PTD-DBM: Complete Guide to the Wnt Pathway Hair Regrowth Peptide

The acronym unpacks to Protein Transduction Domain-fused Dishevelled Binding Motif. Two functional halves stitched into one short peptide. The PTD portion is a carrier sequence that lets the molecule cross cell membranes on its own, no needle into the cytoplasm required, no liposome wrapper. The DBM portion is the business end. It's modeled on a stretch of the CXXC5 protein that normally binds to a signaling hub called Dishevelled (Dvl), and it competes with the real CXXC5 for that same binding site.

That's the whole molecule. A delivery tag fused to a competitive inhibitor.

The compound was developed by Kang-Yell Choi's group at Yonsei University in Seoul, with the first major hair-regrowth paper appearing in the Journal of Investigative Dermatology around 2017 ^[1]. The same laboratory has since built out a small family of follow-on compounds, most notably KY19382, a small molecule that hits both CXXC5 and GSK-3β at once. The broader Yonsei research ecosystem has given rise to commercial ventures including CosmeRNA and Epi Biotech.

What it's not: an FDA-approved drug, an IND-stage candidate with publicly registered human trials, or a peptide with peer-reviewed human safety data. As of mid-2026, every published efficacy study runs on mice and cell culture.

Hair follicles aren't static. They cycle. Each follicle on your scalp moves through three phases, anagen (active growth, lasting years), catagen (a brief regression phase), and telogen (rest, lasting months), and then either restarts anagen or, in pattern baldness, fails to.

What drives the restart is, in large part, Wnt signaling.

Wnt/β-catenin is one of the oldest and most conserved signaling cascades in animal biology. The short version: a Wnt protein binds to a receptor on the cell surface, that signal travels through Dishevelled, β-catenin accumulates in the cytoplasm and then the nucleus, and it switches on a set of genes that tell stem cells to proliferate and differentiate. In the hair follicle, the relevant stem cells live in the bulge region, and the relevant tissue is the next generation of hair shaft.

When Wnt signaling fires, follicles enter anagen. When it's suppressed, they stall in telogen.

In androgenetic alopecia, Wnt signaling is suppressed. Follicles miniaturize over successive cycles, shafts thin, and eventually the cycle arrests. The reasons are tangled. Dihydrotestosterone is upstream of much of it, and DHT appears to drive the elevation of an enzyme product called PGD2 that in turn pushes up CXXC5 expression in scalp tissue ^[2]. So balding scalps have more CXXC5 than haired scalps, and CXXC5 is the thing actively holding the Wnt pathway down.

This is the gap PTD-DBM was designed to fill.

CXXC5 is a negative feedback regulator. When Wnt signaling activates, CXXC5 expression rises, and the CXXC5 protein binds to the PDZ domain of Dishevelled ^[3]. That binding event blocks the signal from propagating downstream. It's a brake.

In a healthy follicle entering anagen, the brake comes off when it needs to. In a balding follicle, CXXC5 levels are elevated and the brake stays on.

PTD-DBM's job is to pop CXXC5 off Dvl. The DBM portion of the peptide is essentially a decoy. It binds to the same PDZ pocket on Dvl that CXXC5 binds to, but more avidly, displacing the endogenous inhibitor. Once CXXC5 is off, Dvl resumes signaling. β-catenin accumulates. Wnt target genes switch on. Follicles that were stuck in telogen get a push toward anagen, and in mouse skin, even regions without existing follicles can develop new ones, a phenomenon called wound-induced hair follicle neogenesis, or WIHN.

The Choi lab has shown each step of this chain in mouse and cell models. Pull-down assays show PTD-DBM displacing CXXC5 from Dvl. Reporter assays show downstream Wnt target gene activation. Histology shows follicle neogenesis in treated mouse skin.

The synergy story matters too. Valproic acid, an old anticonvulsant that happens to inhibit GSK-3β (another Wnt brake, working further downstream), produces additive hair regrowth when combined with PTD-DBM in mice ^[4]. KY19382, the small molecule that hits CXXC5 and GSK-3β simultaneously, produces more complete β-catenin rescue than either approach alone ^[5]. The laboratory's working hypothesis: you need to release both brakes for full pathway reactivation.

The drugs that dominate hair loss medicine, finasteride and minoxidil for pattern hair loss, JAK inhibitors for alopecia areata, work nowhere near where PTD-DBM works.

Finasteride is a 5α-reductase inhibitor ^[6]. It blocks the conversion of testosterone to DHT, which in turn lowers PGD2, which downstream may lower CXXC5. So there's a mechanistic thread that connects them. But finasteride acts at the top of the hormonal cascade. PTD-DBM acts at the bottom, on the downstream effector. In principle, that distinction matters for patients whose hair loss isn't primarily DHT-driven, or for patients who can't tolerate systemic anti-androgens.

Minoxidil is a different animal entirely. It opens KATP channels and acts as a vasodilator ^[7], with some sulfated metabolite activity at the follicle that nobody has fully characterized in forty years of use. It doesn't touch Wnt signaling directly. Patients respond to it or they don't, and the responders mostly thicken existing miniaturized follicles rather than grow new ones.

JAK inhibitors like baricitinib and ritlecitinib are immunosuppressants that target the JAK/STAT pathway ^[8]. They work in alopecia areata, an autoimmune attack on follicles, by quieting the immune assault. Useless for androgenetic alopecia.

PTD-DBM sits in a fourth category. It's a developmental signaling reactivator. The closest conceptual cousin is probably the early-stage Wnt agonist clinical work in other tissues (bone, gut), not the existing hair loss armamentarium.

What's striking, at least on paper, is that PTD-DBM could plausibly work alongside any of these. Minoxidil widens the plumbing, finasteride lowers the hormonal drag, and PTD-DBM tells the follicle to grow. The combinations haven't been tested in humans. They haven't really been tested in mice either, beyond the VPA work.

The mouse data are the strongest part of PTD-DBM's case. They're worth describing in some detail because they're also the entire case.

The foundational regrowth study applied topical PTD-DBM to depilated mouse skin for 28 days ^[9]. Treated animals showed accelerated hair regrowth compared to vehicle controls, and, more striking, formation of new follicles in skin that hadn't had them. Pull-down assays from the same paper confirmed that PTD-DBM was displacing CXXC5 from Dvl in vivo, and nuclear β-catenin staining showed the downstream signal was firing.

The 2023 Cells paper (PMID 36831222) tightened the mechanistic story ^[10]. The group induced alopecia in mice using either PGD2 or DHT directly, then showed that PTD-DBM treatment rescued hair growth in both models. They also used CXXC5 knockout mice as a genetic control: knockouts were resistant to PGD2- and DHT-induced hair loss ^[11], mimicking the pharmacological effect of PTD-DBM. The convergence of pharmacological and genetic rescue is the kind of evidence that makes a mechanism credible in the preclinical literature.

The same paper compared PTD-DBM head-to-head against valproic acid and against KY19382. KY19382 won. PTD-DBM alone gave partial β-catenin restoration; the dual-target small molecule gave fuller rescue. This is internally honest reporting from the developing lab, and it's also a hint about where their commercial bet is going.

A 2023 Advanced Healthcare Materials paper took the delivery problem seriously ^[12]. Topical peptides are notoriously hard to land in skin at therapeutic concentrations. The group built a pyrogallol-functionalized hyaluronic acid patch loaded with both PTD-DBM and VPA, applied it to mouse wounds, and reported regenerative wound healing with suppressed scarring, less α-SMA expression, and increased collagen III deposition ^[13]. Not strictly a hair study, but a proof that the delivery vehicle problem isn't unsolvable.

What the rodent data don't establish: dose-response in primate skin, human pharmacokinetics, durability of effect after treatment stops, safety beyond a 28-day window, or anything about the kind of androgen-driven follicle miniaturization that defines male and female pattern hair loss in humans. Mouse follicle biology is genuinely different. Mouse hair cycles synchronously across the body; human follicles cycle independently. Mouse skin grows new follicles after wounding (the WIHN phenomenon); adult human skin essentially doesn't.

Mice are not small humans. Wnt biology in mouse skin is informative but not predictive.

Here is the part the brief asks about, and the honest answer has to be flat: there's no peer-reviewed human clinical trial for PTD-DBM.

The brief's framing refers to a "2021 proof-of-concept study." A search of PubMed, ClinicalTrials.gov, and the published record returns no such trial that has cleared peer review. What exists in the public record is a 2025 review in Cells (PMID 40497955) that discusses PTD-DBM alongside other Wnt-targeting strategies ^[14] and uses language about "clinical efficacy," but the underlying citations are the preclinical mouse work, not human trials. Commercial cosmetic launches in South Korea by CosmeRNA and Epi Biotech have been built on preclinical data, with no publicly disclosed human trial results from those companies in peer-reviewed venues.

It's possible that internal corporate data exist. It's possible that small human pilot work has been done and not published. Neither possibility constitutes evidence a reader can act on.

Stated plainly: the human evidence base for PTD-DBM as of 2026 is empty. Mechanism is well-developed. Rodent efficacy is real. Human efficacy is unestablished.

The mechanism points most directly at androgenetic alopecia. CXXC5 is elevated in balding scalp tissue, the DHT→PGD2→CXXC5 axis maps cleanly onto pattern hair loss, and the mouse rescue models the group has built specifically replicate androgen-driven loss. If PTD-DBM works in humans, AGA is where the signal should appear first.

Wound-induced hair follicle neogenesis is the second target, but it's a target with a problem: adult human skin doesn't readily do WIHN. Mouse skin does. The translational gap here is large.

Alopecia areata is mechanistically a poor fit. The disease is autoimmune; the brake on the follicle isn't CXXC5, it's a CD8 T-cell attack on the bulge stem cell niche. PTD-DBM doesn't address the immune component. In theory, it could help reactivate dormant follicles after immune suppression, as an add-on to JAK inhibitor therapy, but this is speculation without any supporting data.

Telogen effluvium, chemotherapy-induced alopecia, scarring alopecias, traction alopecia: all hypothetically responsive at the follicle reactivation step, all entirely untested.

The honest scope: pattern hair loss is the indication that matches the mechanism. Everything else is mechanism-adjacent speculation.

Topical is the assumed route for cosmetic formulations and for most of the preclinical work. The PTD carrier domain was specifically engineered to enable cellular penetration, which is what makes a topical strategy plausible in the first place. Most peptides simply don't cross intact stratum corneum at meaningful concentrations.

Whether the PTD tag is sufficient for human skin is a separate question. Human skin is thicker than mouse skin, and the barrier function is more developed. The 2023 hydrogel patch work suggests that even the developing laboratory thinks topical solutions alone may be inadequate, and that adhesive delivery systems holding the peptide against skin under occlusion may be needed.

Microneedle delivery has been discussed in adjacent literature on cosmetic peptides but not, in the published record, formally tested for PTD-DBM. Intradermal injection has not appeared in any peer-reviewed study.

The practical question, what delivery format actually lands enough peptide at the follicle bulge to do something, has no peer-reviewed answer.

There's no human safety data. None. Not from a phase 1, not from a published pilot, not from a regulatory filing.

The mouse studies report no observed adverse effects at the doses tested over the durations tested. Rodent tolerability is a low bar and doesn't establish human safety.

The theoretical concerns are real enough to take seriously. Wnt/β-catenin is an oncogenic pathway in the colon, the breast, and several other tissues. Disinhibiting it pharmacologically, even locally, even transiently, is the kind of intervention that warrants careful tumor surveillance in any human program. The PTD carrier specifically enhances cellular and likely systemic absorption, so the assumption that a topical peptide stays local doesn't apply with the same confidence it would for a conventional cosmetic peptide. Wnt signaling is also critical to embryonic development, which makes pregnancy and breastfeeding hard contraindications by mechanism alone.

None of these risks have been quantified in humans. They're theoretical. But "theoretical" doesn't mean "negligible." It means the studies that would either confirm or rule them out haven't been done.

For populations of particular concern, patients with active or prior Wnt-driven malignancies and anyone pregnant or trying to conceive, the absence of data is itself the relevant fact.

Not approved by the FDA for any indication. No publicly listed Investigational New Drug application. No NDA or BLA. No DailyMed entry. Not classified for 503A compounding. Not eligible for 503B outsourcing facility compounding, which requires either FDA approval or an active IND ^[15].

In the United States, PTD-DBM can't be legally prescribed, can't be legally compounded by a licensed pharmacy for human therapeutic use, and is not sold by any regulated pharmaceutical pathway. Material sold online labeled as PTD-DBM is unregulated research-grade compound, with no oversight of identity, purity, sterility, or potency. The FDA has continued to update bulk drug substance rules for compounding pharmacies, further closing off the gray-market route.

In South Korea, PTD-DBM is sold in cosmetic formulations under the CosmeRNA brand. Cosmetic regulatory frameworks in South Korea, the EU, and the US don't require demonstration of clinical efficacy and don't establish therapeutic claims. A cosmetic-grade product can contain a peptide; it can't legally claim to treat a disease.

In the EU and Japan, no pharmaceutical regulatory classification exists.

The compound's status, in short: cosmetic in one market, research chemical everywhere else, pharmaceutical nowhere.

A serious clinical development program for PTD-DBM would need to answer questions that nobody has yet asked in humans.

Pharmacokinetics first. How much peptide actually penetrates intact scalp skin from a topical formulation? What fraction enters systemic circulation? How long does it persist locally? These are phase 1 questions and they would need a phase 1 trial to answer.

Then dose-finding. The mouse work uses doses that don't translate directly to human exposures. A dose-ranging study in healthy volunteers, then in AGA patients, would establish whether there's a window where the peptide is doing something at the follicle without producing measurable systemic Wnt activation.

Then efficacy. A phase 2 trial in androgenetic alopecia would need a randomized, vehicle-controlled design, with target hair count and hair density on standardized scalp photography as primary endpoints, the same endpoints the field uses for minoxidil and finasteride trials. Twelve months minimum, ideally longer, because hair cycles are long and short trials systematically misestimate response.

Combination work matters too. PTD-DBM plus minoxidil, plus topical or oral finasteride, plus the various delivery vehicles. These are the studies that would tell patients whether the peptide adds anything to existing therapy or whether it stands alone.

And safety surveillance, especially for malignancy signals, would need to run beyond the trial period. Wnt activation is the kind of intervention where five-year follow-up matters more than five-month.

None of this is happening publicly as of mid-2026. No registered ClinicalTrials.gov entries. No published phase 1 protocol. The translational pipeline from Seoul mouse data to American or European regulatory dossier is, on the visible evidence, not under construction.

The mechanism is genuinely novel. Disrupting the CXXC5-Dvl interaction to release the Wnt pathway brake is a different strategy from anything in approved hair loss medicine, and the rodent data supporting it are internally consistent and replicated across several Choi-laboratory papers.

The human evidence is empty. There's no peer-reviewed human clinical trial. The "clinical efficacy" language that circulates in some review articles refers to mechanistic and preclinical evidence, not to results from people. Commercial availability in South Korean cosmetic channels doesn't constitute clinical validation.

The safety case is theoretical in both directions. No human adverse events have been reported because no human trials have been done. The mechanism-based concerns, Wnt's oncogenic potential and the enhanced absorption from the PTD carrier, remain unquantified rather than disproven.

The regulatory status is restrictive. In the US, PTD-DBM is neither approved nor compoundable, and the products available online sit outside the regulated pharmacy system entirely. A patient ordering PTD-DBM from an online vendor is buying an unregulated research chemical of unverified composition.

The honest read of the data: this is a compelling preclinical compound at the start of what would, in a normal drug development pathway, be a five-to-ten-year human trial program. It's not, in 2026, a treatment.