Failure to respond effectively to replication stress is recognized as a key contributor to developmental defects, premature aging syndromes, and the development of neoplasias. Interstrand crosslinks (ICLs) are particularly dangerous DNA lesions as they are considered absolute blocks to replication, and thus a major challenge to the replication stress response. They are believed to occur as a product of oxidative metabolism, and are also a consequence of treatment with some chemotherapy drugs. Psoralens are photoactive DNA interstrand crosslinkers that have been used clinically for many years. We have synthesized, and demonstrated the activity of, antigen linked psoralens. Laser photoactivation of defined subnuclear regions in cells incubated with the compounds resulted in localized crosslinks. Repair of these adducts was monitored in repair proficient and deficient cells. We are using this approach to follow the recruitment of proteins into sites of crosslink repair. ICLs are repaired in a two cycle process. In the first cycle one DNA strand is uncoupled from the other. In the second cycle the remaining adducted (and still crosslinked) base is removed via conventional Nucleotide Excision Repair. There is uncertainty as to whether repair can occur in G1 phase, in addition to the well established S phase repair. We have shown that crosslinks are repaired in the G1 phase of the cell cycle, in a process that is dependent on NER functions. XPC protein was rapidly recruited to sites of crosslinks and monoadduct. However, the XPE damage binding complex was recruited rapidly to monoadducts and slowly to crosslinks. Recruitment of the XPE complex was dependent on XPC activity, and repair synthesis. Our results support a scenario in which the XPE complex does not recognize the crosslink, but is recruited when the remaining monoadducted base is forced out of the helix after the completion of the first repair cycle. The recruitment of the XPE complex is a marker of completion of the first repair cycle and the start of the second. We have applied this antigen tagged psoralen technology to an examination of the function of FANCD2 in ICL repair. FANCD2 is is the central protein in the Fanconi Anemia pathway. Individuals with deficiencies in this pathway suffer severe developmental defects, and show signs of premature aging during postpartum life. The pathway plays a key role in the response to replication stress. We find that FANCD2 is recruited to laser localized psoralen crosslinks in two modes. One is dependent on components of the well known DNA Damage Response pathway, while the other is dependent on DNA repair functions. Only in the latter pathway does FANCD2 contribute to ICL repair. Entry into the DNA repair pathway is required for the association of FANCD2 with the ICL. Replication independent recruitment of FA proteins to ICLs requires the activity of RNF8, a ubiquitin ligase, and a newly discovered ubiquitin binding protein-FAAP20. This FAAP20-RNF8 ubquitin cascade is important for cellular resistance to genomic stress such as imposed by interstrand crosslinks. Understanding the nature of defects in the Fanconi pathway will provide the basis for developing effective therapies for this disorder. We have also characterized the recruitment and contribution to ICL repair of FAN1, a recently discovered nuclease that associates with FANCD2. Ancestral versions of this protein that lack the protein structural domain involved in FA protein interactions are found in organisms that also lack the FA pathway. It is currently believed that FAN1 recruitment to ICLs is dependent on FANCD2. However we have found that this protein is rapidly recruited to ICLs, in a FANCD2 independent manner. There is a second wave of accumulation that is partially dependent on the association with FANCD2. The protein is also found in replication factories in S phase in the absence of any DNA damage. The results of these and other experiments indicate that the ancestral protein participates in a rapid, multiphasic, response to DNA damage throughout the cell cycle. It also is involved in unstressed replication. The association with the FA pathway appears to reflect the response of FAN1 to stressed replication. FAN1 should be seen as a enzyme that evolved to contribute to many different aspects of DNA metabolism. In order to study the encounter of replication forks with ICLs we have developed a novel single molecule approach for visualizing these events. We have combined well established procedures for displaying replication tracts on DNA fibers with immuno-quantum dot detection of individual antigen tagged psoralen ICLs. We observe single and double fork collisions as well as an unanticipated pattern of DNA synthesis on the side of the ICL distal to the fork encounter. We termed this replication traverse. The Fanconi Anemia translocase, FANCM, is required for the traverse patterns. However, these events are independent of the Fanconi Anemia core complex proteins. As these appear in vertebrate lineages, while FANCM is found in Archaea, we propose that the traverse pathways evolved early in response to major replication challenges. Notably, non ubiquitinated FANCD2 also is required for traverse and functions in the same pathway as FANCM. FANCM associates with a subset of replisomes that encounter the block. These replisomes lose one of the key components of the replication apparatus as a result of binding FANCM. Consequently there is a FANCM dependent remodeling of the replisome as the result of replication challenge. We anticipate that the new technology for visualizing fork encounters with major impediments will be useful for studying replication in tumor cells treated with chemotherapy drugs that induce ICLs.
Showing the most recent 10 out of 31 publications
