ORIGINAL ARTICLE Comparison of polymerase chain reaction systems for detection of different cdt genes in Escherichia coli strains M. Oloomi and S. Bouzari Molecular Biology Unit, Pasteur Institute of Iran, Tehran, Iran Introduction Since cytolethal distending toxin (CDT) was first identi- fied by Johnson and Lior (1987) from Escherichia coli, several studies reported that CDT can be produced by different pathogenic bacteria including Shigella dysenteriae (Okuda et al. 1995), Haemophilus ducreyi (Cope et al. 1997), Actinobacillus actinomycetemcomitans (Sugai et al. 1998) and Campylobacter spp. (Eyigor et al. 1999). The toxin has an activity in culture supernatants of these bac- teria that caused cultured cells to become slowly disten- ded over a period of 2–5 days and eventually die (Peres et al. 1997). For production of the active toxin in E. coli and other bacterial species three genes are required; cdtA, cdtB and cdtC genes are arranged in an operon and are required to produce an active toxin (Scott and Kaper Keywords cdt, cytolethal distending toxin, detection, Escherichia coli, PCR. Correspondence S. Bouzari, Molecular Biology Unit, Pasteur Institute of Iran, Tehran, Iran. E-mail: saeidbouzari@yahoo.com 2005/0683: received 15 June 2005, revised 1 November 2005 and accepted 16 December 2005 doi:10.1111/j.1472-765X.2006.01874.x Abstract Aims: Cytolethal distending toxins (CDT) are tripartite toxins encoded by three adjacent or overlapping genes (cdtA, cdtB, cdtC) and found in multiple patho- gens. The present knowledge regarding heterogeneity of cdt genes and our pre- vious study revealed that the available polymerase chain reaction (PCR) systems lack adequate specificity. The detection of various cdt genes present in Escherichia coli strains, from different geographical regions demands further assays for wide-range coverage. On the basis of these observations, we were prompted to undertake the present study; hence the specificity of existing PCR systems was addressed using E. coli prototype strains with known cdt gene sequences. Methods and Results: A multiplex PCR designed for the detection of E. coli cdt genes was found to be sensitive and specific enough for initial screening. How- ever, for subtyping, the PCR systems yielded nonspecific products upon ampli- fication. These primers are usually designed for sequences of the cdtB locus (the most conserved region of the gene), and since CDT-producing E. coli strains carry different cdt genes, none of the systems are really type specific. Furthermore, PCR systems with type-specific primers for other regions of the gene, i.e. ORF A or ORF C are found to be strain specific and their applica- tions in different geographical regions have limitations. Conclusions: In conclusion, based on our observations, using these available primers, it seems that the existing PCR systems are not sufficiently accurate to differentiate between different types of cdt genes. Significance and Impact of the Study: The results obtained from this study revealed that so-far reported PCR systems are short in specificity. These PCR protocols were not found to be specific enough to detect various cdt genes and have a limited range of application. Moreover, due to similarities in cdt genes the cross-reaction between different sets of primers exists. Hence for epidemio- logical studies, some additional PCR protocols are required for screening clini- cal isolates for cdt genes. Letters in Applied Microbiology ISSN 0266-8254 ª 2006 The Authors Journal compilation ª 2006 The Society for Applied Microbiology, Letters in Applied Microbiology 42 (2006) 445–451 445