Control of RAG Cleavage Activity Contributes to Maintaining Genome Stability During V(D)J Recombination

Abstract 2416

Acute lymphoblastic leukemia (ALL) results from malignancy of lymphoid progenitor cells and affects both adults and children. It is the most common childhood cancer and despite advances in treatment that now result in above 80% cure rates for children, considerable problems remain with current therapies. These include low cure rates in children with high-risk ALL, the complexity and toxic effects of current treatments and the stubbornly poor prognosis of adults with ALL (with a less than 40% long-term survival rate). ALL can be initiated by errors in V(D)J recombination, a process which creates multiple combinations of receptor genes in B and T lymphocytes in order to target foreign pathogens. During recombination, DNA double strand breaks are introduced at the borders of two selected gene segments and repair creates a new gene combination. Chromosomal translocations can occur both by mis-targeting of the RAG recombinase proteins at cryptic recombination signal sequences, as well as illegitimate repair with a DNA break generated by alternative cellular processes. Our work has unveiled a remarkable and previously unknown control step which acts during V(D)J recombination to protect genome stability. We demonstrated that the key DNA damage response factor and serine/threonine kinase ATM (ataxia telangiectasia mutated), prevents aberrant cleavage during V(D)J recombination. In wild-type cells only one of the two homologous Ig alleles is normally cleaved at a time, whereas in ATM deficient cells both Ig alleles can be cleaved simultaneously and chromosomal aberrations are detected on two Ig alleles (Hewitt et al., Nature Immunology 2009). Our recent work has been directed at understanding how ATM and the RAG recombinase (RAG1 and RAG2 proteins) cooperate to implement allelic control of V(D)J recombination. We hypothesized that ATM may act to control RAG cleavage, either directly or indirectly. To test this, we investigated developing B cells from coreRAG1 or coreRAG2 mice; these are the shortest active forms of the proteins but lack regulatory domains. We assessed mono- versus biallelic cleavage using γH2AX to indicate repair foci and as a read-out for DNA double strand breaks. In pre-B cells from coreRAG1 mice, γH2AX foci were predominantly colocalized with only one Igk allele per cell, which indicates monoallelic cleavage. In contrast, biallelic colocalization was highly significant in coreRAG2 expressing pre-B cells. We have analyzed RAG2 mutants to precisely identify the protein motifs that regulate cleavage. These were introduced into Rag2-deficient pre-B cell lines by retroviral infection. Expression of a coreRAG2 construct in these cells recapitulated the biallelic cleavage seen in ex-vivo isolated pre-B cells. We found that mutation of putative serine/threonine phosphorylation motifs also resulted in significant biallelic colocalization of γH2AX with Igk alleles. This suggests that RAG2 performs a similar function to ATM in restricting simultaneous RAG cleavage on the antigen receptor loci and may indeed cooperate with serine/threonine kinases. These data provide a mechanistic basis for the similarities in chromosomal abnormalities between Atm^–/– and coreRag2/p53^–/– lymphomas and will contribute to our understanding of why recurrent chromosomal translocations and lymphoid cancers arise in ATM-deficient mice and humans.