By Minji Kim, Alice Lu, and Rudy Tanzi.
Animal Models and Therapeutics
Cynthia Lemere, Brigham and Women’s Hospital, Boston, discussed the Aβ vaccine approach to AD and progress toward developing a potential vaccine for AD that is both safe and efficacious. She provided an overview of the results of intranasal, subcutaneous, and transcutaneous immunizations using two types of Aβ1-15 immunogens (including dendromeric forms) with various adjuvants, including LT(R192G). She highlighted the promising outcomes, such as high antibody titers, reduced cerebral insoluble Aβ42 level, and lower plaque burden, when using the new immunogens together with various adjuvants. She also warned of recent reports that passive immunization with Aβ could lead to cerebral hemorrhage, especially in cases with high levels of congophylic amyloid angiopathy.
Charlie Glabe, University of California, Irvine, presented data from immunization of triple transgenic mutant mice with oligomeric Aβ antigen covalently coupled to colloidal gold. He indicated several advantages of immunization with the Aβ oligomer over fibrillar Aβ antigen. These included much more restricted immune response specifically to oligomers, a lower inflammatory response in terms of microglial activation, and equally effective reduction of amyloid deposition and improvement of behavioral performance. Targeting amyloid oligomers was suggested to be an effective strategy for developing an AD vaccine. Glabe also presented data on his pan-anti-oligomeric antibody, A11, which is able to detect many different soluble amyloidogenic oligomers (but not monomers) and prevent their toxicity.
Dora Kovacs, Massachusetts General Hospital, Boston, focused on the link between cholesterol and AD pathogenesis. She presented a review of the therapeutic effects of statins and ACAT inhibitors (CP-113,818 and CI-1011) on Aβ generation and Aβ pathology in a transgenic mouse model and cell-based assays. By providing 2D gel data showing decreased APP-binding of ER proteins upon CP-113,818 treatment, she illustrated that lack of ACAT activity perturbed APP processing in the ER, perhaps degrading a fraction of immature APP. She also showed that ACAT inhibitors reduce amyloid load more effectively than do statins in the transgenic mice. In exploring the mechanism by which ACAT inhibition and the resulting decrease in cholesteryl esters lowers Aβ generation, Kovacs found that ACAT inhibition retained APP back at the ER, thus impairing maturation and trafficking through the secretory pathway. This results in diminished secretion of APPs and Aβ. Kovacs also described a novel cleavage of APP in the N-terminus in response to cholesteryl ester levels. While searching for proteases that might carry out this clip, Kovacs discovered that HtrA2 was implicated. It was originally found as an APP interactor that preferentially binds immature APP holoprotein. Levels of immature APP were also found to be increased in HtrA2 knockout cells.
Abraham Fisher, Israel Institute for Biological Research, Ness Ziona, Israel, described the effects of AF267B, an M1 muscarinic agonist, on AD pathology in triple transgenic AD mice. AF267B is CNS-penetrable and attenuated all three major pathological hallmarks of AD in the triple transgenic mice: interneuronal and extraneuronal accumulations of Aβ, tau hyperphosphorylation, and cognitive impairment. After pointing out the uncertainty of the AD etiology and therapeutic need to target all AD hallmarks, Dr. Fisher suggested that AF267B might be useful for both prevention and treatment of AD. The drug would not only treat the symptoms of AD (by ameliorating the cholinergic pathway), but based on the new transgenic data, could potentially delay AD progression by reducing Aβ and tau pathologies. The M1 agonist is now being tested in phase 1 clinical trials for AD (under the name of NGX267) by TorreyPines Therapeutics, Inc, La Jolla, California.
Steven Jacobsen, Wyeth Research, provided an overview of the results of passive immunization against Aβ in PDAPP transgenic mouse. Clearance of amyloid plaques was detected by treatment with 12A11, an anti-Aβ monoclonal antibody (mAβ), but not by mAβ 266. To assess effects on cognition, contextual fear conditioning (CFC) was performed, and revealed that contextual memory (CM) deficits occurred in the pre-plaque and the plaque-bearing transgenic mice; mAβ 266 led to acute reversal of the CM deficits. Effects of the anti-Aβ mAβs on improving CM deficits were different depending on the epitopic region, with mAβs directed at the N-terminal portion of Aβ showing the most efficacy.
In the session’s short talk, Thomas Bayer, University of Saarland, Germany, presented the relationships of intraneuronal Aβ levels, neuron loss, and axonopathy in an APP/PS1 transgenic mouse model. He indicated that plaques did not influence neuronal integrity, while Aβ deposits led to intracellular Aβx-42 aggregates, suggesting a revised amyloid hypothesis.
Therapeutics and Imaging
Eddie Koo, University of California at San Diego, provided an overview of the role of nonsteroidal anti-inflammatory drugs (NSAIDs) in treating AD, demonstrating that a subset of NSAIDs modulate γ-secretase activity by reducing Aβ42, while increasing Aβ38. These effects required relatively high concentrations of NSAIDs and were found to be independent of COX inhibition, the primary mode of action of this family of compounds. NSAIDs had no effect on AICD, NICD, or other γ-secretase substrates at modulatory concentrations. Using PS1 mutations, in vitro γ-secretase assays, and FRET/FLIM, Koo showed that the effect of NSAIDs on Aβ42 might be mediated at the γ-secretase/substrate level, with some caveats. Koo also presented data from recent Phase 1 and 2 trials of the R-enantiomer of flurbiprofen, flurizan (Myriad). This NSAID analog with marginal COX-inhibitory activity modulated Aβ42 levels in cells and animal studies. However, in Phase 1, there was no effect on the measurements of Aβ in CSF (pre- and post-drug), perhaps due to the time of lumbar puncture. In Phase 2, 207 subjects treated for one year showed about a 30 percent slowing in cognitive decline in the 1,600 mg dosing (highest dose) cohort; mild AD patients benefited, while moderate-stage AD patients did not. Mild patients also had statistically significant benefits in Activities of Daily Living and Global Function. A larger trial is currently underway.
Christoph Hock, University of Zurich, Switzerland, provided an update on biomarkers for AD. He summarized the use of measurements of CSF-Aβ, plasma-Aβ, β-amyloid in brain by PET, absolute rCBF, and tau proteins as current biomarkers for AD. CSF levels of Aβ42 were low in AD, while both total tau and phosphorylated tau were increased. However, sensitivity and specificity of these biomarkers, even when combined, did not surpass clinical diagnostic accuracy using psychological testing. Reliable biomarkers for sporadic AD in plasma are not yet available, but strategies to improve diagnostic accuracy include multiplexing arrays, proteomic patterns, and discovery, as well as combinations with imaging measures, are being utilized. Potential proteomic biomarkers mentioned include ApoA1, truncated transthyretin, and glutathione-conjugated transthyretin. Hock also pointed out that auto-antibodies to cross-linked Aβ peptide species (CAPS), which have been reported to be decreased in AD plasma, are worthy of further study. He concluded that direct in vivo measurement of brain β-amyloid and rCBF by PET and brain volume by MRI will also be potentially useful in diagnostic and therapeutic evaluation of AD.
Chester Mathis, University of Pittsburgh Medical Center Presbyterian Hospital, presented on in vivo imaging of β amyloid in brain using 6-OH-BTA-1, or Pittsburgh Compound B (PIB). 6-OH-BTA-1 fulfilled the properties needed for an Aβ PET radioligand: It bound selectively to and had a high affinity for Aβ; it crossed the blood-brain barrier well; it had rapid nonspecific clearance; it had no radio-labeled metabolites in the brain; and it worked in vivo in animal models. In human AD brains, PIB appeared in expected gray matter areas and was absent where there was no amyloid. In contrast, in control brains very little PIB retention was observed with an absence of retention in gray matter. PIB retention controls had no overlap between control and AD groups, but MCI subjects spanned the control and AD groups. Unusual patterns of PIB retention were observed in FAD cases and in Down syndrome, perhaps due to preference of the compound for cerebral blood vessel amyloid.
Ward Pedersen, Creighton University Medical Center, described the occurrence of insulin resistance in aged Tg2576 APPswe mice as a rationale for treating them with rosaglitazone, a thiazolidinedione used to treat insulin resistance in diabetics. He demonstrated that rosaglitazone attenuated the learning and memory deficits of Tg2576 mice. Interestingly, metyrapone, an inhibitor of glucocorticoid production, mimicked the effects of rosaglitazone on the learning and memory performance of Tg2576 mice, in a reversible fashion. Rosaglitazone also normalized IDE mRNA levels in the hippocampi of Tg2576 mice, and reduced Aβ42, but not Aβ40, levels with no effect on amyloid plaque burden. Rosiglitazone is now being tested in clinical trials for AD.