A holistic view of soil ecosystems through integrative eukaryotic and prokaryotic metatranscriptome analysis

Background/Question/Methods

Soil is one of the most diverse ecosystems, composed of organisms that drive decomposition and carbon cycling; processes influenced by the world’s changing climate. Although the long-term effects of soil warming have been monitored, we still know very little about the composition and interactions of soil organisms that contribute to carbon cycling. In most studies to date, interactions between prokaryotes and eukaryotes have been ignored in regard to relative contributions of soil organisms to carbon cycling. Accuracy in annotating the taxonomic affinity of metatranscriptomic sequences is limited by the available sequenced genomes and the speed in which genomic databases can be searched. Soil eukaryotes are poorly represented in these databases, making it challenging to identify organisms within the soil community. Extensive databases of eukaryotic marker genes have been curated, which can be searched faster than using complete databases. However, different scientific disciplines use different marker genes and some marker genes are not represented in mRNA-based metatranscriptomic data, leading to poorly annotated in metatranscriptomic data even at the phylum level.

Results/Conclusions

Metagenomic and metatranscriptomic data provides a foundation for an exciting “reverse ecology” framework for determining underlying networks of interactions within microbial communities. We have developed strategies to effectively integrate computational analyses of species diversity and microbial function, in order to understand how communities differ. Different scientific disciplines use different marker genes for sequence comparison: 16S rRNA in Bacteria and Archaea, the ITS region of rRNA for fungi, mitochondrial coxI for metazoans, chloroplast rbcL for plants, and 18S rRNA for protists. Marker genes from metatranscriptome datasets were filtered for more detailed taxonomic and phylogenetic analysis. Using a new algorithm implemented in DIAMOND for searches using metagenomic and metatranscriptomic data that is many times faster than BLAST, we are re-examining the taxonomic annotation of mRNA sequences from a metatranscriptomics data sets from soil warming experiments at the Harvard Forest. The abundance of eukaryotic transcripts in forest soil at Harvard Forest is corroborated by other results from the National Ecological Observatory Network (NEON).