Alzheimer's Disease as a Neurovascular Inflammatory Disorder

Highlights

The immune system and the microbiome, as well as the vasculature, play causal roles in Alzheimer's disease pathology.

Alzheimer's disease risk genes affect blood brain barrier breakdown.

The gut–brain axis in stroke and dementia

Brain research has long been dominated by a focus on neurons. Increasingly, however, researchers have begun to reassess that single-minded emphasis, branching out to explore how systems outside the brain sometimes wield enormous influence over it, said Costantino Iadecola of Weill Cornell Medicine. As a result, researchers now know that vascular pathology can amplify the negative effects of Alzheimer's disease pathology such as amyloid beta plaque formation, and epidemiological studies point to hypertension, obesity and diabetes as risk factors for dementia as well as for stroke.

But the immune system represents another major contributor to brain health and pathology, and Iadecola described his lab's work on how gut bacteria regulate immune cells in the brain. As a major source of immune cells, as well as of the neurotransmitter serotonin, the gut is also home to a vast community of some 40 trillion bacteria—the so-called gut microbiome—that play a key role.

Because the gut microbiome regulates T cells, one of the immune system's main defenses, Iadecola and his team decided to explore how it affects brain recovery from stroke. What they found was that mice in which antibiotics were used to disrupt the microbiome had more extensive brain injury after a stroke and worse recovery of motor performance than control mice. By transplanting the feces of mice with normal microbiomes into those with altered ones, however, the researchers could transfer the protective effects.

The researchers traced this effect to two types of T cells: T regulatory cells, which released a protective cytokine called IL-10, and Gamma Delta T cells, which released an inflammatory cytokine called IL-17. In the absence of these cell types, the microbiome's protective effects disappeared. Further studies suggested that the microbiome instructs immune cells to balance amounts of these T cells.

Stroke is an extreme form of damage that occurs when blood flow to the brain is cut, but subtler blood flow dysregulation occurring in conditions like hypertension and diabetes may also produce profound consequences. Because epidemiological studies correlate salt consumption in different countries with stroke and dementia incidence, Iadecola's lab studied the effect of salt on brain vasculature by feeding mice high-salt diets (8-16 times higher than normal) for two months.

They found that, while the diet caused no changes in blood pressure, it did affect how well endothelial cells, which line the blood vessels, did their jobs. It also hobbled the animals' performance on behavioral and cognitive tests. Molecular studies revealed a massive accumulation of IL-17 in the gut, followed by the movement of IL-17-producing cells to the brain. Mice genetically engineered not to produce IL-17 showed no cognitive effects from a high salt diet, suggesting that elevated levels of this cytokine impairs cognition.

At the same time, people who develop Alzheimer's disease start to lose weight about a year before clinical dementia sets in, suggesting some miscommunication between the gut and the brain about energy expenditure. Iadecola's team turned its attention towards this phenomenon, conducting studies in a mouse model of Alzheimer's disease that found that neurons in the animals' hypothalamus became insensitive to the hunger-regulating hormone leptin. The researchers saw a similar pattern in patients who don't yet show symptoms of Alzheimer's disease but have biomarkers suggesting they will develop the disease. Because leptin is neuroprotective, Iadecola said, it's possible that it's also playing a role in cognition and other aspects of the disease.

Cerebral vascular dysfunction in aging and Alzheimer's disease

A healthy brain needs healthy blood vessels, and Alzheimer's disease pathologies such as amyloid beta and tau accumulation cannot be studied outside the context of brain vasculature, said Berislav Zlokovic of the University of Southern California. Blood circulating through the brain's vasculature is separated from brain tissue by a membrane called the blood brain barrier (BBB). When the BBB breaks down, neuronal as well as vascular degeneration occurs. Endothelial cells in the BBB and blood vessels are normally chock-full of a protein called Glut1, which carries glucose into the brain. But Zlokovic and colleagues found that in mice with half the normal amount of Glut1, the capillaries begin to deteriorate after two weeks, and by 6 months neurons neural and behavioral deficits set it.

Alzheimer's disease patients also show a downregulation of Glut1, but it's not clear whether the protein is causing disease pathology or acting as an innocent bystander. Most postmortem studies of Alzheimer's disease tissue also show pronounced BBB breakdown. To study BBB breakdown in living humans, Zlokovic and his colleagues developed a magnetic resonance imaging (MRI) method that tracks vasculature in different brain areas. In normal aging, BBB breakdown occurs primarily in the hippocampus, but in patients with mild cognitive impairment it was also present in the cortex.

BBB breakdown was also more prevalent in patients who carry the E4 variant of the APOE gene, a strong genetic risk factor for Alzheimer's disease, as well as in mice engineered to carry APOE4. Vascular pathology is present in these mice by two weeks of age, and neurons begin to accumulate toxins but don't deteriorate functionally until a few months of age, Zlokovic said. He also described his work on contractile cells called pericytes, which wrap around the endothelial cells of small blood vessels, showing that the degeneration of these cells can cause Alzheimer's-like pathology in mice.

Ultimately, blood vessels may be converging points of several pathologies, Zlokovic said, and some recent experimental therapies for Alzheimer's disease may have failed because they did not protect vasculature. The group is currently testing a blood-vessel-protecting therapy for stroke in clinical trials, and some of these same drugs appear to be promising in animal models of Alzheimer's disease as well.

Highlights

Pericytes and smooth muscle cells have different roles in regulating blood flow in brain vasculature.

Endothelial cell activation may be a potent therapeutic target for Alzheimer's disease.

The glymphatic system, which clears waste products from the brain, may be dysregulated in Alzheimer's disease.

Mural cells and brain microvasculature

The brain is an energetically expensive organ—although it is just 2% of a person's body mass, it consumes 20% of the body's energy and blood flow. Moreover, the brain does not store energy, and it can't recruit new capillaries when more oxygen is needed. Consequently, many pathologies are likely related to microvascular disruptions, said Jaime Grutzendler of Yale University.

The brain uses a strategy called neurovsascular coupling—coordinating brain activation with blood flow—to supply specific brain areas with blood when they need it, and researchers have previously speculated that this process breaks down in aging and Alzheimer's disease. Grutzendler's lab uses a variety of imaging and genetic engineering techniques to study the cellular basis of neurovascular coupling in live animals.

In one instance, they investigated how the position and function of mural cells—which form two types of vascular cells, pericytes and smooth muscle cells—affects blood flow. Most mural cells sit on or extend along the vasculature, but about 20% of them form distinct rings around blood vessels. When the researchers probed the physiology of these ringed cells, they found that they carry markers of smooth muscle cells and have the ability to contract when activated. Their contractile properties depend on their location, however—those positioned closer to arteries contract strongly, while those in capillaries don't contract at all. There's a good reason for that, Grutzendler said, as capillaries are so narrow that contracting them can cause occlusions.

The other cell types, which cover 80% of the vasculature, are pericytes. Although they don't contract, they do undergo calcium changes upon activation. This suggests they participate in an as-yet unknown signaling mechanism, perhaps transmitting signals upstream to the smooth muscle cells. Grutzendler's team found a fluorescent marker that can reliably distinguish pericytes in capillaries from other types of mural cells. Mice carrying this marker can be cross-bred with genetic models of Alzheimer's disease to study the role of pericytes in disease progression, Grutzendler said. He also described a mechanism his lab discovered by which cells entrap blood clots that get lodged in the blood vessel and actively dislodge them, re-establishing blood flow, then force them out through the vessel wall over a period of 24–36 hours.

Brain microvasculature as a mediator of inflammation

The lion's share of funding for Alzheimer's disease research has been spent in the pursuit of understanding the role of amyloid beta, but this protein fragment clearly doesn't tell the whole story, said Paula Grammas of the George & Anne Ryan Institute for Neuroscience at the University of Rhode Island. Many of the same mechanisms that drive cardiovascular disease are now thought to play a role in Alzheimer's disease. To help unpack that, her lab studies how abnormal vascular function affects how neurons operate. Although her focus is on endothelial cells, in her talk Grammas stressed that multiple vascular mechanisms are undoubtedly at play.

Amyloid is expressed in both blood vessel cells and neuronal tissue in brain, but mouse strains in which the protein is concentrated in vessels show greater cognitive deficits than those in which it predominates in neurons. This suggests an important link between the vasculature and Alzheimer's disease pathology. Normally, endothelial cells, which make up the blood vessel wall, synthesize bioactive molecules that help nurture neurons, but Grammas' work suggests that endothelial cells in Alzheimer's disease are hyperactivated to overproduce inflammatory molecules.

Her lab developed a technique for isolating small, capillary-sized vessels to study their characteristics in tissue from Alzheimer's disease compared to control tissue. Normally, endothelial cells respond to injury by producing cytokines and other inflammatory proteins. Production of these factors is elevated in Alzheimer's disease tissue, they found. The overexpressed factors were all associated with the process of vascular formation, called angiogenesis, yet no new blood vessels are formed.

Grammas therefore reasoned that, normally, angiogenesis operates on a feedback loop in which the formation of the new vessels would act as a cue to shut off the production of these factors. In Alzheimer's disease, however, that shut-off step simply doesn't trigger. If that's the case, she hypothesized that shutting off the activation step might have therapeutic effects. Indeed, her team found that drugs that inhibit or minimize vascular activation improve cognitive deficits in mouse models of Alzheimer's disease, leading Grammas to conclude that vascular activation could be a potent therapeutic target for the disease.

Impairment of glymphatic function in aging and Alzheimer's disease

A clear link exists between Alzheimer's disease pathology and aging, and an age-related impairment in mechanisms that clear aggregating amyloid beta from brain cells may be one important breakdown that drives Alzheimer's disease pathology. Researchers have traditionally focused on how cells like astrocytes and microglia facilitate amyloid beta degradation, said Jeffrey Iliff of Oregon Health and Science University, but other mechanisms are increasingly recognized as playing a role.

Iliff's lab studies how the cerebral spinal fluid (CSF) and in the fluid existing between cells remove cellular waste, including aggregating amyloid beta. For decades, this process was thought to occur through passive diffusion in the ventricles and around the brain—then recently researchers discovered a brain-wide network associated with the blood vasculature through which CSF travels to help clear cellular waste.

A clear understanding of how this so-called glymphatic system affects brain inflammation and neurodegeneration is still evolving, but what is known is that proteins involved in Alzheimer's disease, such as amyloid beta and tau, are cleared through this pathway. Iliff's team found that the ability of CSF to recirculate along this system decreases with age in mice, and that this breakdown correlates with a decreased efficiency of amyloid beta clearance.

One key protein involved in this process is Aquaporin 4 (AQP4), which regulates water movement throughout the brain. In mice lacking AQP4, CSF moves much slower through the glymphatic system and amyloid clearance is impaired. In normal mice, both CSF recirculation and amyloid beta clearance were also impaired with age. Although the amount of AQP4 actually increases with age, in older animals it fails to localize to its correct position, perhaps decreasing the ability of this channel protein to direct the flow of water—and thus flush cellular waste—through this system.

But now, human data is also emerging. In post-mortem tissue from young, older normal, and older Alzheimer's disease people, Iliff and his colleagues found that AQP4 expression correlated with subjects' age, and the protein's localization correlated with Alzheimer's disease status. In another study, five mutations in the AQP4 gene that have previously been associated with other neurological conditions showed a correlation with cognitive decline.

Highlights

Carrying the Alzheimer's disease risk gene APOE4 and experiencing chronic inflammation weakened and shrunk the blood vessel network in the brain and hobbled cognition in mice.

Imaging venules—small veins—in the brain may provide clues about Alzheimer's disease pathology.

The blood-clotting factor fibrinogen promotes neuroinflammation and demyelination in the brain, and may provide a therapeutic target for neurodegenerative diseases.

Blood vessels may play an important role in amyloid clearance.

The interaction of peripheral inflammation and APOE4

Peripheral inflammation is a major risk factor for Alzheimer's disease, as is carrying the E4 variant of the APOE gene. Both these factors independently associate with cardiovascular disease and with cognitive dysfunction. Felecia M. Marottoli from Leon Tai's lab at the University of Illinois at Chicago used mouse models of Alzheimer's disease that carry APOE4 and APOE3 (an APOE gene variant that is not associated with Alzheimer's disease) to examine whether peripheral inflammation and APOE4 interact to induce cognitive and cerebrovascular deficits.

Chronic peripheral inflammation induced with weekly injections of lipopolysaccharide caused cognitive dysfunction in the APOE4 mice and increased cardiovascular leakiness in their brains. The amount of the brain that blood vessels reached in these mice was also reduced by 25% to 30%, and levels of amyloid beta in their hippocampus was increased. The researchers theorized that cytokine levels were responsible for this difference, but plasma cytokine levels were elevated in both APOE3 and APO4 mice receiving LPS injections, so the mechanism mediating the cognitive and cardiovascular effects are not yet clear. Marottoli and her colleagues believe that in the APOE4 mice, amyloid beta somehow predisposes endothelial cells to damage by peripheral inflammation, which sets off a cycle of further damage.

In vivo imaging of veins in microvasculature

Researchers studying the vascular dimension of Alzheimer's disease pathology have focused on arteries, but have largely overlooked veins, said C. Elizabeth Shaaban from the University of Pittsburgh. One major reason is that traditional neuroimaging methods do not directly image blood vessels but rely on markers that typically ignore the venous side of circulation. However, several characteristics of small veins called venules are important in Alzheimer's disease pathology as well as in a cardiovascular condition called Small Vessel Disease.

Shaaban and her colleagues explored the possibility of using a specialized form of magnetic resonance imaging, called 7T susceptibility weighted MRI, to visualize venous circulation. They used the approach in a study of 53 cognitively normal older adults. The researchers defined a measure they termed the "tortuosity ratio"—essentially the degree to which venules are kinked rather than straight—as an in vivo marker of cerebrovascular integrity in older adults. They found that people who had the APOE4 allele had a higher tortuousity ratio. The results support the link between APOE4 and reduced vascular integrity, Shaaban said. It's not yet clear, though, how tortuosity relates to pathological processes in Alzheimer's disease and Small Vessel Disease.

Fibrinogen as a mediator of neurovascular interactions

It is increasingly clear that neurons and glia cannot be studied in isolation from their extracellular environment, said Katerina Akassoglou of the University of California, San Francisco. Her laboratory studies one such factor, the blood coagulating protein fibrinogen, and how it operates as a causal agent in the pathology of several brain diseases.

Fibrinogen appears to be deposited in human blood vessels and brain tissue, and an overabundance is used as a biomarker in plasma and cerebrospinal fluid for mild cognitive impairment. It also correlates with amyloid plaque deposition and with the activation of central nervous system immune cells like microglia—two key hallmarks of Alzheimer's disease pathology. Fibrinogen thus plays several distinct roles: promoting coagulation, binding to amyloid beta and inhibiting its removal, and promoting inflammation. Depleting it is neuroprotective in animal models of Alzheimer's disease, as well as other neurological conditions like peripheral nerve injury, brain trauma, and multiple sclerosis.

Akassoglou's lab found that fibrinogen activates microglial cells, which are in charge of immune response in the brain. Microglia are shaped like amoebas, with small cell bodies and long tentacle-like processes that scavenge for cell detritus and infectious agents. These cells are positioned around brain blood vessels, constantly extending and retracting their processes around blood vessel walls to conduct surveillance of the vasculature. When fibrinogen is injected into the brain, her team found, these cells surround the injection site within minutes. Microglial activation is sustained, and after seven days is accompanied by demyelination that appears to be driven by the production of reactive oxygen species. The researchers have primarily examined this pathway in models of multiple sclerosis, but have also found evidence of it in animal models of Alzheimer's disease.

The group is now exploring whether fibrinogen can be targeted therapeutically, looking for ways to selectively target the molecule's inflammatory functions while leaving its coagulation properties intact. As a first step, the team developed a monoclonal antibody that specifically and selectively binds to fibrin, a molecule that fibrinogen converts to. This antibody potentially inhibits the expression of inflammatory genes—in studies in live mice, the treatment prevents neurodegeneration without interfering with blood coagulation.

Anti-amyloid immunotherapy

Immunotherapy is a major therapeutic approach in Alzheimer's disease, said Steven Greenberg of Massachusetts General Hospital and Harvard Medical School, whose presentation surveyed recent advancements on that front. The monoclonal antibody aducanumab is the most advanced Alzheimer's drug in clinical trials—in September, its maker, Biogen, reported promising results in a phase 2 trial—and several other immunotherapy agents are also under investigation.

The most common adverse effect of immunotherapy for Alzheimer's disease is so-called amyloid related imaging abnormalities, or ARIAs, Greenberg said. ARIA associated with clinical trials, which produce similar effects, can look indistinguishable from spontaneous inflammation induced by cerebral amyloid angiopathy (CAA), a condition in which amyloid is deposited on arteries in the brain.

The key criterion for probable CAA is the presence of focal or asymmetrical white matter lesions extending to just below the white matter. These look like autoimmune responses to amyloid in the vessel, Greenberg said. Both ARIA and CAA-induced inflammation seems to cause vascular injury, possibly as amyloid is delivered to or removed from blood vessels.

Greenberg also described data from an immunotherapy trial for CAA funded by Pfizer. The drug, ponezumab, was initially developed as an anti-amyloid antibody. But because it targeted a form of amyloid that tends to predominate in blood vessels, and because it was designed to have low immunogenicity (and thus was thought to be unlikely to cause ARIAs), it was tested in CAA. The trial used vascular reactivity—changes in blood flow levels through a vessel—as an imaging biomarker, and in the course of it, subjects received three monthly doses of either a placebo or the drug. Although ponezumab was shown to be safe, no evidence of efficacy emerged. However, because the drug did decrease vascular reactivity, the trial did seem to suggest that the blood vessels play a key role in amyloid clearance, and that vascular reactivity may be an important measure to explore in future studies.

Highlights

The instability of plaques in the carotid artery is directly linked to cognitive decline.

Basic research on the role of cardiovascular, immunological and other non-neurological factors in Alzheimer's disease must be translated into the clinic.

Linking atherosclerosis to cognitive decline

After 35 years as a cardiovascular surgeon, Robert J. Dempsey of the University of Wisconsin feels confident that what stroke victims have to fear most is not paralysis but dementia. For every clinical stroke diagnosed, 5 "silent" ones go unnoticed, and these microvascular insults have massive cognitive effects, he said, though the cognitive component of cardiovascular disease is poorly understood. Alzheimer's disease and cardiovascular disease have identical risk factors and prevalence. The loss of decision-making capacity and creativity is symptomatic of cardiovascular disease, he said, and Alzheimer's-associated memory loss follows about a decade later.

Atherosclerosis in the brain is an active process driven by local gene regulation and microvascular changes, including new formation of microvessels. Dempsey and his colleagues hypothesized that what made plaques pathogenic was whether they were stable or apt to break apart, prompting them to develop a way of using ultrasound to measure plaque instability in the carotid artery. To their surprise, they found that it directly predicts memory defects and other cognitive problems in people even when they have no cardiovascular symptoms. Cognitive decline stalled completely in patients who underwent surgery to remove plaques, and even improved in some parameters.

Plaque instability also correlated with small vessel activation and formation, as well as evidence of subclinical vascular lesions in the brain's white matter, prompting Dempsey to investigate regulators of microvasculature. They started by looking at inflammatory proteins called adipokines, which are released by fat cells and which promote atherosclerosis and stroke. In a genetic screen that looked for proteins whose expression changes in response to stroke in an animal model, Dempsey's team narrowed in on two adipokines, lipocalin-2 and DDPIV.

The presence of lipocalin-2 decreased the viability of neurons grown in a dish, and its levels increased with age and after stroke in rats, suggesting that its regulation may be involved in repair of damaged brain tissue. It was also increased in patients with atherosclerosis on the verge of experiencing strokes. Changing levels of DDPIV, meanwhile, correlated with plaque deposition in blood vessels. Dempsey's lab is exploring both molecules as potential biomarkers or therapeutic targets.

Panel discussion: next steps to new Research and therapeutic opportunities

Zorina Galis of the National Institutes of Health and the National Heart, Lung, and Blood Institute summed up the day's discussion by likening the pathology of Alzheimer's disease to a crime scene, where researchers observe a slew of dead brain cells and must identify suspects. With the field increasingly open to exploring the effect of non-neurological factors such as cardiovascular disease, researchers must look for truly causal mechanisms rather than just correlations, she said. One way to probe for connections is to examine existing animal models of cardiovascular disease for brain effects. Techniques for measuring molecular functions at the single cell level will also move the field forward by allowing researchers to map the spatial and temporal dimensions of the vasculature in normal and pathological states.

Iadecola noted that ongoing clinical trials lag behind the current thinking about Alzheimer's as a multifactorial disease, but that the field is positioned to catch up with new therapies. Throughout the 1960s and 1970s, researchers believed in a vascular cause for dementia and treated patients with vasodilators. Amyloid became the focus after its discovery in the familial form of Alzheimer's disease. All the panelists agreed that there is now a worldwide consensus among researchers studying Alzheimer's disease and related disorders as well as funding agencies supporting such work, that vascular, immunological and other mechanisms play a central role in the disease. Iadecola said that amyloid is still "on the map" mechanistically and as a therapeutic target, but other approaches will soon become translationally viable.

Dempsey discussed the need to think carefully about translating basic research into effective clinical interventions. There is currently such a rich range of biological pathways to pursue that researchers must carefully define the pathways they are working on, be vigilant in recognizing artifacts in animal models, and identify potent biomarkers that can confirm links between preclinical research and human biology. Other panelists and attendees also underscored the importance of studying gender differences in disease mechanisms as well as the timing of the disease, and correspondingly, the timing of when therapies must be administered to be effective. In response to one attendee's question about the role of lifestyle in the disease, Dempsey said that "what your mother told you" is basically right: keeping cardiovascular and other risk factors for Alzheimer's disease and dementia in check involves a dose of genetic luck, moderation in eating, and regular exercise like walking.

The panelists noted that studying the genetics of resilience to Alzheimer's and other neurodegenerative diseases—for example, in people who have mutations in risk genes such as Presinilin yet don't develop the disease—could yield important clues. Much basic research must also be done to understand the basic physiology of the vasculature, in the brain as wells other organ systems. Microvasculature is tough to study because the vessels are very small and their distribution around organs is complex. Another difficulty is knowledge fragmentation: researchers studying vascular dysfunction in the brain rarely exchange knowledge with those studying it in the kidney, yet many commonalities exist, said Galis.

Overall, the panelists said, the explosion of research on the role of cardiovascular and other non-neurological mechanisms in Alzheimer's disease and other dementias, as well as strong interest from the research community, funders, and the public in tackling these conditions, gave them hope that new therapies lie on the horizon.

Can the microbiome be harnessed therapeutically to promote neurovascular health?

How does blood brain barrier integrity affect vascular and neuronal pathology?

How do vascular cells that control blood flow contribute to Alzheimer's disease development?

How does the glymphatic system breakdown affect cognition?

What mediates the interaction between peripheral inflammation and genetic risk for Alzheimer's disease?

What is the role of venules in the cardiovascular side of Alzheimer's disease pathology?

How does fibrinogen activate microglia?

What role does the vasculature play in clearing amyloid, and can it be harnessed by immunotherapy?

What is the mechanism by which circulating plaques cause cognitive decline?