Activation of Gs Signaling in Cortical Astrocytes Does Not Influence Formation of a Persistent Contextual Memory Engram

Abstract

Formation and retrieval of remote contextual memory depends on cortical engram neurons that are defined during learning. Manipulation of astrocytic G_q and G_i associated G-protein coupled receptor (GPCR) signaling has been shown to affect memory processing, but little is known about the role of cortical astrocytic G_s-GPCR signaling in remote memory acquisition and the functioning of cortical engram neurons. We assessed this by chemogenetic manipulation of astrocytes in the medial prefrontal cortex (mPFC) of male mice, during either encoding or consolidation of a contextual fear memory, while simultaneously labeling cortical engram neurons. We found that stimulation of astrocytic G_s signaling during memory encoding and consolidation did not alter remote memory expression. In line with this, the size of the mPFC engram population and the recall-induced reactivation of these neurons was unaffected. Hence, our data indicate that activation of G_s-GPCR signaling in cortical astrocytes is not sufficient to alter memory performance and functioning of cortical engram neurons.

Figure 1.

AAV-mediated rM3Ds expression and functionality in mPFC astrocytes. A, Schematic timeline of rM3Ds functionality experiment. Coronal brain section indicating the mPFC region (green) wherein AAV-GFAP::rM3Ds-EGFP (n = 4) was injected and analyzed. CNO was injected 90 min before mice were perfused. B, Representative example of AAV-GFAP::rM3Ds-EGFP expression (green) in the mPFC stained for DAPI (blue). ACC, anterior cingulate cortex; PLC, prelimbic cortex; ILC, infralimbic cortex; FMI, forceps minor of the corpus callosum. Scale bar, 500 µm. C, Top row, S100B staining confirms astrocyte-specific expression of rM3Ds. Bottom row, NeuN staining demonstrates absence of rM3Ds expression in neurons. White arrowheads indicate GFP expression in cell bodies. Scale bar, 10 µm. D, Quantification of colocalization of EGFP⁺/S100B⁺ cells (n = 4) and EGFP⁺/NeuN⁺ cells (n = 4). E, Representative image of Fos-expressing cells. Arrowheads indicate Fos expression in EGFP⁺/NeuN⁻ cells. Left scale bar, 50 µm. Right scale bar, 10 µm.

Figure 2.

Activation of G_s signaling in mPFC astrocytes during memory encoding. A, Schematic and timeline of the experimental design using TRAP2-tdTomato mice. CNO was injected 30 min before CFC and 4TM was injected 1 h after training to tag neurons activated during encoding with tdTomato. On Day 28 after CFC, mice underwent a remote memory test and were perfused 90 min thereafter. B, Schematic of TRAP2-tdTomato labeling. Neuronal activity will induce the transcription of the Fos and CreERT2 genes indicated with the gray and lilac bars, under control of the Fos promotor shown as white bar. Only in the presence of 4TM, CreERT2 will translocate to the nucleus and remove the stop codon located between two loxP sites placed in front of the effector gene tdTomato, allowing the permanent tagging of active cells. C, G_s-DREADD activation did not affect freezing levels during remote retrieval. Unpaired t test: t = 0.32, p = 0.75, control (n = 9), rM3Ds (n = 11). D, Example image of tdTomato-expressing cells. The white arrowhead indicates a tdTomato⁺/S100B⁺ astrocyte. The empty arrowhead indicates a tdTomato⁺/S100⁻ neuron. Scale bar, 50 µm. Extended Data Figure 2-1 shows that rM3Ds-EGFP is expressed in mPFC astrocytes, but not in neurons. E, An increase in tdTomato expression was detected in astrocytes after rM3Ds activation. Mann–Whitney test: U = 0, ***p < 0.001, control (n = 9), rM3Ds (n = 11). F, Representative images showing tdTomato⁺ and Fos⁺ cells in both groups. White arrowheads indicate reactivated (Fos⁺ in tdTomato⁺) neurons. Empty arrowheads indicate tdTomato⁺ neurons that are Fos⁻. Gray arrowheads indicate tdTomato⁺ astrocytes. Scale bar, 50 µm. Extended Data Figure 2-2E shows a zoom-in of Figure 2F. G, Astrocytic G_s pathway activation did not alter the percentage of activated (tdTomato⁺) mPFC neurons during memory encoding. Unpaired t test: t = 0.66, p = 0.51, control (n = 9; 1.06 ± 0.16%), rM3Ds (n = 11; 0.93 ± 0.12%). H, There was no difference in the percentage of activated (Fos⁺) neurons during the remote memory test. Unpaired t test: t = 0.2, p = 0.85, control 8.9 ± 1.4%; n = 9), rM3Ds (8.5 ± 1.6%; n = 11). I, Percentage of Fos⁺ neurons within the tdTomato⁺ and tdTomato⁻ populations. Two-way repeated-measures ANOVA revealed a population effect: F_(1,18) = 60.33, ***p < 0.001. Post hoc Bonferroni’s test: control tdTomato⁺ versus tdTomato⁻ p < 0.0001; rM3Ds tdTomato⁺ versus tdTomato⁻ p < 0.001. Extended Data Figure 2-2 shows the number of DAPI+, tdTomato+, and Fos+ cells per cubic millimeter of all the conditions.

Figure 3.

G_s pathway activation during memory consolidation. A, Schematic and timeline of the experimental design using TRAP2-tdTomato mice. For rM3Ds activation during consolidation, CNO was first injected immediately after CFC and then administered via the drinking water until Day 7 after conditioning. 4TM was injected 1 h after training to tag neurons activated during CFC. Mice underwent a remote memory test on Day 28 after CFC and were perfused 90 min later. B, Depicted is the amount of consumed CNO via drinking water, with the gray dotted line showing the target amount of 0.1 mg/kg. Two-way repeated-measures ANOVA showed no significant difference between the groups: F_(1,18) = 2.709, p = 0.12. C, G_s-DREADD activation during consolidation did not affect freezing during the remote memory test. Unpaired t test: t = 0.48, p = 0.64, control (n = 10), rM3Ds (n = 10). D, CNO-mediated activation of rM3Ds induced a robust increase in tdTomato⁺ astrocytes compared with the control group. Mann–Whitney test: U = 0, ***p < 0.001, control (n = 10), rM3Ds (n = 10). E, Representative images showing tdTomato⁺ and Fos⁺ cells in both groups. White arrowheads indicate reactivated tdTomato⁺/Fos⁺ neurons. Empty arrowheads indicate tdTomato⁺/Fos⁻ neurons. Gray arrowheads indicate tdTomato⁺ astrocytes. Scale bar, 50. F, Astrocytic G_s pathway activation did not alter the percentage of tagged (tdTomato⁺) neurons. Unpaired t test: t = 0.69 p = 0.50, control (n = 10), rM3Ds (n = 10). G, The percentage of activated (Fos⁺) neurons during remote memory retrieval did not differ between groups. Unpaired t test: t = 1.679, p = 0.11, control (n = 10), rM3Ds (n = 10). H, Percentage of Fos⁺ neurons within the tdTomato⁺ and tdTomato⁻ populations. Two-way repeated-measures ANOVA revealed a significant population effect: F_(1,18) = 139.6, p < 0.001, but no interaction, nor main group effect. Post hoc Bonferroni’s test: control tdTomato⁺ versus tdTomato⁻ ***p < 0.001; rM3Ds tdTomato⁺ versus tdTomato⁻ ***p < 0.001. Extended Data Figure 3-1 shows the number of DAPI+, tdTomato+, and Fos+ cells per cubic millimeter of all the conditions.