Glutamatergic neurons of piriform cortex delay induction of inhalational general anesthesia

Abstract

Since their clinical application in the 1840s, the greatest mystery surrounding general anesthesia (GA) is how different kinds of general anesthetics cause reversible unconsciousness, and the precise neural mechanisms underlying the processes. Over past years, although many studies revealed the roles of cortex, thalamus, brainstem, especially the sleep-wake circuits in GA-induced loss of consciousness (LOC),the full picture of the neural circuit mechanism of GA is still largely unknown. Recent studies have focused on the importance of other brain regions. Here, we report that the activity of glutamatergic (Glu) neurons in the piriform cortex (PC), a critical brain region for odor encoding, began to increase during the LOC of GA and gradually recovered after recovery of consciousness. Chemical lesions of the anterior PC (APC) neurons accelerated the induction time of isoflurane anesthesia. Chemogenetic and optogenetic activation of APC^Glu neurons prolonged isoflurane and sevoflurane anesthesia induction, whereas APC^Glu neuron inhibition displayed the opposite effects. Moreover, the modification of APC^Glu neurons did not affect the induction or emergence time of propofol GA. In addition, odor processing may be partially involved in the induction of isoflurane and sevoflurane GA regulated by APC^Glu neurons. In conclusion, our findings reveal a critical role of APC^Glu neurons in inhalational GA induction.

Graphical abstract

1. Introduction

Since its application in the clinic in 1846, general anesthesia (GA) has enabled patients to undergo thorough surgery unconsciously [1], [2], [3]. General anesthetics cause reversible loss of consciousness (LOC), which is similar to the neural characteristics of non-rapid eye movement sleep [4], [5], [6]. Recently, a growing number of studies have shown that LOC induced by general anesthesia shares parts of the neural networks with natural sleep [7], [8], [9]. Nonetheless, sleep-wake circuits also show different functions in GA, and the neural mechanism of GA-induced consciousness change remains unclear [10,11]. The most popular general anesthetics used in clinical practice, such as isoflurane, sevoflurane, and propofol, can cause reversible unconsciousness. However, they have completely different brain structures, characteristics, and receptors. Whether these anesthetics shared a common neural mechanism in consciousness regulation or not remains a mystery.

The piriform cortex (PC), the largest cortical region that receives direct synaptic input from the olfactory bulb (OB), has been identified as a key brain region for encoding odorous objects [12,13]. The PC is structurally divided into the anterior and posterior piriform cortices (APC and PPC, respectively) and is mainly composed of glutamatergic (Glu) neurons [12,14]. The chemical identity of an odor stimulus is encoded in the APC, whereas the integrated perceptual representation of an odor object is encoded in the PPC [13,15]. Besides odor encoding, the PC is also involved in the regulation of learning and memory [16,17], emotion [18], seizure [19], [20], [21] and sleep [17,[22], [23], [24], [25], [26]. A recent study showed that PC-mPFC synaptic plasticity is a neural basis of social communication [27], which suggested a higher-level function in cognition of the PC. Furthermore, the PC encodes odor identity through the regulation of gamma/beta and theta/delta oscillations [15,28]. Interestingly, PC neurons are activated by isoflurane inhalation [29]. Together with the regulation of theta/delta oscillation, it is suggested that PC neurons may participate in the progress of general anesthetic-induced reversible unconsciousness, but their roles and possible mechanisms remain unclear.

In the present study, we investigated how glutamatergic neurons in the PC regulate GA induced by isoflurane, sevoflurane, and propofol. c-Fos staining was used to monitor the activity of PC neurons during different stages of GA. We further lesioned the PC neurons to clarify the function of the PC in anesthesia. Subsequently, chemogenetic or optogenetic manipulation of PC glutamatergic neurons was performed during isoflurane/sevoflurane/propofol-induced GA. Finally, we investigated whether the odor signal mediated by PC is involved in GA regulation. Together, our findings show a critical role of PC glutamatergic neurons in the modulation of isoflurane- and sevoflurane-mediated unconsciousness, but not propofol-induced anesthesia.

2. Materials and methods

2.1. Animals

This study was performed in accordance with the Guide for the Care and Use of Laboratory Animals of China (No. 14,924, 2001). All animal procedures were designed and conducted according to the guidelines set forth by the Zunyi Medical University. Male C57BL/6 J mice were used for all tests and purchased from Changsha Tianqin Technology Co., Ltd. (Changsha, China).Mice were maintained at the Experimental Animal Center of Zunyi Medical University (License Key. SYXK(Qian) 2021–0004). Mice were kept under temperature-controlled conditions on a 12:12 h light/dark cycle (lights on at 6:00 am) with food and water ad libitum. To avoid possible circadian influence on the experimental results, all behavioral and electroencephalogram (EEG) tests were developed between 18:00 and 24:00 p.m. All animal procedures were complied with the ARRIVE guidelines and were designed and conducted in accordance with the National Institutes of Health guide for the care and use of Laboratory animals (NIH Publications No. 8023, revised 1978) and the guidelines set forth by Zunyi medical university.

2.2. Antibodies and reagents

Isoflurane and Sevoflurane were provided by RWD Life Sciences (Shenzhen, China). Pentobarbital and lidocaine were purchased from Chaohui Pharmaceutical Co., Ltd. (Shanghai, China). Ibotenic acid (IBO) was purchased from MedChemExpress LLC (USA, HY-N2311), and Clozapine N-oxide (CNO) was obtained from Sigma-Aldrich (United States, C0832). Antibodies against c-Fos (226 008, RRID:AB_2891278) were purchased from Synaptic Systems. Antibodies against NeuN (MAB377; RRID:AB_2298772) were obtained from Millipore. Moreover, the Parvalbumin antibody (PV27, RRID:AB_2631173) was purchased from Swant. Alexa Fluor-conjugated secondary antibodies were purchased from Invitrogen (A-11039, RRID:AB_142924; A-11008, RRID:AB_143165; and A-11005, RRID:AB_141372).

2.3. Stereotaxic surgery

The mice were anesthetized with 1.4% isoflurane with oxygen (O₂) at 1 L/min and then placed on a stereotaxic apparatus (RWD Life Science, Shenzhen, China). Lidocaine (1%) was subcutaneously injected for local analgesia before the skull surface was exposed.

For lesion experiments, IBO (200 nl/side) of and an equal volume of saline as control were bilaterally injected (speed:20 nl/min) into APC region ([AP]: +0.40 mm, [ML]: ±3.00 mm, [DV]: −4.50 mm) and PPC region ([AP]: −1.00 m, [ML]: ±3.50 mm, [DV]: −4.90 mm), respectively, through a glass micropipette (1-mmglass stock, tapering to a 10–20-micron tip) using amicro-syringe pump. The pipette was retained at the injection site for 10 min to allow the IBO to diffuse and was then slowly withdrawn. The mice underwent further behavioral tests after 7 days.

For chemogenetic tests, 200 nl virus mixed with rAAV-CamKII-cre and rAAV-EF1α-DIO-hM3Dq-eYFP (hM3Dq), rAAV-EF1α-DIO-hM4Di-eYFP (hM4Di), or rAAV-EF1α-DIO-eYFP (Control) (1:5, Brain-VTA, Wuhan, China) were bilaterally injected into the APC ([AP]: +0.40 mm, [ML]: ±3.00 mm, [DV]: −4.50 mm, 20 nl/min). The micropipette was held in place for 10 min, followed by slow retrieval. After virus injection, the EEG electrodes were placed on the skull (AP: +1.0 mm, ML: ±1.5 mm; AP: −3.5 mm, ML: ±1 mm) [9,30], and firmed with skull screws and dental cement.

For optogenetic activation experiments, 200 nl virus mixed with rAAV-CamKII-cre and rAAV-EF1α-DIO-hChR2-eYFP (ChR2) or rAAV-EF1α-DIO-eYFP (control)(1:5, Brain-VTA, Wuhan, China) were unilaterally injected into the APC ([AP]: +0.40 mm, [ML]: +3.00 mm, [DV]: −4.50 mm, 20 nl/min), followed by the placement of an optical fiber (Inper Ltd, Hangzhou, China) into the same side of APC. For optogenetic inhibition tests, virus mixed with rAAV-CamKII-cre and rAAV-EF1α-DIO—NpHR-eYFP (NpHR) or rAAV-EF1α-DIO-eYFP (control) (1:5) were bilaterally injected into the APC on both sides ([AP]: +0.40 mm, [ML]: ±3.00 mm, [DV]: −4.50 mm, 20 nl/min), followed by the placement of optical fibers into the APC and the subsequent placement of EEG electrodes on the skull (AP: +1.0 mm, ML: ±1.5 mm; AP: −3.5 mm, ML: ±1 mm). All tests were performed after three weeks.

2.4. Behavioral tests

Loss of righting reflex (LORR) and recovery of righting reflex (RORR) times are considered standard indices of GA induction and emergence in mice, respectively.

For isoflurane or sevoflurane GA, mice were placed in a recording chamber (RWD Company, Shenzhen, China) and incubated with 1.4% isoflurane or 2.4% sevoflurane with oxygen at a rate of 1 L/min. The interval from the starting point of isoflurane or sevoflurane application to the time point at which mice lost the righting reflex was regarded as the latency to LORR. The mice were then anesthetized for 30 min and gently removed from the chamber into room air. RORR was then quantified in the supine position. The duration from the end of isoflurane or sevoflurane infusion to RORR was defined as RORR time. For propofol anesthesia, we used two methods to measure the LORR and RORR. For LORR, we pumped propofol through the caudal vein at a rate of 10 mg/kg/min into the mouse, and recorded the time that the mice appeared LORR as the latency to LORR. For RORR, a single dose of 20 mg/kg propofol was administered through the caudal vein. The duration of anesthesia was defined as the RORR time.

In the lesion experiments, we recorded the LORR and RORR under isoflurane and propofol anesthesia (Fig. S1a). In the chemogenetic groups, Clozapine N-oxide (CNO) (1 mg/ml, 1 mg/kg, i.p.) or saline (0.9%, equal volume, i.p.) was injected 1 h before the behavioral tests and EEG recording. The LORR, RORR, and EEG were recorded under isoflurane, sevoflurane, and propofol application (Fig. S1b). For optogenetic experiments, a blue laser (473 nm, 1 Hz, 4 ms)[27] or yellow laser (594 nm, 1 Hz, 4 ms) stimulation was delivered to the animals during induction and emergence (Fig. S1c). For induction, opto-stimulation was administered with isoflurane/sevoflurane inhalation or propofol pumping started and continued until the mice achieved LORR. For emergence, the mice were optically stimulated from the cessation of isoflurane/sevoflurane/propofol application until RORR. All mice were subjected to immunofluorescence analysis to verify the position and efficiency of viral transfection after all experiments were completed.

2.5. EEG recording

EEG was recorded using a multichannel signal acquisition system (Appolo, Bio-Signal Technologies, United States) and then digitized and analyzed using the Spike2 software package (Cambridge Electronic Design, Cambridge, United Kingdom). EEG signals were collected and filtered between 0.1 and 300 Hz. EEG signals were recorded for10 min before induction and were continuously recorded until recovery from isoflurane, sevoflurane, or propofol anesthesia. Relative powers in the different frequency bands were computed by averaging the signal power across the frequency range of each band (delta (δ): 1–4 Hz, theta (θ): 4–8 Hz, alpha (α): 8–12 Hz, beta (β): 12–25 Hz, and gamma (γ): 25–60 Hz) and then dividing by the total power from 1 to 60 Hz, as in previous studies [9,32]. The spectrogram was bandpass filtered from 0.1 to 30 Hz and constructed using NeuroExplorer (Nex Technologies, USA).

2.6. Immunohistochemistry

All mice were sacrificed after behavioral tests and EEG recordings. After fixation in 4% PFA and dehydration in 30% sucrose, the brain tissues were frozen in OCT and sectioned at 30 μm (Leica CM1900, Germany). Sections containing APC or PPC were blocked for 2 h at room temperature (RT) in blocking buffer (10% goat serum, 1% BSA, and 0.3% Triton in PBS). The sections were then incubated with primary antibodies overnight at 4 °C and secondary antibodies for 2 hrs at RT. The primary antibodies used were 1:1000 for c-Fos, 1:200 for NeuN, and 1:1000 for PV. The secondary antibodies were diluted to 1:1000. Sections were mounted using the ProLong Gold Antifade Reagent (Invitrogen). All images were captured using a virtual microscopy system (Olympus BX63) and a laser confocal fluorescence microscope (Olympus FV1000), followed by analysis using ImageJ software (National Institutes of Health).

2.7. Statistical analysis

GraphPadPrism 8.0 (GraphPad Software Inc., USA) was used for statistical analyses. Data were presented as the mean ± SD. For two-group comparison, unpaired two-sided Student's t-test was performed for statistical differences. One-way ANOVA followed by Tukey's post hoc test was used for multiple group comparisons. The accepted level of significance was p < 0.05 in all cases. ‘n’ refers the number of animals we tested in the corresponding experiments.

3. Results and discussion

3.1. PC neuron activity in response to GA

To investigate the activity variation of PC neurons during GA, we first monitored c-Fos expression as a neural activity indicator during the state of wakefulness, isoflurane (or propofol)-induced anesthesia, and recovery from anesthesia in the PC. Compared to those in the wakefulness groups, the c-Fos expression was dramatically increased during isoflurane anesthesia, while decreased c-Fos levels were observed in the recovery groups (Fig. 1a, b). This is consistent with the results of a previous report [29]. Concurrently, we found similar accumulated c-Fos expression in the propofol anesthesia groups compared to the wakefulness and recovery stages (Fig. 1c, d).

Fig. 1.

c-Fos expression in PC during isoflurane and propofol anesthesia. (a) Representative images of c-Fos expression in wake, isoflurane and recovery groups in PC (scale bar, 200 μm). (b) Quantification of c-Fos number (per mm²) in PC (wake: 342 ± 63; isoflurane: 1290 ± 136; recovery: 628 ± 87; n = 5).(c) Representative images of c-Fos expression in wake, propofol and recovery groups in PC (scale bar, 200 μm). (d) Quantification of c-Fos number (per mm²) in PC (wake: 366 ± 57; propofol: 960 ± 54; recovery: 562 ± 56; n = 5). (e) Representative images of c-Fos/vGluT2 staining in wake, isoflurane and recovery groups in PC (scale bar, 50 μm). (f) Representative images of c-Fos/VGAT staining in wake, isoflurane and recovery groups in PC (scale bar, 50 μm). *p < 0.05, **p < 0.01.

PC is mainly composed of glutamatergic neurons and a small number of GABAergic neurons [12,14]. We further investigated whether c-Fos variation occurred in both neurons during anesthesia. First, we injected the AAV9-vGluT2-eGFP virus into the PC to label glutamatergic neurons. We found that during isoflurane anesthesia, the increased c-Fos signal was mainly colocalized with the GFP labeling neurons (Fig. 1e). In contrast, we repeated isoflurane treatment in the VGAT-td mice and found no increase in c-Fos number in td-tomato-labeled GABAergic neurons in the PC (Fig. 1f). These data indicate that the activity of PC glutamatergic neurons accumulates during isoflurane anesthesia. These findings suggest the possible role of PC in the regulation of distinct stages of GA.

3.2. Lesion of PC neurons accelerate the induction of isoflurane/sevoflurane anesthesia

To explore the possible roles of PC in different stages of general anesthesia, we injected IBO into C57BL/6 J mice to selectively kill APC and PPC neurons. Seven days after IBO injection, massive neuronal loss was observed in both APC and PPC in the lesion group (Fig. 2a, b, e, and f). We then compared the LORR and RORR times in the lesion and control groups under different general anesthetic-induced anesthesia. We found a shorter LORR time in the APC lesion mice in isoflurane- and sevoflurane-mediated anesthesia (Fig. 2c, d), and a normal RORR time after the inhalation of these anesthetics (Fig. 2c, d). In contrast, the PPC lesion did not change both the LORR and RORR under isoflurane and sevoflurane anesthesia (Fig. 2g, h). These data suggest that APC lesions accelerated the induction of isoflurane and sevoflurane anesthesia.

Fig. 2.

Effects of chemical lesion of PC neurons on isoflurane, and sevoflurane anesthesia. (a) Representative images of NeuN staining in Ctrl and lesion groups in APC(scale bar, 50 μm). (b) Quantification of NeuN number (per mm²) in APC (Ctrl: 3840 ± 543; lesion: 1040 ± 320; n = 5).(c) The effect of APC lesion on LORR and RORR time under 1.4% isoflurane anesthesia (n = 7). (d) The effect of APC lesion on LORR and RORR time under 2.4% sevoflurane anesthesia (n = 7). (e) Representative images of NeuN staining in Ctrl and lesion groups in PPC (scale bar, 50 μm). (f) Quantification of NeuN number (per mm²) in PPC (Ctrl: 3280 ± 588; lesion: 960 ± 320; n = 5). (g) The effect of PPC lesion on LORR and RORR time under 1.4% isoflurane anesthesia (n = 7). (h) The effect of PPC lesion on LORR and RORR time under 2.4% sevoflurane anesthesia (n = 7). *p < 0.05, **p < 0.01.

Meanwhile, we used a single dose (20 mg/kg) of propofol to observe the effect of APC and PPC lesions on propofol-induced anesthesia. We found no significant difference in RORR time between the lesion and control groups (Fig. S2). A 20 mg/kg propofol injection caused an immediate coma in mice; therefore, we failed to evaluate the effect of APC and PPC lesions on the LORR of propofol anesthesia. Therefore, we set up a pumping procedure to slow down the unconscious process of propofol and monitor both the LORR and RORR of propofol-induced anesthesia [31].

These data suggest that only APC lesions affect the LORR of isoflurane/sevoflurane anesthesia (Fig. 2c, d, g, h). Together with the result that isoflurane only activates glutamatergic neurons in the PC, we will use chemogenetic and optogenetic tools to specifically manipulate the activity of APC glutamatergic neurons (APC^Glu) to determine its precise function under GA.

3.3. Chemogenetic manipulation of APC^Glu neurons controls inhalational GA induction

To modulate the activity of APC^Glu neurons, hM3Dq or hM4Di virus combined with CamKII-cre virus was microinjected into the APC of C57BL/6 J mice.

Immunofluorescence images confirmed viral transfection in the APC^Glu neurons (Fig. 3a). CNO was intraperitoneally injected 1 h before anesthetic application to activate or inhibit the APC^Glu neurons. In the chemogenetic activation experiments, accumulated c-Fos numbers were observed in APC^Glu neurons in hM3Dqmice after CNO treatment (Fig. 3b), which indicated the activation of APC^Glu neurons. For isoflurane anesthesia, APC^Glu neuron activation significantly prolonged the induction time (LORR) (Fig. 3c). However, the recovery time (RORR) was comparable between the hM3Dq-CNO group and the hM3Dq-saline or eGFP-CNO groups (Fig. 3c). EEG recordings were set up to further assess how APC^Glu neurons affect isoflurane anesthesia. In the hM3Dq-CNO group, APC^Glu neuron activation reduced the total power percentage of the δ wave (1–4 Hz) and increased the γ wave (25–60 Hz) proportion during LORR induction (Figs. 3d, e and S3a). We found no alteration in the EEG features during the RORR period in the hM3Dq-CNO group (Figs. 3d,e and S3b),which is consistent with the behavioral results. During sevoflurane anesthesia, prolonged LORR time and normal RORR time were also observed in the hM3Dq-CNO group compared to the hM3Dq-saline and eGFP-CNO groups (Fig. 3f). EEG recordings showed a lower δ wave (1–4 Hz) percentage and a higher percentage of γ wave (25–60 Hz) during LORR induction in the hM3Dq-CNO group (Figs. 3g, h and S3c). A decreased δ-wave ratio was also detected during the RORR period in the hM3Dq-CNO group (Figs. 3g, h, and S3d), which was different from that of isoflurane anesthesia. Finally, we examined the effect of APC^Glu neuronal activation on propofol-induced anesthesia. No significant changes in LORR and RORR times during propofol administration were observed in the hM3Dq-CNO group, and the EEG recordings were comparable between the hM3Dq-CNO group and the hM3Dq-saline group or the eGFP-CNO group (Figs. 3i-k, and S3e, f). These results indicate that activation of APC^Glu neurons slows down the induction of isoflurane and sevoflurane anesthesia, but not propofol.

Fig. 3.

Chemogenetic activation of APC^Glu neurons prolonged the induction time of isoflurane and sevoflurane anesthesia. (a) Left: schematic of virus injection and EEG electrode implantation. Right: representative image of APC^Glu expressing hM3Dq-eYFP(scale bar, 200 μm). (b) Representative images of c-Fos expression in hM3Dq groups after 2 hrs saline and CNO injection respectively (scale bar, 20 μm). (c) The effect of APC^Glu neuron activation by CNO on LORR and RORR time under 1.4% isoflurane anesthesia (n = 6). (d) The power distribution of EEG frequency bands in M3-NS or M3-CNO group under 1.4% isoflurane anesthesia (n = 6). (e) Representative EEG wave forms and spectrograms EEG power of M3-NS or M3-CNO group under 1.4% isoflurane anesthesia. (f) The effect of APC^Glu neuron activation on LORR and RORR time under 2.4% sevoflurane anesthesia (n = 6). (g) The power distribution of EEG frequency bands in M3-NS or M3-CNO group under 2.4% sevoflurane anesthesia (n = 6). (h) Representative EEG wave forms and spectrograms EEG power of M3-NS or M3-CNO group under 2.4% sevoflurane anesthesia. (i) The effect of APC^Glu neuron activation on LORR and RORR time under propofol anesthesia (n = 6). (j) The power distribution of EEG frequency bands in M3-NS or M3-CNO group under propofol anesthesia (n = 6). (k) Representative EEG wave forms and spectrograms EEG power of M3-NS or M3-CNO group under propofol anesthesia.*p< 0.05, **p< 0.01.

Next, we performed a chemogenetic inhibition experiment to study the effect of APC^Glu neuronal repression on GA. The eGFP image and decreased c-Fos number confirmed the virus transfection efficiency and inactivation of APC^Glu neurons in the hM4Dq-CNO group (Figs. 4a and S4a). Behavioral tests showed shorter LORR times for isoflurane and sevoflurane, but not for propofol-mediated anesthesia in the hM4Dq-CNO group (Fig. 4b-d). Similar to the hM3Dq-CNO results, we found no change in the RORR time during these three anesthetics induced anesthesia induction in the hM4Dq-CNO group (Fig. 4b-d). EEG recordings also showed contrary performances: during isoflurane and sevoflurane induction, the percentage of δ wave (1–4 Hz) increased and γ wave (25–60 Hz) decreased in the hM4Dq-CNO group, while no significant change was observed during propofol induction (Fig. 4e-j). These data indicate that inhibition of APC^Glu neurons shortened the induction time of isoflurane and sevoflurane anesthesia. Together with the results of the chemogenetic activation experiments, we suggest that APC^Glu neurons may specifically regulate the induction period of inhalational GA.

Fig. 4.

Chemogenetic inhibition of APC^Glu neurons accelerated the induction time of isoflurane and sevoflurane anesthesia. (a) Representative image of APC^Glu expressing hM4Di-eYFP (scale bar, 200 μm). (b) The effect of APC^Glu neuron inhibition by CNO on LORR and RORR time under 1.4% isoflurane anesthesia (n = 6). (c) The effect of APC^Glu neuron inhibition on LORR and RORR time under 2.4% sevoflurane anesthesia (n = 6). (d) The effect of APC^Glu neuron inhibition on LORR and RORR time under propofol anesthesia (n = 6). (e) The power distribution of EEG frequency bands in M4-NS or M4-CNO group under 1.4% isoflurane anesthesia (n = 6). (f) Representative EEG waveforms and spectrograms EEG power of M4-NS or M4-CNO group under 1.4% isoflurane anesthesia. (g) The power distribution of EEG frequency bands in M4-NS or M4-CNO group under 2.4% sevoflurane anesthesia (n = 6). (h) Representative EEG wave forms and spectrograms EEG power of M4-NS or M4-CNO group under 2.4% sevoflurane anesthesia. (i) The power distribution of EEG frequency bands in M4-NS or M4-CNO group under propofol anesthesia (n = 6). (j) Representative EEG wave forms and spectrograms EEG power of M4-NS or M4-CNO group under propofol anesthesia.*p < 0.05, **p < 0.01.

3.4. Optogenetic manipulation of APC^Glu neurons regulates inhalational GA induction

To further confirm the roles of APC^Glu neurons in inhalational GA induction, we used optogenetic tools to manipulate APC^Glu neuron activity to detect its effect on LORR, RORR, and the EEG signature.

The excitatory optogenetic virus (ChR2) with CamKII-cre was unilaterally injected into the APC, and an optical fiber was placed in the same position to specifically activate APC^Glu neurons (Fig. 5a). Continuous blue laser stimulation was applied during the induction and emergence periods of anesthesia (Fig. S1c) [27]. The efficiency of laser stimulation was monitored using c-Fos staining (Fig. 5b).Optogenetic activation of APC^Glu neurons prolonged the LORR time of both isoflurane and sevoflurane anesthesia, but the RORR time did not change, which is consistent with the findings in the hM3Dq-CNO group (Fig. 5c, f). EEG recordings also showed a lower δ wave (1–4 Hz) percentage and a higher percentage of γ wave (25–60 Hz) during the LORR induction of both isoflurane and sevoflurane anesthesia, but not during the emergence period (Figs. 5d, e, g,h and S5a-d). Finally, optogenetic activation of APC^Glu neurons showed no influence on either the LORR or RORR of propofol-induced anesthesia (Figs. 5i-k and S5e, f).

Fig. 5.

Optogenetic activation of APC^Glu neurons prolonged the induction time of isoflurane and sevoflurane anesthesia. (a) Left: Schematic of virus injection and EEG electrode implantation. Right: representative image of APC^Glu expressing ChR2-eYFP (scale bar, 200 μm). (b) Representative images of c-Fos expression in ChR2 groups after 2 hrs with and without blue laser administration respectively (scale bar, 20 μm). (c) The effect of APC^Glu neuron activation by blue laser on LORR and RORR time under 1.4% isoflurane anesthesia (n = 6). (d) The power distribution of EEG frequency bands in ChR2-off or ChR2-on group under 1.4% isoflurane anesthesia (n = 6). (e) Representative EEG wave forms and spectrograms EEG power of ChR2-off or ChR2-on group under 1.4% isoflurane anesthesia. (f) The effect of APC^Glu neuron activation on LORR and RORR time under 2.4% sevoflurane anesthesia (n = 6). (g) The power distribution of EEG frequency bands in ChR2-off or ChR2-on group under 2.4% sevoflurane anesthesia (n = 6). (h) Representative EEG wave forms and spectrograms EEG power of ChR2-off or ChR2-on group under 2.4% sevoflurane anesthesia. (i) The effect of APC^Glu neuron activation on LORR and RORR time under propofol anesthesia (n = 6). (j) The power distribution of EEG frequency bands in ChR2-off or ChR2-ongroup under propofol anesthesia (n = 6). (k) Representative EEG wave forms and spectrograms EEG power of ChR2-off or ChR2-on group under propofol anesthesia.*p < 0.05, **p < 0.01.

To specifically inhibit APC^Glu neuronal activity, NpHR virus with CamKII-cre was bilaterally injected into the APC, followed by optic fiber implantation into the injection sites (Fig. 6a). Yellow laser light was applied during the induction and emergence stages of anesthesia and monitored by c-Fos imaging (Fig. S4b). Optogenetic inhibition of APC^Glu neurons shortened the LORR time of both isoflurane and sevoflurane anesthesia, but the RORR time did not change, which is consistent with the findings in the hM4Dq-CNO group (Fig. 6b, c). EEG recordings correspondingly showed a higher δ wave (1–4 Hz) percentage and lower γ wave (25–60 Hz) percentage during the LORR induction of both isoflurane and sevoflurane anesthesia (Fig. 6e-h). However, no significant changes in propofol-mediated LORR, RORR, and EEG features were detected when we applied a yellow laser to NpHR mice (Fig. 6d, I, j).

Fig. 6.

Optogenetic inhibition of APC^Glu neurons accelerated the induction time of isoflurane and sevoflurane anesthesia. (a) Representative image of APC^Glu expressing NpHR-eYFP (scale bar, 200 μm). (b) The effect of APC^Glu neuron inhibition by yellow laser on LORR and RORR time under 1.4% isoflurane anesthesia (n = 6). (c) The effect of APC^Glu neuron inhibition on LORR and RORR time under 2.4% sevoflurane anesthesia (n = 6). (d) The effect of APC^Glu neuron inhibition on LORR and RORR time under propofol anesthesia (n = 6). (e) The power distribution of EEG frequency bands in NpHR-off or NpHR-on group under 1.4% isoflurane anesthesia (n = 6). (f) Representative EEG wave forms and spectrograms EEG power of NpHR-off or NpHR-on group under 1.4% isoflurane anesthesia. (g) The power distribution of EEG frequency bands in NpHR-off or NpHR-on group under 2.4% sevoflurane anesthesia (n = 6). (h) Representative EEG wave forms and spectrograms EEG power of NpHR-off or NpHR-on group under 2.4% sevoflurane anesthesia. (i) The power distribution of EEG frequency bands in NpHR-off or NpHR-on group under propofol anesthesia (n = 6). (j) Representative EEG wave forms and spectrograms EEG power of NpHR-off or NpHR-on group under propofol anesthesia.*p < 0.05, **p < 0.01.

Previous studies have shown that the PC encodes odor identity through the regulation of gamma/beta and theta/delta oscillations [15,28], which is consistent with the EEG recording variation during isoflurane and sevoflurane anesthesia induction in both chemogenetic and optogenetic experiments. Together with the data collected from chemogenetic and optogenetic manipulation of APC^Glu neurons, we conclude that APC^Glu neurons may specifically regulate the induction period of inhalational GA, including isoflurane and sevoflurane.

3.5. Odor processing may be involved in the inhalational GA induction regulated by APC^Glu neurons

Since PC is a major region that encodes odor objects [12], and our study showed that only the induction of inhalational GA is influenced by APC^Glu neuron activity, we wondered whether odor is an interfering factor for PC to affect anesthesia. For this purpose, we first checked the isoflurane induction time (LORR) in the APC- and PPC-lesioned mice while we placed a bottle of perfume in the recording chamber. Interestingly, we found that perfume reversed the effect of APC lesions on LORR (Fig. 7a). Furthermore, we performed the tests in optogenetically manipulated mice. Similar to the APC-lesioned mice, we found that perfume reversed the shortened LORR time of isoflurane and sevoflurane anesthesia in the NpHR group, but not in the ChR2 group (Fig. 7b, c). These data suggest that odor may play a role for APC^Glu neurons in regulation of inhalational GA induction.

Fig. 7.

Odor processing reverses the effect of APC^Glu neuron inhibition on induction of isoflurane and sevoflurane anesthesia. (a) The effect of perfume employment on LORR and RORR time under isoflurane anesthesia in the APC and PPC lesion mice (n = 6). (b) The effect of perfume employment on LORR time under isoflurane, sevoflurane or propofol anesthesia in the APC^Glu ChR2 mice (n = 6). (c) The effect of perfume employment on LORR and RORR time under isoflurane, sevoflurane or propofol anesthesia in the APC^Glu NpHR mice (n = 6). (d) Representative images of c-Fos expression in wake, perfume and perfume removed for 2 hrs groups in PC (scale bar, 200 μm). (e) Quantification of c-Fos number (per mm²) in PC (wake: 330 ± 33; perfume: 730 ± 64; perfume removed for 2 hrs: 1020 ± 70; n = 5).*p < 0.05, **p < 0.01.

Furthermore, we performed a similar time course of isoflurane anesthesia to check the effect of perfume on APC activity. Similar to isoflurane, perfume administration for 2 h showed increased number of c-Fos in APC. However, when we placed the mice in a new chamber without perfume for 2 h (as recovery from isoflurane anesthesia), more accumulated c-Fos expression was detected in the APC, whereas the c-Fos level decreased during recovery from isoflurane anesthesia (Fig. 7d). These findings may partly explain the reversible effect of perfumes on APC lesions and NpHR mice. The continuous activation of APC neurons by perfume may be resistant to the APC lesion or inhibition, so that the restoration of LORR was observed in the lesion and NpHR mice under perfume administration, which is consistent with the fact that perfume did not alter the LORR in ChR2 mice because the APC neurons were already activated. However, the different effects of perfume and isoflurane on APC also indicated that odor processing is not the only possible role of APC in regulating inhalational GA induction. PC projections to different nuclei may contribute to the regulation of inhalational GA induction. Besides connections with the olfactory system, such as the OB and olfactory cortex [12], previous studies have shown that PC can directly project to the mPFC and control mPFC—NAc plasticity to regulate social transmission in mice [27]. The mPFC and NAc play critical roles in GA [33], [34], [35], [36], [37], [38]. These findings suggest that the PC-mPFC pathway may be involved in inhalational GA control in addition to odor processing.

Lastly, our data showed that neither APC nor PPC activity affected propofol-induced anesthesia, although propofol increased c-Fos expression in the PC. The possible reasons are as follows: first, the PC regulated anesthesia mainly through odor programing; second, propofol and isoflurane (or sevoflurane) had different receptors in neurons [2,6], and the PC neurons had been reported to contain multiple kinds of receptors [12,14], which indicated that propofol and isoflurane (or sevoflurane) may have different effects on the PC neurons; and finally, although only a small number of inhibitory neurons (IN) were found in the PC, the function of these IN in anesthesia is barely understood. Therefore, more precise tools manipulating the PC neurons and their related neural circuits are useful for clarifying the exact roles of PC in GA.

4. Conclusion

APC^Glu neurons prolong the induction time of isoflurane and sevoflurane anesthesia, but have no effect on emergence time. Manipulation of APC^Glu neuronal activity alters neither the induction nor the emergence time of propofol-mediated anesthesia. In addition, odor encoding may be involved in the regulation of APC^Glu neurons during isoflurane and sevoflurane induction. These findings reveal a crucial role of APC^Glu neurons in the induction of inhalational GA.

Author contributions

All authors contributed to the experimental conception and design of the study. Material preparation, data collection, and analyses were performed by Liang Zhou, Rulan Yi, Huanyao Tang, and Qipeng Ran. The manuscript was written by Liang Zhou, Yu Zhang, and Tian Yu, and all authors commented on the previous versions of the manuscript. All authors have read and approved the final manuscript.

Declaration of competing interest

The authors declare that they have no conflicts of interest in this work.

Biographies

Liang Zhou(BRID: 08229.00.25916) is an associated professor at Key Laboratory of Brain Science; Guizhou Key Laboratory of Anesthesia and Organ Protection, Zunyi Medical University. He received his Ph.D. degree in Sichuan University. His research interests are focused on the molecular and neural circuit mechanisms of general anesthesia.

Tian Yu(BRID: 09187.00.03133) is the principle investigator of Key Laboratory of Brain Science; Guizhou Key Laboratory of Anesthesia and Organ Protection, Zunyi Medical University. Her research interests are focused on the neural circuit mechanisms of general anesthesia.