Genetic variability amongst <i>Streptococcus iniae</i> isolates from Australia

Abstract

Streptococcus iniae is a beta-haemolytic, Gram-positive coccus that causes generalized septicaemia and meningoencephalitis in fish (Perera, Johnson, Collins & Lewis 1994). Significant mortality in a diverse range of commercial fish species has caused severe losses for the aquaculture industries in the US, Japan, Israel, Bahrain, Europe, Indonesia and Australia (Agnew & Barnes 2007; Bromage, Thomas & Owens 1999; Colorni, Diamant, Eldar, Kvitt & Zlotkin 2002; Goh, Driedger, Gillett, Low, Hemmingsen, Amos, Chan, Lovgren, Willey, Shaw & Smith 1998; Kusuda & Kawai 1998; Perera et al. 1994; Shoemaker, Klesius & Evans 2001; Stoffregen, Backman, Perham, Bowser & Babish 1996; Yuasa, Kitancharoen, Kataoka & Al-Murbaty 1999). In Australia, S. iniae causes substantial losses to the barramundi, Lates calcarifer (Bloch), aquaculture industry, resulting in a number of farms closing during recent years. Moreover, recent attempts to control the disease through vaccination have met with limited success. In addition to losses to aquaculture, S. iniae has been reported to cause disease in humans handling fresh infected fish (Lau, Woo, Luk, Fung, Hui, Fong, Chow, Wong & Yuen 2006; Lau, Woo, Tse, Leung, Wong & Yuen 2003; Weinstein, Litt, Kertesz, Wyper, Rose, Coulter, Mcgeer, Facklam, Ostach, Willey, Borczyk & Low 1997) and is thus of increasing clinical significance. Human infections have been attributed to certain genotypes based on pulsed-field gel electrophoresis (Fuller, Bast, Nizet, Low & de Azavedo 2001), but there would appear to be more than one genotype involved in human infections (Lau et al. 2003). The aim of the present study was therefore to characterize Australian isolates of S. iniae genotypically and compare them with reference fish and human isolates from the Centers for Disease Control and Prevention (CDC), Atlanta, USA in order to elucidate the diversity of isolates within the Australian aquaculture industry. Streptococcus iniae isolates were obtained from veterinary laboratories across Australia, and from the CDC, Atlanta, USA (Table 1). Bacterial isolates were routinely cultured on Columbia agar base containing 2% defibrinated sheep blood at 35 °C. Identity of all strains received was determined by Gram stain and API 20 Strep (BioMerieux, Marcy L’Etiole, France). Identity was confirmed by PCR amplification of the lactate oxidase gene (lox) as described previously (Mata, Blanco, Dominguez, Fernandez-Garayzabal & Gibello 2004) from genomic DNA purified as follows: S. iniae from overnight vegetable peptone cultures (10 mL) were washed three times in phosphate-buffered saline (pH 7.4), resuspended in 500 μL lysozyme (100 mg mL−1) and incubated for 1 h at 37 °C. Following addition of 200 μL sodium dodecyl sulphate (20%) and 100 μL proteinase K, cells were lysed overnight at 55 °C. Saturated NaCl was added to one-third of the original volume and the mixture incubated at 4 °C for 20 min. Lysates were cleared by centrifugation at 10 000 g for 1 min and the supernatant transferred to clean 1.5 mL microcentrifuge tubes. DNA was precipitated with three volumes of 95% ethanol and washed in 70% ethanol before drying and resuspending in 0.5 m Tris–EDTA buffer (TE pH 7.0). All isolates were Gram positive and beta-haemolytic, and gave essentially identical profiles on API 20 Strep, being positive for esculin hydrolysis, pyrrolidonyl arylamidase, β-glucuronidase, alkaline phosphatase and leucine arylamidase. All isolates could utilize ribose, mannitol, trehalose, and starch after 24 h incubation. The only variability was found in arginine dihydrolase (ADH) activity where the majority of isolates were positive, but on occasions would give negative results. This has previously been demonstrated to be an artefact of the API ADH assay (Barnes & Ellis 2003). When PCR was performed with lox gene primers an amplicon in the region of 870 bp was detected in line with previous results (Mata et al. 2004). Restriction enzyme (Sma1) digests of genomic DNA were analysed by pulsed-field gel electrophoresis (PFGE) essentially as previously described (Facklam, Elliott, Shewmaker & Reingold 2005). Briefly, colonies of overnight blood culture of S. iniae were suspended in 0.5 mL 10 mm Tris–HCl, 1.0 m NaCl, pH 7.6 and adjusted to MacFarland 4. Aliquots (100 μL) were mixed with an equal volume of 2% low melting temperature agarose (BioRad, Regents Park, Australia) and solidified in CHEF DRIII plug moulds (BioRad). Plugs were then incubated at 35 °C with gentle agitation overnight in lysis solution (6 mm Tris–HCl, 1.0 m NaCl, 100 mm EDTA, 0.5% Brij 58, 0.2% sodium deoxycholate, 0.2% sodium lauroyl sarcosine, 1 mg mL−1 lysozyme, 5 U mL−1 mutanolysin (Sigma, Castle Hill, Australia), followed by incubation for 24 h at 50 °C in 0.5 m EDTA, 1% sodium lauroyl sarcosine, 0.1 mg mL−1 proteinase K, pH 7.6 with gentle agitation in a Thermomixer (Eppendorf, Hamburg, Germany). Plugs were then washed 4 × 2 h in 10 mm Tris, 0.1 mm EDTA, pH 7.6 (TE) and stored in the same buffer at 4 °C until required. Prior to digestion, plugs were equilibrated in 1 mL restriction enzyme buffer (Takara, Shiga, Japan) containing 0.01% bovine serum albumin for 1 h with gentle agitation at room temperature. Digestion by SmaI (Takara, Shiga, Japan) was performed in 0.3 mL restriction enzyme buffer at 30 °C for 24 h. Plugs were washed in TE prior to cutting and loading into wells of a 1% pulsed-field agarose/TBE gel. Electrophoresis was performed in TBE buffer for 20 h using a CHEF DRIII PFGE system (BioRad) with a switch time of 5–40 s, 120° angle and a voltage gradient of 6 V cm−1 at 10 °C, with a concatenated Lambda ladder (BioRad) as a molecular weight marker. The gel was stained with ethidium bromide (0.5 μg mL−1) in distilled water and destained in distilled water. PFGE profiles were interpreted according to standard criteria proposed by Tenover, Arbeit, Goering, Mickelsen, Murray, Persing & Swaminathan (1995). Genotyping by pulsed-field gel electrophoresis separation of SmaI digests of genomic DNA revealed a number of quite distinct profiles. To overcome potential problems of inter-laboratory variation in PFGE profiles, SmaI digests of Australian isolates were run in parallel on the same gel with reference isolates from the CDC, Atlanta, USA (1, 2). The early dolphin strains (Fig. 1, lanes 4 & 5; Fig. 2, lanes 3 & 4), including the type strain, were quite distinct from more recent fish and human isolates, in agreement with previous reports (Facklam et al. 2005; Weinstein et al. 1997). This perhaps reflects the opportunity presented for multiple mutations afforded by subculture to confirm viability over more than 30 years of maintenance in strain collections. There were clear similarities in profile between strains isolated from fish and humans, regardless of geographical origin (1, 2). However, there were also marked differences, which did not fully correlate with geographical origin or species from which the strains were isolated; Queensland isolate QMA0079 (Fig. 1, lane 10) bearing strong homology to isolates from Northern Territory (Fig. 1, lanes 15, 16 and 17) and some Western Australian isolates (Fig. 1, lanes 12, 13, 14). Isolates from New South Wales QMA0156 and QMA0157 (Fig. 2, lanes 17 and 18) were similar to South Australian isolates QMA0159 and QMA0160 (Fig. 2, lanes 8 and 9), and these were identical to an isolate from Western Australia (QMA0087; Fig. 2, lane 5) and Queensland (QMA0079; Fig. 2, lane 10), possibly suggesting a common origin for these isolates. Considering the limited number of hatcheries producing barramundi juveniles, which are then shipped to producers for grow out across Australia, this similarity may be expected. However, isolate QMA0158 from South Australia had additional high-molecular-weight bands compared with the other South Australian isolates (Fig. 2, lanes 7, 8 and 9). Moreover, there was diversity amongst isolates from other common geographical origins including Queensland (Fig. 1, lanes 6–10) and Western Australia (Fig. 1 lanes 11–14). Diversity did not reflect year of isolation, with a very recent isolate from a farm in Queensland (QMA0164, isolated 2006; Fig. 2, lane 6) being identical to some of the oldest isolates in our strain collection (QMA0072, isolated 1995, and QMA0074, isolated 1995; Fig. 2, lanes 12 and 13). Pulsed-field gel electrophoretogram (2% pulsed-field agarose) of Sma1 digests of Streptococcus iniae isolates. M, concatenated Lambda ladder, lane 1, ss1440 (human, Canada); lane 2, 4780-01 (human, USA); lane 3, 105-04 (human, USA); lane 4, ss1056 (dolphin, USA); lane 5, ss1123 (dolphin, USA); lane 6, QMA0072 (Lates calcarifer, Qld); lane 7, QMA0074 (L. calcarifer, Qld); lane 8, QMA0075 (L. calcarifer, Qld); lane 9, QMA0076 (L. calcarifer, Qld); lane 10, QMA0079 (L. calcarifer, Qld); lane 11, QMA0080 (L. calcarifer, WA); lane 12, QMA0083 (L. calcarifer, WA); lane 13, QMA0084 (Pteropus alecto, WA); lane 14, QMA0087 (L. calcarifer, WA); lane 15, QMA0109 (L. calcarifer, NT vaccine strain); lane 16, QMA0115 (L. calcarifer, NT, isolated pre-vaccination); lane 17, QMA0120 (L. calcarifer, NT, isolated post-vaccination). Pulsed-field gel electrophoretogram (2% pulsed-field agarose) of Sma1 digests of Streptococcus iniae isolates. M, concatenated Lambda ladder, lane 1, 105-04 (human, USA); lane 2, ss1440 (human, Canada); lane 3, ss1123 (dolphin, USA); lane 4, ss1056 (dolphin, USA); lane 5, QMA0087 (L. calcarifer, WA); lane 6, QMA0164 (L. calcarifer, Qld); lane 7, QMA0160 (L. calcarifer, SA); lane 8, QMA0159 (L. calcarifer, SA); lane 9, QMA0158 (L. calcarifer, SA); lane 10, QMA0079 (L. calcarifer, Qld); lane 11, QMA0076 (L. calcarifer, Qld); lane 12, QMA0074 (L. calcarifer, Qld); lane 13, QMA0072 (L. calcarifer, Qld); lane 14, QMA0120 (L. calcarifer, NT); lane 15, QMA0151 ((L. calcarifer, NT); lane 16, QMA0109 (L. calcarifer, NT vaccine strain); lane 17, QMA0157 (L. calcarifer, NSW); lane 18, QMA0156 (L. calcarifer, NSW). Whilst distinct genetic profiles have been associated with human isolates of S. iniae in the past (Fuller et al. 2001) along with several RFLP profiles (Eldar, Lawhon, Frelier, Assenta, Simpson, Varner & Bercovier 1997), we found that human isolates from CDC, Atlanta had very similar SmaI restriction profiles when analysed by PFGE to many of the fish isolates from Australia. Although barramundi are sold as both processed (filleted) and round fish into the Australian market, as yet, there are no reports of human infection by S. iniae in Australia. However, this may reflect the observation that commercial diagnostic kits have been variously reported to identify S. iniae as S. uberis, S. dysgalactiae subsp. equisimilis and S. anginosus (Lau et al. 2003, 2006) and thus sporadic human cases in Australia may not yet have been identified in routine clinical diagnosis. It would certainly merit a retrospective investigation of clinical isolates of Streptococcus sp. recovered from the appropriate demographic in Australia. Interestingly, the isolate from a flying fox in Western Australia (Fig. 1, lane 13) produced a similar PFGE profile to fish isolates from Western Australia (Fig. 1, lanes 11, 12 and 14). As flying foxes are herbivores, it is unlikely the animal was infected through consumption of infected fish and this raises the prospect of this animal being a potential vector. Isolates from the Northern Territory were of particular interest, as fish from this state vaccinated with a recent autogenous vaccine succumbed to a subsequent outbreak of S. iniae a matter of weeks post-vaccination. However, in terms of genotype based on SmaI digest of genomic DNA, strains isolated from infected fish post-vaccination (Fig. 1, lanes 16 and 17; Fig. 2, lanes 14 and 15) were identical to the strain used in the vaccine (Fig. 1, lane 15; Fig. 2, lane 16). Whilst there are a number of possible reasons for the vaccine failure, it may be that any genetic differences resulting in serological diversity are not be detected using a rare-cutting restriction enzyme such as SmaI. Further work is therefore required to determine the molecular basis for potential serological variability in order to develop rapid, reliable and reproducible methods for determining surface serotype prior to vaccine development. Further information on the genes associated with surface serotype may also facilitate the development of a generic vaccine to cross-protect against multiple strains of S. iniae in cultured fish and prevent potential transfer into humans during processing. The authors are particularly grateful to Nicky Buller, Annette Thomas, Judy Forbes-Faulkner, Suresh Benedict, Mark White and Lynn Shewmaker for kind donation of the strains used in this study. Roslina Nawawi is supported by a scholarship from the Malaysian government.