A 'hypoxia in a pill' doubled survival in mice with a fatal mitochondrial disease

For a decade, Vamsi Mootha's lab at Massachusetts General Hospital has run mice through a strange ritual: sealing them in chambers where the air is thinned to 11% oxygen, the equivalent of a 5,000-meter peak. The point was not to make them sick. Mice engineered to develop Parkinson's disease held in those chambers did better, not worse: their movement disorder eased, and in some cases the brain degeneration itself began to reverse.

That observation, published this year in Nature Neuroscience, is the seed of a new therapeutic bet: for certain neurodegenerative diseases, the brain may want less oxygen, not more. The follow-up, published in the Journal of Clinical Investigation, turns the chamber into a pill.

The two-drug combination is a deliberate engineering exercise, not a wellness claim. The first drug, osivelotor, was originally developed for sickle cell disease. It shifts hemoglobin's oxygen-dissociation curve so the protein holds onto O2 more tightly, leaving less available to tissue. That is the "high-altitude" effect in pill form. The second drug, PT2399, exists to neutralize a problem the first one would otherwise create: when tissues sense less oxygen, HIF-2α drives the bone marrow to churn out extra red blood cells, a condition called polycythemia that can itself be lethal. PT2399 blocks HIF-2α, suppressing the over-correction.

In mice engineered to model advanced Leigh syndrome, a fatal mitochondrial disorder that destroys the brain stem and typically kills within months, the combination more than doubled survival. Mice that would normally die around day 62 lived well past that mark on the drug pair. The result, reported in the JCI paper, is the first oral program that targets tissue oxygen itself rather than a downstream symptom of mitochondrial failure.

The Parkinson's result sits on a different experimental track. Those mice never received the pill. They lived in 11% O2 chambers for sustained periods, and the researchers measured motor symptoms and brain pathology. Mootha's lab page frames the broader oxygen program as an attempt to find a "Goldilocks zone" for tissue oxygen, with too little being dangerous but, for a narrow set of disorders, a modest dial-down plausibly reducing the oxidative damage that drives neuron loss. That "Goldilocks" framing is a hypothesis, not a settled mechanism. The chamber data are real; the translation to a human pill for Parkinson's is not yet established.

Epidemiology offers a partial anchor. People who live at high elevation long-term have lower average rates of diabetes and cardiovascular disease, and there are anecdotes of Parkinson's symptoms easing at altitude. Mootha has cited those observations as the original seed of the chamber program.

The clinical translation is the part that does not yet exist. No Parkinson's trial with the pill has been run, and Mootha has warned, in coverage of the work, that hypoxia should not be self-administered. Tissue oxygen is a tightrope: too little kills neurons, and the body's defenses against low oxygen include the same HIF-2α polycythemia that the second drug has to suppress. A successful human version of this program would have to thread that needle in patients, not in chamber mice.

The next milestone the field is watching is whether osivelotor, already in human trials for sickle cell, can be repurposed quickly enough to enter a Parkinson's or Leigh syndrome study. The Mootha lab has now converted ten years of chamber biology into a defined two-drug combination with a concrete survival result. The translation is early, the mechanism is partly unresolved, and the safety window is narrow. The bet, that less oxygen can sometimes be more, is no longer just a chamber phenomenon.