Serotonin transporter-ibogaine complexes illuminate mechanisms of inhibition and transport

This study investigated serotonin transporter (SERT) complexes with ibogaine to illustrate structure-based mechanisms for transport in serotonin transporter (SERT). The investigation reported that cryo-electron microscopy structures of SERT–ibogaine complexes captured in outward-open, closed and inward-open conformations with ibogaine binding to the central binding site, and the closing extracellular gate with movements of TMs 1b and 6a. The intracellular gate opening had a hinge-like movement of TM1a and the partial unwinding of TM5 that together built a permeation pathway enabling substrate and ion diffusion to the cytoplasm. These structures show the structural rearrangements which occur from the outward-open to inward-open conformations, and give an important insight into the working mechanism of neurotransmitter transport and ibogaine inhibition.

Abstract

“The serotonin transporter (SERT) regulates neurotransmitter homeostasis through the sodium- and chloride-dependent recycling of serotonin into presynaptic neurons. Major depression and anxiety disorders are treated using selective serotonin reuptake inhibitors—small molecules that competitively block substrate binding and thereby prolong neurotransmitter action. The dopamine and noradrenaline transporters, together with SERT, are members of the neurotransmitter sodium symporter (NSS) family. The transport activities of NSSs can be inhibited or modulated by cocaine and amphetamines, and genetic variants of NSSs are associated with several neuropsychiatric disorders including attention deficit hyperactivity disorder, autism and bipolar disorder. Studies of bacterial NSS homologues—including LeuT—have shown how their transmembrane helices (TMs) undergo conformational changes during the transport cycle, exposing a central binding site to either side of the membrane. However, the conformational changes associated with transport in NSSs remain unknown. To elucidate structure-based mechanisms for transport in SERT we investigated its complexes with ibogaine, a hallucinogenic natural product with psychoactive and anti-addictive properties. Notably, ibogaine is a non-competitive inhibitor of transport but displays competitive binding towards selective serotonin reuptake inhibitors. Here we report cryo-electron microscopy structures of SERT–ibogaine complexes captured in outward-open, occluded and inward-open conformations. Ibogaine binds to the central binding site, and closure of the extracellular gate largely involves movements of TMs 1b and 6a. Opening of the intracellular gate involves a hinge-like movement of TM1a and the partial unwinding of TM5, which together create a permeation pathway that enables substrate and ion diffusion to the cytoplasm. These structures define the structural rearrangements that occur from the outward-open to inward-open conformations, and provide insight into the mechanism of neurotransmitter transport and ibogaine inhibition.“

Authors: Jonathan A. Coleman, Dongxue Yang, Zhiyu Zhao, Po-Chao Wen, Craig Yoshioka, Emad Tajkhorshid & Eric Gouaux

Summary

The serotonin transporter (SERT) regulates neurotransmitter homeostasis through the sodium- and chloride-dependent recycling of serotonin into presynaptic neurons. The SERT, dopamine and noradrenaline transporters are members of the neurotransmitter sodium symporter (NSS) family. We investigated the structure of SERT complexes with ibogaine, a hallucinogenic natural product with psychoactive and anti-addictive properties. The structures provide insight into the mechanism of neurotransmitter transport and ibogaine inhibition.

We used antibody fragments to facilitate cryo-EM reconstruction of SERT, a monomeric membrane protein of approximately 70 kDa. The inhibition of serotonin uptake by ibogaine was determined for the ts2-active and N72/C13 SERT variants, and the consequences of antibody binding were investigated.

We carried out saturation binding experiments with [3H]ibogaine in NaCl to ts2-active SERT without and with the 15B8 Fab, and found that ibogaine binds to the transporter more tightly in the presence of KCl or N-methyl-d-glucamine hydrochloride than in NaCl-containing buffers.

We examined a mutant containing a serine-to-cysteine substitution at residue 277 (S277C). The mutant is more reactive to methanethiosulfonate reagents when bound to ibogaine than when bound to inhibitors that stabilize the outward-open conformation.

We used single-particle cryo-EM to study the structure of the SERT ibogaine-binding site and to determine how ibogaine binding influences the conformation of the transporter. We discovered that the outward-open conformation of SERT is well-fitted by the X-ray structure of the ts2-inactive SERT – 15B8 Fab – 8B6 scFv complex.

We used Fabs that preserve serotonin-uptake activity to elucidate the structure of the SERT – ibogaine complex in NaCl. We found that the complex adopts an occluded conformation, whereas the outward-open conformation is populated by residues located in the cytoplasmic-permeation pathway.

A density map at 3.6 resolution was obtained for the ibogaine – 15B8 Fab – SERT complex in KCl. A distinct density feature associated with TM1a was observed, suggesting that this protein is splayed away from the transporter core.

Fig. 1 shows the chemical structures of ibogaine and serotonin, and the uptake of [14C]5-HT by ts2-active SERT in the absence and presence of 15B8 Fab and 8B6 scFv. [3H]ibogaine bound to ts2-active SERT in the presence of 15B8 Fab, 100 mM KCl, 100 mM NMDG-Cl and 100 mM KCl/NaCl, and the S277C mutant in the presence of 10 mM MTS-ACMA.

Cryo-EM images of SERT bound to 15B8 Fab – 8B6 scFv show outward-open, occluded and inward-open conformations, as well as movements of TM6a and TM1a from outward-open to occluded and inward-open conformations.

We examined the influence of small molecules on the conformation of SERT by reconstructing the complex in 3D and fitting a density map with noribogaine. The inward-open conformation was found to be the best fit.

The density maps enabled the localization of ibogaine at the central site, and computational docking followed by molecular dynamics simulations revealed that the tertiary amine interacts with Asp98 and the tricyclic ring system lodges between the aromatic groups of Tyr176 and Tyr95.

Ibogaine binds to the tryptamine group of the transporter and the aromatic ring of Phe341, and TM1a moves from the outward-open and occluded to the inward-open conformation, allowing ibogaine to move towards the cytoplasmic-permeation pathway.

We assessed the binding pose of ibogaine and found that the side chain of Asn177 resides near the methoxy group of ibogaine. This resulted in a more robust inhibition of 5-HT uptake by ibogaine and a weaker inhibition by noribogaine.

Ibogaine binds to the Asn177 site in the iboga receptor and changes conformation upon isomerization from the outward-open to the occluded and inward-open states. The Asn177 mutants N177V, N177A, and N177T show decreased ibogaine binding and increased inhibition of noribogaine uptake.

The binding of [3H]ibogaine to ts2-active SERT was measured in the outward-open, occluded and inward-open conformations. The distances between extracellular and intracellular gating residues are shown.

We next analysed the position of the extracellular and intracellular gates in SERT and found that the ibogaine-bound, outward-open reconstruction is similar to the X-ray structure of paroxetine-bound SERT28. The core TMs of SERT undergo movements that close the extracellular gate, preventing access to the central binding site. The closure of the extracellular gate in NSSs reduces the likelihood of association of ibogaine or similar small molecules with the allosteric site, thus facilitating the transition to the inward-open conformation and the opening of TM1a.

We found that the most noteworthy structural rearrangements are at the closed extracellular and open intracellular gates, where TM1b shifts and tilts by 5.1 , 22°, TM6a moves by 3.4 , 5° towards the scaffold, and TM2 and TM7 undergo an associated movement of 2.8 , 1.0 , and 7.3° towards the scaffold, respectively. The movement of TM1a is accompanied by structural changes in TM5, which expand laterally into the membrane and shift the transporter from an occluded to an inward-open conformation.

The conformational changes observed in bacterial amino acid transporters are similar to those observed in SERT, although deviations from the prototypical model are also present. In SERT, TM1a samples different orientations upon the rupture of the intracellular gate.

We examined the positions of ions surrounding sodium sites in the outward-open, occluded and inward-open conformations of the protein. The positions suggest that Na1 and Na2 are occupied by two bound sodium ions, and Na2 is directly coupled to substrate transport.

Ibogaine interacts with SERT in outward-open, occluded and inward-open conformations, and may remain bound and enable transporter isomerization. This may explain the non-competitive inhibition of transport displayed by ibogaine.

Ibogaine inhibits SERT by binding to the outward-open conformation followed by stabilization of the occluded or inward-open conformations, or by directly binding to the inward-open conformation.

Ibogaine binds to SERT via a two-step mechanism, in which it binds to an outward-open conformation and stabilizes an occluded or inward-open conformation. This mechanism provides insight into how high-affinity small molecules might be crafted to selectively bind to SERT.

Online content

Singh, S. K. & Pal, A. studied LeuT, a prokaryotic homolog of neurotransmitter sodium symporters, and Kristensen, A. S. studied SLC6 neurotransmitter transporters, and Gether, U. studied neurotransmitter transporters. In neurotransmitter transporters, alternating access is controlled by ligand-dependent alternating access, and substrate-modulated unwinding of transmembrane helices is observed in the NSS transporter LeuT. Ibogaine, a new alkaloid in iboga, has excitement-producing properties, and a mechanistic basis for noncompetitive inhibition of serotonin and dopamine transporters. Fabs enable single particle cryoEM studies of small proteins.

Zhang, Y. W., Rudnick, G., Burtscher, V., Hotka, M., Li, Y., Freissmuth, W. & Sandtner, W. (2018). Fluoxetine (Prozac) binding to serotonin transporter is modulated by chloride and conformational changes, and phosphorylation of threonine residue 276 is required for acute regulation of serotonin transporter by cyclic GMP. Dopamine transporter X-ray structure reveals antidepressant mechanism; serotonin transporter lipid binding modulates conformation, pharmacology, and transport kinetics. The X-ray structure and mechanism of the human serotonin transporter was described, as well as the identification of an allosteric citalopram-binding site at the serotonin transporter, and the thermostabilization and purification of the human dopamine transporter in an inhibitor and allosteric ligand bound conformation.

We thank the National Institute for Drug Abuse, Drug Supply Program, L. Vaskalis, H. Owen, V. Navratna, and M. Whorton for their assistance with this work, and the staff of the Northeastern Collaborative Access Team at the Advanced Photon Source for their support.

Author contributions include D.Y.’s initiating studies on the ibogaine inward-open conformation, J.A.C.’s initiating cryo-EM studies on SERT – antibody complexes, J.A.C.’s and D.Y.’s collecting electron microscopy data, and J.A.C.’s writing the manuscript.

MethodS

Antibody production. The 15B8 Fab was produced by papain digestion of 15B8 mAb and purification by cation-exchange chromatography, using standard methods28, or by isolation of recombinantly expressed Fab from Sf9 supernatant by metal affinity chromatography for crystallization34.

SERT was expressed and purified as described previously28,35,36 using baculovirus-mediated transduction of HEK-293S GnTI cells (ATCC). The purification tags were removed by thrombin digestion, and the SERT protein was dissolved in 1 mM DDM and purified with Strep-Tactin affinity chromatography. SERT, Fab and ScFv were mixed at a 1:1.2:1.2 molar ratio and purified by size-exclusion chromatography on a Superdex 200 column in TBS containing 9 mM nonylmaltoside, 0.2 mM CHS, and 1 mM paroxetine or 10 mM ibogaine.

The 15B8 Fab was crystallized by hanging-drop vapour diffusion and cryoprotected with 25% ethylene glycol before flash-cooling in liquid nitrogen.

SERT – antibody complexes were applied to glow-discharged Quantifoil holey carbon grids and frozen in liquid ethane cooled by liquid nitrogen. Images were acquired using a FEI Titan Krios equipped with a Gatan Image Filter operating at 300 kV or an Arctica transmission electron microscope at 200 kV.

Image processing was performed on micrographs to correct for beam-induced drift, determine the contrast transfer function (CTF) parameters, pick particles, perform reference-free 2D classification, homogenous refinement, and local refinement. The resolution of the reconstructions was assessed using the Fourier shell correlation (FSC) criterion.

A total of 1,278,876 particles were selected from 2,904 micrographs and subjected to two rounds of 2D classification using cryoSPARC. An ab initio model with two classes was generated and further refined using local refinement in cisTEM. For the ts2-active ibogaine Fab – scFv dataset, 592,117 particles were selected from 1,639 micrographs, 153,986 particles had clearly defined features, and 724,394 particles were subjected to local refinement in cisTEM. The optimal sharpening B factor was determined by comparing map features for various sharpening factors in cisTEM.

We used rigid-body fitting to interpret the cryo-EM maps of SERT and antibodies derived from X-ray crystallography. We were able to model the main chain of SERT and position most of the bulky side chains. Model refinement was performed using iterative local rebuilding in Rosetta, followed by combining pieces from multiple templates and refinement in RosettaCM. The paroxetine-bound model was refined separately in RosettaCM starting from SERT (PDB code: 6AWN). MolProbity was used to evaluate the stereochemistry and geometry of the structures, and docking and molecular dynamics simulations were used to further our interpretation of the large-scale rearrangements of structural elements in each conformation.

Measurements were made from C positions to determine the distances between TM helices and to determine the angular change between conformations. The uncertainty of each measurement was calculated from 100 models and real space refined ‘back’ into each map in PHENIX.

The structure of SERT was calculated from residues within 5 of the (S)-citalopram structure, and the protein was prepared for docking and molecular dynamics simulations by removing antibody fragments, adding missing hydrogen atoms and side chains, and by removing CHS from the occluded conformation.

The force field parameters of protonated ibogaine were developed on the basis of the CHARMM General Force Field (CGenFF)58. The partial atomic charges of the aliphatic carbon and hydrogen atoms were assigned according to the convention of CHARMM force fields, and the parameters were further optimized using the Force Field Toolkit plugin of VMD.

The bonded parameters of ibogaine contain 2 novel bonds, 11 novel angles, and 41 novel dihedrals, which were not defined in the standard CGenFF force field. All quantum mechanical calculations were performed using Gaussian 09 (ref. 62).

A workflow was developed to systematically search for optimal binding poses of ibogaine in the outward-open, occluded, and inward-open conformations of SERT, independently. The results were analysed using a hybrid k-centres k-medoids clustering method64 with a 2- cut-off.

A total of 96 independent molecular dynamics simulations were performed for the SERT – ibogaine complex, with 800 snapshots taken per simulation. The ibogaine pose with the highest averaged cross-correlation coefficient was selected as the optimal pose for each SERT conformation.

Molecular dynamics simulations were performed on ligand-bound SERT systems using the Dowser65 plugin of VMD, a lipid bilayer composed of 236 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) molecules, and solvated with NaCl or KCl.

To investigate the stability of the ibogaine binding poses, two 50-ns simulations were performed with each ibogaine-bound conformation in a POPC lipid bilayer, resulting in six trajectories.

All simulations were performed using NAMD263, CHARMM36m, TIP3P, and CGenFF force fields, with a constant temperature and pressure of 310 K and 1.01325 bar, respectively. Non-bonded interactions were calculated in a pairwise manner within the 12-cut-off, and long-range non-bonded interactions were calculated with the PME method.

Data analysis was performed on simulation trajectories using MDToolbox, VMD54, MDAnalysis74,75, and NAMD263. The PIE was calculated using NAMD263.

Reconstituted SERT was mixed with soybean asolectin and MSP1E3D1 and reconstituted into nanodiscs. The reconstituted SERT was incubated with 1 mM ibogaine or 0.2 mM paroxetine for 30 min, followed by labelling with 10 m MTS-ACMA for the indicated time.

To measure uptake, 1 105 HEK-293S GnT I cells were transduced with ts2-active, Asn177 mutants or n72/C13 SERT, and then washed with uptake buffer, antibodies, paroxetine and [3H]5-HT were added to the cells. The amount of labelled 5-HT was measured by counting in a standard 96-well plate.

Competition binding experiments were performed using scintillation proximity assays with SERT, Cu-YSi beads, and antibodies at concentrations ranging from 0.1 nM to 1 mM. Ki values were determined with the Cheng – Prusoff equation78.

Ibogaine binding was measured via SPA using ts2-active SERT purified in SPA buffer, and non-specific binding was estimated by experiments that included 100 m unlabelled paroxetine. Data were analysed using a single-site binding function.

Data availability

The data that support the findings of this study are available from the corresponding author upon request and are deposited in the Protein Data Bank and Electron Microscopy Data Bank.

Thermostabilization, expression, purification, and crystallization of the human serotonin transporter bound to S-citalopram were performed, and the structure of the transporter was determined. Several software tools have been developed to facilitate particle selection in single particle electron microscopy, including Mastronarde, D. N., MotionCor2: anisotropic correction of beam-induced motion for improved cryo-electron microscopy, Zheng, S. Q., Gctf: real-time CTF determination and correction, and RELION: implementation of a Bayesian approach to structure determination.

46. Adams, P. D., Pettersen, E. F., Wang, R. Y., et al. describe systems for macromolecular structure solution using Python. There are many tools for molecular graphics, such as Coot, MolProbity, PyMOL, CAVER 3.0, Humphrey, W., Dalke, A., Freddolino, P. L. & Schulten, K., that can be used for analysis and validation of cryo-EM maps and atomic models. 58. Vanommeslaeghe, K. et al., Automation of the CHARMM General Force Field (CGenFF) II: assignment of bonded parameters and partial atomic charges. Several groups have published work on parameterization of small molecules using the force field toolkit, automation of the CHARMM General Force Field (CGenFF) I: bond perception and atom typing, and a web-based graphical user interface for CHARMM. Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L., developed simple potential functions for simulating liquid water, and Feller, S. E., Zhang, Y., Pastor, R. W. & Brooks, B. R., developed constant pressure molecular dynamics algorithms. MDAnalysis is a python package for the rapid analysis of molecular dynamics simulations. A competitive inhibitor is a compound that inhibits an enzyme by preventing the enzyme from catalyzing a reaction.

Ibogaine inhibits 5-HT transport for wild-type and ts2-active SERT variants using 20 mM [3H]5-HT. Ibogaine competes with paroxetine for binding to the ts2-inactive site and yields a Ki value of 3.2 0.4 M. Ibogaine saturation binding experiments of ts2-inactive and ts2-active 15B8 Fab – 8B6 scFv complex in 100 mM NaCl were performed. MTS-ACMA labelling of S277C in the presence of ibogaine or paroxetine was analysed by SDS-PAGE and visualized by in-gel fluorescence. The experiment was performed three times independently with similar results. Paroxetine bound to SERT is shown in three-dimensional reconstruction and fit to the density map. The fit of paroxetine into the electron-microscopy density map and interacting residues are shown.

Cryo-EM reconstruction of ts2-active SERT – 15B8 Fab – 8B6 scFv – paroxetine complex was performed using cryoSPARC. One out of two predominant classes exhibited a subset of homogeneous particles that were used for further processing and global alignment in cryoSPARC.

A cryo-EM density map was created from the SERT – Fab complex and a 2D class average was obtained after three rounds of classification. A 3D class average was obtained from the SERT alone and did not reveal any substantial differences.

Cryo-EM data processing of the ts2-active 15B8 Fab – 8B6 scFv – ibogaine complex was performed using cryoSPARC, RELION, and Gctf. The final reconstructed volume was sharpened using cisTEM.

The final reconstruction of the SERT-Fab complex was performed using cryo-EM and a spherical mask placed over SERT was used for focused 3D classification with 3 classes. The comparison of the classes did not reveal any substantial differences.

A cryo-EM reconstruction of the n72/C13 SERT – 15B8 Fab – ibogaine complex in NaCl was performed after particle picking, sorting, and 2D classification. The final reconstructed volume was sharpened using cisTEM.

The final map was constructed from Cryo-EM data of the SERT – 15B8 Fab complex, and a spherical mask was placed over SERT to perform 3D classification. The comparison of the classes did not reveal any substantial differences.

Cryo-EM data processing of the ibogaine complex between n72/C13 SERT and 15B8 Fab was performed in KCl. After particle picking, particles were sorted using 2D classification, and then refined in RELION using 3D classification. The cryo-EM micrographs show the individual single particles, the 2D class averages after three rounds of classification, the angular distribution of particles used in the final reconstruction, the FSC curves for cross-validation, the final map, masked SERT – Fab complex, and a mask that isolated SERT.

Cholesteryl hemisuccinate, non-proteinaceous density features near Thr276 and Ser277, and SERT – noribogaine complex were studied in 100 mM NaCl. Noribogaine inhibited 5-HT transport for N72/C13 SERT. Michaelis-Menten plots of 5-HT uptake for the n72/C13 transporter in the absence and presence of 1 M noribogaine are shown, as well as a binding site density map for the n72/C13 SERT – 15B8 – noribogaine complex.

Ibogaine docking and molecular dynamics simulations were performed to study the binding of ibogaine to the SERT transporter in different conformations. The structural stability of bound ibogaine was measured as the mass-weighted r.m.s.d. of the ligand, as well as the Asp98 – ibogaine (O – N) distance.

Fig. 8 shows the inhibition of serotonin transport by ibogaine and noribogaine in ts2 mutants, the effect of thermostabilizing Y110A mutation, and the comparison of EL4 and TM1b in the X-ray structure of the ts3 – paroxetine complex and the cryo-EM structure of the ts2 active – ibogaine complex.

The helical displacement of TMs from the outward-open to the occluded conformation, and from the occluded to the inward-open conformation, are shown, as well as the angular changes of TMs associated with transition from the outward-open to the occluded conformation and from the occluded to the inward-open conformation.

SERT was compared with LeuT, MhsT and ibogaine-bound outward-open and occluded conformations. The r.m.s.d. and angular differences between the outward-open and occluded conformations of SERT and LeuT were calculated.

The X-ray structures of SERT show that the sodium and chloride ion-binding sites and putative substrate and ion-release pathways are located in the inward-open conformation. The minimum radius of the tunnels from the central binding site to the intracellular space is approximately 2.5 .

Cryo-eM data collection, refinement and validation statistics for the ts2-inactive paroxetine 15B8 Fab, the ts2-active ibogaine outward-open 15B8 Fab, the n72/C13 ibogaine occluded 15B8 Fab, and the n72/C13 noribogaine inward-open 15B8 Fab.