The ATPase activity of MLH1 is required to orchestrate DNA double-strand breaks and end processing during class switch recombination

Abstract

Antibody diversification through somatic hypermutation (SHM) and class switch recombination (CSR) are similarly initiated in B cells with the generation of U:G mismatches by activation-induced cytidine deaminase but differ in their subsequent mutagenic consequences. Although SHM relies on the generation of nondeleterious point mutations, CSR depends on the production of DNA double-strand breaks (DSBs) and their adequate recombination through nonhomologous end joining (NHEJ). MLH1, an ATPase member of the mismatch repair (MMR) machinery, is emerging as a likely regulator of whether a U:G mismatch progresses toward mutation or DSB formation. We conducted experiments on cancer modeled ATPase-deficient MLH1G67R knockin mice to determine the function that the ATPase domain of MLH1 mediates in SHM and CSR. Mlh1(GR/GR) mice displayed a significant decrease in CSR, mainly attributed to a reduction in the generation of DSBs and diminished accumulation of 53BP1 at the immunoglobulin switch regions. However, SHM was normal in these mice, which distinguishes MLH1 from upstream members of the MMR pathway and suggests a very specific role of its ATPase-dependent functions during CSR. In addition, we show that the residual switching events still taking place in Mlh1(GR/GR) mice display unique features, suggesting a role for the ATPase activity of MLH1 beyond the activation of the endonuclease functions of its MMR partner PMS2. A preference for switch junctions with longer microhomologies in Mlh1(GR/GR) mice suggests that through its ATPase activity, MLH1 also has an impact in DNA end processing, favoring canonical NHEJ downstream of the DSB. Collectively, our study shows that the ATPase domain of MLH1 is important to transmit the CSR signaling cascade both upstream and downstream of the generation of DSBs.

Figure 1.

The MLH1G67R mutation does not have a significant impact on SHM. (A) Alignment of the GxG motif of MLH1 ATPase domain and a 2.5-Å resolution of the backbone of hMLH1 ATP-binding pocket with ATP (Protein Data Bank accession no. 3NA3). G67 of human MLH1 is highlighted with a gray oval, and the ATP molecule is shown inside the MLH1 ATPase pocket. Asterisks signify that the amino acid conservation is identical, and colons signify similarity at the respective amino acid residues. On the right, a cartoon backbone and a surface representation of the first 347 aa residues of mammalian MLH1 along with an ATP molecule trapped in the ATPase domain are shown. The G67 residue does not seem to be in a region having secondary structures but is in close proximity to the adenine base of the ATP molecule and not to the triphosphate group. Although MLH1 was crystallized as a homodimer, for simplicity, only the monomer is shown here. (B) Global analysis of unique mutation frequencies corrected for base composition according to the SHMTool algorithm. (C) Distribution of mutations in the V_H186.2 sequence (273 bp). (D) The spectrum of base substitutions is expressed as frequencies of mutation (x10⁻² mutations/base) and corrected for base composition. The cumulative spectrum of mutations from G, C, A, or T sites is also shown (Sum). (E) Proportion of sequences with different load of mutations (indicated in each fraction) for antigen and nonantigen selected regions (V_H186.2 and Jh2-Jh4, respectively). The center of the pies represents the total number of sequences analyzed in each group. (F) Relative frequency of mutations at G:C versus A:T sites observed in V_H186.2 and Jh2-Jh4 regions.

Figure 2.

Splenic B cells from Mlh1^GR/GR mice show decreased switching efficiencies in ex vivo cultures. (A) Representative FACS results from WT and Mlh1^GR/GR cells stimulated with LPS or LPS + IL4 for 4 d. Percentages of switched cells are indicated within the gates. (B) The mean efficiency of switching in the WT group within each experiment was defined as 100%. Mean percentages of switching ± SEM are shown. Significance was determined using the two-tailed unpaired Student’s t test (***, P < 0.001). Data correspond to six WT and seven Mlh1^GR/GR mice assayed in two different experiments. (C) Overlay histograms of intracellular CFSE signal from splenic B cells stimulated with LPS + IL4 for different periods of time.

Figure 3.

Impaired generation of DSBs in switching Mlh1^GR/GR B cells. (A) ChIP experiments using anti-53BP1 antibodies were conducted on WT, Mlh1^−/−, and Mlh1^GR/GR splenic B cells activated with LPS + IL4 for 50 h. Specific DNA fragments corresponding to the Sμ, Sγ3, and Sγ1 regions of the IgH locus were quantitated in the immunoprecipitates by real-time PCR relative to input DNA. The bars denote the enrichment of specific DNA fragments present in the corresponding immunoprecipitates in stimulated (t = 50 h) over unstimulated (t = 0 h) cells. Error bars indicate the SD of triple PCRs (*, P < 0.05; ***, P < 0.001). (B) Western blotting analysis was performed on WT and Mlh1^GR/GR splenic B cells activated with either LPS (left) or with LPS + IL4 (right) for 96 h. Levels of γ-H2AX and anti-H2A were compared between the unstimulated (t = 0 h) and stimulated (t = 96 h) cells in the respective genotypes. A representative of three separate experiments is shown.

Figure 4.

Increased microhomologies and mutation frequencies at Sμ-Sγ3 junctions in Mlh1^GR/GR switching B cells. (A) Distribution of breakpoints (open triangles) within Sμ and Sγ3 SRs from 27 and 22 successful recombination events in six WT and seven Mlh1^GR/GR mice, respectively. Arrows indicate the position of the nested primers used to detect Sμ-Sγ3 junctions. (B) Percentage of junctions exhibiting insertions (ins), blunt joins, or microhomologies. (C) Scatter plot representing the length in base pairs of each detected microhomology. Horizontal bars indicate the mean length of microhomology. (D) Frequency and type of mutations accumulated at the Sμ-Sγ3 junctions of WT and Mlh1^GR/GR B cells. Insertions and deletions (ins/del) were computed together. WRC:GYW and WA:TW motifs (W = A or T; R = purine; Y = pyrimidine) correspond to preferred AID and error-prone polymerase Polη hotspots, respectively. (E) Distribution of mutations around the recombination breakpoints. Mutations occurring within the same bin window of 100 bp were counted together and represented in histograms.