Dopamine transporter ligand binding domains. Structural and functional properties revealed by limited proteolysis

Dopamine transporters (DATs) are members of the Na+- and Cl--dependent neurotransmitter and amino acid transporter family predicted by hydrophobicity analysis to have 12 transmembrane-spanning helices. The structure of DAT was studied using the photoaffinity compounds [125I]1-[2-(diphenylmethoxy)-ethyl]-4-[2-(4-azido-3-iodophenyl) ethyl] piperazine ([125I]DEEP), a 1-(2-diphenylmethoxy)-ethyl-4-(3-phenyl propyl)piperazine (GBR analog), and [125I]-3beta-(p-chlorophenyl)tropane-2beta-carboxylic acid, 4'-azido-3'-iodophenylethyl ester ([125I]RTI 82), a cocaine analog, which had been shown in a previous study to become incorporated into different regions of the DAT primary sequence. The proximity of the photolabeled binding sites to integral membrane structures was investigated by subjecting photolabeled membrane suspensions to limited proteolysis with trypsin and separately analyzing the resulting membranes and supernatants for the presence of photolabeled DAT fragments. Trypsin treatment of [125I] DEEP-labeled membranes generated labeled 45- and 14-kDa DAT fragments that immunoprecipitated with an epitope-specific antiserum generated against amino acids 42-59 near the first putative transmembrane domain, whereas [125I]RTI 82 was found in 32- and 16-kDa tryptic fragments that precipitated with an antiserum directed against a sequence near transmembrane domain 4 (amino acids 225-238). All of the photolabeled fragments were recovered in the protease-treated membranes, indicating that they possess integral membrane structures that prevent their release from the membrane as soluble forms. The size of the two smallest fragments in conjunction with their retention in the membrane suggests that incorporation of the photoaffinity ligands occurs in or near membrane spanning regions and delineates the maximum possible distance between the transmembrane structures, incorporated photolabel, and antibody epitopes. Carbohydrate analysis of the fragments identified sialic acids and N-linked oligosaccharides exclusively on the 45-kDa [125I]DEEP-labeled fragment, which, based on size, would be expected to contain four consensus glycosylation sites between putative transmembrane domains 3 and 4. Photoaffinity labeling after trypsin treatment of membranes showed that the larger but not the smaller fragments retain binding capacity, as the 45- and 32-kDa fragments were capable of becoming photolabeled. Binding of photoaffinity ligands at these fragments was displaced with the same pharmacology as that of intact DATs. These results verify numerous aspects of DAT structure and topology heretofore only predicted from theoretical considerations and extend our knowledge of DAT structure-function properties.