Pentameric ligand-gated ion channels from the Cys-loop family mediate fast chemo-electrical transduction, but the mechanisms of ion permeation and gating of these membrane proteins remain elusive. Here we present the X-ray structure at 2.9 A resolution of the bacterial Gloeobacter violaceus pentameric ligand-gated ion channel homologue (GLIC) at pH 4.6 in an apparently open conformation. This cationic channel is known to be permanently activated by protons. The structure is arranged as a funnel-shaped transmembrane pore widely open on the outer side and lined by hydrophobic residues. On the inner side, a 5 A constriction matches with rings of hydrophilic residues that are likely to contribute to the ionic selectivity. Structural comparison with ELIC, a bacterial homologue from Erwinia chrysanthemi solved in a presumed closed conformation, shows a wider pore where the narrow hydrophobic constriction found in ELIC is removed. Comparative analysis of GLIC and ELIC reveals, in concert, a rotation of each extracellular beta-sandwich domain as a rigid body, interface rearrangements, and a reorganization of the transmembrane domain, involving a tilt of the M2 and M3 alpha-helices away from the pore axis. These data are consistent with a model of pore opening based on both quaternary twist and tertiary deformation.

Ligand-gated ion channels (LGICs) mediate excitatory and inhibitory transmission in the nervous system. Among them, the pentameric or 'Cys-loop' receptors (pLGICs) compose a family that until recently was found in only eukaryotes. Yet a recent genome search identified putative homologues of these proteins in several bacterial species. Here we report the cloning, expression and functional identification of one of these putative homologues from the cyanobacterium Gloeobacter violaceus. It was expressed as a homo-oligomer in HEK 293 cells and Xenopus oocytes, generating a transmembrane cationic channel that is opened by extracellular protons and shows slow kinetics of activation, no desensitization and a single channel conductance of 8 pS. Electron microscopy and cross-linking experiments of the protein fused to the maltose-binding protein and expressed in Escherichia coli are consistent with a homo-pentameric organization. Sequence comparison shows that it possesses a compact structure, with the absence of the amino-terminal helix, the canonical disulphide bridge and the large cytoplasmic domain found in eukaryotic pLGICs. Therefore it embodies a minimal structure required for signal transduction. These data establish the prokaryotic origin of the family. Because Gloeobacter violaceus carries out photosynthesis and proton transport at the cytoplasmic membrane, this new proton-gated ion channel might contribute to adaptation to pH change.

Using fluorescence lifetime microspectroscopy and imaging techniques, we have studied the fluorescence of cyan fluorescent protein (CFP) transiently expressed in HEK-293 cells, in the presence or absence of its fluorescence resonance energy transfer (FRET) partner, yellow fluorescent protein (YFP). When the two proteins are attached through a 27-amino-acid linker, a 33 % average efficiency of intramolecular energy transfer is accurately determined inside the cell. Additionally, we observe a systematic quenching of the CFP fluorescence with increasing levels of protein expression. This quenching cannot be accounted for by formation of the previously described dimer of GFP-related proteins, since its magnitude is unchanged when the fluorescent proteins carry the mutation A206K shown to dissociate this dimer in vitro. Even when the intracellular protein concentration largely exceeds the in vitro dissociation constant of the dimer, self-association remains undetectable, either between free proteins or intramolecularly within the CFP-YFP construct. Instead, the detailed concentration effects are satisfactorily accounted for by a model of intermolecular, concentration-dependent energy transfer, arising from molecular proximity and crowding. In the case of CFP alone, we suggest that self-quenching could result from a pseudo-homo FRET mechanism between different, spectrally shifted emissive forms of the protein. These phenomena require careful consideration in intracellular FRET studies.

No abstract