What Happens in the Brain to Cause the Transformation From Healthy Aging to AD?

What Happens in the Brain to Cause the Transformation From Healthy Aging to AD?
As people age, changes occur in all parts of the body, including the brain:

• Some neurons shrink, especially large ones in areas important to learning, memory, planning, and other complex mental activities. This translates into some shrinkage of brain volume over the course of years, even in healthy older people (Resnick et al., 2003).
• Tangles develop inside neurons and plaques develop in the spaces between neurons.
• Damage by free radicals increases (free radicals, also called reactive oxygen species, are a kind of molecule that reacts easily with other molecules).

The impact of these changes differs among people as they age. Some older people may notice only a modest reduction in their ability to learn new things, retrieve information from memory, and plan and make decisions. Their performance on complex tasks of attention, learning, and memory may decline. However, if given enough time, they may ultimately score as well on the task as a younger person. Other people, however, experience much greater declines in these cognitive abilities as they grow older. Understanding the differences between healthy aging and a neurodegenerative process is an important key to unlocking the secrets of AD.

Long ago, before we knew much about AD, many people thought that "senile dementia" was just a part of aging. Now, of course, we know that AD is a distinct disease that affects the brain. Several recent studies and reviews of the scientific literature have provided some evidence that cognitive decline with age and AD are, in some respects, separate entities that follow different pathways as they evolve. For example, an investigator from Washington University in St. Louis recently published a review of a vast array of data on memory and executive function (the cognitive abilities involved in planning, organizing, and decision-making) in aging and AD (Buckner, 2004). The review suggested that factors that influence executive function more commonly falter with normal aging. In contrast, factors that influence long-term memory function are more impaired in AD. Some of the changes that can take place in normal aging may not necessarily be the cause of AD. In a more direct assessment of this idea, this investigator and colleagues at Washington University used MRI to measure volumes of the hippocampal region and the white matter region between the two hemispheres of the brain in 150 people aged 18 to 93 years (Head et al., 2005). They found that early-stage AD did not make age-associated reductions in the white matter region worse. They also found that early-stage AD was characterized by significant reductions in hippocampal volume, whereas age alone was associated with only mild reductions in hippocampal volume. These results suggest that AD manifests itself early and significantly in the area of the brain encompassing the hippocampus, whereas normal aging affects the white matter connecting the front regions of the brain. These frontal white matter reductions may underlie the executive function difficulties that are common in normal aging.

Columbia University scientists funded by the National Center for Research Resources (NCRR) and NIA, in collaboration with researchers at the California National Primate Research Center, conducted neuroimaging studies in older monkeys and rats to distinguish the effects of healthy aging and AD on the structure of the brain (Small et al., 2004). The selection of the animals for this study was important because neither monkeys nor rats naturally develop AD. They found that the brain region known as the dentate gyrus (a region within the hippocampus) was changed by aging in both species. However, AD predominantly affected a different region of the hippocampus, thus indicating that some processes affecting normal aging and AD may be distinct.

Back to Top

New Thinking About Cognitive Reserve
Two questions have fascinated investigators for years:

• Why do some people remain cognitively healthy all their lives while others develop dementia?
• Why do some people remain cognitively healthy even though examination of their brain tissue after death shows significant deposits of plaques and tangles?

One possible explanation revolves around the concept of "cognitive reserve." Reserve refers to the brain's ability to operate effectively even when function is disrupted and to the amount of damage that the brain can sustain before the damage is clinically apparent. Individual variability in reserve may reflect genetic differences or differences in life experiences, such as education, occupational experience, or leisure activities.

Lifelong engagement in activities that help to build and maintain cognitive reserve is generally beneficial to health and may even help keep people cognitively healthy as they age. A number of research teams are looking intensively at these activities. For example, new information from the Religious Orders Study, a long-term study of aging among members of 40 religious communities, has revealed that years of formal education may modify the relationship between level of cognitive functioning and AD pathology (Bennett et al., 2003). After comparing participants' brain tissue after their deaths and the results of earlier cognitive function tests, study investigators at the Rush University Medical Center ADC in Chicago, found that the density of plaques in the brain tissue was linked to cognitive function. This relationship was modified by the number of years of education the person had received, such that a person with a higher level of education retained a higher level of cognitive function even in the presence of AD damage to the brain.

In another study, scientists from Columbia University periodically tested the memory performance of 136 cognitively healthy older people over the course of 5 years (Manly et al., 2003). The scientists accounted for the participants' age at the beginning of the study and their years of education. They found that participants with low levels of literacy (a proxy for quality of education) had a steeper decline in their ability to remember a word list immediately after seeing it, as well as after a delay, as compared to those with a higher literacy level. These findings suggest that literacy may be an important measure of cognitive reserve, or even that literacy itself builds cognitive reserve, protecting against memory decline in older people without dementia.

Rush ADC investigators from the Chicago Health and Aging Project (CHAP), an epidemiologic study of AD risk factors in a racially diverse urban population, looked at the issue of cognitive reserve from a slightly different angle (Barnes et al., 2004). They wondered whether an important life experience, such as the level of engagement in social networks of family and friends, could be related to changes in cognitive function. The 6,102 African-Americans and whites in the study participated in up to three interviews over the course of about 5 years. The researchers found that more social networks and a higher level of social engagement were associated with a higher initial level of cognitive function. These factors also were related to a reduced rate of cognitive decline over time. Additional research is clearly necessary to sort out cause and effect. Does increased involvement with social networks increase cognitive reserve, or are people who later develop dementia less involved because they are already in the early stages of AD, before symptoms are evident?

Findings from recent studies suggest that the brain has an inherent ability to cope with age- and disease-associated changes.

It may attempt to compensate for these changes through various mechanisms, such as the use of alternate brain networks to bypass those that are not functioning or the use of new cognitive strategies (Buckner, 2004). A number of scientists are exploring these compensatory mechanisms. For example, researchers at Columbia University examined whether engaging in various intellectual, social, and physical activities, such as gardening, reading, traveling, and going to the movies, might enhance cognitive reserve (Scarmeas et al., 2003). Cognitively healthy older people and people with early AD were scanned for blood flow in the brain using PET. The researchers found that study participants with AD who had a higher leisure activity score also had prominent deficits in cerebral blood flow, despite being at the same level of clinical disease. Those participants who had engaged in more lifestyle activities before disease onset were able to sustain more pathology, as shown by greater cerebral blood flow deficits.

Another approach to understanding compensatory mechanisms comes from Down syndrome (DS) and AD PET scan studies. National Institute of Child Health and Human Development (NICHD)-funded researchers at the University of California at Irvine tested a hypothesis that dementia starts in the same places in the brain in both DS and AD and progresses through similar patterns of change. Data from the scans, which track glucose use by cells in particular brain regions, allowed the researchers to measure and compare brain activity in adults with DS who did not have dementia, adults with AD, and cognitively healthy adults (Haier et al., 2003). The researchers found that, compared to the healthy adults, both those with DS and those with AD exhibited lower brain activity in one particular region of the brain (the posterior cingulate). However, when the researchers looked at another area of the brain-the inferior temporal/entorhinal cortex (the area known to be the site of the earliest brain damage in AD)-the participants with DS had higher brain activity while those with AD had lower activity compared to the healthy participants. The researchers believe that those with DS were in the very early stages of developing dementia and that their elevated brain activity levels represent an attempt by damaged neurons to work harder to maintain function. They also suggest that this compensatory response may eventually fail and the brain activity rates decrease, leading to degeneration of the neurons and the first clinical signs of dementia.

Understanding that AD is a process that develops over many years and is the result of many factors creates opportunities for early interventions that may prevent or delay the onset of the disease. It may be that because of the physical properties of a person's brain and his or her genetic makeup and life experiences, the person is able to tolerate and adapt to a certain amount of change and damage that occurs to the brain during aging. This tolerance level differs from person to person depending on cognitive reserve and other factors. At some point in the life of some people, the balance may tip in favor of a disease process. For others, the balance may remain in favor of healthy aging. Learning about the earliest developments in the disease process will help researchers understand this complex, lifelong balancing act.

Back to Top

Mild Cognitive Impairment
As some people grow older, they develop memory problems greater than those expected for their age. However, these problems do not necessarily meet all the accepted criteria for AD. These people have a condition called mild cognitive impairment.

Because many, but not all, people with MCI progress to AD, scientists debate whether MCI is an early stage of AD, an entirely distinct condition, or a multifaceted condition in which AD is one of several potential causes. To help settle this question, scientists hope to learn more about the underlying causes and courses of MCI. They have defined subtypes of MCI based on cause (for example, degenerative, vascular, psychiatric [especially depression], and medical), and on which aspects of cognition are predominantly affected.

The subtype that features memory impairment most prominently is called MCI with memory loss, or "amnestic MCI," and is the subtype likely to lead to AD. Individuals with other MCI subtypes may have prominent deficits in other cognitive functions, such as language skills or visuospatial ability. These types of dementia can be caused by other degenerative diseases, such as frontotemporal dementia and dementia with Lewy bodies, or by other conditions such as vascular dementia (Petersen, 2005a).

Being able to describe the essential characteristics of MCI is another important step in understanding this condition. Investigators at the Alzheimer's Disease Cooperative Study (ADCS; see p. 57 for more information) have done just that in a recent study (Grundman et al., 2004). The research team compared participants with amnestic MCI to individuals with AD and determined that the people with this type of MCI had impaired memory but that other elements of cognitive function were not affected. The people with amnestic MCI achieved test scores between healthy people and those with AD on cognitive and functional tests. Of the 214 participants with this type of MCI who progressed to dementia, 212 were classified as having AD after 3 years of follow-up (Petersen et al., 2005b). By so clearly describing this population in terms of the elements and extent of cognitive function affected by amnestic MCI, the study team has made a major contribution to a vital component of AD research-clinical trials. Their definition of amnestic MCI has already been widely used by other research teams working on AD treatment strategies.

	Charting the Course of Healthy Aging, MCI, and AD
	Click Here to Enlarge

The Journey from Healthy Aging to AD: Damage in the Brain
Since the earliest days of AD research, investigators have focused on the basic disease process in the brain-on trying to understand exactly what happens in neurons to damage and ultimately kill them. In the last several years, scientists have made enormous progress in characterizing the individual players in this process, describing their activities, and determining the exact steps in the disease process. These advances mean that scientists are increasingly able to move away from studies of these players in isolation and focus on their complex relationships and how their effects on each other affect the development of AD. This section of the Progress Report describes a number of significant advances in this area.

New Discoveries About Beta-amyloid and Plaque Formation
Recent studies have dramatically improved scientists' understanding of how beta-amyloid plaques are formed. These discoveries have fundamentally changed how scientists think about this critical component of AD pathology.

APP, the starting point for beta-amyloid plaques, is one of many proteins associated with cell membranes, the lipid barrier that encloses the cell. As it is being made inside the cell, APP becomes embedded in the neuron's membrane, 4/5 on the outside and 1/5 on the inside, like a toothpick stuck in an orange. While APP is embedded in the cell membrane, certain enzymes (proteins that cause or speed up a chemical reaction) cleave it into discrete fragments. Several years ago, scientists identified the enzymes respon-sible for cleaving APP into these peptides. These enzymes are called alpha-secretase (a-secretase), beta-secretase (ß-secretase), and gamma-secretase (?-secretase). In a major breakthrough, scientists then discovered that, depending on which enzyme does the cleaving and the segment of APP within which the cleaving occurs, APP processing can follow one of two pathways that have very different consequences.

In one pathway, alpha-secretase cleaves the APP molecule within the portion that has the potential to become beta-amyloid. Cleaving at this site results in the release from the neuron of a fragment called sAPPa. This fragment has beneficial properties, such as promoting neuronal growth and survival. The remaining APP fragment, still tethered in the neuron’s membrane, is then cleaved by gamma-secretase at the end of the beta-amyloid segment. The smaller of the resulting fragments also is released into the space outside the neuron, while the larger fragment remains within the neuron and interacts with factors in the nucleus.

In the second pathway, beta-secretase cleaves the APP molecule at one end of the beta-amyloid peptide, releasing a fragment called sAPPß from the cell. Gamma-secretase then cleaves the resulting fragment at the other end of the beta-amyloid peptide. Following its cleavage at both ends, the beta-amyloid peptide is released into the space outside the neuron and begins to stick to other peptides of beta-amyloid. These small, soluble clumps of two, three, four, or even up to a dozen beta-amyloid peptides are called ADDLs. The number of individual beta-amyloid peptides within ADDLs varies, but collectively, they are referred to as oligomers. It is likely that some oligomers may be cleared from the brain. Those that cannot be cleared clump together with more beta-amyloid peptides and other proteins and cellular material. As the process continues, these oligomers grow larger, becoming increasingly insoluble entities called protofibrils and fibrils. Eventually these entities coalesce into the well-known plaques that are characteristic of AD.

Until recently, scientists thought that fibrils and plaques were somehow responsible for all the neuronal damage and death in AD. Now, new evidence suggests that the formation of plaques may actually be a kind of clearance mechanism that the brain uses to get harmful beta-amyloid clumps away from neurons.

	Two Pathways... Different Outcomes
	Click Here to Enlarge

Researchers also are increasingly convinced that oligomers that are not cleared from the brain and that do not become part of a plaque may be one of the neuron-damaging culprits. A research team at Northwestern University has even suggested how oligomers damage neurons and cause the memory loss that features so prominently in AD (Cleary et al., 2005; Lacor et al., 2004). Working with cell cultures and tissue extracts taken from AD and rat brains, the scientists found that some oligomers attach themselves to the synapses located on neurites (the structures that branch out from the cell body). When this happens, the synapses are not able to function properly and therefore cannot receive messages from other neurons. Unable to communicate, the neuron ultimately ceases to function and dies. As this destructive process accelerates, essential cognitive operations, such as memory formation and retrieval, are disrupted.

This expanding knowledge about the stages of plaque formation and the toxicity of molecules formed at each stage is giving scientists new therapeutic targets. For example, because toxic forms of beta-amyloid build up most rapidly when they are at high concentrations in the brain, one strategy is to investigate whether high levels of beta-amyloid in the brain might be partly due to how rapidly these molecules are removed from brain tissue into the blood stream through the blood-brain barrier. Several groups are trying to understand which receptors on the surface of cells at the blood-brain barrier are responsible for this transport and whether their ability to move beta-amyloid in and out of the brain might be related to AD pathology.

One research team based at the University of Rochester Medical Center is working on receptors that remove beta-amyloid from the brain (Deane et al., 2004). A receptor called low-density lipoprotein receptor-related protein (LRP) was previously found to be primarily responsible for this task. The new studies show that LRP binds to beta-amyloid at the blood-brain barrier. Interestingly, LRP’s efficiency in clearing beta-amyloid from the brain into the blood was greatest for the most soluble form of normal beta-amyloid. In contrast, mutated forms (such as one found in a Dutch family where mutated beta-amyloid accumulated around blood vessels) and less soluble forms of normal beta-amyloid were cleared from the brain much less rapidly. However, at high concentrations, all forms of beta-amyloid directly caused an increased rate of LRP breakdown, reducing the ability of LRP to clear beta-amyloid from the brain. Together, these results indicate that preserving LRP activity at the blood-brain barrier may be an important component of strategies to remove beta-amyloid from the brain.

A receptor called p75NTR also is coming under scientific scrutiny. Among other actions, p75NTR makes many kinds of neurons more susceptible to beta-amyloid. In a recent study funded by NINDS and other organizations, a research team from the University of Rochester studied the effects of beta amyloid on p75NTR levels in cell cultures of human neurons. They expected that increasing p75NTR would make the neurons more susceptible to beta-amyloid. Instead, they found that exposure to beta-amyloid increased p75NTR activity in the neurons and that the increased activity actually protected the cells, even when cells were exposed to levels of beta-amyloid 2,500 times higher than those usually found in people with AD (Zhang et al., 2003). These findings open the door to future studies to examine the possibility that activ-ating the p75NTR receptor may be a useful strategy for treating AD in humans.

Scientists funded by the National Institute of Environmental Health Sciences, NIA, and the Wisconsin Distinguished Rath Graduate Fellowship in Medicine have investigated another protein that may be protective because it binds beta-amyloid (Stein et al., 2004). This research team, from the University of Wisconsin at Madison, studied why transgenic mice that had defective genes from people with early-onset AD inserted into their DNA had high levels of beta-amyloid deposits but did not exhibit any neurodegenerative symptoms. Further investigations led the research team to discover that these mice were producing high levels of transthyretin, a carrier of the thyroid hormone thryroxine. When the mice were given antibodies that prevented transthyretin from interacting with the beta-amyloid protein, the mice showed an increased level of brain cell death. Cell culture studies of human brain cells treated with transthyretin and beta-amyloid showed minimal amounts of cell death while beta-amyloid alone caused significant cell death. These studies indicated that transthyretin can block the progression of AD in this mouse model by inhibiting the effects of beta-amyloid protein. Although we do not yet know whether transthyretin protects against beta-amyloid in humans, this discovery suggests that it may be possible to develop a drug that increases the production of transthyretin and thus protects people at risk of AD, possibly including those who are at higher genetic risk of developing the disease. The findings also may improve the detection of environmental agents that may play a role in the development of AD by allowing scientists to determine which of these agents upsets the balance between transthyretin and beta-amyloid proteins.

Avenues of beta-amyloid research involving new pathways are proving to be rich ground for discovery. The possibility that blocking the production of beta-amyloid in the brain might prevent the development of AD is one such avenue. NIMH- and NIA-funded scientists at the Howard Hughes Medical Institute and the University of Pennsylvania built on previous research suggesting that an enzyme called glycogen synthase kinase-3 alpha (GSK-3a) may be crucial to the development of AD because inhibiting the action of GSK-3a also inhibits the formation of beta-amyloid plaques and neurofibrillary tangles. In the new study, the investigators examined the potential of the mood-stabilizing medication lithium to inhibit GSK-3a and minimize the biological changes leading to AD (Phiel et al., 2003). The researchers first showed that lithium inhibited GSK-3a in cultured cells, reducing the production of beta-amyloid. Then, they demonstrated that administering therapeutic doses of lithium to transgenic mice markedly reduced the accumulation of beta-amyloid in the mouse brains. These results appear promising and are opening new lines of investigation about the association of lithium and AD.

For example, it may be useful to investigate whether patients who have taken lithium for bipolar disorder show a lower incidence of AD. However, even if lithium does prove to be useful as an AD intervention, its known side effects, such as nausea, fatigue, and hand tremor, particularly in older patients, may require the development of new agents that provide the therapeutic effects without the negative side effects.

Back to Top

New Discoveries About Presenilin
At the same time that researchers are making new discoveries about beta-amyloid oligomers, they are gaining insight into another crucial AD player—presenilin. Presenilin is an essential component of the gamma-secretase enzyme responsible for cleaving APP to form beta-amyloid. Certain mutations in the presenilin genes (PS1 and PS2) cause the most common type of familial early-onset AD (FAD). Most of the research on the effect of presenilin mutations on AD has focused on their effects on the gamma-secretase enzyme complex.

Gamma-secretase is an example of an enzyme whose biological activity depends on several proteins working together in a complex. To understand how this important enzyme functions, we need to know what other proteins it contains besides presenilin. Knowing the additional protein components will allow us to answer important questions about how it works, what proteins besides APP it may cleave, and how its activity can be altered. A group of Harvard Medical School investigators used mass spectroscopy, a highly sensitive analytical technique that can identify and quantify both known and unknown compounds, to isolate and characterize gamma-secretase (Fraering et al., 2004). Using this technique, the researchers determined that the gamma-secretase complex contains three other proteins in addition to pre-senilin—nicastrin, Aph-1, and Pen-2—and that all four proteins must be present for gamma-secretase activity. In addition, the researchers found that increased production of beta-amyloid, or production of a slightly longer form of beta-amyloid that was more prone to aggregate into toxic products, could be modified by several factors that influence the structure of the gamma-secretase components. These factors include the lipid composition of the cell membrane and a number of well-characterized gamma-secretase inhibitors. These results provide important clues to possible therapeutic strategies for lowering the production of toxic beta-amyloid by modulating gamma-secretase activity.

	Presenilin
	Click Here to Enlarge

Researchers are now showing that the presenilin proteins appear to be involved in a number of other nervous system functions as well. Examples are clearance of proteins no longer required by the cell, neurogenesis (the generation of new neurons), and cell signaling. Some of these other functions of presenilins are likely to be important in brain aging and development of AD and perhaps in other neurodegenerative diseases as well.

A team of scientists from the Mt. Sinai School of Medicine, in New York, is exploring the possibility that presenilins normally play an important role in preventing other typical AD pathologies, including changes to tau and cell death (Baki et al., 2004). Using cells taken from transgenic mice with both normal and mutated PS1 genes, the researchers found evidence suggesting that PS1 normally activates a cell signaling pathway called PI3K/Akt. This pathway blocks cell death and inhibits tau phosphorylation. It also appears that AD mutations in PS1 may promote AD pathology by inhibiting the PI3K/Akt pathway. The researchers hypothesized that inadequate activation of PI3K/Akt signaling and impaired transmission of survival signals due to familial PS1 AD mutations (or the loss of normal presenilin function) could contribute to AD pathology independent of PS1’s role in cleaving APP.

Back to Top

For some time, scientists have realized that a number of devastating diseases—such as AD, dementia with Lewy bodies, frontotemporal dementia, Parkinson’s disease, Huntington’s disease, and prion diseases—are characterized by aggregations of abnormally folded proteins. In AD, the abnormal proteins are beta-amyloid and tau; in PD, it’s synuclein; and in frontotemporal dementia, it’s tau. Scientists think, therefore, that the pathological process in these diseases must share some characteristics, though these overlaps are not fully understood. For example, research in the past 2 years on beta-amyloid aggregates has provided evidence that the actual structure of oligomers may help to explain the pathology seen in these diseases. In one series of experiments, scientists at the University of California at Irvine made an antibody in rabbits that specifically recognizes soluble beta-amyloid oligomers. The antibody failed to recognize soluble lower-molecular weight and fibrillar forms of this peptide (Kayed et al., 2003). This antibody also reacted with a variety of soluble protein oligomers that are involved in other neuro-degenerative diseases, regardless of their specific amino acid sequence. Importantly, this particular antibody blocked the toxicity of these oligomers on cells in culture, including that of beta-amyloid oligomers. These results suggest that many types of soluble oligomers contain a common structural feature, independent of their amino acid sequence, and that the toxicity and pathogenesis of these oligomers may be mediated by a common mechanism. Scientists also know that many neurodegenerative diseases have some clinical characteristics in common. For example, some people with AD have trouble moving, the most obvious symptom of PD. Many of those with PD also have dementia. Sleep-wake disorders, delusions, psychiatric disturbances, and memory loss occur in all of these diseases. Finally, it is clear that all of these diseases develop over many years and occur as a result of complex interactions of genes, lifestyle and environmental factors, and factors affecting all parts of the body (such as hormonal changes). One way of looking at this constellation of disease pathologies, manifestations, and classifications is shown below. It recognizes that complex and interactive biological processes over time result in a group of early symptoms that signal that some kind of disease is beginning to manifest itself. The result may be any one of a number of diseases that has unique characteristics as well as shared characteristics with other diseases. By investigating these diseases individually and together, scientists hope to shed light on their causes and, possibly, on future common treatment and prevention strategies. Click Here to Enlarge

Other studies have focused on a newly-discovered effect of presenilin mutations on one of the two main “garbage disposal” systems of the cell, the lysosome. Neurons contain a large number of these structures, which eliminate most of the damaged or abnormal proteins that can accumulate in age-related neurodegenerative diseases. Researchers from the Mailman Research Center, McLean Hospital, Belmont, Massachusetts, found that presenilin gene mutations induce more severe damage to lysosomal systems in the brains of people dying with FAD than that seen in brains of individuals with the more common late-onset AD (Cataldo et al., 2004). This difference was also seen in PS1 transgenic mouse models of the disease. Some of the lysosomal pathology seemed to be in reaction to the presence of beta-amyloid, and some occurred independently. These results indicate that presenilin mutations can compromise the major garbage disposal systems of the cell and cause a loss of normal neuronal function.

In another twist to the garbage disposal story, a University of Pennsylvania team of scientists showed that removing the presenilin genes altogether in transgenic mice also interfered with the cell’s lysosomes (Wilson et al., 2004). The investigators showed that the faulty lysosomes in these mice contained very high levels of two proteins, alpha- and beta-synuclein. This is an important finding because the abnormal accumulation and aggregation of alpha-synuclein has been implicated in several neurodegenerative diseases, including Parkinson’s disease. The effects were not due to loss of the gamma-secretase activity itself, but seemed more likely to be due to an effect on calcium metabolism within the cells that lacked PS proteins. Thus, aberrant accumulation of alpha- and beta-synuclein in degradative organelles are novel features of neurons that lack PS1, and similar events may promote the formation of clumps of alpha-synuclein found in several neurodegenerative diseases.

New Discoveries About Tangles and Tau
Even though scientists have agreed for years that plaques and tangles are two of the major features of AD, they’ve disagreed about the relative importance of each. The development of the mouse model of AD that exhibits all the major characteristics of the disease has given scientists further insights into plaques and tangles, their interrelationships, and the roles that both beta-amyloid and abnormal tau play in damaging normal neuronal function.

	Tangles and Tau
	Click Here to Enlarge

A recent study supported by NINDS, NIA, and other organizations also explored the association between beta-amyloid and tau (Gamblin et al., 2003). A team of Northwestern University researchers used test tube studies to demonstrate that certain enzymes called caspases can break the tau protein into segments when they are in neurons that have already been treated with beta-amyloid. The resulting segments have a greater tendency to form damaging tangles than does normal tau. These findings help expand our understanding of the functional relationship between beta-amyloid and tau and suggest that the caspases, which are well known for their involvement in cell death pathways, also may be involved.

Recent studies have examined other aspects of tangles and tau, and they also are helping to explain tau’s role in the AD process. For example, some scientists think that the misfolded shape of the highly phosphorylated tau in the neuron increases its tendency to form neurofibrillary tangles and encourages the loss of stabilization of microtubules. Studies have shown that the molecular chaperones Hsp90 and Hsp70 (proteins that protect cells from the adverse effects of protein misfolding and aggregation), stabilize tau’s normal shape, reduce tau’s abnormal phosphorylation and aggregation, and restore its binding to microtubules (Dou et al., 2003). This research, which was funded by NIA, the Alzheimer’s Association, and the Ellison Foundation, has implications not only for AD and diseases caused by tau mutations (such as FTD and parkinsonism linked to chromosome 17), but also for other neurodegenerative diseases, such as Huntington’s disease and prion diseases that are characterized by abnormal aggregates of other types of protein.

Scientists have known for years that tau is important in maintaining microtubule integrity and, therefore, their ability to transport materials and organelles, such as mitochondria, between the cell body of the neuron and the synapse. However, only more recently has it become clear that the level of tau and normal phosphorylation might also regulate the actual transport of materials down intact microtubules. As with other neuronal systems, a delicate balance must be maintained for optimal function. A research team at the New York State Institute for Basic Research in Developmental Disabilities investigated how high levels of tau actually inhibit axonal transport from the cell body to the synapse (Tatebayashi et al., 2004). They found that the enzyme called GSK3ß, which is regulated by the PS1 gene, also phosphorylates tau under normal conditions. This phosphorylation does not alter tau’s ability to bind to microtubules, but it appears to be required for normal axonal transport in cells. These findings raise the possibility that the physiological phosphorylation of tau by GSK3ß also may be involved in regulating organelle transport. Supporting the idea that either too much or too little of a protein like GSK3ß might adversely affect neuronal function, NICHD-funded researchers from the University of Connecticut Health Center found that another function of PS1 is to maintain physiological levels of GSK3ß (Pigino et al., 2003). When the gene for PS1 was eliminated, GSK3ß activity increased. This caused a critical transport protein to fall off the microtubules, causing a decrease in axonal transport. The investigators then inserted a PS1 gene containing a mutation called M146V, which causes a very aggressive early-onset form of FAD and got the same effect.

In the last several years, researchers have actually begun to test a potential therapy for the loss of neuronal function that occurs when microtubules disintegrate and tau filaments begin to form tangles. In studies funded by NINDS, several other NIH Institutes, and the Institute for the Study of Aging, a research team at the University of Kansas showed that beta-amyloid toxicity can be reduced by pretreating tissue culture cells with agents that stabilize microtubules, such as taxol, a drug used in cancer treatment (Li et al., 2003a). These agents seem to work by indirectly inhibiting an enzyme that abnormally phosphorylates tau. The researchers then generated taxol-like compounds that, unlike taxol, can cross the blood-brain barrier and showed a likely positive effect of these drugs in combating beta-amyloid toxicity in adult mice. Working in the same area, a team of scientists from the University of Pennsylvania School of Medicine treated transgenic mice that develop AD-like tangles with a drug that binds microtubules—in effect, substituting for the tau that should normally be there (Zhang et al., 2005). The investigators found that after 12 weekly treatments, the drug restored microtubule function, increased the number of stable microtubules, and lessened motor impairments in the mice. These findings open the way for additional studies to explore the therapeutic potential of drugs that offset the effects of beta-amyloid toxicity and loss of microtubules.

	Damaged and Healthy Mitochondrion
	Click Here to Enlarge

AD researchers have thought for some time that damage to or mutations of mitochondria could play a role in the early development of AD because they lead in several ways to the death of the cell through a process called apoptosis (Beal, 2000; Melov, 2004). Damage to mitochondria also leads to a rapid increase in the formation of free radicals, which are highly reactive oxygen molecules that can build up in neurons over time. If unchecked, the build-up of these molecules can cause oxidative stress, which damages other cellular molecules such as proteins, lipids, and nucleic acids. Oxidative stress and mitochondrial dysfunction have been implicated in several neurodegenerative diseases, including AD.

Several research teams have recently expanded our understanding of mitochondria, free radicals, and oxidative stress. For example, Columbia University scientists observed that in AD brains, beta-amyloid is bound to a protein called Aß -binding alcohol dehydrogenase (ABAD), which is an integral part of mitochondria (Lustbader et al., 2004). Studies in AD transgenic mice show that the binding of beta-amyloid peptides to ABAD prevents the normal function of ABAD and leads to enhanced production of free radicals followed by increased apoptosis. When the investigators developed transgenic mice that overproduced ABAD as well as developed plaques, they observed increased oxidative damage in their brains. Cognitive tests showed that these mice had impaired memory compared to AD transgenic littermates that had normal levels of ABAD. If these findings are borne out in additional studies, blocking the interaction between ABAD and beta-amyloid could be a possible target for future therapeutic strategies in AD.

A research team from the Buck Institute for Age Research, in Novato, California, took advantage of the availability of genetic models of oxidative stress as well as other technologies to explore the impact of mitochondrial oxidative stress on the activity of individual mitochondrial enzymes (Hinerfeld et al., 2004). For this research, they used mice deficient in a key antioxidant enzyme known as superoxide dismutase type 2 (SOD2). These mice suffer a loss of neurons in particular brain regions. The investigators found that certain parts of the mitochondrial molecular machinery that produces energy are more vulnerable to oxidative stress than are others. Most important from a therapeutic standpoint was the finding that an antioxidant compound was able to prevent the damage of the mitochondrial enzymes, and in doing so, restored some of the neuronal loss caused by the oxidative stress.

NIA researchers have tracked the declines in synapse activity and neuronal energy consumption that occurs during the evolution of the AD process (Rapoport, 2003). They found that changes in the structure and function of synapses resulting from AD pathology reduce neuronal energy demand and lead to potentially reversible decreases in energy production within the mitochondria. PET scans showed that, at this early stage in the disease process, the synapses can almost be normally activated in response to stimulation. As the disease progresses, however, tangles with abnormally phosphorylated tau accumulate within the neuron to the point that they disrupt microtubules and axonal transport. Mitochondria are prevented from traveling along the axon between the cell body and the synapse. The resulting severe energy depletion at the synapse and other pathology leads to the death of the neuron. If these findings are confirmed by further research, strategies to maintain neuronal energy metabolism during the early stages of the disease process may be a potential therapeutic goal.

Other research teams have focused on genetic aspects of mitochondria’s role in AD. Mitochondria possess their own DNA, distinct from the DNA in the cell’s nucleus. Although the amount of DNA in a mitochondrion is very small compared to that in the nucleus, mutations in it can have deleterious effects. Researchers from Cornell University Medical College set out to test whether people with AD have mutations in mitochondrial DNA (Coskun et al., 2004). They found several mutations in the part of mitochondrial DNA that controls the activity of mitochondrial genes. These genetic changes could be found in more than half of the AD brains examined but in none of the brains from cognitively healthy individuals of the same age. These AD-specific mutations in mitochondrial DNA could underlie the mitochondrial failure to produce sufficient energy for proper functioning of neurons, which, in turn, can result in loss of communication between neurons. These investigators speculate that the accumulation of these mutations might be a significant and contributing factor to the development and progression of AD.

Finally, a research group from Oregon Health and Sciences University compared the expression of mitochondrial genes from the brains of cognitively healthy individuals and brains from people at early- or late-stage AD (Manczak et al., 2004). They found AD-related changes in the patterns by which certain genes are expressed. These genes encode the proteins that make up the three major components of mitochondria’s energy-producing machinery. These findings suggest that the ability of the mitochondria to produce energy is compromised in AD and that the mitochondria inside the surviving neurons are trying to compensate for the brain’s loss of energy resources.

	Players on the AD Stage: Putting It All Together
	Click Here to Enlarge

Cell Death. An area that is generating increasing scientific interest is abnormal cell death caused by disruptions in the normal function of key structures in neurons. For example, a proteasome is a large structure, which, along with lysosomes, is responsible for chopping up misfolded or damaged proteins so that the cell can dispose of them. This critical system protects cells from the toxicity of damaged proteins. Increasing evidence suggests that disruptions to proteasome activity occur in normal aging as well as in many neurodegenerative conditions, including AD, PD, and Huntington’s disease. The inhibition of proteasome activity causes many and diverse effects on cells, including cell death, and recent studies suggest that proteasome inhibition may contribute to some disease-related mitochondrial changes. A research team from the University of Kentucky explored the effects of chronic, low-level proteasome inhibition on cell function by looking at changes in the activity of a large number of genes. These changes are called “gene expression” (Ding et al., 2004).

Another type of cell death is the process called apoptosis. Some evidence suggests that cell death in AD occurs through this process. Sometimes, apoptosis involves mitochondria. Other times, the process can be activated by membrane receptors, such as tumor necrosis factor receptor (TNFR). Studies have shown that components of a TNFR-1 pathway were elevated in AD brain tissue, while cellular inhibitors of this pathway were decreased. One, called TRADD, binds to the TNFR-1 receptor, causing cell death. Another, called DENN/MADD, binds to the TNFR-1 receptor and prevents cell death. Researchers at the University of Southern California Keck School of Medicine recently found a reduction in DENN/MADD and an increase in TRADD in the hippocampus of brains from people with AD compared to tissue from cognitively healthy people (Del Villar and Miller, 2004). Similar changes were seen in neurons treated with beta-amyloid and in cortical tissue from transgenic mice that develop AD-like disease. These changes, which may be caused by inflammation or oxidative stress, may be responsible for increased apoptosis in the AD brain.

Inflammation. A dynamic and complex biological process, inflammation affects cells and tissues all through the body. This process occurs in response to many types of injuries or abnormal situations. In some cases, the inflammation reaction occurs as part of a healing process, such as when the body reacts to a simple scrape on the skin.

In other cases, inflammation is a central characteristic of a disease process, such as in rheumatoid arthritis. Inflammation occurs in the AD brain as well, and many scientists are examining its role in the development and progression of the disease. They don’t necessarily agree on its significance—some scientists think that inflammation is part of a vicious cycle that is harmful to neurons. Others think that aspects of inflammation may be valuable to the brain by counteracting the detrimental aspects of the AD process.

Recent inflammation-related research has focused on two kinds of glial cells: astrocytes and microglia. These cells have the capacity to move toward and respond to sites of injury. Beta-amyloid plaques in the AD brain are surrounded by numerous activated astrocytes and microglia. In fact, the presence of large numbers of activated astrocytes surrounding sites of brain injury is one of the earliest manifestations of AD. Scientists speculate that this event could occur in response to degenerating synapses and neurons and to the accumulation of beta-amyloid plaques.

A collaborative team of researchers from Columbia University, Stanford University, and the Gladstone Institute of Neurological Disease in San Francisco demonstrated that astrocytes isolated from adult mouse brains can degrade and remove beta-amyloid peptides (Wyss-Coray et al., 2003). This finding from tissue culture was a surprise because in AD brains, astrocytes surrounding beta-amyloid deposits seem incapable of removing beta-amyloid. Discovering ways to stimulate the ability of astrocytes to clear beta-amyloid may be a fruitful strategy for reducing the toxicity and neurodegeneration associated with plaques and their earlier-stage aggregates.

Other investigators are exploring stra-tegies for stimulating microglia. Like astrocytes, most of the evidence to date shows that microglia can engulf and remove beta-amyloid aggregates in tissue culture, but they fail to do so in living brains. Identifying the molecular details of this “phagocytic” activity is crucial if scientists are to develop microglia that will successfully engulf and remove beta-amyloid aggregates from the brain. Researchers from Cleveland’s Case Western Reserve University School of Medicine used both pharmacological and genetic approaches to discover in tissue culture that beta-amyloid fibrils are engulfed by a receptor composed of several proteins on the surface of microglia. This receptor is entirely different from the typical phagocytic receptors (Bamberger et al., 2003). When each of the receptor components was blocked, the microglia lost the capacity to degrade beta-amyloid (Koenigsknecht and Landreth, 2004). These findings pave the way for researchers to find out why the plaque-associated microglia in living brain tissue are unable to remove the beta-amyloid deposits effectively, despite their physical association with the plaques.

Other scientists have found that microglia also may be responsible for some of the damaging processes that occur in the AD brain. Microglia are rapidly activated in the presence of aggregated beta-amyloid and are responsible for the strong inflammatory response that can encourage the neurodegenerative changes observed in AD. Deciphering the molecular mechanism by which fibrillar beta-amyloid activates this inflammatory response may suggest future therapeutic targets. In trying to address this problem, scientists from Massachusetts General Hospital put beta-amyloid into the brains of normal mice and mice that were deficient in a particular cell surface protein called CD36 (El Khoury et al., 2003). The researchers found that far fewer microglia were recruited to the beta-amyloid site in CD36-deficient mice compared to normal mice. In addition, microglia isolated from CD36 mice could not produce certain toxic molecules in response to aggregated beta-amyloid, which precluded the ability of these molecules to harm neurons.

In an effort to understand how to regulate the negative effects of microglia, Stanford University scientists found that a growth factor, TGF-ß, which regulates cell survival and inflammation throughout the body, has a key role in maintaining neuronal integrity and regulating microglial activity (Brionne et al., 2003). The researchers found that the brains of mice that are deficient in TGF-ß have widespread neuronal degeneration and microgliosis, a process characterized by the presence of numerous microglia that have the capacity to harm neurons. They also examined transgenic mice that overproduced TGF-ß in their brains and found that overproduction of TGF-ß protected against various forms of brain injury. This finding is highly significant in light of the fact that humans have large variations in the levels of TGF-ß in the brain.

Other researchers have been investigating additional factors that may influence brain injury caused by beta-amyloid. A team of scientists at the University of California at Irvine developed transgenic mice that lack a protein called C1q. This protein is involved in the complement cascade, a precise series of events that takes place during the body’s immune response; inflammation is part of this response. The researchers bred these C1q-deficient mice with transgenic mice that develop amyloid plaques and found that the offspring that develop plaques in the absence of C1q do not exhibit the intense inflammatory response in the brain tissue as mice that have C1q (Fonesca et al., 2004). The C1q-deficient mice also showed a slower loss of markers of neuronal communication as the plaque pathology developed. These findings suggest that C1q is part of the molecular signaling that leads to the deterioration of neurons in the presence of beta-amyloid aggregates.

Epidemiologic studies have suggested that long-term use of non-steroidal anti-inflammatory drugs (NSAIDs), such as naproxen, ibuprofen, and indomethacin, are associated with a decreased risk of AD. These studies provide another avenue of research suggesting that an inflammatory process may be involved in AD. However, previous laboratory research by a group of scientists from the Mayo Clinic in Jacksonville, Florida, showed that some NSAIDs could lower beta-amyloid production, hinting that this mode of action, rather than an inflammatory pathway, might be responsible for the possible protective effect of some NSAIDs. To further investigate this connection, the Mayo Clinic researchers used cell cultures and an AD transgenic mouse model to study the effects of a number of common NSAIDs on beta-amyloid production (Eriksen et al., 2003). After screening 20 widely-used NSAIDs, the scientists found that 8 effectively reduced beta-amyloid levels both in the cell cultures and in the mouse models. Moreover, this action occurred through a different mechanism than that usually associated with most NSAIDs. The investigators suggest that some NSAIDs could act to delay the development of AD by inhibiting gamma-secretase, and that it might be useful to screen other types of agents for their effect on this key process in AD neuropathology. So far, no clinical trial has shown a beneficial effect of NSAIDs on AD prevention or progression.

The Brain Protects Itself
Until recently, scientists thought that the brain’s ability to repair itself after injury or neurodegeneration was limited or non-existent. In the past several years, however, with new discoveries about brain anatomy and function and the emerging concept of cognitive reserve, scientists are changing that view.

One exciting recent development is the discovery that the brain can, in fact, generate new neurons in a process called neurogenesis. In the adult, this process occurs in two areas of the brain, the hip-pocampus and the subventricular zone, and possibly in others as well. These brain areas contain special cells called stem cells, which can divide and migrate to other areas of the brain under certain conditions. These adult neural stem cells can make new neurons and glial cells. This discovery raises the prospect that neural stem cells could be harnessed to replace dying cells or to repair damaged cells and neural circuits in the aging or diseased brain.

Many researchers are studying the fundamental biology of neurogenesis and how it may be linked to AD. For example, NIA-funded researchers recently examined whether abnormalities in neural stem cells might contribute to human disorders of learning and memory, including AD (Haughey et al., 2002). If neurogenesis is impaired in AD, then it may be possible to develop therapeutic approaches that enhance neurogenesis. The investigators wanted to determine whether stem cell proliferation, survival, and differentiation were impaired in transgenic mice that developed a form of FAD. They also wanted to assess the effects of beta-amyloid on stem cell survival and neuronal differentiation. They found that proliferation and survival of the stem cells in a certain region of the hippocampus was reduced in the mice. They also found that beta-amyloid impaired proliferation and neuronal differentiation of human and mouse stem cells, and promoted apoptosis of the cells.

Presenilin mutations also have been shown to reduce neurogenesis from stem cells in the brain. Mt. Sinai School of Medicine scientists discovered that the PS1 mutation that causes the most severe and earliest form of FAD had a major inhibitory effect on the survival of newly-formed neuronal cells from stem cells in mutated PS1 transgenic mice compared with transgenic mice that contained the normal PS1 gene (Wen et al., 2004). This could have implications for the physiological functioning of the hippocampus in FAD. In contrast to the impaired neurogenesis caused by PS1 mutations in other transgenic mouse models of AD, the NINDS-and NIA-supported work of researchers at the Buck Institute for Age Research showed that neurogenesis is actually enhanced in the hippocampus of people with AD as well as in a transgenic mouse model with two APP mutations that cause FAD (Jin et al., 2004). It is possible that different forms of AD mutations have different effects on brain neurogenesis and that neurogenesis may be an important mechanism the brain uses to try and repair the early effects of AD. Enhancing neurogenesis could be an effective therapeutic strategy.

Other basic science studies conducted by NIA-funded researchers have focused on a protein called Notch, which resides on the surface of neural stem cells. Previous studies have shown that Notch controls the process of neurogenesis. When the gene for Notch is disabled in mice, the development of their brains is severely abnormal and they die in the womb. A similar abnormality in brain development occurs when the PS1 gene is disabled. Mutations in PS1 cause FAD, suggesting a link between Notch and AD. NIA-funded scientists designed experiments using transgenic mice to determine whether Notch is involved in learning and memory (Wang et al., 2004). These mice, which had reduced levels of Notch, exhibited impaired communication at synapses between hippocampal neurons that are critical for learning and memory. When a protein that activated Notch was applied to the hippocampus, communication at the synapses was enhanced. These findings suggest an important role for Notch in synaptic function associated with learning and memory processes, and it may be a new target for therapeutic interventions to improve learning and memory.

Scientists also are beginning to look at other regulators of neural stem cell biology, especially as they influence aging and neurodegenerative diseases like AD. For example, NIMH- and NIA-funded investigators at the Washington University School of Medicine in St. Louis found that transgenic mice had a decreased capacity to generate new neurons in the hippocampus and this decrease was greatest when they were under stress from being housed in isolation (Dong et al., 2004). These deficits also were associated with deficits in contextual memory, which is the aspect of memory related to experiences such as recognizing physical surroundings. Greater amounts of plaques also were deposited in the stressed mice compared to unstressed mice. However, an antidepressant medication that increases cell proliferation in normal mice also increased cell proliferation and improved contextual memory in the stressed transgenic mice. These results suggest that this AD mouse model has an impaired ability to generate new cells in the hippocampus and that this impairment can be modulated by stress and certain drugs.

A number of important questions are being asked about the role of neurogenesis in counteracting the effects of neuro-degenerative damage. In an effort to find answers, a Mt. Sinai School of Medicine research team studied whether injuries to the entorhinal cortex, which has direct connections to the hippocampus, affected the ability of mice to generate new neurons from stem cells in the hippocampus (Gama Sosa et al., 2004). The development of new neurons and their survival was studied for several weeks after the injury. Results showed that the number of new neurons that were formed in the days immediately after the injury did not increase. However, 6 weeks later, the number of new neurons had almost doubled. This transgenic mouse shows promise as a model for studying the molecular events affecting neurogenesis in the hippocampus.

Researchers also continue to be very interested in examining the potential of neurotrophic factors, which are proteins that stimulate neuronal survival and growth, to prevent or treat neurodegenerative disorders. The problem has been how to deliver these proteins to the brain because they do not readily cross the blood-brain barrier. Another challenge is how to specifically target them to degenerating neurons to avoid unwanted side effects. Gene therapy offers one approach. Recently, a group of researchers from the Salk Institute for Biological Studies in La Jolla, California, developed a gene delivery system to carry the human neprilysin gene, which makes an enzyme that degrades beta-amyloid (Marr et al., 2003). The team put the neprilysin delivery system into brain areas containing beta-amyloid plaques in APP transgenic mice. The production of neprilysin appeared to increase degradation or reduce the growth of existing plaques. The amount of plaque was reduced to less than half that found in untreated areas. This study suggests that boosting normal, protective processes in the nervous system might help prevent or treat degenerative events associated with AD, and this strategy provides a new way of potentially interfering with the disease process.

Back to Top

<< Back | Next >>