4 DNA damage response regulates RAG1/2 expression through ATM-FOXO1 | 103 pre-B cells the stability/turnover of RAG1 was clearly affected by DNA damage, suggesting that a post-translational regulatory mechanism exists that may be cell-type specific. However, in further support of a transcriptional regulatory mechanism, we show that DNA damage abolished FOXO1 binding to the Erag enhancer, which was restored by pre-treatment with the ATM kinase inhibitor. Furthermore, DNA damage resulted in the partial cleavage/degradation of FOXO1. However, the dramatic reduction of RAG1 transcription upon DNA damage suggests that the incomplete loss of FOXO1 protein integrity is a secondary event following the loss of FOXO1 DNA-binding activity. The very short half-life of RAG1 protein and mRNA of approximately 15 minutes and 30 minutes, respectively29,30, allows for the dynamic regulation of expression at the transcriptional level, falling well within the range of 2 hours, after which we analyzed RAG1 protein and mRNA expression upon induction of DNA damage. Also from our experiments, we conclude that caspase-mediated proteolysis is not involved in the regulation of RAG1 expression, as pre-treatment with the pan-caspase inhibitor Q-VD-OPH did not prevent RAG1 downregulation upon DNA damage. In previous studies, it was shown that treatment with PI3K inhibitors or the specific lack of the p85a subunit of PI3K in immature B cells increased the expression of RAG151, whereas the overexpression of myristoylated (constitutively active) AKT suppressed RAG1 expression in developing B cells32. Since the FOXO transcription factors are bona fide phosphorylation targets of the PI3K-AKT kinase pathway, acting downstream from the pre-BCR and BCR51 and resulting in nuclear exclusion and degradation of FOXO, we assessed the role of AKT signaling in the downmodulation of RAG1 expression following DNA damage. We found that both AKT and FOXO1 were phosphorylated in an ATM-dependent manner upon NCS treatment, but surprisingly, inhibition of AKT kinase activity did not prevent the downregulation of RAG1. In addition to AKT, several other kinases have been reported to regulate the subcellular distribution and activity of FOXO1, such as p38 MAPK, IkB kinase ERK, and JNK27,52. Pretreatment of RAG-expressing pre-B cells with small molecule inhibitors for neither of these kinases prior to induction of DNA damage was able to prevent RAG1 down-regulation, suggesting that these kinases are not involved in the regulation of RAG1 expression upon DNA damage (data not shown). Using PLA, we show that the ATM kinase and FOXO1 can be found in close proximity (<30 nm). This interaction was found irrespective of the induction of exogenous DNA damage suggesting a constitutive association. Despite the physical interaction, we found no evidence for the ATM-dependent phosphorylation of FOXO1 upon induction of DNA damage. Using ChIP, we detected phosphorylated ATM at the Erag enhancer region in NCS-treated cells, suggesting that DNA damage activates ATM to locally release FOXO1 binding to Erag. Previously, a similar local effect on the transactivation of RAG expression by FOXO1 was described by Chow et al. who showed that MAPK-activated protein kinase (MAPKAPK5) phosphorylates FOXO1 on Ser215, which is required for binding to the
RkJQdWJsaXNoZXIy MTk4NDMw