- Email: [email protected]
Conformation-driven and semiquinone-gated proton-pump mechanism in the NADH-ubiquinone oxidoreductase (complex I)
FEBS 29814
FEBS Letters 579 (2005) 4555–4561
Hypothesis
Conformation-driven and semiquinone-gated proton-pump mechanism in the NADH-ubiquinone oxidoreductase (complex I) Tomoko Ohnishia,*, John C. Salernob a
Johnson Research Foundation, Department of Biochemistry and Biophysics, University of Pennsylvania, Philadelphia, PA 19104, United States b Department of Biology, Rensselaer Polytechnic Institute, Troy, NY 12181, United States Received 16 June 2005; accepted 27 June 2005 Available online 26 July 2005 Edited by Peter Brzezinski
Abstract A novel mechanism for proton/electron transfer is proposed for NADH-quinone oxidoreductase (complex I) based on the following findings: (1) EPR signals of the protein-bound Þ fast-relaxing semiquinone anion radicals (abbreviated as QNf are observable only in the presence of proton-transmembrane are electrochemical potential; (2) Iron–sulfur cluster N2 and QNf directly spin-coupled; and (3) The projection of the interspin vector ˚ along the membrane normal [Yano, T., Dunham, extends only 5 A W.R. and Ohnishi, T. (2005) Biochemistry, 44, 1744–1754]. We propose that the proton pump is operated by redox-driven conformational changes of the quinone binding protein. In the input state, semiquinone is reduced to quinol, acquiring two protons from the N (matrix) side of the mitochondrial inner membrane and an electron from the low potential (NADH) side of the respiratory chain. A conformational change brings the protons into position for release at the P (inter-membrane space) side of the membrane via a proton-well. Concomitantly, an electron is donated to the quinone pool at the high potential side of the coupling site. The system then returns to the original state to repeat the cycle. This hypothesis provides a useful frame work for further investigation of the mechanism of proton translocation in complex I. 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Keywords: NADH-Q oxidoreductase; Proton-pumping mechanism; Semiquinone-gating
1. Introduction NADH-ubiquinone oxidoreductase (complex I), which is located at the entry point of the mitochondrial respiratory chain, * Corresponding author. Fax: +215 573 3748. E-mail address: [email protected] (T. Ohnishi).
Abbreviations: Complex I, NADH-ubiquinone oxidoreductase; Dw, membrane potential; Dlþ H , proton-electrochemical gradient; Em, redoxmidpoint potential; EPR, electron paramagnetic resonance; N2, iron– sulfur cluster N2; N-side, matrix side of the mitochondrial inner membrane; P-side, inner space side of the mitochondrial inner mem , protein-bound fast-relaxing brane; Q, ubiquinone; QH2, ubiquinol; QNf , protein-bound slow-relaxing ubiubisemiquinone (anion form); QNs semiquinone anion; Q-pool, ubiquinone pool in the mitochondrial inner membrane; SMP, submitochondrial particles; SQ, semiquinone; SQNf, protein-bound fast-relaxing semiquinone (including both anion and neutral forms)
catalyzes electron transfer from NADH to the ubiquinone (Q) pool [1–3], in the reaction: þ þ NADH þ Q þ Hþ þ 4Hþ N $ NAD þ QH2 þ 4HP
In this exergonic redox reaction, the transfer of 2 electrons from 1 mol of NADH to ubiquinone pool in the mitochondrial inner membrane (Q-pool) is coupled to the transfer of 4 protons from the negative (matrix) side (abbreviated by the suffix N) to the positive side (abbreviated by the suffix P) of the mitochondrial inner membrane. The consensus ratio for pumped protons/ electrons is 4H+/2e [4,5]. Currently, complex I is the only mitochondrial respiratory complex without available X-ray crystallographic information. The L-shaped low-resolution complex I structure is composed of nearly perpendicular hydrophilic promontory and hydrophobic membrane arms; the former ex˚ out of the matrix side surface level of the inner tends 160 A membrane (N-side) [6]. One non-covalently bound flavin mononucleotide and eight distinct iron–sulfur clusters [2 binuclear (N1a and N1b) and 6 tetranuclear clusters (N2, N3–N6a, N6b)] reside in the promontory region 1 [3,7,8], which transfer electrons from substrate NADH to N2 (highest redox-midpoint potential (Em) cluster), located close to the junction between the two arms. The membrane-associated region includes seven mitochondrially encoded transmembrane subunits, of which ND2, ND4, and ND5 are candidates for proton transport components. Friedrich and Weiss [9] proposed a modular evolution hypothesis for complex I based on sequence analysis. Hydrophilic 51, 24, and 75 kDa subunits are similar to NAD+-reducing hydrogenases [10]; hydrophobic membrane components are homologous to bacterial Na+/H+ antiporters [11]; and interfacial amphipathic PSST and 49 kDa subunits are similar to bacterial Ni–Fe hydrogenases [12]. Based on the observation of semiquinone electron paramagnetic resonance (EPR) signals differing in spin-relaxation rates [SQNf(fast) and SQNs(slow)] 2 and in rotenone sensitivity, two distinct ubiquinone (Q)-binding sites were assigned within the membrane component of complex I. Previously, both SQNf and SQNs were shown to be anionic forms at around neutral pH [3]. An important difference between these two semiquinone
1 In the limited numbers of bacterial NDH-1 (complex I counter part), the 9th [4Fe–4S] iron–sulfur cluster N7 is present; its function is not clear [13]. 2 Based on recent analysis of the slow relaxing SQ species in the isolated bovine heart complex I, we have eliminated very slowly relaxing SQNx species from intrinsic components of complex I [14].
0014-5793/$30.00 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2005.06.086
4556
species is that SQNf is very sensitive to uncoupler while SQNs is not [14–16]. Functional roles of protein-bound quinone species are well established in many transmembrane proteins which couple electron and proton transfer, including cytochrome bc1 and b6f complex [17,18], and photosynthetic reaction centers [19,20]. Considerable information is available regarding possible quinone-binding complex I subunits (e.g., ND5 [21], ND4 [22], ND2 [23], 49 kDa [24], and PSST [25]). The ND5 subunit of bovine heart complex I has been labeled by a fenpyroximate photo-affinity analogue that inhibits complex I, which is also inhibited by 5-(N-ethyl-N-isopropyl)-amiloride, a Na+/H+ antiporter inhibitor [21]. Q-binding sites were also proposed in ND4 and ND5 subunits based on amino acid sequences and X-ray structures [26]. Based on thermodynamic data, including pH dependence of Em, effects of membrane potential (Dw) and pH gradients on redox equilibria, cluster N2 has been regarded as a key component in complex I proton translocation [27,28]. A proton-pump mechanism, which was essentially identical to that proposed for cytochrome oxidase by Wikstro¨m was independently proposed for complex I [29]. Our attention has been focused on the interface between the hydrophilic and hydrophobic arms, because this region contains both cluster N2 and the protein-bound QNf [15,30,31]. Two types of energy coupling mechanism for complex I have been proposed: direct redox-coupling [32–34] and indirect conformationalcoupling [35–38]. New information about the spatial arrangement of N2 and QNf and the coupling of SQNf stability to proton electrochemical potential ðDlþ H Þ suggests that QNf is involved in proton translocation [16]. Now we propose a novel complex I proton-pumping mechanism, which is conformationally gated by the QNf-binding site, and features different SQ stability constants in its input and output conformations.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
respect to the mitochondrial inner membrane. In particular, the gi = 2.05 axis of N2 is oriented along the normal to the membrane plane. Therefore, the interspin vector is 65 from the membrane normal, i.e., only 25 from the membrane plane; the vector between N2 and SQNf is oriented rather close to the membrane plane. The projection of the interspin ˚ · vector along the membrane normal is r cosh = 12 A ˚ , as illustrated in Fig. 1A. cos 65 = 5 A The strong dependence of the QNf signal on Dlþ H would suggest that concurrent electron and proton transfer may occur along a significant part of the membrane potential Dw. However, the short projection of the interspin vector along the membrane normal indicates that electron transfer occurs ˚ of the low dielectric region of the inner memacross only 5 A ˚ . This is clearly insuffibrane, which spans on average 40 A stabilization through a cient to account for the observed QNf loop mechanism [40,41].
2. Supporting evidence The 33 G splitting of the gi = 2.05 signal of cluster N2 arises from direct spin–spin interaction with a specific protein-bound semiquinone anion ðQNf Þ [16] that also causes fast spin-relaxation of the QNf and splits its g = 2.00 signal with a peak-to-peak separation of 56 G (resolved only below 25 K). The unique doublet features of both N2 and QNf spectra were simulated with the same interaction parameter set: J (exchange interaction), 55 MHz; R (center˚ (corresponding to 16 MHz of dipoto-center distance), 12 A lar coupling); h (polar angle), 65; / (azimuthal angle), 0 [16]. Major conclusions of our EPR studies are: (i) QNf signals can be observed only in the presence of Dlþ H [15,30]. Thus, QNf signals were always spectrally resolved as the difference between spectra of coupled and uncoupled submitochondrial particles (SMP); (ii) In the spin–spin interaction between , the contribution of exchange interaction cluster N2 and QNf is more dominant than that of the dipolar interaction. This suggests that the electron distributions around cluster N2 and QNf overlap at their edges, facilitating electron transfer. (iii) Salerno et al. [39] demonstrated earlier that iron–sulfur clusters associated with complex I are highly oriented with
Fig. 1. Membrane normal projection based on the computer simula species tion of the spin–spin interaction between cluster N2 and SQNf (see the text for details). The program includes contributions from both exchange and dipolar interactions which was written by W.R. Dunham (A). Possible explanations for QNf stabilization through the classical loop formulation of Mitchellian chemiosmosis, combined with the proton-well open to the positive side of the membrane (B) and open to the negative side (C). Arrows into the well show the direction of the back-pressure of Dw to the H+ flow coupled with electron transfer from N2 to SQNf.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
4557
3. Transmembrane electron and proton transfer do not explain the behavior of SQNf The proton electrochemical gradient ðDlþ H Þ is expressed as Dlþ H ¼ Dw zDpH. Since Dw is the dominant component of the Dlþ H (DpH < 0.5) in respiring mitochondria [42], we focus only on Dw for the sake of simplicity. If we assume that SQNf ˚ toward the inner space side of the mitochondrial inner is 5 A membrane (P-side) relative to N2, electron transfer from NADH on the N side of the membrane to QNf (oxidized form of SQNf) via N2 would be favored by Dw; Q is stabilized by Dw with respect to Q. However, since the SQNf signal requires both reduction of N2 and the SQ state, it is more important to stabilize Q with respect to ubiquinol (QH2). As shown in Fig. 1B, if protonic equilibration is obtained via a P side proton well, QH2 formation would be favored because the second electron and two protons would move with the electric field (and, in the case of the protons, with an effective pH gradient set up in the P side proton well). If this were true, it would be impossible to observe stabilized semiquinone radical ðQNf Þ in the presence of Ôback pressureÕ from Dw. Semiquinone stabilization by Dw could result from QNf (localized on the P side) equilibrating through a proton well open to the N side, because the first electron to make Q goes downhill from N2 to the P side, while the two protons to make QH2 go uphill from the N side (Fig 1C). Clearly, N2 and SQNf are unlikely to exchange electrons over a large fraction of the total Dw, be˚ out of the 12 A ˚ separation is transmembranous. cause only 5 A Compression of Dw by neighboring proton wells might allow modest direct SQNf stabilization, but is unlikely to be more effective than scalar effects (high bulk pH). EPR data and available structural information for complex I [16,29] suggest that both N2 and SQNf are located toward the N side of the lipid bilayer.
4. Devising a gating model for proton transport Loop type mechanisms do not readily account for the stabilization of SQNf by DW, but two state gating models provide an alternative. Because of technical difficulties in EPR experiments, thermodynamic parameters for protein-bound SQNf are not yet available. The apparent Em,7.0 of N2 in bovine heart SMP varies between 150 and 50 mV (average of 100 mV). Because of the thermodynamic stability (Kstab = 2.0 at pH 7.8) and Dw insensitivity of the SQNs species [12], the QNs species is considered to be an n = 1 M n = 2 converter between QNf and Q-pool; the Em,7.0 of Q-pool averages +90 ± 15 mV [43]. Therefore, SQNf is observed when the , in Q-pool is essentially fully reduced. Merely stabilizing QNf the sense of increasing the stability constant, is insufficient to explain the observation of SQNf signals only in the Dw poised system; the apparent Em value of the AQ/AQH2 couple must be lowered. How can this be accomplished by Dw? In Fig. 2A, we introduce a model including only Dw effects on the equilibrium between two conformational states (A and B). We first concentrate on the mechanism of the redox gating system. State A has a stable semiquinone and is favored in the presence of Dw; state B has an unstable semiquinone and is favored in the absence of Dw. Each conformational state has three redox states, oxidized Q, semiquinone anion (Q) and fully reduced quinol (QH2). We define six independent
Fig. 2. (A) A two-state simple model for conformation-driven proton transfer system with protein-bound QNf as a switching component. It includes only the effect of Dw on the equilibrium between A and B and is states of quinone-binding protein; state A has more stable QNf favored by Dw while state B has an unstable QNf . We define 6 independent parameters: Em1 and Em2 for Q/Q and Q/QH2 in the B state, K1, K2, and K3 for equilibrium constants of Q, Q, and QH2 forms in the B state, and c is the number of charges times RT/F (see the text for details). We also give the effect of Dw through exp(Dw/c) in the B fi A conformational change. (B) Simulation of the dependence of signal amplitude as a function of the E values. the QNf h
parameters for this model: Em1 and Em2 are the Em values for BQ/BQ and BQ/BQH2 couples, respectively, and K1, K2, and K3 are equilibrium constants for the AQ/BQ, AQ/ BQ, AQH2/BQH2 transitions in the absence of Dw. The effect of Dw is introduced through a factor exp(Dw/c) which modifies the equilibrium constants, where c is the number of charges (translocated against the Dw in the B fi A conformational change) times RT/F (about 59.6 mV at 25 C). For simplicity, the same value was used for all A/B transitions. The introduction of Dw favors state A as long as net charges are translocated uphill in the transition from the state B (this does not need to imply translocation of a charged molecule, but could result from reorientation of dipoles in low dielectric milieu). We assume that Em1 and Em2 are 0 and +120 mV for state B, and both K1 and K2 are 1. In the absence of Dw, K3 favors BQH2; we take K3 as 1 · 104. As a simple case, we also assign c = 1 and Dw = 180 mV. The choice of Em values for state B and the three equilibrium constants forces Em1 and Em2 values of state A to be 0 and 120 mV. Fig. 2B presents a simulation of total semiquinone as a function of Eh, revealing marked thermodynamic stabilization of QNf by Dw. For Dw =
4558
180 mV and Dw = 0 mV both A and B conformations participate in redox equilibria. The gating system of Fig. 2A can readily be coupled to proton translocation, producing a redox-driven pump. Figs. 3A shows an open-box type diagram essentially similar to the one proposed by us in 1982 [29]. Here, ‘‘A’’ signifies the input state conformation, where electrons equilibrate with the low potential side of the coupling site and protons equilibrate with the N side of the membrane. ‘‘B’’ signifies the output state,
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
where electrons equilibrate with the high potential Q-pool and protons equilibrate with the P side of the membrane. AQ is reduced by a low potential electron donor (e.g., cluster . Then, AQ serves as a gate and a carrier N2), forming AQNf Nf for the pump. It accepts two protons from the N-side and one electron from the low potential side, forming AQH2. A conformational change converts AQH2 to BQH2, the reduced output state, which in turn releases two protons to a P side facing proton well, and donates an electron to the Q-pool (presumably
Fig. 3. (A) An open-box model of the conformation-driven proton pumping, in which the proton translocation process is gated by two conformational forms of the QNf-binding protein. Red arrows show the direction of proton movement and blue arrow electron movement. (B) The schematic representation of the pump [29]. (C) A simulation shows the enhancement of the semiquinone signal as a function of redox potential (Eh).
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
via QNs). Our experimental data suggest that the EPR signal of AQNf in the presence of Dlþ H is stronger than that of BQNf ; since the latter does not accumulate during turnover, the returning process to the A state must be relatively rapid. It is interesting that the model has an inherent ratio of (4H+/ 2e) without incorporating additional pumps or channels. Fig. 3B depicts a putative scheme for the entire pump. Simulations similar to those shown in Fig. 2B, but including proton effects, confirmed that QNf stability is also enhanced by Dw in this model (Fig. 3C). Additional information, including thermodynamic and kinetic parameters as well as crystallographic structures, is needed to refine and test this picture.
5. Discussion We devised a transmembrane proton-pump model for complex I (Fig. 3) by connecting the gate described in Fig. 2 to transmembrane subunits (e.g., ND5, ND4, and ND2) which carry transmembrane Na+/H+ antiporter sequence motifs [9– 12]. Hellwig et al. reported that redox changes of cluster N2 induced pKa changes of nearby conserved Glu and Asp, or two conserved Tyr residues [44,45]. These changes may be related to concurrent redox-driven conformational changes of the quinone-binding site. BrandtÕs group recently proposed an indirect mechanism of proton pumping via long range conformational energy transfer [38]. Their proposal, based on EM analysis of Yarrowia lipolytica complex I decorated with monoclonal antibodies to the 49 and 30 kDa subunits, identifies PSST and 49 kDa as key players in ubiquinone reduction and proton pumping and places ˚ away from the negative side membrane surface N2 over 30 A in the hydrophilic arm. The following data instead indicate that the location of N2 is close to the membrane surface: (i) site-directed mutagenesis studies indicate localization of N2 within PSST/NuoB (subunit nomenclature of bovine heart/ Escherichia coli) [46,47]; (ii) PSST/Nqo6 [bovine heart/Paracoccus (P.) denitrificans] [48] and 49 kDa/Nqo4 [49] were cross-linked directly to transmembrane ND3 in P. denitrificans complex I by cross-linkers without a spacer. ND3/Nqo7 is composed of three transmembrane helices with only short extra-membrane loops on the N side of the membrane; (iii) Early work by Ohnishi et al. [50] showed that 50% delipidation of bovine heart complex I dramatically broadened N2 EPR spectra and lowered its apparent Em value, but had little effect on low Em iron–sulfur clusters in the hydrophilic region. Changes in inhibitor-sensitive activity, apparent Em and EPR profile of N2 were largely reversible by reconstitution, suggesting association of N2 with the membrane environment; (iv) Lipid activation study using E. coli complex I (by Sinegina et al. [51]) suggested that the environment of N2 is more hydrophobic than that of other low Em iron–sulfur clusters. These data are consistent with our proposed topographical location of N2 adjacent to the membrane surface. High resolution X-ray crystallographic structures of the whole complex I will resolve this issue in the near future. A hypothetical protein-bound quinone (designated QNy at that time) was previously proposed as a possible complex I proton pump component, operating in tandem with a reductant driven Q-cycle [32]; the former mechanism with 2–4H+/ 2e and the latter with 2H+/2e, respectively. This Q-cycle
4559
mechanism has been disputed based on the absence of Q binding sites in complex I with suitable properties [52]. In this paper, we have not considered a possible involvement of cluster N2 in proton pumping. The ability of Dlþ H to promote reverse electron transfer to NADH suggests that N2 may have some role in the pumping mechanism. We hope to refine the current model as new thermodynamic data become available.
6. Conclusions and future studies The proposed mechanism is conceptually related to previous proposals in cytochrome c oxidase [53] and complex I [29], but is based upon unique properties of quinones as electron and proton carriers. A proton pump based on the Q/QH2 couple is attractive, because it accounts for a (H+/2e) ratio of 4 without additional coupling sites involving flavin (see references cited in [32]) or iron–sulfur linked redox-Bohr protons [34]. Specific thermodynamic parameters for the QNf are urgently needed. Available data suggest that for most bound quinones, pKa < 6 for SQ, and pKa > 10 for quinol [20,34,54]. Since the quinone binding site plays a key role in the proton pump, modification using site-directed mutagenesis is a promising avenue for investigation of the mechanism. This proposal does not rule out a role for cluster N2 in proton pumping, either as a co-gating component or through cooperativity. Acknowledgments: This work is supported by NIH Grant GM 30736 to T.O. The authors thank Dr. S. Tsuyoshi Ohnishi for his stimulating and helpful discussions, and Ms. Mary Leonard for the preparation of the figures.
References [1] Weiss, H., Friedrich, T., Hofhaus, G. and Preis, D. (1991) The respiratory-chain NADH dehydrogenase (complex I) of mitochondria. Eur. J. Biochem. 197, 563–576. [2] Walker, J.E. (1992) The NADH:ubiquinone oxidoreductase (complex I) of respiratory chains. Q. Rev. Biophys. 25, 253–324. [3] Ohnishi, T. (1998) Iron–sulfur clusters/semiquinones in Complex I. Biochim. Biophys. Acta 1364, 186–206. [4] Wikstro¨m, M. (1984) Two protons are pumped from the mitochondrial matrix per electron transferred between NADH and ubiquinone. FEBS Lett. 169, 300–304. [5] Galkin, A.S., Grivennikova, V.G. and Vinogradov, A.D. (1999) H+/2e stoichiometry in NADH-quinone reductase reactions catalyzed by bovine heart submitochondrial particles. FEBS Lett. 451, 157–161. [6] Gue`nebaut, V., Schlitt, A., Weiss, H., Leonard, K.R. and Friedrich, T. (1998) Consistent Structure Between Bacterial and Mitochondrial NADH: Ubiquinone Oxidoreductase (Complex I). J. Mol. Biol. 276, 105–112. [7] Yano, T., Magnitsky, S., SledÕ, V.D., Ohnishi, T. and Yagi, T. (1999) Characterization of the putative 2 · [4Fe-4S] binding Nqo9 subunit of the proton translocating NADH-quinone oxidoreductase (NDH-1) of Paracoccus denitrificans: Expression, reconstitution, and EPR characterization. J. Biol. Chem. 274, 28598–28605. [8] Rasmussen, T., Scheide, D., Brors, B., Kintscher, L., Weiss, H. and Friedrich, T. (2001) Identification of two tetranuclear FeS clusters on the ferredoxin-type subunit of NADH:ubiquinone oxidoreductase (complex I). Biochemistry 40, 6124–6131.
4560 [9] Friedrich, T. and Weiss, H. (1997) Modular evolution of the respiratory NADH:ubiquinone oxidoreductase and the origin of its modules. J. Theor. Biol. 187, 529–540. [10] Pilkington, S.J., Skehel, J.M., Gennis, R.B. and Walker, J.E. (1991) Relationship between mitochondrial NADH-ubiquinone reductase and a bacterial NAD-reducing hydrogenase. Biochemistry 30, 2166–2175. [11] Ito, M., Guffanti, A.A. and Krulwich, T.A. (2001) Mrp-dependent Na+/H+ antiporters of Bacillus exhibit characteristics that are unanticipated for completely secondary active transporters. FEBS Lett. 496, 117–120. [12] Volbeda, A., Charon, M.-H., Piras, C., Hatchikian, E.C., Frey, M. and Fontecilla-Camps, J.C. (1995) Crystal structure of the nickel-iron hydrogenase from Desulfovibrio gigas. Nature 373, 580–587. [13] Nakamaru-Ogiso, E., Yano, T., Ohnishi, T. and Yagi, T. (2002) Characterization of the iron-sulfur cluster coordinated by a cysteine cluster motif (CXXCXXXCX27C) in the Nqo3 subunit in the proton- translocating NADH-quinone oxidoreductase (NDH-1) of Thermus thermophilus HB-8. J. Biol. Chem. 277, 1680–1688. [14] Ohnishi, T., Johnson Jr., J.E., Yano, T., LoBrutto, R. and Widger, R.W. (2005) Thermodynamic and EPR studies of slowly relaxing ubisemiquinone species in the isolated bovine heart complex I. FEBS Lett. 579, 500–5006. [15] Magnitsky, S., Toulokhonova, L., Yano, T., Sled, V.D., Ha¨gerha¨ll, C., Grivennikova, V.G., Burbaev, D.S., Vinogradov, A.D. and Ohnishi, T. (2002) EPR characterization of ubisemiquinones and iron-sulfur cluster N2, central components of the energy coupling in the NADH-ubiquinone oxidoreductase (complex I) in situ. J. Bioenerg. Biomembr. 34, 193–208. [16] Yano, T., Dunham, W.R. and Ohnishi, T. (2005) Characterization of the Dlþ H -sensitive SQ species (SQNf) and the interaction with cluster N2: New insight into the energy-coupled electron transfer in complex I. Biochemistry 44, 1744–1754. [17] Hunte, C., Palsdottir, H. and Trumpower, B.L. (2003) Proton motive pathways and mechanisms in the cytochrome bc1 complex. FEBS Lett. 545, 39–46. [18] Cramer, W.A., Zhang, H., Yan, J., Kurisu, G. and Smith, J.L. (2004) Evolution of photosynthesis: Time-independent structure of the cytochrome b6f complex. Biochemistry 43, 5921–5929. [19] Okamura, M.Y., Paddock, M.L., Graige, M.S. and Feher, G. (2000) Proton and electron transfer in bacterial reaction centers. Biochim. Biophys. Acta 1458, 148–163. [20] Zhu, Z. and Gunner, M.R. (2005) Energetics of QuinoneDependent Electron and Proton Transfers in Rhodobacter spheroides Photosynthetic Reaction Centers. Biochemistry 44, 82–96. [21] Nakamaru-Ogiso, E., Sakamoto, K., Matsuno-Yagi, A., Miyoshi, H. and Yagi, T. (2003) The ND5 subunit was labeled by a photoaffinity analogue of fenpyroximate in bovine mitochondrial complex I. Biochemistry 42, 746–754. [22] Gong, X., Xie, T., Yu, L., Hesterberg, M., Scheide, D., Friedrich, T. and Yu, C.A. (2003) The ubiquinone-binding site in NADH:ubiquinone oxidoreductase from Escherichia coli. J. Biol. Chem. 278, 25731–25737. [23] Amarneh, B. and Vik, S.B. (2003) Mutagenesis of subunit N of the Escherichia coli complex I. Identification of the initiation codon and the sensitivity of mutants to decylubiquinone. Biochemistry 42, 4800–4808. [24] Darrouzet, E. and Dupuis, A. (1997) Genetic evidence for the existence of two quinone related inhibitor binding sites in NADHCoQ reductase. Biochim. Biophys. Acta 1319, 1–4. [25] Schuler, F., Yano, T., Di Bernardo, S., Yagi, T., Yankovskaya, V., Singer, P.T. and Casida, J.E. (1999) NADH-quinone oxidoreductase: PSST subunit couples electron transfer from ironsulfur cluster N2 to quinone. Proc. Natl. Acad. Sci. USA 96, 4149–4153. [26] Fisher, N. and Rich, P.R. (2000) A motif for quinone binding sites in respiratory and photosynthetic systems. J. Mol. Biol. 296, 1153–1162. [27] Ohnishi, T. (1976) Studies on the mechanism of site I energy conservation. Eur. J. Biochem. 64, 91–103. [28] DeVault, D. (1976) Theory of iron-sulfur center N-2 oxidation and reduction by ATP. J. Theor. Biol. 62, 115–139.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561 [29] Ohnishi, T. and Salerno, J.C. (1982) Iron-sulfur clusters in the mitochondrial electron-transport chain (Spiro, T.G., Ed.), Iron– Sulfur Proteins, Vol. 4, pp. 285–327, Wiley Publishing Co., New York. [30] Vinogradov, A.D., Sled, V.D., Burbaev, D.S., Grivennikova, V.G., Moroz, I.A. and Ohnishi, T. (1995) Energy-dependent complex I-associated ubisemiquinones in submitochondrial particles. FEBS Lett. 370, 83–87. [31] van Belzen, R., Kotlyar, A.B., Moon, N., Dunham, W.R. and Albracht, S.P.J. (1997) The iron-sulfur clusters 2 and ubisemiquinone radicals of NADH:ubiquinone oxidoreductase are involved in energy coupling in submitochondrial particles. Biochemistry 36, 886–893. [32] Dutton, P.L., Moser, C.C., Sled, V.D., Daldal, F. and Ohnishi, T. (1998) A reductant-induced oxidation mechanism for Complex I. Biochim. Biophys. Acta 1364, 245–257. [33] Brandt, U. (1997) Proton-translocation by membrane-bound NADH:ubiquinone-oxidoreductase (complex I) through redoxgated ligand conduction. Biochim. Biophys. Acta 1318, 79–91. [34] Brandt, U. (1999) Proton translocation in the respiratory chain involving ubiquinone-a hypothetical semiquinone switch mechanism for complex I. BioFactors 9, 95–101. [35] Belogrudov, G. and Hatefi, Y. (1994) Catalytic sector of complex I (NADH:ubiquinone oxidoreductase): Subunit stoichiometry and substrate-induced conformation changes. Biochemistry 33, 4571–4576. [36] Sazanov, L.A., Peak-Chew, S.Y., Fearnley, I.M. and Walker, J.E. (2000) Resolution of the membrane domain of bovine complex I into subcomplexes: implications for the structural organization of the enzyme. Biochemistry 39, 7229–7235. [37] Yagi, T. and Matsuno-Yagi, A. (2003) The Proton-Translocating NADH-Quinone Oxidoreductase in the Respiratory Chain: The Secret Unlocked. Biochemistry 42, 2266–2274. [38] Brandt, U., Kerscher, S., Drose, S., Zwicker, K. and Zickermann, V. (2003) Proton pumping by NADH:ubiquinone oxidoreductase. A redox driven conformational change mechanism?. FEBS Lett. 545, 9–17. [39] Salerno, J.C., Blum, H. and Ohnishi, T. (1979) The orientation of iron-sulfur clusters and a spin-coupled ubiquinone pair in the mitochondrial membrane. Biochim. Biophys. Acta 547, 270–281. [40] Mitchell, P. (1968) Chemiosmotic Coupling and Energy Transduction, Glynn Research, Bodmin. [41] Mitchell, P. (1979) KeilinÕs respiratory chain concept and its chemiosmotic consequences. Science 206, 1148–1159. [42] Nicholls, D.G. and Ferguson, S.J. (2002) Bioenergetics 3, Academic Press, Inc., London. [43] Ding, H., Moser, C.C., Robertson, D.E., Tokito, M.K., Daldal, F. and Dutton, P.L. (1995) Ubiquinone pair in the Qo site central to the primary energy conversion reactions of cytochrome bc1 complex. Biochemistry 34, 15979–15996. [44] Hellwig, P., Scheide, D., Bungert, S., Mantele, W. and Friedrich, T. (2000) FT-IR spectroscopic characterization of NADH:ubiquinone oxidoreductase (complex I) from Escherichia coli: oxidation of FeS cluster N2 is coupled with the protonation of an aspartate or glutamate side chain. Biochemistry 39, 10884–10891. [45] Flemming, D., Hellwig, P. and Friedrich, T. (2003) Involvement of tyrosines 114 and 139 of subunit NuoB in the proton pathway around cluster N2 in Escherichia coli NADH:ubiquinone oxidoreductase. J. Biol. Chem. 278, 3055–3062. [46] Flemming, D., Schlitt, A., Spehr, V., Bischof, T. and Friedrich, T. (2003) Iron-sulfur cluster N2 of the Escherichia coli NADH: ubiquinone oxidoreductase (complex I) is located on subunit NuoB. J. Biol. Chem. 278, 47602–47609. [47] Gurrath, M. and Friedrich, T. (2004) Adjacent cysteines are capable of ligating the same tetranuclear iron-sulfur cluster. Prot. Struct. Funct. Bioinf. 56, 556–563. [48] Kao, M.C., Di Bernardo, S., Matsuno-Yagi, A. and Yagi, T. (2002) Characterization of the membrane domain Nqo11 subunit of the proton-translocating NADH-quinone oxidoreductase of Paracoccus denitrificans. Biochemistry 41, 4377–4384. [49] Kao, M.C., Matsuno-Yagi, A. and Yagi, T. (2004) Subunit proximity in the H(+)-translocating NADH-quinone oxidoreductase probed by zero-length cross-linking. Biochemistry 43, 3750– 3755.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561 [50] Ohnishi, T., Leigh, J.S., Ragan, C.I. and Racker, E. (1974) Low temperature electron paramagnetic resonance studies on ironsulfur centers in cardiac NADH dehydrogenase. Biochem. Biophys. Res. Commun. 56, 775–782. [51] Sinegina, L., Wikstro¨m, M., Verkhovsky, M.I. and Verkhoskaya, M.L. (2005) Activation of isolated NADH:ubiquinone reductase (complex I) from Escherichia coli by detergent and phospholipids. Recovery of ubiquinone reductase activity and changes in EPR signals of iron–sulfur clusters. Biochemistry 44, 8500–8506.
4561 [52] Okun, J.G., Lummen, P. and Brandt, U. (1999) Three classes of inhibitors share a common binding domain in mitochondrial complex I (NADH:ubiquinone oxidoreductase). J. Biol. Chem. 274, 2625–2630. [53] Wikstro¨m, M., Krab, K. and Saraste, M. (1981) Cytochrome oxidase: A synthesis, Academic Press, London. [54] Rich, P.R. (1984) Electron and proton transfers through quinones and cytochrome bc complexes. Biochim. Biophys. Acta 768, 53– 79.
FEBS Letters 579 (2005) 4555–4561
Hypothesis
Conformation-driven and semiquinone-gated proton-pump mechanism in the NADH-ubiquinone oxidoreductase (complex I) Tomoko Ohnishia,*, John C. Salernob a
Johnson Research Foundation, Department of Biochemistry and Biophysics, University of Pennsylvania, Philadelphia, PA 19104, United States b Department of Biology, Rensselaer Polytechnic Institute, Troy, NY 12181, United States Received 16 June 2005; accepted 27 June 2005 Available online 26 July 2005 Edited by Peter Brzezinski
Abstract A novel mechanism for proton/electron transfer is proposed for NADH-quinone oxidoreductase (complex I) based on the following findings: (1) EPR signals of the protein-bound Þ fast-relaxing semiquinone anion radicals (abbreviated as QNf are observable only in the presence of proton-transmembrane are electrochemical potential; (2) Iron–sulfur cluster N2 and QNf directly spin-coupled; and (3) The projection of the interspin vector ˚ along the membrane normal [Yano, T., Dunham, extends only 5 A W.R. and Ohnishi, T. (2005) Biochemistry, 44, 1744–1754]. We propose that the proton pump is operated by redox-driven conformational changes of the quinone binding protein. In the input state, semiquinone is reduced to quinol, acquiring two protons from the N (matrix) side of the mitochondrial inner membrane and an electron from the low potential (NADH) side of the respiratory chain. A conformational change brings the protons into position for release at the P (inter-membrane space) side of the membrane via a proton-well. Concomitantly, an electron is donated to the quinone pool at the high potential side of the coupling site. The system then returns to the original state to repeat the cycle. This hypothesis provides a useful frame work for further investigation of the mechanism of proton translocation in complex I. 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Keywords: NADH-Q oxidoreductase; Proton-pumping mechanism; Semiquinone-gating
1. Introduction NADH-ubiquinone oxidoreductase (complex I), which is located at the entry point of the mitochondrial respiratory chain, * Corresponding author. Fax: +215 573 3748. E-mail address: [email protected] (T. Ohnishi).
Abbreviations: Complex I, NADH-ubiquinone oxidoreductase; Dw, membrane potential; Dlþ H , proton-electrochemical gradient; Em, redoxmidpoint potential; EPR, electron paramagnetic resonance; N2, iron– sulfur cluster N2; N-side, matrix side of the mitochondrial inner membrane; P-side, inner space side of the mitochondrial inner mem , protein-bound fast-relaxing brane; Q, ubiquinone; QH2, ubiquinol; QNf , protein-bound slow-relaxing ubiubisemiquinone (anion form); QNs semiquinone anion; Q-pool, ubiquinone pool in the mitochondrial inner membrane; SMP, submitochondrial particles; SQ, semiquinone; SQNf, protein-bound fast-relaxing semiquinone (including both anion and neutral forms)
catalyzes electron transfer from NADH to the ubiquinone (Q) pool [1–3], in the reaction: þ þ NADH þ Q þ Hþ þ 4Hþ N $ NAD þ QH2 þ 4HP
In this exergonic redox reaction, the transfer of 2 electrons from 1 mol of NADH to ubiquinone pool in the mitochondrial inner membrane (Q-pool) is coupled to the transfer of 4 protons from the negative (matrix) side (abbreviated by the suffix N) to the positive side (abbreviated by the suffix P) of the mitochondrial inner membrane. The consensus ratio for pumped protons/ electrons is 4H+/2e [4,5]. Currently, complex I is the only mitochondrial respiratory complex without available X-ray crystallographic information. The L-shaped low-resolution complex I structure is composed of nearly perpendicular hydrophilic promontory and hydrophobic membrane arms; the former ex˚ out of the matrix side surface level of the inner tends 160 A membrane (N-side) [6]. One non-covalently bound flavin mononucleotide and eight distinct iron–sulfur clusters [2 binuclear (N1a and N1b) and 6 tetranuclear clusters (N2, N3–N6a, N6b)] reside in the promontory region 1 [3,7,8], which transfer electrons from substrate NADH to N2 (highest redox-midpoint potential (Em) cluster), located close to the junction between the two arms. The membrane-associated region includes seven mitochondrially encoded transmembrane subunits, of which ND2, ND4, and ND5 are candidates for proton transport components. Friedrich and Weiss [9] proposed a modular evolution hypothesis for complex I based on sequence analysis. Hydrophilic 51, 24, and 75 kDa subunits are similar to NAD+-reducing hydrogenases [10]; hydrophobic membrane components are homologous to bacterial Na+/H+ antiporters [11]; and interfacial amphipathic PSST and 49 kDa subunits are similar to bacterial Ni–Fe hydrogenases [12]. Based on the observation of semiquinone electron paramagnetic resonance (EPR) signals differing in spin-relaxation rates [SQNf(fast) and SQNs(slow)] 2 and in rotenone sensitivity, two distinct ubiquinone (Q)-binding sites were assigned within the membrane component of complex I. Previously, both SQNf and SQNs were shown to be anionic forms at around neutral pH [3]. An important difference between these two semiquinone
1 In the limited numbers of bacterial NDH-1 (complex I counter part), the 9th [4Fe–4S] iron–sulfur cluster N7 is present; its function is not clear [13]. 2 Based on recent analysis of the slow relaxing SQ species in the isolated bovine heart complex I, we have eliminated very slowly relaxing SQNx species from intrinsic components of complex I [14].
0014-5793/$30.00 2005 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2005.06.086
4556
species is that SQNf is very sensitive to uncoupler while SQNs is not [14–16]. Functional roles of protein-bound quinone species are well established in many transmembrane proteins which couple electron and proton transfer, including cytochrome bc1 and b6f complex [17,18], and photosynthetic reaction centers [19,20]. Considerable information is available regarding possible quinone-binding complex I subunits (e.g., ND5 [21], ND4 [22], ND2 [23], 49 kDa [24], and PSST [25]). The ND5 subunit of bovine heart complex I has been labeled by a fenpyroximate photo-affinity analogue that inhibits complex I, which is also inhibited by 5-(N-ethyl-N-isopropyl)-amiloride, a Na+/H+ antiporter inhibitor [21]. Q-binding sites were also proposed in ND4 and ND5 subunits based on amino acid sequences and X-ray structures [26]. Based on thermodynamic data, including pH dependence of Em, effects of membrane potential (Dw) and pH gradients on redox equilibria, cluster N2 has been regarded as a key component in complex I proton translocation [27,28]. A proton-pump mechanism, which was essentially identical to that proposed for cytochrome oxidase by Wikstro¨m was independently proposed for complex I [29]. Our attention has been focused on the interface between the hydrophilic and hydrophobic arms, because this region contains both cluster N2 and the protein-bound QNf [15,30,31]. Two types of energy coupling mechanism for complex I have been proposed: direct redox-coupling [32–34] and indirect conformationalcoupling [35–38]. New information about the spatial arrangement of N2 and QNf and the coupling of SQNf stability to proton electrochemical potential ðDlþ H Þ suggests that QNf is involved in proton translocation [16]. Now we propose a novel complex I proton-pumping mechanism, which is conformationally gated by the QNf-binding site, and features different SQ stability constants in its input and output conformations.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
respect to the mitochondrial inner membrane. In particular, the gi = 2.05 axis of N2 is oriented along the normal to the membrane plane. Therefore, the interspin vector is 65 from the membrane normal, i.e., only 25 from the membrane plane; the vector between N2 and SQNf is oriented rather close to the membrane plane. The projection of the interspin ˚ · vector along the membrane normal is r cosh = 12 A ˚ , as illustrated in Fig. 1A. cos 65 = 5 A The strong dependence of the QNf signal on Dlþ H would suggest that concurrent electron and proton transfer may occur along a significant part of the membrane potential Dw. However, the short projection of the interspin vector along the membrane normal indicates that electron transfer occurs ˚ of the low dielectric region of the inner memacross only 5 A ˚ . This is clearly insuffibrane, which spans on average 40 A stabilization through a cient to account for the observed QNf loop mechanism [40,41].
2. Supporting evidence The 33 G splitting of the gi = 2.05 signal of cluster N2 arises from direct spin–spin interaction with a specific protein-bound semiquinone anion ðQNf Þ [16] that also causes fast spin-relaxation of the QNf and splits its g = 2.00 signal with a peak-to-peak separation of 56 G (resolved only below 25 K). The unique doublet features of both N2 and QNf spectra were simulated with the same interaction parameter set: J (exchange interaction), 55 MHz; R (center˚ (corresponding to 16 MHz of dipoto-center distance), 12 A lar coupling); h (polar angle), 65; / (azimuthal angle), 0 [16]. Major conclusions of our EPR studies are: (i) QNf signals can be observed only in the presence of Dlþ H [15,30]. Thus, QNf signals were always spectrally resolved as the difference between spectra of coupled and uncoupled submitochondrial particles (SMP); (ii) In the spin–spin interaction between , the contribution of exchange interaction cluster N2 and QNf is more dominant than that of the dipolar interaction. This suggests that the electron distributions around cluster N2 and QNf overlap at their edges, facilitating electron transfer. (iii) Salerno et al. [39] demonstrated earlier that iron–sulfur clusters associated with complex I are highly oriented with
Fig. 1. Membrane normal projection based on the computer simula species tion of the spin–spin interaction between cluster N2 and SQNf (see the text for details). The program includes contributions from both exchange and dipolar interactions which was written by W.R. Dunham (A). Possible explanations for QNf stabilization through the classical loop formulation of Mitchellian chemiosmosis, combined with the proton-well open to the positive side of the membrane (B) and open to the negative side (C). Arrows into the well show the direction of the back-pressure of Dw to the H+ flow coupled with electron transfer from N2 to SQNf.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
4557
3. Transmembrane electron and proton transfer do not explain the behavior of SQNf The proton electrochemical gradient ðDlþ H Þ is expressed as Dlþ H ¼ Dw zDpH. Since Dw is the dominant component of the Dlþ H (DpH < 0.5) in respiring mitochondria [42], we focus only on Dw for the sake of simplicity. If we assume that SQNf ˚ toward the inner space side of the mitochondrial inner is 5 A membrane (P-side) relative to N2, electron transfer from NADH on the N side of the membrane to QNf (oxidized form of SQNf) via N2 would be favored by Dw; Q is stabilized by Dw with respect to Q. However, since the SQNf signal requires both reduction of N2 and the SQ state, it is more important to stabilize Q with respect to ubiquinol (QH2). As shown in Fig. 1B, if protonic equilibration is obtained via a P side proton well, QH2 formation would be favored because the second electron and two protons would move with the electric field (and, in the case of the protons, with an effective pH gradient set up in the P side proton well). If this were true, it would be impossible to observe stabilized semiquinone radical ðQNf Þ in the presence of Ôback pressureÕ from Dw. Semiquinone stabilization by Dw could result from QNf (localized on the P side) equilibrating through a proton well open to the N side, because the first electron to make Q goes downhill from N2 to the P side, while the two protons to make QH2 go uphill from the N side (Fig 1C). Clearly, N2 and SQNf are unlikely to exchange electrons over a large fraction of the total Dw, be˚ out of the 12 A ˚ separation is transmembranous. cause only 5 A Compression of Dw by neighboring proton wells might allow modest direct SQNf stabilization, but is unlikely to be more effective than scalar effects (high bulk pH). EPR data and available structural information for complex I [16,29] suggest that both N2 and SQNf are located toward the N side of the lipid bilayer.
4. Devising a gating model for proton transport Loop type mechanisms do not readily account for the stabilization of SQNf by DW, but two state gating models provide an alternative. Because of technical difficulties in EPR experiments, thermodynamic parameters for protein-bound SQNf are not yet available. The apparent Em,7.0 of N2 in bovine heart SMP varies between 150 and 50 mV (average of 100 mV). Because of the thermodynamic stability (Kstab = 2.0 at pH 7.8) and Dw insensitivity of the SQNs species [12], the QNs species is considered to be an n = 1 M n = 2 converter between QNf and Q-pool; the Em,7.0 of Q-pool averages +90 ± 15 mV [43]. Therefore, SQNf is observed when the , in Q-pool is essentially fully reduced. Merely stabilizing QNf the sense of increasing the stability constant, is insufficient to explain the observation of SQNf signals only in the Dw poised system; the apparent Em value of the AQ/AQH2 couple must be lowered. How can this be accomplished by Dw? In Fig. 2A, we introduce a model including only Dw effects on the equilibrium between two conformational states (A and B). We first concentrate on the mechanism of the redox gating system. State A has a stable semiquinone and is favored in the presence of Dw; state B has an unstable semiquinone and is favored in the absence of Dw. Each conformational state has three redox states, oxidized Q, semiquinone anion (Q) and fully reduced quinol (QH2). We define six independent
Fig. 2. (A) A two-state simple model for conformation-driven proton transfer system with protein-bound QNf as a switching component. It includes only the effect of Dw on the equilibrium between A and B and is states of quinone-binding protein; state A has more stable QNf favored by Dw while state B has an unstable QNf . We define 6 independent parameters: Em1 and Em2 for Q/Q and Q/QH2 in the B state, K1, K2, and K3 for equilibrium constants of Q, Q, and QH2 forms in the B state, and c is the number of charges times RT/F (see the text for details). We also give the effect of Dw through exp(Dw/c) in the B fi A conformational change. (B) Simulation of the dependence of signal amplitude as a function of the E values. the QNf h
parameters for this model: Em1 and Em2 are the Em values for BQ/BQ and BQ/BQH2 couples, respectively, and K1, K2, and K3 are equilibrium constants for the AQ/BQ, AQ/ BQ, AQH2/BQH2 transitions in the absence of Dw. The effect of Dw is introduced through a factor exp(Dw/c) which modifies the equilibrium constants, where c is the number of charges (translocated against the Dw in the B fi A conformational change) times RT/F (about 59.6 mV at 25 C). For simplicity, the same value was used for all A/B transitions. The introduction of Dw favors state A as long as net charges are translocated uphill in the transition from the state B (this does not need to imply translocation of a charged molecule, but could result from reorientation of dipoles in low dielectric milieu). We assume that Em1 and Em2 are 0 and +120 mV for state B, and both K1 and K2 are 1. In the absence of Dw, K3 favors BQH2; we take K3 as 1 · 104. As a simple case, we also assign c = 1 and Dw = 180 mV. The choice of Em values for state B and the three equilibrium constants forces Em1 and Em2 values of state A to be 0 and 120 mV. Fig. 2B presents a simulation of total semiquinone as a function of Eh, revealing marked thermodynamic stabilization of QNf by Dw. For Dw =
4558
180 mV and Dw = 0 mV both A and B conformations participate in redox equilibria. The gating system of Fig. 2A can readily be coupled to proton translocation, producing a redox-driven pump. Figs. 3A shows an open-box type diagram essentially similar to the one proposed by us in 1982 [29]. Here, ‘‘A’’ signifies the input state conformation, where electrons equilibrate with the low potential side of the coupling site and protons equilibrate with the N side of the membrane. ‘‘B’’ signifies the output state,
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
where electrons equilibrate with the high potential Q-pool and protons equilibrate with the P side of the membrane. AQ is reduced by a low potential electron donor (e.g., cluster . Then, AQ serves as a gate and a carrier N2), forming AQNf Nf for the pump. It accepts two protons from the N-side and one electron from the low potential side, forming AQH2. A conformational change converts AQH2 to BQH2, the reduced output state, which in turn releases two protons to a P side facing proton well, and donates an electron to the Q-pool (presumably
Fig. 3. (A) An open-box model of the conformation-driven proton pumping, in which the proton translocation process is gated by two conformational forms of the QNf-binding protein. Red arrows show the direction of proton movement and blue arrow electron movement. (B) The schematic representation of the pump [29]. (C) A simulation shows the enhancement of the semiquinone signal as a function of redox potential (Eh).
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561
via QNs). Our experimental data suggest that the EPR signal of AQNf in the presence of Dlþ H is stronger than that of BQNf ; since the latter does not accumulate during turnover, the returning process to the A state must be relatively rapid. It is interesting that the model has an inherent ratio of (4H+/ 2e) without incorporating additional pumps or channels. Fig. 3B depicts a putative scheme for the entire pump. Simulations similar to those shown in Fig. 2B, but including proton effects, confirmed that QNf stability is also enhanced by Dw in this model (Fig. 3C). Additional information, including thermodynamic and kinetic parameters as well as crystallographic structures, is needed to refine and test this picture.
5. Discussion We devised a transmembrane proton-pump model for complex I (Fig. 3) by connecting the gate described in Fig. 2 to transmembrane subunits (e.g., ND5, ND4, and ND2) which carry transmembrane Na+/H+ antiporter sequence motifs [9– 12]. Hellwig et al. reported that redox changes of cluster N2 induced pKa changes of nearby conserved Glu and Asp, or two conserved Tyr residues [44,45]. These changes may be related to concurrent redox-driven conformational changes of the quinone-binding site. BrandtÕs group recently proposed an indirect mechanism of proton pumping via long range conformational energy transfer [38]. Their proposal, based on EM analysis of Yarrowia lipolytica complex I decorated with monoclonal antibodies to the 49 and 30 kDa subunits, identifies PSST and 49 kDa as key players in ubiquinone reduction and proton pumping and places ˚ away from the negative side membrane surface N2 over 30 A in the hydrophilic arm. The following data instead indicate that the location of N2 is close to the membrane surface: (i) site-directed mutagenesis studies indicate localization of N2 within PSST/NuoB (subunit nomenclature of bovine heart/ Escherichia coli) [46,47]; (ii) PSST/Nqo6 [bovine heart/Paracoccus (P.) denitrificans] [48] and 49 kDa/Nqo4 [49] were cross-linked directly to transmembrane ND3 in P. denitrificans complex I by cross-linkers without a spacer. ND3/Nqo7 is composed of three transmembrane helices with only short extra-membrane loops on the N side of the membrane; (iii) Early work by Ohnishi et al. [50] showed that 50% delipidation of bovine heart complex I dramatically broadened N2 EPR spectra and lowered its apparent Em value, but had little effect on low Em iron–sulfur clusters in the hydrophilic region. Changes in inhibitor-sensitive activity, apparent Em and EPR profile of N2 were largely reversible by reconstitution, suggesting association of N2 with the membrane environment; (iv) Lipid activation study using E. coli complex I (by Sinegina et al. [51]) suggested that the environment of N2 is more hydrophobic than that of other low Em iron–sulfur clusters. These data are consistent with our proposed topographical location of N2 adjacent to the membrane surface. High resolution X-ray crystallographic structures of the whole complex I will resolve this issue in the near future. A hypothetical protein-bound quinone (designated QNy at that time) was previously proposed as a possible complex I proton pump component, operating in tandem with a reductant driven Q-cycle [32]; the former mechanism with 2–4H+/ 2e and the latter with 2H+/2e, respectively. This Q-cycle
4559
mechanism has been disputed based on the absence of Q binding sites in complex I with suitable properties [52]. In this paper, we have not considered a possible involvement of cluster N2 in proton pumping. The ability of Dlþ H to promote reverse electron transfer to NADH suggests that N2 may have some role in the pumping mechanism. We hope to refine the current model as new thermodynamic data become available.
6. Conclusions and future studies The proposed mechanism is conceptually related to previous proposals in cytochrome c oxidase [53] and complex I [29], but is based upon unique properties of quinones as electron and proton carriers. A proton pump based on the Q/QH2 couple is attractive, because it accounts for a (H+/2e) ratio of 4 without additional coupling sites involving flavin (see references cited in [32]) or iron–sulfur linked redox-Bohr protons [34]. Specific thermodynamic parameters for the QNf are urgently needed. Available data suggest that for most bound quinones, pKa < 6 for SQ, and pKa > 10 for quinol [20,34,54]. Since the quinone binding site plays a key role in the proton pump, modification using site-directed mutagenesis is a promising avenue for investigation of the mechanism. This proposal does not rule out a role for cluster N2 in proton pumping, either as a co-gating component or through cooperativity. Acknowledgments: This work is supported by NIH Grant GM 30736 to T.O. The authors thank Dr. S. Tsuyoshi Ohnishi for his stimulating and helpful discussions, and Ms. Mary Leonard for the preparation of the figures.
References [1] Weiss, H., Friedrich, T., Hofhaus, G. and Preis, D. (1991) The respiratory-chain NADH dehydrogenase (complex I) of mitochondria. Eur. J. Biochem. 197, 563–576. [2] Walker, J.E. (1992) The NADH:ubiquinone oxidoreductase (complex I) of respiratory chains. Q. Rev. Biophys. 25, 253–324. [3] Ohnishi, T. (1998) Iron–sulfur clusters/semiquinones in Complex I. Biochim. Biophys. Acta 1364, 186–206. [4] Wikstro¨m, M. (1984) Two protons are pumped from the mitochondrial matrix per electron transferred between NADH and ubiquinone. FEBS Lett. 169, 300–304. [5] Galkin, A.S., Grivennikova, V.G. and Vinogradov, A.D. (1999) H+/2e stoichiometry in NADH-quinone reductase reactions catalyzed by bovine heart submitochondrial particles. FEBS Lett. 451, 157–161. [6] Gue`nebaut, V., Schlitt, A., Weiss, H., Leonard, K.R. and Friedrich, T. (1998) Consistent Structure Between Bacterial and Mitochondrial NADH: Ubiquinone Oxidoreductase (Complex I). J. Mol. Biol. 276, 105–112. [7] Yano, T., Magnitsky, S., SledÕ, V.D., Ohnishi, T. and Yagi, T. (1999) Characterization of the putative 2 · [4Fe-4S] binding Nqo9 subunit of the proton translocating NADH-quinone oxidoreductase (NDH-1) of Paracoccus denitrificans: Expression, reconstitution, and EPR characterization. J. Biol. Chem. 274, 28598–28605. [8] Rasmussen, T., Scheide, D., Brors, B., Kintscher, L., Weiss, H. and Friedrich, T. (2001) Identification of two tetranuclear FeS clusters on the ferredoxin-type subunit of NADH:ubiquinone oxidoreductase (complex I). Biochemistry 40, 6124–6131.
4560 [9] Friedrich, T. and Weiss, H. (1997) Modular evolution of the respiratory NADH:ubiquinone oxidoreductase and the origin of its modules. J. Theor. Biol. 187, 529–540. [10] Pilkington, S.J., Skehel, J.M., Gennis, R.B. and Walker, J.E. (1991) Relationship between mitochondrial NADH-ubiquinone reductase and a bacterial NAD-reducing hydrogenase. Biochemistry 30, 2166–2175. [11] Ito, M., Guffanti, A.A. and Krulwich, T.A. (2001) Mrp-dependent Na+/H+ antiporters of Bacillus exhibit characteristics that are unanticipated for completely secondary active transporters. FEBS Lett. 496, 117–120. [12] Volbeda, A., Charon, M.-H., Piras, C., Hatchikian, E.C., Frey, M. and Fontecilla-Camps, J.C. (1995) Crystal structure of the nickel-iron hydrogenase from Desulfovibrio gigas. Nature 373, 580–587. [13] Nakamaru-Ogiso, E., Yano, T., Ohnishi, T. and Yagi, T. (2002) Characterization of the iron-sulfur cluster coordinated by a cysteine cluster motif (CXXCXXXCX27C) in the Nqo3 subunit in the proton- translocating NADH-quinone oxidoreductase (NDH-1) of Thermus thermophilus HB-8. J. Biol. Chem. 277, 1680–1688. [14] Ohnishi, T., Johnson Jr., J.E., Yano, T., LoBrutto, R. and Widger, R.W. (2005) Thermodynamic and EPR studies of slowly relaxing ubisemiquinone species in the isolated bovine heart complex I. FEBS Lett. 579, 500–5006. [15] Magnitsky, S., Toulokhonova, L., Yano, T., Sled, V.D., Ha¨gerha¨ll, C., Grivennikova, V.G., Burbaev, D.S., Vinogradov, A.D. and Ohnishi, T. (2002) EPR characterization of ubisemiquinones and iron-sulfur cluster N2, central components of the energy coupling in the NADH-ubiquinone oxidoreductase (complex I) in situ. J. Bioenerg. Biomembr. 34, 193–208. [16] Yano, T., Dunham, W.R. and Ohnishi, T. (2005) Characterization of the Dlþ H -sensitive SQ species (SQNf) and the interaction with cluster N2: New insight into the energy-coupled electron transfer in complex I. Biochemistry 44, 1744–1754. [17] Hunte, C., Palsdottir, H. and Trumpower, B.L. (2003) Proton motive pathways and mechanisms in the cytochrome bc1 complex. FEBS Lett. 545, 39–46. [18] Cramer, W.A., Zhang, H., Yan, J., Kurisu, G. and Smith, J.L. (2004) Evolution of photosynthesis: Time-independent structure of the cytochrome b6f complex. Biochemistry 43, 5921–5929. [19] Okamura, M.Y., Paddock, M.L., Graige, M.S. and Feher, G. (2000) Proton and electron transfer in bacterial reaction centers. Biochim. Biophys. Acta 1458, 148–163. [20] Zhu, Z. and Gunner, M.R. (2005) Energetics of QuinoneDependent Electron and Proton Transfers in Rhodobacter spheroides Photosynthetic Reaction Centers. Biochemistry 44, 82–96. [21] Nakamaru-Ogiso, E., Sakamoto, K., Matsuno-Yagi, A., Miyoshi, H. and Yagi, T. (2003) The ND5 subunit was labeled by a photoaffinity analogue of fenpyroximate in bovine mitochondrial complex I. Biochemistry 42, 746–754. [22] Gong, X., Xie, T., Yu, L., Hesterberg, M., Scheide, D., Friedrich, T. and Yu, C.A. (2003) The ubiquinone-binding site in NADH:ubiquinone oxidoreductase from Escherichia coli. J. Biol. Chem. 278, 25731–25737. [23] Amarneh, B. and Vik, S.B. (2003) Mutagenesis of subunit N of the Escherichia coli complex I. Identification of the initiation codon and the sensitivity of mutants to decylubiquinone. Biochemistry 42, 4800–4808. [24] Darrouzet, E. and Dupuis, A. (1997) Genetic evidence for the existence of two quinone related inhibitor binding sites in NADHCoQ reductase. Biochim. Biophys. Acta 1319, 1–4. [25] Schuler, F., Yano, T., Di Bernardo, S., Yagi, T., Yankovskaya, V., Singer, P.T. and Casida, J.E. (1999) NADH-quinone oxidoreductase: PSST subunit couples electron transfer from ironsulfur cluster N2 to quinone. Proc. Natl. Acad. Sci. USA 96, 4149–4153. [26] Fisher, N. and Rich, P.R. (2000) A motif for quinone binding sites in respiratory and photosynthetic systems. J. Mol. Biol. 296, 1153–1162. [27] Ohnishi, T. (1976) Studies on the mechanism of site I energy conservation. Eur. J. Biochem. 64, 91–103. [28] DeVault, D. (1976) Theory of iron-sulfur center N-2 oxidation and reduction by ATP. J. Theor. Biol. 62, 115–139.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561 [29] Ohnishi, T. and Salerno, J.C. (1982) Iron-sulfur clusters in the mitochondrial electron-transport chain (Spiro, T.G., Ed.), Iron– Sulfur Proteins, Vol. 4, pp. 285–327, Wiley Publishing Co., New York. [30] Vinogradov, A.D., Sled, V.D., Burbaev, D.S., Grivennikova, V.G., Moroz, I.A. and Ohnishi, T. (1995) Energy-dependent complex I-associated ubisemiquinones in submitochondrial particles. FEBS Lett. 370, 83–87. [31] van Belzen, R., Kotlyar, A.B., Moon, N., Dunham, W.R. and Albracht, S.P.J. (1997) The iron-sulfur clusters 2 and ubisemiquinone radicals of NADH:ubiquinone oxidoreductase are involved in energy coupling in submitochondrial particles. Biochemistry 36, 886–893. [32] Dutton, P.L., Moser, C.C., Sled, V.D., Daldal, F. and Ohnishi, T. (1998) A reductant-induced oxidation mechanism for Complex I. Biochim. Biophys. Acta 1364, 245–257. [33] Brandt, U. (1997) Proton-translocation by membrane-bound NADH:ubiquinone-oxidoreductase (complex I) through redoxgated ligand conduction. Biochim. Biophys. Acta 1318, 79–91. [34] Brandt, U. (1999) Proton translocation in the respiratory chain involving ubiquinone-a hypothetical semiquinone switch mechanism for complex I. BioFactors 9, 95–101. [35] Belogrudov, G. and Hatefi, Y. (1994) Catalytic sector of complex I (NADH:ubiquinone oxidoreductase): Subunit stoichiometry and substrate-induced conformation changes. Biochemistry 33, 4571–4576. [36] Sazanov, L.A., Peak-Chew, S.Y., Fearnley, I.M. and Walker, J.E. (2000) Resolution of the membrane domain of bovine complex I into subcomplexes: implications for the structural organization of the enzyme. Biochemistry 39, 7229–7235. [37] Yagi, T. and Matsuno-Yagi, A. (2003) The Proton-Translocating NADH-Quinone Oxidoreductase in the Respiratory Chain: The Secret Unlocked. Biochemistry 42, 2266–2274. [38] Brandt, U., Kerscher, S., Drose, S., Zwicker, K. and Zickermann, V. (2003) Proton pumping by NADH:ubiquinone oxidoreductase. A redox driven conformational change mechanism?. FEBS Lett. 545, 9–17. [39] Salerno, J.C., Blum, H. and Ohnishi, T. (1979) The orientation of iron-sulfur clusters and a spin-coupled ubiquinone pair in the mitochondrial membrane. Biochim. Biophys. Acta 547, 270–281. [40] Mitchell, P. (1968) Chemiosmotic Coupling and Energy Transduction, Glynn Research, Bodmin. [41] Mitchell, P. (1979) KeilinÕs respiratory chain concept and its chemiosmotic consequences. Science 206, 1148–1159. [42] Nicholls, D.G. and Ferguson, S.J. (2002) Bioenergetics 3, Academic Press, Inc., London. [43] Ding, H., Moser, C.C., Robertson, D.E., Tokito, M.K., Daldal, F. and Dutton, P.L. (1995) Ubiquinone pair in the Qo site central to the primary energy conversion reactions of cytochrome bc1 complex. Biochemistry 34, 15979–15996. [44] Hellwig, P., Scheide, D., Bungert, S., Mantele, W. and Friedrich, T. (2000) FT-IR spectroscopic characterization of NADH:ubiquinone oxidoreductase (complex I) from Escherichia coli: oxidation of FeS cluster N2 is coupled with the protonation of an aspartate or glutamate side chain. Biochemistry 39, 10884–10891. [45] Flemming, D., Hellwig, P. and Friedrich, T. (2003) Involvement of tyrosines 114 and 139 of subunit NuoB in the proton pathway around cluster N2 in Escherichia coli NADH:ubiquinone oxidoreductase. J. Biol. Chem. 278, 3055–3062. [46] Flemming, D., Schlitt, A., Spehr, V., Bischof, T. and Friedrich, T. (2003) Iron-sulfur cluster N2 of the Escherichia coli NADH: ubiquinone oxidoreductase (complex I) is located on subunit NuoB. J. Biol. Chem. 278, 47602–47609. [47] Gurrath, M. and Friedrich, T. (2004) Adjacent cysteines are capable of ligating the same tetranuclear iron-sulfur cluster. Prot. Struct. Funct. Bioinf. 56, 556–563. [48] Kao, M.C., Di Bernardo, S., Matsuno-Yagi, A. and Yagi, T. (2002) Characterization of the membrane domain Nqo11 subunit of the proton-translocating NADH-quinone oxidoreductase of Paracoccus denitrificans. Biochemistry 41, 4377–4384. [49] Kao, M.C., Matsuno-Yagi, A. and Yagi, T. (2004) Subunit proximity in the H(+)-translocating NADH-quinone oxidoreductase probed by zero-length cross-linking. Biochemistry 43, 3750– 3755.
T. Ohnishi, J.C. Salerno / FEBS Letters 579 (2005) 4555–4561 [50] Ohnishi, T., Leigh, J.S., Ragan, C.I. and Racker, E. (1974) Low temperature electron paramagnetic resonance studies on ironsulfur centers in cardiac NADH dehydrogenase. Biochem. Biophys. Res. Commun. 56, 775–782. [51] Sinegina, L., Wikstro¨m, M., Verkhovsky, M.I. and Verkhoskaya, M.L. (2005) Activation of isolated NADH:ubiquinone reductase (complex I) from Escherichia coli by detergent and phospholipids. Recovery of ubiquinone reductase activity and changes in EPR signals of iron–sulfur clusters. Biochemistry 44, 8500–8506.
4561 [52] Okun, J.G., Lummen, P. and Brandt, U. (1999) Three classes of inhibitors share a common binding domain in mitochondrial complex I (NADH:ubiquinone oxidoreductase). J. Biol. Chem. 274, 2625–2630. [53] Wikstro¨m, M., Krab, K. and Saraste, M. (1981) Cytochrome oxidase: A synthesis, Academic Press, London. [54] Rich, P.R. (1984) Electron and proton transfers through quinones and cytochrome bc complexes. Biochim. Biophys. Acta 768, 53– 79.