Three of Long COVID's most debilitating symptom clusters — cognitive dysfunction, crushing fatigue, and microclotting — have been studied as separate problems by separate research teams in separate fields. But a single molecule sits at the intersection of all three. The aryl hydrocarbon receptor, or AHR, is a transcription factor that the virus hijacks to create a self-reinforcing metabolic trap. Break it, and you may address brain fog, energy collapse, and thrombosis simultaneously.
This isn't speculation. Each connection has been demonstrated in laboratory studies. One research group has already halted the key enzyme ex vivo using an AHR antagonist. But no clinical trial exists. And the most advanced AHR antagonist in pharmaceutical development was recently discontinued — for cancer, not Long COVID. The science has converged. The medicine hasn't followed.
The Receptor
AHR was originally discovered as the receptor for dioxin. For decades it was studied primarily in toxicology — the protein that mediates the harmful effects of environmental pollutants. But over the past fifteen years, it has been recognized as a master regulator of immune function, cellular differentiation, and metabolism. It responds to a broad range of ligands, including tryptophan metabolites produced by normal human biochemistry.
When AHR binds a ligand, it translocates to the nucleus and activates transcription of target genes. Those targets include enzymes that metabolize tryptophan, inflammatory mediators, coagulation factors, and proteins that regulate mitochondrial function. In health, AHR activation is tightly controlled. In Long COVID, the control is lost.
SARS-CoV-2 activates AHR through multiple pathways — both through the inflammatory cytokine-driven production of kynurenine metabolites (which are AHR ligands) and through at least one IDO1-independent pathway that bypasses kynurenine accumulation entirely. The result is sustained, pathological AHR activation that persists long after acute infection resolves.
Thread One: The Brain Fog Engine
The connection between AHR and cognitive dysfunction runs through tryptophan metabolism — specifically, through an enzyme called IDO2.
Tryptophan is an essential amino acid. Under normal conditions, most of it is used to produce serotonin (which regulates mood, sleep, and gut function) and melatonin. A smaller fraction is metabolized through the kynurenine pathway. In Long COVID, this balance is catastrophically disrupted.
A landmark 2023 study from Amsterdam UMC, published in eBioMedicine (The Lancet), revealed something unexpected. While acute COVID-19 activates IDO1 (the canonical tryptophan-degrading enzyme), chronic Long COVID is driven by IDO2 — a related enzyme that is rarely expressed under normal conditions. IDO2 was found active in PBMCs and brain tissue of PASC patients months after infection. All PASC patients in the study had elevated xanthurenic acid, a downstream kynurenine metabolite.
Here is where AHR enters. IDO2 expression is driven by AHR activation. The kynurenine metabolites produced by IDO2 are themselves AHR ligands. This creates a vicious circle:
Virus → AHR activation → IDO2 expression → kynurenine metabolites → more AHR activation → more IDO2
The loop is self-sustaining. Once established, it no longer requires the virus to be present. The metabolic circuit feeds itself.
The consequences for the brain are severe. Tryptophan diverted through IDO2 produces quinolinic acid — a potent NMDA receptor agonist and neurotoxin. Quinolinic acid drives excitotoxicity in neurons, directly contributing to the cognitive impairment that patients describe as brain fog. Meanwhile, tryptophan diverted away from serotonin synthesis depletes the neurotransmitter essential for mood regulation, sleep architecture, and concentration.
A 2024 meta-analysis confirmed the pattern: the kynurenine-to-tryptophan ratio is significantly elevated in Long COVID patients (SMD = 0.755), with tryptophan itself decreased (SMD = -0.520). The kynurenine pathway is overactive. The serotonin pathway is starved.
The Amsterdam UMC team provided the proof of concept. They cultured PBMCs from five PASC patients with an AHR antagonist. IDO2 expression was inhibited in a dose-dependent manner, nearly eradicated at 5 μM. The loop can be broken.
Thread Two: The NAD+ Drain
The second thread connects AHR to the energy crisis — the profound fatigue and exercise intolerance that define Long COVID for many patients.
The link is PARP — poly(ADP-ribose) polymerase. PARP enzymes are activated by AHR signaling and consume NAD+ as their substrate. NAD+ is the central coenzyme in cellular energy metabolism. It is required for glycolysis, the citric acid cycle, and oxidative phosphorylation. When PARP is chronically activated by sustained AHR signaling, it drains the cellular NAD+ pool.
This explains a clinical puzzle that frustrated researchers: why nicotinamide riboside (NR) supplementation failed in Long COVID. A trial at Massachusetts General Hospital (n=58) found that NR successfully boosted NAD+ levels — but failed to improve symptoms. The supplement was refilling a bucket with an active drain. PARP, driven by AHR, was consuming the NAD+ as fast as it was being replaced.
AHR activation also suppresses melatonin through CYP1B1 induction, reduces sirtuin activity (sirtuins are NAD+-dependent enzymes critical for mitochondrial quality control), and impairs superoxide dismutase (SOD) function. The net effect is a cell that can't produce energy efficiently, can't repair its mitochondria, and can't neutralize oxidative stress.
This connects directly to the mitochondrial dysfunction documented in Long COVID. The WASF3 pathway (endoplasmic reticulum stress disrupting mitochondrial supercomplex assembly), the Complex V reversal (ATP synthase running backward), the elevated oxidative stress in lymphocytes — all of these may be downstream of, or amplified by, chronic AHR-mediated NAD+ depletion. Block the AHR-PARP axis, and you stop the drain. The bucket can fill.
Thread Three: The Clotting Switch
The third thread emerged in February 2026, published by Subramaniam and colleagues in Blood Vessels, Thrombosis & Hemostasis. It connects the AHR-kynurenine axis directly to the microclotting that has been independently documented in Long COVID patients.
Using K18-hACE2 mice infected with SARS-CoV-2, the team found that the virus upregulates IDO-1 in the liver, producing elevated kynurenine levels. These kynurenine metabolites activate AHR in endothelial cells, which in turn induces expression of tissue factor — the primary initiator of the coagulation cascade.
Tissue factor on endothelial cells converts the vessel lining from anticoagulant to procoagulant. Blood that should flow freely begins to clot. The researchers validated this in human samples: kynurenine levels in COVID-19 patient sera correlated directly with tissue factor-inducing activity on endothelial cells. The higher the kynurenine, the greater the prothrombotic drive.
Critically, both an IDO inhibitor and an AHR inhibitor separately suppressed tissue factor activity induced by COVID-19 sera. The pathway is pharmacologically targetable at multiple points.
This finding bridges a gap that had puzzled the field. Microclots in Long COVID — amyloid fibrin deposits resistant to fibrinolysis, stabilized by neutrophil extracellular traps — have been extensively documented by Pretorius, Kell, and others. But the upstream trigger for the prothrombotic state was unclear. The kynurenine-AHR-tissue factor axis provides a mechanistic explanation: the same metabolic loop driving brain fog is simultaneously switching on coagulation.
One Target, Three Symptom Domains
This is the convergence. AHR sits at the center of three pathways that have been studied independently:
- Brain fog: AHR → IDO2 → kynurenine → quinolinic acid (neurotoxic) + serotonin depletion
- Fatigue: AHR → PARP → NAD+ depletion → mitochondrial dysfunction + impaired energy metabolism
- Microclots: AHR → tissue factor expression → prothrombotic endothelium → microclotting
And connecting all three: the self-reinforcing AHR-IDO2-kynurenine loop that sustains itself without requiring ongoing viral presence.
No other single target in Long COVID research links this many symptom domains through experimentally validated mechanisms. IL-6 drives inflammation broadly. IFN-γ connects monocytes to fatigue. Complement links endothelial damage to clotting. But AHR is the only node where brain fog, energy failure, and thrombosis converge through named molecular intermediaries at every step.
The Therapeutic Landscape: A Gap Wide Open
Given this convergence, one might expect AHR antagonists to be racing into Long COVID clinical trials. They are not.
BAY2416964 (Bayer): The most advanced AHR antagonist in clinical development. A Phase 1 monotherapy trial in cancer patients (n=78) completed with modest results — 32.8% stable disease, one partial response. Its combination trial with pembrolizumab (NCT04999202) was terminated in early 2025 after enrollment was cut from 164 to 47 patients. BAY2416964's clinical program appears to be winding down — but its failure was in immuno-oncology, not in the metabolic loop it was never tested against.
IK-175: Another AHR antagonist in Phase 1 for cancer. Limited public data.
Clofazimine: An FDA-approved antibiotic used for leprosy and mycobacterial infections since the 1960s. It has well-characterized AHR antagonist activity at low micromolar concentrations. It also directly inhibits SARS-CoV-2 replication — a 2021 Nature study demonstrated broad-spectrum coronavirus inhibition in cell culture and hamster models, with reduced viral load in lungs and fecal shedding. Clofazimine is cheap, globally available, has decades of safety data, and hits both the AHR loop and the virus itself. It has never been trialed for Long COVID.
Amsterdam UMC: The group that demonstrated ex vivo IDO2 inhibition by AHR antagonism in PASC patients is actively seeking a pharmaceutical partner to run a proof-of-concept clinical trial. As of early 2026, no partner has been announced.
The situation is stark. The strongest single-target convergence point in Long COVID research has zero clinical trials. The most advanced pharmaceutical AHR antagonist was discontinued for an unrelated indication. An FDA-approved drug with the right mechanism sits unused. The academic group with the most relevant data can't find a pharma partner.
Why the Gap Exists
Three factors explain the disconnect between science and clinical action.
First, AHR research spans disciplines that don't typically talk to each other. The IDO2-kynurenine work comes from immunology. The NAD+-PARP connection comes from metabolism. The tissue factor work comes from hematology. No single field “owns” AHR in the context of Long COVID. Convergences like this are invisible when research is organized by organ system or disease mechanism rather than by molecular node.
Second, the pharmaceutical industry's interest in AHR antagonists was driven by immuno-oncology. When the cancer trials underperformed, investment dried up. The fact that AHR antagonism might work for a completely different mechanism — breaking a metabolic loop rather than enhancing anti-tumor immunity — doesn't register in a pipeline organized around therapeutic areas.
Third, Long COVID still has no regulatory pathway, no validated endpoints, and no precedent for approval. Even if a pharma company wanted to trial an AHR antagonist, designing the study is hard: which patients, which outcomes, for how long? The failed trials of other drugs (Paxlovid, BC007, ensitrelvir) have made companies cautious, even when the scientific rationale is strong.
The Serotonin Connection
One additional thread deserves mention. The gut produces approximately 95% of the body's serotonin. Tryptophan diverted through the IDO2-kynurenine pathway is tryptophan not available for serotonin synthesis. Platelet dense granules store 99% of peripheral serotonin, and platelet storage pool deficiency has been documented in post-COVID POTS patients — roughly 2.5 versus 4.3 dense granules per platelet.
Serotonin depletion in platelets may contribute to the orthostatic hypotension and tachycardia that characterize dysautonomia in Long COVID. The AHR-IDO2 loop doesn't just drain tryptophan from the brain. It drains it from the cardiovascular regulatory system. This extends the convergence to a fourth symptom domain — autonomic dysfunction — through the same upstream mechanism.
What Must Happen
The convergence is clear. AHR antagonism addresses brain fog (breaking the IDO2 loop and reducing quinolinic acid), fatigue (stopping the PARP-mediated NAD+ drain), microclotting (reducing tissue factor expression on endothelium), and potentially dysautonomia (restoring tryptophan availability for serotonin synthesis). All through one pharmacological intervention.
What's needed:
- A proof-of-concept trial of clofazimine in Long COVID, measuring kynurenine/tryptophan ratio, NAD+ levels, tissue factor activity, and symptom scores. Clofazimine is FDA-approved, cheap, and available. The barrier to repurposing is regulatory and organizational, not scientific.
- A pharmaceutical partner for Amsterdam UMC to run a proper AHR antagonist trial — ideally with a next-generation compound like DA-4505 (proposed as best-in-class) rather than the discontinued BAY2416964.
- Biomarker-guided patient selection. Not every Long COVID patient may have AHR-driven pathology. The kynurenine/tryptophan ratio, IDO2 expression in PBMCs, and tissue factor activity are all measurable. Trials should enroll patients with confirmed pathway activation, not unselected populations — the lesson every failed Long COVID trial teaches.
The evidence is not theoretical. It is experimental, published, and reproducible. What it lacks is translation — someone willing to move from “this works in a dish” to “does this work in a patient.” Until that happens, the strongest convergence point in Long COVID science remains untouched.
Key Sources
- Amsterdam UMC — eBioMedicine/The Lancet (2023): IDO2 (not IDO1) drives PASC; AHR antagonist halts IDO2 ex vivo in PASC patient PBMCs
- Subramaniam et al. — Blood Vessels, Thrombosis & Hemostasis (2026): Kynurenine-AHR-tissue factor axis; IDO/AHR inhibitors suppress prothrombotic activity of COVID-19 sera
- Quintana et al. — Science Advances (2022): AHR as proviral host factor; AHR antagonists reduce SARS-CoV-2 replication in vivo
- Mazari et al. — Inflammopharmacology (2025): AHR gene expression significantly upregulated in COVID patients; IDO-1 slightly downregulated (consistent with IDO2 being the PASC-relevant enzyme)
- Chatterjee & Maparu — Frontiers in Molecular Medicine (2025): AHR signaling in COVID-19 pathology; IDO1-independent AHR activation pathway identified
- Yuan et al. — Nature (2021): Clofazimine broadly inhibits coronaviruses including SARS-CoV-2 in vitro and in vivo
- BAY2416964 — Journal of Clinical Oncology (ASCO 2023): Phase 1 monotherapy results; combination trial (NCT04999202) subsequently terminated
- Kynurenine pathway meta-analysis — Neuroscience (2024): KYN/TRP ratio elevated (SMD 0.755), TRP decreased (SMD -0.520) in Long COVID
- MGH NR trial (2023, n=58): NAD+ boosted but symptoms unchanged — explained by active PARP drain