Fig 1.
The Orai C-termini of adjacent monomers interact in an anti-parallel coiled-coil.
(A) X-ray crystal structure of Drosophila Orai channel (PDB code ID 4HKR) viewed from the extracellular side. The channel displays six-fold symmetry within the pore formed by TM1 helices (blue) and the concentric layer formed by the TM2 (red) and TM3 (green) helices. At the outermost helical layer formed by the TM4 helices (orange and yellow), only a three-fold symmetry is observed due to the formation of an anti-parallel coiled-coil (box) between the cytosolic extensions of adjacent TM4 monomers. (B) Close-up of the hydrophobic pocket formed by putative interaction between L273 of human Orai1 (hot pink) in one monomer with L276 (teal) of the adjacent monomer. PyMol mutagenesis wizard was used to substitute a leucine residue at position 316 which is an isoleucine in the Drosophila structure. (C) Schematic of the experimental set-up. Exogenous cysteines were introduced at L273 or L276 individually. Diagram shows L273C mutations (created with PyMol mutagenesis wizard) on the adjacent TM4 monomers. If the introduced cysteines are close together, then addition of an oxidizing agent, such as diamide, could induce the formation of a disulfide bond.
Fig 2.
Cross-linking the Orai1 C-termini prevents STIM1 association.
(A) Confocal images showing expression of the Orai-YFP single and double cysteine mutants at residues L273 and L276. Whereas the single L273C and L276C mutants are expressed in the plasma membrane, the L273/276C double mutant was only found in unknown intracellular compartments. Cells were transfected with the indicated Orai1 constructs together with CFP-CAD. (B) Summary of Orai-YFP fluorescence at the plasma membrane analyzed from widefield YFP images acquired from cells transfected with CFP-CAD and the indicated Orai1-YFP constructs. Data points are mean ± SEM of 23–26 cells. (C) Representative traces showing E-FRET changes following store depletion by 4 μM ionomycin in single HEK293 cells co-expressing STIM1-CFP and the indicated Orai1-YFP proteins. (D) Summary of E-FRET values of the Orai1 channels before and after store depletion by ionomycin (IONO). For each individual cell, the E-FRET value was averaged from three frames. Data points are mean ± SEM of 21–48 cells. The post-ionomycin E-FRET values of the L273C and L276C channels are not significantly different than that of WT Orai1 channels. (E) Western blot of cell lysates expressing WT Orai1-YFP, L273C Orai1-YFP, or L276C Orai1-YFP channels exposed to varying concentrations of diamide (0–500 μM). Boxes drawn in the L276C Orai1-YFP 500 μM treatment lane indicate the regions used for the quantitation of monomer (magenta) and dimer (red) values. (F) Western blot quantitation of dimer formation at various diamide concentrations (0–500 μM). The percent of total Orai1 cross-linked was calculated as dimer/(monomer + dimer). Data points are mean ± SEM for 4–7 cells (*** p ≤ 0.001; ** p ≤ 0.01). (G) Steady-state E-FRET values of cells co-expressing Orai1-YFP channels with STIM1-CFP and pre-treated with various concentrations of diamide (0–500 μM) prior to store-depletion. Data points are mean ± SEM for 12–51 cells (*** p ≤ 0.001).
Fig 3.
State dependence of cross-linking: prior CAD binding impairs Orai C-terminal self-association.
(A) Confocal images showing CFP-CAD co-localization with WT and mutant Orai1-YFP channels. (B) Western blot of cell lysates co-expressing YFP-CAD with WT Orai1-YFP, L273C Orai1-YFP, or L276C Orai1-YFP channels exposed to varying concentrations of diamide (0–500 μM). The WT-Orai1 and L276C-Orai1 lanes show blot that was exposed for 5 minutes, while the L273C-Orai1 blot, which showed higher protein expression, was exposed for only 30 seconds. The complete blot for each exposure is shown in S1 Fig. (C) Western blot quantitation of dimer formation at various diamide concentrations (0–500 μM) for L273C (left) and L276C (right) in the presence and absence of YFP-CAD. The percent of total Orai1 cross-linked was calculated as dimer/(monomer + dimer). Data points are mean ± SEM for 3–7 cells. Statistics in these graphs represent significance between diamide treatments in the presence and absence of YFP-CAD (** p ≤ 0.01; * p ≤ 0.05) (D and E) Comparison of positioning of the L273 (pink) and L276 (cyan) side chains in the NMR complex structure (Orai helices are in blue and green, the STIM1 fragments are represented in light pink) (PDB ID 2MAK) and the positions of the Drosophila equivalents I316 (hot pink) and L319 (cyan) in the Drosophila crystal structure (helices shown in yellow and orange) (PDB ID 4HKR). PyMol mutagenesis wizard was used to substitute a leucine residue at position 316 which is an isoleucine in the Drosophila structure.
Fig 4.
Induction of disulfide bonds in STIM1 bound channels diminishes STIM1 association and ICRAC.
(A) E-FRET traces of cells co-expressing the indicated Orai1-YFP constructs with CFP-CAD (left) or STIM1-CFP (right). Cells were treated with 500 μM diamide followed by 5 mM BMS. Data points are mean ± SEM for 4–8 cells. (B) Confocal images showing CFP-CAD localization of cells co-expressing CFP-CAD with the indicated Orai1-YFP channels treated with 500 μM diamide then 5 mM BMS. (C) Time-course of Orai1 currents from cells co-expressing STIM1 with the indicated Orai1-YFP channels. After development of a stable current, 200 μM diamide was applied to induce Orai1 cross-linking. This was followed by application of 5 mM BMS as indicated. The diamide induced inhibition of Orai1 current is largely restored following administration of BMS. Note that because BMS itself slightly blocks ICRAC [40], the full extent to relief by BMS requires its washout. Arrowheads denote the time points where current amplitudes were analyzed for block (D) Summary of block produced by 500 μM diamide. Data points are mean ± SEM for 4 cells (** p ≤ 0.01).
Fig 5.
A conserved motif in the bend region of the Orai1 C-terminus is required for STIM1 binding.
(A) Sequence alignment of the Orai1 C-terminus from various species shows a conserved stretch of amino acids spanning residues 260–268 (human Orai1) that include the SHK residues. (B) Close-up of Drosophila Orai1 C-terminus with the conserved ‘SHK’ bend highlighted in green. Because the side chain for K265 was not modeled into the Drosophila structure, we used the PyMol mutagenesis wizard to generate three different potential rotomer conformations of this residue. (C) Summary of E-FRET for mutants of S263 and K265 Orai1-YFP with CFP-CAD. E-FRET values for each cell were averaged from three frames. Data points are mean ± SEM for 15–42 cells (*** p ≤ 0.001; * p ≤ 0.05). (D) Summary of E-FRET for mutation of S263 in Orai1-YFP to the equivalent residues in Orai2 (arginine, R) and Orai3 (alanine, A) with CFP-CAD. Data points are mean ± SEM for 16–19 cells. (E) Close-up of TM4a ‘SHK’ hairpin bend in one monomer. Highlighted in green are residues S263 and Q271, which were mutated to cysteine. Formation of an intra-monomer disulfide bond between these residues is predicted to restrict conformational changes at the bend. (F) Summary of E-FRET between the S263/Q271C Orai-YFP bend double mutant and CFP-CAD, in the presence of diamide or BMS. E-FRET values for each cell were averaged from three frames. Data points are mean ± SEM for 15–32 cells (*** p ≤ 0.001; ** p ≤ 0.01).