Occurrence of Vibrio parahaemolyticus, V. cholerae, and V. vulnificus in Norwegian Blue Mussels (Mytilus edulis)

Abstract

Vibrio parahaemolyticus, V. cholerae, and V. vulnificus were isolated from 10.3%, 1.0%, and 0.1% of 885 blue mussel samples, respectively. Four of the samples contained trh⁺ V. parahaemolyticus, while no tdh-positive isolates were detected. The V. cholerae isolates were non-O:1/non-O:139 serotypes and were ctxA negative.

Vibrio parahaemolyticus, V. cholerae, and V. vulnificus are the principal Vibrio species linked to seafood-borne infections. These bacteria have mainly been associated with temperate regions, but outbreaks caused by V. parahaemolyticus in local seafood have also recently been recognized in colder parts of the world like Alaska (15) and Chile (10). Diagnosed clinical cases in Norway associated with food-borne Vibrio spp. are rare. However, there have been cases in Norway, and also in Denmark, where local seafood has been suspected as the source of infection (5, 19). V. vulnificus wound infections caused by seawater or seafood handling have been documented in Sweden and Denmark (4, 16, 17).

There is limited knowledge regarding the presence of V. parahaemolyticus, V. cholerae, and V. vulnificus in the Scandinavian coastal environment. The few studies conducted have, however, revealed the presence of V. parahaemolyticus and V. vulnificus (3, 9, 14).

During the last decades, the Norwegian shellfish industry has experienced a considerable expansion. As a hazard identification step in a risk assessment perspective for pathogenic vibrios in Norwegian bivalve molluscs, the Norwegian Food Safety Authorities (NFSA) funded a 2-year monitoring program covering commercially grown blue mussels. The aims were to assess the occurrence of V. parahaemolyticus, V. cholerae, and V. vulnificus in Norwegian blue mussels and the presence of the virulence genes tdh and trh in V. parahaemolyticus and ctxA in V. cholerae.

Samples were collected from July 2002 to September 2004 from a total of 102 production sites authorized by the NFSA, except for five sites that were wild-growing mussel sites along the South coast. These were included to provide a sufficient number of sites and samples from this region, which has higher seawater temperatures. Sampling was evenly spread throughout the year, except for the South coast, where a larger part of the samples was collected during the summer season. The localities were categorized into four different regions: North coast (n = 19), Midcoast (n = 25), West coast (n = 45), and South coast (n = 13) (Fig. 1A). The number of sampling sites and frequency differed during the project period because production activity among producers changed. One to 33 (median, 6) samples were collected from each locality. The samples were sent with cooling elements by overnight express mail to the laboratory. The surface water temperature was recorded at the sample site at the time of collection.

FIG. 1.

(A) Map showing the four different regions from which samples were collected. Sample sites positive for V. parahaemolyticus (V.p.), V. cholerae (V.c.), and V. vulnificus (V.v.) are indicated. (B) Locations along the South coast where the four trh⁺ V. parahaemolyticus isolates were isolated.

Isolation of Vibrio spp. was performed according to Nordic Committee on Food Analysis method no. 156 (17a), which is the method required for monitoring on behalf of the NFSA. Qualitative analysis used enrichment in alkaline peptone water (Sifin, Germany) with 2% NaCl and in salt polymyxin broth (Sifin, Germany) with 25 IU ml⁻¹ polymyxin (Oxoid, United Kingdom) at 41.5 ± 0.5°C for 18 ± 2 h before inoculation onto thiosulfate-citrate-bile-sucrose agar (Oxoid) and incubation at 37 ± 1°C for 24 ± 2 h. Fifty of the samples were incubated at both 41.5 ± 0.5°C and 37.0 ± 0.5°C for 18 ± 2 h and were incubated only in alkaline peptone water. For quantitative analysis, thiosulfate-citrate-bile-sucrose agar plates were inoculated with 100 μl of the sample in alkaline peptone water (1:10) and incubated at 37°C ± 1°C for 48 ± 4 h. Identification of colonies was performed as recommended by the Nordic Committee on Food Analysis and Api 20E. V. cholerae isolates were serotyped by the National Institute of Public Health.

Linear regression using GLM in Stata (Stata Corp., College Station, TX) was applied to look at the region and month/year with respect to differences in water temperatures. Further statistical analysis was undertaken using only data from the South coast and West coast. The relationship between V. parahaemolyticus and predictor variables (region, season, and temperature) was investigated using logistic regression with the logit procedure in Stata 8.0/SE for Windows (Stata Corp., College Station, TX). The model was built using the forward selection procedure. Model fit was assessed using the standard routines as delta-beta plots and the Hosmer-Lemeshow test. The results from a logistic model can be used to calculate the expected probability of finding V. parahaemolyticus given a certain set of levels of the explanatory variables (13).

At least one V. parahaemolyticus isolate from each positive sample was tested for tdh (n = 135) by a colony hybridization method (7) and for trh (n = 138) by PCR (1). At least one colony of V. cholerae (n = 12) from each sample was examined by PCR for the presence of ctxA (8). PCR was performed using Dynazyme II polymerase (Finnzyme, Finland) and a Bio-Rad Ismart cycler. All primers were custom synthesized by Medprobe (Denmark).

V. parahaemolyticus was detected in 91 (10.3%) of the 885 blue mussel samples. The numbers were estimated to be 1,800 CFU g⁻¹ and 200 CFU g⁻¹ in two samples, and the remaining samples contained <100 CFU g⁻¹. Vibrio-positive locations are indicated in Fig. 1A, and distribution of positive samples with respect to month and year is shown in Fig. 2. All except one (from the Midcoast) of the V. parahaemolyticus-positive samples were collected from the South coast (69/89) and West coast (21/352).

FIG. 2.

Total number of samples and total number of V. parahaemolyticus (V.p.)-positive samples collected from July (J) 2002 to October (O) 2004. (A) All coastal regions (n = 885 samples); (B) West coast samples (n = 375); (C) South coast samples (n = 158). —, number of V. parahaemolyticus-positive isolates; - - - - , total number of samples.

No V. parahaemolyticus isolates harbored tdh, but four isolates carried trh. These isolates were also urease positive. The isolates represented three samples from one location at different sampling times and one sample from another location (Fig. 1B). To our knowledge, this is the first time trh⁺ V. parahaemolyticus has been detected in North Europe, but it has previously been detected in France (12, 18). The frequency of tdh⁺ and trh⁺ V. parahaemolyticus in the environment appears to be very low (2, 6, 7, 11).

The lowest water temperature at a location where a V. parahaemolyticus-positive sample was collected was 0.6°C. In both samples where ≥100 CFU g⁻¹ V. parahaemolyticus were detected, the temperature was above 19°C. Statistical analysis indicated that the occurrence of V. parahaemolyticus could be predicted using the following variables: region, season, and water temperature (Table 1). The main finding from the model was a steep increase in the probability of finding V. parahaemolyticus when the temperature increased at the South coast but not at the West coast. The data from 2003 and 2004 suggest that for the South coast, more than 50% of the samples would be V. parahaemolyticus positive if the water temperatures rose above 16°C.

TABLE 1.

Results from the logistic regression analysis of data from the South and West coasts^a

Variable	Coefficient	95% confidence interval
Yr (2002)
Yr (2003)	−3.16	−4.36 to −1.97
Yr (2004)	−3.35	−4.67 to −2.04
South coast
West coast	−3.80	−4.92 to −2.69
Temp	0.26	0.18 to 0.34
Constant	−0.90	−2.34 to 0.55

^aData are presented as regression coefficient with 95% confidence interval. The model classified 88.4% of isolates correctly, and no signs of lack of fit were found.

V. cholerae was detected in nine (1%) of the samples. All isolates were non-O:1/non-O:139 serotypes and negative for ctxA by PCR. The lowest water temperature for isolation of V. cholerae was 2°C. V. vulnificus was only detected in one South coast sample by direct plating. These results indicate that the enrichment conditions used in the present study were too stringent to allow V. vulnificus detection. The predominant Vibrio sp. detected was V. alginolyticus (24.5%).

The water temperature along the South coast was found to be significantly higher (P < 0.001) than temperatures along the West coast, Midcoast, and North coast (Table 2). No statistical difference was observed between the West coast and Midcoast. The North coast had significantly cooler temperatures than all the other regions (P < 0.001). Low temperatures are probably the most important explanation for the low frequency of pathogenic Vibrio spp., and differences in temperatures are also reflected by differences in findings among regions.

TABLE 2.

Results from the linear regression analysis of the relationship between water temperature and region, adjusted for year (not shown) and month (not shown)

Region	Avg temp difference	95% confidence interval
South coast
West coast	−0.67	−0.10 to −1.23
Midcoast	−1.63	−1.04 to −2.21
North coast	−3.16	−2.54 to −3.79

The low isolation rate of Vibrio spp. could, however, also have been affected by the transport and isolation method. The samples had to be cooled for an extended period, and the enrichment temperature for most samples was 41.5°C. However, 13 of the 50 samples enriched at 37°C and 41.5°C were positive for V. parahaemolyticus at both temperatures, and seven and eight samples were positive at only one of the temperatures, respectively, suggesting no difference in the isolation rate at the two temperatures for this species.

The results indicate that potentially pathogenic Vibrio spp. do not pose a considerable hazard to blue mussel consumers in Norway. However, this might change during prolonged periods of high water temperatures.