Sex-based multiomics analysis uncovers metabolic and molecular mediators linking MASH and atherosclerosis

Abstract

Background & aims: Atherosclerotic cardiovascular disease (ASCVD) is the leading cause of death in patients with metabolic dysfunction-associated steatohepatitis (MASH). No therapy targets both diseases simultaneously, and a roadblock for discovering new treatments is the lack of animal models that recapitulate both diseases, especially in females.

Methods: Male and female Ldlr ^-/- mice (n = 8-13) were fed a western diet (WD), modified choline-deficient high-fat diet (mCDHFD), or modified MASH-inducing diet (mMASHD) containing equivalent physiological levels of cholesterol. Comprehensive multiomics including metabolomics, lipidomics, and transcriptomics, alongside histopathological and biochemical analyses, were integrated to characterize concurrent MASH and atherosclerosis. Transcriptomics was validated in other mouse models and integrated with human data (n = 79).

Results: While mCDHFD induced MASH-fibrosis in both sexes, WD was effective only in males, whereas mMASHD primarily affected females. mCDHFD induced concurrent MASH and atherosclerosis in both sexes, while WD effectively recapitulated disease co-occurrence only in males. Correlation analyses highlighted links between MASH and atherosclerosis, identifying circulating cholesterol and C-C motif chemokine ligand 2 (CCL2) as potential predictors of coexisting disease (p <0.04). Integrated metabolomic and transcriptomic analyses identified arginine-proline, glycine-serine, glutathione, and sphingolipid metabolism (p <0.03) as key dysregulated pathways, with sphinganine emerging as a predictor of disease severity. Hepatic itaconate and lactate levels were positively correlated with disease severity, whereas glycine, carnitine, 2-aminomuconic acid, and thiamine pyrophosphate were negatively associated (p <0.04). Lipidomic analyses revealed dysregulated polyunsaturated fatty acid, steryl ester, and dihexosylceramide metabolism. Integration of mouse and human transcriptomes revealed similarities in metabolic and proinflammatory/proatherogenic pathways.

Conclusion: This sex-based multiomics analysis establishes a murine model of concurrent MASH and atherosclerosis, reveals sex-specific dietary responses, and identifies metabolic and transcriptional pathways with potential utility as biomarkers and therapeutic targets.

Impact and implications: This study addresses the critical need for an animal model that replicates both metabolic dysfunction-associated steatohepatitis (MASH) and atherosclerotic cardiovascular disease, particularly in females, to facilitate therapeutic development. Using male and female Ldlr ^-/- mice, we found that different diets containing equivalent physiological levels of cholesterol induce sex-specific responses, with a modified choline-deficient high-fat diet effectively modeling both diseases in both sexes, while a western diet is effective only in males. Multiomics analyses identified key metabolic and inflammatory pathways linking MASH and atherosclerosis that mirror transcriptomic signatures found in humans, and highlight circulating cholesterol, CCL2, and sphinganine as potential biomarkers. These findings establish a translational model and reveal sex-specific metabolic pathways that will advance our understanding of the shared pathophysiology of MASH and atherosclerosis, and facilitate the development of dual therapeutic approaches, addressing an urgent unmet clinical need.

Keywords: Animal models; Atherosclerotic cardiovascular disease; MASH; MASLD; Metabolomics; Transcriptomics.

Graphical abstract

Fig. 1

Diet-induced MASH in atheroprone mice. Male and female (n = 10-12/group) Ldlr^-/- mice were fed a SD, WD, mCDHFD, or mMASHD for 24 weeks. (A) Experimental design. (B, H) Gross and liver morphology in males and females. (C, I) Liver-to-body weight ratio. (D-E, J-K) Plasma ALT and AST. (F, L) H&E and PSR staining and F4/80 immunofluorescence of liver sections (scale bars, 200 μm). (G, M) NAS (steatosis, lobular inflammation, hepatocellular ballooning scores) and fibrosis scores. Data are mean ± SEM. One-way ANOVA with Tukey’s post hoc test or Kruskal-Wallis with Dunn’s post hoc test. ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001; ∗∗∗∗p <0.0001 (SD vs. WD). ^#p <0.05; ^##p <0.01; ^###p <0.001; ^####p <0.0001 (SD vs. mCDHFD). ^@p <0.05; ^@@p <0.01; ^@@@p <0.001; ^@@@@p <0.0001 (SD vs. mMASHD). ^&p <0.05; ^&&p <0.01; ^&&&p <0.001; ^&&&&p <0.0001 (mMASHD vs. WD). ^$p <0.05; ^$$p <0.01; ^$$$p <0.001; ^$$$$p <0.0001 (mMASHD vs. mCDHFD). ^%p <0.05; ^%%p <0.01; ^%%%p <0.001; ^%%%%p <0.0001 (WD vs. mCDHFD). ALT, alanine aminotransferase; AST, aspartate aminotransferase; MASH, metabolic dysfunction–associated steatohepatitis; mCDHFD, modified choline-deficient high-fat diet; mMASHD, modified metabolic dysfunction–associated steatohepatitis diet; NAS, nonalcoholic fatty liver disease activity score; PSR, picrosirius red; SD, standard diet; WD, western diet.

Fig. 2

Effects of MASH-inducing diets on atherosclerosis in atheroprone mice. (A-B, I,J) En face analysis of lesions along the aortic tree in male and female Ldlr^-/- mice (n = 10-13/group). (C, K) H&E, ORO, Van Gieson, PSR staining, and CD68 immunofluorescence in aortic sinuses (scale bars, 200 μm). (D, L) Plaque area. (E, M) ORO⁺ area. (F, N) Acellular necrotic area by Van Gieson staining. (G, O) Fibrous cap thickness normalized to plaque area by PSR staining. (H, P) CD68⁺ area (fold-change from SD). Data are mean±SEM. One-way ANOVA with Tukey’s post hoc test or Kruskal-Wallis with Dunn’s post hoc test. p <0.05 was considered statistically significant. MASH, metabolic dysfunction-associated steatohepatitis; mCDHFD, modified choline-deficient high-fat diet; mMASHD, modified metabolic dysfunction–associated steatohepatitis diet; ORO, Oil red O; PSR, picrosirius red; SD, standard diet; WD, western diet.

Fig. 3

Circulating factors linking MASH and atherosclerosis. (A, I) Fasting blood glucose. (B, J) Plasma total cholesterol, and (C, K) triglycerides. (D, L) Cholesterol and (E, M) triglyceride contents in lipoproteins via FPLC. (F, N) Plasma CCL2 and (G, O) CCL5. (H, P) Correlations between disease indices in liver, plasma, and atherosclerotic plaques. Data are mean ± SEM (n = 10-12/group). One-way ANOVA with Tukey’s post hoc test or Kruskal-Wallis with Dunn’s post hoc test. Pearson and Spearman correlations were used to assess linear and non-linear relationships, respectively. p <0.05 was considered statistically significant. MASH, metabolic dysfunction-associated steatohepatitis; mCDHFD, modified choline-deficient high-fat diet; mMASHD, modified metabolic dysfunction–associated steatohepatitis diet; SD, standard diet; WD, western diet.

Fig. 4

Distinct metabolic signatures in concurrent MASH and atherosclerosis. Polar metabolomics of livers and plasma from male and female Ldlr^-/- mice (n = 4-5/group). PCA of male and female liver (A, G) and plasma (B, H) metabolomes. KEGG-based pathway analysis of metabolic pathways in the liver (C, I) and plasma (D, J). Heatmaps of the top differentially abundant metabolites in liver (E, K) and plasma (F, L). Significance thresholds p <0.05 and fold-change >1.5. PCA-based non-parametric permutational multivariate analysis of variance (PERMANOVA). Pathway enrichment significance assessed by a hypergeometric test. MASH, metabolic dysfunction-associated steatohepatitis; mCDHFD, modified choline-deficient high-fat diet; mMASHD, modified metabolic dysfunction–associated steatohepatitis diet; PCA, principal component analysis; SD, standard diet; WD, western diet.

Fig. 5

Correlations between the hepatic metabolome and MASH and atherosclerosis indices. Pearson and Spearman correlations were computed to assess relationships between the top 50 dysregulated liver metabolites and MASH, plasma, and atherosclerosis disease markers in (A) male and (B) female Ldlr^-/- mice (n = 4-5/group). ALT, alanine aminotransferase; AST, aspartate aminotransferase; MASH, metabolic dysfunction–associated steatohepatitis; NAS, nonalcoholic fatty liver disease activity score; PSR, picrosirius red; TC, total cholesterol; TG, triglyceride.

Fig. 6

Lipidomic signatures in concurrent MASH and atherosclerosis. Untargeted lipidomics of livers from male and female Ldlr^-/- mice (n = 4-5 per group). PCA of liver lipidomes in males (A) and females (F). KEGG-based pathway analysis in males (B) and females (G). Heatmaps of altered triglyceride (C, H), cholesteryl ester (D, I), and sphingomyelin (E, J) species. Significance thresholds p <0.05 and fold-change >1.5. PERMANOVA. Pathway enrichment significance assessed by a hypergeometric test. MASH, metabolic dysfunction-associated steatohepatitis; mCDHFD, modified choline-deficient high-fat diet; mMASHD, modified metabolic dysfunction–associated steatohepatitis diet; PCA, principal component analysis; SD, standard diet; WD, western diet.

Fig. 7

Distinct transcriptomic signatures in concurrent MASH and atherosclerosis. RNA-sequencing of liver samples from male and female Ldlr^-/- mice (n = 4-5 per group). PCA of liver transcriptomes in males (A) and females (H). Volcano plots of DEGs in males (B, D, F) and females (I, K, M) comparing WD, mCDHFD, or mMASHD to SD. Pathway analysis of upregulated (red) and downregulated (blue) pathways in males (C, E, G) and females (J, L, N). Two-sided Wald test for significant DEGs. Significance of enrichment determined by right-tailed Fisher’s exact test and Benjamini-Hochberg multiple testing adjustment. DEGs, differentially expressed genes; mCDHFD, modified choline-deficient high-fat diet; mMASHD, modified metabolic dysfunction–associated steatohepatitis diet; PCA, principal component analysis; SD, standard diet; WD, western diet.

Fig. 8

Integrated multiomics analysis highlights common pathways linking MASH and atherosclerosis in mice and humans. Network analysis using MetaboAnalyst 6.0 for integration of metabolites and DEGs using KEGG metabolic networks in male (A) and female (B) Ldlr^-/- mice fed SD, WD, mCDHFD, or mMASHD for 24 weeks (n = 4-5/group). KEGG pathways with p <0.05 were considered significantly enriched. Heatmap of KEGG pathways comparing RNA-sequencing of livers from male (C) and female (D) mice (n = 4 group) to RNA-sequencing of liver samples obtained from patients with MASH compared to controls. GSE130970 (n = 7 control and n = 17 MASH), GSE239422 (n = 12 control and n = 13 MASH), GSE126848 (n = 14 control and n = 16 MASH). Pathways enriched in the upregulated or downregulated DEGs are plotted in red or blue, respectively. Scale bars represents -log₁₀(p value) ∗sign(NES). Right-tailed Fisher’s exact test followed by Benjamini-Hochberg multiple testing adjustment. DEGs, differentially expressed genes; MASH, metabolic dysfunction–associated steatohepatitis; mCDHFD, modified choline-deficient high-fat diet; mMASHD, modified metabolic dysfunction–associated steatohepatitis diet; NES, normalized enrichment score; PCA, principal component analysis; SD, standard diet; WD, western diet.