PREDICTING, DETECTING AND CHARACTERIZING NUTRIENT EXCHANGES BETWEEN MAIZE AND SYNTHETIC BACTERIAL COMMUNITIES

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: ACTIVE
• Funding Source: AFRI COMPETITIVE GRANT
• Reporting Frequency: Annual
• Accession No.: 1030311
• Grant No.: 2023-67017-39419
• Cumulative Award Amt.: $865,000.00
• Proposal No.: 2022-11095
• Project Start Date: Apr 1, 2023
• Project End Date: Mar 31, 2026
• Grant Year: 2024
• Program Code: [A1402]- Agricultural Microbiomes in Plant Systems and Natural Resources

Recipient Organization
DONALD DANFORTH PLANT SCIENCE CENTER
975 NORTH WARSON ROAD
ST. LOUIS,MO 63132

Performing Department
(N/A)

Non Technical Summary
Plants are part of a community that includes fungi, oomycetes, bacteria, viruses and other complex microbiota. Although the number of microbial cells on a plant or animal often outnumbers the host's own cell count by many-fold, microbes are generally not visible to the naked eye. It has been estimated that a single gram of soil may contain 10 billion bacterial cells and plant roots can harbor thousands of bacterial species. While some microbial interactions are detrimental to plant health, symbiotic relationships with bacteria or fungi supply mineral elements to plants and can also provide tolerance to various abiotic and biotic stresses. Capitalizing on the benefits of a healthy microbiome can be traced back several millennia and currently, hundreds of agricultural products are being sold as 'beneficials'. However, these products show inconsistent performance in the field. Nonetheless, unexplored microbes represent one of the greatest untapped opportunities in plant biology. Plant microbiome research has the potential to positively impact agricultural productivity that will feed future generations while mitigating global climate change, restoring natural ecosystems, and reducing energy intensive fertilizer production. Our ability to manipulate these complex ecosystems for the benefit of the plant and planet is in its infancy.Agricultural productivity depends on intensive farming and management practices including fertilizers. U.S. farmers rely on this technology to meet yield demands. This proposal will partially offset fertilizer demands through plant-microbe based nutrient acquisition. Specifically, this research will yield new strategies to develop microbial products that show robust performance in the field. The key to success will be discovering the fundamental rules that govern colonization and persistence (C&P) of beneficial microbes with their plant hosts. If we are able to develop strategies to increase the C&P of microbial based products, this will help farmers achieve high yields with less external inputs while also protecting their soil health and the environment. The long-term outcome of this research will be to optimally redesign production agriculture through the lens of the 'community of players' that can ensure a sustainable planet.

Animal Health Component

30%

Research Effort Categories

Basic

70%

Applied

30%

Developmental

(N/A)

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
206	0110	1100	100%

Knowledge Area
206 - Basic Plant Biology;

Subject Of Investigation
0110 - Soil;

Field Of Science
1100 - Bacteriology;

Keywords

beneficial bacteria

maize

synthetic communities

Goals / Objectives
Sustainable agricultural practices are necessary to mitigate climate change and produce more food, fiber, and renewable fuels. An untapped frontier in agriculture is to promote beneficial interactions with microbes as a sustainable mechanism of providing nutrients to crops. 'Microbial' products are already available to farmers; however, in practice, these products display highly variable results and current data suggest that this is tightly linked to the ability of the beneficial microbes to colonize and persist with their plant hosts. Plant root exudates are a major factor that determine the colonization and persistence phenotypes of exogenously supplied microbes; however, our knowledge of how plant exudates affect colonization and persistence is too limited to be translated into improved performance of microbial products. This proposal combines field level, computational observations with lab based, reductionist experiments. Together, these approaches will yield new synthetic communities of bacteria with plant growth promoting abilities and fundamental knowledge of how plant root exudates affect the colonization and persistence of the synthetic communities.In the proposed studies, multivariate datasets on root and shoot phenotypes from a field trial of 16 diverse maize lines will be leveraged with 323 candidate beneficial bacteria identified in the same study to computationally predict co-occurring or exclusionary (never co-occurring) microbial community members that lead to enhanced growth and plant productivity. Microbes within the communities will be isolated and characterized by genome sequencing and high throughput bioassays and nutrient exchange between plant roots and the microbial community will be quantified using isotopic labeling and metabolite profiling.The objectives of this proposal are as follows:1: Query existing field datasets to identify key components of robustly colonizing beneficial bacterial communities.2: Targeted culturing of root microbiome constituents and assembly of SynComs.3: In depth identification & metabolic characterization of the community players and interaction with maize over development.

Project Methods
Aim 1:We previously computationally predicted 323 candidate beneficial bacteria ("do-ers"). Next, we will identify potential nitrogen fixers using PCR and primers conserved for the nifH (nitrogenase) gene. We will use qPCR in 384 well format to rapidly identify which tubes contain bacteria with the nifH gene and further, which tubes contain the highest abundance of this important molecular marker. We will search for "helpers" for each "do-er" through co-occurrence and pairwise relative abundance correlations. Correlation analysis will be conducted using the Spearman coefficient, a non-parametric rank-based correlation metric, to accurately compare sample's microbial compositions across a variety of underlying distributions. Highly positive or negative correlation coefficients will be used to identify microbes to examine through network analysis. The network analysis will connect microbes (nodes) with strong correlations (edges) to other microbes to produce a clearer view of how these bacteria form communities in the field. We will also look for metabolite abundance - microbe abundance correlations as a hypothesis generating activity.Aim 2:To obtain pure cultures of the target bacteria, we will look for taxonomically related isolates within our existing culture collections. In addition, to isolate target bacteria, we will query the 16s rRNA metagenomic sequence data from root samples to identify the glycerol stock with the highest percentage of the target microbe. These samples can be used in brute force culturing. Alternatively, recently we have transitioned to a new approach, called 'limiting dilution" culturing. Frozen plant root samples are ground and diluted into the culture media and aliquots of the sample are cultured in 96-well plates. With the optimal dilution, most of the wells contain a pure culture, starting from a single bacterial cell. Each well is used as a template for a two-side barcode PCR, sequenced using NGS and then de-multiplexed to identify the bacteria in each well.Isolated strains will be sequenced and characterized through various assays. We will isolate gDNA using CTAB extraction. Multiple samples will be multiplexed and sequenced to obtain ~50x coverage. Illumina reads will be trimmed with Trimmomatic and assembled using SPAdes. Assembled contigs will be annotated using the KBase implementation of Prokka. Annotated genomes will be analyzed for genes and pathways potentially involved in interaction with plants and/or other bacteria, such as genes involved in nitrogen fixation, biofilm production, etc. Characterization assays include growth rate in culture, growth on low nitrogen media, production of indole-3-acetic acid (IAA), 1-aminocyclopropane-1-carboxylic acid (ACC) deaminase activity, biofilm formation, swarming/motility and a rapid seedling assay.We will collate the data generated and assemble SynComs. Each SynCom will have at least one 'do-er' and many 'helpers.' Community evolution will be assessed over the course of 7 days. Each day, an aliquot will be removed from the SynCom, and relative abundance of each component assessed via 16S rRNA gene profiling. We have developed specific PCR primers for this purpose with unique barcodes and can thus multiplex samples and perform these assays cheaply. These data will provide direct evidence of potential synergy or antagonism between the SynCom strains, albeit in in vitro conditions. In addition to in vitro growth dynamics, we will assess the ability of each SynCom to colonize and persist with maize roots over the course of 7 days using a seedling assay.Aim 3:Seed treatments are the preferred method of introducing beneficial microbes into agriculture and a germinating seed offers a unique opportunity for microbiome establishment. Though the importance of germination is accepted, quantitative mechanistic descriptions of germination are difficult to achieve, much less microbial influences on those processes. To this end, we tested whether deuterated water (D2O) can effectively substitute for H2O, the required 'in planta' substrate to initiate germination. In reasonably short duration experiments (36 hours), D2O was incorporated at significant levels into central intermediates (e.g., malate) of seeds with protruding roots. Here, we will further develop this assay for use with SynComs. We will select the four most promising SynComs based on data collected through Aim 2. SynCom treated seeds will be grown in sterile chambers for minimally 36 hours at 27°C at which time the exposed root will be removed, and seed and root will be separately quenched and extracted to analyze labeling and levels of primary metabolites.We will quantify the differences in metabolite composition in the roots and carbon exudate within maize plants treated with SynComs from Aim 2. Changes in metabolite levels between source and sink indicate differential 'shipping' of resources to the ground for exudation and may reflect the persistence of microbes that utilize the exudate and enable enhanced plant productivity. Specifically, we will characterize root contents (unlabeled and 13C isotopologues) from 13CO2-based isotopic labeling of leaves. Leaf vein and root tips that represent the opposite ends of the phloem concentration gradient will become labeled through time course labeling experiments.Maize seeds will be soaked in SynComs for 24 hours then transferred to soil and grown to age 17 days effectively exhausting seedling reserves and establishing the photosynthetic competency of leaves. Plants will be grown under nitrogen replete and minimal nitrogen conditions. At 17 days after planting, leaves will be isotopically labeled using an in house, custom chamber for 4, 8, 12 and 24 hours consistent with preliminary results that suggest significant labeling within this time frame and can quantitatively describe differences in flux due to environmental or microbial perturbations. Experiments will be performed in four labeling chamber replicates including 4-6 individual plant replicates within each chamber for each labeling duration. Labeled exudate, root, and leaf metabolites will be quantified as described above. Samples will also be collected from roots for microbial profiling to assess colonization and persistence of SynCom strains. As a potential example of the types of changes we anticipate, the ratio of alpha ketoglutarate (an organic acid) to glutamate (an amino acid) is likely to be elevated in the leaf under nitrogen deficient conditions because the conversion between the two involves an aminotransferase and a nitrogen donor. Limited nitrate transport to the leaf results in a reduced amount of ammonia to make glutamine or glutamate or to produce other amino acids. Thus, the availability of nitrogen directly impacts aminotransferase reactions in the leaf and the export of amino acids to sinks including the roots and as exudates. Changes in nitrate and these metabolites would reflect a microbial benefit ascribed to nitrogen nutrition.

Progress 04/01/23 to 03/31/24

Outputs
Target Audience:The long-term goal of this research is to provide sustainable and effective agricultural products to farmers that mitigate the need for synthetic fertilizers that are continually becoming more costly and impact profitability. In addition, current farming practices result in excess nitrogen application that contributes to field runoff, polluting wells, rivers and streams, creating 'dead zones' in the Mississippi Delta, and because the chemical approach to synthetic nitrogen production (I.e., the Haber-Bosch synthesis) is an energetically intensive process, the production of fertilizer reduces the net energy obtained from crops intended to be used as biofuels. The immediate audience for this research is the scientific community working to identify beneficial microbes to promote plant health and reduce the requirement for fertilizer inputs on the farm. Further, commercial interest has emerged as a result of the NIFA grant and is in the early stages of spawning new collaborations that can increase the impact of the work, and hopefully deliver solutions for agriculture in a timely fashion. Changes/Problems:A postdoctoral scientist, Dr. Poonam Jyoti, successfully competed for and was awarded a prestigious Assistant Professor Faculty Fellowship throught INSPIRE program in India. While this is an excellent training outcome for our labs, NIFA, and the Danforth Center, it leaves a vacancy related to aspects of this work. Dr. Jyoti will continue to use her developed expertise related to plant and microbe interactions, metabolism, nutrient exchange, isotopic labeling and metabolic flux analysis in her new professorial role which will magnify our efforts to sustainably address global challenges such as synthetic fertilizer use. She left the lab approximately the end of May and a recent advertisement has identified a candidate to complete additional aspects of the grant. The new hire will join the grant and teams, pending visa and moving to the U.S. What opportunities for training and professional development has the project provided?This project has directly trained two postdoctoral researchers and a graduate student as well as two technicians and two data scientists. How have the results been disseminated to communities of interest?Results have been disseminated through publications and DDPSC community seminars. Presentations given to the broader scientific community as seminars and poster presentations are listed below: R Bart: The Future of Plant-Environment Interactions: Challenges and Opportunities in a Changing Climate, Banbury Conference, Lloyd Harbor, New York October 2023 D Allen: Use of stable isotopes to study plant metabolism and root exudation, DOE-EMSL Workshop on Stable Isotopes, Richland WA January 2024 P Jyoti, S Morley, S Koley: Big Ideas 3.0, Donald Danforth Plant Science Center, St. Louis MO August 2023 S Arnesen: Harnessing plant beneficial rhizosphere microbial communities to improve plant health, Fast Talks with University of Nebraska Students, Donald Danforth Plant Science Center, St. Louis MO March 2024 P Jyoti, KL Chu, S Koley, S Kambhampati, A Onyshchenko, I Moroenyane R Bart, DK Allen: Understanding root exudates using stable isotopes. The 17th Annual Science Retreat of the Donald Danforth Plant Science Center Scientific Retreat, St. Louis MO June 2023 J Zhou, S Kambhampati, DK Allen: SIMPEL 2.0: Automating Untargeted Isotope Labeling Analysis and Pathway Elucidation, The 18th Annual Science Retreat of the Donald Danforth Plant Science Center Scientific Retreat, St. Louis MO June 2024 What do you plan to do during the next reporting period to accomplish the goals?During the next period we will: Continue mining rhizosphere and endosphere samples for addition candidate N-fixing bacteria (Aim 1.1). Test version 2 of our pipeline for identification of candidate "helper" bacterial strains (Aim 1). Characterize our candidate N-fixing isolates and begin assembling and testing syncoms based on candidate ASVs identified in version 2 of our pipeline. Apply the developed ARA method and refine isotope labeling with 15N to more rigorously characterize microbes identified as growing in N-free media and that have nifH expression (Aim 2.2). Further refine our ability to isotopically label and measure exudates from plants (Aim 3.2). Pursue additional isotopic labeling experiments for untargeted labeled metabolomics analysis with SIMPEL 2.0, and refine the software tool to assess a second mass spectrometric dimension (i.e., MS2 measurement from tandem mass spectrometry). The increased information content from MS2, coupled with linkage to KEGG metabolic pathways database, will enable more confident identification of unlabeled metabolites (Aims 1&3).

Impacts
What was accomplished under these goals? 1.1: To identify beneficial bacterial communities that robustly colonize maize roots and potentially enhance plant performance we conducted a comprehensive analysis of field data focusing on the nifH gene, a marker for nitrogen-fixing bacteria. Using qPCR with nifH-specific primers, we screened 1502 samples from field experiments conducted over two seasons (2021 and 2022). Our analysis so far has prioritized the 2022 field data, encompassing 960 samples from 64 plots of 8 maize-teosinte introgression genotypes. From the 2022 dataset, we successfully screened 932 samples for nifH gene abundance, revealing significant variability in nifH abundance across different sample zones (endosphere, rhizosphere, soil) and field sections. nifH abundance correlated positively with certain soil properties (sulfur, soluble salts, soil pH) and negatively with zinc levels. We observed a positive correlation between nifH abundance and several plant phenotypes, including root dry weight, mid-season, kernel weight and flowering time, for at least one of the genotypes analyzed. Samples with high nifH abundance will be leveraged for isolating candidate nitrogen-fixing bacterial strains (Aim 2.1). 1.2: Our initial approach involved running co-occurrence and exclusion analyses to identify microbe-microbe associations that could contribute to the stability of synthetic communities (SynComs). Over the past year, we have compared multiple flavors of network analysis, to better understand microbial interactions. To date, we have not identified a "smoking gun" community that we would predict to significantly impact plant health. However, these computational approaches will continue in the next year. To enhance our analysis, we incorporated an additional year's dataset (2022) to serve as a training dataset, thus addressing concerns about false positives and reproducibility. Our focus shifted to exploring quality control metrics, particularly balancing presence and absence data, to refine our approach and manage the high levels of hits encountered. We evaluated existing R packages for network analysis and are modifying and developing our pipeline to better fit our specific needs. 1.3: Initial efforts focused on analyzing the 2021 dataset, considering target metabolites as phenotypes of interest. However, no strong patterns or relationships between microbial communities and plant phenotypes were observed. Multi-omics approaches, including Procrustes analysis and multiple coinertia analysis, were used to compare the principal components of the microbiome and metabolome, but neither method revealed significant relationships. We then expanded our analysis to include the 2022 dataset. Using XCMS for peak detection and retention time alignment, followed by salt cluster filtering and PQN normalization, we processed metabolomics data from root and shoot samples. Applying the Kruskal-Wallis test and Dunn's test, we identified significant differences in metabolite intensities: 1309 differential features in shoots and 104 in roots. Kendall's tau correlation analysis revealed significant correlations between specific metabolites and plant phenotypes, such as kernel weight, seed count, days to flowering, and plant height. Additionally, we identified significant correlations between metabolite levels and soil properties like phosphorus, potassium, zinc, and soil pH, particularly in shoots. Despite these efforts, no definitive patterns emerged linking microbial communities directly to root metabolic profiles. Feature identification is ongoing, and we are continuing to refine our methods. 2.1: We focused on isolating nitrogen-fixing bacteria from field samples identified as enriched in the nifH gene as part of task 1.1. Instead of using brute force or limiting dilution approaches, we opted for isolation on nitrogen-free (N-free) media to enhance the selection of nitrogen-fixing bacteria. A pipeline was developed for the isolation process. Field samples underwent heat treatment and serial dilutions were then plated on N-free media under both anaerobic and aerobic conditions. This method allowed us to isolate 156 bacterial strains from three field rhizosphere samples. Colonies were subsequently grown in rich media, and their identities were confirmed through 16S sequencing of the V4/V5 region and nifH PCR. The isolation approach leveraging N-free media proved effective, with most wells containing a single genus, indicating successful purification of isolates. Notably, we confirmed the presence of the nifH gene in three new bacterial isolates: Paenibacillus taohuashanense, Pseudomonas granadensis, and Paenibacillus helianthin. Nitrogen fixing ability with be confirmed in these strains and they will be further characterized using in vitro studies and seedling assays, as outlined in subsequent tasks. 2.2: Task 2.2 focuses on characterizing the candidate n-fixing bacteria isolated in Task 2.1, including validation and assessment of nitrogen-fixing abilities. We are adapting a previously established acetylene reduction assay (ARA) to measure the nitrogenase activity of these strains by converting acetylene to ethylene, quantified via gas chromatography. We are optimizing this assay with a known free living diazotroph. Preliminary results are encouraging, but further optimization is required to accurately measure nitrogenase activity in the new isolates (task 2.1). 2.3: Currently, we are still refining our method for identifying 'helper' strains, and therefore, no SynComs have been assembled or tested to date. 3.1: We performed D2O labeling on seeds germinated for 36 hours, comparing mock treatments with BLO402 (a bacterial strain previously shown to be capable of promoting plant growth). Using a qExactive Orbitrap mass spectrometer and our expanded SIMPEL 2.0 software, we conducted untargeted metabolomics analysis, identifying 36,152 metabolomic features, of which 2,770 were enriched metabolites. These grouped into approximately 900 active bins contributing to germinative metabolism. We are currently evaluating the metabolic response to BLO402, with plans to extend this analysis to SynComs from previous aims. 3.2: To study root exudates, we developed a new method to overcome the limitations of hydroponic and natural soil systems. Sterile seeds are grown in a sand-gravel mix for 10 days, followed by a water wash with an internal standard. The exudates are collected, centrifuged to pellet debris, desalted, and analyzed via mass spectrometry. This method, tested with multiple substrates and growth mediums, evaluated target exudates like glucose, phenylalanine, sucrose, serine, and leucine/isoleucine at 7, 10, and 12 days of growth. We found that root exudation peaks mid-day by collecting exudates at six time points over a 24-hour period. These findings show plants start exuding detectable amounts at the early seedling stage, with maximum exudation mid-day. For stable isotope-based labeling, we used 13CO2 to analyze root exudates focused on central metabolites. Initial experiments recovered labeled metabolites in exudates after 12 hours, with enrichment increasing over time. This method distinguishes root exudates from other rhizosphere metabolites. Approximately 20 labeled metabolites were identified after 12 hours of labeling, with similar levels at 24 hours, indicating saturation of leaf photosynthetic contribution relative to remaining seed-derived carbon. These advances in root exudate collection and stable isotope labeling, though still being refined, will enable detailed studies on plant-microbe interactions and potentially enable complementary experiments leveraging the DOE National Lab resources.

Publications

• Type: Journal Articles Status: Published Year Published: 2024 Citation: Kambhampati et al., 2024. SIMPEL: using stable isotopes to elucidate dynamics of context specific metabolism DOI: https://doi.org/10.1038/s42003-024-05844-z
• Type: Conference Papers and Presentations Status: Other Year Published: 2023 Citation: P Jyoti, KL Chu, S Koley, S Kambhampati, A Onyshchenko, I Moroenyane R Bart, DK Allen: Understanding root exudates using stable isotopes. The 17th Annual Science Retreat of the Donald Danforth Plant Science Center Scientific Retreat (June 2023; poster)
• Type: Conference Papers and Presentations Status: Other Year Published: 2024 Citation: J Zhou, S Kambhampati, DK Allen: SIMPEL 2.0: Automating Untargeted Isotope Labeling Analysis and Pathway Elucidation, The 18th Annual Science Retreat of the Donald Danforth Plant Science Center Scientific Retreat (June 2024; poster)
• Type: Conference Papers and Presentations Status: Other Year Published: 2023 Citation: PI Bart presented this work at a Banbury conference entitled The Future of Plant-Environment Interactions: Challenges and Opportunities in a Changing Climate in October 2023.
• Type: Conference Papers and Presentations Status: Other Year Published: 2024 Citation: S Arnesen: Harnessing plant beneficial rhizosphere microbial communities to improve plant health. Fast Talks with University of Nebraska Students (March 2024: Talk).