Activation induced deaminase (AID) is essential for both Ig somatic hypermutation (SHM) and class switch recombination (CSR) in mature B cells. AID deaminates dC to dU. There are four uracil DNA glycosylases, UNG, SMUG1, TDG, and methyl binding domain 4 (MBD4), that are capable of recognizing and removing dU in U:G mismatches. Using the base excision repair (BER) pathway, many (but not all) AID induced dU bases can be excised by a uracil DNA glycosylase (UNG) leaving an abasic site that can then be replicated over by error prone polymerase to produce both transition and transversion mutations. Genetic studies show that UNG deficiency in mice and humans leads to loss of class switch recombination (CSR) and impaired somatic hypermutation (SHM). Evidence indicates that TDG cannot substitute for UNG and SMUG1 plays little natural role in these processes since it is poorly expressed in activated B cells. Other U:G mismatches could be substrates for mismatch repair (MMR) MSH2/MSH6 binding which in turn recruit PMS2-MLH1. PMS2 nicks the DNA and contributes to the induction of DSBs in S regions. MBD4 protein was originally discovered by virtue of its interaction with MLH1, a constituent of the MMR protein complex and MMR is deeply involved in CSR and SHM. We were intrigued by the functional association of MBD4 and AID in the context of active DNA demethylation in zebrafish and have sought to explore this potential interaction in mature B cells engaged in CSR. Previous studies focused on Mbd4 knockout mice indicated no phenotype with regard to CSR and SHM. However, we noticed that there are alternative splice variants of Mbd4 mRNA with open reading frames, which would enable expression of the C-terminal end of the protein even when exons 2-5 are deleted. We constructed another knockout in which Mbd4 exons 6-8 and the 3'UTR were deleted in CH12 cells that are normally capable of inducible CSR. Strikingly we found that in Mbd4 deficient CH12 cells, CSR was significantly impaired even while all other criteria for CSR remain intact. Based on these intriguing new studies we propose to more fully examine MBD4 for functional isoforms and to construct a new mouse in which Mbd4 exons 6-8 and the 3'UTR have been deleted by targeted homologous recombination. Follow-up studies will characterize these mice with respect to CSR and SHM. Long term, this mouse will also be used to investigate the involvement of Mbd4 in mismatch repair and genome stability, keys to understanding oncogenesis.
The Mbd4 gene is mutated and inactivated in a large percentage of primary human colorectal carcinomas. However, the physiological role of MBD4 in specific DNA repair pathways remains unclear. Based on new information suggesting a role for Mbd4 in the immune system, we will study MBD4 for its role in antigen gene rearrangements. This information may clarify the role of MBD4 and further elucidate how the cell keeps the genome intact and what happens when this system fails, as in cancer.