Correspondence: Bruno Gomez-Gil, CIAD, A.C. Mazatlán Unit for Aquaculture and Environmental Management, AP. 711, Mazatlán, Sinaloa, México 82000. Tel.: +52 669 9898700; fax: +52 669 9898701; e-mail: email@example.com
During 2003 and during late September of 2004, more than 1230 cases of gastroenteritis were reported in the south of Sinaloa State, north-western Mexico. All cases were attributed to the consumption of raw or undercooked shrimp collected at the Huizache-Caimanero lagunary system. Vibrio parahaemolyticus was identified by standard biochemical methods, and many strains were positive for PCR amplifications of the tlh and tdh genes and negative for the trh gene. A representative strain belonged to the O3 : K6 serogroup. This is the first outbreak of gastroenteritis caused by the pandemic strains of O3 : K6 V. parahaemolyticus in México.
Strains of Vibrio parahaemolyticus are responsible for gastroenteritis, usually associated with consumption of raw or undercooked seafood in many parts of the world. Oysters are perhaps the main source of infection in North America, due probably to high water temperatures and low salinities (DePaola et al., 2003a).
Clinical signs of infection include gastroenteritis with nausea, watery diarrhea, vomiting, abdominal cramps, low-grade fever, and chills and, rarely, bloody diarrhea (Farmer et al., 2003). The disease is mild and self-limiting, but can be fatal, especially in immuno-compromised patients; onset of the symptoms occurs from a few to 96 h after consumption of infected material (Wong et al., 2000), and can last for three days. There are many strains of V. parahaemolyticus, but only those that produce a thermostable direct hemolysin (TDH) and/or the TDH-related hemolysin (TRH) have the ability to cause gastroenteritis (Nishibuchi et al., 1992), and almost all the strains isolated from clinical samples have either or both genes (tdh and trh) which encode the respective hemolysins. Since 1996 a new serotype (O3 : K6) appeared in Calcutta, India, and rapidly spread to many countries, becoming the first pandemic strain of V. parahaemolyticus (Okuda et al., 1997; Matsumoto et al., 2000; Wong et al., 2000). This strain is positive for the tlh and tdh genes but negative for the trh gene, and also has a genetic marker, the open reading frame 8 (ORF8; Myers et al., 2003) that differentiates it from other pathogenic strains (Nasu et al., 2000).
In Mexico there have been no reports of outbreaks due to V. parahaemolyticus, although isolated cases have been attributed to this bacterium, mainly based on the clinical symptoms. Few environmental tdh+ strains have been isolated from water and fish in Mexico (Cabrera-Garcia et al., 2004), but strains in clinical specimens have not been examined. In the south of Sinaloa State, north-western Mexico, diarrhea cases have occurred sporadically for years during the summer. The symptoms of the patients resembled those of the V. parahaemolyticus infection. This disease is called locally ‘el gaste’, which might be translated from Spanish as ‘the waste’, because of the debilitating nature of the disease. During late September and early October 2003, fewer than 100 cases of diarrhea were reported in local clinical facilities of Mazatlan and El Rosario counties in this area. Almost no stools were sampled for bacterial growth in TCBS agar. In the same period in 2004, more than 1220 cases were recorded in the surrounding area, and probably more in other states where aquatic products were commercialized.
Materials and methods
Samples were taken from stools, shrimp, water, and sediments in October 2003, between September and October 2004, and in March 2005. All samples were processed as recommended by DePaola & Kaysner (2004).
Biochemical tests for identification of Vibrio species were carried out following standard procedures. The strains exhibiting the following characteristics were identified as V. parahaemolyticus: gram-negative rods; oxidase positive; arginine dihydrolase negative; ornithine and lysine decarboxylase positive; growth at 8% NaCl but not at 0%; negative for sucrose, lactose, ONPG, urease, and Voges-Proskauer; positive for arabinose, D-mannitol, D-mannose, indole, gelatinase, and citrate; sensitive to the vibriostatic agent O/129 (Alsina & Blanch, 1994).
DNA was extracted by boiling suspicious TCBS colonies (round green with darker-bluish centers) from the samples in lysis buffer (KCl 50 mM, Tris pH 8.5 10 mM, MgCl2 1.5 mM, gelatin 0.01%, IGPAL 0.45%, and Tween-20 0.45%) at 95°C for 5 min. 153 of the strains were further purified in TCBS and TSA, cryopreserved at −70°C (8), and deposited at the Collection of Aquatic Important Microorganisms (CAIM, http://www.ciad.mx/caim). The DNA of the preserved strains was extracted with a DNA extraction kit (Wizard DNA Genomic DNA Purification System, Promega Corp.). PCR amplification of the tlh, tdh, and trh genes was carried out according to Bej et al. (1999). Vibrio parahaemolyticus type strain CAIM 320T was used as positive control for the tlh gene and negative control for the open reading frame 8 (ORF8), and DNA of strain CNRVC010089 was used as positive control for tlh and trh genes; DNA of strains JKYVp6 and CAIM 729 were used as positive controls for ORF8 (Myers et al., 2003), and also for tlh and tdh genes. PCR products were electrophoresed in a 2.0% agarose gel, stained with ethidium bromide, visualized under UV light, and the image was captured with a digital camera (Epi-Chemi II, UVP).
rep-PCR genomic fingerprinting
Representative strains (53) were fingerprinted with rep-PCR (GTG5 primer; Gomez-Gil et al., 2004). Briefly, DNA was amplified with Taq DNA polymerase (Promega) as described earlier (Wong et al., 2001). Resulting amplicons were resolved in a 1.2 % agarose gel in 1X TAE buffer at 90 V for 140 min, and analyzed with the GelCompar II (ver. 4.5, Applied-Maths, Belgium) software as described by Gomez-Gil et al. (2004).
The test strains were grown and their O : K serovars were determined using specific antisera as described previously (Suthienkul et al., 1995).
A total of 1225 patients reported as having consumed raw or undercooked shrimp taken from the Huizache and Caimanero lagoons (Fig. 1). Cases peaked during or immediately after the weekend when people went to this location to consume these products. Clinical symptoms were watery diarrhea that started three to four hours after consumption of the shrimp, and abdominal cramps, the main reason for which people sought medical help. Other symptoms included chills, headaches, and nausea. The symptoms lasted for 2–3 days and no medication was necessary apart from painkillers; treatment included rehydration.
Only one V. parahaemolyticus strain (CAIM 740) was isolated from stools from the 2003 episode. This strain was identified by the presence of the tlh gene, a molecular marker for the species, and by comparing its rep-PCR pattern. From the 2004 outbreak, 288 Vibrio spp. isolates were collected from stools (141), shrimp (69), seawater (55), and sediments (23). Two more isolates were obtained from shrimps during March 2005.
Of the 288 isolates obtained during the 2004 outbreak, only 162 isolates were identified biochemically; of these, 103 were identified as V. parahaemolyticus (20 were from humans, 32 from seawater, 40 from shrimp, and 11 from sediments). 37 isolates were identified as V. cholerae non O1, 20 as V. fluvialis, and 2 isolates could not be identified.
132 out of the 141 strains isolated from human feces of the 2004 outbreak and the one from the 2003 outbreak were analyzed for the presence of the tlh, tdh, and trh genes. 103 strains (78.0 %), including the one isolated in 2003, were positive for the tlh and tdh genes but negative for the trh gene. The rest were tlh positive (tlh+) but tdh and trh negative (tdh−, trh−). All the stools analyzed had at least a tlh+ and tdh+ strain. One strain isolated from water (CAIM 1397) was tlh+, tdh+, and trh−. Two strains (CAIM 1374 and CAIM 1376) isolated from shrimp were tlh+, tdh+, and trh−. The two isolates from shrimp obtained during 2005 were tlh+, tdh+, and trh+.
A representative strain belonging to the most common rep-PCR pattern (CAIM 1400, Fig. 2 cluster a) and isolated from stools, had the O3 : K6 serotype. One of the two shrimp isolates (CAIM 1772) from 2005 was O5 : K17. Strains CAIM 1400, CAIM 1378 (isolated in October 2004), CAIM 1361 (isolated on 30 September 2004), CAIM 740 (isolated in 2003), and the pandemic reference strain CAIM 729 and DNA of JKYVp6 showed positive results for the presence of the ORF8 of phage f237, a molecular marker of the pandemic O3 : K6 strain (Nasu et al., 2000; Myers et al., 2003). A negative ORF8 result was obtained for strain CAIM 727 (O3 : K5), not a pandemic isolate.
rep-PCR fingerprinting showed that the majority of the isolates from human feces analyzed with this technique had an identical pattern to the pandemic strains (O3 : K6) CAIM 729 (TX 2103), CAIM 1400, and JKYVp6 (Fig. 2 cluster a). Another group of isolates (Fig. 2 cluster b) share an identical pattern among them but different from the pandemic cluster; two of these isolates were obtained from shrimps (CAIM 1374 and CAIM 1376) and the other four from human feces. All were tlh+, tdh+, and trh−; the shrimp isolates were collected at different locations and days after the human isolates (shrimp isolates 3rd Oct. 2004 and human isolates the 30th Sept. 2004). The two shrimp isolates from 2005 were identical between them (clones) but different from all others (Fig. 2).
The absence of the trh gene, along with urease negative reaction, presence of ORF8, identical rep-PCR pattern with pandemic strains (CAIM 729 and JKYVp6, Fig. 2), and O3 : K6 serotype (CAIM 1400) lead us to believe that this outbreak was caused by the V. parahaemolyticus pandemic strain (Okuda et al., 1997; Matsumoto et al., 2000; DePaola et al., 2003b; Myers et al., 2003) that has affected many countries in Asia and the Americas (DePaola et al., 2003c; Gonzalez-Escalona et al., 2005). Although no pandemic strain was isolated from shrimp, sediment, or water, the epidemiological reports that almost invariably point to the shrimp consumption by affected individuals as cause of illness, lead us to believe that these organisms were the most probable responsible vectors. Evidence is presented here that the same strain (although not pandemic) was isolated from shrimp and stools (Fig. 2 cluster b), reinforcing the idea that shrimp are indeed vectors of V. parahaemolyticus and that at least two tdh+ strains were involved in the gastroenteritis outbreak of 2004.
This is the first report of the presence of the V. parahaemolyticus pandemic strain O3 : K6 in the North American Pacific.
Environmental conditions of the lagoon during September and October, elevated temperatures (>30°C) with low salinities (<10 ppt), are appropriate for the proliferation of this pathogen (DePaola et al., 2003a). During the winter months, the water temperature falls and the salinity can rise to more than 90 ppt in some parts, forcing the strain to over-winter in a yet unknown niche (DePaola et al., 2003a). Because the same strain, as analyzed by rep-PCR (CAIM 740), has been found in at least two consecutive years, it is presumed that these events are periodical. During 2005, no cases were reported, although fishing and consumption of shrimp was permitted again in the area. During this year the lagoon and particularly shrimps were monitored, but only two isolates were recovered from shrimp macerates that were potentially pathogenic because of the presence of tlh, tdh, and trh genes (CAIM 1772 and CAIM 1773).
The spread of non-indigenous microorganisms has been attributed to the water ballast of ships (Myers et al., 2003), probably associated with copepods. Assuming that spread of the pandemic strains to Mexico was mediated by ballast water, it may have arrived via the port of Mazatlan. Due to the severity of the outbreaks during 2003 and, particularly, 2004, it is probable that the diarrhea episodes that occurred before 2003 could have been caused by another pathogen or strain (at least two other tdh+/trh+ strains were detected). Unfortunately, although no strains or data are available for the episodes before 2003, one could propose that the pandemic strain may have arrived Mexico just before the outbreak, sometime during 2003 or 2002.
Roxana Atondo-Mexía, Carmen Bolán, Julio Monjardin-Heraldez, Perla Carrillo-Duart, and Lidia Nevarez-Ceniceros for technical help. DNA of strain CNRVC010089 was donated by Dr L. Lizárraga-Partida, CICESE, México, and of strain JKYVp6 by Dr Romilio Espejo, Chile.