A blocking antibody against the chemokine receptor CXCR3 cut T cell infiltration into the brains of tauopathy mice by roughly half and preserved about 40% more tissue in memory centers, without changing tau levels, researchers at Washington University School of Medicine in St. Louis report in Neuron.

The team injected the antibody into young mice with Alzheimer's-like tau accumulation every five days for three and a half months, starting before major brain cell loss had occurred. At the end of treatment, treated animals carried about half as many brain CD4+ and CD8+ T cells as untreated controls, showed less evidence of nerve cell damage, and performed better on a memory test. Tau tangles, measured in both soluble and insoluble fractions, appeared identical in treated and untreated animals.

"In tauopathies, including Alzheimer's disease, there's no treatment right now that actually decreases neurodegeneration," said David M. Holtzman, MD, the Barbara Burton and Reuben M. Morriss III Distinguished Professor in WashU Medicine's Department of Neurology and senior author of the paper. "If we can show that we're really decreasing brain cell death, it's certainly worth trying to pursue that pathway to prevent the most devastating consequences of these diseases."

The study, first-authored by postdoctoral researcher Joshua T. Emmerson, PhD, with co-senior author Jason Ulrich, PhD, professor of neurology, defines what the authors call "the CXCR3 axis as a critical target that, when blocked, mitigates CD4+ and CD8+ T cell infiltration and confers neuroprotection in a model of tau-mediated neurodegeneration in vivo without influencing levels of soluble or insoluble tau."

The mechanism work rests on a chemokine trail. T cells navigate by following chemical signals, and the Holtzman lab had previously found elevated CXCL10—one such chemokine—in their tau mouse model; other groups had reported the same elevation in Alzheimer's patients. Activated T cells carry CXCR3, the receptor that binds CXCL10. In the current study, mice genetically lacking either CXCL10 or CXCR3 on T cells showed no brain infiltration even when researchers deliberately provoked inflammation.

One finding carries particular weight for translational planning: biodistribution tests showed the antibody reached the border of the brain but did not enter brain tissue itself. The therapeutic effect therefore does not require crossing the blood-brain barrier—a property that eliminates one of the costliest failure modes in CNS drug development.

"For this therapeutic approach, if it is safe, you wouldn't have to design the drug to get into the brain—which is a big deal since most molecules don't cross the blood-brain barrier well—and you don't have to get rid of the tau to get this therapeutic effect," said Holtzman, who also directs WashU's Hope Center for Neurological Disorders and the Knight Alzheimer Disease Research Center.

The result extends a research arc from the same lab. A 2023 Nature paper showed T cells flooding into tau-laden mouse brains and driving neurodegeneration; a related paper published earlier this month in Nature Neuroscience traced those T cells' instructions to lymph nodes outside the brain. The question of how the cells physically entered the brain remained open until this CXCR3 work.

The commercial and clinical context sharpens the gap this approach targets. The two marketed Alzheimer's antibodies, lecanemab and donanemab, clear amyloid plaques and slow decline but have not demonstrated an ability to prevent brain cell death. They also do not address primary tauopathies—diseases marked by tau buildup in which amyloid never appears. Alzheimer's is a secondary tauopathy in which both proteins accumulate.

Holtzman points to a near-term repurposing path: existing T cell-directed drugs, including therapies approved for multiple sclerosis and other autoimmune disorders, could be evaluated as Alzheimer's candidates. "Tauopathies aren't thought of as autoimmune disorders, so they haven't been treated the same way, but this study shows for the first time in an animal model that these diseases respond to a specific T-cell therapy," he said.

The usual caveats apply. The data come from a transgenic mouse model, not human tissue; the treatment window began before significant neurodegeneration, and the study does not establish whether intervention works after cell loss is underway. No human dosing, safety, or efficacy data exist yet. The authors state that peripheral CXCR3 inhibition "could be a therapeutic approach in tauopathies"—a projection that will require clinical testing to validate.

Holtzman's group and collaborators will now need to show that T cell blockade can translate from a three-and-a-half-month mouse dosing regimen to a safe, durable therapy in people.