Each was also subject to surface sterilization (designated by an s) to examine just the endophytic community. + indicates if an isolate of that taxa was obtained from a specific sample. Other taxa were isolated from 20% or less of the samples plated (i.e. from just one to four samples) and included various genera that are known plant pathogens (e.g. Agrobacterium, Erwinia,

Leifsonia poae, Xanthomonas) or non-pathogenic symbionts (e.g. Curtobacterium, Massilia, Methylobacterium, Serratia, Stenotrophomonas) [5, 20]. As with Pantoea, these taxa are likely to be specific plant-associated strains, although some of these buy Cilengitide lineages (e.g. Massilia timonae, Serratia, Stenotrophomonas) can include potential human pathogens. Other MDV3100 chemical structure culturable bacteria are probably also present in these samples, given that our isolation strategy focused only on the numerically dominant colonies (i.e. those growing on plates from the greatest dilution), and only on those that appeared morphologically distinct. Use of additional media types may also have led to a greater number of distinct isolates, although the two types of growth medium

used represent both a rich, general purpose media (TSA) and one more commonly used on nutrient poor environmental samples (R2A agar) [24]. That said, while approximately half of the isolates were obtained on R2A agar, all of them were capable of growth on TSA and this medium was eventually used for the maintenance of all cultures. Culture independent analyses A total of 50,339 non-chimeric partial 16S rRNA

gene sequences of >200 bp were obtained from community DNA 454 pyrosequencing. With the use of primers designed to avoid chloroplasts, just 24 of these sequences proved to be chloroplast GSK1120212 clinical trial derived and an additional 16 could FER not be grouped to any recognized bacterial phylum, leaving 50,299 for subsequent analyses, or a mean of 2,515 per sample. Across all samples, a total of 634 OTUs were detected, representing 11 different bacterial phyla (or subphyla in the case of the Proteobacteria; Figure  2). Gammaproteobacteria and Betaproteobacteria were the dominant lineages in almost all leaf vegetable samples, regardless of surface sterilization or agricultural type, and accounted for at least 90% of the sequences obtained in all but three samples (Figure  2). Exceptions were the sample of unsterilized organically grown red leaf lettuce (from which they accounted for 80% sequences obtained), and the samples of both unsterilized and surface sterilized organically grown baby spinach (from which they accounted for 59% and 25% of the sequences, respectively).