On the Structure and Function of Reduced Nicotinamide Adenine ...

European J. Biochem. 3 (1968) 461 -472

On the Structure and Function of Reduced Nicotinamide Adenine Dinucleotide Phosphate-Cytochrome f Reductase of Spinach Chloroplasts G. FORTI and E. STURANI Laboratorio di Fisiologia Vegetale, Istituto di Scienze Botaniche dell’Universitk, Milano (Received August 11,1967)

1. NADPH-cytochrome f reductase from spinach chloroplasts has been purified, and its amino acid composition is reported. 2. The specificity of the enzyme for different electron acceptors has been studied. I n addition to higher plant cytochrome f , Euglena gracilis cytochrome f and plastocyanin are reduced. The reaction with these acceptors occurs through a two-step mechanism, whereas higher plant cytochrome f forms a ternary complex with NADPH and the enzyme. The K , values for NADPH in the reactions with the different electron acceptors, the K,’s for the electron acceptors and the turnover numbers are reported. 3. The sulphydryl reagents slowly inactivate the enzyme and enhance the fluorescence of enzyme-bound FAD and of the tryptophan residues ofthe protein. I n the presence of NADPH, the inactivation is very rapid, and FADH, is split from the apoenzyme. Preincubation of the enzyme with NADPH inactivates reversibly all the reactions except cytochrome f reductase. The results are interpreted in terms of structure to function relationships and the significance of the NADPH-cytochrome f reductase in photosynthetic electron transport is also discussed.

The photosynthetic electron transport systeni of chloroplasts includes the activity of a flavoprotein for the light-dependent reduction of NADP [1,2]. The chloroplast flavoprotein catalyzes the reduction of NADP by ferredoxin [a], the non-heme iron protein photoreduced by the light reaction I of photosynthesis [1,2]. It has been deiiionstrated in this laboratory that the same flavoprotein, a n FAD enzyme, catalyzes the transfer of electrons from NADPH to cytocrome f, [3 -51, the chloroplast cytochrome oxidizcd by light reaction I and reduced by light reaction I1 [6-8]. On the other hand, the recycling of electrons from the electronegative side of light reaction I to the electropositive side of this photochemical reaction through cyt. f , is known to be coupled to ATP formation in vivo [9] as well as in vitro [lo, 111, in the process of “cyclic” photophosphorylation. However, the pathway of electron transport in this cyclic process is still unknown. The NADPHCytochrome f reductase activity of the chloroplasts flavoprotein could account for this important metabolic €unction. It was previously demonstrated [4,5] that the NADPH-cyt. f reductase is identical with the NADPH diaphorase discovered by Amon and JagenNon-Standard Abbreviations. Fd, ferredoxin; PC, plastocyanin; cyt. f, cytochrome f from parsley leaves; cyt.,,,, f type cytochrome from Euglena gracilis; cyt. c, mammalian cytochrome c ; PCMS, parachloromercuriphenylsulfonic acid.

dorf [12] and Mar& and Servettaz [13], with the transhydrogenase of Keister et al. [14] and with NADP reductase. Several methods for the purification of the enzyme have been reported from the various laboratories engaged in the study of the different activities. The method described here is a modification of that previously reported [15]. The NADPH-Cytochrome f reductase has also been described in the diatom Navicula pelliculosa [16,17]. I n previous studies [4,5,15] on the mechanism of action of the chloroplast flavoprotein the differences between the diaphorase and the cyt. f reductase activities from the kinetic point of view have been emphasized, as well as the specificity of higher plants cytochrome f as electron acceptor. The studies presented here have extended this research to other electron acceptors of physiological interest, such as the f-type cytochrome of Euglena gracilis [18] and plastocyanin [19]. Plastocyanin, the copper protein of the photosynthetic apparatus, has been shown to have the same redox potential as cyt. f [20,35] and, like cyt. f , is oxidized by light reaction I and reduced by light reaction I1 [21--231. It was, therefore, of obvious interest t o investigate its reactivity as an electron acceptor for cyt. f reductase. The chemical and physical properties of the flavoprotein have also been further studied and are discussed here in relation to the mechanism of action of the enzyme.

NADPH-Cytochrome f Reductase of Chloroplasts

462

European J. Biochem.

METHODS

RESULTS

ACTIVITY MEASUREMENTS

PURIFICATION O F THE CHLOROPLAST FLAVOPROTEIN

All activity measurements were carried out a t 30°, in a thermostated spectrophotometer (Optica, model CF4) in Tris-HC1 buffer 0.05 M, p H 8.2. The standard reaction mixture for diaphorase activity measurements contained, in 1 ml final volume: NADP 0.5 mM, glucose-6-phosphate 5 mM, glucose6-phosphate dehydrogenase in large excess and ferricyanide 0.7 mM. The reduction of ferricyanide was followed as the decrease of absorbancy a t 420nm, and the molar extinction coefficient of 1,020 cm2 mole-1 was used for the calculations. The cytochrome f reductase activity was measured as the increase of absorbancy a t 554.5nm in a reaction mixture containing NADP and the NADPH generating system as above, and pure parsley cytochrome f+++,obtained as previously described [24]. The oxidation of cytochrome f was performed by dialysis against 0.1 mM ferricyanide, in the presence of 0.05 M Tris buffer, pH 8.0. The same procedure as above was used to measure the reduction of cytochrome 552 from Euglena gracilis, the f-type cytochrome of Euglena, obtained in the pure form according to Perrini et al. 1181, except that the reduction rate was followed at 552nm. The NADPH-plastocyanin reductase activity was measured a t 597 nm in an identical reaction mixture, except that plastocyanin was substituted for the cytochromes. Plastocyanin was purified from spinach leaves according to Katoh and Takamiya [20]. The molar extinction coefficient (reduced minus oxidized) of 19,600 cm2 mole-1 was used for the parsley [24] and Euglena 11181 cytochromes, and the value of 9,800 cm2mole-l [20] was assumed for plastocyanin (oxidized minus reduced). I n all cases, one unit activity is defined as the transfer of 1 pequivalent electron per minute. The phosphodiesterase activity was measured fluorimetrically, as the increase of fluorescence (excitation light of 450 nm, fluorescent emission a t 530nm) due to the formation of FMN from FAD, in the presence of 0.05 M Tris or phosphate buffer, p H 7.6. The fluorescence of a standard of pure FAD, before and after hydrolysis by snake venom (Naja Naja) phosphodiesterase was established for each experiment. BAD (Sigma, St.Louis) was purified according to Massey and Swoboda [25]. An 11-fold increase of fluorescence was observed upon conversion t o FMN, a t pH 7.6, by the snake venom phosphodiesterase. NADP, NAD, glucose-6-phosphate and cytochrome c (horse heart) were obtained from Sigma (St. Louis, No.) ; glucose-6-phosphate dehydrogenase was obtained from Boehringer (Mannheim). All other reagents were analytical grade. Ammonium sulfate was recrystallized in the presence of versene.

The enzyme was purified from spinach leaves by a modification of the previously described method [15]. All operations were carried out in a cold room. The leaves, rinsed in distilled water, were ground in a Waring-Blendor with 130 ml of double-distilled, deionized water for each 100 g of leaves. The homogenate was filtered through a double layer of cheesecloth, and the filtrate was made 0.05 M with Tris-HC1, pH 8. The acetone fractionation was carried out as described by Keister et al. [la], and the fraction, precipitated between 35 and 75°/o acetone, was dialyzed overnight against approx. 10-12 volumes of 5 mM Tris, p H 7.4. The dialyzed enzyme solution, clarified by centrifugation, was made 0.22 M with NaCl and filtered through a DEAE-cellulose column ( 2 . 5 ~ 6 cm, 5 for a preparation from 8-10 kg of leaves) equilibrated with 0.1 M Tris, p H 7.4, containing 0.114 M NaCl[2]. The enzyme was washed out of the column, while ferredoxin remained adsorbed on it and was eluted separately. The column was washed with 1.2-1.3 volumes of the Tris-NaC1 buffer, and the filtered enzyme was subjected to ammonium sulfate fractionation. Solid ammonium sulfate was added slowly, and the plI was maintained between 7.2 and 7.4 by the addition of 1M Tris as required. The fraction precipitated between 40 O l O and 7501, saturation was dialyzed against 0.03 M Tris, pH 7.4, and was then applied on to a DEAEcellulose column ( 5 x 6 0 cm). The amount of total protein a t this step, for a preparation from about 10 kg of leaves, was usually from 3.5-6g. The column was washed with 0.035M Tris, p H 7.4, until the effluent was protein-free and then the enzyme was eluted by a linear gradient of Tris, pH 7.4, from 0.035 to 0.35M. This chromatographic step is illustrated in Fig.1. The best fractions (having more than 100 units of activity per unit of absorbancy a t 275 nm) were pooled. Further purification could be achieved by two alternative procedures. Procedure A: Chromatography on Hydroxyapatite. The enzyme from the previous sttep was dialyzed against phosphate buffer, 0.01 M, p H 7.15 and applied on to a column of hydroxyapatite, equilibrated with the same buffer and preparedaccording to Jenkins [26]. The size of the column can be varied depending on the amount of protein; usually 150 to 1,000mg of protein were applied to columns of 2.5 x 20 cm with satisfactory reproducible results. 'I'hc length of the column was increased to 30 cm for the chromatography of 3,000-3,500 mg of protein. After the column was washed with about 2 volumes of 0.01 M phosphate, pH 7.15, the phosphate concentration was raised to 0.1 M. About one column volume of this buffer would spread the enzyme down the

VOl.3, N0.4, 1965

G. FORTIand E. STURANI

463

0.1 M phosphate, p H 7.6, containing ammonium sulfate 3 (wlv) [28b]. The enzyme prepared by this procedure, although homogeneous when subjected t o disc gel electrophoresis, may contain trace amounts of a phosphodiesterase activity, splitting FAD to yield FMN and AMP. Also NADP is split by the contaminating activity. I n a typical preparation of "pure') enzyme, of specific activity 646 units/mg protein in the diaphorase assay, the phosphodiesterase activity could split 0.028 pmoles of FAD/min/mg protein. The

column, and the elution was then achieved by increasing the phosphate concentration to 0.22 M. The eluted enzyme could be completely purified by chromatography on a column of DEAE-SephadexA 50. The enzyme was dialyzed against Tris buffer 0.13 M, p H 8, and the column (2.5x90 cm, for 500 to 3,000mg protein) was equilibrated against the same buffer. After the enzyme had been applied to the gel, the column was washed with 0.13 M Tris, pH 8.0, until the 275 nm absorbancy of the effluent dropped below 0.1. The enzyme was then eluted with 0.20 M Tris, p H 8.0. The flavoprotein obtained by this procedure is a homogeneous preparation, showing no traces of contaminants on disc-gel or starchgel electrophoresis.

-

A----a

52 3.0-$400t 150- ._ : -G z

200--

o

o.-.-.o

PHOSPHODIESTERASE DIAPHORASE

n

2 2.55'$!. v,

-: 2

3001

a

-

\ E

0

0

.-es

0.9

f

I

0 I

0

0

i

t, 100- 5 2 0 -3g -.E 1. z 5 -Ld 3 m w 200 LL 2 15-cn >- 0 Q

LD

150 k

v

100

5 -

0

- $k

050-

a

-

0

50

E

t

5

1.0-$

n 0

I 6 0.5-

2 100-

0 - 1

a

3

FRACTION NUMBER Fig. 1. Chromatography of the flavoprotein on DEAE-cellulose. Inpure enzyme traction was applied to a 5 x 60 cm DEAEcellulose column and the column washed with 0.035 M Tris, p H 7.4, until the effluent was protein free followed by elution of the enzyme with a linear gradient from 0.035 to 0.35 M Tris, p H 7.4. Fractions were collected and monitored for absorbancy at 275 nm (-) and enzymic activity ( ....)

..

Procedure B: Chromatography on Calcium Phosphate Gel. The column (2.5 x 50 cm) was filled with a mixture of calcium phosphate gel, (1:9, w/w) (prepared accordiiyg to Swingle and Tiselius [27]) and cellulose powder (Whatman). The mixture was deaerated before being poured into the column. Except for the slightly different ratio of gel to cellulose, the preparation of the column is identical t o that described by Massey [28a]. The column was washed with 5 mM Tris, p H 7.4. The enzyme from the DEAE-cellulose step was dialyzed against the same buffer and then applied on t o the column. The flavoprotein was adsorbed on top of the column, and the column was again washed with 5 mM Tris, p H 7.4. The yellow band of the enzyme was then spread along the column by 0.05 M phosphate, p H 7.6. This buffer was allowed to flow until the enzyme had travelled down two thirds the length of the column. Elution was then achieved by

phosphodiesterase activity is proportional to the enzyme concentration and activity was linear for over 2 hours. Added FAD is eventually converted quantitatively into FMN, while the FAD of the enzyme, as discussed later, is not available t o the phosphodiesterase. The contaminating phosphodiesterase, which is inhibited by pyrophosphate, can be completely separated from the flavoprotein by chromatography on the dextran gel Sephadex G-150 or G-100. This final step of the purification is illustrated in Fig.2. As expected no detectable increase of speciflo activity can be obtained; however, the peak of the flavoprotein is free of the contaminating activity. PHYSICAL AND CHEMICAL PROPERTIES

O F T H E ENZYME

Spectral Properties The spectral properties of the enzyme have been previously reported [2,4,29]. Upon reduction with NADPH, the enzyme is reduced to the semiquinone level, FADH', but not further to FADH, [4,41]. The

464


semiquinone, characterized by long wavelength absorption a t 530 nm and above 14,301, has been shown to be formed also upon anaerobic illumination of the enzyme in the presence of a donor such as versene Table 1. Purification of the spinach chloroplast flavoprotein Fraction

Protein

Activity Recovery Specific activity unitslnig protein

units

Crude extract 177,000 100 35-750/0 acetone 12,000a 147,000 83 Ammonium sulfate 4,360a 143,000 81 0.4-0.7 sat. DEAE-cellulose 364a 62,000 35 Hydroxyapatite 112b 57,000 32.2 (procedure A) 62.5b 40,000 22.6 DEAE-Sephadex 71.5b 45,000 25 Calcium phosphate (procedure B)

12.2 32.8

170 509 640c 646c

l’rotcin determined by ultraviolet absorhancy, according to Knlckar [451. b l’rotcin estimated on the basis of F A D content and ultraviolet absorbancy. C Thc specific activity of the enzyme rrported hcrc is about twice thr T d u c previously reported [4,15]. The discrepancy is due to the fact that in the previous reports the protein was estimated by the 280/260 absorbancy ratio according to Xalckar [451, a method which gives erroneous msults with the chloroplasts flavoprotein. The values obtained should be divided by the factor 2.13 in the case of the pure enzyme. a

Table 2. Aminoacid composition of the NADPH-cyt. f reductase An amount of hydrolysate corresponding to 26.2 nmoles and 68.1 nmoles respectively of enzyme-bound FAD was taken for analysis from the samples subjected or not subjected to performic acid oxidation. No correction applied for the losses of labile amino acids Aminoacid

Lysine Histidine Ammonia Arginine Cysteic acid Aspartic acid Threonine Serine Glutamic acid Proline Glycine Alanine Valine Methionine Methionine sulfone Isoleucine Leucine Tyrosine Phenylalanine Tryptophan Molecular weight per molc FAD Plus FAD Total

+

Yolar equivalents per mole of FAD

35.98 5.29 48.82 9.55 6.25 29.37 16.19 15.72 35.72 16.38 27.17 21.77 19.53 11.27 14.50 25.00 11.348 (12.3)b 13.03 5.72 (6.25) 36,668 785 37,453

Determined by the method of Moore et al. L31.321. Determined spectropliotometrically as by Beaven and Holiday[34], aftcr removal of FAD. b


[30]. The previously reported [4]molecular weight of 40,000 is in approximate agreement with the value obtained from the amino acid composition, reported in Table2, as determined by the method of Moore et al. [31,32], after 36 hours hydrolysis with 6N HC1 in a sealed, evacuated tube. Cysteine was determined by performic acid oxidation of the enzyme, followed by hydrolysis according to Schram et al. [33].Tyrosine and tryptophan were determined spectrophotometrically by the method of Beaven and Holiday [34], after removal of FAD by either acid treatment of the enzyme, or reduction with Na,S,O, in the presence of 4 M urea [ 5 ] . The ratio tyrosineltryptophan, in a number of analyses, was found to be 1.97, slightly lower than the one previously reported [15]. The high absorbancy of the enzyme a t 275 nm, i. e. 88,000 cmz mole-l of FAD [El, is explained by the absorbance of the coenzyme a t this wavelength.

Reaction Kinetics As reported previously [3-5,151, the flavoprotein enzyme is responsible for different reactions occurring in the chloroplasts, namely, ferredoxin-NADP reductase, NADPH diaphorase, transhydrogenase and NADPH-cytochrome f reductase. Only this last reaction involves the formation of a ternary complex between the enzyme, NADPH and higher plants’ cytochrome f , as demonstrated by the kinetics of the reaction [4]. The peculiar properties of the reaction with cyt. f as the electron acceptor, and the absolute absence of reaction with cytochrome c, induced us to extend the study of the cytochrome specificity to cyt.,,, from Euglena gracilis, the f-type cytochrome of the photosynthetic apparatus of this organism [18]. Though cyt.,,, of Euglena and cyt. f of higher plants have the same redox potential [18,35], very similar spectral properties, and occupy the same position in the electron transport chain of photosynthesis [7,8,36], their protein moieties differ markedly. Parsley cyt. f is a large protein molecule of molecular weight 250,000 [24], and contains one heme group per 63,000 mol. wt. [24,35]. C Y ~ . of ~,~ Euglena is a small protein, of mol. wt. 13,500, containing one heme group per molecule [18]. Both pigments are acidic proteins. The kinetics of the NADPH-Cytochrome,,, (Euglena) reaction catalyzed by the flavoprotein enzyme from spinach is illustrated in Fig.3. It can be seen that parallel straight lines are obtained in the Lineweaver-Burk plot, with different concentrations of cyt.,,,. This indicates, according to the equations derived by Alberty 1371, a two-step reaction, in which the enzyme is first reduced by the electron donor, which is oxidized, and the reduced enzyme is subsequently oxidized by the electron acceptor. The calculated K , values for NADPH and cyt.,,, are, respectively, 2.27 pM and 47 pM, and the turnover

G. FORTI and E. STURANI

Vol. 3, No. 4, 1968

CYT.552 4.6pM

/

07-1

'

0

0

10 20 MNADPHI (IIM-')

30

Fig. 3. Lineweaver- Burk plot of the NADPH-cyt. 552 reductase reaction. The reaction mixture contained in 1 ml final volume: NADP 0.5 mM, glucose-6-phosphate 5 mill, glucose-6-phosphate dehydrogenase in large excess, Cyt.,,, was present as indicated and its reduction measured as the increase of absorbancy a t 552 nm. Enzyme concentration 11 nM

465

oxidation is achieved by dialysis against ferricyanide, followed by the separation of excess ferricyanide and ferrocyanide by dialysis against buffer. It is essential to avoid prolonged exposure of the cytochromes to ferricyanide concentrations greater than 0.5- 1 mM, and to p H values above 8. If these rules are not observed, some, as yet undefined, denaturation of the cytochromes occurs, which is evidenced by the fact that some spontaneous reduction of the cytochromes by NADPH can be measured. The cytochromes denatured in this way show altered kinetic parameters e. g. K , values for cyt. f as high as 150p M have been measured, with higher turnover numbers. The same has been found to be true for cytochrome,,, (Euglena). The high value previously reported [4] for the K , of the flavoprotein for cyt. f and for the turnover number of NADPH-cyt. f reductase are distorted most probably due to cytochroine f denaturation. The oxidation of cyt. f and cyt.,,, (Euglena) may be carried out satisfactorily in the cold room (at 1--5"), by dialysis against a large excess (usually

CYT f 3.5 pM

w

L

01

05

075

10 1ANADPHI (pM-')

15

20

Fig. 4. Lineweaver-Burk plot of the NADPH-cyt. f reductase reaction. Conditions as for Fig. 3 with cyt.,,, replaced by cyt. f . Reductase activity was measured as the increase of abscrbancy at 554.5 nM. Enzyme: 11 nM

number 2,520. As previously reported [4], the Lineweaver-Burk (l/w us. ~ / N A D P H ) of the reaction NADPH --f cyt. f (parsley) with different concentrations of cyt. f, gave straight lines convergent on the negative side of the abscissae. This indicates, according to Alberty [37], the formation of a ternary complex, NADPH-enzyme-cyt. f. This finding has now been confirmed, and is illustrated in Fig.4. The calculated K , values for NADPH and cyt. f are respectively 1.37 pM and 9.1 pM, and the turnover number is 1,870. It should be emphasized that oxidised cytochrome f and cyt.,,, have to be prepared under strictly controled conditions, in order to prevent any denaturation of these pigments. The cytochromes are extracted and isolated in the reduced form, and the 31 European J. Biochem., Vol. 3

100-500 t o 1, mole/mole) of 0.1 mM ferricyanide, in 0.05 M Tris buffer, pH 7.6 to 8.0. This procedure yields reproducibly over 99O/, oxidized cyt. f and cyt.,,,, which are not reduced to any measurable extent by NADPH in the absence of the flavoprotein enzyme.

Reaction with Plastocyanin Plastocyanin, the blue copper protein of chloroplasts discovered by Katoh [19], has a redox potential of 370 mV [20], identical to the value for cyt. f. Furthermore, plastocyanin has been demonstrated to mediate electron transport in chloroplasts between light reaction I1 and light reaction I [21--231, as is also the case for cytochrome f. The problem of whether cyt. f and plastocyanin are involved "in

4ee

NADPH-Cytochrome f Reductafie of Chloroplasts

series” in the e1ect)ronflow, or “in parallel” in two distinct pathways is still controversial. It seemed of interest, therefore, to investigate the reactivity of plastocyanin as an electron acceptor for the NADPHcyt. f reductase of chloroplast. Fig.5 shows a Line-

-c

0.3

-E

I

PC 59 u M

I

> 5 02

s

-2 -11

01


polated a t infinite concentration of both substrates. It can be seen that the K , value for NADPI-I, when plastocyanin is the electron acceptor, is over 3,000 times higher than when cyt. f is the acceptor. If the kinetics of the plastocyanin reaction is measured with NADPH concentrations in the micromolar range, an inhibition by oxidized plastocyanin, competitive with NADPH, is apparent. This is shown in Fig. 6. This finding, together with the high K , value for NADPH, makes it rather unlikely that under physiological conditions the NADPH + ylastocyanin activity of the flavoprotein plays a relevant role. Sen Pietro and Lazzarini [38] have shown that the chloroplast flavoprotein (“transhydrogenase” in their terminology) can transfer electrons from NADPH to ferredoxin, taking advantage of the fact that ferredoxin ( E i = -4430 mV [39]) is rapidly reoxidized by mammalian cytochroinc c.

-_ 02

0 4 06 08 l/[NADPHI (rnM-’)

10

Fig. 5. Lineweaver-Burk plot of the hTADPH-plastocyanin reductase reaction. Conditions as for Fig.3 except that plastccyanin was substituted for the cyt.,,, and decrease of absorbancy a t 591 nm recorded as a measure cf plastocyanin reduction. Enzyme 11 nM 18 4

PC 283 p M

/

li! 14

sum: NADPH

177pM

10.7pM

0

10

20 30 40

50

60 70

1ANADPHI (mM-’)

Fig. 6. Competitive inhibition of plastocyanin reduction by oxidized plustocyanin. Conditions as for Fig. 5 except for the lowered concentration of NADPH as indicated. Enzyme concentration 11 nM

weaver-Burk plot of the NADPH-plastocyanin reaction of the chloroplast flavoprotein. A two-step reaction is indicated by the kinetic data. The K , values are 5 mM and 55 pM, respectively, for NADPH and plastocyanin. A turnover number of 4,780 equivalent electron per mole of enzyme per minute can be extra-

+ 2 cyt. c oxidified+ NADY + 2 cyt. c

+ H+.

The kinetics of this reaction havc been measured, as the rat,e of cyt. c reduction (AA,,,) and a Lineweaver-Burk plot is shown in Fig.7. It can be seen that NADPH inhibits the reaction competitively with ferredoxin. By extrapolation of the straight part of the lines obtained a t the lower concentrations of NADPH, the K , values of 0.71 pM and 10 pM are found for ferredoxin and NADPH, respectively, and a turnover number of 4,500. A t higher concentrations of cyt. c , an inhibition, competitive to ferredoxin, is observed. Evidently ferredoxin binds to cyt. c , and this complex is not available to the enzyme. The cyt. c. a t the concentration of 90pM gives very little inhibition; however, it is quite possible that the reported K , value for ferredoxin of 0.71 pM is erroneously too high, due to the cyt. c binding. Direct spectrophotometric evidence for the formation of a flavoprotein-ferredoxin complex has been reported [40]. The kinetic parameters of the different reactions catalyzed by the chloroplast flavoprotein are summarized inTable 3. The values found in three different experiments for the NADPH-cyt. f (from parsley) reaction are reported in the Table, the second experiment being the one illustrated in Fig. 4.


Vol.3, No.4, 1968

467

/

Table 3. Kinetic data of the chloroplast flavoprotein Kcaction

NADPH -f [Fe(CN,)I3NADPH + Cvt. " ,f NADPH NADPH NADPH

--f --f

--f

Cyt. 552 Plastocyanin F d --f Cyt. c

Turnover llulnber

K,,, for NAUPH micromolar

K,, for th(. elcctron acceptor, niicroriiolsr

electron equiv./minl m o l e of cnzymc 30'

viiiolar

prriolar

36,800 30.7 1,830 1.60 1,870 1.37 1,800 1.80 2,520 2.27 4,780 5,000 4,500 10

164 14 9.1 20 47 55 0.71

Reductive Inactivation of the Flavoprotein Enzyme It was previously reported that the diaphorase but not the cyt. f reductase activity of the enzyme is inactivated by preincubation with the substrate NADPH [4,5], and that this inactivation decreases as the enzyme concentration is increased [4,15], disappearing completely when the enzyme concentration is 0.6 to 1 pM or above, quite independently of the ratio NADPH/enzyme [4,15]. On the other hand, it is well established that the enzyme is reduced to the FADH' semiquinone level by NADPH and not to the fully reduced state [4,41]. It should be noted, however, that the reduction of the enzyme by NADPH can only be studied spectrophotometrically with enzyme concentrations which are high enough to prevent completely the above reported reductive inactivation. It is therefore quite 31*

possible that when the enzyme is preincubated in the abseiice of any electron acceptor with 0.5 mM NAUPH a t the very low enzyme concentrations (0.005 to 0.02 pM) used to measure its catalytic activities, complete reduction to enzyme-FADH, might occur, thus accounting for the observed reversible loss of activity. The different sensitivity of the diaphorase and cytochrome f reductase activities to reductive inactivation is of particular interest if one considers that the reaction mechanisms involved are different, as indicated by their kinetic properties. A more detailed investigation revealed that all reactions catalyzed by the flavoprotein enzyme, not involving the formation of a ternary complex, are inhibited t o approximately the same extent by a 5minute preincubation of the enzyme with NADPH, while the cyt. f reductase reaction, which involves the formation of a ternary complex, NADPHEnzyme-cyt. f , is not inhibited (see Table 4).

Effect of-XH Reagents It was previously established that both the diaphora.se and cytochrome f reductase activity of the flavoprotein enzyme are not affected by arsenite, even if this reagent is preincubated with the enzyme in the presence of NADPH 1141. This finding has now been confirmed and extended to other conditions such as the simultaneous presence of 3 M urea during the arsenite preincubation. It seems therefore unlikely that a disulfide-dithiol participates in the catalytic cycle of the enzyme. Sulphydril reagents, such as N-ethylmaleimide, inactivate rapidly the enzyme if preincubated in the presence of NADPH or another reductant able to reduce the enzyme [5]. I n the absence of a reductant, the inactivation by


468

Table 4. Inhibition of the flavoprotein reactions by preincubation with NADPH Conditions : The reaction mixture and preincubation contained 0.05 M, Tris buffer, p H 8.2; 0.5 mM NADPH, 5 mM glucose-6-phosphate, glucose-6-phosphate dehydrogenase in excess, and enzyme containing 11 ppmoles of FAD. The preincubated mixture contained no electron acceptor. The electron acceptors werc added to start the catalytic reaction a t the following final concentrations: ferricyanide 0.7 mM; cyt.,,, 8 pM; plastocyanin 16 pM; cyt. f 10 pM; ferredoxin 10 piV1 and cyt. c. 90 pM. Preincubation for 5 minutes at 30" Reaction

Preincubstion tinie

"0

NADPH + Fe(CN),

0 5

100 60

NADPH + cyt.,,,

0 5 0 5 0

100 45 100 51

NADPH + Fd

--f

cyt. c

NADPH + cyt. f

5

100 44

0 5

100 100

Table 5. Inhibition by p-chlorornercuriphenylsulphonate of diaphorase activity Conditions: standard diaphorase test-Enzyme: 10 nM. Preincubation: 5min a t 30'; enzyme lOnM, Tris buffer 0.05 M, pH 8.2, and other additions as indicated Additions to preincubation misture

None None None PCMS 10pM PCMS 100 p M PCMS 1 pM, NADPI10.5 mM PCMS 10 pM, NADPH 0.5 mM PCMS 100 pM, NADPH 0.5 mM NADPH 0.5 mM

PCMS present in reaction mixture

Activity

WM

units

None 10 100 10 100 1 10 100 Kone

of 3/1 to the enzyme (mole/mole) is sufficient to produce complete inactivation in about 90 minutes (Table 6), the inactivation being faster a t higher concentration of PCMS. Upon addition of glutathione in large excess, the activity is partially (25-500/0, in a large number of experiments) restored. Fig.8 shows that the non-fluorescent en.zymic FAD becomes fluorescent and available to added phosphodiesterase upon treatment with PCMS. The FAD

Activity

min

NADPH + Plastocyanin


.280 .280 .247 .220 .135 .046 .010 ,012 .170

Inhibition

o/io

11.8 21.4 52 84 96 96 39

N-ethylenaleimide/occursvery slowly [ 5 ] . The same has now been found to be true with a more efficient/ SH reagent, p-chloromercuriphenylsulfonate(PCMS), as shown in Table 5. It can be seen that 0.1 mM PCMS gives only 120/, inhibition when the enzyme is exposed to the inhibitor in the presence of the electron acceptor (ferricyanide in this case), while 1 pM PCMS is sufficient to produce 84O/, inhibition if the enzyme is pretreated with the inhibitor in the presence of NADPH, i. e. under the conditions required for the reduction of a disulphide bridge. The effects of the treatment of the enzyme with the sulphydril reagent, alone and in combination with reduction by NADPH, are illustrated in Table 6, Fig. 8, and Fig. 9. It can be seen that PCMS in a ratio

Table 6. Inactivation of the enzyme by PCMS Preincubation: enzyme 2.8 pM; pyrophosphate 0.05 M, p H 8.3; PCMS as indicated-At the times indicated, a convenient aliquot of the preincubated enzyme was withdrawn and its diaphorase activity measured. GSH was added in a 600 t o 1 ratio t o PCMS moles PCMS/mole of enzyme

Preincubation time

3

6

30

Enzymic activity

"I, 100 91 62

minutes

0 0.25 3 4 8 14 17 22 30 34 44 85 85 105 225

100 95 74 68 60 47 12

GSH 45 44

-

-

57 46

10 0

-

-

100 85 37

31 16 11

-

-

-

-

added 50 45

44 40

fluorescence rise, concurrent with a small increase in the fluorescence of the tryptophan residue(s) of the protein, is clearly a slower change when compared to the loss of activity, thus indicating the formation of a non-fluorescent but inactive intermediary state during the structural changes of the protein. Fig.9 shows that treatment with PCMS does not cause any shift in the absorption maximum of the enzyme a t 456 nm. When NADPH is added anaerobically in the presence of PCMS, the enzyme is fully reduced (curve 2, Fig.9) and upon reoxidation in air a shift of the absorption maximum to 449 nm is apparent (curve 3, Fig.9), indicating the release of FAD from the apoenzyme. The released FAD was quantitatively recovered either after precipitation of the apoenzyme with ammonium sulfate, or after separation from the apoenzyme by Sephadex 6-100 gel chromatography, as previously demonstrated for theNADPH reduction in the presence of urea [5]. The loss of activity under these conditions (reduction in the presence of PCMS or urea) was immediate and irreversible. All attempts

G . FORTI and E. STURANI

Vol. 3, No. 4, 1068

to reactivate the apoenzyme by adding back FAD, under a different set of conditions, were unsuccessful. The apoenzyme prepared in the described manner does not titrate the rabbit antibody which inhibits all catalytic activities when added in stoichiometric amounts to the enzyme [l,51.

DISCUSSION

The experiments described here leave no doubt that the mechanism of the reaction of the spinach chloroplast flavoprotein with cytochrome f from higher plants is absolutely peculiar and different from the reaction with the other electron acceptors, including cytochrome 552 of Euglena gracilis, the f-type cytochrome from the photosynthetic apparatus of this organism [is]. Indeed the steady-state kinetics experiments reported here (see Fig. 4),in agreement with those previously reported [4], indicate the formation of a ternary complex of NADPH-flavoproteincytochrome f during the catalysis of NADPH-cyt. f reductase reaction, whereas the catalysis of the reduction of plastocyanin, cyt. 552 (Euglena), and ferricyanide [4] occurs through a simple two-step mechanism. This is indicated (on the basis of the equations of Alberty [ X I ) by the fact that the Lineweaver-Burk plots of i / w us. l/[NADPH], a t different concentrations of these electron acceptors, give parallel straight lines (see Figs.3,5,6 and [4]). The interpretation of this kinetic behaviour according to the well-know equations of Alberty [37] has been more recently criticized by Massey [42], who pointed out that getting parallel lines does not ncccssarily imply the simple two step reaction :

0)

t

z 3 W

W

m

469

Q

112 W

TIME ( m i d

Fig.8. Effect of PCMS on the enzyme activity, fluorescence and the accessibility of enzyme-boundFAD to phosphodiesterase. Conditions: 2 mpmoles of enzyme were incubated in the spectrofluorimeter cuvette in 0.05 M Tris p H 8.0, with 26 mpmoles of PCMS, in a final volume of 0.5 ml. At the times indicated, fluorescence measurements were made; 1 p1-samples were taken and their diaphorase activity measured. The fluorescence of a 4 WMolar FAD solution was checked a t each point, and represents the 1000/, value of the scale. The fluorescence of the protein was measured a t 340 nm upon excitation a t 295 nm; under these conditions most of the fluorescence is contributed by tryptophan residues [46]

+ AH, + EAH, + EH, + A EH, + B + EH,B + E + BH, E

According to Massey, the parallel lines are also consistent with a ternary complex mechanism of the type :

E + AH,

!

EH,A

,+EAH, Kl

+ B &A

vg+

EH,AB =K,B

K7

EH,A

1 EA

+ BH,

5051 $04

360

400

440 480 WAVELENGTH (nm)

520

Fig. 9. Spectral changes upon anaerobic incubation of the enzyme with PCMS and NADPH. The anaerobic spectrophotometer cuvette contained: pyrophosphate 0.05 M p H 8.2 and enzyme (curve 1).A 3 fold excess (mole/mole) of PCMS was added with no changes in the spectrum (curve 1). After 60 minutes the mixture of NADPH, glucose-6-P and glucose-6-phosphate dehydrogenase was added from the side arm, and curve 2 was recorded after the spectral change was completed. At this point air was readmitted into the cuvette and, after complete reoxidation, curve 3 was recorded

Provided that K , >))K4, parallel lines are also obtained by this mechanism. Though this argument of Massey is certainly valid, we feel that it is rather unlikely to apply to the case of the chloroplast flavoprotein. Indeed, the values of the K5’s (in the above reaction scheme) should be much higher than the K i s for a number of different elcctron acceptors characterized by widely different turnover numbers such as cyt. 552, plastocyanin and ferricyanide (see Table 3 and c. f. [4]).Though this possibility cannot be excluded, it seems rather improbable. The basic difference between the cyt. f reductasc reaction and the reaction with the other electron acceptors is also indicated by the fact that these latter reactions, but not the former ones, are inactivated by preincubation

470

NBDPH-Cytochrome f Reductase of Chloroplasts

of the enzyme (at low concentration) with NADPH (Table 4). It was previously reported [4,15] that the reversible NADPH inactivation does not occur when the enzyme concentration is high enough for its absorption spectrum to be measured, i. e . 5-10 pM or above, and that under these conditions NADPH does not reduce the FAD beyond the FADH‘ semiquinone level [4,41]. One could speculate that when the enzyme is a t low concentrations, such as required for catalytic activity measurement, i. e. 0.5 to lOnM, it could be fully reduced by NADPH 0.5 mM. The reduction to the FADH, level would imply reversible inactivation, if the FADH’ semiquinone form of the enzyme is admitted to be the reductant for oneelectron acceptors such as ferricyanide, plastocyanin [20] and cyt. 552 [18]. I n the case of cytochrome f from higher plants it is known that one heme group is present for 63,000 grams of protein [35,24], and a molecular weight of 110,000-120,000 has been found from ultracentrifugation studies [35,24], or 250,000 by the gel filtration method [24]. Therefore, cytochrome f , bound to the enzyme in a ternary complex together with NADPH, can be considered a two-electron acceptor, if by this it is meant that two electrons are transferred to the two heme iron atoms present in the ternary complex. With this assumption, it is conceivable that the FADH, state of the enzyme could be involved in the cyt. f reduction reaction and the observed (see Table 4 and [4]) lack of inhibition by NADPH preincubation would be explained. I n other words, the redox state of the enzyme participating in catalysis would be the FADH,-enzyme for cyt. f reduction, and the semiquinonc for all the other reactions. This admittedly speculative hypothesis needs direct demonstration by rapid spectrophotomctric techniques. Nevertheless, it is in agreement with the observations reported here on the steady-state kinetics and the reversible inactivation of the catalytic reactions studied. The comparison of the kinetics of cyt. f reductase activity with the plastocyanin reductase activity (Figs.4 and 5, Table 3) is of interest when one considers that plastocyanin and cyt. f have the same redox potential (El,,= 370 mV a t pH 7 ) and both are oxidized by light reaction I of photosynthesis and reduced by light reaction I1 [21-23, 6-81. The results reported here indicate that cyt. f can also be reduced efficiently by the flavoprotein’s catalytic activity, i . e . by the electrons made available by light reactionI. The reduction of plastocyanin by the flavoprotein, on the other hand, is unlikely to occur in vivo a t any appreciable extent, due to the high K , value of 5 mM (see Table 3 and Fig.5). Cyt. f , but not plastocyanin, is therefore a suitable candidate for a major role in the cyclic electron transport of chloroplasts coupled to ATP formation [9], while the role of plastocyanin is more properly understood as

Zuroilcaii J. Biochein.

related to the “non cyclic” electron transport between the two light reactions [21-231. The inhibitory action of oxidized plastocyanin, competitive with NADPH (Fig.6) for the active site of the flavoprotein, could be a physiologically important regulatory mechanism for the alternative flow of the electrons through the “cyclic” (light reaction I + ferredoxin --f flavoprotein --f cyt. f -+ light reaction I)

N NADPH and the non-cyclic pathway (H,O + light reaction I1 + plastocyanin --f light reaction1 3 Ferredoxin + flavoprotein --f NADPH --f CO, assimilation). Under conditions of active assimilation of CO,, NADPH is continuously reoxidized by the reductive CO, cycle, and plastocyanin therefore shifted towards the oxidized state, i. e. in the state which inhibits the NADPH-cyt. f reductase reaction. As soon as ATP shortage becomes limiting for the assimilation of CO, through the Calvin cycle, NADPH concentration increases and oxidation of plastocyanin by light reaction I becomes limited by the lack of electron acceptors on the electronegativc side of the light reaction. As a consequence, a higher concentration of NADPH and a shift towards plastocyanin reduction is to be expected, i . e. the conditions to permit the electron flow through the cyclic cyt. f pathway. The cyclic electron flow is known to be coupled to ATP synthesis [9-111, and its operation will then reactivate the GO, assimilation process, leading to NADPH utilization and plastocyanin oxidation, and therefore again to inhibition of the cyclic electron flow. The previously reported observations [4,5] and the experiments described here (Fig.!3) point out that complete reduction of the flavoprotein enzyme to FADH, by its substrate NADPH requires either the presence of moderate concentrations of urea 141 or the presence of a sulphydril reagent such as PCMS (Fig.9). I n both cases the rapid, irreversible inactivation, is accompanied by the splitting of FADH, from the apoprotein [ 5 ] (Fig.9). Recently, Foust and Massey reported [40] that the chloroplast flavoprotein (“Ferredoxin-NADPH reductase”) can accept 4 electrons per molecule froin 2 molecules of NADPH (or NADH) in the presence of Neurospora crassn NADase, which destroys NADP. The enzyme therefore can accept 4 electrons per molecule of 40,000 mol. wt., containing 1 molecule of FAD, in a two-steps process. A first, rapid reaction with one mole of NADPH per mole of enzyme leads t o the FADH’ semiquinone level (characterized by absorption i n the region of 520-600 nm [4,30,41], followcd by a second, much slower [4,40] reaction with a second molecule of NADPH to give the FADH, enzynic. This second reaction does not occur unless urea [4] or a sulphydril

Vo1.3, So.4, 1068


reagent (Fig. 9) or Neurospora crassa NADase [40]are present. These three conditions might well act through the same basic mechanism, i . e. the prevention of NADP binding to the enzyme either because NADP (formed upon oxidation of NADPH) is destroyed (in the case of NADase), or because its binding is prevented by urea or by the binding of the mercurial to an SH group essential, directly or undirectly, to the formation of the NADP-enzyme complex [15]. The protective effect of NADP on the native structure of the enzyme required for catalytic activity is also demonstrated by the observation that NADP prevents and also reverses the reductive inactivation [4]. I n this respect, the chloroplast flavoprotein is similar to lipoyl dehydrogenasc [43]and glutathione reductase [44].I n contrast to lipoyl dehydrogenase, however, the chloroplast flavoprotein is unlikely to have a disulfide-dithiol prosthetic group involved in the catalytic cycle, as indicated by the lack of inhibition by arsenite of both diaphorase and cytochrome f reductase activities [4]. The fact that the enzyme requires 4 electrons for full reduction [40]though having only one FAD forces one to search for a second electron acceptor group. This is most likely a disulfide bridge, as indicated by the dramatic increase of the enzyme sensitivity to -SH reagents upon treatment with NADPH (see Table 5). This concept is not in contradiction with the suggested non-participation of a disulfide-dithiol in the catalytic cycle. Indeed, the disulfide may be a two-electron acreptor site in the full reduction of the enzyme requiring 4 electrons, but all evidence indicates that the 4-electrons reduced enzyme is not involved in any one of the catalytic activities. On the contrary, the fully reduced enzyme obtained in the presence of urea [4]or a sulphydryl reagent (Fig.9, Table 5 , also f [ 5 ] ) is irreversibly inactivated and loses its FAD prosthetic group. It is therefore quite clcar that the full reduction of the enzyme labilizes the binding of FAD to the protein moiety of the enzyme, and makes it sensitive to -SH reagents and to low concentrations of urea, which are harmless by theniselves t o the active structure of the enzyme [4].This does not mean that the enzyme is necessarily inactivated if FAD is reduced to FADH,: on the contrary, f d l reduction can be obtained by treatment with excess dithionite, and full activity recovered upon reoxidation in air, provided that exposure t o urea, -SH reagents and transition metal ions, is avoided while the enzyme is in the reduced state. Thanks are due t o Dr. F. R i m , P. Fasella and D. Barra for performing the amino acid analyses in their analyser and t o G. F'oiist and V. Massey for providing information about their Swingle-Tiselius gel chromatography of chloroplast flavoprotein. This work was supported by the Corisiglio Nazionale delle Ricerche of Italy.

471

REFERENCES 1. San Pietro, A,, and Keister, D. L., Arch. Biochem. Biophys. 98 (1962) 235. 2. Shin, &I., Tagtzwa, K., and Amon, D. I., Biochem. 2. 338 (1963) 84. 3. Forti, G., Bertole', M. L., and Parisi, B., Natl. Acad. Sci., Natl. Res. Council, Publ. 1145 (1963) 284. 4. Zanetti, G., and Forti, G., J . Biol. Chem. 241 (1966) 279. 5. Forti, G., and Zanetti, G., I n Biochemistry of Chloroplasts (edited by T. W.Goodwin), Academic Press, Now York 1966, Vol. 11. 6. Hill, R., and Bonner, W. D., I n Light and Life (edited by W. D. McElroy and B. Glass), The Johns Hopkins Press, Baltimore 1961. 7. Amesz, J., and Duysens, L. N. M., Biochim. Biophys. Acta, 64 (1962) 261. 8. Chance, B., and Bonner, W.D., Natl. Acad. Sci., Natl. RPS.Council, Publ. 1145 (1963) 66. 9. Forti, G., and Parisi, B., Biochim. Biophys. Acta, 71 (1963) 1. 10. Amon, D. I., I n Light and Life (edited by W. D. McElroy and B. Glass), The Johns Hopkins Press, Baltimore 1961. 11. Jagendorf, A. T., and Forti, G., I n Light and Life (edited by W. D. McElroy and B. Glass), The Johns Hopkins Press, Baltimore 1961. 12. Avron. M., and Jagendorf, A. T., Arch. Biochem. Biophys. 65 (1956) 475. 13. Marre', E., and Servettaz, O . , Arch. Biochem. Biophys. 75 (1958) 309. 14. Keister, D. L., San Pietro, A., and Stolzenbach, F. E., J . Biol. Chem. 235 (1960) 2898. 15. Forti, G., In Energy Conversion by the Photosynthetic Apparatus, Brookhaven Symposia in Biology, No 19 (1966) 195. 16. Yamanaka, T., and Kamen, M. D., Biochem. Biophys. Research Commun. 19 (1965) 751. 17. Yamanaka, T., and Kamen, M. D., Biochim. Biophys. Acta, 112 (1966) 436. 18. Perini, F., Kamen, M. D., and Schiff, J. A., Biochim. Biophys. Acta, 88 (1964) 74. 19. Katoh, S., Nature, 186 (1960) 533. 20. Hatoh, S., Shiratori, I., and Takamiya, A., J . Biochem. (Tokyo) 51 (1962) 32. 21. Katoh, S., and Takamiya, A., Natl. Acad. Sci., Natl. Res. Council, Publ. 1145 (1963) 262. 22. De Kouchkovsky, Y., and Fork, D. C., Proc. A'atl. Acad. Sci. U.S. 52 (1964) 232. 23. Fork, D. C., and Urbach, W., Proc. Natl. Acad. Sci. U .S. 53 (1965) 1307. 24. Forti, G., BertolB, M. L., and Zanetti, G., Biociienz. Biophys. Acta, 109 (1965) 33. 25. Rlassey, V., and Swoboda, B. E. P., Biochem. Z. 338 (1963) 474. 26. Jenkins, J., Biochem. Prep. 9 (1962) 83. 27. Rwingle, S. M., and Tiselius, A,, Biochenr. J . 48 (1951) 171_. -.

28a. Massey, V., Biochim. Biophys. Acta, 37 (1960) 310. 2Sb. Foust, G., and blassey, V., Personal communication. 29. Avron, M. A., and Jagendorf, A. T., Arch. Biochem. Biophys. 70 (1957) 17. 30. Massey, V., and Palmer, G., Biochemistry 5, (1966) 3181. 31. Moore, S., Spackman, D. H., and Stein, W. H., Anal. Chem. 30 (1958) 1185. 32. Spackman, D. H., Stein, W. H., and Moore, S., Anal. Chem. 30 (1958) 1190. 33. Schram. E., Moore, S., and Bigwood, E. J., Biochem. J . 57 (1954) 33.

472

G. FORTI and E. STURANI:NADPH-Cytochrome f Reductase of Chloroplasts

34. Beaven, G. M., and Holiday, E. R., Advan. Protein Chem. 7 (1952) 319. 35. Davenport, H. E., and Hill, R., Proc. Roy. SOC.,London Ser. B. 139 (1952) 327. 36. Olson, J. M., and Smillie, R. M., Natl. Acad. Sci., Natl. Res. Council, Publ. 1145 (1963) 56. 37. Alberty, R. A., Advan. Enzymol. 17 (1956) 1 . 38. Lazzarini, R. A., and San Pietro, A., Biochim. Biophys. Acta, 62 (1962) 417. 39. Tagawa, K., and Arnon, D. I., Nature, 195 (1962) 537. 40. Foust, G., and Massey, v., Federation Proc. 26 (1967) 732. 41. Shin, M., and Amon, D. I., J . Biol. Chem. 240 (1965) 1405.


42. Massey, V., Personal communication. 43. Massey, V., and Gibson, Q. H., Proc. Intern. Congr. Biochem. 5th 5 (1963) 157. 44. Massey, V., and Williams, C. H. Jr., J. B i d . Ghem. 240 (1965) 4470. 45. Kalckar, H. M., J. Biol. Chem. 167 (1947) 461. 46. Teale, F. W. J., Biochem. J. 76 (1960) 381.

G. Forti and E. Sturani Laboratorio di Fisiologia Vegetale Istituto di Scienze Botaniche dell’Universit& Via Giuseppe Colombo 60, Milano 443, Italy