Summary

Researchers mapped 4,321 microproteins in 608 post-mortem human brain samples, including dozens with altered expression in Alzheimer’s disease. The atlas expands the molecular inventory available for studying neurodegeneration, while the proteins’ biological roles still require experimental validation.

Researchers have identified 4,321 microproteins in human brain tissue, creating a large molecular atlas that includes more than 1,000 proteins previously absent from standard catalogues. Dozens showed altered expression in samples from people with Alzheimer’s disease.

The study, published in Nature Aging on 14 September 2026, analysed 608 post-mortem samples from the dorsolateral prefrontal cortex, a brain region involved in cognitive control. The findings provide a new set of molecular candidates for research into ageing and neurodegeneration. They form a research map rather than a diagnostic or treatment result.

A hidden layer of the brain proteome

Microproteins are proteins made up of fewer than 150 amino acids. Their small size makes them difficult to detect with standard mass spectrometry, a technique widely used to identify the proteins present in cells and tissues.

Some microproteins are produced from parts of the genome previously regarded as non-coding. Others come from genes that also produce larger, better-known proteins. This range of possible origins makes them difficult to capture through conventional RNA sequencing alone.

The research team combined three approaches: mass spectrometry, RNA sequencing and ribosomal profiling. Ribosomes are the cellular machines that build proteins from messenger RNA. Ribosomal profiling examines the RNA strands attached to these machines, providing evidence that particular RNA sequences are actively being used to produce proteins.

This combined analysis identified 4,321 microproteins. Of these, 3,217 had not previously been characterised in UniProtKB/Swiss-Prot, a standard catalogue of human proteins. The researchers then used a deep-learning model to assess 3,001 of the identified microproteins according to the model’s confidence in their mass-spectrometry detection. It ranked 1,067 as having strong detection confidence.

What the Alzheimer’s signal means

The samples came from people with and without Alzheimer’s disease, allowing the researchers to compare microprotein expression between the two groups. Dozens of the small proteins showed altered expression in the Alzheimer’s samples.

That result matters because Alzheimer’s disease is partly associated with proteins that misfold, accumulate or trigger toxic responses in the brain. Mapping a wider range of proteins could reveal molecular changes that are missed when research focuses only on larger, familiar proteins.

The atlas currently identifies candidate proteins and disease-associated expression patterns. Assigning biological functions to the proteins will require further experiments. The researchers have made the dataset publicly available so that other laboratories can study individual sequences and test how the proteins are produced and what they do in brain cells.

The evidence is based on post-mortem tissue from one brain region, so the immediate contribution is a detailed molecular reference for research. Whether particular microproteins help drive Alzheimer’s biology, result from disease-related changes or could eventually become biomarkers or therapeutic targets remains an open question.

Sources