Plant genetic effects on microbial hubs impact host fitness in repeated field trials

Abstract

Although complex interactions between hosts and microbial associates are increasingly well documented, we still know little about how and why hosts shape microbial communities in nature. In addition, host genetic effects on microbial communities vary widely depending on the environment, obscuring conclusions about which microbes are impacted and which plant functions are important. We characterized the leaf microbiota of 200 Arabidopsis thaliana genotypes in eight field experiments and detected consistent host effects on specific, broadly distributed microbial species (operational taxonomic unit [OTUs]). Host genetic effects disproportionately influenced central ecological hubs, with heritability of particular OTUs declining with their distance from the nearest hub within the microbial network. These host effects could reflect either OTUs preferentially associating with specific genotypes or differential microbial success within them. Host genetics associated with microbial hubs explained over 10% of the variation in lifetime seed production among host genotypes across sites and years. We successfully cultured one of these microbial hubs and demonstrated its growth-promoting effects on plants in sterile conditions. Finally, genome-wide association mapping identified many putatively causal genes with small effects on the relative abundance of microbial hubs across sites and years, and these genes were enriched for those involved in the synthesis of specialized metabolites, auxins, and the immune system. Using untargeted metabolomics, we corroborate the consistent association between variation in specialized metabolites and microbial hubs across field sites. Together, our results reveal that host genetic variation impacts the microbial communities in consistent ways across environments and that these effects contribute to fitness variation among host genotypes.

The genetic architecture of adaptation to leaf and root bacterial microbiota inArabidopsis thaliana

“…1B). This is in line with the polygenic architecture reported in GWAS conducted on microbial communities (Horton et al 2014; Walters et al 2018; Bergelson et al 2019; Roman-Reyna et al 2020; Deng et al 2021; Brachi et al 2022; VanWallendael et al 2022). Differences in the number of QTLs among the six ‘plant compartment × seasons’ were barely significant (Generalized Linear Model, GLM, F = 2.30, P = 0.0471).…”

“…The efficiency of the local score approach was demonstrated in GWAS conducted in A. thaliana , with the fine mapping down to the gene level and the functional validation of four QTLs associated with quantitative disease resistance to the bacterial pathogen Ralstonia solanacearum (Aoun et al 2020; Demirjian et al 2022). In addition, the local score approach was successfully applied in a recent GWAS on leaf bacterial communities characterized on 200 Swedish accessions of A. thaliana grown in four native habitats in Sweden (Brachi et al 2022).…”

“…All GWAS performed previously on plant microbial communities were conducted on one specific plant compartment (Walters et al 2018; Roman-Reyna et al 2020; Deng et al 2021; Brachi et al 2022; VanWallendael et al 2022), with the exception of one GWAS conducted on worldwide accessions of A. thaliana at the leaf (Horton et al 2014) and root (Bergelson et al 2019) levels. Similar to our results, this latter GWAS showed a small overlap between the leaf and root compartments in candidate genes associated with descriptors of bacterial and fungal communities (Bergelson et al 2019), which is in line with the adaptive differences in microbial community diversity and composition observed among plant niches, from rhizosphere soils to plant canopies (Müller et al 2016; Cregger et al 2018).…”

“…Merging the lists of candidate genes identified in our study and in two GWAs conducted in ecologically relevant conditions in the native area of A. thaliana (Brachi et al 2022) led to the establishment of a short list of 50 core genes associated with natural variation of microbiota traits (Data Set 15). Of particular interest is the gene FLAGELLIN-SENSITIVE 2 ( FLS2 ) overlapping with 39 top SNPs in our study (Data Set 11).…”

The genetic architecture ofArabidopsis thalianain response to native non-pathogenic leaf bacterial species revealed by GWA mapping in field conditions

“…So far, the five GWAS [83][84][85][86] and the single genome-environment association study (GEAS) (Roux et al 2022) conducted on the leaf compartment and using bacterial community descriptors as phenotypic traits, revealed a polygenic architecture controlling microbiota assembly, which is in line with the small percentage of variance explained by the phenotyping of individual mutant lines [87]. In agreement with those association genetic studies and the two GWAS conducted on A. thaliana in response to a PGPB strain [47,49], we identified a complex genetic architecture for the response of A. thaliana to 13 non-pathogenic bacterial strains.…”