Non-alcoholic fatty liver disease (NAFLD) comprises a group of liver pathologies in which the hallmark is an aberrant accumulation of triglycerides in hepatocytes. These disorders mostly derive from metabolic abnormalities, such as obesity, type II diabetes and from genetic risk factors. Among the factors determing susceptibility and progression across the disease spectrum, the genetic polymorphism of the human Adiponutrin gene rs738409, known also as PNPLA3 I148M, represents one of the most recently studied and of highest relevant determinants for pathogenesis and progression of NAFLD in the last years, since it has been shown to be associated with simple steatosis and progression towards non-alcoholic steatohepatitis (NASH), liver fibrosis and cirrhosis. The enzymatic role of this protein is still uncertain: some studies reported a lipase activity, while others showed lysophosphatidic properties, accordingly the I148M would result either as loss or gain of function. Moreover, both in vitro and in vivo studies have so far been lacking to understand the molecular and cellular mechanisms how the I148M variant contributes to the progression from simple steatosis to more advanced liver injury. Therefore, the aim of my doctoral thesis was to investigate the specific input of PNPLA3 and its genetic variant in the mechanisms of liver fibrogenesis, which represents the key step toward severity of NAFLD development, such as NASH and later cirrhosis. In this view, hepatic stellate cells (HSCs) are the main responsible of the scarring process which characterizes liver fibrosis and, notably, so far nobody investigated on the role of PNPLA3 during HSCs in vitro activation. My research revealed that primary human HSCs require PNPLA3 for achieving a fully activated phenotype and its repression was related to less fibrogenic activity. Moreover, a comparison between primary human HSCs expressing the wild type (WT) PNPLA3 and the I148M isoform showed that cells carrying the variant express significantly higher amount of pro-inflammatory and pro-fibrogenic cytokines, compared to their WT counterpart, such as the chemokine C-C motif ligand 5 (CCL5) and the granulocyte and macrophages colony stimulating factor (GM-CSF), consequently contributing to immune cells recruitment in the site of injury. To confirm our findings in primary cells, we generated a stable cell line overexpressing either the WT or the I148M PNPLA3 isoforms using LX-2 cells, a human immortalized HSC cell line. Interestingly, LX-2 overexpressing the PNPLA3 genetic variant recapped the pro-inflammatory and pro-fibrogenic features showed by primary human HSCs. Moreover, medium cytokine profile analysis revealed that, in addition to CCL5 and GM-CSF, other pivotal pro-inflammatory cytokines are released in the medium collected from LX-2 overexpressing the I148M isoform, such as the chemokine C-C motif ligand 2 (CCL2), the chemokine C-X-C motif ligand 1 and 8 (CXCL1, CXCL8). Interestingly, both primary human HSCs and LX-2 expressing the I148M PNPLA3 showed also significantly higher lipid droplet (LD) amount and lower retinol content, when compared to WT PNPLA3 expressing cells. Therefore, by analyzing the regulation and DNA specific binding of some anti fibrotic nuclear receptors, such as the proliferator-activated receptor gamma (PPAR), the retinoid X receptor (RXR) and the retinoic acid receptor (RAR), we found that both primary HSCs and LX-2 carrying the I148M variant display significantly reduced transcriptional activity of all those corresponding nuclear receptor response elements DR-1, DR-2 and DR-5, as indicated by luciferase and gel-shift assays.