Research ArticlesIon Channels

Ca2+ controls gating of voltage-gated calcium channels by releasing the β2e subunit from the plasma membrane

See allHide authors and affiliations

Science Signaling  05 Jul 2016:
Vol. 9, Issue 435, pp. ra67
DOI: 10.1126/scisignal.aad7247

Kicked off the membrane by calcium

Voltage-gated calcium (Cav) channels govern Ca2+ entry into excitable cells, notably neurons, muscles, and secretory cells. These channels have a pore-forming subunit, an auxiliary subunit, and a regulatory β subunit; the electrophysiological and regulatory characteristics of the channel depend on which subunits are present. Kim et al. found that stimulation of G protein–coupled receptors (GPCRs) coupled to Gq (muscarinic acetylcholine receptors and purinergic receptors), which increase intracellular Ca2+, triggered the release of β2e from the membrane, thereby enhancing the inactivation of Cav2.2 channels. Because cellular excitability must be tightly controlled to mediate appropriate physiological responses, maintain organismal homeostasis, and prevent Ca2+ toxicity, Cav channels are subject to complex regulation. Membrane dissociation of β2e by cytosolic Ca2+, which occurred independently of known regulatory mechanisms, adds GPCR signaling to the complexity. β2e and Cav2.2 are abundant in neurons, and these Gq-coupled GPCRs are important in pain signaling and regulation of the cardiovascular system.

Abstract

Voltage-gated calcium (Cav) channels, which are regulated by membrane potential, cytosolic Ca2+, phosphorylation, and membrane phospholipids, govern Ca2+ entry into excitable cells. Cav channels contain a pore-forming α1 subunit, an auxiliary α2δ subunit, and a regulatory β subunit, each encoded by several genes in mammals. In addition to a domain that interacts with the α1 subunit, β2e and β2a also interact with the cytoplasmic face of the plasma membrane through an electrostatic interaction for β2e and posttranslational acylation for β2a. We found that an increase in cytosolic Ca2+ promoted the release of β2e from the membrane without requiring substantial depletion of the anionic phospholipid phosphatidylinositol 4,5-bisphosphate (PIP2) from the plasma membrane. Experiments with liposomes indicated that Ca2+ disrupted the interaction of the β2e amino-terminal peptide with membranes containing PIP2. Ca2+ binding to calmodulin (CaM) leads to CaM-mediated inactivation of Cav currents. Although Cav2.2 coexpressed with β2a required Ca2+-dependent activation of CaM for Ca2+-mediated reduction in channel activity, Cav2.2 coexpressed with β2e exhibited Ca2+-dependent inactivation of the channel even in the presence of Ca2+-insensitive CaM. Inducible depletion of PIP2 reduced Cav2.2 currents, and in cells coexpressing β2e, but not a form that lacks the polybasic region, increased intracellular Ca2+ further reduced Cav2.2 currents. Many hormone- or neurotransmitter-activated receptors stimulate PIP2 hydrolysis and increase cytosolic Ca2+; thus, our findings suggest that β2e may integrate such receptor-mediated signals to limit Cav activity.

INTRODUCTION

The calcium ion (Ca2+) is a ubiquitous second messenger that regulates most cellular processes (13). Alterations in intracellular Ca2+ concentration are mediated by Ca2+ mobilization from the internal Ca2+ stores, often the endoplasmic reticulum (ER) or Ca2+ influx from the extracellular medium (4). The former is elicited by various plasma membrane–embedded receptors (5), and the latter is primarily mediated by voltage- and ligand-gated calcium channels (6, 7). In unstimulated cells, cytosolic Ca2+ concentration is maintained at ~100 nM by Ca2+ homeostatic mechanisms, including the activity of plasma membrane Ca2+ adenosine triphosphatase (ATPase) (PMCA), sarcoplasmic and endoplasmic reticular ATPase (SERCA), and Na/Ca exchanger (NCX). However, when cells are activated by Gq-coupled GPCRs [heterotrimeric guanine nucleotide–binding protein(G protein)–coupled receptor], the intracellular Ca2+ concentration can increase to 1 μM by Ca2+ efflux through the inositol 1,4,5-trisphosphate (IP3) receptor or the ryanodine receptor on the ER or sarcoplasmic reticulum (SR) (8). The increase in intracellular Ca2+ concentration eventually triggers a multitude of common and cell-specific responses, such as neurotransmitter release, muscle contraction, mitochondrial metabolism, gene expression, cell growth, and proliferation (912). Many Ca2+ signaling cascades involve various Ca2+-binding proteins rather than a direct effect of Ca2+ on downstream molecules (13). These Ca2+-binding proteins not only contribute to the buffering of intracellular Ca2+ but also decode Ca2+ signals in a spatiotemporal-dependent manner. For instance, calmodulin (CaM) is an extensively characterized Ca2+ sensor in cells (14). When Ca2+ binds to the EF-hand motifs on CaM, CaM mediates a wide range of cellular functions, including cell division and differentiation, gene transcription, membrane fusion, and muscle contraction (15).

Although Ca2+-binding proteins are important for Ca2+ homeostasis and Ca2+ signaling, Ca2+ may also directly regulate cellular processes. As a charged ion, Ca2+ can interact with other charged entities in cells, such as the anionic phospholipids in cellular membranes. Several reports have shown a direct effect of Ca2+ on the activity and distribution of proteins in the plasma membrane by binding to charged phospholipids and thereby altering the charge of the inner leaflet of the membrane (1618). These findings suggested that changes in cytosolic Ca2+ concentration play a direct role in various biological processes by mainly modulating the interactions between membrane proteins and membrane lipids.

In the excitable cells, including neurons, skeletal, smooth, and cardiac muscles, and secretory cells, voltage-gated Ca2+ (Cav) channels are the major molecules that mediate Ca2+ influx (6). Cav channels are composed of the pore-forming α1 and the ancillary α2δ and β subunits. In mammals, there are 10 α1, 4 α2δ, and 4 β different forms of each subunit. Cav channels are regulated not only by voltage but also by the phospholipid environment, Ca2+ signals, and phosphorylation (19); physiologically, the excitability of cells is regulated by GPCR ligands, which can regulate Cav channels through any of these mechanisms.

As one of the ancillary subunits, the Cav β subunit plays an important role in trafficking the channels to the plasma membrane, regulating the gating properties, and mediating channel modulation by signaling molecules (20, 21). The β subunits are not transmembrane proteins and instead contain domains that interact with the α-interacting domain of the pore-forming α1 subunit. β2a and β2e also interact with the cytoplasmic side of the membrane through covalent lipid modifications (β2a) and electrostatic and hydrophobic interactions mediated by the amino acid side chains (β2e), enabling these β subunits to localize to the plasma membrane even in the absence of the α1 subunit. Functionally, the interaction of β subunits with the plasma membrane is a major determinant of channel gating, as well as channel modulation by phosphatidylinositol 4,5-bisphosphate (PIP2) (2225). The β subunits that only interact with the α-interacting domain (“cytosolic” β subunits) induce fast inactivation of Cav channels, whereas the membrane-bound β subunits slow the channel inactivation. With regard to the modulation of Cav channels by PIP2, the activity of Cav channels with cytosolic β subunits is strongly inhibited by PIP2 depletion, whereas the effect of PIP2 depletion is much weaker for Cav channels with membrane-tethered β subunits. We focused on the mechanism of β2e-dependent regulation of Cav2.2 channels. The 23–amino acid N-terminal region contains several basic amino acids, targets β2e to the plasma membrane, and thus maintains the channel in a slow inactivation state under basal conditions (2628).

The M1 muscarinic receptor (M1R) is a GPCR coupled to Gq, which stimulates phospholipase C (PLC) activity to cleave PIP2, producing IP3 and diacyglycerol and reducing PIP2 abundance in the plasma membrane. The IP3 stimulates release of Ca2+ from intracellular stores. Thus, this receptor produces multiple signals that can regulate Cav activity. Stimulation of cells overexpressing M1R results in depletion of PIP2 and phosphatidylinositol 4-phosphate (PI4P), disruption of the interaction of β2e with the membrane, and rapid inactivation of the Cav channel (28). However, given the complexity of the signaling pathways of Gq-coupled GPCRs, we investigated the role of changes in Ca2+ and PIP2 or their downstream effectors, regulated β2e localization and channel gating. In particular, we investigated (i) whether Ca2+ directly affected these events by interacting with the anionic phospholipids; (ii) whether CaM, which is negatively charged and can compete with anionic membrane phospholipids or directly bind to basic proteins to interfere with membrane association, was involved; or (iii) whether phosphorylation of β2e altered its interaction with the membrane (29, 30).

By expressing tagged forms of the β2 subunits and various mutants in the absence of the α1 subunit in tsA201 cells [a human embryonic kidney (HEK) 293–derived line], we found that the fluctuation of intracellular Ca2+ dynamically and directly controls the interaction of the β2e subunit with the plasma membrane. An increase of cytosolic Ca2+ concentration resulted in the transient dissociation of the β2e subunit from the plasma membrane and the acceleration of Cav2.2 channel inactivation. The Ca2+ effects were independent of CaM, did not require substantial depletion of PIP2 from the plasma membrane, and were unaffected by mutating potential phosphorylation sites in β2e. Thus, our experimental results indicate that Ca2+, by interfering with tethering the β2e subunit to the plasma membrane, controls Cav channel gating properties and PIP2-mediated regulation of channel activity. Furthermore, this effect of Ca2+ represented a CaM-independent, Ca2+-mediated mechanism to control Cav channel activity.

RESULTS

Gq-coupled GPCR activation reduces the amount of β2e at the plasma membrane

Several types of Gq-coupled GPCRs, including muscarinic and purinergic receptors, are endogenously present in tsA201 cells, which are derived from HEK293 cells (31). These cells do not have Cav channels (32); therefore, we introduced Cav β subunits with or without Cav α1B and Cav α2δ-1 in the cells and then examined the effects of intracellular second messengers on the membrane interaction of Cav β subunits and channel inactivation in response to agonists of the Gq-coupled GPCRs M1R (a muscarinic acetylcholine receptor) or P2Y2 [an adenosine 5′-triphosphate (ATP) receptor]. Membrane tethering of β2a results from palmitoylation on two cysteines of the N terminus (33), whereas β2e is tethered to the membrane through both electrostatic interactions between the polybasic residues of the N terminus and acidic phospholipids and the hydrophobic insertion of tryptophan residue to the plasma membrane (Fig. 1, A and B) (2628). We detected β2a and β2e at the plasma membrane even in the absence of α1B and α2δ-1 (Fig. 1A). Because of the membrane tethering of the β subunit, the cells cotransfected with α1B, α2δ-1, and β2a, or β2e exhibited slow inactivation of Cav currents in response to a voltage step to +10 mV (Fig. 1A).

Fig. 1 Activation of endogenous muscarinic receptor causes shuttling of β2e.

(A) Comparison of the N-terminal sequences of β2a and β2e, their cellular localization, and their effects on Cav2.2 current inactivation. In β2a, red indicates the palmitoylated residues; in β2e, green represents the hydrophobic tryptophan, and blue indicates residues involved in the electrostatic interaction. Confocal images of β2a-GFP and β2e-GFP in the absence of α1 and α2δ-1 subunits. Scale bar, 10 μm. Images are representative of 15 cells in five independent experiments. Current inactivation of Cav2.2 channels in cells transfected with α1B, α2δ-1, and β2a-GFP or β2e-GFP was recorded during a 500-ms test pulse at +10 mV. Currents are representative of 10 cells in three independent experiments. (B) Schematic diagram of the membrane-targeting mechanisms of β2a and β2e. The red squiggly lines indicate the palmitoyl chains that tether β2a to the membrane. Tryptophan and the positively charged residues in the N-terminal region of β2e that enable the hydrophobic and electrostatic interactions are shown. SH, Src homology; GK, guanylate kinase; PM, plasma membrane. (C and F) Confocal images of cells expressing PH-PLCδ-RFP (PH-RFP) and β2a-GFP (C) or β2e-GFP (F) without α1B and α2δ-1 were taken before (Control), during (Oxo-M), and after (Washout) Oxo-M (10 μM) application. Nontransfected cells were loaded with Fluo4-AM for 10 min before imaging in separate experiments. Scale bars, 5 μm. Images are representative of six or seven cells in three independent experiments. Three ROIs were marked in confocal images for analysis of cytosolic fluorescence intensity in a single cell, and one arrow line was marked to measure the fluorescence intensity ratio of the plasma membrane (PM) and the cytosol (Cyto). Note that Fluo4-AM images are at a lower magnification. (D and G) Time courses showing changes of fluorescence intensity in regions of interest in cells expressing PH-RFP and β2a-GFP (D) or β2e-GFP (G) and exposed to Oxo-M (10 μM). Cytosolic fluorescence intensity of PH-RFP and β2a-GFP or β2a-GFP was measured in six (D) or seven (G) cells from three independent experiments. Fluo4-AM Ca2+ signals were measured in 9 (D) or 12 (G) cells from three independent experiments. Data are means ± SEM. (E and H) Quantification of fluorescence intensity ratio of plasma membrane and cytosol (FPM/Fcyto) was calculated from line intensity histograms of cells expressing PH-RFP and β2a-GFP (E) and β2a-GFP (H). n = 6 cells from three independent experiments. ***P < 0.001, with two-way analysis of variance (ANOVA) followed by Tukey’s post hoc test. Data are means ± SEM. (I) Confocal images of cells expressing mCherry-tagged β2e (β2e-mCh) and IP3K-A-GFP without other Cav subunits in the presence or absence of Oxo-M (10 μM). Scale bars, 5 μm. Images are representative of six cells in three independent experiments. (J) Time courses of cytosolic fluorescence intensity changes in cells loaded with Fluo4-AM (left) or in cells expressing β2e-mCherry (right) in response to Oxo-M in the presence or absence of coexpressed IP3K-A. Cytosolic β2e-mCherry fluorescence was measured in six cells from three independent experiments; Ca2+ signal was measured in nine cells from three independent experiments. Data are means ± SEM.

Our previous data show that, in tsA201 cells coexpressing M1R and β2e without the other subunits, the addition of the M1R agonist oxotremorine-M (Oxo-M) triggered translocation of the β2e subunit (28). Here, we examined the effects of the endogenous receptors, such as muscarinic and purinergic receptors, on the interaction of β2 subunits with the plasma membrane by expressing green fluorescent protein (GFP)–tagged β2a (β2a-GFP) or GFP-tagged β2e (β2e-GFP) and PH-PLCδ-RFP [pleckstrin homology–red fluorescent protein (PH-RFP)] to monitor changes in PIP2 abundance in the membrane. PH-RFP fluorescence at the membrane decreases as PIP2 is metabolized. In nontransfected control cells, we detected changes in cytosolic Ca2+ by loading the cells with the Ca2+ indicator dye Fluo4-AM. Despite stimulating a transient increase in intracellular Ca2+ in separate cells, consistent with a previous study (34), Oxo-M did not affect the subcellular location of β2a-GFP (Fig. 1, C to E). Without α1B, Oxo-M induced significant translocation of β2e-GFP from the plasma membrane to the cytosol (Fig. 1, F to H). In neither case did Oxo-M induce a notable PIP2 change in the plasma membrane (Fig. 1, E and H). We also tested the effect of ATP-mediated activation of the endogenous purinergic receptors (35) on Ca2+ signaling, the distribution of β2a-GFP and β2e-GFP, and PIP2 abundance. ATP application did not change the PH-RFP signal or alter the distribution of the β2a-GFP signal (fig. S1A), but did induce an increase in the cytosolic β2e-GFP signal (fig. S1B).

As reported previously (28), we observed an increase in cytosolic fluorescence for PH-RFP, indicating PIP2 hydrolysis, in response to Oxo-M stimulation of M1R-transfected cells, and β2e-GFP also shifted into the cytosol (fig. S2A). We observed a similar response to ATP in cells overexpressing P2Y2 receptors (fig. S2B). These results with the cells overexpressing the GPCRs are consistent with previous data (28). However, the results with cells with the endogenous receptors suggested that even though membrane phosphoinositides (PIs) are required for the membrane targeting of β2e, transient translocation of β2e to the cytosol by endogenous Gq-coupled GPCR activation does not require PIP2 depletion, indicating that other factors can participate in the translocation process of β2e.

Because activation of endogenous receptors elicited an increase in intracellular Ca2+, we examined whether Ca2+ released from the intracellular Ca2+ store contributed to β2e release from the membrane. We expressed mCherry-tagged β2e without the α1B subunit and also expressed IP3 3-kinase-A (IP3K-A) tagged with GFP, which prevents Ca2+ release from ER by converting IP3 to IP4 (36), in tsA201 cells. Upon stimulation of the endogenous muscarinic receptor with Oxo-M, we found that expression of IP3K-A-GFP abolished the Oxo-M–induced increase in cytosolic Ca2+ and also failed to elicit an increase in the cytosolic β2e-mCherry signal (Fig. 1, I and J). Likewise, the translocation of β2e-mCherry by purinergic activation was also abolished in IP3K-A-GFP–expressing cells (fig. S3, A and B). These results indicated that the translocation of β2e depended on the receptor-mediated increase in intracellular Ca2+.

Ca2+ mobilization is sufficient for β2e translocation

To further elucidate the functional effect of Ca2+ ions on the cytosolic translocation of the β2e subunit, we used thapsigargin and the Ca2+ ionophore ionomycin to increase cytosolic Ca2+ concentration. Thapsigargin inhibits the SERCA pump, thus causing the accumulation of Ca2+ in the cytosol (37). Thapsigargin produced a sustained increase in intracellular Ca2+ when extracellular Ca2+ was present and stimulated the cytosolic translocation of β2e (Fig. 2A). Internal Ca2+ responses to thapsigargin in the absence of extracellular Ca2+ were in agreement with those of previous studies conducted on HEK293 cells (34, 38). β2e translocation and the increase in intracellular Ca2+ occurred in two phases after thapsigargin application. In the first phase with Ca2+-free medium, the intracellular Ca2+ concentration increased slightly as Ca2+ leaked out of the ER and a small fraction of β2e translocated to the cytosol. In the second phase with the addition of Ca2+-containing medium, intracellular Ca2+ increased strongly as a result of influx from the extracellular medium, and β2e robustly moved from the plasma membrane to the cytosol. PH-RFP fluorescence was unaffected under either conditions, indicating that membrane PIP2 was not depleted.

Fig. 2 Cytosolic Ca2+ increase triggers cytosolic translocation of β2e from the plasma membrane.

(A and B) Top: Confocal image of cells expressing PH-RFP and β2e-GFP without α1B and α2δ-1 or control cells loaded with Fluo4-AM and exposed to 2 μM thapsigargin (TG) (A) or 2 μM ionomycin (Iono) (B). Black and red bars indicate the concentration of Ca2+ in external medium. The regions marked with white circles were used for time-course analysis of cytosolic fluorescence intensity. Arrows indicate lines that were marked for the analysis of plasma membrane/cytosol intensity ratio. Scale bars, 5 μm. Images are representative of seven cells in three independent experiments. Middle: Time courses showing changes of fluorescence intensity in regions of interest in cells expressing PH-RFP and β2e-GFP or in control cells loaded with Fluo4-AM and exposed to thapsigargin or ionomycin. The fluorescence intensity was taken every 5 s in cells with PH-RFP and β2e-GFP or every 3 s in cells loaded with Fluo4-AM. n = 7 cells from three independent experiments. Data are means ± SEM. Bottom: Plasma membrane/cytosol fluorescence ratio was calculated from line intensity histograms of cells expressing β2e-GFP and PH-RFP and exposed to 2 μM thapsigargin or ionomycin. n = 7 cells from three independent experiments. **P < 0.005 and ***P < 0.001, with two-way ANOVA followed by Tukey’s post hoc test. Data are means ± SEM.

With ionomycin as a Ca2+ ionophore, we also observed a synchronous increase in cytosolic Ca2+ and redistribution of β2e from the membrane to the cytosol (Fig. 2B). At concentrations greater than 5 μM, ionomycin also evokes PIP2 depletion by activating PLCβ (39, 40). However, at the concentration we used (2 μM), ionomycin induced translocation of β2e without PIP2 depletion. The thapsigargin and ionomycin results indicated that an increase in cytosolic Ca2+ is necessary for β2e redistribution from the membrane to the cytosol. However, our data indicated that the redistribution is independent from the origin of Ca2+ ions whether from internal stores or extracellular influx. Moreover, depletion of membrane PIP2 was not involved in the β2e translocation.

β2e translocation is independent of CaM and protein kinases

Considering that cytosolic Ca2+ triggers various biological processes (2, 18), it is possible that signaling molecules, such as the known Cav regulator CaM, activated by Ca2+ ions could be involved in the translocation of β2e. Therefore, we examined the involvement of CaM in the β2e translocation. We coexpressed β2e-GFP with either wild-type CaM or a dominant-negative mutant CaM (DN-CaM), which is insensitive to Ca2+ because of alanine substitutions in the four Ca2+-binding EF-hand domains. We observed a robust redistribution of β2e-GFP from the plasma membrane to the cytosol in response to ATP in cells expressing only β2e-GFP and in cells coexpressing either CaM protein (Fig. 3, A to C).

Fig. 3 Ca2+-mediated β2e translocation is independent of CaM.

(A) Confocal images of cells expressing β2e-GFP alone without α1B and α2δ-1 in the absence (None) and presence of wild-type CaM (WT-CaM, middle) or DN-CaM (bottom) in response to 50 μM ATP. The regions marked with white circles were used for time-course analysis of cytosolic fluorescence intensity. Arrows indicate the lines that were marked for the analysis of plasma membrane/cytosol ratio. Scale bar, 5 μm. Images are representative of six cells in three independent experiments. (B) Time courses showing changes of fluorescence intensity in regions of interest in cells expressing β2e-GFP alone (None) or with WT-CaM or DN-CaM and exposed to ATP. n = 6 cells from three independent experiments. Data are means ± SEM. (C) Quantification of fluorescence intensity ratio of plasma membrane and cytosol was calculated from line intensity histograms of cells expressing β2e-GFP alone or with WT-CaM or DN-CaM, before, during (ATP), and after exposure to 50 μM ATP. n = 6 cells obtained two independent experiments. ***P < 0.001, with two-way ANOVA followed by Tukey’s post hoc test. Data are means ± SEM. (D) Analysis of CaM binding to β2e-GFP and N-del-GFP from transfected HEK293 cells by 3× FLAG-hCaM immunoprecipitation. 3× FLAG-hCaM was immunopurified by anti–FLAG M2 agarose beads. Lysates from cells expressing β2e-GFP and 3× FLAG-hCaM or N-del-GFP and 3× FLAG-hCaM were incubated with M2 agarose beads under Ca2+- or EGTA-containing conditions. GFP-IQ(1–4) was used as a positive control of CaM binding. Data are representative of three independent experiments. EGFP, enhanced GFP.

We also performed an immunoprecipitation assay to detect the formation of a complex between β2e and CaM. We used human CaM (hCaM) with three FLAG tags and the GTPase-activating protein IQGAP1 (tagged with GFP) as a positive control, on the basis of a previous report demonstrating that their interaction was independent of Ca2+ (41). We used a deletion mutant of β2e lacking the N-terminal region (N-del-β2e) and tagged with GFP as a negative control. We immunoprecipitated with the FLAG antibody in the presence and absence of the Ca2+ chelator EDTA and blotted for the GFP or FLAG tag (Fig. 3D). Although IQGAP1 bound CaM in the presence or absence of Ca2+, β2e did not bind to CaM under either condition, suggesting that CaM is not involved in the Ca2+-mediated translocation of the β2e subunit.

Because β2e has three potential sites that could be phosphorylated by protein kinase C (PKC) (fig. S4A), we tested for the involvement of PKC in Ca2+-mediated β2e translocation by stimulating the tsA201 cells with phorbol 12-myristate 13-acetate (PMA), a PKC activator. In hippocampal astrocytes and airway epithelial cells, PMA stimulated the recruitment of the C1 domain of PKCγ domain to the plasma membrane, and the PKC substrate MARCKS (myristoylated alanine-rich C-kinase substrate) translocated to the cytosol (42, 43). We used these responses as positive controls by expressing the C1 domain fused to yellow fluorescent protein and MARCKS fused with GFP and monitoring their distribution after PMA application (fig. S4B). PMA had no effect on β2e distribution (fig. S4B). To further rule out phosphorylation of β2e by PKC, we mutated the potential phosphorylation sites in the N-terminal region of β2e (fig. S4A). ATP induced the cytosolic translocation of these mutants (T4A and S19,29A) (fig. S4C). Together, these results indicated that CaM- and PKC-mediated phosphorylation reactions are not involved in the translocation of β2e.

Peptide binding to liposome membrane is diminished by Ca2+ addition

In our previous study (27), we showed, using peptide-to-liposome fluorescence resonance energy transfer (FRET) assay, that the N-terminal peptide of β2e bound selectively to liposome membranes containing 15% phosphatidylserine (PS) and 1% PIP2, confirming that anionic phospholipids are essential for the binding of β2e to the membrane. Here, we assessed whether Ca2+ disrupted the interaction between the peptide and the liposomes. Tryptophan (W) in the 23-residual peptide of the N terminus of the β2e subunit (MKATWIRLLKRAKGGRLKSSDIC) served as the donor, and dansyl–phosphatidylethanolamine (PE) in the liposome acted as the acceptor. A decrease in the emission of the tryptophan (355 nm) indicates an interaction between the N-terminal peptide and the liposome (Fig. 4A) (27). We monitored the fluorescence intensity of the tryptophan donor in the presence of liposomes and various concentrations of divalent ions. The addition of CaCl2 to a mixture of the N-terminal peptide and liposomes containing PS and PIP2 increased the fluorescence emission of the peptide (intensity at 355 nm) (Fig. 4B), suggesting that Ca2+ ions inhibit the interaction of the peptides with the liposome. To quantify the change of emission, we normalized the tryptophan fluorescence in the presence of liposomes with various Ca2+ concentrations to the initial value without liposome addition. This analysis showed that the increase in Ca2+ concentration reduced the FRET signal (Fig. 4C), consistent with the increase in tryptophan fluorescence intensity at 355 nm (Fig. 4C). Examination of the Ca2+ effect on N-terminal binding at various concentrations showed that Ca2+ had a concentration-dependent effect, exhibiting a substantial effect on the hundred micromolar range, indicating the reduction of binding affinity of the peptide for the liposomes (Fig. 4D).

Fig. 4 Divalent ions interfere with the interaction between the β2e N-terminal peptide and liposomes.

(A) Schematic diagram of the FRET-based peptide-liposome binding assay. Binding was tested using FRET between the Trp5-containing peptide from β2e (W5), which served as the FRET donor, and liposomes labeled with 5% dansyl-PE, which served as the FRET acceptor. The initial spectrum of tryptophan was measured in the absence of liposomes (F0), and the subsequent spectrum was recorded after liposome addition (F) in various concentrations of divalent ions. (B) Fluorescence emission spectrum from tryptophan in the peptide-to-liposome binding assay in the absence or presence of CaCl2. a.u., arbitrary unit. (C) Summary of FRET in various Ca2+ concentrations at 355 nm. Data are means ± SEM of two experiments. ***P < 0.001, one-way ANOVA followed by Tukey’s post hoc test. (D) Concentration-dependent effect of Ca2+ on the binding between peptides and liposomes. Normalized F is represented as a function of concentration of Ca2+ ions, and solid lines indicate the fit of the F0/F signals to the Hill equation. (E) Fluorescence emission spectrum from tryptophan in the peptide-to-liposome binding assay in various Mg2+ concentrations. (F) Summary of FRET in various Mg2+ concentrations. Data are means ± SEM of two experiments. ***P < 0.001, one-way ANOVA followed by Tukey’s post hoc test.

To determine whether a decrease in peptide binding to the liposome was due to membrane charge screening by Ca2+ ions, we performed the assay with Mg2+, another physiological divalent ion. We found that Mg2+ addition also decreased the peptide-liposome interaction (Fig. 4E). The screening effects of 0.1 mM CaCl2 or 3 mM MgCl2 were significant, and the presence of 0.1 mM CaCl2 enhanced the effect of 3 mM MgCl2 (Fig. 4F). Overall, these results demonstrated that Ca2+ disrupts the interaction of the basic N-terminal peptide and acidic liposome. These data support the hypothesis that, in cells, cytosolic Ca2+ directly triggers β2e translocation by screening the charge of the membrane phospholipids.

Ca2+ induces CDI of Cav channels with β2e regardless of CaM

To determine whether the translocation of β2e from the plasma membrane to the cytosol by Ca2+ ions affected the gating properties of the Cav channel, we tested Ca2+-dependent inactivation (CDI) in tsA201 cells transiently transfected with the Cav2.2 channel subunits α1B, α2δ-1, and β2, along with wild-type CaM or DN-CaM. Because CaM functions as the Ca2+ sensor that mediates CDI for most Cav channels, expression of Ca2+-insensitive DN-CaM typically abolishes CDI (44, 45). Therefore, we assessed the ability of Ca2+/CaM to mediate CDI of Cav channels with β2e. For reference, we first measured the conductance of the Cav2.2 channel with β2a in cells coexpressing wild-type CaM or DN-CaM (Fig. 5A). As a quantitative index of CDI, we plotted the fraction of Ca2+ current at 300 ms after the peak during 500-ms depolarization (r300) as a function of voltage (Fig. 5B). Because Ba2+ does not bind CaM and thus cannot trigger CDI through that mechanism, we calculated the differences in conductance (f value) between the Ba2+ and Ca2+ currents as an indication of CDI. In cells expressing wild-type CaM, the Ba2+ current exhibited slow inactivation and minor, monotonic decline of r300, whereas the Ca2+ current exhibited fast inactivation and a U-shaped response of r300 to voltage, which are hallmarks of CDI (Fig. 5, A and B). In contrast, in cells expressing DN-CaM, less difference between Ba2+ and Ca2+ conductance occurred (as indicated by the smaller f value compared with that in the cells coexpressing wild-type CaM), indicating that CaM mediates CDI of Cav2.2 channels with β2a.

Fig. 5 Ca2+ induces fast inactivation of Cav2.2 channels with the β2e subunit independently of CaM.

(A and C) CDI of Cav2.2 channels with the β2a subunit (A) or β2e (C). Top: Voltage protocol and representative whole-cell Ba2+ (black) and Ca2+ (red) currents elicited by 500-ms depolarization step to +10 mV in cells expressing WT-CaM or DN-CaM. Whole-cell Cav2.2 currents were measured in the presence of 10 mM BaCl2 or 10 mM CaCl2 in the external medium. Superimposed current traces were scaled to the peak amplitude. (B and D) Top: Fraction of current remaining after 300-ms depolarization (r300), plotted as a function of step potential (mV). The f value is the difference between r300 relations of Ba2+ and Ca2+ at +10 mV and indicates the strength of CDI. Bottom: Normalized I-V relation taken from peak Ba2+ and Ca2+ currents evoked by voltage steps to the indicated potential (mV). n = 6 cells for both graphs from three independent experiments. (E) Effect of intracellular perfusion of diC8-PIP2 on CDI of Cav2.2 current in the presence of DN-CaM. diC8-PIP2 (200 μM) was added to the pipette solution. Recording was started 5 to 8 min after whole-cell configuration. Each trace is a representative of six from two independent experiments. (F) Top: Fraction of current remaining after 300-ms depolarization (r300) in diC8-PIP2–dialyzed cells, plotted as a function of step potential (mV). Bottom: Normalized I-V relation taken from peak Ba2+ and Ca2+ currents evoked by voltage steps to the indicated potential (mV). n = 6 cells from two independent experiments.

Cav2.2 channels with β2e exhibited similar voltage-dependent Ba2+ current as channels with β2a (Fig. 5, C and D). However, CDI of the Ca2+ current occurred with either wild-type CaM or DN-CaM, indicating that CDI of Cav2.2 channels with β2e was independent of CaM (Fig. 5, C and D). We also tested the Cav2.3 channel with either β2a or β2e in the presence of coexpressed wild-type or DN-CaM (fig. S5). Similarly to the Cav2.2 channels, the Cav2.3 channel with β2a exhibited a dependence on CaM for CDI (fig. S5A), whereas the channel with β2e did not (fig. S5B). Our finding that Ca2+ influx triggered the dissociation of β2e from the plasma membrane and channels with this β subunit exhibited fast inactivation is similar to previous studies (24, 28), which showed that the subcellular localization of β subunits is important for the inactivation of Cav channels.

We examined whether the β2e-mediated CDI resulting from an increase in intracellular Ca2+ depended on depletion of membrane PIP2 by perfusing diC8-PIP2 into the cells to replenish any hydrolyzed PIP2. CDI occurred with β2e-containing Cav2.2 channels in the presence of diC8-PIP2 and DN-CaM (Fig. 5, E and F), suggesting that the β2e-mediated CDI is independent of PIP2 depletion. To further examine whether changes in the abundance of PIP2 at the plasma membrane affected the Ca2+-induced β2e translocation to the cytosol, we expressed β2e-GFP in tsA201 cells with or without PIP5 kinase type Iγ (PIPKIγ), which augments membrane PIP2 concentration by phosphorylating PI4P, and stimulated the cells with ATP to activate purinergic receptors and increase intracellular Ca2+. ATP-stimulated cells induced the plasma membrane to cytosol translocation of β2e in the presence or absence of PIPKIγ (fig. S5, A and B).

Another pathway through which PIP2 can be altered is hydrolysis by PLC, some isoforms of which are activated by Ca2+. To determine whether depolarization-induced activation of Cav2.2 channels triggered a change in membrane PIP2, we coexpressed α1B, α2δ-1, β2e, and PH-RFP in tsA201 cells and then exposed the cells to 30 mM potassium to depolarize the cells and stimulate Ca2+ influx through the Cav2.2 channels. The distribution of PH-RFP in the cells was the same in control and depolarized cells (fig. S7A), which, we confirmed, produced an increase in Ca2+ (fig. S7B), indicating that the Ca2+ signal produced by influx through the Cav channel did not trigger PIP2 degradation in these cells. The experiments with DN-CaM and in which PIP2 was manipulated indicated that internal Ca2+ increase is sufficient to induce β2e translocation without requiring CaM activity or depletion of membrane PIP2.

Membrane tethering of β2e by Lyn abolishes inactivation by Ca2+ influx

To determine whether β2e translocation from the membrane was necessary for this β subunit to mediate CDI, we generated a fusion protein, Lyn-β2e, between β2e and the membrane-targeting sequence of the kinase Lyn. The membrane-targeting sequence of Lyn (Lyn11) was lipidated, thereby tethering the protein to the membrane (24). Therefore, we fused Lyn11 to the N terminus of β2e to anchor the protein to the membrane and confirmed that increasing intracellular Ca2+ with ionomycin application had no effect on the plasma membrane localization of Lyn-β2e-GFP (Fig. 6A). Electrophysiological analysis of tsA201 cells expressing α1B, α2δ1, and Lyn-β2e-GFP showed that membrane anchoring of β2e resulted in CDI that depended on CaM (Fig. 6B). This behavior of the Cav2.2 channel Lyn-β2e resembled that of channels with β2a (Fig. 5A).

Fig. 6 Lipidated β2e restores CaM-mediated CDI of Cav2.2 channels.

(A and C) Confocal images and quantification of membrane colocalization in cells expressing PH-RFP and Lyn-β2e-GFP (A) or N-del-β2e-GFP (C) without α1B and α2δ-1 subunits in response to 2 μM ionomycin. Scale bars, 5 μm. The colocalization coefficients were obtained from merged images of cells expressing PH-RFP and Lyn-β2e-GFP or N-del-β2e-GFP before and after ionomycin addition. n = 5 cells from three independent experiments for both. (B) CDI of Cav2.2 channels with Lyn-β2e-GFP. Top: Voltage protocol and representative whole-cell Ba2+ (black) and Ca2+ (red) currents elicited by 500-ms depolarization step to +10 mV in cells also expressing WT-CaM or DN-CaM. Middle: Fraction of current remaining after 300-ms depolarization (r300), plotted as a function of step potential (mV). Bottom: Normalized I-V relation taken from peak Ba2+ and Ca2+ currents evoked by voltage steps to the indicated potential (mV). For analysis, n = 5 cells from three independent experiments. (D) Representative currents of Cav2.2 channels with N-del-β2e-GFP in the presence of Ba2+ and Ca2+ in cells also expressing WT-CaM or DN-CaM. Currents were measured during 500-ms test pulse at +10 mV. Ca2+ current (red) was scaled up to Ba2+ current (black). Superimposed current traces were scaled to the peak amplitude. (E) Summary of inactivation time constants (τ) for Cav2.2 currents with Lyn-β2e-GFP (B) or N-del-β2e-GFP (D). Data are means ± SEM. n = 6 cells from three independent experiments. ***P < 0.001, with one-way ANOVA followed by Tukey’s post hoc test. Data are means ± SEM.

Although we did not detect an interaction between CaM and β2e (Fig. 3D), the effect of DN-CaM on Cav2.2 channels with β2a (Fig. 5A) or Lyn-β2e (Fig. 6B) suggested that CaM mediates CDI when the β subunit is anchored to the membrane by lipidation. We tested the importance of membrane association for regulatory function by analyzing the effect of the N-terminally deleted β2e (N-del-β2e), which cannot interact with the membrane but retains the interaction with the α1 subunit, on Cav2.2 conductance in the presence of wild-type or DN-CaM. Consistent with the requirement of the N terminus for membrane association, N-del-β2e-GFP localized in the cytosol in the absence of α1B and was unaffected by the ionomycin-induced increase in intracellular Ca2+ (Fig. 6C). Furthermore, Cav2.2 channels with N-del-β2e exhibited fast inactivation in the presence of either wild-type CaM or DN-CaM (Fig. 6, D and E), suggesting that membrane association was necessary for channel-regulating function. The functional role of the N terminus was recovered by tagging Lyn11 directly to the N-del-β2e (Lyn-N-del-β2e). Lyn-N-del-β2e localized to the plasma membrane without α1B (fig. S8A), and the Cav2.2 channels with this subunit displayed CaM-dependent inactivation (fig. S8B), suggesting that membrane tethering of β2e subunit through the native N terminus or Lyn11 is critical for regulating the channel inactivation.

Ca2+ influx augments PIP2 sensitivity to Cav channel

Although we did not find evidence for a requirement of PIP2 depletion in CDI mediated by β2e (Fig. 5E and figs. S6 and S7), channels with β2e may exhibit altered responses to PIP2 depletion. Previous studies reported that the subcellular distribution of the β subunit is a key determinant for Cav channel regulation by PIP2 (24, 46). Specifically, inhibition of Cav channels with β2a or β2e by PIP2 depletion is minor compared to that exhibited by channels with cytosolic β subunits, because membrane-associated β subunits may prevent differences in membrane lipids from affecting the pore-forming subunit directly (27). Because we observed that Ca2+ triggered the membrane dissociation of β2e, we assessed whether this affected the response of the Cav2.2 channel with this β subunit to PIP2 depletion. To selectively remove membrane PIP2, we used the rapamycin-inducible dimerization system, consisting of Lyn11-FRB (LDR) to target the FRB protein to the plasma membrane and cyan fluorescent protein CFP–FKBP-INP54p (CF-Inp), a fusion protein with a polyphosphoinositide 5-phosphatase (INP54p) that hydrolyzes PIP2 and is recruited to the FRB portion of LDR in response to rapamycin (47, 48). In cells expressing LDR, CF-Inp, and PH-GFP, rapamycin stimulated the membrane recruitment of CF-Inp and simultaneously induced the movement of PH-GFP to the cytosol, indicating that CF-Inp translocation to the membrane depleted the membrane of PIP2 (Fig. 7A). We then expressed LDR and CF-Inp in tsA201 cells coexpressing α1B, α2δ1, and β2e-GFP or N-del-β2e-GFP and monitored the Ba2+ current or the Ca2+ current in the presence or absence of rapamycin. Rapamycin-mediated depletion of PIP2 triggered a 16% reduction in the Ba2+ current and a 30% reduction in the Ca2+ current of Cav2.2 channels with β2e-GFP (Fig. 7, B and C). In contrast, Cav2.2 with N-del-β2e exhibited ~30% inhibition of both Ba2+ and Ca2+ currents in response to rapamycin-induced PIP2 depletion (Fig. 7, B and C). Thus, these results indicated that the membrane dissociation of β2e by Ca2+ influx increases the inhibition of Cav2.2 channels by PIP2 depletion.

Fig. 7 PIP2 depletion–mediated current inhibition predominates in the presence of extracellular Ca2+.

(A) Confocal gray scale images of a cell expressing CF-Inp, PH-PLCδ-GFP, and LDR (rapamycin-inducible system) and the response to 1 μM rapamycin (Rapa). Images before and after Rapa addition. Scale bar, 5 μm. Images are representative of 20 cells in six independent experiments. (B) Suppression of Ba2+ and Ca2+ currents by rapamycin-induced PIP2 depletion. Currents were measured in cells expressing the rapamycin-inducible system and Cav2.2 channels with β2e-GFP (top) or N-del-β2e-GFP (bottom) in the presence of Ba2+ or Ca2+ as the charge carrier. Currents were recorded at +10 mV every 4 s. Currents in insets indicate current traces before (a) and after (b) rapamycin addition. (C) Summary of Cav2.2 current inhibition by PIP2 depletion. For analysis, Ba2+ and Ca2+ currents before and after rapamycin application were measured and presented as percent inhibition by rapamycin-mediated PIP2 depletion. n = 7 cells from independent experiments. **P < 0.005, one-way ANOVA followed by Tukey’s post hoc test. Data are means ± SEM.

DISCUSSION

The elaborate inactivation mechanism of Cav channels is important for the tight regulation of the cytosolic Ca2+ concentration, not only because excess Ca2+ is toxic (49) but also because different patterns of Ca2+ signals encode different responses. In particular, the fine-tuning of inhibition of channel activity by β subunit isoforms critically restrains Ca2+ entry. Here, we report Ca2+-mediated control of the subcellular localization of the β2e subunit and how translocation of β2e away from the membrane modulates the gating properties of Cav channels (Fig. 8). Several lines of evidence suggested that Ca2+ modulated Cav channel gating by “screening” the membrane phospholipids and thereby promoting the subcellular redistribution of β2e. We also found that with Ca2+ but not Ba2+ in the extracellular medium, Cav channels with β2e exhibited inactivation even in the presence of Ca2+-insensitive DN-CaM, suggesting that CDI of Cav channels with β2e is governed by Ca2+ rather than by Ca2+/CaM, which is presently the predominant model for inactivation of Cav channels (45). In addition, with the rapamycin-induced dimerization system, we showed that PIP2 depletion produced a greater decrease in Ca2+ currents compared to Ba2+ currents in Cav channels with β2e. Together, the data suggested that intracellular Ca2+ regulates the sensitivity of the channel to changes in PIP2 by controlling the association of β2e with the plasma membrane.

Fig. 8 Schematic model of Ca2+-mediated feedback regulation of Cav channels with β2e.

In the resting state (left), β2e is tethered to the plasma membrane by an electrostatic interaction between N-terminal basic amino acid residues and anionic phospholipids. When the cells are depolarized or Gq-coupled GPCRs are activated (right), Ca2+ ions, which influx from the extracellular medium to the Cav channels or Ca2+ release–activated channels or are released from the ER, block β2e interaction with the plasma membrane by screening the phospholipid negative charges. The dissociation of β2e from the plasma membrane phospholipids promotes the inactivation of Cav currents independently of CaM and augments the PIP2 sensitivity of Cav channels. In this configuration, β2e subunit still remains stably bound to the I-II loop of the α1 subunit. [Ca2+]i, intracellular calcium concentration.

These results have several implications. From the receptor signaling perspective, we used purinergic and muscarinic receptors, which are GPCRs endogenously present in HEK293-derived tsA201 cells (31). In tsA201 cells transfected with M1R, the membrane distribution of the β2e subunit relies on membrane anionic phospholipids, such as PI4P and PIP2; thus, a cycle of their depletion and resynthesis could produce a reversible and transient translocation of the β2e from the membrane to the cytosol (28). However, our current study showed that application of agonists, such as ATP or Oxo-M, to activate endogenous receptors in tsA201 cells had no effect on the depletion of PIP2, which we monitored with the membrane-anchored PH-PLCδ probe. This lack of an effect on PIP2 is consistent with Falkenburger et al. (50). Nevertheless, the agonists triggered the translocation of the β2e subunit, suggesting that PI lipid depletion was not the cause of the translocation of β2e subunits. Like β2e, membrane-anchored proteins with charged domains that interact with the membrane through an electrostatic mechanism can be displaced from the membrane by Gq-coupled GPCR signaling, which can involve activation of protein kinases, lipid-metabolizing enyzmes, and CaM, due to a change of the charge of target molecules or the neutralization of negatively charged phospholipids (2). However, our results indicated that it is less likely that such proteins affected the distribution of β2e in the tsA201 cells.

These results turned our attention to the direct involvement of Ca2+ in the cytosolic translocation of plasma membrane–associated β2e. Analysis of other proteins that exhibit electrostatic membrane-lipid interactions has shown that intracellular-free Ca2+ ions shield membrane anionic phospholipids by binding to the phosphate groups, producing a change in the association of proteins with the membrane and thus regulating the physiological processes mediated by the proteins (16, 17). By manipulating cytosolic Ca2+, we showed fluctuations in the intracellular Ca2+ concentration that reversibly regulated the association of β2e with the plasma membrane. Furthermore, in peptide-liposome binding assays, Ca2+ and Mg2+ decreased the interaction between an N-terminal peptide of β2e and liposome, thus supporting that interpretation that dissociation of β2e from the plasma membrane is primarily due to the phospholipid charge-screening property of divalent cations. This result is consistent with a previous report showing that the rise in cytosolic Ca2+ concentration modulates protein-lipid interactions (17). Although it is technically difficult to measure the direct interaction between Ca2+ and anionic phospholipids, Shi et al. (17) showed by nuclear magnetic resonance that Ca2+ directly and reversibly binds to the phosphate groups of phospholipids.

From the perspective of Cav channel inactivation, CDI mediated by CaM is well established (5153). Although the extent of CDI depends on various factors, such as Ca2+ buffering, cellular or tissue context, regulatory β subunits, and type of Cav channel, a common feature of inactivation mediated by CaM is that it requires the Ca2+-binding lobes of CaM. CaM consists of two lobes, a C-terminal lobe and an N-terminal lobe, each of which binds two Ca2+ (52). Without Ca2+, bound CaM cannot inactivate Cav channels. Most studies of CDI use cells expressing β2a, which leads to slow inactivation compared to inactivation by other β subunits and thus enables the isolation of the effect of Ca2+/CaM on Cav channels (45). However, we found that the Cav channels with β2e exhibit fast inactivation even in the presence of a DN-CaM. This result suggested that robust Ca2+ influx through Cav channels negatively regulates the channel opening by disrupting a link between the plasma membrane and the N terminus of β2e. This type of fast inactivation resulting from disrupted interactions with the membrane is consistent with the previously proposed mechanism that disruption of the palmitoylation of β2a results in Cav channels with increased inactivation (54).

We observed an interesting but puzzling difference between Ca2+ and Ba2+ as charge carriers in mediating β2e translocation-dependent Cav inactivation. Although both have an identical charge, only Ca2+ disrupted the interaction between the plasma membrane and β2e. One answer is that Ca2+ transported through Cav channels interacts more tightly to membrane anionic phospholipids than Ba2+ does. Consistently, Ba2+ is insufficient for screening a negative membrane and thus does not trigger inactivation of the Cav channel (55, 56). Our electrophysiological analysis (Fig. 5) showed that in current-voltage (I-V) curves, compared with Ba2+ current, the peak current with Ca2+ occurred at a more depolarized potential, indicating that Ca2+ reduced membrane negativity through stronger binding to the lipids and thus required more depolarization to activate Cav channels. Again, this analysis is consistent with a previous study showing that the I-V of N-type Cav channels can be influenced by the types of charge carriers (57). In addition to fast inactivation triggered by the dissociation of β2e from the membrane, we found that the current inhibition of Cav2.2 channels with β2e by induced PIP2 depletion was increased by the influx of Ca2+, but not Ba2+. This result corroborates previous findings that cytosolic forms of the β subunit increase the PIP2 dependence of Cav channel for activity, whereas membrane-anchored β subunits attenuate it (24, 46, 58).

In conclusion, our findings expanded regulatory mechanisms of the β subunits on Cav channel activity by showing that the subcellular localization of β2e is dynamically altered by changes in cytosolic Ca2+ and that such a location change affects the gating property and GPCR-mediated modulation of Cav channels. In particular, the direct control of Cav channels by Ca2+ ions represents a simple Ca2+ feedback mechanism to set the gating property of the Cav channel. Considering that the β2e subunit is present in various brain tissues (28) and that intracellular Ca2+ concentrations near the channels could be on the order of several hundred micromolar (5961), it is possible that Cav channel regulation by β2e occurs in multiple areas of the brain. Furthermore, because our results indicated that the CDI mediated by β2e is independent from CaM, β2e could play a redundant role in CDI in cells with other β subunits that mediate CaM-dependent CDI. Our data indicated that lipid turnover would have a predominant effect on the interaction of β2e with the membrane when the PIP2-depleting effect of GPCR signaling is strong, whereas the Ca2+ shielding effect would predominate when the GPCR signal to PIP2 hydrolysis is weak. Therefore, further studies will elucidate the physiological consequences caused by such a regulatory mechanism in neurons and other excitable cells.

MATERIALS AND METHODS

Complementary DNA

Mouse β2e and N-terminally deleted β2e subunits were cloned from mouse brain complementary DNA (cDNA). Cloning of β2e-GFP, β2e-mCherry, and N-del-GFP was previously described (27, 28). For the site-directed mutagenesis of potential phosphorylation sites, point-mutated constructs for potential phosphorylation sites were obtained by polymerase chain reaction (PCR) using a QuikChange Site-Directed Mutagenesis Kit (Agilent Technologies). The following primers were used: T4A mutant, 5′-GGAATTCATGAAGGCCGCCTGGATCAGGCTTCT-3′ (sense) and 5′-AGAAGCCTGATCCAGGCGGCCTTCATGAATTCC-3′ (antisense); S19,20A mutant, 5′-GGGAGGAAGGCTGAAGGCTGCGGACATCTGT-3′ (sense) and 5′-ACAGATGTCCGCAGCCTTCAGCCTTCCTCCC-3′ (antisense). For Lyn-β2e-GFP and Lyn-N-del-β2e construct, Lyn was amplified from LDR using PCR and the following primers: 5′-CAAGCCGCTAGCATGGGATGTATAAAATC-3′ (forward) and 5′-ATCTCGAGAGAGCACTACCAGCACTACC-3′ (reverse). The PCR product was subcloned into β2e-GFP and N-del by Nhe I/Xho I sites.

Cell culture

tsA201 cells were maintained in Dulbecco’s modified Eagle’s medium (HyClone, Thermo Scientific) containing 10% fetal bovine serum (HyClone, Thermo Scientific) and 0.2% penicillin/streptomycin (HyClone, Thermo Scientific) at 37°C under 5% CO2. For transfection, cells were plated in 3.5-cm culture dishes at 70 to 80% confluency. For the expression of Cav channel subunits, cells were transiently cotransfected with various subunits using Lipofectamine 2000 or 3000 (Invitrogen) according to the manufacturer’s protocol. The transfected DNA mixture consisted of plasmids encoding α1, β, and α2δ-1 at a 1:1:1 molar ratio. When required, a plasmid encoding enhanced GFP was also included in the DNA mixtures as an indicator of transfected cells. Cells were plated on poly l-lysine–coated chips the day after transfection. Plated cells were used for a confocal microscope or current recording experiments within 1 to 2 days after transfection.

Electrophysiology

Whole-cell patch-clamp recordings were acquired as previously described (27, 28). Ba2+ currents or Ca2+ currents were recorded in whole-cell configurations using a HEKA EPC-10 patch-clamp amplifier with pulse software (HEKA Elektronik). For analysis, Fit Master software was also used. Pipettes were pulled from a glass micropipette capillary (World Precision Instruments) using a Flaming/Brown microscope puller model P-97 (Sutter Instrument Co.) and have resistances of 1.5 to 2.5 megohms when filled with internal solution. Series resistance errors were compensated for >60%, and fast and slow capacitance was compensated for before the applied test pulse sequences. Ca2+ or Ba2+ currents were sampled at 10 kHz and filtered at 3 kHz. For all recordings, cells were held at −80 mV. All data obtained were leak- and capacitance-subtracted before analysis. The external Ringer’s solution contained 160 mM NaCl, 10 mM BaCl2 or CaCl2, 1 mM MgCl2, 10 mM Hepes, and 8 mM glucose; pH was adjusted to 7.4 with NaOH. The internal solution of the pipette consisted of 175 mM CsCl, 5 mM MgCl2, 5 mM Hepes, 0.1 mM 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA), 3 mM Na2ATP, and 0.1 mM Na3GTP; the pH was adjusted to 7.5 with CsOH. diC8-PIP2 (Echelon Biosciences) was prepared as 1 mM stock in dimethyl sulfoxide (DMSO) and diluted 1:5 in internal solution. The reagents obtained were as follows: CsOH, BAPTA, Na2ATP, and Na3GTP (Sigma-Aldrich), and other chemicals (Merck).

FLAG-CaM immunoprecipitation and Western blotting

For FLAG immunoprecipitation, transfected cells were lysed with a buffer containing 1% Triton X-100, 50 mM tris-HCl (pH 7.5), 150 mM NaCl, and protease inhibitor cocktails (Roche). The resulting cell lysates were equally separated and added to 2 mM CaCl2 or 5 mM EGTA. The cell lysate was incubated with 10 μl of mouse anti–FLAG M2 agarose beads (Sigma-Aldrich) at 4°C for 2 hours. The beads were washed three times with lysis buffer. Beads were incubated with 100 μM 3× FLAG peptides (Sigma-Aldrich) containing lysis buffer for an additional 30 min at 4°C. Finally, the supernatant of the 3× FLAG-hCaM immunoprecipitate was mixed with 5× SDS sample buffer and analyzed by Western blotting. For Western blotting, proteins separated on 10% gels were transferred to a 0.45-μm nitrocellulose membrane. The membrane was blocked with 5% nonfat milk and incubated with the primary antibody in tris-buffered saline with Tween 20 buffer [25 mM tris-HCl (pH 7.4), 150 mM NaCl, and 0.01% Tween 20] containing 5% nonfat milk. The bound primary antibodies were detected using a goat anti-mouse or a goat anti-rabbit immunoglobulin G horseradish peroxidase–conjugated secondary antibody and an ECL detection system (Millipore). The rabbit polyclonal antibody against GFP and the mouse M2 antibody against FLAG were from Invitrogen and Sigma-Aldrich, respectively.

Assay for peptide-liposome binding

For the preparation of liposomes, the methods were previously described (27, 28). Briefly, the lipid mixture was dissolved using a chloroform/methanol mixture in the ratio of 2:1 and was dried under a gentle stream of nitrogen in the hood, thereby generating a lipid film. The film was then dissolved with 100-μl buffer containing 150 mM KCl, 20 mM Hepes/KOH (pH 7.4), and 5% sodium cholate. The lipids were passed over a size exclusion column to remove detergent [Sephadex G50 in 150 mM KCl and 20 mM Hepes (pH 7.4)]. Liposomes consisted of PC (l-α-phosphatidylcholine), PE (l-α-phosphatidylethanolamine), PS (l-α-phosphatidylserine), cholesterol, dansyl-PE, and PIP2 (44:10:15:25:5:1 mole percent) were collected as eluted (~400 μl). Liposomes were detected by ultraviolet, which shows the presence of dansyl-PE. In response to Ca2+ or Mg2+ addition, the binding of the peptide to liposomes was monitored using FRET measurements using a FluoroMax (HORIBA Jobin Yvon). Dansyl-PE incorporated in liposomes quenches the fluorescence of tryptophan in the peptide (62). Assays were performed at 37°C in 1 ml of buffer containing 150 mM KCl and 20 mM Hepes-KOH (pH 7.4). Various concentrations of CaCl2 or MgCl2 were added to the test buffer. The synthesized N-terminal peptide (750 nM) contained one tryptophan residue. The FRET signal was presented as F0/F, where F0 represents the fluorescence intensity at 355 nm before the liposome addition and F indicates the fluorescence intensity after the addition of various concentrations of CaCl2 or MgCl2 with the same amount of liposome.

Confocal imaging

All imaging was performed with a Carl Zeiss LSM 700 confocal microscope. For live cell imaging, a 40× (water) apochromatic objective lens at 1024 × 1024 pixels with digital zoom was used. For time courses, 524 × 524 pixels were used. To analyze the time course of the degree of cytosolic fluorescence of various proteins, images were taken every 5 s in ZEISS ZEN imaging software (every 3 s for cells loading Fluo4-AM). Time-course experiments of cytosolic fluorescence intensity of transfected probes and of Ca2+ signaling with Fluo4-AM were merged to a graph after each experiment was performed separately. Quantification of fluorescence intensity ratio of plasma membrane and cytosol was calculated from line intensity histograms of cells (shown as arrow in each figure) in the whole series of images after acquisition, as reported previously (63). Membrane localization was assessed by forming membrane/cytosol intensity ratio values from line intensity histograms of cells in the whole series of images after acquisition. Regions of interest were selected in the cytosolic regions of cells, and quantitative analysis was performed using the “profile“ and “measure“ tools in ZEN 2012 lite imaging software (Carl Zeiss MicroImaging). All confocal images were transferred from TIFF formats, and raw data from the time course were processed with Microsoft Office Excel 2012 and summarized in Igor Pro (WaveMetrics Inc.).

Chemicals

Oxo-M (Sigma-Aldrich) and ATP disodium salt hydrate (Sigma-Aldrich) were dissolved in sterilized water to make 10 and 50 mM solution stock, respectively. Each solution was diluted further to 1:1000 in standard Ringer’s solution for a working solution. Thapsigargin (Enzo) and ionomycin (Enzo) were prepared as 2 mM stock in DMSO and stored at −20°C, and working solutions were diluted further to 1:1000 in standard Ringer’s solution. Fluo4-AM (Invitrogen) and rapamycin (LC Laboratories) were prepared as 1 mM stocks in DMSO and diluted to 1:250 and 1:1000 in standard Ringer’s solution, respectively. PMA (Sigma-Aldrich) was prepared as 1 mM stock and diluted to 1:1000 in standard Ringer’s solution.

Statistical analysis

All data were analyzed using Excel (Microsoft), IGOR Pro 6.0 (WaveMetrics), or GraphPad Prism 6.0 (GraphPad Software). Statistics in text or figures represent mean ± SEM. Statistical comparisons were made by one-way or two-way ANOVA depending on the number of experimental groups followed by Tukey’s post hoc test. Differences were considered significant at the *P < 0.01, **P < 0.005, and ***P < 0.001.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/9/435/ra67/DC1

Fig. S1. Stimulation of endogenous purinergic receptors with ATP triggers β2e translocation to the cytosol.

Fig. S2. Oxo-M- and ATP-induced translocation of PH-RFP and β2e-GFP in cells transfected with M1R and P2Y2 receptors.

Fig. S3. IP3K-A prevents the ATP-induced translocation of β2e-mCherry.

Fig. S4. Phosphorylation is not a key factor for translocation of β2e.

Fig. S5. β2e-mediated CDI of Cav2.3 channels is independent of CaM.

Fig. S6. Purinergic activation induces translocation of β2e in PIPKIγ-expressing cells.

Fig. S7. Depolarization by 30 mM K+ has no effect on membrane PIP2 distribution.

Fig. S8. Lyn-tagged N-del-β2e is localized to the plasma membrane.

REFERENCES AND NOTES

Acknowledgments: We thank many laboratories for providing plasmids. The following cDNAs were given to us: rat Cav2.2 and α2δ-1 (gift from D. Lipscombe, Brown University, Providence, RI), Cav2.3 (from T. P. Snutch, University of British Columbia), rat β2a (from W. A. Catterall, University of Washington, Seattle, WA), PH-RFP (PLCδ1) (from K.Mackie, University of Washington, Seattle, WA), wild-type CaM and DN-CaM (both are non–FLAG-tagged; from T. Davis, University of Washington, Seattle, WA), rapamycin-inducible dimerization system CF-Inp and Lyn11-FRB (LDR) (from T. Inoue, Johns Hopkins University, MD), IP3K-A (from H. Kim, Seoul, College of Medicine, Korea University, Seoul, Korea), MARCKS (from H. Kim, Yonsei University College of Medicine, Seoul, Korea), 3× FLAG-hCaM (from D.-J.J., Kyungpook National University, Daegu, Korea), and PIPKIγ (from Y. Aikawa and T. F. Martin, University of Wisconsin, Madison, WI). Funding: This work was supported by the National Research Foundation of Korea grant funded by the Korea government (MSIP) (no. 2016R1A2B4014253), the DGIST R&D Program of the Ministry of Science, ICT and Future Planning (no. 16-BD-06), and the Korea Brain Research Institute (KBRI) basic research program funded by the Ministry of Science, ICT and Future Planning (no. 2231-415). Author contributions: B.-C.S. and D.-I.K. designed the research; D.-I.K., H.-J.K., D.-J.J., and Y.P. performed biochemical and electrophysiological research; and D.-I.K. and B.-C.S. wrote the paper. Competing interests: The authors declare that they have no competing interests. Data and materials availability: The plasmids related to β2e subunit require a material transfer agreement from DGIST, Korea.
View Abstract

Navigate This Article