SS-31, also written elamipretide and appearing in older papers as MTP-131 or Bendavia, is a synthetic tetrapeptide with the sequence D-Arg–2′,6′-dimethyltyrosine–Lys–Phe–NH2 [4]. Two of its four residues are not proteinogenic: the N-terminal arginine is the D-enantiomer, and the tyrosine carries methyl groups on the aromatic ring. The residues alternate basic and aromatic, and the C-terminus is amidated, so at neutral pH the molecule is a small, water-soluble polycation. That alternating cationic–aromatic motif is the structural feature nearly all of the mechanistic literature returns to [1, 6].

It belongs to a family described in the literature as Szeto-Schiller peptides, after the two investigators whose laboratories first described them [2]. SS-31 is the member that has been studied most, and it is now routinely used as the benchmark against which newer analogs in the same class are compared [6, 11].

Identity and characterization

The identity fields carried on the Echo Peptides product page are taken from a public chemical record rather than from a supplier datasheet:

The formula accounts for nine nitrogens across four residues, which is what an arginine guanidinium, a lysine side chain and an amidated C-terminus together require. At 639.8 g/mol, a 10 mg quantity is 15.6 µmol, so a 10 mg vial taken up in 1 mL of diluent is 15.6 mM and the same vial in 5 mL is 3.13 mM. Those molarity figures are arithmetic from the molecular weight, not measurements.

Identity numbers are not a purity statement and not a batch statement. A CAS number names a substance; whether a particular vial contains that substance at a stated purity is a question only an assay on that material can answer, and only for the material assayed.

Mechanism as studied

Association with cardiolipin

Cardiolipin is a four-chain anionic phospholipid found in the inner mitochondrial membrane, where it contributes to cristae curvature and to the organization of respiratory complexes into supercomplexes [2]. The central mechanistic proposal for SS-31 is that it partitions preferentially into membranes enriched in this lipid.

The evidence is biophysical and reasonably direct. Using a polarity-sensitive fluorescent analog of SS-31, Birk and colleagues reported high-affinity binding to cardiolipin [3]. In a companion study using defined liposomes and bicelles, SS-31 interacted with vesicles containing cardiolipin but not with phosphatidylcholine-only preparations, at an approximate 1:1 stoichiometry, and NMR indicated that the aromatic residues penetrate deep into cardiolipin-containing bilayers rather than resting on the surface [4]. Later work in model and mitochondrial membranes found that the peptide partitions into the interfacial region of the bilayer with an affinity and binding density that track membrane surface charge, and that at the concentrations tested it did not destabilize lamellar bilayers, though it did produce saturable changes in lipid packing [5].

Two qualifications belong with that. First, the selectivity described is for anionic, cardiolipin-containing membranes over neutral ones — a lipid-composition preference, not a receptor. SS-31 has no identified receptor, and none of the work retrieved here proposes one. Second, the affinity relationship reported by Mitchell and colleagues is with surface charge generally [5], which is a weaker and more general statement than binding to cardiolipin specifically.

The cytochrome c switch

Cytochrome c associates with cardiolipin, and that association is described as determining whether cytochrome c behaves as an electron carrier or acquires peroxidase activity [2]. The SS-31/cardiolipin complex was reported to inhibit cytochrome c peroxidase activity by protecting its heme iron [3]. In the liposome and mitoplast work, SS-31 restored cytochrome c reduction and oxygen consumption in preparations to which cardiolipin had been added, and in freshly isolated mitochondria it increased state 3 respiration and the efficiency of ATP synthesis [4]. The authors' framing is that the peptide suppresses the peroxidase behavior while preserving the electron-carrying behavior [4].

Antioxidant scavenging, and the reframing of it

The earliest characterization was as a targeted antioxidant. Zhao and colleagues reported that peptides of this class are cell-permeable, concentrate roughly 1000-fold in the inner mitochondrial membrane, and reduced intracellular reactive oxygen species and cell death caused by tert-butylhydroperoxide in cultured neuronal N2A cells with EC50 values in the nanomolar range. In isolated mitochondria the same peptides reduced mitochondrial ROS production, inhibited permeability transition and swelling, and prevented calcium-induced cytochrome c release. Analogs lacking dimethyltyrosine did not inhibit mitochondrial ROS generation or swelling, which was read as evidence that the dimethyltyrosine residue supplies the scavenging chemistry [1].

Later biophysical work does not overturn that observation but shifts the emphasis. Mitchell and colleagues proposed that modulation of membrane surface electrostatics is a key component of the mechanism, offering non-exclusive routes by which tuning surface charge could alter the distribution of ions and basic proteins at the membrane interface, and showed as a proof of concept that SS-31 changes divalent cation distribution in the interfacial region [5]. A structure–activity study across four tetrapeptides found that analogs differed markedly in their effect on membrane surface charge, and read that as support for an electrostatic mechanism [6]. These are proposals, stated as such by their authors; the relative contribution of direct radical scavenging and of membrane electrostatics has not been resolved by the work retrieved here.

Protein neighbors

Chemical cross-linking with mass spectrometry in mitochondria identified protein interactors of SS-31 [7]. All of the identified interactors were already-known cardiolipin binders, and they fell into two groups: proteins of the oxidative phosphorylation pathway, and proteins involved in 2-oxoglutarate metabolic processes. Cross-linked residues frequently sat near regions where those proteins contact cardiolipin [7]. This is consistent with a lipid-mediated mechanism and complicates a purely lipid account of it: whatever the peptide does, it does in the immediate company of specific proteins.

Model systems it has actually been examined in

The literature falls into four tiers, and it is worth separating them because they support different kinds of statement.

  • Defined lipid systems. Liposomes and bicelles of controlled composition, examined by fluorescence spectroscopy, turbidity and NMR [4]; model bilayers and isolated mitochondrial membranes examined by biophysical and computational methods [5]; NMR and molecular dynamics structural models of the peptide in a membrane-bound state [6].
  • Isolated mitochondria and mitoplasts. Oxygen consumption, state 3 respiration and ATP synthesis efficiency [4]; permeability transition, swelling and calcium-induced cytochrome c release [1]; cross-linking proteomics [7].
  • Cultured cells. Neuronal N2A cells under chemical oxidant stress [1]; mammalian cell culture assays of permeation and mitochondrial localization, and serum-withdrawal stress models scoring mitochondrial membrane potential, ATP content and cell survival [6]; HeLa 705, U87 and bEnd.3 cells in a comparative uptake study [11].
  • Whole-animal rodent work. Aged mouse skeletal muscle, with mitochondrial energetics measured in vivo by combined optical and 31P magnetic resonance spectroscopy [8]; phosphoproteomic profiling of skeletal muscle in aged female mice [9]; muscle metabolomics following electrically stimulated contraction in young and aged mice of both sexes [10].

The rodent work is where the most-cited readouts come from. In aged mice, resting and maximal mitochondrial ATP production, oxidative phosphorylation coupling (P/O) and cellular energy state (PCr/ATP) were reported at young-animal levels one hour after a single systemic administration, alongside a more reduced glutathione redox status and lower mitochondrial H2O2 emission [8]. In the same study SS-31 had no observable effect on young muscle [8] — a detail that matters more than the positive result, because it says the readouts move only where the measured baseline was already displaced.

What the evidence does not establish

SS-31 is supplied here as a laboratory reference material. It is not for human or veterinary use, and nothing in this guide describes a dose, a schedule, a route or a use in people or animals outside a controlled study. The following gaps are real and are stated plainly rather than smoothed over.

  • No mechanism is settled. Cardiolipin binding, surface-charge modulation and direct radical scavenging are all supported by data and none has been shown to be the operative one. The authors of the electrostatics work describe their own proposals as non-exclusive [5].
  • Molecular readouts do not all move together. In aged mouse muscle, elamipretide partially reversed age-related changes in protein phosphorylation, but those phosphorylation changes did not relate to the S-glutathionylation changes reported previously in the same system [9]. A single upstream mechanism would be easier to argue if they tracked.
  • SS-31 is not an optimum within its own class. In a four-peptide structure–activity study, side-chain composition and sequence register influenced activity, all analogs except SS-31 formed compact reverse-turn conformations when membrane-bound, and the tryptophan-containing analog had the strongest effect on most membrane properties [6]. In a separate comparison, an SS-31-derived peptide showed roughly 35% greater uptake and about twofold greater suppression of reactive oxygen species formation in bEnd.3 cells than SS-31 itself [11].
  • Effects reported in intact tissue are largely restorative rather than additive. The clearest demonstration of this is the absence of effect in young muscle [8]. There is no evidence retrieved here that the compound alters energetics in an unstressed preparation.
  • There is no receptor. Nothing in this literature identifies a binding protein that functions as a receptor. The interactors found by cross-linking are cardiolipin binders, which is a different claim [7].

Handling and stability

The searches run for this guide returned no peer-reviewed characterization of the lyophilized or reconstituted stability of SS-31 — no shelf-life study, no freeze-thaw data, no solution half-life at a stated temperature and pH. That is a genuine absence, not an oversight, and it means the handling guidance below is general peptide practice plus the supplier's own storage statement, not a literature-derived specification.

What the structure does support is narrower. The sequence contains a D-amino acid and a non-proteinogenic dimethylated tyrosine [4], both of which are common design choices in peptides intended to resist proteolysis; the retrieved literature states the composition without measuring the resulting stability, so the inference stops there. The molecule is water-soluble and polybasic [5], which is consistent with reconstitution in aqueous diluent without a co-solvent.

The product page states that the material is supplied as a lyophilized powder and refrigerated at 2–8 °C. Aromatic and guanidinium groups are the usual candidates for photo- and oxidative degradation in peptides of this type, so protection from light and minimal headspace exposure are reasonable defaults — reasonable, not validated. A laboratory that needs a stability figure for its own conditions will have to generate it.

Open questions

Three questions look tractable and unanswered. First, what is the quantitative relationship between cardiolipin content and peptide partitioning in a native inner-membrane preparation, as distinct from a compositionally defined liposome? The affinity relationship reported so far is with surface charge [5], and cardiolipin is not the only source of surface charge in a real membrane. Second, does the peptide's effect on membrane electrostatics account for the respiratory readouts, or are the two independent consequences of the same binding event? A study that dissociated them would be informative in either direction. Third, why do the phosphorylation and redox modification changes in aged muscle fail to correspond [9] — is that a timing artifact, a tissue-heterogeneity artifact, or evidence of more than one mechanism operating on different timescales?

Until those are addressed, the defensible summary is narrow: SS-31 is a cationic aromatic tetrapeptide that partitions into anionic phospholipid membranes, alters their surface electrostatics and lipid packing, associates with cardiolipin and with cardiolipin-binding respiratory proteins, and changes measured respiratory and redox parameters in isolated mitochondria, cultured cells under chemical or nutrient stress, and aged rodent muscle. Everything past that sentence is either hedged in the source or not yet established.