ER-GCaMP6f: An Endoplasmic Reticulum Targeted Genetic Probe to Measure Calcium Activity in Astrocytic Processes

Abstract

Ca²⁺ is a major second messenger involved in cellular and subcellular signaling in a wide range of cells including astrocytes which use calcium ions to communicate with other cells in the brain. Even though a variety of genetically encoded Ca²⁺ indicators have been developed to study astrocyte calcium signaling, understanding the dynamics of endoplasmic reticulum calcium signaling is greatly limited by the currently available tools. To address this, we developed an endoplasmic reticulum targeted calcium indicator, ER-GCaMP6f, which is anchored to the cytosolic side of the organelle and measures signaling occurring in close proximity to the endoplasmic reticulum of astrocytes. Using a combination of confocal and super resolution microscopy techniques, we demonstrate the localization of the indicator in the ER in both the cell soma and the processes of astrocytes. Combining ER-GCaMP6f with total internal reflection fluorescence microscopy, we show that Ca²⁺ fluctuations in small astrocytic processes can be detected which are otherwise not observable with existing indicators and standard widefield and confocal techniques. We also compared the ER-GCaMP6f indicator against currently used plasma membrane tethered and cytosolic GCaMP6f indicators. ER-GCaMP6f identifies dynamics in calcium signaling of ER resident receptors that are missed by plasma membrane anchored indicators. We also generated an adeno associated virus (AAV5) and demonstrate that ER-GCaMP6f can be expressed in vivo and measured calcium activity in brain slices. ER-GCaMP6f provides a powerful tool to study calcium signaling in close proximity to the endoplasmic reticulum in astrocyte cell soma and processes both in culture and in brain slices.

Graphical Abstract

Introduction

Calcium signaling is one of the fundamental features of astroglial biology and physiological activity^{1, 2}. Astrocytes are electrically nonexcitable cells and utilize calcium as a major second messenger through which these cells communicate ³. Astrocytes play a key role in synaptic communication where they reuptake synaptically released glutamate and supply it back to presynaptic neurons using calcium as a second messenger system ^{4, 5}. Astrocyte calcium imaging has been significantly advanced after the development of genetically encoded calcium indicators (GECI) which can measure astrocyte specific calcium signaling both in vitro and in vivo by avoiding major limitations of organic calcium dyes ^{6, 7}. These genetically encoded indicators provide powerful information of astrocyte physiology at the cellular level ^{8, 9}. However, current technology is primarily limited to either soluble indicators that reside in the cytoplasm or those anchored to the plasma membrane which makes it difficult to perform subcellular region-specific calcium activity measurements.

Plasma membrane localized indicators provide a way to measure ion channel activity by monitoring Ca²⁺ flux into the astrocyte, while freely diffusing cytosolic indicators allow for the detection of Ca²⁺ release throughout the cell. While the majority of Ca²⁺ is stored in the endoplasmic reticulum (ER) and astrocyte activity results in the release of Ca²⁺ from these intracellular stores, cytosolic indicators are not targeted to the ER and measure ion activity throughout the cell. The endoplasmic reticulum (ER) calcium concentration is usually several thousand times higher than the cytosolic calcium ¹⁰. Inositol 1,4,5-trisphosphate receptors (IP3Rs) and Ryanodine receptors (RyRs) are believed to be the two major family of receptors that regulate ER calcium ¹¹. Even though the ER is a major subcellular organelle involved in calcium signaling and has been observed in the soma and processes, commonly used calcium indicators in astrocytes that target the cytosol or the plasma membrane ¹² do not specifically capture events in close proximity to the ER and do not detect the full range of events in astrocytic processes¹³. Thus, calcium signaling data is difficult to interpret in terms of assigning activity to different subcellular organelles ¹⁴. Indicators that have been developed to target the ER have shown different dynamics than traditional cytosolic and plasma membrane targeted indicators ^{10, 15-17}. Recently, intraluminal GCaMP probes to measure activity on the luminal side of the ER have been developed and utilized to demonstrate the presence of IP3R independent Ca²⁺ activity in astroyctes^{17, 18}. These measurements take place in the background of the high resting concentration of Ca²⁺ in the ER complicating measurements of low-level calcium release from the ER. Calcium activity centered at mitochondria-ER contacts¹⁹ likely serves as a signaling mechanism between these organelles to regulate cellular processes²⁰. In these regions where the ER and mitochondria are in close contact (<30 nm), calcium exchange between the ER and mitochondria does not cause global calcium fluctuation and leads to small amplitude events^{18, 19}. A probe on the cytoplasmic side of the ER would measure low amplitude events in the low background concentration of Ca²⁺ in the cytoplasm making it easier to distinguish events.

Despite recent advancements in the development of Ca²⁺ sensors there are still major limitations in measuring astrocyte Ca²⁺ activity in subcellular regions and astrocytic processes. Current studies lack the ability to delineate Ca²⁺ activity of ER resident receptors or to determine the pharmacological effects on the Ca²⁺activity of such receptors. To overcome the inability of commonly available calcium indicators to specifically measure near ER dynamics, we developed a genetically encoded calcium indicator, ER-GCaMP6f, that is anchored on the cytosolic side of the ER membrane in astrocytes utilizing an ER retention motif from a cytochrome P450²¹. We then generated an AAV5 vector that could be used to specifically label astrocytes in vitro and in vivo. We demonstrate that this indicator can be utilized to study near ER calcium dynamics in both astrocytic cell soma and processes. Using super resolution microscopy, we showed that our indicator is specifically expressed in the ER of both the astrocytic cell soma and processes. ER-GCaMP6f can be utilized in cell culture, can be expressed in vivo to measure calcium activity in live slices, and can be used in combination with total internal reflection fluorescence (TIRF) microscopy to resolve small fluctuations in calcium signaling otherwise undetected by traditional microscopy

In contrast to existing indicators which target the plasma membrane and cytoplasm, we show that ER-GCaMP6f provides calcium dynamics in close proximity to the ER. A direct comparison of Lck-GCaMP6f (Plasma membrane localized) and cytosolic GCaMP6f revealed that intracellular calcium events detected by both showed similar dynamics. However, ER-GCaMP6f measures near ER calcium signaling revealing different dynamics for both spontaneous and stimulated events than those observed with existing indicators. These findings provide a new tool to study astrocyte physiology, and we believe it will be immensely useful to the study of astrocyte activity.

Experimental Methods

Animals

Wild type C57BL/6 mice and Sprague-Dawley rats were housed with ad libitum water and food access and kept on a 12 hr light/dark cycle. The experimental protocols were all consistent with the guidelines issued by the U.S. National Institutes of Health (NIH) and were approved by the University of Kentucky Institutional Animal Care and Use Committee.

Construction and expression of ER-GCaMP6f in HEK-293 cells

ER-GCaMP6f plasmid was constructed by utilizing the endoplasmic reticulum targeting motif (27 amino acids) of a cytochrome p450 ^{21, 22}. HEK-293 (ATCC, CRL-1573™) were cultured in Matrigel coated flasks and maintained in a culture medium containing DMEM/F-12 supplemented with GlutaMax, 10% FBS and 1% penicillin-streptomycin in an incubator at 37 °C with 5% CO₂. Cells were split with the help of trypsin (Try-pLE™ Express Enzyme, Fisher Scientific: 12604013) when they reached confluence and were diluted at a ratio of 1:10 to start new culture. Two hundred thousand cells were plated per dish in Matrigel coated glass bottom dishes. Cells were transfected with Lipofectamine 2000 (ThermoFisher 11668027) following the manufacturer’s protocol.

Expression of ER-GCaMP6f in primary astrocytes and in vivo

We constructed an adeno associated virus (AAV serotype 5) with truncated glial fibrillary acidic protein ^{23, 24} GfaABC1D as a promoter to incorporate the ER-GCaMP6f plasmid construct. Primary astrocytes were cultured from P2-P4 pups as described elsewhere ^{25, 26}. Transfection of primary astrocytes was performed similarly as transfection of HEK-293 cells (section 2.2). Transduction of cultured astrocytes was performed as described elsewhere ²⁷

For in vivo ER-GCaMP6f expression, following isoflurane anesthesia (2-5% isoflurane in O2) animals were placed in a stereotaxic instrument (Kopf Instruments, Tujunga, CA, USA), and AAV-ER injected at the following stereotaxic coordinates (in mm from bregma): A/P: + 1.0, M/L: ±1.0, D/V: −7.0. 2.0 μl ER-GCaMP6f (full titer 2.86x10^13) virus solution was infused at a rate of 0.2 μl/min using a syringe pump, and the injector was left in place for 5 minutes following each infusion to allow for diffusion. Animals were euthanized, and brain slices were acquired after 4 weeks to allow for GCaMP6f expression^{28, 29}.

Results and Discussion

Construction of ER-GCaMP6f plasmid and AAV

We utilized the endoplasmic reticulum targeting motif (27 amino acids) of a cytochrome p450 ^{21, 22} to anchor the Ca²⁺ indicator ER-GCaMP6f to the ER membrane and facing the cytoplasm (Fig. 1A, Supplementary Fig. 1). This motif has previously been used to anchor proteins to the ER membrane with an outward facing orientation ^{21, 22}. ER-GCaMP6f was cloned into a pcDNA 3.1 vector under a CMV promoter. We used GfaABC1D promoter to pack the plasmid in an AAV serotype 5. We transfected HEK-293 cells with the ER-GCaMP6f plasmid and observed robust expression of the Ca²⁺ indicator. To confirm the ER specific localization, we labelled ER-GCaMP6f transfected HEK-293 cells with ER-Tracker™ Red/BODIPY™ TR Glibenclamide, an ER-specific marker and performed epifluorescence and total internal reflection fluorescence (TIRF) microscopy (Fig. 1B). We saw clear colocalization between ER-GCaMP6f and ER-Tracker™ Red/BODIPY™ TR Glibenclamide fluorescence. In order to visualize structural features within the ER, we also performed structural illumination microscopy (SIM), a super resolution technique capable of ~100 nm resolution (Fig. 1B-1C). We observed a consistent reticulated pattern and overlap between the two channels. In similar experiments, we did not observe colocalization of the ER-GCaMP6f with a mitochondrial marker (Supplementary Fig. 2), further indicating that the expression of ER-GCaMP6f is localized to the endoplasmic reticulum.

FIGURE 1:

Design and expression of ER-GCaMP6f genetic construct AAV5. A) Schematic of ER-GCaMP6f used to measure endoplasmic reticulum specific calcium signaling in astrocytic processes. B) Wide field image of an HEK-293 cell labelled with ER-GCaMP6f, Super resolution (Structured Illumination Microscopy) image of a different HEK-293 cell labelled with ER-GCaMP6f and ER-marker (ER-Tracker™ Red/BODIPY™ TR Glibenclamide) In B). Scale bars = 10 μm. C) 3D SIM image of HEK-293 cell labelled with ER-GCaMP6f, ER-marker and merged image of ER-GCaMP6f and ER-marker (ER-Tracker™ Red/BODIPY™ TR Glibenclamide)

Expression of endoplasmic reticulum (ER) in astrocytic soma and processes.

In order to express ER-GCaMP6f in astrocytes, we transduced astrocytes with ER-GCaMP6f containing AAV5 with GfABC1D promoter. To verify ER specificity in astrocytes, we performed a series of structural illumination microscopy experiments in primary astrocytes. We observed the presence of ER in both cell soma and astrocytic processes (Fig. 2A-B, Supplementary Fig.3). There was clear overlap between ER-GCaMP6f fluorescence and the ER specific marker (ER-Tracker TM Red). Validation of the presence of ER in the processes is a recent development³⁰ where expression was shown in processes and endfeets of astrocytes.

FIGURE 2:

ER-GCaMP6f measures ER specific calcium signaling in astrocytes. A) 3D SIM image of an astrocytic cell soma labelled with ER-GCaMP6f, ER-marker (ER-Tracker™ Red/BODIPY™ TR Glibenclamide) and merged image (nucleus is labelled with DAPI). Scale bars = 10 μm. B) 3D SIM image of an astrocytic process labelled with ER-GCaMP6f, ER-marker (ER-Tracker™ Red/BODIPY™ TR Glibenclamide) and merged image. Scale bars = 10 μm. C) Sample spontaneous calcium activity in astrocytic cell soma and processes expressing ER-GCaMP6f at different time intervals from a widefield microscope. Scalebar = 40 μm. D) Calcium activity in cell soma E) Calcium activity in processes. F) Wide-field image of calcium activity in astrocytic processes. G) TIRF microscopy image of calcium activity in astrocytic processes.

ER-GCaMP6f measures calcium activity in both astrocytic cell soma and processes

To verify that ER-GCaMP6f was able to measure spontaneous Ca²⁺ activity in the cell soma and the processes, we performed fluorescence time-lapse imaging. We observed spontaneous activity in both the cell soma and processes in astrocytes (Fig. 2C, Supplementary Fig. 4). Fig. 2D shows activity in a representative astrocyte soma where each trace is a different ROI within the cell where separate ROIs demark individual microdomains of activity within the cell. Similarly, Fig. 2E shows activity from the same cell in different ROIs in the processes. In the cell soma, the frequency of activity was similar across all the ROIs. Activity in the processes was more heterogeneous with different ROIs showing different frequencies in spontaneous activity. In wide field fluorescence (Fig. 2F), the signal is collected from the entire cell and is dominated by the bulk of ER from the interior of the cell. TIRF microscopy has previously been used to visualize the plasma membrane and the peripheral ER by limiting fluorescence to a region within 150 nm of the cell surface^{31, 32}. TIRF imaging (Fig. 2G) revealed activity of processes that were not visible using wide field microscopy. Comparing spontaneous activity in wide field and TIRF, we observed clear bursts of activity in TIRF but not in epifluorescence. This indicates that even small processes exhibit Ca²⁺ activity and that these small fluctuations are visible using a combination of TIRF microscopy and ER-GCaMP6f expression.

Comparison of ER- GCaMP6f against commonly used astrocyte calcium indicators

We tested the ER-GCaMP6f AAV against commonly used astrocyte specific AAV indicators in cell culture. We chose cytosolic GCaMP6f which is a soluble GCaMP6f expressed throughout the cell ³³ and Lck GCaMP6f (Fig. 3A-B) which is a membrane tethered astrocyte specific GCaMP6f. We performed time lapse fluorescence imaging to compare the activity of the different indicators. In our 2-minute time scale study with a 0.5 second time interval, we found that the number of spontaneous events per 2 minutes were not significantly different among these 3 viruses (p=0.362, n=49) in the cell soma. The cytosolic indicator had 2.2 ±0.4 events, Lck-GCaMP6f had 1.8 ±0.2 events and ER-GCaMP6f had 1.5 ±0.2 events per 2-minute interval (Fig. 3C). However, we observed clear differences in the processes where the cytosolic indicator had 1.3 ±0.2 events, Lck- GCaMP6f had 2.1 ±0.2 events and ER had 2.0 ±0.2 events. In the processes, Lck-GCaMP6f and ER-GCaMP6f detected a significantly higher (p=0.014 for all comparison, n=102) number of events compared to cytosolic GCaMP6f (Fig. 3D). This suggests that cytosolic GCaMP6f adequately measures spontaneous cell soma activity but likely misses activity occurring in the processes. This could be due to the concentration of soluble GCaMP6f entering the processes whereas ER-GCaMP6f and Lck-GCaMP6f are anchored to membranes in these regions.

FIGURE 3:

Comparison of ER-GCaMP6f with Lck and cytosolic GCaMP6: A) Merged wide field image of a time series of calcium activity from astrocytes expressing Lck-GCaMP6. B) Merged wide field image of a time series of calcium activity from astrocytes expressing ER-GCaMP6f. C-H) Comparison of the number of events per 2-minute interval, amplitude of the events, and half width for ER, Lck and cytosolic GCaMP6f.

To further analyze the calcium signals from these indicators, we measured the amplitude and half width of calcium signals (Fig. 3E-F). The amplitudes of calcium events were also not significantly different (p=0.559, n=89) among the three indicators for spontaneous events in the cell soma (Fig. 3E). The cytosolic indicator had an average amplitude of 0.7 ±0.2 in the cell soma, Lck had an amplitude of 0.9 ±0.1, and ER- GCaMP6f had an amplitude of 0.8 ±0.1 in the cell soma. In the processes, however, the amplitudes were significantly different (p<0.001, n=183) (Fig. 3F). Though the cytosolic indicator group had fewer number of events in the processes in comparison to Lck and ER, it had relatively higher amplitude of calcium waves in the processes (p<0.001, n=183). The average amplitude for cytosolic GCaMP6f was 1.1 ±0.2 which was 0.9 ±0.1 for Lck and 0.4 ±0.1 for ER. Thus, the ER- GCaMP6f had significantly lower amplitude in comparison to both the cytosolic indicator (p<0.001) and Lck(p<0.001) (Fig. 3E-F).

As we performed experiments using a 0.5 second time interval to minimize the laser induced effects on amplitude and number of events, we were able to capture the half width of various events (Fig. 3G-H). They were significantly different in the cell soma (p<0.001, n=88) with the cytosolic indicator having a half width of 3.8 ±0.3 seconds, Lck had a half width of 4.8 ±0.4 seconds, and ER had a half width of 8.0 ±1.5 seconds. Cytosolic and Lck indicators were not significantly (p=0.079) different in half width from each other, but they were both significantly shorter than the half width of ER-GCaMP6f (p<0.001). Similar results were observed in the processes, where the average duration of spontaneous signals was longer for the ER indicator as compared to the other two. Cytosolic-GCaMP6f had a half width of 3.7 ±0.4 seconds, Lck- GCaMP6f had a half width of 3.4 ±0.2 seconds and ER- GCaMP6f had a half width of 5.9±0.4 seconds. (Fig. 3G-H). Though the average half width is longer for ER-GCaMP6f calcium measurements as compared to the other indicators, measurements with ER-GCaMP6f were able to detect both very short and relatively longer calcium events in the ER (Fig. 3G-H). This indicates that the ER indicator is detecting Ca2+ activity that is missed by the other indicators and is likely measuring calcium activity near the ER.

These results suggest ER-GCaMP6f can measure spontaneous calcium signaling in both the cell soma and processes of astrocytes. Together with the clear expression of the indicator in the ER, the differences in the number of events, amplitude, and duration supports that ER-GCaMP6f is measuring near ER calcium signaling. It is likely that that Lck and cytosolic GCaMP6f are in part measuring ER resident events but also have strong non-ER component. Using ER-GCaMP6f, the processes were easier to distinguish, and we observed more frequent Ca2+ events in the processes as compared to cytosolic GCaMP6f.

ER-GCaMP6f measures inositol 1,4,5-trisphosphate receptors (IP3Rs) mediated calcium signaling in both the cell soma and processes

While the plasma membrane tethered Lck-GCaMP6f and cytosolic registered events in both the cell soma and processes, there were clear differences from ER-GCaMP6f signals. To determine if these differences were due to ER-GCaMP6f measuring near ER events, we performed a set of experiment using ER resident receptors. Inositol 1,4,5-trisphosphate receptors (IP3Rs) and Ryanodine receptors (RyRs) are the two major receptors involved in endoplasmic reticulum mediated calcium signaling³⁴. To study the role of IP3Rs in ER mediated calcium signaling, we performed time lapse experiments in the presence of an IP3Rs activator and an IP3Rs blocker. ATP is an activator of IP3Rs and have been extensively used in calcium imaging^35-37. We selected regions of interests (ROIs) actively showing spontaneous calcium fluctuations and conducted time lapse imaging for 5 minutes with an interval of 2 seconds. We added ATP at the 5-minute mark to determine how activation of ER resident receptors affected the number and amplitude of the calcium events (Fig 4-5). 20 μM ATP significantly increased (P<0.001, n=30) the number of events in the cell soma of astrocytes containing ER-GCaMP6f from 4.5±0.3 to 11.3±0.9 per 5 minute (Fig. 4C). Similarly, ATP significantly (p<0.001, n=30) increased the number of events from 4.1±0.6 to 10.9±0.7 per 5 minute in the processes (Fig. 4D). ATP also increased the amplitude of events in both the cell soma (p=0.002, n=236) and processes (p=0.042, n=221) (Fig. 4E-F). As anticipated due to the increase in Ca²⁺ release, we observed similar increases in the number and amplitude of events using Lck- GCaMP6f in response to ATP.

FIGURE 4:

Measurement of IP3Rs mediated calcium signaling by ER-GCaMP6f in astrocytes. A-B) Average intensity plot of calcium activity from ER-GCaMP6f expressing astrocytes treated with 20 μM ATP and 50 μM 2-APB. Scalebar = 40 μm. C-J) Comparison of number of events and amplitudes of control and ATP/2-APB treated astrocytes across cell soma and processes

FIGURE 5:

Measurement of RyR mediated calcium signaling by ER-GCaMP6f. A-B) Average intensity plot of calcium activity from ER-GCaMP6f expressing astrocytes treated with 100 μM and 50 mM of caffeine. Scalebar = 40 μm. C-J) Comparison of no of events and amplitudes of control and caffeine treated astrocytes across cell soma and processes

To study the effect of IP3R deactivation in ER mediated calcium signaling we used a common IP3R blocker, 2-APB ³⁸. 2-APB significantly (p=0.024, n=20) decreased the number of events from 4.2±1.0 to 1.7±0.5 per 5 minute in the cell soma (Fig. 4G). It also significantly decreased (p=0.047, n=28) the number of events from 3.6±0.4 to 2.7±0.5 per 5 minute in the processes (Fig. 4H). We observed that 2-APB both decreased the number of detectable low amplitude events and simultaneously decreased the amplitude of high amplitude events resulting from the measurement of IP3R independent events. In the soma the amplitude values were not significantly different (p=0.693, n=48) where non-treated and treated astrocytes had an amplitude of 0.5±0.1. In the processes, the amplitudes were 0.3±0.1 for both untreated and 2-APB treated astrocytes showing no significant difference (p=0.972, n=98) (Fig. 4I-J). We also tested the effect of IP3Rs blockers and activators in Lck-GCaMP6f and observed effects in similar direction on number of events and amplitude (Supplementary Fig. 6A-D, G-H). These results strongly suggest ER-GCaMP6f successfully measures IP3R mediated calcium signaling in astrocytes. Some calcium activity still persisted after deactivating IP3R suggesting ER-GCaMP6f also measures non-IP3R calcium activity.

ER-GCaMP6f measures ryanodine receptor (RyRs) mediated calcium signaling in both the cell soma and processes

Ryanodine receptors (RyRs) are another major class of receptors involved in ER calcium signaling. Caffeine is an activator of RyRs but simultaneously inhibits IP3Rs. These effects have previously been used as a tool to study RyRs mediated calcium signaling^{14, 39}. We treated astrocytes with both low (100 μM) and high concentrations (50 mM) of caffeine to investigate whether ER-GCaMP6f measures ryanodine receptor mediated calcium signaling (Fig. 5). The number of events were significantly different in the cell soma (p=0.024, n=22) between untreated control astrocytes which had 3.8±1.3 events per 5 minute and 100 μM caffeine treated astrocytes which had 6.4±2.1 events per 5 minute (Fig. 5C). The number of events were significantly different (p=0.039, n=22) in processes as well where control astrocytes exhibited 4.4±1.1 events per 5 minute while astrocytes treated with 100 μM caffeine treated part had 6.4±1.6 events per 5 minute (Fig. 5D). The amplitudes, however, were not significantly different between control and astrocytes treated with 100 μM caffeine in either the cell soma (p=0.581, n=109) or the processes (p=0.317, n=123) (Fig. 5E-F). With 50 mM caffeine treatment, the pattern was completely reversed for ER-GCaMP6f measurements. Treatment with 50 mM caffeine significantly (p<0.001, n=56) decreased the number of events from 4.6±0.7 to 1.3±0.3 in the cell soma. Similarly, the number of events in the processes were also decreased by 50 mM of caffeine (p=0.001, n=22) where untreated astrocytes had 4.1±0.9 events per 5 minute and treated astrocytes had 1.6±0.4 events per 5 minute.

We also performed similar studies using Lck-GCaMP6f. The addition of 100 μM caffeine resulted in a distinct effect on Lck-GCaMP6f as it decreased the number of events in both the cell soma (p<0.001, n=20) from 5.1±0.8 to 1.9±0.7 and in the processes from 5.1±0.7 to 2.6±0.5 (p=0.002, n =34) (Supplementary Fig. 6E-F). The amplitude with caffeine was not significantly different in the cell soma (p=0.090, n=59) between control and 100 μM caffeine treated astrocytes. However, 100 μM caffeine significantly reduced the amplitude in processes (p=0.02, n=128) where non-caffeine events had an amplitude of 0.6±0.1 and 100 μM caffeine treated events had an amplitude of 0.3±0.1 (Supplementary Fig. 7D-E). We observed similar effects when astrocytes were treated with 50 mM caffeine (Supplementary Fig. 6G)

The effect of caffeine on ER-GCaMP6f expressing astrocytes was completely reversed when increasing the concentration from 100 μM to 50 mM. At 100 μM, caffeine significantly increased the number of events in both the cell soma and processes. As low concentrations of caffeine are shown to activate ryanodine receptors ⁴⁰, we believe it does not sufficiently block IP3Rs so overall it increases the calcium activity which strongly suggests ER-GCaMP6f successfully measures RyR mediated calcium signaling near the ER of astrocytes. The high concentration of caffeine still activates RyRs ⁴⁰ but has been shown to block IP3Rs more effectively ⁴¹. In our study 50 mM caffeine decreased the number of events in both the cell soma and processes measured by ER-GCaMP6f. Combining these two results suggests ER-GCaMP6f measures both IP3R and RyR mediated calcium signaling in astrocytes. When we added 100 μM caffeine, which should be sufficient to activate RyRs and partially block IP3R, to Lck-GCaMP6f expressing astrocytes, the number of events decreased. This decrease in Lck-GCaMP6f detected events is likely due to a suppression of IP3R signaling and a lack of detection of some RyR events in the ER (Supplementary Fig. 7A-E). This contrasts with the observation of an increase in activity when using ER-GCaMP6f with 100 μM caffeine. This suggests that the ER anchored indicator is able to detect events from a larger population of RyRs than Lck-GCaMP6f.

ER-GCaMP6f can be expressed in vivo to study calcium signaling in brain slices

In order to validate the delivery and expression of ER-GCaMP6f in vivo, we injected the nucleus accumbens (NAc) of rats with an ER-GCaMP6f AAV5 that specifically targets astrocytes and performed calcium imaging using an epifluorescence microscope (Fig. 6A). Time lapse imaging was performed on brain slices from these animals to examine the ability of ER-GCamp6f to detect spontaneous calcium transients. For these animals, one side of the brain was injected with Lck-GCaMP6f and the other side with ER-GCaMP6f. We observed fewer ER-GCaMP6f labelled active ROIs in the Nucleus Accumbens shell with ER-GCaMP6f as compared to the Lck-GCaMP6f labelled brain slices (Fig. 6B, Supplementary Fig. 5). To determine the capability of the ER-GCaMP6f to measure changes in activity, we compared animals subjected to short- and long-term cocaine exposure (Fig. 6, Supplementary Fig. 9). ER-GCaMP6f registered different spontaneous ER activity between the two groups. Animals with extended access exhibited lower duration and amplitude in ER Ca²⁺ activity compared to the short-term access animals (Supplementary Fig. 9). Short access cocaine self-administration has been recently reported to decrease amplitude and duration of astrocytic Ca²⁺ transients recorded either at 1 or 14 days of withdrawal from self-administration in the nucleus accumbens shell²⁸. The data in that publication was acquired using a cytosolic variant of GCaMP6f. Our current data, using plasma membrane and ER variants of GCaMP6f and collected in the same brain area, indicate that amplitude and duration of astrocytic Ca²⁺ signals are also sensitive to length of cocaine self-administration sessions (Fig. 6, Supplementary Fig. 9). Moreover, we find that while Ca²⁺ amplitudes were suppressed in extended access animals when measured with either plasma membrane or ER indicators, Ca²⁺ event durations were suppressed in extended access animals when measured with ER, but not plasma membrane, indicator. With respect to amplitude, these results may indicate that decreased Ca²⁺ at the ER is mirrored as smaller Ca²⁺ elevations at the plasma membrane.

FIGURE 6:

Expression of ER-GCaMP6f in vivo to measure ER-targeted calcium signaling in astrocytes at the nucleus accumbens (NAc). A) An illustration of ER-GCaMP6f expression and imaging in vivo. B-E) Comparison of spontaneous calcium activity in rat brain slices expressing ER-GCaMP6f with Lck-GCaMP6f

This demonstrates that the ER indicator detects physiological differences between conditions and has an additional impact for studies of cocaine-mediated plasticity of astrocyte and neuronal Ca²⁺ signals²⁸. Additionally, the dynamics measured in both groups were different for the ER indicator versus the plasma membrane indicator, where calcium activity in the ER was less frequent and shorter in duration than those at the plasma membrane suggesting that near ER activity is distinct from those occurring at the plasma membrane. (Fig. 6 C-D, Supplementary Fig. 9).

Pharmacological treatment was also applied to further study calcium channel responses. The IP3R signaling blocker (2-APB) diminished the strength of calcium activity in both the ER and the plasma membrane but with different patterns. The frequency of calcium activity through the cell membrane is partially blocked by 2-APB, while ER calcium activity is not This confirms the 2 localized indicators reveal distinct calcium activity. While duration of calcium activity is partially blocked by 2-APB at both the plasma membrane and the ER, a larger effect was seen at the plasma membrane. The amplitude of calcium transients was also partially blocked by 2-APB at both the plasma membrane and the ER (Fig. 6 C-D).

Conclusion

We developed an endoplasmic reticulum targeted GCaMP6f indicator to study calcium signaling in astrocytes. Structural illumination microscopy showed the reticulate pattern indicative of ER strongly suggesting its localization to ER. We showed that ER-GCaMP6f is expressed in both the cell soma and processes of astrocytes and can also be applied to study low amplitude and short duration calcium fluctuations in astrocytic processes by utilizing TIRF microscopy. We validated near ER activity by selectivity measuring IP3R and RyR mediated calcium signaling in astrocytes evidenced by pharmacological treatments with different reagents in cultured astrocytes. ER-GCaMP6f was also expressed in a mouse model to using an AAV5 delivery vector to validate in vivo activity. In vivo expression and calcium imaging of brain slices revealed that ER-GCaMP6f measures IP3R mediated calcium signaling occurs at different locations and with different dynamics as compared to Lck-GCaMP6f. This ER anchored indicator reports distinct calcium transients compared to lck-GCaMP6f, enabling the measurement of near ER intracellular calcium activity.

Key points.

Endoplasmic reticulum is expressed in both astrocytic cell soma and processes.

ER-GCaMP6f reports calcium signaling in close proximity to the endoplasmic reticulum in astrocytic soma and processes.

ER-GCaMP6f measures Inositol triphosphate receptors (IP3Rs) mediated calcium signaling in the endoplasmic reticulum of astrocytes.

ER-GCaMP6f measures Ryanodine receptors (RyRs) mediated calcium signaling in endoplasmic reticulum of astrocytes.

ER-GCaMP6f can be expressed in astrocytes in vivo to measure calcium activity in brain slices.

ABBREVIATIONS

ER

Endoplasmic reticulum

SERCA

Sarco/Endoplasmic reticulum Ca2+-ATPase

IP3R

Inositol 1,4,5-trisphosphate receptors

RyR

Ryanodine receptors

GECI

Genetically encoded calcium indicators

AAV

Adeno Associated Virus

SIM

Structural illumination microscopy

ROI

Region of interest

TIRF

Total internal reflection fluorescence

ATP

Adenosine triphosphate

2-APB

2-Aminoethoxydiphenyl borate