Title of project

SPARK – Systems Proteomics Advancing Research in Diabetic Kidney Disease

Abstract

Kidney complications in Type 2 diabetes mellitus (T2DM) can result from diabetes itself, known as diabetic kidney disease (DKD), or indirectly through multifactorial pathways. Renal biopsy remains the gold standard for DKD diagnosis and differentiating it from other type of Kidney disease, it is an invasive procedure. Traditionally, DKD is seen as a linear complication of chronic hyperglycaemia, starting with albuminuria and progressing to reduced eGFR. However, evidence shows DKD is heterogeneous since kidney function may remain preserved despite albuminuria and some patients present with non albuminuric DKD. Early pathology involves glomerular as well as tubular and interstitial changes, contributing to this heterogeneity. Further, this heterogeneity plays distinct roles in risk prediction and therapeutic intervention.

Against this background, we aim to identify tissue proteomic markers using spatial proteomics in individuals with type 2 diabetes and albuminuria with preserved kidney function, and to investigate corresponding urine proteomic patterns to develop a non-invasive approach for DKD risk stratification. This will enhance diagnostic precision and lay the foundation for personalized treatment approaches in resolving DKD heterogeneity.

Through this project, we propose to to generate a spatially resolved proteomic atlas of DKD(from the PRIMETIME2 study cohort at SDCC, which includes a renal biopsy-proven group of Type 2 Diabetes subjects with proteinuria) and control subjects (from the PRIMETIME4 cohort, in which kidney tissues were collected during nephrectomy at SDCC), mapping protein localization at subcellular resolution to uncover disease-specific alterations in cellular organization. This will allow the identification of tissue-derived signatures for stratifying DKD subjects. Further to identify non-invasive proxies for tissue-based signatures through specific urine proteomic patterns targeted proteomics screening using OLINK proteomics (PEA technology) method will be used. This will potentially serve as a non-invasive tool to provide insights into markers for the precision diagnosis of DKD. Tissue and urine proteomics data will be used to build a machine learning classifier for detecting diabetic kidney disease (DKD) endotypes. Pathological findings and spatial proteomics will complement urine data, enabling development of clinically relevant, non-invasive DKD risk prediction tools. This result will be validated in an external cohort to ensure robustness of the findings. This approach will help to identify DKD outcomes specific clinically relevant non-invasive risk scoring/prediction methods within the clinic.