Functional properties and fatty acids profile of different beans varieties

Abstract Dried seeds of four varieties of Phaseolus vulgaris, three of Vigna unguiculata ssp. unguiculata and two of Vigna angularis grown and marketed in Italy, Mexico, India, Japan, Ghana and Ivory Coast were analysed for fatty acids content. In oils from seeds of P. vulgaris, the main fatty acids were linolenic (34.7–41.5%) and linoleic (30.7–40.3%), followed by palmitic (10.7–16.8%). The first three aforementioned fatty acids in the lipid fraction of V. unguiculata varieties were 28.4, 28.7 and 26.2%, respectively; while in V. angularis varieties, main fatty acids were linoleic (36.4–39.1%) and palmitic (26.9–33.3%), followed by linolenic (17.9–22.2%). Statistical analyses indicate that botanical species play a rule in bean fatty acids distribution, while the same was not verified for geographical origin. Furthermore, the atherogenic index (AI) and the thrombogenic index (TI) were investigated for health and nutritional information. The results showed that these wide spread legumes have functional features to human health.


Introduction
Legumes are increasingly used by developing and underdeveloped countries of the world to circumvent the precarious situation of protein shortage in diets. economically, legumes represent the second most important family of crop plants after Gramineae, accounting for approximately 27% of the world's crop production (Graham & Vance 2003). The common bean (Phaseolus vulgaris L.), one of about 100 species of genus Phaseolus of the Fabaceae (Leguminosae) family, is one of the most important legumes both from economic and from nutritional point of view (Reyes-Martínez et al. 2014). On an equal footing of the common bean, there are the black-eyed beans (Vigna unguiculata ssp. unguiculata L. Walp) and the adzuki or small red beans (Vigna angularis) both species of the Vigna genus of the Fabaceae family.
Inadequate information on beans fatty acid composition are available from literature (Yoshida et al. 2009;Sutivisedsak et al. 2011;Antova et al. 2014). The scientific literature refers to the health benefits of dietary linoleic (C18:2n6) and alpha-linolenic (C18:3n3) acids, due to their therapeutic and preventive effects on coronary heart and other diseases (Nordoy et al. 2001). The correlation between fatty acid content and botanical or geographic origin was verified for different food (Arena et al. 2007;Di Bella et al. 2007), but not for beans.
The main aim of the research was to establish the fatty acid composition of 33 edible beans in an attempt to correlate such composition with the botanical and geographical origin of the samples. The secondary aim was to assess the health and nutritional values of lipid fraction of investigated beans according to indexes of lipid quality.

Results and discussion
The analyses were carried out on 33 samples of dried edible beans from nine varieties (three or six for each one). The list of investigated sample typologies is given in Table S1 (online only), which also shows the country of origin. The oil yield and the fatty acid composition of analysed samples are given in Table S2 (online only). In agreement with the literature data (Yoshida et al. 2009;Sutivisedsak et al. 2011), the oil content varied from 1.09 to 1.89%. In oils from seeds of P. vulgaris varieties, linolenic and linoleic acids dominated, followed by palmitic and oleic acids, except for RF samples for which the amount of linoleic acid was higher than the linolenic one. In agreement with results of this research, references on lipid fraction of varieties of P. vulgaris shown predominance of linoleic and linolenic acids reaching up the 70%, and most often, the linolenic acid was the dominant fatty acid (Yoshida et al. 2005). All lipid fractions from V. unguiculata ssp. unguiculata had about 29% of linoleic acids, 28% of linolenic acid, and 26% of palmitic acid as confirmed by Soris and Mohan (2011). According to Yoshida et al. (2009), the principal fatty acid of V. angularis varieties were linoleic and palmitic acids, followed by linolenic and oleic acids.
In order to find a possible discrimination among samples of different botanical origin, a factor analysis by principal components extraction was carried out. Three principal components with eigenvalues exceeding one (3.944, 2.183 and 1.503) were extracted; those explained the 35.85, 19.85 and 13.66% of total variance, respectively.
The two-dimensional Scatterplot for the 33 bean samples showed three groups clearly distinguished according to the botanical origin ( Figure 1(A)). On the whole, the samples of P. vulgaris showed positive PC1 and negative PC2; thus, they were characterised by higher C18:3n3, C14:0 and C22:2n6 and lower C16:0 content than the others. Samples of V. unguiculata ssp. unguiculata showed positive PC1 and PC2; so, these samples showed the highest C18:0, C16:1n9 and C22:0 concentrations. Samples of V. angularis had negative PC1 and PC2; as a result, these samples showed the highest C16:0, C18:2n6 and lowest C18:3n3 contents. It can also be pointed out that the only two V. angularis samples (RA1 and RA3), with positive PC2 scores and small projections on PC1 (very nearby to samples of P. vulgaris) were those that had the greatest C18:0 concentrations, although results were lower than the minimum value determined within the P. vulgaris samples. To try to accentuate the separation among different samples, the variables significantly different among the groups were researched by Kruskal-Wallis test. Significant differences (α = 0.05) in FA composition were found for C16:0, C18:0, C18:1n9, C18:2n6, C18:3n3 and C22:0 (Table 1). Principal components analysis (PCA) was carried out again considering only these variables. Two principal components (with eigen values 3.128 and 1.913, respectively) were extracted; these explained, overall, 84.017% of the variance, in particular 52.128% for the first component and 31.889% for the second one. The first component showed the highest positive correlation with C18:1n9 (0.918), C18:3n3 (0.917) and, to a lesser extent, with C22:0 (0.674), while negative correlations can be observed for C16:0 (−0.730) and C18:2n6 (−0.674). The dominant variable in the second component was C18:0 (0.925). Figure 1(B) showed the Scatterplot of the bean samples in the plain definite by these components and, as can be observed, the three clusters were clearly distinguishable according to the botanical origin. Using the PCA, an attempt of differentiation among the beans of different geographical origin was performed. In the first approach, all independent variables were entered together and then, in another attempt, only the variable significantly different among groups were used. Unfortunately, as showed in Supplementary Figure S1 (online only), the samples were not distinguishable on the basis of their geographical origin.
In this research, the atherogenic index (AI) and the thrombogenic index (TI) were investigated in order to give information on health and nutritional values of beans lipid fraction. Regarding to the samples of P. vulgaris was calculated for AI and TI, the lowest values equal to 0.18 and 0.10, respectively. The value of AI and TI of the V. unguiculata ssp. unguiculata were 0.50 and 0.32, respectively. Similar results were obtained for V. angularis: 0.54 (AI) and 0.41 (TI). Low values of AI and TI were more beneficial to health according to Benedetti Tonial et al. (2013); therefore, the lipid fraction of beans belonging to P. vulgaris specie seemed of better quality than other considered in this research. The PUFAn6 and PUFAn3 daily intakes for adults are 6.4 g/day and 1.6 g/day, respectively, as established by european Scientific Committee on Food (SCF 1992). In 2009, the european Food Safety Authority (eFSA) published its recommendations for PUFA: a linoleic acid intake of 10 g/day and a linolenic acid intake of 2 g/day (eFSA 2009). Taking into account, the total fat content and the quantity of the above-mentioned fatty acid in each of the three species, it has been observed that the consumption of 7 g/day of beans (FAOSTAT 2014) of the species P. vulgaris meets on average a percentage equal to 0.64% of the recommended amount of PUFAn6, 2.79% of PUFAn3, 0.40% of linoleic acid and 2.23% of linolenic acid. The consumption of same amount of beans of the species V. unguiculata ssp. unguiculata and V. angularis supplies, respectively, the following percentages: 0.48% (PUFAn6), 1.65% (PUFAn3), 0.31% (linoleic acid) and 1.32% (linolenic acid); 0.53% (PUFAn6), 1.14% (PUFAn3), 0.34% (linoleic acid) and 0.91% (linolenic acid). Furthermore, in developing countries, the bean consumption can be up to 30 g/day (FAOSTAT 2014), and therefore, the intake of healthy fatty acid from beans in those countries is higher than the others. For example, for beans of the species P. vulgaris, the percentages of the recommended amount for PUFAn3 and for linoleic acid are equal to 11.34-1.73%, respectively. For beans of the species V. unguiculata ssp. unguiculata the percentage of the recommended amount for linolenic acid is 5.67%, whereas V. unguiculata ssp. unguiculata beans the percentage of the recommended amount for PUFAn6 is 2.28%. In consideration of the foregoing, the lipid fraction of beans evaluated in this research has beneficial effects for the consumer's health, especially that of beans of P. vulgaris species.

Conclusion
The analysis carried out on different varieties of beans belong to P. vulgaris, V. unguiculata ssp. unguiculata and V. angularis species highlighted that all analysed samples showed the presence of 11 different fatty acids, among which the C16:0, C18:1n9, C18:2n6 and C18:3n3 were the most abundant. Research results besides demonstrate that there is a relationship between fatty acids distribution in beans and their botanical species, while no links were verified with their geographical origin. Furthermore, beans appeared a good dietary source of beneficial fatty acids despite their lipid content is low. Accordingly, this very important crop plants acquire an extra nutritional value for people that have a low and middle income in most developing and underdeveloped countries of the world.

Supplementary material
experimental details relating to this paper are available online, alongside Figure S1 and Tables S1 and S2.

Disclosure statement
No potential conflict of interest was reported by the authors.