Presence of Activatable Shiga Toxin Genotype (stx_2d) in Shiga Toxigenic Escherichia coli from Livestock Sources

Abstract

Stx2d is a recently described Shiga toxin whose cytotoxicity is activated 10- to 1,000-fold by the elastase present in mouse or human intestinal mucus. We examined Shiga toxigenic Escherichia coli (STEC) strains isolated from food and livestock sources for the presence of activatable stx_2d. The stx₂ operons of STEC were first analyzed by PCR-restriction fragment length polymorphism (RFLP) analysis and categorized as stx₂, stx_{2c vha}, stx_{2c vhb}, or stx_{2d EH250}. Subsequently, the stx_{2c vha} and stx_{2c vhb} operons were screened for the absence of a PstI site in the stx_2A subunit gene, a restriction site polymorphism which is a predictive indicator for the stx_2d (activatable) genotype. Twelve STEC isolates carrying putative stx_2d operons were identified, and nucleotide sequencing was used to confirm the identification of these operons as stx_2d. The complete nucleotide sequences of seven representative stx_2d operons were determined. Shiga toxin expression in stx_2d isolates was confirmed by immunoblotting. stx_2d isolates were induced for the production of bacteriophages carrying stx. Two isolates were able to produce bacteriophages φ1662a and φ1720a carrying the stx_2d operons. RFLP analysis of bacteriophage genomic DNA revealed that φ1662a and φ1720a were highly related to each other; however, the DNA sequences of these two stx_2d operons were distinct. The STEC strains carrying these operons were isolated from retail ground beef. Surveillance for STEC strains expressing activatable Stx2d Shiga toxin among clinical cases may indicate the significance of this toxin subtype to human health.

Shiga toxigenic Escherichia coli (STEC) isolates are important food-borne pathogens that exist as commensal bacteria of ruminant animals. Shiga toxins (Stxs) comprise an A subunit that carries the toxic function and a B-subunit pentamer that binds the toxin to the eukaryotic cell receptor (for a review, see reference 22). Studies have indicated that Stx type 2 (Stx2), encoded by the stx₂ operon, has an epidemiological relationship with the severe human disease conditions hemolytic-uremic syndrome and hemorrhagic colitis. In addition, toxins Stx2 and Stx2c, encoded by different genotypic subtypes of stx₂, have also been associated with a greater severity of human disease. In contrast, the more recently described stx₂ genotype, stx_{2d EH250} (26), is apparently of lesser clinical significance.

The nomenclature designating the Stx2d EH250 subtype is complicated by the prior use of Stx2d to designate activatable Stx2 (16, 18). Both Stx2d and Stx2d EH250 possess Ser291 and Glu297 residues in the Stx2A subunit toxin-coding region; however, only Stx2d is activatable. While STEC isolates carrying stx_{2d EH250} are commonly found, particularly from sheep (8), very few reports describe activatable stx_2d genotype in STEC.

Activatable stx_2d was originally detected in STEC strain B2F1. Two activatable operons, stx_2d1 and stx_2d2, have been identified in the B2F1 (17). Prior to their designation as stx_2d1 and stx_2d2, these operons were designated stx_{2c vha} and stx_{2c vhb}, respectively, due to the restriction fragment length polymorphisms (RFLPs) present in the stx_2B genes. However, it is now clear that stx_2d1 and stx_2d2 are not synonymous with the stx₂ subtypes stx_{2c vha} and stx_{2c vhb}, respectively; rather, stx_2d1 and stx_2d2 are subsets of stx_{2c vha} and stx_{2c vhb}, respectively, which are further defined by additional mutations in the stx_2A subunit gene encoding the activatable toxin phenotype.

STEC strain B2F1 is extremely virulent in an orally infected streptomycin-treated mouse model, and the Stx2d toxins produced by this strain have increased cytotoxicities for Vero cells after incubation with mouse or human intestinal mucus (17). Stx2dA subunits possess two amino acid substitutions relative to the sequence of classical Stx2, Ser291 and Glu297. These residues are present in the A2 peptide and are believed to be associated with the property of activation. Ser291 and Glu297 contribute to a recognition motif that allows intestinal mucus or elastase purified from intestinal mucus to cleave the A2 peptide between Thr295 and Gly296, decreasing the length of the A2 peptide by two residues (16, 19). The modified A2 peptide is then able to interact with the Stx2B subunit pentamer, producing the activatable phenotype (19). Interestingly, the stx_2d operons described by Piérard et al. (26) (stx_{2d EH250}) also possess Ser291 and Glu297; however, these residues are adjacent to several other amino acid substitutions, which apparently mitigate the property of activation by intestinal mucus. Hence, Stx2d EH250 toxins are not activatable (A. R. Melton-Celsa and A. D. O'Brien, personal communication).

The aims of this study were (i) to examine the distribution of stx₂ subtypes in livestock ruminant animals, (ii) determine if activatable stx_2d operons were present, (iii) characterize stx_2d isolates and operons, and (iv) determine if stx_2d operons were borne on bacteriophages. We have demonstrated the presence of the activatable stx_2d genotype in food and livestock sources and have shown by nucleotide sequencing that this stx₂ subtype may be clearly distinguished from the stx_{2d EH250} subtype. We have further demonstrated that some stx_2d operons are carried by inducible bacteriophages that may be propagated in E. coli K-12.

MATERIALS AND METHODS

Isolation of STEC.

STEC strains were isolated from food and livestock sources following enrichment broth culture, stx-specific PCR, and hydrophobic grid membrane filtration (20). The virulence determinants of each isolate were determined by a multiplex PCR for stx₁, stx₂, eae, and ehxA (24). Isolates carrying stx_2d were kindly serotyped by Roger Johnson, Health Canada, Guelph, Ontario, Canada.

stx₂ characterization and subtyping.

STEC isolates that carried stx₂ were further characterized by PCR-RFLP analysis to determine the presence of stx₂ subtypes. Primers GK5 and GK6 (30) were used to amplify the stx_2B subunit gene. Amplicons were digested with FokI and HaeIII to discriminate stx₂ from stx_2c subtype genes. Use of restriction enzymes NciI and RsaI for PCR-RFLP analysis enabled classification of stx_2c genes as stx_{2c vha}, stx_{2c vhb}, or stx_{2d EH250} (26). Isolates encoding stx_{2c vha} and stx_{2c vhb} were further examined to identify genes of this type which did not possess a PstI site in the 3′ region of the stx_2A subunit gene. stx_{2c vha} and stx_{2c vhb} operons without PstI sites were sequenced in the 3′ region of the stx_2A subunit gene. Putative stx_2d genes showed identity to the sequence of activatable stx_2d1 carried by O91:H21 strain B2F1 (11, 16).

DNA sequencing.

Complete stx_2d operons were amplified with the primers 5′-GATGGCGGTCCATTATC-3′ (25) and 5′-ACTGAATTGTGACACAGATTA-3′ (A. R. Melton-Celsa, personal communication). Both strands of each stx_2d operon amplicon were completely sequenced. Sequencing was performed with a Big Dye Cycle Sequencing kit following the instructions of the manufacturer and an ABI 377 fluorescent sequencer (Applied Biosystems, Foster City, Calif.) at the Australian Genome Research Facility, University of Queensland, Brisbane, Australia. Nucleotide sequence alignments were performed by the using BLAST program (http://www.ncbi.nlm.nih.gov/BLAST/) (1). The identities of any two sequences were compared by the using the BLAST 2 SEQUENCES program (http://www.ncbi.nlm.nih.gov/blast/bl2seq/bl2.html) (34). The predicted amino acid alignments presented in Fig. 1 and 2 were prepared by using the CLUSTALW program (version 1.8) at the Baylor College of Medicine Search Launcher (http://www.dot.imgen.bcm.tmc.edu:9331/multi-align/multi-align.html) and the BOXSHADE program (version 3.21; http://www.ch.embnet.org/software/BOX_form.html).

FIG. 1.

Predicted amino acid sequences of the stx_2d A-subunit genes from isolates identified in this study aligned with the published sequences of Stx2d1 from STEC isolate B2F1 (11, 17) and Stx2 (GenBank accession no. AB035143). Dots indicate residues identical to those in Stx2d1. Residues Ser291 and Glu297, which distinguish the Stx2d1A-subunit carboxy terminus from that of Stx2A, are boldface and underlined.

FIG. 2.

Predicted amino acid sequences of the stx_2d B-subunit genes from isolates identified in this study aligned with the published sequences of Stx2d1 from STEC isolate B2F1 (11, 17) and Stx2 (GenBank accession no. AB035143). Dots indicate residues identical to those of Stx2d1. Residues Asn16 and Asp24, which distinguish the Stx2d1B subunit from that of Stx2B, are boldface and underlined.

Immunoblot detection of Stx expression from isolates with activatable stx_2d.

Expression of stx_2d operons was assayed by immunoblotting (2). Briefly, STEC isolates were patch inoculated onto nitrocellulose membranes placed on an agar plate surface. The production of Stx was determined by using rabbit anti-Stx antibodies reactive with all major Stx types and variants. Stx antisera were kindly provided by Roger Johnson, Health Canada.

Isolation and characterization of stx_2d bacteriophages.

stx_2d isolates were treated with mitomycin C (1 μg/ml) to induce the bacteriophage lytic cycle and production of phage. Lysed culture supernatants were filtered through a 0.45-μm-pore-size filter to ensure a cell-free lysate. The bacteriophage titers in the cell-free lysates were determined by serial dilution and propagation on E. coli Q358 (supE hsdR φ80^r recA⁺) (13). Single phage plaques were picked from appropriately diluted indicator plates, resuspended in 100 μl of sterile H₂O, and confirmed to carry stx₂ by PCR of the plaque supernatant. Several purified single plaques from each isolate were confirmed to carry stx_2d by PCR with primers GK5 and GK6, followed by RFLP analysis of the PCR amplicons.

Nucleotide sequence accession numbers.

The sequences of the complete stx_2d operons from strains EC152a, EC173b, EC604a, EC782, EC1586, EC1720a, and EC1871 have been submitted to GenBank and given the accession numbers AF500187, AF500190, AF500192, AF500193, AF500188, AF500189, and AF500191, respectively.

RESULTS

stx₂ genotypes among STEC isolates from livestock sources.

STEC isolates (n = 311) possessing stx₂ were isolated from a variety of food and livestock sources. Each isolate was then characterized to designate a specific stx₂ subtype (Table 1). The dominant stx₂ subtype found in STEC strains of ovine origin was stx_{2d EH250}, first described by Piérard et al. (26). A total of 79 of 103 (77%) isolates from lamb meat had the stx_{2d EH250} genotype, while 30 of 38 (79%) isolates from sheep carcasses and feces were stx_{2d EH250}. In comparison, classical stx₂ was dominant among STEC strains isolated from beef cattle (26 of 51; 51%) and dairy cattle (33 of 56; 59%) sources. STEC strains isolated from ground beef samples showed greater heterogeneity in stx₂ genotypes, in which stx₂ (29%), stx_{2d EH250} (21%), stx_{2c vha} (11%), and stx_{2c vhb} (11%) were characterized.

TABLE 1.

Genetic subtypes of stx₂ present in STEC isolates of different animal or meat origins

stx2 subtype	% STEC isolates from different sources possessing alternative stx2 genetic subtypesa
	Lamb meat (n = 103)	Sheep (carcasses and feces) (n = 38)	Ground beef (n = 63)	Beef cattle (carcasses and feces) (n = 51)	Dairy cattle sources (n = 56)
stx2	2	3	29	51	59
stx2 and stx2c vha	0	0	0	14	4
stx2 and stx2c vhb	0	0	3	2	11
stx2 and stx2d EH250	3	0	0	0	0
stx2c vha	1	8	11	22	5
stx2c vhb	2	0	11	4	7
stx2d EH250	77	79	21	6	2
stx2c vha and stx2c vhb	1	0	0	2	2
stx2d EH250 and stx2c vha	1	0	0	0	0
stx2d EH250 and stx2c vhb	1	0	0	0	0
Negativeb	8	0	16	0	11
Otherc	5	11	10	0	0

^aPercentages are rounded to the nearest integer value.

^bPrimers GK5 and GK6 did not amplify the correct amplicon from the stx_2B gene.

^cPrimers GK5 and GK6 amplified amplicons of the correct size from the stx_2B gene; however, the amplicons did not digest in a manner consistent with the anticipated RFLP scheme.

Presence of stx_2d in STEC.

Isolates possessing stx₂ were further characterized for the possession of putative stx_2d operons. By PCR-RFLP analysis and preliminary sequencing of the DNA, 12 STEC isolates were found to possess stx₂ operons, with the predicted amino acid sequences of the stx_2A subunit showing carboxy-terminal modifications specific for stx_2d. Seven putative stx_2d operons from isolates that represented the diversity of isolates by original source of isolation and the stx_{2c vha}, stx_{2c vhb}, or non-stx_{2c vha} and stx_{2c vhb} classification were completely sequenced. The DNA sequences of the stx_2d operons from isolates EC152a and EC604a were identical. The DNA sequences of the other five isolates were each unique and conformed with the PCR-RFLP analysis characterization for each stx_2d operon. Isolates identified as carrying stx_2d are listed in Table 2.

TABLE 2.

Characterization of STEC isolates with the stx_2d genotype

Isolate no.	Serotype	Isolate source	Virulence factor genotypea			stx2B subtypeb	Phage recovered
			stx	eae	ehxA
152a	O?:H29	Beef cattle feces	2	−d	−	2	Non-stx
173b	O174:H21	Beef cattle feces	2	−	−	vhb	Non-stx
604a	O2:H29	Dairy farm	2	−	−	2	Non-stx
782	O?:NMc	Beef cattle feces	2	−	−	vha	Non stx
1564b	O1:H20	Ground beef	2	−	+	vhb
1585	O174:H21	Dairy farm	2	−	−	vhb	Non-stx
1586	O174:H8	Ground beef	1 and 2	−	+	2	stx1
1662a	O174:H21	Ground beef	2	−	−	vha	stx2d
1720a	O174:H21	Ground beef	2	−	−	vha	stx2d
1871a	O?:H11	Dairy farm	2	−	+	vhb	Non-stx
1995a	O174:H8	Ground beef	1 and 2	−	+	vhb	stx1
2062a	O8:H19	Lamb meat	2	−	−	vhb

^aSTEC isolates were screened for the virulence factors stx (Stx genes stx₁ and stx₂), eae (E. coli attaching-and-effacing gene), and ehxA (enterohemolysin gene) by multiplex PCR (see Materials and Methods).

^bThe genetic subtype of the toxin B subunit gene for each stx_2d operon was classified by RFLP analysis and nucleotide sequencing (see Materials and Methods).

^cNM, nonmotile.

^d−, negative.

Comparison of the predicted amino acid sequences of the Stx2A subunit with the Stx2d sequence from O91:H21 strain B2F1 (11, 17) indicated that all the stx_2d operons encoded Stx2A with Phe291Ser and Lys297Glu amino acid substitutions, characteristic of activatable Stx2d (17) (Fig. 1). Comparison of the predicted amino acid sequences of the B subunits also showed Asp16Asn and Asp24Ala substitutions characteristic of Stx2dB subunits. The one exception was EC1586, in which Asp24 was conserved in the Stx2dB subunit (Fig. 2). The stx_2dB subunit genes of isolates EC152a, EC604a, and EC1586 did not conform to the stx_{2c vha} or stx_{2c vhb} RFLP designation that has been typical of other stx_2d operons.

The nucleotide sequence of the stx_2d operon from EC782 was identical to that of stx_2d1 (11, 16). The identical nucleotide sequences of the stx_2d operons from EC152a and EC604a differed by one nucleotide from the sequence of stx₂-NV206 (3). Pairwise comparison and alignments of the stx_2A and stx_2B genes from EC173b, EC1586, EC1720a, and EC1871a with homologs from the activatable operons stx_2d1, stx_2d2, and stx₂-NV206 (3, 11, 16) showed 98 to 100% identities and indicated that the new stx_2d operons comprised mosaic sequences.

stx_2d bacteriophage characterization.

For each of the 12 stx_2d-containing isolates identified, the bacteriophage lytic cycle was induced with mitomycin C. EC1564b and EC2062 cultures did not lyse in response to mitomycin C induction, and their culture supernatants did not show plaques on indicator plates. Phage plaques were identified from the induced culture lysates of all stx_2d-carrying isolates except EC1564b and EC2062a. The phages from EC152a, EC173b, EC604a, EC782, EC1585, and EC1871a did not carry stx operons, while those from EC1586 and EC1995 were characterized as carrying stx₁. The stx₂ phages φ1662 and φ1720a were induced from EC1662 and EC1720a, respectively. PCR-RFLP analysis specific for the stx_2B subunit genes of φ1662 and φ1720a showed that the stx₂ operons of these phages corresponded to the stx_2d operons sequenced from their respective STEC host isolates. RFLP analysis of φ1662 and φ1720a genomic DNA with the AvaI restriction enzyme indicated that these two phages were related; however, digestion of the DNA samples was complicated by contaminating nuclease activated during enzyme incubation (Fig. 3).

FIG. 3.

RFLP analysis of stx_2d bacteriophages. Bacteriophage genomic DNA was isolated, digested with AvaI, and electrophoresed. Lanes: 1, undigested φ1720a; 2, φ1720a; 4, undigested φ1662a; and 5, φ1662a. Lane 3, DNA size markers of 23.1, 9.4, 6.6, 4.4, 2.3, and 2.0 kb.

Stx expression from stx_2d strains.

Eleven stx_2d-containing isolates were examined for expression of Stx by immunoblotting (Fig. 4). Isolates EC1586 and EC1995a carried stx₁ genes, in addition to stx_2d, such that Stx expression from these isolates is the cumulative expression of both stx₁ and stx_2d. All other isolates examined carried only stx_2d. Stx expression determined by the diameter and intensity of the zone of staining was variable between isolates. Of particular note were the different levels of Stx2d expression by EC1662a and EC1720a, isolates that were both serotype O174:H21, whose stx_2B subunit genes had vha RFLP patterns, and that produced phages stx_2d possessing similar genome RFLP profiles.

FIG. 4.

Immunoblot analysis of Stx expression from stx_2d-containing isolates. Equivalent inocula of stx_2d-containing isolates were grown on nitrocellulose membranes placed on tryptic soy agar plates and grown overnight. The Stx expressed by bacterial colonies was captured on a capture membrane precoated with rabbit anti-stx antibodies. The capture membrane was probed with alkaline phosphatase-labeled rabbit anti-mouse immunoglobulin G. Triplicate inocula were made for each isolate, as indicated.

DISCUSSION

In Australia, STEC strains that do not possess the locus of enterocyte effacement are the major causes of hemolytic-uremic syndrome and bloody diarrhea (9). We have begun to characterize the stx₂ genotypes of STEC strains isolated from food-producing animal sources to determine if these stx₂ operons are characterized by specific properties. STEC strains possess an array of stx₂ genotypic variants (26, 32, 36); however, the roles of particular genotypic variants in human pathogenesis are not clear. We determined that STEC strains isolated from sheep or sheep meat sources most frequently carry stx_{2d EH250}, while STEC strains isolated from cattle sources most frequently carry stx₂ or stx_2c variant stx_{2c vha} or stx_{2c vhb}. While cattle have been known to be a significant reservoir of STEC isolates, recent studies have also shown that stx_{2d EH250} is routinely associated with ovine STEC isolates (8) (K. S. Gobius et al., unpublished data). The predominance of stx_{2d EH250} among ovine STEC isolates compared with its prevalence among bovine STEC isolates suggests that particular stx₂ subtypes may be associated with specific ruminant species.

Using screening by PCR-RFLP analysis, we were able to identify activatable stx_2d variants in STEC isolates that were predominantly from bovine sources. Partial sequencing of the stx₂ operons from 12 putative stx_2d isolates further confirmed that the operons were consistent with the presence of stx_2d. Further nucleotide sequencing of seven complete stx₂ operons confirmed that they encoded the activatable stx_2d genotype.

Each STEC isolate carrying activatable stx_2d was recovered from a separate sampling location or on independent sampling dates, ensuring that the isolates were not clonal representatives of the same strain. The nucleotide sequence polymorphisms of the stx_2d operons present in all isolates (except EC152a and EC604a, for which the stx_2d operons were identical; data not shown) support the independent origins of these isolates. Similarly, the diversity of serotypes among the strains carrying stx_2d also supports the conclusion that the isolates are not related. While four of the isolates were serotype O174:H21, the stx_2d operons carried by each of the isolates were differentiated by RFLP analysis and/or nucleotide sequencing.

All the serotypes that carried stx_2d in this study have previously been isolated from livestock sources in other countries, including Argentina, Brazil, Canada, France, Germany, New Zealand, and the United Kingdom (2, 4, 6, 7, 23, 28, 29, 31, 39); however, none of these serotypes have previously been associated with stx_2d. Isolates of serotypes O1:H20, O2:H29, O174:H8, and O174:H21, which have been identified to carry stx_2d, have previously been isolated from healthy or symptomatic humans in a variety of countries (14, 15, 27, 33).

Bertin et al. (3) recently described a novel stx₂ subtype, stx₂-NV206, present in STEC isolates of serotype O6:H10 isolated from a healthy cow. stx₂-NV206 operons also possess Ser291 and Glu297; however, the stx_2B gene of this subtype does not have restriction sites characteristic of stx_{2c vha} or stx_{2c vhb} (which have previously been associated with activatable stx_2d operons). In our study, the identical stx_2d operons carried by isolates EC152a and EC604a differed by a single nucleotide from the sequence of stx₂-NV206. Similarly, the sequence of stx_2d from EC782 was identical to that of stx_2d1 from B2F1, originally isolated in North America. The occurrence of identical or nearly identical activatable stx_2d operons in bovine STEC isolates from different global locations is intriguing. Their presence in the geographically separated locations of Australia, France, and North America suggests both conservation of these stx₂ variants and the possibility of their global dissemination.

In contrast to the conserved stx_2d variant operons present in isolates EC152a, EC604a, and EC782, we also identified stx_2d operons whose sequences did not show complete identity to those of previously described stx_2d operons. The stx_2d operons from EC173b, EC1586, EC1720a, and EC1871a showed nucleotide sequence polymorphisms in mosaic blocks of sequence. The mosaic structures of stx bacteriophage genomes, gained through bacteriophage genome shuffling, have been understood for some time (10, 37). However, recombination in stx operons is more difficult to detect by sequence analysis if recombination occurs between identical sequences. Our data provide evidence that alternative activatable stx_2d operons have arisen through recombination of independent stx₂ sequences. Such stx₂ operon recombination may account for the increasing number of stx₂ variants that continue to be detected and described. Recombination in stx₂ operons, coupled with bacteriophage lambda recombination and mosaicism, may create a rich genetic pool of stx phages capable of affecting horizontal gene transfer within the broad population of members of the family Enterobacteriaceae.

stx genes are bacteriophage borne or are associated with defective prophages (21). The stx_2d1 operon from strain B2F1 has been shown to be carried by an inducible bacteriophage that formed small turbid plaques on E. coli K-12 strain DH5α (35). We were able to induce and propagate stx_2d bacteriophages from only 2 of 12 stx_2d-positive STEC isolates. Notwithstanding the limited number of propagating stx_2d phages detected in our study, by infection of an E. coli K-12 indicator strain, we have confirmed that stx_2d operons are potentially transferred by bacteriophage induction and infection of new enterobacterial hosts. In contrast to the work of Teel et al. (35), we did not detect stx phages by induction from strain EC782, which was shown to possess an stx_2d operon identical to stx_2d1. Since the stx_2d bacteriophages were propagated on an E. coli K-12 strain, it is possible that other stx_2d phages may have been induced from additional STEC isolates but that such a phage(s) was unable to propagate on the specific indicator strain used or readily formed lysogens (unable to be detected by plaque formation). James et al. (12) found that the stx phage φ24B::Kan was capable of infecting only a minority of wild E. coli strains tested. However, of the strains able to be infected, the majority were susceptible to lysogeny by this phage. Alternative explanations for the detection of a small number of stx_2d phages following induction in our study may be (i) that the stx_2d operons in these isolates are carried on prophage remnants no longer capable of lytic induction or (ii) stx_2d phages were induced from these STEC isolates and rapidly formed Q358 lysogens so that detection by plaque formation was not possible.

The two stx_2d phages (φ1662a and φ1720a) successfully induced from isolates EC1662a and EC1720a were related when they compared by using phage genome-specific RFLP analysis. The apparent nuclease contamination of phage DNA may be similar to that observed with plasmid DNA isolated from various STEC strains (5). Given that both phages were induced from isolates of the same O174:H21 serotype and that the stx_2d operons showed RFLP patterns characteristic of the stx_{2c vha} genotype, it was surprising that the level of Stx2d expression by EC1720a appeared to be less than that by EC1662a. Recently, Wagner et al. (38) have shown that stx₂ transcription is regulated by the late promoter p_R′ of lambda phages, such that expression of Stx2 is coordinated with induction of the phage lytic cycle. Therefore, we conclude that transcription from p_R′ in φ1662a may be greater than that from p_R′ in φ1720a, leading to a higher level of Stx2d expression by EC1662a.

In conclusion, we have identified the activatable stx_2d genotype in STEC strains isolated from food-producing livestock sources. The STEC isolates carrying stx_2d consisted of multiple serotypes, and the nucleotide sequences of the stx_2d operons varied. In at least two isolates stx_2d was carried by inducible stx phages that could infect and propagate on an E. coli K-12 strain, demonstrating the potential for the horizontal transfer of stx_2d. It has been suggested previously that STEC strains involved in human infection, which carry stx_2d, may be more virulent due to mucus activation of the toxin. The increased virulence of the Stx2d toxin may compensate for the absence of other virulence attributes such as expression of the attaching-and-effacing gene (eae) commonly found in other enterohemorrhagic E. coli strains (17). Therefore, surveillance for STEC strains expressing the activatable Stx Stx2d in human illness may indicate the significance of this toxin subtype to human health.