Comparative genomic analysis reveals the adaptive traits of Ralstonia spp. in aquatic environments

Abstract

Ralstonia spp. are highly adaptable bacteria that are widely distributed across diverse environments. Here, we isolated four Ralstonia pickettii (R. pickettii) genomes from cultures of Dolichospermum spp., and using a comparative genomic framework of 228 Ralstonia genomes. We performed phylogenetic analyses that grouped them into water, soil, plant, and human-associated clades based on their predominant isolation habitats. Fluorescence in situ hybridization revealed minimal physical interactions between R. pickettii and cyanobacterial cells, indicating a commensal or independent ecological relationship. Distinct differences in carbohydrate-active enzymes (CAZymes) and secondary metabolite profiles were observed between water and human-associated dominant groups compared to plant-associated dominant groups, highlighting potential niche-specific adaptations. The water-associated dominant groups harbored antibiotic resistance genes, including CeoB and OXA-type β-lactamase genes. These genes are typically linked to human-associated strains, suggesting potential horizontal gene transfer or shared selective pressures, and the gene content of T3SS is reduced. Notably, water-associated dominant groups exhibited a unique pyrimidine degradation pathway, potentially enabling the utilization of exogenous pyrimidines to support survival in nutrient-limited aquatic environments. We propose that the gene content loss of T3SS and the acquisition of specialized metabolic pathways reflect adaptive strategies of Ralstonia spp. for thriving in aquatic free-living niches.

Keywords: Ralstonia; antibiotic resistance; comparative genomics; microbial evolution; pathogenic microorganisms.

Figure 1

The relative positional relationship between R. pickettii and Dolichospermum sp. in the culture medium. Using a probe designed based on the 16 s rRNA sequence of R. pickettii, light is activated within the 460–500 nm wavelength range (green), whereas the self-luminescence of blue-green algae (red) is activated within the 510–560 nm (G-2A) wavelength range. Maximum intensity projection of stack fluorescence images acquired at magnifications of ×10 (A), ×20 (B), and ×100 (C) is shown.

Figure 2

Genomic features and phylogenetic reconstruction of Ralstonia spp. genomes. (A) Maximum likelihood phylogenetic tree constructed from a concatenated alignment of 421 single-copy orthologous proteins. Nodes with bootstrap values ≥50 are indicated by solid circles. (B) Distributions of genome size, GC content (%), and number of coding sequences (CDSs) across different clades. Statistical significance was determined using the Wilcoxon rank-sum test: p < 0.05 (*), p < 0.01 (**).

Figure 3

Composition of antibiotic resistance genes (ARGs) and bacterial secretion systems in Ralstonia spp. genomes. The figure shows the distribution of ARGs, virulence factors (VFs), type III secretion systems (T3SS), type IV secretion systems (T4SS), type VI secretion systems (T6SS), and other secretion systems in Ralstonia spp. genomes.

Figure 4

Composition and comparative analysis of carbohydrate-active enzymes (CAZymes) in Ralstonia spp. genomes. (A) Diversity and abundance of CAZymes across different clades. CAZyme genes are classified into six categories: PL (polysaccharide lyases), GT (glycosyltransferases), GH (glycoside hydrolases), CE (carbohydrate esterases), CBM (carbohydrate-binding modules), and AA (auxiliary activities). (B) NMDS (non-metric multidimensional scaling) analysis of enzyme composition among clades based on Bray–Curtis distance.

Figure 5

Overview of the metabolic potential across different habitats. Metabolic pathways were inferred from the orthogroup gene count matrix (excluding outgroups) obtained from OrthoFinder software results. Shared genomic metabolic pathways represent highly conserved sequences present in all genomes. Metabolic differences among clades were assessed using Fisher’s exact test (p < 0.05). Incomplete transporters, identified by sequence loss, were aligned to the corresponding E. coli protein sequences using BLASTp, with identity thresholds above 50%.

Figure 6

Prediction of secondary metabolites in Ralstonia spp. from different habitats based on antiSMASH software result. (A) Average number of secondary metabolites predicted per genome across different habitats. (B) NMDS analysis of clades based on secondary metabolite profiles using Bray–Curtis distances.