Voltage-gated sodium (NaV) channels are densely expressed in most excitable cells and activate in response to depolarization, causing a rapid influx of Na+ ions that initiates the action potential. The voltage-dependent activation of NaV channels is followed almost instantaneously by fast inactivation, setting the refractory period of excitable tissues. The gating cycle of NaV channels is subject to tight regulation, with perturbations leading to a range of pathophysiological states. The gating properties of most ion channels are regulated by the membrane phospholipid, phosphatidylinositol (4,5) bisphosphate (PI(4,5)P2). However, it is not known whether PI(4,5)P2 modulates the activity of NaV channels. Here, we utilize optogenetics to activate specific, membrane-associated phosphoinositide (PI)-phosphatases that dephosphorylate PI(4,5)P2 while simultaneously recording NaV1.4 channel currents. We show that dephosphorylating PI(4,5)P2 left-shifts the voltage-dependent gating of NaV1.4 to more hyperpolarized membrane potentials, augments the late current that persists after fast inactivation, and speeds the rate at which channels recover from fast inactivation. These effects are opposed by exogenous diC8PI(4,5)P2. We provide evidence that PI(4,5)P2 is a negative regulator that tunes the gating behavior of NaV1.4 channels.
Introduction
NaV channels initiate and propagate action potentials in excitable cells. Their characteristic gating cycle, including voltage-dependent activation followed instantaneously by fast inactivation, serves as a “timer” to reset the membrane potential, ensuring the generation of recurrent action potentials. Humans express nine NaV channels isoforms (NaV1.1–1.9) that are distinguished by differences in their gating kinetics, tissue distribution, and affinity for the pore-blocker tetrodotoxin (TTX; Goldin et al., 2000; Yu et al., 2005). Each NaV channel is composed of a pore-forming α subunit that associates with one or two β subunits (Alexander et al., 2019). The α subunit has four homologous domains (DI–DIV), each comprising six transmembrane α-helical segments (S1–S6) that incorporate an asymmetric loop located between S5 and S6, which lines the inner channel pore and forms the activation gate. S1–S4 form the voltage-sensing domain (VSD), with S4 carrying voltage-sensing residues, typically positively charged arginines in every third position that move in response to depolarization to gate the channel (Catterall et al., 2020). Domains I–III predominantly participate in the process of activation while DIII and DIV (Cha et al., 1999), along with the DIII–DIV linker, are crucial to the mechanism of fast inactivation (Aldrich et al., 1983; Aldrich and Stevens, 1983; Vandenberg and Horn, 1984; Goldschen-Ohm et al., 2013; Wisedchaisri et al., 2019; Catterall et al., 2020; Jiang et al., 2021).
All aspects of NaV channel gating are subject to stringent regulation to maintain the function of excitable tissues (Lehmann-Horn et al., 1983; Ptacek et al., 1991; Mantegazza et al., 2021; Fouda et al., 2022). NaV channelopathies caused by mutations, aberrant cell signaling pathways, or the off-target actions of drugs typically change the magnitude of Na+ currents. However, the underlying gating dynamics can be complex, in part because mechanical coupling between the domains of the NaV channel means that dysregulation in fast inactivation also shapes activation (Aldrich et al., 1983; Aldrich and Stevens, 1983; Vandenberg and Horn, 1984; Goldschen-Ohm et al., 2013; Wisedchaisri et al., 2019; Jiang et al., 2021). Changes in gating behavior often lead to an increase in the magnitude of the current that persists after fast inactivation (called ILATE), which contributes to the plateau phase of the cardiac action potential (Plant et al., 2006; Makielski, 2016; Plant et al., 2020), resurgent excitability in neurons (Crill, 1996; Raman and Bean, 1997), and the excitability of skeletal muscle fibers (Cannon, 2015). Therefore, even a small increase in ILATE can precipitate diseases including cardiac arrhythmias, pain syndromes, epilepsy, and disorders of skeletal muscle contractility (Lehmann-Horn et al., 1983; Ptacek et al., 1991; Plant et al., 2006; Mantegazza et al., 2021; Fouda et al., 2022).
The phospholipid PI(4,5)P2 is a minor constituent of eukaryotic plasma membranes that is required for the physiological activity of most types of ion channels (Suh and Hille, 2008; Gada and Logothetis, 2022). PI(4,5)P2 is composed of a polar myo-inositol headgroup, coupled via a glycerol-phosphodiester linkage to arachidonic and stearic acid chains that permeate the inner leaflet of the membrane (Balla, 2013). The headgroup carries negatively charged phosphates at the 4 and 5 positions that interact with basic residues on partner proteins with a range of functional consequences (Dickson and Hille, 2019). PI(4,5)P2 is necessary in the gating process of many channel types, but also acts as an allosteric nexus that can regulate channel function by mediating the effects of drugs, post translational modifications, or protein partners (Liang et al., 2014; Xu et al., 2020; Gada and Logothetis, 2022; Xu et al., 2022; Gada et al., 2023a). PI(4,5)P2 levels are regulated by the activities of specific kinases and PI-phosphatases, which contribute to a dynamic equilibrium that modulates cell signaling and excitable behavior.
Although PI(4,5)P2 has emerged as a master regulator of ion channel function, its role in the operation of NaV channel activity is unknown. Here, using optogenetic tools to activate specific membrane PI-phosphatases, we show that PI(4,5)P2 plays a role in the physiological gating behavior of NaV1.4 channels. Dephosphorylating PI(4,5)P2 left-shifts the voltage-dependence of channel activation, speeds recovery from fast inactivation, slows the rate of fast inactivation, and augments ILATE. Our data support the conclusion that PI(4,5)P2 levels tune the activity of NaV1.4 channels.
Materials and methods
Reagents and molecular biology
Purified diC8PI(4,5)P2 was purchased from Echelon Biosciences. For oocyte studies, plasmid vectors were linearized after the Xenopus laevis β-globulin domain of each construct, verified by gel electrophoresis, and transcribed in vitro using the mMESSAGE mMACHINE T7 Transcription Kit (Thermo Fisher Scientific) to generate cRNA, according to the manufacturer’s protocol. Rat NaV1.4 (SCN4A; GenBank accession no. NM_013178.2) was handled in pBud, a CMV-driven vector, and was a gift from the Chanda lab (Washington University, St. Louis, MO). To study NaV1.4 currents, the plasmid was used in concert with NaVβ1 (SCN1B, isoform b; GenBank accession no. NM_001037) in pRAT, a CMV-driven vector, as before (Plant et al., 2006). The inactivation-deficient mutant NaV1.4-WCW was handled in pCap and was a gift from the Lingle lab (Washington University). CIBN-CAAX, mCherry-CRY2-5POCRL, and iRFP-PHPLCδ1 were gifts from the De Camilli lab (Yale University, New Haven, CT). CRY2-Sac2 was a gift from the Idevall lab (Uppsala Universitet, Uppsala, Sweden), and was subcloned into pMAX(+) for the generation of cRNA for use in Xenopus oocytes. The open reading frame of pseudojanin (Hammond et al., 2012) was designed with an N-terminal CRY2-tag in the pMAX(+) vector and was generated by Genscript.
Cell culture
Human embryonic kidney (HEK293T) cells were acquired from American Type Culture Collection (Cat #CRL-3216; ATCC) and were maintained in Dulbecco’s modified Eagle’s medium (ATCC) supplemented with 100 U/ml penicillin, 100 μg/ml streptomycin, and 10% (vol/vol) fetal bovine serum. Cells were routinely confirmed to be mycoplasma free via PCR and Hoescht stain. The cells were incubated in a 37°C, humified incubator supplemented with 5% CO2. For our experiments, cells were seeded on glass coverslips in 35-mm culture dishes at least 1 d before transfection. Cells were transiently transfected with rNaV1.4, NaVβ1, CIBN-CAAX, and CRY2-5-PtaseOCRL or CRY2-pseudojanin, along with the near-infrared PI(4,5)P2 biosensor iRFP-PHPLCδ1 as indicated, in OptiMEM using polyethylenimine (PEI) for 1–2 h at a ratio of 1 μg of DNA to 4 μl PEI. The near red fluorescent protein, mCherry, was typically used as a transfection marker. The cells were studied 24–48 h after transfection.
Total internal reflection fluorescence microscopy (TIRFM)
HEK293T cells seeded on 35-mm, glass-bottomed petri dishes were transfected using PEI (1:4) with 1 μg iRFP-PHPLCδ1, and 0.75 μg of CRY2-PJ/CRY2-5POCRL, with 0.75 µg CIBN-CAAX 48-h before the experiments. Blue-light-activation of CRY2 phosphatases was performed using a 445-nm laser, and iRFP was excited by a 647-nm laser (OBIS, Coherent) in an imaging solution comprising (in mM) 130 NaCl, 4 KCl, 1.2 MgCl2, 2 CaCl2, and 10 HEPES; pH was adjusted to 7.4 with NaOH. Cells were illuminated at 647 nm in TIRF mode to visualize iRFP and at 445 nm to activate the optogenetic machinery. Data were collected at 5 s intervals with a 10 ms exposure time to minimize photobleaching and were saved as stacked TIFF images. Upstream of the laser clean-up filters, the beams were conditioned for coherence using home-made Keplerian beam expanders. A high numerical-aperture apochromatic objective (60×, 1.5 NA; Olympus) installed on an RM21 microscope frame with a piezo-driven nano-positioning stage was illuminated by laser lines calibrated to provide 10 mW of incident light (MadCity Labs; Gada et al., 2023b). The fluorescence from iRFP was captured using a back illuminated sCMOS camera from Teledyne Photometrics. The camera and the lasers were controlled using the free software Micro-Manager (University of California, San Francisco). Tetraspec beads (Thermo Fisher Scientific) were used to calibrate the evanescent field depth to 100 nm and map the sCMOS chip. We used ImageJ to analyze the TIRF data.
Electrophysiology
Whole-cell patch clamp
HEK293T cells were studied by whole cell patch-clamp. Currents were recorded with a Tecella Pico-2 amplifier (Tecella) controlled using WinWCP software (University of Strathclyde). Currents were low-pass (Bessel) filtered at 9 kHz and digitized at 50 kHz. Patch pipettes were pulled from borosilicate glass (Clark Kent), using a vertical puller (Narishige) and had a resistance of 2.5–4 MΩ when filled with internal solution containing (in mM) 60 CsCl, 80 CsF, 2 MgCl2.6H2O, 10 EGTA, 5 HEPES, and 5 Na2ATP. The cells were perfused using a multichannel gravity-driven perfusion manifold (Warner) with an extracellular solution containing (in mM) 130 NaCl, 5 CsCl, 1.2 MgCl2.6H2O, 1.5 CaCl2, 8 glucose, and 10 HEPES. Data were collected from cells with a series resistance <15 MΩ. The junction potential was <3 mV and was not compensated.
Current–voltage (I–V) relationships were evoked from a holding potential of −90 mV by 400 ms test pulses between −80 and 50 mV, in 5 mV increments. Steady-state inactivation (SSI) was studied by holding the cells at −140 mV and then comparing currents evoked by 50-ms test pulses between −140 and 30 mV to those evoked by a 100-ms prepulse to 0 ms. A 10-s interpulse interval was used in both cases. Normalized peak current values are plotted against the prepulse potential (mV). To quantify voltage-dependence, the data were normalized, plotted against the driving force to generate normalized conductance–voltage (G–V) relationships that were fit using a Boltzmann function, where IMAX is the maximum current and K is the slope factor. Recovery from fast inactivation was studied by holding cells at −100 mV and comparing currents evoked by a pair of 50-ms test step to −30 mV separated by an interpulse interval that increased in duration by 5 ms increments per sweep. The time constant for recovery from inactivation (τinact) was obtained from mono-exponential fits of the normalized current amplitude to the recovery time using where A is the amplitude of components and t is time. Whole-cell currents were normalized to cell capacitance. Mean ± SEM capacitance values were 12 ± 4 pF for HEK293T cells.
For simultaneous optogenetic-patch clamp studies, the blue light system was photoactivated in epifluorescence-mode using continuous excitation from a broad-spectrum LED (Excelitas) through a 448/20 nm filter (Chroma), via a 20× objective lens (Olympus). The light output at the sample was measured at 50 mW/cm2 by a photometer (ThorLabs). To avoid pre-activation of CRY2-fusion proteins, cells were incubated in the dark and handled in foil-wrapped dishes prior to the experiments. Cells of interest were identified using mCherry as a transfection marker because its spectral properties fall beyond the activation range of CRY2. Further, we use a transilluminator that includes a bandpass filter to block blue light during brightfield visualization. To generate paired data, the currents were recorded first in the dark and again following a 3-min pre-illumination with epifluorescent blue light that persisted through the remainder of the study. All the experiments were performed at room temperature.
Two-electrode voltage clamp (TEVC)
Oocytes were isolated from Xenopus frogs and surgically isolated, dissociated, and defolliculated using collagenase according to standard protocols. Isolated oocytes were transferred to an Oocyte Ringer’s solution supplemented with Ca2+ and penicillin/streptomycin antibiotics. Oocytes were each selected and injected with 50 nl of cRNA resuspended in RNAse-free water. After injection, oocytes were incubated at 17°C prior to TEVC experiments. Electrodes were pulled from borosilicate glass capillaries using a Flaming-Brown micropipette puller (Sutter Instruments) and filled with a 3M KCl solution containing 1.5% agarose. Pipette resistances were between 0.2 and 1.0 MΩ. Currents were recorded 2 d after injection using a GeneClamp 500 (Molecular Devices) or an OC-725D (Warner) amplifier. Oocytes were perfused with a gravity-driven apparatus with a physiological buffer containing (in mM) 2 KCl, 96 NaCl, 1 MgCl2, and 5 HEPES, buffered to pH 7.4 with NaOH. Blue light was generated using a 470 nm LED (Luminus) that was focused on the oocytes using a collimating lens, as previously described (Gada et al., 2023c). The incident light was powered to 5 mW/cm2, as determined by a photometer (ThorLabs). Currents were evoked from a holding potential of −90 mV by 400 ms step depolarizations between −80 and 50 mV with 5 mV increments and an interpulse interval of 7.5 s. To quantify voltage-dependence, the data were normalized, plotted against the driving force to generate normalized G–V relationships that were fit using a Boltzmann function, where IMAX is the maximum current and K is the slope factor. To generate paired-data sets, the currents were recorded before and again following a 3-min pre-illumination period with continuous illumination through the remainder of the experiment.
Statistics
Data were handled in WinWCP, Clampfit, and Excel software with statistical analysis performed using GraphPad (Prism). The data are presented as mean ± standard error (SEM) with statistical differences between paired groups determined by two-tailed, paired, Student’s t test, unless indicated otherwise. The threshold for significance was determined to be P < 0.05. We used Biorender to generate the schematic diagrams.
Results
Dephosphorylation of PI(4,5)P2 modifies the gating parameters of NaV1.4 channels
To study the role of PI(4,5)P2 in the activity of NaV1.4 channels, we combined whole-cell patch-clamp recording with photoactivation of membrane targeted PI-phosphatases that dephosphorylate PI(4,5)P2. This robust optogenetic platform is based on blue light (∼440–490 nm) photoactivation of Cryptochrome 2 (CRY2) and its protein partner, CIBN (Liu et al., 2008). Fusing CIBN to the membrane anchor CAAX targets CRY2 and its cargo to the inner leaflet of the cell membrane in response to photostimulation (Kennedy et al., 2010; Idevall-Hagren et al., 2012; Gada and Logothetis, 2022). Prior studies using real-time measurements of PI(4,5)P2-regulated channel activity, or PI(4,5)P2 biosensors such as the PH-domain of PLCδ, as reporters (Cui et al., 2022) have shown that the membrane localization of CRY2-fused PI-phosphatases, and subsequent dephosphorylation of PI(4,5)P2 occur within tens of seconds of photoactivation. Although PI(4,5)P2 can be regenerated within minutes, this effect can be mitigated by continued photostimulation (Idevall-Hagren et al., 2012; Xu et al., 2020; Ningoo et al., 2021; Xu et al., 2022).
First, we studied decoupling of the PI(4,5)P2 biosensor iRFP-PHPLCδ1 from the plasma membrane of HEK293T cells following photoactivation of CRY2-fused pseudojanin (CRY2-PJ). Pseudojanin is a fusion construct that encompasses the 5-phosphatase inositol polyphosphate 5-phosphatase E and the 4-phosphatase Sac1, depleting PI(4,5)P2 by generating PI(4)P and then PI (Hammond et al., 2012; Fig. 1 A). Using TIRFM we determined that iRFP-PHPLCδ1 decouples from the membrane rapidly in response to blue-light (BL) illumination and completely within ∼225 s (Fig. 1, B and C). These findings are consistent with our prior observations and the work of others using iRFP-PHPLCδ1 (Idevall-Hagren et al., 2012; Gada et al., 2022).
Using whole-cell patch-clamp, we studied HEK293T cells expressing rat NaV1.4, NaVβ1, CRY2-PJ, and CIBN-CAAX in HEK293T, and compared the currents evoked by a step depolarization protocol before and after photoactivation. Under control conditions, NaV1.4 currents first activated at −50 mV and reached a peak current-density of −312 ± 23 pA/pF at −30 mV (data are mean ± SEM). BL illumination evoked a ∼10 mV leftward shift in the I–V relationship, such that robust activation of NaV1.4 channels was first observed at ∼−55 mV. The current-density peaked at −40 mV and was ∼32% larger than the control at that potential (Fig. 1 D and Table 1). To interrogate this effect further, the data were replotted against the driving force, normalized, and fit with a Boltzmann function to obtain a G–V relationship. Photostimulation of CRY2-PJ caused a −10 ± 2 mV leftward shift in the mean half-maximal activation voltage of NaV1.4 (V½act) without a change in the slope factor, Kact (Fig. 2 A and Table 2).
Next, we studied the effect of PI(4,5)P2 dephosphorylation on the voltage-dependence of SSI using a paired-pulse protocol, where the prepulse was increasingly depolarized and the test-pulse was of fixed amplitude. Activation of CRY2-PJ did not alter the half-maximal voltage-dependence of SSI (V½SSI) but flattened the slope of the curve (KSSI) such that shifts in SSI were more pronounced when the test potential was more depolarized than V½SSI (Fig. 2 A and Table 2). Notably, the changes in V½act and KSSI augmented the magnitude of the “window” current between the voltage-dependent activation and SSI gating processes of NaV1.4 channels (Fig. 2 A, inset).
DiC8PI(4,5)P2 opposes the effects of CRY2-PJ dephosphorylation
Given that dephosphorylating PI(4,5)P2 evoked a left shift in the V½act of NaV1.4 gating, we sought to study the channels in a PI(4,5)P2 enriched environment. To accomplish this, we incorporated a high concentration of the soluble, short-chain PI(4,5)P2 analog diC8PI(4,5)P2 in the patch pipette. Treating cells with 200 µM diC8PI(4,5)P2 evoked a ∼5.5 mV right-shift in V½act and a ∼3 mV right-shift in V½SSI compared to untreated cells, without changing Kact or KSSI (Fig. 2, B and C; and Table 2). In addition, diC8PI(4,5)P2 ameliorated the effects of CRY2-PJ photoactivation on V½act and abolished the increase in the magnitude of the window current (Fig. 2 B, inset). In total, the change in mean V½act between cells treated with 200 µM diC8PI(4,5)P2 and control cells studied following dephosphorylation of endogenous PI(4,5)P2 by CRY2-PJ was −16 ± 2 mV, with a −5.4 ± 2 mV shift in the mean V½SSI (Fig. 2 C). Together, these data indicate that the gating parameters of NaV1.4 channels are modulated by the level of PI(4,5)P2 in the membrane.
DiC8PI(4,5)P2 opposes the effects of CRY2-5POCRL dephosphorylation
Under physiological conditions, dephosphorylation of PI(4,5)P2 occurs primarily via the action of inositol 5-phosphatases, generating PI(4)P (Logothetis et al., 2015). To study the effects of dephosphorylating PI(4,5)P2 at the 5-position, we used CRY2-5POCRL, a well-established optogenetic version of the inositol 5-phosphatase region of Lowe’s oculocerebrorenal protein, OCRL (Fig. 3 A; Idevall-Hagren et al., 2012; Xu et al., 2020; Ningoo et al., 2021; Xu et al., 2022; Gada et al., 2023a; Gada et al., 2023c). TIRFM studies with iRFP-PHPLCδ1 showed that photoactivation of CRY2-5POCRL evoked rapid redistribution of the PI(4,5)P2-biosensor away from the cell membrane, with total depletion occurring within ∼225 s (Fig. 3, B and C). These findings are consistent with the data we obtained using CRY2-PJ and prior reports on CRY2-5POCRL (Idevall-Hagren et al., 2012; Xu et al., 2022; Gada et al., 2023c).
Like CRY2-PJ, photoactivation of CRY2-5POCRL caused NaV1.4 currents to activate ∼5 mV earlier than control cells and augmented the peak current density by ∼10% (Fig. 3 D). This effect translated to a 7.5 ± 3 mV left-shift in the V½act of NaV1.4 and augmented the window current (Fig. 4 A, and Tables 1 and 2). Neither Kact, V½SSI, or KSSI were altered by CRY2-5POCRL. As observed for our studies with CRY2-PJ, including 200 µM diC8PI(4,5)P2 in the recording pipette right shifted V½act by ∼5 mV and precluded the effects of CRY2-5POCRL on the voltage-dependent gating of NaV1.4 channels (Fig. 4 B). The overall excursion in the mean V½act between cells treated with 200 µM diC8PI(4,5)P2 and control cells recorded following photoactivation of CRY2-5POCRL was −12.5 ± 3 mV; 75% of the range observed with CRY2-PJ (Fig. 4 C and Table 2).
Recovery from fast inactivation of NaV1.4 channels is speeded by dephosphorylation of PI(4,5)P2
In addition to the voltage-dependence of channel gating, cellular excitability is shaped by the rate at which NaV channels recover from fast inactivation and become available to initiate an action potential. To test whether dephosphorylation of PI(4,5)P2 alters this key gating parameter, we studied the NaV1.4 channels using paired steps to −30 mV and incrementally increased the duration between the pulses. Both dephosphorylation by CRY2-PJ and CRY2-5POCRL significantly expedited the rate at which NaV1.4 channels recovered from fast inactivation (τrecovery), compared to the control (Fig. 5, A and C, respectively). Including 200 µM diC8PI(4,5)P2 in the pipette precluded the speeding of τrecovery observed when either PI-phosphatase was activated (Fig. 5, B and D, respectively). Together these data indicate that the presence of PI(4,5)P2 in the cell membrane is necessary to maintain the physiological rate at which NaV1.4 channels recover from fast inactivation.
Dephosphorylating PI(4,5)P2 slows fast inactivation and augments ILATE
Our study of voltage-dependent activation and SSI gating revealed that dephosphorylation of PI(4,5)P2 increased the magnitude of the NaV1.4 window current (Fig. 2 A, inset; Fig. 4 A, inset), suggesting an increased probability of channels remaining in the open state at hyperpolarized voltages. Such changes to the biophysical properties of an NaV channel are expected to manifest an increased ILATE that persists following fast inactivation of the peak current. To study the effect of PI(4,5)P2 dephosphorylation on ILATE, we recorded currents evoked by a 400 ms depolarizing pulse to −20 mV before and after illumination with blue light. Photoactivation of CRY2-PJ augmented ILATE by eightfold from control levels of 0.24–∼2% of the peak current (Fig. 6, A and B; and Table 3). We also assessed the rate of fast inactivation of the peak current by fitting the data to a Boltzmann function and extracting tau (τinact) values and determined that photoactivation of CRY2-PJ prolonged τinact from 4.7 ± 0.3 to 9 ± 1 ms (Fig. 6 C and Table 3). Photoactivation of CRY2-5POCRL also evoked an eightfold increase in ILATE and a significant prolongation of τinact (Fig. 6 C and Table 3). Including 200 µM diC8PI(4,5)P2 in the recording pipette ameliorated the impact of both CRY2-PJ and CRY2-5POCRL on ILATE and τinact (Fig. 6, C and E; and Table 3). These findings suggest that intact PI(4,5)P2 modulates the rate and extent of the fast inactivation gating process that transitions the channels from the open to the inactive state.
PI(4,5)P2-dependent activation gating requires the fast inactivation process
Because of the mechanical linkage between the four consecutive domains of NaV channel α-subunits, changes in fast inactivation can also influence the activation process. Therefore, we studied the effects of PI(4,5)P2 dephosphorylation on an NaV1.4 mutant that abolishes fast inactivation. NaV1.4-L435W-L437C-A438W (NaV1.4-WCW) shows robust expression in Xenopus oocytes and has been reported to disrupt the transitions from activation to inactivation (Goldschen-Ohm et al., 2013). We expressed NaV1.4-WCW and NaVβ1, CIBN-CAAX, and CRY2-5POCRL, with or without the 4-phosphatase CRY2-Sac2, as indicated and studied the effect of PI(4,5)P2 on the currents in oocytes, as before (Gada et al., 2023c). Here, we co-express CRY2-5POCRL and CRY2-Sac2 to generate PI, like with CRY2-PJ in mammalian cells. NaV1.4-WCW passed non-inactivating, voltage-dependent currents that reached a peak at approximately −10 mV (Fig. 7 A). CRY2-5POCRL and CRY2-Sac2 did not impact the kinetics, magnitude, or the V½act of the current (Fig. 7, A and B). Similarly, photoactivation of CRY2-5POCRL studied alone did not change the activity of NaV1.4-WCW (Fig. 7 C). Together, these data suggest that the impact of PI(4,5)P2 on the voltage-dependence of NaV1.4 activation requires mechanical coupling to the fast inactivation process and that depleting PI(4,5)P2 uncouples these key gating processes.
Discussion
We show that PI(4,5)P2 is integral to multiple aspects of NaV1.4 channel gating, including regulating the voltage-dependence of activation, the rate of fast inactivation, the rate of recovery from fast inactivation, and the magnitude of the persistent current, ILATE. To the best of our knowledge, this study is the first to reveal a role for PI(4,5)P2 in the function of NaV channels. We find that depleting PI(4,5)P2 augments the overall activity of NaV1.4, while enriching PI(4,5)P2 levels right shifts the voltage-dependence of activation. Together, these findings show that PI(4,5)P2 is negative modulator of NaV1.4 and suggest that PI(4,5)P2 levels at the membrane tune the activity of the channel.
PI(4,5)P2 is an established, inextricable cofactor in the gating of many types of ion channels, adopting a broad range of regulatory roles that vary with the structure and function of the channel under investigation (Suh and Hille, 2008; Gada and Logothetis, 2022). The role of PI(4,5)P2 has been most extensively explored for inward rectifying potassium (Kir) channels (Sui et al., 1998; Rohacs et al., 1999; Hansen et al., 2011; Logothetis et al., 2015). However, within the ion channel superfamily, voltage-gated calcium (CaV) channels are the most structurally homologous to NaV channels (Catterall et al., 2020). Extensive functional studies show that CaV channels exhibit dual regulation by PI(4,5)P2. Thus, PI(4,5)P2 stabilizes CaV channel currents by reducing “rundown” while dephosphorylation of PI(4,5)P2 results in a leftward shift in the voltage-dependence of channel activation and inactivation, in an isoform-dependent manner (Wu et al., 2002; Suh et al., 2010). Recently, structural analysis demonstrated that PI(4,5)P2 associates with DII of CaV2.2, stabilizing the VSD in a “down”-conformation (Gao et al., 2021). PI(4,5)P2 also impedes the rundown of voltage-gated potassium (KV) channels and left-shifts the G–V relationship of KV1 and KV7 channels via interaction between PI(4,5)P2 and the S4–S5 linker in the VSD of each protomer (Zhang et al., 2003; Rodriguez-Menchaca et al., 2012a; Rodríguez Menchaca et al., 2012b; Sun and MacKinnon, 2020). Similarly, we find that dephosphorylation of PI(4,5)P2 left-shifts the voltage-dependent gating of NaV1.4 channels.
Although NaV channels are not typically associated with rundown, this phenomenon has been reported to occur for some NaV isoforms during longer recordings protocols (Hampl et al., 2016), or when channels are studied in off cell patches (Morris and Juranka, 2007). Current rundown has been attributed to the mechanosensitivity of NaV channels or the accumulation of channels in an inactivated state. However, in other ion channel families, rundown indicates a progressive loss of PI(4,5)P2 from the local membrane environment (Rodríguez Menchaca et al., 2012b). While we do not observe overt rundown of NaV1.4 channels, it is notable that the V½act is right shifted when a high concentration of diC8PI(4,5)P2 is included in the recording pipette (Figs. 2 and 4). In general terms, the extent of channel rundown will depend on the affinity of the channel for PI(4,5)P2, the abundance of PI(4,5)P2 in the experimental system, and the ability of other PI-species to act as surrogate. This final parameter has been extensively studied for Kir channels with some members of the family, such as Kir2.1 showing exquisite selectivity for PI(4,5)P2 while gating of other Kir channels, such as Kir6.2 can be supported by multiple PI-species (Rohacs et al., 2003). Our TIRF data show that CRY2-PJ and CRY2-5POCRL are equally effective at decoupling iRPF-PHPLCδ from the cell membrane (Figs. 1 and 3) but that the leftward shift in V½act is ∼25% (−2.5 mV) less for CRY2-5POCRL than CRY2-PJ. Given that CRY2-5POCRL generates PI(4)P and CRY2-PJ generates PI, this apparent difference in V½act might reflect that some aspects channel gating can be supported, at least in part, by PI(4)P.
In skeletal muscle cells, a left shift in the voltage dependence of NaV channel gating is expected to reduce the threshold for an action potential and change firing patterns. Given mechanical coupling between the domains of NaV channels, this excursion in the voltage dependence of activation could also reflect changes in fast inactivation (Aldrich et al., 1983; Goldschen-Ohm et al., 2013). Indeed, our studies with NaV1.4-WCW show that depleting PI(4,5)P2 has limited impact on the G–V relationship of the channel in the absence of fast inactivation (Fig. 7). This finding suggests that PI(4,5)P2 might play a role in the coupling between these processes. Similarly, the increased window current that results from PI(4,5)P2 dephosphorylation suggests that there is a shift in equilibrium from inactivated to the open state of the channel which is corroborated by the increase in ILATE. At the cellular level, the speeded rate of recovery from fast inactivation adds to this paradigm, enabling excitable cells to fire at a higher frequency, effectively shortening the refractory period. These data coalesce to predict a pattern of altered excitability in a PI(4,5)P2 deficient environment, placing NaV1.4 channels under the direct oversight of PI(4,5)P2 levels in the membrane. Although multiple disease-related mutations evoke similar dysregulation of NaV channel gating, whether dephosphorylation or hydrolysis of PI(4,5)P2 plays a role in NaV channelopathies is yet to be determined.
Together, the data presented in this study show that PI(4,5)P2 acts as a negative regulator of NaV1.4 function. Thus, dephosphorylation of PI(4,5)P2 is expected to manifest detrimental changes in excitability. Our findings support the need for a structural understanding of NaV1.4–PI(4,5)P2 association to delineate specific sites of interaction and to further determine the role of PI(4,5)P2 in NaV channelopathies.
Data availability
The data generated in this study are included in the article.
Acknowledgments
Joseph A. Mindell served as editor.
The authors thank Dr. Diomedes Logothetis at Northeastern University for helpful comments on the manuscript. We are grateful to Heikki Vaananen for oocyte preparation. Rat NaV1.4 was a gift from the Chanda lab (Washington University) and NaV1.4-WCW was a gift from the Lingle lab (Washington University). CIBN-CAAX, mCherry-CRY2-5′ptaseOCRL and iRFP-PHPLCδ1 were gifts from the De Camilli lab (Yale University). We thank Andrew Zorn and Dr. Takeharu Kawano for the design and generation of CRY2-pseudojanin for use in these studies.
The work was funded by National Institutes of Health Heart, Lung, and Blood Institute grant R01HL144615 to L.D. Plant.
Author contributions: Conceptualization, visualization, writing: K.D. Gada and L.D. Plant. Investigation: K.D. Gada, J.M. Kamuene, A. Chandrashekar, R.C. Kissell, and A.K. Yauch. All authors designed experiments and analyzed the data. Resources and funding acquisition: L.D. Plant.
References
Author notes
Disclosures: The authors declare no competing interests exist.