Diamond Blackfan anemia (DBA) is an inherited bone marrow failure syndrome most often due to germline pathogenic variants (PVs) in ribosomal protein (RP) genes, with RPS19 and RPL5 being the most frequently mutated. In addition to hypoplastic anemia, individuals have a moderately increased risk of myelodysplastic syndrome, acute myeloid leukemia, and solid cancers, most commonly colorectal cancer and osteosarcoma. A causal link between RP deficiency and cancer predisposition has yet to be firmly established. Our work aims to examine the impact of deficiencies in RPS19 and RPL5 on DNA double-strand break (DSB) repair with the overarching hypothesis that disruptions in DSB repair could increase mutagenesis and ultimately contribute to cancer development in DBA. We found that DBA patient-derived lymphoblastoid cells were hypersensitive to ionizing radiation (IR) and had delayed resolution of the DNA damage marker, γ-H2AX, at IR-induced foci. Additionally, we found that both RPS19- and RPL5-knocked down (KD) bone marrow-derived CD34+ cells had delayed resolution of DSBs after IR in neutral comet assays. These data suggest that RPS19 and RPL5 deficiencies lead to impaired DSB repair. We, therefore, assessed the effect of RPS19- and RPL5-KD on the four main DSB repair pathways using well-established U2OS (osteosarcoma) DSB reporter cell lines. We found decreased homologous recombination (HR) and single-strand annealing (SSA) upon RPS19-KD, whereas the efficiencies of nonhomologous end-joining (NHEJ) and alternative end-joining (alt-EJ) were unchanged. Consistent with an HR defect, RPS19-KD cells were hypersensitive to PARP inhibition. In contrast, RPL5-KD cells had increased NHEJ and alt-EJ, whereas the efficiencies of HR and SSA were unchanged. Together, these results suggest that shifts in DSB repair pathway choice, with less high-fidelity HR, or more error-prone NHEJ or mutagenic alt-EJ, could contribute to mutagenesis in the context of RPS19 or RPL5 PVs. To investigate the mechanism(s) underlying these effects, we examined cell cycle profiles and found decreased G1 and increased G2/M cells following both RPS19- and RPL5-KD, which suggests altered cell cycle profiles are not responsible for DSB repair effects since NHEJ occurs principally in G1 and HR is restricted to G2. Given the essential roles of RPS19 and RPL5 in translation, we also examined levels of crucial proteins in each DSB repair pathway. Surprisingly, we found increased levels of NHEJ proteins DNA ligase IV (LIG4) and XRCC4 in both RPS19 and RPL5-KD cells. However, restoring LIG4/XRCC4 to normal levels via siRNA KD did not normalize the DSB repair efficiencies. Since HR and SSA require extensive resection of the 5' DNA strand at DSBs, we examined whether RPS19-KD impaired this key step by assessing the accumulation of nuclear RPA2, which associates with end-resected DNA. We found nuclear RPA2 levels were not reduced but rather increased following IR compared to the control, indicating the reduction in HR and SSA efficiency in these cells was not due to impaired end resection and may, instead, involve a downstream step in HR that we are currently investigating. To assess if RPS19 and RPL5 might directly contribute to repair, we monitored the recruitment of fluorescently tagged RPS19 and RPL5 to DSBs. We found that these proteins were rapidly recruited to sites of laser microirradiation and sites of chromatin associated Fok1-nuclease (Figure 1), both of which induce DSBs. The recruitment of both proteins was hindered by PARP inhibition and enhanced for RPS19 with PARG inhibition, suggesting RPS19, in particular, may be parylated or interact with parylated proteins. Similar studies were performed in Saos2 cells, which, in contrast to U2OS cells, are p53 null. Whereas RPS19 was recruited to sites of laser microirradiation in Saos2 cells, RPL5 was not. Interestingly, a transcriptionally hyperactive p53 (∆CTD), but not a wild-type or a dominant-negative form of p53 (p.R248Q), restored RPL5 recruitment in Saos2 cells, suggesting a p53 downstream target is required for its recruitment. We are currently investigating proteins that mediate RPS19 and RPL5 DSB recruitment. Together, our results support a model whereby altered DNA DSB repair in RPS19- and RPL5-deficient cells may contribute to cancer development in DBA.
Read full abstract