Slow Metabolism-Driven Amplification of Hepatic PPARγ Agonism Mediates Benzbromarone-Induced Obesity-Specific Liver Injury

Abstract

Obesity and nonalcoholic fatty liver disease (NAFLD) are established risk factors for drug-induced liver injury (DILI). The previous study demonstrates that benzbromarone (BBR), a commonly prescribed pharmaceutical agent for managing gout and hyperuricemia, exacerbates hepatic steatosis and liver injury specifically in obese individuals. However, the precise mechanism underpinning this adverse effect remains incompletely elucidated. Given the significance of BBR and its analogs in anti-gout/hyperuricemia drug discovery, elucidating the mechanism by which BBR exacerbates obesity-specific DILI warrants further investigation. In this study, through a combined multi-omics, pharmacological, and pharmacokinetic approaches, it is found that BBR-induced obesity-specific DILI is primarily through the potentiation of peroxisome proliferator-activated receptor gamma (PPARγ) signaling pathways. Further in vivo and in vitro pharmacokinetic analyses reveal that obese db/db mice exhibited a diminished capacity to metabolize BBR in their livers. This reduction leads to prolonged retention of BBR, subsequently resulting in chronic and sustained hepatic PPARγ agonism. This study demonstrates that a slow metabolism-driven amplification of hepatic PPARγ agonism mediates BBR-induced obesity-specific hepatic steatosis and subsequent DILI, which also emphasizes the importance of the reduced hepatic drug metabolism capacity in patients with obesity or pre-existing NAFLD in both clinical practice and drug discovery processes.

Keywords: PPARγ; benzbromarone; drug metabolism; drug‐induced liver injury; obesity.

Figure 1

Benzbromarone (BBR) aggravates lipid accumulation in primary cultured hepatocytes by enhancing PPAR signaling pathways. A) Primary cultured mouse hepatocytes were incubated with different doses of BBR for 24 h, and the cell viability was measured by CCK8 assay to calculate the 50% inhibitory concentration (IC50). B) After primary hepatocytes were incubated with BBR/oleic acid (OA) as indicated for 16 h, intracellular triglyceride concentrations were measured, ***p < 0.001, n = 6. C) BODIPY staining in different groups; images are shown under 600× or 1800× magnification, scale bar = 100 µm. D) Flow scheme of RNA sequencing. E–H) Differentially expressed genes (DEGs), volcano plot, and KEGG enrichment in hepatocytes treated with BBR in the presence and absence of OA. I) Intersection and heatmap of DEGs in hepatocytes treated with BBR in the presence and absence of OA, against a predefined set of PPAR target genes. J–L) GSEA enrichment in BBR‐treated hepatocytes. M,N) Intersection and heatmap of DEGs in BBR‐treated hepatocytes compared to public transcriptomic datasets and PPAR target genes. O) BBR‐stimulated DEGs related to lipid metabolism in the PPAR signaling pathways.

Figure 2

Benzbromarone exacerbates hepatic steatosis in db/db mice by amplification of PPAR signaling pathways. A) Flow scheme of RNA sequencing of livers from db/db mice. B) Differentially expressed genes (DEGs) and volcano plots for livers from BBR‐treated db/db mice compared to the vehicle control group. C) KEGG enrichment depicting upregulated DEGs. D–F) GSEA enrichment analysis and corresponding heatmaps depicting upregulated DEGs. G–I) Venn diagrams and heatmaps depicting upregulated DEGs in the livers of BBR‐treated db/db mice compared to BBR‐treated primary mouse hepatocytes cultured in vitro, livers from Wy14643‐treated mice, and livers from rosiglitazone‐treated mice.

Figure 3

Lipidomic analysis indicates enhanced lipid synthesis in benzbromarone‐treated hepatocytes. A) Flow scheme of lipidomic analysis. B) Principal component analysis (PCA) between OA‐treated hepatocytes and those cotreated with OA and BBR. C) Volcano plot representing the levels of lipid species shows significant changes between the two groups. D) Levels of individual species of triglycerides (TAG), E) diacylglycerol (DAG), F) free fatty acids (FFA), and fatty acyl‐CoAs (FFA‐CoA), *p < 0.05, **p < 0.01, ***p < 0.001, n = 5. G) Network diagram illustrating the alterations in lipid metabolism pathways within BBR‐treated hepatocytes, with genes exhibiting upregulation highlighted in red font. TCA cycle, tricarboxylic acid cycle.

Figure 4

Inhibition of PPARγ mitigates benzbromarone‐induced hepatic lipid accumulation in vitro. A) Molecular docking of BBR with PPARα, PPARβ/δ, and PPARγ using Glide. B,C) Intracellular triglyceride levels in primary hepatocytes treated with BBR and different PPAR modulators. D–F) Intracellular triglyceride levels and BODIPY staining in hepatocytes treated with BBR and GW9662. G) MST experiment measuring the affinity between rosiglitazone and PPARγ in the presence/absence of BBR. H) Comparison of the PPARγ binding sites of BBR and rosiglitazone. The green residues are amino acids that show the same hydrophobic interactions with BBR and rosiglitazone. I) The mRNA expression of PPARγ target genes in primary hepatocytes after different treatments. *p < 0.05, **p < 0.01, ***p < 0.001, n = 6.

Figure 5

Hepatic knockdown of PPARγ alleviates benzbromarone‐induced liver injury in db/db mice. A) Flow scheme of BBR administration in db/db mice. B) Body weight gain, ***p < 0.001 compared to the vehicle‐treated AAV8‐Vec group; ^# p < 0.05 compared to the BBR‐treated AAV8‐Vec group, n = 10. C) Daily food intake (g·day⁻¹ per mouse). D) Representative images of the liver and other abdominal organs. E) Liver weight normalized by body weight in each group. The serum levels of F) alanine transaminase (ALT), G) aspartate transaminase (AST), H) creatinine, I) triglyceride, J) total cholesterol (TC), K) non‐esterified fatty acid (NEFA), L) very low‐density lipoprotein (VLDL), M) low‐density lipoprotein (LDL), and N) high‐density lipoprotein (HDL). *p < 0.05, **p < 0.01, ***p < 0.001 as indicated, n = 10.

Figure 6

Knockdown of PPARγ restores benzbromarone‐induced aggravation of hepatic steatosis in db/db mice. A) The expression and tissue localization of PPARγ and green fluorescent protein (GFP) were determined in situ by immunohistochemistry, scale bar = 100 µm. B) The expression of PPARγ in livers from each group was detected by western blotting, using GAPDH as an internal control; the blots were repeated 3 times. C) Representative images of hematoxylin‐eosin (HE), Oil Red O, and Masson staining in each group, scale bar = 100 µm. D,E) The levels of triglycerides and total cholesterol (TC) in liver tissue were measured, ***p < 0.001 as indicated, n = 10. F) The Venn diagram and heatmap of DEGs in each group. G) The qPCR assay was utilized to assess the liver tissue expression levels of several representative DEGs, including Pparg, Cidea, Cidec, Plin2, and Hilpda. *p < 0.05, **p < 0.01, ***p < 0.001, n = 5.

Figure 7

The PPARγ‐selective antagonist GW9662 alleviates benzbromarone‐induced aggravation of hepatic steatosis and liver injury in db/db mice. A) Flow scheme of GW9662/BBR co‐treatment in db/db mice. B) Body weight gain, C) daily food intake in different groups. D) Representative images of the livers and other abdominal organs are shown. E) Liver weights and F) liver weights normalized by body weight in each group. G) Serum ALT, H) AST, I) creatinine, J) triglyceride, K) total cholesterol, and L) NEFA levels. *p < 0.05, ***p < 0.001, n = 10. M) Representative image of HE and Oil Red O staining in each group, scale bar = 100 µm.

Figure 8

Slow metabolism of benzbromarone enhances its hepatic steatogenic effect by amplification of PPARγ agonism in obese mice. A) Plots of the plasma pharmacokinetic profiles of BBR in different groups, n = 6 per time point. B) Comparison of the pharmacokinetics parameters, area under the curve (AUC), half‐life (T_1/2), mean residence time (MRT), and clearance (CL) for BBR, n = 6. C) Heatmap depicting the expression of CYP450 enzymes in the DEGs derived from RNA sequencing data of the liver tissue from healthy humans (n = 24), patients with NAFLD (n = 20) or NASH (n = 19), sourced from the publicly available dataset GSE89632. D) Kinetic profiles of CYP450s substrate metabolism after incubation using pooled liver S9 fractions from db/db mice and their lean littermates, n = 5. E) Kinetic profile of BBR after incubation with liver S9 fractions. F) Proposed metabolic pathway of BBR. G) Quantification of intracellular 6‐OH‐BBR concentrations in primary hepatocytes after incubation with BBR for the different times in the presence/absence of OA, n = 6. H) Plots of the plasma pharmacokinetic profiles of 6‐OH‐BBR in different groups, n = 6 per time point. I) Kinetic profiles of 6‐OH‐BBR production in db/db mice and their lean littermates, *p < 0.05, **p < 0.01, ***p < 0.001, n = 6. J) Heatmap of the expression of CYP450 enzymes in liver DEGs from BBR‐treated db/db mice compared to vehicle groups. K) In silico docking of BBR, 6‐hydroxy‐BBR (6‐OH‐BBR), 5,6‐dihydroxy‐BBR (5,6‐2‐OH‐BBR), and 6,7‐dihydroxy‐BBR (6,7‐2‐OH‐BBR) with PPARγ.