Fig 2: Decreased LXR signaling disrupts cholesterol homeostasis in PNPLA3 I148M carrying human HSCs. Human primary HSCs isolated from liver resections and cultivated in vitro and LX-2 stably overexpressing cells (n = 3 for each PNPLA3 genotype) obtained, as described in Materials and Methods. All bar graphs show mean values ± SD. (A,B) Expression of cholesterol efflux transporter ABCA1 and cholesterol-metabolizing enzymes CYP7B1, CYP27A1, and CH25H analyzed by RT-PCR and normalized to 18s in untreated primary HSCs carrying either the WT or the variant of PNPLA3 (n = 3 for each genotype); *P < 0.05, **P < 0.01. (C) TC and FC amounts quantified by using colorimetric quantitative assay on total lipid extracts collected from HSCs carrying either WT or I148M PNPLA3. Data shown are representative of two independent lipid extractions normalized to internal standards (Abcam); **P < 0.01. (D,E) Total protein extracts were isolated from primary HSCs carrying either WT or I148M PNPLA3 (n = 3 for each genotype) and analyzed by western blotting for SREBP-2, LDLR, phospho-HMGCR, and total HMGCR. Densitometry analysis was performed using ImageJ software, and data were normalized to calnexin. Data presented as protein amount relative to calnexin. Phospho-HMGCR was normalized to total HMGCR and expressed as protein ratio. Open bars refer to PNPLA3 WT HSCs and closed bars to I148M HSCs (n = 3 for each genotype); *P < 0.05, **P < 0.01 versus WT carriers. Abbreviations: phospho-, phosphorylated; rel., relative.

Fig 3: LXR agonist T09 alone restores ABCA1 and attenuates inflammatory and fibrogenic markers independently from PPAR? in HSCs with the PNPLA3 I148M variant. LX-2 stable overexpressing WT or I148M PNPLA3 untreated or treated for 24 hours with either T09 (+T09) or the combination T09 and ROSI (+T09+Rosi). mRNA expression of (A) ABCA1, (B) collagen1a1, (C) PPAR?1 and (D) CCL5 and analyzed by RT-PCR and normalized to 18s. White bars show untreated HSCs, gray bars show HSCs treated with T09, and black bars show HSCs treated with T09 or the combination (+T09+Rosi). All bar graphs show mean values ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001 versus untreated WT or I148M HSCs (n = 3 for each genotype).

Fig 4: LXR inverse agonist SR9238 and antagonist GSK2033 blocked T09-induced LXR signaling in human HSCs. Whole mRNA extract collected from untreated (ctrl) stably overexpressing LX-2 carrying either WT or I148M PNPLA3, treated with LXR inverse agonist (SR9238) or LXR antagonist (GSK2033) for 24 hours. Expression of ABCA1 and SREBP-1c analyzed by RT-PCR and normalized to 18s. I148M expression data were normalized to WT ctrl. All bar graphs show mean values ± SD. *P < 0.05, **P < 0.01, and ***P < 0.001 versus untreated WT HSCs or as indicated in the graph. Abbreviation: Ctrl, control.

Fig 5: Summary representation of impaired LXRa signaling resulting in altered TC and FC accumulation in human HSCs carrying I148M PNPLA3. Activated human HSCs carrying I148M have decreased PPAR? transcriptional activity due to phosphorylation on specific inhibitory residue (p-PPAR?), which directly impairs LXR target gene expression involved in cholesterol and lipid metabolism (ABCA1, SREBP-1c, FASN). As such, reduced cholesterol efflux through ABCA1 leads to TC and FC accumulation, accompanied by reduced endogenous cholesterol synthesis (SREBP-2, HMGCR) and decreased extracellular uptake (LDLR). Reduced PPAR/LXR signaling axe decreases the anti-inflammatory ability of these nuclear receptors, thus resulting in higher expression of profibrogenic features (CCL5, CCL2, collagen1a1, a-SMA, increased migration and proliferation).