Posted 15 December 1998
Interviewed by Keith A. Crutcher
We invite you to submit comments and questions on John Trojanowski's remarks.
Q. I appreciate your taking the time to talk to me.
A. I am pleased to do this, since, although my major focus as a biomedical
researcher is to try to unravel the causes and mechanisms of neurodegenerative
disorders such as Alzheimer's disease (AD), I also am dedicated to the mission
of trying to help educate the public about the seriousness of these disorders
and what research needs to be done to develop cures for them. In fact, I
can show you something that was done in an effort to educate the public
about some of these disorders, much of the article is relevant to the questions
that you sent to me. This was a publication (Connections vol. 7, #2) done
by the ADEAR center, the educational arm of the national institute on aging
(NIA), about an AD-like disorder known as dementia with Lewy bodies (LB).
Part of the text came out of a recent review from our group, but this article
was written to make the concepts of the original review much more accessible
to the public. The ADEAR article provides an overview of our concept that
many neurodegenerative diseases including AD are diseases due to "fatal
attractions" between normal or "good" proteins that interact
abnormally, precipitate in the brain and cause problems leading to the malfunction
of brain cells or their degeneration. Indeed, we think that this is a common
pathological or mechanistic theme of many different degenerative diseases
of the brain, including those that are common or rare, as well as sporadic
or hereditary. I hope this kind of article will improve public understanding
of these disorders, since I think it's very important that the public knows
as much about these issues as possible so they share the decision-making
about how we proceed to find ways to treat the many neurodegenerative disease
for which we have no effective therapies at this time.
Q. Yes, and one of the goals for these interviews is to get people who
are really considered to be the thought leaders in the field to talk candidly
about what they think the state of the art is.
Q. So one of the things we were wanting to get, from the horse's mouth
so to speak, is your guiding hypothesis. If you had unlimited resources
right now, what hypothesis would you be testing?
A. Well, this ADEAR review is timely since it focuses on LB diseases such
as Parkinson's disease (PD), diffuse LB disease (also known as dementia
with LB or DLB), etc., and these disorders are prototypical examples of
a number of common and uncommon sporadic and hereditary diseases that we
include among disorders that are due to "fatal attractions" of
brain proteins. Examples of brain proteins that are liable to be involved
in "fatal attractions" and cause disease are the presenilins,
amyloid precursor protein (APP), tau, synuclein and prion proteins. Some
of these proteins have only been discovered very recently because they are
involved in diseases such as AD, but others, like tau, were discovered earlier,
and we know a lot about what they are supposed to do and so we can carry
out more informed experiments of their function-dysfunction when they are
either mutated or altered and form aggregates in a neurodegenerative disease.
To be more specific, the notion of fatal attractions is something I think
links all of these diseases because these proteins somehow become sticky,
bunch together and form aggregates, accumulations of proteins, like a cat
unstringing a ball of twine and making a mess in a house where it would
trip people up and so forth. I think another analogy one could use, is one
that I gave them for the article, and that is the notion that by aggregating,
the proteins "gum up" the functions of the cell by blocking the
normal movement of organelles and proteins around the cell, which is very
fundamental and necessary to the function of a cell much in the way that
the flow of traffic in LA is necessary for the city to function. Thus, when
the traffic is slowed down, this will bring the city to its knees.
So if that's the case, I think one central hypothesis that needs to be
tested is: Are these protein aggregates really the proximal causes of nerve
cell degeneration, and, in some cases glial cell degeneration, or are they
epiphenomena in the disease? We don't think this is the case with the tau
aggregates in the tangles of AD because there are disease causing mutations
in the genes that encode tau, APP, the presenilins, prions, and alpha-synuclein.
Thus, it is reasonable to conclude that these proteins, when abnormal, are
proximal causes of disease. Indeed, if you have these mutations, you're
going to get the disease 100% of the time, if you live long enough, and,
except for those who die early in life, there are no escapees. Thus, there
are several solid reasons that allow us to focus research seriously on these
proteins to understand how they lead to nerve cell death.
Q. What kind of critical experiments you would see needing to be done
either to support or refute the hypothesis?
A. The critical experiments that we and others are trying to do is model
these mutant proteins in cells and transgenic mice to see exactly how death
occurs. The purpose being to provide formal proof in a reductionist model
system, a cell or an animal model, that these proteins may be killers when
they aggregate and/or become dysfunctional. We think they are in humans
but it would be nice to confirm this in a model system where you can exclude
everything else...smoking, lifestyle, what you eat, etc. A lot of people
may not need that convincing proof, but it is necessary for drug discovery,
and the other very important consequence of such models is that they enable
scientist to then dissect the death pathway that is activated by these aggregates
so that you can come up therapies or better diagnostics.
For example, it may be possible to develop drugs that prevent the aggregation
of disease proteins, or that prevent subsequent or downstream effects of
aggregation that activate pathways leading to cell death. Even if such drugs
only delay, rather than arrest the disease process, they may be effective
therapeutically. Indeed, drugs that delay progression or improve the function
of remaining nerve cells in disorders such as AD would be very helpful.
To help people understand this concept, it is useful to think of how we
treat PD. A well-known patient with PD is Janet Reno, the attorney general.
She's had PD for about 5 years I believe, but I see her on TV periodically,
and I find it very difficult when I see her on television to appreciate
that she has PD. Although PD will continue to lead to the death of nerve
cells in her brain that result in motor deficits, and there is no therapy
that blocks the killing of nerve cells in PD, drugs like L-dopa are effective
therapeutically because they enable the remaining cells to pick up the function
of the lost cells. Since we know that it is necessary to lose about 90%
of neurons in the substantia nigra neurons before a patient shows the motor
deficits of PD, if you can make those remaining cells (i.e. 10% of the original
number of such nerve cells), fill in for the function of the lost cells,
then someone like Janet Reno may continue to function very well, and, when
this therapy is working optimally, to all outward appearances a patient
with PD may seem normal.
On the other hand, since there is no drug for AD that is the equivalent
of L-dopa, patients with AD, like former president Ronald Reagan, have a
neurodegenerative disease that progresses just like PD even with L-dopa
treatment, but these AD patients do not yet have the benefit of a therapy
that substantively improves the function of their remaining nerve cells
that are involved in memory or cognitive processes.
Q. Do you think the available drugs for AD, the cholinesterase inhibitors,
are serving a similar function in the case of treating AD symptoms?
A. It looks as though they have an effect, but the effect is rather modest.
Thus, I think there is consensus in the field, that we need better drugs
that target more encompassing or more fundamental aspects of AD.
Q. So, in terms of the hypothesis you are focusing on, is it tau as opposed
to A-beta? I know there is kind of an artificial distinction there.
A. Sure, and we're often put into the "tauist" camp, referring
to researchers who think that tau is the most important molecule to focus
on in AD. However, we are not so narrowly focused in our research, and we
also work on beta-amyloid, very seriously. We also work on alpha-synuclein.
We'd work on more if we had enough brain space to deal with all the molecules
but you can't work on everything and so we have to focus on a couple of
things. We believe there's not enough information at this time to rule out
tau or A-beta in AD. These are all very hot leads that very much need to
be pursued and we do pursue each of these molecules.
Q. So if you were pushed to have to say, "this is what I think the
most likely pathway is..."
A. Well, I think that's erroneous because AD is not one disease. The analogy
that I use is that 50 years ago we thought cancer was one disease, proliferating
cells make a cancer so aren't cancers all alike? Well, no. We know that
there are different genes that unleash this cell proliferation cascade and,
unfortunately, that means that a therapy for breast cancer is different
from prostate cancer, different from brain cancer, etc. I mean there are,
hopefully, strangleholds you can get on the cancer through one therapeutic
approach and the one that was in the news in the summer was angiogenesis.
Tumors, like all growing tissues, need a blood supply so if you can strangle
the blood supply you can theoretically destroy the tumor. So it would be
wonderful if we could find what I call the soft underbelly of the beast
we call AD to which we could direct a therapy and kill this horrendous animal
we call AD. We haven't identified that yet. It may require several different
therapies to block plaques, to block tangles. Certainly for sporadic AD.
Q. In that regard, to what extent do you think the disease, or diseases,
can be classified as primarily genetic?
A. Even in the genetic form of AD, there are three different genes that
harbor disease-causing mutations. Three entirely different genes, and, if
you count trisomy 21, three copies of chromosome 21, then there's yet another
genetic mechanism. So its very important for everyone to realize that effective
therapy for AD is not going to be a single silver bullet, at least I don't
think this will be the case. We're still waiting to see if the angiogenesis
notion with cancer, a single silver bullet to block blood vessel production,
will be effective, and maybe it will be and maybe we'll find something like
that for AD but I think in the interim we have to push on with the molecules
that you know are important in the pathology and the cell death process.
I think it would be foolish of me to say "tau is more important than
We even found in recent studies that over 50% of familial AD (FAD) brains
and over 50% of Downs syndrome brains have Lewy bodies. So now I would also
have to say that synuclein is playing a role in some way. How all of this
gets initiated in FAD, how you get plaques, tangles and LB with an FAD mutation
in presenilin I don't know. But there are certainly three bad proteins to
deal with and if we could sort out what mutations lead to each of these
protein aggregates that would be very very important, but this may not lead
to a therapy that's useful for sporadic AD. I think the scientific community
appreciates that we should think of Alzheimer "diseases" rather
than one disease and the public, I hope, will also appreciate that. There
are many forms of this dementia we call AD.
Q. What would you say is the most common criticism, if there is one,
of your major hypothesis or do you think there is something that still needs
to be done to address the particular idea that you have?
A. Even though we worked on tau and plaques and Lewy bodies, we were most
identified with tau and tangles and there is fashion in science just like
there is in other things. We felt it was important to be highly vocal, almost
combative, about not losing sight of tangles. It was not intended to be
self-serving or promoting our favorite hypothesis but the field deserves
to have the brightest minds in the field working on a diversity of issues.
What we felt was bad for the field, and bad for the patients, was to have
everyone jump on one bandwagon. I think the fastest way to the discovery
of cures or preventatives or agents that will delay disease is a broad front
that addresses a number of possibilities. There are some crazy ideas. There
are ideas that really haven't gone forward, like the notion of aluminum.
We've done some studies ourselves with aluminum, but its hard to see how
to take that forward to a therapy. Certainly there are a number of very
serious notions, anti-inflammation and a number of things that deserve to
be heard and deserve to be funded and pursued.
It was nothing malevolent, people just jump on a bandwagon and think
that by following presenilin 1 or presenilin 2 or APP they will get most
quickly to the cause. If I were an investment banker or stockbroker I would
want to invest in small caps and big caps, perhaps I would invest in the
Asian markets, not today perhaps, but 2 or 3 years ago. In other words,
you can't give a nickel to everyone but you can certainly do something better
than pursuing one molecule. We have been very strong advocates of tau because
we felt it was being underexplored. That has changed dramatically this past
summer with the discovery of tau gene mutations in frontotemporal dementia
and Parkinsonism linked to chromosome 17 (FTDP-17), a group of syndromes
some of which are very much like AD. Some of the families, in fact, that
are now known to have FTDP-17 due to tau mutations were initially considered
to have FAD. They look so similar to a clinician that for all the world
they would appear to be AD. Not only that, but to pathologists they also
have identical tau abnormalities and tau inclusions (neurofibrillary tangles).
So it's as close to sporadic AD, I would argue, as FAD due to mutations
in APP, PS1 and PS2 is to sporadic AD.
What is the argument that those hereditary diseases are not the disease
that Alois Alzheimer described? I think they are close enough and we know
there is a lot of phenotypic variability that shouldn't confuse us. We should
realize that it's plaques, tangles, now Lewy bodies, nerve cell death and
a declining cognitive or motor function that is the essence of these diseases
and whether they occur with an aphasia, a speech problem, or a psychosis
or what have you, we're talking about diseases that are sufficiently similar
that if you solve FTDP-17 I think you're going to solve AD at the same time.
Q. Its interesting to see how some of those in the amyloid camp are backing
off a bit.
A. Yes, there is nothing like a mutation to crystallize your thoughts. That's
the importance of genetics. The fact that the mutation absolutely causes
the disease makes it no longer tenable to argue that tau is not important.
I don't know that we were necessarily prescient, or what have you, we had
amassed ourselves a lot of data which convinced us, so we weren't able to
convince other people, maybe we're not very good in making the case...
Q. Well, I've been rather impressed at how conservative groups are in
terms of once they get an idea its very difficult to budge them.
A. So I think this is going to result...if you look at the neuroscience
program this year you'll find no sessions on tau. For the last 5 or 6 years
we have fallen into "AD: other" and I think next year you'll see
at neuroscience a lot about synuclein, a lot of tau and there'll be tau
session I, II and III because it's just got to be a focus.
Q. Now along that line, would you see tau as being early on in the cascade
because of the mutation?
A. I think you have to. What's the most mysterious, although there's lots
not known about the familial dementing conditions, is how to think about
linking tau and A-beta into a single pathological cascade and that could
be limited imagination on the part of scientists or, more likely, a lack
of sufficient biological understanding of these proteins. So we need to
know a lot more basic biology and that's not an intellectual sandbox for
scientists, it really is the basis from which springs forth new hypotheses
about how to look at stuff. You need to have more fundamental understanding
of the functions of these proteins and what they normally do and what the
dysfunctional consequences of those perturbations are for the cell. That
requires basic studies that will then lead to clinical applications, clinical
studies, drug discovery efforts and so forth.
Q. You mentioned before aluminum. Thinking now of other notions that
are still out there that would be considered even more marginal than tau
used to be, are there any areas that are being currently overlooked?
A. I think so. Once you have a mutation, it's easy to focus on what those
genetic mutations do to a protein. What has been hard for all of us to focus
on are the epigenetics, or environmental causes of AD such as head trauma,
which is a very robust epigenetic risk factor. It may interact with the
apoE4 allele but its still epigenetic. In women who are postmenopausal,
the lack of estrogen. Again, all women will go through menopause, some will
voluntarily take estrogens and some won't, but if you do you're going to
reduce your likelihood of getting AD by 40%. These are complicated issues
to pursue. In our own laboratory we are pursuing the head trauma risk factor.
We looked at this in normal rats as well as in transgenic mice expressing
wild-type and mutant APP. Our hypothesis was that head trauma would accelerate
AD plaque formation and perhaps induce tangles. None of those things have
happened but we've shown that head trauma will induce a massive loss of
neurons in the PDAPP mice, i.e. those that form plaques (as in AD) due to
the overexpression of mutant APP.