Development of an environmental functional gene microarray for soil microbial communities

Abstract

Functional attributes of microbial communities are difficult to study, and most current techniques rely on DNA- and rRNA-based profiling of taxa and genes, including microarrays containing sequences of known microorganisms. To quantify gene expression in environmental samples in a culture-independent manner, we constructed an environmental functional gene microarray (E-FGA) consisting of 13,056 mRNA-enriched anonymous microbial clones from diverse microbial communities to profile microbial gene transcripts. A new normalization method using internal spot standards was devised to overcome spotting and hybridization bias, enabling direct comparisons of microarrays. To evaluate potential applications of this metatranscriptomic approach for studying microbes in environmental samples, we tested the E-FGA by profiling the microbial activity of agricultural soils with a low or high flux of N₂O. A total of 109 genes displayed expression that differed significantly between soils with low and high N₂O emissions. We conclude that mRNA-based approaches such as the one presented here may complement existing techniques for assessing functional attributes of microbial communities.

FIG. 1.

Schematic view of the E-FGA surface allowing simultaneous gene expression profiling and normalization. A cDNA probe with flanking vector sequences bound to the surface of the microarray is shown. The Alexa Fluor 532-labeled cDNA and the Alexa Fluor 635-labeled nPCR can bind to the same spot, allowing accurate calculation of cDNA levels for probes with variable DNA deposition or hybridization efficiencies.

FIG. 2.

Correlations between replicate spots on a single microarray (A) before normalization (raw F532 values) and (B) after normalization using the nPCR product (F532/F635). The axes indicate (A) absolute foreground fluorescence levels and (B) normalized fluorescence values. Histograms show the relative distribution of each respective data set (inset).

FIG. 3.

Microbial expression profiles comparing (A) two high-N₂O-emitting sites (three soil samples and microarrays for each), (B) two low-N₂O-emitting sites (three soil samples and microarrays for each), and (C) average expression levels for high- versus low-N₂O-emitting sites (six soils samples/microarrays per axis). Each point represents a microarray spot. Normalized gene expression levels are shown.

FIG. 4.

Distribution of 32 N₂O-linked genes in COG functional classifications. Genes which have a significantly greater expression level in soil with high and low N₂O emissions are shown above and below the horizontal line, respectively. For details of genes, see Table 3.