Monomeric/dimeric forms of Fgf15/FGF19 show differential activity in hepatocyte proliferation and metabolic function.

Human Fibroblast Growth Factor 19 (FGF19) and mouse ortholog Fgf15 play similar roles in liver regeneration and metabolism via the activation of Fgfr4/b-klotho (Klb). Monomeric FGF19 and dimeric Fgf15 are both necessary for liver regeneration and proper bile acid (BA) metabolism. FGF19 elicits stronger effects than Fgf15 on glucose and fatty acid metabolism and only FGF19 induces hepatocellular carcinoma (HCC). However, inhibiting FGF19/FGFR4 signaling in HCC patients is associated with toxicity due to elevated BA levels. Here, we examine the structure/function relationship in Fgf15/FGF19 to better understand the molecular basis for their distinct functions. We demonstrate that FGF19 is a more effective activator of Fgfr4 and of downstream signaling (Erk, Plcg1) than Fgf15. Furthermore, we use site-directed mutagenesis to show that the presence or absence of an unpaired cysteine in Fgf15/19 modulates ligand structure and determines the ability of these molecules to induce hepatocyte proliferation, with monomers being more potent activators. Consistent with these findings, an engineered dimeric variant of FGF19 is less effective than wild-type FGF19 at inducing liver growth in cooperation with the Wnt-enhancer RSPO3. In contrast to effects on proliferation, monomeric and dimeric ligands equally inhibited the expression of Cyp7a1, the enzyme catalyzing the rate limiting step in BA production. Thus, structure and function of Fgf15/FGF19 are intricately linked, explaining why FGF19, but not Fgf15, induces liver tumorigenesis. Our data provide insight into FGF19/FGFR4 signaling and may inform strategies to target this pathway while limiting on-target toxicity due to dysregulation of BA production or induction of hepatocyte proliferation.