Most of what scientists know about the microscopic structure of Alzheimer disease in the human brain comes from tissue examined after death.
That makes it difficult to study how the disease changes inside living brain tissue while proteins are still being produced, modified and cleared.
New research has now provided a rare glimpse of those processes during life.
A study published in Nature Communications on 24 September 2026 analysed small brain biopsies collected from 18 people undergoing ventriculoperitoneal shunt surgery for suspected normal pressure hydrocephalus.
Amyloid plaques were found in 9 of the 18 biopsies. Mature tau tangles were uncommon, but phosphorylated tau 217 was concentrated around some amyloid plaques.
Stable isotope labelling also showed that newly produced tau could be detected in the brain within about three hours, with an estimated brain tau half-life of roughly 34 days.
By contrast, the amyloid plaques showed little detectable turnover over the period that could be measured.
The findings suggest that tau and amyloid can behave very differently inside living human brain tissue and demonstrate that some of the earliest biochemical stages of Alzheimer pathology can be studied directly during life.
Living brain tissue is rarely available for Alzheimer research
Alzheimer disease develops many years before severe memory loss becomes obvious.
During this long period, amyloid beta begins accumulating outside neurons and abnormal tau changes develop within and around nerve cells.
Researchers can track parts of this process using brain scans, blood tests and cerebrospinal fluid biomarkers.
These tools are extremely valuable, but they cannot reveal every chemical detail occurring at the exact location of a plaque or within the surrounding tissue.
Direct brain tissue provides much richer molecular information.
The problem is that collecting brain tissue from living people purely for research would not normally be ethically acceptable.
The new study used a clinical situation in which a small amount of brain tissue could be obtained during surgery that participants were already undergoing for another medical reason.
The participants were undergoing shunt surgery
The 18 participants had suspected normal pressure hydrocephalus and were undergoing ventriculoperitoneal shunt placement.
Normal pressure hydrocephalus can cause problems with walking, bladder control and cognition, particularly in older adults.
The participants were mostly in their seventies and eighties and all had some cognitive symptoms.
Because Alzheimer pathology becomes increasingly common with age, this group provided an opportunity to study individuals who might have early or coexisting Alzheimer-related changes even though the surgery itself was not being performed to treat Alzheimer disease.
The researchers analysed the brain tissue alongside ventricular and lumbar cerebrospinal fluid and compared the results with post-mortem control tissue.
Half of the biopsies contained amyloid plaques
Amyloid beta pathology was detected in 9 of the 18 living brain biopsies.
The amount and type of amyloid varied between participants.
The tissue included cored plaques, diffuse plaques and neuritic plaques, as well as amyloid deposited in blood-vessel walls in at least one participant.
This diversity matters because amyloid plaques are not chemically or structurally identical.
Different plaque forms may represent different stages of development or interactions with surrounding nerve cells.
Studying them in living tissue allowed the researchers to examine those differences without relying entirely on changes that may occur after death.
Mature tau tangles were rare
One of the most notable findings was that mature tau tangles were uncommon in the biopsied tissue.
This suggests that the sampled tissue captured relatively early stages of Alzheimer-related pathology in several participants.
Tau pathology usually becomes more extensive as Alzheimer disease advances.
Finding amyloid plaques with relatively little mature tangle pathology therefore gave the researchers an opportunity to examine what may happen before widespread tau aggregation is established.
This is particularly important because many treatments are likely to work best before substantial neuronal damage has accumulated.
Phosphorylated tau clustered around amyloid plaques
Although mature tau tangles were rare, phosphorylated tau 217 was not evenly distributed through the tissue.
It was enriched around neuritic amyloid plaques.
The signal was strongest near the plaque and became weaker as the distance from the plaque increased.
This spatial pattern supports the idea that amyloid deposition and abnormal tau phosphorylation are biologically connected.
The study cannot by itself establish every step of that relationship, but seeing the two processes together in living human tissue adds direct evidence to a link that has previously been inferred from imaging, fluid biomarkers, animal models and post-mortem studies.
The researchers mapped plaque chemistry directly inside the tissue
The study used matrix-assisted laser desorption ionisation mass spectrometry imaging to examine the chemical composition of amyloid plaques.
This technique can detect different molecular forms of a protein while preserving information about where those molecules are located in a tissue section.
The researchers could therefore map the chemistry of individual amyloid deposits rather than measuring only the average protein concentration of the entire biopsy.
They identified several modified forms of amyloid beta, including species associated with plaque maturation.
This type of spatial chemical analysis could eventually help researchers understand why some plaques remain relatively stable while others are associated with stronger local neuronal injury.
Stable isotopes allowed the researchers to measure protein turnover
The study did more than describe what the tissue looked like at one moment.
Participants received a stable isotope labelled amino acid before surgery.
As cells produced new proteins, some of the labelled building blocks were incorporated into newly synthesised molecules.
Mass spectrometry could then detect how much labelled protein had appeared over time.
This method gave the researchers a way to estimate how quickly important Alzheimer-related proteins were being produced and replaced inside the living brain.
Newly made tau appeared within about three hours
Tau showed surprisingly rapid incorporation of the stable isotope label.
Newly synthesised tau could be detected in soluble brain extracts within approximately three hours.
The researchers estimated that tau in the brain had a half-life of about 34 days.
A half-life describes the approximate time required for half of a pool of molecules to be replaced or cleared.
This does not mean every tau molecule survives for exactly 34 days.
It provides an estimate of the overall turnover rate within the measured protein population.
The result shows that soluble tau is a dynamic protein pool even in older human brains.
Amyloid plaques behaved very differently
The amyloid plaques showed far less evidence of rapid turnover.
Within the period that could be measured, the deposits appeared comparatively stable.
This contrast is biologically important.
Tau appears to be continuously synthesised and cleared on a timescale that can be measured over weeks.
Established amyloid plaques, by comparison, may persist much longer once they have formed.
This difference may partly explain why changing the amount of soluble protein and removing established deposits can require different therapeutic approaches.
Ventricular fluid more closely resembled the brain than lumbar fluid
The surgery also gave researchers access to cerebrospinal fluid from the brain’s ventricles.
They compared this with cerebrospinal fluid collected from the lower spine.
The tau profile in ventricular cerebrospinal fluid more closely resembled the profile measured directly in brain tissue.
This suggests that the location from which cerebrospinal fluid is sampled can influence what it reveals about brain protein metabolism.
It also provides a useful biological reference for interpreting lumbar cerebrospinal fluid tests, which are much easier to obtain clinically but are physically further from the brain.
The study captured pathology before the most advanced stages
The combination of amyloid plaques, limited mature tau tangles and changing biomarker profiles suggests that several participants were sampled relatively early in the Alzheimer disease continuum.
This is one reason the study is important.
Post-mortem tissue often represents the endpoint of years or decades of disease.
By that stage, it can be difficult to distinguish changes that helped initiate disease from those that appeared only after severe pathology developed.
Living tissue collected earlier may allow researchers to study events closer to the beginning of the pathological process.
This was a proof-of-principle study
The sample was small.
Only 18 living participants contributed biopsies, and only half of those biopsies contained amyloid pathology.
The study was therefore designed primarily to show that this type of analysis is technically possible rather than to estimate how common each pathological feature is in the wider population.
Larger studies would be needed to determine how reliably the measured protein-turnover patterns vary across different stages of Alzheimer disease.
The participants were not a representative Alzheimer cohort
The participants were selected because they were undergoing surgery for suspected normal pressure hydrocephalus.
They were not recruited as a population-based sample of people with Alzheimer disease.
Normal pressure hydrocephalus can itself affect cognition and brain physiology.
Some participants may also have had mixed neurological conditions.
The findings therefore should not be interpreted as describing the exact biology of every person with Alzheimer disease.
A biopsy samples only a tiny part of the brain
Alzheimer pathology is not evenly distributed through the brain.
A small biopsy can reveal tremendous molecular detail, but it represents only the tissue that happened to be sampled.
A participant could have more extensive pathology elsewhere that was not captured by the biopsy.
This is why the tissue findings need to be interpreted alongside imaging, cerebrospinal fluid biomarkers and clinical information.
The strength of the biopsy is its chemical resolution, not its ability to map the entire brain.
The study does not support routine brain biopsy for Alzheimer diagnosis
The research should not be interpreted as a proposal to biopsy the brains of people being assessed for memory problems.
Brain biopsy is invasive and carries risks.
The participants were already undergoing neurosurgery for a separate clinical reason, which created an unusual research opportunity.
Blood tests, cerebrospinal fluid biomarkers and brain imaging remain far more practical methods for diagnosing and monitoring Alzheimer disease.
The value of the biopsy approach is mainly scientific because it can help researchers understand what those less invasive biomarkers actually reflect inside the brain.
Living tissue could help test how future treatments work
The ability to measure protein turnover directly could eventually become useful in small mechanistic studies of new therapies.
A treatment designed to reduce tau production, alter tau clearance or change amyloid-plaque chemistry could potentially be evaluated at a molecular level in living tissue when appropriate surgical opportunities exist.
The researchers also suggest that the approach could be applied to other neurodegenerative diseases.
Conditions involving alpha-synuclein, TDP-43 or other abnormal proteins may contain similarly important differences between soluble proteins and long-lived aggregates.
The bigger lesson is that Alzheimer pathology is dynamic
Amyloid plaques can look static in a microscope, but Alzheimer disease develops through interacting biological processes that unfold over time.
The new study shows that those processes can now be measured directly in living human brain tissue under rare but clinically appropriate circumstances.
The contrast between rapidly turning over tau and comparatively stable amyloid plaques is particularly striking.
It suggests that different components of Alzheimer pathology operate on very different biological timescales.
Understanding those timescales may become increasingly important as treatments move from simply detecting abnormal proteins toward changing how they are produced, modified, accumulated and cleared.
Source Information
Study Title: Spatiotemporal characterization of Alzheimer disease pathology in living human brain tissue
Authors: Aram Aslanyan, Soumya Mukherjee, Eleanor M. Moncur, Alicja Szadziewska, Yingxin He, Przemysław R. Kac, Reid Coyle, Elena Camporesi, Kaj Blennow, Fernando Gonzalez Ortiz, Claire A. Leckey, Tatiana A. Giovannucci, Kanza Tariq, Henrik Zetterberg, Donald L. Elbert, Lewis Thorne, Ahmed K. Toma, Laurence Watkins, Nick C. Fox, Randall J. Bateman, Jörg Hanrieder, Ross W. Paterson and colleagues
Journal: Nature Communications
Published: 24 September 2026
Sample/Dataset: Brain biopsies from 18 people undergoing ventriculoperitoneal shunt surgery for suspected normal pressure hydrocephalus, together with matched ventricular and lumbar cerebrospinal fluid and post-mortem comparison tissue.
Method: The researchers combined immunofluorescence, luminescent amyloid staining, mass spectrometry imaging, biochemical protein analysis and stable isotope labelling to characterise amyloid and tau pathology and estimate protein turnover directly in living human brain tissue.
Main finding: Amyloid plaques were present in 9 of 18 biopsies, while mature tau tangles were rare. Phosphorylated tau 217 was enriched around amyloid plaques. Newly synthesised tau could be detected within about three hours and had an estimated brain half-life of roughly 34 days, while established amyloid plaques showed little detectable turnover during the measurement period.
DOI: 10.1038/s41467-026-76494-4
