Appl Microbiol Biotechnol (1992) 37:626-630

Applied Microbiology Biotechnology © Springer-Verlag 1992

Elimination of by-products in llfl-hydroxylation of Substance S using Curvularia lunata clones regenerated from NTG-treated protoplasts Danuta Wilmafiska 1, Krystyna Milczarek 1, Anna Rumijowska z, Katarzyna Bartnicka I and Leon Sedlaczek 1,2 1 Institute of Microbiology, University of L6d~, PL-90-237 L6d~, ul. Banacha 12, Poland : Microbiology and Virology Centre of the Polish Academy of Sciences, PL-93-231 L6d~, ul. Dgbrowskiego 251, Poland Received 12 February 1992/Accepted 14 April 1992

Summary. Stable mutants showing improved l l-hydroxylation o f Substance S were isolated, following treatment with N-methyl-N'-nitro-N-nitrosoguanidine (NTG) and regeneration of uninucleate protoplasts of the appropriate fungal strains. This procedure was especially suitable for obtaining more directed 1 lfl-hydroxylation of Substance S with Curvularia lunata I M 2901. Apart f r o m producing cortisol (1 lfl-hydroxy-S), the parent strain formed several by-products that significantly lowered the yield of the desired 1 lfl-hydroxyderivative. Isolated mutants of this microorganism carried out directed 1 lfl-hydroxylation with only a small amount of one of the by-products, which resulted in a much higher yield of cortisol.

Introduction A m o n g microbiological transformations of steroids, l lB-hydroxylation of Reichstein's Substance S belongs to the most important ones. It is a direct way of producing cortisol (Colingsworth et al. 1953; Shull and Kita 1955), which, apart f r o m being a finished medical steroid, is also the starting material for the manufacture of several other potent steroids of prednisolone (i.e., 1-dehydrocortisol) structure (Smith 1984). Despite the great interest in directed 1 lfl-hydroxylation resulting in a high yield of the desirable product, the microorganisms capable of l lfl-hydroxylating steroids isolated so far accumulate by-products as well, mainly other hydroxyderivatives, which significantly decrease the economic efficiency of the process. The improvement of cortisol yield by strain selection, changes in media composition and variations in the fermentation conditions proved ineffective (Smith 1984). Since there are some examples of successful application of the classic mutation-selection method in strain development for other steroid transformations, including hydroxylations (see review by Sedlaczek 1988), it Correspondence to: L. Sedlaczek

seemed reasonable to undertake such work with fungi, which are acknowledged as l lfl-hydroxylators of steroids. To increase the probability of improved strain selection, a fraction of uninucleate protoplasts of these organisms was used for mutagenic treatment. Microbial protoplasts and spheroplasts have been shown to be very useful in this kind of procedure, yielding biotechnologically interesting clones for the production of ergot alkaloids (Keller 1983), and antibiotics (Malina et al. 1985; Filippini et al. 1986; Mat~ju et al. 1991) as well as genetically modified Mycobacterium strains for sterol transformation (Jekkel et al. 1989).

Materials and methods Microorganisms. Cunninghamella elegans IM 21Gp, able to carry out 11~- and 1lfl-hydroxylation of Substance S, the 1la-hydroxyderivative being the main product (Sedlaczek et al. 1981), Curvularia lunata 1M 2901 and C. tuberculata IM 4417, which transform Substance S multidirectionally, the l lp-hydroxyderivative being one of the main products, and Cylindrocladium simplex IM 2358, a fungus that 1lc~-hydroxylates Substance S (Sedlaczek et al. 1985), were used in these experiments to compare the effect of protoplasts mutagenizing both the 11~-, and llfl-hydroxylations. All these microorganisms came from the strain collection of the Institute of Microbiology, University of L6dL Trichoderma viride 1131 CBS 354-33, kindly supplied by Prof. L. Ferenczy, University of Szeged, Hungary, was the source of enzymes lyric towards the cell walls of the steroid-transforming fungi. Culture media. Medium PL-2 consisted of (grams per litre): glucose, 20; peptone, 5; yeast extract (Difco), 5; KHzPO4, 5; malt extract 12°Blg 100ml (l°Balling = 1 g of soluble substances extracted from the grain per 100 ml malt extract) Peppler and Perlman 1979, pH 6.0. Medium Z-T consisted of (grams per litre): glucose, 4; yeast extract (Difco), 4; agar, 25; malt extract 6°Blg, up to 11, pH 7.0. Medium Z-T stabilized osmotically contained additionally 0.6 M KC1. T. viride was grown in a medium that contained the following: 100 g wet mycelium of the appropriate steroid-transforming fungus, suspended in 700 ml distilled water and autoclaved for 20 min at 117° C (as the inducer of cell-wall-degrading enzymes), 3 g glucose, and minerals as in the medium used by De Vries and Wessels (1972). Distilled water was added up to 1 1, pH 6.0. Czapek-Dox medium (Prescott and Dunn 1959) was used to isolate auxotrophs.

627

Reagents. Steroids: Substance S of Reichstein (17a,21-dihydroxy-

Auxotrophy determination. To reveal auxotrophic mutants among

4-pregnene-3,20-dione), cortisol (llfl-hydroxy-SubstanceS), and epicortisol (llc~-hydroxy-SubstanceS) were from Koch-Light Lab.; 14a-hydroxy-S, 6/~-hydroxy-S and 20/~-hydroxy-S came from Steraloids Inc. N-methyl-N'-nitro-N-nitrosoguanidine (NTG) was from Merck. Other chemicals were from Serva or Sigma.

clones isolated from the mutagenized protoplasts, the regenerated clones were tested for growth on mineral Czapek-Dox medium.

Preparation of the lytic enzymes. The medium for T. viride cultivation was inoculated with the spores of this microorganism washed from 7-day cultures on Z-T slants. After a 7-day incubation at 28°C on a reciprocal shaker (120 strokes/rain, stroke length 5 cm) the lytic enzymes and other proteins were precipitated from the culture filtrate by adding (NH4)2SO 4 to 80% saturation, centrifuged, dissolved in a small volume of water and dialysed against water. The insoluble material was removed by centrifugation and the supernatant was lyophilized and stored at 4°C as the crude lytic enzyme preparation.

Release of protoplasts. The steroid-transforming fungi were grown in PL-2 medium for 24 h. The mycelia were filtered and washed twice with distilled water, then samples (80-100 rag) were suspended in 5 ml citrate-phosphate buffer, pH 5.6, containing 0.8 M MgSO4. Lytic enzymes (2.5 mg lyophilized preparation/ml) were added and the digestion was terminated when no further increase in the number of protoplasts was observed. The digestion mixture was filtered through a nylon net, and transferred into 0.6 M KC1 (Dtugofiski et al. 1984). To obtain uninucleate protoplasts, the suspension in 0.6 M KC1 was centrifuged at 150g for 5 min. The pellet, which contained multinucleate protoplasts and small fragments of the hyphae, was rejected. The supernatant was centrifuged at 3500g for l0 min. The resulting sediment contained predominantly uninucleate protoplasts, contaminated with no more than 7-10% anucleate forms. This procedure is a modification of a method presented by Nagy et al. (1985 unpublished data). The nuclei of the protoplasts were stained by the method described by Gaugy and Fevre (1982).

Treatment of protoplasts with NTG. Prior to the treatment of protoplasts with the mutagen, the number of colony-forming units other than protoplasts (spores, fragments of the hyphae) was determined in the protoplast suspension. Aliquots of the suspension were diluted in water and 0.6 M KC1, and transferred to the stabilized Z-T plates. The protoplasts were used for further experiments if the number of colonies obtained from the 0.6 M KC1 suspension was at least 100 times bigger. NTG (100 ~g/ml) was added for 15 min to the protoplasts (1 × 107/ml), the suspension was then centrifuged, washed with 0.6 M KCI, and diluted in 0.6 M KC1 for clone regeneration. Protoplast regeneration. The suspension of protoplasts was transferred (0.5 ml) to petri dishes that contained stabilized Z-T medium. After a 7-day incubation, the spores and hyphae from the single colonies (clones) were used for inoculation of the Z-T slants on which the isolated microorganisms were maintained.

Steroid transformation. All the clones obtained from the NTGtreated protoplasts were screened for l lp-hydroxylase activity with Substance S as substrate. The mycelium of each isolated clone (2 g wet biomass), from a 48-h culture on the PL-2 medium, was suspended in 20 ml distilled water to which 10 mg Substance S, dissolved in 0.25 ml 96° ethanol, was added. The transformation was carried out for 6 h at 28° C on a reciprocal shaker. Steroids were then extracted with chloroform:ethanol (9: 1) and analysed. The mutants that showed higher hydroxylating activity than the parent strain Curvularia lunata IM 2901 in the flask experiments, were checked in a laboratory fermentor Bioflo II (New Brunswick): volume of the medium 800 ml, mycelium wet mass 50 g, steroid concentration 1 g/l, dissolved oxygen tension not less than 30% saturation. Steroid analysis. The steroids were analysed using thin-layer chromatography (TLC) on Merck precoated TLC plates, silica gel 90 F-254, and Hewlett-Packard (HP 5890 Ser II) gas chromatography (GC).

Results F o u r steroid-hydroxylating m i c r o o r g a n i s m s were used at the first stage of this study to increase the p r o b a b i l i t y o f i m p r o v e d strain selection following m u t a g e n i c t r e a t m e n t of the protoplasts. The overall n u m b e r o f clones isolated a n d screened for increased steroid h y d r o x y l a t i o n activity, as well as f r e q u e n c y of changes in the m o r p h o logical a n d b i o c h e m i c a l features of the isolated clones, are s h o w n in Table 1. The exposure of protoplasts to N T G resulted in altered p h e n o t y p e s o f all o f the m i c r o o r g a n i s m s used. I n a d d i t i o n to the m o r p h o l o g i c a l changes, the t r a n s f o r m i n g ability of the selected clones was increased. Cunninghamella elegans: w i t h o u t noticeable change in the ratio of the 11o~- to the l lfl-hydroxyderivative, the a m o u n t o f the p r o d u c t s f o r m e d in c o m p a r a b l e c o n d i t i o n s was, o n average, 20% larger t h a n that o f the c o n t r o l strain. Cylindrocladium simplex: as with Cunninghamella elegans, q u a n t i t a t i v e changes in the l l ~ - h y d r o x y p r o d u c t were o b t a i n e d . The increase in epicortisol f o r m a t i o n r a n g e d f r o m 10°70 to 30% d e p e n d i n g o n the tested clone. Curvularia tuberculata: due to the m u l t i d i r e c t i o n a l attack o n the substrate, b o t h q u a n t i t a t i v e a n d qualitative differences were noticed which resulted in m o d i f i c a t i o n of the t r a n s f o r m a t i o n pattern.

Table 1. The number and phenotypic features of clones isolated from mutagenized protoplasts of four fungal strains Feature

Cunninghamella Cylindroeladium Curvularia Curvularia elegans simplex lunata tuberculata

Overall number of regenerated clones tested in biotransformation Morphological changes of the regenerated strains The number of auxotrophic mutants The number of clones with increased steroid-transforming activity

1084

a In surface cultures: variable height and appearance of colonies, lack of aerial hyphae, altered colour of the mycefium and sporangiospores (conidia), loss of sporagniospore spines (Cunninghamel-

6 11

1271 1171 620 Frequent, in about 20% of clones a 1 18 2 19 47 21

la elegans). In submerged cultures: change from filamentous form to pellets, altered pigmentation

628

Substance S - 74~- OH - S

•

3

5 A

- - • o

6fl-OH 77

S --'~

-OH- s

--

•

cortisol 2o;3-oH - s

--

9

C.lunafa parent strain

I

6

•

II

III

6

IV

7

Fig. 1. Transformation patterns of Substance S with Curcularia lunata parent strain and clones isolated from mutagenized protoplasts (I-IV). Copies of original chromatograms developed in methylene chloride : dioxane: water (2 : 1: 1)

The most promising effects were observed with C. lunata. The number of both the auxotrophs and clones with improved 1 lfl-hydroxylation ability exceeded that obtained for the other fungal species. For this reason, more attention was given to this microorganism. The parent strain C. lunata IM 2901 transformed Substance S multidirectionally, as shown in Fig. 1. In addition to the 1lfl-hydroxyderivative (cortisol), several other transformation products were formed, among which 14c~-, OH-S, 6fl-OH-S, and 20fl-OH-S were preliminarily recognized by the comparison of the Rf values and UV254 fluorescence in TLC, and the retention times in GC of the products with those of the authentic samples. This kind of transformation pattern facilitates evaluation of the mutagenic effect because it can be expressed phenotypically as the change in the amount of the particular product as well as in the number of products. Taking into account the transformation patterns, all tested 1171 clones could be divided into five groups. The biggest one included strains without altered transformation ability, producing the same steroids as the parent strain. Other clones lost the l lfl-hydroxylation activity (patterns II and III), and still others produced more cortisol at the expense of by-products, the number of which diminished. This refers especially to the clones of the transformation pattern I, which synthesize the transforming enzyme system of distinctly increased regio- and stereospecificity resulting in the cortisol formation with only a few traces of other products. The clone designated C. lunata CL 102/3, and the mutants derived from it in the second step of N T G mutagenesis, CL 366/102, CL 642/ 102, and CL 377/102, are representatives of these microorganisms. The results obtained by TLC were confirmed by GC analysis (Fig. 2). The retention times of the products and the amounts of some of the compounds produced are shown in Table 2. The RT values of the products from peaks 1, 4, 5, and 6 were the same as those of the standard samples of 20fl-OH-S, 6fl-OH-S, l lfl-OH-S (cortisol) and 14o~-OH-S, respectively. However, only cortisol was further identified by determining the specif-

3

Fig. 2. Gas chromatography separation of steroids produced by the parent strain (A) and the mutants of the CL 102/3 series (B) of C. lunata. Column HP-17, cross-linked phenyl methyl silicone, 10 m × 0.53 mm. Temperatures: injector 250° C, column 240° C, detector 270°C; flow: helium 15ml/min. Samples dissolved in acetone. Retention time of cortisol (peak 5) was 19.779rain (A) and 19.863 min 0l); retention time relative to progesterone was 1.48. Other preliminarily identified products are: 1-20fl-OH-S; 4-6fl-OH-S; 6-14a-OH-S

ic rotation (O/)D , maximum absorption at 240-242 ~tm, melting point, and IR absorption spectrum, which were identical with that of the original product. The amount of cortisol and the presumable recognized by-products in flask experiments was expressed in mol percent, 100% being 1.44 mM (0.5 g/1 of Substance S), according to Hesselink (1988). In addition, the yield of cortisol was estimated in pilot-scale fermentations with 601 medium. In unoptimized preliminary transformations, 17.1 g (28.5%) and 39.0 g (65%) refined cortisol was obtained from 60 g Substance S, using the parent strain and the CL 366/102 clone, respectively. From the numerous by-products accompanying cortisol formation in Substance S transformation by C. lunata IM 2901, only one having the retention time of 12.633 min (A) and 12.720 rain (B) was still produced at a much lower concentration by the mutants of the CL 102/3 series. Due to the greater regio-, and stereospecif-

629 2. The retention times (RT) and amounts of transformation products formed by the parent strain (A) and the mutant (B) of

Table

Curvularia lunata

Peak number

1 2 3 4 5 6 7 8

RT indicated by the integrator a (min) A

B

8.616 10.406 12.633 15.861 19.779 (cortisol) 26.707 29.315 35.414

--12.720 --

19.863

RT relative to progesterone b

Amounts of recognized products (mol %) A

B

0.64 0.78 0.95 1.19 1.48

0.65 nd nd 0.19 24.27

--nd

2.00 2.20 2.65

8.16 nd nd

--

67.95

m

nd, not determined Slight differences in A and B result from hand-made injections of the analysed samples b Progesterone was used as the internal standard

icity of the enzyme system formed by the mutant strains, the yield of cortisol, the desirable transformation product, increased more than twice compared with that of the parent strain. The stability of the new mutants was demonstrated in 20-1 laboratory fermentors as well as in pilot-scale fermentations (2001). With conventional methods for the preservation of biologically active material, the strains of the CL 102/3 series have been maintained in our laboratory since their isolation (i.e. about 3 years) without any detectable changes in their steroid hydroxylation ability.

The results obtained in the present paper indicate that the classic mutation-selection procedure may be very effective in this field if applied to appropriate cell material. The mutagenic treatment of conidia and hyphal fragments of Curvularia lunata IM 2901 proved completely unsuccessful (data not shown), probably due to the use of multinucleate material. This could result in mixed cultures, unsuitable for stable mutant selection, as well as in the loss of potential mutants by the "rescue" repair of the DNA from undamaged nuclei (Rowlands 1983). By treatment of the uninucleate fraction of C. lunata IM 2901 protoplasts with NTG, the desired mutants could be obtained with a relative high frequency. It remains to be explained whether the eliminated byproducts are formed by the action of highly directionspecific independent enzymes or by a single enzyme of diminished specificity. The 19-nortestosterone transforming system of C. lunata NRRL 2380 carrying out hydroxylation in positions 10ft-, llfl-, and 14o~- of the substrate was suggested to be a single hydroxylase that forms different complexes with the same substrate due to minor misalignment or variant geometry of approach and binding of the steroid substrate molecule (Lin and Smith 1970). Obtaining mutants of C. lunata IM 2901, which produce only by-products (Fig. 1, transformation patterns II and III), and those that form mainly the 1 lflhydroxyderivative, may indicate the operation of at least two distinct enzyme systems. However, the by-products could also be formed by individual enzymes if they were encoded by genes occurring in clusters. It is well known that N T G frequently produces closely linked mutations in gene clusters, due to its high specificity for DNA replication fork (Rowlands 1983). Further experiments are being made in our laboratory to gain more data that could clarify this question. Acknowledgement. We are grateful for financial support of this

Discussion

Unlike the microbial l lo~-hydroxylation of progesterone, which can be performed in almost quantitative yields, the introduction of the hydroxyl group into position 1 lfl is a less directed process, and it is always accompanied by undesirable by-product formation (Smith 1984). Because of the great commercial importance of 1lfl-hydroxylation of steroids, much effort has been put into its improvement. However, it was not possible to obtain results as good as for l la-hydroxylation. Most recently genetic engineering has been used to help to solve this problem. Osiewacz and Weber (1989) have developed a transformation system for C. lunata A T 46 that may be useful for future strain improvement. A gene-transfer system has also been described for the transformation of Cochliobolus lunatus m 118, another biotechnologically important l lfi-hydroxylating microorganism (Dermastia et al. 1991). Thus, the first step has been made towards the use of molecular genetics in strain improvement programmes for steroid transformation, particularly fungal 1lfl-hydroxylation.

work by grant CPBP 04.11. From the Ministry of National Education.

References

Colingsworth DR, Karnemaat JN, Hanson FR, Brunner MP, Mann CK, Haines WJ (1953) A partial microbiological synthesis of hydrocortisone. J Biol Chem 203:807-813 Dermastia M, Rozman D, Komel R (1991) Heterologous transformation of Cochliobolus lunatus. FEMS Microbiol Lett 77 : 145-150 DeVries OMH, Wessels JGH (1972) Release of protoplasts from Schizophyllum commune by a lytic enzyme preparation from Trichoderma viride. J Gen Microbiol 73 : 13-22 Dlugofiski J, Sedlaczek L, Jaworski A (1984) Protoplast release from fungi capable of steroid transformation. Can J Microbiol 30: 57-62 Filippini S, Garofano L, Breme U (1986) Nitrozoguanidine mutagenesis on Streptomycesfradiae protoplasts. Fifth International Symposium on the Genetics of Industrial Microorganisms, GIM 86. Book of Abstracts, Split, Yugoslavia, September 1420, p 106 Gaugy D, Fevre M (1982) Protoplast production from Saprolegnia monoica. Microbios 34 : 89-98

630 Hesselink P (1988) Sterol side chain cleavage by Mycobacterium. Characterization, optimization and genetics. PhD thesis, University of Groningen, The Netherlands Jekkel A, Csajagi E, Ilk6y E, Ambrus G (1989) Genetic recombination by spheroplast fusion on sterol-transforming Mycobacterium strains. J Gen Microbiol 135 : 1727-1733 Keller U (1983) Highly efficient mutagenesis of Clavicepspurpurea by using protoplasts. Appl Environ Microbiol 46:580-584 Lin YY, Smith LL (1970) Microbial hydroxylations. VII. Kinetic studies on the hydroxylation of 19-norsteroids by Curvularia lunata. Biochim Biophys Acta 218:515-525 Malina H, Tempete C, Robert-Gero M (1985) Enhanced sinefungin production by medium improvement, mutagenesis, and protoplast regeneration of Streptomyces incarnatus NRRL 8089. J Antibiot 38 : 1204-1210 Mat~ju J, Mar~alkova J, Nohynek M, Steinerova N (1991) Mutant strains of Streptornyces cinnamonensis protoplasts. Cultural and physiological conditions. Folia Microbiol 36:42-48 Osiewacz HD, Weber A (1989) DNA mediated transformation of the filamentous fungus Curvularia lunata using a dominant selectable marker. Appl Microbiol Biotechnol 30:375-380

Peppler H J, Perlman D (1979) Microbial technology. Academic Press, New York, pp 22, 24 Prescott SC, Dunn CC (1959) Industrial microbiology. McGrawHill, New York, p 519 Rowlands AR (1983) Industrial fungal genetics and strain improvement. In: Smith JE, Berry DR (eds) The filamentous fungi, vol 4. Fungal technology. Arnold, London, p 346 Sedlaczek L (1988) Biotransformations of steroids. CRC Crit Rev Biotechnol 7 : 187-236 Sedlaczek L, Jaworski A, Wilmafiska D (1981) Transformation of steroids by fungal spores. I. Chemical changes of Cunninghamella elegans spores and mycelium during cortexolone hydroxylation. Eur J Appl Microbiol Biotechnol 13 : 155-160 Sedlaczek L, Wilmafiska D, Dlugofiski J, Uszycka-Horawa T, Budzyfiska M, Skibifiska M, Jaworska R, Trzcifiska Z, Szczepaniak J, Go~lifiska H (1985) Polish patent no. 123 535 Shull GM, Kita DA (1955) Microbiological conversion of steroids. I. Introduction of 1lfl-hydroxyl group into C2x steroids. J Am Chem Soc 77 : 763 Smith LL (1984) Steroids. In: Rehm H-J, Reed G (eds) Biotechnology, a comprehensive treatise in 8 volumes, vol 6a. Verlag Chemie, Weinheim, p 33