Acrolein Is a Pathogenic Mediator of Alcoholic Liver Disease and the Scavenger Hydralazine Is Protective in Mice

Abstract

Background & aims: Alcoholic liver disease (ALD) remains a major cause of morbidity and mortality, with no Food and Drug Administration-approved therapy. Chronic alcohol consumption causes a pro-oxidant environment and increases hepatic lipid peroxidation, with acrolein being the most reactive/toxic by-product. This study investigated the pathogenic role of acrolein in hepatic endoplasmic reticulum (ER) stress, steatosis, and injury in experimental ALD, and tested acrolein elimination/scavenging (using hydralazine) as a potential therapy in ALD.

Methods: In vitro (rat hepatoma H4IIEC cells) and in vivo (chronic+binge alcohol feeding in C57Bl/6 mice) models were used to examine alcohol-induced acrolein accumulation and consequent hepatic ER stress, apoptosis, and injury. In addition, the potential protective effects of the acrolein scavenger, hydralazine, were examined both in vitro and in vivo.

Results: Alcohol consumption/metabolism resulted in hepatic accumulation of acrolein-protein adducts, by up-regulation of cytochrome P4502E1 and alcohol dehydrogenase, and down-regulation of glutathione-s-transferase-P, which metabolizes/detoxifies acrolein. Alcohol-induced acrolein adduct accumulation led to hepatic ER stress, proapoptotic signaling, steatosis, apoptosis, and liver injury; however, ER-protective/adaptive responses were not induced. Notably, direct exposure to acrolein in vitro mimicked the in vivo effects of alcohol, indicating that acrolein mediates the adverse effects of alcohol. Importantly, hydralazine, a known acrolein scavenger, protected against alcohol-induced ER stress and liver injury, both in vitro and in mice.

Conclusions: Our study shows the following: (1) alcohol consumption triggers pathologic ER stress without ER adaptation/protection; (2) alcohol-induced acrolein is a potential therapeutic target and pathogenic mediator of hepatic ER stress, cell death, and injury; and (3) removal/clearance of acrolein by scavengers may have therapeutic potential in ALD.

Keywords: ADH, alcohol dehydrogenase; ALD, alcoholic liver disease; ALDH, aldehyde dehydrogenase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; ATF, activating transcription factor; Apoptosis; CHOP; CHOP, CCAAT/enhancer-binding protein homologous protein; CYP2E1, cytochrome P4502E1; ER, endoplasmic reticulum; FDP-lysine, Nε-(3-formyl-3,4-dehydropiperidino)lysine; GRP, glucose regulated protein; GSTP, glutathione-s-transferase-Pi; IRE1, inositol-requiring enzyme 1; JNK, cJun N-terminal kinase; LPO, lipid peroxidation; Lipid Peroxidation; NIAAA, National Institute on Alcohol Abuse and Alcoholism; PERK, protein kinase RNA-like endoplasmic reticulum kinase; PUFA, polyunsaturated fatty acids; TRAF, TNF receptor-associated factor; TUNEL, terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling; Therapeutic; UPR, unfolded protein response; XBP1, X-box binding protein-1; mRNA, messenger RNA; siRNA, small interfering RNA.

Figure 1

Alcohol consumption leads to accumulation of acrolein-protein adducts, up-regulation of ADH and CYP2E1, and down-regulation of GSTP in mice livers. (A) Accumulation of acrolein adducts in mice livers by immunohistochemistry using specific FDP-lysine antibodies (magnification, 20× and 80×). (B) Quantification of acrolein adducts observed in panel A. Means ± SEM, n = 6 mice. ***P < .001 compared with control by analysis of variance–Bonferroni analysis. (C) ADH, CYP2E1, and GSTP protein levels in mice livers. Numbers represent mean densitometry ratios normalized to corresponding β-actin levels. (D) ADH, CYP2E1, and GSTP mRNA levels. Means ± SEM, n = 6 mice. ***P < .001 compared with control by analysis of variance–Bonferroni analysis. C, control; E, alcohol.

Figure 2

Effect of pharmacologic modulators of alcohol metabolism on alcohol-induced acrolein adduct accumulation in cultured hepatic cells. (A) Accumulation (magnification, 20×) and quantification of acrolein adducts by immunocytochemistry using specific FDP-lysine antibodies in H4IIEC cells treated for 24 hours as follows: untreated control cells (C) or cells treated with 200 mmol/L alcohol (E), 50 μmol/L acetaldehyde (AA50) or 100 μmol/L acetaldehyde (AA100). Means ± SEM (n = 4). **P < .01 and ***P < .001 compared with control by analysis of variance–Bonferroni analysis. (B) Accumulation of acrolein adducts (magnification, 20×) in H4IIEC cells treated for 24 hours as follows: untreated control cells (C), cells treated with 200 mmol/L alcohol alone (E), or 200 mmol/L alcohol in the presence of 10 μmol/L (N-(1,3-Benzodioxol-5-ylmethyl)-2,6-dichlorobenzamide) ((N-(1,3-Benzodioxol-5-ylmethyl)-2,6-dichlorobenzamide)+E), 10 μmol/L 4-methyl pyrazole (4MP+E), 10 μmol/L Pyrazole (PYR+E), or 10 μmol/L allyl sulfide (AS+E). (C) Quantification of acrolein adducts in panel B. Means ± SEM (n = 4). **P < .01 and ***P < .001 compared with control by analysis of variance–Bonferroni analysis. (D) Inhibition of GSTP mRNA and protein by siRNA transfection in H4IIEC cells. For polymerase chain reaction analysis: means ± SEM by analysis of variance–Bonferroni analysis. *P < .05 and **P < .01 compared with control. For Western blot: numbers represent mean densitometry ratios normalized to β-actin. (E) Accumulation of acrolein adducts (magnification, 20×) in transfected H4IIEC cells. (F) Quantification of acrolein adducts in panel E. Means ± SEM, n = 4. *P < .05 and ***P < .001 compared with control by analysis of variance–Bonferroni analysis. C, control; E, 200 mmol/L alcohol; A, 20 μmol/L acrolein; nt, not transfected; si, GSTP siRNA; and sc, scrambled RNA.

Figure 2

Effect of pharmacologic modulators of alcohol metabolism on alcohol-induced acrolein adduct accumulation in cultured hepatic cells. (A) Accumulation (magnification, 20×) and quantification of acrolein adducts by immunocytochemistry using specific FDP-lysine antibodies in H4IIEC cells treated for 24 hours as follows: untreated control cells (C) or cells treated with 200 mmol/L alcohol (E), 50 μmol/L acetaldehyde (AA50) or 100 μmol/L acetaldehyde (AA100). Means ± SEM (n = 4). **P < .01 and ***P < .001 compared with control by analysis of variance–Bonferroni analysis. (B) Accumulation of acrolein adducts (magnification, 20×) in H4IIEC cells treated for 24 hours as follows: untreated control cells (C), cells treated with 200 mmol/L alcohol alone (E), or 200 mmol/L alcohol in the presence of 10 μmol/L (N-(1,3-Benzodioxol-5-ylmethyl)-2,6-dichlorobenzamide) ((N-(1,3-Benzodioxol-5-ylmethyl)-2,6-dichlorobenzamide)+E), 10 μmol/L 4-methyl pyrazole (4MP+E), 10 μmol/L Pyrazole (PYR+E), or 10 μmol/L allyl sulfide (AS+E). (C) Quantification of acrolein adducts in panel B. Means ± SEM (n = 4). **P < .01 and ***P < .001 compared with control by analysis of variance–Bonferroni analysis. (D) Inhibition of GSTP mRNA and protein by siRNA transfection in H4IIEC cells. For polymerase chain reaction analysis: means ± SEM by analysis of variance–Bonferroni analysis. *P < .05 and **P < .01 compared with control. For Western blot: numbers represent mean densitometry ratios normalized to β-actin. (E) Accumulation of acrolein adducts (magnification, 20×) in transfected H4IIEC cells. (F) Quantification of acrolein adducts in panel E. Means ± SEM, n = 4. *P < .05 and ***P < .001 compared with control by analysis of variance–Bonferroni analysis. C, control; E, 200 mmol/L alcohol; A, 20 μmol/L acrolein; nt, not transfected; si, GSTP siRNA; and sc, scrambled RNA.

Figure 3

Alcohol-induced accumulation of acrolein adducts causes hepatic ER stress in mice livers. (A) ATF3 and ATF4 mRNA levels. Means ± SEM, n = 6 mice. **P < .01 compared with control by the Student t test. (B) ATF3 and ATF4 protein levels. (C) Phospho-PERK and phospho-eukaryotic translation initiation factor 2 α (eIF2α) protein levels. (B and C) Numbers represent mean densitometry ratios normalized to corresponding control proteins (β-actin, total PERK, or eIF2α). C, control; E, alcohol.

Figure 4

Alcohol-induced accumulation of acrolein adducts does not induce UPR adaptive responses in mice livers. (A) GRP78 and GRP94 mRNA levels. Means ± SEM, n = 6 mice. *P < .05 and **P < .01 compared with control by the Student t test. (B) GRP78 and GRP94 protein levels. (C) Protein levels of ATF6 and cleaved/active ATF6, phospho-IRE1, and total IRE1. (D) XBP1 splicing by semiquantitative end point reverse-transcription polymerase chain reaction visualized by agarose gel electrophoresis. (E) Protein levels of XBP1s and XBP1u. (B, C, and E) Numbers represent mean densitometry ratios normalized to corresponding control proteins (β-actin or total IRE1). C, control; E, alcohol.

Figure 5

Alcohol-induced hepatic acrolein build-up and consequent ER stress leads to proapoptotic signaling in mice livers. (A) Phospho-JNK and total JNK protein levels. (B) Pro- and cleaved/active caspase-12 protein levels. (C) CHOP mRNA levels. Means ± SEM, n = 6 mice. **P < .01 compared with control by the Student t test. (D) CHOP protein levels. (A, B, and D) Numbers represent mean densitometry ratios normalized to corresponding control proteins (total JNK or β-actin). C, control; E, alcohol.

Figure 6

Alcohol-induced acrolein and ER stress leads to steatosis, hepatocyte apoptosis, and liver injury in mice livers. (A) Hepatic steatosis (by H&E; magnification, 40×) and Oil Red O staining; magnification, 20×). (B) Apoptosis by TUNEL staining (magnification, 20×), with quantification of apoptosis. Means ± SEM, n = 6 mice. ***P < .001 compared with control by Student t test. (C) Liver injury: serum ALT and AST. Means ± SEM, n = 6 mice. **P < .01 and ***P < .001 compared with control by the Student t test. C, control; E, alcohol.

Figure 7

Acrolein mimics the in vivo effects of alcohol and causes ER stress and cell death in cultured hepatic cells. (A) Accumulation and quantification of acrolein adducts (immunocytochemistry using specific FDP-lysine antibodies; magnification, 20×) in H4IIEC cells treated for 24 hours. Means ± SEM, n = 4. **P < .01 and ***P < .001 compared with control by analysis of variance–Bonferroni analysis. (B) ATF3, ATF4, GRP78, GRP94, and CHOP mRNA levels at 6 hours. Means ± SEM, n = 3 experiments. *P < .05 and **P < .01 compared with control by analysis of variance–Bonferroni analysis. (C) ATF3, ATF4, GRP78, GRP94, and CHOP protein levels from untreated cells (C) or cells treated for 24 hours with alcohol at 50 mmol/L (E50), 100 mmol/L (E100), 200 mmol/L (E200), or acrolein at 20 μmol/L (A20) or 30 μmol/L (A30). Immunoblot analysis was repeated 3 times with similar results, and representative blots are shown. Numbers denote densitometry ratios normalized to corresponding β-actin levels. (D) Cell survival by 3, (4, 5-dimethylthiazol-2-yl) 2, 5-diphenyltetrazolium bromide assay (24 h). Means ± SEM, n = 3 experiments. *P < .05 compared with control by analysis of variance–Bonferroni analysis. C, control; E, 200 mmol/L alcohol; and A, 30 μmol/L acrolein.

Figure 8

Acrolein scavengers show protective effects both in vitro and in vivo. (A) Cell survival in H4IIEC cells by 3, (4, 5-dimethylthiazol-2-yl) 2, 5-diphenyltetrazolium bromide assay (24 h). Means ± SEM, n = 3 experiments. *P < .05 compared with the corresponding treatments of alcohol (E) or acrolein (A) alone, by analysis of variance–Bonferroni analysis. C, control; E, 200 mmol/L alcohol; A, 30 μmol/L acrolein; HYD, 50 μmol/L hydralazine; CAR, 100 μmol/L carnosine. (B–F) Data are from livers of mice. E, alcohol; Hyd+E, hydralazine+alcohol. (B) Hepatic acrolein adduct accumulation in mice (magnification, 20×), hepatic steatosis by H&E (magnification, 80×), and Oil Red O staining (magnification, 20×). (C) ATF3, ATF4, GRP78, GRP94, and CHOP mRNA levels. Data are presented as means ± SEM (n = 6 mice). *P < .05 and ***P < .001 compared with alcohol (E) by analysis of variance–Bonferroni analysis. (D) Protein levels of phospho-JNK and total JNK, pro- and cleaved/active caspase 12, and CHOP. Numbers represent mean densitometry ratios normalized to corresponding control proteins (total JNK or β-actin). (E) Apoptosis by TUNEL staining (magnification, 80×), with quantification of apoptosis. **P < .01 compared with E by analysis of variance–Bonferroni analysis. (F) Liver injury: serum ALT and AST. Means ± SEM, n = 6 mice. **P < .01 and ***P < .001 by analysis of variance–Bonferroni analysis.

Figure 8

Acrolein scavengers show protective effects both in vitro and in vivo. (A) Cell survival in H4IIEC cells by 3, (4, 5-dimethylthiazol-2-yl) 2, 5-diphenyltetrazolium bromide assay (24 h). Means ± SEM, n = 3 experiments. *P < .05 compared with the corresponding treatments of alcohol (E) or acrolein (A) alone, by analysis of variance–Bonferroni analysis. C, control; E, 200 mmol/L alcohol; A, 30 μmol/L acrolein; HYD, 50 μmol/L hydralazine; CAR, 100 μmol/L carnosine. (B–F) Data are from livers of mice. E, alcohol; Hyd+E, hydralazine+alcohol. (B) Hepatic acrolein adduct accumulation in mice (magnification, 20×), hepatic steatosis by H&E (magnification, 80×), and Oil Red O staining (magnification, 20×). (C) ATF3, ATF4, GRP78, GRP94, and CHOP mRNA levels. Data are presented as means ± SEM (n = 6 mice). *P < .05 and ***P < .001 compared with alcohol (E) by analysis of variance–Bonferroni analysis. (D) Protein levels of phospho-JNK and total JNK, pro- and cleaved/active caspase 12, and CHOP. Numbers represent mean densitometry ratios normalized to corresponding control proteins (total JNK or β-actin). (E) Apoptosis by TUNEL staining (magnification, 80×), with quantification of apoptosis. **P < .01 compared with E by analysis of variance–Bonferroni analysis. (F) Liver injury: serum ALT and AST. Means ± SEM, n = 6 mice. **P < .01 and ***P < .001 by analysis of variance–Bonferroni analysis.

Figure 9

Acrolein mediates alcohol-induced hepatic ER stress, apoptosis, and injury in ALD, and the scavenger, hydralazine, prevents these effects.