Journal of Biotechnology, 14 (1990) 345-362

345

Elsevier BIOTEC 00522

Cell growth and enzyme synthesis of a mutant of Arthrobacter sp. (DSM 3747) used for the production of L-amino acids from D,L-5-monosubstituted hydantoins Christoph Syldatk 1, Vera Mackowiak 1, Hartmut H~Ske 2 Christiane Gross a, Giselher Dombach i and Fritz Wagner '] I Institute of Biochemistry and Biotechnology A, Technical Uniuersity of Braunschweig, Braunschweig, F.R.G.," 2 Riitgerswerke AG, Castrop-Rauxel, F.R.G.

(Received 20 March 1989; accepted 7 January 1990)

Summary A m i c r o o r g a n i s m with the a b i l i t y to form L - t r y p t o p h a n from D , L - 5 - ( 3 - i n d o l y l m e t h y l ) h y d a n t o i n ( D , L - 5 - I M H ) was isolated a n d i d e n t i f i e d as A r t h r o b a c t e r sp. ( D S M 3747). A f t e r isolation of a m u t a n t with high t r y p t o p h a n p r o d u c t i o n activity b u t low t r y p t o p h a n d e g r a d a t i o n , cultural c o n d i t i o n s were o p t i m i z e d to achieve high a m o u n t s of b i o m a s s with g o o d specific activities c o n c e r n i n g the e n z y m a t i c hyd a n t o i n - c l e a v i n g reactions. T h e ability of the m i c r o o r g a n i s m to p e r f o r m these b i o c o n v e r s i o n s was f o u n d to be i n d u c i b l e b y D , L - 5 - I M H as well as to be d e p e n d e n t on the presence of M n 2+. The highest specific D , L - 5 - I M H - c l e a v i n g activity of the cells was o b s e r v e d in the e x p o n e n t i a l p h a s e of growth. T h e a d d i t i o n of yeast extract to the m i n e r a l salts m e d i u m was f o u n d to be essential for o b t a i n i n g b i o m a s s c o n c e n t r a t i o n s of a b o u t 25 g 1-] cell d r y m a s s b y b i o r e a c t o r cultivations. I n o r d e r to o b t a i n a c o n s t a n t l y high g r o w t h rate, feeding of the C - s o u r c e was pOz-controlled. T h e i n d u c e r D , L - 5 - I M H h a d to be c o n t i n u o u s l y fed to p r e v e n t a decline of the L - t r y p t o p h a n - f o r m i n g enzyme activities, because it was subjected to

Correspondence to: Dr. C. Syldatk, Institute of Biochemistry and Biotechnology A, Technical University of Braunschweig, Konstantin-Uhde-Strasse 5, D-3300 Braunschweig, F.R.G. Abbreviations: CDM=cell dry mass (g 1-1); D,L-5-IMH=D,L-5-(3-indolylmethyl)hydantoin; NCarbtrp = N-carbamoyltryptophan; Gluc = glucose; pO2 = oxygen dissolved (%); Trp = tryptophan; Y x / s = yield (g g-l); biomass formed per substrate used; /Lm~x = maximal specific growth rate (h-1); N-3-CH3-IMH = D,L-5-(3-indolylmethyl)-3-N-methylhydantoin.

0168-1656/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

346 degradation with the enzymes induced and higher concentrations of D,L-5-IMH aggravated the growth significantly. The synthesis of the enzymes was also inducible, when inducer and Mn 2÷ were not added until the late growth phase. Using this process, the consumption of D,L-5-IMH was reduced remarkably. So, under these conditions biomass concentrations of 25 g 1 i cell dry weight with a specific enzymatic activity of 0.20 mmol g 1 h-1 (tryptophan per dry mass per time) could be obtained within 13 h. Using 1 g 1-1 of the chemically modified inducer D,L-5-(3-indolylmethyl)-3-Nmethylhydantoin, which was not degradable by the microorganisms, a biomass concentration of 28 g 1-1 cell dry weight with a specific activity of 0.34 mmol g-1 h-1 (tryptophan per dry mass per time) could be obtained within 28 h. Arthrobacter; D,L-5-(3-Indolylmethyl)hydantoin; L-Tryptophan; Cell growth; En-

zyme synthesis

Introduction

D,L-5-Monosubstituted hydantoins are important intermediates in the chemical synthesis of many D,L-amino acids, but can also be utilized as substrates for enzymatic cleavages leading to the optically pure D- or L-amino acids. Many papers report on the production of optically pure D-amino acids from these compounds (e.g., Yamada et al., 1978; Olivieri et al., 1981; Sun, 1983; Syldatk et al., 1986; Morin et al., 1986; Yokozeki et al., 1987a). On the other hand, the L-specific hydantoin cleavage has so far only been found in bacteria of the genera Arthrobacter (Guivarch et al., 1980; Miyoshi et al., 1985; Syldatk et al., 1987), Bacillus (Tsugawa et al., 1966), Clostridium (Lieberman and Kornberg, 1954) and Flavobacterium (Sano et al., 1977; Yokozeki et al., 1987b; Nishida et al., 1987), whereby the enzymes of Bacillus and Clostridium have substrate specificities distinctly different from those of Arthrobacter and Flavobacterium. The natural function of the D-specific enzyme D-hydantoinase is probably identical with that of the well-known enzyme dihydropyrimidinase (EC 3.5.2.2) (Yamada et al., 1978), the natural function of the D-N-carbamoylase with that of the enzyme fl-ureidopropionase (EC 3.5.1.6) (Yokozeki and Kubota, 1987). In contrast, the natural function of the enzymes involved in the L-specific cleavage of D,L-5-monosubstituted hydantoins is not yet known. The great advantages of both the D- and the L-specific processes for the production of optically pure amino acids compared with other enzymatic methods are not only their wide substrate specificities - even non-proteinogenic amino acids can be produced by both methods (Takahashi et al., 1979; Cotoras and Wagner, 1984) - but also the fact that most of the substrates actually racemize under mild reaction conditions within a short period of time when the enzymes are still active (Tsugawa et al., 1966; Nishida et al., 1987; Yokozeki et at., 1987a). Therefore the racemic D,L-5-monosubstituted hydantoins can be fully converted into optically pure products without any other chemical reaction step necessary.

347 Starting from investigations of Cotoras and Wagner (1984), the aim of this work was to isolate a microorganism with good ability to form L-tryptophan from D,L-5-indolylmethylhydantoin and with low tryptophan degrading activity, to optimize the cultural conditions of cell growth and the biosynthesis of the Ltryptophan-forming enzymes within this microorganism. The optimization of the enzymatic cleavage of D , L - 5 - I M H by resting cells of a mutant of Arthrobacter sp. (DSM 3747) is described elsewhere (Gross et al., 1987a,b; 1990).

Materials and Methods Microorganisms Microorganisms from public type culture collections as well as newly isolated bacteria from soil and water samples were examined. The best strain, Arthrobacter sp. (DSM 3747), isolated from compost, was used for further investigations. Chemicals D,L-5-(3-Indolylmethyl)hydantoin (D,L-5-IMH) and D,L-5-(3-indolylmethyl)-3N-methylhydantoin (N-3-CH3-IMH) were obtained from Professor K. Krohn, Institute of Organic Chemistry, Technical University of Braunschweig, F.R.G. N-LCarbamoyltryptophan (N-L-Carbtrp) was obtained from Sigma Chemicals (Sigma, St. Louis, U.S.A.). The complex nutrients were obtained from Difco Laboratories (Difco, Detroit, U.S.A.). All other chemicals were commercially available products of reagent grade. Media The following media were used: Medium I used for the screening experiments contained per litre 10.0 g D,L-5-IMH, 0.2 g fructose, 3.9 g (NH4)2SO4, 0.95 g K H 2 P O 4, 2.0 g K 2 H P O 4 • 3H20, 0.2 g MgSO 4 • 7H20, 0.02 g CaC12 - 2 H 2 0 and 10 ml of a trace element solution (see below) at p H 7.0. Medium 11, the basal medium used for the characterization of Arthrobacter sp. (DSM 3747) contained per litre 10.0 g glucose, 2.0 g (NH4)2SO4, 2.0 g K2HPO4- 3H20, 0.95 g K H 2 P O 4, 0.2 g MgSO 4 • 7H20, 0.02 g CaC12 • 2 H 2 0 and 10 ml of a trace element solution at p H 7.0. The trace element solution contained per litre 50 mg H3BO3, 40 mg MnSO4 - 2H20, 40 mg ZnSO 4 • 7H20, 20 mg (NH4)6Mo7024 • 4H20, 20 mg FeC13, 10 mg K I and 4 mg CuSO 4 • 5H20. Variations of medium II concerned the C-source and the N-source (see Table 1). Medium 111, the basal medium for the optimization of cell growth and enzyme synthesis contained per litre 20.0 g glucose, 6.5 g (NH4)2SO 4, 0.2 g MgSO a • 7H20, 0.02 g MnC12 • 4H20, 0.02 g CaC12 • 2H20, 0.02 g FeSO 4 • 7H20, 4.54 g K H 2 P O 4, 7.61 g K 2 H P O 4- 3H20, 0.32 g citrate-l-hydrate and 1.0 g D , L - 5 - I M H at p H 6.8. Variations concerned the C- and the N-source, the concentrations of Mn 2÷ and D,L-5-IMH. Further investigations were performed with the addition of 0.1% of different complex nutrients to medium III.

348

Cultivation conditions Screening experiments, the investigations concerning the characterization of

Arthrobacter sp. (DSM 3747), the optimization of its growth and the synthesis of the D,L-5-IMH-cleaving enzymes were carried out in 500 ml-shake flasks containing 100 ml medium under aerobic conditions at 100 rpm and 30°C. Bioreactor cultivations were carried out under aerobic conditions in 10-20 1 bioreactors (Braun Melsungen, Melsungen, F.R.G.) at 30 ° C with regulation of the pH value by addition of 10% aqueous N H 3. For the feeding experiment s in the bioreactor, the concentrations of Ca 2+, Fe 2+, Mg 2+ and Mn 2÷ were doubled to avoid growth limitations. Glucose (Gluc) was added as a 50% aqueous solution to the growing culture after consumption of the initial substrate. This procedure was automatically controlled by the 02 saturation of the culture medium to prevent an oxygen limitation as well as a limitation of the C-source: the feeding was stopped, when the pO 2 value sank below a limit and started again, when a certain pO 2 value was exceeded. D,L-5-IMH was added as a 10% aqueous solution or as 10% solution in 10% aqueous poly(ethylene glycol) 400 (Serva, Heidelberg, F.R.G.). Physiological activity of the cells was monitored using a pH-electrode, a pO 2 electrode and oxygen and carbon dioxide gas analyzers. The time course of the cultivations was followed by measuring cell dry mass (CDM), the pO 2, and the concentrations of glucose (Gluc) (Miller, 1959) and NH~- (Facwett and Scott, 1960).

Mutagenesis Cells of Arthrobacter sp. (DSM 3747) grown in shake flasks in medium II with 5.6 g 1-1 fructose as C-source were harvested in the late exponential growth phase, washed with 0.9% aqueous NaC1 solution, resuspended in buffer-solution and treated with N-methyl-Nl-nitro-N-nitrosoguanidine ( M N N G ) for 30 min and H N O 2 for 15 min according to Carlton and Brown (1981). An enrichment step in medium II with 4 g 1-a tryptophan as C-source and 10 000 U 1-1 penicillin G was employed before the mutagenized cultures were spread on agar plates and screened for D,L-5-IMH-hydrolyzing activity as previously described (Gross et al., 1987b). Tryptophan degrading activity of the mutants isolated was investigated using the Lederberg-technique by parallel incubation with medium II containing 4 g 1-1 L-tryptophan as the sole C-source. Mutants of Arthrobacter sp. (DSM 3747) which showed no or only poor growth on the tryptophan agar plates were selected for further investigations.

Assay of the microbial D,L-5-IMH cleavage Upon harvesting the wet biomass by centrifugation the bioconversion of D,L-5IMH was performed with resting cells in 10-ml shake flasks containing 100 mg wet biomass and 5 mg D,L-5-IMH in 5 ml 0.1 M glycine-buffer, p H 8.5, at 27 ° C or at 4 8 ° C and 100 rpm for 2 h under nitrogen atmosphere (Factor between 27 and 4 8 ° C was determined to be 7). The concentrations of D,L-5-IMH, N-Carbtrp and tryptophan (Trp) were determined by HPLC on a RP-8-column (Serva, Heidelberg, F.R.G.), using 50 mM KH2PO4/25% methanol as eluent and UV-detection (280

349 nm) (Flow rate: 1.0 ml min-1; retention times: Trp: 6.1 min, N-Carbtrp: 8.5 rain, D,L-5-IMH: 16.7 min). The specific activities of the enzymes were defined as mmol g--1 h-1 (product per CDM per time). The optical purity of the Trp was determined by HPLC on a ChiralValCu = Sil00 column (Serva, Heidelberg, F.R.G.), using 10 mM CUSO4/5% acetonitrile as eluent and UV detection (280 nm) Flow rate: 1.0 ml min-~; retention times: D-Trp: 11 min, L-Trp: 14 rain).

Results

Screening experiments The screening for microorganisms using D,L-5-IMH as C-source resulted in the isolation of 11 bacterial strains. While five of these strains showed only the formation of N-Carbtrp in different enantiomeric purity, the other six were able to produce optically pure L-Trp from D,L-5-IMH under assay conditions (for the detection of the optical purity of the products see: Syldatk et al., 1987). All six strains showed only slight differences in their morphological and physiological characteristics, yet they differed strongly in the ratio of N-Carbtrp to L-Trp formation. The best strain, Arthrobacter sp. (DSM 3747), which had been isolated from compost, showed a ratio of 87% L-Trp to 13% N-Carbtrp under assay conditions and was chosen for further investigations. Identification and taxonomical characterization of Arthrobacter sp. (DSM 3747) All results concerning the characteristics of Arthrobacter sp. (DSM 3747) are summarized in Table la-c. According to the 8th edition of Bergey's Manual of Determinative Bacteriology (Buchanan and Gibbons, 1975), the isolated strain was classified as belonging to the genus Arthrobacter, especially to the Arthrobacter globiformis group. It is of interest that the Arthrobacter sp. (DSM 3747) is able to utilize not only D,L-5-IMH but also natural cyclic amides such as uric acid, allantoin and creatinin as sole N-sources in a mineral salts medium. Orotic acid, hydantoin propionic acid, and dihydrouracil, which are the natural substrates of dihydroorotase (EC 3.5.2.3) (Lieberman and Kornberg, 1954), 5-carboxymethylhydantoin amidohydrolase (EC 3.5.2.4) (Hassal and Greenberg, 1963), and dihydropyrimidinase (EC 3.5.2.2) (Yamada et al., 1978), cannot be utilized as N- or C-sources by this microorganism. Mutagenesis of A rthrobacter sp. (DSM 3 747) The aim of these experiments was to isolate a mutant of Arthrobacter sp. (DSM 3747) with both high specific activity to form L-Trp from D,L-5-IMH and low activity to degrade it. It was not possible to select a strain which was blocked in the first oxidative reaction of the L-Trp degradation pathway. Many strains were isolated showing higher D,L-5-IMH-cleaving enzymatic activities than the wild strain, but at the same time a worse ratio of L-Trp to N-Carbtrp. Therefore, a mutant of Arthrobacter

350 TABLE la Taxonomical characteristics of Arthrobacter sp. (DSM 3747)

Morphological characteristics approx. 1 × 5 # m in cell size, polymorphic, not motile, no flagellum, no spore, negative or slightly positive in Gram staining, not acid-fast

Cultural characteristics Nutrient agar colonies: Moderate growth, pale white, circular, convex and entire colonies Nutrient broth: Moderate growth, uniformly turbid, no growth on the liquid surface, pale yellow in medium III

Physiological characteristics (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15)

Obligate aerobic (O/F-test: negative) Temperature for growth: 22-35 o C (optimum: 27 o C) pH for growth: 5.0-10.0 Catalase reaction: positive Oxidase reaction: negative Urease reaction: positive Gelatine liquefying: positive Phenylalanine deaminase: negative Lysine decarboxylase: negative Ornithine decarboxylase: negative Arginine dihydrolase: negative H2S-production: negative Vitamine requirement: negative MacConkey agar: no growth Denitrification: negative

TABLE l b Utilization of different C-sources by Arthrobacter sp. (DSM 3747) C-source

Growth

C-source

Growth

Glucose Fructose Saccharose Maltose Lactose Melezitose D-Xylose D-Glucuronate N-Acetylglucosamin a-D-Glucosamin

+ + + + + + + + + + + + + + +

2-Oxoglutarate Starch Cellulose Serine Glutamate p-Hydroxybenzoate Tyramine Stearic acid n-Tetradecan n-Hexadecan Allantoin Uric acid D,L-5-1MH Creatinin Orotic acid Dihydrothymin Dihydrouracil

+ + + + + + + + + + -

meso-lnositol Glycerol Methanol Ethanol Butanol Acetate Pyruvate

(The cultivations were carried out in 500 ml-shake flasks at 2 7 ° C containing 100 ml medium II with (NH4)2SO4 as N-source and 10 g 1-1 of different C-sources at 27 ° C under shaking. The cultures were inoculated with 1 ml of a 24 h incubated preculture grown on medium II with glucose as C-source.)

351 Table lc Utilization of different N-sources by Arthrobacter sp. (DSM 3747) N-source

Growth

N-source

Growth

Nitrite Nitrate Ammonium Urea Glutamate Uric acid D,L-5-IMH Orotic acid

+ + + + +

Allantoin Creatinin Imidazole Dihydrothymin Dihydrouracil Histidine Hydantoinpropionic acid

+ + + -

(The cultivations were carried out in 500 ml-shake flasks containing 100 ml medium II with glucose as C-source and 2 g 1-1 of different N-sources at 27 o C under shaking. The cultures were inoculated with 1 ml of a 24 h incubated preculture grown on medium II with (NH4)2SO a as N-source.) sp. ( D S M 3747) was chosen for further e x p e r i m e n t s able to o f L - T r p as the wild strain with a r a t i o of 85% L - T r p to 15% conditions. T h e a s s a y as well as the p r o d u c t i o n of L - T r p N 2 - a t m o s p h e r e to p r e v e n t its e n z y m a t i c d e g r a d a t i o n ( G r o s s

f o r m twice the a m o u n t N - C a r b t r p u n d e r assay were p e r f o r m e d u n d e r et al., 1987a; 1990).

Cultural conditions for growth and synthesis of D,L-5-1MH-hydrolyzing enzymes T h e following e x p e r i m e n t s were carried out in 500 m l - s h a k e flasks in m e d i u m I I I u n d e r the c o n d i t i o n s d e s c r i b e d above.

Effect of different D,L-5-IMH-concentrations It is k n o w n f r o m literature that, of m o r e t h a n thirty h y d a n t o i n derivatives a n d r e l a t e d c o m p o u n d s tested, as well the synthesis of the e n z y m e h y d a n t o i n a s e as the f o r m a t i o n o f the L - N - c a r b a m o y l a s e are s t r o n g l y i n d u c i b l e only b y the c h e m i c a l l y synthesized i n d u c e r D , L - 5 - I M H ( S y l d a t k et al., 1987). D , L - 5 - I M H at different c o n c e n t r a t i o n s was used i n these e x p e r i m e n t s in m e d i u m I I I c o n t a i n i n g 20 g 1-1 glucose a n d 6.5 g 1-1 ( N H 4 ) 2 S O 4 to investigate their effect on cell g r o w t h a n d on the synthesis of the L - t r y p t o p h a n f o r m i n g enzymes. I n c r e a s i n g e n z y m e activities were o b t a i n e d u p to a c o n c e n t r a t i o n of 1.0 g 1-1 D , L - 5 - I M H . H i g h e r c o n c e n t r a t i o n s h a d n o further s t i m u l a t i n g effect, b u t the cell yield was n o t negatively influenced b y c o n c e n t r a t i o n s o f up to 2.5 g 1-1 D , L - 5 - I M H . A l l following e x p e r i m e n t s in shake flasks were c a r r i e d o u t with 1.0 g 1-1 D , L - 5 - I M H as inducer.

Effect of different carbon sources T o the b a s a l m e d i u m I I I c o n t a i n i n g ( N H 4 ) 2 S O 4 a n d 1.0 g 1-1 D , L - 5 - I M H different C-sources were a d d e d at a final c o n c e n t r a t i o n of 20.0 g 1-1. T h e results c o n c e r n i n g cell yield a n d e n z y m e activities after 24 h are s u m m a r i z e d in T a b l e 2. T h e best g r o w t h was o b s e r v e d with glycerol, glucose a n d fructose, while o n l y p o o r g r o w t h was o b t a i n e d after 24 h with ethanol, citrate a n d acetic acid. T h e highest e n z y m e activities were o b t a i n e d b y the a d d i t i o n o f glucose, which was used as c a r b o n source for further experiments.

352 TABLE 2 Effect of carbon sources on cell growth and synthesis of the D,L-5-lMH-hydrolyzing enzymes by a m u t a n t of Arthrobacter sp. (DSM 3747) (for reaction conditions see text) Carbon source

Cell dry mass C D M (g 1-1)

Glucose Glycerol Sucrose Fructose Starch

5.42 5.84 4.90 5.24 3.99

Specific enzyme activities (mmol g - 1 h - 1 ) Carbtrp

Trp

Carbtrp + Trp

0.009 0.019 0.019 0.014 0.007

0.042 0.025 0.011 0.010 0.024

0.051 0.044 0.030 0.024 0.031

Effect of different nitrogen sources Different nitrogen sources at a final concentration of 6.5 g 1 - t were added to medium III which contained glucose and 1.0 g 1-1 D,L-5-IMH. The results of these experiments are shown in Table 3. Cell growth after 24 h was optimal with urea and ammonium-containing N-sources, while the highest L-Trp-forming enzyme activities were observed with urea, sodium nitrate and a m m o n i u m sulfate, the latter being chosen as N-source for further investigations.

Effect of organic nutrients Different complex compounds were added to the basal medium III containing glucose, (NH4)2SO 4 and 1.0 g 1-1 D,L-5-IMH. Their effects on cell growth and enzyme synthesis are summarized in Table 4. While in some cases the addition of organic nutrients had a positive effect on the 24 h cell growth of Arthrobacter sp. (DSM 3747), the L-Trp-forming enzyme activities were lessened concerning the ratio of L-Trp to N-Carbtrp. This relationship was negatively influenced by all components except yeast extract. 1.0 g 1-1 yeast extract was found to be the optimal concentration for microbial growth and enzyme synthesis and was therefore chosen as additional nutrient for bioreactor experiments (see below).

TABLE 3 Effect of nitrogen sources on cell growth and synthesis of the D,L-5-IMH-hydrolyzing enzymes by a m u t a n t of Arthrobacter sp. (DSM 3747) (for reaction conditions see text) Nitrogen source

Cell dry mass C D M (g 1-1)

(NH4)2SO 4 NH4NO 3 NaNO 3 Urea

5.15 5.15 4.20 4.43

Specific enzymatic activities (mmol g - 1 h - 1) Carbtrp

Trp

Carbtrp + Trp

0.012 0.011 0.008 0.029

0.099 0.063 0.104 0.114

0.111 0.074 0.112 0.143

353 TABLE 4 Effect of organic nutrients on cell growth and synthesis of the D,L-5-IMH-hydrolyzingenzymes by a mutant of Arthrobacter sp. (DSM 3747) (for reaction conditions see text) Specific enzymaticactivities (retool g-1 h-1)

Organic nutrient

Cell dry mass CDM (g l- ~)

Carbtrp

Trp

Carbtrp + Trp

Control Corn steep liquor Malt extract Meat extract Peptone Soya flour Tryptone Yeast extract

5.42 5.79 6.22 5.67 5.28 5.56 6.15 5.19

0.009 0.004 0.018 0.008 0.009 0.007 0.007 0.007

0.042 0.024 0.016 0.038 0.030 0.031 0.039 0.043

0.051 0.028 0.034 0.046 0.039 0.038 0.046 0.050

Effect o f the M n 2 + concentration Mn 2+ had been found to be essential for the synthesis and the activity of the enzyme hydantoinase of Arthrobacter aurescens BH20 (Syldatk et al., 1987, 1988). That is why Arthrobacter sp. (DSM 3747) was cultivated for several passages in medium III without Mn z+, before shake flasks were inoculated containing medium III containing all the components mentioned above and different concentrations of Mn 2+.

Mn 2+ had no effect on cell growth in 24 h incubations in shake flasks, whereas it was found to be essential for the synthesis of the D,L-5-IMH-hydrolyzing enzymes. A concentration of 2 mg MnSO n • 2H20 per g C D M proved to be optimal. Bioreactor experiments Starting from the investigations in the shake flasks (see above), bioreactor experiments were performed aiming at high yields of biomass with enhanced enzyme activities for an application as biocatalyst in the production of L-Trp from D,L-5-IMH (see: Gross et al., 1990). A batch cultivation was performed under the optimal shake flask conditions (see above) at pH 6.8 and 3 0 ° C to pursue the time course of growth and enzyme synthesis of the D,L-5-IMH-hydrolyzing enzymes. In medium III, containing initial concentrations of 20.0 g 1-1 glucose, 6.5 g 1-1 (NH4)2SO 4 and 1.0 g 1- t D,L-5-IMH, a biomass concentration of 10.5 g 1- t cell dry mass was reached within 12 h (/~m~x = 0.37; YX/S = 0.44). The highest specific enzyme activities were observed in the exponential growth phase but decreased distinctly after 8 h. The very same procedure, but with additional feeding of glucose under pO 2 control, resulted in a biomass concentration of 14 g 1-1 cell dry mass within 16 h (Yx/s = 0.35). Although the specific growth rate was on a constantly high level, the specific enzyme activities of the D,L-5-IMH-hydrolyzing enzymes again decreased after 8 h. This effect was quite obviously caused by the consumption of the inducer with progressive synthesis of the enzymes as could be confirmed by analysis and repetitive addition of D,L-5-IMH.

354

In this experiment the glucose concentration to be fed was e n h a n c e d to 6% (total quantity: 8% glucose). T h e concentrations of Ca 2÷, Fe z÷, Mg 2÷ and M n 2÷ were d o u b l e d to avoid growth limitations b y a lack of the bivalent cations. The p H value was regulated b y addition of 10% aqueous N H 3 solution to avoid limitation of the N-source. A d d i t i o n a l inducer D , L - 5 - I M H was added to the growing culture in portions of 1 g 1-1 after 13, 15, 17, 20 and 23 h of growth. 73.4 -~3.2 ~3.0

tO0

40

30

20

32~

Olic I~H IIH IIH I~H [~H 90

:I\\A

12.8

2

42.6 -~2.4 -~2.2 -~2.0

70

. 0ft

,o

~ 16

50 o .u_

0.1

o.os I 2

I 4

I 6

I 8

Cultivotion

I 10

I I 12 14

I I I I 16 18 2 0 22

I 21,

time (h)

Fig. 2. Cultivation of a mutant of Arthrobacter sp. (DSM 3747) in a 20-1 bioreactor with feeding of glucose by pO 2 control and continuously enhanced feeding of the inducer D,L-5-IMH. Conditions: medium III with glucose, (NH4)2SO4, 1 g 1-1 yeast extract and 1 g 1- l D,L-5-IMH; 7.5 g 1- l D,L-5-IMH were later added continuously as 10% solution in 10% aqueous poly(ethylene glycol) 400 to the culture; p H control was performed automatically by addition of 10% aqueous N H 3 solution to the culture.

356

The results of this experiment are shown in Figs. l a and b: After 30 h 51 g 1-~ glucose were consumed and a cell dry mass concentration of 25 g 1-~ was obtained (Yx/s = 0.40). In comparison with the cultivations without additional feeding of D,L-5-IMH the specific L-Trp-forming activity of the cell mass distinctly increased every time the inducer had been added to the culture, reached a maximum 1 h after the addition and again decreased. The D,L-5-IMH-concentration seemed to be insufficient to keep the specific enzyme activities on a constantly high level. Fig. 2 shows a growth experiment in which 1.0 g 1-~ D,L-5-IMH was present from the very beginning of the cultivation, but was continuously fed to the growing culture starting after 10 h of growth and enhancing its concentration continuously according to the increase in biomass (total D,L-5-IMH concentration: 8.5 g 1-~). A cell dry mass concentration of 22 g 1-~ was obtained within 20 h (Yx/s = 0.43). The specific D,L-5-IMH-hydrolyzing activ!ty reached a maximum after 6 h, decreased, but could be increased again and held on a high level during the whole cultivation. The very high quantity of the inducer seemed to be insufficient to keep the specific enzyme activity on a constantly high level. Founded on these experiments, the next studies aimed at answering the question whether it was possible to

100

3.4 b

b,O

3.2 9O

35

8O

3.0 2.8 2.6

30

70

0.4

0.35 q

2./-,

0.3

2.2 4 o

25

•

60

20 "S

L

,o 50 ~ ~

1.6

--

1.2 c~"

~, 0.15 o.

30

1.0

c~ 20

o~

15

10

1.8 m >, ~_ .y,

0.8 2O

lO

EE 0.25

0.2

0.1

0.6 0.4

0,05

0.2 ~0

2

4

6

8

10

12

14

0 16

0

I 2

I z,

I 6

I 8

10

I 12

I 14

I 16

CuItlvotion time ( h i Cultivotion time (h) Fig. 3. Cultivation of a mutant of Arthrobacter sp. ( D S M 3747) in a 20-1 bioreactor with feeding of glucose by pO 2 control without inducer at the beginning and portionwise feeding of 1 g 1-1 D , L - 5 - I M H and Mn 2+ starting after a m i n i m u m 9 h of growth. Conditions: medium III with glucose, ( N H 4 ) 2 S O 4 , 1 g 1- ] yeast extract; 1 g 1- ] D , L - 5 - I M H were later added portionwise as 10% aqueous solution to the culture; pH control was performed automatically by addition of 10% aqueous N H 3 solution to the culture; precultures for the bioreactor-inoculation were done without inducer and M n 2+.

357

induce enzyme synthesis not until the late exponential growth phase, i.e., after an inducer- and also Mn2+-free cultivation. Thus timing of inducer and Mn 2÷ addition was varied with the initial concentrations of glucose and yeast extract being 40 g 1-1 and 1 g 1-1; additional glucose was fed at the end of the exponential growth phase. The crucial findings can be drawn from Figs. 3a and b and Table 5. With the inducer and Mn 2÷ added from the onset of cultivation, cell growth and consumption of glucose were significantly retarded by about 5 h as compared with the situation where both effectors were omitted. Remarkably, in either case enzymes could be equally induced under portion-wise addition of 1 g 1-1 inducer for 4 h, reaching a specific activity of 0.20 mmol g-1 h-1 (Trp per CDM per time). Using this novel procedure now (cultivation without Mn 2+ and D,L-5-IMH from the onset, addition not before 8 h of growth, pO2-controlled feeding of glucose) the cultivation time to obtain 25 g 1-1 CDM could be reduced from 22 to 13 h (see Fig. 3a) as compared to cultivations with Mn 2÷ and D,L-5-IMH from the onset, continuously enhanced feeding of D,L-5-IMH and pO2-controlled feeding of glucose starting after 10 h of growth (see Fig. 2a). The quantity of inducer was reduced from 8.5 g 1-1 D,L-5-IMH to 1.0 g 1-1, and the specific enzyme activity of the biomass was enhanced from 0.13 mmol g-1 h-1 (Trp per CDM per time) (see Fig. 2b) to 0.20 mmol g-1 h-1 (Trp per CDM per time) (see Fig. 3b). In parallel experiments, a chemically modified inducer was tested: D,L-5-IMH was methylated at the N-3 position of the hydantoin ring. Arthrobacter sp. (DSM 3747) was not able to hydrolyze this compound enzymatically. Using 1 g 1-1 N-3-CH3-IMH as inducer in bioreactor experiments, a biomass concentration of 28 g 1-1 CDM was obtained within 28 h (see Fig. 4a). The specific enzyme activity of the biomass was significantly enhanced from 0.20 mmol g-1 h - l (see Fig. 3b) to 0.34 mmol g - 1 h - 1 ( T r p per CDM per time) (see Fig. 4b). No drop in the specific enzyme activities of the biomass could be observed.

TABLE 5 Dependence of cell growth and enzyme synthesis of Arthrobacter sp. (DSM 3747) on D,L-5-IMH (I) and N-3-CH3-IMH (NI) as inducer, M n 2÷ (Mn) and yeast extract (YE) (for reaction conditions see text) Parameter

Time of cultivation (h) Induction phase (h) Cell concentration (g 1-1) Cell yield Yx/s (g g - l ) Specific growth rate la a ( h - 1) Specific activity b (mmol g - 1 h - l )

Initial medium composition +I + Mn +YE

-I + Mn +YE

-I - Mn +YE

-1 - Mn -YE

+NI + Mn +YE

16 4 22 0.47 0.32 0.18

14 4 23 0.49 0.28 0.17

13 4 24 0.48 0.37 0.20

16 4 17 0.38 0.22 0.08

28 4 28 0.35 0.32 0.35

a determined during exponential phase of growth. b (Trp per C D M per time) determined at the end of the cultivation.

358 ~3.4

Gt~c t

4C

32

35/

28I

30

24

"

710° 9o

"~3.2 "~3.0

o

25-

C 20 o~

~

~28

-~2.4

J2.2 2.0 41.8

• 20

L

o~

i,~

50 "~ J1.6 40 ~

30 1C

Q.

,

I

6

I

z

"0.2 /0

0

8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 Cultivation

0.4

I'l

i

2

-~1.2 Y "~1.0 ? 10.6 -~0.4

~0 0;

41.4

~0,8

20

5-

0

-~2.6

i '°70

time (h)

b

0,35 03

0.25 E

0.2

~

0.15

t~ 0.1 ol O.O5

0

2 4

6

8

12 14 16 18 20 22 24

Cultivation

2

30 3234

6

time ( h )

Fig. 4. Cultivation of a mutant of Arthrobacter sp. (DSM 3747) in a 20-1 bioreactor with feeding of glucose by pO2 control with 1 g 1-] as inducer N-3-CH3-IMH and Mn2+. Conditions: medium III with glucose, (NH4)2SO 4 and 1 g 1-1 yeast extract; pH control was performed automatically by addition of 10% a q u e o u s N H 3 s o l u t i o n to t h e c u l t u r e .

Discussion In our screening, Arthrobacter sp. ( D S M 3747) was selected as the best strain for the formation of L-Trp from the c h e m i c a l l y synthesized D , L - 5 - I M H . This microorganism is able to produce L-Trp from this c o m p o u n d in at least two e n z y m a t i c reaction steps. W e f o u n d that the ability to form L-Trp from D , L - 5 - I M H is not widely distributed in nature. These results agree with literature m a i n t a i n i n g that

359

D,L-5-IMH-hydrolyzing activity has so far been described in bacteria belonging to the genera Arthrobacter (Guivarch et al., 1980; Syldatk et al., 1987) or Flavobacterium (Nishida et al., 1987; Yokozeki et al., 1987b). Of more than 30 hydantoin derivatives and related compounds tested only D , L - 5 - I M H caused a significant induction of L-Trp-forming enzyme activities in Arthrobacter (Syldatk et al., 1987) and Flavobacterium (Yokozeki et al., 1987b). The optimal concentration for Arthrobacter sp. (DSM 3747) was 1.0 g 1-1 in shake flask experiments, while 3.5-5.0 g 1-1 D , L - 5 - I M H were necessary for Flaoobacterium cells (Yokozeki et al., 1987b). Nishida et al. observed that 5.0 g 1-1 N-carbamoyl-L-tryptophan also caused induction of D,L-5-IMH-hydrolyzing enzymes in Flavobacteriurn sp. I-3, but supposed that it was caused by L-5-IMH formed enzymatically from this compound (Nishida et al., 1987). Arthrobacter sp. (DSM 3747) was able to grow in a simple mineral salts medium and utilize various compounds as C-sources. The inorganic and organic compounds used as N-sources included among others the natural cyclic amides allantoin and creatinin, whereas dihydrouracil and dihydrothymine - the substrates of the enzyme dihydropyrimidinase (EC 3.5.2.2), which is identical with D-hydantoinase (Yamada et al., 1978) - and orotic acid - the substrate of the enzyme dihydroorotase (EC 3.5.2.3) - were not accepted as C- or N-source. Investigations of the purified hydantoinase of Arthrobacter sp. (DSM 3747) will have to show whether or not the natural cyclic amides allantoin and creatinin can be hydrolyzed by this enzyme. Glucose did not inhibit the formation of the D,L-5-IMH-hydrolyzing enzymes, the same having been reported on the D-hydantoinase of Pseudomonas fluorescens (Sun, 1983), but was the best C-source for growth and the synthesis of the D,L-5-IMH-hydrolyzing enzymes of Arthrobacter sp. (DSM 3747). For Flavobacterium sp. 1-3 dextrin and sorbitol as well as sucrose were suitable C-sources (Nishida et al., 1987), the same being true of Flavobacterium sp. AJ 3940 (Yokozeki et al., 1987b). The most suitable N-sources for Arthrobacter sp. (DSM 3747) and also for Flaoobacterium sp. 1-3 and AJ 3940 were urea and (NH4)2SO 4 (Nishida et al., 1987; Yokozeki et al., 1987b). Comparable to Flavobacterium sp. 1-3 (Nishida et al., 1987), yeast extract gave the most favorable results concerning cell growth and enzyme synthesis on the addition of complex nutrients to the mineral salts medium of Arthrobacter sp. (DSM 3747). G o o d results were obtained on the growth of Arthrobacter sp. (DSM 3747) in a bioreactor after optimization of the cultural conditions in shake flasks, but the D,L-5-IMH-hydrolyzing activity of the cells distinctly decreased after 8 h of growth. Further investigations showed that the reason for this was the consumption of the inducer D,L-5-IMH, when it was added only once at the onset of cultivation. Feeding of D , L - 5 - I M H was successful in avoiding a decrease in the specific D,L-5-IMH-hydrolyzing activity of the cells, but the inducer concentration had to be enhanced continuously depending on the cell concentration. More than 8.5 g 1-1 D , L - 5 - I M H were necessary to obtain about 25 g 1-1 cell dry mass with good specific enzyme activity, an adverse effect on the economics of this process. It was possible

360

to reduce the concentration of the inducer by starting the cultivation without inducer and Mn 2÷ and beginning the feeding of both after a minimum of 8 up to 10 h of growth. This procedure resulted in a much shorter cultivation time, since inhibition of cell growth caused by Mn 2+, D , L - 5 - I M H and degradation products could be overcome under Mn 2÷- and inducer-free conditions. The highest activity without any drop was obtained using the chemically modified N-3-CH3-IMH as inducer. This process seems to be the most economic one, because no feeding of inducer was necessary and no degradation products were observed. It will be of interest to investigate whether a cultivation without inducer and Mn 2÷ at the beginning and addition of the new inducer after a minimum of 10 h of growth will also result in a faster growth and a higher activity.

Acknowledgements This work was supported by grants of the Bundesministerium fiJr Forschung und Technik, Bonn, F.R.G., and the RiJtgerswerke AG, Castrop-Rauxel, F.R.G. We thank Professor Krohn and his co-workers from the Institute of Organic Chemistry, Technical University of Braunschweig, F.R.G., for the synthesis of D,L-5-IMH. We thank Mrs. Ulrichs, Mrs. Ruske and Mr. Rasch for their technical assistance.

References Buchanan, R.E. and Gibbons, N.E. (1975) Bergey's manual of determinative bacteriology, 8th Edn. Williams and Wilkins Company, Baltimore. Carlton, B.C. and Brown, B.J. (1981) Gene mutation. In: Gerard, P., (ed.), Manual of methods for general bacteriology. American Society for Microbiology, Washington, pp. 222-242. Cotoras, D. and Wagner, F. (1984) Stereospecific hydrolysis of 5-monosubstituted hydantoins. In: Proc. 3rd European Congress of Biotechnology, Miinchen, Vol. 3, Verlag Chemic, Weinheim, pp. 351-354. Fawett, J.K. and Scott, J.E. (1960) A rapid and precise method for the determination of urea. J. Clin. Pathol.Xl3, 156-159. Gross, C., Syldatk, C. and Wagner, F. (1987a) Development of a process for the production of aromatic L-amino acids from D,L-5-monosubstituted hydantoins. In: Neijssel, O.M., Van der Meer, R.R. and Luyben, K.C.A.M (Eds.), Proceedings 4th European Congress of Biotechnol., Amsterdam, Vol. 2 Elsevier, Amsterdam, pp. 249-252. Gross, C., Syldatk, C. and Wagner, F. (1987b) Screening method for microorganisms producing L-amino acids from D,L-5-monosubstituted hydantoins. Biotechnol. Tech. 1, 85-90. Gross, C., Syldatk, C., Mackowiak, V. and Wagner, F. (1990) Production of L-tryptophan from D,L-5-indolylmethylhydantoin by resting cells of a mutant of Arthrobacter species (DSM 3747). J. Biotechnol., 523. Guivarch, M., Gillonier, C. and Brunie, J.-C. (1980) Obtention d'amino acides optiquement actifs h l'aide d'hydantoinases. Bull. Soc. Chim. Ft. 1-2, 91-95. Hassall, H. and Greenberg, D.M. (1963) The bacterial metabolism of L-hydantoin-5-propionic acid to carbamylglutamic acid and glutamic acid. J. Biol. Chem. 238, 3325-3329. Lieberman, I. and Kornberg, A. (1954) Enzymatic synthesis and breakdown of a pyrimidin, orotic acid. 1I. Dihydroorotic acid, ureidosuccinic acid and 5-carboxymethylhydantoin. J. Biol. Chem. 207, 911-924. Miller, G. (1959) Use of dinitrosalicylic acid reagent for determination of reducing sugars. Anal. Chem. 31,426-428.

361 Miyoshi, T., Hironoshin, K., Masaaki, K. and Susuma, C. (1985) Process for production of L-amino acids. EP 0159 866. Morin, A., Hummel, W. and Kula, M.-R. (1986) Production of hydantoinase from Pseudomonas fluorescens strain DSM 84. Appl. Microbiol. Biotechnol. 25, 91-96. Nishida, Y., Nakamichi, K., Nabe, K. and Tosa, T. (1987) Enzymatic production of L-tryptophan from D,L-5-indolylmethylhydantoin by Flavobacterium species. Enzyme Microb. Technol. 9, 721-725. Olivieri, R., Fascetti, E., Angelini, L. and Degen, L. (1981) Microbial transformation of racemic hydantoins to D-amino acids. Biotechnol. Bioeng. 23, 2173-2183. Sano, R., Yokozeki, K., Eguchi, C., Kagawa, T., Noda, I. and Mitsugi, K. (1977) Enzymatic production of L-tryptophan from L- and D,L-5-indolylmethylhydantoin by newly isolated bacterium. Agric. Biol. Chem. 41,819-825. Sun, W. (1983) Screening of strains producing dihydropyrimidinase and fermentation conditions. Acta Microbiol. Sinica 23, 133-142. Syldatk, C., Cotoras, D., M~511er, A. and Wagner, F. (1986) Microbial enantioselective hydrolysis of D,L-5-monosubstituted hydantoins. BTF-Biotechforum 3, 10-19. Syldatk, C., Cotoras, D., Dombach, G., Gross, C., Kallwass, H. and Wagner, F. (1987) Substrate- and stereospecificity, induction and metallodependence of a microbial hydantoinase. Biotechnol. Lett. 9, 25-30. Syldatk, C., Dombach, G., Gross, C., Miiller, R. and Wagner, F. (1988) Production of D- and L-amino acids from D,L-5-monosubstituted hydantoins. In: Blanch, H.W. and Klibanov, A.M. (eds.), Enzyme Engineering 9, New York Academy of Sciences, New York, pp. 323-329. Takahashi, S., Ohashi, T., Kii, Y., Kumagai, H. and Yamada, H. (1979) Microbial transformation of hydantoins to N-carbamyl-D-amino acids. J. Ferment. Technol. 57, 328-332. Tsugawa, R., Okumura, S., Ito, T. and Katsuga, N. (1966) Production of L-glutamic acid from D,L-5-hydantoin propionic acid by microorganisms. Agric. Biol. Chem. 30, 27-34. Yamada, H., Takahashi, S., Kii, Y. and Kumagai, H. (1978) Distribution of hydantoin hydrolyzing activity in microorganisms. J. Ferment. Technol. 56, 484-491. Yokozeki, K. and Kubota, K. (1987) Mechanism of asymmetric production of D-amino acids from the corresponding hydantoins by Pseudomonas species. Agric. Biol. Chem. 51,721-728. Yokozeki, K., Nakamori, S., Eguchi, C., Yamada, K. and Mitsugi, K. (1987a) Screening of microorganisms producing D-p-hydroxyphenylglycine from D,L-5-(p-hydroxyphenyl)hydantoin. Agric. Biol. Chem. 51, 355-362. Yokozeki, K., Sano, K., Eguchi, C., Yamada, K. and Mitsugi, K. (1987b) Enzymatic production of L-tryptophan from D,L-5-indolylmethylhydantoin by mutants of Flavobacterium species T-523. Agric. Biol. Chem. 51, 819-825.