Available scientific literature reports very few microsatellite markers
derived from tetraploid genomes using de novo transcriptome sequencing,
mostly because their gain usually represents a major computational challenge
due to complicated combinatorics during assembly of sequence reads. Here we
present a novel approach for mining polymorphic microsatellite loci from
transcriptome data in a tetraploid species with no reference genome
available. Pairs of 114 bp long de novo sequenced transcriptome reads of
Centaurium erythraea were merged into short contigs of 170-200 bp each. High
accuracy assembly of the pairs of reads was accomplished by a minimum of 14
bp overlap. Sequential bioinformatics operations involved fully free and
open-source software and were performed using an average personal computer.
Out of the 13 150 candidate contigs harboring SSR motifs obtained in a final
output, we randomly chose 16 putative markers for which we designed primers.
We tested the effectiveness of the established bioinformatics approach by
amplifying them in eight different taxa within the genus Centaurium having
various ploidy levels (diploids, tetraploids and hexaploids). Nine markers
displayed polymorphism and/or transferability among studied taxa. They
provided 54 alleles in total, ranging from 2 to 14 alleles per locus. The
highest number of alleles was observed in C. erythraea, C. littorale and a
hybridogenic taxon C. pannonicum. The developed markers are qualified to be
used in genetic population studies on declining natural populations of
Centaurium species, thus providing valuable information to evolutionary and
conservation biologists. The developed cost-effective methodology provides
abundant de novo assembled short contigs and holds great promise to mine
numerous additional EST-SSR-containing markers for possible use in genetics
population studies of tetraploid taxa within the genus Centaurium.
Pursuit for est microsatellites in a tetraploid model from de novo transcriptome sequencing