The genus Edwardsiella belongs to the Hafniaceae family in Enterobacterales and comprises five valid species: E. anguillarum, E. ictaluri, E. piscicida, E. hoshinae, and E. tarda (Buján et al., 2018). These are major bacterial pathogens causing edwardsiellosis in aquatic animals, responsible for economic losses in aquaculture globally (Park et al., 2012; Soto et al., 2012). These bacteria may cause zoonotic infections in warm-blooded animals and humans (Adepoju et al., 2021; Pham, et al., 2021). Among the five species, E. tarda has been well-studied as it has representative zoonotic pathogenicity (Michael and Abbott, 1993; Xu and Zhang, 2014): studies have shown that the species actually represents three genetically distinct taxa, and almost all the fish-pathogenic isolates were re-assigned as E. anguillarum and E. piscicida (Shao et al., 2015; Buján et al., 2018). Edwardsiella anguillarum is now considered a newly emerging aquatic pathogen, in aquaculture (especially in eel culture) and wild aquatic animals (Oh et al., 2020). Extensive antibiotic usage in global aquaculture has resulted in antimicrobial resistance (AMR) to various agents, as observed in Edwardsiella isolates (Wimalasena et al., 2019; Algammal et al., 2022). Therefore, the identification of potential virulence factors and AMR-associated genes in E. anguillarum is urgently needed, as the mechanisms underlying them remain unclear. To date, only seven genome sequences for E. anguillarum (one complete and six drafts) are available in the GenBank database (accessed in December 2022), mostly isolated from diseased Japanese eels (Anguilla japonica). Here, we present the first complete genome sequence of E. anguillarum isolated from a diseased giant mottled eel (Anguilla marmorata) cultured in Korea.

The moribund juvenile giant mottled eels with petechial hemorrhage on the skin were collected from a private eel culture farm (Gyeonggi-do, Republic of Korea), and the representative C-5-1 strain was isolated from a liver abscess in the fish by culturing overnight on tryptic soy agar (Difco) at 25°C. Based on the gyrB sequence analysis (Pascual et al., 2010), the isolated bacteria showed > 99% nucleotide identity with the E. anguillarum strains available in the GenBank database, including ET080813^T (NZ_CP006664.1), and was finally classified as E. anguillarum. For whole-genome sequencing, genomic DNA was extracted using the QIAamp DNA Mini Kit (Qiagen) and sequenced via a hybrid approach using the PacBio Sequel System (Pacific Biosciences), by constructing a 20-kb SMRTbell^TM template library, and the HiSeq X-10 platform (Illumina), by preparing a DNA library using the TruSeq Nano DNA library prep kit (Illumina). The PacBio reads (599,706,396 bp; 69,247 reads) were assembled with the Microbial Assembly Application in the SMRT Link v8.0 (Ardui et al., 2018). The Illumina paired-end reads (905,820,202 bp; 6,006,018 reads) were used for accurate genome sequencing using Pilon v1.21 (Walker et al., 2014). Genome annotation was performed using the NCBI prokaryotic genome annotation pipeline (http://www.ncbi.nlm.nih.gov/books/NBK174280/).

The C-5-1 genome comprised 4,293,127 bp (58.7% G + C content, 126× coverage), consisting of one chromosome (4,140,843 bp, 58.9% G + C content, 125× coverage) and one plasmid named p1_C-5-1 (152,284 bp, 53.3% G + C content, 156× coverage), encoding 3,771 protein-coding genes, 25 rRNAs, 99 tRNAs, 6 noncoding RNAs, and 109 pseudogenes (Table 1, Supplementary data Fig. S1). The genomic similarity between C-5-1 and other available type strains of Edwardsiella spp. was assessed using the Orthologous ANI tool software (Lee et al., 2016) and the genome was most similar to E. anguillarum ET080813^T (99.96%) than to other Edwardsiella strains (Supplementary data Fig. S2). This was also validated by genome-based phylogeny using the Type (Strain) Genome Server (TYGS) (Meier-Kolthoff et al., 2022) (Supplementary data Fig. S3).

Table 1

Features	Value
Genome size (bp)	4,293,127
Chromosome	4,140,843
Plasmid (p1_C-5-1)	152,284
G + C content (%)	58.7
No. of contigs	2
Plasmid	1
Total genes	4,010
Protein-coding genes	3,771
rRNAs	25
tRNAs	99
ncRNAs	6
Pseudogenes	109

Potential virulence-associated genes in the C-5-1 strain were compared with those of Edwardsiella spp. strains available in the GenBank database while additional virulence and AMR genes were screened as previously described (Lee et al., 2018). Among those reported in E. anguillarum (Zhai et al., 2021), genes encoding for the ABC transporter and flagellar assembly were detected in the C-5-1 chromosome. Several quorum sensing (QS)-related genes (qseB, qseC, qseD, and LuxS/AI-2), previously reported in E. tarda, have also been detected (Wang et al., 2011; Park et al., 2012). This suggests that C-5-1 has the potential ability to use QS to modify bacterial surface structures including flagella and fimbriae, increasing its survival rate in particular environmental niches.

Several AMR-related genes were detected in the C-5-1 chromosome and plasmid (Supplementary data Table S1). Interestingly, tet(D) (tetracycline efflux MFS transporter D) (locus_tag: MUN71_RS19745) was detected in the C-5-1 plasmid with a deduced amino acid sequence > 99% identical to those previously reported for E. piscicida isolated in Korea (Kim et al., 2018), and to some strains of Citrobacter spp. and Alteromonas spp. isolated in Japan (Furushita et al., 2003). This suggests that the gene can be transferred via plasmid between genera, as well as in Edwardsiella spp., and could be associated with Tet(D)-related antimicrobial resistance in the aquaculture industry.

Eight prophage regions were identified in the C-5-1 chromosome using the PHASTER (Phage Search Tool Enhanced Release) server (Arndt et al., 2016), three of which were intact (score > 90) and five incompletes (score < 70) (Supplementary data Table S2). The length of intact prophage regions consisted of 17.6 kb (Region 2), 44.1 kb (Region 6), and 44.5 kb (Region 7). Prophages are known to be involved in evolving the pathogenicity of host bacteria including antibiotic resistance and adaptation to new ecological niches (Fortier and Sekulovic, 2013). In addition, several spacers derived from viral sequences were detected in their genome by CRISPRCasFinder (Grissa et al., 2007). The analysis indicated that a total of 32 spacers, which were closely related to those from other Edwardsiella spp., Citrobacter spp., and Salmonella spp., were found in the C-5-1 chromosome. Based on these results, the strain C-5-1 could survive against various bacteriophage infections and might evolve into a more uncontrollable pathogen due to gene transfer carried out by prophages.

This is the first report on the genome of E. anguillarum in Korea. The results provide important insights into its genomic characteristics, which might lead to the development of strategies to control E. anguillarum prevalence in the Korean aquaculture industry.

Nucleotide sequence accession numbers

Edwardsiella anguillarum C-5-1 was deposited in the Korean Collection for Type Cultures (KCTC) under the accession number KCTC 92163. The C-5-1 chromosome and plasmid were deposited in the DDBJ/END/GenBank under the accession numbers CP095163 and CP095164, respectively.

Edwardsiellosis는 수생 동물, 특히 양식 어류에 집단 폐사를 일으킬 수 있는 질병으로, Edwardsiella 속 세균 종들이 원인균으로 알려져 있다. 현존하는 5종의 Edwardsiella spp. 중 E. anguillarum은 가장 최근에 분류된 종으로 국내·외 뱀장어 양식 산업에서 심각한 위해를 끼치고 있는 것으로 알려져 있다. 본 연구에서 우리는 국내에서 양식되는 무태장어(Anguilla marmorata) 중 병어로부터 E. anguillarum C-5-1을 분리하였으며, 해당 균주의 전장 유전체를 분석하였다. 분석 결과, E. anguillarum C-5-1 균주는 원형의 유전체(4,140,843 bp)와 하나의 플라스미드(p1_C-5-1; 152,284 bp)를 지니며, 3771개의 단백질 코딩 유전자로 구성됨을 확인하였다. 더불어 C-5-1 균주에서 다양한 병원성, 항생제 내성, prophage 영역 및 CRISPR spacer 관련 유전자들의 존재가 확인되었다. 흥미롭게도, p1_ C-5-1 플라스미드에서 한국 및 일본 등에서 보고된 서열과 매우 상동성이 높은 tet(D) 유전자가 확인되었으며, 이는 국내 양식 분야에서 Edwardsiella spp.의 항생제 내성 전파와 밀접한 연관성이 있을 것으로 추정된다. 본 연구 결과는 국내 양식 어류에서 출현하는 잠재적 인수공통 감염능을 지닌 E. anguillarum의 특성과 더불어 향후 해당 세균에 의한 감염증 제어 기법 발굴을 위한 기반 자료로 활용될 것이다.

This research was supported by the National Research Foundation of Korea (NRF-2020R1I1A2068827), funded by the Ministry of Education, and by the Development of Technology for Biomaterialization of Marine Fisheries Byproducts of the Korea Institute of Marine Science & Technology Promotion (KIMST), funded by the Ministry of Oceans and Fisheries (KIMST-20220128).

The authors have no conflict of interest to report.