First Report and Pathogenicity of Vibrio campbellii (VC_AHPND) Isolated in South Korea

1. Introduction

Of the aquatic invertebrate species that are farmed worldwide, shrimp comprise a fast-growing species with a wide global market, with approximately 4 million tons being produced annually [1]. However, infectious diseases can impede productivity in shrimp farms; additionally, novel pathogens arising from mutations can emerge in farm environments. Acute hepatopancreatic necrosis disease (AHPND), a representative bacterial disease, has occurred frequently in shrimp farms over the past 12 years, causing enormous economic losses of more than $1 billion annually throughout Asia, where approximately 70% of global shrimp is produced [2,3,4]. Before being named AHPND, it was known as early mortality syndrome (EMS) [5]. AHPND was first discovered in China in 2009 [6] and has since spread to various countries in Southeast Asia and North America [7,8,9].

In 2013, the main causative agent of AHPND was identified to be Vibrio parahaemolyticus, a gram-negative, rod-shaped bacterium [10]. V. parahaemolyticus contains the pVA1 plasmid that expresses the fatal binary toxin PirA/B, which causes sudden death in shrimp [11]. This plasmid contains the toxin PirA/B genes of Photorhabdus and Xenorhabdus and is homologous to the insect toxin expression plasmid. When V. parahaemolyticus acquires a plasmid expressing the toxin genes PirA and PirB (15.75 and 56.12 kDa, respectively), it causes AHPND in shrimp [12,13,14,15].

V. parahaemolyticus AHPND is called EMS because it causes high mortality rates of 40–100% for approximately 8–30 d in giant tiger prawns (Penaeus monodon) and white-leg shrimp (Penaeus vannamei) in the post-larval stage after stocking [13,16]. However, AHPND has been shown to affect not only post-larval stages but also various growth stages, including the adult stage [17]. Additionally, shrimp affected by AHPND exhibit extensive necrosis and the detachment of tubular epithelial cells, empty digestive tracts, pale-to-white hepatopancreases, lethargy, anorexia, and slow growth [18,19].

Outbreaks of V. parahaemolyticus AHPND first occurred in China in 2009, followed by Vietnam (2010), Malaysia (2011), Thailand (2012), Mexico (2013), the Philippines (2015), and South America (2016) [11,14,19,20]. Initially, V. parahaemolyticus was identified as the only causative agent of AHPND; however, other Vibrio strains carrying a similar pVA1 plasmid containing PirA/B toxin genes have recently been reported in different parts of the world, including in Vibrio campbellii from Latin America [21], V. owensii from China [16], V. harveyi from Malaysia [22], and V. punensis from South America [23].

In South Korea, AHPND outbreaks caused by V. parahaemolyticus strains were reported in a shrimp farm on the west coast in 2016 and have continued to occur [24], but there have been no reports of AHPND outbreaks caused by other Vibrio spp. strains. In this study, we identified a novel Vibrio strain encoding an AHPND-causing toxin gene. The strain was isolated from a domestic vannamei shrimp farm, identified, and evaluated for its characteristics and toxicity.

2. Results

2.1. Bacterial identification via PCR

To identify the collected bacteria, PCR was performed on the HJ-2023 strain isolated from the hepatopancreases of white-leg shrimp and on the control bacteria (VP_AHPND and VC_non-AHPND). PCR analysis was performed using primers targeting the 16S rRNA and rpoD genes that can detect the Vibrio strains HJ-2023, VP_AHPND, and VC_non-AHPND. The toxR gene that detects V. parahaemolyticus was detected only in VP_AHPND, and the vch gene sequence that detects V. campbellii was detected in the isolated HJ-2023 strain and VC_non-AHPND. Similarly, only HJ-2023 and VP_AHPND were positive for gene sequences targeting the virulence genes PirA (AP3) and PirB, which were present in the pVA1 plasmid (Figure 1).

2.2. Phylogenetic tree based on pathogenic rpoD

Phylogenetic analysis of the rpoD gene grouped the isolated HJ-2023 strain with V. campbellii strains registered in GenBank and showed a close relationship between V. campbellii and V. harveyi (Figure 2).

2.3. Biochemical characteristics

The biochemical properties of the isolated HJ-2023 strain were determined using API Tests 20E and 20NE. Biochemical tests revealed differences in ODC, GEL, ARA, PNPG, MNE, and GNT between the AHPND and non-AHPND strains (Table 1). Furthermore, differences were observed in GLU, NAG, ADI, and CIT between the VP_AHPND and HJ-2023 strains. The biochemical reactions of the isolated HJ-2023 strain were different from those of the existing strains; the HJ-2023 strain showed positive reactions for ONPG and ESC (Table 1 and Table 2).

2.4. Antibiotic susceptibility test

An antibiotic susceptibility test of the isolated HJ-2023 strain and the control bacteria VP_AHPND and VC_non-AHPND revealed that the strains were resistant to streptomycin, erythromycin, cephalothin, kanamycin, vancomycin, ampicillin, clindamycin, and lincomycin. Furthermore, the strains exhibited sensitivity to norfloxacin, nalidixic acid, chloramphenicol, oxytetracycline, and tetracycline. The isolated HJ-2023 strain and VP_AHPND showed an intermediate level of resistance to ciprofloxacin, whereas VC_non-AHPND showed sensitivity. The isolated HJ-2023 strain also showed an intermediate level of resistance to gentamicin, whereas the VP_AHPND and VC_non-AHPND strains were gentamicin-resistant (Table 3).

2.5. Comparison of the virulence of the isolated HJ-2023 strain and that of the controls

After inoculating white-leg shrimp with the isolated HJ-2023 strain and the control strains, the cumulative mortality rate was observed for 5 d. In the high challenge concentration (2×10⁶ CFU/50 µL) experiments, the average cumulative mortality rates were 90%, 95%, and 10% in the isolated HJ-2023 strain, VP_AHPND, and VC_non-AHPND groups, respectively, compared to 0% mortality in the control group (no bacteria inoculated) at 12 h after infection. Comparing the virulence of the isolated HJ-2023 strain and VC_non-AHPND groups with that of the PBS control group further showed average cumulative mortality rates of 95% and 85%, respectively, in the high concentration challenge experiment and 30% and 0%, respectively, in the low concentration challenge experiment at 72 h. In the low challenge concentration (2×10⁵ CFU/50 µL) experiments, the average cumulative mortality rates of the shrimp in the HJ-2023, VP_AHPND, and VC_non-AHPND groups were 40%, 55%, and 0%, respectively. The cumulative mortality rate was maintained for 120 h (Figure 3).

3. Discussion

AHPND can cause serious economic losses by affecting major aquaculture species, such as Litopenaeus vannamei, Penaeus monodon, and M. rosenbergii [18], and has been listed by the World Organization for Animal Health in 2017 [10]. In South Korea, AHPND was designated as a statutory infectious disease for aquatic life in species subject to slaughter in 2021 [1]. White-leg shrimp, which is the most important species in the global shrimp farming industry and accounts for the majority of shrimp in South Korea, with high production and consumption rates, is one of the main species affected by AHPND [24].

We found that the bacteria from the white-leg shrimp farm were positive for the PirA/B toxin genes; however, the presence of V. parahaemolyticus could not be detected using PCR. Therefore, we assumed that it was the AHPND of another species and identified and characterized the isolated strain. In the case of Vibrio spp., it is difficult to distinguish between species because of their high genetic similarity. In particular, the 16S rRNA gene sequence of the Harveyi clade, which causes AHPND, shares more than 99% sequence homology with other species [25]. Therefore, we performed a PCR analysis using the rpoD and toxR gene sequences, which showed the minimum intraspecific similarity between V. parahaemolyticus and other Vibrio spp. [26]. It was confirmed that the isolated strain was not V. parahaemolyticus. The isolated strain was suspected to be V. campbellii among various Vibrio spp. Therefore, PCR was performed targeting the Vch gene, a V. campbellii detection gene, and the bacterium was identified as V. campbellii. The presence of the pVA1 plasmid containing the AHPND-inducing toxin genes (PirA/B) was confirmed using PCR followed by gene sequencing analysis. Most of the bacterial strains harboring pVA1 had 99.92–100% nucleotide sequence homology with each other [27]; the gene sequence from the isolated strain was 100% identical to the PirA/B toxin gene sequences of VP_AHPND, the control strain used in this study (data not shown). The isolated strain was named V. campbellii AHPND (HJ-2023), and its characteristics and virulence were studied.

The PirA/B binary toxin is essential for the development of pathogenic lesions in AHPND.

According to reports on the PirA/B toxin genes present in pVA1, V. parahaemolyticus strains lacking PirA/B fail to cause AHPND; however, the strains transformed with plasmids containing PirA/B have been found to show clinical signs of AHPND and cause 100% mortality in shrimp larvae [28,29]. Additionally, pVA1 can be transferred among similar clusters and is capable of self-transfer using the pndA PSK gene system, which is a mobilizing gene associated with conjugative transfer [14]. Therefore, the pVA1 plasmid has the potential to be transferred among different Vibrio spp. or bacterial species through horizontal transfer [14,15,30]. Dong et al. [27] showed that the pVA1 plasmid can be transferred from AHPND-causing V. campbellii to non-AHPND-causing V. owensii in a controlled laboratory environment. The authors further provided direct evidence of horizontal plasmid transfer among Vibrio species. Accordingly, it is highly likely that the pVA1 plasmid from the V. parahaemolyticus was horizontally transferred to the V. campbellii strain.

Based on the phylogenetic analysis, the isolated HJ-2023 strain was classified into the V. campbellii group and confirmed to be biochemically different from the genetically similar V. harveyi group using biochemical tests [31]. The AHPND-causing gene from the isolated HJ-2023 strain also showed 100% similarity to the rpoD gene of VC_AHPND, which was previously reported in China and Thailand [20,32]. To date, the Vibrio spp. that have been reported to cause AHPND are V. campbellii, V. harveyi, and V. owensii, all of which belong to the Harveyi clade of the Vibrionaceae family and are considered as the pathogenic clade of AHPND [33]. However, a recent study reported that V. punensis, a member of the Orientalis clade, was introduced with a toxin plasmid that causes AHPND [23]. Therefore, AHPND is not limited to one bacterial clade, and there is a possibility that the toxin plasmid can be transferred to various bacterial strains. This is likely to increase the virulence and spread of the disease via transfer of the toxin plasmid to potential bacterial pathogens via various routes.

Based on the biochemical characteristics determined using API Test 20E and 20NE and the percentage of identification (% ID), the isolated HJ-2023 strain showed 99.3% and 84.8% similarity to V. vulnificus and V. alginolyticus, respectively. VC_non-AHPND was identified as V. parahaemolyticus (81.8% similarity) and V. alginolyticus (97.9% similarity). Additionally, VP_AHPND was found to be similar to V. parahaemolyticus (99.9%). A previous study reported that the classification of Vibrio species based on the biochemical characteristics of the API test is somewhat inappropriate [34]. As a result of the reading, however, the control VP_AHPND was clearly classified as V. parahaemolyticus, and the isolated HJ-2023 and VC_non-AHPND strains were classified as V. alginolyticus and V. vulnificus. The isolated HJ-2023 strain was confirmed to be V. campbellii via gene sequence identification, but the same result was not obtained in the bacterial identification based on the biochemical characteristics determined using the API tests, as the information on V. campbellii was not registered in the bacterial identification reading software; therefore, it was likely confirmed as a similar Vibrio species. Because of this diversity in biochemical characteristics, when bacteria are isolated from diseased white-leg shrimp using TCBS medium, approximately 84.3% of them form green colonies [22]. V. campbellii also forms green colonies because it lacks a sucrose hydrolase gene, making it difficult to distinguish it from V. parahaemolyticus [20].

In the antibiotic disc susceptibility test, both the isolated HJ-2023 strain and the control bacteria V_PAHPND and VC_non-AHPND showed strong resistance to ampicillin, clindamycin, and lincomycin, with a small inhibition area of 0 mm. Most Vibrio strains are resistant to ampicillin and lincomycin; therefore, they showed similar results [35]. The isolated HJ-2023 strain showed strong resistance against erythromycin, which is effective against gram-positive bacteria, in an area of 0 mm, whereas VP_AHPND and VC_non-AHPND showed weak resistance. The isolated HJ-2023 strain further showed weak sensitivity to gentamycin, but VP_AHPND and VC_non-AHPND showed resistance. Taken together, these results indicate that even within the same Vibrio genus, susceptibility and resistance to antibiotics differ depending on the species [36].

To compare the pathogenicity of the isolated HJ-2023 strain and that of the control bacteria VP_AHPND and VC_non-AHPND, we observed the cumulative mortality rate after infection in white-leg shrimp. When the shrimp were artificially challenged with high concentrations of the isolated HJ-2023 strain and VP_AHPND strains, more than 90% mortality was observed within 12 h. These results were similar to those of multiple studies that showed 100% mortality within 12–36 h in VP_AHPND and VC_AHPND artificial infection experiments [27,30]. In the low concentration experiments, when the concentration was 10-times lower than the high concentration, the cumulative mortality rates of the isolated HJ-2023 strain and VP_AHPND were 30% and 55% within 72 h, respectively. Accordingly, the isolated HJ-2023 strain showed approximately 25% less toxicity than VP_AHPND; however, it is undoubtedly a lethal strain when outbreaks occur in aquaculture fields.

V. campbellii is widely distributed in marine environments and is an important pathogen in marine shrimp and fish [37]. For a long time, V. campbellii was often misidentified because of its high similarity to V. harveyi and was considered non-pathogenic to shrimp. However, according to the results of the present study, although VC_non-AHPND itself causes mortality in shrimp, V. campbellii is conferred this ability by the pVA1 plasmid and appears to potentially cause stronger pathogenicity in shrimp. AHPND-causing V. harveyi is more virulent than the V. harveyi involved in non-AHPND [22]. The present study indicates that when devising measures to prevent the spread of AHPND, quarantine focusing on the V. parahaemolyticus strain, which is believed to be a unique strain of AHPND, is not a clear solution. In the present study, even the same Vibrio spp. showed various differences in their antibiotic susceptibility and biochemical characteristics. These results highlight the complexity of AHPND-causing strains and their virulence through multiple pathways, which may increase the risk of AHPND in the shrimp aquaculture industry.

Additional studies are needed to determine how the biochemical properties of VC_non-AHPND change as toxin plasmids are acquired. The present study may serve as a basis for attaining a better understanding of and preventing AHPND caused by various bacteria in shrimp aquaculture. Based on the results of this study, effective biosecurity measures should be considered to prevent the spread of AHPND caused by various bacterial strains.

4. Materials and methods

4.1. Sampling and polymerase chain reaction (PCR)

Penaeus vannamei was sampled from a shrimp farm in Chungcheongnam-do, Taean-gun, South Korea. The hepatopancreases of five randomly selected shrimp were aseptically obtained and plated on thiosulfate citrate bile salt sucrose (TCBS) (Difco, Franklin Lakes, NJ, USA). The plate was then incubated at 27 °C for 48 h, and a single colony was selected and subjected to PCR. The control strains used included V. parahaemolyticus [VP_AHPND (ATCC 17802)], which causes AHPND, and V. campbellii [VC_non-AHPND (KCTC 2716)], which does not cause AHPND. Primer sets used to identify the Vibrio species are provided in Table 4. For the PCR, a 20-μL reaction was set up with 1 μL of each colony dilution as a template, 1 μL of each primer, and 10 μL of ExPrime Taq DNA Premix (Genetbio, Daejeon, South Korea). The thermal cycling conditions were as follows: initial denaturation at 95 °C for 10 min, followed by 30 cycles of 5 min denaturation at 95 °C, 30 s annealing at 42 °C (for PirA), 45 °C (for PirB), 50 °C (for rpoD and AP3), 52 °C (for vch), and 54 °C (for 16S rRNA and toxR), and extension at 72 °C for 1 min, with a final extension step at 72 °C for 7 min. The PCR products were electrophoresed on 1% agarose gels and analyzed using a Gel-Doc UV transilluminator (Bio-Rad, Hercules, CA, USA). We named the identified Vibrio campaberii AHPND strain as HJ-2023.

4.2. Phylogenetic analysis based on pathogenic rpoD

The rpoD gene amplified via PCR from the isolated HJ-2023 strain was cloned using pGEM®-T Easy Vector Systems (Promega, Madison, WI, USA). Sequencing was performed by COSMOgenetech (Seoul, Korea). The sequences were further compared with those of other available Vibrio core group member (V. campbellii, V. harveyi, V. rotiferianus, V. parahaemolyticus, and V. alginolyticus) rpoD gene sequences using a BLAST search in the National Center for Biotechnology Information GenBank database (www.ncbi.nlm.nih.gov/BLAST). The rpoD gene sequences were then aligned, and a phylogenetic tree was constructed by bootstrapping 1,000 replicates using the neighbor-joining method in Molecular Evolutionary Genetic Analysis 11 software (http://www.megasoftware.net/).

4.3. Biochemical tests

The biochemical identification of the isolated HJ-2023 strain and control bacteria, VP_AHPND and VC_non-AHPND, was performed using commercially available API 20E and API 20NE TEST Kits (BioMérieux, Marcy-l'Étoile, France). Each bacterial suspension was prepared in 1.5% NaCl solution and analyzed at 37 ± 1 °C for 24 h according to the manufacturer's instructions. The results were interpreted on the APIWEB^TM (BioMerieux) website.

4.4. Antimicrobial susceptibility test

For the antibiotic susceptibility test, the isolated HJ-2023 strain and the control bacteria VP_AHPND and VC_non-AHPND, at 0.5 McFarland turbidity and in total volumes of 50 µL, were spread on Luria broth (LB) agar plates containing 1.5% NaCl. Fifteen types of antibiotic discs were then prepared and incubated at 27 °C for 24 h. After the incubation period, the diameters of the transparent area surrounding the antibiotic discs were measured in millimeters to calculate the inhibition area. The strains were classified into three grades—Resistant (R), Intermediate (I), and Sensitive (S)—based on the standards of the Clinical and Laboratory Standards Institute. The types and concentrations of the 15 antibiotic discs used for the strains are as follows: ciprofloxacin (CIP; 5 µg), norfloxacin (NoR; 10 µg), gentamicin (GM; 10 µg), nalidixic acid (NA; 30 µg), chloramphenicol (C; 30 µg), oxytetracycline (T; 30 µg), streptomycin (S; 10 µg), tetracycline (TE; 30 µg), erythromycin (E; 15 µg), cephalothin (CF; 30 µg), kanamycin (K; 30 µg), vancomycin (VA; 30 µg), ampicillin (AM; 10 µg), clindamycin (CC; 2 µg), and lincomycin (L; 2 µg).

4.5. In vivo experiments to assess pathogenicity

To compare the virulence of the isolated HJ-2023 strain and the control bacteria VP_AHPND and VC_non-AHPND, in vivo tests were performed on white-leg shrimp. White-leg shrimp (average weight: 25 ± 5 g) were collected from a prawn farm located in Gunsan-si, Jeollabuk-do, placed in a breeding tank maintained at 25 °C for 1 week, and used for in vivo infection experiments. The water used for breeding the shrimp was maintained with 30% salinity using sea salt (KENT Marine Reef Salt, Franklin, WI, USA). To confirm the presence or absence of an AHPND infection, the hepatopancreases of five white-leg shrimp were isolated, and non-infection was confirmed via PCR. The strains (isolated HJ-2023 and the control bacteria VP_AHPND and VC_non-AHPND) were cultured in LB supplemented with 1.5% NaCl and incubated at 27 °C with shaking until the absorbance at 600 nm (OD₆₀₀) reached 1.1. The shrimp were then divided into two replicates, with seven groups in total (a control group and six experimental groups). Each group comprised 10 randomly assigned white-leg shrimp. Each group was inoculated via intramuscular injection of phosphate-buffered saline (control group) or 2×10⁶ CFU/50 µL (high challenge concentration) and 2×10⁵ CFU/50 µL (low challenge concentration) of bacteria. After inducing the infection, the cumulative mortality rate was observed for 5 d.

5. Patents