TRPC1 Is Required for Functional Store-operated Ca2+ Channels

The exact role of TRPC1 in store-operated calcium influx channel (SOCC) function is not known. We have examined the effect of overexpression of full-length TRPC1, depletion of endogenous TRPC1, and expression of TRPC1 in which the proposed pore region (S5-S6, amino acids (aa) 557–620) was deleted or modified by site-directed mutagenesis on thapsigargin- and carbachol-stimulated SOCC activity in HSG cells. TRPC1 overexpression induced channel activity that was indistinguishable from the endogenous SOCC activity. Transfection with antisense hTRPC1 decreased SOCC activity although characteristics of SOCC-mediated current, ISOC, were not altered. Expression of TRPC1Δ567–793, but not TRPC1Δ664–793, induced a similar decrease in SOCC activity. Furthermore, TRPC1Δ567–793 was co-immunoprecipitated with endogenous TRPC1. Simultaneous substitutions of seven acidic aa in the S5-S6 region (Asp → Asn and Glu → Gln) decreased SOCC-mediated Ca2+, but not Na+, current and induced a left shift in E rev. Similar effects were induced by E576K or D581K, but not D581N or E615K, substitution. Furthermore, expressed TRPC1 proteins interacted with each other. Together, these data demonstrate that TRPC1 is required for generation of functional SOCC in HSG cells. We suggest that TRPC1 monomers co-assemble to form SOCC and that specific acidic aa residues in the proposed pore region of TRPC1 contribute to Ca2+ influx.

vated primarily in response to depletion of Ca 2ϩ from the internal Ca 2ϩ stores (3,4). However, the mechanism(s) underlying activation and gating of these plasma membrane calcium channels as well as their molecular composition have remained elusive for more than a decade (5)(6)(7). The most well characterized SOCC is the Ca 2ϩ release activated Ca 2ϩ channel (CRAC) that mediates the store-operated current I CRAC in mast cells and lymphocytes (8,9). This channel has been reported to have a very high selectivity for Ca 2ϩ and unique inactivation properties. SOCCs have also been described in a number of other cell types. Many of these display characteristics different from CRAC and from each other (7,8). Based on the distinct activities seen in different cell types, it is likely that the molecular composition, or regulation, of their SOCCs is different.
Members of the TRP superfamily of putative ion channel proteins have been suggested as components of SOCCs (10). Although it is possible that some TRPVs or TRPMs might form SOCCs (11)(12)(13), there are sufficient data to demonstrate that the TRPC family is involved in agonist-stimulated Ca 2ϩ signaling (3,4,14). It is now clear that some TRPCs, e.g. TRPC1 and TRPC4, allow plasma membrane Ca 2ϩ influx in response to depletion of the internal Ca 2ϩ store. Others, like TRPC3 and TRPC6, appear to be activated by PIP 2 hydrolysis in a number of cell types. Compared with the other TRPCs, TRPC1 displays a relatively wider tissue distribution (15,16). Furthermore, it is endogenously present and suggested to be involved in SOCE in several cell types as follows: salivary gland cells (17,18), endothelial cells (19,20), vascular smooth muscle cells (21), and DT 40 cells (22) based on the decrease in SOCE induced by transfection with antisense TRPC1 (17,19), addition of the TRPC1 antibody extracellularly (20,21), or knock-out of the TRPC1 gene (22). Although these previous studies provide convincing evidence that TRPC1 has a critical role in SOCE, they do not elucidate whether TRPC1 serves as an essential regulatory subunit that is required for the assembly or regulation of the SOCCs in these cells or contributes more directly to the channel function per se, i.e. as a component of the pore-forming unit of the channel.
We have previously suggested (17) TRPC1 as a candidate for store-operated Ca 2ϩ entry mechanism in the human submandibular gland cell line, HSG. We have also reported that the TRPC1 C terminus mediates Ca 2ϩ -dependent feedback regulation of SOCE in HSG cells (18,23). To elucidate whether TRPC1 is a regulatory subunit of SOCC or contributes more directly to SOCC-mediated Ca 2ϩ influx, here we have measured SOCE and SOCC activity in control HSG cells and in cells overexpressing either TRPC1 or TRPC1 with mutations in the proposed pore region. The data demonstrate that TRPC1 is required for generation of functional SOCC in HSG cells and that specific acidic aa residues in the proposed pore region of TRPC1 contribute to Ca 2ϩ influx. * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18

EXPERIMENTAL PROCEDURES
HSG Cell Culture, Transfection, and Crude Membrane Preparation-HSG cells were cultured and stably transfected as described before (17,18,23). For transient transfection, cells were detached, diluted to desired the concentration, and allowed to attach to tissue culture plates overnight and then co-transfected with pcDNA3.1 encoding the required TRPC1 (1 g) and pcDNA3.1 encoding GFP (0.5 g) or with the GFP plasmid alone. Cells were cultured on coverslips or tissue culture plates for 24 h, and GFP-positive cells were selected for activity measurements.
Site-directed Mutagenesis-The expression plasmid pcDNA3.1 containing the TRPC1 gene was used as a template for PCR. Oligonucleotides were synthesized as required for the mutagenesis, and PCR was performed using the QuickChange Mutagenesis kit (Stratagene). Following PCR, the parental DNA template was removed by DpnI endonuclease digestion, and the remaining PCR product was transformed into Escherichia coli XL1-Blue cells. Colonies were selected, and each mutation was confirmed by sequencing.
Electrophysiology-Cell attached patch clamp measurements were performed as described earlier (24). Pipette solution contained 100 mM Na/HEPES, 1 mM MgCl 2 , 2 mM CaCl 2 (pH 7.2) (HCl). Standard bath solution containing (mM) 145 NaCl, 5 KCl, 1 MgCl 2 , 1 CaCl 2 , 10 HEPES ((pH 7.4) with NaOH) was replaced by a high KCl solution containing (mM) 145 KCl, 5 NaCl, 1 MgCl 2 , 2 CaCl 2 , 10 HEPES ((pH 7.4) with KOH) after the seal was formed. Liquid junction potentials (8.0, 4.3, and 4.2 mV with Ca 2ϩ , Na ϩ , and Ba 2ϩ , respectively) were calculated using P-Clamp7 software. Currents were recorded using an Axopatch 200A amplifier and digitized with Digidata 1200 (Axon Instruments) at a rate of 10 kHz with filtering at 1 kHz. P-Clamp 7 (Axon instrument) and Origin 6 (Microcal) were used for data analyses. The open probability of the channel (NP o ) was defined as the ratio of open channel area to the total area in an all-point amplitude histogram. The mean unitary amplitudes of the single channel currents were determined from allpoint amplitude histogram fit to a sum of Gaussian distributions. Mean open times were calculated by maximum likelihood estimation using pClamp7 software.
The store-operated inward Ca 2ϩ current (I SOC ) was measured using the whole cell patch clamp technique as described before (23,25) with some modifications. The external solution contained 135 mM sodium glutamate, 1 mM MgCl 2 , 2 mM CaCl 2 , 10 mM glucose, 10 mM HEPES (pH 7.4) (NaOH). The pipette solution contained 135 mM cesium glutamate, 3 mM MgCl 2 , 3 mM ATP (Mg), 10 mM HEPES (pH 7.2) (CsOH), 10 mM EGTA, and 3.5 mM CaCl 2 (final [Ca 2ϩ ] ϭ 100 nM). Cells were held at 0 mV and subjected to voltage ramps (Ϫ100 to ϩ100 mV at 2 mV/ms). Cells were stimulated by perfusion with Tg-or CCh-containing medium. Data are presented as mean Ϯ S.E. or percentage of responding cells. Statistical analyses of the data were performed using Student's t test or 2 test, as appropriate. p Ͻ 0.05 was considered significant.
Immunoprecipitation and Western Blotting-Crude membranes were prepared as described previously (26). Protein concentration was determined by using the Bio-Rad protein assay (microassay procedure). Membrane solubilization and immunoprecipitation were performed as described previously (26,27). Immunocomplexes were pulled down with protein A beads (anti-HA was used at 1:100 dilution for IP). Proteins were released with SDS buffer and visualized by the ECL reaction after Western blotting. Anti-HA (Roche Molecular Biochemicals, 3F10) and anti-TRPC1 (16) were used at 1:1000 dilution.
Confocal Microscopy-HSG cells expressing TRPC1⌬567-793 were cultured on coverslips for 24 h. Cells were fixed, permeabilized, and treated with the HA antibody at 1:100 dilution for 1 h. Cells were then washed and probed with rhodamine-linked secondary antibody as described before (16). Confocal images were collected using an MRC 1024 krypton/argon laser scanning confocal equipped with a Nikon Optiphot II photomicroscope.
[Ca 2ϩ ] i Measurements-Fura2 fluorescence in single cells (cultured overnight on glass bottomed microwell dishes, MatTek Corp.) was measured as described earlier (23) by using an SLM 8000/DMX 100 spectrofluorimeter or Polychrome 1V (TILL Photonics) attached to an inverted Nikon Diaphot microscope with a Fluor 40ϫ oil-immersion objective. Images were acquired using an enhanced CCD camera (CCD-72, DAGE-MTI) and either Image-1 or Metafluor software (Universal Imaging Corporation, PA). In cells that were transiently transfected, fura2 fluorescence was measured in GFP-positive cells. Analog plots of the fluorescence ratio (340/380) are shown, and peak values were used for analysis.

SOCC Activity in HSG Cells
Expressing TRPC1-SOCC activity was measured in control HSG cells and TRPC1-overexpressing HSG cells (TRPC1-cells) by using cell attached patch clamp technique as we have described previously (24). Tg and CCh induced burst-like (Fig. 1, A and C) and flickery (E and G) SOCC activities, respectively, in both sets of cells (membrane potential was held at Ϫ40 mV). This difference in the kinetics is due to modulation of the channel by the [Ca 2ϩ ] in the internal store and in the subplasma membrane region near the channel (24). Both CCh-and Tg-induced SOCC activities displayed unitary current amplitudes of about Ϫ0.85 pA, with occasional overlapping channel openings corresponding to multiples of the first level (i.e. about Ϫ1.7 pA, Fig. 1, B, D, F, and H), although SOCC activity in TRPC1-cells was higher than in control cells. The number of events in the major current level was significantly increased from 1115 Ϯ 145 (n ϭ 9) to 2109 Ϯ 237 (n ϭ 12) and 1278 Ϯ 141 (n ϭ 11) to 2318 Ϯ 265 (n ϭ 11) with Tg and CCh, respectively (12 s recording; n ϭ number of cells, p Ͻ 0.05 in both cases). Average NP o values in TRPC1cells 0.43 Ϯ 0.045 with CCh and 0.33 Ϯ 0.036 with Tg were significantly higher than the values in control cells 0.26 Ϯ 0.019 and 0.12 Ϯ 0.015, respectively (p Ͻ 0.05 in both cases, n is given above). The mean open times of SOCC (calculated from the activity at the major current level) with CCh or Tg were 5.5 Ϯ 0.4 ms or 11.6 Ϯ 2.1 ms, respectively, in TRPC1-cells and 3.8 Ϯ 0.4 ms or 6.5 Ϯ 1.2 ms in control cells.
The I-V relationships obtained with 2 mM Ca 2ϩ ϩ 100 mM Na ϩ (Fig. 1, I and J, values were obtained from the major current peak), or 2 mM Ca 2ϩ ϩ 100 mM N-methyl-D-glucamine (data not shown), in the pipette solution were similar in control and TRPC1-cells stimulated either with CCh or Tg. In all cases, the current displayed inward rectification and reversed at about ϩ20 mV. The slope conductance of SOCC with Ca 2ϩ was 20 pS, which was similar in CCh-or Tg-stimulated TRPC1-cells and control cells. Furthermore, as in control cells, SOCC in TRPC1-cells was also blocked by Gd 3ϩ (Fig. 1, I and J, 0/17 cells tested displayed activity), La 3ϩ , and by 2-aminoethoxydiphenylborate (data not shown). Additionally, the kinetics and unitary current amplitudes with Na ϩ (2 pA) or Ba 2ϩ (0.65 pA) as charge carrying ions (Ca 2ϩ was either removed from the pipette solution or replaced with Ba 2ϩ ) were similar in control HSG and TRPC1-cells (membrane potential was held at Ϫ40 mV). The I-V characteristics of SOCC-mediated Na ϩ and Ba 2ϩ currents as well as the slope conductances (44 pS with Na ϩ and 15 pS with Ba 2ϩ ) were also similar in control and TRPC1-cells (current recordings not shown). Note that the Ca 2ϩ conductance of SOCC determined here is close to that estimated earlier for TRPC1 by noise analysis, 16 pS (28). In aggregate, the data in Fig. 1 demonstrate that TRPC1 expression in HSG cells generates channel activity that is indistinguishable from the endogenous SOCC activity. Additionally, and consistent with our previous studies (17), transfection with antisense TRPC1 (TRPC1-as, but not with antisense TRPC3 or TRPC4) eliminated SOCC activity in 12/14 cells (86%). In 2/14 cells (14%) SOCC activity was similar to that in control cells (current traces not shown, data are summarized in Fig. 5). These data demonstrate that endogenous TRPC1 is required for the Tgand CCh-stimulated SOCC activity and SOCE in HSG cells.
Deletion of the TRPC1 Pore Region Decreased SOCC Activity-TRPC proteins have been suggested to have six transmembrane domains. The aa sequence spanning the region between S5 and S6 includes a seventh hydrophobic domain and has been proposed to form the pore of the channel (3,4), although the exact topology of TRPC1 has not yet been confirmed. To elucidate the role of TRPC1 in SOCC function, we have truncated TRPC1 after S5 (aa 567-793) thus deleting the proposed pore region, S6, and the C terminus. Stable expression of TRPC1⌬567-793 exerted an inhibition of endogenous SOCC activity. 8/9 cells (89%) failed to display any channel activity in response to stimulation by either CCh or Tg at both negative and positive membrane potentials (Fig. 2, A and B, cells were held at either ϩ40 or Ϫ40 mV). 1/9 cells (11%) displayed normal SOCC activity with both Tg and CCh. The frequency of detection of channel activity was significantly reduced (p Ͻ 0.01, 2 test) as compared with that in TRPC1-cells (23/31, 74%) and control cells (21/29, 72%). In agreement with these single channel measurements, Tg-or CCh-stimulated Ca 2ϩ influx (Fig. 2, C-F) in cells expressing TRPC1⌬567-793 was significantly reduced (Ͼ70%) as compared with that in control cells and in cells expressing TRPC1 (p Ͻ 0.05, see figure legend for the average fluorescence values). Consistent with our earlier reports (18,23) and in contrast to the effect of TRPC1⌬567-793 on SOCE, expression of TRPC1⌬664 -793 induced an increase in SOCE (Fig. 2, E and F). Thus, the region aa 567-664 is important for SOCE. Fig. 2G shows that the levels of TRPC1, TRPC1⌬664 -793, and TRPC1⌬567-793 were similar. Thus, the observed differences in SOCE in cells expressing these proteins are not due to differences in the levels of expression. Furthermore, deletion of the aa 567-793 of TRPC1 did not affect its localization to the plasma membrane (Fig. 2H). Importantly, TRPC1⌬567-793 was immunoprecipitated with endogenous TRPC1 (Fig. 2I). Together, these data suggest that expressed TRPC1 proteins associate with the endogenous TRPC1, a component of the endogenous SOCC in these cells.
Effect of Mutating Acidic Amino Acid Residues in the Pore Region of TRPC1 on SOCE-To demonstrate directly the role of the TRPC1 pore region in SOCC function, we made Asp 3 Asn and Glu 3 Gln substitutions of the seven negatively charged residues present in the region between S5 and S6 of TRPC1 (aa 557-620, Fig. 3A) and stably expressed this mutant TRPC1 (Mut-pore) in HSG cells. 13/15 cells (87%) expressing this mutant did not display detectable CCh-or Tg-stimulated SOCC activity (Fig. 3, B and C, respectively). This represents a significant decrease in the frequency of channel detection when compared with that in TRPC1-cells or control HSG cells (p Ͻ 0.01, 2 test). 2/15 cells (13%) displayed a few channel openings at very negative membrane potentials (Ϫ80 mV or lower), and the amplitude of the current detected at Ϫ80 mV appeared to be lower than that seen in control HSG cells or in TRPC1-cells (data not shown). Outward current was also decreased. Due to the low frequency of detection, we have not yet characterized these single channel characteristics in detail. These observations are in contrast to cells transfected with antisense TRPC1, where 14% of the cells displayed normal channel activity. Consistent with the effects of Mut-pore on SOCC activity, fura2 fluorescence measurements demonstrated that the CCh-or Tg-stimulated Ca 2ϩ influx was significantly reduced (by 70%) as compared with those in control HSG cells and TRPC1expressing cells (Fig. 3, D-F). Internal Ca 2ϩ release was not changed. Additionally, infection of Mut-pore cells with adenovirus encoding TRPC1 (AdTRPC1) induced a 30% recovery of Tg-stimulated Ca 2ϩ influx. Infection with AdTRPC3 did not increase Ca 2ϩ influx in Mut-pore cells (data not shown), suggesting that full-length TRPC1 can overcome the effects of Mut-pore. Thus the extent of inhibition achieved by Mut-pore depends on the level of expression relative to that of full-length TRPC1.
We could not determine whether Mut-pore interacted with endogenous TRPC1 because the molecular weights were similar and anti-TRPC1 antibody recognized both proteins. However, to determine whether the expressed Mut-pore can interact with other TRP monomers, HA-tagged Mut-pore was co-expressed with FLAG-tagged TRPC1. The two TRPC1 proteins were co-immunoprecipitated with each other (Fig. 3G).
To analyze further the effect of Mut-pore on CCh-and Tgstimulated SOCC activity, we performed whole cell patch clamp experiments. Cells were held at 0 mV and subjected to voltage ramps every second. The traces shown in Fig. 4, A and D, represent CCh-and Tg-stimulated currents that have been reconstructed from the current amplitudes at Ϫ40 mV obtained from voltage ramps (Ϫ100 to ϩ100 mV at 2 mV/ms). The individual ramps obtained at time "I" (before stimulation of cell) are shown in Fig. 4B and at "II" (30 s after stimulation of cell) are shown in Fig. 4, C and E. At time I, i.e. before addition of CCh or Tg to the bath, a small non-selective basal current was detected, which was not changed by expression of TRPC1 or Mut-pore. Also note that there was no change in the current for up to 10 min of recording, unless cells were stimulated (data not shown). Both CCh and Tg stimulated I SOC , a store-operated current that we have described previously (23,25). Under the experimental conditions used in this study, I SOC displayed inward rectification and reversed at about ϩ10 mV (black trace in Fig. 4, C and E, the basal currents have been subtracted from each trace, also see Fig. 5 for additional data). The inward currents are primarily carried by Ca 2ϩ because replacement of Na ϩ in the medium with N-methyl-D-glucamine did not alter this current (trace not shown). In TRPC1-cells the amplitude of the inward current was significantly increased (11.7 Ϯ 0.9 pA in control versus 17.1 Ϯ 1.4 pA in TRPC1-cells at Ϫ40 mV, red trace in Fig. 4, C and E). However, the characteristics of I SOC were not altered. Importantly, in Mut-pore-expressing cells the amplitude of I SOC was significantly lower than that in control HSG cells and TRPC1-cells (green trace in Fig. 4, C and E, also see details in Fig. 5H). Furthermore, there was a significant left shift in the reversal potential of I SOC (see inset in Fig. 4, C and E, also see Fig. 5G for average values and statistical analysis), suggesting a decrease in Ca 2ϩ selectivity of the channel. Consistent with the single channel measurements, the effects induced by antisense TRPC1 on I SOC were different from those seen with Mut-pore; the amplitude of I SOC was significantly decreased, but the reversal potential was not significantly changed (see Fig. 5, G and H).
We also measured I SOC in the absence of external Ca 2ϩ , in which case the inward current is carried by Na ϩ (Fig. 4F, leak current measured prior to stimulation of cells, trace not shown, has been subtracted). In control HSG cells both CCh (data not shown) and Tg (Fig. 4F) activated an inwardly rectifying current that reversed at 0 mV (black trace, 9.2 Ϯ 1.0 pA at Ϫ40 mV). This current was significantly (p Ͻ 0.05) increased in TRPC1-cells (Fig. 4, F, red trace, 16.5 Ϯ 1.2 pA at Ϫ40 mV). Note that TRPC1 expression induced the same increase (1.5-2.0-fold) in the Ca 2ϩ and Na ϩ currents, which is consistent with the 1.5-2.0-fold increase in SOCE detected by fura2 fluorescence measurements. In contrast to the Ca 2ϩ current in Mut-pore cells, which was 67% lower than in TRPC1-cells and 52% lower than in control cells, the Na ϩ current was 17% less than that in TRPC1-cells (green trace, Fig. 4F) and significantly larger than that in control cells. The inset in Fig. 4F shows Mut-pore-mediated Ca 2ϩ and Na ϩ currents relative to that seen in TRPC1-cells. Thus, mutation of the seven acidic aa residues in the TRPC1 pore region induced a much larger decrease of SOCC-mediated Ca 2ϩ influx than SOCC-mediated Na ϩ influx. These data strongly suggest that these mutations do not result in an overall decrease in channel activity.
Identification of Specific Amino Acid Residues in the TRPC1 Pore Region That Are Involved in SOCC Function- Fig. 5A  shows the alignment of aa residues in the pore region of TRPCs 1, 4, and 5. We made individual substitutions of two residues E576K and E615K, which flank the proposed hydrophobic region (shaded region) and are conserved in TRPC1, -4, and -5. Asp-581, which is present only in TRPC1 and is closest to the pore, was also substituted with Lys or Asn. Fura2 fluorescence and whole cell patch clamp measurements were performed with cells expressing these mutant TRPC1 proteins (note that both stable and transient expression of these TRPC1 mutants yielded similar results). Fig. 5B summarizes the Ca 2ϩ influx data. E576K and D581K induced dominant suppression of SOCE, similar to that seen by Mut-pore expression. Although E615K and D581N (data not shown) induced lower Tg-stimulated Ca 2ϩ influx as compared with TRPC1-cells, it was not lower than that in control HSG cells. Fig. 5C shows that the expression levels of these mutant TRPC1 proteins were similar and do not account for their distinct effects on SOCE. Currentvoltage ramps of I SOC generated by expression of these mutants are shown in Fig. 5, D-F. Averages of the reversal potentials and current densities in these cells are shown in Fig. 5, G and H. Importantly, there was maximum left shift of the E rev (shown by the arrows in D-F) in D581K cells, from ϩ10.8 Ϯ 1.3 and ϩ10.2 Ϯ 0.9 mV in TRPC1-cells and control HSG cells, respectively to Ϫ12.2 Ϯ 2.5 mV. E rev was also left-shifted in E576K cells but not to the same extent as in D581K cells. E615K did not induce any shift in E rev . Consistent with the Ca 2ϩ measurements, the Ca 2ϩ currents generated by Tg in D581K and E576K were significantly lower than that in control cells (see Fig. 5H). D581K induced the largest decrease in I SOC . I SOC in E615K, although lower than in TRPC1 cells, was similar to that in control HSG cells. These data demonstrate that two acidic amino acid residues in the TRPC1 pore region, Glu-576 and Asp-581, are involved in Ca 2ϩ influx via SOCC. However, other amino acids in TRPC1 (which were not tested in this study) might also have a role in Ca 2ϩ influx. We have not presently examined changes in the permeability of monovalent cations such as Cs ϩ (which is present in the intracellular solution) in cells expressing D581K, which might account for the negative reversal potential. DISCUSSION Previous studies (17-23) from our laboratory, and others, provide convincing evidence that TRPC1 has a critical role in SOCE. Based on the presently available data two possible roles can be assigned to TRPC1: (i) it is a regulatory subunit of SOCC that is required for its activation or assembly, and (ii) it is a component of the SOCC pore that directly contributes to SOCC-mediated Ca 2ϩ influx. The data presented above suggest that TRPC1 is a functional component of an SOCC in HSG cells that contributes to Ca 2ϩ influx in response to stimulation by either agonist or Tg. We have shown that the characteristics of SOCC activity generated by expression of TRPC1 in HSG cells are similar to the endogenous SOCC activity in these cells. Consistent with this, expression of TRPC1 in HSG cells induces an increase in the amplitude of the SOCC-mediated current, I SOC , without altering its characteristics (23) (Fig. 4). We have also shown here and previously (17) that endogenous TRPC1 in HSG cells is essential for SOCC activity. Tg-or CCh-stimulated SOCC activity, I SOC , and Ca 2ϩ influx, as well as levels of endogenous TRPC1 in HSG cells are decreased by 80% or more in cells transfected with antisense-TRPC1. More significant is our finding that acidic amino acid residues in the proposed pore region of TRPC1 contribute to SOCE. We have shown that expression of TRPC1 with simultaneous mutations (Asp 3 Asn and Glu 3 Gln) of seven acidic aa residues in the proposed pore region suppressed endogenous SOCC activity in more than 80% of the cells. In the few cells expressing this mutant where channel activity was detected, there was an apparent decrease in the Ca 2ϩ conductance. Importantly, the amplitude of I SOC in these cells was decreased, and there was a left shift in the FIG. 4. Whole cell current recordings in cells expressing Mut-pore. A, stimulation of I SOC by CCh in Mut-pore (green traces), control HSG (black traces), and TRPC1-cells (red traces). The current traces have been reconstructed from the amplitudes measured at Ϫ40 mV in voltage ramps and smoothed for the presentation using Origin6. B and C, voltage ramps of currents recorded at time points I (before stimulation) and II (30 s after stimulation). C, inset, reversal potentials of the currents in the three cells. D and E, I SOC measured in Tg-stimulated cells. E, inset, reversal potentials of the currents. F, Tg-stimulated Na ϩ currents measured when Ca 2ϩ is excluded from the external solution. Inset shows the Na ϩ and Ca 2ϩ currents in Mut-pore cells relative to those in TRPC1-cells. **, the decrease in the Ca 2ϩ current was significantly different (p Ͻ 0.05) from the decrease in the Na ϩ current. reversal potential, suggesting a decrease in the Ca 2ϩ permeability of the channel. Interestingly, the Na ϩ current mediated by Mut-pore was similar to that seen in TRPC1-cells. Thus, mutations in the acidic aa residues in the pore region do not appear to decrease SOCC activity per se. We have also demonstrated that two of these seven acidic amino acid residues, Glu-576 and Asp-581, are involved in SOCC-mediated Ca 2ϩ influx. Substitution of either residue, E576K or D581K, induced a suppression of Tg-stimulated SOCE and I SOC . Notably, Glu-576 and Glu-615 are conserved in TRPC1, TRPC4, and TRPC5, although Asp-581 is present only in TRPC1. Our data also rule out the possibility that a decrease in SOCC activity in cells expressing mutant TRPC1 proteins is due to differences in protein expression or altered localization. In aggregate, these data support the proposal that TRPC1 is a component of the functional (i.e. pore-forming) unit of SOCC, rather than an associated regulatory protein. Although it is possible that a single mutation in an extracellular domain of TRPC1 might alter its interaction with an as yet unknown SOCC protein and lead to a decrease in Ca 2ϩ influx via the channel, it is highly unlikely.
Previous studies have shown that Asp 3 Asn substitution in the pore region of TRPV4 decreased Ca 2ϩ influx, although Asp 3 Lys substitutions completely blocked channel function (29). Substitution of conserved hydrophobic residues (LFW) in the TRPC6 pore region eliminated TRPC6-generated channel activity in HEK293 cells (30). However, neither of these channels appear to be involved in SOCE. While this manuscript was in preparation, it was reported (31) that expression of TRPV6 with substitution of FEL to AAA in the proposed pore region induced suppression of endogenous I CRAC in Jurkat cells. However, it is not clear whether the endogenous TRPV6 in Jurkat cells is required for I CRAC . Furthermore, TRPV6 displayed both store-operated and store-independent activities. The data we have presented above are significant because they demonstrate the following: (i) endogenous TRPC1 in HSG cells is required for the endogenous SOCE detected in these cells; (ii) expressed TRPC1 is regulated by store depletion and is not spontaneously active; and (iii) acidic aa residues in the pore region of TRPC1 contribute to SOCE.
Although presently we do not have evidence for homo-multimerization of endogenous TRPC1 in HSG cells, we have shown that the expressed mutant TRPC1 is immunoprecipitated with endogenous TRPC1. This result has major implications because it suggests that TRPC1 monomers interact with each other to form SOCC. Our data are consistent with reports suggesting that Drosophila TRP and TRPC1␤ homomultimerize via N-terminal interactions (32,33). We suggest that endogenous or exogenously expressed TRPC1 monomers associate with each other to form functional SOCCs. Thus, when full-length TRPC1 is expressed channel activity is increased. When mutant TRPC1s are overexpressed relative to the endogenous protein, the probability of mutant TRPC1 monomers associating with each other or with endogenous TRPC1 monomers is relatively high. As a result, aberrant SOCCs are formed which have decreased permeability for calcium. In contrast, in cells transfected with antisense TRPC1 or expressing TRPC1⌬567-793, where either depletion of TRPC1 or the pore region is deleted, respectively, there is a reduction in the total number of functional channels rather than a change in the channel properties. However, other proteins, including other TRPCs, might be co-assembled with endogenous TRPC1 and required for functional SOCC in HSG cells. Studies using heterologous expression have shown that TRPC1 interacts with TRPC3 (33, 34), TRPC6 (30), and TRPC4 and TRPC5 (35). The latter study also showed that endogenous heteromers of either TRPC1 and TRPC4 or TRPC1 and TRPC5 were co-immunoprecipitated from rat brain. TRPC1 expression altered the currents generated when either TRPC5 or TRPC4 was expressed alone, and furthermore, these currents were not activated by store depletion. Although it is unclear why TRPC1 forms different types of channels in different cells, it is important to note that the characteristics of the channel formed by heteromeric TRPC1 channels (35) appear to be considerably different from those of I SOC measured in HSG cells. Further studies will be required to determine which TRPC proteins are endogenously expressed in HSG cells and interact with endogenous TRPC1.
In conclusion, the data presented here suggest that TRPC1 is a component of the functional (pore-forming) unit of SOCC in HSG cells. However, these data do not rule out the possibility that SOCC might be a heteromer of TRPC1 with other TRPCs (5,30,(33)(34)(35) or with other as yet unknown protein(s). We have shown earlier that in HSG cells TRPC1, like the Drosophila TRP (14,36), is assembled in a supramolecular protein complex with key proteins involved in the Ca 2ϩ signaling cascade that leads to SOCC activation (26). We suggest that SOCC activity in any cell type will depend not only on the proteins that constitute its pore-forming unit but also other regulatory proteins that might affect its function, assembly, or localization. It will be important to determine whether differences in the molecular composition of the channel per se, or its regulation, account for the large variation in the characteristics of SOCCs seen in different cell types.