PEG-MGF: What Mechano Growth Factor Is, and What PEGylating a Fragment Changes

PEG-MGF is a synthetic peptide copying the C-terminal E-domain of one splice variant of the IGF-1 gene, with a polyethylene glycol chain attached to slow its clearance. The proposed mechanism is that this E-domain peptide acts independently of the IGF-1 receptor to expand muscle precursor cells and delay their differentiation 2. The evidence is in rodent tissue and cell lines. A pharmaceutical group testing native and stabilised forms of the peptide in mouse and human muscle cells found no effect 4. There are no human data, and no peer-reviewed study of the PEGylated fragment itself could be identified.

Three different things share the name, and most confusion comes from merging them. One is a messenger RNA splice variant, which is real and measurable. The second is a free E-domain peptide proposed to be released from the protein that RNA encodes. Whether it exists naturally is disputed. The third is a synthetic analogue of that peptide with a polymer attached, which has never been studied as such. This article takes each in turn.

One gene, several tails

IGF-1 is made as a precursor. The mature hormone, seventy residues long, is followed by a C-terminal extension called the E-domain, which is cleaved off during processing. Alternative splicing of the final exons produces precursors with different E-domains. All of them share the same mature IGF-1. They differ only in the tail that is normally removed.

In 1999 Goldspink's group reported that stretch and electrical stimulation of rabbit skeletal muscle induced expression of a particular IGF-1 splice variant, alongside structural genes associated with muscle adaptation 1. Because it was induced by mechanical load, they named it mechano growth factor. The observation concerned transcripts: a change in which messenger RNA the muscle made in response to loading, measured in rabbits.

Nomenclature is a common source of error here. The splicing pattern differs between species, so the variant called MGF corresponds to IGF-1Ec in humans and IGF-1Eb in rodents 3. The E-domain sequences also differ between species. A synthetic peptide based on one species' E-domain is not automatically equivalent to another's, and papers do not always state which sequence they used.

The E-domain peptide hypothesis

The step from a transcript to a peptide came in 2002. Yang and Goldspink synthesised a peptide corresponding to the distinctive C-terminal part of the MGF E-domain and applied it to cultured myoblasts. They reported that, unlike mature IGF-1, the E-domain peptide increased myoblast proliferation while inhibiting terminal differentiation into myotubes. They also reported that this effect was not mediated by the IGF-1 receptor, implying a separate receptor 2.

That result made MGF interesting. It suggested that muscle under load makes two signals from one gene: mature IGF-1, which drives differentiation, and a separate E-domain peptide that first expands the pool of precursor cells. It is an elegant idea. It also rests on three assumptions: that the E-domain is released as a free, stable peptide in tissue; that the synthetic sequence reproduces it; and that a receptor for it exists. None of the three had been shown directly.

Does the natural peptide exist?

A 2010 minireview examined the first assumption and found it unsupported. It noted that no analogous peptide product of the IGF-1 gene had been identified in or isolated from cultured cells, their conditioned medium, animal tissues or biological fluids 3. Synthetic E-domain peptides had been shown to promote proliferation and survival of cells in culture. But the endogenous molecule those peptides were meant to imitate had never been found. The review also noted that no receptor for the E-domain peptide had been identified.

This matters for how the whole literature should be read. If the free peptide does not occur naturally, then experiments with synthetic E-domain peptides are pharmacology with a designed molecule, not a study of a physiological signal. Such a molecule might still have useful effects. But the story built on it, that muscle releases this peptide under load, becomes a hypothesis the synthetic experiments cannot test. Expression of the splice variant transcript does not settle it, because the E-domain could remain attached to the precursor or be degraded after cleavage.

The negative result

The most direct test of the proliferation claim came from Novartis in 2014 4. The group applied MGF peptide, in native and stabilised forms, to the C2C12 mouse myoblast line, to primary human skeletal muscle myoblasts and to primary mouse muscle stem cells. The peptide did not increase proliferation in any of them. It also failed to inhibit myoblast differentiation into myotubes. Mature IGF-1 and a full-length IGF-1 precursor, tested in parallel, produced robust proliferative responses in the same cells. The authors concluded that their results called into question whether MGF has a physiological role.

The design makes this result hard to set aside. Positive controls worked, so the cells could respond. Several cell types were used, including primary human cells, not just a line. And the peptide was tested in more than one form, which answers the objection that the native sequence simply degraded before it could act. A null result is not proof of absence. But this is a careful null from a group with no stake in the hypothesis, directly contradicting the founding cell-culture claim.

The cardiac line of work

A separate strand studies the E-domain in heart muscle. In 2013 Mavrommatis and colleagues used a modified synthetic E-domain peptide in rat H9c2 cardiac cells and in a mouse model of myocardial infarction 5. The peptide entered cells rapidly, localised to the nucleus and did not activate the IGF-1 receptor. In stressed cells it prevented collapse of the mitochondrial membrane potential and the activation of caspase-3, a key step in apoptosis. Mice given the peptide after infarction showed better-preserved cardiac contractility, less pathological hypertrophy and fewer apoptotic nuclei than untreated controls.

These results do not conflict with the Novartis data, because they measure something different: survival of stressed cardiac cells, not proliferation of muscle precursors. They are consistent with an intracellular, receptor-independent action, which would also explain why no surface receptor has been found. But they are rodent and cell-line data from one group, the peptide was a modified analogue, and the work does not address the question of whether a free E-domain peptide occurs naturally.

What PEGylating the fragment changes

The rationale for attaching polyethylene glycol is straightforward. A short linear peptide is cleared from plasma within minutes by proteolysis and renal filtration. A PEG chain adds a large hydrated coil that raises the molecule's apparent size above the filtration threshold and physically shields it from proteases. That is the general mechanism of PEGylation, and it is well established for approved protein drugs.

What PEGylation also does is make a different molecule. For a fragment this short, the polymer makes up most of the conjugate's mass. The same shell that blocks proteases and the kidney also hinders the peptide's contact with anything it is meant to bind, so loss of intrinsic activity is the expected cost. The conjugate is also confined differently in the body and distributes more slowly into tissue. Whether any activity of the E-domain peptide survives PEG attachment, and at which attachment site, has not been reported in the peer-reviewed literature.

There is also a mismatch between the modification and the proposed mechanism. The cardiac data suggest the E-domain acts after entering cells 5. A bulky polymer coil is designed to keep a molecule in circulation and reduce uptake. A modification that prevents cellular entry could defeat the very action the peptide was proposed to have. That is a testable question, and it has not been tested. The PEGylated product therefore rests on the evidence for the unmodified fragment, which is contested, plus an untested assumption that PEG leaves its activity intact.

Regulatory position in 2026

PEG-MGF is not approved as a medicine by any major regulator. The FDA sorts bulk substances nominated for compounding under section 503A into categories, and Category 2 holds substances for which it has identified significant safety risks 6. On 15 April 2026 the FDA announced that twelve peptides, including mechano growth factor in its PEGylated form, would be removed from Category 2. PEG-MGF was among five set aside for a later advisory committee review expected before February 2027 7. The removal followed withdrawal of the nominations. It did not place PEG-MGF in Category 1 and does not in itself make it eligible for compounding 8.

What would be needed

1. Direct detection of a free E-domain peptide in tissue or plasma by mass spectrometry, which would establish whether the natural molecule exists.
2. Identification of a receptor or intracellular binding partner, with a demonstration that removing it abolishes the effect.
3. Independent replication of the proliferation claim with positive controls, resolving the conflict with the Novartis data.
4. Characterisation of the PEGylated conjugate: attachment site, retained activity, cellular uptake and pharmacokinetics.
5. Only after these, safety and pharmacokinetic studies in humans.

The fair summary

The splice variant is real: loading changes how muscle splices IGF-1 transcripts, first shown in rabbits 1. The idea that muscle releases a free E-domain peptide with its own signalling role is attractive and unproven, because the natural peptide has not been isolated 3. The founding cell-culture result 2 was not reproduced in a careful study with positive controls 4. The cardiac data point to a different, intracellular action in rodent models 5. The PEGylated fragment adds an untested modification to that contested foundation. There is no human evidence at any of these levels.