3.1. Optimization of the protein extraction conditions
At the beginning of the protein extraction from the common bean genotypes, an optimization study was performed to determine the effects of alkali level (pH) and salt concentration on the protein content of the final concentrate using central composite rotatable design. Table 1 shows the levels of processing parameters and the protein levels of the concentrates at these processing conditions. As is seen, the protein content of the concentrate samples ranged between 62-79.03%. The lowest protein content was observed for the common bean sample exposed to extraction at pH 11 and no salt addition while the highest protein content was determined for the sample extracted at pH 11 and with the addition of 0.4% salt. It was observed that that the protein content was affected by the processing variables. Table 2 shows the effect of pH and salt concentration on the protein content of the final protein concentrates. It was seen that the linear effect of salt concentration showed significant effect on the protein content of the concentrate (p < 0.05) and linear effect of pH had no effect (p > 0.05). But the interaction effect of both salt level and pH showed very significant effect on the final product (p < 0.01). Figure 2 illustrates the change in protein content of the samples as surface plot and it could be seen that the increase in salt concentered increased the protein level while the alkali level showed a slightly change for the response parameter. The counter plot also shows the change in the protein level depending on the processing variables and the increase could be seen towards to the highest level of both salt and pH level. Eq. 1 shows the constructed regression model which could be used efficiently to estimate the protein level of the common bean protein concentrate. As is seen, coefficient of determination for the regression model was calculated as quite close to unity which shows the fitting ability of the model.
Table 2
ANOVA table showing the effect of processing variables on the protein content of the concentrates
|
Sum of
Squares
|
|
Mean
|
F
|
P-value
|
Source
|
df
|
Square
|
Value
|
Prob > F
|
Model
|
167.57
|
5
|
33.51
|
52.11
|
0.0003
|
X1-Salt (%)
|
44.35
|
1
|
44.35
|
68.96
|
0.0004
|
X2-pH
|
0.13
|
1
|
0.13
|
0.21
|
0.6693
|
X1 X2
|
122.69
|
1
|
122.7
|
190.78
|
< 0.0001
|
X12
|
0.39
|
1
|
0.39
|
0.61
|
0.4716
|
X22
|
0.01
|
1
|
0.01
|
0.01
|
0.9153
|
Residual
|
3.22
|
5
|
0.64
|
|
|
Lack of Fit
|
2.67
|
3
|
0.89
|
3.29
|
0.2418
|
Pure Error
|
0.54
|
2
|
0.27
|
|
|
Cor Total
|
170.78
|
10
|
|
|
|
CV (%)
|
1.14
|
|
|
|
|
R2
|
0.962
|
|
|
|
|
Adj R2
|
0.981
|
|
|
|
|
Protein Content (%) = 70.24 + 2.72X1-0.15X2 + 5.54X1X2 + 0.39X12-0.06X22 (R2 = 0.982) (Eq. 1)
To reveal the best processing conditions for the protein extraction from the common beans, an optimization study based on desirabilty function was perfomed. Maximization and minimization process was applied and the minimum and maximum protein levels were calculated for the processing variables. Figure 3 illustrated the levels of processing variable concluding the maximum and minimum protein content. It was resulted that the maxium protein content would be at pH 11 and %0.4 salt concentration while the minimum protein level would be at pH 11 and 0% salt level. As could be seen from the figure, desirability function showing the accuracy of optmization results was calculated as 0.980 and 0.990 for the maximum and minmum values, respectively. Based on the optimization results, protein extraction from the different genotypes of common beans were performed at pH 11 and 0.4% salt concentrations. Similar pH value (pH 11) for the alkali extraction of protein from faba bean seed was used by Samaei et al. [19]. Karaca et al. [20] investigated the technological properties of protein extracted from different pulses and used salt extraction method due to the globulins rich in the legume seeds. Roy et al. [21] reported that the globulins are soluble in salty solutions and approximately 70% of the proteins in legume seeds is composed by globulins. It was also reported that the isoelectric precipitation technique provided the precipitation of globulins [22] while the mixture of globulins and albumins could be precipitated by salt extraction approach [23].
2.6. Physicochemical and technological properties of protein concentrates
Table 3 shows the protein content of the common bean genotypes and protein concentrates and yield values. As is seen, the protein level of the common bean flour ranged between 22-26.93%. The highest protein content was determined for G8 and the lowest was for G7 (Table 3, Fig. 1). The similar results were also reported by other researchers who informed that the protein content of the common beans had 20–30% of protein level and the genotypes affected the protein content of the samples [24]. Protein content of the protein concentrates was in the range of 72.97–77.99% while the highest level was observed for G8 and the lowest one was for G7. The highest protein level of the concentrate was determined for the common bean sample having the highest protein content (G8) while the lowest level of protein of the concentrate was for the common bean having the lowest protein content (G7). In another research conducted by Karaca et al. [20], the protein content of the protein concentrates obtained from lentil and soybean by salt extraction method were reported to be 74.71 and 72.64%, respectively. Aragundade et al. [25] studied the effect of ionic strength and pH precipitation effect on extractability of protein and reported that the protein content of the broad bean protein concentrate was 71.6 which is similar to the result of the current study. Yield of protein concentrates was calculated according to the approach described by Makeri et al. [26] and the values for each common bean genotype were tabulated in Table 3. The yield values of the protein concentrates ranged between 32.79–50.70 %. Gundogan and Karaca [27] reported that the yield values for the local bean samples were in the range of 39–46%. Makeri et al. [26] reported the yield values of protein concentrates from winged and soybean were 50.82 and 57.12%, respectively.
Table 3
Protein content of common bean flour and their protein concentrates
Genotypes
|
Common Bean Flour (%)
|
Protein concentrate (%)
|
Yield (%)
|
G1
|
26.21 ± 0.45
|
77.70 ± 0.47
|
32.79
|
G2
|
26.64 ± 0.25
|
75.88 ± 0.75
|
49.34
|
G3
|
23.93 ± 0.36
|
76.20 ± 0.84
|
50.70
|
G4
|
26.78 ± 0.41
|
77.78 ± 0.95
|
39.66
|
G5
|
26.59 ± 0.15
|
77.11 ± 0.35
|
32.83
|
G6
|
22.39 ± 0.45
|
74.45 ± 0.45
|
42.27
|
G7
|
22.00 ± 0.15
|
72.97 ± 0.62
|
43.82
|
G8
|
26.93 ± 0.52
|
77.99 ± 0.61
|
32.98
|
Table 4 shows the color properties and functional features of the protein concentrates of common bean genotypes. Color properties were recorded as L, a, and b values and the differences among the samples were significant for all color parameters. The lowest L value which means the darkest protein concentrate was G7 while the highest L value was for the genotype coded as G3. Redness (a) values ranged between 0.40–8.15 and the yellowness (b) values were in the range of 11.03–15.53. The highest redness was for G8 and the highest yellowness was also recorded for G8. Wani et al. [28] reported that the color properties of the different cultivars were recorded as L, a and b as in the range of 81.16–83.25, 0.14–1.70 and 10.64–12.57, respectively. In another study conducted by Joshi et al. [29], L, a and b values of lentil protein concentrates were reported as in the range of 69.2–92.0, 6.9–16.3 and 15.7–24.8, respectively. The color of the protein concentrates is affected by the sample type and also the variation due to the genotypic differences and the pigment intense on the seed coat [28].
Table 4
Some physicochemical and technological properties of protein concentrates
Genotypes
|
L
|
a
|
b
|
Solubility
|
WHC
|
OHC
|
EC
|
ES
|
G1
|
55.65 ± 0.07
|
4.26 ± 0.08
|
15.41 ± 0.07
|
90 ± 2.7
|
3.45 ± 0.04
|
3.43 ± 0.18
|
68.64 ± 1.93
|
87.08 ± 0.59
|
G2
|
57.95 ± 0.04
|
4.72 ± 0.05
|
11.67 ± 0.04
|
58 ± 3.5
|
3.71 ± 0.30
|
4.95 ± 0.18
|
96.40 ± 1.64
|
95.26 ± 2.14
|
G3
|
68.16 ± 0.04
|
0.40 ± 0.08
|
11.03 ± 0.02
|
32 ± 2.4
|
4.07 ± 0.10
|
3.45 ± 0.01
|
67.05 ± 3.91
|
82.91 ± 1.54
|
G4
|
65.36 ± 0.01
|
2.79 ± 0.02
|
12.50 ± 0.02
|
84 ± 1.8
|
3.61 ± 0.04
|
4.22 ± 0.03
|
63.81 ± 1.56
|
81.14 ± 3.10
|
G5
|
54.26 ± 0.00
|
4.00 ± 0.05
|
14.16 ± 0.02
|
90 ± 4.1
|
3.38 ± 0.06
|
3.64 ± 0.17
|
59.27 ± 3.01
|
82.16 ± 3.05
|
G6
|
62.76 ± 0.01
|
0.23 ± 0.09
|
11.60 ± 0.02
|
84 ± 2.4
|
3.58 ± 0.06
|
3.52 ± 0.11
|
67.95 ± 1.81
|
85.00 ± 1.41
|
G7
|
27.14 ± 0.04
|
4.74 ± 0.10
|
13.63 ± 0.06
|
76 ± 2.6
|
3.55 ± 0.01
|
3.31 ± 0.16
|
67.17 ± 3.49
|
92.06 ± 5.08
|
G8
|
37.21 ± 0.05
|
8.15 ± 0.02
|
15.53 ± 0.05
|
94 ± 4.1
|
3.55 ± 0.07
|
3.31 ± 0.18
|
63.33 ± 0.00
|
78.79 ± 1.71
|
WHC: Water holding capacity, OHC: Oil holding capacity, EC: Emulsion capacity, ES: Emulsion stability |
Solubility of the protein concentrates was recorded at pH 7 and they were in the range of 32–94% and the highest soluble protein concentrates were obtained by common bean genotype coded as G8 while the genotype coded as G3 showed the weakest solubility character (Table 4). Among the samples, only two genotypes showed quite weak solubility compared to others. Karaca et al. [20] reported that the solubility of the protein changes significantly according to the protein isolation method and they informed the solubility of lentil protein and soy protein concentrates were 89.88 and 96.79% at neutral pH, respectively. Also, using salt for the extraction of the protein from the structure of the protein source such as a pulse flour could be caused a decrease in the solubility of the protein at neutral pH [20]. Kimura et al. [30] informed that the globulin type in the different pulses showed significant effect on the technological characteristics of the final protein concentrates and the distribution of hydrophilic and hydrophobic amino acid residues on the molecular surface as well as amino acid compositions are important determinants of solubility pH dependence. For instance, after hydrolysis by an enzymatic process, polar and non-polar amino acid groups, could be exposed on the surface of protein molecules and these polar amino acids may interact with water molecules easily through hydrogen bonds and electrostatic interactions, and these process could provide the increased protein solubility [28].
Water holding and oil holding capacity values of the protein concentrates ranged between 3.38–4.07 and 3.31–4.95 g/g, respectively. G3 showed the highest WHC while the highest OHC value was calculated for the genotype coded as G2 (Table 4). In another research, kidney bean protein concentrates showed much higher WHC and OHC compared to our results and they were reported that the WHC and OHC of the concentrates ranged between 5.34–5.83 and 5.82–6.92 g/g, respectively [28]. Withana-Gamage et al. [31] informed the WHC of the chickpea protein concentrate as in the range of 2.34–4.31 g/g and WHC for the mung bean concentrates were reported to be ranged between 1.28–2.78 g/g [32]. In general, WHC showed a wide variation depending on the source. Makeri et al. [26] stated that the dissociation temperature level and pH conditions could result the subunits to create the more water binding sites than oligomeric proteins. As compared to the results in the literature, WHC of the protein concentrates from the common bean genotypes are higher than those of most of the protein concentrate resources. WHC of the protein concentrates could be have a technological function in some food formulations need to high level of water absorption capacity such as meat-based products or baked dough [28]. Wit [33] informed that the imbibing of the water has a critical role for the quality and the product yield in sausages, comminuted and pumped meats. OHC of the protein is also important quality character for some food types including oil incorporation in the formulation such as meat, doughnuts, sausages, soups, custards, baked products, dairy products where they can provide some desirable properties such as body, thickening, and viscosity [34].
Two important quality characteristics of protein concentrates namely emulsion capacity (EC) and emulsion stability (ES) were determined and the calculated values for all genotypes were tabulated in Table 4. As is seen, EC values of the protein concentrates ranged between 59.27–96.40% while the ES values were in the range of 78.79–95.26%. The best emulsion forming performance was observed for the sample coded as G2 while the sample coded as G5 showed the weakest capacity. Similarly, the best ES performance was also monitored for the sample of G2. Significant variations among the EC of the common bean protein concentrates were observed. Hojilla-Evangelista et al. [35] informed that the emulsion activity showed significant variation for the common bean genotypes and the emulsion activity values were in the range of 51.8–65.8 m2/g at neutral pH for the samples and also they showed that the pH had a critical role for the EC because these values ranged between 378.8-387.9 m2/g at pH 10. Sahni et al. [36] also reported that the EC of the alfa-alfa protein concentrates were 1.91, 39.16 and 60.41% at pH 4, 7 and 9, respectively. Mao and Hua [37] reported that the EC of walnut protein concentrate (WPI) was 50.01% and they also stated that the pH had a significant effect on the EC. In neutral pH, EC of the WPI was approximately 50% and increase in pH increased the EC of the samples. It was concluded that EC was pH dependent and alkaline pH has an improver effect on the EC of the protein concentrates.
2.7. Bioactive properties of protein concentrates
Table 5 shows the bioactive characteristics of the common bean flours and their protein concentrates. As is seen, the common bean genotypes showed differences for the studied parameters for both flour and protein concentrates. Total phenolic content (TPC) of the flour and concentrates were in the range of 578.9-1355.9 mg GAE/kg and 313.5-1219.1 mg GAE/kg, respectively. The highest TPC was determined in G8 among the flours and G7 among the concentrates while the lowest ones were determined in G4 and G3 for flour and concentrates, respectively. Piñuel et al. [38] characterized the protein concentrates from red bean and reported that the TPC of the samples ranged between 1356–5217 mg GAE/kg. Total flavonoid content (TFC) of the samples ranged between 352.5–1185 mg QE/kg and 392–1253 mg QE/kg for the flour and concentrates, respectively. The highest TFC of the flours was in the genotype coded as G8 and then G7 while the highest TFC for the protein concentrates was observed in same genotypes.
Table 5
Bioactive properties of common bean flour and their protein concentrates
Common bean flours
|
Genotypes
|
TPC
(mg GAE/kg)
|
TFC
(mg QE/kg)
|
ARADPPH (%)
|
ARAABTS.+
(µg TE/g)
|
FRAA
(mg AAE/kg)
|
AA (mg AAE/kg)
|
G1
|
1218.1 ± 9.4
|
982.5 ± 118.7
|
24.54 ± 2.37
|
3.205 ± 0.08
|
1.72 ± 0.06
|
5410.6 ± 254.9
|
G2
|
1058.2 ± 65.4
|
762.5 ± 41.9
|
20.61 ± 1.03
|
3.482 ± 0.45
|
1.69 ± 0.03
|
5109.6 ± 161.2
|
G3
|
996.5 ± 31.0
|
680.0 ± 46.9
|
25.81 ± 1.11
|
3.677 ± 0.43
|
2.00 ± 0.03
|
4792.2 ± 155.8
|
G4
|
651.0 ± 39.1
|
352.5 ± 28.7
|
15.18 ± 0.44
|
2.616 ± 0.34
|
1.45 ± 0.02
|
4262.7 ± 187.4
|
G5
|
578.9 ± 19.2
|
395.0 ± 17.3
|
14.85 ± 0.13
|
1.869 ± 0.17
|
1.35 ± 0.02
|
4314.8 ± 375.9
|
G6
|
1223.5 ± 32.5
|
856.0 ± 38.3
|
36.32 ± 2.37
|
4.473 ± 0.46
|
2.28 ± 0.01
|
4726.3 ± 449.5
|
G7
|
1200.5 ± 45.2
|
1080.0 ± 45.5
|
31.01 ± 1.74
|
4.001 ± 0.36
|
2.09 ± 0.06
|
5482.5 ± 230.3
|
G8
|
1355.9 ± 15.5
|
1185.0 ± 48
|
34.31 ± 1.35
|
4.318 ± 0.17
|
2.16 ± 0.04
|
5065.8 ± 203.1
|
Protein concentrates
|
G1
|
828.1 ± 53.2
|
679.0 ± 89.1
|
29.07 ± 1.78
|
1.898 ± 0.317
|
1.26 ± 0.02
|
2793 ± 305.5
|
G2
|
829.4 ± 48.2
|
497.0 ± 69.3
|
23.28 ± 2.01
|
3.009 ± 0.846
|
1.10 ± 0.05
|
2894.1 ± 179.8
|
G3
|
313.5 ± 63.5
|
392.0 ± 19.8
|
24.80 ± 0.20
|
1.690 ± 1.200
|
0.81 ± 0.05
|
2443.0 ± 268.1
|
G4
|
352.7 ± 26.8
|
616 ± 39.6
|
25.97 ± 0.88
|
1.331 ± 0.046
|
0.89 ± 0.04
|
2985.5 ± 153
|
G5
|
338.8 ± 68.2
|
903 ± 29.7
|
27.91 ± 2.36
|
0.823 ± 0.148
|
0.78 ± 0.01
|
2229.1 ± 96.3
|
G6
|
409.4 ± 21.4
|
462 ± 19.8
|
26.93 ± 0.82
|
1.260 ± 0.074
|
0.85 ± 0.01
|
3016.6 ± 287.3
|
G7
|
1219.1 ± 59.1
|
1036 ± 39.6
|
37.32 ± 1.48
|
3.612 ± 0.334
|
2.21 ± 0.02
|
3533.8 ± 248.3
|
G8
|
1196.4 ± 53.5
|
1253 ± 148.5
|
36.49 ± 1.16
|
3.624 ± 0.448
|
2.14 ± 0.05
|
3590.2 ± 171.3
|
TPC: Total phenolic content, TFC: Total flavonoid content, ARADPPH: DPPH radical scavenging activity, ARAABTS.+ ABTS.+ radical scavenging activity, FRAA: Ferric reducing antioxidant activity, AA: Antioxidant activity by phosphomolybdenum |
Antiradical activity of all samples was evaluated by using both DPPH and ABTS.+ radical substances. The results of the antiradical activities of the samples were tabulated in also Table 2. DPPH radical scavenging performance of the flour and concentrate samples were in the range of 14.85–34.31% and 23.28–37.32% (Table 5), respectively. For the flour samples, the weakest scavenging activity was determined for the samples of G4 and G5 while the lowest antiradical performance was for the sample of G2 for the concentrate samples. Similar trends were also observed for the ABTS.+ radical scavenging activities of the samples. As is seen, ABTS.+ radical scavenging performance of the flour and concentrates ranged between 1.869–4.473 µg TE/g sample and 0.823–3.624 µg TE/g sample, respectively. The prominent flour and concentrate for the highest and lowest radical scavenging activity of ABTS.+ were G6 and G8, and G5 and G5, respectively. Antioxidant activity of both flour and concentrate was characterized by two popular methods namely ferric reducing (FRAA) and phosphomolybdenum antioxidant activity (AA) approaches. Samples showed differences in terms of antioxidant activities and the strongest antioxidant performances by FRAA were seen in G6, G7 and G8 for the flours and G7 and G8 for the protein concentrate samples. By phosphomolybdenum approach, G1, G7 and G8 among the flours and G7 and G8 among the concentrates were the prominent genotypes (Table 5).
2.8. Classification and corrrelation analysis of samples
To determine the similarity and differences among the studied common bean genotypes, PCA was used and the obtained factor loadings for the samples was illustrated in Fig. 4. The parameters whose curves direct close to one another on the loading plot mean positively related, while those whose curves appear in opposite directions are negatively related. And also the distance between the locations of any genotype on the score plot explains degree of difference or similarity between them [39]. As is seen, G2 showed a significantly different characteristics in terms of EC, WHC and OHC values of protein concentrates. G7 and G8 were also located in same group compared to other genotypes and differed from the other genotypes in terms of bioactive properties. Besides solubility and WHC values of concentrates directed oppositely on plot and that situation was also proved in correlation matrix as a negative correlation (Table 6). Same situation was observed in the study of Stone et al. [40] for pea protein concentrate and WHC was negatively correlated with solubility (r = − 0.857; p < 0.01). Other positive correlations were observed between ARADPPH and TPC, TFC; ARAABTS and TPC; AA and TPC, FRAA and ARA; FRAA and TPC, TFC, ARA; a values and EC, ES, ARAABTS; L values and OHC; OHC and EC. Siddiq et al. [41] also observed the positive correlation between emulsion capacity and stability for bean flours. No significant correlation between solubility and emulsion stability or capacity was detected.
Table 6
Correlation matrix for the studied parameters of protein concentrates
|
TPC
|
TFC
|
ARA
DPPH
|
ARA
ABTS.+
|
FRAA
|
AA
|
PC
|
L
|
a
|
b
|
Solubility
|
WHC
|
OHC
|
EC
|
ES
|
TPC
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
TFC
|
0.691
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
|
ARA
DPPH
|
0.783
|
0.890
|
1
|
|
|
|
|
|
|
|
|
|
|
|
|
ARA ABTS.+
|
0.933
|
0.546
|
0.644
|
1
|
|
|
|
|
|
|
|
|
|
|
|
FRAA
|
0.947
|
0.800
|
0.918
|
0.889
|
1
|
|
|
|
|
|
|
|
|
|
|
AA
|
0.801
|
0.554
|
0.733
|
0.792
|
0.851
|
1
|
|
|
|
|
|
|
|
|
|
PC
|
-0.208
|
0.103
|
-0.210
|
-0.267
|
-0.236
|
-0.282
|
1
|
|
|
|
|
|
|
|
|
L
|
-0.194
|
-0.145
|
-0.490
|
-0.086
|
-0.330
|
-0.329
|
0.251
|
1
|
|
|
|
|
|
|
|
a
|
0.624
|
0.023
|
0.043
|
0.734
|
0.403
|
0.319
|
-0.261
|
0.363
|
1
|
|
|
|
|
|
|
b
|
-0.108
|
0.114
|
0.138
|
-0.287
|
-0.039
|
-0.356
|
0.062
|
-0.083
|
-0.254
|
1
|
|
|
|
|
|
Solubility
|
0.269
|
0.603
|
0.478
|
-0.027
|
0.294
|
0.336
|
0.267
|
-0.107
|
-0.320
|
0.090
|
1
|
|
|
|
|
WHC
|
-0.273
|
-0.562
|
-0.411
|
0.043
|
-0.242
|
-0.178
|
-0.131
|
-0.056
|
0.163
|
-0.244
|
-0.932
|
1
|
|
|
|
OHC
|
-0.179
|
-0.439
|
-0.641
|
-0.017
|
-0.375
|
-0.172
|
0.115
|
0.872
|
0.502
|
-0.359
|
-0.242
|
0.141
|
1
|
|
|
EC
|
0.200
|
-0.411
|
-0.427
|
0.358
|
-0.068
|
0.042
|
-0.202
|
0.520
|
0.853
|
-0.466
|
-0.416
|
0.275
|
0.781
|
1
|
|
ES
|
0.357
|
-0.243
|
-0.110
|
0.397
|
0.144
|
0.122
|
-0.610
|
0.292
|
0.811
|
-0.023
|
-0.296
|
0.039
|
0.463
|
0.784
|
1
|
TPC: Total phenolic content, TFC: Total flavonoid content, ARADPPH: DPPH radical scavenging activity, ARAABTS.+ ABTS.+ radical scavenging activity, FRAA: Ferric reducing antioxidant activity, AA: Antioxidant activity by phosphomolybdenum, PC: Protein content, WHC: Water holding capacity, OHC: Oil holding capacity, EC: Emulsion capacity, ES: Emulsion stability |
Values in bold with a significance level alpha = 0,05 |
In summary, firstly protein isolation conditions (salt and pH) were optimized for common bean and then genotypic variation for the important techno-functional and bioactive attributes of bean protein concentrates were identified. These differences should be further explored for more prominent genotype selection in food industry applications.