Incidence, genetic diversity, and antimicrobial resistance profiles of Vibrio parahaemolyticus in seafood in Bangkok and eastern Thailand

Abstract

Background Emergence of Vibrio parahaemolyticus pandemic strain O3:K6 was first documented in 1996. Since then it has been accounted for large outbreaks of diarrhea globally. In Thailand, prior studies on pandemic and non-pandemic V. parahaemolyticus had mostly been done in the south. The incidence and molecular characterization of pandemic and non-pandemic strains in other parts of Thailand have not been fully characterized. This study examined the incidence of V. parahaemolyticus in seafood samples purchased in Bangkok and collected in eastern Thailand and characterized V. parahaemolyticus isolates. Potential virulence genes, VPaI-7, T3SS2, and biofilm were examined. Antimicrobial resistance (AMR) profiles and AMR genes (ARGs) were determined. Methods V. parahaemolyticus was isolated from 190 marketed and farmed seafood samples by a culture method and confirmed by polymerase chain reaction (PCR). The incidence of pandemic and non-pandemic V. parahaemolyticus and VPaI-7, T3SS2, and biofilm genes was examined by PCR. AMR profiles were verified by a broth microdilution technique. The presence of ARGs was verified by genome analysis. V. parahaemolyticus characterization was done by multilocus sequence typing (MLST). A phylogenomic tree was built from nucleotide sequences by UBCG2.0 and RAxML softwares. Results All 50 V. parahaemolyticus isolates including 21 pathogenic and 29 non-pathogenic strains from 190 samples had the toxRS/old sequence, indicating non-pandemic strains. All isolates had biofilm genes (VP0950, VP0952, and VP0962). None carried T3SS2 genes (VP1346 and VP1367), while VPaI-7 gene (VP1321) was seen in two isolates. Antimicrobial susceptibility profiles obtained from 36 V. parahaemolyticus isolates revealed high frequency of resistance to colistin (100%, 36/36) and ampicillin (83%, 30/36), but susceptibility to amoxicillin/clavulanic acid and piperacillin/tazobactam (100%, 36/36). Multidrug resistance (MDR) was seen in 11 isolates (31%, 11/36). Genome analysis revealed ARGs including blaCARB (100%, 36/36), tet(34) (83%, 30/36), tet(35) (42%, 15/36), qnrC (6%, 2/36), dfrA6 (3%, 1/36), and blaCTX-M-55 (3%, 1/36). Phylogenomic and MLST analyses classified 36 V. parahaemolyticus isolates into 5 clades, with 12 known and 13 novel sequence types (STs), suggesting high genetic variation among the isolates. Conclusions Although none V. parahaemolyticus strains isolated from seafood samples purchased in Bangkok and collected in eastern Thailand were pandemic strains, around one third of isolates were MDR V. parahaemolyticus strains. The presence of resistance genes of the first-line antibiotics for V. parahaemolyticus infection raises a major concern for clinical treatment outcome since these resistance genes could be highly expressed under suitable circumstances.

Genomic mining of Vibrio parahaemolyticus highlights prevalence of antimicrobial resistance genes and new genetic markers associated with AHPND and tdh + /trh + genotypes

“…Available literature also suggests a high resistance prevalence towards β-lactams, which we phenotypically confirmed in this study by the resistance of all three strains towards ampicillin and piperacillin. Larger-scale ( n > 36) phenotypical studies are confirming this too, with high reported resistance prevalence towards both penams [ e.g., 45–100% to ampicillin [ 6 , 28 , 30 – 33 ] and 68–100% to penicillin [ 30 – 32 ]] and cephalosporins [ e.g., 49.6–84% to cefazolin [ 31 , 34 ], 52–73% to cefotaxime [ 33 , 35 ] and 24–82% to cephalothin [ 30 , 34 ]].…”

“…This is again reflected in phenotypical studies on V. parahaemolyticus clinical and environmental strains, where low prevalence of resistance towards tetracyclines [ e.g,. 0.3% to doxycycline [ 30 , 31 , 36 ], 0–15% to tetracycline [ 27 , 28 , 30 – 33 , 35 , 36 ]], trimethoprim-sulfamethoxazole [0–7% [ 27 , 30 , 31 , 33 ]], and fluoroquinolones ( e.g,. 0% to ciprofloxacin [ 27 , 30 ] and 0–1% to levofloxacin [ 27 , 33 ]] were reported.…”

“…However, reported resistance prevalence against aminoglycosides is very variable [ e.g,. 0–64% to amikacin [ 27 , 28 , 30 , 33 ] and 15–50% to kanamycin [ 30 , 33 , 34 ]]. It should be noted that resistance to aminoglycosides is often the result of mutations of the ribosome [ 37 ], a factor that was not included in our analysis.…”

“…Hence, using WGS-based ARG genotyping in V. parahaemolyticus seems to be a good predictive tool to identify, study, understand, and follow up on the dissemination of drug-resistance genotypes in the population. Many studies reported the presence of at least one multidrug-resistant isolate like Vp M0904 [ 27 , 28 , 38 ]. These isolates represent an even bigger threat to aquaculture and human health, as they are more difficult to control with antibiotics, and are an important reservoir of ARGs.…”

“…CRP is suggested to be involved in macrolide, penam, and fluoroquinolone resistance [ 41 ], msbA in nitroimidazole resistance [ 42 ], tet(35) in tetracycline resistance and ugd in resistance against peptide antibiotics [ 43 ], but do not seem to induce resistance in V. parahaemolyticus . It is plausible that these are so-called ‘silent’ genes, that might only be activated in specific conditions or expressed at low levels [ 28 , 44 ]. The high prevalences of especially CRP , msbA , and tet(35) (≥ 99%) suggests that these might be intrinsic resistance genes [ 45 ]; however, their relatively low nucleotide identity ( CRP 80.0% ± 0.1%, msbA 66.5% ± 0.2%, tet(35) 82.6% ± 0.2%, and ugd 69.7% ± 0.9%) compared to other ARGs suggests that these genes might be degenerated and have lost functionality over time.…”