If you searched for "HGF" and landed here, the most useful first move is to clear up one misconception. HGF is not a peptide. It is hepatocyte growth factor: a large secreted glycoprotein of about 90 kDa, two-chained, related to the clotting protein plasminogen. It ends up in the same lists as short peptides like BPC-157 through catalogue proximity, not similarity, and that is exactly why it attracts expectations it cannot support.
The strange detail that makes everything click
HGF looks like an enzyme but is not one. Its β-chain carries a domain structurally resembling a serine protease, but that domain is catalytically dead: it lacks the key catalytic-triad residues and cuts nothing. Evolution took a protease scaffold and repurposed it into a signalling molecule.
Another quirk: HGF itself is made as an inactive single-chain precursor and only "switches on" after a different protease cleaves it within the tissue. So unlike a peptide that acts immediately, HGF needs a separate on-site activation step.
One protein, three discoveries
HGF was discovered three times and given three names. Some saw a mitogen that drives liver cells to divide after injury. Others saw a "scatter factor" that made compact cell colonies disperse. Others saw a factor toxic to certain tumour lines. Only cloning in the late 1980s revealed it was the same protein.
This is not trivia. It explains at once why HGF "does everything": division, movement, tissue building, cell survival. It is not a list of different effects but one biological programme that different labs caught from different angles.
The central inconvenience: HGF is the ligand of an oncogene
HGF has exactly one receptor, c-Met, the product of a proto-oncogene. c-Met activation triggers an "invasive growth" programme: cells loosen their junctions, become motile, and invade deeper into tissue. In an embryo and in an injured liver this is repair. In a tumour it is metastasis.
Hence the paradox that simplified "HGF the regeneration factor" descriptions leave out. The pharmaceutical industry has spent years investing in switching c-Met off, not on. The approved drugs on this axis are c-Met inhibitors (capmatinib, tepotinib) for lung cancer. There is no approved HGF agonist anywhere, a direct consequence of the fact that "more growth" is not always "better".
What human studies actually showed
The temptation to present HGF as a proven regenerative agent is strong. The honest picture is more modest. Because the recombinant protein is nearly impossible to give systemically (it survives minutes in blood, sticks in the matrix, and is hard to manufacture), developers took detours.
Gene therapy. VM202 (Engensis) is a plasmid that makes muscle produce HGF locally. In phase 2 for painful diabetic neuropathy the lower dose gave a clear signal: nearly half of participants reached ≥50% pain reduction versus ~18% on placebo. But there are two catches: the higher dose worked worse than the lower one (illogical and concerning), and the decisive phase 3 failed its primary endpoint, superiority over placebo was not demonstrated.
Local protein. The other approach is to place recombinant HGF straight into the spinal fluid in acute spinal cord injury (KP-100IT). A published trial showed acceptable safety, but it is a safety study, not proof that the method restores function.
Unembellished bottom line: tolerability in these studies was reasonable, but efficacy at the highest level of evidence is not yet confirmed. Most of the striking "HGF repairs tissue" results still belong to preclinical work, rodents and cell cultures.
Why this matters if you work with the reagent
Several different molecules are sold under the single name "HGF", and they behave differently: full-length active two-chain HGF; inactive single-chain pro-HGF that still needs activating; truncated NK1/NK2 fragments that can even act as antagonists depending on conditions. The system the protein was made in determines its glycosylation and folding. For a reproducible experiment these parameters must be fixed in the protocol, otherwise data from different lots are not comparable.
Where this fits
HGF is interesting precisely because it marks the limit of the "growth factor = repair" idea. The same molecule that heals a liver in the lab shares a receptor with metastasis. Compare it with other approaches to fibrosis and regeneration in our reference, the anti-fibrotic B7-33 acts through an entirely different receptor (RXFP1), and the neuroregenerative ARA-290 through an innate receptor complex. Different entry points into the same problem, with different risk profiles.
The full technical treatment, structure, primary-source references and a list of open questions, is in the HGF monograph. The reagent itself sits in the healing and regeneration section.
See also: a general overview of HGF: what it is and why it is studied.
RUO. HGF is treated strictly as a research reagent. It is not a medicine or a supplement and is not intended for human consumption. This text describes published research and is not medical advice.