Introduction

While conventional intensive agricultural practices allowed yields to drastically increase to feed a growing population, this relied mainly on plant breeding and a large use of inputs (e.g. fertilizers and pesticides). However, inputs have negatively impacted the environment, biodiversity and human health. Now agricultural production faces the challenge of supplying an increasing world population without hampering ecosystems (Tilman et al., 2002). Within the agroecological transition, one objective is to reduce the use of inputs and increase biological diversity and biological regulation without hampering crop production, the nature-based solution (FAO, 2014).

In this context, there is a growing interest in maximizing ecosystemic services through the promotion of beneficial biological interactions, such as plant–microbiome interactions (Singh et al., 2020). The plant microbiome has been shown to improve plant nutrition and health (Lugtenberg & Kamilova, 2009; Bulgarelli et al., 2013; Pieterse et al., 2014). However, plant microbiome is driven by soil type, agricultural practices, environmental conditions, and biotic interactions (Philippot et al., 2013). Plant–microbiome interactions are thereby complex as a plant can also drive its associated microbiome, especially through plant genetics and molecular mechanisms. For example, it has been shown that plant species, plant genotype (Van Overbeek & Van Elsas, 2008; Micallef et al., 2009), plant compartment (Brown et al., 2020; Trivedi et al., 2020), plant developmental stage (Mougel et al., 2006; Edwards et al., 2018) and root exudates (Zhalnina et al., 2018) affect the plant microbiome. Therefore, increasing our knowledge is needed to identify the genetic bases and molecular mechanisms involved in the microbiome recruitment to be able to harness plant microbiome through plant genetics.

For that, two types of approaches have been developed (Bergelson et al., 2021): targeted approaches using mutants and transgenic lines in specific functions; and untargeted approaches using segregating or natural population and quantitative genetics (Horton et al., 2014; Escudero-Martinez et al., 2022). Thereby, genome-wide association studies (GWAS) have been used to identify new loci and genes impacting the microbiome (especially bacterial and fungal communities) in different model or cultivated plant species (i.e. Arabidospis thaliana, maize, rice, sorghum, foxtail millet and switchgrass) and for different plant compartments, such as leaves, roots and more recently the rhizosphere (Horton et al., 2014; Wallace et al., 2018; Bergelson et al., 2019; Roman-Reyna et al., 2020; Deng et al., 2021; Brachi et al., 2022; Meier et al., 2022; Sutherland et al., 2022; VanWallendael et al., 2022; Wang et al., 2022; Andreo-Jimenez et al., 2023; Su et al., 2024). These studies showed heritability especially for beta-diversity components, specific OTUs/ASVs (operational taxonomic unit/amplicon sequence variant) or functions, but rarely on alpha-diversity metrics. Plant genetic bases of microbial interactions were also recently assessed using microbial networks (He et al., 2021; Li et al., 2022). These studies enabled the identification of plant genes related to defense response, kinase activity, cell-wall integrity, root development, trichome formation and nutrition.

While GWAS allowed identifying genes involved in microbiome recruitment to improve plant growth, nutrition, and health, interestingly few recent studies have assessed the effect of plant genetics on the associated microbiome in relation to plant performance (i.e. for Arabidopsis leaf bacterial and fungal communities on mature stem size by image analysis, a proxy for seed production, Brachi et al., 2022; for maize rhizosphere bacterial communities on 15 plant vigor traits, Meier et al., 2022; for foxtail millet root bacterial communities on 12 plant vigor traits, Wang et al., 2022; and for switchgrass rhizosphere bacterial communities on anthesis date and plant height, Sutherland et al., 2022).

In this study, we conducted the first GWAS analysis on the microbiome associated to the model legume, Medicago truncatula, and also considering both plant growth and plant nutritional strategy. We previously showed that the genotype of M. truncatula is affecting especially the rhizosphere bacterial communities, when analyzing both bacterial and fungal communities in the rhizosphere and in the root compartments (i.e. pooling rhizoplane and endosphere) (Zancarini et al., 2013). Therefore, we decided to conduct these GWAS analyses only on the rhizosphere bacterial communities using 16S rRNA gene sequencing and using a core collection of 155 accessions of M. truncatula grown in a Mediterranean soil under controlled glasshouse conditions. First, we characterized the different genotypes of the M. truncatula core collection for their growth and nutritional strategies, and identified their associated plant genetic loci using GWAS. Second, we described their associated rhizosphere bacterial communities, which can be considered as the ‘extended plant phenotype’. Then, we assessed relationships between the plant ecophysiological traits and their associated rhizosphere bacterial community composition to identify bacterial candidates predicting plant phenotypic traits of interest. Finally, we tested whether plant genetic loci are associated with these individual bacterial candidates through GWAS. Our study linked plant single nucleotide polymorphisms (SNPs), its associated rhizosphere bacterial community and plant growth and nutritional strategy.

中文翻译：

---

苜蓿 truncatula 基因型驱动植物营养策略及其相关的根际细菌群落

 介绍

虽然传统的集约化农业实践可以大幅提高产量以养活不断增长的人口，但这主要依赖于植物育种和大量投入品（例如肥料和杀虫剂）。然而，投入对环境、生物多样性和人类健康产生了负面影响。现在，农业生产面临着在不妨碍生态系统的情况下满足不断增长的世界人口的挑战（Tilman 等，2002 年）。在生态农业转型中，一个目标是减少投入的使用，增加生物多样性和生物调节，而不妨碍作物生产，即基于自然的解决方案（粮农组织，2014）。

在这种情况下，人们越来越关注通过促进有益的生物相互作用（例如植物-微生物组相互作用）来最大化生态系统服务（Singh et al.， 2020）。植物微生物组已被证明可以改善植物营养和健康（Lugtenberg & Kamilova，2009;Bulgarelli et al.， 2013;Pieterse et al.， 2014）。然而，植物微生物组是由土壤类型、农业实践、环境条件和生物相互作用驱动的（Philippot et al.， 2013）。因此，植物与微生物组的相互作用很复杂，因为植物也可以驱动其相关的微生物组，特别是通过植物遗传学和分子机制。例如，已经显示植物种类、植物基因型（Van Overbeek & Van Elsas，2008;Micallef等人，2009 年）、植物隔室（Brown等人，2020 年;Trivedi等人，2020 年）、植物发育阶段（Mougel等人，2006 年;Edwards等人，2018 年）和根系分泌物（Zhalnina 等人，2018 年）影响植物微生物组。因此，需要增加我们的知识来确定微生物组募集所涉及的遗传基础和分子机制，以便能够通过植物遗传学利用植物微生物组。

为此，已经开发了两种类型的方法（Bergelson等人，2021 年）：使用具有特定功能的突变体和转基因品系的靶向方法;以及使用分离或自然种群和定量遗传学的非靶向方法（Horton等人，2014 年;Escudero-Martinez et al.， 2022）。因此，全基因组关联研究 （GWAS） 已被用于确定影响不同模式或栽培植物物种（即 Arabidospis thaliana、玉米、水稻、高粱、谷子和柳枝稷）和不同植物隔室（如叶、根和最近的根际）中影响微生物组（尤其是细菌和真菌群落）的新基因座和基因（Horton 等人， 2014;Wallace等人，2018 年;Bergelson et al.， 2019;Roman-Reyna 等人，2020 年;邓等人，2021 年;Brachi等人，2022 年;Meier et al.， 2022;Sutherland等人，2022 年;VanWallendael等人，2022 年;Wang et al.， 2022;Andreo-Jimenez 等人，2023 年;Su et al.， 2024）。这些研究表明，特别是对于 β 多样性成分、特定 OTU/ASV（操作分类单元/扩增子序列变体）或功能具有遗传力，但在 α 多样性指标上很少。最近还使用微生物网络评估了微生物相互作用的植物遗传基础（He et al.， 2021;Li et al.， 2022）。这些研究能够鉴定与防御反应、激酶活性、细胞壁完整性、根发育、毛状体形成和营养相关的植物基因。

虽然 GWAS 允许识别参与微生物组募集以改善植物生长、营养和健康的基因，但有趣的是，最近的研究很少评估植物遗传学对相关微生物组与植物性能的关系（即通过图像分析对拟南芥叶细菌和真菌群落对成熟茎大小的影响，这是种子生产的代理， Brachi等人，2022 年;对于玉米根际细菌群落的 15 种植物活力性状，Meier et al.， 2022;对于谷子根系细菌群落的 12 个植物活力性状，Wang et al.， 2022;对于开花日期和株高的柳枝稷根际细菌群落，Sutherland等人，2022 年）。

在这项研究中，我们对与模型豆科植物 Medicago truncatula 相关的微生物组进行了首次 GWAS 分析，并考虑了植物生长和植物营养策略。我们之前表明，在分析根际和根区室（即汇集根际和内圈）中的细菌和真菌群落时，M. truncatula 的基因型尤其影响根际细菌群落（Zancarini et al.， 2013）。因此，我们决定仅使用 16S rRNA 基因测序和在受控温室条件下在地中海土壤中生长的 155 个元节分枝杆菌种质的核心集合对根际细菌群落进行这些 GWAS 分析。首先，我们表征了 M. truncatula 核心集合的不同基因型的生长和营养策略，并使用 GWAS 鉴定了它们相关的植物遗传位点。其次，我们描述了它们相关的根际细菌群落，可以认为是“扩展植物表型”。然后，我们评估了植物生理生态性状与其相关的根际细菌群落组成之间的关系，以确定预测感兴趣的植物表型性状的细菌候选者。最后，我们通过 GWAS 测试了植物遗传位点是否与这些单独的候选细菌相关。我们的研究将植物单核苷酸多态性 （SNP）、其相关的根际细菌群落与植物生长和营养策略联系起来。