Prevalence, Antibiotics Resistance and Plasmid Profiling of Vibrio spp. Isolated from Cultured Shrimp in Peninsular Malaysia

Vibrio is the most common bacterium associated with diseases in crustaceans. Outbreaks of vibriosis pose a serious threat to shrimp production. Therefore, antibiotics are commonly used as preventative and therapeutic measures. Unfortunately, improper use of antibiotics leads to antibiotic resistance. Nevertheless, information on the occurrence of Vibrio spp. and antibiotic use in shrimp, particularly in Malaysia, is minimal. This study aimed to provide information on the occurrence of Vibrio spp., its status of antibiotic resistance and the plasmid profiles of Vibrio spp. isolated from cultured shrimp in Peninsular Malaysia. Shrimp were sampled from seven farms that were located in different geographical regions of Peninsular Malaysia. According to the observations, 85% of the shrimp were healthy, whereas 15% were unhealthy. Subsequently, 225 presumptive Vibrio isolates were subjected to biochemical tests and molecular detection using the pyrH gene. The isolates were also tested for antibiotic susceptibility against 16 antibiotics and were subjected to plasmid profiling. Eventually, 13 different Vibrio spp. were successfully isolated and characterized using the pyrH gene. They were the following: V. parahaemolyticus (55%), V. communis (9%), V. campbellii (8%), V. owensii (7%), V. rotiferianus (5%), Vibrio spp. (4%), V. alginolyticus (3%), V. brasiliensis (2%), V. natriegens (2%), V. xuii (1%), V. harveyi (1%), V. hepatarius (0.4%) and P. damselae (3%). Antibiotic susceptibility profiles revealed that all isolates were resistant to penicillin G (100%), but susceptible to norfloxacin (96%). Furthermore, 16% of the isolates revealed MAR of less than 0.2, while 84% were greater than 0.2. A total of 125 isolates harbored plasmids with molecular weights between 1.0 and above 10 kb, detected among the resistant isolates. The resistant isolates were mediated by both chromosomal and plasmid factors. These findings support the use of surveillance data on the emerging patterns of antimicrobial-resistance and plasmid profiles of Vibrio spp. in shrimp farms. The findings from this study can be used to develop a better disease management strategy for shrimp farming.


Introduction
Shrimp is a popular seafood with high demand, especially in southeast Asia, providing a substantial contribution of USD 250 billion to the local, national, and regional economies [1,2]. China, India, Thailand, Indonesia, and Vietnam are the main contributors

Bacterial Isolation and Identification
Samples of hepatopancreases were homogenized in 1.0 mL of 0.9% sterile saline solution, and then serially diluted with sterile saline. One hundred µL of the suspensions were inoculated onto thiosulfate citrate bile salts sucrose (TCBS) agar (Oxoid, Hampshire, UK) and incubated for 24 h at 30 • C. The yellow and green colonies were then recorded and further sub-cultured onto TCBS to obtain a pure culture [22].

Bacterial Identification Using pyrH Gene
Genomic DNA of the suspected Vibrio isolates was extracted using the DNA Purification Kit (Promega, Madison, WI, USA) and stored at −20 • C until use. Identification of the Vibrio isolates was verified based on the partial sequencing of pyrH that produced a PCR product size of around 500 bp. The pyrH forward primer of 5 -GAT CGT ATG GCT CAA GAA G-3 and pyrH reverse primer of 5 -TAG GCA TTT TGT GGT CAC G-3 were used [27]. The PCR cycling condition included an initial denaturation at 94 • C for 3 min, followed by 30 cycles of denaturation at 94 • C for 1 min, annealing at 55.3 • C for 2 min 15 s, extension at 72 • C for 1 min 15 s, and a final denaturation at 94 • C for 1 min 15 s. The PCR mixtures containing 2.5 µL of the DNA template (10-100 ng) were mixed with 12.5 µL of premixed Reddiant PCR MasterMix (First Base, Kuala Lumpur, Malaysia) comprised of 60 U/mL of Taq DNA polymerase, 3 mM MgCl 2 , 400 µM dNTP mix and 10 µM each of forward and reverse primers, in a total volume of 25 µL of PCR mixture. Then, the PCR amplifications were carried out using a PCR thermocycler (Bio-Rad, Hercules, CA, USA). The amplified products were examined using 1% agarose gel electrophoresis that had been pre-added with FloroSafe DNA stain (First Base Laboratories) dye and sequencing was performed (Apical Scientific Sdn. Bhd., Kuala Lumpur, Malaysia).

Sequence Editing and Analysis
The resulting DNA sequences were trimmed of low quality at the frond and end of the sequence by using ChromasPro 2.1.8 software (Technelysium Pty Ltd., South Brisbane, Queensland, Australia). The trimmed sequences were then blasted in the GenBank database using the nucleotide Basic Local Alignment Search Tool (BLASTn) (https://blast.ncbi.nlm. nih.gov/Blast.cgi (accessed on 1 April 2019)) for preliminary identification.

Phylogenetic Analysis
Phylogenetic analysis was conducted using MEGA version 7.0 [28]. After the addition of pyrH genes obtained from 36 Vibrio spp. reference genomes (Table A1), the database of pyrH sequences was aligned using ClustalW and a curated database of the sequences was used for subsequent analyses. A neighbor joining phylogenetic analysis was conducted with the Kimura-2-parameter model with 1000 bootstrap replicates [29]. The reported sequences were deposited in the GenBank nucleotide sequence databases (accession numbers OP198216-OP198433). (Supplementary Materials).

Antibiotic Susceptibility Test
Antibiotic susceptibility of the Vibrio isolates was determined using the disc diffusion method [30]. Sixteen antibiotics (Thermo Fisher Scientific, Waltham, MA, USA) were used, which included the following: tetracycline 30 µg (TET), ampicillin 10 µg (AMP), gentamicin 10 µg (CN), sulfomethiozole-trimethoprim 25 µg (SXT), erythromycin 15 µg (E), chloramphenicol 30 µg (C), norfloxacin 10 µg (NOR), penicillin G 10U (P), cefepime 30 µg (FEP), cefotaxime 30 µg (CTX), ceftazidime 30 µg (CAZ), cephalothin 30 µg (KF), kanamycin 30 µg (K), ciprofloxacin 5 µg (CIP), nitrofurantoin 300 µg (F) and vancomycin 30 µg (VA). Following incubation for 24 h at 30 • C, the isolates were then inoculated in sterile saline water to achieve a bacterial concentration of 1.5 × 10 8 CFU/mL, which was equivalent to 0.5 MacFarland standard. The broth was evenly swabbed onto Mueller Hinton (MH) agar (Oxoid, Hampshire, UK) supplemented with 1.5% of sodium chloride (NaCl) [31]. Antibiotic discs were aseptically placed on the agar and incubated at 35 • C for 18 h before the inhibition zones were measured. Antibiotic susceptibilities, categorized as resistant (R), intermediate (I) or susceptible (S), were assigned using criteria outlined in the Clinical Laboratory Standards Institute (version M45-A2) [32,33]. The multiple antibiotic resistance (MAR) index was calculated using the MAR index value: A/B = MAR index, where A was the number of antibiotics to which the isolate was resistant, and B was the total number of antibiotics used in this study [34]. A MAR index value of less than 0.2 indicated that the isolates were from low-risk sources of contamination. In contrast, a MAR index of greater than 0.2 indicated that the isolates were from high-risk sources of contamination [35].

Plasmid Extraction
Approximately 2.5 mL of bacterial culture from Tryptic Soy Broth (TSB) (Oxoid, Hampshire, UK), supplemented with 1.5% NaCl, was centrifuged at 12,000× g for 3 min. The Vibrio isolate was purified using a GeneJet Plasmid Purification kit (Thermo Fisher Scientific, Waltham, MA, USA). The plasmid-containing supernatant was stored at −20 • C. The presence of plasmid was detected using agarose gel (1% w/v) electrophoresis (Bio-Rad).

Plasmid Curing
Vibrio isolates, that harbored plasmid, were treated with acridine orange (AO) (Thermo Fisher Scientific, Waltham, MA, USA) following modifications of the method by Letchumanan et al. [19]. The isolates were grown on TSB supplemented with 1.5% of NaCl, at 30 • C for 18 h under constant agitation. After incubation, 200 µL of the aliquots were added into the tubes containing Luria Bertani (LB) broth (Oxoid, Hampshire, UK), supplemented with 1.5% of NaCl (control), and into the tubes containing LB broth, supplemented with 1.5% of NaCl and AO. The isolates were incubated at 30 • C for 18 h [36]. The presence of plasmid was detected using agarose gel electrophoresis (1 % w/v) after treatment with AO. In addition, the antibiotic susceptibility test was repeated, as stated earlier, to confirm changes in resistance profiles.

Clinical Examination
Most of the sampled shrimp were healthy (85%), while 15% were unhealthy shrimp. Based on the observations, the hepatopancreas of healthy shrimp was large and black, while the midgut was full (Figure 1a,c). On the other hand, the hepatopancreas of unhealthy shrimp appeared small and pale, while the midgut was empty (Figure 1b,d). The body coloration between healthy and unhealthy seemed to be different, in that unhealthy shrimp had a pale body color (Figure 1b) compared to healthy shrimp (Figure 1a). In fact, the sizes of the unhealthy shrimp were smaller compared to healthy shrimp. The in-situ examination showed that the hepatopancreas and midgut were associated with signs of vibriosis. However, additional tests, in terms of molecular identification and histological examination, are required for verification. The presence of plasmid was detected using agarose gel electrophoresis (1 % w/v) after treatment with AO. In addition, the antibiotic susceptibility test was repeated, as stated earlier, to confirm changes in resistance profiles.

Clinical Examination
Most of the sampled shrimp were healthy (85%), while 15% were unhealthy shrimp. Based on the observations, the hepatopancreas of healthy shrimp was large and black, while the midgut was full (Figure 1a, Figure 1c). On the other hand, the hepatopancreas of unhealthy shrimp appeared small and pale, while the midgut was empty (Figure 1b,  Figure 1d). The body coloration between healthy and unhealthy seemed to be different, in that unhealthy shrimp had a pale body color (Figure 1b) compared to healthy shrimp ( Figure 1a). In fact, the sizes of the unhealthy shrimp were smaller compared to healthy shrimp. The in-situ examination showed that the hepatopancreas and midgut were associated with signs of vibriosis. However, additional tests, in terms of molecular identification and histological examination, are required for verification.

Isolation and Identification of Vibrio spp.
A total of 225 suspected Vibrio isolates were obtained from the 210 sampled shrimp. They were identified by color, shape, and size of the culture on TCBS agar. Among the 225 isolates, 149 (66%) isolates produced green, round, and medium-sized (2-5 mm) colonies, while 76 (34%) isolates produced yellow, round, and medium-sized colonies ( Table  2).

Isolation and Identification of Vibrio spp.
A total of 225 suspected Vibrio isolates were obtained from the 210 sampled shrimp. They were identified by color, shape, and size of the culture on TCBS agar. Among the 225 isolates, 149 (66%) isolates produced green, round, and medium-sized (2-5 mm) colonies, while 76 (34%) isolates produced yellow, round, and medium-sized colonies ( Table 2).

Biochemical Characterization of Vibrio spp.
All 225 isolates (100%) were Gram negative, positive to catalase and oxidase tests, motile but not able to produce sulfide gas. An indole test showed that 95% of the isolates were positive, while 80% of the isolates tested positive for lysine dehydrogenase (LDC). None of the isolates produced gas and hydrogen sulfide. However, 61% of the isolates produced red (alkaline) coloration in the slants and yellow (acid) in the butts. They fermented the glucose. A total of 39% appeared yellow in both the slants and butts (Table 3).

Prevalence of Vibrio spp.
A total of 225 suspected Vibrio spp. were isolated based on the green and yellow colonies appearing on thiosulfate citrate bile salt sucrose (TCBS) agar. They were also pyrH positive producing a single band of 500 bp. Among them, 22% of the isolates were from Banting, Selangor, 16% were from Ayer Hitam, Kedah, 16% were from Marang, Terengganu, 13% were from Manjung, Perak, 12% were from Merlimau, Melaka, 11% were from Sg. Besar, Selangor and 11% were from Mersing, Johor ( Figure 2). As shown in Figure 2 nius was found predominantly in Ayer Hitam, Kedah. It was also found in Sg. Besar in Selangor, Merlimau in Melaka and Manjung in Perak. V. alginolyticus were found in Mersing, Johor and Ayer Hitam, Kedah. V. communis was mostly found in Ayer Hitam Kedah and was also detected in Sg. Besar, Selangor. P.damselae was found in Sg Besar and Banting, Selangor. It was predominantly found in Merlimau, Melaka. Other species, including V. xuii and V. harveyi, were only found in Manjung, Perak. V. brasiliensis was found in Ayer Hitam, Kedah, while V. hepatarius and V. natriegens were only found in Banting, Selangor.

Antibiotics Susceptibility Test
The findings revealed that all the Vibrio isolates (13/13) were found to be resistant to vancomycin (VA), including V. parahaemolyticus, V. campbellii, V.rotiferianus, V. owensii, V. alginolyticus, V. xuii, V. harveyi, V. natriegens, V. hepatarius, V. communis, V. brasiliensis, Vibrio spp. and P. damselae. The findings showed that eight out of thirteen species were highly resistant (100%) to vancomycin (VA). The same applied to penicillin G (P), which all of the Vibrio isolates (13/13) were found to be resistant to. However, seven out of thirteen (7/13) species were highly resistant (100%) to penicillin G. The results also indicated that all Vibrio spp. were resistant to at least one of the 16 antibiotics tested. When compared to other Vibrio spp. isolates, V. parahaemolyticus was found to be resistant to all the tested antibiotics (16/16). Indeed, V. parahaemolyticus was discovered to be the only species that was resistant to norfloxacin (NOR) (1%). The findings also indicated that seven out of thirteen (7/13) of the Vibrio isolates, over 50%, were resistant toward the tested antibiotics, including V. parahaemolyticus, V. campbellii, V. rotiferianus, V. owensii, V. alginolyticus, V. natriegens and V. communis. (Table 4).

Antibiotics Susceptibility Test
The findings revealed that all the Vibrio isolates (13/13) were found to be resistant to vancomycin (VA), including V. parahaemolyticus, V. campbellii, V.rotiferianus, V. owensii, V. alginolyticus, V. xuii, V. harveyi, V. natriegens, V. hepatarius, V. communis, V. brasiliensis, Vibrio spp. and P. damselae. The findings showed that eight out of thirteen species were highly resistant (100%) to vancomycin (VA). The same applied to penicillin G (P), which all of the Vibrio isolates (13/13) were found to be resistant to. However, seven out of thirteen (7/13) species were highly resistant (100%) to penicillin G. The results also indicated that all Vibrio spp. were resistant to at least one of the 16 antibiotics tested. When compared to other Vibrio spp. isolates, V. parahaemolyticus was found to be resistant to all the tested antibiotics (16/16). Indeed, V. parahaemolyticus was discovered to be the only species that was resistant to norfloxacin (NOR) (1%). The findings also indicated that seven out of thirteen (7/13) of the Vibrio isolates, over 50%, were resistant toward the tested antibiotics,

Multiple Antibiotic Resistance (MAR) Index
The MAR index ranged between 0.06 and 0.75 ( Table 5). The average of MAR index was 0.41, and 16% (n = 36) had a MAR index of <0.2, whereas the remaining 84% (n = 189) showed a MAR index of >0.2. In fact, 0.4% had a MAR index of 0.06, indicating that they were antibiotic resistant to at least one type of antibiotic. On the other hand, 0.4% of the isolates showed a MAR index of 0.75, indicating a resistance to 12 antibiotics. Approximately 95% of the isolates were multidrug resistant (MDR), meaning that they were resistant to three or more antibiotics, with a MAR index between 0.19 and 0.75 ( Table 5). The most common MAR index was 0.38, which was found in 22% of the Vibrio isolates, indicating that they were resistant to six different antibiotics. Other MAR indices were 0.13 (4%), 0.19 (12%), 0.25 (15%), 0.31 (16%), 0.44 (13%), 0.50 (6%), 0.56 (6%), 0.63 (2%) and 0.69 (3%).

Plasmid Profiling
Among the 225 Vibrio isolates tested, 125 (55.6%) isolates harbored plasmid with molecular weight from 1.0 to above 10.0 kb. While 100 (44.4%) of the isolates did not harbored any of the plasmid. The Vibrio spp. that were found to contain plasmid including V. parahaemolyticus (80%), V. communis (40%), V. campbellii (44%), V. owensii (60%), V. rotiferanius (42%), V. natriegens (40%), V. brasiliensis (50%), V. alginolyticus (71%), Vibrio spp. (38%) and P. damselae (43%). However, none of the plasmid were found in V. harveyi, V. xuii and V. hepatarius ( Table 6). Most of the isolates, that harbored plasmid, were resistant to ampicillin (AMP) (84%) ( Table 6). The percentages of plasmids discovered in the resistance isolates are shown in Table 7. There were 57.6% that had one plasmid, 15.2% had two plasmids, 18.4% had three plasmids, 3.2% had four and six plasmids and 2.4% had five plasmids. The plasmids were categorized into twenty-three (23) different profiles as follows: 4 (3.2%) of the isolates presented profiles 1 and 23; 1 (0.8%) of the isolates presented profiles 2, 5, 7, 9, 10, 11,13, 15 and 21; 2 (1.6%) of the isolates presented profiles 3, 6, 8, 14, 16, 19, 20 and 22; 65 (52.0%) of the isolates presented profile 4; 9 (7.2%) of the isolates presented profile 12; and 5 (4.0%) of the isolates presented profile 17 (Table 7). From 23 of the plasmid profiles, the profile that formed the largest group was plasmid profile 4 that consisted of one plasmid of above 10.0 kb size (Table 7). V. parahaemolyticus AMP, K, CTX, F, VA, P + 6       Figure 4 shows the number of resistant isolates before and after plasmid curing. The number of resistant isolates were reduced after the curing. Generally, the number of resistant isolates changed to intermediate and susceptible after the curing process with acridine orange. According to Figure 4a, it was shown that before the curing process, 125 of the isolates were resistant to ampicillin (AMP). However, after curing, 110 of the isolates remained resistant to ampicillin (AMP), while 15 of the isolates changed to either intermediate or susceptible, 4 to intermediate and 11 to susceptible (Figure 4b). Similarly, from 125 of the isolates, 123 and 124 of the isolates remained resistant to vancomycin (VA) and penicillin G (P), respectively, after the curing process. On the other hand, the number of isolates that were resistant to cephalothin (KF) drastically dropped from 102 to 57 isolates (Figure 4a (Figure 4b). In addition, a similar pattern of reduction was observed for isolates' resistance against nitrofurantoin (F), from 58 before curing to 31 after curing (Figure 4a), 21 changing to intermediate and 6 to susceptible (Figure 4b). The findings also revealed a dramatic decrease in isolates resistant to ceftazidime (CAZ) and erythromycin (E), from 37 to 7 and 26 to 6, respectively (Figure 4a), with 11 and 19 changing to intermediate and susceptible regarding ceftazidime (CAZ) and 12 and 8 changing to intermediate and susceptible against erythromycin (E) (Figure 4b). All of the isolates resistant to cipfloxacin (CIP) changed to intermediate (7) and susceptible (6) (Figure 4a,b). Based on the gentamicin (CN) results, the number of resistant isolates dropped from 9 to 3 isolates (Figure 4a), 3 changing to intermediate and 3 to susceptible (Figure 4b). The number of the isolates resistant to cefepime (FEP) dropped from 8 to 6 isolates (Figure 4a), 2 changing to susceptible after the curing process. The number of isolates resistant to chloramphenicol (C) dropped from 12 to 11 isolates (Figure 4a), one of the isolates changing to intermediate (Figure 4b). For norfloxacin (NOR), the resistance of the isolates changed to susceptible after the curing process (Figure 4a,b). There was no change in resistance to sulfomethiozole-trimethoprim (SXT) and tetracycline (TET) after plasmid curing, indicating that the resistance isolates were chromosomally mediated (Figure 4a). The result after plasmid curing revealed that when the antibiotic resistance profile was affected, it indicated that the resistance isolate was plasmid mediated, whereas when the antibiotic resistance profile was unaffected, it indicated that the resistance isolate was chromosomal mediated [19].

Discussion
Vibrio spp. are a group of bacteria naturally found in freshwater, estuaries and marine environments [38]. It is commonly known that Vibrio spp. are responsible for numerous human diseases attributed to the natural microbiota of aquatic environments and seafood [39]. In humans, Vibrio spp. are known to cause gastroenteritis, cholera, and septicemia [40]. The human pathogenic Vibrio spp. of clinical relevance include the following: V. parahaemolyticus, V. alginolyticus, V. cholerae, V. vulnificus, V. tubiashi and V. fluvialis [41]. In addition, V. parahaemolyticus, V. vulnificus, and V. mimicus are foodborne pathogens [42]. Meanwhile, the common Vibrio spp. associated with the shrimp diseases are V. harveyi, V. parahaemolyticus, V. alginolyticus, V. anguillarum, V. vulnificus and V. splendidus. V. harveyi is associated with luminescent vibriosis in shrimps, particularly in P. vannamei and P. monodon. V. parahaemolyticus can cause human illness and is frequently associated with food-borne gastroenteritis or diarrhoea [43].
Identification based on biochemical keys has been proposed for Vibrio spp. and is an excellent means to obtain a large number of reference and environmental Vibrio strains [34]. However, conventional phenotyping and biochemical identification techniques are poorly adapted to Vibrio strains isolated from seafood and aquatic environments [44]. The presence of both false positive and false negative results in all the biochemical identification methods leads to difficulties identifying the Vibrio spp. [45]. Therefore, PCR assays are required for their identification and detection [46].
Phylogenetic analyses were carried out using the neighbor-joining method to confirm taxonomic position in the genus Vibrio [47]. Generally, the 16S rRNA gene is the most popular molecular marker for identifying and classifying isolated pure cultures and estimating bacterial density in environmental samples through metagenomic assay [48]. However, the 16S rRNA gene has low discriminatory power, leading to the misidentification of Vibrio spp. compared to other tested genes [37]. In addition, Vibrio spp. share more than 97% similarity with the 16S rRNA resulting in difficulties in differentiating closely related species [49]. Due to this limitation, other housekeeping genes are used as phylogenetic markers to determine the diversity of bacterial species. In this study, phylogenetic analysis using the pyrH gene showed that Vibrio spp. isolated from shrimp belonged to three clades: Harveyi, Nereis and Orientalis. The Harveyi clade consists of eight species, which are V. parahaemolyticus, V. owensii, V. communis, V. rotiferianus, V. campbellii, V. alginolyticus, V. natriegens, and V. harveyi. V. xuii belongs to the Nereis clade. Finally, the Orientalis clade consists of V. brasiliensis and V. hepatarius [37]. The pyrH gene is a housekeeping gene that encodes uridine monophosphate kinase (UMP kinase), which participates in the pyrimidine biosynthesis catalyzing the conversion of UMP into UDP [50]. According to Chimetto et al. [51], the pyrH gene could effectively distinguish the species level of Vibrio, including V. communis, which currently is categorized as Vibrio spp. In fact, pyrH were also among the genes that gave the highest resolution in distinctively differentiating V. harveyi and V. campbellii [52]. In addition, it was also among the genes that gave the highest resolution in distinctively differentiating V. harveyi and V. campbellii [52]. Thus, pyrH gene was chosen, since it was one of the molecular markers suitable for identifying the closely related Vibrio spp. in this investigation [53].
A study performed by Thompson et al. [54] compared the species resolution level with three different housekeeping genes, rpoA, recA and pyrH. The findings revealed that the pyrH gene was a good predictor of Vibrio and a good discriminatory target at the species level. Moreover, they showed stability of this locus, due to high proportions of synonymous mutations leading to the conservation of the amino acid sequence. Furthermore, until other genome-based techniques are proposed, pyrH is the most powerful method for distinguishing Vibrio spp. in biodiversity, population genetics, and evolution studies. It also helps to eliminate the misidentification of Vibrio ancestry clades [37]. In molecular phylogenetics, the use of a minimum gene set in molecular phylogenetics is critical for reducing time and expense, while also improving the accuracy of results. This is especially important when identifying species and understanding population structure and evolution in a large bacterial taxon like the Vibrionaceae family, with over 140 species [37].
In this study, we successfully isolated 225 Vibrio spp. from the hepatopancreases of cultured shrimp. They were V. parahaemolyticus, V. owensii, Vibrio spp., V. communis, V. rotiferianus, V. campbellii, V. alginolyticus, V. brasiliensis, V. xuii, V. harveyi, V. natriegens, V. hepatarius and P. damselae. Vibrio parahaemolyticus was the most prevalent species, isolated from six sampling locations. In fact, V. parahaemolyticus is often isolated from seafood, particularly shellfish or bivalve mollusks, worldwide [55]. In Croatia, 9.4% of sea fish, shrimp, and bivalve mollusks carry V. parahaemolyticus [56]. Similarly, Letchumanan et al. [57] revealed that V. parahaemolyticus could easily be isolated from retail shrimp in Malaysia. Vibrio parahaemolyticus was found to be dominant in shrimp in Ecuador (81%), Sri Lanka (98%), and Egypt (18%) [58][59][60]. Other species, such as V. campbellii and V. harveyi, were common species detectable in tropical marine regions and are among the most important bacterial pathogens of many commercially farmed marine invertebrate and vertebrate species in many Asian countries [61,62]. However, many previous studies have demonstrated the predominance of V. alginolyticus in shrimp or seafood samples [63]. According to Kriem et al. [44], the most common Vibrio spp. in shrimp in Morocco was V. alginolyticus. Similarly, Baffone et al. [64] reported the predominance of V. alginolyticus among fresh seafood products, followed by V. parahaemolyticus and V. cholerae. Unfortunately, contrary to our findings, V. alginolyticus was scarce and could only be found in two sampling locations in Mersing, Johor and Ayer Hitam, Kedah. The source of samples, methods of identification, study area, season, salinity, and temperature during storage or even transportation may influence the variation of Vibrio spp. prevalence in seafood [65,66]. In addition, temperature, pH, salinity, and nutrient levels present in the water column can ultimately affect the abundance of Vibrio [67].
Antibiotic resistance can be transmitted through sequential mutations in chromosomal genes or by acquiring genetic elements, such as plasmids, bacteriophages, or transposons [68]. This study demonstrated that 100% of the isolates were resistant to penicillin G, followed by vancomycin (98%) and ampicillin (84%). Vibrio resistance to ampicillin has been reported since 1978 with ranges between 40% and 90%. The ampicillin-resistant pattern could be related to the abuse of first-generation antibiotics in the environment, which reduced ampicillin susceptibility and efficiency in treating Vibrio infection [69]. Similarly, penicillin-resistant Vibrio has already been reported in different penaeid culture regions [70,71]. Srinivasan and Ramasamy [72] reported 100% penicillin G resistance in India, while Albuquerque et al. [73] also revealed high resistance to penicillin G. This study also reported that the isolates were highly resistant to vancomycin, which was consistent with Noorlis et al. [28]. Chloramphenicol (C) and norfloxacin (NOR) were the antibiotics to which Vibrio spp. were least resistant. The study showed that resistance to chloramphenicol (C) was only found in V. parahaemolyticus (9%) and V. communis (5%), while resistance to norfloxacin was only found in V. parahaemolyticus (1%). The result regarding chloramphenicol was expected, since it was banned due to the risks of human exposure to its residues in food products, which prevents fish farmers from using it [69]. Previous research by Ottaviani et al. [74], Sahilah et al. [75], and Sudha et al. [69] also supported our findings. Hence, regular monitoring on the usage of antibiotics is required to avoid the emergence of new resistant strains.
The Multiple Antibiotic Resistance (MAR) index, which ranges from 0 to 1.0, is a valuable measure for determining health risks. The MAR index value (0.20) is differentiated between low and high risks. A MAR value greater than 0.20 indicates that the samples have an increased risk of source contamination, while a MAR less than 0.2 indicates that the samples have a low risk of source contamination [35]. Our findings revealed a significant frequency of MAR index >0.2 (84%) in the sampling area compared to MAR index <0.2 (16%), thereby indicating that the aquatic environment in the sampling area may be affected and contaminated with antibiotics from human, animal, and fish consumer sources. In this study, although the farmers claimed that they did not use any antibiotics on their farms, the findings revealed that the occurrence of resistance to multiple antibiotics was very high. This might have been due to the sampling site being located in an area near the city, agricultural or industrial sectors. Hence, the resistance of bacteria to antibiotics might have travelled through water from nearby farms to the shrimp ponds. In fact, antibiotics from animal feeds or medications are absorbed into the sediment causing bacterial selection in the nearby environment [76]. Another possibility was that the presence of antibiotic-resistant bacteria in shrimp could be from post larvae, and the associated variety of drugs used in shrimp hatcheries [40]. The use of these drugs causes resistance to certain antimicrobials during post larvae rearing in the hatchery, which persists in the shrimp gut after transfer to the grow out ponds [77]. Furthermore, geographical location differences may influence the resistance level variance depending on sample collection [18].
The findings revealed that more than 55% of the isolates harbored between one to six DNA bands of plasmid, similar to those reported by Zanetti et al. [78] and You et al. [79]. Plasmid profile determination has been found to be very useful in epidemiological studies, and diagnosis and elucidation of mechanisms of drug resistance [80]. Plasmids are extra-chromosomal materials that allow the movement of genetic materials, including antimicrobial-resistant genes, between bacterial species and genera, through gene exchange processes; hence, increasing antibiotic resistance [81]. A study by Devi et al. [30] revealed that there was no correlation between the resistance of Vibrio to antibiotics and the presence of plasmid. Even though an isolate does not exhibit plasmids, the isolate can still show antibiotic resistance [82]. This statement was also supported by Zulkifli et al. [83], whose findings revealed that 53% of the isolates did not harbor any plasmid. However, they also showed multiple antibiotic resistance patterns to high number of antibiotics, which indicated that resistance to most of these antibiotics was of chromosomal origin, or on mobile genetic elements, that might help in the dissemination of resistant genes to other bacteria of human clinical significance. Nevertheless, most Vibrio isolates in this study lost their resistance to antibiotics following plasmid curing, even though some remained resistant. This indicates that the antibiotic resistance genes in Vibrio spp. isolated from cultured shrimp in this study were both chromosomal and plasmid mediated [36,73]. Although antimicrobial resistance in Vibrio spp. is typically acquired through plasmid transfer or antibiotic exposure, the role of plasmids in multiple antimicrobial resistances should be investigated further [74].

Conclusions
The findings thoroughly analyzed the occurrence, antibiotic resistance profiles, and plasmid profiling of Vibrio spp. isolated from cultured shrimp in Malaysia. Hence, antimicrobial resistance surveillance and drug usage monitoring in aquaculture should be encouraged to improve antibiotic management for public health and food safety in the industry. Furthermore, better knowledge of the molecular basis of resistance acquisition and transmission can aid in the development of new strategies to combat vibriosis for sustainable shrimp production.