Smarca4 ^fl (MMRRC stock number 036548-UNC) (Sumi-Ichinose et al., 1997) and Wnt4 ^{tm3(EGFP/cre)Amc} (Wnt4 ^Cre ) (Mugford et al., 2009) and R26-tdTomato (Madisen et al., 2010) mice were maintained on a 129/Sv and C57BL/6J mixed background. Mice were bred using timed mating, and noon on the day of vaginal plug detection was considered as E0.5. All animal experiments were performed in accordance with animal care guidelines, and the protocol was approved by the IACUC of the Icahn School of Medicine at Mount Sinai (#06-822).

Histologic examinations were performed as described previously (Xu et al., 2014). Dissected kidneys were fixed in 4% paraformaldehyde (PFA), embedded in paraffin, and cut into 6–8 μm sections. Whole-mount or section ISH and immunostaining were performed according to standard procedures. For section ISH, kidneys isolated at specific stages were fixed overnight in 4% PFA at 4°C and cryopreserved in 30% sucrose. Tissues were frozen in OCT (Tissue Tek) and sectioned at 9 μm. Antisense and sense RNA probes were generated using T7 or T3 RNA polymerase according to standard procedures.

Lectin: PHA-L (Phaseolus vulgaris leucoagglutinin lectin) and LTL (Lotus Tetragonolobus lectin) (FL-1111-2 and FL-1321-2, Vector Laboratories). Primary antibodies: anti-Smarca4/Brg1 (ab110641, Abcam), -Wt1 (sc192, Santa Cruz Biotechnology); -THP (Santa Cruz, sc-271022), -NCC (EMD Millipore, AB3553), –phosphohistone-H3 (EMD Millipore, 06-570), -Tnnt2 (ab209813, Abcam), -aSMA (clone 1A4 and A5228, Sigma), -Ncam (sc-1507, Santa Cruz), -Vimentin (ab92547, Abcam), -Pttg1 (HPA045034, Atlas antibodies), -Tyrobp (MBS7127061, MyBiosource), -Foxd1 (MBS9204091, MyBiosource), and -Dlk1 (ab210471, Abcam).

For single-cell isolation, kidneys from R26-tdTomato, Wnt4 ^Cre/+ ;R26-tdTomato or Wnt4 ^Cre/+ ;Smarca4 ^fl/fl ;R26-tdTomato embryos at E18.5 were collected and then minced into 1–3 mm cubes, followed by dissociation in PBS containing 1 mg/mL Dispase and 0.7 mg/mL collagenase IV for 20 min at 37°C with pipetting up and down samples every 5 min. Then, the digested tissues were filtered with a 40-μm nylon cell strain (BD Falcon). The filtered cell suspension was centrifuged at 200 g for 5 min. After removal of the supernatant, the cell pellets were washed with PBS twice to remove fragments and then resuspended in PBS containing 1 mM EDTA. To identify and isolate tdTomato⁺ cells, single-cell suspensions of tdTomato⁻ cells from R26-tdTomato was subjected to fluorescence-activated cell sorting (FACS) to establish gates and regions that can distinguish tdTomato⁻ from tdTomato⁺ populations, and subsequently sort and collect tdTomato⁺ cells. We collected control and mutant tdTomato⁺ cells from three embryos (6 kidneys) from 2 different litters respectively.

Individually barcoded scRNA-seq libraries were prepared using the Chromium Platform (10x Genomics) with the 3′ gene expression V3 kit, using an input of approximately 10,000 cells. Briefly, Gel-bead in emulsions (GEM) were generated on the sample chip in the Chromium controller. Barcoded cDNA was extracted from the GEMs by post-GEM RT-cleanup and amplified for 12 cycles. Amplified cDNA was fragmented and subjected to end repair, poly A-tailing, adapter ligation, and 10 ×–specific sample indexing following the manufacturer’s protocol. Libraries were quantified using Bioanalyzer (Agilent Technologies) and QuBit (Thermo Fisher Scientific) analysis. Libraries were sequenced in paired-end mode on a NovaSeq Instrument (Illumina) targeting a depth of 5 × 10⁴-1 × 10⁵ reads per cell. Sequencing data were aligned and quantified using the Cell Ranger Single-Cell Software Suite (version 3.0, 10 × Genomics) against the provided mm10 reference genome.

scRNA-seq raw data were processed using the 10X Genomics Cellranger v3 to obtain gene expression counts of each barcode for the mutant and control, respectively. Cell ranger QC was applied to the count matrices to remove non-cells. Filtered gene and barcode count matrices were used for scRNA-seq analysis with a custom R pipeline centered around the Seurat package v4.0.3 (Hao et al., 2021), in addition to other publicly available R packages (see below). To remove potential artifacts due to low-quality cells and lowly expressed genes, cells expressing less than 500 genes and genes detected in less than 5 cells were excluded from the analysis. Gene counts for each cell are normalized by dividing the total counts for that cell and multiplying a scale factor of 10,000. Natural-log transformation is then applied to the expression levels plus 1. Next, the top 5,000 most variable genes were identified using the variance stabilizing transformation method. To integrate the mutant and control samples, 5,000 anchor genes were identified between the two samples. After the integration, expression levels were centered and scaled by subtracting the average expression and dividing their standard deviations for each feature. Principal Component Analysis (PCA) dimensionality reduction was applied to the scaled data using the variable genes to calculate the top 30 principal components (PCs). To visualize the cells at a low dimensionality, non-linear dimensional reduction approaches UMAP (McInnes et al., 2018) and tSNE (van der Maaten and Hinton, 2008) were performed using the PCs as input. To cluster the cells, a shared nearest neighbor (SNN) graph (Jarvis and Patrick, 1973) was constructed by calculating the neighborhood overlap between each cell and its 20 nearest neighbors using the Jaccard index (Jaccard, 1912). A local moving algorithm (Waltman and van Eck, 2013) was then performed on the global SNN graph to optimize a modularity function to determine clusters at a resolution of 0.5. To study the detailed cell type characteristics at high resolutions, the cell subsets of tubule progenitors, proximal tubule and stroma were extracted from the global analysis and re-clustered at resolutions of 0.5, 0.25 and 0.4, respectively.

To identify conserved cell cluster markers, differentially expressed genes (DEGs) were identified between a cluster and the rest cells for mutant and control, respectively, using the Wilcoxon Rank Sum test. p-values were adjusted by the Benjamini-Hochberg (BH) procedure (Benjamini and Hochberg, 1995). A gene was identified as a potential conserved marker for a cluster if its adjusted p-values were less than 0.05 in both mutant and control. The potential markers were then ranked by average log fold change in mutant and control by descending order. The top 100 markers for each cluster were then manually compared with known kidney cell type markers from the previous kidney studies (Combes et al., 2019; Miao et al., 2021) and the PanglaoDB database (Franzen et al., 2019). Cell clusters were then annotated based on the maximum overlap of the cluster markers and cell type markers.

Heatmaps were plotted with the top 4–6 markers from each cell type. DEGs between mutant and control were identified for each cell type using the DESeq2 (Love et al., 2014). p-values were corrected using the BH procedure (Benjamini and Hochberg, 1995).

GSEA was performed using ranked fold change values (mutant over control). The enrichment score was calculated according to the GSEA algorithm (Subramanian et al., 2005). GO enrichment analysis was performed using http://bioinformatics.sdstate.edu/idep (Ge et al., 2018).

Trajectory analysis was performed on the control cells, mutant cells, and both mutant and control combined using the R package Monocle 3 v1.0.0 (Cao et al., 2019). For the global analysis, the nephron progenitor cells were set as the root cells. For the stroma subclusters, the CD-associated stroma, nephrogenic stromal progenitors or fibroblast cells were set as root cells. For the endothelial cells, endothelial subtype 2 was set as root.

FACS-purified TdTomato⁺ cells from Wnt4 ^Cre/+ control and Smarca4 ^cKO/cKO kidneys collected from E17.5 kidneys were used for total RNA extraction using Trizol Reagents (15596026, Invitrogen). 0.5–1 μg of total RNAs were treated with RNase-Free DNase I Set (79254, QIAGEN) and then used for reverse transcription using a SuperScript IV Reverse Transcriptase (18090010, Thermo Fisher Scientific) for first-Strand cDNA Synthesis. Real-time PCR was performed using iQ SYBR Green Master Mix (4309155, Applied Biosystems). Expression levels of each transcript were normalized using β-actin as an internal control. Each set of experiments was repeated three times, and the DDCT relative quantification method was used to evaluate quantitative variation.

Oligos used for qPCR: Pttg1-froward 5′-GCA​GTG​GGT​GAA​GTT​GAA​CAC-3′ and Pttg1-reverse 5′-CAG​TGG​TTG​ACA​AGT​TAC​TGT-3'; β-actin-forward 5′-CAT​TGT​TAC​CAA​CTG​GGA​CGA-3′ and β-actin-reverse 5′-GAA​GGT​CTC​AAA​CAT​GAT​CTG-3'.

PH3⁺ cells were counted from sections with clear cortical UB, PTA and SSB and medullary structures. PH3⁺ cells were counted in cortical regions around 32 UBs, including peripheral nephrogenic stroma and ventral PTAs, as well as 32 SSBs per kidney and medullary area 20 × 0.01 mm² per kidney, respectively, and 3 kidneys at E18.5 were counted. Values represent the average number of PH3⁺ cells (±standard deviations) per UB-PTA-SSB for the cortical or 0.01 mm² for the medullary area (section thickness is 6 μm).

TUNEL⁺ cells from 10 sections with clear cortical and medullary structures per E18.5 kidney and 3 kidneys were counted from the medullary region extending into the deep cortical region, respectively. Values represent the average number of TUNEL⁺ cells (±standard deviations) per section (6 μm).

Two-tailed Student’s t-test was used for statistical analysis. A value was considered statistically significant if p < 0.05.

The area of the Dll1 ⁺ domain was measured on sections for spatial calibration using ImageJ software (NIH). The area of the Dll1 ⁺ domain from 18 SSBs (8 sections per kidney and 3 kidneys for each sample) was measured. Two-tailed Student’s t-test was used for statistical analysis.

To investigate whether Smarca4 expression in nephron primordia is required for tubulogenesis, we crossed Wnt4 ^Cre/+ mice with Smarca4 ^fl/fl to delete Smarca4 in Wnt4-expressing (Wnt4 ⁺) precursors, resulting in the conditional Smarca4 KO mice (Smarca4cKO). Immunostaining confirmed Smarca4 depletion in CSBs/SSBs of Smarca4cKO kidneys (Supplementary Figure S1A). The mutant kidneys at E18.5 were smaller, with a length approximately 24.9% ± 1.6% shorter than that of control littermates (Supplementary Figure S1B; n = 8 and p = 0.0225). Histological analysis revealed a significantly shortened medullary region with reduced epithelial tubules and increased interstitium (Figure 1A). Quantification of glomerular numbers showed that the number of glomeruli in Smarca4cKO was 58.4% ± 1.5% of that in Wnt4 ^Cre/+ littermates.

Next, we stained kidney sections with antibodies or lectins to label different tubular segments. Wt1 defines podocyte identity by activating other TFs (Dong et al., 2015). As shown by anti-Wt1 immunostaining (Figure 1B), it is expressed in the podocyte primordium of the proximal RV/SSB and in developing podocytes. In Smarca4cKO kidneys, Wt1 expression in these structures was preserved. The mutant podocytes were also marked by lectin PHA-L (Figure 1C). Remarkably, however, PHA-L-marked PTs in the cortex were markedly reduced (Figure 1C). Similarly, lectin LTL-labeled PTs, especially the proximal convoluted tubules (PCT), were significantly decreased (Figure 1D). While both distal convoluted tubules (DCT) marked by NCC (NaCl cotransporter) (Figure 1E) and HLs labeled with THP (Figure 1F) were detectable, the THP⁺ HLs were found in similar locations as the PTs or DCTs and failed to reach the innermost region of the kidney. These results indicate that Smarca4 is essential for proper tubulogenesis.

Since the knowledge of Wnt4 ⁺ cells and the different cell types they may differentiate into is sparse, we set out to comprehensively define the cellular state of Wnt4 ^Cre -labeled cells and further investigate the requirement of Smarca4 during nephron differentiation at a single-cell resolution. We performed scRNA-seq analysis on FACS-purified fresh tdTomato⁺ cells from control or Smarca4cKO kidneys at ∼ E18.0-18.5 (harvested 7 a.m.) (Supplementary Figure S2A). After removing low-quality cells, we obtained transcriptomes from 5,579 control and 6,299 Smarca4cKO cells, respectively. Unsupervised analysis identified 11 major cell types (Supplementary Figure S2B, C), which can be divided into 24 distinct cell clusters (CS) (Figures 2A, B). All cell types largely overlapped between the control and mutant.

Next, we conducted differentially expressed gene (DEG) analysis between each cluster and the rest of the clusters in the control to further characterize Wnt4 ^Cre -labeled kidney cell types and determine the key marker genes for each cell type (full lists of cluster markers are provided in Supplementary File S1). SSB (CS3) and CD-associated stroma (CS17) exhibited high levels of Wnt4 expression (Figures 2C, D), in agreement with previous ISH findings (England et al., 2020). Consistent with a previous report indicating the re-entry of Wnt4 ⁺ PTA cells into the NPCs (Lawlor et al., 2019), we identified Six2 ⁺ Cited1 ⁺ NPCs (CS4) (Figure 2D; Supplementary Figure S2D). Clusters corresponding to major nephron segments (CS5,6,20) and podocytes (podocyte precursors CS11–high levels of Pcp4, mature podocytes CS15–high levels of Nphs1/Nphs2, and immature podocytes CS22–high levels of Nphs1 but low Nphs2) (Supplementary Figure S2D) were identified. Notably, a large proportion of Wnt4 ^Cre -labled cells were stromal cells (Col3a1/Pdgfrα, Figures 2D, E) and Pecam1 ⁺ endothelial cells (ECs) (Figure 2E). The ECs could be separated into capillary (CS2– Rgcc ⁺ Fabp4 ⁺ Fabp5 ⁺ but low Cdk1, Supplementary Figure S2E), proliferative capillary (CS7–Rgcc ⁺ Fabp4 ⁺ Fabp5 ⁺ but high Cdk1, Supplementary Figure S2E), and arterial (CS14–Gja4 ⁺ ) (Figure 2D; Supplementary Figure S2E; Supplementary File S1). We also identified a small population of SMCs (CS19–Tnnt2 ⁺ , Figure 2D), 2 types of immune cells (Tyrobp ⁺, Supplementary Figure S2E)—macrophages (CS12–C1qa ⁺ ) and neutrophils (CS21–Cd52 ⁺) (Figure 2D), principal cells (CS18 –Calb1 ⁺) and a small population of ureteric epithelial cells (CS23–Krt19 ⁺ Igfbp5 ⁺) (Figure 2D). We noticed higher mitochondrial levels in cluster 9 (Figure 2D), we, therefore, excluded this cluster for further analysis.

Wnt4 is expressed in stromal cells around the ureteric tree in the medullary regions (England et al., 2020) and plays a critical role in the SMC fate commitment in the medullary stroma (Itaranta et al., 2006). These Wnt4 ⁺ cells were thought to be a transient stromal cell population because previous studies did not detect Wnt4 ^Cre -labeled SMCs, endothelia or macrophages/neutrophils by expression studies with cell type-specific markers on kidney sections (Shan et al., 2010). However, our deep sequencing analysis clarified the presence of these cell populations as well as cortical stroma and nephrogenic stroma (Figures 2B, E). To validate these genomic findings, we visualized the expression of Foxd1, Pecam1, Tyrobp, and Tnnt2/SMA (smooth muscle actin) by immunostaining and observed that tdTomato⁺ cells overlapped with Foxd1⁺ nephrogenic/cortical stromal cells and podocytes, Tnnt2⁺/SMA⁺ SMCs, Pecam1⁺ ECs, and Tyrobp⁺ immune cells (Figure 2F). Thus, this analysis revealed the heterogeneity of Wnt4 ^Cre -labeled cells in the mouse kidney, including previously unknown cell types.

To determine the effect of Smarca4 loss on Wnt4 ^Cre -labeled cells, we compared datasets between the control and Smarca4cKO. Although all cell types overlapped between the control and mutant, we observed a significant reduction in the number of podocytes and tubular cells in the mutant (Supplementary Figure S2C). This finding aligns with the results from the phenotypic analyses (Figure 1). Conversely, the mutant exhibited an increase in the proportion of ECs, SMCs and stromal cells (Supplementary Figure S2C). The DEG analysis conducted between the control and mutant groups (Supplementary File S2) revealed a widespread upregulation of the housekeeping ribosome gene Rps18 (Gm10260), the actin-binding protein Tpm3-rs7, and the nucleosome Hist1h2ap (Supplementary Figure S3A). These genes are tightly associated with cell proliferation and growth in high-protein synthesis-demanding cells (Kusnadi et al., 2015; Gaviraghi et al., 2019; Ferreira et al., 2020) and are frequently dysregulated in highly proliferative cancer cells (Catez et al., 2019; Gaviraghi et al., 2019; Ferreira et al., 2020). Therefore, the upregulation of these genes likely indicates changes in the cellular state caused by Smarca4 deficiency.

To further investigate the changes in the cellular state in the mutant, we employed pseudotime ordering analysis (Haghverdi et al., 2016) to evaluate cellular developmental trajectories. Consistent with the upregulation of Rps18 (Gm10260), Tpm3-rs7, and Hist1h2ap, this analysis indicated an overall increase in proliferative progenitors in the mutant. For example, three EC clusters shared a common progenitor residing in the proliferating capillary progenitors, but Smarca4cKO produced more proliferative progenitors (Supplementary Figure S2F). The analysis of nephron tubule differentiation revealed that the mutant produced more precursors that continued along the nephron differentiation path, and the differentiation status of podocytes appeared different from control cells (Supplementary Figure S3B). These findings further suggest that Smarca4 deficiency alters the cellular state and fate (e.g., cell proliferation and differentiation). Together, these results provide important insights into the impact of Smarca4 on the developmental trajectory of Wnt4 ^Cre -labeled cells, which could have implications for kidney development and disease.

Our findings reveal that Wnt4 ^Cre -labeled cells comprise multiple distinct stromal populations that are enriched in Smarca4cKO kidneys. However, the origin and regulatory networks that control the development of distinct stromal cells in the kidney are poorly defined. Therefore, we extracted stromal populations along with SMCs and subclustered them into 9 subpopulations (Figure 3A; Supplementary File S3). Heatmap analysis of the top 4 marker genes (Supplementary Figure S4A) revealed differences between ureteric stromal cells in control and Smarca4cKO kidneys, with a subpopulation in the mutant lacking the expression of Mfap5 (Figure 3B) and Tbx18 (Supplementary Figure S4B) but expressing Wnt4 (Figure 4A). These cells also expressed Dlk1 (Figure 3B; Supplementary Figure S4A), Col3a (Figure 3B), and Itm2a (Supplementary Figures S4A, B), as well as all other top DEGs expressed in Mfap5 ⁺ Tbx18 ⁺ ureteric stroma cells (Supplementary File S3).

To confirm this finding, we conducted immunostaining for Dlk1 and observed its expression in the stroma at the ureteropelvic junction and in the ureter, which co-localized with tdTomato⁺ cells (Figure 3C). Furthermore, we performed ISH and found that Dlk1 transcripts were weakly expressed in the ureteric stroma at the ureteropelvic region, but were substantially upregulated in the outer layer in the mutant (Figure 3D). We also observed some Dlk1 expression in the inner stromal cells, which was not significantly increased in the mutant (open arrowheads, Figure 3D). Based on previous lineage tracing studies, the outer adventitial cells and the inner lamina propria fibroblasts adjacent to the ureteric epithelium derive from Tbx18 ⁺ precursors (Bohnenpoll et al., 2017). Therefore, some Tbx18 ⁺ cells in the Dlk1 ⁺ Mfap5 ⁺ subgroup may represent precursors or intermediate cells, while the Dlk1 ^high outer layer of the ureteric stroma in the mutant may be mixed with some Dlk1 ⁺ Mfap5 ^- cells. As Mfap5 ⁺ cells (Figure 3B) were negative for Wnt4 (Figure 4A), increased Wnt4 signaling in some ureteric stromal cells may lead to changes in the expression of specific genes, such as the downregulation or inhibition of Mfap5 expression. We also confirmed increased Col3a1 transcripts in the mutant (Figure 3E), which is consistent with the upregulation of Col3a1 detected by scRNA-seq (Supplementary Figure S3A).

To gain insight into the lineage relationships of different Wnt4 ^Cre -labeled stromal populations, we conducted pseudotime analyses. We selected Wnt4 ^high CD-associated stromal cells (Figure 4A) as the “root” for our model in which Wnt4 ⁺ “progenitors” expand and produce a group of progenitors that differentiate into fibroblasts. These fibroblasts then give rise to precursors in the nephrogenic stroma, which further branches and differentiate into pericytes (Figure 4A). Interestingly, we found no direct lineage connection between the medullary stroma or ureteric fibroblasts and the CD-associated stromal cells, suggesting that these cells may originate from different progenitor cells. However, in Smarca4cKO, the CD-associated Wnt4 ^high progenitors not only produce more cycling/proliferating cells but also generate fibroblasts that can differentiate into the mutant medullary and ureteric stroma. We observed similar linear relationships in a model where the nephrogenic stroma (Figure 4A) or fibroblasts (Supplementary Figure S4C) were chosen as the “root”. However, when using fibroblasts as the root, more cells showed gene expression changes along the trajectory (grey circles, Supplementary Figure S4C). In contrast, none of these models exhibited a direct linear relationship with the cortical stroma or smooth muscle cells (SMCs), indicating that these cell types likely originate from different progenitor cells. Overall, these analyses underscore the intricate development of Wnt4 ^Cre -labeled stromal cells and reveal increased cell proliferation and altered developmental trajectories caused by Smarca4 deficiency.

After applying stringent cutoffs of adjusted p < 0.01 and fold-change >1.5, the DEG analysis identified only a few upregulated genes in Smarca4cKO. Of these genes, Pttg1 was globally upregulated, particularly in nephrogenic stromal cells, pericytes and some CD-associated stroma (Figure 4B; Supplementary File S4). Pttg1 plays a crucial role in sister chromatid segregation during mitosis, and its upregulation is associated with various types of tumors (Pei and Melmed, 1997; Itaranta et al., 2006; Chen et al., 2020). We also detected a similar upregulation of Top2a (Figure 4B), a proliferation marker associated with tumor grade and Ki67 index (Supplementary Figure S4B). Next, we used two independent methods to confirm the upregulation of Pttg1 in the mutant. First, we immunostained for Pttg1 at the earlier stages of E16.5-17.5 and found that the levels of Pttg1 were increased in the mutant kidneys at these stages, especially in the cortical and PTA-SSB regions (Figure 4C). Second, we performed reverse-transcription and real-time PCR (RT-qPCR) using RNA prepared from freshly FACS-purified tdTomato⁺ cells from E16.5 kidneys and found that Pttg1 expression levels in Smarca4cKO were ∼3.2-fold higher than in control cells (Figure 4D). Altogether, our findings demonstrate that Smarca4 deficiency induces Pttg1 upregulation in Wnt4 ^Cre -labeled cells.

We next investigated whether Smarca4 loss resulted in changes in the expression of genes involved in early SSB segmentation and upregulation of Pttg1 in nephron precursor cells. The SSB is patterned into a proximal segment—Bowman’s capsule and visceral podocyte—derived from the proximal domain of the polarized RV, a distal RV-derived part that further subdivides into the future PT, HL, and DT segments followed by CNT fused with CD (Figure 5A). Subclustering analysis of nephron precursor cells (CS3, 4, 16) revealed 7 distinct populations: uncommitted NP, NP-stromal, committing NP, PTA-RV, proximal SSB for podocyte, intermediate SSB for PT, and intermediate-distal SSB for HL and DT (Figure 5B; Supplementary File S5).

We next analyzed changes in the expression of genes involved in early patterning of SSB, including Lhx1, Notch signaling, and other TFs, which was verified by ISH or immunostaining. Lhx1 is required for initial RV polarization and expressed in distal RV and intermediate-distal SSBs that form PT, HL and DT (Kobayashi et al., 2005) (Figures 5C–E). The Notch ligand Dll1 further specifies the Lhx1 ⁺ domain into the PT fate, as Dll1 hypomorphic mice have severely reduced nephron numbers with loss of PT (Cheng et al., 2007). Dll1 is expressed in the distal RV, and its expression persists in the middle region of the SSB where PT and HL are generated (Figures 5C, E). Jag1, another Notch ligand, is highly expressed in the intermediate SSB for PT and HL (Figures 5D, F). The DEG analysis did not detect significant changes in expression levels of these three genes (Figures 5C, D; Supplementary File S6). Consistent with this, no obvious changes in the expression levels of these genes were observed by ISH (Figure 5E) or immunostaining (Figure 5F), although Smarca4cKO kidneys at ∼ E15.0 were smaller (Supplementary Figure S5A), and Dll1-marked region (Figure 5E) or Jag1⁺ domain (Figures 5F, G) was reduced in size. We measured and calibrated the area of the Dll1 ⁺ domain and found that the Dll1-marked area was ∼53.22% ± 1.8% smaller than the control (n = 18, p = 0.02808). In contrast, Notch1 expression was downregulated in Smarca4cKO SSB-PT territory (Supplementary Figure S5B; Supplementary File S6), while Notch2 did not show significance changes (Supplementary Figure S5B; Supplementary File S6). Pax2 also showed segmented expression in the SSB, with lower expression levels in the future PT regions (Figures 5D, F). However, Pax2 expression levels in the mutant PT regions were higher than in the control (Figures 5D, F; Supplementary File S6). Notably, the HL/DT-specific genes, such as Wfdc2 and Ly6a were also upregulated in the mutant prospective PT region (Figure 5D).

Pou3f3 plays a crucial role in determining the fate of intermediate and distal nephrons (Nakai et al., 2003). It is expressed in the RV and broader regions covering the mid-distal SSB (Figure 5E, Supplementary Figure S5A). However, its expression was reduced in some mutant SSBs (arrow, Figure 5E), which is consistent with the findings from the scRNA-seq analysis (Figure 5D). We also investigated the expression of Irx genes, which are essential for the formation of the intermediate region of the nephron (Reggiani et al., 2007). Irx1 expression begins at the CSB stage and is confined to the intermediate region destined to become HL (Figures 5D, E). In Smarca4cKO kidneys, the number of Irx1-labeled nephron primordia was significantly decreased (Supplementary Figure S5A), and the SSBs had smaller Irx1 ⁺ regions and lacked a well-defined Irx1 ^- future PT fate (arrows, Figure 5E). Additionally, the intensity of Irx1 and Irx2 signals appeared to be reduced in the mutant (Figure 5E; Supplementary Figure S5A). These findings align with the scRNA-seq analysis for Irx1 (Figure 5D; Supplementary File S6) and Irx2 (Supplementary File S6). Together, these results indicate that Smarca4 is necessary for establishing distinct nephron tubular regions during SSB segmentation and extension.

Our analyses also indicated upregulation of Pttg1 (Figure 6A, Supplementary Figure S5C) and Top2a (Supplementary Figure S5C) in the mutant SSB. We also observed elevated expression of ECM markers Col3a1/Col1a1/Fn1 in the mutant (Figure 6A), indicating a potential progression towards fibrosis or EMT. These findings underscore the critical role of Smarca4 in regulating the expression of Pttg1 and the cellular state and fate of nephron precursors.

Since Pttg1 is a key regulator of sister chromatid segregation during mitosis and its increased PTTG1 expression is associated with hyperproliferation, we clarified whether Smarca4 loss also leads to increased proliferation. Immunostaining for anti-phospho-Histone H3 (PH3), a marker for M-phase cells, showed a significant increase in PH3⁺ cells in the cortex and medulla of Smarca4cKO kidneys compared to controls (Figure 6B, ∼4.42-fold in cortex and ∼7.5-fold in medulla). This result is consistent with the transcriptomic trajectories showing increased proliferating/cycling cells in the interstitial cells (Figures 4A, B; Supplementary Figure S2F) and nephron precursors (Supplementary Figure S3B) of Smarca4cKO kidneys. Given that PTTG1 as an oncogene known to promote cell division and contribute to the development of tumors and metastasis (Luo et al., 2013; Noll et al., 2015), the elevated levels of Pttg1 and increased cell proliferation rate in Smarca4cKO may lead to upregulation of ECM-related genes in the precursor cells and limit their ability to differentiate into proper tubular cells.

To further investigate the impact of Smarca4 deficiency on nephron differentiation, we re-clustered the nephron tubule populations (CS5,6,20) and identified two progenitor-tubule intermediate populations (IM1, IM2) and seven distinct nephron tubule populations, including EPT, PCT, PST (proximal straight tubule), DCT and CNT, as well as ascending and descending HL (AHL and DHL) (Figure 7A; Supplementary File S7). Since the proportion of EPT cells was more reduced than the PCT and PST, we analyzed the transcriptome for this cluster and identified 337 upregulated and 22 downregulated genes, with many kidney disease-associated genes among the top downregulated genes (Supplementary Figure S6A; Supplementary File S8). GO analysis revealed that genes involved in metabolic processes, cell-cycle/proliferation, cellular component organization, and negative regulation of cellular process and integrin signaling pathway were upregulated (Supplementary Figure S6B).

We found that certain markers for mature PT cells, including Ass1, Aldob, and Slc34a1, were strongly upregulated in EPT-related cells of Smarca4-deficient kidneys. In contrast to control cells, which showed a temporal progression of maturation from EPT to PCT and PST cells based on the upregulation of these genes (Figures 7B, C), Smarca4cKO kidneys failed to generate discernible cellular heterogeneity with clearly demarcated populations of EPT and differentiated PT due to elevated expression of these genes in EPT cells (Figures 7B, C). Moreover, the HL/DT-specific genes such as Wfdc2 (Figure 7C) and Ly6a (Supplementary Figure S6A) were also strongly upregulated in Smarca4cKO EPT, IM1 and DHL cells, along with many cell-cycle regulators, including Ki67, Cdk1 (Figure 7C), Cdk2, Tab1, Pclaf, Igfbp5, Lmo1, and Smc1b (Supplementary Figure S6C; Supplementary File S8), reflecting higher proliferation rates as seen by anti-PH3 immunostaining (Figure 6B). Upregulation of Wnt4 was also observed in the mutant IM1 cluster (Supplementary Figure S6C).

In contrast to the upregulation of Pttg1 in the mutant precursor cells (Figure 6), its expression levels appeared unchanged in these more differentiated tubular cells (Supplementary Figure S6C). However, the expression of ECM genes such as mesenchymal collagens Col3a1 and Col1a1 (Figure 7C) as well as EMT markers found in tumorigenic EMT (Latil et al., 2017) or renal fibrosis, including EMT-associated cytoskeleton Vimentin, Fn1 (Figure 7C) and Sparc (Supplementary Figure S6C), were significantly increased in Smarca4cKO tubular cells. Notably, the hypoxia-inducible TF Hif1a, a key regulator for renal fibrosis under various conditions (Liu et al., 2017) was also upregulated in the EPT, PCT and CNT (Figure 7C). Similar to the upregulation of Col3a1 in the cortical region of Smarca4cKO kidneys (Figure 3E), ISH also revealed increased Fn1 transcripts in the cortical areas and tubular structures (Figure 7D). Co-immunostaining for Vimentin and Ncam (neural cell adhesion molecule) confirmed that Smarca4cKO kidneys lacked Ncam-labeled epithelial tubules but had Vimentin⁺ cells in the tubular structures (Figure 7E), indicating the presence of fibrotic cells and induction of genes related to fibrosis. Taken together, these results suggest that although some tubular differentiation genes are expressed in Smarca4-deficient kidneys, the mutant tubular cells lose their tubular characteristics due to the induction of cell-cycle regulators and non-tubular ECM/EMT-related genes, leading to cell fibrosis and loss of expression of many nephron tubular-specific genes (Supplementary Figure S6A).