and apoE3 in the e3/4 mouse brains. ApoE4 represented 30-40 % of the total apoE. Moreover, the absolute amount of apoE3 per allele was similar between e3/3 and e3/4 mice, implying that the reduced levels of total apoE in e3/4 mice can be explained by the reduction in apoE4 levels. In culture medium from e3/4 human astrocytoma or e3/3, e4/4 and e3/4 primary as- trocytes, apoE4 levels were consistently lower than apoE3. Secreted cho- lesterol levels were also lower from e4/4 astrocytes. Pulse-chase experiments showed an enhanced degradation and reduced half-life of newly synthesized apoE4 compared to apoE3. Taken together these data suggest that astrocytes preferentially degrade apoE4, leading to reduced apoE4 secretion and ultimately to reduced brain apoE levels. Next we ex- amined the physiological relevance of these findings. Firstly, as apoE has been shown to enhance the degradation of Ab, we measured the endoge- nous levels of Ab in e4/4 and e3/3 mice. As previously reported for the human population, e4/4 mice had higher levels of Ab42 in their hippocam- pus and cortex. Next we examined the capacity of astrocytic e3/3 and e4/4 to promote neurite outgrowth. Interestingly, and unlike previous reports, we found that astrocytic lipoproteins prepared from e4/4 mice, was able to promote neurite outgrowth and increase synaptophysin expression, but with 10 fold less potency than particles from e3/3 mice. Conclusions: Taken together these data suggest that the low levels of astrocytic apoE/ cholesterol secreted by e4 carriers may directly contribute to the disease progression by reducing apoE’s capacity to promote synaptic repair and Ab clearance. O2-04-03 APOE GENOTYPE AND AMYLOID BETA EFFECTS ON COGNITION ARE MEDIATED BY CELL INJURY (TAU) AND DEATH (PROGRESSIVE ATROPHY) Charles DeCarli, Owen Carmichael, Danielle Harvey, Evan Fletcher, Dan Mungas, University of California at Davis, Sacramento, CA, USA. Contact e-mail: cdecarli@ucdavis.edu Background: The amyloid hypothesis suggests a primary role for Ab, al- though clinical symptoms are most closely associated with neurofibrillary formation and neuronal injury and loss. We hypothesized that CSF tau would therefore mediate the effect of CSF Ab on cognitive performance (Figure). Methods: 313 subjects from the Alzheimer’s Disease Neuroi- maging Initiative (ADNI) for whom baseline CSF, repeated MRI and re- peated Cognitive measures were evaluated. Subjects included 70 AD, 152, MCI and 91 Cognitively normal individuals mean age 74.8 þ 7.0 years. Baseline brain and hippocampal volumes were calculated as well as yearly rates of both measures using a fluid flow deformation method to determine jacobian determinant expansion or contraction for each image voxel. Cognitive measures included 1-year change in delayed story recall and ADAS-Cog. Multiple regression analysis was used to systematically test the effects of CSF Ab, CSF Tau, baseline and change in brain volume on change in cognition at 1 year. Results: Significant associations were found between the rate of change of cognitive measures and ApoE4 geno- type, CSF Ab, CSF Tau, baseline and rate of whole brain atrophy adjust- ing for age, education and gender (see table). The impact of CSF Ab remained significantly associated with change in cognitive measures in a multiple regression model that included both CSF Ab and ApoE geno- type (p <0.0001). CSF Tau was also significantly associated with change in delayed story recall (p¼0.0002) and ADAS-Cog (p<0.0001) when added to the model. In a model predicting change in ADAS-Cog which included baseline and change in whole brain volume as well as CSF Tau and Ab, CSF Tau (p<0.0002), baseline, longitudinal brain atrophy (p<0.0001), but not CSF Ab were significantly associated with change in ADAS-Cog. Conversely, in a model predicting change in delayed story recall, CSF Ab (p < 0.0001), CSF Tau (p¼0.013), baseline hippocampus (p < 0.0001), change in hippocampal (p ¼0.016) and brain volume (p<0.0011) were each significantly associated with rate of change in de- layed story recall. Conclusions: CSF Ab has both direct and mediated ef- fects on specific cognitive domains. Reductions in both CSF Ab and CSF tau, therefore, should be considered good markers of therapeutic efficacy in AD clinical trials. O2-04-04 CHOLESTEROL IN AMYLOID PLAQUES: ANALYSIS BY LASER CAPTURE MICRODISSECTION COMBINED WITH MASS SPECTROMETRY Maı ¨ Panchal 1 , Jacqueline Loeper 2 , Claire Perruchini 3 , Adina Lazar 3 , Heidi Vitrac 3 , Denis Pompon 2 , Charles Duyckaerts 3 , 1 Ho ˆpital de la Salpe ˆ- trie `re, 75013 Paris, France; 2 CNRS-CGM, Gif-sur-Yvette, France; 3 Ho ˆpital de la Salpe ˆtrie `re, Paris, France. Contact e-mail: panchal.ma@gmail.com Background: The presence of senile plaques (SP) in the brain is one of the pathological hallmarks of Alzheimer’s disease (AD). Extensive knowl- edge of the protein components of the SP has been acquired over the years but the lipids associated with amyloid beta-peptide (Ab) deposits remain unknown. Many evidences suggest that cholesterol contributes to the path- ogenesis of AD. Two studies have reported that SP were enriched in cho- lesterol. We provided data suggesting that these results were obtained with poorly specific methods (Lebouvier et al. 2009). The presence of choles- terol in the senile plaque remains to be demonstrated. Methods: Experi- ments were performed on frozen brain sections (10 um-thick sections from samples of first temporal gyrus) from three sporadic Braak stage Oral O2-04: Disease Mechanisms: APOE P110