Introduction

Ornamental fish culture is a powerful income and employment source in the world. As a non-food activity, ornamental fish trade provides an excellent opportunity in the aquaculture sector. More than 1 billion ornamental fish are traded internationally, making the industry the most important in global trade (John and Hatha 2013; Sreedharan et al. 2013).

Guppy (Poecilia reticulata) is one of the most popular tropical freshwater aquarium fish in many countries, including Korea. Like other ornamental fish, guppies are prone to various types of viral, bacterial and fungal infections. Bacterial diseases especially caused by Gram-negative species are commonly found in ornamental fish (Lewbart 2001). Environmental factors such as stress, poor water quality, unhygienic handling and polluted feeding can increase the chance of infections among the fish in the captive condition (Kigigha et al. 2012). Furthermore, the deficiency in the immune system of fish increases morbidity and mortality rates (Toranzo et al. 2005).

Aeromonas spp. or aeromonads are the natural microflora of freshwater or brackish water environment (Krishnakumar et al. 2009). To date, there are 36 recognized species in the genus Aeromonas (Navarro and Martínez-Murcia 2018). Psychrophilic or non-motile A. salmonicida is mainly associated with furunculosis in fish while mesophilic or motile Aeromonas spp. such as A. veronii, A. hydrophila, A. dhakensis and A. caviae are particularly documented in humans and animals (Teunis and Figueras 2016). They are regarded as important pathogenic bacteria in fish and may cause motile aeromonad septicemia (MAS). In Korea, several studies have been reported Aeromonas infection in farmed Israeli carp (Cyprinus carpio), cyprinid loach (Misgurnus anguillicaudatus) and mud loach (Misgurnus mizolepis) (Yu et al. 2010, 2015; Jun et al. 2010). Aeromonas infection is a serious threat for freshwater fish production which brings enormous loss to the ornamental fish industry (John and Hatha 2013; Krishnakumar et al. 2009).

The identification of virulence factors is a major constituent to determine the potential pathogenicity due to the multifunctional roles of those virulence factors in Aeromonas pathogenesis (Nawaz et al. 2010). The existence of an assortment of virulence factors allows them to colonize, establish, replicate and damage the host tissues (Yu et al. 2005). The pathogenicity of aeromonads is related to several virulence factors including various extracellular enzymes, cytotoxic and cytotonic enterotoxins, hemolytic toxins and type 3 secretion system (T3SS) (Janda and Abbott 2010).

In the ornamental fish industry, antibiotics have been used since 1970s for controlling bacterial diseases by the ornamental fish producers, retailers as well as fish owners (Trust and Whitby 1976). The uncontrolled antibiotic uses the result in the emergence of antibiotic-resistant bacteria and the expansion of multi-drug resistant Aeromonas spp. (Rowe-Magnus et al. 2002).

As far as we know, there is no report about the identification of both phenotypic and genotypic virulence markers as well as antibiotic susceptibility patterns of Aeromonas spp. isolated from ornamental guppy in Korea. Thus, the objective of our study was to characterize Aeromonas isolates from guppy based on phylogenetic grouping, virulence potential and multi-drug resistance profile.

Materials and methods

Isolation and phenotypic identification of Aeromonas spp.

A total of 52 Aeromonas spp. were isolated from 62 healthy guppies with an average length of 2.20–4.77 cm which were randomly purchased from the pet shops in Seoul, Korea. The live guppies were transported from the pet shops to the laboratory in separate sterile polyethylene bags within 2 h.

Each fish was humanely euthanized with an overdose of MS222 (250 mg/l, buffered) in separate test tubes (AVMA 2013). The whole body of the fish was homogenized with 3 ml of sterile physiological saline. One ml of sample was enriched in 9 ml of alkaline peptone water at 35 °C for 24 h. One loopful from each homogenate was streaked onto a plate of Aeromonas agar base (MB Cell, LA, CA) and incubated overnight at 35 °C. The typical green colonies with dark green center assuming as Aeromonas spp. were subcultured on tryptic soy agar (MB Cell, LA, CA) and tested for indole, oxidase, H2S and starch hydrolysis tests (Hossain et al. 2018a). The genomic DNA from phenotypically identified Aeromonas species was extracted using Exgene Cell SV kit (GeneAll, Seoul, Korea). For molecular identification, PCR amplification and sequencing of the gyrB gene were performed. The primers selection, PCR amplification and sequencing were done according to a previous report (Yáñez et al. 2003). Taxonomic identification of the DNA sequences was done by means of the BLAST option in GenBank (http://blast.ncbi.nlm.nih.gov/).

Genetic distance and phylogenetic analysis of gyrB sequences

Multiple sequence alignments were carried out using ClustalW in MEGA 7.0 software program (Kumar et al. 2016). The gyrB sequences from all isolates and their corresponding reference sequences were included in the alignment to determine the genetic distance within and between the species. The genetic distances were obtained using Kimura’s two-parameter model.

To construct an unrooted phylogenetic tree, gyrB reference sequences of A. veronii (Accession no.: AF417626.1), A. hydrophila (Accession no.: AF417622.1), A. caviae (Accession no.: AJ868400.1), A. dhakensis (Accession no.: AM262163.1) and A. enteropelogenes (Accession no.: AJ868392.1) were obtained from NCBI database. The evolutionary tree was created by the neighbor-joining method with 1000 bootstrap replicates.

Phenotypic characterization of virulence factors

To determine the phenotypic virulence factors, several phenotypic pathogenicity tests were carried out at 35 °C for 24 h. β-hemolysis activity was assessed using sheep blood agar (MB Cell, LA, CA). Slime production was assessed by means of tryptic soy agar supplemented with 5% sucrose and 0.08% Congo red (Freeman et al. 1989). Caseinase production was perceived in nutrient agar supplemented with 10% skim milk (Abd-El-Malek, 2017). DNase activity was tested by DNase agar plates adding 1 N HCl. The gelatin media prepared by 3% beef extract, 5% peptone and 15% gelatin with a final pH 7.0 to detect gelatin production. Lipase production was screened using Tryptic soy agar plates supplemented with 1% Tween 80 (v/v) (Harley and Prescott 2002).

Genotypic characterization of virulence factors

PCR was carried out to determine the presence of 11 virulence genes including act, alt, ast, aer, hly, ser, lip, fla, ascV, ahyB and gcaT. The PCR amplification of the virulence genes was done by means of PCR primers and conditions described in Table 1.

Table 1 Primers used for the detection of virulence genes
Full size table

The PCR mixture of 20 µl contained 2 µl Taq buffer, 1.6 µl dNTP, 14.2 µl nuclease-free water, 0.2 µl AmpOne Taq DNA polymerase (GeneAll, Seoul, Korea), 1 µl template and 1 µl of 10 µM of both forward and reverse primers. The PCR products were observed by electrophoresis on 2% (w/v) agarose gel. A 100-bp DNA ladder (Invitrogen, San Jose, CA) was used as a molecular size reference.

Antibiotic susceptibility testing

The antibiotic susceptibility patterns of the isolates were determined by the antibiotic disk diffusion test using commercially available 17 antibiotic disks of 10 antibiotic groups.

The following antibiotics were used: Penicillins; amoxicillin (30 µg), ampicillin (10 µg), Cephalosporins; cephalothin (30 µg), cefotaxime (30 µg), ceftriaxone (30 µg), Carbapenem; imipenem (10 µg), Quinolones; nalidixic acid (30 µg), ofloxacin (5 µg), ciprofloxacin (5 µg), Tetracyclines; tetracycline (30 µg), oxytetracycline (30 µg), Folate pathway inhibitor; trimethoprim/sulfamethoxazole (1.25/23.75 µg), Ansamycin; rifampicin (5 µg), Phenicol; chloramphenicol (30 µg) and Aminoglycosides; kanamycin (30 µg), amikacin (30 µg), gentamicin (10 µg). Escherichia coli ATCC 25922 was used as a quality control strain.

The resistance phenotype was interpreted according to the Clinical and Laboratory Standards Institute, CLSI guidelines (CLSI 2014).

Results

Aeromonas spp. identification

The sequences of gyrB gene demonstrated that the isolates could be identified as 5 species, namely A. veronii (n = 34), A. dhakensis (n = 10), A. hydrophila (n = 3), A. caviae (n = 3) and A. enteropelogenes (n = 2). The gyrB gene sequences of the isolates were deposited in Genbank.

Percentage of nucleotide substitutions were calculated for a continuous segment of 917 nt. The sequence similarities of the Aeromonas species ranged from 97 to 99% in BLAST comparison of the NCBI database. The nucleotide substitution rates in the intraspecies levels varied from 0 to 2.6%. The inter-species sequence divergence ranged from 2.7 to 9.2%.

Phylogenetic analysis

An unrooted neighbor-joining phylogenetic tree was constructed from the nucleotide sequence alignment (Fig. 1). The phylogenetic tree was established which divides into two major clusters. One subsection included A. veronii whereas the other section included A. hydrophila, A. enteropelogenes, A. dhakensis and A. caviae. Every isolate grouped with their respective reference sequence.

Fig. 1
figure 1

Unrooted neighbor-joining phylogenetic tree based on gyrB sequences of Aeromonas spp. isolated from guppy (Poecilia reticulata). The GenBank sequences are labelled as (∆) and the isolates including A. hydrophila, A. veronii, A. caviae, A. dhakensis and A. enteropelogenes are marked as AH, AV, AC, AD and AE, respectively

Full size image

Phenotypic characterization of virulence factors

Phenotypic virulence factors were analyzed among the tested species by several pathogenicity tests including the production of caseinase, DNase, gelatinase and lipase enzymes, β-hemolysis activity and slime production (Table 2).The β-hemolysis, slime, caseinase, DNase, gelatinase and lipase production were observed in 28 (53.9%), 33 (63.5%), 28 (53.9%), 42 (80.8%), 37 (71.2%) and 42 (80.8%) isolates, respectively.

Table 2 Distribution of virulence factors in guppy-borne Aeromonas species
Full size table

Prevalence of virulence genes in Aeromonas species

The presence of 12 virulence genes was analyzed by PCR in this study (Table 2). Each strain contained at least 2 virulence genes. Two A. dhakensis and one A. veronii showed a compliment of 10 virulence genes which is the highest frequency among the isolates (data not shown). The virulence genes including act (84.6%), hly (80.8%), aer (73.1%), lip (73.1%), gcaT (73.1%), ascV (53.8%), ahyB (53.8%) fla (51.9%), alt (48.1%), ast (36.5%) and ser (34.6%) were detected among the isolates.

Antibiotic susceptibility patterns

The antibiotic susceptibility patterns of Aeromonas species are shown in Table 3. All the isolates were resistant to at least four antibiotics belonging to ≥ 4 structural classes.

Table 3 Antibiotic susceptibility patterns of different Aeromonas spp. isolated from guppy
Full size table

All isolates showed resistance to amoxicillin, nalidixic acid and oxytetracycline. The resistance for ampicillin (92.3%), rifampicin (76.9%), imipenem (71.2%) was prevalent among the guppy-borne isolates. The resistance rates of cephalothin, tetracycline, trimethoprim/sulfamethoxazole, gentamicin, kanamycin, chloramphenicol, ciprofloxacin, ofloxacin, cefotaxime and ceftriaxone were 51.9, 51.9, 50.0, 21.2, 13.5, 5.8, 5.8, 3.9, 3.9 and 1.9%, respectively.

Discussion

The taxonomy of genus Aeromonas is complex due to the phenotypic and genotypic discrepancies at the species level. The 16S rRNA sequencing which is the most common molecular tool for taxonomic studies of the bacteria has proven ineffective for distinguishing different Aeromonas species because of low taxonomic resolution (Nagar et al. 2013). On the other hand, some housekeeping genes were reported with considerably higher discriminatory power than 16S rRNA, showing 92% and 89% average sequence similarity for gyrB and rpoD, respectively (Soler et al. 2004; Nagar et al. 2013). In this study, nucleotide sequencing of the gyrB gene has proved to be a suitable tool for taxonomic and phylogenetic studies in Aeromonas spp. (Figueras et al. 2011; Hoel et al. 2019). The intraspecies nucleotide substitution rates of gyrB gene sequences were found as < 3%. Lower nucleotide substitution rates (< 3%) were observed in a few strains in interspecies level and the only exception was A. hydrophila with nucleotide substitution of 2.7%. Nucleotide substitution rates generally range from 0 to 2% in the intraspecies level while the substitution rate > 3% is considered as different species (Soler et al. 2004).

All Aeromonas spp. isolates were assayed for six phenotypic pathogenicity tests (Table 2). The Extracellular lipase plays an important role in the establishment of infections in humans by disturbing immune system functions and DNase allows the propagation of the bacteria (Tomas 2012). According to our results, 42 (80.8%) of the isolates were positive for lipase and DNase enzyme production and 37 (71.2%) isolates were positive for gelatinase activity. The majority of ornamental fish-borne Aeromonas spp. were also found to produce lipase, DNase and gelatinase in a previous study (John and Hatha 2013).

The slime production reflects the bacterial capacity to adhere in the specific host tissues and produce invasive micro-colonies. The slime production was found in 33 (63.5%) of the guppy-borne aeromonads. The slime production is responsible for acting against phagocytosis, chemotaxis and the action of antibiotics (Tomas 2012). Hemolysis activity is strongly related to enterotoxin production (Burke et al. 1983). The production of serine protease (caseinase) is responsible for the proteolysis activity and aid to the maturation of exotoxins (e.g. aerolysin). The current study could detect 28 (53.9%) β-hemolysis and caseinase positive isolates from guppies. Similarly, about 44.2% of β-hemolytic Aeromonas spp. were isolated from fish and beef (Arslan and Kucuksari 2015). In contrast to our results, 95.9% of caseinase producing isolates were reported by the same study.

The cytotoxic enterotoxin related act gene (84.6%) was highly distributed among the isolates (Table 2). The act gene encoding a cytotoxin (Act) was highly found with a high frequency in the Aeromonas isolates from different sources such as seafood (75.0%) and drinking water (70.0%) (Sen and Rodgers 2004). The enterotoxin genes including alt (48.1%) and ast (36.5%) were less prevalent in the isolates which is in agreement with a previous study (Yano et al. 2015). Among the enterotoxins, act gene is one of the most important virulence determinants linked to hemolytic, cytotoxic and enterotoxic actions. The enterotoxin genes, alt and ast were also observed which are mainly related to diarrhea that influence fluid secretion in the intestine of animals (Nawaz et al. 2010). The high occurrence of aer and hly genes was identified in 73.1% and 80.8% isolates, respectively. Wong et al. (1998) proposed a two-toxin model for aerA and hlyA genes in which both genes equally contributed to Aeromonas virulence and they might be knocked out to reduce the virulence. We observed a heterogeneous distribution of all toxin genes in A. veronii, A. dhakensis and A. hydrophila. None of the A. enteropelogenes harbored act, alt, ast and hlyA genes and no hemolysin genes (aer and hlyA) were found in A. caviae isolates. The lack of specific genes might be attributed to certain species while considering the diversity in the distribution of various virulence genes among the isolates from different sources and the possibility of horizontal gene transfer of virulence determinants among the species (Hoel et al. 2019).

The extracellular enzymes of Aeromonas spp. are generally secreted by means of well-categorized type II or general secretory pathway in the establishment of the pathogenicity (Pemberton et al. 1997). The majority of Aeromonas strains exhibited extracellular lipase and glycerophospholipid-cholesterol acyltransferase related lip (73.1%) and gcaT (73.1%) genes. The presence of other extracellular enzyme encoded genes such as elastase (ahyB) (53.8%), T3SS (ascV) (53.8%) and serine-protease (ser) (34.6%) genes were present. These enzymes have been increasingly linked to microbial virulence since they are involved in the activation of aerolysin (aerA), degradation of host cell components, enhancing adhesion, contributing to cell nutrition and evade host cell (Nawaz et al. 2010).

In the present study, polar flagella related fla (51.9%) gene was observed among the strains. Bacterial flagellar motility enabled by a lateral or polar flagellum, have several biological roles in pathogens such as chemotaxis, adhesion and invasion (Haiko and Westerlund-Wikström 2013).

A misalliance between genotypic and phenotypic virulence factors was noticed in the aeromonads isolated from guppies. All A. enteropelogenes produce lipase in the phenotypic pathogenicity tests but failed to amplify lip gene in PCR assay. Also, every A. hydrophila lacked ser gene but could produce caseinase. In contrast, phenotypically negative A. enteropelogenes for β-hemolysis in sheep-blood agar conferred aer gene in PCR amplification. This issue has been reported previously in Aeromonas spp. in which the mismatch was observed with different primer sets (Yano et al. 2015). This discrepancy might be attributed to the genetic variability of virulence determinants in Aeromonas as previously suggested (De Silva et al. 2018). However, further studies are needed for the assessment of the specific mechanisms of genetic variability.

The antibiotic susceptibility patterns of Aeromonas species are shown in Table 3. All isolates were resistant to at least four antibiotics belonging to ≥ 4 structural classes. The high occurrence of β-lactam (penicillin, cephalosporin and carbapenem) antibiotics was observed among the isolates. The higher percentage of β-lactam resistant isolates was found due to the natural production of chromosomal β-lactamases by aeromonads (Tayler et al. 2010). Amoxicillin resistance was displayed in all guppy-borne isolates (Table 3). Resistance to ampicillin was shown by 48 (92.3%) isolates. This result is similar to a recent report where the majority of goldfish-borne Aeromonas spp. showed resistance against amoxicillin and ampicillin (Hossain et al. 2018a). High resistance to first-generation cephalosporin (cephalothin) and a decreased susceptibility to third-generation cephalosporins (cefotaxime and ceftriaxone) were previously reported in fish-borne aeromonads (Rall et al. 1998; Imziln 2001). In our study, the resistance cephalothin has been observed in 27 (51.9%) isolates and A. enteropelogenes (100.0%), while 2 (5.9%) and 1 (2.9%) of A. veronii isolates showed cefotaxime and ceftriaxone, respectively. In addition, imipenem resistance was observed in 37 (71.2%) isolates including A. dhakensis (80.0%), A. veronii (79.4%) and A. hydrophila (66.7%) whereas all A. caviae and A. enteropelogenes isolates were susceptible to imipenem. Castro-Escarpulli et al. (2003) have also reported the occurrence of imipenem resistance among the Aeromonas spp.

In our study, quinolone, folate pathway inhibitor and tetracycline-resistant Aeromonas species were also observed with a high percentage. In the present study, every isolate showed resistance to nalidixic acid. Resistance to ciprofloxacin was identified in A. caviae (33.3%) and A. veronii (5.9%) isolates and ofloxacin resistance was observed in only A. caviae (33.3%) and A. veronii (2.9%). In the case of the tetracycline group, all isolates showed resistance against oxytetracycline, and 52.9% of the isolates were resistant to tetracycline. Also, all A. enteropelogenes showed susceptibility to trimethoprim/sulfamethoxazole while 50.0% of the aeromonads displayed resistance to this antibiotic. Indeed, quinolone, tetracycline and folate pathway inhibitor antibiotic groups had been extensively used in clinical medicine, veterinary and agriculture sectors for decades contributing to a higher level of bacterial resistance. Therefore, the resistance phenotype against these important antibiotic group among the guppy-borne Aeromonas spp. should not be underestimated.

The results for the aminoglycoside group (gentamicin, kanamycin and amikacin) were also observed. None of the isolates were resistance to amikacin. Besides, gentamicin was resistant to 11 (21.2%) isolates except for A. enteropelogenes. Kanamycin resistance was displayed in 7 (13.5%) isolates including A. veronii (2.9%), A. dhakensis (50.0%) and A. hydrophila (33.3%). Similarly, the high incidence of gentamicin and kanamycin resistance was observed in goldfish-borne and zebrafish-borne aeromonads (Hossain et al. 2018a, b). Finally, rifampicin resistance was exhibited by 40.0% of the isolates. However, only 3 A. veronii (8.8%) isolates were resistant to chloramphenicol. In the previous study, 86.2% and 4.6% of the isolates were resistant to rifampicin and chloramphenicol, respectively (Hossain et al. 2018a).

Our results revealed the high prevalence of Aeromonas spp. among the ornamental guppy. All isolates displayed a significant proportion of phenotypic virulence factors and harbored several virulence genes which reveal the virulence potential of guppy-borne Aeromonas spp. Majority of the isolates showed resistance against different structural classes of antibiotics which revealed the indiscriminate use of those antibiotics in the ornamental fish industry. The occurrence of virulent and drug-resistant guppy-borne Aeromonas spp. can cause a serious threat to the ornamental fish industry. Therefore, further studies should be focused on the characterization of virulence and antimicrobial resistance profiles of Aeromonas spp. isolated from other ornamental fish. Considering the rapid growth of the ornamental fish industry, the antibiotic usage for prophylactic purposes must be substituted by better husbandry especially well-maintenance of water quality and good transportation practice.