New perspectives on viable microbial communities in low-biomass cleanroom environments

Abstract

The advent of phylogenetic DNA microarrays and high-throughput pyrosequencing technologies has dramatically increased the resolution and accuracy of detection of distinct microbial lineages in mixed microbial assemblages. Despite an expanding array of approaches for detecting microbes in a given sample, rapid and robust means of assessing the differential viability of these cells, as a function of phylogenetic lineage, remain elusive. In this study, pre-PCR propidium monoazide (PMA) treatment was coupled with downstream pyrosequencing and PhyloChip DNA microarray analyses to better understand the frequency, diversity and distribution of viable bacteria in spacecraft assembly cleanrooms. Sample fractions not treated with PMA, which were indicative of the presence of both live and dead cells, yielded a great abundance of highly diverse bacterial pyrosequences. In contrast, only 1% to 10% of all of the pyrosequencing reads, arising from a few robust bacterial lineages, originated from sample fractions that had been pre-treated with PMA. The results of PhyloChip analyses of PMA-treated and -untreated sample fractions were in agreement with those of pyrosequencing. The viable bacterial population detected in cleanrooms devoid of spacecraft hardware was far more diverse than that observed in cleanrooms that housed mission-critical spacecraft hardware. The latter was dominated by hardy, robust organisms previously reported to survive in oligotrophic cleanroom environments. Presented here are the findings of the first ever comprehensive effort to assess the viability of cells in low-biomass environmental samples, and correlate differential viability with phylogenetic affiliation.

Figure 1

Venn diagram showing the MOTU detected in various samples of the cleanrooms. Comparison (a) between SAF and Bldg 144 cleanroom floor samples and (b) between floor and GSE samples of SAF cleanroom. Parentheses denote total number of pyrosequences generated and the numerals without parentheses are total number of MOTU present in that sample.

Figure 2

Heatmap of PTU that increased in transformed hybridization intensities in PMA-treated samples compared with non-PMA-treated samples and were called present in the PMA-treated sample. An increase in transformed hybridization intensities in PMA-treated sample is reflected as a positive ratio. In total, 801 PTU were identified that fulfilled this requirement, which were grouped into 70 genera. Displayed are representatives of all genera with the most drastic changes for each sample pair. Numbers in parentheses are the total number of PTU.

Figure 3

PCoA based on: (a) number of pyrosequences per MOTU (PCoA1, percentage of explained variance: 22% PCoA2, percentage of explained variance: 17%) and (b) PhyloChip-derived transformed hybridization scores of each PTU (PCoA1, percentage of explained variance: 87% PCoA2, percentage of explained variance: 6%). Open and closed dots represent PMA-treated and PMA-non-treated samples, respectively.