We propose to use new methods to study, for the first time, the proteolysis of neurofilament and other cytoskeletal proteins in postmortem human brain during normal aging and in Alzheimer's diseas (AD). In addition to increasing our understanding of this complex process in normal and aging brain, these experiments will provide information pertinent to the basic mechanism underlying the formation of neurofibrillary lessions and the degeneration of neurons in AD and other human neurodegenerative disorders. We have obtained evidence supporting our hypothesis that proteolysis is defective in AD brain due to abnormal proteolytic enzyme systems and/or to abnormally modified protein substrates that are resistant to degradation. Having developed methods to purify the four major proteinases from human brain (Ca++-activated neural proteinase(s) [CANP]; cathepsins D and B; and Ca++- independent neutral proteinases), we will seek age- and AD-related abnormalities in these enzyme systems by characterizing in detail the structural and enzymatic properties (including denaturation rates) of each purified enzyme and its isoenzymes from matched speciments of postmortem prefrontal isocortex from normal adults, aged individuals (Greater than 80 years) and AD patients. Interactions between various CANP forms and a specific regulatory factor purified from brain will also be investigated. Distribution of the activity and the content of each human brain proteinase will be measured by radioassay and radioimmunoassay respectively, in 15-20 selected regions of control, aged, and AD brains and correlated with the severity of neurofibrillary pathology quantitated morphometrically and biochemically. Based on our findings that paired helical filaments (PHF) in AD brain are resistent to digestion by brain proteinases, we will seek differences between normal and AD brain in the structure of specific cytoskeletal proteins by first studying the kinetics of degradation by each purified proteinase. Using monoclonal and polyclonal antibodies to individual NFPs and novel 2-D immunblot approaches, we will then compare in control and AD cases the patterns of NFP-immunreactive proteolytic products present in unincubated tissue or generated by either purified brain proteinases in vitro or by corresponding neuronal proteinases in situ within intact brain microslices. Finally, the susceptibility of PHF and recently observed PHF-immunoreactive (? precursor) variant forms to purified proteinases will be further investigated by immunoblot analyses using anti-PHF antibodies.
