More than 1,000 previously overlooked ‘microproteins’ have been found in samples of human brain tissue, providing researchers with the most comprehensive database yet to study these hidden molecules inside the human brain. Dozens of the tiny proteins showed altered expression in people with Alzheimer’s disease, which could pave the way to unlocking new mechanisms of ageing and neurodegeneration.

Microproteins are small proteins with fewer than 150 amino acids and are notoriously difficult to detect with standard gene and protein sequencing techniques. In a study published in Nature Aging on 14 September¹, researchers combined several methods to pinpoint the tiny molecules in postmortem samples of the dorsolateral prefrontal cortex — a region of the brain involved in cognitive control — from people with and without Alzheimer’s disease.

The team identified more than 4,300 microproteins, creating the largest atlas of microproteins in Alzheimer’s disease made so far.

“We might be actually missing a whole layer of biology by overlooking these microproteins,” says Bahareh Ajami, a neuroimunologist at Cedars-Sinai Medical Center in Los Angeles, California.

“Alzheimer’s disease is a proteinopathy [wherein] the pathology is in part due to the proteins that have misfolded or accumulated and evoked a toxic response,” says study co-author Brendan Miller, a neuroscientist at the Salk Institute for Biological Studies in San Diego, California. “It should be somewhat urgent to understand the full proteome”, including microproteins, he adds.

A new detection tool

Microproteins are sometimes referred to as “the dark matter of the genome”, says Miller. Because of their short length, the standard method to identify the protein content of cells, known as mass spectrometry, often fails to detect them.

Some microproteins are made by parts of the genome that were thought to be non-coding and unable to produce proteins. But others are made from coding genes that also produce large, known proteins, explains Miller. This has made it difficult to find microproteins using conventional RNA sequencing techniques.

‘Dark proteins’ hiding in our cells could hold clues to cancer and other diseases

To overcome these challenges, Miller and his colleagues used a combination of mass spectrometry, RNA sequencing and ribosomal profiling. The latter tool involves collecting all ribosomes — the cellular machines that build proteins from RNA — in each cell and sequencing the individual strands of messenger RNA that are bound to them.

Using this approach, the team analysed 608 postmortem brain samples from individuals with and without Alzheimer’s disease. They identified 4,321 microproteins, 3,217 of which had not been previously characterized in a standard catalogue of human proteins called UniProtKB/Swiss-Prot.

Miller and his colleagues then applied a deep-learning model to 3,001 of these microproteins, which ranked them depending on the model’s confidence that they had been correctly identified using mass spectrometry data. The model ranked the detection of 1,067 microproteins with a strong level of confidence.

The atlas does not establish the biological function of these proteins, note Miller and his colleagues, who made the data set of microproteins publicly available. “Different labs can go in, download all of our sequences” and design experiments to validate them, says Miller.