Heterocyclic analogues of squamocin as inhibitors of mitochondrial complex I. On the role of the terminal lactone of annonaceous acetogenins

Heterocyclic analogues of squamocin have been semisynthesized by condensation reactions between squamocin-derived alpha-keto esters and heterodinucleophiles. The strong complex I inhibitory potency of squamocin-benzimidazole, a hybrid derivative, illustrates for the first time the functional analogy existing between the terminal butenolide of annonaceous acetogenins and heteroaromatic substructures of classic inhibitors of the enzyme. This finding supports the categorization of this atypical group of inhibitors as antagonists of the ubiquinone substrates. In addition, competition experiments of squamocin-benzimidazole versus squamocin and rolliniastatin-2 suggest that the binding of this hybrid inhibitor is responsible for a negative allosteric effect at the level of the first ubiquinone-binding site (A site) of mitochondrial complex I. This result supports the existence of a large cooperatively regulated inhibitor/ubiquinone-binding pocket located within the catalytic core of the enzyme, consisting of the association of the previously defined affinity sites A and B.     

Hybrid ubiquinone: novel inhibitor of mitochondrial complex I

We synthesized novel ubiquinone analogs by hybridizing the natural ubiquinone ring (2,3-dimethoxy-5-methyl-1,4-benzoquinone) and hydrophobic phenoxybenzamide unit, and named them hybrid ubiquinones (HUs). The HUs worked as electron transfer substrates with bovine heart mitochondrial succinate-ubiquinone oxidoreductase (complex II) and ubiquinol-cytochrome c oxidoreductase (complex III), but not with NADH-ubiquinone oxidoreductase (complex I). With complex I, they acted as inhibitors in a noncompetitive manner against exogenous short-chain ubiquinones irrespective of the presence of the natural ubiquinone ring. Elongation of the distance between the ubiquinone ring and the phenoxybenzamide unit did not recover the electron accepting activity. The structure/activity study showed that high structural specificity of the phenoxybenzamide moiety is required to act as a potent inhibitor of complex I. These findings indicate that binding of the HUs to complex I is mainly decided by some specific interaction of the phenoxybenzamide moiety with the enzyme. It is of interest that an analogous bulky and hydrophobic substructure can be commonly found in recently registered synthetic pesticides the action site of which is mitochondrial complex I.     

Probing the Ubiquinone Reduction Site in Bovine Mitochondrial Complex I Using a Series of Synthetic Ubiquinones and Inhibitors

Studies of the structure–activity relationships of ubiquinones and specific inhibitors are helpful to probe the structural and functional features of the ubiquinone reduction site of bovine heart mitochondrial complex I. Bulky exogenous short-chain ubiquinones serve as sufficient electron acceptors from the physiological ubiquinone reduction site of bovine complex I. This feature is in marked contrast to other respiratory enzymes such as mitochondrial complexes II and III. For various complex I inhibitors, including the most potent inhibitors, acetogenins, the essential structural factors that markedly affect the inhibitory potency are not necessarily obvious. Thus, the loose recognition by the enzyme of substrate and inhibitor structures may reflect the large cavitylike structure of the ubiquinone (or inhibitor) binding domain in the enzyme. On the other hand, several phenomena are difficult to explain by a simple one-catalytic site model for ubiquinone.     

Mode of inhibitory action of Deltalac-acetogenins, a new class of inhibitors of bovine heart mitochondrial complex I

We have revealed that Deltalac-acetogenins, a new class of inhibitors of bovine heart mitochondrial complex I (NADH-ubiquinone oxidoreductase), act differently from ordinary inhibitors such as rotenone and piericidin A [Ichimaru et al. (2005) Biochemistry 44, 816-825]. Since a detailed study of these unique inhibitors might provide new insight into the terminal electron transfer step of the enzyme, we further characterized their inhibitory action using the most potent Deltalac-acetogenin derivative (compound 1). Unlike ordinary complex I inhibitors, 1 had a dose-response curve for inhibition of the reduction of exogenous short-chain ubiquinones that was difficult to explain with a simple bimolecular association model. The inhibitory effect of 1 on ubiquinol-NAD(+) oxidoreductase activity (reverse electron transfer) was much weaker than that on NADH oxidase activity (forward electron transfer), indicating a direction-specific effect. These results suggest that the binding site of 1 is not identical to that of ubiquinone and the binding of 1 to the enzyme secondarily (or indirectly) disturbs the redox reaction of ubiquinone. Using endogenous and exogenous ubiquinone as an electron acceptor of complex I, we investigated the effect of 1 in combination with different ordinary inhibitors on the superoxide production from the enzyme. The results indicated that the level of superoxide production induced by 1 is significantly lower than that induced by ordinary inhibitors probably because of fewer electron leaks from the ubisemiquinone radical to molecular oxygen and that the site of inhibition by 1 is downstream of that by ordinary inhibitors. The unique inhibitory action of hydrophobic Deltalac-acetogenins may be closely associated with the dynamic function of the membrane domain of complex I.     

Kinetic characterization of mitochondrial complex I inhibitors using annonaceous acetogenins.

The NADH:ubiquinone oxidoreductase (complex I) of the mitochondrial respiratory chain is by far the largest and most complicated of the proton-translocating enzymes involved in the oxidative phosphorylation. Many clues regarding the electron pathways from matrix NADH to membrane ubiquinone and the links of this process with the translocation of protons are highly controversial. Different types of inhibitors become valuable tools to dissect the electron and proton pathways of this complex enzyme. Therefore, further knowledge of the mode of action of complex I inhibitors is needed to understand the underlying mechanism of energy conservation. This study presents for the first time a detailed exploration of the inhibitory action of the Annonaceous acetogenins, the most powerful inhibitors of the mammalian enzyme, taking as the head-series rolliniastatin-1, rolliniastatin-2, and corossolin. Despite their close chemical resemblance, each of them inhibits the complex I with different kinetic features reflecting differential binding to the enzyme.     

Exploring the inhibitor binding pocket of respiratory complex I

Numerous hydrophobic and amphipathic compounds including several detergents are known to inhibit the ubiquinone reductase reaction of respiratory chain complex I (proton pumping NADH:ubiquinone oxidoreductase). Guided by the X-ray structure of the peripheral arm of complex I from Thermus thermophilus we have generated a large collection of site-directed mutants in the yeast Yarrowia lipolytica targeting the proposed ubiquinone and inhibitor binding pocket of this huge multiprotein complex at the interface of the 49-kDa and PSST subunits. We could identify a number of residues where mutations changed I(50) values for representatives from all three groups of hydrophobic inhibitors. Many mutations around the domain of the 49-kDa subunit that is homologous to the [NiFe] centre binding region of hydrogenase conferred resistance to DQA (class I/type A) and rotenone (class II/type B) indicating a wider overlap of the binding sites for these two types of inhibitors. In contrast, a region near iron-sulfur cluster N2, where the binding of the n-alkyl-polyoxyethylene-ether detergent C(12)E(8) (type C) was exclusively affected, appeared comparably well separated. Taken together, our data provide structure-based support for the presence of distinct but overlapping binding sites for hydrophobic inhibitors possibly extending into the ubiquinone reduction site of mitochondrial complex I.     

Characterization of the stimulus for reactive oxygen species generation in calcium-overloaded mitochondria

We have used two different probes with distinct detection properties, dichlorodihydrofluorescein diacetate and Amplex Red/horseradish peroxidase, as well as different respiratory substrates and electron transport chain inhibitors, to characterize the reactive oxygen species (ROS) generation by the respiratory chain in calcium-overloaded mitochondria. Regardless of the respiratory substrate, calcium stimulated the mitochondrial generation of ROS, which were released at both the mitochondrial-matrix side and the extra-mitochondrial space, in a way insensitive to the mitochondrial permeability transition pores inhibitor cyclosporine A. In glutamate/malate-energized mitochondria, inhibition at complex I or complex III (ubiquinone cycle) similarly modulated ROS generation at either mitochondrial-matrix side or extra-mitochondrial space; this also occurred when the backflow of electrons to complex I in succinate-energized mitochondria was inhibited. On the other hand, in succinate-energized mitochondria the modulation of ROS generation at mitochondrial-matrix side or extra-mitochondrial space depends on the site of complex III which was inhibited. These results allow a straight comparison between the effects of different respiratory substrates and electron transport chain inhibitors on ROS generation at either mitochondrial-matrix side or extra-mitochondrial space in calcium-overloaded mitochondria.     

Characterization of the stimulus for reactive oxygen species generation in calcium-overloaded mitochondria

Abstract

We have used two different probes with distinct detection properties, dichlorodihydrofluorescein diacetate and Amplex Red/horseradish peroxidase, as well as different respiratory substrates and electron transport chain inhibitors, to characterize the reactive oxygen species (ROS) generation by the respiratory chain in calcium-overloaded mitochondria. Regardless of the respiratory substrate, calcium stimulated the mitochondrial generation of ROS, which were released at both the mitochondrial-matrix side and the extra-mitochondrial space, in a way insensitive to the mitochondrial permeability transition pores inhibitor cyclosporine A. In glutamate/malate-energized mitochondria, inhibition at complex I or complex III (ubiquinone cycle) similarly modulated ROS generation at either mitochondrial-matrix side or extra-mitochondrial space; this also occurred when the backflow of electrons to complex I in succinate-energized mitochondria was inhibited. On the other hand, in succinate-energized mitochondria the modulation of ROS generation at mitochondrial-matrix side or extra-mitochondrial space depends on the site of complex III which was inhibited. These results allow a straight comparison between the effects of different respiratory substrates and electron transport chain inhibitors on ROS generation at either mitochondrial-matrix side or extra-mitochondrial space in calcium-overloaded mitochondria.     

Exploring the binding site of acetogenin in the ND1 subunit of bovine mitochondrial complex I

¹²⁵I-labeled (trifluoromethyl)phenyldiazirinyl acetogenin, [¹²⁵I]TDA, a photoaffinity labeling probe of acetogenin, photo-cross-links to the ND1 subunit of bovine heart mitochondrial NADH-ubiquinone oxidoreductase (complex I) with high specificity [M. Murai, A. Ishihara, T. Nishioka, T. Yagi, and H. Miyoshi, (2007) The ND1 subunit constructs the inhibitor binding domain in bovine heart mitochondrial complex I, Biochemistry 46 6409-6416.]. To identify the binding site of [¹²⁵I]TDA in the ND1 subunit, we carried out limited proteolysis of the subunit cross-linked by [¹²⁵I]TDA using various proteases and carefully analyzed the fragmentation patterns. Our results revealed that the cross-linked residue is located within the region of the 4th to 5th transmembrane helices (Val144-Glu192) of the subunit. It is worth noting that an excess amount of short-chain ubiquinones such as ubiquinone-2 (Q₂) and 2-azido-Q₂ suppressed the cross-linking by [¹²⁵I]TDA in a concentration-dependent way. Although the question of whether the binding sites for ubiquinone and different inhibitors in complex I are identical remains to be answered, the present study provided, for the first time, direct evidence that an inhibitor (acetogenin) and ubiquinone competitively bind to the enzyme. Considering the present results along with earlier photoaffinity labeling studies, we propose that not all inhibitors acting at the terminal electron transfer step of complex I necessarily bind to the ubiquinone binding site itself.     

Efficient large scale purification of his-tagged proton translocating NADH:ubiquinone oxidoreductase (complex I) from the strictly aerobic yeast Yarrowia lipolytica

Proton translocating NADH:ubiquinone oxidoreductase (complex I) is the largest membrane bound multiprotein complex of the respiratory chain and the only one for which no molecular structure is available so far. Thus, information on the mechanism of this central enzyme of aerobic energy metabolism is still very limited. As a new approach to analyze complex I, we have recently established the strictly aerobic yeast Yarrowia lipolytica as a model system that offers a complete set of convenient genetic tools and contains a complex I that is stable after isolation. For crystallization of complex I and to obtain its molecular structure it is a prerequisite to prepare large amounts of highly pure enzyme. Here we present the construction of his-tagged complex I that for the first time allows efficient affinity purification. Our protocol recovers almost 40% of complex I present in Yarrowia mitochondrial membranes. Overall, 40-80 mg highly pure and homogeneous complex I can be obtained from 10 l of an overnight Y. lipolytica culture. After reconstitution into asolectin proteoliposomes, the purified enzyme exhibits full NADH:ubiquinone oxidoreductase activity, is fully sensitive to inhibition by quinone analogue inhibitors and capable of generating a proton-motive force.     

Isolation and characterization of complex I, rotenone-sensitive NADH: ubiquinone oxidoreductase, from the procyclic forms of Trypanosoma brucei.

Additional characterization of complex I, rotenone-sensitive NADH:ubiquinone oxidoreductase, in the mitochondria of Trypanosoma brucei brucei has been obtained. Both proline:cytochrome c reductase and NADH:ubiquinone oxidoreductase of procyclic T. brucei were inhibited by the specific inhibitors of complex I rotenone, piericidin A, and capsaicin. These inhibitors had no effect on succinate: cytochrome c reductase activity. Antimycin A, a specific inhibitor of the cytochrome bc1 complex (ubiquinol:cytochrome c oxidoreductase), blocked almost completely cytochrome c reductase activity with either proline or succinate as electron donor, but had no inhibitory effect on NADH:ubiquinone oxidoreductase activity. The rotenone-sensitive NADH:ubiquinone oxidoreductase of procyclic T. brucei was partially purified by sucrose density centrifugation of mitochondria solubilized with dodecyl-beta-D-maltoside, with an approximately eightfold increase in specific activity compared to that of the mitochondrial membranes. Four polypeptides of the partially purified enzyme were identified as the homologous subunits of complex I (51 kDa, PSST, TYKY, and ND4) by immunoblotting with antibodies raised against subunits of Paracoccus denitrificans and against synthetic peptides predicted from putative complex I subunit genes encoded by mitochondrial and nuclear T. brucei DNA. Blue Native polyacrylamide gel electrophoresis of T. brucei mitochondrial membrane proteins followed by immunoblotting revealed the presence of a putative complex I with a molecular mass of 600 kDa, which contains a minimum of 11 polypeptides determined by second-dimensional Tricine-SDS/PAGE including the 51 kDa, PSST and TYKY subunits.     

Identification of the ubiquinone-binding site of NADH:ubiquinone oxidoreductase (complex I) from Neurospora crassa.

In order to localize the ubiquinone-binding site of complex I (NADH:ubiquinone oxidoreductase), a novel photoreactive ubiquinone analogue (Q0C7ArN3) has been synthesized. It is shown that the direct chemical precursor of this analogue (Q0C7ArNO2) and the analogue itself are accepted as substrates in an enzyme assay utilizing ubiquinone-depleted mitochondrial membranes of Neurospora crassa. The activity of the enzyme applying these derivatives is inhibited by 50% at a concentration of 9 and 20 microM rotenone. Photoaffinity labeling experiments were performed with both isolated complex I and whole mitochondrial membranes of N. crassa under various conditions. In each of these experiments a protein subunit with an apparent molecular mass of about 9.5 kDa was labeled with high specificity. Radioactive labeling was totally prevented by the addition of ubiquinone-2 at concentrations higher than 500 microM but was not affected by comparable concentrations of rotenone or other hydrophobic substances. In the labeling experiments using whole membranes, the labeling signal was dramatically increased in the presence of 1.5 mM NADH. These results strongly suggest that the ubiquinone analogue interacts specifically with the enzyme.     

Tightly-bound ubiquinone in the Escherichia coli respiratory Complex I

NADH:ubiquinone oxidoreductase (Complex I), the electron input enzyme in the respiratory chain of mitochondria and many bacteria, couples electron transport to proton translocation across the membrane. Complex I is a primary proton pump; although its proton translocation mechanism is yet to be known, it is considered radically different from any other mechanism known for redox-driven proton pumps: no redox centers have been found in its membrane domain where the proton translocation takes place. Here we studied the properties and the catalytic role of the enzyme-bound ubiquinone in the solubilized, purified Complex I from Escherichia coli. The ubiquinone content in the enzyme preparations was 1.3±0.1 per bound FMN residue. Rapid mixing of Complex I with NADH, traced optically, demonstrated that both reduction and re-oxidation kinetics of ubiquinone coincide with the respective kinetics of the majority of Fe-S clusters, indicating kinetic competence of the detected ubiquinone. Optical spectroelectrochemical redox titration of Complex I followed at 270-280nm, where the redox changes of ubiquinone contribute, did not reveal any transition within the redox potential range typical for the membrane pool, or loosely bound ubiquinone (ca. +50-+100mV vs. NHE, pH 6.8). The transition is likely to take place at much lower potentials (E(m) ≤-200mV). Such perturbed redox properties of ubiquinone indicate that it is tightly bound to the enzyme's hydrophobic core. The possibility of two ubiquinone-binding sites in Complex I is discussed.     

Quinone binding and reduction by respiratory complex I

Complex I (NADH:ubiquinone oxidoreductase) has a central function in oxidative phosphorylation and hence for efficient ATP production in most prokaryotic and eukaryotic cells. This huge membrane protein complex transfers electrons from NADH to ubiquinone and couples this exergonic redox reaction to endergonic proton pumping across bioenergetic membranes. Although quinone reduction seems to be critical for energy conversion, this part of the reaction is least understood. Here we summarize and discuss experimental evidence indicating that complex I contains an extended ubiquinone binding pocket at the interface of the 49-kDa and PSST subunits. Close to iron–sulfur cluster N2, the proposed immediate electron donor for ubiquinone, a highly conserved tyrosine constitutes a critical element of the quinone reduction site. A possible quinone exchange path leads from cluster N2 to the N-terminal β-sheet of the 49-kDa subunit. We discuss the possible functions of a highly conserved HRGXE motif and a redox–Bohr group associated with cluster N2. Resistance patterns observed with a large number of point mutations suggest that all types of hydrophobic complex I inhibitors also act at the interface of the 49-kDa and the PSST subunit. Finally, current controversies regarding the number of ubiquinone binding sites and the position of the site of ubiquinone reduction are discussed.     

Inhibition of mitochondrial NADH:ubiquinone oxidoreductase by ethoxyformic anhydride

The NADH:ubiquinone, but not the NADH:ferricyanide, reductase activity of mitochondrial complex I (NADH:ubiquinone oxidoreductase) is inhibited by incubation of the enzyme at pH 6.0 and 0 degree C with ethoxyformic anhydride (EFA), and the inhibition is partially reversed by subsequent incubation of EFA-treated complex I with hydroxylamine. These results and spectral changes of EFA-treated complex I in the u.v. region are consistent with modification of essential histidyl or tyrosyl residues between the primary NADH dehydrogenase and the site of ubiquinone reduction. Treatment of complex I with EFA in the presence of high concentrations of Seconal or Demerol did not protect against EFA inactivation, suggesting that the site of EFA modification may not be the same as the inhibiton sites of Seconal and Demerol. However, the presence of NADH during incubation of complex I with EFA greatly enhanced the inhibition rate, indicating that the reduced conformation of complex I is more susceptible to attack by EFA.     

Assembly of NADH: ubiquinone reductase (complex I) in Neurospora mitochondria. Independent pathways of nuclear-encoded and mitochondrially encoded subunits

NADH:ubiquinone reductase, the respiratory chain complex I of mitochondria, consists of some 25 nuclear-encoded and seven mitochondrially encoded subunits, and contains as redox groups one FMN, probably one internal ubiquinone and at least four iron-sulphur clusters. We are studying the assembly of the enzyme in Neurospora crassa. The flux of radioactivity in cells that were pulse-labelled with [35S]methionine was followed through immunoprecipitable assembly intermediates into the holoenzyme. Labelled polypeptides were observed to accumulate transiently in a Mr 350,000 intermediate complex. This complex contains all mitochondrially encoded subunits of the enzyme as well as subunits encoded in the nucleus that have no homologous counterparts in a small, merely nuclear-encoded form of the NADH:ubiquinone reductase made by Neurospora crassa cells poisoned with chloramphenicol. With regard to their subunit compositions, the assembly intermediate and small NADH:ubiquinone reductase complement each other almost perfectly to give the subunit composition of the large complex I. These results suggest that two pathways exist in the assembly of complex I that independently lead to the preassembly of two major parts, which subsequently join to form the complex. One preassembled part is related to the small form of NADH:ubiquinone reductase and contributes most of the nuclear-encoded subunits, FMN, three iron-sulphur clusters and the site for the internal ubiquinone. The other part is the assembly intermediate and contributes all mitochondrially encoded subunits, one iron-sulphur cluster and the catalytic site for the substrate ubiquinone. We discuss the results with regard to the evolution of the electron pathway through complex I.     

A small isoform of NADH:ubiquinone oxidoreductase (complex I) without mitochondrially encoded subunits is made in chloramphenicol-treated Neurospora crassa

In mitochondria of Neurospora crassa grown in the presence of chloramphenicol a small form of NADH:ubiquinone reductase is made in place of the normal electron-transfer-complex I. This smaller enzyme has a molecular mass of approximately 350 kDa and consists of (at least) 13 different subunits which are all synthesized in the cytoplasm. The complex I which is normally found in Neurospora has a molecular mass of approximately 700 kDa and consists of around 30 different subunits, of which at least six are made in the mitochondria. Immunoblotting and peptide mapping suggest that the subunits of the small enzyme are homologous to subunits of the large enzyme, one subunit might even be identical. The small and the large NADH:ubiquinone reductases have the same high-affinity binding site for NADH but the two enzymes differ in the affinity and inhibitor sensitivity of the ubiquinone-binding site. The possibility is discussed that the small NADH:ubiquinone reductase is primitive isoform of complex I.     

Cryo-electron microscopy reveals how acetogenins inhibit mitochondrial respiratory complex I

Mitochondrial complex I (NADH:ubiquinone oxidoreductase), a crucial enzyme in energy metabolism, captures the redox potential energy from NADH oxidation/ubiquinone reduction to create the proton motive force used to drive ATP synthesis in oxidative phosphorylation. High-resolution single-particle electron cryo-EM analyses have provided detailed structural knowledge of the catalytic machinery of complex I, but not of the molecular principles of its energy transduction mechanism. Although ubiquinone is considered to bind in a long channel at the interface of the membrane-embedded and hydrophilic domains, with channel residues likely involved in coupling substrate reduction to proton translocation, no structures with the channel fully occupied have yet been described. Here, we report the structure (determined by cryo-EM) of mouse complex I with a tight-binding natural product acetogenin inhibitor, which resembles the native substrate, bound along the full length of the expected ubiquinone-binding channel. Our structure reveals the mode of acetogenin binding and the molecular basis for structure–activity relationships within the acetogenin family. It also shows that acetogenins are such potent inhibitors because they are highly hydrophobic molecules that contain two specific hydrophilic moieties spaced to lock into two hydrophilic regions of the otherwise hydrophobic channel. The central hydrophilic section of the channel does not favor binding of the isoprenoid chain when the native substrate is fully bound but stabilizes the ubiquinone/ubiquinol headgroup as it transits to/from the active site. Therefore, the amphipathic nature of the channel supports both tight binding of the amphipathic inhibitor and rapid exchange of the ubiquinone/ubiquinol substrate and product.     

Superoxide radical formation by pure complex I (NADH:ubiquinone oxidoreductase) from Yarrowia lipolytica

Generation of reactive oxygen species (ROS) is increasingly recognized as an important cellular process involved in numerous physiological and pathophysiological processes. Complex I (NADH:ubiquinone oxidoreductase) is considered as one of the major sources of ROS within mitochondria. Yet, the exact site and mechanism of superoxide production by this large membrane-bound multiprotein complex has remained controversial. Here we show that isolated complex I from Yarrowia lipolytica forms superoxide at a rate of 0.15% of the rate measured for catalytic turnover. Superoxide production is not inhibited by ubiquinone analogous inhibitors. Because mutant complex I lacking a detectable iron-sulfur cluster N2 exhibited the same rate of ROS production, this terminal redox center could be excluded as a source of electrons. From the effect of different ubiquinone derivatives and pH on this side reaction of complex I we concluded that oxygen accepts electrons from FMNH2 or FMN semiquinone either directly or via more hydrophilic ubiquinone derivatives.     

Inhibitors of the quinone-binding site allow rapid superoxide production from mitochondrial NADH:ubiquinone oxidoreductase (complex I)

Neither the route of electron transport nor the sites or mechanism of superoxide production in mitochondrial complex I has been established. We examined the rates of superoxide generation (measured as hydrogen peroxide production) by rat skeletal muscle mitochondria under a variety of conditions. The rate of superoxide production by complex I during NADH-linked forward electron transport was less than 10% of that during succinate-linked reverse electron transport even when complex I was fully reduced by pyruvate plus malate in the presence of the complex III inhibitor, stigmatellin. This asymmetry was not explained by differences in protonmotive force or its components. However, when inhibitors of the quinone-binding site of complex I were added in the presence of ATP to generate a pH gradient, there was a rapid rate of superoxide production by forward electron transport that was as great as the rate seen with reverse electron transport at the same pH gradient. These observations suggest that quinone-binding site inhibitors can make complex I adopt the highly radical-producing state that occurs during reverse electron transport. Despite complete inhibition of NADH: ubiquinone oxidoreductase activity in each case, different classes of quinone-binding site inhibitor (rotenone, piericidin, and high concentrations of myxothiazol) gave different rates of superoxide production during forward electron transport (the rate with myxothiazol was twice that with rotenone) suggesting that the site of rapid superoxide generation by complex I is in the region of the ubisemiquinone-binding sites and not upstream at the flavin or low potential FeS centers.