Functional analysis of the Shiga toxin and Shiga-like toxin type II variant binding subunits by using site-directed mutagenesis.

The B subunit of Shiga toxin and the Shiga-like toxins (SLTs) mediates receptor binding, cytotoxic specificity, and extracellular localization of the holotoxin. While the functional receptor for Shiga toxin, SLT type I (SLT-I), and SLT-II is the glycolipid designated Gb3, SLT-II variant (SLT-IIv) may use a different glycolipid receptor. To identify the domains responsible for receptor binding, localization in Escherichia coli, and recognition by neutralizing monoclonal antibodies, oligonucleotide-directed site-specific mutagenesis was used to alter amino acid residues in the B subunits of Shiga toxin and SLT-IIv. Mutagenesis of a well-conserved hydrophilic region near the amino terminus of the Shiga toxin B subunit rendered the molecule nontoxic but did not affect immunoreactivity or holotoxin assembly. In addition, elimination of one cysteine residue, as well as truncation of the B polypeptide by 5 amino acids, caused a total loss of activity. Changing a glutamate to a glutamine at the carboxyl terminus of the Shiga toxin B subunit resulted in the loss of receptor binding and immunoreactivity. However, the corresponding mutation in the SLT-IIv B subunit (glutamine to glutamate) did not reduce the levels of cytotoxicity but did affect extracellular localization of the holotoxin in E. coli.     

Cloning and sequencing of a Shiga-like toxin type II variant from Escherichia coli strain responsible for edema disease of swine.

A Shiga-like toxin type II variant (SLT-IIv) is produced by strains of Escherichia coli responsible for edema disease of swine and is antigenically related to Shiga-like toxin type II (SLT-II) of enterohemorrhagic E. coli. However, SLT-IIv is only active against Vero cells, whereas SLT-II is active against both Vero and HeLa cells. The structural genes for SLT-IIv were cloned from E. coli S1191, and the nucleotide sequence was determined and compared with those of other members of the Shiga toxin family. The A subunit genes for SLT-IIv and SLT-II were highly homologous (94%), whereas the B subunit genes were less homologous (79%). The SLT-IIv genes were more distantly related (55 to 60% overall homology) to the genes for Shiga toxin of Shigella dysenteriae type 1 and the nearly identical Shiga-like toxin type I (SLT-I) of enterohemorrhagic E. coli. (These toxins are referred to together as Shiga toxin/SLT-I.) The A subunit of SLT-IIv, like those of other members of this toxin family, had regions of homology with the plant lectin ricin. SLT-IIv did not bind to galactose-alpha 1-4-galactose conjugated to bovine serum albumin, which is an analog of the eucaryotic cell receptor for Shiga toxin/SLT-I and SLT-II. These findings support the hypothesis that SLT-IIv binds to a different cellular receptor than do other members of the Shiga toxin family but has a similar mode of intracellular action. The organization of the SLT-IIv operon was similar to that of other members of the Shiga toxin family. Iron did not suppress SLT-IIv or SLT-II production, in contrast with its effect on Shiga toxin/SLT-I. Therefore, the regulation of synthesis of SLT-IIv and SLT-II differs from that of Shiga toxin/SLT-I.     

Effects of human intravenous immune globulin on diarrhea caused by Shiga-like toxin I and Shiga-like toxin II in infant rabbits.

Shiga toxin and the related Shiga-like toxins (SLT), produced by Escherichia coli, can cause hemorrhagic colitis and hemolytic uremic syndrome (HUS). Human intravenous immune globulin (HIVIg) blocks the cytotoxicity of some SLTs in vitro. To examine the ability of HIVIg to modify disease caused by Shiga-like toxin I or Shiga-like toxin II (SLT-I or SLT-II), we injected 3-day-old rabbits intraperitoneally with SLT-containing cell-free supernatants from Escherichia coli O157: H7. A subset of rabbits was treated with subcutaneous HIVIg. All rabbits given 10(4) CD50 of SLT-I developed severe diarrhea, and 5 died. When HIVIg 500 mg/kg was given in addition to SLT-I, only 6 of 18 rabbits (33.3%) developed diarrhea (P < 0.0001), and 1 died. HIVIg 500 mg/kg or 1,000 mg/kg protected against diarrhea when given one hour prior to toxin. HIVIg 1,000 mg/kg was protective when administered one hour after toxin, but not at 6 or 24 hr. Seventeen of 18 rabbits given 10(6) CD50 of SLT-II developed severe diarrhea, and 4 died. In contrast to SLT-I-associated disease, HIVIg had no effect on diarrhea in rabbits given SLT-II. We conclude that HIVIg protects infant rabbits from diarrhea and death caused by intraperitoneally administered SLT-I, but does not affect the course of SLT-II-associated illness.     

Synthesis of an artificial glycoconjugate polymer carrying Pk-antigenic trisaccharide and its potent neutralization activity against Shiga-like toxin.

Fluorescence-labeled glycoconjugate polymers carrying carbohydrate segments of a globotriaosyl ceramide (Gb3) were synthesized and subjected to biological assays using Escherichia coli O-157 strains and Shiga-like toxins (Stx-I and Stx-II). For the fluorescence labeling, a new polymerizable fluorescent monomer with a TBMB carbonyl chromophore (Ex. 325 nm, Em. 410 nm) was designed. A glycosyl monomer of the trisaccharide segment of Gb3 was prepared from p-nitrophenyl beta-lactoside and copolymerized with acrylamide and the fluorescent monomer to prepare a fluorescence-labeled glycoconjugate copolymer carrying [alpha-D-galactopyranosyl-(1-->4)-beta-D-galactopyranosyl]-(1-->4)-beta- D-glucopyranoside. The polymer showed potent neutralization activity against Stx-I and also binding activity onto E. coli O-157 strains.     

Globotetraosylceramide is recognized by the pig edema disease toxin

The pig edema disease toxin has been shown by a tlc glycolipid binding assay to bind specifically to globotetraosylceramide (Gb4, GalNAc beta 1-3Gal alpha 1-4Gal beta 1-4GlcCer.). Binding was reduced for globotriosylceramide (Gb3, Gal alpha 1-4Gal beta 1-4GlcCer) and more markedly for the Forssman antigen (GalNAc alpha 1-3GalNAc beta 1-3Gal alpha 1-4Gal beta 1-4GlcCer). Paragloboside, blood group A glycolipids, glycolipids terminating in Gal NAc beta 1-4Gal-, and glycolipids in which globoside was present as an internal sequence did not bind the toxin. Isogloboside (GalNAc beta 1-3Gal alpha 1-3Gal beta 1-4GlcCer) was efficiently recognized. This toxin is genetically related to the verotoxin (or Shiga-like) family of toxins for which Gb3 has been shown to be the receptor. The difference in susceptibility of cell lines to the cytotoxicity of the pig edema disease toxin and the Shiga and Shiga-like toxins is consistent with the difference in receptor glycolipid binding.     

Escherichia coli strains isolated from pigs with edema disease produce a variant of Shiga-like toxin II

Escherichia coli O157:H7 strains 933 produces elevated levels of 2 phage-encoded, antigenically distinct cytotoxins designated Shiga-like toxin I (SLT-I) and Shiga-like toxin II (SLT-II). These toxins kill both HeLa and Vero cells. In this report, the relationship between SLTs and a cytotoxin produced by E. coli strains isolated from pigs with edema disease (ED) was examined. Culture filtrates from 72 out of 81 ED strains were cytotoxic for Vero but not HeLa cells. Cytotoxicity was neutralized by antiserum to SLT-II but not by anti-Shiga toxin. No toxin-converting phage were detected in 20 toxigenic ED strains examined. The cytotoxin of the ED-causing strains appears to be a variant of SLT-II and production of this cytotoxin is not phage-mediated.     

Polyisobutylmethacrylate modifies glycolipid binding specificity of verotoxin 1 in thin-layer chromatogram overlay procedures.

Verotoxins (or Shiga-like toxins) are a family of closely related toxins elaborated by Escherichia coli. At least three toxins have been described, VT1, VT2, and SLTII, in addition to Shiga toxin itself, and all bind to globotriaosyl ceramide, Gb3. Some discrepancies exist in the literature regarding the binding of the toxins to Gb4 as monitored by TLC overlay procedures. These procedures are widely used to investigate the specificity of carbohydrate-binding ligands. Polyisobutylmethacrylate, PIBM, is generally used in TLC overlay procedures to prevent silica loss and orient carbohydrate moieties for the binding of various ligands to glycolipids. We now report that pretreatment of chromatograms with PIBM modifies binding of VT1 to include Gb4 and decreases binding to Gb3 and the P1 glycolipid. We suggest that PIBM can alter the conformation of the glycolipid oligosaccharide, and therefore caution is advised in analysis of ligand binding to glycolipids after treatment with this compound.     

Nucleotide sequence analysis and comparison of the structural genes for Shiga-like toxin I and Shiga-like toxin II encoded by bacteriophages from Escherichia coli 933

The nucleotide sequence of the Shiga-like toxin type II (SLT-II) structural genes cloned from bacteriophage 933W of the enterohemorrhagic Escherichia coli O157:H7 strain 933 was determined. This sequence was compared with the published sequence for the structural genes of the antigenically distinct Shiga-like toxin type I (SLT-I) encoded by bacteriophage 933J. The SLT-I and SLT-II structural genes shared 58% overall nucleotide and 56% amino acid sequence homologies. The A and B subunits of SLT-I and SLT-II were nearly identical in size and had similar secondary structures and hydropathy plots. The regulation proposed for the SLT-II operon is similar to that previously proposed for SLT-I.     

Mutational analysis of the Shiga toxin and Shiga-like toxin II enzymatic subunits.

The A-subunit polypeptides of Shiga toxin, the Shiga-like toxins (SLTs), and the plant lectin ricin inactivate eucaryotic ribosomes by enzymatically depurinating 28S rRNA. Comparison of the amino acid sequences of the members of the Shiga toxin family and ricin revealed two regions of significant homology that lie within a proposed active-site cleft of the ricin A chain. In previous studies, these conserved sequences of the SLT-I and ricin A subunits have been implicated as active sites. To establish the importance of these regions of homology, we used site-directed mutagenesis to alter the A-subunit sequences of two members of the Shiga toxin family. Substitution of an aspartic acid for glutamic acid 166 of the Slt-IIA subunit decreased the capacity of the polypeptides to inhibit protein synthesis at least 100-fold in a cell-free translation system. However, this mutation did not prevent the expression of immunoreactive, full-length Slt-IIA. In addition, SLT-II holotoxin containing the mutated A subunit was 1,000-fold less toxic to Vero cells. Finally, site-directed mutagenesis was used to delete sequences encoding amino acids 202 through 213 of the Shiga toxin A subunit. Although this deletion did not prevent holotoxin assembly, it abolished cytotoxic activity.     

Identification of three amino acid residues in the B subunit of Shiga toxin and Shiga-like toxin type II that are essential for holotoxin activity.

Shiga toxin of Shigella dysenteriae type I and Shiga-like toxins I and II (SLT-I and SLT-II, respectively) of enterohemorrhagic Escherichia coli are functionally similar protein cytotoxins. These toxin molecules have a bipartite molecular structure which consists of an enzymatically active A subunit that inhibits protein synthesis in eukaryotic cells and an oligomeric B subunit that binds to globotriaosylceramide glycolipid receptors on eukaryotic cells. Regionally directed chemical mutagenesis of the B subunit of SLT-II was used to identify amino acids in the B subunit that are critical for SLT-II holotoxin cytotoxic activity. Three noncytotoxic mutants were isolated, and their mutations were mapped. The substitutions of arginine with cysteine at codon 32, alanine with threonine at codon 42, and glycine with aspartic acid at codon 59 in the 70-amino-acid mature SLT-II B polypeptide resulted in the complete abolition of cytotoxicity. The analogous arginine, alanine, and glycine residues were conserved at codons 33, 43, and 60 in the 69-amino-acid mature B polypeptide of Shiga toxin. Comparable mutations induced in the B-subunit gene of Shiga toxin by oligonucleotide-directed, site-specific mutagenesis resulted in drastically decreased cytotoxicity (10(3)- to 10(6)-fold) as compared with that of wild-type Shiga toxin. The mutant SLT-II and Shiga toxin B subunits were characterized for stability, receptor binding, immunoreactivity, and ability to be assembled into holotoxin.     

Synthetic oligodeoxyribonucleotide probes to detect verocytotoxin-producing Escherichia coli in diseased pigs

Oligonucleotide probes constructed from the sequences published for Shiga-like toxin I (SLT-I) and Shiga-like toxin II (SLT-II) genes and antibody against the purified toxins were used to study the SLT (SLT-IIp) produced by porcine E. coli O138 and O139 strains. By DNA hybridization assays no homology was observed between SLT-I and SLT-IIp. By contrast the oligonucleotide probe derived from the slt-II A gene detected porcine strains of E. coli producing SLT-IIp and E. coli strains associated with human disease producing SLT-II. Homology of nucleotide sequences between SLT-IIp and SLT-II is reflected by serological cross-reactivity as demonstrated by a dot blot ELISA and neutralization of SLT-IIp with anti-SLT-II. The toxins were distinguishable in their ability to kill HeLa S-3 cells. The oligonucleotide probe and anti-SLT-II can facilitate identification of SLT-IIp producing E. coli to further clarify their role in diseased pigs.     

Shiga toxin, Shiga-like toxin II variant, and ricin are all single-site RNA N-glycosidases of 28 S RNA when microinjected into Xenopus oocytes

Ricin, Shiga toxin, and Shiga-like toxin II (SLT-II, Vero toxin 2) exhibit an RNA N-glycosidase activity which specifically removes a single base near the 3' end of 28 S rRNA in isolated rat liver ribosomes and deproteinized 28 S rRNA (Endo Y., Mitsui, K., Motizuki, M., & Tsurugi, K. (1987) J. Biol. Chem. 262, 5908-5912; Endo Y. & Tsurugi, K. (1987) J. Biol. Chem. 262, 8128-8130, Endo, Y., Tsurugi, K., Yutsudo, T., Takeda, Y., Ogasawara, K. & Igarashi, K. (1988) Eur. J. Biochem. 171, 45-50). These workers identified the single base removed, A-4324, by examining a 28 S rRNA degradation product which was generated by contaminating ribonucleases associated with the ribosomes. To determine whether this N-glycosidase activity applies in living cells, we microinjected ricin into Xenopus oocytes. We also microinjected Shiga toxin and a variant of Shiga-like toxin II (SLT-IIv). All three toxins specifically removed A-3732, located 378 nucleotides from the 3' end of 28 S rRNA. This base is analogous to the site observed in rat 28 S rRNA for ricin, Shiga toxin, and SLT-II. Purified, glycosylated, ricin A chain contains this RNA N-glycosidase activity in oocytes. We also demonstrated that the nonglycosylated A subunit of recombinant ricin exhibits this RNA N-glycosidase activity when injected into Xenopus oocytes. Ricin, Shiga toxin, and SLT-IIv also caused a rapid decline in oocyte protein synthesis for nonsecretory proteins.     

Enzyme-linked immunosorbent assays for detecting antibodies to Shiga-like toxin I,Shiga-like toxin II,andEscherichia coli O157:H7 lipopolysaccharide in human serum

Shiga-like toxin-producingEscherichia coli O157:H7 are important causes of bloody diarrhea and hemolytic uremic syndrome. To facilitate the epidemiologic study of these organisms, we developed enzyme-linked immunosorbent assays (ELISAs) for antibodies to Shiga-like toxin I (SLT I), Shiga-like toxin II (SLT II), andE. coli O157 lipopolysaccharide (LPS). We tested serum samples from 83 patients in two outbreaks ofE. coli O157:H7 diarrhea and from 66 well persons. Forty-three patients (52%) had at least one serum sample positive for anti-O157 LPS antibodies; among 26 culture-confirmed patients, 24 (92%) had at least one positive serum sample. Two (3%) of 66 control sera had positive anti-O157 LPS titers. ELISA results for SLT I and II were compared with those of HeLa cell cytotoxicity neutralization assays on both patient and control sera. Neutralization assays detected anti-SLT I antibodies in at least one serum sample from each of 17 (20%) patients and 7 (10.6%) controls, while 16 (19%) patients and 7 controls had positive titers by anti-SLT I ELISA. Although all serum samples, including control sera, showed nonspecific neutralization of SLT II, no antibody titers to SLT II were detected by either neutralization or ELISA. These results indicate that ELISAs for SLT I and SLT II antibodies are comparable to HeLa cell cytotoxicity neutralization assays. Both the ELISAs and neutralization assays are insensitive in detecting infected patients. However, the ELISA for antibodies toE. coli O157 LPS is both sensitive and specific, and may be more useful than assays for antitoxic antibodies in detecting persons withE. coli O157:H7 infection.     

Shiga toxin binding to glycolipids and glycans

Background

Immunologically distinct forms of Shiga toxin (Stx1 and Stx2) display different potencies and disease outcomes, likely due to differences in host cell binding. The glycolipid globotriaosylceramide (Gb3) has been reported to be the receptor for both toxins. While there is considerable data to suggest that Gb3 can bind Stx1, binding of Stx2 to Gb3 is variable.

Methodology

We used isothermal titration calorimetry (ITC) and enzyme-linked immunosorbent assay (ELISA) to examine binding of Stx1 and Stx2 to various glycans, glycosphingolipids, and glycosphingolipid mixtures in the presence or absence of membrane components, phosphatidylcholine, and cholesterol. We have also assessed the ability of glycolipids mixtures to neutralize Stx-mediated inhibition of protein synthesis in Vero kidney cells.

Results

By ITC, Stx1 bound both Pk (the trisaccharide on Gb3) and P (the tetrasaccharide on globotetraosylceramide, Gb4), while Stx2 did not bind to either glycan. Binding to neutral glycolipids individually and in combination was assessed by ELISA. Stx1 bound to glycolipids Gb3 and Gb4, and Gb3 mixed with other neural glycolipids, while Stx2 only bound to Gb3 mixtures. In the presence of phosphatidylcholine and cholesterol, both Stx1 and Stx2 bound well to Gb3 or Gb4 alone or mixed with other neutral glycolipids. Pre-incubation with Gb3 in the presence of phosphatidylcholine and cholesterol neutralized Stx1, but not Stx2 toxicity to Vero cells.

Conclusions

Stx1 binds primarily to the glycan, but Stx2 binding is influenced by residues in the ceramide portion of Gb3 and the lipid environment. Nanomolar affinities were obtained for both toxins to immobilized glycolipids mixtures, while the effective dose for 50% inhibition (ED₅₀) of protein synthesis was about 10⁻¹¹ M. The failure of preincubation with Gb3 to protect cells from Stx2 suggests that in addition to glycolipid expression, other cellular components contribute to toxin potency.     

Molecular cloning and nucleotide sequence of another variant of the Escherichia coli Shiga-like toxin II family

Escherichia coli strain H.I.8 (O128:B12) produces low levels of a Shiga-like toxin (SLT) which we have called SLTIIva because of its close relationship with SLTIIv. The Vero cell cytotoxicity of SLTIIva is neutralized by antisera against SLTII and SLTIIv but not by antisera against SLTI. These data indicate that the SLT of strain H.I.8 is a member of the SLTII family. Since SLTIIva shares with SLTIIv the property of having low cytotoxicity to HeLa cells compared with Vero cells, it is appropriate to consider both toxins as variants of SLTII. SLTIIva differs from SLTIIv in that it is more heat-stable. Further, SLTIIv-producing strains of E. coli have only been isolated from pigs while the SLTIIva-producing E. coli strain examined in this study was isolated from a human infant with diarrhoea. The genes for this SLT were cloned from a cosmid library of total cellular DNA by screening recombinants for Vero cell toxicity and with a DNA probe derived from SLTIIv structural genes. Nucleotide sequence analysis was performed on a 2.0 kb AvaII-HincII fragment which encodes the toxin gene. The nucleotide sequence data confirm the close relationship between SLTIIva and SLTIIv: they have 98% nucleotide sequence homology in the B subunit gene and 70.6% homology in the A subunit gene. Comparison of DNA sequences indicated that SLTIIva was most closely related to SLTIIv, closely related to SLTII and less closely related to SLTI.     

The observation of multivalent complexes of Shiga-like toxin with globotriaoside and the determination of their stoichiometry by nanoelectrospray Fourier-transform ion cyclotron resonance mass spectrometry.

We show by nanoelectrospray ionization (nanoES) Fourier-transform ion cyclotron resonance mass spectrometry (FT-ICR MS) that it is possible to observe oligosaccharide-protein complexes with dissociation constants in the millimolar range, such as P(k) trisaccharide (globotriaoside) complexed with the Shiga-like toxin (SLT) of pathogenic E. coli. It is further demonstrated that nanoES/FT-ICR MS is an exquisite method to study quantitative aspects of the association of mono- and polyvalent oligosaccharide ligands with multimeric proteins, such as the SLTs. At increasing trisaccharide:protein ratios it was shown that the B(5 )toxin subunit complexes with 5 P(k) trisaccharides and only after all 5 copies of site 2 are essentially filled do any of the remaining 10 receptor sites become occupied. From the distribution of bound P(k)'s at the five binding sites, it was possible to establish association constants for each of the five sites and to confirm that binding occurs noncooperatively, the association constants for each site are identical and that compared to site 1, site 2 exhibits a tenfold higher affinity for the globotriaoside synthetic ligand 1. The facile identification of the occupancy of binding sites represents information that is not readily available by other techniques. This sensitive and rapid estimation of association constants for protein-ligand complexes, which are free of unpredictable secondary effects that plague enzyme linked assays, is likely to find wide application.     

A convenient oxidation of natural glycosphingolipids to their "ceramide acids" for neoglycoconjugation. Bovine serum albumin-glycosylceramide acid conjugates as investigative probes for HIV gp120 coat protein-glycosphingolipid interactions.

A new method to cleave the double bond of sphingolipids has been developed. Using limited concentrations of KMnO4 and an excess of NaIO4, in a neutral aqueous tert-butanol solvent system gave nearly quantitative yields of the oxidized product. A variety of natural glycosphingolipids (GSLs): GlcC, GalC, SGC, LC, Gb3C, Gb4C, Gg4C, Gb5C, and GM1C, gave the corresponding acids: 2-hydroxy-3-(N-acyl)-4-(O-glycosyl)-oxybutyric acids, i.e. "glycosyl ceramide acids" (GSL.CCOOH) in excellent yields (80-90%). Deacyl GSLs (dGSLs) were oxidized to acids containing the oligosaccharides devoid of hydrocarbon chains, i.e. "ceramide oligosaccharides" (dGSL. NRR1CCOOH, where R = R1 = H; R = H, R1 = CH3CO; or R = R1 = Me). The efficacy of this method was demonstrated by transforming natural GSLs: GlcC, GalC, GalS, SGC, LC, Gb3C, and Gb4C into neoglycoproteins via coupling glycosyl ceramide acids (except GalS, which was coupled directly) to bovine serum albumin (BSA). Mass spectroscopic analysis of GalC-BSA conjugates, (GalC.CONH)nBSA and (GalS.NHCO)nBSA gave a value of 9 +/- 1 and 16 +/- 2 for n. Neoglycoconjugates derived from GlcC, GalC (type I and II and the behenic analog), SGC, LC, and Gb3C were recognized by the recombinant human immunodeficiency virus coat protein gp120 (rgp120). The GalS conjugate showed significantly reduced binding, and the Gb4C conjugate showed no binding. Thus, rgp120/GSL-BSA interaction requires a terminal galactose and/or glucose residue. Terminal N-acetylgalactosamine containing GSLs are not bound. The ceramide acid conjugates provide a more effective scaffold for presentation of glycone for rgp120 binding than those derived from dGSLs. The retention of receptor specificity of the glycoconjugates was validated by retention of the expected binding specificity of VT1 and VT2e for Gb3C and Gb4C conjugates, respectively. These studies open a new vista in the generation of glycoconjugates from GSLs and further emphasize the role of aglycone in glycolipid recognition.     

Crystallization of the B chain of Shiga-like toxin I from Escherichia coli

Shiga-like toxin I (SLT-I) is produced by several pathogenic strains of Escherichia coli associated with diarrheal disease. The toxin consists of an A chain, which attacks eukaryotic ribosomes, inhibiting protein synthesis, and multiple copies of a 69 amino acid B chain. The B subunit mediates cell binding and uptake through its interactions with cell surface carbohydrate moieties. Here we report that the B chain has been crystallized in a form suitable for high-resolution X-ray analysis. The space group is P2(1)2(1)2(1), with a = 56.2 A, b = 59.9 A and c = 102.5 A. A rotation function using three-dimensional diffraction data suggests that the asymmetric unit is a tetramer.