Applying Insights from the Study of Normal Aging to Solve Dementia

by Brenda Patoine

May, 2009

/uploadedImages/News_and_Publications/Special_Publications/Articles/1_jessell_thomas.jpg Carol A. Barnes, Ph.D.
Professor of Psychology and Neurology
Research Scientist
University of Arizona

Q: You are examining how memory processes change with normal aging. Why focus on normal memory changes as opposed to Alzheimer’s disease or other pathologies?

A: One reason we are interested in studying normal age-related memory changes is that only about five percent of people over 65 have Alzheimer’s disease (AD). The proportion increases in older age ranges, but the fact remains that most of us will age relatively normally and will not have a dementing illness. The question then becomes: what is normal versus what is pathological in terms of memory changes? To answer that, you have to first know what is normal. Only then can you begin to assess what goes wrong in Alzheimer’s or other dementias. That’s the approach we have taken.

My interest in separating pathology from normal also stems partly from my fascination with the fact that only humans get AD. No other animal gets it. What we’re studying in animal models of memory loss is normal age-related memory loss. We’re trying to use the knowledge we can gain from animals to then define in humans what might be expected at what age range. If we could understand that, it would have a tremendous impact. The population is growing fastest in the higher age ranges, so it would be very good for us to have a clear, long-term view of how the brain processes that underlie memory change with age.

That’s the basic rationale for this kind of work. The ultimate goal is to develop therapeutic interventions that could be used to optimize cognitive performance, in normal aging or in pathological states.

Q: Has the view of what happens to the brain during normal aging changed in recent years?

A: One critical issue revolves around the question of what is happening in the brain networks underlying memory. Do we lose neurons as we age from brain structures that support memory? The dogma until very recently has been, yes, of course, we lose many many neurons as we age. In fact, it was presumed that we lose neurons every day. That has turned out to be false. It is really remarkable that the hippocampus, the structure that’s so important to memory, actually maintains a steady supply of its principal cells across the lifespan. This has now been shown across species, from mice and rats to dogs and primates.

If cell loss is not the issue, then what is it that goes wrong in the aging brain? It turns out that what goes wrong involves changes in the connections between the cells. Even where there is no reduction in the number of cells, there is a loss of synaptic contacts. As a result, the ability of one cell to communicate with another is diminished. This disrupts the networks that serve memory functions.

So we have two sets of broad changes occurring in the aging brain. We know that the synaptic contacts decline; that’s an anatomical change. We also know that the ability of the synapses to be modified—the plasticity of the synapse—is altered with aging. That represents a functional change.

Q: What does this emerging view of normal aging suggest in terms of therapeutic development?

A: I think this is good news in terms of therapeutics. The reason I feel that way is because if the cells are gone, there is very little you can do to replenish them, at least in most parts of the brain. But there may well be something you can do about reconnecting cells. Maintaining these cell-to-cell connections is critical, because if the cells are not communicating as well, synaptic plasticity, and the ability of the circuits to store information will be altered.

The quest to better understand the activity across networks of cells has fostered new approaches to studying brain circuits. For example, so-called ensemble recording methods make it possible to record from multiple single cells simultaneously to track changes in neural networks in real time.

This kind of approach is necessary for decoding high-level brain functions such as memory, which involve complex interactions among various brain regions and populations of cells. Observing the activity of one cell in a particular brain structure is not going to give you a very good idea of what the structure does unless you’re in a very basic structure such as the primary visual area, for example. In a higher-level structure like the hippocampus, which is very complex in terms of its connections with other brain areas, it’s particularly important to be able to monitor the activity of populations of cells.

Q: What have you learned so far from studying the behavior of networks of cells in the hippocampus?

A: We’ve been using ensemble recording to study the activity of a particular type of cell in the hippocampus— the so-called “place cells.” These cells record where an animal is in a particular space and generate what I call a “hippocampal map,” which is essentially the pattern of cell activity that corresponds to a specific environment. If you put an animal into a very familiar environment, the animal will retrieve the corresponding map. You can actually reconstruct where the animal was in the environment just by analyzing the firing patterns of the place cells.

Typically, you can bring the animal back into the same environment time after time after time, even after a delay, and they retrieve the same map. But in older animals, something sometimes goes wrong: about a third of the time they retrieve the wrong hippocampal map. They retrieve a completely different map for a completely identical experience, as though they really think they are somewhere else.

It is analogous to being in Tucson and trying to use a map of Phoenix to find your way around. You’re going to have grave problems in spatial navigation. We think this could be part of the explanation for why older people and older rats more frequently become lost; it may be because they don’t have the right map to help them navigate the space.

Q: What might be impairing the older person’s ability to retrieve the right spatial map from their memories?

A: What we think is going wrong gets back to the concept of plasticity. It turns out that you need a strong capacity for modifying synapses in order for cell-to-cell communication to stabilize the hippocampal map. We think that something goes wrong with cellular plasticity mechanisms—the same mechanisms that underlie learning. We’ve known for many years that plasticity mechanisms are weaker in older animals. It’s not that older animals can’t learn; plasticity does occur, but the durability of the plasticity mechanism declines with age. So the failure to call up the right map may be related to the older animals’ reduced ability to make strong synapses.

Q: What are the next steps in your research?

A: We’re exploring an exciting avenue that allows us to examine large circuits of individual cells spread across the brain to investigate which cells are active in a given experience, and how patterns of activity change with age. Our dream is to be able to do this as a whole-brain image—that is, to look at networks of cell activity throughout the entire brain all at once. Right now, we are settling for hippocampal circuits and some neocortical circuits.

We are looking at dynamic changes in gene expression, and specifically, at the Arc gene, which is necessary for proper synaptic function. Arc expression is an indicator of where activity occurs in a given cell network—we can think of it as a marker of cellular activity. With this method, you can determine, for example, whether the same circuits or different circuits are being engaged in the young brain versus the older brain, just by looking at the patterns of Arc expression. And you can extract temporal information by looking at whether Arc is expressed in the cell nucleus or in the cytoplasm. We call this the catFISH method, an acronym for compartment analysis of temporal gene transcription activity by fluorescence in situ hybridization. catFISH not only gives you the number of cells activated but also the time frame during which those cells were activated.

Using this method, we have shown that the same proportions of a type of hippocampal cell (the CA-1 pyramidal neuron) are active in a given environment in young and old rats alike. In other words, if 30 percent of CA-1 cells are active in a particular behavioral situation in a young rat, the same proportion will be active in an old rat during that same behavioral condition. That fits with the concept of no age-related cell loss in this part of the hippocampus.

You can then perform a PCR analysis [short for Polymerase Chain Reaction, a fast, inexpensive technique for making an unlimited number of copies of any piece of DNA] to obtain levels of Arc RNA and provide a quantitative measure of how much of the Arc gene product is present during the particular behavior. Now we are able to put these two techniques together, using half the brain for the catFISH analysis and the other half for the PCR method. Using this dual approach, we’ve been able to determine that, even though the same numbers of cells express Arc in the old animals as in the young, there is less Arc activated per cell in the older animals. So for some reason, Arc expression is reduced in older animals.

Q: What do you think could be behind the decreased expression of Arc in older animals?

A: We think that this reduction in Arc transcription in the old animals is partly due to a reduction in DNA methylation [a naturally occurring process by which DNA is chemically modified, typically resulting in gene silencing.] This is a very exciting new finding that we are in the process of submitting for peer-review publication. It may be one mechanism responsible for the instability of the hippocampal network in older animals that we discovered in our electrophysiological recordings. It ties together the results we’ve seen from the ensemble recordings and may provide an explanation for those results—the answer to the “why” question.

We are also excited by this finding because we know that if we can understand and target the mechanisms of this altered DNA methylation process, we might be able to pump up cellular levels of Arc to strengthen the synaptic connections and, ultimately, improve memory. That is the hope.

Right now this is all pie in the sky, because we aren’t able to do that yet. We need to carefully analyze exactly where this process is going wrong. There are a number of regions on the Arc gene that could be affected by methylation, so we’re trying to identify the locus, to understand where this is happening. If we can find a very selective locus, we might be able to identify an agent that would reverse that process and therefore reverse some of the age-related memory changes that are seen in animals and humans—that’s the thinking.

To get to that point, it’s critical that we understand the specificity of the changes that reduced methylation produces in the Arc gene, because if you change DNA methylation throughout the brain in a nonspecific manner, you would wreak havoc on all sorts of cell functions. But if you can target a selective region of the gene, then you have some hope of selectively improving cognition and not causing all sorts of horrible side effects. That is one of our goals long term.

Q: What drives your work in your field?

A: I see the suffering of people who have Alzheimer’s disease or other age-related dementias, and there is so much pain for family members. I cannot personally imagine anything more horrible than my memory going. I don’t want to live forever, but I really want to have a functioning brain for as long as possible, and I don’t think I am alone in that desire. What are we besides our memories? When you lose your memories, you lose yourself in a way, and that is just devastating.

Unfortunately, I think the prognosis for finding a cure for Alzheimer’s is rather discouraging right now, so the best we can do is to understand it well enough to at least try to delay it. We really don’t have the best drugs available to us right now to be able to do anything but delay the progression a bit. Some breakthrough could be right around the corner, but in reality, we’re probably looking at a much longer trajectory, on the order of five or ten years. I think there is a lot to be learned from just understanding normal aging well enough. If you can do that, some of those agents that might come out of research on normal memory problems might be applicable to dementia and Alzheimer’s disease.

Q: Does that suggest that Alzheimer’s exists on a continuum with normal aging in terms of memory problems?

A: I really do not see Alzheimer’s and normal age-related memory loss on a continuum at all. I’ve contributed to some of the literature that suggests that they really are distinct phenomena. This has not necessarily been a popular view, but I think more and more people are coming around to having the same viewpoint as the evidence gets stronger.

Rather than being on a continuum, I believe that Alzheimer’s disease is overlaid on normal aging, which partly explains why it is so hard to tease out pathological changes from normal ones. The patterns of changes in normal aging are very distinct from the brain regions that change in AD. There may indeed be similar mechanisms at play, but they are affecting different regions.

For example, in normal aging, monkeys, humans, and rats all seem to be vulnerable in the dentate gyrus, a region of the hippocampus. In AD, there is very little change in the granular cells of the dentate gyrus compared to age-matched controls until very late stages, but the CA-1 region of the hippocampus is particularly devastated. So I think that similar things may be happening mechanistically in those distinct regions. If we can understand what’s going wrong it might be possible to modify those processes, and such modifications might therefore be applicable to both normal aging and Alzheimer’s.