S2-02-04: Amyloid-beta: From physiology to pathology

Symposia S2-02: Disease Mechanisms: APP and Beta-Amyloid irreversible. Results: We first observed close postsynaptic association of calpain activation with A␤ plaque formation in brains from both AD patients and transgenic (Tg) mice overexpressing amyloid precursor protein (APP). The mice also exhibited axonal termini dynamically misdirected to calpain activation-positive A␤ plaques. Consistently, cerebrospinal fluids from Tg mice and AD patients contained greater quantity of calpain-cleaved spectrin than controls. Genetic deficiency of calpastatin (CS), calpain-specific inhibitor protein, augmented A␤ amyloidosis, tau phosphorylation, microgliosis, somato-dendritic dystrophy and mortality in APP-Tg mice. In contrast, brain-specific CS overexpression resulted in opposite phenotypes. Conclusions: These observations indicate that calpain activation may underlie A␤-triggered pathological cascade and thus shall become a relevant pharmacological target in the treatment of AD. S2-02-03

REGULATION OF AICD GENERATION

Frederic Checler, Institut de Pharmacologie Moleculaire et Cellulaire, Valbonne, France. Contact e-mail: [email protected] Background: Amyloid b-peptide (Ab), which plays a central role in Alzheimer Disease, is generated by presenilin-dependent and presenilinindependent g-secretase cleavages of b-amyloid precursor protein (bAPP). We report that the presenilins (PS1 and PS2) also regulate Ab degradation and p53-associated cell death. Methods: Classical cell biology approaches. Results: Presenilin-deficient cells fail to degrade Ab, and have drastic reductions in the transcription, expression and activity of neprilysin, a key Ab-degrading enzyme. Neprilysin activity is restored by transient expression of PS1 or PS2, and by expression of the amyloid intracellular domain (AICD), which is co-generated with Ab, during g-secretase cleavage of bAPP. Neprilysin gene promoters are transactivated by AICDs from APPlike proteins (APP, APLP1 and APLP2). We also demonstrate that depletion of nicastrin, one of components of the presenilin-dependent g-secretase complex, also drastically reduces neprilysin activity and protein and mRNA expressions. Furthermore, we demonstrate that PS deficiency, catalytically inactive PS mutants, g-secretase inhibitors and bAPP or APLP2 depletion, all reduce the expression and activity of p53, and lower the transactivation of its promoter and mRNA expression. p53 expression was also diminished in the brains of PS- or bAPP-deficient mice. The g- and e-secretase-derived C-terminal fragments (AICDC59 and AICDC50, respectively) of bAPP trigger p53-dependent cell death, and increase p53 activity and mRNA. Finally, PS1 mutations enhance p53 activity in HEK293 cells and p53 expression in FAD-affected brains. Conclusions: Thus, our studies show that AICDs control neprilysin and p53 at a transcriptional level, in vitro and in vivo and unravel novel functions for presenilins. S2-02-04

AMYLOID-BETA: FROM PHYSIOLOGY TO PATHOLOGY

Ottavio Arancio, Columbia University, New York, NY, USA. Contact e-mail: [email protected] Background: Various lengths of Amyloid-␤ (A␤) peptides (e.g. A␤42 and A␤40) are produced in abnormally high amounts during Alzheimer’s disease, whereas they are present in low amounts in normal healthy individuals. Methods: We investigated whether and how A␤ peptides interfere with both memory formation and the regulation of hippocampal LTP, an electrophysiological correlate of learning and memory. Results: Starting from the observation that nanomolar levels of A␤42 impair LTP and both spatial and fear memory, I will present data indicating that nanomolar levels of A␤42 reduce phosphorylation of the memory molecule CREB. I will also show that A␤-induced reduction in LTP and memory is rescued by up-regulation of both the adenylyl-cyclase-cAMP-PKA pathway and the NO-synthase-guanylyl cyclase-cGMP-PKG pathway, two enzymatic cascades that are known to increase phospho-CREB. The effects of A␤ peptides, however, are not limited to disruption of synaptic function and memory. We also found that in contrast to nanomolar concentrations, low picomolar

T125

concentrations of A␤42 caused a marked increase of hippocampal LTP and both spatial and contextual fear memory. These positive effects involve an enhancement of the neurotransmitter released during LTP induction and modifications of the mechanisms regulating memory acquisition. Both effects rely on the presence of ␣7-containing-nicotinic-acetylcholine-receptors (␣7nAchRs), as perfusion with picomolar levels of A␤42 fail to enhance LTP and spatial and contextual memory in ␣7nAchR-KO mice. How does it happen that a molecule performing a positive function turns into a bad boy? We are currently investigating the effects of exposures to picomolar concentrations of A␤42 over time. We have found that the spontaneous neurotransmitter release frequency starts to increase after 30 min. Moreover, 24 hr treatment with picomolar A␤42 suppresses different forms of synaptic plasticity including the enhancement in the number of spontaneous transmitter release events, active boutons and immunoreactive puncta for both pre- and post-synaptic markers. Conclusions: We propose a model for A␤ peptide action in which their positive and negative effects on synaptic plasticity and memory represent a continuum, with low concentrations of A␤ playing a positive role enhancing synaptic plasticity and memory, and high concentrations playing a negative role reducing synaptic plasticity and memory. S2-02-05

ANTI-ABETA ANTIBODIES IN ALZHEIMER’S DISEASE MOUSE MODELS

Todd E. Golde, Mayo Clinic College Of Medicine, Jacksonvile, FL, USA. Contact e-mail: [email protected] Background: The pioneering work of Schenk and colleagues demonstrated the therapeutic potential of generating an antibody response capable of recognizing Abeta and Abeta amyloid. Indeed, dozens of studies now show that in mouse models of AD antibodies to Abeta can attenuate Abeta deposition and plaque associated pathologies (e.g. gliosis) and reverse behavioral deficits in these mice. Despite initial setbacks multiple active and passive Abeta immunotherapy trials are now being tested humans. Methods: Both active immunizations with fibrillar or Abeta or fragments of Abeta and passive immunization with anti-Abeta antibodies have been evaluated in mouse models of AD. Endpoints have included Abeta deposition, gliosis, behavioral alterations, and levels of Abeta and antibody in various brain, CSF and plasma. Results: Numerous studies consistently demonstrate that active or passive immunotherapy suppresses Abeta deposition in mouse models of AD. Though diffuse plaques may be cleared by immunotherapy, it is controversial as to whether compact plaques are cleared. In mice only rare side effects have been reported. To date there is no consensus as to mechanisms of action, ideal target epitope, or whether an antibody is truly needed. Our work suggests that efficacy is related to target epitope, is dependent on the levels of Abeta deposition at the time of initiation of therapy, and cannot be easily explained by most of the postulated mechanisms. Novel strategies targetting amyloid versus Abeta will be presented. Conclusions: Abeta immunotherapy remains a promising approach for the treatment of AD. Late stage trials of passive immunotherapy are underway in humans. Though mouse models may be helpful in evaluating the potential efficacy of a given Abeta immunotherapy, they may not be useful in predicting possible side effects. Even in mice the efficacy of immunotherapy is limited with respect to certain endpoints when treatment is initiated in mice with AD like amyloid loads. Further understanding of the mechanisms underlying efficacy are likely to be needed in order to develop an optimized immunologic approach to targeting Abeta in AD. S2-02-06

IMMUNOLOGICAL ANALYSIS OF A␤ OLIGOMER DIVERSITY USING CONFORMATION-DEPENDENT ANTIBODIES

Charles Glabe, University of California, Irvine, Irvine, CA, USA. Contact e-mail: [email protected]