Phylogenetic Analysis and PCR-Restriction Fragment Length Polymorphism Identification of Campylobacter Species Based on Partial groEL Gene Sequences

Bacterial strains.The bacterial strains used in this study are listed in Table 1. Campylobacter strains isolated in Finland were characterized by microscopy (Gram stain, motility); for catalase and oxidase activity, hippurate and indoxyl acetate hydrolysis, nitrate reduction, and H₂S production in triple sugar iron agar; and with the API Campy system (bioMérieux sa, Marcy l'Etoile, France), as appropriate. C. helveticus and C. upsaliensis isolates were identified by species-specific PCR (19).

Human isolate 6871, tentatively identified as C. coli, was further characterized by hippuricase PCR (21); 16S rRNA gene sequencing with primers 27f, 518r, 536f, 1054r, 1073f, and 1492r (10, 26); 23S rRNA gene PCR-RFLP analysis (11); and dot blot hybridization (12). Hybridization was performed with genomic DNA isolated from C. jejuni reference strain NCTC 11168 and labeled with digoxigenin-11-dUTP (DIG-High Prime; Roche Diagnostics GmbH, Mannheim, Germany) as the probe. Genomic DNAs from C. coli CCUG 11283 and C. jejuni NCTC 11168 were used as references for evaluation of the degree of hybridization. The hybridizations were carried out at 58°C.

Isolation of genomic DNA.Genomic DNA was isolated from cultures grown on brucella blood agar for 24 to 48 h at 37°C under microaerobic conditions, as described previously (12, 29). Alternatively, for testing of the rapid PCR-RFLP method, cell lysates were prepared by suspending a 10-μl loopful of growth in 500 μl of sterile distilled water in a microcentrifuge tube. The tubes were heated at 100°C for 10 min and subsequently cooled to 4°C. The tubes were centrifuged at 13,000 rpm (Biofuge 13, rotor 3757; Heraeus Sepatech GmbH, Osterode/Harz, Germany) for 5 min, and the supernatant was stored at −20°C.

PCR amplification and cloning of partial groEL gene.The partial (593-bp) groEL gene was amplified with the degenerate primers H60F (5′-GGN GAY GGN CAN CAN CAN GCN CAN GT-3′) and H60R (5′-TCN CCR AAN CCN GGN GCY TTN CAN GC-3′) (30) (see below for explanation of bases designated Y, N, D, or R). The amplified region corresponds to nucleotides 253 to 845 of the groEL gene in C. jejuni NCTC 11168 (GenBank accession no. AY044099 ). The 50-μl PCR mixture contained 200 ng of genomic DNA (or 5 μl of cell lysate), 100 μM each deoxynucleoside triphosphate, 2.5 U of Taq DNA polymerase (MBI Fermentas, Hanover, Md.), 10 mM Tris-HCl (pH 8.8 at 25°C), 50 mM KCl, 0.08% Nonidet P-40, 1.5 mM MgCl₂, and 2 μM each primer. The PCR thermal cycling conditions were as described earlier (13). PCR products of the expected sizes were purified from 2% NuSieve GTG low-melting-point agarose gels (BioWhittaker Molecular Applications, Rockland, Maine) with a QIAquick gel extraction kit (Qiagen GmbH, Hilden, Germany), cloned into pGEM-T Easy vector (Promega, Madison, Wis.), and transformed into competent JM109 Escherichia coli cells (Promega). The transformants were selected with ampicillin, and recombinants were selected by blue-white differentiation. Plasmids were isolated from several clones with a Qiagen Plasmid Mini kit. To check for the presence of the correct 593-bp insert, plasmids were digested with EcoRI and the restriction products were separated on 1% agarose gels.

DNA sequencing.Nucleotide sequencing of three clones per strain was performed by automated cycle sequencing with Big Dye terminators (ABI 377XL; PE Applied Biosystems, Foster City, Calif.) and primers RP (reverse primer) and UP (universal primer M13-20).

Sequence analysis.The partial groEL gene sequences determined in this study and those available in public databases (Table 1) were analyzed with Bionumerics software (version 3.5; Applied Maths BVBA, Sint-Martens-Latem, Belgium). Distances were calculated by using all 541 bases (with the omission primer sequences). After pairwise and multiple-sequence alignments, a phylogenetic tree was constructed by the neighbor-joining method (31). Distances were corrected for multiple base changes by the method of Jukes and Cantor (14). The topology of the tree was evaluated by 1,000 trials of bootstrap analysis.

The groEL nucleotide sequences were translated to the corresponding amino acid sequences with the Transeq program (EMBOSS [European Molecular Biology Open Software Suite]). The resulting sequences were aligned, and a neighbor-joining tree was calculated with the ClustalW program. The phylogenetic tree based on the amino acid sequences was drawn with the TreeView program (version 1.6.6; Division of Environmental and Evolutionary Biology, University of Glasgow [http://taxonomy.zoology.gla.ac.uk/rod/rod.html ]).

Design of new reverse PCR primer.A new degenerate reverse PCR primer was developed for the conserved protein sequence AEDIEGE. The nucleotide sequence of the new reverse primer H60R1 was 5′-C YTC NCC YTC DAT RTC YTC NGC-3′ (517-bp product, corresponding to nucleotides 253 to 769 of the groEL gene in C. jejuni NCTC 11168, where Y is C or T; N is G, A, T, or C; D is G, T, or A; and R is A or G). The new primer was developed to have a melting temperature closer to that of the forward primer and to increase the specificity of the PCR. The melting temperature, primer-dimer, and secondary structure formation were checked by using the Sigma-Genosys basic oligonucleotide calculator (http://www.sigma-genosys.co.uk/oligos/frameset.html ).

PCR-RFLP typing of Campylobacter isolates.On the basis of computational restriction fragment analysis of the partial groEL gene sequences with the Restriction Mapper program (version 3; http://www.restrictionmapper.org/ ), the AluI and ApoI restriction enzymes were chosen, as they were expected to yield sufficient fragment numbers of sufficient sizes to produce species-specific restriction profiles. Digestions were performed with 20 μl of the PCR products in a total volume of 25 μl with 5 U of AluI or 4 U of ApoI (New England Biolabs Inc., Beverly, Mass.). The resulting fragments were separated electrophoretically (90 V for 3 h) in 4% MetaPhor agarose gels in 1× Tris-acetate-EDTA buffer. The gels were stained with ethidium bromide and visualized under UV light. The resulting patterns were analyzed with BioNumerics software.

Nucleotide sequence accession numbers.The partial groEL and 16S rRNA gene sequences obtained in this study were deposited in GenBank. The accession numbers are shown in Table 1. For comparison, published groEL and 16S rRNA gene sequences were downloaded from GenBank (Table 1).

Phylogenetic analysis of the partial groEL gene sequences compared with 16S rRNA gene sequences.The neighbor-joining tree constructed from the partial groEL gene sequences is shown in Fig. 1. The major topology of the tree based on the partial groEL gene sequences was similar to that based on the 16S rRNA gene sequences. The more detailed similarity analysis of the partial groEL and 16S rRNA gene sequences among Campylobacter species, A. butzleri, and Helicobacter pylori is shown in Table 2. The similarities of the groEL sequence of Campylobacter species to those of A. butzleri strain CCUG 10373 and H. pylori strain 26695 were in the range of 51 to 73% and 56 to 67%, respectively, whereas the 16S rRNA gene sequence similarities were 83 to 85% and 81 to 83%, respectively (Table 2). The interspecies groEL sequence similarities between Campylobacter species ranged from 65% (between C. fetus subsp. fetus and C. rectus) to 94% (between C. jejuni and C. coli) and 100% (between C. hyoilei and C. coli) (Fig. 1). C. upsaliensis and C. helveticus formed a separate branch with high bootstrap support, as did C. coli and C. jejuni as well as C. lanienae, C. fetus subsp. fetus, and both C. hyointestinalis subspecies (Fig. 1). The partial groEL gene sequences of the type strains of C. upsaliensis and C. helveticus showed 91% similarity, whereas the 16S rRNA gene sequences showed 98% similarity (Table 2). The 16S rRNA gene sequence similarity between C. helveticus and C. jejuni was high (97%), but the partial groEL gene similarity was only 84%. The groEL sequence similarity of C. coli and C. jejuni was 91%, whereas the 16S rRNA gene sequence similarity was 98%. C. lari showed 86 and 87% groEL gene sequence similarity to C. jejuni and C. coli, respectively, whereas the 16S rRNA gene sequence similarities were 98 and 97%, respectively. Similarly, the similarity of the C. lanienae groEL gene sequence to those of C. hyointestinalis subsp. hyointestinalis and C. hyointestinalis subsp. lawsonii was 81%, whereas the 16S rRNA gene sequence similarities were 97 and 98%, respectively. The groEL sequence similarity between C. fetus and C. hyointestinalis subsp. hyointestinalis and C. hyointestinalis subsp. lawsonii was 85%.

• Open in new tab
• Download powerpoint

FIG. 1.

Neighbor-joining tree based on partial 541-bp groEL sequences. The tree was rooted with H. pylori. The number at each branch point represents the percentage of bootstrap support calculated from 1,000 trees. The scale bar represents the sequence divergence.

Strain 6871 had a higher level of groEL sequence similarity to the C. jejuni strains (94.6 to 96%) than to the C. coli strains (92.6 to 93.6%). The strain was negative by the hippurate hydrolysis test and hippuricase gene PCR. The partial (1,470-bp) 16S rRNA gene sequence of strain 6871 showed the highest degree of similarity to C. jejuni strains ATCC 43431 (99.9%) and NCTC 11168 (99.7%) in a search with the BLAST algorithm. The level of sequence similarity to the type strain of C. coli CCUG 11283 was, however, also high (98.0%). PCR-RFLP analysis of the 23S rRNA gene resulted in a C. jejuni-specific profile for strain 6871. Dot blot hybridization of the whole genomic DNA of strain 6871 gave a more intense signal with C. jejuni than with the C. coli reference strain.

The intraspecies sequence similarities for C. coli ranged from 99 to 100%, those for C. jejuni ranged from 95% (98% when the hippurate-negative C. jejuni strain 6871 was excluded) to 100%, and those for C. lari ranged from 99 to 100%. The intraspecies sequence similarities between the two strains of C. helveticus and C. upsaliensis were 99.4 and 97.6%. The sequence similarity between C. hyointestinalis subsp. hyointestinalis and C. hyointestinalis subsp. lawsonii was 97.9%. The sequence similarity between C. sputorum biovar sputorum and C. sputorum subsp. bubulus was 99.8%.

The neighbor-joining tree based on the deduced partial GroEL amino acid sequences (180 amino acids) is shown in Fig. 2. Hippurate-negative C. jejuni strain 6871 had 100% amino acid similarity to C. jejuni strain 4117. The amino acid similarities between C. upsaliensis (two strains) and C. hyointestinalis subsp. hyointestinalis and C. hyointestinalis subsp. lawsonii were 100%. The intraspecies amino acid similarities of the other species were equal to the nucleotide sequence similarities, whereas the interspecies amino acid similarities were higher than the nucleotide sequence similarities, ranging from 84% (between C. hyointestinalis and C. lari) to 99% (between C. jejuni and C. coli) and 100% (between C. hyoilei and C. coli). A. butzleri strain CCUG 10373 and H. pylori strain 26695 had GroEL amino acid sequence similarities to the Campylobacter species GroEL amino acid sequences that ranged from 70 to 75% and 77 to 81%, respectively.

• Open in new tab
• Download powerpoint

FIG. 2.

Neighbor-joining tree based on the deduced partial 180-amino-acid GroEL sequences. The tree was rooted with H. pylori. The scale bar represents the sequence divergence.

PCR-RFLP identification of Campylobacter species based on partial 517-bp groEL gene amplicons.The results of the computational restriction fragment length analysis with the AluI and ApoI enzymes are shown in Table 3. According to these results, all Campylobacter species studied had species-specific AluI digestion patterns and all but C. helveticus and C. mucosalis and C. rectus and H. pylori had species-specific ApoI digestion patterns. The AluI and ApoI digestion patterns of the 517-bp groEL gene amplicons of representative Campylobacter strains from 11 species and 3 subspecies or biovars, as well as reference strains of A. butzleri and H. pylori, are shown in Fig. 3. Table 1 lists the profiles obtained for the remaining strains.

• Open in new tab
• Download powerpoint

FIG. 3.

RFLP analysis of the partial 517-bp groEL gene amplicon digested with AluI and ApoI. The combined dendrogram was derived with BioNumerics software by using the unweighted pair group method with arithmetic averages and the Dice coefficient with 1% optimization and tolerance.

Visually, 19 and 10 separate profiles were obtained by AluI and ApoI digestion, respectively. AluI digests alone discriminated among all species included in the study. For C. coli, C. jejuni, C. lari, and C. upsaliensis, three, two, two, and four separate AluI restriction profiles, respectively, were observed. Hippurate-negative strain C. jejuni 6871 was the only strain to yield AluI restriction profile 8b. Two ApoI restriction patterns were observed for C. coli. The ApoI restriction patterns could not differentiate between C. helveticus and C. mucosalis, C. hyointestinalis and C. fetus, or the less common ApoI profile 6 of C. coli (observed in 6 of 21 cases) and C. jejuni.

An atypical restriction profile for C. coli isolate S161R from a pig was observed. After the subculture of separate colonies, this isolate was subsequently identified as a coculture of two C. coli strains, strains S161Ra and S161Rb, representing two separate AluI-ApoI profiles, 10a-8 and 10c-6, respectively.