Structure of bacterial respiratory complex I

    Structure of bacterial respiratory complex I John M. Berrisford, Rozbeh Baradaran, Leonid A. Sazanov PII: DOI: Reference:

S0005-2728(16)30004-4 doi: 10.1016/j.bbabio.2016.01.012 BBABIO 47589

To appear in:

BBA - Bioenergetics

Received date: Revised date: Accepted date:

2 November 2015 18 January 2016 20 January 2016

Please cite this article as: John M. Berrisford, Rozbeh Baradaran, Leonid A. Sazanov, Structure of bacterial respiratory complex I, BBA - Bioenergetics (2016), doi: 10.1016/j.bbabio.2016.01.012

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

PT

Structure of bacterial respiratory complex I

RI

John M. Berrisford1, Rozbeh Baradaran2 and Leonid A. Sazanov*

SC

*Corresponding author. Institute of Science and Technology Austria (IST Austria),

NU

Am Campus 1, 3400 Klosterneuburg, Austria. E-mail: [email protected]

MA

Summary

Complex I (NADH:ubiquinone oxidoreductase) plays a central role in cellular

D

energy production, coupling electron transfer between NADH and quinone to proton

TE

translocation. It is the largest protein assembly of respiratory chains and one of the most elaborate redox membrane proteins known. Bacterial enzyme is about half the

AC CE P

size of mitochondrial and thus provides its important “minimal” model. Dysfunction of mitochondrial complex I is implicated in many human neurodegenerative diseases. The L-shaped complex consists of a hydrophilic arm, where electron transfer occurs, and a membrane arm, where proton translocation takes place. We have solved the crystal structures of the hydrophilic domain of complex I from Thermus thermophilus, the membrane domain from Escherichia coli and recently of the intact, entire complex I from T. thermophilus (536 kDa, 16 subunits, 9 iron-sulfur clusters, 64 transmembrane helices). The 95 Å long electron transfer pathway through the enzyme proceeds from the primary electron acceptor flavin mononucleotide through seven conserved Fe-S clusters to the unusual elongated quinone-binding site at the interface

1

European Bioinformatics Institute, Cambridge CB10 1SD, U.K. 2Memorial SloanKettering Cancer Center, 430 E 67th Street, New York, 10065, USA 1

ACCEPTED MANUSCRIPT with the membrane domain. Four putative proton translocation channels are found in the membrane domain, all linked by the central flexible axis containing charged

PT

residues. The redox energy of electron transfer is coupled to proton translocation by the as yet undefined mechanism proposed to involve long-range conformational

SC

RI

changes.

INTRODUCTION

NU

Complex I catalyses the transfer of two electrons from NADH to quinone,

MA

coupled to the translocation of protons across the membrane, contributing to the proton-motive force required for the synthesis of ATP [1-4]:

TE

D

NADH + Q + H+ + nH+in -> NAD+ + QH2 + nH+out

AC CE P

Here “in” refers to the mitochondrial matrix or bacterial cytoplasm (N-side of the membrane), “out” – to intermembrane space or bacterial periplasm (P-side). The reaction is fully reversible – in the presence of a proton-motive force complex I can reduce NAD+ using quinol as a source of electrons [5]. The physiological role of the reverse electron transport in complex I is not yet established. The number of protons translocated per NADH oxidized (“n”) is currently considered to be 4, although this stoichiometry was studied mostly for mitochondrial enzyme [6-8], and less so for bacterial counterpart [9]. Different quinones are used as electron acceptors by complex I from different species. The mitochondrial enzyme uses ubiquinone, complex I from Thermus thermophilus uses menaquinone and Escherichia coli complex I uses either ubiquinone or menaquinone, depending on the growth conditions encountered by the bacteria [2, 3].

2

ACCEPTED MANUSCRIPT Complex I is one of the largest macromolecular assemblies known and is the largest component of the respiratory chain. The mammalian mitochondrial enzyme

PT

consists of 44 different subunits (45 in total, as the SDAP subunit is present in two copies [10]) of about 1000 kDa in total [11, 12]. The best studied examples are

RI

complex I from bovine heart [1, 12], the obligate aerobic yeast Yarrowia lipolytica

SC

[13, 14] and the fungus Neurospora crassa [15]. The prokaryotic enzyme is simpler and generally consists of 14 conserved “core” subunits with a combined molecular

NU

mass of about 550 kDa [2, 16]. Escherichia coli and few other bacterial enzymes

MA

consist of 13 subunits because two subunits are fused (NuoC and NuoD) [17]. Thermus thermophilus and its closest relatives contain additional frataxin-like and chaperone-like subunits and so have 16 subunits in the entire complex [16, 18]. The

TE

D

best studied examples of bacterial complex I are from E. coli [19], T. thermophilus [16, 20], Paracoccus denitrificans [21, 22] and Aquifex aeolicus [23]. Prokaryotic

AC CE P

complex I is also called NDH-1, standing for NADH dehydrogenase-1, in contrast to the simpler non-proton pumping NDH-2 enzyme [24]. Analogues of all conserved subunits of bacterial complex I are found in the mitochondrial enzyme [1] and they contain equivalent redox components [2]. Both mitochondrial and bacterial enzymes have a characteristic L-shaped structure, with the hydrophobic arm embedded in the membrane and the hydrophilic peripheral arm protruding into the mitochondrial matrix or the bacterial cytoplasm [23, 25, 26]. Thus, the mechanism is likely to be conserved throughout the species and the bacterial enzyme represents a useful ‘minimal’ model of mitochondrial complex I. Apart from canonical complex I, its less well characterized analogues are also found in chloroplasts, cyanobacteria and archaea. In these cases, three-subunit NADH dehydrogenase module of complex I is replaced by F420H2 dehydrogenase in archaea

3

ACCEPTED MANUSCRIPT and by as yet unknown electron input module in chloroplasts and cyanobacteria [27, 28].

PT

Complex I contributes about 40% of the proton flux across inner mitochondrial membranes, used for ATP synthesis, and so is central for energy production in

RI

eukaryotes. Since neuronal tissues rely heavily on respiration for their energy needs,

SC

even relatively small drop in complex I activity due to mutations can lead to severe human neurodegenerative diseases [29]. Complex I has also been suggested to be a

NU

major source of reactive oxygen species (ROS) in mitochondria, which can damage

MA

mtDNA and may be one of the causes of aging [30]. Parkinson’s disease, at least in its sporadic form, may be caused by increased ROS production from malfunctioning complex I [31]. In bacteria, such as E. coli, expression of complex I depends on the

TE

D

growth conditions and can be induced at low oxygen concentrations, when higher effectiveness of the respiratory chain is required [32, 33]. At high oxygen

AC CE P

concentrations complex I may be replaced by faster, non energy-conserving NDH-2 enzyme [33].

Subunit composition

Nomenclature of complex I subunits historically differs for different species, complicating comparisons (Table 1). Using harsh detergents or unfavorable pH, the enzyme can be split into three main domains, likely reflecting its evolutionary origins [19, 27, 34]. Hydrophilic peripheral arm consists of the dehydrogenase domain and connecting domain, which provides a link to the hydrophobic membrane domain. Peripheral arm consists of subunits Nqo1-6 and 9 (T. thermophilus nomenclature, which will be used throughout for simplicity) and a membrane arm consists of Nqo7-8 and Nqo10-14 [27]. The order of genes in bacterial operons encoding complex I

4

ACCEPTED MANUSCRIPT subunits generally reflects these domains, with the exception that subunits Nqo7 and Nqo9 swapped places [35]. All known cofactors - the primary electron acceptor flavin

PT

mononucleotide (FMN) and 8 to 9 iron-sulfur (Fe-S) clusters [2, 36, 37], are found in the hydrophilic arm, comprising the catalytic core of the enzyme. The proton-pumping

RI

machinery must reside in the membrane arm, which contains 63-64 transmembrane

SC

(TM) helices in bacteria [18, 38] and about 78 in mitochondria [10, 39]. The mechanism of coupling between electron transfe[39]r and proton translocation is not

NU

fully established. Two main models are being discussed currently: indirect with some

MA

contributions from direct coupling (redox-driven through chemical intermediates, usually employing modifications of the Q cycle) [40, 41] and purely indirect or

D

conformation-driven coupling [2-4, 18, 39, 42, 43].

TE

In addition to the 14 core subunits, mitochondrial complex I contains many (30

AC CE P

in bovine enzyme[12]) additional subunits, termed “supernumerary” or “accessory”. These subunits are found in both peripheral and membrane domains of the complex and most are smaller than 20 kDa [4, 12, 44]. Many of them have similarities to the proteins of known function, such as acyl-carrier protein or a cell death regulatory protein [45]. We have purified and characterized complex I from Paracoccus denitrificans, a close relative of mitochondria’s ancestors. Surprisingly, this enzyme was found to contain three “accessory” subunits common to all eukaryotes [22], suggesting that the elaboration of complex I started before the symbiosis occurred (consistent with phylogenetic tree analysis placing P. denitrificans subunits closer to the root than mitochondrial subunits; data not shown). The physiological role of the “accessory” subunits is not yet established. Structural knowledge on mitochondrial complex I has advanced dramatically recently, although full atomic structures either of the entire complex or of supernumerary subunits are not yet solved. From the data

5

ACCEPTED MANUSCRIPT available so far [10, 39] it is clear that the structures and key features of the core subunits are very similar between mitochondria and bacteria, validating the use of

PT

bacterial complex I as a “minimal” model for human enzyme. Supernumerary subunits appear to provide a structural scaffold around the core, possibly stabilizing the

RI

complex [10, 39].

SC

Purification of complex I was achieved first with the bovine enzyme,

NU

employing solubilisation of membranes in deoxycholate followed by a series of precipitation steps [46]. This preparation gives a highly active enzyme, sensitive to

MA

the specific inhibitor rotenone. However, it contains significant amount of impurities. More modern preparation using solubilisation in dodecyl-β-D-maltoside (DDM)

D

followed by a series of chromatography steps gives pure, but less active enzyme [44,

TE

47], unless lipids are added to the buffers [48]. Another extensively characterized

AC CE P

mitochondrial enzyme has been purified from the yeast Yarrowia lipolytica, where attachment of a hexa-histidine tag to the NUGM (30 kDa) subunit allowed fast and efficient purification of complex I [49]. Bacterial complex I is generally more fragile than the mitochondrial enzyme, and so its first purification was only achieved relatively recently, employing solubilization of E. coli membranes in alkylglucoside, followed by several chromatography steps and sucrose gradient centrifugation [19]. Increased yield of protein can be obtained by using genetically engineered strains over-expressing the complex [50]. Another way to increase yield is to grow E. coli under low oxygen conditions [32]. The enzyme is stable in the pH range 5.5 to 6.5 when purified in DDM, but will fragment in shorter carbon chain detergents [34]. The T. thermophilus enzyme is stable in similar pH range and can be purified intact in tridecyl-β-Dmaltoside (TDM) at room temperature (it will fragment at 40C) [18, 43]. Electron

6

ACCEPTED MANUSCRIPT microscopy (EM) has shown that the E. coli enzyme has an L-shape morphology similar to the mitochondrial enzyme, with two arms of similar length (about 20 nm

PT

each), but somewhat thinner [25, 26, 51]. The only other intact bacterial complex I’s characterized by EM until now are from A. aeolicus [23] and P. denitrificans [22],

RI

showing familiar L-shape.

SC

Complex I purified from all species so far exists as a monomer in detergent solution. Only under very mild conditions (e.g. using digitonin as solubilising

NU

detergent) it is possible to isolate mitochondrial respiratory supercomplexes

MA

containing complex I in combination with complexes III and IV [52, 53], although the

D

physiological role of such assemblies is not yet established.

TE

Cofactors

The cofactors known to be present in complex I are FMN and Fe-S clusters,

AC CE P

which have relatively low optical extinction with broad, overlapping absorbance bands, giving a light brown colour to the protein. Binuclear Fe-S clusters ([2Fe-2S]) contribute to broad absorbance peaks at 420, 470 and 560 nm, tetranuclear clusters ([4Fe-4S]) – at 420 nm [54], and FMN – at 450 nm [55]. Absorbance in these regions of the spectra decreases upon reduction of the complex [54], but it is not possible to de-convolute spectra of individual redox centers, with the exception of contribution from high-potential cluster N2 [54] and FMN [55]. Due to these limitations, fast kinetic studies in solution, such as those on heme-containing proteins, are not possible and detailed studies of the redox centers in complex I have been performed by electron paramagnetic resonance (EPR) spectroscopy on frozen samples. Extensive EPR studies over the years, particularly in Ohnishi’s group, using bovine complex I as a main model, allowed to establish the presence of EPR signals

7

ACCEPTED MANUSCRIPT (in the g ~ 2 region) from two binuclear (N1a and N1b) and four tetranuclear (N2, N3, N4 and N5) Fe-S clusters in complex I [36]. In this nomenclature, spin relaxation

PT

rates of complex I clusters increase (and optimal EPR sample temperature decreases) with the cluster identification number, from N1 to N5 [36]. The analysis of known

RI

Fe-S cluster binding motifs in the sequence of subunits suggested that complex I

SC

should contain two more tetranuclear clusters [36], indicating that their EPR signals are not observed in the intact complex, due to either very low potentials [56] or

NU

unusual EPR properties. Thus, mitochondrial and bacterial complex I usually contain

MA

8 Fe-S clusters (2 binuclear and 6 tetranuclear), making it one of the most elaborated Fe-S assemblies known. In T. thermophilus, E. coli and some other bacteria there is an additional Fe-S binding motif in subunit Nqo3 (NuoG), coordinating tetranuclear

TE

D

cluster N7. These predictions were confirmed by the X-ray structure of the hydrophilic domain of complex I from T. thermophilus [57].

AC CE P

The assignment of some of EPR signals to the structurally defined clusters (and corresponding coordinating motifs in the sequence) is under discussion [56]. Previously, EPR signals N1 to N7 (used as cluster names), were assigned to sequence motifs in certain subunits (i.e. cysteine residues coordinating clusters) [36]. This assignment was subsequently used for clusters defined in the structure [57]. However, for some of the EPR signals and some of the clusters this assignment is not consistent with emerging mutagenesis data [58], analyses of sub-complexes [59], and new EPR data [60]. The exact EPR signatures for these clusters (N4, N5 and N6a/b) remain to be fully established. To avoid confusion with cluster nomenclature, the traditional N1 to N7 nomenclature can be used as “nicknames” for structurally defined clusters (Fig. 1B).

8

ACCEPTED MANUSCRIPT STRUCTURE Core subunits

PT

The first X-ray crystal structure for complex I was that of the hydrophilic arm of the enzyme from T. thermophilus, determined to 3.1 Å resolution, in both the apo and

RI

NADH bound forms [57, 61]. These structures allowed identification of the subunits

SC

and prosthetic groups within this sub-complex. The X-ray crystal structure of the membrane arm from E. coli to 3.9 Å resolution allowed identification and

NU

determination of the organization of subunits in the membrane arm [43]. Later the

MA

resolution was improved to 3.0 Å resolution and the atomic model of the membrane domain was described, revealing many unusual features of the protein fold [38]. Recent X-ray crystal structure of the entire complex from T. thermophilus to 3.3 Å

[18].

TE

D

resolution finally revealed how the hydrophilic and membrane arms function together

AC CE P

In the overview of this structure, shown in Fig. 1A, the peripheral arm of complex I is a Y-shaped assembly about 140 Å high and the membrane arm is beanshaped, about 180 Å long. The peripheral arm from T. thermophilus contains nine subunits: Nqo1-6, Nqo9 and unexpectedly, two additional subunits not part of the nqo operon, Nqo15 [57] and loosely-bound Nqo16 [18]. One uppermost tip of the peripheral arm is formed by the subunits Nqo1 and Nqo2, and the other by the Cterminal domain of Nqo3. The main stem is formed by the N-terminal domain of Nqo3 and the connecting subunits. Its lower part consists of subunits Nqo4 and Nqo6 (the latter coordinates the terminal Fe-S cluster N2), and it forms an interface with the membrane domain, sitting on top of subunit Nqo8. The position of the peripheral domain within the whole complex is consistent with recent EM data on the complex

9

ACCEPTED MANUSCRIPT from Aquifex aeolicus [23], Neurospora crassa [62], Yarrowia lipolytica [63], Bos Taurus [10, 25, 64] and E. coli [51].

PT

The membrane arm comprises seven subunits: Nqo7-8 and Nqo10-14, with 64 transmembrane (TM) helices, most of them lying normal to the membrane [18, 38,

RI

43]. 16 TM helices are present in subunit Nqo12, 14 in Nqo13, 14 in Nqo14, 5 in

SC

Nqo10, 3 in Nqo11, 3 in Nqo7 and 9 in Nqo8. Subunits Nqo12, Nqo13 and Nqo14 are arranged, like carriages in the train, towards the distal end of the membrane arm, with

NU

Nqo12 furthest away from the hydrophilic arm. These three subunits share sequence

MA

similarity with Na+ / H+ Mrp (Multiple resistance and pH adaptation) antiporter complex subunits (MrpA, MrpD and MrpD, respectively [65, 66]), contain 14 conserved TM helices each and are likely responsible for each pumping a proton

TE

D

across the membrane. A fourth proton channel is formed by subunits Nqo7, 8, 10 and 11 [18], which also contain an elongated cavity for the quinone molecule to enter

AC CE P

close to the hydrophilic arm. Unexpectedly, the C-terminus of Nqo12 forms an extended 110 Å alpha helix, which runs along the membrane surface and terminates adjacent to Nqo14 [18, 38, 43]. This amphipathic helix, termed helix HL, links together antiporter-like subunits at least as a stabilising “strap” [67-69] but may also act as a putative coupling element [70]. Another such element (termed βH) is formed from series of connected β-hairpins and helices on the opposite side of the domain [38]. Most subunits of the complex have structural homology to other proteins, which apparently served as smaller “building blocks” during the evolution of the enzyme. These “blocks”, containing different redox centers, fit together in the peripheral arm in such a way that a continuous electron transfer pathway through the enzyme is formed. The evolutionary origins of complex I can thus be traced to different types of

10

ACCEPTED MANUSCRIPT ferredoxins (subunits Nqo2 and Nqo9), FeFe-hydrogenases (N-terminus of subunit Nqo3), molybdopterin-containing enzymes (C-terminus of subunit Nqo3) and NiFe-

PT

hydrogenases (subunits Nqo4 and Nqo6). Such similarities were noted also from sequence comparisons [1, 4, 27, 71].

RI

Many protein complexes with an as yet unknown structure seem to share bigger

SC

“building blocks” with complex I. Several NAD+-reducing enzymes, for example cytoplasmic NiFe-hydrogenase [1] and formate dehydrogenase [72] from Ralstonia

NU

eutropha contain analogues of subunits Nqo1-3, the “dehydrogenase domain” of

MA

complex I. Membrane-bound NiFe-hydrogenases contain analogues of the “connecting domain” subunits (Nqo4-6, and Nqo9) and of most membrane domain subunits, which may be involved in proton pumping [71]. It is likely that complex I

TE

D

originated from the unification of two pre-evolved complexes, a soluble hydrogenase and an Mrp-like antiporter [73].

AC CE P

Seven iron-sulfur clusters form a chain in the peripheral arm, between the FMN and the quinone-binding site, spanning a distance of some 95 Å (Fig. 1B) [37]. The two electrons extracted from NADH are passed one by one down the iron-sulfur cluster chain to a quinone molecule in the vicinity of cluster N2. Electrons entering the complex from NADH are transferred to FMN, which resides in Nqo1 [36, 59]. At least one electron is then transferred from FMN to a tetranuclear cluster N3 that is also found in Nqo1. From cluster N3 the electron is then passed to the binuclear cluster N1b and then on to tetranuclear clusters N4 and then N5 within Nqo3. The cluster N7 is also present in Nqo3 in some bacterial species but with 20 Å between this cluster and its closest cluster, N4, it is too removed from the redox chain to be part of the electron transfer pathway [37]. As cluster N7 is also only conserved in a small number of bacterial species it may represent an evolutionary remnant in

11

ACCEPTED MANUSCRIPT complex I [74]. From N5 the electrons are transferred to the two tetranuclear clusters in Nqo9: N6a and N6b, before being passed to the final tetranuclear cluster, N2,

PT

within Nqo6. N2 has the highest electron potential of the clusters and passes its electron onto the quinone substrate, ubiquinone or menaquinone depending on the

RI

organism [2]. Cluster N2 lies only 12 Å from quinone head-group binding in a

SC

channel formed between Nqo6 and the adjacent Nqo4 subunit and close to the hydrophobic domain [18].

NU

Adjacent to FMN, within subunit Nqo2, lies the binuclear cluster N1a, which

MA

does not form part of the main redox chain, representing a diversion (Fig. 1B). This cluster is fully conserved and so must be important for function, possibly acting as an

TE

D

anti-oxidant (see below).

Additional subunits

AC CE P

An unexpected find in the structure of the hydrophilic arm of T. thermophilus complex I was the presence of an additional subunit, Nqo15 [16, 57]. This protein is not part of the nqo operon and is only present in a few extremophilic prokaryotic species. Nqo15 has a fold that is highly homologous to frataxin [75, 76], a protein which is involved in the biogenesis of iron-sulfur centers and has no sequence homologue in T. thermophilus. Although the exact role of frataxin is still under some debate it has been shown to form spherical oligomers in vitro, which can bind iron, suggesting a potential role in iron storage akin to ferritin [77]. Studies by Adinolfi et al. have expanded on this role by showing that bacterial frataxin, CyaY, actually inhibits the production of iron-sulfur clusters in an iron dependent manner through its interaction with the cysteine desulfurase IscS [78]. Although the structures of frataxin and Nqo15 are homologous, sequentially these two proteins are very different.

12

ACCEPTED MANUSCRIPT Frataxin contains several fully conserved acidic residues on the edge of the first βstand and α-helix, which are involved in iron binding [79]. These residues are not

PT

present within Nqo15, which instead has series of four histidine residues along a different β-strand on the face of its β-sheet. Within T. thermophilus complex I these

RI

histidines face into a channel formed between subunits Nqo1, 2, 3 and 15. Side chains

SC

of histidines from Nqo1 and Nqo3 also face into this channel. The end of this channel lies in close proximity to the iron-sulfur clusters N3 and N1a. Divalent cations,

NU

including iron mimic manganese, have been shown to bind in this channel [61]. This

MA

leads to the intriguing possibility that this subunit may be involved in iron binding and may have a role in iron-sulfur cluster regeneration. In the structure of the entire T. thermophilus complex, another novel subunit

TE

D

was identified, which is not necessary for oxidoreductase or proton-pumping activity of the enzyme, but is essential for crystallization as it is involved in crystal contacts

AC CE P

[18]. This 14.2 kDa subunit, dubbed Nqo16, belongs to the superfamily DUF3197 (NCBI) of proteins with no known function, and is only found in thermophiles. It has a fold consisting of five-stranded β-sheet flanked by two α-helices on each side, and shares no significant similarity to any known proteins. Nqo16 might play a role as an assembly factor, as the proportion of T. thermophilus complex I containing Nqo16 varies depending on the cell growth [18].

NADH /FMN binding The substrate NADH binds in subunit Nqo1 adjacent to the FMN moiety. The two electrons that are extracted from NADH are first transferred to an adjacent FMN molecule as a hydride ion. Complex I contains a non-traditional Rossmann fold, which binds both FMN and NADH through the incorporation of an additional glycine

13

ACCEPTED MANUSCRIPT rich loop (Fig. 2A) [57, 61]. The residues involved in interactions with both nucleotides are very well conserved suggesting this binding pocket is consistent

PT

between species as varied as bacteria and humans. Within the binding pocket the nicotinamide ring of NADH stacks against the

RI

exposed face of the isoalloxazine ring of the bound FMN, in a manner similar to that

SC

seen in other nucleotide binding flavoenzymes [61]. This positions the B-face of the nicotinamide ring against the re face of the isoalloxazine ring to allow the 4B

NU

hydrogen of NADH to transfer as a hydride to the N5 atom of FMN. The transfer of

MA

the hydride is unusually fast and the distance between these two atoms is shorter than usual, at 3.2 Å. The short distance between the two nucleotides can be explained by the close proximity of the Cβ atom of Nqo1 Glu97 which is within van der Waals

together (Fig. 2A).

TE

D

contact distance with the C4N atom of the bound NADH, forcing the two nucleotides

AC CE P

The binding of the adenine ring of the NADH moiety causes a shift of around 1.5 Å of the loop formed by Nqo1 residues 202 to 207 towards the bound NADH. This positions the side chain of Nqo1 Phe205 1.7 Å closer to the adenine ring promoting a stacking interaction and a hydrogen bond forms between the side chain of Nqo1 Lys202 and the phosphate moiety of the NADH molecule [61]. Complex I is unusual in its ability to use deamino-NADH as a substrate. This is due to NADH binding, unusually, in an extended conformation, so that N6A of the adenine ring (which is an oxygen in deamino-NADH) forms no interactions with the protein. Although there are no large scale changes in the structure of the hydrophilic domain upon reduction by NADH, reproducible shifts of around 1 Å can be seen in a 4-helix bundle in Nqo4 and helices H1 and H2 in Nqo6 [61]. A surprising result was seen in comparison of the coordination of the terminal iron-sulfur cluster N2 between

14

ACCEPTED MANUSCRIPT the oxidized and reduced crystal structures. N2 is coordinated by an unusual tandem cysteine motif consisting of Nqo6 Cys45 and Cys46, with Cys111 and Cys140

PT

completing the ligation [61]. The tandem cysteine motif is fully conserved in complex I and the T. thermophilus structure revealed strained geometry in these residues. In the

RI

oxidized form of the enzyme clear electron density was observed for both the side

SC

chains of Cys45 and Cys46. Upon reduction of the hydrophilic arm by NADH the electron density for the side chain of Cys46 disappears suggesting that this residue has

NU

disconnected from the N2 cluster. However, when crystals of the hydrophilic arm

MA

were reduced with dithionite the electron density for the side chain of the neighboring cysteine, Cys45, was absent and the side chain of Cys46 had clear connecting density to the cluster N2 [61]. In each case it is most likely that the cysteine ligand had been

TE

D

replaced by a solvent molecule and that the disconnected cysteine side chain had become protonated. These results suggest the coordination of N2 is changing

AC CE P

depending on the degree of reduction of the complex (in excess NADH most EPRvisible clusters will be reduced, while in dithionite, which reacts slowly with complex I, some clusters apart from N2 may be oxidised). It is possible to envisage that during the catalytic cycle Cys45 and Cys46 are becoming sequentially protonated and disconnected from N2, before passing their protons, upon reconnection, to the nearby protein side-chains and possibly even to the bound quinone molecule. Additionally, it is likely that these changes in cluster coordination are the driving force for the observed shifts of nearby helices, as we do not see any major changes in the middle of the domain, between NADH site and cluster N2. Whether the movements of the Nqo4 4-helix bundle and helices H1 and H2 in Nqo6 might propagate through into the hydrophobic arm and thus promote proton pumping is still unclear, although these helices do contact several TM helices.

15

ACCEPTED MANUSCRIPT

Quinone binding site

PT

The quinone-binding site, formed between subunits Nqo4, Nqo6, Nqo7 and Nqo8, is unusually elongated and enclosed from the solvent. The hydrophilic head of

RI

the quinone molecule binds in the deep end of a cavity, about 15 Å away from the

SC

membrane surface. It is hydrogen bonded to Tyr87 and His38 from Nqo4; both residues are invariant and essential for activity (Fig. 2B) [18]. The entry point for the

NU

quinone head-group is very narrow (approximately 2-3 × 4-5 Å) and may not allow

MA

any solvent into the cavity when quinone is bound, as its tail will block the entrance. Surprisingly, the chamber is lined mostly by hydrophilic residues, possibly allowing

TE

D

for quinone headgroup to be guided deep into the cavity [18].

FUNCTIONAL IMPLICATIONS

AC CE P

Electron transfer pathway

The electron transfer pathway can be traced unambiguously through the structure, with all redox centres within 14 Å from each other, a maximal distance for electron transfer relevant in biology [80] (Fig. 1B). At pH 7, the two-electron midpoint redox potential (Em) of NADH is about -320 mV, of FMN about -340 mV and of ubiquinone (UQ) about +110 mV (T. thermophilus utilizes menaquinone (MQ), about -80 mV). The one-electron potential of cluster N1a is usually lower, while that of cluster N2 is usually higher than that of most other clusters, which appear to be roughly isopotential at about -250 mV [2, 36]. In the main redox chain, cluster N3 accepts electrons from the flavin, while the high-potential cluster N2 reduces the quinone at the interface with the membrane domain. The free energy available from NADH/UQ pair is about 430 mV, which means that complex I

16

ACCEPTED MANUSCRIPT operates at close to 100% efficiency, achieving stoichiometry of 4H+/2e- (at equilibrium 4∆p=2∆Eh, and ∆p is about 200 mV in mitochondria). Modelling shows

PT

that electron can be transferred from FMN to Q within 50 µs, much faster than about 5 ms required for one catalytic turnover [81]. This is consistent with the fact that most

RI

EPR-visible clusters of complex I are reduced under steady-state NADH oxidation, so

SC

that it is likely that N2 oxidation and quinone binding/release are rate-limiting [82]. Real-time EPR measurements confirmed that first electron is transferred from NADH

NU

to cluster N2 within 90 µs [83]. The modelling calculation above was performed

MA

assuming that all clusters apart from N1a and N2 are isopotential [81], while it is likely that clusters N5 and N6b have very low potentials and so are not observed by

D

EPR, leading to an alternating profile of higher and lower potential clusters along the

TE

chain [56]. Clusters N5 and N6a are separated by the longest distance in the chain (Fig. 1B), so they probably comprise a rate-limiting step in the N3 to N2 pathway. It

AC CE P

remains to be established whether the conserved His ligation of N5 may be related to its position at this “bottleneck” and bestow on it some regulatory properties. Ultrafast freeze quenching experiments suggest that not only N5 to N6a electron transfer is the slowest in the chain, but also that it becomes even slower (~1.2 ms) when cluster N2 is reduced, possibly allowing for synchronization of electron transfer with slower proton-pumping reactions [84].

Cluster N1a and ROS production One intriguing question is the role of cluster N1a, as it is not in the main redox pathway (Fig. 1B). We have suggested that N1a may play the role of an antioxidant, preventing excessive generation of reactive oxygen species (ROS) by complex I [37]. Flavin is now generally regarded as a main source of ROS in complex I [3, 85, 86],

17

ACCEPTED MANUSCRIPT although cluster N2 and (semi)quinone [87] are also being discussed. FMN accepts two electrons simultaneously (as a hydride) from NADH and transfers them one at a

PT

time to one-electron carriers Fe-S clusters. The one-electron redox potential of N1a (~ -380 mV in bovine) is too low for accepting the first electron from reduced FMNH2,

RI

but it is suitable for accepting the second electron, from flavosemiquinone (midpoint

SC

potentials are about -300 mV for FMNH2 / flavosemiquinone and about -390 mV for flavosemiquinone / oxidized flavin [88]). Thus, two electrons from flavin can be

NU

donated nearly simultaneously to two nearby clusters, N3 (first) and N1a (second)

MA

[83, 84]. This mechanism will prevent any significant accumulation of the flavosemiquinone intermediate, which could otherwise react with oxygen, leading to ROS production. The flavin is exposed to the solvent at the deep end of the NADH-

TE

D

binding cavity (Fig. 2A), whereas cluster N1a is shielded and so it is suitable for such a temporary storage of electrons. As oxidation of cluster N2 by quinone is a likely

AC CE P

rate-limiting step, electrons can move from N1a, via FMN, towards cluster N3 as soon as N2 is re-oxidized. The flavosemiquinone formed during this transfer will be very short-lived, as the electron transfer between redox centers is several orders of magnitude faster than quinone binding/release and the flavosemiquinone/FMN pair represents a small, but unfavorable redox potential barrier in this pathway. This is in broad agreement with studies by Ransac et al. who have suggested that N1a is mainly reduced by the flavosemiquinone species and can be oxidized by a different flavosemiquinone species [89]. In their modelling study they showed that this could lead to a decrease in the lifetime of the flavosemiquinone species and thus reduce ROS production. Experimental data suggests that fully reduced flavin is responsible for ROS production in complex I [90], possibly reflecting low concentrations of flavosemiquinone due to the above mechanism. Complex II appears to possess a

18

ACCEPTED MANUSCRIPT similar mechanism (by having an off-pathway heme) for efficient coupling of twoand one- electron transfers, avoiding radical formation [91]. Further studies with site-

PT

directed mutants are required to fully establish the controversial role of N1a ([92] and

RI

our unpublished data).

SC

Proton-translocating channels

Proton pumping occurs in the membrane arm, most likely through the four

NU

channels identified in the structure, as mentioned above – three in antiporter-like

MA

subunits and one at the interface with the hydrophilic domain (Fig. 3). As the protonpumping domains are up to 120 Å away from the quinone binding site this leads to the question of how reduction of the quinone is linked to proton pumping. Most likely

coupling is at play.

TE

D

answer, supported by the structural features, is that long-range conformational

AC CE P

One of the most fascinating features of the structure is the presence of conserved charged and polar residues in the middle of the membrane, which extend from the quinone-binding site at the junction between the two main domains to the tip of the hydrophobic arm subunit Nqo12 (most charged residues are indicated on Fig. 3) [18, 38]. These residues are mostly found in the breaks in discontinuous TM helices and are surrounded by a “river” of water molecules (some modelled and some experimentally observed [38]), which span the entire length of the hydrophobic arm (Fig.3 in ref [18]). The whole arrangement thus represents a flexible hydrophilic central axis of the membrane domain. Many residues from the central axis are key residues in the putative protontranslocating channels. The fold of 14 conserved helices in antiporter-like subunits can be subdivided into a highly conserved core of ten helices (TM 4-13) and the less

19

ACCEPTED MANUSCRIPT conserved TM1-3 and TM14. In the conserved core two sets of five helices are related to each other by internal symmetry, i.e. TMs 4–8 can be superimposed on TMs 9–13.

PT

The symmetry-related helices TM7 and TM12 are interrupted in the middle of the bilayer by an extended loop of 5–7 residues. Such discontinuous helices are normally

RI

found in sites important for ion transport, because they introduce flexibility (due to

SC

disrupted secondary structure) and charge (dipole at the exposed ends of helical fragments) to the middle of the membrane [93, 94]. The broken helices are

NU

strategically located: TM7s contact helix HL, while TM12s are placed at the interfaces of subunits. In addition, TM8s, found in the centre of subunits at the

MA

interface of symmetry related domains, are partly unwound in the middle by π-bulges

D

[95], which are also usually found at protein functional sites. Elevated b-factors near

TE

the breaks in the helices in the refined crystal structures of complex I (PDB 3RKO and 4HEA) are consistent with the increased flexibility around these areas. Such

AC CE P

flexibility would be important in order to facilitate the propagation of any conformation changes during the catalytic cycle as discussed below. Each symmetry-related set of five helices contains an apparent half-channel for proton translocation: cytoplasmic half in TM4-8 and periplasmatic half in TM9-13 [18, 96]. The half-channels are formed by conserved polar residues lining polar cavities. Helix TM7 contains, in its intramembranous loop, a key lysine (termed LysTM7), which is in close proximity to its pKa-modulating glutamate on TM5 (GluTM5), forming the central part of the cytoplasm-linked half-channel. The periplasm-linked half-channels contain a central lysine (replaced by glutamate in Nqo13) within the loop in TM12 (Lys/GluTM12). The half-channels are linked in the middle of the membrane by conserved polar residues, including the lysine from broken TM8 (histidine in Nqo12) [18]. Thus, each antiporter-like subunit contains a

20

ACCEPTED MANUSCRIPT single proton channel formed from two connected half-channels (indicated by blue arrows in Fig. 3).

PT

The fourth proton channel is also formed from two connected half-channels, linked to the cytoplasm via Nqo8 subunit and to the periplasm via subunits Nqo10 and

RI

Nqo11 [18]. The arrangement of the TM helices in Nqo8 is unusual, as nearly all are

SC

dramatically tilted relative to membrane normal (core TM helices 2-6 are tilted by up to 45o), which may confer extra flexibility in this area. Surprisingly, Nqo8 core

NU

presents antiporter-like half-channel fold with Glu130 and Glu163 in the GluTM5

MA

position and Glu213 and Glu248 near the LysTM7 position. However, there are many more charged (conserved) residues in the membrane part of Nqo8 compared to the antiporter-like subunits. In the second half-channel, Glu32 in subunit Nqo11 mimics

TE

D

the GluTM12, which interacts with the conserved essential Tyr59 in subunit Nqo10. A Glu/Asp quartet and putative water molecules in the center of the membrane link

AC CE P

the two half-channels into a single channel [18] (Fig. 3). Thus, this channel is referred to as the “E-channel” due to this abundance of glutamate residues in its center. More detailed recent analysis [97] indicates that additional proton input pathways into the central parts of subunits are possible: one from the cytoplasm roughly along central TM8 and another as a ‘side entry’ from the interface between subunits, through GluTM5 (indicated as dark violet arrows in Fig. 3). Multiple input pathways would enable the effective capture of protons, present in low concentrations in the cytoplasm (which has a high pH). In contrast, an exit pathway into the periplasm seems to be possible only around TM12 and it appears much less conductive. This would be consistent with the necessity for the protein to tightly control ejection of protons against the gradient into the low-pH periplasm. A similar organization is apparent in the E-channel, which has a porous cytoplasmic half but

21

ACCEPTED MANUSCRIPT less clear connection to the periplasm. It is therefore likely that the central hydrophilic axis of complex I is usually poised for action, fully loaded with protons captured from

PT

the cytoplasm and re-distributed between subunits. Once during the catalytic cycle,

into the periplasm from each of the four channels.

RI

the conformation of the membrane arm may be changed so that one proton is ejected

SC

It is energetically very expensive for the protein to maintain such a high concentration of charged residues in the middle of the membrane. The overall

NU

architecture thus suggests that the central hydrophilic axis probably plays a

MA

dominating role in the propagation and coordination of conformational changes

D

during the catalytic cycle.

TE

Coupling between electron transfer and proton translocation The total lack of redox groups in the hydrophobic domain provides implications

AC CE P

for the mechanism of coupling between electron transfer and proton translocation. From the redox potentials of cofactors as noted above it follows that most of the redox energy is released during the quinone chemistry (with smaller part also released during cluster N2 reduction). The structure of T. thermophilus complex I with the bound quinone analogue decyl-ubiquinone confirmed that the unusual enclosed chamber at the interface between the two main domains is indeed the quinone-binding site [18]. This tight enclosure likely allows the protein to control the protonation of the headgroup via charged residues but not the solvent, so that fully reduced ubiquinol (Q2-) or key charged residues nearby remain unprotonated until the energy of electrostatic interactions is used up to drive conformational changes [73]. The most plausible scenario of the reaction mechanism is as following: the transfer of two electrons from NADH to the quinone via the FMN and the Fe-S

22

ACCEPTED MANUSCRIPT clusters results in charged species in the Q chamber, which interact electrostatically with the Glu/Asp quartet via a funnel of charged residues in between, driving

PT

conformational changes first in the E-channel. Shifts of helices observed upon reduction of cluster N2 probably help with these changes, allowing for full redox

RI

energy to be used. The conformational changes in the E-channel then propagate to the

SC

neighbouring antiporter-like subunit Nqo14, and on to distal Nqo13 and Nqo12, all through the flexible central hydrophilic axis (Fig. 4) [18]. The result of these

NU

concerted conformational changes are the changes in exposure to solvent and in pKa

MA

of key residues in the half-channels, resulting in proton translocation. However, the exact extent of such movements is currently unclear. How exactly other coupling (or

D

connecting) elements, i.e. helix HL and βH motif, are involved, also remains to be

TE

established by the determination of structures of different redox states of the complex.

AC CE P

Mutations in complex I and disease Research on complex I has increased in significance since the discovery that many human diseases, mostly neurodegenerative, involve defects at the molecular level of this enzyme complex. Such diseases include the Leigh’s syndrome,[98] encephalomyopathy and cardiomyopathy [99, 100], and the Leber’s hereditary optic neuropathy (LHON) [101-103]. LHON is considered the most common disease caused by the mtDNA mutations, specifically by point mutations in the genes encoding the membrane domain subunits ND1, ND4 and ND6 (Nqo8, Nqo13 and Nqo10 respectively in T. thermophilus) [104, 105]. Leigh’s syndrome is mostly associated with mutations in the nuclear genes encoding the PSST and TYKY (Nqo6 and Nqo9) subunits of complex I [106, 107]. Additionally, mutations in the nuclear gene encoding the 51 kDa (Nqo1) subunit of complex I, which contains the NADH-

23

ACCEPTED MANUSCRIPT binding site, lead to the neurological disorder leukodystrophy and myoclonic epilepsy [108].

PT

Due to high degree of sequence conservation in the core subunits, the availability of bacterial enzyme structure allows us to explain the molecular basis for

RI

many of human pathological mutations, as described in our earlier publications [3, 18,

SC

38]. In particular, two of the three main LHON mutations (in subunits ND4 and ND6) appear to interfere with conformational coupling mechanism and the third one (in

MA

NU

ND1) likely impedes quinone movement in and out of its cavity [97].

Conclusions

Complex I is one of the largest known membrane proteins. The recent crystal

TE

D

structure of the entire, intact complex I from Thermus thermophilus has revealed many unique features of this giant proton pump [18]. We are now beginning to

AC CE P

understand the enormous complexity of this molecular machine. The coupling between the electron transfer and proton pumping through long-range conformational changes can be aided by such unique features as tandem coordination of the terminal Fe-S cluster, enclosed quinone-binding chamber, central hydrophilic axis spanning an entire membrane domain, two additional coupling (connecting) elements, proton channels formed from two connected half-channels with lysines as central residues, etc. The result of global conformational changes in complex I is the effective translocation of four protons across the membrane per cycle, resulting in major contribution to energy production in most cell types.

24

ACCEPTED MANUSCRIPT Acknowledgements

PT

The experimental work discussed in this review was performed while authors were at the Medical Research Council Mitochondrial Biology Unit, Cambridge, UK, funded

SC

RI

by the Medical Research Council.

NU

Figure legends

MA

Fig. 1. (A) Overview of the entire complex I from T. thermophilus (PDB ID: 4HEA). Subunits are coloured differently and labelled. FMN and Fe-S clusters are shown as magenta and red-orange spheres, respectively, with cluster N2 labelled. Key helices

TE

D

(prefix indicates Nqo subunit number) around the entry point (Q) into the quinone reaction chamber, and approximate membrane position are indicated. (B) Positions of

AC CE P

the redox centers in complex I. The distances between the cofactors given in Å were calculated both center-to-center and edge-to-edge (shown in parentheses). Blue arrows show the main electron transfer pathway between FMN and quinone. Green arrow shows a diversion to cluster N1a. The positions of NADH [61] and quinone headgroup [18] are based on experimental data. The entire ubiquinone tail was modeled into the quinone-binding cavity (PDB ID: 4HEA and 3IAM).

Fig. 2. (A) The NADH binding site of T. thermophilus complex I (PDB ID: 3IAM). A view from the solvent-exposed side. FMN and the Nqo1 residues involved in NADH binding site are shown as sticks with carbon in yellow and NADH with carbon in salmon. Hydrogen bonds are represented as dotted green lines, hydrophobic stacking interactions in grey, the hydride (H−) transfer path in red and van der Waals contact

25

ACCEPTED MANUSCRIPT between E97 and C4 of NADH in grey. (B) The quinone-binding site. Hydrophilic arm subunit Nqo4 is shown in green, Nqo6 in red and hydrophobic arm subunit Nqo8

PT

in orange. Iron-sulfur cluster N2 is shown as red-orange spheres. Theoretical model of bound ubiquinone-10 is shown with the headgroup positioned according to

RI

experimental structure with decyl-ubiquinone [18] and the hydrophobic tail modelled

SC

to fit within the cavity. Carbon atom in cyan indicates the 8th isoprenoid unit. The quinone chamber is shown with surface in transparent brown and helices framing its

NU

entry point are labelled (prefix indicates Nqo subunit number). Movable helix 6_H1

MA

[61], interacting with 8_AH1, is also labelled.

Fig. 3. Proton translocation channels. Two sets of five symmetry-related helices in the

TE

D

antiporter-like subunits Nqo12, Nqo13 and Nqo14 each form an apparent half-channel for proton translocation with TM4-8 comprising the cytoplasmic half and TM9-13 the

AC CE P

periplasmic half. These are connected by conserved polar residues including Lys/HisTM8. Polar residues lining the channels are shown as sticks with carbons shown in dark blue for the first (amino-terminal) half-channel, in green for the second (carboxy-terminal) half-channel and in orange for connecting residues. Key residues for proton translocation in antiporter-like subunits, that is GluTM5 and LysTM7 from the first half-channel, Lys or HisTM8 from the connection and Lys or GluTM12 from the second half-channel, are indicated. Residues playing similar roles in the E-channel are also indicated (Glu-Asp quartet comprises Glu213, Glu163 and Glu130 from Nqo8 and Asp72 from Nqo7, labelled in red; 11_E67, 11_E32, 10_Y59 are also important for proton translocation). The quinone-binding cavity is shown in brown, with the modelled ubiquinone molecule shown in cyan and residues connecting the cavity to the E-channel shown in magenta. Previously suggested proton translocation

26

ACCEPTED MANUSCRIPT pathways are indicated by blue arrows, and additional proposed paths (new entry sites

PT

and inter-subunit transfer) by dark violet arrows.

Fig. 4. Suggested coupling mechanism of complex I. Upon electron transfer from

RI

cluster N2, negatively charged quinone initiates a cascade of conformational changes,

SC

propagating from the E-channel (Nqo8/10/11) to the antiporters via the central hydrophilic axis (red arrows). Cluster N2-driven shifts of Nqo4/6 helices [61] (blue

NU

arrows) likely assist overall conformational changes. Helix HL and the βH element

MA

help coordinate conformational changes by linking discontinuous TM helices between the antiporters. In the antiporters, LysTM7 from the first half-channel is assumed to

D

be protonated (via the link to cytoplasm) in the oxidised state.[38] Upon reduction of

TE

quinone and subsequent conformational change, the first half-channel closes to the cytoplasm, GluTM5 moves out and LysTM7 donates its proton to the connecting

AC CE P

Lys/HisTM8 and then onto Lys/GluTM12 from the second half-channel. Lys/GluTM12 ejects its proton into periplasm upon return from reduced to oxidised state. A fourth proton per cycle is translocated in the E-channel in a similar manner. TM helices are numbered and key charged residues (GluTM5, LysTM7, Lys/GluTM12, Lys/HisTM8 from Nqo12-14, 11_Glu67, 11_Glu32, interacting with 10_Tyr59, 8_Glu213 and some residues from the connection to Q cavity) are indicated by red circles for Glu and blue circles for Lys/His. Subunit names are shown for T. thermophilus, E. coli and bovine.

27

ACCEPTED MANUSCRIPT Table 1. Nomenclature of the core subunits of complex I. “Nuo” nomenclature originates from “NADH:Ubiquinone Oxidoreductase”, “Nqo” – from “NADH:Quinone Oxidoreductase” and

SwissProt Homo

taurus

code

sapiens

E. coli

T. thermophilus

(R.

(P.denitrificans,

capsulatus)

A. aeolicus)

Molecular

Cofactors

mass

RI

Bos

PT

“ND” – from “NADH dehydrogenase”.

SC

(T.

NDUFS1

NUBM

NDUFV1

24 kDa

NUHM

Connecting domain

30 kDa TYKY PSST

NUCM

86.5

N1b, N4, N5, (N7)a

Nqo1

48.6

FMN, N3

NDUFV2

NuoE

Nqo2

20.3

N1a

NuoD

Nqo4

46.3

NDUFS2

AC CE P

49 kDa

Nqo3

NuoF

D

51 kDa

NuoG

MA

NUAM

TE

75 kDa

NU

Dehydrogenase domain

thermophilus)

(NuoCDb)

NUGM

NDUFS3

NuoC

Nqo5

23.8

NUIM

NDUFS8

NuoI

Nqo9

20.1

N6a, N6b

NUKM

NDUFS7

NuoB

Nqo6

20.2

N2

Membrane domain ND1

NU1M

ND1

NuoH

Nqo8

41.0

ND2

NU2M

ND2

NuoN

Nqo14

44.9

ND3

NU3M

ND3

NuoA

Nqo7

13.1

ND4

NU4M

ND4

NuoM

Nqo13

49.4

ND4L

NU4LM

ND4L

NuoK

Nqo11

10.0

28

ACCEPTED MANUSCRIPT ND5

NU5M

ND5

NuoL

Nqo12

65.2

ND6

NU6M

ND6

NuoJ

Nqo10

18.4

b

PT

Cluster N7 is present only in some bacteria (E. coli, T. thermophilus, etc.). Subunits NuoC (30 kDa) and NuoD (49 kDa) are fused in E. coli and some other bacteria.

RI

a

SC

References

AC CE P

TE

D

MA

NU

[1] J.E. Walker, The NADH - ubiquinone oxidoreductase (complex I) of respiratory chains, Q. Rev. Biophys., 25 (1992) 253-324. [2] T. Yagi, A. Matsuno-Yagi, The proton-translocating NADH-Quinone oxidoreductase in the respiratory chain: the secret unlocked, Biochemistry, 42 (2003) 2266-2274. [3] L.A. Sazanov, Respiratory complex I: mechanistic and structural insights provided by the crystal structure of the hydrophilic domain, Biochemistry, 46 (2007) 22752288. [4] U. Brandt, Energy converting NADH:quinone oxidoreductase (complex I), Annu. Rev. Biochem., 75 (2006) 69-92. [5] A.D. Vinogradov, Catalytic properties of the mitochondrial NADH-ubiquinone oxidoreductase (complex I) and the pseudo-reversible active/inactive enzyme transition, Biochim. Biophys. Acta, 1364 (1998) 169-185. [6] M. Wikstrom, Two protons are pumped from the mitochondrial matrix per electron transferred between NADH and ubiquinone, FEBS Lett., 169 (1984) 300304. [7] A.S. Galkin, V.G. Grivennikova, A.D. Vinogradov, H+/2e- stoichiometry in NADH-quinone reductase reactions catalyzed by bovine heart submitochondrial particles, FEBS Lett., 451 (1999) 157-161. [8] A. Galkin, S. Drose, U. Brandt, The proton pumping stoichiometry of purified mitochondrial complex I reconstituted into proteoliposomes, Biochim. Biophys. Acta, 1757 (2006) 1575-1581. [9] A.V. Bogachev, R.A. Murtazina, V.P. Skulachev, H+/e- stoichiometry for NADH dehydrogenase I and dimethyl sulfoxide reductase in anaerobically grown Escherichia coli cells, J. Bacteriol., 178 (1996) 6233-6237. [10] K.R. Vinothkumar, J. Zhu, J. Hirst, Architecture of mammalian respiratory complex I, Nature, (2014). [11] J. Carroll, I.M. Fearnley, R.J. Shannon, J. Hirst, J.E. Walker, Analysis of the subunit composition of complex I from bovine heart mitochondria, Mol. Cell. Proteomics, 2 (2003) 117-126. [12] J. Carroll, I.M. Fearnley, J.M. Skehel, R.J. Shannon, J. Hirst, J.E. Walker, Bovine complex I is a complex of 45 different subunits, J. Biol. Chem., 281 (2006) 32724-32727. [13] S. Kerscher, L. Grgic, A. Garofano, U. Brandt, Application of the yeast Yarrowia lipolytica as a model to analyse human pathogenic mutations in mitochondrial complex I (NADH:ubiquinone oxidoreductase), Biochim. Biophys. Acta, 1659 (2004) 197-205.

29

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

[14] C. Hunte, V. Zickermann, U. Brandt, Functional modules and structural basis of conformational coupling in mitochondrial complex I, Science, 329 (2010) 448-451. [15] A. Videira, Complex I from the fungus Neurospora crassa, Biochim. Biophys. Acta, 1364 (1998) 89-100. [16] P. Hinchliffe, J. Carroll, L.A. Sazanov, Identification of a novel subunit of respiratory complex I from Thermus thermophilus, Biochemistry, 45 (2006) 44134420. [17] D. Schneider, T. Pohl, J. Walter, K. Dorner, M. Kohlstadt, A. Berger, V. Spehr, T. Friedrich, Assembly of the Escherichia coli NADH:ubiquinone oxidoreductase (complex I), Biochim. Biophys. Acta, 1777 (2008) 735-739. [18] R. Baradaran, J.M. Berrisford, G.S. Minhas, L.A. Sazanov, Crystal structure of the entire respiratory complex I, Nature, 494 (2013) 443-448. [19] H. Leif, V.D. Sled, T. Ohnishi, H. Weiss, T. Friedrich, Isolation and characterization of the proton-translocating NADH: ubiquinone oxidoreductase from Escherichia coli, Eur. J. Biochem., 230 (1995) 538-548. [20] S.W. Meinhardt, D.C. Wang, K. Hon-nami, T. Yagi, T. Oshima, T. Ohnishi, Studies on the NADH-menaquinone oxidoreductase segment of the respiratory chain in Thermus thermophilus HB-8, J. Biol. Chem., 265 (1990) 1360-1368. [21] T. Yano, J. Sklar, E. Nakamaru-Ogiso, Y. Takahashi, T. Yagi, T. Ohnishi, Characterization of cluster N5 as a fast-relaxing [4Fe-4S] cluster in the Nqo3 subunit of the proton-translocating NADH-ubiquinone oxidoreductase from Paracoccus denitrificans, J. Biol. Chem., 278 (2003) 15514-15522. [22] C.Y. Yip, M.E. Harbour, K. Jayawardena, I.M. Fearnley, L.A. Sazanov, Evolution of respiratory complex I: "supernumerary" subunits are present in the alpha-proteobacterial enzyme, J. Biol. Chem., 286 (2011) 5023-5033. [23] G. Peng, G. Fritzsch, V. Zickermann, H. Schagger, R. Mentele, F. Lottspeich, M. Bostina, M. Radermacher, R. Huber, K.O. Stetter, H. Michel, Isolation, characterization and electron microscopic single particle analysis of the NADH:ubiquinone oxidoreductase (complex I) from the hyperthermophilic eubacterium Aquifex aeolicus, Biochemistry, 42 (2003) 3032-3039. [24] T. Yagi, T. Yano, S. DiBernardo, A. MatsunoYagi, Procaryotic complex I (NDH-1), an overview, Biochim. Biophys. Acta, 1364 (1998) 125-133. [25] D.J. Morgan, L.A. Sazanov, Three-dimensional structure of respiratory complex I from Escherichia coli in ice in the presence of nucleotides, Biochim. Biophys. Acta, 1777 (2008) 711-718. [26] V. Guenebaut, A. Schlitt, H. Weiss, K. Leonard, T. Friedrich, Consistent structure between bacterial and mitochondrial NADH:ubiquinone oxidoreductase (complex I), J. Mol. Biol., 276 (1998) 105-112. [27] T. Friedrich, D. Scheide, The respiratory complex I of bacteria, archaea and eukarya and its module common with membrane-bound multisubunit hydrogenases, FEBS Lett., 479 (2000) 1-5. [28] L.A. Sazanov, P.A. Burrows, P.J. Nixon, The plastid ndh genes code for an NADH-specific dehydrogenase: isolation of a complex I analogue from pea thylakoid membranes, Proc. Natl. Acad. Sci. USA, 95 (1998) 1319-1324. [29] A.H. Schapira, Human complex I defects in neurodegenerative diseases, Biochim. Biophys. Acta, 1364 (1998) 261-270. [30] R.S. Balaban, S. Nemoto, T. Finkel, Mitochondria, oxidants, and aging, Cell, 120 (2005) 483-495. [31] T.M. Dawson, V.L. Dawson, Molecular pathways of neurodegeneration in Parkinson's disease, Science, 302 (2003) 819-822.

30

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

[32] L.A. Sazanov, J. Carroll, P. Holt, L. Toime, I.M. Fearnley, A role for native lipids in the stabilization and two-dimensional crystallization of the Escherichia coli NADH-ubiquinone oxidoreductase (Complex I), J. Biol. Chem., 278 (2003) 1948319491. [33] Q.H. Tran, J. Bongaerts, D. Vlad, G. Unden, Requirement for the protonpumping NADH dehydrogenase I of Escherichia coli in respiration of NADH to fumarate and its bioenergetic implications, Eur. J. Biochem., 244 (1997) 155-160. [34] P.J. Holt, D.J. Morgan, L.A. Sazanov, The location of NuoL and NuoM subunits in the membrane domain of the Escherichia coli complex I: implications for the mechanism of proton pumping, J. Biol. Chem., 278 (2003) 43114-43120. [35] T. Yano, S.S. Chu, V.D. Sled, T. Ohnishi, T. Yagi, The proton-translocating NADH-quinone oxidoreductase (NDH-1) of thermophilic bacterium Thermus thermophilus HB-8. Complete DNA sequence of the gene cluster and thermostable properties of the expressed NQO2 subunit, J. Biol. Chem., 272 (1997) 4201-4211. [36] T. Ohnishi, Iron-sulfur clusters/semiquinones in complex I, Biochim. Biophys. Acta, 1364 (1998) 186-206. [37] P. Hinchliffe, L.A. Sazanov, Organization of iron-sulfur clusters in respiratory complex I, Science, 309 (2005) 771-774. [38] R.G. Efremov, L.A. Sazanov, Structure of the membrane domain of respiratory complex I, Nature, 476 (2011) 414-420. [39] V. Zickermann, C. Wirth, H. Nasiri, K. Siegmund, H. Schwalbe, C. Hunte, U. Brandt, Structural biology. Mechanistic insight from the crystal structure of mitochondrial complex I, Science, 347 (2015) 44-49. [40] P.L. Dutton, C.C. Moser, V.D. Sled, F. Daldal, T. Ohnishi, A reductant-induced oxidation mechanism for complex I, Biochim. Biophys. Acta, 1364 (1998) 245-257. [41] S.T. Ohnishi, J.C. Salerno, T. Ohnishi, Possible roles of two quinone molecules in direct and indirect proton pumps of bovine heart NADH-quinone oxidoreductase (complex I), Biochim. Biophys. Acta, 1797 (2010) 1891-1893. [42] T. Friedrich, Complex I: a chimaera of a redox and conformation-driven proton pump?, J. Bioenerg. Biomembr., 33 (2001) 169-177. [43] R.G. Efremov, R. Baradaran, L.A. Sazanov, The architecture of respiratory complex I, Nature, 465 (2010) 441-445. [44] L.A. Sazanov, S.Y. Peak-Chew, I.M. Fearnley, J.E. Walker, Resolution of the membrane domain of bovine complex I into subcomplexes: implications for the structural organization of the enzyme, Biochemistry, 39 (2000) 7229-7235. [45] I.M. Fearnley, J. Carroll, R.J. Shannon, M.J. Runswick, J.E. Walker, J. Hirst, Grim-19, a cell death regulatory gene product, is a subunit of bovine mitochondrial nadh:ubiquinone oxidoreductase (complex I), J. Biol. Chem., 276 (2001) 3834538348. [46] Y. Hatefi, Preparation and properties of NADH: ubiquinone oxidoreductase (complex I), EC 1.6.5.3, Methods Enzymol., 53 (1978) 11-14. [47] M. Finel, J.M. Skehel, S.P. Albracht, I.M. Fearnley, J.E. Walker, Resolution of NADH:ubiquinone oxidoreductase from bovine heart mitochondria into two subcomplexes, one of which contains the redox centers of the enzyme, Biochemistry, 31 (1992) 11425-11434. [48] M.S. Sharpley, R.J. Shannon, F. Draghi, J. Hirst, Interactions between phospholipids and NADH:ubiquinone oxidoreductase (complex I) from bovine mitochondria, Biochemistry, 45 (2006) 241-248.

31

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

[49] S. Kerscher, S. Drose, K. Zwicker, V. Zickermann, U. Brandt, Yarrowia lipolytica, a yeast genetic system to study mitochondrial complex I, Biochim. Biophys. Acta, 1555 (2002) 83-91. [50] V. Spehr, A. Schlitt, D. Scheide, V. Guenebaut, T. Friedrich, Overexpression of the Escherichia coli nuo-operon and isolation of the overproduced NADH:ubiquinone oxidoreductase (Complex I), Biochemistry, 38 (1999) 16261-16267. [51] A.A. Mamedova, P.J. Holt, J. Carroll, L.A. Sazanov, Substrate-induced conformational change in bacterial complex I, J. Biol. Chem., 279 (2004) 2383023836. [52] E. Schafer, H. Seelert, N.H. Reifschneider, F. Krause, N.A. Dencher, J. Vonck, Architecture of active mammalian respiratory chain supercomplexes, J. Biol. Chem., 281 (2006) 15370-15375. [53] T. Althoff, D.J. Mills, J.L. Popot, W. Kuhlbrandt, Arrangement of electron transport chain components in bovine mitochondrial supercomplex I1III2IV1, EMBO J., 30 (2011) 4652-4664. [54] T. Rasmussen, D. Scheide, B. Brors, L. Kintscher, H. Weiss, T. Friedrich, Identification of two tetranuclear FeS clusters on the ferredoxin-type subunit of NADH:ubiquinone oxidoreductase (complex I), Biochemistry, 40 (2001) 6124-6131. [55] L. Euro, D.A. Bloch, M. Wikstrom, M.I. Verkhovsky, M. Verkhovskaya, Electrostatic interactions between FeS clusters in NADH:ubiquinone oxidoreductase (Complex I) from Escherichia coli, Biochemistry, 47 (2008) 3185-3193. [56] H.R. Bridges, E. Bill, J. Hirst, Mossbauer spectroscopy on respiratory complex I: the iron-sulfur cluster ensemble in the NADH-reduced enzyme is partially oxidized, Biochemistry, 51 (2012) 149-158. [57] L.A. Sazanov, P. Hinchliffe, Structure of the hydrophilic domain of respiratory complex I from Thermus thermophilus, Science, 311 (2006) 1430-1436. [58] G. Belevich, L. Euro, M. Wikstrom, M. Verkhovskaya, Role of the conserved arginine 274 and histidine 224 and 228 residues in the NuoCD subunit of complex I from Escherichia coli, Biochemistry, 46 (2007) 526-533. [59] G. Yakovlev, T. Reda, J. Hirst, Reevaluating the relationship between EPR spectra and enzyme structure for the iron sulfur clusters in NADH:quinone oxidoreductase, Proc Natl Acad Sci U S A, 104 (2007) 12720-12725. [60] M.M. Roessler, M.S. King, A.J. Robinson, F.A. Armstrong, J. Harmer, J. Hirst, Direct assignment of EPR spectra to structurally defined iron-sulfur clusters in complex I by double electron-electron resonance, Proc Natl Acad Sci U S A, 107 (2010) 1930-1935. [61] J.M. Berrisford, L.A. Sazanov, Structural basis for the mechanism of respiratory complex I, J. Biol. Chem., 284 (2009) 29773-29783. [62] V. Guenebaut, R. Vincentelli, D. Mills, H. Weiss, K.R. Leonard, Threedimensional structure of NADH-dehydrogenase from Neurospora crassa by electron microscopy and conical tilt reconstruction, J. Mol. Biol., 265 (1997) 409-418. [63] M. Radermacher, T. Ruiz, T. Clason, S. Benjamin, U. Brandt, V. Zickermann, The three-dimensional structure of complex I from Yarrowia lipolytica: a highly dynamic enzyme, J. Struct. Biol., 154 (2006) 269-279. [64] N. Grigorieff, Three-dimensional structure of bovine NADH:ubiquinone oxidoreductase (complex I) at 22 A in ice, J. Mol. Biol., 277 (1998) 1033-1046. [65] I.M. Fearnley, J.E. Walker, Conservation of sequences of subunits of mitochondrial complex I and their relationships with other proteins, Biochim. Biophys. Acta, 1140 (1992) 105-134.

32

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

[66] C. Mathiesen, C. Hagerhall, Transmembrane topology of the NuoL, M and N subunits of NADH:quinone oxidoreductase and their homologues among membranebound hydrogenases and bona fide antiporters, Biochim. Biophys. Acta, 1556 (2002) 121-132. [67] G. Belevich, J. Knuuti, M.I. Verkhovsky, M. Wikstrom, M. Verkhovskaya, Probing the mechanistic role of the long alpha-helix in subunit L of respiratory Complex I from Escherichia coli by site-directed mutagenesis, Molecular microbiology, 82 (2011) 1086-1095. [68] S. Steimle, C. Schnick, E.M. Burger, F. Nuber, D. Kramer, H. Dawitz, S. Brander, B. Matlosz, J. Schafer, K. Maurer, U. Glessner, T. Friedrich, Cysteine scanning reveals minor local rearrangements of the horizontal helix of respiratory complex I, Mol Microbiol, 98 (2015) 151-161. [69] S. Zhu, S.B. Vik, Constraining the Lateral Helix of Respiratory Complex I by Cross-linking Does Not Impair Enzyme Activity or Proton Translocation, J Biol Chem, 290 (2015) 20761-20773. [70] L.A. Sazanov, The mechanism of coupling between electron transfer and proton translocation in respiratory complex I, J Bioenerg Biomembr, (2014). [71] P.M. Vignais, B. Billoud, J. Meyer, Classification and phylogeny of hydrogenases, FEMS Microbiol. Rev., 25 (2001) 455-501. [72] J.I. Oh, B. Bowien, Structural analysis of the fds operon encoding the NAD+linked formate dehydrogenase of Ralstonia eutropha, J. Biol. Chem., 273 (1998) 26349-26360. [73] R.G. Efremov, L.A. Sazanov, The coupling mechanism of respiratory complex I - A structural and evolutionary perspective, Biochim. Biophys. Acta, (2012). [74] T. Pohl, T. Bauer, K. Dorner, S. Stolpe, P. Sell, G. Zocher, T. Friedrich, Ironsulfur cluster N7 of the NADH:ubiquinone oxidoreductase (complex I) is essential for stability but not involved in electron transfer, Biochemistry, 46 (2007) 6588-6596. [75] S.J. Cho, M.G. Lee, J.K. Yang, J.Y. Lee, H.K. Song, S.W. Suh, Crystal structure of Escherichia coli CyaY protein reveals a previously unidentified fold for the evolutionarily conserved frataxin family, Proc. Natl. Acad. Sci. U S A, 97 (2000) 8932-8937. [76] S. Dhe-Paganon, R. Shigeta, Y.I. Chi, M. Ristow, S.E. Shoelson, Crystal structure of human frataxin, J. Biol. Chem., 275 (2000) 30753-30756. [77] O. Gakh, J. Adamec, A.M. Gacy, R.D. Twesten, W.G. Owen, G. Isaya, Physical evidence that yeast frataxin is an iron storage protein, Biochemistry, 41 (2002) 67986804. [78] S. Adinolfi, C. Iannuzzi, F. Prischi, C. Pastore, S. Iametti, S.R. Martin, F. Bonomi, A. Pastore, Bacterial frataxin CyaY is the gatekeeper of iron-sulfur cluster formation catalyzed by IscS, Nat. Struct. Mol. Biol., 16 (2009) 390-396. [79] Y. He, S.L. Alam, S.V. Proteasa, Y. Zhang, E. Lesuisse, A. Dancis, T.L. Stemmler, Yeast frataxin solution structure, iron binding, and ferrochelatase interaction, Biochemistry, 43 (2004) 16254-16262. [80] C.C. Page, C.C. Moser, X. Chen, P.L. Dutton, Natural engineering principles of electron tunnelling in biological oxidation-reduction, Nature, 402 (1999) 47-52. [81] C.C. Moser, T.A. Farid, S.E. Chobot, P.L. Dutton, Electron tunneling chains of mitochondria, Biochim. Biophys. Acta, 1757 (2006) 1096-1109. [82] A.B. Kotlyar, V.D. Sled, D.S. Burbaev, I.A. Moroz, A.D. Vinogradov, Coupling site I and the rotenone-sensitive ubisemiquinone in tightly coupled submitochondrial particles, FEBS Lett., 264 (1990) 17-20.

33

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

[83] M.L. Verkhovskaya, N. Belevich, L. Euro, M. Wikstrom, M.I. Verkhovsky, Real-time electron transfer in respiratory complex I, Proc. Natl. Acad. Sci. U S A, 105 (2008) 3763-3767. [84] S. de Vries, K. Dorner, M.J. Strampraad, T. Friedrich, Electron tunneling rates in respiratory complex I are tuned for efficient energy conversion, Angew Chem Int Ed Engl, 54 (2015) 2844-2848. [85] L. Kussmaul, J. Hirst, The mechanism of superoxide production by NADH:ubiquinone oxidoreductase (complex I) from bovine heart mitochondria, Proc. Natl. Acad. Sci. U S A, 103 (2006) 7607-7612. [86] A.P. Kudin, N.Y. Bimpong-Buta, S. Vielhaber, C.E. Elger, W.S. Kunz, Characterization of superoxide-producing sites in isolated brain mitochondria, J. Biol. Chem., 279 (2004) 4127-4135. [87] A.J. Lambert, M.D. Brand, Inhibitors of the quinone-binding site allow rapid superoxide production from mitochondrial NADH:ubiquinone oxidoreductase (complex I), J. Biol. Chem., 279 (2004) 39414-39420. [88] V.D. Sled, N.I. Rudnitzky, Y. Hatefi, T. Ohnishi, Thermodynamic analysis of flavin in mitochondrial NADH:ubiquinone oxidoreductase (complex I), Biochemistry, 33 (1994) 10069-10075. [89] S. Ransac, C. Arnarez, J.P. Mazat, The flitting of electrons in complex I: A stochastic approach, Biochim. Biophys. Acta, 1797 (2010) 641-648. [90] K.R. Pryde, J. Hirst, Superoxide is produced by the reduced flavin in mitochondrial complex I: a single, unified mechanism that applies during both forward and reverse electron transfer, J Biol Chem, 286 (2011) 18056-18065. [91] R.F. Anderson, R. Hille, S.S. Shinde, G. Cecchini, Electron transfer within complex II. Succinate:ubiquinone oxidoreductase of Escherichia coli, J. Biol. Chem., 280 (2005) 33331-33337. [92] J.A. Birrell, K. Morina, H.R. Bridges, T. Friedrich, J. Hirst, Investigating the function of [2Fe-2S] cluster N1a, the off-pathway cluster in complex I, by manipulating its reduction potential, Biochem J., 456 (2013) 139-146. [93] K.R. Vinothkumar, R. Henderson, Structures of membrane proteins, Q. Rev. Biophys., 43 (2010) 65-158. [94] E. Screpanti, C. Hunte, Discontinuous membrane helices in transport proteins and their correlation with function, J. Struct. Biol., 159 (2007) 261-267. [95] R.B. Cooley, D.J. Arp, P.A. Karplus, Evolutionary origin of a secondary structure: pi-helices as cryptic but widespread insertional variations of alpha-helices that enhance protein functionality, J. Mol. Biol., 404 (2010) 232-246. [96] R.G. Efremov, L.A. Sazanov, Structure of the membrane domain of respiratory complex I, Nature, (2011). [97] L.A. Sazanov, A giant molecular proton pump: structure and mechanism of respiratory complex I, Nature reviews. Molecular cell biology, 16 (2015) 375-388. [98] T. Fujii, M. Ito, T. Okuno, K. Mutoh, R. Nishikomori, H. Mikawa, Complex I (reduced nicotinamide-adenine dinucleotide-coenzyme Q reductase) deficiency in two patients with probable Leigh syndrome, J Pediatr, 116 (1990) 84-87. [99] H. Bentlage, R. de Coo, H. ter Laak, R. Sengers, F. Trijbels, W. Ruitenbeek, W. Schlote, K. Pfeiffer, S. Gencic, G. von Jagow, et al., Human diseases with defects in oxidative phosphorylation. 1. Decreased amounts of assembled oxidative phosphorylation complexes in mitochondrial encephalomyopathies, Eur J Biochem, 227 (1995) 909-915. [100] M. Nishizawa, K. Tanaka, K. Shinozawa, T. Kuwabara, T. Atsumi, T. Miyatake, E. Ohama, A mitochondrial encephalomyopathy with cardiomyopathy. A

34

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

case revealing a defect of complex I in the respiratory chain, J Neurol Sci, 78 (1987) 189-201. [101] X. Qi, A.S. Lewin, W.W. Hauswirth, J. Guy, Suppression of complex I gene expression induces optic neuropathy, Ann Neurol, 53 (2003) 198-205. [102] M.L. Valentino, P. Avoni, P. Barboni, F. Pallotti, C. Rengo, A. Torroni, M. Bellan, A. Baruzzi, V. Carelli, Mitochondrial DNA nucleotide changes C14482G and C14482A in the ND6 gene are pathogenic for Leber's hereditary optic neuropathy, Ann Neurol, 51 (2002) 774-778. [103] M.L. Valentino, P. Barboni, A. Ghelli, L. Bucchi, C. Rengo, A. Achilli, A. Torroni, A. Lugaresi, R. Lodi, B. Barbiroli, M. Dotti, A. Federico, A. Baruzzi, V. Carelli, The ND1 gene of complex I is a mutational hot spot for Leber's hereditary optic neuropathy, Ann Neurol, 56 (2004) 631-641. [104] P.F. Chinnery, D.T. Brown, R.M. Andrews, R. Singh-Kler, P. Riordan-Eva, J. Lindley, D.A. Applegarth, D.M. Turnbull, N. Howell, The mitochondrial ND6 gene is a hot spot for mutations that cause Leber's hereditary optic neuropathy, Brain, 124 (2001) 209-218. [105] P.Y. Man, P.G. Griffiths, D.T. Brown, N. Howell, D.M. Turnbull, P.F. Chinnery, The epidemiology of Leber hereditary optic neuropathy in the North East of England, Am J Hum Genet, 72 (2003) 333-339. [106] J. Loeffen, J. Smeitink, R. Triepels, R. Smeets, M. Schuelke, R. Sengers, F. Trijbels, B. Hamel, R. Mullaart, L. van den Heuvel, The first nuclear-encoded complex I mutation in a patient with Leigh syndrome, Am J Hum Genet, 63 (1998) 1598-1608. [107] R.H. Triepels, L.P. van den Heuvel, J.L. Loeffen, C.A. Buskens, R.J. Smeets, M.E. Rubio Gozalbo, S.M. Budde, E.C. Mariman, F.A. Wijburg, P.G. Barth, J.M. Trijbels, J.A. Smeitink, Leigh syndrome associated with a mutation in the NDUFS7 (PSST) nuclear encoded subunit of complex I, Ann Neurol, 45 (1999) 787-790. [108] M. Schuelke, J. Smeitink, E. Mariman, J. Loeffen, B. Plecko, F. Trijbels, S. Stockler-Ipsiroglu, L. van den Heuvel, Mutant NDUFV1 subunit of mitochondrial complex I causes leukodystrophy and myoclonic epilepsy, Nat Genet, 21 (1999) 260261.

35

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

36

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

37

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

38

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

39

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

Graphical abstract

40

ACCEPTED MANUSCRIPT Highlights Function and general architecture/organisation of respiratory complex I is described.

PT

Structure of bacterial complex I, representing the “core” of the enzyme, is described. Unusual binding sites for substrates, NADH and quinone, are discussed.

RI

Electron transfer pathway and proton translocation channels are discussed in detail.

AC CE P

TE

D

MA

NU

SC

Putative mechanism of coupling between electron transfer and proton translocation is discussed.

41