The Detection And Characterization Of Laribacter Hongkongensis Among Fishery Products In Guangzhou

Objectives:In order to acquire detailed information of Laribacter hongkongensis in fishery products in Guangzhou, a couple of freshwater fish and edible frogs, popularly consumed in southeast, were sampled to detect for L.hongkongensis. The antibotypes, pulsed field gel electrophoresis (PFGE) and enterobacterial repetitive intergenic consensus sequence (ERIC) were performed to describe the bacteriological characterization of L.hongkongensis isolates, to assess the genetic relationship of isolates and provide evidences for molecular epidemiology of L.hongkongensis. Besides, the role of integron harboring gene cassettes was described in L.hongkongensis and the diversity of resistance patterns were investigated by PFGE.Methods:Biochemical tests and 16S rRNA based PCR were used to detect L.hongkongensis in fishery products. Susceptibility testing to 19 antimicrobial agents was carried out in L. hongkongensis isolates by the disk diffusion method. The genomes of isolates were grouped into different clusters by PFGE and ERIC-PCR. The un-weighted pair-group mean arithmetic (UPGNA) of BioNumerics v4.0 software was used to cluster analysis. Specific primers were used to investigate the intergrase gene and gene cassette in L.hongkongensis isolates. The amplicons amplified were sequenced and analyzed by BLABT programme. Results:1. Grading-up for identification of L. hongkongensisThe 100% sensitivity were found in urease test, sugar oxidation/fermentation test, catalase test, cytochrome oxidase test, arginine dihydrolase test, while urease test was examed to be the highest specificity and positive predictive value,91.4% and 72.2%, respectively. The primer aiming 16S rRNA gene owned the 100% specificity. At the same time, there is no amplicon for non-target bacteria. The highest Kappa value 0.795 appeared in urease test associating with 16S rDNA based PCR.2. The detection of L. hongkongensis in fishery products in Guangzhou2.1 A total of 590 full intestines samples of fishery products were purchased in five farm product markets in Guangzhou from Oct.2008 to Sep.2009. The bacterium was isolated from 69 (17.6%) of 390 freshwater fish, including 66 (28.1%) of the 235 grass carps,3 (11.1%) of the 27 bighead carps. As for frog specimens, the bacterium was isolated from 99 (49.5%) of the 200 edible frogs. L. hongkongensis was not isolated from samples of Tilapia, Johnny carp, Sun fish, Chinese perch, Mud eel and Pond loach. Higher isolation rate was observed in edible frog than freshwater fish (49.5% vs 17.7%,x2=765.673, P<0.001). There was significant difference in the isolation rates of L. hongkongensis among six breeds of edible frogs (x2=14.476, P=0.013). The recovery rate followed in turn:Guenther's frog (63.4%), Non-wide Chinese tiger frog (57.3%), Black-spectacled toad (46.7%), Bull frog (45.5%), Wide Chinese tiger frog (30.0%), and Giant spiny frog (20.0%).2.2 A significant difference in grass carps was obtained in comparison with the isolation rates of L. hongkongensis among different seasons (x2=31.381, P<0.001). The isolation rate in grass carp ranges from 6.3% in winter to 46.7%, 33.3%,37.1%, in spring, summer, autumn, respectively.3. Analysis of antibiotics resistance of L. hongkongensis3.1 As determined by disc-diffusion antibiotic susceptibility testing, the isolates exhibited resistant to 19 antibiotics tested in distinct extent. The antibiotic resistant rates for the 199 L. hongkongensis isolates were as follows:CEF 99.5%, CFZ 97.0%, RIF 74.9%, CFP 61.8%, AMP 59.8%, ERY 22.2%, TET 18.1%, CAZ 12.6%, SXT 12.1%, PIP 10.1%, CIP 6.5%, ATM 6.0%, STR 5.6%, CXM 4.5%, FEP 4.5%, CHL 3.5%, IPM 2.0%, AMK 2.0%, GEN 0.5%. All the L. hongkongensis isolates were found to be resistant to cephalosporin, majority contributed to the first and third generation cephalosporin except for CAZ, but interestingly low resistance rates were found in the second and fourth generation. All strains were resistant to one class antibiotic agent at least. Among them,129 isolates (64.8%) were multidrug-resistant (resistant to 3 or more classes of antibiotics).3.2. Higher resistant rates were acquired in edible frog for quinolones (12.1% vs 1.4%,χ2=6.486, P=0.011), tetracyclines (27.3% vs 10.1%,χ2=7.389, P=0.007), sulfonamides (21.2% vs 2.9%,χ2=11.541, P=0.001) and rifampicine (83.8% vs,69.6%,χ2=4.823, P=0.028). Frog isolates of L. hongkongensis were resistant to chloromycetin, while fish isolates were sensitive to it.3.3 Multi-drug resistances were statistically differences among the strain origins (χ2=156.471, P<0.001). The highest multi-drug resistant rate was found in bighead carp (100.0%), followed by bull frog (83.3%), grass carp (70.0%), non-wide Chinese tiger frog (71.2%), wide Chinese tiger frog (61.5%), black-spectacled toad (45.5%), guenther's frog (42.3%) and giant spiny frog (40.0%).4. Conditions of PFGE and ERIC-PCR for L. hongkongensis4.1 The PFGE conditions adapt to L. hongkongensis as follow:OD 1.8, cell disruption for 4h, SpeⅠ20U cut for 20h, switch time of 2.2 to 54.2 s,120°field angle,6V/cm of electric field intersity, electrophoresis 25h.4.2 DNA extracting kit was better than CTAB method and directly spilling method, ensuring a good ERIC profile. Both primers were superior to a single one. The 48℃, annealing temperature, was determined to get a better profile for analysis.5. PFGE profilesThe patterns obtained from SpeⅠdigestion were stable and reproducible when repeated analyses in every 10th strains were performed under identical digestion and electrophoresis conditions in each run. This SpeI enzyme was discriminated into 8 to13 bands ranging from 50 Kb to 500 Kb. The genotypes of the strains could be grouped into 17 clusters (at a cut-off value of 40%) and 161 types (at a cut-off value of 80%), generating a discrimination index score of 0.997. None of these patterns indicated a predominant genotype. The dendrogram showed 24 types that gathered together only a few isolates. Majority of the types (137 types) had only one isolate. Among 161 types, fish isolates accounted for 53 types, while frog ones did 108 types.6. ERIC-PCR profilesERIC-PCR was repeated at least once for each strain in order to check the reproducibility of the experiment. In all cases, the same pattern of bands was found for each replicate. PCR products amplified with the pair ERIC primers yielded 5 to 10 bands mainly ranging in size from 200 to 2000bp. ERIC-PCR grouped into 19 clusters and 135 types (similarity cut off value equaled to PFGE), therein,109 of the strains (54.8%) into 45 genomic types (E1-E45), and 90 strains (45.2%) showed unique fingerprint profile (E46-E135), corresponding to a discrimination index score of 0.995. Among 135 types, fish isolates accounted for 49 types, while frog ones did 86 types.7. The relationship among antibiotype, PFGE and ERIC profiles7.1 A higher intraspecies gene polymorphism was described by PFGE. Among the 199 isolates studied, PFGE analysis revealed 161 genotypes and ERIC-PCR analysis revealed 135 genotypes. The isolates from different food samples with similar PFGE patterns showed different ERIC-PCR patterns (e.g., W548and Y37), vice versa (e.g., W339 and Y258). The antimicrobial resistance patterns were not related specifically to the genotypes. Most of the strains had the same antibiotic pattern were considered to be similar, the results obtained were inconsistent with those obtained by genotyping. Although fifteen fishery product specimens each yielded more than one isolate, only two (Y330 and Y336) of these shared similar PFGE and ERIC-PCR patterns. 7.2 Basically, the typing results from the three methods were lack of the concordance. The PFGE of genomic macrorestriction fragments showed a relatively high interspecies and intraspecies genetic diversity among L. hongkongensis from different aquatic products.80.9% of isolates with different PFGE patterns also yielded different ERIC-PCR patterns. A total of 19.1% of the isolates that were indistinguishable by their PFGE patterns were distinguishable by ERIC-PCR. However,47.6% of the isolates indiscernible by ERIC-PCR were discerned by PFGE.8. PCR screening for integrase genes and cassettesPCR screening for integrase genes among the 199 L. hongkongensis isolates revealed that 13 isolates (6.5%) could produce 565-bp product, suggesting the presence of class 1 integrase (intl1). No amplicon was observed after PCR targeting the class 2 integrase gene (intI2). All the 13 L. hongkongensis isolates possessing the intI gene were subject to a CS-based PCR assay flanking the variable region of integron for detection of antibiotic resistance gene.9. Sequencing and analysisAfter sequencing analysis, five different gene cassettes arrangements were identified and as follows:dfrA1-orfC, aadA1-dfrA1, dfrA17-aadA5, dfrA14-arr2-cmlA5 and dfrA32-ereA-aadA2. An approximate 1100 bp PCR products were obtained from seven strains (W10, W24, W208, W280, W358, W490 and W496) containing dfrA1-orfC gene cassette. Database search in the GenBank revealed that this cassette array had 99% identity to that of a Salmonella strain (accession no. AY963803), encoding resistance to trimethoprim (dfrA1) and the function of orfC was unknown. In case of two strains (W339 and W350) about 1500 bp amplicons were obtained which contained aadA1-dfrA1 cassette array showing 93% identity to that of a Salmonella strain (accession no. AY123251), and the presence of aminoglycoside adenyl transferase gene (aadA) was known to confer resistance to streptomycin and spectinomycin while dfrA1 was to trimethoprim. Sequencing of similar length amplicon (1500 bp) from another 2 different strains (W400, W424) revealed a gene cassette array with 100% identity to dfrA17-aadA5 gene cassette array, conferring resistance to trimethoprim and aminoglycoside, from a E. coli strain (accession no. GU055937). A newly combined gene cassette array, dfrA14-arr2-cmlA5, having～2800 bp amplicon size was discovered in W64 strain and these genes were known to confer resistance to trimethoprim, rifampin and chloramphenicol, respectively (e.g., E. coli strain accession no. AM932674 and Salmonella strain accession no. GU067640). Another array of gene cassette, dfrA32-ereA-aadA2, was detected in strain Y142 which showed 99% identity to that of a Salmonella strain (accession no. GU067642) and these genes were known to confer resistance to trimethoprim, erythromycin and aminoglycoside, respectively. Interestingly, the trimethoprim resistant genes(dfrAl, dfrA14, dfrA17 and dfrA32) were commonly present within all these strains with integron.10. The relationship between integron and antibioticsThe antibiotics resistances of L. hongkongensis isolates harboring integron were higher than non-harboring isolates for cefepime (61.5% vs 0.5%,x2=104.709, P<0.001), ciprofloxacin (30.8% vs 4.8%,x2=13.381,P<0.001), gentamicin (7.7% vs 0,x2=5.531, P=0.019), tetracycline (69.2% vs 14.5%,x2=24.549, P <0.001), trimethoprim-sulfamethoxazole (84.6% vs 7.0%,x2=69.036, P<0.001), rifampicin (100.0% vs 73.1%,/=4.667, P=0.031). There is no significant in the rest 12 breeds of anbiotics, compared with the resistance rates between isolates with integron and non-integron isolates. Thirteen isolates harboring integron were multi-resisitance (resistant to 3 antibiotics or more).11. PFGE profiles of 13 isolates harboring integronPFGE analysis revealed 13 different banding patterns among the 13 integron-positive strains. The SpeI restriction typing produced 8-13 bands ranging from 50 Kb to 500 Kb. The genotypes of the strains could be grouped into eleven sub-types (at a cut-off value of 80%). No precise correlation could be revealed among genotypes and specific integrons associated phenotypes. For example, some strains carrying dfrAl gene belonged to the highly dissimilar genotypes (A, B, C and D clusters). Even if strains possessing the same gene cassette in their integrons and isolated from the same source (W490 and W496; W339 and W350; W400 and W424) had high genetic diversity. On the contrary, isolates (W64 and Y142) belonging to the same cluster (A) possessed different gene cassettes in their integrons.Conclusions:1. Among the investigated 8 breeds of freshwater fishes, L. hongkongensis was only detected from grass carps and bighead carps. Spring has the highest isolation rate, while winter, the lowest one.2. To our knowledge, this is the first report in Chinese main land showing the detection of L. hongkongensis among six breeds of edible frogs. There is no report about the detection of Guenther's frog, Black-spectacled toad, Bull frog, and Giant spiny frog.3. The positive rate of frog isolates was higher than fish isolates.4. All strains were resistant to one class antibiotic agent at least. All the L. hongkongensis isolates were found to be resistant to cephalosporin, majority contributing to the first and third generation cephalosporin except for CAZ, but interestingly low resistance rates founding in the second and fourth generation. Among them,129 isolates (64.8%) were multidrug-resistant (resistant to 3 or more classes of antibiotics).5. Higher reproducibility and resolving power were found the profiles generated by PFGE and ERIC-PCR. Both methods can act as typing approaches for L. hongkongensis, while PFGE is better.6. The investigation in terms of antibiotypes PFGE and ERIC-PCR profiles seemed to have given sufficient proof to intraspecies gene polymorphism of L. hongkongensis isolates7. Five different gene cassettes arrangements were identified and as follows: dfrA1-orfC, aadA1-dfrA1, dfrA17-aadA5, dfrA14-arr2-cmlA5 and dfrA32-ereA-aadA2. Of the total, dfrA14-arr2-cmlA5 was a novel gene cassette array. 8. Class 1 integron got the most credits in resistance to trimethoprim-sulfamethoxazol among L. hongkongensis. A slack relationship was found between antibiotics resistance and integrons.9. The abundant intraspecies gene polymorphism was discovered due to isolates harboring integrons showing different PFGE profiles.