What the molecule does, where it has been studied, and where the record is thin.

What this page covers

The research record on Thymosin Beta-4 is unusually rich for a peptide that has never been approved for any therapeutic use. Two independent Phase I IV-safety cohorts exist, a Phase III ophthalmic trial produced a statistically significant healing signal, and the rodent literature spans stroke, traumatic brain injury, multiple sclerosis models, peripheral neuropathy, spinal-cord injury, acute kidney injury, pulmonary fibrosis, corneal healing, and cardiac progenitor mobilization. The honest caveat that runs through all of it: the molecule studied in those trials is the 43-amino-acid full-length Tβ4, and TB-500 is a 7-amino-acid synthetic fragment. This page assembles the mechanism first — G-actin sequestration, the AcSDKP cleavage pathway, NF-κB suppression, angiopoietin-1 upregulation — then walks through the animal and human evidence body area by body area, and flags where the rodent signal has and has not translated into larger mammals or people.

Mechanism — G-actin, AcSDKP, NF-κB

Thymosin Beta-4 is the principal cytoplasmic G-actin sequestering peptide in mammalian cells. It binds monomeric actin 1:1 through a central α-helix containing the LKKTET hexapeptide, holding G-actin in a polymerization-incompetent state and creating a mobilizable reservoir that powers rapid actin remodeling during cell migration, wound contraction, and immune-cell trafficking ^[16]. This sequestration shifts the G:F-actin equilibrium and underlies the molecule's role in cytoskeletal repair contexts across organs.

A second mechanism runs through Tβ4's N-terminus. Prolyl-oligopeptidase cleaves the parent peptide to release the N-terminal tetrapeptide AcSDKP — N-acetyl-Ser-Asp-Lys-Pro, also known as Goralatide — which has its own antifibrotic, pro-angiogenic, and hematopoietic-regulating activity. AcSDKP suppresses TGF-β1 / Smad2 phosphorylation, the central profibrotic signaling axis in heart, kidney, and lung. It is cleared almost exclusively by the N-terminal active site of angiotensin-converting enzyme, which is the reason ACE-inhibitor therapy elevates endogenous AcSDKP — a pharmacologic interaction that is a recurring theme in cardiorenal Tβ4 / AcSDKP work ^[20]. Whether the synthetic TB-500 heptapeptide is itself a substrate for prolyl-oligopeptidase cleavage is mechanistically unresolved.

A third arm is anti-inflammatory. Tβ4 binds NF-κB RelA/p65 directly and suppresses TNF-α-driven IL-8 transcription. A fourth is angiogenic: in endothelium and in Schwann cells of the sciatic nerve in db/db diabetic mice, Tβ4 elevates angiopoietin-1 and lowers angiopoietin-2 in a PI3K/Akt-dependent program that restores functional vascular density ^[3]. A fifth is regenerative: in adult mouse hearts, IP Tβ4 at 150 μg every three days reactivates a WT1+/Tbx18+ epicardial progenitor program capable of forming new coronary vessels ^[16]. And a sixth, identified by a 2025 microneedle study, is a previously unknown immune-regulator interaction: Tβ4 binds Vsig4 and IL22Rα2 specifically, both of which are downregulated by injury, expanding the molecule's known target set beyond G-actin and NF-κB ^[18].

The central-nervous-system record

The neurorestorative profile is the most-replicated finding in the Tβ4 literature, and it is consistent across mechanism (delayed dosing rather than acute neuroprotection), endpoint (durable functional recovery rather than acute infarct reduction), and model.

In rats with controlled cortical impact TBI, Xiong and colleagues administered recombinant human Tβ4 at 6 mg/kg intraperitoneally beginning 24 hours after injury and continuing every three days for four additional doses. The intervention reduced hippocampal CA3 neuronal loss, increased BrdU-positive proliferating cells and new oligodendrocytes in the injured cortex and hippocampus, and produced durable improvements in sensorimotor function and Morris water-maze spatial learning that persisted through day 90 ^[1]. The 24-hour start matters: the molecule does not appear to act as a classical neuroprotectant. It acts on the recovery phase.

The same delayed-dosing pattern appears in stroke. Morris and colleagues gave adult rats a single intravenous Tβ4 dose of 3.75 mg/kg 24 hours after embolic middle-cerebral-artery occlusion. Adhesive-removal and modified Neurological Severity Score improvements appeared from day 14 and persisted through day 56, with no change in infarct volume ^[5]. Acute volume — what most classical neuroprotectants target — was not affected. Long-term function was.

In experimental autoimmune encephalomyelitis and cuprizone-diet demyelination, Zhang and colleagues found that daily IP Tβ4 promoted oligodendrocyte progenitor cell proliferation and their maturation into myelinating oligodendrocytes through an EGFR-dependent program; new mature-OL density correlated with EAE clinical score recovery at r = 0.73 (p < 0.05) ^[2]. A foundational mechanistic question — whether Tβ4 can promote remyelination, not merely protect against demyelination — is answered in the affirmative in two independent mouse models.

Spinal-cord injury extends the same pattern. In a rat compression SCI model, Cheng and colleagues administered Tβ4 at 30 minutes, three days, or five days after injury and observed BBB locomotor recovery, preserved oligodendrocytes, myelin basic protein levels 57.8% above saline controls, and reduced lesion cavity area at every dosing window ^[4].

The Alzheimer literature is newer. In 12.5-month-old APP/PS1 mice challenged with systemic LPS, Othman and colleagues found that 5 mg/kg IV Tβ4 prevented the LPS-induced increase in amyloid-plaque burden and attenuated sickness behavior more strongly in APP/PS1 mice than in wildtype littermates ^[6]. A 2025 study by Zeng and colleagues went further: 0.5 μg/mL Tβ4 added to human cerebral organoids derived from familial-AD iPSCs increased neuronal density and lowered intracellular Aβ, while AAV-driven cortical-neuron TMSB4X overexpression in 5xFAD mice rescued neuronal hyperexcitability ^[7]. The organoid-plus-mouse combination identifies Tβ4 as a candidate intervention target for AD, joining the broader CNS pattern.

Peripheral and visceral organs

In db/db type-II diabetic mice, Wang and colleagues administered Tβ4 at 6 mg/kg or 24 mg/kg intraperitoneally and observed restored functional vascular density and regional blood flow in the sciatic nerve, increased intraepidermal nerve fiber density, reversal of diabetes-induced axon-diameter and myelin-thickness reductions, and the angiopoietin-1-up / angiopoietin-2-down PI3K/Akt program in endothelium and Schwann cells ^[3]. The result links Tβ4's angiogenic mechanism to a measurable peripheral-nerve repair phenotype — distinct from its CNS effects but consistent with them.

In a 90-minute ischemia / 3-hour reperfusion rat acute-kidney-injury model, Aksu and colleagues found that Tβ4 administered pre-ischemia or pre-reperfusion reduced caspase-9 activation, matrix metalloproteinase-9 activity, and hyaluronan accumulation, restored renal oxidative-stress and inflammation markers, attenuated tubular histologic injury, and lowered serum urea and creatinine ^[8]. The renal-protective signal in rats has not, as of this writing, been carried into a registered human trial.

In mouse bleomycin-induced pulmonary fibrosis, Yu and colleagues nebulized recombinant human Tβ4 at 20, 100, and 500 μg per dose across early, mid, and late dosing windows. The 100-μg dose was optimal. Hydroxyproline content (the canonical readout of pulmonary collagen accumulation) was reduced, lung function improved, and fibroblast proliferation, migration, and activation were suppressed via TGF-β1 / Smad blunting and EMT inhibition ^[9]. The work is the first demonstration of inhaled Tβ4 in IPF-relevant models.

The corneal data are the most clinically advanced. The foundational mouse work — Sosne and colleagues, 2002 — showed that topical Tβ4 at 5 μg BID accelerated corneal epithelial healing at every measured time point after alkali burn and significantly reduced IL-1β, KC, and MIP-2 chemokine mRNA ^[14]. Two decades later, the same molecule reached Phase III in human neurotrophic keratopathy as RGN-259, the 0.1% Tβ4 ophthalmic solution. In a randomized, placebo-controlled, double-masked trial of 18 patients (NCT02600429), complete corneal healing at day 29 was observed in 60% of treated patients versus 12.5% of placebo (p = 0.066), with statistically significant healing at day 43 (p = 0.036) and durable benefit two weeks after stopping treatment ^[13]. It is the most rigorous published efficacy signal in the entire Tβ4 development program. The same RGN-259 missed the prespecified primary endpoints of the Phase III ARISE-3 dry-eye study.

A 2023 study extends the picture by introducing context-dependence. Hepatic-stellate-cell-conditional deletion of Tβ4 in a CCl4 mouse model REDUCED liver fibrosis — direct evidence that Tβ4 plays a pro-fibrotic role in the hepatic compartment, opposite to its anti-fibrotic role in heart, kidney, and lung ^[21]. Any framing that calls Tβ4 'uniformly regenerative' is incomplete.

Cardiac — the rodent-to-pig translation gap

Smart and colleagues, in a 2007 Nature paper, established the cardiac-progenitor story: adult mouse epicardial progenitor cell (EPDC) mobilization, restored multipotent capacity, and new coronary-vessel formation in the injured adult heart after Tβ4 administered IP at 150 μg every three days ^[16]. The work grounded a substantial body of subsequent epicardial-Tβ4 work and underwrote the clinical hypothesis that systemic Tβ4 could promote cardiac repair after myocardial infarction.

That clinical hypothesis became RGN-352 — an intravenous Tβ4 formulation developed by RegeneRx for post-acute-MI cardiac repair. The Phase II trial (NCT01311518) was designed for approximately 75 patients dosed at 450 mg or 1,200 mg IV daily for three days then weekly for four weeks. Its status on ClinicalTrials.gov is listed as Withdrawn as of September 2021.

The likely reason the program stalled is in plain view in the peer-reviewed record. Wei and colleagues, in a 2016 randomized 24-animal study, administered Tβ4 at 150 μg/kg IV bolus plus maintenance to pigs before or after a closed-chest 90-minute ischemia / 24-hour reperfusion myocardial-injury protocol. Global infarct size measured by TTC and by MRI was not reduced by Tβ4 versus vehicle ^[23]. The rodent cardiac-repair signal did not translate to a large-mammal IR injury model in the way the rodent CNS signals translated to the human eye.

It is worth saying this plainly. The cardiac literature contains a Nature paper showing adult-heart regeneration in mice and a peer-reviewed pig study showing no infarct-size reduction in a model designed to anticipate human Phase II. These two facts coexist.

Human safety — what has been done

Two independent Phase I IV-safety datasets exist for recombinant full-length Tβ4. The first, by Ruff and colleagues, randomized 40 healthy adult volunteers in the United States to single intravenous doses of 42, 140, 420, or 1,260 mg of recombinant Tβ4, with a multiple-dose extension. There were no dose-limiting toxicities and no serious adverse events at doses up to 1,260 mg; reported adverse events were mild to moderate ^[11]. The dataset is the foundational human safety record on which the rest of the clinical program rests.

The second, by Wang and colleagues, enrolled 84 healthy Chinese adults to single IV doses of NL005 recombinant human Tβ4 at 0.05, 0.25, 0.5, 2.0, 5.0, 12.5, and 25.0 μg/kg, plus a multiple-dose 0.5–5.0 μg/kg/day x 10-day extension. Pharmacokinetics were dose-linear in Cmax and AUC, there was no accumulation across the multiple-dose extension, no serious adverse events occurred, and immunogenicity was favorable ^[12]. The two cohorts are separated by a 50-fold difference in tested dose range — Ruff studied a high-dose envelope, Wang a low-dose envelope — and both came out clean.

A third human observational signal comes from sepsis. Belsky and colleagues, studying an emergency-department septic-shock cohort, found that plasma Tβ4 was undetectable while circulating filamentous (F-) actin was elevated ^[10]. The interpretation is that endogenous Tβ4 is consumed during sepsis-associated unregulated actin polymerization, and the authors argued for both a biomarker role and a rationale for testing exogenous Tβ4 supplementation in sepsis. No interventional human sepsis trial of Tβ4 has been published to date.

None of the above involves the synthetic 7-AA TB-500 fragment. No registered human clinical trial of the synthetic fragment has been published. The human safety envelope quoted in research-chemical literature is, in every meaningful respect, the safety envelope of the full-length parent molecule.

Recent work (2023–2025) — direction of travel

The 2023–2025 literature is shifting toward biomaterial and combination delivery rather than free peptide. In 2025, Yu and colleagues developed a HAMA/PLMA dual-photopolymerizable hydrogel loaded with adipose-derived stem-cell exosomes overexpressing Tβ4. Applied to streptozotocin-induced type-1 diabetic mouse wounds, the construct accelerated closure, stimulated CD31+ angiogenesis, increased collagen synthesis, and altered macrophage polarization through PI3K/AKT/mTOR/HIF-1α signaling ^[19]. The same year, He and colleagues built a low-temperature chitosan-sucrose dissolvable microneedle platform delivering 248 μg of Tβ4 per patch, accelerating full-thickness wound closure in mice and identifying Vsig4 and IL22Rα2 — both downregulated by injury — as specific Tβ4-binding immune regulators ^[18]. A 2024 engineered tandem-repeat construct of the Tβ4 actin-binding region (the structural class TB-500 belongs to) accelerated corneal-epithelium wound closure beyond monomeric Tβ4 in rats ^[17], following an earlier 2013 dimeric-Tβ4 demonstration in dermal wounds ^[15]. The direction is multimerization and targeted delivery, not solo IV.

This matters for how the TB-500 record is read. The most active frontier in 2024–2025 Tβ4 research is figuring out how to deliver the molecule better, not how to deliver more of it. The free-peptide IV approach that grounded the original Phase I and Phase II programs is not where the field is investing. That observation alone reframes the entire research-chemical pitch built around free-peptide subcutaneous self-administration.