Dear Editor,
Riluzole (Rilutek) is one of the few drugs prescribed in clinic for the treatment of amyotrophic lateral sclerosis1. In addition, riluzole shows promise in the treatment of psychiatric disorders and neurodegenerative movement disorders due to its neuroprotective effect2,3. It is reported that riluzole can activate the transient receptor potential canonical subfamily member 5 (TRPC5) channel4, which is a nonselective calcium-permeant cation channel5. TRPC5 is widely expressed in the brain and kidney, and serves as a potential therapeutic target for psychiatric disorders and progressive kidney disease6. TRPC5 channel can be activated by diverse stimuli5, including external and internal calcium, trivalent lanthanides (La3+, Gd3+), depletion of intracellular Ca2+ stores, reduced extracellular thioredoxin, hypo-osmotic buffer condition, indirect pathways like the stimulation of membrane receptors (receptor activation), and small-molecule activators like Englerin A and riluzole. Activation of TRPC5 has therapeutic potential for the treatment of kidney cancers6. Among the activators, riluzole is highly specific for TRPC5 and does not activate other TRP channels, including the closest homolog TPRC isoforms4. Although there are reports on the structures of TRPC5 in complex with several inhibitors available7,8,9, how activators, such as riluzole, bind TRPC5 remains unknown, which impedes not only the elucidation of the activation mechanism of TRPC5 but also structure-based drug discovery targeting TRPC5 activation.
Here we purified the truncated functional human TRPC5 (hTRPC51–764) protein9 and supplemented the protein with riluzole for cryo-EM sample preparation. Subsequent single particle analysis resolved the structure of hTRPC5 in complex with riluzole at a global resolution of 2.4 Å (Fig. 1a; Supplementary Figs. S1, S2a–c and Table S1), which is the highest among available TRPC5 structures so far7,8,9. The excellent map quality revealed densities of many lipids and ligands, including previously identified diacylglycerol (DAG)9 (Supplementary Fig. S2d) and riluzole (Supplementary Fig. S2e), as well as calcium ion inside the voltage sensor-like domain (VSLD)9 (Supplementary Fig. S2f). The overall architecture of riluzole-bound hTRPC5 is similar to the apo state or the clemizole (a TRPC5 inhibitor)-bound structures7,9, with a root-mean-square deviation (RMSD) of 0.781 Å and 0.682 Å, respectively. hTRPC5 with riluzole bound shows a four-fold symmetric homotetramer with dimensions of 100 Å × 100 Å × 130 Å. Each monomer of hTRPC5 is comprised of an intracellular cytosolic domain (ICD) and a transmembrane domain (TMD) (Fig. 1a), which can be further divided into ion channel pore (Segment 5–Segment 6, S5–S6 for short) and VSLD (S1–S4) (Supplementary Fig. S2d).
Riluzole binds inside a pocket of VSLD and is adjacent to the Ca2+ ion density with the distance of 4 Å (Fig. 1b, c). Several residues on S1, S2, S3, and S4 are closed to the ligand density of riluzole, indicating a direct interaction between the drug and the channel (Supplementary Fig. S2f). Specifically, Y374 on S1 and F414 on S2 form π–π stacking interaction with benzothiazole ring of riluzole. R492 on S4 forms a cation-π interaction with the benzothiazole ring of riluzole. D439 on S3 and S495 on S4 form hydrogen bonding (H-bond) with the amine and trifluoromethoxy of riluzole, respectively. N443 on S3 shows hydrophobic interaction with the amine of riluzole. M442 on S3, L493 on S4, and L496 on S4 may form Van der Waals interaction with riluzole (Fig. 1d–e). To elucidate the contribution of these residues to the binding of riluzole, we mutated them into alanines or their counterparts in hTRPC3/6/7. These mutants, including F414M, N443L, S495A, and L496A, were previously found to have proper tetramer assembly, unaltered surface expression, and robust calcium influx9. To avoid the interference of Ca2+ modulation, HEK293 cells expressing hTRPC5 mutants were recorded in 0 mM extracellular Ca2+ solution. We found that all these mutations decrease the potency of riluzole to some extent (Fig. 1f; Supplementary Fig. S3, and Table S2). Notably, the corresponding residue of F414 in hTRPC5 is Met in hTRPC1/3/6/7, and N443 in hTRPC5 is Leu in hTRPC3/6/7. Both F414M and N443L decrease the potency of riluzole towards hTRPC5 (Fig. 1f), in agreement with the selectivity of riluzole between hTRPC5 and hTRPC3/6/74.
Riluzole-binding site is close to a calcium ion. We wonder if there is any relationship between riluzole activation and calcium activation. The Ca2+ ion in VSLD interacts with E418 on S2, E421 on S2, and D439 on S3 (Fig. 1d), and has been reported to mediate calcium activation of TRPC59. hTRPC5 triple mutant (E418Q, E421Q, and D439N, the EED mutant) almost totally abolished the activation of TRPC5 by 14 mM extracellular calcium9 (Supplementary Fig. S4a–c). However, the potency of riluzole on this mutant remains unchanged compared to wild-type hTRPC5 (WT) in the presence of 0 mM extracellular Ca2+ (Supplementary Fig. S4d). To further study whether calcium binding in this site would affect the activation of riluzole, we carried out electrophysiological experiments to record the hTRPC5 currents under three conditions: 2 mM extracellular Ca2+; 100 μM riluzole in 0 mM extracellular Ca2+; 100 μM riluzole in 2 mM extracellular Ca2+. 2 mM Ca2+ or 100 μM riluzole in 0 mM extracellular Ca2+ could elicit small hTRPC5 currents. However, the hTRPC5 currents were potentiated by several folds by 100 μM riluzole in 2 mM extracellular Ca2+ (Fig. 1g and Supplementary Fig. S4e). In contrast, although 100 μM riluzole could elicit currents in EED mutant in the absence of extracellular Ca2+, the presence of 2 mM extracellular Ca2+ failed to further potentiate the riluzole-activated current (Fig. 1h and Supplementary Fig. S4f). Statistical analysis showed that the riluzole-activated hTRPC5 currents were potentiated ~7-fold at +100 mV and ~90-fold at −60 mV in 2 mM Ca2+ relative to 0 mM Ca2+. In contrast, the riluzole-activated currents in EED mutant in 2 mM Ca2+ were 0.3-fold at +100 mV and 0.5-fold at −60 mV relative to 0 mM Ca2+ (Fig. 1i, j). These results comprehensively suggest the synergistic effect between riluzole and the nearby calcium on the activation of TRPC5. This is also in agreement with that the EC50 of riluzole is around 210 μM in the absence of extracellular Ca2+ at −100 mV (Fig. 1f and Supplementary Table S2) but decreased to 20.7 μM in the presence of 1.5 mM extracellular Ca2+ at −100 mV4. Besides the Ca2+, La3+ was reported to potentiate the riluzole-activated TRPC5 currents4, but the actual effectiveness is much weaker than expected. The outward currents at +100 mV have no significant change after La3+ stimuli. And the La3+-dependent potentiation of inward current at −100 mV is < 2-fold4. Mutagenesis studies suggest that the La3+-binding site locates at the extracellular loop of S5–S65. The different binding sites and weak synergistic effect infer that riluzole and La3+ activate the TRPC5 by different mechanisms. In contrast, both riluzole and Ca2+ bind inside VSLD and exhibit a remarkable synergistic effect, suggesting that they share a converged activation mechanism.
The ion permeation pathway of riluzole-bound hTRPC5, lined by pore helices, pore loops, and S6 of four protomers (Supplementary Fig. S5a), is constricted at the lower gate formed by I621, N625, and Q629, similar to the apo state and clemizole-bound state (Supplementary Fig. S5b), suggesting that our riluzole-bound structure is in a non-conductive closed state, which correlates with the relatively low open probability of TRPC5 upon riluzole stimulation (NPO is around 0.4 at −80 mV)4. Activators with higher efficacy or gain-of-function mutations might further increase the PO to stabilize the open-state structure of TRPC5.
We previously reported the structure of hTRPC5 in complex with an inhibitor clemizole9. Both clemizole and riluzole are benzothiadiazine derivatives, and bind to the same pocket inside VSLD. The structural comparison showed that the binding pocket of clemizole and riluzole are largely overlapped (Supplementary Fig. S5c). Further analyses on the two structures reveal that their main chains are almost superposable, with a RMSD of 0.682 Å (Supplementary Fig. S5c). All the riluzole-interacting residues are involved in clemizole-binding (Supplementary Fig. S5d). However, because clemizole is larger than riluzole, clemizole interacts with more residues in VSLD than riluzole, including P659, Y446, E418, and G417. These structural observations suggest a competitive binding behavior and are in agreement with previous electrophysiological results10.
The synergistic effect between riluzole and Ca2+ on hTRPC5 activation is akin to that observed for icilin and Ca2+ on the activation of the TRPM8 channel11. Icilin is an agonist of TRPM8 and binds inside the VSLD11. The binding of icilin and Ca2+ in VSLD induces the structural changes of the lower portion of S4 from α helix to 310 helix, which further drive the opening of the TRPM8 channel11 (Supplementary Fig. S6a, b). Notably, another TRPM8 agonist WS-12 also binds inside VSLD but failed to drive the conformational change of VSLD and subsequent opening of the channel11 (Supplementary Fig. S6a, b). In addition, TRPM8 antagonists, such as AMTB and TC-I 2014, all bind inside the VSLD12 (Supplementary Fig. S6a, c). The larger chemical groups of antagonists (molecular weight of AMTB: 431.0 Da and TC-I 2014: 467.4 Da versus WS-12: 289.4 Da and icilin: 311.3 Da) may prevent the activating conformational change of S4 of VSLD and stabilize the closed state of TRPM812 (Supplementary Fig. S6d).
Given the similar structural features of TMD between TRPM8 and TRPC5 and their conserved calcium binding sites in VSLD, we speculate TRPC5 shares a similar activation mechanism to TRPM8. Riluzole might behave like WS-12, an agonist of TRPM8, which has low efficacy, activates the channel to some level and does not prevent the activating conformational changes of VSLD. In contrast, the bulkier clemizole (molecular weight of clemizole: 325.8 Da vs riluzole: 234.2 Da) behaves like AMTB, an antagonist of TRPM8, which also binds inside VSLD but blocks the conformational changes essential for channel activation (Supplementary Fig. S6e).
Besides TRPC5, riluzole is also reported to activate the small conductance, Ca2+-activated K+ channel (SK channel) in cultured hippocampal neurons and in HEK293 cells expressing recombinant SK2 channels. The crystal structure of SK2 C-terminal domain–calmodulin (CaM)–riluzole complex was reported13. The structure revealed that riluzole could bind inside a pocket situated at the intracellular interface of SK2–CaM channel complex13 (Supplementary Fig. S7a–c). Although the riluzole-binding residues are different between hTRPC5 and SK2 (Fig. 1d and Supplementary Fig. S7c), the chemical environments of their binding pockets share some similarities (Supplementary Fig. S7d–g). These two pockets exhibit similar negatively-charged surface but different hydrophobic patches. In detail, both pockets exhibit highly negatively-charged surface at the mouth, binding to the positively-charged amine functional group of riluzole. Whereas inner spaces of the two pockets are less negatively-charged (Supplementary Fig. S7d, e). The hydrophobicity of riluzole-binding pocket seems to be less important for riluzole binding, since the inner space of riluzole-binding pocket in TRPC5 is highly hydrophilic, which contrasts with the highly hydrophobic surface of the inner pocket of the SK2–CaM complex (Supplementary Fig. S7f, g). In addition to SK2 and hTRPC5, riluzole has a wide range of effects on multiple ion channels and receptors, including NaV, KV, CaV, and ligand-gated neurotransmitter receptors14. The diversity of riluzole-binding sites might be due to the low molecular weight (234.2 Da) and small van der Waals volume (164 Å3) of the drug.
In conclusion, our structure of hTRPC5 in complex with riluzole reveals that the riluzole binds inside the VSLD. The binding site of riluzole overlaps with some of the hTRPC5 inhibitors such as clemizole9, and TRPC4 inhibitors such as pyridazinone-based small molecules15, suggesting the VSLD is a key regulatory module for TRPC channel activity and a hot spot for the binding of small molecule gating modifiers. Moreover, we found riluzole-binding site is close to the activating Ca2+ ion-binding site in VSLD. More importantly, riluzole and calcium display a synergistic effect on the activation of hTRPC5. Our studies provide insights into the gating mechanism of hTRPC5 and how riluzole impacts its function.
Data availability
Density maps are deposited at the Electron Microscopy Database (accession codes: EMD-32436) and protein coordinates are deposited at the Protein Data Bank (PDB code: 7WDB).
References
Miller, R. G., Mitchell, J. D. & Moore, D. H. Riluzole for amyotrophic lateral sclerosis (ALS)/motor neuron disease (MND). Cochrane Database Syst. Rev. 2012, CD001447 (2012).
Liu, J. & Wang, L. N. The efficacy and safety of riluzole for neurodegenerative movement disorders: a systematic review with meta-analysis. Drug Deliv. 25, 43–48 (2018).
Zarate, C. A. & Manji, H. K. Riluzole in psychiatry: a systematic review of the literature. Expert. Opin. Drug Metab. Toxicol. 4, 1223–1234 (2008).
Richter, J. M., Schaefer, M. & Hill, K. Riluzole activates TRPC5 channels independently of PLC activity. Br. J. Pharmacol. 171, 158–170 (2014).
Zholos, A. V. TRPC5. Handb. Exp. Pharmacol. 222, 129–156 (2014).
Sharma, S. & Hopkins, C. R. Review of transient receptor potential canonical (TRPC5) channel modulators and diseases. J. Med. Chem. 62, 7589–7602 (2019).
Duan, J. et al. Cryo-EM structure of TRPC5 at 2.8-Ã… resolution reveals unique and conserved structural elements essential for channel function. Sci. Adv. 5, eaaw7935 (2019).
Wright, D. J. et al. Human TRPC5 structures reveal interaction of a xanthine-based TRPC1/4/5 inhibitor with a conserved lipid binding site. Commun. Biol. 3, 704 (2020).
Song, K. et al. Structural basis for human TRPC5 channel inhibition by two distinct inhibitors. Elife 10, e63429 (2021).
Richter, J. M., Schaefer, M. & Hill, K. Clemizole hydrochloride is a novel and potent inhibitor of transient receptor potential channel TRPC5. Mol. Pharmacol. 86, 514–521 (2014).
Yin, Y. et al. Structural basis of cooling agent and lipid sensing by the cold-activated TRPM8 channel. Science 363, eaav9334 (2019).
Diver, M. M., Cheng, Y. & Julius, D. Structural insights into TRPM8 inhibition and desensitization. Science 365, 1434–1440 (2019).
Cho, L. T. et al. An intracellular allosteric modulator binding pocket in SK2 ion channels is shared by multiple chemotypes. Structure 26, 533–544 (2018).
Bellingham, M. C. A review of the neural mechanisms of action and clinical efficiency of riluzole in treating amyotrophic lateral sclerosis: what have we learned in the last decade? CNS Neurosci. Ther. 17, 4–31 (2011).
Vinayagam, D. et al. Structural basis of TRPC4 regulation by calmodulin and pharmacological agents. Elife 9, e60603 (2020).
Acknowledgements
We thank for the helpful suggestions from Wenjun Guo and Kangcheng Song. The cDNAs of hRRPC5 were generously provided from Dr. Xiaolin Zhang. Cryo-EM data were collected on the Cryo-EM platform of Peking University with the assistance of Xuemei Li, Daqi Yu, Xia Pei, Bo Shao, Guopeng Wang, and Zhenxi Guo. This work is supported by grants from Ministry of Science and Technology of China, the National Natural Science Foundation of China (91957201, 31870833, and 31821091 to L.C., 11902021 to Y.Y.) and the Fundamental Research Funds for the Central Universities (YWF-21-BJ-J-1150 to Y.Y.).
Author information
Authors and Affiliations
Contributions
L.C. conceived the project and supervised all the research. M.W. prepared the sample and performed the EM analysis. M.W. performed the structural determination. Y.Y. performed the electrophysiological experiments. Y.Y. and L.C. wrote the manuscript with the help from all authors.
Corresponding author
Ethics declarations
Conflict of interest
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Yang, Y., Wei, M. & Chen, L. Structural identification of riluzole-binding site on human TRPC5. Cell Discov 8, 67 (2022). https://doi.org/10.1038/s41421-022-00410-5
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41421-022-00410-5
This article is cited by
-
Thermosensing ability of TRPC5: current knowledge and unsettled questions
The Journal of Physiological Sciences (2024)
-
Functional determinants of lysophospholipid- and voltage-dependent regulation of TRPC5 channel
Cellular and Molecular Life Sciences (2024)
-
Effect of Clemizole on Alpha-Synuclein-Preformed Fibrils-Induced Parkinson’s Disease Pathology: A Pharmacological Investigation
NeuroMolecular Medicine (2024)