Xenon Attenuates Excitatory Synaptic Transmission in the Rodent Prefrontal Cortex and Spinal Cord Dorsal Horn

The experimental protocols were approved by the Ethical Committee on Animal Care and Use of the Government of Bavaria, Germany. For the experiments in the PFC, 28- to 42-day-old male C57Bl6 mice were killed by cervical dislocation, and the brains were rapidly removed and placed in ice-cold artificial cerebrospinal fluid (ACSF) containing 125 mm NaCl, 2.5 mm KCl, 25 mm NaHCO³, 2 mm CaCl², 2 mm MgCl², 25 mm D-glucose, and 1.25 mm NaH²PO⁴(all from RBI/Sigma, Deisenhofen, Germany). Saturation with a mixture of 95% O²–5% CO²(carbogen gas) led to a pH of 7.4. Sagittal slices (350 μm thick) were prepared using a microtome (HM 650 V; Microm International, Walldorf, Germany). For the experiments in the SG of the spinal cord dorsal horn, 14- to 28-day-old male Wistar rats were killed by cervical dislocation, and coronary slices of lumbar spinal cord were obtained as described previously.20All slices were allowed to recover in a storage chamber (34°C) for at least 1 h before being transferred to the recording chamber.

A platinum ring with nylon filaments was used to fix the slices on the bottom of the recording chamber, which was continuously perfused (2 ml/min) with ACSF. We used infrared videomicroscopy and the gradient contrast system (Zeiss, Oberkochen, Germany; for details, see Dodt et al.  21) to visualize the neurons within layer V of the PFC or within the SG. The SG was clearly identifiable as a translucent band across the superficial dorsal horn.

The patch pipettes were pulled from thin-walled borosilicate glass tubes with inner filament (OD 1.5 mm, ID 1.17 mm, GC150TF-10; Clark Electromedical Instruments, Pangbourne Reading, United Kingdom) and heat polished using a two-step horizontal puller (DMZ-Universal Puller; Zeitz-Instruments, Munich, Germany). Pipettes had an open tip resistance of 4–6 MΩ when filled with a solution containing 130 mm K-D-gluconate, 5 mm KCl, 0.5 mm EGTA, 2 mm MgCl², 10 mm HEPES, 5 mm D-glucose, and 20 mm Na²-phosphocreatine (all from RBI/Sigma). Currents were recorded with a switched voltage clamp amplifier (SEC 10 l; NPI electronic, Tamm, Germany) with switching frequencies of 60–80 kHz (25% duty cycle). Series resistance was monitored continuously and compensated in bridge mode. All patch clamp experiments were performed at room temperature (22°–25°C) to ameliorate oxygenation of the neurons.

Electrically evoked postsynaptic currents (ePSCs) were elicited by square pulse stimuli (6–100 V, 0.05–0.5 ms, interstimulus interval 15 s) delivered via  a bipolar tungsten electrode, which was placed either in layer V near the apical dendrite of the recorded neuron (5-μm electrode tip diameter) or in the dorsal root entry zone of the spinal cord slices (50-μm electrode tip diameter). Preceding each tissue stimulation, neuronal input resistance was determined by injecting a current pulse hyperpolarizing by 10 mV for 200 ms. In the PFC, in a subset of experiments, paired-pulse stimulation was performed by delivering the same stimulus at 50-ms interpulse intervals. The paired-pulse ratio was determined by dividing the amplitude of the second current response by the amplitude of the first one.

Photolytically evoked excitatory currents were induced upon focal photolysis of caged L-glutamate.21For this purpose, the beam of an ultraviolet laser (355 nm wavelength, frequency-tripled Nd:YVO4, 100-kHz pulse repetition rate; DPSS Lasers, San Jose, CA) was focused by the objective (×60, 0.9 numerical aperture; Olympus, Tokyo, Japan) on a small spot (5 μm in diameter) positioned on a dendrite approximately 10–20 μm from the soma. Laser stimulation was delivered alternating with electrical stimulation in intervals of 15 s. When a stable whole cell recording had been obtained, γ-α-carboxy-2-nitrobenzyl caged glutamate at a final concentration of 0.25 mm was added to the recirculating perfusate. Caged glutamate had no discernible effect on neurons per se .21Glutamate was released by Q-switching brief laser pulses (3–5 ms; intensity 1–2 mW), applied at regular intervals of 30 s throughout the experiment.

To isolate specific currents, we used d(−)-2-amino-5-phosphonopentanoic acid (AP5; 50 μm), 1,2,3,4-tetrahydro-6-nitro-2,3-dioxo-benzo[f]quinoxaline-7-sulfonamide (NBQX; 5 μm), 3-amino-propyl(diethoxymethyl)phosphinic acid (CGP35348; 200 μm), and bicuculline methiodide (20 μm). Of these four receptor antagonists, an appropriate cocktail was used, with NBQX omitted for AMPA, AP5 omitted for NMDA, and bicuculline methiodide omitted for GABA^Areceptor–mediated responses. For the experiments in the spinal cord, strychnine at a final concentration of 2 μm was added to the ACSF, and for recordings of NMDA receptor–mediated currents in the spinal cord, MgCl²was reduced to 0.5 mm. For AMPA receptor–mediated currents, the holding potential was set to −70 mV, and for NMDA and GABA^Areceptor–mediated currents, the holding potential was set to −30 and −50 mV, respectively. Miniature excitatory postsynaptic currents (mEPSCs) were continuously recorded (10 min) at −70 mV in the presence of 1 μm tetrodotoxin, 50 μm AP5, 200 μm CGP35348, and 20 μm bicuculline methiodide. mEPSCs were completely blocked in the presence of 5 μm NBQX (data not shown).

The whole cell responses were amplified, low-pass filtered (3 kHz), and then digitized (ITC-16 Computer Interface; Instrutech Corp., Port Washington, NY) with a sampling frequency of 9 kHz and stored to a hard drive (Power Macintosh G3 computer, data acquisition software Pulse version 8.5; Heka electronic GmbH, Lambrecht, Germany). The obtained data were analyzed with IGOR Pro software (WaveMetrics, Portland, OR). For detection of mEPSCs, the amplitude threshold was defined as the threefold amplitude of the baseline variance (noise level), and the events identified were subsequently verified visually. We then quantified the frequency and peak amplitudes of the detected events.

Under control conditions, the superfusing ACSF was exposed to a mixture of 65% N²–30% O²–5% CO²; for xenon application, the ASCF was gassed with a mixture of 65% xenon–30% O²–5% CO². In a subset of experiments, xenon was applied at a lower concentration using a gas mixture of 30% xenon–35% N²–30% O²and 5% CO². Replacement of N²by xenon did not change the pH of the gassed ACSF. We have shown repeatedly13,22that it is feasible to obtain stable synaptic responses from brain slice preparations for at least 40 min under these conditions with relatively low oxygen concentration. All gas mixtures were purchased from Linde AG (Unterschleissheim, Germany) and applied at a flow rate of 0.3–0.5 l/min to the ACSF reservoir. Oxygen and carbon dioxide concentrations were verified with a calibrated gas monitor (Datex Capnomac Ultima, Duisburg, Germany). Tubing was made of polytetrafluoroethylene (Teflon; VWR International, Darmstadt, Germany) to minimize loss of xenon.

Concentration measurements of dissolved xenon were accomplished using headspace gas chromatography (RCC Ltd., Itingen, Switzerland). When gassing the ACSF with the 65% (30%) xenon gas mixture, the concentration of xenon in ACSF was 1.9 ± 0.2 mm (1.1 ± 0.1 mm) (mean ± SD; n = 4). For comparison, other studies describe xenon saturation of 3.9 mm 23or 2.0 mm 4when gassing with pure xenon, 3.4 mm when gassing with 80 vol% xenon,6,7or 3.54 mm when gassing with 84 vol%,8which would result in a calculated mean of approximately 2.3 mm at an assumed 65 vol% xenon saturation. The difference from our obtained concentration of 1.9 mm might be explained by an increased evaporation due to our open-chamber system.

Tetrodotoxin, AP5, NBQX, bicuculline methiodide, strychnine, γ-α-carboxy-2-nitrobenzyl caged glutamate (all from RBI/Sigma), and CGP35348 (Novartis Laboratories, Basel, Switzerland) were bath applied at known concentrations via  the superfusion system.

For all statistical evaluations, SPSS Statistics version 16 (SPSS GmbH Software, Munich, Germany) was used. The 40-min recording time of ePSCs, electrically evoked inhibitory postsynaptic currents (eIPSCs), electrically evoked excitatory postsynaptic currents (eEPSCs), and photolytically evoked currents was first partitioned in equidistant subintervals of 5 min, and then the averaged relative amplitudes in each of them were determined. A two-factorial multivariate analysis of variance (MANOVA) with repeated-measures design was applied on the averaged relative amplitudes with interval as a within-subjects factor and kind of receptor as a between-subjects factor. Differences between the various intervals were tested by tests with contrasts, and differences between the various receptors and stimulation methods were tested by post hoc  tests. MANOVA was also performed for testing differences in the amplitudes and frequency of mEPSCs between the control and xenon recordings. As the nominal level of significance, we accepted α= 0.05. It was corrected (according to Bonferroni procedure) for all a posteriori  tests (tests with contrasts and post hoc  tests). Numerical data are presented as mean ± SEM with the number of experiments (= neurons) indicated, if not stated otherwise. In graphs where error bars are not shown, they are smaller than the size of the symbol.

In a first set of experiments, the effect of 1.9 mm xenon on compound synaptic transmission in the PFC or SG was examined. Compound postsynaptic currents upon electrical stimulation (ePSCs) were recorded from neurons in the PFC (layer V) or the SG. Ten to 15 min after xenon application, the ePSCs recorded from the PFC and SG were reduced to 70.1 ± 10.3% and 74.3 ± 4.2%, respectively (figs. 1A and B). The averaged relative amplitudes in this interval were significantly lower than in the two control intervals (tests with contrasts in MANOVA, PFC: P = 0.022, SG: P = 0.003), with no significant difference between PFC and SG (Wilks multivariate test, P = 0.661). Upon termination of xenon application, the ePSCs nearly reversed to control level. Application of xenon for a longer period (up to 25 min) did not result in an additional reduction of current responses (data not shown). Neither resting membrane potential (PFC: control −71.3 ± 0.9 mV, xenon −71.1 ± 0.2 mV; SG: control −50.6 ± 2.2 mV, xenon −52.8 ± 4.7 mV) nor input resistance (PFC: control 272.2 ± 26.7 MΩ, xenon 294.3 ± 29.9 MΩ; SG: control 823.7 ± 117.8 MΩ, xenon 789.9 ± 117.0 MΩ) of the neurons were changed in the presence of 1.9 mm xenon.

From neurons in the PFC and SG, electrically evoked GABA^Areceptor–mediated eIPSCs (GABA^A-eIPSCs) were recorded in the presence of 50 μm AP5, 5 μm NBQX, and 200 μm CGP35348 at a holding potential of −50 mV. Xenon, 1.9 mm, did not affect the amplitudes of GABA^A-eIPSCs (figs. 1C and D) in the PFC or in the SG. The decay of the GABA^A-eIPSCs was fitted biexponentially, with two time constants, τ^decayfast and τ^decayslow, and total charge transfer was calculated by integrating the area under the GABA^A-eIPSCs. In both CNS regions, we did not detect an apparent influence of xenon on either of these parameters (table 1).

In further sets of experiments, the impact of xenon on NMDA and AMPA receptor–mediated excitatory synaptic transmission in the PFC (fig. 2) and SG (fig. 3) was investigated.

N -methyl-d-aspartate receptor–mediated current responses were evoked using either electrical stimulation (NMDA-eEPSCs) or photolytic uncaging of glutamate (p-NMDA-Cs). Xenon, 1.9 mm, reversibly diminished NMDA-eEPSCs recorded from neurons in the PFC (SG) to 57.7 ± 4.4% (71.9 ± 3.4%) of control (PFC: n = 6, SG: n = 5; tests with contrasts in MANOVA, PFC: P = 0.008, SG: P = 0.004; figs. 2A and 3A). Photolytically evoked p-NMDA-Cs were reduced by xenon to 58.6 ± 3.6% (PFC; n = 6) and 71.5 ± 5.6% (SG; n = 5) of control (tests with contrasts in MANOVA, PFC: P = 0.001, SG: P = 0.002; figs. 2A and 3A). There was no significant difference in the degree of xenon-induced reduction when comparing NMDA-eEPSCs with p-NMDA-Cs in the PFC or in the SG (Bonferroni post hoc  tests in MANOVA, PFC: P = 0.951, SG: P = 0.937). Moreover, when comparing the impact of xenon on NMDA receptor–mediated currents in the PFC with that in the SG, we did not see a significant difference (Wilks multivariate test, NMDA-eEPSC: P = 0.166, p-NMDA-Cs: P = 0.104). The current decay of NMDA-eEPSCs was fitted biexponentially, with two time constants, τ^decayfast and τ^decayslow. In both CNS regions, a clear influence of xenon on deactivation kinetics of NMDA-eEPSCs was not detectable (table 2).

Electrically (AMPA-eEPSCs) and photolytically (p-AMPA-Cs) evoked AMPA receptor–mediated currents were recorded in the PFC and SG. Xenon reduced the amplitudes of AMPA-eEPSCs to 66.9 ± 4.1% (PFC; n = 6; fig. 2B) and 64.5 ± 6.5% (SG; n = 5; fig. 3B). p-AMPA-Cs were reduced in the presence of xenon to 63.7 ± 2.9 (PFC; n = 5; fig. 2B) and 58.2 ± 3.1% (SG; n = 5; fig. 3B) of control. Reduction of the amplitudes of AMPA-eEPSCs and p-AMPA-Cs was not significantly different in the PFC or in the SG (Bonferroni post hoc  tests in MANOVA, PFC: P = 0.826, SG: P = 0.209). Comparison of the xenon effect on AMPA receptor–mediated current responses in the PFC with that in the SG revealed no statistical significant difference (Wilks multivariate test, AMPA-eEPSC: P = 0.452, p-AMPA-Cs: P = 0.640). We did not see a conspicuous influence of 1.9 mm xenon on current decay kinetics of AMPA-eEPSCs (table 2).

In a subset of experiments, we tested whether a lower concentration of xenon (1.1 mm), which most likely corresponds to minimum alveolar concentration (MAC)^awake, affected NMDA and AMPA receptor–mediated current responses in the PFC. Xenon, 1.1 mm, reversibly diminished NMDA-eEPSCs to 76.5 ± 3.7% of control (n = 4; tests with contrasts in MANOVA, P < 0.05) with no significant difference from the reduction of p-NMDA-Cs caused by 1.1. mm xenon (78.0 ± 4.1%; n = 4; fig. 2A, triangles). Likewise, 1.1 mm xenon reduced the amplitudes of AMPA-eEPSCs (83.3 ± 3.4%; n = 4) and p-AMPA-Cs (79.9 ± 4.2%; n = 4) to a similar extent (fig. 2B).

Miniature excitatory postsynaptic currents were recorded in the presence of 1 μm tetrodotoxin (fig. 4). mEPSCs recorded from neurons in the PFC (SG) occurred at a frequency of 9.0 ± 0.7 (4.9 ± 2.1) Hz and had a mean amplitude of 4.2 ± 0.3 (5.7 ± 0.4) pA. Application of 1.9 mm xenon reduced the mEPSC amplitudes in the PFC (fig. 4A) and SG (fig. 4E). Figures 4B and Fshow the effect of xenon on cumulative distributions of mEPSC amplitudes and interevent intervals. Xenon, 1.9 mm, increased the proportion of mEPSCs having smaller amplitudes but had no effect on the distribution of interevent intervals. Figures 4C and Gshow pooled data from five experiments performed in the PFC and five experiments performed in the SG. Xenon reduced the mean mEPSC amplitude to 3.7 ± 0.1 (5.1 ± 0.2) pA, whereas the mean frequency remained unchanged.

In an additional set of experiments, AMPA-eEPSCs upon paired-pulse stimulation were recorded from PFC neurons. The paired-pulse ratio did not change when 1.9 mm xenon was applied (P = 0.111, n = 7; fig. 4D).

Recently, we demonstrated that the inhalational anesthetic xenon depresses excitatory while not affecting inhibitory synaptic transmission in the murine basolateral amygdala.13Because the effects of an anesthetic on the cellular level may vary depending on the local network under consideration,24,25we now investigated the effect of xenon in two additional CNS regions. In the current study, we show that in acute rodent slice preparations of the cortex and the spinal cord, both NMDA and AMPA receptor–mediated synaptic transmission are depressed by xenon, whereas GABA^Areceptor–mediated inhibitory synaptic transmission remains unchanged.

Based on published MAC values of xenon for humans (63–71 vol%)26,27and a solubility coefficient of 0.0887 for xenon at 37°C,28the calculation of MAC equivalents of dissolved xenon reveals a value of 2.2–2.5 mm. Therefore, with a measured xenon concentration of 1.9 ± 0.3 mm dissolved in ACSF at room temperature, the xenon concentration applied during most of our experiments lies close to MAC^immobility, whereas the lower concentration of 1.1 mm, which was applied in a subset of experiments, corresponds closely to MAC^awake.29 

The results provided in the current study further support the hypothesis that xenon does not act via  an enhancement of γ-aminobutyric acid (GABA)–mediated synaptic transmission. For intravenous30,31and volatile anesthetics,32as well as for nitrous oxide,23,33a substantial potentiating effect on GABA-mediated currents has been reported, and for intravenous and volatile anesthetics, there is growing evidence that they mediate their anesthetic properties, at least in part, via  an enhancement of inhibitory synaptic transmission (reviews: Campagna et al. ,34Rudolph and Antkowiak35). In the current study, we observed no effect of xenon on the amplitudes of GABA^A-eIPSCs recorded from native neurons in the PFC or in the SG. A slight effect of xenon on GABA^A-eIPSC deactivation time constants, and total charge transfer, which determines the strength of inhibition,36,37cannot be completely excluded by our data. However, we certainly can exclude a pronounced effect, as it has been described, e.g. , for volatile anesthetics, prolonging IPSC time constants to 260–500%.32,38Our results are in accord with studies5,6showing that xenon exerts no effects on GABA^A-IPSCs evoked from cultured hippocampal neurons, but contrasts with two studies describing a xenon-induced increase of Cl⁻currents through GABA^Areceptors, which were heterologously expressed in Xenopus  oocytes4or human embryonic kidney 293 cells.23In the latter studies, the xenon-induced potentiation occurred only at nonsaturating GABA concentrations, which are below the range that occurs synaptically,36thus possibly explaining the discrepancy with our data. However, the possibility, that extrasynaptic GABA^Areceptors, which are activated at subsaturating GABA concentrations (reviews: e.g. , Walker and Semyanov,39Hemmings et al.  40) and mediate a tonic neuronal inhibition, are affected by xenon, cannot be ruled out by our data.

In both PFC and SG, compound synaptic transmission was reduced under xenon. For a further detailed investigation of the underlying mechanisms, we used pharmacologic isolation to separately investigate the effect of xenon on NMDA and AMPA receptor–mediated synaptic transmission.

In both CNS regions, we observed a pronounced xenon-induced reduction of NMDA receptor–mediated currents without apparent change of current decay kinetics. An NMDA receptor–depressing effect has been described for, e.g. , propofol, nitrous oxide, ketamine, and also some volatile anesthetics (review: Rudolph and Antkowiak35), and a depressant effect on the NMDA receptor has been suggested as the main mechanism for the xenon anesthetic state.5NMDA receptor–mediated currents recorded from amygdalar neurons in a brain slice preparation,13from neurons in culture,5,6and from heterologously expressed NMDA receptors4were reported to be reduced by xenon. In the current study, in the PFC as well as in the SG, photolytically evoked NMDA receptor currents, which are generated beyond the influence of the presynaptic terminal, were depressed by xenon to the same amount as were electrically evoked NMDA receptor–mediated currents. This finding provides strong evidence for a mainly postsynaptic mechanism of xenon on NMDA receptors in the PFC and SG.

Similar to our recently published results in the basolateral amygdala, also AMPA receptor–mediated currents in the PFC and SG were reduced by xenon. These results further support the hypothesis that the xenon anesthetic state is not only due to an NMDA receptor depression, as has been suggested initially.5In fact, evidence for a participation of the AMPA receptor in mediating the xenon anesthetic state derives not only from in vitro  electrophysiologic7,8but also from in vivo  studies using the model organism C. elegans  10or human volunteers.11,41 

Because AMPA receptor currents upon electrical stimulation and upon photolytic uncaging of L-glutamate were reduced to the same amount, we propose a postsynaptic mechanism of xenon action, both in the PFC and in the SG. The analysis of the effect of xenon on AMPA receptor–mediated mEPSCs in the PFC and SG, as well the analysis of the xenon effect on the paired-pulse ratio of AMPA-eEPSCs in the PFC, further support this hypothesis. Xenon reduced mEPSC amplitudes while not affecting mEPSC frequency, which speaks, according to the classic interpretation of mEPSC analysis,42in favor of a postsynaptic mechanism of xenon action. Changes in paired-pulse ratio and neurotransmitter release are inversely correlated.43,44Therefore, our findings of decreased AMPA-eEPSCs with unchanged paired-pulse ratio suggest that xenon inhibits excitatory synaptic transmission without changing the probability of presynaptic neurotransmitter release.

The analysis of current kinetics of GABA^A, NMDA, and AMPA receptor–mediated currents revealed considerably faster current responses recorded from SG neurons compared with neurons in the PFC. Possible explanations for these discrepancies might be species differences and/or the approximately threefold higher input resistance of the SG neurons, which reflects a smaller cell size and thus a less pronounced current distortion typically produced by space clamp deficits.

When comparing the amount of reduction of NMDA and AMPA receptor–mediated currents in the PFC and the SG, we did not see a difference between the two CNS regions. Furthermore, when comparing the results of the current study with the recently published results of the xenon effect in the basolateral amygdala,13no significant difference occurred in terms of the impact of xenon on NMDA or AMPA receptor–mediated currents recorded in the PFC, SG, or amygdala (Wilks multivariate test, P > 0.05). We therefore hypothesize an equipotent xenon action on the cortical, subcortical, and spinal levels.

The PFC is essentially involved in higher cognitive functions such as working memory,45–47attention regulation,48level of wakefulness,49–51and consciousness.17,52,53Therefore, the observed xenon action in PFC neurons might account for the hypnotic properties of xenon, in fact, at concentrations close to MAC^immobilityand close to MAC^awake. Moreover, because layer V pyramidal neurons in the PFC represent the major output neurons of the PFC,54a xenon-induced disruption of excitatory synaptic transmission to these cells might also affect further important subcortical processes regulated by the PFC.

Current knowledge regarding the action of xenon on the spinal cord level is limited to few studies. It has been shown that xenon depresses spinal cord dorsal horn neuronal activity in vivo ,55,56slows ventral root potentials and spinal monosynaptic reflexes in vitro ,57and attenuates long-term potentiation of C-fiber evoked potentials.58However, precise insights into the mechanisms of xenon action on dorsal horn level are missing. In our study, we showed that xenon depresses both NMDA and AMPA receptor–mediated synaptic transmission to SG neurons, presumably via  postsynaptic mechanisms. Nociceptive signals transmitted by thin (Aδ) and unmyelinated (C) primary afferent fibers are conveyed to neurons in the superficial laminae (SG) of the spinal cord dorsal horn as the first site of their synaptic integration,18,19and inhibition of SG neuron activity might be a key mechanism of antinociception.59Therefore, the observed xenon effect on NMDA and AMPA receptor–mediated synaptic transmission might crucially account for the profound analgesic60,61properties of xenon. In addition, another endpoint of general anesthesia, immobility, is largely produced by anesthetic action on the spinal cord level.62–65It is still a matter of debate whether depression of ventral or dorsal horn neuronal activity contributes more to immobility.66,67Although accumulating evidence suggests the ventral horn as the primary site,67,68the involvement of dorsal horn depression to produce immobility cannot be ruled out.66,69Hence, the depression of synaptic transmission to SG neurons observed in our study might also account for the immobilizing properties of xenon.

In summary, we have shown that a clinical relevant concentration of xenon depresses NMDA and AMPA receptor–mediated synaptic transmission in the PFC and the SG, while not affecting GABA-mediated synaptic transmission. Our data provide evidence that the depression might be mainly due to postsynaptic mechanisms. According to the primary functions of the investigated CNS areas, the observed depression of excitatory synaptic transmission might account for the hypnotic and analgesic properties of the general anesthetic xenon.

The authors thank Kurt Biedermann, Ph.D. (Chemist, Department of Analytics, RCC Ltd., Itingen, Switzerland), for concentration measurements of dissolved xenon and Alexander Yassouridis, Ph.D. (Biostatistician, Head of the Research Group Biostatistics, Max Planck Institute of Psychiatry, Munich, Germany), for statistical advice.