Q&A Report: Vascular Contributions to Dementia

Diabetes is a major risk factor for Alzheimer’s disease, and there could be many reasons for the connection between these two diseases. As mentioned, hyperglycemia is associated with increased large artery stiffness. In particular, hyperglycemia increases advanced glycation end-products (AGEs) that cross-link collagen in the wall of arteries leading to increased stiffness. At the same time, hyperglycemia and AGEs also have direct negative impacts on the endothelial cells, neurons, and other brain cells leading to increased oxidative stress and inflammatory signaling.

We don’t know which one comes first but knowing this would support better design of preventions/treatments. Amyloid-beta causes cerebral endothelial cell dysfunction, and cerebrovascular dysfunction leads to increased amyloid-beta accumulation (due to limited blood flow, increased blood barrier permeability, poor removal of amyloid-beta from the brain, etc.). Thus, these two (amyloid-beta and endothelial dysfunction) create a vicious cycle. Philosophically, most treatments are designed to target one or the other, and perhaps targeting both at the same time would have the most benefit.

Hypercapnia is often used in human studies to determine cerebrovascular reactivity. There is some evidence that the actions of hypercapnia are via stimulated nitric oxide production, and thus, this is relevant to our ex vivo artery studies. An advantage of this technique is a catheter is not required (e.g., for drug injections), making this technique more reasonable for long-term repeated measures studies. Also, because this technique uses MRI, we can measure perfusion in deep brain regions that are not accessible with Doppler or microscopy. However, there are limitations of this protocol. In particular, not all brain regions are equally reactive to hypercapnia. In addition, the reaction to hypercapnia is not as physiologically relevant to dementia as the reaction to reaction to neuron stimulation (i.e., measures of neurovascular coupling).

Data from large studies in humans suggests that the appearance of arterial stiffening (about 50 years of age) occurs earlier than the onset of amyloid-beta accumulation (about 60-70 years of age). However, these data are limited by the sensitivity of the measurement techniques, and it possible that amyloid-beta is accumulating in the brain before it can be detected. We believe that increased arterial stiffness leads to cerebrovascular dysfunction and increased neuroinflammation, both of which trigger increased amyloid-beta production. At the same time, it is possible that amyloid-beta accumulation, particularly within the vasculature (e.g., cerebral amyloid angiopathy), increases the susceptibility to detrimental impact of stiffer large arteries. Likely, it is a combination/interaction of the vascular dysfunction and amyloid-beta (and other factors) that are cause late-onset Alzheimer’s disease, a multi-factorial disease.

With the models we used thus far, we do not find accumulation of amyloid-beta around the arteries. However, our studies were performed with the 3xTg-AD mouse, a model not known for vascular amyloid-beta accumulation. Perhaps we would find more cerebrovascular amyloid if we used a model with mutations more commonly associated with cerebral amyloid angiopathy (e.g., APP –Dutch) or a model where mutant proteins (APP or otherwise) are expressed in vascular cells.

Some of these studies were performed only in male mice (e.g., our original work with the Eln+/- mouse), while some use both males and females (e.g., the Eln+/- mouse crossed with the 3xTg-AD). For those studies performed only in male mice, we are working to repeat the studies in females. We do not have enough data yet to identify sex differences in the outcomes.

In a recent study from my lab, we found that female and male mice have a decline in cerebral artery endothelial function with aging. In that study, we related the endothelial dysfunction to frailty index, a marker of “biological age.” We found that cerebral artery endothelial dysfunction was related to a higher frailty index among old females, and this was independent of chronological age. However, there was no relation between cerebral artery endothelial function and frailty among old male mice. This suggests that cerebral artery dysfunction is more related to declines in general health among females compared with males. Here is a link to read more about this study: https://pubmed.ncbi.nlm.nih.gov/34664649/

We also have not examined the effects of estrogen or menopause/ovariectomy with the models of large artery stiffness or high pulse pressure. It is known that large artery stiffness increases post-menopause. There is also evidence that the age-related increase in large artery stiffness occur more rapidly in human females compared with males. The reason for this more rapid onset of arterial stiffness in females may be that estrogen is protective against stiffening in the pre-menopausal state, and the dramatic decline in estrogen with menopause leads to a rapid increase in large artery stiffness. Potentially, this rapid increase in arterial stiffness is more detrimental to the cerebral vasculature due to less time for adaptation compared with the slower increase in stiffness in males. However, this hypothesis remains to be experimentally tested. We recently wrote a mini review on this topic: https://pubmed.ncbi.nlm.nih.gov/35072153/

We assess pulse wave velocity in the aorta of mice. We start by recording the arrival time of the pulse at the aortic arch and the abdominal aorta. We collect the pulse wave forms at these two locations simultaneously using Doppler flow probes. We then calculate pulse wave velocity as the distance between the probes, divided by the difference in arrive time between the probes. A limitation to this measurement is that it must be performed on mice while under anesthesia. This article further describes the methods for measuring arterial stiffness in rodents: https://pubmed.ncbi.nlm.nih.gov/32268787/

We believe that arterial dysfunction is more likely to have a role in late-onset Alzheimer’s disease. Admittedly, our first studies were with the 3xTg-AD mouse, a model that uses mutations associated with early-onset Alzheimer’s disease. We chose this model initially because it allowed us to study the interaction of aberrant amyloid-beta and tau pathology with greater arterial stiffness. In the future, we are interested in increasing the translatability of these findings by using a model more similar to late-onset Alzheimer’s disease, such as the humanized APP mouse.

For C57BL6 mice, we classify “old” as 24-28 months and “young” as 4-8 months of age.

As far as I know, longevity has not been studied in the elastin heterozygote mouse model. For the data I presented, we studied these mice at a young age (4-8 months). Our goal was to model age-related increases in arterial stiffness in a young mouse, thus isolating the effects of arterial stiffness.

It seems likely that increased large artery stiffness and higher pulse pressure should be associated with more microinfarcts, but we have yet to examine this in our models. Given the limitations with measuring microinfarcts by imaging in humans, there is a lack of studies about these associations in humans as well. However, there are a number of studies that find a relation of greater large artery stiffness or higher pulse pressure with other indicators of cerebral small vessel disease, such as white matter hyperintensities. Here are a few examples:
https://pubmed.ncbi.nlm.nih.gov/29549223/
https://pubmed.ncbi.nlm.nih.gov/23172923/
https://pubmed.ncbi.nlm.nih.gov/22075523/
https://pubmed.ncbi.nlm.nih.gov/28389742/

Thus far, in our model of increased large artery stiffness (the Eln+/- mouse) we find endothelial dysfunction in cerebral arteries. However, I would expect that increased large artery stiffness would have a greater impact on the more fragile microvasculature compared with the cerebral arteries. We are performing experiments now to examine the effects on the microvasculature.

With aging, the major extracellular matrix changes to large arteries include increased collagen, increased collagen cross-linking, and fragmentation of elastin (for more information see https://pubmed.ncbi.nlm.nih.gov/34779281/). We use the elastin heterozygote mouse because it is a strong model of greater large artery stiffness with minimal other effects; however, we acknowledge that lower elastin content is not a primary cause of age-related arterial stiffening.

During our ex vivo pulse studies, the myogenic response of the arteries is still present. While we did not measure this myogenic response in these studies, previous work by Springo et. al. indicates that cerebral arteries from old mice have less myogenic tone after exposure to high pulse pressure compared with young mice (see https://pubmed.ncbi.nlm.nih.gov/25605292/). An impaired myogenic response in the old cerebral arteries will all the high pulse pressure to further penetrate the microvasculature, thus exacerbating the negative effects of high pulse pressure on the microvasculature.

We are currently testing interventions to prevent or reverse age-related increases in arterial stiffness. We recently studied pyridoxamine, a form of vitamin B6 that prevents the formation of advanced glycation end-products. We found long-term pyridoxamine treatment prevented age-related increases in large artery stiffness, preserved cerebral artery endothelial function, and had modest benefits for cognitive function (manuscript currently under review). Other potential pharmaceuticals include the collagen cross-link breaker alagebrium. Some of the strongest benefits for preventing age-related increases in arterial stiffness come from lifestyle interventions, such as aerobic exercise and dietary factors (see https://pubmed.ncbi.nlm.nih.gov/30212305/).

In our work, we are more focused on understanding the mechanisms and initiating events between these two. Studies from humans indicate that markers of both arterial stiffness and cerebrovascular function are related to cognitive impairment. While there are studies that suggest that arterial stiffness is a stronger marker for cognitive impairment, other studies suggest that arterial stiffness and endothelial function are equally strong.

While studies have found T-cells present in the brain with aging and Alzheimer’s disease, we have not measured T-cell content in our models of large artery stiffness. We do find increased inflammatory signaling in the cerebral arteries and brain, in particular elevated interleukin-1 beta. In addition, we find more activated microglia in our mouse model of large artery stiffness.

The cognitive impairments most related to arterial stiffness (and pulse pressure) are impairments in memory and learning. Greater large artery stiffness has been associated with impaired executive function and memory in human studies (a recent meta-analysis: https://pubmed.ncbi.nlm.nih.gov/32106748/) and memory in mice (https://pubmed.ncbi.nlm.nih.gov/31057061/).

We do not find that Eln+/- mice have systemic endothelial cell dysfunction compared with Eln+/+ mice. For example, we find that there is no endothelial dysfunction in skeletal muscle feed arteries of Eln+/- mice (https://pubmed.ncbi.nlm.nih.gov/25627876/). Others have found no dysfunction in the aorta or carotid arteries of Eln+/- mice (https://pubmed.ncbi.nlm.nih.gov/14597767/). The Eln+/- mouse was originally created as a model of supravalvular aortic stenosis, and thus, a limitation to these studies is that this model does have impaired heart valve function (https://pubmed.ncbi.nlm.nih.gov/9819363/).