Analysis of the Monoalkylation and Cross-Linking Sequence Specificity of Bizelesin, a Bifunctional Alkylation Agent Related to (+)-CC-1065

The sequence specificity of bizelesin, an interstrand DNA-DNA cross-linker related to the monoalkylating compound (+)-CC-1065, was studied using restriction enzyme fragments. Bizelesin, like (+)-CC-1065, forms monoalkylation adducts through N3 of adenine but can also form DNA-DNA cross-links six base pairs apart on opposite strands. Compared to many other minor groove cross-linking compounds, bizelesin is very efficient at crosslinking DNA. There is a higher than expected proportion of cross-linked adducts based upon the relative number of cross-linked vs monoalkylated adducts. This is rationalized based upon the relative thermodynamic stability of the cross-linked vs monoalkylated species. Where bizelesin monoalkylation occurs, the sequence specificity is significantly higher than those of (+)-CC-1065 and other monoalkylating (+)-CPI analogs. The bizelesin GC tolerance at crosslinking sites is twice as high as for the monoalkylation sites. This increased GC tolerance can be largely explained by the covalent immobilization of the second alkylation arm at sequences that are not normally reactive toward CPI monoalkylation compounds but are made reactive due to a proximity effect. This same rationale can be used to explain the reactivity of the second alkylation arm of bizelesin with guanine, cytosine, and thymine on some sequence. There are some sequences that appear to be unusual in their reactivity with bizelesin in that bizelesin formed cross-linking spanning seven base pairs, and bizelesin forms monoakylation adducts on guanine. In these cases, it is proposed that bizelesin may trap out rare conformational forms during the second alkylation step, or bizelesin may alkylate unusual sites due to the strong precovalent affinity of bizelesin for those sites.