Principles of Molecular Bioenergetics and The Proton Pump of Cytochrome Oxidase

ROBERT B. GENNIS

University of Illinois, Department of Biochemistry, 600 South Mathews Street, Urbana, IL 61801

1 Introduction: General Principles of Bioenergetic Systems

All of the bioenergetic enzymes described in this book can couple an exergonic or free-energy yielding reaction to the electrogenic movement of charged species across the membrane, generating a protonmotive force. In the case of bacteriorhodopsin, the driving reaction is the absorption of a photon, for the bc1 complex, the oxidation of ubiquinol by cytochrome c is the driving force, and for the respiratory oxidases, the reduction of O2 to H2O provides the impetus. In this book, the principles utilized by a number of these systems are detailed with an emphasis on recent structural studies. It is convenient to classify two classes of mechanisms used to generate a trans-membrane voltage:

(1) Mechanisms utilizing an oxidoreduction loop.

(2) True ion (proton) pumps.

1.1 Oxidoreduction Loops

The principle of coupling different chemical reactions is central to biology and is accomplished in a number of ways. Many of the systems that generate a protonmotive force can be understood in terms of Mitchell's chemiosmotic oxidoreduction loop. This is illustrated by the example shown in Figure 1, which shows a redox loop formed from the anaerobic respiratory system comprised of formate dehydrogenase and nitrate reductase enzymes from E. coli. Recently, the structures of each of these two enzymes were determined. The topology of the catalytic active sites assures that the net reaction results in the generation of a protonmotive force. Formate dehydrogenase oxidizes formate on the periplasmic side of the membrane (the positive or P-side) and electrons are delivered through a series of metal centers to a menaquinone reductase site located near the cytoplasmic surface (the negative or N-side of the membrane). The formate dehydrogenase, thus, separates the oxidative and reductive half-reactions on opposite sides of the membrane. Protons are released in the periplasm upon formate oxidation and protons are taken up from the cytoplasm upon the reduction of menaquinone. The actual charge crossing the chemiosmotic barrier is the electron.

Reduced menaquinol is a neutral, hydrophobic compound and can diffuse freely within and across the membrane bilayer. The nitrate reductase enzyme has a menaquinol oxidation site located near the periplasm, whereas the site where nitrate is reduced to nitrite is located on the opposite side of the membrane. Electrons are transferred across the membrane between these active sites to couple the two half-reactions catalyzed by the enzyme (see Figure 1). The full reaction of nitrate reductase, therefore, is coupled to the release of protons in the periplasm, the uptake of protons from the cytoplasm and the transfer of charges, in the form of electrons, across the membrane.

The net reaction of both of these enzymes together results in the transfer of four protons from the cytoplasm to the periplasm for each formate oxidized and nitrate reduced. Points to note are

(1) The actual charges crossing the membrane are electrons and not protons.

(2) The net transfer of protons is due to the vectorial placement of the enzyme active sites so that the oxidation and reduction half-reactions occur on oppo site sides of the membrane.

(3) The protons are directly involved in the substrate chemistry.

(4) The two enzymes are coupled by a neutral, hydrophobic hydrogen carrier, in this case, menaquinol.

(5) The generation of the protonmotive force cannot be decoupled from the chemical reaction without changing the 'wiring'. The topology of the active sites, located on opposite sides of the membrane, and the uptake/release of protons from/to the N/P side of the membrane enforce this coupling so that the chemistry cannot proceed without generating a transmembrane voltage.

These are general features of Mitchell's initial proposal for how the protonmotive force is generated, and the structural and functional studies on these and other systems have supported this proposal.

The photosynthetic reaction center and the bc1 (and b6f) complex can be understood as variations of this same general principle, illustrated schematically in Figure 2. In the case of the reaction center, the absorption of a photon results in electron transfer across the membrane leading to charge separation. The reductive and oxidative reactions that follow occur on opposite sides of the membrane. The direction of electron flow is from the P-side to the N-side of the membrane, as is also the case for nitrate reductase and formate dehydrogenase. The reduction of ubiquinone in the bacterial reaction center utilizes protons from the N-side of the membrane (bacterial cytoplasm for the prokaryotic enzyme) and the protons are delivered to the buried QB active site through a proton-conducting channel. In essence, the proton is transferred electrogenically towards the P-side of the membrane to 'meet' the electron at the quinone reduction site. Both proton and electron transfers contribute to the voltage generation by the reaction. The charge movement across the membrane is, thus, comprised partially by the electron transfer from the P- to the N-side of the membrane and partially by the proton being transferred in the opposite direction.

The topological constraint of the bc1 center is that the quinol oxidation site and the site where cytochrome c is reduced are both located on the P-side of the membrane (Figure 2). Simple electron transfer between these sites would not be coupled to the generation of a protonmotive force. The enzyme has evolved a remarkable 'Q-cycle' to solve this problem. The enzyme transfers one electron from reduced quinol to cytochrome c and uses the free energy of this reaction to drive the second electron across the membrane (P-side to N-side) to a quinol reductase site, located on the opposite side of the membrane. For every two quinols oxidized, one quinol is regenerated, two electrons cross the membrane, four protons are released to the periplasm and two protons are taken up from the cytoplasm. Although more complicated, the principles of how the protonmotive force is generated are those of the oxidoreduction 'loop'. The idea of the 'Q-loop' modification was also proposed long before the structures of the bc1/b6f complexes were known, but the structural data fully support the functional model.

Finally, although the structure of the cytochrome bd quinol oxidase is not known, recent evidence suggests that this represents yet another variation of the oxidoreduction loop (Figure 2). In this case, the two half-reactions, quinol oxidation and O2 reduction to water, also occur on the same side of the membrane. However, although the active site where O2 reduction occurs is near the P-side of the membrane, the protons are delivered to this active site from the opposite side of the membrane. Hence, the proton delivery pathway enforces the coupling between the chemistry catalyzed by the enzyme and the generation of a trans-membrane voltage. The working model is that the voltage is generated virtually entirely by protons crossing the bulk of the membrane to 'meet' the electron at the O2 reduction site.

1.2 Proton Pumps

The transhydrogenase, ATP synthase, bacteriorhodopsin and cytochrome oxidase each also generate a protonmotive force principally or entirely by the transfer of protons across the electrochemical barrier of the membrane. Each of these systems must also have pathways that facilitate proton transfer across the membrane. However, in contrast to the systems described in the previous section, the transhydrogenase, ATP synthase, bacteriorhodopsin and cytochrome oxidase use a mechanism in which the protons transferred across the membrane are not directly involved in the chemical reactions driving the charge separation. One consequence of this is that it is possible in each of these enzymes to find mutants that decouple the proton pump from the chemistry.

In each of these systems, the driving reaction forces the protein into a transient, unstable state with a high chemical potential. This state then relaxes back to the ground state via a specific kinetic pathway which couples the dissipation of the transiently stored free energy to the generation of a protonmotive force. A high energy state of the protein, or a sequence of such states, serves as an intermediate to capture the free energy of one process (e.g. reduction of O2 to water) and then utilize this free energy to drive a second process (e.g. electrogenic proton pumping). We will reserve the term 'proton pump' for the bioenergetic enzymes that operate in this manner and distinguish them from the systems that generate a protonmotive force through an oxidoreductive loop.

By far the best understood proton pump is bacteriorhodopsin. It is useful to briefly review the mechanism of proton pumping by bacteriorhodopsin in order to put the cytochrome oxidase studies in perspective.

Bacteriorhodopsin has at least four protonation sites along a proton-conducting channel (D85, D96, Schiff's base, proton-release cluster). Each of these groups undergo large pK]a changes (5 to 10 pH units) as the protein goes through the sequence of intermediate states after absorption of a photon by retinal. At an early stage following the light-induced isomerization of retinal, an internal proton moves from the Shiff's base to D85, and a proton bound at the release cluster at the periplasmic (P-side) surface is ejected to the aqueous phase. Only after this has occurred is the internal proton on D96 transferred to the Schiff's base and then refilled through a channel from the cytoplasmic aqueous medium. Changes in the proton affinity at different sites within the protein during the photocycle is of obvious importance to understand why the protons shift between internal sites and are taken up/released from the bulk medium at specific times in the photocycle. In addition, the kinetics of proton transfer is equally important, since it is this feature that assures that the pump is unidirectional. The kinetics assure that the Schiff's base is reprotonated from the cytoplasmic side and not the periplasmic side of the membrane. The kinetics, in turn, are determined by structural features that facilitate or retard specific proton transfer reactions.

Recent structural and spectroscopic studies of bacteriorhodopsin have highlighted the role of internal water molecules. As the protein changes conformation, water molecules shift positions, and water may also be taken up from or lost to the bullk medium. The changing hydogen-bonding patterns of internal water molecules not only play a major role in the proton affinities of ionizable residues within the protein, but also provide pathways for proton transfer within the protein. The elemental step of proton transfer is a shift of a proton from a hydrogen-bond donor to a hydrogen-bond acceptor. Transfer over large distances requires movement along a series of hydrogen bonds. Because of this, changing the orientation or position of a single water molecule could have a very large influence on whether a proton pathway is functional. For example, it appears that the movement of water molecules essentially creates a proton-conducting channel from D96 to the Schiff's base in bacteriorhodopsin late in the photocycle, apparently gating proton transfer to re-protonate the Schiff's base.

Although much less is known about the transhydrogenase and ATP synthase proton pumping enzymes, it is clear that each of these proteins also undergoes conformational changes during the catalytic cycle. In the case of the ATP synthase, this is in the form of the rotation of the C-subunit ring and its associated γ-subunit, which distorts the active site where ATP is synthesized. The details of the conformational changes at the active site are provided by the X-ray structure of the F1 portion of the ATPase.

1.3 Cytochrome Oxidase Uses Both Mechanisms of Energy Coupling

Cytochrome oxidase is a unique bioenergetic enzyme in that it generates a trans-membrane voltage using both the oxidoreductive loop principle and a true proton pump mechanism. For each electron passing through the enzyme from reduced cytochrome c to O2, two charges cross the membrane.

The mitochondrial enzyme is just one of a large superfamily referred to as the heme–copper oxidases. All of the heme–copper oxidases catalyze the reduction of O2 to H2O at a bimetallic active site containing a heme and a Cu atom. The heme component has an open coordination site where O2 can bind to the ferrous heme Fe and, subsequently, be reduced to H2O.

O2 + 4e-1 + 4 H+ [??] 2 H2O (1)

Studies with representatives of each of the major sub-classes of the heme-copper oxidases strongly suggest that all the heme–copper oxidases also are proton pumps. The invarant aspects for all these enzymes are as follows:

(1)The heme-copper active site is buried within the membrane, approximately halfway across the dielectric barrier. In the mitochondrial enzyme, the active site components are heme a3 and CuB. The 'a' denotes the chemical structure of the heme and the subscript '3' has come to mean that this is the oxygen binding site.

(2) Electrons are directed to the O2 reductase site from a second active site located at or near the P-side of the membrane where the reduced substrate is oxidised. Most of the superfamily members oxidize ferrous cytochrome c, but a number of heme–copper oxidases oxidize quinol. The most extensively studied quinol oxidase is cyochrome bo from E. coli.

(3) The electrons are transferred between the two active sites by a series of metal redox centers, the last of which is always a six-coordinate heme. Hence, all the heme–copper oxidases contain two hemes. These two hemes are located within the same subunit, with the Fe–Fe distance near to 12 Å and the heme edges as close as 5 Å. In the mitochondrial oxidase, the second heme is heme a, so this enzyme is called an aa3-type cytochrome c oxidase.

(4) Proton input channels assure that all the protons that are used to generate water from O2 come from the N-side of the membrane (mitochondrial matrix or bacterial cytoplasm). Proton channels or pathways also are critical to the transfer of protons all the way across the membrane, which is necessary for the proton pump. Hence, one proton must traverse about half of the distance across the membrane to 'meet' the electron at the oxygen at the heme–copper active site (an oxidoreduction loop), whereas the second proton is driven all the way across the membrane (a proton pump). This is shown schematically in Figure 3. We can take this topology and proton pumping into account in the actual reaction catalyzed by the enzyme.

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] 2)

This chemical reaction indicates that three processes are coupled together by the enzyme: (i) electron transfer to the active site; (ii) proton delivery to the active site; (iii) proton pumping. Furthermore, the reaction occurs in multiple steps. An understanding of the molecular mechanism of the proton pump can only follow an understanding of the chemistry that occurs at the enzyme active site. A brief review of the enzyme structure and the chemical mechanism by which O2 is activated and reduced to water is a necessary prerequisite to discussing the proton pump.