Mitochondrial membrane uncoupling protein 3 (UCP3) is not only expressed in skeletal muscle and heart, but also in brown adipose tissue (BAT) alongside UCP1, which facilitates a proton leak to support non-shivering thermogenesis. In contrast to UCP1, the transport function and molecular mechanism of UCP3 regulation are poorly investigated, although it is generally agreed upon that UCP3, analogous to UCP1, transports protons, is activated by free fatty acids (FFAs) and is inhibited by purine nucleotides (PNs). Because the presence of two similar uncoupling proteins in BAT is surprising, we hypothesized that UCP1 and UCP3 are differently regulated, which may lead to differences in their functions. By combining atomic force microscopy and electrophysiological measurements of recombinant proteins reconstituted in planar bilayer membranes, we compared the level of protein activity with the bond lifetimes between UCPs and PNs. Our data revealed that, in contrast to UCP1, UCP3 can be fully inhibited by all PNs and IC50 increases with a decrease in PN-phosphorylation. Experiments with mutant proteins demonstrated that the conserved arginines in the PN-binding pocket are involved in the inhibition of UCP1 and UCP3 to different extents. Fatty acids compete with all PNs bound to UCP1, but only with ATP bound to UCP3. We identified phosphate as a novel inhibitor of UCP3 and UCP1, which acts independently of PNs. The differences in molecular mechanisms of the inhibition between the highly homologous transporters UCP1 and UCP3 indicate that UCP3 has adapted to fulfill a different role and possibly another transport function in BAT.

Adenine nucleotide translocase (ANT) is a well-known mitochondrial exchanger of ATP against ADP. In contrast, few studies have shown that ANT also mediates proton transport across the inner mitochondrial membrane. The results of these studies are controversial and lead to different hypotheses about molecular transport mechanisms. We hypothesized that the H+-transport mediated by ANT and uncoupling proteins (UCP) has a similar regulation pattern and can be explained by the fatty acid cycling concept. The reconstitution of purified recombinant ANT1 in the planar lipid bilayers allowed us to measure the membrane current after the direct application of transmembrane potential ΔΨ, which would correspond to the mitochondrial states III and IV. Experimental results reveal that ANT1 does not contribute to a basal proton leak. Instead, it mediates H+ transport only in the presence of long-chain fatty acids (FA), as already known for UCPs. It depends on FA chain length and saturation, implying that FA’s transport is confined to the lipid-protein interface. Purine nucleotides with the preference for ATP and ADP inhibited H+ transport. Specific inhibitors of ATP/ADP transport, carboxyatractyloside or bongkrekic acid, also decreased proton transport. The H+ turnover number was calculated based on ANT1 concentration determined by fluorescence correlation spectroscopy and is equal to 14.6 ± 2.5 s−1. Molecular dynamic simulations revealed a large positively charged area at the protein/lipid interface that might facilitate FA anion’s transport across the membrane. ANT’s dual function—ADP/ATP and H+ transport in the presence of FA—may be important for the regulation of mitochondrial membrane potential and thus for potential-dependent processes in mitochondria. Moreover, the expansion of proton-transport modulating drug targets to ANT1 may improve the therapy of obesity, cancer, steatosis, cardiovascular and neurodegenerative diseases.

2,4-Dinitrophenol (DNP) is a classic uncoupler of oxidative phosphorylation in mitochondria which is still used in “diet pills”, despite its high toxicity and lack of antidotes. DNP increases the proton current through pure lipid membranes, similar to other chemical uncouplers. However, the molecular mechanism of its action in the mitochondria is far from being understood. The sensitivity of DNP’s uncoupling action in mitochondria to carboxyatractyloside, a specific inhibitor of adenine nucleotide translocase (ANT), suggests the involvement of ANT and probably other mitochondrial proton-transporting proteins in the DNP’s protonophoric activity. To test this hypothesis, we investigated the contribution of recombinant ANT1 and the uncoupling proteins UCP1-UCP3 to DNP-mediated proton leakage using the well-defined model of planar bilayer lipid membranes. All four proteins significantly enhanced the protonophoric effect of DNP. Notably, only long-chain free fatty acids were previously shown to be co-factors of UCPs and ANT1. Using site-directed mutagenesis and molecular dynamics simulations, we showed that arginine 79 of ANT1 is crucial for the DNP-mediated increase of membrane conductance, implying that this amino acid participates in DNP binding to ANT1.

Several mitochondrial proteins, such as adenine nucleotide translocase (ANT), aspartate/glutamate carrier, dicarboxylate carrier, and uncoupling proteins 2 and 3, are suggested to have dual transport functions. While the transport of charge (protons and anions) is characterized by an alteration in membrane conductance, investigating substrate transport is challenging. Currently, mainly radioactively labeled substrates are used, which are very expensive and require stringent precautions during their preparation and use. We present and evaluate a fluorescence-based method using Magnesium Green (MgGrTM), a Mg2+-sensitive dye suitable for measurement in liposomes. Given the different binding affinities of Mg2+ for ATP and ADP, changes in their concentrations can be detected. We obtained an ADP/ATP exchange rate of 3.49 ± 0.41 mmol/min/g of recombinant ANT1 reconstituted into unilamellar liposomes, which is comparable to values measured in mitochondria and proteoliposomes using a radioactivity assay. ADP/ATP exchange calculated from MgGrTM fluorescence solely depends on the ANT1 content in liposomes and is inhibited by the ANT-specific inhibitors, bongkrekic acid and carboxyatractyloside. The application of MgGrTM to investigate ADP/ATP exchange rates contributes to our understanding of ANT function in mitochondria and paves the way for the design of other substrate transport assays.

Molecular dynamics (MD) simulations of uncoupling proteins (UCP), a class of transmembrane proteins relevant for proton transport across inner mitochondrial membranes, represent a complicated task due to the lack of available structural data. In this work, we use a combination of homology modelling and subsequent microsecond molecular dynamics simulations of UCP2 in the DOPC phospholipid bilayer, starting from the structure of the mitochondrial ATP/ADP carrier (ANT) as a template. We show that this protocol leads to a structure that is impermeable to water, in contrast to MD simulations of UCP2 structures based on the experimental NMR structure. We also show that ATP binding in the UCP2 cavity is tight in the homology modelled structure of UCP2 in agreement with experimental observations. Finally, we corroborate our results with conductance measurements in model membranes, which further suggest that the UCP2 structure modeled from ANT protein possesses additional key functional elements, such as a fatty acid-binding site at the R60 region of the protein, directly related to the proton transport mechanism across inner mitochondrial membranes.