HOMOPOLYMER LENGTH VARIATION IN THE DROSOPHILA GENE MASTERMIND

Runs of identical amino acids encoded by triplet repeats (homopolymers ) are components of numerous proteins, yet their role is poorly unders tood. Large numbers of homopolymers are present in the Drosophila mela nogaster mastermind (mam) protein surrounding several unique charged a mino acid clusters. Comparison of mam sequences from D. virilis and D. melanogaster reveals a high level of amino acid conservation in the c harged clusters. In contrast, significant divergence is found in repet itive regions resulting from numerous amino acid replacements and larg e insertions and deletions. It appears that repetitive regions are und er less selective pressure than unique regions, consistent with the id ea that homopolymers act as flexible spacers separating functional dom ains in proteins. Notwithstanding extensive length variation in interv ening homopolymers, there is extreme conservation of the amino acid sp acing of specific charge clusters. The results support a model where h omopolymer length variability is constrained by natural selection.