I believe that the main mechanism via which ApoE influences AD (and CAA) is that it influences the probability of whether Aβ will aggregate into a β-sheet structure (and subsequent toxicity). I think the different human isoforms alter this probability so that E4>E3>E2. Whether this difference is due to differences in "Aβ clearance" or in "Aβ nucleation" (or both) has not yet been clarified.

View all comments by David Holtzman

I would like to share with you my opinions regarding the issue of how ApoE affects AD.

Question 1: What types of physiological mechanisms are affected by the ApoE genotypes?

ApoE4 is associated with the pathogenesis of numerous neurodegenerative diseases other than AD (e.g., MS, motor neuron disease, Wilson's disease). Based on comparisons of the effects of the ApoE genotypes on the age of onset and progression of these diseases, and on numerous studies with mice transgenic for the different ApoE isoforms, it seems that ApoE plays a key role in neuronal susceptibility to insults and in neuronal repair, and that the pathology associated with ApoE4 is due to impairments in these processes (see Chapman et al., 2001).

Question 2: What are the cellular mechanisms which mediate the effects of the ApoE4 genotype?

Analysis of the ApoE transgenics literature reveals that, depending on the experimental paradigm (e.g., aging, distinct brain lesions, head trauma, ischemia and environmental stimulation), the phenotypic expression of...  Read more

I would like to share with you my opinions regarding the issue of how ApoE affects AD.

Question 1: What types of physiological mechanisms are affected by the ApoE genotypes?

ApoE4 is associated with the pathogenesis of numerous neurodegenerative diseases other than AD (e.g., MS, motor neuron disease, Wilson's disease). Based on comparisons of the effects of the ApoE genotypes on the age of onset and progression of these diseases, and on numerous studies with mice transgenic for the different ApoE isoforms, it seems that ApoE plays a key role in neuronal susceptibility to insults and in neuronal repair, and that the pathology associated with ApoE4 is due to impairments in these processes (see Chapman et al., 2001).

Question 2: What are the cellular mechanisms which mediate the effects of the ApoE4 genotype?

Analysis of the ApoE transgenics literature reveals that, depending on the experimental paradigm (e.g., aging, distinct brain lesions, head trauma, ischemia and environmental stimulation), the phenotypic expression of the ApoE4 genotype can be devided into two classes. A loss of function (i.e., the phenotype of the ApoE4 mice is similar to that of ApoE-deficient mice), and a gain of toxic function (e.g., paradigms in which the ApoE4 mice are more susceptible to the insult than both ApoE3- and ApoE-deficient mice).

Question 3: Which molecular mechanisms mediate the diverse effects of the ApoE isoforms?

Three major molecular mechanisms have been proposed.
(i) Isoform-specific interactions with APP and amyloidosis.
(ii) Interactions with lipid transport and metabolism.
(iii) Isoform-specific and receptor-mediated effects on signal transduction and cellular signaling. These effects are mediated by the interactions of ApoE with the family of ApoE receptors and by cross-talk interactions of these receptors with other receptor systems.

In view of the suggestion that the effects of the ApoE isoforms are mediated by more than one mechanism (see question 2 above), I think that the ultimate answer will be that the above mechanisms, which are not mutually exclusive, are all valid and that their relative importance may vary in different diseases and following exposure to distinct experimental paradigms.

The jury is still out on this issue. However, if pushed to make a definite prediction, I think that the effects of ApoE in AD and neurodegeneration are probably mediated by modulation of cellular signaling. This choice stems from three observations:
(i) ApoE4 plays a role in diseases with no amyloid deposition (though APP could still be valid).
(ii) The average steady-state levels of brain lipids are not markedly affected by the ApoE genotype (although more subtle dependencies may yet be discovered).
(iii) The ApoE receptors comprise a very intricate system whose basic neurobiology is just being worked out. The richness of this system and the discovery of ApoE receptor-mediated and isoform-specific effects of ApoE on other receptor systems and ionic channels led me to "bet" on this mechanism.

With many thanks and best wishes,
Sincerely,
Danny Michaelson

View all comments by Daniel Michaelson

Prior to connecting the cell cycle to gonadotropins, I was trying to answer the question you have raised. This hypothesis is one that no one else has raised and probably for good reason. The best I could come up with was that changes in the ligand-to-receptor ratio affected Aβ production. That is, the smaller the ratio of ApoE to its receptor, the more likely β cleavage would occur.

ApoE4 has been shown to be a risk factor for AD. ApoE3 does not appear to have any effect on risk of AD, while ApoE2 is protective. ApoE genotype also affects overall expression of ApoE. E2 results in the highest levels of ApoE; E3 has somewhat diminished levels; and E4 results in the least expression of ApoE (Pablos-Mendez, 1997).

According to the hypothesis, this difference in ApoE expression would be one explanation of E4's increased risk of AD and E2's decreased risk of AD. Amyloid precursor protein (APP), in addition to probably multiple other functions, could serve as a sensor and messenger regarding available extracellular cholesterol. (I still think the primary driving force is...  Read more

Prior to connecting the cell cycle to gonadotropins, I was trying to answer the question you have raised. This hypothesis is one that no one else has raised and probably for good reason. The best I could come up with was that changes in the ligand-to-receptor ratio affected Aβ production. That is, the smaller the ratio of ApoE to its receptor, the more likely β cleavage would occur.

ApoE4 has been shown to be a risk factor for AD. ApoE3 does not appear to have any effect on risk of AD, while ApoE2 is protective. ApoE genotype also affects overall expression of ApoE. E2 results in the highest levels of ApoE; E3 has somewhat diminished levels; and E4 results in the least expression of ApoE (Pablos-Mendez, 1997).

According to the hypothesis, this difference in ApoE expression would be one explanation of E4's increased risk of AD and E2's decreased risk of AD. Amyloid precursor protein (APP), in addition to probably multiple other functions, could serve as a sensor and messenger regarding available extracellular cholesterol. (I still think the primary driving force is due to an inappropriate mitogenic stimulus. A cell undergoing division would require significant cholesterol for the lipid constituents of the two new daughter cells.) APP possesses a binding domain for ApoE in its β amyloid (Aβ) amino acid sequence (Haas, 1997) and Aβ, once cleaved from APP, retains the ability to bind ApoE (Wisniewski, 1993; Strittmatter, 1993; LaDu, 1994).

Therefore, ApoE can be bound by either APP or Aβ E2>E3>>E4 = affinity of ApoE to Aβ (LaDu, 1994; Yang, 1997) and probably APP. Internalization of the LRP/ApoE receptor/ligand complex might be associated with APP cleavage, and the site of APP cleavage might be dependent on whether the ApoE binding site of APP is occupied. It is possible that if the ApoE site on APP is occupied, α cleavage occurs, and if it is not occupied, β/γ cleavage occurs. Therefore, the more the available ApoE, the less likely β cleavage will occur. If β cleavage occurs, the resulting Aβ is able to bind ApoE, making less ApoE available to bind intact APP and thereby resulting in further β/γ cleavage of APP, producing more Aβ.

Aβ bound to ApoE/LRP complex is internalized (Narita, 1997; Du, 1997) and could serve to signal the cell that there is a shortage of ApoE (i.e., cholesterol). The remaining C-terminus of APP is also internalized and has been shown to affect gene transcription (Cao, 2001). The C-100 fragment resulting from Aβ/Aγ cleavage could also signal the cell that there is a shortage of cholesterol, while the longer fragment resulting from Aα cleavage would signal that adequate cholesterol was available.

STATINS

There is epidemiological evidence that individuals who take statins are less likely to develop AD (Wolozin). Statins are primarily used to treat hypercholesterolemia. Many individuals with hypercholesterolemia have decreased serum HDL cholesterol. The primary apolipoprotein in HDL cholesterol is ApoE. Statins have been shown to increase serum HDL cholesterol. If they are increasing serum ApoE, they may be increasing ApoE in the CNS.

View all comments by Richard Bowen

I am both puzzled and fascinated by the idea of determining the validity of a scientific finding through an opinion poll or a popularity contest. I wonder what would have been the results of a similar opinion poll, by Queen Isabella of Spain, in 1490 concerning questions about the earth being round? History of science is full of examples illustrating the gross failures of the (then) prevailing scientific orthodoxy to gauge accurately the authenticity of new findings or concepts. Perhaps the question should be framed not in terms of opinions but in terms of hypothesis, or evidence in support of a particular theory. Ultimately, the question about the role of ApoE, or any other putative risk factor, in the etiology or pathogenesis of AD will be answered through validated experimental evidence, not by opinion polls.

This might be a bit of a picky point, but it is important for the Alzheimer's Forum to maintain the clear distinctions between marketing research and hypothesis-driven research.

Best regards,
Zaven S. Khachaturian

View all comments by Zaven Khachaturian

Dear Zaven,

Thanks for your feedback on the online poll. This is purposely a bit tongue-in-cheek, amd we certainly aren't pretending to validate any hypothesis by this method! The idea is simply to take the pulse of the research community at a particular moment in time. We assume that most respondents have some hypothesis-driven reason for responding in a certain way. And some have written in with more detailed explanations (which we will post in the next few days.) As you note, the history of science is littered with the corpses of conventional ideas, and we expect the same thing might happen in AD research.... It will be interesting for people to look back on the poll results in 5-10 years' time. Anyway, I hope you don't think we are misguided in having the poll. Scientists need to have some fun too! :)

Cheers,
June

View all comments by June Kinoshita