Compensatory base change

A compensatory base change is a nucleotide change that offsets the effect of a change at an interacting nucleotide position, often by restoring compatible base pairing. In a simple RNA example, a canonical G–C base pair becomes A–C after one substitution and A–U after a second. The final sequence differs from the original while recovering canonical Watson-Crick base pairing.

This supplies a functional test of a proposed RNA interaction. Compare the original sequences, each altered partner separately and the pair carrying both changes. If the single changes impair the readout while the matched double change restores it, that pattern supports a pairing-dependent explanation. Recovery need not be complete, and restoring complementarity does not ensure that every other molecular interaction is restored.

A 1986 experiment tested recognition of a splice donor site by U1 small nuclear RNA. Researchers introduced altered adenovirus E1A and human U1 genes into HeLa cells, then examined whether complementary changes in U1 rescued splicing defects. Rescue was efficient for the tested change at position $+5$ of one donor but very weak for a change at position $+3$ of another.^[ref] The result supported a required base pairing interaction while showing that complementarity alone was insufficient for normal RNA splicing.

Changes observed across related sequences can also suggest conserved pairing. However, shared ancestry and other sequence constraints can produce correlated changes. A designed rescue comparison tests a more specific functional prediction; neither correlation nor one successful rescue establishes every detail of the molecular contact geometry.