Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Dec;31(12):2793-2814.
doi: 10.1681/ASN.2020010052. Epub 2020 Oct 28.

Single-Cell Profiling of AKI in a Murine Model Reveals Novel Transcriptional Signatures, Profibrotic Phenotype, and Epithelial-to-Stromal Crosstalk

Affiliations

Single-Cell Profiling of AKI in a Murine Model Reveals Novel Transcriptional Signatures, Profibrotic Phenotype, and Epithelial-to-Stromal Crosstalk

Valeria Rudman-Melnick et al. J Am Soc Nephrol. 2020 Dec.

Abstract

Background: Current management of AKI, a potentially fatal disorder that can also initiate or exacerbate CKD, is merely supportive. Therefore, deeper understanding of the molecular pathways perturbed in AKI is needed to identify targets with potential to lead to improved treatment.

Methods: We performed single-cell RNA sequencing (scRNA-seq) with the clinically relevant unilateral ischemia-reperfusion murine model of AKI at days 1, 2, 4, 7, 11, and 14 after AKI onset. Using real-time quantitative PCR, immunofluorescence, Western blotting, and both chromogenic and single-molecule in situ hybridizations, we validated AKI signatures in multiple experiments.

Results: Our findings show the time course of changing gene expression patterns for multiple AKI stages and all renal cell types. We observed elevated expression of crucial injury response factors-including kidney injury molecule-1 (Kim1), lipocalin 2 (Lcn2), and keratin 8 (Krt8)-and of several novel genes (Ahnak, Sh3bgrl3, and Col18a1) not previously examined in kidney pathologies. AKI induced proximal tubule dedifferentiation, with a pronounced nephrogenic signature represented by Sox4 and Cd24a. Moreover, AKI caused the formation of "mixed-identity cells" (expressing markers of different renal cell types) that are normally seen only during early kidney development. The injured tubules acquired a proinflammatory and profibrotic phenotype; moreover, AKI dramatically modified ligand-receptor crosstalk, with potential pathologic epithelial-to-stromal interactions. Advancing age in AKI onset was associated with maladaptive response and kidney fibrosis.

Conclusions: The scRNA-seq, comprehensive, cell-specific profiles provide a valuable resource for examining molecular pathways that are perturbed in AKI. The results fully define AKI-associated dedifferentiation programs, potential pathologic ligand-receptor crosstalk, novel genes, and the improved injury response in younger mice, and highlight potential targets of kidney injury.

Keywords: acute kidney injury; cellular crosstalk; renal developmental genes; single-cell.

PubMed Disclaimer

Figures

None
Graphical abstract
Figure 1.
Figure 1.
scRNA-seq reveals proximal tubule dedifferentiation and mixed-identity cells in the ischemia reperfusion–induced AKI. (A) UIR procedure scheme 1. (B) The experimental timeline. (C) UMAP plots show the renal cell populations in the control and UIR day 1. UIR resulted in substantial reduction of mature proximal tubule population. (D) CISH validates the AKI-induced proximal tubule dedifferentiation. The differentiated proximal tubule marker Slc34a1 is abundantly expressed in the control kidney and dramatically reduced in UIR day 1, with some remaining expression in the outer cortex. Original magnification, control and UIR day 1, ×4; zoom into the outer cortex, ×40. (E) Kim1 (green), Aqp2 (red), Slc34a1 (purple) RNAscope. Similar to (D), RNAscope validates the reduced expression of Slc34a1 in UIR day 1 kidneys. Conversely, expression of the injury marker Kim1 is greatly increased at UIR day 1. Original magnification, control and UIR day 1, ×4; zoom, ×60. (F) Heatmap shows the relative marker gene expression in UIR day 1 renal cell types. Yellow color represents expression level above the mean, black color represents the mean, and purple/blue represents expression level below the mean. The heatmap shows genes elevated in renal cell populations relative to each other, based on the z-score. AKI induces the elevated ectopic gene expression in the UIR day 1 cell populations. Note the cluster of mixed-identity cells exhibiting particularly high stochastic expression of markers of many different mature cell types. Original magnification, 4×, 2500 μm; 40×, 100 μm; 60×, 25 μm. Related to Supplemental Figures 14 and Supplemental Tables 2 and 3. Endoth, endothelial; CD, collecting duct; Interc, intercalated. (A) Reprinted from ref. , with permission from Elsevier.
Figure 2.
Figure 2.
AKI induces reactivation of the renal developmental program in the adult kidney. (A) Representative images of Umod (red), Slc34a1 (purple), and DAPI (blue) RNAscope. Note the ectopic expression of the loop of Henle marker gene Umod in the Slc34a1-positive tubules shown with the yellow pointers. Original magnification, ×60, 0.03 μm Nyquist zoom, scale 10 μm, maximal intensity projection from approximately 6-μm Z-stacks. (B and C) Imaris quantification of Umod and Aqp2 transcripts in the Slc34a1-positive UIR day 1 versus control tubules, 12 Z-stacks (50–70 tubules) per group, t test. ****P<0.0001. Data are presented as the transcript number normalized to the tubule volume. For Umod analysis, three control kidneys (12 Z-stacks per animal) and four UIR day 1 kidneys (12 Z-stacks per animal) from two independent batches were analyzed. (D and E) Feature plots show the elevated Cd24a and Sox4 expression in UIR day 1. (F) CISH shows the elevated Cd24a expression in the UIR day 1 renal tubules. Cortex, purple frames (left); medulla, black frames (right). Related to Supplemental Figures 24 and Supplemental Tables 2 and 3. Original magnification, ×40. Scale, 100 μm.
Figure 3.
Figure 3.
scRNA-seq reveals the transcriptional landscapes of AKI recovery. (A) UMAPs show renal cell populations in the UIR day 2, 4, 7, and 14. Day 11 is shown in Supplemental Figure 3, because it was very similar to day 14. (B) Slc34a1 CISH, UIR day 2, 4, 7, and 14. Slc34a1, a marker of differentiated proximal tubules, showed continued, severely reduced, expression at day 2, which steadily recovered by day 14. Original magnification, ×4; zoom into the cortex, ×40. (C) Kim1 (green), Aqp2 (red), Slc34a1 (purple), DAPI (blue) RNAscope, UIR day 2, 4, 7, and 14. The injury marker Kim1 showed strong expression at day 2, less expression at day 4, and very little expression at day 7, which returned to normal levels by day 14. Original magnification, ×4, zoom into the cortex, ×60. (D) UMAP shows the integrated, all-time-point analysis of change in renal cell populations over the AKI course. Note the colocalization of renal cell populations from UIR day 1 and 2, outlining the most prominent injury, and UIR day 11 and 14, which reflect AKI recovery and also colocalize with the control. Renal cell populations from the intermediate stages (UIR day 4 and 7) are localized between the “injured” and “recovered” groups. (E) Trajectory analysis of the control, UIR day 1, 4, 7, and 14 renal cell populations shows the transition from injured to recovered renal tubules. Numbers 1 and 2 represent the significant branch points of differentiation. Scale, ×4, 2500 μm; ×40, 100 μm; ×60, 25 μm. Related to Supplemental Figures 58 and Supplemental Tables 4 and 5.
Figure 4.
Figure 4.
Sox4 and Cd24a label the proximal and distal tubule injury. (A) Venn diagram shows the genes elevated in the injured prox at UIR day 1, 2, 4, and 7. Injured proximal tubule gene expression patterns were compared with control proximal tubule. Note 183 overlapping genes highlighted by the red box. (B) Venn diagram shows the genes elevated in the mixed-identity cells at UIR day 1, 2, and 4. Mixed-identity cell gene expression patterns were compared with control proximal tubule. A total of 99 genes overlapping between the time points are highlighted by the red box. (C) qPCR shows Sox4 and Cd24a expression over the AKI course. These nephrogenic genes showed significant elevation at days 1, 2, and 4. n=4–6 animals per group, analyzed with one-way ANOVA with Bonferroni correction. **P<0.01 compared with control. Scale, ×4, 2500 μm; ×40, 100 μm; ×60, 25 μm. (D) Feature plots show cells expressing Slc34a1 (green), Sox4 (red), and both (blue) in UIR day 4 versus control. In the control, very few proximal tubule cells expressed Sox4 (many with Slc34a1, blue dots). In UIR day 4, Sox4 was predominantly expressed by injured proximal tubules, showing an inverse relationship with Slc34a1. (E) RNAscope with Slc34a1 (pink), Sox4 (green), and Kim1 (white) probes, and DAPI (blue) in the UIR day 4. In the cortical region of the UIR day 4 kidney, the dedifferentiated proximal tubules, with reduced Slc34a1 expression, showed robust expression of Kim1 and Sox4. Original magnification, ×60, 0.14 μm/px Nyquist zoom, maximal intensity projection from Z-stack; scale, 50 μm. (F) Feature plots show Lcn2 (green), Cd24a (red), and Double (blue) positive cells in UIR day 1 versus control. Whereas Lcn2 and Cd24a showed very weak expression in the control kidney, UIR day 1 showed striking colocalization between Lcn2 and Cd24a in the distal tubule, loop of Henle, and collecting duct. Cd24a was also elevated in the injured proximal tubules and mixed-identity cells. The feature plots show gene expression without color gradient outlining the expression level. (G) RNAscope images show Lcn2 (green) and Cd24a (pink) colocalization at UIR day 1, white pointers. Cd24a is also elevated in Slc34a1-positive (cyan) proximal tubules. DAPI, blue. Original magnification, ×60, 0.21 μm/px Nyquist zoom, maximal intensity projection (MaxIP) from Z-stack; scale, 50 μm. Related to Supplemental Figures 912 and Supplemental Table 6.
Figure 5.
Figure 5.
AKI causes extensive Osteopontin and Keratin signaling in the injured kidney. (A) Feature plots show Spp1 expression in control and UIR day 1 cell populations. Color gradient outlines the expression intensity. (B) Spp1 CISH, control versus UIR day 1. Original magnification, ×4; zoom into the cortex, ×40. Scale, 100 μm. (C) Feature plots show Krt8 expression in control and UIR day 1 cell populations. Color gradient outlines the expression intensity. (D) Combined Spp1 CISH (cyan) and Krt8 Immunofluorescence (red) shows the injury-induced Spp1 and Krt8 elevation in UIR day 1 and 4, resolving by UIR day 14. scRNA-seq-predicted Krt8 elevation is reproduced on the protein level. DAPI (blue), UIR day 1. Original magnification, ×20. Scale, 200 μm. TD, transmitted detector shows the chromogenic Spp1 CISH signal. Related to Supplemental Figure 13.
Figure 6.
Figure 6.
scRNA-seq reveals novel gene expression signatures of AKI. (A) Feature plots show Sh3bgrl3 and Ahnak expression in UIR day 1 versus control. Note that although both genes are nearly absent in the control proximal tubules and moderately expressed in the control collecting duct, AKI induces their robust elevation in multiple tubular segments. (B) Sh3bgrl3 and Ahnak CISH, UIR day 1 versus control. Note that Ahnak, which is normally expressed mostly in the glomeruli and stroma, is substantially elevated in the injured renal tubules. Original magnification, zoom into the cortex, ×40. Scale, 100 μm. (C) Myh9 CISH shows robust elevation in UIR day 1 and 2, which steadily declines at day 4 and 7, with reversal to the normal levels by day 14. Note the intratubular (red arrows) and glomerular (black stars) Myh9 expression induced by AKI. Original magnification, ×4; scale, 2500 μm. Zoom into the cortex (black frames), ×40; scale, 100 μm. (D) Feature plots show Myh9 expression in UIR day 1 versus control. Note the elevation in all tubular segments. (E) Myh9 CISH (magenta), Aqp1 IF (green), DAPI (blue), control versus UIR day 1. White pointers show Myh9 within Aqp1-positive tubules. Original magnification, 60× Nyquist 0.07 px/μm zoom, scale 25 μm. (F) qPCR shows Myh9, Sh3bgrl3, and Ahnak expression over AKI course, four to six animals per group, one-way ANOVA with Bonferroni and Holm. *P<0.05, **P<0.01 compared with the control. Related to Supplemental Figure 13.
Figure 7.
Figure 7.
scRNA-seq identifies novel epithelial-to-stromal interactions in adult AKI. (A) Circos plot of ligand-receptor interactions between the proximal (brown) and distal (blue) tubules, the stromal cells (green), and the stromal/pericyte cells (purple) in the normal kidney. The populations producing the putative ligand are labeled; directions of predicted ligand-receptor interactions are shown with the sharp ends of the arrows. The names of all putative ligands and receptors with respect to the cell populations are available in Supplemental Figure 14 and Supplemental Table 7. (B) Circos plot of ligand-receptor interactions between the proximal tubules (brown), injured proximal tubules (red), mixed-identity cells (green), distal tubules (blue), and the stromal cells (teal) in the UIR day 4 kidney. The populations producing the putative ligand are labeled; black arrows show Vim-Cd44 (highlighted in blue), Col18a1-Gpc4, and Col18a1-Itgb1 (highlighted in black) ligand-receptor pairs. Note the dramatic increase in the number of potential interactions compared with the control, using the same filters. The names of all putative ligands and receptors with respect to the cell populations are available in Supplemental Figure 15 and Supplemental Table 7. (C) Representative RNAscope image of control proximal tubules, Vim (green), Col18a1 (red), Slc34a1 (purple). Original magnification, 60×, 0.05 μm/px Nyquist zoom, maximal intensity projection from the Z-stack. Scale, 25 μm. (D) Representative RNAscope image of UIR day 4 proximal tubule, Vim (green), Col18a1 (red), and Slc34a1 (purple). Note the elevated Vim and Col18a1 inside the proximal tubule exhibiting lowered Slc34a1 level. Original magnification, 60×, 0.05 μm/px Nyquist zoom, maximal intensity projection from the Z-stack. Scale, 25 μm. Related to Supplemental Figures 1315, Supplemental Videos 1 and 2, and Supplemental Table 7.
Figure 8.
Figure 8.
Increasing onset age exacerbates AKI outcome. (A) Experimental outline. (B) qPCR shows the fibrosis markers Vim, Col1a1, and Fn1 expression over the AKI course, n=4–6 per group, t test, 4 weeks day 7 versus 10 weeks day 7, 4 weeks day 14 versus 10 weeks day 14. *P<0.05, **P<0.01. (C) Masson trichrome staining shows fibrotic remodeling in the 10-week-old mice at UIR day 14. Note the abundant fibrotic remodeling (blue color on the Masson trichrome staining, yellow pointers) in the UIR day 14 of older mice, while UIR day 14 younger mice exhibit normal kidney histology. Original magnification, ×40. Scale, 100 μm. (D–F) qPCR shows Slc34a1, Sox4, and Cd24a expression over the AKI course, n=4–6 per group, t test, day 7, 4 weeks versus 10 weeks; day 14, 4 weeks versus 10 weeks. *P<0.05, **P<0.01. (G) Cd24a IHC. Note persistent intratubular expression in the UIR 10-week-old mice, resolved in the UIR 4-week-old mice. Original magnification, zoom into the cortex, 40×. Scale, 100 μm. (A) The kidney image is reprinted from ref. , with permission from Elsevier.

Comment in

  • The molecular genetics of AKI.
    Carney EF. Carney EF. Nat Rev Nephrol. 2021 Jan;17(1):14. doi: 10.1038/s41581-020-00378-z. Nat Rev Nephrol. 2021. PMID: 33199853 No abstract available.

References

    1. Harding JL, Li Y, Burrows NR, Bullard KM, Pavkov ME: US trends in hospitalizations for dialysis-requiring acute kidney injury in people with versus without diabetes. Am J Kidney Dis 75: 897–907, 2020. - PMC - PubMed
    1. Hoste EAJ, Kellum JA, Selby NM, Zarbock A, Palevsky PM, Bagshaw SM, et al. .: Global epidemiology and outcomes of acute kidney injury. Nat Rev Nephrol 14: 607–625, 2018. - PubMed
    1. Pavkov ME, Harding JL, Burrows NR: Trends in hospitalizations for acute kidney injury — United States, 2000–2014. MMWR Morb Mortal Wkly Rep 67: 289–293, 2018. - PMC - PubMed
    1. Sawhney S, Fraser SD: Epidemiology of AKI: Utilizing large databases to determine the burden of AKI. Adv Chronic Kidney Dis 24: 194–204, 2017. - PMC - PubMed
    1. Ferenbach DA, Bonventre JV: Mechanisms of maladaptive repair after AKI leading to accelerated kidney ageing and CKD. Nat Rev Nephrol 11: 264–276, 2015. - PMC - PubMed

Publication types