Regulation of mitochondrial respiration and apoptosis throughcell signaling: cytochrome c oxidase and cytochrome c inischemia/reperfusion injury and inflammation

Maik Hüttemanna,b,*, Stefan Hellingc, Thomas H. Sandersonc,b, Christopher Sinklera,Lobelia Samavatia,e, Gargi Mahapatraa, Ashwathy Varughesea, Guorong Lua, Jenney Liua,Rabia Ramzanf, Sebastian Vogtf, Lawrence I. Grossmana,b, Jeffrey W. Doana, KatrinMarcusc, and Icksoo Leea

aCenter for Molecular Medicine and Genetics, Wayne State University School of Medicine,Detroit, MI 48201, USAbCardiovascular Research Institute, Wayne State University School of Medicine, Detroit, MI48201, USAcMedizinisches Proteom-Center, Funktionelle Proteomik, Ruhr-Universität Bochum, Bochum,GermanydDepartment of Emergency Medicine, Wayne State University School of Medicine, Detroit, MI48201, USAeDepartment of Medicine, Division of Pulmonary/Critical Care and Sleep Medicine, Wayne StateUniversity School of Medicine, Detroit, MI 48201, USAfCardiovascular Research Laboratory at the Biomedical Research Center, Philipps-University,D-35032 Marburg, Germany

AbstractCytochrome c (Cytc) and cytochrome c oxidase (COX) catalyze the terminal reaction of themitochondrial electron transport chain (ETC), the reduction of oxygen to water. This irreversiblestep is highly regulated, as indicated by the presence of tissue-specific and developmentallyexpressed isoforms, allosteric regulation, and reversible phosphorylations, which are found in bothCytc and COX. The crucial role of the ETC in health and disease is obvious since it, together withATP synthase, provides the vast majority of cellular energy, which drives all cellular processes.However, under conditions of stress, the ETC generates reactive oxygen species (ROS), whichcause cell damage and trigger death processes. We here discuss current knowledge of theregulation of Cytc and COX with a focus on cell signaling pathways, including cAMP/proteinkinase A and tyrosine kinase signaling. Based on the crystal structures we highlight all identifiedphosphorylation sites on Cytc and COX, and we present a new phosphorylation site, Ser126 onCOX subunit II. We conclude with a model that links cell signaling with the phosphorylation stateof Cytc and COX. This in turn regulates their enzymatic activities, the mitochondrial membranepotential, and the production of ATP and ROS. Our model is discussed through two distincthuman pathologies, acute inflammation as seen in sepsis, where phosphorylation leads to strong

© 2011 Elsevier B.V. All rights reserved.*Corresponding author. Tel.: +1 313-577-9150; fax: +1 313-577-5218. mhuttema@med.wayne.edu (M. Hüttemann).Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to ourcustomers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review ofthe resulting proof before it is published in its final citable form. Please note that during the production process errors may bediscovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

NIH Public AccessAuthor ManuscriptBiochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

Published in final edited form as:Biochim Biophys Acta. 2012 April ; 1817(4): 598–609. doi:10.1016/j.bbabio.2011.07.001.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

COX inhibition followed by energy depletion, and ischemia/reperfusion injury, where hyperactiveETC complexes generate pathologically high mitochondrial membrane potentials, leading toexcessive ROS production. Although operating at opposite poles of the ETC activity spectrum,both conditions can lead to cell death through energy deprivation or ROS-triggered apoptosis.

KeywordsApoptosis; Cancer; Neurodegenerative diseases; Noonan syndrome; Oxidative phosphorylation;Sepsis; Stroke

1. IntroductionIn multicellular organisms communication between cells and within subcellularcompartments is essential to coordinate diverse cellular and overall organ functions. Thecommunication system in higher organisms utilizes a battery of mechanisms such ashormonal regulation through the bloodstream, electrical conduction to transmit signals inneurons, and small molecules or ions inside the cell such as cAMP and calcium. Cellsignaling pathways streamline communication within and among cells. Central roles playreceptors, kinases, phosphatases, and scaffolding proteins that together generate a specificsignal that eventually leads to the chemical modification, i.e., phosphorylation ordephosphorylation of downstream target proteins. A key cellular unit, the mitochondrialoxidative phosphorylation (OxPhos) machinery, is currently being explored for evidence ofregulation by cell signaling pathways.

The primary role of OxPhos is the provision of energy, which drives all cellular processes.An average human synthesizes 65 kg of ATP from ADP and phosphate per day [1],highlighting the importance of this process for organismal health. A secondary role is theparticipation of OxPhos in apoptosis, a concerted sequence of events including changes inOxPhos activity, mitochondrial membrane potential (ΔΨm) hyperpolarization, reactiveoxygen species (ROS) production, release of cytochrome c, formation of the apoptosome,and execution of apoptosis via caspase activation. Given the dual role of OxPhos in cellularlife and death decisions, tight regulation is warranted.

OxPhos generates more than 15 times the amount of ATP compared to anaerobic glycolysis.OxPhos consists of the electron transport chain (ETC) and ATP synthase. The ETC consistsof NADH dehydrogenase (complex I), succinate dehydrogenase (complex II), the non-protein electron carrier ubiquinone, bc1-complex (complex III), cytochrome c (Cytc), andcytochrome c oxidase (COX; complex IV). Electrons enter the ETC primarily via complex Ifrom NADH. Complex II links OxPhos with the citric acid cycle and feeds electrons derivedfrom succinate directly into the ubiquinone/ubiquinol pool. Except for complex II, whichdoes not function as a proton pump, the other ETC complexes pump protons from the matrixinto the intermembrane space, generating the mitochondrial membrane potential (ΔΨm).ΔΨm is utilized by ATP synthase (complex V) to synthesize ATP from ADP and phosphate,driven by the backflow of protons from the mitochondrial intermembrane space to thematrix. In contrast to many unicellular organisms including yeast, which can grow andsurvive in the absence of oxygen, animals fully depend on aerobic energy metabolism, withfew exceptions restricted to parts of the organism, such as the fetus during the first half ofpregnancy [2], and cancer, where the exception is known as the Warburg effect [3, 4].

Hüttemann et al. Page 2

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

2. Regulation of mitochondrial OxPhosIn the following sections we review the multifaceted regulation of mitochondrial respirationthrough respiratory control, tissue-specific isoforms, allosteric regulation, and post-translational modifications, which will be the basis for the subsequent discussion of the roleof OxPhos in selected human diseases.

2.1. The traditional view of the regulation of mitochondrial OxPhos: respiratory controlDepending on the activity of the cell, cellular energy demand varies widely and energyproduction thus has to be adjusted accordingly. The traditional principle by which OxPhos iscontrolled is called respiratory control, first described by Lardy and Wellman [5] asstimulation of mitochondrial respiration by a phosphate acceptor (e.g., ADP), and laterextended by Chance and Williams [6] as an inhibition of respiration during transition fromstate 3 (high rate) to state 4 (low rate), when ADP is converted to ATP.

When cells utilize energy ATP is converted into ADP, which is converted back into ATP byATP synthase using the proton gradient. This reduction in ΔΨm allows the proton pumps toresume electron transfer and to pump protons across the inner membrane. When cells are atrest and the vast majority of ADP has been converted into ATP, ATP synthesis slows, andΔΨm builds up, which inhibits the proton pumps and thus mitochondrial respiration. Whenworking with isolated mitochondria two extreme states are often reported to characterizemitochondria function: state 3 respiration after the addition of ADP and phosphate in thepresence of complex I or II substrates, which gives maximal respiration rates, and state 4,when all ADP has been converted into ATP, which gives low respiration rates (state 1 isbaseline respiration of mitochondria in the absence of substrates; state 2 is defined asrespiration after the addition of ETC substrates such as succinate but in the absence ofexogenous ADP). Analysis of respiratory states of isolated mitochondria is useful todetermine the number of ATP molecules produced per oxygen consumed, which is referredto as the P/O ratio; the ratio of state 3/state 4 respiration serves as a measure ofmitochondrial “tightness” or coupling and is referred to as the respiratory control ratio.

Respiratory control is the most basic level of regulation and is defined by substrateconcentrations. For example, substrates for complexes I and II, and ADP and phosphate forATP synthase, and the reaction intermediates and products, such as ΔΨm generated by theproton pumps and ATP generated by ATP synthase all affect the respiration rate. However,additional regulatory mechanisms exist, as recently established for mammalian COX andCytc, and these will be discussed.

2.2. Tissue-specific isoforms of cytochrome c oxidase and cytochrome cIn contrast to OxPhos complexes I, II, III, and V, for which no tissue specific isoforms havebeen reported, both Cytc and COX occur as distinct tissue-specific isozymes in mammals.This suggests that the terminal step of the ETC, i.e., the electron transfer from Cytc to O2 viaCOX, has to be functionally adapted to tissue-specific energy needs.

2.2.1. Cytochrome c—Mammalian Cytc is a one-electron carrier that shuttles electronsfrom bc1 complex to COX. It contains a covalently attached heme group and 104 aminoacids (Fig. 1, top left). Cytc is highly positively charged with a pI of 9.6 and is located in themitochondrial intermembrane space, where it is associated with negatively chargedphospholipids of the inner membrane, in particular cardiolipin.

In rodents, Cytc occurs as a testes-specific isoform (Cytc-T), which is expressed in thegerminal epithelial cells [7], and a somatic isoform (Cytc), which is expressed in all other

Hüttemann et al. Page 3

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

tissues. In humans there is only a single functional Cytc gene because the Cytc testis isoformgene was lost during primate evolution. The former Cytc-T in humans is now a non-transcribed pseudogene [8, 9].

In mice the two Cytc isoforms share 86% sequence identity but their enzymatic functionsare clearly distinct as discussed elsewhere [10]. For example, Cytc-T shows a threefoldincreased activity to reduce hydrogen peroxide compared with the somatic isoform, but alsoshows an approximately fourfold increased ability to induce apoptosis [11]. The authorsproposed that these functional changes may prevent radical damage of sperm and at thesame time serve as a selective agent by initiating apoptosis in damaged sperm. In addition toits role in respiration, as a ROS scavenger under healthy conditions, and as an initiator ofapoptosis, Cytc participates in several other activities including Erv1-Mia40-coupled proteinimport, cardiolipin oxidation during apoptosis, and p66shc-coupled ROS production(reviewed in [10]), where it would be interesting to assess possible differences between theCytc isoform pair.

2.2.2. Cytochrome c oxidase—COX is the terminal enzyme of the ETC. It acceptselectrons from reduced Cytc and transfers them to molecular oxygen, which is reduced towater. This reaction is coupled to the transfer of protons from the matrix to theintermembrane space by COX. In addition to these pumped protons COX takes up chemicalprotons, which are needed when oxygen is reduced to water. Both ‘pumped’ and ‘chemicalprotons’ contribute to ΔΨm. Cow heart COX has been crystallized as a dimer [12]. Permonomer, mammalian COX is composed of two heme and two copper redox centers locatedin catalytic subunits I and II. COX contains 13 subunits, three of which are encoded by themitochondrial DNA and ten by nuclear DNA (Fig. 1).

For six of the smaller, nuclear-encoded subunits tissue-specific isoforms have beenidentified. All isoforms are encoded by separate genes and are not generated throughdifferential splicing. Three liver- and heart-type isoform pairs of subunits VIa, VIIa, andVIII exist, which have long been known (reviewed in [13]). Liver-type COX is expressed inmost organs such as liver, brain, and kidney, whereas the heart-type isozyme is expressed inheart and skeletal muscle. Heart-type COX is expressed in tissues with a high aerobiccapacity and a large number of mitochondria, whereas other tissues including the braincannot afford such a high mitochondrial load due to their specialized functions. Since thesetissues still fully depend on aerobic energy production they are equipped with a more activeenzyme, the liver-type. In addition, we recently identified three more COX subunit isoforms.These are a lung-specific isoform of COX subunit IV (COX4-2), a testis-specific isoform ofsubunit VIb, and a third isoform of subunit VIII [9, 14, 15].

Functional data including enzyme kinetics have so far been reported only for the better-studied liver, heart, and lung isozymes showing that enzymatic activity follows the orderheart-type < liver-type < lung-type [16, 17]. The reason for lung COX being the highestactivity enzyme is not clear. It was proposed that the expression of COX4-2 in lung, a highlyoxygenated tissue, might drain electrons from the ETC leaving fewer substrate electrons forthe formation of ROS [17, 18].

2.3. Allosteric regulation of cytochrome c and cytochrome c oxidaseAnother regulatory mechanism found in key metabolic enzymes is allosteric regulation,which allows fine-tuning of enzymatic activity based on the concentrations of reactionsubstrates, intermediates, and products. A common mechanism found in COX and Cytc isthat they bind adenine nucleotides. At physiological concentrations, ATP binds to Cytc,which changes the binding of Cytc to COX, leading to an inhibition of the reaction between

Hüttemann et al. Page 4

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Cytc and COX, and the elimination of the low Km phase of the otherwise biphasic kineticswith COX [19].

For COX, the largest nuclear-encoded subunit, subunit IV, has been shown to bind ATP atthe matrix side [20, 21]. ATP binding leads to an allosteric inhibition of enzyme activity athigh intra-mitochondrial ATP/ADP ratios, adjusting energy production to demand. It wasthen proposed that COX subunit IV is a pivotal regulator for COX activity in higherorganisms in contrast to bacteria that lack this subunit [22, 23].

COX subunit Va, which localizes to the matrix (Fig. 1), was shown to bind the thyroidhormone T2. Upon binding, the allosteric ATP inhibition is released, allowing a highturnover even at high ATP/ADP ratios [24].

The muscle specific isoform of subunit VIa was suggested to decrease the pumped proton-to-electron stoichiometry at high ATP/ADP ratios, in contrast to the VIa liver-type-containing isozyme, which pumps protons at a high ratio independently of the ATP/ADPratio [25]. However, liver-type COX shows a reduced proton/electron stoichiometry afteraddition of palmitate [26].

In addition to allosteric regulation COX is inhibited by nitric oxide (NO), which competeswith O2 for binding at the binuclear heme a3 - CuB reaction center [27]. Altered NOconcentrations are found in a variety of human diseases including amyotrophic lateralsclerosis, cancer [28], arthritis [29], sepsis and septic shock [30], trauma [31],neurodegenerative diseases [32], and obesity and diabetes [33]. Elevated levels of NO resultin COX inhibition and overall reduced ETC flux, leading to decreased energy levels, whichmight at least in part explain some aspects of NO-related pathologies.

2.4. Regulation of cytochrome c and cytochrome c oxidase by phosphorylationResearch analyzing other compartments of the cell including the plasma membrane, cytosol,and nucleus has long established the central role of cell signaling in the regulation andfunctioning of the living cell. However, mitochondria have only recently been recognized astargets of cell signaling pathways, and are now rousing from a century-long hibernation.Phosphorylation sites have been identified in all OxPhos complexes, but the signalingpathways involved including kinases and phosphatases remain unknown or uncertain inmost cases [34]. Phosphorylation of Cytc and COX has been most intensively studied, bothstructurally and functionally, and those phosphorylations will be discussed in detail,including a new phosphorylation site in COX subunit II that we present here. In Cytc andCOX, 4 and 14 phosphorylation sites have been mapped, respectively, mainly through massspectrometry (MS) (Fig. 1 and Table 1). MS methodology has become increasingly sensitiveand can be operated in a high throughput manner. More and more such data sets areavailable now, including a study on human skeletal muscle showing phosphorylation of CytcThr28 and Ser47 [35], and a study on human HeLa cells that indicated phosphorylation ofCOX subunit IV-1 Ser67 and Ser136 COX subunit Va Thr35 and Thr38 [36] (see Fig. 1 andTable 1). Such information is important but two points clearly need to be addressed: 1) Whatare the functional consequences of individual phosphorylations, and 2) How much of theprotein is phosphorylated at a given condition, i.e., what are the dynamics of thesephosphorylations. Below we summarize the known phosphorylations on Cytc and COX andfocus our discussion on phosphorylations and signaling pathways that affect the function ofCytc and COX.

2.4.1. Phosphorylation of cytochrome c—After Cytc had been studied for more thana century, we discovered that it is the target of cell signaling pathways. We isolated Cytcfrom cow heart tissue under conditions that preserve the physiological phosphorylation state

Hüttemann et al. Page 5

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

and found, by MS, that it was phosphorylated on Tyr97 (Fig. 1) [37]. The phosphorylatedform showed functional differences, including a shift of the heme iron - methionine 80absorption band from 695 to 687 nm. In addition, phosphorylated Cytc showed enhancedsigmoidal kinetics and inhibition in the reaction with COX by shifting the Km of COX forCytc. Half maximal turnover was observed at a Cytc substrate concentration of 5.5 μMcompared to 2.5 μM for unphosphorylated Cytc.

We later analyzed Cytc phosphorylation in cow liver tissue and, to our surprise, found that itwas phosphorylated on a different residue, Tyr48 [38]. There were no spectral changes, andenzyme kinetics with isolated COX were distinct compared to Tyr97-phosphorylated heartCytc. Tyr48-phosphorylated Cytc produced a hyperbolic response, similar tounphosphorylated Cytc. At maximal turnover, COX activity was more than 50% reducedwith Tyr48-phosphorylated Cytc. The finding of partial (but not full) respiratory inhibitionwith in vivo Tyr97- and Tyr48-phosphorylated Cytc in heart and liver supports our modelthat, under healthy conditions, ETC activity does not operate at full capacity, undoubtedly toprevent ΔΨm hyperpolarization. This prevents the production of significant amounts ofROS, which are generated at high ΔΨm levels, a concept we discuss in section 3.

In a follow-up study we generated and purified phosphomimetic Cytc by replacing CytcTyr48 with Glu, which mimics the negative charge of the phosphate group. Similar effectscompared to in vivo Tyr48-phosphorylated Cytc were observed in the reaction with COX,suggesting that it is a good model for phosphorylated Cytc. Tyr48Glu Cytc showed reducedbinding affinity to cardiolipin, a mitochondria-specific lipid with which it is normallyassociated. Strikingly, the ability of Cytc to trigger down-stream caspase activation wascompletely abolished in the Tyr48Glu mutant suggesting that Cytc phosphorylation mayfunction as a switch for the regulation of apoptosis. This finding may have broader impactfor the molecular mechanism underlying the ability of cancers to evade apoptosis. Althoughthe kinases that mediate Cytc phosphorylation are unknown, the fact alone that Cytc can betyrosine phosphorylated might suggest a link to cancer because receptor and non-receptortyrosine kinase signaling often causes cancer and it may target Cytc.

Two more phosphorylation sites were recently mapped on Cytc from human skeletal muscletissue by high throughput phosphoproteomic MS analysis. Thr28 and Ser47 were identifiedon Cytc [35], suggesting that Cytc in this tissue is targeted for phosphorylation by asignaling pathway that is distinct from heart and liver. The functions of thesephosphorylations remain unknown as does the extent of Cytc phosphorylation.

Interestingly, all four identified phosphorylation sites are located on the right side of Cytc inthe conventional view (Fig. 1). It remains to be seen how phosphorylation of Cytc affects itsother functions, including ROS scavenging under healthy conditions [39, 40] and ROSgeneration under stressed conditions through the p66shc pathway [41].

2.4.2. Phosphorylation of cytochrome c oxidase—In 1997, the first paper waspublished suggesting that COX is phosphorylated on subunit IV by an endogenous kinaseafter incubation of mitochondria with [γ-32P]ATP [42]. After that the cAMP-dependentsignaling pathway was studied by several groups in more detail and therefore deserves an in-depth discussion. The localization of PKA has been proposed by various investigators tooccur in all mitochondrial compartments, including the outer side of the outer mitochondrialmembrane, the intermembrane space, and the matrix, but it still remains a controversial issuedepending on the tissue or cell-type and methodology (reviewed in [34]).

In initial work, the Kadenbach group incubated isolated cow heart COX with PKA, cAMP,and [γ-32P]ATP in vitro [43]. After autoradiography they obtained a signal for subunit Vb

Hüttemann et al. Page 6

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

and a faint signal for subunits II or III. In a follow-up study, using the same approach theyshowed that subunit I was also phosphorylated [44]. It must be emphasized that dataobtained with an in vitro approach should be viewed with great caution, because auxiliarycomponents, such as scaffolding proteins that mediate specificity, are absent. For example,PKA signaling specificity is mediated through A-kinase anchoring proteins (AKAPs) thatprovide a platform for the docking of kinases and other proteins near membranes and theirtargets [45]. Therefore, an in vitro approach should only be considered when similar resultsare obtained in vivo. Based on our experience (e.g., comparing phosphorylation patterns ofCOX after incubation with several kinases in vitro, which did not match results obtained invivo after stimulation or inhibition of the corresponding signaling pathway) we stronglycaution the reader against pursuing an in vitro approach.

Our laboratory later revisited cAMP-dependent phosphorylation of COX in vivo in livertissue, where cAMP serves as a starvation signal, triggered by the starvation hormoneglucagon. We purified COX after treatment with the phosphodiesterase inhibitortheophylline, which generates high cAMP levels. During the purification of mitochondriaand subsequently of COX, we included activated vanadate and fluoride, two unspecifictyrosine and serine/threonine phosphatase inhibitors, respectively. The isolated enzyme wastyrosine phosphorylated on subunit I, whereas the untreated control enzyme did not show asignal. The site of phosphorylation was identified by MS as COX subunit I tyrosine 304.Tyr304 is located adjacent to the oxygen binding center on COX, an ideal site for enzymeregulation (Fig. 1). Indeed, Tyr304-phosphorylated COX was fully or strongly inhibited upto 10 μM Cytc substrate concentrations, even in the presence of allosteric activator ADP[46]. Control experiments using the physiological hormone glucagon, the strongestphysiological signal for increasing intracellular cAMP levels in liver, or the adenylyl cyclaseactivator forskolin, both of which increase intracellular cAMP, showed a similar inhibitoryeffect. Notably, Tyr304 phosphorylation cannot be directly mediated by Ser/Thr-specificPKA, and it was proposed that a downstream tyrosine kinase phosphorylates COX [46].

It should be noted that there are distinct tissue-specific responses to cAMP and probably forother signaling pathways as well. For example, in neuronal tissue theophylline treatmentleads to COX activation [47]. In heart, theophylline showed no effect, and other heart-specific cAMP inducers such as 3-isobutyl-1-methylxanthine (IBMX) did not showsignificant effects (unpublished).

One study proposed that cAMP/PKA signaling leads to phosphorylation of COX in heartafter ischemia. There, COX subunits I, IV, and Vb were found to be phosphorylated afterischemia in rabbit hearts and these phosphorylations were abolished in the presence ofalleged PKA inhibitor H89 [48]. The authors later identified the phosphorylated amino acidsby MS to be Ser115 and Ser116 of subunit I, Thr52 of subunit IV, and Ser40 of subunit Vb[49]. These sites (Fig. 1 and Table 1) are not PKA consensus sites, and it thus seemsunlikely that PKA signaling is directly involved. Kinase inhibitors often target a group ofkinases, and H89 inhibits at least five other kinases [50], one or more of which may beinvolved in the phosphorylations observed after ischemia.

The Manfredi group has recently added another interesting facet to cAMP-dependentsignaling by identifying a matrix-localized carbon dioxide/bicarbonate-regulated adenylylcyclase [51]. Via sensing of CO2 generated by the citric acid cycle, this enzyme linksnutrient availability to OxPhos activity. COX is a target of this signaling, and the authorsshowed that subunits I and IV are targets of a matrix-localized PKA. The phosphorylationsites need to be mapped to provide additional evidence of direct PKA involvement. Theeffect of matrix adenylyl cyclase and PKA-dependent signaling is activation of COX [51],and such activation was shown to improve mitochondrial function in COX-deficient cells

Hüttemann et al. Page 7

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

[52]. Additional support for matrix localized PKA signaling to COX was reported earlier.The RIα regulatory subunit of PKA can bind to matrix-localized subunit Vb of COX [53]. Itis noteworthy that the observed activation of COX through matrix-localized adenylylcyclase and PKA results in the opposite effect compared to cAMP-dependent in vivo Tyr304phosphorylation. This phosphorylation occurs on the intermembrane space side, and in vitrophosphorylation of COX with commercially available PKA, which can presumably occur onthe matrix and intermembrane space sides, also leads to COX inhibition [43]. Mapping ofthe phosphorylation sites in these studies would shed light into this controversial point.

Additional phosphorylation sites have been mapped on COX, but the signaling pathwaysinvolved and their functional consequences are unknown. Those are Tyr11 of subunit IV-1in isolated cow liver [37], Ser67 and Ser136 of subunit IV-1, Thr35 and Thr38 of subunit Vain human HeLa cells by high-throughput MS [36], Ser34 of subunit IV-1, and Ser4 andThr35 of subunit Va in cow heart [54], and Thr11 of subunit VIa heart isoform, which wasidentified in the cow heart crystal structure [55]. The presence of phosphorylation sites intissue-specific isoforms suggests that one of their functions is the provision of a platform forcell signaling that can mediate tissue-specific regulation. This is accompanied by tissue-specific expression of signaling molecules, such as mitochondrial tyrosine phosphatase PTP1B, which is only found in brain but is absent in muscle, heart, and liver tissue [56].

We here present another new phosphorylation, which we identified in cow heart COX (Fig.2). This site, Ser126 on catalytic subunit II, was unambiguously mapped by MS in threeindependent COX isolations, in which the enzyme was isolated from heart tissue that wasimmediately frozen on dry-ice, kept on ice for 20 min, and kept ischemic for 45 min at 37°C prior to enzyme isolation. The phosphorylation site is located on the intermembrane spaceside (Fig. 1). It is adjacent to but not part of the Cytc-COX docking site based on computermodelling [57]. In that model the nearest amino acid of Cytc, Lys27, is located within 11 Åof Ser126. Ser126 is located at the interface of subunits II and IV, with His109 of subunit IVbeing the closest amino acid (within 5 Å). Therefore, this phosphorylation might be involvedin enabling allosteric regulation of COX via ATP and ADP that bind to subunit IV, which islost when COX is fully dephosphorylated [13].

One concern regarding the above studies is that there is little overlap of phosphorylationsites identified by different labs. Likely explanations are that different purification protocolswere used and that treatments differed, including buffer compositions and the use ofphosphatase inhibitors. In addition, MS methodology is problematic with hydrophobicpeptides. For COX as a membrane protein this makes it difficult to identify hydrophobicphospho-epitopes. For example, phospho-Tyr304 identification of hydrophobic subunit Irequired significant optimization of the sample preparation [46].

There is another series of studies suggesting that COX is a target of signaling pathways, butin these cases the phosphorylation sites have not been identified. Non-receptor tyrosinekinase Src has been shown to localize to the mitochondrial intermembrane space [58], andphosphorylates COX subunit II in osteoblasts [59]. This phosphorylation leads to COXactivation and appears to be important for the bone-resorbing activity of these cells, whichmight be altered in rheumatoid arthritis [60]. Another COX regulator is protein kinase Cε(PKCε). In rat neonatal cardiac myocytes treatment with diacylglycerol or 4β-PMA, whichare PKC activators, resulted in the phosphorylation of an 18 kDa protein in vitro [61, 62]. Itwas shown that PKCε co-immunoprecipitated with COX and that the radiolabeled 18 kDaband contained subunit IV. Activation of PKCε resulted in about two- to fourfold increasedCOX activity. Src kinase is also phosphorylated and activated by PKCε [63, 64]; however,no tyrosine phosphorylation of COX subunit II was observed after PKCε stimulation [62].This might be explained with tissue specific differences in cell signaling including presence

Hüttemann et al. Page 8

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

or absence of additional required components such as scaffolding proteins. PKCε isimplicated in ischemia/reperfusion injury, a topic that we will revisit in the next section.

Mitochondrial metabolism shows major alterations in cancer. Most solid tumors show areduction in mitochondrial mass of 25–60% compared to healthy tissue [65]. Cancer cellsrely less on respiration and more on glycolysis, which is known as the Warburg effect [3, 4].The molecular changes that mediate this metabolic switching are now emerging. Often,cancers show increased receptor tyrosine kinase signaling, which promotes proliferation andis anti-apoptotic. One such receptor is epidermal growth factor receptor (EGFR), which isimplicated in numerous cancers including colon, lung, breast, cervical, and head and neckcancers. Recent research has changed the picture of how signaling from the plasmamembrane is transmitted to the mitochondria, which now appears to be much more dynamicthen thought previously. Notably, EGFR, upon stimulation with EGF, translocates to themitochondria where it directly interacts with COX subunit II, as shown by phage displayscreening and co-immunoprecipitation experiments [66]. EGFR binding to COX isdependent on the activation state of the receptor and requires Tyr845 phosphorylation,which is mediated by Src kinase. EGFR in combination with Src leads to an increase inCOX subunit II phosphorylation in vitro as was shown after incubation with [γ-32P]ATPfollowed by autoradiography [67]. However, the phosphorylation sites still have to beidentified. Importantly, the effect of EGFR/Src signaling on COX activity is distinct fromSrc signaling alone discussed above because it leads to 60% inhibition after cells weretreated with EGF for 20 minutes. This finding may explain the mechanism of the Warburgeffect in cancers with upregulated EGFR signaling, and it remains to be seen if a similarsequence of events wherein the receptor is translocated to the mitochondria followed byphosphorylation and partial inhibition of COX also applies to other receptor tyrosinekinases.

The presence of kinases targeting COX implies the presence of phosphatases to reverse suchsignaling. Protein tyrosine phosphatase Shp-2, which is part of the Ras pathway, was thefirst identified tyrosine phosphatase that localizes to the mitochondrial intermembrane spaceand the outer mitochondrial membrane in addition to the cytoplasm [68], and we speculatedthat it might target COX or Cytc, the only two components of the ETC with mapped tyrosinephosphorylation sites (Table 1). Mutations in the Shp-2 encoding gene PTPN11 account forabout 50% of cases with Noonan syndrome [69], an autosomal dominant disordercharacterized by dysmorphic facial features, congenital heart disease, webbed neck, shortstature, chest deformity, and mental retardation. In these Noonan patients Shp-2 isconstitutively active [70]. We analyzed patient and mouse cell lines and observedsignificantly increased COX activity [71]. Among OxPhos complexes, only COX and Cytcprotein levels were down-regulated in the mutant cells, suggesting a compensatorymechanism to counterbalance increased COX activity. Additional work is needed toestablish COX and Cytc as possible direct or indirect targets of Shp-2. Clearly,dephosphorylation of Tyr48 and Tyr97 on Cytc and Tyr304 on COX subunit I would resultin increased respiration rates, and all three epitopes are located in the intermembrane space,as is Shp-2. Alteration in COX/Cytc activity and their total amount could affectmitochondrial energy and ROS production. Indeed, in the Shp-2 mutant cells ATP levelswere reduced by 30% and ROS levels were increased. Both reduced ATP and increasedROS may interfere with organ development as is observed in the patients (for a moredetailed discussion see [13]).

Another important mechanism related to cell signaling that acts on COX may be through thedirect and indirect action of calcium. Calcium is the proposed strongest signal formitochondrial activation [72]. Mammalian COX contains a calcium-sodium exchange site insubunit I [73, 74], which may affect COX activity. In addition, calcium was shown to cause

Hüttemann et al. Page 9

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

dephosphorylation of most mitochondrial proteins [75], likely through calcium-activatedphosphatases. As described above, dephosphorylation of COX (at Tyr304, subunit I) andCytc (Tyr 48 and Tyr97) would result in increased ETC activity. Calcium also plays a keyrole during conditions of cellular stress, where it leads to hyperactive ETC complexes andincreased ΔΨm levels, leading to excessive ROS production as discussed in section 3.

3. The connection between ETC activity, ΔΨm, energy, and ROS in humandisease: acute inflammation and ischemia/reperfusion injury as twodistinct mitochondrial pathologies

The role of OxPhos in human disease can be distilled into two buzzwords, energy and ROS.Many pathologies present with decreased ATP or increased ROS levels, and sometimes acombination of both. We have previously proposed a model that links ETC activity, ΔΨm,and the production of ATP and ROS [13]. Based on our model there are three states,‘hypoactive,’ ‘healthy,’ and ‘hyperactive’ (Fig. 3), which may explain manypathophysiological conditions where mitochondria are involved. In the basic, ‘healthy’ statethe OxPhos complexes are phosphorylated, which reduces their activity. Importantly, thisprevents the generation of ΔΨm levels larger than 140 mV; i.e., physiologically lower ΔΨmlevels are sufficient to inhibit further proton pumping. This regulation ensures that ΔΨm ismaintained around 120 mV under healthy conditions, which is sufficient for efficient ATPproduction [76] but prevents the formation of ROS, which are generated exponentially atΔΨm levels exceeding 150 mV [77].

A further fine-tuning of the regulation of mitochondrial respiration under healthy conditionsis mediated via the allosteric regulation of COX and Cytc by ATP and ADP discussedabove. Respiration rates can be high if ADP is not limiting and no allosteric ATP-inhibitionof COX occurs [21, 23]. Under these conditions the high rate of proton uptake by ATPcontributes to the maintenance of lower ΔΨm levels. When ADP becomes limiting theallosteric ATP-inhibition inhibits respiration and further proton pumping. This occursalready at healthy ΔΨm levels, because maximal rates of ATP synthesis by ATP synthaseoccur at ΔΨm = 100 – 120mV [76]. It should be noted, however, that phosphorylation ofOxPhos components appears to be the overarching regulatory mechanism preventingunhealthy high ΔΨm values even in the presence of ADP. This is so because ΔΨm valuesduring state 3 reported for isolated mitochondria, with presumably dephosphorylatedOxPhos complexes, are often higher (>140 mV) than those reported for intact cells. Forexample, a ΔΨm value of 172 mV was reported for rat liver mitochondria during state 3respiration [78]. During state 4, mitochondria isolated according to traditional protocols withdephosphorylated OxPhos complexes can generate ΔΨm levels of 200 mV and above [13].Isolated mitochondria are morphologically and functionally distinct from mitochondria inintact cells and show unphysiologically high COX activities [79].

To exemplify the two distinct pathological states of our model, ‘hypoactive’ and‘hyperactive,’ we will discuss acute inflammation and ischemia/reperfusion injury. Thehypoactive state seen in acute inflammation results in decreased ΔΨm and energy levels. Incontrast, the hyperactive state during reperfusion following ischemia results in ΔΨmhyperpolarization and ROS production. Both states can result in cell death through thesedistinct pathways.

3.1. Role of cytochrome c oxidase and cytochrome c in acute inflammation: the hypoactivestate and the energy crisis

Sepsis is an acute inflammatory condition that affects the entire body and represents apotentially life-threatening systemic inflammatory state. Sepsis is sometimes referred to as

Hüttemann et al. Page 10

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

blood poisoning in the non-medical literature. In the United States, sepsis develops in750,000 people annually with more than 210,000 deaths, which makes it the leading causeof mortality in patients in intensive care units [80]. Sepsis was formerly attributed to apathogenic infection of the blood, but the definition has changed in modern times to reflectthe body’s inflammatory response to severe insult or injury, including the development ofmultiple organ dysfunction syndrome (MODS) [81]. Therefore, the combination of aninfection with an overly active immune response can lead to organ dysfunction and eventualdeath.

Except for mutations in the pathogen recognition toll-like receptor 4 (TLR4) [82], the onlyother known genetic predictor for survival after sepsis is a particular mitochondrial DNAcomposition, the haplogroup H. Mitochondrial haplogroups are grouped based on arelatively small number of mitochondrial DNA polymorphisms that are preserved in certainpopulations, for example from different continents. Those subtle changes have been linkedto certain phenotypes and pathologies, such as longevity or male fertility [83]. Patientsbelonging to haplogroup H, which is common in Europeans, have a more than twofoldincreased chance of survival after sepsis compared to patients in other haplogroups [84].This finding suggests that mitochondria play a key role in acute inflammation.

Given the role of COX and Cytc in providing energy for all cellular functions and the role ofCytc in the regulation of cellular survival and apoptosis, it is no surprise that both enzymeshave been implicated in the pathology of sepsis. The terminal step of the ETC is a logicalfunctional target: it is characterized by poor tissue oxygen extraction and utilization as seenin patients, suggesting that the ETC is a target of inflammatory signaling. In fact, it washypothesized in the past that dysfunction of mitochondrial electron transport may beresponsible for some of the symptoms of sepsis because direct systemic delivery of oxygenduring the course of sepsis was shown to be ineffective [85]. This indicated that oxygenutilization rather than uptake is impaired, a model referred to as cytopathic hypoxia [86].

Lipopolysaccharide (LPS, or endotoxin) is often used in animal experiments to cause aseptic state and leads to the production of pro-inflammatory cytokines, including tumornecrosis factor α (TNFα) through the activation of TLR4. TNFα is a key cytokine that isinduced as part of the inflammatory cascade that leads to the septic state. TNFα affectsmetabolism in that it induces the production of lactate in vitro and in vivo [87, 88],indicating a switching from aerobic to glycolytic energy metabolism. In LPS-treated rats,ETC complexes I, II, and IV were down-regulated both at the transcript and protein levelwithin 24 h after treatment [89]. It was further demonstrated that early in sepsis oxidation ofCytc by COX is competitively and reversibly inhibited; in later stages (at 48 hours), itbecomes noncompetitive and irreversible [90]. Rats subjected to cecal ligation and puncture,another animal model for sepsis, showed unchanged muscle tissue oxygen levels betweensham control and septic animals but the septic animals had reduced ATP levels [91].Another study performed by the Kozlov group showed that induction of endotoxic shockthrough LPS treatment resulted in an early increase in blood levels of TNFα and 70%decreased ATP levels after 8 h, a critical time at which about 30% of the animals are deadand most of the remaining animals surviving [92].

All the above reports indicate that metabolic changes occur at the level of OxPhos duringsepsis but the molecular mechanism remained unclear. To gain more insight into thisquestion we investigated the effect of TNFα on cow and mouse liver tissue and mousehepatocytes in culture. TNFα treatment of bovine and murine liver homogenates produced a60% reduction of COX activity within 5 min after treatment [93]. Importantly, isolation ofcow COX revealed phosphorylation of tyrosine 304 of subunit I after TNFα, which wasidentified with a phospho-epitope-specific antibody. This is the same site that is targeted for

Hüttemann et al. Page 11

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

phosphorylation by the cAMP-dependent pathway in liver discussed above. TNFα treatmentdecreased the mitochondrial membrane potential and resulted in 35% and 64% reduction ofcellular ATP levels in mouse hepatocytes and H2.35 cells [93]. It will be interesting to see ifCytc is also targeted by TNFα signaling, which could augment the effect of COXphosphorylation.

A question that emerges is: why did such a response evolve, which can lead to death inhumans? A likely explanation is that, in a day-to-day context, the organism constantly has todefeat pathogens at a more local setting, e.g., after a small skin cut. Some pathogens seizethe host infrastructure and energy production system. For example, chlamydiae expresses anumber of nucleotide transporters that facilitate the uptake of molecules such as ATP [94].Therefore, shutting down OxPhos locally at the site of infection slows down pathogenicgrowth because essential metabolites are no longer provided by the host. In the rare situationof a systemic inflammation this response can get out of control, leading to MODS and deathdue to energy failure of entire organs.

In summary, our model proposes that acute inflammatory signaling leads to phosphorylationof COX and strong enzyme inhibition, reduced membrane potentials, and consequentlydecreased energy levels (Fig. 3). If ATP levels drop below a threshold, such as 30% in livertissue [92], organ function cannot be maintained, leading to organ failure.

3.2. Role of cytochrome c oxidase and cytochrome c in brain ischemia/reperfusion injury:the hyperactive state and excessive reactive oxygen species

Cerebral ischemia/reperfusion injury is caused by interruption of blood flow to the brain.This can occur in multiple pathologies, the most common being ischemic stroke. Ischemicstroke occurs when blood flow to a region of the brain is prevented by occlusion of acerebral blood vessel. Restoration of blood flow to ischemic tissue is necessary to limittissue damage; however restoration of oxygen and nutrients to stressed tissue leads toadditional damage. This damage is called reperfusion injury and contributes greatly to theoverall severity of tissue damage. The brain also undergoes ischemia/reperfusion injury inthe setting of cardiac arrest followed by resuscitation. Cardiac failure results in completecerebral ischemia and resuscitation of the arrested patient initiates reperfusion of the brain.Both ischemic stroke and cardiac arrest/resuscitation result in extensive morbidity andmortality throughout the population. Stroke is the 3rd leading cause of death and disability inthe US and only 10% of resuscitated patients leave the hospital neurologically intact and canresume their former lifestyles [95]. These statistics demonstrate the dramatic need forgreater understanding of ischemia/reperfusion injury in the brain and for targeted therapiesto treat these disease processes.

Mitochondria play multiple roles in the setting of cerebral ischemia/reperfusion injury bothas a cause of neuronal damage and a site of intracellular injury. We propose that there arethree distinct phases of altered COX activation during the progression of ischemia/reperfusion injury: 1) the ischemic starvation phase, 2) the reperfusion-inducedhyperactivation phase, and 3) the mitochondrial dysfunction phase. The role of COX andCytc phosphorylation in these distinct activation profiles remains largely unknown.However, many studies have identified critical alterations in COX activity during theprogression of reperfusion injury, suggesting an important role for cell signaling in thesechanges.

During the ischemic starvation phase, tissue becomes oxygen deprived, thereby depletingCOX of its terminal substrate for respiration. In the ischemic brain, this is rapidly followedby ATP depletion, elevated ADP, and, importantly, an increase in mitochondrial calcium[96]. There are multiple cell signaling pathways activated by ischemia that signal to the

Hüttemann et al. Page 12

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

mitochondria to increase ATP production in order to compensate for the rapid loss ofenergy. Many phosphorylation/dephosphorylation events have been shown to occur on COXthat could, in theory, increase ATP production in ischemic tissue if oxygen were present. Ofparticular interest is the role of mitochondrial calcium during ischemia and earlyreperfusion. Mitochondrial calcium concentrations increase during ischemia whenintracellular pumps fail, and mitochondria actively sequester calcium during earlyrestoration of blood flow. This results in dramatically increased mitochondrial calciumconcentrations [97]. As discussed above, calcium is the most important signal formitochondrial activation [72] and induces dephosphorylation of most mitochondrial proteins[75]. Calcium has been shown in vitro to lead to hyperactivation of COX through a higherbasal activity and loss of the allosteric inhibition by ATP [43]. This is likely caused bychanges in the phosphorylation state of COX, which is further supported by the finding thatdephosphorylation of COX makes the enzyme insensitive to ATP and ADP [13]. Thus, theloss of allosteric control does not appear to be a direct effect of calcium on COX, but iscaused indirectly by dephosphorylation of COX, presumably by a calcium-dependentphosphatase [43]. Increased basal COX activity in combination with a loss of allostericinhibition by ATP could cause profound problems when reperfusion of the tissue is initiated,as this would lead to hyperpolarization of mitochondrial ΔΨm (Fig. 3). It was shown thattreatment of brain mitochondria with calcium increased state 4 respiration by 141% [98],which may be explained by COX dephosphorylation. These events would render COX“primed” for hyperactivity resulting in deleterious effects when oxygen and nutrients arerestored upon reperfusion.

Based on our model (Fig. 3), upon restoration of blood flow, oxygen and nutrients reachstressed mitochondria and a COX enzyme that has been post-translationally modified toincrease its activity. These alterations aid in restoration of ΔΨm, and reestablishment ofcellular energy levels [99]. In fact, mitochondria rapidly become re-energized and ΔΨm isquickly restored within 5 min [99]. However, in this hyperactive state, the ETC generatespathologically high ΔΨm levels, and mitochondrial hyperpolarization leads to an exponentialincrease in ROS generation at membrane potentials exceeding 140 mV [77]. Additionally,measurements of ROS following ischemia indicate that the majority of mitochondrial ROSare created during this early reperfusion interval [100], and the ETC is a primary source ofROS during reperfusion [101]. Indeed, Liu and coworkers demonstrated that reperfusion ofischemic brain results in a rapid restoration of ΔΨm, followed by a transienthyperpolarization of ΔΨm, which lasts for about 8 min in the mouse brain [102]. This effectis lost when animals undergo ischemic preconditioning, a method shown to produce a robusteffect in reducing ROS generation by the mitochondria, which makes it profoundlyneuroprotective. These findings position COX as a potential regulatory site that canindirectly control ROS generation by regulating overall ETC flux, thereby controlling ΔΨm.This also suggests that regulation of ETC activity and/or ΔΨm hyperpolarization are targetsfor therapeutic intervention. Neuroprotective therapies designed to prevent ΔΨmhyperpolarization have been effective at both preventing ROS generation by themitochondria and preventing neuronal cell death [103]. In agreement with this hypothesis,mild mitochondrial uncoupling with FCCP reduces ΔΨm hyperpolarization, prevents ROSgeneration, and minimizes cell death in models of cardiac ischemia [104], traumatic braininjury [105], and peroxide-induced neuronal death [106]. Additionally, the converse is alsotrue; promoting hyperpolarization during reperfusion by knocking out mitochondrialuncoupling protein 2 (UCP2) results in increased ROS generation and neuronal damage[103].

The final phase in the progression of brain ischemia/reperfusion injury is known asmitochondrial dysfunction or secondary energy failure. In this phase, COX activity becomesdramatically reduced throughout the progression of reperfusion injury, eventually

Hüttemann et al. Page 13

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

culminating with energy failure and cell death [107–109]. The role of COX phosphorylationin reduced COX activity during this late stage is currently unknown. Multiple studies havesuggested that diminished mitochondrial respiration is likely caused by oxidative damage tothe mitochondria [110, 111]. Interestingly, while COX appears to be a major site ofdepressed respiratory capacity [107], COX is particularly resistant to direct oxidativedamage [111]. Oxidative damage to the mitochondria can occur on proteins involved inrespiration as well as lipids critical for respiratory protein function. Specifically, the innermitochondrial membrane lipid cardiolipin, which is required for proper membrane insertionand function of COX and other OxPhos components, can become oxidized [112]. Oxidativedamage to the mitochondria results in eventual failure of OxPhos. In addition, ROSproduction during the reperfusion phase and subsequent mitochondrial dysfunction activatethe intrinsic apoptotic pathway, thereby committing the cell to death. In summary, thismodel puts forth the hypothesis that ischemia-induced stress leads to changes in thephosphorylation state of COX. This renders COX hyperactive and leads to mitochondrialROS generation during the initial stages of reperfusion. ROS can trigger apoptosis, and theycause damage that initiates ETC and COX hypoactivity and mitochondrial dysfunctionduring later reperfusion. It is thus possible that therapeutic intervention at the level of COXmay be a potent neuroprotective therapy in the context of brain reperfusion injury. If thealterations, i.e., phosphorylations, to COX that increase its activity can be targeted toprevent mitochondrial hyperpolarization and subsequent ROS generation, it may be possibleto prevent the downward spiral of mitochondrial dysfunction and cell death.

The role of Cytc in ischemia/reperfusion injury is twofold. The loss of Cytc function is acontributing factor to mitochondrial dysfunction or secondary energy failure, and Cytcrelease from mitochondria is an initiating event in mitochondrial type II apoptosis and celldeath. Recent identification of phosphorylation sites on Cytc discussed above that regulateits activity in respiration and apoptosis raises the possibility that Cytc is actively targeted bystress signaling during ischemia/reperfusion. Because phosphorylation of Cytc partiallyinhibits respiration, if stress-induced signaling also targets Cytc for dephosphorylation, thiswould lead to increased ETC flux and further contribute to the hyperpolarization of ΔΨm. Insupport of this hypothesis, we found that Cytc isolated from ischemic brain showed no signsof phosphorylation (unpublished data). In contrast to phosphorylated Cytc, thisdephosphoryated Cytc would have the full capability to bind to Apaf1 and triggerdownstream caspase activation [10].

In the progression of brain reperfusion injury, mitochondrial respiration begins to diminish[108, 113], and mitochondrial dysfunction eventually culminates in cell death. Diminishedrespiratory capacity precedes the release of Cytc into the cytosol. Therefore, a lack of Cytcavailability cannot account for impaired respiration during early reperfusion. One possiblefactor in mitochondrial dysfunction is peroxidation of cardiolipin by Cytc, an important stepin the execution of the apoptotic cascade [114]. Cardiolipin tethers Cytc to the innermitochondrial membrane, whereas peroxidation of cardiolipin allows disassociation of Cytcand liberates it into a free-floating pool [114]. Extensive peroxidation of cardiolipin hasbeen demonstrated during brain reperfusion, which is associated with initiation of apoptosis.Interestingly, cardiolipin peroxidation might be regulated by phosphorylation of tyrosinesites on Cytc, as suggested by studies using phosphomimetic mutant Cytc [115].

During the final stages of reperfusion injury, Cytc participates in cell demise by initiatingapoptosis. Multiple studies have demonstrated the neuroprotective and antiapoptotic effectof therapies designed to activate cell survival signaling and prevent apoptotic release of Cytc[116, 117]. However, the idea that stimulation of cell signaling could lead tophosphorylation of Cytc, thereby suppressing apoptosis, has yet to be investigated. In

Hüttemann et al. Page 14

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

summary, phosphorylation events that regulate COX and Cytc may be novel targets fortherapeutic intervention to limit the damage caused by ischemia/reperfusion.

4. ConclusionCOX and Cytc show all three main regulatory features found in key metabolic enzymes:isoform expression, allosteric control, and phosphorylation. This points to the importance ofthe regulation of the terminal step of the ETC, and supports its suggested rate-limiting rolefor overall ETC flux in intact cells. This has direct implications for energy and ROSproduction, which are dysregulated in numerous human diseases.

In higher organisms cell signaling cascades function as communication networks that allowadjustment of cell, organ, and organismal function to varying internal and environmentalconditions, ranging from complete rest (e.g., hibernation) to strenuous exercise. To study theeffect of signaling cascades on mitochondrial function important precautions have to betaken to preserve posttranslational modifications. To maintain the physiologicalphosphorylation state, phosphatase inhibitors have to be included at all steps during theisolation of mitochondrial proteins, and detailed protocols are available for both Cytc andCOX [38, 118].

Another crucial aspect that deserves special attention is that signaling must not begeneralized, and instead has to be carefully evaluated in a tissue-specific context. We havetouched on this topic discussing cAMP-dependent signaling that, depending of the tissue,can lead to distinct and sometimes opposing effects. The expression of tissue-specificisoforms further underlines this notion since it can provide tissue-specific targets forphosphorylation as is the case for COX subunit VIa and likely others that have yet to beidentified.

Phosphorylation sites have been mapped in all OxPhos complexes and it is thereforepossible that strong regulatory effects on enzyme activity, e.g., as seen for COX subunit ITyr304 phosphorylation, will be reported for complexes I, II, III, and V in future work.Identification of kinases and phosphatases that directly act on OxPhos would also makepossible the utilization of genetic methods to further pinpoint the role of posttranslationalregulation. This research direction holds tremendous translational potential because it wouldallow targeted therapeutic intervention in conditions such as sepsis, ischemia/reperfusioninjury, cancer, and many others.

AcknowledgmentsThis work was supported by grant GM089900 from the National Institutes of Health, the Center for MolecularMedicine and Genetics, and the Cardiovascular Research Institute, Wayne State University School of Medicine,Detroit, and a grant from the German Research Foundation (DFG Ka 192/40-1).

References1. Rich P. Chemiosmotic coupling: The cost of living. Nature. 2003; 421:583. [PubMed: 12571574]2. Morriss GM, New DA. Effect of oxygen concentration on morphogenesis of cranial neural folds and

neural crest in cultured rat embryos. J Embryol Exp Morphol. 1979; 54:17–35. [PubMed: 528863]3. Warburg O, Posener K, Negelein E. Über den Stoffwechsel der Carcinomzelle. Biochem Z. 1924;

152:309–344.4. Warburg O. On the origin of cancer cells. Science. 1956; 123:309–314. [PubMed: 13298683]5. Lardy HA, Wellman H. Oxidative phosphorylations; role of inorganic phosphate and acceptor

systems in control of metabolic rates. J Biol Chem. 1952; 195:215–224. [PubMed: 14938372]

Hüttemann et al. Page 15

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

6. Chance B, Williams GR. Respiratory enzymes in oxidative phosphorylation. I. Kinetics of oxygenutilization. J Biol Chem. 1955; 217:383–393. [PubMed: 13271402]

7. Goldberg E, Sberna D, Wheat TE, Urbanski GJ, Margoliash E. Cytochrome c: immunofluorescentlocalization of the testis-specific form. Science. 1977; 196:1010–1012. [PubMed: 193188]

8. Zhang Z, Gerstein M. The human genome has 49 cytochrome c pseudogenes, including a relic of aprimordial gene that still functions in mouse. Gene. 2003; 312:61–72. [PubMed: 12909341]

9. Hüttemann M, Jaradat S, Grossman LI. Cytochrome c oxidase of mammals contains a testes-specificisoform of subunit VIb – the counterpart to testes-specific cytochrome c? Mol Reprod Dev. 2003;66:8–16. [PubMed: 12874793]

10. Hüttemann M, Pecina P, Rainbolt M, Sanderson TH, Kagan VE, Samavati L, Doan JW, Lee I. Themultiple functions of cytochrome c and their regulation in life and death decisions of themammalian cell: From respiration to apoptosis. Mitochondrion. 2011; 11:369–381. [PubMed:21296189]

11. Liu Z, Lin H, Ye S, Liu QY, Meng Z, Zhang CM, Xia Y, Margoliash E, Rao Z, Liu XJ.Remarkably high activities of testicular cytochrome c in destroying reactive oxygen species and intriggering apoptosis. Proc Natl Acad Sci U S A. 2006; 103:8965–8970. [PubMed: 16757556]

12. Tsukihara T, Aoyama H, Yamashita E, Tomizaki T, Yamaguchi H, Shinzawa-Itoh K, NakashimaR, Yaono R, Yoshikawa S. The whole structure of the 13-subunit oxidized cytochrome c oxidaseat 2.8 Å. Science. 1996; 272:1136–1144. [PubMed: 8638158]

13. Hüttemann M, Lee I, Pecinova A, Pecina P, Przyklenk K, Doan JW. Regulation of oxidativephosphorylation, the mitochondrial membrane potential, and their role in human disease. JBioenerg Biomembr. 2008; 40:445–456. [PubMed: 18843528]

14. Hüttemann M, Kadenbach B, Grossman LI. Mammalian subunit IV isoforms of cytochrome coxidase. Gene. 2001; 267:111–123. [PubMed: 11311561]

15. Hüttemann M, Schmidt TR, Grossman LI. A third isoform of cytochrome c oxidase subunit VIII ispresent in mammals. Gene. 2003; 312:95–102. [PubMed: 12909344]

16. Vijayasarathy C, Biunno I, Lenka N, Yang M, Basu A, Hall IP, Avadhani NG. Variations in thesubunit content and catalytic activity of the cytochrome c oxidase complex from different tissuesand different cardiac compartments. Biochim Biophys Acta. 1998; 1371:71–82. [PubMed:9565657]

17. Hüttemann M, Lee I, Liu J, Grossman LI. Transcription of mammalian cytochrome c oxidasesubunit IV-2 is controlled by a novel conserved oxygen responsive element. FEBS J. 2007;274:5737–5748. [PubMed: 17937768]

18. Fukuda R, Zhang H, Kim JW, Shimoda L, Dang CV, Semenza GL. HIF-1 regulates cytochromeoxidase subunits to optimize efficiency of respiration in hypoxic cells. Cell. 2007; 129:111–122.[PubMed: 17418790]

19. Ferguson-Miller S, Brautigan DL, Margoliash E. Correlation of the kinetics of electron transferactivity of various eukaryotic cytochromes c with binding to mitochondrial cytochrome c oxidase.J Biol Chem. 1976; 251:1104–1115. [PubMed: 2600]

20. Napiwotzki J, Shinzawa-Itoh K, Yoshikawa S, Kadenbach B. ATP and ADP bind to cytochrome coxidase and regulate its activity. Biol Chem. 1997; 378:1013–1021. [PubMed: 9348111]

21. Arnold S, Kadenbach B. The intramitochondrial ATP/ADP-ratio controls cytochrome c oxidaseactivity allosterically. FEBS Lett. 1999; 443:105–108. [PubMed: 9989584]

22. Ludwig B, Bender E, Arnold S, Hüttemann M, Lee I, Kadenbach B. Cytochrome c oxidase and theregulation of oxidative phosphorylation. Chembiochem. 2001; 2:392–403. [PubMed: 11828469]

23. Kadenbach B, Ramzan R, Wen L, Vogt S. New extension of the Mitchell Theory for oxidativephosphorylation in mitochondria of living organisms. Biochim Biophys Acta. 2010; 1800:205–212. [PubMed: 19409964]

24. Arnold S, Goglia F, Kadenbach B. 3,5-Diiodothyronine binds to subunit Va of cytochrome-coxidase and abolishes the allosteric inhibition of respiration by ATP. Eur J Biochem. 1998;252:325–330. [PubMed: 9523704]

25. Frank V, Kadenbach B. Regulation of the H+/e− stoichiometry of cytochrome c oxidase frombovine heart by intramitochondrial ATP/ADP ratios. FEBS Lett. 1996; 382:121–124. [PubMed:8612732]

Hüttemann et al. Page 16

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

26. Lee I, Kadenbach B. Palmitate decreases proton pumping of liver-type cytochrome c oxidase. Eur JBiochem. 2001; 268:6329–6334. [PubMed: 11737187]

27. Brookes PS. Mitochondrial nitric oxide synthase. Mitochondrion. 2004; 3:187–204. [PubMed:16120354]

28. Carreras MC, Poderoso JJ. Mitochondrial nitric oxide in the signaling of cell integrated responses.Am J Physiol Cell Physiol. 2007; 292:C1569–1580. [PubMed: 17496232]

29. Cuzzocrea S. Role of nitric oxide and reactive oxygen species in arthritis. Curr Pharm Des. 2006;12:3551–3570. [PubMed: 17017948]

30. Assreuy J. Nitric oxide and cardiovascular dysfunction in sepsis. Endocr Metab Immune DisordDrug Targets. 2006; 6:165–173. [PubMed: 16787291]

31. Steiner J, Rafols D, Park HK, Katar MS, Rafols JA, Petrov T. Attenuation of iNOS mRNAexacerbates hypoperfusion and upregulates endothelin-1 expression in hippocampus and cortexafter brain trauma. Nitric Oxide. 2004; 10:162–169. [PubMed: 15158696]

32. Moncada S, Bolanos JP. Nitric oxide, cell bioenergetics and neurodegeneration. J Neurochem.2006; 97:1676–1689. [PubMed: 16805776]

33. Stepp DW. Impact of obesity and insulin resistance on vasomotor tone: nitric oxide and beyond.Clin Exp Pharmacol Physiol. 2006; 33:407–414. [PubMed: 16700872]

34. Hüttemann M, Lee I, Samavati L, Yu H, Doan JW. Regulation of mitochondrial oxidativephosphorylation through cell signaling. Biochim Biophys Acta. 2007; 1773:1701–1720. [PubMed:18240421]

35. Zhao, Leon IR, Bak S, Mogensen M, Wrzesinski K, Hojlund K, Jensen ON. Phosphoproteomeanalysis of functional mitochondria isolated from resting human muscle reveals extensivephosphorylation of inner membrane protein complexes and enzymes. Mol Cell Proteomics.2010:M110.000299.

36. Olsen JV, Vermeulen M, Santamaria A, Kumar C, Miller ML, Jensen LJ, Gnad F, Cox J, JensenTS, Nigg EA, Brunak S, Mann M. Quantitative phosphoproteomics reveals widespread fullphosphorylation site occupancy during mitosis. Sci Signal. 2010; 3:ra3. [PubMed: 20068231]

37. Lee I, Salomon AR, Yu K, Doan JW, Grossman LI, Hüttemann M. New prospects for an oldenzyme: mammalian cytochrome c is tyrosine-phosphorylated in vivo. Biochemistry. 2006;45:9121–9128. [PubMed: 16866357]

38. Yu H, Lee I, Salomon AR, Yu K, Hüttemann M. Mammalian liver cytochrome c is tyrosine-48phosphorylated in vivo, inhibiting mitochondrial respiration. Biochim Biophys Acta. 2008;1777:1066–1071. [PubMed: 18471988]

39. Korshunov SS, Krasnikov BF, Pereverzev MO, Skulachev VP. The antioxidant functions ofcytochrome c. FEBS Lett. 1999; 462:192–198. [PubMed: 10580118]

40. Wang ZB, Li M, Zhao Y, Xu JX. Cytochrome c is a hydrogen peroxide scavenger in mitochondria.Protein Pept Lett. 2003; 10:247–253. [PubMed: 12871144]

41. Giorgio M, Migliaccio E, Orsini F, Paolucci D, Moroni M, Contursi C, Pelliccia G, Luzi L,Minucci S, Marcaccio M, Pinton P, Rizzuto R, Bernardi P, Paolucci F, Pelicci PG. Electrontransfer between cytochrome c and p66Shc generates reactive oxygen species that triggermitochondrial apoptosis. Cell. 2005; 122:221–233. [PubMed: 16051147]

42. Steenaart NA, Shore GC. Mitochondrial cytochrome c oxidase subunit IV is phosphorylated by anendogenous kinase. FEBS Lett. 1997; 415:294–298. [PubMed: 9357986]

43. Bender E, Kadenbach B. The allosteric ATP-inhibition of cytochrome c oxidase activity isreversibly switched on by cAMP-dependent phosphorylation. FEBS Lett. 2000; 466:130–134.[PubMed: 10648827]

44. Lee I, Bender E, Kadenbach B. Control of mitochondrial membrane potential and ROS formationby reversible phosphorylation of cytochrome c oxidase. Mol Cell Biochem. 2002; 234–235:63–70.

45. Welch EJ, Jones BW, Scott JD. Networking with AKAPs: context-dependent regulation ofanchored enzymes. Mol Interv. 2010; 10:86–97. [PubMed: 20368369]

46. Lee I, Salomon AR, Ficarro S, Mathes I, Lottspeich F, Grossman LI, Hüttemann M. cAMP-dependent tyrosine phosphorylation of subunit I inhibits cytochrome c oxidase activity. J BiolChem. 2005; 280:6094–6100. [PubMed: 15557277]

Hüttemann et al. Page 17

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

47. Hüttemann M, Nantwi KD, Lee I, Liu J, Mohiuddin S, Petrov T. Theophylline treatment improvesmitochondrial function after upper cervical spinal cord hemisection. Exp Neurol. 2010; 223:523–528. [PubMed: 20144890]

48. Prabu SK, Anandatheerthavarada HK, Raza H, Srinivasan S, Spear JF, Avadhani NG. Proteinkinase A-mediated phosphorylation modulates cytochrome c oxidase function and augmentshypoxia and myocardial ischemia-related injury. J Biol Chem. 2006; 281:2061–2070. [PubMed:16303765]

49. Fang JK, Prabu SK, Sepuri NB, Raza H, Anandatheerthavarada HK, Galati D, Spear J, AvadhaniNG. Site Specific Phosphorylation of Cytochrome c Oxidase Subunits I, IVi1 and Vb in RabbitHearts Subjected to Ischemia/Reperfusion. FEBS Lett. 2007; 581:1302–1310. [PubMed:17349628]

50. Bain J, Plater L, Elliott M, Shpiro N, Hastie CJ, McLauchlan H, Klevernic I, Arthur JS, Alessi DR,Cohen P. The selectivity of protein kinase inhibitors: a further update. Biochem J. 2007; 408:297–315. [PubMed: 17850214]

51. Acin-Perez R, Salazar E, Kamenetsky M, Buck J, Levin LR, Manfredi G. Cyclic AMP producedinside mitochondria regulates oxidative phosphorylation. Cell Metab. 2009; 9:265–276. [PubMed:19254571]

52. Acin-Perez R, Salazar E, Brosel S, Yang H, Schon EA, Manfredi G. Modulation of mitochondrialprotein phosphorylation by soluble adenylyl cyclase ameliorates cytochrome oxidase defects.EMBO Mol Med. 2009; 1:392–406. [PubMed: 20049744]

53. Yang WL, Iacono L, Tang WM, Chin KV. Novel function of the regulatory subunit of proteinkinase A: regulation of cytochrome c oxidase activity and cytochrome c release. Biochemistry.1998; 37:14175–14180. [PubMed: 9760254]

54. Helling S, Vogt S, Rhiel A, Ramzan R, Wen L, Marcus K, Kadenbach B. Phosphorylation andkinetics of mammalian cytochrome c oxidase. Mol Cell Proteomics. 2008; 7:1714–1724.[PubMed: 18541608]

55. Tsukihara T, Shimokata K, Katayama Y, Shimada H, Muramoto K, Aoyama H, Mochizuki M,Shinzawa-Itoh K, Yamashita E, Yao M, Ishimura Y, Yoshikawa S. The low-spin heme ofcytochrome c oxidase as the driving element of the proton-pumping process. Proc Natl Acad Sci US A. 2003; 100:15304–15309. [PubMed: 14673090]

56. Augereau O, Claverol S, Boudes N, Basurko MJ, Bonneu M, Rossignol R, Mazat JP, Letellier T,Dachary-Prigent J. Identification of tyrosine-phosphorylated proteins of the mitochondrialoxidative phosphorylation machinery. Cell Mol Life Sci. 2005; 62:1478–1488. [PubMed:15924266]

57. Roberts VA, Pique ME. Definition of the interaction domain for cytochrome c on cytochrome coxidase. III. Prediction of the docked complex by a complete, systematic search. J Biol Chem.1999; 274:38051–38060. [PubMed: 10608874]

58. Salvi M, Brunati AM, Bordin L, La Rocca N, Clari G, Toninello A. Characterization and locationof Src-dependent tyrosine phosphorylation in rat brain mitochondria. Biochim Biophys Acta.2002; 1589:181–195. [PubMed: 12007793]

59. Miyazaki T, Neff L, Tanaka S, Horne WC, Baron R. Regulation of cytochrome c oxidase activityby c-Src in osteoclasts. J Cell Biol. 2003; 160:709–718. [PubMed: 12615910]

60. Miyazaki T, Tanaka S, Sanjay A, Baron R. The role of c-Src kinase in the regulation of osteoclastfunction. Mod Rheumatol. 2006; 16:68–74. [PubMed: 16633924]

61. Ogbi M, Johnson JA. Protein kinase Cε interacts with cytochrome c oxidase subunit IV andenhances cytochrome c oxidase activity in neonatal cardiac myocyte preconditioning. Biochem J.2006; 393:191–199. [PubMed: 16336199]

62. Ogbi M, Chew CS, Pohl J, Stuchlik O, Ogbi S, Johnson JA. Cytochrome c oxidase subunit IV as amarker of protein kinase Cε function in neonatal cardiac myocytes: implications for cytochrome coxidase activity. Biochem J. 2004; 382:923–932. [PubMed: 15339253]

63. Ping P, Zhang J, Zheng YT, Li RC, Dawn B, Tang XL, Takano H, Balafanova Z, Bolli R.Demonstration of selective protein kinase C-dependent activation of Src and Lck tyrosine kinasesduring ischemic preconditioning in conscious rabbits. Circ Res. 1999; 85:542–550. [PubMed:10488057]

Hüttemann et al. Page 18

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

64. Song C, Vondriska TM, Wang GW, Klein JB, Cao X, Zhang J, Kang YJ, D’Souza S, Ping P.Molecular conformation dictates signaling module formation: example of PKCε and Src tyrosinekinase. Am J Physiol Heart Circ Physiol. 2002; 282:H1166–1171. [PubMed: 11834516]

65. Pedersen PL. Tumor mitochondria and the bioenergetics of cancer cells. Prog Exp Tumor Res.1978; 22:190–274. [PubMed: 149996]

66. Boerner JL, Demory ML, Silva C, Parsons SJ. Phosphorylation of Y845 on the Epidermal GrowthFactor Receptor Mediates Binding to the Mitochondrial Protein Cytochrome c Oxidase Subunit II.Mol Cell Biol. 2004; 24:7059–7071. [PubMed: 15282306]

67. Demory ML, Boerner JL, Davidson R, Faust W, Miyake T, Lee I, Hüttemann M, Douglas R,Haddad G, Parsons SJ. Epidermal growth factor receptor translocation to the mitochondria:Regulation and effect. J Biol Chem. 2009; 284:36592–36604. [PubMed: 19840943]

68. Salvi M, Stringaro A, Brunati AM, Agostinelli E, Arancia G, Clari G, Toninello A. Tyrosinephosphatase activity in mitochondria: presence of Shp-2 phosphatase in mitochondria. Cell MolLife Sci. 2004; 61:2393–2404. [PubMed: 15378208]

69. Tartaglia M, Mehler EL, Goldberg R, Zampino G, Brunner HG, Kremer H, van der Burgt I,Crosby AH, Ion A, Jeffery S, Kalidas K, Patton MA, Kucherlapati RS, Gelb BD. Mutations inPTPN11, encoding the protein tyrosine phosphatase SHP-2, cause Noonan syndrome. Nat Genet.2001; 29:465–468. [PubMed: 11704759]

70. Neel BG, Gu H, Pao L. The ‘Shp’ing news: SH2 domain-containing tyrosine phosphatases in cellsignaling. Trends Biochem Sci. 2003; 28:284–293. [PubMed: 12826400]

71. Lee I, Pecinova A, Pecina P, Neel BG, Araki T, Kucherlapati R, Roberts AE, Hüttemann M. Asuggested role for mitochondria in Noonan syndrome. Biochim Biophys Acta. 2009; 1802:275–283. [PubMed: 19835954]

72. Robb-Gaspers LD, Burnett P, Rutter GA, Denton RM, Rizzuto R, Thomas AP. Integratingcytosolic calcium signals into mitochondrial metabolic responses. Embo J. 1998; 17:4987–5000.[PubMed: 9724635]

73. Kirichenko A, Vygodina T, Mkrtchyan HM, Konstantinov A. Specific cation binding site inmammalian cytochrome oxidase. FEBS Lett. 1998; 423:329–333. [PubMed: 9515733]

74. Kirichenko AV, Pfitzner U, Ludwig B, Soares CM, Vygodina TV, Konstantinov AA. Cytochromec oxidase as a calcium binding protein. Studies on the role of a conserved aspartate in helices XI-XII cytoplasmic loop in cation binding. Biochemistry. 2005; 44:12391–12401. [PubMed:16156652]

75. Hopper RK, Carroll S, Aponte AM, Johnson DT, French S, Shen RF, Witzmann FA, Harris RA,Balaban RS. Mitochondrial matrix phosphoproteome: effect of extra mitochondrial calcium.Biochemistry. 2006; 45:2524–2536. [PubMed: 16489745]

76. Kaim G, Dimroth P. ATP synthesis by F-type ATP synthase is obligatorily dependent on thetransmembrane voltage. Embo J. 1999; 18:4118–4127. [PubMed: 10428951]

77. Liu SS. Cooperation of a “reactive oxygen cycle” with the Q cycle and the proton cycle in therespiratory chain--superoxide generating and cycling mechanisms in mitochondria. J BioenergBiomembr. 1999; 31:367–376. [PubMed: 10665526]

78. Cossarizza A, Ceccarelli D, Masini A. Functional heterogeneity of an isolated mitochondrialpopulation revealed by cytofluorometric analysis at the single organelle level. Exp Cell Res. 1996;222:84–94. [PubMed: 8549677]

79. Picard M, Taivassalo T, Ritchie D, Wright KJ, Thomas MM, Romestaing C, Hepple RT.Mitochondrial Structure and Function Are Disrupted by Standard Isolation Methods. PLoS One.2011; 6:e18317. [PubMed: 21512578]

80. Hotchkiss RS, Karl IE. The pathophysiology and treatment of sepsis. N Engl J Med. 2003;348:138–150. [PubMed: 12519925]

81. Ruggieri AJ, Levy RJ, Deutschman CS. Mitochondrial dysfunction and resuscitation in sepsis. CritCare Clin. 2010; 26:567–575. x–xi. [PubMed: 20643307]

82. Duan ZX, Gu W, Zhang LY, Du DY, Hu P, Huang J, Liu Q, Wang ZG, Hao J, Jiang JX. Clinicalrelevance of the TLR4 11367 polymorphism in patients with major trauma. Arch Surg. 2009;144:1144–1148. [PubMed: 20026833]

Hüttemann et al. Page 19

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

83. Herrnstadt C, Howell N. An evolutionary perspective on pathogenic mtDNA mutations:haplogroup associations of clinical disorders. Mitochondrion. 2004; 4:791–798. [PubMed:16120433]

84. Baudouin SV, Saunders D, Tiangyou W, Elson JL, Poynter J, Pyle A, Keers S, Turnbull DM,Howell N, Chinnery PF. Mitochondrial DNA and survival after sepsis: a prospective study.Lancet. 2005; 366:2118–2121. [PubMed: 16360789]

85. Hayes MA, Timmins AC, Yau EH, Palazzo M, Hinds CJ, Watson D. Elevation of systemic oxygendelivery in the treatment of critically ill patients. N Engl J Med. 1994; 330:1717–1722. [PubMed:7993413]

86. Fink MP. Bench-to-bedside review: Cytopathic hypoxia. Crit Care. 2002; 6:491–499. [PubMed:12493070]

87. Lee MD, Zentella A, Vine W, Pekala PH, Cerami A. Effect of endotoxin-induced monokines onglucose metabolism in the muscle cell line L6. Proc Natl Acad Sci U S A. 1987; 84:2590–2594.[PubMed: 3472226]

88. Tracey KJ, Lowry SF, Fahey TJ, Albert JD, Fong Y, Hesse D, Beutler B, Manogue KR, Calvano S,Wei H, Cerami A, Shires GT. Cachectin/tumor necrosis factor induces lethal shock and stresshormone responses in the dog. Surg Gynecol Obstet. 1987; 164:415–422. [PubMed: 3576418]

89. Callahan LA, Supinski GS. Downregulation of diaphragm electron transport chain and glycolyticenzyme gene expression in sepsis. J Appl Physiol. 2005; 99:1120–1126. [PubMed: 16103521]

90. Levy RJ, Vijayasarathy C, Raj NR, Avadhani NG, Deutschman CS. Competitive andnoncompetitive inhibition of myocardial cytochrome C oxidase in sepsis. Shock. 2004; 21:110–114. [PubMed: 14752282]

91. Astiz M, Rackow EC, Weil MH, Schumer W. Early impairment of oxidative metabolism andenergy production in severe sepsis. Circ Shock. 1988; 26:311–320. [PubMed: 3208424]

92. Duvigneau JC, Piskernik C, Haindl S, Kloesch B, Hartl RT, Hüttemann M, Lee I, Ebel T, MoldzioR, Gemeiner M, Redl H, Kozlov AV. A novel endotoxin-induced pathway: upregulation of hemeoxygenase 1, accumulation of free iron, and free iron-mediated mitochondrial dysfunction. LabInvest. 2008; 88:70–77. [PubMed: 17982471]

93. Samavati L, Lee I, Mathes I, Lottspeich F, Hüttemann M. Tumor necrosis factor α inhibitsoxidative phosphorylation through tyrosine phosphorylation at subunit I of cytochrome c oxidase.J Biol Chem. 2008; 283:21134–21144. [PubMed: 18534980]

94. Knab S, Mushak TM, Schmitz-Esser S, Horn M, Haferkamp I. Nucleotide parasitism by Simkanianegevensis (Chlamydiae). J Bacteriol. 2011; 193:225–235. [PubMed: 20971898]

95. Roger VL, Go AS, Lloyd-Jones DM, Adams RJ, Berry JD, Brown TM, Carnethon MR, Dai S, deSimone G, Ford ES, Fox CS, Fullerton HJ, Gillespie C, Greenlund KJ, Hailpern SM, Heit JA, HoPM, Howard VJ, Kissela BM, Kittner SJ, Lackland DT, Lichtman JH, Lisabeth LD, Makuc DM,Marcus GM, Marelli A, Matchar DB, McDermott MM, Meigs JB, Moy CS, Mozaffarian D,Mussolino ME, Nichol G, Paynter NP, Rosamond WD, Sorlie PD, Stafford RS, Turan TN, TurnerMB, Wong ND, Wylie-Rosett J. Heart disease and stroke statistics--2011 update: a report from theAmerican Heart Association. Circulation. 2011; 123:e18–e209. [PubMed: 21160056]

96. Rosenthal RE, Hamud F, Fiskum G, Varghese PJ, Sharpe S. Cerebral ischemia and reperfusion:prevention of brain mitochondrial injury by lidoflazine. J Cereb Blood Flow Metab. 1987; 7:752–758. [PubMed: 3693430]

97. Zaidan E, Sims NR. The calcium content of mitochondria from brain subregions following short-term forebrain ischemia and recirculation in the rat. J Neurochem. 1994; 63:1812–1819. [PubMed:7931337]

98. Vlessis AA, Widener LL, Bartos D. Effect of peroxide, sodium, and calcium on brainmitochondrial respiration in vitro: potential role in cerebral ischemia and reperfusion. JNeurochem. 1990; 54:1412–1418. [PubMed: 2313294]

99. Ekholm A, Katsura K, Kristian T, Liu M, Folbergrova J, Siesjo BK. Coupling of cellular energystate and ion homeostasis during recovery following brain ischemia. Brain Res. 1993; 604:185–191. [PubMed: 8457847]

Hüttemann et al. Page 20

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

100. Fabian RH, DeWitt DS, Kent TA. In vivo detection of superoxide anion production by the brainusing a cytochrome c electrode. J Cereb Blood Flow Metab. 1995; 15:242–247. [PubMed:7860658]

101. Piantadosi CA, Zhang J. Mitochondrial generation of reactive oxygen species after brain ischemiain the rat. Stroke. 1996; 27:327–332. [PubMed: 8571432]

102. Liu RR, Murphy TH. Reversible cyclosporin A-sensitive mitochondrial depolarization occurswithin minutes of stroke onset in mouse somatosensory cortex in vivo: a two-photon imagingstudy. J Biol Chem. 2009; 284:36109–36117. [PubMed: 19892710]

103. Haines BA, Mehta SL, Pratt SM, Warden CH, Li PA. Deletion of mitochondrial uncouplingprotein-2 increases ischemic brain damage after transient focal ischemia by altering geneexpression patterns and enhancing inflammatory cytokines. J Cereb Blood Flow Metab. 2010;30:1825–1833. [PubMed: 20407461]

104. Brennan JP, Southworth R, Medina RA, Davidson SM, Duchen MR, Shattock MJ. Mitochondrialuncoupling, with low concentration FCCP, induces ROS-dependent cardioprotection independentof KATP channel activation. Cardiovasc Res. 2006; 72:313–321. [PubMed: 16950237]

105. Pandya JD, Pauly JR, Sullivan PG. The optimal dosage and window of opportunity to maintainmitochondrial homeostasis following traumatic brain injury using the uncoupler FCCP. ExpNeurol. 2009; 218:381–389. [PubMed: 19477175]

106. Choi K, Kim J, Kim GW, Choi C. Oxidative stress-induced necrotic cell death via mitochondira-dependent burst of reactive oxygen species. Curr Neurovasc Res. 2009; 6:213–222. [PubMed:19807658]

107. Racay P, Tatarkova Z, Chomova M, Hatok J, Kaplan P, Dobrota D. Mitochondrial calciumtransport and mitochondrial dysfunction after global brain ischemia in rat hippocampus.Neurochem Res. 2009; 34:1469–1478. [PubMed: 19252983]

108. Allen KL, Almeida A, Bates TE, Clark JB. Changes of respiratory chain activity in mitochondrialand synaptosomal fractions isolated from the gerbil brain after graded ischaemia. J Neurochem.1995; 64:2222–2229. [PubMed: 7722507]

109. Green DR, Kroemer G. The pathophysiology of mitochondrial cell death. Science. 2004;305:626–629. [PubMed: 15286356]

110. Martin LJ, Brambrink AM, Price AC, Kaiser A, Agnew DM, Ichord RN, Traystman RJ. Neuronaldeath in newborn striatum after hypoxia-ischemia is necrosis and evolves with oxidative stress.Neurobiol Dis. 2000; 7:169–191. [PubMed: 10860783]

111. Zhang Y, Marcillat O, Giulivi C, Ernster L, Davies KJ. The oxidative inactivation ofmitochondrial electron transport chain components and ATPase. J Biol Chem. 1990; 265:16330–16336. [PubMed: 2168888]

112. Zhang M, Mileykovskaya E, Dowhan W. Gluing the respiratory chain together. Cardiolipin isrequired for supercomplex formation in the inner mitochondrial membrane. J Biol Chem. 2002;277:43553–43556. [PubMed: 12364341]

113. Vannucci RC, Towfighi J, Vannucci SJ. Secondary energy failure after cerebral hypoxia-ischemiain the immature rat. J Cereb Blood Flow Metab. 2004; 24:1090–1097. [PubMed: 15529009]

114. Kagan VE, Bayir HA, Belikova NA, Kapralov O, Tyurina YY, Tyurin VA, Jiang J, StoyanovskyDA, Wipf P, Kochanek PM, Greenberger JS, Pitt B, Shvedova AA, Borisenko G. Cytochrome c/cardiolipin relations in mitochondria: a kiss of death. Free Radic Biol Med. 2009; 46:1439–1453.[PubMed: 19285551]

115. Pecina P, Borisenko GG, Belikova NA, Tyurina YY, Pecinova A, Lee I, Samhan-Arias AK,Przyklenk K, Kagan VE, Hüttemann M. Phosphomimetic substitution of cytochrome C tyrosine48 decreases respiration and binding to cardiolipin and abolishes ability to trigger downstreamcaspase activation. Biochemistry. 2010; 49:6705–6714. [PubMed: 20586425]

116. Hetz C, Vitte PA, Bombrun A, Rostovtseva TK, Montessuit S, Hiver A, Schwarz MK, ChurchDJ, Korsmeyer SJ, Martinou JC, Antonsson B. Bax channel inhibitors prevent mitochondrion-mediated apoptosis and protect neurons in a model of global brain ischemia. J Biol Chem. 2005;280:42960–42970. [PubMed: 16219766]

Hüttemann et al. Page 21

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

117. Sanderson TH, Kumar R, Sullivan JM, Krause GS. Insulin blocks cytochrome c release in thereperfused brain through PI3-K signaling and by promoting Bax/Bcl-XL binding. J Neurochem.2008; 106:1248–1258. [PubMed: 18518905]

118. Lee I, Salomon AR, Yu K, Samavati L, Pecina P, Pecinova A, Hüttemann M. Isolation ofregulatory-competent, phosphorylated cytochrome c oxidase. Methods Enzymol. 2009; 457:193–210. [PubMed: 19426869]

119. Sanishvili R, Volz KW, Westbrook EM, Margoliash E. The low ionic strength crystal structure ofhorse cytochrome c at 2.1 Å resolution and comparison with its high ionic strength counterpart.Structure. 1995; 3:707–716. [PubMed: 8591047]

120. Thingholm TE, Jorgensen TJ, Jensen ON, Larsen MR. Highly selective enrichment ofphosphorylated peptides using titanium dioxide. Nat Protoc. 2006; 1:1929–1935. [PubMed:17487178]

Hüttemann et al. Page 22

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig. 1.Mapped phosphorylation sites on cytochrome c and cytochrome c oxidase. Crystal structuredata of horse heart cytochrome c [119] and cow heart COX [12] were used and processedwith the program Swiss-PDBViewer 3.7. Identified phosphorylated amino acids inmammals are shown in sticks. See Table 1 for a detailed description of the sites includingphospho-epitopes and references. Note that Thr52 in rabbit corresponds to Ser52 in cowCOX subunit IV.

Hüttemann et al. Page 23

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig. 2.Nano-LC-ESI-MS/MS analysis of Blue Native-PAGE-separated, BrCN and trypsin cleavedCOX subunit II isolated from cow heart. Prior to analysis by tandem mass spectrometry(MS/MS) the phosphopeptides were enriched with titanium dioxide as described [54, 120].The MS/MS spectrum unambiguously identifies the peptide IPTpSELKPGELR. Thelocalization of Ser126 as the site of phosphorylation was possible after fragmentation of thedoubly charged precursor ion (710.5 Th) with collision-induced dissociation (CID) and thetriply charged ion (474.3 Th) with electron transfer dissociation (ETD). A, the CID-MS/MSspectrum contains a CID-typical neutral loss of phosphoric acid from the phosphoserine ofthe precursor ion. The y10 and b3 fragment ions as well as the neutral loss ions y9 and b5(both -H3PO4) indicate phosphoserine. B, the ETD fragmentation resulted in a spectrum

Hüttemann et al. Page 24

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

specifying the peptide amino acid sequence including the phosphosite-specific C-terminal z8and z9 ions and the corresponding N-terminal c3 and c4 ions. Furthermore, on the basis ofthe fragmentation pattern phosphorylation of Thr125 was excluded by the presence of ionsc2, z9 and z10.

Hüttemann et al. Page 25

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Fig. 3.Proposed model linking COX activity with the mitochondrial membrane potential, ATP andROS production under healthy conditions and in two distinct pathological states. Underhealthy conditions (middle) COX and Cytc are phosphorylated leading to a partial inhibitionof respiration (Cytc is not shown). This generates healthy mitochondrial membranepotentials ΔΨm of about 120 mV, sufficient for efficient ATP generation but too low forROS production. Two distinct pathological conditions are shown, both of which can lead tocell death: 1) acute inflammatory signaling via TNFα causes COX subunit I Tyr304phosphorylation (pY304), leading to strong COX inhibition, decreased ΔΨm levels, andeventually energy depletion. In patients with sepsis this model would explain organ failureand death through energy failure; 2) the opposite mode of action takes place duringischemia/reperfusion. During ischemia, nutrients and oxygen become depleted. This stresscauses excessive calcium release leading to changes in the phosphorylation pattern and/ordephosphorylation of Cytc and COX. The electron transport chain is now ‘primed’ forhyperactivity. During reperfusion, in the presence of oxygen and metabolites, OxPhosresumes to rebuild ΔΨm and ATP. Since the ETC complexes are in a hyperactive state, fluxis maximal, which leads to a hyperpolarization of ΔΨm, reaching levels at which excessiveROS are generated. ROS cause extensive damage to the cell and trigger apoptosis.

Hüttemann et al. Page 26

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

NIH

-PA Author Manuscript

Hüttemann et al. Page 27

Tabl

e 1

Iden

tifie

d ph

osph

o-ep

itope

s in

mam

mal

ian

cyto

chro

me

c an

d cy

toch

rom

e c

oxid

ase1

Enz

yme

Spec

ies &

tiss

uePh

osph

oryl

ated

am

ino

acid

2Ph

osph

o-ep

itope

3R

efer

ence

Met

hod

Kin

ase

pred

ictio

n4

Cyt

cH

uman

skel

etal

mus

cle

Thr2

8EK

GG

KH

KT

GPN

LHG

L[3

7]H

T-M

S5A

MP

kina

se, l

ow c

onfid

ence

Cyt

cH

uman

skel

etal

mus

cle

Ser4

7TG

QA

PGY

SYTA

AN

KN

[37]

HT-

MS

no k

now

n m

otif

Cyt

cC

ow li

ver

Tyr4

8G

QA

PGFS

YTD

AN

KK

N[4

0]M

Sno

kno

wn

mot

if

Cyt

cC

ow h

eart

Tyr9

7ER

EDLI

AY

LKK

ATN

E[3

9]M

Sno

kno

wn

mot

if

CO

XC

ow li

ver

Tyr3

04, S

U I

MD

VD

TRA

YFT

SATM

I[4

8]M

S, c

AM

P-de

pend

ent

no k

now

n m

otif

CO

XC

ow li

ver

Tyr3

04, S

U I

MD

VD

TRA

YFT

SATM

I[9

3]Ph

osph

o-ep

itope

- spe

cific

ant

ibod

y; T

NFα

-de

pend

ent

no k

now

n m

otif

CO

XR

abbi

t hea

rtSe

r115

and

Ser

116,

SU

ISL

HLA

GV

SSIL

GA

INF

[51]

MS;

afte

r isc

hem

ia/re

perf

usio

nno

kno

wn

mot

if

CO

XC

ow h

eart

Ser1

26, S

U II

DSY

MIP

TSEL

KPG

ELTh

is st

udy

MS

no k

now

n m

otif

CO

XC

ow li

ver

Tyr1

1, S

U IV

-1SV

VK

SED

YA

LPSY

VD

[39]

MS

Abl

kin

ase,

low

con

fiden

ce

CO

XC

ow h

eart

Ser3

4, S

U IV

-1V

AH

VK

NLS

ASQ

KA

LK[5

6]M

SC

AM

K2G

, low

con

fiden

ce

CO

XH

eLa

cells

Ser6

7, S

U IV

-1Y

RIK

FKES

FAEM

NR

G[3

8]H

T-M

SPK

Cε

CO

XR

abbi

t hea

rtTh

r52,

SU

IV-1

KA

PWG

SLT

RD

EKV

EL[5

1]M

S; a

fter i

sche

mia

/repe

rfus

ion

DN

A P

K, l

ow c

onfid

ence

CO

XH

eLa

cells

Ser1

36, S

U IV

-1N

PIQ

GLA

SKW

DY

EKN

[38]

HT-

MS

no k

now

n m

otif

CO

XC

ow h

eart

Ser4

, SU

Va

SHG

SHET

DEE

F[5

6]M

Sno

kno

wn

mot

if

CO

XC

ow h

eart

Thr3

5, S

U V

aEL

RK

GM

NT

LVG

YD

LV[5

6]M

SPK

Cε,

low

con

fiden

ce

CO

XH

eLa

cells

Thr3

5 an

d Th

r38,

SU

Va

LRK

GIN

TLV

TY

DM

VPE

[38]

HT-

MS

PKCε

& D

NA

PK

, low

con

fiden

ce

CO

XR

abbi

t hea

rtSe

r40,

SU

Vb

MLP

PKA

ASG

TKED

PN[5

1]M

S; a

fter i

sche

mia

/repe

rfus

ion

no k

now

n m

otif

CO

XC

ow h

eart

Thr1

1, S

U V

IaA

KG

DH

GG

TG

AR

TWR

F[5

7]C

ryst

al st

ruct

ure

no k

now

n m

otif

1 Iden

tifie

d ph

osph

oryl

atio

n si

tes a

re p

rese

nted

in a

n as

cend

ing

amin

o ac

id a

nd su

buni

t (SU

) ord

er.

2 Num

berin

g ac

cord

ing

to m

atur

e pe

ptid

e ba

sed

on c

ow.

3 Sequ

ence

is b

ased

on

the

spec

ies w

here

the

phos

phor

ylat

ion

site

was

iden

tifie

d.

4 Bas

ed o

n th

e Sc

ansi

te M

otif

Scan

web

tool

(http

://sc

ansi

te.m

it.ed

u/m

otifs

can_

seq.

phtm

l). T

he p

redi

ctio

ns a

re p

urel

y sp

ecul

ativ

e an

d ha

ve to

be

view

ed w

ith e

xtre

me

caut

ion.

5 Hig

h th

roug

hput

mas

s spe

ctro

met

ry se

quen

cing

Biochim Biophys Acta. Author manuscript; available in PMC 2013 April 1.