Proteogenomic basis for ecological divergence of closely related bacteria in natural acidophilic microbial communities

Abstract

Bacterial species concepts are controversial. More widely accepted is the need to understand how differences in gene content and sequence lead to ecological divergence. To address this relationship in ecosystem context, we investigated links between genotype and ecology of two genotypic groups of Leptospirillum group II bacteria in comprehensively characterized, natural acidophilic biofilm communities. These groups share 99.7% 16S rRNA gene sequence identity and 95% average amino acid identity between their orthologs. One genotypic group predominates during early colonization, and the other group typically proliferates in later successional stages, forming distinct patches tens to hundreds of micrometers in diameter. Among early colonizing populations, we observed dominance of five genotypes that differed from each other by the extent of recombination with the late colonizing type. Our analyses suggest that the specific recombinant variant within the early colonizing group is selected for by environmental parameters such as temperature, consistent with recombination as a mechanism for ecological fine tuning. Evolutionary signatures, and strain-resolved expression patterns measured via mass spectrometry-based proteomics, indicate increased cobalamin biosynthesis, (de)methylation, and glycine cleavage in the late colonizer. This may suggest environmental changes within the biofilm during development, accompanied by redirection of compatible solutes from osmoprotectants toward metabolism. Across 27 communities, comparative proteogenomic analyses show that differential regulation of shared genes and expression of a small subset of the approximately 15% of genes unique to each genotype are involved in niche partitioning. In summary, the results show how subtle genetic variations can lead to distinct ecological strategies.

Fig. 1.

Evolutionary relationships between the Leptospirillum group II genotypic groups. Genotypic groups are defined based on similarity to the UBA (type VI) and five-way CG variants (type I) for which genome sequences are available (21, 24, 44). A1, common ancestor; A2, ancestor of type I; A3, ancestor of type VI. Types I and VI resulted from recombination events between A2 and A3, whereas types II–V were most likely the result from recombination events between types I and VI.

Fig. 2.

Community composition of biofilms in the Richmond Mine. (A) Layout of Richmond mine tunnel system with indication of the sampling locations (gray dots). Each pie diagram represents FISH data for that sample, numbered as in Table S2 and listed chronologically (left to right, up to down) for each sampling location. Numbers next to the sampling locations indicate estimates for the fraction of the total flow at the mine outlet that passes at that point. Black segments indicate archaeal data were not available for these samples. Examples of FISH imaging for samples 11, 14, 17, and 20 are presented in (B), (C), (D), and (E), respectively (purple, Leptospirillum group II UBA; yellow, Leptospirillum group II five-way; red, other Leptospirilla).

Fig. 3.

Nonmetric multidimensional scaling (NMDS) of the biological data. (A) Leptospirillum group II data (fraction of whole community) for the maximum number of samples (Table 1). Final stress from NMDS, 2.8. Circle diameter is scaled based on the developmental stage (DS) of the biofilm. Samples 2 and 49 are represented as cubes, as they have an atypical biofilm morphology that does not fit the standard succession model. Groups: 5CG-Hi, samples with high levels of Leptospirillum group II five-way CG type; L2-Lo, samples with relatively low levels of Leptospirillum group II; UBA-Only, samples for which the Leptospirillum group II population only contained the UBA type. (B) All data including the quantitative estimates for the recombinants determined by PIGT (Table S2), when including maximum number of samples (Table 1). Final stress from NMDS, 14.8. Triangle color indicates the temperature of the AMD solution at the time of sampling. Groups I–VI indicate the dominant Leptospirillum group II genotype present in these samples. In both ordinations, sample groupings are intended as a guide for interpretation and are not statistically supported.

Fig. 4.

Relationship between dN/dS and protein abundance. Average (and standard error) NSAF for all proteins with a dN/dS = x (black) as well as outliers (red, >5× the average NSAF for all proteins with its respective dN/dS) are presented.

Fig. 5.

Strain-resolved proteomic analysis. Average linkage hierarchical clustering of 10 strain-resolved datasets (horizontal) and proteins (vertical) based on the Kendall Tau distance matrix of the proteomics data (S-NSAFs). Blue and red shadings group strain-resolved protein abundance data sets that were used for SAM analysis (Fig. S5).

Fig. 6.

Fine-scale distribution of the two genotypic groups. FISH on a sectioned biofilm from the C +10 m location (November 2006) to reveal the distribution of the Leptospirillum group II UBA (purple) and five-way CG genotypic groups (yellow). (A) Montage of epifluorescence micrographs. (B and C) Confocal microscopy imaging of boxed regions in (A). Not visible are archaea and fungi, which are mixed throughout the biofilm in stream biofilms (23). The biofilm is partly refolded on itself.