G. Orti et al., PHYLOGENETIC ASSESSMENT OF LENGTH VARIATION AT A MICROSATELLITE LOCUS, Proceedings of the National Academy of Sciences of the United Statesof America, 94(20), 1997, pp. 10745-10749
Sixty-six haplotypes at a locus containing a simple dinucleotide (CA),
microsatellite repeat were isolated by PCR-single-strand conformation
al polymorphism from populations of the horseshoe crab Limulus polyphe
mus. These haplotypes were sequenced to assess nucleotide variation di
rectly. Thirty-four distinct sequences (alleles) were identified in a
region 570 bp long that included the microsatellite motif, In the repe
at region itself, CA-number varied in integer values from 5 to 11 acro
ss alleles, except that a (CA)(8) class was not observed. Differences
among alleles were due also to polymorphisms at 22 sites in regions im
mediately flanking the microsatellite repeats, Nucleotide substitution
s in these regions were used to estimate phylogenetic relationships am
ong alleles, and the gene phylogeny was used to trace the evolution of
length variation and CA repeat numbers. A low correlation between siz
e variation and genealogical relationships among alleles suggests that
absolute fragment size (as normally scored in microsatellite assays)
is an unreliable indicator of historical affinities among alleles, Thi
s finding on the molecular fine structure of microsatellite variation
suggests the need for caution in the use of repeat counts at microsate
llite loci as secure indicators of allelic relationships.