2.1. Preliminary experiments to obtain enzyme with different culture media

Fermentations were conducted with different culture media to determine the best composition of components and increased production of MTGase, A Box-Behnken experimental model was performed for culture media with three factors and three levels of 72 and 96 h of fermentation. Three different concentrations of casein, yeast extract, and DDGS were tested. Concentrations of glucose 30 g/L, MgSO₄ 0.5 g/L, KH₂PO₄ g/L, Na₂HPO₄ g/L, and peptone 10.5 g/L were kept constant in all media.

To measure MTGase activity, the Grossowicz method was used ¹⁶. The procedure for measuring MTGase enzyme activity was as follows: 5 mL of culture medium was taken at different inoculation times (72, 96 and 120 h), and these were centrifuged at 7000 rpm/10 min. An aliquot of 0.2 mL of supernatant was taken from each tube; one tube was placed in a water bath at 75°C to inactivate the enzyme and this was the blank. Subsequently, we added 0.5 mL of the solution of the specific substrate N-CBZ-GLN-GLY (Sigma-C6154, Sigma Aldrich Co., St. Louis, MO) to the tubes with the sample and blank and incubated these at 37°C for 10 min. At the end of this time, the reaction was quenched with 0.2 mL of 5% FeCl3 solution in 0.1 N HCl and the absorbance was measured at 525 nm. Calculations were made to obtain the MTGase units using a calibration curve of monohydroxamate. For this, commercial MTGase (Ajinomoto Co. Inc.) was used as a control. In this method, a unit of enzyme is defined as the formation of 1 mol of hydroxamate per minute at 37°C.

The experimental design followed an incomplete and second-order factor structure. The yeast extract, DDGS, and casein were considered independent variables and their effects on the dependent variables (TGase activity at 72 and 96 h) were calculated. The interaction between medium composition variables and dependent variables can be set through a linear equation including linear, interaction, and second-order terms. In the experimental design, encoded factors were named X1 (casein), X2 (yeast extract), and X3 (DDGS).

2.2. Process simulation

The design and programming of the production plant of microbial transglutaminase from enzymatic sorghum grains hydrolysates was performed with SuperPro Designer v. 7.5 (Intelligen Inc., Scotch Plains, NJ). The general process for obtaining enzyme using enzymatic sorghum flour hydrolysates is as follows: the raw material is treated by washing with running water, then it is dried and the sorghum grains are ground to obtain flour. The bran is separated from the flour and afterwards enzymatic hydrolysis is performed with α-amylase and glucoamylase. After hydrolysis, the glucose juice is separated from the solid waste; glucose is sterilized by autoclaving (121°C for 15 minutes). In another tank, the remaining components of the medium (peptone, yeast extract, casein, MgSO₄, K₂HPO₄, NaH₂PO₄) are mixed. After mixing, these compounds were also sterilized to prevent the Maillard reaction whose products may inhibit bacterial growth. Once sterilized, both parts of the medium are mixed in a fermenter and fermented for 96 hours; the crude extract with the enzyme is separated from the solid residues by centrifugation. After MTGase is separated, it is lyophilized and subsequently packaged (Figure 1 and Table 1). In addition to the simulation of this process, an enzyme production plant was designed with the common medium (Figure 2 and Table 1), which has glycerol as a carbon source; the composition of the medium was glycerol 32 g/L, MgSO4 0.5 g/L, KH₂PO₄ g/L, Na₂HPO₄ g/L, peptone 10.5 g/L, yeast extract 2.5 g/L and casein 38.4 g/L.

Figure 2 Flowsheet for MTGase production with common media 

2.3. Economic Analysis

SuperPro Designer ® Version 7.5 was used for the economic analysis of the process design (Figure 2). The cost of equipment was obtained from the SuperPro-Designer database or at the website http://www.kitmondo.com. The total equipment purchase cost was 4.413 x10⁶ US$ and 3.901 x 10⁶ US$ for plant with hydrolysates and common media, respectively. This was the basis for calculating the cost of other components of the fixed capital investment such as installation, instrumentation, etc. These costs are show in Table 2.

3.1. Preliminary laboratory experiments

The culture media results were analyzed with the statistical program Design expert v7 (Stat-Ease, Inc., Minneapolis, MN). Analysis of variance showed that the casein-yeast extract interaction had a significant effect (p ≤ 0.0275) in the production of the enzyme, and casein was the nutrient with the greatest effect (p ≤ 0.0003) since media without casein had an MTGase activity close to zero. The program also made ​​a prediction based on the experimental values; because DDGS does not influence the model (p ≤ 0.4275), the lowest concentration (2 g/L) can be used. The model predicts that the optimal concentrations are DDGS 2 g/L, yeast extract 2 g/L, and casein 15 g/L. With these conditions, a casein concentration of 0.67 U/mL was obtained.

The experimental data showed an activity of 0.62 U/mL (Table 2), which is consistent with the prediction ¹⁴; therefore, these nutrient concentrations were chosen for the plant design in the industrial simulator. It is important to point out that it is not a replica of the MTGase production plant but a design based on the MTGase production process from sorghum grains. For comparison and validation of the feasibility of production of the enzyme with enzymatic hydrolysate media of sorghum, an MTGase production plant with a common culture media, which has glycerol as a carbon source, was designed in the same simulator.

The laboratory experiments that formed the basis for the design of enzyme production plants yielded an enzyme activity of 0.62 U/mL for the enzyme produced with media based on sorghum hydrolysates (Table 3), while with glycerol a maximal activity of 0.44 U/mL was obtained. However, the commercial enzyme from Ajinomoto Co., Inc. (Tokyo, Japan) has an average activity of 1 U/mL, thus this activity was considered for the enzyme produced in medium with glycerol to compare production with an established company. Ajinomoto Co., Inc. achieves this activity using a genetically modified E. coli and not a wild strain of S. mobaraensis.

Table 1 Key inputs and outputs design MTGase production plant with both media based on enzymatic hydrolysates 

3.2. Process Design and economical evaluation

Plant design was carried out following the above production process, considering from the washing of raw material to the packing of the lyophilized enzyme. The enzyme marketed by Ajinomoto Co. Inc. contains 99% maltodextrin and 1% MTGase with an approximate activity of 100 U/g and the cost of this product is approximately 70 US$/kg. The final product of the MTGase production plant contains about 62 U/g at a cost of 35 US$/kg. For comparative purposes and to truly establish the profitability of the MTGase production plant using glucose as a carbon source obtained from sorghum grains, it was necessary to simulate an enzyme producing plant using the common medium, which has glycerol as the carbon source and a higher concentration of nitrogen sources than the medium supplemented with hydrolysates. Equipment for obtaining sorghum flour was not necessary from this plant. Table 2 to 4 show name, code, inputs and outputs for plants.

The total capital investment was 27.103 x 10⁶ US$ with an equivalent production of 320,158 units/year. Meanwhile, the unit product cost was 16.06 US$/entity with a total revenue of 12.326 x 10⁶ US$ in addition to obtaining 4,347 US$ /year from sales of subproducts (ground sorghum residue, solid residue from hydrolysis, and solid residue from fermentation), for a total annual income of 12.330 x 10⁶ US$. The byproducts from enzyme production contain a high protein and fiber content, which may be used as cattle feed supplement.

The sum of capital investment is analyzed formerly (Table 2) showing the total plant direct costs (TPDC) and indirect costs (TPIC). The TPC is estimated at 22.428 x 10⁶ US$; adding contractor fees and contingency costs results in a direct cost of fixed capital of 25.792 x 10⁶ US$.

The cost of sorghum was consulted at the Secretariat of Agriculture, Livestock, Rural Development, Fisheries and Food (SAGARPA) ¹⁷. The cost of the other reagents from Sigma Aldrich was quoted in dollars. The estimated annual cost of the raw materials was 134,000 US$, and in turn the annual operating cost was 5.140 x 10⁶ US$ (Table 5). This total includes the cost of raw materials, cost of operators, lab fees, waste management, utilities, and the pending issues fund; in this, consumables, transportation, sales strategies expenses, maintenance, and others are considered (Table 4). This plant shows an annual income of 12.36 x 10⁶ US$; of this 12.3 x 10⁶ US$ correspond to MTGase sale and 4,000 US$ to subproduct sales (0,20 US$ /kg). There is a profit margin of 58.31% and a percentage of return on investment of 24.96%. According to all these data, the return on investment time is 4.01 years.

The same process was conducted but with commonly used medium ingredients ⁸. Although this reduces the number of equipment because enzymatic hydrolysis is not required, the total investment does not significantly decrease (Table 4). We estimated savings of 12% in comparison with the equipment estimated in the prior plant. Furthermore, the concentrations of some components of the medium, such as casein, are duplicated, which increases production cost. By comparing the costs of raw materials in both production plants, we have an increase of 244% due to the high cost of carbon and nitrogen sources in the common medium, compared with the medium based on sorghum hydrolysates with DDGS. In the plant with sorghum hydrolysates, the annual cost of raw material is estimated at 134,000 US$, while the other plant's estimated cost is 327,000 US$.

Table 2 Fixed capital estimate summary for the MTGase production (price in US$) 

Table 3 Experimental design with three factors and three levels 

The plant design with common medium generated in the program is displayed in Figure 2. The total equipment purchase cost was 3.901 x 10⁶ US$. As in the previous design, the purchase cost of the equipment was the basis for calculating the cost of other components of the fixed capital investment, such as installation, instrumentation, etc. These costs are shown in Table 2.

The prices of equipment and raw materials were consulted in the same sources as in the previous plant. The same number of operators for the process and the same number of hours and wages were used (Table 5). This plant shows an annual income of 10.836 x 10⁶ US$ of which 10.774 x 10⁶ US$ correspond to MTGase marketing and the remainder to the sale of fermentation by products (0.20 US$/kg). A profit margin of 48.91% and a return on investment of 22.29% were defined. According to all the data, it was determined that the return on investment time is 4.49 years as shown in Table 6.

Table 4 Operating cost summary MTGase production plants (Price in US$) 

Table 5 Important input and output economic parameters used in the SuperPro Designer 

Table 6 Economic comparison of two plants 

In the MTGase production plant based on sorghum hydrolysates, 4,916 kg of enzyme are obtained per year which yields 3,680 units/batch, while the enzyme produced with the common medium produces 4,550 units per batch with a production of 8,505 kg of MTGase per year. Although less equipment is used in the plant that produces the enzyme with glycerol as a carbon source, the cost does not decrease significantly due to the amount of some raw materials such as casein. In the medium with hydrolysates a casein concentration of 19 g/L is used, while in the medium with glycerol 38 g/L is used.

The return on investment in both plants is almost similar, as shown in Table 6, with a difference of 0.5 years less in the hydrolysates based production plant; moreover, the cost of MTGase produced by hydrolysates is less than the enzyme produced with the common medium, therefore it would have a higher demand. It is important to point out that it was assumed that an activity of 1 U/mL would be obtained using the common medium, although fermentations using this medium resulted in a maximum activity of 0.4 U/mL. The value of 1.0 U/mL was used because this is how enzyme activity is currently marketed by Ajinomoto Co. Inc., although the strain used for the marketed enzyme is genetically modified to increase MTGase activity. Even though genetically modified strains increase production and have been shown to be safe in the production of food additives, it is important to have a thorough purification process to avoid the presence of toxic substances that may cause harm to health; therefore, the use of totally harmless wild strains such as S. mobaraensis is preferable for obtaining food additives. The plant that used the hydrolyzed enzyme as a glucose source, reported a higher percentage of profits than the plant in which glycerol was used (58.3% and 48.9%, respectively). For the enzyme produced with enzymatic hydrolysates a cost of 35 US$/kg was obtained; while the enzyme produced with the common medium was 70 US$/kg, which is its current market price. These results confirmed the feasibility of MTGase production using glucose as a carbon source obtained from enzymatic hydrolysis. A culture medium with 600g/L skim milk, 40g/L potato and 5g/L glycerol was evaluated; the maximum activity obtained was 2.95±0.30U/mL at 72h; the economic yield was 9.03 US$ (8.11€) of transglutaminase obtained per US$ spent in nutrients ¹⁸. On the other hand, the production of transglutaminase from enzymatic hydrolysates of sorghum was evaluated, the optimal conditions selected were 10 g/L of yeast extract, 2 g/L of corn steep liquor and 13.7 g/L of sodium caseinate, yielding a maximum activity of 1.128 U/mL and a transglutaminase economic yield of 4.144 US$ (3.72 €) per US$ spent in nutrients ¹⁹. However, the red sorghum is an economical cereal and an alternative to use as carbon source in culture media.