CHLORTETRACYCLIN E A N D PIGMEN T FORMATIO N BY STREPTOMYCE
S AUREOFACIEN S
IN C O N T I N U O U S CULTUR E J. SLEZAK , B. SIKYT A
Antibiotic s Researc h Institute , Roztok y nea r Prague , Czechoslovaki a
In a previous paper concerned with continuous culture of Streptomyces aureofaciens in a synthetic medium (Sikyta, Slezak and Herold, 1961) we have described the observation that, when sucrose is used as limiting substrate, changes in dilution rate result in significant differences in chlortetracycline and pigment formation. In the steady state, both chlortetracycline and the pigments are synthesized only at dilution rates higher than 0.05 h r . and are not fprmed at dilution rates lower than 0.05 h r . . In similar experiments with ammonium sulfate as the limiting substrate, differences were less significant. We went further into this, trying to find an explanation for the above phenomenon. It was found that the changes were reversible and that increasing the dilution rate over 0.05 h r . the formation of both chlortetracycline and pigments was resumed again. Typical changes in the chlortetracycline level, resulting from changes in the dilution rate, are shown in Fig. 1. Two facts emerge on closer observation: Firstly, that the influence of a change in the dilution rate usually does not become immediately apparent, but rather manifests itself only after some delay, and secondly that the increase in the chlortetracycline level following an increase of the "V dilution rate is more rapid than the drop in the level following after the lowering of the dilution rate. For the explanation of these changes in the ability of the mycelium to synthesize chlortetracycline and pigments, we put forward the following possibilities as the most likely: Fig . 1. Tim e dependenc e of chlortetracyclin e concen tratio n an d mycelia l dr y weight following change s 1. The growth of the organism in dilutio n rate . 1 - chlortetracyclin e concentratio n at very low sucrose concentrations ( O , ìg./ml.) , 2 - mycelia l dr y weight ( # , mg./ml.) , (when sucrose is used as limiting 3 - dilutio n rat e ( Ä , h r . - ) , 4 - p H ( O ) . -1
• 1 .5
Tim e (hrs. )
J . Slezak an d B. Sikyt a
substrate). Enzyme surfaces related to chlortetracycline and pigment production may not be saturated at these low concentrations and the production of both types of substances may stop. 2. The production of inhibitors of chlortetracycline and pigment synthesis, which might be accumulated at extremely low dilution rates. This could be the case if the production rate of these inhibitors were not related solely to the growth of the organism, but e. g. to cell concentration, which is not uncommon with some products. 3. Genetic changes characterized by selection of a part of heterogeneous population, assuming that the non-pigmented part would grow faster at a lower sucrose concentration (lower dilution rate) whereas the pigmented part would grow faster at a higher sucrose concentration (higher dilution rate).
Material and methods The cultivation was performed in a 10-liter laboratory fermentor, modified for continuous cultivation as described previously (Sikyta, Slezak and Herold, 1961), in a medium of the following composition: Sucrose Ammonium sulphate Monobasic potassium phosphate Sodium hydroxide Magnesium sulphate Sodium chloride
0.2 % 0.5 % 0.68 % 0.125% 0.05 % 0-2 %
The aerated and agitated culture was maintained at 29° C. The experiments were performed in the following way: after inoculation, the cultivation was allowed to proceed batchwise for 24 hours. By this time, the exponential growthphase was usually nearly finished and sucrose was practically used up. After this, continuous addition of fresh medium from a storage vessel was started and adjusted so as to have a dilution rate of about 0.03 h r . . After a period of time ranging from 50 to 150 hours, the dilution rate was increased to 0.15 h r . . The drop or increase in chlortetracycline level was determined in 3-hour intervals by bioassay (Hess, 1955). Pigment concentration in the medium was estimated by measuring the absorbancy of the filtered liquid at 200-500 ôçì with a Unicam recording spectrophotometer. These absorbancies showed some interesting relationships and will be dealt with later in greater detail. According to visual estimation, pigments appeared to be absent from the medium completely after about 70 hours of continuous growth. Mycelial dry weight remained practically unchanged during continuous growth. Sucrose concentration, both at low and high dilution rates, was below the sensitivity threshold of the analytical method used. During cultivation, samples of the culture were withdrawn for attempts to elucidate the causes for decreased formation rate of the antibiotic and of pigments during cultivation at a low dilution rate. -1
E AN D PIGMEN T
Results and discussion Elimination of the direct influence of low sucrose concentration or of the influence of some inhibiting substance. For our experiments, we used both whole cultures and their filtrates, and also washed mycelium from samples withdrawn from the continuous culture after 150 hours of growth at the dilution rate of 0.03 h r . during which time the culture had already lost its pigmentation ability and did not contain any chlortetracycline. As controls, samples were used from a 24 hr. batch culture grown in a medium of identical composition. The experiments were run in flasks with 80 ml. of nutrient medium. Pigment formation was estimated visually after 7 days of cultivation on a shaking machine. The following experiments were performed: 1. Non-pigmented culture was diluted 1:1, 1:2 and 1:10 with a medium of the same composition as that used in continuous culture experiments, but without sucrose. Sucrose was added separately to a 0.2 % concentration. 2. Non-pigmented washed mycelium was inoculated into a nutrient medium of the same composition as that for continuous culture. The same experiment was repeated using pigmented mycelium from a batch culture. 3. Sucrose was added to a filtrate of a non-pigmented culture to make a 0.2 % concentration, and this solution was then inoculated with pigmented mycelium from a batch culture. A control experiment was run with the exclusion of sucrose in order to eliminate the possibility that pigments might be released into the medium from the mycelium. In this case, the solution remained quite colourless. 4. Sucrose (0.2 %) and either yeast autolyzate or a trace-element mixture were added to a non-pigmented culture. This experiment was performed because—according to some data in the literature (Kautsky and Kautsky, 1951)—these factors might influence pigment formation in some microorganisms. Results of all the experiments were unequivocal: the introduction of non-pigmented mycelium never leads to the production of pigments in the medium or to mycelium pigmentation. When pigmented mycelium from a batch culture was used, the pigment was invariably released into the medium, with the sole exception of the experiment (3) where sucrose was excluded from the medium. To complete the experiment, we added 2 % sucrose directly to a non-pigmented culture at a dilution rate of 0.03 h r . , attempting in this way to eliminate the possibility that the difference between cultivation conditions in flasks and in the fermentor might have a decisive influence. Even under those conditions the culture remained free of any pigment. These results show that the loss of ability to synthesize pigments and chlortetracycline is inherent in the cells and seems to be unaffected by changes in the medium composition during batch cultivation. This loss is evidently a result of long-term cultivation under conditions prevailing during continuous fermentation at low dilution rates; it is impossible to maintain these conditions in batch culture for longer periods of time. Similarly, in order to regain the ability of synthesizing antibiotics and pigments, the cells must be _1
J . Slezak an d B. Sikyt a
cultivated for some time under conditions prevailing in a continuous culture at a high dilution rate. In theory, deeper biochemical changes inside the cells or a genetic change in population might be involved.
Verificatio n of a genetic chang e in populatio n Genetic changes in population have been verified by the following experiments: Samples were withdrawn at regular intervals from the continuous culture and inoculated on agar slants containing a medium with sucrose and potassium nitrate. A control experiment was run simultaneously using a sample from a 24-hour batch culture, i. e. before starting the continuous feeding. Further samples were taken after 72, 96, 144 and 168 hours of growth. The dilution rate was 0.03 h r . ; after 70 hours of growth, the culture appeared to be non-pigmented on visual observation. After sporulation on agar slants the spores were suspended in a physiological saline solution and sown out on Petri dishes containing the same medium as the slants. One week later, the total number of colonies on each Petri dish and, under UV light, the number of pigmented colonies was counted. In Fig. 2 the mean values of the ratio of pigmented colonies vs. total number of colonies are plotted against the time-axis. After 24 hours of cultivation, i. e. before starting the continuous feeding, 90% of the colonies were pigmented, after 168 hours only 14% (at a dilution rate of 0.03 h r . ) . The curve has an approximatively exponential type. In the basis of these results it may be assumed that the true reason for the partial or complete loss of ability of the strain to produce chlortetracycline and pigments (in continuous culture at low dilution rates) is the selection of non-pigmented part of a heterogeneous population with supression of growth of the pigmented part. At higher dilution rates the reverse is true. The selection probably occurs as a result of a different growth response of the pigmented and non-pigmented parts of the population to sucrose concentration. At a given dilution rate (i. e. at a given sucrose concentration) the part of the population which has a lower specific growth rate under such conditions (at the dilution rate) is always washed out. The maximum specific growth rate seems to be the same for both parts of the population, for the pigment production was not resumed either after long-term cultivation in flasks or after the addition of sucrose ^ , to a non-pigmented continuous culture, even -1
Fig. 2. Relationshi p betwee n th e duratio n of continuou s cultivatio n an d th e relativ e numbe r of pigment-producin g colonies at dilutio n rat e 0.03 h r . - ÷ axis: tim e of cultivation ; y axis: % of pigmen t produc ing colonies vs. tota l numbe r of colonies.
though m both cases the organism seemed to grow at the maximum specific growth rate. In spontaneous mutant
b a c t e r i a
s i m i l a r
c a s e s
selection differing only in specific growth rate
CHLORTETRACYCLIN E AND PIGMEN T FORMATIO N
at low substrate concentrations but otherwise having the same maximum specific growth rate, have been described (Novick and Szilard, 1951). In this way, all previous observations can be readily explained. After a change in the dilution rate has taken place, the following factors seem to be operating: mutual proportion of both fractions of the population prior to changes in the dilution rate, concentration of chlortetracycline and of pigments prior to change, the specific rate of production of chlortetracycline and of the pigments by the producing fraction of the population, and the selection rate. Most of these factors depend on the dilution rate before and after the change and some of them also on the length of time for which the population had been growing at the original dilution rate. An exact calculation of the rate of the decrease of the antibiotic level at a low dilution rate and its increase after a rise in the dilution rate, and the calculation of the time necessary for the effects of the change to become visible in terms of concentration changes, would require a detailed kinetic study, which was not undertaken at this stage of the investigation. A similar drop in the production of chloramphenicol in continuous culture of Streptomyces venezuelae at low dilution rates was observed by Bartlett and Gerhardt (1959); it seems therefore possible that similar selections might take place also with another microorganisms. Reuser (1961) has called the attention to the fact that long-term cultivation, especially in polynuclear organisms, leads to a gradual decrease of antibiotic production.
Study of pigment production The concentration of pigments was estimated in filtrates of the medium by measuring their absorbancies by means of a recording spectrophotometer in the range of200-500çéì . A typical absorbance curve is shown in Fig. 3. We attempted to separate the pigment mixture into components; preliminary experiments have shown the mixture to be composed of at least 5 separate substances. Consequently, an absorbancy curve such as the one in Fig. 3 cannot yield any information as to the concentration of the individual components, being merely a sum of absorbancies of the several substances present. Still, these measurements are a better means of objectively and quantitatively estimating pigmentation than mere visual observation. Samples for estimation were taken during the period of é decreasing antibiotic and pigment concentration at a low \ \ dilution rate and also for a short period after increasing the s. \ dilution rate. As the colour of most of the pigments is \ ^ pH-dependent, each sample was measured at pH 1, 7 and 12. ^ \ / \ The curves had characteristic maxima at 270 and 370m^. | 2. ^ \ Absorbancy values were plotted for these two wavelenghts * \ (Fig. 4). In this Figure the curve of chlortetracycline levels " \ 4
Fig. 3. A typica l absorbanc y curv e of pigment s at p H 7.0. ÷ axis: wavelengt h (ðéì) , y axis: absorbancy .
W a v e lengh t
35 0 (mu )
J. Slezak an d B. Sikyt a
against time is included. After the increase of the dilution rate, absorbancy values keep decreasing for a time, then both the absorbancy values and chlortetracycline start abruptly upwards. By means of these data the previous visual observation of the loss of pigmentation has been put on a quantitative basis.
100 Tim e
Fig. 4. Absorbancie s (p H 7) an d antibioti c concentration s following dilutio n rat e change s in de pendenc e on time . 1 - absorbancy , 2 - Chlortetracyclin e concentratio n ìg./ml . ( O ) - chlortetra cycline, ( O ) - absorbancie s at 270 éçì , ( · ) - absorbancie s at 370 ðéì .
It is worth noting that the antibiotic practically disappears after 60 hours of cultivation, i. e. after 36 hours of continuous growth, whereas the absorbancy values continue to decrease still after 108 hours, i. e. after 84 hours of continuous growth, and that they would apparently go on decreasing even further. Experiments done so far give no indication why this is so. The curves show how wrong it would be to assume that a culture is in steady-state merely because some variables remain constant. Should the dry weight of mycelium betaken as a measure for deciding whether the steady-state has been attained (Fig. 1), this would be found to occur in about 10 hours after starting the continuous feeding of medium; if concentration of the antibiotic were to serve as the criterion, the steady-state would appear to be reached in 36 hours, according to visual estimation of pigmentation in about 46 hours. If, however, the absorbancy of substances with absorption maxima at 270 ôçì were taken as the critical parameter, the steady-state would not evidently be reached at all (Fig. 4). In our case, attention has been drawn to the fact that the culture is not in steady-state by the disappearance of pigment, and prompted us to a further investigation. In other cases, however, changes occuring in the culture might be so slight in character that their detection might be very difficult; this applies both to biochemical changes in the entire culture and especially to genetic changes in the population. It is therefore not correct to characterize some state of the culture as "steady-state" simply because the concentration of the organism and basic nutrients remain unchanged. For some purposes, slight changes are uniportant and might be neglected, but even then it is preferable to state the criterion for the definition of the steady-state (concentration of the organism, of some nutrient, product, etc.).
CHLORTETRACYCLIN E AND PIGMEN T FORMATIO N
Conclusion It has been found that during cultivation of the strain Streptomyces aureofaciens in a synthetic medium with sucrose as limiting factor, the specific rate of synthesis of chlortetracycline and of pigments remains practically constant at dilution rates higher than 0.05 h r . . When the dilution rate drops below this value the concentration of chlortetracycline and of pigments gradually decreases. A more thorough study has shown that this phenomenon is caused by a genetic change in population, namely by selection of one part of the heterogeneous population which is unable to produce tetracycline and pigments. Other possible causes have been eliminated by a series of experiments. The problem of choice of different criteria for the definition of the steady-state in a continuous culture has been discussed. _1
Acknowledgement . Th e author s wish to than k Dr . F . Paleckov a for advic e an d help in th e geneti c experiment s an d Mrs . H . Pafizkov a for spectrophotometri c measurements .
Reference s Bartlett , M . C , Gerhardt , P. : Continuou s Antibioti c Fermentation . Design of a 20 Litre , SingleStag e Pilot Plan t an d Trial s with Tw o Contrastin g Processes . J . Biochem . Microbiol . Techn . Eng . 1:359, 1959. Hess , J. : in Antibiotika , by Milo s Herol d a kolektiv , Sbirk a vedeckyc h prac i 6. Statn i zdravotnick e nakladatelstvi , Prah a 1955. Kautsky , H. , Kautsk y H. , Jr. : Aparatu r zur Erzielun g stationare r Bedingunge n bei der Kultu r von Mikroorganismen . Z. Naturforsch . 6b : 190, 1951. Novick , Á., Szilard , L. : Experiment s with th e Chemosta t on Spontaneou s Mutation s of Bacteria . Proc . Nat . Acad . Sci. 36:708 , 1950. Reusser , F. : Continuou s Fermentatio n of Novobiocin , Appl . Microbiol . 9:366,1961 . Sikyta , B., Slezak , J. , Herold , M. : Growt h of Streptomyce s aureofacien s in Continuou s Culture , Appl . Microbiol . 9:233 , 1961.
Discussio n Prof. Mateles: Since a grea t dea l of sucros e is presen t in th e effluen t fro m th e 1st an d 2nd stag e fermenters , it appear s likely tha t sucros e is not a growth-limitin g nutrient . Hav e you closed an y experiment s to see whethe r th e sucros e concentratio n could be reduce d withou t affectin g th e CT C yield, thu s increasin g th e economi c efficiency ? Dr. Sikyta: W e hav e carrie d out an orientatio n experimen t in which th e concentratio n of sucros e was decrease d b y on e third . Productio n of chlorotetracyclin e was th e same , i. e. 2000 ìg./ml. , as in th e experimen t with a highe r sucros e concentration . Academician Malek: Th e results , presente d by Dr . Sikyta , show especiall y clearl y th e importanc e of a deepe r knowledg e of th e proces s itself an d further , tha t in studyin g th e compositio n of th e fermentatio n medium , especiall y with respec t to th e quantit y of individua l components , we canno t only star t out fro m th e batc h cultures . An d quantit y is ver y importan t in thos e cases, wher e th e fermente d component s of th e mediu m could influenc e produc t formation , an d wher e therefor e we mus t resor t to th e use of long holdin g times . Fro m wha t ha s been said b y Dr . Slezak on th e reversibilit y of change s in th e cultur e (loss of pigment , loss of production ) we woul d rathe r thin k tha t it is not a genetic , bu t a physiologica l change . It is not ver y probabl e tha t a rea l geneti c chang e would revers e so ver y easily even thoug h th e influenc e of selection is important . Especiall y unde r condition s of complicate d limit s we canno t exclud e th e possibilit y tha t th e formatio n of th e substance s investigate d would b e influence d b y some factor s in th e medium , which hav e not been fully recognize d in th e experimenta l material . Thi s is importan t for th e decision ho w t o select th e procedur e for furthe r work . I f it is a wor k of geneti c character , selection condition s mus t b e adjuste d in such a way, as t o b e favourabl e for th e productio n population , an d unfavourabl e for th e non-producin g population . I f it is a physiologica l change , it would b e necessar y to kno w mor e of th e biosyntheti c processes . Dr. Slezak: I n answe r t o Prof . Malek' s first remark : W e believe tha t fro m th e tabl e giving th e compositio n of th e mediu m as was shown her e it follows tha t in our case we wer e dealin g with substrat e limitation . Wit h respec t t o th e geneti c or physiologica l changes , th e following ma y b e said : Th e fact tha t afte r th e dilutio n rat e ha d increased , chlorotetracyclin e an d pigment s agai n bega n to be synthesised , does not exclud e th e possibilit y tha t it was a geneti c change . Th e populatio n is primaril y hetero geneou s an d its par t which synthesize s th e mentione d substance s ha d n o tim e t o b e washe d out in th e cours e of cultivation . As soon as th e dilutio n rat e was increased , th e concentratio n of th e limitin g material—sucrose—increased , mor e favourabl e condition s bega n t o prevai l for th e productiv e par t of th e populatio n which no w bega n t o be selected . Prof. Ierusalimsky: I n our laborator y b y experiment s with B. megatherium we observe d an effect simila r t o tha t foun d b y th e authors . At ver y high dilutio n rate s th e cells bega n t o prolongate . At first thi s effect was reversible , late r th e chang e bega n t o b e stabilized . A ne w for m was obtained , with a change d for m of th e colonies, an d ver y long, nonsporulatin g cells. T o prov e geneti c change s th e author s ough t t o obtai n pur e species or mutant s of th e actinomycetes , which do or do not for m pigment , an d the n t o convinc e themselve s whethe r thes e organism s will maintai n thei r propertie s an d whethe r on e does not chang e to th e other . As long as thi s is not done , th e proble m of whethe r genotypi c or reversibl e physiologica l changes , as assume d b y Prof . Malek , hav e take n place , will remai n open . Dr. Sikyta: Th e finding tha t th e growt h rat e of th e non-pigmentin g par t of th e population , no t producin g th e antibiotic , is highe r at lower concentration s of th e limitin g factor , an d on th e contrar y tha t th e growt h rat e of th e pigmentin g par t is highe r at a highe r concentratio n of th e limitin g substrat e (at a highe r dilutio n rate ) is not isolate d in th e literature .