Rapid Detection and Molecular Differentiation of Toxigenic ...

0 downloads 0 Views 813KB Size Report
Mar 4, 2011 - sogenic beta-corynephages bearing the tox gene. tox encodes the diphtheria toxin (DT), which is responsible for the systemic symptoms of ...
JOURNAL OF CLINICAL MICROBIOLOGY, July 2011, p. 2485–2489 0095-1137/11/$12.00 doi:10.1128/JCM.00452-11 Copyright © 2011, American Society for Microbiology. All Rights Reserved.

Vol. 49, No. 7

Rapid Detection and Molecular Differentiation of Toxigenic Corynebacterium diphtheriae and Corynebacterium ulcerans Strains by LightCycler PCR䌤 Andreas Sing,1§ Anja Berger,1§ Wulf Schneider-Brachert,2 Thomas Holzmann,2 and Udo Reischl2* National Consiliary Laboratory for Diphtheria, Bavarian Health and Food Safety Authority, Veterina ¨rstraße 2, 85764 Oberschleißheim, Germany,1 and Institute of Medical Microbiology and Hygiene, University Hospital Regensburg, Franz-Josef-Strauss Allee 11, 93053 Regensburg, Germany2 Received 4 March 2011/Returned for modification 13 April 2011/Accepted 10 May 2011

The systemic symptoms of diphtheria are caused by the tox-encoded diphtheria toxin (DT) which is produced by toxigenic Corynebacterium spp. Besides the classical agent C. diphtheriae, the zoonotic pathogen C. ulcerans has increasingly been reported as an emerging pathogen for diphtheria. The reliable detection of toxigenic Corynebacterium spp. is of substantial importance for both diphtheria surveillance in the public health sector and the clinical workup of a patient with diphtherialike symptoms. Since the respective tox genes of C. diphtheriae and C. ulcerans differ from each other in both DNA and amino acid sequence, both tox genes should be covered by novel real-time PCR methods. We describe the development and validation of a LightCycler PCR assay which reliably recognizes tox genes from both C. diphtheriae and C. ulcerans and differentiates the respective target genes by fluorescence resonance energy transfer (FRET) hybridization probe melting curve analysis. sition, respectively (22). Moreover, compared to C. diphtheriae tox and DT (10) C. ulcerans tox and DT seem to be much more heterogenous, since the so-far-published 12 sequences of C. ulcerans tox and DT can be subdivided into 5 different tox and DT sequence groups, respectively (8, 17, 19, 20, 22). The differences in C. diphtheriae and C. ulcerans tox allow a differentiation between these two related virulence genes by PCR (22). For the diagnostic workup of a patient with diphtherialike symptoms, however, it is important to have a fast, reliable, and robust method that can detect both tox genes with the same assay. While a conventional C. diphtheriae tox PCR first described in 1993 and used by many laboratories worldwide for the identification of toxigenic corynebacteria amplifies both C. diphtheriae and C. ulcerans tox (5), real-time PCR assays allow a much faster laboratory diagnosis of tox-positive corynebacteria. Recently, we developed and evaluated a TaqMan-based real-time PCR assay that reliably detects both C. diphtheriae and C. ulcerans tox (18), thus overcoming the problem of a previously published TaqMan-based real-time tox PCR that detects only C. diphtheriae tox while missing C. ulcerans tox (3, 9, 24). The two most widely applied real-time PCR detection formats are TaqMan hydrolysis probes and LightCycler hybridization probes. Since diagnostic laboratories often have only one of these two real-time PCR systems available, we decided to develop and evaluate a LightCycler-based real-time PCR protocol allowing both simultaneous C. diphtheriae and C. ulcerans tox gene detection and differentiation of the two tox genes from each other by fluorescence resonance energy transfer (FRET) hybridization probe melting curve analysis in a single assay.

Diphtheria is caused by both toxigenic Corynebacterium diphtheriae and Corynebacterium ulcerans strains harboring lysogenic beta-corynephages bearing the tox gene. tox encodes the diphtheria toxin (DT), which is responsible for the systemic symptoms of diphtheria. Since the beginning of modern bacteriology, C. diphtheriae has been known as the etiologic agent of diphtheria, a disease already described in ancient times. In contrast, only recently has C. ulcerans been recognized as an emerging zoonotic pathogen which has been found more often than C. diphtheriae in reported cases of both diphtheria and diphtherialike disease in several Western industrialized countries with low diphtheria incidence (1, 11, 15, 21, 25). Corynebacterium pseudotuberculosis, the causative agent of caseous lymphadenitis in sheep and goats, may also produce a DT-like toxin. Infection of humans by toxigenic C. pseudotuberculosis is extremely rare (25). Moreover, the complete tox gene of C. pseudotuberculosis has not been sequenced so far. The reliable detection of toxigenic Corynebacterium spp. is of substantial importance for both diphtheria surveillance in the public health sector and the clinical management of the individual patient presenting with diphtheria-related symptoms. While the traditional Elek test for detection of diphtheria toxin is time consuming or not available in many laboratories worldwide due to the lack of either expertise or antitoxin, real-time PCR offers a rapid tool to confirm the presence of the diphtheria toxin-encoding gene tox in an isolate or specimen. Both tox and DT from C. diphtheriae and C. ulcerans differ from each other in about 5% of their base pair and amino acid compo-

* Corresponding author. Mailing address: Institute of Medical Microbiology and Hygiene, University Hospital of Regensburg, Franz-Josef-Strauss Allee 11, 93053 Regensburg, Germany. Phone: 49-941-944-6450. Fax: 49-941-944-6451. E-mail: udo.reischl@klinik .uni-regensburg.de. § Both authors contributed equally to this paper. 䌤 Published ahead of print on 18 May 2011.

MATERIALS AND METHODS Control strains. Twenty tox-bearing strains of C. diphtheriae and C. ulcerans collected by the German Consiliary Laboratory on Diphtheria since 1999 (Table 1) and two C. diphtheriae type strains (the toxigenic strain NCTC 10648 and the

2485

2486

SING ET AL.

J. CLIN. MICROBIOL. TABLE 1. tox-positive Corynebacterium sp. strains and control strains tested in this study Result in:

Strain

1 107 109 110 111 112 114 115 117 126 141 143 170 172 173 179 188 190 200 203 NCTC 10648 NCTC 3984 NCTC 10356 a b

Yr of isolation

Species

C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C. C.

diphtheriae ulcerans ulcerans ulcerans diphtheriae diphtheriae ulcerans diphtheriae diphtheriae ulcerans ulcerans ulcerans diphtheriae ulcerans diphtheriae diphtheriae ulcerans ulcerans ulcerans ulcerans diphtheriae diphtheriae diphtheriae

mitis

mitis gravis gravis

mitis mitis mitis

1997 2007 2007 2007 2007 2007 2007 2007 2007 2007 2009 2009 2009 2009 2009 2010 2010 2010 2010 2010

Source/gender/age (yr)a

Human/M/3 Cat/ND/ND Human/F/ND Human/F/ND Laboratory trial DIPNET Laboratory trial DIPNET Laboratory trial DIPNET Laboratory trial DIPNET Laboratory trial DIPNET Human/F/56 Human/F/62 Dog/ND/ND Human/F/11 Human/M/65 Human/F/49 Human/M/75 Human/M/64 Human/F/61 Human/F/87 Human/M/58 Positive-control strain Positive-control strain Negative-control strain

2007 2007 2007 2007 2007

tox PCRb

LightCycler PCR assay

Reference

Elek test

Positive Positive Positive Positive Negative Positive Positive Negative Positive Positive Positive Positive Positive Negative Positive Positive Negative Positive Positive Positive Positive Positive Negative

Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Negative

Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Positive Negative

14 18 18 18 12 12 12 12 12 16 7 7 7 7 7 7 7 7 7 7

M, male; F, female; ND, not determined; DIPNET, Diphtheria Surveillance Network. The tox PCR was performed as described previously (14).

weakly toxigenic strain NCTC 3984) were included in this study. In addition, 70 DT-negative strains of various Corynebacterium spp., isolated from clinical and veterinary specimens (among them 14 C. diphtheriae subsp. gravis, 29 C. diphtheriae subsp. mitis, 11 C. diphtheriae subsp. belfanti, and one C. ulcerans [18]) and the nontoxigenic type strain C. diphtheriae NCTC 10356 were used to determine the specificity of the real-time PCR. The analytical specificity of the presented assay was further assessed by testing DNA preparations of 118 Gram-positive and Gram-negative bacterial strains other than Corynebacterium spp. (13). All strains were characterized by standard microbiological and biochemical procedures, including the API Coryne system (bioMe´rieux, Nu ¨rtingen, Germany), as described previously (20, 22). Moreover, 16S rRNA gene sequencing was performed to corroborate the species ID results obtained. In addition, a 500-bp fragment of the rpoB gene was sequenced and compared with the GenBank database (6). The presence of the tox gene was investigated by a conventional tox PCR described by Hauser et al. (5). DT production was evaluated by a modified Elek Test as described previously (22, 26). Template DNA extraction. Total DNA preparations of cultured bacterial organisms or clinical specimens were prepared using a High Pure PCR template kit from Roche Diagnostics, Mannheim, Germany. DNA concentrations were determined spectrophotometrically. Real-time PCR assay design. Primer sequences were selected according to the method of Sulakvelidze et al. (23) in order to amplify a 248-bp fragment within the C. diphtheriae tox gene. Additionally, the 1-nt mismatch (Fig. 1, position 135 in the alignment) with recently deposited C. ulcerans tox gene sequences (e.g., GenBank sequence no. FJ858272) that was observed within the annealing region of primer CD-toxF was considered by applying a relatively low annealing temperature (50°C) in the corresponding real-time PCR thermocycle profile. To select LightCycler hybridization probe sequences reliably covering the tox gene in both C. ulcerans and C. diphtheriae, all tox sequences currently available in the GenBank database—eight for C. ulcerans, seven for human-derived C. diphtheriae, and six for DT-encoding prophages—were aligned. To allow for discrimination between the respective tox genes in C. ulcerans and human-derived C. diphtheriae by subsequent LightCycler melting curve analysis, the 1-nt mismatch between the sequences with GenBank sequence accession no. AY820132 (tox from C. diphtheriae) and FJ858272 (tox from C. ulcerans) was considered when designing the corresponding sensor hybridization probe sequence. Oligonucleotides and PCR protocol. Amplification primers CD-toxF (GAA AAC TTT TCT TCG TAC CAC GGG ACT AA, positions 118 to 146 in GenBank sequence no. FJ858272) and CD-toxR (ATC CAC TTT TAG TGC

GAG AAC CTT CGT CA, positions 366 to 338 in GenBank sequence no. FJ858272) and hybridization probes CD-HP-3 (AAT AAA TAC GAC GCT GCG GGA TAC-FL, positions 247 to 270 in GenBank sequence no. FJ858272) and CD-HP-4 (LC Red 640-CTG TAG ATA ATG AAA ACC CGC TC, positions 272 to 294 in GenBank sequence no. FJ858272) were synthesized by Metabion (Munich, Germany) and TIB Molbiol (Berlin, Germany), respectively. Real-time PCR amplification mixtures contained Roche LightCycler FastStart DNA Master HybProbe, 3 mM MgCl2, primer and probes in final concentrations of 500 nM and 200 nM, respectively, and 5 ␮l of template DNA. The LightCycler real-time PCR thermoprofile consisted of a initial denaturation step at 95°C for 10 min, followed by a 50-cycle amplification profile as follows: heating at 20°C/s to 95°C with a 10-s hold, cooling at 20°C/s to 50°C with a 20-s hold, and heating at 20°C/s to 72°C with a 30-s hold. Melting curve analysis (starting at 40°C) was performed with a temperature transition rate of 0.2°C/s to determine the melting temperature (Tm) values for the sequences targeted by the hybridization probes. Within each run, 103 copies of a recombinant plasmid (containing a modified segment of the expected amplicon) were tested in a separate reaction capillary as a positive control. As a negative control, PCR-grade water was used instead of DNA template. To demonstrate the absence of PCR inhibitors, a 5-␮l aliquot of the template DNA prepared from clinical samples was subjected to the LightCycler control kit DNA in a separate PCR. Analytical sensitivity was determined with a 10-fold dilution series of C. diphtheriae NCTC 10648 DNA, diluted in pooled DNA preparations from five throat swabs of C. diphtheriae-negative patients (mimicking the natural background of the PCR assay when applied to clinical specimens) (2).

RESULTS Analytical specificity of the LightCycler PCR assay. All 20 tested tox-positive isolates (8 C. diphtheriae and 12 C. ulcerans) led to positive real-time PCR results, as evidenced by probe fluorescence, whereas the 70 tox-negative Corynebacterium strains remained negative. Even after as few as 35 amplification cycles, the LightCycler curves generated allowed clear discrimination between tox-positive and -negative Corynebacterium strains (Fig. 2). The specificity of this primer and probe

VOL. 49, 2011

DETECTION OF TOXIGENIC CORYNEBACTERIA BY LIGHTCYCLER PCR

2487

FIG. 1. Multiple sequence alignment of C. diphtheriae and C. ulcerans tox genes (GenBank sequence accession no. AY820132 and FJ858272, respectively). Annealing regions of primer and LightCycler hybridization probe sequences of the novel PCR assay are indicated, together with the consistent 1-nt mismatch at position 294 (GenBank sequence accession no. FJ858272) that discriminates between the tox genes of both species. GenBank sequence no. FJ376656 was included as a representative tox gene sequence of the recently described feline C. diphtheriae isolates (4) that will be covered but misidentified as C. ulcerans by Tm analysis of the presented PCR assay. pos., position.

combination was further evaluated with various Gram-positive and -negative organisms other than C. diphtheriae or C. ulcerans (n ⫽ 118). Since all of these isolates tested negative, the analytical specificity was determined to be 100%. Analytical sensitivity. Systematic testing of DNA dilution series revealed an LOD (lower limit of detection) of 100 fg template DNA per 20-␮l PCR for three out of three independent replicates (2). A template input of 100 fg of bacterial chromosomal DNA equals 30 to 40 genome copies (calculation was based on the known genome size of C. diphtheriae). Inhibition events were not observed with any of the investigated DNA preparations (pseudo throat swabs) mimicking the natural background of the PCR assay. In comparison to the conventional tox PCR (5), the presented LightCycler PCR assay turned out to be about 25 times more sensitive. Melting curve analysis. LightCycler melting curve analysis revealed characteristic melting points for tox-positive isolates. A Tm of 62°C was observed for all of the tox-positive C. diphtheriae strains investigated (n ⫽ 8), whereas a Tm of 60°C was

FIG. 2. LightCycler amplification curves of representative tox-positive Corynebacterium sp. strains. Results obtained with a representative collection of 9 tox-positive C. ulcerans strains and 6 tox-positive C. diphtheriae strains, including C. diphtheriae NCTC 10648, 2 C. diphtheriae subsp. gravis, and 2 C. diphtheriae subsp. mitis strains, are depicted. pos., positive; neg., negative.

2488

SING ET AL.

J. CLIN. MICROBIOL.

FIG. 3. Plot of LightCycler melting curve analysis of representative tox-positive Corynebacterium sp. strains. Results obtained with a representative collection of 9 tox-positive C. ulcerans strains (Tm, ⬃60°C) and 6 tox-positive C. diphtheriae strains (Tm, ⬃62°C), including C. diphtheriae NCTC 10648, 2 C. diphtheriae gravis, and 2 C. diphtheriae mitis strains, are depicted.

observed for all of the tox-positive C. ulcerans strains investigated (n ⫽ 12). Figure 3 shows a graphical plot of the LightCycler melting curve analysis of a representative collection of tox-positive isolates. Sequencing of real-time PCR products. The amplicons of the tox-specific real-time PCRs performed on both C. ulcerans and C. diphtheriae strains were subjected to DNA sequencing in order to confirm the results of melting curve analysis at the sequence level. The sequences obtained were analyzed by employing NCBI BLAST search and “pileup” from the HUSAR sequence analysis package (http://genius .embnet.dkfz-heidelberg.de/) to construct a multiple alignment with representative sequences for each of the investigated species. The species identity and the observed nucleotide differences in the respective target sequences were confirmed, and an alignment of selected tox gene sequences is shown in Fig. 1. DISCUSSION The application of current real-time PCR technology allows a much more rapid identification of specific target genes than conventional PCR. As a consequence, many pathogen- or pathogenicity factor-specific PCR protocols have recently been adopted or newly developed for real-time PCR platforms. Especially in the field of medical microbiology, the availability of fast and reliable diagnostic results has considerable impact in guiding therapy and prophylaxis against important infectious diseases. In the last few years, several new sequences of tox genes from C. ulcerans were published in the GenBank database which differed substantially from those of C. diphtheriae tox (19, 22). When trying to amplify tox genes by previously published conventional and TaqMan-based real-time PCR assays, Cassiday et al. found that the conventional tox PCR de-

tects both C. diphtheriae and C. ulcerans tox genes, while their TaqMan-based real-time PCR assay proved to be insufficient for the detection of C. ulcerans tox due to mismatches in primer and probe binding regions (3). These limitations were overcome by a novel TaqMan-based PCR assay recently developed by our group (18). Considering that many diagnostic laboratories are only running one of the two most widely used real-time PCR concepts, i.e., TaqMan and LightCycler PCR, reliable protocols for both detection systems are needed. We therefore developed and evaluated a LightCycler PCR assay which reliably detects all human-derived C. diphtheriae and C. ulcerans tox genes published so far. Our assay is able to detect the presence of either tox gene in clinical strains with a detection limit of less than 40 genome copies. All 20 DT gene-harboring Corynebacterium sp. strains collected at the German Consiliary Laboratory for Diphtheria tested clearly positive in our real-time tox PCR, as well as in the conventional C. diphtheriae tox PCR described by Hauser et al. (5). Moreover, the previously designed C. ulcerans tox-specific PCR (23) yielded positive results for the 12 toxigenic C. ulcerans strains analyzed in this study, while it was negative for all 8 tox-bearing C. diphtheriae strains. In addition, the results of the novel LightCycler PCR and those of the conventional tox PCR by Hauser et al. (5) were congruent for all tox-positive and tox-negative Corynebacterium strains tested. Although the real-time PCR assay described in the present study was found to be a reliable and rapid method for detecting the presence of tox, both physicians and microbiologists need to be aware that an Elek test must be performed on tox-positive isolates to test for the production of a functional DT. Similarly to the long-known tox-bearing but nontoxinogenic C. diphtheriae strains, the existence of tox-positive, Elek test-negative C. ulcerans strains has been described by us (22) and recently confirmed by others (3). Besides the detection of tox from both C. diphtheriae and C. ulcerans, LightCycler melting curve analysis allowed clear differentiation between the C. ulcerans tox gene and at least all thus-far-published human-derived C. diphtheriae tox genes without any extra cost or physical manipulations. Recently, Hall et al. identified a novel group of C. diphtheriae strains isolated from a cluster of four domestic cats within a single household in the United States (4). As these feline isolates harbor a variant C. diphtheriae tox sequence (which has higher sequence identity to the C. ulcerans tox sequence than to tox sequences of human-derived C. diphtheriae strains) and the sequence similarity of the species marker gene rpoB is less than 98% compared to C. diphtheriae type strains, these feline isolates might represent a new subspecies of C. diphtheriae. Due to a 1-nt deletion at position 55 of the tox gene (according to the sequences with GenBank sequence accession no. FJ376656, FJ422272, FJ422273, and FJ422274) and a cytosineto-thymine substitution at position 74, which would prematurely terminate the resulting peptide at amino acid position 25, these novel feline C. diphtheriae strains are not able to functionally express tox, thus being nontoxinogenic (4). Therefore, the clinical significance of these strains in humans is not yet known; due to the lack of diphtheria toxin production, it might be presumed that these strains will not pose a real diagnostic problem in the workup of a patient presenting with diphtherialike symptoms. However, it should be mentioned

VOL. 49, 2011

DETECTION OF TOXIGENIC CORYNEBACTERIA BY LIGHTCYCLER PCR

that its residual gene sequence (which is almost identical to the C. ulcerans tox sequence within the amplified segment between positions 118 to 366 of the respective GenBank entries [see FJ376656 in Fig. 1]) will presumably be covered by the primers of the PCR assay presented in this study, and the resulting amplicons will consequently be misclassified as C. ulcerans tox gene by LightCycler melting curve analysis. Based on the results of our study, the novel LightCycler real-time PCR assay reliably detects the tox genes of C. diphtheriae and C. ulcerans strains. Being about 25 times more sensitive than a conventional tox PCR is certainly a diagnostic benefit when applying the presented real-time PCR assay for direct testing of clinical samples. Moreover, the assay is able to differentiate between C. ulcerans and C. diphtheriae tox genes with the performance of a simple melting curve analysis subsequent to the real-time amplification and detection process. The potential cross-reaction which could be expected with the slightly variant tox gene of the recently described nontoxigenic C. diphtheriae strains is addressed. Assuming that the occurrence of these variant feline strains is still considered a sporadic event and considering that all of the established PCR protocols targeting a segment within the tox gene should be equally affected, a potential misclassification should not interfere with the diagnostic specificity of the presented real-time PCR protocol in routine medical microbiology. In conclusion, our study, as well as the work of other groups, illustrates the need to critically reevaluate the analytical specificity of established real-time PCR assays for the detection of given bacterial targets and to adjust them in the case of any limitations in assay specificity or sensitivity becoming apparent. Since human clinical isolates of toxigenic C. pseudotuberculosis are extremely rare and the sequence of its tox gene is not yet available, we were not able to include toxigenic C. pseudotuberculosis strains in our PCR study. ACKNOWLEDGMENTS We thank Jasmin Fra¨ßdorf, Helga Kocak, Wolfgang Schmidt, Angela Huber, Rebecca Fechter, and Holger Melzl for expert technical assistance. REFERENCES 1. Bonmarin, I., et al. 2009. Diphtheria: a zoonotic disease in France? Vaccine 27:4196–4200. 2. Bustin, S. A., et al. 2009. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 55:611– 622. 3. Cassiday, P. K., L. C. Pawloski, T. Tiwari, G. N. Sanden, and P. P. Wilkins. 2008. Analysis of toxigenic Corynebacterium ulcerans strains revealing potential for false-negative real-time PCR results. J. Clin. Microbiol. 46:331–333.

2489

4. Hall, A. J., et al. 2010. Novel Corynebacterium diphtheriae in domestic cats. Emerg. Infect. Dis. 16:688–691. 5. Hauser, D., M. R. Popoff, M. Kiredjian, P. Boquet, and F. Bimet. 1993. Polymerase chain reaction assay for diagnosis of potentially toxinogenic Corynebacterium diphtheriae strains: correlation with ADP-ribosylation activity assay. J. Clin. Microbiol. 31:2720–2723. 6. Khamis, A., D. Raoult, and B. La Scola. 2004. rpoB gene sequencing for identification of Corynebacterium species. J. Clin. Microbiol. 42:3925–3931. 7. Konrad, R., et al. 2010. Matrix-assisted laser desorption/ionisation time-offlight (MALDI-TOF) mass spectrometry as a tool for rapid diagnosis of potentially toxigenic Corynebacterium species in the laboratory management of diphtheria-associated bacteria. Euro Surveill. 15:pii⫽19699. 8. Lartigue, M. F., et al. 2005. Corynebacterium ulcerans in an immunocompromised patient with diphtheria and her dog. J. Clin. Microbiol. 43:999–1001. 9. Mothershed, E. A., P. K. Cassiday, K. Pierson, L. W. Mayer, and T. Popovic. 2002. Development of a real-time fluorescence PCR assay for rapid detection of the diphtheria toxin gene. J. Clin. Microbiol. 40:4713–4719. 10. Nakao, H., I. K. Mazurova, T. Glushkevich, and T. Popovic. 1997. Analysis of heterogeneity of Corynebacterium diphtheriae toxin gene, tox, and its regulatory element, dtxR, by direct sequencing. Res. Microbiol. 148:45–54. 11. Neal, S., and A. A. Efstratiou. 2007. DIPNET—establishment of a dedicated surveillance network for diphtheria in Europe. Euro Surveill. 12:E9–E10. 12. Neal, S. E., and A. Efstratiou on behalf of DIPNET and International Diphtheria Reference Laboratories. 2009. International external quality assurance for laboratory diagnosis of diphtheria. J. Clin. Microbiol. 47:4037– 4042. 13. Reischl, U., et al. 2002. Real-time fluorescence PCR assays for the detection and characterization of Shiga toxin, intimin and enterohemolysin genes from Shiga toxin-producing Escherichia coli. J. Clin. Microbiol. 40:2555–2565. 14. Robert Koch-Institut (RKI). 1997. Diphtherie: Bericht u ¨ber zwei Erkrankungen. RKI Epidemiol. Bull. 37:255–256. 15. Robert Koch-Institut (RKI). 2008. 10 Jahre Konsiliarlaboratorium fu ¨r Diphtherie: zur Charakterisierung von C.-diphtheriae-verda¨chtigen Isolaten. RKI Epidemiol. Bull. 3:23–25. 16. Schuhegger, R., et al. 2009. Pigs as a source for toxigenic Corynebacterium ulcerans in diphtheria-like disease. Emerg. Infect. Dis. 15:1314–1315. 17. Schuhegger, R., R. Kugler, and A. Sing. 2008. Pitfalls with toxigenic Corynebacterium ulcerans causing diphtheria-like illness. Clin. Infect. Dis. 47:288. 18. Schuhegger, R., et al. 2008. Detection of toxigenic Corynebacterium diphtheriae and Corynebacterium ulcerans strains by a novel real-time PCR. J. Clin. Microbiol. 46:2822–2823. 19. Seto, Y., et al. 2008. Properties of corynephage attachment site and molecular epidemiology of Corynebacterium ulcerans isolated from humans and animals in Japan. Jpn. J. Infect. Dis. 61:116–122. 20. Sing, A., S. Bierschenk, and J. Heesemann. 2005. Classical diphtheria caused by Corynebacterium ulcerans in Germany: amino acid sequence differences between diphtheria toxins from Corynebacterium diphtheriae and C. ulcerans. Clin. Infect. Dis. 40:325–326. 21. Sing, A., and J. Heesemann. 2005. Imported cutaneous diphtheria, Germany, 1997-2003. Emerg. Infect. Dis. 11:343–344. 22. Sing, A., M. Hogardt, S. Bierschenk, and J. Heesemann. 2003. Detection of differences in the nucleotide and amino acid sequences of diphtheria toxin from Corynebacterium diphtheriae and Corynebacterium ulcerans causing extrapharyngeal infections. J. Clin. Microbiol. 41:4848–4851. 23. Sulakvelidze, A., et al. 1999. Diphtheria in the Republic of Georgia: use of molecular typing techniques for characterization of Corynebacterium diphtheriae strains. J. Clin. Microbiol. 37:3265–3270. 24. Tiwari, T. S., et al. 2008. Investigations of 2 cases of diphtheria-like illness due to toxigenic Corynebacterium ulcerans. Clin. Infect. Dis. 46:395–401. 25. Wagner, K. S., et al. 2010. Diphtheria in the United Kingdom, 1986-2008: the increasing role of Corynebacterium ulcerans. Epidemiol. Infect. 138:1519– 1530. 26. Wellinghausen, N., et al. 2002. A fatal case of necrotizing sinusitis due to toxigenic Corynebacterium ulcerans. Int. J. Med. Microbiol. 292:59–63.