Transcriptional regulation in the absence of inositol trisphosphate receptor calcium signaling

Abstract

The activation of IP₃ receptor (IP₃R) Ca²⁺ channels generates agonist-mediated Ca²⁺ signals that are critical for the regulation of a wide range of biological processes. It is therefore surprising that CRISPR induced loss of all three IP₃R isoforms (TKO) in HEK293 and HeLa cell lines yields cells that can survive, grow and divide, albeit more slowly than wild-type cells. In an effort to understand the adaptive mechanisms involved, we have examined the activity of key Ca²⁺ dependent transcription factors (NFAT, CREB and AP-1) and signaling pathways using luciferase-reporter assays, phosphoprotein immunoblots and whole genome transcriptomic studies. In addition, the diacylglycerol arm of the signaling pathway was investigated with protein kinase C (PKC) inhibitors and siRNA knockdown. The data showed that agonist-mediated NFAT activation was lost but CREB activation was maintained in IP₃R TKO cells. Under base-line conditions transcriptome analysis indicated the differential expression of 828 and 311 genes in IP₃R TKO HEK293 or HeLa cells, respectively, with only 18 genes being in common. Three main adaptations in TKO cells were identified in this study: 1) increased basal activity of NFAT, CREB and AP-1; 2) an increased reliance on Ca²⁺- insensitive PKC isoforms; and 3) increased production of reactive oxygen species and upregulation of antioxidant defense enzymes. We suggest that whereas wild-type cells rely on a Ca²⁺ and DAG signal to respond to stimuli, the TKO cells utilize the adaptations to allow key signaling pathways (e.g., PKC, Ras/MAPK, CREB) to transition to the activated state using a DAG signal alone.

Keywords: CREB; Ca2+ dependent transcription; IP3 receptor; NFAT; calcineurin; calcium signaling.

FIGURE 1

NFAT activity in wild-type and IP₃R TKO cells. (A) HEK293 WT (black) and TKO (red) cell lines were co-transfected with fire-fly NFAT-luciferase and Renilla luciferase reporter vectors. After 24h, cells were treated for a further 4 h with either 25 μM carbachol (Cch), 200 nM Ionomycin (Iono) or 100 nM PMA. When used, 0.5 μM cyclosporine A (CsA) was preincubated for 30min. Lysates were assayed for luciferase activity as given in ‘Experimental Procedures’. All fire-fly/Renilla luciferase ratios were normalized to the value obtained for WT cells at baseline. Statistical significance *p < 0.05; ****p < 0.0001; ns = not significant. (B) HEK293 cells were treated with 0.5 µM Thapisgargin (Tg) in the presence or absence of cyclosporine A (CsA) pretreatment as in (A, C) HeLa WT (black) and TKO (red) cell lines were treated with 25 µM Histamine (His) and other reagents as described in (A, D) HEK293 and HeLa WT (black) and TKO (red) cell lines were immunoblotted for RCAN1. In the lower panel RCAN1 levels were normalized to GAPDH and quantitated. The data shown are the mean ± S.E.M.of 3 independent experiments.

FIGURE 2

CREB activity in wild-type and IP₃R TKO cells. (A) HEK293 WT (black) and TKO (red) cell lines were co-transfected with fire-fly CRE-luciferase and Renilla luciferase reporter vectors. Treatment conditions and data analysis were as described for Fig 1A. Vasoactive intestinal peptide (VIP) was used at 0.5 μM. (B) HeLa WT (black) and TKO (red) cell lines were treated with reagents as described in Panel A except 25 μM Histamine (His) was used as agonist. (C) Representative immunoblots for pCREB (ser-133) and total CREB in HEK293 WT and TKO lysates are shown. Cells were treated for 10min with 25 µM carbachol (Cch), 1 μM Ro-31–8220 (Ro-31), a pan-PKC inhibitor, or both. In the lower panel p-CREB data was expressed as a ratio of total CREB and normalized to baseline WT values. (D) Representative immunoblots and quantitation of pCREB in HeLa WT and TKO lysates are shown. Conditions were the same as used for (C) except for the use of 25 μM Histamine (His) as agonist. (E) HEK293 WT and TKO cells were transfected with scrambled siRNA (Scr) or siRNA for PKCδ for 48 h and expression levels of PKCδ was measured by immunoblotting using GAPDH as a loading control. Expression was normalized to WT levels and the data shown are the mean ± S.E.M. of 3 independent experiments. (F) HEK293 WT and TKO cells were transfected with siRNA as described in (E) and then stimulated for 10min with 25 μM carbachol (Cch). Changes in p-CREB and total CREB were measured by immunoblotting. Fold activation by Cch was quantitated in 3 experiments in the lower panel. (G, H) Conditions were the same as described for (E, F) except HeLa WT and TKO cells were used and were stimulated with 25 μM Histamine (His) as agonist. In (C,D,F, H) separate gels were used to measure phospho- and total CREB.

FIGURE 3

AP-1 activity in wild-type and IP₃R TKO cells. (A) HEK293 WT (black) and TKO (red) cell lines were co-transfected with fire-fly AP1-luciferase and Renilla luciferase reporter vectors. Treatment conditions and data analysis were as described for Figure 1A. (B) HeLa WT (black) and TKO (red) cell lines were treated with reagents as described in Panel A except 25 μM Histamine (His) was used as agonist. (C) Lysates were prepared from HEK293 WT and TKO cells treated with the indicated concentrations of Cch for 4 h. Samples were immunoblotted for c-fos and quantitated using GAPDH as a loading control (G). Data were normalized to the WT unstimulated values. (D) fosB; (E) c-jun; (F) junB. The data shown are the mean ± S.E.M. of 3 independent experiments. All data are statistically significant from WT unstimulated values at p < 0.05 with only the data not significantly different marked “ns”.

FIGURE 4

ROS handling is altered in IP₃R TKO cells. (A) The rate of extracellular H₂O₂ generation was measured with Amplex Red in WT and TKO HEK293 cells grown in 96 well plates as described in ‘Experimental Procedures’. Antimycin A (AA; 0.5 μM) was used as a positive control while AA and Rotenone (Rot; 5 µM), diphenyleneiodonium chloride (DPI; 5 µM) and 3-aminotriazole (3-AT; 5 mM) were used to inhibit the electron transport chain, NADPH oxidase and catalase respectively. The data shown are box plots showing mean (small box), median (horizontal line) 25th and 75th percentile (box) and 1.5 x interquartile range (whiskers). Statistics are one-way ANOVA following a normality test and subject to Tukey post hoc analysis. *-**** represents p ≤ 0.05- p ≤ 0.0001 respectively. (B) Conditions were as in (A) but using HeLa WT and TKO cells. (C) Mitochondrial superoxide was measured in WT and TKO HEK293 cells grown in 96 well plates with MitoSox. The data shown are individual measurements made in 3 separate experiments with the data normalized to the mean value of WT cells. (E–H) Key antioxidant enzyme levels were immunoblotted in HEK293 and HeLa lysates using tubulin as the loading control. Data were quantitated and normalized to the WT unstimulated values. The data are the mean ± S.E.M. of 3 independent experiments. All data are statistically significant from WT unstimulated values at p < 0.05 with only the data not significantly different marked “ns”. (E) Superoxide dismutase-1 (SOD1). (F) Superoxide dismutase-2. (G) Catalase. (H) Tubulin. (I) HEK293 and HeLa WT (black) and TKO (red) cell lines were co-transfected with Nrf2/ARE-luciferase and Renilla luciferase reporter vectors. After 24 h, cells lysates were assayed for promoter activity.

FIGURE 5

Analysis of the HEK293 RNAseq data set for Wild-type versus IP₃R TKO cells. (A). Distribution of the 828 differentially expressed genes is shown as a volcano plot. The default cut off for FDR was set at 1% and |log₂ FC| >1. The top 30 genes are indicated. (B). Manually curated gene sets for Ca²⁺ signaling pathways, transcription factors and mitochondrial genes (MitoCarta 3.0) were used to identify hits within the DEG gene list. The significance level of the gene changes measured by FDR is indicated in the inset. Genes in black were upregulated and genes in blue were downregulated. The origin of the curated gene sets are given in “Materials and Methods”. (C). Bubble plot of KEGG enrichment analysis of DEGs. Top 20 KEGG pathways (p ≤ 0.05) are presented. Y-axis represents pathways; X-axis represents rich factor; (rich factor equals the ratio between the DEGs and all annotated genes enriched in the pathway). The color and size of each bubble represent enrichment significance and the number of DEGs enriched in a pathway, respectively.

FIGURE 6

Analysis of the HeLa RNAseq data set for Wild-type versus IP₃R TKO cells. (A). Distribution of the 310 differentially expressed genes is shown as a volcano plot. The default cut off for FDR was set at 1% and |log₂ FC| >1. The top 30 genes are indicated. (B). Manually curated gene sets for Ca²⁺ signaling pathways, transcription factors and mitochondrial genes (MitoCarta 3.0) were used to identify hits within the DEG gene list. The significance level of the gene changes measured by FDR is indicated in the inset. Genes in black were upregulated and genes in blue were downregulated. The origin of the curated gene sets are given in “Materials and Methods”. (C). Bubble plot of KEGG enrichment analysis of DEGs. Top 20 KEGG pathways (p ≤ 0.05) are presented. Y-axis represents pathways; X-axis represents rich factor; (rich factor equals the ratio between the DEGs and all annotated genes enriched in the pathway). The color and size of each bubble represent enrichment significance and the number of DEGs enriched in a pathway, respectively.

FIGURE 7

Common genes or pathways altered in the HEK293 or HeLa data sets. (A) Venn diagram showing the common genes between the HEK293 and HeLa data sets with (B) listing the 18 common genes. The significance level of the gene changes measured by FDR is indicated in the inset. Genes in black were upregulated and genes in blue were downregulated. (C, D) Gene set enrichment analysis (GSEA) was carried out using a Hallmark gene set found in the Molecular Signatures Database (http://www.gsea-msigdb.org/gsea/msigdb/collections.jsp). The statistical criteria for significance was a nominal p val <0.05 and a false discovery rate q-val <0.25.The top 20 enriched or de-enriched categories in the TKO cells based on normalized enrichment score (NES) are shown. Pathways in both data sets that were unique (black), common (red), or common but changing in opposite directions (green) were colored as indicated.

FIGURE 8

Hypothetical scheme showing agonist-mediated activation of signaling in WT and TKO cells In WT cells both Ca²⁺ mobilization and DAG formation mediate the transition of signaling pathways from ‘resting’ to an ‘activated’ state. In TKO cells many signaling pathways have adapted by adjusting their Ca²⁺ sensitivity to be active at resting Ca²⁺ (e.g., NFAT), increasing their reliance on the DAG signal alone (e.g., CREB via PKCδ) and becoming sensitized to increased ROS (e.g., Jnk, AP-1).