ACBP/DBI neutralization for the experimental treatment of fatty liver disease

Abstract

Acyl-CoA binding protein (ACBP), also known as diazepam-binding inhibitor (DBI), is an extracellular checkpoint of autophagy. Here, we report that patients with histologically confirmed metabolic-associated steatohepatitis (MASH) or liver fibrosis exhibit elevated levels of circulating ACBP/DBI protein as compared to non-affected controls. Plasma ACBP/DBI strongly correlated with the NAFLD and FIB4 scores in patients, and these correlations were independent of age and body mass index. We studied the capacity of a monoclonal antibody (mAb) neutralizing mouse ACBP/DBI to combat active liver disease in several mouse models, in which steatohepatitis had been induced by four different protocols, namely, (i) methionine/choline-deficient diet, (ii) Western style diet (WD) alone, (iii) WD combined with the hepatotoxic agent CCl₄, and (iv) a combination of CCl₄ injections and oral ethanol challenge. Injections of anti-ACBP/DBI mAb attenuated histological, enzymological, metabolomic and transcriptomic signs of liver damage in these four models, hence halting or reducing the progression of non-alcoholic and alcoholic liver disease. Steatosis, inflammation, ballooning and fibrosis responded to ACBP/DBI inhibition at the preclinical level. Altogether, these findings support a causal role of ACBP/DBI in MASH and liver fibrosis, as well as the possibility to therapeutically target ACBP/DBI.

Fig. 1. High ACBP/DBI levels in plasma from MASH patients.

ACBP/DBI overabundance in patient diagnosed histologically with MASH (A) and fibrosis (B) shown as violin plots. Statistical comparisons were performed by means of the two-tailed Student’s t-test. Positive Spearman correlations of ACBP/DBI level with the backscatter scores for NAFLD (BS_SC NAFLD) (C) and for liver fibrosis (BS_SC FIB4) (D). ACBP/DBI correlated with several clinical parameters such as body mass index (BMI), age, C reactive protein (CRP), AST (aspartate transaminase), bilirubin, international normalized ratio (INR), albumin, alkaline phosphatase, NAFLD score, and FIB4 score (E).

Fig. 2. ACBP/DBI neutralization in vivo protects against MCD-induced hepatic steatosis, inflammation, and ballooning in curative setting.

A Experimental strategy of the MASH injury induced by MCD for 4 weeks (control with regular chow diet [RCD]) in C57BL/6 mice. The diet was maintained for 2 weeks and α-DBI or IgG (5 μg/g B.W.) was administrated 1 d before week 4 and biweekly during the diet. B–H α-DBI attenuates MASH produced by MCD diet through autophagy activation. Hepatic HES images (B), NAFLD activity score (C), ALT (D) and AST (E) activity in plasma from mice treated with α-DBI or IgG. Representative Western blots (F) and densitometric analysis of p62 (G) normalized against GAPDH and the ratio LC3B II/I (H) (both expressed as relative expression, RE) from liver extracts (n = 5–10 mice per group). Inflammation foci, macrosteatosis, and microsteatosis vesicular were indicated as asterisk, arrows and arrowheads, respectively. Data are represented as means ± SEM. Statistical analyses (p values) were calculated by ANOVA test.

Fig. 3. α-DBI therapeutic administration promotes a differential hepatic gene expression profile from whole transcriptome sequencing.

Heat map of the hierarchical clustering of genes with fold change ≥1.0 or ≤-1.0 and p < 0.05) using Z score for normalized value from mice fed with RCD or MCD and treated with IgG or α-DBI (A). Expression level represented as Volcano plot between mice treated with α-DBI versus IgG and fed with MCD (B) (n = 10 mice in each group). GSEA-based KEGG pathway enrichment analysis from up-regulated genes (C) and downregulated genes (D) after α-DBI administration therapeutically. The top most significant pathways are shown in the figures. E Heat map of the hierarchical clustering from liver (genes manually curated involved in fibrosis and immune pathways with fold change ≥ ±1.0 and with p < 0.05) using Z scores for normalized value from mice fed with RCD or MCD and treated with IgG or α-DBI. Statistical analyses (p value) were calculated using Dunn’s test with Bonferroni correction.

Fig. 4. Therapeutic neutralization of ACBP/DBI attenuates MASH induced by Western diet.

Schematic diagram of hepatic MASH induced by Western-style diet (WD) in C57BL/6 mice for 16 weeks. The diet was kept for 4 weeks and several doses of IgG or α-DBI (5 µg/g B.W.) was administrated from week 16 to 20 (A). B–J α-DBI reduces WD derived-MASH and fibrosis phenotype. Representatives HES images from liver (B), NAFLD activity score (C), ALT (D) and AST activity (E) in plasma, representative pictures of liver fibrosis (F), fibrosis score (G), and Immunoblots (H) and densitometric analysis of p62 (I) normalized against GAPDH and the ratio LC3B II/I (J) (both expressed as RE) from liver extracts. (n = 5–18 mice/group). Inflammation foci, macrosteatosis, and microsteatosis vesicular were indicated as asterisk, arrows, and arrowheads, respectively. Data are displayed as means ± SEM. Statistical analyses (p values) were calculated by ANOVA test.

Fig. 5. The phenotypic derived from Western diets with CCl4 is ameliorated by ACBP/DBI neutralization in vivo treatment.

Schematic diagram of hepatic MASH induced by WD plus CCl₄ in C57BL/6 mice for 10 weeks. The diet plus CCl₄ injection was maintained for 4 weeks and several doses of IgG or α-DBI (5 µg/g) was administrated from week 10 to 14 (A). B–J α-DBI diminished the steatosis, inflammation, ballooning and fibrosis in vivo. Representative pictures of hepatic HES (B), NAFLD activity score (C), ALT (D) and AST (E) activity in plasma, representative images of liver fibrosis by Sirius Red (F), fibrosis score (G), and of Western Blotting (H) and densitometry analysis of Collagen 1A1 (I) and α-SMA (J) normalized against GAPDH (both expressed as relative expression, RE) from liver extracts. (n = 5–10 mice/group). Inflammation foci, ballooning and steatosis vesicular are indicated as asterisk, arrows, and arrowheads, respectively. The results are shown as means ± SEM. Statistical analyses (p values) were calculated by ANOVA test.

Fig. 6. α-ACBP/DBI treatment diminishes the liver damage and fibrosis produced by Ethanol combined with CCl₄.

Schematic diagram of hepatic injury induced by Ethanol diet plus CCl₄ in C57BL/6 mice for 4 weeks. The ethanol diet plus CCl₄ administration was maintained for 4 more weeks and repeated doses of IgG or α-DBI (5 µg/g B.W.) were administrated from weeks 4 to 8 (A). B–J ACBP/DBI neutralization reduces liver damage and fibrosis in vivo. Representatives HES images from liver (B), hepatic damage score (C), plasmatic ALT (D) and AST activity (E), representative hepatic fibrosis images by Sirius Red (F), and fibrosis score (G). Inflammation foci, necrotic cell and steatosis are indicated as asterisk, arrows, and arrowheads, respectively. The results are represented as means ± SEM. Statistical analyses (p values) were calculated by ANOVA test.