Refined NrfA phylogeny improves PCR-based nrfA gene detection

Authors: WELSH, A - University Of Illinois; Chee Sanford, Joanne; LOEFFLER, F - University Of Tennessee; SANFORD, R - University Of Illinois

Submitted to: Applied and Environmental Microbiology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/14/2014
Publication Date: 4/1/2014
Citation: Welsh, A., Chee Sanford, J.C., Loeffler, F., Sanford, R. 2014. Refined NrfA phylogeny improves PCR-based nrfA gene detection. Applied and Environmental Microbiology. 80:2110-2119.

Interpretive Summary: The bacterial process of dissimilatory nitrate reduction to ammonium (DNRA) promotes N-retention in the terrestrial nitrogen- (N-) cycle, however it is generally thought that DNRA is not a significant process in most soils. During DNRA, nitrate is reduced to nitrite, the latter then further reduced to ammonium via the enzyme nitrite reductase (NrfA). The goal of this project was to better characterize and determine the presence of bacterial populations potentially mediating DNRA using the nrfA gene as an indicator of these populations in two agricultural soils in Illinois (Urbana and Havana). The two soil types contrast in water retention and drainage, and carbon and nitrogen content. High-throughput gene sequencing technology was used to characterize the diversity of nrfA genes at both field sites in soil taken in November and June at soil depths of 0-5cm, 5-20cm and 20-30cm at three locations 60 m apart within a field. When we looked at the protein sequences of NrfA grouped according to those that were 85% similar, the sand dominated Havana soil was the most diverse with 795 of 1007 groups found, whereas the Urbana silty loam soil had only 447 groups. There were more shared groups between soil depths in the Urbana soil than Havana demonstrating site conditions affect spatial distribution of the NrfA community. For both soil types, the surface depth (0-5cm) shared the least number of NrfA groups with the deeper layer (20-30cm) and the deeper layer had fewer different types of nrfA genes. NrfA communities in Havana soils were clearly separated from Urbana soil NrfA. While microbial populations harboring NrfA were spatially differentiated regardless of sampling time, seasonal differences were not observed. The most abundant NrfA found were most similar to the ones harbored by common soil bacteria like Chthoniobacter flavus, Holophaga foetida and Anaeromyxobacter sp., suggesting these populations were the most important in these soils. The signficance of this study show that DNRA populations are diverse and present in opposing soil types, an unexpected result where the sandy soil (Havana) with low organic matter and aerated conditions would suggest DNRA may be less significant than in a higher organic content soil (Urbana). Populations potentially mediating DNRA vary greatly at field scale and differ between soil types, and most importantly, DNRA in general may be more significant in agricultural soils than previously thought, suggesting this N-retention process may significantly compete for N-fertilizer along with other processes like denitrification, the latter a N-loss process.

Technical Abstract: Dissimilatory nitrate reduction to ammonium (DNRA) promotes N-retention in the terrestrial nitrogen- (N-) cycle. Respiratory nitrite reduction to ammonium is catalyzed by the nitrite reductase NrfA. Prior phylogenetic analyses showed that NrfA divided into18 distinct clades amongst available sequenced bacterial genomes. The goal of this project was to better understand bacterial populations potentially mediating DNRA in two agricultural soils (Urbana and Havana) that contrast in water retention and drainage, and carbon and nitrogen content. The Roche 454 sequencing platform was used to sample nrfA gene diversity at both field sites using DNA extracted from soil taken in November and June at 0-5cm, 5-20cm and 20-30cm depths and in cores 60 m apart within a field. By clustering sequences into operational taxonomic units (OTUs) at 85% amino acid similarity, the sand dominated Havana soil was the most diverse with 795 of 1007 total OTU groups whereas the Urbana silty clay loam soil had only 447 groups. There were more shared OTU groups between soil depths in the Urbana soil than Havana demonstrating site conditions affect spatial distribution of the NrfA community. For both soil types, the surface depth (0-5cm) shared the least number of OTU groups with the deeper layer (20-30cm) and the deeper layer was leastdiverse. A Unifrac algorithm was used to compare the unique fraction of shared branches in a phylogenetic tree of all soilNrfA pair wise for each soil sample and visualized using non-metric multidimentional scaling. NrfA communities in Havanasoils were clearly separated from Urbana soil NrfA. Permutational multivariate analysis of variance showed that when nested within site, significant factors explaining NrfA community structure were position in the field (R2=0.56, p<0.0001) likely driven by physicochemical variations in Urbana north and south parts of the field and soil depth (R2=0.07, p<0.1), however, season was not significant (p=0.810). NrfA in the 20 most abundant OTU groups were most similar to soil bacteria like Chthoniobacter flavus, Holophaga foetida and Anaeromyxobacter sp. in clades J and K. These data reveal significant differences in nrfA functional gene diversity with depth, position in the field, and soil type suggesting that populations potentially mediating DNRA vary greatly at field scale.