Reply to Bernardi: The mitochondrial permeability transition pore and the ATP synthase

The mitochondrial permeability transition pore (PTP) is a physiological phenomenon lacking a molecular basis. The phenomenon, described over 40 y ago, is that in response to elevated levels of Ca²⁺ ions in the mitochondrial matrix a nonspecific channel opens, water enters the mitochondria, their cristae swell, their membranes rupture, terminating ATP synthesis, and cell death by necrosis ensues (1). Pore opening can be inhibited by the binding of cyclosporin A to cyclophilin D, a prolyl cis–trans isomerase in the mitochondrial matrix that interacts with the pore without necessarily being a component. The open pore allows hydrophilic molecules up to a molecular weight of 1,500 Da to pass (2), providing an estimate of pore diameter of 20 to 30 Å. These are the defining features of the pore. In 2013, Bernardi and coworkers (3) proposed that the pore is associated with the dimeric ATP synthase complex, which forms rows of dimers along the edges of the mitochondrial cristae. We adopted two experimental approaches to test this proposal. One approach has been to remove subunits of the ATP synthase by disrupting the corresponding genes by CRISPR-Cas9 in human cells and then to examine the effects of removal of individual subunits on the structure and assembly of the ATP synthase, and on the defining properties of the pore. The result of all such interventions, including recent experiments where the mitochondria lack an ATP synthase assemblage, plus experiments with ρ⁰ cells which are devoid of mitochondrial DNA, and therefore lack subunits ATP6 and ATP8, is that the defining properties of the pore survive unscathed, including the passage through the pore of polyethylene glycols (PEGs) up to PEG1000, monitored by light absorbed by the mitochondria (4–6). Bernardi has agreed that ρ⁰ cells, where the major dimerizing component ATP6 is absent (7), still have a functional PTP (8), but, despite the retention of the defining properties of the pore, his letter criticizes the recent experiments on the basis that the induction and rate of mitochondrial swelling is “slower” in the cells where subunits of ATP synthase have been removed (9), although the diameter of the pore is unchanged. Two plausible explanations have been provided already. First, the “slower” rates of induction and swelling (more precisely, change in light absorbance) are probably related to changes in the structures of the cristae and mitochondrial networks accompanying the removal of membrane subunits involved in dimerizing ATP synthase, as observed in human and yeast mitochondria (see references cited in ref. 6). Second, because the ATP synthase is defective in the mutant cells, metabolic differences arise, changing NAD(P)H/NAD(P)⁺ and ADP/ATP ratios, thereby generating a less favorable environment for induction of the PTP (10, 11). Collectively, our cell-based experiments provide overwhelming evidence that the PTP is not associated with the ATP synthase. The second approach being followed is to determine a high-resolution structure of the dimeric mammalian ATP synthase complex. Once achieved, this structure should show whether or not the membrane domain of the ATP synthase can form an aqueous channel 20 to 30 Å in diameter.