Similar effects of Brca2 truncation and Rad51 paralog deficiency on immunoglobulin V gene diversification in DT40 cells support an early role for Rad51 paralogs in homologous recombination

Abstract

BRCA2 is a tumor suppressor gene that is linked to hereditary breast and ovarian cancer. Although the Brca2 protein participates in homologous DNA recombination (HR), its precise role remains unclear. From chicken DT40 cells, we generated BRCA2 gene-deficient cells which harbor a truncation at the 3' end of the BRC3 repeat (brca2tr). Comparison of the characteristics of brca2tr cells with those of other HR-deficient DT40 clones revealed marked similarities with rad51 paralog mutants (rad51b, rad51c, rad51d, xrcc2, or xrcc3 cells). The phenotypic similarities include a shift from HR-mediated diversification to single-nucleotide substitutions in the immunoglobulin variable gene segment and the partial reversion of this shift by overexpression of Rad51. Although recent evidence supports at least Xrcc3 and Rad51C playing a role late in HR, our data suggest that Brca2 and the Rad51 paralogs may also contribute to HR at the same early step, with their loss resulting in the stimulation of an alternative, error-prone repair pathway.

FIG.1.

Gene targeting of the BRCA2 loci. (A) Experimental strategy. (B) Schematic representation of part of the BRCA2 locus, the configuration of the targeted alleles, and the gene-disrupted allele after treatment with TAM. The BRCA2 conditional null construct, which contains a bsr selection marker gene flanked by two loxP signals (ploxPbsr) in the upstream of the BRCA2 locus and the third loxP signal in intron 2, was used to target the BRCA2 locus. The promoter and exons 1 and 2, which are flanked by loxP sites, would be deleted upon exposure of the cells to TAM. Knock-in of the BRC truncation construct inserts a termination codon in the 5′ end of the BRC4 repeat, resulting in deletion of nucleotides 4369 to 9955 of the coding sequences of the chicken BRCA2 gene (54). Of note, since BRC3 sequences are not conserved in the chicken BRCA2 gene, the brca2tr cells would have only two functional BRC repeats, BRC1 and BRC2. S indicates relevant SacI sites. The open boxes and arrowheads represent the exons and loxP signals, respectively. (C) Western blot analysis of whole-cell extracts of cell lines with the indicated genotype. Each lane contains 200 μg of protein. The wild-type and brca2tr mutant cells show 340- and 160-kDa bands, respectively. (D) Western blot analysis of gene-disrupted clones that carry the human Rad51 transgene. Human Rad51 migrates slightly faster than the chicken counterpart in sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis. The same membrane was probed with the anti-chicken Ku70 antiserum as a control. WT, wild type.

FIG. 2.

Growth rate and viability of brca2tr cells. (A) The relative rate of cell growth per day is plotted for the indicated genotypes. (B and C) Sensitivity of wild-type and brca2tr cells to DNA-damaging agents. The fractions of surviving colonies after the indicated treatment of cells compared with untreated controls of the same genotype are shown on the y axis on a logarithmic scale. (B) Gamma rays; (C) cisplatin. The plating efficiency was 20 to 30% for brca2tr cells and 100% for wild-type cells. The dose of ¹³⁷Cs gamma rays and concentration of cisplatin are displayed on the x axis on a linear scale in each graph. Data shown are representative of those from at least three independent experiments. WT, wild type. Error bars indicate standard deviations.

FIG. 3.

Immunofluorescence visualization of Rad51 subnuclear foci. Wild-type (A and B), brca2tr (C and D), and xrcc2 (E and F) cells were analyzed at 3 h after treatment with 2 Gy of IR. (G) Average number of Rad51 foci per cell at 3 h after treatment with 2 Gy of IR. At least 100 morphologically intact cells were analyzed. WT, wild type. Error bars indicate standard deviations.

FIG. 3.

Immunofluorescence visualization of Rad51 subnuclear foci. Wild-type (A and B), brca2tr (C and D), and xrcc2 (E and F) cells were analyzed at 3 h after treatment with 2 Gy of IR. (G) Average number of Rad51 foci per cell at 3 h after treatment with 2 Gy of IR. At least 100 morphologically intact cells were analyzed. WT, wild type. Error bars indicate standard deviations.

FIG.4.

Analysis of V_λ sequences cloned from sIgM loss variants. (A) Fluctuation analysis of the generation frequency of sIgM loss variants. The abundance of sIgM loss variants was determined in several parallel cultures derived from sIgM-positive single cells after clonal expansion (3 weeks); median percentages are noted above each data set and are indicated by the dashed lines. WT, wild type. (B) Comparison of V_λ sequences from sIgM loss cells sorted from parental sIgM-positive clones of wild-type or brca2tr cells. Each horizontal line represents the rearranged V_λ1/J_λ (402 bp). Point mutations (lollipop shape) and gene conversion tracts (horizontal bars above lines) are indicated. (C) Proportions of V_λ sequences carrying different number of point mutations (PM), gene conversions (GC), or mutations of ambiguous origin (Amb) among sorted sIgM-negative populations. Segment sizes are proportional to the number of sequences carrying the number of mutations indicated around the peripheries of the pie charts. The total number of V_λ sequences analyzed is indicated in the centers of the charts, with the data compiled from analysis of three brca2tr cells. (D) Nucleotide substitution preferences deduced from point mutations in sequences from brca2tr cells. All of the V sequences having more than two point mutations showed different patterns, while 12 of the 27 analyzed V sequences having a single-base substitution showed identical events. Thus, most of the mutations may represent independent events rather than clonal expansion. (E) Ig diversification patterns in wild-type cells and rad51c, rad51d, and brca2tr mutants. The frequencies of gene conversion tracts and nontemplated point mutations per mutated sequence in sorted sIgM-negative cells are plotted. The ratio of point mutations per gene conversion is shown to the right of each bar. The mutation data are derived from 105 WT sequences from 5 subclones, 38 each of rad51c and rad51d sequences from two subclones, and 40 brca2tr sequences from three subclones. Ambiguous mutations (44) were excluded from this analysis. Data indicated by asterisks are from reference .

FIG. 5.

Fluctuation analysis of the generation frequency of sIgM gain revertants. (A) The abundance of sIgM gain variants was determined in several parallel cultures derived from sIgM-negative single cells after clonal expansion (3 weeks); median percentages are noted above each data set and are indicated by the dashed lines. (B) Gene conversion tract spectra, showing the average tract length in the rearranged V_λ segment from the sIgM-positive revertants. The black bars and open boxes represent the gene conversion tract spectra and one-base deletions, respectively. WT, wild type.