Immunological Characterization of Escherichia coli O157:H7 Intimin γ1

Abstract

Portions of the intimin genes of Escherichia coli O157:H7 strain E319 and of the enteropathogenic E. coli O127:H6 strain E2348/69 were amplified by PCR and cloned into pET-28a(+) expression vectors. The entire 934 amino acids (aa) of E. coli O157:H7 intimin, the C-terminal 306 aa of E. coli O157:H7 intimin, and the C-terminal 311 aa of E. coli O127:H6 intimin were expressed as proteins fused with a six-histidine residue tag (six-His tag) in pET-28a(+). Rabbit antisera raised against the six-His tag-full-length E. coli O157:H7 intimin protein fusion cross-reacted in slot and Western blots with outer membrane protein preparations from the majority of enterohemorrhagic and enteropathogenic E. coli serotypes which have the intimin gene. The E. coli strains tested included isolates from humans and animals which produce intimin typesα (O serogroups 86, 127, and 142), β1 (O serogroups 5, 26, 46, 69, 111, 126, and 128), γ1 (O serogroups 55, 145, and 157), γ2 (O serogroups 111 and 103), and ɛ (O serogroup 103) and a nontypeable intimin (O serogroup 80), results based on intimin type-specific PCR assays. Rabbit antisera raised against the E. coli O157:H7 C-terminal fusion protein were much more intimin type-specific than those raised against the full-length intimin fusion protein, but some cross-reaction with other intimin types was also observed for these antisera. In contrast, the monoclonal antibody Intγ1.C11, raised against the C-terminal E. coli O157 intimin, reacted only with preparations from intimin γ1-producing E. coli strains such as E. coli O157:H7.

Escherichia coli O157:H7 is associated with hemorrhagic colitis, thrombotic thrombocytopenic purpura, and hemolytic-uremic syndrome in humans (21, 22). In addition to producing Shiga-like (Vero) toxin and enterohemolysin (6), E. coli O157:H7 has been shown to attach to the cytoplasmic membranes of intestinal epithelial cells, to efface their microvilli, and to cause actin to accumulate beneath sites of bacterial attachment (8). These features are shared with several other enterohemorrhagic E. coli (EHEC) serotypes and members of the enteropathogenic E. coli (EPEC) group (23, 24, 30).

The eae gene, which has been shown to be necessary for attaching and effacing activity, encodes a 94- to 97-kDa outer membrane protein (OMP) which is termed intimin (20). This gene is located in a chromosomal pathogenicity island also known as the locus of enterocyte effacement (LEE) (9). The entire nucleotide sequences of the 35-kbp LEE of EPEC strain E2348/69 (O127:H6) and the 43-kbp LEE of E. coli O157:H7 strain EDL933 have been determined (11, 36, 46). In addition to encoding intimin, the LEE encodes a number of other proteins which are necessary for intimate attachment of these bacteria to epithelial cells, such as proteins which are part of a type III secretion system (Sep and Esc proteins), the translocated intimin receptor (Tir), CesT (a Tir chaperone protein), and E. coli secreted proteins (EspA, EspB, and EspD).

Analysis of the nucleotide sequences of the intimin genes from different EHEC and EPEC strains has shown a high degree of homology in the 5′ two-thirds of the genes and a significant degree of heterogeneity in the 3′ one-third of the genes (2, 5, 20, 26, 33, 35, 45). Gannon et al. (16) identified five variants of the eae gene in E. coli strains from human and animal sources by examining the restriction fragment length polymorphisms (RFLP) of PCR products obtained from the amplified 5′ conserved region of the gene. Similarly, Boerlin et al. (7) identified six variants of this E. coli gene by another RFLP-PCR approach, and Wieler et al. (45) identified four variants of the gene in Shiga-like-toxin-producing E. coli strains of bovine origin by intimin type-specific (TS) PCR assays. Adu-Bobie et al. (1) reported five distinct types of intimin among the E. coli strains based on TS-PCR assays that used oligonucleotide primers complementary to the 3′ end of specific intimin genes. They noted that among strains of the two phylogenetically defined subgroups of EPEC (clones 1 and 2), clone 1 strains produce a common intimin type, which they designated α, and clone 2 strains produce another common intimin type, which they designated β. They further noted that two strains of E. coli O86:H34 produce a distinct intimin type, which they designated δ, and that E. coli O157:H7 (H⁻) and EPEC O55:H7 strains produce a common and distinct intimin type, which they designated γ. Reid et al. (37) have also described a multiplex PCR assay for differentiation among intimin typesα, β, and γ, which is in agreement with the typing scheme of Adu-Bobie et al. (1).

Recently, Oswald et al. (33) have described a TS-PCR assay which identifies a fifth intimin variant, from E. coli serotype O103:H2, which they have called intimin ɛ. They have also reclassified certain intimin types based on the degree of nucleotide sequence homology and on RFLP-PCR profiles. In their scheme, intimin γ is divided into two subtypes, with intimin γ1 present in strains of EHEC serotypes O145:H⁻ and O157:H7(H⁻) and EPEC serotype O55:H7 and intimin γ2 present in strains of EHEC serotypes O86:H40 and O111:H8(H⁻) and EPEC serotypes O127:H40 and O128:H8(H⁻). Furthermore, the single representative of the intimin δ group (EPEC O86:H34) reported by Adu-Bobie et al. (1) was reclassified as intimin β2 by Oswald et al. (33) based on the similarity of the nucleotide sequence of this intimin type to that of intimin β1. The heterogeneity observed among E. coli intimin genes and their expressed proteins not only suggests distinct phylogenetic lineages for these EHEC and EPEC subgroups but also is likely to be important for the affinity of these adhesins to their receptors. Indeed, considerable variation has also been noted in the Tir genes among EHEC and EPEC subgroups (34). In addition, host immunity to these surface-exposed proteins produced by one E. coli strain may not provide protection against intestinal colonization by E. coli strains which bear distinct intimin or Tir types (10). Furthermore, the intimin or Tir type may also play a role in determining which region of the gut is colonized (44).

Several different E. coli intimin genes have been cloned, and recombinant intimin proteins have been expressed at a high level and purified (3, 4, 12, 13, 27, 28, 47). The antisera raised against these recombinant proteins have been used to examine the degree of immunological relatedness among intimins. These antisera have been helpful in the classification of intimin types (1, 3); however, cross-reactions with other intimin types are known to occur. Zhu et al. (47, 48) reported production of monoclonal antibodies to the intimin of E. coli O45 (typeβ), antibodies which reacted with a homologous intimin type from rabbit diarrheagenic E. coli type 1 (RDEC-1) but not with intimin type α from E. coli O127:H6 strain E2348/69. Unfortunately, the specificities of these monoclonal antibodies with respect to other intimin types were not reported.

In this report, we characterize a monoclonal antibody specific for intimin γ1 and describe the use of this antibody in specifically identifying intimin γ1-producing E. coli strains.

MATERIALS AND METHODS

Bacterial strains.

The bacterial strains used in this study are listed in Table 1. E. coli strains of human origin were kindly provided by M. Karmali (Health Canada, Laboratory Centre for Enteric and Zoonotic Diseases, Guelph, Ontario, Canada), M. Anand (Southern Alberta Provincial Health Laboratory, Calgary, Alberta, Canada), C. Gyles (University of Guelph, Guelph, Ontario, Canada), S. M. Scotland (Central Public Health Laboratories, London, England), and M. Richter (Northern Alberta Provincial Health Laboratory, Edmonton, Alberta, Canada). E. coli strains of porcine origin were supplied by C. Gyles (University of Guelph) and J. M. Fairbrother (University of Montreal, Saint-Hyacinthe, Quebec, Canada).

TABLE 1.

E. coli strains used in the study, intimin genotypes, and reactivities of rabbit antisera raised against recombinant intimin proteins with E. coli lysates

Strain	Serotype	Source	Pathotype	vt PCR resulta	eae PCR resultb	Intimin typec	HaeIII typeb	Reactivity to E. coli lysate of rabbit antiserum raised againstd:
								His-intO157	His-intO157C	His-intO127C
EC226	O86:NM	Human	EPEC	−	+	α	E	+++	+	+
EC990986	O127:NM	Human	EPEC	−	+	α	E	+	+	+++
E2348/69	O127:NM	Human	EPEC	−	+	α	E	+	−	++
EC990987	O142:H34	Human	EPEC	−	+	α	E	+w	−	+w
H19	O26:H11	Human	EHEC	+	+	β1	A	+++	−	−
EM88-3618(1)	O26:H11	Bovine	EHEC	+	+	β1	A	+++	+	−
EA01985-91	O26:H11	Bovine	EHEC	+	+	β1	A	+	−	−
EM88-4256	O26:H11	Bovine	EHEC	+	+	β1	A	+	−	−
41131	O26:H11	Human	EHEC	+	+	β1	A	+++	−	−
EC990983	O26:H11	Human	EPEC	−	+	β1	A	+++	−	−
H30	O26:NM	Human	EHEC	+	+	β1	A	+++	−	−
P86-4220	O45:NM	Porcine	VTEC	−	+	β1	A	+++	−	−
P87-4725	O45:NM	Porcine	VTEC	−	+	β1	A	+++	−	−
EC920142	O69:H11	Human	EHEC	+	+	β1	A	+++	++	−
EM90-2768	O111:H11	Bovine	EHEC	+	+	β1	A	++	−	−
EM88-4108	O111:H11	Bovine	EHEC	+	+	β1	A	+w	−	−
EC200055	O126:H2	Human	EPEC	−	+	β1	A	+++	−	−
H18	O128:NM	Human	EHEC	+	+	β1	A	+++	+	−
EC920234	O80:NM	Bovine	VTEC	+	+	NTe	A	++	−	−
EM88-3620	O5:NM	Bovine	EHEC	+	+	β1	D	+++	+	−
2340	O5:NM	Bovine	EHEC	+	+	β1	D	+++	−	−
5432	O103:H2	Human	EHEC	+	+	ɛ	C	++	−	−
5529	O103:H2	Human	EHEC	+	+	ɛ	C	++	−	−
9291	O103:H2	Human	EHEC	+	+	ɛ	C	++	+w	−
35280	O103:H2	Human	EHEC	+	+	ɛ	C	++	+w	−
EC322	O55:H7	Human	EPEC	−	+	γ1	B	+++	++	−
33264	O145:NM	Human	EHEC	+	+	γ1	B	+++	++	−
E319	O157:H7	Human	EHEC	+	+	γ1	B	+++	++	−
E321	O157:H7	Human	EHEC	+	+	γ1	B	++	++	−
H4420	O157:H7	Bovine	EHEC	+	+	γ1	B	+w	−	−
LRH1	O157:H7	Human	EHEC	+	+	γ1	B	+	+w	−
LRH2	O157:H7	Human	EHEC	+	+	γ1	B	+	+w	−
LRH6	O157:H7	Human	EHEC	+	+	γ1	B	+++	++	−
E32511	O157:NM	Human	EHEC	+	+	γ1	B	+	+w	−
CL8	O157:NM	Human	EHEC	+	+	γ1	B	++	+	−
43426	O103:H25	Human	EHEC	+	+	γ2	B	+w	−	−
EM87-1507	O111	Bovine	EHEC	+	+	γ2	B	−	−	−
44717	O111:H12	Human	EHEC	+	+	γ2	B	−	−	−
EC920018	O111:H8	Human	EHEC	+	+	γ2	B	+w	−	−
52050	O111:NM	Human	EHEC	+	+	γ2	B	−	−	−
K-12	Rough	Lab	E. coli	−	−	NAf	NA	−	−	−
EC990988	O119:H27	Human	EPEC	−	−	NA	NA	−	−	−
DAB	O139:H1	Porcine	VTEC	+	−	NA	NA	−	−	−
B2F1/3	O91:NM	Human	EHEC	+	−	NA	NA	−	−	−

^aSee Gannon et al. (16).

^bSee Gannon et al. (16).

^cSee Oswald et al. (33).

^dReactivity characterized as +++ (strong), ++ (moderate), + (low), +w (weak), or −, (none).

^eNT, nontypeable.

^fNA, not applicable.

PCR amplification, cloning, and nucleotide sequencing of eae genes.

Oligonucleotide primers used in the study were purchased from Canadian Life Technologies (Burlington, Ontario, Canada) and were synthesized according to the nucleotide sequences of the eae genes of E. coli CL8 (O157:H7) (accession no. Z11541) and E. coli E2348/69 (O127:H6) (GenBank accession no. M58154). In order to facilitate cloning of the PCR products, a BamHI restriction site was placed at the 5′ ends of the forward primers EO157 (5′-GGA TCC TTG TGG TGG AGC CAT AAC ATG), EO157C (5′-GGA TCC GGA CAG GTC GTC GTG TCT GCT), and EO127C (5′-GGA TCC GGC CAG GTC GTC GTG TCT GCT), and a HindIII restriction site was placed at the 5′ end of the reverse primer EAE18 (5′-AAG CTT GGT ACC AGC CTC GGG ATT GG). The EO157 forward primer includes the putative eae ribosomal binding and start sites.

For the cloning of eae genes, E. coli DNA was extracted and PCR amplification experiments were carried out as described previously (15). Briefly, amplifications were carried out with a GeneAmp PCR system 9600 V2.01 (Perkin-Elmer Cetus, Rexdale, Ontario, Canada). Ten nanograms of bacterial DNA was transferred into a 100-μl PCR mixture containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 2.0 mM MgCl₂, 1.0 μM each primer, 0.2 mM each deoxynucleoside triphosphate (Canadian Life Technologies), and 2.5 U of Taq DNA polymerase (Gibco/BRL, Gaithersburg, Md.). PCR amplification cycles consisted of 95°C for 15 s, 60°C for 15 s, and 72°C for 90 s for 35 cycles and a final extension cycle of 4 min at 72°C. The ca. 3,200-bp PCR product generated with oligonucleotide primers EO157 and EAE18 and the two 1,300-bp PCR products, one generated with oligonucleotide primers EO157C and EAE18 and the other generated with primers EO127C and EAE18, were cloned into plasmid pCR II (Invitrogen Corp., San Diego, Calif.) by means of the T-A cloning procedure described by the manufacturer. Plasmid-insert ligation mixtures were used to transform competent E. coli DH5α cells (Canadian Life Technologies). Recombinant plasmids were isolated by an alkaline lysis procedure, treated with the Wizard DNA clean-up system (Fisher Scientific, Whitby, Ontario, Canada) as described in the manufacturer’s instructions, and then purified by passage through a Sephadex G-50 spin column (Pharmacia Biotech Inc., Baie d’Urfé, Quebec, Canada) as described by Sambrook et al. (39). The pCR II constructs were then digested with BamHI and HindIII, and the fragments of interest were subcloned into pET 28a(+) (Novagen, Madison, Wis.) in accordance with the manufacturer’s protocol.

The pET 28a(+) ligation mixtures were used to transform competent E. coli DH5α cells. The pET 28a(+) plasmids found to contain insert DNA were then used to transform competent E. coli BL21(DE3) cells for use in expression experiments. The pET 28(+) plasmid constructs containing insert DNA were designated pET-eaeO157, pET-eaeO157C, and pET-eaeO127C (Table 2).

TABLE 2.

Sizes of intimin-encoding DNA fragments and of expressed intimin fusion proteins

Plasmid construct	Size of fragment (bp)	Fusion protein	Protein size (kDa)	E. coli strain
pET-eaeO157C	918	His-intO157C	40	E319
pET-eaeO157	2,806	His-intO157	100	E319
pET-eaeO127C	932	His-intO127C	40	E2348/69

Expression and purification of recombinant proteins.

Recombinant intimin genes in pET 28a(+) were expressed and purified according to the manufacturer’s instructions (Novagen). Briefly,E. coli BL21(DE3) cells harboring pET plasmid constructs were grown in Terrific broth (TB) (12 g of tryptone, 24 g of yeast extract, 4 ml of glycerol, 100 ml of K-phosphate buffer, each per liter) supplemented with kanamycin (50 mg per liter) for 3 h at 37°C. Isopropyl-β-d-thiogalactopyranoside (IPTG) was then added to the cultures to a final concentration of 1 mM, the cultures were incubated for an additional 3.5 h, and the cells were harvested by centrifugation at 6,000 × g for 10 min. The bacterial pellets were resuspended in 30 ml of binding buffer (5 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl [pH 7.9]) and placed on ice. The bacterial cells were then disrupted with a Sonifier Cell Disruptor 200 (Branson) fitted with a microtip (settings, 40% duty cycle and output 6) in four 15-s bursts separated by 15-s intervals. These steps were repeated three more times, and the resulting lysate was spun at 20,000 × g for 20 min. The pellet obtained was suspended in 5 ml of binding buffer containing 8 M urea and was incubated on ice for 1 h. This mixture was then spun at 39,000 × g for 20 min, and the supernatant was recovered and filtered through a 0.45-μm-pore-size membrane (Nuclepore, Cambridge, Mass.). After this, the filtrate was passed through a His-binding resin column, and the column was washed with the following solutions: 25 ml of binding buffer, 7.5 ml of wash buffer 1 (30 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl [pH 7.9]), and 7.5 ml of wash buffer 2 (60 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl [pH 7.9]). The protein was then removed from the column with 15 ml of elution buffer (500 mM imidazole, 0.5 M NaCl, 20 mM Tris-HCl [pH 7.9]). Preparations were dialyzed at 4°C against 0.1 M phosphate-buffered saline (pH 7.4) with six changes of buffer over a 3-day period by using dialysis tubing with a 12- to 14-kDa exclusion limit (Ultrapure; Canadian Life Technologies). The protein preparations were analyzed by electrophoresis on sodium dodecyl sulfate-10% polyacrylamide gel electrophoresis (SDS-10% PAGE) minigels (Bio-Rad).

Intimin gene typing.

Intimin types were determined on the basis of the HaeIII restriction endonuclease digestion profiles of eae PCR products as described by Gannon et al. (16), with some modification. In brief, PCR products of the eae gene were amplified using primer pair AE9-AE14 (16) and were purified using the Concert Rapid PCR purification kit (Gibco/BRL) according to the manufacturer’s instructions. Five micrograms of the purified DNA was digested with 10 U of HaeIII, and the digests were analyzed by electrophoresis on 1.2% agarose gels containing 0.25 μg of ethidium bromide per ml. The digest patterns obtained from these PCR products were assigned to eae PCR-HaeIII RFLP profiles A to E. The intimin type (α, β1, β2 [δ], γ1, γ2, or ɛ) was determined with the oligonucleotide primer sets and methods described by Oswald et al. (33), which are referred to as the TS-PCR assays in Results.

Preparation of antisera.

Approximately 150 μg of each purified six-His-tagged intimin fusion protein (His-intO157, His-intO157C, and His-intO127C) was homogenized 1:1 with incomplete Freund’s adjuvant (Canadian Life Technologies). A 1-ml volume of this mixture was injected into two subscapular and two intramuscular sites of Flemish giant × French lop hybrid rabbits (Biosciences Animal Services, University of Alberta, Edmonton, Canada). Booster doses of this vaccine preparation were given again at 4 to 6 weeks after the first vaccination, and then up to three booster vaccinations were given every 2 weeks thereafter. Two weeks after the last vaccination, the rabbits were exsanguinated and the sera were collected by centrifugation. The antisera were absorbed against heat-killed E. coli BL21(DE3) cells by incubation of the sera with the cells for 4 h at room temperature, followed by incubation of the mixture at 4°C with shaking overnight. This suspension was then spun at 12,000 × g for 10 min to remove the bacteria, and the supernatant was collected and stored at − 70°C.

Hybridoma culture supernatants were prepared by Rita Bigham, S. Druhan, and G. Tiffin and were a kind gift from W. G. Yates of the Animal Diseases Research Institute, Canadian Food Inspection Agency, Lethbridge, Alberta, Canada. The immunoglobulin (Ig) type and subtype of monoclonal antibodies derived from hybridomas were determined by using a mouse monoclonal antibody isotyping kit (Pierce, Rockford, Ill.) (19).

Preparation of E. coli OMPs and lysates.

OMPs were prepared from E. coli strains as described by Agin and Wolf (3), with modification. Briefly, E. coli strains were grown overnight at 37°C in Luria-Bertani broth (LB). These cultures were used for inoculation of 10 ml of one or more of the following media at a ratio of 1:100, inoculum to medium: minimal essential medium (MEM), Dulbecco’s modified Eagle’s medium (DMEM), buffered peptone water (BPW) (1% peptone water [Lab M balanced peptone water no. 1] supplemented with 3.5 g of Na₂HPO₄/liter and 1.5 g of KH₂PO₄/liter), M9 minimal medium (supplemented with 44 mM NaHCO₃, 0.4% glucose, and 0.1% Casamino Acids), rich broth (RB; LB containing 0.2% glucose), TB, and LB. The cultures were grown at 37°C with shaking for 2 to 24 h. Bacterial concentrations were adjusted to an optical density at 600 nm (OD₆₀₀) of 1.0 for OMP preparations and an OD₆₀₀ of 0.2 for simple bacterial lysates, and a 1-ml sample of the concentrated or diluted bacterial growth was harvested by centrifugation at 2,000 × g at 4°C for 10 min. The pellets were resuspended in 10 mM HEPES buffer (pH 7.4) and centrifuged again. This washing step was repeated once more, and then the pellet was resuspended in 1 ml of 10 mM HEPES. The cells were placed on ice and disrupted with a Sonifier Cell Disruptor 200 (Branson) fitted with a microtip probe (settings, 40% duty cycle and output 6) for 40 s. For OMP preparations, debris was removed from the bacterial lysates by centrifugation at 9,000 × g at 4°C for 10 min. The supernatant was removed and centrifuged at 16,000 × g at 4°C for 30 min. The pellet obtained from this centrifugation step was resuspended in 400 μl of 10 mM HEPES (pH 7.4) containing 1% N-lauroylsarcosine (Sigma), and the mixture was incubated at room temperature with shaking for 30 min. Following this, OMPs were collected by centrifugation at 16,000 × g for 30 min at 4°C. The pellet was washed once in HEPES buffer without mixing, resuspended in 100 μl of this buffer, and stored at−20° C.

Slot and Western blot immunoassays.

For slot blot immunoassays, approximately 100 μl of each bacterial lysate was transferred onto nitrocellulose membranes by using a slot blot apparatus (Bio-Rad) according to the manufacturer’s instructions. For Western blot assays, 10 μg of bacterial OMPs was separated by electrophoresis using SDS-10% PAGE gels. Separated proteins were transferred from the gels onto nitrocellulose membranes with a Trans-Blot semi-dry transfer cell (Bio-Rad) by using Towbin’s transfer buffer (25 mM Tris, 192 mM glycine, 20% methanol [vol/vol] [pH 8.3]). Membranes were incubated overnight in the blocking solution, Tris-buffered saline (20 mM Tris-HCl, 500 mM NaCl [pH 7.5]) containing 5% (wt/vol) nonfat powdered milk. They were washed three times in Tris-buffered saline and then incubated in rabbit polyclonal antisera (diluted 1:200 in the blocking solution) or hybridoma supernatant fluid (diluted 1:10 in the blocking solution) for 2 h. After this, the membranes were washed three times for 15 min each time in Tris-buffered saline containing 2% nonfat milk powder. The membranes were incubated for 1 h in blocking solution containing horseradish peroxidase-conjugated goat anti-rabbit or goat anti-mouse IgG (Cedarlane Laboratories, Hornby, Ontario, Canada) at a dilution of 1:2,000, and then the membranes were washed three more times in blocking solution as described above. After this, the membranes were incubated in freshly prepared 4CN color development solution (Bio-Rad) for 10 to 30 min and then rinsed in deionized water to stop the reaction.

RESULTS

Typing of eae genes with TS-PCR primers and HaeIII digestion.

The eae PCR assay results with DNAs from representative E. coli strains are presented in Table 1. An 890-bp PCR fragment was produced with DNAs from 40 of the E. coli strains tested. DNAs from E. coli serotype O91:NM, from an EPEC O119 strain, from a porcine E. coli O139 edema disease strain, and from an E. coli K-12 strain were not amplified in the eae PCR assay.

Intimin types, as determined by TS-PCR amplification with eae-specific oligonucleotide primers (see Materials and Methods) and by HaeIII restriction endonuclease digestion patterns of eae PCR products, are also presented in Table 1. Two EPEC strains of E. coli serogroup O127 and one EPEC strain each of serogroups O86 and O142 had intimin α. All intiminα-producing E. coli strains also were of eae PCR-HaeIII RFLP profile E. Intimin type β1 was present in 16 of 40, or 40%, of the E. coli strains which possessed the eae gene. Intimin β1 was present in verocytotoxin-producing E. coli of serogroups O5 (two of two strains), O26 (six of six), O69 (one of one), O111 (two of five), and O128 (one of one), in the porcine postweaning diarrhea-producing E. coli strains of serogroup O45 (two of two), and in EPEC strains of serogroups O26 (one of one) and O126 (one of one). All E. coli strains with intimin typeβ1, except for the two strains of serogroup O5, were of eae PCR-HaeIII RFLP profile A. The E. coli serogroup O5 strains were of eae PCR-HaeIII RFLP profile D. The single E. coli strain of serogroup O80 which was not amplified by the intimin-TS-PCR assay also was of eae PCR-HaeIII RFLP profile A.

All strains of E. coli O157:H7(H⁻) (eight of eight), E. coli O145:H⁻ (one of one), and E. coli O55:H7 (one of one) possessed intimin type γ1, according to intimin-TS-PCR assays. One strain of serotype O103:H25 and four strains of serogroup O111 were of intimin type γ2. E. coli strains with intimin γ1 and those with intimin γ2 were of eae PCR-HaeIII RFLP profile B. As expected, the four E. coli serotype O103:H2 strains were of intimin type ɛ. These E. coli strains also all had the same unique eae PCR-HaeIII RFLP profile, profile C (Table 1).

Reactivity of intimin polyclonal antisera with E. coli lysates.

Rabbit antisera raised against His-intO157 reacted with lysates from all E. coli strains which produce intimin γ1 [E. coli O157:H7(H⁻), O145:NM, and O55:H7 strains]. However, among the E. coli O157:H7 strains, some strains showed low or weak reactivities (Table 1). These antisera also reacted strongly with lysates from other E. coli strains which possess other intimin types. Rabbit antisera raised against His-intO157C also reacted with all lysates from E. coli strains which produce the homologous intimin γ1, except E. coli O157:H7 strain H4420 (Table 1). However, reactions were also observed for these antisera with lysates from E. coli strains which produce other intimin types. In contrast, rabbit antisera raised against His-intO127C reacted only with lysates from E. coli strains which had intimin type α (EPEC strains of serogroups O86, O127, and O142).

Expression and purification of fusion proteins.

Portions of the eae gene of E. coli O157:H7 strain E319 and E. coli O127:H6 strain E2348/69 were expressed as proteins fused with the six-His tag of pET-28(+) expression vector systems. These constructs contained either the entire E. coli O157:H7 eae gene (pET-eaeO157), the 3′ ca. 918-bp segment of the E. coli O157:H7 eae gene (pET-eaeO157C), or the 3′ 933-bp segment of the E. coli O127:H6 strain E2348/69 eae gene (pET-eaeO127C) (Table 2).

The His-intO157 fusion protein was expressed as a 100-kDa protein (Fig. 1A, lane 2) from pET-eaeO157, and the His-intO157C (Fig. 1A, lane 5) and His-intO127C (Fig. 1A, lane 6) proteins were expressed as 40-kDa proteins from pET-eaeO157C and pET-eaeO127C, respectively. Although all of the His-intimin fusion proteins were of low solubility, they could be readily purified with 8 M urea (Fig. 1A, lanes 3, 6, and 9, respectively) to concentrations of 20 to 30 mg per liter of broth culture (data not shown).

FIG. 1.

Expression of His-intimin fusion proteins by pET-eaeO157, pET-eaeO157C, and pET-eaeO127C. Shown are stained SDS-PAGE gels (A) and corresponding Western blots with His-intO157 rabbit antisera (B) or mouse monoclonal antibody Intγ1.C11 (C). Lanes M, molecular weight markers (Bio-Rad); lanes 1 and 2, uninduced and induced E. coli BL21(DE3) harboring pET-eaeO157, respectively; lanes 3, purified His-intO157 protein; lanes 4 and 5, uninduced and induced E. coli BL21(DE3) harboring pET-eaeO157C, respectively; lanes 6, purified His-intO157C protein; lanes 7 and 8, uninduced and induced E. coli BL21(DE3) harboring pET-eaeO127C, respectively; lanes 9, purified His-intO127C protein.

Rabbit antisera raised against His-intO157 reacted with both the homologous fusion protein His-intO157 (Fig. 1B, lanes 2 and 3) and His-intO157C (Fig. 1B, lanes 5 and 6) and reacted weakly with the His-intO127C fusion protein in Western blots (Fig. 1B, lanes 8 and 9).

Reactivity of the monoclonal antibody.

The mouse monoclonal antibody Intγ1. C11 was of subclass IgG1 (data not shown). It reacted with His-intO157 (Fig. 1C, lane 2 and 3; Fig. 2, position H3) and His-intO157C (Fig. 1C, lanes 5 and 6; Fig. 2, position H4) intimin fusion proteins but did not react with His-intO127C in Western (Fig. 1C, lanes 8 and 9) and slot (Fig. 2) blots. While rabbit antisera raised against His-intO157 and His-intO157C cross-reacted with bacterial lysates from other EHEC and EPEC strains (Table 1), the monoclonal antibody reacted only with preparations from the intimin γ 1-producing E. coli strains of serotypes O157:H7(H⁻) (Fig. 2, positions A1 through B2), O145:NM (Fig. 2, position B3), and O55:H7 (Fig. 2, position B4) in slot blots.

FIG. 2.

Reactivities of monoclonal antibody Intγ1.C11 with bacterial lysates (see Materials and Methods) from E. coli strains in slot blot assays. The following are the position coordinates (A1 to H2) followed by the E. coli strain designation with the serotype or serogroup of the strain shown in parentheses: A1, E319 (O157:H7); A2, E32511 (O157:H⁻); A3, E321 (O157:H7); A4, CL8 (O157:H⁻); A5, LRH1 (O157:H7); A6, LRH2 (O157:H7); B1, LRH6 (O157:H7); B2, H4420 (O157:H7); B3, 33264 (O145:H⁻); B4, EC322 (O55:H7); B5, 2340 (O5:NM); B6, EM88-3613 (O5:NM); C1, H19 (O26:H11); C2, H30 (O26); C3, 41131(O26); C4, EAO1985-91 (O26); C5, EM90-2768 (O26); C6, EM88-4256 (O26); D1, EC920142 (O69:H11); D2, EM88-4108 (O111); D3, EM88-3620 (O5); D4, HI8 (O128); D5, P86-4220 (O45); D6, P87-4725 (O45); (E1, EC990983 (O26); E2, EC200055 (O126); E3, EC990986 (O127); E4, EC320 (O127:H6); E5, EC990987 (O142); E6, EC920234 (O80:NM); F1, 5432 (O103:H2); F2, 5529 (O103:H2); F3, 9291 (O103:H2); F4, 35280 (O103:H2); F5, EC920018 (O111:H8); F6, 52050 (O111:K58); G1, EM87-1507 (O111); G2, 43426 (O?:K?:H21); G3, 44717 (O111); G4, EC226 (O86:NM); G5, B2F1/3 (O91:NM); G6, EC990988 (O119); H1, DAB (O139:K82); H2, E. coli K12. Position H3, purified His-intO157; H4, purified His-intO157C; H5, purified His-intO127C; H6, uninoculated control.

Levels of intimin expression by E. coli O157:H7 strains in different media.

Intimin γ1 expression by E. coli O157:H7 and O157:H⁻ strains varied and was dependent on the culture medium and bacterial strain (Fig. 3A). All strains tested produced intiminγ1 in DMEM and MEM; however, intimin γ1 production was low or not detected in LB, RB, and TB (Fig. 3A). E. coli O157 strain E319 had high intimin γ1 production relative to that of other E. coli O157 strains in all media tested except for RB and TB. E. coli strain E32511 had low intimin γ1 production in MEM and DMEM, but its production of intimin γ1 was higher in BPW and M9 medium than those of other E. coli O157:H7 strains. Among the E. coli O157:H7 strains tested, E. coli O157:H7 strain LRH6 produced the highest level of intimin γ1 in MEM and DMEM but produced little or no intimin γ1 in the other media tested. E. coli O157:H7 strains LRH1, LRH2, and H4420 were low-intimin γ 1-producing strains. Intimin γ1 production by E. coli O157:H7 strain H4420 was detected only in MEM (Fig. 3A).

FIG. 3.

Reactivities of monoclonal antibody Intγ1.C11 in slot blot assays with bacterial lysates (see Materials and Methods) from E. coli O157:H7(H⁻) strains grown under different cultural conditions. (A) Strains H4420, LRH6, LRH2, LRH1, E32511, and E319 were grown in MEM, DMEM, BPW, LB, RB, and TB for 24 h at 37°C. Final cell concentrations were adjusted to an OD₆₀₀ of 0.2, and bacterial lysates were prepared and tested in slot blot assays. (B) E. coli O157:H7 strain E319 was grown in MEM, DMEM, BPW, LB, RB, TB, and M9 medium for 2, 4, 5, 6, and 24 h at 37°C. Final cell concentrations were adjusted to an OD₆₀₀ of 0.2, and bacterial lysates were prepared and tested in slot blot assays.

In MEM, DMEM, and M9 medium, intimin γ1 production by E. coli O157:H7 strain E319 was detected after 2 h of culture, was highest between 4 and 5 h of incubation, and did not increase after this (Fig. 3B). Intimin γ1 was detected after 3 h of incubation in BPW and LB. By contrast, very low levels of intimin γ1 production were detected at 24 h in RB and TB (Fig. 3).

DISCUSSION

In the present study, attempts were made to determine if rabbit polyclonal antisera raised against E. coli O157:H7 intimin would be useful in identifying intimin γ1-producing E. coli strains. Antisera raised against the entire intimin polypeptide of E. coli O157:H7 reacted not only with His-intO157 and His-intO157C fusion proteins on Western blots but also with His-intO127C as well as bacterial lysates from a large number of other intimin-producing E. coli strains. Similar broad-based reactivity with intimins has been recently reported by Batchelor et al. (4) for rabbit polyclonal antisera raised against a conserved portion of intimin from E. coli O157:H7. This result is not surprising, given that the N-terminal and central portions of the protein appear to be reasonably well conserved among intimin types, and it is likely that many epitopes are also shared among these bacterial proteins. It is interesting to note that some of the weakest reactions with these antisera were observed for bacterial lysates from E. coli O157:H7 strains which would presumably produce the homologous intimin type γ1. Consequently, it is possible that the reactions observed may provide more information about the level of intimin γ1 production by the E. coli O157:H7 strains than about antigenic differences among the intimin types.

The rabbit polyclonal antisera raised against the C-terminal portion of intimin γ1 (His-intO157C) appeared to be more type specific than the antisera raised against the full-length His-E. coli O157:H7 intimin fusion protein, but these antisera also reacted with bacterial lysates from E. coli strains which produce other intimin types (e.g., intimins α, β1, and ɛ) but not with intiminγ2-producing E. coli strains. This suggests that there may be common epitopes among some of these other types and intimin γ1. The polyclonal rabbit antisera raised against the C-terminal portion of intimin α (His-intO127C) was much more specific and reacted only with bacterial lysates from E. coli strains producing intimin α (EPEC O86, O127, and O142 strains).

In contrast to the rabbit polyclonal antisera, the mouse monoclonal antibodies to intimin γ1 reacted only with the homologous intimin antigens His-intO157 and His-intO157C and not with His-intO127C in Western (Fig. 1) and slot (Fig. 2) blots. The monoclonal antibody Intγ 1.C11 also reacted only with OMPs and bacterial lysates from E. coli strains which produce intimin γ1 [serotypes O157:H7(H⁻), O145:H⁻, and O55:H7] and did not react with OMPs or bacterial lysates from other eae PCR-positive E. coli strains. It is evident that these monoclonal antibodies are useful in the detection of intiminγ1-producing E. coli strains. In addition, intimin is reported to be highly immunogenic (25, 34), and it is possible that these monoclonal antibodies could be used in a blocking enzyme-linked immunosorbent assay to detect specific antibody responses to this portion of the protein.

Frankel et al. (12) reported that the C-terminal 150 amino acids and the cysteine residue at position 937 of intimin were necessary for the attachment of E. coli strain E2348/69 (O127:H6, intimin α) to HEp-2 cells. Recently, Gansheroff et al. (17) have also shown that rabbit polyclonal antisera raised against the C-terminal one-third of intimin γ1 block the binding of E. coli O157:H7 to HEp-2 cells. Therefore, it is possible that the monoclonal antibodies such as the one described in this study could prevent the attachment of intimin γ1-producing E. coli strains to epithelial cell lines and intestinal cells. If this is true, monoclonal antibodies such as this could be used to prevent intestinal colonization by E. coli in animals and humans. By way of analogy, monoclonal antibodies to the F5 (K99) adhesin have been used to prevent intestinal colonization by enterotoxigenic E. coli strains in calves (41). However, further study is required to determine whether monoclonal antibodies to intimin γ1 could prevent bacterial attachment to the intestine.

Recent work has shown that control of expression of the operons in the LEE pathogenicity islands of EPEC and EHEC is mediated by a complex regulatory cascade. For both EPEC and EHEC, LEE operons 1 through 3 encode components of the type III secretion apparatus, the Tir operon encodes Tir, CesT, and intimin, and operon 4 encodes the secreted Esp proteins. The LEE-encoded regulator (Ler) expressed by the first gene of operon 1 appears to positively regulate expression of genes in LEE operons 2 and 3, the Tir operon, and operon 4 (29, 43). In the EPEC strain E2348/69 (O127:H6), Ler expression has been shown to be positively regulated by least three different factors: (i) the products of the plasmid-borne per locus, (ii) a quorum-sensing mechanism involving autoinducer-2 (43), and (iii) a locus termed the integration host factor (14). By contrast,E. coli O157:H7 and other EHEC isolates possess neither the EPEC plasmid nor the per regulatory genes (18, 29). As with EPEC, quorum sensing involving autoinducer-2 appears to play an important role in the expression of LEE operons 1 and 2 in E. coli O157:H7 (43). In addition, the expression of E. coli O157:H7 LEE operons 3 and Tir may be activated by the alternative ς factor RopS (ς³⁸). Given these differences in LEE regulation between EPEC and E. coli O157:H7, it is not surprising that conditions affecting the expression of intimin by EPEC strain E2348/69 and E. coli O157:H7 strains should also differ. Previous studies have shown that E. coli O157:H7 adheres very poorly to epithelial cell lines such as HEp-2, INT407, and Caco-2 relative to EPEC strains (31, 40, 42). However, as with E. coli strain E2348/69 (O127:H7) (38), intimin expression by E. coli O157:H7 strain E319 is greatest during logarithmic-phase growth. In the present study, E. coli O157:H7 strain E319 intimin was first detected at 2 h after incubation in MEM and M9 medium, was greatest at 4 to 5 h after incubation, and did not appear to increase after this. The level of intimin expression by E. coli O157:H7 strains was also dependent on the culture medium and on the bacterial strain. All E. coli O157:H7(H⁻) strains tested produced intimin in both MEM and DMEM, but intimin was detected in only a few strains grown in BPW, LB, or M9 medium.

Interestingly, little intimin expression was observed for E. coli O157 strains when they were grown in media which allowed rapid bacterial growth, e.g., RB and TB. Nishikawa et al. (31) reported that E. coli O157:H7 and other EHEC strains cultured in peptone water in the absence of mannose adhered well to HEp-2 cells, whereas bacteria grown in the presence of d-mannose or other sugars showed little adherence to the HEp-2 cells. Results of the present study would support the possibility that intimin production in E. coli O157 strains is subject to catabolite repression. It would also seem likely that intimin expression is regulated differently among the E. coli O157:H7 strains examined or that some of the strains maintained in the laboratory may have lost certain regulatory elements. While it may be argued that some of the differences in intimin expression observed among E. coli O157:H7 strains may be artifacts that represent differences in reactivity to the monoclonal antibody, results with polyclonal antibody to the full-length intimin also suggest differences in the levels of intimin expression among these E. coli O157:H7 strains (see Table 1). Recently, Ogierman et al. (32) have reported the isolation of a strain of E. coli O157:H7 from a human patient which lacks attaching and effacing activity and produces very low levels of intimin. These workers demonstrated that this low-level intimin production is attributable to a mutation in the ler gene which results in an amino acid substitution in this important regulatory protein. It is possible that some of the E. coli O157:H7 strains producing low levels of intimin that were examined in this study, e.g., E. coli O157:H7 strain H4420, may also have defects in this or other regulatory genes. Further study is required to characterize the genetic features associated with differences in intimin expression among E. coli O157:H7 strains. This information will be helpful in understanding the pathogenesis of infection caused by this bacterial pathogen and the conditions which would be helpful for the immunological detection of intimin γ1-producing E. coli strains.