Characterisation of PNPLA3 Variant Effects in 3D Decellularized Human Liver Scaffolds

Metabolic dysfunction-associated steatotic liver disease (MASLD) is the most common chronic liver disease worldwide, affecting approximately 38% of the adult population. Its largest single inherited determinant is the rs738409 (c.444C>G) variant of patatin-like phospholipase domain-containing 3 (PNPLA3), which is associated with progression to fibrosis, cirrhosis and hepatocellular carcinoma (HCC). PNPLA3 is expressed both in hepatocytes, where lipids accumulate, and in hepatic stellate cells (HSC), which drive fibrogenesis, and the consequence of the substitution is different in both cell types. Despite extensive study, the mechanism by which the variant causes disease has not been fully clarified. Every cell type of an affected liver carries the same genotype, so the contribution of any one cell type cannot be isolated in a patient. In vitro coculture studies have approached this by varying the genotype of one cell type at a time; the two have not been varied together. Preclinical liver models are also commonly built from cells whose PNPLA3 genotype is neither determined nor reported. This project investigated whether the PNPLA3 genotype of the HSCs and that of the HCC cells when cultured acts independently on fibrogenic gene expression. Three-dimensional scaffolds were obtained by decellularizing healthy and cirrhotic human liver, and the cirrhotic scaffolds were quantified for minimal presence for residual DNA content and maintenance of matrix architecture. Primary human HSCs of known genotype (CC, wild-type; GG, homozygous for the variant) were co-cultured for fourteen days with HCC cell lines of known genotype, in all four combinations, on healthy scaffolds. ACTA2, COL1A1, TGFβ and YAP1, markers for HSC activation and fibrogenesis, were measured by RT-qPCR and analysed by two-way analysis of variance. A second, smaller experiment compared one genotype-mismatched pair on healthy and on cirrhotic scaffolds. The cirrhotic scaffolds met the accepted criteria for decellularization: residual DNA fell below the 50 ng/mg threshold, and the collagen network and the fibrotic septa were retained. The interaction between the two genotypes was significant for all four genes. Where the HSCs were CC, the genotype of the partner changed expression several fold; where HSCs were GG, it changed only COL1A1. This thesis shows that the PNPLA3 genotype of HSCs and of HCC cells do not act independently on fibrogenic gene expression. Therefore, the PNPLA3 genotype of the cells from which a three-dimensional liver model is built and should be determined and reported.