Double-strand breaks at an initiation site for meiotic gene conversion

Synchronized meiosis and recombination in fission yeast: observations with pat1-114 diploid cells

The mutation pat1-114 has been used to synchronize meiosis in the fission yeast Schizosaccharomyces pombe. We have investigated several aspects of such synchronized meiotic cultures. In both pat1-114 and pat1+ diploids, meiotic landmark events are initiated at the same time after meiosis induction, but synchrony is much more pronounced in the pat1-114-driven meiosis. Commitment to recombination and to meiosis have been timed at 2 h after meiotic induction. Due to a seven-fold reduction of intragenic recombination frequency in the ade6 region of pat1-114 diploids, physical analysis of recombination has not been possible. We have distinguished three factors that influence intragenic recombination frequencies: temperature, azygotic versus zygotic meiosis, and the nature of the pat1 allele. Differences and similarities in the timing of meiotic landmarks in S. cerevisiae and S. pombe are discussed.

Characterization of recombination intermediates from DNA injected into Xenopus laevis oocytes: evidence for a nonconservative mechanism of homologous recombination

Homologous recombination between DNA molecules injected into Xenopus laevis oocyte nuclei is extremely efficient if injected molecules have overlapping homologous ends. Earlier work demonstrated that ends of linear molecules are degraded by a 5'----3' exonuclease activity, yielding 3' tails that participate in recombination. Here, we have characterized intermediates further advanced along the recombination pathway. The intermediates were identified by their unique electrophoretic and kinetic properties. Two-dimensional gel electrophoresis and hybridization with oligonucleotide probes showed that the intermediates had heteroduplex junctions within their homologous overlaps in which strands ending 3' were full length and those ending 5' were shortened. Additional characterization suggested that these intermediates had formed by the annealing of complementary 3' tails. Annealed junctions made in vitro were rapidly processed to products, indicating that they are on the normal recombination pathway. These results support a nonconservative, single-strand annealing mode of recombination. This recombination mechanism appears to be shared by many organisms, including bacteria, fungi, plants, and mammals.

Involvement of single-stranded tails in homologous recombination of DNA injected into Xenopus laevis oocyte nuclei

Homologous recombination of DNA molecules injected into Xenopus laevis oocyte nuclei is extremely efficient when those molecules are linear and have overlapping homologous ends. It was previously shown that a 5'----3' exonuclease activity in oocytes attacks injected linear DNAs and leaves them with single-stranded 3' tails. We tested the hypothesis that such tailed molecules are early intermediates on the pathway to recombination products. Substrates with 3' tails were made in vitro and injected into oocytes, where they recombined rapidly and efficiently. In experiments with mixed substrates, molecules with 3' tails entered recombination intermediates and products more rapidly than did molecules with flush ends. Molecules endowed in vitro with 5' tails also recombined efficiently in oocytes, but their rate was not faster than for flush-ended substrates. In most cases, the 5' tails served as templates for resynthesis of the 3' strands, regenerating duplex ends which then entered the normal recombination pathway. In oocytes from one animal, some of the 5' tails were removed, and this was exacerbated when resynthesis was partially blocked. Analysis by two-dimensional gel electrophoresis of recombination intermediates from 5'-tailed substrates confirmed that they had acquired 3' tails as a result of the action of the 5'----3' exonuclease. These results demonstrate that homologous recombination in oocytes proceeds via a pathway that involves single-stranded 3' tails. Molecular models incorporating this feature are discussed.

DNase I-hypersensitive sites and transcription factor-binding motifs within the mouse E beta meiotic recombination hot spot

The second intron of the E beta gene in the mouse major histocompatibility complex is the site of a meiotic recombination hot spot. We detected two DNase I-hypersensitive sites in this intron in meiotic cells isolated from mouse testes. One site appears to be constitutive and is found in other tissues regardless of whether or not they express the E beta gene. Near this hypersensitive site are potential binding motifs for H2TF1/KBF1, NF kappa B, and octamer transcription factors. Gel retardation studies with mouse lymphoma cell nuclear extracts confirmed that each of these motifs is capable of binding protein. The binding of transcription factors may contribute to the enhancement of recombination potential by altering chromatin structure and increasing the accessibility of the DNA to the recombination machinery.

The split-end model for homologous recombination at double-strand breaks and at Chi

In recent years two different styles of model for homologous recombination have been discussed, depending on whether or not the recombination event occurs in the vicinity of a double-strand break in DNA. The models of Holliday and Meselson and Radding exemplify those that do not involve a break whereas the model of Szostak et al is taken as an example of those that do. Recent advances in understanding a prototypic recombination system thought to promote exchange distant from DNA ends, at Chi sites, suggest a mechanism of initiation neither like Holliday/Meselson-Radding nor like Szostak et al. In those models, only one strand of DNA may invade a homologous DNA molecule. We propose a model for Chi in which exonuclease degrades DNA from a double-strand break to the Chi site; the exonuclease is converted into a helicase upon interaction with Chi; unwinding produces a recombinagenic split-end, and both 3'- and 5'-ending strands at the split-end are capable of invading a homologue. Different genetic consequences are proposed to result from invasion by each. We review evidence supporting the split-end model and suggest its application in at least some cases previously considered to proceed via the Meselson/Radding model and by the double-strand-break repair model of Szostak et al.

DNase I-hypersensitive sites and transcription factor-binding motifs within the mouse E beta meiotic recombination hot spot

The second intron of the E beta gene in the mouse major histocompatibility complex is the site of a meiotic recombination hot spot. We detected two DNase I-hypersensitive sites in this intron in meiotic cells isolated from mouse testes. One site appears to be constitutive and is found in other tissues regardless of whether or not they express the E beta gene. Near this hypersensitive site are potential binding motifs for H2TF1/KBF1, NF kappa B, and octamer transcription factors. Gel retardation studies with mouse lymphoma cell nuclear extracts confirmed that each of these motifs is capable of binding protein. The binding of transcription factors may contribute to the enhancement of recombination potential by altering chromatin structure and increasing the accessibility of the DNA to the recombination machinery.

A unique pathway of double-strand break repair operates in tandemly repeated genes

The RAD52 gene product of the yeast Saccharomyces cerevisiae is required for most spontaneous recombination and almost all double-strand break (DSB) repair. In contrast to recombination elsewhere in the genome, recombination in the ribosomal DNA (rDNA) array is RAD52 independent. To determine the fate of a DSB in the rDNA gene array, a cut site for the HO endonuclease was inserted into the rDNA in a strain containing an inducible HO gene. DSBs were efficiently repaired at this site, even in the absence of the RAD52 gene product. Efficient RAD52-independent DSB repair was also observed at another tandem gene array, CUP1, consisting of 18 repeat units. However, in a smaller CUP1 array, consisting of only three units, most DSBs (ca. 80%) were not repaired and resulted in cell death. All RAD52-independent DSB repair events examined resulted in the loss of one or more repeat units. We propose a model for DSB repair in repeated sequences involving the generation of single-stranded tails followed by reannealing.

Extensive 3'-overhanging, single-stranded DNA associated with the meiosis-specific double-strand breaks at the ARG4 recombination initiation site

Meiosis-specific double-strand breaks occur at the initiation site for meiotic gene conversion in the yeast ARG4 gene. Here we show that the break fragments end in extensive 3'-overhanging, single-stranded tails. The single-stranded tails very in length, generating a gradient of single-strandedness that parallels the gradient of gene conversion frequencies in ARG4. In strains carrying a rad50S mutation, which blocks meiotic recombination, the extensive single-stranded tails do not form, suggesting that their generation is an obligatory step in meiotic recombination. Using the rad50S mutant, we have mapped the site of the ARG4 break to a small region within the genetically defined recombination initiation site. These results strongly support the double-strand break model of meiotic recombination.

A unique pathway of double-strand break repair operates in tandemly repeated genes

The RAD52 gene product of the yeast Saccharomyces cerevisiae is required for most spontaneous recombination and almost all double-strand break (DSB) repair. In contrast to recombination elsewhere in the genome, recombination in the ribosomal DNA (rDNA) array is RAD52 independent. To determine the fate of a DSB in the rDNA gene array, a cut site for the HO endonuclease was inserted into the rDNA in a strain containing an inducible HO gene. DSBs were efficiently repaired at this site, even in the absence of the RAD52 gene product. Efficient RAD52-independent DSB repair was also observed at another tandem gene array, CUP1, consisting of 18 repeat units. However, in a smaller CUP1 array, consisting of only three units, most DSBs (ca. 80%) were not repaired and resulted in cell death. All RAD52-independent DSB repair events examined resulted in the loss of one or more repeat units. We propose a model for DSB repair in repeated sequences involving the generation of single-stranded tails followed by reannealing.

A poly(dA.dT) tract is a component of the recombination initiation site at the ARG4 locus in Saccharomyces cerevisiae

An initiation site for meiotic gene conversion is located in the promoter region of the ARG4 locus in Saccharomyces cerevisiae. We have tested the hypothesis that the initiation site is identical with the promoter by making a series of small deletions that remove specific promoter elements. Disruption of most promoter elements does not lower the level of gene conversion in ARG4, and analysis of RNA levels at the time of recombination in meiosis reveals no direct correlation between the level of ARG4 transcript and the level of gene conversion in ARG4. However, deletion of a tract of 14 A residues located at the peak of the gene conversion gradient decreases the number of gene conversion events stimulated by the initiation site to 25 to 35% of the normal level. We conclude that the poly(dA.dT) tract is responsible for most but not all of the high levels of meiotic gene conversion observed in ARG4.

A poly(dA.dT) tract is a component of the recombination initiation site at the ARG4 locus in Saccharomyces cerevisiae

An initiation site for meiotic gene conversion is located in the promoter region of the ARG4 locus in Saccharomyces cerevisiae. We have tested the hypothesis that the initiation site is identical with the promoter by making a series of small deletions that remove specific promoter elements. Disruption of most promoter elements does not lower the level of gene conversion in ARG4, and analysis of RNA levels at the time of recombination in meiosis reveals no direct correlation between the level of ARG4 transcript and the level of gene conversion in ARG4. However, deletion of a tract of 14 A residues located at the peak of the gene conversion gradient decreases the number of gene conversion events stimulated by the initiation site to 25 to 35% of the normal level. We conclude that the poly(dA.dT) tract is responsible for most but not all of the high levels of meiotic gene conversion observed in ARG4.

Chromosome synapsis and genetic recombination: their roles in meiotic chromosome segregation

Chromosome synapsis and genetic recombination ensure the faithful segregation of chromosomes at meiosis I by establishing physical connections between homologs. Recent observations suggest that recombination may also play a role in the homology search process that precedes synapsis.

The role of recombinational hotspots in genome instability in mammalian cells

Genome instability has been associated with progression of transformed cells to high tumorigenicity. Although genome instability may result from a variety of factors, some studies suggest that DNA in the region of a chromosome rearrangement can subsequently have much higher rates of DNA deletions or gene amplification. One approach to studying the factors that produce these high rates of DNA rearrangement is by analysis of unstable integration sites for DNA transfected into mammalian cells. Integrated sequences commonly show a temporary instability, and at rare locations this instability is continuous and can be observed even after multiple subclonings. These continuously unstable locations undergo DNA amplification of both the integrated sequences and the surrounding cell DNA, and it can occur either at the original site or on episomes after looping out from the chromosome. Because the adjacent cell DNA plays a role in this instability, and the region can be shown to be stable before integration, the results indicate that these recombinational hotspots can be formed de novo by the process of integration. Current studies are attempting to determine which sequences are responsible for the high rates of recombination and whether similar types of event are involved in the instability associated with endogenous cellular genes in cancer cells.

Detection of heteroduplex DNA molecules among the products of Saccharomyces cerevisiae meiosis

We have used denaturant-gel electrophoresis to provide a physical demonstration of heteroduplex DNA in the products of yeast meiosis. We examined heteroduplex formation at arg4-nsp, a G.C----C.G transversion that displays a moderately high level of postmeiotic segregation. Of the two possible arg4-nsp/ARG4 mismatches (G.G and C.C), only C.C was detected in spores from mismatch repair-competent (Pms1+) diploids. In contrast, C.C and G.G were present at nearly equal levels in spores from Pms1- diploids. These results confirm previous suggestions that postmeiotic segregation spores contain heteroduplex DNA at the site of the marker in question, that C.C is repaired less frequently than is G.G, and that the PMS1 gene product plays a role in mismatch correction. Combined with the observation that Pms1+ ARG4/arg4-nsp diploids produce 3 times more 3+:5m (wildtype:mutant) tetrads (+, +, +/m, m) than 5+:3m tetrads (+, +/m, m, m), these results indicate that, during meiosis, formation of heteroduplex DNA at ARG4 involves preferential transfer of the sense (nontranscribed) strand of the DNA duplex.

Meiotic recombination between dispersed repeated genes is associated with heteroduplex formation

In Saccharomyces cerevisiae, recombination events occurring between allelic genes located on homologous chromosomes are often associated with heteroduplex formation. We found that recombination events between repeated genes on nonhomologous chromosomes (ectopic events) are also associated with the formation of heteroduplexes, indicating that classical and ectopic recombination events involve similar mechanisms.

Meiotic recombination between dispersed repeated genes is associated with heteroduplex formation

In Saccharomyces cerevisiae, recombination events occurring between allelic genes located on homologous chromosomes are often associated with heteroduplex formation. We found that recombination events between repeated genes on nonhomologous chromosomes (ectopic events) are also associated with the formation of heteroduplexes, indicating that classical and ectopic recombination events involve similar mechanisms.

A pathway for generation and processing of double-strand breaks during meiotic recombination in S. cerevisiae

We have identified and analyzed a meiotic reciprocal recombination hot spot in S. cerevisiae. We find that double-strand breaks occur at two specific sites associated with the hot spot and that occurrence of these breaks depends upon meiotic recombination functions RAD50 and SPO11. Furthermore, these breaks occur in a processed form in wild-type cells and in a discrete, unprocessed form in certain nonnull rad50 mutants, rad50S, which block meiotic prophase at an intermediate stage. The breaks observed in wild-type cells are similar to those identified independently at another recombination hot spot, ARG4. We show here that the breaks at ARG4 also occur in discrete form in rad50S mutants. Occurrence of breaks in rad50S mutants is also dependent upon SPO11 function. These observations provide additional evidence that double-strand breaks are a prominent feature of meiotic recombination in yeast. More importantly, these observations begin to define a pathway for the physical changes in DNA that lead to recombination and to define the roles of meiotic recombination functions in that pathway.

Analysis of wild-type and rad50 mutants of yeast suggests an intimate relationship between meiotic chromosome synapsis and recombination

The RAD50 gene of S. cerevisiae is required during meiosis for both recombination and chromosome synapsis and is also required for repair of double strand breaks during vegetative growth. We present below the isolation and analysis of several types of rad50 mutants. We show that null mutations block both meiotic recombination and formation of synaptonemal complex (SC) at early stages, while nonnull mutations block both processes at intermediate stages. These observations suggest that recombination and SC formation involve a series of intimately related events. Furthermore, all rad50 mutants block formation of tripartite SC structure but permit other aspects of SC development, i.e., formation of axial cores. In light of this and other observations, the meiotic and mitotic defects of rad50 mutants can be accounted for economically by the proposal that meiotic recombination, meiotic chromosome pairing, and vegetative DNA repair all use a common chromosomal homology search that involves RAD50 function.

Influence of cellular sequences on instability of plasmid integration sites in human cells

To learn more about mechanisms of genome instability in human cells, I investigated DNA sequences that promote high rates of recombination by analyzing rare unstable plasmid integration sites in simian virus 40-transformed human fibroblasts. Previous studies had hypothesized that rearrangement or loss of integrated sequences could be attributed to adjacent cellular DNA. Consistent with this interpretation, a cloned fragment containing both the integrated plasmid and 2.0 kb of adjacent cell DNA from one such unstable integration site in the cell line LM205 demonstrated a much higher incidence of rearrangements when integrated into other chromosome locations than did the original plasmid. To further test this hypothesis, portions of cellular DNA from this region were integrated in duplicate in other locations to determine their ability to promote restriction-fragment-length polymorphism, an indicator of high rates of homologous recombination. Although two types of instability were observed, neither could be attributed solely to the cell sequences being tested in the plasmid. The first type of instability was a transient deletion or amplification of the plasmid DNA soon after integration, which appeared to be a general phenomenon often associated with any type of newly integrated sequence. A second type of instability continued indefinitely for many cell generations, as did that observed in cell line LM205. Because this was rare (one of 78 clones tested), it could not be attributed solely to cell sequences contained within the plasmid. However, the rearrangements in this cell clone occurred exclusively within the cell DNA adjacent to the integration site, again suggesting a role for cis-acting cell sequences in this process. The inability to identify specific cell sequences responsible for instability may therefore indicate that a complex combination of sequences is involved, possibly within both the plasmid and cell DNA.

Chicken IgL variable region gene conversions display pseudogene donor preference and 5' to 3' polarity

Chicken immunoglobulin variable region diversity is generated during B-cell development in the bursa of Fabricius by intrachromosomal gene conversion, resulting in the replacement of sequence blocks within the unique rearranged VL1 and VH1 genes with homologous sequences derived from V region pseudogene segments (psi V). In this report, the nucleotide sequences of 217 gene conversion events in 52 random IgL clones were analyzed to characterize the molecular mechanism of gene conversion. The frequency of psi VL usage as gene conversion donors is shown to depend on the proximity of the psi VL segment to VL1, extent of homology with VL1, and relative orientation of the psi VL segments. Gene conversion events are not observed in the 5' region of homology between psi VL segments and VL1, but are distributed throughout the remainder of the VL1 exon. The 5' ends of individual gene conversion events always begin in regions of homology between the donor psi VL and recipient VL1 gene, whereas the 3' ends can occur in regions of nonhomology and often have nucleotide insertions or deletions. These results suggest a 5' to 3' polarity in the gene conversion mechanism. The implications of our data are discussed in relation to current molecular models of gene conversion.

Degradation of linear DNA by a strand-specific exonuclease activity in Xenopus laevis oocytes

Linear DNA injected into Xenopus laevis oocyte nuclei recombines with high efficiency if homologous sequences are present at overlapping molecular ends. We found that injected linear DNA was degraded by a 5'----3' strand-specific exonuclease activity during incubation in the oocyte nucleus to leave a heterogeneous population of 3'-tailed molecules. Decreasing the concentration of DNA injected increased the heterogeneity and the average rate of degradation. The 3' tails created were relatively stable; among molecules persisting after overnight incubation, many had 3' tails intact to within 10 bases of the original ends. DNA molecules that were efficient substrates for homologous recombination in oocytes were also partially degraded, leaving 3' tails. We found no evidence for other potent nuclease activities. If molecules with recessed 3'-OH ends were injected, endogenous polymerase efficiently resynthesized complementary strands before degradation of the 5' tails occurred. 3'-tailed molecules are plausible intermediates in the initiation of homologous recombination events in Xenopus oocyte nuclei.

Mutations in the RB1 gene and their effects on transcription

Inactivation of both alleles of the RB1 gene during normal retinal development initiates the formation of a retinoblastoma (RB) tumor. To identify the mutations which inactivate RB1, 21 RB tumors isolated from 19 patients were analyzed with the polymerase chain reaction or an RNase protection assay or both. Mutations were identified in 13 of 21 RB tumors; in 8 tumors, the precise errors in nucleotide sequence were characterized. Each of four germ line mutations involved a small deletion or duplication, while three somatic mutations were point mutations leading to splice alterations and loss of an exon from the mature RB1 mRNA. We were unable to detect expression of the mutant allele in lymphoblasts of three bilaterally affected patients, although the mutation was present in the genomic DNA and transcripts containing the mutations were obvious in the RB tumors in the absence of a normal RB1 allele. The variations in the level of expression of mutant transcripts suggest deregulation of RB1 transcription in the absence of a functional RB1 gene product.

Genomic sequence and organization of rabbit Tcrb constant region genes

We previously reported that some rabbits have three different copies of T-cell receptor b (Tcrb) constant region genes unlike man, mice, and rats who generally have two copies. Two of these C beta genes were found on an approximately 14 kilobase (kb) and one on an approximately 6 kb Eco RI fragment. The gene on the 6 kb fragment is of beta 2 type. A previously described portion of the 14 kb fragment appeared to have sequences characteristic of C beta 1. We have now shown that the 6 kb fragment is adjacent to and 3' of the 14 kb fragment. Furthermore, the second linked sequence of C beta gene present on the 14 kb fragment resembles to a large extent the C beta 2 gene present on the 6 kb fragment. Moreover, this second C beta gene has a 5' cluster of J beta sequences resembling J beta 2 of other species. However, exon 4 and the 3' untranslated region (3'UT) are of the beta 1 type. Mapping studies using Southern analyses of both genomic DNA and the 14 kb clone have identified another cluster of J beta 2 sequences 5' of the third tandem C beta 2 gene present on the 6 kb Eco RI fragment. Thus, the second gene on the 14 kb fragment appears to be a chimeric genomic Tcrb gene that may have arisen by an unequal crossing-over event analogous to that which may have deleted C beta 1, D beta 2, and J beta 2 in NZW mice.

Mutations in the RB1 gene and their effects on transcription

Inactivation of both alleles of the RB1 gene during normal retinal development initiates the formation of a retinoblastoma (RB) tumor. To identify the mutations which inactivate RB1, 21 RB tumors isolated from 19 patients were analyzed with the polymerase chain reaction or an RNase protection assay or both. Mutations were identified in 13 of 21 RB tumors; in 8 tumors, the precise errors in nucleotide sequence were characterized. Each of four germ line mutations involved a small deletion or duplication, while three somatic mutations were point mutations leading to splice alterations and loss of an exon from the mature RB1 mRNA. We were unable to detect expression of the mutant allele in lymphoblasts of three bilaterally affected patients, although the mutation was present in the genomic DNA and transcripts containing the mutations were obvious in the RB tumors in the absence of a normal RB1 allele. The variations in the level of expression of mutant transcripts suggest deregulation of RB1 transcription in the absence of a functional RB1 gene product.

Degradation of linear DNA by a strand-specific exonuclease activity in Xenopus laevis oocytes

Linear DNA injected into Xenopus laevis oocyte nuclei recombines with high efficiency if homologous sequences are present at overlapping molecular ends. We found that injected linear DNA was degraded by a 5'----3' strand-specific exonuclease activity during incubation in the oocyte nucleus to leave a heterogeneous population of 3'-tailed molecules. Decreasing the concentration of DNA injected increased the heterogeneity and the average rate of degradation. The 3' tails created were relatively stable; among molecules persisting after overnight incubation, many had 3' tails intact to within 10 bases of the original ends. DNA molecules that were efficient substrates for homologous recombination in oocytes were also partially degraded, leaving 3' tails. We found no evidence for other potent nuclease activities. If molecules with recessed 3'-OH ends were injected, endogenous polymerase efficiently resynthesized complementary strands before degradation of the 5' tails occurred. 3'-tailed molecules are plausible intermediates in the initiation of homologous recombination events in Xenopus oocyte nuclei.

A DNA double chain break stimulates triparental recombination in Saccharomyces cerevisiae

Mitotic recombination between his3 heteroalleles on heterologous chromosomes is stimulated by a DNA double chain break delivered in vivo at a site 8.6 kilobase pairs distant from one his3 allele and unlinked to the other. The induced recombination at his3 is accompanied by gap repair at the break site using the uncut homolog as a template. The DNA between the break site and his3 is not deleted in most of the His+ recombinants.