Sex-specific modulation of amyloid-β on tau phosphorylation underlies faster tangle accumulation in females

Abstract

Females are disproportionately affected by dementia due to Alzheimer's disease. Despite a similar amyloid-β (Aβ) load, a higher load of neurofibrillary tangles (NFTs) is seen in females than males. Previous literature has proposed that Aβ and phosphorylated-tau (p-tau) synergism accelerates tau tangle formation, yet the effect of biological sex in this process has been overlooked. In this observational study, we examined longitudinal neuroimaging data from the TRIAD and ADNI cohorts from Canada and USA, respectively. We assessed 457 participants across the clinical spectrum of Alzheimer's disease. All participants underwent baseline multimodal imaging assessment, including MRI and PET, with radioligands targeting Aβ plaques and tau tangles, respectively. CSF data were also collected. Follow-up imaging assessments were conducted at 1- and 2-year intervals for the TRIAD cohort and 1-, 2- and 4-year intervals for the ADNI cohort. The upstream pathological events contributing to faster tau progression in females were investigated-specifically, whether the contribution of Aβ and p-tau synergism to accelerated tau tangle formation is modulated by biological sex. We hypothesized that cortical Aβ predisposes tau phosphorylation and tangle accumulation in a sex-specific manner. Findings revealed that Aβ-positive females presented higher CSF p-tau181 concentrations compared with Aβ-positive males in both the TRIAD (P = 0.04, Cohen's d = 0.51) and ADNI (P = 0.027, Cohen's d = 0.41) cohorts. In addition, Aβ-positive females presented faster NFT accumulation compared with their male counterparts (TRIAD: P = 0.026, Cohen's d = 0.52; ADNI: P = 0.049, Cohen's d = 1.14). Finally, the triple interaction between female sex, Aβ and CSF p-tau181 was revealed as a significant predictor of accelerated tau accumulation at the 2-year follow-up visit (Braak I: P = 0.0067, t = 2.81; Braak III: P = 0.017, t = 2.45; Braak IV: P = 0.002, t = 3.17; Braak V: P = 0.006, t = 2.88; Braak VI: P = 0.0049, t = 2.93). Overall, we report sex-specific modulation of cortical Aβ in tau phosphorylation, consequently facilitating faster NFT progression in female individuals over time. This presents important clinical implications and suggests that early intervention that targets Aβ plaques and tau phosphorylation may be a promising therapeutic strategy in females to prevent the further accumulation and spread of tau aggregates.

Keywords: Alzheimer's disease; amyloid-β; sex difference; tau neurofibrillary tangle progression; tau phsophorylation.

Figure 1

Cortical amyloid-β and tau aggregates strongly correlate with CSF phosphorylated tau₁₈₁ concentration. (A) Sex-disaggregated linear regression analyses were performed in both TRIAD and ADNI cohorts to elucidate the sex differences in the relationships between CSF phosphorylated tau-181 (p-tau₁₈₁) concentrations and regional amyloid-β (Aβ) and tau load. Findings revealed that both male and female subjects presented positive correlations between CSF p-tau₁₈₁ concentration and neocortical Aβ-PET standardized uptake value ratio (SUVR; TRIAD, males: P < 0.0001, R² = 0.43; females: P < 0.0001, R² = 0.52. ADNI, males: P < 0.0001, R² = 0.20; females: P < 0.0001, R² = 0.24). Furthermore, both male and female individuals displayed positive correlations between CSF p-tau₁₈₁ concentrations and regional tau-PET SUVRs. It is noteworthy that this positive correlation appeared to weaken or become non-significant in Braak V-VI regions of interest (ROIs) in male subjects (TRIAD, Braak V: P = 0.0017, R² = 0.16; Braak VI: P = 0.03, R² = 0.075. ADNI, Braak V-VI: not significant), while female subjects continued to demonstrate moderate positive correlations between CSF p-tau₁₈₁ concentrations and regional tau-PET SUVRs in subjects (TRIAD, Braak V: P < 0.0001, R² = 0.5; Braak VI: P < 0.0001, R² = 0.32. ADNI, Braak V-VI: P < 0.0001, R² = 0.29). (B) Sex-disaggregated voxel-based analyses were conducted within the TRIAD cohort to explore the differences between sexes in the associations between CSF p-tau₁₈₁ and cerebral Aβ and tau load at the voxel level. The findings revealed a noteworthy positive correlation between CSF p-tau₁₈₁ concentration and Aβ-PET signal in the temporoparietal cortices of males. Conversely, in females, this positive correlation was observed in multiple brain regions across the brain. When we investigated the relationships between CSF p-tau₁₈₁ and tau-PET data, males exhibited a positive connection between CSF p-tau₁₈₁ concentration and tau-PET signal primarily in the cingulum and temporal cortices. In contrast, females displayed a positive relationship between CSF p-tau₁₈₁ concentration and tau load throughout the entire brain. Images represent voxel-based t-statistical parametric maps overlaid on the structural MRI reference template. Age, APOE ε4 carriage status and pathological status were used as covariates in the model. Results were corrected for multiple comparisons using a false discovery rate (FDR) cluster threshold of P < 0.001. ADNI = Alzheimer’s Disease Neuroimaging Initiative; L = left; R = right; TRIAD = Translational Biomarkers in Aging and Dementia.

Figure 2

CSF phosphorylated-tau predicts tau accumulation in females. (A) Sex-disaggregated linear regression analyses were performed in both TRIAD and ADNI cohorts to examine the differences between sexes in the relationships between CSF phosphorylated tau-181 (p-tau₁₈₁) concentrations and the longitudinal accumulation of neurofibrillary tangles (NFTs), as indicated by changes in tau-PET standardized uptake value ratios (SUVRs). In the TRIAD cohort, two follow-up time points were evaluated. Females displayed positive correlations between baseline CSF p-tau₁₈₁ concentrations and NFT accumulation at both 1-year (P < 0.0001, R² = 0.58) and 2-year (P < 0.0001, R² = 0.56) follow-up visits. In contrast, males only exhibited positive associations at the 2-year follow-up assessment (P = 0.0004, R² = 0.25). In the ADNI cohort, three follow-up time points were assessed. Females demonstrated positive correlations between baseline CSF p-tau₁₈₁ concentrations and NFT accumulation at 2-year (P = 0.0004, R² = 0.21) and 4-year (P = 0.0004, R² = 0.34) follow-up visits. Conversely, males did not exhibit any such associations. (B) Sex-disaggregated voxel-based analyses demonstrated that the concentration of p-tau₁₈₁ in the CSF at the baseline was positively associated with longitudinal NFT accumulation in females. In contrast, males presented almost no association. Images represent voxel-based t-statistical parametric maps overlaid on the structural MRI reference template. Age, APOE ε4 carriage status and pathological status were used as covariates in the models. Results were also corrected for multiple comparisons using a false discovery rate (FDR) cluster threshold of P < 0.001. ADNI = Alzheimer’s Disease Neuroimaging Initiative; L = left; R = right; TRIAD = Translational Biomarkers in Aging and Dementia.

Figure 3

Tau phosphorylation predicts faster tau accumulation in Aβ+ females. (A) Aβ+ females presented higher CSF p-tau₁₈₁ concentrations as compared to Aβ+ males in both the TRIAD cohort (P = 0.04, Cohen's d = 0.51) and ADNI cohort (P = 0.027, Cohen's d = 0.41). (B) Aβ+ females also presented faster neurofibrillary tangle (NFT) accumulation compared with Aβ+ males [TRIAD cohort (1-year): P = 0.026, Cohen's d = 0.52; ADNI cohort (4-year): P = 0.049, Cohen's d = 1.14]. (C) In Aβ+ female subjects, baseline CSF p-tau₁₈₁ concentration was found to associate with the change in tau-PET meta-region of interest (ROI) standardized uptake value ratios (SUVRs) at 1-year (TRIAD: P = 0.05, R² = 0.3), 2-year (TRIAD: P < 0.0001, R² = 0.73; ADNI: P = 0.0025, R² = 0.39) and 4-year follow-up (ADNI: P = 0.0014, R² = 0.84) assessments. (D) Voxel-based linear regression models demonstrated positive correlations between CSF p-tau₁₈₁ concentration and baseline tau-PET SUVR as well as longitudinal NFT accumulation in Aβ+ females. The models were corrected for age and APOE ε4 carriage status, and for multiple comparisons using a false discovery rate (FDR) cluster threshold of P < 0.001. ADNI = Alzheimer's Disease Neuroimaging Initiative; L = left; R = right; TRIAD = Translational Biomarkers in Aging and Dementia.

Figure 4

Sex-specific modulation of cortical amyloid-β on tau phosphorylation underlies faster tau tangle accumulation in females. We proposed a novel model suggesting that males and females react differently to amyloid-β (Aβ) plaques, which consequently leads to differences in CSF phosphorylated tau (p-tau) production and causes the faster neurofibrillary tangle (NFT) accumulation observed in females. Overall, it supports the hypothesis that Aβ-dependent tau phosphorylation is an important driving force behind faster tau progression in females. Image was created with BioRender.com.