7 min read

The Broken Repair Shop: How COVID Destroys the Organelle Your Body Needs to Heal

The Broken Repair Shop: How COVID Destroys the Organelle Your Body Needs to Heal

Everyone knows about mitochondria — the powerhouses of the cell. But there's another organelle quietly running your body's repair infrastructure, and COVID is destroying it. It's called the peroxisome. And its failure may explain why Long COVID patients can't heal.

In 2022, a team at Columbia University found something strange in the blood of patients with ME/CFS — depleted plasmalogens, elevated dicarboxylic acids, reduced carnitines. Three metabolic signatures, all pointing to the same place: damaged peroxisomes. At the time, nobody connected it to COVID.

Then, between 2025 and 2026, three independent research teams studying completely different organ systems — lungs, gut, and brain — found the exact same problem. COVID doesn't just infect your tissues. It systematically dismantles the machinery your body needs to repair them.

The Organelle Nobody Talks About

Peroxisomes are membrane-bound organelles present in virtually every human cell. They metabolize very-long-chain fatty acids, synthesize plasmalogens (critical membrane lipids), and neutralize reactive oxygen species. In immune cells called macrophages, they do something even more important: they feed and regulate mitochondria, control inflammation, and direct tissue regeneration.

Think of them as the supply depot for your cellular repair crews. Mitochondria get the headlines, but peroxisomes keep them fueled. Without functional peroxisomes, mitochondria starve, inflammation runs unchecked, and damaged tissue can't rebuild.

Healthy Cell vs COVID-Damaged Cell HEALTHY Peroxisomes Active Fat metabolism + ROS scavenging Mitochondria Fueled Energy for repair programs Inflammasome Controlled IL-1β kept in check Tissue Heals Normal regeneration COVID-DAMAGED IFN-γ Storm Blocks new + destroys existing Peroxisomes Destroyed Pexophagy + blocked biogenesis Inflammasome Unleashed IL-1β + IL-18 flood tissue Fibrosis + Failed Repair Chronic inflammation loop Source: Wei et al., Science (2025); Wang et al., Developmental Cell (2026) IFN = interferon; IL-1β/IL-18 = inflammasome cytokines; ROS = reactive oxygen species

The ME/CFS Discovery That Predicted Everything

In July 2022, W. Ian Lipkin's group at Columbia published the largest metabolomics study of ME/CFS at the time — 106 patients, 91 controls, 888 metabolites measured. The signature was stark: significantly decreased plasmalogens (p < 0.001), depleted phospholipid ethers, elevated dicarboxylic acids. Machine learning could distinguish ME/CFS from healthy controls with an AUC of 0.873. All three abnormalities pointed to peroxisomal dysfunction.

The finding sat in the literature, largely unconnected to the pandemic raging outside. ME/CFS researchers understood its importance. The broader scientific community did not. Then came three papers that changed everything.

Three Organs, One Broken Machine

Four lines of evidence:

Tissue Study Key Finding
Blood Lipkin 2022 — ME/CFS Depleted plasmalogens, AUC 0.873
Lungs Wei et al. 2025 — Science IFN → pexophagy → IL-1β → fibrosis
Brain Hobman 2025 — Front Mol Neurosci Reduced PEX3, PMP70, PEX14
Gut Wang et al. 2026 — Dev Cell VLCFA disruption → failed repair

Lungs: The Science Paper

In March 2025, Xiaoqin Wei, Jie Sun, and 28 colleagues at the University of Virginia published a landmark study in Science. They found that severe COVID triggers excessive interferon signaling in alveolar macrophages — the immune cells responsible for lung repair. This interferon surge does two things simultaneously: it blocks the construction of new peroxisomes and triggers the autophagy of existing ones (a process called pexophagy).

Without peroxisomes, these macrophages lose control of their inflammasome — the molecular platform that processes inflammatory cytokines. IL-1β floods the tissue unchecked. Instead of healthy lung cells regenerating, damaged progenitor cells accumulate — KRT8-high dysplastic epithelial progenitors that drive fibrotic remodeling. The lung can't heal because its repair infrastructure has been demolished.

Critically, human lung tissue from Long COVID patients showed the same chronic peroxisome impairment seen in the mouse models. This wasn't a rodent artifact. It was happening in people.

Gut: The Developmental Cell Paper

In January 2026, Wang and colleagues published a complementary finding in Developmental Cell. Intestinal SARS-CoV-2 reservoirs disrupt very-long-chain fatty acid (VLCFA) metabolism, suppressing PPAR signaling and reducing peroxisome abundance in the gut. The result: impaired intestinal stem cell differentiation, failed epithelial regeneration, prolonged diarrhea, inflammation, and microbiome dysbiosis.

Different organ. Different research team. Same organelle. Same outcome: the body can't repair itself because the repair shop is closed.

Brain: The Frontiers Paper

Also in 2025, Hobman and Power published in Frontiers in Molecular Neuroscience that COVID-19 brain samples showed reduced levels of PEX3, PMP70, and PEX14 — key peroxisomal proteins — alongside elevated cytokines (IL-8, CXCL10). Peroxisome dysfunction in microglia leads to oxidative stress that damages neurons and oligodendrocytes. This could explain brain fog through the same mechanism playing out in the lungs and gut.

The Pattern: COVID didn't create peroxisome dysfunction — it exposed it. The Lipkin ME/CFS finding in 2022 predates the COVID papers. This is a post-viral syndrome mechanism, not a COVID-specific one. It may explain why ME/CFS and Long COVID share so many symptoms.

The IFN-γ Thread

If you read my previous post on IFN-γ, this connection will hit hard. Krishna et al. found that over 60% of Long COVID patients have persistently elevated interferon-gamma up to 31 months after infection. At the time, I called IFN-γ a smoke alarm — a biomarker that signals ongoing disease.

But the Wei paper reveals something worse: IFN-γ isn't just the alarm. It's the arsonist. Persistent interferon signaling is the upstream driver of peroxisome destruction. Every day that IFN-γ stays elevated, it continues to block new peroxisome construction and trigger the autophagy of existing ones. The biomarker from Post #10 is actively causing the damage described in this post.

THE PEROXISOME DESTRUCTION CASCADE Persistent IFN-γ Blocked Biogenesis New peroxisomes can't be built Pexophagy (Autophagy) Existing peroxisomes destroyed Inflammasome Activation IL-1β → Fibrosis KRT8+ progenitor buildup in tissue IL-18 → Blood Signal Respiratory Long COVID biomarker

And there's a further connection. Gao et al. identified a plasma biomarker signature specifically for respiratory Long COVID symptoms: CCL3, CD40, IKBKG, IL-18, and IRAK1. IL-18, like IL-1β, is an inflammasome product — cleaved by caspase-1 and released through the same Gasdermin D pore. Peroxisome destruction produces both the tissue-damaging cytokine (IL-1β) and the blood biomarker of respiratory LC (IL-18). One organelle's failure explains both the tissue damage and the blood signature.

Two Cheap Drugs Nobody Is Testing

Here is what makes this story both hopeful and infuriating. Two FDA-approved, generic, widely available drugs have already shown they can restore peroxisome function in animal models of COVID-induced damage.

Drug Approved For Mechanism COVID Evidence LC Trial?
Sodium Phenylbutyrate Urea cycle disorders (Buphenyl) Peroxisome biogenesis promoter, chemical chaperone, HDAC inhibitor Restored peroxisome function, reduced lung fibrosis in mice (Wei 2025) None
Fenofibrate Hyperlipidemia (Tricor, generic) PPAR-α agonist → peroxisome proliferation Alleviated GI symptoms, restored epithelial repair (Wang 2026) None

Wei's team showed that sodium phenylbutyrate restored peroxisome function in macrophages, mitigated lung inflammation and fibrosis, and enhanced alveolar regeneration in mice. Wang's team showed that both phenylbutyrate and fenofibrate alleviated gastrointestinal symptoms by promoting peroxisome proliferation and restoring epithelial repair.

These aren't exotic experimental compounds. Phenylbutyrate has been prescribed for urea cycle disorders for decades. Fenofibrate is one of the most commonly prescribed cholesterol drugs in the world. Both are generic. Both are cheap.

Critical Caveat: All of this evidence is from mouse models and human tissue samples. As of March 2026, no clinical trial of either drug for Long COVID has been registered or announced. UVA has filed a patent on the peroxisome-targeting concept and is “collaborating with scientists and physicians” to advance the research — but no trial is underway. The FERMIN trial tested fenofibrate for acute COVID in 701 patients and found no benefit — but acute infection and chronic repair failure are different problems.

Why This Matters for the 44 Million

The peroxisome story is significant for three reasons.

First, it explains multisystem disease. Long COVID patients don't have a lung problem or a gut problem or a brain problem. They have a peroxisome problem — the same organelle failing in different tissues, producing different symptoms depending on where the damage is worst. This maps directly onto the eight subtypes of Long COVID we know exist.

Second, it connects Long COVID to ME/CFS through a shared biological mechanism, not just symptom overlap. If the Lipkin metabolomics data is correct, peroxisome dysfunction was causing post-viral illness long before SARS-CoV-2 arrived. COVID is the largest natural experiment in post-viral disease we've ever seen — and it's confirming what ME/CFS researchers suspected for years.

Third, it offers a therapeutic target with existing drugs. Not a moonshot. Not a ten-year development timeline. Drugs that are already manufactured, already safety-tested, already available at pharmacies. What's missing is the will — and the funding — to run the trials.


4
Research teams
4
Organ systems
2
FDA-approved drugs
0
Clinical trials

Your body knows how to heal. It's been doing it your entire life. But COVID can break the machinery that makes healing possible — tiny organelles called peroxisomes that most people have never heard of. Four independent research teams have now documented this damage in four different organ systems. Two cheap, existing drugs can restore these organelles in animal models.

The repair shops can be rebuilt. The tools exist. What's missing is someone willing to test them in the 44 million Americans — and hundreds of millions worldwide — who are still waiting to heal.

Sources: Wei et al., Science (2025). Wang et al., Developmental Cell (2026). Hobman & Power, Frontiers in Molecular Neuroscience (2025). Lipkin et al., Int J Mol Sci (2022). Krishna et al., Science Advances (2024). Gao et al., Nature Immunology (2025). UVA Newsroom (2025).