Search

Shiga toxin (ST) is a protein toxin of Shigella dysenteriae type 1, a causative agent of severe diarrhoea and dysentery. In this report we describe the gastrointestinal secretory antibody response of mice following orogastric immunization with ST. Gastrointestinal secretions were sampled by a gastrointestinal lavage technique weekly for 5 weeks after initial immunization. Assay of lavage samples by ELISA showed that mice vaccinated orogastrically with various doses of ST developed gastrointestinal antibody to ST in a dose-dependent manner. Serum anti-ST activity developed by 5 weeks after initial immunization. The ability of ST to act as a mucosal immune adjuvant was investigated by coadministration of ST and keyhole limpet haemocyanin. In contrast to cholera toxin, a potent adjuvant, ST did not demonstrate adjuvant activity. The mouse gastrointestinal lavage model could be useful for further analysis of the cellular basis of ST immunogenicity.

Oligonucleotide probes designed on the basis of the N-terminal sequence of Clostridium perfringens beta-toxin were used to isolate the encoding gene (cpb). The nucleotide sequence of cpb was determined, and on the basis of DNA hybridization experiments it was shown that the gene is found only in type B and C strains of C. perfringens. The deduced amino acid sequence of the beta-toxin revealed homology with the alpha-toxin, gamma-toxin, and leukocidin of Staphylococcus aureus. The beta-toxin purified from C. perfringens appeared to exist in monomeric and multimeric forms. Recombinant beta-toxin, produced in Escherichia coli, appeared to be mainly in the multimeric form.

Acute- and convalescent-phase sera from 18 Thai patients and convalescent-phase sera from two Israeli patients and one Bangladeshi patient with Shigella dysenteriae 1 (Shiga) dysentery were tested by enzyme-linked immunosorbent assay to detect antibodies that bind S. dysenteriae lipopolysaccharide (LPS), Shiga holotoxin, or two synthetic peptides representing epitopes from the B subunit of Shiga toxin. Paired sera from 24 Maryland adults with Shigella flexneri 2a or Shigella sonnei diarrhea served as negative controls. Of the 16 paired Thai serum samples tested for immunoglobulin G LPS antibody, 10 had greater than or equal to 4-fold rises (the two subjects with the highest convalescent-phase titers exhibited toxin-neutralizing activity); acute-phase specimens from four of the remaining six individuals already had elevated Shiga LPS titers in their acute specimens ranging from 1:800 to 1:12,800. Similarly, convalescent-phase sera from the two Israeli patients and the Bangladeshi patient revealed LPS titers of 1:800 to 1:3,200. In contrast, none of the Maryland volunteers with S. flexneri or S. sonnei diarrhea manifested rises in Shiga anti-LPS (P less than 0.00001 versus 10 of 16 Thai patients). Only 4 of the 18 Thai patients had significant rise in antibody to purified Shiga toxin, while one of the two Israeli patients and the one Bangladeshi patient had elevated convalescent-phase titers. None of the sera that reacted with Shiga holotoxin had antibody that bound to the peptides. This report, which describes a search for serum antibodies that bind Shiga toxin in patients with Shiga dysentery, demonstrates such antibodies in only a minority of patients with bacteriologically confirmed disease. During Shiga dysentery, Shiga toxin may be elaborated in such small quantities in vivo that it fails to elicit an immune response in most patients even though it may exert biological effects. In this behavior Shiga toxin resembles tetanus toxin, another potent exotoxin that fails to elicit antitoxic responses in people who recover from clinical tetanus.

Specific DNA probes were used to identify Shiga-like toxin I (SLT I)- and SLT II-producing Escherichia coli in vegetables, meats, cattle, and farm animals in Thailand. SLT-producing E. coli was isolated from 9% of market beef specimens, from 8 to 28% of fresh beef specimens at slaughterhouses, and from 11 to 84% of fecal specimens from cattle. Animals were frequently infected with several different SLT-producing E. coli types that hybridized with either the SLT I, SLT II, or both SLT probes. Of 119 SLT-producing E. coli isolates, 24% hybridized with the SLT I probe, 31% hybridized with the SLT II probe, and 44% hybridized with both SLT probes. The enterohemorrhagic E. coli plasmid probe hybridized with 64% (68 of 106) of SLT-producing E. coli isolates from food and cattle and with 8% (17 of 201) of E. coli isolates from pigs. No SLT-producing E. coli was detected in pigs. Seventy-six percent (26 of 34) of E. coli isolates that hybridized with the SLT II probe were cytotoxic to Vero but not to HeLa cells, suggesting that they produced the variant of SLT II. The high prevalence of SLT-producing E. coli in beef-producing animals suggests that exposure to animals and eating beef may pose a health risk for acquiring enterohemorrhagic E. coli infections in Thailand.

Synthetic oligonucleotides, constructed from the nucleotide sequences of genes coding for the A subunit of Shiga-like toxin (SLT) I and the B subunit of SLT-II, were used as probes at different degrees of stringency to identify Escherichia coli producing different types of SLTs. At 45 degrees C, the A-I oligonucleotide probe hybridized with E. coli producing SLT-I, SLT-II, and variant of SLT-II (SLT-IIv). At 53 degrees C, only SLT-I-producing E. coli hybridized with this probe. At 45 degrees C, the B-II oligonucleotide probe hybridized with SLT-II- and SLT-IIv-producing E. coli. At 53 degrees C, this probe hybridized with only SLT-II-producing E. coli. The A-I and B-II oligonucleotide probes were subsequently tested for hybridization with 73 SLT-producing E. coli and 49 non-SLT-producing E. coli isolated in Asia and Canada. At 45 degrees C, the A-I oligomer had a sensitivity of 97% and a specificity of 100% in identifying SLT-producing E. coli. At 53 degrees C, the A-I oligonucleotide probe had a sensitivity of 92% and a specificity of 91% in identifying E. coli containing genes encoding SLT-I. At 45 degrees C, the B-II oligonucleotide had a 100% sensitivity and 97% specificity in identifying E. coli that hybridized with the SLT-II probe. Of 17 E. coli that hybridized only with the SLT-II probe, 10 did not hybridize with the B-II oligonucleotide at 53 degrees C. All 10 isolates were cytotoxic to Vero cells but not to HeLa cells, confirming that the B-II oligonucleotide probe used at 53 degrees C will differentiate isolates producing SLT-II and SLT-IIv.

Shiga toxin has been purified in milligram quantities to near homogeneity from cell lysates of Shigella dysenteriae 1 strain 3818-0. Purification involved an initial ultracentrifugation, ammonium sulfate fractionation, chromatography on DEAE-cellulose and carboxymethyl cellulose, gel filtration, and preparative isoelectric focusing in sucrose gradients. The purified toxin was resolved by discontinuous polyacrylamide gel electrophoresis into a major cytotoxic protein band and a closely migrating, cytotoxic protease-nicked minor band. Antiserum generated by immunization with glutaraldehyde-inactivated toxin was shown to be monospecific against S. dysenteriae cell lysates. This highly purified toxin was cytotoxic to HeLa cells, enterotoxic in rabbit ileal loops, and lethal to mice. Monospecific antiserum to the toxin neutralized completely these toxin activities in both purified toxin preparations and crude shigella cell lysates.

Specific DNA probes were used to identify Shiga-like-toxin (SLT) I- and II-producing Escherichia coli from children less than 5 years of age with bloody diarrhea, in infants with diarrhea, and in controls of the same age without diarrhea in Thailand. At one hospital, SLT-producing E. coli was identified in 4 (7%) of 54 children with bloody diarrhea from whom other enteric pathogens were not identified and from 3 (6%) of 50 children without diarrhea. In the positive specimens, SLT-producing E. coli constituted only 0.3 to 4% of the 100 to 300 colonies on the replica blots. Non-toxin-encoding 933J and 933W bacteriophagelike DNA sequences were detected by colony hybridization with E. coli isolates from 18 (33%) of 54 children with bloody diarrhea and 23 (46%) of 50 controls. At another hospital, SLT-producing E. coli was not identified in 115 infants with diarrhea and 119 controls without diarrhea. One infant with diarrhea was infected with E. coli O76:H7 that hybridized with the enterohemorrhagic E. coli probe but not with the SLT probes. E. coli producing SLT I or SLT II was isolated in small numbers from a similar proportion of Thai children with bloody diarrhea in whom no other enteric pathogen was identified and from controls without diarrhea.

The effect of Shiga toxin, from Shigella dysenteriae 1, on the component reactions of peptide elongation were investigated. Enzymic binding of [3H]phenylalanine-tRNA to reticulocyte ribosomes was inhibited by 50% at 7 nM toxin. Elongation factor 1 (eEF-1)-dependent GTPase activity was also inhibited. Both reactions were not restored by addition of excess eEF-1 protein. In contrast, toxin concentrations of 200 nM were required to inhibit by 50% the elongation factor 2 (eEF-2)-dependent translocation of aminoacyl-tRNA on ribosomes. Addition of excess eEF-2 restored translocation activity. The eEF-2-dependent GTPase activity was unaffected at toxin concentrations below 100 nM, and Shiga-toxin concentrations of up to 1,000 nM did not affect either GTP.eEF-2.ribosome complex-formation or peptidyltransferase activity. Thus Shiga toxin closely resembles alpha-sarcin in action, both being primary inhibitors of eEF-1-dependent reactions. In contrast, the 60 S ribosome inactivators ricin and phytolaccin are primary inhibitors of eEF-2-dependent reactions of peptide elongation.

Heat-labile enterotoxin (LT)-producing Escherichia coli strains, as identified by the Y1 adrenal cell assay, were examined with a DNA probe coding for type I and type II LTs. Of 236 LT-producing E. coli isolates, 60% hybridized with LT-I, 17% hybridized with LT-II, and 23% did not hybridize with either probe and no longer produced LT as determined by the Y1 adrenal cell assay. These isolates presumably lost plasmids coding for LT-I during storage. A total of 75% of LT-producing E. coli isolates (27 of 36) from cows, 64% of LT-producing E. coli isolates (7 of 11) from buffalo, 31% of LT-producing E. coli isolates (4 of 13) from beef obtained in markets, and 2% of LT-producing E. coli isolates (3 of 168) from humans contained genes coding for LT-II. Genes coding for LT-II were not found in 50 LT-I-producing and heat-stable enterotoxin-producing E. coli isolates from 11 children with diarrhea and 44 LT-nonproducing and heat-stable enterotoxin-producing E. coli isolates from 12 other children with diarrhea. A total of 9% of LT-II-producing E. coli isolates (3 of 34) from cows and buffalo hybridized with DNA probes for genes coding for verocytotoxin 2 (VT2), and 18% (6 of 34) hybridized with a DNA probe coding for enterohemorrhagic E. coli (EHEC) adhesin fimbriae. E. coli SA-53, the original isolate in which LT-II was found, contained genes coding for VT2 and EHEC adhesin fimbriae. Five VT-producing, LT-II-producing E. coli isolates that hybridized with the EHEC probe did not contain DNA sequences coding for VT1 or VT2. LT-II-producing E. coli strains were frequently isolated from cattle and buffalo but were rarely isolated from humans.

Toxin A and toxin B preparations of Clostridium difficile have been shown to affect metabolic functions of intact HeLa cells with different kinetics. The cytotoxins were purified from dialyzed filtrates of C. difficile strain VPI 10463 by hydrophobic interaction chromatography and ion-exchange chromatography and were concentrated by dialysis or by ultrafiltration. The toxins, which are immunologically unrelated, were analyzed by polyacrylamide gel electrophoresis and by immunochemistry with the Western blot technique. Toxin A was resolved into one major cytotoxic protein and a minor, rapidly migrating species that did not comigrate with toxin B. Toxin B was resolved into one major and three minor cytotoxic proteins. One protein comigrating with toxin A had no cytotoxic activity. The highly purified toxin A at 1.0 mg/ml caused loss of intracellular K+ and inhibition of protein synthesis in HeLa cells within 1 h. These effects correlated with morphological changes indicating cytotoxicity. At lower protein concentrations of toxin A (10- to 100-fold less), however, cytotoxic effects were seen at 120 min, whereas no changes in K+ levels or protein synthesis were yet evident. The toxin B preparation, 1,000-fold more toxic than toxin A, was diluted to equivalent cytotoxicity as measured in the overnight assay. Toxin B caused loss of K+ and inhibition of protein synthesis well after cytotoxic morphological changes were complete. In contrast, at higher protein concentrations (2- to 2,000-fold more), intracellular K+ was lost completely by 120 min. The effects on cell rounding and protein synthesis were incomplete at 120 min, but increased with the toxin B concentration.

Evidence is presented that endocytosis is involved in the transport to the cytosol of the cytotoxin from Shigella dysenteriae 1, Shiga toxin, which acts by removal of a single adenine residue in 28-S ribosomal RNA. Inhibition of endocytosis by ATP depletion of the cells prevented toxin uptake. Exposure of HeLa S3 and Vero cells to toxin at low extracellular pH, where translocation to the cytosol, but not endocytosis is inhibited, allowed the toxin to accumulate in a compartment where it was protected against antibodies to the toxin. Upon transfer of the cells to normal medium endocytosed toxin entered the cytosol. Electron microscopical studies of cells exposed at 0 degrees C to a toxin-horseradish peroxidase (HRP) conjugate, or to unconjugated toxin followed by horse antitoxin antibodies and then protein G-gold, revealed that the Shiga toxin binding sites were randomly distributed on the cell surface, without any preference to, for example, coated pits. In contrast, when cells were exposed to toxin at 37 degrees C, the binding sites were preferentially localized in coated pits. The Shiga-HRP conjugate was also seen in endosomes, lysosomes, and in the Golgi region. Endocytosis by the coated pit/coated vesicle pathway was selectively inhibited by acidification of the cytosol. Under these conditions, both the uptake of toxin-HRP conjugates and intoxication of the cells were inhibited. Evidence from the literature as well as our own results suggest that Shiga toxin binding sites are glycolipids. Thus, Shiga toxin appears to be the first example of a lipid-binding ligand that is endocytosed from coated pits.

Hybridoma cell lines which produce monoclonal antibodies to Shiga toxin from Shigella dysenteriae 1 were prepared. The monoclonal antibodies were all of the immunoglobulin G1 isotype and differed in their ability to neutralize cytotoxicity and to bind to Shiga toxin in a solid-phase radioimmunoassay. When used for immunoblot analysis, these antibodies were able to identify specifically both nicked and unnicked Shiga toxin in crude lysates of S. dysenteriae.

When Shigella species, Escherichia coli, and five other bacterial enteric pathogens isolated from children with diarrhea in Thailand were tested for hybridization under stringent conditions with probes for Shiga-like toxins I and II, only 30 Shigella dysenteriae 1 hybridized with the Shiga-like toxin I probe.

Although the role of Shiga toxin in dysentery is unknown, the toxin is cytotoxic to HeLa cells, causes fluid secretion in rabbit intestine, and is lethal to rabbits and mice when injected parenterally. In the present study, rabbits received three weekly doses of Shiga toxin directly into chronically isolated ileal loops. Within a week, secretions from these loops contained immunoglobulin A (IgA) anti-Shiga toxin. The titer of IgA anti-Shiga toxin increased after weekly doses 2 and 3. Little IgG anti-Shiga toxin was present in loop secretions, although high titers of IgG anti-Shiga toxin were found in the sera. These loop secretions were able to neutralize the cytotoxic effects of Shiga toxin in the HeLa cell assay. The capacity to neutralize the cytotoxicity of the toxin correlated strongly with the IgA anti-Shiga toxin titer in these same secretions. Pooled immune loop secretions were also able to significantly reduce fluid accumulation in acutely ligated loops in rabbits, while loop secretions from control rabbits could not. Shiga toxin elicited a strong secretory IgA response upon application to the intestine. Further, the mucosal antibodies produced functioned to prevent the toxic effects of Shiga toxin both in vitro and in vivo.

Clostridium difficile toxins A and B were shown to share immunochemical and structural features, including shared sequential epitopes. Nineteen hybridomas generated after immunization of mice with a mixture of toxoids produced monoclonal antibodies, all IgM(x), which bound to toxin A and toxin B in a solid-phase radioimmunoassay (RIA). None of the antibodies neutralized the cytotoxicity of either toxin, alone or in pairs, nor did they neutralize mouse lethality. The antibodies did not inhibit hemagglutination by toxin A, and none of those tested neutralized the toxin's enterotoxic activity. Studies of binding of antibodies to native toxins in the RIA showed that the antibodies differed in their recognition of the toxins. Many of the antibodies bound with higher avidity to toxin A than to toxin B. In Western blots, all the antibodies recognized both toxins in the native state; in addition, some antibodies recognized the minor cytotoxic species of toxin B. When the toxins were denatured and reduced, five antibodies bound to both toxins, five to A only, and nine to neither, demonstrating that the antibodies had different epitope specificities. Further structural comparisons were made by investigation of mol. wts, subunit structures and amino acid compositions. The native mol. wts of toxin A and toxin B, as determined by electrophoresis to equilibrium in 4-30% polyacrylamide gel electrophoresis (PAGE), were 430,000 and 368,000, respectively. Denatured and reduced toxins each had a single subunit of 315,000. Both toxins had about 50% hydrophobic amino acids.

Shiga toxin purified from Shigella dysenteriae 1 was cytotoxic to cultured epithelial cells from human colon and ileum. The cytotoxicity, which affected only about 50% of treated cells, was neutralized by rabbit antiserum monospecific for Shiga toxin and mediated by protein synthesis inhibition.

Inhibition of the peptide elongation cycle of eukaryotic protein synthesis by Shiga toxin from Shigella dysenteriae 1 was examined in toxin-treated reticulocyte lysate mixtures. Peptidyl transferase activity of toxin-treated ribosomes was measured by following the decrease in peptidyl-tRNA concentrations when puromycin was added after incubation with toxin. Concentrations of [3H]leucine-labeled peptidyl-tRNA were measured by extraction with cetyltrimethylammonium bromide. The data suggest that Shiga toxin inhibited aminoacyl-tRNA binding. Toxin-treated ribosomes retained peptidyl transferase activity, and toxin did not block translocation. Furthermore, no inhibition of initiation of protein synthesis could be observed. Finally, Shiga toxin had no detectable nucleolytic effect on polysomal 28S rRNA, nor was hydrolysis of 5.8S or 5S rRNA observed.

Shiga toxin purified to near homogeneity from cell lysates of Shigella dysenteriae 1 inhibited protein and deoxyribonucle acid syntheses in intact HeLa cells. Inhibition was dependent on toxin concentration and time of incubation. A minimal latent period of 30 min was observed with saturating doses of toxin. Ribonucleic acid synthesis, uptake of alpha-aminoisobutyric acid, and maintenance of intracellular K+ concentrations were not affected until well after maximal inhibition of protein and deoxyribonucleic acid syntheses. The inhibitory effect of toxin was sensitive to heat inactivation and was prevented by antibody neutralization. Several cytotoxic components were separated by polyacrylamide gel electrophoresis of the purified toxin preparations; all inhibited protein and deoxyribonucleic acid syntheses equally.

The ionic requirements for entry of Shiga toxin into cells were examined by measuring inhibition of protein synthesis after short-term incubations with toxin. The sensitivity of Vero cells and HeLa cells to Shiga toxin was strongly dependent on the divalent cation present. Vero cells were most sensitive in the presence of CaCl2 and SrCl2, whereas HeLa cells were equally sensitive in the presence of MgCl2, SrCl2, and CaCl2. Both cell lines were protected by BaCl2, CoCl2, and MnCl2. Inhibitors of Ca2+ transport, like verapamil, D600, and Co2+ as well as the calcium-ionophores A23187 and ionomycin, protected both cell lines. HEp-2 cells were protected against Shiga toxin by a high concentration of potassium in the medium as well as by potassium depletion of the cells. Substitution of chloride in the medium with slowly permeable anions, like SO42- and SCN-, protected the cells against Shiga toxin. High concentrations of the ionophore nigericin that increase pH of acidic intracellular vesicles did not protect Vero cells against Shiga toxin. Shiga Toxin X-114 at pH values below 4.5. This binding was shifted to higher pH values after pretreatment of the toxin with dithiothreitol. The results indicate that Ca2+ transport through physiologically occurring Ca2+ channels is required for entry of Shiga toxin into cells. Furthermore, the sensitivity of cells of Shiga toxin is strongly dependent on the anions present.

To help explain a role of the Shiga toxin family in hemorrhagic colitis and hemolytic-uremic syndrome in humans, it has been hypothesized that these toxins cause direct damage to the vascular endothelium. We now report that Shiga toxin purified from Shigella dysenteriae 1 does indeed have a direct cytotoxic effect on vascular endothelial cells in cultures. Human umbilical vein endothelial cells (HUVEC) in confluent monolayers were reduced 50% by 10(-8) M Shiga toxin after a lag period of 48 to 96 h. In comparison, nonconfluent HUVEC were reduced 50% by 10(-10) M Shiga toxin within a 24-h period. These data suggest that dividing endothelial cells are more sensitive to Shiga toxin than are quiescent cells in confluent monolayers. Both confluent and nonconfluent HUVEC specifically bound 125I-Shiga toxin. However, in response to the toxin, rates of incorporation of [3H]leucine into protein were more severely reduced in nonconfluent cells than in confluent cells. Toxin inhibition of protein synthesis preceded detachment of cells from the substratum. The specific binding of 125I-Shiga toxin to human endothelial cells and the cytotoxic response were both toxin dose dependent and neutralized by anti-Shiga toxin antibody. Heat-denatured Shiga toxin was without the cytotoxic effect. In addition, the complete culture system contained less than 0.1 ng of bacterial endotoxin per ml, as measured by the Limulus amoebocyte lysate test.