Humanin

Overview

Humanin (HN) is a small peptide with an unusual origin: rather than being encoded by the nuclear genome like most cellular peptides, its sequence is found within mitochondrial DNA, specifically a short open reading frame in the 16S ribosomal RNA region (the MT-RNR2 gene). It was the first member of a small and growing class of molecules known as mitochondrial-derived peptides (MDPs), a family that also includes MOTS-c and the SHLP peptides, and it remains the best-characterized of the group.

Humanin was discovered in 2001 by Hashimoto and colleagues at Keio University, who were screening for genes that could rescue neurons from cell death caused by familial Alzheimer's disease mutations and amyloid-beta. The peptide they identified did exactly that in cell models, which is why much of the early literature frames humanin as a neuroprotective and cytoprotective factor.

Most of what is known about humanin still comes from cell-culture and animal experiments. In those models it has been described as anti-apoptotic (it helps cells survive a range of stresses) and as a participant in metabolic and aging-related signaling. Human data are largely observational: circulating humanin can be measured in people, and its levels decline with age. These descriptions should be read as research hypotheses supported mainly by preclinical data, not as established human effects.

Chemistry and structure

Humanin is a single short linear peptide. A notable quirk is that it exists in two reported forms depending on where it is translated:

• A 24-amino-acid form (sequence MAPRGFSCLLLLTSEIDLPVKRRA) when produced in the cytosol from a nuclear-style reading of the open reading frame.
• A shorter form (about 21 residues) when translated inside the mitochondrion using the mitochondrial genetic code, which terminates earlier.

Both forms have been reported to have biological activity in experimental systems. The peptide combines a hydrophobic core (a run of leucine residues) with basic residues near the C-terminus, and it contains a cysteine that has been implicated in self-association.

Analogs

Several synthetic analogs have been engineered, the most cited being HNG, also called S14G-humanin, in which serine at position 14 is replaced by glycine. In laboratory assays this single substitution has been reported to make the peptide far more potent than native humanin (figures on the order of a 1000-fold increase in some neuroprotection assays are quoted in the literature). Other analogs (for example colivelin, a humanin-derived fusion peptide) have also been studied preclinically. These analogs are distinct molecules and their data should not be read as data on native humanin.

Mechanism of action

The proposed mechanisms of humanin come almost entirely from in vitro and animal experiments. The most frequently cited themes are:

• Anti-apoptotic (cell-survival) signaling. Humanin has been reported to bind and neutralize pro-apoptotic proteins such as BAX and tBID, helping keep mitochondria from triggering programmed cell death, and to interact with IGFBP-3 in a way that regulates survival versus apoptosis (Zhu et al., 2022).
• A cell-surface receptor complex. Secreted humanin has been described as signaling through a tripartite receptor comprising CNTFR-alpha, WSX-1, and gp130, activating downstream STAT3 and pro-survival kinase pathways such as PI3K/Akt (Hashimoto et al., 2009).
• Cytoprotection against diverse stresses. In cells, humanin has been reported to protect against oxidative stress, serum starvation, hypoxia, and other insults, which is why it is often framed as a general stress-resistance factor (Yen et al., 2013).
• Metabolic signaling. In rodent studies, humanin has been linked to improved hepatic and peripheral insulin sensitivity, in part through hypothalamic STAT3 activation, and to pancreatic beta-cell survival (Cobb et al., 2016; Boutari et al., 2022).

Conceptually this places humanin near other peptides studied for mitochondrial or longevity-related endpoints, such as MOTS-c and the mitochondria-targeted peptide SS-31; each has its own distinct and separately limited evidence base.

Research and evidence

The humanin evidence base is weighted heavily toward cell and animal work, with human data limited to observational and association studies rather than therapeutic trials. The table labels evidence type explicitly.

Neuroprotection — the origin story

Humanin was defined by its neuroprotective activity: in the foundational study it rescued neuronal cells from death induced by a wide spectrum of familial Alzheimer's disease genes and by amyloid-beta (Hashimoto et al., 2001). Subsequent cell and rodent work, much of it using the more potent HNG / S14G analog, reported protection against amyloid-beta toxicity, ischemia-reperfusion injury, and other insults, and improvements in learning and memory measures in animal models (e.g., Zhang et al., 2016). These are model-system findings, not demonstrations of benefit in people with disease.

Metabolism — rodent and cell evidence

In mice and cells, humanin has been associated with improved insulin sensitivity, effects on glucose handling, and beta-cell survival, with one review summarizing that chronic administration delayed or prevented diabetes onset in animal models (Boutari et al., 2022; Cobb et al., 2016). These findings are animal- and cell-based and have not been confirmed in controlled human trials.

Aging and longevity — model organisms plus human associations

Humanin is frequently discussed as a longevity-associated peptide for two reasons. First, circulating humanin declines with age in humans and rodents, and some long-lived models (for example GH-deficient Ames dwarf mice, and reportedly the naked mole-rat) show relatively higher or more stable levels (Cobb et al., 2016; Yen et al., 2020). Second, in model organisms, overexpressing humanin increased lifespan in C. elegans in a daf-16/FOXO-dependent manner, and humanin treatment prevented age-related cognitive decline in mice (Yen et al., 2020). Human work also includes reports of higher humanin in the offspring of centenarians. These are correlational human observations and model-organism experiments. They do not show that administering humanin extends human life or healthspan.

Safety and risks

Reliable human safety information for administered humanin is largely absent.

• Human safety is essentially unknown. Because there are no completed controlled human trials of the administered peptide, there is no robust human safety, tolerability, or pharmacokinetic dataset, and long-term effects are uncharacterized.
• Animal data do not establish human safety. Favorable findings in rodents and cells cannot be assumed to translate to people.
• A note of caution from cancer biology. Because humanin is broadly anti-apoptotic (it helps cells survive), some experimental work has raised the question of whether it could, in principle, also support the survival of unwanted cells; for example, humanin has been reported to promote tumor progression in some experimental cancer models. This is a preclinical signal, not an established human risk, but it shows how little is settled.
• Unregulated product quality. Material sold as humanin is typically supplied as an unregulated research chemical. Identity, purity, dosing accuracy, and contamination are real concerns independent of any intrinsic property of the peptide itself.

Status and regulation

• Drug approval. Humanin is not approved as a medicine by the FDA, EMA, or other major regulators. It is an early-stage research molecule, not a marketed therapeutic.
• Supplement status. Peptides of this type generally do not meet the legal definition of a dietary ingredient, and humanin is not a lawfully marketed dietary supplement. Material is commonly labeled "for research use only."
• Sport. Anti-doping policy on research peptides has been evolving; athletes subject to the anti-doping code should verify the current status of humanin and its analogs directly with WADA/USADA rather than assuming it is permitted.

How it compares

Humanin is usually grouped with other peptides studied in the Longevity & cellular health space. The most common comparisons are to the fellow mitochondrial-derived peptide MOTS-c and the synthetic mitochondria-targeted peptide SS-31 (elamipretide). They are mechanistically distinct, and the strength of their human evidence differs.

Bottom line: Among these three, SS-31 (elamipretide) is by far the most advanced: it has been studied in multiple registered human clinical trials and received FDA accelerated approval in September 2025 for Barth syndrome, although it is not approved for longevity or the other uses discussed here. Humanin and MOTS-c rest on far weaker foundations, mainly animal and observational human data, with no approved use and no completed human therapeutic trials of the administered peptide. Membership in the same "longevity peptide" category does not imply comparable evidence.

Common misconceptions

• "Humanin is a proven anti-aging or Alzheimer's therapy." The supportive evidence is overwhelmingly from cells and animals. There is no completed human trial showing it treats aging, Alzheimer's, diabetes, or any other condition.
• "Human studies prove humanin works." The human data are mostly observational (measuring the body's own humanin and noting that it declines with age) or association studies, for example in centenarian offspring. These are not trials of an administered drug.
• "It's natural, so it must be safe." Humanin is indeed produced by human mitochondria, but administering a synthetic version as a research chemical at non-physiological levels is a different matter, and its human safety is uncharacterized. Its broad anti-apoptotic activity even raises preclinical questions about effects on unwanted cells.
• "Humanin and HNG are the same thing." HNG (S14G-humanin) is a modified analog reported to be much more potent in assays; data on one are not automatically data on the other.
• "All mitochondrial peptides are interchangeable." Humanin, MOTS-c, and the SHLP peptides are distinct molecules with different sequences and proposed roles; conflating them is inaccurate.