Circadian rhythmic transcripts in mouse cultured cortical astrocytes.

(A) Experimental scheme for synchronizing circadian rhythms in mouse cultured cortical astrocytes culture with subsequent RNA-sequencing performed over 2 days. (B) Venn diagram displaying the number of circadian rhythmic transcripts identified by two algorithms (q < 0.05 in MetaCycle or BioCycle). (E) Expression profiles of Plat, Il34, Adora1, and Mybph in SS-synchronized cultured astrocytes obtained from RNA-seq data. The q-values for BioCycle (red) and MetaCycle (blue) are displayed in the bottom right corner. (C) Radial histogram of the distribution of phases of rhythmic genes in the astrocyte transcriptome. TASS was transformed to CT for ease of interpretation (TASS 8 hour = CT0). (D) Top 10 enriched GO Biological Process (BP) terms for significant astrocyte circadian rhythmic genes (p < 0.01) identified by Metascape (https://metascape.org). See also S1-S3 and Table S1.

Herp is rhythmically expressed in mouse cultured astrocytes and its expression is controlled by BMAL1.

(A, B) Select core clock genes exhibiting rhythmic expression in cultured astrocytes. Circadian clock-synchronized cultured astrocytes from WT (black) and Bmal1-/- (orange) mice were harvested at the indicated time (CT0 = TASS 8 hour). Expression was analyzed for rhythmicity using MetaCycle, and expression levels of given genes were quantified using RNA-seq (A) and real-time qRT-PCR (B) data. Values are means ± SEM (n = 2; p-values are indicated by insets in graphs). White and gray backgrounds represent subjective day and subjective night, respectively. (C - E) Assessment of Herp siRNA for use in characterizing anti-HERP antisera. Cultured astrocytes were transfected with the indicated siRNA (20nM) and processed for real-time qRT-PCR (C) and Western blot (D, E) analyses at 48 hours post transfection. (C) siRNA-mediated knockdown of Herp mRNA. Values are means ± SEM (n = 2; *p < 0.05, **p < 0.005, ***p < 0.0005, and ****p < 0.00005; independent t-test). (D) Representative Western blot image of HERP knockdown (from three independent experiments). GAPDH served as loading control. (E) Densitometric quantification of HERP levels, normalized to GAPDH levels. Values are means ± SEM (n = 3; ***p < 0.0005; Student’s t-test). (F–I) BMAL1-dependence of HERP expression. Circadian clock-synchronized cultured astrocytes from WT (F, G) and Bmal1-/- (H, I) mice were harvested at the indicated times (CT0 = TASS 8 hours) and processed for Western blot analysis. HERP levels were normalized to total ERK (tERK), which served as a loading control. tERK values at different times were normalized to those at CT4 (defined as 1). P-values are indicated by insets in graphs. White and gray backgrounds represent subjective day and subjective night, respectively. (F, G) BMAL1 and HERP expression over time in astrocytes from WT mice. (F) Representative Western blot image (from five independent experiments). (G) Densitometric quantification of Western blot data from WT mice. (H, I) BMAL1 and HERP expression over time in astrocytes from Bmal1-/- mice. (H) Representative Western blot image (from two independent experiments). (I) Densitometric quantification of Western blot data from Bmal1-/- mice.

Herp knockdown altered ATP-induced ER Ca2+ response.

(A-I) Imaging of subcellular ATP-induced Ca2+ signals in Herp-knockdown and control cultured astrocytes. Cultured astrocytes were co-transfected with 20 μM non-targeting (CTRL) siRNA or Herp siRNA together with ER (A–C), cytosol (D–F) or mitochondrial (Mito; G–I) compartment-specific Ca2+ indicator (denoted at left). At 48 hours post transfection, cultured astrocytes were stimulated with 100 µM ATP and Ca2+ imaging analysis was performed. Images were acquired every 3 seconds. (A, D, G) Representative time-lapse images of each Ca2+ indicator. (B, E, H) ΔF/F0 values over time following ATP application. (C, F, I) Area under the curve values, calculated from panels B, E, and H. (A–C) CTRL siRNA, n = 19; Herp siRNA, n = 22. (D–F) CTRL siRNA, n = 20; Herp siRNA, n = 25. (G–I) CTRL siRNA, n = 16; Herp siRNA, n = 16. (J–M) ITPR1 and ITPR2 protein levels in control and Herp-knockdown astrocytes. Cultured astrocytes were transfected with the indicated siRNA (20nM) and processed for Western blot analysis 48 hours after transfection. Vinculin and GAPDH served as loading control for ITPRs and HERP, respectively. (J and K) ITPR1 protein levels. Itpr1 siRNA-transfected astrocytes were included as controls. (J) Representative Western blot image (from twelve independent experiments). (K) Densitometric quantification of Western blot data showing relative levels of ITPR1 in Herp siRNA-transfected astrocytes compared with CTRL siRNA transfected astrocytes. (L, M) ITPR2 protein levels. Itpr2 siRNA-transfected astrocytes were included as controls. (L) Representative Western blot image (from five independent experiments). (M) Densitometric quantification of Western blot data showing relative levels of ITPR2 in Herp siRNA-transfected astrocytes compared with CNTL astrocytes. Values in graphs are means ± SEM (*p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.00005; independent t-test). See also S4.

ATP-induced ER Ca2+ release varies according to CT.

(A) Schematic depiction of experimental scheme from transfection to live-cell Ca2+ imaging at different CTs. (B–J) Imaging of subcellular ATP-induced Ca2+ signals in cultured astrocytes as a function of CT. Cultured astrocytes were transfected with ER (B–D), cytosolic (E–G) or mitochondrial (Mito; H–J) compartment-specific Ca2+ indicator (denoted at left) and then their circadian rhythm was synchronized by SS. At the indicated CT, astrocytes were stimulated with 100 µM ATP and Ca2+ imaging was performed. (B, E, H) Representative time-lapse images of each Ca2+ indicator. (C, F, I) ΔF/F0 values over time following ATP application. (D, G, J) Area under the curve values, calculated from panels C, F, and I. (B–D) CT22, n = 24; CT34 n = 19. (E–G) CT22, n = 33; CT34, n = 38. (H–J) CT22, n = 50; CT34, n = 54. (K, L) CT-dependent changes in ITPR1 and ITPR2. Cells were harvested at the indicated CTs and processed for Western blot analysis. Vinculin and GAPDH served as loading control for ITPR and BMAL1, respectively. (K) Representative Western blot image (from six independent experiments). (L) Densitometric quantification of Western blot data showing relative levels of ITPR1 and ITPR2 at different CTs. Values in graphs are means ± SEMs (*p < 0.05, ****p < 0.00005; independent t test). See also S4.

CT-dependent ER Ca2+ release is abolished in cultured astrocytes from Bmal1-/- mice.

Cultured astrocytes from Bmal1-/- mice were transfected with ER Ca2+ indicator and then their circadian rhythm was synchronized by SS. At the indicated CTs, astrocytes were stimulated with 100 μM ATP and Ca2+ imaging was performed. (A) Representative time-lapse images of ER Ca2+ indicator. (B) ΔF/F0 values over time following ATP application. (C) Area under the curve values, calculated from panel B. (A–C) CT22, n = 34; CT34, n = 43. (D, E) Cultured astrocytes from Bmal1-/- mice and WT littermates were synchronized by SS. Cells were harvested at the indicated CTs and processed for Western blot analysis. (D) Representative Western blot image (from six independent experiments). GAPDH served as a loading control. (E) Relative levels of HERP in cultured astrocytes from Bmal1-/- mice and WT littermates at different CTs. Values are means ± SEM (*p < 0.05, **p < 0.005, ***p < 0.0005, ****p < 0.00005; repeated measures two-way ANOVA).

ATP-induced pCx43(S368) levels and the conductance of gap junctions varies according to CT.

CT-dependent changes in phosphorylated Cx43 (pCx43). (A–D) Cultured astrocytes, with (C, D) or without (A, B), circadian clock synchronization were treated with 100 µM ATP and processed for Western blot analysis at the indicated times. Vinculin, GAPDH, and/or β-tubulin (TUBB) served as loading controls. The intensity of pCx43(S368) and Cx43 for each sample was normalized to that of Vinculin. (A, B) Cx43 phosphorylation in unsynchronized cultured astrocytes at different times (0, 5, 15, 30 minutes) after stimulation with ATP. (A) Representative Western blot image (from more than three independent experiments). Vinculin served as a loading control. (B) Densitometric quantification of Western blot data, showing relative pCx43(S368)/Cx43 levels. Values are normalized to those for mock-treated samples at time zero (defined as 1). (C, D) Cx43 phosphorylation in SS-synchronized cultured astrocytes at CT22 and CT34 at different times (0, 5, and 15 minutes) after stimulation with ATP. (C) Representative Western blot image (from more than three independent experiments). Vinculin, GAPDH, and β-tubulin (TUBB) served as loading controls. (D) Densitometric quantification of Western blot data, showing relative pCx43(S368)/Cx43 levels. Values are normalized to those for mock-treated samples at time zero for CT22 (defined as 1). (E, F) Changes in Cx43 phosphorylation in vivo. Mice were entrained to a 12-hour light/dark cycle followed by constant dark conditions. At the indicated CTs, the prefrontal cortex area was dissected and processed for Western blot analysis. (E) Representative Western blot image (from three independent experiments). (E) Densitometric quantification of Western blot data showing relative levels of pCx43(S368). (G-L) Circadian rhythm of cultured astrocytes from (G-I) WT and (J-L) Bmal1-/- mice were synchronized by SS. At the indicated CT, gap-FRAP was performed. (G and J) Representative time-lapse images of prebleaching, bleaching and recovery condition during gap-FRAP analysis at CT22 and CT34. (H and K) Ft/F0 values over time following photobleaching (yellow rectangle). (I and L) Recovery % values, calculated from panel H and L, respectively. (G-L) CT22, n = 15; CT34, n = 11. (J-L) CT22, n = 9; CT34, n = 9. Values in graphs are means ± SEMs (*p < 0.05, **p < 0.005, ***p < 0.0005, and ****p < 0.00005; independent t-test). See also S5-S6.

Calcium ion homeostasis

Cluster analysis of astrocyte rhythmic genes.

(A) Histogram showing the distribution of mean expression levels of transcripts. Sets of transcripts with low (red) and high (green) expression were defined by modeling the distribution with a Gaussian mixture model. Dashed vertical lines represent the possible cut-off points for defining the set of highly expressed transcripts. Maroon line, 99.5% cut-off point (TPM 0.40); blue line, 99.0% cut-off point (TPM 0.57); purple line, 95.0% cut-off point (TPM 1.59). (B) Time-averaged expression levels of marker genes for astrocytes, microglia, oligodendrocytes, neurons, and endothelial cells in astrocyte cultures. Values are mean TPM ± S.E.M (n = 2); *, not detected in this analysis. (C) Expression profiles of Plat, Il34, Adora1, and Mybph in SS-synchronized cultured astrocytes obtained from RNA-seq data. The q-values for BioCycle (red) and MetaCycle (blue) are displayed in the bottom right corner. (D) Heatmap generated by plotting 412 candidate genes obtained from RNA-seq data in ascending order according to their MetaCycle phase. The expression level of each gene was normalized using Min-Max normalization.

Comparative analysis of circadian rhythmic transcripts in mouse cultured astrocytes and various tissues from CircaDB.

(A) Number of transcripts that overlapped with 12 tissues in mouse circadian transcriptome datasets from CircaDB (http://circadb.hogeneschlab.org). (B) Comparison of mean phase (ZT) from CircaDB and phase (TASS) from cultured astrocytes for 14 transcripts that are rhythmic in 10 or more tissues. (C) Scatter plot showing phase in CircaDB and TASS from cultured astrocytes for transcripts that are rhythmic in one tissue.

Cluster analysis of astrocyte rhythmic genes.

Cluster analysis of 412 candidate genes performed using ClustVis (https://biit.cs.ut.ee/clustvis/)1. Expression levels of each gene were normalized using Min-Max normalization. Green bar, Cluster 1; orange bar, Cluster 2.

Expression profiles of Itpr1,Itpr2 and Itpr3 in cultured astrocytes obtained from RNA-seq data.

White and gray backgrounds represent subjective day and subjective night, respectively.

Expression profiles of Cx43 and Cx30 in cultured astrocytes obtained from RNA-seq data.

White and gray backgrounds represent subjective day and subjective night, respectively.

CBX treatment abolished gap junction communication.

20μM carbenoxolone (CBX), a gap junction channel block, was applied to cultured astrocytes for 30 minutes before gap-FRAP analysis. (A) Representative time-lapse images of prebleaching, bleaching and recovery condition during gap-FRAP analysis. (B) Ft/F0 values over time following photobleaching (yellow rectangle). (C) Recovery % values, calculated from panel B. Mock, n = 7; CBX, n = 7. Values in graphs are means ± SEMs (*p < 0.05, **p < 0.005, ***p < 0.0005, and ****p < 0.00005; independent t-test).