Conservation of polyglutamine tract size between mice and humans depends on codon interruption.

Tandem repeats of amino acids are common in eukaryotic proteins (Green and Wang 1994; Karlin and Burge 1996; Albà, Santibáñez-Koref, and Hancock 1999). However, the functional significance of these repeats remains controversial. They have been suggested to play a role in transcriptional activation (Mitchell and Tjian 1989), and changes in glutamine repeat lengths can affect transcriptional activity of proteins (Gerber et al. 1994; Kazemi-Esfarjani, Trifiro, and Pinsky 1995). Glutamines may form polar zippers and connect transcription factors (Perutz et al. 1994). However, the evolutionary lability of these repeats in a variety of genes, including genes causing human neurodegenerative disorders, has been taken to suggest that they are not functionally important and that their size changes are driven by mutational forces (Green and Wang 1994; Rubinsztein et al. 1994, 1995). Comparative studies of repeat sizes between species have dealt with only a small sample of genes. It is therefore unclear whether the rapid change reported previously is representative of tandem amino acid repeats in general or whether conserved tandem amino acid repeats also exist. To address this question, we compared polyglutamine-encoding regions in a larger sample of genes from mice and humans. The GenBank subset of human and mouse sequences was searched for proteins with tracts of six or more glutamines using BLASTP (Altschul et al. 1990). Redundancy in the data was eliminated by running FASTA (Pearson and Lipman 1988). Sequences with 95% identity were considered redundant, and only one representative was used in subsequent analysis. Discrepancies in the lengths of polyglutamine tracts in nearly identical sequences were resolved by taking the sequence with the longest tract. The sample contained 68 different human and 27 different mouse proteins, yielding 96 polyglutamine tracts for humans and 52 for mice. BLASTP was then used to identify homologous sequences from the other species, and sequence similarity was confirmed using PILEUP (Genetics Computer Group 1997). In total, there were 44 different polyglutamine tracts in human proteins homologous to a mouse database entry, and 49 tracts in mouse proteins homologous to a human database entry; 21 were identified in both species.