Monitoring Astrocytic Ca2+ Activity in Freely Behaving Mice

Monitoring astrocytic Ca2+ activity is essential to understand the physiological and pathological roles of astrocytes in the brain. However, previous commonly used methods for studying astrocytic Ca2+ activities can be applied in only anesthetized or head-fixed animals, which significantly affects in vivo astrocytic Ca2+ dynamics. In the current study, we combined optic fiber recordings with genetically encoded Ca2+ indicators (GECIs) to monitor astrocytic activity in freely behaving mice. This approach enabled selective and reliable measurement of astrocytic Ca2+ activity, which was verified by the astrocyte-specific labeling of GECIs and few movement artifacts. Additionally, astrocytic Ca2+ activities induced by locomotion or footshock were stably recorded in the cortices and hippocampi of freely behaving mice. Furthermore, this method allowed for the longitudinal study of astrocytic activities over several weeks. This work provides a powerful approach to record astrocytic activity selectively, stably, and chronically in freely behaving mice.


INTRODUCTION
Brain astrocytes are abundant glial cells that tile the central nervous system (Khakh and Sofroniew, 2015). These ubiquitous cells respond to neuronal activity with increased intracellular Ca 2+ concentrations via the activation of various receptors and then release gliotransmitters, which in turn act on neurons (Haydon and Nedergaard, 2014;Bazargani and Attwell, 2016;Zhang and Chen, 2017). By such direct bidirectional interactions, astrocytic Ca 2+ transients serve as significant markers of astrocyte integration into neuronal networks (Halassa and Haydon, 2010;Santello et al., 2019). Increasing evidence indicates that astrocytic Ca 2+ transients are behaviorally relevant (Shigetomi et al., 2016) and required for cognitive functions (Halassa and Haydon, 2010;Santello et al., 2019). Also, the regulation of Ca 2+ dynamics is dysfunctional in neuronal diseases (Kuchibhotla et al., 2009;Verkhratsky et al., 2017). Hence, recording these transients in freely behaving mice is essential to elucidate the role of astrocytes in physiological and pathological processes and diseases, such as learning and memory (Adamsky et al., 2018), hyperactivity (Nagai et al., 2019), depression  and Alzheimer's disease (Verkhratsky et al., 2017).
In this study, we combined optic fiber-based Ca 2+ recording with GECIs and monitored locomotion and footshock-evoked astrocytic activities in freely behaving mice. Furthermore, we chronically recorded astrocytic Ca 2+ transients in the brains of freely behaving mice for several weeks via this method. The results indicate that optic fiber-based Ca 2+ recording combined with GECIs is an ideal approach to dissect the functions of astrocytes in physiological and pathological brain processes.

Animals
C57BL/6J mice (3-4 months old, male) were used in all experiments and purchased from the Laboratory Animal Center at the Third Military Medical University. Mice were housed in groups under a 12-h light/dark cycle with free access to food and water; mice implanted with optic fibers were individually housed. All experimental procedures were performed according to institutional animal welfare guidelines and were approved by the Third Military Medical University Animal Care and Use Committee.

Surgery and Adeno-Associated Virus (AAV) In vivo Microinjections
The mice were anesthetized with 1-2% isoflurane in oxygen and then placed in a stereotactic head frame (RWD Technology Company Limited, China) with a heating pad (37.5-38 • C) underneath. A small vertical incision was made in the skin, and a craniotomy (0.5 mm × 0.5 mm) was performed with a dental drill. AAV5-GfaABC 1 D-cyto-GCaMP6f-SV40 (Cat# AV-5-52925, UPenn vector core; Shigetomi et al., 2016;Adamsky et al., 2018;Nagai et al., 2019;Yu et al., 2020), which is a genetically encoded Ca 2+ indicator driven by the astrocyte-specific GfaABC 1 D promoter (a shorter 681 bp GFAP promoter; Shigetomi et al., 2016;Nagai et al., 2019;Yu et al., 2020), was injected at a volume of 500 nl/site without dilution, while AAV5-GfaABC 1 D-PI-Lck-GFP-SV40 (Cat# AV-5-PV2369, UPenn vector core) was injected as a control. The coordinates for the auditory cortex were as follows: anteroposterior (AP): −3.1 mm; mediolateral (ML): ±3.8 mm; and dorsoventral (DV): −1.25 mm with a 20 degrees slope below the dura. These coordinates for the hippocampal CA1 region were as follows: AP: −2.18 mm; ML: ±1.60; and DV: 1.2 mm from the dura. The injection coordinates were based on the Paxinos and Franklin (2004). Depending on the spread of the reagent injected, 1-3 injection sites were utilized at the DV axis to achieve sufficient coverage of the desired region. A glass micropipette with a tip diameter of 10-20 µm was inserted to infuse the virus. After each injection, the micropipette was held in place for 5-10 min before being slowly retracted. The scalp incision was closed with tissue glue (3M Animal Care Products, Vetbond), and postinjection analgesics were provided for 3 days to facilitate recovery. The experiments were performed during the time window of 4-6 weeks postinjection.

Optic Fiber Setup
A custom-built fiber photometry setup (Figure 2A) was used for recording the astrocytic Ca 2+ transients throughout the entire study Qin et al., 2018;Yao et al., 2019). The fluorescent genetically encoded Ca 2+ indicator (GCaMP6f) was excited by light (wavelength of 488 nm) delivered by a laser (OBIS 488 LX-50 mW, Coherent, Santa Clara, CA, USA), and the beam intensity of the laser was accurately adjusted by the current. Then, the beam was deflected by a dichroic mirror (Di02-R488, Semrock, West Henrietta, NY, USA) and focused on the end of an optic fiber (200 µm diameter, NA 0.48, Doric lenses, Quebec City, QC, Canada) through a collimator. The beam intensity at the tip of the optic fiber was approximately 0.22 mW/mm 2 . The emitted GCaMP6f fluorescence (green) was guided back using the same fiber and then separated from excitation blue light by the same dichroic mirror and a bandpass emission filter (FF01-535/22, Semrock, West Henrietta, NY, USA). Finally, the GCaMP6f fluorescence emission was detected by an avalanche photodiode (Si APD, S2382, Hamamatsu Photonics K.K., Japan). The digitization of the collected GCaMP6f fluorescence signals was achieved by a multifunction I/O device (USB-6002, National Instruments, Austin, TX, USA) with a sampling frequency of 2000 Hz. Data were acquired by customized software written on the LabVIEW platform (National Instrument, Austin, TX, USA).

Optic Fiber Recordings in Freely Behaving Mice
Mice were anesthetized with 1-1.5% isoflurane and then placed in a stereotactic head frame (RWD Technology Company Limited, China) with a heating pad (37.5-38 • C) underneath.
A small vertical incision was made in the skin, and a craniotomy (1 mm × 1 mm) was performed using a dental drill above the previous virus injection sites in the auditory cortex or hippocampus. An optic fiber (200 µm diameter, NA 0.48, Doric lenses, Quebec City, QC, Canada) was glued into a short cannula (ID. 0.51 mm, OD. 0.82 mm), inserted through the craniotomy, and advanced slowly to 50 µm above the previous virus injection sites. The optic fiber was fixed to the mouse skull using dental cement. After implantation of the optic fibers, mice were allowed to recover for 3-5 days before astrocytic Ca 2+ transient recording. Changes in fiber curvature were checked every 6 h to avoid breakage in freely behaving mice. The mice were put into a white rectangular recording chamber in which they could move freely. Every mouse was continuously recorded for 1 h for locomotion-evoked Ca 2+ transients or 10 trials for footshockinduced astrocytic Ca 2+ transients by customized acquisition software based on the LabVIEW platform (National Instrument, Austin, TX, USA). A camera (Aigo AHD-X9, China) was used to simultaneously record mouse behavior at 30 Hz under a spatial resolution of 1,280 × 720 pixels.

Data Analysis
Astrocytic Ca 2+ transients were acquired at the sampling frequency of 2,000 Hz and low-pass filtered by the Butterworth filter. For the optic fiber recordings in freely behaving mice, the relative fluorescence changes (∆f/f = (f − f baseline )/f baseline ) were calculated for Ca 2+ transients, where f baseline was the baseline fluorescence level taken during the current test recording period Yao et al., 2018). A template-matching algorithm was used, taking into account the properties of the rise and decay times of the Ca 2+ signals, to automatically detect the astrocytic Ca 2+ transients . A Ca 2+ transient was regarded as a signal if its amplitude was three times larger than the standard deviation of the noise band Qin et al., 2018;Yao et al., 2018). We converted the image frames into binary format to obtain the mouse shape according to the image intensities and calculated the mouse movement by the change in video clip image relative to body size. Traces of astrocytic Ca 2+ transients were analyzed around the onset of locomotion/footshock or at randomly selected times within the same assay (shuffled data) as a control. The data analysis software was custom-written in MATLAB 2018b (MathWorks, USA).

Statistical Analysis
Data are expressed as the mean ± SEM. Investigators were blinded to group allocation during data analysis. When comparing three or more groups, we use the Kruskal-Wallis test; n.s., no significant difference.

Specific and Efficient Expression of GECIs in Astrocytes
GCaMP6f GECI was employed because previous in vivo characterization studies have demonstrated its higher sensitivity compared to that of commonly used synthetic Ca 2+ dyes (Chen et al., 2013) and revealed extensive Ca 2+ fluctuations within astrocytes (Shigetomi et al., 2013;Srinivasan et al., 2015;Yu et al., 2020). We used serotype 5 of AAV (Ortinski et al., 2010) and the minimal astrocyte-specific GfaABC 1 D promoter (Xie et al., 2010;Shigetomi et al., 2016;Nagai et al., 2019;Yu et al., 2020) to selectively express cytosolic GCaMP6f (cyto-GCaMP6f) within astrocytes located in the cortex and hippocampus of the adult mouse ( Figure 1A). Twenty-one days after the microinjection of AAV5, the sites of cyto-GCaMP6f expression were confirmed by post hoc histology. Then, optic fiber recordings were applied according to the sites of cyto-GCaMP6f expression in the following experiments ( Figure 1B).
This result indicated that microinjection of AAV5 in vivo resulted in reliable, robust, and mosaic expression of GCaMP6f within astrocytes in the cortices and hippocampal CA1 area of adult mice (Figures 1C-F). Based on immunohistochemical analysis, GCaMP6f expression was almost colocalized with the astrocyte marker s100β (Figures 1C,E) and not detected within neurons (Figures 1D,F) in either the cortex or hippocampus. The statistical results further indicated that the GCaMP6f expression specificity was 94.8 ± 2.9% in the cortex and 85.8 ± 2.4% in the hippocampus (Table 1), while its expression efficiency was 94.2 ± 2.7% in the cortex and 90.4 ± 3.5% in the hippocampus ( Table 1), demonstrating that the combination of AAV5 and the GfaABC 1 D promoter selectively and efficiently targets astrocytes.

Astrocytic Ca 2+ Transient Recording in Freely Behaving Mice
We applied the optic fiber-based approach to record astrocytic Ca 2+ transients in freely behaving mice. This fiber recording device is described in our previous work Yao et al., 2018Yao et al., , 2019 and allows the excitation of GCaMP6f and the collection of emitted light (Figure 2A). An optic fiber with a diameter of 200 µm was implanted above GCaMP6f-positive astrocytes ( Figure 2B) 3 weeks after virus injection (AAV5-GfaABC 1 D-cyto-GCaMP6f, Figure 1A). Each mouse was placed in a white, opaque, rectangular chamber 24 h after optic fiber implantation, and its behavior was recorded using a camera placed above the chamber (Figure 2C).
To visualize the astrocytic Ca 2+ transients recorded by optic fibers in freely behaving mice without undue levels of movement artifacts, we compared the optic fiber recording data between the GFP-injected control group (AAV5-GfaABC 1 D-PI-Lck-GFP-SV40) and the GCaMP6f-injected group (AAV5-GfaABC 1 D-cyto-GCaMP6f-SV40). A representative example revealed that movement-related transients were not observed in the GFP-injected control group (Figure 2D), while astrocytic Ca 2+ transients were obvious and stable in the GCaMP6finjected group during free movement ( Figure 2E). The statistical results further confirmed that the mean value of optic fiber-recorded transients was 15.27% in the GCaMP6finjected group (Figure 2F, blue histogram) and only 0.11% in the GFP-injected control group ( Figure 2F, upper right, green histogram). Also, the distributions of the amplitudes of GCaMP6f ( Figure 2F, blue histogram) and GFP fluorescence ( Figure 2F, upper right, green histogram) indicated that the transients recorded in the GCaMP6f group were several 100 times greater than those in the GFP group ( Figure 2F). These experiments showed that the astrocytic Ca 2+ transients recorded by optic fibers in freely behaving mice were without movement artifacts.
To investigate the role of astrocytic Ca 2+ transients in cognition functions, such as exploring and spatial memory, the astrocytic activities, and the real-time locations of mice have to be recorded at the same time. In the present study, we acquire the simultaneous recordings of astrocytic Ca 2+ transients and the motion of freely behaving mice ( Figure 2C). The specific locations where the astrocytic Ca 2+ transients start/end can be clearly labeled in the trajectories of mice according to the simultaneously recorded videos (Figures 2G,H), which cannot be easily done by the traditional imaging methods, such as in vivo two-photon imaging.

Locomotion-Induced Astrocytic Ca 2+ Transients in the Cortices and Hippocampi of Freely Behaving Mice
Previous two-photon imaging studies in head-strained mice have shown that locomotion triggers widespread astrocytic activation in multiple brain regions (Nimmerjahn et al., 2009;Paukert et al., 2014). Next, we aimed to determine whether locomotioninduced astrocytic Ca 2+ transients could be recorded in the cortical and subcortical areas of freely behaving mice by optic fibers. To avoid Ca 2+ signal contamination in nearby brain areas, we recorded astrocytic activities in two brain regions that are separate from each other (cortical area: auditory cortex, subcortical area: hippocampus). Three weeks after virus injection, an optic fiber with a diameter of 200 µm was implanted above GCaMP6f-positive astrocytes in the auditory cortex ( Figure 3A) or the hippocampal CA1 area (Figure 3B), and recordings were performed at least 24 h after fiber implantation. In this study, we simultaneously recorded both astrocytic Ca 2+ transients and body movements during the freely behaving state (Figure 2C). Mouse body movement was recorded with a camera placed above the recording chamber ( Figure 2C) and is shown as black traces (Figures 3C,E). We found reliable locomotion-induced astrocytic Ca 2+ transients in both the cortex ( Figure 3C) and the hippocampus (Figure 3E). Further analysis indicated that most of the astrocytic Ca 2+ transients started approximately 0.3-1.0 s after the onset of locomotion in the cortex ( Figure 3D) and hippocampus ( Figure 3F). The mean values of Ca 2+ transient latencies were 0.81 s in the cortex and 0.66 s in the hippocampus. These data indicated that the locomotion-induced astrocytic Ca 2+ transients shown by previous two-photon imaging studies (Nimmerjahn et al., 2009;Paukert et al., 2014) can also be recorded by optic fibers in freely behaving mice.

Footshock-Evoked Astrocytic Ca 2+ Transients in the Cortices and Hippocampi of Freely Behaving Mice
In a previous study, we applied in vivo two-photon imaging in head-strained mice and demonstrated that cortical astrocytes could be activated by footshock (Zhang et al., 2016;Zhang and Chen, 2017). To assess the astrocytic Ca 2+ transients evoked by footshock in the cortices and hippocampi of freely behaving mice, we used an optic fiber-based approach to record astrocytic Ca 2+ transients in these two brain regions upon application of the footshock stimuli, revealing reliable footshock-evoked astrocytic Ca 2+ transients in both the cortex (Figure 4A) and the hippocampus (Figure 4C). Further analysis indicated that most of the astrocytic Ca 2+ transients started approximately 0.3-1.0 s after the onset of footshock stimuli in the cortex ( Figure 4B) and hippocampus ( Figure 4D). The mean values of Ca 2+ transient latencies were 0.67 s in the cortex and 0.72 s in the hippocampus. These results indicated that footshock-evoked astrocytic Ca 2+ transients can also be reliably recorded by optic fibers in freely behaving mice.

Chronic Recordings of Astrocytic Ca 2+ Transients in the Brains of Freely Behaving Mice
Many physiological or pathological processes, such as neuromodulation, neural development, and learning/memory, are placed on a long-term axis (Aramuni and Griesbeck, 2013). Thus, to fully understand these processes, data must be extrapolated in the same cells during numerous experimental sessions. Because GECIs may allow the establishment of long-term functional biographies of identified cells (Aramuni and Griesbeck, 2013;Hainmueller and Bartos, 2018), we next combined optic fiber-based recording and GECIs to record the Ca 2+ transients of the same population of astrocytes in the brains of freely behaving mice for several weeks (Figure 5A). Astrocytic Ca 2+ transients induced by footshock can be stably recorded for at least 28 days ( Figure 5B). Also, there was no difference between the amplitudes, latencies, rise time, and decay time of astrocytic Ca 2+ transients in response to footshock on different recording days (Figures 5C-F). Therefore, optic fiber-based recordings combined with GECIs are an ideal method for chronic monitoring of astrocytic Ca 2+ transients in the brains of freely behaving mice.

DISCUSSION
In recent decades, extensive in vivo studies have focused on the role of astrocyte Ca 2+ transients in the function of neural circuits (Bazargani and Attwell, 2016;Shigetomi et al., 2016;Zhang and Chen, 2017;Santello et al., 2019;Yu et al., 2020).
The most relevant experiments were performed in anesthetized or head-fixed awake mice with two-photon microscopy (Hirase et al., 2004;Bekar et al., 2008;Nimmerjahn et al., 2009;Ding et al., 2013;Paukert et al., 2014;Zhang et al., 2016;Oe et al., 2020) and camera-based large-field imaging (Ghosh et al., 2013;Daniel et al., 2015). However, increasing evidence supports that astrocytic Ca 2+ transients are affected by anesthesia (Nimmerjahn et al., 2009;Thrane et al., 2012) and restraintinduced stress (Oliveira et al., 2015;Oe et al., 2020). Here, we  combined optic fiber-based recordings and GECIs to monitor astrocyte Ca 2+ activities in the brains of freely behaving mice. We have shown that both locomotion and footshock-evoked astrocytic Ca 2+ transients can be recorded in not only the cortical area but also the subcortical area (hippocampus) of freely behaving mice by optic fibers. More importantly, combined with GECIs, we applied optic fibers to monitor the Ca 2+ transients of the same population of astrocytes in the brains of freely behaving mice for several weeks. Hence, the combination of optic fiber-based recordings and GECIs provides us with an ideal approach to explore the roles of astrocytic Ca 2+ transients in the behaviors and cognition functions of freely behaving mice.
FIGURE 5 | Long-term recording of astrocytic Ca 2+ transients evoked by footshock stimuli in the cortices of freely behaving mice. (A) Schematic of the optic fiber recording in the freely behaving mouse. (B) Footshock-evoked astrocytic Ca 2+ transients in the same mouse on days 1, 5, and 28. (C) Summary of the amplitudes of astrocytic Ca 2+ transients in response to footshock from day 1 to days 20-28. (D) Summary of the latencies of astrocytic Ca 2+ transients in response to footshock from day 1 to days 20-28. (E) Summary of the raise time of astrocytic Ca 2+ transients in response to footshock from day 1 to days 20-28. (F) Summary of the decay time of astrocytic Ca 2+ transients in response to footshock from day 1 to days 20-28. n = 10-25 trials from five mice (C-F). Kruskal-Wallis test; n.s., no significant difference.
Further dissection of the astrocytic Ca 2+ transients will necessitate simultaneously manipulating and recording astrocytic activities. Recently, we introduced a new variant of fiber photometry that simultaneously enables the optogenetic manipulation and real-time recording of Ca 2+ activities in the same population of cells (Li et al., 2017;Qin et al., 2018). At the same time, optogenetic tools have been developed for selectively stimulating or attenuating astrocyte Ca 2+ activities in vivo (Yu et al., 2020), including the light-gated ionotropic glutamate receptor (LiGluR), Channelrhodopsin 2 (ChR2), Melanopsin, and opto-XRs (Yu et al., 2020). Therefore, the combination of these optogenetic tools and optic fiber-based recordings will provide opportunities to simultaneously and selectively control and monitor astrocytic activities in freely behaving mice, which will support further analysis of the roles of astrocytic activities in physiological and pathological processes.
Astrocytes represent a diverse population of cells and display brain area-specific properties (Bayraktar et al., 2014;Chai et al., 2017;Khakh and Deneen, 2019). One recent in vitro study indicated significant differences between astrocytes located in different brain regions in regards to their electrophysiological properties, Ca 2+ signaling, morphology, and astrocyte-synapse proximity (Chai et al., 2017). These differences indicate that astrocytic Ca 2+ monitoring may be a valuable metric for measuring astrocyte diversity at the functional level. However, until now, no in vivo evidence has shown whether astrocytes within different brain neural circuits are largely distinct. In recent years, our group and others have presented multichannel fibers enabling photometry from several to tens of brain regions simultaneously (Guo et al., 2015;Kim et al., 2016;Qin et al., 2019;Sych et al., 2019). More importantly, this approach enables simultaneous multichannel fiber photometry and optogenetic perturbations across many brain regions (Sych et al., 2019). Therefore, multichannel fiber photometry offers opportunities for characterizing astrocytic activities and investigating the diversity of astrocytic functions in different brain regions during behavioral performances. Also, although fiber implantation may activate local astrocytes (Karve et al., 2016), these cells are limited in the superficial area around the end of the optic fiber. Most astrocytes in the deep layers show healthy morphology (Figures 3A,B). Because the optical fiber-based system is collecting an average fluorescence signal from superficial as well as from deeper layers (Grienberger et al., 2012), this technique is an appropriate in vivo approach to monitoring astrocytic signals.
Although it lacks cellular resolution, the optic fiber has attracted considerable attention because of its simplicity and versatility. In the present study, we demonstrate that the combination of optic fiber recordings and GECIs not only allows the selective monitoring of astrocytic activities but can also be applied to study astrocytic activities for long periods in freely behaving animals. In the future, with the help of optogenetic tools and other sophisticated techniques, we anticipate that optic fiber recordings will become more versatile for investigating the complex functions of astrocytes in the mammalian brain under both health and disease conditions.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by the Institutional Animal Care and Use Committee of Third Military Medical University.

AUTHOR CONTRIBUTIONS
HQ, XC, and KZ contributed to the design of the study and interpretation of the data. WH, CY, JL, and TJ carried out the experiments and acquired the data. HQ, WH, CY, JL, TJ, SL, TC, HF, XL, and KZ processed and analyzed the data. XL and KZ wrote the manuscript with help from all other authors. All authors contributed to the article and approved the submitted version.