Nkd2 as a therapeutic target for renal fibrosis

Abstract

The present invention relates to elucidating the role of proteins derived from the naked cuticle homolog 2 (Nkd2) gene in the development of chronic kidney disease, particularly progressive chronic kidney disease and renal fibrosis. The present invention particularly relates to methods for identifying compounds that bind to the Nkd2 protein and the use of Nkd2 for screening and identifying Nkd2-interacting compounds. The present invention further relates to pharmaceutical compositions for use in treating kidney disease, particularly pharmaceutical compositions comprising agents that bind to and / or inhibit the NKD2 protein (Figure 1a).

Description

[Technical Field]

[0001] The present invention relates to the role of naked cuticle homolog 2 (Nkd2) protein in the development of chronic kidney disease, particularly progressive chronic kidney disease and renal fibrosis. The invention particularly relates to methods for identifying compounds that bind to Nkd2 protein and the use of Nkd2 for screening and identifying Nkd2-interacting compounds. The invention further relates to pharmaceutical compositions for use in treating kidney disease, particularly pharmaceutical compositions comprising agents that bind to and / or inhibit Nkd2 protein. [Background technology]

[0002] Chronic kidney disease (CKD) affects more than 10% of the global population, and its prevalence is increasing. Regardless of the type of initial insult, the final common pathway of kidney injury is renal fibrosis. Renal fibrosis is a hallmark of chronic kidney disease progression, yet no anti-fibrotic therapies currently exist. The degree of renal fibrosis is closely associated with renal functional decline and clinical outcomes in CKD, making renal fibrosis an important therapeutic target for CKD. There are currently no approved therapeutic drugs for renal fibrosis. This is largely due to the fact that the cellular origin, functional heterogeneity, and regulation of scar-forming cells in the human kidney remain unclear and remain a major source of debate in this field (Duffield 2014; Falke et al. 2015). The only available treatments are continuous renal replacement therapy (dialysis) and kidney transplantation. These two options are associated with significant inconvenience for patients and a significant economic burden on national healthcare systems. Therefore, novel therapeutic approaches are highly desirable.

[0003] Renal fibrosis is defined by the excessive deposition of extracellular matrix, which destroys and replaces functional parenchyma, leading to organ failure. The renal histology is broadly divided into three compartments, all of which can be affected by fibrosis: in the glomerulus, termed glomerulosclerosis; in the tubulointerstitium, termed interstitial fibrosis; and in the vasculature, termed arteriosclerosis and perivascular fibrosis (Djudjai and Boor 2019).

[0004] Renal fibrosis is characterized by the high expression, secretion, and accumulation of extracellular matrix (ECM) proteins, such as collagen-1. Myofibroblasts are the primary source of ECM in renal fibrosis, although their cellular origin remains controversial (Duffield 2014; Friedman et al. 2013; Kriz et al. 2011; Kramann and DiRocco 2013). Single-cell RNA sequencing and mapping can dissect the cellular heterogeneity of complex tissues and disease processes, generating novel insights into the cell populations and mechanisms that mediate disease (Ramachandran et al. 2019; Dobie and Henderson 2019). Previously, genetic fate-tracing data in mice were extended by various staining approaches in human tissues, suggesting that a wide range of different cell types are involved in fibrosis, including epithelial, endothelial, and circulating hematopoietic cells, as well as resident mesenchymal cells (Duffield 2014; Friedman et al. 2013; Kramann and DiRocco 2013). Summary of the Invention [Problem to be solved by the invention]

[0005] Therefore, the problem underlying the present invention is to provide agents, compounds and compositions for use in the treatment of chronic kidney disease, as well as methods and means for identifying said agents, compounds and compositions.

[0006] The present invention provides methods and means for identifying drugs, compounds and compositions for use in the treatment of chronic kidney disease, and in particular methods for identifying highly effective drugs, compounds and compositions for use in the treatment of progressive chronic kidney disease and renal fibrosis.

[0007] As disclosed herein, the inventors have found that naked cuticle homolog 2 (NKD2) protein is produced in terminally differentiated myofibroblasts, which are involved in renal fibrosis, but not in cells expressing pericyte and fibroblast marker proteins or only small amounts of extracellular matrix proteins. Furthermore, Nkd2-expressing cells were found to have increased activity of pro-fibrotic signaling pathways. Furthermore, the inventors have found that fibrosis can be suppressed by deletion of the Nkd2 gene or knockdown of Nkd2 expression, demonstrating that this gene is involved in extracellular matrix production and fibrosis. Thus, the inventors have identified Nkd2 as a novel target and, for the first time, have discovered a new therapeutic approach for developing therapeutic agents that inhibit Nkd2 gene expression and / or NKD2 protein activity using small molecule drugs (SMOLs), peptides, or biological agents.

[0008] In view of the prior art, it is therefore an object of the present invention to provide a method for reducing the expression and / or secretion of extracellular matrix (ECM) proteins by a given cell.

[0009] A further object of the present invention is to provide a method for identifying an agent that binds to and / or inhibits naked cuticle homolog 2 (NKD2) protein or a fragment thereof.

[0010] A further object of the present invention is to provide a method for using naked cuticle homolog 2 or a fragment thereof, or a nucleic acid encoding naked cuticle homolog 2 (NKD2) protein or a fragment thereof, to identify drugs that bind to NKD2 or a fragment thereof.

[0011] Based on the above findings, a further object of the present invention was to provide a drug for use in the treatment of chronic kidney disease, particularly a drug for use in the treatment of progressive chronic kidney disease and / or renal fibrosis.

[0012] Based on the above findings, a further object of the present invention is to provide a pharmaceutical composition containing the drug and a method for producing the same.

[0013] These and other objects are achieved according to the invention by the methods and means set forth in the independent claims. The dependent claims relate to specific embodiments.

[0014] The general advantages of the present invention and its features are described in detail below. [Brief explanation of the drawings]

[0015] [Figure 1]a. Schematic diagram of the nephron structure and cell types residing in different niches. b. UMAP embedding of 51,849 MME- (CD10-) single-cell transcriptomes from 15 human kidneys. Colors indicate the five major cell types: epithelial (n = 9,280), endothelial (n = 29,814), immune (n = 9,616), mesenchymal (n = 3,115), and neural (Schwann cells, n = 24). For abbreviations, see 1d. c. Correlation network of single-cell clustering results. Nodes represent cell clusters. Edges (linear connections) represent correlations between clusters. The network layout was determined using a force-directed layout implemented in the ggraph R package (https: / / cran.r-project.org / web / packages / ggraph / index.html). d. Scaled expression of the top 10 specific genes in each cell type / state cluster. Gene rankings per cluster were calculated using genesorteR.Cell cluster labels refer to cell type / status groupings of 29 reference cell types (B cells (B): n = 3,101, T cells (T): n = 484, natural killer cells (NK): n = 740, plasma cells (P): n = 167, mast cells (Mast): n = 142, dendritic cells (DC): n = 841, monocytes (Mono): n = 1,111, macrophage 1 (Mac1): n = 1,476, macrophage 2 (Mac2): n = 615, macrophage 3 (Mac3): n = 939, arterial duct endothelial (Art1): n = 901, glomerular capillary (GC): n = 4377, venous endothelial (Ven): n = 2724, lymphatic endothelial (lEn): n=509, rectal canal 1 (VR1): n=5355, rectal canal 2 (VR2): n=2,023, rectal canal 3 (VR3): n=3,051, rectal canal 4 (VR4): n=726, rectal canal 5 (VR5): n=3,271, rectal canal 6 (VR6): n=4,378, injured endothelial cells (iEn): n=2,499, vascular smooth muscle cells (VSMC): n=426, pericyte 1 (Pe1): n=455, pericyte 2 (Pe2): n=188, fibroblast 1 (Fib1): n=761, fibroblast 2 (Fib2): n=208, fibroblast 3 (Fib3): n=246, Myofibroblast 1a (MF1a): n=525, myofibroblast 1b (MF1b): n=306, proximal tubule (PT): n=917, injured proximal tubule (iPT): n=909, descending thin limb (DTL): n=806, connecting tubule (CNT): n=662, macular degeneration cells (Mdc): n=869, thick ascending limb 2 (TAL2): n=692, thick ascending limb 3 (TAL3): n=315, thick ascending limb 4 (TAL4): n=3 90, Interstitial Cell 3 (IC3): n = 62, Interstitial Cell 4 (IC4): n = 78, Interstitial Cell 5 (IC5): n = 33, Interstitial Cell 6 (IC6): n = 39, Interstitial Cell 7 (IC7): n = 40, Interstitial Cell 9 (IC9): n = 72, Interstitial Cell A (IC-A): n = 754, Interstitial Cell B (IC-B): n = 316, Urothelial Cell (Ure): n = 246, Podocyte (Pod): n = 44, Schwann Cell: n = 24). Cell clusters are shown in columns, and genes are shown in rows. Each column represents the average expression level across all cells in the cluster. Figure 1d_1 depicts the expression of overexpressed genes.Figure 1d_2 depicts the expression of under- and over-expressed genes. e. Stratification of single cells according to patient clinical parameters (CKD = chronic kidney disease, eGFR = estimated glomerular filtration rate). f. UMAP embedding of the transcriptome of 31,875 CD10+ (CD10+) single cells stratified according to patient clinical parameters. g. Log fold change of cell cycle stage assignment frequency in healthy and CKD epithelial cells relative to a random model. Positive values ​​represent enrichment, negative values ​​represent depletion. h. KEGG pathway enrichment for CD10+ cells. i. Cells from chronic kidney disease patients are enriched for cells highly expressing ECM (ECM = extracellular matrix, CD10- cells). Violin plot of major cell types in CD10- cells and ECM scores of cells stratified by healthy / CKD. P values ​​for differences in eGFR categories: mesenchymal (p<0.001), immune (p<0.001), epithelial (p<0.001), endothelial (p<0.001). k. Violin plot of ECM scores for cells identified as mesenchymal (by major cell type, stratified by healthy / CKD). P values ​​for differences in eGFR categories: Fib1 (0.0001), Fib2 (ns), Fib3 (ns), MF1a (ns), MF1b (ns), Pe1 (ns), Pe2 (ns), SMC (ns). l. Cell counts by mesenchymal cell type and clinical parameters. m. UMAP embedding of fibroblasts / pericytes / myofibroblasts from 13 human kidneys (n=2,689). Different cell types are separated by dashed lines. Continuous lines refer to the lineage tree predicted by Slingshot. n. Expression of selected genes shown in the UMAP in Figure b. o. Diffusion map embeddings of pericytes, fibroblasts, and myofibroblasts, along with Col1a1 expression on the embeddings. [Figure 2]a. UMAP embedding of 37,800 PDGFRb+ single-cell transcriptomes from eight human kidneys. Dashed lines delineate the four major cell types: epithelial cells (n = 461), endothelial cells (n = 2,341), immune cells (n = 20,838), and mesenchymal cells (n = 25,385). Cell types / states were identified by unsupervised clustering of single-cell transcriptomes (see Methods): fibroblast 1 (Fib1), fibroblast 2 (Fib2), fibroblast 3 (Fib3), pericytes (Pe), vascular smooth muscle cells (VSMCs), mesangial cells (Mesa), myofibroblast 1 (MF1), myofibroblast 2 (MF2), myofibroblast 3 (MF3), mesangial cells (MESA), macrophages 1 (MC1), macrophages 2 (MC2), dendritic cells (DC), arterial duct endothelial cells (Art), glomerular endothelial cells (GC), ductus rectum (VR), injured endothelium (iEn), proximal tubule (PT), injured proximal tubule (iPT), interstitial cells (IC), collecting duct principal cells (PC), thick ascending limb (TAL). b. Stratification of single cells by patient clinical parameters (CKD = chronic kidney disease, eGFR = estimated glomerular filtration rate). c. Expression levels of selected genes from Figure a shown on the UMAP. d. Scaled gene expression of the top 10 genes in each cell type / state cluster. Gene rankings per cell cluster were determined using genesorteR. Cell cluster labels refer to the cell clusters highlighted in a and b. Cells are columns (100 cells per column), genes are rows. Figure 2d_1 depicts the expression of overexpressed genes. Figure 2d_2 depicts the expression of underexpressed / overexpressed genes. e. Diffusion map embedding of PDGFRb+ fibroblasts / myofibroblasts / pericytes (n=23,883) and expression of selected genes on the same embedding. Lines correspond to the three lineages (lineages L1, L2, and L3) predicted by Slingshot. Representative images of multiplex RNA in situ hybridization of Meg3, Notch3, and Postn in n=35 human kidneys. Scale bars: left 10 μm, right 25 μm.From top to bottom: RNA in situ hybridization images for Meg3, Notch, Postn, and Dapi / detection of Postn only / Postn + Notch-3 / Postn + Meg3 / Postln + DAPI. Quantification of Meg3 / Notch3 double-positive cells. g. Top left: Gene expression dynamics of overexpression along the pseudotime axis in lineage 1 (pericytes to myofibroblasts, see e). Cells (columns) were aligned along the pseudotime axis, and genes (rows) correlated with pseudotime were selected and plotted along pseudotime (see Methods). Ten cells per column were averaged. Genes were classified into seven groups that exhibited pseudotime expression patterns. Selected example genes are shown. Bottom left: Image similar to top left, but depicting low levels of underexpression / expression. Top right: Cell cycle stages along pseudotime as a percentage of 2000 cells per column. Bottom right: Enrichment analysis of the PID signaling pathway was performed along pseudotime. [Figure 3]a. Fate-tracing experimental design (top) and visualization (bottom) of PDGFRbCreER-tdTomato in a UUO (unilateral ureteral obstruction) mouse kidney model compared with sham surgery. Top: PDGFRb-tdtom and DAPI detection; Middle: PDGFRb-tdtom detection; Bottom: DAPI detection. b. Representative images of Col1a1 in-situ hybridization in PDGFRbCreER;tdTomato kidneys after UUO surgery. From top to bottom: PDGFRb-tdtom + Col1 + DAPI detection / Col1 detection / PDGFRb-tdtom + Col1 detection / Col1 + DAPI detection. c. Percentage of Col1a1-mRNA-expressing cells co-expressing tdTomato 10 days after UUO surgery (n=3). d. Time-course UUO experimental design. UUO was performed in PDGRb-eGFP mice, and eGFP-positive single cells were isolated from mouse kidneys at days 0, 2, and 10 after UUO induction and assayed using Smart-Seq v2. UMAP embedding of cells collected during a time-course UUO experiment. Cell types were identified by unsupervised clustering (see Methods) (parietal epithelial cells (PECs): n = 68, matrix-producing cells (MPs): n = 76, injured smooth muscle cells 1 (iSMCs1): n = 112, injured smooth muscle cells 2 (iSMCs2): n = 77, injured smooth muscle cells 3 (iSMCs3): n = 76, mesangial cells (Mesa): n = 74, pericytes 1 (Pe1): n = 113, renin-producing smooth muscle cells (rSMCs): n = 101, smooth muscle cells 1 (SMC1): n = 172, pericytes 2 (Pe2): n = 83). f. Percentage of cells occurring in each cell type per time point. Each column sums to 100. g. Expression levels of selected genes in all 10 cell clusters. Expression on UMAP embeddings of selected genes from he. i. Immunofluorescence (IF) staining of kidneys from sham-operated and UUO (day 10) mice showing Pdgfra expression in a subset of PDGFRbCreER;tdTomato-positive cells (arrows). From left to right: PDGFRbCreERtdTomato+PDGFRa+DAPI detection / PDGFRbCreERtdTomato detection / PDGFRa detection / DAPI detection.j. RNA in situ hybridization showing colocalization of Col1a1 expression in PDGFRa / PDGFRb double-positive cells. Col1a1 / PDFGRa / PDFRb triple-positive cells (arrows) are present exclusively in the interstitium of the kidney. From top to bottom: PDGFRa + Col1a1 + PDGFRb + DAPI detection / PDGFR-a detection / PDGFRa + Col1a1 detection / PDGFRa + PDGFRb detection / PDGFRa + DAPI detection. k. Left: Col1a1 expression and ECM score in CD10-negative cells (Figure 1b-c) stratified according to PDGFRa and PDGFRb expression levels. Right: Percentage of Col1a1-positive and -negative cells in the same dataset stratified as above. Col1a1-negative cells are detected in PDGFRa / b double-negative cells, and Col1a1-positive cells occur predominantly in PDGFRa / b double-positive cells. Group comparison: (other genes) vs. (a / b): p<0.001, (a-) vs. (a / b): p<0.001, (b) vs. (a / b): p<0.001, (other genes) vs. (a): p<0.001, (a) vs. (b): p<0.001, (other genes) vs. (b): p<0.001. Bonferroni-corrected p-values ​​based on Wilcoxon rank-sum test. Distribution of IF / TA scores for 62 patients and representative images of trichrome-stained human kidney tissue microarrays (TMAs) stained with multiplexed RNA in situ hybridization using 62 kidney PDGFRa, PDGFRb, and Col1a1 probes and a nuclear counterstain (DAPI) (left). Mean-scaled Col1a1 expression levels in the in situ hybridization data stratified by PDGFRa / PDGFRb detection levels for the same dataset (middle). Percentages of Col1a1-positive and -negative cells in the same dataset, stratified as above (right). Group comparisons: (a / b) vs. (col1a1): p<0.001, (a / b) vs. (b): p<0.001, (a / -) vs. (a): p<0.001. Bonferroni-corrected p-values ​​based on Wilcoxon rank-sum test. Scale bars: a, 1000 μm, b, 10 μm, i+j, 50 μm, k, 10 μm.From top to bottom, multiplex RNA in situ hybridization: detection of Col1a1 + PDGFRa + PDGFRb + DAPI / detection of Col1a1 / detection of Col1 + PDGFRa / detection of Col1a1 + PDGFRb / detection of Col1 + DAPI. [Figure 4]a. Schematic of the UUO experiment. UUO was performed on PDGRb-eGFP mice. eGFP+ / PDGFRb+ single cells were isolated from mouse kidneys and assayed using 10x Genomics drop-seq on days 0 and 10 after UUO (n = 5 each). b. Flow cytometry quantification of PDGFRa, PDGFRb, and PDGFRa / b-expressing cells on day 10 after UUO surgery compared with sham surgery. *p < 0.05; **p < 0.01 calculated by one-way ANOVA with post-hoc Bonferroni correction. c. Left: UMAP embedding of cells isolated in the UUO experiment (shown in a). (n = 7,245) Four major cell clusters could be identified: epithelial cells (n = 223), endothelial cells (n = 370), immune cells (n = 199), and mesenchymal cells (n = 6,633). Unsupervised clustering of single-cell transcriptomes allowed us to distinguish 10 cell types: fibroblast 1 (Fib1), myofibroblast 1 (MF1), myofibroblast 2 (MF2), myofibroblast 3 (MF3), endothelial cells (EC), injured proximal tubule cells (iPT), unidentified mesenchymal cells (uM), and macrophages / monocytes (MC). Right: Percentage of cells in each cell cluster in sham-operated and UUO mice. d. Expression levels of selected genes in each cell cluster (shown in c). Image of extracellular matrix score (ECM score) visualized on a UMAP embedding from c. f. Violin plot of Col15a1 expression in different cell clusters. Only mesenchymal cells are shown. g. Violin plot of collagen score in different cell clusters. Only mesenchymal cells are shown. The collagen score is the average expression level of core collagen genes provided by Naba et al. Representative image of multiplex RNA in situ hybridization of PDGFRa, PDGFRb, and Meg3 in a human kidney with hn=34. Meg3 colocalizes with PDGFRa and PDGFRb. From top to bottom: Meg3 + PDGFRa + PDGFRb + DAPI / Meg3 detection / Meg3 + PDGFRa detection / Meg3 + PDGFRb detection / Meg3 + DAPI detection.i. Percentage of Meg3-positive cells among PDGFRa / b double-positive cells quantified by RNA in-situ hybridization. j. UMAP (left) and diffusion map (right) embeddings of fibroblasts and myofibroblasts (n=6,557). The cell clusters shown in c. The black line indicates the phylogenetic tree predicted by Slingshot. Bottom: Expression of selected genes visualized in the same UMAP embedding. k. Signaling pathways enriched in the same mesenchymal cell cluster. [Figure 5]a. Nkd2 expression visualized by UMAP embedding in Figure 4c (mouse Pdgfra / b double-positive cells). b. Percentage of Col1a1 positive and negative cells stratified by Pdgfra and Nkd2 expression in the same dataset as in Figure 4c. Col1a1 negative cells are enriched in PDGFRa / Nkd2 double-negative cells, and Col1a1 positive cells are also enriched in PDGFRa / Nkd2 double-positive cells. c. Scaled gene expression of genes identified as correlated (Figure 5c_1) or anti-correlated (Figure 5c_2) with Nkd2 expression in human PDGFRb- cells, as depicted in Figure 2a-c. Representative images of multiplex RNA in situ hybridization of PDGFRa, PDGFRb, and NKD2 in dn=36 human kidneys. From top to bottom: NKD2 + PDGFRa + PDGFRb + DAPI detection / NKD2 detection / NKD2 + PDGFRa detection / NKD2 + PDGFRb detection / NKD2 + DAPI detection. e. Percentage of NKD2+ cells among PDGFRa / PDGFRb double-positive cells. Quantified by RNA in situ hybridization from patients with low or high interstitial fibrosis, scored blindly by a renal pathologist. f. Western blot verification of lentiviral Nkd2 overexpression. The exogenous overexpressed protein was HA-tagged. g. Expression levels of Col1a1, fibronectin (Fn), and Acta2 (aSMA) quantified by qPCR after Nkd2 overexpression in human immortalized PDGFRb+ cells treated with transforming growth factor beta (TGFb) or vehicle (PBS). h. Western blot analysis of Nkd2 knockout in multiple single-cell clones (1, 2, 3) compared to a non-targeting gRNA clone (NTG). Clone 2 completely lacks Nkd2 protein expression due to a large insertion, while clones 1 and 3 only exhibit reduced Nkd2 protein expression due to smaller indel mutations. Figure 1 shows the expression levels of Col1a1, fibronectin (Fn), and Acta2 after Nkd2 knockout in the clones depicted in ig.j. Gene set enrichment analysis (GSEA) of ECM genes in Nkd2-perturbed PDGFRb- kidney cells. "Shallow" indicates that NKD2 protein was still detectable in clones 1+3. "Severe" indicates that NKD2 protein was still detectable in clone 2. k. Changes in the intensity of the PID signaling pathway in PDGFRb+NKD2-KO clones and overexpression clones (up indicates genes whose expression increased under the indicated conditions, and down indicates genes whose expression decreased). l. Representative images of multiplex RNA in-situ hybridization of PDGFRa, PDGFRb, and NKD2 in human iPSC-derived kidney organoids. From top to bottom: Nkd2+PDGFRa+Col1a1+DAPI detection / Nkd2 detection / Nkd2+PDGFRa detection / Nkd2+Col1a1 detection / Nkd2+DAPI detection. m. Quantification of the fluorescence intensity of NKD2 in kidney organoids. n. Immunofluorescence staining of Col1a1 (black) in iPSC-derived kidney organoids. o. Quantification of collagen content in kidney organoids. ##p<0.01, ####p<0.0001 by one-way ANOVA with Bonferroni post-hoc test (vs. control NTG). *p<0.05, **p<0.01, and ***p<0.001, ****p<0.0001 by t-test (e+m) or one-way ANOVA with Bonferroni post-hoc test (g, i vs. TGFb NTG). Data represent mean ± SD. Scale bar: 10 μm for c, 50 μm for l+n. DETAILED DESCRIPTION OF THE INVENTION

[0016] Before describing the present invention in detail, it is to be understood that the present invention is not limited to the specific component parts of the described devices or to the process steps of the described methods, as such devices and methods may vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting. It should be noted that, as used in the specification and the appended claims, the singular forms "a," "an," and "the" include singular and / or plural references unless the context clearly dictates otherwise. Furthermore, when a range of parameters bounded by numerical values ​​is given, it should be understood that the range is intended to be inclusive of those limits.

[0017] Furthermore, it should be understood that the embodiments disclosed herein are not meant to be understood as separate, unrelated embodiments. Features discussed in one embodiment are meant to be disclosed in relation to other embodiments shown herein. In one case, if a particular feature is not disclosed in one embodiment but is disclosed in another embodiment, those skilled in the art will understand that this does not necessarily mean that the feature is not disclosed in the other embodiments. Those skilled in the art will understand that while it is the intent of the present application to disclose the feature in other embodiments as well, this has not been done merely for the sake of clarity and to keep the description manageable.

[0018] Furthermore, the contents of the prior art documents referred to herein are incorporated by reference. This refers especially to prior art documents that disclose standard or conventional methods. In that case, incorporation by reference has the primary purpose of providing a sufficient enabling disclosure and avoiding lengthy repetition.

[0019] According to a first aspect of the present invention, the present invention relates to a method for reducing the expression and / or secretion of extracellular matrix (ECM) proteins by a given cell, the method comprising at least one step selected from the group consisting of: (i) inhibiting or reducing the expression of the nkd2 gene in said cells; (ii) promoting the degradation of NKD2 protein in said cells; and / or (iii) inhibiting or reducing the activity of NKD2 protein in said cells.

[0020] The inhibition or reduction of nkd2 gene expression can be achieved by, for example, nkd2 gene knockdown, knockout, conditional gene knockout, gene modification, RNA interference, siRNA and / or antisense RNA.The inhibition or reduction of nkd2 gene expression can be achieved by, for example, antisense molecules such as antisense oligonucleotides, antisense conjugates, or catalytic nucleic acid molecules such as ribozymes.Such molecules can be produced intracellularly by expression vectors or introduced from outside of cells.

[0021] Antisense oligonucleotides can be chemically modified to enhance their stability and / or binding affinity.The chemical modification of the backbone chemistry of antisense oligonucleotides, for example, by phosphorothioate, phosphorodithioate, phosphoramidite, alkylphosphotriester or borane phosphate, has been described in the prior art (for example, WO00 / 49034A1).

[0022] The promotion of degradation of the NKD2 protein in the cells can be achieved, for example, by proteolytic molecules such as proteases. Such proteases and proteolytic molecules can be heterologously synthesized in the cells using an expression vector, or the amount of synthesis in the cells can be increased by increasing the expression level of a homologous gene of a gene encoding the protease in the cells, or they can be introduced into the cells from outside the cells.

[0023] Said inhibition or reduction of the activity of the NKD2 protein can be achieved by the use of an agent that binds to the naked cuticle homolog 2 (NKD2) protein.

[0024] Preferably, the predetermined cells are kidney cells, more preferably kidney myofibroblasts, and most preferably terminally differentiated kidney myofibroblasts.

[0025] The NKD2 protein has been shown to be a WNT antagonist. The naked cuticle (NKD) family includes Drosophila naked cuticle and its two vertebrate orthologues, naked cuticle homolog 1 (NKD1) and 2 (NKD2). The Nkd2 locus is located on chromosome 5p15.3. Loss of heterozygosity in these regions has been frequently found in various tumor types, including breast cancer. Both NKD1 and NKD2 have been reported to antagonize canonical Wnt signaling by interacting with Dishevelled via an EF-hand-like motif (Hu et al., 2006). Furthermore, NKD2 has been shown to bind to Dishevelled through a TGFα-binding domain (Li et al., 2004). Human NKD1 and 2 share only 40% identity with each other and approximately 70% identity with their respective mouse orthologues.

[0026] Although the C-terminus of NKD2 is highly disordered, the N-terminal region of NKD2 contains most of the functional domains, including myristoylation, EF-hand motifs, Dishevelled-binding region, vesicle recognition, and membrane targeting motifs (Li et al., 2004; Rousset et al., 2001; Zeng et al., 2000). NKD2 has been suggested to function as a switch protein through several of its functional motifs (Hu et al., 2006). In glioblastoma cells, the promoter region of NKD2 is hypermethylated.

[0027] As described herein, the inventors have identified NKD2 as a therapeutic target for the treatment of renal fibrosis. Nkd2 has been found to be expressed exclusively in terminally differentiated PDGFRa+ / PDGFRb+ myofibroblasts, which express high levels of the extracellular matrix protein collagen-1. This matrix protein is primarily produced by myofibroblasts, which arise from differentiation of fibroblasts and pericytes. More than 40% of collagen-1-producing cells have been shown to be Nkd2 / PDGFRa+. Cells expressing pericyte and fibroblast marker proteins and secreting low levels of matrix proteins lack Nkd2 expression. Furthermore, increased activity of pro-fibrotic signaling pathways, including TGF-β, Wnt, and TNFα signaling pathways, was detected in Nkd2-expressing myofibroblasts. Nkd2 overexpression and depletion experiments, respectively, demonstrated that Nkd2 is involved in the production of extracellular matrix proteins. When Nkd2 was overexpressed in human fibroblasts using lentivirus, stimulation with the fibroblast growth factor TGF-β increased the expression of fibroblast growth-promoting matrix proteins such as collagen-1 and fibronectin. CRISPR / Cas9-mediated knockout of Nkd2 significantly reduced the expression of collagen-1, fibronectin, and ACTA2. In human kidney organoids containing all compartments stimulated with IL1-13 to induce fibrosis, knockdown of Nkd2 using siRNA reduced collagen-1 expression and suppressed fibrosis.

[0028] The NKD2 protein may be a mammalian, non-primate, primate, particularly a human NKD2 protein, or a fragment thereof.

[0029] According to another aspect of the present invention, the present invention relates to a method for identifying an agent that binds to naked cuticle homolog 2 (NKD2) protein, or a fragment thereof.

[0030] The method may include at least the following steps: (i) providing an NKD2 protein or a fragment thereof; (ii) adding at least one agent to be screened for binding to the NKD2 protein or a fragment thereof; and (iii) identifying at least one agent that binds to the NKD2 protein or a fragment thereof;

[0031] Preferably, the agents screened and identified according to the present invention are NKD2 inhibitors or antagonists.

[0032] The agent according to the present invention may be selected from the group consisting of small molecule compounds, peptides, and biological agents.

[0033] In the present invention, the terms "small molecule compound," "small molecule" ("smol"), or "chemical" refer to organic compounds with low molecular weights (<1,000 daltons), often on the order of 1 nm in size. Many pharmaceuticals are small molecules. Such small molecules may regulate biological processes. Small molecules may be able to inhibit specific protein functions. In the field of pharmacology, the term "small molecule" specifically refers to molecules that bind to specific biopolymers and act as effectors, altering the activity or function of targets. For example, acetylsalicylic acid (ASA) is considered a small molecule pharmaceutical drug, consisting of 180 daltons and 21 atoms. Such small molecules often have a low ability to provoke an immune response and are relatively stable over time.

[0034] Said "biological product", "biological medicine", "biological therapeutic drug" or "biopharmaceutical drug" according to the present invention is preferably an antibody, or an antigen-binding fragment thereof, or an antigen-binding derivative thereof, or an antibody-like protein, or an aptamer.

[0035] In a preferred embodiment of the method for identifying an agent that binds to the NKD2 protein or a fragment thereof, said agent is a member of a compound library.

[0036] The compound libraries may be composed of, for example, small molecule compounds, peptides, or biological compounds, respectively.

[0037] In the present invention, the term "(combinatorial) compound library" relates to a collection of chemical compounds, small molecules, macromolecules such as peptides or proteins, which contain multiple different combinations of related chemical, peptide or biological species that can be used together in a particular screening assay or identification step.

[0038] According to another aspect of the present invention, the present invention relates to the use of naked cuticle homolog 2 or a fragment thereof, or a nucleic acid encoding naked cuticle homolog 2 (NKD2) protein or a fragment thereof, in a method for identifying an agent that binds to NKD2 or a fragment thereof as described above.

[0039] According to another aspect of the present invention, the present invention relates to an antibody, or an antigen-binding fragment or derivative thereof, or an antibody-like protein that specifically binds to the NKD2 protein.

[0040] Preferably, the antibody, or antigen-binding fragment or derivative thereof, or antibody-like protein inhibits NKD2 activity, ie, acts as an inhibitor or antagonist of NKD2.

[0041] As used herein, the term "antibody" refers to a protein consisting of one or more polypeptide chains encoded by immunoglobulin genes or fragments of immunoglobulin genes, or cDNAs derived therefrom, including the light chain kappa, lambda, and heavy chain alpha, delta, epsilon, gamma, and mu constant region genes, as well as any of a number of different variable region genes.

[0042] The basic immunoglobulin (antibody) structural unit is usually a tetramer consisting of two identical pairs of polypeptide chains: a light chain (L, molecular weight approximately 25 kDa) and a heavy chain (H, molecular weight approximately 50-70 kDa). Each heavy chain contains a heavy chain variable region (VH or VH). H ) and the heavy chain constant region (CH or C H The heavy chain constant region consists of three domains: CH1, CH2, and CH3. Each light chain contains a light chain variable region (VL or V L ) and the light chain constant region (CL or C L The VH and VL regions contain regions of hypervariability, also called complementarity-determining regions (CDRs), which are interposed between more conserved regions called framework regions (FRs). Each VH and VL region is composed of three CDRs and four FRs, arranged from the amino terminus to the carboxy terminus in the order FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4. The variable regions of the heavy and light chains form the binding domain that interacts with an antigen.

[0043] The CDRs are the most important for the binding of an antibody or its antigen-binding portion. The FRs can be replaced with other sequences as long as the three-dimensional structure required for antigen binding is maintained. Structural changes in the construct most often result in a loss of sufficient binding to the antigen.

[0044] The term "antigen-binding portion" of a (monoclonal) antibody refers to one or more fragments of an antibody that retain the ability to specifically bind to an antigen in its native form. Examples of antigen-binding portions of antibodies include a Fab fragment (a monovalent fragment consisting of the VL, VH, CL, and CH1 domains), a F(ab')2 fragment (a bivalent fragment comprising two Fab fragments linked by a disulfide bridge at the hinge region), an Fd fragment consisting of the VH and CH1 domains, an Fv fragment consisting of the VL and VH domains of a single antibody arm, and a dAb fragment consisting of a VH domain and an isolated complementarity-determining region (CDR).

[0045] The antibody or antibody fragment or antibody derivative according to the invention may be a monoclonal antibody. The antibody may be of the IgA, IgD, IgE, IgG or IgM isotype.

[0046] As used herein, the term "monoclonal antibody (mAb)" refers to an antibody composition having a homogeneous antibody population, i.e., a homogeneous population consisting of whole immunoglobulins or fragments or derivatives thereof. Particularly preferred such antibodies are selected from the group consisting of IgG, IgD, IgE, IgA, and / or IgM, or fragments or derivatives thereof.

[0047] The term "fragment" as used herein refers to fragments of such antibodies that retain target binding ability, such as CDRs (complementarity determining regions), hypervariable regions, variable domains (Fv), IgG heavy chains (consisting of VH, CH1, hinge, CH2 and CH3 regions), IgG light chains (consisting of VL and CL regions), and / or Fab and / or F(ab)2.

[0048] As used herein, the term "derivative" refers to protein constructs that are structurally different but still have some structural relatedness, such as scFv, Fab and / or F(ab)2, as well as bispecific, trispecific or higher specificity antibody constructs, all of which are described below.

[0049] Other antibody derivatives known to those skilled in the art are diabodies, camelid antibodies, domain antibodies, bivalent homodimers consisting of two chains of scFvs, IgAs (two IgG structures linked by a J chain and secretory component), shark antibodies, antibodies consisting of a New World primate framework and non-New World primate CDRs, dimeric constructs comprising CH3+VL+VH, other scaffold protein formats containing CDRs, and antibody conjugates.

[0050] As used herein, the term "antibody-like protein" refers to a protein that has been engineered (e.g., by mutagenesis of an Ig loop) to specifically bind to a target molecule. Typically, such antibody-like proteins contain at least one variable peptide loop attached at both ends to a protein scaffold. This double structural constraint greatly increases the binding affinity of the antibody-like protein to a level comparable to that of an antibody. The variable peptide loop is usually 10 to 20 amino acids in length. The scaffold protein may be any protein with good solubility. Preferably, the scaffold protein is a small globular protein. Antibody-like proteins include, but are not limited to, affibodies, anticalins, and engineered ankyrin and affilin proteins. Antibody-like proteins can be obtained from large libraries of mutants, for example, by panning from a large phage display library, and can be isolated similarly to conventional antibodies. Antibody-like binding proteins can also be obtained by combinatorial mutagenesis of surface-exposed residues of globular proteins. As used herein, the term "Fab" refers to an IgG fragment containing the antigen-binding region, said fragment consisting of one constant and one variable domain from each of the heavy and light chains of the antibody.

[0051] As used herein, the term "F(ab)2" refers to an IgG fragment consisting of two Fab fragments linked together by disulfide bonds.

[0052] As used herein, the term "scFv" refers to a fusion of the heavy and light chain variable regions of an immunoglobulin, in which a single variable region fragment is linked together with a short linker, usually containing serine (S) and / or glycine (G) residues. This chimeric molecule retains the specificity of the original immunoglobulin despite the removal of the constant region and the introduction of the linker peptide.

[0053] The modified antibody formats are, for example, bi- or tri-specific antibody constructs, antibody-based fusion proteins, immunoconjugates, and the like.

[0054] IgG, scFv, Fab and / or F(ab)2 are antibody formats well known to those skilled in the art, and the relevant validation techniques are available from the respective textbooks.

[0055] According to a preferred embodiment of the invention, the antibody, or antigen-binding fragment or derivative thereof, is a murine, chimeric, humanized or human antibody, or an antigen-binding fragment or derivative thereof, respectively.

[0056] Monoclonal antibodies (mAbs) derived from mice contain proteins from other species that can induce antibodies, potentially causing undesirable immunological side effects. To overcome this problem, antibody humanization and maturation methods have been designed to generate antibody molecules that ideally retain the specificity and affinity of the nonhuman parent antibody while minimizing immunogenicity when administered to humans. These methods are used, for example, to replace mouse MAb framework regions with corresponding human framework regions (so-called CDR grafting). WO200907861 discloses the creation of humanized forms of mouse antibodies by linking the CDR regions of a nonhuman antibody to human constant regions using recombinant DNA technology. US6548640 by the Medical Research Council describes CDR grafting technology, and US5859205 by Celltech describes the production of humanized antibodies.

[0057] The term "humanized antibody," as used herein, refers to an antibody, fragment, or derivative thereof in which at least a portion of the antibody constant and / or framework regions, and optionally some of the CDR regions, are derived from or aligned to human immunoglobulin sequences.

[0058] In another aspect, the present invention relates to a drug obtainable by the identification method as described above.

[0059] The agent has the ability to specifically bind to naked cuticle homolog naked cuticle homolog 2 (NKD2) protein. In a preferred embodiment, the agent specifically binds to the NKD2 protein or a fragment thereof with high or particularly high affinity and / or avidity. In a preferred embodiment, the agent reduces or inhibits NKD2 activity upon binding to NKD2.

[0060] As used herein, the term "specifically binds" means that the agent has a dissociation constant KD for the NKD2 protein molecule or an epitope thereof of at most about 100 μM. In embodiments, the KD is about 100 μM or less, about 50 μM or less, about 30 μM or less, about 20 μM or less, about 10 μM or less, about 5 μM or less, about 1 μM or less, about 900 nM or less, about 800 nM or less, about 700 nM or less, about 600 nM or less, about 500 nM or less, about 400 nM or less, about 300 nM or less, about 200 nM or less, about 100 nM or less, about 90 nM or less, about 80 nM or less, about 70 nM or less, about 60 nM or less, about 50 nM or less, about 40 nM or less, about 3 It is 0 nM or less, about 20 nM or less, or about 10 nM or less, about 1 nM or less, about 900 pM or less, about 800 pM or less, about 700 pM or less, about 600 pM or less, about 500 pM or less, about 400 pM or less, about 300 pM or less, about 200 pM or less, about 100 pM or less, about 90 pM or less, about 80 pM or less, about 70 pM or less, about 60 pM or less, about 50 pM or less, about 40 pM or less, about 30 pM or less, about 20 pM or less, about 10 pM or less, or about 1 pM or less.

[0061] The medicament may be for use in the treatment of chronic kidney disease, particularly where the chronic kidney disease is progressive chronic kidney disease and / or renal fibrosis.

[0062] The agent may be a small molecule compound (smol), a peptide, or a biological agent, preferably the biological agent is an antibody, or a fragment or derivative thereof, or an antibody-like protein, or an aptamer.

[0063] Small molecule compounds according to the present invention may include, for example, alkyl-, alkenyl-, alkynyl-, alkoxy-, aryl-, alkylene-, arylene groups, amino groups, halogen groups, carboxylate derivative groups, cycloalkyl groups, carbonyl derivative groups, heterocycloalkyl groups, heteroarylene groups, sulfonic acid groups, sulfate groups, phosphonic acid groups, phosphine groups, phosphinoxide groups, among other chemical backbones, substituents, groups or residues.

[0064] According to another aspect of the present invention, the present invention relates to the use of an agent that binds to and / or inhibits naked cuticle homolog 2 (NKD2) protein in a method for treating chronic kidney disease, preferably wherein the chronic kidney disease is progressive chronic kidney disease and / or renal fibrosis. In a preferred embodiment, the agent inhibits NKD2 activity when bound to NKD2.

[0065] The present invention relates to a method for treating or preventing chronic kidney disease, the method comprising administering to a human or animal subject an agent that binds to and / or inhibits naked cuticle homolog 2 (NKD2) protein in a therapeutically effective amount or dose.

[0066] As used herein, the term "effective amount" refers to a dose or amount effective, at dosages and for periods of time necessary, to achieve the desired result. The effective amount may vary depending on factors such as the disease state, age, sex, and / or weight of the subject, pharmaceutical formulation, and the subtype of disease being treated, but can nevertheless be routinely determined by one skilled in the art.

[0067] According to another aspect of the present invention, the present invention relates to a pharmaceutical composition comprising the above-mentioned antibody, or its antigen-binding fragment or derivative, or antibody-like protein, or the above-mentioned agent, and optionally one or more pharmaceutically acceptable excipients. Preferably, the excipients may be selected from the group consisting of pharmaceutically acceptable buffers, surfactants, diluents, carriers, excipients, fillers, binders, lubricants, disintegrants, adsorbents, and / or preservatives.

[0068] According to another aspect of the present invention, the present invention relates to a method for producing a pharmaceutical composition, comprising: (i) a method for identifying an agent that binds to and / or inhibits the NKD2 protein or a fragment thereof as described above, and further (ii) mixing the identified agent with a pharmaceutically acceptable carrier;

[0069] According to another aspect of the present invention, the present invention relates to a composition comprising a combination of (i) the above-mentioned antibody, or its antigen-binding fragment or derivative, or antibody-like protein, or an agent that binds to the above-mentioned naked cuticle homolog 2 (NKD2) protein, or the above-mentioned pharmaceutical composition, and (ii) one or more further therapeutically active compounds.

[0070] The pharmaceutical composition may include one or more pharmaceutically acceptable buffers, surfactants, diluents, carriers, excipients, fillers, binders, lubricants, glidants, disintegrants, adsorbents, and / or preservatives.

[0071] The pharmaceutical compositions may be administered in the form of powders, tablets, pills, capsules, or pearls. In aqueous form, the pharmaceutical formulations may be ready for administration, whereas in lyophilized form, the formulations may be converted into liquid form prior to administration by the addition of water for injection, which may or may not contain preservatives such as, for example, but not limited to, benzyl alcohol, antioxidants such as vitamin A, vitamin E, vitamin C, retinyl palmitate, selenium, the amino acids cysteine ​​and methionine, citric acid and sodium citrate, and synthetic preservatives such as the parabens methylparaben and propylparaben.

[0072] The pharmaceutical formulation may further comprise one or more surfactants, one or more tonicity agents, and / or one or more metal ion chelators, and / or one or more preservatives.

[0073] The pharmaceutical formulations described herein are suitable for at least oral, parenteral, intravenous, intramuscular or subcutaneous administration. Alternatively, the conjugates according to the invention may be provided in a depot formulation, which allows for sustained release of the active agent over a period of time.

[0074] In yet another aspect of the invention, there is provided a primary package such as a pre-filled syringe or pen, a vial, or an infusion bag, which contains the formulation according to the previous aspect of the invention.

[0075] Pre-filled syringes or pens may contain the formulation either in lyophilized form (which must then be dissolved, e.g., with water for injection, before administration) or in aqueous form. The syringes or pens are often single-use, disposable items and may have a volume of 0.1 to 20 ml. However, the syringes or pens may also be multi-use or multi-dose syringes or pens. According to another aspect of the invention, the invention relates to a treatment kit of parts comprising: (i) the pharmaceutical composition described above; (ii) a device for administering the composition; and (iii) Optionally, instructions for use.

[0076] array [Table 1]

[0077] [Example]

[0078] While the invention has been illustrated and described in detail in the drawings and the foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive, and the invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.

[0079] All amino acid sequences disclosed herein are shown from N-terminus to C-terminus; all nucleic acid sequences disclosed herein are shown 5'->3'.

[0080] Example 1 Materials and Methods Human Tissue Processing

[0081] Kidney tissue was collected by surgeons from normal and tumor sites. Tissues were snap-frozen on dry ice or placed in pre-chilled University of Wisconsin solution (#BTLBUW, Bridge to Life Ltd., Columbia, USA) and transported to our laboratory on ice. The tissues were sliced ​​into approximately 0.5-1 mm pieces, transferred to C-tubes (Miltenyi Biotec), and processed on a gentle-MACS (Miltenyi Biotec) using program spleen 4. The tissues were then digested for 30 minutes at 37°C with agitation at 300 RPM in a digestion solution containing 25 μg / ml Liberase TL (Roche) and 50 μg / ml DNase (Sigma) in RPMI (Gibco). After incubation, the samples were processed again on the gentle-MACS (Miltenyi Biotec) using the same program. The resulting suspension was passed through a 70 μm cell strainer (Falcon) and washed with 45 ml of cold PBS before being centrifuged at 500 g for 5 minutes at 4°C. Trypan blue staining was performed, and cells were counted using a hemocytometer. Epithelial cells were further enriched by CD10 staining and PDGFRβ staining for fibroblasts. Viable single cells were enriched by FACS sorting and gating on DAPI-negative cells. The average time from biopsy collection to preparation of a single-cell suspension was 5–6 hours.

[0082] mouse PDGFRβCreERt2 (i.e., B6-Cg-Gt(Pdgfrβ-cre / ERT2)6096Rha / J, JAX Stock #029684) and Rosa26tdTomato (i.e., B6-Cg-Gt(ROSA)26Sorttm(CAG-tdTomato)Hze / J JAX Stock #007909) were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). Offspring were genotyped by PCR according to the Jackson Laboratory protocol. Pdgfrb-BAC-eGFP reporter mice were developed by N. Heintz (The Rockefeller University) in the GENSAT project. All mice were genotyped by PCR. Mice were housed under specific pathogen-free conditions at the University of Edinburgh or RWTH Aachen. UUO was performed as previously described. Briefly, after a flank incision, the left ureter was clamped at the level of the lower pole with two 7.0-mm ties (Ethicon). Mice were sacrificed 10 days after surgery. The animal experimental protocol was approved by the German LANUV-NRW and the UK Home Office regulations. All animal experiments were performed in accordance with their guidelines. PDGFRbeGFP male mice for SMART-Seq2 were used within 10 days of birth and 9–11 weeks of age at the time of surgery. They were sacrificed as described. PDGFRbCreER;tdTomato mice (8 weeks old, 2 males / 3 females) were given three oral doses of tamoxifen (10 mg p.o.). After a 21-day washout period, they underwent UUO surgery or sham surgery (as described above) and were sacrificed 10 days after surgery.

[0083] Single cell isolation in mice Euthanized mice were perfused through the left heart with 20 ml NaCl 0.9% to remove residual blood from the vasculature. Kidneys were surgically removed, cut into small pieces, placed in 15 ml tubes (Falcon), and placed on ice-cold PBS containing 1% FCS. To isolate kidney single cells, a combination of enzymatic and mechanical disruption was used, as described above for the isolation of human single cells. This method resulted in an overall survival rate of over 80%.

[0084] FACS Cells were labeled with the following monoclonal direct fluorochrome-conjugated antibodies: anti-CD10 human (clone HI10a, Biolegend), anti-PDGFRb mouse (clone PR7212, R&D), anti-PDGFRα mouse (clone APA5, Biolegend), anti-CD31 mouse (clone Meg13.3, Biolegend), and anti-CD45 mouse (clone 30_F11). Dissociated cells were resuspended on ice in 1% PBS-FBS to a final concentration of 1 x 10. 7 Cells were preincubated with Fc-Block (TruStainFx human, TruStainFx mouse Clone 91, Biolegend) and then incubated with the above antibodies diluted 1:100 in 2% FBS / PBS for 30 minutes in the dark. For human anti-PDGFRb staining, goat anti-mouse Dyelight 405 (Poly24091, Biolegend) was used as the secondary antibody. All compensation was performed at the time of imaging using single-color staining, negative staining, and fluorescence minus 1 controls. Cells were sorted using a SONY SH800 sorter (Sony Biotechnology; 100 μm nozzle sorting tip, Sony) in semi-pure mode with a target efficiency of 80% or higher. For plate-based sorting for SMART-Seq, cell sorting was performed using a FACS Aria II machine (Becton Dickinson, Basel, Switzerland).

[0085] Single-cell assays including Smart-Seq2 and 10X Genomics 3´ sc-RNA-Seq (V2 and V3). For Smart-Seq2, single cells were processed at the SciLifeLab - Eukaryotic Single Cell Genomic Facility (Karolinska Institute). Prior to shipping, single cells were sorted into wells of a 384-well plate containing pre-prepared lysis buffer. Libraries were sequenced on an Illumina HiSeq 4500. Single-cell solutions of primary human kidney cells and cells were run in parallel on a Chromium Single Cell Chip Kit. Libraries were run using the Chromium Single Cell 3´ library kit V2 and i7 Multiplex kits (PN-120236, PN-120237, PN-120262, 10x Genomics) according to the manufacturer's protocol. Library quality was assessed using a D1000 ScreenTape on a 2200 TapeStation system (Agilent Technologies). Sequencing was performed on an Illumina Novaseq platform using S1 and S2 flow cells (Illumina).

[0086] Assessment of fibrosis in human kidneys PAS-stained sections of the kidneys were analyzed and scored blindly by an experienced renal pathologist. All sections were screened for kidney disease, and no specific glomerular signs of tubulointerstitial or vascular disease were found, except for age-related changes and hypertensive nephropathy. The extent of interstitial fibrosis and tubular atrophy were assessed as two separate parameters, as a percentage of the affected cortical area. The degree of total glomerular sclerosis was estimated as the percentage of glomeruli with glomerular sclerosis among all glomeruli. The degree of arteriosclerosis, i.e., fibroelastic thickening of the intima relative to media thickness, was scored on a scale of 0 to 3, with 0 being absent, 1 being mild (<50%), 2 being moderate (51–100%), and 3 being severe (>100% thickening relative to the media).

[0087] For collagen I and III immunohistochemistry, um sections of formalin-fixed, paraffin-embedded kidney tissue were deparaffinized by incubation in xylene (3 × 5 min), followed by rehydration in a series of gradually increasing concentrations of ethanol (100% ethanol 3 × 2 min, 95% ethanol 2 × 2 min, 70% ethanol 1 × 2 min) for indirect immunohistochemistry. Endogenous peroxidase activity was blocked with 3% H2O2 in distilled water for 10 min at room temperature. After washing (twice with PBS), the sections were incubated with primary antibodies [Collagen I (Southern Biotech) Cat No. 1310-01; Collagen III (Southern Biotech) Cat No. 1330-01] in 1% BSA / PBS in a humidified cabinet for 1 h at room temperature. Slides were then washed twice with PBS for 5 min and biotinylated secondary antibodies were added (30 min). Avidin-biotin complex was added and incubated for 30 minutes, followed by incubation in DAB solution at 37°C for 10 minutes. The reaction was stopped by washing with HO, and the slides were counterstained with methyl green for 4 minutes. Finally, the slides were dehydrated in ascending ethanol and xylene. Fully digitized images of immunohistochemically stained slides were further processed using a whole slide scanner (NanoZoomer HT, Hamamatsu Photonics, Hamamatsu, Japan) and analyzed with the viewing software NDP.view (Hamamatsu Photonics, Hamamatsu, Japan) and ImageJ (National Institutes of Health, Bethesda, MD). The percentage of positively stained area in the kidney cortex was analyzed in a blinded manner.

[0088] Antibodies and immunofluorescence staining Kidney tissue was fixed in 4% formalin for 2 hours at room temperature, dehydrated overnight in 30% sucrose, and then frozen in OCT. 5-10 μm frozen sections were used. Slides were blocked with 5% donkey serum and incubated with primary antibody for 1 hour, followed by three 5-minute washes in PBS and a 45-minute secondary antibody incubation. After staining with DAPI (4',6'-diamidino-2-phenylindole) (Roche, 1:10,000), slides were mounted with ProLong Gold (Invitrogen, #P10144). The following antibodies were used: anti-mouse PDGF-Ra (AF1062, 1:100, R&D), anti-CD10 human (clone HI10a, 1:100, biolegend), anti-HNF4a (clone C11F12, 1:100, Cell Signaling), pan-cytokeratin type I / II (Invitrogen, Ref. MA1-82041), Dach1 (Sigma, HPA012672), Col1a1 (Abcam, ab34710), ERG (Abcam, ab92513), AF488 donkey anti-goat (1:100, Jackson Immuno Research), and AF647 donkey anti-rabbit (1:200, Jackson Immuno Research).

[0089] Confocal imaging Images were acquired using a Nikon A1R confocal microscope with 40X and 60X objectives (Nikon). Raw image data were processed using Nikon Software or ImageJ.

[0090] Human kidney tissue microarray Paraffin-embedded, formalin-fixed kidney specimens were selected from 98 non-tumorous human kidney samples from the Eschweiler / Aachen Biobank based on previously performed PAS staining. Randomly selected areas per sample were used to extract one 2 mm core from each kidney sample using a TMArrayer™ (Pathology Devices, Beecher Instruments, Westminster, USA). Each core was arrayed in a recipient block on a 2 mm grid covering approximately 2.5 cm², and 5-micron-thick sections were cut and processed using standard histological techniques.

[0091] RNA in-situ hybridization In situ hybridization was performed using formalin-fixed, paraffin-embedded tissue samples and the RNAScope Multiplex Detection Kit V2 (RNAScope, #323100) according to the manufacturer's protocol with minor modifications. Antigen retrieval was performed in a water bath at 96°C for 22 minutes instead of 99°C for 15 minutes. After antigen retrieval, 3-5 drops of Pretreatment Solution 1 were incubated at room temperature for 10 minutes. Washing steps were performed three times for 5 minutes. The following probes were used for the RNAscope assay: Hs-PDGFRβ #548991-C1, Hs-PDGFRa #604481-C3, Hs-Col1a1 #401891, Hs-COL1A1 #401891-C2, Hs-MG3 #400821, Hs-NKD2 #581951-C2 (targeting 236-1694 of NM_033120.3), Hs-Postn #409181-C2 and 409181-C3, Hs-Pecam1 #487381-C2, Hs-Cl19 #474361-C3, Hs-Cl21 #474371-C2, Hs-Notch3 #558991-C2, and Mm-Col1a1 #319371, Mm-PDGFRa #480661-C2, Mm-PDGFRb #411381-C3.

[0092] Image quantification - ISH image analysis At least three representative tubulointerstitial regions per image were extracted as subsamples by systematic random sampling. Next, all fluorescent dots (transcripts) were manually annotated using the cell counting tool in Fiji (Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany). Subsequently, watershed detection (limits: 0.1–0.4) for identifying adjacent nuclei, edge detection for incomplete objects, and object size selection (limits: 12–180 μm) were performed. 2 Single nuclei were isolated using a home-developed tool based on the marker (https: / / gitlab.com / mklaus / segment_cells_register_marker). The total number of individual dots was then retried for each isolated nucleus. Due to the complex morphology of the kidney and the high cellular density, it is not possible to identify the origin of non-nuclear transcripts; therefore, dots located outside the nucleus were not included in this analysis. For Meg3 and NKD2 analysis of PDGFRa / b cells, after nuclei were segmented, images were analyzed using QPath, and cells were counted based on one or more positive spots per imaging channel. For Col1a1-IF or NKD2-ISH quantification, images were split by RGB channel, and the integrated fluorescence density was determined for each image using ImageJ.

[0093] Quantitative RT-PCR Cell pellets were collected and washed with PBS. RNA was extracted using the RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. 200 ng of total RNA was reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). qRT-PCR was performed using iTaq Universal SYBR Green Supermix (Biorad) on a Bio-Rad CFX96 Real Time System with a C1000 Touch Thermal Cycler. The cycling conditions were 95°C for 3 minutes, followed by 40 cycles of 95°C for 15 seconds and 60°C for 1 minute, followed by one cycle of 95°C for 10 seconds. GAPDH was used as a housekeeping gene. Data were analyzed using the 2-CT method. The primers used are listed in Table 2.

[0094] RT-PCR primer sequence list (human) [Table 2]

[0095] Generation of human PDGFRb+ cell lines PDGFRb+ cells were isolated from healthy human kidney cortex from a nephrectomy specimen (a 71-year-old male patient) by MACS separation (Miltenyi biotec, autoMACS Pro Separator, #130-092-545, autoMACS Columns #130-021-101) after generating a single-cell suspension (as described above). For isolation, the single-cell suspension was first stained in two steps with a specific PDGFRb antibody (R&D #MAB1263 antibody, dilution 1:100) and then incubated with Anti-Mouse IgG1-MicroBeads solution (Miltenyi, #130-047-102). Following MACS, the cells were cultured for 14 days in DMEM medium (Thermo Fisher #31885) supplemented with 10% FCS and 1% penicillin / streptomycin. They were then immortalized using SV40LT and HTERT as follows: Retroviral particles were produced by transient transfection of HEK293T cells using TransIT-LT (Mirus). Two types of amphotropic particles were generated by cotransfection of the plasmid pBABE-puro-SV40-LT (Addgene #13970) or xlox-dNGFR-TERT (Addgene #69805) in combination with the packaging plasmid pUMVC (Addgene #8449) and the pseudotyped plasmid pMD2.G (Addgene #12259). Forty-eight hours posttransfection, retroviral particles were concentrated 100-fold using a Retro-X concentrator (Clontech). Cell transduction was performed by incubating target cells with serial dilutions of retroviral supernatant (1:1 mixture of concentrated particles containing SV40-LT or, alternatively, hTERT) for 48 hours. Infected PDGFRb+ cells were then selected with 2 μg / ml puromycin for 7 days starting 72 hours post-transfection.

[0096] Cultivation of human induced pluripotent stem cell (iPSC)-derived kidney organoids Human iPSC-15 clone 0001 was received from the Stem Cell Facility at Radboud University Center, Nijmegen, The Netherlands. Human iPSCs were cultured on Geltrex-coated plates in E8 medium (Life Technologies). When they reached 70-80% confluence, iPSCs were detached using 0.5 mM EDTA, and cell aggregates were split 1:3 and replated. Human iPSCs were differentiated using a modified protocol based on Takasato et al. (Nature, 2015) and seeded at a density of 18,000 cells per cm on Geltrex-coated plates (Greiner). Differentiation into intermediate mesoderm was initiated using CHIR99021 (6 μM, Tocris) in E6 medium (Life Technologies) for 3 and 5 days, followed by supplementation with FGF 9 (200 ng / ml, RD Systems) and heparin (1 μg / ml, Sigma-Aldrich) in E6 medium up to 7 days. After 7 days of differentiation, cell aggregates (300,000 cells per organoid, a mix of day 3 and day 5 CHIR-differentiated cells) were cultured on Costar transwell inserts and stimulated self-organizing kidney formation using E6 differentiation medium. On days 7+18, kidney organoids were used for siRNA knockdown experiments, as described below. siRNA knockdown of NKD2 in human iPSC-derived kidney organoids

[0097] NKD2 siRNA knockdown was performed according to the manufacturer's protocol (DharmaFECT transfection reagent and NKD2-specific Smart Pool siRNA, both Horizon Discovery). Transfection master mix and scrambled control were prepared in Essential 6 medium (Gibco) and added to the organoids. After an initial 24-hour incubation, the transfection master mix was refreshed, and IL-1β (Sigma-Aldrich) was added at a concentration of 100 ng / ml to induce fibrosis. IL-1β administration and transfection master mix refresh were repeated every 24 hours for 2 days. 96 hours after the start of transfection, organoids were harvested and processed for paraffin sections. Fluorescence in situ hybridization (FISH) and immunofluorescence staining were performed as described above.

[0098] TGFb-treatment experiments After 24 hours of serum starvation in 0.5% FCS-containing medium, 75% confluent PDGFRb cells were treated with 10 ng / ml TGFb (100-21-10UG, Peprotech) in PBS for 24 hours. For inhibitor experiments using T-5224, the inhibitor (or vehicle) was added to the culture wells 1 hour before the addition of TGFb. All experiments were performed in triplicate.

[0099] AP-1 inhibitor administration T-5224 (c-Fos / AP-1 inhibitor, Cayman Chemicals, #22904) was dissolved in DMSO and stored at -80°C. DMSO was always added to control wells in the same proportion.

[0100] Cell proliferation (WST-1 assay) The WST-1 assay using PDGFRb- cells was performed in a 96-well plate according to the manufacturer's recommendations (Roche Applied Science). Briefly, 1 x 10^4 PDGFRb cells were seeded into each well of a 96-well plate and treated with T-5224 or vehicle (DMSO) at the indicated concentrations in triplicate. Cells were incubated with WST-1 reagent for 2 hours and then harvested at the indicated time points. Absorbance was measured at both 450 nm and 650 nm (reference).

[0101] sgRNA:CRISPR-Cas9 vector construction, virus production, and transduction NKD2-specific guide RNA (forward 5′-CACCGACTCCAGTGCGATGTCTCGG-3′; reverse 5′-AAACCCGAGACATCGCACTGGAGTC-3′) was cloned into pL-CRISPR.EFS.GFP (Addgene #57818) using BsmBI restriction digestion. Lentiviral particles were produced by transiently co-transfecting HEK293T cells with the lentiviral transfer plasmid, packaging plasmid psPAX2 (Addgene #12260), and VSVG packaging plasmid pMD2.G (Addgene #12259) using TransIT-LT (Mirus). Viral supernatants were harvested 48–72 h posttransfection, clarified by centrifugation, supplemented with 10% FCS and Polybrene (Sigma-Aldrich, final concentration 8 μg / ml), and filtered through a 0.45 μm filter (Millipore; SLHP033RS). Cell transduction was performed by incubating PDGFRβ cells with viral supernatant for 48 hours. eGFP-expressing cells were single-cell sorted into 96-well plates. Expanded colonies were assessed for mutations using a mismatch detection assay: gDNA spanning the CRISPR target site was PCR amplified and analyzed by T7EI digestion (T7 Endocnuclease, NEB M0302S). To examine specific mutation events in both alleles within the expanded clones, PCR products were inserted into pCR TMThe clones were subcloned into the 4Blunt-TOPO vector (Thermo Scientific K287520). A minimum of six colonies per CRISPR clone were expanded and submitted for Sanger sequencing (clone C2: 30 colonies sequenced). Western blot analysis confirmed complete knockout of NKD2.

[0102] Western blot Western blotting was performed according to standard protocols. Briefly, cell lysates were prepared in RIPA buffer containing a protease inhibitor cocktail (Roche). The protein concentration of the lysates was quantified using a BCA assay (#23225, Pierce, ThermoScientific). Protein lysates were heated at 95°C for 5 minutes in 4x SDS sample loading buffer (BioRad) and loaded onto a 10% SDS-Page gel. Samples were then transferred to PVDF membranes, and blots were probed with primary antibodies (1:3000 rabbit anti-human NKD2 polyclonal antibody, Invitrogen PA5-61979) in 5% Blotto (Thermo Fisher) for 2 hours, washed, and then incubated with secondary antibodies (1:5000 horseradish peroxidase-HRP conjugate anti-rabbit, Vector Laboratories) for 1 hour. TM Developed using ECL Western Blotting Substrate A and B. Mouse monoclonal anti-GAPDH antibody (Novus Biologicals NB300-320; 1:1000) followed by HRP-conjugated anti-mouse secondary antibody (Vector Laboratories) was used as a loading control.

[0103] Lentiviral overexpression of Nkd2 Construction of NKD2 vector and generation of stable NKD2-expressing cell lines. Human NKD2 cDNA was PCR amplified using the primer sequences 5´-atggggaaactgcagtcgaag-3´ and 5´-ctaggacgggtggaagtggt-3´. Restriction sites and an N-terminal 1xHA tag were introduced into the PCR product using primers 5´-cactcgaggccaccatgtacccatacgatgttccagattacgctgggaaactgcagtcgaag-3´ and 5´-acggaattcctaggacgggtggaagtg-3´. The PCR product was then digested with XhoI and EcoRI and cloned into pMIG (pMIG was a gift from William Hahn (Addgene plasmid # 9044; http: / / n2t.net / addgene:9044; RRID:Addgene_9044)). Retroviral particles were produced by transient transfection using TransIT-LT (Mirus) in combination with the packaging plasmid pUMVC (pUMVC was provided by Bob Weinberg (Addgene plasmid #8449)) and the pseudotyping plasmid pMD2.G (pMD2.G was provided by Didier Trono (Addgene plasmid #12259; http: / / n2t.net / addgene:12259 ; RRID:Addgene_12259)). Viral supernatants were harvested 48–72 h posttransfection, clarified by centrifugation, supplemented with 10% FCS and Polybrene (Sigma-Aldrich, final concentration 8 μg / ml), and filtered through a 0.45 μm filter (Millipore; SLHP033RS). Cell transduction was performed by incubating PDGF-β cells with the viral supernatant for 48 h. eGFP-expressing cells were single-cell sorted.

[0104] Bulk RNA sequencing RNA was extracted using the RNeasy Mini Kit (QIAGEN) according to the manufacturer's instructions. For rRNA-depleted RNA-seq, diluted 1 ng and 10 ng total RNA were used to prepare sequencing libraries using the KAPA RNA HyperPrep Kit with RiboErase (Kapa Biosystems) according to the manufacturer's protocol. Sequencing libraries were quantified using quantitative PCR (New England Biolabs, Ipswich, USA), equimolar pooled, and the final pool was normalized to 1.4 nM. The libraries were denatured using 0.2 N NaOH and neutralized with 400 nM Tris pH 8.0 before sequencing. Final sequencing was performed using the NextSeq platform (Illumina) according to the manufacturer's protocol (Illumina, CA, USA).

[0105] Preparation for ATAC-seq 5000-7500 PDGFRa / b pos cells were FACS-sorted from freshly isolated UUO kidneys as described above, washed twice with cold PBS, and centrifuged at 500g for 5 min. The cell pellet was then lysed in 50 μl of ice-cold lysis buffer (10 mM Tris-HCl, pH 7.5; 10 mM NaCl, 3 mM MgCl2, 0.08% NP40 substitute [74385, Sigma], 0.01% Digitonin [G9441, Promega]) and immediately spun down at 500g for 9 min. The pellet was resuspended in 50 μl of transposase reaction mix, and 25 μl of 2xTD buffer (20 mM Tris-HCl, pH 7.6, 10 mM MgCl2, 20% DMF), 0.5 μl tagment DNA enzyme 1 [15027865, Illumina], and 24.5 μl nuclease-free water were added. The transposition reaction was incubated at 37°C and 350 rpm for 30 minutes in a thermoshaker. The transposed DNA was then purified using the MinElute Reaction Cleanup kit (28204, Qiagen) and eluted in 15 μl of nuclease-free water. The transposed DNA was amplified by PCR (14 cycles total) using custom Nextera PCR primers with NEB Next 2x Master mix (M0541S; New England Biolabs). The first PCR was performed in a volume of 50 μl using NEB Next 2x Master mix and 1.25 μM custom primers for 6 cycles, and the second RT-PCR was performed in a volume of 15 μl using 5 μl (10%) of the preamplification mix and 0.125 μM primers for 20 cycles. The number of additional cycles was determined as described previously. The amplified DNA library was purified using the MinElute PCR Purification kit (28004, Qiagen) and eluted in 20 μl of 10 mM Tris-HCl (pH 8).Library quality was visualized with an Agilent D1000 ScreenTape on a 2200 TapeStation system (Agilent Technologies). ATAC-seq libraries were loaded onto an Illumina NextSeq 500 for 75-bp paired-end sequencing.

[0106] Smart-Seq2 data processing Initial single-cell transcriptome data were processed at the Eukaryotic Single-Cell Genomics Facility at the Science for Life Laboratory in Stockholm, Sweden. The resulting reads were mapped to the mm10 build of the mouse genome (linked to transcripts of eGFP and ERCC spike-insets) to calculate the number of endogenous genes, spike-ins, and eGFP transcripts per cell. Ribosomal RNA genes, ribosomal proteins, and ribosomal pseudogenes were filtered out. We noticed that unsupervised cell clustering (described below) resulted in cells that did not feature alignments assigned to either eGFP or PDGFRb clustering within a single cluster. Therefore, we chose to remove these cells and performed all analyses and clustering without considering them (17 cells).

[0107] 10x single-cell RNA-Seq data processing Fastq files were processed using Alevin and Salmon (Alevin parameters -l ISR, Salmon version 0.13.1) with the Gencode v29 human transcriptome and the Gencode vM20 mouse transcriptome as reference transcriptomes. The expected cell parameter in Alevin was set to three times the number of cells estimated according to the knee-method applied to the read count distribution per cell barcode. Therefore, the UMI count matrix produced by Alevin contained a large number of estimated cells, which could be subsequently filtered (see next paragraph).

[0108] 10x scRNA-Seq cell filtering Ribosomal RNA genes (average 0-1% of the detected RNA amount per cell) and mitochondrial-encoded genes (average 0-80% of the detected RNA amount per cell) were moved from the main gene expression matrix. Mitochondrial-encoded genes were removed to avoid introducing unnecessary cell-to-cell variation that may depend solely on changes in mitochondrial content. Because the log10 (total UMI counts per cell) distributions from the count matrix generated by Alevin (see above) typically exhibited a bimodal distribution, log10 (total UMI counts per cell) was sorted into two clusters using the mclust R package v5.4.3 with the model name "E." Cells belonging to the cluster with the highest counts were retained. Next, cells were filtered based on their mitochondrial RNA content and bias toward highly expressed genes as follows: (1) Cells were classified into two clusters using a bivariate Gaussian mixture with two components trained on log10 (total UMI counts per cell) and the percentage of mitochondrial UMIs per cell. Clustering was performed using the R package Mclust with modelNames set to "EII." Cells falling into the cluster with high mitochondrial content were excluded. This filtering was performed only on libraries showing a clear bimodal distribution of mitochondrial content (only three 10x libraries in this study). (2) The total number of UMIs per cell should be correlated with the total number of uniquely detected genes. Cells that do not follow this relationship (outliers) were filtered by clustering nuclei using a bivariate Gaussian mixture model of log10(total UMI counts) and log10 total uniquely detected genes. The model names were set as "VEV" and "VEE" using the mclust R package. (3) Cells whose ratio of the total counts of the top 500 genes was more than five times the absolute median deviation of all cells were removed. (4) Finally, to exclude cells whose expression ratios of ribosomal proteins and pseudogenes were predominant, cells whose expression ratios of ribosomal proteins and pseudogenes were more than five times the absolute median deviation of other cells were removed. Because libraries from proximal tubule epithelial cells are expected to have a high number of mitochondrial reads, mitochondrial-based filtering was not performed on CD10+ libraries. It should be noted that not all filtering steps were performed on all libraries, as this depends on the quality of each library and the distribution of UMI-cell genes.

[0109] Human 10x single-cell data integration strategy Data analysis revealed several points: (1) Cell types are not guaranteed to be equally represented across patients and conditions (healthy vs. CKD). This is because the cell types captured in a single 10x Chromium run are determined by random cell sampling. (2) Because this classification is based on clinical parameters rather than molecular data or controlled in vitro experiments, samples from both healthy and CKD patients consist of cells in both healthy and diseased states. The proportion of primarily healthy and diseased cell states is expected to vary between healthy and diseased patient samples. (3) Samples from different patients were processed and prepared on different days, dictated by the surgical schedule at Eschweiler Hospital. Therefore, technical (batch) effects could not be controlled experimentally. (4) Because certain cell types are significantly more abundant than others, class imbalance and dataset size (cell count) may affect the clustering results of unsupervised modularization-based graph clustering algorithms, potentially enabling the discovery of high-resolution cell clusters of underrepresented cell types.

[0110] The experimental strategy was designed to obtain separate libraries from CD10+ and CD10- cell fractions to mitigate class imbalance at the cell type capture frequency level using the 10x Chromium protocol. To further mitigate the above, we aimed to (1) cluster data at a local level while preserving global information about the relationships between cell types, and (2) correct for potential technical (batch) variations between samples while preserving important differences, such as differences in cell types and their disease-related status. To achieve this, we developed a strategy with the following steps:

[0111] Step 1. After quality control and cell filtering (see above), cells from each 10x library were clustered separately and each cell cluster was assigned to one of six major cell types: CD10+ epithelial cells, CD10- epithelial cells, immune cells, endothelial cells, mesenchymal cells, and neural cells.

[0112] Step 2. For each of the six major cell types, cells from all 10x libraries were merged together. After correcting for cell-to-cell variations due to technical reasons, the cells were clustered using unsupervised graph clustering. This process resulted in six maps: endothelial, CD10+ epithelial, CD10- epithelial, mesenchymal, immune, and neural. Each map was composed of cells from multiple 10x libraries.

[0113] Step 3. Three single-cell maps were integrated: (1) CD10+ cells (proximal tubule / Figure 1), (2) CD10- cells (proximal tubule depletion / Figure 1), and (3) PDGFRb+ cells (mesenchymal / Figure 2) by combining single-cell expression (UMI count) and clustering information from all major cell type individual maps for each dataset in Step 2. All plots herein are reproduced from these three integrated maps.

[0114] This method achieved local clustering and eliminated technical variability, enabling high-resolution cell state discovery despite the large variation in cell population size. The smallest cluster consisted of 24 cells, while the largest cluster consisted of 5,355 cells. Compared to approaches that use high-level clustering followed by subclustering, our approach generates high-resolution clusters in a data-driven, unbiased manner, completely avoiding the issue of which clusters to subcluster. Zeisel et al. (2017) used a somewhat similar data integration approach.

[0115] Overall, this approach was biologically informed and allowed us to compensate for possible technical effects during cell clustering, such that nearly all cell clusters contained cells from multiple patients / libraries, while preserving interesting differences between patients, such as diseased cell states (e.g., damaged proximal tubule cells), differences in (myo)fibroblast status, and differences in ECM expression.

[0116] Mouse 10x single-cell data integration strategy The mouse 10x data were analyzed and integrated in the same manner as described for the human data. The script used to create the integrated map can be found here: https: / / raw.githubusercontent.com / mahmoudibrahim / KidneyMap / master / make#intergrated#maps / mouse#PDGFRABpositive.r

[0117] Mouse Smart-Seq2 single-cell data integration strategy Single-cell plate sorting was performed to ensure that cells from all three time points were equally represented on all plates, so no further batch effect mitigation was performed during analysis. Variable genes were determined using the decomposeVar function in the Scran R package after running the trendVar function on ERCC transcripts. Genes with an FDR value <0.01 and a biological variance component >1 were retained as highly variable genes. Using these variable genes, we performed the same clustering approach as described for the 10x Chromium data, but with only two clustering iterations and without changing the number of nearest neighbors. The script used to analyze the mouse Smart-Seq2 data can be found here: https: / / github.com / mahmoudibrahim / KidneyMap / blob / master / make#intergrated#maps / mouse#PDGFRBpositive.r .

[0118] Cluster annotation Gene rankings for each cluster were generated using the sortGenes function in the genesorteR R package, with the BinarizeMethod set to "adaptiveMedian" (Smart-Seq2 Data) or "naive" (10x Data). High-resolution cell clusters were then manually annotated based on prior knowledge and literature references. There were 50 such clusters in the CD10- data, 7 in the CD10+ data, 26 in the PDGFRb+ human data, 10 in the mouse Smart-Seq2 data, and 10 in the mouse PDGFRa+ / b+ data. At this high-resolution level (Level 3), cell populations either represent bona fide cell types or distinct cell states. Therefore, these high-resolution cell clusters were grouped into canonical cell types based on their annotation. As a result, 29 cell types were identified in the CD10- map, 1 cell type in the CD10+ map, 16 cell types in the PDGFRb+ map, 5 cell types in the mouse PDGFRa+ / b+ map, and 6 cell types in the Smart-Seq2 mouse PDGFRb+ map. Furthermore, to facilitate data interpretation, cell clusters were classified as epithelial, endothelial, mesenchymal, immune, or neural, and annotated in plots and figures.

[0119] UMAP and Diffusion Map The integrated full-map UMAP projections (Figures 1, 2, 3, 4, and 5) were generated using the UMAP Python package (https: / / github.com / lmcinnes / umap) with the reduced corrected dimension returned by fastMNN, min_dist set to 0.6, and the number of neighbors set to the square root of the number of cells. Local UMAP projections (Figures 1 and 4) were generated with min_dist set to 1, as this parameter tends to produce more geometrically accurate embeddings (see https: / / umap-learn.readthedocs.io / en / latest / ). Diffusion maps were generated using the Destiny R package (https: / / github.com / theislab / destiny) with the reduced dimension returned by fastMNN as input and the number of neighbors set to the square root of the number of cells. We tested various randomization seeds for the UMAP and diffusion maps, as well as various distance metrics for the diffusion maps (as recommended in the Destiny R package manual), and confirmed that they did not result in qualitative differences in the resulting single-cell projections.

[0120] Phylogenetic trees, trajectories, and pseudotime We used the Slingshot R package for phylogenetic tree inference and pseudo-temporal cell order inference based on the UMAP / diffusion map projection. Cell clustering (step 2 from the integration strategy, see above) was used as input cell clusters. The start and end clusters were selected based on reasonable expectations from our prior knowledge, as discussed and recommended by Street et al. (e.g., myofibroblasts are the end cluster for the pericyte / fibroblast / myofibroblast map).

[0121] Gene dynamics along pseudotime Genes whose expression changes with cell order were defined as normalized expression correlated with cell order, quantified by the Spearman correlation coefficient with a Bonferroni-Hochberg-corrected p-value cutoff of 0.001. Gene clusters and expression heatmaps (e.g., Figure 2f, top) were generated by sorting cells along the pseudotime predicted by SlingShot and using the genesorteR function plotMarkerHeat. This function clusters genes using the k-means algorithm, with plots and clustering set to average every 10 cells along the pseudotime. Pathway enrichment and cell cycle analysis were calculated by grouping every 2,000 cells along the pseudotime.

[0122] Pathway enrichment and gene ontology analysis For single-cell data, we used KEGG pathway and PID pathway data downloaded from MSigDB in November 2019 as ".gmt" files. Pathway enrichment analysis was performed using the clusterProfiler R package using the top 100 genes for each cell cluster / group defined by the sortGenes function in the genesorteR package. The enricher function was used with minGSSize set to 10 and maxGSize set to 200. The top five terms by q-value for each cell cluster / group were plotted as a heatmap of -log10(q-values). Gene Ontology Biological Process analysis was performed on the top 200 genes in a similar manner. The enricher function was used with minGSSize set to 100 and maxGSize set to 500. To compare pathway activity between NKD2+ and NKD2- mesenchymal cells, the activities of 14 pathways were estimated on a single-cell basis using PROGENy, using the top 500 most responsive genes from the model, as recommended by benchmark studies.

[0123] Cell cycle analysis Cell cycle analysis followed the method used in Macosko et al. and described in the tutorial by Po-Yuan Tung (https: / / jdblischak.github.io / singleCellSeq / analysis / cell-cycle.html, date: 06-07-2015), using normalized gene expression as input and setting the gene correlation value to 0.1. We used the cell cycle gene set provided by Yang et al. To quantify enrichment / depletion of single-cell cycle assignments (Figure 1g), we plotted the log2 fold change of these frequencies against the average frequency obtained by randomizing the true frequency matrix 1,000 times while keeping row and column sums constant. Randomization was performed using the R package Vegan (https: / / CRAN.R-project.org / package=vegan). Positive numbers indicate enrichment relative to what would be expected by chance, and negative numbers indicate depletion.

[0124] ECM and Collagen Score The expression of core matrisome genes provided by Naba et al. was compiled based on gene expression data normalized in the same way as used for cell cycle analysis.

[0125] Gene expression heatmap Scaled gene expression heatmaps, such as those in Figure 2d, were generated using the plotMarkerHeat and plotTopMarkerHeat functions in the genesorteR R package. Heatmaps of the percentage of expressing cells, such as those in Figure 3d, were generated using the plotBinaryHeat function in the genesorteR R package. Heatmaps showing log2 fold change and enrichment of features, such as those in Figure 5j and 5k, were generated using the ComplexHeatmap R package (v.2.4.2).

[0126] ATAC-Seq analysis Illumina Tn5 adapter sequences were trimmed from ATAC-Seq reads using the bbduk command in the BBmap suite (version 38.32, settings: trimq = 18, k = 20, mink = 5, hdist = 2, hdist2 = 0). ATAC-Seq reads were mapped to the mm10 genome assembly using STAR (version 2.7.0e), retaining only uniquely mapped pairs (settings: AlignEndsType EndToEnd, alignIntronMax 1, alignMatesGapMax 2000, alignEndsProtrude 100 ConcordantPair, outFilterMultimapNmax 1, outFilterScoreMinOverLread 0.9, outFilterMatchNminOverLread 0.9). To remove sequence duplications, we used the MarkDuplicates command in Picard (version 2.18.27) (setting: remove_duplicates=TRUE, http: / / broadinstitute.github.io / picard / ). Next, we used Samtools (version 1.3.1) to remove discordant read pairs from the BAM files. We converted the BAM files to BED files using bedtools (version 2.17.0). Each read was extended 15 bp upstream and 22 bp downstream from the 5´-end of the read, taking into account steric hindrance at Tn5-DNA contacts. We used JAMM (version 1.0.7rev5) to identify open regions in the final BED files, separate the two replicates, and retain all peaks ≥50 bp wide for further analysis (parameters: -r peak, -f 38,38, -e auto, -b 100). The ATAC-Seq signal bigwig file was generated using the JAMM SignalGenerator pipeline (settings: -f 38,38 -n depth).

[0127] To deconvolute the ATAC-Seq signals according to the scRNA-Seq clustering from the bulk ATAC-Seq data, we performed the following strategy. Three major steps of data analysis were taken to deconvolute the ATAC-Seq signals: 1) Each open chromatin peak (where a TF is expected to bind DNA) was first assigned to a specific gene. 2) These genes were ranked by scRNA-Seq cluster (e.g., Fib, MF1 / 2) according to their expression levels in the single-cell RNA-Seq dataset. 3) The top 2000 ATAC peaks were used to identify enriched transcription factor motif sequences.

[0128] More specifically, each open chromatin ATAC-Seq peak was assigned to a gene according to its nearest annotated transcription start site using the bedtools closest function, with 100 kb set as the maximum possible assignment distance. ATAC-Seq peak rankings for each scRNA-Seq cluster were obtained by ranking peaks according to the rank of their assigned genes in the single-cell RNA-Seq cluster. The top 2,000 ATAC-Seq peaks in each scRNA-Seq cluster were selected, and de novo motif searches were performed separately in the open chromatin regions of each scRNA-Seq cluster using the XX motif (settings: --revcomp --merge-motif-threshold MEDIUM). Only motifs with an occurrence rate of 5% or higher, as defined by the XX motif, were retained for further analysis. The occurrence rates of all motifs from the four scRNA-Seq clusters were quantified for peaks assigned to the top 200 genes in each single-cell RNA-Seq cluster using FIMO44 with default parameters (MEME version 5.0.1). This generated a motif frequency matrix for the scRNA-Seq cluster. To quantify the enrichment / depletion of motif occurrence in the scRNA-Seq cluster, we plotted the log2 fold change in their frequency relative to the average frequency obtained by randomizing the true frequency matrix 1,000 times while keeping row and column sums constant. Randomization was performed using the R package Vegan (https: / / CRAN.R-project.org / package=vegan). Positive numbers indicate enrichment relative to what would be expected by chance, while negative numbers indicate depletion (see Figure 4k). Irf8, Nrf, Creb5 / Atf3, Elf / Ets, and Klf were selected for further investigation. Using the bigwig file generated by JAMM as input, DeepTools version 3.3.1 was used to plot the signal of all peaks containing those motifs. The same bigwig file and motif occurrence were visualized using the Integrative Genomics Viewer.

[0129] Other visualization and analysis Heatmaps without quantifying gene expression were generated using the heatmap2 function in the gplots R package (https: / / CRAN.R-project.org / package=gplots). Violin plots were generated using the vioplot R package (https: / / CRAN.R-project.org / package=vioplot).

[0130] Quantification and statistical analysis for non-single cell sequencing data Data are presented as mean ± SEM unless otherwise specified in the legend. Comparisons between two groups were performed using unpaired t-tests. For comparisons between multiple groups, one-way analysis of variance with Bonferroni's multiple comparison test or two-way analysis of variance with Sidak's multiple comparison test was applied. Statistical analysis was performed using GraphPad Prism 8 (GraphPad Software Inc., San Diego, CA). A p value of less than 0.05 was considered significant.

[0131] Gene regulatory network analysis Gene expression was scaled for each gene, and the Pearlson correlation coefficient was calculated between Nkd2 and all other genes along single cell lines in pericytes, fibroblasts, and myofibroblasts. The top 100 correlated and top 100 anti-correlated genes were selected and subjected to pathway enrichment analysis. Furthermore, the single-cell expression of these 200 genes was used as input for the GRNboost2+ python package to predict putative regulatory links between genes. The output network was filtered by removing connections with strength <= 10. The resulting network was plotted as an undirected network using the ggraph package (https: / / cran.r-project.org / web / packages / ggraph / index.html) and clustered into four modules using the Louvain algorithm implemented in the igraph package.

[0132] Predicting transcription factors from single-cell data To obtain transcription factor scores for the distal and proximal regions, the top 200 marker genes for fibroblast, pericyte, and myofibroblast clusters were used as the input gene list for RCisTarget. Analysis was performed with default parameters according to the RCisTarget Vignette (available at https: / / bioconductor.org / packages / release / bioc / vignettes / RcisTarget / inst / doc / RcisTarget.html). To quantify AP1 expression, all Jun and Fos genes were used as the gene set, and the AP1 score was calculated using the same method as for ECM score. To quantify AP1 activity (defined as the expression level of putative target genes), AP1 target genes were defined according to the Dorothea regulon database, and the single-cell AP1 activity score was calculated using the same method as for ECM score.

[0133] Supervised cell classification in mice Using the human PDGFRb+ dataset as a reference, single cells from the mouse PDGFRa+b+ dataset were classified using the CHETAH algorithm with default parameters. Human gene symbols were converted to mouse gene symbols using the biomaRt database.

[0134] CellphoneDB analysis Cell-cell interactions between cell types found in the human CD10- fraction were estimated using CellPhoneDB (v.2.1.1) database version 2.0.0, normalized gene expression data, and default parameters (10% of cells expressing ligand / receptor). Interactions with p-values ​​<0.05 were considered significant. Based on annotations from the database, only ligand-receptor interactions in which at least one partner in the interaction pair was a receptor were considered; receptor-receptor and other interactions with unclear receptors were excluded. Ligand-receptor interactions from pathways involved in kidney fibrosis were selected from the KEGG database for Hedgehog, Notch, TGFβ, and WNT signaling; from MSigDB 3 for EGFR signaling; and through membership in the REACTOME database; and through manual curation for PDGF signaling.

[0135] Bulk RNA-Seq data analysis Gene expression was quantified at the transcriptional level against the human Gencode v29 transcriptome using Salmon v1.1.0 with the --validatMappings and --gcBias parameters turned on. Transcript-level counts were aggregated to gene-level counts using the tximport R package import with countsFromAbundance set to "lengthScaledTPM." Using the Limma R package (v.3.44.1), differences in gene expression between Nkd2-perturbed human kidney PDGFRb+ and their controls were tested using empirical Bayes methods after voom transformation. Results showed that two of the three CRISPR-Cas9 NKD2 knockout clones grouped together in principal component analysis, indicating a shallow phenotype, while the third grouped independently, indicating a more severe phenotype. Therefore, we grouped the first two knockout clones together and performed statistical contrasts between the two independent knockout conditions. Differentially expressed genes were ranked using a moderated t-statistic based on pathway and gene ontology analysis statistical tests. P values ​​were adjusted for multiple testing using the Benjamini & Hochberg method. Genes and pathways with an FDR < 0.05 were considered significant.

[0136] For pathway and gene ontology analysis, we used the clusterProfiler R package with KEGG and PID pathways. We used genes with an adjusted p-value of <0.01 in Nkd2-perturbed cells compared to control, and an absolute log fold change of ≥1 for knockout comparisons (≥0 for overexpression comparisons), ranked by adjusted p-value for up to 200 genes. To examine the enrichment of ECM genes in phenotypes, we used the fgsea R package (v.1.14.0)54 with GSEA-preanked and the MatrisomeDB gene set collection.

[0137] Example 2: Single-cell atlas of human chronic kidney disease To understand which resident cells in the human kidney secrete extracellular matrix during homeostasis and chronic kidney disease, we constructed a single-cell map of the human kidney, focusing specifically on the tubulointerstitium. Because over 80% of renal cortical cells are proximal tubular epithelial cells, previous single-cell maps of the kidney tend to obscure the significant heterogeneity of other kidney cellular compartments. Therefore, we chose a sorting strategy that enriches for live (viable) non-proximal tubular epithelial cells (i.e., CD10-negative cells, also known as membrane metalloendopeptidase (MME)-positive cells). However, to unbiasedly map the entire kidney, we also sorted for the live CD10+ proximal tubular epithelial fraction. Note that CD10 is also expressed on other cell types, so this is a non-exclusive sorting strategy. However, it allows for enrichment / depletion of proximal tubular epithelial cells. We subjected both CD10+ and CD10- fractions from 13 CKD patients with different stages of hypertension-induced nephrosclerosis (n=7; estimated glomerular filtration rate, eGFR >60 and n=6; eGFR <60) to scRNA-seq. We profiled 53,672 CD10- cells from 11 patients (n=7 eGFR >60; n=4 eGFR <60). Patients with eGFR <60 exhibited increased interstitial fibrosis and tubular atrophy. To integrate data across patients, we employed an unsupervised graph-based clustering method (see Methods) and identified 50 distinct CD10- cell clusters (Figure 1a-d) represented in both eGFR groups (Figure 1e). Our sorting and data integration strategy enabled us to understand the full extent of renal interstitium heterogeneity, including the identification of rare cell types, such as Schwann cells, not previously described in kidney single-cell maps (Figure 1a-d).

[0138] A total of 33,690 CD10+ proximal tubule cells were profiled (from eight patients (n = 5; eGFR > 60, n = 3; eGFR < 60)) and divided into seven clusters (Figure 1f). Cell cycle analysis of the CD10+ proximal tubule cell clusters showed increased cell cycle progression in CKD, possibly reflecting the epithelial repair response (Figure 1g). KEGG pathway analysis and gene ontology terms in CD10+ cells suggested increased fatty acid metabolism among various other metabolic pathways in CKD (Figure 1h). Fatty acid metabolism has been reported as a key dysregulated pathway that causes tubular dedifferentiation and fibrosis in human and mouse kidneys (Kang et al., 2015).

[0139] Thus, we employed a sorting strategy that allows us to create a high-resolution map of human kidney homeostasis and CKD and to investigate the cellular origin of ECM in human CKD.

[0140] Example 3: Origin of extracellular matrix in human chronic kidney disease To understand which cell types contribute to extracellular matrix (ECM) production during the progression of human kidney fibrosis, we established a single-cell ECM expression score that summarizes the expression of core ECM molecules, such as collagens, glycoproteins, and proteoglycans. We validated this score in a published dataset of 36 patients with diabetic nephropathy (Fan et al., 2019) and confirmed that ECM score values ​​increased in advanced CKD.

[0141] The ECM score showed a clear shift toward highly ECM-expressing cells in CKD (Figure 1i). Next, we compared the ECM scores of major cell types in homeostasis and CKD. We identified mesenchymal cells as the cells with the highest ECM expression, and their ECM expression levels further increased in CKD. While ECM expression was not significantly increased in any of the mesenchymal subclusters in CKD, all fibroblast and myofibroblast populations expanded in CKD, explaining the overall increase in ECM gene expression observed in mesenchymal cells (Figure 1k-l). Historically, ACTA2 was used as a myofibroblast marker. However, because ECM expression is a hallmark of fibrosis, we defined myofibroblasts as cells expressing most ECM genes. Another important mechanism for increased ECM expression, in addition to the expansion of individual cells, is their differentiation into ECM-rich myofibroblasts. To further investigate the heterogeneity of mesenchymal populations expressing ECM in human CKD and their putative differentiation processes, we created Uniform Manifold Approximation and Projection (UMAP) embeddings of (myofibroblasts) and pericytes (Figure 1m-n). These UMAP embeddings were consistent with the results of unsupervised graph clustering (Figure 1b-c), highlighting the previously underrecognized heterogeneity of human renal mesenchyme (human kidney mesenchymal tissue). Myofibroblasts were clearly identified as a cell group expressing periostin (postn) (Figure 1n). Diffusion mapping is a dimensionality reduction method that assumes that cells are related to each other through a differentiation-like diffusion process. Using this method, we elucidated the putative differentiation mechanism to myofibroblasts. Diffusion map embedding of the mesenchymal cells with the highest ECM expression suggested that myofibroblasts arise from pericytes and fibroblasts (Figure 1o).

[0142] In epithelial cells, a slight upregulation of ECM genes was observed (Figure 1j), suggesting a minor contribution from epithelial-mesenchymal transition (EMT), a long-standing topic of discussion in nephrology. Injured proximal tubule epithelium (iPT) showed the highest expression of ECM genes among CD10- epithelium, and various expressed genes and GO terms suggested dedifferentiation from normal epithelium. A slight increase in ECM expression in CKD was also observed in the CD10+ fraction (all sorted proximal tubule epithelium). Injured cells were defined by the expression of genes reported to be injury-related, such as Sox9, CD24, and CD133 in proximal tubule epithelium and VCAM1 and ACKR1 in endothelium.

[0143] These data indicate that the majority of the ECM generated during fibrosis in the human kidney is derived from multiple distinct mesenchymal cell subtypes, with only a minor contribution from dedifferentiated tubular epithelial cells.

[0144] Example 4: The major source of myofibroblasts in human renal fibrosis is distinct pericyte and fibroblast subpopulations CD10-scRNA-seq data indicated that the majority of Col1a1-expressing cells were PDGFRb+. Therefore, we sorted 37,380 PDGFRb+ cells from eight human kidneys (n=4; eGFR>60, n=4; eGFR<60). Unsupervised clustering identified mesenchymal cell populations and some epithelial, endothelial, and immune cell types (Figure 2a-d), which were annotated according to their correlation with CD10- populations. Collagen and general ECM gene expression was predominant in pericyte, fibroblast, and myofibroblast clusters, consistent with the CD10- data (Figure 1j-k). However, some macrophage / monocyte, endothelial, and injured epithelial populations also expressed collagen1a1 and PDGFRb, although at much lower levels than mesenchymal cells (Figure 2a-c). The computational prediction of the dual likelihood score did not yield particularly high scores for endothelial cells or injured epithelial cells, but did yield slightly elevated scores for macrophage populations. Expression of Col1a1 mRNA in LTA+ proximal tubules, CD68+ macrophages, and Pecam-1+ endothelial cells was confirmed by in situ hybridization (ISH). These data may partially explain the controversy in the literature regarding the contribution of non-mesenchymal cells to the renal myofibroblast pool (Duffield 2014; Wang et al. 2017), because these non-mesenchymal cell types exhibited very little ECM gene expression, even though the majority of ECM gene expression is derived from mesenchymal cells.

[0145] Pseudotime trajectory and diffusion map analysis of the major ECM-expressing cell subtypes from the PDGFRb+ population demonstrated three major sources of myofibroblasts in the human kidney: 1) Notch3+ / RGS5+ / PDGFRa- pericytes, 2) Meg3+ / PDGFRa+ fibroblasts, and 3) Colec11+ / CXCL12+ fibroblasts (Figure 2e). Notably, diffusion mapping revealed that non-CKD cells were primarily located within the low-ECM-expressing pericyte and fibroblast population, suggesting the potential for differentiation of non-CKD mesenchymal cells (pericytes and fibroblasts) with low ECM expression into CKD myofibroblasts with high ECM expression (Figure 2e). UMAP embedding of these mesenchymal cells was also consistent with these results. This is consistent with their differentiation into ECM- and Postn-expressing myofibroblasts in the context of renal fibrosis. We verified this predicted directionality by ISH in human kidneys, confirming that the number of Postn-expressing cells increased and the number of Meg3+ cells decreased in renal fibrosis (Figure 2f). Using ISH, we further verified the major lineages identified in the diffusion map analysis, consisting of Notch3+ pericytes (lineage 1) and Meg3+ fibroblasts (lineage 2) (Figure 2f). As a result, both Meg3+ and Notch3+ cells co-expressed Postn, confirming their differentiation into myofibroblasts (Figure 2f). Interestingly, intermediate-stage cells co-expressing Notch3 / Meg3 / Postn were observed in the center of the diffusion map, likely representing differentiating cells (Figure 2f). Next, we assessed whether the identified mesenchymal subpopulations also differed spatially. Although no clear spatial localization of fibroblasts 1 (Meg3+), pericytes (Notch3+), or fibroblasts 2 (Cxcl12+) was observed, as expected, myofibroblasts 1 (Postn+) were abundant in the fibrotic region. Interestingly, myofibroblasts 3 (Ccl19+ / Ccl21+), which are increased in human kidney fibrosis, showed clear enrichment in the peri-glomerular region.

[0146] Next, we analyzed the gene expression program of pericyte-to-myofibroblast differentiation (lineage 1) (Figure 2g). Cell cycle analysis showed profound changes consistent with both the differentiation process and the expansion of the myofibroblast population (Figure 2g). To better understand pericyte-to-myofibroblast differentiation, we aligned pathway enrichment along pseudotime. Among other pathways, early (canonical Wnt, Myc, AP1), intermediate (ATF2, PDGFRa, Myc), and late (integrin, ECM-receptor interaction, TGFb) signaling were observed (Figure 2g bottom).

[0147] Similar to pericyte-to-myofibroblast differentiation, cell cycle arrest followed by increased proliferation was observed during fibroblast-to-myofibroblast differentiation (lineages 2 and 3). Pseudochronological pathway enrichment highlighted early AP1 signaling, inflammation, and immune cell interaction pathways, followed by integrin signaling, focal adhesion, and ECM interaction pathways.

[0148] Pseudotemporal analysis of lineage 2 revealed that TGFβ signaling was dominant (Figure 2g). Myofibroblast 1, which likely represents a fully differentiated myofibroblast, expressed high levels of TGFβ ligand and low levels of TGFβ receptor. However, the opposite was observed for fibroblast 1, suggesting that there is a mechanism by which myofibroblasts promote fibroblast differentiation.

[0149] Many of the above pathways are known to be key regulators of fibrosis, including integrin 27 and AP1 transcription factor signaling (Wernig et al., 2017), which were consistently and highly active during the early stages of both pericyte and fibroblast-to-myofibroblast differentiation. To further understand the transcriptional regulation of fibroblast and myofibroblast populations, we performed enrichment analysis of transcription factor DNA sequence motifs in the promoters and distal regions of marker genes for various mesenchymal populations. This revealed that AP-1 (Jun / Fos) may play a critical role in fibroblast-to-myofibroblast differentiation. To functionally validate the role of AP1, we generated a novel human PDGFRb+ kidney cell line by lentiviral transduction with hTERT and SV40LT. Pharmacological inhibition of activator protein 1 (AP1) significantly reduced proliferation and decreased osteoglycin (Ogn) expression, while increasing Postn expression, suggesting myofibroblast differentiation of these cells. Notably, in the human PDGFRb data, Ogn marked fibroblasts 1 / 3, and Postn marked myofibroblasts 1. Consistent with these results, AP1 expression negatively correlated with collagen expression in both fibroblasts and myofibroblasts. Interestingly, the expression of putative AP1 target genes positively correlated with collagen expression in myofibroblasts, suggesting that AP1 may act as a repressor. We also performed ligand-receptor analysis (Efremova et al., 2020) to elucidate which cell types interact with the major ECM-expressing mesenchymal cells (fibroblasts, pericytes, and myofibroblasts). While the least signaling was observed from healthy proximal tubule epithelium, injured proximal tubule epithelium was among the top signaling partners to the interstitium, consistent with the tubulointerstitial signaling that is a hallmark of renal fibrosis (Venkatachalam et al., 2015). We noted interactions from pathways that are said to be key players in fibrosis, including TGFβ, PDGFRa / β, Notch, EGFR, and WNT signaling (Kramann and DiRocco, 2013).Among these pathways, Notch, TGFb, Wnt, and PDGFa signaling from injured proximal tubules to mesenchymal profibrotic cells was observed.

[0150] In summary, three cell sources, Notch3+ / PDGFRa- pericytes, Meg3+ / PDGFRa+ fibroblasts, and Colec11+ / Cxcl12+ fibroblasts, were characterized by PDGFRb, shedding light on their differentiation process.

[0151] Example 5: Double-positive PDGFRa+ / PDGFRb+ mesenchymal cells account for the majority of ECM-expressing cells in human and mouse kidney fibrosis We further explored the findings presented above using genetic fate tracking in mice. PDGFRbCreER-tdTomato mice were pulse-treated with tamoxifen, underwent unilateral ureteral obstruction (UUO) surgery, and sacrificed on day 10 (Figure 3a). In situ hybridization (ISH) for Col1a1 mRNA confirmed that virtually all Col1a1-expressing cells in mouse kidney fibrosis were derived from the PDGFRb lineage (Figure 3b-c). Furthermore, immunostaining for the historically used myofibroblast marker aSMA (ACTA2) confirmed that the majority of aSMA-expressing cells were derived from the PDGFRb lineage. Next, we performed a SmartSeq2-based sc-RNA-seq time-course study using PDGFRb-eGFP mice (Picelli et al., 2014) (Figure 3d-e). After UUO, smooth muscle cells and pericytes decreased in abundance over time, whereas mesangial cells and Col1a1+ / PDGFRa+ matrix-producing cell clusters increased significantly over time (see Figure 3f-g). Similar to the human kidney dataset (Figures 1 and 2), the major ECM-expressing cell populations were defined by dual expression of PDGFRa / PDGFRb and expression of decorin (DCN) and periostin (Postn) (Figure 3g-h). Furthermore, pericytes and vascular smooth muscle cells (vSMCs) showed some ECM expression, although at significantly lower levels than the dual PDGFRa / PDGFRb population.

[0152] Immunostaining and ISH in mice confirmed that Col1a1-expressing cells were double-positive for PDGFRa+ and PDGFRb-tdTomato (Figure 3j). This is consistent with the human CD10-mediated enrichment of Col1a1+ cells when selecting for PDGFRa+ / b+ expressing cells, confirming that PDGFRa / PDGFRb-expressing cells are the primary source of ECM expression (Figure 3k). We confirmed this finding in a larger human cohort using multiplex ISH on tissue microarrays from 62 patients (Figure 3l). The diffusion map embedding of matrix-producing cells and pericytes is consistent with our human PDGFRb data, suggesting that pericytes (PDGFRb+, PDGFRa-, Notch3+) are one source of the primary ECM-producing cells (PDGFRb+, PDGFRa+, Col1a1+, Postn+).

[0153] These PDGFRa+ / PDGFRb+ double-positive mesenchymal cells, including populations of fibroblasts and myofibroblasts, including pericyte-derived myofibroblasts but not non-activated pericytes (i.e., pericytes that do not show high ECM gene expression) (Fig. 2e), were found to account for the majority of Col1a1-expressing cells in human and mouse kidney fibrosis.

[0154] Example 6: PDGFRa+ / PDGFRb+ cells are heterogeneous and contain different fibroblast states Next, to gain mechanistic insight into the fibroblast-to-myofibroblast transition and dissect the heterogeneity of the PDGFRa+ / PDGFRb+ population, we performed UUO versus sham surgery in PDGFRb-eGFP mice, followed by sorting of eGFP / PDGFRa double-positive cells. We then generated scRNA-Seq data from 7,245 double-positive PDGFRa+ / PDGFRb+ mouse kidney cells (Figure 4a). Consistent with a rapidly expanding cell population, PDGFRa+ / PDGFRb+ double-positive cells increased in number approximately 140-fold after injury (Figure 4b), consistent with our Smart-Seq2 data (Figure 3f). UMAP embedding of PDGFRa+ / PDGFRb+ cells revealed four major and distinct populations corresponding to mesenchyme (fibroblasts and myofibroblasts), epithelium, endothelium, and immune cells (Figure 4c-d). All of these cell types have previously been discussed as potential cellular origins of kidney fibrosis (Duffield et al., 2014; Wang et al., 2017; Kramann et al., 2018). Of note, because pericytes are PDGFRa- in humans and mice, undifferentiated pericytes were not detected in these PDGFRa / PDGFRb data (Figures 2e and 3g). Non-mesenchymal cells had significantly lower PDGFRb, PDGFRa, ECM, and collagen levels compared to mesenchymal cells (Figures 4d-e), supporting our observation in human data that non-mesenchymal cells contribute less to the scar formation process (Figures 1 and 2). Of note, similar to the human data, the calculated doublet scores do not suggest that these matrix-expressing non-mesenchymal cell populations are likely doublets.

[0155] Unsupervised clustering revealed two major classes of mesenchymal cells in this mouse PDGFRa+ / PDGFRb+ dataset: (1) fibroblast 1, indicated by Scara5 and Meg3 expression, and (2) myofibroblasts, consisting of various myofibroblast subpopulations (Figure 4c-d). In the human data, myofibroblast 1 corresponds to terminally differentiated myofibroblasts with the highest ECM expression, preceding myofibroblast 2 (Ogn+), while fibroblast 1 appears as a "progenitor" non-activated fibroblast population (Figure 2e). Indeed, in the PDGFRa+ / PDGFRb+ data, fibroblast 1 cells and myofibroblasts can be distinguished by three major features: First, the mouse myofibroblast-specific collagen Col15a1 (Figure 3g) was expressed at lower levels in fibroblast 1 than in the myofibroblast cluster (Figure 4f). Next, although Meg3 is also expressed in proximal tubule cells and a portion of the glomerular endothelium, within the mesenchymal population, Meg3 was only detected in fibroblast 1 (Fig. 4d). The presence of a Meg3+ PDGFRa+ / PDGFRb+ mesenchymal subpopulation in the human kidney was verified by in situ hybridization (Fig. 4h-i), suggesting the existence of a fibroblast 1-like subpopulation in the human kidney. Third, fibroblast 1 cells were Scara5+ but Frzb-, again demonstrating their distinctness from myofibroblasts.

[0156] After establishing fibroblast 1 as a distinct fibroblast population, we performed pseudotime analysis by generating UMAP and diffusion map embeddings of all mouse Pdgfra / Pdgfrb mesenchymal cells to gain insight into their lineage relationships (Figure 4j). This analysis suggested that fibroblast 1 (Meg3, Scara5) and myofibroblast 2 (Col14a1, Ogn) represent an early state, myofibroblast 3a represents an intermediate state, and myofibroblast 1a (Nrp3, Nkd2), 1b (Grem2), and 3b (Frzb) represent an end-state state (Figure 4j).

[0157] These data suggest that fibroblast 1 and myofibroblast 2 are the major sources of myofibroblasts in fibrotic mouse kidneys. Myofibroblast 2 (Ogn+ / Col14a1+) may be present in healthy mouse kidneys or may arise as an intermediate state through the differentiation of pericytes to myofibroblasts (Figure 2e, human data). Expression of angiotensin receptor 1 (AGTR1a) is abundant in myofibroblast 2, suggesting its pericyte origin (Figure 4j).

[0158] Analysis of time-course UUO data revealed that Ogn, Scara5, and Pcolce2 are enriched during homeostasis, while naked cuticle homolog 2 (Nkd2) is enriched after injury. This further suggests that fibroblast 1 (Meg3+, Scara5+) and myofibroblast 2 (Ogn+) cell types exist in kidney homeostasis. Furthermore, supervised classification of mouse Pdgfra+ / Pdgfrb+ single-cell data using our human Pdgfrb+ cells as a reference confirmed that fibroblast 1 and myofibroblast 2 are common features of both species.

[0159] Overall, the human and mouse data combined suggest a model in which Pdgfrb+ / Pdgfra+ / Postn+ ECM-high-expressing myofibroblasts (herein referred to as myofibroblast 1) arise from Pdgfrb+ / Pdgfra- / Notch3+ pericytes, Pdgfrb+ / Pdgfra+ / Scara5+ fibroblasts (fibroblast 1), and Pdgfrb+ / Pdgfra+ / Cxcl12+ fibroblasts (fibroblast 2). Pericytes potentially differentiate into myofibroblast 1 via an intermediate state of ECM-expressing Pdgfrb+ / Pdgfra+ / Ogn+ / Col14a1+ (myofibroblast 2).

[0160] Example 7: Distinct fibroblast and myofibroblast states are distinguished by specific transcription factor regulatory programs Next, we aimed to confirm whether the fibroblast and myofibroblast states detected in our data truly represent distinct cell types. Different cell types may be distinguished by both distinct gene expression profiles and distinct transcription factor regulatory programs (Gerstein et al., 2012). We generated bulk ATAC-Seq (Buenrostro et al., 2013) data from Pdgfra+ / Pdgfrb+ mouse kidney cells 10 days after UUO surgery and deconvolved open chromatin region (OCR) signatures from the ATAC-Seq data based on OCR proximity to marker genes identified in the scRNA-Seq clusters. Fibroblast 1 and myofibroblast 2 were distinct from each other and from other myofibroblast populations. Myofibroblast 1a was distinct from myofibroblast 1b and characterized by enrichment for ATF. Myofibroblasts 2 and 3b showed enrichment for the orphan receptor NRF4A1, previously reported as a key regulator of TGFβ signaling and fibrosis (Palumbo-Zerr et al., 2015). Fibroblast 1 was enriched for the AP-1 (Jun / Fos) motif (Figure 4k), consistent with its role inferred from human data. RNA expression of these ATAC-Seq-selected factors was consistent with sequence motif enrichment (Figure 4k), highlighting divergent transcriptional regulation between fibroblast 1, myofibroblast 2, and other myofibroblast populations. Furthermore, we highlight transcription factors, such as Nrf1, Irf8, and Creb5, that may be underrepresented in renal fibrosis. Consistent with the ATAC-Seq data, signaling pathway analysis based on scRNA-Seq data indicated that fibroblast 1 and myofibroblasts are distinct pathway-enriched populations (Figure 4l). Thus, fibroblast and myofibroblast subtypes appear to be distinct ECM-expressing mesenchymal cell types possessing specific transcription factor regulatory programs.

[0161] Example 8: Nkd2 is required for collagen expression in human kidney PDGFRb+ cells and is a potential therapeutic target for renal fibrosis Next, we investigated whether our scRNA-seq data could be used to identify potential therapeutic targets in human renal fibrosis. Nkd2 is specifically expressed in mouse Pdgfra / Pdgfrb terminally differentiated myofibroblasts (Figure 5a), and Nkd2 / PDGFRa double-positive cells account for more than 40% of all Col1a1 cells (Figure 5b). In human PDGFRb cells, NKD2 is a marker of ECM-rich myofibroblasts, and its expression positively correlates with Postn and ECM expression and inversely correlates with genes associated with pericytes and fibroblasts (Figure 5c). Furthermore, NKD2 myofibroblasts were associated with elevated TGFb, Wnt, and TNFa pathway activity compared to NKD2 cells. We confirmed NKD2 expression by multiplex ISH on human kidney tissue microarrays (TMAs) from 36 patients and found that a subpopulation of human PDGFRa / PDGFRb-expressing cells also expressed Nkd2 (Figure 5d-e). Furthermore, the abundance of PDGFRa / PDGFRb / Nkd2 co-expressing cells was higher in patients with more pronounced interstitial fibrosis (Figure 5e).

[0162] Nkd2 has been documented as a modulator of the Wnt pathway and TNFα (Zhao et al., 2015; Hu and Li, 2010; Hu et al., 2010; Li et al., 2004). To understand the mechanism by which Nkd2 regulates renal fibrosis, we used our human PDGFRb+ data to predict a gene regulatory network focusing on genes correlated with Nkd2 using the GRNboost2 framework. The results clustered into four gene regulatory modules: ribosomal proteins (module 1), ECM expression-related genes (module 2), pericyte-related genes (module 3), and non-activated fibroblast-related genes (module 4). This gene cluster included various Wnt modulators and effectors, including Kif26b, Lef1, and Wnt4, in addition to Nkd2. Nkd2 was associated with ECM genes and indirectly connected to Etv1 and Lamp5, and to Col1a1 via Lamp5. This analysis suggests that Nkd2 may be regulated by Etv1 (a member of the Ets factor family) and may act by influencing Lamp5-mediated paracrine signaling.

[0163] Lentiviral overexpression of Nkd2 in our human PDGFRb cell line increased the expression of key profibrotic ECM molecules, such as col1a1 and fibronectin, in response to TGFb (Figure 5f-g). Importantly, CRISPR / Cas9 knockout of Nkd2 significantly reduced the expression of col1a1, fibronectin, and ACTA2, regardless of the presence or absence of TGFb (Figure 5h-i). RNA-seq analysis of Nkd2-overexpressing cells revealed increased expression of ECM regulators and ECM glycoproteins, whereas RNA-seq analysis of Nkd2 knockout clones showed a loss of ECM regulators, ECM glycoproteins, and collagens (Figure 5j). Pathway and gene ontology analysis demonstrated a role for Nkd2 in the ECM expression program and further suggested interactions with AP1 and integrin signaling pathways (Figure 5k). Furthermore, knockout of Nkd2 in vitro significantly altered the expression of Wnt receptors and ligands, suggesting its possible involvement in this pathway.

[0164] To further validate Nkd2 as a therapeutic target, we generated induced pluripotent stem cell (iPSC)-derived kidney organoids containing all major compartments of the human kidney. IL1b is well known to induce fibrosis in iPSC-derived kidney organoids (Lemos et al., 2018). Importantly, siRNA-mediated knockdown of Nkd2 suppressed IL1b-induced Col1a1 expression in kidney organoids (Figure 5l-o). These data confirm that Nkd2 marks myofibroblasts in human and mouse kidney fibrosis and is required for collagen expression in renal myofibroblasts, thus making it a promising potential therapeutic target for treating patients with kidney fibrosis.

[0165] Example 9: Screening for drugs that bind to and / or inhibit NDK2 protein Screening experiments can identify and validate small molecule therapeutics, peptides, and biologics that bind to the NKD2 protein and inhibit its activity.

[0166] A DNA-barcoded compound library is generated and screened as described (Kunig et al. 2018). To this end, recombinant NKD2 protein or its fragments with a His tag are expressed in E. coli, insect cells, or mammalian cells. Purified NKD2 protein is incubated with the compound library and isolated by immunoprecipitation. Compounds that bind to the NKD2 protein are identified by Sanger sequencing of the DNA barcodes. Identified compounds are then tested for their effects on NKD2 function, myofibroblast differentiation, expression and secretion of matrix proteins such as collagen 1, and the development of renal fibrosis. To this end, an in-vivo mouse model of renal fibrosis is employed.

[0167] To identify and validate small molecule therapeutic compounds, peptides, and / or biologics that affect NKD2 expression, in vitro human cell-based fluorescent dye reporter systems, such as those expressing eGFP NKD2 fusion proteins or luciferase-based reporter systems, are established. Compound libraries are screened to identify compounds that reduce eGFP fluorescence or luciferase activity as readouts in 384- to 1,536-well assays. Expression of these human NKD2 fusion reporter constructs in these cells can be achieved, for example, by transfection and selection via a resistance gene cassette or by viral transduction. These assays employ human cell lines, such as 293T cells, as well as established human renal myofibroblast cell lines. In parallel with this screening, cytotoxicity assays are performed to exclude compounds that affect reporter fluorescence or activity through nonspecific toxicity or induction of apoptosis.

[0168] JPEG2023549706000004.jpg211170

[0169] JPEG2023549706000005.jpg201170

[0170] JPEG2023549706000006.jpg211170

[0171] JPEG2023549706000007.jpg200170

[0172] JPEG2023549706000008.jpg201170

[0173] JPEG2023549706000009.jpg205170

[0174] JPEG2023549706000010.jpg39170

Claims

1. An in vitro method for identifying therapeutic agents for chronic kidney disease, progressive chronic kidney disease and / or renal fibrosis, the method comprising reducing the expression and / or secretion of extracellular matrix (ECM) proteins by a given cell, and comprising at least one step selected from the group consisting of: (i) A step of inhibiting or reducing the expression of the nkd2 gene in the cells, (ii) A step of promoting the degradation of the NKD2 protein in the cells, and / or (iii) A step of inhibiting or reducing the NKD2 protein activity in the cells.

2. The method according to claim 1, wherein inhibition or reduction of nkd2 gene expression is achieved by nkd2 gene knockdown, knockout, conditional gene knockout, gene modification, RNA interference, siRNA and / or antisense RNA.

3. The method according to claim 1, wherein inhibition or reduction of NKD2 protein activity is achieved by using a drug that binds to the naked cuticle homolog 2 (NKD2) protein.

4. The method according to any one of claims 1 to 3, wherein the cells are kidney cells, renal myofibroblasts, or terminally differentiated renal myofibroblasts.

5. A method for identifying a therapeutic agent for chronic kidney disease, progressive chronic kidney disease and / or renal fibrosis, the method comprising identifying an agent that inhibits naked cuticle homolog 2 (NKD2) protein activity, (i) A step of providing NKD2 protein, (ii) Adding at least one drug that is screened for inhibiting NKD2 protein activity, and (iii) A step to identify at least one drug that inhibits NKD2 protein activity. Methods that include...

6. The method according to any one of claims 1 to 5, wherein the drug is selected from the group consisting of small molecule compounds, peptides, and biological preparations.

7. The method according to claim 6, wherein the biological preparation is an antibody, an antigen-binding fragment thereof, an antigen-binding derivative thereof, or an aptamer.

8. The method according to any one of claims 5 to 7, wherein the drug is a member of a compound library.

9. Use of a nucleic acid encoding naked cuticle homolog 2, or a naked cuticle homolog 2 (NKD2) protein, in the method according to any one of claims 1 to 4.

10. An antibody, antigen-binding fragment thereof, antigen-binding derivative thereof, or aptamer for use in the treatment of chronic kidney disease, progressive chronic kidney disease, and / or renal fibrosis, which specifically binds to the NKD2 protein and inhibits NKD2 activity.

11. A drug identified by the method of any one of claims 5 to 8, for use in the treatment of chronic kidney disease, progressive chronic kidney disease and / or renal fibrosis, which is a small molecule compound (smol), a peptide, an antibody or an antigen-binding fragment thereof or an antigen-binding derivative thereof, or an aptamer.

12. The agent according to claim 11, which specifically binds to the naked cuticle homolog 2 (NKD2) protein and inhibits NKD2 activity, for use in the treatment of chronic kidney disease, progressive chronic kidney disease and / or renal fibrosis.

13. A pharmaceutical composition for use in the treatment of chronic kidney disease, progressive chronic kidney disease and / or renal fibrosis, comprising the antibody according to claim 10, or its antigen-binding fragment, or its antigen-binding derivative or aptamer, or the agent according to any one of claims 11 to 12, and one or more pharmaceutically acceptable additives.

14. The pharmaceutical composition according to claim 13, wherein the additive is selected from the group consisting of pharmaceutically acceptable buffers, surfactants, diluents, carriers, excipients, fillers, binders, lubricants, lubricants, disintegrants, adsorbents, and / or preservatives.

15. A method for producing a pharmaceutical composition for use in the treatment of chronic kidney disease, progressive chronic kidney disease and / or renal fibrosis, comprising the following: (i) The method according to any one of claims 5 to 8, further (ii) Mixing the identified drug with a pharmaceutically acceptable carrier.

16. (i) the antibody according to claim 10, or its antigen-binding fragment, or its antigen-binding derivative or aptamer, the agent according to any one of claims 11 to 12, or the pharmaceutical composition according to claim 13 or 14, and (ii) One or more additional therapeutically active compounds A composition for use in the treatment of chronic kidney disease, progressive chronic kidney disease and / or renal fibrosis, including a combination of the above.

17. Repair kit for parts including the following: (i) The pharmaceutical composition according to claim 13 or 14, (ii) Apparatus for administering the composition, and (iii) Instructions for use.

Images