N-methyl- D -aspartate receptors and glycinergic transmission, respectively, mediate muscle relaxation and immobility of pentobarbital in mice

gabaculine and sarcosine, respectively, the neuronal nicotinic acetylcholine receptor antagonist mecamylamine, or the N-methyl- D -aspartate receptor channel blocker MK-801 could enhance pentobarbital-induced components of anesthesia. Muscle relaxation, unconsciousness, and immobility were evaluated by grip strength, the righting reflex, and loss of movement in response to nociceptive tail clamping, respectively, in mice. Pentobarbital reduced grip strength, impaired the righting reflex, and induced immobility in a dose-dependent manner. The change in each behavior induced by pentobarbital was roughly consistent with that in electroencephalographic power. A low dose of gabaculine, which significantly increased endogenous GABA levels in the central nervous system but had no effect on behaviors alone, potentiated muscle relaxation, unconsciousness, and immobility induced by low pentobarbital doses. A low dose of MK-801 augmented only the masked muscle-relaxing effects of pentobarbital among these components. Sarcosine enhanced only pentobarbital-induced immobility. Conversely, mecamylamine had no effect on any behavior. These findings suggest that each component of anesthesia induced by pentobarbital is mediated through GABAergic neurons and that pentobarbital-induced muscle relaxation and immobility may partially be associated with N-methyl- D -aspartate receptor antagonism and glycinergic neuron activation, respectively.


Introduction
General anesthesia can be described as a complex state composed of sedation, analgesia, muscle relaxation, unconsciousness, loss of movement in response to noxious stimuli (immobility), and amnesia [1,2]. Many anesthetics, including inhaled and intravenous (i.v.) agents, such as sevoflurane, barbiturates, and propofol, are believed to exert their effects through γ-aminobutyric acid type A (GABA A ) receptors in the central nervous system (CNS). In addition to GABA A receptors, glycine and neuronal nicotinic acetylcholine (nACh) receptors are members of the Cys-loop ligand-gated ion channel superfamily [3]. The barbiturate pentobarbital prolongs spinal glycinergic currents [4]. Sarcosine is a type I glycine transporter inhibitor that increases endogenous glycine levels in the synaptic cleft. Glycine receptors might play a role in mediating the inhibition of responses to noxious stimuli, such as analgesia and immobility among the components of anesthesia, by halothane, another inhaled anesthetic [5]. Subanesthetic concentrations of inhaled anesthetics block some classes of neuronal nACh receptors, which appear to mediate other components of anesthesia, such as amnesia and analgesia [6][7][8]. Mecamylamine is a neuronal nACh receptor antagonist that blocks all known nACh receptor subtypes. Ketamine and the gaseous anesthetics nitrous oxide and xenon block a different type of ligand-gated ion channel, the N-methyl-D-aspartate (NMDA) receptor. Unlike these anesthetics, MK-801 is a selective NMDA receptor channel blocker. In NMDA receptor GluRε1 subunit knockout mice, the general anesthetic potency of pentobarbital is reduced similarly to that of ketamine compared with that of pentobarbital in wildtype animals [9]. NMDA receptors are glutamate-gated cation channels that are involved in long-term potentiation, which induces synaptic plasticity and which is a candidate mechanism of the antagonists for amnesia and unconsciousness [2]. Thus, each component of anesthesia should be explored separately [10].
Barbiturates can depress CNS activity, resulting in effects ranging from sedation to general anesthesia. The barbiturates used for clinical anesthesia include thiopental, which is metabolized to pentobarbital, an active metabolite with a longer half-life. After the administration of barbiturates such as thiopental, pain perception and reaction are relatively unimpaired until the moment of unconsciousness. At higher doses, barbiturates produce unconsciousness and suppress responses to painful stimuli [11,12]. Pentobarbital at high concentrations directly activates Clchannel opening via GABA A receptors, even in the absence of GABA [13], suggesting that pentobarbital-induced anesthesia is mediated by enhancement of GABAergic neurons [11,12,14]. However, it is unclear whether all components of anesthesia induced by pentobarbital are mediated only through the enhancement of GABAergic neurons because GABA neuron stimulation by the GABA transaminase inhibitor gabaculine induced loss of the righting reflex (LORR), which is one marker of unconsciousness, but not immobility [15,16]. Thus, in this study, we examined whether gabaculine, sarcosine, mecamylamine, or MK-801 could enhance components of anesthesia, such as muscle relaxation, unconsciousness, and immobility, induced by pentobarbital in mice.

Animals
Adult male ddY mice (Kyudo Co. Ltd., Kumamoto, Japan) weighing 30-44 g were used (n = 795). Mice were housed (five per cage) in an airconditioned room maintained at 25±1 • C with 50% relative humidity on a 12-h light/dark cycle (lights on at 8:00). Food and water were available ad libitum. Animals were used only once in all experiments. All behavioral experiments were performed during the light cycle. This study was approved by the Hiroshima University Animal Research Committee (approval numbers: A08-89, A11-94-2, A20-58) and conducted in accordance with the Guide for the Hiroshima University Animal Experimentation Regulation.

Drugs
Pentobarbital sodium salt was purchased from Tokyo Chemical Industry Co. Ltd. (Tokyo, Japan). 3-Amino-2,3-dihydrobenzoic acid (gabaculine) hydrochloride, sarcosine, mecamylamine hydrochloride, and (+)-MK-801 hydrogen maleate were obtained from Sigma-Aldrich Co. LLC (St. Louis, MO, USA). All drugs were dissolved in 0.9% saline solution. Each drug was freshly prepared on the day of the experiment. Pentobarbital, gabaculine, mecamylamine, and MK-801 were administered as intraperitoneal (i.p.) injections at a volume of 0.1 mL per 10 g of body weight. Sarcosine was injected i.v. at a volume of 5 mL/kg.

Behavioral studies
Components of general anesthesia were assessed according to three behaviors: grip strength as an index of muscle relaxation, LORR as a marker of unconsciousness, and loss of movement in response to painful stimulation as a measure of immobility.

Grip strength test
Prior to pentobarbital, specific ligand, or vehicle administration, grip strength was measured once at baseline. After pentobarbital injection, grip strength was measured and recorded every 5 min for 60 min postinjection using an apparatus (MK-380 M, Muromachi Kikai, Co., Ltd., Tokyo, Japan) ( Supplementary Fig. 1A.). The mouse was placed on a wire grid (12×18 cm) and pulled backward slowly. The point of force at which the mouse released the grid was determined as the grip strength [17].

Righting reflex test
The righting reflex is a response by which a mouse quickly returns to an upright position even when forced to stay on its side or back. Mice were investigated individually in a circular glass beaker (13.5 cm, diameter × 22.5 cm, height). To assess the righting reflex, the beaker was tilted by hand to an angle of approximately 45 • from the horizontal plane. This was repeated three times for each recording time after drug injection. The righting reflex scores were assessed and recorded every 2 min for 4 h after drug injection by a blinded observer according to the rating scale, as reported previously [18] (Supplementary Fig. 1A.). In brief, a score of 0 indicated a normal righting reflex; a score of +1 indicated that the mouse righted itself within 2 s in all three trials (slightly impaired righting reflex); a score of +2 indicated that the time to righting in the best response among three trials was >2 s but <10 s (moderately or severely impaired righting reflex); and a score of +3 corresponded to absence of the righting reflex (no righting within 10 s in all three trials). Onset of LORR was considered to occur when mice received a score of +3. The time between LORR (a score of +3) and when mice regained righting ability (a score of +2) was considered the duration of LORR. The total score represents the total of the scores recorded every 2 min after drug administration.

Immobility test
To determine immobility, at the time when the drug induced its peak effect on the righting reflex, a tail clamp was applied under LORR with artery forceps approximately 1.0 cm from the base of the tail for 30 s or until the animal moved, as described previously [18]. Pentobarbital produced its peak effect on the righting reflex within 10 min after i.p. administration ( Fig. 1 and Table 1). Therefore, the tail clamp was applied 10 min after pentobarbital administration ( Supplementary Fig.  1A). Purposeful movements of the head and/or legs after tail-clamp stimulation were considered responses. Purposeless movements, such as coughing or hyperventilation, were excluded. The anesthetized animals were kept warm with a heating pad (CMA150, CMA Microdialysis AB, Kista, Sweden).
The number of mice that lost the response among all animals that received a specific treatment was used to calculate the percent effect because the manner of the response was all-or-none. The dose-response was fit to a sigmoidal curve to obtain the 50% effective dose (ED 50 ) using Prism version 7.03 (GraphPad Software, San Diego, CA, USA).

Electroencephalogram (EEG) and electromyogram (EMG) studies
Mice were deeply anesthetized by i.p. administration of chloral hydrate (400 mg/kg) and subcutaneous (s.c.) injection of 0.1 mL of 3% mepivacaine, a local anesthetic, into the head and placed in a stereotaxic frame (KOPF model 900; David Kopf Instruments, Tujunga, CA, USA). Two stainless-steel electrodes were implanted in the skull over the primary visual cortex (AP: − 3.5 mm relative to the bregma, ML: 1.7 mm relative to the midline) according to the atlas of Franklin and Paxinos [19] as positive and reference electrodes for surface EEG recording. A third electrode was implanted on the parietal bone to serve as the relative electrode. An EMG electrode was inserted into the trapezius muscle. After surgery, mice were housed individually and allowed to recover for 1 week. The day before EEG and EMG recordings, animals were placed overnight in an acrylic chamber (30×30×35 cm) shielded with metal mesh to block electromagnetic interference from the surrounding environment. On the following day, the recording began with 30 min of basal measurements followed by 80 min of post-drug injection measurements ( Supplementary Fig. 1B). The measurements were performed during the light period (11:00 a.m. to 4:00p.m.). Cortical EEG and EMG signals were amplified and filtered (EEG, 1-40 Hz; EMG, 20-200 Hz), digitized at a sampling rate of 128 Hz, and recorded using SleepSign software version 3.0 (Kissei Comtec, Nagano, Japan) [20]. Only artifact-free recordings were used in the analysis. EEG and EMG data were automatically scored offline in 10-s epochs as either wake or sleep states using SleepSign software according to standard criteria.
According to Hayashiuchi et al., "Sleep states were characterized by a remarkable reduction of EMG power and high-amplitude, low-frequency EEG. Wakefulness was characterized by a high EMG signal with lowamplitude and mixed-frequency EEG" [21]. However, the EEG patterns and other features of general anesthesia generally differ from those of natural sleep [22]; therefore, EEG power spectra (0.65-40 Hz; μV 2 ) were analyzed by fast Fourier transform using SleepSign software. To compare the time-course differences in EEG band power versus baseline, we calculated the EEG power of each 10-min period in the delta (δ, 0.65-4 Hz), theta (θ, 4-10 Hz), alpha (α, 10-12 Hz), beta (β, 12-20 Hz), and gamma (γ, 20-40 Hz) bands as a percentage versus baseline. The total power in each band for 30 min before pentobarbital administration at a dose of 20 or 40 mg/kg was divided by three, and the averaged 10min power values were considered the baseline. The values of each 10min cumulative power measurement after pentobarbital administration were normalized as percent (%) changes versus the baseline values. The EMG power spectrum  Hz; μV 2 ) was analyzed in the same manner as EEG power spectra.  (10,15,20, and 25 mg/kg, n = 6-7, A; 20, 30, 40, and 50 mg/kg, n = 6, B) and vehicle were administered intraperitoneally (i.p.). Prior to pentobarbital or vehicle administration, grip strength was measured once as baseline. After i.p. pentobarbital or vehicle injection, grip strength was measured and recorded every 5 min for 60 min postinjection (A). The righting reflex was assessed every 2 min for 4 h after i.p. injection of pentobarbital using righting reflex scores. A score of 0 indicated a normal righting reflex; a score of + 1 indicated that the mouse righted itself within 2 s in all three trials (slightly impaired righting reflex); a score of + 2 indicated that the time to righting in the best response among three trials was > 2 s but < 10 s (moderately or severely impaired righting reflex); and a score of + 3 corresponded to absence of the righting reflex (no righting within 10 s in all three trials) (B). Each point represents the mean ± SEM (A) or mean (B). *P < 0.05, **P < 0.01 compared with baseline (time 0), two-way analysis of variance (ANOVA) followed by Dunnett's test (A). Two-way ANOVA (subject × time) revealed that the time factor concerning the change in the grip strength of mice treated with pentobarbital was significant (F(12, 72) = 2.25, P = 0.0176, n = 7 (10 mg/kg); F(12, 72) = 3.94, P = 0.0001, n = 7 (15 mg/kg); F(12, 72) = 9.77, P < 0.0001, n = 7 (20 mg/kg); and F (12, 60) = 32.38, P < 0.0001, n = 6 (25 mg/ kg)), but that with vehicle was not significant (F(12, 72) = 1.62, P > 0.05, n = 7).

Microdialysis study
To measure brain GABA concentrations after i.p. administration of gabaculine, a microdialysis study was conducted, as previously reported [23]. Briefly, a microdialysis guide cannula was implanted into the hippocampus (AP: 2.6 mm, L: 2.0 mm, V: 1.2 mm from the bregma and top of the skull) under deep anesthesia [19]. The animals were allowed to recover from the surgery for at least 7 days before perfusion experiments.
On the day of the experiment, the probes were inserted into the guide cannula and perfused with artificial cerebrospinal fluid at a constant flux rate of 0.5 µL/min with a microperfusion pump. After 3 h of perfusion, the first three samples were collected at 60-min intervals to determine baseline GABA levels in the hippocampus. The samples were collected up to 24 h after gabaculine administration ( Supplementary Fig. 1C). A 10-(n = 5) or 50-mg/kg (n = 4) dose of gabaculine was administered via i.p. injection after stabilization. All microdialysis samples were stored at − 20 • C.
The GABA concentration in each sample was measured by high performance liquid chromatography using a fluorescence detector and a computing integrator/printer after pre-column fluorescence derivatization with o-phthalaldehyde. A sufficient volume (30 μL) was obtained from 24 of the 27 remaining samples. The basal GABA level in each animal was calculated as the mean of the last three consecutive samples immediately before gabaculine administration. The subsequent results were expressed as a percentage of the basal GABA level in each individual animal.

Statistical analysis
The data are presented as the mean ± SEM. Statistical analyses were performed using two-way analysis of variance (ANOVA) followed by Dunnett's test, a paired t-test, or Student's t-test for parametric data, and Kruskal-Wallis one-way ANOVA followed by Dunn's multiple comparisons test or the Mann-Whitney U test was applied for non-parametric data. A probability value (P) of <0.05 was considered significant. All statistical analyses were performed with Prism version 7.03 (GraphPad Software).

Effects of pentobarbital on grip strength, the righting reflex, and immobility
I.p. injection of pentobarbital at doses of 15, 20, and 25 mg/kg significantly reduced grip strength in a dose-dependent manner (P < 0.05), but this effect was not observed at 10 mg/kg (P > 0.05, two-way ANOVA (subject × time) followed by a post-hoc Dunnett's test). The efficacy peaked at 5-10 min after injection and then returned to baseline within 30 min (Fig. 1A).

Effects of pentobarbital on EEG and EMG recordings
As shown in Fig. 2Ab, pentobarbital (20 mg/kg) induced a significant increase in θ power from 10 to 20 min post-injection compared with the baseline value (P < 0.01, two-way ANOVA (subject × time) followed by a post-hoc Dunnett's test). Furthermore, this dose significantly increased α, β, and γ power for 40 min after injection (P < 0.01; Fig. 2Ac, 2Ad, and 2Ae). In contrast, 20 mg/kg of pentobarbital did not affect δ and EMG power (P > 0.05; Fig. 2Aa and Af).
At 40 mg/kg, pentobarbital significantly increased δ power from 10 to 20 min post-injection (P < 0.05) and reduced EMG power over 20 min after administration (P < 0.01), at which time righting reflex scores peaked (Fig. 1B), in contrast to the effect of 20 mg/kg ( Fig. 2Ba and Bf).
At this dose, time-course changes in θ, α, β, and γ power exhibited a similar pattern. These powers gradually and significantly increased, peaking at 50 or 60 min post-injection, excluding the period of 10-20 min after injection when the depth of anesthesia was maximal (P < 0.05; Fig. 2Bb, Bc, Bd and Be).

Effects of gabaculine on hippocampal GABA levels
The time-course effects of gabaculine on extracellular GABA levels are shown in Supplementary Fig. 2. Gabaculine dose-dependently increased extracellular GABA levels, which peaked at approximately 12 h post-injection. Meanwhile, 50 mg/kg of gabaculine induced a steeper increase in GABA levels than 10 mg/kg, reaching a plateau after the peak, and this dose significantly increased GABA levels at 5-24 h after injection versus baseline (P < 0.05, two-way ANOVA (subject × time) followed by a post-hoc Dunnett's test). At the peak, GABA levels were >10-and 30-fold higher than the baseline levels following treatment with 10 and 50 mg/kg of gabaculine, respectively. Gabaculine (50 mg/ kg) was reported to increase GABA concentrations of the rat medulla oblongata >12-fold at 20 h after injection, which plays an important role in consciousness, compared with the control (1.7 μmol/g weight) [24].
Although gabaculine is an irreversible GABA transaminase inhibitor, a low dose induced a significant increase in GABA levels at 9-18 h after Table 1 Effect of pentobarbital on percentage of loss of the righting reflex (LORR), onset of LORR, duration of LORR, and total righting reflex score in mice. The righting reflex was assessed every 2 min for 4 h after the intraperitoneal injection of pentobarbital using righting reflex scores. A score of 0 indicated a normal righting reflex; a score of + 1 indicated that the mouse righted itself within 2 s in all three trials (slightly impaired righting reflex); a score of + 2 indicated that the time to righting in the best response among 3 trials was > 2 s but < 10 s (moderately or severely impaired righting reflex); and a score of + 3 corresponded to absence of the righting reflex (no righting within 10 s in all three trials). The onset of LORR is the beginning time of a score of + 3. The time between LORR (a score of + 3) and the time mice regained that ability (a score of + 2) is the duration of LORR. The total score is the total of the scores recorded every 2 min after drug administration. Data are shown as the mean ± SEM. Parentheses show number of animals used in each group. ND: not detected.
injection (P < 0.05). However, its levels gradually decreased thereafter, and the significance disappeared at 19 h after injection (P > 0.05), suggesting that its effects are reversible. Therefore, in subsequent studies, the behavioral effects of 50 mg/kg gabaculine were examined starting 17 h after injection, when its effects on GABA levels had plateaued ( Supplementary Fig. 1D). Meanwhile, i.v. administration of sarcosine was reported to produce analgesia in mouse neuropathic pain models, and its analgesic effect peaked at 4 h post-injection and persisted until 72 h [25]. In the present study, we tested behaviors from 5 h after i.v. injection of sarcosine. I.p. administration of MK-801 and mecamylamine induced behavioral effects immediately after injection, and the effects peaked within 30 min [26][27][28]. Thus, we began to observe MK-801-and mecamylamineinduced behavior 30 min after i.p. injection (Supplementary Fig. 1D).

Effects of gabaculine, sarcosine, MK-801, and mecamylamine combined with pentobarbital on grip strength
Gabaculine (50 mg/kg) combined with pentobarbital (10 mg/kg) significantly reduced grip strength throughout the recording after pentobarbital injection (P < 0.0001), whereas gabaculine or pentobarbital alone had no such effects (P > 0.05), as analyzed by two-way ANOVA (subject × time) followed by a post-hoc Dunnett's test (Fig. 3A). Sarcosine plus pentobarbital also had no effect on grip strength according to the post-hoc test (P > 0.05) (Fig. 3B). A low dose of MK-801 (0.03 mg/kg), as well as the vehicle control and pentobarbital (10 mg/kg) alone, did not induce muscle relaxation (P > 0.05). However, pentobarbital significantly enhanced the masked muscle-relaxing effects of MK-801 at a low dose (0.03 mg/kg) at 5, 10, and 15 min after pentobarbital injection (P < 0.0001). At a higher dose (0.2 mg/kg), MK-801 alone or combined with pentobarbital (10 mg/kg) significantly suppressed grip strength throughout the recording period, and pentobarbital had an additive effect on the inhibitory effects of MK-801 on grip strength (P < 0.05; Fig. 3C). Treatment with vehicle and mecamylamine had no effect on grip strength (P > 0.05). Mecamylamine combined with pentobarbital significantly induced muscle relaxation within 5 min post-injection (P < 0.01), and pentobarbital alone also weakened grip strength only at 5 min (P < 0.05; Fig. 3D).

Effects of gabaculine, sarcosine, MK-801, and mecamylamine combined with pentobarbital on the righting reflex
Gabaculine (50 mg/kg) significantly shortened the onset of LORR Gabaculine (50 mg/kg, n = 7) or vehicle (n = 7) was administered intraperitoneally (i.p.) 17 h before the i.p. injection of pentobarbital (10 mg/kg, n = 7) or vehicle (n = 5). Sarcosine (10 mg/kg, n = 7) or vehicle (n = 7) was injected intravenously (i.v.) 5 h before the i.p. injection of pentobarbital (10 mg/kg, n = 7) or vehicle (n = 5). MK-801 (0.03 or 0.2 mg/kg, n = 7) or vehicle (n = 7) was administered i.p. 30 min before the i.p. injection of pentobarbital (10 mg/kg, n = 7) or vehicle (n = 5). Mecamylamine (5 mg/kg, n = 7) or vehicle (n = 7) were injected i.p. 30 min before the i.p. injection of pentobarbital (10 mg/kg, n = 7) or vehicle (n = 5). Prior to drug or vehicle administration, grip strength was measured once at baseline. The effects of gabaculine on the grip strength were examined 17 h post-injection, when its efficacy on GABA levels was at a plateau. Sarcosine was tested 5 h post-injection. MK-801 and mecamylamine were examined 30 min post-injection. Grip strength was measured and recorded 5 min before, at, and every 5 min for 60 min after injection of pentobarbital or vehicle. *P < 0.05, **P < 0.01 compared with baseline, two-way analysis of variance (ANOVA) followed by Dunnett's test. For gabaculine, two-way ANOVA (subject × time) revealed that the time factor concerning the change in the grip strength of mice treated with gabaculine + pentobarbital was significant (F(14, 84) = 26.42, P < 0.0001), but that with vehicle + vehicle, vehicle + pentobarbital, and gabaculine + vehicle was not significant (F(14,56)  and enhanced the duration of LORR and the total righting reflex score induced by pentobarbital at 30 mg/kg, which produced LORR in approximately half of the animals (P < 0.05, Mann-Whitney U test), whereas sarcosine (10 mg/kg) had no such effects (P > 0.05). Only a high dose of MK-801 (0.2 mg/kg) significantly increased the duration of LORR and the total righting reflex score induced by pentobarbital (P < 0.01, Kruskal-Wallis one-way ANOVA followed by Dunn's multiple comparisons test), but no such effects were observed with a low dose of MK-801 (0.03 mg/kg) or mecamylamine (5 mg/kg, P > 0.05; Table 2).

Effects of gabaculine, sarcosine, MK-801, and mecamylamine combined with pentobarbital on immobility
As presented in Fig. 4A, pentobarbital alone increased the percent loss of movement in response to noxious stimuli (immobility) in a dosedependent manner, with an ED 50 of 50 mg/kg. Gabaculine (50 mg/kg) shifted the pentobarbital dose-response curve for immobility to the left, and the ED 50 for pentobarbital decreased from 50 to 28 mg/kg. Similarly, sarcosine (10 mg/kg) slightly decreased the ED 50 of immobility for pentobarbital from 53 to 36 mg/kg (Fig. 4B). In addition, a high dose of MK-801 (0.2 mg/kg) reduced the ED 50 for pentobarbital from 58 to 29 mg/kg, but a low dose (0.03 mg/kg) only slightly altered the ED 50 from 58 to 44 mg/kg (Fig. 4C). Conversely, mecamylamine (5 mg/kg) induced no substantial change in the ED 50 for pentobarbital, as it shifted from 56 to 51 mg/kg (Fig. 4D).

Discussion
The present study demonstrated that a low dose of gabaculine (50 mg/kg, i.p.), which significantly increased endogenous GABA levels in the CNS without affecting behaviors per se ( Supplementary Figs. 2 and  3A), enhanced the reduction of grip strength, impairment of the righting reflex, and immobility induced by low doses of pentobarbital (Table 2 and Fig. 3A and Fig. 4A). MK-801 at the lowest dose augmented only the masked muscle-relaxing effect of pentobarbital (Fig. 3C). Sarcosine enhanced pentobarbital-induced immobility alone among components (Fig. 3B). Conversely, mecamylamine had no effect on any pentobarbital-induced behavior ( Table 2 and Fig. 3D and Fig. 4D). These findings suggest that each component of anesthesia induced by pentobarbital is mediated by GABAergic neurons and that muscle relaxation and immobility induced by pentobarbital may partially be associated with NMDA receptor antagonism and activation of spinal glycinergic neurons, respectively.
In this study, components of general anesthesia were evaluated using three behavioral endpoints: grip strength (as an index of muscle relaxation), righting reflex (as a measure of unconsciousness), and loss of movement in response to noxious stimulation (as a surrogate of immobility). I.p. administration of pentobarbital produced these behaviors in a dose-dependent manner. At a dose of 20 mg/kg, pentobarbital significantly reduced grip strength (Fig. 1A), but it did not impair the righting reflex. The mean righting reflex score was less than + 1 throughout the recording period (Fig. 1B). Pentobarbital (40 mg/kg) induced LORR in all tested mice (Table 1). However, even at 45 mg/kg, the anesthetic did not affect immobility in response to noxious stimuli. A high dose (80 mg/kg) induced immobility in all examined animals (Fig. 4). Unconsciousness is defined as a state of lacking normal sensory awareness. As previously stated, we assessed unconsciousness using the righting reflex. Immobility was also examined under LORR. However, the righting reflex depends on both sensory input and motor output in the CNS. To demonstrate unconsciousness devoid of motor confounds, we examined EEG and EMG endpoints together with the righting reflex. EEG changes after the intracerebroventricular infusion of pentobarbital at a rate of 4 μL/min for 50 min in rats were reported to reveal an increase in only α and β power at 3-14.5 min post-infusion, suggesting that this reflects a sedative component of the drug [29]. As shown in Fig. 2, 20 mg/kg pentobarbital significantly increased α and β power for approximately 40 min after injection, but it did not alter EMG power (Fig. 2Ac, Ab and Af). Compared with the reduction in grip strength over 20 min and the slight impairment of the righting reflex within 10 min post-injection (Fig. 1), the increases in α and β power were maintained for 40 min, suggesting that the alteration in EEG power is more sensitive than the examined behaviors. In contrast to the dose of 20 mg/kg, 40 mg/kg increased δ power as well as α and β power and decreased EMG power for approximately 30 min, although a significant increase in δ power was detected only from 10 to 20 min and δ power was increased by 50% from 20 to 30 min without statistical significance after injection (Fig. 2Ba). Jugovac et al. [29] suggested that simultaneous δ, α, and β power increases are correlated with loss of consciousness (although they Table 2 Effects of gabaculine, sarcosine, MK-801, and mecamylamine on percentage of loss of the righting reflex (LORR), onset of LORR, duration of LORR, and total righting reflex score induced by intraperitoneal (i.p.) administration of 30 mg/kg pentobarbital in mice. Gabaculine was administered i.p. 17 h before pentobarbital. Sarcosine was injected intravenously 5 h before pentobarbital. MK-801 and mecamylamine were injected i. p. 30 min before pentobarbital. The righting reflex was assessed every 2 min for 4 h after the i.p. injection of pentobarbital using righting reflex scores. A score of 0 indicated a normal righting reflex; a score of + 1 indicated that the mouse righted itself within 2 s in all three trials (slightly impaired righting reflex); a score of + 2 indicated that the time to righting in the best response among three trials was > 2 s but < 10 s (moderately or severely impaired righting reflex); a score of + 3 corresponded to absence of the righting reflex (no righting within 10 s in all three trials). The onset of LORR is the beginning time of a score of + 3. The time between LORR (a score of + 3) and the time mice regained that ability (a score of + 2) is the duration of LORR. The total score is the total of the scores recorded every 2 min after drug administration. Data are shown as the mean ± SEM. The number of animals used in each group is shown in parentheses. *P < 0.05, **P < 0.01 compared with pentobarbital, Mann-Whitney U test, † †P < 0.01 compared with pentobarbital, Kruskal-Wallis one-way ANOVA followed by Dunn's multiple comparisons test.
did not examine EMG power), which is roughly consistent with our data for the duration of LORR being 28.0±7.0 min (n = 6) ( Table 1 and Fig. 2B). They also suggested that θ and δ power increases reflect surgical anesthesia. At 40 mg/kg, there was no period of simultaneous θ and δ power increases, and this dose induced no immobility. Thus, EEG and EMG power could be indicators of unconsciousness and immobility.
To examine the effect of drugs used in this study on muscle relaxation, we employed the grip strength test. In previous reports, grip strength was assessed using a strain gauge to measure the pull force in grams exerted by mice as an index of muscle relaxation [30]. The muscle relaxants diazepam (benzodiazepine receptor agonist) and baclofen (GABA B receptor agonist) and the neuromuscular blocking agents succinylcholine, pancuronium, and dantrolene dose-dependently reduced grip strength in this test. 2-Amino-7-phosphonoheptanoic acid, a selective NMDA receptor antagonist, also suppressed grip strength [28,29]. These findings indicate the pertinence of the grip strength test as an indicator of muscle relaxation.
In the present study, low doses of gabaculine (50 and 70 mg/kg) did not reduce grip strength, but a high dose (100 mg/kg), which was the ED 50 for LORR as reported previously [15], reduced grip strength ( Supplementary Fig. 3A), suggesting that muscle relaxation induced by this dose reflects impairment of the righting reflex. In contrast, low doses of MK-801 (0.05-0.2 mg/kg) dose-dependently reduced grip strength ( Supplementary Fig. 3C), but even the highest dose (0.2 mg/kg) did not affect the righting reflex. Therefore, both the lowest (0.03 mg/ kg) and highest (0.2 mg/kg) doses of MK-801 were used in the subsequent experiments to examine the effects of MK-801 combined with pentobarbital on behaviors. Therefore, we were able to distinguish the efficacy of MK-801 in nature from its motility disturbance actions because 0.03 mg/kg of the drug had no effect on muscle relaxation ( Supplementary Fig. 3C). Indeed, the lowest dose of MK-801 did not enhance impairment of the righting reflex and immobility induced by pentobarbital, whereas the highest dose enhanced these components ( Table 2 and Fig. 4). In contrast, the lowest dose of MK-801 enhanced pentobarbital (10 mg/kg)-induced muscle relaxation for a short time ( Fig. 3C) but had no effect alone. 2-Amino-7-phosphonoheptanoic acid is a muscle relaxant, and therefore, MK-801 might reduce grip strength via an NMDA receptor antagonistic mechanism in the CNS. Similarly, a low dose of gabaculine potentiated pentobarbital-induced suppression of grip strength (Fig. 3A). Baclofen is a known muscle relaxant [30,31], but it is unknown whether the specific GABA A receptor agonist muscimol is also a muscle relaxant. Gabaculine is an indirect GABA receptor agonist that stimulates both GABA A and GABA B receptors. Of note, a recent lipid raft-protein interaction study revealed that both NMDA and GABA A receptors are considered the main targets of pentobarbital [32], consistent with the findings of a behavioral study [9]. Furthermore, this finding may be supported by the in vitro and in vivo evidence that benzodiazepine-sensitive GABA A receptors with the α1 subunit inhibit the NMDA/nitric oxide pathway [33][34][35][36]. Thus, the potentiation of pentobarbital-induced muscle relaxation by gabaculine may be , and mecamylamine (D) in combination with pentobarbital on immobility in response to noxious stimuli in mice. Gabaculine (50 mg/kg) or vehicle was administered intraperitoneally (i.p.) 17 h before the i.p. injection of pentobarbital at various doses. Sarcosine (10 mg/kg) or vehicle was injected intravenously (i.v.) 5 h before the i.p. injection of pentobarbital at various doses. MK-801 (0.03 or 0.2 mg/kg) or vehicle was administered i.p. 30 min before the i.p. injection of pentobarbital at various doses. Mecamylamine (5 mg/kg) or vehicle was given i.p. 30 min before the i.p. injection of pentobarbital at various doses. Pentobarbital produced its peak effect on the righting reflex within 10 min after the administration. Therefore, the tail clamp was applied 10 min after the pentobarbital injection, and movement in response to tail-clamp stimulation was evaluated at the time. The doses of pentobarbital are plotted against the percent effect of animals that lost the movement response to tail-clamp stimulation on the logarithmic-probability (X-Y axis) scale graph (seven animals per dose, 5-8 doses per dose-response curve). Each point represents the percent effect of 7 animals per dose (see Materials and methods for ED 50 values). mediated through both NMDA receptor blockade and GABA receptor activation.
In addition to gabaculine, sarcosine also potentiated pentobarbitalinduced immobility (Fig. 4B). A review article revealed that conventional volatile anesthetics induce immobility via molecular targets in the spinal cord, including glycine receptors, GABA A receptors, glutamate receptors, and TREK-1 potassium channels. In contrast, i.v. anesthetics cause immobility almost exclusively via GABA A receptors harboring β3 subunits [37]. However, pentobarbital has been reported to enhance glycine-evoked currents at glycine (α1) receptors expressed in Xenopus laevis oocytes [38]. In addition, using the whole-cell patch-clamp technique in rat spinal dorsal horn neurons, pentobarbital was shown to prolong spinal glycinergic miniature inhibitory postsynaptic currents by slowing the desensitization and deactivation of glycine receptors [4]. These studies suggest that pentobarbital-induced immobility involves activation of both GABAergic and glycinergic neurons.
Gabaculine, sarcosine, and MK-801 induced synergistic effects on pentobarbital-induced anesthetic endpoints via their specific actions, i. e., increases in endogenous GABA or glycine in the synaptic clefts or selective NMDA receptor channel blockade, respectively. However, these drugs might directly act on GABA A receptors to produce an additive effect. Nevertheless, no binding assay of these drugs has been reported at present. In addition, neither in vitro binding assays nor in vivo behavioral studies using GABA A receptor subunit knockout mice have been performed to examine the anesthetic effects of these drugs. However, the hypnotic effect of pentobarbital was reduced in GABA A receptor α1 subunit knockout mice compared with wild-type mice [39], but this effect was not observed in β3 subunit knockout mice [40]. Furthermore, in the present study, we did not investigate whether these drugs enhance pentobarbital-induced anesthetic effects through the same site of action, although these drugs act on the same circuitry that is driven by pentobarbital. This effect could be examined using tests with isoboles because when two drugs act together to produce a combined effect that is consistent with their individual potencies (probably via the same action site) it is termed additive; when the effect is greater than additive it is termed synergistic (via different action sites); and when the effect is less than additive it is termed sub-additive (via antagonistic action) [41]. Thus, further studies will be necessary to determine the effect, especially for the righting reflex and immobility tests, for which the response was all-or-none.
In the current study, all drugs were administered systemically. However, after administration of drugs into the brains of conscious animals, there may be significant interest in the observed pharmacological effects. Intracerebral injection procedures provide a better estimation of the central actions of drugs because diffusion through the blood-brain barrier is not involved. However, there is a possibility that leakage of the drug through the needle track followed by absorption from the subarachnoid space into the systemic circulation can occur and that the volume injected is greater than the cerebrospinal fluid volume. Therefore, if the drug leaks from the foramina and passes over the external surface of the brain, it might activate many other sites of nervous activity through direct action [42]. In addition, tissue damage caused by insertion of the needle into the brain may also affect the drug action results. When examining the effect of a drug on LORR, we may choose the ventrolateral preoptic nucleus, tuberomammillary nucleus, mesopontine tegmental anesthesia area, or pontine reticular formation region, all of which are involved in LORR induction [10,[43][44][45], as a target of the intracerebral injection of GABA A receptor ligands. However, the regions responsible for muscle relaxation and immobility are unclear at present. Thus, further studies are required to identify the neuronal circuits involved in the muscle relaxant and immobility actions of general anesthetics. Furthermore, in this study, gabaculine was used as an indirect GABA receptor agonist. Thus, the present results could support future studies using selective ligands of GABA A and GABA B receptors to clarify the role of GABA receptor subtypes.
In conclusion, all of the components of anesthesia induced by pentobarbital, namely muscle relaxation, unconsciousness, and immobility, might be mediated through GABAergic neurons. Muscle relaxation and immobility induced by pentobarbital may partially be associated with NMDA receptor antagonism and spinal glycinergic neuron activation, respectively. These properties may be useful clinically to reduce the adverse effects of another anesthetic, especially at higher concentrations, and induce required components when combined with pentobarbital.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
Data will be made available on request.