COS 69-3 - Symbiont genome decay in response to stress in the free-living environment alters host specificity in a legume-rhizobia mutualism

Tuesday, August 16, 2022

10:30 AM – 10:45 AM EDT

Location: 515B

Presenting Author(s)

AS

Anna Simonsen

Florida International University, Florida, United States

Coauthor(s)

SA

Samuel Andrew

Commonwealth Scientific and Industrial Research Organisation (CSIRO), Australian Capital Territory, Australia
JB

Justin Borevitz

Australian National University, Australian Capital Territory, Australia

Background/Question/Methods

Many soil microbes form non-obligate symbiosis with plant hosts, and are horizontally transmitted to new hosts by surviving as free-living colonies. Since symbiosis confers fitness advantages, such as stable nutrient-rich environments and protection from free-living abiotic (i.e. heat, drought) and biotic stressors (i.e. competitors, viruses), selective pressures in free-living and symbiosis stages will differ greatly. However, eco-evolutionary feedback between these major life-history stages is poorly understood. Using legume-rhizobia mutualisms, we show that microbial evolutionary responses to abiotic stress associated with free-living soil conditions also impacts symbiosis specialisation patterns on numerous legume host species. To examine eco-evolutionary dynamics, 374 strains of a widespread nitrogen-fixing rhizobial species (Bradyrhizobium diazoefficiens, a root symbiont of hyperdiverse legume genus Acacia), were first isolated from soils spanning four natural stress gradients (soil salinity, heat, drought and acidity). Since bacterial populations have flexible pangenomes (i.e. some strains have unique genes absent in other strains), we measured evolutionary responses by quantifying patterns of gene presence/absence from strain genomes across all stress gradients. Evolutionary effects on symbiosis function were tested by inoculating 21 Acacia species (from diverse ecoregions) in the glasshouse with a strain subset (n=220) and quantifying strain occupancy from sequenced root nodules.

Results/Conclusions

Rhizobia strains isolated from higher stress environments tended to have fewer protein coding genes (hence smaller genomes) indicating that all abiotic stressors lead to continuous reductions in genome size. Using gene-level network and duplication properties to predict the distribution of genes across natural stress gradients, we found non-random patterns of gene loss: genes with redundant functions were lost in high stress, while multi-functional role genes were retained. Gene loss was widespread across the genome, with hotspots of low gene loss in close spatial proximity to core genes, suggesting that Bradyrhizobium has evolved to cluster essential-function genes in discrete regions to maintain viability during genomic decay. Follow-up inoculation experiments indicate that strains with smaller genomes exhibit stronger patterns of specialisation to particular Acacia species, while strains with larger genomes infect most Acacia species (a generalist symbiosis pattern). Together, these data suggest that genomic decay (through loss of protein-coding genes) is an important evolutionary response to abiotic free-living stress, which alters symbiosis function. This raises questions about impacts of genome decay on the adaptive capacity of bacterial populations, since abiotic stress appears to reduce a symbionts’ generalist capacity, potentially reducing its flexibility in response to rapidly changing conditions.