Reversal of murine alcoholic steatohepatitis by pepducin-based functional blockade of interleukin-8 receptors

Abstract

Objective: Alcoholic steatohepatitis is a life-threatening condition with short-term mortality up to 40%. It features hepatic neutrophil infiltration and blood neutrophilia, and may evolve from ethanol-induced breakdown of the enteric barrier and consequent bacteraemia. Signalling through CXCR1/2 G-protein-coupled-receptors (GPCRs), the interleukin (IL)-8 receptors, is critical for the recruitment and activation of neutrophils. We have developed short lipopeptides (pepducins), which inhibit post-ligand GPCR activation precisely targeting individual GPCRs.

Design: Experimental alcoholic liver disease was induced by administering alcohol and a Lieber-DeCarli high-fat diet. CXCR1/2 GPCRs were blocked via pepducins either from onset of the experiment or after disease was fully established. Hepatic inflammatory infiltration, hepatocyte lipid accumulation and overall survival were assessed as primary outcome parameters. Neutrophil activation was assessed by myeloperoxidase activity and liver cell damage by aspartate aminotransferase and alanine aminotransferase plasma levels. Chemotaxis assays were performed to identify chemoattractant signals derived from alcohol-exposed hepatocytes.

Results: Here, we show that experimental alcoholic liver disease is driven by CXCR1/2-dependent activation of neutrophils. CXCR1/2-specific pepducins not only protected mice from liver inflammation, weight loss and mortality associated with experimental alcoholic liver disease, but therapeutic administration cured disease and prevented further mortality in fully established disease. Hepatic neutrophil infiltration and triglyceride accumulation was abrogated by CXCR1/2 blockade. Moreover, CXCL-1 plasma levels were decreased with the pepducin therapy as was the transcription of hepatic IL-1β mRNA.

Conclusions: We propose that high circulating IL-8 in human alcoholic hepatitis may cause pathogenic overzealous neutrophil activation, and therapeutic blockade via pepducins merits clinical study.

Keywords: ALCOHOLIC LIVER DISEASE; IMMUNE-MEDIATED LIVER DAMAGE; INTERLEUKIN 8; LEUKOCYTES.

Figure 1

x1/2pal-i1 treatment inhibited the development and progression of alcoholic liver disease. (A) Experimental approach. Mice received the Lieber–DeCarli (LDC) diet. Pepducin therapy was commenced either with the introduction of ethanol (day 5; ‘preventative setting’; 2.5 mg/kg x1/2pal-i1 subcutaneous every other day) or after mice had established disease (day 29, ‘therapeutic setting’; 5.0 mg/kg x1/2pal-i1 subcutaneous once a day). On day 35, mice received LPS (2.5 mg/kg intraperitoneal) and were assessed 24 h later. (B) x1/2pal-i1 prevents development of liver steatosis. Representative liver sections stained with H&E (n=15). (C) Therapeutic x1/2pal-i1 reverts liver steatosis. Representative liver sections stained with H&E (n=15). (D) Histological disease activity. H&E-stained sections for hepatocyte ballooning, leucocyte infiltration and Mallory–Denk bodies, Oil-red-O for steatosis. Statistical analysis: Mann–Whitney U after Kruskal–Wallis; n=15 per group; *p<0.05 (LDC vs LDC-EtOH); †p<0.05 (LDC-EtOH vs LDC-EtOH-x1/2pal-i1). (E) x1/2pal-i1 treatment prevents from alcohol-induced mortality. Mice received 2.5 mg/kg of x1/2pal-i1, scram-i1 (2.5 mg/kg) or vehicle control every other day. Statistical analysis: Mantel–Cox test p<0.0001. n=15. (F) x1/2pal-i1 therapy reduces alcohol-induced mortality and prevents from weight loss. Mice received 5 mg/kg of x1/2pal-i1, scram-i1 or vehicle control every day from day 29 until the conclusion of the experiment. The animals’ weight was taken every other day and weight curves were compared. Statistical analysis: Mantel–Cox test p<0.0001. n=15 (n=7 for scram-i1).

Figure 2

x1/2pal-i1 treatment normalises liver–body weight ratio and prevents from hepatic necrosis. (A) Prophylactic and therapeutic x1/2pal-i1 administration result in a normal liver–body weight ration. Statistical analysis: Mann–Whitney U after Kruskal–Wallis. n=15 per group (n=7 for scram-i1 experiment). (B and C) Prophylactic and therapeutic x1/2pal-i1 treatment prevents from liver cell necrosis. Aspartate aminotransferase (AST) and alanine aminotransferase (ALT) levels as markers of liver cell necrosis. Statistical analysis: Mann–Whitney U after Kruskal–Wallis. n=15 per group (n=7 for scram-i1 experiment).

Figure 3

CXCR2 blockade reduces pro-inflammatory cytokine expression. Mice on Lieber–DeCarli diet were injected with x1/2pal-i1 in a prophylactic or therapeutic mode. (A) x1/2pal-i1 pepducin treatment decreases serum CXCL1. CXCL1 serum levels were measured by ELISA. (B) CXCR1/2 blockade by the x1/2pal-i1 pepducin decreases hepatic CXCL1 mRNA transcription. mRNA levels were measured by SYBR-green real-time PCR. (C) CXCR1/2 blockade by the x1/2pal-i1 pepducin decreases hepatic interleukin (IL)-1β mRNA transcription. mRNA levels were measured by SYBR-green real-time PCR (D) CXCR1/2 blockade by the x1/2pal-i1 pepducin decreases hepatic tumour necrosis factor (TNF)α mRNA transcription. mRNA levels were measured by SYBR-green real-time PCR. Statistical analysis: Mann–Whitney U after Kruskal–Wallis. n=6 (E) CXCR1/2 blockade reduces pro-caspase-1 expression in livers. Liver tissue lysates were immunoblotted for caspase-1 protein expression. Representative western blot analysis, n=3, each lane represents an individual liver specimen.

Figure 4

Pepducin treatment inhibits neutrophil infiltration and activation. (A and B) x1/2pal-i1 prevents accumulation of neutrophils in liver tissue sections of alcoholic mice. To compare the numbers of neutrophils, histological sections were stained with naphthol AS-D chloroacetate. Neutrophil numbers were counted in four representative microscopic fields from 10 individual mice per group. (C) x1/2pal-i1 treatment reduces myeloperoxidase (MPO) levels in liver tissue. Homogenised liver tissue was assessed for MPO levels. Right panel: preventative treatment; left panel: therapeutic treatment. Statistical analysis: Mann–Whitney U after Kruskal–Wallis. n=10. (D) Ethanol induces CXCL1 secretion in Hep3B and HepG2 cells. CXCL1 levels were measured after 24 h of ethanol exposure to the cells. Mann–Whitney U after Kruskal–Wallis, n=3 (E) Ethanol-induced neutrophil chemotaxis is CXCL8 dependent. CXCL8 (IL-8) was neutralised by adding antibodies directed against CXCL8. CXCR1 and CXCR2 receptor signalling was inhibited by x1/2pal-i1. The nonsense pepducin scram-i1 had no effect on neutrophil chemotaxis induced by HepG2 cell supernatants. Chemotaxis experiments were performed in modified Boyden chambers. Statistical analysis: Mann–Whitney U after Kruskal–Wallis, n=3. LDC, Lieber–DeCarli diet.

Figure 5

(A) CXCR1/2 pepducin blockade reverses steatosis in alcoholic mice. Oil-red-O stain of liver sections of mice treated therapeutically. N=15. (B) Blockade of CXCR2 receptors decreases the triglyceride content in livers. Triglycerides from liver tissue of mice treated therapeutically were measured by a photometric assay. N=10. Statistical analysis: Mann–Whitney U test after Kruskal–Wallis. (C–E) The x1/2pal-i1 pepducin has no effect in mice on regular chow diet. Mice received the animal facility's standard chow diet. The animals were injected with either 10% of dimethyl sulfoxide or the x1/2pal-i1 pepducin (5 mg/kg) subcutaneous every other day. After 5 weeks of treatment, the animals were sacrificed and evaluated for (C) liver–body weight ratio, (D) alanine aminotransferase (ALT) and (E) aspartate aminotransferase (AST). N=10. Statistical analysis: Mann–Whitney U after Kruskal–Wallis; n.s.: p>0.05. LDC, Lieber–DeCarli diet.

Figure 6

Proposed mechanism of x1/2pal-i1 therapy. IL, interleukin; MMP, matrix metalloproteinases; MPO, myeloperoxidase; ROS, reactive oxygen species; TNF, tumour necrosis factor.