Disentangling leaf-microbiome interactions in Arabidopsis thaliana by network mapping
The leaf microbiota plays a key role in plant development, but a detailed mechanism of microbe-plant relationships remains elusive. Many genome-wide association studies (GWAS) have begun to map leaf microbes, but few has systematically characterized the genetics of how microbes act and interact. Previously, we integrated behavioral ecology and game theory to define four types of microbial interactions - mutualism, antagonism, aggression, and altruism, in a microbial community assembly. Here, we apply network mapping to identify specific plant genes that mediate the topological architecture of microbial networks. Analyzing leaf microbiome data from an Arabidopsis GWAS, we identify several heritable hub microbes for leaf microbial communities and detect 140-728 SNPs responsible for emergent properties of microbial network. We reconstruct Bayesian genetic networks from which to identify 22-43 hub genes found to code molecular pathways related to leaf growth, abiotic stress responses, disease resistance and nutrition uptake. A further path analysis visualizes how genetic variants of Arabidopsis affect its fecundity through the internal workings of the leaf microbiome. We find that microbial networks and their genetic control vary along spatiotemporal gradients. Our study provides a new avenue to reveal the "endophenotype" role of microbial networks in linking genotype to end-point phenotypes in plants. Our integrative theory model provides a powerful tool to understand the mechanistic basis of structural-functional relationships within the leaf microbiome and supports the need for future research on plant breeding and synthetic microbial consortia with a specific function. IMPORTANCEIt is found that plant genes act as microbiome gatekeepers to select which microbes get to live inside the leaves for health. Many genome-wide association studies (GWAS) have begun to map leaf microbes, but few has systematically characterized the genetics of how microbes act and interact. This work illustrates a more comprehensive picture of the genetic architecture underlying the leaf microbiome by network mapping. This study also dissects how genetic variants affect its fecundity by direct path and indirect path through microbial network, revealing the "endophenotype" role of microbial networks in linking genotype to end-point phenotypes. Future studies could benefit from this work to improve understanding the underlying genetic mechanisms that govern the relationships between plants and their microbiomes, and to manipulate plant genetic system to reconfigure microbiome. Plants could become more efficient at selecting their microbial partners to improve their health, resilience, and productivity.