SS-31 (Elamipretide): A Research Overview
Almost every peptide in a research catalogue works by finding a receptor. SS-31 does not appear to work that way at all. It is a four-residue molecule that crosses cell membranes without a transporter, concentrates several thousand-fold inside mitochondria, and then binds not to a protein but to a lipid — cardiolipin, the signature phospholipid of the inner mitochondrial membrane. Whatever else it does downstream follows from that single physical association.
That mechanism has made SS-31 one of the more heavily studied compounds in mitochondrial bioenergetics over the past fifteen years, and it is also why the literature on it is unusually structural: a good deal of the published work concerns membrane biophysics rather than physiology. This article summarises what that literature reports about SS-31 — its chemistry, the proposed mechanism, the principal experimental findings, and the handling and analytical considerations relevant to its use as a laboratory reference material. Everything described below refers to published work in isolated mitochondria, cell culture, rodent models or registered clinical trials. Nothing here describes use outside those settings.
Chemistry and structure
SS-31 — also catalogued as elamipretide, MTP-131 and Bendavia — is a tetrapeptide with the sequence D-Arg-Dmt-Lys-Phe-NH2, where Dmt denotes 2',6'-dimethyltyrosine. Its molecular formula is C32H49N9O5 and its molecular weight is approximately 639.8 Da. The free base is catalogued under CAS 736992-21-5; the acetate salt, which is the form most commonly supplied, carries CAS 1334953-95-5.
Three features of that short sequence do most of the work. The first is the alternating pattern of cationic and aromatic side chains — arginine, then dimethyltyrosine, then lysine, then phenylalanine. This alternation is the defining motif of the Szeto–Schiller peptide family, named for Hazel Szeto and Peter Schiller, whose laboratories characterised the series. The second is the D-configuration of the N-terminal arginine and the C-terminal amide, both of which substantially reduce susceptibility to exopeptidase cleavage relative to an all-L, free-acid peptide. The third is net charge: with two basic residues and no acidic ones, the molecule carries a net charge of approximately +3 at physiological pH, which is what drives its accumulation in the negatively charged mitochondrial matrix compartment.
It is worth noting that this accumulation is not the classic lipophilic-cation mechanism used by compounds such as MitoQ, which are drawn across the inner membrane by the electrochemical potential and are correspondingly dependent on it. SS-31 uptake has been reported to be largely independent of membrane potential — a distinction that matters experimentally, because it means the peptide can in principle reach depolarised or damaged mitochondria that potential-driven agents cannot.
The cardiolipin hypothesis
Cardiolipin is a dimeric phospholipid with four acyl chains and two negatively charged phosphate groups, found almost exclusively in the inner mitochondrial membrane, where it constitutes roughly a fifth of the phospholipid content. It is structurally required for the tight curvature of cristae membranes, and it acts as a scaffold for the respiratory chain complexes and for cytochrome c.
Birk, Szeto and colleagues, writing in the Journal of the American Society of Nephrology (2013), reported that SS-31 binds cardiolipin with high affinity and that the resulting complex inhibits the peroxidase activity of cytochrome c. That reaction matters because cytochrome c, when bound to cardiolipin, gains peroxidase function and catalyses the peroxidation of cardiolipin itself — an early, self-amplifying step in the mitochondrial damage cascade. The authors proposed that SS-31 interrupts this by shielding the heme iron of cytochrome c within the complex. In a rat model of renal ischaemia, they reported that pretreatment preserved cristae architecture, prevented mitochondrial swelling, and was associated with faster ATP recovery on reperfusion.
Later biophysical work refined the picture. Mitchell and colleagues (Journal of Biological Chemistry, 2020) examined SS-31 interaction with model lipid bilayers and reported that the peptide does not simply dock onto cardiolipin as a discrete ligand–receptor pair. Rather, it partitions into the membrane interface and modulates surface electrostatics — reducing the negative surface potential of anionic bilayers. Their interpretation is that the functional consequence of SS-31 binding is a change in the electrostatic environment in which membrane-associated proteins operate, rather than an allosteric effect on any single target. A follow-up structure–activity study in eLife (2022) examined a series of related tetrapeptides and reported that membrane affinity alone does not predict biological activity, which argues against the simplest version of the electrostatic model.
Mapping the protein interactions
Chavez and colleagues took a different approach in PNAS (2020), using chemical cross-linking coupled to mass spectrometry to identify which mitochondrial proteins sit close enough to a biotinylated SS-31 analogue to be covalently captured. The result is one of the more informative datasets on the compound.
The captured proteins fell into two functional groups: components of oxidative phosphorylation and ATP handling, and enzymes of 2-oxoglutarate metabolism. Specific cross-links were reported between the peptide and lysine residues on both the α- and β-subunits of ATP synthase, as well as to adenine nucleotide translocase and mitochondrial creatine kinase. Critically, every protein identified was already a known cardiolipin binder — which is consistent with the peptide acting on the lipid environment shared by those proteins rather than on the proteins directly. The study did not demonstrate a catalytic interaction with any of them.
Ageing and functional models
The best-known rodent work concerns aged skeletal muscle. Siegel, Marcinek, Rabinovitch and colleagues (Aging Cell, 2013) administered a single intraperitoneal dose of 3 mg/kg to 27-month-old mice and reported that age-related declines in resting and maximal mitochondrial ATP production, in the P/O ratio, and in cellular energy state were reversed within one hour. The same dose produced no measurable change in 5-month-old animals — an internal control that argues the effect depends on a pre-existing deficit rather than being a general stimulant action. Treated aged muscle showed a more reduced glutathione redox status, lower mitochondrial hydrogen peroxide emission, greater fatigue resistance in situ, and, after eight days of dosing, increased whole-animal endurance capacity.
Comparable work in cardiac tissue has been published in eLife (2020), reporting that late-life administration in old mice was associated with improvements in measures of diastolic function. Separate rodent studies have examined the compound in models of diabetic nephropathy (American Journal of Physiology — Renal Physiology, 2015) and in lipopolysaccharide-induced mitochondrial and synaptic dysfunction in mice (2019). The pattern across these papers is consistent: effects are reported in tissue that is already energetically compromised, and are minimal or absent in healthy young controls.
Clinical development
Elamipretide has been carried into registered human trials by Stealth BioTherapeutics across several indications, including primary mitochondrial myopathy (the MMPOWER programme) and Barth syndrome, a rare X-linked disorder caused by mutations in the tafazzin gene that impair cardiolipin remodelling. The Barth syndrome programme is the mechanistically cleanest test of the cardiolipin hypothesis available, since the disease is defined by a cardiolipin defect. The TAZPOWER trial was a randomised, double-blind, placebo-controlled crossover study in twelve genetically confirmed patients, followed by a 168-week open-label extension. The randomised phase did not meet its primary endpoints; the open-label extension reported changes in functional and cardiac measures, and a reduction in the monolysocardiolipin-to-cardiolipin ratio that is the biochemical hallmark of the condition. The compound received FDA accelerated approval for Barth syndrome in September 2025.
That regulatory history is relevant context for anyone reading the SS-31 literature, but it describes a specific pharmaceutical product studied under clinical protocols in a defined patient population. It has no bearing on research-grade material, which is supplied for laboratory use only.
Laboratory handling
SS-31 is supplied as a lyophilised powder. Several handling considerations follow from its composition rather than from generic peptide practice.
- The dimethyltyrosine residue is the chemical weak point. Phenolic side chains are susceptible to oxidation, and 2',6'-dimethyltyrosine is not a standard proteinogenic residue, so its degradation products are less familiar than those of ordinary tyrosine. Minimising exposure to light, oxygen and elevated temperature is proportionately more important.
- Adsorption is a real source of loss. A net charge of +3 on a molecule of only 640 Da gives a high charge-to-mass ratio and promotes binding to glass and to some plastics. Dilute working solutions can lose a meaningful fraction of their nominal content to container surfaces.
- Aliquot after reconstitution. Repeated freeze–thaw cycling degrades peptides generally; dividing reconstituted material avoids cycling a single vial.
- Salt form affects mass calculations. Material supplied as the acetate salt contains counter-ion and residual water that are not peptide. Where the certificate of analysis reports net peptide content separately from gross weight, that figure — not the vial label — is the one to use in concentration calculations.
The general principles are covered in more detail in our guide to peptide storage and handling.
Analytical verification
Reversed-phase HPLC establishes chromatographic purity and mass spectrometry confirms identity against the expected monoisotopic mass. For SS-31 there is one specific point worth attention: because the peptide is short and highly basic, it elutes early on a standard C18 gradient and can co-elute with small polar impurities and with counter-ion. A purity percentage from a poorly resolved early-eluting peak carries less information than the same number from a well-retained one, so the chromatogram itself is worth examining rather than only the summary figure. On the mass spectrum, oxidation of the dimethyltyrosine residue would appear as a +16 Da satellite. Third-party analytical documentation for our catalogue is published on the Lab Results page.
What remains unresolved
The central open question is how binding a lipid produces the reported functional outcomes. The cardiolipin interaction is well established; the causal chain from that interaction to changed ATP synthesis is inferred rather than demonstrated. The 2022 structure–activity work, which found that closely related analogues with similar membrane affinity differ in activity, indicates that the model is incomplete.
Beyond mechanism, three practical gaps persist. Reported effect sizes in rodent work depend heavily on the degree of pre-existing mitochondrial dysfunction, which makes cross-study comparison difficult. Cardiolipin composition varies substantially between tissues and species, so extrapolation from one model system to another is less safe than the shared mechanism might suggest. And the relationship between the acute biophysical effect — measurable within an hour in the Siegel work — and any longer-term structural change in mitochondrial membranes is not established.
For laboratories working with SS-31 as a reference compound, that combination is what makes it useful: a chemically simple, well-characterised molecule with an unusually specific and reproducible molecular target, sitting on top of a mechanistic account that is still visibly unfinished. Researchers interested in the wider class of mitochondria-associated peptides may also find our overview of MOTS-c a useful comparison, since the two compounds approach mitochondrial function from opposite directions — one structural, one transcriptional.
Research use only. Not a medicine, supplement or therapeutic good. Not for human or veterinary consumption.