There’s a grim correlation between diabetes and end-stage kidney disease (ESKD): in fact, diabetes is ESKD's leading cause worldwide (1). Diabetic kidney disease (DKD) also occurs in a significant proportion of diabetic individuals (40%), adding the possibility of kidney failure and artificial blood filtering (dialysis) requirements to the challenges of insulin management that affect diabetic individuals’ quality of life and health outcomes (1).
The good news is there are promising emerging treatments for ESKD and DKD. These include a class of drugs called sodium-glucose cotransporter-2 inhibitors (SGLT2i), which lower blood sugar levels by preventing the kidneys from reabsorbing excess sugar and helping to alleviate kidney damage (2).
Despite these benefits, scientists still only have limited knowledge of the SGLT2i drug class' mechanism of action. Moreover, the cellular and molecular drivers of DKD are blurry, pointing to a great need to both define disease mechanisms and develop treatment programs that can optimally target the underlying biology of DKD (1).
This gap in our current understanding of diabetic kidney disease led a team of researchers from Washington University, St. Louis, in collaboration with Janssen Research & Development (Pharmaceutical Companies of Johnson & Johnson), to perform a 1 million single cell atlasing study of a murine DKD model under different treatment regimens (1). Their study, published in Cell Metabolism, represents the largest single cell atlas of mouse kidney to date. Mice were batched into seven treatment groups, representing disease-free controls, untreated mice, and single treatments or combination treatments using three drugs prescribed for diabetes and kidney disease: lisinopril (ACEi), rosiglitazone (Rosi), and JNJ-39933673, an SGLT2 inhibitor (specifically, ACEi+Rosi and ACEi+SGLT2i combination treatments were tested).
Right away, histological data suggested that combination treatments were the most effective to rescue kidney pathological changes. To go deeper into how treatments were affecting specific cell types within the kidney, the team used single nucleus RNA-sequencing (snRNA-seq) to profile kidney tissues collected two days and two weeks after treatment. They analyzed a total of 946,660 single cells, which were further classified into 18 major kidney cell types and, importantly, provided resolution of rare cell types like macula densa cells, among others (1).
Cell type–specific insights reveal disease mechanisms and treatment responses
Given the substantial heterogeneity of kidney cells, the team first investigated how diabetes pathology uniquely affected specific kidney cell types. They performed differential gene expression analysis between cells from diseased and control mice cohorts, revealing the cell types with the highest number of differentially expressed genes during disease progression: principal cells (PCs), thick ascending limb cells (TALs), parietal epithelial cells (PECs), proximal tubular cells (PTs), and endothelial cells (ECs).
Their analysis also identified a number of upregulated genes that could serve as disease biomarkers for DKD. Mapping disease traits identified by genome-wide association studies back to cell-specific gene expression unexpectedly revealed a strong association between PCs and ECs and the estimated glomerular filtration rate (eGFR), which is a measure of how well the kidneys are clearing excess fluid and waste from the blood. This may point to a more prominent, and previously uncharacterized, role of PCs in diabetes pathology.
The team sought to take their cell type–specific insights one step further and understand how specific kidney cell populations respond to DKD treatment regimens—which could reveal clues to what drug combinations worked best to prevent progression of disease.
Their results reflected, again, a highly heterogeneous and cell-specific transcriptional response: the genes and pathways that were corrected by each unique treatment regimen were very different from one another, with few shared genes between regimens.
They also discovered a population of proximal tubular cells (which optimize reabsorption and secretion of certain glomerular filtrates in the process of making urine) expressing injury markers, including Havcr1. Combination drug treatments were effective to reduce both the expression of the injury marker Havcr1 in DKD and the overall proportion of these injured cells, while single treatments were not. Validated in mice, this powerful finding has implications for human disease manifestations and treatment too, as the team was able to take cell type–specific molecular signatures from their single cell data and confirm the presence of this injured PT cell state in human bulk RNA-seq datasets of DKD (1).
Finding more options for patients with diabetic kidney disease
Though emerging therapies to treat diabetic kidney disease have shown great promise in recent years, reducing ESKD risk by one-third, the remaining two-thirds of ‘‘at risk’’ trial subjects, whether on ACEi or SGLT2i drugs, continue to progress to ESKD (1). This study provides detailed, cell-specific insights into disease mechanisms and treatment responses that justify combination therapies—and can support future development of novel therapies to fill in the gaps of current treatment regimens.
And this is just one study among the many that are ongoing or still to come. Other scientists are adding their findings to build a more detailed understanding of diabetic kidney disease and disease-associated cell states—including a team from the University of Pennsylvania, who released a pre-print featuring single cell RNA-seq, ATAC-seq, and Visium Spatial data on human diabetic kidney tissue.
In honor of World Diabetes Day, we celebrate all of the scientists and medical professionals who are working hard to understand the underlying mechanisms of diabetic kidney disease and find new and improved treatment options for patients.
- Wu H, et al. Mapping the single-cell transcriptomic response of murine diabetic kidney disease to therapies. Cell Metab 34: 1064–1078.e6 (2022). doi: 10.1016/j.cmet.2022.05.010