Characteristics of viable but nonculturable Vibrio parahaemolyticus induced under extended periods of cold-starvation with various NaCl concentrations

This study was undertaken to examine the induction of VBNC states of Vibrio parahaemolyticus under prolonged cold-starvation with various NaCl concentrations and their responsive characteristics to maintain cell viability. V. parahaemolyticus entered the viable but nonculturable (VBNC) state in artificial sea water at 4°C within 80 day and persisted in the VBNC state for 150 days. During cold-starvation, bacterial cells were used to estimate their cell functions, including cytotoxicity, fatty acid composition, membrane potential, and morphology. Cytotoxic effect of V. parahaemolyticus cells against animal cell lines was decreased to below 50% after 80 days. VBNC V. parahaemolyticus cells showed decreasing levels of palmitic, vaccenic, and hexadecenoic acid on membrane, concomitantly with the formation of empty gaps between the cytoplasmic and outer membrane, in comparison with those of the pure cultures. Starvation at 4°C for 30 days resulted in a high increase in N-phenyl-1-napthylamine intensity within V. parahaemolyticus cells. Membrane potential and cellular composition were strongly affected by increasing NaCl contents of the microcosms after its evolution into the VBNC state. VBNC V. parahaemolyticus cells may undergo selected physiological changes such as the modulation of membrane potential and re-arrangement of cellular composition.


Introduction
parahaemolyticus ST550 was induced into the VBNC state in Morita mineral salt solution (MMS) at 4°C for 32 days. Previous studies indicate that approximately 30-70 days were required for V. parahaemolyticus strains to enter the VBNC state at 3-5°C (10,16,17).
Importantly, conversion of VBNC forms can be accelerated more rapidly when V.
parahaemolyticus cells in the stationary growth phase were incubated in ASW microcosms supplemented with high concentrations of NaCl at 4°C (18,19). Cold-starvation in ASW microcosms supplemented with 5%, 10%, and 30% NaCl at 4°C led to the phase transition of V. parahaemolyticus into the VBNC state within 30, 14-21, and 3 days, respectively.
Considering that food preservation processing such as low pH or high NaCl are commonly used to prevent the growth of spoilage and pathogenic organisms on food, the addition of NaCl may shorten the incubation-times that are needed for V. parahaemolyticus to enter the VBNC state. As microorganisms in such a dormant but viable state may be recovered in a favorable environment where provides sufficient energy sources to encourage their biological function and growth, the incidence of VBNC pathogens on food would be closely involved in the food-borne disease outbreaks and pose a potential risk to public health. However, little is known about how high NaCl contents in nutrient-deficient microcosms affect the formation and physiological characteristics of VBNC V. parahaemolyticus induced by prolonged coldstarvation. Exploring cell properties of V. parahaemolyticus in response to various environmental conditions may be critical for better understanding the ecology of this pathogen, as well as its survival mechanisms. In this study, V. parahaemolyticus ATCC 17802, V. parahaemolyticus ATCC 33844, and V. parahaemolyticus ATCC 27969 were incubated in ASW microcosms (pH 6) supplemented with various NaCl concentrations at 4°C until these bacteria were induced into the VBNC state. Physiological properties of VBNC V. parahaemolyticus were characterized by measuring cytotoxic effects to animal cell lines, membrane potential with N -phenyl-1 -napthylamine (NPN) uptake, intracellular leakage of nucleic acid and protein, cell hydrophobicity, and morphological change. Additionally, fatty acid composition on cell membrane of V. parahaemolyticus was analyzed before and after the induction of VBNC cells.

Preparation of microcosm
According to the instructions provided by a reliable supplier, 30  To determine effects of higher NaCl concentrations on the induction of V. parahaemolyticus into the VBNC state, each of microcosms was modified by adding excessive amounts of NaCl (5%, 10%, and 30%) and its acidity was adjusted to pH 6 using membrane-filtered 1 N NaOH (Kanto chemical, Tokyo, Japan). The modified microcosms had different NaCl contents: 0.75% (ASW), 5% (ASW 5 ), 10% (ASW 10 ), and 30% NaCl (ASW 30 ). All microcosms were autoclaved at 121°C for 20 min prior to use. for 24 h. V. parahaemolyticus cells in the stationary phase were harvested by centrifugation at 10,000 × g for 3 min, washed in ASW, and the final pellets were re-suspended in 1 ml of ASW, corresponding to approximately 10 8-9 CFU/ml. Each bacterial suspension was inoculated into ASW, ASW 5 , ASW 10 , and ASW 30 , respectively. The microcosms were kept at 4°C until the culturable counts of V. parahaemolyticus decreased to below the detection limits (< 1.0 log CFU/ml).
parahaemolyticus ATCC 27969 were plate-counted on tryptic soy agar (Difco) amended with 3% NaCl (TSA S ). Decimal dilutions (10 -1 ) were prepared in alkaline peptone water (APW, Difco) consisting 10 g of peptone and 10 g of NaCl in 1 l of sterile DW. Then, 100 µl of these aliquots were spread on TSA S . Each agar plate was incubated at 37°C for 24 h, and colonies shown on the media were enumerated.

Epifluorescence microscopy with SYTO 9 and propidium iodide
Total and viable counts of V. parahaemolyticus were measured using the Live/Dead® BacLight™ Bacterial Viability Kit (Invitrogen, Mount Waverley, Victoria, Australia) containing two nucleic acid stains, SYTO9 and propidium iodide (PI). While SYTO9 has a high affinity for DNA and chromosome and is used for labelling bacterial cells with intact and compromised membranes, PI selectively penetrates bacterial cells with damaged membranes. Briefly, equal volumes (1:1) of SYTO9 and PI were combined, and 3 µl of this mixture were added to 1 ml of the bacterial suspension. After 15 min of incubation at 25°C in the dark, 5-8 µl of the bacterial aliquots were attached on a glass slide. Bacterial images were demonstrated via an electron-fluorescent microscope (TE 2000-U, Nikon, Tokyo, Japan).

Cytotoxicity assay
Caco2 and Vero cell lines were cultured in 5-10 ml of Dulbecco's modified eagle's medium (DMEM, Corning, NY, USA) supplemented with 5% (DMEM 5 ) and 20% (DMEM 20 ) fetal bovine serum (FBS, Corning) at 37°C for 2 days in 5% CO 2 , respectively. After 2 days of incubation, each DMEM solution was removed in a petri-dish and washed in 5 ml of PBS three times. Each culture was added by 5 ml of trypsin (Corning) for cell lysis and incubated at 37°C for 5 min in 5% CO 2 . To alleviate the enzymatic activity caused by trypsin, 2-3 ml of the DMEM media, such as DMEM 5 and DMEM 20 , were added to Caco2 and Vero cells, respectively. Animal cell fluids were further transferred to sterile cap tubes and centrifugated at 15,000 × g for 3 min. The supernatants were eliminated, and cell pellets from Caco2 and Vero were re-suspended, corresponding to the cell density of 10 4 ml -1 , in 5 ml of DMEM 5 and DMEM 20 , respectively. Then, 100 μl of Caco2 and Vero were loaded into 96-well plates containing 100 μl of DMEM 5 and DMEM 20 , respectively. The eukaryotic cell lines were incubated at 37°C for 24 h in 5% CO 2 before use. At regular time-intervals, V. parahaemolyticus cells incubated in ASW microcosms were withdrawn from the incubator.
The bacterial aliquots (100 μl) were added to 96-well plates containing 100 μl of each cell lines and were incubated at 37°C for 24 h in 5% CO 2 . Five mg ml -1 of 3-(4, 5dimethylthiazol-2-yl)-2, 5 diphenyl tetrazolium bromide (MTT, Corning) was added to each well in the 96-well plates, and the cell fluids were incubated at 37°C for 1 h. The culture fluids were added by 100 μl of DMSO (Corning) and were read on a microtiter plate reader at optical densities (ODs) between 570 and 620 nm (Multiskan GO Microplate Spectrophotometer, Thermo Scientific, Vantaa, Finland).

Leakage of cellular components
Bacterial solutions (1.5 ml) of V. parahaemolyticus incubated in ASW, ASW 5 , ASW 10 , and ASW 30 at 4°C for 100 days were transferred to sterile microtubes and were centrifugated at 15,000 × g for 3 min. Each cell supernatant was collected and used to assess the leakage of cellular components such as DNA and protein via a microtiter plate reader (Multiskan GO Microplate Spectrophotometer, Thermo Scientific) at OD 570 nm and OD 620 nm , respectively.

Measurement of enzymatic activity
Catalase activity was measured, using a spectrophotometric H 2 O 2 -degradation assay (CAT100, Sigma-Aldrich). Briefly, the pure cultures of V. parahaemolyticus or VBNC V. parahaemolyticus cells incubated in ASW microcosms at 4°C for 100 days were resuspended in 50 mM potassium phosphate buffer (pH 7) containing 1 g of 3-mm-sized glass bead (Sigma), vortexed for 25 min, and were centrifugated at 15,000 × g for 3 min. The supernatants were separately transferred to sterile microtubes. In a total of 100 μl of volume, 15 μl of the supernatant were mixed with 5 mM H 2 O 2 and incubated at 25°C for 15 min. This reaction ceased by the addition of 900 μl of 15 mM sodium azide. The absorbance was colorimetrically read at 520 nm via a multi-scan Go spectrophotometer (Thermo Scientific Inc.).

Cell hydrophobicity
One ml of V. parahaemolyticus grown overnight in TSB S at 37°C and incubated in ASW microcosms at 4°C for 50 days were centrifugated at 15,000 × g for 3 min, washed in PBS twice, and were re-suspended in PBS to fit an OD of 1.0 (A o ) at 600 nm via a UV-Visible Spectrometer (Multiskan GO Microplate Spectrophotometer, Thermo Scientific). One hundred μl of hexadecane was added to 1 ml of the bacterial solution and was incubated at an ambient temperature for 10 min. ODs of the mixtures in aqueous phase were measured at 600 nm (A 1 ). The degree of hydrophobicity was calculated, following as[1-A 1 /A o ] ×100 (%).

Transmission electron microscopy (TEM)
V. parahaemolyticus cells grown in TSB S at 37°C for 24 h and incubated in ASW and ASW 5 at 4°C for 100 days were centrifugated at 15,000 × g for 3 min, rinsed in 0.1M PBS (pH 7) three times, and were re-suspended in 0.1M PBS, respectively. The cell fluids were pre-fixed in 2% paraformaldehyde overnight at 4°C. The cell solutions were washed in 0.1M PBS, post-fixed in 1% osmium tetroxide, and were serially dehydrated by 30%, 50%, 70%, 95%, and 100% ethanol solutions. Each of them was infiltrated with 2 ml of epoxy resin.
Polymerization of the resins was performed at 60°C for 24 h. The resins were cut (section: approximately 120 nm thickness) and were photographed with a JEOL JEM 1200 EX transmission electron microscope (JEOL USA Inc., Peabody, MA, USA).

Fatty acid composition
Fatty acid analysis was carried out according to the standard protocol provided by the Microbial Identification System (MIDI, Microbial ID Inc., Newark, Del., USA). V. parahaemolyticus was harvested by centrifugation at 15,000 × g for 3 min and was processed by saponification, methylation, and extraction of carboxylic acid derivatives from long-chain aliphatic molecules. The extracted lipids were analyzed by gas chromatography (GC) and identified, using the TSBA6 database of the MIDI system. parahaemolyticus ATCC 33844 and V. parahaemolyticus ATCC 27969 declined slowly in ASW and ASW 5 during the first 28 days at 4°C, but these cells remained culturable until day 42. V. parahaemolyticus strains were also uncultivable in ASW 10 and ASW 30 due to the lack of its culturability on TSA S on day 21. Importantly, the times needed for the complete loss of culturability were lessened with the increasing amounts of NaCl in cold-starvation conditions. While all V. parahaemolyticus strains took ≥2 months in ASW to become the VBNC state, approximately 3-21 days were required to do so in ASW 30 . V. parahaemolyticus ATCC 17802 dropped to the detection limits in ASW, ASW 5 , ASW 10 , and ASW 30 at 4°C for 60, 60, 21, and 4 days, respectively. parahaemolyticus ATCC 17802 remained highly virulent in ASW consistently for killing more than 95% of Caco2 and Vero cell lines (Fig 2). V. parahaemolyticus ATCC 33844 exhibited the decreasing cytotoxic effects to Vero cells at levels of 100%, 63%, 82%, and 39% when maintained in ASW at 4°C for 0, 7, 21, and 80 days. There were gradual decreases in the cytotoxic activities of V. parahaemolyticus ATCC 27969 on the inactivation of Vero cells with the prolonged incubation periods under cold-starvation conditions (data not shown). After resuscitation process, V. parahaemolyticus cells were transformed to a culturable state in TSB S , concomitantly with the recovered cytotoxic activities between 50-100% more than those of these bacteria exposed to cold-starvation until 80 days. In addition, there were no differences in the cytotoxic effects of VBNC V. parahaemolyticus cells with regard to the increasing NaCl contents of the ASW microcosm.

Membrane potential and cellular leakage
Outer membrane permeabilizing activity in VBNC V. parahaemolyticus cells was determined using the NPN assay in Table 1 (Table 2). Initially, DNA and protein leakages of V. parahaemolyticus ATCC 33844 were 0.576-1.562 at 260 nm and 0.466-1.316 at 240 nm, respectively. DNA and protein were found to be released at levels of 2.315-2.683 at 260 nm and 1.218-1.433 at 240 nm from V. parahaemolyticus cells exposed to cold-starvation on day 150, respectively.
Regardless of the bacterial strains used, the cellular leakages from VBNC V.
parahaemolyticus cells were increased with the increasing NaCl concentrations of the ASW microcosms.

Enzymatic activity
After 100 days of cold-starvation, the catalase activity (U/mg) of V. parahaemolyticus was measured as shown in parahaemolyticus to hydrolyze reactive oxygen species (ROS) compounds may be dependent on the bacterial strains used and the length of cold-starvation stress, rather than the different NaCl concentrations. ATCC 17802 were restored to a culturable state, following enrichment in a nutrient-rich medium (TSB S ) at 25°C for 7 days (data not shown). In the recovered cells, the total density of unsaturated fatty acid was increased more than that for the VBNC cells. While lauric acid, 2 -ydroxylauric acid, and unknown ( C14 3OH ) were increased, palmitic acid was decreased remarkably among the total concentrations of saturated fatty acid in the recovered cells. In addition, cis-vaccenic acid was also increased largely, ranging from 20.65% to 32.17% after a stress relief. Interestingly, it was found that some fatty acids, such as 3 -hydroxy-9methyldecanoic acid (C 11 iso 3OH ), cetyl alcohol ( C16 N alcohol ), and cis -11 -palmitoleic acid (C 16:1 w5c ), were synthesized exclusively with the increasing NaCl concentrations during coldstarvation. Cetyl alcohol and cis -11 palmitoleic acid were newly formed when V.

Fatty acid composition
parahaemolyticus ATCC 17082 persisted in ASW microcosms containing ≥5% NaCl at 4°C for 90 days. On the other hand, palmitic acid, (7Z) -13 -methyl-7 -hexadecenoic acid (C 17:1 anteiso ), and cis -Vaccenic acid were increased by the decreasing NaCl amounts of the ASW microcosms. The results indicate that high NaCl concentrations may be an important factor for inducing an alteration in fatty acid profile of V. parahaemolyticus exposed to coldstarvation conditions.

Morphological change
The pure cultures of V. parahaemolyticus ATCC 17802 were filled with lots of granules in cytoplasm and their cell membranes were shown to become intact without minor damages ( Fig 3A). By contrast, VBNC V. parahaemolyticus ATCC 17802 cells had the less organized cytoplasmic layers. Particularly, cell membrane of VBNC V. parahaemolyticus was largely loosened, with the generation of empty gaps between the inner and the outer membranes ( Fig   3B-C). Importantly, V. parahaemolyticus cells acquired the aberrantly-shaped coccal morphologies after the entry into the VBNC state. parahaemolyticus ATCC 33844 grown in tryptic soy broth supplemented with 3% NaCl at 37°C for 24 h (A) or incubated in artificial sea water (ASW; pH 6) (B) and ASW (pH 6) supplemented with 5% NaCl (C) at 4°C for 100 days.

Discussion
After induction of VBNC forms, cell membrane integrity can be measured via epifluorescence microscopy with dual-staining of membrane permeabilizing probes such as SYTO9 and PI (20,21). In this study, 100 days of starvation at 4°C resulted in the inability of V. parahaemolyticus to grow, whereas the cell number with intact membranes was consistently stable over several months of cold-starvation (Fig 1). V. parahaemolyticus was shown to maintain its membrane structure and integrity ranging from 4.8 to 6.5 log CFU/slide during cold-starvation, regardless of the different NaCl contents in ASW microcosms. In particular, V. parahaemolyticus was induced into the VBNC state in ASW microcosms supplemented with higher NaCl contents at 4°C within 21 days and persisted for 150 days in the adverse environments. Although the addition of NaCl in low acidified foods is generally known to preserve and inhibit the growth of spoilage and pathogenic microorganisms, this study may indicate that NaCl can be an important determinant that induces or accelerates the generation of VBNC cells. However, there may be a matter of debate whether V. parahaemolyticus cells would be still alive in microcosms supplemented with >5% NaCl for more than 100 days. V. parahaemolyticus is a moderate halophilic, with optimal growth at 3% NaCl (22), and some strains can grow at 9.6% NaCl (23). VBNC V. cholerae cells within biofilms were recoverable through animal passage challenge even after having been starved at 4°C for more than one year. In this way, resuscitative effects can be one possible explanation for estimating the viability of VBNC cells. After 150 days of persistence in in ASW microcosms (pH 6) at 4°C, VBNC V. parahaemolyticus cells were reverted to a culturable state following temperature upshift in a formulated resuscitation buffer (data not shown). Accordingly, it was found that V. parahaemolyticus was able to enter the VBNC state in low acidified and nutrient-deficient environments containing high NaCl concentrations at 4°C, and VBNC cells retained their membrane structure and integrity under cold-starvation conditions consistently. Until now, many studies were undertaken to investigate phase transition of microorganisms into the VBNC state caused by various environmental stresses, there is still insufficient information to determine whether the complex factors (low temperature, starvation, NaCl, and low pH) trigger the formation of VBNC cells. Thus, further studies are necessary to ensure the accurate and effective identification of VBNC bacteria, as well as their pathogenic potentials.
ROS compounds play an important role on the loss of culturability and formation of VBNC cells (1,2,25,26,27). As aerobic organisms respond to oxidative stress, major cellular components such as polyunsaturated fatty acids and proteins on membrane are directly degraded by ROS compounds. Bacteria may begin to be injured and altered at the essential site of cell membranes when the concentration of active ROS substances increases to a level that exceeds the cell's defense capacity, thereby causing a decrease in membrane fluidity (28,29). In this study, VBNC V. parahaemolyticus ATCC 17802 induced in ASW microcosms (pH 6) at 4°C for 100 days exhibited the increased catalase activities more than those for the actively growing cells (Table 3). In the VBNC forms of V. parahaemolyticus Peculiarly, correlations between cold-starvation and membrane potential of VBNC cells were reported by several studies (13,30). In this study, the increase in fluorescence due to partitioning of NPN uptake into outer membrane was measured by the prolonged incubation of V. parahaemolyticus under cold-starvation conditions. VBNC V. parahaemolyticus ATCC 17802 induced in ASW and ASW 5 at 4°C for 30 days showed the increasing levels of NPN uptake more than those for the pure cultures (Table 1). When Micrococcus luteus was incubated in lactate (0.01%) minimal medium at 4°C, this organism became VBNC after 30 days and exerted a reduction of membrane potential, as evidenced by quantitative flowcytometry with Rhodamine 123 probe that is indicative of viable or non-viable cells (30).
As well-organized in a study of Trevors et al. (14), if bacteria underwent a phase transition to the VBNC state temporarily, membrane became less fluid with an intracellular leakage (K + ) from cytoplasm, supporting our findings. Using radioactive probe  (Table 4). Wong et al.
(2) showed that lauric acid, myristic acid, pentadecanoic acid ( C15 ), and palmitic acid were found to be increased in VBNC V. parahaemolyticus ST550 cells induced in MMS at 4°C for 35 days. Food-isolated strains of V. parahaemolyticus commonly exhibited increased concentrations of decanoic acid ( C10 ), tridecanoic acid ( C13 ), and myristic acid after induction of the VBNC state (10). Gram-negative bacteria typically alter their membrane fluidity with significant changes in the ratio of saturated fatty acid to unsaturated fatty acid, the levels of cyclopropane fatty acid, and cis/trans isomerization in response to external environmental conditions (19). As determined by Chiang et al. (31), who showed that acid-adaptation at pH 5.5 for 90 min increased the ratio of saturated fatty acid/unsaturated fatty acid in V. parahaemolyticus cells, the acidified microcosms used in this study would be linked to the increased concentration of saturated fatty acids in VBNC V. parahaemolyticus cells. An increase in the amount of palmitic acid and stearic acid was shown to be involved in increasing membrane rigidity in V. parahaemolyticus cells (32). Meanwhile, VBNC cells had comparatively increased hydrophobicity (14,33). The increasing membrane rigidity would be closely involved in the maintenance of membrane integrity, thereby making the extraction of DNAs from VBNC cells more difficult.

Conclusion
At the onset of cold-starvation, V. parahaemolyticus used in this study was induced into the VBNC state, while retaining its membrane integrity. The higher the NaCl concentrations, the faster is the shift into the VBNC state. V. parahaemolyticus underwent selected physiological changes, such as modulation of membrane potential, and re-arrangement of fatty acid composition and hydrophobicity that may result in a decrease of cell fluidity, as the cells were induced into the VBNC state during cold-starvation. Theoretically, the physiological modulations may lead to the dwarfing of V. parahaemolyticus cells with the flappy outer membrane out of cytoplasm, thereby minimizing their cell maintenance requirements. V. parahaemolyticus responds to a certain environmental stress such as cold-starvation by inducing its phase transition into a VBNC state.