Cyproheptadine Enhances the I K of Mouse Cortical Neurons through Sigma-1 Receptor-Mediated Intracellular Signal Pathway

Cyproheptadine (CPH) is a histamine- and serotonin-receptor antagonist, and its effects are observed recently in the modulation of multiple intracellular signals. In this study, we used cortical neurons and HEK-293 cells transfected with Kv2.1 α-subunit to address whether CPH modify neural voltage-gated K+ channels by a mechanism independent of its serotonergic and histaminergic properties. Our results demonstrate that intracellularly delivered CPH increased the I K by reducing the activity of protein kinas A (PKA). Inhibition of Gi eliminated the CPH-induced effect on both the I K and PKA. Blocking of 5-HT-, M-, D2-, H1- or H2- type GPCR receptors with relevant antagonists did not eliminate the CPH-induced effect on the I K. Antagonists of the sigma-1 receptor, however, blocked the effect of CPH. Moreover, the inhibition of sigma-1 by siRNA knockdown significantly reduced the CPH-induced effect on the I K. On the contrary, sigma-1 receptor agonist mimicked the effects of CPH on the induction of I K. A ligand-receptor binding assay indicated that CPH bound to the sigma-1 receptor. Similar effect of CPH were obtained from HEK-293 cells transfected with the α-subunit of Kv2.1. In overall, we reveal for the first time that CPH enhances the I K by modulating activity of PKA, and that the associated activation of the sigma-1 receptor/Gi-protein pathway might be involved. Our findings illustrate an uncharacterized effect of CPH on neuron excitability through the I K, which is independent of histamine H1 and serotonin receptors.


Introduction
Cyproheptadine (CPH) is typically used as an antihistamine. CPH is also widely applicable as an H 1 and 5-hydroxytryptamine (5-HT) receptor antagonist for the treatment of allergic reactions, as a prophylactic for migraines and in the symptomatic treatment of metastatic carcinoid syndrome [1]. As a first-generation antihistamine, CPH produces side effects such as sedation and reduced cognitive function [2]. Moreover, CPH is thought to resemble clozapine, which is an atypical antipsychotic that is used to treat schizophrenia [3]. Interestingly, the structure of CPH is similar to that of antidepressants such as amitriptyline, imipramine and N-methylamitriptyline [4]. These data suggest that CPH might play a role in the central nervous system (CNS). However, few studies have explored the cellular or molecular mechanisms by which CPH acts in the CNS or in primary neurons.
CPH possesses multiple pharmacological activities and acts as antagonist for serotonergic, dopaminergic, histaminergic, adrenergic or muscarinic receptors due to its relatively high binding affinity [5]. In addition to the previously mentioned receptordependent effects, CPH also directly inhibits ion channels, such as potassium (K + ), sodium (Na + ) and N-type and L-type calcium (Ca 2+ ) channels in cardiac cells [6]. In addition, it has recently been reported that CPH induces apoptosis in leukemia cells, which is independent of its known activity as a histamine and serotonin receptor antagonist [7]. Similarly, the pancreatic b-cell toxicity that results from CPH exposure appears to be unrelated to its serotonergic or histaminergic properties because pharmacologically inactive metabolites of CPH also exhibit the same toxicity [8]. Moreover, recent studies have suggested that the diabetogenic effect of CPH may be at least partially due to altered regulation of protein synthesis [9]. These data indicate CPH effect independent of serotonergic or histaminergic receptors. Perhaps, CPH may act as an intracellular signaling molecule by binding to intracellular sites, such as one reported on sigma-1 receptor [10]. Notably, previous studies have indicated that sigma-1 receptor is widely distributed in CNS neurons and is associated with multiple intracellular signaling pathways [11]. Therefore, the serotonergic or histaminergic receptor independent effects mediated by CPH may be channeled, directly or indirectly, by targeting the sigma-1 receptor.
Among the voltage-gated ion channels, K + channels are extremely diverse in structure and function and are one of the most important signaling macromolecules in both neuronal and non-neuronal cells. In the CNS, the activity of K + channels determines the frequency and duration of action potentials [12]. Modifying K + channels can alter neuronal excitability and modulating neurotransmitter release [13]. Inhibitors and activators of K + channels have been proposed as potential therapies for a diverse array of neurological and cardiovascular disorders including stroke, hypertension, mental illness and others [14].
Although CPH has previously been reported to directly inhibit ion channels, few studies have addressed the possibility that CPH modifies the activity of neuronal voltage-gated K + (Kv) channels by a mechanism unrelated to its serotonergic and histaminergic properties.
In this study, we investigated the effects of CPH on the delayed rectifier outward potassium current (I K ) in primary cultures of cortical neurons using the whole-cell patch-clamp technique. The I K was significantly enhanced by the intracellular application of CPH through its modulation of the sigma-1 receptor via cAMP/ PKA signaling to specifically regulate Kv2.1 a-subunit.

Results
Both the Intracellular and Extracellular Application of Cyproheptadine Enhance the I K Cortical neurons display two main voltage-dependent outward K + currents, the delayed rectifier potassium current (I K ) and the transient potassium current (I A ), which can be specifically blocked by 20 mM TEA and 5 mM 4-AP, respectively. We first determined whether CPH affected either the I A or the I K using corresponding voltage protocols and specific K + channel blockers. In the presence of 5 mM TEA, the I A was obtained by depolarizing pulses to +40 mV from a holding potential of 2100 mV, and the intracellular application of CPH did not affect the I A (Fig. 1A, left). When 4-AP was added to the bath solution, the I K was evoked by 200 ms of depolarization to +40 mV from the holding potential of 250 mV, and the intracellular application of CPH significantly enhanced the I K (Fig. 1A, right). These data suggest that CPH selectively modulates the I K .
CPH is a lipid-soluble and membrane-permeable molecule, and its effect as a H 1 and serotonin receptor antagonist has recently been reported [7]. We first examined the intracellular effect by the addition of CPH via a micropipette. The I K of cortical neurons reached its maximum amplitude within 30-60 seconds. This CPH-induced effect on the I K was concentration-dependent from 20 nM to 20 mM of CPH. The increases in the I K induced by 20 nM, 200 nM, 2 mM, 10 mM and 20 mM of CPH were 3.161.1% (n = 7), 12.2161.37% (n = 8), 26.1662.19% (n = 7), 36.162.01% (n = 9) and 52.4861.91% (n = 8), respectively. The statistical analysis of these concentration-response data is presented in Fig. 1B. We also tested whether the extracellular application of CPH affected the I K . The results showed that applying CPH by extracellular perfusion induced an increase in the I K that was similar to the intracellular application. However, the maximum effect was observed after 3 min (Fig. 1C). Therefore, to obtain a rapid and stable effect, a CPH solution was applied intracellularly using a micropipette in the following electrophysiology experiment.
Because the I K increase can be modified by membrane channel protein dephosphorylation [15], alkaline phosphatase (ALP) was used to investigate whether the effect of CPH on the I K might be due to a CPH-induced dephosphorylation of the I K channel. As shown in Fig. 2B, ALP (200 unit/L) alone increased the I K by 29.861.4% (n = 6), a result similar with CPH shown in Fig. 2A. Previous studies suggested that protein phosphatase 1 (PP1) contributes to the maintenance of constitutive phosphorylation of Kv2.1 subunit of I K channel in cortical neurons [16]. We used okadaic acid, the PP1 inhibitor, to test this hypothesis. The result presented in Fig. 2C showed that okadaic acid (200 nM) alone increased the I K by 4.9262.0% (n = 6), and in the presence of okadaic acid the 10 mM CPH-induced increase in the I K was significantly reduced from 36.162.01% (CPH alone) to 4.5962.0% (n = 6). A statistical analysis of these data (Fig. 2D) suggested that the effect of CPH on the I K was not direct but might be instead mediated by decreasing K + channel phosphorylation. I A and I K traces recorded intracellularly in the absence and presence of 10 mM cyproheptadine (CPH). The I A was obtained by the application of depolarizing pulses to +40 mV from a holding potential of 2100 mV, and the I K currents were evoked with 200 ms depolarizing pulses to +40 mV from a holding potential of 250 mV. (B) A statistical analysis of the concentration-dependent effect of CPH on the I K values. The data are reported as the mean 6 S.E.M. from 6-10 cells. *P,0.05 when comparing different concentrations of CPH by a one-way ANOVA. (C) The time-course of the I K changes induced by the extracellular application of 10 mM CPH. The insets in the graphs show superimposed I K traces from the initial control levels (after the establishment of the whole-cell configuration) and after the external infusion of CPH. The time points (1, 2, and 3) noted on the curves correspond to the superimposed I K traces illustrated by the insets. doi:10.1371/journal.pone.0041303.g001 The cAMP/PKA Pathway Contributes to the Cyproheptadine-mediated I K Increase Because the PKA pathway plays an important role in channel phosphorylation, the effects of CPH were studied in the presence of forskolin, dibutyryl cAMP (db-cAMP) and Rp-cAMP to address whether the cAMP/PKA pathway contributed to the CPHmediated increase of the I K . Forskolin or db-cAMP was applied to the bath solution to activate adenylate cyclase or PKA, respectively. Perfusion of cortical neuron with 10 mM forskolin provoked a gradual decrease in I K amplitude by 22.963.8% (n = 5), and in the presence of 10 mM forskolin, CPH increased the I K amplitude by only 4.961.0% (n = 8) (Fig. 3A), which is significantly less than CPH alone (36.162.01%). The administration of db-cAMP also eliminated the effect of CPH on the I K (Fig. 3B). A uniform perfusion of 20 mM db-cAMP, produced a significant reduction of 23.462.1% in I K currents, and abolished the CPH-induced inhibitory effect on I K . In the presence of 20 mM db-cAMP, CPH only increased the I K by 4.360.9% (n = 7), which is similar to the forskolin results. However, the application of Rp-cAMP, a membrane-permeable specific PKA inhibitor, mimicked the effect of CPH and increased the I K amplitude by 24.762.32% (n = 5). In the presence of both Rp-cAMP and 10 mM CPH, the I K was further increased only slightly by 2.860.6% (n = 8) relative to the baseline levels (Fig. 3C). Taken together, these results indicate critical contribution of the cAMP/PKA pathway in the CPHinduced I K increase. A statistical analysis of these data is shown in Fig. 3D.
Next, western blot was used to measure the levels of phosphorylated PKA (pPKA) in response to CPH application. The results demonstrated that there was a significant and dosedependent decrease in the pPKA levels of cortical neurons that were incubated with CPH for 5 min (Fig. 4A). In the presence of 2 mM, 10 mM and 20 mM CPH, the activity of pPKA were significantly decreased to 81.366.9%, 76.566.1%, and 71.264.4% of the control values (n = 4), respectively. The results of the quantitative statistical analysis of the pPKA activity are presented in Fig. 4B.

CPH Targets G i Protein-coupled Receptor Signaling
Through Sigma-1 Receptor G i negatively couples with the cAMP/PKA pathway; therefore, selective blockers of G i were used to determine whether the G i protein was associated with the CPH-induced increase in I K and the decrease in PKA activity in cortical neurons. Our results indicated that pre-incubation of cortical neurons with selective blockers of G i (pertussis toxin (PTX) and NF023) significantly attenuated the CPH-induced increase in the I K (Figs. 5A and 5B). 200 ng/ml PTX and 20 mM NF023 alone did not alter the I K amplitude (Fig. 5C). After the cortical neurons were incubated with either 200 ng/ml PTX for 2 hours or 20 mM NF023 for 5 min, the addition of 10 mM CPH only increased the I K by 7.3160.77% (n = 9) and 3.7360.84% (n = 7), respectively. These values were significantly different from those obtained with CPH alone (36.162.01%, Fig. 5C). In addition to affecting the I K recordings, blocking G i with PTX and NF023 also reduced the CPH-induced inhibitory effect on the intracellular pPKA levels. After the cortical neurons were incubated with either 200 ng/ml PTX for 2 hours or 20 mM NF023 for 5 min, 10 mM CPH only decreased the pPKA levels by 2.463.7% and 3.164.3% (n = 6), respectively, which is significantly different from the decrease seen with CPH alone (24.566.1%, Fig. 5D).
Although we applied CPH intracellularly to avoid any receptordependent effects, CPH is a lipid-soluble and membranepermeable molecule and has been used as a serotonin and H 1histamine receptor antagonist [17]. We hypothesized that G i protein-coupled receptors might be involved in the CPH-induced increase in the I K . As shown in Fig. 6A, we administered commonly used G i anatgonists to test this hypothesis. Antagonists of serotonin receptors (risperidone), D 2 -dopamine receptors (sulpiride), muscarinic receptors, H 1 -histamine receptors (orphenadrine) and H 2 -histamine receptors (cimetidine) did not eliminate the CPH-induced I K increase. In the presence of risperidone (20 mM), sulpiride (20 mM), orphenadrine (50 mM) and cimetidine (20 mM), 10 mM CPH increased the I K by 33.4165.19% (n = 5),32.2763.2% (n = 6), 28.7461.28% (n = 6) and 30.1462.67% (n = 5), respectively, which is similar to the effect seen with CPH alone (36.162.01%, Fig. 6A) However, when cortical neurons were incubated with a sigma-1 receptor antagonist (BD1063), the effect of CPH was abolished. The addition of 20 mM BD1063 reduced the 10 mM CPH-induced increase in the I K to 3.460.8% (n = 9), which indicated that the sigma-1 receptor might be involved (Fig. 6A).
Previous studies have indicated that the sigma-1 receptor is widely distributed in CNS neurons and is associated with multiple intracellular signaling pathways [11]. Therefore, a specific agonist and an antagonist of the sigma-1 receptor were used to explore whether the intracellular application of CPH would activate the sigma-1 receptor and induce an increase in the I K . When the sigma-1 receptor was blocked by 20 mM BD1063, the CPHinduced increase in the I K was reduced to 3.460.8% (n = 9) (Fig. 6B). In contrast, the sigma-1 receptor agonist PRE-084 mimicked the effect of CPH on the I K and elicited a gradual increase in the I K amplitude (Fig. 6C). The application of 20 mM PRE-084 increased the I K by 26.863.7% (n = 9), which is similar to the effect produced by CPH (Fig. 6D). Similar with the observation for CPH, okadaic acid also reduced PRE-084-induced I K increase (Fig. 6D). In the presence of 200 nM okadaic acid, 20 uM PRE-084 only increase I K by 6.9460.92% (n = 6).
We also measured the effect of a sigma-1 receptor agonist and antagonist on the activity of pPKA. Consistent with the I K recording, activating the sigma-1 receptor with PRE-084 induced a significant decrease in the intracellular levels of pPKA, and blocking the sigma-1 receptor with BD1063 significantly suppressed the PRE-084-induced decrease of pPKA levels by 4.163.2% (n = 6). Moreover, blocking G i with PTX and NF023 significantly reduced the inhibitory effect of PRE-084 on pPKA levels (Fig. 7A). After the cortical neurons were incubated with 200 ng/ml PTX for 2 hours or 20 mM NF023 for 5 min, 20 mM PRE-084 only decreased the pPKA levels by 5.4163.3% and 3.662.7% (n = 6), respectively. This result was similar to that obtained by the application of CPH alone.
To further confirm that the CPH-induced increase of the I K was associated with the sigma-1 receptor, the affinity of CPH for the sigma-1 receptor was determined by a competitive-binding assay. Sigma-1 receptors were labeled in rat liver homogenates using the radioactive sigma-1 receptor ligand [ 3 H]-(+)-pentazocine (5 nM). Figure 7C shows the competitive binding curves of CPH and haloperidol against [ 3 H]-(+)-pentazocine. The IC 50 value (50% inhibition of the specific binding of [ 3 H]-(+)-pentazocine) for CPH was 2.4960.63 mM, which is in line with the concentration used in the electrophysiology and PKA activity assays. The dose-response curve showed that CPH interacted with the sigma-1 receptor in a competitive manner. Furthermore, the results also showed that the K i of CPH was 0.9360.23 mM, which is consistent with other sigma-1 receptor ligands [10,11]. The K i values were derived from the IC 50 values according to the equation K i = IC 50 /(1+ C/Kd) [18], where C is the concentration of the radioligand [ 3 H]-(+)pentazocine (5 nM) and Kd (3 nM) is the dissociation constant of the corresponding [ 3 H]-(+)-pentazocine binding to the sigma-1 receptor in rat liver homogenates [18].
We further investigated whether the CPH-induced increase in the I K was dependent on the sigma-1 by knocking down the expression of sigma-1 receptor in cortical neurons with interfering RNAs (siRNAs). A plasmid containing a specific sigma-1 receptor siRNA along with enhanced green fluorescent protein was transfected into the cultured mouse cortical neurons. Green fluorescence was observed 48 h after transfection. To avoid a background effect of the transfection process on the current amplitude and CPH, a group of cells that had been transfected with a random sequence and EGFP were used as a negative control. Figures 8A show the typical current recordings in a cortical neuron transfected with siRNA against sigma-1 receptor. When the sigma-1 receptor expression was knocked down, the effect of CPH on the I K was almost eliminated, and the addition of 10 mM CPH only increased the I K by 6.0961.7% (n = 8). The CPH-induced increase of the I K recorded from the eGFP control group was similar to that of the untransfected cells. A statistical analysis of these data is shown in Fig. 8B. Concurrently, the efficiency of the siRNA in HEK-293 cells that endogenously express the sigma-1 receptor was also analyzed. A Western blot showed that siRNAs against the sigma-1 significantly inhibited the level of protein that was expressed in the HEK-293 cells (Figs. 8C). This result suggested that both the sigma-1 receptor is needed to see the effect of CPH on the I K .
The Kv2.1 a-subunit is Involved in the Cyproheptadineinduced I K Increase It has been previously reported that the Kv2.1 a-subunit, which is expressed at high levels in most mammalian CNS neurons, is a major contributor to the I K channels and plays a crucial role in regulating neuronal excitability [19]. To address the potential role of Kv2.1 in the CPH-induced increase of I K channel activity, Jingzhaotoxin-III (JZTX-III), which has been previously demonstrated to specifically inhibit the Kv2.1 a-subunit in HEK-293 cells, was used [20]. 100 nM JZTX-III significantly inhibited 87.2%62.12% (n = 5) of the Kv2.1 transfected HEK293 cells (Fig. 9A). An inhibition of the I K amplitudes by 28.564.4% (n = 5) was also observed in primary cortical neurons after 100 nM JZTX-III was applied to the bath solution (Fig. 9B). In the presence of 100 nM JZTX-III, the increase in the I K that was induced by CPH decreased to 8.263.9% (n = 7), which indicates that the elimination of Kv2.1 activity prevents the CPH-induced increase of the I K in cortical neuron ( Fig. 9C and 9F). Using HEK-293 cells, which endogenously express the sigma-1 receptor and were transfected with the exogenous Kv2.1 a-subunit, the experiments that had already been performed in cortical neurons were repeated. As the control, 10 mM CPH did not affect the small endogenous background current of HEK293 cells which was only transfected with EGFP (Fig. 9D). However, 10 mM CPH significantly increased the Kv2.1 currents in the HEK-293 cells by 36.3263.43% (Fig. 9E and 9G, n = 12). Furthermore, when G i was blocked by 200 ng/ml PTX or 20 mM NF023, the application of 10 mM CPH only increased the Kv2.1 currents by 5.2460.98% (n = 10) and 4.3161.18% (n = 6), respectively. When both the expression of the Kv2.1 a-subunit and either the sigma-1 receptor or the was knocked down in the HEK-293 cells, the CPH-induced increase in Kv2.1 currents was reduced to 5.4661.32% (n = 5) and 5.6860.47% (n = 8), respectively. A statistical analysis of these data is shown in Fig. 9G.

Discussion
Recent studies have examined the effects of CPH independent of its function as a histamine and serotonin receptor antagonist [7,9]. In this study, we investigated the effects of CPH on the I K , and the underlying receptor or signal mechanism involved in its effects on the I K . We demonstrated for the first time that CPH enhances the I K in cultured cortical neurons through a G idependent cAMP/PKA pathway via activation of the sigma-1 receptor. Ion channels as membrane proteins can be modulated through direct protein-protein interactions or by indirect, membranebound, receptor-mediated phosphorylation or dephosphorylation of the channels. Previous studies have shown that CPH directly inhibits K + , Na + and Ca 2+ channels in cardiac cells [6]. In this study, PTX treatment reduced I K responses to CPH, and cells perfused internally with NF023, which is known to be a very potent inhibitor of G i protein-dependent processes [21], eliminated the CPH-induced increase in the I K . Likewise, either blocking constitutive phosphorylation of I K channel with the PP1 inhibitor [22] or inhibiting PKA activity with Rp-cAMP interfered with the CPH-induced increase in I K . Moreover, the measured changes in PKA activity suggest that CPH inhibits the cAMP/PKA pathway. These findings therefore suggest a novel mechanism by which CPH increases voltage-dependent K + channels in cortical neurons that involves both G i protein-coupled receptors and cAMP/PKA inhibition. Since CPH did not alter the gating property of I K , increase of I K might caused by an increase in current density or channel opening rate.
CPH possesses multiple pharmacological activities due to its relatively high affinity for various receptors, including serotonergic, dopaminergic, histaminergic, adrenergic and muscarinic receptors [23]. In this study, we first attempted to avoid the serotonin, dopamine, or histamine receptor-dependent effects by applying CPH intracellularly. In addition, our results demonstrated that the specific receptor antagonists (against serotonin, dopamine, histamine and muscarinic receptors) that are associated with CPH's known activity as well as the G i protein-coupled receptor antagonists all failed to interfere with the CPH-induced increase of the I K . In contrast, applying a sigma-1 receptor antagonist or knocking down the sigma-1 receptor in cortical neurons eliminated the effect of CPH on the I K . Interestingly, the competitive binding assay using the liver tissues which express high abundant of sigma-1 receptor [24] showed that CPH can act as a ligand for the sigma-1 receptor. In our experimental binding system, the K i of haloperidol for the sigma-1 receptor was similar to the results reported by Klouz et al [18]. The sigma-1 receptor is widely expressed in the CNS and is present at high levels in the cortex, the hippocampus and the striatum [25]. It binds to a vast number of synthetic compounds whose features include a common N-substituted trace amine [11]. Because CPH also has this Nsubstituted trace amine, and the sigma-1 receptor binding site is located intracellularly [10], it is highly likely that intracellular CPH might rapidly bind to the sigma-1 receptor.
Given sigma receptors and ion channels are distinct proteins and the cloned sigma-1 receptor does not have the typical structure of a G protein-coupled receptor [26], a relay mechanism will be required to transmit the signal of the target ion channels when the sigma-1 is activated. Earlier studies in melanotroph cells have demonstrated a Gs-dependent modulation of Kv channels by  sigma receptors [27], but whether G protein-coupled receptor is involved or not remain unknown. Kim et al. have recently demonstrated a physical and functional association of the sigma-1 receptor with the m-opioid receptor [28]. They also established that the sigma-1 receptor can directly associate with the m-opioid receptor and that this association allows selective antagonists of the sigma-1 receptor to potentiate m-opioid receptor-induced cell signaling. Coincidentally, the m-opioid receptor is a heterotrimeric G protein-coupled receptor, and m-opioid receptor activation has been shown to inhibit the cAMP/PKA pathway by G i activation [29]. In this study, we also experimentally observed that knocking down m-opioid receptor expression in cortical neurons interfered with the CPH-sigma-1 receptor-mediated increase in the I K . (data not shown) Moreover, activating m-opioid receptor with agonist mimicked the effect of CPH on the I K while the antagonist of mopioid receptor failed (data not shown). It is thus highly likely that both the sigma-1 receptor and the m-opioid receptor were needed to see the effect of CPH on the I K of cortical neurons. However, further study is needed for conforming whether both of sigma-1 and m-opioid receptor was functional coupled in cortical neuron, and CPH as sigma-1 ligand in modulating the m-opioid receptorinduced signaling.
In most studies, the modulation of voltage-gated K + and Ca 2+ channels by the sigma-1 receptor does not involve transduction mechanisms, such as G protein signaling or phosphorylation; instead, a direct interaction between the two proteins has been suggested [30][31][32]. In support of this idea, both Kv1.4 and the sigma-1 receptor can be coimmunoprecipitated from membrane lysates that have been prepared from both rat posterior pituitary cells and mRNA-injected Xenopus oocytes. However, the effect of sigma-1 receptor ligands on voltage-dependent ion channels has been inhibitory in almost all of these previous reports; additionally, most of these studies have been conducted in non-neuronal cells or in Xenopus oocytes [11,30,33]. In contrast, our study found that CPH enhanced the I K via the activation of the sigma-1 receptor and the m-opioid receptor and the inhibition of cAMP/PKA pathway, which is different from previous reports. This discrepancy might result from the cell type tested due to the presence or absence of the endogenous m-opioid receptor. It also might be due to the types of K + channels that were studied. Native neuronal I K channels are mainly composed of members of the Kv2 subfamily, which are easily modified by the cAMP/PKA pathway through phosphorylation or dephosphorylation mechanisms [34][35][36]. This process may explain why the CPH-induced activation of the sigma-1 receptor and the subsequent modulation of the I K in cortical neurons were associated with an indirect mechanism involving the activation of both the m-opioid receptor/G i -protein and the cAMP/PKA pathway in this study. In addition to the mopioid receptor, previous studies by Mavlyutov et al. showed colocalization of both the Kv2.1 potassium channels and the muscarinic m2AChR with the sigma-1 receptor in mouse motoneurones [37,38]. The resolution of the data, as pointed out by Mavlyutov et al., is limited and the detected Kv2.1 potassium channels, the muscarinic m2AChR and the sigma-1 receptor may lie in different membranes. Further functional identification of their direct interactions is warranted.
Electrophysiological studies have revealed that most mammalian neurons express multiple types of voltage-gated K + channels with distinct time-and voltage-dependent properties. The Kv2 (Shab) genes in many species encode components of potassium channels with sustained currents. In particular, the Kv2.1 subunit, which is highly expressed in most mammalian CNS neurons, is a major contributor to the I K channels and plays a crucial role in regulating neuronal excitability. Previous studies and recent works from our laboratory have confirmed by that Kv2.1 is a major component of cortical neuronal I K channels.
In this study, we also reported that CPH might target the Kv2.1 a-subunit, as revealed by the use of a specific toxin (JZTX-III) that has been reported to selectively block Kv2.1 a-subunits [39]. We further ascertained the specificity in a heterologous system using HEK-293 cells, which endogenously express the sigma-1 receptor [11] and the m-opioid receptor [7], by transfecting an expression vector containing the Kv2.1 a-subunit. The results were similar to those observed in the primary cortical neurons as we observed increased Kv2.1 currents by CPH in the HEK-293 cells transfected with the Kv2.1 a-subunit. CPH, however, did not affect Kv2.2 currents in transfected HEK-293 cells (data not shown). Moreover, CPH-induced enhancement of the Kv2.1 channels in transfected HEK-293 cells was also eliminated after blocking G i proteins with PTX and NF023 or knocking down the expression of the sigma-1 receptors. The overall results not only indicated that the Kv2.1 a-subunits were primarily affected by CPH but also supported the idea that CPH enhanced the I K via the activation of the sigma-1 receptor/G i protein and cAMP/PKA pathways.
Although previous studies have demonstrated that CPH modulates membrane ion channels in the kidney and cardiovascular system [6], there have been few studies examining the influence of CPH on voltage-gated K + currents in the nervous system. Because voltage-gated K + channels play a crucial role in maintaining the resting membrane potential of the neuron as well as the duration and frequency of action potentials, modulators of these channels would likely have therapeutic effects in various neurological and psychological disorders [40]. For example, K V channel blockers are expected to restore normal firing in depression and cognitive dysfunction, while K V channel activators should be useful in reducing the pathological hyperexcitability that occurs in epilepsy [41]. In addition, because the sigma-1 receptor has been linked to the modulation of processes involving memory and learning, depression, and psychosis [42], CPH's association with the sigma-1 receptor and the subsequent increase in I K channel activity may contribute to its apparent clinical efficacy as an antidepressant and an antipsychotic.
A proposed model illustrating the signaling cascade that might be associated with CPH in cortical neurons is shown in Fig. 10. We propose that CPH, acting through its association with the sigma-1 receptors at an intracellular binding site, increases the I K via G i activity and cAMP/PKA pathway inhibition.

Ethics Statement
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of the Fudan University (Permit Number: 2007-0002). All surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering.

Cultures of Mouse Cortical Neurons
ICR (Swiss Hauschka) mice were purchased from the Shanghai Laboratory Animal Center at the Chinese Academy of Sciences (Shanghai, China). Primary neuronal cultures were prepared from the cerebral neocortex of 15-day-old embryonic mice as originally described. Briefly, the cerebral neocortices were dissected from embryonic ICR mice. The dissociated cortical neurons were then plated onto poly-L-lysine-coated (10 mg/ml) 35-mm dishes (Costar, Corning, NY) at a density of 1610 5 cells per dish. The cultured cells were incubated at 37uC with 5% CO 2 in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 10% fetal calf serum, 10% heat-inactivated horse serum and a 1% antibioticantimycotic solution. After 24 hours in culture, cytosine 1-b-Darabinofuranoside (5 mM) was applied to the culture medium for 48 hours to eliminate actively proliferating fibroblasts. On day 4, the cultured cells were washed and maintained in fresh medium. All of the experiments were conducted with cortical neurons that had been in culture for 6-10 days.

Patch-clamp Recordings
Whole-cell currents were recorded in cortical neurons using a patch-clamp technique. Prior to the I K recordings, the culture medium was replaced with a bath solution containing 140 mM NaCl, 2.5 mM KCl, 10 mM HEPES, 1 mM MgCl 2 , and 0.001 mM TTX. The pH was adjusted to 7.4 using NaOH. Soft glass recording pipettes were filled with an internal solution containing 135 mM K gluconate, 10 mM KCl, 10 mM HEPES, 1 mM CaCl 2 , 2 mM MgCl 2 , 2 mM ATP?Mg, 0.1 mM GTP?Na 3 , and 10 mM EGTA. The pH of this solution was adjusted to 7.3 using KOH. The pipette resistance was 5-7 MV after it was filled with the internal solution. All of the recordings were performed at room temperature (23-25uC).

DNA Constructs and Cell Transfection
Total RNA was isolated from the primary cultures of mouse cortical neurons using a commercial kit according to the manufacturer's instructions (Qiagen Mini RNeasy, Qiagen, Valencia, CA, USA). First-strand synthesis was performed using SuperScript II reverse transcriptase (Invitrogen, Carlsbad, CA, USA). The following primer set was designed to amplify mouse Kv2.1: a forward primer (59 TGGCTCGAGATGCCGGCGGG-CATG 39) that included an XhoI site and a reverse primer (59 ATACAGAATTCGGATACTCTGATCCCT 39) that included an EcoRI site (GenBank NM_008420). Mouse Kv2.1 cDNA was ligated into the pll3.7 vector using the XhoI and EcoRI restriction sites. Each gene was fused to the N-terminus of EGFP, which was used as a fluorescent marker to identify transfected human embryonic kidney (HEK)-293 cells which purchased form the cell bank of Chinese Academy of Science (Shanghai, China). All of the constructs were verified by DNA sequencing. The plasmids were extracted using a Qiagen plasmid midi kit (Qiagen, Valencia, CA, US). The DNA concentration and purity of each plasmid were determined by measuring the absorbances at 260 and 280 nm, respectively. HEK-293 cells were transfected using the calcium phosphate method. The average transfection efficiency was above 80%. Two days after transfection, the HEK-293 cells were examined for green fluorescence.

RNA Interference Knockdown of Sigma-1 and m-opioid Receptors
The plasmid used to silence the sigma-1 receptor was constructed using the pGPU6/GFP/Neo siRNA vector (Gene-Pharma, Shanghai, China). An siRNA sequence corresponding to nucleotides 500-519 of the human sigma-1 receptor open reading frame (PubMed nucleotide ID: NM005866), which is identical to the mouse sequence in this region, was inserted into the pGPU6/ GFP/Neo vector, and the plasmid was transfected into both primary mouse cortical neurons and mammalian HEK-293 cells. As a negative control, a random siRNA sequence was inserted into the pGPU6/GFP/Neo vector. To silence the m-opioid receptor gene, a plasmid was constructed using the pll3.7 siRNA vector. The siRNA sequence corresponding to nucleotides 392-409 of the mouse MOR-1 (ID: NM_001039652) was inserted into the pll3.7 vector, and the plasmid was transfected into both cortical neurons and HEK-293 cells. Transfection was accomplished using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) and Opti-mem media (Gibco, Grand Island, NY, USA) according to the manufacturers' instructions. Sigma-1 receptor and m-opioid receptor siRNAs were cotransfected with enhanced green fluorescent protein (EGFP) to label the transfected cells. Sigma-1 and mopioid receptor protein levels were detected by an immunoblot with antibodies against both proteins. Anti-Sigma-1 receptor antibody was a gift from Dr. Teruo Hayashi's lab and anti-mopioid receptor antibody was purchased from Millipore (Millipore, Billerica, MA).

Ligand-receptor Binding Assay
Membranes from Sprague-Dawley rat (150-200 g) livers were prepared for the sigma-1 receptor binding assay using a previously described method [24] with minor modifications. The livers were quickly processed and homogenized with a glass homogenizer (8-10 up and down strokes) in 6 ml/g of cold buffer (0.32 M sucrose in 50 mM Tris, pH 7.4) and then homogenized with a Tissue-Tearor (Biospec Products, Inc.). The cell suspension was centrifuged at 1200 g for 10 min at 4uC, and the supernatant was decanted and recentrifuged at 25000 g for 20 min at 4uC. The resulting pellet was then resuspended in 5-6 ml/g of 50 mM Tris (pH 7.4) and centrifuged at 25000 g for 20 min at 4uC. This procedure was repeated twice. The pellet was resuspended in a final volume (5-6 ml/g) of buffer and stored at 280uC until use. The membrane preparation contained approximately 2.5 mg protein/ml, as determined by the Lowry method [44].
The affinity of CPH for the sigma-1 receptor was determined by a competitive binding assay. The membrane preparations, which contained approximately 250-400 mg of protein, were incubated in duplicate for 180 min at 30uC with various concentrations of CPH (1 nM-100 mM) and 5 nM [ 3 H]-(+)-pentazocine (28 Ci/ mmol; Perkin-Elmer, Boston, MA, USA) in a total volume of 200 ml of binding buffer (50 mM Tris and 4 mM MgCl 2 , pH 7.4). The specific binding was assayed in the presence of 10 mM haloperidol. The reaction was stopped by rapid filtration through a Whatman GF/B glass fiber filter and subsequent washing with cold buffer (50 mM Tris and 5 mM EDTA, pH 7.4) using a Brandel 24-well cell harvester. The filters were then dried at 80uC for 30 min in a drying oven and immersed in a scintillation cocktail. The radioactivity of each filter was determined with a MicroBeta liquid scintillation counter. The IC 50 values (i.e., the concentration of unlabeled test compound needed to cause a 50% inhibition of the specific binding of the radiolabeled ligand) from these assays were calculated by nonlinear regression (PRISM, Graphpad, San Diego, CA, USA) using a sigmoidal function. Inhibition constant (K i ) values were calculated from the IC 50 values according to Cheng and Prusoff [45].

Data Acquisition and Analysis
All of the currents were recorded using an Axopatch 200B amplifier (Axon Instruments, Foster City, CA, USA) that was operated in voltage-clamp mode. A Pentium computer was connected to the recording equipment with a Digidata 1300 analog-to-digital (A/D) interface. The current was digitally sampled at 100 ms (10 kHz). The current signals were filtered by a 5-kHz, five-pole Bessel filter. The currents were corrected online for leak and residual capacitance transients by a P/4 protocol. The data acquisition and analysis were performed with pClamp 8.01 software (Axon Instruments, Foster City, CA, USA) and/or Origin6.1 (MicroCal, Northampton, MA, USA). The statistical analysis was performed using Student's t-test with non-paired or paired comparisons when relevant. The values are given as the mean 6 S.E.M. with n representing the number of cells tested. A value of P,0.05 was considered a significant statistical difference between groups. When multiple comparisons were made, the data were analyzed by a one-way ANOVA.