J. Basic Microbiol. 32 (1992) I , 21 -27

(Fachrichtung Biologie der Ernst-Moritz-Arndt-Universitat Greifswald. Institut fur Biochemie)

Purification and characterization of an inducible L-lysine : 2-oxoglutarate 6-aminotransferase from Candida utilis THOMAS HAMMER and RUDIGERBODE (Received 14 Jirrie 199l/Accepted 26 Septutnher 1991)

L-Lysine: 2-oxoglutarate 6-aminotrdnsferase (EC 2.1.6.36) was purified 202-fold from the yeast Canciitlr utilis. The subunit M , estimated by sodium dodecyl sulfate polyacrylamide gel electrophoresis was 40 kDa. The M , of the native enzyme was estimated to be 83 kDa by gel filtration, suggesting a dimeric structure. The enzyme exhibits absorption maxima a t 280, 340 and 420 nm. and binds 2 mol pyridoxal-5-phosphate/mol of the native enzyme. The aminotransferase has a maximum activity at pH 8.5 and at 4 "C. 2-Oxoglutarate is the best amino acceptor with L-lysine as amino donor. Lower activity is observed with oxaloacetate (38%), pyruvate (19%) and 2-oxoadipate (7?'0) as acceptor or with L-thialysine (13%) as donor. The K , values are 2.5 mM for L-lysine, 3.8 mM for 2-oxoglutarate and 0.015 mM for pyridoxal-5-phosphate. The enzyme activity i s induced in cells grown in the presence of L-lysine.

The flux through the lysine pathway in yeast cells is controlled predominantly by regulation of the activity of its first enzyme, homocitrate synthase (BHATTACHARJEE 1985). In contrast, the knowledge on the degradation of this amino acid and its regulation is very poor. Recent investigations have shown that in different yeast species at least three enzymatic steps could be found to initiate lysine degradation (HAMMER et al. 1991). The amino acid is catabolized either via acetylated intermediates by action of acetyl-CoA: L-lysine N-acetyltransferase or is converted to 2-aminoadipate-6-semialdehyde by L-lysine 6dehydrogenase or by L-lysine 6-aminotransferase. Cundidu utilis, an excellent fodder-yeast, grows on media with lysine as sole nitrogen and carbon source. The first enzyme of its lysine catabolism could be identified as an aminotransferase (HAMMER et al. 1991). In the present study, we described the complete purification and enzymatic properties of L-lysine : 2-oxoglutarate 6-aminotransferase (LAT, EC 2.6.1.36) from this yeast. It is the firstly extensive characterization of this enzyme from an eukaryote. Materials and methods Strain and growth conditions: C. utilis H92 is a strain from our laboratory and was pregrown for 20 h at 30 "C on a rotary shaker (1 10 rpm). Minimal medium (TANAKA et rrl. 1967) supplemented per I with 1 mg biotin, 10 g glucose as carbon source and 40 mM NH,H2P0, as nitrogen source was used. The cells were harvested at the end of logarithmic growth phase, washed with distilled water, transferred into minimal medium containing 5 mM L-lysine per I as sole nitrogen source (C-source: 10 g glucose/l) or into medium containing I0 g L-lysine per 1 as sole carbon source (N-source: 40 mM NH,H,PO,). and incubated for 10 h (logarithmic growth phase) a t 30 "C on a rotary shaker. Lysine concentration in the media was determined according to VOCELand SHIMURA (1971). Enzyme preparation: All steps were carried out at 4 " C . Step 1 : The washed cells were resuspended in buffer A (100 mM Tris/HCl, pH 8.5) and disrupted by passing them twice through an X-pressure cell. Crude extract was obtained after removal of cell debris by centrifugation at 20.000 x g for 20 min.

22

T. HAMMER and R. BODE

Step 2 : Pondered ammonium sulfate i t a s added slowly to the crude extract up to 40'% saturation. After 20 min thc solution ivas centrifugated a 5 beforc. More ammonium sulfate was added to the supernatant up to 60% saturation. After further 20 niin the precipitated protein was collected by centrifugation. The precipitate w a s dissolved in ;I minimal amount of buffer A. then the solution was desalted by chromatography on Sephadex G25. Step 3 : The desalted protein solution was loaded onto ti DEAt: cellulose colunin ( 5 x 30 cm). which had been preequilibrated with buffer A. After washing the column with buffer A. the enzyme u a h eluted with ;I linear gradient (500 mi) of 0-0.5 M KCI in buffer A (tlow rate: 80 ml h, 10-ml fractions). The fractions containing enzyme activity were pooled and glycerol \S;IS added to the solution to gikc a concentration or 20%. Step 4: This solution was column ( 2 i20 cm) preequilibrntcd in buffer B (100 inM directly lo;idcd onto ii hydrosylupatite-Utr~)~el Tris HCI. 10 m u potassium phosphate. 20"b glycerol. ptl 8.5). The column was Iirst washed with buffer B. The enrynic \\.as eluted using ii linear gradient (300 ml) 01' 10- 100 niM potassium phosphate i n buffer B ( f l o w rate: 50 nil h. 6-nil fractions). Fractions contoining the cnzyinc activity were pooled. 5 g ammonium sulfate per 10 ml of the solution uere added and after 20 min the solution was ccntrifugated. The precipitate \\as dissolved in 2 i d of buffer B. Step 5 : This enzyme solution was loaded immediately onto a Sephadex G200 column ( 2 x 50 cm), which had been preequilibratcd and eluted i n buffer B (flow rate: 15 mi h . 1 .5-nil fractions). Fractions containing the enzyme werecombined. Step 6: The combined sample was applied to a hydroxylapatite column ( 2 x 2 cm). which had been prcequilibrated in buffer B. The column w a s first washed with buffer B, then the en7yme was eluted using ;I linear gradient (I00 nil) of 10- 150 nib4 pot iuni phosphate i n buffer B (no\ rate: 20 nil,'h. I . 1 -ml fractions). Fractions containing enzyme activity Here pooled and concentrated by Centriprep-I0 centrifugation. The apo-enzyme ~va sprepared by repeated acid dialysis treatment by thc method described b j SkxmA C'I id. (1976). En7ymc-bound pyridoual-5-phosphate determination: The enzyme (0.2 mg) was treated with 1 ml of 35 mhi lH,SO, and heated for 2 h at 120 C. The liberatcd pyridoxal-5-phosphate was determined using thc method described b!, WADAand SUFIt (1961). Enzyme assay: The aminotransferasr w'i\ assayed by determining the amount o f 2-;~niinoadipatcand SHIMLRA1962). The reaction mixture contained 10 mM r-lysine. 6-semialdehyde formed (SAGISAKA 10 niM 2-ozoglutarate. 0. I niit pyridoxal-5-phosphate. 100 nib* Tris, HCI buffer (pH X.5) and enzyme \elution i n a final volume o f I nil. After 2 0 min of incubation at 37 C. I ml of a solution of 2% p-diineth~laminobenzaldeh~de in 2-methouyethanol was added. The mixture was boiled foi- 20 min and ccntrifugated. and the absorbance at 356 nm was determined. The glutamate formed in the transaminase reaction betvreen various amino acids and 2-oxoglutarate was measured by a coupled and N VERMEEKSC'II ( 1987) using reaction ticcording to the procedure described by DERG A R A B E D I A glutamate dehydrogenase. N A D . dichlorophenol-indophenol and phenazine methosulfate. Protein determination: Protein concentration in enzyme extracts was determined by the method described by LOWRY c r ( I / . (1951) using bo\.ine serum albumin as a standard. while those in column effluents \+ere monitored bq OD280. Molecular mass determination: .W, of the nati+c'anirnotransferase was estimated by Scphadex (3200 gel filtration. Thc column ( 2 x 50 cm) w s equilibrated and eluted with buffer A. The tlow rlite was adjusted to be I 5 nil h and 0.7 nil-samples were collected. The calibration proteins \\err catalnse (240 kDa). alcohol dehydrogenase ( I 5 0 kDa). bovine serum albumin (68 kDa) and ovalbumin (45 kDa). Disc gel electrophoresis in sodium dodecyl sulfate was performed as described by LAEMMLI (1970). .\Ir tviis calculated uith respect to the following markers: catalase (60 kDa), bovine serum albumin (68 kDa). ovalbumin (35 kD:i) and chynotrypsinogen ( 2 5 kDa). *

Results The lysine aminotransferase (LAT) activity was normally present in only small amount in C'. urdis cells prown in minimal medium. but activity was much higher in extracts from cells grown in the presence of L-lysine (Table I ) . The highest enzyme activity was obtained if Coritiidr cells were cultivated in medium containing L-lysine as the sole carbon or as the sole carbon and nitrogen source. Under these conditions LAT activity was increased more than 8-fold in comparison with that reached in cells grown in minimal medium. In the presence of cycloheximide no noticeable increase of the aminotransferase activity was observed.

23

Lysinc aminotransferase from C u n d i h irrilis

Table 1 Level of lysine aminotransferase in C'. z r t i i i ~ incubated 10 h in minimal medium with diffcrcnt composition Carbon source ( 10 !3/1)

Nitrogen source

Supplemcn t

Glucose Glucose Glucose Glucose Glucose Glucose L-Lysine L-Lysine

40 mM 40 mM 40 mM 40 mM 5 mM 5 mM 40 mM

-

")

Ammonium Ammonium Ammonium Ammonium I -Lysine I -Lysine Ammonium L-Lysine

hizyme acti\)ity") ( pk a t / mg ) 30 34 68 92

1 mM L-Lysine 5 mM L-Lysine I0 mM L-Lysine -

161

0.1 mM Cycloheximide

35

-. 75. 3

-

265

Crude extrdct was uscd

We followed the time course of both enzyme activity and lysine utilization of C. irrilis cells grown with lysine as sole carbon or nitrogen source. As shown in Fig. I , LAT activity strongly increased during growth and the maximum enzyme level was achieved in both cases between 9 and 15 h of growth; thereafter the enzyme activity decreased rapidly during further incubation. To purify and to characterize the LAT. cells were harvested after 10 h of cultivation in medium containing L-lysine as the sole carbon source. The purification starting from 35 g (fresh weight) of C. uti1i.y cells is summarized in Table 2. The preparation started with amnioniuin sulfate precipitation. The desalted 40 - 60% ammoiiiuin sulfate fraction was subjected to chromatography on DEAE cellulose. The enzyme activity began to elute a t a

10

Time (h)

Fig. 1 Time course of LAT activity (0, 0 ) from C'. utilis and lysine concentration in the medium (0. m). Cells were first grown for 16 h in minimal medium, then transferred to mcdium with 5 mM L-lysine as the sole nitrogen source (open symbols) o r to medium with 10 g L-lysine per 1 as the sole carbon source (filled symbols) and incubated at 30 'C

24

T. HAMMER and R . BODE

Table 2 Purification of lysine aminotransferase from c'. irtilis Step

~~

Total protein (mg)

Total activity (nkat)

1.740 454 50

178 250 I44

Spccific activity (nkatlnig)

Recovery

Purification

(YO)

(-fold)

~

Crude extract Ammonium sulfate DEAE cellulose HydrouylapatiteUltrogel Sephader G200 Hydroxylapatite

12 4.6 2.0

I26 88 66

0.16 0.55 2.9 10.2 19.5 32.5

100

90 52 45 32 24

1 3.4 18.1

64 122 202

concentration of 0.18 M KCI in a single peak. The combined fractions were applied to a hydroxylapatite-Ultrogel column after changing from buffer A to buffer B. The presence of glycerol in buffer B protected considerably the enzyme stability at this and following steps. LAT eluted at a concentration of about 20 mM phosphate. After Sephadex G200 filtration the enzyme was chromatographed on hydroxylapatite. The enzyme eluted at 40mM phosphate from the column. Using this purification procedure LAT was finally purified 202-fold from the original extract with a recovery of 24% (Table 2). The enzyme when examined in the presence of Tris buffer has a n optimum reactivity at pH 8.5. 5Ooh of maximum activity were expressed at pH 7.2 and pH 10.2. The aminotransferase reaction was carried out at various temperatures at p H 8.5. The reaction velocity increased linearly at temperatures from 25 to 37 "C. Maximum activity was found at 40 ^ C ; at higher temperatures the activity decreased rapidly. For LAT the activation energy calculated from the ARRHENIUS plot was 26 kJ/mol. The amino acid specificity of LAT was studied using the standard assay system. The synthesized glutamate was determined with glutamate dehydrogenase. The enzyme was highly specific for L-lysine and did not react with following amino acids: D-lysine, 6-acetyl-~-lysine,2-acetyl-L-lysine. L-ornithine, b-alanine, 5-hydroxy-~~-lysine, DL-diaminopiinelate, L-citrulline, L-arginine (each 10 mM). ~-Thialysinecould also serve as ainino donor, but the enzyme activity was only 13% of the activity with L-lysine (each 10 mM). The experimental results concerning amino acceptor specificity of LAT using L-lysine as amino donor are shown in Table 3. The highest activity was obtained with 2-oxoglutarate. but other aliphatic keto acids could also serve as substrate. Table 3 Kcto acid specilicity of lysine aminotransferase Kcto acid

Relative activity

(10 r n M )

( O/O )

2-Oxoglutaratc Oxaloacetate Pyruvate 2-Oxoadipate 2-Oxobutyrate 2-Oxoisovalerate 2-Oxoisocaproate Phenylpyruvate 4-Hydroxyphenylpyruvate

100

38 19

7 0 0 0 0 0

Lysine aminotransferase from Curididu utilis

25

A

V’ 2

1

Fig. 2 Determination of K , values from LAT. (A) Double-reciprocal plots of initial velocity against 2-oxoglutarate concentration at various concentrations of L-lysine; (inset) secondary plot from the intercepts of l / v of A. (B) Double-reciprocal plots of initial velocity against ~-1ysineconcentration at various concentrations of 2-oxoglutarate; (inset) secondary plot from the intercepts of l / v of B

MICHAELIS constants for L-lysine and 2-oxoglutarate were determined either by varying concentrations of L-lysine at several concentrations of 2-oxoglutarate or by varying concentrations of 2-oxoglutarate at several concentrations of L-lysine. The series of parallel lines obtained indicated a ping-pong-bi-bi mechanism for the LAT reaction (Fig. 2). The K, values calculated from the secondary plots of the intercepts were 3.8 mM for 2-oxoglutarate and 2.5 mM for L-lysine. The K, value for pyridoxal-5-phosphate was determined at fixed concentrations of 2-oxoglutarate and L-lysine (each 10 mM) and varying concentrations of pyridoxal-5-phosphate. A K,,, value of 15 PM was calculated. A sample of the protein solution recovered from step 6 of the purification procedure exhibited only one protein band an SDS polyacrylamide gel electrophoresis with a M , of 1 kDa. M , of the native LAT was 83 5 kDa as judged by gel filtration, strongly 40 suggesting that the native enzyme consists of two subunits of identical size. The same conclusion is also possible by determination of the enzyme-bound pyridoxal5-phosphate. The liberated coenzyme was determined to be 2 mol pyridoxal-5-phosphate/mol native LAT. The spectrum of the purified LAT at pH 8.5 is shown in Fig. 3. The enzyme exhibited absorption maxima at 280,340 and 420 nm. The addition of L-lysine (5 mM) to the enzyme solution caused a decrease in the absorption at 420nm and a small increase in that at 340 nm, indicating the conversion of the co-enzyme from its pyridoxal to its pyridoxamine form. The absorption spectrum was, however, not influenced by addition of 5 mM 2-oxoglutarate. Apo-LAT, which is enzymatically inactive, has only one peak at 280 nm.

Discussion

The results reported in this study show that C. utilis is capable of catabolizing L-lysine by LAT that converts Jysine to 2-aminoadipate-&semialdehyde. This compound is probably converted to 2-aminoadipate and glutarate via a pathway similar to that of Pseudomonas

26

0.1 W U C

n L

0

n

< 0.2

Fig. 3 Absorption spccfra of LAT (0.5 mg protein/ml, 0.1 M Tris HC1. p k l 8.5. 22 C ) . Curvc I . holo-enzyme; curve 2. holo-en7gmc nith 5 n i M 1.-lysine: curve 3. apo-enzyme LOO

300

h (nm)

ctcwcgim)str (Foi HERGILL and GUEST1977). The high derepression of the aminotransferase

activity from yeast cells grown in lysine-containing media indicates that the enzyme plays an important role in the catabolism of L-lysine in C. urilis. A significant derepression by lysine was also observed for LAT from other yeasts investigated (HAMMER et al. 1991, KIKZELP I c i / . 1983. SCHMIDT. et a/. 1988). LAT from C. zcti/i.s has been purified about 200-fold with 24% yield. This is the first report on the purification of LAT from a n eukaryote. In previous reports on a LAT from yeast sources. no attempts were made t o purify the enzyme (HAMMEK c't a/. 1991, KINZEL r t crl. 1983. SCHMIDT r t cil. 1988). In contrast. SODAand MISONO(1968) purified the enzyme of F/rrt.ohric,tPrii,nl /iite.scwis 187-fold with a recovery of 20%. The main difference of the enzyme properties from both sources is that the Fltr[.ohacteriitrii LAT can use L-ornithine ( 5 5 % of the activity in comparison with that of L-lysine) as amino donor, whereas in the case of the Ccriididu enzyme no activity with L-ornithine was observed, suggesting that LAT from C. itti/i.s has only one function in the organism. 2-Oxoglutarale is the best amino acceptor for both enzymes. Other keto acids are converted with different activity. From et a/. 20 different yeast species LAT was investigated for its keto acid specificity (HAMMER 1991). The enzyme could be classified into 4 groups. 5% of the yeasts have a LAT that reacted only with pyruvate as amino acceptor. 20% of the enzymes used only 2-oxoglutarate as substrate. The third group ( 5 0 O i 0 ) included yeast strains, whose enzyme reacted with 2-oxoglutarate as well as oxaloacetate. LAT of the remaining strains was capable of using 2-oxoglutarate. oxaloacetate and pyruvate as amino acceptor. Therefore, LAT from C. lrtilis belongs to the latter group. The basic kinetic propertics of the transaininase obtained here (K,n values: L-lysine ) somewhat different from 2.5 nni. 2-oxoglutarate 3.8 inw pyridoxal-5-phosphate 15 p ~ are those reported for the F. lurestwis enzyme (2.8 m M , 0.5 mu, 0.36 PM, respectively; SODAand MISONO1968). Both enzymes showed, however. the typical ping-pong mechanism of the enzyme reaction. Similar to LAT of Flarohtrcteriirnz (SODAand MISONO1968) the enzyme of C. utilis is a dimer of two subunits with identical molecular mass. The iM,of the native Cmdickii enzyme

Lysine aminotransferase from Cundidu

27

irtrlrr

is determined to be 83 kDa and thereby differs from that found for the F/uuobactrriuni transaminase (1 16 kDa). Another yeast, Pichia guillirr-moridii, has a LAT with a M , of 90 kDa (ScHMm'r et a/. 1987). Furthermore, the Candidu transaminase contains 2 mol pyridoxal-5-phosphate per mol of the native enzyme. It can be assumed that 1 mol of the coenzyme binds to each subunit. Possibly cooperative effects, however, based on the occurrence of two subunits were not observed during the kinetic analysis of the enzyme. The visible spectrum of the Candidu enzyme is characterized by a maximum at 420 nm. The shape of the spectrum and its alteration in the presence of the amino donor closely resemble those of other aminotransferases including LAT of F. lutescms (DERGARABEDIAN and VERMEERSCH 1987, SANADA et al. 1976, SODAand MISONO1968, KOIDEet a/. 1980). In summary, the results presented here characterize the enzyme responsible for the ability of C. utilis to use L-lysine as a source of nitrogen as well as of carbon. Prerequisite for its function in vivo is furthermore an effective uptake system for the basic amino acid and a specific induction mechanism of LAT synthesis. At present, we are engaged with both problems. References BHATTACIIARJEE, J. K., 1985. x-Amiiioadipate pathway for the biosynthesis of lysine in lower cukaryotes. CRC Crit. Rev. Microbiol., 12, 131 - 151. DER GARABEDIAN, P. A. and VERMEERSCH, J. J., 1987. Cundidu L-norleucine, leucine: 2-oxoglutarate aminotransferase. Purification and properties. Eur. J . Biochem.. 167. 141 147. FOTHERGILL, J. C. and GUEST,J. R., 1977. Catabolism of 1.-lysinc by Pscwdornorzus ueruginosu. J. Gen. Microbiol., 99, 139- 155. HAMMER. T., BODE,R., SCHMIDT, H. and BIRNBAUM, D., 1991. Distribution of three lysine-catabolizing enzymes in various yeast species. J . Basic Microbiol., 31, 43-49. KINZEL,J. J., WINSTON, M. K. and BHATTACHARJEE, J. K., 1983. Role of 1.-lysinc-x-kctoglutaratc aminotransferase in catabolism of lysine as a nitrogen source for Rhodororulu glutinis. J. Bacteriol., 155, 417-419. KOIDE,Y., HONMA,M. and SHIMOMURA, T., 1980. L-Tryptophan-n-ketoisocaproatcaminotransferase from Pseudomonas sp. Agric. Bid. Chem.. 44, 2013-2019. LAEMMLI, U . K., 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227, 680-685. LOWRY,0. H., ROSEBROUGH, N. J., FARR, A. L. and RANDALL, R. J., 1951. Protein measurement with the Fohn phenol reagent. J. Biol. Chem., 193, 265-275. SAGISAKA, S. and SHIMURA, K., 1962. Studies in lysine biosynthesis. I l l . Enzymatic reduction of sl-aminoadipic acid: isolation and some properties of the enzyme. J. Biochem., 51. 398-4404. Y . ,SHIOTANI, T., OKUNO,E. and KATUNUMA, N., 1976. Coenzyme-dependent conformational SANADA, properties of rat liver ornithine aminotransferase. Fur. J. Biochcm., 69, 507- 51 5. SCHMIDT,H., BODE,R. and BIKNBAUM, D., 1987. Lysine degradation in Pichia guilliermondii: characterization ofa novel enzyme, L-lysine: pyruvate aminotransferase. J. Basic Microbiol., 27,595 -601. SCHMIDT.H., BODE,R . and BIRNBAUM, 0..1988. A novel enzyme, ~-1ysine:pyruvate aminotransferase, catalyses the first step of lysine catabolism in Piclzici guilliermondii. FEMS Microbiol. Lett., 49, 203 - 206. SODA,K. and MISONO,H., 1968.r.-Lysine-x-ketoglutarateaminotransferase. 11. Puritication, crystallization, and properties. Biochemistry. 7. 41 10-4119. TANAKA, A., OHISHI,N. and FUKUI, S., 1967. Studies on the formation of vitamins and their function in hydrocarbon fermentation. Production of vitamin B, by Cundidu alhic,cmsin hydrocarbon medium. J. Ferment. Technol., 45, 617-623. Y . , 1971. Spectrophotometric determination of lysine. Methods Enzymol., VOGEL,H. J . and SIIIMURA, 17B, 228 - 229. WADA,H. and SNELL,E. E., 1961. The enzymatic oxidation of pyridoxinc and pyridoxaniine phosphate. J. Biol. Chem., 236, 2089-2095. -

Mailing address: Doz. Dr. R . BOW, Ernst-Moritz-Arndt-Universitlt Greifswald, Fachrichtung Biologic, Institut fur Biochemie, Jahnstrak 15a. 0-2200 Greifswald. Germany