Proton circuits across the inner mitochondrial membrane link the primary energy generators, namely the complexes of the electron transport chain, to multiple energy utilizing processes, including the ATP synthase, inherent proton leak pathways, metabolite transport and linked circuits of sodium and calcium. These mitochondrial circuits can be monitored in both isolated preparations and intact cells and, for the primary proton circuit techniques, exist to follow both the proton current and proton electrochemical potential components of the circuit in parallel experiments, providing a quantitative means of assessing mitochondrial function and, equally importantly, dysfunction.
The field of mitochondrial bioenergetics has been undergoing an explosive expansion into the fields of physiology and cell biology in the past decade with the realization that the life and death of cells is integrally related to the underlying bioenergetic state of the cells’ mitochondria. Many forms of apoptotic cell death are triggered by the release of cytochrome c from mitochondria, mitochondrial dysfunction lies at the heart of many neurodegenerative disorders such as Parkinson’s disease and stroke, mitochondria play an integral role controlling insulin secretion, while oxidative stress originating at the mitochondria is implicated in certain theories of aging. In addition, great advances are being made in understanding the interactions between the mitochondrion and the rest of the cell, and mechanisms of mitochondrial transport and biosynthesis (99% of the mitochondrion’s proteins are imported from the cytoplasm and integrate with the 1% that are synthesized on the mitochondrial genome). All of this excitement is resulting in the publication of 10000 mitochondrial papers each year, and it is fair to say that an understanding of basic mitochondrial function and bioenergetics is essential for the physiologist and cell biologist.
The elements of mitochondrial bioenergetics are relatively straight-forward. Substrate dehydrogenases deliver electrons to the respiratory chain via the intermediacy of carriers such as NADH. The respiratory chain functions as a linked sequence of three proton-pumping complexes, and the drop in energy as each pair of electrons is transported down the chain is conserved in the pumping of approx. ten protons out of the mitochondrial matrix across the inner membrane. The protons re-enter the matrix driving a fourth, ATP-hydrolysing, proton pump (the ATP synthase) in reverse, resulting in the synthesis of ATP. This ‘proton circuit’ is at the centre of mitochondrial bioenergetics and is the subject, together with other linked ion circuits, of this chapter.
The proton circuit
In 1965 Peter Mitchell, in the first of his famous ‘Little Grey Books’ , published a remarkable hypothetical Figure which is reproduced and updated here as Figure 1(A). In the context of the times where mitochondrial energy metabolism was exclusively thought of in terms of chemical linkages, this remarkably prescient figure proposed a total paradigm shift, whereby the electron transport chain was linked to the ATP synthase by a proton circuit, which also had secondary linkages to the transport of metabolites, calcium and sodium. In contrast to Mitchell’s other seminal proposals on the nature of the constituent proton pumps, which were proposed before protein structure and confirmation were properly understood, and which have been greatly modified and refined in the subsequent literature, this single figure has remarkably withstood the passage of time, and today, after more than 40 years, retains all of the essential information for our current understanding of ion cycling across the mitochondrial inner membrane .
An easy way to visualize the basic mitochondrial proton circuit across the inner membrane is to consider the simple electrical analogy (Figure 1B). All energy fluxes have intensive and extensive components. In the case of a waterfall the intensive component is the height of the fall, the extensive component the water flow, and the power the product of the two. The corresponding components of an electrical circuit are voltage, current and wattage, while in the proton circuit the voltage component is the proton electrochemical potential, or protonmotive force, Δp, across the membrane defined by the relationship: where Δψ is the transmembrane potential, in mV, and ΔpH is the transmembrane pH gradient.
‘Proton voltage’: membrane potential and ΔpH
Since Δψ is the dominant component of the proton electrochemical potential, ΔpH is frequently ignored in studies both with isolated mitochondria and intact cells. However, it must always be borne in mind that doing so induces a significant error, since a typical ΔpH of 0.5 units contributes 30 mV to the full potential. ΔpH can be estimated from the distribution of a weak acid such as acetate, which is only permeable as the neutral protonated acid, and dissociates according to its pK on each side of the membrane . Alternatively, matrix pH can be estimated in cells from the fluorescence properties of matrix-incorporated pH-sensitive GFPs (green fluorescent proteins) , or by 31P-NMR , and the difference from cytoplasmic pH calculated. It is particularly important to monitor ΔpH when the matrix pH can change, for example during the uptake of cations in the presence of limiting phosphate. An example in intact cells where large fluctuations in matrix pH have been reported is the response of insulin-secreting cells to increased glucose availability . Since a central component in GSIS (glucose-stimulated insulin secretion) is an increased Δp , simple measurement of Δψ would substantially underestimate the full bioenergetic hyperpolarization.
Although absolute values of ΔpH may be obtained with reasonable accuracy, it is exceedingly difficult to obtain precise quantitative values for Δψ, particularly for mitochondria in intact cells. All techniques depend on estimating the equilibrium concentration gradient of lipophilic cations across the membrane and substituting this value into the Nernst equation: where X+ is a membrane-permeant monovalent cation. With isolated mitochondria TPP+ (tetraphenylphosphonium cation) is frequently used in combination with an external electrode that monitors the decrease in the concentration of the cation in the medium as it is accumulated by the mitochondria . However, in order to accurately calculate the potential it is necessary to know the matrix volume and to correct for binding of the probe to the membrane itself. In this context it is important to remember that the mitochondrial matrix is far from an ideal solution and, in fact, consists of approx. 50% protein and 50% water, and can be more accurately visualized as a viscous gel, and that simple Nerstian substitutions are approximations.
A variety of membrane-permeant cations have been developed for optical determinations of mitochondrial membrane potential with the isolated organelles and intact cells . Most of these probes showed the phenomenon of fluorescence quenching associated with their aggregation which occurs at a critical concentration that may readily be attained within the mitochondrial matrix. Fluorescence quenching may be exploited with isolated mitochondria in a cuvette, since the total fluorescence will decrease as the probe accumulates in the mitochondrial matrix above this threshold. With both isolated mitochondria and intact cells it is essential to establish that the probe does not itself affect mitochondrial bioenergetics. For example several cyanine dyes used as membrane-potential indicators are potent inhibitors of mitochondrial complex I . Estimation of mitochondrial membrane potential within intact cells is considerably more complex, and virtually all determinations are semiquantitative, at best estimating changes in potential from an assumed initial starting value. While principles similar to isolated mitochondria apply for the distribution of the probe across the inner mitochondrial membrane, permeant cations must first be transported across the plasma membrane. Exogenous probe will therefore be accumulated across the plasma membrane in response to the plasma membrane potential, and then across the inner mitochondrial membrane in response to the mitochondrial membrane potential. The fluorescence signal from these so-called mitochondrial membrane-potential indicators is thus a function of both membrane potentials .
The optical resolution of a fluorescence microscope is sufficient to resolve single mitochondria in thin processes, while a confocal microscope can visualize the three-dimensional array of mitochondria in an intact cell body. From first principles it should be possible to determine directly the difference in membrane potential probe concentration between the mitochondrial matrix and the cytoplasm; however, in practise this demands a high dynamic range and it can be difficult to quantify the faint cytoplasmic signal and in particular to differentiate it from the extracellular background. In addition, mitochondria are mobile and the means must be found to distinguish a true depolarization from the movement of the mitochondrion out of the focal plane of the objective. This last problem can be resolved by the inclusion of a membrane-potential-independent mitochondrial fluorescence that can be used as a reference to ratio the fluorescence of the membrane-potential indicator. The MitoTracker series of fluorescent probes each possess some residual potential dependence and a more robust solution is to transfect the cells with a mitochondrial-targeted fluorescent protein to provide a reference signal that can be ratioed with the potential indicator. It must be emphasized that this method, and indeed most of the published methods, actually just detect changes in mitochondrial membrane potential from an arbitrary starting value rather than reporting absolute values.
While the mitochondrial membrane potential determinations reviewed above provide information on the voltage component of the proton circuit, there are powerful homoeostatic mechanisms that tend to minimize changes in potential with energetic load. Thus respiration increases to match the increased utilization of the proton circuit, while in many cells processes that enhance energy demand activate metabolism in parallel to facilitate electron entry into the respiratory chain and counter any depolarization [11,12]. Changes in mitochondrial membrane potential are therefore rather small under normal metabolic conditions, although the sensitivity with which these may be detected is enhanced by the logarithmic nature of the Nernst equation. In contrast, the proton current quantitatively reports the utilization of the proton circuit by the multiple pathways summarized in Figure 1. Surprisingly it is relatively straightforward to quantify such an elusive component, even with mitochondria in intact cells, since there is a precise stoichiometry between electron flow down the respiratory chain to oxygen and the number of protons pumped out of the matrix by the three energy-conserving complexes, I, III and IV. The proton current can thus be quantified simply by measuring mitochondrial respiration in isolated preparations or in intact cells, correcting for any non-mitochondrial respiration in the latter by the residual respiration in the presence of electron transport chain inhibitors. The oxygen electrode chamber has been the central instrument for the study of isolated mitochondrial bioenergetics for the past 50 years, while recent technical advances have greatly facilitated the measurement of cell respiration in monolayer cultures [13,14].
As Mitchell summarized in his original Figure , there are a number of pathways by which protons re-enter the mitochondrial matrix. The predominant pathway is via the ATP synthase, where the high protonmotive force generated by the respiratory chain proton pumps forces the ATP-hydrolysing proton pump to run backwards in the direction of ATP synthesis. The ATP synthase itself is a complex assembly (reviewed in ), whereby a proton-driven rotor buried within the membrane (Fo) drives the rotation of an eccentric shaft (γ subunit) that induces a cycle of conformational changes in three pairs of alternating α and β subunits arranged radially around the shaft. ADP and phosphate (‘Pi’) bind to a binding site located between an α and β subunit. As the shaft rotates, tightly bound ATP is formed and is then released, this last step requiring the major input of free energy. The three binding sites function sequentially, resulting in the net synthesis of ATP.
In addition to the ATP synthase, all mitochondria possess an inherent proton leak across the inner membrane, the molecular identity of which is uncertain, although the ANT (adenine nucleotide transporter) has been implicated . A proton leak may seem energetically wasteful, and was initially considered to be an artefact of mitochondrial isolation; however, the endogenous proton leak is highly dependent upon the proton electrochemical gradient, being maximal in state 4 where there is no ATP synthesis, and much lower in state 3 conditions of rapid phosphorylation . In the absence of a state 4 leak electrons would be stalled within the electron transport chain with an increased risk of leaking from complex I and III to form the superoxide anion. On the other hand, in state 3 all of the proton current is required for ATP synthesis, and the high voltage-dependence of the leak means that protons are largely channeled efficiently through the ATP synthase with a greatly reduced dissipative leak (Figure 2).
In addition to the endogenous leak possessed by all mitochondria, a number of specific UCPs (uncoupling proteins) have been described. These will be discussed in detail in a subsequent chapter, but the best characterized is UCP1 which is expressed in thermogenic brown adipose tissue . In its normal physiological state UCP1 has a loosely bound purine nucleotide and is inactive. However, the initiation of lipolysis under the control of the sympathetic nervous system generates non-esterifed fatty acids that both activate the conductance of UCP1 and supply the substrate for the mitochondria allowing rapid thermogenic respiration to occur in the absence of stoichiometric ATP turnover . As will be discussed in a subsequent chapter, a number of homologues of UCP1 have been described with a wide tissue distribution, although their physiological functions are still a matter of controversy.
The extensive family of six transmembrane metabolite transporters in the inner mitochondrial membrane are almost all driven indirectly by the proton circuit. Thus the highly active phosphate transporter catalyses the electro-neutral exchange of the monovalent phosphate ion for a hydroxyl ion. The efflux of an hydroxyl ion is formally equivalent to the influx of a proton, and the result is that the monovalent phosphate anion accumulates in the matrix in response to the pH gradient. ATP synthesis in the intact cell involves the import of phosphate to the mitochondrial matrix, and the exchange of ADP for ATP via the ANT. This latter process is electrogenic, with the net uptake of one positive charge. Thus a single cycle of ADP and phosphate entry and ATP extrusion is driven by the full proton electrochemical gradient and utilizes one proton in addition to those entering through the ATP synthase. For this reason the free energy of ATP hydrolysis can be higher in the cytoplasm than in the mitochondrial matrix.
The metabolite transporters will be described in detail in a later chapter, but in terms of linkage to the proton circuit it is interesting that the dicarboxylate carrier responsible for the transport of malate equilibrates the anion in response to the square of the proton gradient, whereas the equilibrium gradient of citrate via the tricarboxylate transporter equals the third power of the pH gradient, being linked indirectly to the influx of two and three protons respectively.
The proton-translocating transhydrogenase
The nicotinamide nucleotide transhydrogenase utilizes the proton electrochemical potential to transfer electrons from NADPH to NADP+ . The likely stoichiometry is one proton per NADP+ reduced, and helps to ensure that the matrix NADP pool is maintained in a highly reduced state, in particular to provide electrons for the regeneration of reduced glutathione by glutathione reductase.
Monovalent cation transport
When Mitchell first proposed his chemiosmotic hypothesis he realized that a high membrane potential could result in the slow inward leakage of cations, and that over time this would result in mitochondrial swelling and ultimate rupture. To counter this he proposed and demonstrated the existence of a monovalent cation/proton antiport that would utilize the pH gradient to expel any potassium or sodium leaking into the matrix . As discussed below, the Na+/H+ exchanger plays an important role in Ca2+ cycling.
The plasma membranes of many cells possess a K+ channel that is inhibited by ATP. This KATP channel is important, for example, in linking enhanced mitochondrial ATP synthesis in pancreatic β-cells in response to increased glucose availability to plasma membrane depolarization and hence insulin secretion. It has been proposed that a similar channel exists in the inner mitochondrial membrane, controlling matrix volume, and that this plays an important role in protecting the heart during cardiac reperfusion after a heart attack. However the channel has only been inferred by the action of inhibitors that affect the plasma membrane channel, and the existence of a separate mitochondrial channel remains controversial.
Mitochondria avidly accumulate calcium when the concentration of the cations in their environment exceeds approx. 0.5 μM. Calcium transport is ultimately driven by the proton circuit, mitochondria possess a calcium uniporter that accumulates the divalent cation driven by the membrane potential. Thermodynamically a 150 mV membrane potential can maintain an equilibrium calcium gradient of 100000; however, this powerful sequestration pathway is balanced by a sodium–calcium exchanger that exports calcium in exchange for sodium, and the sodium–proton exchanger discussed above links the calcium and sodium circuits back to the primary proton circuit (Figure 1).
At first sight the cycling of calcium, sodium and protons across the inner membrane appears be symmetrical. However, the kinetics of the calcium uniporter is such that its activity increases as the 2.5th power of the cytoplasmic free calcium , whereas that of the sodium–calcium exchanger is controlled by the matrix free calcium concentration which, as will be discussed below, is limited and controlled by the formation of the calcium phosphate complex within the matrix . Typically uniporter activity exceeds that of the efflux pathway when the external free calcium concentration exceeds approx. 0.5 μM . In other words mitochondria become net accumulators of calcium above this so-called ‘set point’, and can serve in situ as temporary stores of the cation, in response to elevations in cytoplasmic Ca2+, releasing it to the matrix when the plasma membrane calcium transport pathways lower the cytoplasmic free calcium concentration below the set point. This mode is of particular significance in excitable tissues, such as muscle and brain, where large excursions in cytoplasmic calcium occur; in other tissues where the set-point is rarely exceeded, calcium cycling serves to regulate matrix free calcium as a means of controlling tricarboxylic acid cycle activity .
Phosphate plays an essential role in the accumulation of calcium by mitochondria. In the strict absence of the anion, net calcium accumulation is extremely limited, since the electrogenic accumulation of positively charged calcium is countered by the net extrusion of positively charged protons from the electron transport chain, with the result that a large pH gradient builds up across the membrane. Since the full proton electrochemical potential gradient cannot increase, the membrane potential falls in parallel with the increase in pH gradient, rapidly limiting calcium accumulation. If phosphate is present, the calcium-generated pH gradient drives the parallel uptake of the anion, in exchange for hydroxyl, which effectively limits the pH gradient. When inside the matrix the phosphate forms a readily dissociable complex with calcium that is stabilized by the alkaline matrix pH and buffers the matrix free calcium to a value that is largely independent of the total matrix calcium load .
The mitochondrial proton circuit, and its linkage to metabolite, cation and anion transport, remains essentially as proposed by Peter Mitchell in 1966 . The ability to monitor both the current and potential components of the circuit has provided the backbone for mitochondrial physiology both with the isolated organelle, and increasingly with intact cells, and provides a means of quantifying the rather nebulous concept of mitochondrial function, and more importantly dysfunction, in healthy cells and in the exploding fields of mitochondrial diseases, diabetes, amyotrophic lateral sclerosis, neurodegenerative disorders and normal aging where mitochondria are centre stage.
• Electron transport in mitochondria is coupled to ATP synthesis by a proton circuit whose proton electrochemical potential and current can be quantified.
• Mitochondria possess a voltage-regulated proton leak that limits production of free radicals. Some mitochondria possess in addition specific UCPs.
• The proton translocating transhydrogenase maintains the NADP pool in the matrix highly reduced.
• The distribution of metabolites across the mitochondrial membrane is affected by the proton gradient.
• Calcium cycles across the inner membrane and is linked to the proton circuit via an intermediate sodium circuit.
• Mitochondrial ‘function’ and ‘dysfunction’ can be determined by investigation of the state of the proton circuit.
- © The Authors Journal compilation © 2010 Biochemical Society