청국장에 대한 ε-Poly-L-lysine 혼합제재의 항 미생물 효과 및 혼합비율의 최적화

Raw soybeans were purchased from Dalseong-Nonghyup (Daegu, Korea). Bacillus subtilis KCCM-1089P used in the fermentation of cheonggukjang was obtained from KCCM (Seoul, Korea). ε-Poly-L-lysine and ethanol were obtained from Shinseung Hichem (Seoul, Korea). 98.2% ethanol and 0.8% ε-poly-L-lysine (EPM) were mixed to determine the antibacterial effect upon cheonggukjang.

Three kilograms of soybeans were put in water at 20 ° C for 12 h, steamed for 30 min, and cooled at 40 ° C. 20 mL of B. subtilis and 400 mL of water were blended with the soybeans to produce the culture solution. This solution was cultured at 42 ° C for 24 h and cheonggukjang was subsequently manufactured (Kim et al., 2008). Various amounts of EPM (0-2% by weight of cheonggukjang) were commingled with prepared cheonggukjang.

Ten gram of crushed sample in 90 mL of water was filtered through filter paper (Whatman No.5). pH was measured using an Orion 710 A+ pH meter (Thermo Fisher Scientific, Beverly, MA, USA). The color of the samples was measured using a Colorimeter (JC-801, Color Techno System. Co. Ltd, Tokyo, Japan) and L (lightness) was determined.

Antimicrobial activity was measured using Escherichia coli KCTC-1039 or Bacillus cereus KCTC-1021. 50 μL of 0-2% EPM was spread on an LB agar plate. A 7 mL layer of top agar (0.8%) containing 100 μL each of E. coli or B. cereus in culture medium was then added to the plate and the plate was incubated at 37 ° C until a clear zone appeared.

In order to measure viable cells, 10 g of the sample in 90 mL of water was homogenized using an ACE homogenizer (Ultra- Turrax T25, Janke and Kunkel, Brussels, Belgium) set at 15,000 rpm. 1 mL of the homogenized sample was spread on potato dextrose agar (Difco, Detroit, MI, USA) and colony number was measured after 48 h at 30 ° C.

Sensory analysis was carried out by 10 expert panelists. Each panelist evaluated 10 g of boiled sample to which 1 g of salt had been added. A nine-point hedonic scale of 1 (disliked extremely) to 9 (liked extremely) was used by the panelists to evaluate taste, texture, flavor and overall preference.

The experimental mixture design and statistical analysis were performed using Design Expert software version 7 (Stat- Easy Co., Minneapolis, MN, USA). A D-optimal design consisting of 13 experimental runs, including five replicates at the center point, was chosen to evaluate the combined effect of two independent variables; Cheonggukjang (X₁) and EPM (X₂). Based on preliminary experiments, the ranges of the two samples were 0.98 ≤ X₁≤1 and 0 ≤ X₂≤ 0.02, respectively. The response values utilized were viable cells, pH, texture, and overall preference. The effect and regression coefficients of individual linear, interactive, and cubic terms were determined. Results were considered to be statistically significant if they had a probability (p) of 0.05. The numerical optimization was performed using the canonical model basis with a set response goal area and desirability was predicted.

All data analysis was performed using SPSS (Statistical Package for the Social Sciences v20.0, Chicago, IL, USA). The level of significance among samples was tested with analysis of variance (ANOVA) and Duncan's multiple range tests performed at the p<0.05 level.

pH of cheonggukjang mixed with EPM stored at two temperatures, 15 and 30 ° C, was measured over a 192 h period, as shown in Fig. 1. The pH of cheonggukjang with 2% EPM increased from 6.33 to 6.59 and from 6.31 to 7.19 when stored for 192 h at 15 and 30 ° C, respectively. The pH of the control without EPM was the lowest after 192 h, having decreased from 6.38 to 6.00 and from 6.34 to 5.55 after storage at 15 and 30 ° C, respectively. In general, pH levels were maintained or slightly increased in samples containing between 0.5 and 1.5% EPM, whereas the control samples showed an increase in pH during the initial 48 h period followed by a continuous decrease in pH over the remainder of the 192 h period. This increase in pH during the initial 48 h storage period might be due to the production of ammonia gas during initial fermentation while the subsequent gradual decrease in pH of cheonggukjang may be due to an increase in acidity resulting from the growth of lactic acid bacteria (Suh et al., 1982; Sung et al., 1984). In general, the pH of the cheonggukjang with EPM showed a trend of increasing or maintained pH during storage at either 15 or 30 ° C.

Fig. 1. Changes in pH of cheonggukjang with different EPM concentrations during storage at 15°C (A) and at 30°C (B).

Download Original Figure

The antimicrobial activity of EPM was tested by evaluating the growth of E. coli and B. cereus. Potato dextrose agar plate was used for incubation of those strains. Growth of E. coli was restricted at 0.4 mg/L EPM, whereas 0.2 mg/L was the minimum concentration for the restriction on growth of B. cereus, demonstrating that higher level of EPM was needed for growth restriction of E. coli than of B. cereus. The viable cell counts in Cheonggukjang with 0-2% added EPM stored at two temperatures (15 and 30 ° C) are shown in Fig. 2. Higher EPM concentrations resulted in greater microbial growth inhibition. The most significant reduction in viable cell count was seen with 2% EPM, with a reduction from 6.6×10 ⁵ CFU/mL to 8.95×10 ³ CFU/mL at 15 ° C (Fig. 2A) and 6.6×10 ⁵ CFU/mL to 4.6×10 ⁴ CFU/mL at 30 ° C (Fig. 2B). Addition of 1.5-2% EPM at 15 ° C resulted in viable cell counts below 10 ⁵ CFU/mL. These concentrations of EPM showed relatively higher bacterial growth inhibition than lower EPM concentrations (below 1.5%) at both storage temperatures. When 1.0-1.5% EPM and fermented vinegar mixtures were added to yakbab, yeast growth was most effectively inhibited (Kim, 2004). Park (2000) also reported that polylysine mixtures inhibited microbial growth and that fermentation occurred slowly in kimchi. More viable cells remained at higher EPM concentrations following 30 ° C storage than following 15 ° C storage due to the optimal growth temperature of viable cells (20-40 ° C).

Fig. 2. Viable cell count in cheonggukjang with different EPM concentrations during storage at 15°C (A) and at 30°C (B).

Download Original Figure

The color changes in terms of L values (lightness) of the cheonggukjangs at 15 and 30 ° C are shown in Table 1. L values in samples without EPM decreased from 53.41 to 49.23 at 15 ° C and also decreased from 57.33 to 47.69 at 30 ° C. Furthermore, L values in samples with 2% EPM decreased from 55.40 to 49.69 at 15 ° C and from 56.23 to 47.69 at 30 ° C, indicating that cheonggukjang turned dark during the fermentation time with or without EPM. This is similar to the report of Choi et al. (1998) in that as fermentation time increased, lightness of cheonggukjang decreased (i.e.: turned dark). When the EPM content was increased, L values were not changed substantially except in the 72 h storage time point at 30 ° C. With respect to the effect of temperature, L values at 30 ° C (49.23-52.24 at 192 h) were generally lowered (darker) than at 15 ° C (47.69-50.75 at 192 h).

The mean sensory scores for taste, texture, flavor, and overall preference of cheonggukjang with different EPM concentrations (0-2%) fermented at 30 ° C for 24 h are shown in Table 2. The control samples without EPM received the lowest flavor (2.4) and texture scores (3.4), while increasing EPM levels received increasing flavor and texture scores. 2% EPM received the highest scores (3.5 and 5.3). Consequently, the specific flavor and texture of cheonggukjang was improved by the addition of EPM. Taste was not found to be significantly different among EPM concentrations. A similar finding was reported by Ko & Kim (2004) in that there was no significant difference in taste regardless of the polylysine concentration utilized. But the flavor that was reason for the reluctance of cheonggukjang was improved by the addition of EPM. Overall preference was also not significantly different between different EPM concentrations.

Mixture design is an important methodology for experiments in which the variable factors are the proportions of the components of a mixture and the response variables vary as a function of these proportions (Choi et al., 2006). The range of mixture ratios of cheonggukjang and EPM was 98- 100% and 0-2% (w/w), respectively. The total content of mixed cheonggukjang (sum of all variables) was 100% (w/w), without the inclusion of the fixed variables. In order to allocate the points for the mixture within the feasible design region, a modified distance design was applied. Thirteen total design points were set for the experimental designs: three experimental points, five points for calculating the lack of fit, and five replication points (Table 3). In order to remove division errors for the order of the actual experimental designs, all experimental orders were performed at random. According to a mixture design, four kinds of responses, including viable cell count, pH, texture and overall preference, were chosen.

In order to analyze the interaction effect on each cheonggukjang mixture, modeling was necessary for each response (Han & Kim, 2003). Analysis of the selected models and regressions of the polynomial equations are shown in Table 4. The significance of the selected models was determined by the F-test. Viable cell count and pH were each selected to a quadratic and cubic non-linear model, respectively (0.0025 and 0.0126 probability values). But texture and overall acceptability selected to linear model meant independent contribution. Each coefficient determined at the predicted canonical equation showed the effect of each cheonggukjang on each response as a numerical value.

The optimum mixture ratio of cheonggukjang with EPM was determined using the optimization process suggested by Derringer and Suich (1980). While texture and overall preference were set to the maximized value, viable cells and pH were each set to the minimized value and target 6.5 values, respectively (Table 5). The desirability constraint used the criterion of degree of optimization. From the numerical optimization results, the optimum cheonggukjang formulation was determined to be 98.52% cheonggukjang and 1.48% EPM. The optimum process resulted 72.2.% desirability and the predicted response values for this mixture ratio showed that the viable cell, pH, texture, and overall preference scores were 1.56×10 ⁵ CFU/mL, 6.5, 4.38 and 5.34, respectively (Table 5).