Niclosamide potentiates TMEM16A and induces vasoconstriction

TMEM16A, also known as anoctamin-1 (ANO1), is a bona fide calcium-activated chloride channel (CaCC) (Caputo et al., 2008; Schroeder et al., 2008; Yang et al., 2008). It is widely expressed in various cell types and plays crucial roles in regulating physiological processes such as smooth muscle contraction, gut motility, fluid secretion, cell volume regulation, nociception, and anxiety (Pedemonte and Galietta, 2014; Whitlock and Hartzell, 2017; Kunzelmann et al., 2019; Hawn et al., 2021; Le et al., 2021; Song et al., 2022). Upregulation of TMEM16A has been reported in a wide range of pathological conditions, including inflammatory airway disease, asthma and chronic obstructive pulmonary disease (COPD), cancer, hypertension, and ischemia/reperfusion injury (Ma et al., 2017; Wang et al., 2017; Crottès and Jan, 2019; Papp et al., 2019; Bai et al., 2021; Korte et al., 2022). Recently, TMEM16A variants have been linked to Moyamoya disease, a pediatric cerebrovascular disease characterized by progressive occlusion of carotid arteries and a high incidence of stroke (Pinard et al., 2023). Therefore, pharmacological manipulation of TMEM16A provides a promising new strategy to treat these diseases.

A wide variety of TMEM16A inhibitors have been identified since its molecular cloning in 2008 (Namkung et al., 2011; Peters et al., 2015; Seo et al., 2016; Centeio et al., 2020; Liu et al., 2021; Al-Hosni et al., 2022). Niclosamide, one of the World Health Organization’s essential medicines for treating tapeworm infections, was recently reported to be a TMEM16A inhibitor (Miner et al., 2019; Henckels et al., 2020). This finding subsequently led to the proposal that niclosamide could be repurposed to suppress upregulated TMEM16A activity in inflammatory airway diseases such as asthma and COPD (Cabrita et al., 2019; Ousingsawat et al., 2022). However, other studies are contradictory and do not support niclosamide as a clean TMEM16A inhibitor (Ousingsawat et al., 2022; Genovese et al., 2023). Niclosamide has off-target effects including disruption of calcium signaling by inhibiting calcium store release (Ousingsawat et al., 2022; Genovese et al., 2023).

To investigate the effects of niclosamide on TMEM16A, here, we use patch clamp, molecular docking and simulation, structure-guided mutagenesis, calcium imaging, and pressure myograph to study the effect of niclosamide on heterologously expressed TMEM16A in HEK293T cells and endogenous TMEM16A in vascular smooth muscle cells (vSMCs). We find that, under physiological calcium and voltage, niclosamide acutely potentiates TMEM16A activation without any obvious inhibitory effect. Micromolar niclosamide potentiates TMEM16A by binding to a putative extracellular binding site. We further demonstrate that niclosamide potentiates endogenous TMEM16A in vSMCs, increases intracellular calcium, and constricts the mesenteric artery. Our study, together with another recent report (Danahay et al., 2023), shows no evidence that niclosamide inhibits TMEM16A CaCC when Ca²⁺ and voltages are within a physiological range. The acute potentiation effect of niclosamide on TMEM16A imposes a warning about repurposing niclosamide as a TMEM16A inhibitor to treat human diseases such as asthma and COPD. The putative niclosamide binding site identified in this study may inform the development of more potent and specific TMEM16A potentiators to mitigate diseases that require augmented CaCC activity, including cystic fibrosis and gastroparesis.

Regents used included: Niclosamide (cat#: N3510; Sigma-Aldrich), phenylephrine hydrochloride (PE) (cat#: P6126; Sigma-Aldrich), acetylcholine chloride (ACh) (cat#: A6625; Sigma-Aldrich), amphotericin B (cat#: 15290018; Gibco), and collagenase type 2 (cat#: LS004177; Worthington). Other reagents were purchased from Sigma-Aldrich unless otherwise specified.

Mouse handling and usage were carried out in strict compliance with the protocol approved by the Institutional Animal Care and Use Committee at Duke University, in accordance with National Institute of Health guidelines. C57BL6/J (stock #000664) mice and Acta2-GCaMP8.1/mVermilion (Lee et al., 2021) mice expressing calcium sensor GCaMP8.1 and mVermilion reporter in SMCs were purchased from the Jackson Laboratory (stock #032887).

The pressure myograph experiments were done on the third-order mesenteric arteries following the procedures previously described (Shahid and Buys, 2013; Rode et al., 2017; Turner et al., 2019). Briefly, the third-order mesenteric arteries were dissected and placed immediately into ice-cold HEPES-PSS solution (125 mM NaCl, 3.8 mM KCl, 1.2 mM CaCl₂, 25 mM NaHCO₃, 1.2 mM KH₂PO₄, 1.5 mM MgSO₄, 0.02 mM EDTA, and 8 mM D-glucose, pH 7.4) pre-equilibrated with 95% O₂/5% CO₂ for 15 min. Small segments of 3–5 mm in length with no branching were cut from the mesenteric artery and mounted on glass cannulas in a pressure myograph (model 110p, Danish Myo Technology A/S). Two pressure transducers (P1 on the right and P2 on the left side) were built in to monitor the pressure within the artery lumen. The vessels were equilibrated at 60 mmHg and 37°C for at least 45 min. The bath solution was changed once with prewarmed HEPES during equilibration. The viability of the vessel and integrity of the endothelial cells were assessed by addition of 10 µM PE to constrict and 10 µM ACh to relax the vessel. Only the arteries that were constricted by PE and dilated by ACh were included. The vessel pressure was maintained at a constant 60 mmHg throughout the experiments. The outer arterial diameter was monitored using a Cainda WiFi Digital Microscope and analyzed with ImageJ (Schneider et al., 2012). To change solutions, the myograph bath solution was completely vacuum aspirated and replaced with a new solution from the top. Relaxation percentage was calculated by the formula: $% relaxation=D−DPE/DAch−DPE*100%$⁠, where D is the diameter of the vessel, D_PE is the full constricted diameter of the vessel in response to 10 μM PE, D_Ach is the fully dilated diameter of the vessel in response to 10 µM ACh, and D_basal is the diameter of vessel at the basal level.

Aortic vSMCs were isolated from Acta2-GCaMP8.1/mVermilion mice. Aortic vSMCs were chosen because of their abundance and ease to culture. GCaMP8.1/mVermilion mice express the GCaMP8.1/mVermilion fusion protein in vSMCs which allows us to monitor the Ca²⁺ dynamics in response to different reagents. Isolation of murine aortic smooth cells was performed as described in detail previously (Adhikari et al., 2015). Briefly, after euthanizing the mouse, the aorta was rapidly isolated and transferred to a 6-well plate with complete DMEM solution (10% FBS, 1% penicillin/streptomycin) and amphotericin B (1 μl/ml). The remaining blood was removed by perfusing the aorta using a tuberculin syringe filled with sterile PBS. The surrounding fat tissues were cleaned with fine-tipped forceps. The dissected aorta was then cut into 1–2 mm pieces using fine scissors and placed into 5 ml tubes containing 0.1 ml of complete DMEM with collagenase type 2 (1.42 mg/ml). The aorta was digested at 37°C (with 5 % CO₂) for 6 h. After a 6-h incubation, 3 ml of DMEM plus 10% FBS was added and gently mixed. Cells were centrifuged at 300 × g for 5 min at room temperature, the medium aspirated, and the pellet washed with 3 ml of DMEM + 10% FBS twice. Cells were incubated at 37°C with 5% carbon dioxide, undisturbed for 5 days until confluence. The cells were subcultured by trypsinization (0.25%) for 3–4 min and seeded on poly-L-lysine-coated cover glasses for Ca²⁺ imaging or electrophysiological experiments.

Cultured vSMCs were reseeded on poly-L-lysine-treated cover glasses. After 24–48 h, the cells were placed into HEPES-PSS solution (in mM: 125 NaCl, 3.8 KCl, 1.2 CaCl₂, 25 NaHCO₃, 1.2 KH₂PO₄, 1.5 MgSO₄, 0.02 EDTA, and 8 D-glucose, pH 7.4). The GCaMP8 signal was monitored using the GFP filter set and a 40× objective on an Olympus IX73 inverted epifluorescence microscope. Time-lapse images were acquired by a Prime 95B Scientific CMOS Camera (Photometrics) controlled by MetaFluor software (Molecular Devices).

Untransfected or TMEM16A-transfected HEK293T cells were patch-clamped under whole-cell configuration and held at −60 mV with 500 nM free Ca²⁺ in the pipette solution. Images were taken every 2 s before and after the application of 5 µM niclosamide using an Olympus IX73 inverted microscope. Cell volume was quantified by the cell area measurement function in ImageJ.

The generation and validation of the TMEM16A stable cell line and TMEM16F-deficient (knockout [KO]) HEK293T cell lines have been reported in our recent studies (Huang et al., 2012; Le et al., 2019b; Zhang et al., 2020; Liang and Yang, 2021; Zhang et al., 2022). All the TMEM16A mutants were transiently expressed in TMEM16F-KO HEK293T cells to avoid interference from endogenous TMEM16F. HEK293T cells were cultured with Dulbecco’s modified Eagle’s medium (DMEM; 11995-065; Gibco BRL) supplemented with 10% FBS (F2442; Sigma-Aldrich) and 1% penicillin–streptomycin (15-140-122; Gibco BRL). All cells were cultured in a humidified atmosphere with 5% CO₂ at 37°C.

The murine TMEM16A coding sequence corresponding to the “a” splicing isoform (cDNA 30547439; Open Biosystems) was subcloned in the pEGFP-N1 vector, resulting in eGFP tags on the C-terminus. Single-point mutations of TMEM16A were generated using QuikChange site-directed mutagenesis and the plasmids were subsequently sequenced. The plasmids were transiently transfected to TMEM16F-KO HEK293T cells using X-tremeGENE9 transfection reagent (Sigma-Aldrich). Cells grew on coverslips coated with poly-L-lysine (Sigma-Aldrich). The medium was changed 6 h after transfection with regular medium (11995-065; Gibco BRL). Experiments proceeded 24–48 h after transfection. Experiments performed in the Hartzell lab (Fig. 2) used the TMEM16A (ac) isoform as previously described (Xiao et al., 2011).

Primers for QuikChange mutagenesis are listed in Table 1.

To gain insights into the niclosamide binding site on TMEM16A, we performed docking experiments using the Schrodinger Suite (Schrödinger Release 2022-2: Schrödinger LLC, 2021). Niclosamide (5-chloro-N-(2-chloro-4-nitrophenyl)-2-hydroxybenzamide, PUBCHEM CID 4477) was prepared using Schrodinger LigPrep with the OPLS4 force field (Harder et al., 2016). Ligand tautomeric and ionization and states were generated at pH 7 ± 1 using Epik (Shelley et al., 2007; Greenwood et al., 2010). TMEM16A (PDB accession no. 5OYB) was prepared for docking using the Schrodinger Protein Preparation Wizard in a pH 7.4 environment (Sastry et al., 2013). N- and C-termini were capped with N-acetyl and N-methyl amide, and the missing side chains were added using Prime (Jacobson et al., 2002, 2004). H-bond assignments were optimized using PROPKA at pH 7.4 followed by a restrained molecular dynamics minimalization (Olsson et al., 2011; Søndergaard et al., 2011). The niclosamide ligand was docked onto the TMEM16A receptor using Glide. The ligand was docked within a 47-nm³ receptor grid that was centered on the transmembrane region of one protomer of the TMEM16A dimer (Friesner et al., 2004, 2006; Halgren et al., 2004). Induced fit docking was performed using the Induced Fit Protocol (Farid et al., 2006; Sherman et al., 2006a, 2006b).

All currents were recorded in either inside-out, outside-out, or whole-cell configurations 24–48 h after transfection using an Axopatch 200B amplifier (Molecular Devices) and the pClamp software package (Molecular Devices). Glass pipettes were pulled from borosilicate capillaries (Sutter Instruments) and fire-polished using a microforge (Narishge) to reach a resistance of 2–3 MΩ.

For the inside-out patch, the pipette solution (external) contained (in mM) 140 NaCl, 10 HEPES, and 1 MgCl₂, adjusted to pH 7.3 (NaOH), and the bath solution contained 140 NaCl, 10 HEPES, and 5 EGTA, adjusted to pH 7.3 (NaOH). Intracellular (perfusion) solutions with 0.387 μM free Ca²⁺ were made by adding CaCl₂ to a solution containing 140 mM NaCl, 10 mM HEPES, 5 mM EGTA, and the amount of CaCl₂ added was calculated using WEBMAXC (Bers et al., 2010) to achieve desired free Ca²⁺. For Ca²⁺ sensitivity measurement, 0 or 5 µM niclosamide was included in the pipette solution. The membrane was held at −60 mV, and TMEM16A’s inward currents were elicited by perfusion of EGTA-buffered Ca²⁺ solutions of various free Ca²⁺ concentrations. Steady-state currents were measured for each Ca²⁺ application and normalized to the peak current elicited by 100 μM Ca²⁺ to construct the dose-dependent curve.

For the outside-out patch, the pipette solution (internal) contained (in mM) 140 NaCl, 10 HEPES, and 5 EGTA, and the amount of CaCl₂ added was calculated using WEBMAXC (Bers et al., 2010) to achieve 200 nM free Ca²⁺. The bath solution contained 140 NaCl, 10 HEPES, and 5 EGTA, adjusted to pH 7.3 (NaOH).

For whole-cell recordings on HEK293T cells, the pipette solution (internal) contained (in mM) 140 CsCl, 1 MgCl₂, and 10 HEPES, plus CaCl₂ to obtain the desired free Ca²⁺ concentration using WEBMAXC. For whole-cell recordings on primary vSMC cells from Acta2-GCaMP8.1/mVermilion mice, the pipette solution (internal) contained (in mM) 140 TEA-Cl, 1 MgCl₂, and 10 HEPES, plus CaCl₂ to obtain the desired free Ca²⁺ concentration using WEBMAXC. The bath solution contained 140 NaCl, 10 HEPES, and 5 EGTA, adjusted to pH 7.3 (NaOH).

Procedures for solution application with/without niclosamide were described previously (Le et al., 2019a; Le and Yang, 2020; Liang and Yang, 2021). Briefly, a perfusion manifold with a 100–200-µm tip was packed with eight PE10 tubes. Each tube was under separate valve control (ALA-VM8; ALA Scientific Instruments), and the solution was applied from only one PE10 tube at a time onto the excised patches or whole-cell clamped cells. All experiments were at room temperature (22–25°C). All the chemicals for solution preparation were obtained from Sigma-Aldrich including niclosamide.

For binding site mutational analysis, the currents were generated by a gap-free protocol held at −60 mV in response to desired niclosamide concentrations. Right after the gap-free protocol, a voltage-step protocol was applied from −100 mV to +140 mV in the presence of the highest niclosamide concentration (10 µM). All the currents at −60 mV in response to different niclosamide concentrations were then normalized to the maximum tail current at +140 mV in the presence of 10 µM niclosamide (G_Max).

In the conduct of this study, randomization and blinding procedures were not implemented. This decision was informed by the specific nature of our experimental design, which involved the utilization of an overexpressing system and smooth muscle cells (SMCs) derived from the mice expressing GCaMP8 specifically expressed in SMCs. These systems necessitated the use of a fluorescence marker to identify cells expressing TMEM16A, rendering the randomization and blinding processes unnecessary for this investigation. Despite the absence of these standard protocols, meticulous attention was dedicated to ensuring the accuracy and reliability of our experimental procedures and subsequent data analysis.

Video 1 shows that a patch-clamped HEK293T cell overexpressing TMEM16A did not change cell volume at the holding potential of −60 mV with 500 nM intracellular Ca²⁺. Video 2 shows that niclosamide (5 µM) induced dehydration and volume reduction of a patch-clamped HEK293T cell with TMEM16A overexpression. Video 3 shows niclosamide (5 µM) cannot induce dehydration and volume reduction of a patch-clamped HEK293T cell without TMEM16A overexpression. Video 4 shows niclosamide (5 µM) increases intracellular Ca²⁺ in the primary aortic SMCs from the Acta2-GCaMP8.1/mVermilion mice. Ca²⁺ increase was suppressed by 10 μM Ani9, a TMEM16A inhibitor.

To elucidate the effects of niclosamide on TMEM16A current, we conducted whole-cell patch clamp recording of HEK293T cells overexpressing TMEM16A (Fig. 1). With a physiological intracellular Ca²⁺ concentration of 0.5 µM and a −60 mV holding voltage, exogenous TMEM16A gave rise to characteristic outwardly rectifying currents with slow activation and deactivation kinetics (Fig. 1 B). At a holding potential of −60 mV, TMEM16A current was potentiated at both positive and negative voltages in the presence of 5 µM niclosamide (Fig. 1 A), but the inward current was potentiated to a greater extent than the outward current so that the G-V relationship became nearly linear, unlike the outward rectification typically observed at this Ca²⁺ concentration (Fig. 1, B, D, and E). These changes resemble the TMEM16A current at saturating levels of Ca²⁺ (Xiao et al., 2011; Huang et al., 2012). These effects were observed in the authors’ two laboratories independently using different TMEM16A splicing variants (see Materials and methods for details), ruling out potential systematic error (Fig. 2). Our analysis of the G-V relationships normalized to the maximum tail current amplitude at +140 mV before niclosamide application further demonstrates the potentiation (Fig. 1 C). The potentiation effect of niclosamide is dose-dependent with an EC₅₀ (half-maximal effective concentration) of 1.27 ± 0.02 µM (Fig. 1, E and F) and is completely reversible (Fig. 1, A–C). In addition, niclosamide-induced potentiation can be completely abolished by Ani9 (Fig. 1, G–I), a specific TMEM16A inhibitor (Seo et al., 2016). Our whole-cell patch clamp experiments thus demonstrate that niclosamide potentiates exogenous TMEM16A CaCC under physiological Ca²⁺ and voltage.

In the whole-cell patch clamp experiments, 5 mM EGTA was included in the extracellular solution to rule out the possibility that niclosamide-induced TMEM16A potentiation results from Ca²⁺ influx from the extracellular solution. Furthermore, our Fura2 Ca²⁺ imaging results conclusively demonstrated that niclosamide did not elicit any increase in intracellular Ca²⁺ levels, eliminating the possibility that niclosamide triggered Ca²⁺ release from internal stores (Fig. 1 J). Hence, it is reasonable to assert that the observed potentiation of TMEM16A by niclosamide is more likely attributed to a direct impact on channel activation, as opposed to any indirect influence stemming from Ca²⁺ dynamics.

Because Ca²⁺-activated Cl⁻ efflux via TMEM16A is a potent driving force for water flux and cell volume regulation (Almaça et al., 2009; Takayama et al., 2015), we examined niclosamide’s effect on cell volume. In the absence of niclosamide, no obvious cell volume change occurred during 2 min of whole-cell recording in HEK293T cells overexpressing TMEM16A (Fig. 3 A and Video 1). This is consistent with little or no opening of TMEM16A under conditions of 0.5 µM intracellular Ca²⁺, symmetrical Cl⁻, and a holding voltage of −60 mV (Fig. 1, B and C). On the other hand, 5 µM niclosamide induced dramatic cell shrinkage (Fig. 3 B and Video 2). The niclosamide-induced shrinkage is likely due to its potentiation effect on TMEM16A, which promotes Cl⁻ and water efflux at −60 mV. In stark contrast, 5 µM niclosamide had no obvious effect on the volume of untransfected HEK293T cells lacking TMEM16A overexpression (Fig. 3 C and Video 3). Our cell volume experiments thus further support that niclosamide potentiates TMEM16A.

Niclosamide has been reported to exhibit multiple off-target effects such as interfering with intracellular Ca²⁺ signaling and inhibiting G protein-coupled receptors (GPCRs) (De Filippo et al., 2017; Genovese et al., 2023). To minimize the effects of niclosamide on intracellular signaling, we conducted excised patch clamp experiments under outside-out and inside-out configurations. Consistent with our whole-cell patch clamp recording, the extracellular application of 5 µM niclosamide to outside-out patches markedly potentiates TMEM16A activation and its effect is reversible (Fig. 4, A and B). This further supports our whole-cell patch clamp findings that niclosamide-mediated potentiation is not through Ca²⁺ store release. In contrast, intracellular application of 5 µM niclosamide to inside-out patches has no effect on TMEM16A current (Fig. 4, C and D). In addition, we measured the apparent Ca²⁺ sensitivity of TMEM16A with and without niclosamide with various intracellular Ca²⁺ concentrations (Fig. 4, E–H). We found that niclosamide significantly increased TMEM16A sensitivity to Ca²⁺. Our excised patch clamp experiments thus demonstrate that niclosamide directly potentiates TMEM16A from the extracellular side but not the intracellular side.

To dissect the molecular underpinning of niclosamide-mediated TMEM16A potentiation, we sought to identify the niclosamide binding site on the extracellular side of the channel. Through the combination of molecular docking and all-atom molecular dynamics (MD) simulation (see Materials and methods), we identified an extracellular site that favors niclosamide binding (Fig. 5 A). The highest-scoring pose has a docking score of −8.837. The binding site is a highly hydrophobic pocket near the extracellular side of the membrane that is lined with hydrophobic and positively charged amino acids from transmembrane helices (TM) 5, 9, and 10. Niclosamide is predicted to have strong interactions with R605 and F601 located at the extracellular end of the TM5. The side chains of these residues face toward the dimer interface and away from the ion permeation pathway (Fig. 5 A). The NH₂ group of R605 coordinates the niclosamide via a salt bridge contact with the carboxy oxygen and by hydrogen bonds with the carbonyl oxygen of the niclosamide hydroxybenamide moiety. F601 makes a pi–pi stack with the hydroxybenzamide ring. F781 on TM9 makes a pi–pi stack with the niclosamide nitrophenyl group, whereas the NH₂ group of R864 forms a salt bridge with the nitro group of niclosamide.

To test the significance of these potential interactions, we substituted alanine for each of these residues and measured the apparent sensitivity of these mutants to niclosamide using a whole-cell patch clamp (Fig. 5, B–E). For the wildtype (WT) channel at −60 mV, niclosamide dose-dependently potentiates TMEM16A current with an EC₅₀ of 1.54 ± 0.46 µM (Fig. 5 D). The lowest concentration of niclosamide to produce obvious channel activation occurs at ∼0.1 µM (Fig. 5 B). The F601A, R605A, F781A, and R864A mutations significantly attenuate niclosamide potentiation (Fig. 5, B–E). They increase the minimum niclosamide concentration for TMEM16A activation by >10-fold (>1 µM, Fig. 5 B) and produce a rightward shift the niclosamide dose–response curves at −60 mV with EC₅₀’s of 7.19 ± 0.66 µM, 5.66 ± 0.91 µM, 3.07 ± 0.75, and 13.17 ± 2.46 µM, respectively (Fig. 5 D). The WT TMEM16A current in the presence of 5 µM niclosamide is ∼70% of G_{(+140mV/10 μM Niclo)}, whereas this percentage was decreased significantly to 20–40% for the single alanine mutations (Fig. 5, D and E). The F601A–R605A double mutant channel shows the most rightward-shifted niclosamide dose–response curve with an EC₅₀ of 79.60 ± 3.67 µM (Fig. 5 D) and a potentiation of only ∼10% of G_{(+140 mV/10 μM Niclo)}, suggesting that the effects of mutations on these two residues are synergistic. Potentiation by niclosamide is accompanied by a significant decrease in outward rectification, which is a characteristic of TMEM16A current observed with sub-micromolar intracellular Ca²⁺ (Fig. 1 B and Fig. 5 C). The observation that both the single and double alanine mutations largely preserve TMEM16A outward rectification in the presence of niclosamide (Fig. 5 C), further supports the conclusion that the alanine mutations of the putative niclosamide binding residues attenuate niclosamide-mediated TMEM16A potentiation.

Furthermore, mutating F601 or R605 to tryptophan, an amino acid with a bulkier sidechain, also dramatically attenuated niclosamide-induced potentiation as evidenced by insensitivity to niclosamide <1 µM (Fig. 6 A), preservation of current outward rectification (Fig. 6 B), and the rightward-shift of the dose–response curves (Fig. 6, C and D). Interestingly, the tryptophan mutations are more effective than the corresponding alanine mutations in reducing niclosamide potentiation, suggesting that the niclosamide potentiation effect is highly sensitive to the environment of the putative binding site. The attenuation effect of the charge reversal mutation R605E also supports this conclusion (Fig. 6).

The putative binding site mutations have negligible effects on the voltage-dependent activation of TMEM16A, as evidenced by the identical G-V relationship and the similar half-activation voltage (Fig. 7). This indicates that the effect of the mutations on niclosamide potentiation is not due to their influence on channel activation. Taken together, our computational and functional tests identified a putative niclosamide binding site on the extracellular side of TMEM16A. Residues F601, R605, F781, and R864 within the putative binding site provide potential coordinates for niclosamide binding.

TMEM16A is abundantly expressed in vSMCs and plays a crucial role in regulating their membrane potential, which regulates the activation of voltage-gated calcium channels (VGCCs) and subsequent vSMC contraction and vascular tone (Huang et al., 2009, 2012; Hawn et al., 2021; Wray et al., 2021). To examine the niclosamide effect on endogenous TMEM16A, we studied aortic vSMCs derived from Acta2-GCaMP8.1/mVermilion mice, which express the genetically encoded Ca²⁺ sensor GCaMP8.1 under the control of the Acta2 locus promotor (Lee et al., 2021). The GCaMP8.1 and mVermilion positive vSMCs were recorded under whole-cell configuration with 140 mM TEA-Cl to block K⁺ channels. Endogenous CaCC was activated by 0.5 μM Ca²⁺ in the internal (pipette) solution. Under this condition, we recorded a CaCC current with outward rectification that is a characteristic of TMEM16A (Fig. 8, A and B). This current was greatly potentiated by 5 μM extracellular niclosamide, and the potentiation effect can be completely suppressed by the TMEM16A inhibitor Ani9 (Fig. 8, A–D).

To further examine the impact of niclosamide-mediated potentiation of endogenous TMEM16A on vSMC function, we conducted Ca²⁺ imaging experiments on vSMCs from Acta2-GCaMP8.1/mVermilion mice (Fig. 8, E–G; and Video 4). We found that 5 μM niclosamide triggered a robust increase in intracellular Ca²⁺. Because this acute Ca²⁺ increase was suppressed by Ani9, we concluded that niclosamide stimulates vSMC TMEM16A, mediating Cl⁻ efflux, which presumably leads to membrane depolarization, VGCC activation, and subsequent increases in intracellular Ca²⁺ (Bulley and Jaggar, 2014; Leblanc et al., 2015; Jackson, 2021; Wray et al., 2021).

Our finding that niclosamide elevates cytosolic Ca²⁺ in vSMCs suggests that it could induce vasoconstriction. To test this hypothesis, we used an ex vivo pressure myograph to quantify the effect of niclosamide on the contractility of the third-order mesenteric arteries from C57BL/6J mice (Fig. 9). We found that niclosamide acutely induced a significant constriction of the equilibrated, non-treated vessels (basal). Niclosamide-induced vasoconstriction can be readily reversed by Ani9. Subsequent application of ACh and PE induced normal vasodilation and vasoconstriction, demonstrating that TMEM16A potentiation by niclosamide and inhibition by Ani9 does not alter the response of the vessel to the vasodilators and vasoconstrictors. Our pressure myograph experiment thus supports the conclusion that niclosamide rapidly potentiates TMEM16A CaCC and triggers Ca²⁺ increases in vSMCs (Fig. 8), resulting in acute vasoconstriction (Fig. 9).

This study provides evidence at molecular, cellular, and tissue levels that niclosamide is a potent TMEM16A potentiator under physiological Ca²⁺ and voltage. Niclosamide potentiates TMEM16A from the extracellular side. In vSMCs expressing endogenous TMEM16A, niclosamide stimulates the TMEM16A, which increases cytosolic Ca²⁺ and produces vasoconstriction. Consistent with a recent report (Danahay et al., 2023), we find no indication that niclosamide inhibits TMEM16A CaCC at physiological Ca²⁺ concentrations and voltages.

We have examined the effect of niclosamide in both whole-cell and excised patch clamp configurations. First, our whole-cell patch clamp recording under well-buffered physiological calcium concentration shows that extracellular niclosamide potentiates TMEM16A in an acute and reversible fashion (Fig. 1). Second, we observed dramatic cell shrinkage when TMEM16A-expressing cells were exposed to niclosamide (Fig. 3), indicating that niclosamide-induced TMEM16A potentiation drives cell dehydration through TMEM16A-mediated chloride efflux. Third, our Ca²⁺-imaging (Fig. 1 J) and excised patch clamp experiments (Fig. 4) explicitly demonstrate that niclosamide potentiates TMEM16A from the extracellular side without involving Ca²⁺ release from internal stores. Fourth, mutating the putative niclosamide binding residues, F601, R605, F781, and R864, significantly attenuates niclosamide’s potentiation effect (Figs. 5 and 6), further supporting that niclosamide directly acts on TMEM16A to potentiate the channel. Fifth, extracellular niclosamide acutely potentiates endogenous TMEM16A CaCC in vSMCs, inducing intracellular Ca²⁺ elevation and vasoconstriction (Figs. 8 and 9).

The identification of the putative niclosamide binding site also sheds new light on the allosteric regulation of TMEM16A gating. Previously, we and others found that Ca²⁺ binding to amino acids located adjacent to the pore region serves as a primary stimulus to promote channel opening (Tien et al., 2014; Yu et al., 2014; Paulino et al., 2017). Ca²⁺ binding to another Ca²⁺ binding site, located at the intracellular side of TM2 and TM10 near the dimer interface, allosterically promotes TMEM16A gating (Le and Yang, 2020). In addition, phosphatidylinositol 4,5-bisphosphate (PI(4,5)P₂), by binding to various locations of TMEM16A at the membrane-cytosol interface, also allosterically regulates TMEM16A opening, presumably by stabilizing the open state (Le et al., 2019a; Tembo et al., 2019; Yu et al., 2019; Ko et al., 2020; Jia and Chen, 2021). Distinct from the Ca²⁺ and PI(4,5)P₂ binding sites located toward the intracellular side of the transmembrane region, the putative niclosamide binding pocket uniquely resides at the extracellular side (Fig. 5 A). It is close to the dimer cavity but slightly away from the ion permeation pathway. Niclosamide binding to this site likely triggers allosteric arrangements of the channel gate, thereby lowering the free energy required for channel opening. Future studies are needed to dissect the mechanism underlying this drug-induced TMEM16A potentiation from an extracellular site. Additionally, the extracellular localization of the niclosamide binding pocket offers a promising avenue for in silico drug screening. This approach could pave the way for identifying novel allosteric modulators of TMEM16A, expanding our tools for modulating this crucial channel.

Our data contrast with some prior studies suggesting that niclosamide acts as a TMEM16A antagonist (Miner et al., 2019; Henckels et al., 2020). The discrepancy might be derived from the differences in the functional assays used to evaluate the niclosamide effect. Both studies (Miner et al., 2019; Henckels et al., 2020) used high-throughput screening assays, either imaging-based high-throughput screening and/or high-throughput automated Q-patch. However, our current study used the gold standard of human-controlled patch clamp with both whole-cell and outside-out configurations. A recent study of different TMEM16A isoforms (Danahay et al., 2023) and our independent patch clamp characterizations in two laboratories with two different TMEM16A isoforms rule out the possibility that the reported inhibitory effects of niclosamide are derived from different splicing isoforms.

A plausible explanation for the observed inhibitory effect is likely due to the protocol used to activate TMEM16A as well as niclosamide’s off-target effects, including GPCRs, ion channels, calcium pumps, and mitochondrial proteins (Chen et al., 2009; Jurgeit et al., 2012; Danahay et al., 2023). Indeed, different from our current study that used well-controlled intracellular Ca²⁺ to directly elicit TMEM16A current, previous patch-clamp experiments showing an inhibitory effect of niclosamide effect relied on UTP to indirectly increase intracellular Ca²⁺ and subsequently activate TMEM16A (Cabrita et al., 2019). This indirect approach to activate TMEM16A is susceptible to the non-specific effects of niclosamide, as evidenced by the recent publication showing that niclosamide indeed markedly attenuates UTP-stimulated calcium increase (Danahay et al., 2023; Genovese et al., 2023).

Previous studies also showed inconsistent niclosamide effects at the tissue level (Li et al., 2017; Miner et al., 2019; Danahay et al., 2020, 2023). The discrepancy may also derive from niclosamide’s broad molecular targets. Although niclosamide was shown to relax precontracted human bronchial rings (Miner et al., 2019; Danahay et al., 2020), it paradoxically constricts human pulmonary arteries (Danahay et al., 2020), consistent with the vasoconstriction effect in murine mesenteric arteries observed in our study (Fig. 8). It is unclear why the airway and the arteries show opposite responses to niclosamide. Given the broad targets of niclosamide, it is likely that niclosamide may target different sets of proteins in the two different types of SMCs.

In conclusion, our current study provides compelling evidence to show that niclosamide directly potentiates TMEM16A by binding to a putative extracellular site under physiological Ca²⁺ and voltage. By potentiating TMEM16A, niclosamide depolarizes and acutely induces a Ca²⁺ increase in vSMCs, promoting vasoconstriction. Based on this study and previous investigations, niclosamide is a compound with multifaceted effects on various protein targets. When attempting to repurpose niclosamide as a TMEM16A antagonist to treat asthma, COPD, and hypertension, its acute potentiation effect on TMEM16 CaCCs and its broad impacts on different targets need to be carefully evaluated to avoid triggering severe side effects in human patients.

All study data are included in the article and/or its online supplementary files. The raw data are shared upon reasonable request.

Jeanne M. Nerbonne served as editor.

This research was supported by awards from the National Institute of General Medical Sciences of the National Institutes of Health, DP2-GM126898 and R35-GM153196 (awarded to H. Yang), and GMR01-132598 (awarded to H.C. Hartzell), and American Heart Association postdoctoral fellowship to P. Liang (#903807).

Author contributions: P. Liang: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Writing—original draft, Writing—review & editing, Y.C.S. Wan: Formal analysis, Investigation, K. Yu: Data curation, Formal analysis, H.C. Hartzell: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing—original draft, Writing—review & editing, H. Yang: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Visualization, Writing—original draft, Writing—review & editing