Interchromosomal Crossover in Human Cells Is Associated with Long Gene Conversion Tracts

Abstract

Crossovers have rarely been observed in specific association with interchromosomal gene conversion in mammalian cells. In this investigation two isogenic human B-lymphoblastoid cell lines, TI-112 and TSCER2, were used to select for I-SceI-induced gene conversions that restored function at the selectable thymidine kinase locus. Additionally, a haplotype linkage analysis methodology enabled the rigorous detection of all crossover-associated convertants, whether or not they exhibited loss of heterozygosity. This methodology also permitted characterization of conversion tract length and structure. In TI-112, gene conversion tracts were required to be complex in tract structure and at least 7.0 kb in order to be selectable. The results demonstrated that 85% (39/46) of TI-112 convertants extended more than 11.2 kb and 48% also exhibited a crossover, suggesting a mechanistic link between long tracts and crossover. In contrast, continuous tracts as short as 98 bp are selectable in TSCER2, although selectable gene conversion tracts could include a wide range of lengths. Indeed, only 16% (14/95) of TSCER2 convertants were crossover associated, further suggesting a link between long tracts and crossover. Overall, these results demonstrate that gene conversion tracts can be long in human cells and that crossovers are observable when long tracts are recoverable.In mammalian cells, gene conversion without associated crossover has been widely reported for both intrachromosomal and interchromosomal recombining substrates (13,51,62,80,87). However, most studies to date have not succeeded in associating interchromosomal crossovers with specific conversion tracts (37,67,71,80,83); reports of crossovers are rare (7,72). Nevertheless, presumptive interchromosomal crossovers, in the form of chromosome-scale loss of heterozygosity (LOH), are commonly observed in mammalian cells. LOH is the major contributor to the mutational spectrum in mice and humans (31,33,47,49,78), notably at tumor suppressor loci (3,11,15,25,34,35,40,68,88,89). Crossover products have also been associated with intrachromosomal recombination between repeat sequences in mammalian cells (12,30,52,75), and similar outcomes are observed during the integration of plasmids into host cell chromosomes (9,48,90). Additional evidence for crossovers in mammalian cells is provided by the relatively frequent occurrence of sister chromatid exchange (94), although these products may also result from other mechanisms (56). Since the occurrence of crossovers in mammalian cells has been clearly demonstrated, the absence of crossovers associated with specific interchromosomal gene conversion tracts requires further explanation.Several mechanisms have been proposed to explain the preference for gene conversion without associated crossover in eukaryotic organisms. Crossover is believed to require the formation of Holliday intermediates (17,36,44,76,77) and their resolution through structure-specific endonuclease scissions (79). Theoretically, half of Holliday junctions that are resol...

Interhomolog recombination and loss of heterozygosity in wild-type and Bloom syndrome helicase (BLM)-deficient mammalian cells

“…A previous study examining DSB-induced interhomolog recombination events in wild-type human lymphoblastoid cells reported that COs comprised a somewhat greater fraction of recombinants than we observed, although it is probable that all interhomolog NCOs could not be scored, given the requirement for coconversion of markers (8). An NCO bias also is observed for meiotic interhomolog recombination, but this bias is less pronounced since COs are estimated to comprise ∼10% of events (see ref.…”

“…COs in wild-type human cells similarly were found to be associated with NCOs (8). These results raise two questions: Why are COs frequently associated with a second DSB repair event, and why does this second event occur exclusively by HR?…”

“…DSBs induce mitotic interhomolog HR in mammalian cells (7), but CO-LOH occurs infrequently during DSB repair (5,8), suggesting that factors actively suppress crossing over. Candidate factors include the RecQ helicase deficient in cells from Bloom syndrome patients (BLM) (9).…”

Analysis of Spontaneous Gene Conversion Tracts within and between Mammalian Chromosomes

“…Studies of DSB-induced mitotic intra-and interchromosomal recombination as well as repair of an induced chromosomal DSB via transfected plasmid sequences reveal that gene conversion tracts are generally very short, ranging from a few nucleotides to b 200 bp, [13][14][15][16] although with longer regions of shared homology can occasionally extend to 3 kb. 17 In DSB-induced interchromosomal recombination, Neuwirth et al 18 show that under appropriate selective conditions, long gene conversion tracts (up to 11 kb) can be observed. Although gene conversion occurs spontaneously during mitotic intra-and interchromosomal recombination [19][20][21][22][23][24] and gene targeting, [25][26][27] as well as meiotically in mouse spermatids, 28,29 most events are exceedingly rare, complicating measurement of gene conversion tract length.…”

Assessment of Anti-recombination and Double-strand Break-induced Gene Conversion in Human Cells by a Chromosomal Reporter

“…At regions surrounding a DSB, gene conversion generally occurs as a result of heteroduplex processing by the mismatch repair (MMR) pathway. A number of studies have demonstrated that gene conversion tracts can range from several hundred base pairs to 11.2 kb in mammalian cells (3,6,7). This underscores a necessity for a better understanding of how cells regulate gene conversions at various regions in reference to the locations of DSBs.…”