In-situ characterization of cell injury in Diabetes using tissue cytometry and machine learning
Tarek Ashkar   (Indianapolis, IN)
There is a fundamental gap in understanding the mechanisms and the pathobiology of human diabetic nephropathy (DN). The presence of this knowledge gap presents an important problem because specific therapeutic interventions to treat or slow disease progression cannot be fully accomplished until this gap is filled. Our long-term goal is to characterize key cellular and molecular pathways underlying DN, and to identify patterns of injury that promulgate disease progression so that specific therapeutic interventions could be developed. The objective of this application is to comprehensively characterize cell injury induced by diabetes in situ, using innovative approaches in imaging and image analytics applied on biopsies with diabetic kidney disease. The central hypothesis of this application is that diabetes induces a spatially-anchored and biologically interpretable pattern of cell injury across the entire nephron and the associated interstitium. The rationale for the proposed research is that an imaging-based detection and classification of injury allows the preservation of tissue architecture and defining the spatial context of each cell, thereby enhancing the ability to interpret how specific injury across different cell types is linked to disease. The hypothesis will be tested the following two specific aims: 1) Define the landscape of diabetes-induced cell injury in situ using an imaging-based approach that combines tissue cytometry and machine learning and 2) Establish that tubular injury in diabetes differentially affects thick ascending limb (TAL) cells. At the completion of this project, we expect to define a spatially- anchored extended injury profile for the major cell types in diabetic kidney disease and demonstrate a differential susceptibility to diabetes-induced injury for specific cells in various compartments. These results will have an important positive impact because they are expected to define subpopulations of injured cells induced by diabetes, which could complement other ongoing efforts to define the transcriptomic profile of cell injury in diabetes. These results will fundamentally advance our mechanistic understanding of diabetic kidney disease because of linking cell injury to specific renal microenvironments.