A comprehensive atlas of tiny, previously uncatalogued proteins in the human brain has been created, revealing that many microproteins are altered in Alzheimer’s disease. The study, published in Nature Aging, maps thousands of these short molecules in the frontal cortex and links them to cellular energy production and genetic regulation. Researchers say traditional gene databases may capture only a fraction of the brain’s functional proteins.
Microproteins are typically fewer than 150 amino acids long and arise from small open reading frames embedded within or near larger, well-known genes. Because of their brief length, they have long been difficult to detect with standard laboratory tools, leading to their exclusion from conventional genetic reference databases.
The Salk Institute team, led by Brendan Miller and Alan Saghatelian, built a customised genetic database to search for these hidden molecules in human brain tissue. They performed advanced RNA sequencing on frontal cortex tissue from 12 postmortem brains and appended these instructions to the standard protein reference database before analysing mass spectrometry data from 610 brain samples.
Initial analysis identified 4,321 microproteins, of which more than 3,000 were deemed unreviewed and not present in widely used, manually curated protein databases.
To confirm that these unreviewed microproteins were genuine, the researchers cross-checked their data with ribosome profiling and employed a deep learning tool called PROSIT to assess the quality of the protein fragments. This filtering process yielded 1,067 unreviewed microproteins with strong, highly confident evidence of presence in the brain.
When comparing brains from individuals without Alzheimer’s disease to those with symptomatic disease, dozens of microproteins showed altered levels. For instance, a microprotein derived from the BCL3 gene was less abundant in symptomatic brains, while a microprotein from the RAB3C gene was more abundant. In laboratory-grown human microglia, the brain’s resident immune cells, these microproteins were observed to remain stable and occupy specific intracellular regions.
The team also investigated how these microproteins are regulated. A substantial proportion appear to be produced independently of their neighbouring, larger parent genes. In many cases, the microprotein RNA changed in Alzheimer’s disease even when the parent gene’s RNA did not. This pattern of independent regulation was confirmed in a separate brain tissue dataset from another patient group.
Further analysis pointed to alternative splicing as a likely mechanism behind this separation. The researchers identified motifs near splice sites that attract RNA-binding proteins, suggesting a regulatory process that may go awry in neurodegeneration.
Among the microproteins studied, one encoded by the MKKS gene drew particular attention. Standard databases list the MKKS gene’s major product as a large protein associated with a rare developmental disorder, but in the adult frontal cortex the tiny microprotein appears to be the dominant product. Mass spectrometry data showed thousands of fragments matching the microprotein and very few aligning with the large protein, and its levels were lower in brains with symptomatic Alzheimer’s disease.
To probe its function, researchers mapped the network of genes active alongside the MKKS microprotein and found a strong overlap with mitochondrial function, the cell’s energy factories.
Experiments in microglia used CRISPR to delete the DNA sequence that encodes the MKKS microprotein without altering the main MKKS gene. When subjected to stress tests measuring the cells’ oxygen consumption rate, microglia lacking the microprotein consumed less oxygen, indicating a role in cellular respiration and energy production.
The study also explored the role of pseudogenes, sequences once regarded as nonfunctional. By examining RNA and ribosome data, the researchers found evidence that several pseudogenes are actively translated into microproteins. One such product, derived from the ACTBP7 pseudogene, is related to cytoskeletal proteins; when added to microglia, it altered the distribution of the cells’ structural fibres.
The authors caution that not detecting a microprotein by mass spectrometry does not prove its absence from the brain, as some may lie below current detection limits. Likewise, the exact genetic pathways that generate many short proteins remain to be fully mapped.
Identifying microproteins that change in Alzheimer’s brains does not by itself prove a causal role in the disease. Such alterations may reflect a cellular response to injury or shifts in regulatory networks. The researchers emphasise that future work will need to manipulate individual microproteins in living systems to distinguish drivers of neurodegeneration from markers of cellular stress.
The study, titled A microprotein atlas of the human frontal cortex in Alzheimer’s disease, was led by Brendan Miller, Eduardo Vieira de Souza, Calvin Lau, Joan M. Vaughan, Victor J. Pai, Servando Giraldez, Andréa Rocha, Jolene K. Diedrich, Clodagh C. O’Shea, David A. Bennett and Alan Saghatelian.
