Edit · 2026 01 11 182135 Nature S41589 025 01841 3

Saving needs a token. Log in or register to get one. Edits are public and licensed CC-BY-SA-4.0.

# Persistent activation of TRPM4 triggers necrotic cell death characterized by sodium overload | Nature Chemical Biology

**URL:** https://www.nature.com/articles/s41589-025-01841-3

**Saved:** 2026-01-11T02:21:41.032Z

---

Article
Published: 06 February 2025
Persistent activation of TRPM4 triggers necrotic cell death characterized by sodium overload
Wan Fu, Jianghuang Wang, Tianyu Li, Yuhui Qiao, Zili Zhang, Xiaomin Zhang, Mingkai He, Yan Su, Ziye Zhao, Chen Li, Ronghua Xiao, Yujun Han, Shen Zhang, Zhiqiang Liu, James Lin, Guoqiang Chen, Yang Li & Qing Zhong

Nature Chemical Biology 
volume
 21, 
pages
1238–1249 (2025)Cite this article

11k Accesses

34 Citations

9 Altmetric

Metrics
details

Abstract

Sodium influx and overload are frequently observed in human tissue injuries. Whether sodium overload imposes a causative effect on necrotic cell death and the mechanism involved are unclear. Here we identify necrocide 1 (NC1) as a compound that induces necrotic cell death through sodium overload, termed NECSO for necrosis by sodium overload. NC1 targets the transient receptor potential cation channel subfamily M member 4 (TRPM4), a nonselective monovalent cation channel, to promote Na+ influx and necrosis. TRPM4-deficient cells are resistant to NC1-induced NECSO. NC1 specifically activates human TRPM4, not mouse TRPM4, because of differences in a transmembrane region, as revealed by domain swapping and molecular docking. Gain-of-function mutations in human TRPM4 linked to cardiac arrhythmias show increased vulnerability to NECSO triggered by NC1 or 2-deoxy-D-glucose. Chemical screening identified NECSO inhibitors that block necrosis induced by NC1 or energy depletion. These findings provide insights into regulated Na+ influx-mediated necrosis and its implications for disease.

You have full access to this article via your institution.

Download PDF
Similar content being viewed by others
Sodium disrupts mitochondrial energy metabolism to execute NECSO
Article Open access
13 December 2025
Cardiac function is regulated by the sodium-dependent inhibition of the sodium-calcium exchanger NCX1
Article Open access
07 May 2024
Neocarzilin A induces apoptosis and mitochondrial disturbance by targeting reticulon 4-mediated endoplasmic reticulum stress
Article Open access
16 June 2025
Main

Excessive sodium contributes to severe human diseases such as hypertension, cardiovascular and kidney diseases1,2, brain edema3 and diabetes4. Dysregulated sodium influx disrupts osmotic balance, causing tissue edema5 linked to conditions such as tissue injury6, stroke7 and tumors8. Whether sodium overload directly triggers cell death, as seen with calcium, iron or copper9,10,11, remains unclear.

Sodium concentration is tightly regulated by Na+–K+ adenosine triphosphatase (ATPase) and ion channels, maintaining low intracellular and high extracellular levels12. Sodium influx is essential for membrane potential, action potential propagation and cell communication, regulated by various factors such as ligands and signaling pathways13,14.

We previously identified necrocide 1 (NC1) as a potent inducer of regulated necrosis distinct from known cell death pathways15. Here, we found that NC1 triggers massive sodium influx, rapid membrane depolarization and cell edema. CRISPR–Cas9 screening revealed transient receptor potential cation channel subfamily M member 4 (TRPM4), a nonselective cation channel, as NC1’s target, inducing necrosis through sodium overload. The manipulation of necrosis by sodium overload (NECSO) may benefit in protecting cells or tissues from energy-depletion-induced injuries.

Results
TRPM4 is essential to NC1-induced necrotic cell death

We previously described a lethal chiral compound, NC1 (Extended Data Fig. 1a), which induces robust immunogenic necrotic cell death in specific cancer cell lines through unknown mechanisms15 (Extended Data Fig. 1a–c and Supplementary Video 1). Attempts to suppress NC1-induced necrosis using inhibitors of apoptosis (Z-VAD), necroptosis (Nec-1), autophagy (CQ and 3-MA) or ferroptosis (Fer-1, Liprox-1 and DFO) were unsuccessful (Extended Data Fig. 1d). Consistent with previous observations15, these data support that NC1 induces a distinct cell death pathway.

To uncover its molecular mechanism, we conducted a genome-wide CRISPR–Cas9 screen in NC1-sensitive MCF7 cells stably expressing Cas9 (Supplementary Fig. 1a,b). High-quality single guide RNA (sgRNA) transfection was confirmed through deep sequencing, with 99.9% of sgRNAs successfully transfected (Supplementary Fig. 1c). Screening revealed key genes for NC1 resistance in the NC1-resistant cell pool (Supplementary Fig. 1d), analyzed using the MAGeCKFlute pipeline16. TRPM4 emerged as the top candidate, consistently showing extraordinary enrichment scores (Fig. 1a).

Fig. 1: NC1 triggers a distinct cell death pathway dependent on TRPM4.

a, Beta scores and false discovery rate (FDR) of sgRNA-targeted genes analyzed by the MAGeCKFlute-RRA of NC1 treatment and control in the mentioned CRISPR–Cas9 screening. The horizontal dotted line on the y axis corresponds to a value of 1.30 (−log10(0.05)). The cutoff for the x axis is set at ±3.03. b, Dose-dependent cytotoxicity of NC1 was assessed 24 h after NC1 treatment using the CTG assay in MCF7 cells (control and TRPM4 KO with three different sgRNAs). c, Cell death assay measured by LDH release in MCF7 TRPM4-KO cells treated with NC1 for 24 h. d, Dose-dependent cytotoxicity of NC1 in MCF7 TRPM4-KO cells with re-expression of WT TRPM4 or D984A mutant. e, CETSA of NC1 or NC1i in HEK293T cells stably expressing TRPM4–EGFP. The graph represents the relative TRPM4 band intensities of the indicated samples as the mean ± s.d. from three independent experiments. f, Purification with size exclusion chromatography of hTRPM4 (inset). Purified protein was loaded in dose-dependent concentrations (Conc.) and stained with Coomassie brilliant blue (CBB). NC1 binds to the hTRPM4 protein as shown by SPR measurements. g, The whole-cell current density–voltage relation of MCF7 and MCF7 TRPM4-KO cells activated by NC1. h, Statistics of hTRPM4 current density at +100 mV in g. i, The whole-cell current density–voltage relation of NC1 activating hTRPM4 stably expressed in HEK293T cells with duration recordings. Viability data represent the mean ± s.d. of n = 3 replicates from one representative of three independent experiments. Electrophysiological data represent the mean ± s.e.m. of (n = 6 for MCF7 cells and 4 for MCF7 TRPM4-KO cells in g,h; n = 7 for vehicle, 10 for NC1 5 and 10 min, 7 for NC1 15 min and 5 for NC1 washout in i) independent cells. Immunoblot data show one representative of three replicates from three independent experiments. Statistical analysis was performed using a one-way ANOVA (GraphPad Prism 9.0.0) in e and two-way ANOVA in h.

Source data

Full size image

To validate the effect of TRPM4, TRPM4-knockout (KO) MCF7 cells were generated using three gRNAs. TRPM4 deletion completely blocked NC1-induced cell death, as shown by cell viability and lactate dehydrogenase (LDH) release assays (Fig. 1b,c and Supplementary Fig. 2a). Reintroducing human TRPM4 (hTRPM4) restored NC1 sensitivity, while the TRPM4 inhibitor 9-phenanthrol17 dose-dependently alleviated NC1-induced cell death (Fig. 1d and Supplementary Fig. 2b,c). Furthermore, the ion-impermeable TRPM4 mutant D984A (ref. 18) failed to rescue the necrosis phenotype in TRPM4-KO cells, confirming that TRPM4’s ion channel activity is required for NC1-induced necrosis (Fig. 1d). These results establish that TRPM4 is essential for NC1-induced cell death.

NC1 directly binds to and activates TRPM4

We hypothesized that TRPM4 is the direct target of NC1, given its top ranking in our CRISPR screen. We firstly performed cellular thermal shift assay (CETSA) to analyze the interaction between NC1 and TRPM4 in HEK293T cells stably expressing TRPM4. NC1 stabilized TRPM4 at elevated temperatures, whereas DMSO or its inactive stereoisomer (NC1i) did not (Fig. 1e and Supplementary Fig. 2d). Surface plasmon resonance (SPR) analysis confirmed that NC1 directly binds to TRPM4 with an affinity of 0.88 μM (Fig. 1f and Supplementary Fig. 2e,f). These results suggest that NC1 specifically interacts with TRPM4 to induce cell death.

To investigate NC1’s effect on TRPM4, whole-cell patch-clamp recordings were conducted in HEK293T cells stably expressing TRPM4–EGFP. NC1 robustly induced TRPM4-dependent currents (half-maximal effective concentration (EC50) ≈ 306.3 nM), which were inhibited by the TRPM4 inhibitor 9-phenanthrol, confirming TRPM4 specificity (Extended Data Fig. 2a and Supplementary Fig. 2g,h). In wild-type (WT) MCF7 cells, NC1 elicited strong currents while TRPM4 KO blocked this effect (Fig. 1g,h). Additionally, NC1i failed to activate TRPM4 (Extended Data Fig. 2b,c). In a cell-free system, TRPM4 was gated upon NC1 exposure on the outside-out membrane patch (rather than the inside-out patch) (Extended Data Fig. 2d–g). Remarkably, NC1 triggered a sustained and increased current along with the time course and the current did not diminish even after washout, suggesting the continuous and irreversible activation of TRPM4 upon NC1 exposure (Fig. 1i and Extended Data Fig. 2h). These results indicate that NC1 directly binds to TRPM4 and triggers a persistent activation of TRPM4.

NC1-induced necrosis through TRPM4 is driven by sodium influx

TRPM4 is a calcium-activated, nonselective monovalent cation channel protein19,20. As most cations in vivo are metals, we analyzed major intracellular metal levels using inductively coupled plasma mass spectrometry (ICP-MS), including Na, K, Ca, Mg and Fe. Upon NC1 treatment, sodium levels increased while potassium decreased, whereas Ca, Mg and Fe levels remained stable, suggesting robust sodium influx and potassium efflux (Fig. 2a). Ion-sensitive probes confirmed these changes; CoroNa green fluorescence (sodium) increased and PBFI fluorescence (potassium) decreased in NC1-treated MCF7 cells (Fig. 2b,c). The above results indicate that sodium influx and potassium efflux occur with NC1-induced necrosis.

Fig. 2: Sodium influx is an essential event for NC1-induced cell death.

a, Sodium, potassium, calcium, iron and magnesium levels were assessed by ICP-MS in MCF7 cells treated with 50 nM NC1 for 0 (control cells), 1 and 2.5 h. The data are represented as the mean ± s.d. of triplicate samples. b, MCF7 cells were loaded with sodium indicator CoroNa (green) and potassium indicator PBFI (blue) and then treated with 50 nM NC1 for 0, 1 or 2 h. Scale bars, 50 μm. c, Statistics of fold change of CoroNa or PBFI fluorescence intensity. Data shown are the mean ± s.d. Each dot represents data from an individual image in b. d, Representative time-lapse images of MCF7 cells treated with 50 nM NC1 in different EBSs: normal, increased concentration of KCl to 25 mM, calcium-free and sodium-free (osmotic pressure was maintained by an equal concentration of NMDG-Cl). Scale bars, 25 μm. e, The cell viability of MCF7 cells treated with NC1 for 6 h in different EBSs described in d. f, Sodium and potassium levels were assessed by ICP-MS in MCF7 cells treated with 50 nM NC1 for 3 h in different EBSs (normal, sodium-free and 25 mM KCl). Results are plotted as the fold change over control. The data are represented as the mean ± s.d. of triplicate samples. g, Cell viability of MCF7 cells treated with NC1 for 6 h in EBSs containing varying concentrations of Na+ (osmotic pressure was maintained by an equal concentration of NMDG+). h, The cell viability of MCF7 cells treated with NC1 for 6 h in EBSs containing different monovalent cations (Na+ in EBS was replaced by an equal concentration of K+, Li+, Cs+ or Rb+). i, Intracellular sodium levels measured by CoroNa in MCF7 WT and TRPM4-KO cells treated with 50 nM NC1 for 0, 1 and 3 h. Viability data represent the mean ± s.d. of n = 3 replicates from one representative of three independent experiments. Micrograph and flow cytometry data show one representative of at least three replicates from three independent experiments. Statistical analysis was performed using a one-way ANOVA (GraphPad Prism 9.0.0) in a,c,f.

Source data

Full size image

To determine whether sodium influx or potassium efflux drives NC1-induced necrosis, we assessed its killing activity in isotonic extracellular buffer solution (EBS) devoid of serum and amino acids. The treatment was limited to 6 h to elicit NC1’s effect while minimizing starvation-induced cell death. NC1 exhibited similar cytotoxicity in EBS compared to normal culture medium and calcium depletion from EBS did not affect cell death (Fig. 2d,e). To investigate sodium’s role, NaCl in the medium was replaced with N-methyl-D-glucamine (NMDG)-Cl, a nonpermeable cation that maintains osmolarity. Sodium depletion significantly blocked NC1-induced cell death, indicating sodium influx as a critical factor (Fig. 2d,e). Conversely, raising extracellular potassium concentration from 5 mM to 25 mM to inhibit potassium efflux did not affect cell death, suggesting that potassium efflux is a concomitant not causative event (Fig. 2d,e). Intracellular concentrations of sodium and potassium were confirmed, with ICP-MS indicating the blockage effect of sodium influx and potassium efflux under the above conditions (Fig. 2f). Notably, Na+ depletion with NMDG-Cl blocked sodium influx and cell death without perturbation of potassium efflux (Fig. 2e,f). Potassium channel inhibitors targeting NKCC (Na+–K+–2Cl− cotransporter) and KATP (ATP-sensitive potassium channel) also had no impact on cell death (Extended Data Fig. 3a). To confirm the dependency on sodium influx, reducing sodium concentrations in EBS progressively decreased NC1’s killing activity (Fig. 2g). Substitution of sodium with other monovalent cations inhibited cell death to varying degrees (Fig. 2h), highlighting sodium influx as the primary factor. TRPM4 KO in NC1-treated MCF7 cells abolished sodium influx, confirming TRPM4’s role in mediating this effect (Fig. 2i). Collectively, these data suggest that NC1 triggers TRPM4-dependent necrotic cell death through the induction of sodium influx; therefore, we refer to this specific type of necrosis as necrosis by sodium overload (NECSO).

TRPM4 is a Ca2+-activated cation channel mediating cell membrane depolarization20. We hypothesized that NC1-induced sodium influx causes depolarization. Using a voltage clamp (Extended Data Fig. 3b,c) and the membrane potential probe DiBAC4(3) (Extended Data Fig. 3d,e), we observed membrane depolarization within 5–10 min of NC1 exposure. Because of electroneutrality, sodium influx is typically accompanied by chloride influx. Chloride levels, measured with the chloride quencher MQAE, increased upon NC1 treatment (Extended Data Fig. 3f,g). Additionally, inhibitors targeting anion channels or transporters effectively blocked NECSO to varying degrees (Extended Data Fig. 3h). Sodium overload also increases intracellular osmolarity, leading to water influx, cell swelling and edema. Transmission electron microscopy (TEM) revealed severe swelling of organelles, including the mitochondria, endoplasmic reticulum and Golgi, as well as membrane rupture, in NC1-treated cells (Extended Data Fig. 3i). Treatment with D-mannitol, a nonpermeant oncotic agent, increased extracellular osmolarity and delayed NECSO progression (Extended Data Fig. 3j). These findings indicate that membrane depolarization and osmotic imbalance-induced cell swelling are defining features of NECSO.

NC1 specifically targets hTRPM4

We investigated whether TRPM4 expression correlates with cellular sensitivity to NECSO. Among nine human cancer cell lines, MCF7, MB468, HCC1143 and PC3, which highly expressed TRPM4, were strongly sensitive to NC1. HEK293T, U-2OS and A549 cells, with moderate TRPM4 expression, showed intermediate sensitivity, while SW620 and HeLa cells, with negligible TRPM4 expression, were least sensitive (Fig. 3a,b and Extended Data Fig. 1b). Notably, mouse and hamster cells, including normal and cancer lines, were over 1,000-fold more resistant to NC1 than human cells, despite some mouse cells expressing TRPM4 at levels similar to several human cell lines (Fig. 3a and Extended Data Fig. 4a). This suggests species-specific differences in TRPM4 confer NC1 resistance.

Fig. 3: NC1 specifically targets hTRPM4 through interaction with the TM region.

a, TRPM4 protein expression profiles for human (left) and mouse (right) cell lines and corresponding heat map for sensitivity to NC1-induced cell death. b, Correlation of TRPM4 expression and NC1 sensitivity of human cell lines. Each dot represents one human cell line. TRPM4 band intensities of the indicated samples from a were analyzed using Image J software. IC50 values of NC1 were from Extended Data Fig. 1c. The linear regression was fitted with GraphPad Prism 9.0.0. c, Dose-dependent cytotoxicity assay of NC1 in TRPM4-KO MCF7 cells with re-expression of hTRPM4 or mTRPM4. d, CETSA of NC1 in HEK293T hTRPM4 or mTRPM4 cells. The graph represents the relative TRPM4 band intensities of the indicated samples as the mean ± s.d. from three independent experiments. e, The whole-cell current density–voltage relation of hTRPM4 or mTRPM4 expressed in HeLa cells. f, Schematic diagram of the validation system to verify whether mutant or chimera TRPM4 responds to NC1. g, Left: phylogenetic tree constructed by CLUSTALW multiple-sequence alignments for TRPM4 proteins (with accession numbers of protein sequences) from several species, namely, Homo sapiens (NP_060106.2), Macaca mulatta (XP_028695988.1), Canis lupus familiaris (XP_541500.3), Sus scrofa (XP_020950369.1), Mus musculus (NP_780339.2) and Ochotona princeps (XP_012785770.1) isoforms of TRPM4. Right: summary results of their corresponding sensitivity to NC1-induced channel activation or cell death. h, Phylogenetic tree of TRPM family protein and summary results of their corresponding sensitivity to NC1. i–k, TRPM4 chimeric constructs and their corresponding sensitivity to NC1. hTRPM4 and mTRPM4 are indicated by h and m, respectively. Chimeric constructs were generated by PCR cloning (i). Statistics showing the sensitivity to NC1-induced current densities of TRPM4 chimeras at +100 mV in whole-cell patch-clamp experiments using HeLa cells (j). NC1-induced cell death in cell cytotoxicity experiments using MCF7 TRPM4-KO cells respectively re-expressed with hTRPM4 or chimeric TRPM4 constructs as shown (k). Viability data represent the mean ± s.d. of n = 3 replicates from one representative of three independent experiments. Electrophysiological data represent the mean ± s.e.m. of (n = 6 for hTRPM4 in e; n = 4 for mTRPM4 in e; n = 7 in j) independent cells. Micrograph and immunoblot data show one representative of three replicates from three independent experiments. Statistical analysis was performed using a two-way ANOVA (GraphPad Prism 9.0.0) in j.

Source data

Full size image

To test whether NC1 specifically targets hTRPM4, we re-expressed hTRPM4 or mouse TRPM4 (mTRPM4) in TRPM4-KO MCF7 cells. Only hTRPM4 restored NC1 sensitivity (Fig. 3c). CETSAs confirmed that NC1 binds hTRPM4 but not mTRPM4 (Fig. 3d). Additionally, NC1 activated hTRPM4 channel activity but failed to activate mTRPM4 (Fig. 3e), demonstrating NC1’s specificity for hTRPM4.

We hypothesized that TRPM4 sequence differences among species account for their distinct sensitivities to NC1. To test this, we used HeLa cells, which lack detectable TRPM4 and are resistant to NC1. Transient expression of hTRPM4 restored NECSO, characterized by ‘balloon-like’ necrotic cells, upon NC1 treatment (Fig. 3f). This system allowed us to evaluate the role of TRPM4 variants in NECSO.

TRPM4 homologs from mouse and other species, unlike hTRPM4, showed no sensitivity to NC1-induced NECSO (Fig. 3g and Extended Data Fig. 4b–d). Additionally, other TRPM family members, including the closely related TRPM5, did not respond to NC1 (Fig. 3h and Supplementary Fig. 3a–d). These findings demonstrate that NC1 is a highly specific agonist for hTRPM4, enabling NECSO exclusively in cells expressing hTRPM4.

TRPM4 TM region determines NC1 response

TRP proteins from different species exhibit varied sensitivity to stimuli, aiding in identifying receptor domains involved in ligand binding or activation21,22. hTRPM4 and mTRPM4 share 88.4% sequence similarity and similar tetrameric structures, with conserved regulatory modifications such as phosphorylation, sumoylation and glycosylation23,24,25,26. However, their NC1 sensitivity differs sharply (Fig. 3c). We performed domain-swapping analysis to identify the structural domains required for NC1 sensitivity. Both hTRPM4 and mTRPM4 were divided into three fragments: N terminus (1–767 in human, 1–763 in mouse), transmembrane (TM) (768–1093 in human, 764–1089 in mouse) and C terminus (1094–1214 in human, 1090–1213 in mouse) (Fig. 3i and Extended Data Fig. 5a). The TM fragment of TRPM4 includes the pre-S1 (S0) domain, S1–S6 core TM domains and the TRP domain embedded in the lipid layer (Extended Data Fig. 5a). Swapping the hTRPM4 TM region with the mTRPM4 TM region blocked NC1 sensitivity (hN-mTM-hC), while replacing the N (mN-hTM-hC) or C (hN-hTM-mC) terminus retained sensitivity (Fig. 3i–k, Extended Data Fig. 5b,c and Supplementary Fig. 4a–d). Replacing the mTRPM4 TM region with the hTRPM4 TM (mN-hTM-mC) restored NC1 sensitivity (Fig. 3i–k and Supplementary Fig. 4a–d). NC1-sensitive chimeras fully rescued NECSO phenotypes in TRPM4-KO MCF7 cells (Fig. 3k), confirming the hTRPM4 TM region’s critical role in NC1 sensitivity.

Molecular docking identifies key residues for NC1 activation

Domain-swapping analysis highlighted the importance of the TM region of hTRPM4 in NC1 sensitivity. Molecular docking using the hTRPM4 structure (Protein Data Bank (PDB) 6BQV) revealed that NC1 binds within a cavity in the TM region. This cavity is surrounded by the S3 and S4 helices, the S4–S5 linker from one subunit and the S5 and S6 helices from an adjacent subunit. Known as the vanilloid-binding pocket (VBP), this region is a common ligand-gating site in TRP channels27 but is here reported as a TRPM4 ligand-binding site (Fig. 4a and Extended Data Figs. 6 and 7). Docking predicts that NC1 interacts with key residues in this interfacial cavity near the Ca²⁺-binding site, including F936 (S5), W864 (S3) and V904, L907 and H908 (S4) (Fig. 4b,c), suggesting its role in NC1 gating. H908 forms hydrogen bonds and π–π stacking with NC1’s indole and hydroxyphenyl groups, while W864 forms hydrogen bonds with the hydroxyphenyl moiety. V904, L907 and F936 stabilize NC1’s methylindoline group through hydrophobic interactions (Fig. 4b,c, Extended Data Figs. 6–8, Supplementary Video 2 and Supplementary Data 1). To confirm the importance of this pocket, we substituted key residues. Substitution of W864, H908 or F936 with alanine disrupted NC1 binding, TRPM4 channel activation and NECSO sensitivity (Fig. 4d–g). Substitution of M927 to arginine blocked NC1 access and abolished its effects, while M927A retained NC1 responsiveness (Fig. 4e,f). Electrophysiological analysis showed that these mutants (W864A, H908A and F936A), unresponsive to NC1, retained varying ion permeability compared to the nonfunctional mutant D984A (Fig. 4h). These findings confirm the predicted pocket’s essential role in NC1-induced TRPM4 activation.

Fig. 4: Docking and mutagenesis analysis reveal critical residues of TRPM4 for NC1 activation.

a, Surface representation showing the overall structure of the molecular modeling TRPM4–NC1 complex. MW, molecular weight. Inset: a close-up view of the NC1-binding site. Individual TRPM4 subunits are colored cyan or gray, whereas NC1 is highlighted with yellow. Lower right: chemical structure of NC1. b,c, Specific recognition of NC1-bound hTRPM4. A 3D diagram showing that NC1 binds a deep cavity of TRPM4 between S3, S4, S5′ and S6′ helix; Ca2+ is colored purple (b). A 2D diagram showing that NC1 interacts with the key residues of TRPM4 through a mixture of hydrogen bonds, hydrophobic interactions and π–π stacking (c). d, CETSA of NC1 and TRPM4 mutants of key residues in b,c. ΔTm = Tm(NC1) − Tm(DMSO). Melting temperatures (Tm) were calculated by TRPM4 band intensities of the indicated samples in western blot with GraphPad. The graph represents mean ± s.d. from three independent experiments. e, The current density–voltage relationship of hTRPM4 mutants involving key residues for NC1 binding was examined in HeLa cells. f, Summary plot showing the maximum amplitude of current densities activated by NC1 in the TRPM4 WT channel compared to the aforementioned mutants. g, Dose-dependent cytotoxicity assay of NC1 in TRPM4-KO MCF7 cells with re-expression of WT TRPM4 or TRPM4 mutants (W864A, H908A, F936A). h, Summary data showing the constitutive activity of hTRPM4 mutants (W864A, H908A, F936A) recorded from inside-out patches at −100 mV. WT TRPM4 was shown as positive control and D984A as negative control. Viability data represent the mean ± s.d. of n = 3 replicates from one representative of three independent experiments. Electrophysiological data represent the mean ± s.e.m. of (n = 6 for WT, H908A and W864A, 7 for F936A, 4 for M927A and 3 for M927R in e,f; n = 3 for WT, D984A and M927R, 4 for H908A, 5 for F936A and 6 for W864A in h) independent cells. Immunoblot data show one representative of three replicates from three independent experiments. Statistical analysis was performed using a one-way ANOVA (GraphPad Prism 9.0.0) for d,h and two-way ANOVA (GraphPad Prism 9.0.0) for f.

Source data

Full size image

Molecular dynamics (MD) simulations modeled the NC1–TRPM4 complex in a phospholipid bilayer (Extended Data Fig. 8a). The root-mean-square deviation (r.m.s.d.) stabilized under 5 Å within 100 ns, confirming a stable complex. Interaction energy analysis validated NC1’s stable binding to TRPM4 (Extended Data Fig. 8b–d). Binding energy contributions revealed key stabilizing residues: S863, W864, H908 and F936, consistent with molecular docking predictions (Extended Data Fig. 8e). These findings suggest that a VBP-like cavity participates in NC1 activation of TRPM4.

The TM region of TRPM4 includes six TM helices and a TRP helix. The S5–S6 helices and P-loop form the ion-conducting pore, surrounded by the S1–S4 domain, which contains a Ca²⁺-binding site. The S4–S5 linker connects these regions and is crucial for channel gating. While TRPM4 is impermeable to Ca²⁺, it is activated by intracellular Ca²⁺ (refs. 20,28). To determine whether NC1 affects channel activity through Ca²⁺, we tested its effects with varying EGTA (Ca²⁺ chelator) concentrations. NC1 activated TRPM4 independently of Ca²⁺ (Extended Data Fig. 9a–d). Compared to calcium, NC1 strongly augmented TRPM4 currents and intracellular Ca²⁺ did not further enhance NC1-induced activation (Extended Data Fig. 9e). Additionally, calcium chelation did not inhibit NC1-induced NECSO (Extended Data Fig. 9f,g).

Point substitutions at the calcium-binding site23,29, E828K and E1068Q, were tested for NC1 responsiveness. E828K completely blocked NC1-induced activation, while E1068Q retained reduced responsiveness (Extended Data Fig. 9h,i). In TRPM4-KO cells, these mutants showed partial or no NECSO-inducing effects (Extended Data Fig. 9j). These findings suggest that NC1 and calcium may share common mechanisms in activating TRPM4 channel activity.

Energy-depletion-induced necrosis shares features of NECSO

Dysregulation of TRPM4 is linked to ATP depletion in conditions such as ischemia and spinal cord injury30,31. We hypothesized that NC1-induced NECSO mimics energy-depletion-induced necrosis. Cos7 cells expressing WT TRPM4 or mutants (H159A with reduced ATP inhibition24, channel-deficient D984A and E1068Q with reduced calcium sensitivity) were tested for cell death through LDH release assay upon NC1 exposure or 2-deoxy-D-glucose (2DG) + NaN3 (energy depletion). WT TRPM4 cells showed dose-dependent cell death with NC1 and energy depletion, while H159A cells were more vulnerable to NC1 (Fig. 5a). Conversely, D984A and E1068Q mutants exhibited minimal cell death under either condition. NC1-binding defective mutants (W864A, H908A and F936A) also resisted energy-depletion-induced cell death (Fig. 5b). These findings suggest that energy-depletion-induced necrosis shares a genetic profile with NECSO.

Fig. 5: Energy-depletion-induced cell death shares features of NECSO.

a, Left: heat map depicting the cell death (measured by LDH release) of Cos7 cells stably expressing various hTRPM4 mutants and then treated with 2DG + NaN3 (0, 5, 10 or 20 mM 2DG plus 1 mM NaN3) or NC1 (0, 200, 500 or 1,000 nM) for 16 h. Right: immunoblot analysis of the expression levels of different hTRPM4 mutants in Cos7 cells. b, Heat map depicting the cell death of Cos7 cells expressing various NC1-insensitive hTRPM4 mutants and then treated with 2DG + NaN3. c, The whole-cell current density–voltage relation of Cos7 cells expressing hTRPM4 activated by 2DG. d, Intracellular sodium levels measured with CoroNa in Cos7 cells stably expressing hTRPM4 treated with vehicle or 20 mM 2DG plus 1 mM NaN3 in EBS for 6 h by fluorescence-activated cell sorting (FACS). The sodium in EBS was either maintained at physiological concentration (150 mM) or completely replaced with equal molar NMDG. e, Heat map illustrating the cell death of Cos7 cells transfected with WT hTRPM4 or H159A mutant and subsequently treated with 2DG + NaN3 under sodium-free and sodium-containing conditions. f, Top: heat map depicting the cell death of rat cardiac myoblast H9C2 cells transiently expressing various GOF or LOF mutants of hTRPM4 and then treated with 2DG + NaN3 or NC1. Bottom: immunoblot analysis of the expression levels of different hTRPM4 mutants in H9C2 cells. g, The expression levels of TRPM4 in human failing heart according to a meta-analysis of multiple studies (https://saezlab.shinyapps.io/reheat/). The t value represents the corrected gene expression level, with each data point representing an independent study. The bold dashed line represents the mean value, while the thin dashed lines represent the upper and lower quartiles. Cell death data represent the mean of n = 3 replicates from one representative of three independent experiments. Electrophysiological data represent the mean ± s.e.m. of n = 6 independent cells. Flow cytometry and immunoblot data show one representative of three replicates from three independent experiments.

Source data

Full size image

We examined NECSO under energy depletion. In Cos7 cells with hTRPM4, 2DG activated TRPM4 channel activity (Fig. 5c). Energy depletion triggered sodium influx and necrotic cell death, both counteracted by sodium depletion with NMDG (Fig. 5d,e). These results indicate that energy depletion may also induce NECSO.

GOF mutants of hTRPM4 show NECSO sensitivity

TRPM4 mutations are linked to cardiac arrhythmias, including progressive familial heart block type 1, Brugada syndrome and long QT syndrome, which can lead to sudden cardiac death (SCD)32,33,34. Over 25 mutations have been associated with familial cardiac diseases, with some acting as gain-of-function (GOF) and others as loss-of-function (LOF) mutants. Abnormal energy supply or stress could trigger SCD in these populations (Supplementary Table 1).

TRPM4 is more abundant in Purkinje fibers, which are involved in ventricular arrhythmias35. We hypothesized that disease-related mutations may alter NECSO sensitivity in heart cells, leading to injury. Several GOF and LOF TRPM4 mutants were expressed in rat H9C2 cardiomyoblasts. GOF mutants showed increased sensitivity to NECSO, while LOF mutants did not respond (Fig. 5f and Supplementary Table 1). This suggests that TRPM4 GOF mutations may enhance cell death and injury during energy-depletion conditions. Consistently, TRPM4 KO protected human cardiomyoblasts from NC1 treatment and energy depletion (Supplementary Fig. 5a,b). Additionally, overexpression of TRPM4 but not other TRPM family members was upregulated in left-ventricular samples from people with heart failure across 16 transcriptomic studies (Fig. 5g).

Identification of CLT and DHPs as NECSO inhibitors

To identify small-molecule inhibitors of NECSO, we used in silico screening with the newly identified NC1-binding pocket of TRPM4. Clotrimazole (CLT), an antifungal drug, was found to share a similar structure to NC1 and fit well into the binding cavity. Molecular docking showed that CLT binds through hydrophobic interactions, van der Waals contacts and hydrogen bonds, interacting with residues such as M927 in S5, F931 in S6 and H908 in S4 (Fig. 6a and Supplementary Data 2). These interactions suggest that CLT competes with NC1 to inhibit TRPM4 and NECSO. Testing in MCF7 cells confirmed CLT’s potency, with an EC50 of 4 µM (Fig. 6b). Its inhibitory effect on TRPM4 was also confirmed through electrophysiology (Fig. 6c and Supplementary Fig. 6).

Fig. 6: DHPs and CLT function as small-molecule inhibitors of NECSO.

a, Left: 3D diagram depicting that CLT specifically binds to hTRPM4, recognizing a deep cavity similar to NC1. Right: 2D diagram showing that CLT interacts with the key residues of TRPM4 through a mixture of hydrogen bonds and hydrophobic interactions. b, CLT and DHPs presented an inhibitory effect on NC1-induced cell death. MCF7 cells were pretreated with increasing concentrations of CLT or DHPs followed by 100 nM NC1 treatment and cell viability was assessed using CTG 24 h thereafter. c, The whole-cell current density–voltage relation of HEK293T cells stably expressing hTRPM4. Cells were pretreated with 30 μM DHPs or CLT followed by 1 μM NC1 treatment. d, Top hits of inhibitor screening for NECSO. The DHP dots are highlighted in cyan color. Insets: chemical structures. e, Intracellular sodium levels measured by FACS using CoroNa in MCF7 cells pretreated with 30 μM DHPs or CLT for 1 h and then exposure to 100 nM NC1 for 3 h. f, Heat map illustrating the cell viability of MCF7 or MB468 cells pretreated with various types of calcium-channel inhibitors followed by NC1 exposure. g, Heat map depicting the cell death measured by LDH release of H9C2 cells expressing hTRPM4 mutants treated with 2DG + NaN3 or NC1 combined with treatment of different inhibitors. Cell death data represent the mean of n = 3 replicates from one representative of three independent experiments. Electrophysiological data represent the mean ± s.e.m. of (n = 4 for vehicle, NC1 and CLT and 3 for DHPs) independent cells. Flow cytometry and cell death data show one representative of three replicates from three independent experiments.

Source data

Full size image

In the second strategy, we screened a US Food and Drug Administration-approved drug library of 1,376 molecules to identify inhibitors of NC1-induced cell death in MCF7 cells. The nonspecific TRPM4 antagonist glimepiride36 and mitochondrial permeability transition pore inhibitor cyclosporine A (CsA)37 were selected, validating the screening strategy (Fig. 6d). Dihydropyridines (DHPs), known calcium-channel blockers targeting L-type voltage-gated calcium channels (LTCCs)38, also showed strong NECSO inhibitory effects in MCF7 cells (Fig. 6b). We hypothesized that DHPs and CLT block NECSO through different mechanisms; DHPs indirectly suppress NC1-mediated TRPM4 activation, while CLT directly inhibits TRPM4. Electrophysiological results confirmed that DHPs did not affect TRPM4 channel activity but CLT directly inhibited it (Fig. 6c). Both DHPs and CLT blocked sodium influx induced by NC1 (Fig. 6e). Moreover, we tested whether other calcium channels were involved in NECSO and found that only LTCC blockers efficiently inhibited NECSO in MCF7 and MB468 cells (Fig. 6f). LTCCs are activated by TRPM4-dependent membrane depolarization39, suggesting a feedback loop between these channels regulating NECSO. The above data demonstrate that DHPs inhibit NECSO by targeting LTCCs to prevent TRPM4 activation, while CLT blocks NECSO by directly competing with NC1 for its binding pocket on TRPM4.

DHPs and CLT were tested for their ability to block energy-depletion-induced cell death. Both inhibitors reduced NC1 and energy-depletion-induced NECSO more effectively than glimepiride in Cos7 cells (Extended Data Fig. 10). They also protected cardiomyoblasts expressing TRPM4 GoF mutants from energy depletion (Fig. 6g). These findings suggest that NECSO inhibitors could have therapeutic potential to protect cardiac cells, particularly in conditions with TRPM4 GOF mutations.

Discussion

In this study, we reported that a small-molecule compound elicits necrotic cell death, which is characterized by sodium overload and dependent on monovalent cation channel TRPM4. The small-molecule compound NC1 serves as a highly selective agonist for TRPM4 through a pocket architecturally akin to the ligand-binding regulatory pocket found in other TRP family proteins. NC1 persistently activates TRPM4 channel activity. Consequently, uncontrolled opening of TRPM4 channel causes massive sodium influx and membrane depolarization, leading to necrotic cell death, which we termed as NECSO to separate from other types of necrotic cell death. NECSO could be blocked by calcium-channel blocker indirectly and an antifungus drug CLT as a direct competitor. Energy depletion likely activates NECSO given its shared cell morphology change, genetic and chemical intervention profiles. Cardiac arrhythmias associated with TRPM4 GOF mutants are more vulnerable to NECSO and can be blocked by NECSO inhibitors.

TRPM4 has been implicated in cellular demise in diverse cellular contexts, including neurons31,40,41, astrocytes19,42 and endothelial cells30,42. This demise has long been classified as oncosis43. However, the oncosis classification presents an obscure definition based on morphological changes such as cell swelling and membrane blebbing that often result from cellular stress or damage leading to cell membrane permeability changes. There are many factors that might cause cell membrane permeability changes and, in this study, we described NECSO as a type of necrotic cell death wherein necrotic cell death is incited specifically by the surge of sodium ions mediated by the activation of TRPM4. Although persistent TRPM4 activation might also lead to the influx of monovalent cations such as lithium and cesium, our study highlights sodium influx as the pivotal driver of cellular demise resulting in the most pronounced cell death response. Moreover, the influx of sodium rather than other cations mediated by TRPM4 likely governs the major cellular responsiveness to NC1 in vivo, given the limited physiological concentrations of other monovalent cations in extracellular fluid.

NECSO stands apart in its genetic and biochemical composition, diverging markedly from apoptosis, autophagy and various well-defined programmed necrosis pathways. NECSO likely features a more restrained level of membrane permeability, selectively permitting ion exchange before extensive membrane rupture. The precise mechanism by which sodium influx triggers necrosis remains unclear, although membrane depolarization is thought to bear importance. The participation of the sodium pump (Na+–K+ ATPase) in this process is contemplated. Excessive sodium influx would drive hyperactivation of sodium pumps, which relentlessly hydrolyze ATP to expel sodium from the cell, thus maintaining membrane potential homeostasis. This process, intensively energy consuming, could demand up to 70% of cellular energy, particularly notable in certain neural tissues44. Continuous hyperactivation of sodium pumps would deplete the cellular ATP reservoir, leading to energy exhaustion. Supporting this notion, our observations underscore the influence of mitochondria on NECSO, where CsA, a mitochondrial permeability transition inhibitor, effectively curtails NC1-triggered cell demise15. Notably, mitochondria’s roles in ATP or reactive oxygen species generation could potentially intersect with NECSO regulation, given their documented modulation of TRPM4 channel activity19,45,46. Additionally, excessive sodium might directly impede mitochondrial energy production, as Na+ operates as a secondary messenger modulating the function of the mitochondrial oxidative phosphorylation system by influencing inner mitochondrial membrane fluidity47. These hypotheses will be subjected to further exploration in upcoming investigations.

TRPM4 is a member of the TRP channel family, complex polymodal signal sensors adept at integrating information from physical and/or chemical stimuli to modulate cellular excitability across diverse environmental or physiological contexts48. Different from extensively explored TRPV and TRPA families, which serve as receptors for vital sensory perceptions such as heat, pain and cold, the function and regulatory mechanism of TRPM4 remain largely enigmatic because no natural or physiological ligand of TRPM4 has been reported until now. Compounds such as intracellular calcium, decavanadate, tissue plasminogen activator, U73122 and diazoxide have been documented to activate TRPM4 (ref. 49), likely different from NC1. NC1 potently and persistently activates TRPM4 channel activity and retains its membrane localization to enable a robust influx of sodium ions. Molecular docking has revealed that a cavity in the TM region surrounded by the S3 and S4 helices and S4–S5 linker from one subunit and S5 and S6 helix from the adjacent subunit is crucial for NC1 effect. This pocket exhibits a degree of conservation across TRP channels, which is often referred as the VBP pocket for ligand binding. For example, ligand–receptor pairs including vanilloid–TRPV1 (PDB 5IRX), ECN–TRPV5 (PDB 6B5V) and GNE551–TRPA1 (PDB 6X2J) all bind through VBPs. NC1 is the first chemical ligand shown to activate TRPM4 through the VBP region, which might explain its potent activation of this channel. The docking results may have limitations because the molecular docking was conducted on the basis of the published calcium-bound structure of hTRPM4, which is still a closed state23. However, upon NC1 binding, TRPM4 most likely transformed to an open state and was trapped in that conformation by NC1. The compound might eventually penetrate into a more embedded reservoir and, thus, can no longer be washed out, which can be confirmed through further structural analysis. We also noticed a significant difference between the EC50 of NC1 in cells and the Kd value obtained from the in vitro SPR experiment. We detected a Kd value of 880 nM for the NC1–TRPM4 interaction, with no significant binding observed for NC1i (Fig. 1f and Supplementary Fig. 2e,f). However, the Kd value still exhibits some disparity with the EC50 of NC1. We suspect the lipid binding and likely protein binding of TRPM4 might contribute to this difference. TRPM4 is a TM protein and lipids (phosphatidylinositol 4,5-bisphosphate) on the cell membrane can bind to TRPM4, thus affecting its conformational stability or activation24,50. In SPR experiments, TRPM4 is directly coupled to the chip without lipid binding. Thus, the Kd value might not be as low as the EC50 in the cellular circumstance. Some TRPM4-interacting proteins might also help the binding of TRPM4 with NC1. All these possibilities will be systemically and thoroughly studied in the future.

Interestingly, NC1 specifically activates hTRPM4 but not mTRPM4 despite the fact that hTRPM4 and mTRPM4 share 88% sequence homology. NC1 failed to bind mTRPM4 and mTRPM4 is unable to rescue the deficiency caused by hTRPM4 KO. We demonstrated that the TM domain accounts for the distinct sensitivity to NC1 in hTRPM4 and mTRPM4. In addition, NC1 might stabilize and activate TRPM4 allosterically, which might help to capture dynamic channel activation processes. Structural studies on TRPV1 with double-knot toxin, a peptide spider toxin stabilizing the open state of TRPV1, successfully revealed the dynamic regulation of TRPV1 (ref. 51). Until now, the structure of the open state of TRPM4 has yet to be elucidated23,24 and future studies on the structure of NC1–hTRPM4 might help to resolve this problem. Furthermore, the NC1-binding pocket might be a valuable target for drug design for TRPM4 agonists and antagonists, which is illustrated by the example of CLT, an antifungal agent sharing structural similarities with NC1 and fitting in the NC1–hTRPM4 pocket. CLT serves as an antagonist to restrain NC1-triggered NECSO. CLT is the ring-opened analog of NC1 with a similar three-dimensional (3D) structure containing a combination of hydrophobicity and π–π interactions, a spherical shape and an all-carbon quaternary stereocenter. CLT’s inhibitory effect likely occurs through competition with NC1 because both compounds share a common pocket within TRPM4. The inhibitory effect of CLT on TRPM4 and other TRPM proteins has also been reported by other groups49,52. Notably, CLT’s neuroprotective impact on spinal cord ischemia–reperfusion injury53 likely also arises from TRPM4 inhibition.

Under physiological circumstances, the propensity for TRPM4 to open is heightened by elevated cytoplasmic [Ca2+] and diminished ATP concentrations. Augmented cytosolic Ca2+ levels or profound ATP depletion often triggers TRPM4-dependent cell death in pathological contexts54. Consistently, we observed that energy-depletion-induced necrotic cell death shares largely overlapped morphology and chemical profiles with NECSO, suggesting that NECSO is involved in energy-depletion-induced necrotic cell death. Furthermore, TRPM4 dysfunctions have been implicated in a broad spectrum of pathological contexts involving cell and tissue injuries including stroke, brain or spinal cord injury, infection, cardiac diseases and cancers31,40,41,55.

In this study, we observed that cardiac disease-associated TRPM4 GOF mutations are more vulnerable to NECSO and the inhibitors are able to block the resulting cell death. It is reported that TRPM4 not only has an essential role in cardiac conduction but is also involved in hypertension, hypertrophy and heart failure56,57. TRPM4 KO improves survival rate and enhances β-adrenergic cardiac reserve after ischemic heart failure and TRPM4 inhibitor 9-phenanthrol was also found to reduce ischemia–reperfusion-induced cardiac death in rats58. An increase in expression of TRPM4 in human failing heart tissue was also reported59. Previous investigations of cardiac diseases mostly focused on the electrophysiological function of TRPM4. Our research supplies a mechanism underlying the function of TRPM4 as a cell death mediator. Therefore, development of TRPM4 antagonists might hold promise to protect cells from sodium-overload-associated cell and tissue injuries. On the other hand, NECSO inducers such as NC1 might represent a promising avenue for cancer therapy. TRPM4 dysfunction has been linked to multiple cancers and NC1 derivatives have been shown to present robust anticancer effects on xenograft tumor models60. However, how to balance the killing effects of TRPM4 in normal and cancer cells remains a challenge. Nevertheless, these studies implicate the involvement of NECSO in a broad spectrum of human diseases, the targeting of which might bring beneficial effects to populations if we carefully elucidate the detailed mechanisms and tease apart the critical positive and unwanted side effects in each pathological condition. With the critical genes and chemical tools identified in this study, we can start to address the biological importance of NECSO.

Methods
Cell lines and culture conditions

MCF7, MB468, HeLa, H9c2, HEK293T, SW620, C127, CHO, Cos7 and MEF cells were cultured in DMEM containing 4.5 g L−1 glucose and L-glutamine (Sigma-Aldrich). PC3, HCC1143, 4T-1, EMT-6, CT26WT, B16 and MC38 cells were cultured in RPMI-1640 medium containing high glucose, L-glutamine and HEPES (Sigma-Aldrich). U-2OS cells were cultured in McCoy’s 5A medium (American Type Culture Collection (ATCC)). A549 cells were cultured in Ham’s F-12K medium (BaseMedia). Neuro-2a cells were cultured in Eagle’s minimum essential medium (ATCC). AC16 cells were cultured in DMEM:F-12 medium (ATCC). All media were supplemented with 10% FBS (Gibco) and 1% penicillin–streptomycin. The cells were cultured at 37 °C in a humidified atmosphere with 5% CO2. All cell lines were tested for Mycoplasma and were not authenticated.

Plasmids

Plasmid transfections were performed in HeLa or HEK293T cells with PEI (Polysciences) transfection reagent. Virus was produced by cotransfection of HEK293T cells with GAG, POL and pLenti expression plasmids at a 1:1:1 w/w/w ratio, using the PEI transfection reagent. Medium containing secreted virus was obtained after 48 h and sterile-filtered. CRISPR sgRNA sequences targeting TRPM4 were designed using the CRISPR design tool developed by the Zhang laboratory (http://crispr.mit.edu/). The oligo sequences preceding the protospacer motif were as follows:

5′-GAACCGCTTCCGCTTGCGCC-3′

5′-CCTAGGGCTCGTTCCCTGCG-3′

5′-GTCAACTATGAACGTCGTGC-3′

Western blotting

Proteins were extracted from the cells using radioimmunoprecipitation assay lysis buffer supplemented with protease and phosphatase inhibitors (APEX BIO). The protein concentration was determined using the BCA protein assay kit (Thermo Fisher Scientific). Equal amounts of protein (20 μg) were separated on 10% SDS–PAGE gels (Bio-Rad Laboratories) and transferred onto PVDF membranes (Bio-Rad Laboratories). The membranes were then blocked in 5% nonfat milk for 1 h at room temperature and incubated overnight at 4 °C with primary antibodies against the target proteins. After washing, the membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized using an enhanced chemiluminescence substrate and detected using a chemiluminescence imaging system (Bio-Rad Laboratories). The following blotting reagents and antibodies were used anti-α-tubulin (Sigma-Aldrich, T6199; diluted 1:10,000), anti-β-actin (Sigma-Aldrich, A5441; diluted 1:10,000), anti-TRPM4 (Abclonal, A10146; diluted 1:1,000) and anti-Cas9 (Abcam, ab204448; diluted 1:1,000).

Cell viability and death measurement

The cell viability and death were assessed using the CellTiter-Glo (CTG) assay or LDH release assay. Cells were seeded in a 96-well plate at a density of 1 × 104 cells per well and allowed to adhere overnight. For analyses using EBS, culture medium was replaced with isotonic EBS (150 mM NaCl, 5 mM KCl, 10 mM glucose, 10 mM HEPES, 2 mM CaCl2, 1 mM MgCl2) just before the experimental treatment. After treatment in various experimental conditions, the culture medium was aspirated. For the CTG assay, the intracellular ATP was measured using the CTG luminescent assay kit (Promega, G7570) according to the manufacturer’s instructions as described previously. Briefly, cells were washed with PBS, 100 μl of CTG reagent was added to each well and the plate was gently shaken for 10 min to induce cell lysis and release of intracellular ATP. The luminescence signal, proportional to the amount of ATP present and indicative of cell viability, was measured using a multiwell plate reader. For the LDH release assay, the reaction mixture containing the culture supernatant and LDH assay solution (Promega) was incubated at room temperature for 30 min, followed by the addition of stop solution. The absorbance was measured at 490 nm using a microplate reader. A standard curve was generated using serially diluted LDH control samples to calculate the percentage of LDH release from treated cells. The experiment was performed in triplicate and the results were expressed as the mean ± s.d.

CRISPR–Cas9-based screening and data processing

A schematic of CRISPR screening is shown in Fig. 2a. Specifically, MCF7 cells stably expressing Cas9 (blasticidin selection) were generated and then infected with a whole-genome CRISPR–Cas9 gRNA lentiviral library (human library, GeCKO version 2.0) that targets 19,050 protein-coding genes with 122,411 gRNA sequences (six targeting gRNA sequences per gene). After being treated with 2 µg ml−1 puromycin for 3 days, the infected cells were referred to as ‘day 0’. These cells were subsequently passaged every 2 days for a total of 14 days (denoted as day 14), while maintaining 500-fold coverage. At the end of the 14-day period, the cells were divided into two groups and cultured in medium with DMSO or 2 nM NC1. Samples were collected from the day 0, DMSO and NC1 groups and genomic DNA extraction was performed using a QIAamp DNA blood midi kit (Qiagen). The sgRNA inserts were amplified using NEBNext Q5 hot start HiFi PCR master mix (New England Biolabs). In the second round of PCR, Illumina TruSeq adaptors with i5 and i7 barcodes were added and the PCR products were subjected to next-generation sequencing. The obtained sequencing results were then used for further analysis. The gRNA differences between the DMSO and NC1 groups were calculated and analyzed using the MAGeCKFlute16.

Electrophysiology

All electrophysiological recordings were obtained using patch-clamp recordings for (1) macroscopic current of TRPM4 and its mutants overexpressed in HEK293T, HeLa or Cos7 cells and (2) membrane potential of MCF7 cells. For transfected cells, patch-clamp recordings were conducted 24–48 h after transfection. For calcium-activated TRPM4 whole-cell recording, the bath solution contained 150 mM NaCl, 10 mM HEPES, 1 mM MgCl2 and 2 mM CaCl2, pH 7.3. The pipette solution contained 150 mM KCl, 10 mM HEPES, 5 mM EGTA and 4.45 mM CaCl2, pH 7.3 (1 μM free Ca2+). For the calcium-free condition, 10 mM EGTA was added without CaCl2. In the inside-out or outside-out patch mode, the extracellular solution contained 150 mM NaCl, 10 mM HEPES, 1 mM MgCl2 and 2 mM CaCl2, pH 7.3. The intracellular solution contained 150 mM NaCl, 10 mM HEPES and 5 mM EGTA. The 5 mM EGTA was replaced by 1 mM CaCl2 for calcium-activated TRPM4 inside-out patch. Patch pipettes were pulled from PC-10 (Narishige) and heat-polished to a resistance of 3–10 MΩ. Patch-clamp recordings were performed with Axopatch-200B amplifier and Axon Digidata 1550A driven by Clampex10 software (Molecular Devices). The current signals were filtered at 1 kHz and digitized at a 10-kHz sampling frequency. The whole-cell currents were recorded by voltage steps from −100 mV to +100 mV in 20-mV increments, followed by stepping down to −100 mV. The series resistance was compensated before forming a tight patch seal. The concentration of free Ca2+ was prepared according to MAXCHELATOR. Liquid junction potentials were less than 2 mV, as calculated by JPCalc software. All values were measured using Clampfit 10.6 software.

In the analysis of NC1 activation of TRPM4, the normalized activation current (I/I0) at +100 mV was fitted with the Hill equation:

$$\frac{I}{{I}_{0}}={A}_{1}+\frac{{A}_{2}-{A}_{1}}{1+{10}^{(\log {x}_{0}-x)p}}$$
(1)

where I is the current amplitude after application of activators, A1 and A2 are constants between 0 and 1, x is the concentration of activator, x0 is the EC50 and p is the Hill constant.

In the analysis of CLT inhibition of TRPM4, the fraction of unblocked current (I/I0) at +100 mV was fitted with the Hill equation:

$$\frac{I}{{I}_{0}}={A}_{1}+\frac{{A}_{2}-{A}_{1}}{1+{10}^{(\log {x}_{1}-x)p}}$$
(2)

where I0 and I are the current amplitudes before and after the application of inhibitors, A1 and A2 are constants between 0 and 1, x is the concentration of the inhibitor, x1 is the half-maximal inhibitory concentration (IC50) and p is the Hill constant.

ICP-MS

The ICP-MS was conducted according to a previously published paper61. In brief, after calculation of the cell number, cell pellets were collected directly into the 50-ml metal-free tube and then 2 ml of 69% nitric acid was added. The tubes were placed on the digestion block (DigiPREP Jr., SCP-Science) and the temperature of the block was ramped up to 120 °C and maintained at 120 °C for approximately 1 h until the solution had almost disappeared. Tubes were gently swirled every 20 min during incubation. When the samples had cooled down, 0.5 ml of H2O2 was added and the samples were heated at 120 °C for a further 30 min. After digestion was complete, the tubes were allowed to cool to room temperature. The volume was made up to 20 ml with ultrapure water. Aqueous calibration standards were prepared by diluting 1,000 mg L−1 ICP-MS standard solutions (Spex CertiPrep) with 10% nitric acid. Elements were detected by ICP-MS (NexION2000G, PerkinElmer). Element concentrations were quantified using external calibration with aqueous standard solutions and 103Rh as an internal standard. A certified liver reference standard (T07243QC, FAPAS) was used for quality control.

Flow cytometry

For measurement of intracellular sodium concentration, cells were detached from culture plates using 1× PBS + EDTA. Single-cell suspensions were prepared in 1× PBS with 4% FBS. The cells were then stained with 2 μM CoroNa-AM dye at 37 °C for 30 min, followed by two washes with 1× PBS 4% FBS. Analysis of stained cells was performed using a CytoFlex S (Beckman Coulter) and postacquisition analysis was conducted using FlowJo software. Sodium concentration were measured under FITC (525/50 505LP). To assess intracellular sodium concentration under NC1 or 2DG treatment, cells were subjected to the previously described treatment durations and concentrations before being lifted off the culture plates.

Fluorescence microscopy

For fluorescence microscopy, HeLa cells transiently expressing the hTRPM4–EGFP fusion protein were seeded onto glass coverslips and grown to 70–80% confluency. Followed by appropriate NC1 treatment, the cells were loaded with 2.5 μg ml−1 propidium iodide (PI) at room temperature for 20 min to stain the nuclei. After three washes with PBS, the coverslips were mounted onto glass slides and ready for fluorescence imaging.

For live-cell microscopy, cells were grown in four-well Lab-Tek II chambered coverglass (Thermo Fisher Scientific) imaging chambers. To image membrane potential changes, cells were incubated for 2 h with 5 µM DiBAC4(3) (Dojindo) and then treated with 50 nM NC1 before imaging. To image the effect of ions on NECSO, MCF7 cells were incubated with SytoxGreen (Thermo Fisher Scientific) for 30 min at room temperature and the medium was replaced with another isotonic medium containing different ions as indicated before imaging.

TEM

TEM analyses were performed following established protocols. In brief, MCF7 cells were treated with 100 nM NC1 at different time points as indicated and then fixed with a fixative solution containing 2% paraformaldehyde, 3% glutaraldehyde and 0.1 M cacodylate buffer at pH 7.3. Subsequently, the cells were washed with 0.1 M sodium cacodylate buffer and treated with 0.1% filtered cacodylate-buffered tannic acid. They were then postfixed with 1% buffered osmium and stained with 1% filtered uranyl acetate. After dehydration, the samples were subjected to a graded series of Epon up to 100% concentration for 6 h at 37 °C. They were then embedded in pure Epon and left overnight at 60 °C. Thin sections were obtained using a Leica EM UC7 ultramicrotome, placed on formvar-coated grids (Gilder Grids, AG100N) and stained with 3% uranyl acetate and lead citrate. Finally, the sections were examined using the TEM instrument PHILIPS CM-120 operating at 80 or 120 kV.

Protein expression and purification

hTRPM4 (NM_017636) containing a C-terminal 3× Flag tag was cloned into a PCDNA4.1 vector and heterologously expressed in HEK293S cells (Life Technologies) using the PEI transfection system (Thermo Fisher Scientific). Cells were cultured in suspension at 37 °C for 48 h and harvested by centrifugation at 1,500 rpm for 15 min. The cell pellet was resuspended in buffer A (20 mM HEPES pH 7.4 and 150 mM NaCl) supplemented with a protease inhibitor cocktail (containing 2 μg ml−1 DNase I, 0.5 μg ml−1 pepstatin, 2 μg ml−1 leupeptin, 1 μg ml−1 aprotinin and 0.1 mM PMSF) and homogenized by sonication on ice. TRPM4 was extracted with 2% (w/v) DDM (Goldbio) supplemented with 0.2% (w/v) cholesteryl hemisuccinate (CHS; Sigma-Aldrich) by gentle agitation for 3 h at 4 °C. After extraction, the supernatant was collected following a 40-min centrifugation at 48,000g and incubated with Flag beads resin (Sigma-Aldrich) with gentle agitation overnight. Then, the resin was collected following a 5-min centrifugation at 300g and washed three times in buffer B (buffer A + 0.1% DDM + 0.02% CHS). TRPM4 was eluted with buffer B supplemented with 2 mg ml−1 Flag peptide and further purified by size-exclusion chromatography (SEC) on a Superose 6 10/300 GL column (GE Heathcare) pre-equilibrated with buffer B. The protein peak was collected and concentrated to use.

Protein–ligand interaction

For the CETSA, HEK293T cells stably expressing TRPM4 were cultured and treated with 100 μM NC1 or NC1i for 1 h at room temperature. Following compound treatment, cells were divided into aliquots and subjected to a heat shock at different temperatures ranging from 37 °C to 70 °C for 3 min. The heated samples were then immediately cooled on ice and lysed and the resulting lysates were centrifuged to obtain the soluble fraction. Protein levels in the soluble fraction were assessed with western blotting. The intensity of the protein bands was quantified using image analysis software (Fiji Image J, version 1.53a) and the thermal stability profiles were generated by plotting the protein band intensity against the heat-shock temperatures. A shift in the melting temperature (Tm) of the target protein in the presence of the compound indicates binding and stabilization of the protein. Control samples without compound treatment were included to establish baseline thermal stability profiles. Statistical analysis was conducted using GraphPad Prism 9 software and the significance of Tm shifts was determined using one-way analysis of variance (ANOVA) followed by Dunnett’s post hoc test.

For the SPR-based binding assay, a Biacore 8K instrument (GE Healthcare) was used. The assays were conducted at 25 °C in HBS-EP running buffer (10 mM HEPES pH 7.4, 150 mM NaCl, 0.2% (v/v) surfactant DDM and 2% DMSO). The WT Htrpm4 protein was immobilized onto CM7 sensor chips using a standard amine-coupling procedure in 10 mM sodium acetate (pH 4.5). Compounds were diluted in a serial manner and injected onto the sensor chip at a flow rate of 30 μl min−1 for 120 s (contact phase), followed by a buffer flow of 120 s (dissociation phase). The KD value was determined using Biacore 8K evaluation software (GE Healthcare, version 3.0.12) by analyzing the steady-state data at equilibrium.

Molecular docking simulations

The TRPM4 structure was obtained from the Protein Data Bank (PDB 6BQV), while the two-dimensional (2D) structure of NC1 or CLT was built from the ChemDraw software (CambridgeSoft, version 18.0). Molecular docking was performed using the Schrödinger Maestro software suite (Schrödinger, 2020-3). Before docking, the ligand was optimized using the OPLS3e force field in the LigPrep module. The protein underwent the protein preparation wizard, including the addition of missing residues, removal of waters and optimization of hydrogen bonds and energy using the OPLS3e force field. Protonation states of the protein and ligand at pH 7.4 ± 0.2 were determined using Epik. Docking was performed on a selected residue within a 25 × 25 × 25 Å3 grid box. Extra-precision docking (Glide XP) with flexible ligand sampling was used, followed by postdocking minimization to generate a maximum of ten ligand poses using the Glide program. The docking conformation with the highest score was analyzed to identify critical molecular interactions between the ligand and amino acid residues in the binding pocket, guiding site-directed mutagenesis.

MD simulation

For MD simulation, only the TM domain of TRPM4 was selected. The best pose of the TRPM4–NC1 complex obtained from docking was chosen and the ligand–protein systems were constructed using the CHARMM-GUI62 (https://charmm-gui.org) membrane builder. The complex was inserted into a lipid bilayer composed of 180 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine molecules, solvated with a 22.5-Å-thick layer of 150 mM NaCl aqueous solution. To assess the stability of the ligand–protein poses, 50-ns simulations were performed. After 25,000 steps of minimization, the systems underwent equilibration in six steps according to the default settings from CHARMM-GUI lipid bilayer outputs, followed by a 50-ns production run. All simulations were conducted using GROMACS 2020.3 with CHARMM36m force fields and WYF parameters for the protein, CHARMM general force field for the ligand, CHARMM36 force fields for the lipids and the TIP3P model for water. The simulations used isothermal–isobaric (NPT) ensembles at a constant temperature of 303.15 K. The simulation trajectories were analyzed for structural stability using r.m.s.d. calculations. MD simulation trajectories were also subjected to molecular mechanics energies combined with Poisson–Boltzmann surface area continuum solvation calculations using a gmx_mmpbsa bash script63, with solvent-accessible surface area used as the model for non-polar solvation energy. The best pose of the TRPM4–NC1 complex from molecular docking was selected for binding pocket analysis using Fpocket 2.0 software. Pocket parameter information was generated using the dpocket program with default settings.

Statistics and reproducibility

Statistical analyses, including one-way and two-way ANOVA, were conducted using GraphPad Prism software (version 9.0.0). All statistical data are reported as the mean ± s.d., except for the patch-clamp data, which are presented as the mean ± s.e.m. All microscopic, biochemical and biological assays were independently repeated at least three times with similar results. The mean values of the control and experimental groups were compared for significant differences. Data distribution was assumed to be normal, although formal testing was not performed. P values below 0.05 were considered statistically significant. The sample size was not determined using a statistical method. No data were excluded from the analyses. Illustrations were prepared using Adobe Illustrator and BioRender.com. The experiments were not randomized and the investigators were not blinded to treatment allocation or outcome assessment, unless otherwise specified.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All original data that support the findings of this study are provided in the article and Supplementary Information. All specific and stable reagents generated in this study are available from the corresponding authors without restriction. Source data are provided with this paper.

References

Thornton, S. N. Sodium intake, cardiovascular disease, and physiology. Nat. Rev. Cardiol. 15, 497 (2018).

Article
 
Google Scholar

Friedman, P. A. Codependence of renal calcium and sodium transport. Annu. Rev. Physiol. 60, 179–197 (1998).

Article
 
CAS
 
Google Scholar

Amiry-Moghaddam, M. & Ottersen, O. P. The molecular basis of water transport in the brain. Nat. Rev. Neurosci. 4, 991–1001 (2003).

Article
 
CAS
 
Google Scholar

Hunter, R. W., Dhaun, N. & Bailey, M. A. The impact of excessive salt intake on human health. Nat. Rev. Nephrol. 18, 321–335 (2022).

Article
 
Google Scholar

Liang, D., Bhatta, S., Gerzanich, V. & Simard, J. M. Cytotoxic edema: mechanisms of pathological cell swelling. Neurosurg. Focus 22, E2 (2007).

Article
 
PubMed Central
 
Google Scholar

Rungta, R. L. et al. The cellular mechanisms of neuronal swelling underlying cytotoxic edema. Cell 161, 610–621 (2015).

Article
 
CAS
 
Google Scholar

Wang, Y. et al. Brain tissue sodium is a ticking clock telling time after arterial occlusion in rat focal cerebral ischemia. Stroke 31, 1386–1391 (2000).

Article
 
CAS
 
Google Scholar

Papadopoulos, M. C. et al. Molecular mechanisms of brain tumor edema. Neuroscience 129, 1011–1020 (2004).

Article
 
CAS
 
Google Scholar

Dixon, S. J. et al. Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell 149, 1060–1072 (2012).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Tsvetkov, P. et al. Copper induces cell death by targeting lipoylated TCA cycle proteins. Science 375, 1254–1261 (2022).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Cai, Z. et al. Plasma membrane translocation of trimerized MLKL protein is required for TNF-induced necroptosis. Nat. Cell Biol. 16, 55–65 (2014).

Article
 
CAS
 
Google Scholar

Boron, W. F. & Boulpaep, E. L. Medical Physiology: A Cellular and Molecular Approach 2nd edn (Saunders, 2012).

Kadir, L. A., Stacey, M. & Barrett-Jolley, R. Emerging roles of the membrane potential: action beyond the action potential. Front. Physiol. 9, 1661 (2018).

Article
 
PubMed Central
 
Google Scholar

Murphy, E. & Eisner, D. A. Regulation of intracellular and mitochondrial sodium in health and disease. Circ. Res. 104, 292–303 (2009).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Zhang, J. et al. Necrocide 1 mediates necrotic cell death and immunogenic response in human cancer cells. Cell Death Dis. 14, 238 (2023).

Article
 
PubMed Central
 
Google Scholar

Wang, B. et al. Integrative analysis of pooled CRISPR genetic screens using MAGeCKFlute. Nat. Protoc. 14, 756–780 (2019).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Grand, T. et al. 9-Phenanthrol inhibits human TRPM4 but not TRPM5 cationic channels. Br. J. Pharmacol. 153, 1697–1705 (2008).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Nilius, B. et al. The selectivity filter of the cation channel TRPM4. J. Biol. Chem. 280, 22899–22906 (2005).

Article
 
CAS
 
Google Scholar

Chen, M. & Simard, J. M. Cell swelling and a nonselective cation channel regulated by internal Ca2+ and ATP in native reactive astrocytes from adult rat brain. J. Neurosci. 21, 6512–6521 (2001).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Launay, P. et al. TRPM4 is a Ca2+-activated nonselective cation channel mediating cell membrane depolarization. Cell 109, 397–407 (2002).

Article
 
CAS
 
Google Scholar

Bohlen, C. J. et al. A bivalent tarantula toxin activates the capsaicin receptor, TRPV1, by targeting the outer pore domain. Cell 141, 834–845 (2010).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Lin King, J. V. et al. A cell-penetrating scorpion toxin enables mode-specific modulation of TRPA1 and pain. Cell 178, 1362–1374 (2019).

Article
 
PubMed Central
 
Google Scholar

Autzen, H. E. et al. Structure of the human TRPM4 ion channel in a lipid nanodisc. Science 359, 228–232 (2018).

Article
 
CAS
 
Google Scholar

Guo, J. et al. Structures of the calcium-activated, non-selective cation channel TRPM4. Nature 552, 205–209 (2017).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Nilius, B. et al. Regulation of the Ca2+ sensitivity of the nonselective cation channel TRPM4. J. Biol. Chem. 280, 6423–6433 (2005).

Article
 
CAS
 
Google Scholar

Woo, S. K. et al. Complex N-glycosylation stabilizes surface expression of transient receptor potential melastatin 4b protein. J. Biol. Chem. 288, 36409–36417 (2013).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Cao, E., Liao, M., Cheng, Y. & Julius, D. TRPV1 structures in distinct conformations reveal activation mechanisms. Nature 504, 113–118 (2013).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Nilius, B. et al. Voltage dependence of the Ca2+-activated cation channel TRPM4. J. Biol. Chem. 278, 30813–30820 (2003).

Article
 
CAS
 
Google Scholar

Yamaguchi, S., Tanimoto, A., Otsuguro, K., Hibino, H. & Ito, S. Negatively charged amino acids near and in transient receptor potential (TRP) domain of TRPM4 channel are one determinant of its Ca2+ sensitivity. J. Biol. Chem. 289, 35265–35282 (2014).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Gerzanich, V. et al. De novo expression of Trpm4 initiates secondary hemorrhage in spinal cord injury. Nat. Med. 15, 185–191 (2009).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Yan, J., Bengtson, C. P., Buchthal, B., Hagenston, A. M. & Bading, H. Coupling of NMDA receptors and TRPM4 guides discovery of unconventional neuroprotectants. Science 370, eaay3302 (2020).

Article
 
CAS
 
Google Scholar

Pironet, A., Vandewiele, F. & Vennekens, R. Exploring the role of TRPM4 in calcium-dependent triggered activity and cardiac arrhythmias. J. Physiol. 602, 1605–1621 (2024).

Article
 
CAS
 
Google Scholar

Liu, H. et al. Gain-of-function mutations in TRPM4 cause autosomal dominant isolated cardiac conduction disease. Circ. Cardiovasc. Genet. 3, 374–385 (2010).

Article
 
CAS
 
Google Scholar

Syam, N. et al. Variants of transient receptor potential melastatin member 4 in childhood atrioventricular block. J. Am. Heart Assoc. 5, e001625 (2016).

Article
 
PubMed Central
 
Google Scholar

Hof, T. et al. TRPM4 non-selective cation channels influence action potentials in rabbit Purkinje fibres. J. Physiol. 594, 295–306 (2016).

Article
 
CAS
 
Google Scholar

Guinamard, R. et al. Functional characterization of a Ca2+-activated non-selective cation channel in human atrial cardiomyocytes. J. Physiol. 558, 75–83 (2004).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Halestrap, A. P., Connern, C. P., Griffiths, E. J. & Kerr, P. M. Cyclosporin A binding to mitochondrial cyclophilin inhibits the permeability transition pore and protects hearts from ischaemia/reperfusion injury. Mol. Cell. Biochem. 174, 167–172 (1997).

Article
 
CAS
 
Google Scholar

Hess, P., Lansman, J. B. & Tsien, R. W. Different modes of Ca channel gating behaviour favoured by dihydropyridine Ca agonists and antagonists. Nature 311, 538–544 (1984).

Article
 
CAS
 
Google Scholar

Gonzales, A. L., Amberg, G. C. & Earley, S. Ca2+ release from the sarcoplasmic reticulum is required for sustained TRPM4 activity in cerebral artery smooth muscle cells. Am. J. Physiol. Cell Physiol. 299, C279–C288 (2010).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Simard, J. M. et al. Newly expressed SUR1-regulated NCCa-ATP channel mediates cerebral edema after ischemic stroke. Nat. Med. 12, 433–440 (2006).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Leiva-Salcedo, E., Riquelme, D., Cerda, O. & Stutzin, A. TRPM4 activation by chemically- and oxygen deprivation-induced ischemia and reperfusion triggers neuronal death. Channels (Austin) 11, 624–635 (2017).

Article
 
Google Scholar

Wei, S. et al. Comparison of anti-oncotic effect of TRPM4 blocking antibody in neuron, astrocyte and vascular endothelial cell under hypoxia. Front. Cell Dev. Biol. 8, 562584 (2020).

Article
 
PubMed Central
 
Google Scholar

Loh, K. Y., Wang, Z. & Liao, P. Oncotic cell death in Stroke. Rev. Physiol., Biochem. Pharmacol. 176, 37–64 (2019).

Article
 
CAS
 
Google Scholar

Harris, J. J. & Attwell, D. The energetics of CNS white matter. J. Neurosci. 32, 356–371 (2012).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Nilius, B., Prenen, J., Voets, T. & Droogmans, G. Intracellular nucleotides and polyamines inhibit the Ca2+-activated cation channel TRPM4b. Pflugers Arch. 448, 70–75 (2004).

Article
 
CAS
 
Google Scholar

Simon, F. et al. Hydrogen peroxide removes TRPM4 current desensitization conferring increased vulnerability to necrotic cell death. J. Biol. Chem. 285, 37150–37158 (2010).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Hernansanz-Agustin, P. et al. Na+ controls hypoxic signalling by the mitochondrial respiratory chain. Nature 586, 287–28 (2020).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Diver, M. M., Lin King, J. V., Julius, D. & Cheng, Y. Sensory TRP channels in three dimensions. Annu. Rev. Biochem. 91, 629–649 (2022).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Kovacs, Z. M. et al. Pharmacological modulation and (patho)physiological roles of TRPM4 channel—part 1: modulation of TRPM4. Pharmaceuticals (Basel) 15, 81 (2022).

Article
 
CAS
 
Google Scholar

Zhang, Z., Okawa, H., Wang, Y. & Liman, E. R. Phosphatidylinositol 4,5-bisphosphate rescues TRPM4 channels from desensitization. J. Biol. Chem. 280, 39185–39192 (2005).

Article
 
CAS
 
Google Scholar

Zhang, K., Julius, D. & Cheng, Y. Structural snapshots of TRPV1 reveal mechanism of polymodal functionality. Cell 184, 5138–5150 (2021).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Hill, K., McNulty, S. & Randall, A. D. Inhibition of TRPM2 channels by the antifungal agents clotrimazole and econazole. Naunyn Schmiedebergs Arch. Pharmacol. 370, 227–237 (2004).

Article
 
CAS
 
Google Scholar

Usul, H., Arslan, E., Cansever, T., Cobanoglu, U. & Baykal, S. Effects of clotrimazole on experimental spinal cord ischemia/reperfusion injury in rats. Spine 33, 2863–2867 (2008).

Article
 
Google Scholar

Simard, J. M., Woo, S. K. & Gerzanich, V. Transient receptor potential melastatin 4 and cell death. Pflugers Arch. 464, 573–582 (2012).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Ghosh, S. et al. Plasma membrane channel TRPM4 mediates immunogenic therapy-induced necrosis. Cancer Res. 83, 3115–3130 (2023).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Vandewiele, F. et al. TRPM4 inhibition by meclofenamate suppresses Ca-dependent triggered arrhythmias. Eur. Heart J. 43, 4195–419 (2022).

Article
 
CAS
 
Google Scholar

Mathar, I. et al. Increased catecholamine secretion contributes to hypertension in TRPM4-deficient mice. J. Clin. Invest. 120, 3267–3279 (2010).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Piao, H. L. et al. Transient receptor potential melastatin-4 is involved in hypoxia–reoxygenation injury in the cardiomyocytes. PLoS ONE 10, e0121703 (2015).

Article
 
PubMed Central
 
Google Scholar

Flores, R. O. R. et al. Consensus transcriptional landscape of human end-stage heart failure. J. Am. Heart Assoc. 10, e019667 (2021).

Article
 
Google Scholar

Boudreau, M. W. et al. A small-molecule activator of the unfolded protein response eradicates human breast tumors in mice. Sci. Transl. Med. 13, eabf1383 (2021).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Huang, J. et al. Analysis of ionomic profiles of spinal cords in a rat model with bone cancer pain. J. Pain Res. 17, 1531–1545 (2024).

Article
 
PubMed Central
 
Google Scholar

Jo, S., Lim, J. B., Klauda, J. B. & Im, W. CHARMM-GUI Membrane Builder for mixed bilayers and its application to yeast membranes. Biophys. J. 97, 50–58 (2009).

Article
 
CAS
 
PubMed Central
 
Google Scholar

Valdes-Tresanco, M. S., Valdes-Tresanco, M. E., Valiente, P. A. & Moreno, E. gmx_MMPBSA: a new tool to perform end-state free energy calculations with GROMACS. J. Chem. Theory Comput. 17, 6281–6291 (2021).

Article
 
CAS
 
Google Scholar

Download references

Acknowledgements

We thank the Core Facility of Basic Medical Science, Shanghai Jiaotong University School of Medicine and Instrumental Analysis Center of Shanghai Jiaotong University for technical support. The computations in this paper were run on the π2.0 cluster and supported by bioinformatics platform of the Center for High Performance Computing at Shanghai Jiao Tong University. This work was partially supported by SJTU Kunpeng and Ascend Center of Excellence. We thank Y. Jiang (University of Texas Southwestern Medical Center) for helpful discussion. We thank N. Huang (Tsinghua Institute of Multidisciplinary Biomedical Research, Tsinghua University) for the use of Schrödinger software. We express our gratitude to J. Guo (Zhejiang University), A. C. Y. Chang and B. Li (Shanghai Jiaotong University) for providing reagents. The work was supported in part by grants from the Ministry of Science and Technology of the People’s Republic of China (2023YFA0914900), National Natural Science Foundation of China (92254307, M-0140 and 91957204), the Fundamental Research Funds for the Central Universities, the Shanghai Sailing Program (20YF1422300) to W.F. and Zhiyuan Future Scholar Program (zirc2023-13) to Z. Zhao. The work was also supported by the innovative research team of high-level local universities in Shanghai (SHSMU-ZDCX20212000) and Shanghai Frontiers Science Center of Cellular Homeostasis and Human Diseases.

Author information
Author notes

These authors contributed equally: Wan Fu, Jianghuang Wang, Tianyu Li, Yuhui Qiao.

Authors and Affiliations

Institute for Translational Medicine on Cell Fate and Disease, Shanghai Ninth People’s Hospital, Key Laboratory of Cell Differentiation and Apoptosis of National Ministry of Education, Department of Pathophysiology, Shanghai Jiao Tong University School of Medicine, Shanghai, China

Wan Fu, Jianghuang Wang, Yuhui Qiao, Zili Zhang, Xiaomin Zhang, Mingkai He, Yan Su, Ziye Zhao, Shen Zhang & Qing Zhong

Shanghai Key Laboratory of Maternal Fetal Medicine, Shanghai Institute of Maternal-Fetal Medicine and Gynecologic Oncology, Department of Anesthesiology, Shanghai First Maternity and Infant Hospital, School of Medicine, Tongji University, Shanghai, China

Tianyu Li & Zhiqiang Liu

Anesthesia and Brain Function Research Institute, Tongji University School of Medicine, Shanghai, China

Tianyu Li & Zhiqiang Liu

State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai, China

Tianyu Li & Yang Li

University of Chinese Academy of Sciences, Beijing, China

Tianyu Li & Yang Li

Zhiyuan College, Shanghai Jiao Tong University, Shanghai, China

Ziye Zhao

Network and Information Center, Shanghai Jiao Tong University, Shanghai, China

Chen Li & James Lin

School of Chinese Materia Medica, Nanjing University of Chinese Medicine, Nanjing, China

Ronghua Xiao, Yujun Han & Yang Li

Hainan Academy of Medical Sciences, School of Basic Medicine and Life Science, Hainan Medical University, Haikou, China

Guoqiang Chen

National Clinical Research Center for Aging and Medicine, Huashan Hospital, Fudan University, Shanghai, China

Yang Li

Contributions

W.F. and J.W. performed most of the biological and biochemical experiments. T.L. performed electrophysiological experiments with the help of R.X., Y.H. and Z.L. J.W. performed the molecular docking, MD simulations and mutagenesis analysis. Y.Q. and Z. Zhang performed the chemical inhibitor screening and helped with some of the cell experiments. X.Z., M.H., Y.S. and Z. Zhao helped with the cell biological experiments. C.L. performed the CRISPR screening data analysis with guidance of J.L. S.Z. and G.C. helped in data analysis. W.F. and Q.Z. conceptualized the project and designed the experiments. W.F., J.W. and Q.Z. analyzed the data and wrote the paper with the help of all authors. Y.L. and Q.Z. supervised the project. All authors discussed the results and commented on the paper.

Corresponding authors

Correspondence to Yang Li or Qing Zhong.

Ethics declarations
Competing interests

The authors declare no competing interests.

Peer review
Peer review information

Nature Chemical Biology thanks János Almássy, Kengo Watanabe and the other, anonymous reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data
Extended Data Fig. 1 Different death-inducing effect of the stereoisomers NC1 and NC1i.

a, Chemical structure of compound NC1 and NC1i. b, IC50 of NC1 and NC1i in various cell lines treated for 48 h. c, Live-cell images of MCF7 cells treated with 50 nM NC1 in medium containing SytoxGreen. Scale bars, 25 μm. d, Heatmap of viability of cells pretreated for 1 h with 20 μM Z-VAD-FMK (Z-VAD), 10 μM Necrostatin-1 (Nec-1), 50 μM Chloroquine (CQ), 50 μM 3-Methyladenine (3-MA), 1 μM Ferrostatin-1 (Fer-1), 1 μM Liproxstatin-1 (Liprox-1), or 100 μM Deferoxamine (DFO) and then treated with either 100 nM NC1, RSL3, 50 ng/ml TNF-α + 1 μM Smac Mimetic (TS), or TS + 20 μM Z-VAD (TSZ) for 24 h. Viability data represent mean of n = 3 replicates from one representative of three independent experiments. Micrograph data show one representative out of three replicates from three independent experiments.

Source data

Extended Data Fig. 2 NC1, rather than NC1i, triggers TRPM4 activation.

a, Typical whole-cell current traces recorded from HEK293T cells stably expressing hTRPM4 channel and stimulating voltage steps used in the experiment is on the top (left panel). Subsequently, whole-cell current traces recorded after 1 μM NC1 application (middle panel), followed by inhibition with 50 μM 9-ph inhibitor (right panel). Three figures were recorded from one same patch. b, The current density-voltage relation of hTRPM4 steady-state current with vehicle or NC1i application. c, Statistics of hTRPM4 steady-state whole-cell current density with vehicle, NC1i or NC1 application at +100 mV in b. d-g, Typical current traces of hTRPM4 channel treated vehicle or 1 μM NC1 by using outside-out (d) or inside-out (f) patch. Statistics of hTRPM4 outside-out (e) or inside-out (g) current. h, The time course of TRPM4 current-densities in HEK293T-TRPM4 cells activated by 1 μM NC1 was followed by NC1 washout using bath solution after 420 seconds. The whole-cell currents were recorded at −100 and +100 mV. Electrophysiological data represent mean ± sem of (n = 5 for b, 11 for NC1 and 5 for NC1i in c, 8 for e, 4 for g, 8 for h) independent cells. Two-way ANOVA with post Sidak test (GraphPad Prism 9.0.0).

Source data

Extended Data Fig. 3 NECSO is accompanied with Membrane depolarization and cell edema.

a, Dose-dependent cytotoxicity of NC1 in MCF7 cells pretreated with 20μM NKCC inhibitors (Furosemide, Bumetanid and Azosemide) or KATP inhibitors (PNU 37883A and Glipizide). b, A typical membrane potential of MCF7 cells treated with 1 μM NC1. c, Statistics of membrane potential values of MCF7 cells recorded 10 minutes after exposed to vehicle or NC1 in b. d, Representative confocal images of MCF7 cells loaded with 5 μM DiBAC4(3) followed by treatment with 50 nM NC1 for 1.5 h. e, Statistics of fold change for DiBAC4(3) fluorescence intensity, mean ± s.d. Each dot represents an individual image in d. f, Representative confocal images of MCF7 cells treated with DMSO, 50 or 100 nM NC1 for 3 h and then loaded with chloride indicator MQAE. Scale bars, 10 μm. g, Statistics of fold change MQAE fluorescence intensity, mean ± s.d. Each dot represents an individual image in f. h, Dose-dependent cytotoxicity of NC1 in MCF7 cells pretreated with multiple inhibitors of chloride anion channels. i, Transmission electron microscopy (TEM) of MCF7 cells treated with 100 nM NC1 at different time points as indicated. Scale bars, 500 nm. Red arrowheads in images of 0.5, 1.5 and 2.5 h represented swollen mitochondria, endoplasmic reticulum and Golgi apparatus respectively, and highlighted swollen nuclear membrane (above) and broken cell membrane (below) in images of 4 h. j, Cell death assay measured by LDH release of MCF7 cells treated with NC1 for 24 h in medium containing different doses of osmo-protectant D-mannitol. Viability and cell death data represent mean ± s.d. of n = 3 replicates from one representative of three independent experiments. Electrophysiological data represent mean ± sem of n = 10 independent cells. Micrograph data show one representative out of three replicates from three independent experiments. One-way ANOVA for statistics in c, e, g (GraphPad Prism 9.0.0).

Source data

Extended Data Fig. 4 Different responses to NC1 of TRPM4 from multiple species.

a, IC50 of NC1 in various cell lines of different species treated for 48 h. b, Representative confocal images of HeLa cells with expression of various species TRPM4, treated with 1 μM NC1 for 24 h. PI staining indicates the dead cell. c, Statistics of the percentage of NC1-responsive cells in TRPM4-EGFP transfected cells in b, mean ± s.d. from 50 cells from three independent experiments. d, Summary results of sensitivity to NC1-induced current density of different species TRPM4 at +100 mV in whole-cell patch clamp experiments. e, Representative confocal images of HeLa cells transfected with different TRPM family protein, subsequently treated with 1 μM NC1 for 24 h. PI staining indicates the dead cell. f, Histogram summarizes the percentage of NC1-responsive cells in TRPM4-EGFP transfected cells in e, mean ± s.d. from 50 cells from three independent experiments. Micrograph data show one representative out of three replicates from three independent experiments.

Source data

Extended Data Fig. 5 Distinct responses to NC1 of TRPM4 human-mouse chimeras.

a, Schematic representation of the three major domains and their corresponding amino acid sequence numbers in hTPPM4 (upper panel). Each structural component is labeled and color-coded in detail (lower panel). b, Representative confocal images of HeLa cells transfected with different TRPM4 chimera, treated with 1 μM NC1 for 24 h. c, Histogram summarizes the percentage of NC1-responsive cells with TRPM4-EGFP expression in b, mean ± s.d. from 50 cells from three independent experiments. Micrograph data show one representative out of three replicates from three independent experiments.

Source data

Extended Data Fig. 6 Model representations of ligand binding sites in vanilloid binding pocket (VBP).

The overall architecture of the indicated TRP channels, and one subunit is highlighted. Ligands are presented with spheres (upper panel). Ligands are shown as sticks. Oxygen is shown in red, nitrogen in blue, sulfur in yellow, fluorine in light blue, bromine in dark red, and chloride in green (lower panel).

Extended Data Fig. 7 Different cross-sections of Ligand NC1 binding pocket in hTRPM4 (PDB: 6BQV).

a, b, Overall surface presentation of TRPM4. Ligand NC1 is shown in green spheres presentation. Subunits are individually colored with gray black or gray white. The ligand binding pocket of NC1 is composed of residues from two adjacent subunits. The view is from the front (a) or the top (b).

Extended Data Fig. 8 Molecular dynamics (MD) simulation analyses of the binding of NC1 to TRPM4.

a, The simulated state diagram of the TRPM4-NC1 complex embedded in the cell membrane phospholipid bilayer. b, Summary results of structural stability analyses (root mean square deviation, RMSD) of channel-ligand complex following NC1 binding to TRPM4 channel in 100 ns bouts. c, Interaction energy calculations for NC1 binding to TRPM4 channel on trajectories with convergent conformations. d, Information on the binding pockets of NC1 with the TRPM4 channel was calculated after MD simulations. e, Energy contribution of residues within a 5Å distance of NC1. The mean values of energy are represented by blue bars, while the black lines depict the standard error.

Source data

Extended Data Fig. 9 NC1 activation of TRPM4 per se is not affected by calcium.

a, b, c, The conductance-voltage relation of whole-cell hTRPM4 current in HEK293T-TRPM4-EGFP cells treated vehicle or 1 μM NC1 with different concentrations of EGTA (0, 1, 10 mM) added in the patch pipette solutions. d, Statistics of whole-cell hTRPM4 current density at +100 mV in a, b, c. e, The current density-voltage relation of whole-cell hTRPM4 current in HEK293T-TRPM4 cells treated vehicle or 1 μM NC1 in the presence or absence of 1 μM intracellular Ca2+. NC1 further increase the current amplitude of TRPM4 with the presence of intracellular calcium. f, g, Dose-dependent cytotoxicity of NC1 in MCF7 cells pretreated with calcium ion chelators BAPTA-AM (f) or EGTA-AM (g). h, i, The whole-cell current-voltage relation of hTRPM4 E828K (h) or E1068Q (i) mutant expressed in HeLa cells (right panel). Summary data showing the maximum amplitude of currents activated by NC1 at +100 mV (left panel). j, Dose-dependent cytotoxicity of NC1 in TRPM4-KO MCF7 cells with re-expression of TRPM4 WT or mutants above. Viability data represent mean ± s.d. of n = 3 replicates from one representative of three independent experiments. Electrophysiological data represent mean ± sem of (n = 7 for a, e and i, 5 for b and h, 6 for c) independent cells. Immunoblot data show one representative out of three replicates from three independent experiments. Two-way ANOVA (GraphPad Prism 9.0.0) for statistics in d, h, i.

Source data

Extended Data Fig. 10 NECSO inhibitors protect cells from energy depletion.

Heatmap depicting the cell death (measured by LDH release) of Cos7 cells stably expressing hTRPM4 which were pre-treated with different inhibitors followed by 2DG + NaN3 or NC1 exposure respectively for 16 h (left). Representative bright field images of the cellular morphology corresponding to each treatment condition as described above (right). Scale bars, 20 μm. Cell death and micrograph data show one representative out of three replicates from three independent experiments.

Source data

Supplementary information
Supplementary Information

Supplementary Figs. 1–8 and Tables 1 and 2.

Reporting Summary
Supplementary Video 1

Upon NC1 treating, MCF7 cells exhibited typical morphological characteristics of necrosis.

Supplementary Video 2

The X and Y roll presentation of NC1 binding to the VBP pocket in hTRPM4 (PDB 6BQV).

Supplementary Data 1

The docking model of NC1/CLT binding to hTRPM4.

Supplementary Data 2

The docking model of NC1/CLT binding to hTRPM4.

Supplementary Data 3

Statistical source data for Supplementary Fig. 1.

Supplementary Data 4

Statistical source data for Supplementary Fig. 2.

Supplementary Data 5

Statistical source data for Supplementary Fig. 3.

Supplementary Data 6

Statistical source data for Supplementary Fig. 4.

Supplementary Data 7

Statistical source data for Supplementary Fig. 5.

Supplementary Data 8

Statistical source data for Supplementary Fig. 6.

Supplementary Data 9

Uncropped scans of blots and gels in figures, supplementary figures and extended data.

Source data
Source Data Fig. 1

Statistical source data.

Source Data Fig. 2

Statistical source data.

Source Data Fig. 3

Statistical source data.

Source Data Fig. 4

Statistical source data.

Source Data Fig. 5

Statistical source data.

Source Data Fig. 6

Statistical source data.

Source Data Extended Data Fig. 1

Statistical source data.

Source Data Extended Data Fig. 2

Statistical source data.

Source Data Extended Data Fig. 3

Statistical source data.

Source Data Extended Data Fig. 4

Statistical source data.

Source Data Extended Data Fig. 5

Statistical source data.

Source Data Extended Data Fig. 8

Statistical source data.

Source Data Extended Data Fig. 9

Statistical source data.

Source Data Extended Data Fig. 10

Statistical source data.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article
Cite this article

Fu, W., Wang, J., Li, T. et al. Persistent activation of TRPM4 triggers necrotic cell death characterized by sodium overload. Nat Chem Biol 21, 1238–1249 (2025). https://doi.org/10.1038/s41589-025-01841-3

Download citation

Received
20 October 2023

Accepted
09 January 2025

Published
06 February 2025

Version of record
06 February 2025

Issue date
August 2025

DOI
https://doi.org/10.1038/s41589-025-01841-3

Share this article

Anyone you share the following link with will be able to read this content:

Get shareable link

Provided by the Springer Nature SharedIt content-sharing initiative

Subjects
Cell death
Transient receptor potential channels
This article is cited by
Exploiting metabolic cell death for cancer therapy
Chao MaoDadi JiangBoyi Gan

Nature Reviews Cancer (2026)

Targeting metalloptosis in tumor therapy: from molecular mechanisms to application of metal nanoparticles
Jinxiang WangQin TianYihang Pan

Molecular Cancer (2025)

FGD3 mediates lytic cell death, enhancing efficacy and immunogenicity of chemotherapy agents in breast cancer
Junyao ZhuXinyi DaiDavid J. Shapiro

Journal of Experimental & Clinical Cancer Research (2025)

Sodium disrupts mitochondrial energy metabolism to execute NECSO
Yuhui QiaoJianghuang WangQing Zhong

Nature Communications (2025)

Structural landscape of activation, desensitization and inhibition in the human TRPM4 channel
Celso M. Teixeira-DuarteWeizhong ZengYouxing Jiang

Nature Structural & Molecular Biology (2025)

Cancel

Preview · live