Ceramide synthase 6 (CerS6) is upregulated in alcohol-associated liver disease and exhibits sex-based differences in the regulation of energy homeostasis and lipid droplet accumulation

Abstract

Objective: Alcohol-associated liver disease (ALD) is the leading cause of liver-related mortality worldwide. Current strategies to manage ALD focus largely on advanced stage disease, however, metabolic changes such as glucose intolerance are apparent at the earliest stage of alcoholic steatosis and increase the risk of disease progression. Ceramides impair insulin signaling and accumulate in ALD, and metabolic pathways involving ceramide synthase 6 (CerS6) are perturbed in ALD during hepatic steatosis. In this study, we aimed to investigate the role of CerS6 in ALD development and the relevance of CerS6 to human ALD.

Methods: C57BL/6 WT and CerS6 KO mice of both sexes were fed either a Lieber-DeCarli control (CON) or 15% ethanol (EtOH) diet for six weeks. In vivo metabolic tests including glucose and insulin tolerance tests (GTT and ITT) and energy expenditure were performed. The mice were euthanized, and serum and liver lipids and liver histology were examined. For in vitro studies, CerS6 was deleted in human hepatocytes, VL17A and cells were incubated with EtOH and/or C_16:0-ceramides. RNAseq analysis was performed in livers from mice and human patients with different stages of ALD and diseased controls.

Results: After six weeks on an EtOH diet, CerS6 KO mice had reduced body weight, food intake, and %fat mass compared to WT mice. Energy expenditure increased in both male and female KO mice, however, was only statistically significant in male mice. In response to EtOH, WT mice developed mild hepatic steatosis, while steatosis was ameliorated in KO mice as determined by H&E and ORO staining. KO mice showed significantly decreased long-chain ceramide species, especially C_16:0-ceramides, in the serum and liver tissues compared to WT mice. CerS6 deletion decreased serum TG and NEFA only in male not female mice. CerS6 deletion improved glucose tolerance and insulin resistance in EtOH-fed mice of both sexes. RNAseq analysis revealed that 74 genes are significantly upregulated and 66 genes are downregulated by CerS6 deletion in EtOH-fed male mice, with key network pathways including TG biosynthetic process, positive regulation of lipid localization, and fat cell differentiation. Similar to RNAseq results, absence of CerS6 significantly decreased mRNA expression of lipid droplet associated proteins in EtOH-fed mice. In vitro, EtOH stimulation significantly increased PLIN2 protein expression in VL17A cells while CerS6 deletion inhibited EtOH-mediated PLIN2 upregulation. C_16:0-ceramide treatment significantly increased PLIN2 protein expression compared to CON. Notably, progression of ALD in humans was associated with increased hepatic CerS6 expression.

Conclusions: Our findings demonstrate that CerS6 deletion improves glucose homeostasis in alcohol-fed mice and exhibits sex-based differences in the attenuation of EtOH-induced weight gain and hepatic steatosis. Additionally, we unveil that CerS6 plays a major role as a regulator of lipid droplet biogenesis in alcohol-induced intra-hepatic lipid droplet formation, identifying it as a putative target for early ALD management.

Keywords: Alcohol-associated liver disease; CerS6; Ceramide; Insulin resistance; perilipin2.

Figure 1

CerS6 ablation reduces body weight gain (A), cumulative dietary consumption (B), adiposity (C), epididymal white adipose tissue (eWAT) weight (D), and liver weight (E), and alters metabolic phenotyping (F–I) in the EtOH-fed mice. WT and CerS6 KO male and female mice were fed an EtOH diet for 6 weeks. The Comprehensive Lab Animal Monitoring System was used to measure metabolic phenotyping including CO₂ production, O₂ consumption, energy expenditure, and respiratory energy ratio. Data are presented as mean ± SEM (n = 5–9 per group). Each dot represents data from individual mice. A two-tailed unpaired t-test was used to compare WT and KO mice. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 2

CerS6 ablation improves glucose intolerance and insulin sensitivity in the EtOH-fed mice. (A, B) Insulin tolerance test (ITT) and (C, D) glucose tolerance test (GTT) were performed after 6 weeks of feeding EtOH diet in WT and CerS6 KO mice of both sexes (n = 5 per group). AUC, area under curve. Data are expressed as mean ± SEM. Each dot represents data from individual mice. A two-tailed unpaired t-test was used to compare WT and KO mice. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 3

CerS6 ablation improves EtOH-induced hepatic steatosis and reduces serum lipids in the EtOH-fed mice. WT and CerS6 KO male and female mice were fed an EtOH diet for 6 weeks. (A) Representative histologic images of hematoxylin and eosin (H&E)-stained liver sections with steatosis (%) and Oil Red O (ORO)-stained liver sections. Original magnification ×20; n = 5–9 per group. (B) Serum ALT levels, (C) hepatic TG levels, (D) hepatic FFA levels, (E) serum TG levels, (F) serum NEFA levels of the EtOH-fed mice. Data are expressed as mean ± SEM. Each dot represents data from individual mice. A two-tailed unpaired t-test was used to compare WT and KO mice. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 4

CerS6 ablation alters sphingolipids in the liver and serum of EtOH-fed WT and CerS6 KO mice. WT and CerS6 KO male and female mice were fed an EtOH diet for 6 weeks. (A) In the heatmap, the color code indicates the log2 of the ratio between means of the groups (CerS6 KO/WT) for an individual ceramide. The x axes denote the ceramide species. A more intense red color indicates a greater increase of absolute concentration of the individual ceramide in the liver and serum. (B, C) Dihydroceramide (dhCer), (D) dihydrosphingosine (dhSph), dihydrosphingosine-1-phosphate (dhSph-1P), sphingosine (Sph), and sphingosine-1-phosphate (Sph-1P) levels in the livers and serum of WT and KO mice. Data are expressed as means ± SEM (n = 5). A two-tailed unpaired t-test was used to compare WT and KO mice. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 5

RNA-seq analysis of liver transcriptomes in male WT and CerS6 KO mice fed a CON or an EtOH diet for six weeks. (A) Principal component analysis (PCA) showing global sample distribution profiles. Each dot represents data from individual mice. (B) Heatmap analysis of the differentially expressed genes regulated by genotype and diet in four groups. (C) Volcano plot of the DEGs in the liver of diet-matched WT vs. KO mice. Upregulated genes are in red and downregulated genes are in blue. Vertical lines indicate the threshold for a relative expression fold change of >2 or <−2 compared to WT and horizontal line indicates the threshold of false discovery rate <0.05. (D) Top 10 pathways based on the pathway enrichment analysis results of the genes that are significantly regulated in EtOH-fed KO mice compared to EtOH-fed WT mice, and that are significantly regulated by diet and genotype in four groups. (E–L) CerS6 deletion alters gene expression related to lipid metabolism in the livers of EtOH-fed mice. Relative mRNA expression levels were determined by real-time RT-PCR and were normalized to 36B4 mRNA levels. Relative protein levels by Western Blotting were normalized to GAPDH. Data are expressed as means ± SEM (n = 5/group). A two-tailed unpaired t-test was used to compare WT and KO mice. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001.

Figure 6

In vitro CerS6 ablation alters perilipin2 (PLIN2) expression in human hepatocytes, VL17A. (A) PLIN2 expression in cell lysates of VL17A parental and CerS6 KO cells that were treated with control or 100 mM ethanol for 48hr (n = 3); (B) with control, 100 nM C_16:0-ceramide, or 100 nM C_16:0-ceramide with 100 mM ethanol for 24hr (n = 4). Data are expressed as mean ± SEM. Statistical analyses were performed using a two-tailed unpaired t-test or one-way ANOVA with Tukey's posthoc test. ∗P < 0.05.

Figure 7

Hepatic gene expression in human subjects with different stages of ALD and diseased controls. Early ASH, early acute alcoholic hepatitis (N = 12); Nsev_AH, nonsevere acute alcoholic hepatitis (N = 11); SevResp_AH, severe alcoholic hepatitis responders to medical therapy (N = 9); SevNonR_AH, severe alcoholic hepatitis non-responders to medical therapy (N = 9); Expl_AH, explants from patients with alcoholic hepatitis (N = 11). Gene expression levels are presented in transcripts per million reads (tpm). Data are expressed as mean ± SEM. Statistical analyses were performed using a two-tailed unpaired t-test or one-way ANOVA with Tukey's posthoc test. ∗P < 0.05.