IE62402B1 - Process for Extracting and Purifying Epidermin - Google Patents

Abstract

A process for preparing the polypeptide epidermin with an antibacterial action is described. The process is distinguished by an economic procedure and relatively high epidermin yields.

Description

PROCESS FOR EXTRACTING AND PURIFYING EPIDERMIN . f) The invention relates to a process for obtaining the polypeptide epidermin.

The antibiotic epidermin is known from European Patent Application 85 113 908.9 (Publication No. 181 578).

This specification describes a process for obtaining, isolating and purifying this substance, starting from a culture broth obtained with the aid of a resistant strain of Staphylococcus epidermidis. The resistant strain was deposited on 26.10.1984 at the .

Deutsche Sammlung von Mikroorganismen under the Number DSM 3095. To isolate it, the active component is either concentrated by extracting the culture filtrate, freed from cells and lime, with n-butanol, followed by evaporation of the butanol extract, dissolving the residue in methanol and stirring it into an excessive quantity of cold diethyl ether in order to separate off the lipidic contaminants, the activity remaining in the precipitate, or in order to concentrate the epidermin the centrifuge culture filtrate is adsorbed on Amberlite® XAD-8 or on related types of this polymer based on acrylic ester (made by Serva), the adsorbed epidermin being released from the resin by means of methanol/conc. hydrochloric acid (99:1)' and isolated by evaporation from the hydrochloric acid methanol solution, after neutralisation with ammonia. Subsequent chromatography of the Amberlite® XAD eluate or of the butanol extract freed from lipids, using Sephadex LH-20 with methanol/acetic acid (95:5), separates a large number of small peptides, amino acids and salts from the antibiotic in the medium. In subsequent multiplicative counter-current distribution by the Craig method, the antibiotic is left behind at the start in a first liquid-liquid distribution using the system nbutanol/ethyl acetate/O.l N acetic acid (3:1:3). In a 2 4 0° - 2 second Craig distribution with the neutral system 2butanol/0.05 N ammonium acetate (1:1) the antibiotic is in the centre of the apparatus. The ammonium acetate is removed by lyophilisation under high vacuum, and after freeze-drying the epidermin formed is obtained as a white powder which is uniform in all the thin layer systems used. In this way, 2.6 g of lyophilised epidermin can be obtained from 80 litres of culture filtrate (with adsorption on Amberlite® XAD-8, gel chromatography on Sephadex LH-20 and multiplicative distribution according to Craig).

As described in the above-mentioned specification, the producing organism Staphylococcus epidermidis DSM 3095 is grown aerobically at 37 °C in a complex medium made up of 2 to 4% meat extract, 1 to 3% malt extract and 0.25 to 1% CaCO3 or 0.25 to 0,5% Ca(OH)2 (percent by weight). The maximum antibiotic activity is reached after 18 to 23 hours.

It has been found that the antibiotic epidermin can be obtained in substantially higher yields and by a simpler method if the antibiotic formed in the culture filtrate or in the culture broth, still containing the microorganism, is adsorbed on Amberlite® XAD-1180 or Amberlite® XAD-16 or related types of polymer based on styrene-divinyl copolymers either through a column or, in the case of the culture broth, by adding the resin in small amounts. The active component is released from the resin by elution with methanol/0.01 N hydrochloric acid (9:l,v:v), the eluate is adjusted to a pH of 5.3 to 5.8 and placed on a weak cation exchanger such as Amberlite® IRC-50 or related resins such as Amberlite® IRC-84, for example, whilst the ion exchange may take place in a column or by introduction in small amounts, and in the latter case preferably by twice adding a quantity of 3% by volume of the resin to the XAD eluate within one hour. After adsorption the resin is introduced into a column, non-bound substances are then washed out with a 0.05 N sodium phosphate buffer solution in water at pH 7.0, the epidermin is then eluted with a solution of 80% of the above-mentioned sodium phosphate buffer, which is 1.5 N with regard to common salt and contains 20% by volume of methanol, at a pH between 6.0 and 8.0, preferably at pH 7.0.

To remove the salt the eluate is adjusted to a pH of 6.0 and is then added in small amounts for adsorption to a resin of the above-mentioned styrene-divinyl copolymer type, e.g. Amberlite® XAD-1180. The salts are washed out with water and the bound epidermin is then eluted with methanol/50% acetic acid (9:1, v:v) . After removal of the solvent and freeze drying, a lyophilised crude epidermin is obtained. From this, pure epidermin can be isolated by preparative HPLC chromatography (high performance liquid chromatography) using Nucleosil 100 C-18 (10 /xm) or Lichro-Sorb RP Select B. For this purpose the epidermin is washed out step by step, first with a solvent A consisting of water containing 1% by weight of formic acid, then with a solvent B, consisting of methanol/water (80:20, v/v) with 1% by weight of formic acid.

In terms of the procedure used, the epidermin may be produced by a batch fermentation, a related feeding fermentation in which individual substances are added during fermentation, a fermentation with discontinuous adsorption of the epidermin formed as on-line adsorption, and an on-line adsorption which proceeds continuously.

In so-called batch fermentation, the best results were achieved with a nutrient solution having the following composition: 3.3% of meat extract, 3% of malt extract, 0.37% of calcium hydroxide (% by weight). The yield is advantageously affected by the addition of 1 to 6% by weight of alkali metal chlorides, such as sodium chloride, potassium chloride and 0.001 to 0.002% by weight of iron ions, e.g. in the form of FeCl3 and/or FeSO0, and also ammonium chloride or sulphate (between 57 and 200 mM); the preferred concentrations are 3% by weight of sodium chloride and 0.00125% by weight of ferric chloride. The influence of initial concentrations of 3% common salt and different concentrations of ferric chloride on the epidermin production is illustrated by Fig. 1. Of all the Csources, maltose gave the best results after malt extract. It is advisable to use glucose only in conjunction with other C-sources. A combination of lactose with maltose or galactose with maltose gave equally good results. A combination of glucose with maltose or lactose or galactose gave the same yield as the malt extract. All the other conventional C-sources resulted in either minimal or no epidermin production.

The fermentation is carried out with good ventilation at temperatures of between 34 and 37 °C, preferably 36°C. The best pattern of production is obtained if the pH value is 6.0 to 7.0 before fermentation. In the absence of carbonates or hydroxides of divalent cations, such as calcium carbonate or calcium hydroxide, only a little production occurs. After the addition of calcium - carbonate, for example, the pH value shows a characteristic pattern, falling into the acid range, in which case only slight production occurs. When the pH value subsequently climbed into the alkaline range production started up., Instead of calcium carbonate it is also possible to use magnesium carbonate whilst calcium hydroxide yielded better results than calcium carbonate. The production was increased somewhat by using 50 mM calcium hydroxide rather than 25 mM calcium carbonate. When the C-sourees (sugar) are utilised by the strain, organic acids such as acetic acid are formed, which are complexed by divalent cations whilst at the same time the medium is buffered. The increases in activity at the time of maximum production by the strain DSM 3095, determined by the plate diffusion test (determining the diameter of the inhibitory area in mm against Micrococcus luteus ATCC 9341), using a calibrated line, with the activity in a brain-heart infusion nutrient solution according to EP-A-0 027 710 being taken as 100%, were as follows: by the processes according to the prior art: a. ) brain-heart-infusion-agar according to EP-A-0 027 710 100% b. ) 3% meat extract, 2% malt extract, mM calcium carbonate according to EP-A-0 181 578 200% c. ) 3% meat extract, 2% malt extract, mM calcium hydroxide according to EP-A-0 181 578 320% by the processes according to the invention: d. ) 3.3% meat extract, 3% malt extract, mM calcium hydroxide 440% θ.) 3.3% meat extract, 3% malt extract, mM calcium hydroxide, 3% sodium chloride, gm iron(III)chloride 1,720% f. ) 3.3% meat extract, 3% malt extract, 50 mM calcium hydroxide, with the further addition of glucose, KH2 PO4 and ammonium chloride 1,950% g. ) 3.3% meat extract, 3% malt extract, mM calcium hydroxide, with additional on-line adsorption of the epidermin produced after the first isolation step, and for other additives see point f 2,780%.

Fig. 2 shows the course of fermentation as described under d) above; Fig. 1 shows the dependency of epidermin production on the addition of 3% sodium chloride and of various quantities of iron(III)chloride as specified in e). All the percentages given above are percent by weight. Fig. 3 shows the course of fermentation as described in point f), Fig. 4 shows the same course when on-line adsorption is included, as specified under point g) above.

A comparison of the values achieved by the plate diffusion test or HPLC shows that when the process according to the invention is used together with its variants compared with the processes known per se there is a significant increase in the epidermin yields, namely from 320% up to 2,780%.

The individual steps and conditions for carrying out the process according to the invention will be described more fully hereinafter.

The producing strain is best stored by deepfreezing (-18°C) in a medium containing 3.3% by weight of meat extract, 3% by weight of malt extract, 0.37% by weight of calcium hydroxide in 40% by weight of glycerol (the remainder is water). For each fermentation, a preculture is grown for 18 hours at 36°C on an agar medium (pH 7.2 to 7.4) which contains per litre 8 g of Lab Lemco powder (made by Oxoid), 10 g of peptone, 3 g of common salt, 2 g of Na?HPO4, 15 g of agar and 10 g of sterilised glucose.

Fermentation can be carried out in suitable shaking flasks and in order to produce larger quantities of the substance fermenters with capacities of 200 litres or more may be used.

For flask tests, 500 ml Erlenmeyer flasks with a lateral opening are used. The flasks are filled with 100 ml of nutrient solution and autoclaved for 20 minutes at 121 °C. The inoculant used is 1% of an 8 hour old preculture. Incubation is at 36°C on a shaking machine rotating at 160 revolutions per minute (rpm).

For fermentation on the 15 litre scale, a 20 litre fermenter (type b 20 Braun/Melsungen or Giovanola Freres, Monthey, Switzerland; with a recirculating system) was charged with 15 litres of nutrient solution with the addition of 0.5 ml of polyol and sterilisation was carried in situ at 121°C for 30 minutes. The inoculant used was 150 ml of an 8 hour old preculture. Fermentation was carried out at 36 °C, 0.4 vvm and 900 rpm.

Production of epidermin by batch fermentation: In the bioreactors used, the best growth and the best yield of epidermin were obtained at 36°C, a ventilation rate of 0.4 vvm and stirring at 900 rpm. Reduced ventilation and slower stirring reduced the yield of epidermin (Fig. 5). More powerful ventilation and stirring resulted in a slight improvement in the yield but excessive foaming also occurred. This is because of the powerful surface activity of the antibiotic; the foam could not be suppressed mechanically or chemically without some loss of activity.

Fig. 2 shows the course of a batch fermentation on the 20 litre scale with the vigorously producing strain DSM 3095 in a nutrient solution containing 33 g of meat extract, 30 g of malt extract, 3.8 g of calcium hydroxide in 1 litre. This strain produces epidermin in a quantity up to 80 mg/1. Intensive growth combined with the utilisation of the C-sources provided, with the formation of acetate, can be recognised by the drop in the pH value. A maximum number of cells is achieved after 30 hours. After this time, the chief fermentable C-sources such as glucose and maltose are exhausted.

The phosphate added is used up after 8 hours. The maximum concentration of the antibiotic is achieved after 48 hours and amounts to 80 mg/1.

Both the addition of chlorides such, as sodium chloride or potassium chloride in a quantity up to 70 g per litre and the addition of FeCl3 or FeSO<( up to 150 μΜ increased the yield of epidermin without lengthening the fermentation process (Fig. 1). The maximum yield of epidermin of 310 mg/1 was obtained by simultaneously adding 3% sodium chloride and 75 μΜ iron(III)chloride to the production medium.

This rapid epidermin production constitutes a good possible method of developing a continuous fermentation process with high throughputs of the medium used.

Epidermin production by feeding fermentation: In order to be able to increase the yield of epidermin by extending the growth phase and achieving a higher cell density, higher concentrations both of the C-sources and also of the added phosphate were necessary. Epidermin production is subject to strong Ccatabolic repression (Fig. 6) and is also regulated by the phosphate content of the medium (Fig. 7).

Throughout the entire fermentation process the nitrogen concentration is above 150 mM and the majority of this content cannot be used by microorganisms. In flask cultures, the addition of ammonium salts up to 150 mM, resulted, as shown in Fig. 8 (for ammonium chloride), in a significant stimulation of epidermin production.

Glucose is primarily metabolised to form acetate, as shown by the acidification of the medium. If there is a shortage of the sugars, the organic acids are used as additional C-sources, the pH of the medium changing to the alkaline range. The glucose is therefore added in accordance with the pH during the course of fermentation. The pH is maintained at 6.0 and acidification is not suppressed.

Phosphates are added continuously. The rate of addition was calculated by means of the rate of consumption of the phosphates in batch fermentations in which 10 mM of phosphate were added.

Neither the addition of glucose or phosphate alone nor together resulted in a significantly higher yield of epidermin during fermentation. The higher content of epidermin which was found with the pH-regulated addition of glucose during fermentation, does not depend on additional C-sources but is caused by the slower decomposition of the antibiotic in the acidic medium. This can be demonstrated by fermentations in which the pH value is kept constant by the addition of sulphuric acid. Only the development of combined feeding fermentation with pH-dependent addition of glucose and continuous addition of phosphates and ammonium nitrogen resulted in both a significant increase in the biomass and also an increase in the yield of antibiotic.. Fig. 3 shows the course of a combined feeding fermentation.

The cell mass grows by three times the amount to 2xl011 cells per ml; the maximum epidermin yield is 350 mg/1. Although the maximum cell concentration is reached after 24 hours, the maximum epidermin yield is only achieved during the stationary phase after 72 hours; in the course of feeding fermentation the identity of the growth and production phase which is characteristic of batch fermentation disappears. Nevertheless, epidermin production remains closely linked with the growth of the organism. Firstly, 80% of the total yield are produced during the log phase and secondly the biomass is produced by the number of living cells, which means that the stationary phase can be regarded as a balanced state produced by the growing and lysing cells.

The addition of sodium chloride during feeding fermentation does not produce the same stimulating effect as occurs in batch fermentation. During the first 24 hours the production rate was 40% higher and the growth rate was 70% higher than in the case of feeding fermentation. The specific productivity of the microorganism is therefore 20% lower; after 24 hours the epidermin production ceases, which can be put down either to limitation of nutrients by a hitherto unknown factor or inherent poisoning of the organism by metabolic by-products occurring.

The combined feeding fermentation was also transferred to the 200 litre scale of a pilot plant, using the same ventilation rate, stirrer speed and the same conditions of addition. Without any additional optimisation, epidermin yields of the order of 80 to 90% of those achieved on the 20 litre scale were obtained.

Obtaining epidermin with on-line adsorption of the antibiotic produced: Discontinuous adsorption: In order to prevent feedback inhibition and other possible influences on the producing organisms and in order to protect the epidermin formed from the decomposing influence of proteases and heat, the epidermin was discontinuously removed from the fermentation liquor during fermentation.

In order to do this, the entire fermentation liquor including the producing organisms was sprayed under pressure into a spherical adsorption chamber filled with Amberlite® XAD-1180. This effectively made the resin highly turbulent and resulted in rapid adsorption of the idiolite. The free-flowing resin pellets were filtered off with a screen (diameter 0.25 mm), whilst the fermentation liquor with the biomass was recycled into the reactor (Fig. 9). The course of feeding fermentation with discontinuous adsorption of the epidermin is shown in Fig. 4. The fermentation parameters and conditions of addition were identical to those described above. The maximum cell mass was achieved after 46 hours, at 4 x IO10 cells and the maximum epidermin yield after elution from the resin was 500 mg/l. The total fermentation time was the same as in the feeding fermentation.

The first purification or crude purification is simplified by including the first isolation step in the fermentation process; the epidermin yield after this first purification step is 50% higher compared with the yield in feeding fermentation before purification.

Continuous adsorption: The continuous on-line adsorption of the epidermin during the fermentation process is carried out using a cross-flow filtration apparatus. The material retained is recycled into the reactor whilst the filtrate is adsorbed on an Amberlite® XAD-1180 column. The eluate is also recycled into the reactor (cf. Fig.10). Adsorption is set in operation after 12 to 15 hours' fermentation. Maximum epidermin yields after the antibiotic has been released from the adsorber resin are achieved with quantities of up to 500 mg/l after 80 to 90 hours.

The optimised isolation plan, as shown in Fig. 10, constitutes a major achievement in the attempt to increase the production of epidermin. Finally, a product with a purity of 80% is obtained from the combination of adsorption according to the invention with ion exchange chromatography. The use both of batch and on-line adsorption processes results in a rapid method of purifying epidermin.

In order to monitor the course of fermentation, samples are taken under sterile conditions at various times during fermentation. The samples were evaluated as follows: a) pH values: Measurement with a laboratory pH-meter (Knick pH-mV meter). b) Growth pattern: Growth was monitored by means of the increase in the number of living bacteria. 0.5 ml of culture taken under sterile conditions were diluted in saline and of this 0.1 ml was plated out onto plates (medium: peptone 10 g, meat extract 8 g, common salt 3 g, disodium hydrogen phosphate 2 g, glucose 10 g to 1 litre). After 18 hours' incubation at 37°C the individual colonies could be counted. c) Concentration of antibiotic: The samples were centrifuged in an Eppendorf centrifuge 3200 for 2 minutes and either 10 μΐ of the supernatant were tested in the plate diffusion test or by HPLC as described below. In parallel, a calibrating curve was plotted using known concentrations.

HPLC system: Spray volume: 10 μΐ Eluent: A : water containing 0.05% 70% perchloric acid B : acetonitrile Gradient: Minutes A B 0 77.5 22.5 8 63.0 37.0 8.5 0 100 10 9.5 0 100 10 77.5 22.5 14 77.5 22.5 Flow rate: 2 ml/min. 15 Detection: 210 nm Column: Nucleosil 7 C-18 with associated preliminary column. d) e) 30 f) 35 g) Phosphate measurement: The phosphate content in the culture filtrate was measured using the method of Itaya and Ui (1966) in Clin. Chim. Acta 14 . 361-366.

Measurement of nitrogen: The nitrogen concentration was measured by the Micro-Kjeldahl method using an automatic distillation apparatus (of the Kjeldahl system II type made by Tecator).

Measurement of glucose and maltose: The content of glucose and maltose in the medium was determined enzymatically using a Test Kit made by Boehringer Mannheim.

Acetate measurement: This was carried out by gas chromatography following the instructions of Platen and Schink (1987), Arch. Microbiol. 149. 136-141. h) Epidermin measurement: The epidermin concentration in the culture broth was measured once by bioassay and another time by HPLC according to Fiedler et al., 1987.

Chromatographia 24., 433-438.

After the production peak had been reached the culture liguid was centrifuged off by continuous centrifugation (centrifuge type LA 71b-4, Loher & Sohn.e, Ruhstorf/Rott) at 1380 rpm. In order to achieve optimal separation of the cells the flow rate had to be kept very slow. A first concentration of the active components was achieved by adsorption on styrene-divinyl copolymers as described hereinbefore.

Other methods of adsorption of the antibiotic, for example batch adsorption, discontinuous and continuous adsorption processes with and without separation of the cell mass, have already been described hereinbefore.

The Examples which follow are intended to illustrate the invention: Example 1 Batch fermentation with a strain of Staphylococcus epidermidis DSM 3095: Medium: 3.3% by weight Lab Lemco powder 3.0% by weight malt extract 0.38% by weight calcium hydroxide pH 6.5 (with 3 N H2SOJ Reactor: Type b 20 (Giovanola) with 15 1 of medium Ventilation: 0.4 vvm Stirring rate: 900 rpm Temperature: 36 °C Result: Maximum epidermin yield 80 mg/1 after 48 hours; for results see Fig. 2 Example 2 Feeding fermentation with a strain of Staphylococcus epidermidis DSM 3095 Medium: 3.3% by weight Lab Lemco powder 3.0% by weight malt extract 0.38% by weight calcium hydroxide pH 6.5 (with 3 N H2SOJ Reactor: Type b 20 (Giovanola) with 15 1 of medium Ventilation: 0.4 vvm Stirring rate: 900 rpm Temperature: 36 °C Conditions of addition: Solution 1: Glucose 1 kg, water 1 1, pH 6.0. This solution was added in accordance with the pH and the pH of the fermentation solution was adjusted to 6.0.

Solution 2: NH4C1 500 g, KH2PO4 120 g, water 1.5 1, pH 6.0 (adjusted with anhydrous NaOH). The solution was fed in constantly at a flow rate of 1 ml per litre per hour. The addition began after 4 hours.

Result: Maximum epidermin yield: 350 mg/1 after 72 hours. For results see also Fig. 3.

Example 3 Feeding fermentation with on-line adsorption of the antibiotic: Medium: Reactor: Adsorption Results: See Example 2 See Example 2 On Amberlite® XAD-1180;· 150 g dry weight Maximum epidermin yield: 500 mg/1 after 72 hours; the epidermin yield was measured after the first isolation step (see also Example 4). For results see also Fig. 4.

Example 4 Isolation and purification of the epidermin from a 15 1 culture broth after using on-line adsorption: After on-line adsorption of the epidermin during the fermentation process the resin (cf. Example 3) was washed with 150 1 of water and, if an adsorption chamber was used, transferred into a column for further washing and elution.

Washing: 5 1 methanol/H20 1:1 (v:v) Elution: 5 1 methanol/0.01 N hydrochloric acid 9:1 (v:v) The pH of the eluate was adjusted to 5.5; the epidermin was adsorbed onto this resin by the addition of 2 batches each consisting of 45 g of Amberlite® IRC50 over a period of 1 hour to the eluate. The resin w;as transferred into a column for washing and elution.

Washing: 2 1 0.05 N Na phosphate buffer pH 7.0 Elution: 10 1 0.05 N Na phosphate, 1.5 N common salt in water/methanol 8:2 (v:v), pH 7.0.

In order to remove the salt the pH of the eluate was adjusted to 6.0 and epidermin was adsorbed in batches on Amberlite® XAD-1180, in 2 batches each consisting of 75 g of dry weight within 1 hour. The resin was packed into a column for washing and elution 5 Washing: 20 1 of water Elution: 2.5 acid 1 of methanol/0.01 N hydrochloric 9:1 (v:v).

After evaporation of the solvent and freeze-drying of the eluate, 6500 mg of a substance were obtained containing 5200 mg of epidermin (80% purity). Final purification was carried out by preparative HPLC with gradient elution: Column: Nucleosil 100 C-18 (10 μ) Solvent: A: water containing 1% by weight formic acid B: methanol/water 8:2 containing 1% formic acid.

After preparative HPLC (high performance liquid chromatography) and freeze-drying 4.94 g of pure epidermin were obtained. The antibiotic proved to be uniform in all the tests carried out. The flow diagram is shown in Fig. 10.

During the preparation and isolation of the epidermin, the plate diffusion test was used for monitoring and biological characterisation. Further information on this subject can be found in EP-A-85 113 908.8. In this publication the strain Staphylococcus epidermidis DSM-3095 is defined in more detail.

The culture broth produced by the strain DSM-3095 can also be exchanged for a culture broth produced by the strain NCIB 11536 (deposited at the National Collection of Industrial Bacteria in Aberdeen), in order to produce epidermin therein, isolate and subsequently purify it. However, in terms of epidermin production, the strain NCIB 11536 is inferior to the strain DSM 3095.

Epidermin acts as an antibiotic against skin infections such as eczema, impetigo, cellulitis and acne.

Claims (12)

Claims

1. Process for isolating epidermin from a culture broth or a culture filtrate of a strain of Staphylococcus epidermidis and for purifying this substance, characterised in that a.) the culture filtrate or culture broth is added to a styrene-divinyl copolymer, b. ) the active component is released from the resin by 'θ elution with methanol/diluted hydrochloric acid, c. ) the eluate is adjusted to a pH of 5.3 to 5.8, d. ) the eluate is placed on a weak cation exchanger, e. ) non-bound substances are subsequently washed out with a buffer solution at pH 7, f. ) the active component is eluted out of the cation exchanger with a solution consisting of buffer substance, sodium chloride and methanol at pH 6.0 to 8.0, and for purification g. ) the active substance of the eluate is readsorbed on a styrene-divinyl copolymer, the resin is washed with water in order to remove salts and the epidermin is released from the resin with a methanol/acetic acid mixture and the solution is evaporated or freeze-dried, whilst the epidermin thus obtained may subsequently also be subjected to high performance liquid chromatography for extra purification.

2. Process according to claim 1, characterised in that epidermin is extracted from the culture broth or culture filtrate by adsorption on a styrene-divinyl copolymer, the active substance is liberated from the resin using methanol/0.01 N hydrochloric acid (9:1 v:v), the eluate, adjusted to a pH of between 5.3 and 5.8, is placed on ,a methacrylic acid-divinylbenzene copolymer such as Amberlite® IRC-50 or IRC-84 as cation exchanger, whilst the ion exchange may take place in a column or by introduction in small amounts, non-bound substances are washed out of the resin with a 0.05 N sodium phosphate! buffer at pH 7.0, the active substance is liberated from the resin with a solution consisting of 80% sodium phosphate buffer, 20% methanol, which is 1.5 N with regard to common salt, at pH 7, the eluate thus obtained is adjusted to a pH of 6.0 and added to a styrenedivinyl copolymer resin which is subsequently purified by washing out the salts, the epidermin is released from this resin using methanol/50% acetic acid (9:1 v:v) and the eluate is evaporated or freeze-dried, whilst the epidermin thus obtained may if desired be subjected to final purification by preparative high performance liquid chromatography (HPLC).

3. Process according to claims 1 and 2, characterised in that the epidermin production is carried out by batchwise fermentation in a bioreactor at a temperature of 35 to 37’C, a ventilation rate of 0.2 to 0.6 wm and a stirrer rate of 700 to 1000 rpm with the addition of a chloride, preferably sodium chloride, in ci quantity up to 70 g per litre and/or an iron salt in a quantity of up to 150 μΜ.

4. Process according to claim 3, characterised in that C-sources and/or phosphates and/or ammonici or ammonium salts are continuously or discontinuously fed into the fermentation liquor during fermentation.

5. Process according to claims 1 to 4, characterised in that during the course of fermentation the epidermin formed is discontinuously removed from the fermentation liquor by adsorption on styrene-divinyl copolymers, whilst the fermentation liquor and biomass thus treated are recycled into the fermenter.

6. Process according to claims 1 to 4, characterised in that during the course of fermentation the epidermin formed is removed from the fermentation liquor by continuous on-line adsorption on styrene-divinyl copolymers, whilst the retained material is continuously recycled into the reactor.

7. Process according to claims 1 to 3, characterised in that the culture medium contains 2 to 4% by weight of meat extract, 1 to 3% by weight of malt extract or maltose, galactose, lactose, glucose or mixtures of lactose and maltose, galactose and maltose, and 0.25 to 1% by weight of calcium carbonate or 0.25 to 0.5% by weight of calcium hydroxide, and also optionally 1 to 6% by weight of alkali metal chlorides, 0.001 to 0.002% by weight of iron ions and ammonium salts corresponding to a 57 to 200 mmolar solution at a pH of between 6 and 7.

8. Process according to claim 7, characterised in that the culture medium contains 3.3% by weight of meat extract, 3% by weight of malt extract or maltose, 0.37% by weight of calcium hydroxide and, optionally, 1 to 6% by weight of sodium chloride and/or potassium chloride, 0.001 to 0.002% by weight of iron ions, preferably in the form of ferric chloride or ferrous sulphate, and ammonium salts, preferably ammonium chloride or sulphate, corresponding to a 57 to 200 mmolar solution, optionally together with potassium or sodium dihydrogen phosphate at a pH of 6 to 7.

9. Process according to claims 1 to 8, characterised in that Staphylococcus epidermidis DSM-3095 is used as. the epidermin-producing strain.

10. Process according to claims 1 to 8, characterised in that Staphylococcus epidermidis NCIB-11536 is used as the epidermin-producing strain.

11. Process for isolating epidermin as claimed in claim 1 substantially as described herein with reference to the Examples and/or the accompanying drawings.

12. Epidermin whenever prepared by a process as claimed in any preceding claim.