KY19382 has gone from a laboratory code to a searched name — about 720 US searches a month and rising fast, many of them followed by “reddit” (clickstream estimate, October 2026). This page sets out what it is, what has been published about it and by whom, and what has never been tested, as of 15 October 2026. Two corrections come first. It is not a peptide, despite appearing on lists of hair peptides. And no study has tested it in a human being. Nothing here is advice to use any compound; hair loss is a medical question for a dermatologist. The peptides that have reached human hair trials are on our peptides for hair growth page.
What it is
PubChem records KY19382 as 5,6-dichloroindirubin-3'-methoxime, formula C17H11Cl2N3O2, 360.2 daltons (PubChem CID 148959784). It is one of a series of synthetic analogues of indirubin-3'-monoxime made in Kang-Yell Choi's laboratory at Yonsei University in Seoul. Indirubin itself is a component of a traditional Chinese medicine preparation. A 360-dalton organic molecule with two chlorine atoms is about as far from a peptide — a chain of amino acids — as cosmetic chemistry gets.
How it is meant to work
Hair follicles cycle between growth (anagen), regression and rest, and the Wnt/β-catenin signalling pathway is one of the switches that pushes them into growth. A protein called CXXC5 acts as a brake: it binds a pathway component called Dishevelled and dampens the signal. KY19382 was identified as a compound that blocks that binding, releasing the brake (Ryu et al., British Journal of Pharmacology 2021).
The same laboratory had earlier designed a peptide to do the same job, PTD-DBM, which competes with CXXC5 for its binding site. That is the link between KY19382 and peptide hair lists: same target, same laboratory, different kind of molecule. PTD-DBM's mouse results are summarised on our hair growth hub.
What the eleven papers tested
PubMed returned 11 records mentioning KY19382 on 15 October 2026 (10 with it in the title or abstract), none retracted. The hair work is four papers:
| Paper | What was tested | In what |
|---|---|---|
| Ryu 2021, British Journal of Pharmacology | hair induction markers; hair length; regrowth after shaving; new follicles in wounds | human dermal papilla cells, cultured mouse whiskers and human follicles, mice |
| Ryu 2022, BMB Reports | regrowth in diabetic mice; follicle length in high-glucose culture | diabetic mice, cultured human follicles |
| Ryu 2022, Pharmaceutics | KY19382 combined with a PKM2 activator | mice |
| Seo 2020, Pharmaceutics | a micelle formulation to get an indirubin analogue through skin | laboratory skin-permeation tests |
The 2021 paper is the foundation. Its authors report that KY19382 raised markers of hair induction in human dermal papilla cells, lengthened mouse whiskers and human hair follicles kept alive in culture, sped regrowth on shaved mouse backs, and increased wound-induced hair follicle neogenesis — new follicles forming in healing mouse skin, a phenomenon that happens in mice and has not been shown in adult human scalp. Minoxidil was the paper's positive control in cells and on mouse backs, and the authors report KY19382 promoted more regrowth than minoxidil at 28 days in mice. Their own discussion carries the caution that matters: earlier Wnt activators, such as valproic acid, promoted hair growth in the laboratory but “often show marginal effects in clinical tests”. The other papers extend the same model to diabetic mice and to drug combinations. The remaining records are about wound healing, bone growth and cancer cells, not hair.
The formulation paper matters more than its title suggests. Indirubin analogues dissolve poorly in water, and the authors built a surfactant micelle to push one through skin. Whether any formulation delivers enough KY19382 to a human follicle, through a human scalp, is a question no published study has answered.
Who did the work
Every hair paper comes from the same Yonsei group, and the papers declare the commercial interest. The 2021 paper's conflict statement reads that Choi “is the CEO of CK Biotech. Inc. (Seoul, Korea), which has a licence to develop and use the compounds disclosed in the publication”; the 2022 papers give the company as CK Regeon Inc. The formulation paper notes Choi is an inventor on a patent application covering KY19382 formulations. A developer's own preclinical work is the normal first step for any drug; it is also, by definition, not independent replication, and none exists yet.
The human evidence: none
ClinicalTrials.gov returned no study for KY19382 on 15 October 2026. We found no human trial, case series or safety study in PubMed either. That leaves the gap every mouse hair result faces: mouse fur cycles in synchronised waves and responds to many compounds that do nothing on a balding human scalp, and wound-induced follicle formation is a mouse phenomenon. The ladder a hair drug climbs — human safety, then dose-finding, then hair counts against placebo — has not been started in public. Our page on PP405, another non-peptide hair molecule, shows what even the first rung of that ladder looks like when it is posted.
The safety question nobody has answered in people
Switching on Wnt/β-catenin signalling is a known concern because the same pathway is overactive in several cancers. The developers have addressed it in their own laboratory: a 2025 paper reported that KY19382 and a related compound, KY19334, inhibited the malignant behaviour of human cutaneous squamous cell carcinoma cells and reduced chemically induced skin tumours in mice (Lee et al., Experimental and Molecular Medicine 2025). That result runs against the concern in cells and mice; it is not a human safety record, and there is no data at all on applying the compound to human skin for months.
Why people can find it to buy
PubChem's synonym list for KY19382 includes several chemical-supplier catalogue numbers, such as HY-131447 and s6085 — the form in which compounds are sold to laboratories as research reagents. That is why it can be bought and discussed on forums. A reagent listing is not a product approval: no medicines regulator has approved KY19382 for any use, and the purity, dose and vehicle in anything sold for self-application are unknown. We did not assess forum reports; they are anecdotes without a control, and the absence of a human trial is the checkable fact.
What would change this page
A registered human trial — safety first, then hair counts against a vehicle — or an independent group reproducing the hair results. Until then, the record shows a well-characterised molecule that works on mouse fur and cultured follicles, held under licence by its inventor's company, and untested on people. The established comparison, minoxidil and finasteride, is set out on our hair growth hub.
Sources
Opened 15 October 2026: PubChem compound record and synonym list (CID 148959784); PubMed search KY19382 (11 records) and KY19382[tiab] (10), with a retraction filter (0); abstracts and conflict statements of PMIDs 33751552, 36016500, 36559274, 32093032 and 40887499; ClinicalTrials.gov search KY19382 (0 studies). Corrections go to the contact page.