(−)-Naringenin 4′,7-dimethyl Ether Isolated from Nardostachys jatamansi Relieves Pain through Inhibition of Multiple Channels

(−)-Naringenin 4′,7-dimethyl ether ((−)-NRG-DM) was isolated for the first time by our lab from Nardostachys jatamansi DC, a traditional medicinal plant frequently used to attenuate pain in Asia. As a natural derivative of analgesic, the current study was designed to test the potential analgesic activity of (−)-NRG-DM and its implicated mechanism. The analgesic activity of (−)-NRG-DM was assessed in a formalin-induced mouse inflammatory pain model and mustard oil-induced mouse colorectal pain model, in which the mice were intraperitoneally administrated with vehicle or (−)-NRG-DM (30 or 50 mg/kg) (n = 10 for each group). Our data showed that (−)-NRG-DM can dose dependently (30~50 mg/kg) relieve the pain behaviors. Notably, (−)-NRG-DM did not affect motor coordination in mice evaluated by the rotarod test, in which the animals were intraperitoneally injected with vehicle or (−)-NRG-DM (100, 200, or 400 mg/kg) (n = 10 for each group). In acutely isolated mouse dorsal root ganglion neurons, (−)-NRG-DM (1~30 μM) potently dampened the stimulated firing, reduced the action potential threshold and amplitude. In addition, the neuronal delayed rectifier potassium currents (IK) and voltage-gated sodium currents (INa) were significantly suppressed. Consistently, (−)-NRG-DM dramatically inhibited heterologously expressed Kv2.1 and Nav1.8 channels which represent the major components of the endogenous IK and INa. A pharmacokinetic study revealed the plasma concentration of (−)-NRG-DM is around 7 µM, which was higher than the effective concentrations for the IK and INa. Taken together, our study showed that (−)-NRG-DM is a potential analgesic candidate with inhibition of multiple neuronal channels (mediating IK and INa).


Introduction
Pain is an unpleasant sensory and emotional experience associated with actual or potential tissue damage [1]. It can be subdivided into somatic pain and visceral pain according to the originating tissue, while it can be categorized into acute and chronic pain based on the ongoing time [2]. As a rising health problem, chronic pain is predicted to affect up to 30% of adults worldwide, and about 70% of patients are refractory to the current

Analgesic Effects of (−)-NRG-DM in Formalin-Induced Mouse Inflammatory Pain Model
The formalin test is a classical inflammatory pain model which represents somatic pain. Intraplantar injection of formalin results in a typical two-stage nociceptive behavior, which is characterized by licking and biting of the injected paw. The first phase (0 ~ 10 min) mainly reflects nociceptive pain, while the second phase (11 ~ 60 min) represents the inflammatory responses [36]. In our study, we tested the effects of (−)-NRG-DM on formalin-induced inflammatory pain in mice. (−)-NRG-DM at 30 mg/kg and 50 mg/kg body weight were individually intraperitoneally administrated 30 min before injection of 1% formalin solution. Consequently, (−)-NRG-DM significantly attenuated painful behaviors in a dose-dependent manner during phase I and phase II (Figure 2). At the dose of 50 mg/kg, (−)-NRG-DM significantly attenuated painful behaviors, including the licking time and overall pain score in both phases in formalin-injected mice. While at the dose of 30 mg/kg, (−)-NRG-DM shortened the licking time and pain score of the two behavioral stages, but only had a significant influence in phase I.

Analgesic Effects of (−)-NRG-DM in Formalin-Induced Mouse Inflammatory Pain Model
The formalin test is a classical inflammatory pain model which represents somatic pain. Intraplantar injection of formalin results in a typical two-stage nociceptive behavior, which is characterized by licking and biting of the injected paw. The first phase (0~10 min) mainly reflects nociceptive pain, while the second phase (11~60 min) represents the inflammatory responses [36]. In our study, we tested the effects of (−)-NRG-DM on formalin-induced inflammatory pain in mice. (−)-NRG-DM at 30 mg/kg and 50 mg/kg body weight were individually intraperitoneally administrated 30 min before injection of 1% formalin solution. Consequently, (−)-NRG-DM significantly attenuated painful behaviors in a dose-dependent manner during phase I and phase II (Figure 2). At the dose of 50 mg/kg, (−)-NRG-DM significantly attenuated painful behaviors, including the licking time and overall pain score in both phases in formalin-injected mice. While at the dose of 30 mg/kg, (−)-NRG-DM shortened the licking time and pain score of the two behavioral stages, but only had a significant influence in phase I.
To exclude possible non-specific muscle relaxant or sedative effects, the effects of (−)-NRG-DM on motor performance were evaluated in the rotarod test. (−)-NRG-DM was well tolerated in the rotarod test, with no significant effects on the ability to remain on the rotating rod after intraperitoneal administration of (−)-NRG-DM at 100 mg/kg, 200 mg/kg, or even 400 mg/kg (Table 1). Together, these data showed that (−)-NRG-DM is a well-tolerated natural analgesic compound and can dose-dependently suppress somatic pain in vivo.

Analgesic Effects of (−)-NRG-DM in Mustard Oil-Induced Mouse Colorectal Pain Model
The dose-dependent relief of somatic pain in the formalin model prompted us to ask whether it could attenuate the visceral pain. Thereby, we constructed the mustard oil-induced mouse colorectal pain model as previously described [37]. After intracolonic application of 50 µL 0.75% mustard oil, the mice exhibited pain-related behavior (e.g., licking, stretching, squashing, or retraction of the abdomen) in the next 30 min observation period ( Figure 3A). As those observed in intraplantar formalin-induced pain, (−)-NRG-DM dose dependently relieved intracolonic mustard oil-caused writhing ( Figure 3A). After intraperitoneal application of 30 mg/kg and 50 mg/kg (−)-NRG-DM, the writhing number reduced by approximately 60% and 80%, respectively ( Figure 3B). Our data showed that (−)-NRG-DM could attenuate visceral pain either. To exclude possible non-specific muscle relaxant or sedative effects, the effects of (− NRG-DM on motor performance were evaluated in the rotarod test. (−)-NRG-DM wa well tolerated in the rotarod test, with no significant effects on the ability to remain on th rotating rod after intraperitoneal administration of (−)-NRG-DM at 100 mg/kg, 200 mg/kg or even 400 mg/kg (Table 1). Together, these data showed that (−)-NRG-DM is a wel

Analgesic Effects of (−)-NRG-DM in Mustard Oil-Induced Mouse Colorectal Pain Model
The dose-dependent relief of somatic pain in the formalin model prompted us to as whether it could attenuate the visceral pain. Thereby, we constructed the mustard oi induced mouse colorectal pain model as previously described [37]. After intracolonic ap plication of 50 μL 0.75% mustard oil, the mice exhibited pain-related behavior (e.g., lick ing, stretching, squashing, or retraction of the abdomen) in the next 30 min observatio period ( Figure 3A). As those observed in intraplantar formalin-induced pain, (−)-NRG DM dose dependently relieved intracolonic mustard oil-caused writhing ( Figure 3A). A ter intraperitoneal application of 30 mg/kg and 50 mg/kg (−)-NRG-DM, the writhing num ber reduced by approximately 60% and 80%, respectively ( Figure 3B). Our data showe that (−)-NRG-DM could attenuate visceral pain either.

Inhibitory Effects of (−)-NRG-DM on Action Potential Firing in Mouse DRG Neurons
To investigate the underlying mechanism of (−)-NRG-DM-mediated analgesic activ ity, whole-cell current-clamp technology was applied to examine the effects of (−)-NRG DM on the action potential firing in acutely isolated mouse small-diameter DRG neurons The action potentials were evoked by a current injection of 200 pA for a 500 ms period The threshold was defined by the first amplitude at which an action potential with a mem brane potential larger than 0 mV was produced, and the amplitude was defined as th peak relative to the resting membrane potential [38]. Consistent with its analgesic activ ties in vivo, (−)-NRG-DM dose-dependently inhibited the firing frequency of the DRG neurons ( Figure 4A, B). The amplitudes were reduced from 98.00 ± 6.19 mV to 95.00 ± 5.6 mV, 93.40 ± 8.27 mV, 86.80 ± 9.75 mV, and 74.40 ± 7.08 mV after perfusion of 1 μM, 3 μM 10 μM, and 30 μM (−)-NRG-DM, respectively ( Figure 4C, n = 5). Current threshold, th injection current required to elicit a single all-or-none action potential, was determined b

Inhibitory Effects of (−)-NRG-DM on Action Potential Firing in Mouse DRG Neurons
To investigate the underlying mechanism of (−)-NRG-DM-mediated analgesic activity, whole-cell current-clamp technology was applied to examine the effects of (−)-NRG-DM on the action potential firing in acutely isolated mouse small-diameter DRG neurons. The action potentials were evoked by a current injection of 200 pA for a 500 ms period. The threshold was defined by the first amplitude at which an action potential with a membrane potential larger than 0 mV was produced, and the amplitude was defined as the peak relative to the resting membrane potential [38]. Consistent with its analgesic activities in vivo, (−)-NRG-DM dose-dependently inhibited the firing frequency of the DRG neurons ( Figure 4A,B). The amplitudes were reduced from 98.00 ± 6.19 mV to 95.00 ± 5.65 mV, 93.40 ± 8.27 mV, 86.80 ± 9.75 mV, and 74.40 ± 7.08 mV after perfusion of 1 µM, 3 µM, 10 µM, and 30 µM (−)-NRG-DM, respectively ( Figure 4C, n = 5). Current threshold, the injection current required to elicit a single all-or-none action potential, was determined by applying 500 ms depolarizing currents of increasing magnitude. Surprisingly, the threshold of action potential firing was gradually reduced as the concentration of (−)-NRG-DM increased ( Figure 4D, E, n = 5). The inhibitory effects of (−)-NRG-DM on the firing frequency and the amplitudes of action potentials were partially reversible after washout. These data indicated that (−)-NRG-DM can dampen action potential discharges in nociceptive DRG neurons.
applying 500 ms depolarizing currents of increasing magnitude. Surprisingly, the threshold of action potential firing was gradually reduced as the concentration of (−)-NRG-DM increased ( Figure 4D, E, n = 5). The inhibitory effects of (−)-NRG-DM on the firing frequency and the amplitudes of action potentials were partially reversible after washout. These data indicated that (−)-NRG-DM can dampen action potential discharges in nociceptive DRG neurons.

Inhibitory Effects of (−)-NRG-DM on Neuronal Potassium Currents
(−)-NRG-DM treatment caused a significant decrease in the current threshold for DRG neurons indicating that the compound may inhibit the potassium currents. The currents in mouse DRG neurons can be separated into IA and IK. They could be distinguished by applying voltage steps from a holding potential of −50 mV, at which IA was almost completely inactivated, while IK remained unchanged. Thereby, IA could be separated by subtracting IK from the total potassium currents. According to these electrophysiological characteristics, the effects of (−)-NRG-DM on potassium currents were examined ( Figure  5A). We found that (−)-NRG-DM inhibited potassium currents in a dose-dependent manner, and the IC50 values were 5.10 ± 0.04 μM and 119.90 ± 0.03 μM for IK and IA, respectively ( Figure 5B, n = 5). The data indicated that (−)-NRG-DM is an inhibitor of Kv channels in DRG neurons. Additionally, the effects of (−)-NRG-DM on the kinetics of the Kv channels were further characterized. The Kv currents were elicited by multiple 1500 ms depolarization pulses ranging from −80 mV to +90 mV in 10 mV increments from a holding potential of −80 mV. Congruently, the amplitudes of elicited potassium currents were potently reduced by (−)-NRG-DM at 10 μM, a concentration around IC50 of (−)-NRG-DM on IK currents ( Figure 5C). The activation curves of Kv channels before and after the perfusion of 10 μM (−)-NRG-DM were fitted with the Boltzmann equation, the data showed that (−)-NRG-DM does not affect the activation of the potassium currents ( Figure 5D). The values

Inhibitory Effects of (−)-NRG-DM on Neuronal Potassium Currents
(−)-NRG-DM treatment caused a significant decrease in the current threshold for DRG neurons indicating that the compound may inhibit the potassium currents. The currents in mouse DRG neurons can be separated into I A and I K . They could be distinguished by applying voltage steps from a holding potential of −50 mV, at which I A was almost completely inactivated, while I K remained unchanged. Thereby, I A could be separated by subtracting I K from the total potassium currents. According to these electrophysiological characteristics, the effects of (−)-NRG-DM on potassium currents were examined ( Figure 5A). We found that (−)-NRG-DM inhibited potassium currents in a dose-dependent manner, and the IC 50 values were 5.10 ± 0.04 µM and 119.90 ± 0.03 µM for I K and I A , respectively ( Figure 5B, n = 5). The data indicated that (−)-NRG-DM is an inhibitor of Kv channels in DRG neurons. Additionally, the effects of (−)-NRG-DM on the kinetics of the Kv channels were further characterized. The Kv currents were elicited by multiple 1500 ms depolarization pulses ranging from −80 mV to +90 mV in 10 mV increments from a holding potential of −80 mV. Congruently, the amplitudes of elicited potassium currents were potently reduced by (−)-NRG-DM at 10 µM, a concentration around IC 50 of (−)-NRG-DM on I K currents ( Figure 5C). The activation curves of Kv channels before and after the perfusion of 10 µM (−)-NRG-DM were fitted with the Boltzmann equation, the data showed that (−)-NRG-DM does not affect the activation of the potassium currents ( Figure 5D). The values of V 1/2 in the absence and presence of 10 µM (−)-NRG-DM were −6.43 ± 1.21 mV and −8.08 ± 2.02 mV, respectively ( Figure 5D). These data showed that (−)-NRG-DM is an inhibitor of native potassium currents in DRG neurons. of V1/2 in the absence and presence of 10 μM (−)-NRG-DM were −6.43 ± 1.21 mV and −8.08 ± 2.02 mV, respectively ( Figure 5D). These data showed that (−)-NRG-DM is an inhibitor of native potassium currents in DRG neurons.

Inhibition of (−)-NRG-DM on Neuronal Sodium Currents
A reduction in amplitudes of action potentials manifested that (−)-NRG-DM may inhibit native sodium currents, which are mainly involved in the rising phase of an action potential [39]. To record native sodium currents in acutely isolated DRG neurons, a 50 ms test pulse depolarized to −20 mV from a holding potential of −90 mV was applied. As illustrated in Figure 6A, the amplitudes of peak currents and persistent currents dose-dependently declined as the concentration of (−)-NRG-DM increased. The IC50 value of (−)-NRG-DM on native Nav currents was 34.78 ± 0.14 μM ( Figure 6C, n = 5). To understand how (−)-NRG-DM inhibited Nav channels, the effects of 30 μM compound (−)-NRG-DM on Nav channel kinetics were further characterized. The activation currents were elicited by applying step pluses ranging from −60 mV to +15 mV in 5 mV increments for a 50 ms period from a holding potential of −90 mV ( Figure 6D). The activation curves showed that the half-activation voltage did not change significantly after the perfusion of 30 μM (−)-NRG-DM, and the V1/2 values were −15.34 ± 0.58 mV and −15.23 ± 0.56 mV, respectively ( Figure 6E). The influence of 30 μM (−)-NRG-DM on steady-state inactivation was assessed by 500 ms conditioning pulses ramping from −120 mV to 0 mV in 10 mV increments, followed by a 20 ms test pulse at −20 mV ( Figure 6F). In contrast to the lack of effect on the Nav channel activation, (−)-NRG-DM caused a depolarization shift of the steady-state inactivation. The V1/2 value was shifted from −65.72 ± 1.74 mV to −55.34 ± 2.14 mV by 30 μM (−)-NRG-DM ( Figure 6G). These data showed that (−)-NRG-DM is an inhibitor of native sodium channels in small-diameter DRG neurons.

Inhibition of (−)-NRG-DM on Neuronal Sodium Currents
A reduction in amplitudes of action potentials manifested that (−)-NRG-DM may inhibit native sodium currents, which are mainly involved in the rising phase of an action potential [39]. To record native sodium currents in acutely isolated DRG neurons, a 50 ms test pulse depolarized to −20 mV from a holding potential of −90 mV was applied. As illustrated in Figure 6A, the amplitudes of peak currents and persistent currents dosedependently declined as the concentration of (−)-NRG-DM increased. The IC 50 value of (−)-NRG-DM on native Nav currents was 34.78 ± 0.14 µM ( Figure 6C, n = 5). To understand how (−)-NRG-DM inhibited Nav channels, the effects of 30 µM compound (−)-NRG-DM on Nav channel kinetics were further characterized. The activation currents were elicited by applying step pluses ranging from −60 mV to +15 mV in 5 mV increments for a 50 ms period from a holding potential of −90 mV ( Figure 6D). The activation curves showed that the half-activation voltage did not change significantly after the perfusion of 30 µM (−)-NRG-DM, and the V 1/2 values were −15.34 ± 0.58 mV and −15.23 ± 0.56 mV, respectively ( Figure 6E). The influence of 30 µM (−)-NRG-DM on steady-state inactivation was assessed by 500 ms conditioning pulses ramping from −120 mV to 0 mV in 10 mV increments, followed by a 20 ms test pulse at −20 mV ( Figure 6F). In contrast to the lack of effect on the Nav channel activation, (−)-NRG-DM caused a depolarization shift of the steady-state inactivation. The V 1/2 value was shifted from −65.72 ± 1.74 mV to −55.34 ± 2.14 mV by 30 µM (−)-NRG-DM ( Figure 6G). These data showed that (−)-NRG-DM is an inhibitor of native sodium channels in small-diameter DRG neurons.
The sodium currents in small-diameter DRG neurons could be furtherly subdivided into TTX-sensitive (TTX-S) and TTX-resistant (TTX-R) currents, which contribute to setting the firing threshold and the rising phase of an action potential [40]. To isolate TTX-R currents, whole-cell sodium currents were measured in the presence of 300 nM TTX. Similar to its effect on total sodium currents, (−)-NRG-DM did not affect the activation either ( Figure 6I The sodium currents in small-diameter DRG neurons could be furtherly subdivided into TTX-sensitive (TTX-S) and TTX-resistant (TTX-R) currents, which contribute to setting the firing threshold and the rising phase of an action potential [40]. To isolate TTX-R currents, whole-cell sodium currents were measured in the presence of 300 nM TTX. Similar to its effect on total sodium currents, (−)-NRG-DM did not affect the activation either ( Figure 6I   −80 mV to +110 mV in 10 mV increments for a 1500 ms period from a holding potential of -50 mV ( Figure 7D). Surprisingly, the value of V 1/2 shifted from 22.30 ± 1.12 mV in the control condition to 35.88 ± 1.88 mV in the presence of 20 µM (−)-NRG-DM ( Figure 7E). These data showed that (−)-NRG-DM is an inhibitor of the Kv2.1 channel and dampens channel activation.
whether the natural analgesic compound affects Kv2.1 channels transiently expressed in CHO cells. Congruently, (−)-NRG-DM dose-dependently inhibited Kv2.1 channels with an IC50 value of 21.17 ± 0.11 μM ( Figure 7A, n = 5). The typical Kv2.1 current was elicited by applying a 40 mV depolarization stimulus before and after application of (−)-NRG-DM at 20 μM, a concentration around the IC50 of Kv2.1 channels, as illustrated in Figure 7B. The effects of 20 μM (−)-NRG-DM on the activation of Kv2.1 channels were furtherly evaluated. The activation currents of Kv2.1 were elicited by applying multiple pulses ranging from -80 mV to +110 mV in 10 mV increments for a 1500 ms period from a holding potential of -50 mV ( Figure 7D). Surprisingly, the value of V1/2 shifted from 22.30 ± 1.12 mV in the control condition to 35.88 ± 1.88 mV in the presence of 20 μM (−)-NRG-DM ( Figure  7E). These data showed that (−)-NRG-DM is an inhibitor of the Kv2.1 channel and dampens channel activation.

(−)-NRG-DM Inhibits Nav Channels
The dose-dependent suppression of the sodium currents by (−)-NRG-DM indicated that it should inhibit the Nav channels. As Nav1.7 and Nav1.8 are mainly distributed in the PNS and have been demonstrated to play an important role in pain sensing, we detected the effects of (−)-NRG-DM on Nav1.7 and Nav1.8 channels stably expressed in HEK293 cells. As shown in Figure 8A, we found that 30 μM (−)-NRG-DM, a concentration around the IC50 of native Nav currents in DRG neurons, can potently inhibit Nav1.7 and Nav1.8 currents. The values of IDrug/IControl were 0.61 ± 0.05 and 0.56 ± 0.03, respectively ( Figure 8B). There is no significant difference in the inhibitory efficacy between Nav1.7 and Nav1.8 channels; the data showed that (−)-NRG-DM is a non-selective Nav channel inhibitor ( Figure 8B). To understand how (−)-NRG-DM inhibits Nav1.8 channels, we assessed its impacts on the voltage dependence of steady-state activation and inactivation. The Nav1.8 currents were elicited by applying step pluses ranging from −65 mV to +30 mV for 200 ms in 5 mV increments at a stimulus frequency of 0.5 Hz ( Figure 8C). Activation curves showed that the V1/2 did not change significantly before and after the perfusion of 30 μM (−)-NRG-DM, which was −7.10 ± 1.10 mV and −10.31 ± 1.17 mV, respectively ( Figure 8D). The influence of (−)-NRG-DM on steady-state inactivation was evaluated by

(−)-NRG-DM Inhibits Nav Channels
The dose-dependent suppression of the sodium currents by (−)-NRG-DM indicated that it should inhibit the Nav channels. As Nav1.7 and Nav1.8 are mainly distributed in the PNS and have been demonstrated to play an important role in pain sensing, we detected the effects of (−)-NRG-DM on Nav1.7 and Nav1.8 channels stably expressed in HEK293 cells. As shown in Figure 8A, we found that 30 µM (−)-NRG-DM, a concentration around the IC 50 of native Nav currents in DRG neurons, can potently inhibit Nav1.7 and Nav1.8 currents. The values of I Drug /I Control were 0.61 ± 0.05 and 0.56 ± 0.03, respectively ( Figure 8B). There is no significant difference in the inhibitory efficacy between Nav1.7 and Nav1.8 channels; the data showed that (−)-NRG-DM is a non-selective Nav channel inhibitor ( Figure 8B). To understand how (−)-NRG-DM inhibits Nav1.8 channels, we assessed its impacts on the voltage dependence of steady-state activation and inactivation. The Nav1.8 currents were elicited by applying step pluses ranging from −65 mV to +30 mV for 200 ms in 5 mV increments at a stimulus frequency of 0.5 Hz ( Figure 8C). Activation curves showed that the V 1/2 did not change significantly before and after the perfusion of 30 µM (−)-NRG-DM, which was −7.10 ± 1.10 mV and −10.31 ± 1.17 mV, respectively ( Figure 8D). The influence of (−)-NRG-DM on steady-state inactivation was evaluated by applying a 500 ms conditioning pulse ramping from −110 mV to 10 mV in 10 mV increments, followed by a 20 ms test pulse at −20 mV ( Figure 8E). The values of V 1/2 were −48.58 ± 0.67 mV and −53.37 ± 0.70 mV before and after the application of 30 µM (−)-NRG-DM ( Figure 8F). Similarly, no significant difference was observed. These data showed that (−)-NRG-DM is a nonselective inhibitor of Nav channels. applying a 500 ms conditioning pulse ramping from −110 mV to 10 mV in 10 mV increments, followed by a 20 ms test pulse at −20 mV ( Figure 8E). The values of V1/2 were −48.58 ± 0.67 mV and −53.37 ± 0.70 mV before and after the application of 30 μM (−)-NRG-DM ( Figure 8F). Similarly, no significant difference was observed. These data showed that (−)-NRG-DM is a nonselective inhibitor of Nav channels.

Discussion
In the present study, a naringenin derivative named (−)-NRG-DM was first isolated from N. jatamansi (Figure 1). Intraperitoneal administration of (−)-NRG-DM dose-dependently attenuated pain in a formalin-induced mouse inflammatory pain model and mustard oil-induced mouse colorectal pain model, which, respectively, corresponded to somatic pain and visceral pain (Figures 2 and 3). Notably, (−)-NRG-DM was well tolerated as no significant neurotoxicity was observed at doses of 100 mg/kg, 200 mg/kg, and even 400 mg/kg in the rotarod test (Table 1). All animals displayed no sign of toxicity during the rotarod test (0.5 and 1 h) and 24 h after the drug administration. The data showed that (−)-NRG-DM is an analgesic compound with a wide margin of safety. Combining the data obtained from the acutely isolated mouse small-diameter DRG neurons, heterologous expression system, and pharmacokinetic study, we furtherly elucidated that inhibition of neuronal channels mediating IK and INa currents is implicated in the (−)-NRG-DM-produced analgesic activity.

Discussion
In the present study, a naringenin derivative named (−)-NRG-DM was first isolated from N. jatamansi (Figure 1). Intraperitoneal administration of (−)-NRG-DM dosedependently attenuated pain in a formalin-induced mouse inflammatory pain model and mustard oil-induced mouse colorectal pain model, which, respectively, corresponded to somatic pain and visceral pain (Figures 2 and 3). Notably, (−)-NRG-DM was well tolerated as no significant neurotoxicity was observed at doses of 100 mg/kg, 200 mg/kg, and even 400 mg/kg in the rotarod test (Table 1). All animals displayed no sign of toxicity during the rotarod test (0.5 and 1 h) and 24 h after the drug administration. The data showed that (−)-NRG-DM is an analgesic compound with a wide margin of safety. Combining the data obtained from the acutely isolated mouse small-diameter DRG neurons, heterologous expression system, and pharmacokinetic study, we furtherly elucidated that inhibition of neuronal channels mediating I K and I Na currents is implicated in the (−)-NRG-DMproduced analgesic activity.
The current study showed that the analgesic (−)-NRG-DM directly dampens neuron excitability in acutely isolated mouse small-diameter DRG neurons with a reduced threshold and amplitude of action potential firing ( Figure 4A). Intriguingly, the suppression of firing frequency is tightly accompanied by a depolarized firing threshold ( Figure 4A,D). The significant reduction in the number of action potentials started after 10 µM (−)-NRG-DM was applied, at which the significant difference in the firing threshold occurred ( Figure 4D,E). The data suggested that (−)-NRG-DM may cause suppression through a mechanism similar to neuronal desensitization. The feature is involved in capsaicin, a TRPV1 channel agonist, and an 8% capsaicin patch (Qutenza) has been approved for the treatment of chronic pain in clinics [42]. Analgesic activities produced by a direct desensitization of nociceptive neurons might reduce side effects including gastrointestinal erosions, renal and hepatic insufficiency which are commonly associated with cyclooxygenase inhibitors [43]. As 10 µM (−)-NRG-DM did not affect TRPA1 currents, the neuronal desensitization could not be ascribed to its modulation on the channels ( Figure S1). The suppression of neuron activity produced by (−)-NRG-DM is in agreement with the study of (+)-naringenin 4 ,7-dimethyl ether ((+)-NRG-DM), a naturally occurring naringenin derivative that has been isolated many times from plants, that showed analgesic activity in vivo and that did not influence the production and release of pro-inflammatory factors compared to other naringenin derivatives [44]. Additionally, radioligand binding assay demonstrated that (+)-NRG-DM does not show affinity to endocannabinoid or opioid receptors [22]. Together, our data showed that (−)-NRG-DM causes a direct inhibition of neuron excitability through a mechanism similar to excitatory desensitization in nociceptive neurons.
Voltage-gated potassium currents play a fundamental role in the modulation of resting membrane potential [45]. In the small-diameter DRG neurons, the currents can be divided into two separate components: I A and I K [46]. Due to the low current density and distribution ratio, the I A currents did not seem to play a key role in the excitability of nociceptive neurons. The I K currents contribute to the setting of resting membrane potential and appear to be the main contributor to after-hyperpolarization [20,47]. Consistent with the depolarized threshold of action potential firing, (−)-NRG-DM potently inhibited the I K but not I A currents with IC 50 values of 5.10 ± 0.04 µM and 119.90 ± 0.03 µM, respectively ( Figure 5A,B). Notably, the pharmacokinetic study showed that the plasma concentration of (−)-NRG-DM after 0.5~1.5 h intraperitoneal injection of the compound at 50 mg/kg is around 7 µM, which is higher than the IC 50 value of (−)-NRG-DM on I K currents ( Table 2). The reduction in the I K currents in DRG neurons at this concentration was around 58% ( Figure 5B). Kv2.1 homotetramers or containing heterotetramers represent the major component of I K currents in neurons [18,40]. Consistently, the IC 50 value of (−)-NRG-DM on Kv2.1 channels heterologously expressed in CHO cells was 21.17 ± 0.11 µM (Figure 7A,B). The difference between the IC 50 values obtained from DRG neurons and CHO cells may emerge from species differences in primary pharmacology. One might argue that a reduction in the I K currents has been found in multiple animal models for studying pain, and lesser I K currents are involved in human labor pain [48,49]. The feature might be similar to that found in Nav1.9 channels, in which the gain-of-function mutations caused a loss of pain reception in humans and exhibited a reduced neuronal excitability during the long stimulus due to neuronal desensitization [50]. Thus, our data showed that the inhibition of I K mainly contributes to (−)-NRG-DM-produced pain relief and Kv2.1-containing channels involved in the inhibitory activity. The reduction in the amplitudes of the action potentials after exposure to (−)-NRG-DM was in accordance with its suppression of sodium currents ( Figure 4C). (−)-NRG-DM inhibited Nav currents in DRG neurons in a dose-dependent manner with an IC 50 value of 34.78 ± 0.14 µM ( Figure 6C). Nevertheless, a reduction in the I Na of the DRG neurons was around 28% at the plasma concentration obtained from the pharmacokinetic study ( Figure 6C). Nav1.7 and Nav1.8 are analgesic-related Nav subtypes in the PNS, which separately correspond to the TTX-S and TTX-R currents in small-diameter DRG neurons [46]. The inhibitory ratios of (−)-NRG-DM (30 µM) on Nav1.7 and Nav1.8 were 39 ± 5% and 44 ± 3% respectively, which were very close to those on native sodium currents ( Figure 8A,B). As an enantiomer, NRG-DM extracted from peanut stem and leaf promoted sleep by dampening neuronal excitability. In cortical neurons, the inhibitory activity of NRG-DM on Nav channels appeared to be more potent than that on Kv channels [51]. However, in the current study, due to the higher IC 50 value, our data showed that inhibition of Nav channels might contribute to (−)-NRG-DM-produced analgesic activity but the effect is secondary to that of Kv channels.
In conclusion, our study showed that (−)-NRG-DM is a promising analgesic drug candidate potently attenuating somatic and visceral pain in vivo. The analgesic activity could be ascribed to its direct suppression of nociceptive neuron excitability through a desensitization mechanism. Due to the plasma concentration of (−)-NRG-DM being higher than the effective concentrations for the I K and I Na , our study suggested that the inhibition of neuronal channels mediating these currents contributes to the analgesic activity. (−)-NRG-DM may act as a new structural framework for the subsequent development of analgesic drugs.

Animals
All mice were obtained from the Beijing Vitalriver Laboratory Animal Technology Co., Ltd. (Beijing, China). Mice were housed and assayed under controlled temperature conditions (22 ± 2 • C) and a 12 h light/dark cycle with free access to food and water. All animal procedures were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were strictly followed and approved by the guidelines of the IACUC (Institutional Animal Care and Use Committees). The IACUC checked all protocols and approved this study. The animal experiments were conducted in a blinded manner, i.e., drug administration and behavioral tests were finished by different investigators.

Formalin-Induced Inflammatory Model
Adult male ICR mice weighing 20 ± 2 g were randomly divided into 3 groups (n = 10 for each group) and acclimatized in a transparent observation chamber for at least 30 min before the experiment. Mice were intraperitoneally administrated with vehicle or (−)-NRG-DM (30 or 50 mg/kg) 30 min prior to formalin injection. Then, 1% formalin solution (20 µL per site) was subcutaneously injected into the plantar of the left hind paw to induce acute inflammatory pain. Immediately, mice were put back into the observation chambers. Nociception caused by formalin was assessed by scoring painful behaviors and licking time over a period of 60 min. In the present study, the score represented the sum of weighted formalin-induced pain-related behavior: 1 = flinching, 2 = shaking, and 3 = licking or biting of the injected paw. Phases were defined as follows: the peak time of the early nociceptive response phase (phase I) was 0~10 min, and the late phase (phase II) was 11~60 min after formalin injection.

Mustard Oil-Induced Mouse Colorectal Pain Model
Male C57BL/6 mice, weighing 23~25 g, were randomly assigned into 3 groups (n = 10 for each group) and placed in a transparent observation chamber for at least 20 min prior to the experiment. Mice were intraperitoneally administrated with vehicle or (−)-NRG-DM (30 or 50 mg/kg) 30 min before the injection of diluted mustard. Subsequently, 50 µL of diluted mustard oil solution (0.75% in 70% ethanol) was intracolonically administrated and the number of pain responses was counted for 30 min. In the present study, postures defined as pain-related behaviors were in agreement with previous descriptions: (1) licking of the abdomen, (2) stretching the abdomen, (3) squashing of lower abdomen against the floor, (4) abdominal wall retractions.

Rotarod Test
To determine the neurotoxicity effects of (−)-NRG-DM, the standardized rotarod test was conducted in male ICR mice weighted 20 ± 2 g. The mice were divided at random into 4 groups (n = 10 for each group). The mice were placed on a rotarod appliance (YLS-4C, Bio-will, Shanghai, China) with a rod of 3 cm diameter, rotating at a constant speed of 6 rpm. The day before the compound test, all mice were pre-trained and only the animals able to remain on the rod for at least 1 min every time in three consecutive trials (3 min) were retained. During the test, the mice were measured in the rotarod test 0.5 h and 1 h after intraperitoneal administration of (−)-NRG-DM. The animals unable to remain on the rod for 3 consecutive periods were considered motor coordination impaired.

Pharmacokinetic Study
The pharmacokinetic study was performed in male ICR mice weighted 20 ± 2 g. The mice were fasted for 12 h before intraperitoneal administration of (−)-NRG-DM at 50 mg/kg. Blood samples (0.5 mL) were collected at 0.5, 1, and 1.5 h from the abdominal aorta into the heparinized tubes after drug administration. The plasma was separated by centrifugation (11,000 rpm for 5 min) and then stored at −80 • C until analyzed. Plasma samples (20 µL) were treated with the addition of the internal standard solution (20 µL) and acetonitrile (300 µL), then vortexed for 15 min (1000 rpm, RT), and centrifuged of 15 min (3700 rpm, 4 • C). The supernatants were collected for analysis using LC−MS/MS. The data were processed using Analyst software version 1.6.3 (Sciex, ON, Canada).

Preparation of Dorsal Root Ganglion Neurons
Dorsal root ganglia (DRG) were dissected from male C57BL/6 mice aged 4~6 weeks. The ganglia were first cut into small pieces and then digested at 37 • C for 20 min in DMEM containing 1 mg/mL collagenase type I and 0.25 mg/mL trypsin (Sigma-Aldrich). Subsequently, the digested small fragments were terminated and resuspended with DMEM/F12 growth medium (Gibco) supplemented with 10% fetal bovine serum (FBS) (Gibco). Finally, the dissociated DRG neurons were seeded onto 24-well plates with poly-L-lysine-coated coverslips and placed in a 37 • C, 5% CO 2 incubator for at least 1 h before electrophysiological experiments.

Cell Culture and Transfection
Human embryonic kidney 293 (HEK293) cells stably expressing human Nav1.7 and Nav1.8 channels were grown in a high-glucose DMEM medium (Gibco) containing 10% FBS (Gibco). The media were respectively supplemented with 50 mg/mL and 100 mg/mL hygromycin B (Invitrogen, Carlsbad, CA, USA). The cDNA encoding human Kv2.1 channels was synthesized by Sangon Biotech Co., Ltd. (Shanghai, China) based on the GenBank (Kv2.1 Gene ID: 25736) and was subcloned into the pcDNA3.1(+) vector. Chinese hamster ovary (CHO) cells were cultured in DMEM/F12 medium (Gibco) supplemented with 10% FBS. To transiently express the Kv2.1 channels for electrophysiological studies, the constructs encoding the EGFP and the Kv2.1 were co-transfected into the CHO cells with Lipofectamine 2000 reagent (Invitrogen) referring to the manufacturer's instructions. The transfected cells were seeded onto poly-L-lysine-coated glass coverslips before they were used for the electrophysiological study. All cells were grown under standard tissue culture conditions (5% CO 2 , 37 • C).

Statistics
Patch-clamp data were processed using Clampfit 10.6 (Molecular Device, Sunnyvale, CA, USA) and then analyzed in GraphPad Prism 5 (GraphPad Software, San Diego, CA, USA). Voltage-dependent activation curves were fitted to the Boltzmann equation: G = G min + (G max − G min )/[1 + exp (V − V 1/2 )/S], where G max is the maximum conductance, G min is the minimum conductance, V 1/2 is the voltage to reach 50% of the maximum conductance, and S is the slope factor. Steady-state inactivation curves were constructed by plotting the normalized peak currents during the test pulses as a function of the prepulse potentials. The data were fitted to the Boltzmann equation: I/I max = 1/{1+ exp [(V − V 1/2 )/Ki]}, where I is the amplitude of peak currents at each voltage, I max is the maximal value of peak currents, V and V 1/2 are the prepulse potential and the half-maximal potential for inactivation, respectively, and Ki is the inactivation slope factor. Dose-response curves were fitted with the Hill equation: Y = Bottom + (Top − Bottom)/{1+10[(LogIC 50 − X) × k]}, where Bottom and Top are the minimum and maximum inhibition, respectively; X is the log of the concentration; Y is the value of I Drug /I Control ; IC 50 is the drug concentration producing a half-maximum response, k is the Hill Slope. The data are shown as the mean ± SEM, and the significance was estimated using paired two-tailed Student's t-test unless otherwise stated. Statistical significance: * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.