SS-31 and cardiolipin — the mitochondrial membrane story

Most of the conversation about mitochondrial health stops at ATP. Mitochondria make energy, the thinking goes, so mitochondrial dysfunction means less energy — fatigue, weakness, cognitive fog. That framing isn't wrong, but it misses the structural story underneath it, the one that explains why mitochondrial dysfunction is so difficult to reverse even when you address the obvious variables. The problem isn't just that the factory is running slow. It's that the factory's physical structure has warped in ways that make efficient production geometrically impossible.

The inner mitochondrial membrane is not a flat surface. Under electron microscopy it looks like a crumpled landscape, folded into deep invaginations called cristae. This folding is not incidental to function. The folding is the function. The electron transport chain — the series of protein complexes that actually captures energy from electrons and uses it to pump protons across the membrane — depends on the cristae architecture to operate efficiently. The protein complexes are physically organized within those folds in ways that allow them to hand electrons to each other like relay runners. Flatten the cristae and the relay breaks down. The same enzymes, in the same approximate locations, become dramatically less efficient because the membrane geometry that organized their cooperation is gone.

What maintains the cristae geometry is, in large part, a lipid called cardiolipin.

Cardiolipin is unusual. Most phospholipids in cell membranes have two fatty acid tails. Cardiolipin has four. That double structure gives it a distinctive conical shape, and it's that shape — when cardiolipin is present in sufficient quantity and in the right configuration — that helps stabilize the curved architecture of the cristae. Cardiolipin is found almost exclusively in the inner mitochondrial membrane. It doesn't roam through the cell. It lives in one specific location, doing one specific structural job, and when it's damaged or depleted, the membrane loses the geometry it needs to run the electron transport chain well.

Oxidative stress is what damages cardiolipin. Ironically, the mitochondria themselves are the primary source of reactive oxygen species in the cell — a byproduct of electron transport. Under normal conditions, antioxidant systems manage the damage. Under chronic stress, aging, ischemia, or inherited metabolic dysfunction, the reactive oxygen species exceed what those systems can handle. Cardiolipin's four fatty acid tails, being polyunsaturated, are particularly susceptible to oxidative modification. Once oxidized, cardiolipin no longer holds its shape. The membrane geometry softens. Cristae lose their sharp folds. The electron transport chain complexes drift out of optimal configuration. ATP output falls, and reactive oxygen species increase further — a feedback loop that conventional antioxidants address only partially, because most antioxidants distribute throughout the cell and don't concentrate at the specific location where the damage is happening.

This is the problem Hazel Szeto was trying to solve.

Szeto, a pharmacologist at Cornell and later at Weill Cornell Medical College, had spent years working on peptide therapeutics and was thinking carefully about mitochondria not as a general subject but as a targeting problem. If the damage was specific — cardiolipin oxidation in the inner mitochondrial membrane — then the intervention should be specific too. Scattering antioxidant capacity throughout the cell wasn't the same as protecting the one structure that needed protecting. The question was whether a molecule could be designed that would actually find the inner mitochondrial membrane and concentrate there.

The answer she arrived at, with her colleague Peter Schiller, was a tetrapeptide — a chain of four amino acids — that carries a net positive charge arranged in a pattern that turns out to have a specific affinity for cardiolipin. The peptide is SS-31, named for Szeto-Schiller. Its clinical name is elamipretide. It's also known by the study name MTP-131. The molecule has an alternating aromatic-cationic structure: the positive charges draw it toward mitochondrial membranes, which carry a strong negative charge; once there, the aromatic amino acid dimethyltyrosine interacts directly with cardiolipin's phospholipid head groups. SS-31 doesn't need to be pumped in by a transporter. The charge differential does the work.

The mechanism was counterintuitive to people accustomed to thinking about antioxidants as things that neutralize free radicals in solution. SS-31 does have antioxidant properties — it reduces reactive oxygen species generated at the electron transport chain complexes — but that's not the primary story. The primary story is structural. By binding cardiolipin and stabilizing its interaction with the inner membrane, SS-31 helps maintain cristae curvature, preserves the organization of the electron transport chain supercomplexes, and reduces the feedback loop in which damaged cristae architecture produces more reactive oxygen species, which damage cardiolipin further, which degrades the architecture further. Szeto's framing was always about the membrane geometry first. The antioxidant effect is partly a consequence of keeping the membrane organized enough that normal electron handling can resume.

What's worth pausing on here is the design philosophy, because it's genuinely different from how most mitochondrial interventions are approached. CoQ10, NAD+ precursors, alpha-lipoic acid — these are metabolic substrates and electron carriers, compounds that participate in mitochondrial chemistry. They support the function of mitochondria that have sufficient structural integrity to use them. SS-31 targets the structural integrity itself. It's asking a different prior question: before we ask whether the mitochondria have enough fuel and cofactors, have they maintained the architecture that makes using fuel and cofactors possible? This is the distinction between a car with bad gasoline and a car with a collapsed engine block. The substrate matters, but not if the structure isn't there.

The research that followed from this design philosophy covered a wide range of injury and disease contexts. Ischemia-reperfusion injury — the damage that happens when blood flow is restored to tissue that has been deprived of oxygen — is one of the most studied. The restoration of oxygenated blood after ischemia generates a burst of reactive oxygen species that does immediate damage to cardiolipin in cardiac and renal tissue. SS-31 administered before or at the time of reperfusion significantly reduced this damage in animal models, preserving mitochondrial membrane potential, reducing cell death, and improving functional recovery in hearts and kidneys. The effect was specific enough that it supported the targeting hypothesis: this wasn't general antioxidant activity distributed throughout the tissue, it was protection at the membrane level that translated into preserved organ function.

Cardiolipin is also relevant to aging independent of acute injury. Mitochondrial function declines with age, and part of that decline appears to be structural — a gradual degradation of cristae architecture driven by accumulating cardiolipin oxidation, declining cardiolipin synthesis, and reductions in the enzymes responsible for maintaining cardiolipin's fatty acid composition. In aging skeletal muscle, the loss of mitochondrial function contributes to sarcopenia and impaired exercise tolerance. In aging cardiac muscle, it contributes to the diastolic dysfunction that characterizes heart failure with preserved ejection fraction. In neurons, it contributes to the bioenergetic vulnerability underlying neurodegenerative disease risk. In each of these contexts, the cardiolipin-cristae connection provides a mechanistic through-line that connects aging to functional decline through the specific structural damage that SS-31 was designed to address.

What that means practically — for research programs, for clinical trial design, for anyone thinking about where this compound fits — is that SS-31's utility isn't predicted simply by "does this patient have mitochondrial dysfunction?" but by "does this patient have a condition in which cardiolipin damage is mechanistically central?" That's a more specific question. It rules out conditions where mitochondrial dysfunction is downstream of some other primary pathology. It rules in conditions where the inner membrane structure is the bottleneck.

Barth syndrome is the clearest example of the latter category. It's a genetic condition caused by mutations in the TAZ gene, which encodes an enzyme responsible for cardiolipin remodeling — the process by which cardiolipin's fatty acid composition is maintained in its functional form. In Barth syndrome, cardiolipin remodeling fails, immature forms of cardiolipin accumulate, and the resulting mitochondrial dysfunction is severe and genetically unavoidable. SS-31's conditional FDA approval as Stegazo for Barth syndrome in 2023 represents the first time a regulator has formally agreed that the cardiolipin-targeting mechanism is clinically meaningful — that protecting this specific lipid in this specific membrane location is a valid therapeutic strategy, not just an interesting hypothesis.

That approval matters beyond Barth syndrome. It validates the targeting logic that Szeto and Schiller built the molecule around. It suggests that the inner mitochondrial membrane is a druggable structure, that cardiolipin is a legitimate therapeutic target, and that a tetrapeptide small enough to navigate biological barriers can actually change what happens at the membrane level in ways that affect cell function, organ function, and ultimately clinical outcomes.

The broader question — how far that logic extends, to which conditions and which patient populations, at what doses and for how long — remains genuinely open. Heart failure trials have been mixed. Primary mitochondrial myopathy trials have been more encouraging. The ongoing work in Leber's hereditary optic neuropathy and Friedreich's ataxia is probing the boundaries of where cardiolipin damage is primary enough to respond to a cardiolipin-stabilizing intervention. These are hard questions and they won't resolve quickly.

But the membrane geometry story is not a minor footnote. It is, arguably, one of the more important structural insights in mitochondrial biology of the last thirty years — the recognition that ATP production depends not just on having the right enzymes but on having them geometrically organized within a lipid architecture that makes their cooperation possible, and that this architecture can be lost, that its loss is measurable, and that something designed specifically to address that loss can in some contexts restore it. The factory analogy breaks down eventually, but at least for the inner mitochondrial membrane, it holds long enough to matter.