Humanin and MOTS-c: Mitochondrial-Derived Peptides Explained

Key Takeaways

For decades the mitochondrial genome was the most thoroughly read stretch of DNA in biology: 16,569 base pairs, 37 genes, no introns and, it was assumed, no surprises. Thirteen proteins, twenty-two transfer RNAs, two ribosomal RNAs. Then in 2001 a peptide turned up inside one of the ribosomal RNA genes — a second message written through the first.

That peptide was humanin. MOTS-c followed in 2015, and six more a year later. Together they are the mitochondrial-derived peptides, and they changed what a mitochondrion is understood to be: a signalling organelle as well as a power station.

They are also now sold in vials. This article explains what the class is, what each peptide does in the laboratory, what is actually known in humans — and the single fact that should frame any decision to buy one: neither has ever been given to people in a published trial.

How a Hidden Gene Was Found

Humanin was not found by someone looking for mitochondrial peptides. It was found by a group looking for anything that could keep neurons alive. In a functional expression screen, they searched for genes that rescued neuronal cells from death caused by familial Alzheimer's disease mutations and by amyloid-beta. One short sequence did it across several different mutations. It encoded a small polypeptide, the cell secreted it into the culture medium, and the rescue depended on its exact amino acid sequence.

The sequence mapped to the mitochondrial 16S ribosomal RNA gene. Two years later a Nature paper showed what the peptide does at the molecular level and noted that the mitochondrial genome contains an identical open reading frame.

The implication took a decade to be followed up. If one rRNA gene hid a peptide, others might. In 2015 a search of the 12S rRNA gene produced MOTS-c. In 2016 an in silico search of the same 16S region as humanin produced six more candidates, named small humanin-like peptides, or SHLPs.

The Family at a Glance

Of the SHLPs, two behaved like humanin in cells — reducing apoptosis and reactive oxygen species — and SHLP2 acted as an insulin sensitiser in clamp studies. Like humanin, circulating SHLP2 declines with age. They remain a research footnote for now, and nothing below concerns them.

Humanin: The Protector

It blocks a death signal. Bax is a protein that sits inactive in the cytosol until a cell receives a death stimulus, then moves to the mitochondrial membrane and triggers the release of cytochrome c. Humanin binds Bax and prevents that move. When researchers reduced humanin with small interfering RNA, cells became more sensitive to Bax.

It improves insulin action — through the brain. In rat studies using hyperinsulinaemic clamps, humanin infused into the cerebral ventricles improved whole-body insulin sensitivity, an effect tied to hypothalamic STAT-3 and abolished when STAT-3 was blocked. Intravenous analogues reproduced it, and the effect on the liver turned out to be centrally mediated: humanin had no direct effect on isolated liver cells. A single dose of a potent analogue lowered blood glucose in diabetic rats.

It tracks the pace of ageing. Long-lived, growth-hormone-deficient Ames mice carry elevated humanin. Short-lived GH-transgenic mice carry less. In C. elegans, overexpressing humanin was enough to extend lifespan. Middle-aged mice treated twice weekly with the analogue HNG showed better metabolic healthspan markers and lower inflammatory markers.

Read that last result carefully. It is healthspan markers in mice, using an analogue. The lifespan extension is in a worm.

MOTS-c: The Metabolic Signal

MOTS-c gets a full treatment of its own in MOTS-c: Exercise in a Vial. The short version, for comparison with humanin:

It activates AMPK by an unusual route. MOTS-c inhibits the folate cycle and the purine synthesis tied to it, which leads to AMPK activation. Its primary target appears to be skeletal muscle. In mice it prevented both age-dependent and high-fat-diet-induced insulin resistance, and diet-induced obesity.

It enters the nucleus. Under metabolic stress MOTS-c moves to the nucleus in an AMPK-dependent manner and regulates nuclear genes, including those with antioxidant response elements. This was the first demonstration of a mitochondrial-encoded factor directly regulating the nuclear genome — the two genomes cross-regulating each other.

It improves physical performance in mice of every age. Treatment enhanced performance at 2, 12 and 22 months. Intermittent treatment started late in life — three times a week from 23.5 months — increased physical capacity and healthspan.

A 2026 refinement. New work shows MOTS-c improves muscle mitochondrial efficiency in a PGC-1α- and AMPK-dependent manner without changing respiratory protein content — better mitochondria, not more of them.

What Is Known in Humans

This is the section that matters, and it is short.

Humanin falls with age. Circulating levels decline in humans as in mice.

Children of centenarians have much more of it. Offspring of centenarians — who are more likely to reach extreme age themselves — carry circulating humanin well above age-matched controls.

It is lower in disease. Levels are reduced in Alzheimer's disease and in MELAS, a mitochondrial disorder.

Growth hormone and IGF-1 lower it. Treatment with either reduced circulating humanin in human subjects — the one human finding that comes from an intervention, though the intervention was not humanin.

A MOTS-c variant raises diabetes risk — in sedentary men. An Asian-specific mitochondrial DNA variant changes one amino acid in MOTS-c (K14Q). Across three cohorts totalling 27,527 people, men carrying it had a higher prevalence of type 2 diabetes; women did not. In one cohort the excess appeared only in men in the lowest third of physical activity. The variant peptide failed to protect mice the way normal MOTS-c did. This is the strongest human evidence that MOTS-c matters, and it is genetic, not interventional.

MOTS-c tracks muscle signalling. Plasma MOTS-c correlates inversely with myostatin in human subjects.

Exercise raises both. Acute high-intensity exercise increases humanin and MOTS-c in skeletal muscle and plasma.

What No One Has Tested

• Giving either peptide to a person. No published trial has administered humanin, HNG or MOTS-c to humans. There is no human pharmacokinetic study, no dose-finding study, no safety study.
• Whether low levels cause anything. Lower humanin in Alzheimer's disease and higher humanin in long-lived families are correlations. A marker that follows ageing is not shown to drive it.
• Whether raising levels helps. Reviews in the field say that strategies to increase these peptides "might" have broad benefits. That is a hypothesis.
• Long-term safety of an anti-apoptotic peptide. Blocking Bax is protective in a neuron and an open question in a tumour.

There is even uncertainty about the basic physiology. Whether exercise training changes resting peptide levels is conflicting across studies. And the 2026 muscle study found no arterio-venous difference in MOTS-c across an exercising human leg — meaning muscle may not be where the circulating peptide comes from.

Why the Exercise Finding Changes the Question

The commercial pitch for these peptides is that they deliver the signal of exercise without the exercise. The human data point the other way: exercise is the one thing known to deliver the signal.

The K14Q finding sharpens this. Men with a less active form of MOTS-c had more diabetes — but only if they were inactive. Physical activity appeared to compensate for a weaker peptide. And in mouse work combining exercise with MOTS-c treatment, the two acted through pathways that were partly distinct and partly overlapping. Neither substitutes cleanly for the other.

So the useful question is not "which vial replaces training" but "what, if anything, adds to it" — and that is the question the Advanced Mitochondrial Stack is built to answer.

How They Compare With the Other "Exercise Mimetic"

A small molecule called SLU-PP-332 is marketed in the same breath as these peptides and reaches a related destination by a different road: nuclear receptors called ERRs rather than AMPK. It is not a peptide and not mitochondrial-derived. Its evidence is at the same stage — strong in mice, absent in humans — and is reviewed in SLU-PP-332: The Exercise Mimetic Research Update.

Frequently Asked Questions

Are humanin and MOTS-c the same kind of thing as BPC-157 or other research peptides? No. Most research peptides are fragments or analogues of proteins encoded in nuclear DNA. Humanin and MOTS-c are encoded in the mitochondrial genome itself, which is why they are studied as signals of mitochondrial status.

Can I measure my humanin or MOTS-c level? Not usefully. The assays are research tools without clinical reference ranges, and no study shows that acting on a result changes an outcome.

Is there a human dose for humanin? No. Humanin has no published human administration study. Animal work used the analogue HNG by injection or infused the peptide into the brain.

Is there a human dose for MOTS-c? Not from a trial. The figures in circulation, including the convention recorded on our MOTS-c profile, are extrapolated from mouse work.

Which one is more promising? They answer different questions. Humanin has the stronger ageing-biology story — age decline, centenarian families, a worm lifespan result. MOTS-c has the stronger metabolic story and the only human genetic evidence. Neither has human intervention data.

What should I actually do? Train, with some of it hard. Then, if you want a structured plan that is honest about which layers have evidence, see the Mito-Peptide Advanced protocol.

Related Research

• Humanin
• MOTS-c
• MOTS-c: Exercise in a Vial
• The Advanced Mitochondrial Stack: Mito-Peptides + Cofactors
• SLU-PP-332: The Exercise Mimetic Research Update
• The Mito-Peptide Advanced Protocol