STIM2 protein mediates distinct store-dependent and store-independent modes of CRAC channel activation

Abstract

STIM1 and CRACM1 (or Orai1) are essential molecular components mediating store-operated Ca2+ entry (SOCE) and Ca2+ release-activated Ca2+ (CRAC) currents. Although STIM1 acts as a luminal Ca2+ sensor in the endoplasmic reticulum (ER), the function of STIM2 remains unclear. Here we reveal that STIM2 has two distinct modes of activating CRAC channels: a store-operated mode that is activated through depletion of ER Ca2+ stores by inositol 1,4,5-trisphosphate (InsP3) and store-independent activation that is mediated by cell dialysis during whole-cell perfusion. Both modes are regulated by calmodulin (CaM). The store-operated mode is transient in intact cells, possibly reflecting recruitment of CaM, whereas loss of CaM in perfused cells accounts for the persistence of the store-independent mode. The inhibition by CaM can be reversed by 2-aminoethoxydiphenyl borate (2-APB), resulting in rapid, store-independent activation of CRAC channels. The aminoglycoside antibiotic G418 is a highly specific and potent inhibitor of STIM2-dependent CRAC channel activation. The results reveal a novel bimodal control of CRAC channels by STIM2, the store dependence and CaM regulation, which indicates that the STIM2/CRACM1 complex may be under the control of both luminal and cytoplasmic Ca2+ levels.

Figure 1

Store-independent activation of CRACM1 by STIM2 through 2-APB. In all experiments (except black trace in A and green traces in E and F), data are from HEK293 cells grown in the presence of 500 μg/ml G418. A) Average CRAC current densities at −80 mV in wild-type HEK293 cells (black, n=14) and in cells stably overexpressing STIM1 (blue, n=14) or STIM2 (red, n=8) in response to 20 μM InsP₃. [Ca²⁺]_i was clamped to near zero with 20 mM BAPTA. B) Average CRAC current densities induced by 20 μM InsP₃ + 20 mM BAPTA in STIM1 + CRACM1 cells (black, n=15), STIM2 + CRACM1 cells (blue, n=10), and STIM2 + CRACM1 cells stimulated by 2 μM 2-APB (red, n=9). C) Representative current-voltage (I/V) relationships of CRAC currents in STIM1 + CRACM1-expressing cells or 2 μM 2-APB-induced currents in STIM2 + CRACM1-expressing cells shown in B. D) Average CRAC current densities in HEK293 cells expressing STIM2 alone (purple, n=5; trace is flat around 0 pA/pF and masked by blue trace), CRACM1 alone (black, n=3, trace is flat around 0 pA/pF and masked by blue trace), STIM1 + CRACM1 (blue, n=8), and STIM2 + CRACM1 (red, n=9). In all cells, [Ca²⁺]_i was buffered to 150 nM using 20 mM BAPTA and 8 mM CaCl₂. E) High-resolution average CRAC currents at −80 mV in STIM2 + CRACM1 cells grown in the presence (red, n=7) or absence (green, n=5) of G418 induced by 50 μM 2-APB as in D. F) Comparison of activation kinetics of InsP₃- and 2-APB-induced gating of CRAC channels. Data sets are taken from B and E and plotted on same time scale to illustrate speed of STIM2-dependent and 2-APB-induced activation relative to STIM1-dependent and store-operated activation of CRACM1.

Figure 2

Immunolocalization of STIM and CRACM proteins. Images were obtained by confocal microscopy, while colocalization (yellow) was demonstrated by merging the images using ImageJ software (NIH). A) Cells stably expressing CFP-CRACM1 (green) were transfected with STIM1 (red), treated with 2 μM thapsigargin for 10 min, and then fixed, permeabilized, and blocked. Cells were sequentially double-stained with sheep STIM1-specific and rabbit anti-GFP antibodies (Invitrogen), followed by corresponding 2° antibodies. B) Cells stably expressing STIM2 (red) were transfected with CFP-CRACM1 (green), treated with 2 μM thapsigargin for 10 min, and then fixed, permeabilized, and blocked. Cells were sequentially double-stained with sheep STIM2-specific and rabbit anti-GFP antibodies (Invitrogen), followed by corresponding 2° antibodies. C) Binding of STIM proteins to CRACM1 was assessed after transient transfection of STIM1 or STIM2 into HEK293 cells stably expressing CRACM1. Approximate molecular masses of proteins were 85 kDa for STIM1, 105 kDa for STIM2, and 55 kDa for CFP-CRACM1, the glycosylated form of which is ~70 kDa. After transfection, cells were lysed and the proteins were quantified. For immunoprecipitations, 200 μg of protein/sample were incubated with protein G beads coated with rabbit anti-GFP antibodies (Invitrogen) for 2 h followed by separation with SDS-PAGE and transfer to PVDF membranes. Protein expression was confirmed by loading 5 μg of untransfected lysate. STIM proteins were detected by Western blot using an antibody cross-reacting with both STIM1 and STIM2 (BD Biosciences, San Jose, CA, USA), while pulldown and expression of CFP-CRACM1 was demonstrated by stripping and reprobing a goat-anti-GFP antibody (Abcam). D) STIM2-expressing cells were incubated for 3 wk in the presence or absence of G418. Cells were lysed and the proteins were quantified. Proteins (5 μg/sample) were separated by SDS-PAGE and transferred to PVDF membranes. Levels of STIM expression were demonstrated by Western blot with an antibody cross-reacting to both STIM1 and STIM2 (BD Biosciences).

Figure 3

Store-operated activation of CRACM1 through STIM2. A–F) STIM2-expressing HEK293 cells were grown without G418. A) Average CRAC current densities induced by 20 μM InsP₃ in STIM2-expressing HEK293 cells transfected with CRACM1 (black, n=34), CRACM2 (blue, n=5), and CRACM3 (red, n=5) with [Ca²⁺]_i clamped to near zero with 20 mM BAPTA. B) Average current-voltage (I/V) relationships of CRAC currents extracted from representative STIM2 + CRACM1, 2, and 3 cells shown in A at 300 s into experiment. Data represent leak-subtracted currents evoked by 50 ms voltage ramps from −100 to + 100 mV, normalized to cell capacitance (pA/pF). C) Average CRAC current densities in STIM2-expressing HEK293 cells transfected with CRACM1, where [Ca²⁺]_i was clamped to near zero with 20 mM BAPTA to induce passive store depletion (blue, n=7). In other cells (black, n=6), active store depletion was induced by brief (2 s) application of 2 μM ionomycin (indicated by the arrow). D) Average CRAC current densities in STIM1- (green, n=6) or STIM2-expressing HEK293 cells (red, n=7) transfected with CRACM1, where store depletion was prevented by omission of InsP₃ and buffering [Ca2+]_i to 150 nM with 20 mM BAPTA + 8 mM CaCl₂. E) Average CRAC current densities at − 80 mV in wt HEK293 cells (black, n=14) or cells treated with STIM1-specific siRNA (red, n=6) in response to 20 μM InsP₃. [Ca²⁺]_i was clamped to near zero with 20 mM BAPTA. Data demonstrate that STIM1 siRNA treatment was effective and suppressed native CRAC currents. F) Average CRAC current densities in STIM2 + CRACM1 cells that were cotransfected with siRNA against STIM1 (black, n=5) in parallel and exactly as in E. Average CRAC current densities in CRACM1-expressing cells that were cotransfected with the D80A EF-hand mutant of STIM2 (red, n=5). In both experimental sets, store depletion was induced by 20 μM InsP₃ and [Ca²⁺]_i buffered to near zero by 20 mM BAPTA.

Figure 4

Effects of G418 on STIM1- and STIM2-mediated activation of CRACM1. A) Average CRAC current densities in STIM2 + CRACM1 cells in the absence of G418 (closed circles, n=34) or presence of 300 nM (open circles, n=9), 1 μM (closed squares, n=11), 3 μM (open squares, n=8), or 10 μM (closed triangles, n=8) of G418 in the pipette. B) Dose-response relationship of average CRAC current densities as a function of G418 concentration extracted at 300 s from the recordings shown in E. Data were fitted with a dose-response curve, yielding an IC₅₀ value of 640 nM and a Hill coefficient of 2.6. C) Average CRAC currents in STIM1 + CRACM1 cells (n=7) grown in the presence of G418. Pipette solutions contained InsP₃ (20 μM) and 10 μM G418, demonstrating that these cells are largely insensitive to intracellular G418. Extracellular application of 10 μM G418 caused a small and reversible reduction in CRAC current. D) Effects of extracellularly applied G418 on STIM2 + CRACM1. Average CRAC currents in STIM2 + CRACM1 cells grown without G418. Pipette solutions contained InsP₃ (20 μM) to activate biphasic CRAC currents. Bar indicates extracellular application of various concentrations of G418. These caused a small and reversible reduction in CRAC current: 300 nM (filled circles, n=3), 1 μM (open circles, n=4), and 10 μM (filled squares, n=5).

Figure 5

STIM2 causes transient store-operated activation of CRACM1. A–F) STIM2-expressing HEK293 cells were grown without G418. A) Average CRAC current densities in STIM2 + CRACM1 cells, where store depletion was induced by 20 μM InsP₃ + 20 mM BAPTA. Traces represent binned data of experiments in which pipette series resistances were within the following ranges: 2– 4 MΩ (black, n=17), 5–7 MΩ (blue, n=7), or 7–9 MΩ (green, n=7). B) Average CRAC current densities in STIM2 + CRACM1 cells. CRAC currents developed normally both in the absence of added CaM with [Ca²⁺]_i buffered to 100 nM with 10 mM EGTA + 3.6 mM CaCl₂ (black, n=5) or in its presence (100 μM) with [Ca²⁺]_i buffered to 0 with 10 mM EGTA and no added Ca²⁺ (green, n=9). However, CRAC currents were suppressed by the combined presence of 100 μM CaM and 100 nM [Ca²⁺]_i (blue, n=7; red, n=4). Application of 2 μM ionomycin for 3 s (red) or 5 μM 2-APB for 60 s (blue) is indicated in graph. C) Average CRAC currents in STIM2 + CRACM1 cells induced by 20 μM InsP₃ with [Ca²⁺]_i buffered to 100 nM with 10 mM EGTA + 3.6 mM CaCl₂ in the absence (black, n=7) or additional presence of 100 μM CaM (blue, n=7) in pipette. Application of 5 μM 2-APB for 60 s to CaM-treated cells (blue) is indicated in graph. D) Average CRAC currents in STIM1 + CRACM1 cells induced by 20 μM InsP₃ with [Ca²⁺]_i buffered to 100 nM using 10 mM EGTA + 3.6 mM CaCl₂. CRAC currents developed in the absence of added CaM (black, n=9) but were rapidly suppressed by 100 μM CaM in pipette (red, n=6). Application of 2 μM ionomycin for 3 s (red) reversed the CaM-mediated inhibition of CRAC currents. E) Average CRAC current densities in STIM2-expressing HEK293 cells transfected with CRACM1, where [Ca²⁺]_i was clamped to near zero with 20 mM BAPTA. 2-APB (50 μM) was applied for 30 s as indicated in graph (n=7). F) Average CRAC current densities at − 80 mV in STIM2 + CRACM1 cells in response to 20 μM InsP₃ + 20 mM BAPTA. Pipette solution additionally contained 50 μM 2-APB. At the indicated time, 50 μM 2-APB was applied extracellularly.

Figure 6

STIM2 causes transient SOCE in intact cells. A) Changes in [Ca²⁺]_i measured as ratios of fura-2 fluorescence excited at 340 and 380 nm in STIM2 cells transfected with empty vector in the absence (black) or presence (blue) of 1 mM extracellular Ca²⁺. Carbachol (100 μM) and thapsigargin (2 μM) were applied as indicated by arrow and maintained for duration of experiment. B) Identical experimental conditions as in A, but for STIM1 + CRACM1 cells. C) Identical experimental conditions as in A, but for STIM2 + CRACM1 cells. D) These traces represent subtracted traces from A–C, where [Ca²⁺]_i signal in Ca²⁺-free solution was subtracted from that obtained with Ca²⁺ to yield net Ca²⁺ entry.