Evidence for a recombination-independent pathway for the repair of DNA interstrand cross-links based on a site-specific study with nitrogen mustard.

DNA-DNA interstrand cross-links are thought to be important for the cytotoxicity of many chemotherapeutic agents. To study this more definitively, adduct site-specific methods are used to construct a plasmid with a single nitrogen mustard interstrand cross-link (inter-HN2-pTZSV28). Replication efficiency (RE = [colonies from (inter-HN2-pTZSV28)/(control with no cross-link)]) is approximately 0.3 following transformation into Escherichia coli, implying that the cross-link is repaired. The commonly accepted pathway for cross-link repair, which involves both nucleotide excision repair (NER) and recombination, is ruled out since RE is approximately 0.3 in a delta recA strain. Non-RecA-directed recombination such as copy-choice is also unlikely. However, NER is involved since RE was approximately 0.02 in strains deficient in NER. Base excision repair is not important since RE is approximately 0.3 in strains deficient in 3-methyladenine DNA glycosylases I and II, FAPY DNA glycosylase, both known apurinic/apyrimidinic endonucleases, or DNA deoxyribophosphodiesterase. Another hypothetical repair pathway hinging on a 5' --> 3' exonuclease activity is unlikely since RE is approximately 0.3 in cells deficient in either the 5' --> 3' exonuclease activities of DNA polymerase I, exonuclease VII, or RecJ. Thus, aside from NER, it is unclear what else participates in this recombination-independent repair pathway, although a pathway involing NER followed by replicative bypass of the lesion is the current working hypothesis. Psoralen interstrand cross-links appear not to be repairable by this second pathway, which may have implications for the relative cytotoxicity of interstrand cross-links from different agents.