66 | Chapter 3 DNA damage repair such as PRDM9, EGR1, MAZ or RREB141–44, in p53 signaling or apoptosis such as PATZ1or SP145,46, or in B-cell development and proliferation such as KLF447. RAG-dependent DNA breaks localized close to SSR such as CA- and GC-rich sequences Several of the motifs identified by the unsupervised Simple, Thorough, Rapid, Enriched Motif Elicitation (STREME) analysis indicated an association with simple sequence repeats (SSR). SSRs have previously been associated with the formation of non-canonical (non-B) DNA structures, such as quadruplexes, hairpins, cruciform or triple-stranded DNA48,49, and were suggested to be a possible target of RAG1/2 recombinase50. Therefore, we investigated the presence of GA and CA repeats around the RAG-dependent NBS1 binding sites. Using unsupervised SEA analysis, each dataset was interrogated for the presence of [CACA]1-20 and [GAGA]1-20 repeats, where increases from 1 up to 20, the maximum repeat length corresponding to half of the average NBS1 peak length without any extension. While no notable enrichment for CA or GA repeats was observed in sequences directly under the peaks (Figure 6A), both 500bp and 1000bp-extended NBS1 peaks showed strong enrichment of both CA and GA repeats (Figure 6B and 6C). Next to the SSR, GC-rich regions are also known to assume structures such as hairpins, cruciform or triple-stranded DNA, and the GC-rich motif 5’-GCCGCCGGGCG-3’ was identified as RAG1/2 transposition hotspot51,52. We next investigated the presence of this transposition hotspot motif, and 1 to 5 repeats thereof [5’-GCCGCCGGGCG-3’]1-5, around the putative RAG-off target DSBs. While no enrichment (all q-values > 0.05) has been identified in sequences directly under the RAG-dependent NBS1 peaks (Figure 7A), a moderate enrichment of 1.3-11 fold (q-value < 0.05) was found in the peaks with 500bp offset and 1.5-15 fold in the peaks with 1000bp offset (Figure 7B and 7C), further underscoring the strong association of RAG1/2 (off-) targets with the repeat regions. Figure 5. Motif analysis of RAG1/2-dependent NBS1 peaks motif logos are shown on the left side. For each motif, additional Tomtom analysis was performed to compare the obtained motif to known motifs (graphs on the right-hand side) in not extended NBS1 peak sequences (A), in sequences extended by 500bp on each end (B), and in sequences extended by 1000bp on each end (C).
RkJQdWJsaXNoZXIy MTk4NDMw