Modulators of oxidase Implications for the regulation of mitochondrial respiration Camilla Rydström Lundin Academic dissertation for the Degree of Doctor of Philosophy in Biochemistry at Stockholm University to be publicly defended on Thursday 13 December 2018 at 10.00 in Magnélisalen, Kemiska övningslaboratoriet, Svante Arrhenius väg 16 B.

Abstract Oxidative phosphorylation in mitochondria is performed by complexes and electron carriers that reside in the inner membrane. It is now generally accepted that these respiratory enzyme complexes assemble into larger so-called supercomplexes. However, it is presently not known why, under which conditions or how these supercomplexes form. A number of factors of particular importance for the formation of supercomplexes have been identified, such as the Respiratory supercomplex factors (Rcf1 and Rcf2) and . The work presented in this thesis is focused on the characterization of (CytcO) in mitochondria from Saccharomyces cerevisiae strains in which these components have been removed, with a particular focus on Rcf1. First, we concluded that Rcf1 has an impact on the activity and ligand binding kinetics of CytcO, which upon genetic deletion of rcf1 leads to formation of sub-populations of CytcO with different functionality. Second, we noted that the ability of CytcO to oxidize cytochrome c (cyt. c) depends on the presence of Rcf1. Further, we observed that while CytcO in wild-type mitochondria displayed differences in the oxidation kinetics of cyt. c from horse or S. cerevisiae, with the Δrcf1 mitochondria these differences were lost. This observation suggested that Rcf1 interacts with cyt. c. Furthermore, the data showed that in CytcO from Δrcf1 mitochondria heme a3 was altered while heme a was intact. Using proteo-liposomes of different composition and size we also investigated the influence of lipid head groups on the coupled activity of a quinol oxidase and ATP-synthase. Specifically, we addressed the question if protons are transferred between proton “producers” and “consumers” via lateral proton transfer along the membrane surface or via bulk water. Our data supported the principle of lateral proton transfer. Lastly, we characterized the ligand binding of flavohemoglobin and concluded that the flavohemoglobin has a population that resides in the of mitochondria, not only in matrix and cytosol as previously suggested.

Keywords: cytochrome c oxidase, cytochrome c, OXPHOS, membrane , kinetics, ligand-binding, electron transfer, Rcf1, respiratory supercomplexes, Saccharomyces cerevisiae.

Stockholm 2018 http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-161515

ISBN 978-91-7797-457-4 ISBN 978-91-7797-456-7

Department of Biochemistry and Biophysics

Stockholm University, 106 91 Stockholm

MODULATORS OF SACCHAROMYCES CEREVISIAE CYTOCHROME C OXIDASE

Camilla Rydström Lundin

Modulators of Saccharomyces cerevisiae cytochrome c oxidase

Implications for the regulation of mitochondrial respiration

Camilla Rydström Lundin ©Camilla Rydström Lundin, Stockholm University 2018

ISBN print 978-91-7797-457-4 ISBN PDF 978-91-7797-456-7

Cover image is a schematic of a membrane embedded cytochrome c oxidase.

Printed in Sweden by Universitetsservice US-AB, Stockholm 2018 Min familj, utan er vore jag vilse

List of publications

This thesis is based on the following publications.

(I) Lipid-mediated Protein-protein Interactions Modulate Respiration-driven ATP synthesis Nilsson T, Lundin CR, Nordlund G, Ädelroth P, von Ballmoos C, Brzezinski P Sci. Rep. (2016) 6:24113

(II) Regulatory Role of the Respiratory Supercomplex Factors in Saccharomyces cerevisiae Rydström Lundin C, von Ballmoos C, Ott M, Ädelroth P, Brzezinski P Proc. Natl. Acad. Sci. U S A. (2016) 113:4476-4485

(III) Modulation of the O2-reduction Site in Saccharomyces cerevisiae Cytochrome c Oxidase Rydström Lundin C and Brzezinski P FEBS Lett. (2017) 591:4049-4055

(IV) Reactions of S. cerevisiae mitochondria ligands:

Kinetics of CO and O2-binding to Flavohemoglobin and Cytochrome c Oxidase Björck ML, Zhou S, Rydström Lundin C, Ott M, Ädelroth P, Brzezinski P Biochim. Biophys. Acta. (2017) 1858:182-188

Reprints are made with permission of the publishers.

Abstract

Oxidative phosphorylation in mitochondria is performed by enzyme complexes and electron carriers that reside in the inner membrane. It is now generally accepted that these respiratory enzyme complexes assemble into larger so-called supercomplexes. However, it is presently not known why, under which conditions or how these supercomplexes form.

A number of factors of particular importance for the formation of supercomplexes have been identified, such as the Respiratory supercomplex factors (Rcf1 and Rcf2) and cardiolipin. The work presented in this thesis is focused on the characterization of cytochrome c oxidase (CytcO) in mitochondria from Saccharomyces cerevisiae strains in which these components have been removed, with a particular focus on Rcf1. First, we concluded that Rcf1 has an impact on the activity and ligand binding kinetics of CytcO, which upon genetic deletion of rcf1 leads to formation of sub-populations of CytcO with different functionality. Second, we noted that the ability of CytcO to oxidize cytochrome c (cyt. c) depends on the presence of Rcf1. Further, we observed that while CytcO in wild-type mitochondria displayed differences in the oxidation kinetics of cyt. c from horse heart or S. cerevisiae, with the rcf1 mitochondria these differences were lost. This observation suggested that Rcf1 interacts with cyt. c. Furthermore, the data showed that in CytcO from rcf1 mitochondria heme a3 was altered while heme a was intact.

Using proteo-liposomes of different lipid composition and size we also investigated the influence of lipid head groups on the coupled activity of a quinol oxidase and ATP-synthase. Specifically, we addressed the question if protons are transferred between proton “producers” and “consumers” via lateral proton transfer along the membrane surface or via bulk water. Our data supported the principle of lateral proton transfer.

Lastly, we characterized the ligand binding of yeast flavohemoglobin and concluded that the flavohemoglobin has a population that resides in the intermembrane space of mitochondria, not only in matrix and cytosol as previously suggested.

1

Table of Contents ABSTRACT ...... 1

ABBREVIATIONS ...... 4

1 MAIN FINDINGS ...... 5

2 CELLULAR ENERGY CONVERSION ...... 7

2.1 THE ...... 7 2.2 OXIDATIVE PHOSPHORYLATION ...... 7 2.3 TO RESPIRE OR FERMENT, THAT IS THE QUESTION ...... 8

3 HEME, HEME AND YEAST STRAINS ...... 10

4 CYTOCHROME C OXIDASE ...... 12

4.1 REACTION WITH OXYGEN ...... 12 4.2 FLASH PHOTOLYSIS AND FLOW FLASH ...... 13 4.3 LIGAND BINDING ...... 15

4.3.1 Hemes a and a3 ...... 15 4.3.2 CO-binding to other hemes ...... 17 4.4 STRUCTURE AND FUNCTION OF CYTOCHROME C OXIDASE ...... 18 4.4.1 Cox1 ...... 19 4.4.2 Cox2 ...... 19 4.4.3 Cox3 ...... 20 4.4.4 Cox5 ...... 20 4.4.5 Cox12 (VIb) ...... 22 4.4.6 Cox13 (VIa) ...... 23 4.4.7 Populations of CytcO ...... 23

5 CYTOCHROME C ...... 24

5.1 CYTOCHROME C AND RCF1 ...... 25 5.2 CYTOCHROME C IN OXIDATIVE FOLDING ...... 27

6 THE MEMBRANE ...... 28

6.1 CARDIOLIPIN ...... 28 6.2 LATERAL PROTON TRANSFER ...... 29

7 RESPIRATORY SUPERCOMPLEXES ...... 31

7.1 FACTORS RELATED TO SUPERCOMPLEXES ...... 32 7.1.1 Rcf1 ...... 32 7.1.2 Rcf2 ...... 33 7.1.3 Aac2 ...... 33 7.1.4 Cardiolipin ...... 34

8 QUESTIONS AND SPECULATIONS ...... 35

8.1 HEME SYNTHESIS AND DISTRIBUTION OF RESPIRATORY ...... 35 8.2 THE MIA40 PATHWAY, OXPHOS AND THE TWO-FACED CYT. C ...... 35 8.3 RCF1, CYT. C AND COX5 ...... 35

2

9 POPULÄRVETENSKAPLIG SAMMANFATTNING ...... 36

ACKNOWLEDGEMENTS ...... 38

REFERENCES ...... 40

3

Abbreviations

Aac2 ADP-ATP translocator/carrier CI Complex I, NADH: Ubiquinone oxidoreductase CII Complex II, CIII Complex III, Coenzyme Q-cytochrome c oxidoreductase, cyt. bc1 CIV Complex IV, Cytochrome c oxidase, CytcO CV Complex V, ATP synthase CL Cardiolipin CytcO Cytochrome c oxidase DOPA Dioleoylphosphatidic acid DOPC Dioleoylphosphatidylcholine DOPE Dioleoylphosphatidylethanolamine DOPG Dioleoylphosphatidylglycerol HIGD/Higd Hypoxia-induced domain IDP Intrinsically disordered peptide/protein IMM Inner mitochondrial membrane IMS Inter membrane space OM Outer membrane OXPHOS Oxidative phosphorylation Rcf Respiratory supercomplex factors SC Supercomplexes TM Transmembrane yHb Yeast flavohemoglobin

4

1 Main findings

The papers presented in this Thesis deal with factors that may modulate as evaluated at the level of cytochrome c oxidase (CytcO) from mitochondria or the quinol oxidase from Escherichia coli.

In paper I we investigated whether or not protons propagate along the membrane by means of lateral transfer, or via the bulk water. This problem was addressed by evaluating the influence of lipid composition on the coupled activity of co-reconstituted E. coli bo3 quinol oxidase and ATP-synthase. ATP synthesis was measured as a function of the protein-to-lipid ratio and the lipid composition, i.e. increasing amounts of dioleoylphosphatidyl-glycerol (DOPG), - ethanolamine (DOPE), phosphatidic acid (DOPA) or cardiolipin (CL) added to phosphatidylcholine (DOPC) liposomes. Adding only 2.5 % of CL decreased the coupled activity to about 70 % of that obtained with only DOPC. We also noted that addition of negatively charged (PG, PA or CL) resulted in a greater decrease in the ATP-formation rate than addition of the zwitterionic PE. Upon increasing the liposome size the rate decreased and was insensitive to the lipid composition. The sensitivity to lipids was partly restored by increasing the protein-to-lipid ratio. The data indicate that the average distance between the proton pump and ATP-synthase is important. We concluded that lipid composition determines proton-transfer rates and thereby influences the ATP-production rate.

Because both the medium (lipid surface) and the distance between the proton pump and ATP- synthase determines the ATP-synthesis rate, we concluded that protons are transferred via the membrane surface.

In paper II we moved from the synthetic system of liposomes to mitochondria purified from various deletion strains of Saccharomyces cerevisiae. Results from earlier studies had established that supercomplexes (SC) of respiratory enzymes are held together by the so-called Respiratory supercomplex factors 1 and 2 (Rcf1 and Rcf2) as well as cardiolipin. The previous reports had shown that deleting these factors resulted in diminished activity of the respiratory enzymes, particularly that of CytcO. Thus, the work presented in paper II started with the question if it is the disruption of supercomplexes per se that influences the activity of the separate enzymes or if the activity was altered by other means by the removal of these factors. We aimed for a detailed characterization of CytcO after having deleted Rcf1 and Rcf2.

We noted that deleting Rcf1 had the greatest influence on activity and ligand binding kinetics of CytcO. Hence, the studies were focused on Rcf1.

We used the flow-flash and flash photolysis techniques to evaluate if the diminished activity originates from the entire CytcO population or if there are separate populations being differently influenced by the Rcf1 deletion. Our data point to the co-existence of different populations of CytcO as a result of the genetic deletion.

5

Furthermore, we investigated the coupled activity of cytochrome (cyt.) bc1 and CytcO, by measuring cyt. c reduction by cyt. bc1 and oxidation by CytcO. We used both horse-heart cyt. c and the native, S. cerevisiae cyt. c. To our surprise, and great enthusiasm, we observed significant differences in the cyt. c-reduction levels. While in wild-type mitochondria, CytcO discriminated between horse and S. cerevisiae cyt. c in terms of its ability to oxidize a reduced population of cyt. c, in mitochondria from the rcf1 strain, the species discrimination was essentially absent. This led us to conclude that Rcf1 is involved in the interaction between cyt. c and CytcO.

Another feature of the rcf1 strain was the altered carbon monoxide (CO) recombination, seen as the appearance of an additional rapid kinetic component, not seen with the wild-type CytcO. This information, in combination with the diminished activity offered functional insights, but also new questions appeared. For example, is the catalytic site, or the immediate surrounding perturbed as a result of Rcf1 removal, or could heme a have lost one of its ligands to become able to bind CO (c.f. the rapid CO-recombination component)?

The results described in paper II led us to the work described in paper III, where cyanide was used to block heme a3 yielding CytcO that would be unable to bind CO. Thus, if in the presence of cyanide the rapid kinetic component remained, this would support the hypothesis that heme a was altered. However, interpretation of the ligand binding in the presence of cyanide was not straightforward and unfortunately did not offer a simple answer. Interestingly, in measurements based on reduction/oxidation and spectral properties, we noted that heme a3 of the rcf1 strain was less blocked by cyanide than heme a3 in wild-type CytcO. We interpreted this observation in terms of “electron leakage” resulting from an altered catalytic site in CytcO from rcf1 mitochondria.

In paper IV we characterized CO ligand binding to both CytcO and yeast flavohemoglobin (yHb). Using this approach, we identified a population of yHb in the IMS, which had previously remained undiscovered. Also, we characterized differences in the CO-recombination absorbance changes between yHb and CytcO in order to be able to discriminate between signals originating from these mitochondrial proteins.

We also compared the single turnover reaction with oxygen of CytcO from the two S. cerevisiae strains, BY474 and W303, and found them to display similar reaction kinetics.

6

2 Cellular energy conversion

2.1 The mitochondrion

Cellular energy conversion takes place largely in the mitochondrion. This organelle accommodates the machineries for oxidative phosphorylation (OXPHOS) and the . , on the other hand, takes place in the cytosol. Mitochondria undergo constant morphological changes to accommodate the energetic requirements of the . The mitochondrion is composed of two membranes, an outer and an inner membrane (OM and IMM, respectively) (Fig. 1). Thus, the internal volume of the mitochondrion is separated into two compartments; the intermembrane space (IMS) and the matrix. While the outer membrane contains porins that allow passage of fairly large molecules, the inner membrane is impermeable to most molecules, including protons, which thus require specific transporters to be translocated over the membrane. This property of the IMM, which allows a proton gradient to be established, underpins energy conversion. Here, electrons extracted from e.g. foodstuff are transferred through a number of membrane-protein complexes, which couple this process to proton translocation. The inner membrane is highly curved, which increases the surface area of the inner mitochondrial membrane.

Another interesting feature of the mitochondrion is that it harbors its own genome (mtDNA) which, though small, includes that encode some of the subunits of OXPHOS enzymes. Thus, respiratory enzymes of mitochondria are assembled of peptides of dual genetic origin, some of which are synthesized within the mitochondrion and others that are imported from the cytosol.

2.2 Oxidative phosphorylation

Energy conversion is more efficient when using oxidative phosphorylation (OXPHOS) than glycolysis. The OXPHOS system, or (ETC), is in most organisms composed of five protein complexes (four in S. cerevisiae), situated in the inner mitochondrial membrane through which electrons are transferred to their final target, molecular oxygen (Fig. 1) (see e.g. [1, 2] for reviews on the mitochondrial ETC). In mammalian mitochondria electrons delivered in the form of NADH enter at complex I (CI, NADH: ubiquinone oxidoreductase) and FADH2 at CII (succinate dehydrogenase). From there they are transferred to CIII (coenzyme Q-cytochrome c oxidoreductase, cyt. bc1) via ubiquinone and onwards through cyt. c to CIV (CytcO) where molecular oxygen is reduced to water. Electron transfer through the IMM is coupled to vectorial proton translocation at CI, III and IV. The established electrochemical proton gradient is used by CV (ATP synthase) to produce ATP. S. cerevisiae mitochondria lack CI.

7

Figure 1. Schematic of cristae and the respiratory enzymes of S. cerevisiae. Respiring cells form cristae to a larger extent than non-respiring cells do. CII to IV are positioned along the flat surface while the ATP-synthase is found at the base of the curve. The secluded nature of cristae produces a microenvironment with a somewhat lower pH than the rest of the IMS due to the proton translocation.

CII; succinate dehydrogenase, CIII; cyt. bc1, C; cyt. c, CIV; CytcO, CV; ATP-synthase, Aac2; ADP- ATP carrier, Pic; phosphate carrier, TCA; citric acid cycle, Succ.; succinate, Fum.; fumarate, Q; ubiquinone, OM; outer membrane, IMM; inner mitochondrial membrane, IMS; intermembrane space, P-side; positive side, N-side; negative side.

Two transporters, ADP-ATP translocator (ANT or Aac) and the phosphate carrier (Pic) are used to transport ATP out of, and ADP and Pi into the . The ADP-ATP translocator also binds to the proton-translocating complexes of OXPHOS [3–5].

2.3 To respire or ferment, that is the question

S. cerevisiae has an outstanding ability to use the carbon source that is offered and adapt to different metabolic conditions. Fermenting sugar is a way to survive at low oxygen concentrations, while respiration on, for example, glycerol or ethanol is preferred when oxygen is available. S. cerevisiae is a so-called Crabtree-positive organism [6], which means that it can ferment even in the presence of oxygen. Hence, when grown in the presence of oxygen, growth by or respiration is determined by the carbon source, specifically the level, which represses respiratory genes [7]. Galactose is a respiro-fermenting carbon source on which

8

cells can both ferment and respire. During the fermentative phase, respiratory carbon sources such as ethanol, glycerol or lactate, which are products, build up [8]. When the fermentative carbon source is consumed, the cells shift to respiration, if oxygen is available. This is called the diauxic shift. A culture at stationary phase primarily respires but, more importantly, should not be regarded as a homogenous population in senescence [9]. Rather, many sub-populations of cells are present with different proteome and mitochondrial ultrastructure and . After the diauxic shift, the volume of mitochondria in relation to total cell volume, increases from approximately 3 % in fermenting cells to 10 % in respiring cells [10]. But after the diauxic shift, many populations that are more or less respiring are likely to co-exist [9]. The morphology of the mitochondria also changes in response to respiratory conditions, by for example increased formation of cristae (Fig. 1) [10, 11].

Another feature that distinguishes fermenting cells from respiring is the iron content, or rather, if iron is bound in heme molecules, complexed to proteins, or as non-heme iron species. While in respiring mitochondria almost the entire iron population is found in proteins, during fermentation as much as 75 % is in the free form as various non-heme species [10].

What determines, or influences, the level of heme proteins in mitochondria?

9

3 Heme, heme proteins and yeast strains

The regulation of the diauxic shift lies mainly at the transcriptional level [6] where the heme activator proteins, Hap1 and Hap2/3/4/5, have key roles [12]. If the oxygen tension drops below a certain level, heme synthesis cannot pursue, as it requires oxygen [13, 14]. Thus, when oxygen tension is sufficient, heme is synthesized, which in turn activates heme-activator proteins that regulate the expression of an array of proteins, many of which have various roles during oxidative phosphorylation [15]. Among the genes regulated by Hap1, cyc1 (cyt. c isoform-1), cyc7 (cyt. c isoform-2) and yhb1 (yeast flavohemoglobin) are found. It is worth mentioning that one of the strains used in our laboratory, BY474, is a daughter strain of S288c which has a mutation in the Hap1 gene. It has been argued that daughter strains of S288c might not be the most optimal for work on OXPHOS enzymes, due to low expression levels of these enzymes [16]. Also, previous studies that compared the reaction of CytcO with O2 in strains BY474 with W303 (also used in our laboratory), had suggested a functional difference [17]. Although the expression level of respiratory enzymes does differ between the two strains (higher in W303), it appears not to be associated with the Hap1 mutation and also did not influence our data. Work done in the laboratory of Winge showed that introducing the wild-type Hap1 gene into BY474 did not alter the levels of CytcO or supercomplex formation [18]. In the work presented in paper IV, we set out to compare the kinetics of oxygen reduction by CytcO in the BY474 and W303 strains. Contrary to the results of Trouillard et al., [17] our results did not indicate any significant difference in oxygen reduction between the two.

Hemes are produced in a consecutive order, starting with heme b from which hemes c or o can be synthesized, and then heme a is synthesized from heme o [19, 20]. Heme b is farnesylated by farnesyl transferase (Cox10) to heme o. Heme o in turn undergoes hydroxylation by the monooxygenase Cox15 together with ferredoxin (Yah1) and ferredoxin reductase (Arh1) to produce the formyl characteristic of heme a [21] (Fig. 2). In eukaryotes, CytcO is the only heme a-containing protein. In paper IV, we characterized CO-recombination to yHb by the use of a yah1 null mutant strain lacking CytcO, but with heme-b containing proteins, such as yHb, intact.

10

Figure 2. Heme synthesis. Heme b is farnesylated to form heme o by Cox10. One of the methyl groups of heme o is oxidized to a formyl by Cox15 and Yah1, which results in heme a. Structures of hemes b, o, a and c are based on PDB 3WG7, 1FFT, BBI46762-mmc3 and 1CHH, respectively [22–25].

When the general heme levels are low, proteins having heme b, heme c or heme o are produced to a larger extent than those containing heme a [20, 26]. This is a result of the extra steps required to synthesize heme a from protoporphyrin IX. These steps involve addition of a farnesyl and a formyl by enzymes having a lower affinity for heme than those involved in producing the other hemes [20]. Therefore, under conditions of low heme levels, leading to less heme a, CytcO is found at a lower relative concentration compared to the other heme proteins.

This brings us to CytcO and its activity.

11

4 Cytochrome c oxidase

4.1 Reaction with oxygen

CytcO catalyses reduction of O2 to H2O, linked to proton pumping across the membrane.

2+ + 3+ + O2 + 4 cytc + 8H n  2H2O + 4 cytc + 4H p

The sub-scripts p and n refer to the positive and negative sides of the membrane, respectively.

Four of these protons are transferred to O2 while the other four are pumped. The reduction of O2 is exergonic and the free energy of the reaction contributes to maintaining the proton electrochemical gradient by means of charge separation and the proton pumping. CytcO contains four redox-active cofactors; CuA, hemes a and a3 and CuB. The latter two form the catalytic site (CS) also called the binuclear centre (Fig. 3).

Figure 3. CytcO, front and side view. The subunits that are discussed in the text are indicated. The approximate path for oxygen binding is shown. The figure is based on the yeast model BBI46762-mmc3 [24] which is based on the structure of bovine CytcO.

Electrons are transferred from of cyt. c to CytcO at CuA, in the soluble part of Cox2. Upon reduction of CuA the electron is transferred perpendicularly to the membrane to heme a and onwards to the catalytic site, parallel to the membrane (for reviews see e.g. [27, 28]). Upon reduction of the catalytic site, O2 is bound (denoted RA), with a time constant of ~10 µs at 1 mM O2, and becomes reduced by the electrons that reside at heme a3, CuB and Tyr244 (bovine CytcO numbering), resulting in scission of the O-O bond (AP) with a time constant 12

of ~30 µs. This reaction is followed in time by slower reactions that involve electron transfer from CuA and heme a to the catalytic site where two protons are taken up and one proton per electron is also translocated to the positive side of the membrane (PF and FO with time constants of approximately 100 s and 1 ms, respectively).

Before we discuss ligand binding of CytcO, we go through the tools at our disposal to investigate them and what information we retrieve from studying ligand binding.

4.2 Flash photolysis and flow flash

In flash-photolysis experiments, the sample is incubated in an anaerobic environment containing CO in a closed cuvette. Under these conditions CO binds to the pre-reduced catalytic site. The cuvette is placed in a spectrophotometer at a wavelength that is selected depending on the co-factor that is investigated, and absorbance changes are recorded. Laser-flash- illumination of the CytcO-CO complex results in dissociation of the ligand, which enters the bulk solution and then rebinds (Fig. 4). No electron transfer can occur after flash photolysis of CO from the fully reduced enzyme. The information retrieved from the rate of recombination, number of kinetic components and their respective amplitudes offers structural information on the catalytic site, which in turn may determine the kinetics of oxygen binding.

13

Figure 4. Flash photolysis. A. Experimental setup. The wavelength is chosen with the monochromator. The monochromatic light illuminates the sample, the transmitted light is detected by the photomultiplier tube (PMT) and transformed into voltage that is amplified and monitored using an oscilloscope. The laser flash dissociates the CO ligand from the catalytic site, which alters the spectral properties (see red trace). B. Illustration of heme a and the catalytic site (heme a3 and CuB) starting with pre-bound CO at the catalytic site. The laser flash induces CO-dissociation (2) and CO then rebinds, via CuB (3), seen as the decrease in signal. Structure based on BBI46762-mmc3 [24].

The spectral properties of heme a3 are determined by the redox state, whether or not it interacts with ligands and the identity of the ligand. CuB is spectrally silent. However, because the CO ligand is transferred from heme a3 via CuB, if there is a perturbation at CuB, it is likely to be displayed in altered CO-binding kinetics. The CO-recombination in rcf1 mitochondria (papers II and III) might involve an altered or even lost CuB, since recombination via an intact CuB was expected to be slower.

The laser flash induces an instant initial absorbance peak from un-ligated, reduced, heme a3 at 445 nm. The following CO recombination to heme a3, via CuB, is seen as a decrease in absorbance to the base level (Fig. 4). The amplitude of the signal corresponds to the concentration of the population that reacts with CO. Thus, the amplitudes of different kinetic phases emanate from kinetically distinct populations, likely having structurally different catalytic sites. Quantification of these populations is complicated by a potentially altered, unknown, absorbance coefficient of the novel states.

14

The flow-flash technique is used to characterize electron transfer in the oxidative phase of a single turnover reaction with oxygen (Fig. 5). Here, the CO-bound anaerobic sample is mixed with an oxygenated solution just before the laser flash releases CO, after which oxygen binds. The reaction proceeds as electrons are transferred via the redox centers to oxygen. The associated absorbance changes, recorded at wavelengths corresponding to each redox center, are used to identify the reactions and to determine their rate constants.

Figure 5. The flow-flash technique. Illustration of the experimental set-up used in a flow-flash experiment, which is used to determine electron transfer rates between the different redox centers. The anaerobic sample of CytcO-CO is mixed with an oxygenated buffer after which the laser flash releases the CO-ligand and oxygen binds. Structure based on BBI46762-mmc3 [24].

4.3 Ligand binding

4.3.1 Hemes a and a3 The iron ion of hemes can interact with six ligands, of which four sites are occupied by nitrogens of the protoporphyrin. The remaining two axial positions are often occupied by imidazoles of histidines (His). In heme a, both of these sites are occupied by His residues and the iron is in a hexa-coordinated, low-spin state. In heme a3, only one of these sites is occupied by His, and the iron is thus penta-coordinated, with the iron in a high-spin state (Fig. 6). O2 binds at the free position where other ligands, such as CO, NO or CN- also can bind [29].

15

Figure 6. Heme a and the catalytic site. Hemes a and a3 in red, the iron in brown and CuB as orange sphere with their respective ligands indicated. Structure based on BBI46762-mmc3 [24].

Why is heme coordination relevant for this work? Because a hexa-coordinated heme can in principle loose one of its axial ligands rendering it able to bind an external ligand. In paper II we characterized the ligand binding of CytcO in wild-type and rcf1 mitochondria. In the rcf1 mitochondria, our probe for ligand binding to the catalytic site, CO, bound approximately 100 fold faster than to the wild-type CytcO. The rapid CO-recombination originated from a fraction of the CytcO, while the kinetics of the other population was the same as that of the wild-type CytcO. The CO-recombination to these two populations is referred to as the “initial fast phase” and the “slow phase”, respectively. Data presented in an earlier publication suggested that, upon addition of Higd1a (a protein which belongs to the hypoxia inducible gene domain family), the mammalian homologue of Rcf1, to bovine CytcO, heme a turned penta-coordinated [30]. Therefore, in paper III we explored the hypothesis that heme a had lost an axial ligand and - was able to bind CO. To this end we used CN as a tool. Cyanide binds to oxidized heme a3 with high affinity [31] possibly by bridging heme a3 and CuB [32]. In this way, heme a3 is locked in its oxidized state and unable to bind oxygen or CO. Thus, if the fast phase remained upon addition of CO it would suggest that another heme than heme a3, was able to bind CO.

Changes in the spin-state are reflected in changes in absorbance spectra. For example, the peak at 424 nm of oxidized bovine CytcO shifts to 428 nm upon CN- binding, reflecting the shift from high- to low spin state [33]. However, binding of a ligand does not necessarily result in a shift of the spin state, nor is it always easily visible in the UV-Vis spectra.

16

4.3.2 CO-binding to other hemes There are challenges associated with investigating an altered ligand binding to CytcO in mitochondria. One obstacle is the many spectrally active co-factors present, which makes a subtle alteration difficult to detect. Also, if there is indeed a structural alteration around heme a, this is likely to alter its spectral properties, i.e. the absorption coefficient.

As mentioned above, there are other heme proteins in mitochondria that may also bind CO thereby complicating the data interpretation. One is the heme b-containing yHb (Fig. 7). The presence of yHb in the IMS resulted in observation of kinetic components associated with very slow CO-recombination, much slower than that with CytcO. Characterization of the location of yHb and kinetics of CO ligand binding to the protein is presented in paper IV. On this note, yHb was previously shown to be located in the cytosol and mitochondrial matrix [34]. However, we found that the protein is also located in the IMS (paper IV).

Figure 7. Structure of yHb (PDB 4G1V) [35]. Heme b in red and the FAD co-factor in green.

We will now look at the different subunits of CytcO and their functions.

17

4.4 Structure and function of cytochrome c oxidase

Cytochrome c oxidases from different organisms are composed of a variable number of subunits ranging from three to thirteen. The S. cerevisiae CytcO has eleven subunits (see Table 1 for their nomenclature). Eukaryotic CytcOs have dual genetic origin with the core subunits (Cox1- 3) encoded by the mtDNA and translated on mito-ribosomes while the others are imported from the cytosol. In the following description, the subunits are referred to according to S. cerevisiae numbering as Cox1-13, and the mammalian homologues in Roman numerals, CoxI-VIII. Subunits of bacterial homologues are denoted SU (subunit) I-IV. The residue numbering refers to the equivalent numbering in S. cerevisiae yeast.

Table 1. Subunit composition of S. cerevisiae and bovine CytcO. (Numbering as in [24])

S. cerevisiae Bovine (Cox nr) (roman numeral) 1 I 2 II 3 III 4 V b 5a IV/1 5b IV/2) 6 Va 7 VIIa (VIIa-1 and VIIa-2) 8 VIIc 9 VIc 12 VIb (VIb-1 and VIb-2) 13 VIa (VIa-1 and VIa-2) - VIIb - VIII

18

4.4.1 Cox1 This mitochondrially-encoded core subunit has 12 transmembrane (TM) -helices and a molecular weight of about 59 kDa. It is the first subunit to be assembled and holds the catalytic site, the D proton pathway as well as most of the K-pathway residues. The entry site for the D- pathway starts at the interface of Cox1 and Cox3 [36, 37] whereas the K-pathway starts at a residue located in Cox2 [38].

4.4.2 Cox2 Cox2 has two TM -helices with the N-and C-terminals positioned in the IMS where the redox center CuA is situated. CuA is composed of two copper ions. Their insertion requires a large number of assembly factors, among them Cox17 and Sco1 (Suppressor of cytochrome oxidase deficiency 1) [39]. CuA is ligated to two cysteines, two histidines, one methionine and one glutamate (Fig. 8).

Figure 8. CuA with ligands. The figure is based on the bovine CytcO structure PDB 3WG7 [40], colors according to atom code.

The interaction between cyt. c and CytcO is mainly mediated by Cox2, primarily via electrostatic interactions [41] and important residues are aspartates and glutamates (Fig. 9) [42, 43].

19

Figure 9. CytcO and cyt. bc1. Illustration of the cyt. bc1 (dimer) – CytcO complex and the approximate path for electron transfer, mediated by cyt. c between cyt. c1 and CuA (dotted black line). Conserved

Cox5/IV residues proposed to mediate the cyt. bc1 - CytcO interaction inside black circle. Charged CytcO residues (based on bovine structure) proposed to interact with cyt. c shown as surface in red [43].

CuA as orange spheres indicated with white arrow. The figure is based on bovine CytcO (PDB 3WG7)

[40] and cyt. bc1 (1BGY) [22]. Note that CytcO subunits VIIb and VIII that are lacking in S. cerevisiae have been removed.

4.4.3 Cox3 This subunit has been ascribed a large number of different functions, but these have not been well characterized. It is one of the core subunits, but still one of the first to detach upon detergent treatment, together with Cox13/VIa, Cox12/VIb and VIIa. It is proposed to provide structural support to the catalytic site via bound lipids that are important for the interaction between Cox1/SUI and Cox3/SUIII. In the R. sphaeroides CytcO these are PE [44] while in the bovine CytcO they are PG, suggested to function as proton traps close to the D-pathway orifice [45]. SUIII also presumably forms the putative oxygen channel [46]. In the context of yeast mitochondrial supercomplexes, Cox3 is found to associate strongly with Rcf1 [3–5]. In the bovine respirasome, which includes CI, Cox3 is situated at the interface between cyt. bc1 and CytcO [47].

4.4.4 Cox5 Cox5 is the largest mammalian and least conserved [48] of the nuclear-encoded subunits. It has only one TM -helix and the C-and N-terminals are found in the IMS and matrix, respectively. In S. cerevisiae Cox5 has normoxic and hypoxic isoforms (5a and 5b, respectively) that are co- expressed with cyt. c isoforms 1 and 2, respectively. Cox5 and cyt. c isoforms are regulated on the transcriptional level by oxygen [49], heme [50] as well as NO [51]. On the proteome level, Cox5 isoforms are regulated by proteolytic cleavage of the “wrong” isoform [52]. The oxygen-

20

dependent isoform expression of Cox5 is why this subunit has been ascribed a central role in the regulation of OXPHOS [53].

Figure 10. S. cerevisiae CytcO. Cox1 and Cox5 are indicated. Note the proximity of Cox5 (blue) and the TM -helices XI and XII of Cox1 (green). C-terminal residues of Cox5 in orange circle. The figure is based on BBI46762-mmc3 [24].

The TM -helix of Cox5 is juxtaposed to TM -helices XI and XII of Cox1 (Fig. 10) and also harbors an ATP-binding site close to TM10 of Cox1 [13]. The C terminus is highly conserved between the two Cox5 isoforms and a CytcO lacking the last 10 residues in the C terminus of Cox5 is completely inactive [13]. This finding implies that the functional difference does not originate from the C-terminal residues. Cox5 is one of the subunits that may undergo posttranslational modifications, such as phosphorylations. Residues that undergo phosphorylations are primarily N-terminal serines situated in the matrix [54]. In addition, ATP binding to bovine CytcO increases the Km for cyt. c [55] mediated by Cox5/IV [56].

Support for a functionally important role of Cox5 comes from FTIR experiments in which CO- ligation to the catalytic site of CytcO having either Cox5a or Cox5b was investigated [57]. CytcO with 5a was found to have two distinct populations, while the one with Cox5b was homogenous (see the section on “Populations of CytcO”).

Intriguingly, during CytcO assembly Cox5 is incorporated just after Cox1, even before the core subunits Cox2 and Cox3 [58, 59]. It has also been shown that in a yeast strain lacking CL the fraction of supercomplexes was decreased, as was the CytcO expression level and the CytcO was less active. Addition of an expression vector with Cox5a restored the level of CytcO, its activity and association into supercomplexes [18]. It must be mentioned that in a model of the yeast supercomplex, Cox5 mediates the interaction between cyt. bc1 and CytcO [60].

21

Importantly, in a study that compared the activity of CytcO having either one of the two yeast Cox5 isoforms showed no effect of Cox5 substitution under aerobic conditions and when the genetic conditions were such that expression levels were similar [61]. The authors did, however, measure a more than two-fold increase in oxygen-reduction rate of CytcO with Cox5b, compared to Cox5a, when the genetic conditions were such that Cox5b was less expressed compared to Cox5a. In this strain, Cox5a and Rox (the suppressor of Cox5b under normoxic conditions) were genetically deleted, allowing an unmodulated expression of Cox5b under its own promoter (otherwise placed under the promoter of Cox5a). The authors discussed their finding in the context of supercomplex formation and they speculated that formation of supercomplexes and Cox5 isoform expression could potentially go hand in hand. This conclusion was also supported by results from other studies [51, 56]. From a structural point of view, this conclusion does not seem unlikely, given the position of Cox5 at the interface in the yeast supercomplex [60].

One intriguing feature of CytcO having Cox5b is its reported ability to reduce nitrite to nitric oxide at low oxygen tension or anoxia [51]. This makes nitrite an alternative electron acceptor for CytcO when oxygen is absent, thereby enabling a continued energy conversion. In addition, the product nitric oxide functions as a signaling molecule which up-regulates the expression of hypoxic genes, such as CYC7 (cyt. c isoform-2) [51], which is co-expressed with Cox5b. Thus a feed-back mechanism is established to further promote the expression of hypoxic genes. NO which is toxic is kept at low levels by yHb.

4.4.5 Cox12 (VIb) Cox12 is small, highly conserved [48], completely IMS-located and, in crystal structures, it provides a dimerization-site for CytcO monomers, together with Cox13 [43, 63]. Cox12 is easily lost during purification or upon removal of CL [64]. The resulting CytcO does not appear to be inhibited in any way, apart from losing its ability to dimerise. The residues that interact with the other monomer are found in a loop of conserved residues, one of them being glycine 47 (Gly47) which is found in all taxa [48]. If, however, Cox12 is genetically deleted, the CytcO level is severely diminished, presumably due to an impaired assembly [65]. In support of a role during assembly, Cox12 was found to be involved in copper insertion into Cox2 [66]. Interestingly, Cox12 appears to be an interaction partner for cyt. c [3, 4, 43, 48] which would not be surprising given that Cox2 and Cox12 are found next to each other in the CytcO structure (see Fig. 9). That the association between Cox12 and cyt. c could be functionally relevant finds support in the identification of a testes-specific Cox12/VIb-T isoform [67] suggested to be co- expressed with a testis-specific isoform of cyt. c [68], the only mammalian cyt. c isoform [69]. The mammalian cyt. c isoforms differ primarily in their C-terminal regions. This tissue- specificity of cyt. c suggests a potential adaptation to environmental conditions of CytcO, manifested via changes in Cox12.

22

4.4.6 Cox13 (VIa) This subunit has been shown to be particularly important for formation of the tetrameric supercomplex composed of III2/IV2 [3-5]. Purified yeast CytcO was found to exist in two populations, defined by the presence or absence of Cox13 [4]. Cox13, in turn, is lost upon genetic removal of Rcf1 [4]. Cox13 is not required for the assembly of CytcO [70, 71] but mediates, together with Cox12, its dimerization, though the details of this are elusive [63]. The first N-terminal residues of bovine Cox13/VIa mediate the interaction with Cox1 on the other monomer, but only one of these residues is well conserved [48]. Bovine Cox13/VIa requires CL to associate with CytcO [57, 58, 66]. In terms of regulation, CytcO can be both activated and inhibited by ATP in ways suggested to involve Cox13/VIa [71]. A cluster of conserved residues in the cytosolic region are suggested to form the ATP-binding motif [48].

It is worth noting that the gene for Cox13 contains a conserved binding site for the transcriptional regulator Hap4, involved in up-regulation of nuclear-encoded genes needed for the diauxic shift [12]. In line with this, Cox13 is up-regulated by the respiratory carbon source glycerol [73].

4.4.7 Populations of CytcO Distinct populations of CytcO having different ligand binding kinetics and spectral properties have been described for different organisms [28, 69, 70, 71, 72, 74]. The two populations are in the oxidized enzyme termed pulsed/resting or fast/slow, defined based on the kinetics of cyanide inhibition. The populations of the reduced enzyme are characterized by their different spectral properties when bound to CO and called conformers I and II [13]. The various forms are suggested to result from a structural flexibility around the catalytic site [77].

Let us continue on our path into the IMS and cytochrome c.

23

5 Cytochrome c

Cytochrome c is a small protein with a molecular weight just below13 kDa. In mitochondria cyt. c is a hub that links OXPHOS, and mitochondrial protein import linked to oxidative folding. Cytochrome cs across species have a number of highly conserved residues [79], which include that mediate the interaction of cyt. c and CytcO [41] (Fig. 11 A). Heme c is distinguished from other hemes in that it is covalently attached to the peptide via conserved cysteines [79]. The heme axial ligands are methionine and histidine. Note that the heme molecule resides in a cavity near the interacting surface which places the methionine and the cysteines that co-ordinate the heme molecule close to the area of protein-protein interaction.

Figure 11. S. cerevisiae cyt. c isoforms-1 and 2. A Surface structure. Positive residues ( or arginine) important for the interaction with CytcO in blue, the fungi-specific N-terminal residues in light grey, the isoform-2 specific first N-terminal residues in dark grey and the C-terminal in light green. B Sequence alignment of isoforms 1 and 2. Cysteines in yellow (the first two make the covalent attachment of heme to the protein and the third, Cys102, is specific for isoform-1). Heme axial ligands His18 and methionine 80 (Met80) in cyan. The N-terminal residues in light and dark grey are the same as in A. Black arrow indicates the first residue in other species. Structures from PDB 1CHH (isoform-1) [25] and 3CXH (isoform-2) [80].

As already discussed above, S. cerevisiae has two oxygen-dependent isoforms of cyt. c (co- expressed with Cox5a or 5b, respectively) that differ primarily in their N-terminal regions

24

where isoform-2 has four residues that are not present in isoform-1 (Fig. 11 B). The N-terminal residues that are found only in isoform-2 were proposed to explain (at least in part) the selective ubiquitin-mediated degradation of apo-cyt. c isoform-1, which is not involved in the degradation of isoform-2 [81]. Also, isoform-1 has a cysteine 102 (Cys102) in the C-terminal which makes it more prone to dimerize [41]. Interestingly, the nine N-terminal residues just mentioned are specific for fungi.

Synthesis of cyt. c and CytcO are intimately connected. Cells that lack cyt. c cannot assemble CytcO [82]. Intriguingly, the mechanism appears to be structural, rather than functional. This conclusion was illustrated by using a “catalytically dead” cyt. c mutant where a highly conserved residue, tryptophan 65 (Trp65) necessary for electron transfer, was replaced by a serine. While a cyt. c null mutant completely lacked heme a, the redox-inactive mutant was able to partially support CytcO assembly. Also, expression of CytcO did not require stoichiometric amounts of cyt. c.

5.1 Cytochrome c and Rcf1

In paper II we argued that Rcf1 is involved in the interaction between CytcO and cyt. c. The basis for our conclusion was primarily the finding that a greater fraction of cyt. c was oxidized when Rcf1 was genetically removed, despite CytcO showing a decreased oxygen-reduction rate in the rcf1 strain. We observed that in the presence of Rcf1, after addition of quinol, the cyt. c pool was reduced and remained reduced. This behavior was interpreted in terms of an initial reduction of the cyt. c pool by cyt. bc1 after which electrons were transferred via a pre-bound cyt. c directly to CytcO. In the absence of Rcf1, after an initial reduction of the cyt. c pool by cyt. bc1, cyt. c was re-oxidized by CytcO. Hence, the data indicate that Rcf1 binds cyt. c.

Binding of Rcf1 to cyt. c is also supported by recent solution NMR studies [83] (Fig. 12 B). Other (unpublished) FCS (fluorescence correlation spectroscopy) data from our laboratory show a physical interaction between cyt. c and Rcf1. In those experiments Rcf1 was reconstituted in GUVs (giant unilamellar vesicles) with cyt. c added. Although cyt. c did bind to the vesicles in the absence of Rcf1, the interaction was greatly increased in the presence of Rcf1. In yet another unpublished study CytcO from the rcf1 strain was reconstituted into liposomes and purified Rcf1 was introduced. The activity of the otherwise less active CytcO was restored to wild-type level, though the mode of interaction remains to be resolved. Collectively these data present strong support for an interaction between Rcf1 and cyt. c.

25

Our conclusions also find support in another publication that compared the oxidation of cyt. c by CytcO where cyt. c was either wild-type or modified at tyrosine 48 (Tyr48) [84]. In these experiments the involvement of Rcf1 in the oxidation of cyt. c by CytcO in yeast wild-type or double Rcf1/Rcf2 null mutant strains were studied. Furthermore, the authors performed experiments where mammalian Higd1a or Higd2a was added to purified bovine CytcO where they noted an increased oxidation of cyt. c by CytcO. The effect was greater with the unmodified cyt. c. Hence, these data also point to Rcf1/Higd proteins as modulators of CytcO activity in a way that involves Tyr48 of cyt. c. Data from the Stuart laboratory showed that in a truncated variant of Rcf1, lacking the large C-terminal, the effect of Rcf1 on supercomplex formation was mediated by the conserved Hig-domain [3] (see Fig. 12 A for a schematic of Rcf1) To summarize, (i) Rcf1 may be involved in the assembly of CytcO [85], (ii) cyt. c is structurally essential for assembly of CytcO [82] and (iii) Rcf1 binds cyt. c [83]. When combining these findings one may speculate that Rcf1 and cyt. c together participate in the biogenesis of CytcO.

Figure 12. Structure of Rcf1. A Predicted structure having two TM -helices and the C-terminal as an IDP (intrinsically disordered protein). The predicted orientation was also supported biochemically [3]. B Structure as determined by solution NMR, with 5 TM helices and three charged residues in the last soluble domain proposed to be part of the interaction with cyt. c. (modified from [83]).

As mentioned above, cyt. c is also involved in import of proteins destined for the IMS that go via the Mia40 pathway, another mitochondrial electron transport chain we will have a glimpse at now.

26

5.2 Cytochrome c in oxidative folding

Mitochondria have different pathways for protein import, depending on their target location. One is the Mia40 pathway, which employs so-called oxidative folding for import of IMS- targeted soluble proteins that have a twin-cysteine motif (either CX9C or CX3C) [48, 79, 87]. Briefly, Mia40 oxidizes the substrates by creating disulfides between the cysteines of the motif. The disulfides, at the same time, assist in folding the proteins and capture them in the IMS. The Mia40 pathway is influenced by the oxygen level, and requires the activity of respiratory enzymes [86]. After Mia40 has oxidized its substrate, it is re-oxidised by Erv1 (Essential for respiration and vegetative growth protein 1). This is an essential flavin-containing sulfhydryl oxidase which delivers the electrons from Mia40 to either oxygen directly or, more commonly under physiological conditions, via cyt. c to either CytcO or cyt. c peroxidase [53, 87, 88]. Note that among the proteins imported via this route several CytcO assembly factors e.g. Cox17, Sco1 and cmc1 (Cx9C mitochondrial protein necessary for full assembly of cytochrome c oxidase) are found. The soluble IMS located subunit Cox12, is also imported via the Mia40 pathway [89]. In yeast, but not in higher eukaryotes, a ternary complex of Mia40, Erv1 and cyt. c has been described and shown to be anchored to the IMM via Mia40 [53, 87]. This places cyt. c and CytcO in a really exciting cellular context because they define the intersection where the electron transfer-chains of OXPHOS and oxidative folding physically converge.

Before we continue with respiratory supercomplexes we discuss the mitochondrial membrane, which is at the core of respiration with influence on both the stability and activity of the respective proteins and their propensity to form larger assemblies.

27

6 The membrane

The lipid composition determines, for example, membrane fluidity, thickness, surface charge, packing and intrinsic curvature [90]. Lipids can be cone-shaped to various degrees, which make them more or less prone to form bilayers. Cone shape also positively correlates with membrane fluidity [90] and influences protein mobility [91]. The protein/lipid ratio of the IMM around cristae, is 75%/25%, where respiratory enzymes may constitute up to 80% of total protein content [92]. The crowded nature of the IMM is sometimes considered a reason for assembly of respiratory enzymes into supercomplexes. This argument is however refuted by others because e.g. beef heart mitochondria, which have a higher protein density than yeast mitochondria, maintain their respiratory enzymes as separate entities to a larger extent than yeast does. Roughly 20% of mammalian CytcO is found in supercomplexes, while in S. cerevisiae, almost the entire population is in complex with cyt. bc1 [93].

Cardiolipin is unique in terms of structure, charge, cellular location and synthesis, let us therefore explore it some more.

6.1 Cardiolipin

CL is a cone-shaped lipid, which results from the four unsaturated acyl chains that take up more space in relation to the smaller head group, which makes it so-called non-bilayer forming. Cone- shaped lipids put a strain on the membrane that at the same time supports the embedded proteins [94, 95]. The stabilizing role is particularly important for oligomeric integral proteins that switch between conformations where the transition between conformations may be induced by the lateral pressure of the membrane [95, 96]. Estimating a pKa to the two ionizable phosphates of the head group is difficult. Recent titration assays indicate that CL is a di-anion at physiological pH [97]. In the presence of di-valent cations (e.g. Ca2+), which reduce repulsive forces of the negatively charged head groups, CL can form higher order structures, such as inverted hexagonal phases [98] suggested to facilitate fusion and fission of mitochondria [90]. In fact, in proteo-liposomes of high CL content and low CytcO content, CL would associate into the inverted hexagonal phase when exposed to Ca2+. This is counteracted by increased amount of the integral membrane protein [98].

Cardiolipin is plentiful in tissues with high energetic demands, such as beef heart mitochondria, from where it was first characterized. Specifically, CL resides in the energy-transducing highly curved IMM where it constitutes 5-25% of the total lipid content (data varies between sources [99, 100].

In yeast cultures the CL content is altered according to metabolism such that it is mostly synthesized during the mid-exponential phase and its steady-state level is highest during early stationary phase, which is when cells respire [101, 102]. The requirement of CL for the activity

28

of CytcO differs between organisms. While in mammalian cells lack of CL influences activity and structure of CytcO, in R. sphaeroides [103] or S. cerevisiae [104, 105], it is not required. In yeast, however, lack of CL induces elevated levels of PG for compensation and yeast cells that lack both CL and PG are not viable [104]. Although CL, PE or PG are not required for viability under normal conditions, when placed under stress-conditions, the need for specific lipids to retain structure and function increases [106]. Like CL, PE is a non-bilayer forming lipid, the most common in yeast, but unlike CL it is zwitterionic [90] (Fig. 13 B).

Either CL or PE is required to maintain an intact mitochondrial [94, 107]. Interestingly, the activity of CL synthase (Crd1), which is a membrane protein situated on the inner leaflet of the IMM [108], is dependent on the electrochemical potential [101]. The activity of CytcO lacking CL is reduced to the extent that the proton gradient is diminished, which impairs membrane potential dependent protein import [94].

Now I would like to introduce the concept of lateral proton transfer, a mechanism by which protons can move between proton “producers” and “consumers”.

6.2 Lateral proton transfer

Lateral proton transfer via protonatable groups at the surface [109] provides a pathway for the protons along the surfaces of both membranes and proteins. Protonatable groups on the protein surface allow faster proton transfer to pathways that are used for proton translocation, as shown for R. sphaeroides CytcO [110, 111]. But protonation of protein residues is also influenced by the surrounding environment, as seen in the increased rate of protonation on CytcO residues near the membrane interface when reconstituted in a membrane, as compared to a detergent- solubilized enzyme [112].

In paper I we asked if protons move between membrane-bound proton “producers” and “consumers”, by lateral transfer along the membrane surface, or in exchange with the bulk. This problem was addressed using proteo-liposomes of different lipid composition with co- reconstituted E. coli bo3 quinol oxidase and ATP synthase, and measuring the ATP-production rate (Fig. 13 A). Introducing CL into the liposomes had greatest effect on the ATP-production rate followed by DOPG and DOPA.

29

Figure 13 A. Experimental system used to investigate lateral proton transfer (paper I). B Structures of lipids used in the study. Acyl chain length and number of unsaturations vary greatly both between and within species. PC and PE are zwitterionic (neutral), PA and PG are negatively charged (-1) and CL has the ability to form either one or two negative charges, depending on pH. (Structures are from Avanti Polar Lipids web page: Avantilipids.com).

We will now continue for a discussion on supercomplexes and their assembly, where lipid composition also plays a role.

30

7 Respiratory supercomplexes

We have now arrived at the respiratory supercomplexes, whose existence for long was debated. With new techniques to visualize them, the questions asked today do not relate to their existence but rather why they exist and what mediates their formation.

Respiratory enzymes are highly conserved, yet their assembly into supercomplexes varies between organisms. The need for and mechanisms of metabolic adaptation differ between organisms and also vary between tissues of higher eukaryotes. Recently, structures of mammalian respirasomes have shed some light on their organization [113–115]. A model of S. cerevisiae supercomplex based on single particle cryo-EM fitted with structures of yeast cyt. bc1 and bovine CytcO is also available [60] (Fig. 14).

Figure 14. Model of the tetrameric CIII2CIV2 supercomplex of S. cerevisiae. Position of complexes based on [60]. Upper panel “front view”, bottom panel “top view”. CytcO is based on BBI46762-mmc3,

[24] and cyt. bc1 on PDB 1BGY [22].

31

The function and role of the supercomplexes has long been debated. Some suggested roles include: substrate channeling, minimization of so-called ROS, kinetic advantages, stability of individual complexes [116, 117]. Supercomplexes have also been proposed to form under metabolic stress conditions [118]. The argument of protein stability primarily refers to mammalian CI that requires CIII and CIV for its stability [68]. In human and bovine muscle cells, 95% of CI is found in SCs. Importantly, CytcO is the only OXPHOS enzyme with adaptive isoforms [119] making it the central player, with which the other complexes associate. This role of CytcO is emphasized in the title of a review by Susanne Arnold: “The power of life- cytochrome c oxidase takes center stage in metabolic control, cell signaling and survival” [120].

In mammals, different metabolic requirements are manifested at the level of tissue or developmental stages, which they meet by expressing specific isoforms of CytcO [93] or by a flexible organization into supercomplexes. Mammalian CytcO appears mainly as free enzymes with around 20% in SC [121] but endurance exercise, where oxygen consumption is high, and the resulting oxygen level low, results in redistribution with less free CytcO and more in SC [122].

Yeast may undergo more swift adaptations in response to, for example, a change in carbon source or altered oxygen tension. The organism meets these changes in conditions on several levels, from altered transcription profile to spatial reorganization of respiratory enzymes. In contrast to mammalian CytcO, the yeast CytcO is primarily found in complex with a dimer of cyt. bc1. The stoichiometry of complexes III and IV is a response to the environmental conditions. Respiratory growth on lactate or glycerol produces mostly the tetrameric III2/IV2 SC, while growth on glucose to a larger extent results in a III2/IV1 stoichiometry and free cyt. bc1 dimers [93]. The III2/IV1 stoichiometry is also seen in rcf1 cells [3–5] and cells lacking CL [123], both of which exhibit a dysfunctional respiration.

7.1 Factors related to supercomplexes

7.1.1 Rcf1 Rcf1 is an 18 kDa peptide of 159 residues with two predicted transmembrane helices located in the IMM [3–5]. The N-terminal first 96 residues of Rcf1 form the Hig-domain which is highly conserved throughout the kingdoms of life (discussed above and illustrated in Fig. 12). Its C-terminal, on the other hand, is fungi-specific and, according to predictions, an intrinsically disordered peptide (IDP) with a very high content of charged residues. There is no consensus on the role of Rcf1 as being an assembly factor for supercomplexes or for CytcO. The latter is suggested based on the observation that Rcf1 is required for formation of the fully assembled CytcO [85, 118].

32

Rcf1 interacts with both cyt. bc1 and CytcO, but binds stronger to CytcO [4, 5]. Cells having a truncated version of Rcf1, consisting only of the Hig-domain, still have functional CytcO and supercomplexes, indicating that independently of the role of the C-terminal, it is not required for activity of supercomplexes or CytcO [3]. Deletion of Rcf1 appears to put the cell in a respiratory-deficient mode, as manifested in the vastly reduced growth rate of the rcf1 strain on glycerol as compared to galactose [5] (and our unpublished data) and the diminished CytcO activity of the rcf1 strain [3, 4, 124].

But the function of Rcf1 reaches beyond its role in CytcO assembly and function; it also associates with the ADP-ATP translocator Aac2, via its conserved Hig-domain. Their association persists when OXPHOS enzymes are removed, as concluded from work on rho0 mutants lacking mitochondrial DNA (mtDNA) and thereby the OXPHOS enzymes [3, 125]. The function of the soluble fungi-specific domain remains to be determined. Even though the data collectively point to the Hig-domain as the functional unit, the observation that cyt. c isoform-2 also has an extended tail, present only in fungi, suggests that a putative interaction should be investigated.

7.1.2 Rcf2 Another member of the Hig-family of proteins is the fungi-specific Rcf2, which has also been implicated in formation of supercomplexes. However, genetic removal of Rcf2 does not influence CytcO to the same extent as removing Rcf1 [3– 5, 118]. In paper II we also characterized the effects on CytcO activity and ligand binding upon genetic deletion of Rcf2 as well as the double deletion of Rcf1 and Rcf2. In these experiments the deletion of Rcf2 did not influence CytcO to the same extent as was seen in the rcf1 strain. This observation was in agreement with results from previous studies [3, 4]. Therefore, our work was primarily focused on Rcf1.

7.1.3 Aac2 Although in the work presented in this thesis we did not investigate the ADP-ATP translocator Aac2, I have chosen to discuss it here. The reason is the involvement of Aac2 in supercomplex formation and its physical association with Rcf1, both of which are topics of this thesis. The expression of Aac2 is induced by oxygen and regulated by Hap2/3/4/5 [126], that also upregulate Cox13. Aac2 is responsible for the export of produced ATP and import of ADP, used as substrate by ATP synthase, and requires CL for function [127]. The transporter utilizes the electrochemical proton gradient for ADP/ATP transport.

Aac2 associates with supercomplexes and specifically associates with Rcf1 in CIII2/CIV2 supercomplexes [4, 121, 128]. Upon its removal fewer tetrameric SC are formed, while the trimeric form remains [4, 121]. Deletion of Aac2 also leads to impaired synthesis of CytcO

33

subunits Cox2, Cox4 and Cox5 [125]. Hence, lack of CytcO potentially explains the decrease in tetrameric supercomplexes. The complex of Aac2 with respiratory supercomplexes has also been shown to include the inner mitochondrial membrane translocase, Tim23 [125], which facilitates protein import in a membrane potential-dependent manner [16].

Hence, by removing Rcf1, an interaction partner for several proteins (cyt. c, CytcO, Aac2 and also the translocase Tim23) active in processes of energy conversion and protein import, is lost, with potential downstream effects we have yet to understand.

7.1.4 Cardiolipin The distribution of small or large supercomplexes correlates with the CL content. This has been shown in yeast [123] and studies on synthetic systems [129]. Supercomplexes composed of III2/IV2 from S. cerevisiae were found to associate with around 50 CL molecules. These molecules were of the same acyl chain composition as the general CL population in the IMM, which suggested that the interaction does not require a specific CL species [60]. CL mediates supercomplex interaction by binding to lysines or arginines found at the cyt. bc1- CytcO interface [130]. Despite the stabilizing function of CL on CytcO, it is not essential for either fermentative or respiratory growth [123]. As mentioned, PE is a zwitterionic non-bilayer forming lipid [90] found in all membranes of the cell. Interestingly, PE and CL have opposing effects on SC stability, where PE rather stabilizes the complexes in their free form [94]. The CL/PE double mutant is lethal [131].

34

8 Questions and speculations

In this section I would like to discuss some points mentioned above that I find particularly intriguing and thought-provoking. But, again, my thoughts on these issues go beyond the boundaries of the papers included. Still, I take the opportunity provided with this chapter to elaborate some on them.

8.1 Heme synthesis and distribution of respiratory enzymes Recall that the heme synthesis is related to metabolism and the oxygen tension. At conditions of low heme levels, less CytcO is produced, not only in absolute numbers, but also relative to the other respiratory enzymes. Given that the amount and stoichiometry of supercomplexes seemingly relate to the level of fully assembled and functional CytcO, it is tempting to speculate that under conditions of low heme synthesis formation of supercomplexes is a way to direct electron transfer to those CytcO that are fully assembled and functional by linking these fully assembled CytcO with cyt. bc1.

8.2 The Mia40 pathway, OXPHOS and the two-faced cyt. c I find the multi-tasking role of cyt. c to be particularly interesting, being involved in three seemingly unrelated processes. The Mia40 pathway, which facilitates import of small soluble IMS-located proteins, including Cox12 and several CytcO assembly factors, uses cyt. c as an electron mediator, which in turn delivers the electrons to CytcO. As mentioned above, Cox12 has been found to be required for CytcO formation and suggested to participate in the assembly of Cox2. Importantly, the activity of the fully assembled CytcO is not impaired when Cox12 is lost. In addition, cyt. c is structurally required for synthesis of CytcO. The communication between the Mia40 pathway and OXPHOS deserves further investigation.

8.3 Rcf1, cyt. c and Cox5 There are a few points that may very well be completely circumstantial, but intriguing. As discussed above, there is support for an interaction between Rcf1 and cyt. c. Also, specific isoforms of cyt. c are co-expressed with the equivalent isoforms of Cox5. While the Hig-domain is well conserved between species, the C-terminus is fungi-specific. N-terminal residues of cyt. c isoform-2 are also fungi-specific. These fungi-specific termini of both Rcf1 and cyt. c are found externally. Rcf1 is implicated in formation of supercomplexes. In structures of supercomplexes, Cox5 is found where CytcO and cyt. bc1 interact. Collectively there appears to be a link/interaction between Rcf1, cyt. c and Cox5.

35

9 Populärvetenskaplig sammanfattning

Barriärer är en förutsättning för liv. I cellen utgörs barriärer (membran) av ett tunt lager fett (lipider). Membraner skapar en avgränsning av olika rum vilken medför att passage av olika molekyler, ofta joner, kan regleras där enbart de med behörighet tillåts, eller ens kan, passera. Med membraner följer även att koncentrationen av en viss molekyl kan vara större på ena sidan, jämfört med på den andra, det vill säga, en gradient kan byggas upp.

Ett av cellens många rum, mitokondrien, omvandlar energi vi får i oss med maten till en för cellen tillgänglig energi-molekyl, ATP. Det som extraheras ur maten för att ombildas till energi i form av ATP är elektroner (negativ laddning). Grunden för ATP-produktion baseras på att protoner (positiv laddning) pumpas av biologiska maskiner, enzymer, från ett rum i mitokondrien till ett annat. Denna pumpning drivs med hjälp av elektroner från maten. Slutstationen för elektronerna är syret vi andats in. Den protongradient som byggts upp, vilken kan jämföras med ett batteri som laddats, kan cellen sedan vid behov använda sig av. Processen följer samma princip som vattenflödet i ett kraftverk. Vatten som ansamlats kommer när porten öppnats att driva en turbin med resultatet att energi omvandlas till en för oss användbar form. Med samma princip kommer flödet av protoner, när de släpps tillbaka genom ATP-syntas (som syntetiserar ATP) få ATP-syntaset att snurra och ATP att bildas. Processen som omvandlar energi på detta vis kallas oxidativ fosforylering och utförs av fem proteinkomplex, numrerade komplex I-V. I det fjärde komplexet, cytokrom c oxidas, omvandlas syre, elektroner och protoner till vatten. Elektronerna transporteras mellan komplex III och IV med hjälp av ett litet protein, cytokrom c. Komplex IV kan även binda andra molekyler, såsom kolmonoxid eller cyanid. Denna förmåga är användbar när man undersöker dess egenskaper på molekylär nivå.

Proteiner som sitter inbäddade i membraner påverkas av membranets egenskaper. I avhandlingens första artikel undersökte vi huruvida de pumpade protonerna som rör sig mellan de olika komplexen gör det nära membranets yta eller om de sprids ut i den omgivande lösningen. För detta skapade vi membran med olika ytladdning där vi satte in komplex IV och V samt varierade avståndet dem emellan. Våra data, som visade att de pumpade protonerna påverkades av dessa parametrar, var till stöd för teorin att protonerna rör sig i membranytans absoluta närhet.

Många fundamentala biologiska processer skiljer sig förvånansvärt lite åt mellan till synes vitt skilda organismer. Bakjäst, Saccharomyces cerevisiae, är en encellig organism som uppvisar stora likheter med däggdjur. Detta, samt att den enkelt kan manipuleras med genteknik, gör den till ett effektivt forskningsverktyg. I arbetet som presenteras i artiklarna II-IV använde vi oss av bakjäst.

Under senare år har kunskap om interaktionerna mellan de proteiner som tillsammans utför oxidativ fosforylering ökat. Man har noterat att komplexen går samman i så kallade superkomplex samt identifierat ett fåtal aktörer med betydelse för denna samverkan. Dessa är främst små proteiner som heter Rcf1 och Rcf2 (eng. Respiratory supercomplex factor) samt en

36

för mitokondrien unik fettmolekyl, kardiolipin. Eliminering av dessa, framförallt Rcf1, har visat sig ha betydande negativ inverkan inte bara på komplexens förmåga att interagera, utan även på cytokrom c oxidasets aktivitet.

I den andra artikeln undersökte vi betydelsen av Rcf1 och Rcf2 för elektronöverföringen till och i cytokrom c oxidaset. Vi noterade att borttagande av dessa resulterar i bildandet av olika populationer, vilka skiljer sig åt i flera hänseenden. Tillsammans pekar våra data på att det område i oxidaset där syret omvandlas till vatten får en annan struktur när Rcf1 tas bort. Vidare fann vi att elektronernas väg från cytokrom c till syret, det vill säga genom cytokrom c oxidaset, påverkas av Rcf1. Vår tolkning är att Rcf1 är involverad i interaktionen mellan cytokrom c och cytokrom c oxidaset.

I den tredje artikeln undersökte vi närmare hur cytokrom c oxidasets struktur ändras när Rcf1 tas bort. Vi använde oss av cyanid som binder till samma plats som syre men ”låser” oxidaset i en inaktiv form. När oxidaset är ”låst” kan inte syre binda och elektroner kan inte heller passera. Våra data visar att när Rcf1 tas bort, förmår cyanid inte låsa oxidaset såsom förväntat. Detta antyder att strukturen där syre binder på något vis förändras när Rcf1 tas bort.

Slutligen karakteriserade vi ett annat enzym, flavohemoglobin, samt identifierade dess fördelning i mitokondrien. Vidare jämförde vi reaktionen med syre hos cytokrom c oxidas från två olika jäststammar som skiljer sig åt i sin arvsmassa.

Sammantaget har våra studier lett till ökad kunskap om samspelet mellan olika aktörer som utför oxidativ fosforylering och bidrar därmed till en djupare förståelse för de molekylära mekanismerna bakom denna energiomvandling.

37

Acknowledgements

Peter, a great thank you for the wonderful years you have given me by including me in your research group. The many exciting moments I had were possible because of your enthusiasm and positive spirit that allows the scientific process to be creative and joyful. I also thank you for allowing me to be a bit “all over the place” while keeping your door open and a port to return to when my scattered thoughts and data needed a focus. The confidence you showed was equally valuable as appreciated. Also, how you take your time to assist us in writing papers and theses add a huge educational value to such processes. It’s funny to think of how those years ago, when I first visited you, I thought I was applying for a short-term job doing lab-dishes. This topped that by far .

Christoph, although you left our lab physically you remained as my co-supervisor, for which I am very glad. In the advent of my PhD studies, when I was scared of even entering the lab, you encouraged me to go explore and to just try things out. You telling me that failure is unavoidably part of science was surprisingly reassuring. I have really enjoyed our discussions, scientific and general. (You clearly did not contact PubMed to have me banned from their site, for which I am also very glad).

Pia, thank you for willingly answer any questions I might have had. But more importantly, thank you for asking the questions I never wanted to have. Though my relief was beyond proportion any time you could not participate in our data up-dates, I am grateful for all the times you did. I suspect few occasions are as preparatory for the D-day as meetings with you present.

Elzbieta, I want to thank you for encouraging me to stand my ground in times of conflicts. Despite the many opportunities for women of our time, you remain an inspiration.

I wish to thank the administrative and technical staff- you have been great and so helpful at all times. You are the oil that keeps DBB running. - Maria, I cherish our many and joyful talks.

A profound thank you to SU avhandlingssupport. Your professionalism and confident approach gave me a portion of serenity during weeks of great stress.

Matt, you already know how much your pushing helped when my fears got hold of me. I thank you for believing in me when I could not myself, that made a difference.

Linda H, I am so glad to have got to know you. Studying for the big evil with you, and sharing both laughter and despair, made it bearable. Also a huge thank you for saving me from my table of contents-crisis…

38

To members of the PBPÄ group; thank you all for making our group such a pleasant one.

I had so much fun sharing office with you Axel, before you left for the land of chocolates (I get why, since those mitochondria of yours must be seriously uncoupled). Then you Max took over that desk, which was quite the same; lots of fun and heaps of sweets (again, more mitochondria to kill for). Fede, we did not share office that long but it was equally appreciated as those lovely cup-cakes of yours. Irina, I wish I could say that sharing office with you was nice but you know I would lie. Our climate preferences are simply incompatible. But we are not. I am glad our relation got from below frosty to really warm and all it took was for me to shout “Back off Irina!” at you. Your reply “Let’s hug Camilla” really was quite unexpected. Thank you for providing a shoulder when my temper or nerves go haywire. I am certain you have done enough good this time around to not end up a worm on your next. I would like to give you a piece of good advice; listen to all the good advice you give yourself. They carry mountains of wisdom ;)

Markus, thank you for our discussions, article and buffer-sharing and for being a generally helpful and kind person. I do hope neither of us ever have to spend a night again on a Reykjavik airport bench, or any other airport, or bench for that matter. Johanna and Ingrid, I have had so much fun with you guys and I am glad to have got to know you. The years at DBB would not have been the same without your company. That includes you too Johan, and a special thank you for saving me from horrid computer-panic before presentations but also for spreading your positive vibes. Linda, thank you for being the kind empathic and encouraging person that you are and for generously sharing that with those around you. Your character is invaluable to any community. Olga, we never got around to having our lunch-discussions, which I’m sure would have been really interesting. Your many smiles have been much appreciated. Jacob, thanks for the fun times over the years and for being there to share the frustrations of teaching. Tobias, keep up your good spirit and continue to spread jokes and laughter around you. Johannes, we did not meet that much, but I always appreciated chatting with you and being in the peaceful zone that surrounds you. Agnes, you are fairly new in the group but it will definitely benefit from having you in it, both socially and scientific-wise. I wish you all the best for your years to come at DBB. Shu, we did not speak much but I want to thank you for all your scientific contributions. The efficiency with which you supply data is astonishing and fruitful for anyone involved.

To our previous, most recent and collaborating members; Gustav, Emelie, Wataru, Sylwia, Finja, Dan, Ram, Michael and Mama- thank you all for contributing to making our workplace pleasant and joyful. That includes all other members of DBB that I have had the pleasure of talking to, including those that I undoubtedly have forgotten to mention.

39

References

[1] P. R. Rich and A. Maréchal, “The mitochondrial respiratory chain,” Essays Biochem., vol. 47, pp. 1–23, 2010.

[2] R. Guo, J. Gu, S. Zong, M. Wu and M. Yang, “Structure and mechanism of mitochondrial electron transport chain,” Biomed. J., vol. 41, no. 1, pp. 9–20, 2018.

[3] V. Strogolova, A. Furness, M. Robb-McGrath, J. Garlich and R. A. Stuart, “Rcf1 and Rcf2, members of the hypoxia-induced gene 1 protein family, are critical components of the mitochondrial cytochrome bc1- cytochrome c oxidase supercomplex,” Mol. Cell. Biol., vol. 32, no. 8, pp. 1363–73, 2012.

[4] M. Vukotic, S. Oeljeklaus, S. Wiese, F. N. Vögtle, C. Meisinger, E. Meyer, A. Zieseniss, D. M. Katschinski, D. C. Jans, S. Jakobs, B. Warscheid, P. Rehling and M. Deckers, “Rcf1 mediates cytochrome oxidase assembly and respirasome formation, revealing heterogeneity of the enzyme complex,” Cell Metab., vol. 15, no. 3, pp. 336–47, 2012.

[5] Y. Chen, E. B. Taylor, N. Dephoure and A. Tonhato, “Identification of a protein mediating respiratory supercomplex stability,” Cell Metab, vol. 15, no. 3, pp. 348–360, 2012.

[6] M. Klein, S. Swinnen, J. M. Thevelein and E. Nevoigt, “Glycerol metabolism and transport in yeast and fungi: Established knowledge and ambiguities,” Environ. Microbiol., vol. 19, no. 3, pp. 878–893, 2017.

[7] K. Otterstedt, C. Larsson, R. M. Bill, A. Ståhlberg, E. Boles, S. Hohmann and L. Gustafsson, “Switching the mode of metabolism in the yeast Saccharomyces cerevisiae,” EMBO Rep., vol. 5, no. 5, pp. 532–537, 2004.

[8] J. V Gray, G. A. Petsko, G. C. Johnston, D. Ringe, R. A. Singer and M. Werner-washburne, “´Sleeping Beauty´: Quiescence in Saccharomyces cerevisiae” Microbiol. Mol. Biol. Rev., vol. 68, no. 2, pp. 187– 206, 2004.

[9] M. Werner-Washburne, S. Roy and G. S. Davidson, “Aging Research in Yeast,” Subcell Biochem, vol. 57, pp. 123–143, 2012.

[10] J. Garber Morales, G. P. Holmes-Hampton, R. Miao, Y. Guo, E. Münck and P. A. Lindahl, “Biophysical characterization of iron in mitochondria isolated from respiring and fermenting yeast,” Biochemistry, vol. 49, no. 26, pp. 5436–5444, 2010.

[11] C. R. Hackenbrock, “Ultrastructural bases for metabolically linked mechanical activity in mitochondria. II. Electron transport-linked ultrastructural transformations in mitochondria,” J. Cell Biol., vol. 37, no. 2, pp. 345–369, 1968.

[12] F. Fontanesi, I. C. Soto and A. Barrientos, “Cytochrome c Oxidase Biogenesis: New levels of Regulation,” IUBMB Life vol. 60, no. September, pp. 557–568, 2008.

[13] P. V. Burke and R. O. Poyton, “Structure/function of oxygen-regulated isoforms in cytochrome c oxidase,” J. Exp. Biol., vol. 201, pp. 1163–1175, 1998.

[14] T. Hon, A. Dodd, R. Dirmeier, N. Gorman, P. R. Sinclair, L. Zhang and R. O. Poyton, “A mechanism of oxygen sensing in yeast: Multiple oxygen-responsive steps in the heme biosynthetic pathway affect Hap1 activity,” J. Biol. Chem., vol. 278, no. 50, pp. 50771–50780, 2003.

40

[15] M. J. Hickman and F. Winston, “Heme levels switch the function of Hap1 of Saccharomyces cerevisiae between transcriptional activator and transcriptional repressor,” Mol. Cell. Biol., vol. 27, no. 21, pp. 7414–24, 2007.

[16] M. G. Baile and S. M. Claypool, “The power of yeast to model of the powerhouse of the cell,” Front Biosci (Landmark Ed.), no. 7, pp. 241–278, 2013.

[17] M. Trouillard, B. Meunier and F. Rappaport, “Questioning the functional relevance of mitochondrial supercomplexes by time-resolved analysis of the respiratory chain,” Proc. Natl. Acad. Sci. U. S. A., vol. 108, no. 45, pp. E1027-34, 2011.

[18] T. Z. Cui, A. Conte, J. L. Fox, V. Zara and D. R. Winge, “Modulation of the respiratory supercomplexes in yeast: Enhanced formation of cytochrome oxidase increases the stability and abundance of respiratory supercomplexes,” J. Biol. Chem., vol. 289, no. 9, pp. 6133–6141, 2014.

[19] T. Mogi, K. Saiki and Y. Anraku, “Biosynthesis and functional role of haem O and haem A, ” Mol. Microbiol., vol. 14, no. 3, pp. 391–398, 1994.

[20] H. Atamna, P. B. Walter and B. N. Ames, “The role of heme and iron-sulfur clusters in mitochondrial biogenesis, maintenance, and decay with age,” Arch. Biochem. Biophys., vol. 397, no. 2, pp. 345–353, 2002.

[21] M. H. Barros, F. G. Nobrega and A. Tzagoloff, “Mitochondrial ferredoxin is required for heme a synthesis in Saccharomyces cerevisiae,” J. Biol. Chem., vol. 277, no. 12, pp. 9997–10002, 2002.

[22] S. Iwata, J. W. Lee, K. Okada, J. K. Lee, M. Iwata, B. Rasmussen, T. A. Link, S. Ramaswamy and B. K. Jap, “Complete structure of the 11-subunit bovine mitochondrial cytochrome bc1 complex,” Science, vol. 281, no. 5373, pp. 64–71, 1998.

[23] J. Abramson, S. Riistama, G. Larsson, A. Jasaitis, M. Svensson-ek, L. Laakkonen, A. Puustinen, S. Iwata and M. Wikström, “The structure of the ubiquinol oxidase from Escherichia coli and its ubiquinone binding site,” Nat. Struct. Biol., vol. 7, no. 10, pp. 910-917, 2000.

[24] A. Maréchal, B. Meunier, D. Lee, C. Orengo and P. R. Rich, “Yeast cytochrome c oxidase: A model system to study mitochondrial forms of the haem-copper oxidase superfamily,” Biochim. Biophys. Acta - Bioenerg, vol. 1817, no. 4, pp. 620–628, 2012.

[25] T. Lo, J. Guillemette, G. Louie, M. Smith and G. Brayer, “Structural studies of the roles of residues 82 and 85 at the interactive face of cytochrome c,” Biochemistry, vol. 34, pp. 163–171, 1995.

[26] S. M. Mense and L. Zhang, “Heme: a versatile signaling molecule controlling the activities of diverse regulators ranging from transcription factors to MAP kinases,” Cell Res., vol. 16, no. 8, pp. 681–92, 2006.

[27] J. P. Hosler, S. Ferguson-Miller and D. A. Mills, “Energy Transduction: Proton transfer through the respiratory complexes,” Annu. Rev. Biochem., vol. 75, no. 1, pp. 165–187, 2006.

[28] V. R. I. Kaila, M. I. Verkhovsky and M. Wikström, “Proton-coupled electron transfer in cytochrome oxidase,” Chem. Rev., vol. 110, no. 12, pp. 7062–7081, 2010.

[29] I. Belevich and M. I. Verkhovsky, “Molecular mechanism of proton translocation by cytochrome c oxidase,” Antioxid. Redox Signal., vol. 10, no. 1, pp. 1–30, 2008.

41

[30] T. Hayashi, Y. Asano, Y. Shintani, H. Aoyama, H. Kioka, O. Tsukamoto, M. Hikita, K. Shinzawa-Itoh, K. Takafuji, S. Higo, H. Kato, S. Yamazaki, K. Matsuoka, A. Nakano, H. Asanuma, M. Asakura, T. Minamino, Y. Goto, T. Ogura, M. Kitakaze, I. Komuro, Y. Sakata, T. Tsukihara, S. Yoshikawa and S. Takashima, “Higd1a is a positive regulator of cytochrome c oxidase,” Proc. Natl. Acad. Sci. U. S. A., vol. 112, no. 5, pp. 1553–8, 2015.

[31] P. Jensen, M. T. Wilson, R. Aasa and B. G. Malmström, “Cyanide inhibition of cytochrome c oxidase. A rapid-freeze e.p.r. investigation,” Biochem. J., vol. 224, pp. 829–837, 1984.

[32] B. C, Hill, T. Brittain, D. G. Eglinton, P. M. Gadsby, C. Greenwood, P. Nicholls, J. Peterson, A. J. Thomson and T. C. Woon, “Low-spin ferric forms of cytochrome a3 in mixed-ligand and partially reduced cyanide- bound derivatives of cytochrome c oxidase,” Biochem. j., vol. 215, no. 1, pp. 57–66, 1983.

[33] D. Parul, G. Palmer and M. Fabian, “Ligand trapping by cytochrome c oxidase: Implications for gating at the catalytic center,” J. Biol. Chem., vol. 285, no. 7, pp. 4536–4543, 2010.

[34] N. Cassanova, K. M. O’Brien, B. T. Stahl, T. McClure and R. O. Poyton, “Yeast flavohemoglobin, a nitric oxide oxidoreductase, is located in both the cytosol and the mitochondrial matrix: Effects of respiration, anoxia, and the mitochondrial genome on its intracellular level and distribution,” J. Biol. Chem., vol. 280, no. 9, pp. 7645–7653, 2005.

[35] E. El Hammi, E. Warkentin, U. Demmer, N. M. Marzouki, U. Ermler and L. Baciou, “Active site analysis of yeast flavohemoglobin based on its structure with a small ligand or econazole,” FEBS, vol. 279, pp. 4565–4575, 2012.

[36] J. R. Fetter, J. Qian, J. Shapleigh, J. W. Thomas, A. García-Horsman, E. Schmidt, J. Hosler, G. T. Babcock, R. B. Gennis and S. Ferguson-Miller, “Possible proton relay pathways in cytochrome c oxidase,” Proc. Natl. Acad. Sci. U. S. A., vol. 92, no. 5, pp. 1604–1608, 1995.

[37] L. Varanasi and J. P. Hosler, “Subunit III-depleted cytochrome c oxidase provides insight into the process of proton uptake by proteins,” Biochim. Biophys. Acta - Bioenerg., vol. 1817, no. 4, pp. 545– 551, 2012.

[38] J. P. Hosler, J. P. Shapleigh, D. M. Mitchell, Y. Kim, M. A. Pressler, C. Georgiou, G. T. Babcock, J. O. Alben, S. Ferguson-Miller and R. B. Gennis, “Polar residues in helix VIII of subunit I of cytochrome c oxidase influence the activity and the structure of the active site,” Biochemistry, vol. 35, no. 33, pp. 10776– 10783, 1996.

[39] L. Banci, I. Bertini, S. Ciofi-Baffoni, T. Hadjiloi, M. Martinelli and P. Palumaa, “Mitochondrial copper(I) transfer from Cox17 to Sco1 is coupled to electron transfer,” Proc. Natl. Acad. Sci., vol. 105, no. 19, pp. 6803–6808, 2008.

[40] K. Hirata, K. Shinzawa-Itoh, N. Yano, S. Takemura, K. Kato, M. Hatanaka, K. Muramoto, T. Kawahara, T. Tsukihara, E. Yamashita, K. Tono, G. Ueno, T. Hikima, H. Murakami, Y. Inubushi, M. Yabashi, T. Ishikawa, M. Yamamoto, T. Ogura, Takashi, H. Sugimoto, J-R. Shen, S. Yoshikawa and H. Ago, “Determination of damage-free crystal structure of an X-ray-sensitive protein using an XFEL,” Nat. Methods, vol. 11, no. 7, pp. 734-736, 2014.

[41] S. Döpner, P. Hildebrandt, F. I. Rosell, A. G. Mauk, M. Von Walter, G. Buse and T. Soulimane, “The structural and functional role of lysine residues in the binding domain of cytochrome c in the electron transfer to cytochrome c oxidase,” Eur. J. Biochem. vol. 391, pp. 379–391, 1999.

[42] Y. Zhen, C. W. Hoganson, G. T. Babcock and S. Ferguson-miller, “Definition of the interaction domain for cytochrome c on cytochrome c oxidase,” J. Biol. Chem., vol. 274, no. 53, pp. 38032–38041, 1999.

42

[43] T. Tsukihara, H. Aoyama, E. Yamashita, T. Tomizaki, H. Yamaguchi, K. Shinzawa-Itoh, R. Nakashima, R. Yaono and S. Yoshikawa, “The whole structure of the 13-subunit oxidized cytochrome c oxidase at 2.8 A,” Science, vol. 272, no. 31, pp. 1136–1144, 1996.

[44] L. Varanasi, D. Mills, A. Murphree, J. Gray, C. Purser, R. Baker and J. Hosler, “Altering conserved lipid binding sites in cytochrome c oxidase of Rhodobacter sphaeroides perturbs the interaction between subunits I and III and promotes suicide inactivation of the enzyme,” Biochemistry, vol. 45, no. 50, pp. 14896–14907, 2006.

[45] V. Sharma, P. Ala-Vannesluoma, I. Vattulainen, M. Wikström and T. Róg, “Role of subunit III and its lipids in the molecular mechanism of cytochrome c oxidase,” Biochim. Biophys. Acta - Bioenerg., vol. 1847, no. 8, pp. 690–697, 2015.

[46] S. Riistama, A. Puustinen, A. García-Horsman, S. Iwata, H. Michel and M. Wikström, “Channelling of dioxygen into the respiratory enzyme,” Biochim. Biophys. Acta - Bioenerg., vol. 1275, no. 1–2, pp. 1–4, 1996.

[47] T. Althoff, D. J. Mills, J. L. Popot and W. Kühlbrandt, “Arrangement of electron transport chain components in bovine mitochondrial supercomplex I1III2IV1,” EMBO J., vol. 30, no. 22, pp. 4652–4664, 2011.

[48] J. Das, S. T. Miller and D. L. Stern, “Comparison of diverse protein sequences of the nuclear-encoded subunits of cytochrome c oxidase suggests conservation of structure underlies evolving functional sites,” Mol. Biol. Evol., vol. 21, no. 8, pp. 1572–1582, 2004.

[49] J. Liu and A. Barrientos, “Transcriptional regulation of yeast oxidative phosphorylation hypoxic genes by ,” Antioxid. Redox Signal., vol. 19, no. 16, pp. 1916–1927, 2013.

[50] M. R. Hodge, G. Kim, K. Singh and M. G. Cumsky, “Inverse regulation of the yeast COX5 genes by oxygen and heme,” Mol. Cell. Biol., vol. 9, no. 5, pp. 1958–1964, 1989.

[51] P. R. Castello, P. S. David, T. McClure, Z. Crook and R. O. Poyton, “Mitochondrial cytochrome oxidase produces nitric oxide under hypoxic conditions: Implications for oxygen sensing and hypoxic signaling in eukaryotes,” Cell Metab., vol. 3, no. 4, pp. 277–287, 2006.

[52] A. F. Dubinski, R. Camasta, T. G. B. Soule, B. H. Reed and D. M. Glerum, “Consequences of cytochrome c oxidase assembly defects for the yeast stationary phase,” Biochim. Biophys. Acta - Bioenerg., vol. 1859, no. 6, pp. 445–458, 2018.

[53] M. Bourens, D. V. Dabir, H. L. Tienson, I. Sorokina, C. M. Koehler and A. Barrientos, “Role of twin Cys- Xaa9-Cys motif cysteines in mitochondrial import of the cytochrome c oxidase biogenesis factor Cmc1,” J. Biol. Chem., vol. 287, no. 37, pp. 31258–31269, 2012.

[54] I. Liko, M. T. Degiacomi, S. Mohammed, S. Yoshikawa, C. Schmidt and C. V. Robinson, “Dimer interface of bovine cytochrome c oxidase is influenced by local posttranslational modifications and lipid binding,” Proc. Natl. Acad. Sci., vol. 113, no. 29, pp. 8230–8235, 2016.

[55] S. Ferguson-Miller, D. L. Brautigan and E. Margoliash, “Correlation of the kinetics of electron transfer activity various eukaryotic cytochromes c with binding to mitochondrial cytochrome c oxidase”J. Biol. Chem., vol.251, no.4, pp.1104-15, 1976.

[56] R. Bisson, G. Schiavo and C. Montecucco, “ATP induces conformational changes in mitochondrial cytochrome c oxidase. Effect on the cytochrome c binding site,” J. Biol. Chem., vol. 262, no. 13, pp. 5992–5998, 1987.

43

[57] L. A. Allen, X. J. Zhao, W. Caughey and R. O. Poyton, “Isoforms of yeast cytochrome c oxidase subunit V affect the binuclear reaction center and alter the kinetics of interaction with the isoforms of yeast cytochrome c,” Journal of Biological Chemistry, vol. 270, no. 1. pp. 110–118, 1995.

[58] J. E. McEwen, C. Ko, B. Kloeckner-Gruissem and R. O. Poyton, “Nuclear functions required for cytochrome c oxidase biogenesis in Saccharomyces cerevisiae. Characterization of mutants in 34 complementation groups,” J. Biol. Chem., vol. 261, no. 25, pp. 11872–11879, 1986.

[59] L. G. J. Nijtmans, J. W. Taanman, A. O. Muijsers, D. Speijer and C. Van Den Bogert, “Assembly of cytochrome c oxidase in cultured human cells,” Eur. J. Biochem., vol. 254, no. 2, pp. 389–394, 1998.

[60] E. Mileykovskaya, P. A. Penczek, J. Fang, V. K. P. S. Mallampalli, G. C. Sparagna and W. Dowhan, “Arrangement of the respiratory chain complexes in Saccharomyces cerevisiae supercomplex III2IV2 revealed by single particle cryo-electron microscopy,” J. Biol. Chem., vol. 287, no. 27, pp. 23095–23103, 2012.

[61] R. Dodia, B. Meunier, C. W. M. Kay and P. R. Rich, “Comparisons of subunit 5A and 5B isoenzymes of yeast cytochrome c oxidase,” Biochem. J., vol. 464, no. 3, pp. 335–342, 2014.

[62] R. A. Waterland, A. Basu, B. Chance and R. O. Poyton, “The isoforms of yeast cytochrome c oxidase subunit V alter the in vivo kinetic properties of the holoenzyme,” J. Biol. Chem., vol. 266, no. 7, pp. 4180–4186, 1991.

[63] A. Musatov and N. C. Robinson, “Bound cardiolipin is essential for cytochrome c oxidase proton translocation,” Biochimie, vol. 105, pp. 159–164, 2014.

[64] E. Sedlák and N. C. Robinson, “Phospholipase A2 digestion of cardiolipin bound to bovine cytochrome c oxidase alters both activity and quaternary structure,” Biochemistry, vol. 38, no. 45, pp. 14966–14972, 1999.

[65] A. E. P. LaMarche, M. I. Abate, S. H. P. Chan and B. L. Trumpower, “Isolation and characterization of COX12, the nuclear gene for a previously unrecognized subunit of Saccharomyces cerevisiae cytochrome c oxidase,” J. Biol. Chem., vol. 267, no. 31, pp. 22473–22480, 1992.

[66] A. Ghosh, A. T. Pratt, S. Soma, S. G. Theriault, A. T. Griffin, P. P. Trivedi and V. M. Gohil, “Mitochondrial genes COA6, COX6B and SCO2 have overlapping roles in COX2 biogenesis,” Hum. Mol. Genet., vol. 25, no. 4, pp. 660–671, 2016.

[67] M. Hüttemann, S. Jaradat, and L. I. Grossman, “Cytochrome c oxidase of mammals contains a testes- specific isoform of subunit VIb - The counterpart to testes-specific cytochrome c?,” Mol. Reprod. Dev., vol. 66, no. 1, pp. 8–16, 2003.

[68] U. D. Vempati, X. Han and C. T. Moraes, “Lack of cytochrome c in mouse fibroblasts disrupts assembly/stability of respiratory complexes I and IV,” J. Biol. Chem., vol. 284, no. 7, pp. 4383–4391, 2009.

[69] B. Hennig, “Change of cytochrome c structure during development of the mouse,” Eur. J. Biochem., vol. 55, no. 1, pp. 167–83, 1975.

[70] J. W. Taanman and R. A. Capaldi, “Subunit VIa of yeast cytochrome c oxidase is not necessary for assembly of the enzyme complex but modulates the enzyme activity. Isolation and characterization of the nuclear-coded gene,” J. Biol. Chem., vol. 268, no. 25, pp. 18754–18761, 1993.

[71] B. Beauvoit, O. Bunoust, B. Guérin and M. Rigoulet, “ATP-regulation of cytochrome oxidase in yeast mitochondria. Role of subunit Vla,” Eur. J. Biochem., vol. 263, no. 1, pp. 118–127, 1999.

44

[72] C. Arnarez, S. J. Marrink and X. Periole, "Identification of cardiolipin binding sites on cytochrome c oxidase at the entrance of proton channels", Sci. rep., vol. 3, pp. 1–8, 2013.

[73] M. Malecki, D. A. Bitton, M. Rodríguez-lópez, C. Rallis, N. G. Calavia,G. C. Smith and J. Bähler, “Functional and regulatory profiling of energy metabolism in fission yeast,” Genome Biol., pp. 1–18, 2016.

[74] E. Antonini, M. Brunori, A. Colosimo, C. Greenwoodt and M. T. Wilson, “Oxygen "pulsed" cytochrome c oxidase: Functional properties and catalytic relevance," Biochemistry, vol. 74, no. 8, pp. 3128–3132, 1977.

[75] A. Moody John, U. Brandt and P. R. Rich, “Single electron reduction of ‘ slow ’ and ‘ fast ’ cytochrome- c,” FEBS, vol. 293, pp. 101–105, 1991.

[76] M. Assempour, D. Lim and B. C. Hill, “Electron transfer kinetics during the reduction and turnover of the cytochrome caa3 complex from Bacillus subtilis,” Biochemistry, vol. 37, pp. 9991–9998, 1998.

[77] B. C. Hill and J. Peterson, “Spectral and cyanide binding properties of the cytochrome aa3 (600 nm) complex from Bacillus subtilis,” Arch. Biochem. Biophys., vol. 350, no. 2, pp. 273–282, 1998.

[78] J. Moody,"’ As prepared ’ forms of fully oxidised haem/Cu terminal oxidases,” Biochim. Biophys. Acta vol. 8, pp. 6–20, 1996.

[79] S. Zaidi, M. I. Hassan, A. Islam and F. Ahmad, “The role of key residues in structure, function, and stability of cytochrome c,” Cell. Mol. Life Sci., vol. 71, no. 2, pp. 229–255, 2013.

[80] S. R. N. Solmaz and C. Hunte, “Structure of complex III with bound cytochrome c in reduced state and definition of a minimal core interface for electron transfer,” J. Biol. Chem., vol. 283, no. 25, pp. 17542– 17549, 2008.

[81] D. A. Pearce and F. Sherman, “Differential ubiquitin-dependent degradation of the yeast apo- cytochrome c isozymes,” J. Biol. Chem., vol. 272, no. 50, pp. 31829–31836, 1997.

[82] A. Barrientos, D. Pierre, J. Lee and A. Tzagoloff, “Cytochrome oxidase assembly does not require catalytically active cytochrome c,” J. Biol. Chem., vol. 278, no. 11, pp. 8881–8887, 2003.

[83] S. Zhou, P. Pettersson, J. Huang, J. Sjöholm, D. Sjöstrand, R. Pomès, M. Högbom, P. Brzezinski, L. Mäler and P. Ädelroth, “Solution NMR structure of yeast Rcf1, a protein involved in respiratory supercomplex formation,” Proc. Natl. Acad. Sci., vol. 115, no. 12, pp. 3048–3053, 2018.

[84] B. Moreno-Beltrán, A. Guerra-Castellano, A. Díaz-Quintana, R. Del Conte, S. M. García-Mauriño, S. Díaz- Moreno, K. González-Arzola, C. Santos-Ocaña, A. Velázquez-Campoy, M. A. De la Rosa, P. Turano and I. Díaz-Moreno, “Structural basis of mitochondrial dysfunction in response to cytochrome c phosphorylation at tyrosine 48,” Proc. Natl. Acad. Sci., vol. 114, no. 15, pp. E3041–E3050, 2017.

[85] J. Garlich, V. Strecker, I. Wittig and R. A. Stuart, “Mutational analysis of the QRRQ motif in the yeast hig1 type 2 protein Rcf1 reveals a regulatory role for the cytochrome c oxidase complex,” J. Biol. Chem., vol. 292, no. 13, pp. 5216–5226, 2017.

[86] K. Bihlmaier, N. Mesecke, C. Kloeppel, and J. M. Herrmann, “The disulfide relay of the intermembrane space of mitochondria: An oxygen-sensing system?,” Ann. N. Y. Acad. Sci., vol. 1147, pp. 293–302, 2008.

[87] K. Hell, “The Erv1-Mia40 disulfide relay system in the intermembrane space of mitochondria,” Biochim. Biophys. Acta - Mol. Cell Res., vol. 1783, no. 4, pp. 601–609, 2008.

45

[88] S. E. Neal, D. V. Dabir, H. L. Tienson, D. M. Horn, K. Glaeser, L. Ogozalek, R. Rachel, A. Barrientos and C. M. Koehler, “Mia40 protein serves as an electron sink in the Mia40-Erv1 import pathway,” J. Biol. Chem., vol. 290, no. 34, pp. 20804–20814, 2015.

[89] L. Kowalski, P. Bragoszewski, A. Khmelinskii, E. Glow, M. Knop and A. Chacinska, “Determinants of the cytosolic turnover of mitochondrial intermembrane space proteins,” BMC Biol., vol. 16, no. 1, pp. 1–22, 2018.

[90] A. I. P. M. de Kroon, P. J. Rijken and C. H. De Smet, “Checks and balances in membrane phospholipid class and acyl chain , the yeast perspective,” Prog. Lipid. Res., vol. 52, no. 4, pp. 374–94, 2013.

[91] S. Ramadurai, R. Duurkens, V. V Krasnikov, and B. Poolman, “Lateral diffusion of membrane proteins: Consequences of hydrophobic mismatch and lipid composition,” Biophys. J., vol. 99, no. 5, pp. 1482– 1489, 2010.

[92] G. Benard and R. Rossignol, “Ultrastructure of the mitochondrion and its bearing on function and bioenergetics,” Antioxid. Redox Signal., vol. 10, no. 8, pp. 1313–1342, 2008.

[93] H. Schägger and K. Pfeiffer, “Supercomplexes in the respiratory chains of yeast and mammalian mitochondria,” EMBO J., vol. 19, no. 8, pp. 1777-1783, 2000.

[94] L. Böttinger, S. E. Horvath, T. Kleinschroth, C. Hunte, G. Daum, N. Pfanner and T. Becker, “Phosphatidylethanolamine and cardiolipin differentially affect the stability of mitochondrial respiratory chain supercomplexes,” J. Mol. Biol., vol. 423, no. 5, pp. 677–686, 2012.

[95] E. Van Den Brink-Van Der Laan, J. Antoinette Killian and B. De Kruijff, “Nonbilayer lipids affect peripheral and integral membrane proteins via changes in the lateral pressure profile,” Biochim. Biophys. Acta - Biomembr., vol. 1666, no. 1–2, pp. 275–288, 2004.

[96] A. G. Lee, “How lipids affect the activities of integral membrane proteins,” Biochim. Biophys. Acta - Biomembr., vol. 1666, no. 1–2, pp. 62–87, 2004.

[97] M. Sathappa and N. N. Alder, “The ionization properties of cardiolipin and its variants in model bilayers,” Biochim. Biophys. Acta - Biomembr., vol. 1858, no. 6, pp. 1362–1372, 2016.

[98] A. Rietveld, T. J. J. M. van Kemenade, T. Hak, A. J. Verkleij and B. De Kruijff, “The effect of cytochrome c oxidase on lipid polymorphism of model membranes containing cardiolipin,” Eur. J. Biochem., vol. 164, no. 1, pp. 137–140, 1987.

[99] B. H. Robinson, “Human cytochrome oxidase deficiency,” Pediatr. Res., vol. 48, no. 5, pp. 581–585, 2000.

[100] N. Khalifat, J.-B. Fournier, M. I. Angelova and N. Puff, “Lipid packing variations induced by pH in cardiolipin-containing bilayers: The driving force for the cristae-like shape instability,” Biochim. Biophys. Acta, vol. 1808, no. 11, pp. 2724–33, 2011.

[101] V. M. Gohil, P. Hayes, S. Matsuyama, H. Schägger, M. Schlame and M. L. Greenberg, “Cardiolipin biosynthesis and mitochondrial respiratory chain function are interdependent,” J. Biol. Chem., vol. 279, no. 41, pp. 42612–8, 2004.

[102] S. Jakovcic, G. S. Getz, M. Rabinowitz, H. Jakob and H. Swift, “Cardiolipin content of wild type and mutant in relation to mitochondrial function and development,” J. Cell Biol., vol. 48, no. 3, pp. 490–502, 1971.

46

[103] X. Zhang, B. Tamot, C. Hiser, G. E. Reid, C. Benning and S. Ferguson-Miller, “Cardiolipin deficiency in Rhodobacter sphaeroides alters the lipid profile of membranes and of crystallized cytochrome oxidase, but structure and function are maintained,” Biochemistry, vol. 50, no. 19, pp. 3879–3890, 2011.

[104] S.-C. Chang, P. N. Heacock, E. Mileykovskaya, D. R. Voelker and W. Dowhan, “Isolation and Characterization of the Gene (CLS1) Encoding Cardiolipin Synthase in Saccharomyces cerevisiae,” J. Biol. Chem., vol. 273, no. 24, pp. 14933–14941, 1998.

[105] F. Jiang, H. S. Rizavi and M. L. Greenberg, “Cardiolipin is not essential for the growth of Saccharomyces cerevisiae on fermentable or non-fermentable carbon sources,” Mol. Microbiol., vol. 26, no. 3, pp. 481– 91, 1997.

[106] A. Musatov, R. Varhač, J. P. Hosler and E. Sedlák, “Delipidation of cytochrome c oxidase from Rhodobacter sphaeroides destabilizes its quaternary structure,” Biochimie, vol. 125, pp. 23–31, 2016.

[107] F. Jiang, M. T. Ryan, M. Schlame, M. Zhao, Z. Gu, M. Klingenberg, N. Pfanner and M. L. Greenberg, “Absence of cardiolipin in the crd1 null mutant results in decreased mitochondrial membrane potential and reduced mitochondrial function,” J. Biol. Chem., vol. 275, no. 29, pp. 22387–22394, 2000.

[108] C. Osman, D. R. Voelker and T. Langer, “Making heads or tails of phospholipids in mitochondria,” J. Cell Biol., vol. 192, no. 1, pp. 7–16, 2011.

[109] S. Checover, E. Nachliel, N. A. Dencher and M. Gutman, “Mechanism of proton entry into the cytoplasmic section of the proton- conducting channel of bacteriorhodopsin,” Biochemistry, vol. 36, no. 45, pp. 13919–13928, 1997.

[110] A. Namslauer, A. Aagaard, A. Katsonouri and P. Brzezinski, “Intramolecular proton-transfer reactions in a membrane-bound proton pump: The effect of pH on the peroxy to ferryl transition in cytochrome c oxidase,” Biochemistry, vol. 42, no. 6, pp. 1488–1498, 2003.

[111] L. Öjemyr, T. Sanden, J. Widengren and P. Brzezinski, “Lateral proton transfer between the membrane and a membrane protein,” Biochemistry, vol. 48, pp. 2173–2179, 2009.

[112] L. Näsvik Öjemyr, C. Von Ballmoos, R. B. Gennis, S. G. Sligar and P. Brzezinski, “Reconstitution of respiratory oxidases in membrane nanodiscs for investigation of proton-coupled electron transfer,” FEBS Lett., vol. 586, no. 5, pp. 640–645, 2012.

[113] J. A. Letts, K. Fiedorczuk and L. A. Sazanov, “The architecture of respiratory supercomplexes,” Nature, vol. 537, no. 7622, pp. 644–648, 2016.

[114] J. S. Sousa, D. J. Mills, J. Vonck and W. Kuhlbrandt, “Functional asymmetry and electron flow in the bovine respirasome,” Elife, vol. 5, pp. 1–17, 2016.

[115] M. Wu, J. Gu, R. Guo, Y. Huang and M. Yang, “Structure of mammalian respiratory supercomplex I1III2IV1,” Cell, vol. 167, no. 6, pp. 1598–1609, 2016.

[116] J. N. Blaza, R. Serreli, A. J. Y. Jones, K. Mohammed and J. Hirst, “Kinetic evidence against partitioning of the ubiquinone pool and the catalytic relevance of respiratory-chain supercomplexes,” Proc. Natl. Acad. Sci. U. S. A., vol. 111, no. 44, pp. 15735–40, 2014.

[117] J. A. Enriquez and G. Lenaz, “Coenzyme q and the respiratory chain: coenzyme q pool and mitochondrial supercomplexes,” Mol. Syndromol., vol. 5, no. 3–4, pp. 119–40, 2014.

47

[118] C. A. Porras and Y. Bai, “Respiratory supercomplexes: Plasticity and implications,” Front. Biosci (Landmark Ed), pp. 621–634, 2015.

[119] S. Helling, M. Hüttemann, R. Ramzan, S. H. Kim, I. Lee, T. Müller, E. Langenfeld, H. E. Meyer, B. Kadenbach, S. Vogt and K. Marcus,, “Multiple phosphorylations of cytochrome c oxidase and their functions,” Proteomics, vol. 12, no. 7, pp. 950–959, 2012.

[120] S. Arnold, “The power of life-cytochrome c oxidase takes center stage in metabolic control, cell signalling and survival,” Mitochondrion, vol. 12, no. 1, pp. 46–56, 2012.

[121] H. Schägger and K. Pfeiffer, “The ratio of oxidative phosphorylation complexes I-V in bovine heart mitochondria and the composition of respiratory chain supercomplexes,” JBC, no. 858, 2001.

[122] C. Greggio, P. Jha, S. S. Kulkarni, S. Lagarrigue, N. T. Broskey, M. Boutant, X. Wang, S. Conde Alonso, E. Ofori, J. Auwerx, C. Cantó and F. Amati, “Enhanced respiratory chain supercomplex formation in response to exercise in human skeletal muscle,” Cell Metab., vol. 25, no. 2, pp. 301–311, 2017.

[123] K. Pfeiffer, V. Gohil, R. Stuart, RC. Hunte, U. Brandt, M. L. Greenberg and H. Schägger, “Cardiolipin stabilizes respiratory chain supercomplexes,” J. Biol. Chem., vol. 278, no. 52, pp. 52873–80, 2003.

[124] C. Rydström Lundin, C. von Ballmoos, M. Ott, P. Ädelroth and P. Brzezinski, “Regulatory role of the respiratory supercomplex factors in Saccharomyces cerevisiae,” Proc. Natl. Acad. Sci., vol. 113, no. 31, pp. E4476–E4485, 2016.

[125] M. K. Dienhart and R. A. Stuart, “The yeast Aac2 protein exists in physical association with the cytochrome bc 1 -COX supercomplex and the TIM23 machinery,” Mol. Biol. Cell vol. 19, pp. 3934–3943, 2008.

[126] K. E. Kwast, P. V. Burke, B. T. Staahl and R. O. Poyton, “Oxygen sensing in yeast: Evidence for the involvement of the respiratory chain in regulating the transcription of a subset of hypoxic genes,” vol. 96, pp. 5446–5451, 1999.

[127] B. Hoffmann, A. Stöckl, M. Schlame, K. Beyer and M. Klingenberg, “The reconstituted ADP/ATP carrier activity has an absolute requirement for cardiolipin as shown in cysteine mutants,” J. Biol. Chem., vol. 269, no. 3, pp. 1940–1944, 1994.

[128] S. M. Claypool, Y. Oktay, P. Boontheung, J. A. Loo and C. M. Koehler, “Cardiolipin defines the interactome of the major ADP/ATP carrier protein of the mitochondrial inner membrane,” J. Cell Biol., vol. 182, no. 5, pp. 937–950, 2008.

[129] S. Bazán, E. Mileykovskaya, V. K. P. S. Mallampalli, P. Heacock, G. C. Sparagna and W. Dowhan, “Cardiolipin-dependent reconstitution of respiratory supercomplexes from purified Saccharomyces cerevisiae complexes III and IV,” J. Biol. Chem., vol. 288, pp. 401–11, 2013.

[130] T. Wenz, R. Hielscher, P. Hellwig, H. Schägger, S. Richers and C. Hunte, “Role of phospholipids in respiratory cytochrome bc1 complex catalysis and supercomplex formation,” Biochim. Biophys. Acta - Bioenerg., vol. 1787, no. 6, pp. 609–616, 2009.

[131] S. M. Claypool, “Cardiolipin, a critical determinant of mitochondrial carrier protein assembly and function,” Biochim. Biophys. Acta - Biomembr., vol. 1788, no. 10, pp. 2059–2068, 2009.

48