2.1 Animals

This study was performed in accordance with the Guide for the Care and Use of Laboratory Animals of China (no. 14924, 2001). All animal procedures were designed and conducted according to the guidelines set forth by the Zunyi Medical University. Male C57BL/6J mice (7–15 weeks old, weighing 20–30 g) were used for all experiments and purchased from Jiangsu Huachuang Cigna Medical Technology Co., Ltd. (Taizhou, Jiangsu, China). All animals were housed in the Laboratory Animal Center of Zunyi Medical University, with a constant temperature of 24°C (±1°C) and a relative humidity of 55% (±2%) on a 12:12 h light/dark cycle (lights on at 6:00 am), while having free access to food and water. To minimize the impact of room temperature on experimental results, all behavioral and electroencephalogram (EEG) tests were conducted at a controlled temperature of 24°C (±2°C).

2.2 Antibodies and reagents

The primary antibodies against MBP (MAB386, RRID: AB_94975), and NeuN (MAB377, RRID: AB_2298772) were obtained from Millipore (Billerica, MA, USA). c-Fos antibody (226008, RRID: AB_2891278) was from Synaptic Systems (Goettingen, Germany). 5-HT antibody (ab211528, RRID: AB_3344910) was from Abcam (Cambridge, UK). Alexa Fluor-conjugated secondary antibodies (A-11039, RRID: AB_142924; A-11008, RRID: AB_143165; and A-11005, RRID: AB_141372) were purchased from Invitrogen (Carlsbad, CA, USA). Lysolecithin (LPC, L4129) was from Sigma (TUCSON, AZ, USA). Ibotenic acid (IBO, HY-N2311) was purchased from MedChemExpress LLC (USA).

2.3 Surgery

The mice were anesthetized with isoflurane throughout, secured in a stereotaxic apparatus (RWD Life Science Co. Ltd., Shenzhen China), and their eyes were covered with erythromycin eye ointment. The scalp was cut to expose the skull, and the level of the skull was adjusted until leveled. According to the mouse brain stereotaxic atlas (Watson and Paxinos), the Bregma serves as the origin with mm as the unit, and specific brain regions are located using the following stereotaxic coordinates [anterior–posterior (AP), medial-lateral (ML), dorsal-ventral (DV)]: the lateral habenula (LHb; AP—1.7, ML ± 0.45, DV—2.85), the paraventricular nucleus (PVT; AP—1.3, ML ± 0.0, DV—3.0), the dorsal raphe nucleus (DRN; AP—4.6, ML ± 0.0, DV—3.05). Use a microsyringe pump (Oran Science and Technology Co., Ltd., USA) to inject drugs or viruses through a glass microelectrode (Taimeng Software Co., Ltd., Chengdu, Sichuan, China). After injection completed, retain the microelectrode at the injection site for 10 min to ensure full diffusion of the drugs or the viruses before gradually withdrawn.

For LPC induced demyelination experiments, LPC (10 mg/mL) was injected into LHb (0.5 μL/side), PVT (1 μL) and DRN (1 μL) at a rate of 100 nL / min to generate the LPC induced demyelination model (LPC group), and the same dose of normal saline (NS) was injected into control group (control group). After drug injection, four screws were inserted into the skull (AP: + 1.0 mm, ML: ± 1.5 mm; AP: − 3.5 mm, ML: ± 1 mm). The EEG electrodes were placed and secured with dental cement and glue. Behavioral experiments and EEG recording were conducted after 7 days.

For neuronal lesion experiments, IBO was injected into LHb (100 nL/side), PVT (200 nL) and DRN (200 nL) at a rate of 20 nL/min to establish the IBO group, and the same dose of NS was injected into the control group. Behavioral experiments were conducted after 7 days.

For in vivo fiber calcium imaging experiment in the demyelinated mice, LPC (10 mg/mL) was injected into LHb (0.5 μL/side), PVT (1 μL) and DRN (1 μL) at a rate of 100 nL/min, and the same dose of NS was injected into the control group. Then the GCaM6s viruses were injected into LHb (50 nL), PVT (80 nL), and DRN (80 nL), respectively (LHb and PVT: rAAV-CaMKIIα-GCaM6s; DRN: rAAV-hsyn-SV40-NLS-Cre and rAAV-EF1α-DIO-GCaMP6s) (BrainVTA Co., Ltd., Wuhan, Hubei, China). Finally, the optical fiber (Inper Ltd, Hangzhou, Zhejiang, China) was inserted 0.1 mm shallower than the injection sites. Calcium recordings were carried out after 3 weeks.

For chemogenetic activation experiment in the demyelinated mice, LPC (10 mg/mL) was injected into LHb (0.5 μL/side), PVT (1 μL) and DRN (1 μL) at a rate of 100 nL/min to generate the demyelination model, and the same dose of NS was injected into the control group. Then the hM3D(Gq) viruses were injected into LHb (50 nL), PVT (80 nL), and DRN (80 nL), respectively (LHb and PVT: rAAV-CaMKIIα-hM3D(Gq); DRN: rAAV-hsyn-SV40-NLS-Cre and rAAV-EF1α-DIO- hM3D(Gq)) (BrainVTA Co., Ltd., Wuhan, Hubei China). Behavioral experiments were conducted after 3 weeks.

2.4 Cuprizone (CPZ) induced demyelination

For effective demyelination, 7-week-old C57BL/6J male mice were fed with 0.3% CPZ (C9012, Sigma, TUCSON, AZ, USA) diet for 4 consecutive weeks to establish a demyelination model. The control groups were fed with routine diet.

2.5 Behavioral tests of isoflurane anesthesia

The time taken from the initiation of anesthesia to loss of righting reflex (LORR) and from the termination of anesthesia to recovery of righting reflex (RORR) were utilized for quantifying the induction and the emergence time of anesthesia, respectively. The mice were gently placed in an induction box (RWD Life Science Co. Ltd., Shenzhen China), acclimated for 5 min, and then the inhalation of 1 L/min of 1% or 1.4% isoflurane with 100% O₂ was initiated to measure the LORR time. After a 20-min maintenance of isoflurane inhalation, the isoflurane administration was discontinued and the RORR time was recorded. An anesthesia monitor (Vamos; Drager Company, Germany) was connected to detect the concentration of isoflurane in the anesthesia chamber and an electric blanket with a rectal temperature probe was used to the bottom of the anesthesia chamber and was controlled at 37.5°C in the whole experiment. The mice were removed after recording, then the induction box was cleaned with 75% alcohol to remove odors. All mice were placed in the behavioral room at least 1 h in advance.

2.6 EEG recording

Electroencephalogram signals were collected using an Apollo II multi-channel neural signal data acquisition system (Bio-Signal Technologies, Nanjing, Jiangsu, China). EEG signals were recorded for 5 min before induction and were continuously recorded until recovery from isoflurane anesthesia. Then, the Spike2 software (Cambridge Electronic Design Limited Technical Centre, Cambridgeshire, England, UK) was used to analyze the percentage of power in different frequency bands of EEG during different periods. The frequency bands can be divided as follows: 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. Finally, the NeuroExplorer software (Nex Technology, Littleton, USA) was utilized to generate EEG power spectrogram.

2.7 Calcium fiber photometry

Fluorescence signal of GCaMP6s was acquired using a dual-color fiber recording device (Inper Ltd, Hangzhou, Zhejiang, China) with a sampling frequency of 100 Hz, an exposure time of 10 ms, and a gain setting of 0. The 410 nm channel was used as the reference channel, while the 470 nm channel was utilized for recording calcium signals. The fiber output power of both channels was regulated within the range of 10–20 μW and 20–40 μW, respectively. The mice were placed in the dark experimental room for a one-h acclimation period prior to the experiment. Following 5 min of baseline recording after adaption in the induction box, isoflurane was continuously delivered for 10 min. The recording was continued for another 5 min after turning off the isoflurane. The data were analyzed by InperDataProcessV0.7.2 software (Inper Ltd., Hangzhou, Zhejiang, China). We first preprocessed the original data, including baseline correction and marker increase and decrease before further analyzing the data to obtain the heat map and linear map. Furthermore, ∆F/F was analyzed [(F − F0)/F = ∆F/F, was the change of original fluorescence signal (F) relative to F0; F, original fluorescence signal; F0, the mean value of original fluorescence signal]. The ΔF/F were then compared between the LPC and Control groups during each period.

2.8 Immunohistochemistry

After all tests completed, mice were anesthetized, sacrificed and fixed. The heart was exposed, and the blood was washed out by rapid perfusion through the heart with phosphate-buffered saline (PBS). When the liver turned white, 4% Paraformaldehyde (PFA) was continuously injected for fixation. After perfusion, the whole brain was dissected out and soaked in 4% PFA for 24 h. Then the brain was dehydrated by 30% sucrose. Brain tissues were frozen in OCT and sectioned with a thickness of 30 μm using a cryostat microtome (Leica CM1900, Germany). The expression of MBP, NeuN, 5-HT, and calcium signal virus in brain slices were observed using immunofluorescence staining. The main procedures were as follows: brain slices were blocked with blocking buffer (10% goat serum, 1% BSA, and 0.3% Triton in PBS) for 2 h at room temperature. The primary antibodies were incubated overnight at 4°C. Then the secondary antibodies were then incubated for 2 h at room temperature in the dark. The dilution ratios of primary antibodies were 1:500 for MBP; 1:200 for NeuN and 5-HT; and 1:1000 for c-Fos and the secondary antibodies were diluted at 1:500. Finally, the nuclei were stained with DAPI Fluoromount-G Reagent (SBA-0100-20, Southern Biotech, Birmingham, AL, USA). Immunofluorescence images were captured using a fluorescence microscope (Olympus, Tokyo, Japan) and subsequently analyzed with ImageJ 1.42q (National Institutes of Health, Bethesda, MD, USA).

2.9 Statistical analysis

Statistical analysis was performed using SPSS 29.0 (IBM Corp., Armonk, NY, USA, Version 29.0), and the data are presented as mean ± SD. The two groups were compared using an unpaired Student's t-test. The effect of LPC-induced demyelination on neuron activation by chemogenetics during isoflurane anesthesia was analyzed by two-way ANOVA. GraphPad Prism 8.0 software (GraphPad Software Inc., San Diego, CA, USA) was used for statistical chart production after data analysis. p < 0.05 was considered statistically significant, and “n” represents the number of animals used in the corresponding experiment.

3.1 Cuprizone-induced demyelination delays the emergence from isoflurane anesthesia

To investigate the role of myelin in isoflurane anesthesia, we first induced demyelination in the CNS of C57BL/6J male mice through continuous feeding of a diet containing cuprizone (CPZ) (Figure 1A). After 4 weeks, MBP expression was significantly reduced in the CPZ group compared to the control group, which indicates successful establishment of the demyelination model (Figure 1B,C). Behavioral tests were conducted under 1.4% isoflurane anesthesia (Figure 1D). The results showed a prolonged RORR time in the CPZ group compared to the control group, while the LORR time was comparable between the CPZ and control groups (Figure 1E). These data indicated that myelin destruction in the CNS may affect the process of isoflurane anesthesia.

3.2 LHb demyelination promotes the emergence from isoflurane anesthesia

Recent research has revealed that myelin in specific nuclei regulate higher-order brain functions.³⁵ However, whether myelin in the distinct nucleus is involved in the regulation of isoflurane anesthesia remains unknown. LHb was reported to promote the induction of general anesthesia.³⁶ So, we wonder that myelin in the LHb can participate in the progression of isoflurane anesthesia. We bilaterally injected LPC (the LPC group) or saline (the control group) into the LHb to generate demyelination in the LHb (Figure 2A). After 1 week, dramatically decreased MBP expression in the LHb of the LPC group indicated the myelin destruction in the LHb (Figure 2C,D). In the behavior test of 1.4% isoflurane anesthesia, we found that the RORR time in the LPC group was shorter compared to the control group, while no significant difference was observed in the LORR time (Figure 2B,E). These findings suggest that demyelination in the LHb promotes the emergence from isoflurane anesthesia. The EEG results were consistent with the behavioral results. Compared to the control group, the power in the theta band was decreased and the power in the beta band was increased in the LPC group during RORR (Figure 2F,G). Notably, the power of theta band in the LPC group was lower than that of the control group during the WAKE phase prior to anesthesia (Figure 2G). Similarly, analysis of both raw EEG traces and EEG power spectrograms indicated higher levels of activity and wakefulness in the LPC group compared to the control group (Figure 2F). Additionally, we performed the behavior test under 1% isoflurane anesthesia and found that demyelination in the LHb also promoted the emergence from 1% isoflurane anesthesia (Figure S1A,B,E).

Next, we compared the effects of myelin damage and neuron lesion in the LHb on isoflurane anesthesia. IBO (the IBO group) or saline (the control group) was bilaterally injected into the LHb region to kill the LHb neurons (Figure 2H,I). During 1.4% isoflurane anesthesia, the LORR time was prolonged and the RORR time was shortened in the IBO group (Figure 2J). This suggests that depletion of LHb neurons inhibits the anesthetic effect of isoflurane. Taken together, we conclude that both demyelination and neuronal depletion in the LHb contribute to emergence from anesthesia.

3.3 PVT demyelination facilitates the induction of isoflurane anesthesia

Next, we selected the PVT, which has been demonstrated to plays a crucial role in facilitating recovery from anesthesia,⁹ to study whether myelin in the PVT is involved in the regulation of isoflurane anesthesia. LPC was injected into the PVT to induce the demyelination. MBP expression in the PVT was reduced in the LPC group compared to the control group (Figure 3A,B). The behavioral tests of 1.4% isoflurane anesthesia showed the shortened LORR time in the LPC group and comparable RORR time between the LPC group and the control group (Figure 3C). The EEG recording revealed that the LPC group exhibited a higher power of the delta band compared to the control group during LORR, suggesting an enhanced effect of isoflurane anesthesia following PVT demyelination (Figure 3E). Compared with the control group, the power of the delta band was increased and the power of the alpha band was decreased in the LPC group during the WAKE period, suggesting that PVT demyelination reduced awaken activity of EEG in mice (Figure 3E). From the raw EEG traces and EEG power spectrograms, it can also be observed that mice in the LPC group transitioned more easily from an awake state to an anesthesia state. Additionally, the overall excitability of cortical brain electrical activity was markedly lower in the LPC group compared to the control group, indicating that demyelination of the PVT puts a sedation-like state in mice (Figure 3D). Furthermore, the LPC group showed shortened LORR time and prolonged RORR time during 1% isoflurane anesthesia (Figure S1A,C,F).

IBO-induced neuronal depletion in the PVT was also performed to compare the effects of myelin and neurons in the PVT on isoflurane anesthesia (Figure 3F,G). Normal LORR time and prolonged RORR time under 1.4% isoflurane anesthesia were observed in the IBO group, indicating that the arousal-promotion effect of the PVT was suppressed by neuron depletion (Figure 3H). It is suggested that both demyelination and neuronal depletion in the PVT contribute to the action of isoflurane, whereas the process may differ. These findings suggest that the myelin in specific nuclei may play an essential role in regulating consciousness change during isoflurane anesthesia. Furthermore, our data also indicated that the mechanisms underlying general anesthesia may differ between processes involving the myelin and neurons, even in the same nucleus.

3.4 The effect of isoflurane anesthesia is not affected by DRN demyelination

Interestingly, both LHb and PVT are mainly composed of a single type of neurons. The potential reason behind our findings may be attributed to the neuronal homogeneity within the nucleus, as their specific influence on GA is clearly defined. Given the distinct roles of different types of neurons in the nucleus with a diversity of neuron types, we were intrigued by the impact of demyelination in heterogeneous nuclei on the modulation of general anesthesia. The dorsal raphe nucleus (DRN), which is composed of diverse subtypes of neurons, plays multiple roles in general anesthesia, making it the target of our study.^37-40 LPC injection was performed to induce demyelination in the DRN and the decrease MBP expression confirmed the efficiency of myelin defect in the DRN (Figure 4A,B). The LPC group and the control group showed the comparable LORR and RORR time during 1.4% isoflurane anesthesia, indicating that DRN demyelination did not affect the progress of isoflurane anesthesia (Figure 4C). There was no significant difference in the percentage of EEG power in different frequency bands between the two groups, which is consistent with the behavioral results (Figure 4D,E). Correspondingly, demyelination in the DRN exhibited no significant effect on LORR or RORR time during 1% isoflurane anesthesia (Figure S1A,D,G).

Furthermore, IBO-induced neuron deletion in the DRN resulted in a significant reduction in both 5-HT and NeuN neuron populations (Figure 4F,G). The RORR time was prolonged in the IBO group under 1.4% isoflurane anesthesia, but there was no difference in LORR time between the two groups (Figure 4H). These findings are consistent with the former study that reported the arousal effect of DRN in GA.^{41, 42} Together, demyelination in the DRN did not affect the anesthetic effect, while neuron depletion in the DRN enhanced the anesthetic effect.

3.5 Demyelination causes decreased neuronal activity in the awake mice

Overall, our findings demonstrate that myelin in the certain nuclei participate in the regulation of isoflurane anesthesia, although its' specific effect may differ from the roles of neurons in the corresponding nucleus. The different roles of myelin and neurons in GA modulation may due to the effect of myelin destruction on neuronal activity. To determine the relationship between demyelination and neuronal activity in the corresponding nucleus, we set up c-Fos straining to monitor the neural activity after demyelination in the LHb, PVT and DRN, respectively. We found decreased c-Fos in the awake mice after demyelination (Figure 5A,B), demonstrating that demyelination reduces neuronal activity, which suggested that myelin may be involved in the regulation of GA, potentially by influencing neuronal activity.

3.6 Myelin loss causes failure response of the neurons to GA

Next, we conducted real-time calcium fiber recordings to monitor the variation of neuronal activity during isoflurane anesthesia in the LHb, PVT or DRN-demyelinated mice (Figure 6A,B). Myelin loss was confirmed by the MBP expression and the fluorescence signal of GCaMP6s virus showed the fiber photometry recording site in the LHb, PVT and DRN, respectively (Figure 6C,D). The averaged traces plot of calcium signal revealed that the fluorescence intensity of calcium signals in the LPC group was lower than that observed in the control group during the wake period before isoflurane anesthesia (Figure 7A,D,G). The statistical analysis revealed a significant reduction in the calcium signal peak in the LPC group compared to those in the control group, indicating that demyelination leads to a decline in neuronal basal activity (Figure 7C,F,I). During isoflurane anesthesia, the amplitude of changes in calcium signal fluorescence intensity in the LPC group was significantly lower than that in the control group, which proved that the response ability of neurons to isoflurane stimulation after demyelination was greatly reduced, and the neuronal activity was barely altered during the process of isoflurane anesthesia (Figure 7A,B,D,E,G,H). These findings illustrate that the involvement of the myelin in regulating isoflurane anesthesia may be mediated through the impact on neuronal activity.

3.7 Demyelination in the LHb and PVT blunts the effects of neuron activation on isoflurane anesthesia

Finally, we tested the effect of demyelination in the LHb, PVT or DRN on isoflurane anesthesia when we activated the neurons in the corresponding nucleus by chemogenetic approaches (Figure 8A,B). In the LHb, hM3D(Gq)-mediated neuron activation showed a prolonged RORR time, while LPC-induced demyelination blocked this effect (Figure 8C, Figure S2A,B). Moreover, demyelination in the PVT showed similar effect on isoflurane anesthesia after chemogenetic activation of the PVT neurons (Figure 8D, Figure S2C,D). However, no influence on isoflurane anesthesia was observed in the DRN-demyelinated mice after chemogenetic activation of the DRN neurons (Figure 8E, Figure S2E,F). Taken together, our findings demonstrated that myelin in the LHb and PVT regulates the progress of isoflurane anesthesia through their impact on neuronal activity.