Novel Bacillussubtilis IND19 cell factory for the simultaneous production of carboxy methyl cellulase and protease using cow dung substrate in solid-substrate fermentation

Background Hydrolytic enzymes, such as cellulases and proteases, have various applications, including bioethanol production, extraction of fruit and vegetable juice, detergent formulation, and leather processing. Solid-substrate fermentation has been an emerging method to utilize low-cost agricultural residues for the production of these enzymes. Although the production of carboxy methyl cellulase (CMCase) and protease in solid state fermentation (SSF) have been studied extensively, research investigating multienzyme production in a single fermentation process is limited. The production of multienzymes from a single fermentation system could reduce the overall production cost of enzymes. In order to achieve enhanced production of enzymes, the response surface methodology (RSM) was applied. Results Bacillus subtilis IND19 utilized cow dung substrates for the production of CMCase and protease. A central composite design and a RSM were used to determine the optimal concentrations of peptone, NaH2PO4, and medium pH. Maximum productions of CMCase and protease were observed at 0.9 % peptone, 0.78 % NaH2PO4, and medium pH of 8.41, and 1 % peptone, 0.72 % NaH2PO4, and medium pH of 8.11, respectively. Under the optimized conditions, the experimental yield of CMCase and protease reached 473.01 and 4643 U/g, which were notably close to the predicted response (485.05 and 4710 U/g). These findings corresponded to an overall increase of 2.1- and 2.5-fold in CMCase and protease productions, respectively. Conclusions Utilization of cow dung for the production of enzymes is critical to producing multienzymes in a single fermentation step. Cow dung is available in large quantity throughout the year. This report is the first to describe simultaneous production of CMCase and protease using cow dung. This substrate could be directly used as the culture medium without any pretreatment for the production of these enzymes at an industrial scale.


Background
Cellulases catalyze the hydrolysis of cellulose, and many microorganisms, including fungi, bacteria, and protozoans, to produce cellulase [1]. In recent years, cellulolytic enzymes from Saccharomyces cerevisiae [2], Talaromyces cellulolyticus [3], and S. cerevisiae TJ14 [4] have been identified and characterized for various biotechnological processes. These enzymes have many useful applications in the paper industry, bioethanol generation, extraction of fruit and vegetable juice, textiles, the detergent industry, and animal feed production [5][6][7]. Proteases are an important group of industrial enzymes and are widely used in the food, chemical, pharmaceutical, and leather processing industries [8]. The global market for these enzymes could reach $4.4 billion by the year 2015, and the maximum sales of industrial enzymes came from the leather and bioethanol market [9]. It was previously reported that the cost of growth medium covered approximately 30-40 % of production cost of industrial enzymes [10]. Hence, simultaneous production of cellulase and protease could help to reduce cost. Research examining novel substrates for the production of cellulase and protease has been a continuous effort.
SSF has been an emerging method to utilize the costeffective agro-residues to produce cellulases and proteases [11,12]. In the last two decades, SSF has attracted attention in Western countries due to its advantages in the production of secondary metabolites, enzymes, and novel foods [13]. In SSF, the cheap substrates, such as banana fruit stalk, wheat straw, paddy straw, apple pomace, sugarcane bagasse, oil palm empty fruit bunch, green gram husk, Imperata cylindrical grass and potato peel, and pigeon pea, have been utilized for the production of cellulase and protease [14][15][16][17][18][19][20][21][22][23]. Although these agroresidues were regarded as the potential substrates in SSF, their availability is largely seasonal. The ideal substrate should be available throughout the year and be cheap. Therefore, cow dung is a possible substrate. Cow dung is rich in cellulose (35.4 %), hemicelluloses (32.6 %), ash (13.3 %), nitrogen (1.4 %), and traces of minerals, such as nitrogen, potassium, and sulphur, and traces of phosphate, iron, cobalt, magnesium, potassium, chloride, and manganese [24].
Most cellulolytic enzymes used in industry are of fungal origin; however, these enzymes lack stability at high temperatures. Because many industrial processes are carried out at high temperatures, there is a need for thermostable enzymes from other sources [25]. Cellulases of bacterial origin have potent activity with crystalline celluloses. These enzymes showed high activity and stability towards alkaline pH and are thermostable in nature compared with the fungal cellulases [26]. Cellulases produced by bacteria are notably high in quantity, whereas the fungal cellulases are mostly inducible in nature [27]. Likewise, a wide range of bacteria are known to produce proteases; a large proportion of the commercially available proteolytic enzymes are derived from the genus Bacillus because of their capacity to produce large amounts of alkaline proteases with significant activity and stability at high temperature and pH [8,28].
The traditional method to evaluate the optimal conditions for enzyme production is based on one-variable-ata-time approach. However, this approach fails to reflect the interactive effects among the selected factors or variables and it is a time-consuming process and requires multiple experimental runs. Additionally, this method does not guarantee to find accurate optimal conditions. However, statistical methods, such as response surface methodology (RSM), have been greatly used to determine the optimum level of factors in a bioprocess [29,30]. RSM is a collection of statistical techniques for designing experiments, searching the significant factors, and evaluating optimum conditions, that has been successfully used in the optimization of many bioprocesses [31]. In RSM, 3D plots help to better identify the maximum response and interactions among the tested variables [32]. There have been many studies on RSM-mediated optimization of enzyme production from various microorganisms [33][34][35][36].
Well-established enzyme engineering is required for the effective and simultaneous production of multienzymes in a single fermentation [37,38]. In a multienzyme production system, the supplement of various nutrients are critical, and not all nutrients may enhance the simultaneous production of all enzymes [39]. More than two or three enzymes have been produced in a particular environmental condition by microorganisms, specifically Bacillus sp. Multienzyme production is a complex process that is associated with complex patterns of repression and induction resulting from the mixed substrate environment, pH, moisture content, fermentation time, and inoculum concentration in SSF [40]. The interaction among these factors becomes the key aspect for investigation in the multienzyme production in SSF. Several reports are available for Bacillus sp. for the production of concomitant enzyme production, including lipase and protease [41], amylase and protease [42], proteases and amylases [43]. However, the reports on simultaneous production of CMCase and proteases from Bacillus sp. are limited and perhaps not available. Recently, cow dung was used as the solid substrate for the production of protease [12] and CMCase [36]. To the best of our knowledge, the current study is the first to report simultaneous production of CMCase and protease using cow dung substrate in SSF. Considering the production cost of CMCase and protease, this paper identified the optimum conditions for the production of these enzymes by Bacillus subtilis IND19. A statistical approach was employed to identify the significant factors and RSM was used to obtain the optimized conditions for CMCase and protease production in SSF utilizing cow dung substrate.

Screening of B. subtilis IND19 for cellulolytic and proteolytic activity
In the present study, seven potential cellulolytic bacterial strains were used, which hydrolysed CMC with the zone range of 3.0-6.0 mm. The bacterial isolates, such as VA1, VA2, VA4, VA5, VA6, and VA7, hydrolysed 5, 3, 4, 5, 3, 3 mm, respectively, on CMC agar plates. The CMCase activity of B. subtilis IND19 was higher (6 mm) than the other screened bacterial isolates. Cellulase production of the bacterial strains from the genus Bacillus has been reported by various studies [44][45][46]. The cellulolytic enzyme-producing bacterial isolates, such as VA1, VA2, VA3, VA4, VA5, VA6, and VA7, were evaluated for protease production on skimmed milk agar plates. Among the tested bacterial strain, B. subtilis IND19 showed the maximum production of protease on skimmed milk agar plates (12 mm). The other tested isolates showed hydrolytic zone ranging from 3 to 11 mm. Hence, B. subtilis IND19 was selected for simultaneous production of CMCase and protease.

Cow dung is a substrate of choice for simultaneous production of CMCase and protease
In this paper, cow dung was explored as the low-cost substrate for the simultaneous production of CMCase and protease. This low-cost substrate could lower the production cost of enzymes. Because the production of hydrolytic enzymes using different fermentation processes is notably expensive, and the simultaneous production of several industrial enzymes in a single fermentation medium is a great challenge [47]. Cow dung was attempted for enzyme production. The selection of suitable solid waste for any enzyme production in an SSF process mainly depends on the cost and availability of the substrate material [48]. In recent years, many substrates have been reported for the production of CMCase and protease [17,20,23,36,49]. Considering availability and cost, cow dung is a suitable substrate for the production of cellulase and protease. Reports on SSF of cow dung for the simultaneous production of cellulolytic and proteolytic enzymes using bacteria are limited or perhaps not available. This report could be the first to describe the simultaneous production of CMCase and protease in SSF using cow dung substrate.

Screening variables for the production of CMCase and protease by statistical approach
Initial screening of medium components indicated that carbon source (sucrose), nitrogen source (peptone), addition of salt solution (NaH 2 PO 4 ), and variation of medium pH induced the CMCase and protease production. A statistical approach (2 5 full factorial design) was used to identify the most effective variables affecting CMCase and protease production. All experiments were carried out under SSF for 72 h at 37 °C in duplicates. The experimental values of two-level full factorial design for the production of CMCase and protease are given in Table 1. CMCase production varied between 41.5 and 497.4 U/g and protease yield varied from 206.5 to 4778.2 U/g. The variability in the yield of enzyme production in this paper provides space for the optimization of enzyme production. The F values of this model for CMCase and protease activities were 49.75 and 75.06 U/g, respectively, which were statistically significant at the 5 % level. In this paper, sucrose, peptone, NaH 2 PO 4 , and the initial pH and moisture content of the culture medium significantly influenced the production of both enzymes ( Table 2). These results were in accordance with the observations made with Chaetomium sp. on cellulase production in SSF [50], suggesting that sucrose was the best carbon source for cellulase production. However, cellulose was demonstrated to be the best carbon source for cellulase production from Bacillus sp. [51]. Addition of peptone to the cow dung medium positively influenced both CMCase and protease production. Umikalsom et al. [52] recorded peptone as the suitable nitrogen source for the production of cellulase by Chaetomium globosum in SSF using delignified oil empty fruit bunch fibre as substrate. Likewise, another report also suggested peptone as the best nitrogen source for the cellulase production from Marinobacter sp. MSI032 [53]. In this paper, protease production was enhanced by the supplement of sucrose as the carbon source. This result was in accordance with the observations made with Yarrowia lipolytica [54] and Bacillus sp. [55]. The R 2 of the model values for the production of CMCase and protease were 0.9970 and 0.9954, and the adjusted R 2 was 0.977 and 0.9821, respectively. The regression equation coefficients of the 2 5 full factorial models were calculated and the data were well fitted.
Final equations in terms of coded factors.

CMCase activity
Enzyme

Central composite design
Optimizing process parameters was carried out using RSM. The factors-namely, pH, peptone, and NaH 2 PO 4 , which significantly influenced both CMCase and protease production-were selected for further optimization using central composite design (CCD) to maximize the CMCase and protease production. Our findings showed that peptone, NaH 2 PO 4 , and pH positively influenced CMCase and protease production. However, an excessive concentration of NaH 2 PO 4 had a negative effect on protease production. Most cellulases and proteases are inducible enzymes and addition of carbon sources, such as sucrose, mannitol, and maltose, enhanced the production of cellulolytic and proteolytic enzymes [56,57]. It was previously reported that the production of protease was enhanced by the addition of nitrogen sources, such as tryptone, peptone, yeast extract, skimmed milk, and soybean meal [58]. The observed response in the production of CMCase and protease is shown in Table 3. As shown in Table 4, the p value of the model generated was <0.05, suggesting the CMCase and protease activity could be well-described by this model. Analysis of variance (ANOVA) was carried out to establish a response surface quadratic model. The model F values of 67.14 and 197.54 implied that both quadratic models for the production of CMCase and protease were significant. The model terms, such as A, B, C, AB, A 2 , B 2 , and C 2 , were significant for the production of CMCase; B, C, AB, AC, BC, A 2 , and B 2 were significant for protease production. For CMCase production, the R 2 of the model was 0.9837, indicating that the experimental data agreed well with the model prediction.
The model could explain 98.37 % variability observed in the data [59]. In the case of protease production, the R 2 value was 0.9944. The model can explain 99.44 % variability observed in the data. The lack of fit values were 3.17 and 0.9658 for CMCase and protease production, respectively, which were not significant. The signal-to-noise  The 3D response surface curves in Fig. 1a-c show the interactions among pH, peptone, and NaH 2 PO 4 . These 3D graphs are helpful to identify the interaction between the variables and their levels. The increase in CMCase production was observed in peptone and pH (Fig. 1a), NaH 2 PO 4 and pH (Fig. 1b), and NaH 2 PO 4 and peptone (Fig. 1c). However, further increase in all of these three variables beyond the optimized level decreased the production of enzymes. Similarly, the protease production was increased by increasing the concentrations of peptone and NaH 2 PO 4 (Fig. 2a-c) and was decreased after optimum concentrations of these factors. This was consistent with the fact that CMCase and protease were generally induced in the presence of carbon, nitrogen, minerals, and alteration of pH [18,23,36].
RSM has been widely used for the production of enzymes in SSF by various studies [60][61][62][63]. It helps to identify the interactive effects of selected parameters and requires the minimum number of experimental runs [64]. In this paper, the maximum CMCase and protease production were observed at 0.9 % peptone, 0.78 % NaH 2 PO 4 , and a substrate pH 8.41, and 1 % peptone, 0.72 % NaH 2 PO 4 , and a substrate pH of 8.11, respectively. Under the optimized conditions, the experimental yield of CMCase and protease reached 473.01 and 4643 U/g, which corresponded to the increase of 2.1-fold and 2.5fold in CMCase and protease production. This finding could be observed because cow dung is a complex biomass already containing essential nutrients for the growth of microbes [24]. Hence, the addition of nutrient sources merely increased approximately twofold on CMCase and protease production.

Validation of the experimental model
The response surface model was validated with triplicate experiments under the predicted experimental conditions. The predicted response for CMCase production was 485.05 U/g, which was very close to the experimental Fig. 1  value (473.01 U/g), thereby validating this model. The predicted response of the model for the production of protease was 4710 U/g, and the experimental value was 4643 U/g, which validated the model.

SDS-PAGE analysis of the extracellular protein from B. subtilis IND19
SDS-PAGE analysis revealed the protein pattern from the crude extract of B. subtilis IND19 (Fig. 3a). Zymogram analysis of the crude CMCase exhibited a band which corresponds to 44 kDa (Fig. 3b). This result was in accordance with the observations made with other Bacillus sp. [65]. The molecular weight of the protease was calculated and was found to be approximately 36.12 kDa (Fig. 3c). The molecular weight of protease was similar to that of the previous reports. Generally, the molecular masses of proteases from various Bacillus species range between 17 and 44 kDa [66,67].

Conclusions
This study aimed to optimize the simultaneous production of CMCase and protease by Bacillus subtilis IND19 with RSM. This report describes the first time that cow dung was applied as the substrate for the simultaneous production of these two enzymes in a single fermentation system. This cheap substrate could be useful for the production of CMCase and protease at industrial scale. RSM-mediated experimental design exhibited an increase of 2.1-and 2.5-fold, respectively, for CMCase and protease compared to non-optimized medium. This paper revealed that RSM is a suitable statistical tool in optimizing enzyme production with minimum experimental runs.

Microorganism
The CMCase-and protease-producing B. subtilis IND19 was isolated from the soil sample. The isolated B. subtilis IND19 was maintained on nutrient agar slants (in g/l) (peptic digest of animal tissue, 5.0; beef extract, 1.5; yeast extract, 1.5; sodium chloride, 5.0; and agar, 15) and stored at 4 °C for further experiments. This organism was subcultured every 30 days.

Screening of B. subtilis IND19 for cellulolytic and proteolytic enzyme
CMCase screening was carried out using carboxy methyl cellulose (CMC) agar medium (in g/l) (beef extract, 5.0; peptic digest of animal tissue, 5.0; yeast extract, 1.5; sodium chloride, 5.0; agar, 15; and CMC, 10). The bacterial growth was visible on these plates after 48 h incubation at 37 °C. To visualize the hydrolysis of CMC agar medium, the plate was stained with Gram's iodine solution. This formed a bluish black complex with CMC and gave a distinct zone after 5 min [68]. The cellulolytic bacterial isolate, B. subtilis IND19 was further grown on skimmed milk agar medium (in g/l) (agar, 15; yeast extract, 5; peptone, 5; KH 2 PO 4 , 1.0; MgSO 4 , 0.2; NaCl, 10; skimmed milk, 10, and pH 10.0). The maximum enzymeproducing bacterial isolate was selected for further studies.
The following conditions were employed while amplifying DNA: denaturation at 95 °C for 3 min followed by 30 cycles at 95 °C for 1 min, 55 °C for 30 s, and 72 °C for 1 min and 50 s. The amplified 16S rDNA PCR product was sequenced. Further, the identity of the sequences was checked by BLAST through the NCBI server. The 831 bp 16S rDNA sequences of the bacterial isolate were

Substrate
Cow dung was collected from a farm house (Nagercoil, Kanyakumari, Tamil Nadu, India). It was dried for 10 days and powdered. It was stored in an air tight container before further use.

Production of CMCase and protease under SSF
Cow dung substrate (5 g) was weighed in Erlenmeyer flask (250 ml) and a buffer solution (pH of 8.0, Tris-HCl buffer, 0.1 M) was added to maintain moisture content of the substrate and initial medium pH. Initial moisture content of the medium was maintained as 90 % (v/w). The solid substrate was mixed carefully with buffer and autoclaved at 15 lbs for 30 min and cooled. Then, 10 % inoculum (0.653 OD at 600 nm) was added to the culture medium. The contents were further mixed and incubated for 72 h under 37 °C.

Enzyme extraction
The fermented medium was stirred with double distilled water (1:10 ratio) and shaken at 175 rpm for 30 min in a rotary shaker. The mixed slurry was then completely filtered using cotton, followed by centrifugation at 10,000 rpm at 4 °C for 10 min. The cell free extract was used as the crude enzyme [49].

CMCase assay
CMCase activity was assayed using CMC as the substrate. Hundred microliter of crude enzyme was mixed with 100 μL of 1 % (w/v) CMC (pH 7.5) and incubated at 37 °C for 30 min. Next, 1.5 ml of dinitrosalicylate reagent was added, and the mixture was incubated at 100 °C for 10 min. The mixture was cooled, and the absorbance was measured against the reagent blank at 540 nm. One unit of CMCase activity was defined as the amount of enzyme that liberated 1 μmol of reducing sugars per minute under the above conditions [70].

Protease assay
Casein was used as the substrate for the determination of protease activity. The reaction mixture contained 1.0 ml casein which was prepared in Tris-HCl buffer (0.05 M, pH 8.0) and 0.1 ml of enzyme solution [71]. This mixture was incubated for 30 min at 37 °C and 2.5 ml trichloroacetic acid (0.11 M) was added to terminate the enzyme reaction. It was centrifuged at 10,000 g for 10 min, and the absorbance of the sample was read against sample blank at 280 nm. One unit of the protease activity was defined as 1 μg of tyrosine liberated min −1 under standard assay conditions.

Elucidation of significant factors affecting CMCase and protease production by statistical approach
Two-level full factorial design (2 5 ) was used to identify the significant factors relative to CMCase and protease yield. In this paper, two important physical factors and three nutritional factors were selected. These variables and the selected ranges were based on the results obtained from one-variable-at-a-time approach. The factors selected were sucrose (carbon source), peptone (nitrogen source), NaH 2 PO 4 (mineral), pH, and moisture (physical factors). Each variable was tested at two levels [high (+) and low (−1)]. In two-level full factorial design (2 5 ), a total of 32 experimental runs were generated and the enzyme activities (CMCase and protease) were determined from the crude sample. The variables and their levels are shown in Table 5. The other factors, namely, inoculum size and fermentation period, were kept at optimum level. Two-level full factorial design was based on the χ.
where α ij , α ijk , α ijkl , and α ijklm are the ijth, ijkth, ijklth, and ijklmth interaction coefficients, respectively, α i is the ith linear coefficient, and α 0 is an intercept. Assays of CMCase and protease were carried out in triplicates, and the mean value was taken as response (Y) ( Table 1). ANOVA was used to evaluate the significance of these models, and the p value <0.05 indicated that the model terms were significant. Statistical software Design-Expert 9.0.6.2 was used to design the experiments and analyse the results.

Central composite design and response surface methodology
The CCD was used to identify the optimum concentrations of the factors in order to obtain the maximum CMCase and protease production. The variables selected were analysed at five levels (−α, −1, 0, +1, +α) ( Table 6). According to the Design-Expert 9.0.6.2, for these variables CCD consists of 20 experimental runs including, eight factorial, six axial, and six centre points. Five gram of substrate was taken in 250-ml Erlenmeyer flask, and the required quantities of peptone and NaH 2 PO 4 were added according to the model. The pH of the medium was maintained according to the model design. The substrate and the supplemented nutrients were mixed carefully, sterilized (121 ± 1 °C for 20 min), and cooled. The Erlenmeyer flasks were inoculated with a 0.5-ml of inoculum (10 %, v/w) and incubated at 37 °C for 72 h. The enzyme was extracted as described previously in the materials and methods. After which, CMCase and protease assays were carried out individually in triplicate. The mean value of the experimental results was considered as response Y (Table 3). The fact that values of Prob(>F) are smaller than 0.05 would signify that the model terms were significant ( Table 4). The experimental results of the CCD were fitted with a following secondorder polynomial equation.
Response surface graphs were plotted to determine the optimum concentration of factors for the production of CMCase and protease. The fitted polynomial equation was expressed as 3D surface plots to visualize the relation between responses and the experimental levels of each factor used in the design. Validation of the model was performed under the conditions predicted by the model. The predicted response of the model was validated experimentally. Experiments were carried out in triplicates and validated.