The present invention relates to a method for increasing or prolonging the production of a secondary metabolite by a microorganism in a batch culture. Microorganisms, including both prokaryotic and eu aryotic microorganisms grow in nutrient solutions until the nutrients required for the growth of the organism, i.e. the production of bio ass, have been consumed. This is generally termed the trophophase. When the growth nutrients have been consumed the organisms generally enter a stationary phase, the idiophase, in which there is relatively little biomass production but in which the existing cells survive and metabolise. A number of microorganisms produce secondary metabolites during such an idiophase. These secondary metabolites are generally compounds not essential for normal growth of the organism and are only produced when the organism is in a stationary phase or when there is some restriction on the metabolic pathways associated with unrestricted growth. Examples of secondary metabolites are the antibiotics, pharmacologically active compounds, and toxins.

BACKGROUND ART It has been found that in batch culture, the specific rate of secondary metabolite production drops off rapidly after an early maximum. Attempts to sustain the early maximum rate by the continuous addition of those nutrients found to be essential for the production of the secondary metabolites showed only a limited improvement in the long term rate of secondary metabolite production.

DISCLOSURE OF THE INVENTION The present inventors have found that further substantial improvements in the long term rate of secondary metabolite production can be achieved by the periodic addition to the culture medium of those nutrients required for secondary metabolite production. The present invention consists in a method for

increasing the production of a secondary metabolite by a microorganism, comprising the step of cultivating a suitable microorganism in a culture medium containing those nutrients essential for the production of the desired secondary metabolite, and adding additional amounts of those nutrients essential for the production of the desired metabolites at discrete intervals spaced apart in time.

The present invention further consists in secondary metabolites produced by the foregoing method. While the full reason for the advantage gained by the periodic feeding of secondary metabolite precursors has not been fully elucidated it is thought that the following explanation may be correct. This explanation is given to assist in an understanding of the invention and is not to be taken as limiting the scope of the present invention. It has been found that during the trophophase there is a repression of secondary metabolite synthesis and that during the idiophase the rate of secondary metabolite synthesis peaks and then falls off. It has generally been assumed that this fall off in rate of production of the secondary metabolite is due to the depletion of the precursors for the secondary metabolites. Previous proposals have suggested that the rate of secondary metabolite synthesis could be maintained at a high level for a prolonged time by the continuous feeding of the necessary precursors into the culture medium to maintain a small concentration of those precursor compounds in the reaction medium. While such continuous feeding does increase the overall production of secondary metabolites over a defined time the improvement has not been as great as can be obtained using the process according to this invention. It is thought that the cyclic or periodic feeding of the precursor compounds to the culture medium gets a high level of nutrients into the cells quickly and the cells so charged with nutrients can then revert to vigorous secondary metabolite synthesis once the catabolite repression caused by

the feeding of the precursor compounds has been reduced following the ingestion of the precursor compounds by the cells. It is thought that this increase of secondary metabolite synthesis by the use of cyclical periods of catabolite repression and derepression will have applicability over a wide range of microorganisms producing secondary metabolites as virtually all of the secondary metabolite producing systems studied the microorganisms show catabolite. repression of the secondary metabolite production. The method according to the present invention is applicable to any microorganism capable of producing secondary metabolites. Such microorganisms include filamentous fungi, the filamentous prokaryotes such as the streptomyces and certain of the unicellular bacteria. . Many such organisms are listed in the article "Pharmacologically Active Agents from Microbial Sources" by Dr. S.L. Neidlman contained in "Handbook of Microbiology" Vol. Ill p.p.999-1006, Chemical Rubber Company, Cleveland, Ohio, USA 1973. The present invention is particularly applicable to the production of antibiotic secondary metabolites and more particularly to the production of macrolide antibiotics. The macrolide antibiotics are a structurally related group of antibiotics produced by species of Streptomyces and which all contain a large lactone ring containing 12 to 22 atoms and having few double bonds and no nitrogen atoms. These antibiotics generally have one or more sugar residues attached to the lactone ring.

In respect of any particular organism and any particular secondary metabolite produced by that organism it is necessary to determine those nutrients which regulate the synthesis of the secondary metabolite. This may be done by routine chemostate culture of the organism using a chemically defined medium. By altering the medium composition it is possible to change the kinetic pattern of

production of the secondary metabolites and to observe those cases where high initial values of the specific production rate of the secondary metabolites are observed. An example of such a study is reported in Biotech-Bioeng, 2, 1785-1804 (1980). Once these nutrients are determined the method according to this invention may be carried out.

The method may also be adapted to the production of secondary metabolites by microorganisms growing on complex media as it is not uncommon for complex media to be used in the trophophase of secondary metabolite production the additional defined nutrients required for secondary metabolite synthesis can be fed during the idiophase. If the requirement of the organism for these supplemental feeds have been determined then the present invention can be used to generate cyclical feeds and stimulate production of the secondary metabolite.

In the case of tylosin production by Streptomyces fradiae growing on complex media mass balances on substrate uptake and oxygen and carbon dioxide production need to be carried out during the trophophase and idiophase to determine the metabolic state of the microogranism. The present invention can be used to feed a carbon source, such as glucose, and an amino acid, or another source of nitrogen, in cyclical fashion assuming the lipid source is always present in excess. The data presented in the present specification can then be applied in the formulation of addition rates and amounts in conjunction with the knowledge of the basal uptake rates provided by the complex medium.

The method may be carried out in a strictly batch process in which the nutrients essential for the secondary metabolite production are introduced into the culture medium in a concentrated form and without removal of any of the culture medium. Alternatively a proportion of the culture medium may be removed at the time of addition of the nutrients.

The amount of the cyclical nutrient addition should be adjusted such that there is not a residual amount of that nutrient present in the culture medium during the period between the cyclical feeding of the nutrients. It will also be recognised that not all required nutrients need be fed cyclically, however, it is preferred that all those nutrients which generate catabolite repression be fed cyclically.

The optimum frequency of the nutrient addition must be determined for each organism/secondary metabolite system, however, it seems preferable to add the nutrients at intervals of from 6 to 48 hours, preferably 24 to 48 hours. BRIEF DESCRIPTION OF THE DRAWINGS Figure 1 shows the cell dry weight, glutamate uptake tylosin production and rate of tylosin production in a batch fermentation;

Figure 2 shows the parameters of Figure 1 and glucose uptake for a cyclic fed-batch fermentation according to this invention fed at intervals of 24 hours.

Figure 3 shows the parameters of Figure 2 for a cyclic fed-batch fermentation according to this invention fed at intervals of 36 hours.

Figure 4 shows the effect of varying the amplitude of feeding in a cyclic fed batch fermentaiton according to this invention; and Figure 5 shows the effect of varying the period of feeding in a cyclic fed-batch fermentation according to this invention.

BEST MODE OF CARRYING OUT THE INVENTION Hereinafter given by way of example only are the results of an experiment carried out to show the increased production of the antibiotic tylosin.

MATERIALS AND METHODS The organisms used in this study was Streptomyces fradiae NRRL 2702. The conditions of growth and analytical procedures were described in Biotech, Bioeng., 22, 1785,

(1980) . The medium used contained in one ^" litre of ^' distilled water; Sodium chloride, lg; magnesium sulphate, 2.5g; cobalt chloride, 0.0005g; zinc sulphate, 0.0005g; ferric ammonium citrate, 1.5g; betaine hydrochloride, 2.5g; L-sodium glutamate, 17.5g; dipotassium hydrogen phosphate, 1.15g. The normal fed-batch fermentation was carried out by continuously feeding the culture with glutamate and glucose solutions. For the cyclic fed-batch experiments glutamate and glucose were fed at predetermined and constant time intervals in the different fermentation runs. The time intervals between these additions varied from 12 to 48 hours. The glutamate and glucose were fed using a Watson-Marlow pump (Model 501) and a syringe pump (Sage Instruments, Model 352). Additional methyloleate was added to maintain the concentration above 5 g/1 at all times during the fermentation. The feed cycles were controlled by an Apple microcomputer.

RESULTS The pattern of the batch fermentation of tylosin is shown in Fig. 1 and Table 1 (Control Batch column) . The rate of tylosin synthesis was maximal between 24 and 48 hours of fermentation when the culture was actively growing. The glucose and sodium glutamate were rapidly metabolised by 48 hours and methyloleate was utilised as soon as glucose and glutamate were exhausted. The q - . was as high as 1.2 mq/g/h during growth phase but decreased with time to 0.05 mg/g/h by the end of the fermentation. In an effort to prolong the period of high specific production rates, sodium glutamate and glucose were linearly fed from 48 hours until the end of the fermentation. The feed rates of sodium glutamate and glucose were 0.8 ml/h—(0.18 g/ml of culture fluid) and 0.4 ml/h (0.025 g/ml of culture fluid) respectively. The results of this experiment were shown in Table 1 (Linear Feed column) though the kinetic pattern of the fed-batch fermentation was similar to that of the normal

batch, the productivity was improved and the speci ic rate of tylosin synthesis reached the value of 0.28 mg/g/h by the end of the ferementation, which is about six-fold higher than the specific rate of the normal batch. Further improvement was obtained by cyclic fed-batch fermentation in which sodium glutamate and glucose were cyclic fed at various cycles from 48 hours to the end of the fermentation. At each cycle 20 ml of glutamate (0.18 g/ml) and 10 ml of glucose (0.025 g/ml) were fed into the fermenter during 0.5-hour and 0.25-hour periods respectively. The result of the 24-hour cycle fed-batch fementation is shown in Fig. 2 and Table 1. The total concentration of -tylosin increased nearly 100% above the control. The specific rate of tylosin synthesis was maintained at about 0.5 mg/g/h which was 60% of the maximal value obtained in the batch culture. By the end of the fermentation the specific rate of tylosin syntheses in the cyclic fed-batch culture was two-fold and ten-fold higher than that in the linear fed-batch and control batch cultures respectively. In Table 1 values of q _+i vlosin observed in cyclic fed-batch fermentations at various cycles are tabulated. The highest values of ^tv _l osin ^were obtained when the cyclic fed-batch fermentation was operated using a 36-hour cycle.

Table 1 Values of observed in batch, normal and cyclic fed-batch fermentations.

Control Batch Intervals in Cyclic Fed-batch (hrs) Batch (Linear 12 24 36 48

Feed)

Time ^q tylosin ^g tylosin/g/h) (hrs)

12 0 0 0 0 0 0

36 1.1 1.0 1.08 0.92 1.1 1.06

60 0.7 0.7 0.7 0.6 0.86 0.64

84 0.3 0.5 0.56 0.53 0.76 0.4

108 0.2 0.4 0.48 0.5 0.71 0.36

132 0.16 0.36 0.44 0.49 0.7 0.35

156 0.09 0.32 0.40 0.45 0.61 0.3

180 0.06 0.3 0.40 0.48 0.61 0.3

204 0.05 0.28 0.40 0.5 0.52 0.3

228 0.05 0.28 0.40 0.5 0.61 0.3

Figure 3 shows the cell dry weight, glucose concentration, glutamate concentration and tylosin concentration of the fermentation described above but fed with glucose and glutamate on a 36 hour cycle. In each case the glucose was fed into the culture medium for 0.25 hour at the rate of 250 mg/l/hr each 36 hours and the monosodium glutamate was fed for 0.5 hour at the rate of 1800 mg/l/hr each 36 hours .

Tylosin concentration was measured as total tylosin by chloroform extractions as described ^"" in Antimicrobiol Agents and Chemotheraphy Apr. 1980 p.p. 519-525. This extract contained, in addition to tylosin, certain other macrolide antibiotics such as relomycin. The pure tylosin was recovered by high pressure liquid chromatography.

Figure 4 shows the effect of varying the amplitude of

the cyclic feeding of glucose and monosodium glutamate on the total tylosin production over a 10 day period in the above described fermentation. It will be seen that as the amount of glucose rises or falls from 0.04 g/l per 24 hours and as the amount of monosodium glutamate rises or falls from 0.6 g/l per 24 hours so the tylosin concentration rises or falls.

Figure 5 shows the effect of varying the period of the cyclic feeding on the total tylosin concentration over a 10 day period in the above described fermentation. Monosodium glutamate was fed at the rate of 0.9 g/l per cycle and glucose was fed at the rate of 0.065 g/l per cycle. It can be seen that a cycle of 36 hours produces a maximum tylosin production of 2.1 g/l. This is to be compared with a normal batch feremtation which produces 0.9 g/l tylosin and a linear feed situation in which 0.9 g/l monosodium glutamate and 0.065 g/1 glucose were fed to the culture medium for a 24 hour period which produced 1.2 g/1 tylosin.