In-depth chracterization of bean (Phaseolus vulgaris L.) and lentil (Lens culinaris Medik.) landraces from Appenine Italian regions

Alessandra Renella ¹, Martina Falcione ¹, Massimiliano Corso ², Stéphanie Boutet ², Gabriella Stefania Scippa ¹, Dalila Trupiano ¹

Department of Biosciences and Territory, University of Molise, 86090, Pesche (Is), Italy ¹, Institut Jean-Pierre Bourgin, Universite Paris-Saclay, INRAE, AgroParisTech, 78000, Versailles, France ²

Autochthonous landraces, also known as local varieties, are plant genetic resources characterized by high genetic variability, specific adaptation to environmental growth conditions, and the presence of specialized metabolites such as terpenoids, flavonoids, and alkaloids, which provide health-promoting effects for humans. Despite their importance, landraces are gravely threatened by extinction, primarily due to their replacement by commercial varieties. The identification of genetic resources' distinctive characteristics is crucial to draw the interest of local farmers and consumers that are involved in their conservation. Untargeted metabolomics is a powerful approach that enables the analysis of a wide range of metabolite classes and provides useful information for the discrimination of local varieties, identifying their metabolic signatures and supporting their valorization and conservation activities. In this context, the present study aimed to provide a global view of the metabolite diversity of three lentil landraces from different villages of Molise region (Italy) – Capracotta, CA; Rionero Sannitico, RS; and Agnone, A – in comparison to one landrace from Umbria region (Italy) – Castelluccio di Norcia, CS –, and three bean landrace from Molise region (Italy) – Ciliegino bean, CI -, Basilicata region (Italy) – San Michele Rosso bean, SMR -, and Spain – Spanish Ciliegino bean, CI SP -. A quadrupole time of flight mass spectrometer-based liquid chromatography (LC-MS/MS) metabolomics technique allowed the detection of 407 and 544 differential accumulated metabolic features (DAMfs) for bean (Be) and lentil (Le) landraces, respectively. The annotation of the DAMfs, by consulting home-made spectral and MS libraries (e.g., the GNPS Public Spectral Library) and a molecular network approach (MetGem and Cytoscape software), allowed 53% and 54% of them to be assigned to different metabolic categories (10 for Be and 13 for Le). Flavonoids were the most represented metabolic category (66 for Be, 87 for Le), followed by amino acids and derivatives (59 for Be, 61 for Le), and cinnamic acids (22 for Be, 50 for Le). However, the ten most important metabolic features for the discrimination of populations identified by the variable importance in the projection (VIP) plot of PLS-DA belong to flavonoids, amino acids, and terpenoids for Le and flavonoids for Be. According to univariate and multivariate statistical analyses, significant variations in metabolite composition were observed among landraces. In the PCA scores plot, A, CA, and RS were separated from CS by the PC1 (56 % of variance). Similarly, CI and CI SP were separated from SMR by the PC1 (58 % of variance). These results have been confirmed by the heatmap, allowing the selection of landraces-specific metabolic features. The enrichment analysis is in progress to assess the presence of enriched metabolomic categories for each landrace, and thus improve the annotation process. To shed more light on the beneficial/health-promoting effect of some bioactive compounds, the antimicrobial and antioxidant activity of both legume extracts were also evaluated and discussed according to the differential metabolic profiles characterizing each landrace.

Main author career stage: Postdoc / Fellow

Contribution type: Poster

First choice session: 3. Biodiversity and global change

Second choice session: 5. Genetics, genomics, and bioinformatics