Microbiological Characteristics and Behavior of Staphylococcus aureus , Salmonella spp., Listeria monocytogenes and Staphylococcal Toxin during Making and Maturing Cotija Cheese

: Cotija cheese is an artisanal matured Mexican cheese from unpasteurized milk. This work determined the microbiological characteristics and behavior of Staphylococcus aureus , Salmonella spp., Listeria monocytogenes and staphylococcal toxin during cheese elaboration and ripening. Sixteen 20-kg cheeses were used, eight inoculated with 6 log CFU/mL of each pathogen, and eight uninoculated, and samples were taken at each stage of the process. In the uninoculated samples, the survival of S. aureus and L. monocytogenes decreased after 30 days of ripening. The average counts of S. aureus in milk, curd, and serum were 7 log MPN /mL, and 8.7 log MPN /g in cheese, decreasing from day 15. Salmonella spp. counts (initial inoculum: 7.2 log MPN /mL) decreased after 24 h, and L. monocytogenes counts (8.7 log MPN/g at 24 h) decreased from day 15 in the cheese. Salmonella spp. and L. monocytogenes were not detected in any sample after 60 days of ripening, unlike S. aureus , which was detected at the end of the study. Lactic acid bacteria counts were 9 log CFU/mL in milk and whey and 7.2 log CFU/g in cheese. Pathogens behavior during the ripening process reduces the health risks by consuming products made with unpasteurized milk when subjected to ripening.


Introduction
The prevalence of pathogens in milk is influenced by numerous factors, such as the size of the farm, the number of animals, hygiene, variability in the sampling, and the techniques used for sample analysis, geographic location, and the season of the year. However, within the variability that these factors may have, milk is one of the most important sources of pathogens that cause foodborne diseases [1]. In Mexico, most fresh cheeses are handmade using rustic methods in micro-industries located in small cities. These industries generally have little or no quality control and high composition and sensory parameters variability, leading to a limited shelf life [2][3][4]. However, artisanal matured cheeses show a different reality due to the fermentative processes during the maturation time, giving rise to the

Milk, Curd, and Whey Analysis
Before the beginning of the cheese production, the natural presence of Staphylococcus aureus, Salmonella spp., Listeria monocytogenes and staphylococcal toxin was analyzed, as well as lactic acid bacteria (LAB), aerobic mesophilic bacteria (AMB), coliform organisms (CO), molds and yeasts (M/Y) indicators in milk, uninoculated curd and whey. Only S. aureus and Listeria were recovered in milk, curd, and whey (Table 1), and two species were identified, L. moncytogenes and L. ivanovii. Enterotoxin (<1 ng/mL) and Salmonella were not detected in any non-inoculated samples.  ND  ND  ND  ND  ND  60 d  ND  ND  ND  ND  ND  ND  75 d  ND  ND  ND  ND  ND  ND  90 d  ND  ND  ND  ND ND ND d = days of maturation; ND = not detected; * Salmonella and staphylococci enterotoxin were not detected. Values with different superscripts ( a,b or A,B ) in the rows, are significantly different between adjacent columns of the same pathogen (p < 0.05).
Staphylococcus aureus. Once the cheese was molded, it was left in the ripening chamber for 24 h, and then the native S. aureus was enumerated, finding values of 7.4 log MPN/g   ND  ND  ND  ND  ND  60 d  ND  ND  ND  ND  ND  ND  75 d  ND  ND  ND  ND  ND  ND  90 d  ND  ND  ND  ND ND ND d = days of maturation; ND = not detected; * Salmonella and staphylococci enterotoxin were not detected. Values with different superscripts ( a,b or A,B ) in the rows, are significantly different between adjacent columns of the same pathogen (p < 0.05).
Staphylococcus aureus. Once the cheese was molded, it was left in the ripening chamber for 24 h, and then the native S. aureus was enumerated, finding values of 7.4 log MPN/g in the non-inoculated cheese. Table 1 shows the behavior of S. aureus in the different sections and sampling times of the non-inoculated cheese.
The statistical analysis of the non-inoculated cheese did not show a significant difference in the distribution of S. aureus regarding the analyzed sampled area up to 45 days (p = 0.343), the last sample obtained in which the bacteria was recovered. In addition, during the ripening process, no traces of staphylococcal enterotoxin were detected (<1 ng/mL).
Listeria monocytogenes. In cheese made with non-inoculated milk, native L. monocytogenes counts increased at 24 h, reaching a population of 7.2 log MPN/g and then drastically decreased on day 15, when samples from the core had 2.9 log MPN/g, and later, no L. monocytogenes was detected. On the other hand, the main matrix and surface samples allowed the pathogen to survive until day 30, with a recovery of 0.3 and 1 log MPN/g, respectively (Table 1).

Microbiological Parameters in Inoculated Samples
Staphylococcus aureus. In milk inoculated with 7 log CFU of S. aureus ATCC 13565 per mL, pathogen counts in curd and serum were 7.8 log MPN/mL and 7.1 log MPN/mL, respectively, while in the 24 h cheese, 8.7 log MPN/mL were recovered.
After seven days of maturation of the inoculated cheese, microorganisms from samples taken from the core survived until day 45 (0.15 log MPN/g), unlike the main matrix samples where it was only possible to detect them after 15 days (2.55 log MPN/g). The area where the longest survival time of S. aureus was observed was the surface, since counts were detected until day 90 with 3.6 log MPN/g. The statistical analysis of the inoculated cheese showed no significant differences regarding the recovery of the microorganisms on day seven (p = 0.326). From the second sampling time (15 days), a statistically significant difference was observed in the distribution of S. aureus (p = 0.001) in cheese.
All the samples, regardless of the sampling area, were negative for staphylococcal enterotoxin (≥1 ng/mL) in inoculated cheese after 24 h. Initially was detected in milk, rennet, and whey (Table 2). Salmonella Typhimurium. In the inoculated milk, S. Typhimurium ATCC 14028 showed a gradual and continuous decrease in all samples from the 7.2 log MPN/mL initial inoculum. After 24 h of storage, cheese samples had a reduction of 2.5 log (4.8 log MPN/g of cheese). In the following samplings, the concentration of the pathogen gradually decreased until it stopped recovering on day 45 (core) and 60 (main matrix and surface) ( Table 2).
Listeria monocytogenes. Regarding cheese made with inoculated milk (7.3 log of MPN/mL), a similar behavior was observed compared to the uninoculated milk. In cheese, a pathogen concentration increase was observed (8.7 log of MPN/g) after 24 h of storage, then its presence decreased. In the core and the main matrix, only 5.3 and 5.9 log of MPN/g were recovered until day 15, respectively, while on the surface, L. monocytogenes ATCC BAA-751 survived longer since its last recovery was on day 45 with a 4.5 log of MPN/g. Table 2 shows the behavior of each pathogen in the different parts of Cotija cheese that were analyzed during its elaboration and maturation process.
Indicator groups. Lactic acid bacteria (LAB), aerobic mesophilic bacteria (AMB), coliform organisms (CO), molds and yeasts (M/Y) were quantified. Table 3 shows the microbiota content in milk, curd, whey, and cheese with 24 h of maturation, without inoculation. No significant differences were observed in LAB, CO, M/Y of milk, curd, and whey (p > 0.05). Only the AMB had a higher content in milk and cheese (p < 0.05). The four types of samples showed <10 Log CFU/ g or mL of molds and yeasts ( Figure 2). then its presence decreased. In the core and the main matrix, only 5.3 and 5.9 log of MPN/g were recovered until day 15, respectively, while on the surface, L. monocytogenes ATCC BAA-751 survived longer since its last recovery was on day 45 with a 4.5 log of MPN/g. Table 2 shows the behavior of each pathogen in the different parts of Cotija cheese that were analyzed during its elaboration and maturation process. Indicator groups. Lactic acid bacteria (LAB), aerobic mesophilic bacteria (AMB), coliform organisms (CO), molds and yeasts (M/Y) were quantified. Table 3 shows the microbiota content in milk, curd, whey, and cheese with 24 h of maturation, without inoculation. No significant differences were observed in LAB, CO, M/Y of milk, curd, and whey (p > 0.05). Only the AMB had a higher content in milk and cheese (p < 0.05). The four types of samples showed <10 Log CFU/ g or mL of molds and yeasts ( Figure 2). Table 3. Indicator groups in milk, curd and whey not inoculated with pathogens.

Lactic Acid Bacteria Aerobic Mesophilic Bacteria Coliforms Molds and Yeast
Milk * 9.1 a 8.0 a 6.5 a 3.0 b Curd ** 9.4 a 6.9 b 6.1 a 3.1 b Whey * 8.9 a 6.9 b 5.8 a 3.0 b Cheese ** 7.2 a 8.2 a 7.2 a 4.2 b * Log of CFU/mL; ** Log of CFU/g. Values with different superscripts ( a,b ) in the columns, are significantly different between rows (p < 0.05). The indicator groups in cheese showed that after seven days of maturation, LAB continued to show a homogeneous behavior in the microbial content of the different areas analyzed (core, main matrix, and surface) and between the non-inoculated and inoculated cheese. In non-inoculated cheese, indicator groups maintained an average of around 9.0 log of CFU/g until day 45, then decreased to 1 log CFU/g from day 60, maintaining those count values until the end of this work. In the inoculated Cotija cheese, the LAB behavior was similar. However, it was maintained with an 8.8 log CFU/g until the end of the study. The indicator groups in cheese showed that after seven days of maturation, LAB continued to show a homogeneous behavior in the microbial content of the different areas analyzed (core, main matrix, and surface) and between the non-inoculated and inoculated cheese. In non-inoculated cheese, indicator groups maintained an average of around 9.0 log of CFU/g until day 45, then decreased to 1 log CFU/g from day 60, maintaining those count values until the end of this work. In the inoculated Cotija cheese, the LAB behavior was similar. However, it was maintained with an 8.8 log CFU/g until the end of the study.
The statistical analysis showed a significant difference between the main matrix and the core in the indicator groups at 15, 75 and 90 days (p < 0.05). There was no difference in either of the two groups, inoculated and not inoculated (p > 0.05).
From cheeses with seven days of maturation, main matrix and surface samples from days 7 and 15 had the highest content of indicator groups with an average of 5.9 and 6.4 log of CFU/g, respectively, while the core had 5.0 log of CFU/g. After 30 days, a gradual decrease in the microbial load was observed, more critical in the main matrix and the core. At 90 days, the AMB content was 2.4 log CFU/g in the core, 2.2 log CFU/g in the main matrix and 3.4 log CFU/g in the surface. Statistically, there was a difference in the AMB content in the surface regarding the main matrix and the core on days 7, 30, 45, 60, Appl. Sci. 2021, 11, 8154 6 of 13 75 and 90, while on day 15; the ABM content in the core, main matrix and surface was different (p < 0.05). The coliform content gradually decreased in the non-inoculated and the inoculated cheeses until it practically disappeared.
In the non-inoculated cheese, the main matrix and the core had 3 log CFU/g at 7 days, ceasing to recover at 90 and 60 days in the main matrix and core, respectively, while in the surface, the CO content at 7 days was 5.8 log CFU/g, decreasing to 0.2 log CFU/g on day 90. In the inoculated cheese, the initial count of the core, main matrix and surface was 3 log CFU/g on day 7, and at the end of the study, the counts decreased to 0.3, 0.5, and <1 log CFU/g in the main matrix, surface, and core. A significant difference was observed on days 7, 30 and 60 between the surface, core and main matrix of non-inoculated cheese. Inoculated cheese showed a significant difference in surface compared to main matrix and core on days 7, 15 and 45; and between the core compared to the surface and main matrix on days 30 and 60 (p < 0.05).
In the case of yeasts, the content was heterogeneous in both types of cheese, noninoculated and inoculated. The initial load (7 days) in the non-inoculated cheese was 1.4, 6.2 and 7.5 log CFU/g for the main matrix, surface, and core. At the end of the study, 1.4 log CFU/g were found in the main matrix, 4.4 log CFU/g in the surface and 1.2 log of CFU/g in the core. Yeast content in inoculated cheese was higher on the surface with 5.5 log CFU/g and lower in the main matrix and core (3.7 log CFU/g in both).
At 90 days, the content was higher in the surface (4.4 log CFU/g) and lower in the main matrix and core (1 log CFU/g). Statistically, all sampling times showed differences between the non-inoculated and the inoculated cheeses (p < 0.05). Finally, molds appeared after 60 days in not inoculated and inoculated cheeses. In non-inoculated cheese, no molds were recovered from the core in any sample (<1 log CFU/g). The main matrix had a mold content of 2.8, 2.1, and 0.67 log CFU/g on days 60, 75 and 90, respectively, while the surface presented 1.3, 2.3 and 2.9 log CFU/g on days 60, 75 and 90, respectively. In inoculated cheese, the main matrix had an increase from 0.3 log CFU/g at 60 days to 1 log CFU/g on days 75 and 90, while in the surface, mold increased by 1 log CFU/g on day 60 at 3.2 log CFU/g on day 75, to 3.5 log CFU/g on day 90. In the core, molds were only detected on days 75 and 90 (5.1 and 5.2 log CFU/g, respectively). The statistical analysis showed a significant difference between the three types of samples (main matrix, surface, and core) and the three sampling times in which the presence of molds in the non-inoculated cheese was detected. In contrast, in the inoculated cheese, there was only a significant difference on days 75 and 90 between the core, the main matrix, and the surface (p < 0.05).

Physicochemical Parameters
Changes were observed during the maturation time in titratable acidity, pH, water activity (Wa), and sodium chloride percentage (%) within the cheese's physicochemical parameters analyzed in different sections. The changes of each variable had a different dynamic and gave rise to a distinguishing composition between each cheese section: (A) Core (10 cm in diameter); (B) Main matrix (Section between the core and the surface), and (C) Surface (Obtained 2-cm deep from the surface) ( Figure 1).
The surface had a constant increase of 0.22% of lactic acid, increasing to 1.1% at the end of the study (90 days); pH dropped from 5.9 after seven days of maturation to 5.5 at day 90; Wa decreased from 0.96 to 0.86, and sodium chloride increased from 3.2 to 4.06 due to moisture loss on the surface. In the cheese main matrix, the initial lactic acid was 0.4% and gradually increased to 0.9% at day 90, while the pH dropped slightly from 5.5 to 5.01; the Wa decreased from 0.91 to 0.89 and sodium chloride increased from 3 to 3.4. The samples from the core of the cheese showed the most contrasting changes compared to the surface; the acidity content increased from 0.32% (day 7) to 1.52% on day 90; the pH decreased from 5.9 at day 7 of maturation to 5.0 at day 90; Wa also decreased from 0.92 on day 7 to 0.89, and the sodium chloride content increased from 2.9 to 3.5 (Table 4). Table 4. Physicochemical parameters during the cheese maturation process.

Discussion
Dairy production follows several stages, ranging from obtaining the milk on the farm to further processing in the dairy company or, in many cases, on the farm itself. During this process, the risk of contamination can enter at various points along the supply chain, compromising food safety [16]. During cheese making, microbial contamination can occur at any stage from cheese production to consumption, making cheese a significant vehicle for foodborne illness [4,17].
In the case of Cotija cheese, it is a Mexican artisan product made from raw cow milk without any heat treatment. The ripening process occurs spontaneously and is influenced by environmental conditions [12,18].
The fermentation of Cotija cheese is carried out by the native biota present in the milk; since starter cultures are not added, the initial microbiota is diverse and will influence its sensory characteristics and the safety of the final product. Because of the above, we decided to carry out this project in the region of Cotija. It was used the milk produced in the area. There was a space to work with inoculated milk, and the cheeses were stored for 90 days for their maturation, such as the traditional process of Cotija cheesemaking.
Therefore, the native biota as control (not inoculated) and inoculated cheeses had the same diversity of microorganisms. This allowed us to know the natural behavior that the microbiota has throughout the cheese maturation process and the microbial diversity in the milk obtained in the Cotija region, Michoacán, to the final product generated after 90 days of maturation.
The use of unpasteurized milk has represented a debate in the dairy industry due to the risk posed by the possibility that it is contaminated with pathogens, causing health problems for consumers. There are multiple reports of products made with raw milk in various outbreaks [19][20][21]. However, this does not mean that all products, particularly cheeses made with raw milk, pose a health risk to consumers.
The presence of pathogens in cheese depends on several factors: the type of cheese, the microbiological quality of the milk used, and the storage conditions. Fresh cheeses made with unpasteurized milk represent a greater risk because nothing inhibits the pathogen that may be present in the raw milk, favoring the development and/or survival of pathogens [2,4,20].
The difference with a fermented cheese is that the same fermentation process causes the production of compounds and changes in some parameters that can inhibit the development of the pathogens, like pH, lactic acid and the bacteriocins produced by some of the lactic acid bacteria that participate in the fermentation and cheese ripening [18]. It is important to mention that this study was done in the Cotija region using milk sourced from the dairy herd, which is regularly used for cheesemaking under the same environmental conditions as Cotija ripening.
In the present work, the inoculated pathogens: S. aureus, S. Typhimurium and L. monocytogenes were no longer detected in the different sections of the cheese after day 45 of ripening. However, S. aureus, was detected until the end of the maturation process (90 days) in surface samples, probably because of their halotolerant characteristics [22]. Cheese native S. aureus from uninoculated cheese was not detected at 45 days. Meanwhile, a different behavior was observed in S. aureus (ATCC 13565) from inoculated cheese which was still being detected until the end of the study (90 days). In both cases, they remained longer on the surface of the cheese. It may be due to the increased sensitivity of the native strain to the metabolites of cheese microbiota or the lower halotolerance in comparison to ATCC strain. S. aureus was recovered in the surface in concentrations significantly higher than in the core and the main matrix (p < 0.05). The behavior of the distribution of S. aureus in the inoculated and non-inoculated cheese did not show differences (p > 0.05); both were decreasing as time passed. On the surface of both cheeses, inoculated and non-inoculated, always remained in higher concentration (2 log MPN/g). Behavior of native L. monocytogenes of no inoculated cheese was like inoculated strain; they were not detected to 30 and 40 days, respectively. On the other hand, the distribution of L. monocytogenes in the inoculated cheese showed significant differences (p = 0.0045). The surface showed 3 log more with respect to the core and the main matrix. The distribution of L. monocytogenes in the inoculated and non-inoculated cheese did not show significant differences (p > 0.05). In both cheeses, inoculated and non-inoculated, the surface had the highest concentration of this pathogen. While the distribution of S. Typhimurium in the inoculated cheese, showed significant differences (p = 0.0005). It was observed that it was similar in the main matrix and the core, meanwhile, the surface showed 1.5 log more than the main matrix.
The inhibition of pathogens in ripened cheeses made with raw milk, like Cotija, has also been reported by other authors [23] who have shown the gradual demise of pathogens such as L. monocytogenes [24], and the inhibition of other pathogens in fermented cheeses [25,26]. The absence of inoculated pathogens is associated with the metabolic processes that occurred during cheese maturation and the inhibitory compounds that the other microorganisms have, such as lactic acid bacteria [18,26,27]. García-Cano et al. [25], reported production of bacteriocins from Enterococcus faecium and E. faecalis recovered from Cotija cheese. This author showed inhibition of S. aureus, Yersinia enterocolitica, S. Typhimurium and Pseudomonas aeruginosa. Escobar-Zepeda et al. [18] did not detect pathogenic bacteria like Salmonella, Listeria monocytogenes, Brucella, or Mycobacterium in a metagenomic analysis performed on Cotija cheese. They reported three predominant phyla (Lactobacillus, Leuconostoc and Weissella) associated with the production of various metabolites with antimicrobial activity. Within the complex microbiota in Cotija cheese, the presence of lactic acid bacteria can inhibit the inoculated pathogens, and the yeast present could have participated in this process [28][29][30]. Chombo et al. [12] observed that surface of Cotija cheese showed higher counts of yeasts, which is like our study. These microorganisms could contribute to the inactivation of pathogens observed in cheese's surface, in conjunction with physicochemical factors. Also, yeasts can be present in different cheeses [28,31].
As previously described, the metabolic activity of the microbiota present during the fermentation process generates many compounds ranging from peptides, like bacteriocins, to peroxide and organic acids, among other metabolites that determine the predominance of some microbial groups such as lactic acid bacteria, responsible for the synthesis of these compounds, as well as the typical organoleptic characteristics of Cotija cheese [7,18,32]. Even though only the production of lactic acid was quantified in the present investigation, it should be noted that the absence of pathogens coincided with the increase of lactic acid and the consequent drop in pH. Water activity and its decrease also influenced, although to a lesser degree, to the inactivation of the inoculated pathogens. Lactic acid concentration and pH had the highest impact on the decrease of S. Typhimurium and S. aureus. Both pathogens showed a gradual decrease associated with increasing acidity and decreasing pH. On the other hand, the inactivation of L. monocytogenes showed no relationship with the parameters determined since it occurred drastically.
The present study shows the importance of interaction between microorganisms that constitute the complex microbiota ecosystem of cheese made with no pasteurized milk.
The same microbiota present in cheese is used for its production. In Cotija cheese, the robustness of its native ecology is decisive to ensure its quality. Quality control considers the maturity stage, which ensures the absence of pathogens and takes advantage of the characteristics per se of Cotija cheese, including its functional properties provided by its characteristic microbiota [12]. Likewise, the quality of raw material in Cotija cheese is important because it is the one that will determine the type of microorganisms that will carry out the fermentation process and give the desired sensory characteristics. However, the raw material and milk must be free of pathogens like Brucella and Mycobacterium [18].

Bacterial Strains
Strains used in inoculated cheese were typified: Staphylococcus aureus ATCC 13565, Salmonella Typhimurium ATCC 14028 and Listeria monocytogenes ATCC BAA-751. In uninoculated cheese, native bacteria were isolated from fresh cheese and biochemically characterized.

Inoculum Preparation
All strains were separately reactivated in trypticase soy broth with yeast extract (CS-TEL) twice a row. Subsequently, each bacterium was inoculated in Roux bottles containing 200 mL of trypticase soy agar with yeast extract and 20 mL of CSTEL, then incubated at 35 • C for 18 h. The biomass obtained was separated from the agar using sterile glass beads and extracted with a pipette. Physiological saline solution was added to the biomass obtained to a final volume of 40 mL. From this suspension, the count was carried out in selective agar for each microorganism (Bright Green Agar, Oxford Agar and Baird Parker agar, respectively) to know the bacterial concentration. Additionally, in the culture of S. aureus, the presence of staphylococcal toxin was verified by visual immunoassay (Tecra 3M).

Cheesemaking
The fresh raw milk (180 L) was filtered through a clean gauze cloth to remove foreign matter, and once it reached a temperature between 30 to 33 • C, rennet was added, leaving it to rest until it formed a firm curd, which was cut using lira to obtain approximately 1 cm 3 cube.
After a few minutes of standing, the whey was drained, and the curd was transferred to a stainless-steel table and mixed with 5-10% NaCl until forming a homogeneous mass. The molding of the dough was done in previously washed and disinfected cylindrical hoop-shaped stainless-steel molds, provided with an ixtle or cotton blanket. The mass was pressed for 24 h, and the main matrix was removed from the press. The mold and the blanket with which the main matrix was wrapped allow it to give the traditional cylindrical shape of the cheese and print the characteristic mark on the rind. The blanket was removed, and the hoop was left for the next 15 days. Every 18-24 h, the cheese was removed from the mold, turned over, and the surface cleaned with a clean cotton cloth. This process was carried out daily for the first eight days and then alternates every third day until 15 days were completed; the goal was to have complete drainage and a sturdy piece that could be handled without the shaping ring. Finally, the cheese was stored to begin its ripening process. Ripening was carried out for three months within the same geographical area of Cotija, Michoacán, Mexico, to provide the environmental conditions (Temperature < 28 • C, Relative humidity: 60 to 95% and Altitude: 700 to 1700 m above the level of the sea) that give the cheese its specific organoleptic characteristics (Figure 3).
Appl. Sci. 2021, 11, x FOR PEER REVIEW 10 of 14 ripening process. Ripening was carried out for three months within the same geographical area of Cotija, Michoacán, Mexico, to provide the environmental conditions (Temperature < 28 °C, Relative humidity: 60 to 95% and Altitude: 700 to 1700 m above the level of the sea) that give the cheese its specific organoleptic characteristics ( Figure 3).

Milk
The uninoculated milk was analyzed for S. aureus, staphylococcal toxin, lactic acid bacteria, aerobic mesophilic bacteria, coliforms, molds, yeasts, pH, and titratable acidity. In the case of the previously inoculated milk, each pathogen was quantified, and the presence of staphylococcal toxin was confirmed.

Cheese
For the first sampling (24 h of maturation), only one sample was taken that ranged from the core to the cheese surface. From the 7th ripening sampling: each cheese was divided into three uniform vertical sections to take a sample from each section: (A) core (10

Milk
The uninoculated milk was analyzed for S. aureus, staphylococcal toxin, lactic acid bacteria, aerobic mesophilic bacteria, coliforms, molds, yeasts, pH, and titratable acidity. In the case of the previously inoculated milk, each pathogen was quantified, and the presence of staphylococcal toxin was confirmed.

Cheese
For the first sampling (24 h of maturation), only one sample was taken that ranged from the core to the cheese surface. From the 7th ripening sampling: each cheese was divided into three uniform vertical sections to take a sample from each section: (A) core (10 cm in diameter); (B) main matrix (Section between the core and the surface), and (C) surface (obtained 2-cm deep from the surface) ( Figure 1). The samples were placed in separate sterile bags (A, B and C, respectively) where they were homogenized manually.

Pathogen Count in Cheese
Each pathogen was counted on days 1, 7, 15, 30, 45, 60, 75 and 90, employing traditional counting per culture. In addition, at time 7, lactic bacteria, aerobic mesophilic, coliforms, molds, yeasts, pH, and titratable acidity were quantified in both cheeses, inoculated and non-inoculated. In all cases, the samples were analyzed in duplicate.
The most probable number technique (MPN) was used to enumerate the pathogens in the cheese. It was weighed 25 g of each sample, then 225 mL of CSTEL were added, homogenized for 1 min in a Stomacher (Lab Blender 400), and subsequently seven dilutions were made. Tubes were inoculated with 9 mL of CSTEL and incubated at 35 • C for 24 h from each dilution. Baird Parker agar (ABP) plates for S. aureus were inoculated from the tubes that showed bacterial growth; Bright green agar (AVB) for Salmonella and Oxford agar (OXA) for Listeria monocytogenes to confirm the development of each pathogen.
The growth of each pathogen on selective agar was first confirmed by Gram staining and the biochemical tests corresponding to each microorganism through the API system. Additionally, specific tests were carried out for each bacterium. For Salmonella strains, serological tests were performed with antiserum O Anti-serum Poly A-I & Vi (BD Difco TM), and the hemolysis test and CAMP (Cristie-Atkins-Munch-Peterson Test) for L. monocytogenes.

Staphylococcal Toxin
The presence of staphylococcal toxin was investigated by the visual immunoassay method (Tecra 3M). For which 10 mL of milk or 10 g of cheese were taken and homogenized with 20 mL of sterile water for 3 min, a 3 mL aliquot was taken. The pH was adjusted to 4; next, the aliquot was centrifuged at 1000× g, and then, the liquid was passed through a 45 µm filter, and the pH was adjusted between 6-7. Subsequently, 1 mL was placed in an Eppendorf tube, and 50 µL of the additive was added and shaken manually 200 µL of this mixture were plated in 96-well plates, incubated at 35 • C for 2 h. The liquid was discarded, and four washes were performed with the washing solution. Plates were allowed to dry, and 200 µL of the conjugate was added to each well and incubated for 1 h at room temperature (20-25 • C). The liquid was poured out, and they were carefully washed. Finally, 200 µL of substrate were added to each well and incubated at room temperature (20-25 • C) for 30 min and then the samples were read. The support was placed on a white background, and the reading was done visually by comparing it with the color card.

Lactic Bacteria Count
To 25 g of sample, 225 mL Man-Rogosa-Sharpe (MRS) broth were added twice the concentration with aerobic lactic acid bacteria broth (3M) supplement and homogenized for 1 min in a Stomacher (Lab Blender 400). Subsequently, seven decimal dilutions were made. The Petri film plates (3M) for anaerobic bacteria were inoculated with 1 mL of each dilution and incubated at 37 • C for 48-72 h in anaerobic jars provided with a gas pack.

Indicator Groups
The determination of indicator groups (aerobic mesophilic bacteria, coliforms, molds, and yeasts) was made from 25 g of each sample and homogenized with 225 mL of peptone diluent for 2 min in a Stomacher (Lab Blender 400). From here, seven decimal dilutions were made. From each dilution, Petri-film plates (3M) were inoculated with Aerobic Count and E. coli/Coliform Count, both incubated at 35 • C for 24 h, and Yeast and Mold Count incubated at 20 • C for 5 to 7 days.

Physicochemical Parameters
In addition to the microbiological determinations, pH, titratable acidity, water activity and chlorides were quantified from non-inoculated cheese samples. The pH was measured in 10 g of sample, placing them in a 100 mL volumetric flask with distilled water; subsequently, it was filtered, and pH was determined with an Orion3 Star pH Benchtop meter (Thermo Fisher Scientific, Waltham, MA, USA. In acidity, this was determined by titration from 10 g of sample, using 0.1 N NaOH and phenolphthalein as indicators. The test was performed by titration in triplicate. Water activity determination was carried out using 3TE Series AquaLab equipment (Decagon Device Inc., Pullman, Washington, USA). To determine sodium chloride, the AOAC 983.14 "Chloride in Cheese" method was used, weighing 2 g of ground cheese, and titrating the samples with a standard solution of 0.0856 M AgNO 3 .

Statistical Analysis
A one-factor analysis of variance (ANOVA) with an alpha level of α = 0.05 was performed on the results of the indicator groups, behavior, and distribution of pathogen in cheese. LSD Fisher method (Least significant difference) was performed for homogeneous groups. Statistical analysis was performed with Statgraphics Centurion XIX version 19.2.01 software.

Conclusions
This work shows important results on the behavior of pathogens and the dynamics that follow during the ripening process directly in the place where Cotija cheese is traditionally made. In addition, it shows the reduction of the risk represented by the consumption of products made with unpasteurized milk by being subjected to a maturation process, without neglecting good hygiene practices from obtaining the milk to the end of the cheese fermentation process and its subsequent marketing.