Protein hydrolysates and phenolic compounds from fermented black beans inhibit markers related to obesity and type‐2 diabetes

Obesity and type 2 diabetes mellitus (T2DM) represent an epidemic and public health problem. Bioactive compounds from foods could be used as adjuvants in the prevention of these diseases. Processes such as solid‐state fermentation (SSF) have demonstrated to improve the release of bioactive compounds from food matrix. This project aimed to evaluate the effect of fermented black bean (Phaseolus vulgaris L.) phenolic compounds and protein hydrolysates on markers associated with obesity and T2DM. Twenty‐eight peptide sequences and 12 phenolic compounds were identified on black bean samples. From in silico assays, sequenced peptides showed theoretical binding energies up to −9.8 kcal/mol and phenolic compounds up to −10.1 kcal/mol for monoacylglycerol lipase. Protein hydrolysates from SSF 96 h (12 peptides) showed the potential to inhibit α‐amylase with inhibitory concentration 50 (IC50) 0.57 mg protein/ml and 5.55 mg protein/ml for α‐glucosidase. Phenolic compounds from SSF 48 h blocked α‐glucosidase IC50 0.353 mg GAE/ml. The antioxidant capacity was maintained in phenolic compounds after fermentation for 48 h and improved in the protein hydrolysates. Raw bean protein hydrolysates (0.1 mg protein/ml) presented the potential to inhibit lipid accumulation in 3T3‐L1 cell line (27.9%). SSF is a processing method for generating functional ingredients rich in bioactive compounds capable of acting on biological markers related to the treatment or prevention of obesity and T2DM.

Bioactive peptides are sequences of amino acids that are found within the primary structure of proteins, generated from the hydrolysis of protein, usually by proteolytic enzymes, which have shown biological properties (Li-Chan, 2015). Protein hydrolysates have demonstrated effects to inhibit markers related to noncommunicable diseases such as T2DM and obesity (Kim, Kim, Choi, Lee, & Nam, 2015;Mojica et al., 2016).
Processes such as solid-state fermentation (SSF) have the potential to increase the release of bioactive compounds from the food matrix. During SSF, enzymes hydrolyze trypsin inhibitors, phytic acids among other antinutritional components from the seeds, increasing their protein value. Furthermore, Bacillus subtilis have shown an enhanced content of antioxidant potential of beans  and generation of peptides with antioxidant potential (He et al., 2012).
The objective of this work was to evaluate the effect of SSF with B. subtilis on black beans proteins and phenolic compounds and their potential for inhibiting markers associated with obesity and T2DM.

| Processing and characterization
Black bean samples were analyzed as raw, cooked, and fermented.
Phenolic compounds and protein were extracted. Protein isolates were hydrolyzed using a simulated gastrointestinal digestion assay (pepsin and pancreatin). Phenolic compounds extraction was performed using acidified ethanol. Phenolic extracts and protein hydrolysates were fully characterized. The biological potential was evaluated using in silico, biochemical and in vitro assays ( Figure 1).

| Raw and cooked beans
Black beans (Phaseolus vulgaris L.) were acquired from the local market and cleaned before use. Black beans were used as cooked and uncooked controls. For cooked control, black beans were suspended in purified water (1:6 w/v) and cooked for 70 min in a pressure cooker at low heat. The sample was dried using a convection oven at 60°C, milled and sieved using 40 mesh, and stored at −20°C until use. For raw control, black beans were milled, sieved using 40 mesh, and stored at −20°C until use.

| Solid-state fermentation
B. subtilis was cultivated in Luria-Bertani broth at 30°C for 24 h. The microorganisms were counted using a Neubauer chamber adjusted to a concentration of 10 7 cells/ml in sterile distilled water. The B. subtilis suspension was used to inoculate chopped and sterilized black beans and start the fermentation. SSF was carried out in stainless steel trays in relation (1:2 w/v) black bean and inoculum solution (concentration of 10 7 cells/ml). Fermentation was continued during 48 and 96 h at 30°C and 90% relative humidity in an automated temperature and humidity-controlled chamber. After the incubation time, samples were collected, freeze-dried, milled, sieved using 40 mesh, and stored at −20°C until use.

| Protein extraction
Protein extraction was carried out following Mojica, Chen, and de Mejía (2015). Fermented bean, raw bean, and cooked bean samples were suspended at a 1:10 bean/water ratio, pH 8.0, adjusted with 0.1 M NaOH, and temperature of the 35°C in constant agitation for protein extraction during 1 h. The mix was centrifuged at 4,000 RPM for 15 min at 25°C. The supernatant was adjusted to pH 4.3 with 0.1 M HCl to precipitate proteins, followed by centrifugation at 4,000 RPM for 15 min at 4°C. The liquid supernatant was removed and consequently discarded, the pellet was freeze-dried, and common bean protein isolates (BPI) were stored at −20°C for further analysis.

| Soluble protein
Soluble protein was determined using the DC protein assay kit (BIO-RAD). Five microliters of hydrolysate or blank was added in a 96-well plate with 25 μl reagent A and 200 μl reagent B, incubated for 15 min, and read at 690 nm. Protein concentration was calculated using a bovine serum albumin standard curve (10-1,200 μg/ml).

| Protein hydrolysis
Protein hydrolysis was performed using simulated gastrointestinal digestion (GI) following the method reported by Mojica, Chen, and de Mejía (2015). Protein was incorporated in distilled water (1:20 w/v) and solubilized by sonication for 15 min. Hydrolysis was carried out with pepsin in ration 1:20 (weight enzyme/dry weight protein), for 2 h, at 37°C, and pH 2.0 was adjusted with 0.1 N HCl. Subsequently, pancreatin was added in ration 1:20 (weight enzyme/dry weight protein), for 2 h, 37°C, and pH 7.5, which was adjusted with 0.1 N NaOH.
Enzymes were inactivated in a water bath at 70°C for 20 min. Hydrolysates were centrifuged at 4,000 RPM for 15 min; the supernatant was freeze-dried and stored at −20°C.

| Extraction of phenolic compounds
Phenolic compounds from fermented, raw, and cooked black beans were extracted using the methodology reported by Mojica, Berhow, et al. (2017). Bean flours were suspended (1:40 w/v) in acidified ethanol solution (23% ethanol and 2% formic acid) and stirred during 2 h at 30 ± 5°C in beakers protected from light. After extraction samples were centrifuged 4,000 RPM for 15 min and 25°C, the supernatant was evaporated at 40°C using a vacuum evaporator and stored at −20°C until analysis.

| Total phenolic compounds
Total phenolic compounds from fermented, raw, and cooked black beans were quantified using the Folin-Ciocalteu method with some modifications (Johnson, Lucius, Meyer, & Mejia, 2011). To a 96-well plate, 50 μl (ethanol extract) of either bean polyphenols extracts, standard curve (40-200 μg gallic acid/ml) or blank were added with 50 μl of 1 N Folin-Ciocalteu's phenol reagent. The mixture was allowed to stand for 5 min and subsequently was added 100 μl of 20% Na 2 CO 3 .
The solution was allowed to stand for 10 min before reading the absorbance at 690 nm. The results were expressed as mg of gallic acid (GA) equivalents per gram of bean. where A is the absorbance and DF is the dilution factor.

| Total flavonoids
Samples (Phenolic compounds from fermented, raw, and cooked black beans) (125 μl) and the quercetin standard curve (50-350 μg/ml) were added to a 96-well plate. Then 7.5 μl 5% NaNO 2 was added and incubated for 5 min at room temperature. Next, 7.5 μl of 10% AlCl3 was added and incubated for 6 min. Finally, 50 μl 1 N NaOH and 60 μl distilled water were added to each well and the absorbance was read at 510 nm. Results were expressed as mg quercetin equivalents (QE) per gram of bean (Mojica, Meyer, Berhow, González, & Mejía, 2015).
Absorbance was read at 500 nm, using a Tecan infinite M200 pro. The concentration of condensed tannins was calculated and expressed as mg catechin equivalents (CAE) per gram of bean.

| Identification of phenolic compounds by ESI-QToF-MS
The 12 reference standards and the methanolic extracts were analyzed by direct injection into the TOF mass spectrometer following the methodology of Chen, Wortmann, Zhang, and Zenobi (2007) with some modifications. The mass spectra were acquired using a Xevo-G2-S QToF mass spectrometer (TOF-MS, Waters, UK) equipped with an ESI interface. The optimized mass spectrometric parameters were as follows: in positive ion mode (ESI + ), capillary voltage 3,000 V, source temperature 150°C, desolvation temperature 450°C, desolvation gas 900 L/h, gas flow in the cone 50 ml/min, and the flow of the sample 5 μl/min; in negative ion mode (ESI − ), capillary voltage 2,500 V, source temperature 130°C, desolvation temperature 500°C, desolvation gas 1,000 L/h, gas flow in the cone 50 ml/min, and the flow of the sample 5 μl/min. The mass scale was calibrated using the calibration solution provided by the manufacturer between m/z 50 and 1,500 Da. The protonated molecule

| Identification and quantification of anthocyanins by HPLC-DAD
In the case of anthocyanins, a sample of 15 μl of the crude acidic methanolic extract was injected in the HPLC-DAD system (1260 Infinity, Agilent Technologies, Santa Clara, CA, USA) equipped with a column analytical reverse-phase C-18 Zorbax SB rapid resolution (4.6 × 50 mm, 1.8 μm particle size). The mobile phase used was composed of 0.5% phosphoric acid in water (A) and 0.1% acetic acid in acetonitrile (B) with a flow rate of 1 ml/min; column temperature, 25°C; analysis time, 15 min; wavelength, 520 nm. In this case, the anthocyanin quantification was performed using a calibration curve for each anthocyanin. The concentration was expressed as μg of each anthocyanin individual per gram of dry tissue.

Peptides sequences were drawn and 3D performed in
MarvinSketch software version 17.29.0 and converted to PDB extension using the Discovery Studio Visualizer software version 17.2.0.16349. Flexible torsions, charges, and grid box size (20-30) were assigned using AutoDock Tools. Binding energies were performed using AutoDock Vina (Trott & Olson, 2010). Moreover, the binding pose with the lowest binding energy was selected as a representative to visualize in the Discovery Studio Visualizer software. Peptide physicochemical properties were predicted using the PepDraw tool (http://www.tulane. edu/?biochem/WW/PepDraw/), and protein sources were obtained from protein blast (https://blast.ncbi.nlm.nih.gov/Blast.cgi).

| α-Amylase and α-glucosidase inhibition
The α-amylase and α-glucosidase assays were carried out following (Johnson et al., 2011). Fifty-microliters of samples, positive control (1 mM acarbose) or negative control (distilled water), were added to 50 μl of 13 U/ml α-amylase solution (type VI-B from porcine pancreas in 0.02 M sodium phosphate buffer, pH 6.9) or 100 μl of a 1-U/ml αglucosidase solution (in 0.1 M sodium phosphate buffer, pH 6.9) and incubated at 37°C for 10 min. For the α-amylase assay, 50 μl of 1% soluble starch solution (previously dissolved in 0.02 M sodium phosphate buffer, pH 6.9 and boiled during 10 min) was added to each tube and then leave incubating for 10 min. Finally, 200 μl of dinitrosalicylic acid color reagent was added, and the tubes were placed in a boiling water bath for 5 min. The mixture was diluted with 1 ml of distilled water, and absorbance was read at 520 nm using a clear microplate. For the α-glucosidase assay, a 50 μl of 5-mM pnitrophenyl-α-D-glucopyranoside solution (in 0.1 M sodium phosphate buffer, pH 6.9) was added and incubated at 37°C for 5 min; subsequently, the absorbance was read at 405 nm. For both assays, the results are presented as a percent of inhibition relative to the negative control.

| Determination of cellular lipid accumulation
Lipid accumulation assays were performed following the method reported by Luna-Vital et al. (2017). Cells were seeded in 96-well plate and induced differentiation as described in Section 2.

| Statistical analysis
Statistical analysis was performed using SPSS v20. The data were obtained by least three independent replications, analyzed using analysis of variance (ANOVA) with Tukey test to identify significant differences among treatments, and the differences were considered significant at p < 0.05. The IC 50 was calculated using Gra-

| Degree of hydrolysis
The degree of hydrolysis (DH) is represented in Figure 3.

| In silico assays
Peptides sequences presented a negative theoretical affinity for αamylase ranging from −0.3 to −8.3 kcal/mol (Table 3). Peptide SSVPW from fermented 96 h presented the lowest predicted binding energy.
In the case of α-glucosidase, the predicted binding energies ranged from +0.1 to −5.5 kcal/mol, the peptide NPTPAGPVAPA from the raw sample presented the lowest binding for this enzyme. Binding energies for PLA2 ranged from −4.7 to −8.3 kcal/mol; the sequence SGGGF from fermented 96 h showed the lowest binding value.
Predicted binding energies for MGL ranged from +3.6 to −9.8. The peptide TKPGGGAGP from fermented 96 h was the most potent ( Figure 5). The PPAR-γ theoretical binding ranged from −0.3 to −7.8; the peptide SVGGGTA from fermented 96 h presented the highest inhibitory potential.
On the other hand, phenolic compounds from beans were good inhibitors of selected molecular markers (Table 4). Interaction with αamylase catalytic site presented negative affinity values, −6.1 kcal/mol for rutin and −8.9 kcal/mol for quercetin. For α-glucosidase, the binding energies were −4.2 kcal/mol for ferulic acid and gallic acid and   (Figure 6a).
On the other hand, α-glucosidase (Figure 6b)  The α-amylase inhibition of phenolic compounds (Figure 6c) ranged from 0.104 mg to 0.178 mg GAE/ml with no significant differences among treatments. The inhibition of α-glucosidase using phenolic compounds (Figure 6d)

| Antioxidant capacity
Antioxidant capacity of protein hydrolysates associated with the radical ABTS (Figure 7a)
treatments, lipid accumulation inhibition effect was observed at the same concentration (16.2% and 6.5%, respectively) ( Figure 8a (2015), where pepsin-pancreatin completely hydrolyzed precooked black bean proteins. The cooking process was not able to breakdown phaseolin; however, it facilitated its complete hydrolysis during the GI process. This could be due to a greater unfolding of the tertiary structure of proteins that allowed higher enzymatic cleavage.  (Joehnke et al., 2018). Also, Al-Ruwaih, Ahmed, Mulla, and Arfat (2019) reported an increment in the DH after hydrolysis of kidney bean protein isolates using alcalase.

| Polyphenols characterization
The content of total polyphenols, flavonoids, anthocyanins, and tannins of raw black bean correspond to data reported for Mojica, Meyer, et al. (2015) (Figure 4), where the largest fraction corresponds to tan- The IC 50 was calculated using a nonlinear regression curve fit F I G U R E 8 Lipid accumulation assay by protein hydrolysates and phenolic compounds in 3T3L1 assays. (a) Lipid accumulation inhibition assay for protein hydrolysates; (b) lipid accumulation assay for phenolic compounds. Data were analyzed using ANOVA with post hoc Tukey test; differences were considered significant at p<0.05 -Jáuregui, & Frutos, 2017). Also, synthesis of β-glucosidase hydrolyzes polysaccharides and disaccharides to monosaccharides (Chamoli, Kumar, Navani, & Verma, 2016

| Lipid accumulation
Protein hydrolysates at a concentration of 0.1 mg/ml presented reduction lipid accumulation. This result is comparable with Oseguera Toledo, Gonzalez de Mejia, Sivaguru, and Amaya-Llano (2016); they reported that pinto Durango alcalase hydrolysates at a concentration of 0.1 mg/ml produced a 28% of inhibition in lipid accumulation.
However, at a concentration of 0.5 mg/ml, lipid accumulation was promoted. Therefore, at concentrations equal or lower than 0.1 mg/ml is probably to observe a positive effect on lipid metabolism, potentially by interaction with key enzymes or transcription factors associated with the differentiation of adipocytes, such as PPAR-γ.
Regarding phenolic compounds, they promoted lipid accumulation at the concentrations used. Luna-Vital et al. (2017) found a higher production of intracellular triglycerides by anthocyanins rich extract from purple corn pericarp in mature adipocytes. However, in the case of lipid accumulation during adipocyte differentiation, anthocyaninsrich extract from purple corn pericarp were able to inhibit lipid accumulation and adipocyte differentiation.
Molecular docking of phenolic compounds showed high affinity for PPAR-γ, which modulate adipocyte differentiation. However, this study was performed in mature adipocytes. Phenolic probably mimics the mechanism of action of TDZs such as pioglitazone increasing insulin sensitivity and then promoting glucose uptake and lipid accumulation (Luna-Vital et al., 2017).

| CONCLUSION
SSF is a promising black bean processing method for the generation of functional ingredients with the potential to modulate markers related to type 2 diabetes and obesity. This process could increase the antioxidant and biological potential of common beans. Solid-state fermented black bean flours rich bioactive components could be used in functional foods formulation. However, further in vivo assays are needed to validate solid-state fermented black bean flours biological potential. Legume functional ingredients could be in the prevention of non-communicable diseases.

ACKNOWLEDGMENT
This study was supported by CONACYT, Scientific Projects to Address National Problems 2016 Grant No. 2081.

CONFLICT OF INTEREST
Authors have no conflict of interest to declare.

ETHICS STATEMENT
This manuscript does not contain any studies with human or animal subjects.

S. A. Flores-Medellín conducted the experiments and helped to design
the study, data analysis, and manuscript writing. R. M. Camacho-Ruiz, C. Guízar-González, E. A. Rivera-Leon, and I. M. Llamas-Covarrubias were collaborators of the project who helped to design the study, supervised sample processing, and revised the manuscript. L. Mojica designed the study, provide guidance, manuscript review, and editing.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.