Brain-protecting research peptide (GK-2)
A lab-made dipeptide that mimics nerve growth factor to protect brain cells; studied in animal models of stroke and brain injury; experimental, not an approved drug.
A researcher, an agent, or an algorithm wrote down the sequence and picked a target to hit.
An AI model like OpenFold3 or AlphaFold built a 3D structure and scored how well it fits the binding site.
A second contributor repeated the computation on their own hardware and the scores matched.
A chemistry service or a researcher ordered the sequence, it was manufactured, and mass spectrometry confirmed the right molecule was produced.
A binding or activity measurement confirmed that it actually does what the computer predicted — or didn't.
Snapshot
Class: NGF mimetic dipeptide
Evidence tier: Animal-only evidence
Status: Not approved in any jurisdiction; preclinical research compound only
Best-supported effect: Neuroprotective activity in rodent models of stroke, traumatic brain injury, and Parkinson's-like lesions (animal studies, single research program)
Main caveat: No completed Western-published human trials; virtually all pharmacology originates from one Russian research group; independent replication is limited
What this is
GK-2 is a synthetic dimeric dipeptide designed to mimic the fourth beta-turn loop of nerve growth factor (NGF) — the region of the protein that contacts the TrkA receptor. It was developed by the medicinal chemistry group of Tatiana Gudasheva and Sergei Seredenin at the V.V. Zakusov Research Institute of Pharmacology (Russian Academy of Medical Sciences, Moscow). The design goal was to reproduce NGF's neurotrophic and neuroprotective signaling in a small, drug-like molecule while avoiding the hyperalgesia — pain sensitization — that has limited full-length recombinant NGF as a clinical drug candidate. In published animal work, GK-2 activates the TrkA receptor and downstream survival pathways and shows protective effects in rodent models of ischemia, Parkinsonism, amyloid toxicity, and depression. It is not an approved medicine in any jurisdiction and is not in published Western clinical trials.
Evidence map
| Evidence layer | Grade | What it supports |
|---|---|---|
| Human | None | No human trial data is present |
| Animal | Moderate | Neuroprotective effects in rodent stroke, traumatic brain injury, MPTP-induced parkinsonism, Alzheimer's-relevant amyloid, and forced-swim depression models; most studies originate from a single Russian research program — independent replication outside that lineage is limited |
| In vitro | Moderate | TrkA phosphorylation, PI3K/Akt and MAPK activation, and neuroprotective properties in cell-based assays |
| Computational | None identified | No docking, structure prediction, or design-scoring data are identified |
| Mechanism | Plausible | TrkA-biased dipeptide mimetic concept is chemically coherent and supported by receptor activation and downstream signaling data in cell and animal systems; mechanism not validated in human tissue |
The evidence base is substantially concentrated in the Gudasheva and Seredenin laboratory at the Zakusov Institute. Independent reproduction in Western or non-affiliated groups is limited. This is a material limitation of the current evidence base and is reflected in the confidence rating throughout this card.
Claim check
| Claim | Verdict | Evidence layer | Confidence |
|---|---|---|---|
| Neuroprotection in rodent models of stroke and traumatic brain injury | Supported (animal) | Animal | Medium — multiple published studies, but single research program; no independent Western replication |
| Neuroprotection or disease modification in human stroke or TBI | Not established | Human | Low — no human trial data present |
| Parkinson's-like neurodegeneration protection in animal models | Supported (animal) | Animal | Medium — rodent MPTP model data from the same research program; single-lab limitation applies |
| Antidepressant-like effects in animal models | Supported (animal) | Animal | Medium — forced-swim and related behavioral paradigms; single-lab limitation applies |
| TrkA-selective activation without p75NTR-mediated hyperalgesia | Supported (in vitro / animal) | In vitro / animal | Medium — proposed on the basis of receptor selectivity assays and absence of pain-sensitization in rodent work; not validated in human cells or primate pain models |
| Cognitive improvement or Alzheimer's disease treatment in humans | Not established | Human | Low — no human trial data present |
Experimental exposure
This section reports exposure used in published animal experiments. It does not establish human dosing.
| Context | System | Experimental exposure | Duration | Endpoint | Limitation |
|---|---|---|---|---|---|
| Rodent stroke model (cerebral ischemia) | Rats | Low milligram-per-kilogram doses, intraperitoneal or intravenous; exact dose values not individually extracted | Study-specific acute or subacute period | Neurogenesis, synaptogenesis, histological and behavioral neuroprotection markers | Animal model only; human translation not established; single research program |
| Rodent traumatic brain injury model | Rats | Parenteral route (IP or IV); exact dose not individually extracted | Study-specific | Behavioral and neuroprotective endpoints | Animal model only; human translation not established |
| Rodent MPTP parkinsonism model | Rats or mice | Parenteral route; exact dose not individually extracted | Study-specific | Dopaminergic lesion endpoints | Animal model only; human translation not established |
| Cell-based neuroprotection assays | Neuronal or related cell lines | Concentration range not individually extracted | Assay-specific | Cell survival and TrkA activation markers | In vitro only; does not establish systemic exposure |
Preclinical safety signals
| Signal | System | Notes |
|---|---|---|
| Human side-effect profile | Not characterized | No completed human trials present, human adverse-event profile is unknown |
| TrkA signaling — oncology concern | Theoretical (source-noted) | TrkA contributes to proliferation in some tumor types including TrkA-expressing neuroblastoma and certain adult cancers; formal carcinogenicity and tumor-promotion studies in susceptible animal models are described as absent from the public literature |
| p75NTR engagement | Preclinical / assay | Lower p75NTR activation relative to full-length NGF is a claimed property; extent and clinical implications in humans are not characterized |
| Research-chemical supply quality | Not a clinical compound | GK-2 is not available as pharmaceutical-grade material; supply chain quality and purity are not pharmaceutical-regulated |
| Long-term safety | Not characterized | Chronic animal toxicology and human safety data are absent from the available literature |
Regulatory status
| Region / body | Status | Notes |
|---|---|---|
| US | Not approved | Not FDA-approved; not a controlled substance; not a recognized dietary supplement ingredient; not available as a compounded medicine; regarded as a preclinical research compound |
| EU | Not approved | Per available sources, no approved status in the European Union |
| Russia | Not approved | Per available sources, GK-2 is not registered as a medicine in Russia; other compounds from the broader Gudasheva dipeptide-mimetic program have moved toward Russian clinical evaluation, but GK-2 has not |
| UK, Canada, Australia | Not approved | Per available sources, no approved status in these jurisdictions |
| WADA | per available sources: likely prohibited under S0 | GK-2 is not listed by name on the WADA Prohibited List as reported by source; however, WADA's S0 category prohibits any substance not currently approved by a governmental regulatory authority for human therapeutic use — a description that applies to GK-2; current WADA list status not independently refreshed in this card |
No approved therapeutic status is identified for any jurisdiction.
Mechanism
GK-2 is a dimeric dipeptide — two monosuccinyl-glutamyl-lysine units joined by a hexamethylenediamine linker — modeled on the fourth beta-turn loop of NGF. This loop is the region of the NGF protein most directly associated with TrkA receptor contact. In cell and animal studies from the Gudasheva and Seredenin laboratory, GK-2 engages TrkA, promotes its phosphorylation, and activates the downstream PI3K/Akt and Ras/MAPK survival pathways that characterize full-length NGF signaling.
The central mechanistic claim is that GK-2 is biased toward TrkA over p75NTR engagement. Full-length NGF activates both receptors; p75NTR activation is implicated in NGF-induced hyperalgesia in peripheral tissue. The proposed selective TrkA bias is the pharmacological basis for the argument that GK-2 can deliver NGF-like neuroprotective signaling without the pain sensitization that has complicated clinical development of recombinant NGF.
Both TrkA bias and downstream neuroprotective signaling are characterized in vitro and in rodent models. The mechanism has not been validated in human tissue, human neurons, or primate pain models, and target selectivity has not been independently confirmed outside the originating research program.
Chemistry
| Field | Value |
|---|---|
| Structure type | Dimeric dipeptide |
| Design basis | Synthetic mimetic of the fourth beta-turn loop of nerve growth factor (NGF) |
| Monomer unit | Monosuccinyl-L-glutamyl-L-lysine |
| Linker | Hexamethylenediamine |
| Topology | Two monosuccinyl-glutamyl-lysine fragments joined via hexamethylenediamine linker |
| Amino-acid count | 4 residues total (2 Glu + 2 Lys across two fragments) |
| Full systematic name | Hexamethylenediamide of bis-(N-monosuccinyl-L-glutamyl-L-lysine) |
| Sequence confidence | Needs review — no primary sequence record independently verified in this card |
No molecular weight, formula, CAS number, or sequence string in standard single-letter or three-letter notation was individually extracted from the available literature.
Open questions
- Human pharmacology: No human pharmacokinetic, tolerability, or efficacy data are available. The most urgent gap is a basic Phase I characterization — whether GK-2 reaches the brain, what its half-life is, and whether it is tolerated in humans at all.
- Independent replication: Virtually all published pharmacology originates from the Gudasheva and Seredenin laboratory at the Zakusov Institute. Replication by independent Western groups or non-affiliated Russian groups is limited and is a fundamental requirement before the preclinical record can be treated as broadly credible.
- TrkA selectivity in human models: The p75NTR-sparing argument — which is the central selling point for the no-hyperalgesia claim — has not been rigorously tested in human cells or in primate pain models. This is a key gap between the mechanism hypothesis and a human-relevant safety claim.
- Long-term oncology safety: TrkA signaling contributes to proliferation in neuroblastoma and some adult cancers. Formal carcinogenicity and tumor-promotion studies in susceptible models are absent from the public literature.
- Route and formulation: All published animal work uses intraperitoneal or intravenous parenteral dosing. No validated oral, intranasal, or other patient-friendly route has been published.
- Comparative effectiveness: Direct comparisons of GK-2 versus other neurotrophic-mimetic candidates (Cerebrolysin, Dihexa, Semax, P21, GSB-106) in shared preclinical paradigms are sparse, making relative efficacy difficult to assess.
Research directions for this peptide, selected from the current sources — hypotheses you can explore and model. None of it is proven yet; tap any one to see the full thinking.
Does GK-2 miss the pain-causing nerve cells because those cells have their receptors too far apart for the small double-armed molecule to grab both at once?
Nerve growth factor drugs have been stuck in development for Alzheimer's because they cause severe pain. If GK-2 is naturally selective for brain cells over pain cells because of how its two arms are spaced, it could be developed into a safe Alzheimer's therapy without the pain liability that has stopped its predecessors.
Is glycine in GK-2 necessary because it is the only amino acid that can bend in exactly the right way to position the active lysine for receptor contact?
Understanding why each amino acid in GK-2 is necessary, or replaceable, guides chemists in building improved versions. If glycine turns out to be irreplaceable, it tells developers exactly what any modified version must keep, saving years of trial and error in drug optimization.
Does GK-2 protect brain cells from Alzheimer's-related damage by activating a survival pathway that switches off the enzyme responsible for creating the toxic tau tangles seen in the disease?
If confirmed, GK-2 could be combined with existing anti-amyloid drugs to attack Alzheimer's from two directions at once, potentially slowing the disease more effectively than either approach alone. This would matter for the 50 million people worldwide living with Alzheimer's and the families caring for them.
Is the positive charge on GK-2's lysine amino acid the essential feature that allows this tiny two-residue peptide to activate the TrkA receptor normally switched on by full nerve growth factor?
If true, scientists could design extremely simple, cheap molecules that mimic nerve growth factor by carrying just the right charge, without the complex structure of the full protein. This would lower costs and speed up the development of drugs for nerve damage and Alzheimer's disease.
▸3-letter notation
▸citationbibtex
@peptide{pep10957,
sequence = {GK},
target = {},
author = {peptidemodel},
year = {2026},
status = {designed}
}