Background Soil microbiomes are considered a cornerstone of the next green revolution, and plant growth-promoting bacteria (PGPB) are critical for microbiome engineering. However, taking plant-beneficial microorganisms from discovery to agricultural application remains challenging, as the mechanisms underlying the interactions between beneficial strains and plants in native soils are still largely unknown. Increasing numbers of studies have indicated that strains introduced to manipulate microbiomes are usually eliminated in soils, while others have reported that application of PGPB as inocula significantly improves plant growth. This contradiction suggests the need for a deeper understanding of the mechanisms underlying microbe-induced growth promotion. Results We showed PGPB-induced long-term plant growth promotion after elimination of the PGPB inoculum in soils and explored the three-way interactions among the exogenous inoculum, indigenous microbiome, and plant, which were key elements of the plant growth-promoting process. We found the rhizosphere microbiome assembly was mainly driven by plant development and root recruitments greatly attenuated the influence of inocula on the rhizosphere microbiome. Neither changes in the rhizosphere microbiome nor colonization of inocula in roots was necessary for plant growth promotion. In roots, modification of DNA methylation in response to inoculation affects gene expression related to PGPB-induced growth promotion, and disruptions of the inoculation-induced DNA methylation patterns greatly weakened the plant growth promotion. Together, our results showed PGPB-induced DNA methylation modifications in roots mediated the promotion process and these modifications remained functional after elimination of the inoculum from the microbiome. Conclusion This study suggests a new mechanism in which PGPB affect DNA methylation in roots to promote plant growth, which provides important insights into microbiome–plant interactions and offers new strategies for plant microbiome engineering beyond the perspective of maintaining inoculum persistence in soils. Graphical abstract

DNA methylation is a rapid response strategy promoting plant survival under heavy metal (HM) stress. However, the roles of DNA methylation underlying plant adaptation to HM stress remain largely unknown. Here, we used pokeweed, a hyperaccumulator of manganese (Mn) and cadmium (Cd), to explore responses of plant to HM stress at phenotypic, transcriptional and DNA methylation levels. Mn- and Cd-specific response patterns were detected in pokeweed. The growth of pokeweed was both inhibited with exposure to excess Mn/Cd, but pokeweed distinguished Mn and Cd with different subcellular distributions, ROS scavenging systems, transcriptional patterns including genes involved in DNA methylation, and differentially methylated loci (DML). The number of DML between Mn/Cd treated and untreated samples increased with increased Mn/Cd concentrations. Meanwhile, pretreatment with NADPH oxidase inhibitors prior to HM exposure markedly reduced HM-induced reactive oxygen species (ROS), which caused reductions in expressions of DNA methylase and demethylase in pretreated samples. The increased levels of HM-induced demethylation were suppressed with alleviated ROS stress, and a series of HM-related methylated loci were also ROS-related. Taken together, our study demonstrates that different HMs affect different DNA methylation sites in a dose-dependent manner and changes in DNA methylation under Mn/Cd stress are partly mediated by HM-induced ROS.

Aspergillus includes both plant pathogenic and beneficial fungi. Although endophytes beneficial to plants have high potential for plant growth promotion and improving stress tolerance, studies on endophytic lifestyles and endophyte-plant interactions are still limited. Here, three endophytes belonging to Aspergillus, AS31, AS33, and AS42, were isolated. They could successfully colonize rice roots and significantly improved rice growth. The genomes of strains AS31, AS33, and AS42 were sequenced and compared with other Aspergillus species covering both pathogens and endophytes. The genomes of AS31, AS33, and AS42 were 36.8, 34.8, and 35.3 Mb, respectively. The endophytic genomes had more genes encoding carbohydrate-active enzymes (CAZymes) and small secreted proteins (SSPs) and secondary metabolism gene clusters involved in indole metabolism than the pathogens. In addition, these endophytes were able to improve Pi (phosphorus) accumulation and transport in rice by inducing the expression of Pi transport genes in rice. Specifically, inoculation with endophytes significantly increased Pi contents in roots at the early stage, while the Pi contents in inoculated shoots were significantly increased at the late stage. Our results not only provide important insights into endophyte-plant interactions but also provide strain and genome resources, paving the way for the agricultural application of Aspergillus endophytes.