Researchers at UCLA have identified a coronavirus protein that appears to do two opposite things at once: quiet the body's earliest antiviral alarm while cranking up the inflammatory machinery that causes tissue damage later.
The protein is nucleocapsid, a structural protein that packages and protects the virus's genetic material and that has drawn far less attention than spike. In laboratory work published in Science Advances, a team led by Melody Li, an associate professor of microbiology, immunology and molecular genetics at UCLA, found that nucleocapsid amplified inflammatory pathways inside macrophages, the immune cells that patrol tissue and coordinate early defense.
The essential caveat belongs at the top. This is laboratory work in engineered cells and stem cell-based models. It identifies a plausible contributor to inflammation seen in COVID-19. It does not demonstrate the cause of any individual patient's long COVID symptoms, and no patients were studied.
Macrophages Are Meant to Sound an Alarm, Not Hold It
Understanding why this matters requires knowing what these cells normally do.
Macrophages sit in tissues throughout the body looking for signs of infection. When they detect a pathogen, they release signaling molecules called cytokines and chemokines that recruit other immune cells and coordinate a response. That inflammatory burst is protective when it is proportionate and brief.
When macrophages become overactivated, the same signals damage the tissue they were meant to defend. Sustained inflammatory signaling is one of the leading hypotheses for long COVID, alongside viral persistence, autoimmunity, microclotting and reactivation of other latent viruses. None of these has been established as the explanation, and they are not mutually exclusive.
The UCLA team set out expecting nucleocapsid to suppress immune responses, as many coronavirus proteins do. Zhenlan Yao, co-first author and a former postdoctoral researcher in Li's lab, said they found the opposite. Li described the protein as a double-edged sword: it still dampened early antiviral signals while simultaneously amplifying inflammatory pathways.
That pattern lines up with the clinical course clinicians observed throughout the pandemic, in which the virus blunts the immune response early and overdrives it later, when much of the tissue damage occurs. Li likened it to a thief who trips the alarm instead of staying quiet, and suggested the effect looks like an unintended byproduct rather than a viral strategy.
The Blood Vessel Finding Is the Most Concrete Result
To test whether overactivated macrophages could affect distant tissue, the researchers used two human cell-based models: a stem cell-derived model of the blood-brain barrier and a model of the coronary artery lining. Both are built from endothelial cells, which line blood vessels and control what passes from the bloodstream into surrounding tissue.
They exposed both to fluid containing signals from macrophages producing the Delta variant's nucleocapsid protein. Using fluid rather than direct infection isolates the effect of the signals themselves. The heart barrier broke down significantly, a phenomenon known as vascular leakage.
That specificity is worth noting. The reported breakdown was in the coronary model, and the finding points to a possible mechanism for the cardiac injury documented in severe COVID-19 rather than a general collapse of every barrier tested.
The team compared nucleocapsid proteins from SARS-CoV-1, MERS-CoV and several SARS-CoV-2 variants, and found the pro-inflammatory effect was conserved across pathogenic coronaviruses, with Delta's version the most inflammatory by far. That is a laboratory observation about protein behavior rather than a claim about how any currently circulating variant behaves in people.
The Distance Between This and a Treatment
The researchers suggest the finding points toward more precise therapy, and that reasoning is worth laying out honestly.
Severe COVID-19 is currently treated with broad anti-inflammatory drugs such as corticosteroids. Those dampen harmful inflammation but suppress immune function generally and do not target the viral mechanism driving it. Li suggested that a therapy or vaccine targeting the nucleocapsid protein could in principle rein in hyperinflammation more precisely.
That is a hypothesis about a direction, not a candidate drug. No compound has been announced, and the path from a mechanism in cultured cells to a treatment in patients typically takes many years and fails more often than it succeeds.
Pablo Alvarez, co-first author and a former graduate student in the lab, noted that continued COVID-19 research matters because not everyone responds well to vaccines and immunocompromised people often have limited treatment options, and because the work helps prepare for future coronavirus outbreaks.
Because macrophages play a similar double-edged role in many infections, the same mechanism could prove relevant beyond this virus. The lab has reported a comparable pattern in an unrelated mosquito-borne virus, suggesting viral structural proteins may modulate immunity more broadly than assumed. The work was supported by the National Institute of Allergy and Infectious Diseases, the W.M. Keck Foundation and the American Heart Association, with additional support from the UCLA Broad Stem Cell Research Center.
What This Changes for People Living with Long COVID
For patients, the honest answer is that this changes nothing about care today, and that is worth saying plainly rather than implying otherwise.
There is no test derived from this finding, no treatment, and no way to determine whether nucleocapsid-driven inflammation is contributing to any particular person's symptoms. Long COVID remains a clinical diagnosis based on persistent symptoms following infection.
What research like this does provide is a target for future work and a partial explanation for why symptoms can persist after the virus is cleared. For patients who have been told their symptoms are not physiological, mechanistic findings of this kind matter, even when they do not yet help.
People with ongoing symptoms after COVID-19 should work with a clinician on symptom management, and where available, a long COVID clinic. Pacing strategies for post-exertional symptom worsening, evaluation for treatable contributors such as orthostatic intolerance, and cardiac assessment for chest symptoms all have clinical value now.
Be cautious about clinics selling unproven long COVID treatments, particularly infusions and blood-filtering procedures marketed on inflammation theories. Mechanistic plausibility is not evidence of benefit. This article is general information and is not medical advice.
Frequently Asked Questions
What did the researchers find? The SARS-CoV-2 nucleocapsid protein amplified inflammatory signaling in macrophages while still suppressing early antiviral responses.
Does this explain long COVID? No. It identifies a possible contributor to inflammation. It does not demonstrate the cause of any patient's symptoms.
Was this done in people? No. The work used engineered macrophages and stem cell-based models of the blood-brain barrier and coronary artery lining.
What is vascular leakage? Breakdown of the tight blood vessel lining that normally controls what passes between blood and tissue. The reported breakdown was in the heart model.
Which variant was most inflammatory? Delta's nucleocapsid protein, though the pro-inflammatory effect was conserved across pathogenic coronaviruses including SARS-CoV-1 and MERS-CoV.
Is a treatment coming? No candidate has been announced. This is early mechanistic research, and the path to a therapy is long and uncertain.
What should long COVID patients do now? Work with a clinician on symptom management, and be wary of clinics selling unproven anti-inflammatory treatments.
