Kidney cell response to acute cardiorenal and isolated kidney ischemia-reperfusion injury

Abstract

Acute cardiorenal syndrome (CRS) represents a critical intersection of cardiac and renal dysfunction with profound clinical implications. Despite its significance, the molecular underpinnings that mediate cellular responses within the kidney during CRS remain inadequately understood. We used single nucleus RNA sequencing (snRNAseq) to dissect the cellular transcriptomic landscape of the kidney following a translational model of CRS, cardiac arrest/cardiopulmonary resuscitation (CA/CPR) in comparison to ischemia-reperfusion injury (IRI). In each dataset, we found that proximal tubule (PT) cells of the kidney undergo significant gene expression changes, with decreased expression of genes critically important for cell identity and function, indicative of dedifferentiation. Based on this, we created a novel score to capture the dedifferentiation state of each kidney cell population and found that certain epithelial cell populations, such as the PT S1 and S2 segments, as well as the distal convoluted tubule, exhibited significant dedifferentiation response. Interestingly, the dedifferentiation response in the distal nephron differed in magnitude between IRI and CA/CPR. Gene set enrichment analysis (GSEA) of PT response to IRI and CA/CPR revealed similarities between the two models and key differences, including enrichment of immune system process genes. Transcriptional changes in both mouse models of acute kidney injury (AKI) highly correlated with a dataset of human biopsies from patients diagnosed with AKI. This comprehensive single-nucleus transcriptomic profiling provides valuable insights into the cellular mechanisms driving CRS.NEW & NOTEWORTHY Cardiac dysfunction is a common cause of acute kidney injury in a malady called acute cardiorenal syndrome. In a mouse model of acute cardiorenal syndrome called cardiac arrest/cardiopulmonary resuscitation, we characterized, for the first time, the kidney transcriptional landscape at the single-cell level. We developed a novel method for quantifying cell response to injury and found that cells adapted through dedifferentiation, the magnitude of which varied depending on cell type.

Keywords: acute cardiorenal syndrome; acute kidney injury; ischemia-reperfusion; single nucleus RNA sequencing; transcriptomics.

Figure 1.. CA/CPR experimental model and parameters.

A) Cardiac arrest was induced with potassium chloride (KCl) in instrumented mice and continued for 8 minutes. Core temperature was maintained between 36.5 and 37.5°C during cardiac arrest. Resuscitation with epinephrine and chest compressions (CP) resulted in return of spontaneous circulation (ROSC), after which mice were survived for 24h prior to kidney harvest. B-D) Body weight, duration of chest compressions (CPR time), and epinephrine dose for the 2 animals used for this study are superimposed (red dots) on data from >200 mice subjected to this model by the same surgeon. E) Medulla-cortex immunofluorescence micrographs from representative mouse kidney 24h after sham or CA/CPR demonstrate that CA/CPR sharply increases cortical kidney injury molecule 1 (KIM-1). Proximal tubule denoted by lotus tetragonolobus lectin stain (LTL). Scale bars=100 μm.

Figure 2.. Transcriptional Landscape of CA/CPR.

A) UMAP of kidney cell populations captured by snRNASeq. B) Multidimensional dotplot of cell populations and markers used for identification. C) UMAP grouped by Sham or CACPR sample. D) UMAP split by Sham and CACPR sample. E) Feature plot for Havcr1 split by sham or CACPR. F) Violin plot for Havcr1 split by sham and CACPR.

Figure 3.. Comparison of CA/CPR to IRI.

A) UMAP of kidney cell populations captured by snRNASeq for the combined dataset of CA/CPR and IRI experiments. B) Multidimensional dotplot of cell populations and markers used to identify them. C) Feature plot for Havcr1 split by CA/CPR sham, CACPR, IRI sham, or IRI sample. D) Violin plot for Havcr1 split by CA/CPR sham, CACPR, IRI sham, or IRI sample.

Figure 4.. Cluster and Condition Comparison Between CA/CPR and IRI.

A) Heatmap of the correlation coefficient for the differentially expressed genes that defined each cluster between CA/CPR and IRI samples. B) Representative scatter plot of the log2FC of genes that defined each cluster for Proximal Tubule (PT) clusters between CACPR and IRI. C) Heatmap of the correlation coefficient for the differentially expressed genes that defined each cells response to injury (Condition) CA/CPR and IRI samples. D) Representative scatter plot of the log2FC of genes that defined each cells response to injury for PT clusters between CACPR and IRI. Genes that have log2FC > 1 are colored green and genes that have a log2FC < 1 are colored orange.

Figure 5.. Proximal Tubule Injury in CA/CPR and IRI.

A) UMAP of PT cells in CA/CPR and IRI experiments. B) Multidimensional dotplot of PT cell populations and the markers used to identify them. C) UMAP proximal tubule CA/CPR and IRI experiments split by CA/CPR sham, CACPR, IRI sham, or IRI. D) Proportion table of PT populations by CA/CPR sham, CACPR, IRI sham, or IRI. E) Feature plots for Lrp2, Hsp1a, and Havcr1 split by CA/CPR sham, CACPR, IRI sham, or IRI sample E) Violin plot for Havcr1 split by CA/CPR sham, CACPR, IRI sham, or IRI sample.

Figure 6.. Pathways analysis of CA/CPR and IRI Proximal Tubule.

A) Results of Gene Set Enrichment Analysis for CA/CPR and IRI PT cells. B) Venn Diagram of pathways that were shared or unique to CA/CPR (Red) or IRI (Purple). C) List of the top 10 pathways that were either unique or shared between CA/CPR and IRI.

Figure 7.. Comparison of CA/CPR and IRI PT DEGs to KPMP.

A) Scatter plot of the log2FC of genes that defined each cells response to injury for PT between CA/CPR and the KPMP Acute Kidney Injury (AKI) snRNAseq dataset. B) Scatter plot of the log2FC of genes that defined each cells response to injury for PT between IRI and the KPMP AKI snRNAseq dataset. Genes that have log2FC > 1 are colored green and genes that have a log2FC < 1 are colored orange.

Figure 8.. Cell State Differentiation after CA/CPR and IRI.

A) Diagram of the characterization of the dedifferentiation score. B) Proportion plot of CA/CPR cell populations according to their dedifferentiation state. C) Proportion plot of IRI cell populations according to their dedifferentiation state. D) UMAP of CA/CPR Sham and CA/CPR datasets colored by dedifferentiation category. E) UMAP of IRI Sham and IRI datasets colored by dedifferentiation category. F) Diagram of the characterization of the Adaptation score. G) Proportion plot of CA/CPR cell populations according to their adaptation state. H) Proportion plot of IRI cell populations according to their dedifferentiation state. I) UMAP of CA/CPR Sham and CA/CPR datasets colored by dedifferentiation category. J) UMAP of IRI Sham and IRI datasets colored by adaptation category.