Back

The diversity of disease-associated natural human antibody specificities declines dramatically in Alzheimers disease.

Reyes-Ruiz, J. M.; Zhang, J.; Netanel, J.; Yu, Z.; Glabe, C. G.

2026-07-24 neuroscience
10.64898/2026.07.21.739870 bioRxiv
Show abstract

Alzheimers disease (AD) is the leading cause of dementia and administration of monoclonal antibodies against A{beta} amyloid are associated with disease modifying effects leading to FDA approval for some antibodies. Naturally-occurring human antibodies (NAbs) are believed to be the front line of defense against a variety of diseases, so we investigated the specificity of NAbs in AD and cognitively normal (NC) individuals using epitomic profiling (EP). We found that the diversity of antibody-specific peptide epitope segments (ES) declines dramatically (26-fold) in the AD compared to NC populations. Many of these differentially expressed ES (DEES) map to amyloid sequences that accumulate in AD, such as A{beta}, tau, -synuclein, TDP43 and TMEM106b. NAbs that target amino terminal A{beta} epitopes are elevated in AD (residues 1-6 and 18-21) and NAbs that target different A{beta} epitopes are elevated in NC (residues 5-10 and 35-41). DEES that map to 2N4R tau are highly cross correlated in the population implying that that the antibodies may be polyspecific or that a set of antibodies are coordinately regulated. We used weighted gene correlation network analysis (WGCNA) to identify co-expressed DEES and UMAP to embed DEES in both AD and NC populations that bind to the same antibody or co-expressed set of antibodies. An average of approximately 25 DEES cluster in each WGCNA group. A major WGCNA cluster associated with AD contains 3 groups of overlapping DEES that assemble into longer linear epitopes. One of these 3 groups contain sequences that map to the amino terminus of A{beta}. DEES elevated in AD may be useful blood biomarkers for AD. The DEES clusters elevated in NC may be associated with protection against misfolded protein amyloid seeding and propagation and maintaining the resilience of neuronal populations.

Matching journals

The top 11 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.