Portuguese Common Bean Natural Variation Helps to Clarify the Genetic Architecture of the Legume's Nutritional Composition and Protein Quality

Abstract

Common bean is a nutritious food legume widely appreciated by consumers worldwide. It is a staple food in Latin America, and a component of the Mediterranean diet, being an affordable source of protein with high potential as a gourmet food. Breeding for nutritional quality, including both macro and micronutrients, and meeting organoleptic consumers' preferences is a difficult task which is facilitated by uncovering the genetic basis of related traits. This study explored the diversity of 106 Portuguese common bean accessions, under two contrasting environments, to gain insight into the genetic basis of nutritional composition (ash, carbohydrates, fat, fiber, moisture, protein, and resistant starch contents) and protein quality (amino acid contents and trypsin inhibitor activity) traits through a genome-wide association study. Single-nucleotide polymorphism-trait associations were tested using linear mixed models accounting for the accessions' genetic relatedness. Mapping resolution to the gene level was achieved in 56% of the cases, with 102 candidate genes proposed for 136 genomic regions associated with trait variation. Only one marker-trait association was stable across environments, highlighting the associations' environment-specific nature and the importance of genotype × environment interaction for crops' local adaptation and quality. This study provides novel information to better understand the molecular mechanisms regulating the nutritional quality in common bean and promising molecular tools to aid future breeding efforts to answer consumers' concerns.

Keywords: GWAS; Phaseolus vulgaris; amino acids; ash; carbohydrates; fat; fiber; nutritional quality; resistant starch; trypsin inhibitor.

Figure 1

Variance components for the nutritional composition related traits measured in a collection of 106 Portuguese common bean accessions grown in two contrasting environments (Cabrela with a mild climate, and Córdoba, a heat stress prone region). The “-t” after the trait’s name indicates that data was transformed following a Box-Cox transformation. Genotype (G), environment (E), genotype by environment interaction (G×E), block, and residual (error), carbohydrates (CH), resistant starch (RS).

Figure 2

Principal component analysis based on BLUEs of nutritional composition and protein quality-related traits measured in a collection of 106 Portuguese common bean accessions. (A) Biplot representing accessions grown in Cabrela (purple) and accessions grown in Córdoba (orange). Trait loading vectors of the seven nutritional composition-related traits are represented by arrows. Relevant accessions are identified by their accession numbers followed by 1 or 2 according to the corresponding environment: 1-Cabrela; 2-Córdoba. (B) Biplot representing accessions grown in Córdoba (orange). Trait loading vectors of the 24 nutritional composition and protein quality-related traits are represented by arrows. CH-carbohydrates; RS-resistant starch; Ala-alanine; Arg-arginine; Asp-aspartic acid; Glu-glutamic acid; Gly-glycine; His-histidine; Ile-isoleucine; Leu-leucine; Lys-lysine; Met-methionine; Phe-phenylalanine; Pro-proline; Ser-serine; Thr-threonine; Tyr-tyrosine; Val-valine; TIA-trypsin inhibitor activity. The “-t” after the trait’s name indicates that data was transformed following a Box-Cox transformation.

Figure 3

Manhattan plot depicting the genome-wide association results for protein content in common bean using 78 Portuguese accessions grown in the Cabrela environment (left) and 94 Portuguese accessions grown in the Córdoba environment (right). The y-axis represents the −log₁₀ (p-value) of 9601 SNPs and the x-axis shows their chromosomal positions across the common bean genome. The horizontal red line indicates the significance threshold (p-value = 10⁻³).

Figure 4

Network analysis of the candidate genes proposed for the nutritional quality traits using 106 Portuguese common bean accessions grown in two contrasting environments (Cabrela and Córdoba), using Cytoscape software. Traits represented as rectangles; genes represented as diamonds. Genes identified by green diamonds correspond to candidate genes for the SNP markers associated with the highest P-value for each trait. Traits are identified as: 2014-Cabrela; 2015-Córdoba; 2014–2015 — both environments; “-t” — trait data transformed following a Box-Cox transformation. CH-carbohydrates; RS-resistant starch; Ala-alanine; Arg-arginine; Asp-aspartic acid; Lys-lysine; Phe-phenylalanine; Tyr-tyrosine; Val−valine; TIA-trypsin inhibitor activity.

Figure 5

MapMan functional categories of the candidate genes associated with the nutritional composition and protein quality-related traits in 106 Portuguese common bean accessions grown in two contrasting environments. Candidate genes for amino acid contents were pooled together. The nine bar charts represent the number of candidate genes of a given functional category associated with ash, fat, fiber, carbohydrate, moisture, protein, resistant starch (RS), trypsin inhibitor activity (TIA), and amino acid contents.