Activation-induced cytidine deaminase acts on double-strand breaks in vitro

Activation-induced cytidine deaminase (AID) is likely responsible for DNA cytidine deamination, although it may also act as an RNA deaminase. It functions on single-stranded DNA, the non-template strand in double-stranded DNA during transcription, or both strands in supercoiled DNA. To ask whether AID is able to deaminate cytidine at DNA breaks, plasmids, containing a SnaBI site (TAC down arrow GTA) that forms blunt ends after digestion with SnaBI, were generated. If AID deaminates cytidine at the upstream blunt end, the ATG start codon in either of two drug resistance genes will be regenerated after ligation and replication in UDG-null E. coli cells. This study shows that AID targets cytidine at the break. The extent of deamination activity beyond the break is correlated with the base composition in the break region. If the break region is A, T-rich, C > T transitions are extensive. However, when the break region is not A, T-rich, mutations are mainly restricted to the break, similar to findings in vivo. The results indicate that AID has activity on double strand breaks (DSBs). Based on previous and current findings, a somatic hypermutation (SHM) model is proposed, in which collision between the transcription apparatus and the replication fork generates DSBs. After AID acts on break ends, the error-prone DNA repair machinery fixes and creates mutations. (c) 2006 Elsevier Ltd. All rights reserved.