PKC-dependent coupling of calcium permeation through transient receptor potential canonical 3 (TRPC3) to calcineurin signaling in HL-1 myocytes

Identification of E630 as a Critical Residue in the TRPC3 Selectivity Filter.

In an attempt to obtain TRPC mutants with altered ion selectivity, we initially generated a structural model of the TRPC3 pore region by using a recently developed alignment strategy (13) and structure information available for KcsA as well as a Kv1.2-Kv1.3 chimeric pore. Our hypothetical pore model is illustrated in Fig. S1A along with the sequence comparison of TRPC3 and template pore structures (Fig. S1B). Three of the five negative residues were predicted to be accessible and exposed to the permeation pathway. According to our molecular model, only one glutamate (E630) was localized within the central part of the permeation pathway, whereas the other two residues (E616 and D639) were predicted as part of the extracellular vestibule. Mutagenesis and functional analysis confirmed a critical role of the central glutamate in position 630. Charge inversion (E630K) yielded a nonfunctional channel (Fig. S2), whereas neutralization (E630Q) produced a channel that displayed moderately altered current to voltage (I-V) relation in normal extracellular (Na⁺ plus Ca²⁺ containing; Fig. 1 A and C) solution with a relative increase in the conductance at neutral potential. The I-V relation of the TRPC3-E630Q mutant was virtually insensitive to changes in extracellular Ca²⁺ (Fig. S3). Inspection of I-V relations with Ca²⁺ as the sole extracellular cationic charge carrier and BAPTA in the pipette solution to eliminate indirect, Ca²⁺-mediated currents revealed that the E630Q mutation profoundly reduced Ca²⁺ permeation although the channel complex (Fig. 2 B and D). Inward currents were essentially small or lacking with Ca²⁺ as a charge carrier even at large hyperpolarizing potentials, and reversal potentials were difficult to determine in most experiments. Nonetheless, from six experiments a mean of −79.6 ± 6.3 mV was calculated, demonstrating a substantial shift in reversal potential compared with wild-type channels (1.9 ± 1.4; n = 7). The Ca²⁺/Cs⁺ permeability ratio was reduced from 4.2 to less than 0.02 in the mutant. Thus, our experiments identified a key amino acid within the cation permeation structure of TRPC3. The negative charge in position 630 is apparently essential for divalent permeation but not for transition of monovalent cations through the TRPC3 pore. Our finding is in line with previous reports on the role of negatively charged residues in Ca²⁺ transition through TRP pores (14–17) and, specifically, with a prediction of critical residues in the TRPC selectivity filter obtained by Liu et al. in a study with the prototypical Drosophila TRP channel (14). It is of note that the identified negatively charged residue is conserved in the TRPC3/6/7 subfamily but absent in more distant relatives of TRPC3. Therefore, Ca²⁺ permeation in within the TRPC family of cation channels may involve distinctly different molecular mechanisms. Our results support the notion that monovalent and divalent permeation through nonselective TRP cation channels may involve separate, specific interaction sites within the pore, which combine to a “nonselective” pathway that conducts both types of charge carriers. Based on our observation that a single point mutation within the TRPC3 pore causes specific elimination of Ca²⁺ permeation, we set out to use this genetically engineered cation channel to explore the cellular impact of the TRPC3-mediated monovalent conductance and to identify downstream signaling pathways that are specifically linked to either the monovalent transport or to Ca²⁺ entry through the channel pore. Because the TRPC monovalent conductance is considered of particular importance in excitable cells, and because recent studies have demonstrated the relevance of TRPC channels in cardiac pathophysiology (2, 4, 7, 18) we focused on the cardiac system, using the HL-1 murine atrial cell line. Initially, we compared basic properties of TRPC3 signaling in HL-1 cells with those in the well characterized electrically nonexcitable HEK293 system.

TRPC3 Mediates Agonist-Induced Ca²⁺ Signals in HEK293 and HL-1 Cells by Divergent Mechanisms.

So far, the relative contribution of direct Ca²⁺ permeation through the TRPC3 pore and indirect mechanisms involving TRPC-mediated changes in membrane potential and voltage-dependent signaling partners such as NCX1 has not been evaluated in HEK293 cells. Expression levels of endogenous voltage-gated Ca²⁺ channels are below the detection threshold, and NCX1 expression is typically moderate to low. Hence, only a minor fraction of the TRPC3-mediated Ca²⁺ signal is expected to involve indirect mechanisms in HEK293. Pharmacological characterization of the TRPC3-mediated Ca²⁺ entry pathway in HEK293 and HL-1 cells, determined by using a classical Ca²⁺ readdition protocol, revealed that global Ca²⁺ signals were based on distinctly different mechanisms (Fig. S4). Electrophysiological experiments confirmed that TRPC3 channels were active when Ca²⁺ was elevated after activation by agonist administration in Ca²⁺-free solution (Fig. S5). TRPC3 overexpressing HEK293 as well as HL-1 displayed Ca²⁺ entry that was highly sensitive to inhibition by the TRPC3 blocker Pyr3 (19). KB-R7943, an inhibitor of NCX reverse mode operation, suppressed Ca²⁺ entry only moderately in HEK293 cells and lacked inhibitory effects in HL-1 cells. Block of voltage-gated, L-type Ca²⁺ channels by nifedipine strongly suppressed Ca²⁺ entry into endothelin-stimulated HL-1 cells but had no effect on the Ca²⁺ signal in HEK293 cells (Fig. S4). These results indicate that TRPC3 is effectively linked to voltage-gated Ca²⁺ signaling in cardiac cells and are able to produce large global cytosolic Ca²⁺ rises via promotion of Ca²⁺ entry through Ca_V1.2 channels. The observed lack of NCX-mediated Ca²⁺ entry into HL-1 cells may be explained by predominant forward mode operation in these cardiac cells, based on a tight functional coupling to voltage-gated Ca²⁺ entry channels as well as more negative membrane potentials compared with HEK293 cells. Thus, HL-1 myocytes represent an electrically excitable cell type that displays functional cross-talk and signaling partnership between nonselective TRPC conductances and voltage-gated Ca²⁺ channels. Thereby, TRPC3 signaling in HL-1 cells is distinctly different from that in the nonelectrically excitable HEK293 cell system.

As a next step, we aimed to delineate the cellular role of direct Ca²⁺ permeation through TRPC3 channels in these cells by characterizing coupling between Ca²⁺ entry and the NFAT downstream effector system for wild type and pore mutants (E630Q and E630K) of TRPC3.

Ca²⁺ Permeation Through the Pore of TRPC3 Channels Is Essential for Activation of the NFAT Pathway.

Carbachol-induced Ca²⁺ entry as well as NFAT translocation was strongly reduced by expression of either the Ca²⁺ permeation-deficient mutant (E630Q) or a pore-dead mutant (E630K) (Fig. 2). Basal Ca²⁺ entry into nonstimulated cells was reduced when expressing either of the TRPC3 pore mutants to the level of vector-transfected controls (Fig. 2B). As expected from the proposed NCX1-mediated Ca²⁺ entry contribution, TRPC3-E630Q did not fully eliminate the Ca²⁺ entry signal. Ca²⁺ entry into cells expressing TRPC3-E630Q remained significantly higher than in cells transfected to express TRPC3-E630K. Nonetheless, NFAT translocation in HEK293 cells was completely suppressed with either pore mutation of TRPC3 (Fig. 2 C and D). This finding indicated that Ca²⁺ permeation through the pore is essential to initiate NFAT translocation, whereas indirect NCX-mediated signaling was barely involved because the NCX inhibitor KB-R7943 (5 μM) failed to prevent NFAT translocation (Fig. S6).

A possible dual Ca²⁺ signaling function of TRPC3 was further investigated in the electrically excitable cardiac HL-1 cell line. As illustrated in Fig. 3, Ca²⁺ entry into endothelin-stimulated cells was slightly enhanced compared with vector-transfected controls (Fig. 3A) by expression of either wild-type TRPC3 (Fig. 3B) or TRPC3-E630Q (Fig. 3C) but reduced down to basal (nonstimulated) level with TRPC3-E630K (Fig. 3D). Expression of wild-type TRPC3 (Fig. 3B) or TRPC3-E630Q (Fig. 3C) generated a Ca²⁺ signal that was mainly based on voltage-gated Ca_V1.2 channels as evident by its sensitivity to nifedipine (3 μM). In clear contrast to the observed Ca²⁺ signals, cells expressing TRPC3-E630Q lacked endothelin-stimulated NFAT translocation (Fig. 3C). Thus, the TRPC3 mutant with impaired divalent conductance (E630Q) is able to initiate a large global Ca²⁺ signal via promotion of voltage-gated Ca²⁺ entry, but this signal is not translated into NFAT activation. Importantly, even endogenous TRPC3 channels appear sufficient to exert a significant impact on NFAT translocation as evident from vector-transfected controls displaying higher translocation than cells expression the pore-dead dominant-negative E630K mutant. It is of note that, in contrast to HEK293 cells, NFAT translocation in HL-1 cells was not significantly promoted in response to depletion of intracellular stores with thapsigargin (Fig. S7), indicating that the channels involved in NFAT signaling of HL-1 are not classical store-operated Ca²⁺ entry channels. Opening of TRPC3 channels resulted in NFAT activation, but also generation of an additional intracellular Ca²⁺ signal mediated by voltage-gated Ca²⁺ channels that was fairly well segregated from the NFAT pathway. Notably, TRPC3-mediated NFAT activation can occur at barely detectable global Ca²⁺ changes such as in the presence of nifedipine (3 μM) in control cells or cells overexpressing TRPC3 (Fig. 3 A and B), therefore we hypothesized that the triggering Ca²⁺ elevation is likely to take place in a restricted signaling microdomain at the TRPC3 channel complex, containing essential downstream signaling components such as calmodulin and calcineurin (CaN) to allow specific transduction of this local Ca²⁺ signal. Indeed, direct association of TRPC3 with CaN have been demonstrated (20, 21) and the existence of a dynamic TRPC/CaN signaling complexes have been proposed.

Coupling of Cardiac TRPC3 Signaling to Activation of the NFAT Pathway Involves PKC-Dependent Phosphorylation.

Previous investigations demonstrated assembly of CaN along with immunophyllins (FKBP12) into TRPC6 signalplexes and dependency of this process on protein kinase C-mediated phosphorylation (21). PKC-mediated phosphorylation of TRPC3 appears essential for both recruitment of CaN into TRPC complexes and inhibitory regulation (22, 23). This result prompted us to hypothesize that suppression of PKC phosphorylation may disrupt the functional TRPC3/CaN signaling unit without preventing channel function. Consequently, we set out to test whether TRPC3 linkage to NFAT nuclear translocation depends on regulation of the channel complex by PKC. To suppress TRPC3 phosphorylation by PKC isoenzymes, we performed experiments with GF109203X, a compound that inhibits conventional PKC isoforms including PKC-γ, as one essential player in the control of TRPC3 channels (23). Because CaN has been shown to associate with TRPC3/6 channels in a manner dependent on phosphorylation by PKC (21), we speculated that prevention of PKC phosphorylation may disrupt TRPC/CaN complexes. Indeed, immunoprecipitation experiments in HEK293 cells confirmed that the PKC inhibitor prevents association of CaN into TRPC3 complexes along with reduction of threonine phosphorylation of the channel protein (Fig. 4). Alternatively, a mutant that is defective in PKC-γ–mediated inhibitory modulation, i.e., T573A corresponding to the murine TRPC3-moonwalker (Mwk) mutation, was expressed in HEK293 and HL-1 cells. It is of note that overexpression of the phosphorylation-deficient TRPC3-Mwk by itself was barely tolerated by the cells, presumably due to a gain in function leading to Ca²⁺ overload. Therefore, we transfected cells to overexpress the T573A mutant along with wild-type TRPC3 (DNA ratio 3:1). This transfection resulted in a heteromeric TRPC3 conductance larger than that generated by wild-type TRPC3 alone (Fig. S8) but essentially tolerated by the host cells. The derived heteromers appeared correctly targeted into the membrane as indicated by fluorescence microscopy. Interestingly, currents through TRPC3-Mwk/TRPC3-WT channels were rather stable during continuous agonist stimulation, most likely due to lack of inhibitory regulation of the channels by PKC and recordings in Na⁺-free extracellular solution confirmed Ca²⁺ permeability of the channels.

Either treatment of cells expressing TRPC3-WT with GF109203X (2 μM) or transfection of cells with TRPC3-Mwk/TRPC3-WT produced a similar cellular phenotype, displaying large-agonist–induced Ca²⁺ entry signals that were not accompanied by significant NFAT nuclear accumulation (Fig. 5). This “transcriptionally silent” Ca²⁺ entry into HL-1 cells was for a large part mediated by voltage-gated L-type Ca²⁺ channels as indicated by sensitivity to nifedipine (3 μM). Our results demonstrate disruption of transcriptional TRPC3 signaling in response to reduced PKC-dependent phosphorylation of the channel or as a consequence of the TRPC3-Mwk mutation. This fact may be considered as part of the molecular mechanism underlying the moonwalker pathophysiology (23).

Impaired PKC regulation of cardiac TRPC3 is shown to result in uncoupling from the NFAT pathway without disrupting the linkage between TRPC3 and global myocyte Ca²⁺ via voltage-gated Ca²⁺ entry. We provide evidence that TRPC3-CaN/NFAT signaling takes place in a restricted microdomain and requires both direct Ca²⁺ permeation through the TRPC3 pore as well as CaN targeting into the signal complex. Cardiac TRPC3 complexes are shown to produce Ca²⁺ signals both via direct Ca²⁺ transport and by control of voltage-dependent Ca²⁺ entry. Our results demonstrate the ability of cardiac TRPC3 channels to switch in a phosphorylation-dependent manner between a transcriptionally active and a transcriptionally silent signaling mode (Fig. 6). Because TRPC3 is likely to change in expression along with other TRPC species during pathophysiologic stress, the formation of divergent heteromeric TRPC3 channel complexes may be anticipated. The observed dominant negative effect of the phosphorylation-deficient moonwalker mutation on CaN/NFAT activation suggests a prominent role of phosphorylation in maintenance of transcriptionally active TRPC complexes in the heart. Situations of hampered PKC phopshorylation or promoted dephosphorylation may convert TRPC complexes into a cardiac Ca²⁺ signaling unit that is functionally isolated from the CaN/NFAT activation pathway.

Our findings highlight the key role of nonselective TRPC channels in the control of transcriptional programs and extend this concept by demonstrating a pivotal role of Ca²⁺ transport trough the TRP pore structure along with a unique phosphorylation-dependent molecular switch that allows efficient control of cardiac gene transcription by neurotransmitters and hormones.

Homology Modeling.

For details on sequence alignment, homology modeling and model evaluation, see SI Materials and Methods.

DNA and Mutagenesis.

Site-directed mutagenesis was performed with standard protocols. Details on cDNA constructs and cloning procedures are provided in SI Materials and Methods.

Cell Culture and Transfection.

Cell lines were cultured at 37 °C and 5% CO₂. For HEK293 cells DMEM (Invitrogen) supplied with 10% FCS and for HL-1 atrial myocytes Claycomb medium (Sigma) supplied with 100 μM norepinephrin, 4 mM l-glutamin and 10% FCS were used. Lipofection was used for gene transfer; for details on DNA amounts and reagents, see SI Materials and Methods.

Electrophysiology.

Standard patch clamp protocols were used (SI Materials and Methods). Standard bath solutions contained 140 or 0 mM NaCl, 0 or 140 mM NMDG, 2 mM MgCl₂, 10 mM glucose, 10 mM Hepes, 2 or 0 mM CaCl₂, and 0 or 2 mM BaCl₂ at pH adjusted to 7.4 with NaOH or NMDG. Pipette solution contained 120 mM cesium methanesulfonate, 20 mM CsCl, 15 mM Hepes, 5 mM MgCl₂, and 3 mM EGTA, at pH adjusted to pH 7.3 with CsOH. For delineation of Ca²⁺ permeability of TRPC3 mutants, a bath solution containing 132 mM NMDG, 2 mM MgCl₂, 10 mM Glucose, 10 mM Hepes, 3 mM CaCl₂, 7 mM Ca-Gluconate, at pH adjusted to 7.4 with methanesulfonic acid and a pipette solution composed of 140 mM cesium methanesulfonate, 15 mM Hepes, 5 mM MgCl₂, and 10 mM BAPTA at pH 7.3 was used.

Measurement of NFAT-Translocation.

Cells were transfected to express an N-terminally GFP-tagged NFATc1 fusion (15) and plated on coverslips. For buffers and solutions see SI Materials and Methods. Agonists as well as inhibitors (Pyr3, GFX109203, nifedipine, or KB-R7943) remained present continuously after administration. GFP-NFAT translocation was monitored (488 nm excitation) with standard fluorescence microscopy (Zeiss Axiovert equipped with Coolsnap HQ). GFP-NFAT and YFP-TRPC3-WT/-mutant fluorescence were discriminated by specific cellular localization. Nuclear/cytosol fluorescence intensity ratios of cells were calculated with ImageJ software.

Measurement of Intracellular Ca²⁺ Signaling.

For details on fura-2 calcium imaging experiments see SI Materials and Methods.

Immunoprecipation.

In short, protein-A- or protein-G-bead-precleared supernatants of lysates from stimulated HEK293 cells were incubated with precipitating antibody overnight. After the addition of respective protein-A- or protein-G-beads, washing, and denaturation in Lämmli buffer, the immunocomplexes were separated by SDS/PAGE and subjected to Western blotting. See SI Materials and Methods for details.

Reagents.

Chemicals, reagents, and antibodies were purchased from Sigma Aldrich. KB-R7943 and GF109203X were from Tocris Biosciences. The TRPC3 pore blocker Pyr3 was synthesized as published (16).

Statistics.

Data are presented as mean values ± SEM and was tested for statistical significance by using the Student t test (*P < 0.05).