The sequencing and comparison of vertebrate genomes have enabled the
identification of widely conserved genomic elements. Chief among these are genes
and cis-regulatory regions, which are often under selective
constraints that promote their retention in related organisms. The conservation
of elements that either lack function or whose functions are yet to be ascribed
has been relatively little investigated. In particular, microsatellites, a class
of highly polymorphic repetitive sequences considered by most to be neutrally
evolving junk DNA that is too labile to be maintained in distant species, have
not been comprehensively studied in a comparative genomic framework. Here, we
used the UCSC alignment of the human genome against those of 11 mammalian and
five nonmammalian vertebrates to identify and examine the extent of conservation
of human microsatellites in vertebrate genomes. Out of 696,016 microsatellites
found in human sequences, 85.39% were conserved in at least one other species,
whereas 28.65% and 5.98% were found in at least one and three nonprimate
species, respectively. An exponential decline of microsatellite conservation
with increasing evolutionary time, a comparable distribution of conserved versus
nonconserved microsatellites in the human genome, and a positive correlation
between microsatellite conservation and overall sequence conservation, all
suggest that most microsatellites are only maintained in genomes by chance,
although exceptionally conserved human microsatellites were also found in
distant mammals and other vertebrates. Our findings provide the first
comprehensive survey of microsatellite conservation across deep evolutionary
timescales, in this case 450 Myr of vertebrate evolution, and provide new tools
for the identification of functional conserved microsatellites, the development
of cross-species microsatellite markers and the study of microsatellite
evolution above the species level.