Production of Escherichia coli Heat-labile Enterotoxin in Fermenter Dialysis Culture

Production of Escherichia coli heat-labile enterotoxin was investigated with one porcine and one human strain in three different media under different cultivation conditions. Cultivation in aerated fermenters at pH 7·0 yielded 10–20 times more enterotoxin/ml of culture fluid than cultivation in shake flasks. A trypton-yeast extract medium was optimal in fermenter cultures. Comparatively good yields of enterotoxin in fermenters were also obtained in a glucose-salts medium. Continuous feeding of glucose and salts during fermenter cultivation resulted in a lower production of enterotoxin/mg of bacterial cells. Since this decrease in specific yield could be reversed by using dialysis culture, it was concluded that inhibition of toxin formation was due to the accumulation of extracellular low molecular weight metabolites. The highest yield of enterotoxin in dialysis culture was 80 ED₅₀ ml⁻¹ (rabbit jejunal loop test) which is at least eight times more toxin than in ordinary fermenter culture and 80 times more toxin than in shake flask cultures.     

Response of Campylobacter jejuni to combinations of ferrous sulphate and cadmium chloride

On Mueller Hinton (MH) agar, Campylobacter jejuni showed 20.0 and 30.9mm zones of inhibition surrounding discs impregnated with 2.5 and 20 μg CdCl₂ respectively. The minimal inhibitory concentration (MIC) ranged from 0.64 to 3.2 μg CdCl₂/ml of MH agar for four C. jejuni strains. In the presence of 23 μg FeSO₄/ml of MH the MIC increased to a range of 1.5–5.4 μg CdCl₂/ml of MH. Moreover, the numbers of colonies present on MH supplemented with FeSO₄ were greater than on MH without iron. The growth response of C. jejuni in the presence of 0.025% (w/v) FeSO₄ in MH broth was increased about 10000 fold in three of four strains when compared with the growth in unsupplemented MH broth. Zones of inhibition surrounding 20 μg discs of CdCl₂ were 50.6 and 24.4 mm on MH and Campy-BAP media respectively, with cells grown on MH. These results suggest that the blood-containing medium 'neutralized' the biocidal influence of the CdCl₂. In comparison, C. jejuni inoculum from fluid thioglycollate (FT) medium showed smaller zones of inhibition. These decreased from 34.9 mm on MH agar to 19.6 mm on Campy-BAP agar, suggesting that components in the FT growth medium ameliorated the toxic influence of CdCl₂. Atomic absorption spectroscopy analysis indicated mean values (mg/100 g dry weight) of selected metals bound by C. jejuni as: Cu, 10.4; Mg, 146; Na, 2385; Fe, 45.1; Zn, 13.0; and K, 172.     

Immunological properties and ganglioside recognitions by Campylobacter jejuni-enterotoxin and cholera toxin

Abstract The immunological properties of Campylobacter jejuni enterotoxin (CJT) and cholera toxin (CT) were compared by enzyme-linked immunosorbent assay (ELISA) and Western blotting analysis with antiserum against each toxin. Antibody against CJT recognized the 68, 54 and 43 kDa polypeptides of CJT and the 11 kDa subunit of CT, whereas antibody against CT recognized the 68 and 54 kDa polypeptides of CJT and 11 kDa subunit of CT. The immunological reactions between the heterogenous combinations of toxins and the antibodies were weaker than those between the homogenous systems. Thus, different antigenicity was found in CJT and CT at the subunit level, although they possessed cross-reactive epitope(s). The binding of CJT and CT to gangliosides was also examined. CJT and CT bound to GM₁ ganglioside preferentially than to other ganglioside species. However, CJT did not bind to GD₁b in spite of the fact that CT preferred GD₁b. This suggests that both toxins recognize different receptors on the surface of the target cell. This study is the first demonstration of the different properties between CJT and CT in immunological character and ganglioside recognition.     

A recombinant Escherichia coli heat-stable enterotoxin b (STb) fusion protein eliciting neutralizing antibodies

Abstract STb is a heat-stable enterotoxin elaborated by enterotoxigenic Escherichia coli strains associated with weaning piglets and is responsible for diarrhoea in those animals. The maltose binding protein (MBP) of E. coli was used as a carrier for STb, a poorly immunogenic molecule. Constructions were produced where the gene coding for mature STb toxin (MBP-STb) and a fragment of the gene spanning the major epitopic region of STb (AA8–AA30)(MBP-STb₂) were fused to malE gene coding for MBP. The fusion proteins accumulated in the periplasm and were detected with a polyclonal antibody raised against the purified toxin. MBP-STb induced secretion in the biological model whereas MBP-STb₂ was non-toxic. Immunization of rabbits evoked an antibody response to STb for these two fusion proteins. However, only MBP-STb elicited antibodies that effectively neutralized the toxicity of pure STb toxin as determined in the rat loop assay.     

Survival of Staphylococcus aureus and enterotoxin A in glassy and rubbery states of gelatin

About 1% of Staphylococcus aureus cells survived the production of gelatin sheets containing nutrient broth. Those cells which survived showed no evidence of injury. Growth occurred in rubbery state gelatin with a _w values of 0.98 and 0.93; viability decreased during storage at a _w values of 0.89, 0.62 and 0.36 but there was little loss of viability in gelatin at an a _w of 0.25 over 27 d storage at 26°C. Assays for enterotoxin A detected no synthesis of new toxin but no loss in pre-formed toxin. The results suggest that high levels of Staph. aureus and its toxins should be excluded from glassy and rubbery state food products in order to prevent illness.     

Characterization of immunogenic p230 as the toxin A of Clostridium difficile

Abstract For the identification of toxin A of Clostridium difficile , a 2-dimensional gel system was used. In its first dimension, samples were separated in the absence of reducing and dissociating agents, conditions which maintained the activity of the enterotoxin. This was followed by reduction and dissociation in the second dimension where a 230 kDa polypeptide was electroeluted. Rabbits were immunized with polyacrylamide gel slices containing entrapped native toxin A and the denatured 230 kDa protein. As revealed by immunoblotting, neutralizing antisera derived from native protein samples recognized the native toxin, the denatured 230 kDa protein and another polypeptide of about M _r 35 000. Using both types of antisera as probes the pI of the enterotoxin was about 5.9. Preliminary evidence suggests that the enterotoxin is a multimeric protein of 230 kDa and 35 kDa subunits.     

The relationship between cytotoxic effect and binding to mammalian cultured cells of Clostridium perfringens enterotoxin

Abstract The relationship between the cytotoxic effect and binding to different cell lines of Clostridium perfringens enterotoxin was investigated. The enterotoxin released ⁵¹Cr from Vero and MDCK cells labeled with Na₂-⁵¹CrO₄. The effect varied depending upon the dose of enterotoxin and the duration and temperature of the interaction. The enterotoxin gave no effect on FL, KB, or L-929 cells. [¹²⁵I]Enterotoxin bound specifically to Vero and MDCK cells via a binding site of distinct nature, but not to FL, KB, or L-929 cells. The number of the binding sites located on one MDCK cell (1.98 × 10⁶ sites/cell) was three times that on one Vero cell (5.64 × 10⁵ sites/cell), although the binding affinity of MDCK cell ( K _a/ 3.76 × 10⁷ M⁻¹) was 0.1 that of Vero cells ( K _a/ 3.23 × 10⁸ M⁻¹). Binding of the enterotoxin to susceptible cells was temperature-independent.     

Enterotoxin and cytotoxin production by Salmonella enteritidis strains isolated from gastroenteritis outbreaks

Seventy-six Salmonella enteritidis , three Salmonella virchow and one Salmonella braedenrup strains were screened for enterotoxigenicity by using the Chinese hamster ovary (CHO), Y1 adrenal, Vero and HeLa cell tests. All the strains gave positive reactions for enterotoxin production, except one, and the relative sensitivity to the toxin exhibited by the different cell lines was evaluated. An enterotoxic activity has been identified in sonicated extracts of Salm. enteritidis: This enterotoxin was purified on Agarose A-5m (Bio-Rad) and Superose 12 HR 10/30 column. The enterotoxic activity was eluted from the Superose column in the first peak. Like Vibrio cholerae toxin CT and Escherichia coli enterotoxin LT, it was blocked by GM₁ ganglioside, but at a higher concentration. In addition, a cytotoxic factor has been partially identified. The procedure for isolating the cytotoxin included ammonium sulphate precipitation, size-exclusion chromatography and anion exchange chromatography. This cytotoxin factor caused inhibition of protein synthesis in cultured cells, as determined by flow cytometry and [³H]-leucine incorporation. Flow cytometry analysis also showed an activation of CHO cells when exposed to this cytotoxic factor resulting in a state of active growth. Cytotoxic activity was not blocked by gangliosides.     

The amino acid sequence of the β-subunit: Of porcine enterotoxigenic Escherichia coli enterotoxin — Analysis and comparison with literature data

Abstract The amino acid composition and sequence of the β-subunit of heat-labile enterotoxin (LT) purified from a porcine (LT_p) strain, WT-1, of enterotoxigenic Escherichia coli was analysed, and the result was compared with that reported by Dallas and Falkow [Nature 288 (1980) 499-501] who deduced the amino acid sequence of LT_p from data on the DNA sequence of a porcine strain, EWD299. The purified β-subunit of the LT_p of WT-1 was carboxymethylated, succinylated, digested with chymotrypsin and subjected to high performance liquid chromatography (HPLC). The amino acid composition of the peptide peaks from the column were analysed and compared with the data reported by Dallas and Falkow. Only one fraction differed in amino acid composition from that reported, containing lysine instead of methionine. This fraction was found to consist of two peptides with the sequences Lys-Ser-Gly-Glu-Thr-Phe and Arg-Ile-Thr-Tyr. The former peptide is reported to have the sequence Met-Ser-Gly-Glu-Thr-Phe. Thus, the amino acid at position 43 from the N terminus of the β-subunit of LT_p is lysine, not methionine as reported. This is the first report which studied the amino acid sequence of LT_p analysed by protein toxin itself, not by DNA sequence analysis.     

Involvement of prostaglandin E2 synthesis in the intestinal secretory action of Escherichia coli heat-stable enterotoxin II

Abstract The prostaglandin response of mouse intestinal epithelial cells after exposure to Escherichia coli heat-stable enterotoxin II was examined. The quantity of prostaglandin E₂ produced by the intestinal cells was directly related to the dose of heat-stable enterotoxin II. The change in the amount of prostaglandin E₂ over time correlated to that of the volume of fluid released into the intestinal lumen. We then demonstrated that administration of heat-stable enterotoxin II into the intestinal loops of mice induced elevation of arachidonic acid and phosphatidic acid levels in intestinal epithelial cells. These results show that heat-stable enterotoxin II stimulates arachidonic acid metabolism in intestinal epithelial cells and that the synthesized prostaglandin E₂ functions as a mediator of fluid secretion induced by this enterotoxin.     

Production and characterisation of Campylobacter jejuni enterotoxin in a synthethic medium and its assay in rat ileal loops

A synthetic medium for production of Campylobacter jejuni enterotoxin was developed for the purposes of its purification by modifying syncase medium, replacing sucrose with glucose, and supplementing with 0.025% sodium pyruvate, 0.25% sodium metabisulphite, 0.001% ferric chloride and 0.1% L-cysteine, adjusted to pH 6.7. Culture filtrates of a human diarrhoeal and a chicken isolate, grown in this medium caused fluid accumulation ranging between 0.50-0.70 ml/cm of rat ileal loop. The kinetics of toxin production indicated a peak at 36 h and decline by 72 h, coinciding with the period of release of protease by the organism. At least 0.4 rat ileal loop units of enterotoxic activity was recovered per ml of culture filtrates and one unit of this toxin contained only 14 micrograms of protein. The toxin is heat-labile, pH dependent, nonhaemolytic, resistant to trypsin, sensitive to papain and pronase and may show subunit molecular weight analogy with CT subunits.     

Hemagglutinating activity of the heat-labile enterotoxin isolated from porcine enterotoxigenic Escherichia coli

Abstract The hemagglutinating activity of the heat-labile enterotoxin (LT_p) isolated from porcine enterotoxigenic Escherichia coli was studied by hemagglutination inhibition. The hemagglutinating activity of LT_p was enhanced 64–512-fold with pronase- and neuraminidase-treated human erythrocytes although both intact human and sheep erythrocytes were not agglutinated by LT_p at the highest concentration used. No enhancement was found in hemagglutination of neuraminidase-treated sheep erythrocytes by LT_p. Hemagglutination of pronase-treated human type A erythrocytes induced by LT_p was inhibited by melibiose and galactose among mono-, di-, and polysaccharides used as inhibitors. Galactose was a slightly better inhibitor than melibiose. These findings suggest that LT_p is a bacterial lectin specific for galactose.     

Procedures for large-scale production and purification of Clostridium botulinum C1 toxin for preparation of toxoid

Abstract Fortified cooked meat medium containing calcium carbonate (CaCO₃-FCM) supported toxin production of a strain of Clostridium botulinum type C to a level of 2 × 10⁶ mouse i.p. LD₅₀/ml. C1 toxin was purified by sequential steps of acid precipitation from 5-fold diluted culture supernatant in the presence of RNA, 2nd acid precipitation by dialysis, removal of RNA by protamine treatment, removal of excess protamine and bufferisation by ultrafiltration through Amicon PM-30 membrane, sulphopropyl-Sephadex chromatography, and Sephadex G-200 gel filtration. By these procedures, 25 mg or more of highly purified C1 toxin was constantly obtained from a lot of 600-ml culture.     

Serum antibodies reacting with Escherichia coli heat-labile enterotoxin in sera of patients suffering from Campylobacter jejuni infection

Abstract Using an enzyme-linked immunosorbent assay (ELISA) with purified Escherichia coli heat-labile enterotoxin (LT), an increase in serum antibody against LT was demonstrated in 49% of patients infected with Campylobacter jejuni tested. The antibody titers, however, were not as high as those in patients with cholera. This finding suggests that some strains of C. jejuni in patients with diseases due to C. jejuni produce a toxin (or substance) immunologically related to LT in the intestine.     

Effects of pH on the binding of Clostridium botulinum type E derivative toxin to gangliosides and phospholipids

Abstract Clostridium botulinum type E derivative toxin directly bound to gangliosides G_T1b, G_D1a, and G_Q1b but not to G_M1 or G_D1b at pH 5.0 or above, At the same pH values, it bound to negatively charged phospholipids but not to noncharged ones. At pH 4.0, it bound to any of gangliosides and phospholipids including G_M1, G_D1a, and non-charged phospholipids. It bound to ceramide, a hydrophobic component of ganglioside and also to sphingomyelin, a phospholipid containing a ceramide moiety, only at pH 4.0. It bound to ceramide and sphingomyelin less firmly than to other phospholipids at pH 4.0. We assume that botulinum toxin adheres to the neural cell surface mainly by sialic acid-specific and charge-dependent binding possibly aided by nonspecific hydrophobic(toxin)-hydrophobic(lipids, mainly phospholipids) interaction.     

Synthesis on nontoxic, antibody-binding Escherichia coli heat-stable enterotoxin (STa) peptides

Abstract Modified Escherichia coli heat-stable enterotoxin (ST_a) peptides have been synthesized to identify structures that retain the immunological properties of native ST_a but lack toxicity. Two synthetic peptides, corresponding to the 15 C-terminal amino acid residues of ST_a (ST_h) except for the replacement of one or two Cys residues by Ala, had ≥ 35 000-fold reduced toxicity in infant mice despite almost intact activity as compared to native ST_a to inhibit monoclonal anti-ST antibody binding to solid phase ST_a. Three shorter peptides of 6–8 amino acid residues also did not manifest any toxic activity and showed significant but much reduced reactivity with anti-ST_a antibodies.     

A cytotonic, cholera toxin-like protein produced by Campylobacter jejuni.

1. Campylobacter jejuni is a major cause of gastroenteric infection. 2. This organism appears to produce both cytotonic and cytotoxic virulence factors. 3. We report here that culture filtrates of some clinical isolates of C. jejuni induce elongation of Chinese hamster ovary (CHO) cells in vitro but do not cause inhibition of fluid absorption in the rat ileum. 4. These culture filtrates contain low levels of a protein which cross-reacts immunologically with the cholera toxin. 5. The cholera toxin-like protein of C. jejuni behaved identically to cholera toxin on non-denaturing polyacrylamide gel electrophoresis. 6. Under denaturing conditions, however, this protein displayed no subunit structure and a molecular weight of approximately 50 kDa with many higher molecular weight aggregates. 7. In conclusion, isolates of C. jejuni produced small amounts of enterotoxin when grown in vitro. 8. The toxin cross-reacted immunologically with cholera toxin and has a similar native structure, but does not appear to possess subunits.     

Fusarium response to oxidative stress by H2O2 is trichothecene chemotype-dependent

The present study aims at clarifying the impact of oxidative stress on type B trichothecene production. The responses to hydrogen peroxide (H₂O₂) of an array of Fusarium graminearum and Fusarium culmorum strains were compared, both species carrying either the chemotype deoxynivalenol (DON) or nivalenol (NIV). In both cases, levels of in vitro toxin production are greatly influenced by the oxidative parameters of the medium. A 0.5 mM H₂O₂ stress induces a two- to 50-fold enhancement of DON and acetyldeoxynivalenol production, whereas the same treatment results in a 2.4- to sevenfold decrease in NIV and fusarenone X accumulation. Different effects of oxidative stress on toxin production are the result of a variation in Fusarium 's antioxidant defence responses according to the chemotype of the isolate. Compared with DON strains, NIV isolates have a higher H₂O₂-destroying capacity, which partially results from a significant enhancement of catalase activity induced by peroxide stress. A 0.5 mM H₂O₂ treatment leads to a 1.3- to 1.7-fold increase in the catalase activity of NIV isolates. Our data, which show the higher adaptation to oxidative stress developed by NIV isolates, are consistent with the higher virulence of these Fusarium strains on maize compared with DON isolates.     

Improvements in the microtitre GM1 ganglioside enzyme-linked immunosorbent assay for Escherichia coli heat-labile enterotoxin

A variant of the microtitre G_M1-ELISA for Escherichia coli heat-labile enterotoxin was studied. The test was improved by both reducing the assay time from 2½ d to 8 h and by determining the most appropriate G_M1 coating concentration. Coating the plates with >3 μg of G_M1/ml yielded a maximal sensitivity and ensured a linear relationship between the enterotoxin concentration and the extinction observed when using the final assay-procedure. Thus an optimal accuracy was obtained. This ELISA was 4- to 8-times more sensitive than the Vero cell monolayer assay. The sensitivity of this ELISA and of the Chinese hamster ovary cell monolayer assay were identical.