Locus-specific control of DNA resection and suppression of subtelomeric VSG recombination by HAT3 in the African trypanosome

Abstract

The African trypanosome, Trypanosoma brucei, is a parasitic protozoan that achieves antigenic variation through DNA-repair processes involving Variant Surface Glycoprotein (VSG) gene rearrangements at subtelomeres. Subtelomeric suppression of DNA repair operates in eukaryotes but little is known about these controls in trypanosomes. Here, we identify a trypanosome histone acetyltransferase (HAT3) and a deacetylase (SIR2rp1) required for efficient RAD51-dependent homologous recombination. HAT3 and SIR2rp1 were required for RAD51-focus assembly and disassembly, respectively, at a chromosome-internal locus and a synthetic defect indicated distinct contributions to DNA repair. Although HAT3 promoted chromosome-internal recombination, it suppressed subtelomeric VSG recombination, and these locus-specific effects were mediated through differential production of ssDNA by DNA resection; HAT3 promoted chromosome-internal resection but suppressed subtelomeric resection. Consistent with the resection defect, HAT3 was specifically required for the G2-checkpoint response at a chromosome-internal locus. HAT3 also promoted resection at a second chromosome-internal locus comprising tandem-duplicated genes. We conclude that HAT3 and SIR2rp1 can facilitate temporally distinct steps in DNA repair. HAT3 promotes ssDNA formation and recombination at chromosome-internal sites but has the opposite effect at a subtelomeric VSG. These locus-specific controls reveal compartmentalization of the T. brucei genome in terms of the DNA-damage response and suppression of antigenic variation by HAT3.

Figure 1.

Repair at the INT locus is primarily RAD51 dependent. (A) The schematic illustrates the Tb927.11.4530/4540 locus on chromosome 11 in the INT strain; following insertion of an I-SceI site within a red fluorescent protein (RFP)–puromycin N-acetyltransferase (PAC) fusion gene (R:P); the I-SceI site is indicated. (B) Western blotting validates the INT^rad51 (Tb927.11.8190) null cells. Coo., Coomassie-stained gel, which serves as a loading control. (C) Southern blot analysis of DNA repair following an I-SceI induced break in the INT and INT^rad51 strains. Strains were grown in the presence or absence of tetracycline for 1 week. Genomic DNA was digested with Bsp120I and HindIII and hybridized with a ‘4530’ probe (see (A)). In the INT strain, the modified 5.2-kb ‘a’ allele (see panel (A)) is primarily converted back to ‘wild type’ by HR using the ‘b’ allele as a template. In the INT^rad51 strain, a distinct mechanism (MMEJ) generates a fragment at ∼5 kb. WT, wild-type cells; a, chromosome 11a allele; b, chromosome 11b allele. The schematic indicates the observed fragments. (D) Cloning efficiency in the INT strain and rad51-null derivative. The assay was carried out using medium with or without tetracycline to induce the DSB. Proportions of cells that recover from an induced DSB were derived by dividing the induced value by the un-induced value. Error bars, SD.

Figure 2.

HAT3 and SIR2rp1 facilitate repair at a chromosome-internal locus. (A) Cloning efficiency in the INT strain. hat3, Tb927.10.8310-null INT strain; sir2rp1, Tb927.7.1690-null INT strain. A Student's t-test was used to derive P values: *P < 0.001 and **P < 0.0001. Other details are as in the legend to Figure 1D. (B) Monitoring of RAD51 foci. Nuclei with detectable RAD51 foci were assessed over a time course following induction of I-SceI expression. n = 200 at each time point. Error bars, SD. (C) Immunofluorescence analysis of RAD51 foci in INT (left-hand panel) and INT^hat3 (right-hand panel) cells following induction of I-SceI expression (+Tet for 12 h). RAD51, red; DNA counter-stained with DAPI, blue; N, nucleus; K, kinetoplast.

Figure 3.

A synthetic defect in hat3-sir2rp1 double-null T. brucei. (A) Growth of wild-type, hat3, sir2rp1 and hat3:sir2rp1 double-null strains. Error bars, SD. (B) Cloning efficiency of the wild-type (WT) hat3, sir2rp1 and hat3-sir2rp1 double-null strains. Error bars, SD. (C) Monitoring of γH2A foci in the hat3, sir2rp1 and hat3-sir2rp1 double-null strains. Nuclei with detectable γH2A foci were counted, n = 200 for each sample. Error bars, SD. Inset: immunofluorescence image of a γH2A focus (green); DNA was counter-stained with DAPI, blue.

Figure 4.

HAT3 suppresses subtelomeric VSG recombination. (A) The schematic illustrates the subtelomeric VSG221 locus on chromosome 6a in the TEL strain after insertion of the ^ScePAC cassette; the I-SceI site is indicated; black circle, centromere; striped box, 70-bp repeats; arrowheads, telomeric repeats. (B) Cloning efficiency in the TEL strain and hat3 and sir2rp1-null derivatives. *P < 0.01 and **P < 0.001. Other details are as in the legend to Figure 1D. (C) Survivors from the clonogenic assays were scored by VSG221 immunofluorescence microscopy. hat3,n = 8; sir2rp1,n = 25. The VSG221 immunofluorescence images show an example of a switched survivor. DNA was counter-stained with DAPI.

Figure 5.

Locus-dependent control of DNA resection by HAT3. Accumulation of ssDNA adjacent to a DSB was monitored by slot-blot analysis in the INT (A) and TEL (C) strains. Genomic DNA was extracted at the times indicated following I-SceI induction. Ninety percent of each sample was ‘native’ (n; where a hybridization signal with native DNA indicates the presence of ssDNA) and the remainder was denatured (d). The probes used on each blot are indicated on the right (see the schematic maps in Figures 1A and 4A). ssDNA signals from the INT (B) and TEL (D) strains were quantified by phosphorimager analysis as previously described (10).

Figure 6.

DSB repair within the tubulin/TAN gene array (Tb927.1.2340-2400). (A) The schematic illustrates the TUB locus on chromosome 1 after replacement of an α-TUB gene with the RFP:PAC cassette. The I-SceI recognition site is indicated; black circle, centromere; β, β-TUB genes. (B) The schematic illustrates analysis of the TUB alleles on chromosome 1; the polymorphic wild-type (WT) a and b alleles (37) and the b allele after insertion of the R^sP cassette (TAN, b+) are shown. The approximate sizes of the NcoI (N) fragments expected on Southern blots are indicated below the maps. For Southern blotting, genomic DNA was digested with Nco1 and subjected to pulsed-field gel electrophoresis. Bands corresponding to the WT and b+ alleles are present. (C) Western blotting validates the TAN^rad51 null cells. Coo., Coomassie-stained gel, which serves as a loading control. (D) Cloning efficiency in the TAN^rad51-null strain. Other details are as in the legend to Figure 1D. (E) DSB-repair survivors were cloned in medium containing tetracycline and genomic DNA was analysed by Southern blotting as described in (B) above. Bands corresponding to the ‘WT’-b and truncated b alleles are recovered. The schematic illustrates repair, which we predict is via either allelic homologous recombination (survivors 2, 4 and 5–7) or single-strand annealing (survivors 1, 3 and 6).

Figure 7.

HAT3 and SIR2rp1 facilitate DNA resection at a chromosome-internal tandem gene array. (A) Cloning efficiency in the TAN strain, hat3-null and sir2rp1-null derivatives. Other details are as in the legend to Figure 1D. (B/C) Accumulation of ssDNA adjacent to the DSB was monitored as described in the legend to Figure 5A/C. The loading control was prepared using a ‘4250’ probe also from chr. 1 (see the Materials and Methods section). (C) Signals derived from the TAN strains were quantified by phosphorimager analysis as described (10).

Figure 8.

Impact of HAT3 or SIR2rp1 on γH2A foci and the G₂ DNA-damage checkpoint response. (A) Monitoring of γH2A foci in the INT, TEL or TAN strains with or without HAT3 or SIR2rp1. Nuclei with detectable γH2A foci were counted 12 h after I-SceI induction. See the legend to Figure 3C for other details. (B) Monitoring of nuclei in the G₂-phase of the cell cycle in the INT, TEL or TAN strains with or without HAT3 or SIR2rp1. G₂ cells contain a single nucleus and two kinetoplasts. Counts were taken 0, 12 and 24 h after I-SceI induction (n ≥ 90 cells at each time point).