A dopaminergic projection from the ventral tegmental area to the dorsal raphe nucleus critically regulates propofol anesthesia in mice
Junjie He, Tianxi Yao, Huan Guo, Haichuan He, Zhuangzhuang Tian, Ketao Ma, Jiangwen Yin, Yan Li
TL;DR
This study identifies a brain circuit connecting the VTA and DRN that controls propofol anesthesia in mice, offering new insights into how consciousness is regulated during anesthesia.
Contribution
The study reveals a novel VTADA→DRN5-HT circuit that bidirectionally regulates propofol anesthesia in mice.
Findings
VTADA terminals and DRN5-HT neurons are anatomically connected and functionally linked.
Activation of VTADA or DRN5-HT neurons shortens propofol anesthesia, while inhibition prolongs it.
Optogenetic activation of VTADA terminals rapidly induces arousal from propofol anesthesia.
Abstract
General anesthesia induces a reversible loss of consciousness, yet the precise neural circuits mediating this state transition remain incompletely understood. The ventral tegmental area (VTA) dopaminergic (DA) neurons and the dorsal raphe nucleus (DRN) serotonergic (5-HT) neurons are key components of the ascending arousal system. This study investigated the existence and functional role of VTADA→DRN5-HT circuit in regulating propofol anesthesia. Using virus-mediated neural circuit tracing in DAT-Cre and TPH2-CreER transgenic mice, we demonstrated an anatomically connected circuit appositions between VTADA axon terminals and DRN5-HT neurons. In vivo fiber photometry revealed that the activity of both presynaptic VTADA terminals in the DRN and postsynaptic DRN5-HT neurons was significantly suppressed under propofol anesthesia. Chemogenetic activation of either VTADA or DRN5-HT neurons…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5- —the National Natural Science Foundation of China
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAnesthesia and Sedative Agents · Sleep and Wakefulness Research · Anesthesia and Neurotoxicity Research
Introduction
General anesthesia (GA) is an indispensable technique in modern medicine, inducing the reversible and controlled state of unconsciousness required for surgical and clinical procedures [1]. Despite more than 150 years of clinical application, the precise neurobiological mechanisms underlying anesthesia—particularly how consciousness is generated, maintained, temporarily abolished and restored—remain among the most challenging frontiers in neuroscience. Propofol, the most widely used intravenous anesthetic, is characterized by rapid onset, quick recovery, and minimal side effects, making it an ideal model for studying the neural circuitry mechanisms of anesthesia-induced loss of consciousness (LOC) and recovery of consciousness (ROC) [2].
Previous studies have demonstrated that anesthetic states do not involve simple global cortical suppression but rather complex processes involving dynamic reorganization of functional connectivity between specific subcortical nuclei and cortical regions [3]. Multiple ascending arousal systems (AASs), including cholinergic, monoaminergic (noradrenergic, dopaminergic, serotonergic, and histaminergic), and glutamatergic parallel neuromodulatory pathways that project extensively to the thalamus and cerebral cortex, play central roles in maintaining wakefulness and regulating consciousness levels [4]. Anesthetics are believed to induce unconsciousness by enhancing GABAergic inhibition or reducing excitatory neurotransmission, thereby suppressing these critical ascending arousal nuclei [5]. For example, disruption of thalamocortical circuits is considered a key event in anesthesia-induced LOC [6], whereas basal forebrain cholinergic neurons are involved in anesthetic emergence [7].
Among the ascending arousal systems, midbrain monoaminergic nuclei have attracted considerable attention because of their critical roles in sleep‒wake cycles, emotion, and cognition. Dopaminergic (DA) neurons in the ventral tegmental area (VTA), termed VTA^DA^ neurons, are traditionally considered the core of the “reward center,” but mounting evidence indicates their crucial involvement in regulating arousal and motivation and maintaining cortical activation [4]. Optogenetic or chemogenetic activation of VTA^DA^ neurons effectively reversed the anesthetic states induced by isoflurane, sevoflurane, or propofol or accelerated emergence [8]. These findings strongly suggest that VTA^DA^ neurons serve as key arousal-promoting nodes and that their suppressed activity may be fundamental to anesthesia-induced LOC.
The dorsal raphe nucleus (DRN), another core monoaminergic system component located in the brainstem, contains the largest cluster of serotonergic (5-HT) neurons in the central nervous system. DRN serotonergic neurons (DRN^5-HT^) precisely regulate multiple higher-order brain functions, including sleep, emotion, learning and memory, and sensory processing through extensive axonal projections [9]. Recent research has revealed the role of DRN^5-HT^ neurons in anesthesia and arousal regulation. Various volatile anesthetics (such as isoflurane) significantly suppress DRN^5-HT^ neuronal firing activity, whereas specific optogenetic or chemogenetic activation of these neurons significantly reduces anesthetic emergence time, demonstrating potent arousal-promoting effects [10]. Thus, DRN^5-HT^ neurons constitute another component of the ascending arousal circuitry.
Both VTA^DA^ and DRN^5-HT^ neurons, as key nodes in the ascending arousal system, play important facilitative roles in anesthesia‒arousal transitions. Although their individual functions have been extensively investigated, whether direct functional neural connections exist between these two important monoaminergic nuclei and whether they function as coordinated neural circuits in regulating anesthesia-induced consciousness state transitions remain unclear. Anatomical studies have suggested reciprocal projections between the VTA and DRN, but the cell-type specificity, synaptic properties, and functional significance of these connections in specific physiological or pathological states, particularly whether VTA^DA^ neurons directly innervate DRN^5-HT^ neurons, lack direct functional evidence [11]. Therefore, elucidating the interactions between these core arousal nuclei is crucial for deepening our understanding of anesthetic neural mechanisms and integrating existing monoaminergic anesthesia theories [4].
On the basis of this background, this study aims to explore a previously underexplored neural circuit connecting two key monoaminergic nuclei—the projection from VTA dopaminergic neurons to DRN serotonergic neurons (VTA^DA^→DRN^5-HT^)—and systematically investigate the dynamic activity changes and functional roles of this circuit during propofol anesthesia-induced LOC and recovery processes. We hypothesize that VTA^DA^ neurons form direct excitatory synaptic connections with DRN^5-HT^ neurons and that this circuit’s activity is suppressed during propofol anesthesia, thereby participating in consciousness state regulation [4]. To this end, we combined targeted circuit tracing, in vivo activity monitoring [12], and causal manipulation during propofol induction and emergence to elucidate how VTA^DA^→DRN^5-HT^ signaling modulates anesthesia-related consciousness transitions.
Materials and methods
Mice
TPH2-CreER mice (Cyagen Biosciences, Suzhou, China) selectively express tamoxifen-inducible Cre recombinase in 5-HT neurons, as previously validated. DAT-Cre mice were obtained from Cyagen Biosciences (Suzhou, China) and selectively expressed tamoxifen-inducible Cre recombinase in DA neurons, as previously validated. Compound TPH2-CreER/DAT-Cre mice and wild-type littermate controls were generated for optogenetic/chemogenetic studies. For other in vivo experiments, C57Bl/6j mice (Cyagen Biosciences, Suzhou) were crossed with DAT-Cre mice to generate DAT-Cre and wild-type littermate controls; C57Bl/6j mice were also crossed with TPH2-CreER mice to generate TPH2-CreER and wild-type littermate controls. All the mice received tamoxifen at 8 weeks of age (0.2 mg g⁻¹, i.p., MCE, New Jersey, USA ). At the end of the experiment, the mice were perfused with 4% formaldehyde. Brain sections were collected to examine EGFP expression in the DRN. Only mice with specific EGFP expression in the DRN were included in the data analysis. All the breeding mice were backcrossed to the C57BL6J background for > 12 generations. The mice were group-housed (2–5 per cage) in temperature-controlled environments at 22–24 °C with 12-hour light‒dark cycles. Unless otherwise specified, the mice had ad libitum access to standard chow and water. All efforts were made to minimize the number of animals used and suffering in the experimental design.
Stereotactic surgery
The mice were anesthetized in an induction chamber with 5% isoflurane (in 100% O_2_, 1L/min flow rate) and secured in a stereotactic apparatus (RWD Life Science, Inc.) with continuous 1.5–2% isoflurane delivered via a face mask. Body temperature was maintained at 37.0 ± 0.5 °C via a heating pad. Scalp preparation included hair removal and triple disinfection with iodophor and 75% ethanol. The injection coordinates were determined according to Paxinos and Franklin’s “The Mouse Brain in Stereotaxic Coordinates” (4th Edition). VTA injections: Bilateral VTA viral injections were performed at coordinates of AP -3.0 mm, ML ± 0.5 mm, and DV -4.3 mm from bregma. Using glass microelectrodes connected to microsyringe pumps, 200 nL of viral suspension was slowly injected per side at 20 nL/min. The needles remained in position for 5 min post-injection to ensure viral diffusion and minimize backflow. The DRN injection coordinates were as follows: AP, -4.65 mm; ML, 0 mm; and DV, -3.3 mm from the bregma. The DRN was angled 10° sagittally to avoid the superior sagittal sinus, with 400 nL injection volumes at identical rates and wait times.
Optical fiber and EEG electrode implantation
Following viral injections, optical fibers or EEG electrodes were implanted according to the experimental groups. Optical fiber implantation: For the mice that received in situ viral expression in the DRN, ceramic ferrule optical fibers (200 μm diameter, NA 0.37) were implanted above the DRN targets during the same surgery at the coordinates AP -4.65 mm, ML 0 mm, and DV -3.1 mm. For mice receiving retrograde viral expression in the DRN, ceramic ferrule optical fibers (200 μm diameter, NA 0.37) were implanted unilaterally in the VTA at coordinates AP -3.0 mm, ML 0.5 mm, and DV -4.1 mm. EEG electrode implantation: Miniature stainless steel screw electrodes were implanted in mice requiring EEG recording. Recording electrodes were placed in the frontal cortex (AP + 1.5 mm, ML + 1.5 mm) and parietal cortex (AP -2.0 mm, ML + 2.0 mm). Reference and ground electrodes were positioned in the skull overlying the cerebellum. All the implants were secured with dental cement. Postsurgery, the mice were allowed to recover on heating pads under close health monitoring for 3 consecutive days. All the animals had ≥ 3 weeks of recovery and a viral expression period before the subsequent experiments.
In vivo fiber photometry recording
After sufficient viral expression (> 3 weeks post-surgery), fiber photometry recordings were conducted in fiber-implanted mice via a recording system (RWD R821 multichannel fiber recording system, Shenzhen, China) containing multichannel LED sources and high-sensitivity photodetectors. A 470 nm excitation light (GCaMP excitation) and 405 nm isosbestic control light (correcting motion artifact-induced nonspecific fluorescence changes) were coupled through dichroic mirrors into fiber patch cables connected to head-mounted fiber implants. The emitted fluorescence was returned through the same fiber, filtered, and detected by photodetectors. Fiber photometry data were analyzed via MATLAB 2016a (MathWorks, Cambridge, UK). The fluorescence intensity changes (ΔF/F) were calculated as (F-F₀)/F, where F represents the test fluorescence and F₀ represents the baseline fluorescence intensity.
Electroencephalogram (EEG) recording
EEG signals were acquired via the Pinnacle 8200 signal acquisition system (Pinnacle Technology, Kansas, USA) at a 400 Hz sampling frequency. EEG recording commenced ≥ 5 days post-electrode implantation surgery. The mice were placed in observation chambers for EEG recording. Power spectral analysis was performed on the propofol anesthesia induction and recovery cycle data. Artifact-containing data were removed, and 50 Hz AC interference was filtered. The data were filtered between 0.25 and 60 Hz (basic FIR filter, order 36, EEGLAB, USA). The frequency band power percentages were calculated by dividing the average signal power for each band (δ: 0.25–4 Hz, θ: 4–8 Hz, α: 8–13 Hz, β: 13–25 Hz, γ: 25–60 Hz) by the total power from 0.25 to 60 Hz. Time‒frequency plots were bandpass filtered between 0.25 and 60 Hz and constructed via multitaper methods implemented in the EEGLAB toolbox in MATLAB 2016a (MathWorks, Cambridge, USA). Parameters: sampling frequency (Fs) = 400; nFFT = 2nextpow2(L); window = Hanning; overlap = (length[window])/2; window duration = 20.48 s; window step = 0.64 s.
Propofol anesthesia and behavioral assessment
For continuous tail vein infusion, each mouse was placed in an infusion cage (Fig. 1A) with a 4.5-gauge needle connected to transparent tubing (0.45 mm inner diameter) inserted into the tail vein. The needle was secured to the tail via cyanoacrylate adhesive, and the tubing was connected to a microinfusion pump for continuous propofol infusion (10 mg ml⁻¹ lipid emulsion; Fresenius Kabi Austria GmbH, Graz, Austria). The loss of righting reflex (LORR) serves as a behavioral indicator for assessing the hypnotic properties of propofol. Following propofol administration, the infusion cage was gently rotated every 15 s to position the mouse in the supine position, and the righting ability was evaluated. The LORR time was defined as the first instance when the mouse failed to right itself for > 30 s and was recorded as the LORR time at 30 s. The return of righting reflex (RORR) time was defined as the moment when the mouse could right itself (all four paws on the ground). To examine the induction and emergence times, propofol was infused at 10 mg kg⁻¹ min⁻¹ for 5.5 min to induce LORR, followed by maintenance at 3.0 mg kg⁻¹ min⁻¹ for 30 min on a heating pad to maintain body temperature. The induction time was the interval from the start of propofol infusion to the LORR, whereas the emergence time was the duration from propofol cessation to the RORR. Additionally, loss of consciousness (LOC) and return of consciousness (ROC) were used to assess the hypnotic properties of propofol. For the fiber photometry experiments, propofol was infused at 10 mg kg⁻¹ min⁻¹ to induce LOC. LOC onset was defined as EMG activity cessation when the mouse assumed lateral recumbency, with EEG patterns transitioning from low-voltage fast waves to high-voltage slow waves. Propofol infusion was terminated 30 s after LOC onset, maintaining lateral recumbency. During emergence, the brain state transitioned from the LOC to the ROC before righting occurred. ROC onset was defined as the first appearance of sustained EMG activity (> 20 s) and low-voltage fast-wave EEG. The righting duration was the time from ROC onset to the first righting movement. For the optogenetic inhibition experiments, 10 mg/kg propofol was infused for 5.5 min. In optogenetic activation experiments, propofol was induced at 10 mg kg⁻¹ min⁻¹ and then maintained at 3 mg kg⁻¹ min⁻¹ for steady-state general anesthesia. During steady-state anesthesia, optogenetically induced ROC onset was defined as the appearance of low-amplitude high-frequency EEG, active EMG, or behavioral responses. ROC termination was defined as the resumption of high-voltage slow-wave EEG with an EMG absence > 20 s. The ROC duration was measured from onset to termination.
Chemogenetics
For chemogenetic manipulations, chemogenetic activation experiments employed clozapine N-oxide (CNO, 0.3 mg kg⁻¹) or saline via intravenous administration for 1 min; chemogenetic inhibition experiments used CNO (1 mg kg⁻¹) or saline intravenously for 1 min. After 20 min, propofol was continuously infused at 10 mg·kg⁻¹·min⁻¹ for 5.5 min to induce loss of righting reflex (LORR), followed by maintenance at 3.0 mg·kg⁻¹·min⁻¹ for 30 min. During propofol infusion, the infusion cage was gently rotated every 15 s to position mice on their backs for righting reflex assessment. Each mouse was tested at 5-day intervals to eliminate residual effects of CNO and propofol.
Histology and immunofluorescence staining
Postexperiment, the mice were deeply anesthetized with an intraperitoneal pentobarbital overdose and transcardially perfused with phosphate-buffered saline (PBS) followed by 4% paraformaldehyde (PFA) for fixation. The brain tissues were extracted, fixed in 4% PFA for 4–6 h, and then dehydrated in 30% sucrose until they sank. Coronal brain Sect. (25 μm) were prepared via a cryostat (Leica).
To verify the synaptic connections between VTA^DA^ neuronal axon terminals and DRN^5-HT^ neurons, immunofluorescence double-labeling was performed. The sections were blocked in PBS containing 0.3% Triton X-100 and 10% normal goat serum for 1 h at room temperature. The sections were then incubated overnight at 4 °C with a primary antibody (rabbit anti-TPH2, identifying 5-HT neurons, 1:50 dilution; Abcam #ab288067). After being washed with PBS, the sections were incubated with a fluorescent secondary antibody (goat anti-rabbit-DyLight 647, 1:100 dilution, Boster) for 2 h at room temperature in the dark. Since VTA^DA^ neurons already expressed jGCaMP7f (green fluorescence) via viral delivery, additional staining was unnecessary. The sections were coverslipped with DAPI-containing mounting medium. High-resolution imaging was performed via laser confocal microscopy (Leica, Germany) to confirm the viral expression sites and fiber/electrode placement accuracy and to observe the anatomical relationships between VTA^DA^ axon terminals (green) and DRN^5-HT^ neuronal somata (red).
To verify the Cre-dependent viral transfection efficiency, immunofluorescence double-labeling was conducted. DAT-Cre mouse sections were colabeled with TH primary antibody (rabbit anti-TH, identifying DA neurons, 1:50 dilution, Abcam #ab137869), whereas TPH2-CreER mouse sections were colabeled with TPH2 antibody. The secondary antibodies used were goat anti-rabbit-DyLight 647 (1:100, Boster) and goat anti-rabbit-DyLight 488 (1:100, Boster) according to identical staining protocols. Animals with poor viral transfection or fiber placement > 100 micrometers from target regions were excluded from analysis.
Data analysis and statistics
The data were analyzed via GraphPad Prism 8.0. (GraphPad Software, CA, USA). The quantitative data are presented as the means ± standard errors of the means (SEMs). The Shapiro‒Wilk test was used to determine data normality. For normally distributed two-group comparisons, unpaired t tests were used; for nonnormally distributed data, Mann‒Whitney U tests were used. For multigroup comparisons, one-way or two-way ANOVA with Tukey’s or Sidak’s post hoc multiple comparison tests were used as appropriate. All tests were two-tailed, with P < 0.05 considered statistically significant.
Results
Synaptic connections between VTADA neurons and DRN5-HT neurons
To investigate direct projections from VTA^DA^ neurons to DRN^5-HT^ neurons, neural circuit tracing experiments were conducted. DAT-Cre transgenic mice received stereotaxic injections of the Cre-dependent anterograde viral rAAV-CAG-FLEX-jGCaMP7f (serotype AAV2/1) labeling of VTA dopaminergic axons. After three weeks, the mice underwent perfusion fixation, and DRN sections were immunostained with a TPH2 antibody (the rate-limiting enzyme for 5-HT synthesis) to label 5-HTergic neurons (Fig. 1A). High-resolution confocal microscopy revealed clear green fluorescent signals from VTA^DA^ neuronal axons and terminals distributed throughout the DRN. Importantly, these VTA-derived green axonal terminals formed close appositions with TPH2-positive red 5-HT neuronal somata and dendrites (Fig. 1B). These morphological findings provide strong anatomical evidence for synaptic inputs from VTA^DA^ neurons to DRN^5-HT^ neurons.
Fig. 1. Synaptic connections exist between VTA^DA^ neurons and DRN^5-HT^ neurons. (A) Viral strategy using cre-dependent anterograde trans-synaptic virus (AAV2/1 serotype) to specifically label VTA^DA^ neuron axons and terminals targeting DRN. (B) Representative immunofluorescence images of jGCaMP7f, Tph2, DAPI and their colocalization in DRN. Scale bar = 100 μm at 100 × magnification. Scale bar = 50 μm at 200 × magnification. Data from 3 different mice were quantified
Activity of the VTADA→DRN circuit is suppressed during propofol anesthesia
Following the confirmation of anatomical connectivity between the VTA^DA^ and DRN^5-HT^ neurons, in vivo fiber photometry was employed to monitor real-time neural activity changes during the induction and maintenance of propofol anesthesia. To monitor activity of DRN-projecting VTA^DA^ neurons, DAT-Cre mice received stereotaxic injections of the retrograde virus rAAV-EF1α-DIO-GCaMP6f (serotype AAV2/R) into the DRN region, enabling real-time detection of VTA^DA^ neuronal activity via fiber photometry while simultaneously recording EEG and EMG signals (n = 6 mice) (Fig. 2A and Extended Data Fig. 1A). Stable calcium signal fluctuations were observed during baseline awake states. To better simulate clinical propofol administration, a continuous tail vein infusion anesthesia model was established (Fig. 2B), which provides more stable propofol distribution and metabolism than traditional intraperitoneal or single intravenous injections do. Propofol was infused at 10 mg kg⁻¹ min⁻¹ to induce loss of consciousness (LOC), with recovery of consciousness (ROC) occurring upon cessation (Figs. 2C-D). Brain state transitions were assessed by EEG pattern changes from low-voltage fast waves to high-voltage slow waves and EMG transitions from high activity to complete suppression for LOC, with opposite patterns for ROC. During propofol induction, DRN-projecting VTA^DA^ population activity significantly decreased 10 s before LOC (P = 0.0031), further decreased during the transition period (10 s before to 30 s after LOC; P < 0.0001), and remained low throughout anesthesia (Fig. 2E-G). During emergence, DRN-projecting dopaminergic population activity significantly increased 20 s before ROC (P = 0.0097), peaked rapidly at consciousness recovery (P < 0.0001), and maintained elevated levels during the righting reflex (Figs. 2H-J). These results suggest that DRN-projecting dopaminergic neurons participate in propofol-induced LOC and ROC processes.
Postexperiment, VTA sections underwent perfusion fixation and immunofluorescence staining with a tyrosine hydroxylase (TH) antibody to label dopaminergic neurons. High-resolution confocal microscopy revealed retrograde GCaMP6f green fluorescent signals from the DRN to VTA somata, These green signals were widely distributed in the VTA and colocalized with TH signals (Extended Data Fig. 1A); mice lacking green fluorescent signals were excluded. These findings suggest that DRN-projecting dopaminergic neurons play important roles in propofol-induced LOC and ROC processes. Fiber photometry recordings further demonstrated that propofol-induced unconsciousness is closely correlated with overall VTA^DA^→DRN circuit activity suppression.
Fig. 2VTA^DA^→DRN circuit activity is significantly suppressed under propofol anesthesia. (A) Viral strategy using cre-dependent retrograde virus GCaMP6f (AAV2/R serotype) to specifically label VTA^DA^ neurons projecting to DRN. (B) Setup for fiber photometry and EEG/EMG recordings during continuous propofol infusion via mouse tail vein. Inset: lateral view of infusion cage. (C-F) Calcium signals from GCaMP6f-expressing VTA^DA^ neurons projecting to DRN, corresponding to EEG/EMG, during propofol anesthesia (C-F) induction and (D-I) emergence. Propofol infusion at 10 mg kg⁻¹ min⁻¹ induced LOC and was terminated 30s after LOC onset (D-I). (C and D) Representative EEG/EMG traces and relative EEG power. (E and H) Mean fluorescence signals shown as mean (bold line) with standard deviation (shaded area). (F and I) GCaMP fluorescence signals, n = 6 mice. (G and J) Mean fluorescence signals across different time periods. B.S. represents (-100s to -50s) in (G) and (-120s to -100s) in (J). Paired t-test or Wilcoxon matched-pairs signed-rank test was used. Time = 0 represents LOC or ROC onset. LOC, loss of consciousness; LORR, loss of righting reflex; VTA, ventral tegmental area; DRN, dorsal raphe nucleus; ROC, return of consciousness
DRN5-HT neurons bidirectionally modulate propofol anesthesia
Based on the aforementioned anatomical evidence indicating potential connectivity between VTA^DA^ neurons and DRN^5-HT^ neurons, we subsequently monitored the activity dynamics of DRN^5-HT^ neuronal populations during anesthesia. TPH2-CreER transgenic mice (n = 6) received DRN injections of AAV2/9-rAAV-CAG-FLEX-jGCaMP7f and fiber implantation for real-time monitoring of DRN^5-HT^ neuronal activity via fiber photometry, with concurrent EEG and EMG recordings (Fig. 3A). Postexperiment, DRN regions were processed through perfusion fixation and sectioning, followed by immunofluorescence staining with a TPH2 antibody and high-resolution confocal microscopy to exclude noncolabeled virus-transfected neurons (Fig. 3B). The fiber photometry results were highly consistent with the upstream signal changes. During propofol induction, calcium signals from DRN^5-HT^ somata significantly decreased 10 s before LOC (P = 0.0062), further decreased during the transition period (LOC to 30 s post-LOC; P = 0.0023), and remained low throughout anesthesia (Fig. 3C-E). During emergence, DRN^5-HT^ neuronal activity began rising 20 s before ROC (P = 0.0252), rapidly peaked at ROC onset (P < 0.0001), and maintained high activity during righting behavior (Fig. 3F-H).
To examine the roles of DRN^5-HT^ neurons in anesthetic behavior, chemogenetic activation or inhibition was used to assess the effects on the propofol induction time (LORR) and emergence time (RORR). TPH2-CreER mice received DRN injections of excitatory (hM3Dq) or inhibitory (hM4Di) DREADD viruses (n = 12 per group). CNO or saline was administered intraperitoneally 30 min before anesthesia. Compared with saline controls, chemogenetic activation of DRN^5-HT^ neurons significantly prolonged the propofol induction time (LORR: 231.2 ± 16.28 s vs. 186.0 ± 18.39 s; paired t test, P < 0.0001) while shortening the emergence time (RORR: 675.6 ± 93.04 s vs. 1275 ± 83.02 s; P < 0.001), demonstrating pro-arousal effects(Fig. 3J). Conversely, DRN^5-HT^ neuronal inhibition shortened the LORR (137.8 ± 43.73 s vs. 205.7 ± 31.26 s; P = 0.0006) and prolonged the RORR (1675 ± 463.1 s vs. 1247 ± 120.5 s; P = 0.0040), extending the duration of anesthesia(Fig. 3L). These behavioral data suggest that DRN^5-HT^ neuronal activity bidirectionally regulates mouse sensitivity to propofol anesthesia and emergence capacity.
Fig. 3DRN^5-HT^ neurons bidirectionally modulate propofol anesthetic effects. (A) Viral strategy using AAV-jGCaMP7fto monitor DRN^5-HT^ neuronal activity. (B) Representative microscopic images of jGCaMP7f, Tph2, DAPI and their colocalization in DRN. Scale bar = 50 μm at 200× magnification. (C-D-F-G) Calcium signals of DRN^5-HT^ neurons expressing jGCaMP7f during propofol anesthesia (C-D-E) induction and (F-G-H) emergence. (C and F) Mean fluorescence signals shown as mean (bold line) with standard deviation (shaded area). (D and G) GCaMP fluorescence signals, n = 6 mice. (E and H) Mean fluorescence signals at different time periods. B.S. represents (-300s to -250s) in (C) and (-120s to -100s) in (F). Paired t-test or Wilcoxon matched-pairs signed-rank test was used. Time = 0 represents the moment of loss of righting reflex or return of righting reflex. (I) Viral strategy using Cre-dependent hM3Dq to activate DRN^5-HT^ neurons, with inset showing representative immunofluorescence images of EGFP expression in DRN. Scale bar = 50 μm at 200× magnification. (J) Propofol anesthesia induction and emergence times after intravenous injection of CNO or saline in hM3Dq-expressing mice. Data presented as mean (standard deviation), paired t-test used. (K) Viral strategy using hM4Di to inhibit DRN^5-HT^ neurons, with inset showing representative immunofluorescence image of EYFP expression in DRN. Scale bar = 50 μm at 200× magnification. (L) Propofol anesthesia induction and emergence times after intravenous injection of CNO or saline in hM4Di-expressing mice. Data presented as mean (standard deviation), paired t-test used. LORR, loss of righting reflex; RORR, return of righting reflex; CNO, clozapine N-oxide
Activation of VTADA→DRN projections induces transient arousal from propofol anesthesia
While chemogenetic experiments confirmed downstream neuronal function, direct circuit involvement in anesthetic behavior remains unestablished. To investigate the VTA^DA^→DRN projection function, optogenetics combined with EEG recording was employed. DAT-Cre mice received VTA injections of Cre-dependent ChR2 virus (rAAV-EF1α-DIO-hChR2-EYFP) with DRN fiber and EEG electrode implantation (n = 6). During stable propofol maintenance (persistent loss of the righting reflex and high-amplitude, low-frequency EEG), 473 nm blue light (30 Hz, 5 ms pulses, 30 s) was delivered via fibers to specifically activate VTA^DA^ axon terminals projecting to the DRN (Fig. 4A). Postexperiment DRN sections were subjected to TH antibody immunofluorescence staining and high-resolution confocal microscopy to exclude noncolabeled virus-transfected neurons (Fig. 4B). Within seconds of light onset, anesthetized mice rapidly exhibited signs of arousal, including head lifting, limb movement, and transient righting behavior. EEG power density analysis of ChR2-EYFP mice revealed that optogenetic activation of VTA^DA^→DRN projections significantly reduced delta (0.5–4 Hz, 27.167 ± 1.365% vs. 19.533 ± 2.473%; P < 0.0001) and theta power (4–10 Hz, 39.617 ± 1.797% vs. 32.083 ± 1.123%; P < 0.0001) while significantly increasing alpha and beta power (10–25 Hz, 33.200 ± 1.705% vs. 48.4 ± 3.034%; P < 0.001) (Fig. 4C-E). The mice rapidly returned to their previous anesthetic state after light cessation. EYFP mice showed no significant differences between baseline and light exposure (P > 0.05) (Fig. 4F-H).
To further validate the necessity of projection, another group of DAT-Cre mice received VTA injections of blue light-inhibited virus (rAAV-EF1α-DIO-SwiChRca-TS-EYFP) with DRN fiber implantation (n = 8) (Fig. 4I). Postexperiment DRN processing included TH antibody immunofluorescence staining and high-resolution confocal microscopy, excluding mice with noncolabeled virus-transfected neurons (Fig. 4J). Under propofol anesthesia, blue light application (473 nm, 30 Hz, 5 ms pulses, 30 s) continuously inhibited VTA dopaminergic projection terminals in the DRN. Compared with the yellow light controls, the propofol-induced LOC time was significantly shorter (149.9 ± 13.2 s vs. 219.8 ± 17.90 s; P = 0.0002), whereas the ROC time after propofol cessation was significantly prolonged (379.0 ± 23.33 s vs. 234.9 ± 11.78 s; P < 0.0001). EEG power density spectral analysis within 20 s post-LOC and ROC revealed no significant differences between blue and yellow light illumination (Fig. 4K-M). These experiments further confirmed the critical role of the pathway in maintaining normal arousal processes. The results demonstrated that specific activation of VTA^DA^→DRN projections sufficiently disrupted propofol-maintained anesthesia and induced arousal at both the behavioral and neurophysiological levels.
Fig. 4. Specific activation of VTA^DA^→DRN neural projections induces transient arousal from propofol anesthesia. (A) Viral strategy using Cre-dependent hChR2 to activate VTA^DA^ axon terminals projecting to DRN. (B) Representative immunofluorescence images of hChR2, TH, DAPI and their colocalization in VTA. Scale bar = 50 μm at 200× magnification. (C-F) Blue light pulses (5 ms pulses) activating ChR2-expressing VTA^DA^ axon terminals projecting to DRN, showing representative EMG/EEG traces and EEG power spectra during optical stimulation (30 Hz; 5 ms pulses, 30 s duration) in ChR2-EYFP mice (C) or EYFP mice (F) under steady-state propofol anesthesia. (D-E-G-H) Difference in EEG relative power between 30 s before optical stimulation and during optical stimulation in (D and E) ChR2-EYFP mice and (G and H) EYFP mice, n = 6 mice, :P < 0.001, paired t-test. (I) Viral strategy using Cre-dependent SwiChRca (blue light-induced sustained hyperpolarization) to inhibit VTA^DA^ axon terminals projecting to DRN. (J) Representative immunofluorescence images of SwiChRca, TH, DAPI and their colocalization in VTA. Scale bar = 50 μm at 200× magnification. (K) Representative EMG/EEG traces and EEG power spectra. Propofol 10 mg·kg^-1·min^-1 continuous infusion for 5.5 min. Blue light (top; B-Light) or yellow light (bottom; Y-Light) pulses (30 Hz; 5 ms pulses) given for 20 min, 5 min before propofol infusion. Translucent blue and yellow vertical bars represent 20 s periods after LOC or ROC onset. (L and M) Time to reach (L) LOC and (M) ROC (left) and relative total EEG power within 20 s (right), n = 8 mice. Mean (standard deviation), paired t-test. rAAV, recombinant adeno-associated virus; LOC, loss of consciousness; VTA, ventral tegmental area; DRN, dorsal raphe nucleus; ROC, return of consciousness
The arousal-promoting effect on VTADA neurons requires downstream DRN5-HT neuron activity
To further elucidate the functional interdependence between VTA dopaminergic neurons and DRN 5-HTergic neurons, we designed an experiment in which optogenetic activation was combined with chemogenetic inhibition. DAT-Cre/TPH2-CreER double transgenic mice were generated, with VTA dopaminergic neurons expressing the optogenetic activator ChR2 and DRN serotonergic neurons expressing the chemogenetic inhibitor hM4Di (Fig. 5A-B). The mice were divided into two groups (n = 6/group): one group received saline injection before light stimulation (control group), while the other received CNO to specifically inhibit DRN^5-HT^ neuronal activity (experimental group). During propofol anesthesia maintenance, both groups received identical DRN light stimulation (activating VTA^DA^ terminals).
The results clearly revealed the interdependence between these systems. In the control group (Saline + Light), light stimulation significantly promoted arousal, as expected. EEG power density comparisons before and after stimulation revealed that optogenetic activation of the VTA^DA^→DRN^5-HT^ pathway reduced delta power (0.5–4 Hz, 28.200 ± 1.012% vs. 19.033 ± 0.977%; P < 0.0001) and theta power (4–10 Hz, 40.150 ± 1.305% vs. 33.900 ± 1.047%; P < 0.0001) while increasing alpha and beta power (10–25 Hz, 31.650 ± 0.459% vs. 47.067 ± 1.829%; P < 0.001). The mice quickly returned to their previous anesthetic state after light stimulation ceased (Fig. 5C-E). However, in the experimental group (CNO + Light), where DRN^5-HT^ neuronal activity was chemogenetically suppressed, identical light stimulation almost completely lost its arousal-promoting effect (Fig. 5F-H).
This critical finding demonstrates that VTA^DA^ neuron-mediated arousal effects are not independent but are highly dependent on the normal function of their downstream targets—DRN^5-HT^ neurons. When DRN^5-HT^ neurons are inhibited, upstream VTA^DA^ activation signals cannot be effectively transmitted or executed, thereby disrupting the arousal-promoting function of the entire circuit. This provides decisive evidence that the VTA^DA^→DRN^5-HT^ circuit operates as a functional unit in anesthetic regulation.
Fig. 5. The arousal-promoting function of VTA^DA^ neurons depends on normal activity of downstream DRN^5-HT^ neurons. (A) Strategy for activating VTA^DA^→DRN projections while simultaneously inhibiting DRN^5-HT^ neurons. (B) Representative immunofluorescence images of viral injections in VTA (left) and DRN (right). Scale bar = 100 μm at 100× magnification. (C-F) Blue light pulses (5 ms pulses) activate ChR2-expressing VTA^DA^ axon terminals projecting to DRN. Representative EMG/EEG traces and EEG power spectra during photostimulation (30 Hz; 5 ms pulses, 30 s duration) in mice under steady-state propofol anesthesia: Saline+Light group (C) or CNO+Light group (F). (D, E, G, H) Differences in EEG relative power between 30 s before and during photostimulation in (D, E) Saline+Light group and (G, H) CNO+Light group mice, n = 6 mice, ****P < 0.001, paired t-test. (I) Schematic diagram. rAAV, recombinant adeno-associated virus; LOC, loss of consciousness; VTA, ventral tegmental area; DRN, dorsal raphe nucleus
Discussion
This study systematically revealed a previously underrecognized neural pathway—the projection from dopaminergic neurons in the ventral tegmental area (VTA) to serotonergic neurons in the dorsal raphe nucleus (DRN)—that plays a crucial regulatory role in propofol anesthesia-induced consciousness state transitions (Fig. 5I). The findings not only provide direct anatomical evidence for this circuit but also elucidate its dynamic functions during anesthesia induction, maintenance, and emergence through multiple techniques, including in vivo fiber photometry, chemogenetics, and optogenetics, offering new insights into the neurobiological basis of anesthesia.
Arousal and sleep regulation depend on complex networks comprising multiple brain regions, including the brainstem, hypothalamus, and cortex. The VTA dopamine system and DRN^5-HT^ system, as core components of the ascending reticular activating system, have long been recognized as crucial for maintaining wakefulness and regulating emotion and motivational behavior [13]. Previous studies have demonstrated that VTA dopaminergic neuron activity is closely related to the emergence of propofol anesthesia, with optogenetic activation significantly accelerating mouse recovery from propofol anesthesia [14]. Moreover, DRN^5-HT^ neurons were confirmed to be important nodes that promote arousal, with their activity suppressed by various anesthetics (including propofol), whereas their activation produces significant arousal-promoting effects [10]. Although the independent roles of these systems in anesthesia regulation have been established, whether direct functional connections exist between them to coordinately regulate anesthetic state transitions remains unclear.
This study provides the first definitive anatomical evidence for a direct projection from VTA^DA^ neurons to DRN^5-HT^ neurons. Through anterograde transsynaptic virus injection in the VTA region of DAT-Cre mice, axon terminals from VTA^DA^ neurons formed close contacts with the cell bodies and dendrites of TPH2-positive 5-HTergic neurons in the DRN region (Fig. 1A-B). While these findings support an anatomically connected and functionally important direct pathway, the definitive monosynaptic nature of this projection awaits future verification using techniques such as retrograde rabies virus tracing or electrophysiological assays with TTX/4AP protocols. This discovery not only fills an important gap in anesthesia regulatory neural circuit mapping but also establishes the foundation for subsequent functional studies. Notably, this anatomical connection suggests that the VTA dopamine system may participate in global consciousness state regulation by directly influencing downstream DRN^5-HT^ system activity, thereby indirectly modulating broader cortical regions. The DRN^5-HT^ system sends extensive fiber projections throughout the brain, particularly to regions associated with greater cognition and arousal, such as the prefrontal cortex, thalamus, and basal forebrain [15, 16]. Therefore, regulatory signals from the VTA may be amplified and transmitted brain-wide through the DRN as a “relay station,” forming an efficient arousal regulatory network.
After confirming the VTA^DA^→DRN^5-HT^ anatomical connection, in vivo fiber photometry was employed to monitor real-time circuit activity dynamics during propofol anesthesia and recovery cycles. Compared with traditional intraperitoneal or single intravenous injection, the established continuous tail vein propofol infusion model (Fig. 2B) more accurately mimics clinical anesthesia practice, providing more stable blood drug concentrations and anesthetic depth, thereby ensuring reliable neural activity recordings.
Our results clearly demonstrate that calcium activity in both VTA^DA^ neuron terminals projecting to the DRN and DRN^5-HT^ neuron populations is significantly and synchronously suppressed during propofol-induced loss of consciousness (LOC). Specifically, neural activity in both upstream VTA^DA^ axon terminals (Fig. 2E-G) and downstream DRN^5-HT^ neuron cell bodies (Fig. 3C-E) begins to decline sharply approximately 10 s before LOC and remains at low levels throughout anesthesia maintenance. Conversely, during anesthesia recovery of consciousness (ROC), neural activity at both circuit nodes significantly increases approximately 20 s before consciousness recovery and peaks when animals regain the righting reflex (Figs. 2H-J and 3F-H). This high synchrony of upstream and downstream neuron activity and its tight temporal correlation with anesthetic behavioral states strongly indicate that overall VTA^DA^→DRN^5-HT^ circuit activity plays a key role in regulating propofol anesthesia depth and consciousness state transitions.
Notably, a decrease in neural activity occurs before LOC behavioral indicators, whereas recovery precedes ROC occurrence. This “leading” nature of neural activity dynamics supports the circuit’s potential role as a “driver” rather than “responder” in consciousness state transitions. This predictive activity pattern aligns with previous findings in other arousal-promoting circuits, such as glutamatergic neurons in the paraventricular thalamus (PVT), which exhibit increased activity before recovery from propofol anesthesia. These results collectively point to a core mechanism: general anesthetics may preferentially suppress these key arousal-promoting “upstream” nodes, triggering cascade reactions that ultimately lead to brain-wide functional suppression and consciousness loss.
During data analysis, particular attention was given to excluding motion artifact effects on fiber photometry signals, especially during the behaviorally dynamic anesthesia induction and recovery phases, where animal movement could introduce significant signal noise. Standard signal processing procedures were employed, including the use of 405 nm wavelength light as a reference signal to correct nonspecific fluorescence changes caused by movement or fiber bending and the application of polynomial fitting algorithms for signal debleaching and baseline correction to ensure ΔF/F value accuracy [17]. These rigorous data processing methods enhance the reliability of the conclusions.
To verify the roles of DRN^5-HT^ neurons in propofol anesthesia, chemogenetic techniques were used for specific regulation. After excitatory (hM3Dq) or inhibitory (hM4Di) DREADDs were expressed in the DRN region of TPH2-CreER mice, the bidirectional regulation of DRN^5-HT^ neurons significantly altered the sensitivity of the mice to propofol anesthesia. Chemogenetic activation of DRN^5-HT^ neurons significantly prolonged the anesthesia induction time (LORR time) and shortened the recovery time (RORR time), demonstrating clear antianaesthetic and emergence-promoting effects. Conversely, inhibiting these neurons accelerated anesthesia induction and significantly delayed recovery (Fig. 3I-J).
These behavioral results strongly suggest that DRN^5-HT^ neuron activity levels are key factors affecting the depth and duration of propofol anesthesia, which is consistent with the findings of fiber photometry recordings showing that neural activity is correlated with anesthetic states. This finding aligns with previous studies indicating that activating DRN 5-HTergic neurons can promote isoflurane anesthesia recovery,^10^ while this study extends this function to the most widely used clinical intravenous anesthetic—propofol.
It is important to note a recognized limitation of the DREADD approach. Systemically administered CNO can be back-metabolized to clozapine, which at low nanomolar concentrations may bind to endogenous dopaminergic, serotonergic, muscarinic, and histaminergic receptors, potentially influencing arousal and EEG activity independent of DREADD expression. To mitigate this concern, we employed the lowest effective systemic CNO dose and included rigorous within-subject vehicle controls as well as DREADD-negative/Cre-negative and fluorescent protein-only control groups [18]. Under these controlled conditions, we did not detect significant effects of CNO or vehicle on baseline behaviors, EEG spectra, or propofol anesthesia sensitivity in the absence of DREADD expression. This supports the interpretation that the observed chemogenetic effects are more likely due to specific manipulation of the target circuit. However, we acknowledge that we did not directly measure clozapine levels or systematically compare the effects of low-dose clozapine to CNO in this study. Therefore, we cannot entirely rule out a minor contribution from back-metabolized clozapine. Future studies could further minimize this confound by employing next-generation DREADD agonists with improved pharmacokinetics (e.g., JHU37160) or by using intracranial ligand delivery to the DRN.
Although chemogenetic experiments confirmed the functionality of downstream DRN^5-HT^ neuronal populations, they cannot directly demonstrate the causal relationship of the specific VTA^DA^→DRN^5-HT^ projection pathway in anesthetic behavior. Therefore, optogenetic technology with higher spatiotemporal resolution was employed. The ChR2-expressing virus was injected into the VTA region of DAT-Cre mice, and optical fibers were implanted in the downstream DRN region, enabling specific activation of axon terminals projecting from the VTA to the DRN under stable anesthesia.
The experimental results revealed that during stable propofol anesthesia maintenance, 30-second blue light stimulation induced transient but distinct arousal behaviors, including head lifting and limb movement, accompanied by an EEG transition from high-amplitude slow waves to low-amplitude fast waves characteristic of awakening (Fig. 4A). This “optogenetic arousal” phenomenon directly demonstrates that the VTA^DA^→DRN^5-HT^ circuit possesses significant arousal-promoting capacity, temporarily “overriding” anesthetic suppression and restoring consciousness from an unconscious state. This finding reveals the “switching” mechanism of anesthetic emergence and provides potential neural circuit targets for the development of novel arousal strategies.
This study reveals the structural and functional importance of the VTA^DA^→DRN^5-HT^ circuit, yet the detailed neurotransmitter mechanisms require further investigation. VTA dopaminergic neurons are renowned for dopamine release, but mounting evidence indicates that some VTA neurons, particularly subpopulations projecting to specific brain regions, corelease glutamate or GABA [19, 20]. Therefore, VTA^DA^→DRN signaling may not depend solely on dopamine. Future studies should combine pharmacological blockade, electrophysiological recording, and genetic labeling to identify the primary neurotransmitters (dopamine, glutamate, or their synergistic action) and corresponding receptor subtypes (D1/D2 or NMDA/AMPA receptors) [21, 22]. This can be achieved through optogenetic activation in brain slices while recording postsynaptic currents in downstream DRN^5-HT^ neurons and applying different neurotransmitter receptor antagonists to characterize synaptic transmission [23, 24].
Additionally, the downstream molecular signaling pathways involved in DRN^5-HT^ neuron activation remain unclear. Whether DRN^5-HT^ neuron activation regulates gene expression and neuronal plasticity through the classical cAMP‒PKA or MAPK signaling pathways, thereby affecting anesthetic state persistence, requires investigation. These questions can be explored by rapidly extracting DRN tissue following chemogenetic manipulation and detecting the phosphorylation levels of key signaling pathway proteins via Western blotting or ELISA [25, 26].
From a clinical translation perspective, the revealed VTA^DA^→DRN^5-HT^ circuit may serve as a novel therapeutic target for optimizing anesthetic management. Circuit-specific drugs could enable precise anesthetic depth control, accelerate postoperative recovery, and potentially prevent complications such as post-operative cognitive dysfunction (POCD) [27, 28]. Designing preclinical animal models to evaluate long-term, specific circuit modulation effects on postoperative cognitive behavior (assessed through a water maze and novel object recognition tests) under simulated surgical and anesthetic conditions will be crucial for validating the clinical potential of these models.
Conclusion
In summary, using multiple complementary approaches, this study investigated the VTA^DA^→DRN^5-HT^ circuit in the context of propofol anesthesia. We provide evidence consistent with functional connectivity between these regions, describe activity changes across anesthesia–arousal cycles, and show that targeted manipulations of this circuit can influence arousal in the experimental model used. These findings contribute to ongoing efforts to understand neural mechanisms underlying anesthetic states. However, the results are context- and model-specific, and further work is needed to determine the circuit’s necessity and generalizability across anesthetics, species, and clinical settings before drawing conclusions about clinical applicability or therapeutic targets.
Supplementary Information
Below is the link to the electronic supplementary material.
Supplementary Material 1
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Brown EN, Lydic R, Schiff ND. General anesthesia, sleep, and coma. Schwartz RS, ed. N Engl J Med. 2010;363(27):2638–2650. 10.1056/NEJ Mra 0808281.10.1056/NEJ Mra 0808281 PMC 316262221190458 · doi ↗ · pubmed ↗
- 2Zingg B, Chou X, lin, Zhang Z et al. gang,. AAV-mediated anterograde transsynaptic tagging: mapping corticocollicular input-defined neural pathways for defense behaviors. Neuron. 2017;93(1):33–47. 10.1016/j.neuron.2016.11.045.10.1016/j.neuron.2016.11.045PMC 553879427989459 · doi ↗ · pubmed ↗