All organisms detect DSBs and correct them efficiently. It is important to fully understand how DSBs and other types of lesions contribute to genome instability. REPAIR OF DSBS IN YEAST. Using budding yeast, we were the first to directly characterize DSBs, their repair and genetic control over 35 years ago. Recently, we modified systems for detection and revealed opportunities to address an early step in repair, resection. Since yeast chromosomes can be displayed as individual bands according to size using pulse-field gel electrophoresis (PFGE), repair of individual chromosomes can be addressed. There was little if any restitution of full size chromosomal molecules in G1 diploid cells following ionizing radiation (IR). However, DSBs induced in G2 cells were rapidly repaired: >90% DSBs within 2 hr. Repair requires RAD50, -51 and -52. A critical early step in DSB repair and genome stability is resection of ends. While many studies with yeast characterized resection at a unique DSB using site-specific endonucleases, it has been a challenge to address end events at random, dirty-ended DSBs. We developed a novel approach to address resection of IR induced DSBs. Circular chromosomes linearized by a single, random DSB migrate as a unique band during PFGE; however, within 10 min the band shifts to slower mobility and by 1 hr the apparent size increases 75 kb. This PFGE-shift was identical in WT, rad52 and rad51 strains. Mung bean nuclease digestion revealed the shift was due to resection. There was 1 to 2 kb resection per DSB end during repair in WT cells. In rad52 cells the resection rate was similar. However, in a rad50 mutant lacking the MRX complex, resection of radiation- and HO-induced DSBs was drastically reduced. Importantly, using a 2-D PFGE modification our approach has allowed us to identify for the first time resection at 1 or 2 ends of a DSB. In light of difficulties of precise assessment of events at random DSBs, we set out to examine resection at a defined DSB following expression of endonuclease I-SceI. This led to development of 2-D PFGE, which is useful in addressing many aspects of resection. Using this approach, it is possible to determine whether the in vivo resection giving rise to an intermediate band known as m* is due to limited resection at the 2 ends of a unique DSB or just longer resection at either end. Thus we could establish the role of various repair components in coordinating resection at both ends of a DSB. Using our PFGE approaches, we provide the first system to distinguish resection at 0, 1 or both ends of DSBs. The 0- and 1-end resections predominate in MRX-null, Sae2, and Mre11 nuclease mutants, suggesting new roles for the cancer-related proteins (Ctp1 and MRN in humans) in repair, namely, efficient and coincident/coordinated resection at both ends of a DSB. We proposed that the structural features of the MRX complex are consistent with coincident/coordinated resection being due to an ability to interact with both DSB ends.In the absence of MRX or SAE2, there is loss of coordination of end-resection. Similar to results with a defined DSB, we found resection is a 2-step process: a) initiation of a short, single-strand 3 tail (100 bases) which is determined by the MRX complex and Sae2, and b) processive 5 degradation carried out by Exo1 and Sgs1/DNA2. In sae2 mutants, initiation is reduced dramatically, but this results in only a 2-fold reduction in rate of repair of IR-DSBs and only a modest reduction in survival. Coincident resection at a clean I-SceI-induced break is much less dependent on Mre11 nuclease or Sae2, contrary to strong dependence on MRX complex. These results suggest a differential role for these functions at dirty and clean DSB ends. The observed MRX coordination of end-resection fits nicely with our earlier finding of a role for MRX in physically holding ends of a break in close proximity. Loss of either EXO1 or SGS1 reduces processivity of resection about 2-fold with little effect on DSB repair or survival. However, resection length is severely reduced in an exo1 sgs1 double mutant. Yet, similar to the sae2 mutant, repair of IR-DSBs is only decreased 2-fold and survival remains high, especially as compared to an MRX mutant. Thus, resection appears to be much greater than what is needed for efficient DSB repair, even under conditions of multiple DSBs, and is not rate limiting in overall repair of IR-induced DSBs. Possibly its most important role is providing large ssDNA regions for signaling to prevent cell progression. In a collaboration with the Williams lab, we extended our approaches to the role of the Sae2 homologue Ctp1 and determined the role of various physically characterized mutants on DSB repair in S. pombe. We established that the dimer-of-dimers (THDD) and the C-terminal 'RHR' DNA-interaction motifs are required for efficient DSB repair in S. pombe. TELOMERE PROCESSING. Ends of chromosomes can appear as DSBs including 3 protruding ends. Telomere specific proteins protect ends from resection. In budding yeast, when telomeres become uncapped in a cdc13-1 mutant at the restrictive temperature, the amount of resection can be large and was important in our studiesof mutagenesis in ssDNA, while in humans resection is much smaller (<1 kb). Our approach to addressing 1- and 2-end resections and the ability to quantitate resection length with nucleases provides a useful tool to understanding components determining resection. We initiated studies to investigate telomere resection on individual chromosomes, multiple chromosomes and the role of nucleases. Using our PFGE shift approach, we can address roles for endcapping protein Cdc13 and Ku in preventing resection of 1 or 2 ends of a chromosome, similar to that described above for a DSB in a circle. This provides unique opportunities to address differences in stability of telomere ends among chromosomes, roles of various components in endprocessing (described above) such as Mre11 and Sae2 (Ctp1) and resection lengths, when combined with in vitro nuclease treatment. We recently established that Ku prevents extensive resection from nearly all telomere ends in the genome. Much of the resection in the absence of Ku is 1-end, but our technique allows us to identify resection at both ends as well. There appear to be chromosomal differences in the 1 and 2 end events that is independent of chromosome size. We anticipate being able to integrate commonalities in processing of very different kinds of DSB ends in the same cells: telomeres, clean and dirty DSBs. We have initiated studies with the Bernstein lab at U. Pittsburgh to characterize proteins binding to ends using fluorescent labeled proteins of genes involved in replication and repair. We establish that many ends of the resected chromosomes are can be distinguished, suggesting that differences between DSB ends, which tend to occupy a small space, and telomere ends. GENERALIZING PFGE-SHIFT TO ADDRESS RESECTION IN OTHER ORGANISMS. Prior to our PFGE-shift approach, there were no systems that provid a means for rapid quantitation of resection at random DSBs across the genome, although there were several qualitative foci-based approaches (such as RPA or BRDU) to identify resection in a variety of cell types, including human. Given the large impact of relatively small single-strand tails (i.e., a few 100 bases or more), we have begun to develop a 2-D approach that captures the PFGE-shifting effect of ssDNA tails to identify and quantitate resected molecules across the genome in other organisms without using a circular chromosome. This would provide opportunities to examine components of endprocessing as well as pathway decision-making via resection such as end-joining or recombinational repair. This approach is being applied to fission yeast and eventually to human cells.
Showing the most recent 10 out of 16 publications
