. Using the pTR5 sequence as a probe in colony hybridization, 136 cognate yeast artificial chromosome (YAC) clones were detected among a collection of 820 containing about three genomic equivalents of the Xq24-q28 DNA. The YACs were then grouped into 48 contigs and single clones containing one or more of the TaqI fragments. Overlaps were confirmed both by fingerprinting YACs with AluI and L1 probes and by additional information. A less complete analysis was also carried out with a second moderately repetitive sequence, LF1, and some smaller contigs were merged into larger ones. Moderately repetitive sequences can thus be used as probes for multiple loci in single hybridization experiments and can help to organize and confirm YAC overlaps during the development of maps with long-range contiguity.

The iduronate sulfatase gene: isolation of a 1.2-Mb YAC contig spanning the entire gene and identification of heterogeneous deletions in patients with Hunter syndrome

A recently isolated cDNA clone from the iduronate sulfatase (IDS) gene has been used both to seed a contig of overlapping yeast artificial chromosomes (YACs) and to investigate the molecular defect in patients with Hunter syndrome (MPS II). Six YAC clones were found to span the IDS gene, and those and 14 other YACs were assembled into a 1.2-Mb contig around the gene in Xq27-q28. The physical map of the region identifies several putative CpG islands, suggesting the presence of other genes in the vicinity. DNA from a patient with a translocation breakpoint in the gene also permitted the orientation of the contig in the chromosome. Southern analysis of DNA from 25 unrelated Italian Hunter syndrome patients revealed 4 with deletions or rearrangements in the IDS gene.

Yeast artificial chromosomes spanning 8 megabases and 10-15 centimorgans of human cytogenetic band Xq26

A successful test is reported to generate long-range contiguous coverage of DNA from a human cytogenetic band in overlapping yeast artificial chromosomes (YACs). Seed YACs in band Xq26 were recovered from a targeted library of clones from Xq24-q28 with 14 probes, including probes for the hypoxanthine guanine phosphoribosyltransferase- and coagulation factor IX-encoding genes and nine probes used in linkage mapping. Neighboring YACs were then identified by 25 "walking" steps with end-clones, and the content of 71 probes in cognate YACs was verified by further hybridization analyses. The resultant contig extends across 8 million base pairs, including most of band Xq26, with an order of markers consistent with linkage data. YAC-based mapping, thus, permits steps toward a fully integrated physical and genetic map and is probably adequate to sustain most of the human genome project.

Yeast artificial chromosome-based genome mapping: some lessons from Xq24-q28

Yeast artificial chromosomes (YACs) have recently provided a potential route to long-range coverage of complex genomes in contiguous cloned DNA. In a pilot project for 50 Mb (1.5% of the human genome), a variety of techniques have been applied to assemble Xq24-q28 YAC contigs up to 8 Mb in length and assess their quality. The results indicate the relative strength of several approaches and support the adequacy of YAC-based methods for mapping the human genome.

Vectors for inserting selectable markers in vector arms and human DNA inserts of yeast artificial chromosomes (YACs)

To facilitate studies of gene expression and homologous recombination, plasmids have been developed which permit the insertion of neomycin resistance-encoding gene (NmR) into either the human DNA insert or the vector arm of a yeast artificial chromosome (YAC). To integrate into the YAC arm, the plasmid pRV1 contains a LYS2 (encoding alpha-aminoadipate reductase) gene for selection in the yeast host, and a NmR gene for subsequent selection after transfection of mammalian cells. These two sequences are bracketed by fragments of the URA3 gene (encoding orotidine-5'-phosphate decarboxylase) that can disrupt the URA3 gene in the YAC arm by homologous recombination in yeast. To integrate a selectable marker into the insert, the plasmid pRV2 contains a NmR gene and an intact copy of the URA3 gene, bracketed by segments of an L1 (LINEs) repetitive element. In this case, the vector has been designed for use with YACs that have already been fitted in the vector arm with a different marker (i.e., TK) that has disrupted the URA3 gene in the vector arm. Selection is for the restoration of URA3 gene activity attendant on recombination into an L1 element in the YAC insert. Use of the vectors is illustrated with a YAC clone containing ribosomal DNA.

Sequence of human glucose-6-phosphate dehydrogenase cloned in plasmids and a yeast artificial chromosome

The sequence of 20,114 bp of DNA including the human glucose-6-phosphate dehydrogenase (G6PD) gene was determined. The region included a prominent CpG island, starting about 680 nucleotides upstream of the transcription start site, extending about 1050 nucleotides downstream of the start site, and ending just at the start of the first intron. The transcribed region from the start site to the poly(A) addition site covers 15,860 bp. The sequence of the 13 exons agreed with published cDNA sequence and for the 11 exons tested, with the corresponding sequence in a yeast artificial chromosome (YAC). The latter confirms YAC cloning fidelity at the DNA sequence level. Sixteen Alu sequences constitute 24% of the total sequence tract. Four were outside the borders of the mRNA transcript of the gene; all the others were found in a large (9858 bp) intron between exons 2 and 3. Two Alu clusters each contain Alus lying between the monomers of another.