Rat heart mitochondria became permeabilized to sucrose when incubated with 100 nmol of Ca2+/mg of protein in the presence of Pi. Ca2+ chelation with EGTA restored impermeability to sucrose, which became entrapped in the matrix space. t-Butylhydroperoxide markedly promoted permeabilization in the presence of Ca2+ but not in its absence, and Ca2+-plus-t-butylhydroperoxide-induced permeabilization was reversed by EGTA. The data suggest that Ca2+ and oxidative stress synergistically promote the reversible opening of an inner membrane pore.

Evidence that the CaZ +-induced permeabilization of mitochondria is attributable to a reversible Ca2+-activated pore [A1 Nasser & Crompton (1986) Biochem. J. 239,[19][20][21][22][23][24][25][26][27][28][29] CoA, Mg2 ', AMP and also ATP, when account is taken of ADP arising by hydrolysis, are essentially ineffective.It is concluded that heart mitochondria do contain a pore whose permeation state is controlled over an approximate 1000-fold range by Ca2 ' and other factors including phosphate, oxidative stress and ADP. The possible involvement of the pore in reoxygenation-induced injury in heart is discussed.It has long been known that isolated mitochondria become leaky on accumulation of excess Ca2+, especially in the presence of Pi and the absence of addcd adenine nucleotides, but the molecular basis of this phenomenon has remained obscure (review [l]). Measurements of light-scattering changes as an index of matrix shrinkage lead Haworth and Hunter [2, 31 to suggest the presence of a channel regulated by CaZf and nucleotides, although in our view those studies do not discriminate against the changes in mitochondrial conformation induced by nucleotides (Discussion) or against membrane damage and general leakiness. Ca2 +-induced leakiness could arise most conceivably via the action of Ca2+-dependent phospholipase A2, some of which is located on the matrix face of the inner membrane [4 -61. Indeed, Pfeiffer and co-workers [7, 81 (and references therein) subsequently provided evidence for cyclic deacylation/reacylation of membrane phospholipids, and they proposed that Ca2+-induced displacement of the cycle towards deacylation is the root cause of the increased nonspecific permeability caused by CaZ +. According to this model, the various agents that promote or inhibit the phenomenon do so by affecting the component enzymes of the cycle, viz phospholipase A2, acyl-CoA synthetase and lysophospholipid acyltransferase.

The inner membrane of rat liver mitochondria contains a reversible Ca2+-dependent pore, opening of which is largely blocked by cyclosporin A. Analyses of [3H]cyclosporin binding to rat liver mitochondria demonstrate two classes of high-affinity binding site with capacities of < 5 pmol and approximately 60 pmol cyclosporin . mg mitochondrial protein-' in addition to partitioning into membrane phospholipids (0.03 pmol . mg mitochondrial protein . nM-'). Direct measurement [ ''C]sucrose entry into the matrix space indicates that cyclosporin A inhibits pore opening by interacting with the low-capacity sites. The same low-capacity sites (Kd cyclosporin, 8 nM) are possibly attributable to peptidylprolyl cis-trans-isomerase, although investigation of pore state interconversion from the rapid kinetics of [14C]sucrose entrapment in the matrix space does not indicate that cyclosporin-sensitive prolyl isomerization occurs at the actual step of pore opening/closure. It is suggested that the low-capacity cyclosporin-binding component may stabilize the open pore state; this is supported by the observations that Ca2 + decreases cyclosporin binding to this component and that cyclosporin brings about closure of the pre-opened pore. The implications for the possible number of functional pores in mitochondria are discussed.It is well established that tissue reperfusion injury after a period of ischaemia is associated with oxidative stress, with increased tissue Pi arising from adenine nucleotide breakdown, and with deranged Ca2+ metabolism leading to tissue Ca2+ overload; there is considerable evidence that these changes are critical in the pathogenesis of this form of injury [I -41. It has also become evident that mitochondrial energy transduction is susceptible to the very same factors. Studies with isolated mitochondria indicate that the specific lesion is an inner membrane pore which is normally closed but which opens under the synergistic influence of high intramitochondrial Ca2+, oxidative stress and Pi [5 -111, when cellular ATP is sufficiently depleted [12]. Pore opening is fully reversed on Ca2+ removal [7,13] and this has enabled the development of rapid-mixing (pulsed-flow) solute entrapment techniques to investigate pore properties from radiolabelled solute fluxes [lo]. These have shown, from the relative permeabilities to solutes of varying size, that the open pore is a large structure of about 2 nm internal diameter and that pore closure occurs rapidly (< 60 ms) when matrix Ca2+ is chelated [ll]. Since pore opening completely uncouples mitochondrial energy transduction and allows nonspecific permeation of low-molecular-mass solutes, the consequences of pore opening for cell viability would be severe. One hypothesis of reperfusion injury is that increase in resting cytosolic free Ca2+ after prolonged ischaemia leads to excessive mitochondrial Ca2 + uptake and pore opening on reperfusion (reoxygenation); the consequent impairment of cellular ATP synthesis prevents reestablishment of Ca2+ homeostasis so that irreversibl...

The permeability properties of a putative Ca2(+)-activated pore in heart mitochondria, of possible relevance to re-perfusion-induced injury, have been investigated by a pulsed-flow solute-entrapment technique. The relative permeabilities of [14C]mannitol, [14C]sucrose and arsenazo III are consistent with permeation via a pore of about 2.3 nm diameter. Ca2+ removal with EGTA induced pore closure, and the mitochondria became 'resealed'. The permeability of the unresealed mitochondria during resealing was markedly stimulated by 200 microM-ADP, and the relative permeabilities to solutes of different size were stimulated equally, indicating an increase in open-pore number, rather than an increase in pore dimensions. This is paradoxical, since ADP also stimulated the rate of resealing. The rate of EGTA-induced resealing was also stimulated by the Ca2+ ionophore A23187, which indicates that the rate of removal of matrix free Ca2+ is limiting for pore closure. An explanation for the paradox is suggested in which ADP facilitates pore interconversion between the closed and open states in permeabilized mitochondria, and pore closure in Ca2(+)-free mitochondria occurs much faster than previously thought.

By using an assay based on the precipitation of intact 14C-labelled substrate, an activity has been located in the mitochondrial fraction of rat liver which selectively hydrolyses the 3′ ester link in the fairly recently discovered oligomeric tetraphosphate derivative of ATP and glyceric acid for which the structure 3-phospho[glyceroyl-gamma-triphospho-5′-adenosine-3′-3-phospho]n-glyceroyl- gamma-triphospho-5′-adenosine has been proposed [Hutchinson, Morris & Mowbray (1986) Biochem. J. 234, 623-627]. This enzyme activity (Mr 85,000) has been purified approx. 30-fold from washed mitochondria by (NH4)2SO4 precipitation and f.p.l.c. The apparent Km for substrate (adenosine equivalents) is around 35 microM. The recovery of total activity is about 20%, and this, allied to the relatively low Vmax. found in contrast with the rapid turnover of oligomer seen in post-ischaemic tissues, suggests that some activating factors have been lost in purification. Percoll-gradient studies confirm that the activity is mitochondrial and not lysosomal or endoplasmic-reticular. The activity is latent in intact mitochondria; it is not, however, associated with intact inner-membrane vesicles but released during their preparation, implying an intermembrane-space location. The product of the enzyme is proposed to be the monomeric unit 3-phosphoglyceroyl-gamma-triphospho-5′-adenosine, from which digestion with snake-venom phosphodiesterase releases ADP.