1. Introduction
General anesthesia has been used since the 1840s and is a necessary safety guarantee for most surgeries today. From destroying the lipid bilayer on the cell surface
[1] to protein targets
[2] and the specific molecular sites on specific receptors
[3], the mechanisms of general anesthetics have been well documented. However, the neural circuits underlying general anesthesia remain relatively unclear, compared with the protein and molecular targets of general anesthetics. Growing evidence suggests that many neural circuits that regulate the sleep–wake cycle are involved in the general-anesthesia effect as well. For example, varieties of neurons and their projections that are important to promote wakefulness, including monoaminergic
[4][5], cholinergic
[6], glutamatergic
[7], and orexinergic neurons
[8], participate in promoting emergence from general anesthesia. The γ-aminobutyric acid (GABA) neurons in the ventrolateral preoptic nucleus (VLPO) extensively innervate and suppress multiple arousal-promoting brain regions
[9][10]. The VLPO is vital for both the initiation and maintenance of sleep
[11]. It is reported that VLPO is necessary for propofol-induced inhibition of locus coeruleus (LC) activity
[12]. Whereas directly specific activation of GABAergic in the VLPO modulates sleep–wake architecture but not anesthetic-state transitions
[13]. Therefore, although there may be some common pathways between general anesthesia and sleep, we are still a long way from fully understanding the neural circuitry of general anesthesia.
An explicit exploration of the role of specific neural circuits under general anesthesia requires advanced neural labeling and modulation technologies. In the following contents, in vivo calcium imaging are introduced. In vivo calcium imaging records the activity of specific neurons and neural circuits in target brain regions
[14].
2. In Vivo Calcium Imaging
Calcium ions are popular targets used to detect neuronal activities that link circuit dynamics to behaviors in modern neuroscience research
[15][16]. Exploiting calcium ion properties, two main categories of calcium-ion indicators have been developed: chemical calcium dyes
[17] and genetically encoded calcium indicators (GECIs)
[18][19]. However, chemical calcium dyes, such as Fluo-4 and Oregon Green BAPTA-1 (OGB-1), are delivered through cell permeabilization, which can damage the cell integrity. Additionally, chemical calcium dyes are normally only capable of recording neuronal activity for several hours. These limitations constrain subsequent imaging conditions
[20][21]. In contrast, GECIs could be easily expressed via virus-delivery methods, such as for adeno-associated-virus or lentivirus vectors, and cause the least cellular damage. Furthermore, GECIs could even be successfully expressed by transgenic methods without invasive procedures
[22][23]. Thus, GECIs can be stably expressed in neurons and allow recording of the neural-firing patterns over a long period of time. Most importantly, GECIs are capable of selective yet unbiased labeling of neuronal types through their specific gene promoters, enabling research on the activities of each neuronal cell type. Moreover, GECIs can also be expressed at the nerve-projection terminals, so in vivo calcium imaging can directly monitor the activity of specific neural circuits
[24]. Thus, in vivo calcium imaging has several advantages over traditional in vivo electrophysiological recording in studying neural endpoints and circuits under general anesthesia. Nowadays, GECIs, such as the GCaMP6s series, which possess a high time sensitivity and fluorescence signal-to-noise ratio
[25], have become one of the most widely used calcium-ion detection tools
[26][27].
GECIs bind to calcium ions and emit fluorescence signals that can be used to determine intracellular calcium concentration
[19] (
Figure 1A). These GECI-based signals can be detected by several methods, including optical-fiber photometry, miniscope imaging and two-photon imaging (
Figure 1B). GECIs fluorescence is stimulated through an optical fiber in optical-fiber photometry. It has the advantages of not restricting animal movements and of monitoring deep-brain regions. Thus, optical-fiber photometry coupled with the GECI technology allows the activities of neurons and neural projections in any deep-brain area related to a specific behavior to be captured in freely moving animals
[28]. Among the three in vivo calcium imaging technologies, optical-fiber photometry is currently the most widely used one in the study of neural-circuit mechanisms in general anesthesia. However, despite many advantages, optical-fiber photometry has its disadvantages. For example, compared with other in vivo calcium imaging techniques, its spatial resolution is relatively low, and it can only monitor changes in fluorescence intensity in neuron populations rather than at the single-neuron level. Therefore, it is difficult to detect relatively tiny excitability changes in cell groups using optical-fiber photometry. Further, the GECI-delivery methods, such as viral transfection or dyes, and the optical-fiber implantation, inevitably cause damage to the brain tissue, especially when targeting deep-brain regions.
Figure 1. Schematic diagram of imaging principles of in vivo calcium imaging. (A) In basal condition: CaM and M13 are not bound to each other, and the fluorescence intensity of EGFP generally remains constant and low; in stimulated condition: the fluorescence intensity of EGFP increases significantly when calcium ions bind directly to EGFP. CaM, calmodulin; M13, calmodulin-binding peptide; EGFP, enhanced green fluorescent protein. (B) The schematic diagram of imaging device principles (top) and calcium-signal characterization for the three in vivo calcium imaging techniques (bottom).
Miniscope imaging has a high spatial resolution, and individual neuronal activity can be readily monitored
[29]. The development of a head-mounted miniscope combined with Gradient Index (GRIN) lens-implantation technology also enables neural activities to be monitored in the deep-brain regions of freely moving animals
[30]. However, viral transfection or lens implantation could also cause brain tissue damage. Additionally, this method is relatively more complex than fiber photometry in both surgical operation and imaging
[31]. Two-photon microscopic-imaging technology with higher resolution is also applied, to identify the neural-population codes underlying complex brain functions
[32]. Traditionally, the monitoring depth of this method is relatively shallow and generally limited to the study of neuronal activity in the cortex or hippocampus. In addition, it requires animal-head fixation, restricting the animal’s free movement
[33][34]. These shortcomings may limit its wide application in the study of anesthesia neural circuitry to some extent. Fortunately, a newly developed miniature two-photon miniscope for large-scale calcium imaging in freely moving mice allows stable simultaneous recording of neuronal dynamics of densely active cortical regions in several behavioral tasks, without impediment to the animal’s behavior
[35]. In summary, despite these limitations, with the continuous improvement of in vivo calcium imaging technology, it remains a valuable fundamental research technique with tremendous potential, especially in its application prospect in anesthesiology, such as the study of neural targets of analgesics and sedatives commonly used in the operating room.
In vivo calcium imaging technologies are widely applied to investigate whether specific neural endpoints or projections associated with the sleep–wake cycle participate in general anesthesia. The cortex is organized into six layers, and excitatory cells within each layer receive inputs from other excitatory cells in the same layer and from inhibitory cells. Simultaneously, the cerebral cortex receives a lot of subcortical projections, whose activity is critical for consciousness
[36][37]. A recent two-photon calcium imaging study found that different anesthetics selectively synchronized activity in cortical pyramidal neurons within layer 5, which may contribute to the loss of consciousness induced by general anesthesia
[38].
Monoaminergic neurons that drive arousal produce dopamine, serotonin, histamine, or norepinephrine, and they extensively innervate many brain regions. These monoaminergic cell groups share similar activity patterns, exhibiting high firing rates during wakefulness and slow firing rates during sleep
[39]. The ventral tegmental area (VTA) mainly contains dopaminergic neurons and projects to abundant arousal-promoting brain areas
[5]. Using real-time in vivo fiber photometry, researchers have shown that calcium signals of VTA dopaminergic neurons significantly decline after sevoflurane-induced loss of righting reflex (LORR) and robustly increase due to recovery of righting reflex (RORR)
[40]. The nucleus accumbens (NAc) is situated in the ventral striatum and receives projections from dopaminergic neurons of VTA
[41]. NAc neurons are similarly inhibited in the induction phase of propofol anesthesia and are markedly activated during recovery, and these effects are mediated by dopamine receptor 1
[40][42]. Moreover, the calcium signals of the ventral periaqueductal gray (vPAG) dopaminergic neurons also decrease during induction and increase during emergence, respectively, with isoflurane anesthesia
[43]. Serotonergic (5-HTergic) neurons in the dorsal raphe nucleus (DRN), which project heavily to the midbrain and forebrain, are implicated in the modulation of the sleep–wake transition
[44]. The calcium activity of DRN 5-HT neurons gradually declines after the initiation of isoflurane administration and begins to restore after the termination of anesthetics inhalation
[45].
Together, these results suggest that similar to the role of promoting wakefulness during the sleep–wake cycle, the dopaminergic and serotonergic neurons in the brain may contribute to the consciousness transition that occurs during general anesthesia with various anesthetics. Two additional important awakening monoaminergic cell groups include the histaminergic neurons in the tuberopapillary nucleus and norepinephrine neurons in the LC. The actions of these neuron populations have not yet been reported using in vivo calcium imaging under general anesthesia. However, recently in a larval-fish model, a two-photon laser-based ablation study indicated that the LC neurons play a regulatory role in both the induction of and emergence from intravenous general anesthesia
[46].
The basal forebrain (BF) is a heterogeneous region containing cholinergic, GABAergic, and glutamatergic neurons, and these BF neurons heavily innervate the cortex. In addition, it is worth noting that BF GABAergic neurons are functionally heterogeneous. Some BF GABAergic neurons are primarily active during the awake state, while others are more active during sleep. However, as a whole, BF plays a vital role in promoting quick cortical activity and arousal
[47][48]. During both isoflurane and propofol anesthesia, calcium signals in BF cholinergic neurons decrease gradually during the induction period, begin to rise during the pre-awakening stage, and peak almost at the moment of RORR. Due to the reliable time accuracy and record stability advantages of optical-fiber photometry technology, researchers have observed that neuronal activity changes always precede behavioral changes. Therefore, BF cholinergic neuronal events may be the impetus for changes in the state of consciousness rather than the target of changes
[49]. Given that the function of GABAergic BF neurons is clearly distinct in the sleep–wake cycle, it is odd that there have been no in vivo calcium imaging studies of these cell groups in general anesthesia. Future experiments are needed to target BF GABA neurons in both the induction to and emergence from general anesthesia.
Glutamatergic neurons in the parabrachial nucleus (PBN) are vital in contributing to both behavioral and cortical electroencephalogram arousal
[50]. An optical-fiber photometry study in rats showed an obvious increase in the activity of PBN neurons during emergence from both isoflurane and propofol anesthesia but no significant change in the induction period
[51]. On the other hand, the lateral habenula (LHb), another brain region clustered with glutamatergic neurons, plays an important role in promoting sleep but not arousal
[52]. The average calcium activity of LHb is significantly increased during isoflurane anesthesia maintenance and begins to decline during RORR. Calcium signals of glutamatergic neurons in LHb show no change during the induction stage
[53]. These interesting findings suggest that the glutamatergic neurons may play distinct roles in general anesthesia and the sleep–wake cycle. Compared with the dopaminergic and cholinergic systems, the glutamatergic system may only participate in the recovery phase, but not the induction phase of general anesthesia. Future research should focus on the responses of glutamatergic neurons and their projections in other brain nuclei during the induction phase of general anesthesia.
The lateral septum contains GABAergic neurons that project to multiple wakefulness-promoting subregions. The calcium activity of the dorsal–intermediate lateral septal GABAergic neuron changes in both the processes of induction and emergence, similar to the trend observed in the sleep–wake cycle
[54]. Collectively, these in vivo calcium imaging findings suggest that multiple brain nuclei and neural circuits known to regulate the sleep–wake cycle play a similar role in general anesthesia (
Table 1).
Table 1. Main findings with the three technologies in the neural nuclei and circuits of general anesthesia.
In vivo calcium imaging has recently been used to explore the neural responses and other effects of anesthetics and to study the neural-circuitry mechanisms during a loss of consciousness induced by general anesthesia. For example, Qiu et al. recently used fiber photometry to explore the role of the VTA in dexmedetomidine-induced sedation
[55]. They demonstrated that selective activation of dopaminergic neurons in the VTA attenuates the depth of sedation in mice. Another study successfully employed optical-fiber photometry and miniscope technology to show that different general anesthetic drugs activate a shared population of central amygdala neurons to potently suppress pain reflexes
[56]. Moreover, optical-fiber photometry and two-photon imaging have been successfully applied to study complications of brain dysfunction caused by different anesthetics
[57][58]. Additionally, based on the principle of genetically encoded calcium ion indicators, several fluorescent proteins that can be used to characterize specific neurotransmitter concentrations have been developed, including the dopamine neurotransmitter probe
[55], glutamatergic neurotransmitter probe
[59], adenosine neurotransmitter probe
[60] and orexin sensor probe
[34]. These improvements provide a broader potential for the application of in vivo calcium imaging in many fields in the future.