Environmental Investigations of Vibrio parahaemolyticus in Oysters after Outbreaks in Washington, Texas, and New York (1997 and 1998) (2024)

Abstract

Total Vibrio parahaemolyticus densities and the occurrence of pathogenic strains in shellfish were determined following outbreaks in Washington, Texas, and New York. Recently developed nonradioactive DNA probes were utilized for the first time for direct enumeration of V. parahaemolyticus in environmental shellfish samples. V. parahaemolyticus was prevalent in oysters from Puget Sound, Wash.; Galveston Bay, Tex.; and Long Island Sound, N.Y., in the weeks following shellfish-associated outbreaks linked to these areas. However, only two samples (one each from Washington and Texas) were found to harbor total V. parahaemolyticus densities exceeding the level of concern of 10,000 g⁻¹. Pathogenic strains, defined as those hybridizing with tdh and/or trh probes, were detected in a few samples, mostly Puget Sound oysters, and at low densities (usually <10 g⁻¹). Intensive sampling in Galveston Bay demonstrated relatively constant water temperature (27.8 to 31.7°C) and V. parahaemolyticus levels (100 to 1,000 g⁻¹) during the summer. Salinity varied from 14.9 to 29.3 ppt. A slight but significant (P < 0.05) negative correlation (−0.25) was observed between V. parahaemolyticus density and salinity. Based on our data, findings of more than 10,000 g⁻¹ total V. parahaemolyticus or >10 g⁻¹ tdh- and/or trh-positive V. parahaemolyticus in environmental oysters should be considered extraordinary.

Vibrio parahaemolyticus is a gram-negative halophilic bacterium distributed in temperate and tropical coastal waters throughout the world and is a leading cause of foodborne gastroenteritis (15). Until recently, U.S. illnesses were limited to sporadic cases associated with consumption of raw shellfish (12, 13) or with small outbreaks due to recontamination of cooked or processed seafood (3).

During the summer of 1997, the first confirmed oyster-associated outbreak caused by V. parahaemolyticus in the United States, as defined by the National Shellfish Sanitation Program, occurred in the Pacific Northwest (5). During this outbreak, 209 culture-confirmed cases were reported in North America, and nearly all were associated with shellfish from Washington and British Columbia. Multiple serotypes of V. parahaemolyticus were isolated from the stools of ill persons. A smaller oyster-associated outbreak (43 culture-confirmed cases) occurred in Washington in 1998 (Ned Therien, Washington State Department of Health, personal communication, February 1999 and February 2000).

The largest V. parahaemolyticus outbreak reported in the United States (416 cases, 98 culture confirmed) was linked to consumption of raw oysters from Galveston Bay, Tex. (7; S. S. Barth, L. S. Del Rosario, T. Baldwin, M. Kingsley, V. Headley, B. Ray, K. Wiles, A. DePaola, D. Cook, C. Kaysner, N. Puhr, N. Daniels, L. Kornstein, and M. Nishibuchi, Abstr. 99th Gen. Meet. Am. Soc. Microbiol., abstr. C-57, 1999). This outbreak lasted from May to July 1998 and was distinguished by an extremely high attack rate and by the fact that all clinical isolates belonged to a single clone of the O3:K6 serotype. This clone apparently emerged in India around 1995, becoming endemic in much of Asia; it is the most prevalent strain associated with V. parahaemolyticus illness in Asia (2, 22). It appears that this O3:K6 clone has become pandemic, and there is concern that this may increase the risk of V. parahaemolyticus infections from consumption of U.S. shellfish.

V. parahaemolyticus O3:K6 was subsequently linked to a small outbreak of eight V. parahaemolyticus cases associated with shellfish harvested from Oyster Bay off New York's Long Island Sound from July to September 1998 (6). In each outbreak, state health officials closed affected areas to shellfish harvesting and requested the assistance of the U.S. Food and Drug Administration (FDA) with monitoring of shellfish for abundance of V. parahaemolyticus. Areas in Galveston Bay remained closed until November based on historical seasonal epidemiological data.

The distribution of V. parahaemolyticus in the environment and foods has been studied extensively in Japan (19, 27) and to a lesser degree in the United States (9, 16, 17, 18, 25). It is found in Pacific, Gulf, and Atlantic coastal waters and fauna, but there are few quantitative data on seasonal or geographical distribution. A nationwide survey of shellfish and overlying waters was conducted in 1983, but sampling frequency was only once per season for 1 year (9). Levels in shellfish were found to be 200-fold higher than in overlying waters; the highest densities were observed in the late spring and early summer. Water temperature was positively correlated with V. parahaemolyticus abundance, but no clear relation was observed with salinity or fecal coliform levels. The amount of sampling was limited by the available methodology. The most-probable-number (MPN) procedure, which relies on biochemical identification of suspect isolates (11), is laborious and expensive.

Pathogenic strains of V. parahaemolyticus generally produce a thermostable direct hemolysin (TDH) that is associated with the Kanagawa phenomenon (K⁺) and/or a TDH-related hemolysin (TRH) (14). The genes tdh and trh code for TDH and TRH, respectively; the tdh gene has been used as the target of DNA probes (9). One or both of these genes are detected in most clinical strains of V. parahaemolyticus but are uncommon in environmental and food isolates (9, 17, 27). All strains of V. parahaemolyticus produce a thermolabile direct hemolysin, which reportedly is species specific (24). Recently, alkaline phosphatase- and digoxigenin-labeled oligonucleotide probes for detection of tlh were evaluated, and their results were shown to be in agreement with standard biochemical identification assays (20). Replacement of biochemical tests for bacterial identification with DNA probe hybridization substantially reduces the time and labor required for sample analysis. Vibrio vulnificus enumeration by DNA probe identification of colony lifts from direct plating of oyster hom*ogenates was equivalent to MPN analysis and more rapid and precise (10), but this approach has not been reported for V. parahaemolyticus.

This paper reports total V. parahaemolyticus densities and the occurrence of pathogenic strains in shellfish following outbreaks in Washington, Texas, and New York. Recently developed nonradioactive DNA probes were utilized for the first time for direct enumeration of V. parahaemolyticus in environmental shellfish samples.

MATERIALS AND METHODS

RESULTS

V. parahaemolyticus was recovered from most (77%) of the Pacific Northwest samples during the late summers of 1997 and 1998 (Table ​(Table1).1). Densities varied considerably (<3 to 46,000 g⁻¹) in Washington oysters over a short period from 20 August to 3 September 1997. Densities ranged from 29 to 2,300 g⁻¹ on 26 August 1997 in different areas of Quilcene Bay. Similar variation was seen at the Twanoh State Park sampling site in Hood Canal from one week to the next. Strains with trh and/or tdh were detected at densities of <10 g⁻¹ in 15% of the 1997 samples but were not detected in 1998 samples. All strains that were tdh positive were also K⁺, trh positive, and urease positive; one strain was trh positive and tdh negative. Several O serotypes serotypes (1, 4, and 5) were found among these potentially pathogenic strains.

V. parahaemolyticus was detected in all 106 Texas oyster samples, and counts ranged from 40 to 23,000 g⁻¹. V. parahaemolyticus densities in most samples ranged between 100 and 1,000 g⁻¹ (15.1% were >1,000 g⁻¹ and 9.5% were <100 g⁻¹). Figure ​Figure11 indicates little weekly variation in V. parahaemolyticus levels in Texas oysters except for 29 June 1998, when a density of 23,000 g⁻¹ was found with one sample; this was the only sample from Texas that exceeded 10,000 g⁻¹. Mean V. parahaemolyticus densities at the three to five sites sampled throughout the study were similar to those observed with the expanded sampling set (20 to 30 sites) in August 1998. Greater variation was observed with the MPN analysis than with the direct plating procedure.

Table ​Table22 shows V. parahaemolyticus levels in oysters and salinity at 30 Galveston Bay leases from 17 August to 8 September 1999. Approximately a one-log range (2.07 to 3.16) in mean (log₁₀) V. parahaemolyticus densities was observed among the sites. Salinity ranged from 14.9 to 29.3 ppt. Water temperature varied little (27.8 to 31.7°C) and is not shown. Higher V. parahaemolyticus densities and lower salinities were found in the Smith Point and East Bay areas than in the Ship Channel and Hanna Reef areas. A slight but significant (P < 0.05) negative correlation (−0.25) was observed between salinity and log₁₀ V. parahaemolyticus density. We recovered one tdh-positive isolate, serotype O4:K-untypeable. Two of the three samples collected on 21 September 1998 yielded colonies hybridizing with the tdh DNA probe, but no isolates were cultured for subtyping.

Table ​Table33 lists densities of V. parahaemolyticus in Oyster Bay, N.Y., oysters and method variability. Densities ranged from <10 to 120 g⁻¹ in oysters from various sites in Oyster Bay. There was no apparent trend in V. parahaemolyticus densities between sites or collection dates and little variation in temperature (17.4 to 17.8°C) or salinity (25.5 to 26.2 ppt) of the harvest waters. Strains hybridizing with the tdh probe were not detected. Total variance (total V. parahaemolyticus density determined by direct plating and identification by the alkaline phosphatase-labeled tlh probe) from sample to sample was 0.13 (log₁₀), while method error was minimal (0.06).

DISCUSSION

This paper reports V. parahaemolyticus levels in oysters following four shellfish outbreaks and provides the most extensive quantitative data for V. parahaemolyticus in U.S. shellfish to date. V. parahaemolyticus was prevalent in oysters from Puget Sound, Wash.; Galveston Bay, Tex.; and Long Island Sound, N.Y., in the weeks following shellfish-associated outbreaks linked to these areas. V. parahaemolyticus densities exceeded the FDA level of concern of 10,000 g⁻¹ (26) in two samples, one each from Washington and Texas. Pathogenic strains (those hybridizing with the tdh and/or trh probe) were detected in a few samples, mostly Puget Sound oysters, and at low densities (usually <10 g⁻¹). The isolation of K⁺ strains from incriminated food or the environment associated with V. parahaemolyticus outbreaks has not been reported previously in the United States (3).

While much of the variation in V. parahaemolyticus levels in the present study may be attributed to seasonal and regional differences, the findings may also have been influenced by temporal or spatial proximity of samples to incriminated shellfish, sampling protocols, and bacteriological procedures for each outbreak. With the exception of Washington State, V. parahaemolyticus surveillance was not initiated until shellfish harvesting areas were closed. Washington State shellfish were linked to 68 illnesses from May to September 1997, with the peak period for onset of cases from 10 to 23 August (35 reported cases). Environmental monitoring began on 20 August 1997, prior to closure of oyster harvesting areas on 28 to 29 August 1997 by the Washington Department of Health; monitoring extended through 7 September 1997 (T. Sample and C. Swanson, Vibrio parahaemolyticus Workshop, U.S. Food and Drug Administration, 1997). While many of the samples were collected after closure of the harvest areas, the V. parahaemolyticus levels in these samples were similar to those in samples collected prior to closure near the peak of the outbreak. V. parahaemolyticus densities also varied more in Puget Sound shellfish than in those from Galveston Bay and Long Island Sound. Some shellfish-growing areas in the Puget Sound are exposed during low tide, which may elevate the temperature of emerged oysters, because air temperatures are typically much warmer than water temperatures in the late summer. Since V. parahaemolyticus growth is favored by warmer temperatures, it would probably multiply more rapidly in emerged oysters than in those submerged. In many instances, Washington and Oregon oyster samples were collected from harvesters at the dock or from retail markets; water temperature, salinity, water depth, precise harvest time, and postharvest handling data were not available, but these factors may influence V. parahaemolyticus levels and variability. Results from our laboratory indicate that V. parahaemolyticus densities can increase 100-fold in live oysters within 10 h of harvest at 26°C (J. A. Gooch, A. DePaola, C. A. Kaysner, and D. L. Marshall, Abstr. 99th Gen. Meet. Am. Soc. Microbiol., abstr. P-52, 1999). MPN analysis was used for enumeration of V. parahaemolyticus in 1997 Washington samples and has been shown to be much less precise than direct plating for enumeration of V. vulnificus from oysters (10).

Levels of K⁺ V. parahaemolyticus in this study were well below the infectious dose (10⁵) observed in feeding experiments with human volunteers (23). The V. parahaemolyticus clinical strains from the Pacific Northwest are nearly always urease positive and usually have both the tdh and trh virulence genes. These strains may be more virulent than those used in early feeding experiments with human volunteers (23), described as K⁺ without information on urease and trh. The low densities (<10 g⁻¹) in environmental and market oysters suggest that illness occurs in some individuals at doses well below the 10⁵ to 10⁷ range observed in feeding trials (23). The contribution of postharvest temperature abuse in this outbreak was not determined and requires further study.

V. parahaemolyticus levels in Galveston Bay oysters were higher and less variable than those in Washington State oysters; the higher counts were probably due to warmer Gulf waters. The lower variability was attributed to nearly constant water temperature during the study and to careful sample handling procedures from harvest to analysis, as these samples were collected in the field and cooled immediately rather than collected from the market as in Washington. Unpublished results from our laboratory indicate that the cooling and shipping protocol used with the Texas and New York samples does not affect V. parahaemolyticus numbers in oysters from harvest to analysis within 24 h. Texas samples were analyzed by direct plating methods instead of MPN, further reducing variability. The only count that exceeded 10,000 g⁻¹ was for a sample collected on 29 June 1998; relatively high densities from this harvest date were observed in two other samples (930 and 4,300 g⁻¹). The higher counts in these samples could reflect their closer temporal proximity to the time of incriminated-oyster harvest. Only three samples yielded colonies that hybridized with the tdh probe. Isolates were not cultured from two of the samples. A single isolate from the third sample was not the O3:K6 serotype associated with clinical samples. Harvest for raw consumption resumed in November 1998; no new cases were reported.

Oysters and clams from the Long Island Sound were incriminated in 16 V. parahaemolyticus cases from 21 July to 27 September 1998; O3:K6 was the predominant serotype (6). On 10 September 1998, the New York State Department of Environmental Conservation closed Oyster Bay to harvesting of shellfish. Samples from New York were collected on 12 October (8 samples) and 14 October (7 samples) 1998, more than 1 month after closure. The low V. parahaemolyticus densities (generally <100 g⁻¹) in these samples may reflect the cooling water temperatures (17°C). To increase sensitivity to 1 g⁻¹, 10 plates were spread with replicate 0.1-g portions of oyster hom*ogenate from each sample. While no O3:K6 strains were detected, the low method variance with the New York samples did demonstrate the precision of the direct plating procedure.

This study demonstrates the abundance of V. parahaemolyticus in U.S. oysters following outbreaks on the Pacific, Gulf, and Atlantic coasts. Based on our data, findings of more than 10,000 total V. parahaemolyticus or >10 tdh- and/or trh-positive V. parahaemolyticus per g in environmental oysters should be considered extraordinary. Monitoring total and pathogenic V. parahaemolyticus levels during warmer periods of the year could provide valuable information on conditions leading to an outbreak and may be useful in forecasting outbreaks or developing reliable criteria for reopening shellfish beds after an outbreak. Analysis of incriminated market shellfish would help resolve questions regarding the possibility of increased strain virulence and the importance for risk assessment of postharvest multiplication.

ACKNOWLEDGMENTS

We thank the Washington State Health Department, the Texas Department of Health, and the New York State Department of Environmental Conservation for collection and shipment of oyster samples. We also appreciate the analytical assistance of the many FDA microbiologists at the regional laboratories in Bothell, Wash.; Denver, Colo.; and Atlanta, Ga.

REFERENCES