NO753578L - - Google Patents

Description

"Procedure for the manufacture of toxins".

The present invention relates to the purification of E. coli enterotoxin

in connection with the use of purified heat-labile E. coli enterotoxin in vaccines for active immunization and for the production of sera for passive protection against E. coli infections, and for immunoprophylaxis against cholera.

E. coli enterotoxins have long been recognized as important factors in E. coli infections, especially with regard to the large losses of water and electrolytes that occur in diarrheal diseases caused by coli bacteria. The possibility of synthesis of these enterotoxins is not strain-specific, but is controlled by a transferable plasmid which can be transferred by conjugation from a pathogenic E. coli strain to non-pathogenic strains. The appearance of new enteropathogenic serotypes can thus be explained.

Coli infections have so far been treated with coli vaccines

or chemotherapeutic, especially with antibiotics. However, the use of coli vaccines is generally directed against particular serotypes of E. coli and not against their exotoxins, especially enterotoxins.

This method thus provides anti-bacterial protection, but

not an anti-toxic protection. The plasmid-directed mechanism of synthesis also explains the frequent appearance of anti-biotic resistance among enteropathogenic E. coli strains, as R-factors are transferred by the same mechanism as enterotoxin plasmids.

The plasmid-directed synthesis of two forms of E. coli enterotoxin, namely heat-labile (LT) enterotoxin and heat-stable (ST) enterotoxin has been described in the literature (eg, Smith, H.W. and Gyles, C.L., J. Med. Microbiol. 3, 387 (1970)). These forms differ primarily with respect to their acid lability, heat lability and antigenicity, as e.g. indicated in the following table, which also shows a number of known common characteristics:

However, attempts at further investigation of aeoli enterotoxins and their role in coli infections have so far been hindered by the inability to obtain the enterotoxins with a sufficiently high degree of purity. A prerequisite for achieving the required purity is a reproducible and accurate test system for localization and quantification of enterotoxin activity, e.g. after chromatographic separation, but the methods used so far, e.g. the digestive tract models mentioned above have not proven to be sufficient in these respects, as enterotoxins identified by these methods e.g. are found to have a wide distribution of molecular weights. These known test systems are also time-consuming and can only be used for investigations on a small scale.

A more accurate and simpler test of the enterotoxin activity is used in connection with the present invention and is based on the fact that the stimulation of the adenylate cyclase activity, which is indicated by an accumulation of cyclic 3',5'-adenosine monophosphate (cAMP) in cats -myocardial adenylate cyclase preparations are an accurate measure of enterotoxin activity and concentration.

This method for detecting and/or determining E. coli enterotoxin activity comprises incubating a preparation of feline myocardial adenylate cyclase for a predetermined period of time in the presence of a material to be tested for E. coli enterotoxin activity, and measuring the increase in adenylate cyclase activity in the resulting product relative to a control sample.

More particularly, the incubation can be carried out in the presence of a substrate whose conversion to cyclic 3'5<1->adenosine monophosphate is catalyzed by adenylate cyclase, the increase in adenylate cyclase activity being determined by measuring the increase in the concentration of cyclic 3<1>,5'-adenosine monophosphate in the resulting product relative to a control sample.

The feline myocardial adenylate cyclase preparation can be obtained from feline myocardial tissue in a known manner, such as e.g. as described by Levey et al. (G.S. Levey and E. Epstein, Biochem. Biophys. Res. Commun. 33, 990-995 (1968) and 38, 86-92 (1970)).

The preparation may be in particulate form or in solubilized form obtained by treatment with a nonionic surfactant, particularly "Lubrol-PX" (ICI America Inc.) as described by Levey (supra)

and busy in e.g. a sucrose buffer. In the case of solubilized preparations, however, it is necessary to restore the adenylate cyclase activity apparently destroyed by the surfactant. This can easily be achieved by adding phospholipids, e.g. phosphatidylserine.

The adenylate cyclase catalyzed conversion of substrates to cyclic 3<1>,5<1->adenosine monophosphate is well known. Suitable substrates include adenosine triphosphate (ATP) or preferably 5'-adenyl imidodiphosphate (AMP-PNP). The incubation can be carried out in a conventional manner, e.g. as described in the above Levey references. The reaction medium advantageously comprises magnesium ions, e.g. in the form of magnesium chloride, as well as other materials known to promote the conversion of ATP to CAMP and in the subsequent determination of cAMP, e.g. bovine serum albumin and theophylline plus buffer, e.g. Tris-HCl buffer, as well as the substrate, e.g. AMP-PNP, enzyme and test material or control sample. The incubation is advantageously initiated by adding the enzyme, in particulate form or solubilized form as described above, to the remaining reaction medium, and can conveniently be carried out at 37°C and for a period of 3 to 45 minutes, preferably 5 to 30 minutes. The conversion can then be stopped in a known manner, e.g. by addition

of trichloroacetic acid and by maintaining the mixture at 4°C, e.g. for 10 minutes or more. In the resulting product

cAMP can be isolated in a conventional manner. E.g. the precipitate can be removed by centrifugation and the overlying liquid layer is extracted several times with diethyl ether/water and the aqueous layers are evaporated to dryness. The rest can suitably be recorded in a buffer, e.g. acetate buffer pH 4 and sample portions can then be used for subsequent cAMP determination.

Before the processing of cAMP, a known amount of tritiated is added

cAMP appropriate to the mixture (or at least a sample thereof) to determine the correction for extraction recovery in determining the amount of tritiated cAMP in the isolated product.

The determination of cAMP can be carried out in accordance with which

any of the many known methods for this, the preferred method being the radioisotope dilution test system as described by Gllman A.G., Proe. Nat. Acad. Pollock. 67, 305-312 (1970) and as modified as described above, this system being commercially available5 in the form of a sampling kit (Boehringer, Mannheim, Germany).

This test method enables fractions of material from which enterotoxin is to be isolated to be separated and easily tested for enterotoxin activity. If necessary, in order to achieve greater purity, further separations can then be carried out with the active fractions, suitable after combining them, and the active fractions re-identified. The steps of separation and identification can thus be carried out until a homogeneous or sufficiently pure product is obtained.

The use of the enterotoxin determination described has more particularly led to the development of methods according to the invention for the production of E. coli enterotoxin with a high degree of purity. In particular, it has been found that E. coli enterotoxin can be isolated and purified to a high degree by isotachophoresis or affinity chromatography.

The present invention thus provides a method for isolating and purifying E. coli enterotoxin from an impure or pre-purified cell-free culture filtrate from a culture in a fermentation vessel of an enteropathogenic E. coli-

strain where the culture filtrate is subjected to isotachophoretic or affinity chromatographic separation.

The resulting separated fractions, or at least them

containing protein, can then be fixed on a protoxin activity, in particular by means of the described method, and subjected to such further separation steps as may be necessary to achieve the required degree of purity. It will be clear that the active fractions can be determined in advance, e.g. in previous separations carried out under identical conditions, so that the testing step will then not be necessary.

The enterotoxin-containing cell-free culture filtrate can be obtained

in a known manner. E.g. can the enteropathogenic E. coli strain whose enterotoxin is to be isolated be cultivated e.g. for three generations at 37°C in appropriate medium, e.g. trypticase soy broth

(BBL, Lockeysville, Maryland, USA) or preferred one

modified trypticase soy broth obtained by ultrafiltration of the trypticase soy broth, e.g. a 30% (weight/volume) solution thereof. The first stage of the cultivation is suitably carried out in a relatively short period of time, e.g. 4 to 6 hours, the second stage e.g. until the cultivation reaches the middle of the logarithmic phase, and the final stage in e.g. 8 to 10 hours.

The cells can then be harvested in a conventional way, e.g. by centrifugation, and the overlying liquid layer can be filtered and examined for sterility. The crude sterile broth filtrate is then suitably concentrated by ultrafiltration, e.g.

through "Diaflo PM-30" membranes, de-salted by treatment with a mixed layer ion exchange resin, e.g. "AG 501-X8",

freeze-dried (lyophilized) and stored for later use, e.g.

at -20 to -70°C. During this first concentration process, the bacterial products are suitably protected against proteolytic degradation by the addition of a protease inhibitor, e.g. pentamidine isothionate.

The initially concentrated lyophilised sterile cell filtrate can suitably be pre-purified by means of one or more gel filtration steps before the main separation step, especially when this is to take place by means of isotachophoresis.

The gel filtration step or steps may be carried out

in a conventional manner, the appropriate gels and eluents being determined e.g. in previous experiments with the aim of establishing the approximate molecular weights with which the enterotoxin activity is associated. In this connection, it has been found to be advantageous to carry out a first gel filtration separation using "Biogel A-5m" as carrier (separation range 5,000,000 - 80,000) and an ammonium bicarbonate buffer, pH 7.9 - 8.0, as eluent. After determination of the enterotoxin-active fractions, e.g. by the described test method,

the desired fractions can then be pooled and a subsequent gel filtration separation carried out, suitably on a "Sephadex G-75" column (separation range 70,000 to 3,000) using the same buffer or a Tris-HCl buffer (pH 8.0). The active fractions can then be re-determined and merged.

The combined fractions are then desirably further concentrated by dialysis, e.g. against 0.1 M Tris £ -amino-caproate buffer (pH 8.9) to facilitate the preferred subsequent separation by isotachophoresis..

The isotachophoretic separation can be carried out by the method of Svendson and Rose, Science Tools 17, 13 (1970) using polyacrylamide gel as carrier medium and "Ampholdne" carrier ampholytes as buffer and spacer substances. The separation is preferably carried out in simple buffered gel columns, the "Ampholine" carrier ampholytes being mixed with the sample and the conducting electrolyte, which is suitably Tris~£. -amino caproate buffer (pH 8.9). The stopping electrolyte is arranged in a layer on top of the gel and Tris-sulphate buffer (pH 7.1) is advantageously used as elution electrolyte. The separation sample, particularly the concentrated "Sephadex G-75" gel filtration product is suitably bled with sucrose to increase its viscosity, diluted with the •\termination electrolyte and mixed with the carrier ampholytes, and introduced through a capillary tube through the upper layer of the termination electrolyte to form a layer on top of the gel. The gel is advantageously prepared from standard solutions according to Davis, Ann. NEW. Acad. Pollock. 121, 404 (1964.) under preferred

o use of tris-phosphate pH 8.1 as gel buffer, polymerized only by means of photopolymerization.

The eluted fractions can then be tested for enterotoxin activity, e.g. by means of the described method, and the desired fractions can be suitably pooled and re-concentrated by dialysis against tris-^-aminocaproate buffer pH 8.9.

If desired, a further isotachophoretic separation can be carried out to achieve still greater unity, this time using a different molar ratio between lead ion and counter ion, or preferably another elution electrolyte, e.g. tris-2-(N-morpholino)ethanesulfonic acid pH 6.2.

Active fractions can be re-pooled and concentrated, suitably by dialysis as described above. The final product is then subjected to appropriate gel filtration on "Sephadex G-50" to separate the low molecular weight "Ampholine" carrier ampholytes from the isotachophoretically fractionated protein.

An alternative method for isolating the enterotoxin from to

starting with concentrated sterile cell filtrate or from cell filtrate pre-purified by gel filtration is by means of affinity chromatography. Affinity chromatography is a known technique for separating e.g. proteins, which utilize the biological properties of these materials to bind gigands specifically and reversibly. A solution containing the material to be purified is passed through a column containing

an insoluble polymer or gel to which a specific inhibitor or other ligand has been covalently attached. Proteins that have no particular affinity for the ligand will pass through the column unhindered, while those that have an affinity for the inhibitor, on the other hand, will be slowed down in relation to their affinity. The specifically absorbed protein can then be eluted by changing the composition of the solvent so that dissociation occurs.

The separation according to the invention can be carried out on

a way conventional for affinity chromatographic separation, e.g. as described by Cuatrecasas et al., Methods of Enzymology XXII, 345 (1971). For the purposes of the present invention, suitable binding ligands include amphipathic substances, such as e.g. phospholipids or glycolipids with free carboxyl groups, especially gangliosides. These ligands can be bound to an insoluble matrix, which is suitably an agarose derivative, such as e.g. a derivative prepared by attaching an amino copolymer, e.g. poly-L-lysyl-DL-alanine to cyanobromide-activated agarose in a known manner (eg Cuatrecasas, P., J.

Biol. Chem. 245, 3059 (1970) and Sica V. et al., Nature (London), New Biology 244, 36 (1973)), or albumin-agarose. The ligand can be attached to the matrix in a conventional manner. E.g. in the case of gangliosides, the carboxyl group of the terminal sialic acid is linked to the amino groups of the above agarose derivatives, in the presence of a water-soluble carbodiimide reagent, e.g. 1-ethyl-3-dimethyl-3-(3-dimethylamino-propyl)carbodiimide or dicyclohexylcarbodiimide. For the coupling, the gangliosides are suitably used in the form of their N-hydroxysuccinimide esters or their activated mixed anhydrides. The reaction is conveniently carried out in a solvent or solvent mixture, e.g. a cyclic ether, e.g.

dioxane or a mixture of water/dioxane.

Other suitable ligand matrix systems for use as affinity resins in the separation include fatty acid-aminoalkylamino-anarose complexes. These can be produced in a conventional way, e.g. as described in the first Cuatrecasas reference mentioned above. Suitable alkylenediamines for forming the aminoalkylamino-agarose derivatives include hexanediamine, decanediamine and ethylenediamine. Suitable fatty acids include palmitic acid.

The affinity resins described above are then suitably placed in an affinity column and the initially concentrated lyophilized sterile cell filtrate is preferably taken up in a buffer, e.g. Krebs-Ringer bicarbonate buffer pH 7.4, is transferred to the column and developed, suitably under cooling at 4°C. The elution of the enterotoxins can e.g. is carried out with sodium dodecyl sulphate. The subsequent removal of the sodium dodecyl sulfate and the necessary renaturation of the eluted coli enterotoxin can be carried out in a known manner

(eg, K. Weber et al, J. Biol. Chem. 246, 4504 (1971)).

For the production of enterotoxins according to the invention

any enteropathogenic E. coli strains, or non-pathogenic strains made pathogenic by plasmid transfer can be used. Preferred strains include porcine pathogenic E. coli strains P 263, serotype 08:K87, K88ab:Hl9,

P 155, serotype 0149:K91:88a,c, and P 307, serotype 08:K87,K88ab, human pathogenic strains H10407, serotype 078K and H19,

serotype 026:K60:H11, as well as an Ent<+>and Ent" E. coli K12 strain to which the Ent plasmid from the human strain H 19 has been transferred.

In the insulation process according to the invention it is

possible to obtain E. coli enterotoxin with a purity of at least 99%. The method also enables the special forms of enterotoxin, i.e. heat-labile (LT) form and heat-stable (ST)

form can be isolated as these forms both show enterotoxigenic activity in the described sample, although they differ with respect to physical properties, e.g. the molecular weight, and thus appear in different fractions or the separation step, e.g. the separations by gel filtration or affinity chromatography.

The invention provides a heat-labile (LT) E. coli enterotoxin with a molecular weight of 102,000 + 3,000 which has been isolated by the method according to the invention,

especially from cultures of the E. coli strain P 263, and as before

has not been published.

E. coli strains are known to promote diarrheal diseases in humans

and animals, especially in young animals such as foals, 3am and calves. Coli infections are considered particularly responsible for childhood enteritis, most forms of travel sickness and acute tropical diarrhoea. In farm animals, especially in young animals as mentioned, coli infections are among the most significant diseases. In pigs, coli infections occur mainly in connection with disturbances in the walls of the gastrointestinal tract, especially in newborn piglets and in weaned piglets. A massive multiplication of certain serotypes of haemolysing coliforms then occurs.

The purified heat-labile (LT) E. coli enterotoxins obtained

by the invention, especially with a molecular weight of 102,000

+ 3,000 are useful for protection against E. coli infections as demonstrated by their activity in the rabbit and pig digestive tract test (W. Burrows, G.H. Musteikis, J. Infect. Dis. 116, 183

(1966) and H.W. Smith, CL. Gyles, J. Med. Microbiol 3, 403, (1970)), by the rabbit skin test, (J.O. Craig, Proe. Cholera Res. Symp.

Honolulu 24, U.S. Government Printing Office (1965)) and in the fat cell lipase test (P. Cuatrecasas, Biochem. 12, 3567/

(1973)) and by the adenylate cyclase stimulating test as here

is described.

The purified heat-labile (LT) enterotoxins produced by

the present invention is also useful for protection against cholera as shown by the fact that there is a close serological relationship or cross-immunity between E. coli

(LT) enterotoxin and cholera enterotoxin, which have previously been produced in pure form and thoroughly characterized with regard to their physico-chemical properties.

The invention thus enables the use of the purified heat-labile (LT) E. coli enterotoxins in vaccines for active protection and in the production of sera for passive protection against E. coli infections and in immunoprophylaxis against cholera,

in this regard, enterotoxins from different E. coli strains have been found to be immunologically identical or

so closely related that there is cross-immunity, which is a prerequisite for some immunological applications, e.g. for the passive acquisition of anti-sera, obtained in another animal species.

For some purposes, especially for oral administration and for the preparation of a serum, it is necessary that the heat labile (LT) E. coli enterotoxin is present in the form of a stable toxoid, i.e. a preparation that retains the antigenicity of the enterotoxin, but if toxicity is reduced or eliminated. Such toxoids can be prepared by conventional means (eg, R.S. Rappaport et al, Inf. Imm. 9, 304 (1974)). E.g. can conventional cross-linking agents, such as e.g. formaldehyde or preferably glutaraldehyde is added, it being ensured that the immunological properties are maintained during the treatment. E.g. the purified heat-labile (LT) enterotoxin produced by the invention can be incubated with formaldehyde at an elevated temperature, e.g. 35 to 40°C over an extended period of time, e.g. 2 to 6 weeks, or with glutaraldehyde at room temperature to 40°C, using a glutaraldehyde concentration of 0.001 to 0.05 M.

The invention further enables the production of a vaccine comprising an immunologically effective amount of a purified heat-labile (LT)

E. coli enterotoxin or a stable toxoid thereof, for use in immunization against E. coli infections or cholera, wherein an immunologically effective amount of a purified heat labile (LT)

E. coli enterotoxin or a stable toxoid thereof is added.

The invention also enables the production of a serum comprising antibodies induced by administration of a purified heat-labile (LT) E. coli enterotoxin or a stable toxoid thereof, for use in immunization against E. coli infections or cholera by administration of an immunologically effective amount of such a serum.

For use in vaccines, the dosage of the purified heat-labile (LT) E. coli enterotoxin or stable toxoid thereof will of course vary depending on the method of administration, the host individual and the desired treatment. In general, however, satisfactory results are obtained with a dose of from about 0.015 µg to 10 µg per kg of the animal's body weight. For larger mammals, the total daily dose is in the range of about 1 µg to 1 mg and dosage forms suitable for oral administration comprise from about 100 µg to 1 mg of the stable toxoid. The toxin or toxoid is preferably freeze-dried.

The amounts of serum administered for passive immunization will of course be equivalent to the effective dose of toxin or toxoid as indicated above and will of course depend on the antibody content in the serum. These can be determined by measuring the amount of serum necessary to neutralize the amounts of toxin in question, the residual activity being determined in the digestive tract tests as mentioned earlier, further by the previously mentioned adenylate cyclase stimulating test, the passive hemagglutination microtest (S.V. Boyden, J. Exp. Med. 93, 107

(1951) ; and A.B. Stavitski, J. Immunol. 72, 368 (1954)) and the Bentonite flocculation microtest (H.C. Goodman, J. Bozicevich, Immunochemical Methods, 1964, 43).

The vaccines can be composed in a conventional manner, e.g. with or without additives. When additives are used, however, these advantageously include aluminum compounds, e.g. aluminum hydroxide or phosphate or "Alhydrogel" and water-in-oil emulsions, such as e.g. "Adjuvant 65" (a water-in-oil emulsion of antigen in peanut oil emulsified with mannitol monooleate and stabilized with aluminum stearate) and for animal use, Freund's adjuvant.

These vaccines and sera can be used for the protection of

E. coli infections and cholera in the conventional way. E.g.

protection for children and adults against coli-induced diarrhea can be achieved for children by maternal vaccination associated with a vaccine program for children before the end of passive protection, or generally by active immunization with toxoid.

Protection of animals against coli infections is particularly important in

young animals. Because the newborn agricultural animals can pick up antibodies with the mother's colostrum only in the first few hours or days after birth, it is important for their protection

that this colostrum has a high content of specific anti-

bodies. This can be achieved by repeated immunization of the mother

with the appropriate enterotoxin. Immediately in the first few weeks of life, this protection with maternal antibodies is necessary because during this time infections in young animals with enterotoxin-producing E. coli strains can occur very frequently and can be fatal. A further possibility for the protection of young animals is the supply of specific heterologous antibodies, e.g.

in the form of a serum obtained from coli enterotoxin immunized animals.

The following examples shall illustrate the invention.

EXAMPLE 1: Production of coli enterotoxins

a) _ Medium

"Trypticase soy broth" is used as medium. The Peptone Water Medium

contains per liter 17g trypticase peptone, 3 g phyton peptone,

5 g NaCl, 2.5 g dextrose and 2.5 g K2HP04.3H20 and has a pH of 7.3.

Alternatively, a "modified trypticase soy broth", with

the same composition as described above, but where the peptone has previously been ultrafiltered through "Diaflo PM-10" membranes (Amicon) is used as medium. A 30% (weight/volume) solution of "trypticase soy broth" is ultrafiltered, the remainder is discarded and the filtrate is used further. It is advantageous to use these relatively small peptone molecules instead of the untreated product, in order to facilitate the subsequent separation of unchanged medium from the bacterial products in the impure supernatant liquid of the E. coli cultures.

b) _ 2Y£lSiD2_

The freeze-dried E. coli strain P. 263, Serotype 08:K87, K88ab:Hl9

taken up in water, spread on blood agar plates and cultured at 37°C for 12 hours. Cells from the surface growth are transferred by

using a platinum wire loop to a liter Erlenmeyer flask containing 200 ml of the medium (unmodified or modified trypticase-soy broth) on a rotary shaker. The bacteria

grown at 37°C for 4 hours and used for inoculation of a 2 liter shaking culture. When this culture reaches the middle of the logarithmic phase, it is used for inoculation of a 20 liter fermentation vessel in which the medium is stirred at 500 revolutions/minute and aerated at 10 l/minute. To avoid foaming, 1 ml of a 25% (w/v) solution of an anti-foaming agent ("Glanapon 2000", Gross-Busetti, Vienna) is added. The fermentation vessel is incubated at 37°C for 9 hours and the cells are harvested by centrifugation in a cooled ultracentrifuge at 4°C. The overlying liquid is filtered through a 0.45 µm membrane filter and checked for sterility (by smearing on blood agar plates). The sterile filtrate is desalted by treatment with a mixed layer ion exchange resin (eg "AG-501-X8"). Lyophilize and store for further use at -20 to -70°C.

In an analogous manner as described above, the cell-free culture filtrates of other E. coli strains are prepared, e.g. pig-

pathogenic strains P. 155, serotype<1>0149:K91:88a,c and P 307, serotype 08:K87 K88a,b as well as the human pathogenic strains H104 07, serotype 078K? and H19, serotype 026:K60:H11, and also an Ent<+>

and an Ent E. coli K12 strain that has received the Ent plasmid from the human pathogenic strain H19.

EXAMPLE 2: Purification of coli enterotoxin by gel filtration.

The lyophilisate obtained in Example 1 (from which

preferably of the aforementioned strains) are taken up in 0.1 M NH4HC03~buffer (pH 7.9)

(eluent) and the chromatographic separation is carried out on a "Bio Gel A-5m" column (separation range 5,000,000 - 80,000). When using this column (2.5 x 100 cm) with a total volume of 500 ml and approximate void volume of 19.0 ml, a substance peak is eluted at approximately 220 ml corresponding to a theoretical molecular weight of approximately 10 daltons on the basis of Ev/Eo = 1.15. A further substance peak with a distinct shoulder is eluted at 420 to 480 ml which, on the same basis, corresponds to a molecular weight of 4 5

about 9 x 10 to 1.2 x 10 daltons. The heat-labile activity can clearly be found, e.g. using the adenylate cyclase test,

in the shoulder of this top.

The lyophilized active fractions can be pooled and absorbed

in 0.1 M NH4HCO3 buffer (pH 7.9) and a further gel filtration is carried out using a "Sephadex G-75" column (separation range 70,000 - 3,000). The eluted material

in the void volume is clearly distinguished from other proteins and shows biological activity. The active fractions (identified by e.g. the adenylate cyclase test) are combined and lyophilized.

EXAMPLE 3; Purification of coli enterotoxin by isotachophoresis

The cell filtrate (from example 1) or coli (LT) enterotoxin preparation obtained in example 2 can then be subjected to preparative isotachophoresis. For the isotachophoresis separation, a vertical electrophoresis apparatus can be used with 3.3% polyacrylamide gel (e.g. in accordance with P.J. Svendson and Carsten Rose, Science Tools 17 (1), 13 (1970)) as carrier medium. A single buffer gel column is advantageously used as the ampholine carrier ampholytes are mixed with the protein sample as well as the final electrolyte (Tris-3-aminocaproate buffer is pH 8.9). The final one

electrolyte, which connects the system to the upper electrode (cathode) is arranged in layers above the gel. In the lower electrode section of the column (anode), tris-sulphate buffer (pH 7.1) is used as elution electrolyte. The buffer solution is included in an external reservoir and maintained in circulation with an electric differential pump. The sample, mixed with the tris-aminocaproate buffer, is introduced into the column above the gel by means of a capillary tube that protrudes through the upper layer of tris-aminocaproate buffer. To facilitate the introduction of the sample, its viscosity is increased by adding 3% sucrose.

The gel is prepared from a standard solution in accordance with the method of B.J. Davies (Ann. N.Y. Acad. Sei. 121, 404 (1964)) using tris-phosphate (pH 8.1) as gel buffer, and polymerizes only by photopolymerization. The active material is eluted in the "Ampholyte" carrier ampholyte system shortly before or together with the closing electrolyte.

The eluate is transferred via a UV absorption meter to a fraction collector to give an initial assessment of the separation

(proteinaceous fractions). The active fractions are finally determined by the adeholite cyclase test.

EXAMPLE 4; Purification of coli enterotoxin by affinity chromatography.

The column (10 ml of gel in a glass column (1 x 20 cm)) is washed at room temperature, first for two days with 50% (vol/vol) methanol then for 6 hours with 200 ml of 6 M guanidine hydrochloride and then for 3 hours with 100 ml Krebs-Ringer bicarbonate buffer pH 7.4. 50 g of lyophilized culture filtrate is taken up in 150 ml of Krebs-Ringer bicarbonate buffer and thoroughly dialyzed at 4°C against the buffer. The sample is fed to the developing column at a flow rate of 5 ml/hour at 4°C. It is washed with 500 ml of 0.1 M NH4HCO3 buffer pH 8.0. The elution of the biological activity is carried out at room temperature with 0.1 M NH4HC03 buffer pH 8.0, 0.1% sodium dodecyl sulfate (SDS). The SDS-eluted material is separately subjected to the subsequent treatment, for separation of SDS and renaturation, with a urea solution (D. Weber, D.J. Kuter., J. Biol. Chem. 246 4504 (1971)).

The protein solution is brought to 6 M urea by addition

of solid urea, incubated for 30 minutes, and dialyzed against the buffer solution A (0.05 M tris-acetate pH 7.8, 6 M urea). SDS is removed on a column "Dowex" 1x2, equilibrated with buffer A. The development of the column is carried out with the same buffer. The recovery of the protein from the urea solution is carried out

.at room temperature by dropwise dilution in a 0.05 M tris-acetate buffer pH 7.8, so that the final dilution is at least 10-fold. The solution is then concentrated over a "Diaflo UM 10" filter at 4°C in an "Amicon" cell and dialyzed at 4°C

against 0.05 M tris-acetate buffer pH 7.8 to remove the rest of the urea. The column gel can be prepared in the following way: 4

Albumin-agarose (25 ml) is washed with water and suspended in a

50% (vol/vol) aqueous dioxane solution. 50 mg of the desired ganglioside, e.g. monosialosylganglioside GM-, (GGNSLC) or

^3(SGGNSSLC) is added to this suspension and the mixt

Street

shake for 15 minutes at room temperature. After addition of 100 mg of 1-ethyl-3-dimethyl-3(3-dimethylamino-propyl)carbodiimide, the suspension is shaken for 6 hours at room temperature, after which a further 100 mg of carbodiimide is added. After shaking for a further 12 hours, the gel is washed with 500 ml of water, 500 ml of 75% (v/v) aqueous methanol, 250 ml of 6 M guanidine hydrochloride and a further 500 ml of water. The ganglioside content, determined from the difference between added ganglioside and recovered, unreacted ganglioside, is approximately 0.5 mg/ml gel.

The coupling can also be carried out by reacting 20 mg of the ganglioside and 5 mg of dicyclohexylcarbodiimide in 10 ml of dioxane for 30 minutes at 15°C. The mixture is then added to 20 ml of albumin agarose suspended in dioxane to a total volume of 40 ml.

After 15 hours at room temperature, the gel is washed with 500 ml of dioxane, 500 ml of 90% (volume/volume) methanol and finally with 250 ml of 6M guanidine hydrochloride.

o

B)_ IS2^1i!}2_§Y_^Z^Y^£23S§Y§H£2i2iSi^§§£er_of_gangliosi

20 mg of the ganglioside, e.g. as under A) above, is reacted with 2.5 mg of N-hydroxysuccinimide and 2.5 mg of dicyclohexylcarbodiimide in 10 ml of dioxane for 30 minutes at room temperature. The solution is then added to 20 ml of albumin agarose suspended in dioxane to a final volume of 40 ml. After 15 hours of shaking at 4°C, the gel is washed with 500 ml of dioxane, 500 ml of 90% (volume/volume) methanol and with 250 ml of 6M guanidine hydrochloride.

G) ABY§5^§i§e._§Y_fei525i§£_§I}l}Y^Ei^_§Y_2aD2ii2§i^

100 µl of 0.1 M N-methylmorpholine in tetrahydrofuran is added

a solution of 20 mg of the ganglioside, e.g. as under A) above, in anhydrous tetrahydrofuran. The mixture is stirred for 10 minutes at 0°C and 100 µl of 0.1 M isobutyl chloroformate in tetrahydrofuran is added. The mixture is kept at 0°C for 20 minutes and then added to 20 ml of albumin agarose suspended in dioxane to a final volume of 40 ml. The mixture is then kept at room temperature for 15 hours and the resulting gel is washed as described under A)

above.

In an analogous manner to examples A, B and C above, gels can be prepared by using a copolymer of poly-L-lysyl-DL-

alahin and cyanobromine deactivated agarose instead of albumin-agarose.

D)_ First il lin2_ay_f ett sy £e-a22§£2§§I^f EiY§t3S2mEl§]S§

Aminoethylamino, aminohexylamino and aminodecylaminoagarose derivatives can be prepared according to the method of Cuatrecasas and Anfinsen Methods in Enzymology 22, 345 (1971) by activating "Sepharose 4B" with 200 mg CNBr/ml resin at pH 11 - 11.5,

and stirring at pH 10 and 3°C for 15 hours with the corresponding diamine. The molar ratio of hexane- or decane-diamine to CNBr is 1, while the ratio of ethylenediamine to CNBr is 5. The aminoalkylaminoagarose derivative matures for 5 days and reacts

with IM 2-aminoethanol at room temperature to saturate any remaining activated groups of the "Sepharose" substance. The resulting derivative contains approximately 20 mol of free amino group/

ml of agarose. The desired fatty acid, e.g. palmic acid,

is coupled by stirring for 3 days of a mixture of agarose derivatives in 1.5 times its volume of a 0.1 M solution of the sodium soap of the fatty acid at pH 10 and 37°C, in the presence of 50 mg of ethyl-3-(3-dimethylaminopropyl)carbodiimide per ml "Sepharose". 25%

ethanol is added to improve the solubility of the soap. The product is washed at 37°C with 50% (volume/volume) ethanol, ethanol-0.075 M sodium phosphate (1:1), pH 2.4, and with ethanol 0.05 N NaOH.

Remaining unreacted groups are saturated by acetylation with acetic anhydride.

EXAMPLE 5: Molecular weight determination by gel filtration

For determining the molecular weight of E. coli P263 (LT) enterotoxin, the method of gel filtration can be used. Single column with

"Sephadex G 150" is calibrated by the method of P. Andrews (Biochem. J. 96, 595 (1965)) using horse heart cytochrome C (molecular weight 1.3 x 10 4 ), bovine pancreatic chymotrypsinogen A, (2.4 x 10 4 ), chicken albumin (4.5 x 10 4 ), bovine serum albumin

(6.8 x 10 4 ), rabbit muscle addolase (1.47 x 10 5 ) and bovine liver catalase (2.2 x 10^) as comparison proteins (markers).

The proteins are monitored by their transmission at 280 mH.

For the preparation of the standard curve, the tritium is further monitored

water (1.11 x 10 4 dpm) by liquid scintillation and a saturated solution of blue dextran is used and its optical density at 620 nyu is used for the detection.

The "Sephadex" column is equilibrated with 50 mM trischloride and pH

7.5, 100 mM KC1 and 0.01 M 2-mercaptoethanol.. Each of these materials, with a total volume of 0.7 ml, is transferred to the column and eluted in 0.9 ml fractions. The experiment is carried out at 4°C. The molecular weight of E. coli P263 (LT) enterotoxin is determined as 102,000

+ 3,000 (the limits of experimental error).

EXAMPLE 6: Molecular weight determination by polyacrylamide gel

electrophoresis in sodium dodecyl sulfate

Polyacrylamide gel electrophoresis in 0.1% sodium dodecyl sulfate is carried out at pH 7.2 in 7.5% acrylamide by the method according to A.L. Shapiro, E. Vinuela, J. Maizel (Biochem.

Biophys. Res. Commun. 28, 215 (1967)) and J. Weber, M. Osborn (J.B.C. 244, 4006 (1969)). The following proteins are used as standards: -

Bovine pancreatic trypsinogen (molecular weight of polypeptide chains =

2.4 x 10 4 ), bovine cardiac lactate dehydrogenase (3.6 x 10 4 ), bovine serum albumin (6.8 x 10 4 ), rabbit muscle phosphorylase

(9.4 x 10^) and E. coli @ -galactosidase (1.3 x 10^).

A 1% solution of the sample (E. coli P 263 (LT) enterotoxin)

in sodium dodecyl sulfate and 2-mercaptoethanol is boiled for 3 minutes to remove the possibility of proteolytic degradation. The sample is then dialyzed against a solution consisting of 0.1% in sodium dodecyl sulfate and 2-mercaptoethanol, 0.01 M sodium phosphate,

pH 7.1. The comparison proteins are treated in the same way.

The molecular weight of E. coli (LT) enterotoxin is determined by this method as 102,000 + 3,000 (the limits of experimental error).

EXAMPLE 7: Determination of isoelectric point by isoelectric

focusing in sucrose density gradient

To determine the isoelectric point for purified enterotoxin, isoelectric focusing is carried out in a glass column

(type 8100, 110 ml) according to the method of 0. Vesterberg and

H. Svensson (Chem. Scand.. 20, 820 (1966)). The anode electrolyte (phosphoric acid) is located at the bottom of the device. The "Ampholin"-bearing ampholytes cover a pH range from 3 to 10. Optical density is measured at 280 µm by emptying the column through the measuring cell of a "Uvicord II" UV absorpti6nmeter. The isoelectric point of the protein is determined at 4°C by measuring the pH of the eluted fractions with a combined microelectrode.

The isoelectric point of E. coli P263 (LT) enterotoxin is determined by this method as 6.95.

EXAMPLE 8; Determination of the isoelectric point at

isoelectric focusing in thin-layer polyacrylamide gel.

Enterotoxin is isofocused on a 5% polyacrylamide gel plate i

according to Z.L. Awdeh et al. (Nature (London) 219, 66 1966))

as modified by J. Bours and W.J. van Doorenmaalen

(Science Tools 17, 36 (1970)). A polyacrylamide gel

26.x 12.5 cm and 1 mm thickness is produced with a final concentration of 2%. "Ampholin" carrier ampholytes with a pH range of 3 - 5 5 - 7 and 7 - 10 in the ratio 1:1:1. The gel is polymerized for 2 hours with riboflavin as photocatalyst. 10 to 25 ml of a 0.1% solution of the sample in 2% "Ampholin", pH 3 - 10, is placed on 10 x 5 mm Whatman 3 MM paper strips which are layered on the gel. The isoelectric focusing is carried out at 10°C in an electrophoresis apparatus. Before the gel plates are attached to the electrodes, the electrodes are covered with 0.5 x 25 cm filter paper strips that are moistened with 0.1% (volume/volume) ethylenediamine solution (cathode). The output power is 50 mA (and nacsets to 28 mA) for 2.5 hours at 210 volts (raised to 1080 volts). The pH gradient is determined using a combined microsurface electrode. After isofucusing, the proteins are precipitated with 14% trichloroacetic acid at 60°C. The gel is then washed at room temperature with 10%, 5% and 3% trichloroacetic acid for 24 hours to remove the carrier ampholytes.

The gel is then stained with a 0.05% solution of "Coomassie Brilliant Blue R-250" (dissolved in methanol/acetic acid/water-45/9/46) for 2 hours, followed by washing in the same solvent. For renewed swelling, the gel is placed in acetic acid/water ^(9:91) and then photographed.

The isoelectric point of E. coli P263 (LT) enterotoxin by this method is 6.95.

EXAMPLE 9: Effect of pH.■

6 mg servings of coli enterotoxin (especially E. coli P263

(LT) enterotoxin) in 24 ml Tyrode's solution is placed in 6 test tubes. The contents of the tubes are brought to the respective pH 1,3,5,6,7

and9 with concentrated HCl or 6N NaOH. The test tubes are incubated at 4°C and each tube is brought to pH 7. The contents of each tube are then tested in rabbit and pig digestive tract experiments at 2 ml (0.5 mg) per intestinal loop.

The enterotoxigenic activity of E. coli P263 (LT) enterotoxin is highly acid labile. The activity is completely destroyed at pH4 or less and is already affected at pH6.

EXAMPLE 10: Effect of temperature.

5 mg portions of the purified enterotoxin (especially P263 (LT) enterotoxin) are each taken up in 20 ml of thyrode solution and heated at 40, 50, 60, 65 and 70 and 100°C for 30 minutes. The heated solutions are tested in rabbit and pig digestive tract experiments with an untreated solution as a control. 2.0 ml (0.5 mg) is added per intestinal loop.

/The biological activity of the enterotoxin is completely destroyed by heating for 30 minutes at 65°C. The heat lability is already evident at 50°C and its complete destruction of

the activity mainly takes place already at 60°C. -

EXAMPLE 11; Demonstration of immunological identity or

cross-immunity.

Freeze-dried E. coli strain P 155 serotype 0149:K01, K88a,

c or human pathogenic E. coli strain H 10407, or Ent E. coli strain K 12 and Ent E. coli strain K 12 are used

to produce a cell-free culture filtrate as described above for strain P 263. The cell filtrate serves as a sample with an unknown composition of antigen Ag-sX*

The identification and quantification is carried out using a reference antigen, Ref-Ag (the pure coli enterotoxin from strain P 263 or H 19) and a reference antiserum Ref-Ab (monovalent antiserum against colienterotoxin).

The two questions, whether Ag-rX contains an antigen that is "identical" to Ref-Ag, and if so how the concentration of this antigen can be compared with the known concentration of antigen Ref-Ag, can be answered by "cross-immunoelectrophoresis".

(H.G.M. Clarke and T. Freeman, Clin. Sei. 35, 403 (1968)) and "tandem cross electrophoresis" (Kroll J., A manual of Quantitative Immunoelectrophoresis, N.H. Axelsen et. al. Universitetsforlaget, Oslo, page 57 (1973) , suppl. 1/1973, Scandinavian J. Immun).

With reference to the attached drawings, fig. 1 a "cross-immunoelectrophoresis" image of Ref-Ag, developed against Ref-Ag. Only a precipitate appears.

In fig. 2 Ag-vX, is used; instead of Ref-Ag. On the basis of the position of the deposit in fig. 2 it can be seen that Ag-X contains an antigen identical to Ref-Ag. This conclusion is reinforced by reaction of the "identity" using "pooled precipitates". This experiment can be carried out using "tandem cross-immunoelectrophoresis" (Fig. 3) .

By comparing fig. 3 with fig. 1 it can be seen that Ag-X contains Ref-Ag and that the concentration of the latter in Ag-X is •

less than in Ref-Ag. The quantification of the merged peaks can be carried out by standard retodes (J. Kroll, J.Clin.

Lab. Invest 22, 112 (1968));

If the culture filtrate Ent E. coli K 12 strain is used i

this experiment, instead of Ag-X, no precipitate occurs in fig.

2. This confirms that the sediment that appears in fig. 2 is associated with the enterotoxins in pathogenic E. coli strains.

The same experiments can be carried out to demonstrate the immunological identity or cross-immunity of human-pathogenic and porphin-pathogenic E. coli enterotoxins.

In this test, pure E. Coli P 263 (LT) enterotoxin serves as the Ref-Ag and the monovalent antiserum against P 263 enterotoxin as the Ref-Ag. The human pathogenic enterotoxin H 19 serves as a sample with an unknown composition of Agr)X\ As described for fig.. 2

a single precipitate also appears in the image for "cross-immunoelectrophoresis" of Ag-X developed against Ref-Ag in the same position as in fig. 1. On the basis of the position of the precipitate, it can be concluded that Ag-X contains an antigen identical to Ref-Ag. The "identity" reaction using pooled precipitates in "tandem cross-immunoelectrophoresis" shows (as in Fig. 3) that Ag-X and Ref-Ag are serologically identical.

Based on this, human pathogens or swine pathogens can

coli enterotoxins are used as antigens for immunization or for the production of toxoids.

By applying the same experiment, the question of whether

where there is a partial or complete immunological identity between E. coli enterotoxin and V. cholerae enterotoxin is investigated. The pure E. coli P 263 (LT) enterotoxin serves as Ref-Ag while E. coli P 263 antitoxin is used as Ref-Ab and

the pure V. cholerae toxin serves as the antigen sample Ag-X.

The results are shown in fig. 4, 5 and 6. I. fig. 4 it appears

as previously a single precipitate. In fig. 5 a simple precipitate appears again, but in a different position and with a different form

from the one in fig. 4.

From fig..6 ("tandem-cross-immunoelectrophoresis") it is clear

that there is a distinct and at least partial identity between Ref-Ag and Ag-X (E. coli enterotoxin and V. cholera enterotoxin).

" EXAMPLE 12: Active immunization.

Breeding sows are artificially inseminated and immunized, for the first time one month before the estimated time of birth: 0.95 mg of purified E. coli enterotoxin (especially P 263 (LT) enterotoxin) together with 15 ml of complete Freund's adjuvant (7.5 ml of additive + 7.5 ml of NaCl- buffer 0.05 M, 10~<2>M ^-mercaptoethanol) is added i.m. After 2 weeks, a further 0.35 mg enterotoxin is added in 5 ml 0.9% NaCl solution. Lv.

Due to the serological identity between porcine and human pathogenic coli enterotoxins, the purified (LT) enterotoxin of strain H 19 can e.g. is used alternatively.

6, 12 and 24 hours after the breeding sows have given birth to the piglets, the colostrum is aspirated and collected. 56 hours after birth, a blood sample is taken for a serum sample.

The determination of the anti-body content can be carried out by bentonite flocculation microtest (SBF test). The following results are obtained:

A diluted (1:2) porcine serum can be used in neutralization studies in the rabbit alimentary canal loop test. After using a toxin dose of 0.25 mg/loop (ie 3 toxin units) more than 60% inhibition is achieved. Similar experiment with colostrum

(6 hours post partem) shows in the rabbit model a complete neutralization of the toxin activity.

(1 toxin unit is defined as ED^q in the alimentary canal loop test or in the adenylate cyclase system).

EXAMPLE 13: Preparation of a toxoid.

To achieve optimal conditions for the inactivation, on the basis

of a molecular weight of 102,000 for the toxin molecule, the concentrations of glutaraldehyde are initially varied from 50 to 1000 mol/mol toxin. The reaction is carried out using a toxin concentration range of 100 to 1000 µg/ml, at 30°C in a 0.065 M PBS buffer pH 7.8 and for different incubation times. The reactions are stopped by dialysis against 100 times larger volumes of PBS buffer pH 7.8. At a toxin concentration of over 600 µg/ml, the reaction product begins to become insoluble with increasing concentration of glutaraldehyde. If the toxin concentration is kept between 100 and 600 µg/ml and the glutaraldehyde concentration is varied between 50 and 400 mol/mol toxin, the reaction product remains in soluble form.

The adenylate cyclase test shows that optimal detoxification is achieved in the pH range 7.8 to 8.2. If a toxin concentration of 400 to 600 mg/ml is chosen, residual toxicity is reduced

(measured after incubation for 72 hours at 30°C) to the tenth

by a two-fold increase in the glutaraldehyde concentration between 50 and 200 mol/mol toxin. If the glutaraldehyde concentration is raised from 200 to 400 mol/mol toxin, the residual activity cannot be detected even with the most sensitive methods (the adenylate cyclase test).

To determine preparations of inactivated coli enterotoxin (toxoid) quantitatively, toxin units are determined by the method of Craig (J.P. Craig 1971, Cholera toxins, pages 189 - 254 in S. Kadis et. al., Microbial Toxins 2, Academic Press Inc.,

New York).

This determination is based on the toxins' ability to react

in vitro with the antitoxin. The test determines the reduction in neutralizing capacity for a known amount of antitoxin, after incubation with toxoid. A series of test tubes, each containing the same amount of antitoxin (2 AU/ml) and an equivalent dilution of toxoid are each incubated at 37°C for 30 minutes.

A further series of test tubes serves as a control, in that

each tube contains the same amount of antitoxin (2 AU/ml) to which buffer has been added and which is incubated under the same conditions. After the incubation, identical volumes with a 0.15 log serial dilution of toxin are added to the test tubes and the determination is carried out using the adenylate cyclase test.

The shift of the 50% values on the dose-activity curve is then used to determine the amount of free and bound antitoxins in the antitoxin-toxoid mixture. A TU is the amount of toxoid that binds one AU of the antitoxin (AU = antitoxin unit).

EXAMPLE 14: The adenylate cyclase test.

Preparation of the adenylate cyclase preparation in particulate form is carried out by the method of Levey (G.S. Levey and E. Epstein, Biochem. Biophys. Res. Commun. 33, 990-995 (1968) and 38, 86-92 (1970)) using cats -myocardial tissue. The activation of adenylate cyclase by the coli enterotoxin is determined by the radioisotope dilution test (A.G. Gilman, Proe. Nat. Acad. Sic. 67, 305-312 (1970)) using 5-adenylimido-diphosphate (AMP-PNP) as substrate instead of ATP. AMP-PNP is purified before use by passing it through a column with "Dowex"

50 x 2 in H+<->form.

The reaction medium comprises 50 mM Tris-HCl buffer (pH 7.4),

5mM MgCl2, 10-mM theophylline, 0.1% bovine serum albumin and 2mM AMP-PNP. The incubation volume amounts to 0.1 ml in total and the enzyme concentration is 0.15 mg protein per test. The incubation is started by adding the enzyme and is carried out for 30 minutes at 37°C. To terminate the reaction, 500 ml of 7.5% trichloroacetic acid is added and the mixture is kept at 4°C for

10 minutes. The precipitate is separated by centrifugation and the overlying liquid layer is added to 5 ml of diethyl ether saturated with water. The layers are separated by centrifugation and the ether phase is removed. After a total of three ether extractions, the aqueous is evaporated

solution to dryness at 60 to 70°C on a water bath with an air current. The residue is dissolved in 1 ml of 0.2 M acetate buffer (pH 4.0). The amount of vcAMP formed, which is a direct measure of the degree of activation of adenylate cyclase by coli enterotoxin, is determined by means of the previously mentioned test (available as a test kit from Boehringer Mannheim, West Germany.

EXAMPLE 15: Preparation of a vaccine.

An enterotoxoid solution (especially E. coli P 263 (LT) enterotoxoid) (lmg/ml) is thoroughly dialyzed against a solution ay,

NaHPO 4 (0.07 M) and filtered on a Seitz sterilizing filter.

The same volume of a sterile 0.07 M CaC^ solution is added

then to produce a vaccine absorbed on calcium phosphate with a pH of 6.8 to 7. For inoculation, the vaccine is diluted and added in doses of e.g. 1 to lOO^ug.

EXAMPLE 16: Oral delivery form.

Freeze-dried quantities of enterotoxoid (specifically P 263(LT) enterotoxoid) (eg, 100 µg-1 mg) are pressed into a dragee core with a suitable filter (which may also serve as an additive) and coated with an anionic polymer. from methacrylic acid and methacrylic acid esters (e.g. "Eudragit L and S") to which a plasticizer has been added. Coatings of this type protect against the effects of gastric juices because they are insoluble in the acidic pH range and water-impermeable. They are thus resistant to the gastric juices. After passing through the stomach, the coatings will dissolve.

"Eudragit C" films will e.g. dissolves in the neutral pH range from pH 6 (duodenal juices), while "Eudragit-S" films first dissolve in the lower small intestinal areas (from pH 7).

Claims (33)

1. Method for isolating and purifying E. coli enterotoxin from an impure or contaminated cell-free culture filtrate of an enteropathogenic E. coli strain from cultivation in a fermentation vessel, characterized in that the culture filtrate is subjected to isotachophoretic separation. 2. Method as stated in claim 1, characterized in that the culture filtrate is subjected to isotachophoretic separation using polyacrylamide gel as carrier medium and "ampholine" carrier ampholytes as buffer and spacer substances. 3. Method as stated in claim 2, characterized in that the closing electrolyte is tris-£-aminocaproate buffer pH 8.9 and the elution electrolyte is tris-sulfate buffer pH 7.1. 4. Method as stated in claims 1 - 3, characterized in that the resulting separated fractions are tested for E. coli enterotoxin activity, one or more fractions with the desired activity being subjected to a further isotachophoretic separation. 5. Method as stated in claim 4, characterized in that the further isotachophoretic separation is carried out using polyacrylamide gel as carrier medium, "Ampholine" carrier ampholytes as buffer and spacer substances, tris-^-aminocaproate buffer pH 8.9 as final electrolyte and tris- 2-(N-morpholino)ethanesulfonic acid pH 6.2 as elution electrolyte. 6. Method as stated in claims 1 - 5, characterized in that the resulting E. coli enterotoxin-containing fractions are gel filtered to remove ampholine carrier ampholytes. 7. Method as stated in claim 6, characterized in that the fraction is gel filtered over "Sephadex G-50". 8. Method for isolating and purifying E. coli enterotoxin from an impure or pre-purified cell-free culture filtrate of an enteropathogenic E. coli strain from a culture in a fermentation vessel, characterized in that the culture filtrate is subjected to affinity-chromatographic separation. 9. Procedure as stated in claim 8, characterized in that the affinity ligand is a phospholipid or glycolipid with free carboxyl groups. 10. Method as stated in claim 9, characterized in that the affinity ligand is a ganglioside. • 11. Method as stated in claim 10, characterized in that the ganglioside is ganglioside GM^ (GGNSLC) or GQt t 3 (SGGNSSLC). 12. Method as set forth in claims 8-11, characterized in that the insoluble matrix to which the affinity ligand is bound to form the affinity resin is albuminag^ arose or an agarose derivative prepared by linking a poly-L-lysyl-DL-alanine to cyanobromide-activated agarose.. 13. Method as stated in claim 12, characterized in that the affinity resin is a fatty acid aminoalkylaminoaggarose complex. 14. Method as stated in claim 13, characterized in that the affinity resin is a complex of palmitic acid and an aminoalkylamino agarose derivative formed from Sepharose and ethylenediaminehexanediamine or decandiamine. 15. Method as stated in claims 8-14, characterized in that the elution is carried out with sodium dodecyl sulfate, with the resulting eluted enterotoxin then being renatured. 16. Method as stated in claims 8-15, characterized in that the impure cell-free culture filtrate is pre-purified by gel filtration. 17. Method as stated in claim 16, characterized in that the impure cell-free filtrate is pre-purified by gel filtration and 'Biogel A-5m'. 18. Method as stated in claim 17, characterized in that the gel filtration over "Biogel A-5m" is carried out using an alkaline buffer. 19. Method as stated in claim 18, characterized in that the buffer is ammonium bicarbonate buffer pH 7.9 to 8.0. 20. Method as stated in claims 17 to 19, characterized in that the resulting fractions are tested for enterotoxin activity, one or more fractions with the desired activity then being subjected to a further pre-purification by gel filtration. 21. Method as stated in claim 20, characterized in that the further pre-purification by means of gel filtration is carried out over "Sephadex G-75" using an alkaline buffer. 22. Method as stated in claim 21, characterized in that the alkaline buffer is Tris-HCl buffer pH 8.0. 23. Method as stated in claims 1 to 22, characterized in that - the enterotoxin-containing cell-free culture filtrate is obtained by cultivation for 3 generations of enteropathogenic E. coli strain in a trypticase soy broth or an ultrafiltered trypticase soy broth as medium.. 24. Method as stated in claim 23, characterized in that the first step in the cultivation is carried out for 4 to 6 hours, the second step until the cultivation reaches the middle of the logarithmic phase and the third step for 8 to 10 hours. 25. Method as stated in claim 23 or 24, characterized in that the crude broth filtrate is ultrafiltered in the presence of pentamidine isothionate and then desalted. 26. Method as stated in claims 1 to 25, characterized in that the separated fractions are sealed with respect to E. coli enterotoxin activity by incubating a preparation of cat myocardial aladenylate cyclase for a predetermined period of time in the presence of the sample to be tested, and the increase in adenylate cyclase activity in the resulting product is measured relative to a control sample. 27. Method as stated in claim 26, characterized in that the incubation is carried out in the presence of a substrate whose conversion to cyclic -3'-5'-adenosine monophosphate is catalysed "by adenylate cyclase, the increase in adenylate cyclase activity being determined by measuring the increase in the amount of cyclic 2'-5'-adenosine monophosphate in the resulting product, relative to a control. 28. Method as stated in claim 27, characterized in that the substrate is adenosine triphosphate or 5 <1-> adenylimidodiphosphate. 29. Method as stated in claim 27 or 28, characterized in that the incubation is carried out in a reaction medium comprising Tris-HCl buffer pH 7.4, magnesium chloride, bovine serum albumin and theophylline. 30. Method as stated in claims 26 to 29, characterized in that the incubation is carried out at 37°C for 5 to 30 minutes. 31. Method as stated in claims 26 to 30, characterized in that the preparation of cat myocardial adenylate cyclase is obtained in particulate form from cat myocardial tissue. 32. Method as stated in claims 1 to 31, characterized in that the enteropathogenic E. coli strain is P 263, P 155, P 307, H 10407, H 19 or Ent<+> or Ent K12., the Ent plasmid having been transferred from H19. 33. Method as stated in claims 1-32, characterized in that the enteropathogenic E. coli strain is P 263.