An Addendum to the Chemiosmotic Theory of Mitochondrial Activity: The Role of RNA as a Proton Sink

While readers are referred to scholarly reviews detailing respiratory control mechanisms [45,46,47], the overview provided here is a prelude to the discussion that follows. The first mechanism of respiratory control emerged from the observation that addition of ADP stimulates oxygen uptake of isolated mitochondria (i.e., state 3 respiration), leading to a reduction in Δp [48,49]. This phase is followed by inhibition of respiration due to transformation of ADP into ATP (i.e., state 4 respiration) and a concomitant increase in Δp [48,49]. According to classic chemiosmotic theory, this happens because a high Δp inhibits the activity of proton pumps. However, several conflicting lines of evidence [47], including an unchanged mitochondrial membrane potential according to the rate of cell respiration [50], hinted at an alternative molecular basis for the first regulatory mechanism of respiratory control. In the revised chemiosmotic theory, consumption of Pi to generate ATP will induce a switch from exonuclease activity of PNPase to a polymerising activity once the concentration of Pi falls below a threshold level (see Equations (1) and (6)). Therefore, the available ADP will be diverted to the synthesis of poly-(A) homopolymers in IMS rather than generation of ATP in matrix.

The second mechanism of respiratory control emerged from studies by Kadenbach et al., who discovered allosteric inhibition of cytochrome c oxidase at a high ATP/ADP ratio [45,46]. ATP exerts this effect by binding to the matrix domain of subunit IV of cytochrome c oxidase [51]. The allosteric inhibition of cytochrome c oxidase by ATP is switched on by cAMP-dependent phosphorylation of the complex and switched off by Ca2+-activated dephosphorylation of the protein [52]. In a dephosphorylated excited state, the activity of the complex increases by five- to ten-fold [53], and generation of reactive oxygen species (ROS) by the electron transport chain occurs at a higher rate [45]. The revised chemiosmotic theory interfaces with and complements the second mechanism of respiratory control at multiple levels. At an effector level, amplified ROS generation is expected to trigger oxidation of RNA. The higher propensity of guanine for oxidisation compared to adenine (≈3 times higher propensity [54]), combined with a higher affinity of PNPase to oxidised RNA [55], means that ROS could potentially accelerate the exonuclease activity of the enzyme, thus invoking a high rate of ATP synthesis. This enhanced activity of cytochrome c oxidase comes at a cost of a somewhat reduced efficiency, as a fraction of ADPs will be oxidised and excluded from the pool of usable dinucleotides. The enhanced activity of cytochrome c oxidase upon availability of calcium is important to complement other mitochondrial impacts of this ion. Calcium facilitates dephosphorylation of phosphorylated pyruvate dehydrogenase, thereby activating the enzyme [56]. Calcium also stimulates the entire oxidative phosphorylation cascade [57] within a specific range of ionic concentration [58].

Therefore, it can be argued that the oxidation of RNA within IMS via ROS, which occurs within seconds, primes mitochondria for amplified activity of the electron transport chain in anticipation of an energised citric acid cycle. On the other hand, the reduced activity of PNPase in a ROSlow milieu (i.e., phosphorylated cytochrome c oxidase and a high ATP/ADP ratio) will reduce the measurable component of ΔpRNA (i.e., Δpfree as per Equation (8a)), manifesting as a reduced membrane potential in this relaxed state. Based on this discussion, it can be said that the revised chemiosmotic theory complements the first and the second respiratory control mechanisms while providing explanations for the role of unexplained features such as amplified ROS production.

An aspect of the proposal that requires further experimentation is potential association of RNA with the lipid bilayer of mitochondrial inner membrane. While there is no direct evidence for such an association, certain inferences can be made based on published reports. Within mitochondrial cristae, cardiolipin stabilises respiratory chain supercomplexes [59] and the lipid bilayer [60]. Recent studies revealed that cardiolipin and phosphatidylethanolamine segregate to the negatively curved monolayer leaflet facing the crista lumen, while the opposing, positively curved, matrix-facing monolayer leaflet contains, predominantly, phosphatidylcholine [61]. It has been suggested that cardiolipin headgroups may function as a proton trap, increasing the concentration of protons in proximity of ATP synthase [62]. However, anionic lipid headgroups of cardiolipin could also engage in anti-electrostatic hydrogen bonding [63] with other anionic groups (e.g., phosphate) [64,65]. Accordingly, potential association of RNA with cardiolipin by forming anti-electrostatic hydrogen bonds provides an alternative mechanism by which cardiolipin could enrich the proton in vicinity of ATP synthase. Considering the evidence for functional association of PNPase and ATP synthase [15,66], an alternative hypothesis is that PNPase associates with ATP synthase directly or indirectly and thus localises the RNA into vicinity of ATP synthase.

Experimental validation of the revised chemiosmotic hypothesis requires experiments in which critical parameters of the model are rigorously controlled. Aside from elimination of any RNase activity, the critical parameters of the proposed model include the concentrations of Pi and ADP within IMS. Also of importance, the incubation time is determined by the concentration of Pi in IMS (see Equations (4) and (6)). The concentration of free ADP within IMS is negligible compared to that supplied exogenously. Hence, controlling the concentration of Pi appears to be pivotal for experimental validation of the proposed model.

In summary, a revised framework for chemiosmotic production of ATP is proposed herein, which entails an intermediary step of storing the protons generated via ETC in an RNA-based proton sink residing in IMS. The proposed proton sink not only sequesters protons transiently but also coordinates the release of trapped protons with the availability of ADP and inorganic Pi, the key substrates of ATP synthase. Regulation of the proton sink is achieved via bimodal activity of PNPase, which can operate both as a polymerase and an exonuclease by receiving input from inorganic phosphate.