Journal of Medical Microbiology (2014), 63, 659–666

DOI 10.1099/jmm.0.071498-0

A PCR-RFLP assay for the detection and differentiation of Campylobacter jejuni, C. coli, C. fetus, C. hyointestinalis, C. lari, C. helveticus and C. upsaliensis Kazumasa Kamei,1,2 Masahiro Asakura,1 Srinuan Somroop,1 Noritoshi Hatanaka,1 Atsushi Hinenoya,1 Akira Nagita,3 Naoaki Misawa,4 Motoo Matsuda,5 Shinsaku Nakagawa2 and Shinji Yamasaki1 Correspondence

1

Shinji Yamasaki

2

[email protected]

Graduate School of Life and Environmental Sciences, Osaka Prefecture University, Japan Graduate School of Pharmaceutical Sciences, Osaka University, Japan

3

Department of Pediatrics, Mizushima General Hospital, Okayama, Japan

4

Faculty of Agriculture, University of Miyazaki, Japan

5

Graduate School of Environmental Health Sciences, Azabu University, Kanagawa, Japan

Received 6 December 2013 Accepted 21 February 2014

Although Campylobacter jejuni and Campylobacter coli are the most common bacterial causes of human gastrointestinal diseases, other Campylobacter species are also involved in human and animal infections. In this study, we developed a cytolethal distending toxin (cdt) gene-based PCRRFLP assay for the detection and differentiation of C. jejuni, C. coli, C. fetus, C. hyointestinalis, C. lari, C. helveticus and C. upsaliensis. Previously designed common primers, which can amplify the cdtB gene of C. jejuni, C. coli and C. fetus, were used for detecting seven Campylobacter species and differentiating between them by restriction digestion. The PCR-RFLP assay was validated with 277 strains, including 35 C. jejuni, 19 C. coli, 20 C. fetus, 24 C. hyointestinalis, 13 C. lari, 2 C. helveticus, 22 C. upsaliensis, 3 other Campylobacter spp. and 17 other species associated with human diseases. Sensitivity and specificity of the PCR-RFLP assay were 100 % except for C. hyointestinalis (88 % sensitivity). Furthermore, the PCR-RFLP assay successfully detected and differentiated C. jejuni, C. coli and C. fetus in clinical and animal samples. The results indicate that the PCR-RFLP assay is useful for the detection and differentiation of seven Campylobacter species important for human and animal diseases.

INTRODUCTION Campylobacters normally reside in the intestine of animals such as poultry, pigs, cattle, etc., without causing any symptoms (Man, 2011). Human infection is mostly attributed to the consumption of contaminated meats such as chicken, and to contact with companion animals (Carbonero et al., 2012; Hald et al., 2004; Steinhauserova et al., 2000; Young et al., 2007). Campylobacters are considered to be a leading cause of zoonotic disease in humans worldwide and are also associated with septicaemia and severe complications caused by post-infection, such as pancreatitis, reactive arthritis, haemolytic–uraemic syndrome and Guillain–Barre´ syndrome (Man, 2011; SivadonTardy et al., 2013; Yuki & Hartung, 2012). The GenBank/EMBL/DDBJ accession numbers for the cdtB gene sequences analysed in this study are given in Methods. A supplementary table is available with the online version of this paper.

071498 G 2014 The Authors

Printed in Great Britain

The genus Campylobacter currently comprises 25 species, 12 of which have been isolated from patients with gastroenteritis (Man, 2011). Among these 12 species, Campylobacter jejuni and Campylobacter coli are the most common bacterial causes of human gastrointestinal diseases (Man, 2011). Recently, a growing number of Campylobacter spp. other than C. jejuni and C. coli have also been recognized as important pathogens in humans and animals (Bullman et al., 2011; Inglis et al., 2011; Man, 2011). However, non-C. jejuni/C. coli strains are difficult to isolate by currently available methods. Most of the culture methods currently in use favour the isolation of C. jejuni and C. coli. For example, antimicrobial agents are often used to suppress the growth of non-C. jejuni/ C. coli species. Some campylobacters, such as Campylobacter upsaliensis, Campylobacter hyointestinalis and Campylobacter fetus, cannot grow well in the presence of cephem antibiotics such as cefoperazone. This antibiotic is included in 659

K. Kamei and others

Butzler medium, modified charcoal cefoperazone deoxycholate agar (mCCDA) and Bolton broth. In fact, C. fetus, C. hyointestinalis, Campylobacter lari and C. upsaliensis, in addition to C. jejuni and C. coli, have successfully been isolated from patients with diarrhoea by using non-selective media, for example in the Cape Town protocol (Lastovica & le Roux, 2000; Prouzet-Maule´on et al., 2006; Vandenberg et al., 2004). Furthermore, a hydrogen-enriched atmosphere (generally 3–7 % hydrogen) is required for the growth of some Campylobacter species, such as Campylobacter concisus and some C. hyointestinalis, which is not routinely used in most laboratories. Therefore, information on the species present in a clinical sample, as might be gained by PCR, may aid selection of appropriate culture methods for its isolation. Cytolethal distending toxin (CDT) is known to be a possible virulence factor in many pathogenic Gram-negative bacteria, including Campylobacter spp. (Yamasaki et al., 2006). The cdt gene cluster, consisting of cdtA, cdtB and cdtC, was shown to be ubiquitously present in C. jejuni, C. coli and C. fetus in a species-specific manner (Asakura et al., 2007). Furthermore, cdt genes are also known to be present in C. lari (Matsuda et al., 2008), C. upsaliensis (Fouts et al., 2005), C. hyointestinalis (W. Samosornsuk and others, unpublished observation) and C. helveticus (S. Somroop and others, unpublished observation) indicating that cdt may be one of the best target genes to identify Campylobacter species that are important for human and animal diseases. In this study, we have successfully developed a PCR– restriction fragment length polymorphism (PCR-RFLP) assay using a common primer set designed from the conserved regions of the cdtB gene for the detection and differentiation of these seven Campylobacter species and evaluated it with 277 strains, including 35, 19, 20, 24, 13, 2, 22 and 4 of C. jejuni, C. coli, C. fetus, C. hyointestinalis, C. lari, C. helveticus, C. upsaliensis and three other species of Campylobacter, respectively, and 17 other bacterial species belonging to 10 different genera associated with humans and animals. Furthermore, 21 clinical specimens from diarrhoeal patients and 10 bovine bile samples were also analysed by this PCR-RFLP assay.

METHODS Bacterial strains, media and growth conditions. Details of

bacterial strains used in this study are summarized in Table 1. Campylobacter strains isolated from clinical samples were identified as Campylobacter spp. by culture methods, Gram staining, morphology under microscopy and biochemical tests (catalase and oxidase activities). If necessary, biochemical tests using API Campy (bioMe´rieux), 16S rRNA gene sequencing, detection of the hipO gene by PCR and the cdt gene-based multiplex PCR assay were used for the identification at species level (Asakura et al., 2008; Shiramaru et al., 2012). Campylobacter spp., Arcobacter spp. and Helicobacter spp. except C. concisus, Campylobacter curvus and Campylobacter hominis, were grown on blood agar [blood base agar no. 2 (Oxoid) supplemented with 5 % (v/v) defibrinated horse blood (Nippon Bio-Supp.)] under microaerobic conditions (5 % O2, 7.5 % CO2, 7.5 % H2, 80 % N2) at 37 uC for 2 days or more. C. concisus, C. curvus and C. hominis strains 660

were grown on blood agar containing 6 % formate and fumarate under anaerobic conditions (80 % N2, 10 % CO2, 10 % H2). Aggregatibacter actinomycetemcomitans was grown on trypticase soy agar (TSA; Becton Dickinson) supplemented with 0.6 % yeast extract (Becton Dickinson) at 37 uC for 2 days under 10 % CO2 in 90 % air. Shigella spp., Vibrio cholerae, Providencia alcalifaciens and Escherichia coli were grown in Luria–Bertani broth (LB; Becton Dickinson) at 37 uC overnight. Escherichia albertii and Salmonella enterica were grown in trypticase soy broth (TSB; Becton Dickinson) at 37 uC overnight. Yersinia enterocolitica was grown in TSB at 30 uC for 2 days. Vibrio parahaemolyticus was grown in alkaline peptone water (Nissui) containing 3 % sodium chloride at 37 uC overnight. Haemophilus ducreyi was grown in brain–heart infusion broth (Becton Dickinson) supplemented with 10 % (v/v) fetal bovine serum (Life Technologies), 10 g haemoglobin l21 (Becton Dickinson) and 1 % (v/v) IsoVitalex (Becton Dickinson) at 37 uC for 5 days under microaerobic conditions. DNA preparation. Template DNA for PCR assay was prepared by the boiling method as previously described (Hoshino et al., 1998), with slight modifications. Briefly, bacteria were grown on appropriate agar plates and suspended in 1 ml TE buffer (10 mM Tris/HCl, 1 mM EDTA pH 8.0), or 50 ml mid-exponential phase bacterial culture was added to 450 ml TE buffer. Alternatively, 180 ml clinical and animal samples were mixed with 20 ml 106 TE buffer. The suspension was boiled for 10 min, kept on ice for 5 min, and centrifuged at 12 800 g for 5 min. Supernatant was collected and stored at 220 uC until use. Supernatant (1 ml) was used as DNA template for PCR assay. PCR method. C. jejuni strain 81-176 (Pickett et al., 1996), C. coli strain Co1-243 (Asakura et al., 2008), C. fetus strain ATCC 27374T (Asakura et al., 2008), C. hyointestinalis strain Ch022 (W. Samosornsuk and others, unpublished observation), C. lari strain 298 (Matsuda et al., 2008), C. helveticus strain ATCC 51209T (Stanley et al., 1992) and C. upsaliensis strain ATCC 43954T (Standstedt & Ursing, 1991) were used as positive controls (reference strains) and E. coli strain C600 was used as a negative control. PCR was performed with a common primer set, C-CdtBcom1 and C-CdtBcom2, as described by Asakura et al. (2007), using a TaKaRa PCR Thermal Cycler (Takara Bio) or Applied Biosystems GeneAmp PCR 9700 (Life Technologies). Oligonucleotide primers were from GeneDesign (Osaka). PCR mixture contained 0.5 mM each of C-CdtBcom1 and C-CdtBcom2 primers, 1 ml template DNA, 0.2 mM each of dATP, dCTP, dGTP and dTTP, 16 Ex Taq DNA polymerase buffer, and 1.0 U Ex Taq DNA polymerase (Takara Bio) in a 40 ml reaction volume. The samples were subjected to an initial denaturation at 94 uC for 3 min, followed by 30 amplification cycles, each consisting of 94 uC for 30 s, 50 uC for 30 s and 72 uC for 30 s. A final primer extension at 72 uC for 3 min was included. PCR products were analysed by 2 % agarose gel electrophoresis and bands were visualized with UV light after staining with ethidium bromide. Images were captured on a Bio-Rad Gel Doc system (Bio-Rad). DNA sequence analysis. The DNA sequence of each of the PCR

products was determined using an ABI PRISM 3100-Avant Genetic Analyzer (Life Technologies). Sequencing was performed by the chain-termination method with the BigDye terminator v1.1 cycle sequencing kit (Life Technologies). The sequences obtained in this study were analysed using the DNA Lasergene software package (DNASTAR). Sequence homology search was performed using BLASTN (National Center for Biotechnology Information). PCR-RFLP assay. PCR products (1 to 5 ml) were digested with

5 U DdeI (New England Biolabs) and/or 7.5 U EcoRI (Takara Bio) in a final volume of 10 ml at 37 uC overnight. The resulting fragments were analysed by 4 % agarose gel electrophoresis as described above. Journal of Medical Microbiology 63

Detection and differentiation of seven campylobacters

Detection limit of the cdtB gene-specific PCR. Campylobacter

cells freshly grown on blood agar were suspended in sterile PBS and each bacterial suspension was 10-fold serially diluted using sterile PBS. Each dilution was spread on blood agar to determine viable bacterial cell count and also used to prepare template DNA by the boiling method as described above. For PCR assay only 1 ml of the prepared DNA was used as template. Analysis of clinical samples. Rectal swabs were collected from 21

children with diarrhoea (from 3 months to 13 years and 9 months of age) who visited the paediatric department of Mizushima Central Hospitals, Okayama, Japan. The specimen was suspended in 500 ml sterile saline. DNA template was prepared by the boiling method as described above and 1 ml template DNA was used for PCR-RFLP assay. One loop of stool suspension (approx. 10 ml) was inoculated onto mCCDA (Oxoid) agar. A 100 ml portion of the suspension was also applied onto a mixed cellulose ester-type membrane filter (0.45 mm pore size, 47 mm diameter; Advantec Toyo) placed on a blood agar plate, and the filter was removed after 30 min of incubation at room temperature (Shiramaru et al., 2012). Both of the agar plates were incubated under microaerobic conditions at 37 uC for 2–3 days. Suspected colonies were subcultured on blood agar plates, followed by Gram staining, examination of morphology by microscopy, and the cdt gene-based multiplex PCR (Asakura et al., 2008). Analysis of animal samples. Bile was randomly collected from ten

healthy cattle. Preston broth (CM0067, Oxoid) supplemented with antibiotics (SR0117, Oxoid), growth supplement (SR0232, Oxoid) and 5 % (v/v) defibrinated horse blood was prepared according to the manufacturer’s recommendation. For enrichment culture, 1 ml bile was mixed with 4 ml 1.256 Preston broth and incubated at 37 uC under microaerobic conditions for 24 h. A 1 ml aliquot of bile with or without enrichment culture was centrifuged at 200 g for 10 min. The supernatant was collected and centrifuged at 12 800 g for 10 min. The pellet was washed once with sterile PBS and suspended in 1 ml sterile PBS. The suspension was used for the preparation of DNA template and isolation of campylobacters. DNA template was prepared by the boiling method as described above and 1 ml template DNA was used for PCR-RFLP assay. Bile samples without enrichment were inoculated onto mCCDA and blood agars using the filtration method, and bile samples with enrichment were inoculated onto mCCDA as described above. Culture plates were incubated at 37 uC under microaerobic conditions for 3 days. Suspected colonies were subcultured and species were identified as described above. Spike experiment. A stool specimen was obtained from a healthy

person. A 2 g portion of this stool specimen was suspended in 8 ml sterile PBS. The suspension was enriched with Preston broth and the stool samples with or without enrichment culture were inoculated onto mCCDA agar plates or filter membranes on blood agar plates as described above. Culture plates were incubated at 37 uC under microaerobic conditions for 5 days. Portions (0.2 g) of the healthy stool specimen were spiked with 103 to 108 c.f.u. of each reference strain, such as C. jejuni strain 81-176, C. coli strain Co1-243 and C. hyointestinalis strain Ch022. The spiked sample was suspended in 800 ml sterile PBS and the suspension was centrifuged at 200 g for 10 min. The supernatant was collected and centrifuged at 12 800 g for 10 min. The pellet was suspended in 1 ml sterile PBS, and the suspension was centrifuged again under the same conditions. The pellet was suspended in 500 ml sterile PBS. The sample was boiled for 10 min, kept on ice for 5 min, and centrifuged at 12 800 g for 5 min. Supernatant was collected and treated with phenol/chloroform/isoamyl alcohol (25 : 24 : 1, by vol.). DNA was precipitated with ethanol and http://jmm.sgmjournals.org

suspended in 200 ml TE. The suspension was treated with RNase A (50 mg ml21) at 37 uC for 30 min, and phenol/chloroform/ isoamyl alcohol, and the DNA was precipitated with ethanol. Finally, the purified DNA was suspended in 30 ml TE and 1 ml was used as DNA template for PCR-RFLP assay. Nucleotide sequence accession numbers. The cdtB gene sequences

analysed in this study have been registered in the DNA Data Bank of Japan (DDBJ), and their details (strain/accession no.) are as follows. C. jejuni: Co1-008/AB872826, Co1-119/AB872827, Co1-126/AB872828, Co2-127/AB872829, Co2-128/AB872830, Co2-132/AB872831, Co2-193/ AB872832, Co2-200/AB872833, Co2-214/AB872834, Co3-007/AB872835, Co3-011/AB872836, Co3-012/AB872837, Co3-024/AB872838, Co3-036/ AB872839, Co3-072/AB872840, Co3-078/AB872841, Co3-082/AB872842, B86/AB872843, 8214c/AB872844, 8215a/AB872845, 8414c/AB872846, 9914b/AB872847, 10114a/AB872848, 10114c/AB872849. C. coli: Co1-071/ AB872850, Co1-124/AB872851, Co1-130/AB872852, Co1-194/AB872853, Co1-245/AB872854, Co2-082/AB872855, Co2-173/AB872856, Co2-218/ AB872857, Co3-134/AB872858. C. fetus: 2-1/AB872859, 7/AB872860, 86c/AB872861, 7915b/AB872862, 8013a/AB872863, 8013c/AB872864, 8512a/AB872865, 8614c/AB872866, 8813a/AB872867, 8813c/AB872868. C. hyointestinalis: 1-1/AB872869, 10-1/AB872870, 87-4/AB872871, 946/AB872872, 2003/AB872873, 2030/AB872874, 2032/AB872875, 2033/ AB872876, 2034/AB872877, 2035/AB872878, 2037/AB872879, 2039/ AB872880, 2973/AB872881, 3014/AB872882, 3158/AB872883, 3197/ AB872884, 3477/AB872885, 3535/AB872886, 3839/AB872887, 3857/ AB872888. C. upsaliensis: ATCC 43954/AB872889, 12-1/AB872890, 371/AB872891, 13-1/AB872892, 21-1 /AB872893, 26-4/AB872894, 40-1/ AB872895, 41-2/AB872896, 42-3/AB872897, 48-1/AB872898, 49-1-1/ AB872899, 60-1/AB872900, 66-1/AB872901, 68-3/AB872902, 70-3/ AB872903, 99-1/AB872904, 101-1/AB872905, 102-1/AB872906, 104-1/ AB872907, 115-1/AB872908, G1104/AB872909, Maririn/872910. C. lari: ATCC 43675/AB872911, 918/AB904782, 1500/AB904783, 1502/ AB904784, 1504/AB904785, 1578/AB904786, 1601/AB904787. C. helveticus: ATCC 51209T/AB872912.

RESULTS Evaluation of cdtB gene-based PCR using a common primer set A common primer set, C-CdtBcom1 and C-CdtBcom2, was designed from the conserved regions of the cdtB gene in C. jejuni, C. coli and C. fetus as described by Asakura et al. (2007) and these regions were also found to be relatively conserved in the cdtB gene of C. lari (Matsuda et al., 2008), C. upsaliensis (Fouts et al., 2005), C. hyointestinalis (W. Samosornsuk and others, unpublished observation) and C. helveticus (S. Somroop and others, unpublished observation). The sensitivity and specificity of PCR using the common primer set were evaluated with Campylobacter and other bacterial strains listed in Table 1. The expected size of bands (711–723 bp) was obtained from all the reference strains, including C. jejuni, C. coli, C. fetus, C. hyointestinalis, C. lari, C. helveticus and C. upsaliensis (data not shown). No PCR products were obtained from E. coli strain C600 used as a negative control. By DNA sequence analysis, these PCR products were confirmed to be highly homologous to the cdtB gene of each species (Table S1, available in the online Supplementary Material). The sensitivity of the PCR using this common primer set was evaluated with the strains (n5135) belonging to the seven target Campylobacter 661

K. Kamei and others

Table 1. Bacterial strains used in this study and detection of the cdtB gene Bacterial species (n)*

Campylobacter jejuni (35)

Campylobacter coli (19)

Campylobacter fetus (20)

Campylobacter hyointestinalis (24)

Campylobacter lari (13)

Campylobacter helveticus (2) Campylobacter upsaliensis (22) Campylobacter concisus (2)D Campylobacter curvus (1)D Campylobacter hominis (1)D Helicobacter hepaticus (1) Haemophilus ducreyi (1) Aggregatibacter actinomycetemcomitans (1) Shigella dysenteriae (13) Shigella sonnei (3) Escherichia coli (22) Escherichia albertii (1) Providencia alcalifaciens (1) Arcobacter butzleri (1)D Arcobacter skirrowii (1)D Helicobacter pylori (2)D Helicobacter fennelliae (1)D Shigella flexneri (5)D Salmonella spp. (33)D Yersinia enterocolitica (1)D Vibrio cholerae (23)D Vibrio parahaemolyticus (28)D

Origin (n)

Human (23) Animal (8) 81-176 ATCC 33560T ATCC 43432 ATCC 700819 Human (16) Co1-243 ATCC 33559T ATCC 43478 Human (2) Animal (16) ATCC 27374T ATCC 19438T Animal (22) Ch022 ATCC 35217T Human (2) Mussel (2) Seawater (6)** 298 ATCC 43675 JCM2530T Animal (1) ATCC 51209T Animal (21) ATCC 43954T ATCC 33237T ATCC 51562 ATCC 35224T ATCC BAA-381T ATCC 51448T ATCC 700724 Human (1) Human (13)d Human (3)§ Human (14)|| Animal (8) JCM17328T Human (1) ATCC 49616T ATCC 51132T ATCC 43504T ATCC 43629 ATCC 35684T Human (5) Human (33) Human (1) Human (23)# Shrimp (28)

cdtB gene-based PCR

PCR–RFLP

No. positive

Positive rate (%)

Pattern (n)

Identification rate (%)

23 8 1 1 1 1 16 1 1 1 2 16 1 1 20 1 0 2 2 6 1 1 1 1 1 21 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

100

Cj (23) Cj (8) Cj (1) Cj (1) Cj (1) Cj (1) Cc-I (14), Cc-II (1), Cc-III (1) Cc-I (1) Cc-I (1) Cc-I (1) Cf (2) Cf (16) Cf (1) Cf (1) Chy (20) Chy (1) – Cl (2) Cl (2) Cl (6) Cl (1) Cl (1) Cl (1) Che (1) Che (1) Cu-I (15), Cu-II (4), Cu-III (2) Cu-I (1) – – – – – – – – – – – – – – – – – – – – – – –

100

100

100

88

100

100 100 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

100

100

88

100

100 100 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

The reference strains are underlined. T, type strain; –, not done. *Number in parentheses indicates the number of strains analysed. DThe presence of the cdt gene was not reported. 662

Journal of Medical Microbiology 63

Detection and differentiation of seven campylobacters

Table 1. cont. dSeven strains are cdt-gene negative. §Two strains are cdt-gene negative. ||Four strains are Cdt-I positive, one strain is Cdt-II positive, two strains are Cdt-III positive, four strains are Cdt-IV positive and three strains are Cdt-V positive. One strain is Cdt-I positive, one strain is Cdt-II positive, three strains are Cdt-III positive, one strain is Cdt-IV positive and two strains are Cdt-V positive. #Thirteen strains are O1 and ten strains are non-O1/non-O139. **These strains were provided by Junko Isobe, Toyama Institute of Health, Toyama, Japan.

species. A specific PCR product was obtained from each of the strains of these seven species except for three strains of C. hyointestinalis (Table 1). The sensitivity was determined to be 88 % (21/24) in the case of C. hyointestinalis and 100 % for all other target species. Interestingly, crude cell lysates prepared from three PCR-negative C. hyointestinalis strains showed cytotoxicity to HeLa cells (data not shown), indicating that cdt-like genes might be present in these PCR-negative C. hyointestinalis strains. Specificity of the assay, evaluated with other Campylobacter species, such as C. concisus, C. curvus and C. hominis, and 10 different genera including 17 different species given in Table 1, was found to be 100 %. Irrespective of cdt gene cluster possession, cdtB gene-specific PCR products were not obtained from other Campylobacter and bacterial species tested (Table 1). The detection limit was evaluated with each reference strain and determined to be 10 to 102 c.f.u. per tube for C. jejuni, C. coli, C. fetus, C. lari, C. helveticus and C. upsaliensis, and 103 to 104 c.f.u. per tube for C. hyointestinalis (data not shown). Development and evaluation of a PCR-RFLP assay DNA sequences of the cdtB gene of reference strains (Table 1) were compared to see whether or not appropriate restriction sites are available for the RFLP analysis to differentiate the seven target species. It was found that the combination of DdeI and EcoRI restriction enzymes would generate speciesspecific RFLP patterns. To assess the utility of the PCR-RFLP assay for the species identification, restriction digestion with both DdeI and EcoRI or single digestion was carried out with PCR products obtained using genomic DNAs of 132 Campylobacter strains (Table 1). Restriction digestion revealed that the RFLP patterns of reference strains could clearly differentiate all the seven target species of Campylobacter (Fig. 1a, lanes 2, 3, 6–10). Furthermore, RFLP patterns of all C. jejuni, C. fetus, C. hyointestinalis, C. lari and C. helveticus strains, 17 out of 19 C. coli strains and 16 out of 22 C. upsaliensis strains were the same as each of the reference strains used (Table 1). However, the RFLP patterns of two C. coli and six C. upsaliensis strains were found to be different from their corresponding reference strains (Fig. 1a, lanes 4, 5 and 11, 12, respectively). To examine whether the PCR products resembled the cdtB gene of C. coli or C. upsaliensis, http://jmm.sgmjournals.org

nucleotide sequence analysis was done for all these eight PCR products. This indicated nearly identical sequence to that of the cdtB gene of C. coli or C. upsaliensis (Table S1). However, some mutations resulted in loss or gain of DdeI or EcoRI restriction site (data not shown). Therefore, the PCR-RFLP patterns obtained from C. coli strains Co1-243, Co1-194, Co2-173, and C. upsaliensis strains ATCC 43954T, 40-1, 99-1 were designated Cc-I, Cc-II, Cc-III, and Cu-I, Cu-II, Cu-III, respectively (Fig. 1a, lanes 3–5 and 10–12, respectively). RFLP patterns of C. coli Co1-194 and C. helveticus ATCC 51209T strains (Cc-II and Che) were found to be very similar (Fig. 1b, lanes 6–8). However, single digestion with either DdeI or EcoRI could clearly differentiate these two species (Fig. 1b, lanes 2–5). Differentiation of Campylobacter species by PCR product sequencing Since PCR-RFLP assay did not differentiate very clearly between C. coli and C. helveticus, we tested whether sequencing of the PCR products could directly differentiate the seven Campylobacter species. Among 132 Campylobacter strains examined in this study, the sequences of the cdtB genes of 39 strains have already been reported (Asakura et al., 2007; Matsuda et al., 2008; Parkhill et al., 2000; Pickett et al., 1996; W. Samosornsuk and others, unpublished observation; S. Somroop and others, unpublished observation). Therefore, the PCR products (each about 720 bp in size) of the remaining 93 strains were sequenced to confirm the species by comparison with the cdtB gene of each reference strain. Nucleotide sequences of all PCR products were matched with that of either C. jejuni (92.9–100 % identity), C. coli (99.2–100 %), C. fetus (99.3–100 %), C. hyointestinalis (97.9–100 %), C. lari (99.2–100 %), C. helveticus (100 %) or C. upsaliensis (96.8–100 %) (Table S1). All results obtained by the PCR-RFLP assay were completely identical to those obtained by sequencing. The results suggest that sequencing of PCR products can also be utilized for differentiation of the seven species of Campylobacter. Application of PCR-RFLP assay to human and animal samples To evaluate whether PCR-RFLP assay is applicable to human samples, we used the PCR-RFLP assay developed in 663

K. Kamei and others (a) bp

1

2

3

4

5

6

7

8

9 10 11 12 13 14

(b) bp

800 700 600 500

800 700 600 500

400

400

300

300

200

200

100

100

1

2

3

4

5

6

7

8

9

Fig. 1. PCR-RFLP analysis using representative strains of seven Campylobacter species with typical RFLP patterns. (a) DdeI and EcoRI double-digested PCR product. Lanes: 1 and 14, 100 bp DNA ladder; 2, C. jejuni (strain 81-176); 3, C. coli (Co1243); 4, C. coli (Co1-194); 5, C. coli (Co2-173); 6, C. fetus (ATCC 27374T); 7, C. hyointestinalis (Ch022); 8, C. lari (298); 9, C. helveticus (ATCC 51209T); 10, C. upsaliensis (ATCC 43954T); 11, C. upsaliensis (40-1); 12, C. upsaliensis (99-1); 13, C. upsaliensis (99-1, undigested). (b) PCR products of C. coli (Co1-194, lanes 2, 4, 6, 8) or C. helveticus (ATCC 51209T, lanes 3, 5, 7) were digested with DdeI (lanes 2, 3) or EcoRI (lanes 4, 5), or both (lanes 6–8). In lanes 1 and 9, a 100 bp DNA ladder was used as molecular mass marker.

this study with DNA templates prepared from clinical specimens of diarrhoeal patients. We tested 7 and 14 Campylobacter culture-positive and -negative stool samples, respectively (Table 2). The PCR-RFLP assay detected campylobacters from all seven stool samples that were culture-positive (C. jejuni, C. coli and C. fetus were isolated from five, one and one samples, respectively), indicating that the sensitivity was 100 %. No PCR bands were obtained from the 14 stool samples that were culture-negative, indicating that the specificity of the PCR-RFLP assay was also 100 %. We further evaluated the PCR-RFLP assay with ten bovine bile samples. The PCR-RFLP assay detected C. jejuni directly from three bile samples and from two bile samples that were enriched (Table 2). Among ten bile samples, C. jejuni was isolated from six by the mCCDA culture method. Thus, the sensitivity was determined to be 83 % (five PCR-RFLP-positives versus six culture-positives). Since no PCR products were obtained from the other four samples, which were also Campylobacter culture-negative, the specificity of the PCR-RFLP assay was thus 100 %. Detection limit of campylobacters in stool specimens To examine the detection limit of campylobacters in stool specimens, the PCR-RFLP assay was evaluated by the ‘spike experiment’ (see Methods). A stool specimen, in which campylobacters were undetectable by culture methods, was spiked with reference strains of C. jejuni, C. coli or C. hyointestinalis. The DNA templates prepared from the spiked samples were used for the PCR-RFLP assay and the 664

detection limit was determined to be 36106 c.f.u. g21 for C. jejuni, 26106 c.f.u. g21 for C. coli and 26107 c.f.u. g21 for C. hyointestinalis.

DISCUSSION The campylobacters most frequently isolated from human diarrhoeal stool specimens are C. jejuni and C. coli (Bullman et al., 2011; Inglis et al., 2011). Culture methods currently in use for isolation of Campylobacter species are suitable for cephem-resistant species such as C. jejuni and C. coli (Gharst et al., 2013; Noormohamed & Fakhr, 2012). Campylobacter species susceptible to cephem antibiotics, such as C. fetus, C. hyointestinalis, C. helveticus and C. upsaliensis, might be underestimated. Indeed, non-C. jejuni/C. coli strains such as C. fetus, C. upsaliensis, C. hyointestinalis, have been successfully isolated from patients with diarrhoea by means of non-selective media (Lastovica & le Roux, 2000; ProuzetMaule´on et al., 2006; Vandenberg et al., 2004). If we knew which Campylobacter species is present in human clinical specimens or animal samples such as meat and stool specimens before isolation, we could choose an appropriate selective medium and culture condition. This could increase the isolation rate of non-C. jejuni/C. coli strains, resulting in a better understanding of the real picture of Campylobacter infections. For this reason, in the present study, we attempted to develop and evaluate a PCR-RFLP assay targeting the cdtB genes of seven Campylobacter species for detection and differentiation at the species level. Asakura et al. (2007) have reported that cdt genes are ubiquitously present in C. jejuni, C. coli and C. fetus in a Journal of Medical Microbiology 63

Detection and differentiation of seven campylobacters

Table 2. Identification of Campylobacter spp. from human and animal samples Sample

ID

Culture isolation

PCR–RFLP

Filtration mCCDA Human faeces

Bovine bile

P6496 P6497 P6498 P6506 P6507 P6508 P6509 P6510 P6511 P6512 P6513 P6514 P6515 P6516 P6517 P6518 P6519 P6520 P6521 P6538 P6539 130911-1 130911-2 130911-3 130911-4 130911-5 130918-1 130918-2 130918-3 130918-4 130918-5

– C. jejuni – – C. jejuni C. fetus – – – – – – – C. coli C. jejuni C. jejuni – – – – – – – – – – – – – – –

– – – – – C. fetus – – – – – – C. jejuni C. coli C. jejuni C. jejuni – – – – – C. jejuni C. jejuni – C. jejuni – – C. jejuni C. jejuni C. jejuni –

– C. jejuni – – C. jejuni C. fetus – – – – – – C. jejuni C. coli C. jejuni C. jejuni – – – – – C. jejuni* C. jejuni – C. jejuni* – – –– C. jejuni C. jejuni –

–, Campylobacter was not isolated. *C. jejuni was identified from enrichment cultures.

species-specific manner, and developed a cdt gene-based multiplex PCR (Asakura et al., 2008) which successfully identified these three campylobacters at the species level (Kabir et al., 2011; Samosornsuk et al., 2007; Shiramaru et al., 2012). Recently, the entire nucleotide sequences of the cdt gene cluster of C. upsaliensis (Fouts et al., 2005), C. lari (Matsuda et al., 2008), C. hyointestinalis (W. Samosornsuk and others, unpublished observation) and C. helveticus (S. Somroop and others, unpublished observation) have also been determined. Comparison of the nucleotide sequences of these cdtB genes revealed that a common primer set, which can amplify the cdtB gene of C. jejuni, C. coli or C. fetus, could also amplify the cdtB gene present in the other four species of Campylobacter. In this study, we clearly demonstrated that the common primers could amplify the cdtB gene of seven Campylobacter species, and cdtB genes of these four species are also ubiquitously present in a http://jmm.sgmjournals.org

species-specific manner. Furthermore, restriction digestions of the PCR products clearly differentiated species of campylobacters (Fig. 1). However, RFLP patterns of some C. coli (e.g. Col-194) and C. helveticus (e.g. ATCC 51209T) strains were very similar (Fig. 1b, lanes 6–8). In this case, single digestion with either DdeI or EcoRI could clearly differentiate these two species (Fig. 1b, lanes 2–5). Alternatively, sequencing of the cdt gene-specific amplified PCR product may be used to differentiate Campylobacter species. As shown in this study, the nucleotide sequence of the cdtB gene is very well conserved among Campylobacter species. If an ambiguous result was obtained by RFLP analysis, we could sequence the PCR product to identify the species. However, the major disadvantages are that sequencing reagents are expensive and that an automated DNA sequencer may not be available in an ordinary clinical laboratory involved in routine microbiological examinations. We further evaluated whether the PCR-RFLP assay is applicable to human and animal samples. DNA templates prepared from Campylobacter culture-positive clinical specimens of diarrhoeal patients and bovine bile followed by PCR-RFLP assay successfully identified C. jejuni, C. coli and C. fetus (Table 2). The sensitivity of the assay was 100 % and 83 % for clinical and animal samples, respectively, and the specificity was 100 % in both cases. Furthermore, Campylobacter-like organisms isolated from the digestive tracts of two cattle samples were not identified by the cdtB gene-based multiplex PCR (Asakura et al., 2008), but the PCR-RFLP assay successfully identified these two bacteria as C. hyointestinalis (data not shown). These data indicate that the PCR-RFLP assay developed in this study is applicable not only to bacterial isolates but also to samples of human and animal origin. However, the number of strains and samples tested in this study is not sufficient and, therefore, further studies are needed to evaluate the distribution of cdt genes in the four Campylobacter species (C. hyointestinalis, C. lari, C. helveticus and C. upsaliensis) and the utility of the PCR-RFLP assay with more human and animal samples. The PCR-RFLP assay developed in this study was useful in the examination of clinical and animal samples for the presence of seven Campylobacter species, but the detection limit of C. hyointestinalis was lower than that of other campylobacters. The lower detection limit of C. hyointestinalis may be affected by some mutations of primer binding sites because mismatch annealing is found to have the strongest influence on amplification efficiency. One and four mismatches were found in C-CdtBcom1 and CCdtBcom2 primers, respectively, in comparison with the cdtB gene of the C. hyointestinalis reference strain (data not shown). The detection limit may be improved by further modification of common primers. In conclusion, the cdtB gene-based simple PCR-RFLP assay developed in this study will be useful for the detection and differentiation of seven different Campylobacter species that are involved in human and animal diseases. As mentioned, 665

K. Kamei and others

some of the species which are really difficult to isolate by applying routinely used selective media in a clinical laboratory were also easily detected and differentiated by the PCR-RFLP assay developed in this study. Thus, cdt genes may serve as good targets to detect and differentiate different important campylobacters involved in human and animal diseases at the species level.

Kabir, S. M., Kikuchi, K., Asakura, M., Shiramaru, S., Tsuruoka, N., Goto, A., Hinenoya, A. & Yamasaki, S. (2011). Evaluation of a

cytolethal distending toxin (cdt) gene-based species-specific multiplex PCR assay for the identification of Campylobacter strains isolated from diarrheal patients in Japan. Jpn J Infect Dis 64, 19–27. Lastovica, A. J. & le Roux, E. (2000). Efficient isolation of

campylobacteria from stools. J Clin Microbiol 38, 2798–2799. S. M. (2011). The clinical importance of emerging Campylobacter species. Nat Rev Gastroenterol Hepatol 8, 669–685. Man,

ACKNOWLEDGEMENTS

Matsuda, M., Shigematsu, M., Tazumi, A., Sekizuka, T., Takamiya, S., Millar, B. C., Taneike, I. & Moore, J. E. (2008). Cloning and structural

We thank Drs R. K. Bhadra (CSIR-Indian Institute of Chemical Biology, India) and S. B. Neogi (International Centre for Diarrhoeal Diseases Research, Bangladesh) for critical reading of the manuscript. This study was performed in partial fulfilment of the requirements of a PhD thesis for K. K. from the Graduate School of Pharmaceutical Sciences, Osaka University, Osaka, Japan. This work was supported by a Grant-in-Aid (B) of the Ministry of Education, Culture, Sports, Science and Technology of Japan.

analysis of the full-length cytolethal distending toxin (cdt) gene operon from Campylobacter lari. Br J Biomed Sci 65, 195–199. Noormohamed, A. & Fakhr, M. K. (2012). Incidence and antimicro-

bial resistance profiling of Campylobacter in retail chicken livers and gizzards. Foodborne Pathog Dis 9, 617–624. Parkhill, J., Wren, B. W., Mungall, K., Ketley, J. M., Churcher, C., Basham, D., Chillingworth, T., Davies, R. M., Feltwell, T. & other authors (2000).

The genome sequence of the food-borne pathogen Campylobacter jejuni reveals hypervariable sequences. Nature 403, 665–668.

REFERENCES

Pickett, C. L., Pesci, E. C., Cottle, D. L., Russell, G., Erdem, A. N. & Zeytin, H. (1996). Prevalence of cytolethal distending toxin produc-

Asakura, M., Samosornsuk, W., Taguchi, M., Kobayashi, K., Misawa, N., Kusumoto, M., Nishimura, K., Matsuhisa, A. & Yamasaki, S. (2007). Comparative analysis of cytolethal distending toxin (cdt)

tion in Campylobacter jejuni and relatedness of Campylobacter sp. cdtB gene. Infect Immun 64, 2070–2078.

genes among Campylobacter jejuni, C. coli and C. fetus strains. Microb Pathog 42, 174–183.

Prouzet-Maule´on, V., Labadi, L., Bouges, N., Me´nard, A. & Me´graud, F. (2006). Arcobacter butzleri: underestimated enteropathogen. Emerg

Infect Dis 12, 307–309.

Asakura, M., Samosornsuk, W., Hinenoya, A., Misawa, N., Nishimura, K., Matsuhisa, A. & Yamasaki, S. (2008). Development

Samosornsuk, W., Asakura, M., Yoshida, E., Taguchi, T., Nishimura, K., Eampokalap, B., Phongsisay, V., Chaicumpa, W. & Yamasaki, S. (2007).

of a cytolethal distending toxin (cdt) gene-based species-specific multiplex PCR assay for the detection and identification of Campylobacter jejuni, Campylobacter coli and Campylobacter fetus. FEMS Immunol Med Microbiol 52, 260–266.

Evaluation of a cytolethal distending toxin (cdt) gene-based speciesspecific multiplex PCR assay for the identification of Campylobacter strains isolated from poultry in Thailand. Microbiol Immunol 51, 909–917.

Bullman, S., Corcoran, D., O’Leary, J., O’Hare, D., Lucey, B. & Sleator, R. D. (2011). Emerging dynamics of human campylobacter-

iosis in Southern Ireland. FEMS Immunol Med Microbiol 63, 248–253. Carbonero, A., Torralbo, A., Borge, C., Garcı´a-Bocanegra, I., Arenas, A. & Perea, A. (2012). Campylobacter spp., C. jejuni and C. upsaliensis

infection-associated factors in healthy and ill dogs from clinics in Cordoba, Spain. Screening tests for antimicrobial susceptibility. Comp Immunol Microbiol Infect Dis 35, 505–512. Fouts, D. E., Mongodin, E. F., Mandrell, R. E., Miller, W. G., Rasko, D. A., Ravel, J., Brinkac, L. M., DeBoy, R. T., Parker, C. T. & other authors (2005). Major structural differences and novel potential

virulence mechanisms from the genomes of multiple Campylobacter species. PLoS Biol 3, e15. Gharst, G., Oyarzabal, O. A. & Hussain, S. K. (2013). Review of

current methodologies to isolate and identify Campylobacter spp. from foods. J Microbiol Methods 95, 84–92. Hald, B., Pedersen, K., Wainø, M., Jørgensen, J. C. & Madsen, M. (2004). Longitudinal study of the excretion patterns of thermophilic

Campylobacter spp. in young pet dogs in Denmark. J Clin Microbiol 42, 2003–2012. Hoshino, K., Yamasaki, S., Mukhopadhyay, A. K., Chakraborty, S., Basu, A., Bhattacharya, S. K., Nair, G. B., Shimada, T. & Takeda, Y. (1998). Development and evaluation of a multiplex PCR assay for

Sandstedt, K. & Ursing, J. (1991). Description of Campylobacter upsaliensis sp. nov. previously known as the CNW group. Syst Appl Microbiol 14, 39–45. Shiramaru, S., Asakura, M., Inoue, H., Nagita, A., Matsuhisa, A. & Yamasaki, S. (2012). A cytolethal distending toxin gene-based multiplex

PCR assay for detection of Campylobacter spp. in stool specimens and comparison with culture method. J Vet Med Sci 74, 857–862. Sivadon-Tardy, V., Porcher, R., Orlikowski, D., Ronco, E., Gault, E., Roussi, J., Durand, M. C., Sharshar, T., Annane, D. & other authors (2013). Increased incidence of Campylobacter jejuni-associated Guillain-

Barre´ syndromes in the Greater Paris area. Epidemiol Infect 10, 1–5. Stanley, J., Burnens, A. P., Linton, D., On, S. L., Costas, M. & Owen, R. J. (1992). Campylobacter helveticus sp. nov., a new thermophilic

species from domestic animals: characterization, and cloning of a species-specific DNA probe. J Gen Microbiol 138, 2293–2303. Steinhauserova, I., Fojtikova, K. & Klimes, J. (2000). The incidence

and PCR detection of Campylobacter upsaliensis in dogs and cats. Lett Appl Microbiol 31, 209–212. Vandenberg, O., Dediste, A., Houf, K., Ibekwem, S., Souayah, H., Cadranel, S., Douat, N., Zissis, G., Butzler, J. P. & Vandamme, P. (2004). Arcobacter species in humans. Emerg Infect Dis 10, 1863–1867. Yamasaki, S., Asakura, M., Tsukamoto, T., Faruque, S. M., Deb, R. & Ramamurthy, T. (2006). Cytolethal distending toxin (CDT): genetic

rapid detection of toxigenic Vibrio cholerae O1 and O139. FEMS Immunol Med Microbiol 20, 201–207.

diversity, structure and role in diarrheal disease. Toxin Reviews 25, 61–88.

Inglis, G. D., Boras, V. F. & Houde, A. (2011). Enteric campylobacteria

Young, K. T., Davis, L. M. & Dirita, V. J. (2007). Campylobacter jejuni:

and RNA viruses associated with healthy and diarrheic humans in the Chinook health region of southwestern Alberta, Canada. J Clin Microbiol 49, 209–219.

Yuki, N. & Hartung, H. P. (2012). Guillain-Barre´ syndrome. N Engl J

666

molecular biology and pathogenesis. Nat Rev Microbiol 5, 665–679. Med 366, 2294–2304. Journal of Medical Microbiology 63