Structural and functional analysis of Apo-states of the model Cys-loop receptor GluCl from Caenorhabditis elegans
Biochemistry
Biophysics
Molecular Biology and Physiology of Neurons and Glial Cells
Final Report Abstract
Cys‐loop receptors are a major family of pentameric ligand‐gated ion channels that mediate fast neurotransmission at both excitatory and inhibitory chemical synapses. In humans they are involved in various neurological diseases and serve as targets for clinical drugs and substances of drug abuse. However, so far the details of their activation and gating mechanism have been poorly understood as only single structures of individual family members were available. In the course of this project two new structures of the anion‐selective model Cys‐loop receptor GluCl from Caenorhabditis elegans, were solved in the absence of the allosteric agonist ivermectin that was previously used to solve a structure of the receptor in an open conformation. The new structures were obtained by crystallizing the protein in high concentrations of the lipid POPC and the detergent dodecylmaltoside. The first structure at 3.6 Å resolution clearly exhibited an obstructed ion channel pore with the tightest constriction found at the position of the 9’ leucines whose side chains pointed towards the center of pore. Based on the comparison of the structures a mechanism for the channel activation and gating was proposed that involved mainly rigid‐body motions of the extracellular ligand binding domain against the transmembrane domain. Ivermectin inserts between the transmembrane helices, pushing the pore lining M2 helix sideways and slightly up to open the channel pore. Binding of the allosteric agonist also leads to a closing of the extracellular agonist binding site explaining the higher affinity for glutamate in the presence of ivermectin. At the same time the extracellular domain tilts inward and down, locking the pore in the open conformation. Further research will be required to elucidate how the signal of ligand binding in the extracellular domain is transmitted to the transmembrane domain. Despite evidence from the new structure the location of the channel gate still remains somewhat controversial. Mutation of the 9’ leucines to bulkier phenylalanine did not show the expected effect of a permanently blocked channel in Xenopus laevis oocytes. And mutation to smaller alanines did not lead to a constitutively open channel. Instead, both mutants behaved like wild‐type GluCl. The second structure was solved at 3.2 Å resolution after prolonged incubation of the receptor in the presence of lipid and detergent. In this case the ion channel pore was more open to allow passage of ions, yet in a different conformation compared to the ivermectin‐bound structure. Most interestingly it showed electron densities resembling lipid molecules inserted between the transmembrane helices of adjacent subunits, partially overlapping with the ivermectin binding site. Overall the POPC‐bound was in an intermediate conformation between the apo‐closed and ivermectin‐bound structures. This was particularly intriguing at the glutamate binding site. It was therefore concluded that lipids are natural allosteric agonists that “prime” GluCl for activation by glutamate. Indeed POPS, but surprisingly not POPC, as well as cholesteryl hemisuccinate (CHS) competed with ivermectin binding with affinities in the micromolar (POPS) and low millimolar (CHS) range, respectively. Furthermore, binding of glutamate was also enhanced from virtually no binding in the absence to low micromolar affinities in the presence of lipids. Unfortunately, further characterization of the functional role of lipids on the activity of GluCl by reconstitution into liposomes was hampered by the low efficiency of the incorporation. Also the presence of both POPS and cholesterol was required so that no control vesicles without these two lipids could be prepared. Structures of GluCl in the presence of POPS or cholesterol/CHS could provide more insights into the role of these lipids on the receptor function, but crystals obtained so far did not diffract beyond ~5 Å. An explanation could be that these lipids lead to more flexibility of the receptor preventing the formation of well‐ordered crystals. Additionally, a mammalian expression system for GluCl was established that provides much higher yields of more homogeneous protein compared to the insect cell system used before. Surprisingly this required the re‐insertion of a glycosylation site that was expendable in the insect cells. Taken together, this project provided important new insights into the gating mechanism and the role of lipids in the function of GluCl that might be applicable to Cys‐loop receptors in general. Some controversies and open questions remain that will need to be addressed by different approaches that are currently under development.
Publications
- (2014). Screening and large‐scale expression of membrane proteins in mammalian cells for structural studies. Nat Protoc 9: 2574‐2585
Goehring, A., Lee, C.‐H., Wang, K.W., Michel, J.C., Claxton, D.P., Baconguis, I., Althoff, T., Fischer. S., Garcia, K.C., Gouaux, E.
(See online at https://doi.org/10.1038/nprot.2014.173) - (2014). X‐ray structures of GluCl in apo states reveal a gating mechanism of Cys‐loop receptors. Nature 512: 333‐337
Althoff, T., Hibbs, R.E., Banerjee S., Gouaux, E.
(See online at https://doi.org/10.1038/nature13669)
