For research and educational purposes only · Not medical advice · Consult a qualified physician before any human use
Humanin is a 24-amino acid mitochondria-derived peptide (MDP) encoded within the 16S rRNA gene (MT-RNR2) of mitochondrial DNA, discovered in 2001 simultaneously by three independent research groups. It is the founding member of the MDP family, which also includes MOTS-c and the SHLP peptides. Humanin has documented cytoprotective and neuroprotective effects across dozens of preclinical studies using cell culture and animal models. Human data is limited to observational biomarker studies: circulating humanin declines with age, is elevated in children of centenarians, and is associated with preserved coronary endothelial function. No human intervention trial of any kind has been completed. No IND has been filed. All dosing in community use is extrapolated from animal models without clinical validation.
The defining fact about humanin in 2026: despite 25 years of preclinical research and an impressive mechanistic case, there are zero completed human clinical trials. Not Phase 1. Not Phase 2. No completed human dosing study at any level. No IND application on public record. This is a fundamental gap that separates humanin from virtually every other compound in this database. The human evidence base consists entirely of biomarker correlation studies -- associating endogenous humanin levels with health outcomes -- not intervention data.
The preclinical evidence is genuine and consistent. Dozens of independent laboratories across multiple continents have documented humanin's cytoprotective effects in cell culture and rodent models for Alzheimer's disease, stroke, cardiovascular disease, diabetes, and aging. The three-lab simultaneous independent discovery in 2001 is often cited as strong evidence for biological importance -- when three different groups looking for three completely different things find the same molecule, the underlying biology is almost certainly real. But biological reality in preclinical models has a poor track record of predicting human clinical benefit in complex systemic diseases.
A significant theoretical safety concern exists that must be noted: humanin's mechanism involves blocking apoptosis -- programmed cell death. This is protective in neurons and cardiac cells under stress. In cells that have already undergone oncogenic transformation, blocking apoptosis could potentially enable tumor cell survival and proliferation. This concern is not theoretical: some published studies report that humanin is upregulated in gastric cancer and may contribute to chemotherapy resistance in certain cancer cell lines. This does not mean humanin causes cancer, but it does mean that in any population with pre-existing cancer or high cancer risk, chronic exogenous humanin supplementation carries a theoretical risk that has not been studied in humans.
Humanin's sibling peptide MOTS-c (profiled separately in this database at profile #10) has proceeded further toward clinical translation with an analog (CB4211) in clinical trials for metabolic disease. Humanin itself remains entirely at the bench.
The humanin discovery story in 2001 is one of the more striking episodes in modern peptide biology. The Nishimoto laboratory in Japan was screening a cDNA library constructed from the surviving (non-damaged) brain regions of an Alzheimer's disease patient, looking for factors that could protect neurons from amyloid-beta toxicity. They identified a 75 base pair open reading frame encoding a 24-amino acid peptide in the 16S rRNA region of mitochondrial DNA -- a region previously thought to contain only structural RNA, not protein-coding sequences. Professor Nishimoto named it 'humanin' to denote the potential of this molecule to restore the 'humanity' of Alzheimer's patients, before his premature death from cancer. He named it with hope.
Simultaneously, the Reed laboratory at the Burnham Institute was screening for proteins that could interact with Bax, a key pro-apoptotic protein. They identified humanin as a Bax-binding inhibitor of cell death. The Cohen laboratory at the University of Southern California (later USC and Cedars-Sinai) was screening for IGFBP-3 binding proteins and found humanin independently. Three different proteins, three different experimental systems, one peptide: the convergent discovery strongly implied that humanin was a real and important biological signal rather than an artifact.
The subsequent decade established humanin as the founding member of a new class of signaling molecules: mitochondria-derived peptides (MDPs). The mitochondrial genome was previously considered to encode only 13 protein-coding genes (all components of the oxidative phosphorylation chain), 22 tRNAs, and 2 rRNAs. The discovery that the 16S rRNA gene also harbored a small open reading frame encoding a secreted bioactive peptide opened an entirely new field of mitochondrial biology. Additional MDPs -- including MOTS-c (encoded by the 12S rRNA gene) and the SHLP 1 to 6 family (also from the 16S rRNA region) -- have since been identified, all with overlapping cytoprotective properties.
Pinchas Cohen's laboratory at USC and later Cedars-Sinai has been the primary engine of humanin research, characterizing its IGF-1 signaling interactions, producing the potent HNG analog (S14G substitution, approximately 1000-fold more potent than native humanin for neuroprotection), and generating the human epidemiological data associating circulating humanin levels with aging and longevity phenotypes. The Cohen and Yen laboratories have maintained the primary academic development program, with support from the National Institute on Aging.
Humanin acts through multiple receptor and intracellular mechanisms, which is unusual for a 24-amino acid peptide and reflects the redundancy built into its cytoprotective function.
Extracellular receptor signaling: Humanin binds a trimeric receptor complex consisting of ciliary neurotrophic factor receptor (CNTFR), WSX-1 (IL-27 receptor alpha), and gp130 (IL-6 signal transducer). This complex activates JAK2 kinase, which phosphorylates STAT3. Humanin's activation of pSTAT3 drives transcription of anti-apoptotic and pro-survival genes. This pathway explains humanin's protective effects across multiple cell types and its ability to regulate inflammation, as gp130 is a shared signaling component of multiple cytokine receptors. Humanin also acts through formyl peptide receptors FPRL1 and FPRL2 (FPR2/FPR3), activating ERK1/2 signaling -- a pathway involved in cell survival and neuroprotection.
Intracellular anti-apoptotic mechanisms: Inside the cell, humanin directly binds and sequesters Bax, a key pro-apoptotic Bcl-2 family protein. By physically blocking Bax from inserting into the outer mitochondrial membrane, humanin prevents the mitochondrial permeability transition, cytochrome c release, and downstream caspase activation that commit a cell to apoptosis. Humanin also binds IGFBP-3 (IGF-I binding protein 3), preventing its nuclear translocation and the apoptotic pathway it triggers in the nucleus. This IGFBP-3 interaction creates a functional antagonism between humanin and the IGF-1 signaling axis -- a relationship that is central to understanding humanin's complex metabolic effects.
Metabolic and insulin-sensitizing effects: Through STAT3 signaling in the hypothalamus, humanin influences insulin sensitivity and hepatic glucose production. Animal studies show that humanin suppresses hepatic glucose output and increases peripheral glucose uptake, acting as an insulin sensitizer via central mechanisms. These metabolic effects are not simply secondary to its anti-apoptotic activity -- they represent a distinct functional role as a metabolic regulator.
Age-related changes: Circulating humanin levels decline with age in multiple species including mice, monkeys, and humans. Endogenous humanin is regulated by IGF-1: higher IGF-1 levels (characteristic of growth and younger physiological states) suppress humanin expression, creating a reciprocal relationship between the anabolic IGF-1 axis and the cytoprotective humanin system. Long-lived mouse models (Ames dwarf, calorie-restricted) have higher circulating humanin levels. This inverse relationship between the growth axis and humanin has been proposed as a conserved longevity mechanism, where reduced growth signaling permits higher humanin expression and greater cellular stress resistance.
The most compelling human-relevant evidence for humanin's role in aging comes from two sources. First, circulating humanin levels are significantly lower in older adults compared to younger adults across multiple human cohort studies. This age-dependent decline has been confirmed in both rodents and primates, establishing it as a conserved mammalian phenomenon rather than an artifact.
Second, the children of centenarians -- a population enriched for heritable longevity factors and who themselves are more likely to become centenarians -- have significantly higher circulating humanin levels than age-matched controls. This observation from Yen et al. (2020) is cross-sectional and correlational; it establishes association but not causation. Higher humanin could be a cause of the longevity phenotype, a consequence of whatever genes drive longevity, or simply a correlated biomarker.
The Ames dwarf mouse data provides mechanistic context: Ames dwarfs lack GH and have very low IGF-1 levels, but have 50% higher circulating humanin and live 50 to 70% longer than wild-type. The inverse relationship between the GH/IGF-1 axis (associated with growth but not longevity in multiple species) and humanin (associated with cellular stress resistance and possibly longevity) appears to be conserved. The naked mole-rat data is particularly striking: in this rodent known for negligible senescence and exceptional longevity for its size, humanin levels remain stable with age rather than declining.
Animal intervention data from Yen et al. (2020): HNG treatment of 18-month-old mice twice weekly for 14 months produced significant reductions in visceral fat (p less than 0.05) and significant increases in lean body mass (p less than 0.05) compared to vehicle-treated mice, without changing total body weight at endpoint. Inflammatory markers were reduced. Lifespan itself was not significantly extended. These are healthspan improvements (metabolic and inflammatory parameters) rather than longevity extension.
The centenarian association and age-decline biomarker data are the strongest human-relevant findings in the humanin literature. They are observational -- no intervention data exists in humans. The animal data is consistent and biologically plausible but has not translated to clinical trials. The absence of lifespan extension in the 14-month mouse treatment study despite positive healthspan parameters is a cautionary note for longevity claims.
The original humanin discovery in 2001 showed that this peptide could protect neurons from death induced by three distinct familial Alzheimer's disease gene mutations and by amyloid-beta itself. The protection was structure-specific: scrambled humanin sequence had no protective effect; single amino acid mutations in key positions abolished protection. This specificity argues strongly that the protective effect is mediated by specific molecular interactions rather than non-specific physical properties.
In aged mice, HNG administration improved both spatial memory (Morris water maze performance) and object recognition memory compared to vehicle-treated age-matched controls. The cognitive improvements were accompanied by increased phosphorylation of survival-promoting kinases (AKT, ERK1/2) specifically in the hippocampus -- with the interesting finding that this signaling activation was seen only in old mice, not young mice given the same treatment, suggesting an age-specific responsiveness that aligns with the context of declining endogenous humanin levels with age.
The human genetic association (Bhatt et al., 2018) is the most directly clinically relevant finding for Alzheimer's disease. The rs2854128 SNP in the humanin-coding region was associated with: lower circulating humanin levels in carriers vs. non-carriers; accelerated cognitive aging in a nationally-representative cohort of older adults, with the effect more pronounced in African-Americans (who also had lower baseline circulating humanin levels than Caucasians in this cohort). This is association data -- it does not prove that humanin causes cognitive aging rate differences, but it is consistent with the mechanistic hypothesis and is derived from human population data.
The gap between this evidence and a clinical application is fundamental: there are no human intervention studies of humanin or HNG for Alzheimer's disease. Not Phase 1 for safety. Not Phase 2 for preliminary efficacy. No dosing study in humans at all. The preclinical evidence base is strong and consistent, but the translational gap has not been bridged in over two decades of research.
The neuroprotection evidence base is the most developed in the humanin literature. The original discovery finding (protecting neurons from amyloid-beta toxicity), the HNG animal cognitive data, and the human genetic association study form a coherent narrative. But 25 years after discovery, there is still no human intervention data. The gap between bench and clinical development reflects the field's difficulty in translating MDP biology into trials, compounded by humanin's short half-life and the need to characterize stable analogs.
Widmer et al. (2013) is the landmark human cardiovascular observation. Patients with coronary endothelial dysfunction had significantly lower serum humanin levels (1.3 plus or minus 1.1 vs. 2.2 plus or minus 1.5 ng/mL, p=0.03) compared with those with normal coronary endothelial function. Coronary microvascular endothelial function was significantly and positively correlated with humanin levels (R2=0.17, p=0.009). Coronary epicardial endothelial function showed a trend toward correlation but did not reach significance (R2=0.09, p=0.053). Notably, the traditional inflammatory biomarkers (CRP, Lp-PLA2, homocysteine) were not associated with humanin levels, suggesting that humanin's relationship with endothelial function is independent of conventional inflammatory markers.
The Lu et al. (2021) study of n=327 subjects showed that circulating humanin was significantly reduced in both unstable angina patients (157.77 plus or minus 99.93 pg/mL, p less than 0.05 vs. control) and substantially lower in MI patients (67.17 plus or minus 24.35 pg/mL, p less than 0.05 vs. control). Humanin was inversely correlated with markers of oxidative stress (TBARS), and low humanin levels were associated with worse MACE outcomes during follow-up, establishing it as a potential prognostic biomarker for mild coronary artery disease.
These findings are correlational. They establish that lower endogenous humanin is associated with worse cardiovascular phenotypes in humans. They do not establish that exogenous humanin supplementation would improve cardiovascular outcomes. A healthy person with intact coronary endothelial function has higher circulating humanin -- this could reflect that healthy vasculature produces more humanin (reverse causation) rather than that high humanin protects the vasculature.
The cardiovascular biomarker data in humans is the most directly measurable human evidence in the humanin literature. The Mayo Clinic coronary endothelial function study is well-designed for an observational study. The CAD prognostic biomarker data from China adds prospective outcome correlation. Neither study involves any intervention. The mechanistic animal data (humanin reduces cardiac fibrosis, protects against ischemia-reperfusion injury) is consistent but not translatable without human trials.
The same mechanism that makes humanin cytoprotective -- blocking Bax-mediated apoptosis and IGFBP-3-induced cell death -- creates a theoretical risk in the context of cancer. Apoptosis is one of the body's primary defenses against tumor development: cells with oncogenic mutations are normally eliminated via programmed cell death before they can proliferate into tumors. By blocking multiple apoptotic pathways, exogenous humanin supplementation could potentially extend the survival of pre-cancerous or cancerous cells.
This concern is not entirely theoretical. Humanin is overexpressed in some cancer types. Published reports document humanin upregulation in gastric cancer and bladder cancer tissue compared to adjacent normal tissue. Cell line studies show humanin promotes survival of certain cancer cells and may contribute to chemotherapy resistance in glioblastoma and other tumors by blocking the apoptotic pathways that cytotoxic chemotherapy attempts to activate.
The picture is complicated by contradictory reports showing anti-tumor effects of humanin in other cancer models. The relationship appears to be cell-type dependent: in some contexts humanin's inflammatory modulation and immune effects produce anti-tumor outcomes, while in others the anti-apoptotic effects dominate. The Alzheimer's Disease Discovery Foundation noted: 'There are conflicting reports regarding humanin's role in tumor progression and chemoresistance.'
No human data on cancer risk with exogenous humanin supplementation exists -- there have been no human trials. The cancer concern is drawn from preclinical mechanistic data. However, for any reader considering exogenous humanin use: this is a legitimate and unresolved safety question, not a dismissed theoretical concern.
The cancer risk concern for humanin is mechanistically grounded and supported by preclinical data showing humanin overexpression in certain cancers and pro-survival effects on cancer cell lines. It is counterbalanced by other preclinical data showing anti-tumor effects in different contexts. No human safety data exists to resolve this question. This concern does not mean humanin should not be studied in humans -- it means that cancer risk monitoring should be an explicit component of any future human trial design.
Animal studies consistently show humanin and HNG improve insulin sensitivity and metabolic parameters. The Muzumdar et al. (2009) study established the mechanism: humanin acts in the hypothalamus to suppress hepatic glucose production, similar to the central actions of insulin and leptin. Intracerebroventricular humanin infusion produced significantly greater metabolic effects than peripheral infusion, confirming the central component of this action. This finding has implications for understanding the pharmacokinetics of any therapeutic application -- central delivery vs. peripheral delivery may produce very different metabolic outcomes.
The long-term HNG treatment study in Yen et al. (2020) showed that 14 months of twice-weekly HNG injections in middle-aged mice reduced visceral fat (p less than 0.05) and increased lean body mass (p less than 0.05) without changing total body weight at the final measurement point (28 months). Food intake did not differ between groups, suggesting the body composition changes reflected metabolic remodeling rather than appetite suppression. Circulating inflammatory markers were reduced. These are meaningful metabolic healthspan outcomes in aging mice.
The metabolic effects are mechanistically linked to humanin's interaction with the IGF-1 axis. Humanin antagonizes IGFBP-3 and modulates IGF-1 signaling. IGF-1 itself is closely tied to insulin signaling and insulin resistance. The complex bidirectional relationship between humanin and IGF-1 means the metabolic effects of exogenous humanin are likely dose-dependent, context-dependent, and age-dependent in ways that have not been characterized in humans.
Metabolic insulin-sensitizing effects of humanin are consistently demonstrated in animal models across multiple laboratories. The mechanism (central hypothalamic action suppressing hepatic glucose output) is pharmacologically plausible. The long-term healthspan data in aging mice shows visceral fat reduction and lean mass preservation. Zero human metabolic data exists from any intervention study. This application would require a well-designed human dose-finding and safety study before any metabolic efficacy claims could be evaluated.
Weight-based 0.1mg/kg in research. Very limited human data.
Humanin has no validated human dosing. All community dosing is extrapolated from animal studies without any human pharmacokinetic, safety, or efficacy data. The HNG analog (S14G substitution, approximately 1000-fold more potent) is used in most animal longevity research and is the form typically available from research chemical suppliers. Native humanin and HNG have distinct potency profiles that cannot be simply dose-converted. The short peptide half-life in circulation (estimated minutes based on animal data) means multiple daily injections would likely be required for sustained plasma levels -- a practical barrier that the animal twice-weekly IP injection protocols do not reflect. The FDA has not approved humanin for any use and no IND application is on public record. WADA does not list humanin on the 2026 Prohibited List. Cancer risk from the anti-apoptotic mechanism is an unresolved theoretical and preclinical concern that must be disclosed for any use. Any human use at this stage is experimental without clinical validation of any kind.
Humanin is the founding member of the mitochondria-derived peptide (MDP) family. Its closest database relative is MOTS-c (profile #10), which shares the mitochondrial genome origin and overlapping metabolic/longevity biology, but acts through distinct mechanisms (MOTS-c targets AMPK and the folate cycle) and has advanced further toward clinical translation -- the MOTS-c analog CB4211 has entered clinical trials for metabolic disease. SS-31 (profile #2) is another mitochondria-targeted peptide in this database; unlike humanin (which is encoded by mitochondrial DNA and circulates as an endocrine signal), SS-31 is a synthetic peptide that targets cardiolipin on the inner mitochondrial membrane. In the broader neuroprotection space, Dihexa (profile #16) and Semax (profile #11) both have stronger clinical evidence for cognitive applications, though humanin's mechanism (blocking multiple apoptotic pathways) is distinct from either. Humanin represents the longevity-biology end of the mitochondrial peptide spectrum, with MOTS-c representing the metabolic-exercise end.
Human safety data for exogenous humanin does not exist. No human dosing study has been conducted. All safety considerations are either theoretical or derived from animal models.
Animal safety: Mouse studies using HNG at 4 mg/kg twice weekly intraperitoneally for up to 14 months in middle-aged and old mice (Yen et al., 2020) reported no significant adverse events, no differences in food intake, and no mortality attributable to treatment. Body weight was reduced in treated animals without changes in food consumption, suggesting a metabolic effect. Long-term HNG treatment improved multiple healthspan parameters in mice without apparent toxicity at this dose and schedule.
Theoretical cancer concern: The most important safety consideration is the theoretical oncogenic risk from humanin's anti-apoptotic mechanism. Apoptosis is a key tumor suppressor mechanism -- it eliminates cells that have accumulated DNA damage or oncogenic mutations. By blocking apoptosis through Bax antagonism and IGFBP-3 inhibition, exogenous humanin could theoretically extend the survival of pre-cancerous or cancerous cells. This concern is supported by published preclinical data: humanin is upregulated in gastric cancer tissue compared to normal tissue, and some studies show humanin promotes survival of cancer cells and may contribute to chemotherapy resistance in certain tumor models. The relationship between humanin and cancer is bidirectional and context-dependent -- some studies show anti-tumor effects in other cancer models. No human data exists to resolve this question.
IGF-1 axis interaction: Humanin's antagonism of IGFBP-3 and its reciprocal relationship with IGF-1 signaling creates potential for hormonal interactions that are not characterized in humans. In developmental contexts and young animals, excessive humanin impairs growth and reproductive function by suppressing the IGF-1 axis. In middle-aged and older adults this may represent a beneficial effect, but the dose-response and individual variation are not characterized.
Short half-life and stability issues: Native humanin is rapidly degraded in circulation, making consistent delivery pharmacologically challenging. This has driven development of the HNG analog and other modified versions, but the safety profile of each analog must be separately characterized.
Humanin is the most biologically compelling compound in this database that has not advanced to any human clinical testing. Discovered in 2001 in a convergent triple-lab finding, it has accumulated 25 years of preclinical evidence across neuroprotection, cardiovascular protection, metabolic regulation, and aging. The epidemiological associations with longevity (centenarian offspring have higher levels; naked mole-rats don't lose it with age) are striking. The human genetic association linking lower humanin to accelerated cognitive aging adds a layer of human-relevant evidence. And yet: zero human intervention trials.
This is not an accident of timing or funding. It reflects genuine obstacles: humanin's short peptide half-life makes consistent therapeutic delivery difficult; the potent HNG analog has improved stability but requires its own safety characterization; the cancer risk concern from the anti-apoptotic mechanism has complicated IND development; and the development of the MOTS-c sibling peptide (which has advanced to clinical trials for metabolic disease) may have attracted more commercial interest. None of these obstacles are insurmountable -- they are simply unresolved.
The human observational data for cardiovascular biomarker correlation (Widmer et al., 2013) represents the strongest direct human evidence and establishes that circulating humanin levels are clinically meaningful in humans, not just in mouse models. The centenarian offspring data adds weight to the longevity hypothesis. But observational data is not intervention data, and association is not causation.
For researchers and practitioners: humanin is a legitimate scientific target with a strong preclinical rationale and suggestive human biomarker data. It is not ready for clinical use. The cancer risk concern is real and unresolved. Any human use at this stage is experimental in the most literal sense -- there is no validated dose, no characterized safety profile, and no evidence of human efficacy.
For research and educational purposes only · Not medical advice · Consult a qualified physician before any human use