MOTS-c Explained: Mitochondrial Science and the Human-Evidence Gap

Three takeaways

1. MOTS-c is real. It is a 16-amino-acid peptide encoded inside mitochondrial DNA, discovered in 2015, that activates the cell’s energy sensor AMPK and acts mainly on skeletal muscle.
2. The human evidence is endogenous, not interventional. Human MOTS-c appears in blood, responds to exercise, and a natural genetic variant is associated with type 2 diabetes risk in Japanese men. But no published, peer-reviewed human trial of administered MOTS-c had reported results as of our check on 13 October 2026.
3. Commercial injections are far outside verified evidence — and flagged. The FDA’s compounding page publishes potential-significant-safety-risk language for MOTS-c (immunogenicity, impurities, no human-exposure data); a July 2026 advisory vote to permit compounding is non-binding and pending final FDA action; MOTS-c is also prohibited in sport.

Evidence and status panel

Table of contents

1. [A peptide hiding in mitochondrial DNA](#1)
2. [What endogenous MOTS-c does](#2)
3. [What laboratory and animal studies show](#3)
4. [What the human association studies show](#4)
5. [Giving MOTS-c: what has actually been tested in humans](#5)
6. [The safety gaps](#6)
7. [Where this leaves the reader](#7)

Main article

1. A peptide hiding in mitochondrial DNA

Mitochondria are best known as the cell’s power plants, and most mitochondrial genes build that machinery. The idea that mitochondria also signal is less famous than the supplement ads, but it has real primary sources. In 2001, researchers reported humanin, a protective peptide encoded in what was thought to be a non-coding stretch of mitochondrial DNA [11]. In 2016, the same research lineage described small humanin-like peptides, or SHLPs [10]. And in 2015, Changhan Lee, Pinchas Cohen and colleagues reported the discovery in Cell Metabolism: a short open reading frame inside the gene for mitochondrial 12S ribosomal RNA that encodes a 16-amino-acid peptide they named MOTS-c — mitochondrial open reading frame of the 12S rRNA type-c [1].

Two details matter for reading everything else about MOTS-c. First, it is tiny — 16 amino acids, sequence MRWQEMGYIFYPRKLR [17] — small enough to be synthesised commercially, which is exactly what happened after the paper appeared. Second, the discovery paper’s own framing was metabolic: MOTS-c’s primary target organ appeared to be skeletal muscle, and its cellular action was to inhibit the folate cycle and its linked de novo purine synthesis, leading to activation of AMPK, a master metabolic sensor [1]. That mechanism is well replicated in cells. What “therefore injections will fix your metabolism” is not.

2. What endogenous MOTS-c does

MOTS-c circulates in measurable amounts in blood [1][4][5] and changes with the body’s energy state. The strongest human data concern exercise. A 2021 Nature Communications study from the discovery group reported that exercise induces MOTS-c expression in human skeletal muscle and circulation [2]; within the same paper, treatment of mice — including mice that were 22 months old — with MOTS-c significantly enhanced physical performance, and intermittent treatment started late in life increased physical capacity and healthspan [2]. Separately, a randomised Swedish study measured MOTS-c after acute exercise in 30 healthy adults: after 45 minutes of cycling, circulating MOTS-c showed a trend to increase, while the related peptide humanin rose clearly and resistance exercise did neither [3]. In that cohort, resting MOTS-c levels did not correlate with fitness measures such as VO2max or muscle mitochondrial DNA copy number [3].

So “exercise raises MOTS-c” is a fair summary of the human signal, with two honest caveats: the circulating response is small and inconsistent across the small studies to date, and — critically — endogenous MOTS-c is a correlate of exercise, not proof that injecting MOTS-c replicates exercise. Confusing those two ideas is the load-bearing error behind most commercial claims. Exercise has decades of human outcome evidence behind it; MOTS-c injections have none in the published record as of our check date (see section 5).

3. What laboratory and animal studies show

The mouse work is consistent and, by peptide-treatment standards, quite varied:

• The 2015 discovery study reported that MOTS-c treatment prevented age-dependent and high-fat-diet-induced insulin resistance and diet-induced obesity in mice [1].
• The 2021 exercise paper extended this to physical performance across young, middle-aged and old mice, with late-life treatment improving capacity and healthspan [2].
• Models of oestrogen loss: MOTS-c treatment prevented ovariectomy-induced weight gain and insulin dysfunction while activating brown fat [8], and a separate study reported suppressed ovariectomy-induced bone loss via AMPK [9].

Readers should hold three cautions in view: these are mice given synthetic peptide by injection on researcher-chosen schedules; effects were sex-specific in the variant study below [6]; and no mouse work validates long-term human exposure. The discovery lineage is also commercially entangled: the 2015 and 2021 papers disclose that key authors held consultancy and shareholder roles at CohBar, a company formed to develop mitochondrial-peptide drugs [2][6]. Disclosed interests are not disqualifying, but independent replication remains thin. A mechanism in a mouse answers a mechanism question, not a treatment question — a distinction our guide to reading a peptide study unpacks in detail (article 02).

4. What the human association studies show

Human observational work on naturally circulating MOTS-c exists — and it is messy, which is the most instructive part.

• A Chilean cross-sectional study (10 lean, 10 obese adults) found plasma MOTS-c unchanged in obesity overall, with the correlation between MOTS-c and insulin-resistance indexes present only in lean participants [4].
• A 2025–2026 US study of 54 adults found circulating MOTS-c higher in adults with obesity than in lean controls, with BMI and insulin resistance independently predicting levels, a biphasic relationship with insulin resistance, and — notably — no change in MOTS-c six months after bariatric weight loss. The authors read this as a possible compensatory response, and labelled their study exploratory [5].
• Two such small studies cannot settle the direction of the relationship, and their absolute values are not comparable across assays.
• The strongest human dataset is genetic. The m.1382A>C mitochondrial variant, found in Northeast Asian populations, swaps the 14th amino acid of MOTS-c (K14Q). In a meta-analysis of three cohorts totalling 27,527 people, men carrying the variant had higher prevalence of type 2 diabetes, with the effect concentrated in men with low physical activity; female carriers were unaffected. In mice, the K14Q version of the peptide was less effective at improving metabolic markers than the standard MOTS-c sequence [6].

Notice what this does and does not show. It shows that MOTS-c biology plausibly matters to human metabolism — people whose MOTS-c sequence works less well carry more diabetes risk in specific subgroups. It does not show that raising MOTS-c levels in someone else would reduce their risk, let alone that a compounded injection does. It also corrects an earlier story: a 2015 hypothesis paper proposed this same Japanese-enriched variant as a contributor to Japanese longevity [7]; the larger later work reframed it as diabetes-associated in men [6]. Early enthusiasm was premature. That correction pattern — hypothesis, then hype, then a humbling cohort study — is exactly the pattern the approved-versus-research distinction is designed to catch (article 03).

5. Giving MOTS-c: what has actually been tested in humans

This is where most MOTS-c material online goes quiet, so it deserves care. As of 13 October 2026, we found no published, peer-reviewed trial of administered MOTS-c (or an analogue) reporting metabolic outcomes in humans. What the trial registry and available records show is narrower [12]:

Study card — NCT07505745 (ClinicalTrials.gov)
Design: Phase 2a, randomised, double-blind, placebo-controlled. Participants: 120 adults (up to age 65) with prediabetes and overweight/obesity. Condition: prediabetes with overweight/obesity. Intervention: investigational MOTS-c, subcutaneous, fixed daily schedule for 12 weeks (dose value listed in registry record). Comparator: placebo, quadruple masking; both arms receive standardised lifestyle counselling. Duration: 12 weeks; safety follow-up to week 16. Primary endpoints: change in OGTT-derived insulin sensitivity (Matsuda index); incidence of treatment-emergent adverse events. Secondary: HbA1c, fasting and 2-hour glucose, and immunogenicity (anti-drug antibodies). Finding: no results posted — recruiting; estimated completion February 2027. Adverse events: a prospectively tracked primary endpoint; none reported yet. Limitations: sponsor-run (Hudson Biotech); single site in Guangdong, China; results pending. Funding: sponsor as listed on registry. Citation: ClinicalTrials.gov NCT07505745, last updated 1 April 2026 [12].

Study card — NCT03998514 (CB4211, a MOTS-c analogue)
Design: Phase 1a/1b, randomised, double-blind, placebo-controlled, single and multiple ascending doses. Participants: 88 — healthy non-obese adults and adults with non-alcoholic fatty liver disease, US sites. Intervention: CB4211, a modified analogue of MOTS-c, subcutaneous. Comparator: placebo. Endpoints: adverse events, tolerability, pharmacokinetics. Finding: completed 19 April 2021; no results posted on the registry and no peer-reviewed publication found on our check. Adverse events: the independent ADDF Cognitive Vitality report describes the short study as safe overall but common persistent injection-site reactions [17]. Funding: CohBar, Inc. Limitations: an analogue is not native MOTS-c; unpublished results cannot be weighed; the sponsor’s MOTS-c-based development was later discontinued with the company’s dissolution [17]. Citation: ClinicalTrials.gov NCT03998514 [12]; ADDF report [17].

One distinction matters for reading online claims: the mouse experiments used synthetic versions of the native MOTS-c sequence or the natural K14Q variant, while the only completed human dosing study to date involved a modified analogue (CB4211), and the ongoing Phase 2a is the first registered MOTS-c dosing study we located. Commercial vials sold today are neither of these verified contexts. Per the FDA’s current assessment, it has identified no human-exposure data for MOTS-c products made by compounding [13]. Equating people’s naturally circulating MOTS-c — or any mouse result — with the claimed benefits of commercially injected MOTS-c is a category error this article exists to prevent.

6. The safety gaps

• The FDA compounding flag. FDA’s compounding page, current as of 22 April 2026 and checked by us on 13 October 2026, states that compounded drugs containing MOTS-c “may pose significant risk for immunogenicity,” carry complexities from peptide-related impurities and active-ingredient characterisation, and that the agency has identified no human exposure data and “lacks important information regarding any safety issues raised by MOTS-c, including whether it would cause harm if administered to humans” [13].
• Where compounding stands now. MOTS-c had sat on FDA’s Category 2 list of substances with significant safety risks; in April 2026 the agency removed twelve peptide substances from that list and referred them for review [16]. FDA staff briefing documents for the July 2026 committee meeting recommended against adding any of the seven peptides, including MOTS-c, to the 503A Bulks List [16]; the committee nonetheless voted, non-bindingly, to recommend six of the seven — MOTS-c included — for the Bulks List [14][16][18]. A committee recommendation is not authorisation: final FDA rulemaking had not taken effect as of our check date, and independent commentary has questioned the review process itself [18]. Until an official FDA list change is published, USADA’s guidance puts it plainly: during evaluation, “there is no legal basis to include MOTS-c in any compounded medication” [15].
• Prohibited in sport. MOTS-c is banned at all times under WADA’s Prohibited List (section 4.4, AMPK activators), with USADA noting that a therapeutic-use exemption is unlikely because there is no approved therapeutic use [15].
• The open questions. No long-term human safety data exist. Dose–response in humans is unknown. Doses that sensitise insulin signalling raise theoretical interaction questions with glucose-lowering medication that no study addresses. Immunogenicity — one of FDA’s stated risk categories — is precisely what the ongoing trial monitors [12][13]. And “research use only” labelling, as our regulatory explainer explains, is a purchase disclaimer, not evidence of consumer safety (article 03).

7. Where this leaves the reader

MOTS-c earns the “promising but uncertain” label honestly: a real molecule, a plausible and partly replicated mechanism, consistent mouse benefits spanning metabolism, physical capacity, bone and oestrogen-loss models, and genuinely intriguing human genetics. What it does not yet have is any published human trial showing that administering it helps anyone — and the regulator closest to the current market says it lacks the basic information to know whether compounded MOTS-c would harm people [13].

For readers drawn by the exercise story, the verified direction runs the other way: exercise raises MOTS-c and has direct human outcome evidence. Both halves of that equation are covered in our guide to mitochondria, exercise and metabolic health (article 09). We track the MOTS-c evidence in this publication’s mitochondrial science coverage; the mitochondrial science topic hub collects related guides and future updates to this article.

Known limitations

• The animal findings rest on mouse models with researcher-chosen doses, synthetic peptide and short-to-medium exposure windows; effects were sex-specific in the variant study.
• Human association data come from small cross-sectional cohorts whose absolute values are not comparable across assays, and whose direction is inconsistent.
• The human genetics signal (K14Q) is population-specific (Northeast Asian cohorts) and observational.
• The only completed human dosing study involved an analogue (CB4211) with unpublished results; the first registered MOTS-c dosing trial had not reported as of 13 October 2026.
• The regulatory picture is time-limited: checked 13 October 2026, against a FDA compounding page dated 22 April 2026 and a non-binding July 2026 committee vote; South African status was not verified in this article.

FAQ

Is MOTS-c approved as a medicine anywhere? We found no approval for MOTS-c as a human medicine in any jurisdiction as of 13 October 2026, and USADA notes there is no approved therapeutic use [15]. In the US, its compounding status remains unresolved pending FDA rulemaking after a non-binding July 2026 advisory recommendation [12][14].

Have humans been given MOTS-c in a trial? A Phase 1a/1b trial of CB4211, a modified analogue, completed in 2021 without published or registry-posted results [12][17]. A Phase 2a trial of MOTS-c itself in adults with prediabetes and overweight/obesity was recruiting in China as of our check, with results not expected before 2027 [12].

If exercise raises MOTS-c, why not just inject it? Because the equivalence is asserted, not demonstrated. Exercise changed MOTS-c modestly in small human studies and delivers extensive proven benefits; injections of MOTS-c have no published human outcome data and an active safety flag at FDA [3][13].

What does “research use only” mean? It is wording on products sold for laboratory use, not a safety certification. It does not establish that consumer use is legal, safe or effective, and it does not exempt sellers or users from medicine or anti-doping rules [14][15].

Sources

1. Lee C, et al. Cell Metabolism 2015;21(3):443–454. PubMed 25738459
2. Reynolds JC, et al. Nature Communications 2021;12:470. PubMed 33473109
3. von Walden F, et al. Journal of Applied Physiology 2021;131(3):1035–1042. PubMed 34351816
4. Cataldo LR, et al. J Investig Med 2018;66(6):1019–1022. PubMed 29593067
5. Yoon SH, et al. J Clin Transl Endocrinol 2026;43:100429. PubMed 41551324
6. Zempo H, et al. Aging (Albany NY) 2021;13(2):1692–1717. PubMed 33468709
7. Fuku N, et al. Aging Cell 2015;14(6):921–923. PubMed 26289118
8. Lu H, et al. J Mol Med 2019;97(4):473–485. PubMed 30725119
9. Ming W, et al. Biochem Biophys Res Commun 2016;476(4):412–419. PubMed 27237975
10. Cobb LJ, et al. Aging (Albany NY) 2016;8(4):796–809. PubMed 27070352
11. Lee C, Kim KH, Cohen P. Free Radic Biol Med 2016;100:182–187 (humanin reported in 2001 per this review). PubMed 27216708
12. ClinicalTrials.gov — NCT07505745 and NCT03998514, accessed 13 October 2026.
13. FDA, Certain Bulk Drug Substances Used in Compounding That May Present Significant Safety Risks (content current as of 22 April 2026; checked 13 October 2026).
14. FDA, July 23–24, 2026 PCAC meeting record; committee vote summaries via NCPA (31 July 2026) and The BMJ 2026;394:bmj-2026-100422. Checked 13 October 2026.
15. USADA, “What is the MOTS-c peptide?” (checked 13 October 2026).
16. Drug Topics, “FDA staff review recommends against adding 7 peptides to the 503A bulk list” (July 2026); regulatory timeline reporting of April 2026 Category 2 removals. Checked 13 October 2026.
17. Alzheimer’s Drug Discovery Foundation, Cognitive Vitality report: MOTS-c (updated 17 September 2025).
18. Health Affairs, “FDA Advisory Committee’s Vote May Open A Drug-Compounding Back Door For Unapproved Peptides” (2026). Checked 13 October 2026.

Found an error? Report a correction.

Related articles

• Mitochondria, Exercise and Metabolic Health: What We Know — exercise adaptations with human outcome evidence (article 09).
• How to Read a Peptide Study Without Falling for the Hype — cell/animal/human hierarchy and study cards (article 02).
• Approved Medicines, Clinical Trials and “Research Use Only” — why an investigational label is not consumer safety (article 03).

Newsletter: The Peptide Science Brief — fortnightly research updates and evidence checks. Sign up at /newsletter/. No personalised medical advice; see medical disclaimer and editorial policy.

Suggest a correction