Follistatin-344 is not a 15-amino-acid recovery peptide. It is a full-length splice variant of a secreted glycoprotein that binds myostatin and related TGF-beta ligands. The muscle-growth papers people cite almost all delivered a gene, AAV1-FS344, so muscle kept making the protein for months.

TL;DR

  • Myostatin (GDF-8) is a muscle-made brake. Knock it out and individual mouse muscles weigh 2 to 3 times wild type.
  • Follistatin binds myostatin at the activin type II receptor. Overexpress it in a myostatin-null mouse and muscle mass can reach about four times wild type.
  • Researchers picked FS344 because it is processed to circulating FS315, which hits muscle with less activin and FSH off-target activity than heparin-binding FS288.
  • Human data: six men with Becker muscular dystrophy received intramuscular AAV1.CMV.FS344. Four of six walked farther on a 6-minute walk test. No adverse effects were reported in that study.
  • Catalog “follistatin-344” is not that vector. There is no FDA-approved follistatin drug. WADA lists follistatin under S4.3.

Educational only. Not medical advice, not a protocol, and not a reason to treat a research protein as gene therapy.

What is follistatin-344 actually?

Follistatin is a secreted binding protein, not a growth-hormone secretagogue and not an androgen. It sits on TGF-beta ligands, including myostatin and activin, and keeps them from activating activin receptor type IIB (ActRIIB) on muscle (Lee and McPherron, 2001).

The gene is alternatively spliced. FS344 includes a C-terminal acidic tail and is cleaved to circulating FS315.

The other common product, FS288, sticks to heparin on tissue. FS315 has about a 10-fold lower affinity for activin than FS288. Gene-therapy groups used FS344 to get a serum isoform with less spillover into the pituitary activin-inhibin axis that controls FSH (Al-Zaidy et al., 2015).

If “peptide” still means a short amino-acid chain, start with Peptides 101. FS344 is a several-hundred-residue glycoprotein. Calling it a peptide is catalog language, not size language.

How does blocking myostatin grow muscle?

In 1997, McPherron, Lawler, and Lee knocked out GDF-8 in mice. Null animals were larger. Individual muscles weighed 2 to 3 times those of wild-type littermates, from both more fibers and bigger fibers (McPherron et al., 1997).

That is the brake. Follistatin is one way to sit on it.

Lee and McPherron then made muscle-specific transgenics that overexpressed follistatin, the myostatin propeptide, or a dominant-negative ActRIIB. Each line put on muscle in the same neighborhood as the knockout (Lee and McPherron, 2001).

Follistatin is not only a myostatin sponge. When Lee put a follistatin transgene onto a myostatin-null background, Mstn-/- mice carrying it had about four times the muscle mass of wild type.

Other TGF-beta ligands, likely activins, still limit growth after myostatin is gone (Lee, 2007). That is why FS344 is more interesting than a pure anti-myostatin antibody, and why isoform choice mattered.

What did the gene therapy studies actually measure?

The fitness conversation skips the delivery method. The papers did not.

Study Model What was delivered What happened
McPherron 1997 GDF-8 knockout mice Germline gene deletion Individual muscles 2 to 3 times wild-type weight
Lee 2007 Mstn-/- plus follistatin transgene Lifelong overexpression About 4 times wild-type muscle mass
Haidet 2008 Normal and mdx mice One-time AAV of myostatin inhibitors Mass and strength gains lasting more than 2 years
Kota 2009 Cynomolgus macaques Intramuscular AAV1-FS344 Durable increases in size and strength
Mendell 2015 6 men with Becker MD AAV1.CMV.FS344 into both quadriceps 4 of 6 improved 6-minute walk distance

Haidet’s group showed a single postnatal AAV injection of myostatin-inhibitor proteins could raise mass and strength for more than two years in mice, including aged and dystrophic animals (Haidet et al., 2008). That is continuous local production, not a short plasma spike.

Kota then took AAV1-FS344 into cynomolgus macaque quadriceps. The isoform was chosen because it affects skeletal muscle with only minimal effects on nonmuscle cells.

They put on size and strength, without abnormal key-organ changes reported in that study (Kota et al., 2009).

The human study is small and specific. Mendell injected AAV1.CMV.FS344 into both quadriceps of six ambulatory men with Becker muscular dystrophy (Mendell et al., 2015).

  • Cohort 1 (3 x 10^11 vg/kg/leg): two patients gained 58 m and 125 m on the 6-minute walk test. One did not change.
  • Cohort 2 (6 x 10^11 vg/kg/leg): two patients gained 108 m and 29 m. One did not improve.
  • Safety in this trial: no adverse effects were encountered. Biopsies showed less endomysial fibrosis, fewer centrally nucleated fibers, and more hypertrophy at the higher dose.

That is a proof-of-principle gene-transfer study in a dystrophin-deficient disease, not a hypertrophy trial in trained lifters. Four of six walk-test responders is encouraging biology. It is not a gym protocol.

Is the research vial the same experiment?

No.

The papers delivered a viral vector that kept FS344 being transcribed in muscle. A research chemical labeled follistatin-344 is, at best, a bolus of recombinant protein with no AAV cassette behind it.

WADA already treats black-market FS-344 and FS-315 as real doping products. A funded study noted that no approved follistatin pharmaceutical exists, and that two groups of follistatins are sold on the black market (WADA project 19C09CR). Identity is a COA question.

The provider directory and price index cover pharmacy and research-chemical markets, not gene-therapy vectors. Handling education lives on the reconstitution guide and the calculator. Neither turns a glycoprotein into AAV1-FS344.

Related reading: IGF-1 LR3, PEG-MGF, and peptides vs. steroids.

What does WADA say?

Tested athletes do not get a naming loophole. The 2026 Prohibited List, section S4.3, bans agents that prevent ActRIIB activation, including myostatin-binding proteins such as follistatin, neutralizing antibodies, and decoy receptors such as ACE-031 (WADA, 2026).

The class is prohibited at all times. Clinicians who talk about myostatin as a disease target, not a cycle, are the voices worth more than forum charts. That is the filter behind our experts page.

Myostatin is a real brake, and follistatin is a real ligand trap. None of that is evidence that a research vial recreates a transgene.

Bottom line: Follistatin-344 earned its reputation in knockout mice, macaques, and a six-person AAV trial, not in a reconstituted peptide. If the paper used a virus, the vial is not the paper.

Educational only. Not medical advice. Not a recommendation to use research-grade follistatin or any gene-therapy vector.