172 | Chapter 7 where the Igh recombination already took place while the Igl are still in their germline configuration. An accumulation of RAG-specific NBS1 peaks was observed at the Igk locus, as expected. However, we also observed an accumulation of RAG1/2-dependent DNA breaks at the Ig heavy chain (Igh) locus on chromosome 12. The abundant binding of RAG1/2-dependent NBS1 to Igh suggests concomitant induction of RAG1/2-mediated DNA breaks at Igh and Igl, potentially breaching the concept of strictly serial fashion of gene recombination18. Some earlier studies propose that while v-Abl transformed cells are able to initiate gene recombination, they fail to terminate the process. Perhaps due to the v-Abl presence, which antagonizes the pre-B cell receptor (pre-BCR) signaling and causes the pre-B cells with one functional Igh allele to persistently transcribe and rearrange the non-recombined allele19,20. Another explanation for the continuous presence of RAG-dependent NBS1 binding to Igh could come from the previous work on “receptor editing” showing that functional B-cell receptors can re-engage in further recombination. This may result in the editing of the receptor’s specificity, allowing the cells with auto-reactive B-cell receptors to escape elimination21. The discovery that human and mouse genomes are densely populated with sequences that resemble RSS, so-called cryptic RSS22,23, begs the question of whether the RAG1/2 recombinase complex is able to target and cleave such sequences in regions outside of the Ig loci. Several studies have demonstrated that in extrachromosomal assays RAG1/2 was able to cleave cRSS, and similar observations were made in a v-Abl pre-B cell lines transformed with constructs containing cRSS-substrates24,25. However, the evidence for RAG1/2 off-target activity on the chromatin in living cells is challenging to capture. We investigated the presence of RSS motifs in the proximity of the RAG-dependent NBS1 binding sites and found enrichment for canonical RSS around the NBS1 peaks at the Igk locus, which served as a positive control. Outside of the Ig loci, in the 500-1000bp proximity to RAG-dependent NBS1 binding sites, we observed an enrichment for nonamers, but not for heptamers. Similar observations were made in another study that showed an enrichment for nonamers, but not heptamers in the proximity of RAG1 binding sites in mouse v-Abl cell lines and in mouse thymocytes17. We have also examined the presence of several selected cRSS26,27 in the proximity of RAG-dependent NBS1 binding sites but found no enrichment (unpublished data, data not shown). However, there are hundreds of possible cRSS sequences, therefore a more systematic approach to investigating the cRSS enrichment around RAG-dependent DNA breaks would be more appropriate in a follow-up study. In addition to the RSS motif analysis, we performed an unsupervised motif analysis to determine if there are any other motifs enriched in the vicinity of the RAG-dependent NBS1 binding sites. The motifs we have identified were predominantly zinc-finger-binding sequences, abundant in eukaryotic cells. Intriguingly, several of these motifs corresponded to known binding sites of proteins such as PR domain zinc finger protein 9 (PRDM9), early
RkJQdWJsaXNoZXIy MTk4NDMw