Expression of Trp3 Determines Sensitivity of Capacitative Ca2+ Entry to Nitric Oxide and Mitochondrial Ca2+ Handling

The role of Trp3 in cellular regulation of Ca2+ entry by NO was studied in human embryonic kidney (HEK) 293 cells. In vector-transfected HEK293 cells (controls), thapsigargin (TG)-induced (capacitative Ca2+ entry (CCE)-mediated) intracellular Ca2+signals and Mn2+ entry were markedly suppressed by the NO donor 2-(N,N-diethylamino)diazenolate-2-oxide sodium salt (3 μm) or by authentic NO (100 μm). In cells overexpressing Trp3 (T3-9), TG-induced intracellular Ca2+ signals exhibited an amplitude similar to that of controls but lacked sensitivity to inhibition by NO. Consistently, NO inhibited TG-induced Mn2+ entry in controls but not in T3-9 cells. Moreover, CCE-mediated Mn2+ entry into T3-9 cells exhibited a striking sensitivity to inhibition by extracellular Ca2+, which was not detectable in controls. Suppression of mitochondrial Ca2+ handling with the uncouplers carbonyl cyanide m-chlorophenyl hydrazone (300 nm) or antimycin A1 (-AA1) mimicked the inhibitory effect of NO on CCE in controls but barely affected CCE in T3-9 cells. T3-9 cells exhibited enhanced carbachol-stimulated Ca2+entry and clearly detectable cation currents through Trp3 cation channels. NO as well as carbonyl cyanide m-chlorophenyl hydrazone slightly promoted carbachol-induced Ca2+ entry into T3-9 cells. Simultaneous measurement of cytoplasmic Ca2+ and membrane currents revealed that Trp3 cation currents are inhibited during Ca2+ entry-induced elevation of cytoplasmic Ca2+, and that this negative feedback regulation is blunted by NO. Our results demonstrate that overexpression of Trp3 generates phospholipase C-regulated cation channels, which exhibit regulatory properties different from those of endogenous CCE channels. Moreover, we show for the first time that Trp3 expression determines biophysical properties as well as regulation of CCE channels by NO and mitochondrial Ca2+ handling. Thus, we propose Trp3 as a subunit of CCE channels.


The role of Trp3 in cellular regulation of Ca 2؉ entry by NO was studied in human embryonic kidney (HEK) 293 cells. In vector-transfected HEK293 cells (controls), thapsigargin (TG)-induced (capacitative Ca 2؉ entry (CCE)-mediated) intracellular Ca 2؉ signals and Mn 2؉ entry were markedly suppressed by the NO donor 2-(N,N-diethylamino)diazenolate-2-oxide sodium salt (3 M) or by authentic NO (100 M). In cells overexpressing
Trp3 (T3-9), TG-induced intracellular Ca 2؉ signals exhibited an amplitude similar to that of controls but lacked sensitivity to inhibition by NO. Consistently, NO inhibited TG-induced Mn 2؉ entry in controls but not in T3-9 cells. Moreover, CCE-mediated Mn 2؉ entry into T3-9 cells exhibited a striking sensitivity to inhibition by extracellular Ca 2؉ , which was not detectable in controls. Suppression of mitochondrial Ca 2؉ handling with the uncouplers carbonyl cyanide m-chlorophenyl hydrazone (300 nM) or antimycin A 1 (-AA 1 ) mimicked the inhibitory effect of NO on CCE in controls but barely affected CCE in T3-9 cells. T3-9 cells exhibited enhanced carbachol-stimulated Ca 2؉ entry and clearly detectable cation currents through Trp3 cation channels. NO as well as carbonyl cyanide m-chlorophenyl hydrazone slightly promoted carbachol-induced Ca 2؉ entry into T3-9 cells. Simultaneous measurement of cytoplasmic Ca 2؉ and membrane currents revealed that Trp3 cation currents are inhibited during Ca 2؉ entry-induced elevation of cytoplasmic Ca 2؉ , and that this negative feedback regulation is blunted by NO. Our results demonstrate that overexpression of Trp3 generates phospholipase C-regulated cation channels, which exhibit regulatory properties different from those of endogenous CCE channels. Moreover, we show for the first time that Trp3 expression determines biophysical properties as well as regulation of CCE channels by NO and mitochondrial Ca 2؉ handling. Thus, we propose Trp3 as a subunit of CCE channels.
Depletion of intracellular Ca 2ϩ stores by IP 3 -dependent 1 or IP 3 -independent pathways stimulates a Ca 2ϩ influx phenomenon that is termed Ca 2ϩ release-activated Ca 2ϩ entry, storeoperated Ca 2ϩ entry, or capacitative Ca 2ϩ entry (CCE) (1). The molecular nature of CCE channels is still not clearly resolved. Nonetheless, increasing evidence suggests a role of members of the Trp protein family in the formation of CCE channel complexes (2,3). Various cellular mechanisms have been proposed for activation of CCE channels (4), including a conformational coupling between IP 3 receptors and CCE channels (5). A direct gating by interaction with a specific cytosolic domain of the IP 3 receptor has been demonstrated for cation channels derived by overexpression of Trp3 protein, and activation of this Trp channel was found to depend on IP 3 (6,7). However, recent studies with cells lacking expression of all three types of IP 3 receptors question a general role of IP 3 in activation of CCE channels but confirm requirement of basal PLC activity (8,9). Therefore, Trp species that are sensitive to PLC-derived signals such as IP 3 and diacylglycerol appear as attractive candidates for proteins forming CCE channel complexes. Although heterologous overexpression of Trp3 or Trp6 clearly generates phospholipase C/IP 3 -dependent Ca 2ϩ entry pathways (6,7,10) and sensitivity to activation by diacylglycerols (11,12), these proteins fail to promote the classical CCE phenomenon upon overexpression (10,12). Thus, the cation channels derived by overexpression of Trp3 or Trp6 resemble native CCE channels in terms of their principle PLC dependence, but apparently lack the distinctive ability of CCE channels to sense the filling state of the Ca 2ϩ stores. Nonetheless, Trp3 or Trp6 proteins may function as subunits of multimeric CCE channel complexes. This hypothesis has been put forward by the observation that Trp3 overexpression not only increases phospholipase C/IP 3 -dependent Ca 2ϩ entry but in addition changes its pharmacological properties, leading to loss of sensitivity to block by Gd 3ϩ (10). These results were interpreted as the ability of Trp3 proteins to serve as functional subunits of CCE channel complexes (10,13).
The present work was designed to test whether Trp3 interferes with the regulatory properties of CCE channels such as sensing local intracellular Ca 2ϩ concentrations that are controlled by mitochondria (14). Another important mechanism of CCE regulation is its inhibition by nitric oxide (NO), which has been proposed to promote refilling of the endoplasmic reticulum resulting in reduced CCE (15). NO donors such as DEANO have been demonstrated to inhibit CCE into platelets (16) and smooth muscle cells (15,17). Interestingly, the lipophilic NO donor GEA3162 has been demonstrated to enhance CCE into mouse parotid acini (18) and to activate Ca 2ϩ entry into human embryonic kidney (HEK) 293 cells stably expressing Trp3 (19).
Thus, it appears reasonable to speculate that variable NO sensitivity might exist among CCE channel complexes, depending on the subunit stoichiometry and contribution of specific Trp species. With this study, we demonstrate for the first time that the expression level of Trp3 affects two distinctly different Ca 2ϩ entry pathways. Trp3 overexpression in HEK293 cells promotes a phospholipase C-dependent Ca 2ϩ entry pathway and eliminates the sensitivity of classical CCE channels to inhibition by NO as well as mitochondrial Ca 2ϩ handling.

MATERIALS AND METHODS
Cell Culture-HEK293 cells were stably transfected cells with either the coding region of hTrp3 (GenBank accession no. U47050) (10), designated as T3-9 cells, or with the neomycin resistance cassette, designated as HEK VT (controls). Cells were cultured in DMEM supplemented with 10% fetal calf serum and 0.178 g/liter Geneticin. In case of control cells, different clones were pooled together and cultured as specified above.
SDS-PAGE and Immunoblotting-T3-9 cells were grown to confluence. Carbonate-extracted cell lysates were separated on discontinuous sucrose gradients, and fractions were collected after ultracentrifugation. 25% of the fraction was used for Bradford assay, and the remaining 75% of the fraction was used for Western blot analysis. The fraction was separated on SDS-PAGE (8 or 12% SDS gels; 50 min, 180 mV) and afterward transferred to nitrocellulose membranes (90 min, 200 mA) using the Bio-Rad Protean III and the Mini Transblot. The blots were incubated in TBST buffer (pH 7.4) containing 1% Bio-Rad Blocking reagent overnight at 4°C. Antibodies were diluted in 1% blocking solution (anti-hemagglutinin, 1:1000), and blots were incubated for 2 h with the first antibody, washed with TBST buffer, followed by 60-min incubation with the second antibody (horseradish-conjugated anti-rabbit from Sigma). After washing the blot with TBST buffer, immunoreactivity was detected using the ECL detection system (Amersham Pharmacia Biotech).
Measurement of Cytoplasmic Ca 2ϩ -Intracellular Ca 2ϩ was measured with fura-2. fura-2/AM was initially dissolved in Me 2 SO at 2 mM and used at a final concentration of 2 M. Confluent cells were harvested by enzymatic digestion (0.25% trypsin), suspended in 5 ml of culture medium (Dulbecco's modified Eagle's medium) without serum, and loaded with fura-2/AM for 60 min at 37°C and 5% CO 2 . Thereafter, cells were washed once with Ca 2ϩ containing Tris buffer (50 mM Tris, 2.5 mM CaCl 2 , 1 mM MgCl 2 , pH 7.4), incubated for 20 min, and washed in Tris buffer without Ca 2ϩ (50 mM Tris, 1 mM MgCl 2 , 100 M EGTA, pH 7.4). Fluorescent measurements were carried out with a dual wavelength spectrofluorimeter (Hitachi F2000). Cells were maintained at 37°C, and emission was collected at 510 nm at excitation of 340 and 380 nm, respectively. For store depletion 100 nM thapsigargin (TG) or 200 M carbachol (CCh) were added, and extracellular Ca 2ϩ was elevated subsequently by adding 1 mM Ca 2ϩ to induce Ca 2ϩ entry. [Ca 2ϩ ] i was determined from the fluorescence ratio F 340 /F 380 according to Ref. 20. The fluorescence after sequential addition of 0.1% Triton X-100 and 50 mM EGTA to the cells provided the maximum fluorescence ratio (R max ) and the minimum fluorescence ratio (R min ), respectively. [Ca 2ϩ ] i was calculated using the formula shown in Equation 1.
␤ is the ratio of the emission fluorescence at 380 nm excitation in the presence of Triton X-100 and Triton X-100 plus EGTA, respectively. In experiments performed in the presence of DEANO, CCCP, or other agents that affect mitochondrial function, the compound was added 1 min prior to addition of TG or CCh. Alternative to fura-2 experiments, the intracellular free Ca 2ϩ was determined by loading the cells with Fluo-3 AM (2 M). Cell seeded on coverslips (6 ϫ 6 mm; poly-L-lysinecoated) were loaded with Fluo-3 in DMEM without fetal calf serum and incubated at 37°C under 5% CO 2 for 60 min. Fluo-3 loaded cells were washed twice with phosphate-buffered saline and kept in DMEM with-out fetal calf serum. Coverslips were then transferred to the perfusion chamber mounted on the microscope stage (Nikon Diaphot 300). Experiments were performed at room temperature. The cells were excited with a Polychrome II at 488 nm, and the emission was collected at 535 nm. Fluo-3 fluorescence was monitored on with a CCD camera system (Sensicam) and analyzed using Axon Imaging Workbench 2.1 software. Measurement of Mn 2ϩ Entry-For Mn 2ϩ quench experiments, the cells were loaded with fura-2/AM in DMEM without serum as described above. Cells were washed once with Tris buffer with 2.5 mM Ca 2ϩ , incubated for 20 min, and washed again with nominally Ca 2ϩ -free TRIS buffer. The experiments were conducted either in nominally Ca 2ϩ -free or in extracellular Ca 2ϩ -chelated condition as specified. Mn 2ϩ -induced quench of fura-2 fluorescence was measured at 360 nm (isosbestic point), and Mn 2ϩ entry was initiated by addition of 100 M Mn 2ϩ at 60 s. 3 M DEANO or 300 nM CCCP was added 200 s before stimulation of Mn 2ϩ entry by thapsigargin (100 nM). Control experiments were performed with vehicle only. Mn 2ϩ quench was calculated by measuring the quench in fluorescence per second caused by Mn 2ϩ entry and represented as change on fluorescence (F 360 ) per second.
Measurement of Mitochondrial Ca 2ϩ -For measurements of intramitochondrial Ca 2ϩ , cells were seeded on poly-L-lysine-coated coverslips (6 ϫ 6 mm) and loaded with rhod-2/AM in DMEM without fetal calf serum, at a concentration of 2 M for 45 min. at 37°C. Cells were illuminated with a Polychrome II at 540 nm, and fluorescence emission was collected at 605 nm. The rhod-2 fluorescence was monitored on a Nikon Diaphot 300 microscope equipped with a CCD camera system (Sensicam) and analyzed using Axon Imaging Workbench 2.1 software.
Electrophysiology-All patch-clamp experiments were performed using a extracellular solution containing (in mM) 50 Tris, 137 NaCl, 65 KCl, and 0.1 EGTA or 2.5 CaCl 2, pH adjusted to 7.4 with HCl. The pipette solution for whole cell experiments contained (in mM) 145 potassium gluconate, 15 KCl, 15 HEPES, 5 MgCl 2 , pH adjusted to 7.4 with N-methyl-D-glucamine. For perforated patch recordings and simultaneous current/fluorescence measurement, amphotericin B was added to the pipette solution to a final concentration of 100 g/ml (0.5% Me 2 SO). Measurement of Ca 2ϩ in fluo-3-loaded cells was performed simultaneously with current measurements in voltage clamped cells as described above. Pipettes were pulled from borosilicate glass (Clark Electromedical Instruments, Pangbourne, United Kingdom) and finally firepolished. The pipette resistance (filled) was in the range of 2-5 megohms. Recordings were performed in a 200-l bath chamber and with continuous perfusion by a gravity-driven system.
Voltage clamp and current amplification was performed with a List EPC-7 (List, Darmstadt, Germany) patch-clamp amplifier. Voltage ramp protocols covering a voltage range from Ϫ100 to ϩ80 mV with durations of 500 or 1000 ms were applied. Signals were low passfiltered at 1 kHz and digitized with 5 kHz. Stimulation protocols and recordings were controlled by Axon pCLAMP software (Axon Instruments, Foster City, CA) via an Axon Digidata 1200 computer interface.
Chemicals-Chemicals were purchased from the following suppliers,: Tissue culture medium was from Life Technologies, Inc. (Vienna, Austria); fura-2/AM and rhod-2/AM were from Molecular Probes (Leiden, Netherlands); DEANO was from Alexis (Switzerland); and all other chemicals were purchased from Sigma (Vienna, Austria).
Statistical Analysis-All the individual experiments were averaged, and the mean of the time courses is represented as trace. Results obtained at specific time points were expressed as mean value Ϯ S.E. Differences were considered statistically significant at p Ͻ 0.05 using Student's t test for unpaired values. Fig. 1 shows a comparison of TG-and CCh-induced Ca 2ϩ entry into vector-transfected (HEK VT ) and Trp3-overexpressing HEK293 cells (T3-9) based on fura-2 fluorescence measurements in cell populations. Expression of the hemagglutinin-tagged Trp3 channel protein in the cells was verified by Western blot experiments, which demonstrated the presence of the ϳ100-kDa hemagglutinin-tagged protein in membranes of T3-9 cells (data not shown). To activate the classical CCE pathway, we depleted intracellular stores with thapsigargin (100 nM) in Ca 2ϩ -free solution. Alternatively, carbachol (200 M) was used to induce phospholipase C/IP 3 -mediated release of Ca 2ϩ from intracellular stores. Ca 2ϩ entry was initiated by re-addition of extracellular Ca 2ϩ . As shown in Fig.  1A, elevation of extracellular Ca 2ϩ induced a peak intracellular Ca 2ϩ level of 1156 Ϯ 36 nM (n ϭ 6) in TG-stimulated cells, which was 3-fold higher than that measured in carbachol stimulated cells (373 Ϯ 14 nM; n ϭ 7). In Trp3-overexpressing cells, this relationship was distorted, in that TG-induced Ca 2ϩ signals were slightly reduced (1064 Ϯ 64 nM, n ϭ 10) whereas Ca 2ϩ entry in carbachol-stimulated cells (750 Ϯ 32 nM, n ϭ 12) was enhanced as shown in Fig. 1B. Thus, Trp3 overexpression clearly promoted or generated a phospholipase C-dependent Ca 2ϩ entry pathway but not a classical CCE.

Trp3 Overexpression Fails to Promote Thapsigargin-induced CCE but Generates a Carbachol-stimulated Ca 2ϩ Entry Pathway in HEK293 Cells-
NO Inhibits Thapsigargin-induced CCE in Vector-transfected HEK293 Cells but Not in T3-9 Cells- Fig. 2 (A-D) illustrates the effects of the NO donor DEANO (3 M) on thapsigargin-induced CCE in HEK VT and T3-9 cells in suspension. CCE into HEK VT was markedly reduced in the presence of the NO donor DEANO (3 M), resulting in a reduction of peak intracellular Ca 2ϩ levels from 1156 Ϯ 36 nM to 425 Ϯ 41 nM (n ϭ 6), as shown in Fig. 2A. By contrast, thapsigargin-induced CCE into T3-9 cells was completely insensitive to DEANO (988 Ϯ 223 nM, n ϭ 8, Fig. 2B). Similar results were obtained with authentic NO at a concentration of 100 M (data not shown). 8-Bromo-cGMP, a cell permeant cGMP analogue, did not mimic the effects of NO in respect to Ca 2ϩ mobilization and inhibition of CCE (1146 Ϯ 41 nM, n ϭ 5), suggesting that the effect of NO was independent of cGMP. The action of NO was typically associated with a slight increase in cytoplasmic Ca 2ϩ that was transient in nature and did not require extracellular Ca 2ϩ (see Fig. 2A). Fig. 2 (C and D) illustrates the effects of DEANO on TGinduced Mn 2ϩ entry in HEK VT and T3-9 cells, respectively. DEANO clearly suppressed TG-induced Mn 2ϩ entry into HEK VT cells (Fig. 2C) but not in T3-9 cells (Fig. 2D). The observed lack of sensitivity to NO in T3-9 cells indicates that Trp3 is able to modify the regulatory property of CCE channels. Trp3-induced elimination of NO sensitivity of CCE was confirmed in single fluo-3-loaded HEK293 cells, as shown in Fig. 2 (E and F). Thapsigargin-stimulated Ca 2ϩ entry was again markedly inhibited by the NO donor in vector-transfected controls but not in T3-9 cells.
Trp3 Overexpression Generates Sensitivity of TG-induced Mn 2ϩ Entry to Extracellular Ca 2ϩ -The Mn 2ϩ quench experi-ments illustrated in Fig. 2 were performed in a buffer containing EGTA (100 M), resulting in free Ca 2ϩ levels below 1 M. T3-9 cells exhibited reduced TG-induced Mn 2ϩ quench in nominally Ca 2ϩ -free solutions. As shown in Fig. 3, a clear dependence of TG-induced Mn 2ϩ entry on extracellular Ca 2ϩ was observed in T3-9 cells but not in HEK VT cells. Chelation of extracellular Ca 2ϩ with EGTA restored Mn 2ϩ entry in T3-9 cells to a level comparable with that of controls, indicating that Mn 2ϩ permeation through CCE channels is affected by overexpression of Trp3 in that the TG-sensitive Mn 2ϩ entry pathway gained sensitivity to extracellular Ca 2ϩ .
Inhibition of Thapsigargin-induced CCE by Mitochondrial Uncoupling Is Impaired in T3-9 Cells-The inhibitory action of NO was mimicked by the mitochondrial uncoupler CCCP. TGstimulated CCE into control cells (HEK VT ) was reduced from 1156 Ϯ 36 nM to 416 Ϯ 38 nM in the presence of CCCP (300 nM; Fig. 4A), corresponding to an inhibition of CCE by 64%. Similar to NO, CCCP by itself induced a modest, transient elevation of cytoplasmic Ca 2ϩ . CCE into T3-9 cells was less sensitive to inhibition by CCCP as compared with control cells (Fig. 4B). CCCP-induced inhibition amounted to 28% in T3-9 cells. CCCP inhibited TG-induced Mn 2ϩ entry into control as well as in T3-9 cells (Fig. 4, C and D). Nonetheless, inhibition of Mn 2ϩ entry into T3-9 cells was less pronounced than in vector-transfected controls (41% versus 63%, respectively). Fig. 5 (A and B) illustrates the effects of the mitochondrial uncouplers antimycin A 1 (2 M) and oligomycin (6 M) on TGinduced Ca 2ϩ signals in single fluo-3-loaded cells. Interestingly, antimycin A 1 which prevents the generation of mitochondrial H ϩ gradients and rapidly disturbs Ca 2ϩ uptake into mitochondria (21), was most effective as an inhibitor of CCE, whereas oligomycin, which primarily inhibits ATP synthase (22), failed to inhibit CCE significantly, indicating a key role for mitochondrial Ca 2ϩ handling. Fig. 5C compares the sensitivity of Ca 2ϩ entry into HEK VT and T3-9 cells to inhibition by DEANO, CCCP, antimycin A 1 , and oligomycin. In T3-9 cells, the sensitivity of CCE to inhibition by the NO donor as well as by the mitochondrial uncouplers CCCP and antimycin A 1 were strongly reduced or eliminated.
The similarities between the effects of NO and those of mitochondrial uncouplers supported the hypothesis of mito- chondria as a primary target of NO action. Therefore, we tested the effects of NO on Ca 2ϩ handling by mitochondria.
NO Prevents Ca 2ϩ Sequestration by Mitochondria-DEANO and CCCP induced a moderate rise in basal intracellular Ca 2ϩ concentrations in the absence of extracellular Ca 2ϩ , as evident from the experiments shown in Figs. 2 and 4. The mobilization of intracellular Ca 2ϩ by CCCP is known to result mostly from depolarization of mitochondria and the consequent depletion of mitochondrial Ca 2ϩ . To test whether NO mimics CCCP in modulation of mitochondrial Ca 2ϩ handling, we performed experiments with rhod-2 to measure changes in mitochondrial Ca 2ϩ . As shown in Fig. 6A, rhod-2 fluorescence was localized within the cells in discrete spots and increased upon stimulation with CCh (200 M), whereas rhod-2 fluorescence decreased significantly upon administration of either DEANO (3 M) or CCCP (300 nM). Moreover, CCh-induced increase in rhod-2 fluorescence was effectively prevented by both agents. Averaged time courses of mitochondrial rhod-2 fluorescence during exposure of cells to DEANO, CCCP, and CCh are given in Fig.  6 (C, F, and I), illustrating CCh-induced changes in rhod-2 fluorescence in the absence of drugs (n ϭ 55) and in the presence of DEANO (n ϭ 66) or CCCP (n ϭ 59), respectively.
NO as Well as CCCP Promote CCh-activated Ca 2ϩ Entry into T3-9 Cells-Trp3 overexpression has been demonstrated previously to create a PLC-and IP 3 -dependent Ca 2ϩ entry pathway (6). This function of Trp3 was again confirmed by the promotion of CCh-induced Ca 2ϩ signaling observed in the present study (Fig. 1). Fig. 7 illustrates that DEANO (3 M) as well as CCCP (300 nM) slightly inhibited CCh-induced Ca 2ϩ signaling in control cells (Fig. 7, A and C) but promoted CCh-induced Ca 2ϩ entry into T3-9 cells (Fig. 7, B and D). In T3-9 cells, DEANO (3 M), authentic NO (100 M; n ϭ 4; data not shown) as well as CCCP (300 nM) increased the CCh-induced Ca 2ϩ entry signals measured in cell populations. Both NO as well as CCCP reduced the CCh-induced Ca 2ϩ mobilization in Ca 2ϩfree solution but augmented the Ca 2ϩ signals induced by subsequent Ca 2ϩ entry. Promotion of CCh-induced Ca 2ϩ entry into T3-9 cells was associated with a change in kinetics of the Ca 2ϩ signals. In the presence of DEANO or CCCP Ca 2ϩ entry-induced intracellular Ca 2ϩ levels did not decline but remained elevated during prolonged rises in extracellular Ca 2ϩ . The concentration of Ca 2ϩ measured at 200 s after initiation of Ca 2ϩ entry was 475 Ϯ 14 nM (n ϭ 12) in the presence of vehicle and 758 Ϯ 15 nM (n ϭ 10) in the presence of DEANO. Similarly, CCCP promoted Ca 2ϩ signals in T3-9 cells, resulting in significantly elevated intracellular Ca 2ϩ levels (968 Ϯ 94 nM, n ϭ 6, Fig. 7D) at 200 s after initiation of entry. Thus, the Trp3derived PLC-regulated Ca 2ϩ entry pathway and CCE were divergently regulated by NO. Because Trp3 channels were found previously to exhibit Ca 2ϩ -mediated negative feedback modulation similar to CCE channels, it was of interest to investigate the effects of NO on the relation between bulk cytoplasmic Ca 2ϩ and Trp currents in single adherent HEK293 cells. Fig. 8 shows time courses of intracellular Ca 2ϩ and membrane currents measured simultaneously in single T3-9 cells, which were stimulated with CCh in Ca 2ϩ -free solution and subjected to subsequent elevation of extracellular Ca 2ϩ in the absence (Fig. 8A) and presence (Fig. 8B)  extracellular Ca 2ϩ . This CCh-stimulated membrane conductance was not detected in vector-transfected control cells (n ϭ 3, data not shown). Two distinct modes of regulation of Trp3 channels were evident. (i) Trp3 channels were activated in response to stimulation of muscarinic receptors, a situation that is characterized by depletion of intracellular Ca 2ϩ stores and a transient rise in cytoplasmic Ca 2ϩ ; and (ii) the channels were suppressed during subsequent elevation of extracellular Ca 2ϩ , which caused large rises in cytosolic Ca 2ϩ via Ca 2ϩ entry. Fig. 9 shows a comparison of intracellular Ca 2ϩ levels and current densities obtained in parallel experiments. In the absence of DEANO, Ca 2ϩ re-addition was associated with a sub- stantial and significant reduction in membrane currents. This Ca 2ϩ entry-induced suppression of Trp3 currents was less pronounced in the presence of DEANO. By contrast, Ca 2ϩ entryinduced increments in cytoplasmic Ca 2ϩ were significantly larger in the presence of the NO donor (⌬F 488 ϭ 50 Ϯ 2) as compared with controls (⌬F 488 ϭ 36 Ϯ 1). Thus, NO affects Ca 2ϩ -dependent autoregulation of PLC-regulated Trp3 channels in a manner oppositional to that of native CCE channels.

DISCUSSION
The present study demonstrates that expression of Trp3 modifies essential regulatory properties of CCE channels in HEK293 cells, resulting in reduced sensitivity of the CCE pathway to NO as well as to uncoupling of mitochondrial respiration. Our findings favor a model in which Trp3 proteins function as a regulatory subunit of CCE channels. We provide evidence for a dual effect of Trp3 expression on cellular Ca 2ϩ entry pathways, i.e. generation of a PLC-regulated Ca 2ϩ entry pathway and modification of regulatory and biophysical properties of CCE.
Previous reports have shown that overexpression of Trp3 generates Ca 2ϩ entry channels, which are activated by stimulation of phospholipase C-coupled receptors (5,6). Consistently, phospholipase C-derived second messengers, IP 3 , and/or diacylglycerols, were found to activate Trp3-derived cation channels (11,12). Although CCE appears to be tightly associated with stimulation of phospholipase C, and Trp3 appears as an attractive link, which transmits changes in the Ca 2ϩ content of intracellular stores to the plasma membrane because of its ability to bind to IP 3 receptor proteins (7), the role of Trp3 in CCE is uncertain. Trp3 overexpression failed to promote the classical CCE pathway in several previous studies (5,10). These findings were confirmed by our present study. Nonetheless, Trp3 expression was reported to produce changes in a pharmacological property of CCE channels, e.g. the sensitivity to block by Gd 3ϩ (10), suggesting that Trp3 interferes with endogenous CCE channels to form heteromeric CCE channel complexes of specific properties. The present study was designed to test whether Trp3 overexpression modifies specific regulatory properties of CCE channels. We focused on the regulation by NO, which was demonstrated to inhibit native CCE channels (15,17), and on the sensitivity to mitochondrial function, which has been shown to control CCE by determining negative feedback mechanisms (14).
Inhibition of CCE Channels by NO-Nitric oxide is a potent modulator of cellular Ca 2ϩ homeostasis because of cGMP-dependent and cGMP-independent mechanisms such as S-ni-trosylation of proteins (23). Here we demonstrate that NO, administrated as authentic NO in solution or by use of the NO donor DEANO, substantially inhibits endogenous CCE of HEK293 cells. Similar inhibitory effects of NO on CCE have been reported previously for platelets (15) and smooth muscle cells (17), and have been attributed to promotion of uptake of Ca 2ϩ into the ER (17). Enhanced refilling of the ER as the basis of the inhibitory action of NO observed here appears unlikely because CCE was activated by exposure of cells to thapsigargin, which effectively prevents refilling of the regulatory pool of the ER. In contrast to promotion of refilling of intracellular Ca 2ϩ pools, we observed a slight, distinct release of Ca 2ϩ from intracellular storage sites by NO. The NO-sensitive Ca 2ϩ store was identified as mitochondria. This effect of NO on intracellular Ca 2ϩ handling by mitochondria is likely to be involved in suppression of CCE because Ca 2ϩ sequestration by mitochondria has been demonstrated as a determinant of CCE (14). NO-mediated suppression of mitochondrial function as well as respiration has been observed in various tissues (14,24,25), and this phenomenon has been attributed to direct, cGMPindependent effects of NO on components of the mitochondrial respiratory chain, in particular on cytochrome oxidase (24,25). Consistently, inhibition of CCE by NO was found independent of cGMP, as 8-bromo-cGMP failed to mimic NO in terms of Ca 2ϩ mobilization or inhibition of CCE.
Role of Mitochondria-Our results demonstrate that NO and the mitochondrial uncoupler CCCP exert similar effects on Ca 2ϩ signaling in HEK293 cells. Both compounds were found to inhibit CCE at concentrations that mobilize Ca 2ϩ from mitochondria and prevent Ca 2ϩ uptake into this organelle during IP 3 -induced Ca 2ϩ release. A link between impairment of mitochondrial Ca 2ϩ sequestration ability and inhibition of CCE has been demonstrated previously for T-lymphocytes (14). Thus, mitochondria may well represent the primary target of NO within the mechanism leading to modulation of Ca 2ϩ entry. This role of mitochondria was further confirmed by experiments with the mitochondrial uncoupler antimycin A 1 , which exerted inhibitory effects on CCE in vector-transfected HEK293 cells. Antimycin A 1 suppresses the build-up of mitochondrial H ϩ gradients, resulting in a rapid reduction of mitochondrial membrane potential and impaired Ca 2ϩ uptake (21). By contrast, oligomycin, an inhibitor of ATP synthase that fails to rapidly depolarize mitochondria (22), was barely effective as an inhibitor of CCE. Our results strongly suggest a role of mitochondrial membrane potential and Ca 2ϩ handling in the observed regulation of CCE.
Modulation of the NO Sensitivity of CCE Channels by Trp3-In Trp3-overexpressing cells, the regulatory properties of CCE were strikingly altered, in that CCE lost its sensitivity to NO. This substantial change in cellular regulation of CCE may be explained by either a change in the CCE ion channel complex itself or by a change in the regulatory mechanisms involved. It is tempting to speculate that Trp3 interferes with the formation of CCE channel complexes, which, in wild type HEK293 cells, are composed of as yet unidentified channel proteins. Overexpression of Trp3 protein may alter the stoichiometry of these CCE channels, resulting in heteromultimeric complexes containing Trp3 in addition to the generation of Trp3 homomultimers. A respective model, which is consistent with the observation of three Ca 2ϩ entry pathways of different regulatory properties, is depicted in Fig. 10. The ability of Trp proteins to form functional heteromultimers has repeatedly been demonstrated (12,26), and it is reasonable to speculate that the endogenous CCE channel proteins are members of the Trp family. Alternatively, we cannot exclude at present that Trp3 functions as a regulatory protein of CCE channels without being a component of the pore-forming channel complex itself. Nonetheless, in case that Trp3 is able to contribute to a CCE channel complex, e.g. a tetrameric ion channel as illustrated in the model shown in Fig. 10, it appears likely that Trp3 alters the biophysical and pharmacological properties of CCE. Indeed, a Trp3-induced change in the sensitivity of CCE channels to the pore blocker Gd 3ϩ has been reported previously (10). Interestingly, Mn 2ϩ quench experiments revealed a Trp3-induced change in the sensitivity of CCE channels to extracellular Ca 2ϩ . Mn 2ϩ entry was barely affected by micromolar (10 M) concentrations of extracellular Ca 2ϩ in vector-transfected cells but significantly suppressed by 10 M extracellular Ca 2ϩ in Trp3-overexpressing cells. Hence, Trp3 expression resulted in a unique Ca 2ϩ sensitivity of the thapsigargin-stimulated Mn 2ϩ entry, suggesting a change in the divalent binding properties and a contribution of Trp3 to the pore-forming complex of CCE channels.
Promotion of Phospholipase C-regulated Trp3 Channel Activity by NO-It has been demonstrated repeatedly that overexpression of Trp3 generates a phospholipase C-dependent Ca 2ϩ entry pathway. We report here that this Ca 2ϩ entry pathway is insensitive to inhibition by NO or mitochondrial uncouplers. Interestingly, carbachol-induced Ca 2ϩ mobilization from intracellular stores was blunted by NO as well as CCCP in T3-9 but not in control cells. The mechanism by which mitochondrial uncoupling interferes with IP 3 -induced Ca 2ϩ mobilization is unlikely to involve depletion of an endoplasmic reticulum Ca 2ϩ pool because thapsigargin-induced Ca 2ϩ release was not affected by either CCCP or NO. Despite suppression of Ca 2ϩ release, both NO and CCCP failed to inhibit but rather augmented CCh-induced Ca 2ϩ entry. Oppositional regulation of CCh-stimulated Trp3 channels and CCE channels was surprising because our results suggest that NO as well as CCCP acts by alteration of local Ca 2ϩ feedback mechanisms, and inhibitory autoregulation by Ca 2ϩ has been reported for both Trp3 channels as well as CCE channels. Inhibition of Trp3 by cytoplasmic Ca 2ϩ was again confirmed in the present study by simultaneous measurements of intracellular Ca 2ϩ and membrane currents. In single voltage-clamped cells, Trp3-mediated cation currents measured during re-addition of extracellular Ca 2ϩ were promoted, and the Trp3-mediated Ca 2ϩ entry signal was augmented by NO. Elevation of extracellular Ca 2ϩ from micromolar to millimolar concentration is known to suppress the Trp3-mediated cation conductance because of extra-cellular as well as intracellular inhibitory effects of Ca 2ϩ (12). This Ca 2ϩ -mediated inhibitory modulation of Trp3 channels was suppressed by NO, a phenomenon that is in clear contrast to the observed inhibitory effects of NO and CCCP on endogenous CCE channels in vector-transfected controls. The divergent regulation of these two Ca 2ϩ entry channels by NO and CCCP indicates a distinctly different coupling of the endogenous CCE channels and the phospholipase C-sensitive Trp3 channels to mitochondria. We suggest that a specific property of Trp3 enables up-regulation of channel activity in response to impaired mitochondrial Ca 2ϩ sequestration. It is tempting to speculate that the specific functional interaction of Trp3 proteins with IP 3 receptors that have been reported to communicate tightly with mitochondria (27,28) may provide the basis of this specific coupling. Alternatively, PLC-regulated Trp3 channels and endogenous CCE channels may be targeted to distinct microdomains of the plasma membrane, which may accommodate divergent local changes in cytoplasmic Ca 2ϩ during mitochondrial uncoupling. The molecular property that renders Trp3 channels insensitive to inhibition by NO and mitochondrial uncoupling remains to be clarified. Nonetheless, our study clearly demonstrates that overexpression of Trp3 confers insensitivity to inhibition by NO, as a Trp3-specific regulatory property, to endogenous CCE channels.
In summary, our results demonstrate that Trp3 expression determines Ca 2ϩ entry by generation and modification of two distinct channel types, i.e. phospholipase C-controlled Ca 2ϩ entry channels and classical CCE channels. These two channel types are oppositionally controlled by nitric oxide and mitochondria. Our results provide evidence for a close relationship between Trp3 expression and the function of CCE channels. We suggest that Trp3 determines physiologically important properties of CCE channels such as sensitivity to regulation by NO and mitochondrial function. Thereby, variations in the expression of Trp3 may provide the basis of diverse properties of CCE channels in different tissues and cell types. The model is based on the assumption of tetrameric Ca 2ϩ entry channel complexes and accounts for the observation that Trp3 overexpression produces two Ca 2ϩ entry pathways with regulatory properties different from that of the Ca 2ϩ entry pathways in native HEK293 cells. Channels' sensitivity to store-depletion, NO/mitochondrial uncoupling and PLC-stimulation are given in boxes. ϩ and -signs represent promotion and inhibition respectively. Dashed arrow with cross represents insensitivity.