Not all controls are made equal: Definition of human kidney reference samples by single cell gene expression profiles.

Unlabelled: Identifying kidney disease mechanisms often requires comparing samples from disease states with healthy reference tissues. However, the effect of variations in sample procurement, storage and donor baseline characteristics of reference samples has thus far not been evaluated. Three distinct kidney reference sample types were evaluated for integrity and injury biomarkers and in their ability to define differentially expressed genes (DEGs) when compared to three different diabetic kidney disease (DKD) states. Unaffected parts of tumor nephrectomies (TN), pre-transplant living donor biopsies (LD), and percutaneous kidney research biopsies from healthy volunteers (HC) served as sources for reference tissue. Single cell gene expression profiles showed differences in the expression of injury or disease markers and the proportion of immune and proximal cell states. TN exhibited the highest expression of early stress response genes. A gene set associated with procurement effect in post-operative biopsies (LD and TN) was identified. An age-associated transcriptional signature was extracted from the reference data. Providing these tools to control for age and tissue procurement effects, immune-related pathways were found to be most enriched in DKD when compared to HC. Energy-related processes were enriched in DEGs from DKD versus LD. TN samples exhibited more underlying pathology than LD. The pathway analyses using the DEGs underscore the importance of accounting for appropriate confounding factors in differential expression analyses between disease and reference samples. Comparable controls are essential for appropriate molecular evaluation of pathologic tissues.

Translational Statement: Integrated single-cell data analysis of three reference sample types-needle biopsy from young healthy kidney tissue, pre-perfusion biopsy from transplant kidneys, and cancer-free tissue from tumor-nephrectomies-revealed distinct transcriptional profiles influenced by the biopsy procurement method and age. These differences impacted findings in diabetes-related kidney disease versus reference comparisons highlighting the need and providing the tools to account for these differences in interpreting analyses and identifying disease mechanisms.