news & views MITOCHONDRIA Mitochondrial H + permeability through the ADP/ATP carrier Mitochondrial H + leak, which is responsible for basal respiration, appears to be a transport process mediated by the ADP/ATP carrier and regulated by fatty acids and adenine nucleotides. Paolo Bernardi T he mitochondrial inner membrane has a low but measurable permeability to H + , which is commonly referred to as ‘H + leak’. As the term itself suggests, the H + leak was widely assumed to be the result of passive H + permeation through the lipid bilayer, and ‘basal’ respiration (i.e., oxygen consumption not coupled to ATP synthesis) was assumed to compensate for this backflow of H + . In a recent study in Nature, Bertholet et al. demonstrate that a sizeable fraction of the H + leak may actually take place through the ADP/ATP carrier (AAC) and that H + transport by the AAC requires fatty acids 1 , just like uncoupling protein 1 (UCP1) 2,3 . Consequently, H + backflow could be regulated in all cells, thus allowing for control of thermogenesis in tissues not endowed with UCP1 (Fig. 1). AAC belongs to the SLC25 superfamily of mitochondrial solute carriers and is structurally related to UCP1 (ref. 4 ). UCP1 is an H + channel expressed in brown fat, where it short-circuits the H + gradient, thus maximally stimulating respiration to produce heat 5,6 , whereas the established physiological role of the AAC is to exchange ADP for ATP across the inner mitochondrial membrane, with net translocation of one negative charge 7 . During respiration, ADP enters the mitochondrial matrix while ATP is released, at the cost of one charge per nucleotide exchanged, thus adding to the eight H + necessary in mammals for the synthesis of three ATP molecules through one 360° rotation of the ATP synthase 8 . Fatty acids have long been known to increase respiration with a matching decrease in adenine nucleotide translocation 9 , and increased respiration is inhibited by ADP or AAC inhibitors but not by GDP 10 , which is known to inhibit UCP1 (refs. 5,6,11 ). On the basis of these findings, Skulachev and co-workers have suggested that H + transport is mediated by the AAC either through an allosteric effect of the fatty acid or through AAC-mediated transport of the fatty acyl anion coupled to passive diffusion of the protonated fatty acid through the lipid bilayer 10 . The former hypothesis appears to be correct, because the elegant electrophysiological data of Bertholet et al. show that the AAC acts as a H + channel, and fatty acids play an essential role in H + translocation as cofactors rather than as the transported species 1 . This finding contrasts with those for the UCP1 pathway, in which H + is transported by fatty acid anion shuttling within UCP1 itself 3 (Fig. 1). In this new study, the authors performed patch-clamp recordings of mitoplasts— mitochondria from which the outer membrane has been removed—from tissues that do not express UCP1. They observed that arachidonic, palmitic and lauric acid activated currents that were inhibited by selective AAC inhibitors 1 . Of note, currents were still observed in mitoplasts lacking UCP2 and UCP3, structural analogues of UCP1 that are not considered to be bona fide uncoupling proteins 12 . A convincing demonstration that currents are mediated by AAC came from patch-clamp experiments in mitoplasts from tissues in which the AAC1 or AAC2 isoforms were absent, a condition that matched the lack of fatty- acid-induced currents 1 . This set of findings is consistent with the lower basal oxygen consumption rate of cells with AAC1/ AAC2 double knockout, which could not be stimulated by palmitate yet reached the same maximal respiration after treatment with a mitochondrial uncoupling agent, thus indicating normal respiratory-chain function 1 . A lack of respiratory stimulation by fatty acids was also observed in mitochondria from the hearts of AAC1-null mice, but in that case, basal respiration was actually higher than that in mitochondria from wild-type mice 1 . The authors explain this unexpected finding by higher levels of oxidative stress and possibly increased opening of the permeability transition pore, which is also activated by fatty acids 13 ; however it is difficult to see why a similar condition should not have been encountered in the AAC1/AAC2-null cells. However, it is possible that the discrepancy is only apparent, because the difference in basal respiration disappears if the data Intermembrane space Matrix LCFA AAC H + UCP1 H + H + H + – ADP GDP ADP Fig. 1 | Mechanism of H + transport by the adenine nucleotide carrier and uncoupling protein 1. In spite of the common requirement for long-chain fatty acids (LCFA), the mechanism of H + transport by the AAC and UCP1 differs and best fits the H + channel mode for AAC (left) and the H + buffering mode for UCP1 (right). H + conductance is partially inhibited by ADP in both cases, whereas GDP selectively blocks UCP1. NATURE METABOLISM | www.nature.com/natmetab