MOTS-c: The Exercise Mimetic That Outlived Its Sponsor

A 2 a.m. PubMed epiphany, sold for $90 a month.

The viral X post arrived in the usual costume. Someone discovered something at 2 a.m. on PubMed and it "changed everything." Sixteen amino acids. Old mice running twice as far. A peptide your mitochondria already make. The closest thing to exercise in a needle.

The peptide is MOTS-c. The 2015 paper is real.¹ The 2021 follow-up showing aged mice doubling their treadmill time is also real.² The human data showing skeletal muscle MOTS-c rising 11.9-fold after a bout of high-intensity cycling is real, too.² And underneath all of it is the line the influencer post never quite gets to: the only company that ever tried to bring MOTS-c to humans as a drug went bankrupt in 2023, and their one placebo-controlled trial in 20 patients delivered a 4-week liver-fat change of negative five percent on drug versus negative four-point-nine percent on placebo.⁵

That is the entire MOTS-c story, compressed. Beautiful mouse biology. A 16-amino-acid signaling peptide with a genuinely novel mitochondrial-to-nuclear mechanism. And, sitting next to it, a near-total absence of credible human therapeutic data, a sponsor that no longer exists, and a gray market quietly selling vials of "research-only" peptide to people who saw a thread.

I want to walk through what is actually known, what is being aggressively oversold, and why this peptide has become a kind of Rorschach test for how the longevity space metabolizes preclinical data.

A peptide encoded in mitochondrial DNA, which is itself a surprise.

MOTS-c stands for "Mitochondrial Open Reading frame of the 12S rRNA-c." Pinchas Cohen's lab at USC discovered it in 2015, embedded inside a small open reading frame within the mitochondrial 12S ribosomal RNA gene.¹ Sixteen amino acids long. Sequence MRWQEMGYIFYPRKLR. And — this is the interesting part — its genetic instructions live not in the cell's nucleus, where almost all human proteins are encoded, but in the tiny circular genome of the mitochondrion itself.

This places MOTS-c in a small and contested category called mitochondrial-derived peptides, or MDPs. The first one, humanin, was discovered in 2001 and has spent the subsequent 25 years failing to become a therapeutic.¹¹ MOTS-c is the most studied of the next generation. There is also a family of SHLPs — small humanin-like peptides — that you can mostly ignore unless you want to read review articles.

What MOTS-c does, mechanically, is moderately elegant. It interferes with the folate-methionine cycle, which is a one-carbon metabolism pathway that ordinarily helps build new purine bases for DNA. When MOTS-c jams this pathway, a metabolic intermediate called AICAR accumulates. AICAR is the prototype activator of AMPK — the cellular energy sensor that exercise, metformin, and caloric restriction all converge on.¹ The pharmacology is essentially "endogenous methotrexate, but with AMPK activation as the prize instead of a chemotherapy side effect."

Under metabolic stress, MOTS-c also translocates from the cytoplasm into the nucleus, where it interacts with NRF2 and binds antioxidant response elements to switch on stress-response genes including HO-1, NQO1, and the glutathione machinery.³ This is the part of the X post that gets called "reprograms your genes," and to be fair, the mechanism is genuinely there. A peptide whose instructions live in mitochondrial DNA, but which physically enters the nucleus to modulate nuclear gene expression, is a real and novel piece of cell biology. The retrograde signal from the mitochondrion to the nucleus has been a major theme in mitochondrial biology for two decades, and MOTS-c is one of the most concrete molecular candidates for that role.

So when I say the biology is interesting, I mean it. The problem is not the biology. The problem is what happens between mice and humans.

Two papers, one lab, and the legend behind "running capacity doubled."

The viral claim that "running capacity doubled in old mice" is real. It just is not in the paper everyone cites.

The 2015 Lee et al. paper in Cell Metabolism — the PubMed ID people throw around in posts — does not contain treadmill data.¹ What it contains is something different and arguably more important: a mouse model of high-fat-diet-induced obesity in which 0.5 milligrams per kilogram per day of MOTS-c, injected intraperitoneally for seven to eight weeks, prevented weight gain despite identical caloric intake. The same regimen prevented the development of insulin resistance, improved glucose tolerance, and reduced fasting insulin. The mechanism was traced through the folate-AICAR-AMPK axis I described above. Cohorts were on the order of eight to ten mice per group, which is standard for that journal and era.

The "doubled running capacity" headline comes from the 2021 Reynolds et al. paper in Nature Communications, also from the Cohen lab.² Young, middle-aged, and 22-month-old mice — the rodent equivalent of late-life humans — were treated with MOTS-c and run to exhaustion on a treadmill. The aged mice approximately doubled their time-to-exhaustion versus vehicle controls. The effect was largest in the oldest cohort, which is the part influencers correctly identify as the most provocative finding in the field.

Both papers are from the same lab. The mouse data, taken on its own terms, is real and reasonably well-controlled. What it is not is a substitute for a human clinical trial of administered MOTS-c, and the field has consistently elided that distinction.

One trial of 20 patients. The result tied with placebo.

This is the part of the story that almost never appears in the threads. There is, as of May 2026, exactly one completed, placebo-controlled human study of a MOTS-c-related drug. It was run by a company called CohBar, the molecule was a modified analog called CB4211, and the result was, by any honest reading, a failure.

CohBar dissolved in late 2023.⁶ A reverse merger attempt with a private oncology company collapsed when Nasdaq rejected the listing structure. The MOTS-c clinical program ended with the company. There is, to my knowledge, one small academic trial of MOTS-c for insulin sensitivity registered as of 2025, but no completed Phase 2 or Phase 3 RCT of any MOTS-c product exists. There has never been an FDA-approved MOTS-c product.

The other piece of the human dataset is the observation, repeated across several small studies, that plasma MOTS-c levels appear to be modestly lower in older adults than in younger ones.^7,8 The published effect sizes are on the order of 21% lower in older versus younger cohorts — not a "crash." Skeletal muscle MOTS-c, paradoxically, has been reported to increase with age in healthy men, possibly reflecting a shift in muscle fiber composition with sarcopenia.⁷ Whatever the true age-decline is, it has been measured almost entirely by ELISA, and ELISA quantification of MOTS-c has not always been confirmable by mass spectrometry. The peptide is also extremely unstable — degradation of 85% to 90% within two to three hours at room temperature has been reported.¹⁰ If your assay is detecting something at all, what it is detecting and how reliably it is detecting it remain open questions.

The discoverer, the patents, the company, the silence.

This is where I get less polite. Pinchas Cohen is the dean of the Leonard Davis School of Gerontology at USC. He is also the senior author on the foundational MOTS-c papers, the inventor on the foundational MOTS-c patents, and the co-founder of CohBar — the publicly traded biotech that licensed those patents and tried to bring MOTS-c to humans as a drug. None of this is hidden. It is all disclosed in the academic papers, in the regulatory filings, and in the company prospectus.

The problem is that it is essentially never disclosed in the popular coverage. Influencer threads cite the 2015 paper as if it were arms-length academic science. They cite the 2021 paper the same way. They do not mention that the lab that did the work, the company that commercialized it, and the patents that protect it share a common founder. They especially do not mention that this same chain of commercial interest produced exactly one placebo-controlled human trial, and that the trial's headline efficacy result was indistinguishable from placebo. That is the relevant clinical fact about MOTS-c.

I am not making a corruption argument here. Translational research depends on commercial development, and basic-science labs spinning out drug companies is how most therapeutics reach humans. What I am making is a credibility argument: when the only company that ever ran a controlled human trial of your peptide produces a result that ties with placebo, and then goes bankrupt the next year, the appropriate update to your priors is not "exciting opportunity." The appropriate update is "the human data did not survive the human trial."

$60 a vial, "not for human use," and the regulatory shrug.

None of this has slowed the gray market. MOTS-c is sold widely as a "research peptide, not for human use" on a long list of online vendors. Common pricing in 2026 runs roughly $6.50 to $12 per milligram, with 10 mg vials at $60 to $120 and larger vials proportionally. At the community-standard dose of about 10 mg per week, the monthly cost lands somewhere between $40 and $95.

The FDA does not consider MOTS-c an approved drug, and it cannot be obtained from US 503A compounding pharmacies. During the FDA's 2023 peptide compounding review — which I covered in Issue 012 — MOTS-c was nominated for evaluation, but the nomination was subsequently withdrawn. It therefore sits in an odd regulatory limbo: not formally banned for compounding, not legitimately available through compounding, and not approved through the standard drug pathway.¹⁴

WADA, the World Anti-Doping Agency, is more decisive. MOTS-c is explicitly prohibited at all times for athletes under Section S4.4.1 of the 2026 prohibited list, which covers AMPK activators and metabolic modulators.¹³ USADA has issued specific athlete advisories about it. No therapeutic use exemption is possible because there is no approved therapeutic use.

Quality on the gray market is, as ever, a function of which vendor you trust. The better ones publish third-party certificates of analysis from Janoshik or MZ Biolabs. The worse ones publish nothing. A peptide that degrades 85% within hours at room temperature is unusually unforgiving of bad cold-chain handling, and there is no oversight here at all. The handful of MOTS-c-related adverse events that have been described — mostly persistent injection-site reactions — track with what CohBar saw in CB4211. There is no long-term human safety data because no long-term human trial has been done.

A beautiful piece of mouse biology that has not become medicine.