Effect of food matrix type on growth characteristics and hemolysin production of Vibrio alginolyticus

Background: Vibrio alginolyticus is an important seafood-borne pathogen. There is increasing evidence that V.alginolyticus can also contaminate non-seafood by cross contamination and thereby cause food poisoning in humans. The growth and hemolysin production of V. alginolyticus at 30 °C in briny tilapia, shrimp, scallop, oyster, pork, chicken, freshwater sh and egg fried rice were investigated. Bacterial counts were enumerated by plate counting. Hemolysin production was evaluated by blood agar and hemolytic titer tests. Results: Based on the goodness of t primary model statistics (R 2 , MSE, BF, AF), the modied Gompertz model was a better t to V. alginolyticus growth in foods than the logistic model. Growth kinetic parameters of V. alginolyticus displayed a higher μ max and shorter λ in briny tilapia > shrimp > freshwater sh > egg fried rice > scallop > oyster > chicken > pork. It was notable that the V. alginolyticus counts were similar at the stationary phase, with no signicant growth behavior difference between raw and cooked foods. However, higher thermostable direct hemolysin activity and hemolytic titer were observed in briny tilapia > egg fried rice > shrimp > freshwater sh > chicken > scallop > oyster > pork. Conclusion: V. alginolyticus growth was good in all food matrix types tested. Contrary to current belief, V. alginolyticus displayed a higher hemolytic activity in some non-seafoods (freshwater sh, egg fried rice and chicken) than in scallop or oyster. This is the rst report of growth and toxicity of V. alginolyticus in different food matrices and conrmation that some non-seafood contaminated with V. alginolyticus can be even more pathogenic. This study will enhance the awareness of non-seafood safety and improve the V. alginolyticus risk assessment accuracy.


Background
Vibrio alginolyticus, a Gram-negative, halophilic bacterium, is one of about a dozen of Vibrio species that cause signi cant morbidity and mortality in humans especially in those with low gastric acidity, immunode ciency and liver disease [1,2,3]. Cholera and other Vibrio illness surveillance system data analyses have shown a dramatic increase in the incidence of V. alginolyticus infection, the third most common vibriosis related disease [4]. People come into contact with this bacterium via contaminated water (85 %) or seafood (15 %) [5]. Although foodborne transmission is rare, several cases of food poisoning have been reported in China, Japan and America [6,7].
Foodborne illnesses caused by V. alginolyticus is most common during warmer (≧30 o C) summer and early autumn periods. Gastroenteritis is the most common syndrome with symptoms including fever, nausea, watery diarrhea and abdominal cramps [8,9]. The bacterial growth and hemolysin production of V. alginolyticus contribute to its pathogenicity, causing invasive tissue damage and cytotoxicity, similar to the pathogenesis of other Vibrio species [10,11]. So, if the growth characteristics and hemolysin production of V. alginolyticus in foods could be identi ed, it would help to develop measures to prevent this bacterium multiplying and producing hemolysin, thereby reducing pathogen concentration in foods and the occurrence of V. alginolyticus related food poisoning outbreaks.
V. alginolyticus is widely distributed in nature and has been frequently isolated from a variety of seafood including raw and processed briny tilapia, shrimp, oyster, scallop and food processing environments [12]. The wet environment in restaurant food processing is conducive to V. alginolyticus growth. Therefore, food processing gloves and cooking utensils are potential carriers of V. alginolyticus through contact with contaminated seafood surfaces and transfer of the organism to different types of food matrices, especially raw/cooked non-seafood. There is growing evidence that V. alginolyticus can also contaminate pork, poultry, egg and their products such as egg fried rice by cross-contamination [13,14,15,16]. Thus, there is a need to study the growth characteristics and hemolysin production of V. alginolyticus in seafood and non-seafood so that the risk posed by different food matrices could be accurately assessed.
V. alginolyticus was initially recognized as a biological type of V. parahaemolyticus but later characterized as an independent species [17]. Many studies have described the growth and hemolysin production of V. parahaemolyticus in various food matrices [18,19] but not of V. alginolyticus. To date, most of the V. alginolyticus studies have focused on its prevalence [20], rapid detection [21] and pathogenic mechanisms [22]. To our knowledge, growth and hemolysin production by V. alginolyticus in foods including non-seafood has not been reported.
To better assess the risk of V. alginolyticus in seafood and non-seafood, a clear understanding of the growth characteristics and hemolysin production in varied food matrices is essential. In this study, we quanti ed the growth of pathogenic V. alginolyticus in four seafood (briny tilapia, shrimp, scallop, oyster), three non-seafood (pork, chicken, and freshwater sh) and three cooked food (pork, chicken, egg-fried-rice) types at 30 °C, then the data was used to establish the best growth model by comparing different mathematical equations incorporating with bacterial growth. The thermostable direct hemolysin (TDH) and total hemolytic activity were examined by using blood plate and hemolytic titers.

Results
Growth characteristics analysis of V. alginolyticus in different food types The modi ed Gompertz and logistic models were applied to predict V. alginolyticus counts and to evaluate the model suitability to study the growth pattern in seafood and non-seafood matrices. Performance statistics of these two primary growth models are shown in Table 1, the R 2 values was > 0.98 and MSE value > 0.104 lg CFU/mL in each of the two models. Besides, nine data sets of each food matrix were determined to compare observed values with model predictive values, and the bias factor (BF) and accuracy factor (AF) values calculated by Eqs. (4) and (5) to assess the performance of the two growth models. As seen in Table 1, all the BF and AF values of the modi ed Gompertz model was within the limits of 1.0 ≤ BF ≤ AF ≤ 1.1 but not with the logistic model.
All growth curves were of sigmoidal shape with the initial concentrations smoothly changing to the exponential phase and stabilizing at the stationary phase ( Fig. 1, Table 2). The growth kinetic parameters (Table 3) showed that the V. alginolyticus HY9901 strain initially inoculated at a concentration of 3.24 ± 0.24 lg CFU/g results in a maximum cell number of ~ 7.98 ± 0.53 lg CFU/g, regardless of the of food matrix type. The maximum speci c growth rate (μ max ) of the HY9901 strain changed with food matrix type and varied between 0.76 and 1.62 h -1 and the lag time (λ) varied between 2.27 and 3.43. V. alginolyticus exhibited higher μ max values and lower λ values in briny tilapia > shrimp > freshwater sh > egg fried rice > scallop > oyster > chicken > pork. The μ max values and λ values of V. alginolyticus grown in raw pork and raw chicken were similar to those in cooked pork and cooked chicken.
Hemolytic activity

TDH activity
The TDH activity of V. alginolyticus HY9901 in seafood and non-seafood matrices by measuring the hemolytic circle diameter. The hemolytic zone diameter of briny tilapia was signi cantly higher (p < 0.05) than in freshwater sh > shrimp > chicken > egg fried rice > scallop > oyster > pork matrices' ltrates ( Fig. 2). Besides, the TDH activity of V. alginolyticus in scallop, oyster and pork was lower than in the LB medium and there were no signi cant differences in TDH activity of V. alginolyticus in fresh and cooked chicken and pork.

Hemolytic titer
Following incubation for 24 h, the hemolytic titer of V. alginolyticus in briny tilapia and egg fried rice was >1200 U, signi cantly higher (p < 0.05) than in shrimp > freshwater sh > chicken > scallop > oyster > pork. This result was consistent with the TDH activity of V. alginolyticus in above foods (Fig. 2). The hemolytic titer of V. alginolyticus in most food matrices was higher than (p < 0.05) in scallop, oyster and pork matrices in the LB medium indicating that scallop, oyster and pork may not be conducive for hemolytic activity of V. alginolyticus. There was no signi cant difference in hemolytic titer of V. alginolyticus in fresh and cooked pork and chicken (p > 0.05) (Fig. 3).

Discussion
The incidence of food poisoning caused by V. alginolyticus was positively correlated with temperature and global warming has increased it [23,24]. Risk assessment investigations of V. alginolyticus in food matrices during warm temperature seasons are rare in the literature. Speci cally, the comparison of growth characteristics of V. alginolyticus on seafood and non-seafood and the relative risk posed to consumers has not been reported. In order to explore the growth behavior and hemolysins production of V. alginolyticus in different food matrices during summer and early autumn seasons, the behavior of pathogenic V. alginolyticus HY9901 in terms of changes in cell counts and hemolytic activity in briny tilapia, shrimp, scallop, oyster, freshwater sh, pork, chicken and egg fried rice during incubation time at constant 30 o C were examined.
Results of goodness-of-t primary model (Table 1) showed that the Gompertz model and logistic model met the de ned criteria [25]. Compared with the logistic model, the Gompertz model showed a good statistical t to the observed data and its R 2 values were closer to 1. The MSE values were also low within the precision of microbial enumeration indicating that the modi ed Gompertz model was a better t to V. alginolyticus growth in foods than the logistic model. Besides, the BF and AF values were calculated and also used to assess the performance of the two growth models. The indices bias and accuracy provide an objective indication of model performance. In general, a BF value < 1 is considered unacceptable [26]. However, the BF does not indicate the average accuracy of estimates because under and over predictions tend to cancel out [27]. Therefore, AF should be calculated, which is the sum of the absolute difference between predicted and observed, and describes the overall model error. The higher the AF value, lower is the accuracy of the estimate [28]. As shown in Table 1, the average AF of the modi ed Gompertz and logistic models were1.03 and 1.15. respectively. Results suggest that the predictions were almost identical with observations in the modi ed Gompertz model, and the predicted curves accurately describing the growth of V. alginolyticus in different food matrices at 30 o C. Ma and co-workers [29] showed similar results with the modi ed Gompertz model far more accurate than the logistic model in tting Vibrio growth curves in shrimp and often been used to describe bacterial growth of different foods [30].
V. alginolyticus could grow well in all food matrix types tested and has a similar cell counts at the stationary phase ( Fig. 1). Compared with the growth of other pathogenic organisms in seafood and meat such as Listeria monocytogenes [31], Clostridium botulinum [32] and Salmonella [33], V. alginolyticus exhibited a similar growth capacity. This means that it can survive in our daily food and rapidly proliferate when conditions such as temperature, salinity and also pH become favorable. According to Alam and co-workers [34], the ideal way to inactivate Vibrios in food products is to lower the storage temperature to < 8 °C which would impede the growth of pathogenic and non-pathogenic bacteria and thereby reduce the risk of pathogenic infections. V. alginolyticus exhibited higher μ max values and lower λ values in briny tilapia > shrimp > freshwater sh > egg fried rice > scallop > oyster > chicken > pork (Table 3). V. alginolyticus is ubiquitous in brackish marine waters and in aquatic species. Its intrinsic characteristics have allowed it to adapt easily for growth in aquatic matrices but less in other meat types such as pork and chicken. The slow growth of V. alginolyticus in some non-aquatic foods may be due to the lack of water activity and different pHs, both of which are not suited for pathogenic V. alginolyticus growth in the initial 2 h. However, if egg fried rice is contaminated by other raw seafood, V. alginolyticus can grow quickly in < 2 h and thus increase the risk of human infection.
It is worth noting, the μ max values and λ values of V. alginolyticus grown in raw pork and raw chicken were similar to those in cooked state ( Table 3) which means that the undercooked and cooked state of food has no bearing on the V. alginolyticus growth characteristics.
TDH controls a variety of biological activities including hemolytic activity, and cyto-, entero-and lethal-toxicities [35,36]. TDH activity is the direct cause of the Kanagawa phenomenon in the Wagatsuma agar medium and it has been considered a major virulence determinant of the Vibrio species [37]. Results showed that the hemolytic zone diameter of briny tilapia was signi cantly higher (p < 0.05) than other matrices' ltrates ( Fig. 2) which means that briny tilapia matrix can promote V. alginolyticus to produce more TDH than in other food matrices. Several studies [38,39] have shown that V. alginolyticus induced serious damage to different sh types. The pathogenic V. alginolyticus strain HY9901 used in this study was isolated from diseased sh. We believe that the briny tilapia matrix was most bene cial to the growth (fast growth as shown in Fig. 1) and TDH production of V. alginolyticus and hence people who consume briny tilapia contaminated by V. alginolyticus are more likely to experience a serious food infection.
Interestingly, TDH activity of V. alginolyticus in scallop and oyster was not as high as in freshwater sh, chicken and egg red rice and is most frequently isolated from shell sh [40]. It appears that the scallop and oyster matrices cannot provide su cient nutrients required for TDH production in contrast to some non-seafoods and hence some non-seafood cross-contaminated by V. alginolyticus may pose a higher risk to humans. Therefore, we suggest inclusion of non-seafood products also in V. alginolyticus risk assessment.
The TDH activity of V. alginolyticus in scallop, oyster and pork was lower than in the LB medium which means that these food matrices are not ideal for V. alginolyticus to produce TDH as shown by the growth behavior pro les (Fig.  1). Besides, there were no signi cant differences in TDH activity of V. alginolyticus in fresh and cooked chicken and pork. Hence we believe that the TDH-producing capability of V. alginolyticus is similar in both raw and cooked foods.
Hemolytic titer re ects the total hemolytic activity and the degree of harm caused by V. alginolyticus extracellular products [41]. These extracellular products possess strong phospholipase and/or membrane pore-forming capacity and are cytotoxic to erythrocytes and many other cell types [42]. Results showed that the hemolytic titer of V. alginolyticus in briny tilapia and egg fried rice was signi cantly higher (p < 0.05) than in shrimp > freshwater sh > chicken > scallop > oyster > pork. It 's consistent with the regular of TDH activity of V. alginolyticus in above foods (Fig. 2) and suggests that the briny tilapia and egg fried rice are more conducive than shrimp, chicken, scallop, oyster and pork for V. alginolyticus to produce hemolysis. Because some components of extracellular products are heatstable even after 10 min at 100 °C [43], cooking at a lower temperature while it may kill V. alginolyticus bacteria, it only partly breaks down virulence factors in food products and hence the residual virulence factors may be harmful to human health. This means that when people consume unrefrigerated briny tilapia, egg fried rice, shrimp or freshwater sh contaminated by V. alginolyticus and not heated to at least 100 °C, food poisoning may occur. Seafood products have long been regarded as the only carrier of V. alginolyticus and have been the focus for its toxic effects with less attention paid to non-seafood products. Our research suggests that if some non-seafood (freshwater sh, egg fried rice and chicken) are cross-contaminated by V. alginolyticus, these may also pose a higher risk than some seafood (e.g., scallop, oyster). Therefore, it is necessary to pay attention to non-seafood products also when risk assessment of V. alginolyticus is performed.
It is apparent that V. alginolyticus associated hemolytic titers in different food matrices are determined by the natural components in each food type. We hypothesize that intrinsic factors in egg fried rice promote V. alginolyticus to produce a concentration of the hemolytic factor, related to the thermolabile hemolysin (TLH). Jia and co-workers [44] have shown that the TLH gene from V. alginolyticus shared 94 % identity with the lecithin-dependent hemolysin of V. parahaemolyticus which can damage the cell membranes including that of ounder red blood cells. Thus, we deduce that the lecithin in egg fried rice stimulates the V. alginolyticus TLH gene resulting in an increase in the hemolytic activity. In the human Vibrio infections reported in more recent years, the growth characteristics of V. alginolyticus in non-seafood has been least studied, As shown in our study, an important nding is that V. alginolyticus can grow and show intense virulence in some non-seafood. Wang and co-workers [45,46] reported a higher hemolytic activity of V. parahaemolyticus in egg fried rice than in shrimp or freshwater sh. It appears that V. alginolyticus and V. parahaemolyticus might possess the same virulence and pathogenic toxicity mechanisms. However, further studies are required to determine toxin-producing mechanism(s) and which factor(s) have the most in uence on the production of hemolytic products on different foods.

Conclusions
This study investigated the growth of pathogenic V. alginolyticus in briny tilapia, shrimp, scallop, oyster, freshwater sh, pork, chicken and egg fried rice at 30 °C. Overall, the modi ed Gompertz model well tted the growth characteristics of V. alginolyticus in seafood and non-seafood matrices. V. alginolyticus grew faster in briny tilapia, shrimp and egg-fried-rice with larger μ max values and shorter λ values but all food matrices had a similar cell density of V. alginolyticus at the stable growth phase. Higher TDH activity and hemolytic titers were observed in briny tilapia > egg fried rice > shrimp > freshwater sh > chicken > scallop > oyster > pork. Previous studies initially classi ed V. alginolyticus as a biological type of V. parahaemolyticus. So, the focus has been on the growth and pathogenicity of V. parahaemolyticus in different seafood and not on the growth behavior and hemolysins production of V. alginolyticus in seafood matrices. In addition, the growth in non-seafood products and potential cross contamination was completely ignored. Thus, we may have over the years underestimated the toxicity risk of V. alginolyticus in seafood and non-seafood matrices. In this study, we report on the growth characteristics and hemolytic activity of V. alginolyticus on different food matrices for a more accurate approach to human risk assessment. Because prepared foods may be exposed to a wide temperature range, future research on the growth and hemolytic activity at different temperatures would further improve risk assessment for V. alginolyticus.

Bacterial preparation
Vibrio alginolyticus strain HY9901 [11] (tdh gene positive) was originally isolated from spoiled Lutjnaus erythropterus which cause food poisoning. This strain was kindly provided by the College of Fisheries, Guangdong Ocean University (Zhanjiang, China). The strain was con rmed using PCR by ampli cation of a hypervariable region of the 16S rDNA gene and preserved in Tryptone Soya Broth ( Huangkai, Guangzhou, China) supplemented with 2 % NaCl and 20 % glycerol at -80 °C. The strain was selected on thiosulfate citrate bile salt sucrose (TCBS) agar and grown with agitation at 30 °C for 24 h in Luria-Bertani (LB) medium (Beijing Land Bridge Technology Co., Beijing, China) supplemented with 2 % (w/v) NaCl. The bacterial cells were centrifuged at 2,500 g for 5 min and resuspended in LB medium. The bacterial concentration was con rmed by plate count and adjusted to 10 5 CFU/mL prior to inoculation of different food matrices.

Food matrices preparation and bacterial inoculation
Four raw seafood [shrimp (Litopenaeus vannamei), briny tilapia (Oreochromis mossambicus), scallop (Argopecten irradians), oyster (Crassostrea gigas)] and three raw non-seafood [pork, chicken, freshwater sh (Ctenopharyngodon idellus)] types were obtained from a local supermarket in Zhanjiang, China, and stored at -20 °C. The meat (muscle) from these animals/ sh was used in the study. For cooked foods, egg-fried-rice was prepared by mixing 50 g of egg with 50 g of boiled rice and cooked at 85 °C for 10 min, and the thawed pork and chicken separately added to boiling water and left for 20 min according to the method of Xie and co-workers [47]. Then the cooked egg-fried-rice, pork and chicken were transferred into a biosafety hood and left to cool to room temperature before subsequent treatment.
The number of each food matrix used in the study was eleven (n=11). Test portions, 10 ± 1 g each of raw briny sh, shrimp, scallop, oyster, pork, chicken and freshwater sh, were separately soaked in sterile water containing 100 ppm chlorine at 15 °C, gently shaken for 5 min and washed 10 times with sterile water to inactivate the native bacteria [48]. Salt at 2 % was added to all sterilized raw and cooked food matrices and transferred to sterile Erlenmeyer asks. Next, each sample was inoculated with 1 ml of V. alginolyticus and mixed thoroughly in a vortex mixer (XW-80A, Qilinbei, Haimen, China) for 10 min to ensure uniform distribution of ~10 4 CFU/g V. alginolyticus in the samples.

Model tting
The experimental data obtained at different time points at 30 °C and conditions often used to describe the bacterial growth curves of foods were tted to a modi ed Gompertz model [ Equation 1] and a logistic model using Origin Pro.
9.0. The best tting model was used to describe the growth characteristics of V. alginolyticus on different food matrices.

Evaluation of model performance
In order to evaluate the goodness-of-t of the modi ed Gompertz and logistic models, coe cient of determination (R 2 ), bias factor (BF), accuracy factor (AF) and the mean square error (MSE) were calculated. Goodness-of-t of primary model was evaluated using the adjusted R 2 . The MSE was used to evaluate the difference between the growth data estimated by the model with that measured experimentally, with the MSE values approaching zero indicating a closer t of the data for the model. Besides, validation experiments were carried out to evaluate the models by AF and BF. AF indicates the spread of the results around the predicted values. BF measures the relative average deviation of the predicted and observed V. alginolyticus growth. In the study, the predictions exceeding observed data and < 10% on average in terms of lg (CFU/g), were considered to be accurate. That is, 1.0 < BF < AF < 1.1 was de ned as a satisfactory limit. R 2 , MSE, BF and AF were de ned by the following equations 2-5 [49,50,51,52,53].
In the above equations, N predicted is the predicted bacterial number, N observed the observed bacterial number (lg CFU/g), μ observed the observed speci c growth rate, μ predicted the predicted speci c growth rate (h -1 ), n the number of observations and m the number of parameters of the model.  [56,57,58]. The blank control of the different food matrices were subjected to the same procedure except they were not inoculated with V. alginolyticus. The V. alginolyticus cultured in LB medium was used as the control group.

TDH activity test
Kanagawa phenomenon was tested as previously described by Takeda [59] with slight modi cations. The test plate consisted of 5 % rabbit red blood cells (RBCs), 5 μg/mL chloramphenicol and 0.

Hemolytic titer test
Total hemolytic products in the food samples were measured using a hemolytic titer assay [60]. Brie y, rabbit RBCs were extracted by centrifugation of blood (3,500 rpm for 10 min, Thermo Lynx 6000) three times (washing with PBS each time) and diluted to 2 % with PBS. The ltrate (100 μL) of each food matrix was mixed with 100 μL of PBS

Consent for publication
Not applicable.

Availability of data and material
The datasets used and/or analyzed during the current study are available from the corresponding author on request.

Competing interests
The authors declare that they have no competing interests.

Funding
The design of the study, experimentation, and interpretation of the data was funded by Talent Project of Lingnan Note: R 2 (coefficient of determination), MSE (the mean square error), BF (bias factor) and AF (accuracy factor).   Hemolytic zone diameter of V. alginolyticus HY9901 in different food matrix ltrates. Means ± SE with different lowercase letters are signi cantly different (p < 0.05) among different food matrices. LB medium refers to liquid broth medium.