Plasma tau biomarkers are distinctly associated with tau tangles and decreased with Lewy body pathology

Abstract

Introduction: Plasma-to-autopsy studies are essential to understand how tau blood biomarkers change in relation to Alzheimer's disease (AD) brain pathology and how they are influenced by brain co-pathologies and comorbidities.

Methods: Plasma samples from 102 brain donors of the Vallecas Alzheimer Reina Sofia cohort were analyzed using a mass spectrometry method to measure the levels of six phosphorylated and two non-phosphorylated tau biomarkers.

Results: In cases with high pathological burden of AD, phosphorylated tau (p-tau)217 showed associations with neurofibrillary tangle counts across all regions, while p-tau205 was primarily linked to the frontal cortex, and the non-phosphorylated tau peptides were linked to the temporal cortex. Plasma tau levels decreased as Lewy body pathology progressed, even among individuals at the same tau Braak stage. All plasma biomarkers correlated with creatinine levels, as a marker of renal dysfunction, but this effect was mitigated when using the ratios phospho/non-phospho.

Discussion: Understanding how plasma tau biomarkers are influenced by co-pathologies and comorbidities is crucial for their accurate implementation.

Highlights: The plasma phosphorylated tau (p-tau)217 ratio, followed by the p-tau205 ratio and p-tau217, were the best-performing biomarkers for detecting neuropathologically confirmed Alzheimer's disease (AD). In high AD cases, p-tau217 showed associations with neurofibrillary tangle (NFT) counts across all regions, while p-tau205 was primarily linked to the frontal cortex, and non-phosphorylated tau peptides were linked to the temporal cortex. We observed a decline in plasma tau levels as Lewy body pathology progressed, even among individuals at the same tau Braak stage. All plasma biomarkers correlated with creatinine levels, as a marker of renal dysfunction, but this effect was mitigated when using the ratios. Plasma p-tau199 displayed a distinct behavior compared to other p-tau species, showing no association with NFT counts in any brain region, but demonstrating significant correlations with brain volume and an effect on survival time.

Keywords: blood; mass spectrometry; neuropathological examination; tau phosphorylation.

FIGURE 1

Tau and phosphorylated tau species with higher performance differentiate patients according to their neuropathological assessment. Unpaired two‐sample Wilcoxon testing differences of tau variants’ concentrations by clinical diagnosis (left side) and neuropathological post mortem assessment (right side). AD, Alzheimer's disease; p‐tau, phosphorylated tau. Distribution of p‐tau181 according to the clinical diagnosis (A) or to the post‐mortem neuropathological assessment (B). Distribution of p‐tau217 according to the clinical diagnosis (C) or to the post‐mortem neuropathological assessment (D). Distribution of p‐tau205 according to the clinical diagnosis (E) or to the post‐mortem neuropathological assessment (F). Distribution of p‐tau199 according to the clinical diagnosis (G) or the post‐mortem neuropathological assessment (H). Distribution of p‐tau231 according to the clinical diagnosis (I) or to the post‐mortem neuropathological assessment (J). Distribution of the p‐tau217 ratio according to the clinical diagnosis (K) or to the post‐mortem neuropathological assessment (L). Distribution of the p‐tau205 ratio according to the clinical diagnosis (M) or to the post‐mortem neuropathological assessment (N). Distribution of the p‐tau199 ratio according to the clinical diagnosis (O) or to the post‐mortem neuropathological assessment (P).

FIGURE 2

Specific association between tau species in plasma to tau pathology assessed in post mortem tissue. A complete sample was analyzed using partial Spearman ρ correlation to test the association between each tau variant in plasma and the post mortem number of neurofibrillary tangles in three cortical regions, plus the sum of the three areas. The ρ coefficients were estimated after adjustment by amyloid total area and for covariates (age at plasma collection, sex, age at onset). The percentual values (%) correspond to the percentage of the specific partial Spearman ρ;, it has been calculated with the formula: 100*[partial ρ / (partial ρ_NFTs+ partial ρ_Aβ)]. +p< 0.10; *p< 0.05; ** p< 0.01; *** p< 0.001. p‐tau, phosphorylated tau.

FIGURE 3

Tau variants distribution according to combined AD and Lewy body pathologies. Mixed pathology groups were established based on Braak stages and Lewy pathology consensus criteria (LPC score); AD‐Asyn group was composed of cases with AD plus non‐α‐synuclein deposits (LPC = 0) or deposits in the olfactory bulb (LPC = 1); AD+Asyn‐Amy group comprised of cases with AD plus α‐synuclein deposits in the amygdala (LPC = 2); AD+Asyn‐Lim/Neo involving cases with AD plus α‐synuclein deposits in limbic and neocortical regions (LPC = 4&5). LPC 0: no deposition (n = 45. 45.9%); LPC 1: deposition in the olfactory bulb (n = 11. 11.2%); LPC 2: deposition in the amygdala (n = 18. 18.4%); LPC3: deposition in substantia nigra (n = 1. 1.0%); LPC 4:. deposition in the cingulate cortex (n = 10. 10.2%); LPC 5: deposition in the frontal cortex (n = 13. 13.3%). Due to our sample only including one subject in the LPC = 3 score, it was excluded from the analysis. AD, Alzheimer's disease; p‐tau, phosphorylated tau.

FIGURE 4

Effect of plasma tau species on the patients’ survival by neuropathological groups. Multiple linear regression was conducted with days elapsed between the blood extraction and death date as the dependent variable, and the standardized score of tau species as a predictor variable in independent analysis. The covariates included were disease progression (years elapsed between age at onset and age at blood extraction), age at blood extraction, and sex. Regression coefficients ordered by their magnitude in the AD‐predominant group. Significant B coefficients and their respective model‐adjusted R ² are shown. AD, Alzheimer's disease; p‐tau, phosphorylated tau.