Conference Index | Round Tables | Selected Abstracts | News Summaries

Molecular Interactions in Alzheimer's Disease Pathogenesis: Diagnostic and Therapeutic Implications

Reported by Keith A. Crutcher

July 20, 1998 The "great amyloid debate" appeared to resolve at least one issue in the rather contentious arena of defining the critical neuropathological events in AD. Amyloid per se, i.e., "extracellular, fibrillar, congophilic deposits" (as stated by Robert Terry), is unlikely to be central to the disease process (not implying that APP or the A-beta peptide are not important). Of course this doesn't narrow the possibilities down very much, but it does give those who have not devoted their lives to the study of amyloid some hope that their efforts are not in vain. (It wasn't that long ago that one of the leaders in AD research quipped that "if you weren't working on amyloid, you weren't studying Alzheimer's disease"). In this workshop, half of the presentations related to the potential role of apoE (reviewed in the plenary lecture by Dr. Mahley on Sunday), two others focused on A-beta peptide effects and one reported on tau-microtubule interactions.

ApoE is a bit like an unwelcome dinner guest, not exactly a pariah, but, rather, someone who is well-known but just doesn't belong at the event. Its well-documented role in lipid transport seemed to secure a comfortable place in biology before it barged into the field of AD research. But that's exactly what it did when the group at Duke University (Roses et al.) showed the greater epidemiological risk of AD with the apoE4 genotype.

The three presentations on apoE were all based on in vitro models of apoE function. D. Holtzman (abst. 876) provided an update on studies of the effects of apoE on neurite outgrowth using cultured astrocytes from transgenic mice expressing human apoE4 or apoE3 under the control of the GFAP promoter. This leads to high levels of apoE in the brain, primarily localized to astrocytes, which also secrete high levels of apoE in culture. The secreted apoE is found in association with HDL-like particles and cells secreting E3 promote more neurite outgrowth than E4-secreting cells, outgrowth that is blocked by antibodies to LRP. Entorhinal cortex lesions in apoE-deficient mice reveal slowed clearance of degenerating fibers, suggesting a role for apoE in this process. More recent unpublished data reveal that the apoE3 particles look similar to the mouse wild type particles but have reduced lipid content. ApoE-deficient astrocytes do not appear to secrete HDL-like particles. Holtzman summarized with a suggestion that apoE may be able to deliver or remove cholesterol from neurons, thereby accounting for possible roles in neurite outgrowth and clearance of degenerating fibers, respectively.

M. LaDu (abst. 877) continued the apoE theme but was more focused on the interaction between apoE and A-beta. She reviewed published data demonstrating that apoE2 and apoE3 form stable SDS-resistant complexes with A-beta whereas apoE4 does not. This isoform difference in complex formation is dependent on the presence of lipid. This suggests that the type of lipid particle is important in determining the activity of the apoE. In fact, lipid-associated apoE3 prevents A-beta neurotoxicity, but apoE4 does not (in fact, apoE4 exhibits slight neurotoxicity by itself). Furthermore, both isoforms show the ability to block the activation of astrocytes by A-beta. This is a transient effect and may involve a signaling pathway. LaDu suggested that the interaction of apoE with A-beta might ultimately lead to differences in the extent to which A-beta is deposited as plaques (along with apoE). In the presence of apoE4, higher levels of soluble A-beta may result in greater activation of astrocytes and cytotoxicity. She emphasized the importance of characterizing the lipid environment of apoE since this seems to be critical in explaining the isoform-specific differences.

The third apoE-related talk was presented by W. Strittmatter (abstr. 879), who, as part of the Duke group, was one of the first to propose that apoE may play a critical role in AD through interactions with the cytoskeleton. He noted that isoform influences of apoE are not restricted to AD since there is considerable evidence in the literature for greater deficits following a variety of neurological insults in the presence of the apoE4 allele. In order to uncover possible biological functions of apoE, apoE-deficient mice have been studied for alterations in peripheral nerves. In fact, nerves from such mice exhibit more irregular profiles of unmyelinated axons (decreased circularity) as well as a decrease in their number. This may potentially relate to the fact that apoE is normally expressed in non-myelinating Schwann cells. These mice also exhibit decreased nociceptive function. Another alteration is an increase in the type of tau ("big tau") normally found in peripheral nerve. Strittmatter suggested that this may reflect alterations in the cytoskeleton arising from the loss of a normal interaction of apoE with tau or MAP2c.

E.-M. Mandelkow (abstr. 878) presented a fast-paced overview of the "tau hypothesis" and the potential role played by MARKs (microtubule-affinity-regulating-kinases). She reviewed evidence for the importance of phosphorylation sites on tau (especially serine-214 and serine-262), which appear to regulate the interaction of tau with microtubules. Overexpression of MARKs in CHO cells, for examples, leads to disruption of microtubules. Mandelkow presented an interesting hypothesis regarding the role of MARKs in maintaining neuronal polarity. A homologue of MARK, par-1, appears to play a critical role in establishing polarity in the c. elegans zygote and in epithelial cell polarity. By analogy, Mandelkow suggested that the primary alteration in AD neurons is also a loss of polarity. In fact, an antibody that detects tau phosphorylated at serine-262 reveals otherwise apparently-healthy neurons prior to the development of neurofibrillary changes.

The final two talks related to effects of the A-beta peptide. D. Small (abstr. 880) presented data on the effects of this peptide on AChE expression. He noted that although there is a general decline in AChE in the AD brain, there is an increase around plaques. A-beta peptides that form fibrils cause increased AChE expression in a carcinoma cell line (P19 cells), an effect that is mediated by L-type calcium channels. In transgenic mice expressing CT100 there is also high production of A-beta, providing a model for examining the effects of increased levels of A-beta on AChE in the brain. In fact, these mice have higher AChE levels when extracted under non-detergent conditions. Since there are different forms of AChE, depending on the extent of glycosylation, the possibility of differential effects on AChE type was studied. A greater amount of abnormally-glycosylated AChE is present in both the transgenic mice and in AD brain tissue. Furthermore, elevated levels of this form of AChE are present in postmortem CSF samples from AD patients, but not controls or patients with non-AD neurological disease, suggesting potential utility for diagnostic purposes.

S. Yan (abstr. 881) reviewed evidence for possible mechanisms mediating A-beta toxicity. She noted that at low levels of A-beta, its effects may be mediated by different mechanisms than when present at high, fibrillar levels, the latter perhaps causing more non-specific changes. One receptor candidate for mediating extracellular effects of A-beta is RAGE (Receptor for Advanced Glycation End-products), a member of the immunoglobulin superfamily that is expressed at high levels in the AD brain and is expressed on cells that are affected by A-beta. Expression of dominant-negative RAGE in neuron-like cells blocks the activation of ERK1/2 and apoptosis. A candidate for mediating intracellular effects of A-beta is ERAB (Endoplasmic Reticulum A-beta-Binding Protein), which appears to promote A-beta toxicity, as evidenced by increased apoptosis in COS cells expressing mutant ERAB. Future studies are directed to clarifying the role of these candidate mediators of A-beta toxicity through transgenic models.

Unfortunately, the subsequent discussion period, in which a number of questions from the audience were fielded by the panel of speakers, was not much more illuminating than the individual presentations, suggesting that apoE, tau, and A-beta, all of which deserve to be invited dinner guests, still have not revealed their true identify and remain somewhat enigmatic players in a very complicated drama.