Species-specific PCR detection of the fish pathogen, Vibrio anguillarum, using the amiB gene, which encodes N-acetylmuramoyl-l-alanine amidase


  • Editor: Jeff Cole

Correspondence: In-Soo Kong, Department of Biotechnology and Bioengineering, Pukyong National University, Busan 608-737, Korea. Tel.: +82 51 620 6185; fax: +82 51 620 6180; e-mail: iskong@pknu.ac.kr


Vibrio anguillarum is the causative agent of the fish disease vibriosis and is the most intensely studied species of Vibrio. In the present study, specific primers and a PCR assay were designed to detect V. anguillarum. The primers were designed to amplify a 429-bp internal region of the V. anguillarum amiB gene, which encodes the peptidoglycan hydrolase N-acetylmuramoyl-l-alanine amidase. PCR specificity was demonstrated by successful amplification of DNA from V. anguillarum and by the absence of a PCR product from 25 other Vibrio strains and various enteric bacteria. The PCR produced a 429-bp amplified fragment from as little as 1 pg of V. anguillarum DNA. The limit of detection for this PCR technique was c. 20 bacterial colonies in 25 mg of infected flounder tissue. These results suggest that this PCR system is a sensitive and species-specific detection method, and is possible to use as a diagnostic tool to detect V. anguillarum.


Vibrio anguillarum is a Gram-negative bacterium that causes hemorrhagic septicemia in fish, a disease that leads to great economic losses in fish farming worldwide. Although this bacterium has been reclassified as Listonella anguillarum based on 5S rRNA gene sequence analysis, it is still commonly referred to as V. anguillarum (MacDonell & Colwell, 1984).

To date, knowledge of the biochemical and molecular biological aspects of V. anguillarum pathogenesis has been rather limited compared with that for Vibrio species that cause human illnesses, such as Vibrio cholerae, Vibrio parahaemolyticus, and Vibrio vulnificus. The major virulence factors of V. anguillarum are the hemolysin, protease, outer membrane protein, flagellin, and hemagglutinating proteins (Toranzo et al., 1983, 1993; Kodama et al., 1984). Recently, putative virulence genes of V. anguillarum were identified by random genome sequencing (Rodkhum et al., 2006a). Lorenzo et al. (2003) have also determined the complete sequence of the virulence plasmid pJM1 from V. anguillarum, which affects marine fish.

The morphological and biochemical identification of V. anguillarum from colonies of mixed Vibrio species isolated from marine environments is both time-consuming and inaccurate. Several methods are commonly used to identify V. anguillarum in environmental samples, including the API 20E system, BioLOG fingerprinting, and Biotype-100 biotyping, although it has been reported that diverse V. anguillarum strains are misidentified as other bacteria when these methods are used (Grisez et al., 1991; Austin et al., 1995; Kühn et al., 1996). Other methods, which include plasmid analysis, ribotyping characterization, pulsed-field gel electrophoresis analysis, and DNA hybridization, are also used to identify V. anguillarum (Skov et al., 1995; Martinez-Picado et al., 1996; Austin et al., 1997; Thompson et al., 2004). PCR methods, because of their relative rapidity and simplicity, have also been developed to identify and detect V. anguillarum. The phylogenetic relationships between V. anguillarum and other Vibrio species have been reported, based on comparisons of the DNA sequences of PCR amplicons generated using specific primers that target the recA and 16S rRNA genes (Dorsch et al., 1992; Kita-Tsukamoto et al., 1993; Urakawa et al., 1997; Thompson et al., 2004). However, these genes are not useful in discriminating between closely related strains, due to the very high degrees of sequence identity among these strains.

Recently, two PCR methods that target the sigma factor σ54 (rpoN) and the hemolysin genes were reported for the specific identification of V. anguillarum (Gonzalez et al., 2003; Rodkhum et al., 2006b). Unfortunately, these methods are also not absolutely specific. Gonzalez et al. (2003) have demonstrated that the annealing temperature is very important in detecting the PCR product using specific primers for the rpoN gene. The expected band also appeared at the normal annealing temperature with Vibrio ordalii, which is known to be a very difficult strain to differentiate from V. anguillarum. A multiplex PCR has been reported for the specific detection of V. anguillarum using primers that target five hemolysin genes (Rodkhum et al., 2006b). This method also fails to discriminate V. ordalii from V. anguillarum reliably. Consequently, there remains a need for specific primers for PCR detection of V. anguillarum strains.

The present study aimed to evaluate the sensitivity and specificity of PCR primers designed to target the amiB gene for the detection of V. anguillarum. In a previous study, the amiB gene, which encodes N-acetylmuramoyl-l-alanine amidase, was isolated from V. anguillarum and characterized (Ahn et al., 2006). This enzyme catalyzes the cleavage of peptide bonds in peptidoglycan. By comparing this sequence with the amiB sequences reported for other Vibrio species, a variable region was identified that should enable specific detection of V. anguillarum. Fish infection studies were also conducted to evaluate this method.

Materials and methods

Bacterial strains, media, and culture conditions

In total, 41 strains from 36 bacterial species were used in this study (Table 1). All of the strains were routinely cultured at their optimum temperatures on brain heart infusion (BHI; Difco) and Luria–Bertani agar (LB; Difco). The strains were stored in BHI and LB with 20% glycerol at−70°C.

Table 1.   Strains used in this study
SpeciesSource of reference
  • *

    E, environmental strain kindly provided by Dr S.I. Park (Pukyong National University, Busan, Korea).

  • L, laboratory collection; E, environmental source.

V. alginolyticusATCC 17749
V. anguillarum (serotype O1)Holmstrøm et al. (2003)
V. anguillarum NB10Milton et al. (1997)
V. anguillarum (J-O-3)*E
V. anguillarum (J-O-2)*E
V. anguillarum (serotype O2)ATCC 19264
V. anguillarum (YT)E
V. campbelliiATCC 25920
V. carchariaeLMG 7890
V. choleraeATCC 14035
V. cincinnatiensisNCTC 12012
V. damselaeE
V. diazotrophicusATCC 33466
V. fluvialisATCC 33809
V. furnissiiATCC 35016
V. harveyiNCMB 1280
V. hollisaeLMG 17719
V. logeiATCC 29985
V. mediterraneiNCTC 11946
V. metschnikoviiLMG 11664
V. mimicusATCC 33653
V. natriegensATCC 14048
V. navarrensisATCC 51183
V. nereisATCC 25917
V. ordaliiATCC 33509
V. orientalisNCMB 2195
V. parahaemolyticusATCC 17802
V. proteolyticusNCMB 1326
V. salmonicidaNCMB 2262
V. tubiashiiATCC 19109
V. vulnificusATCC 27562
Aeromonas hydrophilaNCTC 8049
Escherichia coli BL21 (DE3)L
E. coli XL1-blueL
Enterobacter cloacaeE
Edwardsiella tardaE
Klebsiella oxytocaE
Klebsiella pneumoniaeE
Salmonella typhiE
Shigella flexneriE
S. sonneiE

Design of specific primers for V. anguillarum detection

A segment of the amiB gene sequence was used as the PCR target for the specific detection of V. anguillarum. To design a primer pair, a sequence comparison was made using the known N-acetylmuramoyl-l-alanine amidase genes of Vibrio spp. (cholerae O1, parahaemolyticus, alginolyticus, vulnificus fischeri, splendidus, and angustum), which were retrieved from the Entrez database using the National Center for Biotechnology Information GenBank database and the blast search program. The forward primer van-ami8 (5′-ACAT CATCCATTTGTTAC-3′, positions 8–25 in the V. anguillarum amiB gene), and the reverse primer van-ami417 (5′-CCTTATCACTATCCAAATTG-3′, positions 417–436 in the V. anguillarum amiB gene) were used (Fig. 1), and 16S rRNA- and the rpoN gene-specific primers were used as positive controls (Gonzalez et al., 2003).

Figure 1.

 Alignment of N-acetylmuramoyl-l-alanine amidase amino acid sequences (amiB sequences) of Vibrio anguillarum (accession no. ABD85291), Vibrio cholerae O1 (AAF93517), Vibrio parahaemolyticus (BAC61083), Vibrio alginolyticus (ZP01261820), Vibrio fischeri (YP205709), Vibrio vulnificus (AA009746), Vibrio splendidus (ZP00991953), and Vibrio angustum (ZP01237176). Primer regions are indicated by boxes, and primers are indicated by arrows.

PCR analysis

PCR amplification was performed on colonies of bacteria and DNA purified from each strain as templates. The PCR was carried out in a 50-μL reaction mixture that contained 250 μM of each deoxyribonucleoside triphosphate (dNTP), 10 pmol of each primer, 5 μL of 10 × Taq buffer with MgCl2, 0.5 U of Taq DNA polymerase (Takara Bio, Japan), and distilled water up to 50 μL. The PCR thermocycling with van-ami8 and 417 primer pair involved one initial cycle of denaturation at 95°C for 10 min, followed by 25 cycles of 95°C for 30 s, 56°C for 30 s, and 72°C for 30 s, and finally, one cycle of 72°C for 7 min. When the PCR assay was conducted with a primer set corresponding to the rpoN gene, the annealing temperature was 62°C. The PCR products were analyzed by 1.5% agarose gel electrophoresis in 1 × TAE buffer. A 100-bp DNA ladder (Cosmo Genetech, Korea) was used as the molecular weight marker.

Fish infection and DNA isolation from tissues

One hundred microliters of V. anguillarum O1 suspension (6.4 × 109 CFU mL−1) was injected into the abdomen of healthy flounders (Paralichthys olivaceus) with a body weight of 100 g. Viable cells were determined by cell counting from the internal organs of the fish after injection. The tissues (kidney, liver and intestine) were removed, homogenized in sterile phosphate-buffered saline (PBS) using a homogenizer at 9700 r.p.m. for 20∼30 s, serially diluted 10-fold in saline solution, and dropped onto TCBS (Difco) and BHI agar plates. Colonies were counted after incubation at 25°C for 24 h. For PCR amplification, total DNA was purified from 25 mg of each tissue using the QIAamp DNA Mini Kit (Qiagen, Germany), and 0.2 μg DNA was used as the PCR template.


Sequence analysis and specific primer design

The amiB gene from serotype O1 of V. anguillarum has recently been cloned and sequenced by Ahn et al. (2006). Alignment of the encoded amino acid sequence with the N-acetylmuramoyl-l-alanine amidases from V. cholerae O1, V. parahaemolyticus, Vibrio alginolyticus, Vibrio fischeri, V. vulnificus, Vibrio splendidus, and Vibrio angustum showed 77%, 67%, 67%, 66%, 65%, 65%, and 56% sequence identity, respectively. To design a primer pair, the nucleotide and amino acid sequences were compared with other amiB gene sequences described for the genus Vibrio using the clustal w program (Fig. 1). Regions of high sequence variability among the amiB genes were used for the design of specific primer sets. Among the several primer pairs tested, the van-ami8 and van-ami417 primer pair, which is located between nucleotides eight and 417 of the amiB gene, showed the highest specificity in the PCR assay.

Specificity of PCR detection of V. anguillarum

PCR amplification with colonies of six strains of V. anguillarum using the van-ami8 and van-ami417 primers resulted in a product of the predicted length (429 bp), while no products were obtained from other Vibrio strains (Fig. 2a). When the same experiment was conducted with the enteric bacteria listed in Table 1, no amplified band was seen. 16S rRNA- and the rpoN gene-specific primers were selected for positive controls. The expected 1465-bp amplicon for 16S rRNA gene was seen with every strain used in this study (data not shown). However, in case of the rpoN gene, a 519 bp amplicon only appeared in six strains of V. anguillarum (Fig. 2b).

Figure 2.

 (a) Agarose gel electrophoresis of Vibrio anguillarum-specific DNA products amplified by PCR using the van-ami8 and van-ami417 primers. M, 100-bp DNA ladder; lane 1, Vibrio alginolyticus (ATCC 17749); lanes 2–7, Vibrio anguillarum (O1 type, NB10, J-O-3, J-O-2, ATCC 19264, and YT); lane 8, Vibrio campbellii (ATCC 25920); lane 9, Vibrio carchariae (LMG 7890); lane 10, Vibrio cholerae (ATCC 14035); lane 11, Vibrio cincinnatiensis (NCTC 12012); lane 12, Vibrio damselae; lane 13, Vibrio diazotrophicus (ATCC 33466); lane 14, Vibrio fluvialis (ATCC 33809); lane 15, Vibrio furnissii (ATCC 35016); lane 16, Vibrio harveyi (NCMB 1280); lane 17, Vibrio hollisae (LMG 17719); lane 18, Vibrio logei (ATCC 29985); lane 19, Vibrio mediterranei (NCTC 11946); lane 20, Vibrio metschnikovii (LMG 11664); lane 21, Vibrio mimicus (ATCC 33653); lane 22, Vibrio natriegens (ATCC 14048); lane 23, Vibrio navarrensis (ATCC 51183); lane 24, Vibrio nereis (ATCC 25917); lane 25, Vibrio ordalii (ATCC 33509); lane 26, Vibrio orientalis (NCMB 2195); lane 27, Vibrio pharahaemolyticus (ATCC 17802); lane 28, Vibrio proteolyticus (NCMB 1326); lane 29, Vibrio salmonicida (NCMB 2262); lane 30, Vibrio tubiashii (ATCC 19109); lane 31, Vibrio vulnificus (ATCC 27562). (b) Agarose gel electrophoresis of V. anguillarum-specific DNA products amplified by PCR with the rpoN gene-specific primers. M, 100-bp DNA ladder; lane 1–6, V. anguillarum (O1 type, NB10, J-O-3, J-O-2, ATCC 19264, and YT).

Sensitivity of PCR detection of V. anguillarum

Serially diluted samples were used as templates for 30 cycles of PCR amplification. The van-ami8 and van-ami417 primers were tested for amplification while varying the DNA template concentrations from 1 μg μL−1 to 0.1 pg μL−1. The level of product detectable was tested by agarose gel electrophoresis of an amplified amiB gene band of the expected size (429 bp). The expected PCR product was seen for reactions that contained as little as 1 pg of genomic DNA template (Fig. 3).

Figure 3.

 Sensitivity of the van-ami8 and van-ami417 PCR primer pair for amplification. M, 100-bp DNA ladder; the following lanes contain Vibrio anguillarum DNA: lane 1, 1 μg; lane 2, 0.1 μg; lane 3, 0.01 μg; lane 4, 1 ng; lane 5, 0.1 ng; lane 6, 0.01 ng; lane 7, 0.001 ng; lane 8, 0.0001 ng.

Bacterial cell counting in flounder tissues

Detection of V. anguillarum from internal organs was attempted after injecting the fish with V. anguillarum. A liquid extract of the homogenized tissue was dropped onto TCBS and BHI agar plates and the plates were incubated. The colony counts are expressed as the logarithm of CFU mg−1 of tissue. In the control group, no bacteria were seen on the TCBS plates, whereas colonies, ranging from 1.0 to 1.5 log CFU mg−1 of tissue, did appear on the BHI plates (Table 2). At 24 h postinjection, the numbers of bacterial colonies were 3.9, 4.2, and 3.7 log CFU mg−1 of the kidney, intestine, and liver, respectively, on TCBS plates. To evaluate whether specific detection could be achieved with tissues contaminated by V. anguillarum, PCR amplification was performed using total DNA purified from each infected organ. As shown in Fig. 4, all of the DNA samples purified from infected tissues generated a clear band of 429 bp.

Table 2. Vibrio anguillarum CFU counts in homogenates of infected organs
Sample groupSample fromPlate culture (log CFU mg−1)
Control fish (PBS treatment)Kidney1.40
Infected fish (after 24 h)Kidney5.83.6
Figure 4.

 Agarose gel electrophoresis of PCR products from chromosomal DNA recovered from noninjected or infected tissues of flounder. M, 100-bp ladder; lane 1, 2, and 3, liver, kidney, and intestine tissues from normal flounder without infection; lane 4, 5, and 6, liver, kidney, and intestine tissues from flounder infected with Vibrio anguillarum; lane 7, PCR product with chromosomal DNA purified from V. anguillarum for a positive control.


The availability of rapid and specific diagnostic methods for the detection of the fish pathogen V. anguillarum is important in aquaculture because V. anguillarum causes vibriosis in fish, which results in large economic losses. Many attempts have been made to develop methods for the specific detection of V. anguillarum.

To date, all of the reported methods, including various PCR protocols, have certain problems in specifically detecting V. anguillarum. An important issue is discrimination between V. anguillarum and V. ordalii, which is the most closely related strain to V. anguillarum. The API 20E system has been used in biochemical and serological analyses to type Vibrio species. However, it has been reported that this system cannot distinguish V. anguillarum from V. ordalii. Austin et al. (1995) have reported that V. anguillarum can be distinguished from V. ordalii by ribotyping with an acetylaminofluorene-labeled ribosomal probe. However, in plasmid profiling, V. ordalii has a 32-kb plasmid that corresponds to pHJ101, which is present in V. anguillarum biotype 2. Urakawa et al. (1997) have analyzed 35 Vibrio strains by 16S rRNA gene genotyping and they have reported that the 16S rRNA gene sequences of the Vibrionaceae family show more than 90% homology. In particular, V. anguillarum (ATCC 19264) and V. ordalii (ATCC 35509) share 97% identity. Thompson et al. (2004) have performed a phylogenetic analysis of the Vibrionaceae family that involved the genera Vibrio, Photobacterium, and Grimontia using the recA gene sequence. Their results showed that the recA sequence similarity of V. anguillarum (LMG4437) and V. ordalii (LMG13544) is almost 98%. This high degree of sequence identity does not allow reliable discrimination of specific strains using PCR. To address this problem, the rpoN gene, which encodes sigma factor σ54, can be used to identify V. anguillarum in turbot tissue and blood (Gonzalez et al., 2003). However, even using that sequence, V. anguillarum could not be distinguished from V. ordalii at all annealing temperatures. Using an annealing temperature of 58°C, a band was detectable not only in V. anguillarum but also in V. ordalii (NCIMB 2167). Thus, a better technique is required for the reliable identification of V. anguillarum.

In this study, it was examined whether the amiB gene, which encodes the enzyme N-acetylmuramoyl-l-alanine amidase, could be used to design PCR primers to detect V. anguillarum in a specific fashion. By comparing the available amiB gene sequences from the GenBank database, as a target the sequence that encodes the N-terminal region of the AmiB protein was identified, which shows high variability among strains (Fig. 1). The primers that were designed could be used to generate successfully the predicted PCR product from six strains of V. anguillarum, but not from the other strains listed in Table 1. As mentioned above, it has been reported that when using PCR primers that target the rpoN gene, the annealing temperature of 62°C is very important, to exclude detection of a specific band from V. ordalii. However, in this study, the predicted PCR product was produced only from V. anguillarum, even using an annealing temperature of 55°C (Fig. 2). This result suggests that the sequence similarity of the N-terminal region may be low, although no amiB gene sequence from V. ordalii is yet available.

To investigate whether this PCR system could be used in practical situations, the V. anguillarum O1 strain was injected into flounders. The colony number injected into fish was 6.4±109 CFU mL−1, which is close to the value of LD50 (5.4±109 CFU mL−1). After 24 h, the kidney, intestine, and liver were removed from the injected fish, homogenized, and plated onto BHI and TCBS plates to recover V. anguillarum. The results showed that V. anguillarum proliferates well in these fish organs. The PCR assays were performed using 0.2 μg of purified total DNA extracted from 25 mg of the infected tissues. The 429-bp amplicon was obtained for every sample of infected tissue (Fig. 4). Furthermore, the PCR products were cleanly amplified without interference from other sequences in the total DNA of the fish tissues. To determine the detection limit of the PCR assay, 0.2 μg of purified DNA from the infected flounder tissue was serially diluted and PCR was performed. After PCR amplification, the detectable bands were obtained from 0.2 μg to 1 ng of DNA (data not shown). The templates used for this study contained chromosomal DNAs of V. anguillarum together with flounder tissue. It could be estimated that the amount of chromosomal DNA of V. anguillarum in 1 ng of template corresponded to that of about 20 colonies. Gonzalez et al. (2003) reported that the detection limit of PCR using the rpoN gene was 50–500 colonies per 25 mg of tissue. Thus, the result in this study was more sensitive than the PCR assay with the rpoN gene to detect V. anguillarum.

In conclusion, it is possible to apply the described PCR method for the direct detection of V. anguillarum in marine flounder without the time-consuming bacterial cultivation step. Investigation of the suitability of this assay system for the analysis of other marine organisms and seafood is being continued.


This work was supported by a grant (B-2004-10) from the Marine Bioprocess Research Center of the Marine Bio 21 Center, funded by the Ministry of Maritime Affairs and Fisheries, Republic of Korea. The authors thank D.L. Milton, Department of Cell and Molecular Biology, Umeå University, Sweden, for providing V. anguillarum NB10, and K. Holmstrøm, Biotechnological Institute, Department of Molecular Characterization, Denmark, for providing V. anguillarum serotype O1.