Nucleotide binding to human uncoupling protein-2 refolded from bacterial inclusion bodies

Experiments were performed to test the hypothesis that recombinant human uncoupling protein-2 (UCP2) ectopically expressed in bacterial inclusion bodies binds nucleotides in a manner identical with the nucleotide-inhibited uncoupling that is observed in kidney mitochondria. For this, sarkosyl-solubilized UCP2 inclusion bodies were treated with the polyoxyethylene ether detergent C12E9 and hydroxyapatite. Protein recovered from hydroxyapatite chromatography was approx. 90% pure UCP2, as judged by Coomassie Blue and silver staining of polyacrylamide gels. Using fluorescence resonance energy transfer, N-methylanthraniloyl-tagged purine nucleoside di- and tri-phosphates exhibited enhanced fluorescence with purified UCP2. Dissociation constants determined by least-squares non-linear regression indicated that the affinity of UCP2 for these fluorescently tagged nucleotides was 3-5 microM or perhaps an order of magnitude stronger, depending on the model used. Competition experiments with [8-14C]ATP demonstrated that UCP2 binds unmodified purine and pyrimidine nucleoside triphosphates with 2-5 microM affinity. Affinities for ADP and GDP were approx. 10-fold lower. These data indicate that: UCP2 (a) is at least partially refolded from sarkosyl-solubilized bacterial inclusion bodies by a two-step treatment with C12E9 detergent and hydroxyapatite; (b) binds purine and pyrimidine nucleoside triphosphates with low micromolar affinity; (c) binds GDP with the same affinity as GDP inhibits superoxide-stimulated uncoupling by kidney mitochondria; and (d) exhibits a different nucleotide preference than kidney mitochondria.