Mechanism of Enhanced Superoxide Production in the Cytochrome b₆f Complex of Oxygenic Photosynthesis

Abstract

The specific rate of superoxide (O₂∸) production in purified active crystallizable cytochrome b₆f complex, normalized to the rate of electron transport, has been found to be an order of magnitude greater than that measured in isolated yeast respiratory bc₁ complex. The biochemical and structural basis for the enhanced production of O₂∸ in the cytochrome b₆f compared to the bc₁ complex is discussed. The larger rate of superoxide production in the b₆f complex could be a consequence of an increased residence time of plastosemiquinone/plastoquinol in its binding niche near the Rieske protein iron-sulfur cluster, resulting from (i) occlusion of the quinone portal by the phytyl chain of the unique bound chlorophyll, (ii) an altered environment of the proton-accepting glutamate believed to be a proton acceptor from semiquinone, or (iii) a more negative redox potential of the heme b_p on the electrochemically positive side of the complex. The enhanced rate of superoxide production in the b₆f complex is physiologically significant as chloroplast-generated ROS functions in the regulation of excess excitation energy, is a source of oxidative damage inflicted during photosynthetic reactions, and is a major source of ROS in plant cells. Altered levels of ROS production are believed to convey redox signaling from the organelle to the cytosol and nucleus.

Introduction

The protein subunits and prosthetic groups in the structure from the cyanobacterium, M. laminosus ⁽¹⁾ are shown (Fig. 1). The role of the cytochrome bc₁ complex in O₂∸ generation by the mitochondrial respiratory electron transport pathway has been described and reviewed ^(2-8). By analogy with the bc₁ complex, the site of production of superoxide in the cytochrome b₆f complex is shown (Fig. 2), with an emphasis on the electrochemically positive (p, lumen)-side plastoquinone/ol binding site which, by analogy with the mechanism proposed for the bc₁ complex, is close to the site of oxygen reduction by plastosemiquinone and resulting superoxide formation. The production of reactive oxygen species (ROS) in chloroplasts not only underlies oxidative damage inflicted in photosynthetic electron transport ⁽⁹⁾, but also functions in redox signaling from the organelle to the cytosol and nucleus ⁽¹⁰⁾, and in the regulation and dissipation of excess excitation energy ⁽¹¹⁾. Although O₂∸ production by the cytochrome b₆f complex of oxygenic photosynthesis has been detected by EPR spectroscopy ⁽¹²⁾, details on rates and constraints have not yet been determined. The present studies provide quantitative information on the level of O₂∸ generated in electron transport through the b₆f complex that mediates these signaling processes.

Figure 1.

The dimeric cytochrome b₆f complex (PDB ID 4H44). Prosthetic groups in the cytochrome b₆f complex. Hemes b_p (red), b_n (red) and c_n (black) are redox-active prosthetic groups located within the trans-membrane domain, and constitute the low-potential chain. Heme f (black) and [2Fe-2S] cluster (orange/yellow spheres) form the high-potential chain, and are associated with the p-side extrinsic domains of cytochrome f and ISP, respectively. A chlorophyll-a (green) and a β-carotene (yellow) are also associated with the complex. Polypeptides are shown as ribbons. Color code: cytochrome b₆ (cyt b₆), wheat; subunit IV (subIV), orange; cytochrome f (cyt f), cyan; iron-sulfur protein (ISP), yellow; PetL, red; PetM, green; PetG, blue; PetN, gray.

Figure 2.

Trans-membrane electron transfer and Q-cycle mechanism in a schematic of the cytochrome b₆f complex, indicating the branching of electrons from the anionic semiquinone reductant, PQ_p∸ and heme b_p to O₂, the latter reactions responsible for formation of superoxide, O₂∸.

O₂∸ can be formed through a one electron reduction of the oxygen molecule, with a midpoint redox potential -0.14 V in the aqueous phase ^{(13, 14)}.

$${\mathrm{O}}_2+{\mathrm{e}}^{-}\to {\mathrm{O}}_2\dot{-}$$ (i)

Ubisemiquinone, UQ_p∸, formed on the electrochemically positive, p-side, of the complex (reaction iia) is a reductant for oxygen proximal to the bc₁ complex in mitochondrial and photosynthetic bacterial membranes. Because ubisemiquinol is a reductant of the low potential heme b_p in the bc₁ complex ⁽¹⁵⁾, it has been inferred that the ubisemiquinone formed in the bc₁ complex through quinol oxidation by the high potential segment of the electron transport chain has a sufficiently reducing potential to form superoxide, O₂∸ (reaction iia). Based on the similarity of the crystal structures of the protein core ⁽¹⁶⁾, and of the midpoint redox potentials, E_m7 = +80 mV (plastoquinone/quinol, ⁽¹⁷⁾ and +60 mV, ubiquinone/quinol ⁽¹⁸⁾), it is inferred that plastosemiquinone, PQ_p∸ can serve as the reductant for generation of superoxide in the b₆f complex (Fig. 2B, reaction iib):

$${\text{UQ}}_{\mathrm{p}}\dot{-}+{\mathrm{O}}_2\to \text{UQ}(\text{ox})+{\mathrm{O}}_2\dot{-}$$ (iia)

$${\text{PQ}}_{\mathrm{p}}\dot{-}+{\mathrm{O}}_2\to \text{PQ}(\text{ox})+{\mathrm{O}}_2\dot{-}$$ (iib)

The p-side semiquinone, Q_p∸, is generated through the one electron oxidation of ubiquinol or plastoquinol, QH₂, by the iron-sulfur [FeS] cluster of the high potential Rieske iron-sulfur protein subunit of the complex:

$${\text{QH}}_2+[\text{FeS}](\text{ox})\to {{\text{QH}}_{\mathrm{p}}}^{\cdot }+[\text{FeS}](\text{red})+{\mathrm{H}}^{+}$$ (iii)

QH_p∸ generated in reaction (iii) transfers a proton to the p-side aqueous phase through an intra-protein pathway in the cytochrome bc₁ complex ⁽¹⁹⁾. A similar p-side proton release pathway, which forms the p-side anionic plastosemiquinone, PQ_p∸, has recently been defined by crystallographic analysis of the cytochrome b₆f complex ⁽²⁰⁾.

In a Q-cycle mechanism ^{(15, 21-24)}, an electron is transferred from the deprotonated Q_p∸, to the p-side heme b_p (reaction iv), and then across the hydrophobic domain of the membrane through the n-side heme, b_n to reduce quinone specifically bound on the ‘n’ side of the complex.

$${\mathrm{Q}}_{\mathrm{p}}\dot{-}+b_{\mathrm{p}}(\text{ox})\to b_{\mathrm{p}}(\text{red})+{\mathrm{Q}}_{\mathrm{p}}(\text{ox})$$ (iv)

Reduced heme b_p may be an alternative source of electrons for the reduction of O₂ to O₂∸ ⁽²⁵⁾:

$$b_{\mathrm{p}}(\text{red})+{\mathrm{O}}_2\to b_{\mathrm{p}}(\text{ox})+{\mathrm{O}}_2\dot{-}$$ (v)

It has been inferred ^{(2, 8)} that superoxide production in the cytochrome bc₁ complex occurs as a by-pass of step iv above of the Q-cycle by reactions iia, b or v, the latter reaction having been suggested as dominant ^{(4, 25)}. The participation of reactions iib and v in super-oxide generation in the cytochrome b₆f complex is shown in Q-cycle scheme (Fig. 2). The n-side ubiquinone reduction in the bc₁ complex is blocked by the quinone analogue inhibitor antimycin A, which occupies the n-side quinone binding site with high affinity ⁽²⁶⁾. A consequence of this inhibition of the trans-membrane electron transfer chain would be accumulation of the p-side semiquinone electron donor, Q_p∸ (reaction iv). The probability of electron transfer through the other branch of the UQ_p∸ oxidation pathway that forms superoxide (reaction iia) would then be increased. Thus, the specific n-side quinone analogue inhibitor, antimycin A, causes a large increase in the specific rate of formation of O₂∸, relative to the electron transport rate, by the bc₁ complex of yeast or bovine mitochondria ^{(4, 7, 8, 27, 28)}, and the photosynthetic bacterium, Rb. sphaeroides ⁽²⁹⁾. No n-side inhibitor comparable in efficacy to that of antimycin. has been found for the b₆f complex, because the unique heme c_n occupies the quinone binding site homologous to that of antimycin ⁽³⁰⁾, which results in an altered binding-interaction of an n-side quinone analogue inhibitor ⁽³¹⁾. Although the quinone analogue inhibitor NQNO, binds specifically to the n-side heme c_n ⁽¹⁾, and inhibits the oxidation of heme b_n ^(32-34), unlike the inhibition of the respiratory chain resulting from the action of antimycin, the extent of inhibition of linear electron transport, by NQNO, is small ⁽³²⁾.

It has been proposed that O₂∸ production occurs in the cytochrome b₆f complex ⁽²⁾ by a mechanism (reaction iib) similar to the alternative pathway for the bc₁ complex (reaction iia) ^{(2, 3, 5, 8)}. However, experimental details on the rate of O₂∸ generation in the b₆f complex have not been described. The present study compares the specific rate of O₂∸ generation in cytochrome b₆f and bc₁ complexes, and proposes that the level of O₂∸ production by the b₆f and bc₁ complexes is dependent upon the residence time of reduced quinone species within the Q_p-portal.

Materials and Methods

Materials

ADPH and H₂O₂ were purchased from Anaspec (Fremont, CA), equine heart cytochrome c, HRP and SOD enzymes, decyl-ubiquinol and decyl-plastoquinol from Sigma-Aldrich (St. Louis, MO), and DDM/UDM from Anatrace (Maumee, OH), and other reagents from Mallinckrodt/Baker, Inc. (Phillipsburg, NJ).

Preparation of yeast mitochondrial cytochrome bc₁ complex

Cytochrome bc₁ complex from yeast was purified in the laboratory of B. L. Trumpower, as previously described ⁽²⁸⁾.

Preparation of thylakoid membranes and purification of cytochrome b₆f complex

Cyanobacterial thylakoid membranes were prepared and b₆f complex purified as described ⁽³⁵⁾, as were spinach chloroplast thylakoid membranes and cytochrome b₆f complex ^{(35, 36)}. Membranes were re-suspended (chlorophyll a concentration, 2 mg/ml) in TNE (30 mM Tris-HCl, pH 7.5, 50 mM NaCl, 1 mM EDTA, 0.3 M sucrose, with protease inhibitors, benzamidine (2 mM) and ε-amino-caproic acid (2 mM).

Purification of plastocyanin

Plastocyanin from Nostoc ⁽³⁶⁾ and spinach ⁽³⁷⁾ was purified as described and, for the latter, a modified procedure included size exclusion and anion-exchange chromatography.

Activities for ubiquinol-cytochrome c and plasto-quinol-plastocyanin, oxido-reductase were assayed, respectively, with purified bc₁ and b₆f complex. For the bc₁ complex, the assay mixture contained 50 μM equine heart cyt c, 50 mM MOPS, pH 6.9, 0.4 mM DDM, 1 mM KCN, and 0.05 mM decyl-ubiquinol. Reduction of cyt c, initiated by addition of 5 or 45 nM cytochrome bc₁ for uninhibited and inhibited activity, respectively, was monitored through the absorbance change at 551 nm and the activity was based on an extinction coefficient of 2.1 × 10⁽⁴⁾ M^-1cm^{-1 (38)}. For the plastoquinol-plastocyanin oxido-reductase activity of b₆f complex or thylakoid membranes, the assay mixture contained plastocyanin (10^-2 M) from Nostoc or spinach, 50 mM Tris-HCl, pH 7.5, 1 mM UDM, and 20 μM decyl-plastoquinol. Absorbance changes were assayed on a Cary 4000 spectrophotometer. The reaction was initiated by addition of 5 nM cyt b₆f complex or 3 μM Chl-a equivalent of thylakoid membranes, and activity monitored as the change of plastocyanin absorbance at 597 nm (ε_mM = 4.9 mM^-1cm^{-1 (39)}). Inhibitors were added as ethanol stocks (less than 0.5% of total reaction volume). Initial electron transport rates were determined using the SLOPE function in the support software of a Cary 4000 spectrophotometer, and calculated as moles electrons transferred per mol bc complex sec^-1, or per 10⁽³⁾ mol Chl-a equivalent of thylakoid membranes.

Superoxide production assay

Superoxide released from cytochrome bc complexes was measured fluorometrically via formation of H₂O₂. In the presence of saturating levels of SOD, superoxide released from cytochrome bc complexes is converted to H₂O₂. ADPH reacts with H₂O₂ in a 1:1 stoichiometry, in the presence of HRP as a catalyst, to produce the fluorescent oxidation product, resorufin. The reaction mixture contained: (i) for yeast bc₁ complex- 50 mM MOPS, pH 6.9, 0.4 mM DDM, 50 μM ADHP, 1 U/mL HRP, equine heart cyt c 50 μM, 50 μM decyl-ubiquinol; and for b₆f complex- 50 mM MOPS, pH 6.9, 1 mM UDM, 40 μM ADPH, 20 μM decyl-plastoquinol, 1 U/mL HRP, 300 U/ml SOD, plastocyanin (10 mM) from Nostoc or spinach, or 25 μM cyt c from equine heart. Inhibitors were added as ethanol stocks (less than 0.5% of total reaction volume). The reaction was started by adding bc₁ complex to a final concentration of 5 or 45 nM for measurement of uninhibited and inhibited activity, respectively, or b₆f complex to 5 nM (3 μM Chl-a equivalent of thylakoid membranes). The fluorescence indicator of superoxide production was measured by excitation and emission at 530 nm (2 nm band width) and 590 nm (1 nm band width) respectively, utilizing a FluoroMax-3 fluorimeter (Horiba Jobin Yvon Inc., Edison, NJ), and calibrated using H₂O₂ standards. The initial rate of superoxide production was estimated using the LINEST function of MS Office Excel software and calculated as moles superoxide generated·sec^-1 mol^-1 bc complex, or per 10⁽³⁾ mol Chl-a equivalent of thylakoid membranes.

Sequence alignment was performed with Clustal-Omega using default parameters ^{(40, 41)}. Cytochrome b₆ polypeptide sequences were obtained from the NCBI database, with accession numbers: Synechocystis sp. PCC 6803, BAA10149.1; Synechococcus elongatus PCC 6301, BAD79961.1; Nostoc sp. PCC 7120, BAB75120.1; Mastigocladus laminosus, AAR26242.1; Chlamydomonas reinhardtii, CAA44690.1; Arabidopsis thaliana, NP_051088.1; Spinacea oleracea, CAA30128.1.

Figure 4A, B was generated in Pymol (www.pymol.org), from structural superposition of cytochrome bc₁ complex crystal structures (PDB IDs 1NTZ ⁽⁴²⁾ and 3CX5 ⁽⁴³⁾), and cytochrome b₆f complex crystal structures (PDB IDs 4H44 and 4H13 ⁽²⁰⁾).

Figure 4.

Role of conserved Glu (Pro-Glu-Trp-Tyr) in semiquinone deprotonation in cyochrome bc complexes. (A) In the bc₁ complex, the Glu271 residue (green/red stcks, labeled in blue) of the PEWY sequence is found to occupy two distinct locations-quinone proximal (PDB ID 3CX5) and heme b_p-proximal (PDB ID 1NTZ). In the heme b_p-proximal position, Glu271 inteacts with Tyr131 (cytochrome b polypeptide). The figure was generated by superposition of cytochrome bc₁ crystal structures (PDB IDs 1NTZ and 3CX5). The STG molecule and the quinone/STG-proximal Glu271 orientation were obtained from PDB ID 3CX5. (B) In the cytochrome b₆f complex crystal structure (PDB ID 4H44), Glu78 (green/red sticks, labeled in blue) of subunitIV, homologous to Glu271 in the bc₁ complex, is located in the heme b_p-proximal orientation. In this position, Glu78 interacts with Arg87 of cytochrome b₆ (b₆f), through a distance of 2.8 Å. Arg87 (b₆f) is replaced by Ala84 (cyan sticks) in the bc₁ complex. Arg83 of cytochrome b₆ (b₆f) and Arg80 of cytochrome b (bc₁) are conserved in their location. The quinone analog inhibitor TDS has been inserted into the figure from PDB ID 4H13 to mark the Q_p-site. The transmembrane helices (a-g) are labeled. The polypeptides are shown as ribbons. (C) Multiple sequence alignment of the cytochrome b₆ subunit of the cytochrome b₆f complex from prokaryotic cyanobacteria (Synechocystis PCC 6803, Synechococcus elongates PCC 6301, Nostoc PCC 7120, M. laminosus), a eukaryotic alga (C. reinhardtii), and higher plants (Arabidopsis thaliana and Spinacea oleracea). Arg83, Arg87 (trans-membrane helix B) and Tyr136 (trans-membrane helix C) are conserved in cytochrome b₆ polypeptide.

Results

Superoxide production in the cytochrome bc₁ complex

The rate of superoxide (O₂∸) production in the yeast mitochondrial bc₁ complex, using decyl-ubiquinol as the electron donor, is 0.1-0.2 % of the electron transport rate (Table 1), a value similar to results in the literature obtained with bc₁ complex from yeast mitochondria ^{(5, 7, 8, 44)}. This is approximately the same rate, essentially a background rate, obtained in the presence of the p-side quinone analogue inhibitor DBMIB (Table 1), which prevents quinol oxidation by the [2Fe-2S] cluster ⁽⁴⁵⁾. Higher rates have been reported for mitochondrial bc₁ complexes, on the basis of difference in measurements with and without superoxide dismutase to assay O₂∸ production ⁽⁴⁴⁾. O₂∸ generation has also been measured with bc₁ complex obtained from the purple photosynthetic bacterium, Rb. sphaeroides ^{(29, 46)}, for which the rate of O₂∸ production relative to the electron transport rate was not reported. In the present experiments with the yeast bc₁ complex, the specific rate of O₂∸ production, normalized to the electron transport rate, is, however, increased greatly (> 40-fold) in the presence of antimycin A (Table 1), in qualitative agreement with data obtained previously for the bc₁ complex ^{(4, 5, 8, 27-29)}. The electron transport rate of isolated b₆f complex or membranes from spinach and M. laminosus, and O₂∸ generation, was measured, respectively, by the rate of reduction of plastocyanin, and by the rate of change of fluorescence emission intensity from the resorufin product of ADHP and H₂O₂ (Materials and Methods). Rates of electron transport (Fig. 3A) and O₂∸ production (Fig. 3B) obtained with yeast bc₁ complex and b₆f complex from spinach thylakoids and M. laminosus are shown, from which electron transport rates were derived for the yeast bc₁ complex (Table 1), purified b₆f complex isolated from spinach thylakoid membranes (Table 2), isolated M. laminosus cyanobacterial b₆f complex (Table 3), and membranes of M. laminosus (Table 4). The rate of O₂∸ production by electron transfer from the decyl-plastoquinol electron donor to b₆f complex (Table 2) is considerably higher, a factor of 10-20, than measured for the mitochondrial bc₁ complex (Table 1). Similar results were obtained with b₆f complex purified from the cyanobacterium, M. laminosus, using cyanobacterial plastocyanin (Table 3). Using intact membranes from M. laminosus (Table 4), the specific rate of superoxide production by the b₆f complex is larger that measured in isolated bc₁ complex by approximately an order of magnitude but, for reasons not understood, this rate is somewhat smaller that measured with isolated bc₁ complex (Table. 1).

Table 1. Superoxide Production by Cytochrome bc₁ Complex from Yeast Mitochondria.

	Electron transfer rate, Cyt bc-1 · s-1	Superoxide production rate Cyt bc1-1 · s-1	Superoxide production (% electron transfer rate)
Control, no inhibitor 1	151 ± 6.1 (n=5)	0.21 ± 0.13 (n=8)	0.14 ± 0.06
Antimycin A, 20 μM	1.2 ± 0.1 (n=4)	0.13 ± 0.03 (n=6)	10.6 ± 2.3
DBMIB, 2 μM	28.3 ± 3.1 (n=4)	0.04 ± 0.01 (n=3)	0.14 ± 0.02

¹for reactions in the absence and presence of inhibitor, bc₁ complex was used at 5 nM and 45 nM, respectively, with cytochrome c as electron acceptor; n, number of trials.

Figure 3.

Representative original data traces for determination of rates of (A, upper traces) superoxide generation and (B, lower traces) electron transfer in isolated bc₁ and b₆f complexes. The electron transfer and superoxide generation activities were measured as described in Materials and Methods using cytochrome c as the electron acceptor for the yeast cyt bc₁, cyt b₆f from M. laminosus, and cyanobacterial plastocyanin for the b₆f complex from spinach. Traces averaged from two to four measurements are shown. The background fluorescence change measured for the reaction mixture before the addition of the enzyme was subtracted from the superoxide generation activity data (B, lower traces). The arrow indicates addition of 5 nM and 45 nM bc₁ complex, respectively, for reactions in the absence and presence of inhibitor, and 5 nM b₆f complex.

Table 2. Superoxide Production by Isolated Spinach Cytochrome b₆f Complex; Electron Acceptor, Plastocyanin.

	Electron transfer rate, Cyt b6f-1 · s-1	Superoxide production rate Cyt b6f-1 · s-1	Superoxide production (% electron transfer rate)
Control, no inhibitor	245 ± 25 (n=10)	4.5 ± 0.6 (n=10)	1.9 ± 0.3
NQNO, 5 μM	196 ± 7 (n=3)	3.5 ± 0.4 (n=3)	1.8 ± 0.2
DBMIB, 4 μM	n.d. (n=4)	n.d. (n=3)	-

n, number of trials; n.d., rate too small to allow accurate determination.

Table 3. Superoxide Production by Isolated Cytochrome b₆f Complex from the Cyanobacterium, M. laminosus; Electron Acceptor, Cyanobacterial Plastocyanin.

	Electron transfer rate, Cyt b6f-1 · s-1	Superoxide production rate, Cyt b6f-1 · s-1	Superoxide production rate (% electron transfer rate)
Control, no inhibitor	292 (n=2)	6.3 (n=2)	2.2 (n=2)
NQNO, 5 μM	174 (n=1)	2.3 (n=1)	2.2 (n=1)
DBMIB, 2 μM	6.5 ± 1.0 (n=3)	n.d. (n=6)	-
DBMIB, 5 μM	0.5 ± 0.7 (n=3)	n.d. (n=4)	-

n, number of trials; n .d., not determined.

Table 4. Superoxide Production by Cytochromeb₆f Complex in Thylakoid Membranes from the Cyanobacterium, M. laminosus; Electron Acceptor, Cytochrome c.

	Electron transfer rate, 103 chl a-1 · s-1	Superoxide production rate 103 chl a-1 · s-1	Superoxide production rate (% electron transfer rate)
Control, no inhibitor	201 ± 3 (n=3)	2.6 ± 0.6 (n=4)	1.3 ± 0.3
NQNO, 5 μM	151 ± 8 (n=3)	1.5 ± 0.1 (n=3)	0.9 ± 0.1
DBMIB, 2 μM	11 ± 2 (n=3)	n.d. (n=3)	-
DBMIB, 5 μM	0.2 ± 0.4 (n=3)	n.d. (n=3)	-

n, number of trials; n .d., not determined.

Properties of inhibitors

(a) NQNO is an n-side quinone (Q_n) ligand of heme c_n ⁽¹⁾, and it has been shown to inhibit oxidation of heme b_n ^(32-34). NQNO binding to the b₆f complex in thylakoid membranes increases the amplitude of the flash-induced heme b_n reduction by a factor of 2-3 without significant inhibition of linear electron transport and oxygen evolution. In the present studies, concentrations of NQNO that affect the amplitude of heme b_n reduction only partly inhibit linear electron transport, and NQNO does not increase O₂∸ production as does addition of antimycin A to bc₁ complex (Tables 2-4 compared to Table 1). The difference in binding of n-side quinone analog inhibitors in the bc₁ and b₆f complexes resides in the multiple interactions with the amino acid environment that stabilize ligand-binding at the Q_n site of the bc₁ complex, whereas in the b₆f complex where heme c_n occludes access to heme b_n, the Q_n site protrudes into the inter-monomer cavity, with relatively few stabilizing interactions from coordinating amino acid residues. As a result, binding of a ligand such as NQNO at the Q_n-site is expected to be substantially weaker in the cytochrome b₆f complex than in the bc₁ complex. (b) DBMIB. Halogenated quinone analogs and in particular the dibromo-derivative DBMIB ⁽⁴⁷⁾ have been described as potent inhibitors of the chloroplast cytochrome b₆f complex ⁽⁴⁵⁾, and a less efficient inhibitor of the bc₁ complex from mitochondria ⁽⁴⁸⁾. DBMIB has been shown to bind at a position distal to the iron-sulfur binding site and also at this site ⁽⁴⁵⁾. In the present study, 2 μM DBMIB partially inhibits linear electron transport, and O₂∸ production of bc₁ complex to approximately 20% of the uninhibited rate (Table 1). Efficient inhibition is observed for cytochrome b₆f complex with rates of electron transport varying from undetectable to a few percent of the control (Tables 2-4), and no detectable O₂∸ production above the background signal.

Discussion

1. Origin of Elevated Superoxide Production in Cytochrome b₆f Complex

(A) More favorable redox potential of heme b_p in b₆f complex

(i) There is a large variation in the literature of midpoint redox potentials for heme b_p in the isolated complex: (i) -90 mV, pH 7 ⁽⁴⁹⁾; (ii) -172 mV, pH 6.5 ⁽⁵⁰⁾; (iii) -80 mV, pH 6 ⁽⁵¹⁾; -158 mV, pH 8.0 ⁽⁵²⁾. For the bc₁ complex, reported mid-point potentials of heme b_p are: approximately -30 mV in yeast and mammals ^{(53, 54)}, -90/+50 mV in purple bacteria ^{(55, 56)}, and 0 mV ⁽⁵⁷⁾. It is possible, but not determined by existing data, that a difference in redox potentials, e.g., a more negative potential of heme b_p in the b₆f complex, could contribute to the difference in efficiency of superoxide formation. However, the spread of data in the literature on the heme b_p redox potentials is presently too disparate to allow this explanation.

(B) Longer residence time

The higher rate of superoxide production in the b₆f compared to the bc₁ complex could arise from a longer residence time of the semiquinone in its binding niche (“Q_p pocket”) proximal to the iron-sulfur cluster of the Rieske iron-sulfur protein. The 10-20 fold higher rate of O₂∸ production by the cytochrome b₆f complex compared to that of the bc₁ complex (Tables 2-4 vs. Table 1) implies that the rate constant in the b₆f complex for reactions (iib), (iii) or (v) above is increased relative to that for (iv). The source of this difference can be the PQ_p∸ generated in the b₆f complex having a longer life-time in the Q_p pocket relative to UQ_p∸ in the bc₁ complex. An identifiable structure-based cause of this longer residence time could be the partial occlusion of the Q_p-portal for quinone entry/exit by the phytyl chain of the unique chlorophyll-a molecule embedded in each monomer of the complex ⁽³¹⁾. This occlusion does not affect the overall rate of electron transfer through the b₆f complex, as the isolated b₆f complex supports an electron transfer rate of ∼200 electrons monomer^-1 sec^-1 at room temperature ^(58-60), which is comparable to the activity of the bc₁ complex measured in ⁽⁶¹⁾, and larger in the present experiments, than the activity of isolated bc₁ complex (Tables 2, 3 vs. Table 1). Therefore, quinol passage through the chlorophyll-obstructed portal is not the rate-limiting step in the photosynthetic linear electron transport chain.

(C) Role of the Glutamate in the Conserved PEWY Sequence in Semiquinone Formation

The neutral semiquinone species bound within the Q_p-site of cytochrome bc complexes undergoes deprotonation to form an anionic semiquinone (formula iii), which can be a proton donor to the low potential chain. It has been suggested that the Glu residue (Glu78 in subunit IV of cytochrome b₆f complex; Glu271 or Glu272 in the cytochrome b polypeptide in the bc₁ complex) of the conserved Pro-Glu-Trp-Tyr (PEWY) motif located on the p-side ef-loop is involved in deprotonation of the neutral semiquinone to the anionic semiquinone ⁽⁶²⁾ (Fig. 4A, B), which acts as the electron donor to heme b_p (formula iib). Motion of the Glu271 residue in the cytochrome bc₁ complex during proton-transfer has been demonstrated ^{(19, 63-66)}. The Glu residue undergoes a rotation from a proton-extracting Q_p-niche proximal position, to a proton-releasing heme b_p-proximal position, where it interacts with the Tyr131 residue of cytochrome b (Fig. 4A). Hence, motion of the Glu residue constitutes an important step in the transfer of semiquinone from the Q_p-site. The kinetics of these reactions have recently been described ⁽⁶⁷⁾.

In the cytochrome b₆f complex, electron density for the Glu78 residue side chain, homologous to the Glu271 residue in the bc₁ complex, was assigned recently in the 2.7 Å resolution crystal structure (PDB ID 4H44) ⁽²⁰⁾. The Glu78 residue side-chain is found in a heme b_p-proximal position, in which it interacts with a conserved Arg87 of the cytochrome b₆ polypeptide through a short distance of 2.8 Å (Fig. 4B). In this position, the Glu78 side chain does not interact with the quinone analog inhibitor TDS (inserted into Q_p-site from PDB ID 4H13). The side chain of the conserved residue Tyr136 (cytochrome b₆ in the b₆f complex, trans-membrane helix C, Figs. 4B, C), which replaces Tyr131 of cytochrome b (bc₁ complex, Fig. 4A), does not interact with Glu78 in the heme b_p-proximal position. It has been previously reported that the conserved Arg87 of the trans-membrane helix B of the cytochrome b₆ polypeptide (b₆f complex, Fig. 4C) is replaced by a small uncharged residue, such as Ala84 (or Ala83), in cytochrome b (bc₁ complex, trans-membrane helix B) ⁽⁶⁸⁾. An analysis of the amino acid environment of Glu271 of cytochrome bc₁ shows that the Glu271 side chain is separated from the Ala84 residue by a distance of 7.3 Å (Fig. 4A). Hence, unlike the Glu78 residue in the b₆f complex that is inferred to participate in a relatively strong interaction with Arg87 in the heme b_p-proximal position, Glu271 in the bc₁ complex does not interact with a basic residue from the trans-membrane helix B in the heme b_p-proximal position. Thus, Glu271 may be relatively free to undergo motion in the cytochrome bc₁ complex, from a Q_p-niche proximal position to a heme b_p-proximal position, making proton extraction and translocation from the semiquinone a relatively efficient process. On the other hand, in the b₆f complex, motion of Glu78 is expected to be comparatively restricted due to the interaction with Arg87, thereby making proton translocation a less efficient process. As a consequence, the life-time of the neutral semiquinone is expected to be larger within the Q_p-site of the cytochrome b₆f complex.

2. Further consequences of increased semiquinone retention time

Photosynthetic state transitions depend on the redox poise of the thylakoid membrane quinone pool ^(69-71). The presence of reduced plastoquinone in the Q_p-site activates a LHCII kinase bound to the b₆f complex ^{(72, 73)}. Regarding the application of this signaling process to cyanobacteria, although the latter possess state transitions that involve mobility of the phycobilisomes, this does not involve kinase activation, but is dependent on redox equilibria involving plastoquinone ⁽¹¹⁾.

3. Chloroplasts as a major source of ROS in plant cells

High irradiance, and other stress conditions that affect photosynthetic electron transport rate and create hyperoxic conditions result in a transient increase in ROS in chloroplasts ^{(9, 74)}. The major sources of ROS generated in chloroplasts are PSII (singlet oxygen) and PSI (O₂∸). O₂∸ production by the cytochrome b₆f complex, as described in the present study, has been detected by EPR spectroscopy ⁽¹²⁾. Accumulated O₂∸ can also be metabolized to another more stable form of ROS, hydrogen peroxide ^{(11, 74)}.

The production of ROS in chloroplasts not only underlies oxidative damage inflicted during photosynthetic reactions ⁽⁷⁵⁾, but the redox state of photosynthetic electron transport components conveys information about environmental light conditions. Thus, chloroplasts can function as mediators of environmental signals ⁽⁷⁶⁾. Light and stress induced generation of ROS contributes to redox signaling inside the chloroplast and from the organelle to the cytosol and nucleus ^{(10, 77, 78)}. ROS function in chloroplast signaling was demonstrated in relation to regulation dissipation and avoidance of excess excitation energy mechanisms ⁽¹¹⁾. The ROS generated in chloroplasts act as a retrograde signal to the nucleus for regulation of plant responses to environmental stress and pathogen defense responses ^{(79, 80)}. Increases in H₂O₂ concentrations have been shown to be important for the induction of the ascorbate peroxidase gene, APX2, and for the expression of a number of genes involved in plant development and stress responses ⁽⁸¹⁾. There is evidence that in plant disease resistance responses, ROS produced in chloroplasts interplay with signal from other intracellular and extracellular ROS sources in the modulation of pathogen-induced hypersensitive response ^{(82, 83)}.

Abbreviations

ADPH

10-acetyl-3,7-dihydroxyphenoxazine

b_p and b_n

b-hemes on electrochemically positive and negative sides of b₆f complex

Chl-a

chlorophyll

DBMIB

2,5-dibromo-3-methyl-6-isopropyl-p-benzoquinone

DDM, UDM

n-dodecyl- or undecyl-β-D-maltopyranoside

Δμ˜_h⁺

trans-membrane proton electrochemical potential gradient

E_m7

midpoint oxidation-reduction potential at pH 7

ISP

iron-sulfur protein

HRP

horseradish peroxidase

MOPS

3-(N-morpholino) propanesulfonic acid

NQNO

2-n-nonyl-4-hydroxyquinoline-N-oxide

p, n

electrochemically positive, negative side of membrane and cytochrome bc complexes

PC

plastocyanin

PSI

II, photosystem I, II

Q

quinone

Q_p

p-side quinone binding site

PQ/PQH₂

plastoquinone/-ol

PQ_p∸

plastosemiquinone

ROS

reactive oxygen species

O₂∸

superoxide

SOD

superoxide dismutase

STG

stigmatellin

TDS

tridecyl-stigmatellin

UQ/UQH₂

ubiquinone/-ol

UQ∸

ubisemiquinone