MGF / PEG-MGF studies and research data

Mechano growth factor (MGF) was first described by Geoffrey Goldspink and colleagues. They found that stretched rabbit muscle makes an extra IGF-1 mRNA. A 52 base pair insert in exon 5 of the rabbit gene (49 base pairs in the human gene) changes the reading frame, so the precursor protein ends in a different E domain (PMID 8884603, PMID 16244184). In humans this transcript is called IGF-1Ec. In rodents the same splice form is called IGF-1Eb, and the rodent peptide has 25 residues (PMID 24434354). The mature IGF-I part of every isoform is the same 70 residues (UniProt P05019 residues 49-118). The 24 residues at the end of the human IGF-1Ec precursor are YQPPSTNKNTKSQRRKGSTFEERK, residues 135-158 of UniProt isoform P05019-4. This 24-residue stretch is the 'MGF E peptide' that is made synthetically for research and sold. The name is used for three different things: the IGF-1Ec transcript, the full precursor protein, and the synthetic 24-residue peptide. Matheny and colleagues state that no matching peptide product of the Igf1 gene has been isolated from cells, culture medium, animal tissue or body fluids (PMID 20130113). Reported effects come from cell and animal work with the synthetic peptide. Examples are more myoblast proliferation and delayed differentiation in cultured myoblasts (PMID 12095637), protection of neurons in a gerbil ischemia model (PMID 16144956), better heart function after myocardial infarction in mice with a stabilised E-domain analogue (PMID 23712705), and more migration of rat bone marrow stem cells that depended on the IGF-1 receptor in that study (PMID 24323763). The 2002 muscle result was not reproduced by two companies that tested peptide up to 500 ng/ml on C2C12 cells and primary human myoblasts, and also tested primary mouse muscle stem cells (PMID 24253050). Whether the E peptide acts through the IGF-1 receptor is disputed (PMID 12095637 against PMID 24323763). PEG-MGF is sold as the peptide joined to a polyethylene glycol chain, meant to slow clearance. No peer-reviewed paper defines it. Sellers list different conjugate sizes, from about 5,000 Da to about 9.8 kDa, and one forum user wrote 'Peg-1900'. A 2026 narrative review puts PEG-MGF in its lowest evidence tier, with no peer-reviewed human studies (PMID 42395176). Products seized from the market have differed from the human sequence: two were C-terminal amidated analogues (PMID 22328223), one vial was 'MGF R23H' (PMID 28035768), and a 'full-length MGF' of 12,264.9 Da carried an R109H change (PMID 25466910). MGF is not an approved medicine in the US or EU. WADA names 'Mechano growth factors (MGFs)' in S2.3 of the 2026 and 2027 Prohibited Lists. FDA lists MGF in category 3 of the section 503A bulk substances list and lists PEG-MGF among nominated substances whose nomination was withdrawn, with a safety warning about immunogenicity.

Kind

Synthetic fragment · 24 residues

Identifiers

UniProtP05019 (opens in a new tab)PubChem CID175675731 (opens in a new tab)UNIIQ86M4KXC2P (opens in a new tab)Y65LH7J9BL (opens in a new tab)

Parent

Fragment of Insulin-like growth factor I precursor, human IGF-1Ec isoform (UniProt P05019, isoform 4 = P05019-4)The 24 residues equal the C-terminal end of the human IGF-1Ec precursor (UniProt P05019-4, residues 135-158); a script compared the isoform sequence with the GSRS sequence and found them identical. No seed record exists for IGF-1, so slug_ref is null. 'fragment_of' is the closest allowed relation; the record also covers PEG-MGF (a modified form of the peptide) and the name MGF is used for the whole IGF-1Ec isoform too. Matheny 2010 (PMID 20130113) reports that no matching peptide product of the Igf1 gene has been isolated from cells, medium, tissue or fluids, so the peptide is a synthetic copy of a predicted precursor region. uniprot_positions is null because the last 7 residues are not in the canonical entry.

Development

PreclinicalNo company holds an approval or a registered human programme. The splice form was described by Geoffrey Goldspink's group at University College London (Yang et al. 1996, J Muscle Res Cell Motil, PMID 8884603). The peptide and PEG-MGF are sold as research chemicals.ClinicalTrials.gov API v2, checked 2026-10-13. No study tests MGF or PEG-MGF as an intervention: query.intr with the exact phrase PEG-MGF in quotation marks gives 0 records (the unquoted term gives 1 record, NCT02375620, which is about PEG-somatropin), and none of the records that contain the text 'MGF' or 'mechano growth factor' give MGF as a treatment. Six studies name MGF or IGF-1Ec only as something measured: NCT07151807 (observational, serum IGF-1 and IGF-1Ec in solid tumours), NCT00635440, NCT03087292, NCT01431625, NCT03440879 and NCT04321772 (exercise or muscle studies measuring MGF mRNA or protein). Why the candidate sheet count of 106 is not reliable: query.intr 'mechano growth factor' returns 105 studies and query.intr 'MGF' returns 102 studies, and query.term returns 492 and 636. ClinicalTrials.gov expands the terms to synonyms (the cause is inferred; the lists are about IGF-1, mecasermin, acromegaly, growth hormone and stem cell factor). A script read all 105 records of query.intr 'mechano growth factor': none contains the text 'mechano growth' or 'MGF'. Of the 102 records of query.intr 'MGF', 5 contain the string MGF, and only 2 of those mean mechano growth factor (NCT07151807, NCT00635440). I could not reproduce exactly 106. The stored count is 0 because no study tests the substance. Evidence is cell and animal work, for example PMID 12095637, PMID 16144956, PMID 23712705 and PMID 24323763. The 2026 review (PMID 42395176) places PEG-MGF in its lowest evidence tier and states that no peer-reviewed human studies exist. Source (opens in a new tab)

• BPC-157 evidence
• Thymosin beta-4 (human) evidence
• GHK-Cu evidence
• TB-500 evidence