Fig 1.
CLCA2 enhances ER calcium stores and SOCE.
Calcium imaging studies using HEK293 cells loaded with calcium fluorophore Fluo-4. (A) Cells were treated with 2 micromolar ionomycin in the absence of extracellular calcium to allow ER depletion, then calcium was added to allow SOCE. For CLCA2, n = 52; for control, n = 49. (B) CLCA2-expressing cells were treated with ionomycin, and the SOCE inhibitor BTP-2 (10 micromolar) was added after extracellular calcium (n = 121). Control cells expressed CLCA2 but were untreated with BTP-2 (n = 124). The temporal displacement in the SOCE peaks was due to a minor difference in timing of calcium addition. (C) The same cell lines were treated with SERCA inhibitor CPA (15 micromolar). For CLCA2, n = 246; for control, n = 240. (D) Bar graph showing enhancement by CLCA2 of cytosolic calcium from the ER (Peak 1) and SOCE (Peak 2) in A and C. The height of each peak minus background is plotted. P-values were determined for pairwise comparisons between CLCA2 and control peaks. For ionomycin, Peak 1 p = 0.001, and Peak 2 p = 4.58x10-11. For CPA, Peak 1, p = 2.63 x10-21, and Peak 2 p = 9.23 x10-10. For A and C, traces represent the mean of three independent experiments.
Fig 2.
Effect of CLCA2 knockdown on calcium trafficking.
(A) Calcium imaging in MCF10A mammary epithelial cells with CLCA2 knockdown (shRNA; n = 134) or vector (control; n = 196). Cells were treated with ionomycin as in Fig 1 except that calcium was withdrawn after appearance of Peak 2 as indicated by bars at top. (B) The height of each peak minus background is plotted. Peak 1, p = 1.77 x10-9. Peak 2, p = 6.10 x10-18.
Fig 3.
A second family member, CLCA1, also enhances ER calcium release and SOCE.
(A) Cells stably expressing CLCA1 (n = 228 cells) or vector (n = 219) and loaded with Fluo-4 were treated with ionomycin and cytosolic calcium was measured over time as in Fig 2. (B) Amplitudes of peaks 1 and 2 were plotted. For pairwise comparisons between CLCA1 and vector, the p-value of Peak 1 was 3.2 x10-87 and the p-value of Peak 2 was 5.07 x10-84.
Fig 4.
CLCA2 forms a stable complex with STIM-1 calcium sensor and ORAI-1 calcium channel but not with Best1.
(A) Immunoprecipitations and immunoblots from HEK293 cells transduced with CLCA2-flag or vector. CLCA2 (h2) and Bestrophin1 antibodies (B1) failed to precipitate the other protein. (B) Immunoblot showing equal expression of STIM-1 and ORAI-1 irrespective of ectopic CLCA2 in HEK293. (C) Co-immunoprecipitation of CLCA2 and ORAI-1. The percentage of co-precipitation compared to immunoprecipitation is indicated below each band after subtracting non-specific background. Signal intensities were determined by fluorimetric scans (Licor Odyssey). O1, ORAI-1 antibody; S1, STIM-1 antibody. (D) Co-immunoprecipitation of CLCA2 and STIM-1. Lanes 2 and 3 of the STIM-1 panel were underexposed to avoid saturation. Quantification is based on the initial scan. (E) Immunoprecipitates from MCF10A cells expressing vector or shRNA targeting CLCA2 (KD). Endogenous CLCA2 was detected with Sigma Prestige antibody. TD, transduce. TF, transfect. Co-precipitation of CLCA2 with ORAI-1 was detected four times; with STIM-1, six times.
Fig 5.
Cell surface CLCA2 co-immunoprecipitates and co-localizes with intracellular STIM-1.
(A) Immunoblots of surface-biotinylated cells expressing CLCA2. Biotinylated proteins were detected by neutravidin-Alexa 680. (B) Confocal micrographs of HEK293 cells that stably expressed CLCA2 (green) showed colocalization (white) with STIM-1 (magenta). Top, optical cross-section of xy image (bottom). Scale bar, 10 microns.
Fig 6.
CLCA2 enhances IClCa in TMEM16A-transduced HEK293 cells.
Representative whole-cell chloride currents from cells stably expressing vector (A), CLCA2 (B), TMEM16A plus vector (D) or CLCA2 plus TMEM16A (E) are shown. ICaCC was induced by application of 2 micromolar ionomycin in the presence of extracellular calcium. When the current peaked, a voltage ramp from -60 to +60mV was applied and the current response was recorded. In A and B, current was blocked by DIDS, 100 micromolar. In both D and E, current was fully inhibited by substitution of sodium gluconate (S.G.) for chloride. C, F, relative current amplitudes at +60mV +/- S.E.M. C, n = 9 for CLCA2; n = 7 for vector control; **, p<0.01. F, n = 4 cells for each condition; *, p = 0.02; **, p<0.01. The reversal potential was -8mV in agreement with the Nernst prediction of -7.56mV, based on an excess of 34mM chloride in the extracellular buffer.
Fig 7.
CLCA2 and TMEM16A do not directly interact.
(A) Immunoblots of cells transfected with CLCA2, TMEM16A, or both. Proteins were immunoprecipitated and detected with antibodies specific for either protein. (B) Cells stably expressing vector or CLCA2 were transiently transfected with TMEM16A (TMEM) and treated 48h later with the protein cross-linker DSS, followed by immunoprecipitation and immunoblot. TMEM16A was detected with anti-Flag tag antibody, and CLCA2 was detected using TVE20 antibody. In indicated experiments, cells were treated with 1 micromolar ionomycin 5min before adding cross-linker. 1, position of protein monomer. For CLCA2, this includes the 130 kDa precursor and the 100 kDa N-terminal product. 2, apparent multimers. Right, size marker positions are indicated.
Fig 8.
CLCA2 does not increase TMEM16A stability or surface occupancy.
(A) Immunofluorescence micrographs showing HEK293 cells stably expressing vector (upper row) or CLCA2 (lower row) and probed for transduced TMEM16A or CLCA2. Cells were fixed and stained with anti-Myc tag antibody for TMEM16A (red) or anti-FLAG antibody for CLCA2 (green). No difference in TMEM16A localization was observed. Scale bar, 20 microns. (B) Immunoblots from surface-biotinylated HEK293 cells stably expressing untagged CLCA2 or vector and transfected with TMEM16A-Flag. TMEM16A was immunoprecipitated, and the blot was probed with Alexa 680-tagged streptavidin. A second blot was probed for TMEM16A. Beta tubulin served as a control for protein concentration. To quantify expression, the fluorescence intensity was determined for each band and normalized to vector control using Licor software. To confirm that biotin labeling was confined to the surface, biotinylated proteins were precipitated with avidin-coated beads, blotted, and probed for beta tubulin, which produced only weak bands.