There’s a protein in your immune cells that most biochemistry textbooks treat as well-understood. It’s called inducible nitric oxide synthase, or iNOS, and for decades its job description has been simple: when macrophages detect a threat, iNOS produces nitric oxide (NO) to help kill pathogens and signal to other immune cells. That’s the standard story. But a paper published in April 2026 in Nature Metabolism shows the textbook is missing a chapter.
It turns out iNOS has been doing something else entirely on the side. Inside mitochondria, it physically binds to another enzyme called IRG1 and shuts down that enzyme’s ability to produce itaconate, a metabolite that shapes how macrophages handle inflammation. This interaction happens through direct protein-to-protein contact, not through any chemical signal iNOS releases. The finding changes how we think about two of the more important immune regulators in the cell.
Why itaconate matters
When macrophages get activated during inflammation, one of the metabolites that shoots up fastest is itaconate, sometimes by orders of magnitude. IRG1 (also called ACOD1 or CAD) makes it by pulling cis-aconitate out of the citric acid cycle and decarboxylating it.
Itaconate doesn’t just sit there. It modifies enzymes through thiol chemistry, including GAPDH and LDHA, two enzymes that drive glycolysis. It competitively inhibits succinate dehydrogenase, which reins in oxidative stress. It touches transcription factors like ATF3 and signaling molecules like Jak1. The net effect is that itaconate nudges macrophages toward a less inflammatory, more regulatory state. It’s part of the braking system that stops immune responses from running off the rails.
So what controls how much itaconate a macrophage makes?
The clue that didn’t fit
The answer came from a puzzling observation. Macrophages missing iNOS (mice with the Nos2 gene knocked out) accumulated far more itaconate than wild-type cells after inflammatory stimulation. Not a small difference either. After eighteen hours, itaconate in iNOS-deficient macrophages was about fifteen times higher than normal.
The obvious explanation was nitric oxide itself: iNOS makes NO, NO is reactive, reactive molecules mess with enzymes. Reasonable assumption. But when researchers tested it, the data said otherwise.
Testing the NO hypothesis
They started in HEK-293T cells, which don’t naturally express either protein. Co-transfecting iNOS and IRG1 cDNAs gave a clean result: itaconate production dropped to nearly zero, while NO production kept going. That alone suggested something besides NO was at work.
Next they tried adding nitric oxide donors, chemicals like NOC18 and SIN-1 that release NO directly. No effect on itaconate. Neither peroxynitrite donors, hydrogen peroxide, S-nitrosoglutathione, nor any other reactive nitrogen or oxygen species they tested made a dent. If NO were the suppressor, something should have worked.
They then purified recombinant IRG1 enzyme and ran in vitro activity assays, directly exposing it to NO donors, hydrogen peroxide, and other reactive molecules. Nothing inhibited the enzyme. IRG1 kept making itaconate regardless.
The cysteine angle didn’t pan out either. Since itaconate itself modifies proteins through cysteine residues, the team generated six IRG1 cysteine-to-alanine mutants. Every mutant produced itaconate at wild-type levels, and every one was still suppressed by iNOS.
So iNOS was suppressing itaconate, but not through nitric oxide, not through reactive species, and not through cysteine chemistry. It had to be doing something more direct.
A protein-to-protein handshake
Co-immunoprecipitation paired with mass spectrometry gave the answer. In wild-type macrophages, only seven proteins showed significant association with IRG1, and the most abundant by far was iNOS itself, enriched over elevenfold compared to the next protein on the list.
Without iNOS, the picture flipped. IRG1 associated with forty-three different proteins: glycolytic enzymes like pyruvate kinase and aldolase A, oxidative stress regulators like superoxide dismutase, even signaling molecules like interleukin-1α. It looked like iNOS had been acting as a gatekeeper, occupying IRG1 and blocking it from engaging with a much larger network of interaction partners.
Immunofluorescence imaging confirmed that both proteins co-localize within mitochondria of activated macrophages. This wasn’t a random cytoplasmic encounter. It was happening in the same compartment where IRG1 actually works.
Building a structural model
The team used AlphaFold-Multimer to predict how iNOS and IRG1 might physically interact. The model suggested a heterotetramer: two IRG1 molecules paired with two iNOS molecules, oriented roughly perpendicular to each other along a defined interaction groove.
Three-hundred-nanosecond molecular dynamics simulations showed the interface holding steady. MM/GBSA free energy calculations landed around -152 kcal/mol, a strong binding interaction by computational standards. The human versions of both proteins produced similar predictions and binding energies, which suggests the interaction has been conserved through evolution.
Surface plasmon resonance provided the physical measurement. Purified human iNOS bound human IRG1 with a dissociation constant of 189 nM; the mouse proteins came in at 174 nM. Those are solid nanomolar affinities. When they tested endothelial NOS (eNOS), no binding at all. This is an iNOS-specific interaction, not a general NOS family thing.
The BH4 requirement
Tetrahydrobiopterin (BH4) is a cofactor iNOS needs for its nitric oxide synthase activity. The researchers mutated a tryptophan residue (W457) essential for BH4 binding. The W457A mutant, which effectively eliminates BH4 binding, could no longer suppress itaconate. Neither could the W457F mutant, which partially preserves BH4 binding. Both mutants were expressed at normal protein levels.
What this means is that BH4 binding does more than support catalysis. It stabilizes iNOS in a conformation that lets it grab onto and inhibit IRG1. Strip away BH4, and iNOS shifts into a different shape. It still sits in the mitochondria, but it can’t do this particular job.
Blue native gel electrophoresis added another layer. Aminoguanidine, a pharmacological iNOS inhibitor, locked iNOS in its dimeric form. In that frozen state, iNOS kept suppressing itaconate even though NO production was completely blocked. The dimer conformation itself, stabilized by BH4, is what matters, not the catalytic activity.
What this changes
This finding moves iNOS from the category of “simple NO-producing enzyme” to something closer to a metabolic signaling hub. When a macrophage encounters an inflammatory stimulus, iNOS gets induced and imported into mitochondria, where it runs two parallel operations: producing NO and physically binding IRG1 to suppress itaconate production. The mutagenesis experiments show these two outputs can be decoupled from each other.
The researchers propose a timing argument. Early in an inflammatory response, iNOS suppression of itaconate may keep macrophages in a pro-inflammatory, pathogen-fighting state. Itaconate is broadly anti-inflammatory, so holding it back while you’re trying to clear an infection has a certain logic to it. As the response matures and the threat clears, dialing down this axis could help shift things toward resolution.
On the therapeutic side, the iNOS-IRG1 interface is a genuinely new target. Current anti-inflammatory strategies built around itaconate rely on cell-permeable derivatives like 4-octyl itaconate or dimethyl itaconate, which artificially boost itaconate signaling. This work opens an alternative: design peptides or small molecules that specifically disrupt the iNOS-IRG1 interaction, letting IRG1 produce more endogenous itaconate. Since eNOS doesn’t bind IRG1, there’s at least a plausible path to iNOS-specific effects without touching the cardiovascular functions of other NOS family members.
The human question
Translating mouse immunology to humans is always a headache. Human monocyte-derived macrophages often don’t express NOS2 under standard culture conditions because the gene gets epigenetically silenced. That’s caused real debate about whether mouse iNOS findings apply to human disease at all.
The team tackled this with a human iPSC-derived bone marrow organoid system. These organoids produce more mature monocytes that do express both NOS2 and ACOD1 when stimulated. The structural predictions, binding affinities, and functional interactions all held up in the human proteins, which makes the case that this mechanism isn’t just a mouse artifact.
Open questions
A few things still need sorting out. The exact contact residues on each protein, how binding dynamics shift in different cellular states, and whether generating an iNOS mutant that produces NO but can’t suppress IRG1 is even possible. That last one would be a powerful experimental tool for teasing apart the relative contribution of each function.
There’s a bigger question too. The co-immunoprecipitation data showed IRG1’s interaction network expanding dramatically when iNOS is absent. It’s possible iNOS organizes more than just one enzyme in the mitochondria during inflammation, maybe shaping a larger chunk of the metabolic landscape than we currently appreciate.
What’s clear right now is that a protein the field thought it understood has been running a second operation, hidden inside mitochondria. Well-studied systems still have surprises, and the tools of structural biology and computational modeling are finally sharp enough to catch them.
Frequently Asked Questions
What is iNOS?
Inducible nitric oxide synthase (iNOS) is an enzyme expressed by immune cells, particularly macrophages, in response to inflammatory signals. Its best-known function is producing nitric oxide, which helps the immune system fight infections. The research discussed here reveals a second function that operates independently of NO production.
What is itaconate and why does it matter?
Itaconate is a metabolite produced by the enzyme IRG1 during inflammation. It rises sharply in activated macrophages and acts as an anti-inflammatory signal by modifying enzymes involved in glycolysis, inhibiting succinate dehydrogenase, and influencing several transcription factors. In broad terms, itaconate pushes macrophages toward a regulatory, inflammation-dampening state.
How does iNOS suppress itaconate production?
iNOS physically binds to IRG1 inside mitochondria through direct protein-to-protein interaction. This binding blocks IRG1’s enzymatic activity, reducing itaconate output. The suppression does not require iNOS to produce nitric oxide. Instead, it depends on BH4 (tetrahydrobiopterin) binding to iNOS, which stabilizes the enzyme in a dimeric conformation capable of grabbing onto IRG1.
Does nitric oxide itself inhibit itaconate?
No. Multiple experiments showed that nitric oxide donors, peroxynitrite, hydrogen peroxide, and other reactive species do not inhibit purified IRG1 enzyme or reduce itaconate levels in cells. The suppression mechanism is purely structural: iNOS blocks IRG1 by binding to it, not by releasing a chemical signal.
Is this mechanism the same in humans and mice?
The structural predictions, binding affinities, and functional interactions have been confirmed in both mouse and human proteins. Human iNOS binds human IRG1 with a dissociation constant of 189 nM (mouse: 174 nM). The researchers also validated the mechanism in human iPSC-derived bone marrow organoids, which express both NOS2 and ACOD1 upon stimulation.
Does eNOS or nNOS also bind IRG1?
No. Surface plasmon resonance testing showed that endothelial NOS (eNOS) does not bind IRG1. This interaction appears to be specific to iNOS, which matters for drug development because it means you could potentially target the iNOS-IRG1 interface without interfering with the cardiovascular functions of other NOS isoforms.
What are the therapeutic implications?
The iNOS-IRG1 interface represents a new drug target for inflammatory diseases. Current approaches use cell-permeable itaconate derivatives (like 4-octyl itaconate) to artificially boost itaconate signaling. An alternative strategy would be to design peptides or small molecules that disrupt the iNOS-IRG1 interaction, allowing the body to produce more endogenous itaconate. This could offer a more specific way to modulate macrophage inflammation.
What is BH4’s role in all of this?
BH4 (tetrahydrobiopterin) is a cofactor that iNOS binds as part of its normal nitric oxide synthase function. The research shows BH4 binding is also required for iNOS to suppress itaconate. When BH4 binding is disrupted through mutation, iNOS can no longer inhibit IRG1, even though the protein is still expressed. BH4 stabilizes iNOS in the correct dimeric shape for the IRG1 interaction.
What tools were used to study the iNOS-IRG1 interaction?
The research combined co-immunoprecipitation and mass spectrometry (to identify binding partners), AlphaFold-Multimer structure prediction, molecular dynamics simulations, MM/GBSA free energy calculations, surface plasmon resonance (to measure binding affinity), blue native gel electrophoresis (to assess conformational states), immunofluorescence microscopy, and both mouse and human cell models including iPSC-derived organoids.