O1-06-07

S22 O1-06-07

Plenary PL2: IMMUNOTHERAPY WITH ANTIBODIES TO NTERMINAL AMYLOID-␤ PEPTIDES REDUCES CEREBRAL AMYLOID ANGIOPATHY IN PDAPP MICE

Sally Schroeter, Karen Khan, Robin Barbour, Tam Doan, Ming Chen, Terry Guido, Guriq Basi, Peter Seubert, Dora Games, Elan Pharmaceuticals, South San Francisco, CA, USA. Contact e-mail: [email protected] Passive immunotherapy with amyloid-␤ (A␤) antibodies reduces a number of AD pathologies including parenchymal A␤ plaques, neuritic dystrophy, synaptic loss, and memory function in mouse models, and many of these effects are beginning to be seen in AD patients. In addition to parenchymal deposition of amyloid, AD is also characterized by amyloid fibrils in the cerebral vasculature in the majority of patients, a condition called cerebral amyloid angiopathy (CAA). Previous studies have suggested that immunotherapy does not clear and may increase CAA and the incidence of microhemorrhage. In the present study we demonstrate that certain immunotherapies are surprisingly highly efficacious in removing CAA and that the incidence of microhemorrhage can be modulated by dosage. We examined the effects of passive immunization on CAA in 2 PDAPP mouse studies: one a comparison of antibodies to different A␤ epitopes (3D6, A␤1-5; 12A11 and 10D5, A␤3-7; 266 A␤16-23) at one dose, and a 3D6 dose-response study at 3 doses. CAA clearance and microhemorrhage incidence were assessed after 6 months using A␤ IHC and hemosiderin detection. We found that 3D6 at the highest dose completely cleared CAA, with nearly 100% clearance observed in both studies. In addition, partial clearance was evident at both lower doses as well as with 10D5 and 12A11 treatments. 266 showed no effect on CAA or plaque clearance. Sites of hemosiderin staining were limited to focal, perivascular regions and were almost entirely restricted to meningeal vasculature and sagittal sinus vessels. Lower dose 3D6 groups had significantly reduced hemosiderin ratings compared to the highest dose group, and the 266 and 10D5 groups did not significantly differ from the control. Hemosiderin staining co-localized with A␤ at sites of ongoing clearance in the lower dose 3D6, 12A11, and 10D5 groups. Collectively, these observations showed that removal of CAA at lower doses of 3D6 minimized microhemorrhage over the 6-month study period. These observations suggest that passive immunization can reduce CAA, and modulating antibody dose and duration can significantly mitigate the incidence of microhemorrhage. Immunotherapy can then potentially remove or reverse the CAA that is thought to lead to dysfunction of the microvasculature in AD. O1-06-08

ACTIVATION OF MICROGLIA BY NASAL VACCINATION WITH A PROTEOSOME BASED ADJUVANT AND GLATIRAMER ACETATE CLEARS A-BETA IN AN ANTIBODY INDEPENDENT FASHION IN A MOUSE MODEL OF ALZHEIMER’S DISEASE

Howard L. Weiner1, Sanja Petrovic1, Weiming Xia1, Matthew Goldberg1, Jose Vega1, David S. Burt2, Dan Frenkel1, 1 Brigham and Women’s Hospital and Harvard Medical School, Boston, MA, USA; 2Glaxo Smith Kline, Laval, PQ, Canada. Contact e-mail: [email protected] Background: Immune therapy in Alzheimer’s Disease (AD) models by Ab immunization or administration of anti-Ab antibodies reduces Ab levels and reverses behavioral impairment. A human trial of Ab immunization led to meningoencephalitis in some patients and was discontinued. Objectives: We tested whether a nasal vaccination strategy that did not involve immunization with amyloid or depend on anti- Ab antibodies and was designed to activate microglia could clear Ab when given as prevention to young mice. Methods: Age- and sex-matched littermates (APP-J20 mice) received a weekly nasal treatment with Protollin (a proteosome-based adju-

vant used in human influenza vaccination) 1ug/treatment, plus glatiramer acetate (GA, an approved drug for multiple sclerosis) 25ug/treatment, beginning at age 5 months for a duration of 8 months. Results: We found a significant reduction in the level of amyloid fibril (⬃80%) in the brains of GA⫹Protollin treated mice versus controls (P ⬍ 0.001). Furthermore, the reduction of amyloid load involved the reduction of both (50%) soluble Ab and (49%) insoluble form of Ab (p⬍0.02) as measured by ELISA and was associated with elevation of total serum Ab levels. The treated animals exhibited no toxicity as measured by body weight, eating habits, tail tone, or mobility. Clearance of Ab was also observed with nasal Protollin alone and direct intrahippocampal injection of Protollin activated microglia and reduced Ab plaques. We also found reduction in astrocytosis following nasal vaccination with GA⫹Protollin or Protollin alone. We previously found that nasal vaccination with GA⫹Protollin, decreased Ab plaques in 14 month old APP Tg mice; vaccinated animals developed activated microglia (CD11b⫹ cells) that co-localized with Ab fibrils, and the extent of microglial activation correlated with the decrease in Ab fibrils (JCI: 115:2423, 2005). In this study, after chronic treatment for 8 months we observed no difference in microglial activation, suggesting that once Ab is cleared there is downregulation of microglial activation. Conclusion: Our results demonstrate an antibody-independent therapeutic approach for the treatment of AD which targets microglial cells and is also effective, without toxicity, when given chronically to young mice as prevention. MONDAY, JULY 17, 2006 PLENARY PL2 PL-02-01

DEVELOPMENT OF BIOLOGICAL MARKERS OF ALZHEIMER’S DISEASE

Harald Hampel, Alzheimer Memorial Center, University of Munich, Munich, Germany. Contact e-mail: [email protected] Objectives: The pathophysiologic process leading to neurodegeneration in Alzheimer’s disease (AD) is thought to begin long before clinical symptoms develop. Existing therapeutics for AD improve symptoms, but increasing efforts are being directed toward the development of therapies to impede the pathologic progression of the disease. Although these medications must ultimately demonstrate efficacy in slowing clinical decline, there is a critical need for biomarkers that will stratify pre-clinical and clinical patient populations for trials, indicate whether a candidate disease-modifying therapeutic agent is actually altering the underlying degenerative process. Methods and Results: A number of in vivo neurochemistry and neuroimaging techniques, which can reliably assess aspects of physiology, pathology, chemistry, and neuroanatomy, hold promise as biomarkers. These neurobiological measures appear to relate closely to pathophysiological, neuropathological and clinical data, such as hyperphosphorylation of tau, abeta metabolism, rate of atrophy and cognitive decline, as well as risk of future decline. As this work has considerably matured, it has become clear that biological measures may serve a variety of potential roles in early clinical and pre-clinical diagnosis, clinical trials of candidate therapeutic agents for AD, depending in part on the question of interest and phase of drug development. Conclusions: In this presentation, the conceptual framework of current multimodal biomarker research is reviewed, as well as data related to the range of core feasible neurochemical, as well as neuroimaging markers of AD and potential applications of these techniques in future clinical practice and in clinical studies, particularly with respect to early diagnosis, patient stratification, classification, prediction, as well as the monitoring of disease progression in trials of disease-modifying therapies. REFERENCES 1) Blennow K, Hampel H, Lancet Neurol. 2003,10,2(10):605-613 2) Frank R et al., Neurobiol Aging. 2003,7-8, 24(4):521-36

Symposia S2-01: Progress in Genetic Research 3) Frank R, Hargreaves, Nat Rev Drug Discov. 2003,7; 2(7):566-80 4) Smith D, Proc Natl Acad Sci U S A. 2002,4,2; 99(7): 4135-4137 5) Ashburner J et al., Lancet Neurol. 2003, 2;2(2):79-88 PL-02-02

BETA-SECRETASE AS A THERAPEUTIC TARGET

Martin O. Citron, Amgen, Thousand Oaks, CA, USA. Contact e-mail: [email protected] Finding inhibitors of A␤42 generation is a major goal of Alzheimer’s disease drug development. Two target protease activities, ␤-and ␥-secretase, were operationally defined in the early 90s, but progress in this area was slow, because the actual enzymes were not understood at the molecular level. Some years ago we identified a novel membrane bound aspartic protease, BACE1, as ␤-secretase. This finding has been confirmed and BACE1 and its homolog BACE2 have been characterized in detail by many groups. Major progress has been made in two areas: First, the x-ray crystal structure, which is critical for rational inhibitor design, has been solved and shown to be similar to that of other pepsin family members. Second, knockout studies show that BACE1 is critical for A␤ generation, but the knockout mice show an otherwise normal phenotype, raising the possibility that therapeutic BACE1 inhibition could be accomplished without major mechanism based toxicity. However, target-mediated toxicity of ␤-secretase inhibition cannot be ruled out based on the currently available data alone. While various peptidic ␤-secretase inhibitors have been published, the key challenge now is the generation of more drug-like compounds that could be developed for therapeutic purposes. Other current areas of investigation, including identification of additional BACE1 substrates, the potential role of BACE1 overexpression in AD and the phenotype of BACE2 knockout mice will be discussed. PL-02-03

TAU AND TAUOPATHIES

Jesus Avila, Universidad Autonoma de Madrid, Madrid, Spain. Contact e-mail: [email protected] Background: Tau, a microtubule associated protein, can aberrantly polymerize, in phosphorylated form, yielding the paired helical filaments found in the brain of Alzheimer’s disease patients. Objective(s): Our purpose is to know tau pathology related with its phosphorylation and its assembly. Methods: In vitro analysis and the use of mouse models. Conclusions: Tau assembly can be reproduced in vitro by mixing tau protein with polymerization inducers like heparin or Coenzyme Q0, being the assembly of phosphotau facilitated in the presence of Co.Q0. Polymerization of tau has been also mimified by using transgenic mouse models. In these models, human tau, bearing some of the mutations found in patients with FTDP-17, was expressed. Mutations on APP and/or PS-1 will facilitate tau phosphorylation by kinases like GSK3. Thus, a transgenic mouse model overexpressing GSK3 was also characterized. In some of these mouse models, a link between tau phosphorylation and tau assembly has been established. Finally, the possible toxic effect of phosphotau or tau aggregates will be discussed. MONDAY, JULY 17, 2006 SYMPOSIA S2-01 PROGRESS IN GENETIC RESEARCH S2-01-01

SYSTEMATIC META-ANALYSIS OF ALZHEIMER’S DISEASE GENETIC ASSOCIATION STUDIES - THE ALZGENE DATABASE

S23

suggest that the common late-onset form of AD is caused by a variety of independent genetic (and non-genetic) risk factors. In the past decade, literally hundreds of reports have been published claiming or refuting genetic association with putative AD genes, and currently nearly 10 association studies are published each month. For the research community as well as the public this wealth of information has become increasingly difficult to follow, evaluate and—most importantly—to interpret. To better understand these findings, we have created a comprehensive and publicly available database (“AlzGene”; hosted by the Alzheimer Research Forum, available at www.alzgene.org) which systematically collects and summarizes all genetic association studies in the field of AD. In addition to displaying essential details for each study (e.g., sample sizes, onset age, genotype numbers), the database provides up-to-date summary effect size estimates and assessments of publication bias for each polymorphism across all studies. The results of these meta-analyses should help to better distinguish the relevant AD risk genes from false-positive reports as well as those with negligible effects. Currently, the AlzGene database includes nearly 1,000 individual studies of over 300 different candidate genes. At the meeting we will present a comprehensive summary of these metaanalyses with a particular focus on genes which significantly increase or decrease the risk for AD. S2-01-02

IDENTIFICATION OF NOVEL GENES AND MUTATIONS IN ALZHEIMER DISEASE

Christine Van Broeckhoven, VIB8 - Department of Molecular Genetics, Neurodegenerative Brain Diseases Group, University of Antwerp, Antwerpen, Belgium. Contact e-mail: [email protected] Background: Alzheimer disease (AD) is genetically heterogeneous with both causal and susceptibility genes contributing to its genetic etiology. The current genes and mutations explain approximately 80%, indicating that several other genetic factors remain to be identified. Objectives: To identify novel genes and mutation mechanisms underlying the risk for both early- and late-onset AD. Methods: We systematically sample patients with AD using a standardized protocol for phenotyping including an extensive clinical examination, memory tests and brain imaging. Also biochemical data is collected of biomarkers analyzed in cerebrospinal fluid and serum, and in autopsied brain. In all patients we determine the APOE genotype. In patients with onset age ⬍70 years mutation analysis is performed of 5 dementia genes: APP, PSEN1, PSEN2, MAPT and PRNP. Of each index patient we sample the spouse and children, and in case of a positive family history all cooperative family members are sampled for genome-wide segregation and association studies using STR and/or SNP markers. Results: In a series of 180 early-onset AD patients we identified 1 APP mutation (Val717Ile, 0.5%), 3 APP promoter mutations that increase APP transcription (1.7%) but no APP genomic duplication mutations. Of some milder APP promoter mutations the frequency increased with onset age. We also identified 6 mutations in PSEN1 and PSEN2 (3.3%). In 1 AD patient, with an onset age of 69 years, we detected a PRNP octapeptide insertion and in MAPT we found the intron 10 (IVS10⫹29G⬎A) variant of which the pathogenic nature is still controversial. Conclusions: Together, our data suggest that the frequency of mutations in the five dementia genes is around 6%. Of interest is the observation of several mutations in patients with onset ⬎65 years and duration of disease in old age, suggesting that some mutations have a milder pathogenic effect. S2-01-03

SYSTEMATIC SCREENS IDENTIFY NOVEL GENES FOR LOAD SUSCEPTIBILITY ON CHROMOSOMES 9 AND 10

Lars Bertram1, Matthew B. McQueen2, Kristina Mullin1, Deborah Blacker1,2, Rudolph E. Tanzi1, 1Massachusetts General Hospital, Charlestown, MA, USA; 2Harvard School of Public Health, Boston, MA, USA. Contact e-mail: [email protected]

Alison M. Goate, Washington University School of Medicine, St. Louis, MO, USA. Contact e-mail: [email protected]

Alzheimer’s disease (AD) is a genetically complex and heterogeneous disorder showing an age-dependent dichotomy. Several lines of evidence

Background: Late-onset AD (LOAD) probably results from the combined effects of variation in a number of genes as well as environmental factors.