ABSTRACT
Muscarinic receptor stimulation results in activation of nonselective cation (NSC) channels in guinea pig adrenal medullary (AM) cells. The biophysical and pharmacological properties of the NSC channel suggest the involvement of heteromeric channels of TRPC1 with TRPC4 or TRPC5. This possibility was explored in PC12 cells and guinea pig AM cells. Proximity ligation assay (PLA) revealed that when exogenously expressed in PC12 cells, TRPC1 forms a heteromeric channel with TRPC4, but not with TRPC5, in a STIM1-dependent manner. The heteromeric TRPC1–TRPC4 channel was also observed in AM cells and trafficked to the cell periphery in response to muscarine stimulation. To explore whether heteromeric channels are inserted into the cell membrane, tags were attached to the extracellular domains of TRPC1 and TRPC4. PLA products developed between the tags in cells stimulated by muscarine, but not in resting cells, indicating that muscarinic stimulation results in the membrane insertion of channels. This membrane insertion required expression of full-length STIM1. We conclude that muscarinic receptor stimulation results in the insertion of heteromeric TRPC1–TRPC4 channels into the cell membrane in PC12 cells and guinea pig AM cells.
INTRODUCTION
Acetylcholine released from the sympathetic preganglionic nerve fiber has been shown to mediate neuronal transmission through binding to the families of muscarinic acetylcholine receptors (mAChR) and nicotinic acetylcholine receptors (nAChR) in guinea pig adrenal medullary (AM) cells (Inoue et al., 2012). In addition, muscarinic receptor stimulation in AM cells of various mammals results in catecholamine secretion (Olivos and Artalejo, 2008; Inoue et al., 2018). The ionic mechanisms for muscarinic receptor-mediated excitation in AM cells differ among species of mammals, and their details have not yet been sufficiently elucidated (Inoue et al., 2018). In rat AM cells, muscarinic M1 receptor (also known as Chrm1) stimulation (Harada et al., 2015) induces the endocytosis of TASK1 (also known as Kcnk3) channels with a consequent decrease in K+ channel activity (Inoue et al., 2008; Matsuoka and Inoue, 2017; Inoue et al., 2019b), whereas in guinea pig AM cells the muscarinic receptor-mediated excitation is ascribed to not only TASK1 channel inhibition, but also to nonselective cation (NSC) channel activation (Inoue and Kuriyama, 1991; Inoue et al., 2012). This muscarinic receptor-regulated NSC channel has reversal potential of 0 mV and exhibits a conductance decrease at membrane potentials below −50 mV: i.e. the current–voltage (I–V) curve shows a negative slope at membrane potentials below −50 mV (Inoue and Kuriyama, 1991; Inoue et al., 2012). Moreover, La3+ has a double effect of facilitation and inhibition on the muscarinic NSC channel (Inoue et al., 2012). These properties of the NSC channel indicate the possible involvement of heteromeric channels of TRPC1 with TRPC4 or TRPC5 (Strübing et al., 2001; Clapham, 2003; Jung et al., 2003; Semtner et al., 2007). Indeed, the expression of TRPC1, TRPC4, and TRPC5 in guinea pig AM cells have been confirmed at the protein level: TRPC1 and TRPC4 are mainly located in the cytoplasm, whereas TRPC5 is present in the vicinity of, or at, the cell membrane.
The role of STIM1 in the regulation of TRPC channel activity has been disputed (Yuan et al., 2007; DeHaven et al., 2009). Thus, whether or not STIM1 plays an essential role in muscarinic receptor-regulated NSC channel activation in guinea pig AM cells needs to be investigated. STIM1 is expressed in guinea pig AM cells (Inoue et al., 2012), whereas it is present in rat adrenal cortical cells, but not AM cells (Matsuoka et al., 2009). What is interesting in rat AM cells is that although several forms of TRPC channels are present, as detected at the protein and/or mRNA levels, muscarinic receptor stimulation does not result in an apparent activation of NSC channels (Inoue et al., 2008; Matsuoka et al., 2009).
The findings that in guinea pig AM cells, TRPC1 and TRPC4 are localized inside the cells whereas TRPC5 is at the cell periphery, suggest that TRPC1 may form a heteromeric channel with TRPC4, but not TRPC5. If that is the case, muscarinic receptor-mediated excitation in guinea pig AM cells could be ascribed to insertion of heteromeric TRPC1–TRPC4 channels into the cell membrane. Indeed, several isoforms of TRPC channel proteins endogenously or exogenously expressed, such as TRPC3 (Goel et al., 2007), TRPC5 (Bezzerides et al., 2004) and TRPC6 (Cayouette et al., 2004), have been demonstrated to be inserted into the cell membrane in response to G protein-coupled receptor (GPCR) or receptor tyrosine kinase stimulation. The present study aimed initially to investigate whether the TRPC1 isoform forms a heteromeric channel with TRPC4 or TRPC5 in guinea pig AM cells and PC12 cells, a cell line derived from rat AM cells (Greene and Tischler, 1976). Once this proved to be the case, we next aimed to elucidate the role of STIM1 in the formation of these heteromeric channels. Finally, we used immunocytochemical and functional analyses to explore whether muscarinic receptor stimulation facilitates insertion of heteromeric channels into the cell membrane.
RESULTS
Proximity ligation assay
Proximity ligation assay (PLA) was employed to elucidate whether TRPC1 and TRPC4 form a heteromeric channel in PC12 cells, an immortalized rat AM cell line (Greene and Tischler, 1976). This method depends on the specificity of antibodies used. Thus, to explore the specificities of anti-TRPC1 and anti-TRPC4 antibodies, TRPC1–GFP and TRPC4–GFP were exogenously expressed in PC12 cells with STIM1–myc, and PLA between TRPC1 and TRPC4 was carried out with a combination of either mouse anti-TRPC1 and rabbit anti-TRPC4 antibodies or rabbit anti-TRPC1 and mouse anti-TRPC4 antibodies. As shown in Fig. 1A, PLA products mainly developed in PC12 cells expressing exogenous proteins with either of the antibody combinations, indicating that these antibodies were selective and that exogenous TRPC1 and TRPC4 form a heteromeric channel. This notion was further examined with exogenous expression of TRPC1–GFP and untagged TRPC4, and a combination of rabbit anti-GFP and mouse anti-TRPC4 antibodies. PLA products again developed almost exclusively in PC12 cells expressing exogenous proteins (Fig. 1A). Next, whether TRPC1 selectively forms a complex with TRPC4 rather than with TRPC5 was examined in PC12 cells where untagged TRPC1 and STIM1–myc were expressed together with TRPC4–GFP or TRPC5–GFP, and PLA was carried out with a combination of mouse anti-TRPC1 and rabbit anti-GFP antibodies. As shown in Fig. 1B, TRPC4–GFP proteins were diffusely distributed in the cytoplasm, especially with high clustering near the nucleus, whereas TRPC5–GFP proteins were present both in the cytoplasm and at the cell periphery. Although part of TRPC5–GFP was diffusely present in the cytoplasm, PLA products conspicuously developed in cells expressing TRPC1 with TRPC4–GFP, but not with TRPC5–GFP (Fig. 1B,C). Lastly, heteromer formation was biochemically examined (Fig. 1D). TRPC1 was detected in immunoprecipitates obtained with anti-GFP antibody from lysates of PC12 cells expressing TRPC4–GFP, but not GFP alone.
To elucidate whether or not STIM1 is obligatory for G protein-coupled receptor-mediated activation of TRPC channels (Yuan et al., 2007; Dehaven et al., 2009), the extent of heteromer formation was compared between control and exogenous STIM1-expressing PC12 cells. As shown in Fig. 2A,B, the simultaneous expression of STIM1–myc with TRPC1 and TRPC4 resulted in enhancement of PLA reactions between TRPC1 and TRPC4. Levels of rhodamine-like fluorescence (representing PLA products) were plotted against levels of FITC-like fluorescence (reflecting expression levels of TRPC1–GFP and TRPC4–GFP) (Fig. 2C). Levels of detected PLA products increased alongside levels of GFP fusion proteins, and the rate of increase in PLA products in cells expressing STIM–myc was significantly larger than in cells without it (Fig. 2D). In contrast to the previously reported effect on TRPC1–TRPC3 heteromer formation in HEK293 cells (Yuan et al., 2007), muscarinic stimulation did not affect the levels of PLA products between TRPC1 and TRPC4 (Fig. 2C,D). Intriguingly, some PLA products were translocated to the cell periphery in response to muscarinic stimulation (Fig. 2B,E).
STIM1 domains
STIM1 comprises several functional domains (Fig. 3A) (Lewis, 2011; Soboloff et al., 2012). Thus, we examined which domains were responsible for the facilitation of heteromer formation with the simultaneous expression of epitope-tagged STIM1 mutants and TRPC–GFP proteins. The STIM1ΔCt mutant contained STIM1 N-terminal residues 1–234, encompassing the transmembrane domain but not the CAD (CRAC activating domain), P/S (proline/serine-rich domain), TRIP (EBI binding sequence) or K (polybasic domain). The STIMCT (C-terminal) mutant contained residues 235–685 of STIM1, while the STIM1_1-448 mutant contained the N-terminus, transmembrane domain and CAD domain. PLA reactions between TRPC1 and TRPC4 scarcely occurred in PC12 cells expressing STIM1ΔCt, whereas the level of PLA products in cells expressing STM1_1-448 did not differ from that in cells expressing wild-type STIM1 (Fig. 3B,C). Intriguingly, expression of STIM1CT did not reproduce the PLA enhancement seen in cells expressing STIM1 or STM1_1-448. STIM1CT lacks the initial 234 amino acids of STIM1, corresponding to the intralumenal region and the transmembrane domain. Because of the deficit of this N-terminal region, STIM1CT exhibited a diffuse distribution in the cytoplasm, whereas wild-type STIM1 and the other two mutants were distributed in a reticular pattern suggesting their localization in the endoplasmic reticulum (ER). These results suggest that STIM1 (Lewis, 2011) helps TRPC1 and TRPC4 to form a heteromeric channel in the ER. This notion was further examined in cells where STIM1 was exogenously expressed with either TRPC1–GFP or TRPC4–GFP. As shown in Fig. 3D,E, PLA reactions between GFP and STIM1 occurred irrespective of whether cells expressed TRPC1–GFP (n=20) or TRPC4–GFP (n=16). These results suggest that there is a common sequence in TRPC1 and TRPC4 that binds to STIM1 in the ER (Huang et al., 2006).
TRPC1, TRPC4 and TRPC5 have been shown to bind to the STIM–Orai activating region (SOAR; amino acids 344–442) of STIM1, which almost coincides with the CAD region (amino acids 342-448). Thus, whether the CAD region is also involved in heteromeric TRPC1–TRPC4 channel formation was explored (Fig. 4A,B). When TRPC1–GFP and TRPC4–GFP were co-expressed with STIM1 CAD region construct CFP–CAD, the levels of PLA products for the heteromeric channel formation tended to decrease, but this decrease was not statistically significant. However, when heteromeric channel formation was augmented by the exogenous expression of STIM1–myc, the simultaneous expression of CFP–CAD resulted in a significant suppression of the heteromeric channel formation. These results support our notion that STIM1 facilitate heteromeric TRPC1–TRPC4 channel formation in the ER through its CAD region.
Insertion of TRPC channels into cell membrane
The translocation of TRPC1–TRPC4 channels to the cell periphery in response to muscarinic receptor stimulation raises the possibility that heteromeric channels are inserted into the cell membrane upon muscarine stimulation. This possibility was examined by transfecting PC12 cells with plasmids encoding mutated TRPC1–GFP and TRPC4–GFP proteins, in which FLAG and HA tags, respectively, were inserted into the extracellular domains of these channel proteins. The mutant proteins were expressed in PC12 cells with the same efficiency as the naïve TRPC–GFP proteins. Anti-HA and anti-FLAG antibodies were employed for PLA reactions to elucidate heteromeric channel formation and its membrane insertion. When PC12 cells expressing GFP proteins were not permeabilized (Fig. 5A,F), PLA products developed at the cell periphery in some of the stimulated cells, but never in the control cells. These findings were consistently observed in all the five trials where stimulated and non-stimulated cells were simultaneously examined each time (sign test, P<0.05). In 74 stimulated cells, 59.5% exhibited PLA products at the cell periphery, whereas none of the 49 non-stimulated cells did. In contrast to non-permeabilized cells, similar levels of PLA products were observed in permeabilized stimulated (n=6) and non-stimulated cells (n=22; Fig. 5B). It is worth noting that at least some of the PLA products formed between the two different tags were trafficked to the cell periphery in response to muscarine stimulation (Fig. 5B).
Whether STIM1 is also involved in muscarine-induced trafficking of the heteromeric channel to the cell membrane was next explored in PC12 cells expressing mutated TRPC1–GFP and TRPC4–GFP proteins. As shown in Fig. 5D,F, PLA products between FLAG- and HA-tagged TRPC proteins did not develop at the cell periphery in response to muscarine stimulation in cells expressing STM1_1-448 (n=10). This STIM1 mutant does not contain several domains that are important for STIM1 to interact with membrane lipids (Lewis, 2011) or the cytoskeleton (Grigoriev et al., 2008). The result suggests that TRPC1–TRPC4 heteromeric channels are trafficked to the cell membrane through the C-terminus of STIM1.
Muscarinic receptor stimulation in PC12 cells is expected to produce an increase in intracellular calcium levels ([Ca2+]i) through mobilizing Ca2+ from Ca2+ store sites (Kim and Saffen, 2005; Ebihara et al., 2006). This Ca2+ mobilization might result in activation of store-operated Ca2+ entry (SOCE). Whether this process might be involved in membrane insertion of TRPC1–TRPC4 channels was investigated with the expression of a STIM1D76A mutant, which mimics Ca2+ depletion in Ca2+ store sites (Liou et al., 2005). As expected, the simultaneous expression of the STIM1 mutant with the HA- and FLAG-tagged TRPC channels in unstimulated cells resulted in translocation of PLA products to the cell periphery (n=8; Fig. 5E); however, it did not lead to insertion of the channels (n=54; Fig. 5C). Interestingly, muscarine-induced insertion of channels was significantly facilitated in cells expressing STIM1D76A: 89.5% of GFP-positive cells (n=38) exhibited a PLA reaction at the cell surface and the mean±s.e.m. of PLA products in such cells was also increased to 1.18±0.12, compared to 0.85±0.10 in wild-type STIM1-expressing cells (Fig. 5F).
Insertion of the STIM1 and TRPC4 complex into the cell membrane
The findings that STIM1 forms a complex with TRPC1 and TRPC4 raise the possibility that STIM1 is also trafficked to the cell membrane with the TRPC1–TRPC4 heteromeric channel. This possibility was examined with a combination of anti-STIM1 and anti-HA antibodies. If the complex of STIM1 and TRPC4–HA–GFP is inserted into the cell membrane, the extracellular domains of TRPC4 and the intralumenal domain of STIM1, where an epitope for the antibody is located, should be exposed to the extracellular space. We found that PLA products between STIM1 and TRPC4–HA–GFP developed exclusively in PC12 cells stimulated by muscarine but not under basal conditions (Fig. 5G,H).
Functional analysis of membrane insertion
The trafficking of TRPC1–TRPC4 heteromeric channels to the cell membrane in response to muscarine stimulation was functionally examined with a Ca2+ indicator. After PC12 cells were transfected with plasmids encoding TRPC1–GFP, TRPC4–GFP and STIM1–myc, they were loaded with Fura-2 to measure changes in [Ca2+]i. As shown in Fig. 6A,B, the increase in peak amplitudes of Ca2+ induced by treatment with 30 µM muscarine in PC12 cells expressing exogenous STIM1 and TRPC–GFP proteins did not differ from that in control cells, whereas sustained Ca2+ levels in the former were significantly larger than those in the latter (Fig. 6A,C). In addition, the resting Fura-2 ratio was 0.963±0.021 (n=15) in cells expressing TRPC–GFP proteins and STIM1, which did not differ significantly from 0.899±0.024 (n=17) in control cells. These results suggest that exogenously expressed TRPC1–TRPC4 channels have no channel activity in resting conditions and become active upon muscarinic receptor stimulation. This notion was further explored with ML204, which is a specific inhibitor of TRPC4 and has no action on voltage-dependent Ca2+ channels (Miller et al., 2011). We have previously reported that in guinea pig AM cells muscarine failed to activate NSC channels in the presence of 10 µM ML204 (Inoue et al., 2019a). The sustained levels of muscarine-induced Ca2+ increases were significantly suppressed by the simultaneous application of ML204 with muscarine in PC12 cells expressing TRPC1–GFP and TRPC4–GFP (Fig. 6A,C).
Analyses in guinea pig AM cells
Our previous immunocytochemical studies have revealed that TRPC1- and TRPC4-like immunoreactive (IR) material is mainly present in the cytoplasm in guinea pig AM cells (Inoue et al., 2012). If TRPC1–TRPC4 heteromeric channels are involved in NSC currents activated by muscarinic agonists, they should be trafficked to the cell membrane in response to muscarine stimulation. Thus, this notion was immunocytochemically examined. As shown in Fig. 7A, TRPC1- and TRPC4-like IR material present in the cytoplasm was trafficked to the cell periphery upon muscarine stimulation and co-localized with Na+/K+-ATPase α1 subunit (α1, also known as ATP1A1)-like IR material as a marker of the cell membrane. The level of α1-like IR material co-localized with TRPC1-like IR material (as a percentage of total TRPC1-like IR material) increased from 18.3% to 42.3%, whereas that of α1-like IR material co-localized with TRPC4-like IR material (as a percentage of the total TRPC4-like IR material) also increased from 24.0% to 37.7% (Fig. 7B). Furthermore, TRPC1-like IR material was coincident with TRPC4-like IR material (Fig. 7C), with TRPC1/TRPC4 and TRPC4/TRPC1 coincidence rates of 42.2% and 45.1%, respectively, and the values were not affected by muscarine stimulation (Fig. 7D). The coincidence of TRPC1-like and TRPC4-like IR material suggests that TRPC1 and TRPC4 form a heteromeric channel in guinea pig AM cells.
As discussed above, experiments in PC12 cells demonstrated that the anti-TRPC1 and anti-TRPC4 antibodies used were specific and useful for PLA. Thus, these reagents were employed to investigate directly whether TRPC1 forms a heteromeric channel with TRPC4 in guinea pig AM cells. As shown in Fig. 8A,B, PLA revealed that TRPC1 and TRPC4 form a heteromer, and that some of the heteromers were apparently trafficked to the cell membrane in response to stimulation with 30 µM muscarine. While TRPC5-like IR material has previously been detected at the cell periphery (Inoue et al., 2012), PLA reaction between TRPC5 and TRPC1 was scarcely found to occur (Fig. 8B,C).
Finally, an electrophysiological approach was used to examine whether TRPC4 or TRPC5 is inserted into the cell membrane in response to muscarine stimulation in guinea pig AM cells. One of the properties of TRPC4 and TRPC5 is that lantanides, such as La3+, have a double action on channel activity, i.e. enhancement and suppression (Jung et al., 2003; Semtner et al., 2007). Thus, if the channels are located at the cell membrane, La3+ application is expected to induce an inward current at negative membrane potentials (Bezzerides et al., 2004). As shown in Fig. 8D, exposure to 600 µM La3+ resulted in development of an inward current at −60 mV in 64% of AM cells (n=11), a current which was sustained during La3+ treatment. By contrast, inward currents induced by 10 µM muscarine were transiently enhanced in the presence of La3+. The peak amplitude of current evoked by the first application of muscarine in the presence of La3+ was 286.0±67.4% (n=6) of that in its absence. The extent of this enhancement was successively diminished upon repeated application of muscarine (Inoue et al., 2012), and the level of inward current at the end (40–50 s) of a third application in the presence of La3+ was 30.0±8.9% (n=6) of that in its absence. What is more noteworthy is that a noise level was markedly diminished during the development of inward currents, whereas the muscarine-induced currents were associated with an increase in current noise. To explore the ionic mechanism for La3+-induced inward currents, I–V curves were examined with 50 ms pulses before and during La3+ application (Fig. 8E). The I–V curve for La3+-induced current (Fig. 8F) showed that the current had an inwardly rectifying property with a reversal potential of −81.5±1.2 mV (n=4), a value which is close to the equilibrium potential for K+ (−83.6 mV), suggesting that the current is at least in part due to inwardly rectifying K+ channel inhibition (Inoue and Imanaga, 1993) and not to NSC channel activation. Taken together with the immunocytochemical findings, the electrophysiological results strongly suggest that muscarinic receptor stimulation in guinea pig AM cells results in the insertion of heteromeric TRPC1–TRPC4 channels into the cell membrane.
DISCUSSION
Activation mechanism
The TRPC family comprises seven isoforms, which are divided into three subgroups (Clapham, 2003): subgroup one comprises TRPC1, TRPC4 and TRPC5; subgroup two consists of TRPC3, TRPC6 and TRPC7; while TRPC2, which is a pseudogene in humans, forms subgroup three. Although stimulation of receptors coupled with Pertussis toxin-insensitive G proteins is known to activate a heteromer of TRPC1 and TRPC4 or TRPC5 (Strübing et al., 2001; Kim et al., 2014), the signal transduction mechanism is not yet sufficiently elucidated. In the present study, TRPC1 was found to conspicuously form a complex with TRPC4, but not TRPC5, in PC12 cells and guinea pig AM cells, and this heteromer was present in the cytoplasm under resting conditions and trafficked to the cell periphery in response to muscarine stimulations. In contrast to PC12 cells, TRPC4–GFP channels expressed in HEK293 cells have previously been shown to be located at the cell periphery (Schaefer et al., 2000), and TRPC4 proteins in HEK293 cells were surface labeled (Kim et al., 2014). These findings indicate that at least some TRPC4 proteins exogenously expressed in HEK293 cells are located at the cell membrane. Thus, interesting cell type-specific trafficking patterns exist. While TRPC4 and TRPC5 both were localized at the cell periphery in HEK293 cells, in PC12 cells (present results) and guinea pig AM cells (Inoue et al., 2012) TRPC4 was mainly present in the cytoplasm and TRPC5 at the cell periphery. Thus, AM cells and PC12 cells might lack the constitutively active machinery to transport TRPC4 proteins to the cell membrane. Alternatively, because TRPC1 and TRPC4 form a heteromeric channel in PC12 cells in a STIM1-dependent manner, TRPC1 or STIM1, resident proteins of the ER membrane (Lewis, 2011), might hinder the transport of TRPC4 to the cell membrane. As the expression level of STIM1 in HEK293 cells increases, the amount of TRPC5 immunoprecipitated together with TRPC1 has been shown to decrease (Alicia et al., 2008). This result raises the possibility that STIM1 hinders the heteromer formation of TRPC1 with TRPC5, possibly in the ER. Otherwise, TRPC1 would be trafficked to the cell membrane as a component of the heteromeric channel (Alfonso et al., 2008). Because PLA products between STIM1 and TRPC1 or TRPC4 developed in PC12 cells, it would be reasonable to conclude that TRPC1–TRPC4 heteromeric channels in PC12 cells and guinea pig AM cells are mainly present in the ER, along with STIM1.
The findings that mAChR stimulation in PC12 cells induced the trafficking of TRPC1–TRPC4 heteromeric channels to the cell membrane and an increase in [Ca2+]i suggest that membrane trafficking results in the insertion of heteromeric channels into the cell membrane with consequent depolarization. Indeed, the experiment with extracellularly tagged TRPC channels clearly demonstrated the insertion of TRPC1–TRPC4 heteromeric channels into the cell membrane in response to muscarinic receptor stimulation. TRPC1 channels in the cytoplasm have been shown to be recruited to the cell membrane in response to Ca2+ store depletion in HEK293T cells (Alicia et al., 2008) and HSG cells, a cell line originating from a human salivary gland (Cheng et al., 2011), and TRPC1–TRPC4 heteromeric channels function as a store-operated Ca2+ entry channel in endothelial cells (Sundivakkam et al., 2012). Although it is possible that muscarinic receptor stimulation recruits TRPC1–TRPC4 channels to the cell membrane as a result of Ca2+ store depletion (Sundivakkam et al., 2012), this is unlikely as exposure to caffeine has previously been shown not to result in the development of an inward current in guinea pig AM cells (Inoue and Imanaga, 1998). What is more important is that expression of the constitutively active STIM1 mutant, STIM1D76A, did not result in membrane insertion of TRPC1–TRPC4 channels in PC12 cells under resting conditions, but facilitated insertion of the channel into the cell membrane in response to muscarinic receptor stimulation. When STIM1D76A was expressed in HEK293 cells, the ER became tubular and endoplasmic reticulum–plasma membrane (ER–PM) contacts were greatly expanded (Grigoriev et al., 2008). Although this expansion has been thought to result in the insertion of TRPC1 channels into the cell membrane (Alicia et al., 2008), TRPC1–TRPC4 heteromeric channels in PC12 cells were not inserted into the cell membrane. What allows for this difference between TRPC1 in HEK293 cells and TRPC1–TRPC4 in PC12 cells remains to be explored. At this stage, it would be rational to conclude that ER Ca2+ depletion itself does not result in the membrane insertion of TRPC1–TRPC4 channels, but leads to their translocation to the vicinity of the cell membrane, probably as a result of the expansion of ER–PM contacts. This juxtaposition of TRPC1–TRPC4 and the cell membrane may allow for facilitation of membrane insertion in response to muscarinic receptor stimulation. Based on current concepts in cell biology, the ER membrane is not thought to directly fuse with the plasma membrane (Saheki and De Camilli, 2017). Thus, the putative mechanism for channel insertion would be exocytotic fusion of vesicles budded off from the ER. This notion may be consistent with the finding that muscarinic receptor stimulation is mandatory for the membrane insertion of TRPC1–TRPC4 in PC12 cells and probably guinea pig AM cells.
The TRPC4 channel, when exogenously expressed in COS-7 cells, has been reported to be trafficked to the cell membrane upon epidermal growth factor (EGF) stimulation, and this EGF-induced membrane insertion was proposed to be mediated by phosphorylation of tyrosine residues near the C-terminus of TRPC4 and the consequent facilitation of Na+/H+ exchange regulatory cofactor (NHERF, also known as SLC9A3R1) binding (Odell et al., 2005). We have recently reported that in rat AM cells and PC12 cells, muscarinic M1 receptor stimulation rapidly induces endocytosis of TASK1 channels through a signal pathway comprising phospholipase C, protein kinase C and Src (Matsuoka and Inoue, 2017). Thus, it is possible that the muscarine-induced trafficking of TRPC1–TRPC4 heteromeric channels is mediated by a similar signaling pathway. However, efficacy and potency of muscarinic agonists to inhibit TASK channels differed from those that induce the trafficking of TRPC1–TRPC4 to the cell periphery (Inoue et al., 2019a). These pharmacological findings are difficult to reconcile with the involvement of a similar signaling pathway. Further studies will be needed to elucidate the molecular mechanism for muscarinic receptor-mediated trafficking of TRPC1–TRPC4 channels.
Roles of STIM1
The roles of STIM1 in GPCR-mediated activation of TRPC channels are complicated and seem to depend on the expression levels of TRPC channels. For example, when TRPC3 proteins are highly expressed, STIM1 is not obligatory for GPCR-mediated activation; however, STIM1 is obligatory in the case of low expression of TRPC3 (Yuan et al., 2007). The present study clearly demonstrated that the simultaneous expression of exogenous STIM1 enhances heteromer formation of TRPC1 and TRPC4. The heteromer formation in PC12 cells expressing no exogenous STIM1 might be ascribed to the presence of endogenous STIM1 (Wang et al., 2015). The intriguing findings here are that heteromer formation was not facilitated by expression of either STIM1ΔCt or STIM1CT mutants, and fully developed with the expression of STIM1_1-448, which contains a CAD region. Furthermore, CAD acted as a dominant negative for STIM1-dependent heteromeric channel formation. Basic residues in CAD (amino acids 342–448) have been shown to bind to acidic residues in the vicinity of the C-terminal end of Orai1 (Park et al., 2009; Lewis, 2011), whereas an extended region (amino acids 251–535) including CAD is also known to bind to TRPC1, TRPC4 or TRPC5 (Yuan et al., 2007; Lee et al., 2010). The sequence near the C-terminus of such TRPC channels is also rich in acidic residues, as is noted with that of Orai1. Thus, it is likely that basic residues in CAD also bind to acidic residues clustered near the C-terminus of TRPC1 or TRPC4. The fact that STIM1CT, comprising the cytoplasmic segment of STIM1, did not facilitate heteromer formation suggests that STIM1 may facilitate heteromer formation of TRPC1 with TRPC4 in the ER. It was recently shown that a coiled-coil domain in the N-terminus and a segment of 20 amino acids in the C-terminus of each of TRPC1 and TRPC4 are prerequisites for heteromer formation (Myeong et al., 2016). Thus, TRPC1 and TRPC4 may be accumulated in the ER through each binding to STIM1, which is known to multimerize, where they then interact to form heteromeric channels. Another interesting finding with STIM1 mutants is that muscarine stimulation failed to induce the trafficking of heteromeric channels in PC12 cells expressing STIM1_1-448. This mutant lacks a motif (TRIP) for binding to end-binding protein 1 (EB1, also known as MAPRE1), which plays an important role in the targeting to microtubule ends of microtubule plus-end tracking proteins, such as STIM1 (Honnappa et al., 2009). Muscarinic receptor stimulation might extend the ER toward the cell membrane through the interaction between STIM1 and EB1 (Várnai et al., 2008) and then facilitate budding from the ER with the consequent fusion of vesicles to the cell membrane. The present results pave the way to exploring how muscarinic receptor stimulation activates TRPC1–TRPC4 heteromeric channels.
In conclusion, the present study demonstrated two obligatory roles of STIM1 for muscarinic activation of TRPC1–TRPC4 channels in PC12 cells: one is to facilitate heteromeric channel formation of TRPC1 with TRPC4, but not TRPC5, in PC12 cells and guinea pig AM cells; the second is to facilitate the insertion of TRPC1–TRPC4 channels into the cell membrane in response to muscarinic receptor stimulation.
MATERIALS AND METHODS
Male Hartley 1-to-2-month-old guinea pigs (Cavia porcellus) were used. All procedures for the care and treatment of animals were carried out according to the Japanese Act on the Welfare and Management of Animals and the Guidelines for the Proper Conduct of Animal Experiments issued by the Science Council of Japan. The experimental protocols were approved by the Institutional Animal Care and Use Committee of the University of Occupational and Environmental Health (AE07-012). All efforts were made to minimize suffering and to reduce the number of animals used in this study.
Immunocytochemistry
Immunocytochemical staining in an acutely isolated AM cell was performed as described elsewhere (Inoue et al., 2000). Briefly, the animals were killed by cervical dislocation, and adrenal glands were excised and immediately put into ice-cold Ca2+-deficient saline in which 1.8 mM CaCl2 was omitted from standard saline. The standard saline contained 137 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl2, 0.5 mM MgCl2, 0.53 mM NaH2PO4, 5 mM D-glucose, and 4 mM NaOH (pH 7.4). Microscissors and forceps were used to remove the adrenal cortex from the adrenal medulla under stereoscopic observations. The adrenal medulla was cut into half and then treated with 10 mg of collagenase (Yakult Pharmaceutical Industry, Kunitachi, Japan) dissolved in 1.5 ml of Ca2+-deficient saline for 15 min while the preparations were gently stirred with bubbles of O2 gas. This procedure of digestion was repeated twice with a fresh enzyme solution. After enzymatic digestion, one or two pieces of adrenal medullae in Ca2+-deficient saline were placed on a glass-bottomed dish (Matek, Ashland, MA, USA) and then dissociated using fine needles under microscopic observations. The resulting isolated AM cells were left for 30 min to settle on the glass bottom before muscarine stimulation, and then the cells were fixed in 4% paraformaldehyde (PFA) for 1 h, unless otherwise noted. The fixed cells were incubated in phosphate-buffered saline (PBS) with 5% fetal bovine serum and 0.3% Triton X-100 for 30 min. For indirect immunofluorescence studies, cells were treated overnight with primary antibodies (see ‘Reagent details’ below). After incubation, the cells were washed three times in PBS and then treated with secondary antibodies conjugated with Alexa Fluor 488, 546 or 633 (1:200; Molecular Probes Thermo Fisher Scientific, Tokyo, Japan). The immunostaining was observed with a confocal laser scanning microscope (LSM5Pascal, Carl Zeiss, Tokyo, Japan). Excitation wavelength and emission filters were a 458 nm laser and 475–525 nm filter for cyan fluorescent protein (CFP) (CFP-like fluorescence), a 488 nm laser and 510–560 nm filter for Alexa 488 and green fluorescent protein (GFP) (FITC-like fluorescence), 543 nm laser and 560 nm long pass filter for Alexa 546 (rhodamine-like fluorescence), and a 633 nm laser and 650 nm long pass filter for Alexa 633 (Alexa 633-like fluorescence), respectively. Experiments were carried out at 26±2°C.
Voltage-clamp recording
The perforated patch clamp method was used to record the whole-cell current in isolated guinea pig AM cells, as described elsewhere (Inoue et al., 2008). Briefly, digested adrenal medullae in Ca2+-deficient saline were placed in a bath apparatus on the stage of an inverted microscope, and AM cells were dissociated mechanically with fine needles and left for 30 min to settle on the glass bottom before the start of perfusion with the standard saline. The pipette solution contained 120 mM potassium isethionate, 20 mM KCl, 10 mM NaCl, 10 mM HEPES and 2.6 mM KOH (pH 7.2). On the day of the experiment, nystatin dissolved in dimethyl sulfoxide (5 mg in 10 µl) was added to the pipette solution at a final concentration of 100 µg ml−1. The membrane potential was corrected for a liquid junction potential of −3 mV between the pipette solution and standard saline. The current was recorded with an Axopath 200A amplifier (Axon, Foster City, CA, USA) and then fed into a thermal recorder after low-pass filtering at 15 Hz and into a data recorder. To study the I–V curve, 50 ms square pulses were applied in steps of 10 mV from a holding potential of −60 mV, and the current level at the end of pulses was measured and plotted against the membrane potential. The effects of La3+ were examined in a saline in which NaH2PO4 was omitted from the standard saline.
Plasmid construction
For the examination of membrane insertion of TRPC channels by muscarinic stimulation, FLAG and hemagglutinin (HA) tags were inserted into the first outer loop of TRPC1–GFP (TRPC1–FLAG–GFP) and the third outer loop of TRPC4–GFP (TRPC4–HA–GFP), respectively. Both constructs were created by two-step nucleotide insertion. For TRPC1–FLAG–GFP, the first 12 nucleotides (corresponding to four amino acids, DYKD) of FLAG tag were inserted between Gly359 and Arg360, then the last 12 nucleotides (corresponding to four amino acids, DDDK) were inserted right after the first-inserted nucleotides by PCR-based method. For TRPC4–HA–GFP, the first 15 nucleotides (corresponding to five amino acids, YPYDV) of HA tag were inserted between Cys554 and Glu555, then the last 12 nucleotides (corresponding to four amino acids, PDYA) were inserted right after the first-inserted nucleotides. The following pairs of oligonucleotide primers were utilized: 5′-tttggcgattataaagatagaatcattcacacacct-3′ and 5′-gattctatctttataatcgccaaattgagatttggg-3′ for the first half of FLAG, 5′-aaagatgatgatgactatagaatcattcacacacct-3′ and 5′-gattctatagtcatcatcatctttataatcgccaaa-3′ for the second half of FLAG; 5′-cggtgctatccatatgatgttgagaaacagaacaacgcg-3′ and 5′-tttctcaacatcatatggatagcaccggatgcctttgca-3′ for the first half of HA, 5′-gatgttcctgactaagcggagaaacagaacaacgcg-3′ and 5′-tttctccgcttagtcaggaacatcatatggatagca-3′ for the second half of HA. To create a constitutively active form of STIM1, Asp76 on STIM1–myc was substituted with alanine by a sequential, overlapping, PCR-based method, and the following pair of oligonucleotide primers was utilized: 5′-ctgatggccgacgatgccaatggtgat-3′ and 5′-atcgtcggccatcagcttatggatgtt-3′.
Cell culture and transfection
Cell culture and transfection were performed as described elsewhere (Matsuoka and Inoue, 2017). Briefly, PC12 cells, provided without authentication by Dr K. Mizuno (Tohoku University, Sendai, Japan), were cultured in Dulbecco's modified Eagle's medium (Invitrogen Life Technologies, Tokyo, Japan) supplemented with 10% fetal bovine serum (Nichirei, Tokyo, Japan) at 37°C in an atmosphere of humidified air (95%) and CO2 (5%). Lipofectamine 2000 (Invitrogen Life Technologies) was used to transfect PC12 cells with plasmids encoding TRPC1–GFP, TRPC4–GFP, TRPC5–GFP (Shimizu et al., 2006), STIM1–myc (Oh-Hora et al., 2008), YFP–STIM1ΔCT (Covington et al., 2010), mCherry–STIM1_1-448 (Convington et al., 2010), myc–STIM1CT (Huang et al., 2006), and/or CFP–CAD (Park et al., 2009), according to the manufacturer's instructions. The cells were fixed with 4% PFA.
Immunoprecipitation analyses
PC12 cells were seeded in 100-mm dishes and cultured to 70–80% confluence before transfection with plasmids encoding TRPC4–GFP or GFP. The cells were lysed with ice-cold TNE buffer (10 mM Tris-HCl pH 7.4, 1 mM EDTA and 150 mM NaCl), to which 1% Nonidet P-40, 1 mM Na3VO4, and a protease inhibitor cocktail (Calbiochem Merk, Tokyo, Japan) were added, and then subjected to centrifugation at 12,000 g for 30 min at 4°C. The supernatant was collected and used as cell lysate (total cell lysate, TCL). For immunoprecipitation assays, cell lysates were incubated with mouse anti-GFP antibody (1:200; sc-9996; Santa Cruz Biotechnology, Santa Cruz, CA, USA) coupled with protein G-Sepharose (GE Healthcare Bio-Sciences, Tokyo, Japan) at 4°C for 3 h. The beads were washed three times with TNE buffer and then the proteins were eluted with SDS buffer (125 mM Tris-HCl pH 6.8, 2% SDS, 10% glycerol). After addition of 2-mercaptoethanol [final content 5% (vol/vol)] and Bromophenol Blue [0.05% (vol/vol)], the same amount of proteins was fractionated by 10% SDS-polyacrylamide gel electrophoresis and transferred to a polyvinylidene difluoride membrane and then subjected to immunoblot analysis with mouse anti-TRPC1 antibody (1:100; sc-133076; Santa Cruz Biotechnology). The immunoblot was repeated three times.
Proximity ligation assay (PLA)
Fixed PC12 cells expressing exogenous proteins, or AM cells were treated with a combination of mouse and rabbit antibodies. The PLA reaction was detected with the Duolink In Situ kit (Olink Bioscience, Uppsala, Sweden) according to the manufacturer's instructions. The PLA reactions occur between target proteins that are located in close proximity to each other (< 40 nm) (Söderberg et al., 2006). The fluorescent signals were observed as rhodamine-like fluorescence using a confocal laser scanning microscope.
Ca2+ measurement
The ratiometric method was used to investigate effects on intracellular Ca2+ concentration ([Ca2+]i). PC12 cells were maintained in high-osmolarity saline where 3% sucrose was added to standard saline. The cells were incubated in 5 µM Fura-2 acetoxymethyl ester-containing high osmolarity saline for 1 h and then for 30 min in Fura-2 acetoxymethyl ester-free solution. To measure a change in [Ca2+]i, an inverted microscope (IX70; Olympus, Tokyo, Japan) equipped with a xenon burner light source and power supply (AH2-RX; Olympus) was used under operation by the image acquisition software HCImage (Hamamatsu Photonics KK, Hamamatsu, Japan). Fura-2 was alternatively excited at 340 and 380 nm each for 300 ms and emission at 510 nm was measured, and then the Fura-2 ratio (emission evoked at 340 nm/emission evoked at 380 nm) was calculated. This protocol was applied every 900 ms. To stimulate cells, 100 µl of standard saline containing 300 µM muscarine with or without 100 µM ML204, a specific inhibitor for TRPC4 and TRPC5 (Miller et al., 2011), was added to 900 µl of dish solution ∼1 min after the start of measurement.
Statistics
All statistical analyses were performed with Prism (v6.07; GraphPad, La Jolla, CA, USA) or Sigma Plot (v13.0; Systat Software, San Jose, CA, USA). The data are presented as means±s.e.m. and n represents the number of cells examined. When data had been shown to have a normal distribution (Shapiro–Wilk), statistical difference was evaluated with a two-tailed Student's t-test or a one-way ANOVA followed by a post hoc test (Tukey's multiple comparison test). Difference was considered significant when P<0.05. Statistical significance is indicated as *P<0.05, **P<0.01 and ***P<0.001.
Reagent details
Muscarine chloride and ML204 were obtained from Sigma-Aldrich (Tokyo, Japan). Fura-2 acetoxymethyl ester was from Dojindo (Kumamoto, Japan). Mouse (1:200; sc-133076) and rabbit (1:200; sc-20110) anti-TRPC1, mouse anti-myc (1:100; sc-40), mouse anti-FLAG (1:100; sc-166355), mouse anti-HA (1:100; sc-7392), rabbit anti-GFP (1:100; sc-8334) antibodies were from Santa Cruz Biotechnology; mouse anti-Na pump α1 subunit antibody (1:100; 05-369) was from Upstate Biotechnology (Lake Placid, NY,USA); mouse anti-STIM1 antibody (1:100; 610954) was from BD Transduction Laboratories (San Jose, CA, USA); rabbit anti-HA antibody (1:100; A190-108A) was from Bethyl Laboratories (Montgomery, TX, USA); mouse anti-TRPC4 (1:200; 75-119) and mouse anti-TRPC5 (1:200; 75-104) antibodies were from Antibodies Incorporated (Davis, CA, USA); rabbit anti-TRPC4 antibody (1:200; ACC-018) was from Alomone Labs (Jerusalem, Israel). A plasmid encoding mouse STIM1–myc was created by Dr A. Rao (Harvard University, Cambridge, MA, USA) and purchased from Addgene (88415; Cambridge, MA, USA); plasmids encoding yellow fluorescent protein (YFP)-tagged human STIM1ΔCt, mCherry-tagged human STIM1_1-448, and CFP-tagged CAD were gifts from Dr R. S. Lewis (Stanford University, Stanford, CA, USA); a plasmid encoding myc-tagged human STIM1CT was a gift from Dr P. Worley (Johns Hopkins University, Baltimore, MD, USA); mouse TRPC1α, mouse TRPC1α–GFP, mouse TRPC4β, mouse TRPC4β–GFP, and mouse TRPC5–GFP constructs were gifts from Dr Y. Mori (Kyoto University, Kyoto, Japan).
Acknowledgements
The authors are grateful to Drs R. S. Lewis (Stanford University), P. Worley (Johns Hopkins University), and Y. Mori (Kyoto University) for the generous gifts of cDNA constructs.
Footnotes
Author contributions
Conceptualization: K.H., H.M., M.I.; Methodology: K.H.; Validation: M.I.; Formal analysis: K.H.; Investigation: K.H., H.M., M.I.; Resources: K.H.; Writing - original draft: M.I.; Supervision: M.I.; Funding acquisition: H.M., M.I.
Funding
This study was supported, in part, by the Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research (KAKENHI; 17K0855 to M.I. and 18K06865 to H.M.).
References
Competing interests
The authors declare no competing or financial interests.