Anesthesia and Surgery Induce a Functional Decrease Selectively in Excitatory Synaptic Transmission in Mouse Prefrontal Cortex Neurons

Postoperative delirium (POD), a syndrome of confusion and inattention, frequently occurs after anesthesia and surgery. However, the neuropathogenesis of POD remains mostly unknown. The prefrontal cortex (PFC) plays an essential role in cognitive processes. We therefore investigated how anesthesia and surgery induce neurofunctional changes in the PFC, and assessed whether intraoperative administration of dexmedetomidine, an alpha-2 agonist, could prevent the functional changes in the PFC. Laparotomy was performed in mice under isourane anesthesia. After a battery of behavioral tests measuring natural behaviors and learning and anxiety levels, PFC neuronal activities were recorded using whole-cell patch-clamp recordings. Effects of intraoperative dexmedetomidine were also examined. In the anesthesia/surgery group, the frequency of excitatory synaptic responses in PFC pyramidal neurons was decreased after the surgery without any changes in the response kinetics. On the other hand, neuronal intrinsic properties and inhibitory synaptic responses were not changed. Intraoperative dexmedetomidine or glutamate receptor antagonists, prevented the excitatory synaptic dysfunction induced by anesthesia and surgery. These ndings suggest that anesthesia and surgery induce functional reductions selectively in excitatory synaptic responses in PFC pyramidal neurons, and intraoperative dexmedetomidine inhibits the plastic change in the PFC synaptic transmission. excitatory synaptic PFC pyramidal neurons the synaptic response kinetics. We further compared inhibitory synaptic input to PFC pyramidal neurons by assessing spontaneous inhibitory postsynaptic currents (IPSCs). Spontaneous IPSCs elicited in pyramidal neurons in the PFC were not different between the control and anesthesia/surgery groups. No signicant differences were detected in the frequency (control, 7.3 ± 2.6 Hz; n = 10; anesthesia/surgery, 8.9 ± 3.7 Hz; n = 11; p = 0.29) and amplitude (control, 52.5 ± 12.0 pA; n = 10; anesthesia/surgery, 46.1 ± 26.5 pA; n = 11; p = 0.51) of spontaneous IPSCs (Fig. 3). The rise time (control, 2.1 ± 0.4 ms; n = 10; anesthesia/surgery, 2.6 ± 0.6 ms; n = 11; p = 0.08) and decay time (control, 4.5 ± 0.5 ms; n = 10; anesthesia/surgery, 5.4 ± 1.5 ms; n = 11; p = 0.10) of spontaneous IPSCs also did not change. These results suggest that the surgery does not change inhibitory synaptic inputs to PFC pyramidal neurons.


Introduction
Perioperative neurocognitive disorders, including delirium (postoperative), have become the most common complications after anesthesia and surgery. 1 Postoperative delirium (POD), a syndrome of confusion and inattention, occurs in up to 65% of elderly patients postoperatively. 2 POD is associated with increased morbidity and mortality, prolonged hospitalization, poor functional recovery, decreased life independence, and long-term decline in cognitive function. 3,4 However, the pathogenesis of POD is not still fully understood. Although various pathophysiological mechanisms for POD, such as aberrant oxidative cellular metabolism, sympathetic nervous-system stress, perioperative neuroin ammation, and acceleration of Alzheimer's disease, have been proposed, 5 the mechanisms underlying the neurofunctional changes occurring in POD remain mostly unknown. Since POD presents with cognitive disorders, it is thought to involve functional changes in the CNS such as neurodegeneration and synaptic dysfunction.
The prefrontal cortex (PFC) plays an essential role in cognitive processes such as attention, decisionmaking, and behavioral exibility. 6,7 Although the PFC has been reported to be involved in POD, 8 the postoperative changes in neuronal functions in the PFC require elucidation. Therefore, the primary aim of the current study was to investigate whether anesthesia and surgery induce functional changes in PFC neuronal activities. After a battery of behavior tests to detect delirium-like behavior changes, 9 we assessed ring properties and subthreshold synaptic transmissions in the PFC using patch-clamp recordings from acute brain slices, and found that the excitatory synaptic transmission was decreased following anesthesia and surgery. Then, based on the fact that there is no approved effective medication to prevent or treat POD, our secondary aim was to examine whether intraoperative administration of dexmedetomidine could prevent such functional synaptic decrease in the PFC following anesthesia and surgery. Dexmedetomidine is a highly selective α2 adrenoreceptor agonist that provides anxiolytic, sedative, and modest analgesic effects with minimal respiratory depression, and is widely used in clinical practice. 10 Although in both preclinical and clinical studies, administration of dexmedetomidine "after" surgery has been reported potentially to prevent occurrence of POD, it is still not approved for use "during" general anesthesia. 11,12 The present ndings may facilitate identi cation of new underlying mechanisms and preventive measures for POD during anesthesia.

Results
Anesthesia/surgery did not change the ring pattern of PFC pyramidal neurons postoperatively To investigate whether the neuronal activities of PFC pyramidal neurons were altered after anesthesia and surgery, we performed in vitro whole-cell patch-clamp recordings from PFC neurons in brain slice preparations. PFC neurons elicited action potentials in response to 500-ms depolarizing current injections through the recording electrode (Fig. 1A). There were no signi cant differences in the frequency-current relationship between the control and anesthesia/surgery groups (Fig. 1B). The threshold (control, -38.5 ± 2.3 mV, n = 10; anesthesia/surgery, -38.5 ± 2.0 mV, n = 10; p = 0.98) and the amplitude (control, 77.7 ± 1.8 mV, n = 10; anesthesia/surgery, 78.9 ± 7.2 mV, n = 10; p = 0.71) of the action potentials were not different between the two groups. Resting membrane potentials were not also different (control, -71.4 ± 5.6 mV, n = 10; anesthesia/surgery, -66.8 ± 4.6 mV, n = 10; p = 0.08). These results suggest that anesthesia and surgery do not change the neuronal intrinsic activities of PFC pyramidal neurons.

Dexmedetomidine partially alleviated the behavioral impairments induced by Anesthesia/surgery
We examined behavioral changes in buried food, open eld, and Y maze tests at 24 h before (baseline), and 6, 9, and 24 h after iso urane 2 h anesthesia (Fig. 4A). In mice that did not receive the anesthesia and surgery (control group), the relative values of the latency to nd the buried food 6 h after the anesthesia to that of baseline did not change (112.1 ± 81.8%, n = 18). In mice given a laparotomy during the same period of anesthesia (anesthesia/surgery group), the relative latency signi cantly increased 6 h after the anesthesia (322.1 ± 408.2%; n = 19, p = 0.04 versus the control group, see left graph in Fig. 4B). The anesthesia/surgery group still showed a similar tendency in the relative latencies 9 and 24 h after the anesthesia, but there were no signi cant differences between the groups. In mice that underwent a laparotomy under anesthesia with dexmedetomidine at 5 µg/kg (anesthesia/surgery (dexmedetomidine) group, n = 18), the relative latencies to nd buried food 6-24 h after the anesthesia were almost similar to those in the control group, and there were no signi cant differences between the control and anesthesia/surgery (dexmedetomidine) groups (Fig. 4B). However, there was a signi cant difference between the latencies 9 h after the anesthesia in the anesthesia/surgery and anesthesia/surgery (dexmedetomidine) group (p = 0.006, see middle graph in Fig. 4B). In the open eld test, the total distances travelled in a square chamber were not different between the three groups of mice at 6 h (control, 74.6 ± 19.4%, n = 18; anesthesia/surgery, 51.1 ± 19.6%, n = 19; and anesthesia/surgery (dexmedetomidine), 53.9 ± 19.3%, n = 18; p > 0.05), 9 h (control, 49.1 ± 26.0%, n = 18; anesthesia/surgery, 43.9 ± 32.0%, n = 19; anesthesia/surgery (dexmedetomidine), 36.1 ± 15.2%, n = 18; p > 0.05), and 24 h (control, 52.7 ± 21.5%, n = 18, anesthesia/surgery, 45.2 ± 21.4%, n = 19; anesthesia/surgery (dexmedetomidine), 43.8 ± 15.2%, n = 18; p > 0.05) after the anesthesia (Fig. 4C). However, the time spent in the center was different between anesthesia/surgery and control groups 6 h after the anesthesia (control, 70.4 ± 49.1%, n = 18; anesthesia/surgery, 39.5 ± 33.4%, n = 19; p = 0.03) (Fig. 4C). No signi cant differences were observed in the freezing time and latency to the center among the three groups. In the Y maze test, no signi cant differences were seen among the three groups except between the control and anesthesia/surgery (dexmedetomidine) groups 24 h after the anesthesia. These results obtained from the battery of behavioral tests suggest that general anesthesia during a laparotomy impaired the natural (buried food test and open eld test) behaviors of mice in an acute and uctuating manner, but intraoperative administration of dexmedetomidine did not change the natural behaviors.
Intraoperative administration of dexmedetomidine or ionotropic glutamate receptor antagonists did not change the frequency of excitatory synaptic transmission in PFC pyramidal neurons Lastly, we examined whether dexmedetomidine or the glutamate receptor antagonists MK-801 and CP 465022, could prevent the plastic changes in excitatory synaptic transmission in the PFC after the surgery, and compared the frequency of spontaneous EPSCs among the two groups and the anesthesia/surgery group. As shown in Fig. 5A and B, the frequencies of spontaneous EPSCs in both anesthesia/surgery (5 mg/kg dexmedetomidine) and anesthesia/surgery (0.5 mg/kg MK-801 and 10 mg/kg CP 465022) groups were signi cantly greater than that in the anesthesia/surgery group

Discussion
Here we revealed that a laparotomy under general anesthesia induced a decrease in glutamatergic synaptic inputs to pyramidal neurons of the PFC without changing inhibitory synaptic transmission. Administration of dexmedetomidine or glutamate receptor antagonists during the surgery prevented the plastic change in PFC excitatory synaptic transmission.
We rst compared neuronal activities in the PFC between the control and anesthesia/surgery groups.
Neuronal electrical activity is primarily dependent on three components: (1) transmitter release from the presynaptic neuron; (2) synaptic responses in the postsynaptic neurons; and (3) ring properties of the postsynaptic neurons. Our electrophysiological results showed that the frequency of spontaneous EPSCs, but not IPSCs, in the PFC decreased in the anesthesia/surgery group, suggesting that presynaptic glutamatergic transmitter release to PFC neurons is selectively weakened. However, the amplitude, rise time and decay time of spontaneous EPSCs did not change. These ndings suggest that ionotropic glutamate receptors (α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid receptors) expressed in the postsynaptic membrane of the PFC neurons were not altered postoperatively. The ring properties (frequency-current relationship, action potential threshold, and amplitude) of the PFC neurons in addition to resting membrane potentials did not change between the control and anesthesia/surgery groups. Thus, anesthesia/surgery could not impair the postsynaptic neurotransmitter receptor activities, and ring properties, but it modi ed the excitatory presynaptic inputs to the PFC pyramidal neurons.
Next, we con rmed whether a laparotomy under iso urane anesthesia induced the behavioral changes associated with delirium. In clinical practice, the confusion assessment method (CAM) is widely used to diagnose the presence of delirium in patients. 13 The CAM consists of the following four clinical features; acute onset and uctuating course, inattention, disorganized thinking, and altered level of consciousness. Although it is di cult to detect delirium-like behaviors in animals, Peng et al. developed a battery of behavioral tests in female mice to assess delirium from the viewpoint of CAM. 9 They reported that an extended latency to nd buried food and a reduction in the time spent in the center in the open eld test for mice were regarded as the behavioral changes associated with delirium because such parameters depend on the presence and intactness of attention, organized thinking, and consciousness. As reported, in this study anesthesia/surgery also extended the latency to nd buried food and decreased the time spent in the center of the open eld test, although it did not cause behavioral changes in the Y maze test.
The PFC plays important roles in the functions associated with the symptoms of delirium, including regulation of attention 14 and organizing thought. 15,16 PFC pyramidal neurons are reported to regulate attention and behavior. 17 Patients with attention de cit hyperactivity disorder (ADHD), whose main symptom is inattention, also exhibited reduced activity of the PFC. 18 Therefore, behavioral changes such as the delirium-like attention de cits detected in our study may be explained by the synaptic dysfunction (i.e., reduction in the frequency of excitatory synaptic transmission) of PFC pyramidal neurons.
Interestingly, dexmedetomidine partially decreased the buried food latency in the anesthesia/surgery group, and we were not able to detect any differences between the anesthesia/surgery (dexmedetomidine) and control groups in the behavioral tests. These ndings suggest that administration of dexmedetomidine during the surgery may inhibit the postoperative changes in behaviors associated with attention, thinking, consciousness, and anxiety.
Then, we investigated whether intraoperative dexmedetomidine could prevent such plastic changes in excitatory synaptic transmission in the PFC after the surgery. Moreover, we also studied whether intraoperative administration of MK-801 and CP 465022, which are antagonists for ionotropic glutamate receptors (N-methyl-d-aspartate and α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid) could prevent the synaptic dysfunction in the PFC. In the PFC, pyramidal cells are the most abundant neurons (estimated to represent ~ 80% or more of the total neuronal population), 19 and abnormal activations of pyramidal neurons cause excessive glutamate release to induce neurofunctional damage in the CNS. 20,21 Therefore, we hypothesized that an approach speci cally targeting inhibition of abnormal activations in neurons should be effective on preventing the synaptic dysfunction after surgery. The abnormal activations mediated by these glutamate receptors is the main pathological process in several neurological diseases such as stroke, epilepsy, and neurodegenerative diseases. Qian et al. showed that systemic administration of MK-801 (0.5 mg/kg) prevented the brain damage caused by excessive release of glutamate and N-methyl-d-aspartate receptor-related excitotoxicity in stroke. 22 Menniti et al. also showed that systemic administration of CP 465022 (10 mg/kg) potently and effectively inhibited αamino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor-mediated synaptic transmission in the brain. 23 Moreover, Arias et al. reported that a combination treatment of N-methyl-d-aspartate and αamino-3-hydroxy-5-methylisoxazole-4-propionic acid receptor noncompetitive antagonists produced a synergistic neuroprotective effect in the hippocampus. 24 In current clinical practice, a level of antinociception during general anesthesia is usually thought to be su cient when clinical responses, such as body movement and heart rate or blood pressure elevations, no longer occur during surgical stimulation. Opioids are usually used for general anesthesia in conjunction with sedatives to achieve the su cient level of general anesthesia. However, recent studies have reported that the neuronal activations during surgery in the human brain, particularly in the frontal cortex, cannot be suppressed despite the absence of such clinical responses while propofol and remifentanil are administered together. 25 Therefore, some supplemental way to inhibit the abnormal activations in the brain could be required to prevent synaptic dysfunction. Actually, our preliminary experiments using in vivo extracellular recordings, showed that neuronal activity in the rat PFC was increased by surgery-like stimuli even under anesthesia condition, and dexmedetomidine were able to suppress its abnormal neuronal activations in the PFC (data not shown). We also found that intraoperative administration of dexmedetomidine or MK-801 and CP 465022 prevented the decrease in glutamate-mediated excitatory synaptic transmission in pyramidal neurons of the PFC. In both preclinical research and clinical studies, administration of dexmedetomidine "after" surgery has been reported to potentially prevent the occurrence of POD. 11,12,26 Animal studies demonstrated that the mechanism for reduction of cognitive dysfunction was the anti-neuroin ammatory effect of dexmedetomidine. [27][28][29] In addition, because dexmedetomidine exerts its effects at both spinal and supraspinal sites to modulate nociceptive input and transmission, and provides analgesia, 10 it could suppress the neuronal activities elicited by surgery before reaching its nociceptive input to the PFC. Taken together, our study suggests that sequential dexmedetomidine treatment from "during" to "after" surgery may be an effective option for prevention of POD. Moreover, inhibition of glutamate synaptic transmission elicited during surgery might be a novel approach for prevention of POD.
In summary, we found that glutamate-mediated excitatory synaptic transmission in PFC pyramidal neurons decreased "after" anesthesia/surgery, and administration of dexmedetomidine "during " surgery could prevent the synaptic change of PFC neurons.

Animals
Female animals are known to exhibit less anxiety than male animals in behavioral assessments performed in an open space. 30 We rst performed open eld test and compared time spent in the center of the open eld test (see below) between C57BL/6 male and female mice. In male mice, the time spent in the center of the 2nd trial was decreased to less than half of it (48 ± 12.5 % of the rst trail, n = 6). That of female mice was not largely changed (74.6% of the rst trial, see Results), suggesting that anxiety in male mice is not stable even without anesthesia and surgery. To detect behavioral changes including anxiety following anesthesia and surgery, a recent study developed a battery of behavioral analyses using female animals. 9 In this study, we therefore used female animals (C57BL/6 mice (SLC, Hamamatsu, Japan)) to detect behavioral and neuronal changes following anesthesia and surgery. The animals were housed in cages with food and water available ad libitum. The room was maintained at 20°C with a 12-h light/dark cycle. All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee of Hyogo College of Medicine, and were performed in accordance with the institutional guidelines for animal experiments and carried out in accordance with the ARRIVE guidelines 2.0. Every effort was made to reduce the number of animals used.

Behavioral tests and electrophysiological recordings from mice after anesthesia and surgery
Mice were randomly divided into three groups: control, Anesthesia/surgery, and anesthesia/surgery (dexmedetomidine). Mice in the anesthesia/surgery and anesthesia/surgery (dexmedetomidine) groups received a laparotomy under general anesthesia (see below). In the two groups, a battery of behavior tests was conducted 6, 9, and 24 h after surgery. The same battery of behavioral tests was also performed 24 h before the surgery to obtain baseline behavioral data. In the control group, mice were placed in their home cages, and the battery of behavioral tests was administered at the same four time points as in other two groups. We performed in vitro whole-cell patch-clamp recordings for PFC neurons in acute brain slice preparations from the mice of all three groups within 14 days after anesthesia and surgery.

Anesthesia and surgery
Laparotomy was performed in 8-10-week-old mice under iso urane anesthesia, similar to previous studies. 9,31 Anesthesia was induced and maintained using 1.5% iso urane in 100% O 2 in a transparent acrylic box for 15 min, after which anesthesia was maintained via a cone-shaped mask over the nose. A longitudinal incision was made from the xiphoid to pubic symphysis (0.5 cm) on the skin, abdominal muscles, and peritoneum, after which the incision was sutured with 5-0 vicryl thread. The procedure required approximately 10 min. The mouse was then transferred back into the box, and was under anesthesia for 2 h in total. EMLA cream® (2.5% lidocaine and 2.5% propitocaine, Astra Pharmaceuticals) was applied onto the incision at the end of the procedure and at 8-h intervals for one day.

Buried food test
The buried food test was performed as described in previous studies with modi cations. 32,33 We used ten 45-mg pieces of a puri ed pellet of F-0021J (BrainScience idea Co., Ltd.). The mice were habituated to the testing room for 1 h. The test cages (30 × 5 × 15 cm) were prepared with clean bedding (height, 3 cm) and a round-shaped case (diameter, 2 cm) containing the pellets. The case was buried in the bedding (0.5 cm below to the surface), and the location of the case in the test cage was changed at random. The mouse was placed in the center of the test cage, and the time to start eating was measured. If the mouse did not nd the pellet in 5 min, we ended the trial and considered 300 s as the latency period. We used a new test cage, bedding, and gloves for each mouse.

Open eld test
Open eld test was performed as described in previous studies with modi cations. 9,34 The open eld apparatus was a white plexiglass square box (40 × 40 × 40 cm). The mouse was gently placed at the center of the box under dim lighting and was allowed to move freely for 5 min. The mouse in the box was monitored by a video camera, and the behavioral trajectory was analyzed with Any-maze Y maze test Y maze test was performed as described in previous studies, with modi cations. 9,35,36 The Y maze apparatus was made of black-colored plexiglass. The apparatus consisted of three arms (30 × 5 × 15 cm) with an angle of 120° between each arm. The three arms were the start arm, in which the mouse starts to explore (always open); the novel arm, which is blocked at the rst trial, but opened at the second trial; and the other arm (always open). The start arm and the other arm were randomly designed to avoid the spatial memory error. The Y maze test consisted of 2 trials separated by an inter-trial interval (ITI). The rst trial (training) was performed for 10 min and allowed the mouse to explore 2 arms (start arm and other arm) of the maze, with the novel arm being blocked. After an ITI of 1 h, the second trial (retention) was performed. For the second trial, the mouse was placed back in the maze in the same starting arm, with free access to all three arms for 5 min. The maze was placed in a quiet and illustrated room. The behavior was recorded by a video camera mounted above the maze, and the number of entries and the time spent in each arm were assessed with the Any-maze behavior tracking system software. The time spent in and entries into the novel arm were used as an indication of spatial recognition memory. Each of the arms of the Y maze was cleaned with 70% ethanol solution before testing.
In vitro whole-cell patch-clamp recording Mice that underwent the behavioral tests were rapidly decapitated after cervical dislocation, and the brain was quickly removed and submerged in a cold arti cial cerebrospinal uid (ACSF) solution containing (in mM) 124 NaCl, 2.5 KCl, 2 CaCl 2 , 1 MgSO 4 , 25 NaHCO 3 , 1 NaH 2 PO 4 , and 10 glucose equilibrated with 95% O 2 -5% CO 2 . Coronal brain slices (300 µm) at the level of the PFC were prepared using a vibratome. [37][38][39] The brain slices were transferred to a submerged recovery chamber with the ACSF solution at room temperature for at least 1 h. Individual slices were then put into a recording chamber in which the ACSF solution was continuously perfused at 32-33 ℃. Whole-cell patch-clamp recordings were made from single layer / pyramidal neurons visualized with infrared differential interference contrast optics (BX50WI, Olympus, Tokyo, Japan) 39  were recorded under current-clamp conditions. APs were evoked by 500-ms depolarizing pulses from 0 to 150 pA with 30-pA steps through the recording electrode from a membrane potential of -70 mV. The access resistance (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) was monitored throughout the experiment. Data were discarded if the resistance changed by more than 20% during the recording. The recording signals were ampli ed with a patch-clamp ampli er (MultiClamp 700A; Molecular Devices, Sunnyvale, CA, USA), low-pass ltered at 5 kHz, digitized with an analog-to-digital converter (Digidata 1321A; Molecular Devices), and stored on a personal computer using a data acquisition program (pCLAMP version 12.3; Molecular Devices) with a sampling frequency of 10-20 kHz. Synaptic events were analyzed using a software (Minianalysis version 6.0.7; Synaptosoft, Fort Lee, NJ, USA). Only cells exhibiting a stable resting membrane potential (less than -50 mV) were used in this study. The following criteria were used to measure action potential threshold and amplitude. The rst action potential evoked by the minimum injection current was used for all measurements. Action potential threshold is the voltage where a sharp upward rise of the depolarizing phase of the action potential occurs, while action potential amplitude is the difference in voltage between the threshold and its peak amplitude Drugs Dexmedetomidine was obtained from Maruishi Pharmaceutical Co., LTD. (Osaka, Japan) and was dissolved in saline (1 µg/mL). MK-801 and CP 465022 were obtained from Tocris Bioscience (Bristol,

Statistical analysis
All numerical data are shown as mean ± SD. Statistical analyses were performed with EZR 40 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). More precisely, it is a modi ed version of R commander designed to perform statistical functions frequently used in biostatistics. Student's unpaired t tests was used to compare the ring properties, the parameters of EPSCs and IPSCs between control and anesthesia/surgery group. Student's paired t tests was used to compare the ring frequency between before and after stimuli. Two-way repeated analysis of variance (ANOVA) followed by post hoc Bonferroni test was used to compare the frequency-current relationship of action potentials and the cumulative distributions of inter-event interval of spontaneous EPSCs. The Kruskal-Wallis test followed by post hoc Steel-Dwass was used to analyze the behavioral test results of the three groups (control, anesthesia/surgery, anesthesia/surgery (dexmedetomidine)) and the frequency of spontaneous EPSCs for the three groups (anesthesia/surgery, anesthesia/surgery (MK-801 and CP465022), anesthesia/surgery (dexmedetomidine)). In all cases, p < 0.05 was considered to be statically signi cant. "n" refers to the number of animals or neurons. Figure 1 Anesthesia and surgery did not change the ring properties of pyramidal neurons in the PFC (A)

Declarations
Examples of ring properties of PFC pyramidal neurons in control and anesthesia/surgery groups. In response to current injections applied through the recording pipette, pyramidal neurons in both groups showed a tonic ring. (B) Relationships between the frequency of action potentials and amplitudes of the current injected. There were no signi cant differences in the frequency-current relationships between control and anesthesia/surgery groups (control, n = 10; anesthesia/surgery, n = 10; p = 0.28).  latency obtained from the same mice 24 h before anesthesia and surgery. The relative latency 6 h after anesthesia and surgery was higher in the anesthesia/surgery group than in the control group. There were no differences between anesthesia/surgery (dexmedetomidine) (5 µg/kg) and control groups, and the latency 9 h after anesthesia and surgery was lower in the anesthesia/surgery (dexmedetomidine) group in comparison with the anesthesia/surgery group. Control group, n = 18; anesthesia/surgery group, n = 19; anesthesia/surgery (dexmedetomidine) group, n = 18; * p < 0.05, ** p < 0.01. (C) Summary showing relative values of the total distance of the open eld test after anesthesia and surgery to the baseline total distance obtained from the same mice 24 h before anesthesia and surgery (upper three graphs). Anesthesia/surgery group did not show differences in the total distance in the open eld in comparison with the control group at 6, 9, and 24 h after anesthesia and surgery. Lower three graphs showing that anesthesia/surgery group showed lower relative values of time spent in the center of the open eld in compared with the control group at 6, but not at 9 or 24 h after anesthesia and surgery. Control group, n = 18; anesthesia/surgery group, n = 19; anesthesia/surgery (dexmedetomidine) group, n = 18, * p < 0.05.