Posted 26 August 2006

E-mail discussion Printable version

Return to Discussion Text

John Q. Trojanowski and Mark P. Mattson led this live discussion on 22 October 2003 at 12noon EST to discuss the major commonality between many neurodegenerative diseases.

Note: The transcript has been edited for clarity and accuracy.

John Q. Trojanowski
Well, I will try to get things going by asking a question, which is: Do folks think that we underestimate the burden of misfolded proteins in neurodegenerative diseases with inclusions by relying too much on microscopy, since much of the abnormal protein accumulations may be in the form of oligomers that are not evident by microscopy, but could be detected biochemically?

Mark Mattson
One issue I am interested in is why dopaminergic neurons seem exquisitely sensitive to overload of the proteasome. The emerging findings suggest that simply increasing levels of wild-type synuclein is sufficient to cause PD. It seems as though the proteasome of dopaminergic neurons is sitting on the edge of a cliff with regards to the amount of ubiquitinated synuclein it can handle.

John Q. Trojanowski
In response to Mark, do we really know that dopaminergic neurons are so much more sensitive than other neurons to the accumulation of misfolded proteins?

Mark Mattson
We need some way of "titrating" abnormal protein accumulations and then monitoring the consequences with regards to various endpoints of interest—in my case, calcium regulation and oxidative stress.

Gabrielle Strobel
Can I put a question out to everyone, especially John and Mark: How common are double and triple neurodegenerative brain amyloidoses? Do you have a sense of that?

John Q. Trojanowski
As we look in greater detail, especially with biochemistry allowing formic acid insolubility to be a surrogate for what is likely to be an amyloid fibril-containing protein pool, I think double and triple brain amyloidoses may be the rule rather than the exception.

Gabrielle Strobel
Fascinating, John. So standard examination of pathological samples simply tended to overlook "other" amyloids? Certainly Kurt Jellinger has long pointed to overlapping pathologies among dementias, right?

John Q. Trojanowski
Using more traditional morphological criteria, AD is the most common triple brain amyloidosis, since there is Aβ plaque amyloid and tau tangle amyloid in all cases, and over 50 percent also have α-synuclein Lewy body amyloid.

Alexei Koudinov
Hello, everyone. But till now, people rarely called tau changes in AD "amyloid," right, John?

John Q. Trojanowski
Terminology has varied but there is consensus now, I think, that tangles, plaques, Lewy bodies, and other fibrillar deposits are all amyloids formed by different building-block proteins.

Alexei Koudinov
In biochemical/biophysical terms (β-pleated secondary structure, etc.), does tau chemistry in AD fit the amyloidosis definition? Literature avoided such definitions in the past, am I right?

John Q. Trojanowski
If you look at recent work from Goedert and Crowther in PNAS, you will find their paper with elegant data proving that tau fibrils are amyloids, and the same has been reported by many labs for α-synuclein fibrils.

Gabrielle Strobel
What could be the mechanisms underlying the convergence of more than one brain amyloid in many neurodegenerative diseases?

John Q. Trojanowski
In mutation-bearing people, it is a mutation, but in sporadic disease it is less clear what the causes of protein misfolding and amyloidosis are.

Mark Mattson
The mechanisms are likely multifactorial. For example, oxidative stress can promote abnormal folding of proteins and protein aggregation. Overload of the proteasome seems important, and, of course, the data from studies of disease-causing mutations have provided important clues—overproduction of the long form of Aβ in AD, impaired ubiquitin-mediated proteolysis of synuclein in PD, etc. As aging is the major risk factor for the "double and triple amyloidoses," a focus should be on the age-related molecular changes.

John Q. Trojanowski
I agree with Mark on possible mechanisms in sporadic and possibly familial neurodegenerative diseases with brain amyloid deposition.

Gabrielle Strobel
Mark, you are interested in effects of diet on aging and neurodegeneration. Do you have data on links between dietary compounds and protein misfolding/aggregation?

Mark Mattson
Gabrielle, not yet. However, Tuck Finch has recently shown that Aβ deposition is decreased in APP mutant mice on a reduced calorie diet. We have found that dietary restriction upregulates protein chaperones in neurons, which would be expected to enhance their ability to deal with damaged and abnormal proteins such as Aβ, tau, and synuclein. We will be testing the latter hypothesis in the coming months/years.

Edward Zamrini
Do we know 1) if/how much any of the novel, in-vivo amyloid markers detect intracellular vs. extracellular amyloid/fibrillar deposits and 2) the relative strength of detection of one type vs. another?

Gabrielle Strobel
I also wonder if this is another way to look at oxidative stress. I am often puzzled because it seems so obviously to spur neurodegeneration, but specific in-vivo links to, say, AD pathogenesis, are less clear to me.

John Q. Trojanowski
Oxidation could alter protein conformation leading to misfolding, and if sufficient amounts of the misfolded protein accumulate, amyloidosis could ensue. Indeed, folks like Chris Dobson might say, give me any protein and tell me how much you want converted into β-pleated sheet-containing fibrils and I can deliver them to you under the right, in-vitro conditions.

Marcos Marques
Mark, looking the other way around, how would protein misfolding, particularly amyloidosis, affect neuronal metabolism?

Mark Mattson
Marcos, we know that protein misfolding can affect neuronal metabolism, but the mechanisms are in most cases unclear. Prion proteins are good examples. Of course, patients with AD and PD have impaired cellular energy metabolism, but that this is the direct result of protein misfolding has certainly not been established.

John Q. Trojanowski
Misfolded proteins may acquire a toxic function and have deleterious consequences thereby, or by accumulating in cells, they could pull down other proteins that are taken out of action, like the misfolded protein, leading to several losses of function due to the "sidelining" of the proteins in "garbage heaps" inside cells.

Alexei Koudinov

Mark and Gabrielle, regarding the above point on oxidative stress, I think it is very important to look at proteins (that we call amyloid here today) as normal functional elements of brain chemistry, as I pointed out in my (prediscussion comment). Similarly, oxidative stress may well serve to modulate synaptic function and plasticity. This was proposed in several papers (see Kamsler and Segal., 2003 and Berezov et al., 2003). If so, all elements should be considered as one complex mechanism that we should attempt to understand.

John Q. Trojanowski
Again, I concur with Mark that factors which upregulate chaperones could protect cells from the toxicity associated with accumulations of misfolded proteins, be this amyloidosis or some other form of toxicity.

Mikolaj Pawlak
Regarding the relation between diet and AD, isn't it so that we are observing end results of diet applied for years and that's why it is difficult to observe a relationship between diet and disease progression in real time?

Gabrielle Strobel
Mark and John, all, there is a paper in Science this Friday about C. elegans living six times their normal life span, and being highly active (see related news story). They have a few mutations, mostly in insulin-related signaling. Does this suggest anything as to which signaling pathways should be checked for changes in human aging and neurodegeneration (although I am not aware of links between aging pathways and protein aggregation)?

John Q. Trojanowski
I am not certain which of the signaling pathways in the worms would be worth pursuing in human neurodegenerative diseases, but that is a good thought to consider.

Alexei Koudinov
Gabrielle, there is another recent paper (not in Science) which shows that a particular profile of lipoproteins is associated with extreme longevity. This may well be of importance in terms of a role for lipids and cholesterol in AD (see our ARF hypothesis page) and LP signaling (see Herz and Strickland, 2001).

Mark Mattson
Gabrielle, mutations in the insulin signaling pathway increase life span and stress resistance in C. elegans. Apparently, the mutations relieve suppression of a forkhead transcription factor, resulting in increased expression of antioxidant enzymes and perhaps proteins involved in preventing protein damage and/or removing damaged/misfolded proteins.

Gabrielle Strobel
Fascinating, Mark. Gene expression studies of aging humans are coming along, and consistently seem to show differences (downregulation) of genes involved in stress response, DNA repair etc.... All this calls for more work on chaperones. As far as I know, chaperone genes tend not to come up much in screens for genes involved in neurodegeneration in, say, Drosophila or worms. How could one study their role in these multiple amyloidoses better?

Pete Nelson
One question seems to be: Which are chickens and which are eggs?

Diego Forero
Pete Nelson's question is very important. What is the primary mechanism in the amyloidosis? It may be only the visible consequences of other underlying pathogenic pathways.

John Q. Trojanowski
As to the chicken and egg question, I think things may go in either direction, since if one is born with a mutation in tau, this determines subsequent formation of tau amyloid, so you consider tau amyloid the egg because it can come first, i.e., disease begins at conception. But in sporadic disease, there may be many chickens laying eggs that break, damaging tau and precipitating aggregation to form tangles. Sorry to overdo the chicken/egg analogy, but that was how the question was framed.

Pete Nelson
Sorry to use a riddle for a metaphor!

Gabrielle Strobel
Both riddles and metaphors are fair game, Pete!

Mark Mattson
Understanding the normal functions of synuclein, tau, and Aβ is important. It is of considerable interest, in this regard, that synuclein, tau, and APP are axonal proteins. Their normal functions in axons and presynaptic terminals may provide important clues as to the earliest events in disease pathogenesis, as well as to how the abnormal protein aggregates arise.

Gabrielle Strobel
Does this point to axonal transport, then, as one possible unifying theme of what goes awry?

Mark Mattson
Yep, axonal transport and synaptic vesicle recycling might be adversely affected early on in the course of the disease process.

John Q. Trojanowski
This is a good point on the possibility that in all of these diseases there could be a disruption of transport because the disease proteins bind or perturb motor proteins, disassemble microtubules, or physically block traffic when large amyloid accumulations formed by tau or α-synuclein develop in axons as dystrophic neurites or Lewy neurites.

Diego Forero
The topics pointed out by Dr. Koudinov are also very important. We may try to understand the possible pathogenic pathways in amyloidosis, taking into account that these proteins are key regulators of neural function, not only "bad" amyloids.

Alexei Koudinov
To Diego and Pete on primary factor: I believe that I attempted to address this question through the prism of functional role for "amyloid" proteins in normal synaptic function/plasticity (see comment on the ARF live discussion). For example, with regard to oxidative stress and normal/pathological amyloid biochemistry, an interesting sequence of events is proposed by A. Kontush (see abstract), which is that Aβ initially serves as an antioxidant, but its increased production as antioxidant leads to the peptide-aggregated pro-oxidative form.

Marcos Marques
Did anybody look at amyloid animal models for changes in parameters like diet and exercise as a means to reverse the neuronal metabolism imbalance, and at the molecular level, is it possible to determine, then, how amyloid affects neuronal metabolism?

Gabrielle Strobel
John, as an esteemed pathologist, what do you think of Larry Goldstein's argument that many of the swellings one commonly sees in neurodegenerative diseases are axonal blockages that induce, not accompany, damage?

John Q. Trojanowski
Mark, what has become of "synaptosis"? This was the concept that the disease protein may lead to programmed death of processes, and I think you and others proposed this possibility a while ago.

Gabrielle Strobel
John, do you mean a local apoptosis program in dendrites and terminals?

John Q. Trojanowski
Gabrielle, as you may recall from my recent commentary on the papers from the Brady and Goldstein labs, I was pleased to see the concept of axonal transport-induced degeneration extended from AD, where it has been around for over 10 years, to other disorders such as polyQ diseases, and I think it plausible that impairments in axonal transport could be drivers for degeneration in multiple aging-related neurodegenerative disorders.

Mark Mattson
The evidence that activation of apoptotic cascades occurs in synapses, axons, and dendrites is quite strong, and it is clear that these cascades can propagate to the cell body, culminating in death. It is also becoming evident that apoptotic cascades can have local effects on synaptic function and structural remodeling in the absence of cell death. The possible links between the amyloid proteins under consideration here and apoptotic processes remain to be determined, although we do know that Aβ can induce "synaptic apoptosis," at least in cell culture and synaptosome preparations.

Diego Forero
Here's a reference about the concept of synaptic apoptosis developed by Mattson et al., 1998

John Q. Trojanowski
One would assume a progression from oligomers, to protofibrils, to fully formed amyloid fibrils based on in-vitro studies, and this process could unfold anywhere in a cell. Indeed, we have shown by stereology (Mitchell et al., 2000) that only five percent of the area occupied by abnormal tau immunoreactivity in AD is in tangles, while 95 percent is in the dystrophic processes, which means that it is surprising that tangles correlate with dementia because most of the tau pathology is in processes which are not normally counted in correlative studies.

Gabrielle Strobel
Charlie Glabe reported an antibody that appears to recognize mid-size oligomers of a number of fibrillogenic proteins (see related news story). Would that be a useful tool to study the synergy you propose between different aggregation-prone proteins? Does the antibody stain in-vivo sections of different disease brains?

Alexei Koudinov
I was recently reviewing literature on a functional role of Aβ/APP (with regard to a unifying role for cholesterol in synaptic degeneration), and found Askansas et al., 1992 and Torroja et al., 1999. Do the above imply that we miss a more general point while talking about a neurodegeneration commonality—a role for proteins in synaptic machinery?

Gabrielle Strobel
John and Mark, does your thinking say anything about where along the way of aggregation the damage happens to the synapse? Oligomers? Protofibrils? Fibrils? Do you consider this question important at all?

Alexei Koudinov
Gabrielle, we come ourselves to the conclusion that amyloid fibrils can damage synaptic plasticity and that diffuse amyloid has no such effect. For me, the major question is what causes the change in Aβ biology. When we understand this, we will be able to reverse amyloid by affecting the primary cause.

Diego Forero
Some recent references about the possible neuroprotective roles of amyloid proteins: (Plant et al., 2003 and Arendt et al., 2003).

John Q. Trojanowski
And this "synaptic apoptotic" damage could be compounded by impairments in axonal transport caused by an accumulation of insoluble tau or α-synuclein, because transport of key synaptic proteins would not occur normally, nor would trophic factors picked up at terminals be transported back to the perikaryon to sustain the viability of affected neurons.

Gabrielle Strobel
What is the role of ubiquitin-proteasome degradation in this? It would seem to be a candidate for a shared mechanism, as it pops up in some way in all these diseases. But I see no overarching theme yet. Have I missed it? One interesting lead might be that Mike Ehlers studies proteasome degradation in dendrites as a mechanism involved in activity-dependent turnover of postsynaptic proteins.

Mark Mattson
Ubiquitin-mediated protein degradation seems to be at the heart of the problem in Parkinson's disease as parkin is an E3 ligase and synuclein a substrate. In the case of Alzheimer's, Aβ, and tau, the importance of the proteasome is less clear.

John Q. Trojanowski
Amyloidosis is a product of several mechanisms, including the oversupply of substrates for amyloidogenesis and the diminished clearance of substrates allowing accumulations that precipitate into amyloid fibrils under the appropriate in-vitro conditions.

Davi Bock
John, regarding distribution of abnormal tau: Perhaps the tangles are more inflammatory than the dystrophic processes. One oft-neglected clue is that APP has an iron response element (Rogers et al., 2002) and an IL-1 element (Rogers et al., 1999) in its 5' UTR. I wonder if a unifying feature of these amyloidogenic proteins is that they are upregulated by inflammatory processes, and so when they themselves instigate an inflammatory process (due to stochastic misfolding which increases with age, for example), a positive feedback loop is initiated?

John Q. Trojanowski
In response to your query, oxidate/nitrative stress could be another chicken/egg phenomenon, in that it may increase with the accumulation of misfolded proteins, or if increased for other reasons, it could contribute to amyloidogenesis.

Alexei Koudinov
Diego's reference reminded me of another major observation: that PHF-like tau change occurs normally during the short ontogenic period of intense (and membrane cholesterol-demanding) nerve growth (see Koudinov and Koudinova 2003). Can we call this condition amyloidosis? Nature developed it to serve neural/synaptic function. In the disease, therefore, this machinery may serve to compensate synaptic failure. Can it be called a pathological event? Or compensation would be a better definition. This will be available soon as a peer-reviewed publication.

Mark Mattson
The developmental changes in tau involve hyperphosphorylation and not aggregation of tau into filaments. It is clear that hyperphosphorylation of tau, pre se, is not harmful to neurons, as it occurs normally in development. This is important, as it tends to dissociate phosphorylation from the pathogenic process in AD and related tauopathies.

Alexei Koudinov
Mark, there is certainly a borderline between natural normal compensation and a disease-locked condition. Therefore, one should better and strictly define new amyloid proteins, so there will be no confusion (see the beginning of this discussion).

John Q. Trojanowski
Mark, hyperphosphorylation or any phosphorylation of tau is not needed for in-vitro tau amyloid fibril formation, and the tau gene mutations argue that abnormal tau phosphorylation is not the most upstream event in familial tauopathies. However, since the increasing phosphorylation decreases microtubule binding, the excess phosphorylation of tau would disengage it from microtubules and this will destabilize microtubules to impair axonal transport, while leading to pools of unbound tau that could reach concentration thresholds which result in tau fibrillization.

Mikolaj Pawlak
If you can't stop amyloidogenesis, can you at least slow it down by limiting substrates? If yes, then how?

John Q. Trojanowski
Yes, and that is the purpose of β-secretase inhibitor therapies that are intended to reduce the levels of Aβ peptides, which are the substrates for amyloid fibril formation. Alternatively, vaccine therapies, and gene therapy to increase expression of neprylisin, would be other avenues of therapeutic intervention to reduce Aβ plaques in AD.

Mikolaj Pawlak
I see. Thanks. How about physiological processes influencing this pathway?

Gabrielle Strobel
John and Mark, can you tell us more about these aggregation-busting compounds you mention in the article? Sorry for the barrage of questions—this is so interesting!

Alexei Koudinov
John, with regard to Aβ-lowering therapies, one should be sure that no normal pathway is affected—a subject deserving study.

Gabrielle Strobel
The new drug Velcade is a proteasome inhibitor and now in trials for prostate cancer. I wonder what would happen if it got into the brain of people with breaks in the BBB?

Mark Mattson
Yes, if the Velcade gets into the brain, it would likely be bad for neurons. Of course, if you have life-threatening cancer, you might not be too worried about increased risk of neurodegenerative disorders.

Gabrielle Strobel
Mark, I totally agree. Indolent prostate cancer, though, is a different beast from active multiple myeloma....

Diego Forero
Two interesting articles about proteasome dysfunction and AD: Hope et al., 2003 and Keck et al., 2003.

Gabrielle Strobel
Just to rock the boat at the end of the hour: I find the notion of a shared mechanism the least convincing in AD because of the spatial separation between tangles and plaques. Do you think that Aβ misfolding and fibrillization begins intraneuronally, where perhaps its aggregation could interact with tau and α-synuclein?

Pete Nelson
My own $0.02, vis-à-vis Gabrielle's observation: Neuritic plaques seem to me to represent an important nidus of pathology, as (extracellular) amyloid plaques are directly apposed to (intracellular) neurofibrillary pathology, and degenerating neurites. Somehow, one is very directly interacting with the other!

Gabrielle Strobel
Yes, but Pete, how can they interact "very directly" when there is a cell membrane in the middle?

Pete Nelson
The cell membranes are compromised in neuritic plaques, as shown early on by Terry, Wisniewski, and other ultrastructural microscopists.

John Q. Trojanowski
On rocking boats, I would emphasize that what we see in the way of amyloid deposits through a microscope is the tip of the iceberg, in my view, since I expect that the AD brain is awash in variable levels of Aβ, tau, and often α-synuclein oligomers, so interactions could well take place outside the field of view of a microscope.

Angela Biggs
What if accumulation was due to the neuron being in the growth mode of the axon...assuming all three—tau, synuclein, and APP—were involved in growth...and perhaps the nerve became stuck in the growth mode? Decrease the amount of substrate buildup by learning how to get the nerve out of the growth mode? Does that make sense?

Diego Forero
Interesting point, Angela. I think that it may be the case, although now there is only indirect evidence for it.

Angela Biggs
I've been working on an idea whereby serpins stop growth and serine proteases trigger growth...and I find it suggestive that AβPP is a serine protease receptor—nexin. Then, insertion of neuroserpin stops the extension of axons, right?

Gabrielle Strobel
Can I ask another question about drug discovery based on your hypothesis, John and Mark? Does it open new avenues to make stronger drugs than vitamin E out of the oxidative damage knowledge?

Mark Mattson
Antioxidants continue to hold therapeutic potential and many labs are working to identify novel antioxidants which easily enter the brain and scavenge radicals. It's hard to say whether they will have a major impact on the disease process, although they are likely to have some benefit with few side effects.

Pete Nelson
I have to run. Thanks so much to you, Gabrielle and ARF, and much thanks to Drs. Mattson and Trojanowski, and the others. I enjoyed the hour!

John Q. Trojanowski
To paraphrase Mae West: So many fascinating questions and hypotheses, so little time, and it is the end of the hour for me, so I will send greetings to all and thanks for the lively chat!

Gabrielle Strobel
John, on rocking boats: Aha, thanks. I am beginning to see this majority view now, that there is this sea of different aggregating species, few of which are visible with a microscope.

Alexei Koudinov
I agree with John about the iceberg tip, and look forward to more great data by scientists in many related fields. Thank you to all for the discussion and to ARF for organizing/holding it.

Gabrielle Strobel
Let me thank you all for coming and for this fascinating discussion. We clearly need to revisit this topic.

Diego Forero
Thank you again, ARF team, congratulations!!

Mikolaj Pawlak
Thank you.

Edward Zamrini
For John, perhaps for a later time regarding the tip of the iceberg. So, could any of the new amyloid markers detect levels of oligomers heretofore undetectable by microscopy? Bye, everyone.

Gabrielle Strobel
One hopes PIB will, and new antibodies. So, Mark, should we go on calorie restriction in the meantime?

Mark Mattson
Gabrielle, thanks for the opportunity. John, thank you for your efforts on the special issue of the journal and this online chat. Hope to see you in New Orleans. Gabrielle, yes, it's very likely that each of our brains would benefit from smaller or less frequent meals. In addition to helping your neurons deal with damaged proteins, dietary restriction upregulates expression of neurotrophic factors, particularly BDNF. Our data suggest that BDNF mediates several beneficial effects of dietary restriction in the brain, including neuroprotection and stimulation of neurogenesis.

Diego Forero
Bye, everyone.

Gabrielle Strobel
Bye, everyone, and please come back.