4 DNA damage response regulates RAG1/2 expression through ATM-FOXO1 | 97 (Idelalisib) did not prevent NCS-induced down-regulation of RAG1 protein either (data not shown). To firmly establish that AKT is not involved in the regulation of RAG1 expression following DNA damage we show that RAG1 protein is decreased equivalently in BV173 cells overexpressing WT-AKT or a dominant-negative kinase-dead mutant of AKT (K179M/ T308A/S473A) upon NCS treatment (Figure 4D). These combined results clearly indicate that in pre-B cells the effect of DNA damage on RAG1 expression and FOXO1 stability does not involve the PI3K-AKT pathway. ATM interacts with FOXO1 but does not phosphorylate FOXO1 on SQ/TQ residues To gain more insight into how the stability and DNA binding activity of FOXO1 is regulated by ATM upon induction of DNA damage, we assessed whether ATM physically interacted with FOXO1 in RAG1-expressing pre-B cells using the proximity ligation assay (PLA), which detects two proteins in close spatial proximity (<30 nm distance)42. We observed specific interactions between activated ATM (phosphorylated on Ser1981) and FOXO1 in NCS-treated BV173 cells in which RAG1 expression was induced by STI571 (Figure 5A). Strikingly, using antibodies that detected both phosphorylated and unphosphorylated (inactive) ATM we found specific interactions between ATM and FOXO1 regardless of whether DNA damage was induced (Figure 5B), suggesting that FOXO1 and ATM constitutively interact in RAG1-expressing pre-B cells, and that autophosphorylated ATM is associated with FOXO1 upon induction of DNA damage. In conjunction with the ChIP results that showed enrichment of phosphorylated ATM at the Erag enhancer in NCS-treated cells and loss of FOXO1 binding (Figure 2E), these data suggest that phosphorylated ATM could be involved in the release of FOXO1 from Erag. To confirm the specificity of the PLA experiments we assessed the interaction between ATM and phosphorylated p53 (Ser15; an ATM kinase substrate), and as expected the interaction was induced by DNA damage and inhibited by KU55933 pre-treatment (Figure 5C). In addition, for each of the antibody combinations, PLA experiments in which one of the antibodies was omitted were performed as negative controls, showing no specific PLA signals (SUPPL Figure 2). A functional consequence of the interaction between ATM and FOXO1 might be the ATM-dependent phosphorylation of FOXO1, which may regulate its DNA-binding capacity and stability. Interestingly, a previously performed proteomic screen for ATM and ATR (ATM and Rad3-related) substrates in HEK293T cells that were exposed to 10 Gy of g-irradiation suggested that FOXO1 may be phosphorylated on residue Ser509 by ATM and/or ATR43. To explore whether FOXO1 is phosphorylated by ATM/ATR in RAG1-expressing BV173 cells following induction of DNA damage we performed immunoprecipitations with the ATM/ ATR substrate-specific antibody (phospho-SQ/TQ). However, using this approach we could not detect SQ/TQ-phosphorylated FOXO1 upon NCS treatment, whereas ATM-dependent phosphorylation of an established ATM/ATR target such as Nijmegen Breakage Syndrome
RkJQdWJsaXNoZXIy MTk4NDMw