Genome evolution following an ecological shift in nectar-dwelling Acinetobacter

Abstract

The bacterial genus Acinetobacter includes species found in environmental habitats like soil and water, as well as taxa adapted to be host-associated or pathogenic. High genetic diversity may allow for this habitat flexibility, but the specific genes underlying switches between habitats are poorly understood. One lineage of Acinetobacter has undergone a substantial habitat change by evolving from a presumed soil-dwelling ancestral state to thrive in floral nectar. Here, we compared the genomes of floral-dwelling and pollinator-associated Acinetobacter, including newly described species, with genomes from relatives found in other environments to determine the genomic changes associated with this ecological shift. Following one evolutionary origin of floral nectar adaptation, nectar-dwelling Acinetobacter taxa have undergone reduction in genome size compared with relatives and have experienced dynamic gene gains and losses as they diversified. Gene content changes suggest a shift to metabolism of monosaccharides rather than diverse carbohydrates, and scavenging of nitrogen sources, which we predict to be beneficial in nectar environments. Gene gains appear to result from duplication events, evolutionary divergence, and horizontal gene transfer. Most notably, nectar-dwelling Acinetobacter acquired the ability to degrade pectin from plant pathogens, and the genes underlying this ability have duplicated and are under selection within the clade. We hypothesize that this ability was a key trait for adaptation to floral nectar, as it could improve access to nutrients in the nutritionally unbalanced habitat of nectar. These results identify the genomic changes and traits coinciding with a dramatic habitat switch from soil to floral nectar.

Importance: Many bacteria, including the genus Acinetobacter, commonly evolve to exploit new habitats. However, the genetic changes that underlie habitat switches are often unknown. Floral nectar is home to specialized microbes that can grow in this nutritionally unbalanced habitat. Several specialized Acinetobacter species evolved from soil-dwelling relatives to become common and abundant in floral nectar. Here, we investigate the genomic adaptations required to successfully colonize a novel habitat like floral nectar. We performed comparative genomics analyses between nectar-dwelling Acinetobacter and Acinetobacter species from other environments, like soil and water. We find that although gene loss coincided with the switch to living in nectar, gains of specific genes from other bacteria may have been particularly important for this ecological change. Acinetobacter living in nectar gained genes for degrading pectin, a plant polysaccharide, which may improve access to nutrients in their environment. These findings shed light on how evolutionary novelty evolves in bacteria.

Keywords: Acinetobacter; evolution; genomics; microbial ecology; plant–microbe interactions.

Fig 1

Acinetobacter species maximum-likelihood phylogenomic tree based on 399 conserved protein sequences. All nodes have a bootstrap value of 100. Ancestral state reconstruction of ortholog gains and losses at nodes within the nectar-dwelling clade is shown. Nodes are labeled numerically, and a pie chart at each node shows the proportion of ortholog gains (yellow) versus losses (blue). Pie size is scaled by the total number of gain and loss events. Ortholog gains and losses at tips are shown in Fig. S1).

Fig 2

Heatmap of the number of orthologs in each functional category across all genomes. Category assignments are based on RAST annotations (FA = fatty acid metabolism, TE = transposable elements), and the phylogenomic species tree (Fig. 1) shows nectar-dwelling versus environmental isolates. Each row is colored independently based on the variance of ortholog number within the category. The color scale has yellow indicating the highest number of genes and deep blue indicating the lowest number of genes in each row. Genes of unknown function or categories with <6 mean orthologs per genome were excluded, as were uncategorized orthologs and the category of miscellaneous. The range of ortholog numbers across genomes for each category is shown in the right-side column. Specific orthologs are given in Table S2, and statistical analysis of ortholog differences is given in Table S4.

Fig 3

Maximum-likelihood phylogenetic trees of (A) polygalacturonase genes and (B) pectin lyase genes from nectar-dwelling Acinetobacter isolates and the most closely related orthologs from plant pathogens (Table S6). Red bars indicate likely duplication events, and blue bars indicate horizontal gene transfer events. Bootstrap values below 80 are displayed at nodes. Asterisks show nodes and tips with significant positive selection. Outgroup orthologs represent the best BLAST hits in GenBank databases.

Fig 4

Protein structure predictions of polygalacturonase genes under selection in nectar-dwelling Acinetobacter. Shown are representative structures from four clades (Fig. 3) with sites under positive selection detected in either specific orthologs (sites shown in green with an *) or ancestral nodes (sites shown in pink): (A) A. pollinis FNA3 (locus_tag I2F29_RS12745), (B) A. apis ANC 5114 (locus_tag CFY84_RS01715), (C) A. pollinis FNA3 (locus_tag I2F29_RS12925), and (D) A. pollinis FNA3 (locus_tag I2F29_RS02465). Conserved active site motifs, from published work (82), are highlighted in blue. Secretion signal tags, based on protein domain predictions from Uniprot, are shown in gray, and substrate-binding clefts are shown in yellow (83). Proteins are rotated to best show active sites and sites under selection. For A–C, insets show zoomed in views of the binding cleft and sites under selection.