Hepatocyte growth factor (HGF) is secreted as an inactive single-chain HGF (scHGF); however, only proteolytically processed two-chain HGF (tcHGF) can activate the MET receptor. We investigated the localization of tcHGF and activated/phosphorylated MET (pMET) using a tcHGF-specific antibody. In day 16.5 mouse embryos, total HGF (scHGF + tcHGF) was mainly localized in smooth muscle cells close to, but separate from, MET-positive epithelial cells in endodermal organs, including the stomach. In the adult stomach, total HGF was localized in smooth muscle cells, and tcHGF was mainly localized in the glandular base region. Immunostaining for pMET and Lgr5-driven green fluorescent protein (GFP) indicated that pMET localization overlapped with Lgr5+ gastric stem cells. HGF promoted organoid formation similar to EGF, indicating the potential for HGF to promote the survival and growth of gastric stem cells. pMET and tcHGF localizations changed during regeneration following gastric injury. These results indicate that MET is constantly activated in gastric stem cells and that the localization of pMET differs from the primary localization of precursor HGF but has a close relationship to tcHGF. Our results suggest the importance of the microenvironmental generation of tcHGF in the regulation of development, regeneration, and stem cell behavior.
Using a random non-standard peptide integrated discovery system, we obtained cyclic peptides that bind to hepatocyte growth factor (HGF) or mesenchymal-epithelial transition factor. (MET) HGF-inhibitory peptide-8 (HiP-8) selectively bound to two-chain active HGF, but not to single-chain precursor HGF. HGF showed a dynamic change in its molecular shape in atomic force microscopy, but HiP-8 inhibited dynamic change in the molecular shape into a static status. The inhibition of the molecular dynamics of HGF by HiP-8 was associated with the loss of the ability to bind MET. HiP-8 could selectively detect active HGF in cancer tissues, and active HGF probed by HiP-8 showed co-localization with activated MET. Using HiP-8, cancer tissues with active HGF could be detected by positron emission tomography. HiP-8 seems to be applicable for the diagnosis and treatment of cancers. In contrast, based on the receptor dimerization as an essential process for activation, the cross-linking of the cyclic peptides that bind to the extracellular region of MET successfully generated an artificial ligand to MET. The synthetic MET agonists activated MET and exhibited biological activities which were indistinguishable from the effects of HGF. MET agonists composed of cyclic peptides can be manufactured by chemical synthesis but not recombinant protein expression, and thus are expected to be new biologics that are applicable to therapeutics and regenerative medicine.
Non-native ligands for growth factor receptors with distinct chemical properties and different biological activities have the potential to become therapeutic applications. We previously generated MET/hepatocyte growth factor (HGF) receptor agonists using bivalent macrocyclic peptides. The highest MET-activating agonists exhibited biological activity that was indistinguishable from the effects of HGF. In this study, we investigated MET activation, signal characteristics, and biological responses induced by a macrocyclic peptide partial agonist known as aML5-PEG11. aML5-PEG11 induced weak tyrosine phosphorylation of MET while enhancing cell migration with potency comparable to HGF. aML5-PEG11 induced marked AKT (protein kinase B) and ERK (extracellular signal-regulated kinase) activation at a comparable potency and time-dependency to HGF, which suggests that enhancement of cell motility is attributable to activation of these molecules. In a 3-D culture of bile duct cancer cells in collagen gel, HGF induced robust activation of MET, ERK, and AKT, which was associated with enhanced expression of genes involved in bile duct development and subsequent branching of tubulogenesis. In contrast, aML5-PEG11 induced marginal activation of MET, ERK, and AKT (levels near the detection limits), which was associated with failure to enhance the expression of genes involved in bile duct development and a lack of tubulogenic response. Thus, MET activation by aML5-PEG11 couples to biological responses differently from HGF in an extracellular context-dependent manner.
NK1, a splicing variant of hepatocyte growth factor (HGF), binds to and activates Met receptor by forming an NK1 dimer and 2:2 complex with Met. Although the structural mechanism underlying Met activation by HGF remains incompletely resolved, it has been proposed that the NK1 dimer structure participates in this activation. We investigated the NK1 dimer interface’s role in Met activation by HGF. Because N127, V140, and K144 are closely involved in the head-to-tail NK1 dimer formation, mutant NK1 proteins with replacement of these residues by alanine were prepared. In Met tyrosine phosphorylation assays, N127-NK1, V140-NK1, and K144-NK1 showed 8.3%, 23.8%, and 52.2% activity, respectively, compared with wild-type NK1. Although wild-type NK1 promoted cell migration and scattering, N127-NK1, V140-NK1, and K144-NK1 hardly or marginally promoted them, indicating loss of activity of these mutant NK1 proteins to activate Met. In contrast, mutant HGFs (N127-HGF, V140-HGF, and K144-HGF) with the same amino acid replacements as in NK1 induced Met tyrosine phosphorylation and biological responses at levels comparable to those of wild-type HGF. These results indicate that the structural basis responsible for NK1-dependent Met dimer formation and activation differs from, or is at least distinguishable from, the structural basis responsible for HGF-dependent Met activation.