Amyloid-β as a Key Player in Cerebrovascular Dysfunction in Alzheimer's Disease

Alzheimer's disease (AD) is the most prevalent cause of dementia in the elderly. AD neuropathology is mainly characterized by the accumulation of intracellular aggregates of the protein tau and extracellular accumulation of the amyloid-β (Aβ) peptide, in the form of soluble oligomers and insoluble plaques. Cerebrovascular disorder (CVD) is a further hallmark of AD and causes decreased blood flow and perfusion, deterioration of the blood–brain barrier (BBB), and loss of the microvasculature (Zlokovic, 2011). Risk factors for AD and CVD substantially overlap and include diabetes and hypertension. The two-hit hypothesis for AD suggests that damage to the brain vessels (hit one) can lead to BBB dysfunction, hypoperfusion, decreased availability of energy substrates, and increased production and reduced clearance of Aβ—leading to Aβ accumulation in the brain (hit two), thereby rendering the brain milieu vulnerable to neuronal dysfunction and neurodegeneration (Zlokovic, 2011). In line with this hypothesis, changes in cerebral blood flow and local hypoxia/ischemia have been observed in early AD patients and in patients with mild cognitive impairment (MCI), who are at high risk for AD (Zlokovic, 2011). Moreover, hypoxia signaling increases the transcription and activity of β-secretase and γ-secretase, the two enzymes involved in Aβ production (reviewed in Salminen et al., 2017).

Notably, both Aβ and tau have been related to morphological and functional changes in the vasculature of AD patients and mouse models of AD. Aβ deposits within cerebral small vessels lead to cerebral amyloid angiopathy (CAA), a vascular feature of AD. Conversely, CAA may aggravate hypoxia and impair clearance of protein aggregates, further accelerating AD pathology (Zlokovic, 2011; Salminen et al., 2017). Aged mice overexpressing tau P301L (Tg4510), a tau variant associated with frontotemporal dementia, show abnormal blood vessel morphology, including reduced diameter and increased density, and reduced blood flow in the cortex (Bennett et al., 2018). Moreover, soluble tau has been shown to accumulate in the BBB of Tg4510 mice and AD patients - although reducing tau expression in aged Tg4510 mice did not reduce tau accumulation or alleviated vascular dysfunction (Bennett et al., 2020). Furthermore, positron emission tomography (PET) data from patients with AD or MCI revealed a negative correlation between cerebral blood flow and tau burden. The same study reported that Aβ positivity strengthened the correlations among tau, blood flow, and cognitive function, suggesting that tau and Aβ may synergistically impair vascular function during AD.

Recent transcriptomic data from brain vasculature-associated cells has offered further insight regarding cerebrovascular dysfunction in AD. Single-nucleus transcriptional analysis of brain vessel-associated cells from cognitively normal and AD patients revealed 30 previously known AD-risk genes enriched in the vasculature, in addition to loss of diversity of pericytes and vascular smooth muscle cells (mural cells) in AD brains. Extracellular matrix-maintaining pericytes (M-pericytes) appear to be selectively vulnerable, which may be related to loss of BBB integrity in AD (Yang et al., 2022). Transcriptional changes in vasculature-associated cells show region heterogeneity and are more strongly correlated with Aβ and CAA than tau pathology or APOE status, suggesting a more central role for amyloid pathology (Bryant et al., 2023). Interestingly, amyloid plaques and CAA are associated with distinct gene ontology terms, suggesting an interplay of diverse mechanisms (Bryant et al., 2023).

Microglia also play a key role in the interaction between Aβ pathology and CVD. Besides their role in neuroinflammation during both AD and CVD, microglia have an essential role in the fate of Aβ. Depletion of microglial cells in the 5xFAD mouse model of Aβ overexpression almost completely blocked amyloid plaque formation. Remarkably, the absence of microglia increased amyloid deposition in vessels, characteristic of CAA, indicating that the Aβ that escapes from microglia may be the source of CAA in AD (Spangenberg et al., 2019).

Because PET/MRI and histological analyses have limited ability to reveal hemodynamic changes in the brain, Lowerison et al. (2024) investigated the relationship between AD and microvascular changes using ultrasound localization microscopy (ULM) in vivo in 5xFAD mice. ULM involves tracking microbubbles of contrast agents using ultrasound imaging with super-resolution, and it provides vascular structural and functional information down to the capillary scale (Renaudin et al., 2022). It permits analysis of venous and arterial blood flow dynamics in deep tissue without losing resolution, providing whole-brain microvasculature reconstruction. Its higher resolution makes ULM a more suitable technique to evaluate structural and functional parameters of the microvasculature than contrast-enhanced ultrasound (CEUS) imaging and advanced Doppler ultrasound. It also allows in vivo analysis with little to no tissue damage, making it highly translatable to clinical and preclinical investigations of microvascular changes.

Lowerison et al. (2024) investigated vascular alterations in brain regions highly affected by AD progression, namely, the hippocampus, entorhinal cortex, and isocortex, in 5xFAD mice. They performed ultrasound imaging in 3- and 6-month-old mice to allow correlation of Aβ deposition across time from early AD to symptomatic AD. ULM imaging showed that younger 5xFAD mice exhibited hypoperfusion in some brain regions compared with age-matched control mice; this was aggravated in the hippocampus and entorhinal cortex in older mice. Analysis of cerebrovascular function revealed decreased velocity only in the entorhinal cortex, not isocortex, of 3-month-old mice. In contrast, 6-month-old 5xFAD mice showed a significant decrease in flow velocity in all investigated regions. Notably, morphological differences in vascularity and tortuosity were absent in 3-month-old 5xFAD mice, but there was decreased vascularity in the hippocampus and increased tortuosity in the isocortex and entorhinal cortex of the older mice. These data indicate that functional changes precede morphological changes.

By performing a ULM-to-histology correlation analysis in the isocortex, entorhinal cortex, and hippocampus, Lowerison et al. (2024) found negative correlations between Aβ and vascular density and flow velocity. They further reported that relevant microvascular parameters are altered in AD-related regions even in early-stage disease, including deep brain regions that were poorly explored before the advent of ULM. In addition, Aβ deposition, as assessed by immunohistology, was associated with both functional and structural microvascular changes, worsening as AD progresses. Comparing the parameters obtained with ULM in 3- and 6-month-old 5xFAD mice, it is apparent that microvascular dysfunction precedes structural changes and that this is more evident in the hippocampus and entorhinal cortex. Moreover, considering that brain regions with higher amyloid burden appear to be more strongly associated with general hypoperfusion, Aβ deposition seems to aggravate vascular dysfunctions and structure.

The interaction of microglia and Aβ may be a key step in the vascular alterations of AD and CAA pathology (Spangenberg et al., 2019). It is known that damage to the brain vasculature leads to hypoxia, hypoperfusion, BBB breakdown, and Aβ production. This excessive Aβ will build up in the brain parenchyma and trigger AD pathology along with microglial activation and phagocytic activity. Concomitantly, Aβ that escapes or overloads microglial clearance may reach brain vessels and directly interact with endothelial cells and pericytes, causing further cellular dysfunction and aggravating vascular damage, eventually causing morphological vascular alterations. This process also leads to increased risk for new CVD events, creating a feedforward cycle (Fig. 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Summary figure showing the feedforward loop of CVD and AD pathologies. Damage to the brain vasculature caused by CVD leads to BBB breakdown, hypoperfusion, hypoxia, and increased Aβ production. This initial event can cause functional changes in the vasculature, such as decreased flow velocity. The buildup of Aβ in the brain parenchyma triggers AD pathology, with the formation of soluble and insoluble Aβ aggregates and microglial reactivity. Aβ species that escape or overload microglial clearance reach brain vessels and directly interact with endothelial cells and pericytes, causing further cellular dysfunction, amyloid angiopathy, and vascular damage. The dysfunctional state will lead to morphological changes, such as decreased vascularity and increased tortuosity, and increase the risk for new CVD—creating a feedforward cycle. Figure created using BioRender.com.

Lowerison et al. (2024) leveraged the power of ULM to provide important information about how Aβ may be associated with cerebrovascular dysfunction in an in vivo model of AD. In line with previous literature (Yang et al., 2022; Bryant et al., 2023), these results support the association of Aβ pathology with vascular dysfunction and morphological changes in AD-affected brain regions, as proposed by the two-hit hypothesis (Zlokovic, 2011). Since ULM is readily adaptable to clinical settings, longitudinal studies using this technique to assess vascular function in deep brain regions of patients at risk for AD are warranted. Such studies could provide valuable data regarding cerebrovascular dysfunction in human patients, especially during early AD. Further investigations of the impact of AD pathology and risk factors on endothelial cells and cellular components of the BBB are promising approaches that can provide strategies to mitigate vascular dysfunction and break the damage cycle during AD.