Bempedoic acid suppresses diet-induced hepatic steatosis independently of ATP-citrate lyase

Abstract

ATP citrate lyase (ACLY) synthesizes acetyl-CoA for de novo lipogenesis (DNL), which is elevated in metabolic dysfunction-associated steatotic liver disease. Hepatic ACLY is inhibited by the LDL-cholesterol-lowering drug bempedoic acid (BPA), which also improves steatosis in mice. While BPA potently suppresses hepatic DNL and increases fat catabolism, it is unclear if ACLY is its primary molecular target in reducing liver triglyceride. We show that on a Western diet, loss of hepatic ACLY alone or together with the acetyl-CoA synthetase ACSS2 unexpectedly exacerbates steatosis, linked to reduced PPARα target gene expression and fatty acid oxidation. Importantly, BPA treatment ameliorates Western diet-mediated triacylglyceride accumulation in both WT and liver ACLY knockout mice, indicating that its primary effects on hepatic steatosis are ACLY independent. Together, these data indicate that hepatic ACLY plays an unexpected role in restraining diet-dependent lipid accumulation and that BPA exerts substantial effects on hepatic lipid metabolism independently of ACLY.

Keywords: ACLY; ACSS2; PPARα; bempedoic acid; lipid metabolism; metabolic dysfunction-associated steatotic liver disease.

Figure 1.. Bempedoic acid suppresses fatty acid synthesis and promotes fatty acid oxidation.

(A-D) WT C57BL6 male mice were given 15% fructose : 15% glucose drinking water for 4 weeks and gavaged with 30 mg/kg bempedoic acid (BPA) or vehicle (veh) daily for 4 weeks. Mice received 6% deuterium labeled water for 36 hours prior to tissue collection. (A) Schematic of experimental design for panels B-D. Abundance of newly synthesized hepatic (B) palmitate (C) stearate and (D) oleate in veh or BPA treated mice. Molar rates of synthesis are reported for palmitate only, since a labeled palmitate standard was included, and relative rates are reported for panels C and D. (E-H) Relative abundance of (E) hepatic acetylcarnitine and (F) plasma beta-hydroxybutyrate in WT C57BL6 mice given 15% fructose: 15% glucose drinking water for 4 weeks and treated with 30 mg/kg BPA (n=5 per group) (G-H) ¹³C enrichment in (G) hepatic acetylcarnitine and (H) plasma beta-hydroxybutyrate in mice following ¹³C-oleate gavage at 480 mg/kg body weight. Natom equivalents of ¹³C was measured by multiplying mole percent enrichment with abundance. Statistical significance was calculated by two-sided t-tests. Each point represents a biological replicate and error bars represent SD. *P ≤ 0.05; **P ≤ 0.01. Panel A made using Biorender.

Figure 2.. Hepatic ACLY loss exacerbates steatosis on a Western Diet independent of ACSS2 upregulation

A-M) Acly f/f (WT) or Alb-Cre; Acly f/f (Liver-specific ACLY KO, LAKO) 6-8 week old female mice were fed a Western diet (WD) containing either 0.2% or 1.25% cholesterol (chol) for 4 weeks. For 0.2% chol WD, WT n = 5, LAKO n = 6. For 1.25% chol WD, WT n = 5, LAKO n = 6. (A) Schematic of experimental design (B) Western blot in liver lysates after 4 weeks of 0.2% chol WD (C) Body weights over time (D) Tissue weights of mice after 4 weeks of 0.2% chol WD (E) Representative hematoxylin and eosin (H&E) and Oil Red O (ORO) histological stains of livers from mice fed 0.2% chol WD for 4 weeks (F-G) Western blot of ACLY and ACSS2 in liver lysates with tubulin loading control in (F) LAKO or (G) WT mice after 4 weeks of 0.2% or 1.25% chol WD. For the 0.2% chol WD samples, the same samples as in panel B were re-run for comparison to 1.25% chol WD samples. (H) Representative H&E and ORO histological stains of livers from WT or LAKO mice fed 1.25% chol WD for 4 weeks. (I-M) Thin layer chromatography (TLC) of extracted liver lipids from mice (n = 3 per group) and quantification (CE, cholesterol esters; TG, triglycerides; FFA, free fatty acids; Chol, free cholesterol; STD, standard). Statistical significance was calculated by two-sided t-tests. Each point represents a biological replicate (1 mouse) and error bars represent SD. *P ≤ 0.05; **P ≤ 0.01;***P ≤ 0.001. Panel A made using Biorender. See also Figure S1.

Figure 3.. Triacylglycerides, diacylglycerides, and cholesterol esters are elevated in Western diet-fed hepatic ACLY KO mice independent of ACSS2

(A-F) Male mice with deletion of hepatic ACLY (ACLY KO), ACLY and ACSS2 (double KO, DKO), and respective controls (wildtype, WT) were generated by tail vein injection of Acly f/f or Acly f/f; Acss2 f/f mice with AAV8-TBG-Cre or AAV8-TBG-GFP. Mice were fed 0.2% cholesterol containing WD starting 1 week after injection for 6 weeks. For Acly f/f groups, n = 7. For Acly f/f; Acss2 f/f groups, n = 6 (A) Schematic of experimental design; (B) Western blot of ACLY, ACSS2, and GFP in liver lysates with tubulin loading control; (C) Body and (D) liver weights collected after 6 weeks on diet; (E) Representative images of liver H&E and Oil Red O neutral lipid staining; (F) Total liver triglycerides measured by LC-MS; (G) Male Acly f/f mice were administered tail vein injection of AAV8-TBG-Cre or AAV8-TBG-GFP and were fed 0.2% chol containing WD for 12 weeks. Plasma triglyceride concentration was measured using a TAG enzyme assay kit following poloxamer-407 intraperitoneal injection. The 120 min time point was extrapolated since values fell outside of standard curve; (H-R) Female mice of indicated genotypes were generated as described above and were fed on 0.2% chol WD for 22 weeks; (H-J) Body, liver, and gonadal white adipose (GWAT) tissue weights; (K) Serum TAGs measured by enzyme assay kit at endpoint (L) Representative images of liver H&E; (M) Liver TAGs measured by enzyme assay kit (N-P) Liver lipid quantification by LC-MS of (N) Diacylglycerols (O) Cholesterol esters (P) Phosphatidylcholines (Q, R) H&E images (5 per genotype) were scored by a veterinarian pathologist on a scale of 0-4, with 4 being most severe, for (Q) macrovesicular steatosis and (R) microvesicular steatosis. Each point represents a biological replicate (1 mouse) and error bars represent SD. Statistical significance for (D), and (H-J) was calculated by two-sided t-tests. For (F), (K), (M-P) statistical significance was calculated using two-way ANOVA with Šidák’s correction for multiple comparisons. *p ≤ 0.05; **p ≤ 0.01;***p ≤ 0.001; ****p ≤ 0.0001. Panel A made using Biorender. See also Figure S2 and S3.

Figure 4.. ACLY deficiency suppresses fatty acid oxidation and ketone production genes on Western diet

(A-G) Male mice with deletion of hepatic ACLY and ACSS2 (double KO, DKO) and control (WT) were fed 0.2% chol WD for 6 weeks. For each WT and DKO, n = 6 (A) gProfiler functional enrichment analysis of genes downregulated in livers of DKO compared to WT mice, log2 fold change >1 and adjusted p value < 0.1 (B-E) Heatmaps of gene expression from DKO and WT livers for genes involved in select pathways based on curated lists in Rakhshandehroo et al. expressed as row Z-score. (F, G) mRNA expression of PPARα target genes in the livers of mice with deletion of hepatic ACLY (ACLY KO), ACLY and ACSS2 (double KO, DKO) and control (WT) mice fed (F) 0.2% chol WD for 6 weeks or (G) 15% glucose: 15% fructose drinking water for 2 weeks. For WD groups, n = 4-5. For sweetened drinking water groups, n=3-5. Tissues were collected at ZT8-10. Each point represents a biological replicate (1 mouse) and error bars represent SD. *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001. See also Figure S4.

Figure 5.. ACLY deficiency suppresses fatty acid oxidation and endogenous PPARα ligand abundance

(A, B, D-G, I) Mice of indicated genotypes were administered AAV8-TBG-Cre or AAV-TBG-GFP, and 1 week later fed 0.2% cholesterol containing WD (A) Female mice were fed WD for 22 weeks and liver acylcarnitine measurements by LC-MS were done at endpoint, collected at ZT8-10, n = 6-7; (B) Male mice were fed WD for 6 weeks; β-hydroxybutyrate in liver was measured using GC-MS, n = 6-7; (C) Male mice of indicated genotypes were administered AAV8-TBG-Cre or AAV-TBG-GFP, and 1 week later fed 15% glucose : 15% fructose drinking water for 2 weeks. Liver β-hydroxybutyrate was measured using GC-MS, n=4-5/ group; (D) Fatty acid oxidation (³H-palmitate) in liver lysates from female mice fed a WD and tissues collected at ZT8-10, n=10-11/genotype. Data was combined from 3 smaller cohorts, 1 on diet for 4 weeks, 1 for 10 weeks, and 1 for 23 weeks, and each data point was normalized to average control value within the cohort, which ranged from 8.5-10.5 nmol/min/g specific activity. (E) Female mice were fed WD for 22 weeks and hepatic PC 16:0/18:1 was measured via LC-MS at endpoint, with tissues collected at ZT8-10; (F) Mice were fed 0.2% cholesterol containing WD for 19 weeks, fasted for 5 hours and blood glucose was measured using a glucometer; (G) Glucose tolerance tests of female mice fed WD for 19 weeks. (H) Lactate/pyruvate tolerance test of female mice fed WD for 22 weeks and fasted for 5 hours, n=5/genotype. (I) Insulin tolerance tests of female mice fed WD for 22 weeks. Each point represents a biological replicate (1 mouse) and error bars represent SD. Statistical significance for panels F, G, and H were calculated by two-sided t-tests, for panel D by unpaired Welch’s test, and for panels A, B, C, and E using two-way ANOVA with Šidák’s correction for multiple comparisons. *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001.

Figure 6.. Bempedoic acid ameliorates diet-induced lipid accumulation independent of hepatic ACLY

(A,B) Male WT or LAKO mice were treated with 30 mg/kg bempedoic acid (BPA) or vehicle (veh) daily for 8 days. Mice were euthanized and tissue was collected 4 hours after the last dose on day 8. For all groups, n = 4-5. Bempedoyl-CoA abundance per mg liver (liv) or brown adipose (BAT) tissue; (B) Western blot of ACSS2 in liver lysates with ribosomal S6 loading control; (C-O) Mixed male and female WT or LAKO mice on 0.2% chol WD for 6 weeks were gavaged with 10 mg/kg BPA or veh daily for the latter 3 weeks. For all groups, n = 5-7 (C) Schematic of experimental design (D-E) Representative (D) H&E and (E) Oil Red O histological stains of livers from female WT and LAKO mice treated with veh or BPA (F) Liver to body weight ratio (G) Weight of extracted and dried liver lipids normalized to tissue weight (H-L) Total abundance of (H) hepatic triglycerides (TAGs) (I) hepatic diacylglycerol (DAGs) (J) hepatic phosphotidylcholines (PCs) and (K) hepatic phosphotidylethanolamines and (L) hepatic PC 16:0/18:1 measured by LC-MS; (M) PPARα target gene expression in WT mice (N) Hepatic acetylcarnitine abundance (O) Hepatic butyrylcarnitine abundance. (P) Model. Statistical significance was calculated by two-sided t-tests in (F), (G), and (M) and by two-way ANOVA with Šidák’s correction for multiple comparisons (H-L), and (N), (O). Each point represents a biological replicate (1 mouse) and error bars represent SD. *P ≤ 0.05; **P ≤ 0.01;***P ≤ 0.001; ****P ≤ 0.0001. Panels C and P made using Biorender. See also Figure S5 and S6.