Astrocytic PERK deficiency drives prefrontal circuit dysfunction and depressive-like behaviors
Data files
Dec 16, 2025 version files 2.76 MB
-
README.md
7.03 KB
-
SourceData_Fig_1.xlsx
20.79 KB
-
SourceData_Fig_2.xlsx
20.44 KB
-
SourceData_Fig_3.xlsx
22.69 KB
-
SourceData_Fig_4.xlsx
17.11 KB
-
SourceData_Fig_5.xlsx
25.54 KB
-
SourceData_Fig_6.xlsx
2.59 MB
-
SourceData_Fig_7.xlsx
38.08 KB
-
SourceData_Fig_8.xlsx
20.09 KB
Abstract
Major depressive disorder (MDD) is associated with dysfunction in prefrontal cortex (PFC) circuits, yet the glial mechanisms underlying these abnormalities remain unclear. Here, downregulation of the endoplasmic reticulum (ER) stress sensor PERK in PFC astrocytes is identified as a mechanistic contributor to depression‐related phenotypes. PERK expression is markedly reduced in PFC astrocytes from individuals with MDD and in two chronic‐stress mouse models. Astrocyte‐specific PERK deletion in stress‐naïve mice is sufficient to induce robust depressive‐like behaviors and widespread PFC circuit pathology, including dendritic spine loss, pyramidal neuron hypoactivity, and weakened functional connectivity. Mechanistically, PERK‐deficient astrocytes display reduced Nrf2 abundance, dysregulated ER and cytosolic Ca2+ dynamics, and decreased expression of the synaptogenic protein thrombospondin‐1 (TSP1). Restoring astrocytic TSP1 via a blood‐brain barrier‐penetrant adeno‐associated virus rescues PFC circuit function and reverses depressive‐like behaviors. These findings establish astrocytic PERK deficiency as a sufficient driver of synaptic and network dysfunction underlying depressive phenotypes and highlight astrocyte‐directed TSP1 augmentation as a potential therapeutic strategy for MDD.
Dataset DOI: 10.5061/dryad.mkkwh71dq
Description of the data and file structure
This dataset contains the individual source data supporting figures and analyses in the associated article.
Files and variables
File: SourceData_Fig_1.xlsx
Description: Source data for Figure 1, panels C, D, F–H, and J–K.
Sheet Fig. 1C: quantification of GFAP puncta number per astrocyte and EIF2AK3 puncta number per GFAP+ astrocyte in control and MDD samples.
Sheet Fig. 1D: quantification of EIF2AK3 puncta number in SNAP25+ neurons and AIF1+ microglia in control and MDD samples.
Sheet Fig. 1F: quantification of Eif2ak3 puncta number in S100β+ astrocytes of WT mice treated with vehicle or corticosterone (CORT).
Sheet Fig. 1G: immunoblot analysis of PERK protein levels (fold change) in astrocytes isolated from the mouse mPFC after chronic CORT treatment or chronic restraint stress (CRS).
Sheet Fig. 1H_EPM: Pearson correlation between astrocytic PERK protein levels (%) and the animals’ anxiety-like behavior, as measured by time (seconds) spent in the open arms of the elevated plus maze (EPM).
Sheet Fig. 1H_TST: Pearson correlation between astrocytic PERK protein levels (%) and the animals’ depressive-like behavior, as measured by immobility time (seconds) in the tail suspension test (TST).
Sheet Fig. 1H_FST: Pearson correlation between astrocytic PERK protein levels (%) and the animals’ depressive-like behavior, as measured by immobility time (seconds) in the forced swim test (FST).
Sheet Fig. 1J: immunoblot analysis of total eIF2α and Nrf2 levels (fold change) in mPFC astrocytes of mice treated with vehicle or CORT.
Sheet Fig. 1K: immunofluorescent detection of puromycin incorporation (fold change) in mPFC astrocytes as a readout of de novo protein synthesis.
File: SourceData_Fig_2.xlsx
Description: Source data for Figure 2, panels B, C, D, and F.
Sheet Fig. 2B: quantification of Eif2ak3 puncta number per Aldh1l1+ astrocyte in WT and astrocytic PERK cKO mice (n = 5 mice per group).
Sheet Fig. 2C: immunoblot analysis of Nrf2 protein levels (fold change) in astrocytes isolated from the mPFC of WT and PERK cKO mice (n = 5 mice per group).
Sheet Fig. 2D: behavioral performance of WT (n = 12) and PERK cKO (n = 14) mice in the open field (total distance traveled, meter; center time, second), elevated plus maze (open-arm time, second), tail suspension (immobility, second), forced swim (immobility, second), and sucrose preference (%) tests.
Sheet Fig. 2F: behavioral outcomes after PL-mPFC-targeted astrocytic PERK deletion (n = 8 mice per group) in the elevated plus maze, tail suspension, forced swim, and sucrose preference tests.
File: SourceData_Fig_3.xlsx
Description: Source data for Figure 3, panels B–D, F, and G.
Sheet Fig. 3B: quantification of baseline endoplasmic reticulum (ER) Ca2+ levels (arbitrary units) in astrocytes with and without PERK cKO.
Sheet Fig. 3C: quantification of ER Ca2+ levels in astrocytes from mice exposed to vehicle or chronic CORT, as measured by G-CEPIA1er fluorescence (arbitrary units).
Sheet Fig. 3D: quantification of peak ER Ca2+ response amplitude (ΔF/F₀) in WT and PERK cKO mice.
Sheet Fig. 3F: frequency (per minute) and amplitude (peak ΔF/F₀) of astrocytic cytosolic Ca2+ events in WT and PERK cKO mice at rest.
Sheet Fig. 3G: quantification of amplitude (peak ΔF/F₀) and area under the curve (AUC) for air-puff-evoked astrocytic cytosolic Ca2+ responses in WT and PERK cKO mice.
File: SourceData_Fig_4.xlsx
Description: Source data for Figure 4, panels C, E, and I–L.
Sheet Fig. 4C: quantification of perisynaptic astrocyte Ca2+ activity (2-min AUC, %) within 2.5 µm of dendritic spine heads.
Sheet Fig. 4E: normalized changes in spine-head size during a 60-min imaging session.
Sheet Fig. 4I: dendritic spine elimination and formation (%) during P40–P50, P50–P60, and P60–P70 (n = 5 mice per group).
Sheet Fig. 4J: net dendritic spine loss (%) across P40–P70.
Sheet Fig. 4K: quantification of PSD95 puncta (fold change) in the mPFC of WT and PERK cKO mice.
Sheet Fig. 4L: quantification of Homer1 puncta (fold change) in the mPFC of WT and PERK cKO mice.
File: SourceData_Fig_5.xlsx
Description: Source data for Figure 5, panels D, G, H, and J.
Sheet Fig. 5D: quantification of dendritic Ca2+ transient frequency (per min) in WT and PERK cKO mice.
Sheet Fig. 5G: quantification of somatic Ca2+ transient frequency (per min) in WT and PERK cKO mice.
Sheet Fig. 5H: mean pairwise correlation coefficients of somatic Ca2+ transients in WT and PERK cKO mice.
Sheet Fig. 5J: peak amplitude (ΔF/F₀) and AUC (4-s window) of population neuronal Ca2+ responses at struggle onset in WT and PERK cKO mice.
File: SourceData_Fig_6.xlsx
Description: Source data for Figure 6, panels B–E and G–J.
Sheet Fig. 6_BulkRNAseq: bulk RNA-seq raw data collected from the mPFC of WT and astrocytic PERK cKO mice.
Sheet Fig. 6G: RT-qPCR analysis of Thbs1 mRNA expression in WT and PERK-deficient astrocytes.
Sheet Fig. 6H: quantification of Thbs1 puncta number per S100β⁺ astrocyte in the mPFC of WT and PERK cKO mice.
Sheet Fig. 6I: quantification of THBS1 puncta number per GFAP⁺ astrocyte in the anterior PFC of control and MDD subjects.
Sheet Fig. 6J: quantification of Thbs1 puncta number per S100β⁺ astrocyte in the mPFC of WT mice exposed to vehicle or CORT.
File: SourceData_Fig_7.xlsx
Description: Source data for Figure 7, panels B, C, and F–H.
Sheet Fig. 7B: quantification of astrocytic TSP1 expression (fold change) in the mPFC of AAV5-Gfap-TSP1-treated mice versus EGFP controls.
Sheet Fig. 7C: behavioral outcomes in the open field, elevated plus maze, tail suspension, and forced swim tests.
Sheet Fig. 7F: quantification of neuronal Ca2+ transient frequency (per min) across groups.
Sheet Fig. 7G: quantification of mean pairwise correlation coefficients of neuronal Ca2+ transients across groups.
Sheet Fig. 7H: behavioral outcomes in the open field, elevated plus maze, tail suspension, and forced swim tests.
File: SourceData_Fig_8.xlsx
Description: Source data for Figure 8, panels A–C.
Sheet Fig. 8A: behavioral performance of CORT-treated mice receiving AAV-PHP.eB-Gfap-TSP1 or EGFP before CORT administration in the open field, elevated plus maze, tail suspension, and forced swim tests.
Sheet Fig. 8B: behavioral performance of CRS-exposed mice receiving AAV-PHP.eB-Gfap-TSP1 or EGFP before CRS in the open field, elevated plus maze, tail suspension, and forced swim tests.
Sheet Fig. 8C: behavioral performance of CORT-treated mice receiving AAV-PHP.eB-Gfap-TSP1 or EGFP after CORT administration in the open field, elevated plus maze, tail suspension, and forced swim tests.
Detailed descriptions of data collection and analysis methods are provided in the associated article.