Vibrio vulnificus in Denmark

5. Characterization of Vibrio vulnificus

5. 1 Genotypic characterization

Genotypic techniques used for typing V. vulnificus include randomly amplified polymorphic DNA polymerase chain reaction (RAPD-PCR) (Aznar et al., 1993; Høi et al., 1997), pulsed field gel electrophoresis (PFGE) (Buchrieser et al., 1995; Tamplin et al., 1996), ribotyping (Aznar et al., 1993; Dalsgaard et al., 1996b; Tamplin et al., 1996; Høi et al., 1997), and plasmid profiling (Davidson & Oliver, 1986). RAPD-PCR and PFGE results have shown that V. vulnificus biotype 1 is genetically very heterogeneous (Buchrieser et al., 1995; Aznar et al., 1993; Tamplin et al., 1996; Høi et al., 1997). PFGE and RAPD-PCR may be of use in retrospective epidemiological investigations and in ecological studies, but these techniques are too discriminatory for typing V. vulnificus biotype 1 strains (Høi et al., 1997). Buchrieser et al. (1995) used PFGE to characterize V. vulnificus populations isolated from three single oysters. Ninety-five typeable V. vulnificus isolates were divided into 60 genotypes; isolates within a single oyster did not show higher similarity than isolates from different oysters. The study illustrates the high genotypic diversity among V. vulnificus isolates in the oysters (Buchrieser et al., 1995).

A high degree of genotypic diversity was also observed when ribotyping and RAPD-PCR were used to characterize isolates from Danish coastal areas and sediment (Høi et al., 1997). The Danish clinical and environmental strains could not be distinguished by ribotyping which indicates that no specific genomic characteristics are associated with clinical or environmental strains (Høi et al., 1997). Ribotyping has been reported to be useful for differentiating biotype 1 and 2 isolates and for differentiating strains of different geographical origin (Aznar et al., 1993; Høi et al., 1996; Høi et al., 1997; Arias et al., 1997).

Plasmid profiling

The use of plasmid profiling for epidemiological purposes has only been investigated in a few studies. One study reported a low incidence of plasmids in 42 environmental and clinical V. vulnificus strains from the US whereas 8 of 11 strains isolated from wound infections in Denmark harbored plasmids of variable sizes (Davidson & Oliver, 1986; Dalsgaard et al., 1996b). The plasmid content of V. vulnificus strains isolated from cases of primary septicaemia have so far not been studied. In addition, the methods for plasmid extraction and size measurements need improvements as several nucleases affects plasmid quality and indications of plasmid sizes vary considerably in repeated testing (unpublished results). As mentioned in Section 4.4.2, most V. vulnificus biotype 2 strains harbor at least two high molecular weight plasmids which may be associated with eel virulence (Biosca et al., 1996c; Biosca et al., 1997a; Høi et al., 1998b; Dalsgaard et al., in press;). Curing and transfer experiments with these plasmids are necessary to determine their role in pathogenicity.

Prediction of virulence impossible

Jackson et al. (1997) reported that 50 isolates obtained from blood from each of two clinical cases showed 100% homology in Pulsed Field Gel Electrophoresis (PFGE) analysis with four different restriction enzymes. This suggest that a single V. vulnificus strain was responsible for each of the clinical cases. No methods is at the present time able to predict which strain is the Amost@ virulent or which persons within an at-risk population are the most susceptible to V. vulnificus infections. Host specificity has been hypothesized to be a key element of V. vulnificus diseases, since only 5 to 10 infections are reported in Florida each year among an at-risk liver-diseased oyster-eating population estimated to be over 70,000 (Jackson et al., 1997).

Future research should include host specificity and not solely focus on determining virulence markers in V. vulnificus by various typing techniques.

5.1.1 Origin of V. vulnificus biotype 2

Biotype 1 strains are genetically heterogenous compared to biotype 2 strains from a variety of countries, e.g. Denmark, Japan, Spain, Norway, Sweden, and Taiwan (Biosca et al., 1997a; Aznar et al., 1993a; Arias et al., 1998a; Høi et al.,1997). An eel-pathogenic clone of V. vulnificus may have acquired eel-virulence-genes⁶ from other bacterial species or the genes may have evolved by positive selection (i.e. strains with the Abest@ virulence genes will outgrow strains with less efficient virulence genes) in Japan in 1975. The eel-virulence-genes might also have been acquired through transduction by a lysogenic bacteriophage which at the same time could have mediated a serotype conversion, a phenomenon that is described for V. cholerae (Manning et al., 1994).

Role of eels

The annual spawning migrating routes of the Japanese eel (A. japonica) and the European eel (A. anguilla) do not intersect, therefore, the clone of eel-pathogenic V. vulnificus strains have most likely not been transferred directly from A. japonica to A. anguilla in the environment (Nielsen, 1997). A more likely explanation could be that this clone has spread to Europe with an import of live elvers and eels from Asia (Nagasawa et al., 1994). The clone could have been transferred through Europe by shipments of live eels. This hypothesis does not explain how this clone has Aresisted@ genetic changes and retained a homogenous genetic profile since 1975. If strains belonging to the eel-pathogenic clone of V. vulnificus undergo naturally spontaneous mutations they might loose their capability to cause infections in eels and thereby Abe excluded@ of the clone. The genetic homogeneity of these strains might ensure that the eel-virulence-genes remain intact.

5.2 Serotyping, capsule typing and phage typing

Two serological typing schemes have been proposed for V. vulnificus (Martin & Siebeling, 1991; Shimada & Sakazaki, 1984). The typing scheme of Shimada et al. (1984) is based on direct agglutination of heat-killed V. vulnificus cells by polyclonal rabbit antisera that recognize 7 O serovars. Five different O serovars and 10 different capsule types are identified among V. vulnificus strains in the typing scheme proposed by Martin and Simonson which was applied to a collection of Danish clinical and environmental V. vulnificus isolates (Martin & Siebeling, 1991; Simonson & Siebeling, 1993). Anti-capsule reagents which detected 50% of clinical isolates recovered in the US did not recognize any capsule antigens of the Danish isolates (other than a few biotype 2) nor did anti-LPS MAbs react with the Danish isolates suggesting different V. vulnificus populations in the US and Denmark (Danieu et al., 1996). Capsule types have also been studied by using high-performance anion exchange chromatography (HPAE) and nuclear magnetic resonance (Bush et al., 1997; Hayat et al., 1993) and these studies also indicate that numerous capsule types exist in V. vulnificus. Typing of V. vulnificus biotype 1 using reagents raised against different LPS types and capsule types are of limited value because of the high heterogeneity among this species which would require expensive production of many new antisera. Thirteen phages specific for V. vulnificus have been isolated from oysters collected in Louisiana, Alabama and Florida and used for phage typing of V. vulnificus from diseased eels (DePaola et al., 1997a; DePaola et al.,1998; Høi et al., 1998b). The majority of V. vulnificus isolates from Danish eels were lysed by one or more of these phages which suggests a relatedness in V. vulnificus isolates from Gulf of Mexico and Denmark. One to six phage types were seen in each outbreak and none of the phages were specific for certain LPS or capsule types. No apparent correlation was observed between phage typing, ribotyping, and serotyping (Høi et al., 1998b).

5.3 Substitution of the taxon "biotype" to "serovar"

Bisoca et al. (1997a) proposed to rename those strains previously classified as biotype 2 to that of serovar E based on three criteria: (i) serovar E strains express a homogeneous lipopolysaccharide (LPS)-based O serovar while biotype 1 strains are serologically heterogenous; (ii) the majority of serovar E strains are indole-negative and the majority of biotype 1 strains are indole-positive; and (iii) only serovar E strains are virulent for eels. V. vulnificus biotype 2 strains are genetically homogenous compared to biotype 1 strains and they usually harbor high molecular weight plasmids, which may be associated with virulence (Biosca et al., 1996c).

Høi et al. (1998b) recently reported that V. vulnificus isolated from diseased Danish eels are more heterogeneous as shown by O-serovars, capsule types, ribotyping, phage typing, and plasmid profiling than those from Japan and Spain. At least three LPS-associated serovars were isolated from Danish diseased eels which indicate that grouping V. vulnificus biotype 2 into a single serovar is not adequate.

Indole reaction

The indole reaction has previously been used to distinguish between V. vulnificus biotype 1 and 2 (Biosca et al., 1996c). One biotype 2 strain has previously been reported to be indole positive, and data from Denmark further demonstrate that indole production is not a reliable marker for biotype 2 since the majority of isolates from diseased eels in Denmark were indole positive (Høi et al., 1998b). Eel-pathogenic V. vulnificus isolates, which are indole positive, seem to be unique to Danish eel farms and may have evolved from the "original" eel-pathogenic clone (Høi et al., 1998b).

5.4 Conclusions from Chapter 5

Characterization of V. vulnificus has not identified particular traits that predict pathogenicity for humans but has illustrated the high genetic diversity among environmental V. vulnificus isolates and that high molecular weight plasmids may play a role in eel virulence.

V. vulnificus was originally divided into two biotypes differing phenotypically, serologically, and in host range but recent research has indicated that grouping into serovars is more appropriate.