Quantitative Genetics of the Maize Leaf Microbiome

Abstract

The degree to which the genotype of an organism can affect the composition of its associated microbial communities ("microbiome") varies by organism and habitat, and in many cases is unknown. We analyzed the metabolically active bacteria of maize leaves across 300 diverse maize lines growing in a common environment. We performed comprehensive heritability analysis for 49 community diversity metrics, 380 bacterial clades, and 9,042 predicted metagenomic functions. We find that only a few bacterial clades (5) and diversity metrics (2) are significantly heritable, while a much larger number of metabolic functions (200) are. Many of these associations appear to be driven by the Methylobacteria in each sample. Among these heritable metabolic traits, Fisher's exact test identifies significant overrepresentation of traits relating to short-chain carbon metabolism, secretion, and nitrotoluene degradation. Genome-wide association analysis identified a small number of associated loci for these heritable traits, including two that affect multiple traits. Our results indicate that while most of the maize leaf microbiome composition is driven by environmental factors and/or stochastic founder events, a subset of bacterial taxa and metabolic functions is nonetheless significantly impacted by host genetics. Additional work will be needed to identify the exact nature of these interactions and what effects they may have on their host.