Humanin: natural cell-protection signal made by mitochondria
A small peptide produced inside your cells' energy factories that protects brain cells and may slow aging; experimental, not yet an approved drug.
- Class
- Mitochondrial-derived peptide (MDP)
- Status
- No approved therapeutic status in any jurisdiction
- Best-supported effect
- Cytoprotection and anti-apoptotic activity via BAX inhibition in cell and animal models; neuroprotection against amyloid-beta toxicity characterized in preclinical systems
- Main caveat
- No interventional human efficacy data for exogenous Humanin; observational human data reflect endogenous biomarker associations, not therapeutic causality
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.
Named peptide fragment — synthesized for research; ClinicalTrials.gov trials registered for parent compound or class
Fork this card to add platform evidence →
Endogenous peptide fragment — receptor binding/activity established in published literature; CT.gov evidence
Fork this card to add platform evidence →
Snapshot
Class: Mitochondrial-derived peptide (MDP)
Evidence tier: Animal-only evidence
Status: No approved therapeutic status in any jurisdiction
Best-supported effect: Cytoprotection and anti-apoptotic activity in cell and animal models; neuroprotection against amyloid-beta toxicity characterized in preclinical systems
Main caveat: No interventional human efficacy data for exogenous Humanin; observational human data reflect endogenous biomarker associations, not therapeutic causality
What this is
Humanin is a 24-amino-acid peptide encoded by a short open reading frame within the mitochondrial 16S rRNA gene. It was first described in 2001 through independent work identifying a neuroprotective factor in surviving neurons from Alzheimer's disease brain tissue. Humanin became the founding member of the mitochondrial-derived peptide (MDP) family, which later expanded to include MOTS-c and the SHLP peptides. Endogenous Humanin levels decline with age, and observational human data show correlations between circulating levels and metabolic and cognitive parameters — though causality has not been established.
The peptide has well-characterized cytoprotective biology in cell and animal systems: BAX inhibition, IGFBP-3 binding, and STAT3 activation are replicated mechanisms. What has not been established is any therapeutic effect of administering exogenous Humanin in humans. Most preclinical research uses the synthetic HNG (S14G-Humanin) analog, which is approximately 1000-fold more potent than wild-type Humanin in some neuroprotection models; results from HNG studies do not translate directly to wild-type Humanin at equivalent mass doses.
Evidence map
| Evidence layer | Grade | What it supports |
|---|---|---|
| Human | Observational / biomarker | Circulating Humanin levels correlate with age, metabolic parameters, and cardiovascular markers; no interventional human efficacy data for exogenous Humanin identified in the available literature |
| Animal | Moderate–strong | Neuroprotection in Alzheimer's models, healthspan extension, cardioprotection in ischemia-reperfusion, and insulin sensitivity in rodent and other animal experiments |
| In vitro | Strong | BAX inhibition, IGFBP-3 binding, STAT3 activation, amyloid-beta toxicity reduction, retinal pigment epithelium protection — well-replicated in cell systems |
| Computational | Not present | No computational or structural prediction data identified |
| Mechanism | Strong | Multiple independent pathways verified in cell and preclinical systems; mechanism plausibility is the strongest pillar of the Humanin evidence base |
Most published RCTs in the reference set measured endogenous Humanin levels as a biomarker response to exercise, radiation, or other interventions — these are biomarker association studies, not exogenous Humanin therapy trials. A large share of the mechanistic and preclinical evidence originates from or has been advanced substantially by one research network (Pinchas Cohen's laboratory, USC); independent replication depth varies across individual claims.
Claim check
| Claim | Verdict | Evidence layer | Confidence |
|---|---|---|---|
| Cytoprotection and anti-apoptotic activity via BAX inhibition | Supported (preclinical / in vitro) | In vitro, animal | High — extensively replicated in cell systems and animal models |
| Neuroprotection against amyloid-beta toxicity | Supported (preclinical) | In vitro, animal | Medium — strong cell-model evidence; animal model replication present; no human interventional data |
| Improved insulin sensitivity | Supported (preclinical) | Animal | Medium — animal model evidence present; human interventional evidence absent |
| Cardioprotection in ischemia-reperfusion | Supported (preclinical) | Animal | Medium — animal model evidence; no human trial data attached |
| Observational association between endogenous Humanin levels and health outcomes | Supported (observational) | Human | Medium — causality not established; direction of association does not confirm that exogenous supplementation produces the same effect |
| Therapeutic efficacy of exogenous Humanin in humans for any indication | Not established | Human | High confidence in the verdict — no completed Phase II or controlled human trial of exogenous Humanin identified in the available literature |
| Exogenous Humanin is equivalent to HNG (S14G-Humanin) at the same mass dose | Not established | Animal | High confidence in the verdict — HNG is approximately 1000-fold more potent than wild-type Humanin in some preclinical neuroprotection models; the two are not interchangeable |
| Humanin supplementation restores youthful function by reversing age-related decline | Not established | None | High confidence in the verdict — age-associated decline in an endogenous biomarker does not imply that exogenous restoration produces equivalent effects; no interventional human data |
Experimental exposure
This section reports exposure used in animal experiments and preclinical studies. It does not establish human dosing.
| Context | System | Experimental exposure | Duration | Endpoint | Limitation |
|---|---|---|---|---|---|
| Rodent neuroprotection studies | Mice / rat models | Broad range reported across studies (source describes 0.1–10 mg/kg in rodents, often using HNG analog) | Days to weeks depending on endpoint | Amyloid-beta toxicity markers, survival, behavioral endpoints | HNG analog used preferentially for potency reasons; results do not directly translate to wild-type Humanin; no human PK established |
| Healthspan / lifespan studies | Rodent models | Long-duration administration; exact regimen not individually extracted | Extended (weeks to months) | Healthspan and lifespan markers | Rodent duration does not translate to human protocol; no human replication |
| Insulin sensitivity experiments | Rodent metabolic models | Study-specific dose; exact regimen not individually extracted | Study-specific | Insulin sensitivity markers, AMPK activation | No human translation established |
No human pharmacokinetic data for exogenous Humanin are available in the available literature. All dose-related information describes animal experimental contexts or forum-reported extrapolations. Forum-described community protocols (subcutaneous injection, research-chemical sourcing) are not included in this section because they represent community use patterns without clinical or experimental anchor — they are not studied exposure.
Preclinical safety signals
| Signal | System | Notes |
|---|---|---|
| Generally well tolerated in animal and cell studies | Rodent / cell systems | available literature reports no major toxicity signals in preclinical studies; duration and species limitations apply |
| Anti-apoptotic mechanism and theoretical cancer-cell survival concern | Theoretical / mechanistic | BAX inhibition that protects normal cells could theoretically support tumor cell survival; this concern has not been resolved by clinical data and is flagged explicitly in available literature |
| Theoretical interaction with pro-apoptotic cancer therapies | Theoretical / mechanistic | Concurrent use with chemotherapy or radiation sensitizers that depend on inducing apoptosis in target cells could in principle oppose therapeutic intent; no clinical interaction data exist |
| Theoretical additive effect with insulin-sensitizing agents | Theoretical / mechanistic | Preclinical insulin-sensitizing effects raise theoretical concern for additive effects with insulin, sulfonylureas, or GLP-1 agonists; not clinically characterized |
| Long-term safety | Not established | No chronic human safety data; no human pharmacovigilance data |
| Human reproductive and developmental safety | Not established | No reproductive toxicology data identified; no pediatric safety data |
Research the following populations as having specific theoretical or preclinical concerns: active or recent-history cancer, pregnancy, breastfeeding, pediatric use, and concurrent use of pro-apoptotic therapies. These are reported as source-identified contraindications, not clinically validated exclusion criteria.
Regulatory status
| Region / body | Status | Notes |
|---|---|---|
| US (FDA) | Not approved | Not approved for any indication; not a recognized dietary supplement ingredient; no established compounding pathway; sold through research-chemical suppliers not authorized for human use |
| EU | Not approved | Not approved as a medicine in any EU jurisdiction identified in the available literature |
| UK | Not approved | Not approved; availability through research-chemical channels subject to local rules |
| Canada | Not approved | Not approved in any jurisdiction identified in the available literature |
| Russia / other | Not approved | Unlike some peptides approved regionally (e.g., Cerebrolysin, Selank), Humanin has no clinical regulatory approval in any jurisdiction per the available literature |
| WADA | Not listed by name; likely prohibited under S0 | per available sources: Humanin is not named on the WADA Prohibited List; because it is not approved by any governmental health authority for human therapeutic use, Research it likely falls under WADA's S0 catch-all category prohibiting unapproved substances. Current list status not independently refreshed in this card. |
Mechanism
Humanin exerts cytoprotective effects through several independently characterized pathways in cell and animal systems.
BAX inhibition: Humanin directly interacts with BAX to prevent mitochondrial membrane permeabilization, blocking a key step in the intrinsic apoptosis pathway. This mechanism is among the most replicated in the preclinical literature and is the basis for the anti-apoptotic characterization.
IGFBP-3 binding: Humanin binds insulin-like growth factor binding protein-3 (IGFBP-3), modulating IGF-1 signaling. This interaction places Humanin within the IGF-1 axis and raises mechanistic interactions with growth hormone and IGF-1-related therapies.
STAT3 activation: Humanin activates the STAT3 signaling pathway, which has downstream effects on cell survival, inflammation, and immune function. STAT3 involvement introduces a broad theoretical interaction surface with inflammatory and immune-modulating therapies.
AMPK activation / insulin sensitivity: In metabolic model systems, Humanin has been shown to improve insulin sensitivity through AMPK activation and reduce inflammatory cytokine production.
Amyloid-beta toxicity reduction: In neuronal cell systems, Humanin reduces toxicity associated with amyloid-beta peptide accumulation — the primary mechanistic rationale for neuroprotection research in Alzheimer's disease models.
All of these mechanisms are characterized in cell or animal systems. Target confidence for each pathway is verified (in vitro / preclinical); clinical translation in humans has not been established for any pathway.
Chemistry
| Field | Value |
|---|---|
| Amino-acid chain | 24-amino acid sequence; exact one-letter sequence not individually extracted's source |
| Length | 24 amino acids |
| Topology | Linear |
| Origin | Encoded by a short open reading frame within the mitochondrial 16S rRNA gene (16S rDNA) |
| Key analog | HNG (S14G-Humanin): glycine substituted at position 14 for serine; approximately 1000-fold more potent than wild-type in some preclinical neuroprotection assays |
| Sequence confidence | Needs review — source does not provide one-letter sequence; sequence should be verified against primary chemistry source before population |
| Molecular weight | Not extracted from source |
| CAS / formula | Not extracted from source |
Open questions
- Human interventional evidence: No completed Phase II or controlled human trial of exogenous Humanin for any clinical endpoint has been identified. The distance between well-characterized endogenous biology and validated exogenous therapeutic use is currently unbridged.
- Human pharmacokinetics: Absorption, distribution, clearance, and bioavailability of exogenous Humanin by subcutaneous or other routes are not established in humans. Without PK data, dose extrapolations from animal studies have no validated anchor.
- Analog equivalence in humans: Most preclinical research uses HNG (S14G-Humanin) at doses exploiting its ~1000-fold potency advantage. Whether wild-type Humanin produces comparable effects at achievable exogenous doses in humans is unresolved. The research-chemical market conflates the two molecules in ways the evidence does not support.
- Cancer-risk signal: The BAX-inhibiting anti-apoptotic mechanism raises a specific theoretical question about tumor-cell survival. This has not been resolved by human pharmacovigilance data, and the concern is explicitly flagged in available literature.
- Causal direction of observational associations: Observational studies show Humanin levels correlate with age, metabolic, and cognitive parameters. Whether declining Humanin is a cause, consequence, or bystander marker of aging biology is not established. Supplementation logic based on correlation alone is premature.
- Dose-response in humans: No dose-finding studies in humans have been completed. Forum protocols extrapolated from rodent studies are not a validated basis for human dosing.
- Long-term safety: Chronic exposure data in humans are absent. No pharmacovigilance program has been established for exogenous Humanin use.
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 the age-related drop in Humanin come before, and actually cause, the neuron death seen in Alzheimer's disease?
If Humanin loss is the first domino, replenishing it early in aging could prevent the chain of events that leads to dementia, giving millions of people a way to protect their brain health before symptoms appear.
Does the sulfur-containing amino acid in Humanin cause it to lose its protective power when the cell environment is most damaged?
If true, a simple chemical tweak removing that vulnerable amino acid could produce a version of Humanin that stays active during heart attacks or strokes, potentially saving more tissue and benefiting patients in intensive care or surgery.
Does Humanin only protect cells that have stopped dividing, making it safer to use without accidentally helping tumors grow?
If Humanin spares cancer cells from protection, it could be used to shield healthy organs during chemotherapy or after a heart attack without the risk of promoting tumor survival, a major concern with any anti-cell-death drug.
Does the extra segment on the cytosolic form of Humanin make it better at blocking the protein that triggers cell death?
If true, doctors and researchers could target the more potent form specifically, potentially reducing cell loss in conditions like Alzheimer's disease or heart attack with fewer side effects. This could help patients whose cells are dying prematurely due to stress or disease.
▸full evidence table1 metrics
| metric | value | tool |
|---|---|---|
| ranking score | 0.71978759765625 | boltz-2 |
▸3-letter notation
▸recipeboltz-2 2.2.1
| parameter | value |
|---|---|
| model | boltz-2 2.2.1 |
| weights | — |
| hardware | vast_v100_32gb |
| mlx version | — |
| python | — |
| random seed | 1 |
| msa strategy | none_monomer |
| runtime | — |
| predicted by | — |
| predicted at | 2026-05-23 |
▸citationbibtex
@peptide{pep10776,
sequence = {MAPRGFSCLLLLTSEIDLPVKRRA},
target = {longevity},
author = {peptidemodel},
year = {2026},
status = {bioassayed}
}