Carcinogenesis is linked to alterations in the sequence and organization of the genome, and high genomic instability is a hallmark of cancer cells. Such high level of DNA alterations can be explained by a mutator-based mechanism: inactivation of genes required to preserve the genome (mostly repair or checkpoint ) leads to clonal populations of cells that exhibit a dramatic increase in mutation and acquire a fully malignant phenotype. The DNA damage checkpoint coordinates cell- cycle progression, DNA repair, replication and recombination in response to DNA damage. Defects in this surveillance mechanism lead to genomic instability, cancer susceptibility, ageing and several human pathologies. Moreover, a large part of the therapy for cancer is based upon genotoxic agents. Understanding the mechanistic details of the cellular response to DNA damage is thus critical for the comprehension of tumor development and for the design of better therapeutic approaches. We have been concentrating our research on the damage generated to UV light, which are repaired by NER, as a paradigm for all bulky DNA lesions. We recently showed that in yeast UV lesions cannot trigger a checkpoint response, unless they are first processed by NER. Subsequently, only if the Exo1 nuclease can gain access to the NER intermediates, long ssDNA gaps are produced and the checkpoint response is triggered. We have suggested that closely opposing lesions (two closely spaced lesions affecting the two DNA strands) are possibly responsible for preventing the completion of NER and thus allow Exo1-dependent processing. This mechanism would guarantee that the checkpoint is triggered only if repair is not fully efficient. In our first aim, we propose to determine the nature and genomic locations of the lesions that are specifically processed by Exo1 and to define the correlations of this process with post-replication repair mechanisms. Furthermore, we plan to extend all this analysis to human cells, taking advantage of an ample set of NER- deficient XP mutated cells. Preliminary evidence strongly suggest that human cells adopted a similar strategy. This part of the proposal will also be integrated by a computational modeling approach to describe the molecular interactions and define the dynamics of post-replication repair. Finally, biochemical and genetic screens will be performed to identify new genes and pathways affecting genome stability. The second aim is devoted to study the correlation between defects in RNaseH and genome instability. RNaseH is a family of nucleases that process specifically RNA:DNA hybrid molecules, which are formed during normal cellular events (e.g. transcription, replication) or during potentially pathogenic events (e.g retroviral infection). Our preliminary data indicate that accumulation of these structures affects genome integrity. We have results obtained in yeast that correlate RNaseH to post-replication repair and we propose to exploit the yeast model system to determine the exact mechanisms underlying these effects, and to transfer such knowledge to human cells. The abundance of preliminary evidence and our expertise in the field support our optimistic view on the projects described. If successful, this proposal will provide results that likely have relevant biomedical implications, improving our knowledge on the molecular basis of cancer- prone NER syndromes and on the mechanisms linking control of genome stability to cancerogenesis.