Optimization of fermentation conditions for physcion production of Aspergillus chevalieri BYST01 by response surface methodology

Strains

Aspergillus chevalieri BYST01 was early isolated from the gut of termite and deposited at China Center for Type Culture Collection with the number CCTCC2021285 (Zhang et al., 2024).

Culture media

According to the methods detailed previously (Suwannarach et al., 2019; Zhao et al., 2017), a piece of fungal mycelial from the slant was inoculated on PDA plate at 28 °C for 1 week, and then the periphery of the growing colony was cut into homogeneous plugs with a diameter of 5 mm. Five pieces of fresh plugs were added to flasks containing different liquid media to select the basal medium for the optimization of physcion production. The cultures were then fermented at 28 °C in a shaker rotating at 180 rpm for 11 days.

Nine liquid media, including potato dextrose broth (PDB, 200 g/L fresh potato and 20 g/L glucose, pH 7.0), malt-extract broth (ME, 20 g/L malt extract, 20 g/L glucose and 1 g/L peptone, pH 7.0), sabouraud dextrose broth (SDB, 10 g/L tryptone and 40 g/L glucose; pH 5.6), Oatmeal medium (OA, 30 g/L oatmeal), Czapek yeast extract broth (CZB, 3 g/L NaNO₃, 1 g/L KH₂PO₄, 0.5 g/L MgSO₄•7H₂O, 0.5 g/L KCL, 0.01 g/L FeSO₄ •7H₂O and 30 g/L glucose, pH 7.0), Gause medium G-1 (GA1, 20 g/L soluble starch, 0.5 g/L NaCl, 1 g/L KNO₃, 0.5 g/L K₂HPO₄·3H₂O, 0.5 g/L MgSO₄·7H₂O and 0.01 g/L FeSO₄·7H₂O; pH 7.4), Takashio medium (Taka, 1 g/L KH₂PO₄, 1.0 g/L MgSO₄·7H₂O, 2 g/L glucose and 1 g/L tryptone, pH 7.0), yeast extract phosphate broth (YPB, 1 g/L yeast extract, 6.0 g/L KH₂PO₄, 4 g/L NaH₂PO₄ and 1 g/L NH₄OH, pH 7.0) and dextrose peptone medium (DPM, 60 g glucose, 20 g/L peptone, 2 g/L KH₂PO₄, and 1 g/L MgSO₄•7H₂O, pH 7.0) were used for screening the basal medium.

Quantification of physcion by HPLC-DAD analyses

For shake flask fermentation, the fungus A. chevalieri BYST01 was cultivated on PDA plate at 28 °C for 7 days and about five plugs with diameter of 5 mm were transferred into the nine media described above each 250 mL Erlenmeyer flask containing 100 mL liquid medium. After cultivating at 28 °C and 180 rpm for 11 days, the yield of physcion in different medium was observed by HPLC analyses.

After fermentation, 2 mL of chromatographic methanol was added to equal volume of fermentation broth, and then the mixture was centrifuged for 10 min. Further filter the supernatant with 0.22 μm membrane filter. The filtrate is used for subsequent HPLC analysis. Chromatographic analysis was performed by using a C18 reverse-phase analytical column: Fisher Wharton Xbridge C18, 5 μm, 10 mm * 250 mm. SHIMADZU LC16 parameters were as follows: the column oven temperature was set at 40 °C, the injection volume was 5 μL with a flow rate of 1.0 mL/min. The mobile phase consisted of deionized water and 100% methanol at a ratio of 15% to 85% (v/v) for 20 min. For quantification, standard physcion (0.50 mg) was dissolved in 1.0 ml of DMSO and used as a standard stock solution for generating calibration curves. The stock solution was diluted 1/2, 1/4, 1/8, 1/16, 1/32 and 1/64 in DMSO solution to afford gradient solutions. The calibration curves were made from the physcion solutions using a linear fit for the relationship of area vs. concentration of physcion at UV 254 nm.

Single factor experiments for physcion production

The PDB medium was chosen as the basal medium for further optimization for physcion production under different medium compositions (carbon source, nitrogen source and inorganic salts, etc.). Carbon source, such as glucose, sucrose, maltose, xylose and fructose were tested to examine their effects on the physcion production. The glucose (20 g/L) in the normal PDB medium was replaced by these carbons at the same concentration. And the influence of different concentrations of the carbon (glucose, 10–40 g/L) were also analyzed. Different nitrogen source supplements and inorganic salts source were set to evaluate their effects on the physcion production. The extra nitrogen source, such as glycine, urea, NH₄Cl, (NH₄)₂SO₄ and yeast extract were individually added at 0.3% (w/v) concentration. The extra inorganic salts source (CH₃COONa, MnCl₂, NaCl, GaSO₄, KCl and FeCl₃) were individually added at 0.05% (w/v) concentration.

Moreover, to improve the physcion production, the fermentation conditions were optimized at different fermentation time (7, 9, 11, 13, 15 and 17 d), temperature (20, 22, 24, 26, 28, 30 and 32 °C), the initial pH (4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5 and 8.0) and the speed of the rotary shaker (90, 120, 150, 180, 210 and 240 rpm).

The fermentations were performed in 250 mL of medium containing of 100 mL medium broth. The experiments were performed in triplicate, and the mean values were used to minimize variations.

Optimization by Plackett-Burman design (PBD)

The Plackett-Burman design (PBD) is a normal method used to identify the important factors in the early stage of response surface methodology (RSM) (Wu et al., 2022; Zaki et al., 2019). An initial screen was carried out for above five single factors (different fermentation time, glucose concentration, temperature, pH and the speed of the rotary shaker) using a PBD experimental design (Design-Expert 10.0.3). The design set was shown in Table 1 and statistical data was shown in Table 2. Each independent variable is represented in two levels, high and low, which are denoted by (1) and (−1), respectively. Twelve runs of various levels of factors were formulated by the software and the output response represented the yield of physcion. The t test was used to determine the significance of the regression coefficients. The variable with a P value less than 0.05 was considered as a significant factor (Tajik et al., 2024).

Optimization by response surface methodology (RSM)

Following the three variables (temperature, initial pH and the speed of a rotary shaker) that are capable of significantly influencing the physcion production were screened by the PBD, a useful response surface methodology (the Box-Behnken experimental design, BBD) was applied using the Design-Expert software to optimize the physcion production conditions. The three independent variables were analyzed at three levels (−1, 0, and 1) with 17 experimental runs in Table 3. Based on BBD, the optimized objective was to achieve the maximum physcion production using this software. Based on BBD, experiments with different fermentation conditions (temperature, initial pH and speed of a rotary shake) under a constant culture medium were conducted. The physcion yield was subjected to analysis of variance (ANOVA), and the experimental data of BBD were fit with the following second-order polynomial equation: Y = β₀ − β₁A − β₂B + β₃C + β₁₂AB − β₁₃AC − β₂₃BC − β₁₁A² − β₂₂B² − β₃₃C². Where Y is the predicted physcion yield, A (temperature), B (speed of a rotary shake) and C (initial pH), are the initial independent variables; β₁, β₂ and β₃ are the linear coefficients; β₁₂, β₁₃ and β₂₃ are the interaction coefficients; and β₁₁, β₂₂ and β₃₃ are the quadratic coefficients.

Based on the results of the suggested optimal culture conditions by the soft, a new experiment was performed under these optimal parameters. The target physcion yield obtained using these conditions present in Table 3 was compared with that obtained using the second-order polynomial equation calculated.

Bioreactor fermentation

According to the shaking flask experiment results, the optimized fermentation medium was further verified in 5.0 L stirred fermenter (BIOTECH-5BG; Shanghai Baoxing Bio-Engineering Equipment Co., Ltd., Shanghai, China) filled with 2.0 L broth. The strain plugs were cultured in PDB for 3 days prepare seed culture. Then 20% seed culture was transferred to the optimized media. The fermentation was incubated at 28 °C with a stirring speed of 180 rpm and pH 6.6. The physcion yields obtained in fermenter and flask were then compared.

The basal medium selected for optimization

Total of nine media were used to select the basal medium for optimization. After a 11 d fermentation progress, HPLC quantification of production showed that PDB was the optimal medium for production of physcion with a yield of 28.0 mg/L. The yield of physcion in ME (18.4 mg/mL) and SDB (15.8 mg/mL), which were lower than that of PDB (Fig. 1A). Only trace physcion was detected in GA1 and Takashio medium, while no physcion was detected in YPB and OA media (Fig. 1A). The biomass of the strain cultured in different media were evaluated and showed no significant correlation with production of physcion (Fig. 1B). Thus, PDB was designated as the basal medium for following fermentation conditions optimization.

The effect of single factor experiments of physcion production

Common parameters (such as carbon source, nitrogen source, inorganic salts, temperature, initial pH, fermentation time and the speed of the rotary shaker) have a certain effect on physcion production, and these different parameters were tested by one single factor method.

As showed in Fig. 2A, a significant difference of the physcion yield was observed by the PDB consisting with different carbon source at the same concentration of 20 g/L. As the glucose was used as the carbon source added to the PDB, the maximum physcion production (30.0 mg/L) was presented. In addition, the glucose concentration of the PDB had an important role in producing physcion. Results in Fig. 2B indicated that the physcion yield reached its highest productivity (43.0 mg/L) at glucose concentration of 30 g/L. Figure 2C indicated that extra nitrogen source adding to PDB had a negative impact on production of physcion. The physcion productivity of BYST01 were decreased under these adding extra nitrogen source (glycine, urea, NH₄Cl, (NH₄)₂SO₄ and yeast extract) with 3.5–12.7 mg/L, which were much lower than that of normal PDB (28.0 mg/L). In addition, the tested inorganic salts (CH₃COONa, MnCl₂, NaCl, GaSO4, KCl and FeCl₃) also inhibited the physcion production (Fig. 2D). The fermentation time had a certain influence on the accumulation of physcion. Early termination of fermentation process will affect the physcion production, while long-term fermentation will cause waste of resources and even lead to the metabolism and decomposition of physcion, thus reducing the production. As presented in Fig. 2E, the yield of physcion reached a peak at 11 d with a value of 40.0 mg/L in normal PDB consisting with 30 g/L glucose, and then decreased.

For screening the optimal temperature for fermentation, a series of temperature was set. The production of physcion was sensitive to the temperature. It could be seen in Fig. 2F that with the increase of temperature, the yield of physcion increased gradually, and reached the maximum at 28 °C, with the yield of 40.0 mg/L. When the temperature was over 28 °C, the yield of physcion decreased. The production of physcion was evaluated at different pH. The results showed that with the gradual increase of pH, the yield of physcion also gradually increased. The maximum yield was reached at pH 6.5 with a value of 74.0 mg/L, and this pH was the optimal pH for production of physcion (Fig. 2G). Following the same procedure, the influence of agitation rate on the accumulation of physcion was observed. The highest physcion production was achieved at 180 rpm. (Fig. 2H).

Thus, the optimal fermentation parameters were fermented in PDB consisting with 30 g/L glucose at 28 °C and pH 6.5 with continuous shaking at 180 rpm for 11 days by single factor experiments.

Plackett-Burman design (PBD)

Based on the results of the single-factor experiments, the main factors, such as glucose concentration, temperature, initial pH, fermentation time, and the speed of the rotary shaker, that may play an important role in the production of physcion, were further analyzed using PBD experiments (Table 1).

The results of the PBD were statistically analyzed, as shown in Table 2. To determine the optimum response, a fitted first-order model for physcion production was obtained from the PBD as follows: Y (“Y” represents physcion yield, mg/L) = 9.63 + 3.61A + 2.24B − 1.28C + 1.64D + 0.29E, where Y is the predicted physcion yield; A, B, C, D, and E, are the coded factors of temperature (°C), rotating speed (rpm), glucose concentration (g/L), pH, and fermentation time (d), respectively. A factor with a confidence level greater than 95% (P < 0.05) was considered to have a significant effect on physcion yield and was selected for further study. The linear regression coefficient (R²) was 0.9160, the adjusted R² was 0.8360, and the P-value was 0.0042 (P < 0.05) for the model. These results indicated that the model is suitable for the PBD experimental design. Based on the PBD results, among the five studied variables, temperature, initial pH, and the speed of the rotary shaker were the three factors significantly correlated with the yield of physcion (P < 0.05).

Response surface methodology (RSM)

Three combinations of components, temperature, initial pH, and rotating speed were selected for optimization using RSM, and the experimental physcion yields were recorded. The results of 17 runs from the Box-Behnken experiments for the effects of three independent variables are represented in Table 3. The maximum experimental value for physcion production was 82.25 mg/L based on the RSM. The predicted response (Y) for physcion production obtained from further linear multiple regression analysis of the experimental data was described as follows: Y = 79.04 − 11.67A − 6.23B + 5.2C + 2.33AB − 8.76AC − 7.48BC − 28.09A² − 33.86B² − 29.33C², where Y is the predicated physcion yield, and A–C are the temperature, rotation speed, and initial pH, respectively. The statistical significance of this equation was confirmed using Fisher’s F-test. The model F-value of 61.52 indicated that the model was significant (Table 4). There was only a 0.01% chance that such a large F-value could occur due to noise. The linear regression coefficient R² was 0.9875, and the adjusted R² was 0.9715 for the model. These results indicated that the model was suitable for explaining the experimental parameters and their interactions. Of the first terms (A, B, and C), temperature (A), rotation speed (B), and initial pH (C) had significant effects on the physcion yield (P < 0.05), with the P-values of 0.0004, 0.0122 and 0.0267, respectively. Of the interaction terms (AB, AC, BC), temperature (A) interacting with initial pH (C) and rotating speed (B) interacting with initial pH (C) had significant effects on physcion yield (P < 0.05). The quadratic terms (A², B², and C²) had an extremely significant effect on physcion yield (P < 0.001).

To better understand the effects of the variables and their interactions on physcion yield, three-dimensional and two-dimensional response surface curves were exhibited in Fig. 3, based on the constructed model. Three-dimensional graphs were generated for the combination of the two variables, keeping the other one at the optimum level determined by the path of the steepest ascent for physcion production by BYST01. The response surface was convex, suggesting that the optimum conditions were well defined, and there was a maximum for each variable. Response surfaces were generated as an interaction function of temperature, rotation speed, and pH and are presented as 3D surface responses in Figs. 3A–3C, respectively. The effects of temperature, rotation speed, and pH on physcion production are also presented in a 2D contour form (Figs. 3A–3C). Based on the RSM analysis, the predicted maximum production of physcion would be 81.12 mg/L, when the uncoded levels of the fermentation conditions were set at 27.5 °C with a rotary shacking speed of 176.5 rpm and an initial pH of 6.57. Validation of these predictions related to physcion production was confirmed by triplicate laboratory trials in flask culture, and the yield of physcion was 82.0 mg/L under the modified conditions (28 °C, 177 rpm, and pH 6.6). There was almost no difference from the predicted values.

Production of physcion at bioreactor scale

To evaluate the possibility of using the optimized fermentation medium for physcion production at a larger bioreactor scale, fermentation of BYST01 was carried out in a 5 L bioreactor. In this study, the optimized medium supported the maximum productivity of physcion at 85.2 mg/L on the eighth day of fermentation, which was consistent with that of the shaker flask. This indicated that the optimum culture conditions obtained by the RSM experimental design in submerged fermentation supported the enhanced production of physcion in a bioreactor.