Thymosin Beta-4 (TB-500): Actin Binding, Structure and Recent Studies

Thymosin beta-4 is a 43-amino-acid peptide of the beta-thymosin family, first sequenced from calf thymus in 1981 by Low, Hu and Goldstein. Thymosin beta-4 occurs in the cytoplasm of nearly all nucleated vertebrate cells and is encoded in humans by TMSB4X on the X chromosome, with a Y-linked paralogue designated TMSB4Y.

Its cellular abundance is what distinguishes it from most signalling peptides. Weber and colleagues measured concentrations of 560 micromolar thymosin beta-4 against 280 micromolar monomeric actin inside resting human platelets, meaning the peptide is present at roughly twice the molar amount of the species it binds. At those concentrations it is a bulk buffering component of the cytoskeleton rather than a trace regulator.

The name is a historical artefact. The molecule was isolated from a crude thymic preparation known as fraction 5 and characterized as a thymic hormone, and only later was the same chain shown to be indistinguishable from Fx, the actin-sequestering activity that cell biologists had been tracking independently. Both literatures describe one molecule.

UniProt annotates the human entry with N-acetylserine at position 2 of the translated sequence, phosphothreonine at positions 23 and 34, phosphoserine at positions 2 and 31, and several acetyl-lysine sites. Synthetic and recombinant preparations reproduce the backbone; whether they reproduce the modification pattern depends entirely on the production route.

Thymosin beta-4 binds a single monomer of G-actin and blocks its addition to a filament. Xue and colleagues measured a dissociation constant of 0.32 plus or minus 0.02 micromolar by fluorescence anisotropy in 2014. Earlier work by Weber and colleagues gave 0.4 to 0.7 micromolar against platelet actin and 2 to 3 micromolar against skeletal muscle actin.

Nucleotide state governs the interaction. Reported constants cluster between 0.1 and 3.9 micromolar for ATP-loaded actin and between 80 and 100 micromolar for ADP-loaded actin, a roughly hundred-fold discrimination. The consequence is a reservoir of polymerization-competent ATP-actin held in reserve, with spent ADP-actin released preferentially.

Two crystal structures published by Xue, Leyrat, Grimes and Robinson in 2014 resolved the geometry, one at 2.3 angstroms using a Pichia actin hybrid and one at 2.0 angstroms using rabbit skeletal muscle actin with a Cobl peptide. Thymosin beta-4 grips the monomer with two short helices: an N-terminal helix spanning aspartate-5 to lysine-11 that contacts the barbed face, and a C-terminal helix from lysine-31 to glutamine-39 that bridges the nucleotide cleft. Total buried interface was about 1,627 square angstroms, of which the C-terminal helix alone contributed 392.

The same structures explain how actin is handed onward to profilin. The C-terminal helix of thymosin beta-4 stabilizes a closed, flattened nucleotide cleft, whereas profilin favours an open cleft, so the two ligands stabilize incompatible conformations. Affinity of thymosin beta-4 for a profilin-actin complex fell about 25-fold relative to free actin, and a ternary intermediate allows transfer without a fully dissociated monomer.

In free solution the peptide has no fixed fold. Czisch and colleagues showed by NMR spectroscopy in 1993 that thymosin beta-4 lacks a uniquely folded conformation in water and carries only transient helical preference; defined helices covering residues 5 to 16 and 31 to 39 appear when the peptide is placed in helix-promoting solvent such as trifluoroethanol. Structure is induced by the binding partner, not carried into the encounter.

Thymosin beta-4 and TB-500 are not the same chemical entity, although catalogue listings routinely conflate them. Thymosin beta-4 is the full 43-residue chain of 4963 g/mol. TB-500 material has been analytically identified as the acetylated heptapeptide Ac-LKKTETQ, corresponding to positions 17 through 23 of the parent chain and weighing roughly 889 g/mol, under a fifth of the parent mass.

The distinction matters when reading mechanism papers. Crystallography, anisotropy titrations, the profilin exchange work, and essentially all the cardiac, renal and ocular preclinical literature summarized below used full-length protein, whether synthetic or recombinant. Those measurements describe the 43-residue chain. A seven-residue fragment lacks both terminal helices resolved in the crystal structures, so the actin geometry established for the parent does not transfer to it automatically.

Some published work does bridge the two, having localized angiogenesis-associated activity in specific assays to the actin-binding motif that the short fragment reproduces. That is a narrower claim than equivalence, and the two entities carry different CAS numbers, different formulae and different PubChem records.

Thymosin beta-4 is the physiological precursor of Ac-SDKP, an N-acetylated tetrapeptide corresponding to its own first four residues, serine-aspartate-lysine-proline. Release is a two-enzyme process: meprin-alpha first cuts the 43-residue chain into intermediates shorter than 30 residues, then prolyl oligopeptidase liberates the tetrapeptide, which prolyl oligopeptidase cannot excise from the intact chain directly.

Clearance runs through a well-characterized enzyme. Ac-SDKP is hydrolysed by the N-terminal catalytic domain of angiotensin-converting enzyme, and circulating concentrations of the tetrapeptide rise roughly five-fold when that enzyme is inhibited. This places a fragment of thymosin beta-4 inside a pathway that ordinary cardiovascular pharmacology already perturbs.

Wang, Jia and Zhang reviewed this axis in 2022 across hepatic, renal, cardiac and pulmonary fibrosis models, describing antifibrotic activity attributed to the tetrapeptide in each organ system. For anyone reading thymosin beta-4 papers, the practical implication is that observed effects may belong to the intact chain, to Ac-SDKP, or to both, and few study designs separate them.

Recent thymosin beta-4 research has concentrated on ocular surface repair, cardiac ischaemia, and inflammatory injury models, with a parallel strand of protein-engineering work aimed at the peptide's short circulating lifetime. Six studies published between November 2025 and 2026 illustrate the current shape of the field.

In ophthalmology, Nguyen and colleagues reported in Investigative Ophthalmology and Visual Science in 2025 that fusing two copies of the peptide into a tandem construct with two G-actin binding sites raised viability and migration of human telomerase-immortalized corneal epithelial cells above the single-copy peptide, and produced a thicker, continuous epithelial layer in an alkali-injured murine cornea. A 2026 report from Ebrahim and colleagues in the same journal examined adjunctive use alongside ciprofloxacin in Pseudomonas aeruginosa keratitis in C57BL/6 mice, reporting restoration of corneal nerve density toward uninfected control levels.

In cardiac work, Zhang and colleagues published in Cardiovascular Research in 2025 a study pairing a mouse ischaemia-reperfusion model with a randomized, placebo-controlled human trial enrolling 96 people after ST-segment elevation myocardial infarction. In mice, a seven-day course limited cardiac dysfunction and fibrosis assessed 28 days after surgery, with RNA sequencing implicating ErbB2 and Raf1 signalling. In the human arm, infarcted area at 90 days was reduced in the active group, but the overall difference did not reach statistical significance, a result the authors report plainly.

Three 2026 papers extend the inflammatory-injury literature. Ouyang and colleagues reported in Clinical Science that circulating thymosin beta-4 fell in lipopolysaccharide-induced sepsis-associated acute kidney injury and that supplying the peptide suppressed phosphorylation of JNK1/2, p38 MAPK and ERK1/2, with pharmacological MAPK activation partly reversing the effect in HK-2 cells. Ye and colleagues reported in Biomolecules that recombinant human thymosin beta-4 raised survival and lowered circulating pro-inflammatory cytokines in endotoxaemia models, with TLR4 and NF-kappa-B suppression and transcriptomic downregulation of LPAR3. Li and colleagues reported in the Chinese Journal of Contemporary Pediatrics that in BV2 microglial cells challenged with lipopolysaccharide and nigericin, the peptide lowered NLRP3, cleaved caspase-1, GSDMD-N, phospho-STING and phospho-IRF3 protein levels along with mitochondrial reactive oxygen species.

The musculoskeletal evidence is thinner than its popular profile suggests. McGuire, Hughes, Maak and Cushman screened the literature through March 2026 for a scoping review in Applied Sciences and included 80 studies, finding the corpus weighted toward in vitro and mixed designs, concentrated in dermal, vascular and ocular tissue, and sparse in direct musculoskeletal work outside animal models. Biçer and colleagues published one such animal study in 2026, a 32-rat Achilles transection model in which the short fragment group showed significantly higher maximum load to failure at four weeks.

Thymosin beta-4 is obtained either by solid-phase peptide synthesis or by recombinant expression in Escherichia coli, and the recombinant route is the one with published process figures. Li and colleagues expressed the peptide in soluble form as a DsbA and hexahistidine fusion, reaching about 50 grams per litre dry cell weight with the fusion at 40 percent of total protein.

Downstream processing in that report used a five-step column sequence combining thermal denaturation, nickel-resin affinity capture and preparative high-performance liquid chromatography, returning material above 98 percent purity with endotoxin, host cell protein and residual DNA inside the limits set by the Chinese Pharmacopoeia. Those are the most concrete purity and impurity numbers in the open literature for this peptide.

Identity confirmation for a 4963 g/mol chain is straightforward by electrospray or quadrupole time-of-flight mass spectrometry after desalting, with reversed-phase C18 chromatography under an acetonitrile gradient used for purity. Two checks matter more than they would for a shorter peptide: whether the N-terminal serine is acetylated, which shifts the observed mass by 42 units, and whether methionine-6 has oxidized, which shifts it by 16.

A recurring theme in the 2025 and 2026 engineering literature is that the native chain does not persist long in circulation. The tandem-construct work cited short circulating lifetime and synthesis cost as the explicit motivation, and a separate 2026 report described a thiol-site-specific PEGylated variant built for the same reason. Both are indirect evidence that the unmodified peptide is cleared rapidly.

Published handling data for thymosin beta-4 is thinner than the volume of biological literature suggests. Chemical catalogues list the material as a lyophilized powder with aqueous solubility around 50 mg/mL, equivalent to roughly 10 millimolar at a formula weight of 4963.4, and recommend storage below minus 20 degrees Celsius in the dark under inert gas. These figures are vendor specifications; no peer-reviewed solubility profile across pH or formal shelf-life study was located.

The sequence itself indicates the chemistry to watch. There is exactly one methionine, at position 6, and no cysteine anywhere in the 43 residues. Disulfide scrambling is therefore not available as a degradation route, and methionine sulfoxide formation is the principal covalent liability to monitor, detectable as a 16-unit mass shift by electrospray or quadrupole time-of-flight mass spectrometry.

Two identity checks are worth building into any incoming-material protocol. The first is acetylation state at the N-terminal serine, which separates the native modified chain from an unacetylated synthetic one by 42 mass units. The second is whether the material is the 43-residue chain at all rather than the acetylated heptapeptide, since the two differ by more than 4,000 g/mol and are trivially separated by mass spectrometry despite sharing catalogue names.

Because the peptide is intrinsically disordered when free, circular dichroism of the material in plain aqueous buffer is expected to read as predominantly random coil, and a strongly helical spectrum in water would itself be an anomaly worth investigating. Recombinant preparations reported above 98 percent purity by the five-step process cited here, which is a reasonable published benchmark for comparison against a certificate of analysis.