Compartment-specific analysis reveals disrupted astrocytic calcium transients in anesthetized Alzheimer's disease mice

Experimental animals

This study utilized a transgenic mouse line expressing the Swedish mutation of the amyloid precursor protein and the deltaE9 mutation in presenilin 1 (APPswe/PS1dE9; APP/PS1 mice, stock #034829, The Jackson Laboratory) as well as age-matched nontransgenic littermate controls (NTG mice). ⁴² Animals were 4–6 months old and included both males and females. All procedures involving mice were performed with the approval of the Massachusetts General Hospital (MGH) Institutional Animal Care and Use Committees (IACUC, protocol # 2012N000085) and followed the relevant guidelines and regulations, including the ARRIVE guidelines. The mice were housed in a pathogen-free environment with four animals per cage and had access to food and water ad libitum. The facility lighting was maintained on an automated 12-h light/12-h dark cycle.

Stereotaxic injection of adeno-associated viral vectors (AAVs) targeting Yellow Cameleon 36

4–6-month-old APP/PS1 and NTG mice were used for astrocytic calcium activity measurements. The animals were anesthetized with isoflurane inhalation (5% induction, 1.75% maintenance), and each were placed into a stereotaxic instrument. The body temperature was maintained with a heating pad during injections. To prevent dryness, eye ointment was carefully applied to the eyes of the mice, and their fur was removed from the scalp using a trimmer. Liquid iodine was then applied to the exposed skin and wiped clean with 70% isopropyl alcohol swabs three times. An incision was made with sterile surgical scissors to expose the skull, and a small burr hole was drilled into the injection site with the following coordinates: anterior-posterior (AP) −1.5 mm, lateral-medial (ML) −1.5 mm, and dorsal-ventral (DV) −0.8 mm. The mice received intracortical viral injections of 3 µL of AAV2/5-gfa2-YC3.6, which specifically targeted Yellow Cameleon 3.6 (YC 3.6, titer=4 × 10¹² vg/ml) to astrocytes in the right somatosensory cortex. While 3 µL is relatively large for an intracortical injection, this volume was necessary to achieve sufficient astrocytic expression for in vivo imaging. The injections were performed using a 33-gauge metal needle and a 10 µl glass syringe controlled by a stereotaxic injector at a rate of 100 nl/min to minimize tissue damage. Following each intracortical injection, the scalp was sutured, and the mice were allowed to recover from anesthesia on a heating pad until they were awake and freely moving. The viral vector was allowed to express for at least four weeks prior to imaging, during which the mice were observed to behave with no observable behavioral abnormalities.

Cranial window installation and multiphoton imaging

After receiving an AAV2/5-gfa2-YC3.6 injection, APP/PS1 and NTG mice underwent cranial window implantation procedures over the injection sites in the right somatosensory cortices. These procedures allowed the monitoring of astrocytic calcium transients using multiphoton microscopy. The animals were anesthetized with isoflurane inhalation (5% induction, 1.75% maintenance), and a round ∼5 mm craniotomy was performed using a dental drill. A 5 mm glass coverslip was then placed over the cortex, and the coverslip was secured to the skull with a mixture of Krazy glue and dental cement. A mode-locked titanium/sapphire laser (Mai Tai; Spectra-Physics, Fremont, CA) was used to generate two-photon fluorescence with 860 nm excitation, and three photomultiplier tubes (Hamamatsu Photonics, Japan) were used to detect fluorescence emissions in the range of 380–480 nm, 500–540 nm, and 560–650 nm. A 2-photon microscope (Olympus) was used to image YC 3.6-expressing astrocytes with a 25X water immersion objective (NA = 1.05, Olympus). Time-lapse images were acquired immediately following craniotomy from the somatosensory cortex (layer II/III) at a resolution of 256 × 256 pixels and a sampling rate of 2 µs per pixel (2 frames per second). Fields of view (FOVs) were selected using a standardized protocol across all animals and groups, based on clear astrocytic labeling and signal quality, without prior knowledge of activity patterns. Regions with artifacts were excluded. The same number of FOVs per animal and identical imaging parameters were maintained, with spatial separation to avoid repeated sampling.

To minimize phototoxicity, the laser power was kept below 50 mW as measured at the back focal plane of the objective. 150 µl of Texas red dextran (70 kDa; 12.5 mg/ml in PBS; Invitrogen) was administered by retroorbital injections to label the vasculature and imaged using multiphoton microscopy through cranial windows. Mice were euthanized using CO₂ inhalation after acquiring multiphoton images.

Image analysis

ImageJ/FIJI (NIH) was used to process timelapse images of compartment-specific YC 3.6-expression within cortical astrocytes. First, each timelapse (250 s long) was loaded into ImageJ and split into the YFP and CFP channels. Regions of interest (ROIs) containing either somas, processes, microdomains, or endfeet were selected manually. Astrocytic processes were identified as thin extensions emanating from the soma. Endfeet were specifically identified by their close apposition to blood vessels, which were visualized via intravascular labeling with Texas Red dextran, enabling unambiguous delineation of perivascular astrocytic structures. ROIs were manually segmented on time-averaged images to ensure structural clarity and consistency, with care taken to avoid overlap between compartments. The same criteria were applied uniformly across all animals and groups. The raw mean grey values for each ROI of the YFP and CFP channels were measured. The YC 3.6 ratios were calculated by dividing the emitted fluorescence intensity of YFP by CFP. A custom MATLAB script was used to generate calcium traces (plotted as ΔR/R_o over time) for each ROI. To determine the ΔR/R_o values, baseline ratio (R_o) was calculated using the asymmetric least-squares mean smoothing method,^43,44 which allowed us to find a variable baseline R_o without requiring prior information about the peak areas of the signal. The ratio change over the baseline was calculated using the following equation: ΔR/R_o = (R − R_o)/R_o, where R is instantaneous ratio. Ca²⁺ transients were identified as changes in ΔR/R⁺ exceeding four times the standard deviation (σ) of noise band. Noise band was estimated from ΔR/R_o values within the 20th to 50th percentile range, ⁴⁵ depending on compartment activity, to minimize contamination from transients. The resulting detection threshold was conservatively set at 4σ, typically corresponding to ∼0.2 ΔR/R_o. The active cell compartment was defined as having activity above this threshold. The AUC was calculated as the sum of the areas under all calcium transient events within each ROI over the recording period. Events were defined as ΔR/R_o signals exceeding a threshold. For each event, the AUC was computed by integrating the signal above this threshold from event onset to return below threshold, and total AUC per ROI was obtained by summing across all detected events. A similar approach was used to calculate activity duration, defined as the total time the signal remained above threshold. Peak amplitude values were calculated using MATLAB's findpeaks function, which returned the local maxima of the trace, and the number of detected peaks was defined as the event count. Calcium activity was analyzed continuously over the full 250-s recording as a timelapse and was not segmented into discrete intervals. Event counts were normalized to recording duration and expressed as event rate (events per 60 s). Similarly, AUC and cumulative activity durations were normalized to recording length by dividing by 250 s and scaling to a 60-s period (i.e., value ÷ 250 × 60). All quantitative analyses and statistical comparisons were performed on these per-ROI, time-normalized values, preserving the full temporal resolution of the recording and ensuring that reported timelapse statistics reflect overall activity throughout the recording. Pairwise correlations were calculated for each combination of somas within a timelapse using Pearson's Linear Correlation Coefficient and were compared between the two groups. To determine the distances between the somas, the centroids of the somas were exported from ImageJ to MATLAB script to calculate the Euclidean distance. Pairwise correlation values between somas were binned over a distance of 10 μm, and a relationship between pairwise correlation and distance was determined.

Statistical analyses

All statistical analyses were performed using MATLAB (version R2023a). The Shapiro-Wilk normality test was performed to assess the data distribution. Data sets comparing NTG and APP/PS1 were analyzed using a two-sample Student's t-test. Pearson's Linear Correlation Coefficient was used to measure the correlation between distance and pairwise correlations. The Kolmogorov-Smirnov (K-S) test was used to compare pairwise correlations between the groups. Statistical significance was set at p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001.

Calcium transients in the astrocytic soma of APP/PS1 mice exhibited a longer activity duration and increased amplitude compared to those of NTG controls

We investigated whether cortical astrocytes exhibited altered spontaneous calcium activity in APP/PS1 mice compared to astrocytes in NTG littermate controls. Astrocytes were targeted with the ratiometric calcium indicator Yellow Cameleon 3.6 (YC 3.6) via the GFAP promoter in the somatosensory cortex. We assessed the expression of YC 3.6 and confirmed its presence in the somas, processes, microdomains, and endfeet of NTG and APP/PS1 mice (Figure 1A-C). Astrocytic calcium transients were visualized using multiphoton microscopy in anesthetized mice. Here we included both low frequency, high amplitude ²⁰ and high frequency (limited to 1 Hz), low amplitude astrocytic calcium transients ²¹ in our analyses. The high frequency, low amplitude calcium transients within astrocytes are often dismissed as noise in the literature, however these calcium events have physiological relevance to sleep-dependent brain rhythms.^21,46,47 Thus, high frequency, low amplitude calcium transients were included here.

OPEN IN VIEWER

Figure 1.

Yellow Cameleon 3.6 (YC 3.6) was expressed within different astrocytic compartments in nontransgenic (NTG) and APP/PS1 mice. (A) YC 3.6 expression (green) in astrocytic soma, processes, microdomains and endfeet in NTG mice. (B) YC 3.6 expression (green) in astrocytic soma, processes, microdomains and endfeet in APP/PS1 mice. Dextran red shows the blood vessels. Scale bar, 50 μm. (C) The higher magnification image of a single astrocyte shows the area bounded by dashed white box in (B). ROI 1 corresponds to the astrocyte soma; ROIs 2–5 capture processes; ROIs 6–8 represent microdomains; and ROIs 9–10 correspond to endfeet. Scale bar, 20 μm. (D) Representative calcium traces that include both high- and low-amplitude events analyzed in this study.

To develop a comprehensive understanding of calcium activity at subcellular resolution, we selected the ROIs pertaining to distinct astrocytic compartments, including somas, processes, microdomains, as well as endfeet (Figure 1C). Astrocytic calcium transients were analyzed separately in each compartment. Representative calcium traces (Figure 1D) show both high- and low-amplitude events, reflecting the detection of high-amplitude, low-frequency and low-amplitude, high-frequency astrocytic calcium activity. Supplemental Figure 1 illustrates the conversion of YFP/CFP ratios to ΔR/R_o and provides an example of the calculated detection threshold for a representative trace.

To monitor calcium transients in astrocytic somas, we manually selected the somas as ROIs (Figure 2A). A total of 480 somas were identified across 64 timelapses in seven APP/PS1 mice. 448 somas were identified across 64 timelapses in seven NTG mice. We generated astrocytic calcium activity traces, calcium transients, measured as ΔR/R_o over time for each ROI, where ΔR is the difference between the R ratio of YC 3.6 and R_o, while R_o is the baseline ratio (Figure 2B, C). ΔR/R_o traces for active (Figure 2B) and inactive somas (Figure 2C) were thresholded, and calcium transient analysis was performed (see Methods for more details). We determined percentages of active and inactive somas. We found that the percentages of active somas were comparable in NTG and APP/PS1 mice (64.38% active somas in NTG mice and 67.19% active somas in APP/PS1 mice, Chi-square (χ²) test, χ²= 0.83, p = 0.37) (Figure 2D). We also analyzed calcium events by calculating event rate, cumulative activity duration, cumulative area under the curve, and peak amplitude. The cumulative activity duration of astrocytic somas was significantly higher in APP/PS1 mice than the cumulative activity duration in NTG mice (42.37 [41.54, 43.24] seconds and 42.24 [41.41, 43.1] seconds, respectively, median [Q1 (1st Quartile), Q3 (3rd Quartile)], two-sample student's t-test p = 0.013) (Figure 2F). Similarly, the peak amplitude values of astrocytic somas in APP/PS1 mice were significantly higher than those in NTG mice (0.35 [0.3, 0.46] and 0.29 [0.23, 0.42] respectively, two-sample Student's t-test p = 0.035) (Figure 2H). In contrast, there was no significant difference in the event rate of astrocytic somas (1.44 [0, 9.48] transients/min and 1.68 [0, 9.6] transients/min, respectively, two-sample Student's t-test, p = 0.93) between APP/PS1 and NTG mice (Figure 2E). Furthermore, there was no significant difference in the cumulative area under the curve of astrocytic somas (0.03 [0, 0.3] s and 0.03 [0, 0.26] s, respectively, two-sample Student's t-test, p = 0.06) between APP/PS1 and NTG mice (Figure 2G). Based on these results, we concluded that astrocytic soma in APP/PS1 mice displayed calcium signals with extended duration and heightened magnitude when compared to those in NTG counterparts (Figure 2B).

OPEN IN VIEWER

Figure 2.

Astrocytic calcium activity in the soma of APP/PS1 mice showed elevated activity duration and higher amplitude of calcium events compared to NTG controls. (A) Astrocyte schematic outlining the region of interest (ROI: soma) where activity was measured. Representative active (B) and inactive (C) traces of astrocytic Ca²⁺ activity in the soma of the NTG (blue) and APP/PS1 (black) mice. Red line represents threshold. Traces above the threshold are considered active, while those below are considered inactive. (D) Percentage of active somas in NTG control and APP/PS1 mice. The chi-square test was used to measure statistical significance. (E) Box plot of the event rate of astrocytic Ca²⁺ activity in the soma. The event rates were calculated using a ΔR/R threshold of 0.2 (see materials and methods). (F) Box plot of the cumulative duration of astrocytic Ca²⁺ activity in the somas. (G) Box plot of the area under the curve of astrocytic Ca²⁺ activity in somas. (H) Box plot of ΔR/R peak values from astrocytic Ca²⁺ activity in somas. Statistical significance for panels E–H was assessed using a two-sample Student's t-test, *p < 0.05, ***p < 0.001, n.s, not significant.

Calcium transients in the astrocytic processes of APP/PS1 mice exhibited a higher frequency, shorter activity duration, decreased intensity (AUC), and reduced amplitude compared to those of NTG controls

We analyzed spontaneous calcium transients in the astrocytic processes (Figure 3). We confirmed the YC 3.6 expression in the astrocytic processes (Figure 1) and manually selected the processes as ROIs (Figure 3A). We identified 423 processes across 64 timelapses in seven NTG mice, and 257 processes across 64 timeplases in seven APP/PS1 mice. We analyzed the calcium traces (ΔR/R_o) for active and inactive process ROIs (Figure 3B, C). The percentages of active processes were high and comparable in NTG and APP/PS1 mice (99.05% active in NTG mice and 97.67% processes active in APP/PS1 mice, Chi-square test, χ² = 2.13, p = 0.14) (Figure 3D). The cumulative activity duration in astrocytic processes was significantly lower in APP/PS1 mice than that in NTG mice (7.5 [3.41, 13.57] seconds and 12.65 [5.0, 19.71] seconds, respectively, two-sample student's t-test p = 3.27 × 10⁻⁹) (Figure 3F). Similarly, the cumulative area under the curve was significantly lower in APP/PS1 mice than in NTG controls (1.04 [0.37, 2.52] s and 0.25 [0.32, 0.42] s, respectively, two-sample Student's t-test p = 1.87 × 10⁻⁸) (Figure 3G). The peak amplitude values of astrocytic processes in APP/PS1 mice were significantly lower than those in NTG mice (0.29 [0.24, 0.37] and 0.25 [0.32, 0.42], respectively, two-sample student's t-test p = 8.09 × 10⁻⁵) (Figure 3H). However, the event rate of astrocytic processes was significantly higher in APP/PS1 mice than that in NTG mice (27.36 [19.74, 29.16] transients/min and 19.68 [13.26, 24.63] transients/min, respectively, two-sample Student's t-test p = 8.09 × 10⁻⁵) (Figure 3E). These findings suggested that calcium transients in astrocytic processes of APP/PS1 mice exhibited increased frequency, shorter duration, lower intensity (AUC), and smaller amplitude compared to astrocytic calcium transients in NTG controls (Figure 3B).

OPEN IN VIEWER

Figure 3.

Calcium transients in astrocytic processes of APP/PS1 mice display a higher frequency, shorter duration, lower intensity (AUC), and reduced amplitude compared to those in NTG controls. (A) Astrocyte schematic outlining the regions of interest (processes) where the activity was measured. Representative active (B) and inactive (C) traces of astrocytic Ca²⁺ activity in the processes from NTG (blue) and APP/PS1 (black) mice. Red line represents threshold. (D) The percentage of active processes in NTG controls and APP/PS1 mice. The chi-square test was used to measure statistical significance. (E) Box plot of the event rate based on astrocytic Ca²⁺ activity. The event rates were calculated using a ΔR/R threshold of 0.2. (F) Box plot of the cumulative activity duration from astrocytic Ca²⁺ activity in the processes. (G) Box plot of the area under the curve from astrocytic Ca²⁺ activity in the processes. (H) Box plot of ΔR/R peak values from astrocytic Ca²⁺ activity in the processes. Statistical significance for panels E–H was assessed using a two-sample Student's t-test, ***p < 0.001, n.s, not significant.

Calcium transients in the astrocytic microdomains of APP/PS1 mice exhibited higher frequency, shorter activity duration, decreased intensity, and reduced amplitude compared to those of NTG controls

We analyzed spontaneous calcium transients in the astrocytic microdomains (Figure 4). First, we confirmed that YC 3.6 expression was present in the astrocytic microdomains (Figure 1), and manually selected the microdomains as ROIs (Figure 4A). We identified 799 microdomains across 64 timelapses in seven NTG mice, and 276 microdomains across 64 timelapses in seven APP/PS1 mice. We analyzed the calcium traces (ΔR/R) for each microdomain ROI (Figure 4B). We found that all or nearly all microdomains were active in NTG and APP/PS1 mice (100% active in NTG mice and 99.64% active in APP/PS1 mice) (Figure 4C). The percentage of active microdomains did not differ significantly between NTG and APP/PS1 mice (Chi-square test, χ² = 2.90, p = 0.09) (Figure 4C). The cumulative activity duration in astrocytic microdomains was significantly lower in APP/PS1 mice than in NTG mice (11.41 [6.87, 17.03] s and 23.72 [17.75, 27.62] s, respectively, two-sample Student's t-test p = 2.97 × 10⁻⁶⁶) (Figure 4E). Similarly, the cumulative area under the curve was significantly lower in APP/PS1 mice than that in NTG controls (1.62 [0.8, 3.51] s and 7.94 [4.32, 12.9] s, respectively; two-sample Student's t-test, p = 2.49 × 10⁻⁴³) (Figure 4F). The peak amplitude values of astrocytic microdomains in APP/PS1 mice were significantly lower than those in NTG mice (0.33 [0.26, 0.44] and 0.46 [0.33, 0.66], respectively, two-sample student's t-test p = 8.83 × 10⁻²⁴⁶) (Figure 4G). In contrast, the event rate of astrocytic microdomains was significantly higher in APP/PS1 mice than that in NTG mice (27 [24.86, 28.56] transients/min and 20.28 [17.76, 22.08] transients/min, respectively; two-sample Student's t-test p = 0.004) (Figure 4D). Taken together, these results indicated that calcium transients in astrocytic microdomains occurred more frequently, had shorter durations, lower intensity (AUC), and smaller amplitude in APP/PS1 mice compared to astrocytic calcium transients in NTG controls (Figure 4B).

OPEN IN VIEWER

Figure 4.

Calcium transients in astrocytic microdomains of APP/PS1 mice, occur more frequently, with shorter duration, lower intensity (AUC), and diminished amplitude compared to those in NTG controls. (A) Astrocyte schematic outlining the region of interest (microdomains) where activity was measured. (B) Representative active traces of astrocytic Ca²⁺ activity in the processes from NTG (blue) and APP/PS1 (black) mice. Red line represents threshold. (C) The percentage of active microdomains in NTG and APP/PS1 mice. The chi-square test was used to measure statistical significance. (D) Box plot of the event rate from astrocytic Ca²⁺ activity in microdomains. The event rates were calculated using a ΔR/R threshold of 0.2. (E) Box plot of the cumulative activity duration from astrocytic Ca²⁺ activity in the microdomains. (F) Box plot of the area under the curve of astrocytic Ca²⁺ activity in microdomains. (G) Box plot of ΔR/R peak values from astrocytic Ca²⁺ activity in microdomains. Statistical significance for panels E–G was assessed using a two-sample Student's t-test, **p < 0.01, ***p < 0.001, n.s, not significant.

Calcium transients in the astrocytic endfeet of APP/PS1 mice exhibited lower frequency, shorter activity duration, decreased intensity (AUC), and reduced amplitude compared to those of NTG controls

Next, we analyzed spontaneous calcium transients in astrocytic endfeet (Figure 5) in an analogous manner. First, we confirmed YC 3.6 expression in astrocytic endfeet (Figure 1), and manually selected the endfeet as ROIs (Figure 5A). We identified 210 endfeet across 64 timelapses in seven NTG mice. 212 endfeet across 64 timelapses were identified in seven APP/PS1 mice. We analyzed the calcium traces (ΔR/R) for active and inactive endfoot ROIs (Figure 5B, C). The percentages of active endfeet were comparable in NTG and APP/PS1 mice (89.52% active endfeet in NTG mice and 84.43% active endfeet in APP/PS1 mice, Chi-square test, χ²= 2.41, p = 0.12) (Figure 5D). The event rate of calcium transients in astrocytic endfeet of APP/PS1 mice was significantly lower than that in NTG mice (8.28 [0.96, 21.24] transients/min and 13.08 [2.16, 25.14] transients/min, respectively, two-sample Student's t-test p = 0.013 (Figure 5E). The cumulative activity duration in astrocytic endfeet was also significantly lower in APP/PS1 mice than that in NTG mice (1.31 [0.08, 7.41] s and 3.06 [0.31, 9.96] s, respectively; two-sample Student's t-test, p = 0.003) (Figure 5F). Similarly, the cumulative area under the curve was significantly lower in APP/PS1 mice than in NTG controls (0.15 [0.01, 1.01] s and 0.4 [0.04, 1.53] s, respectively; two-sample Student's t-test, p = 0.005) (Figure 5G). Finally, the peak amplitude values of astrocytic endfeet in APP/PS1 mice were significantly lower than those in NTG mice (0.27 [0.23, 0.34] and 0.28 [0.24, 0.38], respectively; two-sample Student's t-test p = 0.02) (Figure 5H). Based on these findings, we concluded that calcium transients in astrocytic endfeet of APP/PS1 mice occurred less frequently, had shorter durations, lower intensity (AUC), and smaller peak amplitude compared to those in NTG controls (Figure 5B).

OPEN IN VIEWER

Figure 5.

Calcium transients in the astrocytic endfeet of APP/PS1 mice, show a lower frequency, reduced duration, decreased intensity (AUC), and smaller amplitude compared to NTG controls. (A) Astrocyte schematic outlining the region of interest (endfeet) where activity was measured. Representative active (B) and inactive (C) traces of astrocytic Ca²⁺ activity in the endfeet of NTG (blue) and APP/PS1 (black) mice. Red line represents threshold. (D) The percentage of active microdomains in NTG and APP/PS1 mice. The chi-square test was used to measure statistical significance. (E) Box plot of the event rate from astrocytic Ca²⁺ activity in the endfeet. The event rates were calculated using a ΔR/R threshold of 0.2. (F) Box plot of the cumulative activity duration from astrocytic Ca²⁺ activity in the endfeet. (G) Box plot of the area under the curve of astrocytic Ca²⁺ activity in the endfeet. (H) Box plot of ΔR/R peak values from astrocytic Ca²⁺ activity in the endfeet. Statistical significance for panels E–H was assessed using a two-sample Student's t-test, *p < 0.05, **p < 0.01, n.s, not significant.

The inverse relationship between pairwise correlation of astrocytic activity and astrocyte-to-astrocyte distance was impaired in APP/PS1 mice

Finally, we investigated the degree of synchronization in calcium activity of astrocytic somas in NTG and APP/PS1 mice. We calculated pairwise correlations as a measure of somal pair synchronization for each of the 64 timelapses (n = 2808 soma pairs in seven NTG mice; n = 2613 soma pairs in seven APP/PS1 mice). The pairwise correlations between soma pairs were significantly higher in APP/PS1 mice than those in NTG mice (1.2 [0.29, 3.8] × 10⁻³ and 1.07 [0.2, 3.12] × 10⁻³, the Kolmogorov Smirnov (K-S) test, p = 0.002) (Figure 6A). Thus, astrocytic somas were more synchronized in APP/PS1 mice (Figure 6A, B). Notably, when determining the effect of distances between two somas on their pairwise correlations, we found that NTG mice followed an inverse relationship between cell-to-cell distances and pairwise correlations. This indicated that as the distance between two somas increased, their pairwise correlation decreased (Pearson's Linear Correlation Coefficient, R² = 0.36, p = 0.001 in NTG controls.) (Figure 6B). However, this relationship was absent in APP/PS1 mice (Pearson's Linear Correlation Coefficient, R² = 0.06, p = 0.25) (Figure 6B). This loss of statistical significance indicated that the effect of distance on the pairwise correlation of astrocytes was impaired in APP/PS1 mice.

OPEN IN VIEWER

Figure 6.

Pairwise correlation of astrocytic Ca²⁺ activity in somas. (A) Box plot of pairwise correlations (rho-values) between each pair of somas. The Kolmogorov–Smirnov (K-S) test was used to measure statistical significance, **p < 0.01. (B) Scatter plot representing the relationship between the distance between two somas and their pairwise correlation in NTG (blue) and APP/PS1 (black) mice. Dashed lines show the linear regression of the pairwise correlation between soma pairs. Pearson's Linear Correlation Coefficient was used to measure statistical significance.