Comparative analysis of brain-derived tau oligomer interactomes in Alzheimer's disease, non-demented with Alzheimer's neuropathology, and primary age-related tauopathy: Implications for neurodegeneration and cognitive resilience

Abstract

BackgroundIn Alzheimer's disease (AD), soluble tau oligomers are central to neurodegeneration and cognitive decline. Resilient individuals, such as those with non-demented Alzheimer's neuropathology (NDAN) or primary age-related tauopathy (PART), offer critical insights into protective mechanisms against tau-mediated neurodegeneration. NDAN individuals exhibit AD neuropathology without cognitive impairment or neurodegeneration, while PART, characterized by hippocampal- and entorhinal-restricted tau pathology, manifests with minimal-to-no amnestic changes. Brain-derived tau oligomers (BDTO) from these cohorts provide a unique platform to explore molecular pathways underlying both vulnerability and resilience to tau pathology.ObjectiveTo identify vulnerability- and resilience-associated pathways by comparing BDTO interactomes across AD, NDAN, and PART.MethodsBDTO were isolated from AD (n = 4; 2M, 2F), NDAN (n = 4; 2M, 2F), and PART (n = 4; 1M, 3F) hippocampal autopsy specimens using co-immunoprecipitation. Proteins were identified via liquid chromatography-tandem mass spectrometry, and non-specific interactors were filtered using SAINTq. Interactome networks and enrichment analyses were performed using Metascape. Findings were cross-referenced with the Neuropro database and existing literature on tangle-associated proteins. Key interactors were validated through reverse co-immunoprecipitation.ResultsA total of 203 proteins were identified, including eight novel interactors not previously linked to AD. All interactomes were enriched in proteins related to tau physiology and lysosomal degradation. NDAN and PART interactomes showed unique enrichment in proteins involved in cellular responses to reactive oxygen species.ConclusionsOne vulnerability-associated and 18 resilience-associated pathways that may mitigate tau-mediated neurodegeneration were identified, laying the groundwork for novel diagnostic and therapeutic strategies targeting pathological tau oligomers.

Keywords: Alzheimer's disease; cognitive dysfunction; hippocampus; tandem mass spectrometry; tauopathies.

Figure 1.

Comparative clinical and histopathological framework of AD, NDAN, and PART. (a) Clinical and histopathological features of Alzheimer's disease (AD), non-demented with Alzheimer's neuropathology (NDAN), and primary age-related tauopathy (PART), as classified by the established A/T/N (Amyloid/Tau/Neurodegeneration) biomarker framework. (b) Conceptual diagram illustrating the continuum of tau pathology extending from PART (intermediate resilience, mixed features) toward either AD (vulnerable/toxic) or NDAN (resilient/non-toxic) states, based on the presence or absence of amyloid-beta (Aβ). Progression along this continuum is associated with differential degrees of tau spreading, neurodegeneration, inflammation, and dementia.

Figure 2.

Experimental workflow schematic. Overview of sample selection, brain-derived tau oligomer (BDTO) isolation, and subsequent proteomic analysis, culminating in a multi-faceted, data-driven evaluation of disease-specific protein signatures. Created with BioRender.com.

Figure 3.

Sequential tau screening across stages of BDTO isolation. The experimental step and corresponding brain tissue fraction are specified above each panel. (a) PBS-soluble homogenates from AD, NDAN, and PART hippocampal specimens (n = 4 each) were screened for total tau (Ab: Tau5) via western blot analysis, with lanes labeled by sample ID as detailed in Table 1. (b) Subsequent analysis on co-immunoprecipitated BDTO fractions were performed using the TOMA2 antibody to probe tauO, focusing on the sample with the highest protein concentration (Supplemental Table 1). Total protein staining for the respective immunoblots are shown in Supplemental Figure 3.

Figure 4.

Specific proteins inform differential BDTO interactomes in AD, NDAN, and PART. (a) Venn diagram depicting the unique and shared proteins across BDTO interactomes, with AD (red), NDAN (blue), and PART (green), illustrating the proteomic diversity inherent to each condition. (b) Accompanying lists detail the specific proteins enumerated in the Venn diagram. All proteins were cross-referenced with the NeuroPro database and individual AP-MS experiments related to Alzheimer's disease. Proteins highlighted in yellow represent novel tau interactors not previously identified in any referenced sources.

Figure 5.

DisGeNET pathway analysis of BDTO-associated proteins reveals enrichment of relevant neurodegenerative disease phenotypes. Cluster heatmap generated from DisGeNET ontologies showcases enriched neurodegenerative disease pathways across AD, NDAN, and PART BDTO. Tau pathology-associated terms/conditions/pathways are indicated with a red asterisk. Selected terms meet defined significance thresholds (p-value < 0.01, minimum count ≥ 3, enrichment factor > 1.5) and are organized into clusters based on kappa similarity scores (> 0.3), with each row representing the most significant term of a cluster. The color-coded p-value scale at the top guides the interpretation of enrichment levels, with gray indicating non-significance. Created in Metascape.

Figure 6.

Identifying hubs within the fibrillar tau and BDTO interactomes through PPI network analysis. Protein-protein interaction (PPI) networks, derived from high-confidence physical interaction data, reveal functional protein networks within fibrillar tau from AD (a) and BDTOs from AD (b), NDAN (c), and PART (d). Each panel represents the PPI network filtered to include only physically interacting proteins, thereby enhancing the biological relevance, and reducing noise. The left side of each panel displays the full fibrillar tau or BDTO PPI (‘hairball’) network for each condition; the right illustrates sub-networks of densely connected proteins, indicative of functional modules, identified using the MCODE algorithm. Created in Metascape.

Figure 7.

Validation of BDTO interactome findings via reverse co-immunoprecipitation. (a) PBS-soluble fractions of hippocampal total homogenates from AD, NDAN, and PART specimens (n = 4 each) were analyzed for the presence of CCT7 and S100β via Western blotting. Lanes are labeled by sample ID. Bar graphs represent quantified protein density normalized to the total protein signal (Supplemental Figure 5), showing no significant differences in CCT7 (p = 0.14) and S100β (p = 0.37) levels between groups. Statistical analysis was performed using one-way ANOVA (α = 0.05) with Tukey HSD post-hoc test. (b) Co-IP of CCT7 and S100β was performed to validate their selective associations found with BDTO. S100β was co-immunoprecipitated from pooled AD brain homogenates, and CCT7 was co-immunoprecipitated from pooled NDAN and PART brain homogenates. Western blot analysis confirms successful co-IP of these proteins, as well as the presence of tau oligomers (Ab: T22) associated with both CCT7 and S100β across all three groups.