HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Best, E L Articles by Bolton, F J Search for Related Content PubMed PubMed Citation Articles by Best, E L Articles by Bolton, F J Agricola Articles by Best, E L Articles by Bolton, F J

Real-time single-nucleotide polymorphism profiling using Taqman technology for rapid recognition of Campylobacter jejuni clonal complexes

E L Best¹, A J Fox², J A Frost¹ and F J Bolton²

¹Campylobacter and Helicobacter Reference Unit, Laboratory of Enteric Pathogens, Centre For Infections, Health Protection Agency, 61 Colindale Avenue, London NW9 5HT, UK ²Health Protection Agency North West Laboratory, Manchester Medical Microbiology Partnership, Manchester Royal Infirmary, Manchester, UK

Correspondence E. L. Best emma.best{at}hpa.org.uk

Received December 2, 2004
Accepted June 22, 2005

The rapid identification of Campylobacter jejuni isolates to strain level would significantly inform the public health investigation of C. jejuni infection. Conceptual advances provided by multilocus sequence typing (MLST) have established the clonal complex as an important epidemiological group at the strain level, enabling accurate and phylogenetically valid strain identification for C. jejuni. The development of real-time PCR assays for allelic discrimination of strain-associated single-nucleotide polymorphisms (SNPs) based upon MLST locus alleles offers one possible approach for rapid strain detection. SNPs defining key alleles diagnostic for the most prevalent clonal complexes were identified following a detailed analysis of the available MLST data. Real-time Taqman allelic discrimination assays designed to detect the SNPs specific for six major clonal complexes, ST-21, ST-45, ST-48, ST-61, ST-206 and ST-257, were developed, allowing the rapid detection of C. jejuni isolates and preliminary strain identification. This will provide an important complementary technique to sequence typing for rapid detection and strain characterization to inform in real-time the public health management and investigation of C. jejuni infections.

Abbreviations: C_T, threshold cycle; MLST, multilocus sequence typing; SNP, single-nucleotide polymorphism.

	INTRODUCTION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
The genotyping system known as multilocus sequence typing (MLST) (Maiden et al., 1998) has been developed for Campylobacter jejuni. It has the advantages of using a methodology that provides a discriminatory molecular profile, being reproducible with simple interpretation and providing data that are directly comparable between laboratories using the internet (Dingle et al., 2001; Taylor & Fisher, 2003; Schouls et al., 2003; Sails et al., 2003a; http://pubmlst.org/campylobacter). This has resulted in the recognition of major genetic lineages or clonal complexes in C. jejuni populations from human infections, and animal and environmental sources (Dingle et al., 2002; Manning et al., 2003), which has provided major conceptual advances in our understanding of the population biology of C. jejuni.

Currently, 24 C. jejuni clonal complexes have been described, with the ST-21 clonal complex being the largest, containing 24 % of (submitted) isolates, and 54 % of all (submitted) isolates being assigned to one of the following six major complexes, ST-21, ST-45, ST-48, ST-61, ST-206 and ST-257 (http://pubmlst.org/campylobacter). The data from preliminary MLST studies of C. jejuni isolates from animals and human cases of infection have established the concept of host associations between clonal complexes of C. jejuni associated with poultry or cattle, which also cause human infections (Dingle et al., 2002). For example, clonal complexes ST-45 and ST-257 are reported to contain isolates predominantly of poultry and human origin, while ST-61 has been associated with bovine, ovine and human isolates (Colles et al., 2003; Dingle et al., 2002; Manning et al., 2003).

C. jejuni has a complex biology in terms of its genetic diversity and high recombination rate (Wassenaar & Newell 2000); however, the members of a lineage (clonal complex) share common alleles in which conserved single-nucleotide polymorphisms (SNPs) are identifiable. These can be representative of the allelic profile and therefore can be used as a shortcut for strain profiling, enabling preliminary identification of the MLST clonal complex. This provides a potential supplementary strategy for rapid identification of isolates belonging to the specific clonal complexes associated with human infection, but not as a replacement method for MLST. The ability to rapidly identify C. jejuni strain types, either directly from a variety of specimens or following short enrichment incubations, to clonal complex levels that have established host associations has major implications for tracing sources of infection and consequently improved public health responses. Furthermore, clonal complex identification opens the possibility of targeted sequencing of polymorphic loci of clonal complex sequence types for rapid case cluster recognition.

Previous findings demonstrated the applicability of the strategy based upon detection of one clonal complex (Best et al., 2004). Based upon this intelligence, the aims of this study were to identify informative SNPs from the common MLST alleles of the major clonal complexes, develop real-time Taqman allelic discrimination assays for six major clonal complexes associated with human infections and verify their usefulness in a strategy for profiling 236 isolates.

	METHODS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Strain selection and preparation.

For specificity studies other Campylobacter species and related organisms were used, including Campylobacter coli (NCTC 11353), Campylobacter sputorum biovar fecalis (NCTC 11415), Campylobacter lari (NCTC 11352), Campylobacter concisus (NCTC 11485), Campylobacter fetus (NCTC 10842), Campylobacter hyointestinalis (NCTC 11608), Campylobacter helveticus (NCTC 12470), Campylobacter curvus (NCTC 11649), Campylobacter sputorum subsp. mucosalis (NCTC 11000), Campylobacter sputorum subsp. bubulus (NCTC 11367), Campylobacter sputorum (NCTC 11528), Campylobacter upsaliensis (NCTC 11541), Escherichia coli (NCTC 09001), Salmonella Typhi (ATCC 4543), Arcobacter butzleri (NCTC 12481), Arcobacter cryaerophilus (NCTC 11885), Arcobacter skirrowii (NCTC 12713), Arcobacter nitrofigilis (NCTC 12251), Helicobacter pylori (NCTC 11637), Helicobacter pullorum (C77339) and Helicobacter canadensis (C78661). Two hundred and twenty-one C. jejuni isolates received by the Health Protection Agency (London) that had also been sequence typed by MLST (Best et al., 2004) were characterized for SNP profiles along with the 15 founder strain reference isolates for C. jejuni MLST as described by Wareing et al. (2003).

C. jejuni isolates were inoculated onto Colombia blood agar containing 5 % defibrinated horse blood and incubated for 24 h at 37 °C in anaerobic jars under microaerobic conditions (5 % CO₂, 5 % O₂, 3 % H₂ and 87 % N₂). DNA was isolated using MagNApure (Roche) with the bacterial DNA isolation kit according to manufacturer's instructions.

MLST.

MLST was carried out as described by Dingle et al. (2001) and sequenced products were separated and detected using an ABI Prism 3700 or a Beckman CEQ 8000 capillary sequencer. Contigs were assembled and edited by use of the Sequence Typing Analysis and Retrieval System (STARS; Man-Suen Chan and Nicki Ventress, University of Oxford), and allele numbers, sequence type and clonal complex were assigned by interrogation of the MLST website (http://pubmlst.org/campylobacter).

Identification of alleles for SNP assay design.

The most common alleles at each locus within every clonal complex were identified from the PubMLST database (http://pubmlst.org.uk/campylobacter). Using the ‘search database – advanced queries’ function with the ‘search by clonal complex’ box checked, allele distributions for each clonal complex were investigated. The alleles that were most specific for the six clonal complexes ST-21, ST-45, ST-48, ST-61, ST-206 and ST-257 were selected by high positive predicted value. This gave an indication of the occurrence of the allele within the clonal complex, where values closest to 1 were ‘predictive’ alleles, occurring most often within the clonal complex (Table 1). Two predictive alleles identified for each of the above clonal complexes were verified for exclusivity in combination for the clonal complex by using the ‘search database – advanced queries’ function with the ‘id’ fields set as the alleles and allele numbers entered.

Downloading of MLST alleles and identification of SNPs.

All locus alleles were downloaded from the C. jejuni website (http://pubmlst.org/campylobacter) into Bioedit Sequence Alignment Editor version 4.0.9 (Tom Hall, Dept of Microbiology, North Carolina State University). SNPs unique for each chosen allele and in a suitable location to meet the primer and probe design parameters (Primer Express Software Version 2, supplied by Applied Biosystems) were identified from the alignments.

Design and application of allelic discrimination assays.

Primers and probes were designed using Primer Express and were synthesized by Applied Biosystems. Each 25 µl PCR reaction consisted of 2.5 µl MagNApure DNA extract, 300 nM forward and reverse primers (Table 2), 100 nM Minor Groove Binding (MGB) probe (Applied Biosystems, 2002) and 1x Taqman universal mastermix, all added into Taqman 96-well reaction plates and sealed with adhesive covers. Taqman cycling consisted of 10 min at 50 °C and 10 min at 95 °C, followed by 40 cycles of 95 °C for 1 min and 60 °C for 1 min.

Data analysis for determination of SNPs.

PCR products were detected directly by monitoring the increase in fluorescence. Real-time Taqman analysis monitored the accumulation of the fluorescence throughout the 40 cycles, producing an amplification plot over the entire course of the reaction. Results were displayed as a C_T (threshold cycle) where the presence of the informative SNP was recognizable with a C_T number within the range 14–22 and a signal strength R_n > 1.

	RESULTS AND DISCUSSION

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Identification of predictive alleles

The aim of this study was to identify SNPs that could ultimately be used as a basis for the real-time identification of six C. jejuni clonal complexes. The strategy involved the identification of allele combinations predictive for clonal complexes by high positive predictive value so that all allele combinations used would enable identification of the clonal complex. For example, within clonal complex ST-257 the alleles glyA62 and pgm-4 had high positive predictive values (0.85 and 0.95, respectively) and were exclusive for this clonal complex (Table 1). In the case of clonal complexes ST-61 and ST-48 the allele glnA4 was used for initial determination of both clonal complexes, with the additional use of two further alleles uncA17 and uncA5 for ST-61 and ST-48, respectively. Likewise the allele tkt-1 was used for the determination of clonal complex ST-206 and also for the determination of clonal complex ST-21. Alternative alleles could have been used in addition to the ones described here; however, the alleles used were more easily distinguishable by the presence of particular SNPs and were more suitable for primer and probe design.

Informative SNP identification

Alignments in Bioedit identified SNPs that were capable of differentiating the predictive allele from the remaining alleles at the locus (Table 2). A strategy for identification of the six clonal complexes based on the presence of specific SNPs was developed with the following allele and SNP positions: clonal complex ST-21, glnA1 108, glnA1 267 and tkt-1 330; ST-45, gltA10 201, gltA10 225 and tkt-7 138, 141; ST-48, glnA4 18, glnA4 202 and uncA5 186, 189; ST-61, glnA4 18, glnA4 202 and uncA17 336; ST-206, tkt-1 330 (from ST-21) and glnA21 18, 33; and ST-257, glyA62 483 and pgm-4 165. Specific SNPs were identified from each of the alleles, for example allele uncA17, which was used for the detection of clonal complex ST-61, was particularly divergent, making the design of specific primers and probes relatively straightforward. This allele has been described as possibly originating from C. coli (Dingle et al., 2002).

Confirmation of specificity of SNP assays

The Taqman fluorogenic assays have been demonstrated to be reliable and accurate, with a low C_T value indicative of the presence of the SNP and results being obtained within 2 h. The results were consistent with the corresponding MLST profiles for these six clonal complexes. The specificity of the assays was confirmed with the C. jejuni founder reference strain set, other Campylobacter species and related organisms. Results showed that the assays were 100 % specific for C. jejuni and no positive results were shown for related or other species of Campylobacter (results not shown). Additionally when the MLST founder reference strains (Wareing et al., 2003) were tested, the assays correctly identified the reference isolates belonging to the specific clonal complexes ST-21, ST-45, ST-48, ST-61, ST-206 and ST-257, with C_T values within the specified range. No positive results were seen for the founder strain reference isolates assigned to other clonal complexes.

Large-scale strain profiling

Table 3 shows the results of the assays when applied to a larger collection of 221 isolates. The assays were shown to be specific and data obtained were consistent with the MLST data. The data showed the need for a minimum of two alleles in order to confirm specificity for these six clonal complexes; for example in the SNP assay for clonal complex ST-257, the first predictive allele, glyA62, was specific for the clonal complex, except for one isolate that was assigned to clonal complex ST-45 [ST-756, allelic profile 24 (asp), 7 (gln), 10 (glt), 62 (gly), 42 (pgm), 7 (tkt), 1 (unc)], but which also had the allele glyA62. In this isolate, however, the second ST-257 allele pgm-4 SNP was not present, as it had allele pgm-42. Similarly, in the assay for clonal complexes ST-61 and ST-48 some isolates assigned to other clonal complexes (ST-45, ST-257 and ST-206) possessed the allele glnA4. However, the inclusion of SNPs specific for ST-48 (uncA5 186, 189) or ST-61 (uncA17 336) meant isolates were assigned correctly. Conversely, isolates within clonal complexes ST-21, with sequence type 21 (allelic profile 2, 1, 1, 3, 2, 1, 5), and ST-206, with sequence type 206 (allelic profile 2, 21, 5, 37, 2, 1, 5), had the allele uncA5 but not the allele glnA4. Certain alleles used were highly specific for the clonal complex, especially in the cases of the allele uncA17, which was only present in clonal complex ST-61 isolates, the allele glnA21, which was only present within the ST-206-assigned isolates, and a small number of ST-403-assigned isolates due to the presence of the glnA27 allele.

Advantages of the SNP strategy

This strategy for the characterization of C. jejuni isolates has many important advantages for timely strain characterization and public health investigations. It is based upon existing nomenclature, whereby all results can be compared with those already in the MLST database without the complication of recharacterizing isolates. Therefore the strategy could be used to rapidly screen for detection or confirmation of the presence of particular C. jejuni clonal complexes, which may have important host associations, and then MLST or targeted sequence typing could be implemented where necessary. All the assays can be performed under the same run conditions and on the same Taqman plate, and therefore detection of the six clonal complexes can be achieved from DNA extracts within 2 h, compared to a total time of 24–36 h for MLST (including time for data analysis). Moreover, this approach has significant cost savings at a fifth of the price of in-house MLST for preliminary screening of isolates.

The real-time PCR approach offers an increased level of sensitivity that is not achievable with the conventional PCR as used in MLST. This provides the potential for direct detection in a variety of specimen types, such as faecal specimens, where low quantities of C. jejuni DNA might be present. This would be especially important in a case where no culture was obtainable but low quantities of DNA may be detectable by the SNP strategy. The MLST SNP assays are based upon the existing data within the PubMLST database (which is representative of isolates that have been sequence typed and submitted). However, as the database increases in size due to the addition of more isolates with new alleles, there is the potential for allelic divergence, which is not covered by the SNPs included in the current assay panel. In order to keep abreast of possible new sequence types the allele specificities need to be monitored regularly, as was done during the course of this investigation. It is also useful to point out as mentioned previously that alternative allele-specific SNPs have already been identified, which would allow adjustment of the assays to improve the sequence type coverage within a clonal complex. A SNP-based strategy would never equal the resolution obtained by full MLST characterization and would therefore not be applicable for studies of C. jejuni population genetics, but the strategy provides a complementary technique able to provide real-time information for action and intervention against a gastrointestinal infection of major public health importance.

Disadvantages of the SNP strategy

There are a few negative aspects of the SNP strategy described. Firstly if a member of a clonal complex lacked the predictive alleles then it would not be assigned by the SNP approach. In this instance it would be likely to have been obtained from a non-human source and would potentially warrant further investigation by MLST. From our experience with MLST on human isolates many isolates within the six clonal complexes have common allelic profiles containing the predictive alleles, which assign successfully. There are a number of isolates in the MLST database belonging to one of the six clonal complexes that do lack the target alleles; however, these are all from non-human sources, e.g. chicken, meat and offal, and are therefore unlikely to be encountered in a clinical microbiology laboratory. For this reason, the motivation for the strategy was to enable rapid detection of the commonest clonal complexes likely to occur within human/clinical isolates and not the more diverse and unusual allelic profiles. It would not be practical to develop a SNP scheme to encompass all the diversity seen within the C. jejuni population, especially when some isolates within the database have not been assigned to a clonal complex and there may be many C. jejuni clones yet to be identified. The system is not just another typing method (Achtman, 1996) but a simplistic means to supplement MLST with reliable data to provide a preliminary clonal complex designation in significantly reduced time and with decreased cost.

In some instances where MLST would preferentially be used, the SNP strategy may not be applicable, for example in investigations of outbreaks where more strain detail than the clonal complex was required (Sails et al., 2003b) or for very large sample sets when a strategy utilizing a microarray system as described for S. aureus may be more time and cost effective (van Leeuwen et al., 2003). Further development of the SNP approach by expansion of the strategy to include other C. jejuni clonal complexes may offer added benefits. Possible extension to other Campylobacter species where MLST schemes are available, for example C. coli, C. lari or C. fetus, may also be useful for detection purposes. Additionally any strains identified with atypical SNPs would require further investigation by MLST followed by a reiterative procedure to determine novel SNPs. The approach based upon the identification of SNPs within defined alleles could be applied to sequenced targets other than those included in the current MLST schemes. Similar approaches have been described for other organisms where rapid typing or detection methods have been implemented based on the presence of SNPs, for example in Listeria monocytogenes (Moorhead et al., 2003), Neisseria meningitidis, Staphylococcus aureus (Robertson et al., 2004) and within the Listeria genus (Koo & Jaykus, 2002). This approach for the identification of SNPs within alleles represents a significant generic method that could be applied for delivery of timely epidemiology for other pathogens.

	ACKNOWLEDGEMENTS

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
The authors wish to acknowledge the Health Protection Agency for funding through a HPA studentship; Martin Maiden, Frances Colles and Roisin Ure (Peter Medawar Building for Pathogen Research and Department of Zoology, University of Oxford, UK) for help in setting up MLST and use of sequencing facilities; the staff of the Molecular Epidemiology Department, HPA North West, Manchester, UK; the Campylobacter and Helicobacter Reference Unit, Laboratory of Enteric Pathogens, HPA, London, UK. This publication made use of the Campylobacter Multi Locus Sequence Typing website (http://pubmlst.org/campylobacter) developed by Dr Man-Suen Chan and Dr Keith Jolley and sited at the University of Oxford. Initial development of this site was funded by the Wellcome Trust; maintenance is funded by DEFRA.

	REFERENCES

TOP INTRODUCTION METHODS RESULTS AND DISCUSSION ACKNOWLEDGEMENTS REFERENCES