The solution was placed in an open column with silica gel. The following reagents were added and the fractions were collected after each passage: (1) 100% petroleum ether, (2) petroleum ether:ethyl Doxorubicin ic50 acetate (1:1, v/v), (3) 100% ethyl acetate. The solvents used for fractions 4–19 were ethyl acetate with a gradient of increasing concentration (2%, 5% 8% 12% 15% 20% 25% 30% 35% 40% 45%, 50%, 60%, 70%, 80%, 90%) and methanol/water (1:1, v/v). The fraction (20) contained methanol and water (1:1, v/v), (21) 100% methanol, and (22) only water. The extracts were

grouped in the following order: fraction A (1–3), fraction B (4–8), fraction C (9–12), fraction D (13–17), fraction E (18–20) and fraction F (21, 22) according to results presented by Aparicio-Fernandez et al., 2005 and Aparicio-Fernandez et al., 2005. In order to evaporate, the mixtures were placed in a balloon on a rotary evaporator. The digestibility of the protein was determined by the method of Akeson and Stahmann (1964), which is assessed in vitro by determining the rate of enzymatic hydrolysis through the associations of pepsin and pancreatin in order to simulate the conditions existing in the gastrointestinal tract. Initially, 0.05 g of phaseolin were weighed and added to 3.3 ml of an acidic solution of pepsin. The samples were maintained for 3 h at 37 °C in a shaking water bath. Then, the

samples were neutralised with 3.3 ml ZD6474 in vivo of 0.1 N NaOH and added to 3.3 ml of pancreatin. The samples were kept for 24 h at 37 °C in a shaking water bath. In the next stage, 2 ml of the mixture were withdrawn and transferred to a centrifuge tube. Added to this mixture were 3.3 ml of picric acid (1%). The material was centrifuged for 30 min at 13950g. The Bradford method for protein determination was then used by pipetting 20 μl of the sample into a quartz cuvette and adding 1 ml of Bradford reagent solution. After

2 min, a reading was obtained on the spectrophotometer at 595 nm. The analysis of digestibility was originally done only with phaseolin, and was later repeated with the addition of polyphenolic extracts, being first added to 2.5 mg of polyphenolic crude extract GNAT2 and in the following analysis, being added to 2.5 mg of the polyphenol fractions of phaseolin. The electrophoresis were performed in polyacrylamide gel at a concentration of 10%. Added to the gel were 20 mg of phaseolin. For the preparation of the polyphenol–phaseolin mixture, 2 mg of polyphenols (dissolved in 10% ethanol) and 20 mg of phaseolin were added. The gels were stained in a solution of Coomassie Brilliant Blue R250 for 2 h and then bleached in a solution of methanol and acetic acid. The trials were randomised. For the results, we used the SAS software (1996) for analysis of variance by F test and comparison of means by the Tukey test (p ⩽ 0.05). Protein digestibility is a nutritional parameter that evaluates the use of a protein source.