A MULTI-OMICS PERSPECTIVE ON COMMON BEANS: METABOLOMIC AND LIPIDOMIC ANALYSIS OF PHASEOLUS VULGARIS REVEALS BIOCHEMICAL DIVERSITY AND FUNCTIONAL COMPONENTS

Phaseolus vulgaris L., from the Fabaceae family, is one of the most significant legume crops produced and consumed globally. In addition to its macronutrient composition, the common bean is a significant source of phytochemicals, particularly polyphenols, which contribute to its broad antioxidant activity. Moreover, the lipid composition of P. vulgaris L., while quantitatively limited, is qualitatively rich and of considerable nutritional and physiological relevance. In this work, we used liquid chromatography–mass spectrometry (LC-MS) to assess the metabolic diversity in seeds of P. vulgaris accessions derived from the domestication of Andean and Mesoamerican gene pools and grown in experimental trials in three representative European environments (Italy, Spain, and Poland). We were able to confidently identify 715 metabolites based on comparison with the in-house library covering major nutritional and anti-nutritional compounds. Specifically, the metabolomics atlas generated for common bean seeds within this thesis includes 498 secondary metabolites and 217 lipid molecules. Lipid fractions were analyzed in both negative and positive ionization mode to generate a comprehensive profile of the seeds' fatty acid content. The annotated metabolites displayed broad chemical diversity and were categorized into several classes. Notably, flavonoids and phenolic compounds were the most diverse in terms of annotated features, with 72 and 43 compounds respectively, accounting for a combined total of 115 distinct metabolites. However, these classes contributed only 3.5% of the total relative abundance. In contrast, saponins, despite having slightly fewer annotated features (only 100 compounds), represented the most abundant class by concentration, making up 34% of the total abundance. This was followed by peptides, contributing 30% of the abundance with 24 features, and dipeptides, which accounted for 20% of the abundance across 52 annotated compounds. Among lipids, triacylglycerols were the most prevalent in both number of features (70) and relative abundance (74%), followed by phospholipids (20.99%) and sphingolipids (2.54%). Moreover, the data generated in this study highlights wide metabolic diversification between Andean and Mesoamerican genepools, and how this is shaped by the interaction with the growing environment. In detail, results display that phenolic and flavonoid accumulation patterns are linked to genetic origin, while saponins are more abundant in Mesoamerican genotypes and significantly correlated with seed weight. Amino acid– and peptide–related metabolites are largely enriched in Mesoamerican gene pool and strongly influenced by genotype × environment interactions. Importantly, G×E effects were also evident for phenolic compounds, whose levels varied not only between gene pools but also across cultivation sites, suggesting environment-dependent modulation of antioxidant metabolism. Similarly, saponin abundance, although generally higher in Mesoamerican lines, displayed site-specific trends that point to environmental cues affecting secondary metabolite biosynthesis. Peptide-related metabolites also showed marked plasticity, reflecting how nitrogen metabolism and protein turnover are tightly regulated by both genetic background and growth conditions. Lastly, lipidomic profiles indicates that lipid composition is shaped by both the genetic background and cultivation site, with Andean genotypes showing more unsaturated TAGs and Polish-grown samples accumulating higher levels of phospholipid, likely reflecting the environmental adaptation linked to temperature and photoperiod. The newly developed metabolomic atlas of common bean seeds will provide a valuable resource for the scientific community, supporting future integration into metabolomics-assisted crop breeding and aiding in the selection of metabolites for the development of plant-based functional foods.