Brief alcohol exposure alters transcription in astrocytes via the heat shock pathway

Abstract

Astrocytes are critical for maintaining homeostasis in the central nervous system (CNS), and also participate in the genomic response of the brain to drugs of abuse, including alcohol. In this study, we investigated ethanol regulation of gene expression in astrocytes. A microarray screen revealed that a brief exposure of cortical astrocytes to ethanol increased the expression of a large number of genes. Among the alcohol-responsive genes (ARGs) are glial-specific immune response genes, as well as genes involved in the regulation of transcription, cell proliferation, and differentiation, and genes of the cytoskeleton and extracellular matrix. Genes involved in metabolism were also upregulated by alcohol exposure, including genes associated with oxidoreductase activity, insulin-like growth factor signaling, acetyl-CoA, and lipid metabolism. Previous microarray studies performed on ethanol-treated hepatocyte cultures and mouse liver tissue revealed the induction of almost identical classes of genes to those identified in our microarray experiments, suggesting that alcohol induces similar signaling mechanisms in the brain and liver. We found that acute ethanol exposure activated heat shock factor 1 (HSF1) in astrocytes, as demonstrated by the translocation of this transcription factor to the nucleus and the induction of a family of known HSF1-dependent genes, the heat shock proteins (Hsps). Transfection of a constitutively transcriptionally active Hsf1 construct into astrocytes induced many of the ARGs identified in our microarray study supporting the hypothesis that HSF1 transcriptional activity, as part of the heat shock cascade, may mediate the ethanol induction of these genes. These data indicate that acute ethanol exposure alters gene expression in astrocytes, in part via the activation of HSF1 and the heat shock cascade.

Keywords: Alcohol; alcohol response element; astrocytes; gene expression; glia; heat shock factor 1; microarray.

Figure 1

Hierarchical clustering by squared Euclidean distance algorithm on differentially expressed genes and Venn diagram of ethanol- and heat-induced genes in primary astrocyte culture. (A) The graph shows hierarchical clustering of the gene expression pattern after treatment as analyzed by squared Euclidean distance algorithm on differentially expressed genes detected by adjusted ANOVA test (post hoc adjusted by Tukey test, P < 0.05, FDR <0.05). Samples were treated for 1 h with 60 mmol/L ethanol (EtOH) or 42°C heat stress (HS). The columns represent the individual samples and the color scale on top represents the log transformed relative change in expression (red indicates gene induction and blue downregulation of genes). The samples grouped according to treatment and EtOH-induced genes largely overlapped with those induced by HS. (B) Venn diagram of significant EtOH- and HS-induced genes versus controls in primary culture of astrocytes. The intersection denotes the number of genes responsive to both treatments.

Figure 2

Heat-map of Gene Ontology categories enrichment analysis across the ethanol (EtOH) and heat shock (HS) treatments. Only categories with an adjusted FDR-q-value of less than 0.25 in at least one condition are shown in the figure. Colors indicate downregulation (green) or upregulation (red), and values = 1 − FDR-q (with downregulation given negative values). Note that genes induced by EtOH or HS treatments tend to belong to the same GO categories while the enrichment results are more heterogeneous among the downregulated genes.

Figure 3

Ethanol induces heat shock factor 1 (HSF1) protein translocation into the nucleus of cortical astrocytes. (A) Ethanol (EtOH) and heat shock (HS) treatment caused the translocation of HSF1 into the nucleus of primary cultured astrocytes. Immunostaining was performed with an HSF1-specific antibody (red) and DAPI nuclear staining (blue). Cells were also positive for a marker of mature astrocytes, glial fibrillary acidic protein (GFAP; green). (B) Quantification was performed by Pearson's correlation coefficient of pixel intensity scatter plots. The colocalization of HSF1 and DAPI signals increases with EtOH and HS treatment of primary astrocytes. All data are the mean ± SEM of n ≥ 30 cells from two independent cultures and were compared with control by one-way ANOVA with Dunnett's multiple comparison post hoc test (significantly different at the level of ***P < 0.001). The scale bar represents 20 μm.

Figure 4

Ethanol activates the transcription of heat shock protein (Hsp) genes in primary astrocyte culture. (A–F) Increase in Cryab, Hsp27, Hsp40, Hsp70, Hsp90, and Hsp110 mRNA after treatment for 1 h with 60 mmol/L ethanol (EtOH) or heat (HS), as measured by Q-PCR. The data were normalized to Actb mRNA and compared with control samples by one-way ANOVA with Dunnett's multiple comparison post hoc test, n ≥ 8. All data are the mean ± SEM (significantly different at the level of ***P < 0.001).

Figure 5

Ethanol induces the expression of heat shock proteins (HSPs) in primary astrocyte culture. (A–E) Increase in αβ-crystallin, HSP40, HSP70, HSP90, and HSP110 protein levels after treatment for 2 h with 60 mmol/L ethanol (EtOH) or 42°C heat shock (HS) in primary astrocyte culture. Representative immunoblots are shown, with the proteins eIF4E or α-tubulin used as internal standards. The bar graphs to the right of the immunoblots represent the quantification of immunoreactive bands intensities normalized to the internal standard, expressed in arbitrary optical density (OD) units. The data are the mean ± SEM of normalized relative OD values analyzed by one-way ANOVA with Dunnett's multiple comparison post hoc test, n ≥ 3 (significantly differently at the level of *P < 0.05, **P < 0.01, ***P < 0.001).

Figure 6

Induction of ethanol- and heat shock-responsive genes by activated heat shock factor 1 (HSF1). (A–H) Increase in Igfbpl1, Igfbp2, Ctgf, Acas21, Acot11, Aldh1l1, Gas6, and Acta2 mRNA after treatment for 1 h with 60 mmol/L ethanol (EtOH) or 42°C heat stress (HS), or transfection with a constitutively active HSF1 construct (Hsf1-act) as measured by Q-PCR. The data were normalized to Actb mRNA and compared with control samples by one-way ANOVA with Dunnett's multiple comparison post hoc test, n ≥ 6. All data are the mean ± SEM (significantly different at the level of *P < 0.01, **P < 0.005, ***P < 0.001).

Figure 7

Schematic representation of the location and number of alcohol response elements (ARE) in some of the genes sensitive to both ethanol and heat stress treatments in cortical astrocytes. Note the presence of several ARE either in the proximal 5′ region or downstream in the intron/exon region. The relative position of the introns, exons, and ARE has been conserved in the illustration.