Susceptibility of Campylobacter jejuni to Stressors in Agrifood Systems and Induction of a Viable-but-Nonculturable State

ABSTRACT

Many bacteria can become viable but nonculturable (VBNC) in response to stressors commonly identified in agrifood systems. Campylobacter is able to enter the VBNC state to evade unfavorable environmental conditions, but how food processing can induce Campylobacter jejuni to enter this state and the potential role of foods in inducing the VBNC state in C. jejuni remains largely unknown. In this study, the culturability and viability of C. jejuni cells were investigated under chlorine treatment (25 ppm), aerobic stress (atmospheric condition), and low-temperature (4°C) conditions that mimicked food processing. In addition, the behaviors of C. jejuni cells in ultrahigh-temperature (UHT) and pasteurized milk were also monitored during refrigerated storage. The numbers of viable and culturable C. jejuni cells in both the pure bacterial culture and food matrices were separately determined by propidium monoazide (PMA)-quantitative PCR (qPCR) and plating assay. The C. jejuni cells lost their culturability but partially retained their viability (1% to 10%) once mixed with chlorine. In comparison, ~10% of C. jejuni cells were induced to enter the VBNC state after 24 h and 20 days under aerobic and low-temperature conditions, respectively. The viability of the C. jejuni cells remained stable during the induction process in UHT (>10%) and pasteurized (>10%) milk. The number of culturable C. jejuni cells decreased quickly in pasteurized milk, but culturable cells could still be detected in the end (day 21). In contrast, the number of culturable C. jejuni cells slowly decreased, and they became undetectable after >42 days in UHT milk. The C. jejuni cells responded differently to various stress conditions and survived in high numbers in the VBNC state in agrifood systems.

IMPORTANCE The VBNC state of pathogens can pose risks to food safety and public health because the pathogens cannot be detected using conventional microbiological culture-based methods but can resuscitate under favorable conditions to develop virulence. As a leading cause of human gastroenteritis worldwide, C. jejuni can enter the VBNC state to survive in the environment and food-processing chain with high prevalence. In this study, the effect of food-processing conditions and food products on the development of VBNC state in C. jejuni was investigated, providing a better understanding of the interaction between C. jejuni and the agroecosystem. The knowledge elicited from this study can aid in developing novel intervention strategies to reduce the food safety risks associated with this microbe.

INTRODUCTION

Many bacteria can enter a viable-but-nonculturable (VBNC) state to combat adverse environmental conditions, such as nutrient starvation, osmotic stress, and shifts in temperature and pH (1, 2). VBNC bacteria fail to form colonies on conventional culture media, but they maintain membrane integrity and low metabolic activity (2). Although pathogenic bacteria are unlikely to cause diseases in the VBNC state, they are potentially virulent after resuscitation under favorable conditions, posing a threat to food safety and public health (3–5). The risk is particularly accentuated, considering that VBNC bacteria cannot be detected using conventional microbiological culture-based methods.

Foods are frequently exposed to complex environmental conditions before, during, and after processing and preservation, thus providing stress to induce bacteria to enter the VBNC state. Several major foodborne pathogens have been reported to enter the VBNC state under conditions that mimic those of food processing. For example, nonthermal processing, such as high-pressure CO₂, could induce Escherichia coli cells to enter the VBNC state (6). In addition, E. coli and Salmonella enterica serotype Typhimurium cells were reported to survive in the VBNC state after chlorination treatment of drinking water (7). Staphylococcus aureus cells were able to enter the VBNC state under strong acid conditions, along with adequate nutrients, which is similar to conditions in foods with acidic additives (8).

Campylobacter is the major cause of foodborne bacterial gastroenteritis in developed countries (9). Raw milk, undercooked poultry products, and contaminated drinking water have been identified as potential sources of human campylobacteriosis (10). Although Campylobacter infections typically cause self-limiting human gastroenteritis, including diarrhea, fever, and abdominal cramps, they can also trigger prolonged postinfectious complications such as Guillain-Barré syndrome, Crohn’s disease, and septicemia in immunocompromised individuals (11). According to the Centers for Disease Control, members of the genus Campylobacter (mainly Campylobacter jejuni [>85%] and Campylobacter coli [5% to 10%]) cause an estimated 1.5 million cases of foodborne diseases annually (12, 13), accounting for 5% of deaths and 15% of hospitalizations due to common foodborne pathogens (14, 15). The economic cost of Campylobacter infections was estimated to be $2.2 billion in 2018 by the U.S. Department of Agriculture Economic Research Service (16).

Although C. jejuni has fastidious growth requirements (e.g., microaerobic, 32°C to 47°C) and an unusual sensitivity toward environmental stress (17), it can survive under aerobic conditions with high prevalence by either forming biofilms or entering the VBNC state (17, 18). The formation of VBNC C. jejuni in response to adverse conditions has been described previously. For example, a large percentage of C. jejuni cells were induced to enter the VBNC state after 30 days of incubation at 4°C in microcosm water (19). In addition, exposing C. jejuni cells to high osmotic pressure (7% NaCl) induced the VBNC state after 24 h (20). Moreover, C. jejuni cells inoculated into acidic Mueller-Hinton (MH) broth became nonculturable but remained viable after 2 h of incubation (21).

However, little is known about how food processing can induce C. jejuni to enter the VBNC state. The potential role of foods in inducing a VBNC state in C. jejuni also remains largely unknown. In the current study, the culturability and viability of C. jejuni cells were monitored under conditions mimicking those of food processing, including chlorine treatment, aerobic stress, and low temperatures. In addition, the effect of a food product (i.e., milk) on the development of VBNC state in C. jejuni cells was also investigated. The findings from this study provide a better understanding of the induction of a VBNC state in C. jejuni in the agroecosystem and will aid in the development of innovative mitigation strategies to reduce the health risks associated with this microbe.

RESULTS AND DISCUSSION

A schematic of the overall workflow of this study is shown in Fig. 1.

FIG 1.

Schematic workflow of this study. PMA-qPCR and the plating assay were performed to monitor the viability and culturability of C. jejuni cells under conditions that mimic those during food processing and in foods: chlorine treatment (A), aerobic stress (B), low temperature (C), UHT milk (D) and pasteurized milk (E). For samples A to D, the plating assay was conducted on MH blood agar (MHBA), as no background microflora existed. For sample E, Campy-Cefex agar was used to inhibit the microbiota in pasteurized milk and selectively cultivate inoculated C. jejuni bacteria. All samples were pretreated with PMA, followed by DNA extraction by either boiling (samples A to C) or a commercial DNA extraction kit (samples D and E). The number of viable cells was obtained by qPCR analysis of the extracted DNA.

VBNC induction by chlorine treatment.

Chlorine-based sanitizers have been widely used in the food industry for decontamination of agricultural products and agrifood processing equipment (22). However, chlorine treatment may not only inactivate bacterial cells but also lead to the induction of VBNC bacteria in some cases (23). For example, VBNC E. coli, S. aureus, and Listeria monocytogenes cells were detected in the processing water of chlorine-disinfected fresh produce (24, 25). Chicken meat is usually treated with chlorine to mitigate the load of Campylobacter and other foodborne pathogenic bacteria. Therefore, Campylobacter bacteria might be induced into the VBNC state during chlorine treatment of poultry products. An initial chlorine concentration of 50 ppm in chlorinated water is recommended by the USDA for poultry carcass wash applications, and this concentration was used as the upper limit in various studies (26–29). Therefore, NaClO solution with 25 ppm of free chlorine was selected to induce the VBNC state in C. jejuni in the current study.

The initial concentration of C. jejuni cells was 8.23 log CFU/mL, and the dynamics of the culturable and viable cells under chlorine treatment over 24 h are shown in Fig. 2. The C. jejuni cells were susceptible to the chlorine treatment, and all four strains quickly lost their culturability (<0.5 log CFU/mL) upon the addition of chlorine. The number of viable cells decreased by 1 to 2 log CFU/mL once the bacterial suspension was added to the NaClO solution, and it remained constant during the induction. Therefore, around 1% to 10% of the C. jejuni population was induced to enter the VBNC state under chlorine treatment. Similarly, in a previous study, L. monocytogenes and Salmonella enterica serovar Thompson cells quickly transited to the VBNC state, as no culturable cells could be detected after a 2-min exposure to either 3 or 12 ppm chlorine (25). In another study, Chen and coauthors treated E. coli cells with increased concentrations of chorine and investigated the impact of chlorine disinfection on the culturability and viability of the E. coli cells (23). Approximately 3 to 5 log CFU/mL E. coli cells were induced to enter the VBNC state after 5 min to 2 h of chlorine treatment. Longer exposure to chlorine, over 24 h, did not cause a significant reduction in the viable E. coli cells.

FIG 2.

Culturable and viable cell counts of C. jejuni strains with chlorine treatment (25 ppm): F38011 (a), NCTC 11168 (b), 81-116 (c), and ATCC 33560^T (d). Black squares represent viable cell counts determined by PMA-qPCR, and red circles represent culturable cell counts determined by plating assay. Each data point represents three replicates.

HOCl is the active component of NaClO solution, and it can inactivate bacteria by interacting with membrane components and changing the membrane permeability, finally leading to the leakage of cellular contents, including DNA and proteins (23). Bacteria develop various defense systems against reactive chlorine species. For example, chaperone Hsp33 and the HOCl-sensing transcription factor YjiE were upregulated when E. coli cells encountered HOCl stress (30). In addition, stress resistance genes (rpoS, marA, ygfA, and relE) and antibiotic resistance genes were expressed at higher levels in chlorine-induced VBNC E. coli cells compared with their culturable counterparts (31). In C. jejuni cells, genes related to bacterial-type flagellum-dependent cell motility, the tricarboxylic acid cycle, cellular respiration, and membrane proteins were upregulated upon treatment with chlorine (32). However, the mechanisms involved in the development of a chlorine-induced VBNC state in C. jejuni cells remain unknown and need to be investigated in future studies.

VBNC induction under aerobic conditions.

C. jejuni is generally considered a microaerophilic microbe and is highly susceptible to oxygen and its reduction products, such as superoxide anion radical, hydroxyl radical, and hydrogen peroxide (17, 33). Therefore, oxygen in the environment and food-processing chain is a primary stress to the survival of C. jejuni (34). Entering into the VBNC state has been recognized as one of the strategies for C. jejuni to counteract oxidative stress (35). In this study, the effect of oxygen on the induction of VBNC for four C. jejuni strains was investigated, and the response of the strains under aerobic conditions is shown in Fig. 3. C. jejuni strains F38011, NCTC 11168, 81-116, and ATCC 33560^T gradually lost their culturability and could not be detected after 24 h of incubation. The viability of all four strains was retained, and >10% of C. jejuni cells were induced to enter the VBNC state in the end. To compare the effects of different oxygen levels on the induction of the VBNC state in C. jejuni cells, Yagi and coauthors monitored the number of culturable cells under aerobic, microaerobic, and anaerobic conditions (36). No colonies were detected after 45, 60, and 60 days of incubation under aerobic, microaerobic, and anaerobic conditions, respectively, indicating that aerobic conditions were the most effective in inducing the VBNC state in C. jejuni bacteria. The difference in the development of the VBNC state in C. jejuni bacteria under aerobic stress between our current study and the previous study (36) might be due to the use of different bacterial strains. In addition, the constant shaking used in the current study could have dissolved more oxygen in the media to facilitate the induction process.

FIG 3.

Culturable and viable cell counts of C. jejuni strains under aerobic conditions: F38011 (a), NCTC 11168 (b), 81-116 (c), and ATCC 33560^T (d). Black squares represent viable cell counts determined by PMA-qPCR, and red circles represent culturable cell counts determined by plating assay. Each data point represents three replicates.

Exposure of bacterial species to aerobic conditions increases the level of intracellular reactive oxygen species (ROS), such as hydrogen peroxide, hydroxyl radicals, and superoxide anion radicals (17). Bacteria have specific oxidative stress genes to regulate ROS-detoxification enzymes (e.g., alkyl hydroperoxide reductase, catalase, and superoxide dismutase) (34). For example, various Gram-negative pathogenic bacteria, including E. coli and S. enterica, harbor oxidative regulators (e.g., OxyR and SoxR) to survive in oxygen-rich conditions (37, 38). However, C. jejuni lacks these regulators and defends against oxidative stress mainly using PerR (peroxide resistance regulator), which is usually found in Gram-positive bacteria (39). In addition, C. jejuni only possesses single copies of oxidative stress defense genes (aphC, sodB, and katA). The transcription levels of these genes were increased when C. jejuni encountered aerobic conditions, and deletion of these genes increased the sensitivity of C. jejuni toward atmospheric oxygen (35). Further study is needed to identify the potential roles of these genes and investigate the underlying mechanisms in inducing the VBNC state in C. jejuni bacteria under aerobic conditions.

VBNC induction under low-temperature conditions.

Low-temperature food preservation is a common method, and it has been validated as a major cause of inducing various pathogenic bacteria such as E. coli, Staphylococcus, and Vibrio into the VBNC state (8, 40, 41). In the current study, the refrigeration temperature (4°C) used for processing, transportation, and storage of animal food products was used to induce C. jejuni cells to enter the VBNC state.

The culturability and viability of C. jejuni cells were evaluated to monitor the response of the cells to low-temperature conditions (Fig. 4). The four C. jejuni strains showed similar trends during induction. The number of culturable cells gradually decreased and reached the detection limit (LOD) on day 20, indicating that all viable cells had entered the VBNC state. In comparison, although a ~1-log CFU/mL decrease was observed in the viable cells, they maintained their viability in the end. The induction of the VBNC state in C. jejuni cells under low-temperature conditions was reported previously. Baffone and coauthors monitored the culturability and viability of 10 C. jejuni strains in artificial seawater at 4°C (5). Different induction periods (12 to 15 days, 26 days, and 35 days) were required for these strains to lose their culturability, indicating strain-specific differences. The difference in the induction periods under low-temperature conditions among these studies might be due to the variation in bacterial strains, nutrient levels, and atmospheric conditions (36). In the current study, the induction time that the C. jejuni cells required to fully enter into the VBNC state under low-temperature conditions (20 days) was much longer than that required under chlorine (immediate) and aerobic stress (24 h) conditions. Similar results were observed for other bacteria, such as E. coli and Vibrio. In a study conducted by Falcioni and collaborators, successful induction of the VBNC state in Vibrio parahaemolyticus was achieved by incubating culturable cells in artificial seawater at 4°C for 69 days (42). Wei and Zhao used low temperatures (4°C and −20°C) to induce E. coli O157:H7 cells into the VBNC state in saline, distilled water, and LB broth (40). The culturability of the E. coli cells was maintained at a high level even after incubation in LB broth and distilled water for over 180 days.

FIG 4.

Culturable and viable cell counts of C. jejuni strains under low-temperature conditions (4°C): F38011 (a), NCTC 11168 (b), 81-116 (c), and ATCC 33560^T (d). Black squares represent viable cell counts determined by PMA-qPCR, and red circles represent culturable cell counts determined by plating assay. Each data point represents three replicates.

Many bacteria such as E. coli, Bacillus subtilis, and S. Typhimurium produce cold shock proteins in response to rapid temperature decrease (43). Cold shock proteins mediate membrane adaptation, cold signal sensing, and translation device alteration, all of which facilitate the growth of bacteria below the optimal growth temperature (44). However, cold shock genes, such as cspA and its homologs, were not detected in the C. jejuni genomes, possibly explaining the inability of C. jejuni cells to grow at temperatures below 30°C. However, metabolic activities such as oxygen consumption, catalase activity, ATP generation, and protein synthesis were still observed in C. jejuni cells, even when the temperature was decreased to 4°C (45). Further study is required to investigate cold shock-related genes in C. jejuni for elucidating the mechanisms for development of the VBNC state under low-temperature conditions.

VBNC induction in milk.

Raw milk should be free of microorganisms if it is produced by a healthy cow (46). However, microorganisms can be transmitted to raw milk via milking equipment, mastitis, and fecal contamination (47). Numerous C. jejuni outbreaks have been linked with the consumption of unpasteurized or inadequately pasteurized milk, and these products are identified as important vehicles for C. jejuni infections (48–50). The survival of C. jejuni bacteria in raw/pasteurized milk at low temperatures has been studied (51, 52). However, these studies mainly used culture-based methods to quantify the number of C. jejuni cells, resulting in potential underestimation of the risks of VBNC C. jejuni. To evaluate whether C. jejuni cells could be induced into the VBNC state in milk at low temperatures and to understand the role of milk in the induction process, our current study investigated the dynamics of culturable and viable C. jejuni cells in both UHT and pasteurized milk during refrigerated storage.

The number of viable C. jejuni cells in UHT milk remained stable over the entire course of 9 weeks, with a <1-log CFU/mL reduction in the end. In contrast to the rapid population reduction observed in MH broth, where no culturable C. jejuni cells were detected after 20 days of induction, the population decreased gradually in refrigerated UHT milk, and culturable C. jejuni cells became undetectable on day 42 or later (Fig. 5). The pronounced robustness of C. jejuni in UHT milk might be due to the protective effect of the proteins in milk. According to a previous study, the addition of milk proteins promoted the survival of E. coli cells in milk at low temperature (53). In another study, Rubin concluded that casein micelles contributed to the prolonged survival of pathogens in cheese (54). Strain-specific differences were also observed in the induction process in UHT milk, as the reduction in culturable cells was different among the four strains. C. jejuni strains F38011, NCTC 11168, 81-116, and ATCC 33560^T entered the VBNC state after 56, 56, 56, and 42 days of storage, respectively, indicating the relatively high susceptibility of C. jejuni ATCC 33560^T to cold stress. Similarly, as reported in a study by Feng and colleagues, more C. jejuni ATCC 33560^T cells lysed under aerobic and starvation conditions than did cells of other strains (e.g., C. jejuni F38011, NTCC 11168, etc.) (55). In addition, C. jejuni ATCC 33560^T was observed to be susceptible to various antibiotics, such as amoxicillin, ciprofloxacin, erythromycin, gentamicin, and tetracycline (56). According to sequencing analysis, C. jejuni ATCC 33560^T has fewer point mutations associated with antibiotic resistance. Moreover, no polymorphisms were observed in the regulatory region of cmeABC. CmeABC has been recognized as a multidrug efflux pump contributing to the antimicrobial resistance of C. jejuni. The transcriptional repressor CmeR can specifically bind to the regulatory region of cmeABC to modulate its expression level, so as to determine the antimicrobial susceptibility of C. jejuni (57). Taken together, these findings demonstrated the vulnerability of C. jejuni ATCC 33560^T to physical and chemical stresses. The mechanisms underlying the high susceptibility of this strain in refrigerated milk remain largely unknown and should be investigated in future studies.

FIG 5.

Culturable and viable cell counts of C. jejuni strains in UHT milk under low-temperature conditions (4°C): F38011 (a), NCTC 11168 (b), 81-116 (c), and ATCC 33560^T (d). Black squares represent viable cell counts determined by PMA-qPCR, and red circles represent culturable cell counts determined by plating assay. Each data point represents three replicates.

The survival rates of C. jejuni cells in pasteurized milk are shown in Fig. 6. Although the number of culturable cells decreased during refrigeration storage, all four strains remained their culturability in the end, with the number of culturable cells ranging from 1.5 to 5 log CFU/mL on day 21. Similar to induction of the VBNC state in UHT milk, C. jejuni strains F38011, NCTC 11168, and 81-116 showed relatively more robustness, while C. jejuni ATCC 33560^T was susceptible to stressors in pasteurized milk. The inactivation of culturable C. jejuni cells in pasteurized milk was greater than that in UHT milk, suggesting the potential presence of antimicrobials. Pasteurized milk contains the lactoperoxidase system (i.e., lactoperoxidase, thiocyanate, and hydrogen peroxide), which can generate antimicrobial intermediate products against a wide range of Gram-negative bacteria, including E. coli, L. monocytogenes, S. aureus, and C. jejuni (58–61). Moreover, the background microflora in pasteurized milk can grow rapidly in the late stages of storage and produce metabolites to inhibit C. jejuni growth (51). In contrast, UHT treatment completely inactivates lactoperoxidase and microbes in milk, resulting in fewer stressors for C. jejuni bacteria to encounter. A previous study reported a slower decline in the culturability of C. jejuni cells in pasteurized milk, in which a reduction of <1 log CFU/mL was observed over 14 days of storage (62). The difference between the process of induction of C. jejuni cells into the VBNC state in pasteurized milk in the current study and this previous study might be due to the use of different C. jejuni strains and different antimicrobials in milk (e.g., lactoperoxidase system, metabolites produced by background microflora, etc.). Although a rapid decrease was observed in the culturable population, there was a <1-log CFU/mL reduction in viable C. jejuni cells during the induction process in the current study. Thus, C. jejuni cells in contaminated milk can be induced to enter the VBNC state, posing a threat to food safety and public health.

FIG 6.

Culturable and viable cell counts of C. jejuni strains in pasteurized milk under low-temperature conditions (4°C): F38011 (a), NCTC 11168 (b), 81-116 (c), and ATCC 33560^T (d). Black squares represent viable cell counts determined by PMA-qPCR, and red circles represent culturable cell counts determined by plating assay. Each data point represents three replicates.

Conclusion.

In conclusion, stressors in food-processing conditions and food products could trigger C. jejuni cells to enter the VBNC state. A certain number of C. jejuni cells transitioned into the VBNC state, although they responded differently to various stress conditions. In addition, the heterogeneous behaviors among the tested strains under the same stress conditions indicated the existence of strain-specific difference. Further study is required to decipher the molecular mechanisms underlying the formation of the VBNC state under various stress conditions and the strain-specific differences that existed among different C. jejuni strains. Taken together, this study provides insight into the induction and persistence of VBNC Campylobacter bacteria in the environment and aids in the development of innovative mitigation strategies to reduce the prevalence of this microbe in the agroecosystem.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Two C. jejuni human clinical isolates (F38011 and 81-116) and two reference strains (ATCC 33560^T, of bovine origin, and NCTC 11168, of human origin) were used in this study. All strains were routinely cultivated on MH agar (BD Difco, Franklin Lakes, NJ, USA) containing 5% (vol/vol) defibrinated sheep blood (Cedarlane, Burlington, ON, Canada) under microaerobic conditions (85% N₂, 10% CO₂, and 5% O₂) at 37°C. A single colony of each strain from the agar plates was transferred to MH broth (BD Difco) and cultivated under microaerobic conditions at 37°C with constant shaking at 175 rpm. The bacterial suspension was diluted with MH broth to obtain a final optical density at 600 nm (OD₆₀₀) of 0.3 (~log 9 CFU/mL), after incubation for 16 to 18 h to the late log phase.

Induction of the VBNC state in C. jejuni cells under food-processing conditions.

Different food-processing conditions, including chlorine treatment, aerobic stress, and low temperatures, were used to induce C. jejuni cells to enter the VBNC state. For chlorine treatment, an overnight C. jejuni culture was centrifuged at 15,000 × g for 5 min at room temperature and washed with phosphate-buffered saline (PBS). The obtained cell pellet was then resuspended in 25 ppm NaClO solution (Sigma-Aldrich, St. Louis, MO, USA) to a final concentration of 8 ± 0.5 log CFU/mL and incubated under microaerobic conditions at 37°C. For aerobic stress and low-temperature treatment, overnight C. jejuni cultures were adjusted to a final concentration of 8 ± 0.5 log CFU/mL in MH broth and incubated under atmospheric conditions at either 37°C or 4°C. All bacterial samples were under constant shaking at 175 rpm during the process of VBNC induction.

The dynamics of the culturable and viable cells in the bacterial population were separately monitored by plating assay and PMA-qPCR at a predetermined sampling schedule, namely, (i) chlorine treatment for 0, 3, 6, 9, 12, and 24 h; (ii) aerobic stress for 0, 4, 8, 12, 24, and 36 h; and (iii) low temperature every 4 days until day 24. The population of VBNC C. jejuni cells was calculated by subtracting the culturable cell counts from the viable cell counts. Bacterial culture (2 mL) was transferred onto each MH blood agar (MHBA) plate to confirm the absence of culturable cells. Thus, the detection limit (LOD) was determined to be 0.5 CFU/mL. When the concentration of culturable cells was below the LOD, all viable cells were considered to be in the VBNC state.

Induction of the VBNC state in C. jejuni cells in milk.

Milk was selected as a representative food model to investigate its role in inducing the VBNC state in this microbe, because the consumption of contaminated milk is one of the major routes of human campylobacteriosis transmission (63). Considering that induction of C. jejuni bacteria into the VBNC state requires a relatively long time under low-temperature conditions, ultrahigh-temperature (UHT) milk was used, due to its long shelf life through which the transition of C. jejuni cells from a culturable state to the VBNC state could be monitored. However, UHT milk is considered sterile, and there is a relatively low possibility of Campylobacter contamination during storage. To better mimic a real scenario, pasteurized milk was also selected to evaluate the process of bacterial entry into the VBNC state as it may carry microorganisms, including C. jejuni bacteria, caused by inadequate pasteurization or preservation.

Both UHT milk (2% fat) and pasteurized milk (3.8% fat) were purchased from a local grocery store in Montreal. A C. jejuni culture (10 mL) was separately added into each milk sample (90 mL) to obtain a final concentration of 8.0 ± 0.5 log CFU/mL, followed by storage at 4°C to facilitate the induction of VBNC bacteria. To be consistent with the induction process using a pure bacterial culture under low-temperature conditions, the UHT milk samples were incubated with constant shaking at 175 rpm. Meanwhile, the pasteurized milk samples were kept static to mimic real storage conditions for milk. The numbers of culturable and viable C. jejuni cells in the milk samples were separately monitored by plating assay and propidium monoazide (PMA)-quantitative PCR (qPCR). The sampling time for UHT milk was set as every 7 days until day 63, and the sampling time for pasteurized milk was set as every 3 days until day 21. Considering the existence of a background microbiome in pasteurized milk, Campy-Cefex agar was prepared and used to selectively cultivate C. jejuni cells (64), and the agar plates were incubated at 37°C under microaerobic conditions for 48 h before enumeration.

PMA treatment.

PMA-qPCR has been widely used for the detection and quantification of VBNC bacteria (4, 8, 65). PMA is a DNA intercalating dye that can only penetrate bacterial cells with compromised cell membranes and covalently binds to double-stranded DNA upon photoactivation. As the amplification signals from dead cells are blocked, only DNA from viable cells can be amplified during the subsequent qPCR process, enabling the quantification of VBNC cells.

The number of viable cells in the bacterial population was determined by PMA-qPCR, with modifications to the procedures based on the method developed by Lv and others (20). Briefly, 1 mL of each bacterial suspension under different induction conditions was centrifuged at 15,000 × g for 5 min and resuspended in PBS with the same volume. A total of 395 μL of the bacterial suspension was first mixed with 100 μL of PMA enhancer for Gram-negative bacteria (Biotium Inc., Hayward, CA, USA) to improve the affinity of PMA to the DNA of the dead cells. Afterward, PMA solution (Biotium Inc.) was added to the cell mixture at a final concentration of 15 μM in a transparent microcentrifuge tube (Froggabio, Concord, ON, Canada). Following incubation in the dark at room temperature for 10 min, the samples were exposed to a 600-W halogen lamp (Smith Victor, Bartlett, IL, USA) for another 10 min. During light exposure, all the samples were placed horizontally on ice at a distance of 20 cm from the light to avoid excessive heating. The complete mixing and cross-linking of PMA and the DNA of dead cells were achieved by constant agitation during the entire PMA treatment. Before DNA extraction, the PMA-treated samples were centrifuged at 15,000 × g for 5 min and washed once with double-deionized water (ddH₂O) to remove the unbound PMA.

DNA extraction and qPCR.

A thermal treatment was used for the rapid extraction of bacterial DNA from nonfood samples. In brief, genomic DNA was released from the bacterial cells by boiling them at 100°C for 10 min, followed by cooling them on ice for another 10 min. To obtain DNA samples with a high yield and purity for downstream analysis, C. jejuni genomic DNA was extracted from the spiked milk using the PowerFood microbial kit (Qiagen, Germantown, MD, USA), according to the manufacturer’s instructions. The extracted DNA was stored at −20°C until qPCR amplification.

The primer set (forward, 5′-GAGTAAGCTTGCTAAGATTAAAG-3′; reverse, 5′-AAGAAGTTTTAGAGTTTCTCC-3′) targeting the DNA-directed RNA polymerase of C. jejuni was selected for amplification; its specificity was validated in a previous study performed by Lv and others (20). qPCR was performed on an Mx3005P real-time PCR system (Agilent, Santa Clara, CA, USA) with a total volume of 20 μL: 2 μL of DNA template, 10 μL of SensiFAST SYBR Lo-ROX kit reagent (Bioline, Memphis, TN, USA), 0.2 μL of each primer (100 nM), and 7.6 μL of sterile ddH₂O. The amplification program was developed as follows: 1 cycle of 50°C for 2 min and then 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. At the end of each extension step, the fluorescence density of the SYBR green was measured and used to calculate the cycle threshold (C_T) value of each sample. Samples were tested in triplicate, and DNase-free sterile ddH₂O was included as the negative control in each run of the experiment.

The sensitivity of the PMA-qPCR assay was evaluated in a live-dead bacterial mixture to mimic practical conditions, where bacterial cells exist in different states. The bacterial mixture was prepared by adding a dead C. jejuni cell culture (6 log CFU/mL; heat inactivated at 90°C for 5 min) to a 10-fold serially diluted live cell culture (ranging from 2 to 8 log CFU/mL). The aforementioned PMA treatment was applied to the mixture, and bacterial genomic DNA was extracted using the boiling method. A standard curve was obtained by correlating the C_T values against known concentrations of C. jejuni cells and used to quantify the number of viable cells. The LOD of this assay was defined as the lowest concentration of bacterial cells that generated a C_T value of >35.

Statistical analysis.

At least three biological replicates were performed for all experiments, with 2 to 4 technical replicates. The data were expressed as the mean value of 3 independent replicates ± standard deviation. Comparison among the different groups was conducted using analysis of variance (ANOVA), followed by Duncan’s multiple range test, using SPSS 20.0 software (SPSS Inc., Chicago, IL, USA), and the significance level was set at P < 0.05.