TB-500

Thymosin Beta-4 fragment

TB-500 is a synthetic fragment of thymosin beta-4 (Tβ4), a naturally occurring peptide that regulates actin — the protein scaffolding cells use to move. It's one of the two flagship 'healing peptides' (alongside BPC-157) and is researched for soft-tissue, tendon, muscle and cardiac repair, plus flexibility and recovery. This guide covers how TB-500 actually works, what the research does and doesn't show, how it's referenced in the literature, its safety and legal status (including its WADA ban in sport), and how it compares to and stacks with BPC-157.

How TB-500 works

TB-500's core mechanism is actin regulation. It's the main G-actin-sequestering peptide in cells: it binds monomeric actin and holds a ready reservoir that cells can rapidly deploy to build the filaments they need to migrate. That matters because healing depends on fibroblasts, keratinocytes and endothelial cells physically moving into injured tissue — and Tβ4 has been shown to stimulate keratinocyte migration 2–3 fold in research models.

Beyond actin, TB-500 promotes angiogenesis (new blood-vessel growth, partly via VEGF), exerts anti-inflammatory effects through NF-κB modulation, and appears to recruit stem/progenitor cells to damaged areas. The combination — cell migration + blood supply + reduced inflammation — is why it's studied so broadly across wound, tendon, muscle and heart tissue.

What the research shows — benefits by use case

Wound & soft-tissue healing. Animal models show faster wound closure, angiogenesis and keratinocyte migration — the best-established effect.

Tendon, muscle & flexibility. TB-500 is popular among athletes for recovery and a subjective sense of improved flexibility/mobility, thought to relate to its systemic cell-migration effects. Human evidence here is anecdotal rather than trial-proven.

Cardiac repair. In animal models, thymosin beta-4 reduced infarct size and improved cardiac function after heart attack — a striking preclinical signal that continues to drive research interest.

Honesty check: as with BPC-157, most TB-500 evidence is preclinical (animal/cell). It's mechanistically compelling but not the same as proven human efficacy.

The honest state of human evidence

The most advanced human work on thymosin beta-4 is topical, not injectable: the ophthalmic formulation RGN-259 reached Phase II trials for dry-eye and corneal healing with encouraging results. For the injectable, systemic use most people are interested in, robust human trial data is limited — the enthusiasm is built largely on animal studies plus athlete anecdote. Treat it as a promising research compound, not a proven therapy.

TB-500 dosing (research reference)

The research and practitioner literature commonly references TB-500 dosed weekly, often with a higher 'loading' phase for several weeks followed by a lower maintenance dose, reflecting its long systemic action (it doesn't need daily dosing the way BPC-157 often is). It's supplied as a lyophilized powder and reconstituted with bacteriostatic water — the reconstitution calculator above converts a research amount into syringe units. This summarizes existing references and is not dosing advice or a recommendation for human use.

Safety & side effects

In animal studies and anecdotal human reports, TB-500 is generally described as well tolerated, with few acute effects noted. But long-term human safety data is lacking, and — as with any peptide that promotes angiogenesis and cell migration — the mechanism is worth understanding rather than assuming it's risk-free. Research-market purity varies widely; a Certificate of Analysis and third-party testing matter.

BPC-157 vs TB-500 — and stacking

BPC-157 and TB-500 are the classic healing-peptide duo, and they work through different mechanisms: BPC-157 leans toward local vascular and gut repair (VEGF, nitric oxide), while TB-500 works more systemically via actin-driven cell migration and flexibility. Because their pathways are complementary, they're frequently run together — the basis of the popular Wolverine blend.

Legal & regulatory status

TB-500 is not FDA-approved and is sold as a research compound (research use only). Notably, thymosin beta-4 / TB-500 is on the WADA Prohibited List — it is banned in competitive sport at all times — so athletes subject to drug testing should be aware. Its compounding status, like other peptides, remains under ongoing FDA review. Follow the laws and rules that apply to you.

Where to get TB-500

TB-500 is sold in 3 forms. They are not interchangeable — the dose and the route differ, so pick the one this page describes unless you know why you want another.

TB-500 reconstitution calculator

The evidence for TB-500

Graded by what exists behind each claim.

Human clinical evidence

• No randomized human trials — a research compound with no commercial route to funding one, so the correlative and theoretical tiers are the evidence base rather than a consolation prize.

📊 Correlative data

• Thymosin beta-4 — the full protein TB-500 is a fragment of — reached phase 2 human trials for dry eye and for pressure ulcers, and was tolerated. Those trials tested the parent molecule by a different route for different endpoints, so they establish that the biology is safe to touch in humans and very little else.
• Widely used alongside BPC-157 for soft-tissue injury; the reported pattern is slower onset and a broader, more systemic feel than BPC.

🧪 How the mechanism reads

• An actin-binding fragment. Its mechanism is cell migration — it promotes the movement of the cells that do repair work into the site that needs them, which is a genuinely different mechanism from BPC-157's angiogenic one and the reason the two are stacked rather than swapped.
• The same pro-migratory, pro-angiogenic signaling carries the same unanswerable long-term question as BPC-157.

Why an empty tier is not a verdict →

How to read these tiers: they say how much human evidence exists, not how well something works — and ✗ flags harm, never a disappointing trial. How the evidence tiers work →

What TB-500 actually does

TB-500 is LKKTETQ, usually sold with an acetyl cap as Ac-Leu-Lys-Lys-Thr-Glu-Thr-Gln. Seven residues, cut out of the middle of a 43-residue protein. Almost everything interesting about this compound follows from asking which seven, and what got left behind.

Where the fragment sits. Mature human thymosin beta-4 is 43 residues, and counting along it from the recorded parent sequence in this site's own sequence file puts LKKTETQ at positions 17 to 23. That is the central actin-binding motif, and the authoritative description of the parent protein's active sites credits the short sequence containing LKKTETQ — the actin-binding domain at residues 17 to 23 plus one additional amino acid — with promoting angiogenesis, wound healing and cell migration Sosne 2010.

And here is what the same paper says about the parts that are not in the vial. Thymosin beta-4 has several active sites with distinct functions. The amino-terminal 4 residues, Ac-SDKP, generally block inflammation and reduce fibrosis. A 15-residue amino-terminal site that includes Ac-SDKP promotes cell survival and blocks apoptosis Sosne 2010. Counting from the same parent sequence, SDKP is residues 1 to 4. Neither of those sites is inside LKKTETQ. So the anti-fibrotic and anti-apoptotic activities that this compound is most often bought for — less scar tissue, protection of injured cells — belong to a different fragment of thymosin beta-4, and TB-500 does not contain it. That sentence is checkable against the sequence and it appears on no vendor page.

The actin question, which is the mechanism everyone quotes and almost nobody has read. Thymosin beta-4 binds G-actin in a 1:1 ratio and prevents its aggregation into F-actin by sequestration, and single-residue changes in the N-terminal stretch 1 to 22 alter that interaction Zoubek 2007. Two results from that work decide how much of the parent's mechanism the fragment can carry. First, a thymosin beta-4 variant missing residues 17 to 23 shows no interaction with G-actin at all. The motif is necessary. Second, the variant with lysines 14, 16, 18 and 19 replaced by alanine forms a complex that is 15 times less stable. Two of those four lysines, K18 and K19, are inside LKKTETQ. Two, K14 and K16, are outside it. So the fragment carries the motif that is required for actin binding and roughly half of the residue set that makes the binding strong.

Necessary is not sufficient, and no published experiment has asked whether the heptapeptide alone sequesters actin. The deletion study establishes that removing LKKTETQ destroys binding; it does not establish that LKKTETQ on its own reconstitutes it. Every actin-sequestration number in the literature was measured on the intact 43-residue protein or on point mutants of it. Describing TB-500 as an actin-sequestering peptide is therefore an inference from a deletion, not a measurement on the molecule in the vial.

The electrostatics change completely when you cut, and this is computable rather than arguable. Run the parent and the fragment through this site's peptide chemistry: acetylated thymosin beta-4 has an isoelectric point of 4.50 and a net charge of about −3.00 at blood pH, a clear polyanion, with a mass near 4963.5 Da. Acetylated LKKTETQ has an isoelectric point of 7.36, a net charge of about 0.00 at blood pH, and a mass of 889.0 Da. Without the acetyl cap the same 7 residues are a cation: isoelectric point 9.54, net charge about +0.94. Three molecules sold under one name, with three different charge states at the pH of plasma, and a certificate of analysis quoting purity says nothing about which one arrived.

Cell, rodent, human — and where it stops

Step one, in a tube. The actin work above is protein biochemistry with purified G-actin and bacterially expressed thymosin beta-4 variants Zoubek 2007. It is the cleanest evidence in this compound's file and it is about the parent, not the fragment.

Step two, in rodents, and this is the one study that actually used the fragment. Thymosin beta-4 and a synthetic peptide containing its actin-binding domain promoted dermal wound repair in db/db diabetic mice and in aged mice Philp 2003. That is a real result for the LKKTETQ-containing sequence and it deserves stating plainly. It is also a dermal wound-healing model: a skin wound, treated locally, in an impaired-healing mouse. It is not a tendon, not a systemic injection, and not a healthy animal.

Step three, in humans, topically, twice. A phase 2 randomized double-masked placebo-controlled trial of thymosin beta-4 eye drops at 0.1%, dosed 6 times daily for 28 days in 9 patients with severe dry eye, reported at day 56 a 35.1% reduction in ocular discomfort against vehicle, P equal to 0.0141, and a 59.1% reduction in total corneal fluorescein staining, P equal to 0.0108 Sosne 2015. Twelve treated eyes against 6 control eyes. Real, randomized, tiny, and topical to the ocular surface.

The venous ulcer study that gets cited alongside it is worth reading before citing. The paper describes a double-blind, placebo-controlled, dose-escalation design in 3 sequential groups of 24 patients across 10 sites in Italy and Poland, 72 patients treated topically for 84 days — and it reports that the study is ongoing and 21 patients have been enrolled so far Guarnera 2007. It is a protocol description, not a result. Anybody citing it as a positive phase 2 in wound healing has cited a study design.

Step four, in humans, systemically, once. A first-in-human phase 1 of recombinant human thymosin beta-4 gave single intravenous doses across 7 cohorts and 54 healthy subjects, at 0.05, 0.25, 0.5, 2.0, 5.0, 12.5 and 25.0 micrograms per kilogram, then dosed 30 more subjects daily for 10 days at 0.5, 2.0 and 5.0 micrograms per kilogram, with 28 days of observation Wang 2021. Adverse events were mild to moderate; there were no dose-limiting toxicities and no serious adverse events; maximum concentration and exposure rose with dose and there was no obvious accumulation over 10 days.

Now put the doses side by side, because this is the comparison that is never made. The highest single systemic dose ever given to a human in a registered study of this biology is 25.0 micrograms per kilogram Wang 2021 — roughly 1.75 milligrams in a 70-kilogram adult, intravenously, under monitoring. This site's own protocol for TB-500 is 2 to 5 milligrams, subcutaneously, once or twice a week. The community dose of the fragment therefore sits at or above the ceiling dose of the parent protein's first-in-human trial. Different molecule, different route, no monitoring.

The obstacles, one at a time. (1) Every human study used the 43-residue parent protein, never the heptapeptide. There are 0 published human studies of LKKTETQ by any route at any dose. (2) Two of the three human studies were topical, to an eye surface and to an open ulcer, where the peptide is applied directly to the tissue it is meant to act on and systemic pharmacokinetics are irrelevant. (3) The one systemic human study was intravenous, at microgram-per-kilogram doses, in healthy volunteers, with safety rather than efficacy as its endpoint. (4) The one animal study that used the fragment used a dermal wound in an impaired-healing mouse Philp 2003, which is a different tissue and a different problem from a chronic tendinopathy in a training adult. (5) The activities most often claimed for this compound — less fibrosis, protection of injured cells — are credited to the N-terminal sites of the parent Sosne 2010, which the fragment does not contain.

TB-500 pharmacokinetics — how much of it actually gets in

This site's card says roughly 2 to 3 hours in plasma, with tissue effects lasting days and dosing once or twice a week. That is a real tension and it is worth resolving rather than restating.

What degrades it. Ac-LKKTETQ is 7 residues, 889.0 Da, with an acetyl cap at the N-terminus and a free carboxylate at the C-terminus. The cap blocks aminopeptidase N, which is the fastest route of attack on a short peptide and is abundant on the vascular endothelium a subcutaneous dose has to cross. The free acid leaves carboxypeptidases an open handle at the other end, and there are no D-amino acids anywhere in the sequence to interrupt endopeptidase recognition. So the molecule is protected at one end and undefended at the other, and the expected plasma residence for something in that class is minutes to a small number of hours.

What has actually been measured, and on what. There are 0 published pharmacokinetic measurements of LKKTETQ in any species. The only human pharmacokinetic data in this whole biology belongs to the 43-residue parent protein given intravenously, where maximum concentration and exposure rose with dose across 0.05 to 25.0 micrograms per kilogram and 10 days of daily dosing produced no obvious accumulation Wang 2021. A 4963.5 Da protein and an 889.0 Da peptide do not share a clearance mechanism, so that number constrains nothing about the fragment.

How the minutes-versus-weeks gap can be honest rather than hand-waved. If this peptide's action is to start something — cell migration and angiogenesis are the activities credited to the LKKTETQ-containing sequence Sosne 2010 — then the peptide does not need to be present while the process runs. Endothelial and progenitor cells migrating into a repair site take days, and they do it whether or not the trigger is still in the blood. That is a coherent explanation for a short half-life and a weekly dose, it is the only one the mechanism supports, and it is a hypothesis: no study has shown that a single exposure to this fragment initiates a process that continues after clearance.

The oral barrier. Swallowed, an acetylated heptapeptide with a free C-terminus meets gastric acid, then pancreatic proteases, then brush-border peptidases, and anything absorbed intact faces hepatic first-pass extraction. No oral bioavailability figure has ever been published for it. The nasal preparation this site lists is subject to the same objection in a milder form: the nasal epithelium carries its own peptidases, and there is no published nasal pharmacokinetic study of this peptide either.

The injectable comparator. Subcutaneous injection removes the gut and the first-pass liver and is the route essentially all reported use takes. It does not remove plasma peptidases, and it introduces an absorption phase whose rate has never been measured for this molecule, which means the 2-to-3-hour figure on the card is an expectation rather than an observation.

What would have to be true, and how you would know it was not

Three predictions. The first says the marker most often suggested for this compound should not move, which is the opposite of what the shared block implies.

1. hs-CRP should not change, and a change means look elsewhere. Draw hs-CRP (High-Sensitivity C-Reactive Protein) at baseline and at 6 to 8 weeks. The prediction is no meaningful movement, for a reason specific to the injury this compound gets bought for: an isolated tendon or soft-tissue injury usually does not raise systemic C-reactive protein in the first place, so there is no elevation for a repair peptide to reduce. A fall from a normal baseline is regression to the mean. A rise is worth chasing, because it points at an infection, a dosing-site reaction, or something else in the stack — and an ESR alongside it separates an acute process from a chronic one. This is a prediction that makes the compound look less impressive and it is what the biology actually implies.

2. The falsification a single person can run: a within-person crossover, and name the thing that will fool you. Chronic tendinopathy improves on its own over months, and almost nobody starts a peptide without also starting or intensifying rehabilitation in the same week. So the design that means something is a crossover with a washout: 6 weeks on, 6 weeks off, 6 weeks on, with the loading program held constant throughout and a numeric pain-on-loading score recorded at the same time of day. What will fool you is the natural history plus the rehab you started at the same time; a single on-period cannot distinguish them and a crossover can. How long before it means anything: 6 weeks per arm, because tendon adaptation is slower than that of muscle.

3. The identity check, and it is unusually decisive here. Acetylated LKKTETQ has a mass of 889.0 Da; the intact acetylated parent protein is 4963.5 Da. Those are not close, and any mass spectrometer separates them instantly. Vendors use the name TB-500 for both molecules. If you can obtain a mass on a lot, that single number tells you which product you have, and it is the only measurement on this page that resolves the question completely. A CBC and a urinalysis are the sensible background draws for anyone injecting an unverified peptide weekly, and neither will tell you anything about the peptide itself.

What nobody has tested yet

Five experiments, and the first one is a gap so large it is easy to miss.

1. No human has ever been given LKKTETQ in a published study. Not intravenously, not subcutaneously, not topically, at any dose, in any indication. Every human datum in this compound's file belongs to the 43-residue parent protein Sosne 2015 Guarnera 2007 Wang 2021. That is the single most important fact about TB-500 and it is not stated anywhere it is sold.

2. Nobody has asked whether the heptapeptide sequesters actin. The published experiment shows that deleting residues 17 to 23 abolishes G-actin binding Zoubek 2007, which proves the motif is necessary. The complementary experiment — put Ac-LKKTETQ into the same G-actin complex-stability assay and ask whether it binds on its own — uses the same purified reagents and the same instrument, and has never been reported. Given that two of the four stability-critical lysines lie outside the fragment, the answer is not obvious in either direction.

3. Nobody has run a mass spectrum on a vendor vial. An 889.0 Da heptapeptide and a 4963.5 Da protein are trivially distinguishable, and the market sells both under one name. Three lots from three vendors would produce the first compositional data this product has ever had, and would also reveal whether the acetyl cap is present — which decides whether the molecule is electrically neutral or a cation at blood pH.

4. Nobody has compared the fragment against Ac-SDKP for the endpoint people actually buy it for. The anti-fibrotic activity of thymosin beta-4 is credited to its N-terminal site, not to the actin-binding one Sosne 2010. A head-to-head in any fibrosis model, with both fragments and the intact protein as arms, would tell the repair-peptide market whether it has been buying the wrong 7 residues for 20 years. That comparison has not been published.

5. Nobody has measured it in plasma after a subcutaneous dose. No assay for LKKTETQ in human plasma has been published, so there is no time to peak, no maximum concentration and no half-life for the route everyone uses. Doping-control laboratories have the hardest part of this problem solved already, because detecting the peptide is exactly their job; the pharmacokinetic curve is a by-product nobody has chosen to publish.

TB-500 — its own safety story, not its class's

The class block on this page is the shared repair-peptide warning about angiogenesis, and it is worth saying that on this compound the concern is less generic than usual, for a specific reason.

1. The angiogenesis claim and the angiogenesis worry are the same sentence. The activity credited to the LKKTETQ-containing sequence is, verbatim, promotion of angiogenesis, wound healing and cell migration Sosne 2010. On most repair peptides the theoretical oncological concern is an extrapolation from a class. Here it attaches to the published description of the exact fragment being sold, and new vessel growth plus cell migration is a fair summary of what a tumor needs to progress. There is no evidence this peptide causes or accelerates cancer in anything, and there is also no study in which that could have been observed, because no human has been given the fragment in a published trial. The honest position is that the mechanism and the worry are the same mechanism, and nobody has looked.

2. The dose people use has no safety precedent, and the comparison is stark. The registered first-in-human study of the parent protein topped out at 25.0 micrograms per kilogram intravenously in 54 healthy subjects, with 30 more taking up to 5.0 micrograms per kilogram daily for 10 days Wang 2021. The community protocol is 2 to 5 milligrams subcutaneously, once or twice a week, indefinitely. Even allowing for the different molecule and the different route, nothing published describes what happens above that phase 1 ceiling, and the fragment's own safety file is empty.

3. What the existing safety data is actually about. Mild to moderate adverse events, no dose-limiting toxicities and no serious adverse events, in an intravenous phase 1 of a recombinant protein Wang 2021; a 9-patient topical eye trial Sosne 2015; and an ulcer study whose published paper reports a design rather than an outcome Guarnera 2007. Those are the grounds for the widespread claim that this compound is well tolerated. They are reasonable grounds for the protein, by those routes, at those doses. They are not evidence about a weekly subcutaneous heptapeptide of unverified acetylation state.

4. The specific practical hazard is the vial, not the molecule. Because the fragment and the parent protein are sold under one name and differ 5.6-fold in mass, a person reconstituting ‘5 mg of TB-500’ has no way of knowing whether they are measuring out a heptapeptide or a protein, and the two have different molar amounts, different charge at blood pH, and different clearance. That ambiguity is unusual even in this market, and it is the reason the mass number in the identity check above is worth more than any certificate.

Sources read for this page

• Sosne G, Qiu P, Goldstein AL, Wheater M. Biological activities of thymosin beta4 defined by active sites in short peptide sequences. FASEB Journal 2010 · PMID 20179146
• Zoubek RE, Hannappel E. Influence of the N terminus and the actin-binding motif of thymosin beta4 on its interaction with G-actin. Annals of the New York Academy of Sciences 2007 · PMID 17495251
• Philp D, Badamchian M, Scheremeta B, Nguyen M, Goldstein AL, Kleinman HK. Thymosin beta 4 and a synthetic peptide containing its actin-binding domain promote dermal wound repair in db/db diabetic mice and in aged mice. Wound Repair and Regeneration 2003 · PMID 12581423
• Sosne G, et al. Thymosin beta4 significantly improves signs and symptoms of severe dry eye in a phase 2 randomized trial. Cornea 2015 · PMID 25826322
• Guarnera G, et al. Thymosin beta-4 and venous ulcers: clinical remarks on a European prospective, randomized study on safety, tolerability, and enhancement on healing. Annals of the New York Academy of Sciences 2007 · PMID 17495250
• Wang X, Liu L, Qi L, Lei C, Li P, Wang Y, Liu C, Bai H, Han C, Sun Y, Liu J. A first-in-human, randomized, double-blind, single- and multiple-dose, phase I study of recombinant human thymosin beta4 in healthy Chinese volunteers. Journal of Cellular and Molecular Medicine 2021 · PMID 34346165

TB-500 — safety, predicted from mechanism

TB-500 — interference & stacking

Bloodwork to run alongside TB-500

Baseline first, then again at 8–12 weeks.

The Inflammation Deep Dive panel covers these in one order — 10 markers, $248.35 with the discount applied.

Check results you already have → · All 103 markers A–Z

TB-500 — frequently asked questions

References & further reading

1. Thymosin beta-4 and its actin-binding domain in dermal wound repair (Philp 2003)
2. Thymosin beta-4 in severe dry eye — phase 2 randomized trial (Sosne 2015)
3. First-in-human phase I of recombinant human thymosin beta-4 (Wang 2021)

TB-500 inside a finished plan

One arm of 2 Protocol Blueprints, free to read in full.

What TB-500 is used for

TB-500 appears under 3 goals in the goal router.

Related Healing & Recovery compounds

← Explore the full Protocol Vault

↑ Back to on this page