Prevalence, Antimicrobial Resistance and Resistance Gene Cassettes Detection in Bacterial Pathogens Isolated from Freshwater Ornamental Fishes

Ornamental fish production is a major source of income and employment globally. More than one billion ornamental fish are traded internationally and have become the most important global trade sector (Hatha and Nifty, 2012). Guppy and Molly are the most popular and commonly available pet fishes in tropical countries because of their beautiful coloration and easy maintenance. Like other ornamental fishes, these fishes are also prone to bacterial, fungal, viral and parasitic infections. Most bacterial infections in aquaculture are especially caused by Gram-negative species (Lewbart, 2001). Poor water quality, improper handling and stress are the common factors that cause diseases in fishes cultured in captive conditions. This enforces the farmers to use antibiotics in the aquaculture system therapeutically and prophylactically to control bacterial infections. Erythromycin, penicillin, ampicillin, tetracycline, oxytetracycline, gentamicin, neomycin, kanamycin, amikacin, nalidixic acid, oxolinic acid, nitrofurantoin, nitrofurazone, furanace, furazolidone, ormetoprim and sulfadimethoxine are commonly used in the ornamental sector for treating bacterial diseases (Yanong, 2011). This unregulated use of antimicrobials in the system leads to the emergence of antimicrobial resistance (AMR) (Hemamalini et al., 2021). The emergence of AMR pathogens has been increasingly reported, which decreases the ability to treat the infections. The resistant determinants from fish pathogens can be transmitted to clinically important pathogens via horizontal gene transfer (HGT). AMR gene dissemination is mainly mediated by integron gene cassettes such as class 1, 2 and 3 (Soufi et al., 2011). Of the 2152 studies published by Indian institutions on AMR, only 11 were on fishes (Taneja and Sharma, 2019). To control the AMR, considerable baseline data are essential. In Tamil Nadu, not many studies were carried out for detecting AMR in theaquaculture sector. Therefore, the present study was carried out to determine the bacterial diversity, antimicrobial susceptibility pattern, and AMR gene cassette detection from infected Guppy and Molly.

Bacterial isolation and dendrogram construction
 
Infected Guppy (Poecilia reticulata) and Molly (P. sphenops) were collected (n=30 each) from ornamental fish farms in Chennai, India. Clinical signs were observed and the infected tissue samples were pooled and inoculated into the nutrient broth for bacterial isolation. Bacterial isolation was carried out by the conventional spread plate method. Bacterial colonies were randomly selected based on colony morphology and sub-cultured. Biochemical tests such as gram staining, motility, catalase, oxidase, sugar fermentation, indole, Voges-Proskauer and citrate utilization were performed. The results of the biochemical tests were recorded as numerical values for dendrogram construction using NTedit, version 1.2 and NTSYSpc, version 2.10e (Exeter Software, Setauket, NY, USA). Shannon Weiner diversity index was calculated using Primer-E software (Clarke and Gorley, 2015).
 
Molecular characterization and phylogenetic tree construction
 
Genomic DNA from representative isolates of each cluster was extracted following the salting-out method (Miller et al., 1988). Polymerase chain reaction (PCR) was performed with 16SrRNA gene universal primers fD1 (5¢CCG AAT TCG TCG ACA ACA GAG TTT GAT CCT GGC TCA G 3') and rD1 (5' CCC GGG ATC CAA GCT TAA GGA GGT GAT CCA GCC 3') (Weisburg et al., 1991) for bacterial identification. The PCR products were sequenced by Sanger dideoxy chain termination method at Eurofins Genomics India Pvt. Ltd., Karnataka, India. The raw sequence data were trimmed and aligned using BioEdit Alignment Editor, version 7.2.5 (Hall, 1999). The sequences were identified by comparing the sequences in the NCBI database using BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) and the final sequences were deposited to GenBank. MEGA7.0 software was used for determining the distance matrices of 16SrRNA sequences and phylogenetic trees were constructed using UPGMA statistical method with Kimura 2- parameter substitution model (Kumar et al., 2016).
 
Antibiotic susceptibility test (ABST) and detection of AMR gene cassettes
 
Thirty-six antibiotics were used to detect the antimicrobial susceptibility pattern of resolved isolates. ABST was performed following the disc diffusion method on Muller Hinton agar (MHA) plates. 0.5 McFarland standards (106 CFU/ml) were prepared for all the isolates and the bacterial lawn was prepared using sterile cotton swabs (Himedia, India). Escherichia coli ATCC 25922, CLSI reference strain, was used as a positive control. The zone of inhibition (ZOI) was measured and interpreted by Clsi (2018) and manufacturer guidelines. The multiple antibiotic resistance (MAR) index was calculated as the ratio of the number of antibiotics to which the bacterium was resistant to the total number of antibiotics used in the study (Krumperman, 1983).

The PCR was performed for the detection of class 1 and class 2 integron gene cassettes from genomic and plasmid DNA using specific primers [class 1 integron primers 5'CSF (5'GGCATCCAAGCAGCAAG3') and 3'CSR (5'AAGCAGACTTGACCTGA3') and class 2 integron primers Hep74F (5’CGGGATCCCCGGCATGCACGATTTGTA3') and Hep51R (5'GATGCCATCGCAAGTACGAG3')]. Plasmid DNA from bacterial isolates was extracted using the HiPurA plasmid DNA miniprep purification kit (Himedia, India). Class 1 and class 2 integron gene cassettes were detected following the PCR conditions suggested by Levesque et al., (1995) and White et al., (2001). The PCR products class 1 and class 2 integron gene cassettes were detected in 1% agarose gel.

Bacterial diversity
 
Clinical signs from infected fishes were observed, including skin ulcers, hemorrhagic patches, fin rot, lesions on internal organs (liver, spleen and kidney) and ascetic fluid accumulation in the abdomen on infected Guppy and Molly. The conventional spread plate technique provided more than 100 colonies for all the fish samples at 10-6 dilution. Dendrogram construction using biochemical test results generated 4 and 9 clusters for Guppy and Molly, respectively. Shannon Weiner diversity index was calculated based on species richness, dominance and evenness (Kim et al., 2017). The Shannon Weiner diversity was calculated as 1.366 and 2.101 for Guppy and Molly, respectively. Knowing the bacterial diversity in infected fishes will help to improve disease treatment methods.
 
Molecular identification of bacterial isolates
 
16SrRNA sequences of recovered isolates were compared with NCBI databases using BLAST and the sequences were identified as Enterobacter cloacae, Aeromonas veronii, A. hydrophila and Enterococcus faecalis in Guppy; A. dhakensis, B. subtilis, E. faecalis, Kurthia gibsonii, Comamonas testosteroni, A. sobria, Edwardsiella tarda, E. cloacae and Klebsiella aerogenes in Molly. The diversity of bacterial isolates derived from infected fish samples are depicted in Fig 1. The 16SrRNA sequences were submitted to GenBank and the accession numbers were given in Table 1. Most of the isolates derived from Guppy and Molly are gram-negative. It was reported that most of the infections caused in aquaculture are mainly by gram-negative bacteria, while some diseases were caused by gram-positive microbes such as Streptococcus sp. and Staphylococcus sp. (Lewbart, 2001). In the present study, enteric bacteria such as  E. cloacae, K. aerogenes and E. tarda were reported in Guppy and Molly. Human enteric pathogens such as E. cloacae could lead to severe mortality in cultured fishes. The source of these pathogens may be fecal contamination, which poses a high health risk to humans and cultured animals (Gufe et al., 2019). The presence of enteric bacteria in the current study might be due to the fecal matter of birds and animals.



A wide range of diseases in aquaculture is caused by Aeromonas sp., including A. hydrophila, A. veronii and A. caviae (Lewbart, 2001). Aeromonas infection in ornamental and food fishes significantly threatens human health. The first report of A. dhakensis pathogenicity in Nile tilapia was reported by Soto-rodriguez et al., (2013). Gram-positive bacteria, B. subtilis was isolated in this study. K. gibsonii was isolated from infected Molly; although this bacterium is not the main pathogenic bacteria, it can enter fish via skin ulceration and may aggravate the host’s condition. No information was available to us about the pathogenicity of C. testosteroni in fish. However, it can cause human diseases (Tsui et al., 2011).

Phylogenetic trees of resolved isolates were constructed using MEGA 7.0 software. It was revealed that the cladogram consisted of 4 and 9 operational taxonomic units (OTUs) with the corresponding bacterial strains from GenBank for the bacterial isolates derived from Guppy and Molly, respectively. The distance coefficient of phylogenetic trees ranged from 0.10 to 0.0 for Guppy and 0.12 to 0.0 for Molly. The phylogenetic tree generated for all the bacterial isolates from Guppy and Molly mainly consists of two branches. In Guppy, two branches were separated at the distance coefficient of 0.12r (Fig 2). In Molly, branch 1, containing 6 clusters, was separated from branch 2, having 3 clusters at the distance coefficient of 0.140r (Fig 3). It was found that the mean distance between the 16SrRNA gene sequences of Guppy and Molly isolates was calculated as 0.02.



​
 
Antimicrobial susceptibility testing
 
Antibiogram profiling of bacterial isolates was performed using 36 antibiotics. Bacterial isolates from Guppy were resistant to minimum of 5 and maximum of 14 antibiotics, isolates from Molly were resistant to minimum of 3 and maximum of 16 antibiotics tested. A. veronii (0.39) isolated from infected Guppy and K. aerogenes (0.45) from Molly has the highest MAR index. K. aerogenes isolated from moribund goldfish collected from ornamental fish farms in Kerala and Tamil Nadu have a higher MAR index (0.67) (Preena et al., 2020). MAR index, resistant antibiotic and respective antibiotic classes are listed in Table 1. The MAR index of >0.2 indicates a higher risk of AMR and antibiotic contamination in aquaculture systems (Krumperman, 1983). In the present study, all the isolates possessed a MAR index >0.2 except A. hydrophila, A. dhakensis, K. gibsonii and E. tarda. This highlights the heavy usage of antibiotics and the occurrence of AMR pathogens in the ornamental fish culture system. Similar to the present study, the highest MAR profiles have been detected by Preena et al., (2020) in goldfish.



All the isolates from Guppy were resistant to at least one antibiotic from cephalosporin 1st generation and polypeptides class. More than 50% of isolates from Guppy were resistant to at least one antibiotic of penicillin, cephalosporin 3rd generation, monobactam, aminoglycoside, sulphonamides, quinolone, nitrofuran, glycopeptides and rifamycin class. In contrast, all the isolates were susceptible to phenicol, macrolides, and fluoroquinolones class antibiotics. In Molly, the highest resistance (78%) was recorded in the polypeptide class and more than 50% of isolates were resistant to minimum of one antibiotic from penicillin, cephalosporin 1^st and 3^rd generation, sulphonamide and rifamycin class. The percentage of antimicrobial resistance exhibited by bacterial isolates toward different antibiotic classes is depicted in Fig 4.



All the isolates were susceptible to sulphafurazole, enrofloxacin, norfloxacin and ciprofloxacin. Nalidixic acid from the quinolone class was ineffective against >20% of isolates from Guppy and Molly. Contrary results were obtained by Preena et al., (2020), where nalidixic acid was ineffective against almost 62% of the tested isolates. The percentage of antibiotic resistant isolates against 36 antibiotics tested are listed in Table 2. E. faecalis from Guppy, B. subtilis, E. faecalis and K. aerogenes from Molly have exhibited resistance against 4^th generation cephalosporin antibiotic cefepime. Similar to our results, K. aerogenes isolated from infected goldfish also exhibited resistance to fourth-generation cephalosporin, cefepime (Preena et al., 2020). The production of extended spectrum beta-lactamases in fish pathogens results in the resistance toward new generation cephalosporins, making it difficult to control the diseases (Verner-jeffreys et al., 2009). Thus, the ineffectiveness of new generation antibiotics towards the fish pathogens could raise major challenges in the aquaculture sector. The emergence of AMR towards new generation antibiotics such as cephalosporins increases the risk and forces the development of better alternatives to antibiotics.



In Guppy, E. faecalis was susceptible to all the penicillin antibiotics. A. hydrophila was susceptible to all the antibiotics tested in cephalosporin 1^st, 2^nd, 3^rd, 4th generation, monobactam, aminoglycoside, phenicol, sulphonamide, macrolide, quinolone, fluoroquinolone and rifamycin group. A. hydrophila is the most common opportunistic pathogen in aquaculture and is associated with water quality-relateddiseases in cultured fish. E. cloacae was sensitive against the antibiotics belonging to cephalosporin 2^nd, 3^rd and 4^th generation, phenicol, sulphonamide, quinolone, fluoroquinolone and nitrofuran class, whereas antibiotics tested from monobactam, macrolides, glycopeptides and rifamycin exhibited resistance. All the bacterial isolates derived from Guppy were sensitive to all the antibiotics of the fluoroquinolone group. In Molly, E. faecalis, K. gibsonii, E. tarda, E. cloacae exhibited susceptibility towards all the antibiotics in the penicillin group. K. aerogenes exhibited resistance and K. gibsonii, A. sobria and E. tarda exhibited susceptibility towards all the antibiotics belonging to the cephalosporin 1st generation class. The MAR index of E. tarda was minimal (Aoki and Kitao, 1981).

A. dhakensis, K. gibsonii and A. sobria were susceptible to all the antibiotics belonging to the cephalosporin 3rd and 4th generation class. A. dhakensis, B. subtilis, K. gibsonii, A. sobria, E. tarda and K. aerogenes were susceptible to all the antibiotics belonging to the aminoglycoside class. A. dhakensis, E. faecalis, E. tarda and E. cloacae were susceptible to all the antibiotics in the sulphonamide group. A. dhakensis, B. subtilis, E. faecalis, K. gibsonii, A. sobria and E. cloacae were susceptible and K. aerogenes exhibited resistance to antibiotics of macrolides. Except for E. cloacae, all the other isolates were sensitive to fluoroquinolone antibiotics. B. subtilis, E. faecalis, E. tarda, E. cloacae and K. aerogenes were susceptible and C. testosteroni were resistant to all nitrofuran antibiotics. K. gibsonii, E. cloacae and K. aerogenes were susceptible to all the antibiotics from the polypeptide group. Thus, this study provides information about AMR in the aquaculture system, which helps to formulate alternative measures to control the disease in the aquaculture system.
 
Screening of Integron gene cassettes
 
In this study, class 1 integron was detected from all the bacterial isolates derived from Guppy and Molly. Two different product sizes of class 1 integron (650 bp and 1700 bp) were detected (Fig 5). Similarly, class 1 integrons were detected in respective bacterial plasmid DNA. Meanwhile, class 2 integron was not detected in the genomic and plasmid DNA of all the isolates from Guppy and Molly. Similar results were reported in fish pathogens isolated from infected Guppy collected from an ornamental fish farm, Cochin (Preena et al., 2019a). It was reported that the prevalence of class 1 integron in aquaculture systems is higher than class 2 and class 3 integron (Stalder et al., 2012). Both plasmid and genomic DNA were found to have class 1 integron indicating the AMR genes are plasmid-borne. The presence of AMR genes in mobile genetic elements indicated the risk of gaining new resistant determinants from other species and enabled potential gene transfer to other clinically important pathogens via horizontal gene transfer (Jacobs and Chenia, 2007). The AMR genes integrated into gene cassettes may be detected by further molecular characterization. However, integrons with empty gene cassettes were also detected in Y. ruckeri (Balta et al., 2010) isolated from rainbow trout. Hence, further molecular characterization of integron is necessary for confirmation.

The present study was conducted to isolate AMR fish pathogens from infected Guppy and Molly. ABST of resolved isolates revealed that most fish pathogens exhibited a MAR index of >0.2, indicating the heavy usage of antibiotics in aquaculture systems of Tamil Nadu and increases the risk of AMR in fish and human health. Some of the isolates from all the three fishes exhibited resistance to new generation antibiotic cefepime, which reminds the need for surveillance and continuous monitoring programs. Meanwhile, all the recovered isolates were susceptible to sulphafurazole, enrofloxacin, norfloxacin and ciprofloxacin. Based on the results, these antibiotics can be used in the ornamental fish culture system. The presence of integron gene cassettes in both genomic and plasmid DNA increases the risk of AMR gene dissemination to clinically important bacteria via gene transfer. Our study provides baseline data on the AMR level in fish pathogens from ornamental fish, which will be essential to mitigate the potential risks to human and fish health.

The authors acknowledge the Institute of Fisheries Post Graduate Studies, TNJFU, for the support.