Research ArticleIon Channels

Transmembrane helix connectivity in Orai1 controls two gates for calcium-dependent transcription

See allHide authors and affiliations

Science Signaling  28 Nov 2017:
Vol. 10, Issue 507, eaao0358
DOI: 10.1126/scisignal.aao0358
  • Fig. 1 Constitutively active Orai1 Ca2+ signals and NFAT activation induced by mutations detected through cancer database screening.

    (A) Model of the human Orai1 protein structure, with Orai1 single-point mutants found in human tumors in conserved positions highlighted, as determined from large-scale cancer genomics data sets. (B) Percentage of RBL cells that express nuclear factor of activated T cells (NFAT)–driven red fluorescent protein (RFP) (average + SEM), measured 24 hours after cotransfection of the NFAT reporter and cancer-associated Orai1 mutants from (A), wild-type (WT) Orai1, or constitutively active Orai1-V102A and Orai1-L138F mutants. Significantly increased number of RBL cells expressing NFAT-driven RFP (P < 0.05 by t test) is indicated by the red bars and a star. Analogous experiments were performed in a nominally Ca2+-free bath solution for coexpression of NFAT reporter and WT Orai1 or Orai1-L138F, indicated by blue bars (n = 3 to 8 cell images, from at least three separate transfections). (C) Representative images of the coexpression of two cancer-associated Orai1 mutants (Orai1-A137V and Orai1-G247S) and WT Orai1 with the NFAT reporter in RBL mast cells. Scale bars, 10 μm. (D) Resting cytosolic Ca2+ concentrations in Fura-2–loaded human embryonic kidney (HEK) cells overexpressing WT Orai1 or cancer-associated Orai1 mutants (n = 31 to 70 cells, from 3 to 4 individual transfections). Significantly increased Ca2+ concentrations (P < 0.05 by t test) are indicated by red bars and a star. (E) Time course of whole-cell patch-clamp experiment in HEK cells coexpressing STIM1 and Orai1 (black), Orai1-A137V (blue), or Orai1-A137V alone (red) in a 10 mM extracellular Ca2+ (n = 7 to 10 cells, from at least two individual transfections). Store-operated currents were activated by 20 mM EGTA in patch pipette. (F to H) Representative current-voltage relationships for STIM1- and Orai1-expressing cells (F), Orai1-A137V (G), or STIM1 and Orai1-A137V (H) in a 10 mM Ca2+ (black) or Na+-based, divalent-free (Na-DVF, red) solution.

  • Fig. 2 Interaction between a central TM2 segment and the pore helix controls Orai1 channel gating.

    (A) Details of the central transmembrane 2 (TM2) segment in Orai1 mutants, including pathophysiological and cancer-associated mutations (A137V, L138F, and M139V), near TM1 and TM3 (Gly183). A hydrogen bond is observed between His134 and Ser93. (B and C) Number of hydrogen bonds between (B) Ser93 and His134 and (C) Ser97 and His134 in WT Orai1 as determined by molecular dynamics simulations. (D) Top view of a representative image of molecular dynamics simulations of the WT Orai1 pore, with the side chains of Arg91, His134, and Leu138 highlighted. (E, G, and H) Similar snapshots are shown for molecular dynamics simulations of Orai1-H134A (E), Orai1-L138F (G), and Orai1-A137V (H). (F) Percentage of RBL cells that express NFAT-driven RFP (average + SEM), measured 24 hours after cotransfection of the NFAT reporter and various Orai1-His134 mutants or WT Orai1. Cells were in a bath solution containing 2 mM Ca2+. Significantly increased numbers of cells (P < 0.05 by t test) positive for NFAT-driven RFP are indicated by a red bar and a star. Analogous experiments were performed in a nominally Ca2+-free bath solution to detect the coexpression of the NFAT reporter and WT Orai1 or Orai1-H134A, indicated by blue bars (n = 3 to 5 cell images, each from a separate transfection).

  • Fig. 3 His134 mutations switch Orai1 between generating constitutively active, store-dependent, or suppressed currents.

    (A) Time course of whole-cell patch-clamp recordings of HEK cells overexpressing Orai1-His134 (Ala, Cys, Glu, Gly, Met, and Val) mutants. These mutants show constitutive activity (n = 7 to 14 cells, from at least two individual transfections). (B) Analogous experiments to those conducted on Orai1-His134 mutants as in (A) or WT Orai1 coexpressed with STIM1 (n = 7 to 14 cells, from at least two individual transfections). (C and D) Time courses for store-dependent activation for Orai1-His134 (Gln and Asn) mutants or WT Orai1 coexpressed with STIM1 (D) or reduced currents generated by Orai1-His134 (Phe, Trp, Tyr, and Leu) coexpressed with STIM1 (n = 6 to 14 cells, from at least two individual transfections). All currents were recorded at −86 mV in a 10 mM Ca2+-containing bath solution, and store depletion was induced by 20 mM EGTA in the pipette. (E) Representative images of cyan fluorescent protein (CFP)–ORAI1-activating small fragment (OASF) and yellow fluorescent protein (YFP)–Orai1 WT (top) and Orai1-H134W mutant (bottom). Scale bars, 10 μm. (F) Intensity plots for regions close to the plasma membrane measured at the dashed lines for individual cells (E) (n = 4 to 16 cells, from at least two individual transfections). (G) Representative current-voltage relationships for maximum currents of Orai1-His134 mutants of (A).

  • Fig. 4 Activation of autophagy and mitophagy transcription factor by Orai1 mutants identified via cancer database screening and myopathy mutants.

    (A) Time course of transcription factor activation by thapsigargin (TG) treatment (1 μM, 5 min) as measured by cytosol-to-nuclear translocation for NFAT (red), transcription factor EB (TFEB) (blue), and microphthalmia-associated transcription factor (MITF) (black) in STIM1 and Orai1 coexpressing HEK cells (n = 6 to 8 cell images, each from three separate transfections). (B and C) Representative images of HEK cells coexpressing YFP-tagged Orai1-A137V (B) or YFP-tagged Orai1-H134A (C) and CFP-MITF in the absence or presence of 2 mM extracellular Ca2+. Scale bars, 10 μm. (D) Average number of HEK cells exhibiting nuclear MITF localization upon coexpression with cancer-associated Orai1 mutants, a myopathy-associated Orai1-L138F mutant, or the Orai1-H134A mutant determined after 24 hours with or without 2 mM Ca2+ in the media (n = 6 to 8 cell images, each from three to five separate transfections). Additional bar indicates similar experiments, in which cells were exposed to Ca2+-containing media for 4 hours (n = 6 to 8 cell images, each from three separate experiments). Significantly increased (P < 0.05 by t test) nuclear localization of MITF induced by Orai1 mutants with respect to Ca2+ in the media is indicated by a star. (E and F) Representative images of HEK cells coexpressing YFP-tagged Orai1-A137V (E) or Orai1-H134A (F) and CFP-TFEB in the absence or presence of 2 mM extracellular Ca2+. (G) Average number of HEK cells exhibiting nuclear TFEB localization upon coexpression with WT Orai1, Orai1-L138F, Orai1-A137V, or Orai1-H134A determined after 24 hours with or without 2 mM Ca2+ in the media (n = 3 to 5 cell images, each of individual transfection). Additional bar indicates similar experiments in which cells were exposed to Ca2+-containing media for 4 hours (n = 6 to 8 cell images, each from three separate transfections). Significantly increased nuclear localization (P < 0.05 by t test) of TFEB induced by Orai1 mutants with respect to Ca2+ in the media is indicated by a star. (H) Autophagosomes were visualized by green fluorescence protein (GFP)–2xFYVE in HEK cells expressing YFP-tagged WT Orai1, Orai1-H134A, or Orai1-A137V. Scale bars, 10 μm. (I) Quantification of the average number of GFP-2xFYVE clusters in the presence or absence of 2 mM Ca2+ in the bath (n = 70 to 90 cells, from three individual transfections). Significantly increased (P < 0.05 by t test) autophagosome formation is indicated by a star.

  • Fig. 5 Increased pore size in the open conformation of Orai1-H134A channels.

    (A and C) Representative snapshot (time point at 192 ns) of the equilibrated part of 200-ns-long molecular dynamics simulations for WT Orai1 (A) or Orai1-H134A (C) showing the pore-forming TM1 helices (four of six TM1 helices shown in red) and pore-lining residues from Glu106 to Trp76. The pore surface is shown in blue (radius > 1.15 Å) and in green (radius: 0.6 to 1.15 Å). (B) Average pore radius of WT Orai1 (red and magenta) and Orai1-H134A (blue) for a 2-ns bin corresponding to the time points shown in (A) and (C). (D) Fluctuations of individual TM1 residues of (black) WT Orai1 and (red) Orai1-H134A are measured as the root mean square (RMS) deviation of Cα atom for the last 50 ns of the 200-ns-long molecular dynamics simulations. (E) Representative Western blot of a cross-linking experiment overexpressing Orai1-R91C and Orai1-H134A-R91C showing monomer and dimer formation. (F and G) Dimerization efficiency in % of cysteine cross-linking for engineered cysteines (R83C to E106C) in the (F) WT Orai1 or (G) Orai1-H134A TM1 pore segment. A cysteine-free Orai1 background was used. For each cysteine position, parallel experiments (n = 5 to 8 individual transfections for each cysteine position) with WT Orai1 and Orai1-H134A background were performed on the same day, and significant differences (P < 0.05 by t test) are indicated by a star.

  • Fig. 6 Side-chain twist of the Arg91 gate induced by constitutively active Orai1-H134A channels.

    (A and B) Representative snapshot of the equilibrated part of the 200-ns-long molecular dynamics simulations for WT Orai1 (A) and Orai1-H134A (B) showing the position of Arg91 residues and Ser90 in the TM1 helices, which are surrounded by TM2 helices (His134 or Ala134 respectively shown) from a top view. (C and D) Hydrogen bonds between Arg91 and Ser90 in WT Orai1 (C) and the Orai1-H134A simulations (D) in a time course (150 to 200 ns). (E and F) Comparison of pore surfaces of WT Orai1 (E) and Orai1-H134A (F) before and after Ca2+ is pulled through the pore. The pore surface is shown in blue (radius > 1.15 Å) and green (radius: 0.6 to 1.15 Å). (G and H) Representative snapshots of the start and end of 200-ns-long molecular dynamics simulations for WT Orai1 (G) and Orai1-H134A (H) illustrate the position of Phe99 residues in the pore helix.

  • Fig. 7 Formation of a chain of water molecules in the hydrophobic pore segment of Orai1-H134A.

    (A, E, and I) Representative snapshots of the equilibrated part of 200-ns-long molecular dynamics simulations for WT Orai1 (A), Orai1-R91G (E), and Orai1-H134A (I) showing the pore-forming TM1 helices (four of six TM1 helices shown in red) and pore-lining residues from Glu106 to Lys87. The pore surface is shown in blue (radius > 1.15 Å) and green (radius: 0.6 to 1.15 Å). (B, F, and J) Similar snapshots for WT Orai1 (B), Orai1-R91G (F), and Orai1-H134A (I) showing the water molecules, Ca2+ (yellow ball), Na+ (orange ball), and Cl (green ball) in the pore. Acidic residues, nonpolar residues, basic residues, and polar residues are depicted in red, orange, blue, and green, respectively. (C, G, and K) Average number and SD (shaded area) of water molecules (over the last 100 ns of respective simulations) inside the pores of WT Orai1 (C), Orai1-R91G (G), and Orai1-H134A (K) channels are calculated. (D, H, L) Water orientation inside the pores of WT Orai1 (D), Orai1-R91G (H), and Orai1-H134A (L) channels is determined as defined by the angle between the dipole vector of the water molecule and the z axis. Water molecules that locally are similarly oriented within the respective pores are depicted in brighter colors.

  • Fig. 8 Pore rearrangements in Orai1 mutants related to myopathy and identified by cancer database screening.

    (A) Representative snapshots of the equilibrated part of 200-ns-long molecular dynamics simulations for WT Orai1, Orai1-R91G, and Orai1-H134A showing the pore-forming TM1 helices (four of six TM1 helices shown in red) and pore-lining residues from Glu106 to Lys87. The pore surface is shown in blue (radius > 1.15 Å) and green (radius: 0.6 to 1.15 Å). (B) Differences among individual TM1 residues for WT Orai1, Orai1-A137V, and Orai1-L138F are measured as RMS deviation of Cα atom for the last 50 ns of 200-ns-long molecular dynamics simulations. (C and D) Dimerization efficiency (%) of cysteine cross-linking for engineered cysteines (S90C to E106C) in the WT Orai1 (C) and Orai1-L138F (D) TM1 pore segment. For each cysteine position, WT Orai1 and Orai1-L138F mutations were performed on the same day (n = 5 to 8 transfections for each cysteine position), and significant differences (P < 0.05 by t test) are indicated by a star. (E) Cross-linking results of Orai1-H134A, Orai1-L138F, Orai1-V102A, and Orai1-A137V as compared to WT Orai1. “+” and green, significantly increased cross-linking. “−” and yellow, significantly decreased cross-linking. (F and G) The time course depicts the number of hydrogen bonds formed between residues Arg91 and Ser90 for the last 50 ns of the Orai1-L138F (F) and Orai1-A137V (G) simulations.

  • Fig. 9 Water permeation in Orai1 mutants related to myopathy and identified through cancer database screening.

    (A, D, and G) Representative snapshots of the equilibrated part of 200-ns-long molecular dynamics simulations for WT Orai1 (A), Orai1-L138F (D), and Orai1-A137V (G) illustrate the pore-forming TM1 helices (four of six TM1 helices) and pore-lining residues from Glu106 to Phe76. Water molecules, Ca2+ (yellow ball), Na+ (orange ball), and Cl (green ball) are shown in the respective pores. (B, E, and H) Average number and SD (shaded area) of water molecules (over the last 100 ns of respective simulations) lining the pores of WT Orai1 (B), Orai1-L138F (E), and Orai1-A137V (H) are calculated. (C, F, and I) Water orientation lining the pores of WT Orai1 (C), Orai1-L138F (F), and Orai1-A137V (I) is determined as defined by the angle between the dipole vector of the water molecule and the z axis. Water molecules that are locally oriented in a similar manner within the respective pores are indicated by brighter colors.

Supplementary Materials

  • www.sciencesignaling.org/cgi/content/full/10/507/eaao0358/DC1

    Fig. S1. Constitutive activity and cellular localization of cancer database and myopathy Orai1 mutants.

    Fig. S2. Interaction between TM2 and pore helix residues.

    Fig. S3. STIM1 and Orai1-His134 mutants mediate Ca2+-selective currents.

    Fig. S4. Starvation induces nuclear translocation of the transcription factors MITF and TFEB.

    Fig. S5. Pharmacological regulation and fast inactivation of the Orai1-H134A mutant.

    Fig. S6. Pore profile of human and Drosophila Orai.

    Fig. S7. Regulation of the Arg91 gate in closed and constitutively open Orai1.

    Fig. S8. Pulled Ca2+ ion through the Orai1 pore extends the Arg91 gate.

    Fig. S9. The hydrophobic gate forms a barrier for water molecules in Orai1 channels.

    Fig. S10. Cysteine cross-linking of constitutively active Orai1 mutants.

    Fig. S11. Dynamics of water molecules in closed and constitutively active Orai1 channels.

    Table S1. Selectivity of Orai1 mutants.

    Movie S1. Ca2+ permeation through the Orai1 channel.

  • Supplementary Materials for:

    Transmembrane helix connectivity in Orai1 controls two gates for calcium-dependent transcription

    Irene Frischauf, Monika Litviňuková, Romana Schober, Vasilina Zayats, Barbora Svobodová, Daniel Bonhenry, Victoria Lunz, Sabrina Cappello, Laura Tociu, David Reha, Amrutha Stallinger, Anna Hochreiter, Teresa Pammer, Carmen Butorac, Martin Muik, Klaus Groschner, Ivan Bogeski, Rüdiger H. Ettrich, Christoph Romanin, Rainer Schindl*

    *Corresponding author. Email: rainer.schindl{at}medunigraz.at

    This PDF file includes:

    • Fig. S1. Constitutive activity and cellular localization of cancer database and myopathy Orai1 mutants.
    • Fig. S2. Interaction between TM2 and pore helix residues.
    • Fig. S3. STIM1 and Orai1-His134 mutants mediate Ca2+-selective currents.
    • Fig. S4. Starvation induces nuclear translocation of the transcription factors MITF and TFEB.
    • Fig. S5. Pharmacological regulation and fast inactivation of the Orai1-H134A mutant.
    • Fig. S6. Pore profile of human and Drosophila Orai.
    • Fig. S7. Regulation of the Arg91 gate in closed and constitutively open Orai1.
    • Fig. S8. Pulled Ca2+ ion through the Orai1 pore extends the Arg91 gate.
    • Fig. S9. The hydrophobic gate forms a barrier for water molecules in Orai1 channels.
    • Fig. S10. Cysteine cross-linking of constitutively active Orai1 mutants.
    • Fig. S11. Dynamics of water molecules in closed and constitutively active Orai1 channels.
    • Table S1. Selectivity of Orai1 mutants.
    • Legend for movie S1

    [Download PDF]

    Technical Details

    Format: Adobe Acrobat PDF

    Size: 2.98 MB

    Other Supplementary Material for this manuscript includes the following:

    • Movie S1 (.avi format). Ca2+ permeation through the Orai1 channel.

    Citation: I. Frischauf, M. Litviňuková, R. Schober, V. Zayats, B. Svobodová, D. Bonhenry, V. Lunz, S. Cappello, L. Tociu, D. Reha, A. Stallinger, A. Hochreiter, T. Pammer, C. Butorac, M. Muik, K. Groschner, I. Bogeski, R. H. Ettrich, C. Romanin, R. Schindl, Transmembrane helix connectivity in Orai1 controls two gates for calcium-dependent transcription. Sci. Signal. 10, eaao0358 (2017).

    © 2017 American Association for the Advancement of Science

Stay Connected to Science Signaling

Navigate This Article