Benzoic acid, propionic acid, and sorbic acid are food preservatives that extend the shelf life of food by preventing the deterioration of quality by microorganisms (Silva and Lidon, 2016). Some preservatives are naturally contained in raw food materials or may be introduced into the food by careless handling or fermentation (Jang et al., 2020; Kim et al., 2018; Lee et al., 2013; Lim et al., 2013; Park et al., 2008; Yun et al., 2017; Yun et al., 2019). However, it is difficult to distinguish between intentionally added preservatives in the food and the preservatives naturally produced in food (Park et al., 2008).
The World Health Organization (WHO) reported that benzoic acid is produced by many plants as an intermediate product in the formation of other compounds, and is detected in high concentrations in berries and in animals (WHO, 2000). Several studies have shown that benzoic acid is frequently detected in dairy products (Cakir and Cagri-Mehmetoglu, 2013; Qi et al., 2009). Benzoic acid in dairy products may be produced by lactic acid bacteria or an anaerobic metabolism of phenols in cheese (Sieber et al., 1995). Kurisaki et al. (1973) showe d that benzoic acid can be produced from phenylalanine in yeast-ripened cheese. Another study has reported that yeast-mold counts affect the formation of benzoic acid (Yerlikaya et al., 2021).
Although propionic acid is not a component of fats or oils, it has been reported to occur as an intermediate metabolite by oxidation of fatty acids (FAO and WHO, 1974), and the Code of Federal Regulation specified that propionic acid is produced by chemical synthesis or bacterial fermentation (FDA, 2022). The Environmental Protection Agency (EPA) also reported that propionic acid is a common intermediate metabolite in the living body, and is one of the metabolites produced by the decomposition of several amino acids (EPA, 1991). Thus, the European Food Safety Authority (EFSA) published a scientific opinion reevaluating propionic acid as a naturally occurring substance (EFSA, 2014). Sorbic acid is naturally found in the oil of ash tree berries in 1859 (Sofos, 1989). Kim et al. (1999) reported the contents of benzoic acid and sorbic acid in 39 plants used as tea or spices in Korea, the content of benzoic acid in spices and the content of sorbic acid in teas or spices were less than 10 ppm. Yun et al. (2017) reported the levels of natural preservatives of sorbic acid in spices. Sorbic acid was found in 88 samples from a total of 493 samples with a concentration of not detected-57.70 mg/L.
Many countries have regulations to limit the concentrations of benzoic acid, sorbic acid, and propionic acid in food for intentional addition. However as described above, the natural production of these preservatives cannot be distinguished from the current technology. If the preservatives are added intentionally to food, their purpose is to inhibit microbial growth. Notably, preservative concentration below minimal inhibitory concentration (MIC) in food could be due to natural production. Various studies on MIC of preservatives against microorganisms have been conducted (Haque et al., 2009; Stanojevic et al., 2009; Warth, 1985; Warth, 1986). However, these studies usually used broth media rather than food matrices. In addition, the previous studies examined one microorganism. Because of the reasons, the results from the studies were not appropriate to be used for microbial standards. If MIC for preservatives are determined with a mixture of microorganisms, which are the most sensitive against the preservatives, in food matrices, the results could be used for establishing microbial standards. In this case, even though the food preservatives are detected in food, if the concentration is below the MIC, the food preservatives might be produced naturally rather than intentional addition, because people do not add the preservatives below the MIC determined with the most sensitive microorganism.
Therefore, the objective of this study was to determine the MIC of propionic acid, sorbic acid, and benzoic acid to the most sensitive microorganisms in animal products, to be used as a standard for determining if the preservatives in food are natural production or intended addition.
Materials and Methods
Unprocessed animal products and processed animal products were selected based on following criteria; i) cases of research on natural preservatives, ii) food items and raw materials with high consumption (MFDS, 2020), iii) fat content. For unprocessed animal products, eggs, chicken breast, chicken legs, pork ribs, pork sirloin, beef ribs, beef chuck, and milk samples were used. For processed animal products, processed butter, fermented milk, ground meat product, natural cheese, and smoked egg samples were used. These samples were purchased from local supermarkets and butcher shops.
Considering the strain variation of microorganisms, a strain mixture for each microorganism was prepared as inoculum. Bacteria strains were cultured in 10 mL of culture media at optimal incubation temperature for 24 h. Aliquots (0.1 mL) of the cultures were inoculated in 10 mL fresh culture media and subcultured at optimal temperature for 24 h. Yeast and mold strains were cultured in 10 mL of culture media at optimal incubation temperature for 24–48 h. Aliquots (0.1 mL) of the cultures were inoculated in 10 mL fresh culture media and subcultured at optimal temperature for 24–48 h. The cultures of the strains for each microorganism species were mixed. Each mixture was then centrifuged at 1,912×g and 15 min for 4°C, and the cell pellets were washed twice with phosphate-buffered saline [PBS; KH2PO4 0.2 g, Na2HPO4 1.5 g, NaCl 8.0 g, KCl 0.2 g, 1 L of distilled water (DW), pH 7.4]. For the bacteria and yeast inocula, cell pellets were diluted with PBS to have 6 Log CFU/mL. For the mold inocula, the resulting suspensions of conidia were vigorously vortexed, and sterile DW was added to the suspension to have 5 Log CFU/mL. Mold cell counts were measured by a hemacytometer, which was confirmed by a serial dilution plate count. The microorganism strains and culture media used in this study were presented in Table 1.
MIC were determined by a broth microdilution method according to the recommendation of the CLSI M07-A, M27-A, and M38-A (Balouiri et al., 2016; CLSI, 2002; CLSI, 2008; CLSI, 2012). Mueller Hinton Broth (MHB; Becton Dickinson, Franklin Lakes, NJ, USA) was used for bacterial cultures, and RPMI-1640 medium (Thermo Fisher Scientific, Waltham, MA, USA) was used for yeast and mold cultures. The pH of MHB was adjusted to pH 7.0 using HCl and NaOH, and the pH of RPMI-1640 medium was adjusted to pH 7.0 with 0.165M MOPS (M1254, Sigma-Aldrich, Gillingham, UK). Preservatives examined were extra pure grade propionic acid (Daejung, Siheung, Korea), food-grade benzoic acid (W213101, Sigma-Aldrich), sorbic acid (W392103, Sigma-Aldrich), calcium propionate (Niacet B.V., Tiel, Netherlands), sodium propionate (Niacet B.V.), sodium benzoate (Wuhan Youji Industries, Hubei, China), and potassium sorbate (Ningbo Wanglong Technology, Zhejiang, China). The stock solution of the preservative was dissolved in MHB and RPMI-1640 medium, and they were two-fold diluted serially with MHB and RPMI-1640 medium. The tests were performed in 96 well-microtiter plates, and 180 μL of diluted preservative solutions with different concentrations were placed in the wells. Each well was inoculated with 20 μL of the inocula at 4 Log CFU/mL. The 96 well microtiter plates were incubated at 35°C for 24 h for the growth of the bacteria and yeast, and at 35°C for more than 48 h for the growth of the fungi. Positive control was the media inoculated with bacteria without a preservative, and negative control was media only. Concentrations at which no optical turbidity was observed after incubation were considered MIC.
To examine the antimicrobial effect of preservatives at low pH, five bacteria that were the most sensitive to the preservatives at pH 7.0 were subjected to propionic acid, benzoic acid, and sorbic acid in MHB at pH 4.5, 5.5, and 6.0. To determine MIC according to the method described in the section of ‘Minimum inhibitory concentrations of preservatives for microorganisms at pH 7.0’, the pH of MHB was adjusted with HCl.
Bacteria that were the most sensitive to propionic acid, benzoic acid, and sorbic acid were used to determine MIC of preservatives in unprocessed animal products (eggs, chicken breast, chicken legs, pork ribs, pork sirloin, beef ribs, beef chunk, and milk) and processed animal products (processed butter, ground meat product, natural cheese, and smoked eggs). The selected bacteria were Campylobacter coli ATCC33559, Campylobacter jejuni ATCC33560, Erwinia carotovora KCCM11319, Micrococcus luteus KCCM11211, and Moraxella catarrhalis KCCM42707. A mixture of the bacteria was prepared according to the procedure described in the section of ‘Inoculum preparation’. Inoculum 0.1 mL was inoculated to 25 g of food sample in a sample bag to obtain a concentration of 4 Log CFU/g. A hundred microliters of the preservatives were then spiked in samples to have 0, 100, 500, 1,000, and 1,500 (1,200 ppm for sorbic acid) ppm. Pork ribs, pork loin, beef ribs, beef chunks, milk, processed butter, fermented milk, and natural cheese were stored at 10°C. Poultry and processed meat products were stored at 5°C, and smoked eggs were stored at 25°C. The sample (25 g) was aseptically transferred to a sample bag containing 225 mL of buffered peptone water (BPW; Becton Dickinson, Sparks, MD, USA), and the sample was pummeled for 60 s in a pummeler (BagMixer® 400, Interscience, Saint Nom la Bretehe, France). One milliliter of the homogenate was serially diluted with BPW, and the homogenates were dispensed on an aerobic bacteria count plate (AC Petrifilm; 3MTM Petrifilm aerobic count plate, 3M, St. Paul, MN, USA) to quantify the total bacteria. The AC Petrifilms were incubated at 35°C for 48 h, and the colonies were then manually counted. The end time of the storage was determined as the time when the bacterial cell counts in the 0-ppm sample increased to 6 Log CFU/g. This experiment was repeated three times. The bacterial cell counts for each concentration of preservatives at the end of the storage were compared to the cell counts on day 0. This comparison was conducted by pairwise t-test at α=0.05 with the general linear model procedure (proc glm) of SAS® (ver.9.4, SAS Institute, Cary, NC, USA). If the difference was not significant, the concentration was determined as MIC per each replication. Among the MIC of 3 replications, the lowest MIC was determined as a final MIC.
Results and Discussion
Control of microorganism growth in raw food materials and products is important in ensuring product safety, shelf life, and consumers’ health. In meat, Pseudomonas, Acinetobacter, and Brochothrix mainly affect the quality and may cause spoilage (Liang et al., 2021; Wei et al., 2021). Also, pathogenic bacteria such as Escherichia coli, Salmonella, Campylobacter, Listeria monocytogenes, and Staphylococcus aureus are frequently detected in meat (Kim et al., 2020; Lee and Yoon, 2021; Park et al., 2021; Yang et al., 2022). Spoilage yeasts mainly include Zygosaccharomyces, Saccharomyces, Candida and Brettanomyces, and spoilage molds include Zygomycetes, Penicillium, Aspergillus, etc. (Blackburn, 2006). Especially, spoiled meats and cheeses often have high cell counts of Debaryomyces, Yarrowia, and Rhodotorula (Blackburn, 2006). The MIC of propionic acid, sorbic acid, and benzoic acid to these microorganisms in broth media were determined at pH 7.0 (Table 2). To increase the solubility of preservatives, salts were combined with the preservatives. Calcium propionate, sodium propionate, sodium benzoate, and potassium sorbate were also examined, and they had higher MIC than acid-type preservatives (Table 2). C. coli, C. jejuni, M. catarrhalis, E. carotovora, and M. luteus had lower MIC for the preservatives (propionic acid, benzoic acid, and sorbic acid), compared to other microorganisms. The preservative used in this study is a weak-acid type, which increases the number of non-dissociated molecules, when the pH is lowered. Thus, the molecules easily penetrate the microbial cell membrane or protoplasm, which prevents microbial growth (Theron and Lues, 2007). Unlike the acidic-preservatives, salt preservatives are considered to have a high MIC, because their pH were close to neutral. To investigate the antibacterial activity of preservatives according to pH, MIC of the preservatives were investigated by adjusting the pH of the medium to 4.5, 5.5, and 6.0. The five bacterial strains showed lower MIC of the preservative at lower pH (Table 3). The MIC of the preservative for E. carotovora were 50 ppm for propionic acid, 25 ppm for sorbic acid, and 50 ppm for benzoic acid at pH 5.5, which were lower MIC than those at pH 6.0. These results confirmed that the microbial growth prevention efficacy of the weak-acid type preservatives increased at low pH as presented in other research.
Unprocessed animal products were inoculated with a mixture of the most sensitive foodborne bacteria selected by MIC to the preservatives, and the samples were stored at 10°C until the bacterial cell counts of the control increased to >106 CFU/g, which is considered to be the level that the spoilage started. At this time the total bacteria in other samples were counted.
The MIC of preservatives in animal products are presented in Table 4. The MIC of propionic acid were 100 ppm in chicken legs, pork ribs, pork sirloin and beef ribs, 500 ppm in chicken breast, beef chunk and milk, and 1,500 ppm in eggs. The MIC of benzoic acid were 100 ppm in chicken legs, pork ribs, and pork sirloin, 500 ppm in chicken breast, beef ribs, beef chunk, and milk, and 1,500 ppm in eggs. The MIC of sorbic acid were 100 ppm in chicken breast, chicken legs, pork ribs, pork sirloin, beef ribs, and beef chunk, and 500 ppm in milk, and 1,200 ppm in eggs. The MIC of propionic acid, benzoic acid, and sorbic acid in processed butter and natural cheese were 100 ppm. In smoked eggs, MIC of propionic acid were 1,000 ppm, and MIC of benzoic acid and sorbic acid were 500 ppm. In our study, the MIC investigated in food were higher than pH in broth media. Specifically, the pH of ground meat was close to 6.0 and the MIC of propionic acid, benzoic acid, and sorbic acid were 1,500, >1,500, and >1,500 ppm, respectively. However, the MIC in the broth of the five strains of microorganisms used as inoculum were below 500 ppm at pH 6.0.
Preservatives are food additives that inhibit microbial growth in food, but most studies have identified MIC in microbiological media rather than food. Although few studies have evaluated the MIC of preservatives in food, it is known that the MIC of preservatives in food were higher than those in microbiological media (Brocklehurst et al., 1995; Weiss et al., 2015). While the media have homogeneous structure and consist of simple composition, the food consists of various components (fat, protein, fiber, and antibacterial substances) and structures (Weiss et al., 2015). Lipid content and preservative activity are correlated (Glass and Johnson, 2004; Weiss et al., 2015). Organic acids such as propionic acid bind to phospholipids in the bacterial cell membrane. However, the fat component in food also competitively binds to lipophilic molecules, making it difficult for preservatives to bind to bacteria. Electrostatic and hydrophobic interactions also significantly affect the activity of acid-type preservatives that are dissociated (Weiss et al., 2015). These reasons may also have caused the differences in MIC between the broth media and animal products in our study.
Many studies evaluated MIC in broth media rather than in food matrix. In our study showed that MIC were higher in animal products than in the broth media. Thus, the case of the MIC determined in the animal products might be appropriate to be determine if the detected preservatives in food are added intentionally or not, because preservatives are added to inhibit microbial growth, and thus, the concentrations should higher than the MIC.