Role of the Supramammillary Nucleus–Medial Septum Glutamatergic Pathway in Mediating the Effects of Isoflurane Anesthesia

Abstract

Background:

Glutamatergic neurons in the supramammillary nucleus (SuM) have been recently identified as a key node in arousal system, yet their role in regulating general anesthesia remains unclear. The aim of the current study is to examine the role of the glutamatergic supramammillary neurons and their projections to the medial septum in mediating the effects of isoflurane anesthesia.

Methods:

Fiber photometry recording was used to determine the changes in calcium signals of glutamatergic neurons in the SuM during isoflurane anesthesia. Optogenetic and chemogenetic approaches were employed to manipulate SuM glutamatergic neuron activity, and the effects on cortical activity, behavioral responses, and physiologic parameters—including pupil diameter, respiratory rate, and blood pressure—were examined in anesthetized mice. Both male and female mice were used in this study.

Results:

The activities of SuM glutamatergic neurons decreased during isoflurane anesthesia and recovered after the emergence. Optogenetic activation of these neurons enhanced cortical activity, decreasing electroencephalogram delta power (mean ± SD, prestimulation vs. stimulation: 51.35 ± 7.26% vs. 32.08 ± 10.48%, n = 8, P = 0.002) and burst suppression ratio (81.82 ± 7.83% vs. 44.53 ± 28.62%, n = 8, P = 0.002). Furthermore, optogenetic activation altered physiologic parameters including enlarged pupil diameter (prestimulation vs. stimulation: 1.05 ± 0.08% vs. 1.95 ± 0.46%, n = 8, P < 0.001), increased respiratory rate (0.98 ± 0.08% vs. 1.57 ± 0.39%, n = 10, P < 0.001) and elevated blood pressure and induced behavioral responses including increased arousal scores and accelerated emergence (light off vs. light on, 171.40 ± 56.39 s to 59.88 ± 27.18 s, n = 8, P = 0.007). Moreover, chemogenetic activation produced similar effects, whereas inhibition led to opposite effects. Finally, optogenetically activating SuM glutamatergic terminals projecting to the medial septum mimicked the effects of activating SuM glutamatergic soma and increased the activity of medial septum glutamatergic neurons.

Conclusions:

This study identifies glutamatergic neurons of the SuM as key neural substrates regulating isoflurane anesthesia and facilitating emergence through their projections to the medial septum.

In mice, synchronous fiber photometry and electroencephalogram/electromyogram recordings revealed that activity of glutamatergic neurons in the supramamillary nucleus decreased during isoflurane anesthesia and recovered after emergence. Both optogenetic and chemoigenetic activation of supramamillary glutamatergic neurons increased arousal scores and accelerated emergence from isoflurane anesthesia, while chemogenetic inhibition of these neurons had opposite effects. Optogenetic activation of axonal terminals of supramamillary glutamatergic neurons projecting to the medial septum mimicked the effects of activating the soma of these neurons and increased the activity of glutamatergic neurons in the medial septum.

Video Abstract

Video Abstract.

Download video file ^{(38.4MB, mp4)}

Editor’s Perspective

What We Already Know about This Topic

• Glutamatergic neurons of the supramammillary nucleus are involved in the regulation of sleep–wake cycles through their projections to the medial septum

• The question of whether the supramammillary nucleus–medial septum pathway is also implicated in general anesthesia mechanisms of actions remains incompletely explored

What This Article Tells Us That Is New

• In mice, synchronous fiber photometry and electroencephalogram/electromyogram recordings revealed that activity of glutamatergic neurons in the supramammillary nucleus decreased during isoflurane anesthesia and recovered after emergence

• Both optogenetic and chemogenetic activation of supramammillary glutamatergic neurons increased arousal scores and accelerated emergence from isoflurane anesthesia, while chemogenetic inhibition of these neurons had opposite effects

• Optogenetic activation of axonal terminals of supramammillary glutamatergic neurons projecting to the medial septum mimicked the effects of activating the soma of these neurons and increased the activity of glutamatergic neurons in the medial septum

General anesthesia induces reversible unconsciousness and prevents patients from feeling pain during surgery, benefiting millions of patients each year. However, the precise neural mechanisms underlying general anesthesia remain largely unclear. Studying the neural circuits underlying anesthesia is essential for improving anesthesia depth control and enhancing patient safety during surgical procedures. Considering that substantial neurophysiologic similarities exist between general anesthesia and natural sleep, such as the reversible loss of consciousness, local electroencephalogram (EEG) signatures, similar behavior transitions, and decreased autonomic functions, general anesthesia and natural sleep are presumed to share a portion of neural circuits.^1,2 Recent evidence shows that general anesthetics induce hypnosis by acting on neural substrates regulating sleep–wake behavior, like the basal forebrain and parabrachial nucleus.^3–6 Further research is required to explore whether other sleep–wake substrates contribute to the modulation of general anesthesia.

The supramammillary nucleus (SuM) is a ventromedial posterior hypothalamic region lying above the mammillary body. The SuM primarily contains glutamatergic neurons, a small part of which coexpress neurotransmitter γ-aminobutyric acid (GABA).⁷ Recent studies prove that the SuM is a key node in the sleep–wake system.^8,9 Chemogenetic activation of SuM glutamatergic neurons produces sustained behavioral and cortical arousal in mice, while chemogenetic inhibition of these neurons results in fragmentation of wakefulness and an increase in sleep.⁸ Optogenetic activation of the SuM axon terminals innervating the dentate gyrus during slow-wave sleep significantly increases the probability of awakening in mice, accompanied by increased gamma power and decreased delta power in the EEG.⁹ Considering the key role of SuM glutamatergic neurons in regulating sleep–wake behavior, we hypothesized that SuM glutamatergic neurons may also be involved in the regulation of general anesthesia.

Neuroanatomical studies have shown that the SuM is widely connected to various brain structures associated with general anesthesia.^10,11 Among these brain structures, medial septum (MS) is particularly noteworthy. Findings from electrolytic and pharmacologic lesions,^12,13 as well as pharmacologic activation and inactivation,^14,15 demonstrate that the MS plays a crucial role in regulating sleep–wake cycles and general anesthesia. Notably, recent findings showed that optogenetic activation of SuM glutamatergic projections in the MS induced rapid transitions to wakefulness from both rapid eye movement (REM) and non-REM sleep in mice.¹⁶ Chemogenetic activation of MS-projecting glutamatergic SuM neurons increased wakefulness amount and prolonged sleep latency, while chemogenetic inhibition of these neurons reduced wakefulness amount and shortened sleep latency.¹⁶ Based on those findings, we hypothesized that SuM glutamatergic neurons may modulate general anesthesia through their projection to the MS.

To test this hypothesis, synchronous fiber photometry and EEG/electromyogram (EMG) recording were first used to explored the activities of SuM glutamatergic neurons during isoflurane anesthesia in mice. Then, optogenetic and chemogenetic approaches were used to examine the effects of manipulating SuM glutamatergic neurons on cortical activation, physiologic parameters, and behavioral responses during isoflurane anesthesia. Finally, we combined optogenetics and fiber photometry recording to identify the downstream targets of SuM glutamatergic neurons in regulating isoflurane anesthesia. Our studies enhance the understanding of neuroanatomical basis of general anesthesia and identify a potential target for accelerating emergence and improving anesthetic safety in clinical practice.

Materials and Methods

Mice

This study utilized adult male and female C57BL/6J mice aged 8 to 10 weeks (GemPharmatech Co., Ltd., China) weighing 22 to 26 g. All experimental animals were housed in the standard environment (temperature: 25° ± 0.5°C, humidity: 55 ± 5%) with a 12-h automatic light/dark cycle (lights on at 7:00 am) and free access to food and water. The mice were randomly assigned to different treatment groups using a random number table to minimize the effect of systematic factors on experimental results. SuM soma experiments include fiber photometry recording, optogenetic activation, chemogenetic activation, chemogenetic inhibition, and optogenetic inhibition. In the fiber photometry recording experiment, the animals expressing GCaMP6m (6 or 8 mice) were used to monitor calcium signaling activity fluctuations during isoflurane exposure. In the optogenetic activation experiment, the animals were divided into ChR2 and mCherry groups (8 or 10 mice per group). In the chemogenetic activation experiment, the animals are divided into hM3Dq and mCherry groups (8 to 14 mice per group). In the chemogenetic inhibition experiment, the animals are divided into hM4Di and mCherry groups (14 mice per group). In the optogenetic inhibition experiment, the animals expressing stGtACR2 (9 mice) were stimulated with blue light, while red light served as the control. In the SuM–MS experiments, fiber photometry recording, optogenetic activation were included. In the fiber photometry recording experiment, the animals expressing GCaMP6m (6 or 8 mice) were used to monitor calcium signaling activity fluctuations during isoflurane exposure. In the optogenetic activation experiment, the animals were divided into ChR2 and mCherry groups (8 to 11 mice per group). The experimenters who conducted behavioral tests were blinded to group allocation. Data analysis was performed by a researcher who was also blinded to the group allocations. All experimental procedures were approved by the Institutional Animal Care and Use Committee of Fujian Medical University (approval No. 2024-0340).

Virus Injection, Optic Fiber Implantation, and Electrode Fixation

The mice were anesthetized with 2% isoflurane and placed on a stereotaxic apparatus (RWD Life Science, China). Then, we cut the skin along the center line and used the skull drill to open a small craniotomy hole above the SuM for virus injection and optic fiber implantation. The appropriate viruses were injected (40 nl/min) bilaterally into the SuM (anteroposterior = −2.30 mm, mediolateral = ±0.30 mm, dorsoventral = −5.15 mm) and MS (anteroposterior = +1.10 mm, mediolateral = 0.00 mm, dorsoventral = −4.60 mm) according to the different groups. After the injection of AAV (100 to 150 nl for each site),¹⁷ the glass pipette was left in the injection location for 10 min to allow the virus diffusion. After the injection, the EEG/EMG electrodes were implanted in the skull and fixed with dental cement. Optical fibers were implanted 0.2 mm above the virus injection site and fixed with dental cement. After surgery, the mice were housed in an appropriate environment until recovery. The EEG/EMG recording and behavioral experiments were performed about 21 days later. Carprofen (4 mg/kg) was used for postoperative pain, dissolved in physiologic saline and injected subcutaneously every 12 h for 2 days after surgery. Body temperature was maintained at 36.5° ± 0.5°C throughout the surgical procedure via a heating pad positioned beneath the chamber.

All viruses used in the experiment were AAV-CaMKII-GCaMP6m (BrainVTA, China), AAV-CaMKII-ChR2-mCherry (Taitool, China), AAV-hSyn-DIO-stGtACR2-eGFP (Taitool), AAV-hSyn-DIO-ChrimsonR-mCherry (BrainVTA), AAV-CaMKII-hM3Dq-mCherry (BrainVTA), AAV-CaMKII-hM4Di-mCherry (Taitool), AAV-CaMKII-Cre (BrainVTA), and AAV-CaMKII-mCherry (HANBIO, China). All viruses were packaged into adeno-associated virus serotype 2/9 vectors with titers of 2 to 5 × 10¹² viral particles per milliliter.

Fiber Photometry Recording

Before the experiment, the mice were allowed to acclimatize in an anesthesia induction chamber (length × width × height = 23 × 10 × 17 cm) made of acrylic material for 10 min with a continuous flow of pure oxygen at 1.5 l/min. Changes in calcium signals were recorded for a total of 90 min (30 min before, 30 min during, and 30 min after 1.4% isoflurane anesthesia), and the values of ΔF/F were calculated for these three periods. In the experiment combining optogenetics with in vivo fiber photometry recording, 1.4% isoflurane was maintained for 30 min, and then red light stimulation (30 Hz, 10 ms, 20 s) was given.^18,19 The concentration of 1.4% isoflurane was chosen because it corresponds to approximately 1.0 minimum alveolar concentration.^20–22

EEG/EMG recording was utilized to determine the time point of the occurrence of burst suppression (BS) onset and BS termination, and the corresponding changes in calcium signals were further analyzed. We analyzed changes in calcium signals across three consecutive periods relative to BS onset (from −300 to 600 s, where 0 s represents the timeof isoflurane administration): −300 to 0 s (preanesthesia period), 0 to BS onset (pre-BS onset period), and BS onset to 600 s (post-BS onset period). In addition, we analyzed changes in calcium signals during three consecutive periods relative to BS termination (from −300 to 600 s, where 0 s represents the time of isoflurane cessation): −300 to 0 s (anesthesia period), 0 to BS termination (pre-BS termination period), and BS termination to 600 s (post-BS termination period). For simultaneously recorded calcium signals in the SuM and MS, we analyzed the correlation of calcium changes within 5-s periods after the time of isoflurane on, BS onset, isoflurane off, and BS termination, respectively. During the induction and emergence of isoflurane anesthesia, we recorded the changes in calcium signals before and after loss of righting reflex (LORR) and recovery of righting reflex (RORR). For further analysis, two consecutive time periods (from −30 to 30s, where 0 s represents the occurrence of LORR or RORR) were examined.

Arousal Scoring during Isoflurane Anesthesia

The mice were acclimated in an anesthesia induction chamber for 5 min, and the optogenetic system was connected via a fiber optic cable. The mice were exposed to 1.4% isoflurane for 10 min, and then the isoflurane concentration was reduced to 0.6%. If the mouse showed any signs of RORR, the isoflurane concentration was increased by 0.1% until the mice continuously maintained LORR for 30 min. Then a 60-s photostimulation (30 Hz, 10 ms pulse width) was given to the mice, and the behavioral responses were monitored throughout the light stimulation period. The arousal levels of the mice were assessed based on observed behavioral responses and established evaluation criteria.¹⁹

Estimation of Induction and Emergence from Isoflurane Anesthesia

The LORR and RORR in rodents are regarded as surrogates for loss and recovery of consciousness in humans.²⁰ In the optogenetic behavioral experiment, mice were given photostimulation and exposed to 1.4% isoflurane with pure oxygen at a flow rate of 1.5 l/min. The isoflurane concentration in the chamber was monitored using an anesthesia monitor (Bene View T5; Mindray, China). The chamber was gently rotated by 90 degrees with a 10-s interval. The mice were considered to show LORR if they could not turn over within 30 s. After a 30-min stabilization period under 1.4% isoflurane anesthesia, the delivery of isoflurane was discontinued. Then the mice were exposed to room air and given photostimulation to determine the time to the onset of RORR. In the chemogenetic behavioral experiments, the mice were injected with 1.0 mg/kg clozapine N-oxide (CNO) or vehicle intraperitoneally 1 h before the test. The mice were then exposed to 1.4% isoflurane and the time to the onset of LORR was recorded. After stabilizing for 30 min during 1.4% isoflurane anesthesia, the mice were exposed to room air and the time to the onset of RORR was recorded. In the LORR dose–response experiments, isoflurane concentration began at 0.4% and was progressively increased by 0.1% every 15 min until LORR was observed. For the RORR dose–response experiments, isoflurane concentration started at 1.4% and was reduced by 0.1% every 15 min until RORR occurred.

Spectrum and Burst Suppression Ratio Experiments with Optogenetic Manipulations

The EEG/EMG electrodes were connected to the recording system, and optical fibers were connected to a light source for photostimulation. EEG/EMG signals were transmitted to an amplifier and collected at a sampling rate of 128 Hz. The recorded signals were subsequently analyzed using software SleepSign 2 (Kissei Comtec, Japan). The mice were acclimatized in the induction chamber for 5 min, after which baseline EEG/EMG signals were recorded for 5 min. For the optogenetic spectrum experiment, 0.8% isoflurane was delivered for induction, and after 30 min of continuous delivery, photostimulation (30 Hz, 10 ms, 120 s) was given to the mice. After the termination of the isoflurane for 10 min, the mice were removed from the anesthesia chamber. For the burst suppression ratio (BSR) experiment, 1.4% isoflurane was delivered for induction and maintained for 30 min. After ensuring that the mice had entered a stable burst suppression pattern for at least 5 min, photostimulation (30 Hz, 10 ms, 120 s) was given. In the chemogenetic spectrum and BSR experiments, the mice were intraperitoneally injected with 1 mg/kg CNO or vehicle 1 h before the experiment.

For the spectrum and BSR analysis, the EEG/EMG signals were amplified and sampled at a rate of 128 Hz. For the spectrum analysis, four frequency bands were analyzed (delta: 0.5 to 4.0 Hz, theta: 4.0 to 7.0 Hz, alpha: 8.0 to 15.0 Hz, and beta: 16.0 to 30.0 Hz). Fast Fourier transform were utilized to determine the relative change in total power of SuM glutamatergic neurons before, during, and after light stimulation in the optogenetic experiments. In the chemogenetic experiments, the band powers were calculated during the 5 min of isoflurane administration and the 5 min before and after isoflurane cessation. For the BSR analysis, raw EEG data were converted to text format for subsequent amplitude analysis of the EEG signals using MATLAB R2019b (MathWorks, USA). In the optogenetic experiments, changes in BSR were calculated for 120 s before, during, and after photostimulation. In the chemogenetic experiments, BSR changes were calculated during the 20- to 30-min period of isoflurane administration.

Pupil Size Measurement

The experiments were conducted using an infrared camera (REE-USB4KHDR01; Rervision, China) to record the pupil size at a frame rate of 30 Hz. Bonsai 2.8.1 was employed to estimate the pupil size by detecting the pupil's edge through pixel analysis in each frame. The pupil diameter was analyzed within a defined region frame by frame, and the data were exported for further analysis. Before the experiment, the mice were anesthetized with 1.4% isoflurane for 5 min, and their heads were secured in a stereotaxic apparatus. Isoflurane anesthesia was maintained throughout the experimental procedure. The ambient light was turned off to minimize environmental influences on pupil size, and baseline values were recorded after the pupil size had stabilized for 2 min. The baseline size was defined as the pupil size 60 s before light administration, and changes in pupil size were analyzed during the 60-s photostimulation period (30 Hz, 10 ms) and 60 s after light stimulation.

Analysis of the Respiratory Rate

Respiratory signals were recorded via a whole-body plethysmograph system (WBP-4M; TOW-INT TECH, China). The mice were placed in plethysmography chambers that connected to respiratory transducers, and the transducers converted the change in chamber volume into electrical signals and transmitted to a computer. The respiratory rate was calculated by analyzing and processing the data using the accompanying software. Before the experiment, the mice adapted to the body box for 5 min, and 1.2% isoflurane was delivered. After the mice were anesthetized, the respiratory rate was recorded for 60 s before photostimulation as the baseline. We analyzed the changes in the respiratory rate during the 60 s of light stimulation (30 Hz, 10 ms), as well as the rate during the 60 s after the light stimulation.

Heart Rate and Blood Pressure Measurement

The heart rate and tail artery blood pressure of anesthetized mice were measured using a noninvasive blood pressure system (BP-2010A; Softron Biotechnology, China). Mice under 1.4% isoflurane anesthesia received optogenetic stimulation (30 Hz, 10 ms, 60 s). The heart rate, diastolic blood pressure, and systolic blood pressure were continuously recorded throughout the experiment. Core body temperature was maintained at 36.5° ± 0.5°C using a heating pad.

Statistical Analysis

In this study, we used 183 male and female mice for SuM soma experiments. Histologic analyses were performed at the end of all experiments to examine the location of virus expression. A total of 76 mice with inaccurate location of virus injection, inaccurate location of optic fibers, and poor recovery or death after virus injection were excluded, and the remaining 107 mice were included in the data analysis. In the SuM–MS experiments, we used 58 male mice. We excluded 22 mice with inaccurate injection sites or poor recovery and used 36 mice for data analysis. GraphPad Prism 8.0 (GraphPad Software, USA) and MATLAB R2019b (MathWorks) software were used for statistical analyses. In this study, we did not perform an a priori statistical power calculation. Instead, sample sizes were determined based on our previous experience with assay variance and the feasibility of the experiment.¹⁹ The experimenters conducting all behavioral tests were blinded to group allocation. The normality of the data was assessed using the Shapiro–Wilk test. Based on the results of this test, either parametric or nonparametric tests were applied. Paired Student’s t test was used to compare the differences of neuronal calcium signal changes in two consecutive time sections during LORR and RORR between groups, as well as the difference in burst suppression ratio in chemogenetic activation experiments. The Wilcoxon signed-rank test was used to analyze the differences of heart rate and systolic and diastolic blood pressure before and during optogenetic stimulation, as well as to analyze isoflurane concentrations for loss of righting reflex in optogenetic and chemogenetic activation experiments. The Mann–Whitney rank sum test was used to analyze the differences in arousal scoring between the groups. One-way repeated-measure ANOVA was used to analyze the differences in neuronal calcium signal changes across three consecutive time sections during the induction or emergence periods of 1.4% isoflurane anesthesia, changes in pupil size, respiratory rate, and the burst suppression ratio. Two-way repeated-measure ANOVA was used to analyze the relative EEG power before, during, and after optogenetic stimulation or chemogenetic activation experiments, as well as to compare the differences in induction time and emergence time between groups in optogenetic and chemogenetic experiments. In the ANOVA post hoc analysis, we used the Bonferroni post hoc test with adjusted P value. Data from repeated measurements within the same animal were averaged before statistical analysis. All presented data in this study are shown as means ± SD or mean (95% CI). In all figures, “n” refers to the number of biologically independent animals per group, unless otherwise specified. Significance was defined as P < 0.05 (*), P < 0.01 (**), or P < 0.001 (***). All figures were created using Adobe Illustrator (USA).

Results

Population Activity of SuM Glutamatergic Neurons Change during the Induction and Emergence from Isoflurane Anesthesia

To investigate whether the activity of SuM glutamatergic neurons is involved in isoflurane anesthesia, we employed fiber photometry to assess calcium signal dynamics during both the induction and the emergence process in mice (fig. 1, A to C). The AAV-CaMKII-GCaMP6m was injected into the SuM, and an optical fiber was planted above the SuM (fig. 1, A and B). Electrodes were positioned in the cerebral cortex and posterior cervical muscle to monitor EEG/EMG activity. After the administration of 1.4% isoflurane, we observed a biphasic decline in the activity of SuM glutamatergic neurons, characterized by an initial rapid decline followed by a slower decline (fig. 1, D and E). Representative EEG/EMG signals synchronized with calcium signal dynamics are provided in supplemental fig. S1, A and B (https://links.lww.com/ALN/E32).

Fig. 1.

Change of calcium signals of supramammillary nucleus glutamatergic neurons during isoflurane anesthesia. (A) Schematic diagram of AAV-CaMKII-GCaMP6m injected into the SuM of C57BL/6J mice. (B) Representative immunofluorescence image of GCaMP6m expression in the SuM after 4 weeks after virus transfection. (C) Schematic diagram of EEG/EMG and fiber photometry recording. (D) Heat maps of calcium signals change of the SuM glutamatergic neurons before, during, and after exposure to 1.4% isoflurane (n = 8); 0 represents the moment of 1.4% isoflurane on; 30 represents the moment of 1.4% isoflurane off. (E) Peri-event plot of the average Ca²⁺ transient aligned to 1.4% isoflurane anesthesia (n = 8); the thick line indicates the mean, and the shaded area indicates SD. The light blue shadow indicates exposure to isoflurane. (F) Statistical chart of calcium signal before, during, and after 1.4% isoflurane exposure (n = 8). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (G) Calcium signal changes of SuM glutamatergic neurons aligned to BS onset during 1.4% isoflurane exposure. (Left) Heat maps (top) and peri-event plot (bottom) of calcium signals aligned to BS onset (n = 8). (Right) Diagram showing the calcium signal trace aligned to BS onset (top) and statistical results showing calcium signals in preanesthesia, pre-BS onset, and post-BS onset stages (bottom). The shadow indicates exposure to isoflurane. Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (H) Calcium signal changes of SuM glutamatergic neurons aligned to BS termination during 1.4% isoflurane exposure. (Left) Heat maps (top) and peri-event plot (bottom) of calcium signals aligned to BS termination (n = 8). (Right) Diagram showing the calcium signal trace aligned to BS termination (top). (Bottom right) Statistical results showing changes of calcium signals in preanesthesia, pre-BS termination, and post-BS termination stages. The shadow indicates exposure to isoflurane. Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (I) Calcium signal changes of SuM glutamatergic neurons aligned to loss of righting reflex (LORR) during 1.4% isoflurane exposure. (Left) Heat maps (top) and peri-event plot (bottom) of calcium signals aligned to LORR (n = 6). (Right) Statistical results showing changes of calcium signals pre- and post-LORR. Statistical analysis was performed using paired Student’s t tests. (J) Calcium signal changes of SuM glutamatergic neurons aligned to recovery of righting reflex (RORR) during 1.4% isoflurane exposure. (Left) Heat maps (top) and peri-event plot (bottom) of calcium signals aligned to RORR (n = 6). (Right) Statistical results showing changes of calcium signals pre- and post-RORR. Statistical analysis was performed using paired Student’s t tests. *P < 0.05; **P < 0.01; ***P < 0.001. Experimental data are expressed as means ± SD. BS, burst suppression; EEG, electroencephalogram; EMG, electromyogram; Iso, isoflurane; LORR, loss of righting reflex; RORR, return of righting reflex; SuM, supramammillary nucleus.

Notably, after 30 min of isoflurane exposure, we observed an approximate 30% reduction in calcium signals (fig. 1E). After the cessation of isoflurane delivery, the calcium signals gradually recovered back to baseline levels (fig. 1E). Statistical analysis indicated a significant reduction in calcium signals during the exposure of isoflurane (mean of differences [95% CI]: during vs. pre, −24.94% [−32.98 to −16.90%], n = 8; P < 0.001), and a significant increase in calcium signals after the terminal of isoflurane exposure (mean of differences [95% CI]: post vs. during, 12.97% [4.93 to 21.01%], n = 8, P=0.001; fig. 1F).

Cortical EEG can effectively reflect the state of cortical activity.²³ BS is an EEG pattern characterized by alternating high-amplitude slow waves (burst waves) and nearly flat, low-amplitude waves (suppression waves).²⁴ The BS is a hallmark parameters of deep general anesthesia.²⁵ We subsequently investigated activity changes of SuM glutamatergic neurons during the onset and termination phases of isoflurane-induced BS. After exposure to isoflurane, EMG signals showed a significant decline, while EEG patterns shifted from high-frequency, low-amplitude oscillations to low-frequency, high-amplitude oscillations. The heat map revealed a decline in GCaMP6m signals before the onset of BS (fig. 1G, left), with the signals continuing to decline until stabilized at a steady level. We divided the induction phase into three consecutive periods: the preanesthesia (−300 to 0 s; with 0 s denoting the initiation of 1.4% isoflurane), pre-BS onset (0 s to the time point at which BS occurred), and post-BS onset (from BS onset to 600 s) periods. Statistical analysis indicated that calcium signals significantly decreased after the occurrence of BS (mean of differences [95% CI]: post-BS onset vs. pre-BS onset, −11.47% [−15.50 to −7.44%], n = 8, P < 0.001; fig. 1G, bottom right). After the termination of isoflurane exposure, the EMG signals significantly increased, while EEG patterns shifted from high-amplitude, low-frequency oscillations to low-amplitude, high-frequency oscillations. The heat map revealed a rise in GCaMP6m signals after BS termination (fig. 1H, left), with the signals continuing to rise until stabilized at a steady level. We divided the emergence phase into three consecutive periods: the anesthesia period (−300 to 0 s; 0 s marking the cessation of isoflurane), pre-BS termination (0 s to BS termination), and post-BS termination (from BS termination onset to 600 s). Statistical analyses showed that calcium signals increased significantly after BS termination (mean of differences [95% CI]: post-BS termination vs. pre-BS termination, 21.40% [12.15 to 30.66%], n = 8, P < 0.001; fig. 1H, bottom right).

Moreover, we further analyzed changes in calcium signals in SuM glutamatergic neurons during isoflurane-induced LORR and RORR. We analyzed the changes in calcium signals before and after LORR (from −30 to 30s, where 0 s represents the occurrence of LORR). The results showed a gradual decrease in calcium signal before the occurrence of LORR and a further decrease after LORR occurred (fig. 1I, left). Statistical analysis showed a significant decrease in SuM glutamatergic neuronal activity after the occurrence of LORR (mean of differences [95% CI]: post-LORR vs. pre-LORR, −7.45% [−10.12 to −4.78%], n = 6, P < 0.001; fig. 1I, right). Similarly, we analyzed the changes in calcium signals before and after RORR (from −30 s to 30 s, where 0 s represents the occurrence of RORR). The calcium signals gradually increased before the occurrence of RORR and reached a peak after RORR occurred (fig. 1J, left). Statistical analysis showed a significant increase in calcium signals after the occurrence of RORR (mean of differences [95% CI]: post-RORR vs. pre-RORR, 8.51% [2.68 to 14.35%], n = 6, P = 0.01; fig. 1J, right). Considering the potential influence of sex differences on sensitivity to isoflurane anesthesia, we also examined the activity of SuM glutamatergic neurons in female mice under the same experimental conditions. The experimental results showed that the calcium signaling of SuM glutamatergic neurons in female mice exhibit similar dynamic changes with those observed in male mice (supplemental fig. S2, https://links.lww.com/ALN/E34). Taken together, these results indicated that the activity of SuM glutamatergic neurons is closely related to the induction and emergence from isoflurane anesthesia.

Optogenetic Stimulation of SuM Glutamatergic Neurons Enhances Cortical Activation during Isoflurane Anesthesia

After establishing the correlation between SuM glutamatergic neurons and isoflurane anesthesia, we employed an optogenetic approach to investigate the role of these neurons in cortical activation during different stages of isoflurane anesthesia. AAV-CaMKII-ChR2-mCherry was bilaterally injected into the SuM, and EEG/EMG electrodes were planted (fig. 2A). Four weeks later, we confirmed the expression of ChR2-mCherry in the SuM (fig. 2B). We first applied photostimulation (30 Hz, 10 ms, 120 s) to SuM glutamatergic neurons in a subanesthetic state induced by 0.8% isoflurane and analyzed the change of EEG spectra. In the ChR2 group, the EEG immediately shifted to a high-frequency, low-amplitude pattern, accompanied by enhanced EMG signals after photostimulation (fig. 2C; supplemental video S1, https://links.lww.com/ALN/E40). EEG spectral analysis showed that photostimulation significantly decreased delta power (prestimulation vs. stimulation: 51.35 ± 7.26% vs. 32.08 ± 10.48%, P = 0.002), while increased alpha (prestimulation vs. stimulation: 11.42 ± 2.48% vs. 17.76 ± 4.50%, P = 0.003) and beta power (prestimulation vs. stimulation: 6.71 ± 1.96% vs. 20.28 ± 6.38%, n = 8, P < 0.001; fig. 2E). In the mCherry group, EEG signals did not show obvious changes during photostimulation (fig. 2D; supplemental video S2, https://links.lww.com/ALN/E41), and photostimulation did not induce significant changes in EEG delta, alpha, or beta power (fig. 2F). We further conducted a detailed analysis of the normalized power density of EEG signals to illustrate the changes across different frequency bands. There is a trend of reduced EEG power density from 0 to 7 Hz in the ChR2 group but not in the mCherry group (supplemental fig. S3, A and B, https://links.lww.com/ALN/E35).

Fig. 2.

Optogenetic activation of supramammillary nucleus glutamatergic neurons induces cortical activation during isoflurane anesthesia. (A) Schematic diagram showing the injection of AAV-CaMKII-ChR2-mCherry into the SuM of C57BL/6J mice. (B) Representative immunofluorescence image of ChR2 expression in the SuM. (C) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of ChR2 mouse in the pre-stim, stim (30 Hz, 10 ms, 120 s), and post-stim phases under 0.8% isoflurane anesthesia. (D) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of mCherry mouse in the pre-stim, stim(30 Hz, 10 ms, 120 s), and post-stim phases under 0.8% isoflurane anesthesia. (E) Statistical results of the EEG power of ChR2 mice in the pre-stim, stim (30 Hz, 10 ms, 120 s), and post-stim phases under 0.8% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (F) Statistical results of EEG power of pre-stim, stim, and post-stim (30 Hz, 10 ms, 120 s) in mCherry mice during 0.8% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (G) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of ChR2 mouse in the pre-stim, stim (30 Hz, 10 ms, 120 s), and post-stim phases under 1.4% isoflurane anesthesia. (H) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of mCherry mouse in the pre-stim, stim (30 Hz, 10 ms, 120 s), and post-stim phases under 1.4% isoflurane anesthesia. (I) Statistical results showing the effect of 30Hz blue light stimulation on the BSR of ChR2 mice under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (J) Statistical results showing the effect of 30Hz blue light stimulation on the BSR of mCherry mice under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. Experimental data are expressed as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. BSR, burst suppression ratio; EEG, electroencephalogram; EMG, electromyogram; Stim, stimulation; SuM, supramammillary nucleus.

Deep anesthesia induced by isoflurane leads to burst suppression oscillations in the cortical EEG, and the BSR is a well-established parameter for monitoring the depth of anesthesia.^26,27 We examined changes in BSR after photostimulation (30 Hz, 10 ms, 120 s) of SuM glutamatergic neurons in deep anesthetic state induced by 1.4% isoflurane. Our results showed that photostimulation disrupted burst suppression oscillations, causing the EEG to transition to a low-voltage fast activity pattern (fig. 2G; supplemental video S3, https://links.lww.com/ALN/E42). After photostimulation ceased, the EEG gradually returned to burst suppression oscillations, similar to the prestimulation state. Quantitative analysis of BSR revealed a significant reduction during photostimulation in the ChR2 group (prestimulation vs. stimulation: 81.82 ± 7.83% vs. 44.53 ± 28.62%, n = 8, P = 0.002; fig. 2I), followed by a gradual recovery after photostimulation. In the mCherry group, photostimulation did not affect the burst suppression oscillations of the mice (fig. 2H; supplemental video S4, https://links.lww.com/ALN/E43), and no significant changes in BSR were observed (fig. 2J). These findings suggest that optogenetic activation of SuM glutamatergic neurons strongly enhances cortical activation during isoflurane anesthesia, leading to a marked decrease in EEG delta power and BSR.

Optogenetic Activation of SuM Glutamatergic Neurons Induces Physiologic Activation during Isoflurane Anesthesia

As a crucial physiologic parameter during anesthesia, pupil diameter reflects cortical activation and arousal levels and is positively correlated with emergence from general anesthesia.^28,29 To investigate the effects of activating SuM glutamatergic neurons on arousal levels, we measured the pupil dilation response elicited by photostimulation (30 Hz, 10 ms, 60 s) of these neurons under 1.4% isoflurane anesthesia. Our results showed a rapid pupil dilation after light stimulation, with peak dilation occurring around 30 s poststimulation (fig. 3C; supplemental video S5, https://links.lww.com/ALN/E44). After the cessation of light stimulation, pupil diameter gradually returned to prestimulation levels (fig. 3C). Statistical analysis indicated a significant increase in pupil diameter during stimulation compared to prestimulation (prestimulation vs. stimulation: 1.05 ± 0.08% vs. 1.95 ± 0.46%, P < 0.001), and the pupil diameter was restored after the end of stimulation (stimulation vs. poststimulation: 1.95 ± 0.46% vs. 1.39 ± 0.30%, n = 8, P = 0.006; fig. 3D). In the control group, pupil size did not significantly change during photostimulation (supplemental video S6, https://links.lww.com/ALN/E45).

Fig. 3.

Optogenetic activation of supramammillary nucleus glutamatergic neurons induces physiologic activation during isoflurane anesthesia. (A) Schematic diagram for measuring pupil diameter of mice during 1.4% isoflurane anesthesia. (B) Representative images of pupil during the pre-stim, stim, and post-stim stages. (C) Time courses of pupil changes in pre-stim, stim (30 Hz, 10 ms, 60 s), and post-stim stages under 1.4% isoflurane anesthesia. The thick line indicates the mean, and the shaded area indicates SD. The shadow indicates light stimulation. (D) Quantitative analysis of pupil size in the pre-stim, stim, and post-stim stages under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (E) Schematic diagram of the measurement of respiration rate of mice during 1.2% isoflurane anesthesia. (F) Representative graphs illustrating the change in respiratory rate after optogenetic stimulation. The shadow indicates light stimulation. (G) Effect of optogenetic activation of SuM glutamatergic neurons on the respiratory rate under 1.2% isoflurane anesthesia. The shadow indicates light stimulation. (H) Quantitative analysis of respiratory rate in the pre-stim , stim, and post-stim stages of 1.2% isoflurane anesthesia (n = 10). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (I) Schematic diagram of the measurement of heart rate and blood pressure of mice during 1.4% isoflurane anesthesia. (J to L) Quantitative analysis of HR (J), SBP (K), and DBP (L) in the pre-stim and stim stages during 1.4% isoflurane anesthesia (n = 8). Statistical analyses were performed using the Wilcoxon signed-rank test. Experimental data are expressed as means ± SD. *P < 0.05; **P < 0.01; ***P < 0.001. DBP, diastolic blood pressure; HR, heart rate; SBP, systolic blood pressure; Stim, stimulation.

Respiratory rate is a key physiologic parameters during general anesthesia and is strongly associated with emergence.³⁰ We measured the effects of photostimulation of SuM glutamatergic neurons on respiratory rate during isoflurane anesthesia. Our results showed that photostimulation (30 Hz, 10 ms, 60 s) quickly elevated the respiratory rate during 1.2% isoflurane anesthesia, with the respiratory rate reaching its peak around 30 s (fig. 3G). Statistical analysis indicated a significant acceleration of respiratory rate during stimulation (prestimulation vs. stimulation: 0.98 ± 0.08% vs. 1.57 ± 0.39%, P < 0.001), and the respiratory rate recovered after the end of stimulation (stimulation vs. poststimulation: 1.57 ± 0.39% vs. 1.01 ± 0.26%, n = 10, P < 0.001; fig. 3H).

Blood pressure and heart rate reflect cardiac function during general anesthesia and are strongly associated with emergence.^30,31 We assessed the effects of photostimulation (30 Hz, 10 ms, 60 s) of SuM glutamatergic neurons on blood pressure and heart rate under 1.4% isoflurane. Statistical analysis revealed that photostimulation significantly increased both systolic (prestimulation vs. stimulation: 81.25 ± 6.36 mmHg vs. 94.79 ± 7.79 mmHg, P = 0.008, n = 8; fig. 3K) and diastolic blood pressure (prestimulation vs. stimulation: 48.50 ± 5.54 mmHg vs. 60.62 ± 6.17 mmHg, n = 8, P = 0.008; fig. 3L) but did not significantly change the heart rate (fig. 3J). Taken together, these findings indicate that optogenetic activation of SuM glutamatergic neurons induces physiologic activation during isoflurane anesthesia, including increased pupil dilation, accelerated respiratory rate, and elevated blood pressure.

Optogenetic Activation of SuM Glutamatergic Neurons Facilitates Behavioral Emergence from Isoflurane Anesthesia

Considering that elevated cortical activity and physiologic parameters are often associated with behavioral emergence from general anesthesia, we further examined the role of SuM glutamatergic neurons in regulating behavioral changes during isoflurane anesthesia. Our results showed that photostimulation of SuM glutamatergic neurons potently induces behavioral responses and significantly increased arousal scores in the ChR2 group. Statistical analysis showed a significant increase in arousal scores in the ChR2 group after photostimulation (mCherry vs. ChR2: 0.50 ± 0.76 vs. 9.75 ± 0.46, n = 8, P < 0.001; fig. 4, A and B). Specifically, body movements (including limbs, head, and tail) and RORR were observed in all ChR2 mice (eight of eight), while crawling was observed in most ChR2 mice (six of eight mice; supplemental table S1, https://links.lww.com/ALN/E31; supplemental video S7, https://links.lww.com/ALN/E46). In the mCherry group, photostimulation did not induce obvious behavioral change (supplemental video S8, https://links.lww.com/ALN/E47).

Fig. 4.

Optogenetic activation of supramammillary nucleus glutamatergic neurons facilitates behavioral emergence from isoflurane anesthesia. (A) Effect of optogenetic stimulation (30 Hz, 10 ms, 60 s) of SuM glutamatergic neurons on the arousal score in mCherry and ChR2 mice (n = 8). Statistical analysis was performed using the Mann–Whitney rank sum test. (B) Pie chart showing the proportions of regaining righting reflex behavior after optogenetic stimulation in mCherry and ChR2 mice (n = 8). (C) Effect of optogenetic stimulation of SuM glutamatergic neurons on LORR time in mCherry and ChR2 groups mice under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (D) Effect of optogenetic stimulation of SuM glutamatergic neurons on RORR time in mCherry and ChR2 groups mice under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (E) Statistical results showing the effects of optogenetic activation of SuM glutamatergic neurons on the isoflurane concentration at which LORR occurred in mice (n = 8). Statistical analysis was performed using Wilcoxon signed-rank test. (F) Dose–response curves showing the effects of blue light stimulation on the EC₅₀ of LORR (n = 8). (G) Statistical results showing the effects of optogenetic activation of SuM glutamatergic neurons on the isoflurane concentration at which RORR occurred in mice (n = 8). Statistical analysis was performed using Wilcoxon signed-rank test. (H) Dose–response curves showing the effects of blue light stimulation on the EC₅₀ of RORR (n = 8). Experimental data are expressed as means ± SD. *P < 0.05; **P < 0.01; ***P < 0.001. LORR, loss of righting reflex; RORR, recovery of righting reflex; SuM, supramammillary nucleus.

We next assessed the effect of activating SuM glutamatergic neuron on induction time and emergence time under isoflurane anesthesia. The mice were exposed to 1.4% isoflurane for 30 min, and the time to LORR and the time to RORR were measured with and without photostimulation. Our results showed that photostimulation significantly shortened the RORR time from 171.40 ± 56.39 s to 59.88 ± 27.18 s (light off vs. light on, n = 8, P = 0.007; fig. 4D) in the ChR2 group, although it did not significantly change the LORR time (fig. 4C). We further examined the effect of SuM glutamatergic activation on isoflurane sensitivity by gradually adjusting the isoflurane concentration. Photostimulation significantly increased the isoflurane concentration required to induce LORR, from 0.79 ± 0.10% to 0.94 ± 0.07% (n = 8, P = 0.008; fig. 4E), and shifted the LORR dose–response curve to the right, with the EC₅₀ increasing from 0.75% (95% CI, 0.735 to 0.758%) to 0.90% (95% CI, 0.887 to 0.904%; light off vs. light on; fig. 4F). Photostimulation significantly increased the isoflurane concentration required to induce RORR, from 0.64 ± 0.10% to 0.79 ± 0.08% (n = 8, P = 0.008; fig. 4G), and shifted the RORR dose–response curve to the right, with the EC₅₀ increasing from 0.70% (95% CI, 0.673 to 0.739%) to 0.84% (95% CI, 0.805 to 0.863%; light off vs. light on; fig. 4H). Taken together, these findings indicate that optogenetic activation of SuM glutamatergic neurons facilitates behavioral emergence from isoflurane anesthesia.

Chemogenetic Activation of SuM Glutamatergic Neurons Enhances Cortical Activation during Isoflurane Anesthesia

To further investigate the role of SuM glutamatergic neurons in regulating isoflurane anesthesia, we used a chemogenetic approach to manipulate the activity of these neurons. AAV-CaMKII-hM3Dq-mCherry was injected into the SuM, and EEG/EMG electrodes were planted (fig. 5A). Robust expression of hM3Dq-mCherry was observed in the SuM after 4 weeks after virus transfection (fig. 5B). One hour after the administration of 1.0 mg/kg CNO, the mice were exposed to 0.8% isoflurane. During the anesthesia induction phase, CNO administration delayed the increase in EEG amplitude and the decrease in EMG activity compared to vehicle administration (fig. 5, C and D). EEG spectral analysis revealed that CNO administration significantly reduced delta power within 5 min after isoflurane administration (vehicle vs. CNO: 43.61 ± 8.00% vs. 30.68 ± 11.02%, n = 10, P < 0.001; fig. 5G). During the anesthesia maintenance phase, CNO administration produced a significant decrease in delta power (vehicle vs. CNO: 55.35 ± 6.92% vs. 43.41 ± 10.59%; P < 0.001) and an increase in theta power (vehicle vs. CNO: 21.95 ± 3.90% vs. 35.19 ± 12.88%, n = 10, P < 0.001; fig. 5H). During the anesthesia emergence phase, CNO administration accelerated the decrease of EEG amplitude and the increase of EMG activity compared to vehicle administration (fig. 5, E and F). EEG spectral analysis revealed that CNO administration significantly decreased delta power (vehicle vs. CNO: 44.10 ± 8.61% vs. 28.97 ± 9.76%, P < 0.001), while it increased alpha power (vehicle vs. CNO: 10.76 ± 2.66% vs. 18.05 ± 5.44%, n = 10, P = 0.04; fig. 5I) within 5 min after isoflurane discontinuation. We analyzed the normalized power density of EEG signals to further illustrate changes across different frequency bands during the induction, maintenance, and emergence phases (supplemental fig. S4, A to C, https://links.lww.com/ALN/E36).

Fig. 5.

Chemogenetic activation of supramammillary nucleus glutamatergic neurons induces cortical activation during isoflurane anesthesia. (A) Schematic of the injection of AAV-CaMKII-hM3Dq-mCherry into the SuM of C57BL/6J mice. (B) Representative image showing the expression of AAV-CaMKII-hM3Dq-mCherry in the SuM. (C to F) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) during the induction and emergence period after administering vehicle or CNO under 0.8% isoflurane anesthesia. (C) Effect of vehicle administration on EEG/EMG changes during the induction period. (D) Effect of vehicle administration on EEG/EMG changes during the emergence period. (E) Effect of CNO administration on EEG/EMG changes during the induction period. (F) Effect of CNO administration on EEG/EMG changes during the emergence period. The shadow indicates exposure to 0.8% isoflurane. (G) Quantitative analysis of the normalized power spectral density of the mice in the induction period after vehicle and CNO administration during 0.8% isoflurane anesthesia. The EEG signals from 0 to 5 min after the initiation of isoflurane anesthesia were analyzed (n = 10). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (H) Quantitative analysis of the normalized power spectral density of the mice in the maintenance period after vehicle and CNO administration during 0.8% isoflurane anesthesia. The EEG signals from 25 to 30 min after the initiation of isoflurane anesthesia were analyzed (n = 10). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (I) Quantitative analysis of the normalized power spectral density of the mice in the emergence periods after vehicle and CNO administration during 0.8% isoflurane anesthesia. The EEG signals from 0 to 5 min after the termination of isoflurane were analyzed (n = 10). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (J) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of vehicle treated mice under 1.4% isoflurane anesthesia. (K) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of CNO treated mice under 1.4% isoflurane anesthesia. (L) Quantitative analysis of the burst suppression ratio after administration of vehicle and the CNO under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using paired Student’s t tests. Experimental data are expressed as means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. BSR, burst suppression ratio; CNO, clozapine N-oxide; EEG, electroencephalogram; EMG, electromyogram; SuM, supramammillary nucleus.

Additionally, we investigated the effect of chemogenetic activation of SuM glutamatergic neurons under 1.4% isoflurane anesthesia. Our results showed that CNO administration significantly disrupted burst suppression oscillations (fig. 5, J and K) and reduced the BSR (vehicle vs. CNO: 74.15 ± 6.88% vs. 37.70 ± 10.12%, n = 8, P < 0.001; fig. 5L), indicating a decrease in anesthetic depth. Collectively, these findings indicate that chemogenetic activation of SuM glutamatergic neurons enhances cortical activation during isoflurane anesthesia.

Chemogenetic Activation of SuM Glutamatergic Neurons Facilitate Behavioral Emergence from Isoflurane Anesthesia

We further investigated the effect of chemogenetic manipulation of SuM glutamatergic neurons on behavioral responses during isoflurane anesthesia. The mice were administrated with 1.0 mg/kg CNO and exposed to 1.4% isoflurane 1 h later. Our results showed that CNO administration significantly prolonged the time to LORR in the hM3Dq group (vehicle vs. CNO: 58.71 ± 8.20 s vs. 70.57 ± 9.46 s, n = 14, P = 0.002; supplemental fig. S5A, https://links.lww.com/ALN/E37). For the anesthesia emergence phase, CNO administration significantly shortened the time to RORR (vehicle vs. CNO: 113.10 ± 66.47 s to 33.29 ± 20.78 s, n = 14, P = 0.004; supplemental fig. S5B, https://links.lww.com/ALN/E37). In the mCherry group, CNO administration did not significantly change the time to LORR or RORR (supplemental fig. S5, A and B, https://links.lww.com/ALN/E37). We assessed the effect of glutamatergic SuM activation on isoflurane sensitivity by gradually changing the isoflurane concentration. Our results showed that CNO administration significantly raised the isoflurane concentration required to induce LORR from 0.73 ± 0.10% to 0.84 ± 0.07% (n = 9, P = 0.008; supplemental fig. S5C, https://links.lww.com/ALN/E37) and shifted the LORR dose–response curve to the right, with the EC₅₀ increasing from 0.70% (95% CI, 0.688 to 0.718%) to 0.81% (95% CI, 0.795 to 0.813%; vehicle vs. CNO; supplemental fig. S5D, https://links.lww.com/ALN/E37). After CNO administration, RORR occurred in mice at higher isoflurane concentrations (vehicle vs. CNO: 0.59 ± 0.06% to 0.89 ± 0.08%, n = 9, P = 0.004; supplemental fig. S5E, https://links.lww.com/ALN/E37). The RORR dose–response curve shifted to the right, with the EC₅₀ increasing from 0.64% (95% CI, 0.636 to 0.639%) to 0.94% (95% CI, 0.929 to 0.945%; vehicle vs. CNO; supplemental fig. S5F, https://links.lww.com/ALN/E37). Collectively, these findings indicate that chemogenetic activation of SuM glutamatergic neurons facilitates behavioral emergence from isoflurane anesthesia.

Additionally, we accessed the effect of inhibiting SuM glutamatergic neurons on cortical activities and behavioral responses during isoflurane anesthesia. A mixture of AAV-CaMKII-Cre and AAV-hSyn-DIO-stGtACR2-eGFP, an inhibitory optogenetic viral vector, was injected into the SuM, and an optical fiber was planted above the SuM (supplemental fig. S6A, https://links.lww.com/ALN/E38). The effect of optogenetic inhibition of SuM glutamatergic neurons on cortical activities during isoflurane anesthesia was tested. After the mice were exposed to 1.0% isoflurane, the time for the occurrence of BS onset and BS termination was recorded according to EEG/EMG signals. Our results showed that blue light stimulation (30 Hz, 10 ms) significantly shortened the time to BS onset from 148.40 ± 35.47 s to 114.20 ± 24.75 s compared to control red light stimulation (n = 9, P = 0.02; supplemental fig. S6B, https://links.lww.com/ALN/E38), while it prolonged the time to BS termination from 104.00 ± 76.84 s to 210.70 ± 112.20 s (n = 9, P = 0.04; supplemental fig. S6C, https://links.lww.com/ALN/E38). Then, we tested the effect of chemogenetic inhibition of SuM glutamatergic neurons on behavioral responses during isoflurane anesthesia. An inhibitory chemogenetic viral vector, AAV-CaMKII-hM4Di-mCherry, was injected into the SuM of mice (supplemental fig. S6D, https://links.lww.com/ALN/E38). The mice were administrated with 1.0 mg/kg CNO and exposed to 1.4% isoflurane 1 h later. Our results show that CNO administration significantly prolonged the RORR time in the hM4Di group (vehicle vs. CNO: 128.00 ± 59.27 s to 312.60 ± 174.70 s, n = 14, P < 0.001; supplemental fig. S6F, https://links.lww.com/ALN/E38), although it did not significantly change the time to LORR (supplemental fig. S6E, https://links.lww.com/ALN/E38).

Optogenetic Activation of Glutamatergic SuM–MS Pathway Enhances Cortical Activation during Isoflurane Anesthesia

Neuroanatomical studies have shown that the glutamatergic SuM projects to several brain structures involved in general anesthesia.^32,33 Among these brain regions, the MS is particularly notable for its substantial projections from the glutamatergic SuM and its recently identified role in modulating general anesthesia.^34–36 Thus, we hypothesize that the MS may be involved in the regulation of isoflurane anesthesia by SuM glutamatergic neurons. To test our hypothesis, we examined the effect of activating the glutamatergic SuM–MS pathway on isoflurane anesthesia. AAV-CaMKII-ChR2-mCherry was injected into the SuM, and an optical fiber was implanted above the MS (fig. 6, A and B). We first examined the effect of optogenetic activation of this pathway on cortical activity during 0.8% isoflurane anesthesia. Our results showed that similar to the effect of opto-stimulating SuM soma, opto-stimulation of MS terminals rapidly induced cortical activation with a significant enhancement of EMG signals (fig. 6C). EEG spectral analysis indicated that light stimulation significantly decreased delta power (prestimulation vs. stimulation: 52.55 ± 4.81% vs. 36.86 ± 12.26%, n = 8, P = 0.01; fig. 6E) but did not significantly change the theta, alpha, and beta power. In the mCherry group, photostimulation did not obviously change EEG and EMG signals (fig. 6D) or significantly change EEG delta power (fig. 6F).

Fig. 6.

Optogenetic activation of glutamatergic supramammillary nucleus-medial septum pathway enhance cortical activation and facilitate behavioral emergence during isoflurane anesthesia. (A) Schematic of optogenetic activation of the glutamatergic SuM–MS pathway. (B) Schematic of coronal section illustrating the location of fiber optic implantation in the MS and terminal projection in the SuM. (C) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of ChR2 mouse in the pre-stim, stim, and post-stim phases of SuM–MS pathway optogenetic stimulation under 0.8% isoflurane anesthesia. (D) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of mCherry mouse in the pre-stim, stim, and post-stim phases of SuM–MS pathway optogenetic stimulation under 0.8% isoflurane anesthesia. (E) Statistical results of the EEG power of ChR2 mice in the pre-stim, stim, and post-stim phases under 0.8% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (F) Statistical results of EEG power of mCherry mice in the pre-stim, stim, and post-stim phases of SuM–MS pathway optogenetic stimulation under 0.8% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (G) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of ChR2 mouse in the pre-stim, stim, and post-stim phases of SuM–MS pathway optogenetic stimulation under 1.4% isoflurane anesthesia. (H) Statistical results showing the effect of optogenetic activation of SuM–MS pathway on the BSR of mCherry mice under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (I, top) Representative images of pupil during the pre-stim, stim, and post-stim stages. (Bottom) Quantitative analysis of pupil size under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (J, top) Representative graphs illustrating the change in respiratory rate after optogenetic stimulation. The shadow indicates light stimulation. (Bottom) Quantitative analysis of respiratory rate in the pre-stim, stim, and post-stim stages of 1.2% isoflurane anesthesia (n = 9). Statistical analysis was performed using one-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (K, L) Quantitative analysis of SBP (K) and DBP (L) in the pre-stim and stim stages of 1.4% isoflurane anesthesia (n = 11). Statistical analysis was performed using paired Student’s t tests (K) or Wilcoxon signed-rank test (L). (M) Effect of optogenetic stimulation (30 Hz, 10 ms, 60 s) of the glutamatergic SuM–MS pathway on the arousal score in mCherry and ChR2 mice (n = 8). Statistical analysis was performed using the Mann–Whitney rank sum test. (N) Pie chart showing the proportions of regaining righting reflex behavior after optogenetic stimulation in mCherry and ChR2 mice (n = 8). (O) Effect of optogenetic activation of the SuM–MS pathway on the LORR in mCherry and ChR2 mice under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (P) Effect of optogenetic activation of the SuM–MS pathway on the RORR in mCherry and ChR2 mice under 1.4% isoflurane anesthesia (n = 8). Statistical analysis was performed using two-way repeated-measure ANOVA followed by the Bonferroni post hoc test. (Q, right) Statistical results showing the effect of optogenetic activation of SuM–MS pathway on the concentration of isoflurane that induced LORR (n = 8). (Left) Dose–response curves showing the effects of no light or blue light stimulation on the EC₅₀ of LORR (n = 8). Statistical analyses were performed using Wilcoxon signed-rank test. (R, right) Statistical results showing the effect of optogenetic activation of SuM–MS pathway on the concentration of isoflurane that induced RORR (n = 9). (Left) Dose–response curves showing the effect of optogenetic activation of SuM–MS pathway on the EC₅₀ of RORR (n = 9). Statistical analysis was performed using paired Student’s t tests. Experimental data are expressed as means ± SD. *P < 0.05; **P < 0.01; ***P < 0.001. BSR, burst suppression ratio; DBP, diastolic blood pressure; Freq, frequency; LORR, loss of righting reflex; MS, medial septum; RORR, recovery of righting reflex; SBP, systolic blood pressure; SuM, supramammillary nucleus.

Then, we examined the effect of activating the SuM–MS pathway on anesthetic depth under 1.4% isoflurane. Our results showed that photostimulation of the MS terminals rapidly interrupted burst suppression oscillations and shifted the EEG to a high-frequency, low-amplitude pattern (fig. 6G). Statistical analysis revealed that photostimulation significantly reduced the BSR (prestimulation vs. stimulation: 72.97 ± 12.03% vs. 32.76 ± 8.17%, n = 8, P < 0.001; fig. 6H). In the mCherry group, photostimulation did not disrupted burst suppression oscillations (supplemental fig. S7A, https://links.lww.com/ALN/E39) or significantly change the BSR (supplemental fig. S7B, https://links.lww.com/ALN/E39). These results indicate that activating the glutamatergic SuM–MS pathway effectively enhances cortical activation during isoflurane anesthesia.

Optogenetic Activation of Glutamatergic SuM–MS Pathway Induces Physiologic Activation during Isoflurane Anesthesia

Next, we assessed the effect of activating the glutamatergic SuM–MS pathway on physiologic parameters during isoflurane anesthesia. We first assessed the effect of optogenetic activation of the SuM–MS pathway on pupil dilation. Our results show that photostimulation of the SuM–MS pathway induces a significant increase in pupil diameter, with this effect lasting for a long time (fig. 6I; supplemental fig. S7C, https://links.lww.com/ALN/E39). Statistical analysis demonstrated photostimulation significantly increased pupil diameter (prestimulation vs. stimulation: 1.03 ± 0.10% vs. 1.34 ± 0.18%, n = 8, P = 0.04; fig. 6I). Compared to photostimulation of the SuM soma, activation of the SuM–MS pathway induced a weaker pupil dilation, but the effect was sustained for a longer duration. Then, we examined the effect of activating the SuM–MS pathway on respiratory rate under 1.2% isoflurane anesthesia. Our results showed that the respiratory rate gradually increased and then returned to prestimulation levels during the photostimulation period (fig. 6J; supplemental fig. S7D, https://links.lww.com/ALN/E39). Statistical analysis revealed a significant increase in respiratory rate during the photostimulation period (prestimulation vs. stimulation: 0.96 ± 0.02% vs. 1.11 ± 0.12%, n = 9, P = 0.001; fig. 6J), although the increase was smaller compared to photostimulation of the SuM soma. Finally, we investigated the effect of activating SuM–MS pathway on blood pressure and heart rate during isoflurane anesthesia. Our results showed that photostimulation significantly increased the systolic (prestimulation vs. stimulation: 73.76 ± 9.22 mmHg vs. 78.76 ± 7.59 mmHg, n = 11, P = 0.006; fig. 6K) and diastolic blood pressure (prestimulation vs. stimulation: 43.94 ± 7.96 mmHg vs. 49.18 ± 7.54 mmHg, n = 11, P = 0.03; fig. 6L) but did not significantly affect heart rate (supplemental fig. S7E, https://links.lww.com/ALN/E39), which is consistent with the effects of photostimulation of the SuM soma. These findings indicate that optogenetic activation of the glutamatergic SuM–MS pathway induces physiologic activation during isoflurane anesthesia.

Optogenetic Activation of Glutamatergic SuM–MS Pathway Facilitates Behavioral Emergence from Isoflurane Anesthesia

We investigated the effect of activating the glutamatergic SuM–MS pathway on behavioral responses during isoflurane anesthesia. Our results demonstrated that light stimulation significantly increased the arousal scores in the ChR2 group (mCherry vs. ChR2: 0.50 ± 0.76 vs. 9.50 ± 0.76, n = 8, P < 0.001; fig. 6, M and N). Body movements were observed in all ChR2 mice (eight of eight mice), while RORR was observed in seven of eight mice, and crawling was observed in six of eight mice (supplemental table S2, https://links.lww.com/ALN/E33). We next analyzed the effects of SuM–MS pathway activation on the times of anesthesia induction and emergence. Photostimulation significantly prolonged the time to LORR (light off vs. light on: 53.25 ± 5.18 s vs. 63.25 ± 8.01 s, n = 8, P = 0.008; fig. 6O) but reduced the time to RORR (light off vs. light on: 137.30 ± 29.62 s vs. 47.00 ± 34.90 s, n = 8, P < 0.001; fig. 6P) in the ChR2 group but not in the mCherry control group. We further investigated the effect of SuM–MS pathway activation on isoflurane sensitivity. Photostimulation significantly increased the isoflurane concentration required to induce LORR, from 0.76 ± 0.07% to 0.95 ± 0.12% (n = 8, P = 0.02; fig. 6Q, left), and shifted the LORR dose–response curve to the right, with the EC₅₀ rising from 0.70% (95% CI, 0.695 to 0.713%) to 0.90% (95% CI, 0.883 to 0.917%, light off vs. light on; fig. 6Q, right). Photostimulation significantly increased the isoflurane concentration required to induce RORR, from 0.63 ± 0.10% to 0.82 ± 0.10% (n = 9, P = 0.004; fig. 6R, left), and shifted the RORR dose–response curve to the right, with the EC₅₀ rising from 0.70% (95% CI, 0.697 to 0.703%) to 0.87% (95% CI, 0.862 to 0.880%, light off vs. light on; fig. 6R, right). Collectively, these findings indicate that optogenetic activation of the glutamatergic SuM–MS pathway impedes anesthesia induction and facilitates emergence from isoflurane anesthesia.

Glutamatergic SuM–MS Pathway Regulates Isoflurane Anesthesia via MS Glutamatergic Neurons

Considering recent studies that highlight the crucial role of glutamatergic neurons in the MS in regulating general anesthesia,³⁴ we hypothesized that MS glutamatergic neurons may mediate the emergence-promoting effects of the SuM–MS pathway during isoflurane anesthesia. To test our hypothesis, we first examined the correlation between the activity of SuM glutamatergic neurons and MS glutamatergic neurons during isoflurane anesthesia. AAV-CaMKII-GCaMP6m was injected into both the SuM and MS, with fiber optics implanted above these regions (fig. 7A). Four weeks after transfection, GCaMP6m expression was confirmed in both regions (fig. 7B). Calcium signals of SuM and MS glutamatergic neurons were simultaneously monitored with multichannel fiber photometry. Our results showed that calcium signals of MS glutamatergic neurons decreased obviously during isoflurane exposure and increased obviously after exposure (fig. 7C). The changes in calcium signals of MS glutamatergic neurons closely resembled those of SuM glutamatergic neurons (fig. 7C).

Fig. 7.

The glutamatergic supramammillary nucleus–medial septum pathway regulates isoflurane anesthesia via medial septum glutamatergic neurons. (A) Schematic illustrating the simultaneous recording of calcium signals in the SuM and MS glutamatergic neurons by multichannel fiber photometry. (B) Representative image of GCaMP6m immunofluorescence in SuM and MS. (C) Time courses of calcium signals in the SuM and MS during 1.4% isoflurane anesthesia (n = 9 trails from 7 mice). The thick lines indicate the mean, and the area of the shadow indicates SD. (D) Time courses of calcium signals in the SuM and MS during isoflurane administration period (n = 12 trails from 7 mice). Time 0 s represents the moment of isoflurane administration, and the shaded area indicates SD. (E) Pearson correlation results between the calcium signals of SuM and MS glutamatergic neurons within 5 s after isoflurane administration (n = 12 trails from 7 mice). (F) Time courses of calcium signals in the SuM and MS during BS onset period (n = 12 trails from 7 mice). Time 0 s represents the moment of BS onset, and the shaded area indicates SD. (G) Pearson correlation results between the calcium signals of SuM and MS glutamatergic neurons within 5 s after the occurrence of BS onset (n = 12 trails from 7 mice). (H) Time courses of calcium signals in the SuM and MS during isoflurane termination period (n = 12 trails from 7 mice). Time 0 s represents the moment of isoflurane termination, and the shaded area indicates SD. (I) Pearson correlation results between the calcium signals of SuM and MS glutamatergic neurons within 5 s after isoflurane termination (n = 12 trails from 7 mice). (J) Time courses of calcium signals in the SuM and MS during BS termination period (n = 12 trails from 7 mice). Time 0 s represents the moment of BS termination, and the shaded area indicates SD. (K) Pearson correlation results between the calcium signals of SuM and MS glutamatergic neurons within 5 s after the occurrence of BS termination (n = 12 trails from 7 mice). (L) Schematic illustrating the recording of calcium signals from MS glutamatergic neurons during optogenetic activation of the glutamatergic SuM–MS pathway. (M) Representative EEG/EMG traces (top) and EEG power spectrograms (bottom) of mice before, during, and after optogenetic stimulation (30 Hz, 10 ms, 20 s) during 1.4% isoflurane. (N, O) Heat maps (N) and mean curve (O) of calcium signals in MS glutamatergic neurons before, during, and after optogenetic stimulation during 1.4% isoflurane (n = 18 trails from 7 mice). Red light stimulation is given at 0 s. Correlation analyses were performed using the Pearson correlation coefficient (E, G, I, and K). Experimental data are expressed as means ± SD. BS, burst suppression; EEG, electroencephalogram; EMG, electromyogram; Iso, isoflurane; MS, medial septum; SuM, supramammillary nucleus.

We further analyzed the correlation between the calcium signals of SuM and MS glutamatergic neurons during four periods: isoflurane administration (Iso on), BS onset, isoflurane termination (Iso off), and BS termination period (fig. 7, D, F, H, and J). The analysis revealed a strong correlation in calcium signals between SuM and MS glutamatergic neurons during isoflurane administration, BS onset, and BS termination periods and a moderate correlation during the isoflurane termination period (fig. 7, E, G, I, and K). The correlation coefficients were 0.9970 during the isoflurane administration period (P < 0.001; fig. 7E), 0.9347 during the BS onset period (P < 0.001; fig. 7G), 0.6428 during the isoflurane termination period (P < 0.001; fig. 7I), and 0.9793 during the BS termination period (P < 0.001; fig. 7K).

We further tested the effects of activating the glutamatergic SuM–MS pathway on the activity of MS glutamatergic neuron during isoflurane anesthesia, combining fiber photometry and an optogenetic approach. A mixture of AAV-CaMKII-Cre and AAV-hSyn-DIO-ChrimsonR-mCherry was injected into the SuM, and AAV-CaMKII-GCaMP6m was injected into the MS. Fiber optics were implanted above the MS for photostimulation and fiber photometry recording (fig. 7L). Acute red light stimulation (30 Hz, 10 ms, 20 s) of the MS immediately interrupted burst suppression oscillations and induced high-frequency EEG activity (fig. 7M), indicating the activation of glutamatergic SuM–MS pathway by ChrimsonR protein. Heat map analysis revealed a significant increase in calcium signals in MS glutamatergic neurons during the light stimulation period (fig. 7N), with statistical analysis showing an approximately 15% increase in calcium signals due to the activation of the SuM–MS pathway (fig. 7O). These findings suggest that MS glutamatergic neurons mediate the emergence-promoting effects of the glutamatergic SuM–MS pathway during isoflurane anesthesia.

Discussion

Our current study demonstrates that SuM glutamatergic neurons play an important role in regulating isoflurane anesthesia and facilitate anesthesia emergence via MS glutamatergic neurons. Specifically, the activity of SuM glutamatergic neurons closely correlates with isoflurane anesthesia, decreasing during induction and recovering during emergence. Optogenetic and chemogenetic manipulation verified the regulatory effects of SuM glutamatergic neurons on cortical activation, physiologic parameters, and behavioral responses during isoflurane anesthesia. Moreover, we identified the emergence-promoting downstream targets of SuM glutamatergic neurons, namely, the MS. Finally, by combining optogenetics and fiber photometry, we showed that photostimulation of glutamatergic SuM–MS pathway significantly increases the activity of MS glutamatergic neurons.

An increasing body of evidence suggests that neural substrates regulating sleep–wake behavior are involved in the regulation of general anesthesia, such as the basal forebrain,³ ventral tegmental area,³⁷ and parabrachial nucleus.⁶ Previous studies have indicated the functional role of SuM glutamatergic neurons in sleep–wake regulation.^8,38 Chemogenetic activation of SuM glutamatergic neurons produced sustained behavioral arousal with increases in theta and gamma EEG activity, while chemogenetic inhibition led to fragmentation of wakefulness and somnolence in mice.⁸ In the current study, our results demonstrated that the SuM glutamatergic neurons are involved in the regulation of general anesthesia. The activity of SuM glutamatergic neurons closely correlates with isoflurane anesthesia, and activation of these neurons potently facilitated behavioral emergence in anesthetized mice. Our results further support the hypothesis of overlapping neural circuits between general anesthesia and natural sleep, suggesting that general anesthetics exert their hypnotic effects through shared circuits involved in sleep–wake regulation. Furthermore, our results indicated that SuM glutamatergic neurons serve as a key neural substrate in modulating anesthesia depth. Optogenetic activation of SuM glutamatergic neurons enhanced cortical activity during isoflurane anesthesia, reducing EEG delta power and BSR at 0.8% and 1.4% isoflurane concentrations, respectively. General anesthesia can affect physiologic parameters, such as blood pressure and respiration, potentially leading to adverse reactions, which can be mitigated by reducing the depth of anesthesia. Our results showed that optogenetic activation of SuM glutamatergic neurons significantly altered physiologic parameters in anesthetized mice, such as enlarged pupil size, accelerated respiratory rate, and elevated blood pressure. Our findings suggest that the glutamatergic SuM may be a potential target for accelerating emergence from general anesthesia and improving its safety in clinical practice.

In the current study, our results demonstrate that SuM glutamatergic neurons are a key neural substrate regulating isoflurane anesthesia and facilitating emergence through their projections to the MS. Optogenetically activating the SuM–MS pathway enhanced cortical activity, altered several physiologic parameters including enlarged pupil diameter, increased respiratory rate and elevated blood pressure, and induced behavioral responses, which mimicked the effects of activating SuM glutamatergic soma. This canonical SuM–MS pathway serves as an important indirect bypass from the SuM to the hippocampus and has been shown to regulate various physiologic functions, including arousal. In addition to this canonical pathway, neuroanatomical evidence suggests the presence of a noncanonical antidromic pathway projecting from the MS to the SuM.^39,40 Interestingly, recent studies report that this noncanonical pathway also facilitates emergence from general anesthesia, similar to the canonical SuM–MS circuit.⁴¹ Wu et al.⁴¹ showed that optogenetic activation of this MS–SuM pathway induced cortical activation during sevoflurane anesthesia. These structural and functional findings suggest the existence of a positive feedback loop between the canonical SuM–MS pathway and the noncanonical MS–SuM pathway. We speculate that this positive feedback loop could rapidly enhance the excitability of both the SuM and MS during anesthetic emergence, potentially serving as a key mechanism underlying the rapid transition from the anesthetized to the awake state.

The MS is primarily composed of three types of neurons: glutamatergic, cholinergic, and GABAergic neurons.⁴² A recent study demonstrated that MS glutamatergic neurons are involved in the regulation of general anesthesia.³⁴ Fiber photometry recordings showed that the activity of MS glutamatergic neurons decreased during sevoflurane anesthesia induction and recovered during emergence.³⁴ Chemogenetic activation of MS glutamatergic neurons prolonged the induction time and decreased the emergence time, while lesions or chemogenetic inhibition of these neurons produced the opposite effects.³⁴ Moreover, MS glutamatergic neurons have been shown to receive projections from SuM glutamatergic neurons.⁴³ Optogenetic activation of SuM glutamatergic neurons projecting to MS glutamatergic neurons reinforced the motivation for behavioral interaction with environment in mice.⁴³ In the current study, our results showed that the calcium signals of SuM glutamatergic neurons were synchronized with those of MS glutamatergic neurons during the isoflurane anesthesia induction and emergence. Additionally, activation of the glutamatergic SuM–MS pathway increased calcium signals in MS glutamatergic neurons during isoflurane anesthesia. Based on the experimental results above and our findings in the current study, we hypothesize that SuM glutamatergic neurons regulate general anesthesia via their projections to MS glutamatergic neurons. It is worth noting that, in addition to glutamatergic neurons, the MS also contains a significant number of cholinergic and GABAergic neurons.⁴² In vivo electrophysiologic recordings showed that the firing of MS cholinergic and GABAergic neurons is closely associated with general anesthesia.³⁶ The study of Tai et al.³⁵ showed that MS cholinergic neurons modulate sensitivity to isoflurane anesthesia, and lesioning these neurons facilitated anesthesia induction and delayed emergence in rats. Two-photon calcium imaging of axonal boutons revealed that MS GABAergic boutons in the hippocampal CA1 were activated during salient sensory events and that sensory responses were abolished by anesthesia.⁴⁴ The modulation of MS GABAergic and cholinergic neurons by SuM glutamatergic neurons, as well as their role in the regulation of general anesthesia, remains to be further elucidated.

In this study, our results indicate that SuM glutamatergic neurons facilitate anesthesia emergence via their projections to the MS. However, the effect of activating the glutamatergic SuM–MS pathway does not fully replicate the effect of activating glutamatergic SuM. Activating the glutamatergic SuM–MS pathway resulted in a weaker effect on physiologic parameters, including pupil dilation, respiratory rate, and blood pressure, compared to activating glutamatergic SuM neurons. This suggests that SuM glutamatergic neurons may regulate anesthesia emergence through additional pathways. In fact, in addition to the MS, SuM glutamatergic neurons also project to several brain structures involved in the regulation of general anesthesia, such as the basal forebrain, parabrachial nucleus, and locus coeruleus.¹⁰ Among these brain structures, the basal forebrain has recently been shown to play an important role in regulating the process of general anesthesia.^3,45 Optogenetic activation of GABAergic neurons in the basal forebrain has been shown to facilitate behavioral and cortical emergence from isoflurane anesthesia.³ Chemogenetic activation of cholinergic neurons in the basal forebrain significantly prolonged anesthesia induction, shortened anesthesia emergence, and reduced the EEG delta power during general anesthesia.⁴⁶ It is likely that the glutamatergic SuM–basal forebrain pathway may also contribute to the regulation of anesthesia emergence by SuM glutamatergic neurons. We hypothesize that SuM glutamatergic neurons may regulate general anesthesia through multiple downstream pathways, thereby exerting a strong influence on promoting anesthesia emergence, although the specific mechanisms underlying this regulation remain to be further elucidated.

Numerous studies have shown that general anesthesia can lead to cognitive dysfunction in patients.^47–49 Moller et al.⁵⁰ found that approximately a quarter of older patients developed postoperative cognitive dysfunction within 1 week of surgery, with around 10% exhibiting cognitive dysfunction 3 months postsurgery. Hou et al.⁵¹ showed that the incidence of postoperative cognitive dysfunction in elderly patients is related to anesthetic depth, with deep anesthesia resulting in a higher incidence compared to light anesthesia. In addition, studies have shown that general anesthesia also cause memory impairment.^52,53 A clinical study of 10,149 middle-aged Chinese individuals found that the incidence of memory impairment was 57.7% at 7 days after cardiac surgery with general anesthesia and 58.9% 12 months later.⁵² However, the specific neural mechanisms underlying cognitive dysfunction and memory impairment caused by general anesthesia remain unclear. Neuroanatomical evidence indicates that SuM glutamatergic neurons project extensively to the dentate gyrus and the CA2 region of the hippocampus.¹¹ Recent studies indicate that the activity of SuM glutamatergic neurons is closely associated with memory and cognition.^54,55 Chen et al.⁵⁴ reported that dentate gyrus-projecting SuM neurons are activated by contextual novelty, while CA2-projecting SuM neurons are preferentially activated by novel social encounters. In addition, optogenetic manipulation of the SuM–dentate gyrus and SuM–CA2 pathways modified hippocampal contextual and social memory, respectively.⁵⁴ Li et al.⁵⁵ demonstrated that chemogenetic activation of SuM glutamatergic neurons increased calcium activity in dentate gyrus granule cells, while inhibition of these neurons reduced c-Fos expression in dentate gyrus granule cells and impaired spatial memory retrieval in mice. In the current study, our results demonstrated that the activity of SuM glutamatergic neurons was significantly suppressed during general anesthesia. Therefore, we speculate that the inhibition of SuM glutamatergic neurons could be a critical mechanism contributing to cognitive and memory impairments after general anesthesia. Selective activation of SuM glutamatergic neurons may offer a potential strategy for mitigating cognitive and memory impairments after general anesthesia.

Research Support

Supported by grant Nos. 82471503 (to Dr. Chen) and 82271529 (to Dr. Cai) from the National Natural Science Foundation of China, Beijing, China; by grant Nos. 2024Y9225 (to Dr. Cai) and 2021Y9005 (to Dr. Chen) from the Joint Funds for the Innovation of Science and Technology in Fujian Province, Fuzhou, China; and by grant Nos. 2024J01582 (to Dr. Cai) and 2025J01701 (to Dr. Chen) from the Natural Science Foundation of Fujian Province, Fuzhou, China.

Competing Interests

The authors declare no competing interests.

Supplemental Digital Content

Supplemental Digital Content 1. Table S1. Arousal scores of SuM opto-stimulation, https://links.lww.com/ALN/E31

Supplemental Digital Content 2. Figure S1. Representative EEG/EMG signals synchronized with calcium signal dynamics, https://links.lww.com/ALN/E32

Supplemental Digital Content 3. Table S2. Arousal scores of SuM–MS opto-stimulation, https://links.lww.com/ALN/E33

Supplemental Digital Content 4. Figure S2. Activity change of the SuM glutamatergic neurons in female mice, https://links.lww.com/ALN/E34

Supplemental Digital Content 5. Figure S3. EEG power density of SuM opto-stimulation, https://links.lww.com/ALN/E35

Supplemental Digital Content 6. Figure S4. EEG power density of SuM chemogenetic activation, https://links.lww.com/ALN/E36

Supplemental Digital Content 7. Figure S5. Chemogenetic activation of SuM glutamatergic neurons, https://links.lww.com/ALN/E37

Supplemental Digital Content 8. Figure S6. Optogenetic and chemogenetic inhibition of SuM glutamatergic neurons, https://links.lww.com/ALN/E38

Supplemental Digital Content 9. Figure S7. Optogenetic activation of glutamatergic SuM–MS pathway induces physiologic activation, https://links.lww.com/ALN/E39

Supplemental Video File 1. EEG signals of SuM ChR2 mouse, https://links.lww.com/ALN/E40

Supplemental Video File 2. EEG signals of SuM control mouse, https://links.lww.com/ALN/E41

Supplemental Video File 3. BSR of SuM ChR2 mouse, https://links.lww.com/ALN/E42

Supplemental Video File 4. BSR of SuM control mouse, https://links.lww.com/ALN/E43

Supplemental Video File 5. Pupil dilation rate of SuM ChR2 mouse, https://links.lww.com/ALN/E44

Supplemental Video File 6. Pupil dilation rate of SuM control mouse, https://links.lww.com/ALN/E45

Supplemental Video File 7. Arousal score of SuM ChR2 mouse, https://links.lww.com/ALN/E46

Supplemental Video File 8. Arousal score of SuM control mouse, https://links.lww.com/ALN/E47

Supplementary Material

aln-143-944-s002.pdf ^{(88.7KB, pdf)}

aln-143-944-s003.pdf ^{(611.3KB, pdf)}

aln-143-944-s004.pdf ^{(89.4KB, pdf)}

aln-143-944-s005.pdf ^{(489.1KB, pdf)}

aln-143-944-s006.pdf ^{(315.1KB, pdf)}

aln-143-944-s007.pdf ^{(230.9KB, pdf)}

aln-143-944-s008.pdf ^{(404.8KB, pdf)}

aln-143-944-s009.pdf ^{(435.7KB, pdf)}

aln-143-944-s010.pdf ^{(516.6KB, pdf)}

Download video file ^{(33MB, mp4)}

Download video file ^{(19.7MB, mp4)}

Download video file ^{(30.9MB, mp4)}

Download video file ^{(25.9MB, mp4)}

Download video file ^{(17.2MB, mp4)}

Download video file ^{(19.6MB, mp4)}

Download video file ^{(17.2MB, mp4)}

Download video file ^{(19.6MB, mp4)}