Endoplasmic reticulum stress in lipid-laden hepatocytes activates Caspase-2, which proteolytically activates Site-1 protease (S1P). Active S1P cleaves and releases sterol regulatory element-binding proteins (SREBPs) from the ER membrane, allowing their nuclear translocation and transcriptional upregulation of de novo lipogenesis genes including FASN, ACC1, and HMGCR. The resulting lipid accumulation re-amplifies ER stress, sustaining a Casp2-S1P-SREBP feed-forward cycle of progressive steatosis.

Free cholesterol accumulation in stressed hepatocytes inhibits TAZ (WWTR1) degradation through the Hippo kinase cascade, allowing nuclear TAZ to bind TEAD consensus elements in the Ihh promoter and transcriptionally induce Indian Hedgehog. Secreted Ihh acts in a paracrine manner on hepatic stellate cells, driving their activation into collagen-secreting myofibroblasts. As the resulting ECM stiffens, integrin-linked mechanosensing further stabilizes YAP/TAZ, creating a stiffness-dependent amplification loop that sustains fibrosis independently of the initial lipotoxic trigger.

MASH-HCC accumulates CD8+ PD1+TOX+ T cells that are terminally exhausted and auto-aggressive, producing TNF that damages hepatocytes rather than eliminating tumor cells. IgA+ plasmocytes generated via IL-21R-STAT1-c-Jun/c-Fos class switching express PD-L1 and secrete IL-10, suppressing any residual CTL activation. ApoE from tumor-infiltrating immune cells further activates PI3K-AKT-NF-kB oncogenic signaling in HCC cells. Together these mechanisms create an immunologically cold TME where anti-PD-1 blockade has no functional effector T cells to reinvigorate.

METTL3, the catalytic subunit of the m6A RNA methyltransferase complex, modifies SCAP mRNA with N6-methyladenosine marks that enhance its translational efficiency. SCAP is the escort protein that chaperoles SREBPs from the ER to the Golgi for activating cleavage; increased SCAP abundance promotes SREBP nuclear translocation and upregulates cholesterol biosynthesis, directly linking RNA epigenetic modification to the lipotoxic cascade driving MASLD-to-HCC progression.

NETs in the MASH liver bind TLR4 on naive CD4+ T cells and shift their metabolic program toward mitochondrial oxidative phosphorylation (OXPHOS), a bioenergetic state that favors differentiation into FoxP3+ regulatory T cells over effector Th1 lineages. This NET-TLR4-Treg axis reduces anti-tumor immune surveillance and adds a neutrophil-mediated immunosuppressive layer to the CD8+ T cell exhaustion and IgA+ B cell suppression already present in MASH-HCC.

In early MASLD, ECM1 sequesters TGF-beta1 in its latent form by preventing activation by thrombospondins and MMPs; active TGF-beta1/Smad2/3 signaling in this context restrains hepatocyte proliferation and promotes apoptosis of damaged cells, functioning as a tumor suppressor. As fibrosis progresses and ECM1 is lost, constitutive TGF-beta1 activation drives epithelial-mesenchymal transition through ZEB1 and SNAI1 upregulation, metastatic invasion, and HCC progression, switching to a tumor-promoting role in established MASH-HCC.

miR-33 post-transcriptionally represses CPT1a (carnitine palmitoyltransferase 1a) and PGC1-alpha, limiting the hepatocyte's capacity for mitochondrial fatty acid oxidation. In MASH, miR-33 activity combines with reduced AMPKalpha signaling to severely impair fatty acid disposal, trapping hepatocytes in a lipid-accumulating state. Deletion of miR-33 in experimental models relieves this repression, restoring PGC1-alpha and MFN2 expression, improving mitochondrial dynamics, and reducing steatosis, positioning miR-33 as a targetable microRNA regulator of hepatic metabolic homeostasis.