CSF Proteins Spot Alzheimer’s Mutation Carriers Decades Before Symptoms

Long before carriers of autosomal-dominant AD mutations notice the first hints of memory loss, changes in their brains are well underway at the protein level. In a study published September 26 in Cell, scientists led by Carlos Cruchaga at Washington University in St. Louis provide further support for this concept.

  • CSF proteomics pegs 137 proteins that diverge in ADAD mutation carriers along AD continuum.
  • Twelve changed before classical AD biomarkers did.
  • Six—GFAP, NPTX2, PEA15, SMOC1, SMOC2, and TNFRSF1B—collectively predicted mutation status better than did classical AD biomarkers.

Measuring more than 6,000 proteins in the cerebrospinal fluid of participants in the Dominantly Inherited Alzheimer’s Network (DIAN), the researchers identified 137 harbingers of AD pathophysiology. A dozen of these changed even before conventional AD biomarkers Aβ or p-tau did. Together, six of them better predicted a person’s mutation carrier status. Notably, all 137 were significantly skewed in the CSF of people with sporadic AD, too, suggesting the findings apply beyond the most deterministic familial forms.

Alzforum covered the findings when they debuted in a medRxiv preprint earlier this year (Feb 2024 news). Since then, the researchers have reanalyzed their proteomics trove with fresh statistical methods, and replicated them using other proteomics techniques. Read on for a few highlights from the updated findings.

Co-first authors Yuanyuan Shen and Jigyasha Timsina used the Somascan proteomics platform to measure 6,163 proteins in CSF of 286 mutation carriers and 184 noncarriers. The researchers placed each participant along a pseudotime trajectory of disease progression, using estimated years from onset (EYO) as a timescale. They split the samples into discovery and replication cohorts, ultimately identifying 137 proteins that distinguished carriers from noncarriers at some point along the trajectory. In the LOAD experiment, the concentration of each of the 137 was either up or down in the CSF of 848 people with biomarker-confirmed sporadic AD relative to 915 healthy controls.

Trifecta of Trouble. As a mutation carrier approaches the EYO, three phases of CSF proteomic changes unfold, each marked by distinct biological pathways. [Courtesy of Shen at al., Cell, 2024.]

The changes hint that multiple biological pathways get successively involved. This would unfold in three phases as participants approach their EYO: an early stage marked by changed stress responses, glutamate metabolism, and neuron mitochondrial damage; a middle stage featuring apoptosis proteins; and a late presymptomatic stage, when microglial and cell communication proteins start to go awry.

While p-tau, total tau, and Aβ42 changed at 17, 12, and 11 years prior to EYO, respectively, 12 other proteins changed even earlier. These included the extracellular matrix protein SMOC1 at 31 years, followed by DNAJB9 at 26 years, nd SMOC2 at 20 years before EYO.

Could any of these proteins predict mutation carrier status? Using a machine-learning approach and separating the cohort into training and testing samples, the scientists identified a cadre of six that did the job. They are GFAP, NPTX2, PEA15, SMOC1, SMOC2, and TNFRSF1B. With 91 percent accuracy, this group boasted stronger predictive power than any classic AD markers, none of which reached 80 percent.

The researchers put this predictive panel to the test using other proteomics methods. Using O-link, an antibody-based quantitative technique, they were able to measure all but PEA15. The remaining five proteins predicted mutation carrier status with an accuracy of 83.7 percent, and distinguished between presymptomatic and symptomatic carriers with 87.9 percent accuracy. Three of the proteins—SMOC1, GFAP, and NPTX2—are included in Alamar’s CNS Nulisa platform, a new technique that use nucleic acid tags to amplify antibody signals and detect vanishingly minute amounts of protein (Aug 2024 conference news). Even with just these three proteins, Nulisa measurements teased out mutation carriers, and their clinical status, with 78.1 and 87.6 percent accuracy, respectively.—Jessica Shugart

Comments

No Available Comments

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Proteomics Uncovers Potential Markers of Early Autosomal Dominant AD
  2. NULISA—A New Proteomic Method to Revamp Biomarker Analysis

Further Reading

Primary Papers

  1. Shen Y, Ali M, Timsina J, Wang C, Do A, Western D, Liu M, Gorijala P, Budde J, Liu H, Gordon B, McDade E, Morris JC, Llibre-Guerra JJ, Bateman RJ, Joseph-Mathurin N, Perrin RJ, Maschi D, Wyss-Coray T, Pastor P, Goate A, Renton AE, Surace EI, Johnson EC, Levey AI, Alvarez I, Levin J, Ringman JM, Allegri RF, Seyfried N, Day GS, Wu Q, Fernández MV, Dominantly Inherited Alzheimer Network, Ibanez L, Sung YJ, Cruchaga C. Systematic proteomics in Autosomal dominant Alzheimer's disease reveals decades-early changes of CSF proteins in neuronal death, and immune pathways. medRxiv. 2024 Jan 13; PubMed.

Annotate

To make an annotation you must Login or Register.