Enhancement of low‐fat Feta cheese characteristics using probiotic bacteria

Abstract The objective of this study was to manufacture low‐fat Feta cheese (LFC) using different types of starter cultures, such as yogurt (Y) cultures (Streptococcus thermophilus and Lactobacillus bulgaricus), bifidobacterium (B) cultures (Bifidobacterium bifidum and Bifidobacterium longum), and mixed of them (Y + B) at different rates (0.4, 0.5, and 0.6%). The Y + B cultures improved the flavor and body and texture of LFC, especially at a ratio of 0.4 + 0.6% and 0.5 + 0.5%, which is similar to the typical full‐fat Feta cheese. Also, the LFC maintained a higher number of probiotics and lactic acid bacteria after 30 d of storage at a range of 5 to 7 log cfu/g.


| INTRODUC TI ON
Feta cheese is a brined soft cheese and it is widely popular in many countries in Africa, Europe, and other regions. It was manufactured from goat milk, but nowadays different types of milk, including sheep, cow, and buffalo milk are utilized to produce this type of cheese.
Traditionally, Feta cheese was made from raw milk to decrease the ripening period, as well as, having the intense flavor that results from the natural microflora. However, milk is pasteurized nowadays before being used in making Feta cheese to assure the safety of consumers and to maintain the typical characteristics of cheese. Manufacturing of Feta cheese from pasteurized milk has various challenges due to the lack of flavor intensity as compared to Feta cheese made from raw milk (Bintsis & Robinson, 2004).
Consequently, several studies have focused on producing Feta cheese from pasteurized milk using different types of starter cultures, additives, various sources of milk and different processing conditions to stimulate the flavor and texture of the raw Feta cheese.
Feta cheese is typically manufactured using lactic acid bacteria (LAB). The count of LAB in Feta cheese is increasing with elevating the acidity at the beginning of the ripening at 5-7°C, and then the count of LAB becomes constant up to 60 d (Bintsis & Robinson, 2004). On the other hand, the count of mesophilic starter cultures is decreasing at the prematuration of Feta cheese, especially at a higher salt content (6%-8%) and pH of < 5.0. As a result, thermophilic and probiotic bacteria have been used as adjunct starter cultures to enhance the flavor of Feta cheese.
The microflora in cheese produces the typical flavor by their metabolic activities, such as proteolysis, lipolysis, and lactose fermentation.
Several probiotic cultures, such as bifidobacterium and streptococci (Peirotén, Gaya, Arqués, Medina, & Rodríguez, 2019), have been utilized recently in the manufacture of cheese due to their ability to produce the desirable flavors (Champagne, Gagnon, St-Gelais, & Vuillemard, 2009). Furthermore, probiotic bacteria can provide health benefits to the consumers at the rate of 5-7 log cfu/g present in the product (Roy, 2001).
Full-fat Feta cheese (FFC) can result in several health issues, such as high blood pressure, cardiovascular diseases, increased cholesterol levels, and obesity due to the high fat content.
The lowfat Feta cheese (LFC) is healthy and beneficial for dieters and heart patients, especially when it is manufactured with probiotics that have several promising health benefits. Although the production of LFC has many challenges due to the lack of fat which in turn has a significant effect on the flavor and texture characteristics of cheese, the sales and market growth of LFC have increased recently due to the negative effect of FFC on health (Michaelidou, Katsiari, Kondyli, Voutsinas, & Alichanidis, 2003). As a result, many researchers have been working on manufacturing of LFC by mimicking its flavor and texture characteristics to that of the FFC (Bintsis & Robinson, 2004). Several modifications have been applied to the traditional process of making LFC to compensate for the lack of fat, such as using different starter cultures and increas- ing the moisture content of cheese to make the texture smoother.

| Manufacturing of low-fat Feta cheese with probiotics
The LFC was manufactured from a mix of cow and buffalo skim milk. Fresh milk was obtained from the Animal Production Farm (Faculty of Agriculture, Assiut University, Assiut, Egypt) and was separated at 20°C. Cow skim milk and buffalo skim milk were mixed at a ratio of 1:1, heated at 73°C/16 s, and then cooled to 40°C. The skim milk was divided into sixteen lots, and then 3% of sodium chloride was added to each part. Then, the starter cultures were added for each lot as shown in Table 1  The inoculated skim milk was left for 30 min at 40°C until complete coagulation. Then, the curd was cut, packed in cheesecloth, and left for draining overnight at 5°C. The cheese was removed from the cheesecloth on the following day, cut into cubes, and pickled in sterilized glass containers contained whey that produced during making the cheese. Cheeses from different treatments were stored at 6 ± 2°C and analyzed when fresh (d = 0), and after 7, 15, 21, and 30 d. This experiment was repeated three times.

| Microbiological analyses
One gram of the cheese samples was weighed under an aseptic environment and transferred into a sterilized jar. Subsequently, 9 ml of a sterile phosphate buffer was added and evenly mixed to have the 1:10 dilution, which was further used to prepare the sequence of dilutions (Mehta, Metzger, Hassan, Nelson, & Patel, 2019). Total bacterial count (TBC) was enumerated by using the standard plate count technique (Wehr & Frank, 2004 (Wehr & Frank, 2004). The plates were anaerobically incubated at 40°C for 48 hr. Bifidobacterium count (BC) was enumerated using a fluid thioglycollate medium using a pouring plat method (Brewer, 1940). The plates were anaerobically incubated at 37°C for 48 hr. The coliform count was enumerated on MacConkey broth media, and tubes were incubated at 32°C ± 1°C for 24 hr (Ashenafi, 1990). Yeast and mold count were also enumerated (Wehr & Frank, 2004) using potato dextrose agar media, and plates were incubated at 25°C ± 1°C for 5 d. The microbiological analyses were performed at 0, 7, 15, 21, and 30 d.

| Sensory evaluation
The sensory characteristics of LFC samples were evaluated according to 10-15 trained panelists from the Dairy Science Department, Assiut University. The LFC was examined as described by Mehta with some modifications (Mehta, Kumar, & Sabikhi, 2017). Samples were evaluated for color and appearance (15 points), flavor (50 points), and body and texture (35 points) to have 100 points overall.
The organoleptic characteristics were evaluated at 0, 15, and 30 d of storage.

| Statistical analysis
Results were analyzed by R software (R x64-3.3.3, 9,205 NW 101st St, Miami, Florida, United States). All data were analyzed by ANOVA using a GLM for each variable to study the effect of probiotic bacteria and time on the characteristics of LFC. Mean separation was done using the least significant difference (LSD) comparison test when significant differences were detected at p < .05.  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14  T15 T2, and T3) had higher TBC as compared to other groups. By the end of ripening, the highest TBC was 7.87 log cfu/g in control while the lowest was 4.81 log cfu/g for T6 as compared to other treatments.

| Microbiological analyses
The TBC decreased with the increase of starter cultures, and this could be caused by acidity development (Ahmed et al., 2020), which led to the inhibition of bacteria in LFC (Charlier et al., 2008).
The TBC in group B cheese (T4, T5, and T6) was the lowest as compared to other groups, and this can be related to the medium which is not suitable as a selective media for bifidobacterium or the poor growth of bifidobacteria (Peirotén et al., 2019). On the other hand, mixing groups Y and B did not markedly increase (p > .05) the TBC. This may be due to the high acidity levels in this group during the storage period which limits the growth of microorganisms (Ahmed et al., 2020).
These results were in agreement with other studies that reported the TBC in soft cheese made with probiotic increased up to 15 d of storage, and then, the number decreased until the end of storage (Hassanien, Mahgoub, & El-Zahar, 2014 (Gomes & Malcata, 1998). The decline in BC at the end of the storage time could be resulted from the low resistance of bifidobacteria with increasing the acidity values after 30 d, as we reported in a previous study (Ahmed et al., 2020). However, the BC ranged from approximately 5 to 7 log cfu/g, which is required to consider the LFC as a probiotic product.  Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14 Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14 Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14 Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14  T15 Yeast and mold count (log cfu/g) of probiotic LFC is presented in Figure 5. The yeast and mold count in control was detected after cheeses. This might be caused by the commitment to hygienic prac-

F I G U R E 2
tices during the manufacturing of cheese. The number of yeast and mold initiated to increase by extending the storage period as a result of acidity development (Makhal, Kanawjia, & Giri, 2015). Generally, these microorganisms are proliferated in the manufacturing environment, which can allow them to reach the cheese samples during manufacture (Vinderola, Prosello, Ghiberto, & Reinheimer, 2000).
Tests were carried out to estimate the coliform group in all LFC treatments. The results were negative in all those tests for all treatments during the 30 d of storage period. This is due to the pasteurization of milk before the manufacturing of cheese, which eliminates coliform bacteria. Moreover, the hygienic practices during the experimental procedure of probiotic LFC for all treatments inhibit the growth of coliforms. These results are in agreement with the finding of several investigators (Ordóñez, Ibáñez, Torre, & Barcina, 1999).

| Color and appearance
The color and appearance scores of LFC with different starter cultures are presented in Figure 6. The color and appearance of all cheeses did not markedly affect (p > .05) by adding starter cultures since the cheese was kept at 4°C. These results were in agreement with other studies, which reported that using Bifidobacterium bifidum in Cheddar cheese manufacturing did not increase the metabolic action and did not affect the cheese color during the 24 weeks of ripening (Dinakar & Mistry, 1994).

| Body and texture
The body and texture scores of LFC with different starter cultures are presented in Figure 7. The starter cultures had a slight effect on the body and texture of LFC which gained higher scores during the ripening time. During storage, the hard casein matrix turned to soft texture because of the slow proteolysis under the action of the endogenous enzymes, rennet enzymes, and also from the added starters proteolytic enzymes. The LFC made with mixed groups (Y + B) gained the highest body and texture scores as compared to other groups (Figure 7), which may be due to the low moisture content in that group (Ahmed et al., 2020). The body and texture of group B were comparatively softer than group Y and group Y + B. This could be due to the low acidity levels in group B that led to retain the high moisture content during the 30 d of storage (Ahmed et al., 2020).
It has been reported that using mixed cultures of Streptococcus thermophilus and Bifidobacterium improved the body and texture of Damiata cheese (Feta cheese type) during the storage period (Boylston, Vinderola, Ghoddusi, & Reinheimer, 2004).

| Flavor
The  Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14 Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14 Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14 Control  T1  T2  T3  T4  T5  T6  T7  T8  T9  T10  T11  T12  T13  T14  T15 reported that the flavor of Ras cheese made using Streptococcus thermophilus and Lactobacillus bulgaricus improved during the ripening period. Group B cheeses had gained the lowest flavor and aroma scores since bifidobacteria do not produce flavor components but acetate (Ong, Henriksson, & Shah, 2007). It is also known that bifidobacteria are poorly grown in milk and its products (Peirotén et al., 2019). The panelists tasted the acetic acid flavor, which proportionally increased as the percentage of bifidobacteria starter cultures elevated (Ong et al., 2007). The panelists did not taste the acetic acid flavor in the cheese produced from a mix of Y + B which resulted in cheese with a higher flavor score as compared to group B that had lowest flavor scores. It has been reported that the flavor of Festivo low-fat cheese improved using bifidobacteria and Streptococcus thermophilus (Ryhänen, Pihlanto-Leppälä, & Pahkala, 2001). It was expected that rennet and indigenous enzymes had an important role to produce the flavor in LFC during cheese ripening (Pereira, Gomes, Gomes, & Malcata, 2008). The overall acceptability of LFC is shown in Figure 9. The highest scores (92/100) were found in T11 after 30 d of storage, followed by T9 (90/100), T12 (88/100), and lastly for the rest of cheeses produced with group Y + B starter cultures.

| CON CLUS ION
The LFC was manufactured using Y, B, and Y + B cultures at rates of 0.4, 0.5, and 0.6%, respectively. We concluded that a mix of Y + B cultures improved the flavor and texture of LFC, especially at 0.4 + 0.6% and 0.5 + 0.5% proportions which can be used to produce LFC with a flavor profile similar to FFC. Additionally, the Y + B starter cultures resulted in LFC with higher lactic acid bacteria up to 30 d of storage. Probiotic bacteria ranged from 5 to 7 log cfu/g in the LFC, so it can provide potential health benefits to consumers. This study indicates that the LFC can be produced with a typical FFC flavor profile by selecting an appropriate ratio of Y + B starter cultures.

ACK N OWLED G M ENTS
The authors would like to thank the dairy plant of Assiut University at the Faculty of Agriculture for providing the essential materials and equipment. Authors are also grateful to Dr. Fathy El-Gazzar and Dr.
Ibrahim Abo El-Naga for their time and help during conducting this experiment.

CO N FLI C T O F I NTE R E S T
The authors have no conflict of interest to declare.