Nonmuscle myosin II isoforms interact with sodium channel alpha subunits.

Sodium channels play pivotal roles in health and diseases due to their ability to control cellular excitability. The pore-forming α-subunits (sodium channel alpha subunits) of the voltage-sensitive channels (i.e., Nav1.1-1.9) and the nonvoltage-dependent channel (i.e., Nax) share a common structural motif and selectivity for sodium ions. We hypothesized that the actin-based nonmuscle myosin II motor proteins, nonmuscle myosin heavy chain-IIA/myh9, and nonmuscle myosin heavy chain-IIB/myh10 might interact with sodium channel alpha subunits to play an important role in their transport, trafficking, and/or function. Immunochemical and electrophysiological assays were conducted using rodent nervous (brain and dorsal root ganglia) tissues and ND7/23 cells coexpressing Nav subunits and recombinant myosins. Immunoprecipitation of myh9 and myh10 from rodent brain tissues led to the coimmunoprecipitation of Nax, Nav1.2, and Nav1.3 subunits, but not Nav1.1 and Nav1.6 subunits, expressed there. Similarly, immunoprecipitation of myh9 and myh10 from rodent dorsal root ganglia tissues led to the coimmunoprecipitation of Nav1.7 and Nav1.8 subunits, but not Nav1.9 subunits, expressed there. The functional implication of one of these interactions was assessed by coexpressing myh10 along with Nav1.8 subunits in ND7/23 cells. Myh10 overexpression led to three-fold increase ( P < 0.01) in the current density of Nav1.8 channels expressed in ND7/23 cells. Myh10 coexpression also hyperpolarized voltage-dependent activation and steady-state fast inactivation of Nav1.8 channels. In addition, coexpression of myh10 reduced ( P < 0.01) the offset of fast inactivation and the amplitude of the ramp currents of Nav1.8 channels. These results indicate that nonmuscle myosin heavy chain-IIs interact with sodium channel alpha subunits subunits in an isoform-dependent manner and influence their functional properties.

Introduction

Sodium channels are integral membrane proteins found in the cells of all higher eukaryotes. They are made up of both alpha (α) and beta (β) subunits. Nine sodium channel alpha subunits (Naα) are voltage gated (i.e., Navα: Nav1.1, Nav1.2, Nav1.3, Nav1.4, Nav1.5, Nav1.6, Nav1.7, Nav1.8, and Nav1.9), and only one Naα subunit (i.e., Nax) is gated by cellular sodium gradient.[1] Navα subunits are functional on their own, but they tend to associate with four known β subunits (i.e., Navβ1, Navβ2, Navβ3, and Navβ4), which are not functional on their own, to modulate their own functional properties.[2] Navα subunits mostly occur at central nervous system neurons (i.e., Nav1.1, Nav1.2, Nav1.3, and Nav1.6), skeletal muscles (SKMs) (i.e., Nav1.4), uninnervated SKM (i.e., Nav1.5), heart (i.e., Nav1.5), parasymphathetic nervous system neurons (i.e., Nav1.7), dorsal root ganglion (DRG) neurons (i.e., Nav1.7, Nav1.8, and Nav1.9), astrocytes (i.e., Nax), hypothalamus (Nax), and so on.[1] Voltage-gated sodium channels (Navs) utilize cellular membrane potential to function, influx extracellular sodium into the cells, and fire action potentials in excitable cells such as neurons and myocytes (skeletal and cardiac).

The mechanisms by which neuronal Naα subunits are trafficked to the cell surface are not well understood though considerable knowledge has been garnered over the years regarding their clustering at the axon initial segment and the nodes of Ranvier.[3-6] Generally, vesicles carrying membrane proteins traffic from the intracellular pools to the plasma membranes. This involves their transport by kinesin family of motor proteins (KIF) along the microtubules and/or by myosin family of motor proteins (myo or myh) along the actin filaments. Kinesin heavy chain-1 (KIF5B) has been shown to be involved in the trafficking of Nav1.2[7] and Nav1.8[8] subunits. Myosin Va (myoVa) is also shown to be involved in the trafficking of Nav1.2 subunits.[9]

In recent years, class II nonmuscle myosins (i.e., NM-IIs) are shown to interact with various membrane proteins including chemokine receptor CXCR4 (in T lymphocytes),[10] epidermal growth factor receptor (EGFR),[11] N-methyl-d-aspartate receptor (NMDAR),[12] α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor (AMPAR),[13] the pore-forming subunit of P/Q-type calcium channels (Cav2.1),[14] a lysosomal membrane protein (Cln3),[15] Kv2.1 channels,[16] and so on. Class II NM-IIs, like class II muscle myosins (i.e., myh1, myh2, myh3, etc.), are hexameric molecules comprising two NM-II heavy chains (i.e., NMHC-IIA/myh9, NMHC-IIB/myh10, or NMHC-IIC/myh14), two myosin essential light chains (ELCs), and two myosin regulatory light chains (MRLCs).[17,18] The heavy chain comprises a globular head domain (the site for interaction with actin and ATP), a neck region (site for interaction with ELC and MRLC), and a tail region which homodimerizes in a helical fashion and is a possible site for interaction with the cargo.[17,18] NM-IIs interact only transiently with actin and typically spend most of their kinetic cycle detached from actin.[19] In contrast, myosin V moves processively along actin filaments and associates strongly with actin in the presence of ATP. These characteristics enable myoVa to function as an intracellular cargo motor.[19] Because of their known association with multiple classes of membrane proteins including ion channels, we hypothesized that NM-IIs might play an important role in the trafficking of Naα subunits expressed in the neuronal tissues. Our results indicate that two NM-II isoforms, myh9 and myh10, interact with sodium channel subunits expressed in nervous tissues in an isoform-specific manner. We also show that the functional properties of Nav1.8 channels are modulated by myh10. Hence, there appear to exist multiple pathways for trafficking of Naα subunits to the cell membrane by two independent cellular motor proteins, myosins and kinesins.

Materials and methods

Plasmids

Green fluorescent protein (GFP) tagged NMHC II-B (i.e., myh10; National Center for Biotechnology Information reference: NM_005964/NP_005955; full length: 1976 amino acids) and AnkG270-mCherry were obtained from Addgene.[20,21] These constructs were subcloned or modified as needed. All other constructs were available in our laboratory. The Nav1.6 construct harbors a mutation (Tyr371Ser) that renders it resistant to tetrodotoxin (TTX).

Cell culture and transfection

Human embryonic 293 cells (HEK293 cells), ND7/23 cells, and Nav1.6-GFP stable cell lines in HEK293 background were cultured according to standard procedures. Optifect (Thermo Fisher Scientific, Waltham, MA) or LipoJet (SignaGen Laboratories, Rockville, MD) transfection reagents were used according to manufacturer’s instructions for transient transfections. ND7/23 cells (i.e., a hybrid of mouse neuroblastoma and rat DRG neurone) endogenously express Nav1.3, Nav1.6, and Nav1.7 subunits and are considered as a model neuronal cell line for functional expression Nav subunits.[22-25]

Antibodies

Various immunoglobulin (IgG) isotypes, mouse monoclonal antibodies, and rabbit polyclonal antibodies used for immunoprecipitation (IP), and/or immunoblot (IB) assays are provided in as Supplemental Tables (Tables S1 and S2). Antibody dilutions and/or concentrations used for IP and/or IB assays along with the molecular weight (∼kDa) of the antigens detected by these antibodies are also provided (Table S2).

Immunoblots were incubated with monoclonal antibodies against myh9 (1:500; Abcam), myh10 (1:1000; Abcam), MRLC (1:200; Santa Cruz Biotechnology), Nav subunits (pan-Navα; 1:1000; Sigma), Nav1.1 (1:500; NeuroMab), ankyrin-G (AnkG) (1:500; NeuroMab), ankyrin-B (1:500; NeuroMab), ankyrin-R (AnkR) (1 µg/mL; Santa Cruz Biotechnology), β-I spectrin (1:100; Abcam), α-II spectrin (1:500; Abcam), β-IV spectrin (1:500; NeuroMab), and β-actin (1:10,000; Sigma).

Polyclonal antibodies against Nav1.2 (1:200; Alomone), Nav1.3 (1:200; Alomone), Nav1.6 (PN4; 1:200; Sigma), Nav1.7 (1:200; Alomone), Nav1.8 (1:200; Alomone), Nav1.9 (1:200; Alomone), Nax (1:1000; Abcam), β-II spectrin (1:2000; Abcam), β-III spectrin (1 µg/mL; Santa Cruz Biotechnology), and GFP (1:1000; Invitrogen) were also used for immunoblotting.

The MRLC antibody (sc-28329; clone E4; Santa Cruz Biotechnology) is already shown to detect various MRLCs and hence considered as a pan-MRLC in some quarters.[26] It is claimed to recognize the MRLCs from human (i.e., MRCL3, MRLC2, MYL9, and LOC391722: MRLC 12B-like), mouse (i.e., Mylc2b, Myl9, and Myl12a), and rat (i.e., Mrlcb and Myl9) tissues. In our hand, it poorly detects the MRLCs from cell lysates, but it detects the MRLCs as a coimmunoprecipitate from precipitation of NMHC-IIs very well (Figures 1 to 3).

Figure 1.

Interaction of myh9 with Naα subunits expressed in the mouse brain tissues. WT type (In, lane 1 in (a) and (b)) or Nav1.6 KO (Scn8a; In, lane 1 in (c)) mouse brain tissue lysates were PC with mouse IgG2b isotypes (PC, lane 2 in (a), (b), and (c)) prior to IP using mouse anti-myh9 antibodies (IP, lane 3 in (a), (b), and (c)) of the IgG2b isotypes. Myh9 coimmunoprecipitated pan-Nav ((i) in (a)), Nax ((ii) in (a)), and Nav1.2 ((iv) in (a)) subunits, but not Nav1.1 subunits ((iii) in (a)), expressed in mouse brain tissues. Nav1.6 subunits ((i) in (b)) were not coimmunoprecipitated with myh9 expressed in the mouse brain tissues. As expected, there was not any Nav1.6 immunoreactivity from Scn8a mouse brain tissues ((i) in (c)). Myh9 also coimmunoprecipitated β-actin ((v) in (a) and (ii) in (b) and (c)) and MRLCs ((iii) in (b) and (c)) expressed in mouse brain tissues. The MRLC immunoreactive signal from the mouse brain tissue lysates is barely detectable. Mouse IgG-HC (panel (ii) in (b) and (c)) and IgG-LC (panel (iii) in (b) and (c)), which are separated from their intact immunoglobulins (i.e., used for PC or IP) upon denaturation, could be seen as these blot sections are probed with mouse (anti-β-actin and anti-MRLC) antibodies. Additional information is available in Figure S1. IgG-LC: immunoglobulin light chain; IgG-HC; immunoglobulin heavy chain; In: lysate input; IP: immunoprecipitation; KO: knockout; mIgG2b: mouse immunoglobulin isotype 2b; MRLC: myosin regulatory light chain; myh: myosin heavy chain; Nav: voltage-sensitive sodium channel; Nax: nonvoltage-dependent sodium channel; PC: precleared; WT: wild type.

Figure 2.

Interaction of myh9 and myh10 with Nav subunits expressed in the rat brain tissues. (a) Myh9 and myh10 interact with pan-Nav subunits expressed in the rat brain tissues. Rat brain tissue lysates (In, lanes 1 and 9) were PC with mouse IgG2b isotypes (PC, lanes 2 and 8) and mouse IgG1 isotypes (PC, lane 6) prior to IP using indicated antibodies (lane 3 = myh9, lanes 4 and 10 = myh10, lane 5 = KIF5B, and lane 7 = AnkG) of the same isotypes as those were used for preclearing. IP complexes in the gel were loaded following the loading of their respective isotype (PC) complexes. Pan-Nav subunits (i) were coimmunoprecipitated with myh9, myh10, KIF5B, and AnkG expressed in the rat brain tissues. CoIP of pan-Nav subunits by AnkG and KIF5B served as positive controls. Myh9 and myh10 coimmunoprecipitated β-actin (ii) and MRLCs (iii) expressed in the rat brain tissues. Anti-MRLC antibodies poorly detect their antigens from rat brain tissue lysates (lane 1 in (iii)). AnkG also coimmunoprecipitated β-actin and MRLCs expressed in the rat brain tissues. Denatured mouse IgG-HC (iii) separated from their intact immunoglobulins (i.e., used for PC or IP) could be seen in the immunoblot, as this section of the blot was probed with mouse anti-β-actin antibodies. (b) Interaction of myh10 with Naα subunits expressed in the adult rat brains. Rat brain tissue lysates (In, lane 1) were PC with mouse IgG1 isotypes (PC, lane 2) and mouse IgG2b isotypes (PC, lane 4) prior to IP using antibodies for AnkG (IP, lane 3) and myh10 (IP, lane 5) of the of the same isotypes as those were used for preclearing. Loading of isotype (PC) complexes in the gel preceded those of the IP complexes. Myh10 coimmunoprecipitated Nax (i) and Nav1.2 (iii) subunits, but not Nav1.1 (ii) and Nav1.6 (iv) subunits, from rat brain tissues. As expected, AnkG coimmunoprecipitated Nav1.1 (ii), Nav1.2 (iii), and Nav1.6 (iv) subunits, but not Nax (i) subunits, expressed in rat brain tissues. AnkG: ankyrin-G; IgG-HC: immunoglobulin heavy chain; In: lysate input; IP: immunoprecipitation; KIF5B: kinesin family member 5B; mIgG1: mouse immunoglobulin isotype 1; mIgG2b: mouse immunoglobulin isotype 2b; MRLC: myosin regulatory light chain; myh: myosin heavy chain; Nav: voltage-sensitive sodium channel; Nax: nonvoltage-dependent sodium channel; PC: precleared.

Figure 3.

Interaction of myh9 and myh10 with Nav1.3 subunits expressed in the postnatal day 3 (P3) rat brain tissues. P3 rat brain tissue lysates (In, lane 1) were PC with mouse IgG2b isotypes (PC, lane 2) prior to IP using mouse anti-myh9 (IP, lane 3) and mouse anti-myh10 (IP, lane 4) antibodies of the IgG2b isotypes. Loading of mIgG2b (PC) complexes in the gel preceded those of the IP complexes. Myh9 and myh10 coimmunoprecipitated Nav1.3 subunits (i) expressed in P3 rat brain tissues. Myh9 and myh10 also coimmunoprecipitated β-actin (ii) and MRLCs (iii) expressed in P3 rat brain tissues. The MRLC immunoreactive signal from the rat tissue lysates is barely detectable and is indicated with an asterisk (*). The denatured mouse IgG-LC (iii) separated from their intact immunoglobulins (i.e., used for PC or IP) could be seen in the immunoblot, as this section of the blot is probed with mouse anti-MRLC antibodies. IgG-LC: immunoglobulin light chain; In: lysate input; IP: immunoprecipitation; mIgG2b: mouse immunoglobulin isotype 2b; MRLC: myosin regulatory light chain; myh: myosin heavy chain; Nav: voltage-sensitive sodium channel; PC: precleared.

The aggregate pan-Nav channel co-IP signals would not reflect any contribution from the Nax subunit, as the pan-Nav antibody (Sigma: Clone K58/35) is raised against an 18 amino acids epitope that is present only in the intracellular III to IV loop (i.e., TEEQKKYYNAMKKLGSKK) of the Nav subunits.

Preparation of cell and tissue lysates or membrane fractions

All animal care and experimental studies were approved by the Veterans Administration Connecticut Healthcare System Institutional Animal Care and Use Committee. We followed the protocols published elsewhere with some modifications to prepare adult rat (Sprague-Dawley), postnatal day 3 rat (Sprague-Dawley), adult mouse (C57/BL6), and Scn8a mouse (i.e., Nav1.6 null homozygous; postnatal day 12–14) brain or DRG tissue lysates for IP and immunoblotting.[7] This is because Nav1.3 proteins are relatively better expressed in embryonic and neonatal brains. We also used young animal’s brains as Nav1.6 knockout (KO; Scn8a) mouse only survives up to two to three weeks after birth. Adult rodents were used for all other applications because they provide larger starting material for the biochemical assays, thus allowing us to reduce the number of animals used in these studies.

The lysis or IP buffer was made of 20 mM Tris-Cl (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1 mM DTT, 10 mM EGTA, and 2× complete protease inhibitor cocktail (Roche Diagnostics Corporation, Indianapolis, IN). A whole animal (male or female) brain was homogenized in pieces in a tissue grinder (Qiagen, Valencia, CA) to a final volume of ∼50 mL lysis buffer. Homogenates were solubilized for 2 h at 4°C and centrifuged at 50,000 g for 30 min at 4°C using a Beckman Coulter Optima® ultracentrifuge to collect the supernatants for IP and immunoblotting. The rat DRG lysates were prepared in the similar manner except that the largest DRG pairs (three pairs: L4–L6) from an adult rat (male or female) was homogenized to a final volume of 1 mL lysis buffer. Nontransfected HEK293 cells (control), cells transiently transfected with plasmid constructs, and Nav1.6-GFP cells were collected by centrifugation at 500 g for 5 min at 4°C upon trypsinization. These pellets were washed twice with ice-cold phosphate-buffered saline by centrifugation at 500 g for 5 min at 4°C before lysis using the IP buffer. Cell supernatants were obtained by centrifugation at 15,000 g for 20 min at 4°C for IP and immunoblotting.

Solubilized mouse brain membrane fractions for antibody characterization (Figure S2) were prepared according to the protocol reported elsewhere.[27] Protein concentrations in the tissue lysates and membrane preparations were determined using the Bradford reagent (Bio-Rad, Hercules, CA).

Immunoprecipitation

For IP experiment, HEK293 cell or animal tissue supernatants containing 0.5 to 4 mg protein (in ∼1 mL lysate) was precleared (PC) for 1 to 4 h at 4°C with 5 to 10 µg of suitable mouse antibody isotypes or rabbit IgG and 80 to 100 µL of Dynabead® protein G (Thermo Fisher Scientific). PC supernatants were incubated (overnight, 4°C) with 5 to 10 µg of desired IP antibody (Table S1) and 80 to 100 µL of Dynabead® protein G. Dynabead® protein G beads bound to control antibody isotypes (i.e., PC complexes) or desired primary antibodies (i.e., IP complexes) were washed for five times with IP buffer or wash buffer supplied by the vendor (Thermo Fisher Scientific) and eluted with NuPAGE® LDS Sample Buffer (Thermo Fisher Scientific) in the presence of NuPAGE® Sample Reducing Agent (Thermo Fisher Scientific).

Western blotting

About 20 to 50 µg of HEK293 cell or animal tissue lysates were denatured using NuPAGE® Sample Reducing Agent in the presence of NuPAGE® LDS Sample Buffer to serve as input (In) sample for Western blotting. The input samples, PC complexes, IP complexes, and/or at times denatured depleted supernatants were resolved on NuPAGE® Novex® 4% to 12% Bis-Tris Gels (1.0 mm, 12 well) and transferred to a nitrocellulose membrane. Membranes were blocked using a blocking buffer (5% nonfat dry milk and 1% bovine serum albumin (BSA) in 0.1% tris-buffered saline with Tween 20 (TBST) or 5% BSA in 0.1% TBST) for 1 h, washed, and incubated overnight with desired primary antibodies (Table S2) diluted in the blocking buffer. The blots were washed and incubated in horseradish peroxidase-conjugated goat antimouse (1:10,000; Dako, Santa Clara, CA) or goat antirabbit (1:10,000; Dako, Santa Clara, CA) for 1 h. The blots were washed extensively and developed for 1 to 10 min with the Perkin Elmer Western Lightning Plus-enhanced chemiluminescence kit using a Bio-Rad ChemiDoc XRS+ or ChemiDoc Imaging System. At times, IBs were stripped using a stripping buffer (Thermo Fisher Scientific) to reprobe with another primary antibody. We usually cut through the IgG-heavy chain (HC) and/or IgG-light chain (LC) regions of the Ponceau S (Sigma) stained nitrocellulose membranes for probing different section of the membrane with different antibodies. Therefore, cut marks could be seen across the IgG-HC and/IgG-LC immunoreactive bands.

Voltage-clamp analysis

ND7/23 cells were cultured on 12-mm glass coverslips coated with poly-D-lysine/laminin (BD Biosciences, San Jose, CA) and transfected with Nav1.8 cDNA and GFP-myh10 (or GFP control) using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). After 48 h, cells with robust green fluorescence were selected for recording. Whole-cell voltage-clamp recordings of ND7/23 cells were performed at room temperature (20°C–22°C) using an EPC-9 amplifier (HEKA Electronics, Lambrecht/Pfalz, Germany). Fire-polished electrodes (0.6–1.5 MΩ) were fabricated from 1.6 mm outer diameter borosilicate glass micropipettes (World Precision Instruments, Sarasota, FL). The pipette potential was adjusted to zero before seal formation, and liquid junction potential was not corrected. Capacity transients were cancelled, and voltage errors were minimized with 80% to 90% series resistance compensation. Currents were acquired with Pulse Software (HEKA Electronics), 5 min after establishing whole-cell configuration, sampled at a rate of 50 kHz, and filtered at 2.9 kHz. For current–voltage relationships, cells were held at −80 mV and stepped to a range of potentials (−60 to +50 mV in 5 mV increments) for 100 ms. Peak inward currents (I) were plotted as a function of depolarization potential to generate I–V curves. Activation curves were obtained by converting I to conductance (G) at each voltage (V) using the equation G = I/(V−Vrev), where Vrev is the reversal potential which was determined for each cell individually. Activation curves were then fit with Boltzmann functions in the form of G = Gmax/{1+ exp[(V1/2,act− V)/k]}, where Gmax is the maximal sodium conductance, V1/2,act is the potential at which activation is half maximal, V is the test potential, and k is the slope factor. Steady-state fast inactivation was achieved with a series of 500-ms prepulses (−90 to 10 mV in 5 mV increments), and the remaining noninactivated channels were activated by a 40-ms step depolarization to 10 mV. Peak inward currents obtained from steady-state fast inactivation protocols were normalized to the maximal peak current (Imax) and fit with Boltzman functions: I =A + (1 − A)/(1 + exp[(V − V1/2,inact)/k)], where V represents the inactivating prepulse membrane potential, and V1/2,inact represents the midpoint of inactivation. Ramp currents were elicited with a slow depolarizing voltage ramp from −80 to 40mV at a rate of 0.2 mV/ms. The absolute ramp current amplitude was normalized to the maximal peak current obtained by I–V protocol. The pipette solution contained (in mM) 140 CsF, 10 NaCl, 1 EGTA, and 10 HEPES, pH 7.30 (adjusted with CsOH). Osmolarity was adjusted to 310 mOsmol/L with dextrose. The extracellular bath solution contained (in mM) 140 NaCl, 3 KCl, 1 MgCl2, 1 CaCl2, 10 HEPES, 5 CsCl, and 20 tetraethylammonium chloride, pH 7.3 with NaOH (327 mOsmol/L). TTX (500 nM) was added to the bath solution to block the endogenous TTX-sensitive voltage-gated sodium currents from ND7/23 cells.

Data analysis

Data were analyzed using FitMaster (HEKA Electronics) and Origin 8.5 professional (Microcal Software, Northampton, MA) and presented as means ± standard error of the mean. Except where noted, statistical significance was determined by unpaired Student’s t tests.

Results

Interaction of myh9 with Naα subunits expressed in the mouse brain

We hypothesized that NM-IIs might interact with the Naα subunits. One of the specific hypotheses we tested was that NM-IIA (i.e., myh9) might interact with Nav1.6 subunits expressed in the brain tissues. We tested this hypothesis using mouse brain tissues as a suitable control (i.e., Scn8a/Nav1.6 KO mouse) was available to us. Results indicated a lack of interaction of myh9 with Nav1.6 expressed in mouse brain tissues (Figure 1(a) and (c); Figure S1). As expected, myh9 coimmunoprecipitates its partner proteins β-actin and MRLCs from both wild type and Nav1.6 KO mouse tissues (Figure 1(b) and (c); Figure S1). These results did not indicate whether myh9 interacts with other Naα subunits expressed in mouse brain tissues. As a first step, we immunoprecipitated myh9 from adult mouse brain tissues and probed for co-IP of pan-Navα subunits expressed there (Figure 1(a)). Co-IP of pan-Nav subunits was observed with myh9 expressed in mouse brain tissues (Figure 1(a)). The pan-Nav co-IP signal observed could be as a result of interaction of one or more Nav isoforms (i.e., Nav1.1, Nav1.2, and/or Nav1.6) with myh9 expressed in mouse brain tissues. Further analysis indicated that myh9 coimmunoprecipitate Nax subunits, Nav1.2 subunits, but not Nav1.1 subunits, expressed in adult mouse brain tissues (Figure 1(a)). Nav1.2 co-IP signals from mouse brain tissues (and also from rat brain tissues in Figure 2; see next section) revealed multiple immunoreactive bands possibly indicating that Nav1.2 subunits exist in multiple modified states as a result of posttranslational modifications, degradation; and so on.[28]

Interaction of myh9 and myh10 with Naα subunits expressed in the rat brain

We wanted to recapture the myh9 and Nav interaction observed in the mouse brain tissues in rat brain tissues. We also wanted to know whether myh10, another nonmuscle myosin isoform, would interact with the same Nav subunits those interact with myh9. To this end, both myh9 and myh10 were immunoprecipitated from adult rat brain tissues and probed for co-IP of pan-Navα subunits. (Figure 2(a)). We also immunoprecipitated AnkG and kinesin-1 heavy chain (i.e., KIF5B) to control for co-IP of pan-Nav subunits, as they are known to interact with various Nav subunits.[4,7,29,30] Results indicated co-IP of pan-Nav subunits with both myh9 and myh10 expressed in rat brain tissues (Figure 2(a)). As expected, both myh9 and myh10, but not KIF5B, coimmunoprecipitated β-actin and MRLCs expressed in the rat brain tissues (Figure 2(a)). Both KIF5B and AnkG also coimmunoprecipitated pan-Nav subunits from rat brain tissues (Figure 2(a)).

This AnkG antibody also pulled down β-actin from rodent brain tissues which could be attributed to its interaction with actin-binding proteins such as spectrins (Figure 2).[31] We also observed that AnkG coimmunoprecipitated MRLCs with (Figure 2) which could be as a result of direct interaction of AnkG with MRLCs or its interaction with proteins (e.g., myosins) which interact with MRLCs.[32-34] In addition, we show that IP of AnkG using the same antibody led to the IP of its cognate antigens (i.e., all the three isoforms of 480, 270, and 190 kDa) and co-IP of Nav1.1 subunits, Nav1.2 subunits, Nav1.6 subunits, α-II spectrin, β-II spectrin, β-III spectrin, and β-IV spectrin (all known to interact with AnkG), but not βI-spectrin and the Nax subunits (which are known not to partner with AnkG), expressed in rat brain tissues (Figure 2; Figure S3).[4,29,30]

The pan-Nav immunoreactive co-IP signals indicate that one or more Nav isoforms expressed in the adult rat brain tissues (i.e., Nav1.1, Nav1.2, and/or Nav1.6) might be interacting with both of these myosins. Hence, we wanted to know which Nav isoforms interact with NM-IIs expressed in the rat brain tissues. Results indicated co-IP of Nav1.2 subunits but not Nav1.1 and Nav1.6 subunits with myh10 expressed in adult rat brain tissues (Figure 2(b)). Myh10 also coimmunoprecipitated the atypical sodium channel, Nax, subunits expressed in rat brain tissues (Figure 2(b), (i)).

Then, we wanted to determine whether Nav1.3 subunits expressed in rat brain tissues would interact with myh9 and/or myh10. To this end, we immunoprecipitated myh9 and myh10 from postnatal day 3 (P3) rat brain tissues, as these tissues are known to express Nav1.3 subunits. IP of myh9 and myh10 from P3 rat brain tissues led to the co-IP of Nav1.3 subunits expressed there (Figure 3).

Reciprocally, we wanted to know whether IP of pan-Nav subunits from adult rat brain tissues would lead to co-IP of myh9 and/or myh10 expressed there (Figure S4). Pan-Nav antibodies could not coimmunoprecipitate myh9 expressed in rat brain tissues. There appears to be some degree of co-IP of myh10 by pan-Nav subunits expressed in rat brain tissues (Figure S4). As expected, pan-Nav subunits coimmunoprecipitated AnkG (which occurs as three different isoforms of 190, 270, and 480 kDa) and AnkR[5] from rat brain tissues (Figure S4).

Interaction of myh9 and myh10 with Nav subunits expressed in the rat DRGs

Nav1.7, Nav1.8, and Nav1.9 subunits are predominantly expressed in adult DRG tissues. Therefore, we wanted to determine whether any of these Nav subunits expressed in the rat DRGs would coimmunoprecipitate with myh9 and myh10 expressed there.[35] We also immunoprecipitated KIF5B to control for co-IP of Nav1.8 subunits as a positive interaction for the two proteins has been reported previously.[8] IP of myh9 and myh10 led to the co-IP of Nav1.7 and Nav1.8 subunits but not Nav1.9 subunits expressed in rat DRGs (Figure 4(a)). As expected, KIF5B also coimmunoprecipitated Nav1.18 subunits expressed in rat DRGs. These results also indicated that KIF5B do not interact with Nav1.7 and Nav1.9 subunits expressed in rat DRGs.

Figure 4.

Interaction of myh9 and myh10 with Nav subunits expressed in adult rat DRG tissues. (a) Interaction of myh9 and myh10 with Nav subunits expressed in adult rat DRG tissues. Adult rat DRG tissue lysates (In, lane 1) were PC with mouse IgG2b isotypes (PC, lane 2) prior to IP using mouse anti-myh9 (IP, lane 3), anti-myh10 (IP, lane 4), and anti-KIF5B antibodies of IgG2b isotypes. Loading of IgG2b (PC) complexes in the gel preceded those of the IP complexes. Myh9 and myh10 coimmunoprecipitated Nav1.7 (i) and Nav1.8 (ii) subunits, but not Nav1.9 (iii) subunits, expressed in rat DRG tissues. As expected, both myh9 and myh10, but not KIF5B, coimmunoprecipitated β-actin expressed in rat DRG tissues. KIF5B also coimmunoprecipitated Nav1.8 subunits, but not Nav1.7 and Nav1.9 subunits, expressed in rat DRG tissues. Mouse IgG-HC (iv) and IgG-LC (iv), which are separated from their intact immunoglobulins upon denaturation, are seen, as this blot section was probed with mouse (anti-β-actin) antibodies. (b) Interaction of Nav1.7 and Nav1.8 subunits with myh9 and myh10 expressed in adult rat DRG tissues. Adult rat DRG tissue lysates (In, lanes 2 and 4) were PC with mouse IgG1 isotypes (PC, lane 1) and rabbit IgG (PC, lane 5) prior to IP using mouse anti-Nav1.7 (IP, lane 3) antibodies of IgG1 isotypes and rabbit anti-Nav1.8 (IP, lane 6) antibodies. Loading of PC complexes in the gel preceded those of the IP complexes. Nav1.7 and Nav1.8 subunits coimmunoprecipitated myh9 (i) and myh10 (ii) expressed in rat DRG tissues. Probing with pan-Nav antibodies revealed the IP signals for Nav1.7 and Nav1.8 subunits (iii). An asterisk (*) indicates loss of input signals following stripping the blot to probe for Nav1.8 IP signals using pan-Nav antibodies. DRG: dorsal root ganglia; IgG-HC: immunoglobulin heavy chain; In: lysate input; IP: immunoprecipitation; KIF5B: kinesin family member 5B; mIgG1: mouse immunoglobulin isotype 1; mIgG2b: mouse immunoglobulin isotype 2b; myh: myosin heavy chain; Nav: voltage-sensitive sodium channel; PC: precleared; rIgG: rabbit immunoglobulins.

On a reciprocal basis, we wanted to know whether Nav1.7 or Nav1.8 subunits would coimmunoprecipitate myh9 and/or myh10 expressed in the DRGs (Figure 4(b)). We immunoprecipitated Nav1.7 and Nav1.8 subunits from rat DRG tissues using antibodies already reported suitable for such reactions.[36-38] Nav1.7 subunits coimmunoprecipitated myh9 and myh10 from rat DRGs (Figure 4(b)). Also, myh9 and myh10 were coimmunoprecipitated with Nav1.8 subunits expressed in rat DRGs (Figure 4(b)).

Functional effects of myh10 overexpression on the electrophysiological properties of Nav1.8 channels heterologously expressed in ND7/23 cells

Earlier we demonstrated that both myh9 and myh10 interact with Nav1.8 subunits. We investigated whether such interaction would influence the electrophysiological properties of Nav1.8 channels. To this end, we coexpressed human Nav1.8 subunits along with GFP (Figure 5(a)) or GFP-myh10 (Figure 5(b)) in ND7/23 cells. Representative Nav1.8 sodium channel currents recorded from cells expressing GFP or myo10-GFP in the presence of Nav1.8 subunits are shown in Figure 5(a) and (b), respectively. Myh10 significantly increased the current density of Nav1.8 channel by three-fold (GFP control: 50.9 ± 10.3 pA/Pf, n = 17; myh10: 135 ± 26 pA/pF, n =16, P < 0.01). To examine the voltage dependence of activation, cells were held at −80 mV and stepped to a range of potentials (−60 to +50 mV in 5 mV increments) for 100 ms. As shown in Figure 5(c), myh10 coexpression hyperpolarized voltage-dependent activation of Nav1.8 channels. When fitted with Boltzmann plots, the midpoint of activation was significantly more negative for myh10 (4.8 ± 1.6 mV, n = 11) than for GFP control (10.9 ± 0.7 mV, P < 0.01; n = 13), a shift of −6.1 mV in the hyperpolarizing direction. The slope for the activation curve for the myh10 condition (8.1 ± 0.3, n = 11) was also significantly different from that for the GFP control (9.6 ± 0.2, n = 13, P < 0.001). Myh10 also affected the steady-state fast inactivation of Nav1.8 channels (Figure 5(d)). Myh10 hyperpolarized the fast inactivation curve. Although the middle point of fast inactivation for myh10 condition (−54.7 ± 1.2 mV, n = 9) was not significantly different from that of the GFP control (−51.7 ± 2.1 mV, n = 10), the slope for the myh10 condition (9.3 ± 0.2, n = 9) was significantly smaller than that for the GFP control (12.3 ± 0.6, n = 10, P < 0.001). In addition, coexpression of myh10 reduced the offset of fast inactivation (GFP control: 16.2 ± 2.0%, n = 10; myo10: 8.9 ± 0.8, n = 9, P < 0.01). We also investigated the effect of myh10 on Nav1.8 ramp currents. As shown in Figure 5(e), myh10 significantly reduced the amplitude of the ramp currents (14.1 ± 0.6%, n = 10, P < 0.01) compared with the GFP control (25.4 ± 2.4%, n = 9).

Figure 5.

Coexpression of myh10 increases peak current density of Nav1.8 channels heterologously expressed in ND7/23 cells. Representative sodium currents were recorded from ND7/23 cells transiently expressing Nav1.8 in the presence of GFP (a) or GFP-myh10 (b). Cells were held at −80 mV, and sodium currents were elicited by a series of step depolarizations from −60 to +50 mV in 5-mV increments. (c) Coexpression of myh10 hyperpolarized activation of Nav1.8. (d) Coexpression of myh10 enhanced steady-state fast inactivation of Nav1.8. (e) Coexpression of myh10 reduced the amplitude of the ramp currents. GFP: green fluorescent protein; myh: myosin heavy chain.

Discussion

Class II myosins, with 15 members (i.e., myh1, myh2, myh3, myh4, myh6, myh7, myh7b, myh8, myh9, myh10, myh11, myh13, myh14, myh15, and myh16) in this group, are the largest class of myosins in vertebrates and are traditionally known as “conventional myosins.” Most of these class II myosins are expressed in muscle (skeletal, smooth, and/or cardiac) cells. However, myh9, myh10, and myh14 are expressed widely in nonmuscle myosins and are known as nonmuscle class II myosin (NM-IIs or NMHC-IIs). Myh10 is more abundant than any other NM-IIs in neuronal tissues and thus has garnered much attention to for its role in brain structure and function.[13,39-41] In this work, we show that both myh9 and myh10 interact with Nax, Nav1.2, Nav1.3, Nav1.7, and Nav1.8 subunits expressed in the nervous tissues. Interaction of myh9 with Nax, Nav1.2, Nav1.3, Nav1.7, and Nav1.8 subunits led to the discovery of their interaction with myh10. This was anticipated because both myh10 and myh9 are homologous protein molecules possibly possessing some of the same structural and functional features for their interaction with Naα subunits, although they are engaged in unique and diverse functions.[13,41-43] For the same reason, we also anticipated that a lack of interaction of Nav1.1, Nav1.6, or Nav1.9 subunits with myh9 would lead to a similar outcome involving myh10. We could not assess the interaction Naα subunits with myh14, the third member of the NM-II group, due to lack of suitable antibodies for its IP. The interaction of myh9 and myh10 with only a subset of channel isoforms increases confidence that these interactions are specific and not an artifact of the biochemical assay.

In general, in some of the IBs, the pan-Nav and Nav subunit-specific immunoreactive bands appear different in the input and/or IP lanes. Several factors could contribute to these observations including the amount of total and target proteins in the sample. It could be seen that the recombinant Nav1.6 (detected by three different antibodies in lanes 3, 6, and 9) appears to run evenly and homogenously compared to Navs detected from native tissues (Figure S2). This could be due to the fact that native Navs are known to be heavily glycosylated and at least two additional isoforms, Nav1.1 and Nav1.2, are present in brain tissues. It is also possible that posttranslational modifications of Navs might occur in varying degrees in different tissue types (native tissues vs. HEK293 cells) which might contribute to their different mobilities on the IBs.[44,45]

The interaction of NM-IIs with Navs is consistent with previous observations that NM-IIs associate with membrane proteins and regulate their trafficking and/or function. For example, myh9 is associated with the trafficking and/or function of chemokine receptor CXCR4 (in T lymphocytes)[10] and EGFR.[11] Similarly, myh10 is associated with the trafficking and/or function of two members of the ionotropic glutamate receptors (i.e., NMDAR[12] and AMPAR[13]), Cav2.1 (the pore-forming subunit of P/Q-type calcium channels),[14] Cln3 (juvenile Batten disease protein/battenin/formerly known as juvenile neuronal ceroid lipofuscinosis: a lysosomal membrane protein),[15] and so on. A recent report also implicates NM-II in the sorting and post-Golgi dendritic trafficking of Kv2.1 channels.[16] Based on these results, we hypothesize that both myh9 and myh10 might be involved in the trafficking of the Naα subunits with which they interact.

Most myosins, if not all, use adaptor or partner proteins to move or interact with their cargo.[11,46,47] Nav subunits use various adaptor proteins such as AnkG,[48] AnkR,[5] syntrophin family of dystrophin-associated proteins,[49] 14-3-3,[50] plakophilin-2,[51] and so on, for their targeting, trafficking, membrane retention and organization, and regulation of their functional properties. It is possible that NM-IIs are also interacting with any of these or here-to-unknown adaptor proteins, thereby indirectly establishing contact with the Naα subunits with which they interact (Figure 6). Hence, the formation of a myosin-adaptor-Naα tripartite complex is plausible. This would be similar to the observation made recently that KIF5B uses AnkG as an adaptor to transport Nav1.2 subunits.[7] Preliminary results from our work (Soc. Neurosci. Abst. (2016) 501.02/G39) also indicate that both myh9 and myh10 interact with AnkG giving credence to the idea that there possibly exist a myosin–ankyrin–Naα pathway in neuronal and/or nonneuronal tissues (Figure 6). It appears that both myh9 and myh10 are using a common adaptor molecule (such as AnkG) to interact with some of the Nav subunits (such as Nav1.2 and Nav1.8) and other unique or common adaptor molecules for their interaction with Nav1.3, Nav1.7, and Nax subunits.

Figure 6.

A model for interaction for NM-IIs with mammalian sodium channel α subunits. NM-IIs could be interacting with Naα subunits directly (i) or via adaptor proteins (such as ankyrins) (ii). Also shown in the model the canonical spectrin–ankyrin–Nav signaling network. NM-II: nonmuscle myosin II.

Our functional data show that coexpression of myh10 affects the current density and gating properties of the Nav1.8 channels. Previous studies have not addressed whether NM-IIs would have any effect on the gating properties of the ion channels with which they interact.[12-14,16] In this study, the changes in current density rather than the changes in gating properties of Nav1.8 channels as a result myh10 overexpression may be its major contribution. The large (three-fold) increase in Nav1.8 current density could be partly ascribed to an increase in number of functional ion channels on the cell surface as a result of myh10 overexpression. The effect of myh10 overexpression on the current density of Nav1.8 might also be underestimated, as there is robust natural expression myh10 in ND7/23 cells These results are analogous to the observation that KIF5B increases the current density and surface expression of Nav1.8 channels in both cultured DRG neurons and ND7/23 cells.[8] These authors[8] also concluded that a stretch of 10 amino acid sequence (i.e., from 511–620) in the stalk domain of KIF5B mediates its interaction with Nav1.8 subunit (via its N-terminus). It is plausible that both myh9 and myh10 are using a conserved region or motif located away from their divergent and nonhomologous tail regions for their interaction with Naα subunits. These data support the idea that interaction of NM-IIs with Nav subunits could be functionally important.

Here, we provide evidence for the first time that class II nonmuscle myosins (myh9 and myh10) interact with a subset of Naα (both Nav and Nax). Our data also show that the interaction of NM-IIs and Nav subunits can be important for regulation of channel functions. Overall, our data support the conclusion that myh9 and myh10 interact specifically with a subset of sodium channels and suggest a model in which the actin–myosin network is involved in the regulation of function of sodium channels expressed in neurons, including DRG sensory neurons.

Click here for additional data file.

Supplemental material for Nonmuscle myosin II isoforms interact with sodium channel alpha subunits by Bhagirathi Dash, Chongyang Han, Stephen G Waxman and Sulayman D Dib-Hajj in Molecular Pain