Asparagine protects pericentral hepatocytes during acute liver injury

Abstract

The nonessential amino acid asparagine can only be synthesized de novo by the enzymatic activity of asparagine synthetase (ASNS). While ASNS and asparagine have been implicated in the response to numerous metabolic stressors in cultured cells, the in vivo relevance of this enzyme in stress-related pathways remains unexplored. Here, we found ASNS to be expressed in pericentral hepatocytes, a population of hepatic cells specialized in xenobiotic detoxification. ASNS expression was strongly enhanced in 2 models of acute liver injury: carbon tetrachloride (CCl4) and acetaminophen. We found that mice with hepatocyte-specific Asns deletion were more prone to pericentral liver damage than their control littermates after toxin exposure. This phenotype could be reverted by i.v. administration of asparagine. Unexpectedly, the stress-induced upregulation of ASNS involved an ATF4-independent, noncanonical pathway mediated by the nuclear receptor, liver receptor homolog 1 (LRH-1; NR5A2). Altogether, our data indicate that the induction of the asparagine-producing enzyme ASNS acts as an adaptive mechanism to constrain the necrotic wave that follows toxin administration and provide proof of concept that i.v. delivery of asparagine can dampen hepatotoxin-induced pericentral hepatocellular death.

Keywords: Amino acid metabolism; Cell stress; Hepatology; Metabolism.

Figure 1. ASNS is expressed in GLUL⁺ pericentral hepatocytes and induced upon CCl₄ treatment.

(A) Immunofluorescent staining for GLUL and DAPI (nuclei, blue) in liver of a C57BL/6J mouse. cv, central vein; pv, portal vein. Scale bar: 100 μm. (B) Workflow of GLUL⁺ and GLUL^– hepatocytes isolation. FACS, fluorescence-activated cell sorting; Ab, antibody. (C) Scheme of glutamine metabolism in liver. (D) mRNA expression and protein levels of glutamine-related transporters and enzymes in sorted GLUL⁺ and GLUL^– hepatocytes. n = 4 animals for each group. (E) mRNA and western blotting analyses of total liver lysates from C57BL/6J mice collected at indicated time points after CCl₄ treatment. n = 4 animals for each group. Veh, vehicle. (F) Immunofluorescent costaining for ASNS and GLUL with DAPI in livers treated with or without CCl₄ for 24 hours. Scale bar: 100 μm. Error bars denote SEM. Statistical analysis was performed by unpaired t test (D) and 1-way ANOVA followed by Bonferroni’s posthoc test (E). *P < 0.05; ***P < 0.001.

Figure 2. Loss of ASNS leads to enhanced CCl₄-induced acute liver injury.

(A and B) mRNA and Western blotting analyses of Asns^hep+/+ and Asns^hep–/– mice treated with corn oil (veh) or CCl₄ for 24 hours. n = 6 (Asns^hep+/+ veh); n = 5 (Asns^hep–/– veh); and n = 10 (all other groups). (C) ALT activity in serum from vehicle or CCl₄-treated mice in (A). (D–F) Representative images of H&E staining (D), TUNEL assay (E) and immunohistochemistry analysis of phospho-H2A.X (p-H2A.X) (F) in livers from (A). Damaged areas are outlined in white lines (D). Scale bar: 100 μm. cv, central vein. Quantification results are indicated on the right. Error bars denote SEM. Statistical analysis was performed by 2-way ANOVA followed by Bonferroni’s posthoc test (A, C–F). *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 3. ASNS protects against APAP-induced acute liver injury and its expression is induced by various liver stressors.

(A and B) H&E staining (A) and TUNEL assay (B) in livers of Asns^hep+/+ and Asns^hep–/– mice treated with PBS (veh) or APAP for 24 hours. n = 6 (Asns^hep+/+ veh); n = 5 (Asns^hep–/– veh); and n = 8 (all other groups). Scale bar: 100 μm. (C) Serum ALT activity in mice from (A). (D and E) mRNA and protein analyses of total cell lysates in livers from (A). (F) Heatmap showing logFC expression of genes in each microarray between treated versus untreated human liver slices (GSE54255, n = 5 for each group), or diseased versus healthy human livers (GSE38941, n = 17 for HBV and n = 10 for normal livers; and GSE28619, n = 15 for AH and n = 7 for normal livers). P values are adjusted by Benjamini & Hochberg. DCF, diclofenac; HBV, hepatitis B virus; AH, alcoholic hepatitis. Error bars denote SEM. Statistical analysis was performed by 2-way ANOVA followed by Bonferroni’s posthoc test (A–D). *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 4. Asns is a direct LRH-1 targeted pericentral gene.

(A) Western blotting analysis of total liver lysates from C57BL/6J mice collected at indicated time points after CCl₄ treatment. (B) Heatmap showing the expression levels of Asns and known LRH-1 targets in publicly available data sets (GSE59305 and GSE59304). (C) mRNA and protein analyses of total cell lysates from livers of Lrh-1^hep+/+ and Lrh-1^hep–/– mice, or Lrh-1^wt and Lrh-1^K289R mice. n = 4 (Lrh-1^hep+/+, Lrh-1^wt, and Lrh-1^K289R) and n = 5 (Lrh-1^hep–/–). (D) Representative images of immunofluorescent staining for ASNS and GLUL in livers from the indicated genetically modified mouse lines. Scale bar: 100 μm. (E) Transcription factor binding site analysis of mouse Asns promoter sequence showed 1 ATF4 binding site and 4 predicted LRH-1 binding sites. Numbers indicate distance from transcription start site (TSS). (F) Binding of LRH-1 to the 4 Asns promoter sites assessed by ChIP analysis using genomic DNA from livers of Lrh-1^hep+/+ and Lrh-1^hep–/– mice treated with or without CCl₄ for 24 hours. n = 4 animals for each group. Error bars denote SEM. Statistical analysis was performed by unpaired t test (C) and 2-way ANOVA followed by Bonferroni’s post-hoc test (F). *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 5. Loss of LRH-1 exacerbates acute liver injury triggered by CCl₄ or APAP.

(A) Western blotting analysis of total cell lysates or chromatin fractions from livers of Lrh-1^hep+/+ and Lrh-1^hep–/– mice treated or untreated with CCl₄ for 24 hours. Quantification of blotting analysis showing the relative levels of ASNS protein compared with the loading control TUBULIN. (B) mRNA expression levels of Asns in CCl₄-treated livers from (A). n = 6 (Lrh-1^hep+/+ veh, Lrh-1^hep–/– CCl₄); n = 5 (Lrh-1^hep–/– veh); and n = 7 (Lrh-1^hep+/+ CCl₄). (C) Serum ALT activity in mice from (B). (D) TUNEL assay in livers from (B). Scale bar: 100 μm. (E) mRNA analysis of livers from Lrh-1^hep+/+ and Lrh-1^hep–/– mice treated with or without APAP for 24 hours. n = 4 (Lrh-1^hep+/+ veh, Lrh-1^hep–/– veh); n = 6 (Lrh-1^hep+/+ APAP); and n = 5 (Lrh-1^hep–/– APAP). (F) ALT activity in serum and H&E staining in livers from mice in (E). Scale bar: 100 μm. Error bars denote SEM. Statistical analysis was performed by 2-way ANOVA followed by Bonferroni’s posthoc test (A–F). *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 6. Activation of LRH-1 protects the mice against CCl₄-induced acute liver injury.

(A) Serum ALT activity of Lrh-1^wt and Lrh-1^K289R mice treated 24 hours with or without CCl₄. n = 5 (Lrh-1^wt veh, Lrh-1^K289R veh); n = 6 (Lrh-1^wt CCl₄); and n = 7 (Lrh-1^K289R CCl₄). (B) Representative images and quantification results of TUNEL staining in livers from (A). (C and D) Western blotting and mRNA analyses of total liver lysates from (A). (E) Serum ALT activity of Shp^hep+/+ and Shp^hep–/– mice treated with or without CCl₄ for 24 hours. n = 5 (Shp^hep+/+ veh, Shp^hep–/– veh); n = 7 (Shp^hep+/+ CCl₄); and n = 6 (Shp^hep–/– CCl₄). (F) Representative images of TUNEL staining of livers from (E). Quantification results are indicated on the right. Scale bar: 100 μm (B and F). Error bars denote SEM. Statistical analysis was performed by 2-way ANOVA followed by Bonferroni’s posthoc test (A, B, and D–F). ***P < 0.001.

Figure 7. Asparagine treatment rescues Asns depletion-induced cell death and liver damage.

(A) Asparagine (Asn) over aspartate (Asp) ratio and glutamate (Glu) over glutamine (Gln) ratio in livers of untreated Asns^hep+/+ and Asns^hep–/– mice. n = 6 animals for each group. (B) Workflow of asparagine delivery upon CCl₄ treatment. Mice were i.p. injected with CCl₄ followed by 2 i.v. injections of 240 mg/kg asparagine (Asn) or PBS 1 hour and 8 hours later. (C) Serum ALT activity of Asns^hep+/+ and Asns^hep–/– mice subjected to the treatment described in (B). n = 6 (CCl₄ of Asns^hep+/+ and Asns^hep–/–) and n = 7 (CCl₄ + Asn of Asns^hep+/+ and Asns^hep–/–). (D) Representative images of TUNEL assay and immunohistochemistry analysis of p-H2A.X in livers from (C). Scale bar: 100 μm. Quantification results are indicated on the right. (E–F) Serum ALT activity and quantification results of TUNEL staining in livers from CCl₄-treated Asns^hep+/+ and Asns^hep–/– mice, followed by i.v. injection of glutamate (Glu) or valine (Val). n = 6 (Asns^hep+/+ CCl₄, CCl₄ + Glu and CCl₄ + Val); n = 7 (CCl₄ of Asns^hep–/–); and n = 5 (Asns^hep–/– CCl₄ + Glu and CCl₄ + Val). (G) Representative images and quantification results of TUNEL assay in livers from APAP-treated Asns^hep+/+ and Asns^hep–/– mice, followed by asparagine (Asn) i.v. injection. n = 6 animals for each group. Error bars denote SEM. Statistical analysis was performed by unpaired t test (A) and 2-way ANOVA followed by Bonferroni’s posthoc test (C–G). *P < 0.05; **P < 0.01.