ous CT carriers. PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19657278 Consistent with previous reports, no significant CLU genotype effects were observed on CSF Ab42 levels within AD patients. In addition, CSF AD biomarkers Ab42 and Tau were not affected by the CLU genotypes in either apoE4 or nonapoE4 control subjects. While CSF CLU protein levels were not impacted by CLU genotypes, Tau/Ab42-confirmed AD patients had significantly higher CSF CLU than non-demented controls, consistent with previous literature findings. Based on these initial results, we hypothesized a potential link between CLU and Tau pathology in AD. CLU is upregulated in the brain of Tg4510 Tau mouse model We next evaluated the expression pattern of CLU MRT-67307 web isoforms in the hippocampus and cortex of Tg4510 mice overexpressing the human mutant P301L Tau associated with frontotemporal dementia. While 5.5 month-old Tg4510 mice displaying Tau pathology and neurodegeneration showed marked upregulation of sCLU protein in the hippocampus, we found that a truncated iCLU form was significantly elevated in the hippocampus of both 2 month-old pre-tangle and 5.5 month-old tangle-bearing Tg4510 in comparison to agematched wild type mice. Unlike 
the clear changes in brain CLU observed with Tau overexpression, Western blot evaluation of hippocampus from Tg2576 amyloid mouse model showed no age-dependent changes in either forms of CLU compared to WT littermate controls. These findings further substantiated our hypothesis for a potential physiologically relevant connection between iCLU and Tau. iCLU is localized to the cytoskeleton fraction and interacts with Tau Because iCLU had been previously detected in both cytosol and nucleus, we first examined the subcellular fractionation of CLU following transfection of full-length or truncated CLU cDNA constructs in HEK 293T cells. Interestingly, we found that most of the,52 kD iCLU was detected in the cytoskeleton/ insoluble fraction. The precursor CLU protein was fractioned primarily to the cytosol, as predicted by ER/Golgi localization. While the majority of sCLU was localized to the membrane fraction, likely associated with secretory Golgi vesicles, some sCLU was also detected in the cytosolic fraction. The primary localization of overexpressed iCLU to the cytoskeleton fraction led us to postulate the possibility of a direct interaction between iCLU and the microtubule-associated protein Tau. To further investigate a potential interaction between CLU and Tau, we used a doxycycline -inducible wild-type Tau HEK 293T cell line transfected with either full-length or truncated CLU cDNA constructs. By overexpressing the truncated CLU construct designed to generate only the iCLU 50 kD isoform followed by pull-down with anti-Tau antibody, we were able to coprecipitate iCLU from iTau-HEK cell lysates. Induction of Tau expression with dox increased Tau interaction with iCLU. In reciprocal co-immunoprecipitation experiments, we also evaluated whether Tau harboring the P301L mutation linked to frontotemporal dementia and overexpressed in Tg4510 mouse Results CLU rs11136000 alters CSF Tau in AD patients To assess potential mechanisms by which CLU confers risk for AD, we examined effects of the CLU rs11136000 SNP on CSF Ab42 and Tau levels, which are reliable disease biomarkers recently used as endophenotypes in AD genetic studies. For this analysis, the AD population was enriched with patients presenting a disease CSF Tau/Ab42 profile, which has a high diagnostic accuracy for AD over other types of dementi
