A role for error-prone polymerases in B cell somatic hypermutation has been hypothesized but not proven. Recent data published in Nature Immunology has focused on one particular candidate, polymerase η (pol η). Rogozin et al.1 reported the results of a mutational analysis of 15 sets of immunoglobulin variable (IgV) genes and identified that the consensus motifs RGYW and WA were targeted by somatic hypermutation. The underlined nucleotide indicates the most likely target of mutations. Interestingly, WA sequences were preferentially mutated on one DNA strand, whereas RGYWs were mutated on both strands. The strand polarity of the WA mutations suggested a role for an error-prone polymerase that synthesizes the nontranscribed DNA strand in somatic hypermutation. Analysis of the nucleotide substitution pattern of human pol η suggested that it could play a role in TA mutations of IgV regions. Another study2, in the same issue of Nature Immunology, which analyzed IgV gene rearrangements of Xeroderma pigmentosum variant (XP-V) patients with defects in pol η, reported a G/C bias in mutations that could have resulted from loss of the normal activity of pol η in hypermutation.
That pol η can cause a biased TA mutational pattern has been established1. In addition, pol η is up-regulated in murine germinal centers2, whereas it is paradoxically down-regulated by B cell receptor engagement in human B cells3. However, the contribution of pol η to somatic hypermutation remains to be determined. To address this, we reanalyzed the mutational pattern in a set of 37 human nonproductive VH gene rearrangements. Mutations in the nonproductive VH rearrangements represented the immediate impact of hypermutations without subsequent selective influences, as previously characterized4,5. Targeting to RGYW motifs on both DNA strands was found in these sequences. To address the potential role of pol η, all mutations that affected the prominently mutated TA and AA (WA sequences) were assessed. Altogether, 69 of 357 mutations (19.3%) occurred in TA dinucleotides. Within these 69 TA mutations, 42 (60.9%) were in an RGYW motif or its inverse repeat, WRCY, which suggested that the primary underlying process is mutational targeting of RGYW on both strands. It should be noted that TA is an inverse repeat of itself and as part of an RGYW/WRCY can be potentially targeted on either DNA strand. When all mutations are analyzed, there is frequently an imbalance between the mutations of T versus A that can be explained by the higher frequency of A in RGYW motifs. As an example, 24 mutations occurred in TAGN sequences, of which 21 mutations were part of an RGYW motif. Notably, only 27 mutations of TA (39.1%), but only 7.6% of all mutations, occurred in TA dinucleotides outside an RGYW/WRCY motif and could be candidates for targeting by pol η. Analysis of the nature of the 27 mutations of TA dinucleotides outside of an RGYW/WRCY motif revealed that A was mutated 21 times (replaced by G eight times, by C six times and by T seven times) and T was mutated six times (replaced by C five times and by A one time), consistent with the known WA bias of pol η. This suggests that pol η may play a role in somatic hypermutation, but its impact is likely to be less than that estimated by Rogozin et al.1.
This is a preview of subscription content, access via your institution