Optimization of physicochemical and textural properties of pizza cheese fortified with soybean oil and carrot extract

Abstract Response surface methodology (RSM) was used to optimize pizza cheese containing carrot extract. The effects of two important independent variables including soybean oil (5%–20%) and carrot extract (5%–20%) were studied on physicochemical and textural properties of pizza cheese containing carrot extract. According to the results, RSM was successfully used for optimizing formulation of pizza cheese containing carrot juice. Results of this study revealed that oil (A), carrot (B), AB, square term of carrot (B2), B, AB, square term of oil (A2), B2, AB, AB, A2B, A2, A2, A, A2, A2, AB, and AB 2 had the most effect on moisture, acidity, stretch, L*, a*, b*, hardness, meltability, springiness, peroxide value (PV), cohesiveness, chewiness, gumminess, fracture force, adhesiveness force, stiffness, flavor, and overall acceptability, respectively. A formulation upon 20% oil and 10.88% carrot extract was found as the optimal formulation for pizza cheese containing carrot extract. At the optimal formulation, PV, L*, a*, b*, meltability, stretch, cohesiveness, springiness, gumminess, chewiness, adhesive force, flavor, texture, and overall acceptability at the optimum formulation were measured 2.23, 82.51, −3.69, 18.05, 17.86, 85.61, 0.41, 7.874, 23.7, 0.27, 0.61, 3.50, 3.95, and 3.65, respectively.

nique to optimize processes and formulation, which recently have been widespread used in the food industry to optimize different processes (Yolmeh & Jafari, 2017). RSM can decrease the number of experiments and properly shows the interactive effect of independent variables on responses, which are used to optimize cheese formulations (Farbod et al., 2014;Jooyandeh, Goudarzi, Rostamabadi, & Hojjati, 2017;and Khetra, Kanawjia, & Puri, 2016).
The lactic method is the most important production method to produce pizza cheese in the world. In addition, the citric and Cheese blend methods are used to produce this cheese. Uncertainty of production formula, lack of knowledge about the role and impact of the ingredients, low quality raw materials, production traditional method, optimization of the process, improper packaging, and lack of standard methods to examine the physical properties of the product are the difficulties of producing pizza cheese (Law & Tamime, 2011).
The objectives of this study were as follows: (1) development of pizza cheese containing carrot extract; (2) measurement of physicochemical and textural properties; (3) determination of an optimum formulation for this pizza cheese through RSM.

| Materials
Primary cheese was purchased from Chenaran dairy Industry Co.
Mashhad in the Khorasan-e-Razavi province. Soybean oil was procured from Behpak Co. (Behshahr, Iran). All chemicals and solvents used in this study were of analytical reagent grade and prepared by Merck (Germany) and Sigma-Aldrich (USA) Chemical Companies.

| Carrot extract
The extract was prepared from carrot, using Bush extracts (CNCJ03 model). The extract was then concentrated from 8.6 to 34.5 Bx at 20°C and 85 mmHg, using a rotary evaporator (Heidolph Co., Germany Laborota 4003 model).

| Preparation of pizza cheese
Pizza cheese was produced using the method described by Farkye, N. Y., & Yim, B. (2003) with some modification. Briefly, primary cheese was poured into the cooking pot, and then the temperature was increased up to 60°C to separate remaining whey. After separating whey, half of the other ingredients containing soybean oil, carrot extract, cream, salt, sodium nitrate, and sodium phosphate were added to the cheese at 80°C. After 15 min and good mixing the above ingredients with primary cheese, other ingredients were added to cheese.
After 15 min, the temperature was decreased to 75°C. The prepared pizza cheeses were packed into polypropylene bags and stored in a refrigerator at 4°C.

| Moisture
Pizza cheese samples were dried in the vacuum oven and moisture content was measured, using the method described by Daniela, Gustavo, and Barbosa (2012).

| Acidity
Acidity of pizza cheese was reported as oleic acid and measured by the following formula (AOAC, 1990): V: volume of used NaOH (mL); W: weight of the sample (g); N: normality of the used NaOH.

| Color analysis
The color properties of pizza cheese samples (L*, a*, and b* values) were evaluated, using the method developed by Zahedi and Mazaheri-Tehrani (2012).

| Peroxide value (PV)
The spectrophotometric method of the International Dairy Federation as described by Azizpour, Najafzadeh, Yolmeh, and Sangatash (2017) was used to determine PV. The PV of pizza cheese samples were calculated, using the following formula: where V, N, W represent volume of used sodium thiosulfate (ml), normality of sodium, and weight of the sample (g), respectively.

| Textural analysis
Melting, stretch-ability, and texture profile analyses were carried out using the method described by Zisu et al. (2007

| Sensory evaluation
The sensory properties of the pizza cheese samples, namely flavor, stiffness, and overall acceptability were evaluated by 20-member trained panelists (10 females and 10 males) took part in the descriptive analysis. The evaluation was done in a climate-controlled sensory evaluation laboratory. The panelists washed their palates between samples with water. The samples were served at room temperature (24 ± 1°C) and analyses were performed under normal lighting conditions on a 5-point hedonic scale (from dislike extremely = 1 to like extremely = 5) (Sun & Brosnan, 2003).

| Experimental design
A three level, two variable box-behnken design was employed to optimize with respect to two independent variables oil (%) and carrot (%). The independent variables and their levels are shown in Table 1.
Regression analysis was done on the data of dependent variables. In order to analyze the obtained result design, expert software version 8 was used.

| Statistical analysis
The multiple regression equation was employed to fit the secondorder polynomial equation based on the observed results as follows: where Y represents the predicted response; β k0 , β ki , β kii , and β kij represent regression coefficients; and x i , x j are the coded independent factors. The models were compared based on R 2 , R 2 -adj, and R 2 -pred. R 2 values closer to 1, indicate that the model is more accurate (Ghorbannezhad, Bay, Yolmeh, Yadollahi, & Moghadam, 2016). After selecting the most accurate model, the analysis of variance (ANOVA) was used to investigate the statistical significance of the regression coefficients by Dunkan's test at 95% confidence level. The interactive effects of the factors were studied, using surface plots derived from the selected model (Yolmeh & Sadeghi Mahoonak, 2016).
The aim of the optimizing formulation of pizza cheese was to maximize the meltability, stretch, L*, b*, cohesiveness, springiness, flavor, texture, overall acceptability (OA), and minimize gumminess, chewiness, Adhesive Force, and PV with the same weight (w = 1). The credibility of the optimum formulation was diagnosed by the desirability values of the responses that range from 0 to 1. The closer values of desirability to 1 showed the more credible and desirable optimal formulation.

| Fitting the response surface models
According to the used design, 13 experiments were performed thrice and the obtained results are shown in Table 2.
The values of R 2 , R 2 -adj, and R 2 -pred revealed that the linear model was more adequate than other models for moisture and fracture force values of pizza cheese samples; however for acidity, stretch, a*, PV, cohesiveness, flavor, and overall acceptability, 2FI model was suitable.
The quadratic model had more accuracy on L*, b*, hardness, chewiness, gumminess, adhesiveness force, and stiffness of pizza cheese containing carrot extract. However, for meltability and springiness, the cubic model was more adequate (Table 3). The selected models are as follows: +0.11AB Oil (%) X 1 5 12.5 20 Carrot (%) X 2 5 12.5 20 In addition, the lack-of-fit of the selected models were insignificant (p > .05), which shows a high suitability of the models to predict the dependent variables.
The analysis of variance (ANOVA) was used to appraise the significance of the quadratic polynomial models. For each terms in the models, a small p-value and a large F-value shows a more signifi-  carrot (AB) had the most effect on L*, springiness, PV, flavor of the samples. The interaction between the quadratic term oil and carrot, and the quadratic term of carrot (B 2 ) had the most effect on cohesiveness and L*, respectively (Table 4).

| Moisture
The moisture content of pizza cheese containing carrot extract was decreased by adding oil and the extract so that the lowest moisture content (41.09) was observed at 20% of oil and carrot extract ( Figure 1). Similarly, Ghanbari, Khosroshahi, Mortazavi, and Tavakolipour (2012) reported the same findings for Iranian low-fat white cheese. Romeih, Michaelidou, Biliaderis, and Zerfiridis (2002) reported that the water holding capacity (WHC) of the matrix of casein was increased by decreasing oil, which leads to an increase in moisture content.

| Acidity
The acidity of pizza cheese samples was decreased by increasing oil, especially at low levels of carrot. On the other hand, the acidity of pizza cheese samples was significantly decreased by increasing carrot at low levels of oil. However, at high levels of oil, the acidity was gently increased (Figure 2). This is in agreement with findings of Katsiari et al. (2000), Katsiari et al. (2002), and Shahab- Lavasani et al. (2012) about Feta, low-fat Kefalograviera and UF white cheeses, respectively. Guinee, Feeney, Auty, and Fox (2002) reported that this increasing acidity could be attributed to microbial growth.
T A B L E 3 The statistics of the four fitted models  .023 Oil (

| Peroxide value
The PV of pizza cheese samples was increased by increasing the carrot extract content but subsequently decreased. At high contents of carrot extract, the PV was decreased by decreasing oil content ( Figure 3). As is shown in Figure 1c, the lowest PV was observed at 5% oil and 20% carrot extract. Mortensen, Sørensen, and Stapelfeldt (2002) observed that PV of semihard cheeses was increased by increasing the oil content. This result is in agreement with observations of Christensen, Povlsen, and Sørensen (2003) that evaluated processed cheese, using fluorescence spectroscopy and chemometrics during storage.

| Color properties
Increasing oil content up to about 10% reduced L* value of pizza cheese samples, but at a higher oil content, the L* value significantly increased. The L* value was gently increased by adding carrot extract; however, at a higher content of carrot extract, the L* value significantly decreased, which is due to high concentrations of carotenoids Here, the a*value of pizza cheese containing carrot extract was increased by increasing oil and carrot extract so that the lowest degree of redness was observed at 5% oil and carrot extract content ( Figure 5).
Here, the b* value of pizza cheese samples was increased by increasing the carrot extract content. On the other hand, at low carrot extract content, the b* value was increased by increasing oil content; however, the opposite is true when increasing oil content at high carrot extract content ( Figure 6). The highest b* value was observed at 12.5% oil and 20% carrot extract. Similarly, Ghanbari et al. (2012) reported that b* value of cheese was increased by increasing oil content.

| Textural properties
Cohesiveness At low content of carrot extract, cohesiveness of pizza cheese samples was initially increased by adding oil up to 9%, but subsequently reduced. On the other hand, the cohesiveness was increased by increasing the carrot extract. The highest cohesiveness pizza cheese containing carrot extract was observed at 5% of oil and carrot extract content (Figure 7). Rashidi, Mazaheri Tehrani, Razavi, and Ghods Rohany (2011) reported that cohesiveness of UF-Feta Cheese was F I G U R E 4 The interactive effect of soybean oil and carrot extract on L* value of pizza cheese F I G U R E 3 The interactive effect of soybean oil and carrot extract on peroxide value of pizza cheese increased by increasing oil content. Zisu and Shah (2005) demonstrated that cheeses containing a high moisture content (low oil content) have weak internal linkages resulting in softer texture. Beigomi et al. (2013) reported that cohesiveness of cheese based on fungal rennet was increased during storage time; however, cohesiveness of cheese based on plant rennet was increased by the 20th day but subsequently decreased.
Gumminess Figure 8 shows the interactive effect between oil and carrot extract on gumminess of pizza cheese containing carrot extract. The gumminess was initially increased by adding oil and carrot extract up to about 16% and 12.5%, respectively; but it decreased at higher content. So, the lowest gumminess was observed at about 20% oil and 20% carrot extract content. The result of this study is in agreement with the findings of Koca and Metin (2004) for low-fat Kashar cheese.
They reported that gumminess low-fat fresh kashar cheese was increased by decreasing the oil content. In addition to oil content, moisture content, and protein to moisture ratio has an effect on textural properties of cheeses (Koca & Metin, 2004). Chewiness The chewiness of pizza cheese containing carrot extract was increased by increasing the oil and carrot extract content. The lowest chewiness of pizza cheese samples was observed at 5% oil and carrot extract (Figure 9). Similarly, Koca and Metin (2004) reported that chewiness of low-fat Kashar cheese was reduced by decreasing the oil content. Adhesiveness The adhesiveness of pizza cheese containing carrot extract was increased by increasing the carrot extract. The adhesiveness was initially increased by adding oil up to 12.5%, but subsequently reduced.
The lowest adhesiveness of pizza cheese samples was observed at 5% oil and carrot extract content ( Figure 10). Dimitreli and Thomareis (2007) reported that high oil content makes the protein matrix weak, resulting in increased adhesiveness.

Springiness
According to Figure 11, the springiness of pizza cheese samples was increased by increasing the carrot extract and decreasing the oil content.
The highest springiness of pizza cheese samples was observed at 5% oil and 20% carrot extract ( Figure 11). The result of this study is in agreement with the findings of Zisu and Shah (2005).
Stretch Figure 12 shows the interactive effect between oil and carrot extract on a stretch of pizza cheese containing carrot extract. The stretch was increased by increasing the carrot extract and decreasing the oil content.
The highest stretch of pizza cheese samples was observed at 5% oil and 20% carrot extract ( Figure 12). Mizuno, R., & Lucey, J. A. (2005) reported that stretch and melt- ability of nonfat pasta filata cheese was increased by adding trisodium citrate. Fox, Guinee, Cogan, and McSweeney (2000) demonstrated that the stretch of cheese is dependent to the casein-associated calcium content so that too low or too much of calcium reduces the stretch.

Meltability
According to Figure 13, meltability of pizza cheese samples was increased by decreasing oil content. On the other hand, the meltability F I G U R E 7 The interactive effect of soybean oil and carrot extract on cohesiveness of pizza cheese F I G U R E 8 The interactive effect of soybean oil and carrot extract on gumminess of pizza cheese was initially increased by adding carrot extract up to 12.5%, but subsequently decreased.
The highest meltability of pizza cheese containing carrot extract was observed at 5% oil and 12.5% carrot extract ( Figure 13). The result of this study is in agreement with findings of Zalazar et al. (2002), and Hassan and Abd-El-Gawad (2000) for mozzarella cheeses. Shirashoji et al. (2006) reported that meltability of processed cheese was decreased by increasing concentration of trisodium citrate.

| Sensorial properties
Flavor Figure 14 shows the interactive effect between oil and carrot extract on flavor desirability of pizza cheese samples. The flavor desirability was initially increased by adding carrot extract up to 12.5%, but subsequently reduced. The flavor desirability was increased by decreasing oil content. The highest flavor desirability of pizza cheese containing carrot extract was observed at 5% oil and 12.5% carrot extract ( Figure 14). Taghvaie, Taslimi, and Mazloumi (2006) reported that partial replacement of milk fat with sunflower oil improved the flavor of cheese. Yu and Hammond (2000) observed the same results for Swiss cheese.
Texture Texture desirability of pizza cheese containing carrot extract was increased by decreasing carrot extract content and increasing the oil content. The highest textural desirability was observed at 5% oil and 20% carrot extract (Figure 15). Similarly, Sipahioglu, Alvarez, and Solano-Lopez (1999) reported that textural desirability of feta cheese was decreased by increasing fat. tract on OA of pizza cheese samples. At low carrot extract content, the OA was increased by decreasing oil content; however, the opposite is true when decreasing oil content at high carrot extract content.
The OA was initially increased by adding carrot extract up to 12.5%, but subsequently reduced. The highest OA was observed at 5% oil and 12.5% carrot extract (Figure 16).

| Optimization
The numerical optimization technique performed to optimize the formulation, when weight and importance value for all of the responses were considered equal . The PV, L*, a*, b*, meltability, stretch, cohesiveness, springiness, gumminess, chewiness, adhesive force, flavor, texture, and OA attributes were considered for the optimization formulation of pizza cheese. The formulation upon 20% oil and 10.88% carrot extract was found as the optimal formulation for pizza cheese containing carrot extract.

| CONCLUSIONS
RSM was successfully used for optimizing formulation of pizza cheese containing carrot juice. Results of this study revealed that the linear model was more adequate than other models for moisture and frac- ture force values of pizza cheese samples; however for acidity, stretch, a*, PV, cohesiveness, flavor, and overall acceptability, the 2FI model was suitable. The quadratic model had more accuracy on L*, b*, hardness, chewiness, gumminess, adhesiveness force, and stiffness of pizza. However, the cubic model was more adequate for meltability and springiness. A formulation upon 20% oil and 10.88% carrot extract was found as the optimal formulation for pizza cheese containing carrot extract. At the optimal formulation, PV, L*, a*, b*, meltability, stretch, cohesiveness, springiness, gumminess, chewiness, adhesive force, flavor, texture, and OA at the optimum formulation were measured 2. 23, 82.51, −3.69, 18.05, 17.86, 85.61, 0.41, 7.874, 23.7, 0.27, 0.61, 3.50, 3.95, and 3.65, respectively.