25 (Liu & Muse, 2005). Sequences were deposited in the GenBank database (JQ901106–JQ901377,

Supporting Information, Table S1). Cross-priming was tested for amplification on 10 other strains belonging to 10 Agaricus species (Table 2). PCR conditions were the same as previously described. Zhao et al. (2011) [JF797194] this study [JQ824135] Zhao et al. (2012) [JN204430] this study [JQ824134] Zhao et al. Roscovitine molecular weight (2011) [JF797195] This study [JQ824136] Zhao et al. (2012) [JN204434] Zhao et al. (2011) [JF797188] Kerrigan et al. (2005) [AY899263] A total of 61 757 reads with an average 283 bp were obtained (NCBI SRA accession number SRA050786). Of them, 866 (1.4%) qualified sequences which were non-redundant, longer than 80 bp and containing at least one microsatellite motif with flanking region suitable for primers design, were released. The design of primer pairs was successful for 305 candidate microsatellites (258 perfects, 47 compounds, 0.49% of the starting number of reads) in 272 sequences. This result was lower than those observed in the foundation paper (Malausa et al., 2011) reporting between 1 and 8% of theoretically amplifiable markers in the obtained sequences. This percentage was clearly species-dependent.

We have little hindsight on the efficiency of such an approach on fungi. Only two fungal species were studied in Malausa et al. AUY-922 chemical structure (2011), a basidiomycete Armillaria ostoyae and an oomycete Phytophtora alni many for which 0.93 and 0.7% of amplifiable markers in the obtained sequences were described, respectively. Our results were slightly higher than those obtained for

arbitrary 454 shotgun library (Abdelkrim et al., 2009; Gardner et al., 2011) and may suggest some failure in the enrichment process. However, regarding the distribution of the patterns observed among the 866 qualified sequences, the most commonly found were in agreement with those expected according to the library enrichment with, for example, 36.3 and 27.6% of (AG)n and (AC)n motifs, respectively (Table S2). About 18% of motif types did not match any used for enrichment, but this number was in the same order of magnitude as those described in Malausa et al. (2011). Focusing on AG and AC motifs, the average number of repeats was 6.9 for AG and 6.7 for AC. Whatever the length of the motif, 90.5% of the microsatellite showed a number of repeats lower than 10. This was consistent with previous reports on fungal microsatellite with, for example, 6.2 repeats per locus in A. bisporus (Foulongne-Oriol et al., 2009). The shortness of microsatellite loci in fungi, together with their weak representation in fungal genomes render their isolation arduous (Dutech et al., 2007) and may explain our results. An adaptation of the enrichment protocol with shorter probes could enhance the efficiency of the technique.