cAMP/PKA Pathways and S56 Phosphorylation Are Involved in AA/PGE2-Induced Increases in rNaV1.4 Current

Arachidonic acid (AA) and its metabolites are important second messengers for ion channel modulation. The effects of extracellular application of AA and its non-metabolized analogue on muscle rNaV1.4 Na+ current has been studied, but little is known about the effects of intracellular application of AA on this channel isoform. Here, we report that intracellular application of AA significantly augmented the rNaV1.4 current peak without modulating the steady-state activation and inactivation properties of the rNaV1.4 channel. These results differed from the effects of extracellular application of AA on rNaV1.4 current. The effects of intracellular AA were mimicked by prostaglandin E2 but not eicosatetraynoic acid (ETYA), the non-metabolized analogue of AA, and were eliminated by treatment with cyclooxygenase inhibitors, flufenamic acid, or indomethacin. AA/PGE2-induced activation of rNaV1.4 channels was mimicked by a cAMP analogue (db-cAMP) and eliminated by a PKA inhibitor, PKAi. Furthermore, inhibition of EP2 and EP4 (PGE2 receptors) with AH6809 and AH23848 reduced the intracellular AA/PGE2-induced increase of rNaV1.4 current. Two mutated channels, rNaV1.4S56A and rNaV1.4T21A, were designed to investigate the role of predicted phosphorylation sites in the AA/PGE2–mediated regulation of rNaV1.4 currents. In rNaV1.4S56A, the effects of intracellular db-cAMP, AA, and PGE2 were significantly reduced. The results of the present study suggest that intracellular AA augments rNaV1.4 current by PGE2/EP receptor-mediated activation of the cAMP/PKA pathway, and that the S56 residue on the channel protein is important for this process.


Introduction
Arachidonic acid (AA) is a polyunsaturated fatty acid cleaved from cell membrane phospholipid molecules via the action of the enzyme PLA 2 . AA is a biologically active signaling molecule that plays important roles in neurons and muscle under both physiological and pathological conditions [1,2]. Its effects include modulation of the activity of protein kinases, elevation of intracellular Ca 2+ levels, and regulation of neuronal excitability [3][4][5]. AA and its metabolites have been shown to modulate ligand-and voltage-gated ion channels, such as NMDA and AMPA receptor channels, voltage-gated Na + and K + channels, and acid-sensing ion channels [6][7][8][9]. Regulation of ion channel activity by AA may occur via direct effects, where AA interacts directly with ion channel proteins, or through perturbation of the plasma membrane [7,10,11]. AA metabolites have been reported to indirectly modulate ion channels through oxygenases or cellular signal transduction pathways [12]. With the exception of our recent study in rat cerebellar granule neurons, most studies have focused on the effects of extracellularly-applied AA and have not investigated the effects of intracellular AA. Free intracellular AA serves as a key transient cell signaling intermediate and undergoes rapid enzymatic conversion to diverse metabolites, including prostaglandins (PGs, such as PGD 2 , PGE 2 and thromboxane A 2 ) and the leukotriene/lipoxin (LX) families of eicosanoids [13]. It would be interesting to compare the effects of extracellular versus intracellular AA application and to investigate the underlying mechanisms of cell response to intracellular application of AA.
Voltage-gated sodium channels (Na V ) are one of the primary classes of ion channels responsible for driving cellular excitability in the nervous system and in skeletal and cardiac muscle. Na V are important clinically because they play a central role in neuronal activity and in a number of disease pathologies [14]. The Na + channel consists of one large α subunit, which creates a functional membrane channel, and small β subunits, which modulate the voltage-dependent Na + channel [15]. To date, ten isoforms of the Na + channel α subunit have been cloned and characterized (Nav1.1-1.9 and Nax). The majority of sodium currents in the brain neurons are mediated by Na V 1.1-1.3 and Na V 1.6 [16,17], and it controls axonal action potential conduction and neurotransmitter release in presynaptic terminals [18]. Na V 1.4 is the predominant voltage-gated Na + channel isoform in skeletal muscle [19]. Mutations in the gene encoding Na V 1.4 have been associated with non-dystrophic skeletal muscle pathologies, including paramyotonia congenita, hyperkalaemic periodic paralysis, and potassium-aggravated myotonia [20]. Therefore, understanding the mechanisms of regulation of Na V 1.4 channel activity is of clinical importance.
Our previous studies indicated that AA activates or inhibits sodium channel current (I Na ) which is mainly composed of the Na V 1.2 α-subunit [21] in rat cerebellar granule cells (GCs) depending on the route of AA application, and that AA acts via either non-metabolic or metabolic pathways [22]. Extracellular AA directly inhibits I Na and modifies Na V channel kinetics, similar to results observed with the Na V 1.4 channel [23]. In contrast, intracellular AA augments the I Na current via metabolic pathways involving PGE 2 -mediated activation of cAMP/ PKA pathways. However, the effect of intracellular AA on Na V 1.4 channels is unclear. Moreover, the precise site phosphorylated by PGE 2 -mediated activation of cAMP/PKA pathways is lacked. Although Na V 1.2 and Na V 1.4 α subunit isoforms have greater than 60% amino acid sequence identity, these channels exhibit functional differences in gating, conductance and ion selectivity, which result in tissue-specific physiological functions and subtle differences in their pharmacological properties [24,25]. Thus, further exploration of the effect of intracellular AA in muscle Na + channels is required for a comprehensive understanding of the physiological role of AA.
In the present study, we used the whole-cell patch-clamp technique to investigate the effects of intracellular AA on the major α subunit of the Na + channel protein complex from skeletal muscle (rNa V 1.4). rNa V 1.4 channels were transiently transfected into stable cultured human embryonic kidney cells (HEK 293), which lack voltage-dependent Na + channels. Additionally, two site-directed mutants were designed to determine whether there is a phosphorylation site for PKA. Our studies indicate that intracellular AA activates rNa V 1.4 currents via the AA!P-GE 2 !EP receptor!cAMP/PKA pathway, and that the PKA phosphorylation site at S56 on the channel protein may be involved in the activity of AA.

Materials and Methods
Cell culture and rNa V 1.4 transfection HEK 293 cells were cultured in 1640 medium supplemented with 10% fetal calf serum in a 5% CO 2 incubator at 37°C. Cells were trypsinized, diluted with culture medium, and grown in 35-mm dishes. The rat Na V 1.4 sodium channel α subunit was cloned into pEGFP-N1 (Clontech Laboratories, Inc., Mountain View, CA, USA), as described previously [23]. Transient transfections were carried out using Lipofectamine TM 2000 Transfection Reagent (Life Technologies Corporation, Carlsbad, CA, USA) when cells reached 40 to 60% confluence. The cells were used for electrophysiological recordings four to five days after transfection. Individual transfected cells were visualized based on their expression of green fluorescent protein.

Patch clamp recordings
Whole-cell currents of HEK 293 cells were recorded using a conventional patch-clamp technique. Prior to current recordings, the culture medium was replaced with a bath solution containing (in mM): NaCl 145, KCl 2.5, HEPES 10, MgCl 2 1, and glucose 10 (pH adjusted to 7.4 using NaOH). Soft glass recording pipettes were filled with an internal solution containing (in mM): CsCl 142, NaCl 3, HEPES 10, MgCl 2 2, and EGTA 5 (pH adjusted to 7.3 using CsOH). The pipette resistance was 4-7 MO after being filling with the internal solution. All recordings were performed at room temperature (23-25C°).

Data acquisition and analysis
All currents were recorded using an Axopatch 200B amplifier (Axon Instrument, Foster City, CA, USA) operated in voltage-clamp mode. The computer was connected to the recording equipment with a Digidata 1300 analog-to-digital (A/D) interface. Currents were digitally sampled at 100 μs (10 kHz) and were filtered with a 3-kHz, three-pole Bessel filter. Currents were corrected on-line for leak and residual capacitance transients by a P/4 protocol. Data acquisition and analysis were performed with pClamp 8.01 software (Axon Instruments, Union City, CA, USA) and/or Origin 6.1 (Origin Lab, Northampton, MA, USA). Values are given as mean ± S.E.M., with n representing the number of cells. Statistical analysis was performed using the Student's t-test with paired comparisons where relevant. When multiple comparisons were made, data were analyzed by a one-way ANOVA test, followed by Tukey's test when significant differences were observed.

Prediction of phosphorylation sites and site-directed mutagenesis
The online program Scansite (http://scansite.mit.edu) was used to predict potential PKA phosphorylation sites in rNa V 1.4. Optimal phosphorylation sites for particular protein Ser/Thr kinases were predicted using the matrix of selectivity values for amino acids at each position relative to the phosphorylation site as determined from the oriented peptide library technique. We evaluated the surface accessibility, score, and percentile parameters of all candidate PKA phosphorylation sites and chose two candidates, S56 and T21, for further study. We performed site-directed mutagenesis at the S56 and T21 sites using the Quikchange II-E Site-Directed Mutagenesis Kit (Agilent Technologies). The PCR primers used are as follows:

Chemicals
All drugs, with the exception of fetal calf serum, were purchased from Sigma-Aldrich (St. Louis, MO, USA). 1640 culture medium and antibiotic-antimycotic solution were obtained from Gibco Life Technologies (Grand Island, NY, USA). Arachidonic acid, 5,8,11,14-eicosatetraynoic acid, flufenamic acid, indomethacin, prostaglandin E 2 , AH6809 and AH23848 were first dissolved in DMSO and then diluted in extracellular or intracellular solution, with a final DMSO concentration <0.2%. At these low doses, DMSO alone does not modulate I Na .

Intracellular AA increased rNa V 1.4 current amplitude without modifying steady-state activation and inactivation properties of the channel
The effect of intracellular AA on rNa V 1.4 current amplitude was first investigated. The rNa V 1.4 current evoked by a depolarization step to -10 mV from a holding potential of -100 mV for evoking the large current amplitude. When 10 μM AA was delivered intracellularly via pipette, the rNa V 1.4 current evoked increased with time after the establishment of the wholecell configuration. Moreover, the AA-induced increase of the rNa V 1.4 current was concentration-dependent ( Fig 1A). The rNa V 1.4 current amplitude recorded 2 minutes after the establishment of the whole-cell configuration increased by 5.27 ± 1.35% (n = 5), 11.72 ± 1.04% (n = 6), 30.72 ± 1.83% (n = 10) and 42.37 ± 6.49% (n = 4) with intracellular AA at concentrations of 0.1 μM, 1 μM, 10 μM, and 50 μM, respectively (P<0.05 by the one-way ANOVA test). To rule out the possibility that DMSO might contribute to the rNa V 1.4 current response, a control solution containing 0.2% DMSO was applied and the current amplitude lightly decrease 1.275 ± 1.015%, which was not a significant difference comparing with control (n = 5, P > 0.05). Statistical analysis of these data is shown in Fig 1B. The effect of intracellular AA on the steady-state activation properties of the rNa V 1.4 channel was then studied using appropriate voltage protocols. rNa V 1.4 currents were evoked by 20 msec depolarizing pulses from a holding potential of -100 mV to potentials between -70 mV to +60 mV in steps of 5 mV at intervals of 10 sec (Fig 2A). The voltage-current curve presented in Fig 2B shows that the maximum activation potential changed from -17.14 ± 3.60 mV to -15.71 ± 3.69 mV in the presence of 10 μM intracellular AA (P>0.05, n = 7 for control and n = 7 for AA). We obtained steady-state activation of rNa V 1.4 by normalizing the conductance as a function of the command potential. The data points were calculated using the equation g Na = I Na / (V m -V rev ), where g Na is the membrane Na + conductance, V m is the membrane potential, and V rev is the reversal potential for Na + . For both the control and AA treatments, the ratios were fitted to a Boltzmann function in the form of g Na /g Na-max = 1/ {1+exp [(V m1/2 -V m )/k]}, and the half-activation potential was obtained ( Fig 2C). According to these curves, the currents were half-activated at -32.43 ± 1.11 mV and -30.08 ± 1.02 mV (n = 26 for control and 24 for AA, P>0.05), the slope factor was 3.0 ± 0.48mV and 3.32 ± 0.33 mV (P>0.05) in the absence and presence of 10 μM AA, suggesting that intracellular AA did not significantly shift the voltage-dependence property of the steady-state activation of rNa V 1.4.
We also studied the effect of AA on the voltage dependence of the steady-state inactivation of rNa V 1.4. Current was elicited using 1s conditioning pre-pulses ranging from -130 mV to -10 mV in steps of 10 mV prior to a -20 mV test pulse. Fig 3A shows the currents obtained from cell recordings with normal intracellular solution (Control) or with 10 μM AA in the intracellular solution (AA). A curve illustrating the relationship between the rNa V 1.4 current peak and the pre-pulse potential is shown in Fig 3B. The steady-state inactivation curve was fitted using the Boltzmann equation In the 25 cells studied (n = 15 for control and n = 10 for AA), half-inactivated potential was -70.72 ± 0.86 mV and -67.91 ± 1.11 mV, slope factor was 4.36 ± 0.42 and 5.90 ±0.25 in the absence and presence of AA, respectively (P>0.05). Therefore, intracellular AA did not significantly shift the steadystate inactivation curve.
Our previous study revealed that extracellular AA had an inhibitory effect on the rNa V 1.4 current [23]. We thus tested whether intracellular application of AA could affect the inhibitory effect induced by extracellular AA. Fig 4A illustrates a representative experiment in which external AA was applied twice to the same cell while 10 μM AA was added to the pipette solution. The current amplitude increased over time with intracellular AA. After a stable rNa V 1.4 current had been obtained, external application of AA (50 xM) still caused a reversible decrease in the current amplitude by 23.05 ± 5.75% (P<0.01, n = 5), which was similar with external AA alone [23]. Statistical analysis of these data is shown in Fig 4B. Intracellular AA increases the rNa V 1.4 current via its metabolic product, PGE 2 , and via the EP receptor AA is known to be consumed through its conversion to prostaglandins and other products by the cyclooxygenase, lipoxygenase, and monooxygenase pathways [13]. Since the COX-2 metabolite prostaglandin E 2 (PGE 2 ) is involved in the intracellular AA-induced modulation of neuronal Na + current, we replaced intracellular AA with PGE 2 or eicosatetraynoic acid (ETYA), the non-metabolized analogue of AA to determine whether it is AA itself or its metabolic products that directly modulate(s) the rNa V 1.4 current (Fig 5A and 5B). Application of 10 μM PGE 2 in the pipette solution induced a similar effect as AA on the rNa V 1.4 current. The current amplitude increased by approximately 61.94 ± 4.96% (n = 10, P<0.01). In contrast, application of 1 μM ETYA to the intracellular solution gradually lowered the rNa V 1.4 current by  approximately 9.30 ± 1.18%. The normal intracellular solution alone did not affect rNa V 1.4 current during the two-minute recording. Statistical analysis of these data is shown in Fig 5B. To further address the role of PGE 2 in the intracellular AA-induced increase of the rNa V 1.4 current, the effects of PGE 2 and ETYA on the steady-state activation and inactivation properties of the rNa V 1.4 channel were compared to that of AA using the same voltage protocols as above (Fig 2A and Fig 3A). The half-activation potential was -39.56 ± 2.11 mV in the presence of 10 μM PGE 2 (n = 14 for ctrl, n = 10 for PGE 2 ), similar to the -34.98 ± 1.62 mV half-activation potential obtained in the presence of intracellular AA (P>0.05). In contrast, the half-activation potential was -45.41 ± 1.81 mV in the presence of 1 μM ETYA (P<0.05, n = 8). There was a significant difference between the effects of ETYA and intracellular AA on the voltagedependence of the steady-state activation of rNa V 1.4 (Fig 5C and 5D). However, there was no significant shift in the steady-state inactivation curve in the present of intracellular 10 μM PGE 2 or 1 μM ETYA (Fig 5E and 5F). Statistical analysis of these data is shown in Table 1 and  Table 2.
Next, flufenamic acid and indomethacin, two cyclooxygenase inhibitors, were used to confirm that AA metabolites mediate the effect of intracellular AA on the rNa V 1.4 current. Application of 5 μM flufenamic acid or 20 μM indomethacin alone in the bath solution did not significantly modify the rNa V 1.4 current amplitude. Within two minutes after the establishment of the whole cell configuration, the current amplitude was only reduced by 3.70 ± 1.66% (n = 8, P>0.05) and 1.83 ± 1.34% (n = 8, P>0.05), respectively. Adding AA to the pipette solution together with flufenamic acid or indomethacin eliminated the AA-induced increase of the rNa V 1.4 current (Fig 6A and 6B). When AA was applied intracellularly, the current amplitude was reduced by 4.06 ± 1.37% (n = 15, P>0.05) and 5.32 ± 1.97% (n = 7, P>0.05) in the presence of flufenamic acid or indomethacin, respectively.
We then investigated whether PGE 2 receptors were associated with the AA/PGE 2 -induced increase of the rNa V 1.4 current. Reverse transcriptase-mediated PCR analysis was first used to detect the presence of EP receptor mRNA in HEK 293 cells. PCR results showed that EP2 and EP4 are the major EP receptors expressed in HEK 293 cells, while EP3 mRNA is less abundant (Fig 7A). Thus, EP2 and EP4 receptor antagonists AH6809 and AH23848 were used for further exploration. Bath infusion of AH6809 or AH23848 significantly inhibited the effects of AA or PGE 2 on the rNa V 1.4 current (Fig 7B and 7C). In the presence of AH6809 or AH23848, the AA-induced increase of the rNa V 1.4 current was significantly reduced to 9.7±4.35% (n = 8,  and AA on the rNa V 1.4 current were investigated in the presence of a PKA activator or inhibitor. Intracellular application of 10 μM db-cAMP (dibutyryl cAMP), a membrane permeable cAMP analogue, mimicked the AA-or PGE 2 -induced effect on rNa V 1.4 ( Fig 8A) and significantly increased the rNa V 1.4 current amplitude by 51.4±11.3% (n = 7, P<0.05). Furthermore, the AA-or PGE 2 -induced increase in rNa V 1.4 current amplitude was attenuated by applying 50 nM PKAi, a bioactive peptide fragment 6-22 amide which inhibited PKA activity (Fig 8C  and 8D). In the presence of 50 nM PKAi, the PGE 2 -induced increase in rNa V 1.4 current was significantly reduced from 61.94 ± 4.96% (n = 10) to 24.2% (n = 8, P<0.05). Similarly, application of 50 nM PKAi reduced the AA-induced increase of rNa V 1.4 current from 47.443±7.12% (n = 13) to 1.82±3.52% (n = 11, P<0.05). Intracellular application of 50 nM PKAi alone did not affect the rNa V 1.4 current amplitude (Fig 8B), Statistical analysis of these results is shown in Fig 8E. These results suggest that the cAMP/PKA pathway is associated with the AA/PGE 2induced increase of the rNa V 1.4 current. Skeletal muscle sodium channels are phosphorylated by PKA despite lacking the large loop I-II that is the target of PKA in other sodium channels [26]. To detect potential PKA phosphorylation sites on rNav1.4, we performed a Scansite search that revealed four potential PKA sites with low percentile scores: T21, S56, S251, and S899. Both S251 and S899 are located in the hydrophobic transmembrane region of the channel protein, a region that is difficult to phosphorylate. Therefore, we chose to investigate the other two sites, S56 and T21. We designed mutated rNav1.4 channel proteins with single-point mutations at these sites and investigated the effects of AA/PGE 2 on these mutated channels.
The two site-directed mutants, rNa V 1.4S56A and rNa V 1.4T21A, were designed to have their S56 serine or T21 threonine residue replaced with an uncharged alanine residue. Fig 9A and 9B show that intracellular 10 μM db-cAMP significantly reduced the rNa V 1.4S56A current and rNa V 1.4T21A current to 4.06±2.67% (n = 18, P>0.05) and 23.65±7.72% (n = 13, P<0.05), respectively. Moreover, the effect of intracellular AA and PGE 2 on rNa V 1.4T21A and rNa V 1.4S56A was significantly diminished (Fig 10A and 10B). When 10 μM AA was delivered intracellularly via pipette, the rNa V 1.4T21A current amplitude was only increased by 2.6±2.4% (n = 6). Furthermore, application of 10 μM AA only slightly decreased the current amplitude of rNa V 1.4S56A by 22.7±4.3% (n = 4, P<0.05, compared with the effect of 10 μM AA on normal rNa V 1.4 with the Student's t test). This result was significantly different than the effect of intracellular application of 10 μM AA on normal rNa V 1.4 current. Similarly, intracellular application of 10 μM PGE 2 only increased the rNa V 1.4T21A current and rNa V 1.4S56A current by 7.2±4.5% (n = 4) and 2.8±1.5% (n = 11), respectively. These results are significantly different than the effect of intracellular application of 10 μM PGE 2 on normal rNa V 1.4 current (P<0.05, compared with the effect of 10 μM PGE 2 on normal rNa V 1.4 with the Student's t-test). Statistical analysis of these results is shown in Fig 10C and 10D.

Discussion
We previously reported that extracellular AA directly inhibited rNa V 1.4 current due to the activity of the AA molecule itself [23]. Here, our results indicate that intracellular AA increased the rNa V 1.4 current via metabolic pathways involving EP receptor-mediated activation of cAMP/PKA, and that the PKA phosphorylation site S56 on the rNa V 1.4 channel protein may play a pivotal role in this regulation. These observations were largely consistent with our previous data in neuronal I Na channels, which are mainly composed of the Na V 1.2 α subunit [22]. However, we observed that AA significantly modified the maximum activation potential and the steady-state activation of neuronal I Na , but not that of Na V 1.4, suggesting an isoform-specific effect. We revealed the pivotal role of the amino acid S56 in establishing the specific channel properties of Na V 1.4. We also demonstrated that PKA phosphorylation is mechanistically involved in the modulation of rNa V 1.4 by intracellular AA. AA is known to directly inhibit several ion channels, including neuronal Na + and skeletal Na + channels, as we have previously reported [7,10,[27][28][29][30]. This inhibitory effect is regarded as a consequence of the direct action of the AA molecule itself because 1) AA was applied extracellularly and 2) the inhibitory effects could be mimicked by EYTA, an AA analogue that cannot be metabolized to generate biologically active products. In this study, we delivered AA into the intracellular space by pipette solution and observed an augmented rNa V 1.4 current. This phenomenon is in contrast to the effects of extracellular AA [23]. The augmented current induced by intracellular AA could be mimicked by PGE 2 , a metabolite of AA produced by cyclooxygenase activity, but not by EYTA. Moreover, the increased rNa V 1.4 current could be eliminated by blocking AA conversion to PGE 2 via cyclooxygenase inhibition. In addition, while rNa V 1.4 currents were increased by intracellular AA, extracellular application of AA reduced rNa V 1.4 current amplitudes, suggesting that intracellular and extracellular application of AA act through different mechanisms to elicit different effects on rNa V 1.4 currents. Taken together, our data indicate that intracellular application of AA increases rNa V 1.4 current via its metabolic product, PGE 2 , as was previously reported for neuronal I Na [22]. These results further confirm that regulation of sodium channels by excessive intracellular AA occurs through the action of PGE 2 , a metabolite of AA, and that this regulation is a universal phenomenon, having been observed in both nerve tissue and skeletal muscle. PGE 2 is known to act through the G-protein coupled receptors EP1-4. Among the four EP receptors, only EP2 and EP4 active adenylate cyclase and induce cAMP elevation. EP3 activation can result in the inhibition or (rarely) elevation of cAMP levels, and EP1 is coupled to calcium mobilization [31]. We observed that HEK 293 cells mainly express three types of PGE 2 receptors. The increase in rNa V 1.4 current induced by intracellular AA/PGE 2 could be mimicked by intracellular cAMP, but not PKC activation or Ca 2+ release (data not shown), so we conclude that the mechanism of the PGE 2 effect involves EP2 and EP4 receptors. EP2/EP4 antagonists significantly eliminated the increased rNav1.4 current induced by intracellular AA/ PGE 2 , suggesting that intracellular AA is converted to PGE 2 , which activates the cAMP/PKA pathway via EP2/EP4 receptors. Although HEK 293 cells and cerebellar granule neurons [22] express different types of PGE 2 receptors, EP2 and EP4 are expressed in abundance by both cell types. Additionally, rNav1.4 currents and neuronal I Na in the two cell types respond similarly to intracellular AA/PGE 2 . These data suggest that EP2 and EP4 play an important role in the regulation of sodium channels by intracellular AA/PGE 2 . A previous study reported that sodium channels can be modulated by phosphorylation via the cAMP/PKA pathway because their α-subunits are the preferred substrates of PKA [25]. cAMP/PKA-dependent phosphorylation of rat brain voltage-gated sodium channels was reported to be restricted to five sites within the linker region between channel subunits I and II [32]. Although the sodium channel from rat skeletal muscle lacks this 210-amino acid linker alone on the rNa V 1.4 current, and the effect of PKAi on PGE 2 -induced increases in rNa V 1.4 current. Results are mean ± S.E.M. from four to eight cells. *, P<0.05 compared to PKAi group; #, P<0.05 compared to intracellular application of cAMP, using the Student's t test.
doi:10.1371/journal.pone.0140715.g008  region, its α-subunit is still reported to be an excellent substrate for PKA phosphorylation [26]. In this study, we investigated the specific amino acid residues that were responsible for modulation of rNav1.4 by intracellular cAMP and AA/PGE 2 . Based on our evaluation of the PKA phosphorylation sites predicted by Scansite, we generated two site-directed mutations in the N-terminal region of rNav1.4: rNa V 1.4S56A and rNa V 1.4T21A. rNa V 1.4S56A current showed no significant change when cAMP was applied. Moreover, the AA/PGE 2 -induced increase in rNav1.4 current was abolished by the S56A mutation. When AA was applied intracellularly to HEK 293 cells expressing rNa V 1.4S56A, current amplitudes were significantly inhibited, similar to the effect of extracellular AA on wild-type rNa V 1.4 channels. These data suggest that AA and PGE 2 can modulate sodium channels by different mechanisms, resulting in opposing effects. The rNa V 1.4T21A mutation, on the other hand, did not completely eliminate the cAMP-induced increase in current.
The rNa V 1.4T21A channel still responded to AA/PGE 2 application despite elimination of the predicted T21 phosphorylation site. Altogether, our results indicate that although rNav1.4 lacks a linker region enriched with PKA phosphorylation sites, its S56 residue plays a pivotal role in intracellular AA/PEG 2 -mediated augmentation of rNa V 1.4 current by activation of the cAMP/PKA pathway. These results are consistent with Yang and Barchi's report, in which they detected rat skeletal muscle Na + channel protein with [ 32 P]phosphate labeling and forskolin stimulation, and found that phosphorylation via the PKA pathway was limited to serine residues [26].
Phosphorylation of ion channels causes changes in gating, resulting in acute increases and decreases in current [33]. In this study, we observed that although AA/PGE 2 induced similar effects on neuronal I Na and rNa V 1.4 currents and activated the same pathways, their effects on the gating kinetics of steady-state activation were different in the two isoforms. AA/PGE 2 increased neuronal I Na amplitude by modulation of its steady-state activation [22]. However, neither steady-state activation nor steady-state inactivation was altered by intracellular AA/ PGE 2 in rNav1.4 channels. This difference might be due to the fact that the α subunits of neuronal Na + channels such as Na V 1.2 possess a loop I-II region that contains PKA phosphorylation sites, while rNa V 1.4 does not possess this region [25]. In addition, the observed differences may be due to the fact that in our study, we did not co-transfect the rNav1.4 β subunit, an accessory protein of the sodium channel that may modulate its gating kinetics. Additional studies are needed to determine the role of S56 in rNa V 1.4 channel activity and the role of rNav1.4 β subunits.
Na V channels are not only responsible for the initiation and propagation of action potentials in a wide variety of excitable cells including skeletal muscle, but also contribute to many other cellular processes, including apoptosis, motility, and secretory membrane activity [34][35][36]. Diseases such as paramyotonia congenita and hyperkalemic periodic paralysis are examples of well-studied genetic ion channel diseases affecting skeletal muscle Na + channels [37][38][39]. Massive AA release from membrane phospholipids and subsequent intracellular accumulation of free AA is a key cellular process in the control of inflammation and cell signaling [13]. A recent study indicated that heightened AA availability via supplementation enhanced endogenous PG synthesis and stimulated net skeletal muscle cell hypertrophy via a COX-2-dependent pathway [2]. Our study provides a new perspective to investigate the underlying mechanism of AA/ PGE 2 -mediated regulation of skeletal muscle Na + channels. Future work is needed to explore how AA/PGE 2 may contribute to physiological or pathological abnormalities in skeletal muscle cells.