St. Jude Scientists Build First Cell-Type-Specific Alzheimer’s Protein Atlas

More than 3,000 brain proteins show altered levels in Alzheimer’s disease, and 78% of those changes had gone undetected until now.

That is the scale of the first cell-type-specific protein atlas, built by scientists at St. Jude Children’s Research Hospital and published this week in Neuron, News-Medical reported.

Alzheimer’s affects nearly the whole brain, yet which cell types carry which molecular damage has remained unsettled. The new atlas, assembled from the brains of two mouse models of the disease, attaches specific protein changes to specific cells instead of averaging them across whole tissue. Researchers identified the changes by layering three technologies: next-generation mass spectrometry, single-cell RNA sequencing and spatial omics.

The reason it matters is disarmingly simple. RNA sequencing, which measures gene activity, tells researchers what instructions a cell is reading. It cannot tell which proteins those instructions actually produce, or where those proteins end up. Protein analysis fills that gap, but until now it has mostly been done in bulk — grinding up tissue and losing all sense of which cell made what.

“This is the first cell-type-resolved proteomic analysis we are aware of in the Alzheimer’s disease field,” said Junmin Peng, PhD, of St. Jude’s Departments of Structural Biology and Developmental Neurobiology and a co-corresponding author of the paper.

“When we compared our data to bulk datasets, we found that a large portion of the protein level changes we detected are new. We also identified many potential disease drivers that can be targeted in future follow-up studies.”

St. Jude Scientists Build First Cell-Type-Specific Alzheimer's Protein Atlas
Photo: stjude.org

Proteins move between different brain cell types

  • Stay-level changes were hiding in plain sight: proteins can be made by one brain cell type and do their work in another — a pattern bulk analysis smoothed over entirely.
  • Roughly four in five of the detected protein changes (78%) had never been captured before, meaning current drug discovery may have been aiming at a partial picture.
  • Nothing here changes treatment today. This is an early-stage research resource; the atlas and its datasets are free for any researcher to use.

Protein levels mismatch RNA data

The team, led by Peng and co-corresponding author Jiyang Yu, PhD, interim chair of St. Jude’s Department of Computational Biology, profiled more than 13,000 proteins across multiple ages and cell types in two Alzheimer’s mouse models, News-Medical reported. Matching that catalogue against single-cell RNA data let them compare what each cell type says with what it actually builds.

The mismatch they found is the study’s sharpest insight. Pleiotrophin, a secreted protein, showed high RNA levels in astrocytes — the brain’s support cells — but high protein levels in microglia, the brain’s immune cells, which carry many known Alzheimer’s risk genes. The gene is read in one cell; the finished protein works in another.

“Because those proteins are secreted, this suggested an interesting form of intercellular communication,” Peng said. “Gene regulation may occur in astrocytes, but after translation and secretion, the proteins function in microglia.”

A focused look at microglia drove the headline figure: almost all of the protein-level changes to that cell type had been previously undetected. Microglia are where the strongest genetic risk signals for Alzheimer’s already cluster, so an almost entirely unseen layer of change there is a substantial blind spot now partly filled.

Researchers benefit from the public PANNDA portal

The immediate beneficiaries are researchers rather than patients. The atlas and accompanying datasets are publicly available at the team’s PANNDA web portal, giving anyone a benchmark for moving from bulk analysis to single-cell work.

What the atlas covers Figure
Proteins profiled across ages and cell types over 13,000
Proteins with altered levels across cell types over 3,000
Share of altered proteins not previously captured 78%

“Because our lab specializes in deep proteomics, we were able to generate exceptionally robust datasets,” said Xue Zhang, PhD, a postdoctoral researcher in Peng’s lab and co-first author. “That depth allowed meaningful comparisons with RNA data. Without sufficient coverage, such integration and analysis would not be possible.”

Yu framed the payoff in terms of cellular crosstalk. “These technologies enable us to understand disease mechanisms more comprehensively and identify specific cellular interactions, for example between microglia and astrocytes,” he said, per St. Jude’s news release.

What this cannot tell patients yet

News-Medical and Medical Xpress reported the study. No drug, diagnostic or prevention strategy comes from this study. Its stated fruits are “potential disease drivers that can be targeted in future follow-up studies,” in Peng’s words — candidate targets, not treatments.

Laboratory research does not change clinical protocols

Because this is basic laboratory research, there is no clinical protocol here, no intervention to adopt and none to avoid. No contraindae apply, because nothing is being prescribed.

The general caution this news does support: readers should not interpret protein-level research, or any supplement marketed on the back of it, as disease-modifying. Alzheimer’s risk assessment, cognitive concerns or memory symptoms warrant evaluation by a physician or neurologist, and established diagnostic and supportive care pathways remain unchanged. Patients already in treatment should not alter anything based on this study.

A blueprint for what comes next

Co-first author Jiyuan Yang, PhD, a senior computational research scientist in Yu’s lab, offered the methodological lesson. “True integration has to start with experimental design, not after the data are generated,” Yang said. “It also requires a clear understanding of how different types of data can complement and inform each other. We hope our study provides a useful blueprint for future multi-omics research.”

The study was supported by the National Institutes of Health under four grant numbers (RF1AG068581, R01AG092468, RF1AG064909 and U19AG069701) and by ALSAC, St. Jude’s fundraising organization — funding the institution disclosed in its release, stjude.org reported.

St. Jude lists co-corresponding author Zhexing Wen of Emory University and co-first authors Kaiwen Yu and Him Shrestha of St. Jude and Ying Zhou of Emory University alongside the team of 24.

What follows in the source’s own words is Peng’s identification of “potential disease drivers that can be targeted in future follow-up studies” — follow-up studies that, as of this week, now have a public atlas to start from.

References

Disclaimer: This article reports on published laboratory research and contains no treatment guidance. For personal medical concerns, consult a qualified physician.

Photo of author

Health Desk

Health Desk at Archyde.

DWTS Season 35: Harry Shum Jr. and Jenna Johnson Shockingly Eliminated After Perfect Semi-Final Scores