Mitochondrial peptides: a complete guide to MOTS-c & SS-31

What are mitochondrial and longevity peptides?

Four very different molecules are sold under the ‘longevity peptides’ label, and they share one argument: aging comes with a cellular energy deficit, and the mitochondrion is where to intervene¹⁰. MOTS-c and humanin are peptides encoded in mitochondrial DNA itself¹². SS-31 (elamipretide) is a synthetic tetrapeptide designed to bind the inner mitochondrial membrane¹⁰¹². Epitalon is a synthetic tetrapeptide modeled on a pineal gland extract²⁶. And NAD+ is not a peptide at all but a dinucleotide, a coenzyme whose levels fall with age and that people try to restore with precursors or infusions¹⁹²⁴.

This guide brings together the essentials for each: mechanism of action, what has been measured in humans, how they compare, what is known about safety and product quality, and the questions still open. For a shorter critical read of all four together, see Epitalon, MOTS-c, NAD+ and SS-31: longevity between promise and proof.

The difference that matters most to a physician is the level of evidence. Only one of the four, elamipretide, has a regulatory approval, and it is for a rare genetic mitochondrial disease, not for aging¹¹. The others rely mainly on cell and animal studies, small human trials and biochemical rather than clinical endpoints⁸²⁴²⁶.

Mitochondria-derived peptides: from humanin to MOTS-c

The mitochondrion keeps a small genome of its own, and for decades it was thought to encode only respiratory chain components and their translation machinery. In 2001 a functional screen identified humanin, a short polypeptide that, in culture, abolished neuronal death caused by several familial Alzheimer’s disease genes and by amyloid beta¹. Humanin was later described as encoded in the 16S ribosomal RNA region of mitochondrial DNA, and an in silico search found six more peptides in the same region, named SHLPs (small humanin-like peptides)³.

In that work, SHLP2 and SHLP3 reduced apoptosis and reactive oxygen species generation in vitro; in hyperinsulinemic-euglycemic clamp studies, SHLP2 infused into the brain increased glucose uptake and suppressed hepatic glucose production; and circulating SHLP2 levels, like those of humanin, fell with age³. This is the basis of the idea that some mitochondrial peptides act as metabolic signals that are lost with aging.

In 2015 the group of Pinchas Cohen and Changhan Lee described MOTS-c, a 16-amino-acid peptide encoded by a short open reading frame within the mitochondrial 12S rRNA². Its main target organ appears to be skeletal muscle: it inhibits the folate cycle and de novo purine synthesis, which activates AMPK, and in mice it prevented age-related and high-fat-diet insulin resistance as well as diet-induced obesity².

In 2018 the same group showed that, under metabolic stress, MOTS-c moves to the nucleus in an AMPK-dependent way, regulates a broad set of genes with antioxidant response elements and interacts with transcription factors such as NRF2⁴. That is the field’s central argument: the mitochondrion does not only make energy, it also sends signals that change nuclear gene expression⁴. These mechanisms are covered in detail in MOTS-c and the mitochondrial peptides: what is known and where the evidence ends.

MOTS-c in humans: exercise, genetics and a phase 1 analog

Human data on MOTS-c are mostly observational. In the 2021 Nature Communications study, exercise increased endogenous MOTS-c expression in human skeletal muscle and in the circulation; the gains in physical capacity with administered MOTS-c, including intermittent treatment started very late in life, were seen in mice⁵. Several authors in this line of research disclosed being consultants and shareholders of CohBar, the company that was developing MOTS-c analogs⁴⁵.

Genetics offers another clue. The m.1382A>C variant, specific to Northeast Asian populations, sits in the MOTS-c coding sequence and was proposed as one possible mechanism behind exceptional longevity in Japan⁶. A later analysis showed that the variant changes one amino acid in the peptide (K14Q); in a meta-analysis of three cohorts with 27,527 people, male carriers of the C allele had a higher prevalence of type 2 diabetes, and in mice K14Q-MOTS-c lost its effect on weight and glucose tolerance⁷. It is one of the few human data points that suggest a causal, if indirect, role.

The only human intervention data come from CB4211, a MOTS-c analog, not the native peptide. According to topline results the company released in 2021, phase 1a enrolled 65 healthy adults and phase 1b enrolled 20 adults with obesity and nonalcoholic fatty liver disease for four weeks. In phase 1b, ALT fell 21% versus a 4% rise on placebo, AST fell 28% versus 11%, and glucose fell 6% versus 0%; liver fat by MRI fell similarly in both groups (−5.03 versus −4.88 percentage points), and for body weight only a trend toward lower weight was reported, without figures⁸. The most common adverse event was injection site reactions⁸. This is a company press release, not a peer-reviewed paper.

Measuring MOTS-c is not trivial either. A doping control laboratory validated a liquid chromatography–mass spectrometry method with a 100 pg/mL detection limit and, when it compared the method with a commercial ELISA, found considerable differences in measured levels⁹. That calls for caution with studies that compare circulating MOTS-c concentrations between patient groups or before and after an intervention.

SS-31 (elamipretide): mechanism and the group’s only approved drug

SS-31 belongs to the Szeto-Schiller peptides, designed to concentrate selectively in the inner mitochondrial membrane¹⁰. It binds cardiolipin, that membrane’s signature phospholipid, through electrostatic and hydrophobic interactions; this keeps cardiolipin from turning cytochrome c into a peroxidase, preserves its electron-carrier function, protects cristae structure and supports oxidative phosphorylation¹⁰. Barth syndrome, caused by TAZ gene defects that disrupt cardiolipin, is the most direct clinical model of that mechanism¹².

Its clinical program is the largest in the group and also the most instructive, because it includes several negative results:

• Barth syndrome (TAZPOWER): in the placebo-controlled crossover phase, with 12 patients, neither primary endpoint was met; in the open-label extension the 6-minute walk improved 95.9 m at 36 weeks¹² and a cumulative 96.1 m at 168 weeks in the 8 patients who reached that visit, with injection site reactions as the most common adverse event¹³.
• Primary mitochondrial myopathy (MMPOWER-3): 218 participants randomized for 24 weeks; the primary endpoints were not met (6-minute walk difference of −3.2 m versus placebo; p = 0.69), although the drug was well tolerated¹⁴.
• Heart failure with reduced ejection fraction: a single infusion reduced left ventricular volumes in the highest-dose cohort at the end of the infusion¹⁵, but in PROGRESS-HF, with 71 patients over 28 days, it did not improve end-systolic volume at 4 weeks¹⁶.
• ST-elevation myocardial infarction (EMBRACE STEMI): the intravenous form was safe and well tolerated but did not reduce infarct size measured by CK-MB¹⁷.
• Geographic atrophy from macular degeneration (ReCLAIM-2): 176 patients over 48 weeks; it missed its primary endpoints, although it was associated with less ellipsoid zone degradation, a surrogate marker of photoreceptor damage¹⁸.

With that record, on September 19, 2025 the FDA granted accelerated approval to elamipretide (Forzinity) for Barth syndrome in patients weighing at least 30 kg, based on improved knee extensor strength, and required a randomized, double-blind, placebo-controlled confirmatory trial¹¹. The most common adverse reactions were mild-to-moderate injection site reactions, and the agency noted that serious reactions have also been reported¹¹. It is an approval for a rare genetic disorder of cardiolipin metabolism, not for aging or physical performance. More detail in SS-31 (elamipretide): the mitochondrial peptide that reached the FDA.

NAD+: what the precursor and infusion trials measure

NAD+ is a central coenzyme of energy metabolism and the substrate of enzymes such as the sirtuins. In experimental models, expression and activity of the NADase CD38 rise with age and are required for the age-related NAD decline and mitochondrial dysfunction; CD38 is also the main enzyme that degrades the precursor nicotinamide mononucleotide (NMN) in vivo¹⁹. That is why the best-studied strategy is not giving NAD+ directly but giving precursors: nicotinamide riboside (NR) and NMN.

Human trials show that the precursors raise NAD+ and are well tolerated. In a 2 × 6-week randomized crossover trial in healthy middle-aged and older adults, NR was well tolerated and stimulated NAD+ metabolism; the authors proposed studying blood pressure and arterial stiffness in future trials²⁰. In postmenopausal women with prediabetes and overweight or obesity, 10 weeks of NMN increased muscle insulin sensitivity measured by hyperinsulinemic-euglycemic clamp²¹. In NADPARK, with 30 patients with newly diagnosed Parkinson’s disease, NR for 30 days was well tolerated and raised brain NAD significantly but variably²².

The problem is the step from biochemistry to the clinic. A 2026 systematic review pooled 113 studies, 33 of them human intervention studies (28 randomized): oral precursors consistently moved NAD metabolites, but functional, metabolic and vascular effects were heterogeneous and often null, and it found no clinical outcome trial of intravenous or intramuscular NAD+ for anti-aging or wellness²⁴. A 2025 meta-analysis concluded that current evidence does not support NMN or NR for preserving muscle mass and function in adults with a mean age over 60²⁵.

On infusions, the most cited pharmacokinetic data come from a 6-hour pilot study: for the first 2 hours plasma NAD+ and its metabolites did not change, indicating that infused NAD+ is removed from plasma rapidly and completely, and NAD+ and methylnicotinamide later appeared in urine²³. The full analysis is in NAD+ in humans: precursors, infusions and what the trials measure.

Epitalon: telomerase, melatonin and the Khavinson program

Epitalon (Ala-Glu-Asp-Gly) is a tetrapeptide synthesized from the amino acid composition of epithalamin, a bovine pineal gland extract²⁶. A 2025 review summarizes 25 years of in vitro, in vivo and in silico studies, with reported effects on melatonin synthesis, interleukin-2 mRNA levels and the activity of enzymes such as telomerase, but warns that it is unclear whether these are its only mechanisms and that physicochemical and structural studies of the peptide are scarce²⁶.

The finding that made it famous dates from 2003: in telomerase-negative human fetal fibroblasts, epitalon induced expression of the catalytic subunit, enzyme activity and telomere elongation²⁷. In female SHR mice treated for life, 54 per group, it did not change mean lifespan but increased the lifespan of the last 10% of survivors by 13.3% and maximum lifespan by 12.3%, without changing total spontaneous tumor incidence²⁸. In humans, the most cited data point is a 12-year randomized study of epithalamin, that is, the extract rather than the tetrapeptide, in older adults with coronary disease: mortality was 28% lower in the treated group²⁹.

The clinical and animal studies cited come from the same research group that developed the peptide, led by Vladimir Khavinson²⁷²⁸²⁹. Independent confirmation arrived in 2025, but in cells: a British laboratory saw concentration-dependent telomere lengthening in normal epithelial cells and fibroblasts through increased hTERT mRNA and telomerase activity³⁰. The full story is in Epitalon: Khavinson’s program, telomerase and what still needs replication.

MOTS-c vs SS-31 vs NAD+ vs epitalon: how they compare

• MOTS-c: endogenous 16-amino-acid peptide encoded in the mitochondrion; acts on the folate–purine–AMPK axis and as a nuclear signal²⁴. Human evidence: observational data from exercise and genetics, and one analog (CB4211) in phase 1 with results not published in a journal⁵⁷⁸.
• SS-31 (elamipretide): synthetic tetrapeptide that binds cardiolipin¹⁰. Human evidence: several randomized trials, most negative on their primary endpoint^14161718; accelerated approval only for Barth syndrome¹¹.
• NAD+: a coenzyme, not a peptide¹⁹. Human evidence: dozens of oral precursor trials confirming that NAD+ rises, with inconsistent clinical benefits; infusions lack outcome trials²³²⁴.
• Epitalon: synthetic tetrapeptide modeled on a pineal extract²⁶. Evidence: telomerase in vitro²⁷³⁰ and maximum lifespan in mice²⁸; human data are with the extract and from a single group²⁹.
• Humanin and SHLPs: endogenous mitochondrial peptides with cytoprotective and metabolic effects in cells and animals; no human intervention trials among the sources in this guide¹³.

Ranked by the strength of the human evidence, first place goes to elamipretide in one narrow rare-disease indication, followed by NAD+ precursors with many small, short trials; MOTS-c and epitalon come last, with interesting biology but very few controlled clinical data^1124829.

Safety, adverse effects and product quality

The best-known safety profile is elamipretide’s, because it is the only agent with phase 3 trials and a label: injection site reactions were the most common adverse event in the 168-week extension and in the approval, and in MMPOWER-3 most events were mild to moderate¹¹¹³¹⁴. Oral NAD+ precursors have generally been well tolerated over weeks to months²⁰²⁴. For the MOTS-c analog, only four weeks of exposure are known, with local reactions as the most common event⁸.

Epitalon has no modern safety trials. The most relevant theoretical question is telomerase: activating it in somatic cells is the proposed mechanism²⁷, and in the independent 2025 study epitalon also lengthened telomeres in two breast cancer cell lines, in that case through the alternative ALT mechanism³⁰. In mice treated for life it did not raise total tumor incidence²⁸, but long-term human safety data with the tetrapeptide do not exist in the literature cited here.

Product quality shapes every result. Most peptides are made by solid-phase synthesis, which can generate impurities from amino acid deletion or insertion, racemization, protecting-group adducts, oxidation, dimers and oligomers, counter-ions such as trifluoroacetate, and contamination with unrelated peptides when good manufacturing practice is lacking; degradation products such as diketopiperazine, pyroglutamate or succinimide formation add to this³¹. These impurities can lead to wrong conclusions even in laboratory studies³¹. For MOTS-c, the doping control method had to include metabolites and two oxidation products of the peptide⁹. How to check purity and identity is explained in How to read a peptide certificate of analysis, and storage in Peptide stability: from lyophilized powder to reconstituted vial.

Open questions about longevity peptides

• Does MOTS-c given to humans reproduce what was seen in mice? That needs a published randomized trial, with the native peptide or an analog⁵⁸.
• Do commercial ELISAs measure circulating MOTS-c accurately? The differences from mass spectrometry suggest not always⁹.
• Does the knee strength gain with elamipretide translate into functional benefit? That is what the confirmatory trial must show¹¹.
• Does raising NAD+ produce any lasting clinical benefit, and what does the intravenous route add over the oral one? So far there are no outcome trials of infusions²³²⁴.
• Will epitalon’s clinical results be replicated independently, and is long-term telomerase activation safe?²⁶³⁰
• Does mitochondrial genetics change the response to these peptides, as the MOTS-c K14Q variant suggests?⁷