Humanin

Humanin is a small peptide that is unusual for a simple reason: its gene does not sit in the nucleus. The sequence lies inside the mitochondrial genome, within the MT-RNR2 region that also encodes the 16S ribosomal RNA. That makes humanin the founding member of a class of molecules now called mitochondrial-derived peptides (MDPs).

The form most often used in laboratory work is 24 amino acids long (MAPRGFSCLLLLTSEIDLPVKRRA) and corresponds to translation in the cytosol. A shorter 21 amino acid form is described when the open reading frame is translated inside the mitochondrion itself, using the mitochondrial genetic code. Both forms appear in the literature, which is one reason published values for molecular weight and potency are not always directly comparable.

Functionally, humanin is usually presented as a retrograde signal: a message sent from the mitochondria outward, to the rest of the cell and, because the peptide is detectable in plasma and cerebrospinal fluid, potentially to other tissues. Researchers sometimes use the informal term "mitokine" for this idea. The concept matters because it reframes mitochondria as signaling organelles rather than passive power plants, a shift described in the review by Lee, Yen and Cohen in Trends in Endocrinology and Metabolism.

What follows is a factual summary of what has been published about humanin, including where the evidence is strong (cell biology, rodent physiology) and where it is absent (controlled human trials). This article is for educational purposes only. Humanin is a research peptide and is not an approved medicine in any major jurisdiction.

The discovery is reported in a 2001 paper in the Proceedings of the National Academy of Sciences by Yuichi Hashimoto and colleagues, working in the laboratory of Ikuo Nishimoto at Keio University. The team was not looking for a mitochondrial peptide. They were looking for anything that could keep neuronal cells alive when familial Alzheimer's disease genes were expressed.

Their approach was a functional screen of a cDNA library prepared from the occipital lobe of a patient with Alzheimer's disease, a brain region that is relatively spared in the disease. From that library they isolated a cDNA that abolished the death of neuronal cells caused by a wide spectrum of familial Alzheimer's disease genes and by amyloid beta. They named the encoded polypeptide Humanin.

Two details from the original paper are worth keeping in mind. First, the transfected cDNA was transcribed into the corresponding polypeptide and then secreted into the culture medium, which supported the idea that humanin acts as an extracellular signal rather than purely inside the cell. Second, the rescue action clearly depended on the primary structure of the peptide, because altering key residues abolished the protective effect. That structure dependence is what later made analogue design possible.

The 2001 authors framed humanin as a molecular clue for developing neuroprotective strategies, not as a finished therapy. Twenty-five years later, that framing still describes the state of the field accurately. The mitochondrial origin of the sequence was recognized afterwards, and it is that observation, more than the Alzheimer's connection, that gave humanin its lasting scientific importance.

Humanin opened a category. Once researchers accepted that the mitochondrial genome contained short open reading frames producing functional peptides, they went looking for more. The best known second member is MOTS-c, described by Lee and colleagues in Cell Metabolism in 2015 and encoded within the mitochondrial 12S rRNA region. In that work, MOTS-c was reported to interfere with the folate cycle and de novo purine biosynthesis, leading to accumulation of AICAR and activation of AMPK, with prevention of diet-induced weight gain and insulin resistance in mice.

A third group arrived with the 2016 paper by Cobb and colleagues in Aging, which reported six additional peptides from the same region of mitochondrial DNA, named small humanin-like peptides (SHLP1 to SHLP6). SHLP2 and SHLP3 shared protective effects with humanin in vitro, reducing apoptosis and reactive oxygen species generation while improving mitochondrial metabolism.

Broadly, the literature positions humanin as the cytoprotective and neuroprotective member of the family, and MOTS-c as the metabolic and exercise-associated member, with considerable overlap in aging biology. If you want a side-by-side view of that second peptide, see the dedicated MOTS-c guide and the comparison in Klow vs MOTS-c.

It is worth separating MDPs from a different category that is often confused with them. Peptides such as SS-31 (elamipretide) are synthetic molecules designed to concentrate in mitochondria and interact with cardiolipin in the inner membrane. They are mitochondria-targeted, not mitochondria-encoded. Humanin and MOTS-c come out of the organelle; SS-31 is sent into it. The two research programs answer different questions.

The best characterized extracellular mechanism involves a cytokine-type receptor complex. Work from Hashimoto and colleagues published in Molecular Biology of the Cell in 2009 reported that humanin inhibits neuronal cell death by interacting with a receptor complex involving CNTF receptor alpha, WSX-1 and gp130, which places humanin signaling in the same family as interleukin-6 type cytokine signaling. Downstream of gp130, the expected pathway is JAK/STAT activation.

That connects directly to the metabolic literature. In the 2009 PLoS One study by Muzumdar and colleagues, the central effects of humanin on insulin action in rodents were associated with activation of hypothalamic STAT-3 signaling. The same intracellular node therefore appears in both the neuronal survival work and the whole-body metabolic work, which is one reason the peptide attracts attention from the aging field.

Other mechanisms have been proposed and are described more cautiously in reviews: binding to formyl peptide receptor family members, interactions with pro-apoptotic proteins inside the cell, modulation of reactive oxygen species, and chaperone-like activity reported for humanin analogues. These proposals come from different experimental systems and have not been unified into a single accepted model.

Two methodological cautions apply across this literature. Many in vitro experiments use peptide concentrations well above what circulates physiologically, so a mechanism demonstrated in a dish does not automatically operate at endogenous levels. And because the different humanin forms and analogues differ in potency, a mechanism established with HNG is not necessarily the mechanism of native 24 amino acid humanin.

The cell-culture evidence is the oldest and the most consistent. The founding 2001 experiments showed protection of neuronal cells against toxicity from multiple familial Alzheimer's disease genes and from amyloid beta. Later cell work extended the same cytoprotective theme to non-neuronal contexts, including retinal pigment epithelium cells exposed to oxidative stress, where humanin has been reported to limit senescence markers and mitochondrial dysfunction.

The animal evidence centers mostly on the analogue rather than the native peptide. In a 2012 study published in Pharmacology Biochemistry and Behavior, Zhang and colleagues reported that three months of intraperitoneal S14G-humanin (HNG) administration improved spatial learning and memory deficits in middle-aged APPswe/PS1dE9 mice, and reduced cerebral amyloid beta plaque deposition, insoluble amyloid beta levels and neuroinflammatory responses compared with vehicle.

Normal aging has also been studied. Yen and colleagues reported in Scientific Reports in 2018 that humanin was neuroprotective in human cell culture models and was sufficient to improve cognition in aged mice, and they described an association between humanin and cognitive measures in humans. The human part of that work is an association in a cohort, not a demonstration that administering humanin improves cognition in people.

The honest summary is that humanin protects cells against several insults reproducibly, improves outcomes in some rodent models of amyloid pathology and aging, and has never been tested as an intervention in a human neurological trial. Alzheimer's models in mice have a long history of producing positive results that do not survive translation, and nothing in the humanin literature exempts it from that pattern. Anyone concerned about cognitive symptoms should speak with a qualified healthcare professional rather than with a peptide vendor.

The metabolic thread starts with the observation that insulin resistance and Alzheimer's disease are epidemiologically linked. Muzumdar and colleagues tested whether a peptide known for neuroprotection also touched glucose metabolism. In their 2009 PLoS One paper, continuous intracerebroventricular infusion of humanin significantly improved overall insulin sensitivity in rodents, with the effect associated with hypothalamic STAT-3 activation. The route of administration matters: the key experiment delivered the peptide directly into the brain ventricles, which is a laboratory technique, not a clinical one.

Later work with the analogue HNG looked at systemic outcomes in aging animals. In the 2020 Aging paper by Yen and colleagues, treating middle-aged mice with HNG improved metabolic healthspan parameters and reduced inflammatory markers. The same group had previously reported that humanin and the related SHLPs behave as age-dependent regulators of apoptosis, insulin sensitivity and inflammatory markers.

In humans, the metabolic data are observational. Circulating humanin has been measured in cohorts with metabolic conditions and compared against controls, and it appears in the literature as a candidate biomarker of mitochondrial stress or mitochondrial reserve. Those studies tell you that humanin levels differ between groups. They do not tell you which direction the causal arrow points, and a peptide released in response to mitochondrial stress could plausibly rise or fall for reasons that have nothing to do with benefit.

So the metabolic claim that is defensible is narrow: in rodents, humanin and HNG influence insulin action and inflammatory markers, with a central nervous system component. The claim that is not defensible is that taking humanin improves insulin sensitivity in people. No trial has tested it. Readers comparing candidates in this space may find the broader survey in best peptides for anti-aging research useful for context on how thin the human evidence base is across the whole category.

Three independent observations make humanin interesting to gerontologists. The first is that circulating humanin generally declines with age, reported in both mice and humans. The second, from the 2020 Aging paper, is that levels are described as surprisingly stable in the naked mole-rat, a rodent used as a model of negligible senescence. The third, from the same paper, is that in children of centenarians, who are themselves more likely to reach extreme old age, circulating humanin levels were much greater than in age-matched control subjects.

There is also a mechanistic link to a pathway already associated with lifespan. Lee and colleagues reported in Aging Cell in 2014 that the growth hormone and IGF-1 axis negatively regulates humanin. Long-lived growth-hormone-deficient Ames mice displayed elevated humanin levels, short-lived growth-hormone-transgenic mice had reduced levels, and treatment with growth hormone or IGF-I lowered circulating humanin in both mice and human subjects.

The most direct causal evidence comes from invertebrates. In the 2020 study, overexpression of humanin in C. elegans was sufficient to increase lifespan, and that effect was dependent on daf-16/FOXO, the worm orthologue of the FOXO transcription factors. Humanin transgenic mice were reported to share several phenotypes with the worm model and to show increased protection against toxic insults.

None of this establishes that raising humanin would extend human life or healthspan. High humanin in centenarian offspring is a correlation, and a peptide that tracks mitochondrial health could easily be a marker of a favorable phenotype rather than its cause. The lifespan extension demonstrated so far is in a nematode. That is a meaningful scientific result and a weak basis for any personal decision.

Native humanin has practical drawbacks as a research tool. It is a short peptide with a limited circulating half-life, it contains a cysteine at position 8 that is prone to oxidation, and its potency in cell assays is modest. Medicinal chemistry responded in the usual way, by producing variants.

The best known is S14G-humanin, commonly abbreviated HNG, in which the serine at position 14 is replaced by glycine. It was reported in early work to be considerably more potent than the wild-type peptide in neuroprotection assays, and it has since become the default molecule in rodent experiments, including the APPswe/PS1dE9 study and the middle-aged mouse healthspan work cited above. Variants substituting the cysteine at position 8 were developed to improve stability. A fusion peptide called colivelin, which attaches an activity-dependent neurotrophic factor fragment to a humanin derivative, was reported to act at femtomolar concentrations in Alzheimer's-relevant models.

The important interpretive point is that an analogue optimized for potency in a cell assay carries no guarantee of a usable safety profile in a living human. Greater potency at a receptor complex that also governs cytokine signaling and cell survival is a reason for more caution, not less, and none of these molecules has completed the pharmacokinetic and toxicology work that precedes human administration.

There are no published interventional clinical trials in which humanin or a humanin analogue was administered to humans to treat or prevent a condition. Where humanin appears in human research, it appears as a measured variable. Registered studies have, for example, quantified circulating mitochondrial-derived peptides in people with chronic obstructive pulmonary disease, and measured humanin expression in women with and without polycystic ovary syndrome. Those are observational biomarker studies, a different thing entirely from a therapeutic trial.

Consequently, humanin has no approval from the FDA, the EMA or any comparable agency. Material sold under the name humanin or HNG is sold as a research chemical, labeled for research use only. That label carries real consequences: no regulatory oversight of identity, purity, endotoxin content or sterility, no pharmacovigilance, and no recourse if a vial contains something other than what the label claims.

Legal status varies by jurisdiction. In many countries, possessing a research chemical is treated differently from supplying it for human use, and marketing it for human use is generally prohibited. Competitive athletes face an additional issue: substances without marketing approval for human therapeutic use fall under the World Anti-Doping Agency's general prohibition of non-approved substances, regardless of whether the specific peptide is named on a list.

This guide deliberately contains no dosing information, no reconstitution instructions and no protocols. That is not an oversight. For a molecule with zero human efficacy data and zero human safety data, publishing numbers would imply a level of knowledge that does not exist. If you are considering any peptide for a health concern, the appropriate step is a conversation with a licensed healthcare professional who can review your situation and the options that do have evidence behind them.

The first limitation is the species gap. Nearly everything known about humanin function comes from immortalized cell lines, nematodes, mice and rats. Cross-species differences in mitochondrial signaling, receptor expression and peptide clearance are substantial, and the mitochondrial-derived peptide field is young enough that these differences have not been systematically mapped.

The second is measurement. Circulating humanin is present at low concentrations and is assayed with antibody-based methods that differ between laboratories. Absolute values from different studies are therefore difficult to compare directly, which complicates any attempt to define a normal range, a deficiency threshold or a target level. Claims that a person's humanin is "low" should be read with that caveat in mind.

The third is the safety question raised by the biology itself. Humanin is anti-apoptotic and cytoprotective. Those properties are desirable in a stressed neuron and undesirable in a transformed cell, and the literature contains discussion of mitochondrial-derived peptides in cancer cell survival and chemoresistance. Without human toxicology data, the net effect of sustained exogenous administration in an adult human is simply unknown, and the absence of reported harms reflects the absence of trials rather than a record of safety.

What would move the field forward is unglamorous: standardized assays, pharmacokinetic characterization of humanin and its analogues in humans, formal toxicology, and then small, well-controlled phase 1 studies. Until that work exists, humanin is best understood as an important discovery in mitochondrial biology and a genuinely promising research target, not as an intervention. This article is for educational purposes only and is not medical advice. Humanin is a research peptide that is not approved for human use. Consult a healthcare professional before making any health decision.