Introduction

This live discussion is an update of last year's initial Live Discussion of axonal transport as an underlying factor in neurodegeneration.

Jorge Busciglio and Scott Brady led this live discussion on 17 June 2003. Readers are invited to submit additional comments by using our Comments form at the bottom of the page.

Transcript:

Jorge Busciglio and Scott Brady led this live discussion, an update of last year's initial Live Discussion of axonal transport as an underlying factor in neurodegeneration.

Participants: Akihiko Takashima, Institute of Physical and Chemical Sciences (RIKEN), Japan; Peter Klein, University of Pennsylvania; Leo Kim, University of Pennsylvania; Frank LaFerla, University of California, Irvine; Changiz Geula, Beth Israel Deaconess Medical Center in Boston; Jorge Busciglio, University of California, Irvine; Scott Brady, University of Illinois at Chicago College of Medicine; Martha Stokely, University of Florida Gainesville; Nikolaos Tezapsidis, Columbia University; Gunnar Gouras, Cornell University; Gabrielle Strobel, ARF; Keith Crutcher, University of Cincinnati; Diego Forero, National University of Colombia, Bogota.

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

Gabrielle Strobel
Hi, and welcome everyone. I am the editor of Alzforum and pleased to moderate today. Perhaps Jorge could start things off with a brief restatement of his hypothesis?

Jorge Busciglio
We are studying the role of PS1 in axonal transport due to its interaction and regulation of GSK-3b activity, which modulates kinesin phosphorylation and anterograde axonal transport.... PS1 mutations appear to alter GSK-3b and kinesin phosphorylation states.

Changiz Geula
Jorge and Scott, is there evidence that GSK-3b directly phosphorylates kinesin in vivo?

Scott Brady
Yes. We published an extensive study in EMBO J last year. GSK-3 specifically phosphorylates kinesin light chains, which then recruit the chaperone HSC70 to the vesicle and lead to removal of kinesin from the vesicle. (Morfini et al., 2002).

Jorge Busciglio
GSK-3b phosphorylates kinesin light chain.

Gabrielle Strobel
What causes the proposed imbalance of kinase/phosphatase activity in aging humans? I understand you propose that FAD mutations disturb GSK-3 activity, but how about the majority of nonfamilial LOAD cases?

Jorge Busciglio
Excellent question Gabrielle, but a difficult one to answer....

Akihiko Takashima
Gabrielle, we have some results on GSK-3b protein being upregulated during aging (Planel, 2002, DrugDevResearch,56,491). PP2A mRNA levels are also downregulated during aging and in AD....

Gabrielle Strobel
Akihiko, what is PP2A? Does it regulate GSK-3 levels?

Akihiko Takashima
Gabrielle, PP2A (protein phosphatase 2a) regulates GSK-3b activity by dephosphorylating S9.

Scott Brady
Akihiko, other phosphatases can also dephosphorylate GSK-3. We have evidence of a role for PP1.

Akihiko Takashima
Scott, there was a nice article on interaction of PP1 and GSK-3b in JBC recently.

Peter Klein
Scott, we published a paper recently in JBC on PP1 dephosphorylation of serine 9 (Zhang et al., 2003).

Akihiko Takashima
Peter, that's a very nice paper!

Peter Klein
Thanks for the compliment.

Scott Brady
Gabrielle, there may be a variety of pathways to misregulate GSK-3.

Gabrielle Strobel
What are those pathways?

Scott Brady
For example, CDK5 activity is required to keep GSK-3 off. We also have evidence that tau may play a role in this.

Gabrielle Strobel
Are any of the several CDK5, tau, or GSK mouse models suitable to addressing the question of when axonal transport begins to change?

Jorge Busciglio
Frank LaFerla
Jorge, your study on PS1 mutations altering GSK-3b and kinesin phosphorylation states was done on embryonic neurons, correct? If so, does this not mean that disruption of axonal transport is a very early event?

Jorge Busciglio
Frank, we used embryonic neurons that differentiate in culture; it is difficult to say when the impairment in transport starts....

Changiz Geula
Scott, what is the evidence for the participation of tau?

Scott Brady
The evidence for tau is several-fold and some of it is unpublished. However, GSK-3 will bind to tau, and in some conformations we have evidence that tau filaments can directly activate it.

Jorge Busciglio
First, let me tell you that postmortem intervals may have confounding effects on kinase/phosphatase activities in AD brains....

Akihiko Takashima
Jorge, yep, GSK-3b activity is impossible to determine in postmortem tissues because serine 9 is extremely labile.

Nikolaos Tezapsidis
Kinesin-driven anterograde fast and slow axonal (and to a lesser degree dendritic?) transport may be equally important as dynein-driven retrograde dendritic (and to a lesser degree axonal) transport of both cargo vesicles and multimeric proteins.

Jorge Busciglio
Yes, Nikolaos. What is the evidence for altered retrograde transport in AD?

Nikolaos Tezapsidis
Jorge, there isn't much; we've started doing some studies. What I was suggesting is that we should look into it more....

Jorge Busciglio
Absolutely, Nikolaos, we will be looking at transport in KIm146V cultured neurons....

Nikolaos Tezapsidis
Jorge, perhaps we can collaborate.

Changiz Geula
Jorge and Nikolaos, disruption of retrograde transport in AD would be consistent with a host of anatomical abnormalities. Most neurotrophins get to the cell body via retrograde transport.

Jorge Busciglio
Yes Changiz, and as you know, there are a number of motor abnormalities in AD patients....

Scott Brady
Nikolaos and Changiz, regarding retrograde, the GSK-3 effect is purely on anterograde transport, but we have just submitted a manuscript with the Sisodia lab that shows a decrement in the transport of Trk [tyrosine receptor kinase] receptors with FAD PS1 mutants. This will lead to reduced levels of receptors to bind neurotrophins and return them to the cell body.

Nikolaos Tezapsidis
Scott, can I have a copy?

Peter Klein
Jorge and Scott, have you tried any GSK-3 inhibitors (other than the CREB peptide) to test whether they affect kinesin light chain (KLC) phosphorylation or FAT?

Jorge Busciglio
Peter, how does GSK-3a affect APP processing?

Peter Klein
Jorge, we are working on it. We don't know yet, but it does not appear to be through a direct effect on g-secretase activity.

Akihiko Takashima
Jorge, according to Peter's paper, decrease in GSK-3b would affect GSK-3a activity and vice versa....

Scott Brady
Peter, we were aware of your paper, but I was referring to a pathway we have defined that involves PP1 in neurons. Nikolaos, once the paper is through review, I will send you a preprint if you send me an e-mail.

Nikolaos Tezapsidis
Will do. I would also like your comments on our paper describing the PS1/CLIP-170 interaction and how it may relate to your study.

Scott Brady
Nikolaos, will do. We don't see any direct connection, but cannot rule it out.

Akihiko Takashima
Gabrielle, does anybody know if axonal transport is inhibited in GSK-3b transgenic mice?

Gabrielle Strobel
Akihiko, I don't know, but I think it ought to be checked.

Diego Forero
I want to contribute a recent reference that shows alterations in axonal transport in AD patients in vitro—Dai et al., 2002.

Jorge Busciglio
Peter, are there GSK-3a consensus sites in nicastrin, Pen-2, or Aph-1?

Peter Klein
Jorge, not in Pen-2 or Aph-1—not sure about nicastrin.

Changiz Geula
Jorge and Scott, in terms of the relationship between disruption of anterograde transport and Ab, the disruption should cause accumulation of APP in the cell and thereby result in increased intracellular Ab. Is this your interpretation? Do you see increased Ab in your cultured cells?

Jorge Busciglio
Yes, Changiz, that is part of our hypothesis. A recent paper by Hoshi et al. also shows that Ab neurotoxicity increases GSK-3 activity, so it could be a positive feedback loop.... PS1 mutations increase Ab production, and the Met146Val is no exception....

Gunnar Gouras
Jorge, we tended to see Ab accumulation in distal processes—how could this fit in?

Jorge Busciglio
Gunnar, decreased transport of mitochondria may affect processing at the tips....

Akihiko Takashima
Jorge, this was demonstrated by our group in 1993-94....

Nikolaos Tezapsidis
Scott and Jorge, we have shown that the FAD mutations on PS1 are associated with a tighter binding of PS1 to CLIP-170, and that the two proteins interact prior to PS1's cleavage and presumably prior to its incorporation into the g-secretase multimeric high molecular weight complex. Whether the tighter binding of PS1/APP vesicles at the microtubule tip facilitates the assembly process of this complex, which generates Ab, or whether it makes it harder for the cargo to move towards a degradation pathway (endosomes) remains to be determined.

Scott Brady
Nikolaos, we need to know more about the nature of the binding of GSK-3 to PS1 before we can make a connection with the CLIP-170 result.

Nikolaos Tezapsidis
Scott and Jorge, the constant motion of cargos is partially regulated by the ability of those motors to attach both to the cargo vesicles and the microtubule track. CLIP-170 is believed to serve such a regulatory role, acting as a linker of endosomes to the plus end of the microtubules.

Scott Brady
Nikolaos, the role you propose for CLIP-170 has been suggested, but not demonstrated directly. We don't see much evidence for this kind of regulation of kinesin-based motility in neurons.

Nikolaos Tezapsidis
Scott, our data are consistent with the initial proposed role for CLIP-170; however, we are the first to explore this in pathways related to AD.

Gabrielle Strobel
Nikolaos, would it be possible to do your experiments in primary neurons, in addition to neuroblastoma lines?

Nikolaos Tezapsidis
Gabrielle, that, we were planning to do next....

Peter Klein
To anyone, do the FAD mutations in PS1 affect GSK-3 binding?

Jorge Busciglio
Peter, we will be mutating the binding domains to answer that question.... Takashima's results indicate that PS1 mutations indeed affect binding.

Akihiko Takashima
Peter, we did it. It seems to be the same affinity, regardless of the mutation.

Jorge Busciglio
…or maybe not?

Akihiko Takashima
Jorge, sorry, I just checked my PNAS paper: GSK-3 binding to PS1 changes with PS1 mutations, but no difference in PS1 and tau binding. You were right :-).

Jorge Busciglio
I knew it, Dr. Takashima!

Akihiko Takashima
Everybody, there were questions proposed to be discussed in the chat by Jorge and Scott. Could we go through them one by one?

Gabrielle Strobel
Thanks, I was waiting for a slow moment to feed them in, but the chat is so lively I have not, so far, wanted to cut it off....

Diego Forero
Dr. Busciglio, what are the results with NT2 cells transfected with PS1 constructs? I ask because your recent article in J Neurosci didn't show the effects on neurite outgrowth in differentiated hNT neurons.

Gabrielle Strobel
As far as I know, GSK-3 has been a potential drug target for a long time, but the work never really took off. Is this now changing with Peter's, Akihiho's, and Jorge/Scott's work?

Peter Klein
Gabrielle, GSK-3 has been a drug target for quite a while because of tau phosphorylation as well as potential applications in diabetes and bipolar disorder. I hope attention is rekindled on targeting GSK-3, especially a, to inhibit APP processing.

Gabrielle Strobel
Peter, perhaps your experiments would lend themselves to the development of a suitable assay for compound screening?

Akihiko Takashima
Peter, according to your paper, inhibiting either GSK-3 will affect the other, so how can you have a GSK-3a-specific inhibitor?

Peter Klein
Akihiko, that is an excellent point. However, we don't actually know the magnitude of GSK-3 inhibition based solely on the phosphorylation, since we do not know what fraction of total GSK-3 is phosphorylated. It might be possible to inhibit GSK-3a without completely inhibiting GSK-3b. Gabrielle, I agree. It would be great to identify GSK-3a-specific inhibitors.

Gabrielle Strobel
Peter, I am asking also because Alzforum is thinking about what sorts of information resources we could compile to support academic researchers who are venturing into drug discovery, or at least target validation, themselves.

Jorge Busciglio
Peter, how specific is kenpaullone?

Nikolaos Tezapsidis
Peter, lithium can be used...also has been used for other brain disorders though (bipolar disorder and schizoaffective disorder).

Peter Klein
Nikolaos, yes, perhaps lithium can be used, as it is for bipolar disorder. A drug with fewer side effects would be preferable in an older population, though.

Gabrielle Strobel
Peter, this has nothing to do with the fact that many AD patients have psychotic symptoms, however? Wrong?

Gunnar Gouras
Jorge and Scott, could you explore more whether APP/Ab are intermediates or unnecessary for your PS/GSK effects?

Scott Brady
Peter, as you point out, there are other mechanisms for regulating GSK-3 activity, including various binding proteins. Much remains to be understood about the microscopic regulation of GSK-3 in different cell types.

Jorge Busciglio
Peter, again, how does kenpaullone look from a therapeutic point of view?

Peter Klein
Jorge, kenpaullone does not distinguish between a and b, and also inhibits CDKs. However, in a study from Phil Cohen's lab, where they examined many protein kinases, there was no overlap between lithium-sensitive kinases (which, in any event, required significantly higher concentrations of lithium) and kenpaullone-sensitive kinases.

Jorge Busciglio
Peter, are you planning lithium treatment and behavioral experiments in the mouse model?

Diego Forero
I believe there are many things to study with lithium as an inhibitor of GSK-3.... There are many unresolved questions

Changiz Geula
Jorge and Scott, what is the magnitude of the presenilin effect on axonal transport? To be a therapeutic target, the effect must be shown to be substantial.

Jorge Busciglio
Changiz, the effect might be chronic and subtle, and impair neuronal homeostasis in the long term. I'm not sure you need a substantial effect; that may not be compatible with survival....

Scott Brady
Changiz. as we discussed in the paper, modest changes in transport may have long-term consequences by reducing the efficiency of the transport and leading to increased vulnerability of the neuron. Transport declines with age and, therefore, the ability of the neuron to maintain itself is reduced. The PS1 effect reduces that still further. The longtime course for development of pathology suggests that relatively modest changes in transport may be highly significant.

Peter Klein
Scott, right you are. Jorge, kenpaullone is being studied as an anticancer therapy, but I do not know offhand if it has been used clinically or what its overall toxicity is. Diego, I hope lithium keeps us in business for a long time.

Diego Forero
Of course, Peter.

Akihiko Takashima
Jorge, we are trying to publish a paper (with great difficulties) demonstrating that lithium can rescue the Ab-induced tau pathology and memory loss in tau Tg mice....

Peter Klein
Jorge, we are studying behavior in lithium-treated mice right now, but more from our interest in bipolar disorder.

Diego Forero
Dr. Takashima, this paper you're talking about sounds very interesting.

Akihiko Takashima
Diego, thank you. We talked about it in the last AD conference....

Gabrielle Strobel
Dr. Takashima, very interesting. Are there differences of lithium effect in tau-APP double transgenic vs. just APP transgenic? This could separate effects of lithium on Ab generation vs. GSK3/tau.

Akihiko Takashima
Gabrielle, we are doing it. Our results so far are with Ab injection....

Gabrielle Strobel
Oh, you use your tau transgenics and inject Ab to make the pathology worse? That is interesting, too.

Akihiko Takashima
Gabrielle, yes, as published by Gotz in Science (see ARF related news story).

Nikolaos Tezapsidis
Gunnar, Ab can be accumulated within neurons at the synapses either by deficient export or hyperactive import if it comes down to kinetics. Alternatively, a mere overproduction/delivery to the neuronal tips or deficient neuronal clearance. Jorge, Scott, any comments on the similarity of KO and mutant PS1 effect vs. amyloid cascade hypothesis.

Jorge Busciglio
Nikolaos, KO and PS1 mutants' similar effect suggest that mutations impair the ability of PS1 to keep GSK-3B activity down (at least when KLC is the target). In that sense, amyloid production may be subjected to a different type of regulatory mechanism.

Nikolaos Tezapsidis
Jorge, this would be consistent with a loss of beneficial function for PS1. This agrees with our data derived from the neuropathological examination of AD brains that showed preservation of neurons with PS immunoreactivity.

Jorge Busciglio
Nikolaos, that is a very interesting observation.

Nikolaos Tezapsidis
Jorge, this raises another very interesting point in this post-genomic era. Proteins and their biological activities, which ultimately confer the clinical phenotype, are much more difficult to study. A higher level of creativity and ingenuity than technology can offer will be required....

Martha Stokely
Scott, perhaps you would like to comment on the possibility of multiple insults to axonal transport in addition to aging.

Jorge Busciglio
Martha, many different insults may converge in transport deficits, and aging may be one of the primary factors affecting transport by increases in oxidative stress and energy deficits....

Scott Brady
Martha, diseases of aging always involve compromises in a wide range of cellular activities. Peter, at the time we did the study, we lacked the tools to test that rigorously, but the availability of GSK-3a-specific probes will allow us to see whether there is a difference. I suspect that both will be able to phosphorylate KLC. It is probably differential regulation and differential compartmentation that determines whether GSK-3a or b is responsible.

Peter Klein
Scott and Jorge, have you looked at GSK-3a phosphorylation in PS KO or KI settings? And do you know if GSK-3a interacts with or phosphorylates KLC?

Diego Forero
Jorge, and what about the other binding proteins of PS1 (Notch, cadherins) and axonal transport? Have you looked for it? It would be very interesting....

Jorge Busciglio
Diego, we have not looked into that yet....

Nikolaos Tezapsidis
Jorge, and even though we should in a "funny" way consider ourselves lucky to have in our hands the FAD-linked mutations, it is still debatable whether we have a safe target.

Jorge Busciglio
Nikolaos, I agree.

Gabrielle Strobel
Many APP transgenics are now being made available by Jackson labs. Alzforum has links to that resource.

Jorge Busciglio
Dr. Takashima, have you worked further on the PS1-GSK-3B binding region?

Akihiko Takashima
Jorge, unfortunately, no.... We focus on tau.

Gabrielle Strobel
Would anyone like to take on question 4 of the backgrounder? How could the relative roles of Ab generation and transport disruption by FAD PS1 be assessed? Video microscopy of FAD transgenic mouse neurons? What other methods are suitable?

Jorge Busciglio
Yes, Gabrielle, video microscopy on living cultured neurons is what we plan to do to directly answer that question.

Diego Forero
The most specific and sensible available approach.

Gabrielle Strobel
Jorge, can this be combined with clever calcium imaging agents? I wish more people incorporated those in their AD-related microscopy.

Jorge Busciglio
Gabrielle, also classic transport studies in vivo will be important to determine the nature of the transport defect....

Gabrielle Strobel
Can confocal in living mice reveal such things? Or injecting labeled stuff and measuring how much got to the target in mouse models at different time points?

Diego Forero
If you consider Dai et al., 2002, it will be possible to do transport studies in patients with PS1 mutations.

Jorge Busciglio
Gabrielle, calcium imaging may be important since PS1 mutations appear to deregulate calcium homeostasis. I think there may be ways we could use both techniques simultaneously.

Gabrielle Strobel
I think Frank's department is a good home in that regard. There is also one talented Bernardo Sabatini who is interested in such methods. I'd be happy to introduce you offline.

Jorge Busciglio
Gabrielle, sure, thank you!

Gabrielle Strobel
Let's address another background question. Here's number 1: The amyloid cascade hypothesis still draws majority support in the field as the predominant, if incomplete, explanation for AD. Can we integrate the axonal transport and the amyloid hypotheses? Do they fit together and how?

Diego Forero
The new amyloid hypothesis (amyloid and synaptic plasticity) could be related with axonal transport very easily....

Leo Kim
Diego, could you please explain more on the topic?

Diego Forero
Hi, Leo, as shown by Kamenetz et al., 2003, amyloid is a key regulator of synaptic function, and as shown by Busciglio et al., changes in PS1 and axonal transport regulate the transport of APP....

Leo Kim
Thank you, Diego.

Nikolaos Tezapsidis
Gabrielle, our paper also supports that notion. We've observed a reduction in Ab production and Ab uptake by intercepting the interaction of PS1 with CLIP-170, a microtubule-interacting protein, implying transport.

Jorge Busciglio
Gabrielle, I think that both the amyloid hypothesis and the impaired axonal transport are perfectly compatible and may both be related. It is clear that Ab will impair transport, and impaired transport may enhance Ab production.

Scott Brady
Gabrielle, axonal transport helps to establish a steady state for a wide range of neuronal functions. Disruption of transport alters those homeostatic mechanisms. The result is that one can have misregulation of many processes as a downstream consequence.

Akihiko Takashima
To answer Gabrielle's question on amyloid hypothesis and axonal transport: Braak and Braak investigated Ab deposition and neurofibrillar changes in thousands of brains at different ages. According to the study of Braak, neurofibrillar changes in the entorhinal cortex precede Ab deposition. Yet, neurofibrillar changes in the limbic and cortex regions follow Ab deposition. The entorhinal stage (Braak stage I, II) marks normal aging, while the limbic (Braak stage III, IV) and the isocortex stages mark diseased states; and neurofibrillar changes in the entorhinal cortex occurs in 100% of the aged population. Thus, NFTs in the entorhinal cortex occur during brain aging, and Ab accelerates NFT formation in the limbic and isocortical regions, resulting in AD. Therefore, disruption of axonal transport might occur during aging and before Ab deposition because of tau dysfunction.

Changiz Geula
Dr. Takashima, would you equate disruption of axonal transport with tangle formation? Can these be independent processes?

Akihiko Takashima
Changiz, from in-vitro and in-vivo studies, tau accumulation in the cytoplasm disrupts axonal transport by the inhibition of association between kinesin and microtubule. Free tau is preferable to phosphorylated.

Jorge Busciglio
Dr. Takashima, do you think that tau phosphorylation precedes detachment of the MT?

Akihiko Takashima
Jorge, yes, I think that tau phosphorylation impairs its binding to MT.

Scott Brady
Gabrielle, the age of onset and slow development of pathology may mean that the two phenomena are not directly related. The experiments that Jorge proposes will help get at that. As I noted above, there may be many things going awry because of the disruption of homeostatic mechanisms in the neuron. This is why the correlation between some pathological features and some of the clinical symptoms is poor. A more detailed analysis of transport will help with this, as Jorge says.

Diego Forero
I am in accordance with Scott. Gabrielle, as shown by Stamer et al., 2002, manipulations that alter the axonal transport in a nonspecific manner have profound effects on cell function.

Gabrielle Strobel
Scott, this appears to call for quite nonspecific therapeutics aimed broadly at maintaining/restoring normal transport flows. Is anything like that conceivable today?

Scott Brady
Gabrielle, in the case of AD, the various strategies for modulating GSK-3 activity may be helpful. The key is not to eliminate activity. These kinases have too many functions for that to be a good idea. The key may be to adjust basal activity while not disrupting other regulatory mechanisms. I think this is, in principle, possible, but will require a better understanding of GSK-3 regulation and compartmentation.

Peter Klein
Scott, I agree completely with your view that much more information on subcellular distribution of GSK-3a and b is needed.

Keith Crutcher
Folks, I have been following the discussion with some interest and confusion. I am not a transport aficionado, but I am having a hard time understanding the underlying hypothesis and how this relates to the vast majority of AD cases. Do any of the results suggest that transport defects are a cause rather than a consequence of AD pathology? Is there a testable hypothesis here?

Jorge Busciglio
Keith, synaptic dysfunction, dystrophy, and neuronal loss can all be the direct result of transport defects.

Akihiko Takashima
Jorge, yes, I agree.

Keith Crutcher
Jorge, fair enough, but why would one expect to see the regional pathology of AD from such changes?

Jorge Busciglio
Keith, many insults including Ab toxicity and tau mutations may converge in transport deficiency. Some neurons are more vulnerable than others....

Keith Crutcher
Jorge, okay, I can accept that, but I am wondering what hypothesis we can generate to pull these observations together.

Jorge Busciglio
Keith, the regional pathology may also be related with differences in Ab processing in distinct brain regions.

Changiz Geula
Keith, the pathology of AD, at least tangle pathology, seems to affect primarily neurons with very long axons, which also happen to be connected with the cortex. Transport defects would be consistent with problems in long axons. However, widespread cortical connectivity appears to be another predisposing factor.

Keith Crutcher
Changiz, Good point, although there are neurons with even longer projections such as Betz neurons which I don't think show major changes...or am I wrong on that?

Changiz Geula
Keith, you are correct. But Betz cells take a very direct route and have very little cortical connections (as least in terms of their axons). This is where widespread cortical connectivity comes in.

Scott Brady
Keith, the hypothesis is that reduction of fast anterograde transport by misregulation of GSK-3 in neurons is indeed underlying most, if not all, AD cases. There are a number of pathways that can influence these kinases. The regional specificity can arise in any one of several ways. There may be differences in the regulation of GSK-3 between neuronal populations. There may be compensatory pathways developed differentially in different neurons. Alternatively, neurons vary in their dependence on neurotrophin support and activity. Some neurons may just live closer to the edge in GSK-3 sensitive pathways.

Diego Forero
Keith and Jorge, the hippocampal cells, may have a greater dependence on axonal transport, because their greater synaptic plasticity burdens them.

Jorge Busciglio
Diego, good observation!

Diego Forero
Thanks, Jorge.

Scott Brady
Keith, the diversity of the nervous system is such that for virtually any insult, neuronal populations exhibit differential vulnerability. This is seen in spinal cord injury and ischemia, as well as toxic treatments and neurological diseases like AD.

Keith Crutcher
Okay, interesting idea...suggesting that it is the total axonal arbor and/or amount of turnover?

Gabrielle Strobel
I sense a consensus here that we are not at a point where therapy development can begin in earnest, except for established targets such as PS and GSK-3b. GSK-3a seems to be a new target that has people intrigued. Anything else?

Keith Crutcher
Thanks for letting me chime in on a subject I know little about. Maybe we can get you folks into a debate at one of our Challenging Views meetings?!

Gabrielle Strobel
We have reached the end of the hour. Please chat away as long as you like past one o'clock, but before more people start leaving, I want to thank you all very much for coming and making this such a lively discussion.

Akihiko Takashima
What about axonal transport during embryonic development? GSK-3b is present in axons during embryogenesis, but not anymore in the adult.

Diego Forero
A reference to keep in mind about axonal transport and enhancement of memory: Wong et al., 2002.

Background

Background Text
By Jorge Busciglio and Scott Brady

Neuronal cells are highly sensitive to transport defects because of their highly polarized morphology and large number of specialized microdomains. For their survival and proper function, neurons depend on the efficient delivery of proteins from the cell body to neuritic processes. Axons in particular are highly susceptible to transport deficiencies because they lack the elements necessary for protein synthesis. In this context, we have suggested that defects in protein transport play a critical role in Alzheimer's disease and other neurodegenerative conditions (Morfini et al., 2002a). Experimental evidence indicates that kinase and phosphatase activities are key regulators of fast axonal transport. Two major serine-threonine protein kinases, glycogen synthase kinase 3b (GSK3b) and cyclin-dependent kinase 5 (CDK5), have been implicated as major kinases responsible for both normal and pathological phosphorylation of tau protein in AD. Moreover, GSK3a, which is highly homologous to GSK3b, has been recently implicated in the modulation of Ab production (see ARF related news story). Both CDK5 and GSK3b have also been shown to regulate kinesin-driven motility. Specifically, GSK3b phosphorylates kinesin light chains in vivo and causes the release of kinesin from membrane-bound organelles (MBOs), leading to a reduction in kinesin-I-driven motility (Morfini et al., 2002b). Given the essential role of axonal transport in neuronal function, a misregulation of transport induced by an imbalance in specific kinase/phosphatase activities within neurons may represent an early and critical step of neuronal pathology.

Presenilin Mutations: They Do More Than Increase Ab

Significant evidence indicates that presenilin-1 (PS1) is essential for g-secretase activity. At the same time, there is also considerable evidence to suggest that PS1 has additional physiological functions, including control of calcium homeostasis, cell-cycle regulation, neurite outgrowth, apoptosis, membrane trafficking, and synaptic plasticity. In particular, PS1 has been implicated in regulating intracellular trafficking, maturation, and delivery to the cell surface of selected transmembrane proteins. Such effects of PS1 have been shown for the membrane proteins AbPP, TrkB and ICAM-5/telencephalin (Cai et al., 2002; Naruse et al., 1998; ARF related news story). One way in which PS1 might modulate intracellular protein trafficking is by regulating kinesin-based motility.

To determine whether PS1 alters kinesin-based protein transport, we utilized presenilin-1 knockout (PS1-/-) and mutant human PS1 knock-in M146V (KIM146V) mice and cultured their cells. We show that PS1 and GSK3b coimmunoprecipitate and colocalize in specific neuronal compartments, particularly growth cones (Pigino et al., 2003). Both FAD mutations in PS1, or the absence of PS1, increased relative levels of GSK3b activity. One possibility is that both PS1 and GSK3ß may be components of a trafficking regulatory complex at specific subcellular locations that is misregulated by the absence of PS1 or by the presence of PS1 mutations. Concomitant with increased GSK3b activity, PS1 deletion or PS1 mutations increased relative levels of kinesin light chain phosphorylation, and markedly reduced the amount of kinesin bound to MBOs. Consistent with a deficit in kinesin-mediated fast axonal transport, densities of synaptophysin and syntaxin-I containing vesicles and mitochondria were reduced in neuritic processes, but not cell bodies, of KIM146V hippocampal neurons. Similarly, we found reduced levels of PS1, AbPP, and synaptophysin in sciatic nerves of KIM146V mice. These results suggest that PS1 modulates GSK3b activity and normally affects the release of kinesin from MBOs at sites of vesicle delivery and membrane insertion.

In summary, we propose a model whereby mutations in PS1 compromise neuronal function by misregulating GSK3 activity, which would cause premature release of kinesin cargoes and impaired delivery of MBOs such as mitochondria to appropriate neuronal compartments. This ARF discussion will focus on the molecular mechanisms that may lead to disturbances of axonal transport and neurodegeneration in AD, including PS1 mutations, altered GSK3 kinase activity, tau hyperphosphorylation, and Ab production.

I suggest we discuss these questions during the chat:

1. The amyloid cascade hypothesis still draws majority support in the field as the predominant, if incomplete, explanation for AD. Can we integrate the axonal transport and the amyloid hypotheses? If yes, how do they fit together?

2. What is/are the earliest event(s) that disrupt axonal transport? Elevated GSK/CDK5 activity? Do we know it is occurring in AD? How can we find out?

3. In FAD, PS mutations would not only increase Ab production, but also increase GSK activity and disrupt transport. How about LOAD?

4. How could the relative roles of Ab generation and transport disruption by FAD PS1 be assessed? Video microscopy of FAD transgenic mouse neurons? What other methods are suitable?

5. How does PS1 modulate GSK3b? Most FAD PS1 mutations cause a gain of its AbPP-proteolytic function. If a different mechanism is at play, what is it and how do we find out?

6. How about GSK3a?

7. Are there therapeutic implications to this further development of the axonal transport hypothesis of AD?

References:
Cai D, Leem JY, Greenfield JP, Wang P, Kim BS, Wang R, Lopes KO, Kim SH, Zheng H, Greengard P, Sisodia SS, Thinakaran G, Xu H. Presenilin-1 regulates intracellular trafficking and cell surface delivery of beta-amyloid precursor protein. J Biol Chem. 2003 Jan 31;278(5):3446-54. Epub 2002 Nov 14. Abstract

Morfini G, Pigino G, Beffert U, Busciglio J, Brady ST. Fast axonal transport misregulation and Alzheimer's disease. Neuromolecular Med. 2002a;2(2):89-99. Review. Abstract

Morfini G, Szebenyi G, Elluru R, Ratner N, Brady ST. Glycogen synthase kinase 3 phosphorylates kinesin light chains and negatively regulates kinesin-based motility. EMBO J. 2002b Feb 1;21(3):281-93. Abstract

Naruse S, Thinakaran G, Luo JJ, Kusiak JW, Tomita T, Iwatsubo T, Qian X, Ginty DD, Price DL, Borchelt DR, Wong PC, Sisodia SS. Effects of PS1 deficiency on membrane protein trafficking in neurons. Neuron. 1998 Nov;21(5):1213-21. Abstract

Pigino G, Morfini G, Pelsman A, Mattson, MM, Brady ST and Busciglio J. Alzheimer's presenilin-1 mutations impair kinesin-based axonal transport. J Neurosci, 1 June 2003. 23(11). Abstract

Comments

  1. I would like to invite the audience to broaden their consideration of the topic of presenilin and axonal transport by including our current findings (Tezapsidis et al. 2003). Full text of the electronic version is publicly available. In our previous study (Johnsingh et al. 2000), we had demonstrated that presenilin 1 (PS1) interacts with the cytoplasmic linker protein 170/ Restin (CLIP-170). CLIP-170 is a microtubule plus-associated protein that directs minus-end movement of cargo-loaded vesicles. In our current paper, we disrupted this interaction in the neuronal cell lines SY5Y and N2a by transfecting vectors that drive the expression of peptide fragments corresponding to their binding domains (BDPs). Interestingly, disrupting the PS1/CLIP-170 complex is associated with both decreased secretion of endogenous Ab and decreased uptake of exogenous Ab from the medium.

    BDP-expressing cells were also more resistant to the surges of Ab secretion that are normally induced by thapsigargin and ionomycin, which elevate intracellular calcium concentrations, and by PS1 mutations linked to familial Alzheimer's disease. Uptake of Ab by SY5Y cells was amplified when preincubated with ApoE and was mediated through lipoprotein receptor-related protein (LRP). BDP-expressing cells or cells treated with PS1 antisense oligonucleotides took up less Ab from the medium compared to controls, indicating that the PS1/CLIP-170 interaction is involved and that PS1 cannot be substituted. In this study, we also mapped the minimum binding domains (mBDPs) of PS1 and CLIP-70 to regions corresponding to the N-terminal end of the large cytoplasmic loop of PS1 and to the C-terminal end of CLIP-170 that contains its metal-binding motif.

    The connection to axonal transport arises because our data from in-vitro taxol-polymerization of tubulin and confocal immmunofluorescence suggest that PS1, via CLIP-170, may serve as an anchor to microtubules for specific subcellular fractions containing amyloidogenic fragments.

    Interestingly, Notch is absent from this population of microtubule-binding subcellular fractions, and its cleavage was unaffected in cells transfected with the PS1-based BDP. This raises the possibility that the interaction of PS1 with CLIP-170 could provide a conceptual basis for antiamyloidogenic therapeutic strategies with improved specificity. However, this approach may be hampered by low efficiency, since it may also block Ab clearance from the interstitial space of the CNS.

    Our study is complementary to Jorge Busciglio's current paper. For example, CLIP-170 as a PS1 partner is subject to regulation by phosphorylation. GSK3b could well be among the various kinases/phosphatases capable of either promoting or inhibiting the binding of CLIP-170 to the microtubules. GSK3b inhibition by lithium is known to significantly increase the presence of CLIP-170-associated proteins (CLASP2) at distal MT ends, which should attract more CLIP-170, which should attract more PS1. This could be investigated more thoroughly.

    Minus-end directed motion within cells (i.e., retrograde transport) generally uses the motor protein dynein, while kinesin does the job in the other direction (anterograde transport). Even so, there is evidence suggesting that organelles of the same microtubule track can move in both directions (see Ma and Chisholm, 2002). At the molecular level, dynactin, previously thought to be exclusively part of the dynein/dynactin complex, can also bind kinesin II. Thus, there is room for both Busciglio's suggestion that PS1 regulates anterograde transport and our suggestion that PS1 regulates retrograde transport. We reported that mutated PS1 binds more tightly to CLIP-170 and this brings about more Ab. Busciglio and colleagues reported that mutated PS1 affects GSK3b activity, which slows down axonal transport. Perhaps these are linked; in fact, there is an abnormal cycle of increased retrograde vs. anterograde motion within AD neurons which results in making more Ab. Such a proposition would suggest that cytoskeletal abnormalities (i.e., tau phosphorylation) may actually precede amyloid build-up.

    In summary, these are the major points of our study:

    1. Prior to its cleavage to become part of the g-secretase complex, PS1 interacts with CLIP-170. This interaction may be relevant to the docking step that Berezovska et al. describe in the same issue of J. Neurosci that features Jorge Busciglio's current paper. (See also schematic diagram in our paper.)

    2. Neuronal cell lines take up Ab following its binding to ApoE through the LRP. This is confirmation of Kang et al., 2000, who worked with a different cell type.

    3. Disruption of the PS1/CLIP-170 interaction led to decreased Ab production, decreased Ab uptake, but no effect on Notch cleavage. Thus, targeting this interaction can be more specific than targeting g-secretase activity and might avoid Notch-related side effects. Here is an interesting twist: How important is Ab uptake by neurons? Is it a clearance mechanism, a signaling event, or does it contribute to neurotoxic build-up inside? Before these questions are answered, it will not be known whether blocking the PS1/CLIP-170 interaction could be efficient as a therapeutic strategy.

    4. PS1 may be involved in the docking step of intracellular vesicle transport toward the minus end of microtubules. A tighter binding of mutated PS1 to CLIP-170 at the tip of the microtubules may create conditions that favor the assembly of the g-secretase complex.

    References:

    . Microtubular interactions of presenilin direct kinesis of Abeta peptide and its precursors. FASEB J. 2003 Jul;17(10):1322-4. PubMed.

    . Altered binding of mutated presenilin with cytoskeleton-interacting proteins. FEBS Lett. 2000 Jan 7;465(1):53-8. PubMed.

    . Modulation of amyloid beta-protein clearance and Alzheimer's disease susceptibility by the LDL receptor-related protein pathway. J Clin Invest. 2000 Nov;106(9):1159-66. PubMed.

Make a Comment

To make a comment you must login or register.

References

News Citations

  1. Lithium Hinders Aβ Generation, Buffing Up GSK as Drug Target
  2. Yet Another Molecular Partner for Presenilins Identified
  3. Finally United? Aβ Found to Influence Tangle Formation

Webinar Citations

  1. From Here to There: AβPP as an Axonal Transport Receptor—How Could This Explain Neurodegeneration in AD?
  2. Axonal Transport Hypothesis Moves On to Implicate Presenilin

Paper Citations

  1. . Fast axonal transport misregulation and Alzheimer's disease. Neuromolecular Med. 2002;2(2):89-99. PubMed.
  2. . Glycogen synthase kinase 3 phosphorylates kinesin light chains and negatively regulates kinesin-based motility. EMBO J. 2002 Feb 1;21(3):281-93. PubMed.
  3. . Presenilin-1 regulates intracellular trafficking and cell surface delivery of beta-amyloid precursor protein. J Biol Chem. 2003 Jan 31;278(5):3446-54. PubMed.
  4. . Effects of PS1 deficiency on membrane protein trafficking in neurons. Neuron. 1998 Nov;21(5):1213-21. PubMed.
  5. . Alzheimer's presenilin 1 mutations impair kinesin-based axonal transport. J Neurosci. 2003 Jun 1;23(11):4499-508. PubMed.
  6. . Inhibitory phosphorylation of glycogen synthase kinase-3 (GSK-3) in response to lithium. Evidence for autoregulation of GSK-3. J Biol Chem. 2003 Aug 29;278(35):33067-77. PubMed.
  7. . Impaired axonal transport of cortical neurons in Alzheimer's disease is associated with neuropathological changes. Brain Res. 2002 Sep 6;948(1-2):138-44. PubMed.
  8. . Spherical aggregates of beta-amyloid (amylospheroid) show high neurotoxicity and activate tau protein kinase I/glycogen synthase kinase-3beta. Proc Natl Acad Sci U S A. 2003 May 27;100(11):6370-5. PubMed.
  9. . Presenilin 1 associates with glycogen synthase kinase-3beta and its substrate tau. Proc Natl Acad Sci U S A. 1998 Aug 4;95(16):9637-41. PubMed.
  10. . Aberrant activation of focal adhesion proteins mediates fibrillar amyloid beta-induced neuronal dystrophy. J Neurosci. 2003 Jan 15;23(2):493-502. PubMed.
  11. . APP processing and synaptic function. Neuron. 2003 Mar 27;37(6):925-37. PubMed.
  12. . Tau blocks traffic of organelles, neurofilaments, and APP vesicles in neurons and enhances oxidative stress. J Cell Biol. 2002 Mar 18;156(6):1051-63. PubMed.
  13. . Overexpression of motor protein KIF17 enhances spatial and working memory in transgenic mice. Proc Natl Acad Sci U S A. 2002 Oct 29;99(22):14500-5. PubMed.

Other Citations

  1. Jorge Busciglio

Further Reading

No Available Further Reading