Expression, purification, and biochemical characterization of enteroaggregative Escherichia coli heat-stable enterotoxin 1

Enteroaggregative Escherichia coli (EAEC) heat-stable enterotoxin 1 (EAST1) is a 4.1k Da protein originally discovered in EAEC but known to be scattered in other diarrheagenic E. coli as well, possibly causing diarrhea in humans and animals. We report for the first time a method to express and purify EAST1 using the Glutathione S-transferase (GST) fusion system. The gst and astA genes were fused together on a pGEX-2T plasmid vector to produce a GST-EAST1 fusion protein. Using Glutathione Sepharose affinity chromatography and C(8) reverse phase high pressure liquid chromatography, EAST1 was purified to homogeneity with a yield of 0.29 mg/L of culture. The protein purified by this method was confirmed to be EAST1 by NH(2)-terminal sequencing and mass spectrometry. The molecular weight of EAST1 is 4104.0 Da, confirming a 38 amino acid peptide as predicted by the astA gene sequence. Mass spectrometry analysis of EAST1 and of two generated peptides established the presence and suggested the position of two disulfide bridges of EAST1 in the conformations C1-C2 and C3-C4. Polyclonal antibodies were raised against EAST1 in New Zealand white rabbits to a titer of 1:8000 using affinity-purified GST-EAST1 fusion protein and to a titer of 1:100 using HPLC-purified EAST1. The biological activity of various EAST1 preparations was tested using the suckling mouse assay with CD-1 and CFW mice strains, but failed to produce fluid accumulation in the intestine.     

A modified bioassay for heat-stable Escherichia coli enterotoxin1.

Infant mice were injected orally with preparations containing Escherichia coli heat-stable enterotoxin (ST) and Evans blue dye, and incubated at 22 degrees C. With enterotoxin-positive samples, the stomach was distended and contained essentially all of the dye. With enterotoxin-negative samples, the stomach remained normal in size and the dye passed freely into the intestines. The time required to obtain the maximum ratio of gut weight to body weight varied from 30 to 90 min and was dependent upon the concentration of enterotoxin. Heat-labile enterotoxin (LT) had no effect during this period. Based on these findings, the mouse incubation time was reduced from 4 h to 90 min, and the heating of test samples was retained only for confirmation of ST. The location of the dye and stomach distention served as an indicator of positive responses to ST. Incubation of the mice at room temperature (22 degrees C) was found satisfactory.     

Catabolite repression of Escherichia coli heat-stable enterotoxin activity.

The Escherichia coli heat-stable enterotoxin (ST) coded for by plasmid pYK007 (Apr ST+) showed a dependence for cyclic adenosine 3',5'-monophosphate (cAMP) to express ST activity in an adenyl cyclase (cya) deletion mutant; no ST activity was detected in the presence of cAMP in a cAMP receptor protein (crp) deletion mutant or in a double deletion mutant (delta cya delta crp). The cya-crp effect on ST activity could not be accounted for by a modification of the copy number of plasmid deoxyribonucleic acid per chromosome equivalent or by an alteration in the secretion of an active intracellular enterotoxin.     

Thermoactivation of a periplasmic heat-stable enterotoxin of Escherichia coli.

Strains of Escherichia coli that host a plasmid that codes for the heat-stable (ST) enterotoxin showed 160 times more extracellular enterotoxin than intracellular activity. However, when washed bacteria were sonicated and incubated at between 50 and 85 degrees C, an activity similar to that of the ST enterotoxin was detected. No such effect was present in strains lacking the plasmid, in a plasmid ST- mutant, or in chromosomal mutants that lack a cyclic AMP-linked positive regulatory system which previously were shown to yield an ST- phenotype. The thermoactivation was inhibited by iodoacetamide and N-ethylmaleimide; chloramphenicol did not affect the phenomenon. The heat-activated ST-like enterotoxin was localized in the periplasmic space. The results are discussed in relation to the export of the toxin from the periplasm to the outside of the cell.     

Purification and characterization of Escherichia coli heat-stable enterotoxin II.

Escherichia coli heat-stable enterotoxin II (STII) was purified to homogeneity by successive column chromatographies from the culture supernatant of a strain harboring the plasmid encoding the STII gene. The purified STII evoked a secretory response in the suckling mouse assay and ligated rat intestinal loop assay in the presence of protease inhibitor, but the response was not observed in the absence of the inhibitor. Analyses of the peptide by the Edman degradation method and fast atom bombardment mass spectrometry revealed that purified STII is composed of 48 amino acid residues and that its amino acid sequence was identical to the 48 carboxy-terminal amino acids of STII predicted from the DNA sequence (C. H. Lee, S. L. Mosely, H. W. Moon, S. C. Whipp, C. L. Gyles, and M. So, Infect. Immun. 42:264-268, 1983). STII has four cysteine residues which form two intramolecular disulfide bonds. Two disulfide bonds were determined to be formed between Cys-10-Cys-48 and Cys-21-Cys-36 by analyzing tryptic hydrolysates of STII.     

GM1 ELISA method for demonstration of Escherichia coli heat-stable enterotoxin

Abstract A GM1-ELISA for detection of the methanol soluble, heat-stable enterotoxin (ST_a) produced by many enterotoxinogenic E. coli strains has been developed. This ST-GM1-ELISA, which is based on inhibition of binding of anti-ST antibody to GM1-bound ST-cholera B subunit conjugates, is relatively simple and possible to perform with stable reagents and without any complicated equipment. By this method ST_a could be detected in culture filtrates of human E. coli isolates with 100% sensitivity and specificity. The sensitivity of the method for purified ST is considerably higher than that of the conventional infant mouse test and comparable to that of recently described radioimmunoassays for ST.     

Identification of copper-zinc superoxide dismutase gene from enteroaggregative Escherichia coli.

We describe here the identification of sodC gene from enteroaggregative Escherichia coli (EAggEC). A 294 bp gene-specific fragment was amplified from the organism by DNA as well as RT-PCR using primers from bacterial sodC sequences. The metal co-factor present in the protein was confirmed by running samples in native gels and inhibiting with 2 mM potassium cyanide. However, the nonpathogenic E. coli possesses the gene but does not express it. Thus, the presence of copper-zinc superoxide dismutase encoded by sodC was demonstrated for the first time in EAggEC, which means it could be a novel candidate for a virulence marker.     

Affinity purification of functional receptors for Escherichia coli heat-stable enterotoxin from rat intestine.

Active receptors for Escherichia coli heat-stable enterotoxin (ST) were partially purified by ligand-affinity chromatography. The affinity column was prepared by coupling ST to biotin derivatized with an extended N-hydroxysuccinylated spacer arm prior to binding to monomeric avidin immobilized on agarose. Detergent extracts of rat intestinal mucosa membranes were quantitatively depleted of ST binding activity when chromatographed on this affinity matrix. Biotinylated ST-receptor complexes were eluted from affinity columns with 2 mM biotin and these complexes quantitatively dissociated with bile salts. Using this technique, functional ST receptors were purified maximally about 2000-fold, with about 3% of the total activity in crude extracts recovered in these purified preparations. Analysis of affinity-purified preparations by polyacrylamide gel electrophoresis and silver staining demonstrated a major protein subunit of 74 kDa. Affinity cross-linking of these preparations to 125I-ST demonstrated specific labeling predominantly of the 74-kDa subunit. In addition, lower amounts of labeled ST were incorporated into subunits of 164 and 45 kDa, confirming the heterogeneous nature of ST receptors. Purified receptors bound ST in a concentration-dependent fashion, with an IC50 of 10(-9) M. These studies demonstrate that ligand-affinity chromatography can be employed to purify ST receptors. The availability of purified receptors will facilitate further studies of mechanisms underlying ST-induced intestinal secretion.     

Characterization of the recombinant human receptor for Escherichia coli heat-stable enterotoxin.

We report here the molecular characterization of a recombinant cell line (293-STaR) expressing the heat-stable enterotoxin receptor (STaR) from human intestine. We have compared the 293-STaR cell line with the human colonic cell line T84 that endogenously expresses STa binding sites. Scatchard analysis of displacement binding studies revealed a single STa binding site with an affinity (Ki) of 97 pM in 293-STaR compared with 55 pM in T84 cells. Saturation isotherms of STa binding gave a Kd of 94 pM for the cloned receptor expressed in 293 cells and 166 pM for the receptor present in T84 cells. Kinetic measurements of STa binding to 293-STaR gave an association rate constant, K1, of 2.4 x 10(8) M-1 min-1 and a dissociation rate constant, K2, of 0.016 min-1. The half-time of dissociation was 43 min, and the Kd calculated from the ratio of the kinetic constants was 67 pM. The pH profile of STa binding showed that the number of STa binding sites is increased 3-fold at pH 4.0 compared with pH 7.0, with no effect on binding affinity. A polyclonal antibody directed against the extracellular domain of STaR immunoprecipitated two proteins of approximately 140 and 160 kDa from both 293-STaR and T84 cells. Cross-linking of 125I-STa to 293-STaR cells resulted in the labeling of proteins with a molecular mass of approximately 153, 133, 81, 68, 56, and 49 kDa, the two smallest being the more abundant. Similar results have been reported for the STaR present on rat brush border membranes. These data suggest that the STaR-guanylyl cyclase identified by molecular cloning is the only receptor for STa present in T84 cells.     

Disulfide bond formation and secretion of Escherichia coli heat-stable enterotoxin II.

The Escherichia coli heat-stable enterotoxin II (STII) is a typical extracellular toxin consisting of 48 amino acid residues, of which 4 are cysteine. There are two disulfide bonds, one between Cys-10 and Cys-48 and one between Cys-21 and Cys-36. We examined the involvement of DsbA in the formation of the disulfide bonds of STII and the role of each in the secretion of STII. A dsbA mutant was transformed with a plasmid harboring the STII gene, and STII was not detected either in the cells or in the culture supernatant. Reducing the level of STII brought about the dsbA mutation restored by introducing the wild-type dsbA gene into the mutant strain. These results showed that DsbA is involved in forming the disulfide bonds of STII and that STII without these disulfide bonds is degraded during secretion. We substituted these four cysteine residues in vivo by oligonucleotide-directed site-specific mutagenesis. The amino acid sequence of the purified STII (C48S) and pulse-chase studies revealed that two intermolecular disulfide bonds must be formed to be efficiently secreted and that cleavage between amino acid residues 14 and 15 is probably the first step in the proteolytic degradation of STII.     

A recombinant Escherichia coli heat-stable enterotoxin (STa) fusion protein eliciting anti-STa neutralizing antibodies

A recombinant fusion protein consisting of native Escherichia coli heat-stable enterotoxin (STa) and a dimer of a synthetic IgG-binding fragment (ZZ), derived from Staphylococcus aureus protein A was produced in E. coli. The fusion protein (ZZSTa) was secreted in large quantities into the growth medium and recovered by affinity chromatography on IgG-Sepharose. Rabbits immunized with the fusion protein responded by producing high serum levels of anti-STa antibodies that also effectively neutralized STa toxicity in infant mice. The fusion peptide ZZSTa had a substantially decreased toxicity as compared with native STa. A polymeric form of ZZSTa separated by size fractionation was about 100 times less toxic than the monomeric fusion protein, yet both forms had the same capacity to induce neutralizing antibodies. This suggests that modified non-toxic forms of ZZSTa with retained immunogenicity may be produced and tested for their usefulness as functional components in a vaccine against diarrhoea caused by enterotoxigenic E. coli.     

The heat-stable toxin I gene from Escherichia coli 18D.

The heat-stable toxin I gene in the human Escherichia coli isolate 18D is the estA1 allele. The gene is not part of a composite transposon, but inspection of the flanking DNA sequences suggests that it was at one time part of a transposon. The hypothetical transposon originated from an event other than the occurrence that formed Tn1681.     

Synthesis and biological properties of carba-analogs of heat-stable enterotoxin (ST) produced by enterotoxigenic Escherichia coli.

Analogs of a heat-stable enterotoxin (ST) that have a CH2-S linkage instead of an S-S linkage in the molecule were synthesized by conventional methods. The synthetic peptides showed toxicity, assayed as induction of fluid secretion in suckling mice, although their toxicities were hundredth that of native ST. This finding implies that ST is not recognized by its receptor protein through an exchange reaction between its disulfide linkages and thiol-groups of its receptor protein(s), but through hydrophobic or electrostatic interactions.     

Characterization of the receptor for heat-stable enterotoxin from Escherichia coli in rat intestine

The receptor for the heat-stable enterotoxin (ST) from Escherichia coli was solubilized with Lubrol-PX from rat intestinal brush-border membranes and characterized. The binding kinetics and analog specificity of the solubilized receptor were virtually identical to those obtained with the membrane-bound receptor. Furthermore, the regulation of the receptor's affinity by cations was also maintained after solubilization, indicating a conservation of the toxin-binding site after removal of the receptor from its membrane environment. Gel filtration and sucrose density gradient sedimentation studies gave a Stokes radius of 5.5 nm and a sedimentation coefficient of 7.0 S for the solubilized receptor. The isoelectric point of the receptor was determined as 5.5 using Sephadex isoelectric focusing electrophoresis. In all of these separation techniques, the ST receptor showed a single peak of activity that was clearly separated from that of guanylate cyclase. When 125I-ST was cross-linked to brush-border membranes with disuccinimidyl suberate, the affinity-labeled receptor solubilized with 0.1% Lubrol-PX eluted at a similar position as the native receptor on gel filtration chromatography. Analysis of the affinity-labeled receptor by sodium dodecyl sulfate-polyacrylamide gel electrophoresis in the presence of reducing agent and by autoradiography revealed the presence of three specifically labeled polypeptides with apparent molecular weights of 80,000, 68,000, and 60,000. These results suggest that the ST receptor is solubilized by Lubrol-PX in an active form with preservation of its regulation by cations. Also, the ST receptor is separable from particulate guanylate cyclase indicating that the receptor is coupled to the activation of guanylate cyclase by an as yet undefined mechanism. Three subunit peptides may constitute a binding region of the receptor.