Subtype-specific modulation of inhibitory interneurons by general anesthetics
Taisuke Sugino, Takuya Okada, Yuki Nomura, Riko Nakayama, Midori Harada, Nobuhiro Nakai, Norihiko Obata, Satoshi Mizobuchi
TL;DR
This study shows how different anesthetics affect brain cells, with ketamine uniquely activating certain neurons unlike other anesthetics.
Contribution
The study reveals ketamine's unique activation of excitatory and SST interneurons compared to isoflurane and propofol.
Findings
All anesthetics suppress excitatory and inhibitory neuronal activity in the cortex.
Ketamine preferentially activates SST interneurons compared to isoflurane and propofol.
Ketamine induces a distinct cortical reconfiguration compared to GABAergic anesthetics.
Abstract
The mechanisms by which general anesthetics induce loss of consciousness remain unclear, particularly regarding their cell-type-specific effects on cortical circuits. Using in vivo two-photon calcium imaging, we compared the effects of isoflurane, propofol, and ketamine—administered at equivalent sedative depth based on behavioral reflex suppression—on excitatory neurons and inhibitory interneuron subtypes (PV and SST) in the mouse somatosensory cortex. All anesthetics suppressed excitatory and inhibitory neuronal activity at the population level. However, ketamine uniquely increased the activity of a larger fraction of excitatory neurons (17.6 ± 3.1%) than isoflurane (4.0 ± 1.9%) or propofol (4.1 ± 2.1%). Similar effects were observed in inhibitory neurons. While PV interneurons showed no anesthetic-specific differences, SST interneurons were preferentially activated by ketamine (25.9…
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
Figure 6
Figure 7
Figure 8Peer 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
TopicsNeuroscience and Neuropharmacology Research · Cancer, Stress, Anesthesia, and Immune Response · Neuropeptides and Animal Physiology
Introduction
General anesthesia induces loss of consciousness, analgesia, and immobility, thereby protecting the body from physiological stress during surgical procedures.1^,^2 Inhalational and intravenous anesthetics are thought to abolish consciousness primarily by suppressing neural activity within the central nervous system. However, the precise mechanisms—particularly at the level of neural circuits—remain incompletely understood.
Recent studies have increasingly focused on how general anesthetics affect excitatory and inhibitory interneurons in the cerebral cortex.3^,^4 Among these, GABAergic interneurons expressing parvalbumin (PV) and somatostatin (SST) play a critical role in finely regulating the cortical excitation-inhibition (E/I) balance in cortical networks.5^,^6 PV interneurons predominantly target the perisomatic regions and axon initial segments, exerting strong and temporally precise inhibitory control.7^,^8 In contrast, SST interneurons preferentially innervate dendritic compartments, providing input-specific inhibition.9
Accumulating evidence suggests that different anesthetics may exert distinct effects on these inhibitory interneuron subtypes. For instance, isoflurane has been reported to suppress the activity of a broad range of neuronal populations, including both PV and SST interneurons, through the activation of GABA_A_ receptors, leading to widespread cortical network suppression.10 Conversely, ketamine acts primarily via NMDA receptor antagonism and has been shown to selectively suppress inhibitory interneuron activity—particularly that of PV neurons—resulting in the disinhibition of excitatory neurons.11 More recently, ketamine has also been reported to suppress spontaneously active excitatory neurons while concurrently activating previously silent ones.12 These cell type-specific effects may significantly reshape cortical circuit dynamics and modulate states of consciousness.
However, many of these findings are based on in vitro experiments or population-level recordings. Studies that directly observe the activity of specific neuronal subtypes at single-cell resolution in the intact brain remain scarce.
In this study, we used in vivo two-photon calcium imaging to investigate how three anesthetics with distinct mechanisms of action— isoflurane, propofol, and ketamine—modulate the activity of excitatory neurons and PV- and SST-positive inhibitory interneurons in the primary somatosensory cortex (S1) of mice. By visualizing cell type-specific and time-resolved neuronal activity patterns in real time, we aimed to reveal the diverse and selective effects of anesthetics on cortical circuits.
This study provides a foundation for understanding how general anesthetics alter cortical function at the level of specific neuronal populations and offers novel insights into the circuit-level mechanisms underlying anesthesia-induced unconsciousness.
Results
Behavioral assessment of anesthetic depth
To investigate the effects of different general anesthetics on anesthetic depth in mice, behavioral assessments were first performed. The proportions of mice exhibiting loss of righting reflex and loss of pedal withdrawal reflex were evaluated at 5, 10, and 15 min after the administration of each anesthetic.13^,^14^,^15 Behavioral assessments were performed on separate days for each anesthetic (Figures 1A and 1B). Based on previous studies,12^,^13^,^16^,^17 we hypothesized that 1% inhaled isoflurane, 150 mg/kg intraperitoneal propofol, and 50 mg/kg intraperitoneal ketamine would produce comparable anesthetic depth. To validate this, anesthetic depth was behaviorally assessed under different dosing regimens: isoflurane at 0.5%, 1%, and 2% (inhalation); propofol at 75, 150, and 300 mg/kg (i.p.); and ketamine at 25, 50, and 100 mg/kg (i.p.). The results confirmed that 1% isoflurane, 150 mg/kg propofol, and 50 mg/kg ketamine induced comparable levels of anesthetic depth, corresponding to a stable sedative state that was slightly lighter than full surgical anesthesia (Figures 1C, 1D, and S1A–S1D), and these doses were subsequently used in all imaging experiments.Figure 1. Experimental setup and anesthetic-induced behavioral endpoints(A) Schematic representation of the experimental setup for two-photon imaging combined with EEG/EMG recordings during general anesthesia.(B) Experimental timeline from cranial window surgery to two-photon imaging and behavioral assessments. Imaging and behavioral tests for each anesthetic were conducted at least 24 h apart to minimize potential carryover effects.(C) Probability of loss of righting reflex (LORR) at 5, 10, and 15 min after the administration of isoflurane (1%), propofol (150 mg/kg), or ketamine (50 mg/kg).(D) Probability of loss of pedal withdrawal reflex (LOPWR) at 5, 10, and 15 min after the administration of isoflurane (1%), propofol (150 mg/kg), or ketamine (50 mg/kg). (C and D) n = 4 trials from 8 mice. Statistical significance was assessed using one-way ANOVA with Tukey’s multiple-comparisons test. ns, not significant. Error bars represent mean ± SEM. Iso, isoflurane; Ket, ketamine; Prop, propofol; PWR, pedal withdrawal reflex; RR, righting reflex.
Effects of general anesthetic administration on excitatory neuronal activity in the cerebral cortex
To investigate the effects of general anesthetics on excitatory neuronal activity in the cerebral cortex, we performed in vivo two-photon calcium imaging of L2/3 neurons in the S1. To enable the expression of GCaMP6f in excitatory neurons, wild-type (WT) C57BL/6 mice were injected with AAV1-CaMKII-GCaMP6f in the S1 region prior to imaging (Figure 2A). CaMKII-positive neurons accounted for approximately 83.7 ± 1.5% of the total neuronal population (Figures 2B and 2C).Figure 2. Isoflurane, propofol, and ketamine suppress excitatory neuronal activity in S1(A) Schematic representation of virus injection (left) and two-photon imaging (right). Scale bars, 50 μm.(B) GCaMP6f expression (green) and NeuN immunostaining (red) in a wild-type (WT) mouse injected with AAV1-CaMKII-GCaMP6f-WPRE-SV40. Fluorescence blur was corrected using the haze reduction function of the fluorescence microscope. Scale bars, 50 μm.(C) Proportion of GCaMP6f-positive neurons among NeuN-positive neurons (n = 8 slices from 4 mice).(D–G) Analysis of excitatory neurons during isoflurane administration (649 neurons from 8 mice).(H–K) Analysis of excitatory neurons during propofol administration (637 neurons from 8 mice).(L–O) Analysis of excitatory neurons during ketamine administration (666 neurons from 8 mice). (D, H, and L) Top: Representative calcium traces from five neurons before and after anesthetic administration. Traces before and after anesthesia are concatenated segments and are not continuous recordings. Bottom: Representative EEG and EMG recordings before and after anesthetic administration. (E, I, and M) Mean powers of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (F, J, and N) Mean correlation coefficients (C.C.) of paired neurons (n = 8 mice) before and after anesthetic administration. (G, K, and O) Mean firing frequencies of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (E–G, I–K, and M–O) Statistical significance was assessed using the Wilcoxon signed-rank test. ∗p < 0.05; ∗∗p < 0.01; ns, not significant. Error bars represent mean ± SEM. AAV, adeno-associated virus; EEG, electroencephalography; EMG, electromyography; Iso, isoflurane; Ket, ketamine; Prop, propofol.
When the data were analyzed using mean values of pCa^2+^, correlation coefficient (C.C.), and firing frequency calculated for each mouse, inhalation of isoflurane significantly reduced the power of calcium transients (pCa^2+^) from 5.5 ± 0.84 to 0.47 ± 0.083 (p = 0.0078), and firing frequency from 0.032 ± 0.0020 to 0.0022 ± 0.00064 Hz (p = 0.0078). The correlation coefficient (C.C.) decreased from 0.58 ± 0.096 to 0.44 ± 0.074, although this change was not statistically significant (p = 0.109) (Figures 2D–2G). In this study, “C.C.” refers to the correlation coefficient calculated between neuronal calcium activity traces, which reflects the degree of synchronous activity across neuronal populations. Because calcium imaging data include long periods of neuronal silence, cosine correlation was used instead of the conventional Pearson’s correlation to more robustly capture temporal similarity in neuronal activity patterns. Similarly, the intraperitoneal administration of propofol significantly decreased pCa^2+^ from 4.3 ± 0.70 to 0.47 ± 0.055 (p = 0.0078), C.C. from 0.50 ± 0.065 to 0.29 ± 0.053 (p = 0.016), and firing frequency from 0.038 ± 0.0030 to 0.0018 ± 0.00053 Hz (p = 0.0078) (Figures 2H–2K). In addition, the intraperitoneal administration of ketamine significantly reduced pCa^2+^ from 3.3 ± 0.49 to 0.94 ± 0.12 (p = 0.0078) and firing frequency from 0.029 ± 0.0016 to 0.0079 ± 0.0012 Hz (p = 0.0078). C.C. decreased from 0.35 ± 0.080 to 0.16 ± 0.071, without statistical significance (p = 0.148) (Figures 2L–2O). As a control, the intraperitoneal administration of physiological saline (0.9%, 10 mL/kg) produced no significant changes in pCa^2+^, C.C., or firing frequency (Figures S2A–S2D).
Effects of general anesthetic administration on inhibitory neuronal activity in the cerebral cortex
Subsequently, to elucidate the effects of general anesthetics on inhibitory neuronal activity in the cerebral cortex, we conducted in vivo calcium imaging of L2/3 neurons in the S1. GCaMP6f expression was selectively driven in inhibitory neurons by injecting AAV1-mDlx-GCaMP6f into the S1 of WT mice prior to imaging (Figure 3A). The imaged mDlx-positive neurons accounted for approximately 13.0 ± 1.3% of the total neuronal population (Figures 3B and 3C).Figure 3. Isoflurane, propofol, and ketamine suppress inhibitory neuronal activity in S1(A) Schematic representation of virus injection (left) and two-photon imaging (right). Scale bars, 50 μm.(B) GCaMP6f expression (green) and NeuN immunostaining (red) in a WT mouse injected with AAV1-mDlx-GCaMP6f-Fischell-2. Fluorescence blur was corrected using the haze reduction function of the fluorescence microscope. Scale bars, 50 μm.(C) Proportion of GCaMP6f-positive neurons among NeuN-positive neurons (n = 8 slices from 4 mice).(D–G) Analysis of inhibitory neurons during isoflurane administration (632 neurons from 8 mice).(H–K) Analysis of inhibitory neurons during propofol administration (644 neurons from 8 mice).(L–O) Analysis of inhibitory neurons during ketamine administration (661 neurons from 8 mice). (D, H, and L) Top: Representative calcium traces from five neurons before and after anesthetic administration. Traces before and after anesthesia are concatenated segments and are not continuous recordings. Bottom: Representative EEG and EMG recordings before and after anesthetic administration. (E, I, and M) Mean powers of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (F, J, and N) Mean correlation coefficients (C.C.) of paired neurons (n = 8 mice) before and after anesthetic administration. (G, K, and O) Mean firing frequencies of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (E–G, I–K, and M–O) Statistical significance was assessed using the Wilcoxon signed-rank test. ∗∗p < 0.01. Error bars represent mean ± SEM. AAV, adeno-associated virus; EEG, electroencephalography; EMG, electromyography; Iso, isoflurane; Ket, ketamine; Prop, propofol.
When the data were analyzed using the mean values of pCa^2+^, correlation coefficient (C.C.), and firing frequency calculated for each mouse. Isoflurane significantly reduced pCa^2+^ from 3.5 ± 0.72 to 0.38 ± 0.054 (p = 0.0078), C.C. from 0.77 ± 0.055 to 0.38 ± 0.057 (p = 0.0078), and firing frequency from 0.038 ± 0.0039 to 0.0028 ± 0.0010 Hz (p = 0.0078) (Figures 3D–3G). Similarly, propofol significantly decreased pCa^2+^ from 4.4 ± 0.54 to 0.34 ± 0.078 (p = 0.0078), C.C. from 0.79 ± 0.024 to 0.40 ± 0.040 (p = 0.016), and firing frequency from 0.036 ± 0.0019 to 0.0010 ± 0.00037 Hz (p = 0.0078) (Figures 3H–3K). Ketamine also significantly reduced pCa^2+^ from 4.1 ± 0.55 to 0.67 ± 0.092 (p = 0.0078) and firing frequency from 0.040 ± 0.0022 to 0.0076 ± 0.0015 Hz (p = 0.0078). Although the C.C. dropped from 0.69 ± 0.037 to 0.29 ± 0.041, this reduction did not reach statistical significance (p = 0.148; Figures 3L–3O). In the control experiment, saline resulted in no significant alteration in any parameter (Figures S3A–S3D).
Ketamine elicits a higher proportion of activated neurons compared to other general anesthetics
Mathematical analysis revealed that a small subset of neurons in the cerebral cortex exhibited increased activity following the administration of general anesthetics (Figure 4A). Neurons were defined as “activated” when they showed an increase in the power of Ca^2+^ transients after anesthetic administration compared with baseline, and the proportion of activated neurons was calculated as the number of activated ROIs divided by the total number of detected ROIs. Because GCaMP6f fluorescence underestimates true spike frequency, power was used as the primary index of neuronal activity rather than frequency. Among excitatory neurons, isoflurane activated 4.0 ± 1.9% of the total population, propofol activated 4.1 ± 2.1%, and ketamine activated 17.6 ± 3.1%. The proportion under ketamine was significantly higher than that observed with isoflurane (p = 0.010) or propofol (p = 0.018) (Figures 4B–4F).Figure 4. Subsets of excitatory and inhibitory neurons are activated by general anesthetics, particularly ketamine(A) Classification of neuronal calcium activity with representative ΔF/F traces. Neurons suppressed by anesthesia are indicated by blue circles, and activated neurons are indicated by red circles. Scale bars, 50 μm.(B–D) Line scatterplots showing changes in activity of individual excitatory neurons before and after the administration of isoflurane (B), propofol (C), and ketamine (D).(E) Percentages of excitatory neurons activated by each anesthetic (n = 8 mice per group).(F) Pie charts showing proportions of activated (ΔPower > 0) and suppressed (ΔPower < 0) excitatory neurons following anesthetic administration. ΔPower was calculated as the difference in pCa^2+^ values before and after anesthesia.(G–I) Line scatterplots showing changes in activity of individual inhibitory neurons before and after the administration of isoflurane (G), propofol (H), and ketamine (I).(J) Percentages of inhibitory neurons activated by each anesthetic (n = 8 mice per group).(K) Pie charts showing proportions of activated (ΔPower > 0) and suppressed (ΔPower < 0) inhibitory neurons following anesthetic administration. (E and J) Statistical significance was assessed using the Kruskal–Wallis test followed by Dunn’s multiple-comparisons test (∗p < 0.05; ∗∗p < 0.01; ns, not significant). Error bars represent mean ± SEM. Iso, isoflurane; Ket, ketamine; Prop, propofol.
Similarly, among inhibitory neurons, isoflurane activated 2.7 ± 0.90% of the population, propofol activated 1.8 ± 0.86%, and ketamine activated 11.0 ± 1.8%. The proportion of activated inhibitory neurons under ketamine was significantly higher than under isoflurane (p = 0.018) or propofol (p = 0.002; Figures 4G–4K).
Effects of general anesthetics on parvalbumin-positive neuronal activity
To further assess the effects of general anesthetics on a key subtype of cortical inhibitory neurons, we examined the activity of PV-positive interneurons using in vivo calcium imaging in L2/3 of the S1 cortex. PV-Cre mice were used, and GCaMP6f was selectively expressed in PV-positive neurons by injecting AAV1-CAG-Flex-GCaMP6f into the S1 region prior to imaging (Figure 5A). The imaged PV-positive neurons constituted approximately 32.5 ± 2.5% of the total GABAergic neuronal population (Figures 5B and 5C).Figure 5. Isoflurane, propofol, and ketamine suppress parvalbumin (PV) neuronal activity in S1(A) Schematic representation of virus injection (left) and two-photon imaging (right). Scale bars, 50 μm.(B) GCaMP6f expression (green) and GABA immunostaining (red) in a PV-Cre mouse injected with AAV1-CAG-flex-GCaMP6f-WPRE-SV40. Fluorescence blur was corrected using the haze reduction function of the fluorescence microscope. Scale bars, 50 μm.(C) Proportion of GCaMP6f-positive neurons among NeuN-positive neurons (n = 8 slices from 4 mice).(D–G) Analysis of PV neurons during isoflurane administration (601 neurons from 8 mice).(H–K) Analysis of PV neurons during propofol administration (608 neurons from 8 mice).(L–O) Analysis of PV neurons during ketamine administration (615 neurons from 8 mice). (D, H, and L) Top: Representative calcium traces from five neurons before and after anesthetic administration. Traces before and after anesthesia are concatenated segments and are not continuous recordings. Bottom: Representative EEG and EMG recordings before and after anesthetic administration. (E, I, and M) Mean powers of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (F, J, and N) Mean correlation coefficients (C.C.) of paired neurons (n = 8 mice) before and after anesthetic administration. (G, K, and O) Mean firing frequencies of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (E–G, I–K, and M–O) Statistical significance was assessed using the Wilcoxon signed-rank test. ∗∗p < 0.01. Error bars represent mean ± SEM. AAV, adeno-associated virus; EEG, electroencephalography; EMG, electromyography; Iso, isoflurane; Ket, ketamine; Prop, propofol; PV, parvalbumin.
When the data were analyzed using mean values of pCa^2+^, correlation coefficient (C.C.), and firing frequency calculated for each mouse, isoflurane resulted in significant reductions in all three parameters compared to pre-anesthetic baseline levels: pCa^2+^ decreased from 5.2 ± 1.2 to 0.49 ± 0.086 (p = 0.0078), C.C. of neuronal activity decreased from 0.74 ± 0.030 to 0.39 ± 0.056 (p = 0.0078), and firing frequency decreased from 0.042 ± 0.0031 to 0.0029 ± 0.00087 Hz (p = 0.0078) (Figures 5D–5G). Similarly, propofol led to significant decreases in neuronal activity parameters relative to baseline levels: pCa^2+^ decreased from 3.5 ± 0.80 to 0.46 ± 0.39 (p = 0.0078), C.C. decreased from 0.69 ± 0.061 to 0.39 ± 0.081 (p = 0.0078), and firing frequency decreased from 0.033 ± 0.0023 to 0.0038 ± 0.0015 Hz (p = 0.0078) (Figures 5H–5K). Likewise, ketamine significantly reduced pCa^2+^ from 5.9 ± 1.1 to 0.80 ± 0.24 (p = 0.0078), C.C. from 0.66 ± 0.035 to 0.30 ± 0.069 (p = 0.0078), and firing frequency from 0.035 ± 0.0015 to 0.0040 ± 0.0012 Hz (p = 0.0078) (Figures 5L–5O). In the control experiment, saline resulted in no significant alteration in any parameter (Figures S4A–S4D).
Effects of general anesthetics on somatostatin-positive neuronal activity
In addition, to investigate the effects of general anesthetics on SST-positive inhibitory neurons, a key inhibitory neuronal subtype in the cerebral cortex, we performed in vivo calcium imaging of L2/3 neurons in the S1. SST-Cre mice were used, and GCaMP6f was specifically expressed in SST-positive neurons via AAV1-CAG-Flex-GCaMP6f injection into the S1 region prior to imaging (Figure 6A). The imaged SST-positive neurons constituted approximately 30.3 ± 1.2% of the total GABAergic neuronal population (Figures 6B and 6C).Figure 6. Isoflurane, propofol, and ketamine suppress somatostatin (SST) neuronal activity in S1(A) Schematic representation of virus injection (left) and two-photon imaging (right). Scale bars, 50 μm.(B) GCaMP6f expression (green) and GABA immunostaining (red) in an SST-Cre mouse injected with AAV1-CAG-flex-GCaMP6f-WPRE-SV40. Fluorescence blur was corrected using the haze reduction function of the fluorescence microscope. Scale bars, 50 μm.(C) Proportion of GCaMP6f-positive neurons among NeuN-positive neurons (n = 8 slices from 4 mice).(D–G) Analysis of SST neurons during isoflurane administration (460 neurons from 8 mice).(H–K) Analysis of SST neurons during propofol administration (470 neurons from 8 mice).(L–O) Analysis of SST neurons during ketamine administration (540 neurons from 8 mice). (D, H, and L) Top: Representative calcium traces from five neurons before and after anesthetic administration. Traces before and after anesthesia are concatenated segments and are not continuous recordings. Bottom: Representative EEG and EMG recordings before and after anesthetic administration. (E, I, and M) Mean powers of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (F, J, and N) Mean correlation coefficients (C.C.) of paired neurons (n = 8 mice) before and after anesthetic administration. (G, K, and O) Mean firing frequencies of Ca^2+^ transients (n = 8 mice) before and after anesthetic administration. (E–G, I–K, and M–O) Statistical significance was assessed using the Wilcoxon signed-rank test. ∗p < 0.05; ∗∗p < 0.01; ns, not significant. Error bars represent mean ± SEM. AAV, adeno-associated virus; EEG, electroencephalography; EMG, electromyography; Iso, isoflurane; Ket, ketamine; Prop, propofol; SST, somatostatin.
When the data were analyzed using mean values of pCa^2+^, correlation coefficient (C.C.), and firing frequency calculated for each mouse, isoflurane resulted in significant reductions in all three parameters: pCa^2+^ decreased from 2.9 ± 0.42 to 0.40 ± 0.084 (p = 0.0078), C.C. decreased from 0.73 ± 0.055 to 0.31 ± 0.052 (p = 0.0078), and firing frequency decreased from 0.030 ± 0.0054 to 0.0023 ± 0.00096 Hz (p = 0.0078) (Figures 6D–6G). Similarly, propofol led to significant reductions in pCa^2+^ from 2.1 ± 0.42 to 0.26 ± 0.038 (p = 0.0078), C.C. from 0.75 ± 0.030 to 0.28 ± 0.058 (p = 0.0078), and firing frequency from 0.033 ± 0.0057 to 0.0018 ± 0.00051 Hz (p = 0.0078) (Figures 6H–6K). In contrast, ketamine significantly decreased pCa^2+^ from 2.2 ± 0.47 to 0.98 ± 0.18 (p = 0.0156) and C.C. from 0.43 ± 0.073 to 0.27 ± 0.076 (p = 0.0391), whereas firing frequency remained unchanged (0.021 ± 0.0022 to 0.017 ± 0.0038 Hz; p = 0.547) (Figures 6L–6O). In the control experiment, saline resulted in no significant alteration in any parameter (Figures S5A–S5D).
Somatostatin-positive rather than parvalbumin-positive neurons drive ketamine-induced inhibitory activation
To determine which inhibitory neuronal subtype contributes to the increased activity observed following ketamine administration, we quantified the proportion of activated PV- and SST-positive neurons under different anesthetic conditions.
Among PV-positive neurons, isoflurane activated 3.9 ± 0.72% of the population, propofol activated 6.2 ± 3.2%, and ketamine activated 6.2 ± 2.3%. There were no significant differences in the proportions of activated PV neurons among the three anesthetic conditions (all p > 0.999) (Figures 7A–7E).Figure 7. Ketamine preferentially activates SST neurons compared to PV neurons(A–C) Line scatterplots showing changes in PV neuron activity before and after the administration of isoflurane (A), propofol (B), and ketamine (C).(D) Percentages of PV neurons activated by each anesthetic (n = 8 mice per group).(E) Pie charts showing proportions of activated (ΔPower > 0) and suppressed (ΔPower < 0) PV neurons following anesthetic administration.(F–H) Line scatterplots showing changes in SST neuron activity before and after the administration of isoflurane (F), propofol (G), and ketamine (H).(I) Percentages of SST neurons activated by each anesthetic (n = 8 mice per group).(J) Pie charts showing proportions of activated (ΔPower > 0) and suppressed (ΔPower < 0) SST neurons following anesthetic administration.(K–M) Comparison of percentages of activated PV and SST neurons following anesthetic administration. Statistical significance was assessed using the Mann–Whitney U test (∗p < 0.05; ns, not significant). Error bars represent mean ± SEM.(N) Schematic summary of cortical circuit effects of isoflurane, propofol, and ketamine. Isoflurane and propofol uniformly suppress inhibitory interneurons (SST and PV) and excitatory pyramidal neurons (PYR). In contrast, ketamine preferentially activates SST interneurons and partially activates a subset of pyramidal neurons, while overall neuronal activity remains suppressed. Arrows indicate increases or decreases in activity. (D and I) Statistical significance was assessed using the Kruskal–Wallis test followed by Dunn’s multiple-comparisons test (∗p < 0.05; ∗∗p < 0.01; ns, not significant). Error bars represent mean ± SEM. Iso, isoflurane; Ket, ketamine; Prop, propofol; PV, parvalbumin; SST, somatostatin.
In contrast, among SST-positive neurons, isoflurane activated 4.0 ± 1.8%, propofol activated 1.6 ± 0.60%, and ketamine activated 25.9 ± 6.0% of the population. The proportion of activated SST neurons was significantly higher under ketamine compared to both isoflurane (p = 0.029) or propofol (p = 0.003) (Figures 7F–7J).
These findings suggest that the increased inhibitory activity following ketamine administration (Figures 4G–4K) is primarily attributable to SST-positive neurons rather than PV-positive neurons. Direct comparison between PV and SST populations under each anesthetic condition further supports this conclusion: ketamine significantly increased activation in SST-positive neurons compared to PV-positive neurons (p = 0.037), whereas no such differences were observed under isoflurane (p = 0.572) or propofol (p = 0.957) (Figures 7K–7N).
Discussion
In this study, we used in vivo two-photon calcium imaging to examine how three general anesthetics—isoflurane, propofol, and ketamine—modulate cortical neuronal activity in mice, focusing on excitatory neurons and inhibitory interneurons expressing PV or SST. All anesthetics broadly suppressed neuronal activity and reduced population-level synchrony. However, ketamine uniquely increased the proportion of activated neurons, particularly among excitatory neurons and SST-positive interneurons.
These findings challenge the conventional notion that general anesthetics uniformly suppress cortical activity. While overall activity was markedly reduced, ketamine produced nonuniform, cell-type-specific alterations within the cortical network. The increased activity observed in subsets of excitatory and SST neurons likely reflects locally heterogeneous modulation within cortical circuits, rather than large-scale network reorganization.
At first glance, the enhanced SST activity observed under ketamine seems inconsistent with previous studies reporting SST suppression.12^,^18^,^19 However, these earlier studies primarily used subanesthetic doses (3–10 mg/kg) in antidepressant paradigms. In contrast, we administered a surgical anesthetic dose (50 mg/kg), which induces a widespread suppression of cortical activity across the imaged population. While a previous study by Cichon et al. (2022) used a similar dose, their analysis was based on population-averaged data. In contrast, our single-cell-level approach revealed that, despite global suppression, distinct subsets of SST neurons became activated under ketamine.
The proportion of activated pyramidal neurons observed in our study was smaller than previously reported under similar conditions.10^,^12 This difference may arise from methodological and temporal factors. Our analysis focused on a relatively stable anesthetic state (5–15 min post-administration), during which cortical activity is largely suppressed, whereas prior studies may have included more dynamic early induction phases. Differences in anesthetic depth, administration routes, or cortical region could also contribute to variability. Nonetheless, by analyzing a well-defined, stable anesthetic phase in the primary somatosensory cortex (S1)—a region with well-characterized architecture and functional relevance to sensory processing—our study provides a consistent framework for evaluating cell-type-specific modulation of cortical activity under anesthesia.
SST interneurons, which modulate dendritic excitatory input and express high levels of NMDA receptors, are particularly sensitive to input-dependent modulation.5^,^20 Although NMDA receptor blockade by ketamine may directly suppress SST activity, secondary effects such as altered intracortical inhibition or extrinsic input changes could lead to disinhibition and paradoxical activation in specific subsets. Similar mechanisms may underlie the heterogeneous responses among excitatory neurons. In contrast, PV interneurons were uniformly suppressed, possibly reflecting their distinct circuit roles. Together, these findings indicate that ketamine modulates cortical activity through cell type-specific and locally mediated circuit interactions, rather than uniform network-level effects.
The differential effects observed among anesthetics likely reflect their distinct pharmacological mechanisms. Isoflurane and propofol primarily potentiate GABA_A_ receptor-mediated inhibition, while also modulating potassium conductance and synaptic transmission, thereby reducing both excitatory drive and network synchrony.21^,^22 In contrast, ketamine acts as an NMDA receptor antagonist but additionally influences AMPA receptor signaling, monoaminergic and opioid pathways, and HCN channels.23^,^24 These broad interactions may account for the locally heterogeneous activation patterns observed here, distinguishing ketamine from classical GABAergic anesthetics.
General anesthetics reduced neuronal synchrony, as indicated by decreased cosine correlation values. This suggests a transition toward more asynchronous firing and reduced temporal coordination within local cortical populations. Some prior studies reported anesthesia-induced increases in synchrony, but these often relied on Pearson’s correlation, which can overestimate coactivity by including silent periods.4^,^25 In contrast, cosine correlation emphasized active co-firing, providing a complementary measure of network coordination. Notably, synchrony increases have been reported at higher isoflurane concentrations (>1.5%), whereas the lower concentration (1%) used here may not elicit the same effect.25 Moreover, synchrony may vary across cortical layers, as previously reported.4 Although two-photon microscopy allows imaging down to ∼500–600 μm (layer 5), our analyses focused on layer 2/3 neurons, where calcium signals are most stable and reliably detected. Accordingly, the present findings primarily reflect activity in superficial cortical layers.
In conclusion, ketamine globally suppresses cortical activity yet selectively enhances the activity of specific subsets of excitatory and SST-positive neurons. These cell-type-specific and locally heterogeneous alterations may reflect unique microcircuit mechanisms driven by NMDA receptor blockade. Together, our findings advance the understanding of cellular and network mechanisms underlying general anesthesia and provide insight into ketamine’s distinct neurophysiological profile among anesthetic agents.
Limitations of the study
This study has several limitations. First, as an observational study, we did not directly explore the mechanisms underlying the differential effects of ketamine. Although our findings point toward cell-type- and circuit-dependent modulation, causality remains speculative. Moreover, while we focused on ketamine’s NMDA receptor antagonism, the drug also interacts with AMPA receptors, monoaminergic systems, GABAergic pathways, calcium channels, and HCN channels.23^,^24 Future studies incorporating optogenetic or chemogenetic approaches combined with pharmacological dissection will be essential to identify specific mechanisms.
Second, we could not capture the earliest phase of anesthetic induction (0–5 min post-injection) due to the time required to configure the imaging system. As a result, our data reflect only the stable phase of anesthesia, potentially overlooking transient activity dynamics crucial for understanding the transition into unconsciousness. Prior studies have shown that neural activity during the induction phase differs substantially from the stable state: Mukamel et al. (2014) and Lee et al. (2017) demonstrated that propofol induction involves rapid spectral and connectivity changes before reaching steady unconsciousness, while Scaglione et al. (2024) reported evolving cortical synchrony and activity transitions under isoflurane.26^,^27^,^28 Together, these findings indicate that the early post-administration period represents an unstable neural state, distinct from the later steady condition. Therefore, our analyses primarily reflect stable anesthesia, during which behavioral suppression and anesthetic depth were consistent, and transient variability was minimized. Future work combining real-time imaging with behavioral monitoring from the onset of anesthesia will be valuable to link transient network dynamics with loss of consciousness. Neuronal activity was analyzed during the 5–15 min post-administration period, which we defined as a relatively stable anesthetic phase based on behavioral stabilization and consistency with prior studies.12^,^29^,^30^,^31 Although propofol has not been systematically evaluated regarding its stabilization window in two-photon imaging paradigms, our behavioral assessments (LORR and LOPWR) confirmed that 150 mg/kg propofol produced a hypnotic depth comparable to that of isoflurane and ketamine. Nonetheless, pharmacokinetic differences could still result in subtle discrepancies in anesthetic stability across agents, which we acknowledge as a limitation.
Third, our imaging was limited to the S1 cortex due to its accessibility and stability for two-photon microscopy. While S1 has established relevance to sensory awareness and anesthetic-induced unconsciousness,3^,^32 it remains unclear whether similar effects occur in prefrontal or association cortices that more directly contribute to conscious processing.33^,^34
Fourth, the behavioral and imaging experiments were conducted under different movement conditions. During behavioral assays, mice were head-fixed while anesthetics were administered—using the same delivery systems, concentrations, and routes as in the imaging sessions—and the fixation was briefly released only for assessing reflex responses (LORR and LOPWR). Thus, anesthetic exposure was consistent across conditions. Nevertheless, minor physiological differences—such as posture or respiratory variation—between the transient unrestrained state and the continuous head-fixed imaging state could have influenced anesthetic effects, which we acknowledge as a methodological limitation.
Finally, anesthetic doses were not individually titrated to behavioral thresholds but selected based on widely used surgical-level concentrations. Although these doses reliably induced anesthesia, subtle differences in depth across agents cannot be excluded. Future studies should include dose-response calibration to ensure equivalent anesthetic depth among conditions.
Resource availability
Lead contact
Further information and requests for resources and reagents should be directed to the lead contact, Takuya Okada ([email protected]).
Materials availability
This study did not generate new unique reagents. All mouse lines were obtained from The Jackson Laboratory, all adeno-associated viruses were obtained from Addgene, and all antibodies used were commercially available.
Data and code availability
- •All data generated and analyzed during this study are available from the lead contact upon request.
- •Custom scripts used for data analysis in GraphPad Prism and MATLAB are available from the lead contact upon request.
- •Some analyses were performed using proprietary software (GraphPad Prism and MATLAB); access to these software tools may be required to replicate the results.
- •Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.
- •This article does not report original code.
Acknowledgments
We thank James M. Wilson for providing AAV1-CaMKII-GCaMP6f-WPRE-SV40, Gordon Fishell for providing AAV1-mDlx-GCaMP6f-Fishell-2, and Douglas Kim for providing AAV1-CAG-flex-GCaMP6f-WPRE-SV40. This research was supported by 10.13039/501100001691JSPS KAKENHI Grant Numbers 22K09022, 23K08380, and 25K12169.
Author contributions
T.S., T.O., and S.M. designed the research. T.S., T.O., R.N., M.H, N.N., and N.O. performed research and analyzed the data. T.S., T.O., Y.N., and S.M. wrote the article.
Declaration of interests
The authors declare no competing interests.
Declaration of generative AI and AI-assisted technologies in the writing process
During the preparation of this work, the authors used ChatGPT (OpenAI) for English language editing. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
STAR★Methods
Key resources table
REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesRabbit monoclonal anti-NeuNAbcamCat# ab177487; RRID: AB_2532109Rabbit polyclonal anti-GABASigma-AldrichCat#A2052; RRID: AB_477652Alexa Fluor^TM^ 647 goat anti-rabbit IgGThermo Fisher ScientificCat#A-21244; RRID: AB_2535812Bacterial and virus strainsAAV1-CaMKII-GCaMP6f-WPRE-SV40AddgeneAddgene plasmid #100834; AAV1AAV1-mDlx-GCaMP6f-Fischell-2AddgeneAddgene plasmid #83899;AAV1-CAG-flex-GCaMP6f-WPRE-SV40AddgeneAddgene plasmid #100835;Chemicals, peptides, and recombinant proteinsIsofluraneViatris Healthcare901036504KetamineDaiichi SankyoN01AX03XylazineBayerN/APropofolFresenius Kabi JapanN/ANormal salineOtsukaN/AVectaShield with DAPIVector LabsCat#H-1200; RRID: AB_2336790Experimental models: Organisms/strainsMouse: C57BL/6JThe Jackson LaboratoryStrain#000664; RRID: IMSR_JAX:000664Mouse: Pvalb CreTanahira et al.35N/AMouse: Sst-IRES-CreThe Jackson LaboratoryStrain#013044; RRID: IMSR_JAX:013044Software and algorithmsCaImAnGiovannucci et al., 2018http://github.com/simonsfoundation/CaImAnMATLAB_R2022aMathworksRRID: SCR_001622Custom code (MATLAB) to analyze dataOkada et al.36N/AEEG/EMG data collection systemUnique MedicalUAS-308SSleepSignKissei ComtechRRID: SCR_018200ImageJNational Institutes of Health (NIH)RRID: SCR_003070GraphPad Prism 10GraphPad Software. Inc.RRID:SCR_002798Adobe IllustratorAdobe SystemsRRID: SCR_010279OtherSomnoSuite Low-Flow Anesthesia SystemKent Scientifichttps://www.kentscientific.com/products/somnosuite/EEG/EMG amplifierUnique MedicalEBA-100Stereotaxic deviceNarishigeSR-5M-HTStereotaxic micromanipulatorNarishigeSM-15R, SM15M-2Micropipette pullerNarishigePC-100Glass Capillary, 1 mm diameterNarishigeGDC-1MicroinjectorBEXBJ-120, BT-200Coverslip Glass, 2- and 4.5- mm diameterMatsunami GlassN/AMultiphoton microscopeNikonA1R MP systemNikon 25× water immersion objective (1.10 NA)NikonN/ATi:sapphire laserChamereon, Coherenthttps://www.coherent.com/lasers/oscillators/chameleon-ultraFluorescence microscopeKeyenceBZ-X700 All-in-one FluorescenceIR cameraSriHomeN/A
Experimental model and study participant details
All experiments were conducted using male mice older than 8 weeks at the time of surgery. Animals were housed under a 12-h light/dark cycle with ad libitum access to food and water. All experiments, including surgeries, behavioral testing, and two-photon imaging, were performed during the light phase (08:00–20:00). For two-photon imaging, 7–8 mice were used for each neuronal population (excitatory, inhibitory, PV-positive, and SST-positive). For behavioral experiments, each anesthetic group consisted of 8 mice, and each mouse performed four behavioral trials.
The following mouse strains were used: wild type C57BL/6 mice (WT; JAX, #000664), PV-Cre mice (C57BL/6 background35), and Sst-IRES-Cre mice (JAX, #013044). All experimental procedures and animal care protocols were approved by the Animal Care and Use Committees of Kobe University Graduate School of Medicine (Approval No. P211002-R4) and conducted in accordance with institutional and national guidelines.
Method details
Surgery and AAV injection
Surgical procedures were performed under ketamine (74 mg/kg, i.p.) and xylazine (10 mg/kg, i.p.) anesthesia. After skin disinfection and skull exposure, the surface was sealed with tissue adhesive (Vetbond, 3M), and a custom head plate was affixed using dental cement (G-CEM ONE, GC, Japan). A circular craniotomy (2–3 mm diameter) was made over the S1 (0.5 mm posterior, 1.5 mm lateral to bregma) under 1% isoflurane, 1–2 days later, as described previously.37 AAV vectors encoding GCaMP6f were injected into layer 2/3 (150–200 μm depth) at three sites using a glass micropipette (tip diameter, 10 μm). For targeting excitatory neurons, AAV1-CaMKII-GCaMP6f (Addgene; 1.0 × 10^13^ vector genomes/mL, diluted 1:1 in phosphate-buffered saline [PBS], #100834) was used. For GABAergic interneurons, AAV1-mDlx-GCaMP6f-Fischell-2 (Addgene; 7.0 × 10^12^ vg/mL, diluted 1:1 in PBS, #83899) was used. In Cre-driver lines, Cre-dependent expression of GCaMP6f was induced using AAV1-CAG-Flex-GCaMP6f (Addgene; 7.0 × 10^12^ vg/mL, diluted 1:1 in PBS, #100835). Throughout the craniotomy, mice were placed on a heating pad maintained at approximately 40°C to prevent hypothermia.
Drug administration
Anesthetic doses were selected based on previous studies38^,^39 and validated through behavioral assessments (see below). For inhalational anesthesia, mice were exposed to 0.5%, 1%, or 2% isoflurane (Viatris Healthcare) via a nose cone using the SomnoSuite Low-flow Anesthesia System (Kent Scientific). The isoflurane dosing regimen was guided by previous in vivo imaging studies using similar concentrations for induction and maintenance.25^,^40^,^41^,^42 During imaging sessions, isoflurane anesthesia was induced at 2% for 10 min and then maintained at 1% during two-photon imaging. During behavioral experiments, isoflurane was sequentially administered at 2%, 1%, and 0.5% for 10 min each. For intravenous anesthesia, mice received intraperitoneal injection of propofol (10 mg/mL, Fresenius Kabi Japan) at 75, 150, or 300 mg/kg, or racemic ketamine (50 mg/mL, Daiichi Sankyo) at 25, 50, or 100 mg/kg. The ketamine doses were selected with reference to prior behavioral and pharmacological studies demonstrating dose-dependent loss of righting reflex and analgesic effects.43^,^44 For both imaging and behavioral sessions, experiments were initiated 5 min after propofol or ketamine injection, which also allowed sufficient time to adjust the microscope focus and acquisition parameters following intraperitoneal injection. All experiments, including those involving isoflurane, were conducted in ambient room air (∼21% O_2_) without supplemental oxygen. The control group received intraperitoneal injections of normal saline (Otsuka) to control for potential effects of the injection procedure. In some experiments, different anesthetic agents were administered to the same animal. To minimize potential carry-over effects from residual drugs or their metabolites, at least 24 h were allowed between behavioral testing or imaging sessions involving different anesthetics.
Behavioral assessment
To evaluate anesthetic-induced hypnosis and depth, we assessed the loss of righting reflex (LORR) and pedal withdrawal reflex (LOPWR), as described previously. Behavioral tests were conducted on freely moving mice fitted with custom-made head plates. LORR was defined as the inability to right the body within 30 s after being placed in the supine position; LOPWR was defined as the absence of hindlimb withdrawal in response to metacarpal stimulation. Reflexes were assessed at 5, 10, and 15 min following anesthetic administration (n = 8 per dose; three doses per anesthetic). All experiments were repeated four times.
During isoflurane experiments, mice were head-fixed under a nose cone for anesthetic administration using the same setup as in the imaging sessions, and the fixation was briefly released only to assess reflex responses.
EEG/EMG recording
EEG and EMG recordings were performed in separate sessions from the two-photon imaging experiments and were used solely as auxiliary indicators to confirm anesthetic depth and behavioral state. For EEG and EMG recordings, a stainless-steel screw electrode was implanted over the frontal cortex, with a reference screw placed above the right cerebellum. Two insulated EMG wires were inserted into the neck muscles. All implants were secured using dental cement. EEG and EMG signals were amplified, filtered (EEG: 0.5–60 Hz; EMG: 5–1000 Hz), and digitized at a sampling rate of 1000 Hz using an EBA-100 amplifier (Unique Medical, Japan). Sleep and anesthesia states were classified using SleepSign for Animals software (Kissei Comtech, Japan) based on spectral analysis of polygraphic recordings. We confirmed alterations in EEG and EMG signals during anesthesia, such as high-amplitude, low-frequency EEG and reduced EMG activity, as shown in Figures 2, 3, 5, and 6.
Two-photon imaging
Two-photon imaging experiments were initiated after a minimum four-week interval to allow for sufficient expression of adeno-associated viruses. Imaging was performed in the left S1 using a laser scanning microscope (Nikon, A1R MP system) equipped with a water-immersion objective lens (25x, numerical aperture (N.A.) 1.10; Nikon) and a Ti: sapphire laser operating at a 920-nm wavelength. Images were acquired from a field measuring 519.85 μm × 519.85 μm (original scan) at a depth of 150–200 μm below the cortical surface. The pixel size was 1.015 μm. Image frames were acquired every 500 ms, with a total of 1000 consecutive frames collected both before and after anesthetic administration. Throughout imaging sessions, mice were placed on a heating pad maintained at approximately 40°C to prevent hypothermia. Mouse behavior during imaging was monitored using an infrared camera.
Image analysis
Image processing and analysis were conducted using ImageJ (National Institutes of Health) and MATLAB (MathWorks) software packages. Motion correction for focal plane shifts was applied using the TurboReg plugin in ImageJ45 on 500 stitched video frames acquired before and after anesthetic administration. To assess neuronal activity in S1, regions of interest (ROIs) were automatically identified using the CaImAn algorithm (http://github.com/simonsfoundation/CaImAn), which defines ROIs as spatially discrete regions exhibiting fluorescence fluctuations.46 Completely silent neurons (and/or those without GCaMP6f expression) were not detected by this algorithm. Using this approach, we extracted an average of 81 excitatory neurons (Figure 2), 81 inhibitory neurons (Figure 3), 76 PV-positive neurons (Figure 5), and 61 SST-positive neurons (Figure 6) per field of view.
The baseline Ca^2+^ fluorescence (F0) was defined as the 35th percentile of the fluorescence intensity histogram obtained over the entire imaging period, following previously established methods.36^,^47^,^48 Fluorescence traces were calculated as ΔF/F0 (ΔF = F – F0), where F represents the instantaneous fluorescent signal. Ca^2+^ transients were defined as events in which ΔF exceeded 2 standard deviations of F0.47 Due to the limited scanning speed (2 Hz) and the prolonged decay kinetics of GCaMP6 fluorescence, our calcium imaging could not accurately resolve actual neuronal firing rates, and it is therefore likely that some neuronal activity events went undetected across all general anesthetic conditions.
The power of Ca^2+^ transients (pCa^2+^) was defined as the mean peak amplitude of ΔF/F0 events for each neuron, representing the strength of individual neuronal activity.36 Ca^2+^ transient frequency was calculated as the total number of transients divided by the imaging duration.
To quantify population synchrony, the correlation coefficient (C.C.) was calculated between all pairs of neuronal ΔF/F0 traces using cosine correlation, which reflects the similarity of temporal activity patterns irrespective of signal amplitude. This method was adopted following Okada et al. (2021), because cosine correlation provides a more stable measure of functional similarity under sparse neuronal activity than Pearson’s correlation. Prior to this analysis, calcium traces were normalized as ΔF/F0 but were not Z-scored. The average of all pairwise C.C. values within each imaging field was used as the representative C.C. for each mouse.36
Neurons were classified as “activated” if they showed an increase in the power of Ca^2+^ transients after anesthetic administration compared to baseline. The proportion of activated neurons was calculated as the number of activated ROIs divided by the total number of extracted ROIs. Because GCaMP6f fluorescence underestimates true spike frequency due to its temporal limitations, power was used as the primary index of neuronal activity rather than frequency.
Immunohistochemistry
Immunohistochemistry was performed to confirm AAV expression and to quantify excitatory, inhibitory, PV-positive, and SST-positive neuronal populations in the S1. Mice were deeply anesthetized with isoflurane and transcardially perfused with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS). Brain samples were obtained from four mice per group, and eight coronal slices were prepared from each brain for analysis. Brains were post-fixed overnight at 4°C in 4% PFA, cryoprotected sequentially in 20% and 30% sucrose/PBS, and sectioned at 20 μm using a cryostat (Leica Microsystems, CM3050S). Sections were permeabilized with 0.5% Triton X-100/PBS, blocked with 5% goat serum/PBS, and incubated for 2 h at room temperature with primary antibodies: anti-NeuN (1:1000, Abcam, ab177487) and anti-GABA (1:1000, Sigma, A2052). After PBS washes, sections were incubated for 1 h with Alexa Fluor 647-conjugated goat anti-rabbit IgG (1:1000, Thermo Fisher Scientific, A-21244). Stained sections were mounted using VECTASHIELD mounting medium (Vector Laboratories, H-1200), imaged with a fluorescence microscope (Keyence BZ-X700×, Keyence, 20× objective), and analyzed using ImageJ (NIH).
Quantification and statistical analysis
All statistical analyses were performed using GraphPad Prism 10 (GraphPad Software, La Jolla, CA, USA). Data are presented as mean ± standard error of the mean (SEM). For behavioral data, one-way analysis of variance (ANOVA) was used. Imaging data were analyzed using the Wilcoxon signed-rank test, Kruskal–Wallis test, or Mann–Whitney U test, as appropriate.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Urban B.W.Bleckwenn M.Concepts and correlations relevant to general anaesthesia Br. J. Anaesth.89200231610.1093/bja/aef 16412173238 · doi ↗ · pubmed ↗
- 2Rudolph U.Antkowiak B.Molecular and neuronal substrates for general anaesthetics Nat. Rev. Neurosci.5200470972010.1038/nrn 149615322529 · doi ↗ · pubmed ↗
- 3Suzuki M.Larkum M.E.General Anesthesia Decouples Cortical Pyramidal Neurons Cell 1802020666676.e 1310.1016/j.cell.2020.01.02432084339 · doi ↗ · pubmed ↗
- 4Bharioke A.Munz M.Brignall A.Kosche G.Eizinger M.F.Ledergerber N.Hillier D.Gross-Scherf B.Conzelmann K.-K.MacéE.Roska B.General anesthesia globally synchronizes activity selectively in layer 5 cortical pyramidal neurons Neuron 110202220242040.e 1010.1016/j.neuron.2022.03.03235452606 PMC 9235854 · doi ↗ · pubmed ↗
- 5Tremblay R.Lee S.Rudy B.GAB Aergic Interneurons in the Neocortex: From Cellular Properties to Circuits Neuron 91201626029210.1016/j.neuron.2016.06.03327477017 PMC 4980915 · doi ↗ · pubmed ↗
- 6Niethard N.Hasegawa M.Itokazu T.Oyanedel C.N.Born J.Sato T.R.Sleep-Stage-Specific Regulation of Cortical Excitation and Inhibition Curr. Biol.2620162739274910.1016/j.cub.2016.08.03527693142 · doi ↗ · pubmed ↗
- 7Hu H.Gan J.Jonas P.Fast-spiking, parvalbumin+ GAB Aergic interneurons: From cellular design to microcircuit function Science 3452014125526310.1126/science.125526325082707 · doi ↗ · pubmed ↗
- 8Ferguson B.R.Gao W.-J.PV Interneurons: Critical Regulators of E/I Balance for Prefrontal Cortex-Dependent Behavior and Psychiatric Disorders Front. Neural Circ.1220183710.3389/fncir.2018.00037 PMC 596420329867371 · doi ↗ · pubmed ↗