STUDIES ON THE COSUBSTRATE SITE OF PROTEASE SOLUBILIZED NADPH-CYTOCHROME P450 REDUCTASE

Int. J. Peptide Protein Rex 16, 1980, 83-96

S T U D I E S ON THE C O S U B S T R A T E SITE O F PROTEASE S O L U B I L I Z E D NADPH-CYTOCHROME P 4 5 0 REDUCTASE

LUDWIG LUMPER, FRIEDHELM BUSCH, SEHERAZADE DiELI6, JORC HENNING and TIBOR LAZAR*

Center for Biochemistry, Justus Liebig-University of Giessen, Giessen, W . Germany

Received 6 November, accepted for publication 20 November 1979

Modification o f the protease solubilized NADPH-cytochrome P450 reductase (= NADPH-cytochrome c reductase) a t the critical S H group in the cosubstrate binding site affects K g A D p H but not V for the cytochrome c reduction. The increase of KEADpH is dependent on the size and the charge o f the substituent introduced. Substitution o f the cosubstrate site SH by the CN-, S2 O3 - and the (N-ethyl) succinimido group effects a 3-, 7- and 23-fold increase of K:ADpH, respectively. The critical SH group in the NADPH binding region can he specifically radiolabeled by N-ethyl (2,3-'4CJ maleimide after preincubation o f the reductase with unlabeled NEM in the presence of 1 mM NADP+. The selective reaction at the essential cysteine in the cosubstrate site is demonstrated by peptide mapping of the thermolytic digest and urea SDS gel electrophoresis o f the cyanogen bromide fragments o f the reductase. Protease solubilized NADPH- cytochrome P450 reductase is inactivated by reagents directed to histidine, rrginine and lysine residues. NADP (H) ( 1 mMJ and 2I-AMP (1 mM) give effective protection only for the reaction of 1,2-~yclohexanedione (12 m M ) . The functional role of the basic amino acid residues for the cosuhstrate binding by the NADPH-cytochrome P450 reductase cannot be established therefore by the modification experiments described.

The number of NADPH binding sites in the NADPH-cytochrome P450 reductase is determined to one sitelmol reductase by titration of the enzyme with NADP+ monitored by CD-spectroscopy.

Key words; basic amino acid residues; CD-spectroscopy; cosubstrate binding site; modification of a critical SH-group.

Abbreviations: Buffer A, 0.05 M potassium phosphate/l mM Titriplex 111 pH = 7.5;Nbs2, 5,5'-dithiobis (2-nitrobenzoic acid); Nbs-, 2-thio-nitrobenzoate anion; Nbs-. l-carboxy-2-nitrophenyl-5-thio-; NEM, N-ethyl- maleimide; l 4 C-NEM, N-ethy1(2,3-l4 C)maleimide; p - CMB, p-chloromercuribenzoate; SDS, sodium dodecyl $ulfate; BrCN, cyanogen bromide.

This paper forms part of the doctoral thesis of Mr. Tibor Lazar (Fachbereich Chemie Justus-Liebig- Universitat Gieoen 1979).

NADPH-cytochrome c reductase is the flavocoenzyme-containing fragment of the amphipathic membrane protein NADPH-cytochrome P-450 reductase (EC 1.6.2.4.) obtained by proteolytic digestion of liver microsomes. The enzymatic activity of the NADPH-cytochrome P 4 5 0 reductase is conserved in the hydrophilic domain NADPH-cytochrome c reductase with the exception that this fragment lacking the hydrophobic membrane anchoring peptide is

0367-8377/80/060083-14 $02.00/0 0 1980 Munksgaard, Copenhagen 83

L LL \IPr R I T .4L

unable to reduce cytochrome P450. As com- parative studies demonstrated, the kinetic parameters for the binding of the cosubstrate NADPH are identical in both forms of the enzyme.

The hydrophilic domain NADPH-cytochrome c reductase contains. like the native membrane protein, six thiol groups/mol protein (Lazar et al., 1977; Knapp et al., 1977). Modification of the NADPH-cytochrome c reductase by 5 3 ' - dithiobis(2-nitrobenzoate) (= NBs2) (1 mM; + 4") causes a complete loss of the enzymatic activip by a slow reaction with three accessible SH-groups under the formation of the (S( 5-thio -2-nit r ob enzo at e))3 -red uct ase (( Nb s ) ~ - reductase)(Lazar er al., 1977). In the presence of cosubstrate or its analogues (NADP+, 2'-AMP) one cysteinyl residue is protected against the attack of 5,5'dithiobis(2-nitrobenzoate) (Nbs,). NADPH-cytochrome c reductase substituted at the two other accessible cysteinyl residues with the Nbs-group ( ( N ~ S ) ~ -reductase) shows the specific activity of the unmodified enzyme and is completely inactivated by renewed treatment with Nbs, under the formation of the (Nbs),- reductase. Lazar er al. (1977) obtained kinetic evidence for a concurrent reaction between the irreversible inhibitor Nbsz and the cosubstrate analogues (NADP+, 2'-AMP) for the Same site of the enzyme, suggesting that there is an 'essential' cysteine residue at or near the NADPH-binding site of the NADPH-cytochrome c reductase.

In this paper we present direct evidence that the modification at this 'essential' cysteinyl residue affects the attachment of the cosubstrate NADP (H) by steric hindrance and electrostatic interactions of the groups introduced with the NADPH-binding site.

Blockage of the essential cysteinyl residue with the Nbs-group proved to be an unsuitable method for the specific introduction of a label stable under the conditions of structural studies. The observation that this special SH-group shows an increased reactivity in the (Nbs),- reductase opens a new route for the specific labeling of the NADPH-cytochrome c reductase.

A more detailed characterization of the NADPH-binding site in the NADPH-cytochrome c reductase is intended by studying the effects of histidine and tyrosine specific reagents and

the modification of basic amino acid residues on the NADPH-cytochrome c reductase activity.

MATERIALS

Enzymes, substrates, coenzymes and inhibitors were obtained from commercial sources at the highest purity quality available and were used without further purification. Sephadex G-50 and G-150, 2',5'-ADP-Sepharo= 4B and Octyl- Sepharose CL4B were products of Deutsche Pharmacia GmbH (Freiburg, F.R.G.). Potassium (14 C)cyanide (61 mCi/mmol), standardized (14C) labeled toluene (1.118 x lo6 d.p.m./g toluene) and N-ethyl(2 ,3-'4C)maleimide (8.4 mCi/mmol) were delivered from Amersham Buchler GmbH (Braunschweig, F.R.G . The premixed scintillation cocktail Riasolve' was a product of W. Zinsser Scintillators (Frankfurt, F.R.G.). Thin layer sheets Polygram CEL 300 and Polygram SIL G (Macherey and Nagel, F.R.G.) were used for peptide mapping.

METHODS

Analytical methods Assay of NADPH-cytochrome c reductase activity, spectrophotometric determination of the NADPH-cytochrome c reductase, estimation of the number of sulfhydryl groups and the flavin content in the enzyme was performed as reported earlier (Lazar er al., 1977).

Polyacrylatnide gel electrophoresis Purity of the enzyme preparations was monitored by disc eiectrophoresis in a medium pore polyacrylamids gel (total monomer concentration: 7.5% (w/v); degree of cross linkage: 2.6%; pH of separation: 9.5). Sodium dodecyl sulfate disc electrophoresis in gels polymerized from 10% (w/v) acrylamide were performed as described by Weber & Osborne (1969).

Amino acid analysis Samples for amino acid analysis were hydrolyzed 22 or 77 h at 105" with 6.7 N HCl in an evacu- ated sealed tube and analyzed with a Beckman Model 120 automatic amino acid analyzer equipped with high sensitivity cuvets and recorders.

Established techniques were used for the determination of the tryptophan content by

84

COSUBSTRATE BINDING SITE

N-bromosuccinimide (Spande & Witkop, 1967) and the N-terminal amino acid of the reductase (Gray, 1967).

Peptide mapping Digestion of the reductase (250-500 pg) with thermolysin was carried out at 50" for 8 h in 0.5% NH4HC03/1 mM CaCl, using an enzyme to substrate ratio = 1 : l O O (wlw). Peptide mapping of the digests was done as described by Chen (1976) using Polygram SIL G thin layer plates. The spots were visualized by fluorescamine (Brosius, 1978). Radioautogra- phies of the peptide maps with Polaroid film Type 766 were performed using a 0-camera LB 290 A (Laboratorium Prof. Berthold, Wildbad, F.R.G.).

Purification of NADPH-cytochrorne c reductase The isolation of the enzyme has been described previously (Lazar etal., 1977). Final purification of the reductase was achieved by adsorption on a 2',5'-ADP-Sepharose 4B column (1.4 x 20 cm) equilibrated with 50 mM potassium phosphate/ 1 mM Titriplex 111, pH = 7.5 (= buffer A) and eluted b y a linear gradient, 0-2 mM 2'-AMP in buffer A.

Preparation of ( N ~ S ) ~ -reductase Reductase ( 1 4 ~ ~ ) was incubated with 1 mM Nbs, in buffer A for 2 4 h at +4". After this time the modified enzyme was separated from excess reagent b y gel filtration through a Sephadex G-50 column (2.5 x 30cm) using buffer A as an eluant.

Preparation of ( S - ~ y a n o ) ~ -reductase by treatment o f (NbsJ3-reductase with K I4CN (NBs),-reductase ( 1 4 . 7 ~ ~ ) was treated with a 50nM solution of K14CN (spec. radioact.: 0.3 mCi/mmol) in 0.1 M potassium phosphate/ 1 nibf Titriplex 111, pH = 7.5. During the reaction aliquots were removed for the determination of the enzymatic activity and the amount of 5-thio-2-nitrobenzoate anion liberated/mol protein. After a reaction time of about 60min the sample was applied t o a Sephadex G-50 column (2.5 x 30cm). Both equilibration and elution were performed with 0.1 M potassium phosphate/l mM Titriplex 111, pH = 7.5. Elution was continuously monitored by measuring the

absorbancies at 280 and 412nm. The protein eluates were collected and concentrated by ultrafiltration. The radioactivity was determined by mixing 50 and loop1 of the sample with l O m l of Riasolve and counting in a Nuclear Chicago Mark I1 scintillation counter. Counting efficiency was determined by addition of lOpl of l4 C-labeled toluene standard t o each counting vial.

Cleavage of (S-cyano) 3-reducrase The solution of (S-cyano), -reductase ( 1 0 ~ ~ ) in 0.1 M potassium phosphatell mbf Titriplex 111, p H = 7.5 containing SOmbi KCN was brought to pH = 9.0 by addition of 2 N NaOH. After addition of 20% (w/v) sodium dodecyl sulfate in 0.05 M borate buffer, pH = 8.5 up t o a final concentration of 1% (w/v), the enzyme solution was incubated at 37" for 20 h (Kuehl et al., 1976). In some experiments the reaction mixture contained 0.1 1 mM p-CMB (molar ratio enzyme : reagent = 1 : 12).

Cyanogen bromide cleavage of NADPH-cytochrome c reductase S-carboxymethylation of NADPH-cytochrome c reductase was performed following the procedure of Kuhn e t al. (1 974). 0.03 pniol carboxy- methylated protein was dissolved in 3 ml of 70% formic acid and a 2000-fold excess of cyanogen bromide was added. The reaction was allowed to proceed at room temperature in the dark for 16 h. The reaction mixture was freeze- dried and the lyophilized material redissolved in 1 ml 0.01 bf H 3 P 0 4 / l % (w/v) SDS/8 M urea/ 1% (w/v) 0-niercaptoethanol (pH = 6.8). The CNBr-peptides were separated by urea-SDS gel electrophoresis (1 2.5% polyacrylamide; bisacrylamide : acrylamide = 1 : 10 (w/w); gel length: 7.5 cm)(Swank & Munkres. 1971). The gels were stained with Comassie brilliant blue R-250. The gels were sliced into 1-mm fractions, dissolved in 0.2ml of Soluene@ overnight at 25", and counted after addition of 2 n d Riasolve.

Modification o f arginine residues by 2.3-butanedione, 1,2-c~~clohexanedione NADPH-Cytochrome c reductase was incubated with either 1,2-cyclohexanedione or phenyl- glyoxal in 0.1 M borate buffer pH = 8.2 k 0.2

85

L. LUMPER ET AL.

(Yankeelov. 1972). The final enzyme concentration was 1 7 ~ ~ . During the course of the reactions aliquots were removed for assays of enzyme activity.

Ph o tooxida tion of NA DPH-cy ro chrome c reductase Test solution in a jacketed reaction vessel connected to a circulating water bath was illumin- ated with a 500 W lamp placed 20 cm from the sample. The emitted light was focused by a lens. The irradiated solution was stirred continuously by a small magnet. Photochemical reactions were performed at 0 f 1". The photosensitizers were added to the enzyme solution in the dark; the sample was then divided into equal parts, one was kept in the dark as a control, and the other was exposed to light.

CD-spectroscopy CD-spectra were moritored with a Cary 61 spectropolarimeter (Varian) equipped with a Wang 2000 for signal averaging. The elhpticities between 200 and 240nm were calculated as mean residue ellipticities (MRW: 116), for other wavelength ranges as molar ellipticities (MW: 6.8 x lo4).

The CD spectra of the enzyme were corrected for the contribution of the nucleotide at each concentration used in the titration; it was assumed that the spectrum of the nucleotide did not alter upon interaction with the enzyme.

The content of secondary structures was estimated according to Greenfield & Fasman (1969).

RESULTS

Characterization of the NADPH-cytochrome c reductase from pig liver microsomes The amino acid analysis of the unmodified NADPH-cytochrome c reductase purified from pig liver microsomes differs distinctly from the data reported by Baggott & Langdon (1970) but is in good agreement with the results published for the homologous enzyme from rat liver (Gum & Strobel, 1979)(Table 1). Spectro- photometric titration of the tryptophan residues by N-bromosuccinimide indicated that porcine NADPHcytochrome c reductase contains 7.5 f 0.5 residues of Trp/mol. This value is corrected

86

TABLk 1

Amino acid composition of the pig liver NADPH- cyrochrome c reductase

Mol '7c No. of residues/mol

Lysine Histidine Arginine Aspartate Threonine Serine Glutarnate Proline Glycine Alanine Valine Methionine Isoleucine Leucine Tyrosine Pheny lalanine Tryptophana Cysteineb

6.2 3.8 8.7 9.5 4.1 4.0

14.3 4.6 3.4 5.2 5.9 2.8 3.8 9.3 1.2 5.5 1.9 0.9

33 19 31 56 21 31 75 3 2 38 50 43 14 23 56 30 25

7-8 6

a Determined by titration of the NADPH-cytochrome c reductase with N-brornosuccinimide.

Determined by the reaction of 1 rnM Nbs, with NADPH-cytochrorne c reductase in the presence of 2% SDS.

for the N-bromosuccinimide consumption by thiol group oxidation. The COOH-terminal sequence cf pig liver NADPHcytochrome c reductase could be determined to {Val, Leu, Asp)-Ser COOH.

The NH2-terininal residue is identified as glutamic acid by the dansyl chloride technique.

Number of NADPH binding sites in the NADPH-cytochrome c reductase The identification of a single cysteinyl residue/ mol reductase essential for NADPH binding by the enzyme leads to the suggestion that only one NADPH binding site exists in the NADPH- cytochrome c reductase (Lazar et al., 1977). Further evidence for this assumption is obtained by changes of the CD spectrum of the NADPH- cytochrome c reductase induced by NADP' (Ki = 2.3 PM) (Fig. 1). Titration of the enzyme with NADP' flattens the CD band with negative ellipticity at 262nm. The transition reaches a maximum at 1 : l molar ratio of enzyme to

L U b U M S I KAI I?, BINUING SITE

f

1 240 260 280 300 nm

, ,q A, '+

B /

FIGURE 1 CD spectrum of NADPH-cytochrome c reductase (1OpM) in buffer A (pH = 7.5) alone (1) and after titration with 6.45 (2), 14.5 (3) and 38.7 (4) p M NADPH. In- set, the change in ellipticity at 255nm as a function of nucleotide concentration (see Methods and Fig. 2).

nucleotide (bound) as is shown for (0)255m in stoichiometry between the amount of 5-thio-2- the inset of Fig. 1. nitrobenzoate eliminated and the thiocyanate

Binding studies with e-NADP' synthesized groups formed in the reductase. In contrast to following the procedure of Neef & Huennekens the results obtained with ( N ~ s ) ~ -reductase only (1976) failed because of the low affinity of this 0.2-0.3 mol ('4C)cyanide/mol unmodified coenzyme analog to the reductase (Ki = 1 x enzyme are incorporated under identical experi- 1 0 - ~ MI. mental conditions.

Conformation of the ( N b ~ / ~ - r e d u c t a s e (Nbs), -reductase is completely inactive (Lazar et al., 1977). No significant changes of conformation compared to that of the unmodified reductase can be detected by CD-spectroscopy. The secondary structure of the enzyme remains unchanged ((0)pzmm = 7500 deg x cm2 x dmol-' ; % helix: 19.5). The ellipticities of the flavin chromophores between 350 and 500nm demonstrate that the flavin binding is not impaired by the introduction of the (Nbs)- groups (Fig. 2). The CD spectrum of the (Nbs)3 - reductase exhibits a changed dichroitic absorption exclusively in the wavelength region, where the (Nbs)-groups contribute to the absorption spectrum of the modified enzyme. The dichroitic absorption of the (Nbs)-groups in the modified reductase seems to consist of two bands centered at about 3 15 nm (negative) and 385 nm (positive).

Cyanylation of the (Nbs13-reductase The (Nbs)-group is quantitatively displaced from the ( N ~ s ) ~ -reductase during the reaction with 50mM cyanide. The liberation of 3 f 0.2mol Nbs-/mol modified reductase and the radioactivity incorporated (3.5 mol 14CN/mol reductase) are in good agreement with a 1:l

N ADPH-cytochrorne c reductase (unmodified) 3 * 0.2 I5 5 5 0.72

(S~Nethy1)succinimido)- reductase b 0.78

(Nbs), -reductase 1.3 i 0.1

25 0.4

30 0.3 k 0.1

0.75

0.25

0.3 (Nbs), -reductasec 0.7 0 0 0

Reaction ccnditions: a Release of Nbs-lmol is estimated by incubation of the (un)modified enzyme with 1 mM Nbs, (+ 4") for 24 h.

(S~Nethyl)succinimido)-reduaase is prepared by incubation of the unmodified reductase with 1 mM N-ethyl- rnaleimide for 48 h in the presence of 1 mM NADP' (pH = 7.5; i 4").

in (Nbs), -reductase the essential SH-group is completely blocked.

cyanide. The (S -~yano)~ -reductase formed is : I

reduction (Fig. 3) . The Kf;IADPH = 22.2 * 5 1

shown to be 100% as active as the unmodified enzyme with an unaltered V for cytochrome c

0.5 pM of the (S-cyano)3 -reductase is, however, increased about threefold compared with the K2ADPH value of the unmodified enzyme

:I (7.8 tial SH-group 5 0.3 p ~ ) to by the the SCN-group. conversion of Blocking the essen- of

,+;:/x2c

the reductase at the two accessible but nonessential SH-groups with the Nbs-group does not, however,affect the KgADPH value (Fig. 3).

~ 4-'/ /

/ I W I O P H I . 0 ' Y

Cleavage of the (S-cyano)3-reductase a t the modified cysteinyl residues (S-cyano), -reductase is cleaved by incubation at pH = 9.0 under denaturing conditions in the presence of 1% SDS. The staining patterns of the fragments electrophoresed in an SDS gel demonstrate incomplete cleavage of the (S- cyano)3-reductase (Fig. 4). The gel scan shows that approximately 8070 of the protein is not split into peptides. The presence of p-CMB does not lead to changes in the pattern of the splitting products, indicating that a transfer of the CN groups t o free SH groups in the protein plays no role.

Modification of the NADPH-cytochrome c reductase with N-ethylrnaleirnide The labeling of the NADPHcytochrome c

FIGURE 3 Lineweaver-Burk plots of l/(un)modified NADPH- cytochrome c reductase initial velocity against 1/NADPH. Reaction conditions: 8.1 X U of enzyme, 60pM cytochrome c, buffer A (pH= 7.51, I = 0.1 3. Curves: A- (S-cyano), -1eductase;- .- .- (Nbs), -1eductase; x-x-x unmodified reductase.

reductase by Nbsz introduces no group at the essential cysteinyl residue stable during enzymatic proteolysis or cyanogen bromide cleavage.

NEM (1 mM) slowly inactivates the unmodified reductase during an incubation period of 24h at pH = 7.5 and +4" (buffer A). Under this set of experimental conditions an average of 3.0 ? 0.2 mol ofNethyl (2.3-l4C) maleimide/ mol reductase is incorporated. In contrast to the complete destruction of the reductase

88

reductase. Side reactions of NEM with lysyl residues cannot be detected by amino acid analysis and are, according. to Smyth et al. (1964), unlikely. Unmodified NADPH-cytochrome c reductase is bound to 2’,5’-ADP- Sepharose and eluted by 2’-AMP. Elution of NADP(H)dependent enzymes bound to adsorb-

activity by 1 mM Nbs,, the extent of inactivation does not exceed 25-35%.

According to the results obtained for the modification of the reductase by Nbs, (Lazar et ol., 1977), the fraction 7f the enzyme inactivated by NEM should correlate to the fraction of the essential SH group modified in a linear relationship. On this basis the data listed in Table 2 strongly imply that the degree of modification of the essential thiol group in the NEM modified reductase is 0.25. This is consistent with four further observations: (1) the K2ADPH of the NEM modified reductase is identical with that shown by the unmodified enzyme; (2) NEM modified reductase is completely inactivated by 1 mM Nbs, under the release of 0.72 mol Nbs-fmol modified enzyme (Table 2); (3) in the presence of 1 mM NADP+ the incorporation of 2.7 f 0.2 mol of l4 C-NEM proceeds into the unmodified reductase without affecting its activity; and (4) (Nbs)~ -reductase incorporates 0.7 mol l4 C-NEMlmol. The difference in the amount of 14C-NEM bound per mol enzyme between (Nbs), - and ( N ~ S ) ~ -reductase is in good agreement with the fraction of the essential SH group modified by NEM in the ( N ~ S ) ~ -reductase, with the loss of activity observed during the modification of (Nbs), - reductase and as well with the number of SH

cosubstrate site and secondly to separate modified reductase blocked at the essential thiol group from other derivates. ( N ~ S ) ~ -reductase is completely bound to 2’,5’-ADP-Sepharose and eluted by 2‘-AMP (I mM), whereas ( N ~ s ) ~ - reductase is not bound. This result demonstrates in another way the blockage of the cosubstrate site in the ( N ~ s ) ~ -reductase.

(S(Nethy1)succinimido)-reductase synthesized by incubation with 1 mM NADP’ can be modified under formation of the (S<N- ethyl)succinimido)ha&e-reductase to the level of 25-35% residual activity by a second reaction with NEM (1 mM) after removal of the protective reagent (‘two step modification’). Essen- tially complete inacthation (95%) of the (S{Nethyl)-succinimido)bee-reductase is ac- complished by the reaction with Nbs, (1 mM) (Fig. 5). These results raise the question of whether the NADPH-cytochrome c reductase blocked at the essential cysteine by the (Nethy1)succinimido-group shows any enzymatic activity. The following data are obtained by affmity chromatography of NEM-modified reductases on 2’ ,5’-ADP-Sepharose (Fig. 5): (1) (S-(Nethyl)succinimido)-reductase is completely adsorbed on 2’,5’-ADP-Sepharose and eluted by 2‘-AMP (1 mM). (2) NADPH-cytochrome c reductase modified by the two step

a9

L. LUMPER ET AL.

IS INQihvl lwcc~n~mtdo) ~ M Y C ~ K

Soec 8nwlrv 20 6 Ulmg Total activity 179 SH grOuo%l-rible9 1 0 T o ~ I p r o l e i n 5 9 m g A , 31A.s3 7 7 lnBctlYB~~o0 by 1 mM

Nbr, lRes act 5$1

1 mM NEM I N E M Redumare 1151

~2~~~~ - 7 *M

Chant tatwe adsorpiton on 2 5 ADP Sepharmc

After 48 n remowel 01 N E M by gel iilrral,on

1 IS INQrhvliruccinrmidoJ,,.,,,,, ieductaie

spec ac1 "Ilk 9 2 U h g K k A D P H 6 2 r ~ Tola1 acr iu~ty 44 U SH B~OUPI accessible 0 4 8 Total ~ r m e i n 4 9 mg Inamvation by 1 mM Nbr

A A.,$ 6 4 i AFFINITY CHROMATOGRAPHY O N 2 5 ADPSEPHAROSE

FRACLON __i____? I F R A C T I O N I1

(eluted by buffer A1 ladsorbed and eluted by 1 mM 2 AMP /n

buffer A1

SH growor aFesstblet 0 16 SH 9rOuPs accessibleC 1 04 Spec activity 16 7 Ulmg Swc. a c r ~ ~ ~ l v 21 Ulmg Total activity 26 U Total ~ l w w 12 U Total protein 2.8mg Total protein 0 . 6 m g A > . > l A + ~ , 7 2 A,.,lA4,, 4 .8 No ,nactwsf!an by 1 mM Nhr, lnaCtivat#on bv 1 mM Nhi,

IS lN~lhylli~~~inimidol -reducrare

FIGURE 5 Inactivation of the (S-(Nethyl)succinimido)-reductase by NEM and affinity chromatography of the inactivated enzyme on 2',5'-ADP-Sepharose (4 X 1 cm column, equilibrated with buffer A, flow rate: 2.7ml/min). (S-(,~~thyl)succinimido)-reductase is prepared by incubation of the unmodified enzyme by 1 mM NEM in the presence of 1 mM NADP' (pH = 7.5; + 4", 48 h). 9: The number of the SH groups accessible is tested by reaction of the enzyme with 1 mM Nbs, a t + 4". Spec. act. are calculated as maximum velocities from Lineweaver-Burk plots.

modification technique with NEM can be separated into two components. The fraction of the (S(N-ethyl)succinimido)-&e-reductase not bound by 2',5'-ADP-Sepharose shows the same specific activity measured for the unmodified enzyme and cannot be inactivated by Nbsz (Fig. 5). This indicates a complete blockage of the essential cysteine by the (Nethy1)succinimido group, which produces about a 23-fold increase in ICE*"'" (Fig. 5). (3) The fraction of the (S(Nethyl)succinimido)-&e-reductase bound to 2',5'-ADP-Sepharose and eluted by 2'-AMP shows identical kinetic properties as the unmodified reductase (Fig. 5). This component of the @(Net hy.I)succinimido)k&e -reductase is identical with (S(Nethy1)succinimido) reductase not reacted with NEM at the essential SH-group.

Peptide mapping of the (S-{N-eth 1d{2,3-'4 C) succinimido)vl,ti,-reductase

The two step modification of the NADPH- cytochrome c reductase by NEM was designed to radiolabel the enzyme specifically at the essential cysteine byNethyl(2,3-I4 C)maleimide.

The nonessential thiol groups are exhaus- tively blocked by prolonged incubation with 1 mM NEM for 48 h (buffer A, pH = 7.5, +4") in the presence of 1 mM NADP'. After removal of excess reagent and inhibitor by gel filtration on Sephadex G-50. Nethyl(2,3-'4 C)maleimide (3.7 mCi/mmol, final concentration: 1 mM) is added and the reaction mixture maintained for further 24h at + 4". An average of 0.78 5 0.13 Nethy1(2,3-I4 C)maleimide/mol reductase is incorporated under these conditions. The (S(N- ethy1(2,3-'4C)succinimidohc,-reductase is cleaved by thermolysin and the resulting peptides are separated by two-dimensional electrophoresis and chromatography on thin layer sheets (Polygram Sil G). As shown by the autoradiographes of the peptide maps the thermolytic digests yield two I4C-labeled peptides (Fig. 6) . The extent of the specific labeling (total counts minus nonspecific labeling) in the negatively charged peptide 1 (electrophoretic mobility expressed relative to Asp: 0.71; RF = 0.67) is approximately thrice that obtained in peptide 2 (RF = 0.55). Cyanogen bromide cleavage o i Scsrboxymethylated (S(Nethy1- ( 2 ,%I4 C)succinirnidoL,.,-reductase under conditions effecting complete conversion of methionine to homoserine and homoserine lactone as measdred by amino acid analysis yields 10-1 2 discernible peptides on SDS-urea- acrylamide electrophoresis (Fig. 7). This result is in good agreement with the number of 14 methionine residues determined/mol (unmodified) reductase (Table 2). Most of the radioactivity (85%) comigrates in SDS-urea-acrylamide electrophoresis with a peptide band of low mobility (Fig. 7).

Analysis of the thermolytic peptides and CNBr fragments provides good evidence that the (S - (N- ethy1(2,3 -I4 C)succinimido)-- reductase is nearly exclusively labeled at one site by Nethy1(2,3-14C)maleimide, which is identical with the accessible cysteine in the cosubstrate binding region.

90


a 0

0 0 0

1

0

0

ORIGIN 0

* + - E C E C T R O P H O R E S E S

FIGURE 6 Fingerprint of the thermolytic peptides obtained from (S+"thyl)-(2,3 c)succinimido)i,,tiv,-reductase. The peptides are mapped using thin layer electrophoresis (Polygram SIL G plates 20 X 20 cm) in the first dimension (pyridinelacetic acid/H, 0 = 1:2:48 (v/v/v) pH = 4.4; 25 V/cm; 110 min) and chromatography (butanol-(l)/pyridine/acetic acid/H, 0 ='37.5:2.5:7.5: 30 (v/v/v/v)) in the second. Peptide spots detected by fluorescence are outlined, The cross hatching indicates the radiolabeled peptides.

Reaction of tetrathionate with NADFH-cytochrome c reductase Treatment of the NADPH-cytochrome c reductase by 0.1 -5 mM tetrathionate during 24h at +4" destroys 70% of the original activity. The modified enzyme is not bound b y 2'3'- ADP-Sepharose and shows a KZADPH 35 x

M. Similar results are obtained for the S4 O;--modified ( N ~ S ) ~ -reductase (Kf;IADPH = 25 x M , residual activity: 30%). Not consistent with these observations is the inactivation of tetrathionate treated derivates by Nbsz (1 mM) under release of 1.4mol NbS-/mOl reductase and 0.4mol Nbs-/mol (Nbs)z- reductase.

Modification of!he NADFH-cytochrome c reductase b y 2,3- bu tanedione and 1,2-~.~clohexanedione** X-ray diffraction studies of the NADPH dependent dihydrofolate reductase (Lactobacillus casei) showed that the 2'-phosphate of NADPH serves as an important anchor for the AMN ribose component of this cosubstrate. All three available oxygens are involved in charge-charge or hydrogen-bonded interactions with basic or hydroxyl groups of the enzyme (Matthews et al., 1979). Neufeld et al. (1955) reported that only cosubstrate analogs with a 2'-phosphate group are competitive inhibitors of the NADPHcytochrome c reductase. These observations led t o the suggestion that, as in other NADPH dependent enzymes, arginyl residues play an essential role in cosubstrate binding by NADPHcytochrome c reductase.

NADPHcytochrorne c reductase loses more than 50% of its activity within 3 h when incubated in 0.1 M borate, pH = 8.2 at 25". NADPH (I-lOmM) and 2'-AMP (1 mM) restrict this spontaneous decrease of activity to about 15%. Therefore the effect of 2,3-butanedione is studied in the presence of cosubstrate or competitive inhibitor. The rapid inactivation of the enzyme by 2,3-butanedione (12 mM)(O.I M borate, pH = 8.2, 25') demonstrates that the cosubstrate or its analogs do not protect the reductase against modifying reagent (Fig. 8). Within 60min a maximum number of 14 k 1 arginines/mol reductase is modified under these conditions. Reductase completely inactivated by 2,3-butanedione is not bound by 2',5'-ADP- Sepharose. This result indicates that the cosubstrate binding region is structurally altered. The level of reactivation observed after removal of modifying reagent and exchange of borate buffer against phosphate buffer (0.05 M , pH = 7.5) by gel filtration through Biogel P-10 does not exceed 30% of the original activity even when incubated for 18 h at 25". 0.5 M hydroxyl- ammonium chloride does not influence the yield of reactivated enzyme.

In the absence of NADP(H) or ?'-AMP, incubation of NADPH-cytochrome c reductase

** Part of the doctoral thesis of MI. Jorg Henning (in preparation).

91

L. LUMPER ET AL.

A

m m FROM TOP OF GEL

20 40 60 80 m m FROM TOP OF GEL

FIGURE 7 Urea SDS gel electrophoresis of the CNBr peptides obtained from (S-(N-ethy1(2,3-14 C))succinimido)hactive- reductase (spec. radioact.: 4.7 mCi/mmol protein). A: Radioactivity distribution. B: Densitometer tracing of the gel stained with Comassie Blue (200 pg digest/ gel).

with 0.5-12mM 2,3-butanedione at 25" results in pseudo first' order inactivation (Fig. 9). Because of the considerable rate of spontaneous inactivation the activity remaining is calculated as the difference between the activity in the

92

complete test system including 2,3-butanedione and the activity of a control subjected to the Same conditions at a fured time interval. Pseudo first order rate constants are derived from the slope of semilog plots activity remaining versus time. Following the procedure of Levy el ~ l . (1963) an apparent second order rate constant is estimated to 7.4f 0.3M-I min-' and a reaction order of 1.1 f 0.1 mol of 2,3-butanedione/mol NADPH-cytochrome c reductase is estimated at pH = 8.0 (25"). Semilog plots of the inactivation kinetics for the reaction between the NADPH-cytochrome c reductase and 2,3-butanedione in the presence of 2'-AMP are non-linear. 'The steepness of the curves increases with time. The form of the semilog plots indicates at least a partial abolition of the protective effect of 2'-AMP during the first phase of the inactivation (pH = 8.5).

No reaction of lysine concomitant with arginine modification occurs as judged by amino acid analysis. 1,2Cyclohexanedione causes slow inactivation of enzyme activity; at 11.7 mM 1,2cyclohexanedione (reagent: reductase = 1.4 x lo4 (mol/mol)) 26% residual activity was achieved after 8 h of reaction. This effect exceeds the spontaneous inactivation by a factor of 1.8 under the experimental conditions (0.2 M borate pH = 8.0, 25"). In the presence of 1 m~ NADP+ 1,2-cyclohexanedione does not increase the rate observed for the spontaneous destruction of reductase activity.

Inhibizion of tke NADPH-cytochrome c reductase by pyridoxal-5'-phosphate To answer the question of whether lysine residues are critically connected with cosubstrate binding by the NADPHcytochrome c reductase, we tested the effect of pyridoxal-5'-phosphate on the activity. The enzyme is completely inactivated (< 1%) by 1mM pyridoxal 5'- phosphate (molar ratio protein to reagent = 1500) at pH = 8.4 and 25" within 6h. At 4" only 49% of the activity is lost. Lowering the pH to 7.5 slows down the activity loss (residual activity: 50% after 6 h at 25"). The cosubstrate NADPH (1 mM) decreases the rate of inactivation substantially (loss of activity/6 h: 60% (pH = 8.0, 25')) but does not afford complete protection.


Time hours)

FIGURE 8 Inactivation of the NADPH-cytochrome c reductase by 2,3-butanedione in the presence of NADPH and cosubstrate analogs. Reaction conditions: 3 PM NADPH-cytochrome c reductase, 12 mM 2,3-butanedione, 0.1 M borate buffer (pH = 8.2), 25". Curves: - butanedione and cosubstrate omitted; - + 1 mM 2'-AMP, 2,3-butanedione omitted; - cosubstrate omitted; w + 1 mM 2'-AMP; x-x-x + 1 mM NADPH; - . - . - + 1 mM NADP'.

Reaction of the NADPH-cytochrome c reductase with Nacetylimidazole Incubation of the NADPH-cytochrome c reductase (5 PM) with 5 mM N-acetylimidazole (reagent: reductase = 1000 mol/mol) up to 2 4 h at pH = 7.5 (buffer A) does not change the activity of the enzyme. The result obtained is independent of temperature in the interval -I- 4 to + 25".

Reaction of the NADPH-cytochrome c reductase with carbodiimidelglycine e thy l ester Modification of the NADPH-cytochrome c reductase (5 PM) by Nethyl-N(3dimethyl)- aminopropyl carbodiimide (12 rnM)/glyche

6 m M \ \

\ \ ' 2 m M

2 3 4

1 m e lhourr l

FIGURE 9 Butanedione concentration dependence of the inactivation of NADPH-cytochrome c reductase. Reaction conditions: 3 pM NADPHcytochrome c reductase, 0.1 M borate buffer, pH = 8.2, 25". The butanedione millimolar concentrations are indicated in the figure. Time zero is defined by the addition of butanedione to the reductase solution. The activity remaining is calculated as difference between the activities of the reaction mixture and a control (without butanedione).

ethyl ester (45 mM) results in an activity loss of 80% within l00min (pH = 7.5; 4-4'). In the absence of glycine ethyl ester the decrease of activity is reduced to 40% residual activity. Because of the high instability of the reductase at pH values below 7.0 the modification with the carbodiimide/glycine ethyl ester system is performed under conditions where side reactions other than the amidation of carboxyl groups have to be excluded. For this reason (Nbs),- reductase is treated by carbodiimide/glycine ethyl ester and reduced by dithioerythritol ( 1 0 m ~ ) after removal of the reagents on Sephadex G-25. The reduced enzyme showed the full specific activity of the unmodified

93

L. LUMPER ET AL.

enzyme. Therefore a functional role of carboxyl groups (reactive at pH = 7.5) cannot be established. A possible explanation for the inactivation of the NADPH-cytochrome c reductase by carbodiimide!glycine ethyl ester is the modification of the essential thiol group. This result is consistent with a 90% protection of unmodified reductase by 1 mM NADP+.

Modification of histidyl residues in the NA DPH-cy toc hrom e c red uctase * Diethyl pyrocarbonate (3 mM) destroys about 75% of the reductase activity within 50 min at pH = 6.75 and within 20 min at pH = 7.5 (+ 4’; reductase :reagent = 1 :250 mol/mol). The loss of enzyme activity during reaction with excess diethyl pyrocarbonate can be analyzed as a pseudo first order process for at least 85% of the total reaction at pH = 7.5. NADP+ at 1 m~ gives no protection against inactivation. Even after decrease of the reagent concentration to 0.3 mM no effect of the cosubstrate is observed. With 3 m ~ diethyl pyrocarbonate and 1 8 . 4 ~ ~ enzyme the change in A235 per 20min corre- sponds to the reaction of 3.7 histidines/mol reductase (loo). The increase in A235 is completely reversed by incubation with 1 M hy- droxylamine within 5min. Enzyme activity is restored by incubation of the diethyl pyrocarbonate treated enzyme with 1 M hydroxyl- amine for 20 h (pH = 7.5;25O). (Nb~)~-reductase (13 yM) cannot be reactivated after reaction with 3 mM diethyl pyrocarbonate (pH = 6.75; + 4’) by reduction with 15 mM dithioerythritol. The ratio A273/A455 is not changed during the modificaticn reaction. A substantial loss of flavin can therefore be excluded. In summary the data presented support the theory that the inactivation by diethyl pyrocarbonate is the consequence of histidine modification and not of accessible SH-groups or tyrosine residues.

Irreversible loss of the catalytic activity is observed during illumination of the NADPH- cytochrome c reductase ( 1 2 . 8 ~ ~ ) in the presence of 3.4mM rose bengal at pH = 7.5. The plot of the logarithms of residual activity as a function of the time is apparently not consistent with first order kinetics over at least two

* Part of the doctoral thesis of Mrs. Seherezade DZeliC (in preparation).

94

orders of magnitude (to a residual activity of about 1%). An activity loss of more than 98% is observed under our experimental conditions during an illumination period of 3.5 h. The presence of 1 mM NADP’ does not prevent the photoinactivation of the reductase. The rate of the reaction seems, however, to be reduced. Modification of the enzyme at the three accessible SH-groups by the Nbs-moiety does not protect against photoinactivation. NADPH- cytochrome c reductase and its (Nbs), derivate are not affected by exposition to light for times up to 8 h under the conditions used for photooxidation in the absence of dye. Photoinacti- vated reductase with 1% residual activity does not exhibit the typical flavoprotein spectrum with absorption bands at 385 and 455nm, indicating a complete loss of the flavin chromophores. Amino acid analysis of the sample showed a complete oxidation of the histidines. Tyrosine and methionine were found as virtually unaltered by photooxidation with rose bengal.

Using 1 mM pyridoxal 5’-phosphate as sensi- tizer NADPHcytochrome c reductase lost about 95% of its activity within 7-8 h under identical conditions (pH = 7.5). The change in A273/&J from 6.5 (native enzyme) to 16 in the spectrum of the irradiated protein indicates a substantial release of flavocoenzyme. A concomitant degradation of 50% of the histidines is shorn by amino acid analysis. This value is equivalent to the oxidation of about 6-7 histidineslmol reductase. Semilogarithmic plots of the res-dual activity as a function of illumination time in the presence of pyridoxal 5’-phosphate dre apparently linear over the reaction time. 1 mM pyridoxal 5’-phosphate does not inhibit the reductase under the conditions used for photooxidation.

DISCUSSION

The inactivation of the protease solubilized NADPHcytochrome c reductase (= NADPH- cytochrome c reductase) by sulfhydryl reagents is the result ofthe modification at one accessible SH group. Lazar er QZ. (1977) demonstrated that this special functional group is located in the NADPH binding center of the NADPH. cytochrome P450 reductase. The successfu’ restoration of the full activity by transformatior


of the (Nbs), -reductase to the (Scyano), - reductase clearly excludes the participation of the critical thiol group in the catalytic me&- anism or as a direct binding group for the cosubstrate. The increase in K2ADPH induced by the substitution of the critical cysteine with the small and uncharged CN group, however, confirms our suggestion that an accessible SH group is part of the cosubstrate site of the NADPH-cytochrome P450 reductase.

The adenine rings of NADPH bind in hydrophobic clefts of the oxidoreductases (Matthews er al., 1977). It is reasonable to suppose an homologous structural architecture for the cosubstrate binding region of the NADPH- cytochrome P450 reductase. Quinoline deri- vatives with long alkyl chains are competitive inhibitors of the NADPH-cytochrome c reductase with respect to the cosubstrate site (Ichikawa & Yarnano, 1969). This observation is an indication for the presence of the adenine cleft.

Minimal structural requirement for the competitive inhibitors derived from the cosubstrate is the 2'-phosphoadenosine moiety. The complete protection of the essential cysteine against modification by NADP' implies that the critical SH group is located in the adenine binding hydrophobic cleft. As is to be expected the decrease in cosubstrate binding affhity by substitution of the essential SH group is correlated with the increasing space filling of the groups introduced. This relation is illustrated by the observation that modification of the cosubstrate site SH group by the CN- or the (S(Nethy1)- succinimido)-group increases the KZADPH of the NADPH-cytochrome c reductase by a factor 3 or 23, respectively. The introduction of negatively charged groups (e.g. sulfo-, thio- sulfato- or Nbs--) enhances the steric effects of neutral substituents by the local electrostatic repulsion of the cosubstrate.

In the last few years the functional role of the arginyl residues in the cosubstrate binding sites of NADPH dependent oxidoreductases has been demonstrated (Matthews et d., 1979). Analysis of the inactivation kinetics of the NADPH-cytochrome c reductase by 2,3- butanedione showed that the loss of activity is caused by the modification of one argvlyl residue. NADPH or 2'-AMP, however, do not offer

complete protection against activity destruction. The time course of the inactivation induced by 2,3-butanedione in the presence of ?'-AMP indicates that the protective effect of the cosubstrate analog is abolished possibly by conformational changes of the reductase following the modification of peripheral arginine groups. In contrast to these results cosubstrate completely preventsinactivation of the NADPH- cytochrome c reductase by 1,2-cyclohexanedione. It remains to be clarified whether the different reactivities of the two arginine specific reagents are due to side reactions of 2,3- butanedione or a more selective attack of 1,2- cyclohexanedione to arginyl residues in the reductase. The low degree of reversibility found for the modification by 2,3-butanedione indicates either side reactions with lysyl residues or, as preliminary experiments suggest, release of flavin. Although no modification of lysyl residues by 2,3-butanedione or 1,2-cyclohexanedione can be detected by amino acid analysis, the critical role of (one) arginyl residue(s) for cosubstrate binding by the NADPHcytochrome c reductase cannot be definitely stated

Hiwatashi & Ichikawa (1978, 1979) claimed that a histidyl residue is essential for the cosubstrate binding site of NADPH-cytochrome P450 reductase. In contrast to their results we are unable to confirm cosubstrate protection against the inactivation of the NADPH-cytochrome c reductase by diethyl pyrocarbonate. In view of our experiments more information is needed to establish the functional role of histidyl residues in the NADPH-cytochrome P450 reductase. A similar situation is found for the lysyl residues.

Mapping of the thermolytic peptides and urea SDS gel electrophoresis of the BrCN fragments shows a preferential incorporation of (Nethy1(2,3-14 C)maleimide into one site of the reductase by the 'two step modification technique'. The effects of this procedure on the reductase activity and the result of the affinity chromatography of (S-(Nethy1)succinimldo) inactive reductase on 2',5'-ADP-Sepharose give strong evidence that the critical SH group is modified by I4C-NEM. The radioactive peptides contain therefore sequences of the NADPH binding site. Sequencing should give infor- mations on the structure of the cosubstrate

95

L. LUMPER ET AL.

binding site in the NADPH-cytochrome P450 reductase not available by modification studies.

ACKNOWLEDGMENTS

The expert technical assistance of Mrs. M. Wagner is gratefully acknowledged. We are indebted to Dr. J. Fohles (Aachen) and Dr. D. Linder (Giessen) for performing the amino acid analysis.

REFERENCES

Baggott, J.P. & Langdon, R.G. (197O)J. Biol. Chem.

Brosius, J. (1978) Biochemisrry 17,501-508 Carraway, K.L. & Koshland, D.E., Jr. (1972) Merhods

Chen, R. (1976) Hoppe-Seylers 2. Physiol. Chem. 357,

Gray, W.R. (1967)MerhodsEnzymoZ. 11,139-151 Greenfield, N. & Fasman, G.D. (1969) Biochemistry 8,

Gum, J.R. & Strobel, H.W. (1979)J. Biol. Chem. 254,

Hiwatashi. A. & Ichikawa, Y. (1978) J. Biochem. 84,

Hiwatashi, A. & Ichikawa, Y . (1979) Biochim. Bio-

Ichikawa, Y. & Yamano, T. (1969) J . Biochem. 66,

Knapp, J.A., Dignam, J.D. & Strobel, H.W. (1977)

Kuehl. G.V., Lee, M. & Muench, K.H. (1976) J. Biol.

245,5888-5896

Enzymol. 25,616-623

873-886

41 08-41 16

4177-4185

1071 -1 086

phys. Acta 580,44-63

35 1-360

J. Biol. Chem. 252,437-443

Chem. 251,3254-3260

Kuhn, R.W., Walsh, K.A. & Neurath. H. (1974) Bio- chernisfr.y 13, 3871 -3877

Larar, T., Ehrig. H. & Lumper, L. (1977) European J. Biochern. 76, 365-371

Levy. H.M., Leber, P.D. & Ryan, E.M. (1963) J, Biol. Cliern. 238, 3654-3659

Matthews, D . k , Atden, R.A., Freer, S.T., Nguyen-huu Xuong & Kraut, J . (1979) J. Biol. Chem. 254,

Neef, V.C. & Huennekens, F . M . (1976) Biochemistry

Neufeld, E.F., Kaplan, N.O. & Colowick, S.P. (1955)

Seelig, G.F. & Colman, R.F. (1977) J. Biol. Chem.

Smyth, D.G., Blumenfeld, 0.0. & Konigsberg, W.

Spande, T.F. & Witko?, B. (1967) Merhods Enzymol.

Swank, R.T. & Munkres, K.D. (1971) Anal. Biochem.

Weber, K. & Osborne, M. (1969) J. Biol. Chem. 244,

Yankeelov, J.A., Jr. (1972) Methods Enzymol. 25,

4144-4 15 1

15,4042-4047

Biochim Biophys. Acta 17,525-535

252,3671 -3678

(1964) Biochem. J. 91,589-595

11,506-532

39,462-477

4406-4412

566 -5 79

Address: Prof. Dr. Dr. Ludwlg Lumper Zentrum fiu Biochemie der Justus Liebig -

Friedrichstrde 24 6300 Giessen W. Germanv

Universitat GieDen

96

Documents

STUDIES ON THE COSUBSTRATE SITE OF PROTEASE SOLUBILIZED NADPH-CYTOCHROME P450 REDUCTASE