Metabolomic profiling of forages to identify anti-methanogenic compounds targeting enteric methane for sustainable livestock systems

Objective. Ruminant methane (CH₄) emissions are a major contributor to global greenhouse gas levels, highlighting the urgent need for sustainable mitigation strategies. This study, conducted as part of the Anti-Methanogenic Feedstock for Livestock Systems in the Global South project led by CIAT, aims to identify anti-methanogenic compounds (AMCs) in forages using an integrative metabolomics approach combined with in vitro CH₄ production assays. Identifying AMCs in livestock feed can support CH₄ mitigation—through selection or genetic optimization—while maintaining animal productivity. Methods. We employ advanced normalized untargeted metabolomics, specifically LC/MS-based techniques (liquid chromatography coupled with mass spectrometry), in conjunction with in vitro emissions data, to identify molecular targets with methane-reducing potential. To pinpoint relevant AMCs, we apply statistical analyses, unsupervised classification, and differential analysis of metabolomic profiles from contrasting high- and low-CH₄-producing samples. This enables the identification of under- or over-expressed metabolites (by mass) associated with methanogenic activity. A critical step in this process is the structural and functional annotation of individual compounds, inferred from LC/MS mass spectra. However, only a small fraction (~5–6%) of the compounds in these forage samples have been structurally resolved and annotated in public databases. To address this bottleneck, we have developed a machine-learning-based pipeline to predict, validate, and identify previously uncharacterized metabolites. This involves: predicting potential molecular structures that match the experimental mass spectra using a trained neural network; generating in silico single-electron impact MS for each predicted structure; and comparing the predicted spectra with the experimental data using dot-product similarity scoring to confirm the best structural matches. We are currently testing and refining this tool to improve the fraction of fully resolved compounds in our forage metabolite database. Progress in the development and application of this pipeline will be reported during this session. Results. Preliminary results from the analysis of three forage species—Leucaena diversifolia, Cratylia argentea, and Urochloa humidicola—revealed different flavonoids and saponins previously reported in the literature as AMCs. Notably, Leucaena diversifolia exhibited high flavonoid levels, which correlated with reduced in vitro CH₄ production, positioning it as a promising candidate for further study. Additionally, across these three species, we found that lyophilization significantly enhances the preservation of volatile and thermolabile metabolites compared to heat-drying, as evidenced by higher peak intensities in the mass spectra (p < 0.05). To broaden the search space across plant families, species, phenotypes, and potential AMCs, we have obtained LC/MS data from 26 samples representing 13 forage species, collected from three CGIAR centers: ICARDA, ILRI, and CIAT. Within the species Trifolium, Leucaena, Erythrina, Panicum, Stylosanthes, Lathyrus, Vicia, Desmodium, Chloris, Cenchrus, Flemingia, Cajanus, and Hedysarum, we selected sample pairs with contrasting CH₄ production profiles. In this talk, we will present a list of metabolites identified in these forages, associated with either positive or negative methanogenic activity. Implications. We expect this comprehensive approach will guide the development of forage-based feeding strategies and gene-edited forage varieties, ultimately contributing to sustainable livestock systems and addressing the dual challenges of climate change and food security in the Global South and beyond.

Clavijo-Buriticá, D.C.; Vélez, G.; Chaura, J.; Riccio, C.; Lara, G.; Villegas, J.; Marin Gomez, A.; Makkar, H.; Arango, J.; Jaramillo-Botero, A.