Disparate roles of ATR and ATM in immunoglobulin class switch recombination and somatic hypermutation

Abstract

Class switch recombination (CSR) and somatic hypermutation (SHM) are mechanistically related processes initiated by activation-induced cytidine deaminase. Here, we have studied the role of ataxia telangiectasia and Rad3-related protein (ATR) in CSR by analyzing the recombinational junctions, resulting from in vivo switching, in cells from patients with mutations in the ATR gene. The proportion of cells that have switched to immunoglobulin (Ig)A and IgG in the peripheral blood seems to be normal in ATR-deficient (ATRD) patients and the recombined S regions show a normal "blunt end-joining," but impaired end joining with partially complementary (1-3 bp) DNA ends. There was also an increased usage of microhomology at the mu-alpha switch junctions, but only up to 9 bp, suggesting that the end-joining pathway requiring longer microhomologies (> or =10 bp) may be ATR dependent. The SHM pattern in the Ig variable heavy chain genes is altered, with fewer mutations occurring at A and more mutations at T residues and thus a loss of strand bias in targeting A/T pairs within certain hotspots. These data suggest that the role of ATR is partially overlapping with that of ataxia telangiectasia-mutated protein, but that the former is also endowed with unique functional properties in the repair processes during CSR and SHM.

Figure 1.

Amplification of Sμ-Sα breakpoints. A typical run for control and ATRD patient. The numbers of μ-α switch fragments were determined from 10 PCR reactions run in parallel (lanes 1–10). 11 frag ments can be seen in the control (lane 9 has two distinct bands) and 10 in ATRD1 (lane 5 has none and lane 10 has two distinct bands). M, molecular weight marker.

Figure 2.

Microhomology usage at Sμ-Sα and Sμ-Sγ junctions. (A) Pie charts demonstrate the perfectly matched short homology usage at Sμ-Sα and Sμ-Sγ junctions in controls, ATRD, and A-T patients. The proportion of switch junctions with a given size of perfectly matched short homology is indicated by the size of the slices. For clarity, the 4–6 bp homology portion, which significantly differs between ATRD and controls, is separated from the pie. (B) Accumulative curves indicate the microhomology usage at Sμ-Sα junctions in controls, ATRD and A-T patients. The percentage of all switch junctions with a given size, using perfectly matched short homology, is plotted.

Figure 3.

Pattern of mutations introduced in the VH3-23-Cγ transcripts. (A) Pie charts demonstrate the proportion of clones carrying the indicated number of mutations. (B) Distribution of mutations in VH3-23-Cγ transcripts in controls (above the zero line) and ATRD patients (below the zero line). Nucleotides are numbered from the third base of codon 10 of the VH3-23 coding region. The number of mutations at each position is plotted and the major hot spots are highlighed. (C) Nature of base substitutions in the VH3-23 genes. The germline sequence used for comparison contains 21.4% A, 31.7% G, 21.4% T, and 25.5% C. Correction: values presented in the column were corrected to represent a sequence with equal amounts of the four nucleotides. Statistical analysis was performed using χ² test. Numbers that are significantly different from controls are bold. *P < 0.05, **P < 0.01.