Trevor J. Gould

Application of Illumina next-generation sequencing to characterize the bacterial community of the Upper Mississippi River.

A next‐generation, Illumina‐based sequencing approach was used to characterize the bacterial community at ten sites along the Upper Mississippi River to evaluate shifts in the community potentially resulting from upstream inputs and land use changes. Furthermore, methodological parameters including filter size, sample volume and sample reproducibility were evaluated to determine the best sampling practices for community characterization.

Systematic improvement of amplicon marker gene methods for increased accuracy in microbiome studies

Amplicon-based marker gene surveys form the basis of most microbiome and other microbial community studies. Such PCR-based methods have multiple steps, each of which is susceptible to error and bias. Variance in results has also arisen through the use of multiple methods of next-generation sequencing (NGS) amplicon library preparation. Here we formally characterized errors and biases by comparing different methods of amplicon-based NGS library preparation. Using mock community standards, we analyzed the amplification process to reveal insights into sources of experimental error and bias in amplicon-based microbial community and microbiome experiments. We present a method that improves on the current best practices and enables the detection of taxonomic groups that often go undetected with existing methods.

Species sorting and seasonal dynamics primarily shape bacterial communities in the Upper Mississippi River.

Bacterial community structure (BCS) in freshwater ecosystems varies seasonally and due to physicochemical gradients, but metacommunity structure of a major river remains understudied. Here we characterize the BCS along the Mississippi River and contributing rivers in Minnesota over three years using Illumina next-generation sequencing, to determine how changes in environmental conditions as well as inputs from surrounding land and confluences impacted community structure. Contributions of sediment to water microbial diversity were also evaluated. Long-term variation in community membership was observed, and significant shifts in relative abundances of major freshwater taxa, including α-Proteobacteria, Burkholderiales, and Actinomycetales, were observed due to temporal and spatial variations. Environmental parameters (e.g. temperature, rainfall, and nutrient concentrations) primarily contributed to differences in phyla abundances (88% of variance), with minimal influence from spatial distance alone (<1% of variance). Furthermore, an annually-recurrent BCS was observed in late summer, further suggesting that seasonal dynamics strongly influence community composition. Sediment communities differed from those in the water, but contributed up to 50% to community composition in the water column. Among water sampling sites, 34% showed significant variability in BCS of replicate samples indicating variability among riverine communities due to heterogeneity in the water column. Results of this study highlight the need for a better understanding of spatial and temporal variations in riverine bacterial diversity associated with physicochemical gradients and reveal how communities in sediments, and potentially other environmental reservoirs, impact waterborne BCS. Techniques used in this study may prove useful to determine sources of microbes from sediments and soils to waterways, which will facilitate best management practices and total maximum daily load determinations.

Bacterial community structure is indicative of chemical inputs in the Upper Mississippi River

Local and regional associations between bacterial communities and nutrient and chemical concentrations were assessed in the Upper Mississippi River in Minnesota to determine if community structure was associated with discrete types of chemical inputs associated with different land cover. Bacterial communities were characterized by Illumina sequencing of the V6 region of 16S rDNA and compared to >40 chemical and nutrient concentrations. Local bacterial community structure was shaped primarily by associations among bacterial orders. However, order abundances were correlated regionally with nutrient and chemical concentrations, and were also related to major land coverage types. Total organic carbon and total dissolved solids were among the primary abiotic factors associated with local community composition and co-varied with land cover. Escherichia coli concentration was poorly related to community composition or nutrient concentrations. Abundances of 14 bacterial orders were related to land coverage type, and seven showed significant differences in abundance (P ≤ 0.046) between forested or anthropogenically-impacted sites. This study identifies specific bacterial orders that were associated with chemicals and nutrients derived from specific land cover types and may be useful in assessing water quality. Results of this study reveal the need to investigate community dynamics at both the local and regional scales and to identify shifts in taxonomic community structure that may be useful in determining sources of pollution in the Upper Mississippi River.

Core functional traits of bacterial communities in the Upper Mississippi River show limited variation in response to land cover.

Taxonomic characterization of environmental microbial communities via high-throughput DNA sequencing has revealed that patterns in microbial biogeography affect community structure. However, shifts in functional diversity related to variation in taxonomic composition are poorly understood. To overcome limitations due to the prohibitive cost of high-depth metagenomic sequencing, tools to infer functional diversity based on phylogenetic distributions of functional traits have been developed. In this study we characterized functional microbial diversity at 11 sites along the Mississippi River in Minnesota using both metagenomic sequencing and functional-inference-based (PICRUSt) approaches. This allowed us to determine how distance and variation in land cover throughout the river influenced the distribution of functional traits, as well as to validate PICRUSt inferences. The distribution and abundance of functional traits, by metagenomic analysis, were similar among sites, with a median standard deviation of 0.0002% among tier 3 functions in KEGG. Overall inferred functional variation was significantly different (P ≤ 0.035) between two water basins surrounded by agricultural vs. developed land cover, and abundances of bacterial orders that correlated with functional traits by metagenomic analysis were greater where abundances of the trait were inferred to be higher. PICRUSt inferences were significantly correlated (r = 0.147, P = 1.80 × 10−30) with metagenomic annotations. Discrepancies between metagenomic and PICRUSt taxonomic-functional relationships, however, suggested potential functional redundancy among abundant and rare taxa that impeded the ability to accurately assess unique functional traits among rare taxa at this sequencing depth. Results of this study suggest that a suite of “core functional traits” is conserved throughout the river and distributions of functional traits, rather than specific taxa, may shift in response to environmental heterogeneity.

Evaluation of water sampling methodologies for amplicon-based characterization of bacterial community structure

Reduction in costs of next-generation sequencing technologies has allowed unprecedented characterization of bacterial communities from environmental samples including aquatic ecosystems. However, the extent to which extrinsic factors including sampling volume, sample replication, DNA extraction kits, and sequencing target affect the community structure inferred are poorly explored. Here, triplicate 1, 2, and 6L volume water samples from the Upper Mississippi River were processed to determine variation among replicates and sample volumes. Replicate variability significantly influenced differences in the community α-diversity (P=0.046), while volume significantly changed β-diversity (P=0.037). Differences in phylogenetic and taxonomic community structure differed both among triplicate samples and among the volumes filtered. Communities from 2L and 6L water samples showed similar clustering via discriminant analysis. To assess variation due to DNA extraction method, DNA was extracted from triplicate cell pellets from four sites along the Upper Mississippi River using the Epicentre Metagenomic DNA Isolation Kit for Water and MoBio PowerSoil kit. Operational taxonomic units representing ≤14% of sequence reads differed significantly among all sites and extraction kits used, although differences in diversity and community coverage were not significant (P≥0.057). Samples characterized using only the V6 region had significantly higher coverage and lower richness and α-diversity than those characterized using V4-V6 regions (P<0.001). Triplicate sampling of at least 2L of water provides robust representation of community variability, and these results indicate that DNA extraction kit and sequencing target displayed taxonomic biases that did not affect the overall biological conclusions drawn.

Associations between soil bacterial community structure and nutrient cycling functions in long-term organic farm soils following cover crop and organic fertilizer amendment

Agricultural management practices can produce changes in soil microbial populations whose functions are crucial to crop production and may be detectable using high-throughput sequencing of bacterial 16S rRNA. To apply sequencing-derived bacterial community structure data to on-farm decision-making will require a better understanding of the complex associations between soil microbial community structure and soil function. Here 16S rRNA sequencing was used to profile soil bacterial communities following application of cover crops and organic fertilizer treatments in certified organic field cropping systems. Amendment treatments were hairy vetch (Vicia villosa), winter rye (Secale cereale), oilseed radish (Raphanus sativus), buckwheat (Fagopyrum esculentum), beef manure, pelleted poultry manure, Sustane(®) 8-2-4, and a no-amendment control. Enzyme activities, net N mineralization, soil respiration, and soil physicochemical properties including nutrient levels, organic matter (OM) and pH were measured. Relationships between these functional and physicochemical parameters and soil bacterial community structure were assessed using multivariate methods including redundancy analysis, discriminant analysis, and Bayesian inference. Several cover crops and fertilizers affected soil functions including N-acetyl-β-d-glucosaminidase and β-glucosidase activity. Effects, however, were not consistent across locations and sampling timepoints. Correlations were observed among functional parameters and relative abundances of individual bacterial families and phyla. Bayesian analysis inferred no directional relationships between functional activities, bacterial families, and physicochemical parameters. Soil functional profiles were more strongly predicted by location than by treatment, and differences were largely explained by soil physicochemical parameters. Composition of soil bacterial communities was predictive of soil functional profiles. Differences in soil function were better explained using both soil physicochemical test values and bacterial community structure data than using soil tests alone. Pursuing a better understanding of bacterial community composition and how it is affected by farming practices is a promising avenue for increasing our ability to predict the impact of management practices on important soil functions.

Structure of bacterial communities in soil following cover crop and organic fertilizer incorporation

Incorporation of organic material into soils is an important element of organic farming practices that can affect the composition of the soil bacterial communities that carry out nutrient cycling and other functions crucial to crop health and growth. We conducted a field experiment to determine the effects of cover crops and fertilizers on bacterial community structure in agricultural soils under long-term organic management. Illumina sequencing of 16S rDNA revealed diverse communities comprising 45 bacterial phyla in corn rhizosphere and bulk field soil. Community structure was most affected by location and by the rhizosphere effect, followed by sampling time and amendment treatment. These effects were associated with soil physicochemical properties, including pH, moisture, organic matter, and nutrient levels. Treatment differences were apparent in bulk and rhizosphere soils at the time of peak corn growth in the season following cover crop and fertilizer application. Cover crop and fertilizer treatments tended to lower alpha diversity in early season samples. However, winter rye, oilseed radish, and buckwheat cover crop treatments increased alpha diversity in some later season samples compared to a no-amendment control. Fertilizer treatments and some cover crops decreased relative abundance of members of the ammonia-oxidizing family Nitrosomonadaceae. Pelleted poultry manure and Sustane® (a commercial fertilizer) decreased the relative abundance of Rhizobiales. Our data point to a need for future research exploring how (1) cover crops influence bacterial community structure and functions, (2) these effects differ with biomass composition and quantity, and (3) existing soil conditions and microbial community composition influence how soil microbial populations respond to agricultural management practices.

Frequencies of heavy metal resistance are associated with land cover type in the Upper Mississippi River

Taxonomic compositions of freshwater bacterial communities have been well-characterized via metagenomic and amplicon-based approaches, especially next-generation sequencing. However, functional diversity of these communities remains less well-studied. Various anthropogenic sources are known to impact the bacterial community composition in freshwater riverine systems and potentially alter functional diversity. In this study, high-throughput functional screening of large (~10,000 clones) fosmid libraries representing communities in the Upper Mississippi River revealed low frequencies of resistance to heavy metals in the following order: Mn2+>Cr3+>Zn2+>Cd2+>Hg2+. No resistance to Cu2+ was detected. Significant, but weak, correlations were observed between resistance frequencies of Cd and Cr with developed land cover (r2=0.08, P=0.016 and r=0.07, P=0.037, respectively). While discriminant function analyses further supported these associations, redundancy analysis further indicated associations with forested land cover and greater resistance to Hg and Zn. Nutrient and metal ion concentrations and abundances of bacterial orders were poorly correlated with heavy metal resistance, except for an association of Pseudomonadales abundance and resistance to Hg and Zn. Taken together, results of this study suggest that allochthonous bacteria contributed from specific land cover types influence the patterns of metal resistance throughout this river.

The lung tissue microbiota of mild and moderate chronic obstructive pulmonary disease

BackgroundOral taxa are often found in the chronic obstructive pulmonary disease (COPD) lung microbiota, but it is not clear if this is due to a physiologic process such as aspiration or experimental contamination at the time of specimen collection.MethodsMicrobiota samples were obtained from nine subjects with mild or moderate COPD by swabbing lung tissue and upper airway sites during lung lobectomy. Lung specimens were not contaminated with upper airway taxa since they were obtained surgically. The microbiota were analyzed with 16S rRNA gene qPCR and 16S rRNA gene hypervariable region 3 (V3) sequencing. Data analyses were performed using QIIME, SourceTracker, and R.ResultsStreptococcus was the most common genus in the oral, bronchial, and lung tissue samples, and multiple other taxa were present in both the upper and lower airways. Each subject’s own bronchial and lung tissue microbiota were more similar to each other than were the bronchial and lung tissue microbiota of two different subjects (permutation test, p = 0.0139), indicating more within-subject similarity than between-subject similarity at these two lung sites. Principal coordinate analysis of all subject samples revealed clustering by anatomic sampling site (PERMANOVA, p = 0.001), but not by subject. SourceTracker analysis found that the sources of the lung tissue microbiota were 21.1% (mean) oral microbiota, 8.7% nasal microbiota, and 70.1% unknown. An analysis using the neutral theory of community ecology revealed that the lung tissue microbiota closely reflects the bronchial, oral, and nasal microbiota (immigration parameter estimates 0.69, 0.62, and 0.74, respectively), with some evidence of ecologic drift occurring in the lung tissue.ConclusionThis is the first study to evaluate the mild-moderate COPD lung tissue microbiota without potential for upper airway contamination of the lung samples. In our small study of subjects with COPD, we found oral and nasal bacteria in the lung tissue microbiota, confirming that aspiration is a source of the COPD lung microbiota.