mismatch repair

Even a superb proofreading copy machine makes mistakes, and DNA polymerase is no exception: now and then it inserts a wrong base that escapes its own proofreading, leaving a mismatched pair like a G sitting opposite a T. Mismatch repair is the cleanup crew that walks behind replication, finds these fresh errors, and fixes them — raising the overall accuracy of copying by another hundredfold or so.

The hard part is not spotting the mismatch but knowing which of the two strands is wrong. Both bases are perfectly normal DNA letters; the only thing that distinguishes the new, error-bearing strand from the correct old template is that the new strand was just made. The repair system exploits a temporary mark of newness — in bacteria the new strand is briefly unmethylated, in humans the system recognises the still-open nicks in the freshly synthesised strand. Knowing which strand is new, the system removes a stretch of it spanning the mismatch and resynthesises that stretch correctly against the trustworthy old strand.

Mismatch repair is also the cell's defence against slippage in repetitive sequences, where the polymerase tends to stutter and add or drop a unit, so its failure shows up as unstable, shifting repeat lengths — a fingerprint called microsatellite instability. Inheriting a broken mismatch-repair gene causes Lynch syndrome, the most common form of hereditary colorectal cancer, vividly illustrating the rule that a cell which cannot correct its own copying errors accumulates mutations fast and slides toward cancer.