Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Aug 8;32(8):1137-1149.e4.
doi: 10.1016/j.str.2024.05.005. Epub 2024 May 29.

Structural basis for inhibition of the lysosomal two-pore channel TPC2 by a small molecule antagonist

Affiliations

Structural basis for inhibition of the lysosomal two-pore channel TPC2 by a small molecule antagonist

Gamma Chi et al. Structure. .

Abstract

Two pore channels are lysosomal cation channels with crucial roles in tumor angiogenesis and viral release from endosomes. Inhibition of the two-pore channel 2 (TPC2) has emerged as potential therapeutic strategy for the treatment of cancers and viral infections, including Ebola and COVID-19. Here, we demonstrate that antagonist SG-094, a synthetic analog of the Chinese alkaloid medicine tetrandrine with increased potency and reduced toxicity, induces asymmetrical structural changes leading to a single binding pocket at only one intersubunit interface within the asymmetrical dimer. Supported by functional characterization of mutants by Ca2+ imaging and patch clamp experiments, we identify key residues in S1 and S4 involved in compound binding to the voltage sensing domain II. SG-094 arrests IIS4 in a downward shifted state which prevents pore opening via the IIS4/S5 linker, hence resembling gating modifiers of canonical VGICs. These findings may guide the rational development of new therapeutics antagonizing TPC2 activity.

Keywords: SG-094; TPC2; antagonist; cryo-EM; electrophysiology; ion channel; structural biology; two-pore channel; voltage-sensing domain.

PubMed Disclaimer

Conflict of interest statement

Declaration of interests The authors declare no competing interests.

Figures

None
Graphical abstract
Figure 1
Figure 1
Structural overview of HsTPC2 in complex with (S)-SG-094 (A and B) Overall structure of HsTPC2 in complex with (S)-SG-094. Light green, Subunit A (in complex with SG-094); Light cyan, Subunit B without (S)-SG-094; Gray – HsTPC2 in apo state (PDB: 6NQ1). For the cytoplasmic view (B), structural differences between apo and (S)-SG-094-bound HsTPC2 are marked with arrows. (C) Close-up view of the (S)-SG-094 binding pocket located at the interface between S5 in the pore domain I of subunit B (light cyan) and S1 from voltage-sensing domain II of subunit A (light green). (S)-SG-094 (pink stick representation) is fitted into ESP map (blue mesh, σ = 5.0). Subunit A residues in proximity of SG-094 are shown as light green sticks. (D and E) Sequence alignment of regions IIS1 (D) and IIS4 (E) in VSDII with other members of the TPC family. Gating charge residues R1-R5 are highlighted in light pink, and residues in the (S)-SG-094 binding pocket are highlighted in blue. (F) Chemical structure of SG-094.
Figure 2
Figure 2
SG-094 traps HsTPC2 in a closed state through rearrangements of IIS4 in VSD II and the IIS4/S5 linker (A–G) Comparison of the VSD II arrangement (A and D–G) and S4/S5 linker II arrangement (B and C) of the inhibited subunit A (light green), the inhibitor-free subunit B (light cyan) in the (S)-SG-094-bound HsTPC2 structure and the respective arrangements observed in the two identical subunits of the closed (apo, yellow, PDB: 6NQ1) or PI(3,5)P2-bound activated structure (orange, PDB: 6NQ0). (A) Comparison of the lumenal side of VSD of HsTPC2 structures. Green, subunit A; Light cyan, subunit B; Yellow, Apo-HsTPC2 (PDB: 6NQ1). For the IIS3/4 loop region, missing model due to flexibility is marked with blue (subunit B) or orange (apo state) triangles. (B) Cytoplasmic view of subunit A (green) shows IIS4/5 in closed-like state. Yellow, Apo-HsTPC2; Orange, PI(3,5)P2-bound HsTPC2 in open state (PDB: 6NQ0). (C) Cytoplasmic view of subunit B (blue) shows IIS4/5 in open-like state. (D) VSD II domain of subunit A (light green) shows shift of IIS4 helix with voltage-sensing residues (R545–R554) C-terminally shifting by a full turn compared to apo-HsTPC2 (light gray). (E) VSD II of subunit B (light cyan) is in the same state as apo-HsTPC2 (light gray). (F and G) Schematic diagrams for simplified views of HsTPC2’s VSD II in resting state including apo, PI(3,5)P2-bound closed and PI(3,5)P2-bound open states (F), and SG-094-inhibited state (G). IIS1 is omitted for clarity.
Figure 3
Figure 3
Comparison of the SG-094 binding site in HsTPC2 to other small molecule modulators targeting VSDs in TPC, Nav, and Kv channels (A) Antagonist trans-Ned-19 in AtTPC1 (PDB: 5DQQ) binds on the other side (extracellular) of VSD II/pore domain interface compared to (S)-SG-094. Agonists PI(4,5)P2 in HsKv7.4 (PDB: 8BYL) and LuAG00563 in HsKv3.1 (PDB: 7PQU) bind at the same site as (S)-SG-094. Inhibitors GX-936 for HsNav1.7 (PDB: 5EK0) and ProTx2 for HsNav1.7 (PDB: 6N4R) bind at S2/S3 pocket on the extracellular side. (B and C) (S)-SG-094 binds at the same site at VSD II as PI(4,5)P2 does at the VSD of Kv7.4. (D and E) Experimentally determined (S)-SG-094 model (pink) matches closely with in silico docked models of (S)-SG-094 (teal, cyan). (F) Extracellular view onto VSDII in the HsNav1.7/NavAb chimera (PDB: 5EK0), illustrating binding of antagonist GX-936 to the cleft between IIS1 and IIS3. (G) Side view of VSDII of HsNav1.7/NavAb in complex with GX-936 superposed with VSDII from the HsTPC2/(S)-SG-094 complex structure. (H) Cartoon schematic of VSDII of HsNav1.7/NavAb in complex with GX-936, illustrating the positioning of voltage-sensing residues R1-R5 (in IIS4) with respect to F1547 of the CTC in IIS2. In contrast to SG-094 which stabilizes IIS4 of TPC2 in a downward-shifted state, GX-936 arrests Nav1.7 in a fully activated (upward shifted) state of IIS4, which leads to Nav1.7 antagonism via inactivation. (I) Cytoplasmic view of VSD of HsKv3.1 in complex with LuAG00563. (J) Side view of VSD of HsKv3.1 in complex with LuAG00563. (K) Cartoon schematic of VSD of HsKv3.1 in complex with LuAG00563, illustrating the positioning of voltage-sensing residues R1-R5 with respect to F256. S4 is in upward shifted state, in line with LuAG00563’s positive modulatory effect on HsKv3.1.
Figure 4
Figure 4
Decreased HsTPC2 inhibition by SG-094 for binding site mutations Y436A, N439A, and F555A in Ca2+ imaging experiments (A–F) The curve graphs are representative Ca2+ signals recorded from HEK293 cells transiently transfected with plasma-membrane HsTPC2L11A/L12A-eYFP variants. The cells were loaded with the ratiometric Ca2+ indicator Fura-2 and stimulated with either 10 μM TPC2-A1-N (A, B, E, and F) or 30 μM TPC2-A1-P (C and D). Blue curves represent control measurements (mean values), where 0.1% DMSO was applied instead of SG-094 before stimulation, orange curves demonstrate the effects caused by the addition of 20 μM SG-094 (mean values). Transfected single-cell traces are shown in light gray, untransfected cell traces are shown in dark gray. Experiments were performed at least in triplicates and statistical analyses of the maximal changes in the Fura-2 ratio (mean ± SEM, unpaired t test using GraphPad Prism 9.0.2, ∗∗∗p < 0.001) are shown within the bar charts.
Figure 5
Figure 5
Loss of TPC2 current inhibition by SG-094 for binding site mutant Y436A in whole-cell patch-clamp recordings Representative current density-voltage (I/Cm-V) relation of transiently expressed, plasma-membrane-targeted HsTPC2L11A/L12A-eYFP variants, WT (A) and Y436A mutant (B). Channels were activated by application of TPC2-A1-N (10 μM, blue traces), followed by application of the TPC antagonist SG-094 (10 μM, orange traces). Statistical analysis of experiments is shown in (C), with each dot representing mean of 5–10 technical replicate measurements (mean ± SEM; n = 7–8 independent experiments; one-way ANOVA, Tukey’s post hoc test using GraphPad Prism 9.0.2, ∗∗∗∗p < 0.0001, n.s. - not significant).
Figure 6
Figure 6
SG-094 binding causes asymmetrical reorganization of VSD II resulting in a single antagonist binding cleft within the HsTPC2 dimer (A) Several residues (Y432, Y436, M562, and K563) in S4 and S1 of the VSD II of both subunits reposition in response to (S)-SG-094 binding to HsTPC2. Gray, apo-TPC2 (PDB: 6NQ1); Green, SG-094-bound subunit A; Light blue, subunit B without (S)-SG-094. Red arrows, residue displacement from apo state to subunit A of (S)-SG-094-bound state; orange arrows, residue displacement from apo state to subunit B of (S)-SG-094-bound state. Electrostatic/steric clash between K563’s amine group and (S)-SG-094’s tertiary amine is marked with blue triangle. (B) Interaction between II S1/2 (light green) and II S3/4 (dark green) loops on the extracellular side of VSD II in subunit A. (C–E) APBS-generated electrostatic surface representations (−5.0 to 5.0 kT/e) of (S)-SG-094-binding sites. (C) In subunit A, the binding pocket is open to accommodate (S)-SG-094, and the surface is generally electroneutral. (D) In apo-TPC2, Y432 and K563 would sterically clash with hypothetical (S)-SG-094, and the environment is more hydrophilic than (S)-SG-094-bound state. (E) In subunit B, the binding site has fully closed due to the movement of M562 and K563, making it sterically difficult for (S)-SG-094 to bind.
Figure 7
Figure 7
Comparison of (S)-SG-094-bound HsTPC2 to open and closed state structures of HsTPC2 (A) Overall structural comparison between the inhibitor-bound HsTPC2 (subunit A: green, subunit B: cyan), closed state HsTPC2 (PDB: 6NQ0; yellow), and open state HsTPC2 (PDB: 6NQ1; orange). Key sites of interest highlighted in the next panels are marked with arrows and labels. (B) Close-up comparison of the IS6 and EF hand domain of HsTPC2 structures. Only open state TPC2 has continuous helix from IS6 to EF1 in response to PI(3,5)P2 binding, with the other three being similar. (C) Close-up comparison of VSD I domain. All four compared structures have the same conformation. (D) Close-up comparison of pore region. All three compared structures have similar conformations, with only the open state having slightly dilated IS6 toward the cytoplasmic side. (E) Close-up views of VSD for uninhibited subunit B against open and closed states. Voltage sensor residues for all three structures are in same positions. (F) Close-up views of the lumenal side of VSD for subunit B against open and closed states. All three have similar structures. (G) Close-up views of the cytoplasmic side of VSD for subunit B against open and closed states. IIS0 for subunit B is shorter than the other two models and has slightly rotated away from the protein center (blue). IIS4/5 helix of subunit B (blue) is more aligned with IIS4/5 of open state model (darker orange) than the closed one (light yellow). (H) Close-up views of VSD for (S)-SG-094-bound subunit A against open and closed states. Voltage sensor residues for subunit A has shifted downward by one turn. (I) Close-up views of the lumenal side of VSD for subunit A against open and closed states. IIS3/4 linker of subunit A (bright green) is ordered whereas the models for the other two structures are missing due to flexibility. (J) Close-up views of the cytoplasmic side of VSD for subunit A against open and closed states. IIS0 for all three structures are in similar states. IIS4/5 helix of subunit A aligns well with the closed state model, and has an extra turn on the N-terminal side.
Figure 8
Figure 8
SG-094-bound HsTPC2 reveals a second lipid binding site, distinct from PI(3,5)P2 (agonist) binding site (A) Overall structure of HsTPC2 denoting the locations of the two phospholipids (PL1, black; PL2, gray). Light green, (S)-SG-094-bound subunit A; Light cyan, subunit B without SG-094; Cyan, VSD I domain. (B) Cytoplasmic view of the interface between VSD I and pore domain shows the head groups of PL1 and PL2 forming hydrophilic interactions with HsTPC2 residues. (C) PL1’s hydrophobic tail extends to a hydrophobic pocket formed by I S4/5, I S5, and II S5. (D) PL2’s hydrophobic tail extends to a hydrophobic region bounded by I S1, I S4, and II S5. (E and F) Overlaying PI(3,5)P2 in an open HsTPC2 structure (PDB: 6NQ0) shows it with a similar binding mode to PL1.

References

    1. Pitt S.J., Reilly-O'Donnell B., Sitsapesan R. Exploring the biophysical evidence that mammalian two-pore channels are NAADP-activated calcium-permeable channels. J. Physiol. 2016;594:4171–4179. doi: 10.1113/jp270936. - DOI - PMC - PubMed
    1. Grimm C., Hassan S., Wahl-Schott C., Biel M. Role of TRPML and two-pore channels in endolysosomal cation homeostasis. J. Pharmacol. Exp. Ther. 2012;342:236–244. doi: 10.1124/jpet.112.192880. - DOI - PubMed
    1. Pitt S.J., Funnell T.M., Sitsapesan M., Venturi E., Rietdorf K., Ruas M., Ganesan A., Gosain R., Churchill G.C., Zhu M.X., et al. TPC2 is a novel NAADP-sensitive Ca2+ release channel, operating as a dual sensor of luminal pH and Ca2+ J. Biol. Chem. 2010;285:35039–35046. doi: 10.1074/jbc.M110.156927. - DOI - PMC - PubMed
    1. García-Rúa V., Feijóo-Bandín S., Rodríguez-Penas D., Mosquera-Leal A., Abu-Assi E., Beiras A., María Seoane L., Lear P., Parrington J., Portolés M., et al. Endolysosomal two-pore channels regulate autophagy in cardiomyocytes. J. Physiol. 2016;594:3061–3077. doi: 10.1113/jp271332. - DOI - PMC - PubMed
    1. Lin P.H., Duann P., Komazaki S., Park K.H., Li H., Sun M., Sermersheim M., Gumpper K., Parrington J., Galione A., et al. Lysosomal two-pore channel subtype 2 (TPC2) regulates skeletal muscle autophagic signaling. J. Biol. Chem. 2015;290:3377–3389. doi: 10.1074/jbc.M114.608471. - DOI - PMC - PubMed

Publication types

LinkOut - more resources