PROJECT SUMMARY The first two cycles of our R01 entitled, “Understanding the role of TDP-43 in Alzheimer's disease and FTLD” were very successful with over 50 peer-reviewed publications and many novel discoveries. In the 2nd cycle we focused mainly on how TDP-43, tau and beta-amyloid contributed to neurodegeneration in those with Alzheimer's disease neuropathologic changes (ADNC). Our studies were conducted on a cohort of 756 ADNC cases. We discovered that TDP-43 deposition in ADNC was heterogeneous and for the first time showed that there are two distinct types of TDP-43 deposition in ADNC that we termed TDP type-α and type-β. What is most interesting about this discovery is that TDP type-α have strikingly similar features to one pathological type of FTLD-TDP (FTLD-TDP type A) while TDP type-β is strongly associated with the presence of neurofibrillary tangles, and hence tau. A second discovery was that different inclusions associated with the different pathological types of FTLD-TDP (type A, type B and type C) had different molecular compositions of TDP-43 specie. For example, while most inclusions consisted of C-terminal fragments of TDP-43, pre-inclusions and perivascular inclusions consisted of a greater burden of full-length TDP-43 than C-terminal fragments. Therefore, one of the main goals of the 3rd cycle is to further assess how ADNC TDP types (type-α and type-β) are related to FTLD-TDP types (type-A, type-B, type-C) by investigating associations with the different molecular specie of TDP-43 including C-terminal fragments, full-length and phosphorylated TDP-43. A second goal of the R01 is to further our findings from the 2nd cycle in order to better understand how TDP types (type-α and type-β) modifies the associations between TDP-43 and neurodegeneration. We will investigate, specifically, how TDP, including TDP types, tau and beta-amyloid and other pathological and genetic factors are associated with trajectories of volume loss of hippocampus and neocortex over time. Our cohort will consist of cases with multiple head MRI scans (range: 2-13 yearly MRI scans; total scans=2316) prior to death. Last, but not least, we will address one of the biggest knowledge gaps in the field, the lack of a biomarker that can help predict the presence of TDP-43 in ADNC during life. We will use [18F]fluorodeoxyglucose PET, which we have shown is superior to MRI, to predict TDP-43. We will first develop the biomarker and then test it in an independent cohort. In order to accomplish all the aims of the 3rd cycle we will upgrade our cohort size to 1303 pathologically confirmed cases of ADNC and FTLD. Findings from this 3rd cycle will significantly improve understanding of TDP-43, and TDP-43 types, and help to address the nagging issue of where the boundary lies between TDP-43 in ADNC and FTLD-TDP. Given that TDP-43 is a potential therapeutic target for the treatment of FTLD and now Alzheimer's disease, the 3rd cycle will likely also have a significant impact...