Gα12 overexpression in hepatocytes by ER stress exacerbates acute liver injury via ROCK1-mediated miR-15a and ALOX12 dysregulation

Abstract

Rationale: Liver injury must be further characterized to identify novel therapeutic approaches. Endoplasmic reticulum (ER) stress may cause hepatocyte death. Gα₁₂ affects cell viability and its expression varies depending on physiological conditions. This study investigated whether hepatocyte-specific Gα₁₂ overexpression affects acute liver injury, and if so, what the underlying mechanisms and treatment strategies are. Methods: All experiments were performed using human liver, hepatocytes, and toxicant injury models with Gna12 KO and/or hepatocyte-specific Gα₁₂ overexpression. RNA-sequencing, immunoblotting, immunohistochemistry, reporter assays, and mutation assays were conducted. Results: Hepatic Gα₁₂ was overexpressed in mice challenged with acetaminophen or other ER stress inducers or in patients with acute liver injury or fibrosis/cirrhosis. Several Gα₁₂ and ER-associated pathways were identified using transcriptomic analysis. Acetaminophen intoxication was characterized by lipid peroxide-induced ferroptosis and was less severe in Gα₁₂-deficient animals and cells. Conversely, Gα₁₂ overexpression in wild-type or Gna12 KO hepatocytes increased hepatotoxicity, promoting lipid peroxidation, inflammation, and ferroptosis. IRE1α-dependent Xbp1 transactivated Gna12. Moreover, Gα₁₂ overexpression enhanced the ability of acetaminophen to induce ALOX12, while downregulating GPX4. The level of miR-15a, herein identified as an ALOX12 inhibitor, was decreased. siRNA knockdown or pharmacological inhibition of ROCK1 prevented dysregulation of ALOX12 and GPX4, rescuing animals from toxicant-induced ferroptosis. These changes or correlations among the targets were confirmed in human liver specimens and datasets of livers exposed to other injurious medications. Conclusions: Gα₁₂ overexpression by ER stress facilitates hepatocyte ferroptosis through ROCK1-mediated dysregulation of ALOX12, and miR-15a, supporting the concept that inhibition of Gα₁₂ overexpression and/or ROCK1 axis may constitute a promising strategy for acute liver injury.

Keywords: ALOX12; GPX4; Gα12; acetaminophen-induced liver injury; miR-15a.

Figure 1

APAP-induced Gα₁₂ overexpression in hepatocytes in association with the unfolded protein response. (A) Heatmap and hierarchical correlation analysis for differentially expressed genes in liver tissue between vehicle- and APAP-treated mice using RNA-seq data (left, n = 3 each). Volcano plot of RNA-seq data. Horizontal and vertical lines indicate the filtering criteria (absolute fold-change ≥ 1.5 and P < 0.05, respectively). Red and blue dots show differentially expressed genes (DEGs) upregulated (382 genes) or downregulated (161 genes) in response to APAP treatment (right, n = 3 each). (B) GSEA enrichment plot of the Biocarta category showing the GPCR pathway (left, NES = 1.6, FDR < 0.1) and Rho/ROCK-related signaling pathways positively correlated with APAP-exposed WT mice (300 mg/kg BW, 6 h) in the Reactome pathway (right, n = 3 each). NES and FDR are shown in the bar graph. (C) Heatmap of Gna subunits in primary hepatocytes treated with APAP or vehicle using microarray dataset (n = 5 each). (D) The levels of major GNA transcripts in HepG2 cells treated with APAP (2 mM, 24 h) relative to vehicle control derived from a public dataset (GSE74000). (E) Gα₁₂ and GRP78 immunohistochemistry and histopathology analyses in mouse liver samples (H&E; scale bar: 200 μm) (left). Percent areas of Gα₁₂, GRP78, and damaged areas were assessed using the Image J program (right). The mice were injected with a single dose of APAP (300 mg/kg BW, 6 h) (n = 3 each). (F) qRT-PCR assays for Gna12 and Hspa5 in APAP-treated WT mice (n = 5 or 7 each). (G) Immunoblottings (upper) for Gα₁₂ and GRP78 in APAP-treated WT mice. The band intensities represent values relative to each respective control (lower) (n = 5 or 6 each). (H) qRT-PCR assays for Gna12 and Hspa5 in AML12 cells treated with APAP (10 mM, 12 h) (n = 3 or 4 each). (I) Immunoblottings (left) for Gα₁₂ and GRP78 in AML12 cells treated with APAP (10 mM, 12 h). The band intensities represent values relative to each respective control (right, n = 3 or 4 each). (J) Immunoblottings for Gα₁₂ and GRP78 in mouse primary hepatocytes treated with APAP (10 mM, 12 h) (n = 3 each). For D-I, the reported values represent mean ± SD (*P < 0.05, **P < 0.01). Statistical significance was tested via two-tailed Student's t-tests.

Figure 2

Xbp1s-dependent Gna12 transactivation in hepatocytes. (A) Ιmmunohistochemistry for Gα₁₂ and GRP78 in the liver of mice treated with a single dose of Tm (2 mg/kg BW, i.p., 72 h) (left). Percent areas of Gα₁₂ and GRP78 were assessed using the Image J program (right, n = 5 or 6 each). scale bar, 200 μm. (B) Immunoblottings for Gα₁₂ and GRP78 in the same samples as in A (left). Densitometric band intensities represent values relative to respective control (right, n = 5 each). (C) Immunoblottings for Gα₁₂ and GRP78 in mouse primary hepatocytes treated with Tm (2 μg/ml, 12 h) (upper left) and densitometric band intensities relative to respective control (upper right, n = 3 each). Immunoblottings for Gα₁₂ and GRP78 in AML12 cells were treated with Tm (2 μg/ml) for the indicated times (lower, n = 3 each; repeated 3 times with similar results). (D) Immunoblottings for Gα₁₂ and IRE1α in AML12 cells treated with tunicamycin (Tm; 2 μg/ml, 12 h) after transfection with IRE1α siRNA (or siCon) (100 nM, 24 h) (upper) and WT-IRE1α (or mock vector) (1 μg, 24 h) (middle, n = 3 or 4 each). Immunoblottings for Gα₁₂, p-PERK, and ATF6 in AML12 cells treated with Tm (2 μg/ml, 12 h) after transfection with siCon or siPERK or siATF6 (100 nM each, 24 h) (lower, n = 3 each; repeated 3 times with similar results; scan values were shown for representative blots). (E) Levels of major transcription factor mRNAs in primary MEF cells treated with Tm (2 μg/ml, 4 h) relative to the vehicle control derived from a publicly available dataset (GSE2082) (n = 3 or 4 each). (F) Immunoblotting analyses for Gα₁₂ and XBP1s in AML12 cells treated with Tm (2 μg/ml, 12 h) after transfection with siXBP1s (or siCon) (100 nM, 24 h) and WT-XBP1s (or mock vector) (1 μg, 24 h) (n = 3 each). (G) qRT-PCR assays for Gna12 and Xbp1s in AML12 cells treated as in F (n = 3 or 4 each). (H) Gna12 promoter-reporter assays. Luciferase activity was measured in AML12 cells after transfection with a Gna12 luciferase construct and a plasmid encoding XBP1s (mock or active form, 0.5 μg, 24 h) (left, n = 4 each). The cells were treated with Tm or a vehicle after transfection with an Xbp1-RE WT or Xbp1-RE mutant luciferase construct (1 μg, 24 h) (right, n = 3 each). For A-E, G, and H, the values represent the mean ± SD (*P < 0.05, **P < 0.01). Statistical significance was tested via two-tailed Student's t-tests or one-way ANOVA with the LSD multiple comparison procedure, where appropriate.

Figure 3

Hepatocyte ferroptosis mediated by Gα₁₂-dependent peroxidation upon APAP exposure. (A) Volcano plot representation of differentially expressed genes between APAP-treated WT and Gna12 KO mice using RNA-seq analysis. The horizontal and vertical lines indicate the filtering criteria (absolute fold-change ≥ 1.5 and raw P < 0.05). The red and blue dots indicate the upregulated (509 genes) or downregulated (190 genes) genes (left, n = 3 each). GO terms associated with ROS production were obtained between WT and Gna12 KO mice treated with APAP (300 mg/kg BW, 6 h) (right, n = 3 each). NES and FDR are shown in the bar graph. (B) GPX4, FTH1, and HMGB1 transcript levels in patients with acute liver failure. Hepatic transcripts levels were assessed using a publicly available dataset (GSE74000) from healthy individuals (normal) or patients with APAP-induced acute liver failure (n = 2 or 3 each). (C) Immunohistochemistry for GPX4 and 4-HNE (scale bar: 200 μm). Percent areas of GPX4 and 4-HNE staining were assessed in the livers of WT or Gna12 KO mice treated with APAP (300 mg/kg BW, 6 h) using the Image J program (n = 4 or 5 each). (D) Immunoblottings for GPX4, 3-NT, and Gα₁₂ in the livers of the same mice as in C. The band intensities represent values relative to their respective control (n = 3 or 4 each). (E) Immunoblottings for GPX4, 3-NT, and Gα₁₂ in WT or Gna12 KO primary hepatocytes treated with APAP (10 mM, 12 h). The band intensities represent values relative to the respective controls (n = 3 each). (F) Histograms of BODIPY 581/591 C11 fluorescence at a 530 nm emission wavelength (FL1 channel) in AML12 cells treated with vehicle or APAP (10 mM, 24 h) after transfection with siCon or siGα₁₂ (100 nM each, 24 h; the experiments were simultaneously performed). Averages of M1 proportion (%) in the FL1 channel were also measured (n = 3 each). For B-F, values were expressed as mean ± SD (*P < 0.05, **P < 0.01). Statistical significance was tested via two-tailed Student's t-tests or one-way ANOVA coupled with Tukey's HSD or the LSD multiple comparison procedure, where appropriate.

Figure 4

Gα₁₂-mediated induction of ALOX12 by APAP intoxication. (A) Venn diagrams showing the overlap of commonly upregulated genes in the liver of mice treated with APAP (300 mg/kg BW, 6 h) versus vehicle treatment (fold change ≥ 2.5 and raw P < 0.05) and WT mice treated with APAP (300 mg/kg BW, 6 h) versus Gna12 KO mice treated with APAP (fold change ≥ 2 and raw P < 0.05) (left, n = 3 each). GO term enrichment (molecular function) of the genes highlighted for functions such as arachidonic acid binding (second rank) (right). (B) KEGG pathway analysis using a cDNA microarray dataset (GSE51694) obtained from the liver of WT and Gna12 KO mice. Ferroptosis-related pathways are indicated in red (left and middle). Heatmap of Alox isoform transcript levels in the livers of WT and Gna12 KO mice (right, n = 3 each). (C) Immunohistochemistry for ALOX12 in the livers of WT or Gna12 KO mice treated with APAP (300 mg/kg BW, 6 h) (n = 3 or 4 each). (D) Immunoblotting for ALOX12 in the liver of WT or Gna12 KO mice treated with APAP (300 mg/kg BW, 6 h). The densitometric band intensities represent values relative to the respective control (n = 3 or 4 each). (E) Immunoblotting for ALOX12 in WT or Gna12 KO primary hepatocytes treated with APAP (10 mM, 12 h) (n = 3). (F) A schematic showing a construct encoding Gα₁₂-WT (Lv-Alb-Gα₁₂) downstream from the albumin promoter. The mice were injected with a single dose of albumin promoter-Gα₁₂-WT lentivirus (or Lv-control) through the tail vein. After one week, the mice were fasted overnight prior to APAP treatment and sacrificed 6 h thereafter. (G) Representative immunohistochemistry (upper) and immunoblottings (lower) for ALOX12 and Gα₁₂ in WT mice treated with APAP (300 mg/kg BW, 6 h) one week after injection with Lv-con or Lv-Gα₁₂ via the tail vein (n = 3 or 4 each). (H) Representative immunohistochemistry for ALOX12 (upper) and immunoblottings (lower) for ALOX12 and Gα₁₂ in WT mice and Gna12 KO mice treated with APAP (300 mg/kg BW, 6 h) one week after injection with Lv-con or Lv-Gα₁₂ via the tail vein (n = 4-6 each). For D, values were expressed as mean ± SD (**P < 0.01). Statistical significance was tested via one-way ANOVA coupled with Bonferroni's method, where appropriate. scale bar: 200 μm.

Figure 5

Identification of miR-15a as an inhibitor of ALOX12 downstream from Gα_12. (A) Venn diagrams showing miR-15a expression among the downregulated miRNAs in the liver of mice treated with APAP (300 mg/kg, 24 h) (Wang K, et al. PNAS, 2009), Huh7 cells with Gα₁₂QL transfection (GSE44079), and those predicted to inhibit ALOX12 based on our Target Scan database analyses; 18 miRNAs were downregulated in the first set, whereas a total of 15 downregulated miRNAs were identified in the second set. Seven miRNAs were predicted to target ALOX12. (B) qRT-PCR assays for miR-15a using the same mice as in Fig. 4G (n = 4-6 each) or Fig. 4H (n = 3-5 each). Experiments were done at the same time and the marked control group (▲) was shared for statistical analysis. (C) qRT-PCR assays for miR-15a in AML12 cells treated with APAP (10 mM, 12 h) after transfection with siGα₁₂ (or siCon) (100 nM, 24 h, n = 5 each) (left), or the plasmid encoding for Gα₁₂ or a mock vector (1 μg, 24 h, n = 5 each) (right). (D) Heatmap of differentially expressed miR-15a target genes based on the GSE51694 dataset from the livers of WT and Gna12 KO mice; Alox12, 6 pro-inflammatory genes, and 8 lipid metabolism genes (n = 3 each). (E) Prediction of miR-15a binding to the 3'-UTR of Alox12 mRNA (upper). Alox12-3'-UTR luciferase assays in AML12 cells transfected with miR-15a ASO (or control ASO) (100 nM, 48 h), or miR-15a mimic (or control mimic) (100 nM, 24 h) (lower, n = 3 or 4 each). (F) Immunoblottings for ALOX12 in AML12 cells transfected with miR-15a ASO (or control ASO) (100 nM, 48 h), or miR-15a mimic (or control mimic) (upper, 100 nM, 24 h). The densitometric band intensities represent values relative to the respective control (lower, n = 3 or 4 each). For B, C, E, and F, values were expressed as mean ± SD (*P < 0.05, **P < 0.01). Statistical significance was tested via two-tailed Student's t-tests or one-way ANOVA with the LSD multiple comparison procedure, where appropriate.

Figure 6

Inhibition of APAP-induced ferroptosis by ROCK1 inhibitors. (A) Immunoblottings for ALOX12, GPX4, and ROCK1 in mouse primary hepatocytes treated with APAP (10 mM, 12 h) after transfection with siROCK1 (or siCon) (100 nM, 24 h). (B) Immunoblottings for ALOX12, GPX4, and ROCK2 in mouse primary hepatocytes treated with APAP (10 mM, 12 h) after transfection with siROCK2 (or siCon) (100 nM, 24 h). (C) Immunoblotting of ALOX12 in AML12 cells transfected with miR-15a ASO (or control ASO) (100 nM, 48 h) after transfection with siROCK1 (100 nM, 24 h). (D) Immunoblotting of ALOX12 in AML12 cells transfected with miR-15a mimic (or control mimic) (100 nM, 24 h) after transfection with Gα₁₂ (1 μg, 24 h). (E) A schematic showing APAP and ripasudil treatment (left). Liver histopathology in WT mice treated with ripasudil (50 mg/kg BW, 5 h) 1 h after APAP treatment (300 mg/kg BW, 6 h) (right, n = 5-7 each). scale bar: 100 μm. (F) Serum ALT and AST activities in the same mice as in E. (G) Immunoblottings for 3-NT, ALOX12, and GPX4 in the same mice as in E. (H) Immunoblotting analyses for ALOX12, GPX4, and p-MLC in WT primary hepatocytes treated with APAP (10 mM, 12 h) 1 h after treatment with ripasudil (50 μM), Y-27632 (10 μM), or netarsudil (2 μM) (n = 3 each). For A-D and G, the band intensities represent values relative to the respective control (n = 3-6 each). For A-D, F, and G, values were expressed as mean ± SD (*P < 0.05, **P < 0.01). Statistical significance was tested via one-way ANOVA coupled with Tukey's HSD or the LSD multiple comparison procedure, where appropriate.

Figure 7

Increase in inflammatory injuries of hepatocytes by Gα₁₂ overexpression. (A) Liver histopathology (H&E), serum alanine transaminase (ALT), and serum aspartate transaminase (AST) activities in WT or Gna12 KO mice treated with APAP (300 mg/kg BW, 6 h) (n = 4-7 each). (B) GSEA plot of GO categories, leading-edge genes (upper, n = 3 each) (NES = 1.94, FDR < 0.1), and qRT-PCR assays in the same mice as in A (lower, n = 3-7 each). (C) Liver histopathology, ALT, and AST activities in the same mice as in Fig. 4G (n = 5 or 6 each). (D) qRT-PCR assays for inflammatory cytokines in the same mice as in Fig. 4G (n = 3-6 each). (E) Liver histopathology (left), ALT and AST activities (right) in the same mice as in Fig. 4H. For C and E, experiments were done at the same time and the marked control groups (▲ and △) were respectively shared for statistical analyses. (F) Flow cytometric analyses for fluorescein isothiocyanate-annexin V and propidium iodide (upper). The average proportion (%) of the upper right portion was measured (lower). AML12 cells were treated with APAP (10 mM, 24 h) after transfection with siCon or siGα₁₂ (100 nM each, 24 h) (n = 3 each). For A-F, values represent the mean ± SD (*P < 0.05, **P < 0.01). Statistical significance was tested via one-way ANOVA coupled with Tukey's HSD or the LSD multiple comparison procedure, where appropriate. scale bar: 100 μm.

Figure 8

Correlations between Gα₁₂-ROCK1, ALOX12, miR-15a, and GPX4 in liver specimens of patients with acute liver injury or liver fibrosis. (A) ROCK1 and ROCK2 transcript levels in the livers of healthy individuals or patients with HBV-acute liver failure (ALF). MHN, massive hepatic necrosis; SHN, submassive hepatic necrosis. The data are reported as a box-and-whisker plot. Box, interquartile range (IQR); whiskers, 5-95 percentiles; horizontal line within the box, median (n = 10 samples from 10 individual normal subjects; n = 8 samples from 2 patients with SHN-ALF; n = 9 samples from 2 patients with MHN-ALF; as described in the GSE38941 database). (B) GNA12, ALOX12, and miR-15a transcript levels in the livers of healthy individuals (n = 5) or patients with DILI (n = 21 for GNA12 and ALOX12, and n = 16 for mir-15a). In some of the patient samples, GNA12 (n = 1), ALOX12 (n = 1) and miR-15a (n = 6) levels were undetectable. (C) Correlations between GNA12 and ALOX12 (left, n = 26), or GNA12 and miR-15a (middle, n = 20), or miR-15a and ALOX12 (right, n = 20) transcripts. In some samples, GNA12 (n = 1), ALOX12 (n = 1) and miR-15a (n = 6) levels were undetectable. (D) Immunoblottings for Gα₁₂ and ALOX12 in healthy individuals (n = 5) or patients with DILI (left, n = 22), and their quantifications (right, n = 27). Representative blots were shown. Short and long exposures were shown for comparison. (E) ROCK1 and ROCK2 transcript levels in the livers of healthy individuals or a large cohort of patients with liver fibrosis (GSE25097, n = 46). (F) Correlations between GNA12 and ALOX12 (left, n = 47), or GNA12 and miR-15a (right, n = 38) transcripts in patients with no fibrosis and liver fibrosis. MiR-15a levels in 9 patients with fibrosis were undetectable. (G) Immunoblottings for Gα₁₂ and ALOX12 in patients with no fibrosis (Metavir score F0) or patients with the portal (Metavir score F1, 2) or septal (Metavir score F3) fibrosis, or those with cirrhotic (Metavir score F4) lesion (upper), and their quantifications (lower, n = 47). Fibrosis was scored as F0 (absent), F1 (portal fibrosis), F2 (portal fibrosis with few septa), F3 (septal fibrosis), and F4 (cirrhosis) according to the Metavir Score System. (H) Immunohistochemistry for Gα₁₂, ALOX12, and GPX4 in patients without fibrosis (n = 2) or septal fibrosis (right, n = 10). Representative liver sections are shown. scale bar: 200 μm. For A and E, Values were expressed as mean ± SD (*P < 0.05, **P < 0.01). For B, D, and G, violin plots show all independent biological replicates with the median as a red straight line and the upper/lower quartiles as black straight lines (*P < 0.05, **P < 0.01). Statistical significance was tested via two-tailed Student's t-tests, Pearson correlation, or one-way ANOVA coupled with Tukey's HSD or the LSD multiple comparison procedure, where appropriate.