Development and Application of a New Scheme for Typing ...

JOURNAL OF CLINICAL MICROBIOLOGY, May 2002, p. 1791–1797 0095-1137/02/$04.00⫹0 DOI: 10.1128/JCM.40.5.1791–1797.2002 Copyright © 2002, American Society for Microbiology. All Rights Reserved.

Vol. 40, No. 5

Development and Application of a New Scheme for Typing Campylobacter jejuni and Campylobacter coli by PCR-Based Restriction Fragment Length Polymorphism Analysis Feng Shi,1* Yuen Yuen Chen,1 Trudy M. Wassenaar,2 Walter H. Woods,3 Peter J. Coloe,1 and Benjamin N. Fry1 Department of Biotechnology and Environmental Biology, Royal Melbourne Institute of Technology University, Melbourne, Victoria 3083,1 and School of Pharmacy and Medical Sciences, University of South Australia, Adelaide, South Australia, 5001,3 Australia, and Molecular Microbiology and Genomics Consultants, Zotzenheim, Germany2 Received 28 November 2001/Returned for modification 10 January 2002/Accepted 3 March 2002

A molecular typing approach for Campylobacter jejuni and Campylobacter coli was developed with restriction fragment length polymorphism analysis of a 9.6-kb PCR-amplified portion of the lipopolysaccharide gene cluster. Sixty-one Penner serotype reference strains were analyzed with this new genotyping scheme, and 32 genogroups were found. Eleven additional genogroups were obtained from 87 clinical C. jejuni strains tested. This molecular typing method shows a correlation with the Penner heat-stable serotyping method, a phenotypic typing method based on lipopolysaccharide structures that is often used as a “gold standard” for subtyping Campylobacter spp. This strong correlation suggests that the data obtained can be directly compared with epidemiological data collected in the past by classical serotyping of C. jejuni and C. coli. In contrast to the high percentage of nontypeability by phenotyping, this molecular typing method results in 100% typeability and provides a superior alternative to serotyping. Campylobacter jejuni and Campylobacter coli are recognized as two of the most common causes of food-borne bacterial gastroenteritis. Furthermore, C. jejuni has been implicated as a frequent antecedent to the development of the neurologic diseases Guillain-Barré syndrome (GBS) (20) and Miller Fisher syndrome (37). Numerous subtyping methods have been developed to differentiate Campylobacter strains for epidemiologic purposes in the past two decades. More than 30 current typing methods have been reviewed elsewhere (27, 28, 29, 38). The various typing systems can be placed in two categories: phenotypic methods, which are based on expressed features such as somatic antigens or enzymatic activity, and genotypic methods, which are based on specific molecular features of chromosomal or plasmid DNA. Two serotyping schemes have been used exclusively for phenotypic typing in the past, the scheme developed by Penner and Hennessy, which detects heat-stable (HS) antigens (31), and the one developed by Lior et al., which detects heat-labile antigens (16). The former is the most widely accepted and well evaluated phenotypic method. The molecular basis for the HS antigenic diversity in C. jejuni and C. coli is the expression of somatic (O) lipopolysaccharide (LPS) (17, 18, 21, 22, 33, 34, 35). LPS is a major constituent of the outer membrane in gram-negative bacteria and comprises three covalently linked regions: lipid A, core oligosaccharide (inner core and outer core), and O polysaccharide. The variability of the Campylobacter LPS outer core and O polysaccharide is thought to

contribute to the antigenic basis of the Penner serotyping system. Serotyping methods like these are time consuming and technically demanding, and antisera are costly to produce, which limits the use of these typing systems to specialized diagnostic laboratories. Furthermore, phenotypes can be unstable, resulting in nonreproducible results or nontypeable strains as well as antiserum cross-reactivity, which hampers the interpretation. Genotyping methods are independent of expressed features and are therefore a better alternative for typing. Several genotyping methods have recently been developed, such as pulsed-field gel electrophoresis (11, 12, 40), amplified fragment length polymorphism (5, 15), flagella gene PCR-restriction fragment length polymorphism (PCR-RFLP) (1, 4, 23, 24, 26), ribotyping (6, 25, 36), and random amplified polymorphic DNA analysis (7, 10, 30). These genotyping systems are more generally available and applicable than phenotypic methods. However, most of these techniques still have their own drawbacks, such as less discriminatory power, poor reproducibility, and complex methodology. The preferred method in terms of handling, costs, and time, is RFLP analysis of PCR products. Such a method has been described for the flagellin genes (23, 26). However, when this method was applied, no correlation could be detected between flagellin genotypes and HS serotypes (23). This greatly reduces the application of flagellin genotyping in long-term epidemiological studies. More importantly, none of these methods correlate well with the serotyping scheme used in past decades, so historical epidemiological trends cannot be determined. For these reasons, genotypic subtyping methods have not been widely used in epidemiological practice and remain to be developed and improved. The LPS biosynthesis gene cluster of C. jejuni 81116 has recently been characterized in our laboratory (9). In this study,

* Corresponding author. Mailing address: Dept. of Biotechnol. and Environ. Biol., Royal Melbourne Inst. of Technol. Univ., Bundoora W. Campus, Bldg. 223, Lvl. 1, Plenty Rd., Bundoora 3083 VIC, Melbourne, Australia. Phone: 61-3-992-57132. Fax: 61-3-992-57110. Email: [email protected]. 1791

1792

SHI ET AL.

J. CLIN. MICROBIOL.

TABLE 1. Bacterial strains used in this study Organism and strain(s)

E. coli O111 S. enterica serovar Typhimurium C. lari C. hyoilei C. hyointestinalis C. sputorum C. fetus C. upsaliensis Campylobacter concisus C. jejuni 81116 C. jejuni clinical strains C. jejuni clinical strains C. jejuni clinical strains C. jejuni clinical O:19 strains C. jejuni clinical strains C. jejuni strains Penner serotype strains

Source(s) or reference

9 Melbourne, Australia, human Melbourne, Australia, human Melbourne, Australia, human Melbourne, Australia, human Melbourne, Australia, human Melbourne, Australia, human Melbourne, Australia, human Melbourne, Australia, human United Kingdom, human Adelaide, Australia, human Canberra, Australia, human Sydney, Australia, human GBS China, Mexico, and Japan, human South Africa, human South Africa, chicken and ostrich 32

No. of strains

1 1 1 1 1 1 1 1 1 1 67 6 1 6 6 1 61

a new genotyping scheme, the LG (LPS genes) genotyping system, which is based on PCR-RFLP of this gene cluster, has been established. The application value of this new system was also evaluated by typing the reference Penner serotype strains and a number of C. jejuni and C. coli clinical isolates. This typing scheme is the first genotyping scheme to our knowledge that has the same molecular basis as the Penner serotyping scheme. MATERIALS AND METHODS Bacterial strains and growth conditions. The strains used in this study are listed in Table 1. Cultures were stored at ⫺70°C in heart infusion broth (Difco) containing 50% glycerol. Bacteria were grown on Columbia agar (Oxoid) supplemented with 5% (vol/vol) defibrinated horse blood and Campylobacter selective supplement (Skirrow) (Oxoid) for 48 h at 42°C under microaerophilic conditions (CO2, 10.5%; O2, 5%; balance, N2). Chromosomal DNA preparation. DNA was isolated from pure cultures by the cetyltrimethylammonium procedure (3). Briefly, one lawn plate of Campylobacter was grown overnight, harvested in 9.4 ml of Tris-EDTA buffer and 0.1 ml of 0.5 M EDTA, and lysed with 0.5 ml of 10% (wt/vol) sodium dodecyl sulfate. Proteinase K was added to a final concentration of 0.1 mg/ml, and the mixture was incubated at 37°C for 2 h. Then 1.8 ml of 5 M NaCl and 1.5 ml of 10% (wt/vol) cetyltrimethylammonium in 0.7 M NaCl were added, and the mixture was incubated for 30 min at 65°C. After the addition of 5 ml of 24:1 chloroform-isoamyl alcohol, the mixture was centrifuged for 10 min at 4°C. The aqueous layer was removed, and 5 ml of 25:24:1 phenol-chloroform-isoamyl alcohol was added. The aqueous layer was removed after centrifugation. The DNA was precipitated with a 0.6-volume of isopropanol and dissolved in water. The concentration of the DNA was measured by spectrophotometric absorbance at 260 nm. Gene distribution analysis. The gene distribution of the wla cluster was analyzed by PCR. Primer sets used for amplification of the individual genes in the wla cluster are given in Table 2. PCR was performed in a 50-␮l reaction volume with a Perkin Elmer GeneAMP 2400 thermal cycler. The reaction mixtures consisted of 1⫻ reaction buffer with 2.0 mM MgSO4 (Promega), 2.5 U of Taq DNA polymerase (Promega), 100 ng of forward and reverse primer, 0.2 mM concentrations of deoxynucleoside triphosphates, and 100 ng of template DNA. The reaction included an initial denaturation of DNA at 94°C for 1 min and then 35 cycles of consecutive denaturation (30 s, 94°C), primer annealing (30 s, 60°C), and chain extension (based on a rate of 1 kb/minute, 72°C). A final elongation step was performed for 10 min at 72°C. Penner serotyping. Penner serotypes of Campylobacter spp. were detected by using the passive hemagglutination technique described by Penner and Hennessy (31), with a heated supernatant from the bacterial culture as the antigen. Antisera were prepared by injecting rabbits intravenously with saline bacterial suspensions of Campylobacter serostrains. Some antisera were absorbed with cross-

reacting HS reference strains to clarify the serotyping results; in particular, antiserum prepared against HS-4 was absorbed with HS-16 bacteria to remove minor reactions with HS-13 and HS-16. PCR-RFLP analysis. The primers galE1 and wlaH3 (Table 2) were used to amplify a 9.6-kb fragment. PCRs were performed as stated above except that a MgSO4 concentration of 2.0 mM was used with 2.5 U of Pfu DNA polymerase (Promega) instead of Taq and a chain extension of 15 min at 72°C per cycle. After amplification, 10 ␮l of PCR product was digested with 10 U of restriction enzymes (HhaI and NlaIII [New England Biolabs] and DdeI and HindIII [Promega]) in a total volume of 20 ␮l with 2 ␮g of bovine serum albumin for more than 3 h at 37°C. The digest was analyzed by electrophoresis by using a 1.5% agarose gel and stained with ethidium bromide. Lambda DNA digested with PstI was used as a reference size marker.

RESULTS Amplification analysis of LPS gene cluster. All 12 genes in the wla cluster were tested for their distribution among selected reference serotype C. jejuni strains. In order to detect duplications, gene order rearrangements, or deletions, PCR amplification was carried out for each single gene as well as across two genes that were found adjacent in the wla cluster of C. jejuni 81116. Primers used in the distribution analysis were designed from the published sequence of the LPS gene cluster

TABLE 2. Primers used for gene distribution analysis and PCR-RFLP Target gene (amplicon size [bp])

Primera

Positionb

Sequence

galE (911)

galE1 galE2

1185–1206 2077–9096

GCGGTGGTGCAGGTTATATAGG TAGGCTGCCAAGAAGTAAGA

wlaB (1,419)

wlaB1 wlaB2

2372–2391 3772–3791

TTGGAGTGGGGCTTATTGTT TGCGTTATCGTAGAAAGGCG

wlaC (599)

wlaC1 wlaC2

4167–4186 4747–4766

ACTCCACTCATCATTAGCGA TTCATCATCACAACCTACCA

wlaD (396)

wlaD1 wlaD2

5146–5165 5521–5542

TTAGATCAAGCAAGTGGCGA CTAAAGTCACGCTATCTTGCCT

wlaE (969)

wlaE1 wlaE2

5858–5877 6808–6827

TTTAGGAAGTGGTGGTGCTG ATCGTCTTCAAGCATAGTCT

wlaF (1,163)

wlaF1 wlaF2

7176–7195 8318–8339

TCCGCGCTTACTTATTGGCT CACATCGCTATAATAACGCACA

wlaG (824)

wlaG1 wlaG2

9186–9205 9991–10010

CCGCAAGATGAATACACGCA AGCCTCTACACAACCTTCAC

wlaH (232)

wlaH1 wlaH2 wlaH3

10388–10407 10601–10620 10783–10806

TGAGCGATGAAAGAGATGAG CGCCCATCCTGTTATACCTG TCAGTTCTTGCCATTAAATTTCTC

wlaI (245)

wlaI1 wlaI2

10820–10839 11046–11065

TATGGTGCTAGTGGTCATGG CACGCTTGCACTAGGACTTA

wlaK (631)

wlaK1 wlaK2

11868–11887 12478–12499

CATCTTTATGGCAATGCGGC ACTTCTTGAGTGTGCATAGCTT

wlaL (569)

wlaL1 wlaL2

13312–13331 13862–13881

GAAGAACTTGTAGCTTATGG CTCACAAACTCTCTTAGTGC

wlaM (738)

wlaM1 wlaM2

14507–14526 15226–15245

ACTTCCTGATATTCACCTTG GCTCACTCCACCGATAAGAT

a

All the primers were designed within the genes. Nucleotide positions are according to the National Center for Biotechnology Information published sequence number Y11648. b

VOL. 40, 2002

CAMPYLOBACTER PCR-RFLP TYPING SCHEME

1793

FIG. 1. (a) Sequence alignment analysis of part of the galE gene from C. jejuni 81116 (HS-6) (9), C. jejuni 11168gp (HS-2) (Sanger Centre website [ftp://ftp.sanger.ac.uk/pub/pathogens/cj/]), and C. jejuni HB9313 (HS-19) (unpublished data). (b) Sequence alignment analysis of part of the wlaH gene from C. jejuni 81116 (HS-6), C. jejuni 11351 (HS-23) (39), C. jejuni 11168gp (HS-2), C. jejuni 11168db (HS-2) (39), and C. jejuni HB9313 (HS-19). The bold characters indicate the primer sequences. The arrows indicate the orientation of the primers.

from C. jejuni 81116 (9) and listed in Table 2. The individual genes were amplified by primers designed within the gene; for example, the galE gene was analyzed by the primer pair galE1galE2. Two adjacent genes were analyzed by a primer combination of the upstream primer of the upstream gene and the downstream primer of the downstream gene; for example, the adjacent genes galE and wlaB were amplified by the primer pair galE1-wlaB2. All of the 12 genes were detected in the 12 C. jejuni strains investigated (data not shown), which indicated that the genes in the wla cluster were conserved among C. jejuni serotype reference strains. Of the 12 strains tested, 7 strains gave a PCR fragment of 2.3 kb with primers wlaI1 and wlaK2, which was compared to the 1.7-kb fragment obtained for the other strains (data not shown). This indicated that an extra fragment, possibly carrying a gene, is present between the wlaI and the wlaK genes in these strains. A gene named wlaJ was described at this position by Wood et al. (39) in 54% of C. jejuni isolates. The highly conserved nature of the wla gene cluster in C. jejuni makes it a suitable target for the development of a gene typing system. Of the 12 genes in the wla cluster, the galE and wlaH genes are most similar to genes found in other gram-negative bacteria, with ⬎52 and ⬎53% amino acid similarity, respectively (9). Therefore, it was reasoned that these genes would be the most

conserved within C. jejuni strains. Sequence alignment of the galE gene from three C. jejuni strains showed that, indeed, the galE gene is very conserved among C. jejuni strains (Fig. 1a). Also, the wlaH gene was found to be conserved when aligning this gene with five C. jejuni strains (Fig. 1b). The most conserved regions within the galE and the wlaH genes were used to design the primers galE1 and wlaH3 to amplify part of the LPS gene cluster (Fig. 1). These two primers gave a 9.6-kb product when used in a PCR, which spans eight genes in the wla cluster. All 61 reference Penner serotype Campylobacter strains were tested and gave the expected products (data not shown). The use of strongly conserved genes increases the risk that other species would be detected by the PCR. Therefore, we tested the cross-reactivity of the PCR primers on other bacteria within or outside the Campylobacter genus. No PCR products could be obtained from the DNA of Campylobacter lari, Campylobacter hyoilei, Campylobacter hyointestinalis, Campylobacter sputorum, Campylobacter fetus, and Campylobacter upsaliensis or from the DNA of the enterobacteria Escherichia coli and Salmonella enterica serovar Typhimurium. The result remained negative when the annealing temperature was lowered to 58°C (data not shown). This indicates that this typing scheme is specific for C. jejuni and C. coli.

1794

J. CLIN. MICROBIOL.

SHI ET AL.

FIG. 2. PCR-RFLP patterns of HhaI from part of reference HS serotype strains. Lane numbers in the figure indicate C. jejuni reference HS serotype strain numbers. Lane M, ␭ DNA digested with PstI.

RFLP within the amplified PCR product. The presence and degree of conservation of restriction enzyme recognition sites was evaluated for individual wla genes in a pilot study (data not shown). As a result of this inventory, the enzymes HindIII, HhaI, DdeI, and NlaIII were chosen to digest the obtained 9.6-kb amplified amplicon from all Penner serotype reference strains. This resulted in RFLP profiles of which typical examples are shown in Fig. 2. Eleven distinct HindIII patterns, 30 HhaI patterns, 25 DdeI patterns, and 19 NlaIII patterns were detected. The combination of the HhaI and DdeI patterns resulted in 32 separate genogroups (LG1 to LG32) for the 61 reference Penner serotype strains, and this enzyme combination was the method of choice for the LG genotyping scheme (Table 3). In most cases, the HhaI patterns will decide which LG genogroup an unknown strain belongs to. DdeI patterns need only be tested if the HhaI pattern is not unique, such as for the patterns Hh3, 10, 20, 26, and 33. All obtained profiles are available at the website http://www.bh.rmit.edu.au/abbt /campylobacter/typing.html. This new typing scheme was then tested on 87 clinical C. jejuni strains, among which 80 strains were HS serotyped. This resulted in an additional 11 new LG genogroups, LG33 to LG43 (Table 3). The results of the LG genotyping of the 80 HS serotyped clinical C. jejuni strains were compared with the Penner serotyping (Table 4). When two or more strains from the same Penner serogroup were typed, there was a correlation with the LG system, except for HS-4. DISCUSSION As for other infectious agents, subtyping methods are required for Campylobacter spp. to recognize temporal and regional trends and to identify and recognize pathotypes. The most common phenotypic subtyping scheme for C. jejuni is the

Penner serotyping system. This method was recently modified by Frost et al. (8) to improve the method and to make it less technically demanding. Despite this improvement, serotyping is still time consuming and costly because of the amount and numbers of the sera needed to be produced, maintained, and quality controlled. Although a number of genotyping methods have recently been established and are regarded as more advantageous than phenotypic methods due to better reliability, higher discriminatory power, better typeability, and general accessibility, serotyping methods still could not be replaced completely as they have been widely used for many years. The serological heterogeneity of C. jejuni strains is determined by variation in LPS structures for which the most likely molecular basis is DNA polymorphism in the genes encoding the LPS biosynthetic enzymes. Therefore, a molecular typing method has been developed based on the wla locus which combines the advantage of a simple and fast PCR-RFLP-based technique with a high typeability and a strong correlation to the HS serotyping scheme used in the past. The wla gene cluster, involved in LPS biosynthesis, is highly conserved in Campylobacter strains, and within this cluster, the galE and the wlaH genes show a high degree of similarity to genes in other gramnegative bacteria (9). Therefore, these genes are likely to be highly conserved among different C. jejuni strains and were used to design primers to amplify the intragenic region. Our results show that the method of LG genotyping is specific for C. jejuni and C. coli and has a higher discriminatory power than the Penner serotyping method (Simpson’s index is 0.872 for the LG typing method and 0.842 for the Penner typing method). Sixty-one reference strains representing different Penner serotypes of C. jejuni and C. coli and 87 clinical C. jejuni strains were included for LG analysis. All tested strains yielded a PCR product of consistent size (9.6 kb) that could be digested by the restriction enzymes HhaI and DdeI to achieve 100% typeability. The RFLP patterns have a reason-

CAMPYLOBACTER PCR-RFLP TYPING SCHEME

VOL. 40, 2002

1795

TABLE 3. LG genotyping systema LG group

HhaI pattern

DdeI pattern

Name

Observed with strain(s):

Name

LG1 LG2 LG3 LG4 LG5 LG6 LG7 LG8 LG9

Hh1 Hh2 Hh3 Hh4 Hh5 Hh6 Hh7 Hh8 Hh9

Dd1 Dd2 Dd3 Dd4 Dd5 Dd6 Dd7 Dd8 Dd9

LG10 LG11 LG12 LG13 LG14 LG15 LG16

Hh10 Hh11 Hh10 Hh12 Hh13 Hh14 Hh15

HS-1, 8 HS-2, 4, 13, 16 HS-3, 31 HS-5 HS-6, 9, 12, 21, 27, 38, 45, 55, 57, 60 HS-7 HS-10 HS-11, 20b HS-14,b 25,b 26,b 28,b 30,b 39,b 48,b 49,b 54,b 56,b 59,b 66b HS-15, 18, 33, 44 HS-17 HS-15, 18, 33, 44 HS-19 HS-22 HS-23 HS-24b

LG17 LG18 LG19 LG20

Hh16 Hh3 Hh17 Hh18

HS-29, 36 HS-3, 31 HS-32 HS-34b

Dd15 Dd16 Dd17 Dd9

LG21 LG22 LG23 LG24 LG25 LG26

Hh19 Hh20 Hh21 Hh22 Hh23 Hh24

HS-35 HS-37 HS-40, 41 HS-42 HS-43 HS-46b

Dd18 Dd19 Dd20 Dd21 Dd22 Dd9

LG27

Hh25

HS-47b

Dd9

LG28 LG29 LG30 LG31 LG32 LG33 LG34 LG35 LG36 LG37 LG38 LG39 LG40 LG41 LG42 LG43

Hh26 Hh27 Hh28 Hh29 Hh30 Hh10 Hh20 Hh26 Hh3 Hh31 Hh32 Hh33 Hh33 Hh34 Hh35 Hh36

HS-50 HS-51b HS-52 HS-53 HS-58 HS-15, 33, 44 HS-37 HS-50 HS-3, 31 514c 2506c 3499,c 109–24–613d 3499,c 109–24–613d 3783c 109–27–310d 104–54–427d

Dd2 Dd23 Dd20 Dd24 Dd25 Dd14 Dd16 Dd7 Dd27 Dd26 Dd25 Dd7 Dd29 Dd28 Dd27 Dd19

Dd10 Dd11 Dd12 Dd13 Dd14 Dd14 Dd9

Observed with strain(s):

HS-1, 8 HS-2, 4, 13, 16, 50 HS-3 HS-5 HS-6, 9, 12, 21, 27, 38, 45, 55, 57, 60 HS-7 HS-10 HS-11, 20b HS-14,b 24,b 25,b 26,b 28,b 30,b 34,b 39,b 47,b 48,b, 49,b 54,b 56,b 59,b 66b HS-15, 33, 44 HS-17 HS-18 HS-19 HS-22, 23 HS-22, 23 HS-14,b 24,b 25,b 26,b 28,b 30,b 34,b 39,b 47,b 48,b 49,b 54,b 56,b 59,b 66b HS-29, 36 HS-31 HS-32 HS-14,b 24,b 25,b 26,b 28,b 30,b 34,b 39,b 47,b 48,b 49,b 54,b 56,b 59,b 66b HS-35 HS-37 HS-40, 41, 52 HS-42 HS-43 HS-14,b 24,b 25,b 26,b 28,b 30,b 34,b 39,b 47,b 48,b 49,b 54,b 56,b 59,b 66b HS-14,b 24,b 25,b 26,b 28,b 30,b 34,b 39,b 47,b 48,b 49,b 54,b 56,b 59,b 66b HS-2, 4, 13, 16, 50 HS-51b HS-40, 41, 52 HS-53 HS-58 HS-22, 23 HS-31 HS-10 109–27–310d 514c HS-58c HS-10 109–24–613d 3783c 109–27–310d HS-37

46,b

46,b

46,b

46,b 46,b

a The LG groups with their denominated HhaI and DdeI patterns. Each pattern name is followed by a list of HS strains in which the pattern is observed. Bold Hh pattern names indicate nonunique patterns for which determination of the DdeI pattern is compulsive. b C. coli strain. c Clinical isolate from Adelaide, Australia. d Clinical isolate from Canberra, Australia.

able size range which can be separated by a 1.5% midi agarose gel. The patterns are very easy to read and can be digitized for comparison between laboratories. A total of 43 LG profile types were identified among the 61 reference HS serostrains of C. jejuni and C. coli and the 87 clinical C. jejuni strains. A comparison of the LG typing data of 80 C. jejuni strains with the Penner serotyping data showed, as expected, a strong association between these two typing methods. All strains belonging to Penner groups 3, 6, 9, 12, 19, 29, 31, and 45 shared their LG banding patterns with the respective Penner reference strains (Table 4). Notably, Penner group 19, which has been recognized to have a strong association with

GBS, is recognized as a separate LG genotype with conserved banding patterns for the 9 clinical isolates tested, originating from different sources. Strains belonging to the Penner groups 1 and 2 had identical LG patterns for 95 and 84% of the investigated strains, respectively. In contrast, the predominant LG genotype of strains belonging to the HS-4 complex serogroup was only found in 38% of the strains tested. Penner serogroups 1, 2, 3, and especially 4, which together comprise a large proportion of clinical isolates in certain parts of the world, are not clonal but represent a genetically diverse population (2, 13, 14, 19, 27, 32). In this respect, the strong correlation between LG genotype and HS-1, HS-2, and HS-3

1796

SHI ET AL.

J. CLIN. MICROBIOL. TABLE 4. Comparison of LG and HS typing methods

Penner (HS) serogroup

Predominant LG genogroup (% of strains)b

No. of strains tested

No. of strains that match predominant LG

LG serogroup found in minority of strains (no. of strains/total no. of strains)

HS-1 HS-2

LG1 (95) LG2 (84)

19 19

18 16

HS-3 HS-4

LG3 LG28 (38)

2 16

2 6a

LG33 (1/19) LG3 (1/19) LG35 (1/19) LG37 (1/19)

HS-5 HS-6 HS-7 HS-9 HS-12 HS-13 HS-17 HS-18 HS-19 HS-29 HS-31 HS-33 HS-42 HS-44 HS-45

LG33 LG5 LG39 LG5 LG5 LG28 LG41 LG17 LG13 LG17 LG36 LG5 LG5 LG1 LG5

1 2 1 1 1 1 1 1 9 1 1 1 1 1 1

0 2 0 1 1 0 0 0 9 1 0 0 0 0 1

a b

LG10 (5/16) LG34 (3/16) LG2 (1/16) LG38 (1/16)

Only for HS-4 does the predominant LG genogroup differ from that found with the serotype reference strain. Percentage given only if not 100 or 0%.

serotype is remarkable. The strains from the HS-4 group more often belonged to genotypes LG10, 28, or 34, than to LG2, which is the LG genotype for the HS-4 reference serotype strain. This indicates that the serotyping reference strain does not have the most common DNA polymorphism for this genetically diverse serogroup. In conclusion, we believe that this fast and simple method is a suitable alternative to serotyping as an application for typing clinical and food isolates of C. jejuni and as a method for facilitating epidemiological research of Campylobacter spp. The added value of the proposed LG typing scheme to the genotyping schemes already existing for C. jejuni is the strong correlation to the classical HS serotyping scheme, which allows comparison with historical data. In addition, LG typing allows for the classification of strains that are nontypeable by serotyping, since such strains are likely to be LG typeable. The observed discriminatory power of LG is slightly higher than that of serotyping, especially for serotype HS-4. In conclusion, we believe that this fast and simple method is a suitable alternative to serotyping and a valuable addition to the genotyping methods available for typing clinical and food isolates of C. jejuni and facilitating epidemiological research of Campylobacter spp. ACKNOWLEDGMENTS We thank Irving Nachamkin for providing some of the clinical C. jejuni strains for this study. We also gratefully acknowledge the cooperation of diagnostic laboratories throughout South Australia in providing isolates for typing. REFERENCES 1. Alm, R., P. Guerry, and T. J. Trust. 1993. Distribution and polymorphism of the flagellin genes from isolates of Campylobacter coli and Campylobacter jejuni. J. Bacteriol. 175:3051–3057.

2. Asrat, D. A., A. Hathaway, E. S. Sjögren, E. Ekwall, and B. Kaijser. 1997. The serotype distribution of Campylobacter jejuni and Campylobacter coli isolated from patients with diarrhea and controls in Tikur Anbassa Hospital, Addis Ababa, Ethiopia. Epidemiol. Infect. 118:222–226. 3. Ausubel, F. M., R. Brent, R. E. Kingston, D. M. Moore, J. G. Seidman, J. A. Smith, and K. Struhl (ed.). 1995. Current protocols in molecular biology. John Wiley & Sons, New York, N.Y. 4. Ayling, R. D., M. J. Woodward, S. Evans, and D. G. Newell. 1996. Restriction fragment length polymorphism of polymerase chain reaction products applied to the differentiation of poultry campylobacters for epidemiological investigations. Res. Vet. Sci. 60:168–172. 5. Duim, B., T. M. Wassenaar, A. Rigter, and J. Wagenaar. 1999. High-resolution genotyping of Campylobacter strains isolated from poultry and humans with amplified fragment length polymorphism fingerprinting. Appl. Environ. Microbiol. 65:2369–2375. 6. Fayos, A., R. J. Owen, M. Desai, and J. Hernandez. 1992. Ribosomal RNA gene restriction fragment diversity amongst Lior biotypes and Penner serotypes of Campylobacter jejuni and Campylobacter coli. FEMS Microbiol. Lett. 95:87–94. 7. Fayos, A., R. J. Owen, J. Hernandez, C. Jones, and A. Lastovica. 1993. Molecular subtyping by genome and plasmid analysis of Campylobacter jejuni serogroups O1 and O2 (Penner) from sporadic and outbreak cases of human diarrhoea. Epidemiol. Infect. 111:415–427. 8. Frost, J. A., A. N. Oza, R. T. Thwaites, and B. Rowe. 1998. Serotyping scheme for Campylobacter jejuni and Campylobacter coli based on direct agglutination of heat-stable antigen. J. Clin. Microbiol. 36:335–339. 9. Fry, B. N., V. Korolik, J. A. ten Brinke, M. T. T. Pennings, R. Zalm, B. J. J. Teunis, P. J. Coloe and B. A. M. van der Zeijst. 1998. The lipopolysaccharide biosynthesis locus of Campylobacter jejuni 81116. Microbiol. 144:2049–2061. 10. Fujimoto, S., B. M. Allos, N. Misawa, C. M. Patton, and M. J. Blaser. 1997. Restriction fragment length polymorphism analysis and random amplified polymorphic DNA analysis of Campylobacter jejuni strains isolated from patients with Guillain-Barré syndrome. J. Infect. Dis. 176:1105–1108. 11. Gibson, J. R., C. Fitzgerland, and J. Owen. 1995. Comparison of PFGE, ribotyping and phage-typing in the epidemiological analysis of Campylobacter jejuni serotype HS2 infections. Epidemiol. Infect. 115:215–225. 12. Imai, Y., M. Kikuchi, M. Matsuda, M. Honda, M. Fukuyama, M. Tsukada, and C. Kaneuchi. 1994. Macro-fingerprinting analysis at the chromosomal genomic DNA level of isolates of thermophilic Campylobacter coli and Campylobacter jejuni, by pulsed-field gel electrophoresis. Cytobios 78:115– 122. 13. Jones, D. M., J. D. Abbott, M. J. Painter, and E. M. Sutcliffe. 1984. A comparison of biotypes and serotypes of Campylobacter sp. isolated from patients with enteritis and from animal and environmental sources. J. Infect. 9:51–58.

VOL. 40, 2002 14. Karmali, M. A., J. L. Penner, P. C. Fleming, A. Williams, and J. N. Hennessy. 1983. The serotype and biotype distribution of clinical isolates of Campylobacter jejuni and Campylobacter coli over a three-year period. J. Infect. Dis. 147:243–246. 15. Kokotovic, B., and S. L. W. On. 1999. High-resolution genomic fingerprinting of Campylobacter jejuni and Campylobacter coli by analysis of amplified fragment length polymorphisms. FEMS Microbiol. Lett. 173:77–84. 16. Lior, H., D. L. Woodward, J. A. Edgar, L. J. Laroche, and P. Gill. 1982. Serotyping of Campylobacter jejuni by slide agglutination based on heatlabile antigenic factors. J. Clin. Microbiol. 15:761–768. 17. Mandatori, R., and J. L. Penner. 1989. Structural and antigenic properties of Campylobacter coli. Infect. Immun. 57:3506–3511. 18. Mills, S. D., W. C. Bradbury, and J. L. Penner. 1985. Basis for serological heterogeneity of thermostable antigens of Campylobacter jejuni. Infect. Immun. 50:284–291. 19. Mills, S. D., G. O. Aspinall, A. G. McDonald, T. S. Raju, L. A. Kurjanczyk, and J. L. Penner. 1992. Lipopolysaccharide antigens of Campylobacter jejuni, p. 223–229. In I. Nachamkin, M. J. Blaser, and L. S. Tompkins. (ed.), Campylobacter jejuni: current status and future trends. American Society for Microbiology, Washington, D.C. 20. Mishu, B., and M. J. Blaser. 1993. Role of infection due to Campylobacter jejuni in the initiation of Guillain-Barré syndrome. Clin. Infect. Dis. 17:104– 108. 21. Moran, A. P., and T. U. Kosunen. 1989. Serological analysis of the heatstable antigens involved in serotyping Campylobacter jejuni and Campylobacter coli. APMIS 97:253–260. 22. Moran, A. P., and J. L. Penner. 1999. Serotyping of Campylobacter jejuni based on heat-stable antigens: relevance, molecular basis and implications in pathogenesis. J. Appl. Microbiol. 86:361–377. 23. Nachamkin, I., K. Bohachick, and C. M. Patton. 1993. Flagellin gene typing of Campylobacter jejuni by restriction fragment length polymorphism analysis. J. Clin. Microbiol. 31:1531–1536. 24. Nachamkin, I., H. Ung, and C. M. Patton. 1996. Analysis of HL and O serotypes of Campylobacter strains by the flagellin gene typing system. J. Clin. Microbiol. 34:277–281. 25. Owen, R. J., M. Desai, and S. Garcia. 1993. Molecular typing of thermotolerant species of Campylobacter with ribosomal RNA gene patterns. Res. Microbiol. 144:709–720. 26. Owen, R. J., A. Fayos, J. Hernandez, and A. Lastovica. 1993. PCR-based restriction fragment length polymorphism analysis of DNA sequence diversity of flagellin genes of Campylobacter jejuni and allied species. Mol. Cell. Probes 7:471–480. 27. Owen, R. J., and J. R. Gibson. 1995. Update on epidemiological typing of Campylobacter. PHLS Microbiol. Dig. 12:2–6.

CAMPYLOBACTER PCR-RFLP TYPING SCHEME

1797

28. Patton, C. M., I. K. Wachsmuth, G. M. Evins, J. A. Kiehlbauch, B. D. Plikaytis, N. Troup, L. Tompkins, and H. Lior. 1991. Evaluation of 10 methods to distinguish epidemic-associated Campylobacter strains. J. Clin. Microbiol. 29:680–688. 29. Patton, C. M., and I. K. Wachsmuth. 1992. Typing schemes—are current methods useful?, p. 110–128. In I. Nachamkin, M. J. Blaser, and L. S. Tompkins (ed.), Campylobacter jejuni: current status and future trends. American Society for Microbiology, Washington, D.C. 30. Payne, R. E., M. D. Lee, D. W. Dreesen, and H. M. Barnhart. 1999. Molecular epidemiology of Campylobacter jejuni in Broiler flocks using randomly amplified polymorphic DNA-PCR and 23s rRNA-PCR and role of litter in its transmission. Appl. Environ. Microbiol. 65:260–263. 31. Penner, J. L., N. Hennessy. 1980. Passive hemagglutination technique for serotyping Campylobacter fetus subsp. jejuni on the basis of soluble heatstable antigens. J. Clin. Microbiol. 12:732–737. 32. Penner, J. L., J. N. Hennessy, and R. V. Congi. 1983. Serotyping of Campylobacter jejuni and Campylobacter coli on basis of thermostable antigens. Eur. J. Clin. Microbiol. 2:378–383. 33. Perez Perez, G. I., and M. J. Blaser. 1985. Lipopolysaccharide characteristics of pathogenic campylobacters. Infect. Immun. 47:353–359. 34. Perez Perez, G. I., J. A. Hopkins, and M. J. Blaser. 1985. Antigenic heterogeneity of lipopolysaccharides from Campylobacter jejuni and Campylobacter fetus. Infect. Immun. 48:528–533. 35. Preston, M. A., and J. L. Penner. 1987. Structural and antigenic properties of lipopolysaccharides from serotype reference strains of Campylobacter jejuni. Infect. Immun. 55:1806–1812. 36. Russell, R. G., J. A. Kiehlbauch, J. I. Sarmiento, P. Panigrahi, D. C. Blake, and R. Haberbager. 1994. Ribosomal RNA patterns identify additional strains of Campylobacter jejuni and C. coli among isolates serotyped by heat-stable and heat-labile antigens. Lab. Anim. Sci. 44:579–583. 37. Salloway, S., L. A. Mermel, M. Seamans, G. O. Aspinall, J. E. Nam Shin, L. A. Kurjanczyk, and J. L. Penner. 1996. Miller-Fisher syndrome associated with Campylobacter jejuni bearing lipopolysaccharide molecules that mimic human ganglioside GD3. Infect. Immun. 64:2945–2949. 38. Wassenaar, T. M., and D. G. Newell. 2000. Genotyping of Campylobacter spp. Appl. Environ. Microbiol. 66:1–9. 39. Wood, A. C., N. J. Oldfield, C. A. O’Dwyer, and J. M. Ketley. 1999. Cloning, mutation and distribution of a putative lipopolysaccharide biosynthesis locus in Campylobacter jejuni. Microbiology 145:379–388. 40. Yan, W., N. Chang, and D. E. Taylor. 1991. Pulsed-field gel electrophoresis of Campylobacter jejuni and Campylobacter coli genomic DNA and its epidemiologic application. J. Infect. Dis. 163:1068–1072.