Genome-resolved metatranscriptomics reveals conserved root colonization determinants in a synthetic microbiota

Article https://doi.org/10.1038/s41467-023-43688-z Genome-resolved metatranscriptomics reveals conserved root colonization determinants in a synthetic microbiota Received: 9 May 2023 Check for updates 1234567890():,; 1234567890():,; Accepted: 16 November 2023 Nathan Vannier1,3, Fantin Mesny 1,4,5, Felix Getzke1,5, Guillaume Chesneau1, Laura Dethier1, Jana Ordon 1, Thorsten Thiergart1 & Stéphane Hacquard 1,2 The identification of processes activated by specific microbes during microbiota colonization of plant roots has been hampered by technical constraints in metatranscriptomics. These include lack of reference genomes, high representation of host or microbial rRNA sequences in datasets, or difficulty to experimentally validate gene functions. Here, we recolonized germ-free Arabidopsis thaliana with a synthetic, yet representative root microbiota comprising 106 genome-sequenced bacterial and fungal isolates. We used multikingdom rRNA depletion, deep RNA-sequencing and read mapping against reference microbial genomes to analyse the in planta metatranscriptome of abundant colonizers. We identified over 3,000 microbial genes that were differentially regulated at the soil-root interface. Translation and energy production processes were consistently activated in planta, and their induction correlated with bacterial strains’ abundance in roots. Finally, we used targeted mutagenesis to show that several genes consistently induced by multiple bacteria are required for root colonization in one of the abundant bacterial strains (a genetically tractable Rhodanobacter). Our results indicate that microbiota members activate strain-specific processes but also common gene sets to colonize plant roots. As photoautotrophs, plants assimilate carbon into complex organic compounds and invest a substantial amount of these photoassimilates in the rhizosphere, thereby sustaining microbial growth1. Consequently, they host a large diversity of microbes within and on their roots that include primarily bacteria and fungi2–4. Microbiota establishment in roots is governed by a complex interplay between the microbiota, the host, and the environment5–12, as well as by microbe–microbe interactions13–16. Because root-associated microbes have been interacting with host-derived cues for ~450 million years, they have likely evolved complex mechanisms to colonize, survive and thrive in the root environment17,18. While the mechanisms by which model mutualistic and pathogenic microorganisms colonize root tissues have been intensively studied19–21, our knowledge of the processes activated by phylogenetically diverse root commensals to colonize and persist in roots within a microbial community remains fragmented. Bacterial mechanisms of root colonization were recently identified in individual root isolates using multiple approaches, including experimental evolution22, comparative genomics23, randomly-barcoded transposon mutagenesis sequencing24,25 (RB-TnSeq) or transposon insertion sequencing26 (IN-Seq). These bacterial genetic determinants required for root/rhizosphere colonization are primarily involved in chemotaxis towards the roots, primary and secondary attachment to the root surface, and to root hairs27. However, it remains unclear whether these 1 Department of Plant Microbe Interactions, Max Planck Institute for Plant Breeding Research, 50829 Cologne, Germany. 2Cluster of Excellence on Plant Sciences, Max Planck Institute for Plant Breeding Research, 50829 Cologne, Germany. 3Present address: IGEPP, INRAE, Institut Agro, Univ Rennes, 35653 Le Rheu, France. 4Present address: Institute for Plant Sciences, University of Cologne, 50923 Cologne, Germany. 5These authors contributed equally: Fantin e-mail: Mesny, Felix Getzke. Nature Communications | (2023)14:8274 1 Article processes are activated in a community context during root colonization and more importantly, whether diverse root colonizers deploy similar or dissimilar strategies to colonize plant roots. Development of metatranscriptomics-based approaches were successfully used to identify microbial functions expressed ex planta in rhizosphere/rhizoplane habitats under different ecological contexts28–32. However, bacterial transcriptome profiling remains more challenging in planta and has been hampered by numerous technical constraints, already in mono-association experiments (see recent examples for leaf commensals and pathogens)33–35. These include (i) the dominance of plant-derived rRNA and mRNA and (ii) the complexity and diversity of the root microbiota that limit transcriptome coverage of individual strains and prevent unambiguous read mapping between closely related species. These constraints, together with difficulties linked to genome annotations from complex samples, have hampered the investigation of microbial functions and molecular mechanisms concomitantly deployed by multiple strains during root microbiota establishment. The recent establishment of comprehensive culture collections of root-associated bacteria and fungi that are representative of naturally occurring root microbiota13,36,37 offers the opportunity to design multikingdom synthetic communities (SynComs) and to reconstitute the root microbiota of healthy plants under strictly controlled environments13. The use of genome-sequenced microbial SynComs has two main advantages compared to natural samples regarding metatranscriptome profiling: (i) the control over the community composition allows building taxonomically diverse communities with reduced complexity thus yielding higher average transcriptome coverage and (ii) the genomes of the cultured organisms are used as reference for read mapping. Reference-based metatranscriptomics has multiple advantages compared to metatranscriptome assembly including better detection of genes (as transcripts with a low number of reads cannot be assembled), a more accurate read assignment to the different microbiota members and consequently an overall higher number of reads mapped. We hypothesized that this strategy could enable the simultaneous inspection of multi-kingdom transcriptomes (bacteria, fungi, and the host plant) at strain-level resolution in a community context. Here, we repopulated roots of germ-free Arabidopsis thaliana with a multi-kingdom microbial SynCom composed of 84 bacteria and 22 fungi in the gnotobiotic FlowPot system38 and profiled changes in both community structure and the transcriptomes of bacteria and fungi at the soil–root interface. We analyzed microbial transcriptional reprogramming at both strain- and community-level resolution and identified microbial processes induced during root colonization. We experimentally validated that several bacterial genes induced upon host contact by multiple strains contribute to bacterial proliferation at roots. This study identifies conserved and strain-specific processes required for microbial adaptation to the root environment in a community context. Results High-resolution (...truncated)