Hydrophobic interactions within the C terminus pole helices tunnel regulate calcium-dependent inactivation of TRPC3 in a calmodulin-dependent manner

Abstract

Recent structural studies have shown that the carboxyl-terminus of many TRP channels, including TRPC3, are folded into a horizontal rib helix that is connected to the vertical pole helix, which play roles in inter-structural interactions and multimerization. In a previous work we identified I807 located in the pole helix with a role in regulation of TRPC3 by STIM1 (Lee K.P. et. al., 2014, Liu H. et al, 2022). To further determine the role of the pole helix in TRPC3 function, here we identified key hydrophobic residues in the pole helix that form tight tunnel-like structure and used mutations to probe their role in TRPC3 regulation by Ca²⁺ and Calmodulin. Our findings suggest that the hydrophobic starch formed by the I807-L818 residues has several roles, it modulates gating of TRPC3 by Ca²⁺, affects channel selectivity and the channel Ca²⁺ permeability. Mutations of I807, I811, L814 and L818 all attenuated the Ca²⁺-dependent inactivation (CDI) of TRPC3, with I807 having the most prominent effect. The extent of modulation of the CDI depended on the degree of hydrophobicity of I807. Moreover, the TRPC3(I807S) mutant showed altered channel monovalent ion selectivity and increased Ca²⁺ permeability, without affecting the channel permeability to Mg²⁺ and Ba²⁺ and without changing the pore diameter. The CDI of TRPC3 was reduced by an inactive calmodulin mutant and by a pharmacological inhibitor of calmodulin, which was eliminated by the I807S mutation. Notably, deletion of STIM1 caused similar alteration of TRPC3 properties. Taken together, these findings reveal a role of the pole helix in CDI, in addition to its potential role in channel multimerization that required gating of TRPC3 by STIM1. Since all TRPC and most TRP channels have pole helix structures, our findings raise the possibility that the pole helix may have similar roles in all the TRP family.

Graphical abstract

1. Introduction

The super family of TRP channels functions as Ca²⁺ permeable cation channels that mediate numerous physiological functions [1–3]. The canonical TRP (TRPC) channels are a subfamily of TRP channels that in human consists of 6 members. TRPC channels are grouped into TRPC1/4/5 and TRPC3/6/7 subgroups based on their homology [3]. All TRPC channels are activated in response to stimulation of Gq/11 protein coupled receptors and mediate significant portion of the receptor mediated Ca²⁺ influx [3]. Excessive Ca²⁺ influx is highly toxic to cells and an important mechanism guarding against Ca²⁺ toxicity is inhibition of TRP channels by Ca²⁺ itself, including TRPC channels [3]. Therefore, understanding regulation of TRPC channels by Ca²⁺ is of paramount physiological significance. In the present work we selected the canonical transient receptor potential channel 3 (TRPC3) as a model to better understand regulation of TRP channels by Ca²⁺. TRPC3 is expressed and plays key roles in the nervous, endothelial, smooth muscle and secretory gland tissues [4–8]. Change in TRPC3 functional expression or mutations that modify gating of TRPC3 are associated with various diseases. For example, TRPC3 over-expression has been reported in human ovarian adenocarcinoma tissue [9] and overactivity of TRPC3 in moonwalker mice result in abnormal Purkinje cell development and cerebellar ataxia [10]. In secretory epithelia over activation of TRPC3 contribute to acute pancreatitis [7, 8].

TRPC3 is activated by receptor stimulation and is regulated by STIM1 in response to ER Ca²⁺ store depletion. Receptor-mediated activation starts by hydrolysis of PI(4,5)P₂ to generate diacylglycerol (DAG), which directly activates TRPC3 [11], likely by interacting with lipid site 2 in the Cryo-EM structure of TRPC3 [12–14], in which DAG was identified as the lipid in the structures of TRPC5 [15, 16] and TRPC6 [17]. DAG interaction with lipid 2 site also induce a sensitized channel state [18]. In addition to activation by DAG, TRPC3 requires activation by STIM1 in response to store depletion [19, 20]. STIM1 appears to directly activate TRPC3 by interacting with the channel [20] and by targeting TRPC3 to PI(4,5)P₂-rich domains at the ER/PM junctions [21]. Overactivation of TRPC3 is prevented by the cytosolic Ca²⁺-binding protein, calmodulin (CaM). When bound to Ca²⁺, CaM competes with IP₃ receptors (IP₃Rs) for binding to an overlapping region on TRPC3 to inactivate TRPC3 currents [22]. This region was termed the CaM/IP₃Rs binding (CIRB) domain, which extends from residue 761 to 785, spanning a flexible region and the horizontal helix coiled-coil domain (rib helix) of TRPC3 [23]. In this manner, CaM can provide a negative feedback mechanism to control cytoplasmic Ca²⁺ levels ([Ca²⁺]_i) by inactivating TRPC3 when bound to Ca²⁺ entering the cell via TRPC3 or released from the ER. However, the molecular mechanism and structural requirements by which CaM inhibits TRPC3 has not yet been fully elucidated.

TRPC3 has six transmembrane domains, of which 5 and 6 form the pore region and 1–4 resembles a voltage sensor domain and assemble into a functional tetrameric ion channel [12, 14]. The Cryo-EM structures show estimated pore diameter of <5.8 Å with close state of <1.0 Å [12–14, 24]. The cytoplasmic N-terminus (NT) consists of four pairs of ankyrin repeat domains (ARDs) that are followed by helical linker domains. The carboxyl terminus (CT) consists of a TRP box, an unsolved disordered region and terminates with two helical coiled-coil domains (CCDs), a horizontal rib helix and a connected vertical pole helix. The flexible disordered region between the TRP box and the rib helix appears to participate in multiple interactions and may regulates the gating of TRPC3 [13, 24]. The rib and pole helices form multiple interactions with the NT ARDs, and the pole helix was suggested to play a role in subunits multimerization [13, 25]. The four pole helices interact to form a central tunnel under the cytoplasmic exit of the pore domain. The functional role of the tunnel is not known and was examined at the present work.

Early structures suggested that the acidic E630 acts as a selective Ca²⁺ filter [12, 13, 24]. The same authors recently reported Cryo-EM structures of TRPC3 and TRPC6 in the presence and in the absence (EGTA+EDTA) of Ca²⁺ that revealed compact structure of the pole helices in the presence of Ca²⁺ and loose structure in the absence of Ca²⁺ [14]. The compact structure contained three potential Ca²⁺ bindings sites (CBS), with CBS1 located between ankyrin repeat 2 and the pole helix and mutation of D798A at this site reduce inhibition of TRPC3 by intracellular Ca²⁺ ([Ca²⁺]_i). CBS2 was in a negatively charged region at the top of the pole helix, but mutation analysis suggested no role in regulation of TRPC3 by Ca²⁺. Finally, CBS3 was located within the voltage sensor-like domain that may function to activate rather than inhibit TRPC3 [14]. When bound to the CBS1 in each of the 4 subunits, the Ca²⁺ ions surround the tunnel form by the pole helices. Communication between CBS1 and the tunnel formed by the pole helices in channel gating and regulation by Ca²⁺ have not been thoroughly investigated. In previous studies we reported that TRPC3 NT and CT CCDs are involved in channel gating. The studies revealed that mutation of the hydrophobic I807 (I807S) that partially disrupts the pole helix is involved in the regulation of TRPC3 by STIM1 [20, 21] and participates in multimerization of TRPC3 with TRPC1 [26, 27]. In the present study we took advantage of these findings to examine the role of the hydrophobic tunnel formed by the pole helix to examine its role in TRPC3 gating by Ca²⁺, likely through influencing the function of CBS1.

2. Materials and Method

2.1. Cell culture and transfection

Human embryonic kidney 293 (HEK293) cells and STIM1^−/− HEK293 cells were cultured in 12-well cell culture plates at 37 °C and 5% CO₂ in Dulbecco’s modified Eagle medium (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with antibiotic-antimycotic reagent (Life Technologies, Grand Island, NY, USA) and 10% fetal bovine serum (Gibco). Cells were transiently transfected using Lipofectamine 2000 Transfection Reagent (Life Technologies), as recommended by the manufacturer.

2.2. Constructs and mutagenesis

The human TRPC3 transcript variant 2 (NM_003305) cDNA, tagged with yellow fluorescent protein, was inserted into pcDNA3.1 (pcDNA3.1-hTRPC3-YFP) as detailed before [20, 21]. The mutagenic primers were designed to contain unmodified nucleotide sequences flanking the desired mutation site. The primers used to generate the various mutants are listed in supplementary table 1. PCR was performed using the BioFACT^™ Lamp Pfu DNA Polymerase kit (Biofact, Daejeon, South Korea) with a Bio-Rad T100 Thermocycler (Bio-Rad Laboratories, Hercules, CA, USA), according to the protocol provided by the manufacturer.

2.3. Current measurement

The patch pipettes were pulled from thin-wall filament glass capillaries GC 150TF-7.5 (Harvard Apparatus, Holliston, MA, USA) to a resistance of 3–4 MΩ using a vertical pipette puller PC-10 (Narishige Group Products, Tokyo, Japan). An inverted microscope (Nikon ECLIPSE Ti, Tokyo, Japan) was used to identify positively transfected cells based on their yellow fluorescence when illuminated at 514 nm. Whole-cell voltage-clamp experiments were performed at room temperature using an Axopatch 200 B capacitor-feedback patch-clamp amplifier connected to a Digidata-1440A Digitizer (Molecular Devices). Currents were elicited by applying linear voltage-ramp protocols ranging from −100 to +100 mV from a holding potential at −60 mV. The rectification score was determined from the relative change in the inward current at −80mV compared to the outward current at +80mV. The percentage current remaining after exposure of the cells to at 2 mM external Ca²⁺ for 30–60 seconds was used to estimate the inhibition by Ca²⁺, by calculating the fraction of current at −60mV remaining in the presence of 2 mM external calcium compared with the current in the fully activated state under divalent-free conditions (DVF).

2.4. Solutions and materials

sThe standard divalent free extracellular solution contained (in mM): NaCl 140, KCl 5, 4-(2-Hydroxyethyl)piperazine-1-ethanesulfonic a (HEPES) 10, and Ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA)0.5 (pH7.4 using NaOH). The extracellular solution used for response for external calcium contained (in mM): NaCl 140, KCl 5, HEPES 10, MgCl₂ 1 and CaCl₂ X (pH7.4 using NaOH). CaCl₂ concentration was adjusted as indicated in the Figures. External solution used for pore size measurements contained (in mM): HEPES 10, EDTA 10 and Cation Salt 150 (pH7.4 using HCl or NMDG). Cation salts used were NaCl, methylamine, dimethyl amine, trimethyl amine and (1-Deoxy-1-(methylamino)-D-glucitol, Meglumine) NMDG. External saline used for permeability measurements contained (in mM): XCl 150, MgCl2 1, HEPES 10 and Glucose 10 (7.4 pH with XOH) (X being Li, Na, Rb and Cs). All saline solutions were within 290–320 mosM. The standard intracellular solution contained (in mM): CsCl 140, MgCl₂ 3, ATP 1, CaCl₂ 1.5 (to clamp free Ca²⁺ at 70 nM), HEPES 10 and EGTA 5 titrated to pH 7.2 with CsOH unless specified. Calmidazolium chloride, carbamoylcholine chloride (CCh), methyl ammonium, dimethyl ammonium, trimethylammonium, tetramethylammonium, N-methyl-d-glucamine, and 1-oleoyl-2-acetyl-sn-glycerol were purchased from Sigma–Aldrich (Seoul, Korea).

2.5. Structural analysis

The Marcoil 1.0 coiled-coil software predicted the CCD based on a hidden Markov model [28]. The first CCD was predicted to span between residues E240 and E284, and the second CCD spans between Q777 and L836 with an initial coiled-coil probability of 11.9% at K789. The probability increases along with the predicted CCD domain, with the peak probability values (p) greater than 89% as calculated from L804 to the end of the protein sequence. The hydrophobic line in the pole helix in the TRPC3 structure (PDB; 6CUD) was visualized by PyMOL 2.2 (The PyMOL Molecular Graphics System)

2.6. Statistical Analysis

Statistical significance of the data was calculated using one-way analysis of variance with Bonferroni correction (Origin Pro 8.1, Northampton, Massachusetts, USA) when comparing three or more dataset groups, and the unpaired two-tailed Student’s t-test was used when comparing two datasets. The linear fit, Y = mX + c, was used to fit the data for a linear relationship. Data are expressed as mean ± SEM; statistical significance is shown as *p < 0.05, **p < 0.01, and ***p < 0.001.

3. Results

3.1. Disrupting hydrophobic interactions within the pole helix tunnel hinders Ca²⁺ dependent inactivation of TRPC3

Previous studies on the regulation of TRPC3 by Ca²⁺ used activation of TRPC3 by OAG [29] and the measurement of Ca²⁺-dependent inactivation (CDI) associated with the structural studies relied on spontaneous activity of TRPC3 [14]. To examine the physiological role of the regulation of TRPC3 by Ca²⁺, we activated TRPC3 by stimulation of the muscarinic M3 receptors. To demonstrate regulation of TRPC3 by Ca²⁺, HEK cells expressing TRPC3 and the M3Rs were stimulated with 100 μM CCh while the cells were perfused with divalent ions free solution (Fig. 1). The stimulation resulted in a large inward Na⁺ current that was promptly and markedly inhibited upon addition of 2 mM Ca²⁺ to the bath perfusate (Fig. 1A). This form of CDI was proposed to result from interaction of cytoplasmic Ca²⁺ ([Ca²⁺]_i) with Ca²⁺ binding site 1 (CBS1) of the three TRPC3 CBS, based on measurement of spontaneous activity of TRPC3 [14]. We extended these findings to show similar CDI for the receptor-activated TRPC3 (n = 8). Disrupting the pole helix by mutating the hydrophilic amino acid I807 (I807S) modified the inhibition by Ca²⁺, causing rapid transient increase in the current and significantly reduced inactivation rate and extent. Interestingly, buffering the cytosol with 5 mM EGTA and even 10 mM BAPTA failed to prevent the CDI (small residual current after addition of 2 mM external Ca²⁺) of TRPC3, although BAPTA further reduced CDI of I807S, suggesting that even the I807S mutants retained some CDI that requires Ca²⁺ at the mouth of the channel (Fig. 1B).

Fig 1. Hydrophobic stretch of the pole helix controls Ca²⁺-dependent inactivation of TRPC3.

Panel (A): HEK cells transfected with TRPC3 (black trace) or TRPC3(I807S) (green trace) in divalent free (DVF) solution were stimulated with 100 μM carbachol (CCH). When maximum current was attained, the cells were exposed to media containing 2 mM Ca²⁺. (B): The percentage of residual current after 10 seconds exposure to external Ca²⁺ was determined in pipette solutions containing 5 mM EGTA or 10 mM BAPTA. (C): Representative IV curves recorded with DVF media (left) and in the presence of 2 mM Ca²⁺ (right). (D): shown is the channel rectification score determined as the ratio of current at −80mV/+80mV for TRPC3 (black) and TRPC3(I807S) (green). (E): Structure of the human TRPC3 pole helix (PDB 6CUD) visualized with hydrophobic side chains in the pole helix interface between two subunits. (F): Summary of residual current after exposure to 2 mM external Ca²⁺ of the indicated TRPC3 pole helix mutants. (G): Summary of rectification score of the indicated the same TRPC3 mutants as in F. (H): Correlation of hydrophobicity with residual current measured with the indicated I807 mutants (I): Correlation of hydrophobicity with current rectification measured with the indicated I807 mutants. (J): Representative IV curves of the indicated I807 mutants. Results are presented as mean ± SEM with ∗p < 0.05 and ∗∗∗p < 0.001.

Summary of the current density of TRPC3 and TRPC3(I807S) under the various conditions (Supplementary Fig. S1) showed that external Ca²⁺ and the I807S mutation mainly affected the inward current, with minimal effect on the outward current. A trend of small reduction in outward current was the same for TRPC3 and TRPC3(I807S), but a did not reach statistical significance. However, the I897S mutation affected channel rectification defined as the ratio of the currents at −80mV/+80mV, and the shape of the IV curve only when Ca²⁺ was present in the extracellular perfusate (Fig. 1C). Notably, the rectification score was 0.36 ± 0.03 for TRPC3 and 0.63 ± 0.046 (p < 0.001) for TRPC3(I807S) in the presence of 2 mM extracellular Ca²⁺ (Fig 1D). Buffering [Ca²⁺]_i had no effect on TRPC3 and TRPC3(I807S) rectification or the increased rectification by TRPC3(I807S) (Fig. 1C), suggesting that the rectification may involve one of the additional CBS that mediates inhibition of TRPC3 by Ca²⁺ that cannot be accessed by cytoplasmic BAPTA.

I807 is located at the top of the pole helix and is part of a hydrophobic tunnel at the bottom of the pore presumed ionic exit site [14]. To analyze the role of the pole helix in gating of TRPC3 by Ca²⁺, first we truncated the residues below I807. Truncation of the pole helix below L804 resulted in impaired TRPC3 current (supplementary Fig. S2). In fact, truncation of only the last 8 residues of the pole helix was sufficient to inhibit channel conductance [21], highlighting the importance of the pole helix to channel function. In another approach, we substituted the hydrophobic amino acids at the pole helix with hydrophilic amino acids that are predicted to disrupt the helical structure. The region at the pole helix (red) in which the hydrophobic residues (green) in each subunit appose each other is shown in Fig. 1E. Mutations of residues upstream of I807 had minimal effect on CDI, as determined by the residual current after inhibition by 2 mM external Ca²⁺ (Fig. 1F). The same assay showed that CDI gradually decreased along the pole helix from its peak at I807 and down to L819 (Fig. 1F). Mutation of the same residues also showed the highest rectification score (Fig. 1G). Therefore, to further explore the importance of the pole helix hydrophobicity we substituted I807 with amino acids with a gradual change in hydrophobicity and compared their CDI. Fig. 1H shows linear relationship between the pole helix hydrophobicity and CDI and Fig. 1I shows that the rectification score of the I807 mutants followed a linear relationship with reduced hydrophobicity (R² = 0.997, with intercept 0.55 ± 0.006, slope – 0.048 ± 0.002, n = 7–12). Example I/Vs are shown in Fig. 1J.

3.2. TRPC3-mediated calcium current is required for calcium-depended inhibition

In principle, the CDI triggered by addition of extracellular Ca²⁺ can be caused either by interaction of Ca²⁺ with an extracellular site or by Ca²⁺ entering the cells through TRPC3. To distinguish between the two possibilities, we took advantage of the Ca²⁺-impermeable TRPC3(E630Q) mutant [30]. Figs. 2A and 2B show that TRPC3(E630Q) lacks CDI to an extent comparable to that of TRPC3(I807S), which is also apparent in the I/Vs in Figs. 2C and 2D. Similarly, the TRPC3(E630Q) mutation increased the rectification score to a level similar to that obtained for TRPC3(I807S) (Figs. 2E). To provide further evidence that the CDI of TRPC3 was due to inward Ca²⁺ current, we characterized the properties of a TRPC3(E630Q/I807S) double mutant. Analyzing the current within the first seconds of addition of external Ca²⁺ revealed the I807S mutation increased the relative Ca²⁺/Na⁺ permeability (calculated as in [31]) of both TRPC3 and TRPC3(E630Q) (Supplementary Fig. S3). Yet, TRPC3(E630Q/I807S) still lacked CDI and displayed rectification score comparable to TRPC3(I807S) (Figs. 2A-E). Finally, to study the specific role of Ca²⁺ in the CDI, we replaced extracellular Ca²⁺ with Ba²⁺. Addition of 2 mM external Ba²⁺ failed to inactivate TRPC3 and TRPC3(I807S) (time course in Fig. 2F and averages in Fig. 2G). The I/Vs in Fig. 2H for Ca²⁺ and 2I for Ba²⁺ and the rectification scores in Fig. 2J further show that CDI is specific for Ca²⁺ and cannot be mediated by Ba²⁺ that is often used as a Ca²⁺ surrogate with Ca²⁺ channels. When taken together these findings demonstrated that the inactivation of TRPC3 is specifically mediated by Ca²⁺ influx through TRPC3. However, the question remains as to whether the lack of CDI by the pole helix TRPC3 mutants is due to either the lack of Ca²⁺ permeability or altered sensitivity of the mutants to Ca²⁺. This is addressed below.

Fig 2. Ca²⁺ influx through TRPC3 is required for Ca²⁺-dependent inactivation of TRPC3.

Panel (A): Time course of the inward current measured with TRPC3 (black), TRPC3(I807S) (green), the Ca²⁺ impermeable TRPC3(E630Q) (red) and the double mutant TRPC3(E630Q/I807S) (blue). Note restoration of Ca²⁺ influx to TRPC3(E630Q) by the I807S mutation (green and blue traces and I/V in D). (B): Residual current after exposure to 2 mM external Ca²⁺. (C, D): Example IV curves recorder in DVF solution (C) and in the presence of 2 mM Ca²⁺ (D). (E): rectification score of the indicated mutants. (F): Time course of TRPC3 and TRPC3(I807S) currents measured in DVF solution. Where indicated the cells were exposed to 2 mM Ca²⁺ (top traces) or 2 mM Ba²⁺ (lower traces). (G): Summary of experiments as in F showing the effect of Ca²⁺ and Ba²⁺ on the residual current. (H, I) Example I/Vs measured with Ca²⁺ (H) and Ba²⁺ (I). (J): The rectification score measured in DVF solutions and solutions containing 2 mM Ca²⁺ or 2 mM Ba²⁺ measured with TRPC3 and TRPC3(I807S). Results are presented as mean ± SEM. ∗p < 0.05, ∗∗ p < 0.01, and ∗∗∗p < 0.001.

3.3. Mutations in the pole helix increase the Ca²⁺ permeability of TRPC3.

To address the question if TRPC3(I807S) mediates Ca²⁺ influx, first we measured Ca²⁺ influx by selectively stimulating TRPC3 and TRPC3(I807S) with the channel ligand GSK that activates the channel like DAG [32, 33]. Figs. 3A and 3B show that 10 μM GSK increased [Ca²⁺]_i in cells transfected with TRPC3 and TRPC3(I807S) and TRPC3(I807S) was significantly more permeable to Ca²⁺ than TRPC3, suggesting that the reduced TRPC3(I807S) CDI was not due to reduced Ca²⁺ influx.

Fig 3. Ca²⁺ influx by TRPC3(I807S) and effect of the I807 and E630 mutants on TRPC3 selectivity and pore size.

Panel (A, B): Ca²⁺ influx was measured in HEK cells transfected with TRPC3 (black) or TRPC3(I807S) (red) and stimulated with 10 μM GSK in Ca²⁺ containing solution and then perfused with Ca²⁺ free solution (A). The peak currents are shown in (B). (C): The reversal potentials were measured in solution when the only permeable extracellular ions were Na⁺, Ca²⁺, Mg²⁺ or NMDG with the indicated TRPC3 variants. (D): Magnified portion of the I/V curves measured with the indicated mutants and external ions. (E): TRPC3 and TRPC3(I807S) currents were measured with the indicated methylated ammonium ions. The results are presented as mean ± SEM. ∗p < 0.05, ∗∗ p < 0.01, and ∗∗∗p < 0.001

Second, we determined the selectivity and ionic permeability of TRPC3 and the mutants and how they affect divalent ions permeability. The divalent ion permeability of TRPC3, TRPC3(I807S), TRPC3(E630Q) and TRPC3(E630Q/I807S) in the presence Na⁺, Ca²⁺ and Mg²⁺ as the only permeable cations are shown in Figs. 3C and 3D. As we reported recently [21], TRPC3(I807S) showed reduced Na permeability that was maintained by TRPC3(E630Q/I807S), as shown by the more negative reversal potentials. On the other hand, TRPC3(I807S) had higher Ca²⁺ permeability than TRPC3 and prominently increased the Ca²⁺ permeability of the Ca²⁺ impermeable TRPC3(E630Q) mutant. The reversal potentials measured by addition of external Mg²⁺ and by replacing external Na⁺ with NMDG shifted to a more negative value, and was the same for all constructs, indicating that all TRPC3 variants showed low permeability for Mg²⁺ and NMDG. Hence, the finding that the I807S mutation markedly shifted the reversal potential of TRPC3 towards a positive potential when Ca²⁺ was the only permeable external ion, indicate that disruption of the pole helix increased TRPC3 Ca²⁺ permeability, consistent with the increased [Ca²⁺]_i when TRPC3(I807S) is activated by GSK (Fig. 3A, B).

The surprising finding of increased Ca²⁺ permeability due to disruption of hydrophobic interactions within the pole helix led us to investigate whether the mutants alter the TRPC3 pore diameter. To this end, we compared the permeation of TRPC3 and TRPC3(I807S) to methylated ammonium ions of increased size, mono, di, tri, and tetra methyl ammonium. Both TRPC3 and TRPC3(I807S) displayed similar reversal potentials to the methyl ammonium ions (Figs. 3E and 3F), indicating that the increased TRPC3(I807S) Ca²⁺ permeability is not due to increased pore diameter. Together, the Ca²⁺ influx and current analysis suggest that the loss of CDI phenotype of the pole helix mutants (Figs. 1) is due to disruption of Ca²⁺ binding by TRPC3. This conclusion is in agreement with a recent report that the absence of Ca²⁺ and mutations in CBS1 that prevent CDI disrupt the compactness of the pole helix [14].

3.4. Ca/CaM contributes to CDI of TRPC3

The TRPC3 rib helix forms part of the CIRB domain that was shown to interact with calmodulin (CaM) to inactivate TRPC3 [22, 34]. Recent structural analysis of drosophila TRP channel C terminal domains identified similar CaM binding site at the rib helix and a site at the pole helix [35]. The TRPC6 CIRB domain, a close relative of TRPC3, interacts with CaM in a Ca²⁺-dependent manner to affect TRPC6 activity [36]. It is therefore possible that binding of CaM to the CIRB domain and rib helix affects the structure of the pole helix to regulate CDI. To test this hypothesis, we evaluated the effect of a CaM inhibitor calmidazolium chloride (CMZ) and of the dominant negative CaM EF hand mutant (CaM_EF) on TRPC3 function. Figs. 4A-E show the effect of CaM and of the CaM_EF mutant and Figs. 4F-J show the effect of CMZ on TRPC3 function. Transfection of CaM had no effect the CDI of TRPC3 or TRPC3(I807S), the time course of activation, residual currents, I/V or rectification scores. This is likely because TRPC3 is saturated with endogenous CaM, which in HEK cells can range from 6–10 μM [37]. By contract, transfection of the CaM_EF mutant increased the residual current of TRPC3 after exposure to 2 mM external Ca²⁺ while CaM_EF had no effect on the residual current of TRPC3(I807S), suggesting that TRPC3(I807S) has reduced sensitive to or lost interaction with Ca/CaM (Fig. 4B). In line with the CaM_EF findings, Figs. 4F-J show that incubating the cells with 1 μM CMZ hindered the CDI of TRPC3 to a level measured with TRPC3(I807S) by increasing the residual TRPC3 current, while having no effect on the residual current of TRPC3(I807S) (Fig. 4G). Similarly, CMZ increased TRPC3 rectification without affecting the rectification of TRPC3(I807S) (Fig. 4J). These findings suggest functional coupling between the pole helix and the CIRB domain that affect Ca²⁺ regulation of TRPC3.

Fig 4. Effect of Calmodulin, the Calmodulin EF hand mutant and calmidazolium on Ca²⁺-dependent inactivation of TRPC3 and TRPC3(I807S) currents.

Panel (A): Time course of TRPC3 (black traces) and TRPC3(I807S) (green traces) expressed alone or together with calmodulin (CaM) or the calmodulin EF mutant (CaMEF) measured in DVF solution and where indicated exposed to 2 mM Ca²⁺. (B): Summary of residual current after exposure to 2 mM external Ca²⁺ measured as in A. (C, D): Example I/Vs for TRPC3 (C) and TRPC3(I807S) (D). (E): Rectification score for the indicated conditions. (F): Time course of TRPC3 and TRPC3(I807S) current in control cells and cells treated with 1 μM calmidazolium for 5 min and stimulated by 100 μM CCh. (G): Summary of residual current after exposure to 2 mM external Ca²⁺ measured as in F. (H, I): Example I/Vs for TRPC3 (H) and TRPC3(I807S) (I). (J): Rectification score for the indicated conditions. Results are presented as mean ± SEM. ∗p < 0.05, ∗∗p < 0.01, and ∗∗∗p < 0.001

3.5. Localization at the STIM1-formed ER/PM junction is required for CDI

TRPC channels, including TRPC3, interacts with and are regulated by STIM1 [19, 20, 38]. Localization of TRPC3 in STIM1-formed ER/PM junction is required for regulation of TRPC3 by lipids and modulates TRPC3 ionic selectivity [21]. Therefore, we asked whether CDI of TRPC3 is affected by STIM1. This was tested by expressing TRPC3 and TRPC3(I807S) in STIM1^−/− cells and measuring the effect of external Ca²⁺ on their activity (Fig. 5). To be able to compare TRPC3 behavior in the presence and absence of STIM1, the cells were maximally stimulated with 100 μM CCh since STIM1 affect the apparent Michaelis-Menton constant (Km) but not the maximal velocity (Vmax) for stimulation with CCh [21]. The behavior of TRPC3 and TRPC3(I807S) were similar in STIM1^−/− cells, with the addition of 2 mM external Ca²⁺ causing an initial increase in the current (Fig. 5A), although the increased current was lower with TRPC3 (Fig. 5B) and prominently and similarly reduced CDI (Fig. 5C). At present, the reason for the reduced increased current with TRPC3 compared to TRPC3(I807S) upon addition of external Ca²⁺ is not known. Among the possibilities can be faster TRPC3 inactivation of better access of Ca²⁺ to TRPC3 CBS1.

Fig 5. Localization in the STIM1-formed junctions in required for CDI.

TRPC3 (black) and TRPC3(I807S) (green) were transfected in STIM1^−/− cells and current was measured in response to stimulation with 100 μM CCh in DFM and in the presence of 2 mM external Ca²⁺, as indicated. Panel (A) is the time course, Panel (B) the maximal increase increased current in pA/pF measured on addition of Ca²⁺, panel (C) is the residual current after exposure to 2 mM external Ca²⁺ and panel (D) are example I/V.

4. Discussion

Excess [Ca²⁺]_i is highly toxic and many cellular mechanisms have been evolved to ensure that [Ca²⁺]_i will be maintained within physiologically beneficial levels. One of the key mechanism involves the acute limit of the period of the [Ca²⁺]_i increase. This is achieved by Ca²⁺ clearance by the SERCA and PMCA pumps that are Ca²⁺-activated transporters [39]. However, most of the Ca²⁺ toxicity is due to excessive Ca²⁺ influx from the virtually unlimited source of extracellular Ca²⁺ that enters the cells through Ca²⁺ permeable channels [40]. An effective mechanism in controlling Ca²⁺ influx channels is their inhibition by Ca²⁺. This is well documented for voltage gated Ca²⁺ channels [41], the Orai channels [42] and the TRP channels [3]. However, the molecular details of the inhibition and its regulation by channel domains are not fully understood. As for the TRPC channels, a major strid was made recently by obtaining the structure of TRPC3 and TRPC6 in the presence and absence of Ca²⁺ [14]. The Ca²⁺ bound structure is compact and with 3 Ca²⁺ ions bound to three separate sites CBS1–3. However, binding of Ca²⁺ to CBS1 appears to be the main site regulating inhibition of TRPC3 by Ca²⁺ [14].

A major issue not addressed before is how binding of Ca²⁺ to TRPC channels CBS1 is modulated by the channel domains and by localization of the channels at the ER/PM junctions. A unique structure found in all TRP channels is the connected pole and rib helices [43]. The pole helices of the four protomers are assembled into a tunnel just under the presumed cytoplasmic ion exit site. The inner portion of the protomers interact with each other, as shown in Figure 1, to form a hydrophobic domain close to CBS1 while the rib helices together with an unstructured domain form the site at which IP₃ receptors and CaM compete for interaction. This compact rib-pole structure is held together by Ca²⁺ binding to TRPC3 and is disrupted when Ca²⁺ is removed from all three sites [14], although the contribution of each of the three CBSs to the structure is not known at present. In the present studies, we examined the role of the pole helix and its hydrophobic nature in inhibition of TRPC3 by Ca²⁺. TRPC3 is directly activated by DAG and is regulated by STIM1 [23, 26] and by STIM1-dependent localization at the ER/PM junctions [21]. In previous studies we have shown that the carboxyl terminus CCD that includes the pole helix is required for interaction of TRPC3 with TRPC1 and subsequent activation by STIM1 [26]. In the present studies we extended these findings to show that pole helix modulates the CDI of TRPC3 and regulate channel ion selectivity and gating. Moreover, regulation by CDI and by CaM requires localization of TRPC3 in the STIM1 formed ER/PM junctions.

The pole helix has an essential role in TRPC3 function since deletion of residues below L804 (Fig. S2) or even the last 8 residues [21] resulted in inactive channel. Moreover, the pole helix controls channel gating and selectivity. It limits the Ca²⁺ permeability of TRPC3 and confer higher Na⁺/Ca²⁺ selectivity (Figs. 3, 5). The effect of the pole helix of channel selectivity is independent of Ca²⁺ influx, and thus intracellular Ca²⁺, since it takes place, and it is similar with Ca²⁺ permeable and impermeable TRPC3 (Fig. S3). Finally, the pole helix is essential for a robust CDI of TRPC3, as shown by the marked reduced CDI upon disruption of pole helix hydrophobicity (Fig. 1). Together, these effects limit Ca²⁺ influx into cells with activated TRPC3 to guard against Ca²⁺ toxicity.

The CDI of TRPC3 is mediated by Ca²⁺ binding to CBS1 [14]. The present findings show that CDI is modulated by the pole helix and by Ca²⁺/CaM. A surprising finding is that even 10 mM cytoplasmic BAPTA could not prevent the strong CDI, which suggests close vicinity of CBS1 to the channel pore and tight binding of Ca²⁺ to CBS1. Nevertheless, CDI appears to be modulated by Ca²⁺/CaM that was shown to interact with the TRPC3 [22, 34] and TRPC6 [36] CIRB domains (Fig. 4) located in the unstructured loop between the rib helix and TRP box. How CaM exert its effect needs further examination. Measurement of TRPC3 current in inside-out patches showed that the current was inhibited by Ca²⁺ and the inhibition was not affected by CaM or the CaM mutant [14]. It is possible that regulation by CaM was lost when the patches were isolated, or the regulation required cell stimulation as was used in the present studies. Indeed, when rhabdomeric drosophila TRP was activated by light, CaM affected the current measured at low external Ca²⁺, although not during large Ca²⁺ influx [35]. CaM interacted with two sites in the C terminus of drosophila TRP, with the CaM N lob bound to an upstream site and the CaM C lob bound to a site formed partially by the pole helix. The site at the pole helix is highly conserved among the drosophila and mammalian TRPs and includes the hydrophobic stretch examined in the present studies [35]. Indeed, the I807S mutation that disrupts this site also eliminated regulation by CaM (Figure 4). The somewhat variable observation on regulation of TRP channels by CaM that appears to depend on the channel isoform studied and the method used and thus required further analysis.

The hydrophobic interface within the pole helix has several roles in channel function. Disruption of the hydrophobicity even by mutation of a single residue, I807, change TRPC3 ionic selectivity (Fig. S3), Ca²⁺ permeability (Fig. 3), rectification (Fig. 1C, D) and inhibition by Ca²⁺ that requires Ca²⁺ binding to CBS1 (Fig. 1A, B). These findings indicate that the hydrophobic interface within the pole helix is required for the pole helix to provide an anchor that allows the transduction of conformational movement of the rib helix and the disordered flexible region to the channel pore. The flexible disordered region appears important for channel function and likely for polymodal regulation of TRPC3 by intracellular signals and transfer of conformational changes. Accordingly, truncations and extension of the flexible loop by introduction of glycines increased and decreased channel function when activated by GSK [24]. Clearly additional structural and functional studies are required to fully understand the conformational effect of the pole and rib helices in channel gating. The present findings highlight the multiple roles of these structures in TRPC3 function. The presence of similar rib and pole helices that appears to exist in different conformation suggest that our findings are relevant to other members of the TRP channels family.

Regulation of TRPC3 by lipids, channel selectivity [21] and CDI (Fig. 5), all required localization of TRPC3 in STIM1-formed ER/PM junctions. Similar regulatory mechanism by Ca²⁺ at the STIM1-formed ER/PM junction may be common to all TRP channels that contain pole-rib helices structure to limit Ca²⁺ influx through the channels. Interestingly, Ca²⁺-dependent inhibition of Orai1 that is mediated by the SARAF protein also required localization at the ER/PM junctions [44]. Ca²⁺-dependent inactivation of voltage-gated Ca²⁺ channel take place in junctional domains [45]. Thus, disruption of the junctions should lead to uncontrolled Ca²⁺ influx and Ca²⁺ toxicity. It should be informative to explore further the loss of junctional control of CDI as part of the pathology of cellular Ca²⁺ toxicity and cell death.

Highlights.

• Hydrophobic interaction in the pole helix in TRPC3 is important for TRPC channel function.

• Mutation interrupting Pole helix hydrophobic interaction attenuates calcium dependent inactivation.

• TRPC3 has calcium dependent feedback regulation either by direct biding of calcium to CBSs and Calcium/Calmodulin dependent regulation to CIRB.

Abbreviations

ARDs

ankyrin repeat domains

BAPTA

1,2-bis(2-aminophenoxy)ethane-N,N,N,N-tetraacetic acid

CaM

calmodulin

CIRB

CaM/IP3R binding

CCD

coiled-coil domain

CMZ

calmidazolium chloride

CT

carboxyl terminus

DAG

diacylglycerol

ER

endoplasmic reticulum

HEK293

human embryonic kidney 293

HH

horizontal helix

NT

N-terminal

STIM1

stromal interaction molecule 1

TRPC

canonical transient receptor potential channel

VH

vertical helix