Propagation of some human DNA sequences in bacteriophage lambda vectors requires mutant Escherichia coli hosts.

Abstract

The growth of clones of human genomic DNA fragments in a bacteriophage X vector has been examined in a number of different Escherichia coli hosts. A large proportion (8.9%) of the phages carrying different fragments of the human genome fail to grow on standard rec' hosts but will grow on hosts carrying mutations in the recB, recC, and sbcB genes. Heteroduplex analysis in the electron microscope of DNA from four of these phages revealed substantial secondary structure, including snap-back regions 200-500 base pairs in length. Such structures were not found in phages from the same DNA library that grow in rec' hosts. Recombinant DNA libraries in bacteriophage X vectors are generally found to contain most of the sequences of the genomes from which they are derived (1, 2). Nevertheless, anecdotal reports of sequences that cannot be found in libraries of eukaryotic genomes are fairly common. One striking example is the hypervariable region 3' to the a-globin locus, which has resisted strenuous efforts at cloning (S. Goodbourn and S. Orkin, personal communications). Another example is the junction fragment from a translocation between chromosomes 6 and 10 in a murine plasmacytoma (3). A related phenomenon is the observation of discrepancies between cloned and genomic sequences that are the result of deletions occurring during cloning of mammalian DNA (3)(4)(5). Only some of these deletions can be prevented by growth on a recA host.For the past few years we have been studying a highly polymorphic locus in humans, now known as D14S1, which was identified by hybridization of gel transfers of genomic DNA using an adjacent single-copy fragment as probe (6-8).Efforts to obtain clones of the sequences surrounding and including the polymorphic site itself from the Maniatis human genomic library (2) and from libraries made in our laboratory, using conventional cloning conditions as well as hosts carrying mutations in the recA gene, were all unsuccessful. Given the rarity in general of highly polymorphic DNA sequences in the human genome and the striking difficulty in recovering the variable segments in two of the most studied ones (a globin and D14S1), we became concerned that there might be a systematic difficulty in cloning certain human DNA sequences. Further, we speculated that the difficulty might sometimes be associated with a high degree of polymorphism.Recently, Leach and Stahl (9) reported that inverted repetitions cannot be cloned into phage X vectors unless mutant hosts are used. We report here that such a host, which contains mutant recB, recC, and sbcB genes, will propagate many phage X clones containing random human DNA fragments, including D14S1, that cannot be grown on the Escherichia coli hosts commonly used in the selection and propagation of recombinant DNA genome libraries. MATERIALS AND METHODSBacterial Strains. With the exception of strains BNN45 and SF8, all strains used are derivatives of the E. coli K-12 strain AB1157; they have the following markers in common: leu6 aral4 his4 thrl thil lac Y1...

Structural organization and functional analysis of centromeric DNA in the fission yeast Schizosaccharomyces pombe.

“…Since most inverted repeat sequences cannot be cloned in E. coli (11,22), it is possible that heterochromatic centromeric DNA sequences in other organisms are similarly organized, and recombinant plasmids or phage in genomic libraries would lack cloned segments spanning the centers of the repeats. In that regard, it has been reported that a sizable fraction (about 9%) of the human genome cannot be cloned in standard rec+ E. coli hosts (39). We have been unable to generate E. coli clones bearing recombinant cosmids or plasmids that contain DNA sequences spanning the centromeric inverted repeat, even using mutant strains that are reported to propagate small inverted repeat sequences (22,39).…”

“…In that regard, it has been reported that a sizable fraction (about 9%) of the human genome cannot be cloned in standard rec+ E. coli hosts (39). We have been unable to generate E. coli clones bearing recombinant cosmids or plasmids that contain DNA sequences spanning the centromeric inverted repeat, even using mutant strains that are reported to propagate small inverted repeat sequences (22,39). It seems likely that recent technology developed to clone long segments of DNA in S. cerevisiae (4) will be successfully applied to the problem of characterizing the sequence arrangements of repetitive DNA elements in the centromere regions of higher eucaryotic chromosomes.…”

Wolffish antifreeze protein genes are primarily organized as tandem repeats that each contain two genes in inverted orientation.

“…One clone, XWr6, contained a variant symmetric restriction pattern which appeared to be the result of a recombination event between the "left gene" of one repeat and the "right gene" of another to generate an inverted repeat containing two "right" genes. The combination of direct and inverted repeats in these clones might explain why major-component genes are underrepresented in library screens relative to their copy number in the genome (23,25).…”

RecA-independent high-frequency deletion of recombinant cosmid DNA in Escherichia coli

“…In both ;1 and plasmid cloning, a recB sbcB host and a recB recC sbcB host have been reported to prevent the deletions of some kinds of recombinant DNAs that were not stably maintained in usual rec+ or recA hosts (Collins et al, 1982;Leach & Stahl, 1983;Wyman et al, 1985;Nader et al, 1985). Therefore, we examined the effects of such E. coli hosts on the cosmid deletion events.…”