Almost every compound on this shelf was discovered as a molecule and named afterwards: isolated, sequenced, labelled. MGF ran the other way. "Mechano growth factor" was first the name of a messenger RNA — a splice variant of the IGF-1 transcript that appears in muscle after mechanical loading — and the peptide sold under that name is a synthetic 24-mer inferred from what that transcript's protein product would be cut into.
That distinction is not pedantry. It sits underneath everything else here: an unidentified receptor, a failed replication from two pharmaceutical companies, and a supporting literature that often used a chemically modified analog rather than the native sequence.
The splicing story the name came from
The IGF-1 gene has six exons, and alternative splicing across exons 4, 5 and 6 produces several pro-IGF-1 isoforms. All share the same 70-amino-acid mature IGF-1 core — the part that binds IGF-1R — and differ in the C-terminal extension called the E-domain: IGF-1Ea, IGF-1Eb, IGF-1Ec.
In humans, a 49-base-pair insert in exon 5 causes a reading-frame shift, generating the distinct Ec E-domain. Rodents have a 52-bp insert producing the analogous form, called IGF-1Eb by rodent convention. The same entity carries different letters in different species, which is why the literature reads as confusing: the replication paper below says "IGF-1Eb," the Goldspink-lineage papers say "IGF-1Ec," same splice event, different animals.
The load-bearing observation was about transcripts, not peptides. After muscle injury or loading, IGF-1Eb/Ec mRNA rises first, correlating with satellite cell and myoblast proliferation; IGF-1Ea rises later, correlating with differentiation into mature fibres. That pattern earned the transcript its nickname. The leap — and it is a leap — was to synthesize the 24 C-terminal residues of the pro-peptide and treat them as the active agent.
What is actually in the vial
The peptide is a 24-residue fragment corresponding to the C-terminal end of the human IGF-1Ec pro-peptide, containing no part of mature IGF-1. At that length, solid-phase synthesis is the only plausible route.
Reading the composition per our sequence primer: one tyrosine, no tryptophan, so A280 quantitation is possible but weak; no methionine or cysteine, so oxidation and disulfide scrambling are off its liability list; and six or seven basic residues in 24, predicting substantial counterion mass and low net peptide content by composition rather than defect.
There is one documentation wrinkle worth knowing before comparing a certificate of analysis against a theoretical mass. The current reference-database sequence and the sequence used throughout the published MGF literature differ at one position. UniProt (P05019-4) and NCBI RefSeq (NP_001104753.1) both give the human IGF-1Ec C-terminus ending …STFEERK. The synthetic peptide described in the MGF literature, and the sequence generally circulated for the catalog compound, ends …STFEEHK.
One residue, arginine versus histidine — roughly 19 Da on a ~2,850 Da peptide. Trivially resolvable by mass spectrometry, and completely invisible if nobody states which target mass the observed one was compared against. We are not adjudicating which is correct; the useful point is that this is a molecule where "the mass matched" means nothing until you know what it was matched to.
The proposed mechanism, and the receptor that was never found
The founding functional paper is Yang and Goldspink, FEBS Letters 2002;522(1-3):156-160. In cultured myoblasts, the isolated E-domain increased proliferation while inhibiting terminal differentiation — the opposite division of labour from mature IGF-1, which drives differentiation and fusion. The mechanistic claim rested on an antibody experiment: blocking IGF-1R did not abolish the effect, which the authors read as evidence the E-peptide acts through a different receptor.
Note what that establishes: a negative result about one receptor, not the identification of a positive one. In the two decades since, no receptor for the MGF E-peptide has been established. As with FOXO4-DRI and MOTS-c, the vocabulary of receptor pharmacology — affinity, efficacy, selectivity, bias — has nothing to attach to. Vendor copy quoting a binding affinity for MGF is quoting something unpublished.
A follow-on human-cell study (Kandalla et al., Mechanisms of Ageing and Development 2011;132(4):154-162) reported increased proliferative lifespan and delayed senescence in primary human muscle progenitor cells from neonatal and young adult donors — but not from old adult donors. That age-dependence is usually dropped when the paper is cited in support of sarcopenia research.
The replication problem — the most important paper on this compound
In 2014, two pharmaceutical companies tried to reproduce the core myoblast findings and could not. Fornaro et al., American Journal of Physiology — Endocrinology and Metabolism 2014;306(2):E150-6 (PMID 24253050), is a joint Novartis/GSK paper whose title states the result: "Mechano-growth factor peptide… has no apparent effect on myoblasts or primary muscle stem cells."
The design is unusually thorough for a negative report:
- C2C12 cells and primary human myoblasts — no increase in proliferation up to 500 ng/mL (an assay ceiling, not a dose), and no inhibition of differentiation into myotubes
- Positive controls worked — the same cells responded to mature IGF-1 and to full-length IGF-1Eb, so the systems were not simply unresponsive
- Primary mouse muscle stem cells, tested in case cultured lines had lost responsiveness — also null
- Cardiac myocyte p-ERK activation, a separately published effect, tested with both the native and a stabilized peptide — no response, while IGF-1 gave a robust one
Their conclusion: the results "call into question whether there is a physiological role for MGF."
Read it alongside a quieter observation. A 2010 Endocrinology minireview (Matheny, Nindl & Adamo, 2010;151(3):865-875) notes that no analogous peptide product of the Igf1 gene has been identified in, or isolated from, cultured cells, their conditioned medium, or animal tissues or biological fluids. The synthetic 24-mer is well-characterized; the endogenous free E-peptide it models has never been shown to exist as a separate molecular species.
A failed replication is not proof of absence, and the field has not uniformly accepted Fornaro's conclusion. But taken with Matheny's point it inverts the burden: the question is no longer only what does MGF do, but is there a molecule for it to be a copy of. On our evidence hierarchy, that is a rung-1 dispute never resolved before the compound reached a catalog.
The stabilized-analog footnote
Several cardiac MGF papers did not use the native sequence. Mavrommatis et al. (Molecular and Cellular Biochemistry 2013;381(1):69-83) describe stabilizing their peptide by amidating the C-terminus and switching arginines 14 and 15 to the D-stereoisomer.
Those are the anti-peptidase modifications catalogued in our peptidase and clearance explainer, and they mean the molecule tested was not the one the splice variant encodes but a designed analog with better protease resistance.
The analytical consequence is familiar: D- and L-amino acids are isobaric, so a mass spectrum cannot distinguish the native peptide from a D-substituted version, and a C-terminal amide shifts the mass by only ~1 Da against the free acid. If a supplier's sequence line carries no D-prefixes and no amidation annotation, the product is the native form — and papers using the stabilized analog are not straightforwardly about it.
What PEGylation is doing — and what it does to the paperwork
PEG-MGF is the same E-domain peptide with a polyethylene glycol chain attached, for half-life extension. That rationale is directionally sound but usually described wrongly. PEGylation is primarily an anti-filtration strategy, not an anti-protease one: a bare ~2.8 kDa 24-mer sits far below the glomerular filtration cutoff and clears renally almost immediately, and PEG slows that by raising hydrodynamic radius. Same category as albumin binding on CJC-1295 DAC — and the mirror image of IGF-1 LR3, where engineering away from a carrier produced measurably faster clearance.
What is not sound is the arithmetic that accompanies it. The half-life figures attached to PEG-MGF in vendor copy — minutes for MGF, roughly a day for PEG-MGF — do not come from published pharmacokinetic measurement of either compound. No human PK study of either exists. A plausible mechanism is being reported as a result.
PEGylation also changes what a certificate of analysis can establish:
- PEG is polydisperse. The reagent is a distribution of chain lengths, not one molecule, so a PEGylated peptide has no single intact mass — the spectrum shows peaks spaced by the 44 Da ethylene-oxide repeat unit, not the clean [M+H]⁺ our chromatogram-and-spectrum piece describes.
- PEG size and attachment chemistry are rarely disclosed. A 5 kDa and a 20 kDa conjugate of the same peptide are different products with different clearance behaviour, and the name does not distinguish them.
- The conjugation site is usually unstated. With multiple lysines plus a free N-terminus, amine-directed chemistry can produce positional isomers — same mass, different attachment point, inseparable by intact mass.
- PEG raw material varies between suppliers. A 2024 comparison of PEG-lipid raw materials reported measurable vendor-to-vendor differences in impurity profiles and chain-length polydispersity: the polymer is a reagent with its own quality file.
A purity percentage on a PEGylated product is therefore answering a harder question than the same number on a plain peptide, and the documentation rarely says how.
Where the human file stands
Empty. A ClinicalTrials.gov search returns no interventional study in which MGF or PEG-MGF was administered to human participants — the hits are unrelated records where "growth factor" appears in another context. No human PK, safety or efficacy data exists.
It was named in the mainstream literature this year: a June 2026 narrative review in Frontiers in Endocrinology (2026;17:1822475) on peptides modulating the GH–IGF-1 axis stratifies compounds "from regulatory-grade randomized trial data to a complete absence of human studies," placing PEG-MGF alongside IGF-1 LR3 in the analogue group — a freely citable placement, at the bottom of the ladder.
The 2024–2026 preclinical literature is diffuse rather than convergent, and an earlier MGF paper on intervertebral disc cells was retracted in 2024. None of it addresses the replication question, and none of it is in humans.
FAQ
Is MGF a fragment of IGF-1? Not of mature IGF-1. It corresponds to the C-terminal E-domain of the IGF-1Ec pro-peptide — a region removed during processing — and contains none of the 70-residue sequence that binds IGF-1R. That is why it does not sit on the same mechanistic tier as IGF-1 LR3, despite both being filed under growth hormone.
Does the failed replication mean MGF does nothing? No — one negative paper does not erase a literature. It shifts where the uncertainty sits: two industrial labs, multiple cell types, working positive controls, native and stabilized peptides, no effect. Combined with the fact that a free endogenous E-peptide has never been isolated from any tissue, the honest description is disputed at the most basic level — unusual even for a preclinical-only compound.
Is PEG-MGF better characterized than MGF? No — less. It inherits every open question about the peptide and adds polydispersity, undisclosed PEG size, and conjugation-site heterogeneity. More engineering, not more evidence. See /quality/ and /library/.
This article is educational and for the laboratory research community. Trulogic Labs products are sold for laboratory and research use only and are not for human consumption.