Differences in the cerebral amyloid angiopathy proteome in Alzheimer's disease and mild cognitive impairment.

Cerebral amyloid angiopathy (CAA) is characterized by amyloid beta (Aß) deposition in cerebrovasculature. It is prevalent with aging and Alzheimer's disease (AD), associated with intracerebral hemorrhage, and contributes to cognitive deficits. To better understand molecular mechanisms, CAA(+) and CAA(-) vessels were microdissected from paraffin-embedded autopsy temporal cortex of age-matched Control (n = 10), mild cognitive impairment (MCI; n = 4), and sporadic AD (n = 6) cases, followed by label-free quantitative mass spectrometry. 257 proteins were differentially abundant in CAA(+) vessels compared to neighboring CAA(-) vessels in MCI, and 289 in AD (p < 0.05, fold-change > 1.5). 84 proteins changed in the same direction in both groups, and many changed in the same direction among proteins significant in at least one group (p < 0.0001, R2 = 0.62). In CAA(+) vessels, proteins significantly increased in both AD and MCI were particularly associated with collagen-containing extracellular matrix, while proteins associated with ribonucleoprotein complex were significantly decreased in both AD and MCI. In neighboring CAA(-) vessels, 61 proteins were differentially abundant in MCI, and 112 in AD when compared to Control cases. Increased proteins in CAA(-) vessels were associated with extracellular matrix, external encapsulating structure, and collagen-containing extracellular matrix in MCI; collagen trimer in AD. Twenty two proteins were increased in CAA(-) vessels of both AD and MCI. Comparison of the CAA proteome with published amyloid-plaque proteomic datasets identified many proteins similarly enriched in CAA and plaques, as well as a protein subset hypothesized as preferentially enriched in CAA when compared to plaques. SEMA3G emerged as a CAA specific marker, validated immunohistochemically and with correlation to pathology levels (p < 0.0001; R2 = 0.90). Overall, the CAA(-) vessel proteomes indicated changes in vessel integrity in AD and MCI in the absence of Aß, and the CAA(+) vessel proteome was similar in MCI and AD, which was associated with vascular matrix reorganization, protein translation deficits, and blood brain barrier breakdown.

Comparison of the Amyloid Plaque Proteome in Down Syndrome, Early-Onset Alzheimer's Disease and Late-Onset Alzheimer's Disease.

Background: Down syndrome (DS) is strongly associated with Alzheimer's disease (AD), attributable to APP overexpression. DS exhibits Amyloid-ß (Aß) and Tau pathology similar to early-onset AD (EOAD) and late-onset AD (LOAD). The study aimed to evaluate the Aß plaque proteome of DS, EOAD and LOAD. Methods: Using unbiased localized proteomics, we analyzed amyloid plaques and adjacent plaque-devoid tissue ('non-plaque') from post-mortem paraffin-embedded tissues in four cohorts (n = 20/group): DS (59.8 ± 4.99 y/o), EOAD (63 ± 4.07 y/o), LOAD (82.1 ± 6.37 y/o) and controls (66.4 ± 13.04). We assessed functional associations using Gene Ontology (GO) enrichment and protein interaction networks. Results: We identified differentially abundant Aß plaque proteins vs. non-plaques (FDR < 5%, fold-change > 1.5) in DS (n = 132), EOAD (n = 192) and in LOAD (n = 128); there were 43 plaque-associated proteins shared between all groups. Positive correlations (p < 0.0001) were observed between plaque-associated proteins in DS and EOAD (R2 = 0.77), DS and LOAD (R2 = 0.73), and EOAD vs. LOAD (R2 = 0.67). Top Biological process (BP) GO terms (p < 0.0001) included lysosomal transport for DS, immune system regulation for EOAD, and lysosome organization for LOAD. Protein networks revealed a plaque enriched signature across all cohorts involving APP metabolism, immune response, and lysosomal functions. In DS, EOAD and LOAD non-plaque vs. control tissue, we identified 263, 269, and 301 differentially abundant proteins, including 65 altered non-plaque proteins across all cohorts. Differentially abundant non-plaque proteins in DS showed a significant (p < 0.0001) but weaker positive correlation with EOAD (R2 = 0.59) and LOAD (R2 = 0.33) compared to the stronger correlation between EOAD and LOAD (R2 = 0.79). The top BP GO term for all groups was chromatin remodeling (DS p = 0.0013, EOAD p = 5.79×10- 9, and LOAD p = 1.69×10- 10). Additional GO terms for DS included extracellular matrix (p = 0.0068), while EOAD and LOAD were associated with protein-DNA complexes and gene expression regulation (p < 0.0001). Conclusions: We found strong similarities among the Aß plaque proteomes in individuals with DS, EOAD and LOAD, and a robust association between the plaque proteomes and lysosomal and immune-related pathways. Further, non-plaque proteomes highlighted altered pathways related to chromatin structure and extracellular matrix (ECM), the latter particularly associated with DS. We identified novel Aß plaque proteins, which may serve as biomarkers or therapeutic targets.