Study of Mechanisms of Homologous Recombination in Mammalian Cells

Andreas Sturzenegger
Almost every cell of the human body is constantly exposed to DNA damaging agents of exogenous and endogenous origin, which can cause various lesions in the genome. These DNA damages can have mutagenic potential and lead to genomic instability that is one of the major drivers of carcinogenesis. To deal with DNA alterations, cells have developed highly specialized DNA repair pathways. Two distinct mechanisms, termed non-homologous end joining (NHEJ) and homologous recombination (HR), deal with
more » ... (HR), deal with repair of DNA double-strand breaks (DSBs), the most dangerous form of DNA damages. HR is an accurate repair mechanism, but it is restricted to S and G2 phases of the cell cycle. On the other hand, NHEJ can take place throughout the cell cycle and is error-prone. HR reaction relies on an initial 5' to 3' DNA end resection step generating 3' ssDNA overhangs. In yeast, the Exonuclease 1 and the helicase/nuclease Dna2 in conjunction with the RecQ-type helicase Sgs1 constitute two independent pathways for long-range DNA-end resection. In human cells, there are five RecQ homologs known, namely RECQL1, BLM, WRN, RECQL4 and RECQL5. It has already been demonstrated that human BLM can mediate DNA-end resection in conjunction with DNA2. However, data obtained from a study in Xenopus egg extracts identified WRN as the major helicase that works in complex with DNA2 to catalyse DNAend resection. Therefore, we wanted to investigate whether WRN can mediate DNA-end resection in conjunction with DNA2 in human cells. Within this thesis work it is demonstrated that indeed WRN can promote DNA2-mediated DNA end processing and that this process is dependent on the presence of RPA and ATP. In addition, this reaction occurs in 5' to 3' direction generating the anticipated 3' ssDNA overhang required for HR repair. Further, we provide evidence for the physical interaction of WRN and DNA2 in vitro and in vivo. Moreover, we found that WRN and BLM stimulate extensive DNA end resection by DNA2 in vivo. Finally, we show that BLM mediates DNA-end resection as part of the BLM-TOPOIII-RMI1-RMI2 (BTRR) complex. Taken together our data provide evidence that in human cells, DNA2 resects broken DNA ends in conjunction with either WRN or BTRR complex. The 3' ssDNA tail generated by DNA-end resection is subsequently coated with RPA and does not only serve as a substrate for the loading of RAD51 recombinase, but also provides a platform for the activation of ATR kinase that is recruited to RPA-coated ssDNA trough its interacting protein ATRIP. Activated ATR phosphorylates CHK1 thereby inducing the DNA damage checkpoint that leads to a cell cycle stop. Although previous studies show that MutS, a heterodimeric mismatch repair protein composed of MSH2 and MSH3, is involved in the HR process, its exact function is unclear. In this thesis work, it is demonstrated that MutS serves as a mediator in ATR activation upon DNA DSB induction in human cells. We provide evidence that MSH2, MSH3 and ATR/ATRIP form a complex in cells. In addition, we show that siRNA-mediated depletion of MSH2 or MSH3 impairs the phosphorylation of ATR targets and the formation of ATRIP foci in response to replication-associated DNA DSBs. Further, mutations in the mismatch-binding domain of MSH3 diminished the binding of MutS to persistent hairpin loops in RPA-coated ssDNA and compromised ATR activation in vivo. Thus, our results demonstrate that MutS binds to hairpin loop structures persisting in RPA-coated ssDNA at sites of DNA damage and mediates recruitment of the ATR-ATRIP complex for its activation. HR has the potential to generate chromosomal rearrangements trough crossover formation. While crossovers are required in meiotic cells for proper chromosome segregation, in mitotic cells, HR predominantly proceeds via the synthesis-dependent strand annealing (SDSA) pathway that always leads to non-crossover products. This implies that mitotic cells possess a stringent control system for HR subpathway selection. However the underlying molecular mechanism is not well defined. Here, we present evidence that siRNA-mediated knockdown of RECQ5 impairs DNA DSB repair by SDSA in human cells. Further, we provide in vivo and in vitro data demonstrating that RECQ5 helicase counteracts the inhibitory effect of RAD51 on RAD52-mediated ssDNA annealing, a key step of SDSA. Finally, we show that RECQ5 suppresses sister chromatid exchanges in human cells in a complementary manner with the BTRR complex that mediates the dissolution of recombination intermediates to prevent crossovers. These results suggest that RECQ5 acts during the post-synaptic phase of SDSA to prevent formation of aberrant RAD51 filaments on the extended invading strand, thus limiting its channeling into potentially hazardous crossover pathway. Taken together, the results of this thesis work identify new key players and their exact working mechanism in the cellular response to DNA DSBs and thus provide potential new therapeutic targets for cancer therapy.
doi:10.5167/uzh-126045 fatcat:uwg7jyipq5ftvlxezpjqnrqn54