Henry C. Powell, MD, PhD, DSci, FRCPath - US grants

The pathology of experimental diabetic neuropathy can be divided into an early phase characterized by reduced motor nerve conduction velocity, reduction in diameter of nerve fibers and late changes associated with degeneration and loss of myelinated nerve fibers and microangiopathy. To explain early changes, occurring within weeks of induction of streptozotocin diabetes, we suggest that increased glucose and sorbitol induce osmotic and electrolyte disturbances which affect nerve conduction while the late changes (1-2 years) may be substantially attributable to ischemia caused by thickening of walls of small blood vessels (microangiopathy) and increased blood viscosity, the combined effects of which reduce nerve blood flow. To test the first hypothesis, we will measure endoneurial fluid electrolyte concentration with a new technique developed for this purpose, relating the findings to chemical, electrophysiologic and morphologic parameters. Since early diabetes affects the nerve interstitium, analysis of endoneurial fluid concentrations of sorbitol and albumin are necessary to detect changes which can produce edema. Glycosylation of proteins occurs in diabetic tissue and may affect the nerve in two ways; glycosylated albumin causes avid pinocytosis allowing it to cross the blood nerve barrier and penetrate the endoneurium, glycosylation of myelin may produce structural and functional changes affecting conduction. To evaluate the ischemic hypothesis, we have developed a method for measuring nerve blood flow by a non-invasive technique suitable for small volumes such as the rat sciatic nerve. By correlating in vivo measurements of nerve blood flow with postmortem morphologic analysis, we hope to assess the significance of the ischemia in chronic diabetic neuropathy. Morphologic studies of microangiopathy will be continued in human diabetic nerve, from which our preliminary evidence suggests that the unique anatomic configuration of nerve blood supply further complicates vascular narrowing due to endothelial proliferation and basal laminar thickening. By coordinating new experimental techniques with established methods for investigating neuropathy, we hope to understand how metabolic and vascular changes in the interstitial microenvironment interact, causing complex alterations of function and structure known as diabetic neuropathy.

One of the major objectives of the present study is to isolate the putative beta protein precursor(s) from Alzheimer's disease, Down's syndrome and normal sera, cerebrospinal fluid, urine or tissue and to define if chemical differences exist, i.e. an "amyloidogenic" isotype in Alzheimer's disease and Down's syndrome. The identification of such an isotype may lead to a specific diagnostic serum or CSF test for Alzheimer's disease by the use of a radioimmunoassay. Another major objective is to define the proteolytic enzyme characteristics of cerebral vessels that are implicated in the cleavage of beta protein precursor to form amyloid fibrils in cerebral vessel walls and those responsible for its proteolysis to form the amyloid fibrils of "senile" plaques, in order to define proteolytic enzyme differences between cerebral parenchymal cells, e.g. microglia, and endothelial cells, that may explain the differences in vessel and plaque amyloid fibril chemical composition. This study will utilize biochemical procedures extensively for the isolation and characterization of the vascular complement of lysosomal and other proteolytic enzymes. Organotypic endothelial cell cultures will also be employed in this investigation. In the absence of a chemically characterized native amyloid beta protein precursor, the beta protein gene product and its hydrolytic cleavage derivatives will be used as enzyme substrates. The ultimate purpose of this study is to develop reagents selective for the inhibition of cerebral amyloid fibril formation by the prevention of beta protein precursor hydrolysis and, thus, lead to an approach to the therapy of Alzheimer's disease. In addition in vitro radioautographic receptor binding methods will be employed to determine whether the immunohistochemical definition of the beta protein precursor as a receptor-binding protein can be verified. A final objective is to define by chemical, immunochemical and immunohistochemical methods the nature and origin of the paired helical filaments of the neurofibrillary tangles in Alzheimer's disease.