Epidemiological study on Yersinia ruckeri isolates from rainbow trout (Oncorhynchus mykiss, Walbaum) in North West Germany

Yersinia ruckeri is the causative agent of Enteric Redmouth Disease, one of the most important infectious diseases in rainbow trout hatcheries in Europe. Although generally Y. ruckeri was well controlled by vaccination and antibiotic treatment, outbreaks of this disease were still periodically observed, especially in endemic areas. Moreover, some Y. ruckeri strains seem not to be affected by several of the commercial vaccines. All these strains were lacking motility, while strains from previous outbreaks were all motile. It was hypothesized that vaccination exerted a selective pressure that enabled the emergence of non-motile strains that are resistant to the commercial vaccines. Therefore, there is a high risk that these non-motile vaccine-resistant strains spread and originate severe outbreaks of disease in trout farms. The aim of this project was to investigate the genotypic and phenotypic diversity of Y. ruckeri isolates collected in North West Germany to figure out the relationship according to the colony development, compare intraspecies differences with in vitro cytotoxicity to different fish cell lines, and the expression of pathogenic factors, especially between motile and non-motile isolates, and study their susceptibility to different antimicrobial agents and the genetic determents of antimicrobial resistance to understand the situation trout farming in North Western Germany now is facing. Since 2011, 48 Y. ruckeri isolates were collected from 12 trout farms in North-Rheine-Westphalia during different seasons. Additional 33 isolates, offered by LAVES Hannover and the fish disease service of Hessen, were from clinical cases in Lower Saxony and Hessen. One Reference strain was provided by the Clinic for Poultry at the University of Veterinary Medicine, Hannover, and one typical non-motile strain was offered by Dr. Gould from MSD Animal Health. The isolates were characterized by biochemical tests, pulsed-field gel electrophoresis (PFGE) and repetitive sequence-based PCR assays, including (GTG)5-PCR, BOX-PCR, ERIC-PCR and REP-PCR. Whole cell fatty acid composition was analyzed by gas chromatography. Cytotoxicity assays were performed using Common carp brain (CCB), epithelioma papulosum cyprini (EPC), fathead minnow epithelial cell (FHM) and rainbow trout gonad-2 (RTG-2) cells at different incubation times and temperatures. A multiplex PCR was established to detect 10 virulence factor genes (hemolysin genes yhlAB, ruckerbactin genes rucC and rupG, ABC exporter protein system genes yrp1 and yrpDEF , flagellar secretion chaperones gene flgA, flagellar secretion apparatus gene flhA) found in Y. ruckeri. The expression of these genes was analyzed and compared by semi-quantitative PCR. Susceptibility tests to antimicrobial substances and genetic determents for a resistance to these substances were detected by microdilution test and PCR. Eight different API 20E profiles were obtained, which were different from that of the Y. ruckeri reference strain DSM18506. Five isolates were confirmed as Y. ruckeri by 16S rDNA sequencing. 17 isolates hydrolyzed Tween 80/20. All Y. ruckeri isolates exhibited 15 major fatty acids, including 12:0, 13:0, 13.957 (equivalent chain length, ECL unknown), 14:0, 14.502 (ECL unknown), 15:0, 16:1ω5c, 16:0, 17:1ω8c, 17:0 CYCLO, 17:0, 16:1 2OH, 18:1ω9c, 18:1ω7c and 18:0. In addition to 17 PFGE patterns, two different REP-PCR patterns, five ERIC-PCR patterns, four (GTG)5-PCR patterns and three BOX-PCR patterns were obtained. According to the results of API 20E, PFGE and the various PCR assays, the isolates could be divided into 27 different groups. In rainbow trout hatcheries, isolates from more than two different typing groups were present in the same fish farm. Two groups of isolates were found to be present in all three federal states in North West Germany. Non-motile strains were more active to cause in vitro cytotoxicity to fish cell lines than motile strain in the first 24 h, but without significant difference after longer incubation at the lower temperature. All studied virulence genes were found present in both motile and non-motile strains of Y. ruckeri. In addition, no significant differences were found in the expression of these genes. For most of the antimicrobial agents tested, a unimodal distribution of MIC values was observed. In general, the MIC values increased with the incubation time. For enrofloxacin and nalidixic acid, a bimodal MIC distribution was observed with one population showing MICs of 0.008-0.015 mg/L of enrofloxacin and 0.25-0.5 mg/L of nalidixic acid, respectively, and the other subpopulation exhibiting MICs of 0.06-0.25 mg/L of enrofloxacin and 8-64 mg/L of nalidixic acid, respectively. While isolates (n=3) from the subpopulation with the lower (fluoro)quinolone MICs showed the non-mutated gyrA wildtype QRDR sequence, another ten isolates from the subpopulation with the higher (fluoro)quinolone MICs exhibited different single bp mutations that resulted in single amino acid substitutions in the GyrA protein: Ser ® Arg or Ser ® Ile (at position 83) or Asn ® Tyr (at position 87) [Escherichia coli numbering]. A single isolate showed high MIC values for sulfamethoxazole and sulfamethoxazole/trimethoprim. A ~8.9 kb plasmid was found in this isolate, which carried the genes sul2, strB and a dfrA14 gene cassette integrated into the strA gene. Neither class 1 nor class 2 integrons were found in this isolate. In conclusion, there was a variety of Y. ruckeri isolates in North West German aquaculture, dominated by non-motile strain. In most farms from the field study, two or more different Y. ruckeri isolates were present. It was easier for non-motile strains to cause outbreaks in winter, since they were more active than motile strain at lower temperature. The fatty acid composition of Y. ruckeri isolates was homogenous and there was no obvious link between the fatty acid profiles and the epidemiological parameters. Non-motile strains of Y. ruckeri could not be differentiated from motile strains by either the presence of virulence genes detected in this study or by the expression of these genes. Incubation for 24h-28h at 22±2 °C appears to be sufficient for the testing of Y. ruckeri for susceptibility to antimicrobial substances. The largely unimodal MIC distribution suggests that most of the isolates tested represent the wildtype population that has not acquired resistance mechanisms. Exceptions were (i) the finding of isolates with elevated (fluoro)quinolone MICs that also had resistance-mediating mutations in the QRDR region of gyrA, and (ii) the single isolate that carried plasmid-borne sul2, dfrA14, ΔstrA and strB genes. This observation confirmed that Y. ruckeri is able to develop mutations for (fluoro)quinolone resistance, but can also acquire plasmid-borne resistance genes.