Tau protein changes reveal new targets for neurodegenerative diseases

Tau protein aggregation is a defining feature of more than 20 neurodegenerative diseases, collectively known as tauopathies. These include Alzheimer’s disease and chronic traumatic encephalopathy (CTE). However, new research led by Boston Children’s Hospital suggests that treating these disorders as molecularly similar may be wrong.

The study challenges the long-standing assumption that tauopathies can be diagnosed and treated using a single framework. Instead, the findings demonstrate that each disease is characterised by a distinct chemical signature of tau protein modifications, potentially leading to the development of more precise diagnostics and targeted therapies.

Mapping tau across 203 brains

The research team was led by senior authors Dr Judith Steen and Dr Hanno Steen, with experimental work carried out by co–first authors Dr Mukesh Kumar, Dr Christoph Schlaffner, Dr Shaojun Tang and Maaike Beuvink. The scientists analysed post-mortem brain tissue from 203 patients representing multiple tauopathies.

To do so, they used a novel mass spectrometry platform known as FLEXITau. This tool allows absolute quantification of pathological tau species, measuring both the identity and abundance of disease-relevant chemical modifications. Unlike traditional approaches, FLEXITau captures detailed molecular changes rather than broad structural features.

From Alzheimer’s to a broader atlas

The work builds on the team’s earlier studies in Alzheimer’s disease, where they showed that tau chemistry evolves as the disease progresses. In that research, the p217 tau modification emerged as the most accurate diagnostic marker for Alzheimer’s disease and has since become FDA-approved.

Applying the same technology across multiple tauopathies, the researchers identified 145 post-translational modifications and 195 cleavage sites on tau. Machine-learning models were then used to rank which molecular features best distinguished each disease.

“For the first time, we can tell diagnostics and drug developers exactly which post-translational modifications to target across tauopathies, where they are on the protein, and how abundant they are in each disease,” said Dr Steen, Director of the Neuroproteomics Laboratory at Boston Children’s. “Instead of guessing which tau forms matter, we now have a precise molecular roadmap.”

Machine learning highlights what matters most

While cryo-electron microscopy has previously revealed disease-specific tau structures, the chemical composition of tau is poorly understood. The FLEXITau approach bridges that gap by providing quantitative chemical data suitable for advanced computational analysis.

“The machine learning analysis ranks modifications by importance to disease,” said Dr Steen. “This provides a priority list for diagnostics and drug development – the modifications that matter most. Machine learning and other AI tools require high-quality data and standards, and this method, called FLEXIQuant, can standardise measurements of any protein of interest, whether in neurodegeneration or cancer.”

Implications for future therapies

Quantifying how much of each molecular target exists also has direct implications for drug development.

“Knowing how much of a molecular target exists is essential for diagnostic or drug design,” said Dr Steen. “If a modification is rare or low abundance, it’s not a viable target. FLEXITau gives us the quantitative data needed to model dosing, pharmacokinetics and therapeutic feasibility.”

The findings further suggest that distinct enzymatic ‘writer’ and ‘eraser’ pathways drive tau pathology in different diseases. “The chemical signatures reflect specific enzymatic activities,” noted Dr Steen. “This opens new avenues for targeting the enzymes that generate disease-specific tau forms.”

Validated in an independent cohort, the resulting atlas provides a foundation for precision diagnostics, imaging and therapeutics across tau-mediated neurodegeneration. The FLEXIQuant platform can also be extended to other disease-related proteins, including synuclein in Parkinson’s disease or TDP43 in ALS. By pinpointing which tau modifications are most relevant in each disease, the findings could help to inform future drug discovery, enabling researchers to design therapies that precisely target the molecular drivers of neurodegeneration.