ABSTRACT
Cytolethal distending toxin (Cdt) is produced by a variety of pathogenic bacteria, including pathogenic serotypes of Shiga toxin-producing Escherichia coli (STEC). The Cdt family comprises five variants (Cdt-I to Cdt-V) encoded by three genes located within the chromosome or plasmids or, in the case of Cdt-I, within bacteriophages. In this study, we evaluated the occurrence of the cdt gene in a collection of 140 environmental STEC isolates. cdt was detected in 12.1% of strains, of which five strains carried inducible bacteriophages containing the Cdt-V variant. Two Cdt-V phages of the Siphoviridae morphology lysogenized Shigella sonnei, generating two lysogens: a single Cdt phage lysogen and a double lysogen, containing a Cdt phage and an Stx phage, both from the wild-type strain. The rates of induction of Cdt phages were evaluated by quantitative PCR, and spontaneous induction of Cdt-V phage was observed, whereas induction of Stx phage in the double lysogen was mitomycin C dependent. The Cdt distending effect was observed in HeLa cells inoculated with the supernatant of the Cdt-V phage lysogen. A ClaI fragment containing the cdt-V gene of one phage was cloned, and sequencing confirmed the presence of Cdt-V, as well as a fragment downstream from the cdt homolog to gpA, encoding a replication protein of bacteriophage P2. Evaluation of Cdt-V phages in nonclinical water samples showed densities of 102 to 109 gene copies in 100 ml, suggesting the high prevalence of Cdt phages in nonclinical environments.
INTRODUCTION
Cytolethal distending toxin (Cdt) was first described in 1987 (21) as a new toxin produced by Escherichia coli strains. Bacterial Cdts belong to the AB2-type of toxins that block the G2 and early M phases during mitosis (30, 42). Consequently, the cells do not divide, but since they continue to grow, distension to five times their normal size can be observed before their disintegration. Cdt is a virulence factor that benefits bacterial survival and enhances microbial pathogenicity (31). Current experimental evidence supports the important role of Cdts in the in vivo pathogenesis of Cdt-producing bacteria, which can induce hepatocellular carcinoma and inflammation of the stomach, intestine, and liver in susceptible mouse strains during chronic infection (15).
Multiple pathogenic Gram-negative bacteria have been found to produce Cdt, including E. coli (which produces five distinct variants of Cdt), Shigella spp., Campylobacter spp., Aggregatibacter (formerly Actinobacillus) actinomycetemcomitans, Haemophilus ducreyi, Helicobacter spp., and Salmonella enterica serovar Typhi (S. Typhi) (17, 31, 38). Cdt is encoded by three adjacent genes (cdtA, cdtB, and cdtC), which express three subunits: CdtA, CdtB, and CdtC. The catalytic subunit CdtB has DNase I-like activity, whereas CdtA and CdtC are binding proteins that deliver CdtB into target cells, with cytotoxic effects (5, 15). All three cdt genes must be expressed for Cdt to initiate cellular toxicity, and all are found in these species, with the exception of S. Typhi, which only contains CdtB (36), although the role of CdtB in Salmonella pathogenesis is unclear (17). The Cdt subunits of a given genus show various degrees of similarity to the Cdt proteins of other genera.
The cdt clusters are generally located on the chromosome of Cdt-producing bacteria. However, the five variants of Cdt produced by E. coli are an exception since within this genus, Cdts can be encoded by chromosomal genes (EcCdt-II) or chromosomal genes flanked by bacteriophage P2 and lambda-like sequences (EcCdt-V), lambdoid prophage genes (Cdt-IV), and also by inducible lambdoid prophages (Cdt-I) (1) or a pVir plasmid (Cdt-III) (36).
Current information about Cdt-producing strains pertains to clinical isolates. Nevertheless, outside the clinical setting, gene exchange can lead to the emergence of new pathogenic strains, which can subsequently be found in hospitals (7, 8). This exchange is mostly conducted by elements of horizontal gene transfer, that is, genomic islands, plasmids, bacteriophages, and transposons. In recent decades, bacteriophages have increasingly been recognized as elements of horizontal gene transfer (6, 26, 41), especially since studies on microbial genomics have shown that a significant proportion of bacterial DNA corresponds to phage DNA (6). With regard to Cdt, the presence of the cdt genes within mobile genetic elements has been described previously (1, 40), and the data suggest that the cdt-I and cdt-IV genes might have been acquired from a common ancestor by phage transduction and then evolved in their bacterial hosts. In all cases, cdt genes are located downstream of the prophage head and tail genomic regions. cdtC is located upstream from a P4-like prophage integrase gene (intB) (40), and a protease gene is located downstream of cdt-IV genes.
There is a clear relationship between Cdt production and pathogenic serotypes of Shiga toxin-producing E. coli (STEC) of clinical origin, including O157 (12) and non-O157 serotypes (3, 4). Although information is available about the occurrence of other virulence factors, such as STEC and Stx phages, in nonclinical settings (13, 14, 18), the same kind of study has not been reported for cdt. The present study was conducted (i) to determine the prevalence of cdt in STEC strains of nonclinical origin, (ii) to compare this incidence with non-STEC strains of nonclinical origin, and (iii) to elucidate whether bacteriophages, as free phages or associated with E. coli, could act as a reservoir and vehicle of transmission of cdt genes.
MATERIALS AND METHODS
Bacterial strains, serotyping, bacteriophages, and media.Cdt was evaluated in a collection of 140 stx2-positive E. coli strains available in our laboratory. These strains were isolated from fecally polluted environments (domestic sewage, cattle, pigs, and poultry wastewater) according to previously described methods (13). Serotyping of O and H antigens was carried out according to the method described by Guinée et al. (16), using all available O (O1 to O181) and H (H1 to H56) antisera. All antisera were obtained and absorbed with the corresponding cross-reacting antigens to remove nonspecific agglutinins. The O antisera were produced in the Laboratory Reference for Escherichia coli, and the H antisera were obtained from the Statens Serum Institut (Copenhagen, Denmark). The collection comprised 11 O157:H7 strains, 1 O157:H− strain, and 128 strains of other serotypes.
E. coli laboratory strain DH5α, E. coli strain C600, and a Shigella sonnei clinical isolate 866 were used as host strains in the experiments described below and for propagation of the bacteriophages induced from the strains studied. Luria-Bertani (LB) broth or LB agar was used to culture the bacteria. When necessary, media were supplemented with ampicillin at 100 μg/ml or chloramphenicol at 25 μg/ml (Sigma-Aldrich, Steinheim, Germany).
Standard DNA techniques.Chromosomal DNA was prepared from 40-ml cultures as previously described (28). Chromosomal DNA was digested with XmnI restriction endonuclease (Promega, Madison, WI). This endonuclease does not cut cdt. Phage DNA was digested with ClaI and EcoRV restriction endonucleases (Promega). Restriction fragments were separated on 0.8% agarose gels in Tris-borate-EDTA buffer and stained with ethidium bromide. PCR products were purified using a PCR purification kit (Qiagen, Inc., Valencia, CA).
PCR studies.PCRs were performed with a GeneAmp PCR system 2400 (Perkin-Elmer/PE Applied Biosystems, Barcelona, Spain) with the oligonucleotides listed in Table 1. Screening for cdt in the isolates was performed by using multiplex CDT PCR (Table 1), and the different variants were later analyzed with the specific oligonucleotides. Similarly, the presence of stx2 was confirmed using the primers UP378/LP378 and S2Aup/lp, and the variants were confirmed by sequencing. Purified bacterial or phage DNA was diluted 1:20 in double-distilled water. The PCR product was analyzed by gel electrophoresis, and bands were visualized by ethidium bromide staining. When necessary, PCR products were purified by using a PCR purification kit (Qiagen).
Oligonucleotides used in this study
Real-time qPCR. (i) Preparation of standard curves.A plasmid construct was used to generate standards for the quantitative PCR (qPCR) assays. The 466-bp fragment of cdtB obtained by conventional PCR (Table 1) and purified as described above was cloned by using a pGEM-T Easy vector, into which the PCR products were inserted, according to the manufacturer's instructions (Promega, Barcelona, Spain). The construct was transformed by electroporation into E. coli DH5α electrocompetent cells. Cells were electroporated at 2.5 kV, 25-F capacitance, and 200-Ω resistance.
Colonies containing the vector were screened by conventional PCR to identify those containing the vector plus insert. The presence of the insert in the vector and its orientation were assessed by conventional PCR and sequencing by using the primers listed in Table 1. The vector containing the insert was purified from the positive colonies using a Qiagen plasmid midi purification kit (Qiagen), and the concentration of the vector was quantified by using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies/Thermoscientifics, Wilmington, DE). The reaction product was linearized by digestion with XmnI restriction endonuclease (Promega). The restricted product was purified and quantified.
The following formula was used to calculate the number of construct gene copies (GC): [concentration of the pGEM-T Easy::insert (ng/μl)/molecular weight (ng/mol)] × 6.022 × 1023 molecules/mol=the number molecules of pGEM-T-Easy::insert/μl. The number of GC/μl of the stock prepared for each gene was calculated. Serial decimal dilutions of this stock were made in double-distilled water to prepare the standard curve for qPCR. The standard dilutions were then divided into aliquots and stored at −80°C until use. Three replicates of each dilution were added for each qPCR.
A Custom TaqMan (Applied Biosystems, Spain) set of primers and probe was designed. The forward qcdtB-F/R primer and qcdt-probe with a FAM reporter (FAM, 6-carboxyfluorescein) and an NFQ quencher (nonfluorescent quencher) were used under standard conditions in a Step-One RT-PCR system (Applied Biosystems). cdtB genes were amplified in a 20-μl reaction mixture using the TaqMan environmental real-time PCR Master Mix 2.0 (Applied Biosystems). The reaction contained 2 μl of the DNA sample or quantified plasmid DNA. All samples were run in triplicate, along with the standards and positive and negative controls, the number of GC being the average of the three replicates. A 1:10,000 dilution of positive bacterial DNA was used as a positive control.
(ii) Control of nonphage DNA.Controls were performed to rule out the presence of bacterial or nonencapsidated DNA. After DNase treatment, but before release of phage DNA from the phage capsid, the samples were used as templates for conventional PCR of eubacterial 16S rRNA genes and for qPCR of the cdt (see Table 3). Negative results confirmed that all nonviral DNA was eliminated from the sample and that only genes in phage DNA would be detected.
To screen for PCR inhibition, dilutions of the standard were spiked in the samples containing phage DNA, and the experimental difference was compared to the true copies of the target genes in the standards. Inhibition of the PCR by environmental DNA was not detected.
Isolation of temperate Cdt bacteriophages and preparation of phage lysates.cdt-positive E. coli isolates were grown in 200-ml cultures from single colonies in LB broth at 37°C until reaching the exponential growth phase, determined by an optical density at 600 nm (OD600) of 0.3 measured using a spectrophotometer (Spectronic 501; Milton Roy, Belgium). Mitomycin C was added to the cultures to a final concentration of 0.5 μg/ml. These cultures were incubated overnight at 37°C with shaking in the dark. The induced cultures were centrifuged at 10,000 × g for 10 min, and the supernatants were filtered through low-protein-binding 0.22-μm-pore-size membrane filters (Millex-GP; Millipore, Bedford, MA) and treated with DNase (100 U/ml; Sigma-Aldrich, Spain).
Isolation of phage DNA. (i) Extraction of phage DNA from phage lysates.Phage lysates were concentrated 100-fold by means of protein concentrators (100-kDa Amicon Ultra centrifugal filter units, Millipore, Bedford, MA) to a final volume of 2.0 ml. Phage DNA was then isolated from the concentrated phage lysates by phenol-chloroform extraction as described previously (32). DNA quantification was performed using a NanoDrop ND-1000 spectrophotometer. The phage DNA concentration was calculated based on three isolations of phage DNA in three independent experiments.
(ii) Extraction of phage DNA from environmental samples.For extraction of phage DNA, 50-ml portions of each water sample were passed through low-protein-binding 0.22-μm-pore-size membrane filters (Millex-GP) and concentrated 100-fold using protein concentrators (100-kDa Amicon Ultra centrifugal filter units; Millipore) to a final volume of 0.5 ml (18). Samples were treated with DNase (100 U/ml of the phage lysate) to eliminate free DNA outside the phage particles. Phage DNA was then extracted as described previously (18).
Purified DNA was eluted in a final volume of 50 μl and evaluated by agarose (0.8%) gel electrophoresis. The bands were viewed following ethidium bromide staining. The concentration and purity of the DNA extracted were determined using a NanoDrop ND-1000 spectrophotometer.
(iii) Extraction of bacterial DNA from environmental samples.Portions (50 ml) of each water sample were passed through 0.45-μm-pore-size polyvinylidene difluoride (PVDF) Durapore membrane filters (Millipore), described by the manufacturer as low-protein-binding membranes. These filters allowed the phages to pass through while bacteria were retained on the surface of the filter. To remove phages retained on the filters, 10 ml of phosphate-buffered saline was added to the surface of the filter, gently agitated, and then removed by filtration. Two washing steps allowed high (99%) phage reduction without significant loss of bacteria. The membrane containing retained bacteria was recovered in 4 ml of LB. The suspension was centrifuged at 3,000 × g for 10 min. To recover DNA from both Gram-positive and Gram-negative bacteria, the pellet was suspended in 180 μl of enzymatic solution (20 mg of lysozyme/ml, 25 mg of lysostaphin/ml, 20 mM Tris-HCl [pH 8.0], 2 mM EDTA, 1.2% Triton), followed by incubation for 30 min at 37°C. DNA was then extracted using a QIAamp DNA stool minikit (Qiagen) according to the manufacturer's instructions.
Preparation of DIG-labeled cdtB-specific gene probes.A 466-bp fragment corresponding to cdtB resulting from amplification with the primers CDT-up1 and CDT-lp1 (Table 1) was labeled with digoxigenin (DIG) by incorporating DIG-11-deoxy-uridine-triphosphate (Roche Diagnostics, Barcelona, Spain) during PCR as described previously (29) and used as a probe. A DIG-labeled stx probe (28) was used to detect double lysogens.
Infectivity of Cdt phages induced from environmental E. coli strains.To evaluate the ability of the induced Cdt phages to infect diverse host strains, E. coli DH5α, E. coli C600, and S. sonnei 866 were used as hosts. A drop of phage suspension, prepared as described above, was spotted onto a monolayer. The monolayer was prepared with 1 ml of each host strain grown at an OD600 of 0.3, mixed with 2.5 ml of LB soft agar (LB broth with 0.7% agarose), poured onto LB agar plates, and incubated at 37°C overnight. After incubation, the spot test was transferred to a nylon membrane (Hybond N+; Amersham Pharmacia Biotech, Spain) and hybridized with the cdtB-specific probe. Hybridization was performed at 64°C, according to the standard procedure (32) using a DIG-labeled cdtB probe prepared as described above. Stringent hybridization was achieved with a DIG-DNA labeling and detection kit (Roche Diagnostics) according to the manufacturer's instructions.
Construction of lysogens.Lysogenic phages present in the cdt-positive environmental strains were induced as described above. The phage lysates were used to lysogenize the host strains described above (E. coli DH5α and C600 and S. sonnei 866). A portion (1 ml) of the phage lysate was dropped onto a monolayer containing the host strains prepared as described above. The plates were incubated at 37°C for 18 h. After incubation, bacteria were recovered from the colonies that grew within the drop and then harvested in 1 ml of SM buffer, 10-fold diluted, plated on LB agar, and incubated at 37°C for 18 h. The presence of cdt in the colonies was tested by use of a colony hybridization assay (29) with a DIG-labeled cdtB probe and confirmed by PCR. The lysogens were subcultured to confirm their stability.
Lysogens were tested for their ability to produce phages after induction. Phages induced from the Cdt lysogens were genotypically and morphologically characterized and compared to those isolated from the wild-type E. coli strains using the same procedures.
Analyses of cdt in bacteriophage and bacterial DNA.Purified phage DNA was digested with ClaI and EcoRI, and purified bacterial DNA was digested with XmnI. The fragments were separated by agarose (0.8%) gel electrophoresis, and bands were visualized by ethidium bromide staining. After electrophoresis, DNA was transferred to nylon membranes (Hybond N+) by capillary blotting (32). Southern blot analysis was performed with a DIG-labeled cdtB fragment probe as described above.
Sequencing of cdt-V and flanking regions in phage DNA.A 3,211-bp ClaI restriction fragment containing cdt-V detected in phage DNA was ligated with a ClaI-digested pBC-SK+ vector and transformed by electroporation into E. coli DH5α electrocompetent cells as described above.
Nucleotide sequencing of the cloned fragment was performed in triplicate with a model 377 automated DNA sequencer (Perkin-Elmer Applied Biosystems, Weiterstadt, Germany). Sequencing was carried out with an ABI Prism BigDye 3.1 terminator cycle sequencing ready reaction kit (Perkin-Elmer), plus Cdt primers and primers for the pBC-SK+ vector (Table 1). Sequencing was performed in triplicate. The sequences were assembled by using the software DNA Baser v3.2. Nucleotide sequence analysis and searches for open reading frames (ORFs) and homologous DNA sequences in the EMBL and GenBank database libraries were performed with the tools available from the National Center for Biotechnology Information (www.ncbi.nlm.nih.gov).
Electron microscopy.The cdt-converting bacteriophages present in 20 ml of culture obtained after induction of each strain were purified by CsCl centrifugation (32). The easily visible gray band in which the bacteriophages were expected (11, 32), corresponding to a density of 1.45 ± 0.02 g/ml, was collected and dialyzed to remove the CsCl. A drop of this phage suspension was deposited on copper grids with carbon-coated Formvar films and stained with 2% KOH phosphotungstic acid (pH 7.2) for 2.5 min. Samples were examined in a Hitachi EM 800 electron microscope operating at 80 kV.
Evaluation of Cdt production.To determine the production of Cdt by the lysogens, a cell culture assay was used. The Cdt activity of the lysogens in HeLa cells was determined by using various dilutions (1/2, 1/4, 1/8, 1/10, 1/20, and 1/50) of filtered supernatants from the overnight cultures of strains in Trypticase soy broth. The distending effect was evaluated at the highest dilution that caused 50% of the HeLa cells in a well to be distended up to 96 h of incubation.
Production of toxin was compared for each strain with or without mitomycin C induction and with S. sonnei 866 as the negative control. For this purpose, bacteria were grown from single colonies in LB at 37°C to the exponential growth phase, measured as described above. At this point, each culture was divided into two aliquots, and mitomycin C was added to one of the aliquots as described above for bacteriophage induction. Bacteria were then incubated at 37°C for 18 h. The cultures were centrifuged at 12,000 × g for 10 min, and the supernatant was filtered through 0.22-μm-pore-size membrane filters (Millex-GP).
Environmental samples. (i) Urban sewage.We used 34 sewage samples collected from the influent of an urban sewage plant that serves the urban area of Barcelona, including a number of cities and towns, of approximately 500,000 inhabitants. No animal fecal contamination was expected in these samples. A total of 16 of these samples were used for phage DNA analysis; the remaining 18 samples were used for bacterial DNA analysis.
(ii) Cattle wastewater.Fifteen samples containing exclusively fecal contaminants of cattle origin were collected from slaughterhouse wastewater effluents. No human fecal contamination was expected in these samples.
The urban sewage and animal samples were collected regularly approximately every 15 days over 7 months. Portions (50 ml) of each sample were analyzed. Bacterial and phage DNA was isolated from these samples as described above, in order to determine the presence of cdt-V in the bacteriophage and bacterial fractions of the environmental samples.
Isolation of cdt-positive E. coli isolates.The prevalence of cdt in non-Shiga-toxin-producing E. coli isolates was also assessed. For this, serial decimal dilutions of sewage samples were filtered through 0.45-μm-pore-size membranes (EZ-Pak Filters; Millipore). The membranes were then transferred onto Chromocult coliform agar (Merck, Darmstadt, Germany) and incubated at 37°C for 24 h. Enumeration of the E. coli bacteria was performed by counting the blue colonies grown on the plate. Later, plates from dilutions presenting heavy but nonconfluent colony growth were selected for colony transfer, which was performed by carefully placing a nylon membrane onto the surface of the agar and quickly peeling it from the plate. The bacterial cells on the membranes were then lysed and fixed, and the membranes were hybridized, as described previously (13), using DIG-labeled cdtB. Colonies showing a positive hybridization signal were confirmed by PCR for the presence of cdt.
Nucleotide sequence accession number.The nucleotide sequence of the 3,211-bp fragment of phage Φ125 containing the cdt-V gene cluster was submitted to the GenBank database library and assigned accession number JF461073.
RESULTS
Screening for Cdt phages in the environmental STEC collection.The presence of cdt and different cdtB variants in an stx2-positive E. coli collection of 140 strains isolated from different nonclinical water samples containing human and cattle fecal pollution was screened using a multiplex PCR (Table 1). The prevalence of cdt in the collection was 12.1% (Table 2). The presence of cdt was confirmed by colony blot hybridization with the DIG-labeled cdtB probe (Fig. 1A). Specific primers for variants Cdt-I to Cdt-V (Table 1) were used to characterize the cdt variant in each strain, and the results (Table 2) showed a higher prevalence (89.5%) of cdt-V among the isolates. The strains were also characterized for the presence of Stx2 variants, as previously described (13).
Characterization of cdt-positive STEC environmental isolatesa
Colony hybridization showing the cdt-positive strains (47, 62, 78, 115, and 125) and the Cdt lysogens S(Φ62) and S(Φ125) obtained using S. sonnei as the host strain (A) and plaque blot hybridization with the cdtB-DIG probe of phages spotted onto the S. sonnei host strain (B).
To determine whether cdt was located in the genome of an inducible prophage, phages were induced with mitomycin C. Induction was measured by a reduction of the OD600 of the bacterial culture, which indicates activation of the lytic cycle of induced prophages (Table 2). In parallel, since the strains were stx2 positive, the presence of inducible Stx2 bacteriophages was also evaluated by PCR (Table 2).
Many strains showed a reduction of the optical density after mitomycin C treatment, indicating induction of the lytic cycle by activation of temperate phages present in the strains. This reduction can be caused by Cdt phages or other inducible bacteriophages (such as Stx phages or others). Accordingly, an area of lysis corresponding to the drop of the phage lysate obtained from cultures of the cdt-positive strains was observed on an agar monolayer containing bacterial host strains E. coli C600, E. coli DH5α, and S. sonnei 866. To confirm the presence of inducible Cdt phages in the phage suspension, real-time qPCR demonstrated the presence of inducible Cdt phages from the strains although differential rates of induction were observed among the strains. Moreover, to detect infectious Cdt phages, we hybridized the lytic spot with the DIG-labeled cdtB probe. Infectious Cdt phages were confirmed (Table 2) in five cdt-positive environmental strains (strain 47, serotype O91:H21; strain 62, serotype O22:H8; strain 78, serotype O89:H19; strain 115, serotype O91:H21; and strain 125, serotype O157:H7), giving a rate of detection of inducible phages of 29.4%. The cdt variant in these phages was again confirmed as cdt-V using the specific set of primers and sequencing, with the present study providing the first description of inducible bacteriophages carrying the cdt-V variant. None of the five phages produced clear lytic areas on the agar monolayer with S. sonnei (Table 2), but all showed positive signals after hybridization (Fig. 1B), some signals being stronger (phages Φ62, Φ115, and Φ125) than others (phages Φ47 and Φ78). Positive hybridization signals were not observed for the E. coli strains. The levels of Cdt phage induction of the cdt-positive strains were evaluated by real-time qPCR.
Isolation of Cdt phages from the STEC strains and generation of lysogens.Cdt phages were propagated after induction of 200-ml cultures of strains 47, 62, 78, 115, and 125. Lysogens were generated to check the ability of Cdt phages to integrate within a host and to transduce cdt. Working with lysogens also prevents possible interference from other phages, which might be present in the wild-type strains, as the experiments described below revealed. We tried to produce lysogens for the Cdt phages and we were successful in producing cdt lysogens for two of these, corresponding to phages Φ62 and Φ125, using S. sonnei as the recipient strain, generating lysogens S(Φ62) and S(Φ125) (Fig. 1A). After several attempts, no lysogens were obtained from E. coli strain C600. Lysogens S(Φ62) and S(Φ125) were again able to produce infectious Cdt phages, which showed positive hybridization signals in S. sonnei after induction (Fig. 1B). Lysogens S(Φ62) and S(Φ125) were subcultured to confirm their stability, and our observations indicated the stability of the phage within the lysogen.
Morphological characterization of Cdt phages.Electron microscopy observations showed the presence of phages in both lysogens S(Φ62) and S(Φ125). These phages corresponded to the Siphoviridae morphology (11) with a capsid of 45 ± 2 nm in diameter and a noncontractile tail of 150 × 10 nm (Fig. 2A). However, during evaluation of the phage lysate induced from lysogen S(Φ62), a second phage, of a different morphological type, was observed. This second phage was present in lower numbers (ratio 1/100 compared to the first one) and corresponded to the Podoviridae morphology (11) with a capsid of 51 ± 2 nm in diameter and a short tail (Fig. 2B).
Micrographs of Cdt phages and Stx phages induced from lysogens. (A) Siphoviridae morphological type of bacteriophages induced from lysogens S(Φ62) and S(Φ125). (B) Podoviridae morphological type of bacteriophages induced from lysogen S(Φ62). Bar, 100 nm.
Double lysogeny in strain S(Φ62) and spontaneous induction of Cdt phages.Double lysogeny (Cdt phage and Stx phage) in strain S(Φ62) was suspected after electron microscopy examination. Since the wild-type strain 62 also harbored an inducible Stx phage, it is likely that this phage was transferred at the same time as the Cdt phage to the host strain. PCR analysis with the primers S2Aup/lp and UP378/LP378 (Table 1) confirmed the presence of stx in lysogen S(Φ62) but not in lysogen S(Φ125). This is the first description of two phages harboring two different toxins being simultaneously transferred to a host strain without using any kind of selection.
Differential induction rates of Cdt phages from the two lysogens, S(Φ62) and S(Φ125), was evaluated by real-time qPCR (primers and probes in Table 1), designed to detect cdtB-V. After mitomycin C induction a reduction in the absorbance of the cultures (OD600) was observed. The wild-type strains (62 and 125) and lysogen S(Φ62) showed a reduction in absorbance from 0.7 to 1.2 U but no differences in absorbance (<0.2 U) were observed for lysogen S(Φ125). Phage DNA was extracted from the phages present in the supernatant of cultures of the wild-type strains and their lysogens, with or without mitomycin C treatment. A portion (10 μl) of this DNA was used as a template to quantify the number of GC of cdtB. Since Cdt phages are only known to carry one cdt copy, the cdt GC values can be extrapolated to the number of Cdt phages in each sample. The number of Cdt phages induced from the two lysogens was compared to those from the respective wild-type strains and the induced and noninduced cultures (Fig. 3). The presence of cdt-V in phage DNA was clearly observed but induction appeared to be independent of the presence of mitomycin C, since no significant differences (Student t test, P > 0.05) in the densities of Cdt phages were observed between induced and noninduced cultures. Cdt phages seem to be released spontaneously.
Densities of Cdt phages (log10 GC) and Stx phages (log10 GC) with or without mitomycin C induction from wild-type strains 62 and 125 and lysogens S(Φ62) and S(Φ125), as evaluated by real-time qPCR.
To confirm the simultaneous transduction of Stx and Cdt phages in lysogen S(Φ62) and to evaluate the rates of induction of both phages, analysis by qPCR with a set designed for stx (Table 1) was carried out in parallel to detect stx genes in phage DNA from the wild types and the lysogens. The presence of inducible Stx phages was confirmed in the wild-type strains 62 and 125, as well as in lysogen S(Φ62). As expected, Stx phages were not detected in lysogen S(Φ125). In this case, in contrast to Cdt phages, the densities of Stx phages (GC) showed a significant (Student t test P < 0.05) increase after mitomycin C induction (Fig. 3).
Cdt-V phages insert into a different site than the CDT-1Φ phage.To identify the putative site of insertion of Cdt-V phages, we analyzed a previously described insertion site for phage CDT-1Φ in both the wild-type strain and the lysogen. The results indicate that Cdt-V phages were not inserted into the same site, since this chromosomal site, evaluated using the primers B1 and B2 (Table 1), appeared intact. Moreover, these phages have different attP sequences than phage CDT-1Φ since no amplimer was obtained with primers P1 and P2 (Table 1), as previously described by Asakura et al. (1).
Sequencing of cdt flanking regions.Since lysogen S(Φ62) harbors two phages, characterization of a Cdt-V phage continued using phage Φ125 induced from the single lysogen S(Φ125). The insertion of this phage in the lysogen was confirmed by Southern blot analysis of XmnI restricted chromosomal DNA of the lysogen and hybridization with the cdt probe (Fig. 4A). In order to sequence the complete cdt-V and, if possible, its flanking regions, phage Φ125 DNA was also analyzed by Southern blot hybridization (Fig. 4B). ClaI restriction of DNA of phage Φ125 identified a 3,211-bp band containing the cdt gene. This fragment was cloned within a pBC-SK vector and sequenced. The sequence of the fragment confirmed that this phage harbored cdt-V (Fig. 4C and Table 3). The cdtA, cdtB, and cdtC genes showed a similarity of 100% with previous cdt sequences, some of them identified as cdt-V (4, 19). The fragment also contained a truncated ORF downstream from cdtC that shows homology with the replication protein A of bacteriophage P2 (24).
Southern blot analysis for detection of cdt fragment in an XmnI-restricted bacterial DNA from S. sonnei 866, lysogens S(Φ62) and S(Φ125) (A); in ClaI- and EcoRI-restricted phage DNA obtained from lysogen S(Φ125) (B); and genetic organization of the 3,211-bp ClaI fragment of phage Φ125 (C).
Sequence comparison of ORFs detected in the 3,122-bp fragment sequenced from phage Φ125 with previously published sequences
Expression of Cdt by lysogens evaluated by HeLa cell assay.After 5 days of incubation, sterile filtrates from lysogen S(Φ125) were analyzed using HeLa cells. Distension of the cells allowed confirmation of the expression of Cdt in the supernatant of lysogen S(Φ125). Cultures were grown in the presence or absence of mitomycin C, and a similar distending effect was observed in both cultures (Fig. 5). No morphological changes were observed in the controls of the host strain S. sonnei (Fig. 5). In addition, we performed dilutions 1/2, 1/4, 1/8, 1/10, 1/20, and 1/50 of the supernatants of lysogen S(Φ125) grown in the presence or absence of mitomycin C. The distending effect was observed in the cells from dilution 1/8 onward in both cultures, suggesting that the amounts of Cdt present in the supernatants with or without mitomycin C were similar.
Distending effect of sterile culture filtrates from lysogen S(Φ125) on HeLa cells after 96 h of incubation. (A and C) Lysogen S(Φ125) with and without mitomycin C induction, respectively. (B) S. sonnei strain 866. (D) HeLa cells incubated for 96 h. Bar, 200 μm.
Evaluation of Cdt expression in the wild-type strain 125, strain 62, and lysogen S(Φ62) was not possible since the distending effects were masked by the cytotoxic effect occurring after 48 to 72 h, probably caused by the Stx produced by these strains, and therefore the cell culture was destroyed before the Cdt distending effect was visible.
Quantification of free Cdt phages and bacteria in environmental samples.Since Cdt temperate phages have been isolated from strains of environmental origin, we decided to evaluate their presence in nonclinical environments. To evaluate this question, real-time qPCR, which allows detection of the B subunit of the cdt-V variant, was set up to evaluate its presence in phage DNA isolated from human urban sewage and wastewater containing cattle fecal pollution.
The cdt gene was detected in phage DNA isolated from more than 62.5% of the samples of urban sewage and in 73.3% of the cattle wastewater analyzed, with an average density of 103 GC/100 ml for both types of sample (Fig. 6). The percentage of positive samples in this set of experiments, as well as the high maximal densities of cdt-V in the phage DNA of some of the samples, indicates that free Cdt phages are quite common in fecally polluted samples, although the densities of cdt-V GC found in the same type of samples were not homogeneous. Controls confirmed that this was not due to contamination of DNA other than phage DNA and enabled us to rule out the presence of inhibitors in the environmental samples analyzed.
The left chart shows the densities of cdt genes (log10 GC/100 ml) in phage DNA isolated from 16 urban sewage samples and 15 wastewater samples from cattle. The left table summarizes the percentages of positive samples, geometric means, and maximum and minimum values of cdt-V (GC/100 ml) in phage DNA isolated from the human and cattle samples. The right chart and table show the densities of cdt genes (log10 GC/100 ml) in bacterial DNA isolated from 18 urban sewage samples.
cdt was also present in the bacterial fraction of the sewage samples analyzed. cdt was detected in bacterial DNA isolated in >88% of the urban sewage samples, and only two samples were negative for bacterial DNA analysis. The densities of cdt in bacterial DNA were more homogeneous than those observed in phage DNA, with an average density of 3.5 × 103 GC/100 ml (Fig. 6).
Isolation of cdt in environmental E. coli isolates.To elucidate the prevalence of cdt genes and Cdt inducible phages in wild-type isolates of E. coli in the environment, independently of the presence of stx, we searched for cdt-positive E. coli in eight independent sewage samples by colony blotting. We isolated six E. coli strains (Table 4), which means a percentage of cdt in non-STEC E. coli isolates between 0.05 and 1.18% of the densities of E. coli grown in Chromocult agar. Among these, the variant most frequently detected was cdt-IV. The presence of inducible prophages was deduced by reduction of the absorbance of the cultures, and the presence of inducible Stx phages was confirmed by real-time qPCR. Cdt phage induction was independent of mitomycin C since no differences in the cdt copies in phage DNA were observed when comparing cultures treated with or without mitomycin C. However, none of these strains carried Cdt phages capable of infecting and generating lysogens using S. sonnei as the host strain.
Occurrence of cdt-positive E. coli environmental isolatesa
DISCUSSION
The STEC strains analyzed in this study are nonclinical isolates, isolated from fecally polluted water on the basis of the presence of Stx. The incidence of cdt in the environmental collection was similar to that in clinical isolates (3, 4, 12, 19), showing that Cdt-producing STEC strains, which may represent a potentially novel virulence clone (40), are present in the environment. The prevalence of cdt-V in our cdt-positive strains was very high. Among these, the prevalence of cdt-V in non-O157 serotypes (8.6%) was similar to previous data (12), although it should be noted that our collection was mainly composed of non-O157 serotypes (only 11 O157:H7 among the 140 isolates), since serotype O157:H7 is not commonly found in environmental settings (13). Despite the low frequency of O157:H7 in our collection, four of these strains harbored cdt (36.4%). Janka et al. (19) reported that the cdt-V allele is present in the majority of sorbitol-fermenting E. coli O157:NM strains. Later on, cdt-V was also reported in 5% of non-O157 enterohemorrhagic E. coli (EHEC) clinical isolates (4). In our study, cdt-V was also found in three isolates of serotype O91:H1. There were five isolates of serotype O91:H21 in our collection (data not shown), and although the number of isolates was low, cdt showed a prevalence of 60% in this serotype, which is in accordance with previous reports (3, 22) that showed a prevalence of 70% in larger strain collections. Some authors have suggested that some serotypes lacking important virulence factors harbor cdt and that cdt may therefore contribute to the pathogenicity of these strains. Our STEC strains, like those described previously (12), lack many virulence genes, such as eae (data not shown), which encodes intimin (20). As suggested in the latter report, this underlines the virulence potential of Cdt-V.
A novelty of the present study is that cdt-V was found in an inducible Cdt bacteriophage. The location of cdt-V in phages, although not demonstrated experimentally, has been suggested previously (12, 40) and is supported by the P2 sequences or lambdoid phage sequences found in the flanking regions of cdt-V in several bacterial isolates. Inducible phages were detected in all cdt-positive strains in the present study, but the numbers of Cdt phages detected after induction were different. The isolates inducing more phages were those generating infectious Cdt phages detected with S. sonnei, and two of them generated lysogens, perhaps because the number of phages induced increases the chances of infection and subsequent transduction. Difficulties with regard to isolating and detecting Cdt phages were confirmed here, and we were only able to obtain good results when we used the S. sonnei strain as a host. The lack of signals when using E. coli as the host strain, as well as the weak signals obtained with some Cdt phages in S. sonnei, indicates that the presence of Cdt-inducible bacteriophages in previous studies could have been underestimated. Although we do not know the reason why this strain is a good host for temperate phages, this S. sonnei strain had previously shown good abilities to generate lysogens of temperate Stx bacteriophages (28, 33). This is the reason why we used it here for the detection of Cdt phages and generation of lysogens.
The Cdt-V phages analyzed here were produced at high rates in the absence of an inducing agent such as mitomycin C; therefore, they are produced by spontaneous induction (25). This was observed in the wild-type strain, as well as in the lysogens. Variation in the rates of phage induction was observed as a reduction in the absorbance units when adding mitomycin C to the cultures; however, this approach is not conclusive when strains carry more than one temperate phage. This was the case with the wild type and with lysogen S(Φ62), which both carry a Stx phage, together with the Cdt phage. In contrast, no variation in the optical density was observed for lysogen S(Φ125), apparently only carrying a Cdt phage. Confirmation of the spontaneous release of Cdt phages was achieved after evaluation of Cdt phage induction by qPCR. This mitomycin C-independent induction contrasts with the induction of phage CDT-1Φ (1), which was only produced by mitomycin C-induced cultures and not in the absence of mitomycin C.
Accordingly, expression of the Cdt occurred in the S(Φ125) lysogen, independently of mitomycin C induction, since the same distending effect was observed in the presence or absence of mitomycin C and at the same dilution of the supernatant. It was not possible to observe the effect of Cdt in lysogen S(Φ62) or in the STEC wild-type strains, since the cytotoxic effect of Stx destroyed the cell culture before any distending effect could be observed.
If we assume that stx is mobilized by bacteriophages, and given that the present study and previous reports suggest that many cdt variants are also located in the genome of inducible or noninducible bacteriophages, then it is reasonable to think that in those strains harboring both genes, double lysogeny has occurred at some point during their evolution. Since many STEC strains are reported to harbor cdt (3, 4, 12), double lysogeny of Cdt and Stx phages should be a common event. However, this simultaneous transduction of both phages without any kind of selection has never been demonstrated experimentally before now. Previous reports from our group indicated that double lysogeny, proposed as a mechanism to increase genetic variability among strains, seems to occur at higher rates than single lysogeny (34). A second phage seems to be introduced into a previously lysogenized strain more easily than in a nonlysogenic strain. Moreover, triple lysogeny can be achieved more easily than double lysogeny (R. Serra-Moreno, unpublished data). This applies to different Stx phages but also to the same Stx phage, which conflicts with the outcome expected according to the model of immunity to superinfection proposed for phage λ. This has also been demonstrated by other authors (10), indicating that double and triple lysogens of a single phage are routinely detected from a single infection of a double lysogen.
Once in the same background, both phages are induced, although at different rates and obviously through different pathways. The Stx phage is activated through the SOS system (23, 43), and although SOS-independent spontaneous induction has also been demonstrated (25), the number of Stx phages increased after SOS activation. The Cdt-V phage described here seemed to be induced independent of mitomycin C treatment in the same bacterial background, indicating a spontaneous release, a mechanism independent of inducing agent (25), that could involve the bacterial integration site or the phage biology itself.
Characterization of the Cdt-V phage showed a phage of Siphoviridae morphology, similar to the CDT-1Φ phage described by Asakura et al. (1). The Cdt-V operon in this phage is flanked by a bacteriophage region showing homology with a fragment of the gpA encoding a protein from phage P2. The gpA is truncated in the 3′ region by the insertion of the cdt operon, suggesting an insertion event that must cause the disruption of gpA. Homology with P2 in the flanking regions of the cdt-V cluster has been widely reported (12, 40), indicating a common trend for this cluster. On the contrary, the cdt-I and cdt-IV flanking sequences are not related to P2 phage genes (1, 40). For example, in CDT-1Φ, the cdtI cluster is located adjacent to and upstream of the integrase gene. Another difference in Cdt-V phages from our study is that they do not insert within the S. sonnei or E. coli genome at the same site as described for CDT-1Φ, since the site was not occupied in either strain containing the cdt-V phage. Moreover, the primers used to amplify the attP regions in CDT-1Φ are not useful for Cdt-V phages. Variability among phages harboring the same toxin is not surprising, because the mosaic structure of bacteriophages means that many of them could interchange different fragments of diverse origin.
Given that Cdt phages were easily induced from STEC strains present in nonclinical settings, free Cdt phage particles could be found among the healthy human population. Therefore, we screened wastewater samples containing human and cattle fecal pollution in which free Cdt-V phages could be expected. These values were compared to the prevalence of cdt in bacterial DNA of the sewage samples. Our approach allowed us to isolate DNA from Gram-positives and Gram-negatives, which include E. coli as well as other genera. Urban sewage was selected based on its relationship with human pathogenicity, and cattle wastewater was selected since cattle are a major source of STEC, to which cdt-V is related. The presence of cdt in the viral DNA fraction of the environmental samples suggests that phages act as a reservoir of cdt genes, as previously shown for Stx phages (18, 27, 37) or phages carrying antibiotic resistance genes (26). It seems likely that Cdt phages, found as free particles in wastewater, could behave as reservoirs of cdt in the environment. Moreover, if some of these phages are spontaneously induced, they could be released from bacterial strains after induction and found as free particles even in the absence of an SOS response-inducing agent. cdt has also been detected in bacterial DNA from environmental samples. Densities are similar to those of phages, but with more homogeneous values among samples. Evaluation of cdt-positive, non-STEC, E. coli environmental isolates showed a lower prevalence of cdt (0.3 to 1.2%) than in our collection of STEC strains (12%). However, the prevalence of STEC in the same type of urban sewage, which is 0.1% (13), must be considered. Differences among these studies and real-time qPCR results can be explained by the fact that the previous approach is limited to those E. coli strains able to grow on Chromocult, while real-time qPCR evaluates any kind of bacterial DNA independently of the viability of the strain and its ability to grow in Chromocult.
It is interesting that in the six non-STEC E. coli environmental isolates, Cdt-IV was the predominant variant, rather than Cdt-V, which appears to be more prevalent among STEC strains. However, the number of cdt-positive, non-STEC environmental isolates is too low to consider these data to be any more than a mere indication. These isolates were able or not able to induce Cdt phages at high rates according to the real-time qPCR results. Consequently, no lysogens were obtained with these phages. The significant presence of Cdt-V phages in environmental strains correlates with the fact that Cdt-V phages are present as free particles in wastewater, in which phages are known to be highly resistant to inactivation processes (2, 35) and from where they can infect and transduce cdt-V to environmental strains, causing the emergence of new Cdt-producing bacteria.
ACKNOWLEDGMENTS
We thank Andreu García Vilanova and Aiora Aregita for excellent technical assistance.
This study was supported by the Generalitat de Catalunya (2009SGR1043), by the Spanish Ministry of Education and Science (AGL2009-07576), and the Xarxa de Referència en Biotecnologia. A.A.-G. was supported by a grant FI from the Generalitat de Catalunya, Catalonia, Spain.
FOOTNOTES
- Received 15 March 2011.
- Returned for modification 13 April 2011.
- Accepted 26 May 2011.
- Accepted manuscript posted online 6 June 2011.
- Copyright © 2011, American Society for Microbiology. All Rights Reserved.
