Peripheral immune priming worsened tau-linked brain damage in mice

A study published September 3 in *Nature Neuroscience* reports that an immune process occurring largely outside the brain helped drive neurodegeneration in mice with tau pathology. The work connects specialized dendritic cells in peripheral tissues to the CD8 T cells that accumulate in affected brains.

Mice genetically lacking conventional type 1 dendritic cells, known as cDC1s, showed less brain atrophy, fewer infiltrating CD8 T cells and greater preservation of neurons. A second genetic experiment that retained the dendritic cells but disabled their ability to cross-present antigens produced similar protection.

The findings refine an emerging view of tauopathy as involving interactions between the nervous and immune systems. They do not establish that suppressing dendritic cells or T cells would safely treat Alzheimer’s disease in people.

How the pathway was tested

The principal model combined a human P301S tau transgene with human APOE4. At approximately 9.5 months, these TE4 mice develop substantial tau-associated inflammation and tissue loss.

Researchers compared tau-expressing and non-tau mice with or without an enhancer required for cDC1 development. The analysis included male and female animals and measured regional brain volume, dentate-gyrus neurons, plasma neurofilament light, nest-building behavior, glial reactivity and immune-cell populations.

Compared with cDC1-intact TE4 mice, animals lacking cDC1s had better-preserved hippocampal and cortical volumes, more dentate-gyrus neurons and higher nest-building scores. The protective pattern appeared without a substantial reduction in insoluble or tissue-stained phosphorylated tau, suggesting that the immune pathway affected damage downstream of, or alongside, the underlying tau burden.

Single-cell and T-cell-receptor sequencing showed fewer activated and clonally expanded CD8 T cells in the brains of cDC1-deficient mice. Removing cDC1s also reduced several measures of microglial and astrocyte activation, although some effects differed by sex. Plasma neurofilament light was significantly lower in deficient males but not females, and several microglial markers were not significantly reduced in females.

Evidence for peripheral activation

Dendritic cells can cross-present material they acquire to CD8 T cells, prompting those T cells to become activated. The researchers detected presentation of a brain-derived experimental antigen in secondary lymphoid tissues and found that disrupting the cross-presentation protein WDFY4 protected tauopathy mice even though cDC1s remained present.

The team also transferred T cells from older TE4 mice into the hippocampus of younger TE4 mice. Eight weeks later, recipients showed increased glial activation, but not greater tau accumulation or measurable hippocampal atrophy. This supports an inflammatory effect while stopping short of showing that transferred cells alone caused neuronal loss during that interval.

Mass-spectrometry profiling identified thousands of peptides bound to antigen-presenting molecules in brain tissue. Tauopathy brains contained increased representation of peptides derived from tau and other neuronal proteins. The study did not determine which specific antigen, if any, drove the damaging T-cell clones.

Human findings remain preliminary

Postmortem frontal-lobe tissue from eight people with primary tauopathies showed more CD8 T cells in gray and white matter than tissue from five controls. Analyses of existing cerebrospinal-fluid and meningeal datasets did not identify clear disease-associated changes in dendritic-cell abundance or antigen-presentation gene expression.

Those observations show that T-cell infiltration is not unique to the mouse model, but they do not demonstrate the proposed cDC1 mechanism in humans. The human sample was small, cross-sectional and drawn from people who already had advanced disease.

What the study does not establish

The genetic alterations were present throughout development rather than introduced as treatments after disease began. P301S tau with APOE4 models selected features of tauopathy and cannot reproduce the full biology of sporadic Alzheimer’s disease. Broad interference with dendritic-cell or T-cell function could also impair infection control, tumor surveillance and vaccine responses.

The practical result is therefore mechanistic: peripheral antigen presentation appears necessary for much of the damaging CD8 T-cell response in this model. Identifying the relevant antigens, confirming the pathway in living people and finding a selective way to modify it remain separate challenges.