Transcript of Live Discussion: The Role of ApoE in Alzheimer's Disease: An Alternative View

Discussion Held on 11 December 2001. Posted 10 February 2002.

Participants included Keith Crutcher, Nico Stanculescu, June Kinoshita, Bruce Teter, Marcos Marques, Alexei Koudinov, Jacob Raber, Bill Rebeck, Steve Rockwood.

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

June: Greetings everyone, why don't we begin today's discussion. Keith, would you like to make an opening statement?

Keith Crutcher: I think I have posted a fairly complete summary of the main idea, which really boils down to the notion that apoE is likely to be a much more important contributor to AD than just a contributor to amyloid deposition (for which there is already good evidence). I think the most important feature of the hypothesis is that it potentially provides an explanation for both plaque and tangle pathology as well as neuronal degeneration.

June: All of you others should be thinking up great questions to ask!

Keith Crutcher: I should note that my interest in apoE arose partly from my disappointment with amyloid as an explanatory mechanism. However, I don't think this is the time to point out the many weaknesses (IMHO) of the amyloid hypothesis.

Alexei Koudinov: The isoform-specific role of ApoE in defining aggregation properties of Ab may have to do with initial ignoring of the major function of apoE, namely its participation in lipid-, particularly cholesterol-, metabolism

June: What's the most damning evidence AGAINST the amyloid hypothesis, in your view?

Keith Crutcher: For one thing, I think the epidemiological evidence implicates apoE much more than AbPP.

Jacob Raber: We definitely see cognitive impairments in hAbPP-transgenic mice before they develop plaques, but ADDLs could contribute to the cognitive deficits.

Keith Crutcher: It is more helpful to think of criticisms of the apoE hypothesis.

Alexei Koudinov: Let's talk creatively.

June: So what is wrong with the dominant ApoE hypothesis, apart from its convergence on Ab as the disease mechanism?

Keith Crutcher: I'm not sure I know what the dominant apoE hypothesis is...effect on amyloid deposition? There seems to be an emerging consensus that amyloid deposition is an insufficient basis for understanding AD. I also think it is important to keep in mind that apoE may be contributing to AD regardless of isoform.

Alexei Koudinov: I would recommend the following references on this point: Joseph et al. 2001; Mesulam, 1999; Lue et al, 1999; Koudinov et al, 2001.) Also, a recent paper by a Scottish group (White, et al. 2001) points to importance of both apoE and apoJ in lipid redistribution under neuronal tissue repair.

Keith Crutcher: Yes, I think the idea of relative contributions to repair is still reasonable. But the dominant negative effect of E4 would suggest to me that this is not a sufficient explanation.

Alexei Koudinov: We discuss several possibilities for the role of apolipoproteins in our recent article (Koudinov et al, 2001b.)

Keith Crutcher: Yes, Alexei, and I wonder which of those possibilities you favor and why.

Alexei Koudinov: I mean not only apoE isoform differences, but also the pathways apoE could be involved with in the brain. They are those discussed by many groups. For example, cholesterol redistribution via brain tissue lipoproteins and mediated by apolipoproteins and lipoprotein receptors, ApoE's importance for synaptic plasticity, lipoprotein cholesterol removal from the brain and CSF to periphery. See (Shibata et al. 2000, Demeester et al. 2000, Rebeck et al. 1998, Koudinov et al., 2002.)

Marcos Marques: What did you think of the recent Science paper by Frank Pfrieger's group (see ARF story) claiming that cholesterol/apoE is involved in synaptic development?

Alexei Koudinov: Regarding this article, there are several background observations that are critical to it. I recommend these as further reading: (Teter et al., 1999; Fan et al. 2002; Champagne et al., Soc. Neurosci Abstract 26, 1546; Matthies et al. 1997).

Keith Crutcher: I think the evidence for cholesterol involvement in synaptogenesis is very strong, but I don't think this calls into question the possibility of apoE contributing to pathology. I think one issue is that we still don't know enough about what apoE is doing "normally".

Bill Rebeck: Do we think apoE has a job "normally", or that it only shows up during some kind of damage?

Alexei Koudinov: I think there is a normal job, see references above.

Keith Crutcher: Hi Bill, good question. I think apoE does have "normal" functions and, in fact, my suspicion is that what is happening in AD is a reflection of its normal function.

Bill Rebeck: Which is? Cholesterol delivery? Don't cells make enough of their own?

Keith Crutcher: One possibility is plasticity, as others have suggested. But plasticity involves both the generation and the removal of connections.

Alexei Koudinov: It must be to ensure synaptic plasticity structure-functional rearrangements (details in Koudinov & Koudinova, 2001). Both synthesis and distribution seems critical. Several studies point to this, see references above.

Keith Crutcher: Alexei, yes, I agree that both synthesis and distribution are relevant.I don't think cholesterol delivery is its only role, but I think the idea that apoE's interaction with lipids is critical to understanding its metabolism is highly relevant.

Bill Rebeck: Is there any evidence that apoE4 people have worse plasticity, outside of AD?

Keith Crutcher: Another good question. I don't know of direct data on this, nor how to really acquire it. Arendt's data in AD brains (Arendt et al, 1997) suggests this possibility but I don't know any other data.

Marcos Marques: Bill, I think there is, like the pugilistc dementia and neuropathology in E4 AIDS patients. (Corder et al., 1998; Dunlop et al, 1997).

Keith Crutcher: Marcos, but this would assume that the relevant variable is plasticity, for which there is no direct evidence, I don't think.

Alexei Koudinov: As for apoE4 isoforms, I recall Jude Poirier's group presenting a poster on behavioral analysis in mice in San Diego. (Find abstract.)

Keith Crutcher: Jude is a strong proponent of the idea that apoE is good, regardless of isoform, and we just need to increase apoE.

Bill Rebeck: What about the stroke data? The epidemiology about apE isoforms and stroke is weak, I think, but is there any neuropathology?

Bruce Teter: Plasticity outside of AD has not been examined directly by synaptic markers, although accumulating evidence for preclinical manifestations of the presence of E4 continue to show up (in brain imaging).

Keith Crutcher: Having worked in the plasticity field as well, I am also skeptical of the idea that just increasing plasticity is going to improve things.

Marcos Marques: I agree there is no direct evidence of worse plasticity being the major player in AD. However, I think what I cited above supports a worse plasticity in apoE4 people, potentially explaining why apoE4 people have higher cognitive impairment or poorer neuronal survival under "stress/disease". My assumption here is that apoE has a normal neurotoxicity role involved in plasticity, but the aging process or stress (such as head injury/virus infection) could lead to overly active neurotoxicity, consequently resulting in AD. In that regard the higher toxicity of the E4 isoform over that of E3 could explain its genetic risk and at the same time explain why E3 people still could get AD.

Alexei Koudinov: Joachim Herz discusses a possible signaling function for apoE LRP receptor in several articles. (Herz, 2001; see also our commentary to it, at Koudinov and Koudinova, 2001c). And there is a Japanese article on detailed characterization of apoE knockins in mice, (Hamanaka et al., 2000.) We also refer to this in our articles "Drs. Koudinov Alzheimer Bibliography."

Keith Crutcher: I think the evidence for signaling effects is especially interesting and relevant.

Keith Crutcher: I also think the evidence for a role for apoE as an anti-infective agent needs to be explored.

June: Are you referring to Ruth Itzhaki's findings?

Keith Crutcher: June, I was thinking of her data but also the data coming from apoE KO mice, which are much more susceptible to infections.

Keith Crutcher: Like in any multifactorial disease, what needs to be defined are the pleiotropic effects of specific genes.

Keith Crutcher: I think the apoE field has been dominated by this cholesterol transport hypothesis to the exclusion of other possible roles of this protein.

Bill Rebeck: I think we need to start organizing the functions of apoE into some coherent patterns: delivery of lipids, activation of specific receptor-mediated processes. It's too confusing right now.

Keith Crutcher: I agree, Bill.

Alexei Koudinov: You are right, Bill

Bill Rebeck: I also think the knockout mice are difficult to interpret if apoE has roles in development.

Keith Crutcher: Bill, yes, that is a good point, as is the case for all transgenic data.

Alexei Koudinov: I agree.

Jacob Raber: Although it may give you some hints regarding what to look for when apoE is present.

Keith Crutcher: Nevertheless, the observations are intriguing and probably relate to well-known interactions with heparan sulfate proteoglycans, which serve as receptors for various infective agents.

June: Re: ApoE knockouts, it is time to do conditional knockouts! Anyone working on this?

Keith Crutcher: June, I think some groups are working on conditional knockouts. I also think that the epidemiology for apoE effects in humans is relatively unexplored for many putative roles of apoE.

Alexei Koudinov: I am not aware. And what about knocking out the function of a selected brain area, as done in (Riedel et al., 1999)?

June: Keith and I have talked about constructing a pathway diagram to depict all of the reported possible mechanisms. I'm wondering if there is some way to construct such a pathway diagram through a collective effort.

Alexei Koudinov: Good idea

Bruce Teter: Pathway diagram- wonderful idea, let's set up a common site, start with a basic structure, and have everyone be able to add to it.

Bill Rebeck: I love the idea of a pathway chart, but can it be one where every thought does not get put in until at least several people agree on it? Too many pathways just puts off research....

Keith Crutcher: Bill, yes I think that is a potential problem. Maybe a pathway could be indicated with links to supporters. (In my case, the neurotoxic hypothesis may only have one link!)

June: Each pathway needs to be annotated and have a "weight" assigned to it that represents the strength of the evidence.

Keith Crutcher: I also would like to see a clear distinction made between levels of analysis for such a diagram.

Bruce Teter: Any ideas for how this could work - web site, graphics, moderator?

Keith Crutcher: ARF is the most logical location for this.

June: I can set up a pathway site on the Alzheimer Research Forum web site. It won't be interactive (not unless I invest a huge amount in software development), but people can comment, and I can make changes to the diagram.

Alexei Koudinov: By the way how many hits does ARF collect per week?

June: Alexei, we get around 14,000 visits per week. I estimate that around 60-70 percent are researchers.

Alexei Koudinov: Great

Keith Crutcher: I think the amyloid field would benefit from such a diagram as well.

June: I've thought about that as well. But that's an even bigger challenge. ApoE might provide a nice test bed for the system.

Keith Crutcher: Well, apoE is more important anyway :-)

Marcos Marques: Interesting too, is how people will evolve in their acceptance of new ideas, like for example the neurotoxic role for apoE, which is part of today's discussion.

Bill Rebeck: I know of two or three other groups who talk about apoE4 toxicity. Including me.

Keith Crutcher: I think indicating the weight of evidence might be difficult but indicating the amount of support from the scientific community could be illuminating in many ways. Well, we have at least two or three links to the neurotoxic hypothesis! And I know that Mary Jo LaDu, who was an early contributor to the idea, hasn't completely abandoned it (even though she is focused on Ab-derived diffusible ligands now.)

Alexei Koudinov: What is your opinion, Marcos?

Marcos Marques: Alex/Bill, it is comforting that when confronted with experimental evidence, people can change their opinion, maybe we have some examples in the room :-)

Alexei Koudinov: By the way, another cross of apoE and Ab could be through their competition for lipoprotein receptors, as they both bind to it and both are players on lipid metabolism (Winkler et al. 1999; Zlokovic et al., 1996; Koudinov et al., 1998).

Keith Crutcher: Alexei, I agree that the interaction with receptors is likely to be very important.

June: Regarding the neurotoxicity hypothesis, what are the most compelling data to support it?

Keith Crutcher: I tried to summarize the most relevant data in my overview. I think both in vitro and in vivo data are consistent with the idea that apoE4 has greater neurotoxicity.

Bill Rebeck: I like the data linking apoE to changes in intracellular calcium and then downstream effects (Ohkubo et al., 2001.)

Alexei Koudinov: Another good point

Keith Crutcher: Yes, we had also seen effects of truncated apoE on intracellular calcium that seemed to correlate with neurotoxicity (Tolar et al. 1999). Neurotoxicity! The Japanese group provided quite impressive evidence for a role of intracellular signaling pathways.

Bill Rebeck: I have been trying to argue that apoE normally has short-term effects but then is cleared, but in AD it gets stuck on those plaques for a long time, having chronic effects.

Keith Crutcher: Bill, yes I think this idea of apoE hanging around and persistently activating receptors is very nice.

Alexei Koudinov: We discuss in Koudinov & Koudinova, 2001, that during the synaptic efficacy amplification, apoE may act through lipoprotein receptors in a short-time period of the long-term potentiation onset.

Marcos Marques: I think one interesting aspect about a putative apoE isoform trans-signal effect is that it may be related not only to differential isoform receptor affinity (as observed in neuronal cell lines (Ji et al., 1998) but, more importantly, to a differential apoE isoform intracellular fate, particularly in real neurons (Dekroon et al., 2001, Neurobiol disease). To complicate matters, the specific lipids that can be associated with apoE can interfere with apoE proteolysis (Bradley et al., 1984). I speculate that apoE proteolysis could be involved in trans-signal effects of apoE via a feedback effect that depends on isoform specificity and/or lipid association. Thus, the LRP/NMDA trans-signal effect demonstrated in Bill Rebeck's lab [Bacskai et al., 2000] should also be observed for apoE, but I predict that this is related to the generation of 22 kDa apoE, which shows calcium influx mediated by LRP/NMDA as we previously demonstrated (Tolar, 1999).

Keith Crutcher: In the Tolar et al. paper, we had also found a linkage between the truncated-apoE effect and NMDA receptors, something that really surprised us, but we were happy to find that Bill's group found similar things.

Bill Rebeck: Did anyone see that data from Ulrike Beisiegel talking about apoE recycling after endocytosis? Another mechanism that may differ between isoforms. (Heeren J, 2001).

Keith Crutcher: Yes, the Beisiegl data point to another possible mechanism.

Keith Crutcher: Sounds like we don't have strong skeptics here.

Alexei Koudinov: ... so, the change in apoE recycling can make a very important difference. It also may change the recycling of cholesterol that apoE normally mediates.

Marcos Marques: The isoform-differential intracellular fate of apoE could also be related back to intracellular apoE proteolysis and its "potential" neurotoxic effect.

Bill Rebeck: If you buy that apoE receptors have signaling functions, it's not a large step to apoE sometimes having toxic effects. The question is when and how strong.

Steve Rockwood: {enters}

June: Hi Steve!

Steve Rockwood: Hi, June.

Alexei Koudinov: I also mean apoE receptors as a way to deliver lipoprotein cholesterol for structural, and likely immediate, plasticity needs.

Keith Crutcher: Yes, that is one of the points I tried to make. The fact that apoE can activate intracellular pathways opens up a whole new vista on possible effects on neurons. The fact that it seems to be linked to NMDA, with its well-known excitotoxic links, is especially intriguing.

Bill Rebeck: I think the presence of excess cholesterol signals nearby cells to branch out. Those cells need the cholesterol and the signal.

Alexei Koudinov: ... cholesterol to get integrated structurally in new dendritic processes. Thanks for bringing up this issue...

Keith Crutcher: One relatively unexplored avenue is the extent to which lipids might modify these signaling effects.

Bill Rebeck: True, Keith. It's just hard working with those lipids.

Keith Crutcher: Yes, indeed, but maybe that is where some of the action is? Bill, what kind of signal do you envision for this?

Bill Rebeck: I think the first step is those proteins that bind the cytoplasmic tail of apoE recepotors. Disabled, Fe65, etc.

Keith Crutcher: Do you think these proteins will serve as linkages to the NMDA receptor as well?

Bill Rebeck: Yes, I think the NMDA linkage is real. I think LRP and NMDA are both right there in the synapse.

Alexei Koudinov: …as supported by highest expression of lipoprotein receptors in pyramidal cell layers of CA1 and the granule cells of the dentate gyrus in the hippocampus.

Keith Crutcher: Bill...I agree. This is probably a very important direction to pursue.

Bill Rebeck: Maybe linking activity to plasticity.

June: Keith, have the proteases that cleave apoE into your N- and C-terminal fragments been identified?

Keith Crutcher: As far as I know, the question of which proteases are involved hasn't been answered. However, a number of proteases have been shown to be able to cleave apoE into the two major fragments (Wetterau et al. 1988; Aggerbeck et al. 1988). What has impressed me from the outset is the presence of the protease-sensitive hinge domain.

Bruce Teter: The effect of proteolysis on apoE trafficking - isoform dependence, recycling, intracellular fate etc.- all that has not been examined.

Keith Crutcher: Yes, Bruce, this is an interesting point.

Bill Rebeck: I need to leave. Thanks everyone!

Bruce Teter: Bye Bill

Keith Crutcher: See you Bill! Thanks for checking in.

Nico: Thanks, Bill

June: Bye Bill! I'll circulate a transcript, so you can add further comments.

June: So Keith, there's so much going on with ApoE, it's hard to get a sense of what's potentially most important for AD. What are the priority questions for further investigation?

Keith Crutcher: I think it is fair to say that apoE is probably doing something of interest...what exactly remains to be determined. But I think as more studies are done, specific hypotheses can be tested.

June: Do you think the cleavage products are critical?

Keith Crutcher: The evidence we have points to a role for the fragments.

Keith Crutcher: I think we need to know whether E4 toxicity is relevant or just a tissue culture artifact (like amyloid;->)

Bruce Teter: Keith and all - I must run, thanks for organizing this - a good start. Will send specific comments by email.

June: Thanks, Bruce! I'll look for your email.

Keith Crutcher: Thanks Bruce...see you

June: Is the E4 toxicity issue best addressed through transgenic models?

Keith Crutcher: Transgenics might help. Jacob Raber's and Lennart Mucke's mice certainly are suggestive (Raber et al, 1998; Buttini M et al. 2000; also note Raffai & Weissgraber 2002).

June: Which mice, specifically?

Jacob Raber: GFAP- and NSE-apoE mice (Hartman et al. 2001.)

Keith Crutcher: We proposed some time ago to generate mice expressing the truncated form of apoE (one of the many triaged grant proposals). Jacob's mice show evidence of proteolysis.

Marcos Marques: Jacob, has anyone done apoE knockin over an AbPP knockout, do you think that could clarify or separate the role of amyloid and apoE?

Jacob Raber: Not as far as I know. You would need to knock APLP2 out as well.

Keith Crutcher: I need to head out. Sorry for all the trouble folks had with the software today. Thanks for coming. Thank you June and Nico.

Nico: Thanks, Keith, and thanks to all!

Jacob Raber: Best regards.

Steve Rockwood: Sorry to have come in so late--thanks for letting me skulk!

June: Thank you all for participating today. You may send additional questions and comments to me by email. You're also welcome to keep chatting as long as you like.

Keith Crutcher: Yes, I will stay logged on for a while but I may not be able to respond. So don't mind me if I don't reply. Please feel free to send me questions by email.