High frequency of microsatellites in Drosophila pseudoobscura
Genes & Genetic Systems
Using 30,000 bp of anonymous sequence data, we note that dinucleotide repeat arrays appear to be much more common in Drosophila pseudoobscura than in D. melanogaster or D. simulans. Repeat arrays bearing five or more units are situated on average once every 3000 bp in D. pseudoobscura, and repeat arrays bearing ten or more units are situated on average once every 7500 bp. We did not detect an association between microsatellite presence and GC-content of flanking regions. While developing
... e developing molecular markers for quantitative trait locus (QTL) mapping in Drosophila pseudoobscura, we incidentally noted the presence of numerous dinucleotide repeat arrays. Repetitive elements in DNA sequences, and microsatellites in particular, have attracted much recent attention, both because of their unusual patterns of mutation and because of their utility as genetic markers (e.g., Jarne and Lagoda, 1996; Goldstein and Schlotterer, 1999) . Recent studies in Drosophila melanogaster and other Drosophila species have shown that microsatellites generally have a lower mutation rate than observed in mammals (Schug et al., 1997; Wetterstrand, 1997; Schug et al., 1998a; Noor et al., 2000) . They are distributed throughout the D. melanogaster genome at an average frequency of about one every 60 kilobases (Schug et al., 1998b) and in the D. simulans genome at an average frequency of about one every 290 kilobases (Hutter et al., 1998) . The length, presence, and/ or variability of microsatellites has been shown to correlate with %GC content in alligators, humans, and Drosophila melanogaster (Glenn et al., 1996; Bachtrog et al., 1999; Brock et al., 1999) , but not in wasps (Ezenwa et al., 1998) . Given the stronger selective constraints imposed by coding regions of the genome than noncoding regions, one might predict that, among randomly selected genomic regions, microsatellites would be more common in low %GC (typically noncoding) regions than high %GC (coding) regions. We had initially sought to develop new codominant molecular markers in Drosophila pseudoobscura using a modification of the method of Karl and Avise (1993) . After this work was begun and numerous sequences were obtained from D. pseudoobscura, more efficient means of developing new markers were published (e.g., , and we abandoned this objective. However, we present our observations from the extensive random sequence data we obtained regarding the frequency of dinucleotide repeat arrays in D. pseudoobscura and the relationship between %GC content and microsatellite presence.