Samples

The common bean samples were provided directly by farmers and were obtained from the 2016-2018 harvests at 13 different locations of the state of Durango, Mexico. Nine samples corresponded to Pinto Saltillo variety and were obtained from the towns of Alvaro Obregon (AO-PS), Calixto Contreras (CC-PS), Dr. Castillo del Valle (DCV-PS), Estación Progreso (EP-PS), Ignacio Ramírez (IR-PS), Jesús Agustín Castro (JAC-PS), Luis Moya (LM-PS), Ramón Corona (RC-PS) and Santa Catalina de Sena (SCS-PS). Four samples corresponded to the Black variety, obtained from Calixto Contreras (CC-B), Durango (DD-B) and Ignacio Allende (IA-B1 and IA-B2, harvested in 2016 and 2017, respectively). Once received, samples were stored in hermetic bags and refrigerated at 4 °C until analysis and processing.

Sample preparation

At least 1.5 Kg of bean grains from each locality were manually selected to remove external material, immature seeds, and damaged grains. Subsequently, they were washed with tap water and dried at room temperature. Each sample was divided into two batches; the first batch, without heat treatment (raw), was ground (IKA Werke Mill, MF10 Basic, 1 mm mesh) to obtain a flour with a particle size of 1 mm, and the second batch was subjected to heat treatment (cooked) in a conventional pressure cooker (14.7 psi) for 60 min (based in preliminary work). The cooked beans were frozen and lyophilised (Freeze Dry System, FreeZone 45, Labconco, USA; -45 ° C, 0.133 mBar) and subsequently ground to a flour with a particle size of 1 mm.

Chemical determinations

The following measurements were performed in raw and cooked flour samples. Nitrogen content was determined by the Kjeldahj method (digestion, distillation and titration) (method A.O.A.C 979.09; A.O.A.C., 2005) using a micro Kjeldahl Labconco equipment. Protein content was determined using the conversion factor of 6.25. Moisture content was determined by the thermobalance method (NMX-F-428-1982). The contents of resistant starch (RS) and total starch (TS) were determined enzymatically (Megazyme enzyme kit K-RSTAR-100A 08/1, Megazyme, Ireland). The contents of total dietary fiber, soluble fiber and insoluble fiber were also determined enzymatically (Megazyme enzyme kit K-TDFR-200A 04/17, Megazyme, Ireland). Water absorption capacity (WAC), water absorption index (WAI) and oil absorption capacity (OAC) were determined at room temperature according to ^{Rocha-Guzmán et al. (2008)}.

Statistical analysis

Each analysis was performed with at least two and in some cases up to four replicates. The results were subjected to an analysis of variance (ANOVA) and the comparison of means was performed by Tukey´s test, using the software package Minitab 18 (Version 18.1 for Microsoft Windows 10) with p < 0.05.

Chemical composition

The contents of protein and moisture in raw- and cooked-bean flours are presented in Table 1. Protein content in Pinto Saltillo and Black flours ranged from 13.0 to 16.7 g/100 g. Cooking caused a slight increase (12.9 to 19.0 %) in protein content. Similar results have been reported by ^{Campos-Vega et al. (2009)}, and ^{Wang et al. (2010)}. The high-est protein content was found in the Pinto Saltillo variety, specifically in RC-PS (raw) and JAC-PS (cooked) samples. Heat treatment of legumes (such as cooking) improves protein quality due to inactivation of thermolabile anti-nutritional factors and the heat-induced structural changes that facilitate proteolysis (^{Wainaina et al., 2021}). The protein content increase in cooked-bean flours is attributed to the loss of soluble solids during cooking, increasing the availability of protein (^{Wang et al., 2010}).

The highest moisture content was observed in sample CC-PS (11.7 %), followed by JAC-PS (11.2 %). ^{Senthilkumar et al. (2018)} reported a moisture content of up to 15 % as the recommended optimum for storing raw beans for up to 1 year, which was met in our study. The differences found in moisture content are most likely related to crop, environmental conditions and soil characteristics (García-Díaz et al., 2018).

The contents of resistant starch (RS) and total starch (TS) in raw- and cooked-bean flours are presented in Table 2. The highest values of resistant and total starch were found in the samples of the Pinto Saltillo variety. The samples with the highest contents of resistant starch were RC-PS (raw) and IR-PS (cooked). The RC-PS and CC-PS samples had the highest contents of total starch, in raw and cooked form, respectively. The content of resistant starch was higher in raw samples. This suggests that heat generated during seed grinding was insufficient for starch gelatinization (^{García-Alonso et al., 1998}). The content of resistant starch in raw and cooked flours varied from 10.7 to 37.2 g/100 g and from 6.6 to 9.3 g/100 g, respectively. The total starch content ranged from 15.4 to 51 g/100 g in untreated samples and from 34.9 to 51.3 g/100 g in heat-treated samples, values similar to those reported previously (^{Wang et al., 2010}; ^{Fabbri et al., 2016}).

The general increase in most starch fractions and reduction in resistant starch could be attributed to an improved availability and digestibility by effect of heat treatment, which caused deterioration or disruption of cell walls, resulting in gelatinization and dispersion of starch molecules making them more prone to digestive enzymes attack (Eyaru et al., 2009; ^{Singh et al., 2016}). The compositional variability observed for the tested samples might also be associated with differences in growing conditions. Furthermore, interactions between proteins and starch could further contribute to lower digestibility of legumes (^{Rebello et al., 2014}). Cooking reduces the resistant starch content and increases the Rapid Digestible Starch content (^{Wainaina et al., 2021}), however, the extent of this reduction varies with cooking methods. ^{Yadav et al. (2010)} reported a lower resistant starch content in pressure cooked legumes compared to boiled legumes because it caused a more uniform and complete gelatinization.

Figure 1 shows the contents of soluble (SDF), insoluble (IDF) and total dietary fiber (TDF) in raw- and cooked-bean flours. The highest SDF content was found in the Pinto Saltillo variety, specifically in samples SCS-PS (raw) and AO -PS (cooked). The highest IDF content was found in the Black variety, specifically in samples IA-B2 and CC-B (raw and cooked, respectively), and the highest TDF content was found in the Black variety, in samples IA-B2 (raw) and CC-B (cooked).

Figura 1 Contenido de fibra dietaria soluble (FDS), insoluble (FDI) y total (FDT) en harinas de frijol crudo y cocido de las variedades Pinto Saltillo y Negro. 

Data indicate the mean ± standard deviation. Analyzes were performed in duplicate (AO-PS: Alvaro Obregón-Pinto Saltillo; CC-PS: Calixto Contreras-Pinto Saltillo; DCV-PS: Dr. Castillo del Valle-Pinto Saltillo; EP-PS: Estación Progreso-Pinto Saltillo; IR-PS: Ignacio Ramírez-Pinto Saltillo; JAC-PS: Jesús Agustín Castro-Pinto Saltillo; LM-PS: Luis Moya-Pinto Saltillo; RC-PS: Ramón Corona-Pinto Saltillo; SCS-PS: Santa Catalina de Sena-Pinto Saltillo; CC-B: Calixto Contreras-Black; DD-B: Durango-Black; IA-B1: Ignacio Allende-Black 2016; IA-B2: Ignacio Allende-Black 2017).

Figure 1 Content of Soluble (SDF), Insoluble (IDF) and Total Dietary Fiber (TDF) in raw and cooked bean flours of the Pinto Saltillo and Black beans varieties. 

Cooked samples had an SDF content of 4.7 to 8.9 %, which was higher than that of raw samples. Heat treatment induces structural disintegration and a high degree of fragmentation of the food matrix, which allows more water molecules to interact with other compounds through hydrogen bonds, improving compound solubility, water retention and swelling capacity (^{Chen et al., 2014}).

The IDF contents in raw- and cooked-samples were 35.7-51.0 % and 24.7-51.8 %, respectively, showing a tendency to decrease with heat processing except for the JAC-PS sample. IDF content represents around 93-97 % of the TDF content in legumes, and therefore, SDF represents a smaller proportion (3-7 %) of the dietary fiber content in beans. Therefore, if an increasing or decreasing trend is observed for the insoluble fraction, in general this trend will be observed in the TDF content in legumes (^{Martín-Cabrejas et al., 2008}). An appropriate heat pretreatment can expand the fiber compact structure by transforming hemicellulose, resulting in an increase in porosity and a reduction of the mechanical strength of the polymer matrix, facilitating the decrease in the insoluble fiber content. In addition, heat treatment can cause changes in the physical characteristics and chemical composition of dietary fiber by inducing partial degradation of the insoluble fraction components (cellulose and hemicellulose) to simple carbohydrates (^{Ullah et al., 2018}). An increase in insoluble fiber content, as an effect of heat treatment, can be associated with protein-fiber complexes formed after a chemical modification induced by seed cooking (^{Bressani, 1993}).

The high values of dietary fiber can be attributed to the different crop and growth conditions of the different samples. ^{Kutoš et al. (2003)} reported a higher dietary fiber content in cooked bean samples without prior soaking, as was done in this research, suggesting a higher nutritional value than for soaked and cooked beans. They also observed that changes in dietary fiber content due to thermal processing of beans are highly complex, and depend on the type of bean, type of processing and treatment duration. ^{Campos-Vega et al. (2009)} analyzed the soluble and insoluble fiber content of different varieties of Mexican beans (raw and cooked) and reported a higher content of both fibers in the cooked samples, with values of up to 14 % for SDF and 41 % for IDF.

^{Kutoš et al. (2003)} and ^{Londero et al. (2005)} suggested that fiber contents may vary with the analytical method used, impeding proper comparison for values obtained with different methods. ^{Picolli and Ciocca (1999)} reported that the most appropriate method is that which considers dietary fiber and quantifies the different fiber fractions (soluble and insoluble) using enzymes, generating results that are similar to those generated under physiological conditions in the gastrointestinal tract.

Functional properties

The physicochemical properties of foods with high starch and protein contents are important, particularly water absorption capacity (WAC), which is indicative of the ability of the macromolecule to interact with water, and the water absorption index (WAI), which depends on the availability of hydrophilic groups to bind with water molecules. The oil absorption capacity (OAC) is associated with the physical entrapment of the oil and with the accessibility to non-polar sites in the protein polypeptide chain; it is an indirect measurement of protein denaturation (^{Singh et al., 2005}). In general, the WAC, WAI and OAC values of cooked bean flours were higher than in raw samples. The WAC values (Table 3) for raw bean flours ranged from 1.0 to 2.4 mL/g, with significant differences between samples. The sample CC-PS showed the highest value. Heat treatment resulted in WAC values between 2.9 and 3.6 mL/g. ^{Peyrano et al. (2016)} have reported that heat treatment causes protein denaturation, increasing accessibility to the proteins polar amino acid groups, improving its affinity for water and increasing water absorption capacity. Thermal processing promotes gelatinization of starch (including amylose and amylopectin chains dissociation) and a high content of raw fiber that undergoes swelling, this structure modification exposes new binding sites that can interact with water molecules contributing to the increase in water absorption capacity (^{BeMiller and Huber, 2007}).

WAI ranged from 1.5 to 2.1 g/g for raw samples and from 2.8 to 3.4 g/g for heat-treated samples. Water absorption is one of the most important parameters and WAI values of 2.6 to 3.7 g/g indicates that the bean is well cooked (^{Granito et al., 2004}). Oil absorption capacity ranged from 0.8 to 1.8 g/g for raw beans and from 1.3 to 2.0 g/g for heat-treated samples. These results are compatible with those reported by ^{Rocha-Guzmán et al. (2008)}. The OAC is a product of the physical entrapment of fats by proteins, through the formation of micelles. Heat treatment can affect the composition and profile of polar and non-polar amino acids; the content of polar amino acids decreases after thermal processing while the content of non-polar amino acids increases. A higher proportion of non-polar groups on the surface of the protein could be responsible for an enhanced OAC (^{Lin and Fernández-Fraguas, 2020}).

The high WAC values can be attributed to the hydrophilic nature of proteins, and high OAC values are related to a greater availability of non-polar side chains in the molecules. A WAC value greater than 90 % can be considered adequate to obtain cooking times of less than 110 min (^{Pérez-Herrera et al., 2002}).