Reflections on the state of play in somatic hypermutation

Somatic hypermutation (SHM) of rearranged variable genes proceeds in two phases. Phase I which is well understood is triggered by activation-induced cytidine deaminase (AID) and targets mutations at C:G base pairs equally on both DNA strands. Phase II, is less well understood, and targets A:T base pairs by coopting DNA polymerase-eta and acts in a strand biased fashion such that mutations off A exceed mutations of T by two- to threefold. Current molecular models attempting to explain A:T targeted Phase 11 are critically reviewed. It is the author's viewpoint that the 'RT-model', which invokes both transcription-coupled DNA and RNA deamination together with error-prone reverse transcription via Pol-eta, is the best explanation of current somatic hypermutation data. (c) 2008 Elsevier Ltd. All rights reserved.