Soil metagenomics: In silico and functional analysis of nifH sequences

CALDEROLI PRISCILA; COLLAVINO M.M; AGUILAR O.M

Mar del Plata

Congreso; VIII Congreso Argentino de Microbiología General (SAMIGE); 2012

SAMIGE

Metagenomics is important to understand the ecology and evolution of microbial ecosystems. In this study, we describe the use of bioinformatics tools for the assessment of nifH gene biodiversity in soil metagenomes. Metagenomic libraries were generated from Argentina Pampean Region soils subjected to different agricultural practices (crop rotation with nutrient amendment, monocropping without nutrient reposition, and non-cultivated soil). Additionally, we examined the diversity of the diazotrophics assessing the nifH mRNAs in the RNA isolated directly from soils. To retrieve nifH orthologous genes from the meta-databases two different in silico approaches were applied. Firstly, the construction of a nifH database with all amino acid sequence records nifH related entries from the Genbank Database. The nifH sequences were aligned with ClustalW, visualized and manually corrected using Seaview, and used to build a Hidden Markov model (HMM). Secondly, a functional search of the annotated metagenomes in the MG-RAST metagenomics analysis server was performed. Retrieved hits obtained were visually inspected to remove poorly aligned sequences. Both searches resulted in 22 sequences apparently nifH orthologous with e values < 1e-12. After blastn and tblastx analysis in the NCBI resulted in only 6 nifH sequences, which were compared to the phylogenetic tree by Zher et al., 2003 to assign a phylogenetic position. Altogether these results suggested that the functional groups such as nitrogen fixers are low represented in the metagenomic datasets we examined. Recovered sequences fell into three different groups, two related with the subcluster 1K, and one with 1A. Most (3) 1K sequences were similar to uncultivated soil bacteria. The other 1K subgroup was closely related to Bradyrhizobium sp. Two other sequences were similar to Anaeromyxobacter sequences. About 50 clones carrying nifH cDNA from three differently managed soils were analysed. The nucleotide sequences were confirmed as nifH sequence by using blast programs, and imported into the ARB program to compare with the publically available nifH database (Zehr et al., 2003). We found that the nifH sequences of uncultivated soil showed high diversity and distributed among several subclusters in group I, in particular subgroups 1P, 1A and 1B (69%, 15% and 15% respectively). In contrast, the sequences from the cultivated soils showed lower diversity, mostly similar to the orders Rhizobiales and Burkholderiales, within the sublclusters 1J and 1K. In general, the sequences are grouped according to soil agronomic practices, suggesting that the composition of the active diazotrophic community is strongly affected by agronomic practice. In conclusion, these results show that the diazotrophic community can be studied not only by an in silico approach but also by functional methodology, and that in future both methods should be complementary to assess nitrogen fixation communities in environmental samples.