A Role for Small RNA Molecules during the DNA Repair Phase of Somatic Hypermutation

Somatic hypermutation (SHM) is a natural process that introduces point mutations into immunoglobulin (Ig) genes during antibody affinity maturation. In addition to this fundamental role in immune diversification, aberrant targeting of SHM contributes to translocations and point mutations of proto-oncogenes associated with B cell malignancy. During the first phase of SHM, the enzyme activation induced cytidine deaminase (AID) converts cytosine (C) to uracil (U) to result in a U-G mismatch. Spontaneous U-G mismatches are normally corrected by high fidelity DNA repair pathways. However, during the second phase of SHM, U-G mismatches are processed by low fidelity base excision and mismatch repair pathways to yield mutations. These second phase pathways are initiated by recognition of the uracil by uracil DNA glycosylase (UNG) and MSH2/MSH6. As a DNA mutator, AID poses a direct threat to genomic integrity but the mechanisms responsible for guiding AID to its genetic target in the first phase and determining whether a mutation will be repaired in a high fidelity or low fidelity manner during the second phase are not understood. Numerous cellular processes are regulated by small inhibitory RNA (siRNA) and microRNA (miRNA) molecules through a mechanism known as RNA interference (RNAi). To test the hypothesis that small RNA molecules are involved in SHM, we introduced two natural inhibitors of RNAi into the DT40 B cell line. DT40 cells express high levels of AID and mutate their Ig loci with high frequency. The protein 3′hExo is a siRNase that has been found to be a general inhibitor of RNAi. VA1 is a non-coding adenoviral RNA that disrupts RNAi by inhibiting nuclear export of pre-miRNA and by direct inhibition of Dicer. Following retroviral infection of 3′hExo or VA1 and empty vector controls into DT40 cells, 24 clones from each group were expanded from single cells over 28 days. DT40 cells are IgM⁻ due to a mutation in the Ig heavy chain gene. When this mutation is corrected by AID activity, they become IgM⁺, allowing mutation frequency to be followed by flow cytometric analysis of IgM reversion. Compared with mock infected cells and cells infected with empty vector, cells that over-expressed 3′hExo or VA1 demonstrated a 3 and 2-fold reduction in IgM reversion, respectively. Similarly, sequencing of the Ig light chain gene from a minimum of three clones from each group revealed mutation frequencies of 6.2×10⁻⁴ mutations/nucleotide (nt) in mock infected cells, 6.8×10⁻⁴ mutations/nt in empty vector controls, 1.8×10⁻⁴ mutations/nt in cells over-expressing 3′hExo, and 2.7×10⁻⁴ mutations/nt in cells over-expressing VA1. No difference in the type of mutations or nucleotide bias was observed. Comparable reductions in RNAi activity with over-expression of 3′hExo or VA1 in DT40 cells were observed in siRNA reporter assays. Thus, general inhibition of RNAi in DT40 cells caused a 2 to 3-fold reduction in AID-mediated mutation events (P<0.001). Neither 3′hExo nor VA1 had a significant effect on the rate of cell proliferation, and Ig gene expression and AID expression were equivalent in all experimental groups. To distinguish a role for small RNA molecules in AID targeting versus repair of the resulting U-G mismatch, we performed the same experiments in UNG-deficient DT40 cells. In the absence of UNG, U-G mismatches are not recognized or repaired but are simply replicated, revealing the footprint of AID activity in the form of C to T transition mutations. In contrast to wild type DT40 cells, over-expression of 3′hExo or VA1 in UNG-deficient DT40 cells did not result in a reduction in mutation frequency and, as expected, all observed mutations were C to T transitions. Thus, in UNG-deficient cells, AID targeting was intact despite abrogation of the RNAi pathway by 3′hExo and VA1 (confirmed by siRNA reporter assay). This suggests that the reduced mutation frequency when 3′hExo or VA1 is over-expressed in wild type DT40 cells may be due to augmented high fidelity DNA repair rather than reduced AID targeting. Based on these results, we propose that small RNA molecules may play a role in mediating the balance of high fidelity and low fidelity DNA repair during SHM. This balance must be carefully maintained in order to preserve genomic integrity without compromising immune diversification. Further characterization of this mechanism is critical to our understanding of how these processes break down during malignant transformation of B cells.