Behavior and Cellular Evidence for Propofol-induced Hypnosis Involving Brain Glycine Receptors

The experimental protocol conformed to the Guide of National Institutes of Health for the Care and Use of Laboratory Animals and was fully approved by the Institutional Animal Care and Use Committee of the University of Medicine and Dentistry of New Jersey, Newark, Jersey.

For in vivo  experiments, female adult Sprague-Dawley rats (250–350 g) were individually housed under controlled conditions (20–22°C), with plentiful access to water and food ad libitum .

The primary endpoint for evaluation of the hypnotic state was the LORR, which was defined as the inability of animals to right themselves when positioned in a supine position. Specifically, after injection of propofol or ketamine, the rats were gently turned onto their backs immediately, and thereafter at 5-min intervals. If they did not regain their posture within 10 s, then they were recorded as having lost their righting reflex. The estimation of LORR was made by observers who did not know what previous drug treatment the animals had received. In experiments measuring hypnosis, the righting reflex was considered to be restored when the animals first regained an upright position by standing on their feet. The onset time, duration, and percentage of animals exhibiting LORR were measured.

Rats were anesthetized using sodium pentobarbital (20 mg/kg, intraperitoneal injection), prepared for aseptic surgery, and secured into a stereotaxis frame. Cannulae (22-gauge, 11-mm length) (Plastics One Inc., Roanoke, VA) were positioned for injection into the intracerebroventricular (ICV) space at 5.2 mm anteroposterior, -1.0 mm mediolateral, and −9.1 mm dorsoventrical from the bregma.16Cannulae were affixed with dental resin (orthodontic resin, Caulk Company, Mitford, DE), and the animals were allowed to recover for at least 7 days. After completion of all experiments, the animals were sacrificed by administering an overdose of sodium pentobarbital and were perfused through the heart with 10% buffered formalin. Frozen sections of the brain were cut on a cryostat (50 μm) and stained with cresyl violet to assess the implantation position of the cannulae.

The brain slices were prepared as described previously.10,17In brief, rats, aged 12–24 days, were anesthetized and then sacrificed by decapitation. The brain was quickly excised and coronally sliced (300 μm) with a VF-200 Slicer (Precisionary Instruments Inc., Greenville, NC). This was done in icecold modified glycerol-based artificial cerebrospinal fluid saturated with 95% O²/5% CO²(carbogen) containing (in mm): 250 glycerol, 2.5 KCl, 1.2 NaH²PO⁴, 1.2 MgCl², 2.4 CaCl², 26 NaHCO³, and 11 glucose.17Midbrain slices were then kept in carbogen-saturated regular artificial cerebrospinal fluid at room temperature (22–24°C) for at least 1 h before use. Regular artificial cerebrospinal fluid has the same composition as glycerol-based artificial cerebrospinal fluid, except that glycerol was replaced with 125 mm NaCl.

For cell isolation, midbrain slices containing the posterior hypothalamus were first incubated in oxygenated standard extracellular solution containing 0.3 mg/ml papain (from papaya latex; Sigma Aldrich, St. Louis, MO) at room temperature for 15 min. The standard extracellular solution contained (in mm): 140 NaCl, 5 KCl, 2 CaCl², 1 MgCl², 10 HEPES, and 10 glucose (320 mOsm, pH set to 7.3 with Tris base). The slices were then incubated in enzyme-free standard extracellular solution. The posterior hypothalamic region was dissected out under an inverted microscope, and single cells were dissociated by trituration using two fire-polished glass pipettes with gradually narrowing diameters. The cells settled to the bottom of the culture dish within 20 min and were ready for electrophysiological recordings.

Whole cell currents were recorded at a holding potential of 0 mV using an Axopatch 200B amplifier, via  a Digidata 1322A analog-to-digital converter, and pClamp 9.2 software (Molecular Devices Co., Foster City, CA) at room temperature (22–24°C), using a pipette solution containing (in mm): 135 CsF, 5 KCl, 2 MgCl², 10 HEPES, 2 Mg adenosine triphosphate, 0.2 guanosine 5'-triphosphate, pH 7.2, osmolarity 280–300 mOsm.

Propofol, ketamine, glycine, strychnine, and GABAzine (SR-95531) (Sigma Aldrich) were prepared in normal saline solutions. Solutions were applied to a dissociated neuron with a superfusion system via  a Y-tube perfusion system.18The propofol used for intravenous (IV) administration was from AstraZeneca Pharmaceuticals LP (Wilmington, DE).

Data were statistically compared using a Student t  test at a significance level of P < 0.05 using Sigma plot (Systat Software Inc., San Jose, CA) for statistical analyses. For all experiments, average values are expressed as mean ± SEM. Dose-response data were fitted as previously described3,19to a logistic equation of the following form:

P = 100 Dⁿ/ (Dⁿ+ (ED⁵⁰)ⁿ)

where P is the percent of the population anesthetized, D is the drug dose, n is the slope parameter, and ED⁵⁰is the drug dose for a half-maximal effect.

We first tested whether systemic (intraperitoneal) administration of strychnine could attenuate propofol-induced hypnosis. We used the LORR score as our primary measure for hypnosis, because the concentrations of anesthetics that are necessary to produce the hypnotic state in humans are similar to those needed to induce LORR in rodents.1,3,20,21As expected, strychnine (intraperitoneal) dose-dependently reduced the percentage of rats exhibiting LORR in response to propofol (intraperitoneal). Specifically, 0.1, 0.3, 0.5, and 0.75 mg/kg strychnine reduced the LORR induced by 100 mg/kg propofol to 100%, 100%, 75%, and 56%, respectively. A comparable result was observed when strychnine was given subcutaneously. Strychnine (0.75 mg/kg, intraperitoneal) induced a large rightward shift of the LORR dose–response curve to propofol and significantly increased the median effective dose (ED⁵⁰) of propofol from 69.0 ± 0.8 mg/kg (mean ± SEM) to 96.0 ± 2.3 mg/kg (P = 0.048, fig. 1A). The subcutaneous administration of GABAzine also reduced the percentage of rats exhibiting LORR in response to propofol (intraperitoneal). The effect was dependent on the concentrations of GABAzine; 3 and 5 mg/kg GABAzine reduced the LORR induced by 100 mg/kg propofol to 66.6% and 33.3%, respectively. The systemic administration of GABAzine (5 mg/kg, subcutaneous injection) resulted in a rightward shift of the LORR dose–response curve to propofol and increased the ED⁵⁰of propofol to 104.5 ± 5.1 mg/kg (P = 0.028, fig. 1A). There was no significant difference between the dose–response curve to propofol generated in the presence of GABAzine (5 mg/kg, subcutaneous injection) and that in the presence of strychnine (0.75 mg/kg, intraperitoneal) (P = 0.39). Ketamine exemplifies another class of anesthetics which has no effect on GlyRs, but reduces excitatory neurotransmission by inhibiting the N-methyl-D-aspartic acid subtype of glutamate receptors.22The percentage of animals exhibiting LORR in response to 35, 50, and 150 mg/kg (intraperitoneal) of ketamine was not changed in the absence or presence of strychnine (0.75 mg/kg, intraperitoneal, fig. 1B). Accordingly, the ED⁵⁰of ketamine was virtually the same in the absence (33.3 ± 0.5 mg/kg) and the presence (33.2 ± 10.4 mg/kg) of strychnine (P > 0.8, fig. 1B). This finding shows that systemic strychnine does not act “nonspecifically” to attenuate responses to all anesthetic agents; for example, by causing a generalized increase in neuronal excitability. These data indicate that propofol is less effective as a hypnotic agent in the presence of a GlyR antagonist and probably exerts its hypnotic effect, in part, through GlyRs. Of note, strychnine reduced LORR in 25% of 2 of the 8 rats receiving 100 mg/kg ketamine. The underlying mechanism for this effect warrants further investigation.

To confirm that GlyRs located in the brain are responsible for the effect of strychnine, strychnine was injected into the ICV space, and propofol (IV) was used to induce LORR in a dose-dependent manner (fig. 2A1). Strychnine (50 μg/25 μl, ICV) induced a rightward shift in the percentage of rats exhibiting LORR induced by propofol (intraperitoneal) and increased the median effective dose (ED⁵⁰) of propofol from 5.18 ± 0.02 mg/kg to 5.53 ± 0.07 mg/kg (P = 0.017, fig. 2A1).

In addition, strychnine (50 μg/25 μl, ICV) significantly prolonged the onset time (control, 4.6 ± 0.2 s; strychnine, 6.0 ± 0.3 s); P < 0.01, n = 6, fig. 2A2), and reduced the duration (control, 19.03 ± 3.04 min; strychnine, 10.77 ± 1.39 min; P < 0.05, n = 6) of LORR induced by 10 mg/kg propofol (IV, fig. 2A3). Similar results were observed when other doses (9 and 12 mg/kg) of propofol were used (fig. 2A2 and 2A3). Thus, propofol (IV) is less effective as a hypnotic in the presence of strychnine (ICV).

As a comparison, we tested the LORR induced by ketamine (IV) in the absence and presence of strychnine (50 μg/25 μl, ICV). As expected, ketamine (20–30 mg/kg, IV) induced LORR in a dose-dependent manner. The percent LORR, the onset time, and the duration of LORR induced by ketamine (IV) were essentially the same with or without strychnine (fig. 2B1–3).

These data indicate that the attenuation of propofol-induced hypnosis by strychnine is a result of blocking the brain GlyRs, instead of nonspecific excitatory effects in the central nervous system.

The above in vivo  experiments indicate that propofol-induced hypnosis involves brain GlyRs. We next examined the effects of propofol on I^Glyusing patch clamp techniques. The I^Glywere elicited by the application of glycine to neurons isolated from the posterior hypothalamus of rat brains. All neurons tested responded to the application of glycine. The posterior hypothalamus is a key region in the brain sleep pathway.3,23At a holding potential of 0 mV, application of glycine (30 μm) elicited an outward current, which was completely abolished by strychnine (fig. 3A), indicating the presence of strychnine-sensitive GlyRs in the posterior hypothalamic neurons. When applied together with propofol, the current was substantially larger. However, strychnine eliminated this current, indicating that it was mediated by strychnine-sensitive GlyRs. We next measured the I^Glyelicited by 10 μm glycine in the presence of varying concentrations of propofol. When applied separately, glycine and propofol each elicited an outward current (fig. 3B). To quantify the propofol-glycine interaction, we first normalized the peak current amplitude elicited by 10 μm propofol, or by the mixture of glycine plus propofol to the peak current induced by 10 μm glycine. On average, the peak current amplitude elicited by 10 μm propofol was 108.5 ± 7.5% (n = 7) of that elicited by 10 μm glycine. The peak amplitude elicited by the mixture (10 μm glycine and 10 μm propofol) was 382.2 ± 98.5% (n = 7) of that elicited by 10 μm glycine. This value is considerably greater than the sum of 10 μm glycine and 10 μm propofol (100 + 109 = 209). This result indicates that propofol and glycine have a synergistic effect. Figure 3Csummarizes the result of 3 to 7 cells, indicating that the peak amplitude of the I^Glyincreased with the increasing concentrations of propofol. That is, the normalized values of peak amplitude elicited by 10 μm glycine in the presence of 0, 3, 10, 30, and 100 μm propofol were 0, 54.8 ± 5.2% (n = 3, P = 0.06), 382.2 ± 98.5% (n = 7, P < 0.01), 385.6 ± 128.9% (n = 5, P < 0.05), and 431.3 ± 29.2% (n = 3, P < 0.01), respectively. The apparent EC⁵⁰for propofol was 5.4 ± 0.6 μm.

To further characterize the propofol potentiation of I^Gly, we evaluated the effects of propofol on currents elicited by varying concentrations of glycine (3, 10, 15, 30, and 100 μm). While propofol profoundly increased the current elicited by 3, 10, and 15 μm glycine, it had no significant effect on the current elicited by 30 and 100 μm glycine (fig. 4, A1–A3, B1, and B2). Propofol (10 μm) shifted the concentration–response curve of glycine to the right and decreased the EC⁵⁰for glycine from 35.7 ± 3.1 μm in controls to 11.1 ± 2.9 μm in the propofol group (fig. 4A3). Similarly, 1 μm of propofol enhanced the currents induced by 10, 15, 30, and 100 μm glycine by 1395 ± 567% (n = 4, P = 0.03), 741 ± 144% (n = 5, P = 0.01), 17 ± 19% (n = 5, P = 0.5), and 8 ± 15% (n = 5, P = 0.6), respectively. These data suggest that propofol acts on the GlyR to increase its affinity for the agonist or it functions as a positive allosteric modulator of GlyRs.12Similarly, propofol (1 μm) enhanced currents induced by 0.1, 1, 10, and 100 μm GABA by 293 ± 158% (n = 3, P = 0.2), 497 ± 114% (n = 3, P = 0.04), 55 ± 42% (n = 4, P = 0.16), and -44 ± 16% (n = 3, P = 0.14), respectively (fig. 4, C1 and C2).

Propofol alone dose-dependently induced an outward current in hypothalamic neurons (fig. 5A1). Propofol (1, 3, 10, 30, and 100 μm) induced currents with the amplitudes of 15 ± 5 (n = 5), 35 ± 18 (n = 5), 174 ± 49 (n = 11), 441 ± 85 (n = 11), and 877 ± 199 pA (n = 4), respectively. The apparent EC⁵⁰of propofol alone was 26.4 ± 3.4 μm (fig. 5A2). Strychnine (1 μm) and bicuculline (10 μm) suppressed the current induced by 30 μm propofol by 46.6 ± 3.7% (n = 11, P < 0.01) and by 58.9 ± 4.9% (n = 4, P < 0.05), respectively (fig. 5B1–B3).

Our major finding is that the GlyRs in the brain contribute to propofol-induced hypnosis. Our result shows that strychnine reduced the percentage of animals exhibiting LORR in response to propofol, prolonged the onset time but reduced the duration of LORR induced by propofol, and did not alter the LORR induced by ketamine. These results indicate that strychnine attenuation of propofol-induced LORR is a result of the blockade of brain GlyRs, instead of a generalized increase in neuronal excitability. Finally, in keeping with in vivo  observations, our in vitro  patch–clamp data indicate that propofol potentiates the I^Glyfrom posterior hypothalamic neurons. Thus, the current investigation provides the first behavioral and cellular evidence indicating that brain GlyRs contribute to the hypnotic effect of propofol.

Interestingly, systematic strychnine produced an effect comparable to that of systemic GABAzine regarding the attenuation of LORR induced by propofol (fig. 1A). These data indicate that both GlyRs and GABA^Areceptors play a crucial role in the propofol-induced hypnotic state. Our data are in general agreement with a previous study showing that GABA^Areceptor blockade antagonizes the action of propofol but not ketamine.3,24Based on our observations, we hypothesize that propofol potentiates the function of GlyRs in neurons within the sleep pathway, increases the influx of chloride ions, and hyperpolarizes neurons. Thus, these appear to be, at least in part, the cellular and molecular consequences of propofol administration that contribute to its hypnotic effect.

Moreover, our in vivo  observations are supported by the results of our in vitro  experiments. Our patch–clamp data indicate that functional GlyRs exist in posterior hypothalamic neurons. Propofol significantly enhanced I^Gly, indicating that these GlyRs are sensitive to propofol. Our in vitro  observation is consistent with previous studies regarding propofol potentiation of I^Gly.12,13In addition, we observed that propofol substantially enhanced currents induced by GABA, which is consistent with previous reports.11–13,15,25Interestingly, the I^PROof neurons from different brain areas appear to have significantly different sensitivities to strychnine. Strychnine (1 μm) produced a 47% inhibition of the I^PROof posterior hypothalamic neurons, but only a 10% inhibition of the I^PROof hypothalamic paraventricular nucleus neurons.15We interpret these data to signify differences among brain regions.

GlyRs are modulated by a number of drugs, including volatile and IV anesthetics.26Volatile anesthetics, such as halothane, chloroform, and ether enhance the function of GlyRs in rat medullary neurons,27in rat hippocampal neurons,28in recombinant systems with transiently transfected cells,29and in Xenopus  oocytes.27,30GlyRs in the spinal cord have been recognized as the most credible candidates for mediating immobility caused by volatile anesthetics.31,32Furthermore, some IV anesthetics such as propofol and pentobarbital potentiate the cellular response to glycine in a homomeric expression system30and in spinal dorsal horn neurons.12Our current investigation provides compelling evidence that brain GlyRs play a considerable role in propofol-induced hypnosis.

The authors thank John Le, M.S. (New Jersey Graduate School, Newark, New Jersey); Urvi Bhavsar, B.A. (Rutgers University, Newark, New Jersey); Kimberly Sokol, M.S. (New Jersey Graduate School, New Jersey); Radhika Shah, B.A. (New Jersey Medical School, Newark, New Jersey); Yan Gao, B.A. (New Jersey Medical School, Newark, New Jersey); Linda Huang, B.A. (New Jersey Medical School, Newark, New Jersey); Ho Yin Aaron Shiu, B.A. (New York College of Osteopathic Medicine of the New York Institute of Technology, New York, New York); Victor La (Student, John Hopkins University, Baltimore, Maryland); Barrett Warden (Student, University of Pittsburgh, Pittsburgh, Pennsylvania); and Devang Patel (Student, North Bergen High School, North Bergen, New Jersey) for their technical help.