Class 1 integrons and plasmid-mediated multiple resistance genes of the Campylobacter species from pediatric patient of a university hospital in Taiwan

The 14 Campylobacter strains used in this study were isolated from feces sample of pediatric diarrhea patient in China Medical University Hospital storage bank from January 2014 to February 2017. The Campylobacter species was cultured on Campy-BAP plates (Becton–Dickinson and Company, USA) containing five anti-microbial agents (amphotericin B, polymyxin B, cephalothin, trimethoprim, and vancomycin) and 10% sheep blood overnight at 42 °C and in an air condition of 5% O₂, 10% CO₂, and 85% N₂. Subsequently, the strains were all re-cultivated under microaerophilic conditions. The isolated strains were then placed in storage in skim milk medium containing 20% glycerol stock at a temperature of −80 °C [17, 18].

The cultured bacteria were suspended in 300 μl of bi-distilled water and mixed with 900 ml of ethanol (Carl Roth GmbH). The measurement of the sample was then conducted following procedures previously described by El-Ashker et al. [19]. Measurement was performed with an Ultraflex III TOF/TOF mass spectrometer (Bruker Daltonics) equipped with a 200-Hz smartbeam 1 laser. The parameter setting was as follows: delay, 80 ns; ion source, 1 V, 25 kV; ion source, 2 V, 23.4 kV; lens voltage, 6 kV; and mass range 0–20 137 kDa. The raw spectra was analyzed by MALDI BIOTYPER 2.0 software (Bruker Daltonics, Billerica, MA, USA) using the default settings. The procedure of the MALDI-TOF mass spectrometry measurement was performed automatically without any user intervention. The software generated a list of peaks up to 100. The peaks with a mass-to-charge ratio difference of <250 ppm were considered to be identical. The peak list that was generated was used for matches against the reference library by directly using the integrated pattern-matching algorithms of the software. All parameters were the same regardless of the presumptive bacterial species analyzed. The BIOTYPER 2.0 database was composed using only Campylobacter and related species [20].

The Kirby–Bauer disc diffusion test and Epsilometer test (Etest) were used to determine the anti-microbial susceptibility of the Campylobacter spp. isolates [21‐23]. The susceptibility of each isolate to each antibiotic was determined according to the Neo-Sensitabs™ “user’s guide” for susceptibility testing and the latest guidelines of the Clinical and Laboratory Standards Institute (CLSI). To conduct the Kirby–Bauer disc diffusion test, sixteen antibiotic discs were chosen, including amikacin (30 μg/disc), ampicillin (30 μg/disc), cefepime (30 μg/disc), cefotaxime (30 μg/disc), chloramphenicol (30 μg/disc), ciprofloxacin (5 μg/disc), erythromycin (15 μg/disc), gentamicin (10 μg/disc), imipenem (10 μg/disc), kanamycin (5 μg/disc), nalidixic acid (30 μg/disc), streptomycin (30 μg/disc), sulfamethoxazole/trimethoprim (1.25/23.75 μg/disc), tetracycline (5 μg/disc) and tobramycin (10 μg/disc). All of which were purchased from Mast Group Ltd. (Merseyside, UK). The Campy-BAP plates were prepared and inoculated with the Campylobacter spp. isolates according to the procedures recommended by the NCCLS. The plates were incubated at 37 °C under the same microaerophilic conditions already described in the “Isolation and identification of bacterial strains” section above for 48 h.

For the Epsilometer test (Etest), five antibiotic Epsilometer strips were used, including strips containing amikacin, ciprofloxacin, imipenem, piperacillin/tazobactam, and tetracycline. For each tested organism, a Campy-BAP plate was inoculated by swabbing the plate evenly in three directions with a 1.0 McFarland standard of the test organism. The Etest strips were then applied to the surface of the plate and were incubated under the same conditions as indicated above for the disc diffusion test [22].

The plasmid DNA was extracted using a Plasmid Miniprep Purification kit (GeneMark, GMbiolab, Taichung, Taiwan) and the genomic DNA extract was prepared with the PureLink Genomic DNA Mini kit (Carlsbad, CA, USA). The polymerase chain reaction (PCR) assay was applied to detect tetracycline-resistant genes reported by Abdi-Hachesoo et al. [24, 25]. The 9 pairs of primers were used to detect tet (A), tet (B), tet (K), tet (L), tet (M), tet (O), tet (Q), tet (S), and tet (W) gene in plasmid and genomic DNA of Campylobacter isolates, respectively. The specific genes, applicons, primers, and annealing temperatures are listed in Additional file 1: Table S1. Briefly, we performed the PCR with O’In 1 DNA polymerase (Yeastern Biotech, Taipei, Taiwan) with 10 pmol of forward and reverse primers (Mission Biotech Co., Ltd., Taipei, Taiwan) and 60 ng of the extracted plasmid or chromosomal DNA. The Campylobacter jejunii isolate 3 was positive for tet (A), tet (L), tet (M), tet (O), and tet (Q), we further identified the PCR product by DNA sequencing. We used the Campylobacter jejunii-isolate 3 as positive control for these genes. The erythromycin-resistant (erm) genes were detected, and the PCR-amplification of primers for erm (A), erm (B), erm (C), and erm (F), respectively, were used to detect any Erm-R genes in the individual Campylobacter isolates. Each gene of the PCR conditions, primers, and size that was listed in Additional file 1: Table S1.

The gene cassettes of integron were harbored in all the integrase-positive isolates. In order to screen these isolates for the presence of the class 1 integrase gene, the specific primers IntI1F and IntI1R were used [26]. The specific oligonucleotide PCR primer used in this study is listed in Additional file 1: Table S1. Following the manufacturer’s instructions, template DNA from all the isolates was prepared using Plasmid Miniprep Purification kits (GeneMark). PCR was then performed in a total volume of 25 μl consisting of 1 μl of target DNA, 17 μl of distilled water, 1 μl of each primer (10 μM), and 5 μl of 5× Master Mix (PCR Master Mix Kit, GeneMark), which in turn consisted of 0.75 U Taq Plus DNA polymerase, 250 μM dNTP, 2 mM MgCl₂, and PCR buffer. More specifically, the amplicons that were used for the gene cassette of class I integron analysis were determined by the specific primers Cassette F and Cassette R [27], while the specific oligonucleotide PCR primer used is listed in Additional file 1: Table S1. The PCR procedures were executed under the same conditions as indicated above for the integrase gene detection (except that the annealing temperature was set at 55 °C for 30 s for detecting the intI1 gene while it was set at 57 °C for 45 s for detecting the gene cassette). PCR products of integron gene cassette were then sequenced from both sides. The same company (i.e., Mission Biotech Co., Ltd.) also synthesized the oligonucleotide primers used for the DNA sequencing. The nucleotide sequences were analyzed using the Blast program available on the National Center for Biotechnology Information (NCBI) website (http://blast.ncbi.nlm.nih.gov/Blast.cgi) in order to search for the closest sequences in the GenBank database (NCBI, US National Library of Medicine).

The genotyping of the 14 Campylobacter species strains was performed according to a standard operating procedure which was recommended by the Centers for Disease Control and Prevention (CDC) for PulseNet PFGE of Campylobacter spp. (http://www.cdc.gov/pulsenet). The DNA of agarose plugs was performed with SmaI restriction enzyme. The electrophoresis conditions consisted of an initial switch time of 6.8 s and a final switch time of 35.4 s and a gradient of 6 V/cm for 19 h on CHEF-DR III System (Bio-Rad, Hercules, CA, USA). The PFGE patterns were analyzed using Gel Compare II version 6.5 software (Applied Maths, Sint-Matenslatem, Belgium). The DNA size marker was used by standard Salmonella serovar Branderup H9812 strain (digested with XbaI restriction enzyme). The pair comparisons of types and cluster analyses were performed by the Dice correlation coefficient and UPGMA (unweighted pair group method with arithmetic averages) clustering algorithm.

According to the MALDI-TOF mass spectrometry results, the detection ratio of the clinical isolates for all the patient treated from January 2015 to December 2015 was 5%. The bacterial species of clinical isolates were identified by the adequate level (score ≥2.0; the addition of biochemical tests for score ≤2.0). In total samples, ten C. jejuni strains, three C. coli strains, and one C. fetus strain were isolated. Overall, the respective population rates for the C. jejuni, C. coli, and C. fetus strains were 71% (10/14), 21% (3/14), and 8% (1/14). These results indicate that C. jejuni, C. coli, and C. fetus strains were the major dominant Campylobacter species found in the samples from the diarrhea of pediatric patient.

The Kriby–Bauer disc diffusion test was used in this experiment. More specifically, a variation of the standard disc diffusion method that has been approved by the CLSI and EUCAST as an acceptable means of screening for the susceptibility of Campylobacter isolates so as to be treated with fifteen antibiotics, and the result is summarized in Table 1. It should be noted that because the breakpoints for amikacin, cefepime, streptomycin, and tobramycin have not yet been defined by the CLSI, the suggested E. coli breakpoint was used for those particular antibiotics instead. All the Campylobacter isolates exhibited full resistance to cefepime, streptomycin, and trimethoprim/sulfamethoxazole, as well as potential resistance to ampicillin, ciprofloxacin, cefotaxime, kanamycin, nalidixic acid, tobramycin, and tetracycline. Conversely, all the isolates were susceptible to amikacin, chloramphenicol, erythromycin, gentamicin, and imipenem.

All the isolates were tested by Epsilometer test (Etest) for their susceptibility to four antibiotics (amikacin, ciprofloxacin, imipenem and tetracycline), and the result is summarized in Additional file 1: Table S2. Because the breakpoints for amikacin and imipenem have not yet been defined by the CLSI, the suggested E. coli breakpoint from the same reference paper was used instead. All the isolates showed potential resistance to ciprofloxacin and tetracycline, while only five isolates exhibited resistance to amikacin, and only two isolates exhibited resistance to imipenem. Our prediction is that the will be isolates susceptible to amikacin and imipenem.

All Tet-R genes were found in plasmid, but not in genomic DNA, of tested Campylobacter spp. isolates. According the results of the screening for tetracycline-resistant genes, 79% (11/14) of the Campylobacter spp. isolates tested positive for tet (A) and tet (O) total, with tet (A) being seen in 89% (8/9) of the C. jejuni isolates and 75% (3/4) of the C. coli isolates. The tet (O) was found in 89% (8/9) of the C. jejuni isolates, 50% of the C. coli isolates, and 100% of the C. fetus isolates. Ten of the Campylobacter spp. isolates (71%) total tested positive for tet (M): it was detected in 78% of C. jejuni isolates (7/9) and 75% of the C. coli isolates (3/4). The tet (Q) resistant genes were seen in 36% of these Campylobacter spp. isolates total, being found in 33% of the C. jejuni isolates (3/9) and 50% of the C. coli isolates (2/4). The tet (L) (Fig. 1), which had the lowest prevalence rate, was found in only three of the Campylobacter spp. isolates total (21%), being detected in 22% of the C. jejuni isolates (2/9) and 25% of C. coli isolates (1/4). The Campylobacter isolates were screened to determine which harbored any of nine tetracycline-resistant genes, and the result is summarized in Additional file 1: Table S3. The tet (B), tet (K), tet (S) and tet (W) resistant genes were not identified in any of the tested Campylobacter spp. isolates.

Bild vergrößern

The result of the PCR detect Erm-R genes in 36% (5/14) erythromycin-resistant isolates showed that only the erm (B) (Fig. 1) (2 C. jejuni isolates, 22%; 1 C. coli isolate, 25%) was found (Additional file 1: Table S4). The erm (A), erm (C), and erm (F) resistant genes were not identified in any of the tested Campylobacter spp. isolates.

In Additional file 1: Table S5, we show the characteristics of various gene cassette types in class 1 integron of Campylobacter species isolates. The integrase gene (intI1) was detected in 86% of the fourteen Campylobacter isolates (12/14) identified in this study. Among the intI1-positive strains, twelve isolates (86%) harbored class 1 integron containing various sizes of gene cassettes ranging from 750 to 1907 b.p. in length.

The results of the PCR testing to detect Class I integron cassette genes indicated that 86% (12/14) of the Campylobacter spp. isolates tested positive for Class I integron cassette genes (Fig. 2). According to the sequencing analysis, two gene cassette types were identified in this study (namely, type I: dfr12-gcuF-aadA2 genes and type II: dfrA7 gene). Each cassette type had a unique inserted gene cassette pattern harboring various antibiotic-resistant genes. The most common gene cassette type (type I: dfr12-gcuF-aadA2 genes) was found in 92% of the 12 intI1-positive strains (9 C. jejuni strains and 3 C. coli strains) and was responsible for trimethoprim, streptomycin, and spectinomycin resistance in those strains. The other cassette type (type II: dfrA7 gene) was found in only 8% of the intI1-positive strains (i.e., one C. coli strain) and was responsible for trimethoprim resistance in that strain.

Bild vergrößern

In Fig. 3, among 14 Campylobacter isolates were typeable by PFGE, that the guidelines of PFGE patterns were applied by Tenover [28]. PFGE analysis was performed in order to determine the genetic diversity of these isolates and the relationship between the isolates retrieved from different patients. Analysis of the PFGE profiles of the Campylobacter isolates suggested that the isolates possessed diverse genotypes, when comparing profiles of isolates belonging to different species. The similarity of the isolates was from 50 to 100%. These 14 isolates had 13 different PFGE genotypes, only one PFGE type was recovered between two C. jejuni isolates. Furthermore, by using a cut-off similarity value of 75%, profiles of the C. jejuni were classified to 6 clusters and C. coli were classified to 2 clusters.

Bild vergrößern