Finding. Two chemically unrelated, brain-penetrant inhibitors of the mitochondrial permeability transition pore bind NLRX1. Binding potency tracks pore blockade. Deleting NLRX1 in human cells or mouse tissue raises the calcium threshold for opening, independently of cyclophilin D; overexpressing it lowers that threshold. The oral lead, GSK900, is brain-penetrant and active in an mPTP-sensitive neurological injury model.
Why this paper matters
The mitochondrial permeability transition pore is the inner-membrane catastrophe of calcium-overloaded mitochondria: membrane potential collapses, solutes flood out, and the cell often dies. For thirty years the field has argued over whether the pore is ATP synthase, the adenine nucleotide translocase, a lipid defect, or something else. Cyclophilin D is the one agreed sensitizer, which is why cyclosporin A became the textbook inhibitor. The problem is practical as well as ontological. Several compounds that close the pore still work when CypD is gone. Their target was unnamed.
Peltier-Heap, Frederick, Booty and colleagues stop treating that residue as a nuisance and use it as a handle. They take two optimized, brain-penetrant chemotypes that inhibit the pore and ask what protein they share. Chemoproteomics answers NLRX1: the one NOD-like receptor that lives on mitochondria. That is a different kind of claim from another candidate subunit. It says a mitochondrial innate-immune sensor is required for a bioenergetic death channel, and that you can drug it.
What they actually measured
The logic is chemical first, genetic second, physiological third.
Affinity-based chemoproteomics with both chemotypes recovers NLRX1. Across a compound series, how tightly a molecule binds NLRX1 predicts how well it blocks the pore. That structure-activity lock is the difference between a sticky off-target and a pathway.
CRISPR-Cas9 editing in human cells and tissues from Nlrx1 knockout mice then test necessity. Loss of NLRX1 raises the calcium load required to open the pore. Overexpression does the opposite. Both effects persist without CypD, which is the sentence that moves this paper out of the cyclosporin literature and into a new regulatory layer.
NLRX1 is not floating free of the usual suspects. It associates with ATP synthase and the adenine nucleotide translocase, and those associations are compound-sensitive. The inhibitors, in other words, do not merely occupy an isolated pocket. They remodel a complex the field already suspected was the pore, or next to it. Over longer timescales the same protein helps keep the mitochondrial proteome in register, which may be a separate job or the slow echo of a leakier inner membrane.
GSK900 is the translational object. It is orally bioavailable, it reaches brain, and it is active in an mPTP-sensitive neurological injury model. The abstract does not name the model or the effect size. The existence of a brain-penetrant, oral probe against a newly named regulator is still the first thing a neurodegeneration or stroke lab should know this week.
How to read the score
This scores at the top of the weekly set because it changes what the pore is, not only how to inhibit it. Dual chemotypes, binding-activity tracking, human-cell and mouse-tissue genetics, CypD-independence, physical ties to ATP synthase and ANT, and a brain-penetrant lead are a lot of independent supports for one protein.
Confidence is high for the claim as stated in the abstract: NLRX1 is necessary for normal calcium-induced opening and is the shared target of these inhibitors. Confidence is not high that NLRX1 is the pore-forming subunit, that GSK900 is a drug, or that the neurological model will survive a full methods read. Preprint, unnamed injury model, no binding-site structure. Those are real gaps. They do not make the genetic and chemical coincidence small.
What to do with it
If you model permeability transition, stop treating NLRX1 as an innate-immune side quest. Put it on the same board as CypD, ATP synthase, and ANT. When the files post, take the chemoproteomic table, the compound-sensitive interaction data, and the GSK900 exposure numbers. Do not write a clinical tweet that NLRX1 blockade treats stroke. Do write that the pore now has a CypD-independent, druggable regulator with a brain-penetrant probe, and that this preprint is how that sentence became sayable.