Photoinduced intracomplex electron transfer between cytochrome c oxidase and TUPS-modified cytochrome c

Authors


A. Kotlyar, Department of Biochemistry, George S. Wise Faculty of Life Sciences, Tel Aviv University, Ramat Aviv, Israel. Fax: +972 3 640 6834, E-mail: sasha@hemi.tau.ac.il

Abstract

A novel method for initiating intramolecular electron transfer in cytochrome c oxidase is reported. The method is based upon photoreduction of cytochrome c labeled with thiouredopyrene-3,6,8-trisulfonate in complex with cytochrome oxidase. The thiouredopyrene-3,6,8-trisulfonate-labeled cytochrome c was prepared by incubating the thiol reactive form of the dye with yeast iso-1-cytochrome c, containing a single cysteine residue. Laser pulse excitation of a stoichiometrical complex between thiouredopyrene-3,6,8-trisulfonate-cytochrome c and bovine heart cytochrome oxidase at low ionic strength resulted in the reduction of cytochrome c by the excited form of thiouredopyrene-3,6,8-trisulfonate and subsequent intramolecular electron transfer from the reduced cytochrome c to cytochrome oxidase. The maximum efficiency by a single laser pulse resulted in the reduction of ≈ 17% of cytochrome a, and was achieved only at a 1 : 1 ratio of cytochrome c to cytochrome oxidase. At higher cytochrome c to cytochrome oxidase ratios the heme a reduction was strongly suppressed.

Abbreviations
IPTS

1-isothiocyanatopyrene-3,6,8-trisulfonate

Nbs

5-mercapto-2-nitrobenzoate

Nbs2

5,5′-dithiobis-(2-nitrobenzoic acid)

TUPS*

excited triplet state of thiouredopyrene-3,6,8-trisulfonate

TUPS

thiouredopyrene-3,6,8-trisulfonate

Cytochrome c oxidase catalyzes electron transfer from ferrocytochrome c to dioxygen and couples this process to proton translocation across the inner mitochondrial membrane, thus converting the free energy of the redox process to an electrochemical proton gradient [1]. The three-dimensional structure of the enzyme, as determined recently by X-ray diffraction [2], consists of four redox active centers: two hemes (heme a and a3), CuB, which is closely associated with heme a3, and a mixed-valence binuclear CuA center.

Estimation of electron transfer rates in nonphotosynthetic electron transferring proteins requires the use of photochemical techniques, which enable the initiation of electron transfer reactions within microseconds and submicroseconds. This has been achieved for cytochrome c oxidase by the flow-flash method of Greenwood & Gibson [3], or by using external photosensitive elements as electron donors [4–21]. Current experimental methods employ excited flavines [4,5], ruthenium complexes [6,7] or zinc-substituted heme proteins [8–10] as light-driven photo-reductants. Of these, the ruthenium complexes have proven to be most practical. Pulse irradiation of Ru(II) complexes results in the formation of a low potential excited state (Eo ≈ −720 mV), capable of reducing many different compounds of chemical and biological nature. This method has been utilized to study electron transfer in cytochrome c[6] and within complexes of cytochrome c with other proteins [14–19]. It was recently used successfully to initiate electron transfer in cytochrome c oxidase [17–19]. To initiate the electron transfer, specially designed Ru-cytochrome c derivatives were synthesized and complexed with cytochrome c oxidase in an analogous fashion to the unmodified cytochrome c[19]. Following photoactivation of Ru(II), the rate of electron transfer from ferrocytochrome c to heme a via CuA of the oxidized resting form of cytochrome c oxidase was measured. However, the efficiency of this method is rather low; < 1% of the total cytochrome c population can be perturbed, and the small magnitude of the resulting signal limits the time resolution of the monitoring system.

We have recently introduced a new method for studying electron transfer in nonphotosynthetic proteins with submicrosecond time resolution [20]. The initiation of electron transfer was achieved by photoexciting a thiouredopyrene-3,6,8-trisulfonate (TUPS) molecule covalently bound to cytochrome c. Single-photon excitation of TUPS yields the low potential triplet state of the dye with high quantum efficiency [20]. The high yield and long lifetime of the triplet state make TUPS an efficient initiator of electron transfer processes. In our previous work, the TUPS moiety was covalently attached to eight different lysine residues of cytochrome c[21]. Laser pulse photoexcitation of the dye resulted in fast (µs) and highly efficient reduction of heme c.

In this study, we prepared a new TUPS(Cys102)–cytochrome c derivative to measure electron transfer reactions in the complex of cytochrome c with cytochrome oxidase. The derivative was obtained by modifying the single cysteine residue of the yeast iso-1-cytochrome c with a thiol-specific TUPS reagent. The derivative was shown to interact with cytochrome oxidase in the same fashion as the unlabeled cytochrome c. The TUPS(Cys102)–cytochrome c derivative was used to measure the apparent rate constant of intracomplex electron transfer from cytochrome c to cytochrome oxidase. Between 15 and 20% of the heme a of cytochrome oxidase was reduced by a single laser pulse.

Materials and methods

Materials

Yeast iso-1-cytochrome c was obtained from Sigma. 1-Isothiocyanatopyrene-3,6,8-trisulfonate (IPTS) was purchased from Lambda Fluorescence (Graz, Austria).

Preparation of TUPS–cystamine

The thiol-specific TUPS derivative was prepared by incubating IPTS (20 m m final concentration) with 200 m m cystamine at pH 9.0 for 10 min at room temperature. The reaction of IPTS with the amino group of cystamine resulted in the formation of the TUPS–cystamine derivative. One milliliter of the mixture was loaded onto a Sephadex G-25 (medium) column (1.0 × 5.3 cm) equilibrated with 10 m m Hepes (pH 7.5). During the course of our studies, we observed that the TUPS dye had a high affinity for the Sephadex matrix in the presence of salts and could be eluted from the column with only deionized water. This property of the dye was used to separate the dye–cystamine derivative from the nonbound cystamine. The latter was washed off the column with 10 mL of buffer, while the TUPS–cystamine derivative was eluted from the column with water. Fractions containing TUPS–cystamine were used for SH-specific modification of yeast iso-1-cytochrome c.

Preparation of TUPS(Cys102)–cytochrome c derivative

Yeast iso-1-cytochrome c (0.5 m m) was reduced with 1 m m ascorbate and incubated with TUPS–cystamine (1.5 m m) in 150 m m Hepes (pH 8.5), containing 0.2 m KCl for 4 h at 25 °C. The mixture was passed through a Sephadex G-25 column, equilibrated with 5 m m Hepes (pH 8.0), in order to get rid of salts and nonbound TUPS–cystamine. The fractions that eluted in the void-volume were collected and loaded onto a CM-Sephadex column (1.6 × 10 cm) equilibrated with 10 m m Hepes (pH 8.0) containing 100 µm ascorbate. The TUPS–cytochrome c derivative was eluted by a linear KCl gradient from 0 to 0.5 m over 8 h. The main fraction, eluting at 100–150 m m KCl, contains the TUPS–cytochrome c derivative. Spectral analysis, carried out as described in Kotlyar et al. [21], of the fraction showed that it contained a single equivalent of TUPS. The unlabeled cytochrome c was eluted at ≈ 300 m m KCl. The labeled cytochrome c was concentrated by lyophilization. The lyophilized cytochrome c was dissolved in a small volume of water and passed through a Sephadex G-25 column equilibrated with 5 m m Hepes (pH 7.5). The labeled cytochrome c was stored at −20 °C.

The content of free SH-groups in the dye-containing fraction of cytochrome c was determined by monitoring spectrophotometrically the titration of this fraction with 5,5′-dithiobis-(2-nitrobenzoic acis) (Nbs2) at 450 nm using a milimolar extinction coefficient of 7.6 m m−1·cm−1. Formation of 5-mercapto-2-nitrobenzoate (Nbs) was followed at 450 nm instead of 412 nm (absorbance maximum of Nbs) to rule out absorbance contribution by the Soret band of cytochrome c. The reaction was initiated by addition of 1 m m Nbs2 to a 1-mL sample containing 10 µm cytochrome c in 50 m m phosphate buffer, pH 8.0.

Reactivity with cytochrome oxidase

Cytochrome oxidase was isolated from bovine hearts as described by Yoshikawa et al. [22]. The enzyme concentration was determined spectrophotometricaly using a milimolar extinction coefficient of 39.8 m m−1·cm−1 for the dithionite-reduced enzyme at 604 nm. Cytochrome c oxidase activity was followed spectrophotometricaly at 550 nm by monitoring the oxidation of ferrocytochrome c. The reaction mixture contained 20 m m potassium phosphate (pH 7.0), 0.5% Tween 80 and ferrocytochrome c. The reduced cytochrome c was prepared by incubating ferricytochrome c with ascorbate and subsequently removing the reductant by gel filtration. The enzymatic reaction was initiated by the addition of an appropriate amount of cytochrome oxidase. The initial rates were measured at ferrocytochrome c concentrations ranging from 0.5 to 20 µm. The dependence of the initial rate on the substrate concentration was analyzed by a Lineweaver-Burk plot to obtain the Km and Vmax values. The enzyme exhibited a maximum activity of 10 µmol per min per mg of protein.

Transient absorption kinetics

The sample (1.5 mL), containing 5 m m Mes (pH 6.0), TUPS–cytochrome c and cytochrome oxidase was placed in a 1-cm pathlength quartz cuvette. Glucose oxidase, catalase and glucose at final concentrations of 200 µg·mL−1, 20 µg·mL−1 and 20 m m, respectively, were added to the cuvette under constant Ar flow and the reaction mixture was incubated for an additional 10 min in order to remove oxygen from the solution. Excitation of the dye was initiated by the third harmonic frequency of a Nd : Yag laser (355 nm, 2 ns full-width at half-maximum, 3 mJ per pulse), which was focused on a spot with a surface area of 0.3 cm2. The redox state of cytochrome c was monitored at 550 and 556 nm (556 nm is the isosbestic point of oxidized and reduced cytochrome c), and the redox state of heme a was monitored at 604 and 585 nm. The probe beam was generated by a ‘Coherent’ CR-599 dye laser with rhodamine-110 (550 and 556 nm) and rodamine 6G (604 and 585 nm) dyes, and the beam crossed the pulse irradiated face of the cuvette perpendicular to the excitation beam. The transmitted probe beam was passed through a monochromator/photomultiplier assembly, and transients were stored and averaged by a Tektronix TDS 520A digital oscilloscope as described previously [21]. The response time of the detection system was 20 ns. The transients are the average of 20 pulses collected at a frequency of 0.02 Hz.

Results

Preparation of TUPS(Cys102)–cytochrome c derivative

The thiol–disulfide exchange reaction between the single cysteine residue of cytochrome c and the cystamine disulfide results in the formation of TUPS(Cys102)–cytochrome c derivative. At high TUPS–cystamine to cytochrome c ratio, the yield of modified cytochrome c is 40–50%. Labeled cytochrome c was separated from unlabeled on a CM-Sephadex column ( Fig. 1). The derivative, carrying a negatively charged TUPS molecule, was eluted first from the column, at a lower salt concentration than the unlabeled cytochrome c. Spectral analysis showed that the fraction that eluted first contained a single equivalent of TUPS. Titration of this fraction with Nbs2 showed no free SH-groups in the protein, confirming that the dye is linked to cytochrome c via its cysteine residue. A control reaction with 10 µm unmodified iso-1-cytochrome c resulted in the formation of 10 µm Nbs.

Figure 1.

Separation of the TUPS(Cys102)–cytochrome c derivative (first peak) from nonmodified cytochrome c (second peak). Yeast iso-1-cytochrome c (0.5 m m) was labeled with TUPS–cystamine as described in Materials and methods. The crude mixture containing the modified and the unlabeled cytochromes (10 mg) was chromatographed on a CM-Sephadex column (1.6 × 10 cm) equilibrated with 10 m m Hepes (pH 8.0), containing 100 µm ascorbate. Cytochromes were eluted from the column with a linear gradient of 0–500 m m KCl at a flow rate of 0.2 mL·min−1. Cytochrome elution was followed by the absorbance at 280 nm.

Photoinduced electron transfer kinetics in TUPS(Cys102)–cytochrome c

We have shown that photoexcitation of TUPS generates the triplet state of the dye (TUPS*) with high quantum efficiency [20]. The triplet, being a strong reductant, is capable of reducing the heme group of cytochrome c. Figure 2 depicts the 550–556 nm transient absorbance changes resulting from cytochrome c reduction by TUPS*. The single-exponential rise has an apparent rate constant, kapp = (7.0 ± 0.3) × 104 s−1. The apparent rate constant was independent of the concentration of the TUPS–cytochrome c derivative from 10 to 100 µm, indicating an intramolecular electron transfer mechanism. This reductive branch of the photochemical process generates two redox-active species: the reduced cytochrome c and the oxidized form of TUPS. Reoxidation of cytochrome c by the oxidized dye brings the system to its initial prepulsed state. The monophasic decay proceeds with an apparent rate constant of (3.1 ± 0.3) × 103 s−1, which was independent of the concentration of the TUPS–cytochrome c derivative. The redox cycle initiated by photoexcitation of TUPS can be repeated numerous times without a decrease in the signal amplitude. The efficiency is high, and between 15 and 20% of the cytochrome c molecules undergo intramolecular reduction in a single pulse.

Figure 2.

Transient kinetics of the TUPS(Cys102)–cytochrome c derivative. TUPS(Cys102)–cytochrome c derivative (10 µm) in a solution containing 5 m m Mes (pH 6.0) was treated with glucose oxidase (200 µg·mL−1), catalase (20 µg·mL−1) and glucose (20 m m) under constant Ar flow for 10 min in order to remove oxygen from the solution. The sample was irradiated by a 6-mJ pulse (Nd : Yag, third harmonic frequency) as described in Materials and methods. The absorbance changes were monitored at 550 and 556 nm. The transient accounts for the reduction and subsequent reoxidation of the protein heme group. The transient was fitted (continuous curve) by a sum of a single exponential rise having the apparent rate constant kapp = (7.0 ± 0.3) × 104 s−1 and a single exponential decay having an apparent rate constant kapp = (3.1 ± 0.3) × 103 s−1.

Activity of xytochrome c oxidase in complex with TUPS–cytochrome c

In order to be useful for elucidating the electron transfer dynamics of cytochrome oxidase, the TUPS–cytochrome c derivative must interact with cytochrome oxidase in the same fashion as the unlabeled cytochrome c. To test whether this was the case, the kinetic parameters (Km and Vmax) for the labeled cytochrome c were measured and compared with those of the native protein. A Lineweaver–Burk plot of the oxidation of TUPS(Cys102)–cytochrome c derivative ( Fig. 3) yielded a Km of 1.5 µm, which is identical within the accuracy of the analysis to the value of 2 µm estimated for the unlabeled protein. The data clearly indicate that modification of the cysteine residue of cytochrome c does not reduce the affinity of cytochrome oxidase for the protein.

Figure 3.

.Oxidation of TUPS(Cys102)–cytochrome c derivative by cytochrome oxidase. Lineweaver–Burk plot of the initial rates at different concentrations (0.5–10 µm) of unlabeled cytochrome c (●) and TUPS(Cys102)–cytochrome c derivative (○). Activity assays were carried out in 20 m m potassium phosphate (pH 7.0) and were followed spectrophotometrically at 550 nm by monitoring the oxidation of ferrocytochrome c (see Materials and methods). V is expressed in µmole of cytochrome c oxidized per min per mg of protein.

Photoinduced intracomplex electron transfer between TUPS(Cys102)–cytochrome c and cytochrome oxidase

When cytochrome oxidase was added to the solution at low ionic strength, the apparent rate of cytochrome c reoxidation increased as a result of electron transfer to cytochrome oxidase. The 550–556 nm transient in Fig. 4 shows the kinetics of cytochrome c photoreduction and reoxidation by cytochrome oxidase. The reduction of cytochrome c occurred with an apparent rate constant, kapp = (8.0 ± 0.4) × 104 s−1. The reoxidation of cytochrome c in the presence of the enzyme was monophasic and proceeded with a rate constant of (1.1 ± 0.3) × 104 s−1. The 604–585 nm transient shows the reduction of heme a ( Fig. 4) of cytochrome oxidase induced by photoexcitation of TUPS–cytochrome c. The apparent rates of heme a reduction and cytochrome c reoxidation were similar and independent of the concentration of the TUPS–cytochrome c and cytochrome oxidase over the range 10–30 µm, which is consistent with an intramolecular mechanism.

Figure 4.

Photoinduced electron transfer from TUPS(Cys102)–cytochrome c derivative to cytochrome oxidase. TUPS(Cys102)–cytochrome c derivative (10 µm) and cytochrome oxidase (15 µm) in a solution containing 5 m m Mes (pH 6.0) was treated with glucose oxidase (200 µg·mL−1), catalase (20 µg·mL−1) and glucose (20 m m) under constant Ar flow for 10 min. The sample was irradiated by a 6-mJ pulse (Nd : Yag third harmonic frequency) as described in Materials and methods. The kinetics were monitored at 550 and 556 nm (curve 1) and at 604 and 585 nm (curve 2). The 550–556 nm transient accounts for the reduction and subsequent reoxidation of cytochrome c. The 550–556 nm transient was fitted (continuous curve) by a sum of single exponential rise having the apparent rate constant kapp = (8.0 ± 0.4) × 104 s−1 and a single exponential decay having an apparent rate constant kapp = (1.1 ± 0.1) × 104 s−1. The 604–585 nm transient reflects the heme a reduction.

The efficiency of heme a reduction within the cytochrome oxidase/TUPS–cytochrome c complex depends strongly on the ratio between the two proteins. The 604–585 nm transients measured at different cytochrome oxidase to TUPS(Cys102)–cytochrome c ratios are shown in Fig. 5 . It is clear that the maximum reduction of heme a is achieved at equal concentrations of the cytochrome c derivative and cytochrome oxidase. At higher cytochrome c to cytochrome oxidase ratios, the reduction of the heme a is strongly suppressed. The decrease in the amplitude of the signal is not due to lower quantum efficiency at higher cytochrome c concentrations. When the ratio between cytochrome oxidase and cytochrome c was kept constant a linear increase in the signal amplitude with cytochrome c concentration in the range 10–30 µm was observed (data not shown).

Figure 5.

Dependence of the yield of cytochrome oxidase reduction on the ratio between cytochrome c and cytochrome oxidase. The transients were measured at 604–585 nm as described in Fig. 4 at 10 µm cytochrome oxidase and at the indicated TUPS(Cys102)–cytochrome c derivative concentrations.

Discussion

The TUPS(Cys102)–cytochrome c derivative used here has two advantages which contribute to its usefulness for driving electron transfer in cytochrome oxidase: the photoexcitation of TUPS results in efficient reduction of cytochrome c with between 15 and 20% of the cytochrome c molecules undergoing intramolecular reduction in a single pulse; and the derivative interacts with cytochrome oxidase in an identical manner to that of the unlabeled native cytochrome c, as reflected by the same affinity of cytochrome oxidase (Km) for both proteins. The high affinity of cytochrome oxidase for the labeled cytochrome c makes it possible to keep the proteins bound to one another under the conditions of the laser experiment and to study intramolecular electron transfer reactions within the complex of cytochrome oxidase and the modified cytochrome c.

Laser pulse excitation of the TUPS–cytochrome c complex with cytochrome oxidase results in efficient electron transfer to the enzyme. The reduction of heme a of cytochrome oxidase achieved within a single pulse exceeded 15%. The efficiency of cytochrome oxidase reduction in complex with the TUPS–cytochrome c derivative is much higher than that reported for reduction of the enzyme by Ru-modified cytochrome c derivatives [17,19].

The kinetics at 550–556 nm and 604–585 nm do not provide direct evidence for the involvement of CuA in the electron transfer pathway between cytochrome c and heme a, and the apparent rate constant of the cytochrome c oxidation is the same within experimental error to that of heme a reduction. However, previous studies [12,17,19] have provided convincing evidence that CuA is the initial electron acceptor of electrons from cytochrome c. Studies using Ru-labeled cytochrome c have shown that electron transfer from cytochrome c to the CuA occurs with a rate constant of 6.0 × 104 s−1[17], which is almost equal to the apparent rate constant for the cytochrome c reduction in the TUPS(Cys102)–cytochrome c derivative ( Fig. 2). The inability to resolve the reaction associated with CuA reduction, in this study, is due mainly to the rather slow electron transfer rate from the excited TUPS to the heme in the TUPS(Cys102)–cytochrome c derivative. Maximum reduction of heme a occurred at equimolar concentrations of cytochrome c and cytochrome oxidase suggesting that the functionally active electron transfer complex consists of one molecule of cytochrome c and one molecule of cytochrome oxidase. At ratios > 1, reduction was suppressed ( Fig. 5). These findings may be rationalized by assuming the existence of high- and low-affinity binding sites for cytochrome c on cytochrome oxidase having different effects on electron transfer activity. The high affinity site acts as an active electron transferring site from which electrons enter cytochrome oxidase. Binding of cytochrome c to the low-affinity site inhibits electron transfer from cytochrome c bound at the high-affinity site. When both sites are saturated with cytochrome c no efficient electron transfer from ferrocytochrome to cytochrome oxidase takes place. This assumption is supported by previous studies that have shown that there are two binding sites for horse heart [23] and yeast cytochrome c[24] on bovine heart cytochrome oxidase.

Acknowledgements

This work was supported by the Israel United States Binational Science Foundation (538/95).

Ancillary