First report of Edwardsiellosis in cage-cultured sharpsnout sea bream, Diplodus puntazzo from the Mediterranean

The disease has affected cultured sharpsnout sea breams of a commercial fish farm
in a single location in East Greece (Saronikos bay). Two populations of sharpsnout
sea breams stocked in two consecutive years into 4 floating cages presented signs
of disease. Each cage was stocked with approximately 50–55.000 fish. Two outbreaks
of Edwardsiellosis were recorded, the first in October 2013 and the second in July
2014. Table 1 summarizes the production data of the affected stocks and the mortalities observed.

Table 1. Production data and mortalities. Production data of the affected sharpsnout sea breams
and mortality observed during the two outbreaks of Edwardsiellosis

Diseased fish exhibited nodules and abscesses in spleen and kidney (Fig. 1). Pure colonies with identical morphology were obtained at the site by the fish vet
of the fish farm in 10 % horse blood Agar (Oxoid) from the kidneys of approximately
30 fish exhibiting signs of distress in each incidence. The initial presumptive diagnosis
was performed at the fish farm using API20E bacterial identification kit (BioMerieux,
France) on 8–10 pure bacterial colonies from each incidence (n?=?18). The API profile of all isolates was similar with variability only in the H
₂
S production. Other apparently healthy fish species cultured separately in different
cages at the same site of the fish farm including gilthead sea bream, Sparus aurata, European sea bass, Dicentrarchus labrax, meagre, Argyrosomus regius and red porgy, Pagrus pagrus were also sampled (n?=?20 per fish species) for bacteriological examination (bacterial cultures in blood
agar) during the outbreaks and were negative for E. tarda. The fish were treated successfully following oral administration of oxolinic acid
in feed (50 mg kg
^?1
biomass for 7 consecutive days).

Fig. 1. Nodules in spleen. Enlarged spleen of affected fish with multiple whitish nodules

Pure cultures of two strains (EtS1 and EtS2) isolated from sharpsnout sea bream from
each two incidences were sent to the laboratories of HCMR for further characterization.
Both isolates formed small distinct colonies. The bacteria were characterized and
identified biochemically with BIOLOG GEN III and API20E systems according to manufacturers’
instruction and molecularly using PCR amplifying Edwardsiella sp. specific gyrB gene coupled with 16 s sequencing 15], 16]. The 16 s rRNA sequence data of the two strains have been deposited in NCBI GenBank
under the accession numbers KP729431 and KP729432. In addition, the two strains were
screened for the presence of selected virulence genes (citC, fimA, gadB, katB, mukF and esrB) with PCR 17], 18]. All primers used are presented in Table 2.

Table 2. Primers. List of primers used in this study

The isolates were Gram negative, non-motile, oxidase negative, rods, identified as
E. tarda with 55 % probability using BIOLOG GENIII. They could utilize several carbon sources
such as sugars (maltose, glucose, N-Acetyl-D-glycosamine, N-Acetyl-neuraminic acid,
mannose, fructose, glucose-6-PO
₄
, fructose-6-PO
₄
, serine), aminoacids (galacturonic acid, gluconic acid, glucuronic acid) and caroboxylic
acid (lactic acid). Both strains could not utilize arabinose, mannitol and sucrose
and exhibited variability in H
₂
S production. They could not grow in salinity exceeding 4 % NaCl and they were resistant
in acidic pH. They could grow in 20 and 25 °C with best growth at 30 °C.

Identification was further validated after amplification of gyrB gene which resulted in a PCR product at the expected size. All virulence genes assessed
were present in the isolates (Fig. 2). Haemolysis was tested on horse and sharpsnout sea bream-blood agar. The later was
obtained from disease-free broodstock fish of the Institute of Marine Biology, Biotechnology
and Aquaculture, HCMR. Strains E. tarda (DSM 30052) and E. hoshinae (DSM 13771) were used as negative controls and Aeromonas veronii strain (Aero1- HCMR pathogen collection) previously shown to be strongly haemolytic
on fish blood as positive control. Following 48 h incubation at 25 °C only the Aeromonas veronii control strain was positive for haemolysis.

Fig. 2. Virulence genes. PCR amplification of selected virulence genes for EtS1. Lanes 1–7:
esrB, katB, gadB, citC, mukF, fimA, gyrB

Antibiotic susceptibility of the two isolates was determined with a disk diffusion
method 19] on Mueller-Hinton agar for ampicillin (10 ?g), oxytetracycline (30 ?g), oxolinic
acid (2 ?g), flumequine (30 ?g), florfenicol (30 ?g) and sulphamethoxazole/trimethoprim
(25 ?g) as suggested for fish pathogens 20]. Inhibition diameter was recorded after 24 h incubation at 25 °C. Response was determined
according to the breakpoints shown in Table 3.

Table 3. Antiobiotic sensitivity. Sensitivity of the two Edwardsiella isolates against the antibiotics commonly used in aquaculture (r: radius, S: sensitive,
I: intermediate, R: resistant)

Phylogenetic relationships of the two isolates were studied using 16 s rRNA gene.
The taxonomic status, strain collection numbers and GenBank Accession numbers (http://www.ncbi.nlm.nih.gov/) of all used sequences are presented in Table 4. A total length of 721 bp sequences was used for the phylogenetic analyses. Genetic
distances and Neighbor-joining analysis 21] were respectively estimated and performed in MEGA 22] under Tamura-Nei 23] model of evolution. Confidence of tree nodes was tested by bootstrap analyses with
1000 replicates. Both isolates were grouped with E. ictaluri and E. tarda strains isolated from fish (Fig. 3). Genetic distances between strains EtS1 and EtS2 was 0 %, while the mean genetic
distance within the fish-clade was 0.002 % and overall mean distance for the genus
was 0.005 %.

Table 4. Bacterial strains and codes. Taxonomy, collection codes and accession numbers from
NCBI GenBank of the bacterial strains used in this study

Fig. 3. Phylogenetic analysis. Phylogenetic relationships of Edwardsiella strains studied as derived from Neighbour-Joining analysis for 16S gene. Numbers
on clades indicate the bootstrap values