The findings, says Aggarwal, “explain the two mysteries of this polymerase: why it works so well with template G, and why it only puts C opposite it.” The group determined the crystal structure of yeast Rev1 in complex with template DNA and its favorite incoming nucleotide, dCTP. They found several features that distinguish Rev1 from standard eukaryotic polymerases.
First, Rev1 is its own template. An arginine residue within Rev1 acts like a surrogate template G by forming hydrogen bonds with the incoming C. Any other base results in steric hindrance and unfavorable electrostatic interactions. “The paradigm is that coding is provided by the DNA sequence,” says Aggarwal. “Here, the protein dictates what comes in.”
In fact, the incoming C initially does not even contact the template G, which the group found is rotated out of the DNA helix by Rev1. The correct template is ensured, however, by hydrogen bonding between this twisted G and a part of Rev1 called the G-loop. Bases other than G would create steric hindrance, although an empty sugar backbone would not, which is consistent with the known ability of Rev1 to add a C opposite an abasic site.
The twist of the G also explains how Rev1 is able to polymerize through damaged DNA containing N2-adducted Gs, as the N2 group is turned away from Rev1. These adducts are created by common carcinogens, including those in cigarette smoke. Defects in human Rev1 might be associated with increased stalling of replication at these adducts, resulting in DNA breaks and eventually leading to cancers.